36 Hydrocarbons
Figure 8. Production of alkylbenzenes, process of Hüls AG [165]
[163]. The removal of these compounds from
monoalkylbenzenes is difficult. Formation of
these byproducts can be largely avoided by
carrying out the alkylation of benzene with
olefins from the dehydrochlorination, where di-
chloroalkanes are dehydrochlorinated to give
tar-like products that are easily separated. In in-
dustrial production the reaction of C10 – C13 -
olefins with benezene is carried out with a
1 : 10 molar ratio in the presence of anhy-
drous hydrofluoric acid at 10 ◦ C, a process
that dates back to 1949 [164]. The indus-
trial process using the steps paraffin chlori-
nation – dehydrochlorination – alkylation in the
presence of hydrofluoric acid was first realized
by Chemische Werk Hüls in 1962 [162]. This
process is shown in Figure 8. In this way a quality
standard for alkylbenzenes for use in detergents
was created; in ecological terms, this standard is
still valid.
3.11. Diphenylmethane
Diphenylmethane [101-81-5] benzylbenzene,
C13 H12 , M r 168.24, was first prepared in 1871
by Zincke from benzyl chloride and benzene in
the presence of zinc dust or copper (I) oxide.
Physical Properties. Diphenylmethane
forms colorless prismatic needles with a harsh
herbaceous odor reminiscent of geranium
leaves. It is insoluble in water, soluble in al-
cohol, ether, chloroform, and benzene.
mp 26 – 27 ◦ C
bp
101.3 kPa 261 – 262 ◦ C
1.33 kPa 125.5 ◦ C
d 20 1.006
4
Flash point ca. 130 ◦ C
Enthalpy of formation (room temperature) 109.6 kJ/mol
Enthalpy of combustion (20 ◦ C) 41.2×103 kJ/kg
Chemical Properties. As an araliphatic hy-
drocarbon, diphenylmethane displays the chem-
ical properties of both aromatic and aliphatic
compounds. It is difficult to limit substitution
in the phenyl radicals to one of the rings; there-
fore, controlled syntheses of particular unsym-
metrically substituted diphenylmethanes must
be carried out by different paths. Under suit-
able reaction conditions the methylene group
can also be substituted, e.g., by halogen or ni-
tro groups. Benzophenone can be produced by
oxidation, for example, with oxygen over vari-
ous catalysts, chromic acid or dilute nitric acid.
Hydrogenation in ethanol over a nickel catalyst
leads to dicyclohexylmethane. When diphenyl-
methane is passed over platinum – charcoal at
300 ◦ C, cyclization to fluorene takes place.
 Hydrocarbons 37

Production. Of the many known methods to polyphenyls (n > 2) are also known and may be
synthesize diphenylmethane only a few are used isolated from coal tar or from aryl halides by the
commercially. Ullmann reaction.
The Friedel – Craft reaction of benzyl chlo-
ride with benzene in the presence of aluminum
chloride is an important method for production 4.1. Biphenyl
of diphenylmethane:
Biphenyl [92-52-4], phenylbenzene, diphenyl,
C12 H10 , M r 154.2, was discovered in 1862 by
Fittig by the reaction of bromobenzene with
Other known catalysts of this process are el- sodium metal. Berthelot obtained the com-
emental aluminum, iron(III) chloride, antimony pound in 1867 by passing benzene vapor through
pentachloride, zinc chloride – hydrogen chloride a heated tube. In 1875 Büchner demonstrated
and boron fluoride dihydrate. When sulfuric acid the presence of biphenyl in coal tar [177].
is used as a condensation agent diphenylmethane
yields are 83 % [166] and with polyphosphoric
acid yields are over 90 % [167].
Condensation of benzene with formaldehyde
in presence of 85 % sulfuric acid yields 79 %
diphenylmethane [168].
Physical Properties. Pure biphenyl is a
white solid; when only slightly impure it has a
yellow tint. It crystallizes from solution results
as glossy plates or monoclinic prisms. Biphenyl
Quality Specifications. Commercial di- is almost insoluble in water, but is easily soluble
phenylmethane should be at least 97 % pure in organic solvents such as ethanol, diethyl ether,
(GC). It must be free of halogens and have a and benzene. The dipole moment and X-ray
minimum melting point of 24 ◦ C. measurements demonstrate that the two benzene
rings are coplanar in the solid state. In the melt,
Uses. Diphenylmethane is used in the fra- in solution, and in the vapor free rotation about
grance industry both as a fixative and in scenting the central C-C bond can take place; in substi-
soaps. It can serve as a synergist for pyrethrin in tuted biphenyls, however, this rotation may be
pesticides and insecticides [169, 170]. Diphenyl- severely restricted, e.g., in dinitrodiphenic acid,
methane is recommended as a plasticizer to im- leading to the existence of optically active iso-
prove the dyeing properties [171], as a solvent mers. The most important physical properties
for dyes [172], and as a dye carrier for print- are listed below:
ing with disperse dyes [173]. The addition of
diphenylmethane to saturated, linear polyesters mp 69.2 ◦ C [178]
bp (101.3 kPa) 255.2 ◦ C [178]
improves their thermal stability [174], and its d 20
4 (solid) 1.041
addition to jet fuels increases their stability and d 77
4 (liquid) 0.9896 [179]
77
lubricating properties [175]. nD 1.5873
Substituted diphenylmethanes are used as T crit 515.7 ◦ C [180]
pcrit 4.05 mPa
solvents for pressure-sensitive recording mate- Flash point 113 ◦ C [181]
rials [176]. Ignition temperature 570 ◦ C
Heat of combustion 6243.2 kJ/mol
Heat of vaporization 53.9 kJ/mol
Heat of fusion 18.60 kJ/mol
4. Biphenyls and Polyphenyls
In the series of aromatic hydrocarbons with The variation of some physical properties
polyphenyls or polyaryls of the general formula with temperature is given in Table 18 (see next
C6 H5 (C6 H4 )n C6 H5 , only the lowest members page).
(n = 0, 1) are economically important. Higher
38 Hydrocarbons
Table 18. Temperature dependence of some physical properties of
biphenyl [179]
t, ◦ C 100 200 300
Vapor pressure, kPa 25.43 246.77
Liquid density, g/cm3 0.97 0.889 0.801
Heat capacity, J/g 1.786 2.129 2.468
Chemical Properties. Biphenyl is a very sta-
ble organic compound; in an inert atmosphere
it remains unchanged even at a temperature
>300 ◦ C. The compound sublimes, distills with
steam, and undergoes various substitution reac-
tions; the reactivity is similar to, but less than that
of benzene. Substitution reactions, e.g., chlori-
nation, nitration, and sulfonation, occur at the
2- and 4-positions, the latter being the preferred
site of attack in Friedel – Crafts reactions.
Production. Hydrodealkylation of Benzene.
Biphenyl is currently obtained mainly as a by-
product in the production of benzene by ther-
mal or catalytic hydrodealkylation of toluene
(→Benzene, Chap. 5.3.1.) [182–185]. The reac-
tion is carried out in a hot tube reactor at 700 ◦ C
with a hydrogen pressure of 4 MPa; the ratio of
hydrogen to toluene is 4 : 1. Toluene conver-
sion ranges from 35 to 85 %, with methane being
formed at 2.5 mol % above the stoichiometric.
Biphenyl is obtained as the pot residue, fol-
lowing removal of gas and distillation of ben-
zene and toluene; about 1 kg of biphenyl is ob-
tained per 100 kg benzene. The biphenyl can be
enriched to a purity of 93 – 97 % by distillation.
Thermal Dehydrocondensation of Benzene.
Until the early 1970s, biphenyl was produced
exclusively by specific thermal dehydroconden-
sation of benzene [186]. This process yields a
higher quality product than the toluene route,
e.g., distilled material with a purity >99.5 %;
varying quantities of terphenyl isomers and sev-
eral higher homologues are also obtained.
The slightly endothermic dehydrocondensa-
tion (∆H = + 8 kJ/mol [187]) of benzene to bi-
350
phenyl and terphenyl has been investigated in-
558.06 tensively [188, 189]. The formation of biphenyl
0.751 from benzene starts at ca. 480 ◦ C and increases
2.640
with increasing temperature and residence time.
Industrially, the reaction is carried out in a
multitube reactor, heated electrically [190] or
by heat exchanger, at 700 – 850 ◦ C using resi-
dence times of the order of seconds. An optimal
tube length of 45 – 75 m and diameter of 100 –
120 mm have been reported [191–194]. None
of the published procedures work in equilib-
rium because carbon black formation increases
with increasing temperature and longer resi-
dence time.
The preferred residence time is ca. 10 s,
which results in a benzene conversion of ca.
10 % [192, 195]. By recycling the benzene,
1 kg of starting material may yield 0.8 – 0.85 kg
biphenyl, 0.07 – 0.12 kg terphenyl, and some
higher polyphenyls. The major technical prob-
lem in the production of biphenyl by dehydro-
condensation of benzene is carbonization at the
high temperature required [196]. The carbon de-
posit on the reactor tube walls and the release of
large carbon flakes at bends or narrow passages
within the tubes may lead to blockages. Counter-
measures include addition of about 0.1 % oxy-
gen or sulfur compounds, such as methanol, eth-
anol, acetone, or carbon disulfide [197–199] to
the benzene feedstock, the use of special alloys
[200, 201], uniform heating of the reactor by
means of circulating gases, and high feedstock
velocity, and the use of turbulent flow [192]. A
proposal for further improving the process is to
carry out the reaction in fluidized beds. Quartz
sand [202] or sintered corundum [203] is used
as energy carrier in this process; these materials
are recycled after the carbon deposits have been
removed by burning them outside of the reactor.
Other Synthetic Routes. In addition to these two
industrial processes for the production of biphe-
nyl, other synthetic routes have been described:
1) Ullmann reaction: heating halogenoben-
zenes with copper powder [204].
2) Pyrolysis of coal tar in the presence of molyb-
denum oxide – alumina catalyst [205, 206].
3) Reduction of phenyldiazonium chloride so-
lution with copper and zinc [207].

4) Heating phenylmagnesium halides with cop-
per(II) salts [208].
5) Oxidative dehydrocondensation of benzene
[209–214]. Benzene is oxidized to biphenyl
by platinum- or palladium salts in glacial
acetic acid at 120 ◦ C; the salts are reduced
to the pure metals, and stoichiometric quan-
tities of the salts are required.
6) Dehydrocondensation of benzene in the pres-
ence of catalytic quantities of noble metal
salts requires reoxidation of the metals dur-
ing the course of the reaction. This may be
done with pure oxygen, but a partial pres-
sure of 6 MPa is required [215]. Additives
such as β-diketones [216], heteropolyacids
[217], or organic copper salts [218] are inca-
pable of significantly lowering the required
oxygen pressure, so that a potentially explo-
sive regime cannot be avoided [217]. Yields
of biphenyl are ca. 260 mol %, calculated on
the palladium salt charge.
7) Hydrodimerization of benzene to phenylcy-
clohexane, followed by dehydrogenation to
biphenyl [219–223]:
Suitable catalysts for this dimerization are
noble metals on molecular sieve, alkali met-
als on alumina or carbon, and transition met-
als such as nickel, tungsten, or zinc on the
zeolites faujasite or mordenite. A selectiv-
ity of up to 87 % has been claimed. Sub-
sequent dehydrogenation is carried out at
about 400 ◦ C over Pt – Al2 O3 [223], Cr2 O3 –
Na2 O – Al2 O3 [222, 224], Cr2 O3 – K2 CO3 –
Fe2 O3 [225], or Cr2 O3 – MgO – Al2 O3 [226]
giving yields of up to 99 %. Data concern-
ing residence times over the catalyst are not
available, either for the hydrodimerization or
the dehydrogenation processes.
Analysis. Gas chromatography is used to an-
alyze biphenyl and its derivatives.
Storage and Transportation. Biphenyl is
marketed as a solid in the form of plates, flakes,
or pellets, containing <10 ppm benzene and
Hydrocarbons 39
with a biphenyl content of >99.5 %; it is trans-
ported in bags or fiber drums; in the molten state
(ca. 120 ◦ C) in tank cars. It is not a regulated ma-
terial for transportation, on account of its toxi-
city; there are no restrictions for transportation
by sea, land, or air. The greatest hazards in han-
dling biphenyl are the risk of dust explosions
and the ignition of biphenyl vapor – air mixtures
over the molten product.
Uses. Biphenyl is an important heat transfer
agent, because of its high thermal stability. It is
used as a composite with diphenyl ether, espe-
cially as a melting point depressant. The eutectic
mixture of 26.5 % biphenyl and 73.5 % diphenyl
ether is marketed under the trade names Diphyl
(Bayer), Dowtherm A (Dow), Thermic (ICI),
Gilotherm (Rhône-Poulenc), Therm S 300 (Nip-
pon Steel), Santotherm VP and Therminol VP-1
(both Monsanto) for use at up to 400 ◦ C. The eu-
tectic begins to boil at 256 ◦ C; the vapor pressure
at the maximum operating temperature (400 ◦ C)
is ca. 1.1 MPa.
A further large quantity of biphenyl is used as
a dye carrier [227, 228]. Until a few years ago,
polychlorinated biphenyls (PCBs), generated by
chlorination of biphenyl, were produced in large
quantities for use as nonflammable hydraulic flu-
ids and as transformer dielectrics. Production
has largely ceased on ecological considerations
(→Chlorinated Hydrocarbons). Among the hy-
droxy derivatives, o- and p-hydroxybiphenyl and
p,p -dihydroxybiphenyl are used industrially. o-
Hydroxybiphenyl is used as a preservative and
fungicide. It is synthesized mainly from cyclo-
hexanone (→Phenol Derivatives). p-Hydroxybi-
phenyl and p,p -dihydroxybiphenyl are gener-
ated by sulfonation of biphenyl, followed by fu-
sion with alkali.
Economic Aspects. According to a 1976 es-
timate, 67 % of the annual United States pro-
duction of 40 000 t biphenyl was produced by
dealkylation of toluene and 33 % by thermal de-
hydrocondensation of benzene [229].
The annual production for 1984 is estimated
at 15 000 t, with an increasing proportion from
dealkylation of toluene. The production capaci-
ties of the major producers, Monsanto, (United
States), Bayer (Federal Republic of Germany),
and Monsanto (United Kingdom) have not been
40 Hydrocarbons
Table 19. Physical properties of terphenyls

Property o-Terphenyl m-Terphenyl p-Terphenyl Reference

[84-15-1] [92-06-8] [92-94-4]

mp, ◦ C 59 87 212 [233]

bp (101.3 kPa), ◦ C 332 365 376 [234]
Flash point, ◦ C 171 206 210 [236]
T crit , ◦ C 613 603 621 [237]
Pcrit , MPa 3.903 3.503 3.330 [238]
Heat of vaporization (at bp),
kJ/mol 58.3 64.2 62.6 [234]
Vapor pressure, kPa
93 ◦ C 0.01172 0.00165 [233, 239]
204 ◦ C 2.834 0.8274 0.789
315.6 ◦ C 64.4 27.3 24.56
426.7 ◦ C 439.9 240.6 174.4

disclosed. Based on the legal restrictions im-
posed on the use of PCBs, capacity should far
exceed current demand.
4.2. Terphenyls
Physical Properties. The three isomeric ter-
phenyls, C18 H14 , M r 230.29, are colorless to
pale yellow crystalline solids [230–235]. Some
physical properties are listed in Table 19.
steam volatile, and show the typical substitu-
tion reactions of aromatic hydrocarbons, such
as bromination and nitration. The substitution
pattern for the preferred positions of the three
terphenyls is shown in Table 20.
Production. Terphenyls are byproducts in
the production of biphenyl by dehydroconden-
sation of benzene; they are found in the high-
boiling fraction of the pyrolysis products. The
pot residue has the following approximate com-
position [240]:
3 – 8 %o-Terphenyl
44 %m – Terphenyl
24 %p – Terphenyl
1.5 %Triphenylene
22 – 27 %Higher polyphenylenes and tar.

When, in the early 1960s, terphenyl was pro-

posed to be used as coolant and moderator in
nuclear reactors [241], numerous reports were
published of attempts to enhance the yield of
terphenyl in dehydrocondensation reactions. In-
creasing the temperature or residence time was
ruled out because of increased carbonization.
Table 20. Terphenyl substitution pattern
These procedures also enhanced the biphenyl
Position for substitution fraction, so that the latter had to be recycled
First Second Third Fourth
to the reactor [193, 197, 242–244]. Recycling a
concentration of up to 30 % biphenyl has been
o-Terphenyl 4- 4 - 4 - 5 - described [193]. A practical level appears to be
m-Terphenyl 4 - 4- 4 -
p-Terphenyl 4- 4 - 2 - using benzene with 10 % biphenyl [210]. At-
tempts have also been made to recycle the ex-
tremely high melting p-terphenyl, after isolation
from the reaction mixture, to obtain a lower melt-
Chemical Properties. Terphenyls, like bi- ing terphenyl mixture [242]. The o-derivative
phenyls, are thermally extremely stable organic is readily separated from the terphenyl isomer
compounds, again offering a potential for use as mixture by distillation; m- and p-terphenyl dis-
heat transfer media. Terphenyls are sublimable, till together and the pure isomers can be ob-

tained by zone refining [245]. Terphenyl iso-
mers, including hydrogenated terphenyl, can be
analyzed by gas chromatography. The sole pro-
ducer of terphenyl in the United States is Mon-
santo. The market amounts to several thousand
tons per year. Terphenyl is available in Europe
from Bayer.
Storage and Transportation. Solid ter-
phenyl is shipped as flakes in laminated bags or
in fiber drums, liquid, hydrogenated terphenyl
in tank cars, steel drums, or barrels. Terphenyls
and hydrogenated terphenyls possess a low toxi-
city, and are therefore not regarded as dangerous
goods for the purpose of transportation.
Uses. Terphenyl is used as a heat transfer
agent because of its excellent thermal stabil-
ity; a mixture with a composition of 2 – 10 % o-
terphenyl, 45 – 49 % m-terphenyl, 25 – 35 % p-
terphenyl, and 2 – 18 % higher polyphenyls is
used. This composition approximates the pot
residue from the distillation of biphenyl. This
mixture has bp ca. 360 ◦ C; a disadvantage is
the high mp, 145 ◦ C. The largest fraction is par-
tially hydrogenated to yield a clear oil, mis-
cible with hydrocarbons and chlorinated hy-
drocarbons. Hydrogenation lowers the setting
point and the viscosity but unfortunately also
reduces the thermal stability. Partially hydro-
genated terphenyl [246] is used mainly as a
dye carrier for pressure sensitive recording ma-
terials and copy paper, and as a heat transfer
oil. Heat transfer oils consisting of partially hy-
drogenated terphenyl isomers are marketed un-
der the following trade names: Santotherm 66
or 88, Therminol (Monsanto, United States),
Gilotherm (Rhône-Poulenc, France), and Therm
S 900 (Nippon Steel, Japan). Their usage range
extends to 340 ◦ C.
4.3. Polyphenyls.
Polyphenyls with four or more phenyl residues
have generally been synthesized as p-linked iso-
mers and are used as scintillators [247]. They
are economically unimportant.
A symmetrical hydrocarbon within this class
is 1,3,5-triphenylbenzene, C24 H18 , mp 174 ◦ C,
prepared by condensation of three molecules of
acetophenone.
Hydrocarbons 41
5. Hydrocarbons from Coal Tar
Anthracene and Naphthalene and Hydronaph-
thalenes are separate keywords.
Hydrocarbons from coal tar are generally
recovered in large-scale operations by distil-
lation and crystallization, after the polar, co-
boiling components, phenols and nitrogen bases,
have been separated by extraction (→Tar and
Pitch). Supplies of coal tar as a raw material
have been sufficient to date to meet the demand
for these hydrocarbons, so that no other produc-
tion processes, such as syntheses, have yet been
developed to the stage of commerciality. Hydro-
carbons from coal tar are used predominantly in
the synthesis of dyes, plastics, and pharmaceu-
ticals.
5.1. Acenaphthene
Acenaphthene [83-32-9], C12 H10 , M r 154.21,
was discovered in coal tar in 1867 by Berth-
elot.
Physical Properties. Acenaphthene forms
colorless needles; readily soluble in chloroform,
toluene, and hot ethanol; soluble in cold ethanol,
hot diethyl ether, and benzene; insoluble in wa-
ter.
mp 95.4 ◦ C
bp (101.3 kPa) 277.5 ◦ C
 (25 ◦ C) 1.2195 g/cm3
n100 1.6048
D
T crit 530 ◦ C
Heat of fusion 1.344×105 J/kg
Heat of vaporization (40 ◦ C) 3.48×105 J/kg
Heat of combustion (25 ◦ C) 4.033×107 J/kg
Chemical Properties. Acenaphthene reacts
with halogens, preferentially at 3-, 5-, and 6-
positions, or at the 1-position when irradi-
ated with visible light. Nitration and sulfona-
tion occur at the 3-, 5-, or 6-positions. Cat-
alytic hydrogenation leads to tetra- and decahy-
droacenaphthene. Decahydroacenaphthene can
42 Hydrocarbons
be isomerized in the presence of Lewis acids
to 1,3-dimethyladamantane. Oxidation of ace-
naphthene gives naphthalic acid or its anhydride,
acenaphthenequinone, acenaphthenol, and ace-
naphthenone.
Production. High-temperature coal tar con-
tains, on average, 0.3 % of acenaphthene; in ad-
dition, it is formed under the conditions of coal
tar distillation by hydrogenation of acenaphthy-
lene, which is found in crude tar at a level of 2 %.
Acenaphthene is concentrated to ca. 25 % in the
coal tar fraction boiling between 230 and 290 ◦ C
(wash oil), which is recovered at ca. 7 % in con-
tinuous tar distillation. The readily crystallizable
acenaphthene fraction, boiling at 270 – 275 ◦ C,
is obtained from the wash oil by redistillation,
or by removing it directly during primary tar
distillation. From this, technical acenaphthene
(95 – 99 % pure) is produced by crystallization
in agitated coolers and centrifuges, or by contin-
uous counter-current crystallization [256, 259].
The pure compound is obtained by further dis-
tillation and recrystallization.
Uses. Acenaphthene is used on a large-scale
for synthesis of naphthalic anhydride by gas-
phase with air oxidation at 300 – 400 ◦ C, in the
presence of vanadium containing catalysts [258,
259]. Liquid-phase oxidation with chromate or
air in the presence of cobalt- or manganese
acetate at 200 ◦ C [260], or cobalt resinate at
120 ◦ C [261] produces naphthalic acid (→Car-
boxylic Acids, Aromatic, Chap. 4.4.). Produc-
tion of naphthalene-1,4,5,8-tetracarboxylic acid
from acenaphthene by condensation with mal-
onic acid dinitrile [262] is at present not indus-
trially important. A technically feasible proce-
dure has been developed using 5-methyl-peri-
acenaphthindene-7-one [42937-13-3] (by re-
action of acenaphthene with diketene in HF)
(→Carboxylic Acids, Aromatic, Chap. 4.7.),
[263]. Naphthalic anhydride and naphthalene-
1,4,5,8-tetracarboxylic acid are used as inter-
mediates for production of perylene and peri-
none pigments (→Pigments, Organic) [262].
Other valuable intermediates, particularly for
the production of dyes and optical brighteners,
and also for pharmaceuticals and pesticides,
are produced from, e.g., 4-bromacenaphthene
[264] (for 4-bromonaphthalic anhydride
or naphthalic anhydride-4-sulfonic acid),
from 5,6-dichloracenaphthene (for 4,5-di-
chloronaphthalic anhydride) [265] and from
3,5,6-trichloracenaphthene [266]. Nitration
and oxidation of acenaphthene produce 4-
nitronaphthalic anhydride, from which the flu-
orescent pigment Solvent Yellow 44 can be
made [267]. Reaction of acenaphthalene with
sodium cyanate in hydrofluoric acid gives
acenaphthene-5,6-dicarboxylic acid imide,
which can be used as an intermediate for dyes
[268].
Bromination of acenaphthene in the pres-
ence of, e.g., iron(III)- or aluminum chloride,
followed by further bromination by the addi-
tion of a radical initiator, such as azobisisobu-
tyronitrile (AIBN), and subsequent dehydro-
bromination produces condensed bromacenaph-
thene, which is particularly well-suited for ren-
dering plastics nonflammable and radiation-
resistant [269]. Thermally stable plastics can
be obtained from acenaphthene in the follow-
ing ways: 1) by condensation with formalde-
hyde; 2) oxidation and reaction with an aromatic
polyamine [270]; by cocondensation with phe-
nol and formaldehyde [271]; and by conden-
sation with terephthaloyl chloride [272]. Ace-
naphthene also serves as a feedstock for the pro-
duction of acenaphthenequinone, acenaphthene,
and dimethyladamantane [273].
Derivatives.
Acenaphthenequinone [82-86-0], C12 H6 O2 ,
M r 182.17; mp 261 ◦ C; yellow needles; slightly
soluble in ethanol; soluble in hot benzene and
toluene. It is formed by liquid-phase oxidation of
acenaphthene with hydrogen peroxide or dichro-
mate in acetic acid, by chlorination of dibro-
macenaphthylene followed by acid hydrolysis,
or by condensation of acenaphthene with ethyl
nitrite and separation of the dioxime. Acenaph-
thenequinone is suitable for use as an insecticide
and fungicide and can be converted into vat dyes.
Acenaphthylene [208-96-8], C12 H8 , M r
152.18; mp 92 – 93 ◦ C; yellow prisms; soluble
in ethanol, diethyl ether, and benzene; insoluble
in water. It is found in coal tar at a concentra-

tion of 2 % and can be produced industrially by
catalytic dehydrogenation of acenaphthene in
the gas phase. Thermal, ionic, radical, or radio-
chemically induced polymerization produces
polyacenaphthylene [274]. Polyacenaphthylene
and copolymers with other olefins are charac-
terized by high thermal and mechanical stability
together with good electrical insulation prop-
erties. The thermal stability of polystyrene can
be significantly increased by copolymerisation
with acenapthylene [274]. Polymerisation of
acenaphthylene with acetylene in the presence
of a Lewis acid catalyst gives electrically con-
ductive polymers; they are used for electronic
engineering and for the antistatic finishing of
plastics [275]. Thermosetting resins with good
resistance to heat and chemicals can be obtained
by cocondensation with phenol and formalde-
hyde [271]. Acenaphthylene possesses excellent
properties as an antioxidant in cross-linked poly-
ethylene and ethylene – propylene rubber [274].
Thermal trimerization of acenaphthylene leads
to decacyclene, which can be further processed
to sulfur dyes [276].
5.2. Indene
Indene [95-13-6], C9 H8 , M r 116.16, was disco-
vered in coal tar in 1890 by Kraemer and
Spilker.
Physical Properties. Indene is a colorless
liquid; insoluble in water; soluble in ethanol;
it is miscible with diethyl ether, naphtha, pyri-
dine, carbon tetrachloride, chloroform, and car-
bon disulfide.
mp −1.5 ◦ C
bp (101.3 kPa) 183 ◦ C
 (20 ◦ C) 0.9966 g/cm3
n20
D 1.5763
Heat capacity (25 ◦ C) 1.610×103 J/kg
Heat of fusion 8.33×104 J/kg
Heat of combustion (25 ◦ C) 4.132×107 J/kg
Dipole moment (in CCl4 ) 0.85 D
Chemical Properties. Indene polymerizes
readily at ambient temperature and in the
Hydrocarbons 43
dark. Polymerization is accelerated by heat,
acid, and Friedel – Crafts catalysts. Depending
on reaction conditions, oxidation leads to di-
hydroxyindan, homophthalic acid, or phthalic
acid. Hydrogenation produces indan. Halogena-
tion occurs at the 2,3-double bond. Diene
adducts are formed with maleic anhydride and
hexachlorocyclopentadiene.
Production. Indene is found in high-tempe-
rature coal tar at an average concentration of
1 %. Pyrolysis residual oils from olefin produc-
tion contain varying amounts of indene [277],
which can be separated by extractive distilla-
tion with N-methyl-2-pyrrolidone [278] or by
crystallization [279]. Indene is recovered indus-
trially from dephenolated and debased tar light
oil by rectifying distillation followed by crystal-
lization. When the phenol extraction is omitted
a highly concentrated indene fraction can also
be obtained by azeotropic distillation of an in-
dene – phenol fraction with water, whereby the
phenol is separated as a bottom product [280].
Uses. Indene is used as a comonomer in coal-
tar derived indene – coumarone and petroleum-
derived aromatic hydrocarbon resins (→Resins,
Synthetic). Technical pure indene is used for the
production of indan and its derivatives. Esters of
indene-1-carboxylic acid, which are effective as
acaricides, can be synthesized from indene for
use in, e.g., tick collars [281].
Derivatives.
Indan [496-11-7], hydrindene, C9 H10 , M r
118.18;
mp −51.4 ◦ C
bp (1013 kPa) 176 ◦ C
 (20 ◦ C) 0.9644 g/cm3
Indan is a colorless liquid; insoluble in water;
soluble in ethanol and diethyl ether. It is found at
a concentration of 0.1 % in coal tar, but is mainly
produced by catalytic hydrogenation of indene.
44 Hydrocarbons
Alkylated indans are used in the production of
synthetic lubricants (→Lubricants and Related
Products). Indan is used for the synthesis of in-
danols (→Phenol Derivatives).
2-Indanol [4254-29-9], C9 H10 O, M r 134.18;
mp 69 ◦ C; colorless crystals. It is obtained by hy-
drogenation of 2-indanone prepared by reaction
of indene with hydrogen peroxide and boiling in
sulfuric acid over a platinum catalyst in ethanol.
2-indanol is used in the synthesis of the coro-
nary therapeutic, Aprindine [37640-71-4], N-
(3-diethylaminopropyl)-N-indan-2-ylaniline;
(→Antiarrhythmic Drugs), [282] and can be
employed as an intermediate for other pharma-
ceuticals, e.g., for antihypertonics [283].
5-Indanol [1470-94-6], C9 H10 O, Mr
134.18; mp 55 ◦ C; colorless crystals. It is pro-
duced by fusion of indan-5-sulfonic acid with
potassium hydroxide [284]. 5-Indanol is used in
the synthesis of the multi-purpose antibiotic,
Carindacillin [26605-69-6] (6-[2-phenyl-2-
(5-indanyloxycarbonyl)acetamido] penicillanic
acid; →Antibiotics), [285]. 5-Indanylesters of
substituted picolinic acids can be used as anti-
hypertonic [286]. 5-Indanol itself may be used
as an antidandruff additive in hair shampoos
(→Hair Preparations) [287].
4-Indanol [1641-41-4], C9 H10 O, M r 134.18;
mp 49 – 50 ◦ C; colorless crystals. Analogous to
5-indanol it is synthesized by potassium hydrox-
ide fusion of indan-4-sulfonic acid, which is
produced as a byproduct in indan sulfonation.
Catalytic gas-phase amination of 4-indanol pro-
duces 4-aminoindan [288], which is required
as an intermediate, e.g., for pharmaceuticals.
Clorindanol (7-chloro-4-indanol) is used as an
antiseptic.
5.3. Fluorene
Fluorene [244-36-0], C13 H10 , M r 166.22, was
discovered in coal tar in 1867 by Berthelot.
Physical Properties. Fluorene forms color-
less flakes with slight violet fluorescence; sub-
limes readily; insoluble in water; soluble in ben-
zene, diethyl ether, carbon disulfide, and hot eth-
anol; soluble with difficulty in cold ethanol.
mp 115 ◦ C
bp (101.3 kPa) 298 ◦ C
 (20 ◦ C) 1.181 g/cm3
Heat of fusion 1.21×105 J/kg
Heat of combustion (25 ◦ C) 3.994×107 J/kg
Chemical Properties. Fluorene is capa-
ble of numerous chemical reactions, through
both the aromatic rings and especially the reac-
tive methylene group. Oxidation produces flu-
orenone. Reaction with dialkyl carbonate and
alkali metal hydride or alcoholate, followed by
neutralization and saponification of the resulting
ester produces fluorene-9-carboxylic acid; the
latter is also produced by metalation with butyl
lithium followed by carboxylation. Nitration
leads predominantly to 2-nitrofluorene, chlo-
romethylation to 2-chlormethylfluorene. Halo-
genation, depending on the reaction conditions,
gives the mono-, di-, or tri-haloderivatives, sub-
stitution taking place at the 2-, 2,7-, or 2,4,7-
positions; in the presence of catalysts waxy
products containing ca. 50 wt % chlorine are
formed.
Production. High-temperature coal tar con-
tains on average, 2 % fluorene. The latter is reco-
vered industrially by redistillation of coal tar
wash oil or by direct removal of a fluorene frac-
tion during primary distillation of coal tar, fol-
lowed by recrystallization of the fluorene frac-
tion (dephenolated and debased if necessary)
e.g., from solvent naphtha or naphtha [289]. Flu-
orene can be isolated from a ≥60 % fluorene
fraction by continuous counter-current crystal-
lization [290]. It is possible to refine the tech-
nical product further via the sodium compound
(by reaction with sodium or sodium amide), or
by sulfuric acid purification and crystallization
from methanol.
Uses. Fluorene is the starting material for
production of fluorenone and fluorene-9-carbox-
ylic acid. Fluorene-9-carboxylic acid is an in-
termediate in the production of antidiabetic and
antiarrhythmic drugs, and plant growth regula-
tors [291]. Plant growth regulators are also ob-
tained by reaction of fluorene with phthalic an-
hydride [292]. Fluorene-2-acetic acid (by acyla-
tion of fluorene) can be used as a precursor in

the synthesis of anti-inflammatory drugs [293].
Reaction of fluorene with sulfur produces elec-
trically conductive polymers [294]. The azine
dye Sirius Light Violet FRL is produced from 2-
aminofluorene (→Azine Dyes, Chap. 5.2.). The
fluorene analog of malachite green dye is ob-
tainable from 3,6-bis-(dimethylaminofluorene)
[295]. 2,7-Diiodofluorene can be employed in
the production of styryl dyes for optical bright-
eners [296]; heptabromofluorene can be used as
a fire-retardant for plastics [297]. Fluorene it-
self can be used in admixture with biphenyl and
phenyltoluene as a carrier for dyeing polyester
fibers [298].
Derivatives.
9-Fluorenone [486-25-9], C13 H8 O, Mr
180.21;
mp 85 – 86 ◦ C
bp 101.3 kPa) 341.5 ◦ C
 (99.4 ◦ C) 1.130 g/cm3
nD (99.4 ◦ C) 1.6369
9-Fluorenone forms yellow crystals; insol-
uble in water; soluble in hot sulfuric acid;
very readily soluble in ethanol and diethylether;
volatile in steam. Fluorenone can be produced by
catalytic oxidation of fluorene [261, 299–302],
or of fluorene fractions in the presence of a qua-
ternary ammonium salt [303], or by catalytic ox-
idative cracking (oxicracking) of suitable aro-
matic [304].
Di- and trinitrofluorenones are synthesized
from fluorenone for use as electron accep-
tors in electrophotography, e.g., in copying
systems [305]. 2,4,7-Trinitrofluorenone is used
as an analytical reagent for polynuclear aro-
matic hydrocarbons. Fluorenone can be used
as starting material for the synthesis of im-
idazolylfluorene salts with antimycotic activ-
ity [306]. 9,9-Bis(4-hydroxylphenyl)fluorene
(from fluorenone by reaction with phenol [307])
Hydrocarbons 45
can be used as starting material for ther-
mally stable plastics [308]. 9,9-Bis(4-methyl-
aminophenyl)fluorene (from fluorenone by re-
action with phenyl methyl amine) has been sug-
gested as a hardener for epoxy resins [309]. Flu-
orenone is also suitable as an oxidizing agent in
the Oppenauer reaction.
5.4. Fluoranthene
Fluoranthene [206-44-0], C16 H10 , M r 206.26,
was isolated from coal tar by Fittig and Geb-
hardt in 1878.
Physical Properties. Colorless crystals with
light blue fluorescence; insoluble in water; read-
ily soluble in diethyl ether, boiling ethanol, chlo-
roform, carbon disulfide, and glacial acetic acid.
mp 110 ◦ C
bp (101.3 kPa) 384 ◦ C
 (20 ◦ C) 1.236 g/cm3
Heat of vaporization 3.29×105 J/kg
Heat of combustion (25 ◦ C) 3.913×107 J/kg
Chemical Properties. Oxidation of
fluoranthene with chromic acid leads to
fluorenone-1-carboxylic acid via fluoranthene-
2,3-quinone; hydrogenation leads to 1,2,3,10 b-
tetrahydrofluoranthene, and to perhydroflu-
oranthene via decahydrofluoranthene. 2,3-
Dihydrofluoranthene is obtained by metalation
with sodium. Halogenation, nitration, and sul-
fonation take place predominantly at the 4-
position. Condensation with phthalic anhydride
in the presence of aluminum chloride gives a
mixture of 7,8- and 8,9-phthaloyl fluoranthene.
Production. Fluoranthene belongs to the
main constituents of high-temperature coal tar,
which contains, on average, 3.3 % of fluoran-
thene. It is recovered from the fluoranthene frac-
tion, which boils at 373 – 385 ◦ C, and is obtained
from the high-boiling anthracene oil II and from
pitch distillate by redistillation. Subsequent re-
crystallization of the fluoranthene fraction from
46 Hydrocarbons
solvent naphtha gives 95 % pure fluoranthene.
Further refining is carried out by recrystalliza-
tion from xylene with simultaneous partial sul-
fonation of the impurities with ca. 1 % of con-
centrated sulfuric acid, or by recrystallization
from a pyridine – water mixture.
Uses. Fluoranthene is used in the pro-
duction of fluorescent dyes (→Fluorescent
Dyes). Yellow vat dyes are obtained by
condensation with phthalic anhydride, via
phthaloyl fluoranthenes, whereas bromina-
tion [310] and condensation with 1-amino-4-
benzoyl-amino anthraquinone produces olive
dyes [311]. 2-Benzylfluoranthene (from 2,3-
dihydrofluoranthene [312] and benzaldehyde)
can also be used for the production of dye
intermediates [313]. Fluoranthene and alkyl
fluoranthenes can be used as stabilizers in
epoxy resin adhesives [314], as additives in
electrically-insulating epoxy resin compositions
[315], and in electrically-insulating oils [316].
2,3-Dihydrofluoranthene and 1,2,3,10 b-tetra-
hydrofluoranthene [317] are used for numerous
derivatives, some of which (e.g., their deriva-
tives of phenylethyl amine) are of pharmaceuti-
cal interest. 7-Fluoranthenyl aminoalcohols (for
cytostatics, anthelmintics, or bactericides [318])
and bis-aminoketones, e.g., 3,8-bis(4-piperi-
dinobutyryl)fluoranthene (as antiviral agents
[319]) can be synthesized from fluoranthene.
5.5. Phenanthrene
Phenanthrene [85-01-8], C14 H10 , M r 178.24
was discovered in coal tar by Fittig and Os-
termayer in 1872.
Physical Properties. Colorless flakes with
weak blue-violet fluorescence; readily sub-
limable; insoluble in water; slightly soluble in
ethanol; readily soluble in diethyl ether, ben-
zene, carbon disulfide, and glacial acetic acid.
mp 101 ◦ C
bp (101.3 kPa) 338.5 ◦ C
 (20 ◦ C) 1.172 g/cm3
Heat capacity (25 ◦ C) 1.361×103 J/kg
Heat of fusion 1.04×105 J/kg
Heat of combustion (25 ◦ C) 3.982×107 J/kg
Chemical Properties. Phenanthrene can
be oxidized to phenanthrenequinone, diphenic
acid, or to phthalic acid, depending on the ox-
idant. Hydrogenation produces 9,10-di-, tetra-,
octa-, or perhydrophenanthrene. Phenanthrene
is halogenated predominantly in the 9,10-
positions, and nitrated in the 9-position. Sul-
fonation leads to mixtures of 2-, 3-, and 9-
phenanthrene sulfonic acids. For acylation see
→Acylation and Alkylation, Chap. 2.2.2.
Production. Phenanthrene, at a concentra-
tion of 5 %, is the second most important coal
tar constituent in terms of quantity after naph-
thalene. During primary distillation of coal tar,
it is concentrated in the anthracene oil fraction.
After crystallization of the anthracene residues
the phenanthrene is recovered as a fraction from
the filtrate of this crystallization, or from the top
fraction of crude anthracene distillation, by re-
distillation. Technically pure grades of phenan-
threne are obtained by sulfuric acid refining and
recrystallization from methanol, or by repeated
rectification of the phenanthrene fraction. The
accompanying substances can be separated ei-
ther by partial sulfonation, or by partial conden-
sation with formaldehyde and hydrogen chlo-
ride. The most persistent accompanying sub-
stance, diphenylene sulfide, can be completely
removed by treating the melt with sodium and
maleic anhydride.
Uses. Phenanthrene forms the basis for pro-
duction of 9,10-phenanthrenequinone and 2,2 -
diphenic acid (see Chap. 4) [320]. It can be
used to synthesize anthracene, via the isomer-
ization product of sym-octahydrophenanthrene
(→Anthracene). Electrically conductive sub-
stances, e.g., for use in batteries and solar cells,
can be produced by the electrochemical con-
version of phenanthrene diazonium salts in a
solvent containing a conductive salt, and sub-
sequent doping with various ions (e.g., Na+ ,
Ba2+ , H+ , etc.) [321]. Liquid-crystalline 7-
n-alkyl-9,10-dihydrophenanthrene-2-carboxyl-
ic acid ester, used for optical-electronic ap-
 Hydrocarbons 47

plications, can be synthesized from 9,10-di- pharmaceuticals (e.g., immuno-suppressives)

hydrophenanthrene [322]. By cross-linking or fungicides [344, 345].
with p-xylylene glycol and 4-toluenesulfonic
acid, polycondensed thermosetting resins
are obtained for composites or temperature-
resistant, electrically insulating coatings [323].
A polyamide – polyimide resin can be produced
by oxidation of phenanthrene to phenanthrene-
9,10-quinone and -9,10-diol, condensation with
formaldehyde, oxidation to the polycarbox- 5.6. Pyrene
ylic acid, formation of the anhydride and fi-
nally reaction with an aromatic diamine. This Pyrene [129-00-0], C16 H10 , M r 202.26, was
resin is suitable for use in high-temperature found in coal tar by Graebe in 1871.
insulators, printed circuit boards, and lami-
nates [324]. Phenanthrene has been proposed
as a plasticizer for plastics and molding com-
pounds [325], phenanthrene and alkylphenan-
threnes have been suggested as stabilizers for
mineral oil products [326, 327]. Deep khaki
vat dyes are prepared by heating a mixture of Physical Properties. Colorless crystals with
phenanthrene and anthracene with sulfur [328]. blue fluorescence; insoluble in water; very read-
[(Phenanthrylmethyl)amino]-propandiol, which ily soluble in diethyl ether, benzene, and carbon
possesses antitumor activity, can be synthesized disulfide; poorly soluble in ethanol.
via phenanthrene-9-carboxaldehyde [329].
mp 150 ◦ C
Derivatives. bp (101.3 kPa) 393.5 ◦ C
 (23 ◦ C) 1.271 g/cm3
Phenanthrene-9-10-quinone [84-11-7], Heat capacity (18 ◦ C) 1.126×107 J/kg
C14 H8 O2 , M r 208.91; mp 217 ◦ C; orange nee- Heat of vaporization 3.21×105 J/kg
dles; poorly soluble in water; slightly soluble Heat of combustion (25 ◦ C) 3.878×107 J/kg
in ethanol; soluble in boiling water and in di- Dielectric constant (17 ◦ C) 3.21
ethyl ether; readily soluble in hot glacial acetic
acid. Phenanthrene-9,10-quinone is produced
Chemical Properties. Pyrene is oxidized
by liquid-phase oxidation of phenanthrene with
by mild oxidants such as chromic acid to
chromates [330], catalytically with oxygen
pyrene-1,6- and 1,8-quinone; further oxida-
[331], with tert-butylhydroperoxide [332, 333]
tion leads to perinaphthenone [548-39-0] and
or via phenanthrene-9,10-oxide with nitric acid
napthalene-1,4,5,8-tetracarboxylic acid. Hydro-
or hypochlorite [334, 335]. Further oxidation
genation produces, depending on the reaction
transforms the quinone into 9-hydroxyfluorene-
conditions, ditetra-, hexa- or decahydropyrenes;
9-carboxylic acid, which is used in the form of
halogenation products substituted in the 1-, 1,6-,
its salts or esters (in combination with the herbi-
1,8-, 1,3,6- and 1,3,6,8-positions. The substitu-
cide methyl chlorophenoxyacetic acid (MCPA),
tion pattern for halogenation can be changed in
trade name: Aniten) or as the 2-chloro derivative
favor of the 4-position by Diels – Alder reac-
(trade name: Maintain) as plant growth regula-
tion with hexachlorocyclopentadiene [77-47-4]
tors [336, 337]. Phenanthrene-9,10-quinone has
in the presence of iron powder or thallium
been suggested as an additive for photographic
acetate, followed by retro-Diels – Alder reac-
or electrophotographic applications [308, 338,
tion. With nitration and sulfonation the 1-, 1,6-,
339], in UV-curable coatings and adhesives
and 1,3,6,8-substitution products are formed.
[340, 341] and in production of cellulose by
Pyrene can be condensed with phthalic anhy-
wood pulping [342]. An intermediate for azo
dride by Friedel – Crafts reaction to give diph-
pigments can be obtained by condensation with
thaloyl pyrenes, and with benzoyl chloride to
hydrazinobenzoic acid [343]. Condensation
the dye pyranthrone [128-70-1].
with aromatic amines gives intermediates for
48 Hydrocarbons
Production. High-temperature coal tar con-
tains an average of ca. 2 % of pyrene. It is
recovered from a fraction crystallizing above
110 ◦ C, which is obtained by redistillation of
the high-boiling anthracene oil II or pitch dis-
tillate. Pure pyrene is produced by recrystal-
lization, e.g., from solvent naphtha [346] or by
fractional crystallization from the melt [256],
dephenolation and debasing, and by refining
with 80 % sulfuric acid. As an alternative to
sulfuric acid refining, the pyrene-accompanying
brasane (2,3-benzodiphenylene oxide) can also
be separated by recrystallization from xylene by
adding iron(III) chloride. Traces of tetracene are
removed by reaction with maleic anhydride.
Uses. The dye intermediate naphthalene-
1,4,5,8-tetracarboxylic acid [128-97-2] can be
produced by halogenation of pyrene [347,
348] (halogenation of the pyrene fraction is
also possible [349]) followed by oxidation
(→Carboxylic Acids, Aromatic, Chap. 4.7.),
[350]. Other pyrene-based dyes are Sirius Light
Blue F 3 GL from 3-aminopyrene (→Azine
Dyes, Chap. 5.2.) and the green fluorescent
dye, pyranin (Solvent Green 7), from pyrene-
1,3,6,8-sulfonic acid [351]. Anthraquinone dyes
(e.g., pyranthrone) may be obtained from diben-
zoyl pyrene or diphthaloyl pyrene. Optical
brighteners may be synthesized by reaction of
pyrene with a complex of cyanuric chloride
and aluminum chloride (→Optical Brighteners)
[352]. By analogy to fluoranthene, pyrene and
alkylpyrenes can be used as additives in electro-
insulating oils [316], and in epoxy resins for
electrical-insulation [315]. In a similar man-
ner to phenanthrene, thermosetting resins may
be formed with p-xylene glycol and 4-toluene-
sulfonic acid [323]. Condensation products of
1-bromopyrene or nitropyrene with formalde-
hyde are used as photoconductive substances in
electrophotography [305, 353, 354]. 1,2,3,6,7,8-
Hexahydropyrene is effective as a synergist for
dialkylsulfide antioxidants in lubricants [355].
On the basis of pyrene, pyrenyl aminoalcohols
can be synthesized as cytostatics, anthelmintics,
or bactericides [356]. Pyrene itself can serve as
an electron donor to increase the blackness in
pencil leads [357].
6. Toxicology and Occupational
Health
6.1. Alkanes
Methane is toxicologically virtually inert.
In very high concentrations (80 – 90 vol %) it
causes respiratory arrest [358]. The bulk of an
inhaled dose is exhaled unchanged [359].
Ethane easily causes dyspnea in laboratory
animals, beginning at concentrations of approx-
imately 2 – 5 vol % in the air inhaled [360].
At high concentration respiratory arrest occurs
[361]. A slight sensitization of the myocard to
catecholamines has been described at a concen-
tration of 15 – 19 vol % [362].
Propane has narcotic properties [361]. It is
neither a skin nor a mucosal irritant. At very
high concentration it also causes respiratory ar-
rest [363]. In dogs adverse effects on circulatory
function were found with inspiratory concentra-
tions of 1 % or greater. A negative inotropic ef-
fect, a drop in the mean aortic pressure, and an
increase in pulmonary vascular resistance were
found [364].
MAK (1987): 1000 ppm.
Butane causes slight drowsiness beginning
at a concentration of 10 000 ppm in the air in-
haled [365]. In the mouse the inhalatory LC50
is 680 g/m3 [366]. It has been proven in the dog
that butane also sensitizes the myocardium to
the effects of catecholamines [364].
MAK (1987): 1000 ppm.
n-Pentane has a less pronounced narcotic ef-
fect than the C1 – C4 hydrocarbons. In humans
the lowest lethal dose after acute inhalatory ex-
posure is apparently 130 000 ppm, the lowest
toxic dose 90 000 ppm [367]. Pentane apparently
possesses a neurotoxic effect, although it is not
very pronounced [368, 369].
n-Hexane possesses a marked neurotoxic ef-
fect that distinguishes it from the other alkanes.
The causal factor for peripheral neuropathies
is the oxidation product (cytochrome P 450-
dependent oxidases) 2,5-hexadione. Toxicity
manifests itself clinically in polyneuropathies
[370, 371]. Hexane is also a skin irritant. The
LC50 (inhalation mice) is 120 g/m3 [372]. In rats
the acute oral LD50 is 24 – 49 mL/kg [373]. Hex-
ane is absorbed through the skin. Dermal expo-
sure can lead to poisoning (LD dermal in rab-
bits ca. 5 mL/kg) [374]. In rats subchronic expo-

sure to a concentration of 400 – 600 ppm leads to
neuropathies [375], which are characterized by
degeneration of both myelin sheaths and axons
[376–378].
MAK (1987): 50 ppm.
n-Heptane has narcotic properties. In hu-
mans 0.1 % in the inhaled air leads to dizzi-
ness and 0.5 % to equilibrial disturbances with
loss of motor coordination [365]. Within 3 min
4.8 vol % in the air inhaled leads to asphyxia
[379]. Like hexane, heptane is a skin and mu-
cosal irritant [365]. The biotransformation of n-
heptane, as with hexane, is by oxidation. Hep-
tane apparently also possesses neurotoxic prop-
erties [380]. Myocardial sensitization to cate-
cholamines has also been found [381].
MAK (1987): 500 ppm.
n-Octane is similar to n-heptane in regard to
its narcotic effect; however, it does not seem to
cause any other effects on the nervous system
[372]. Octane is also metabolized by oxidation
[382].
MAK (1987): 500 ppm
Only limited toxicological data are available
on the higher molecular mass alkanes.
6.2. Alkenes
The alkenes under consideration are toxicolog-
ically not very active. Higher molecular mass
compounds possess narcotic properties. The
alkenes are not neurotoxic. The α-olefins gener-
ally seem to be more reactive and toxic than the
β-isomers.
6.3. Alkylbenzenes
Trimethylbenzenes. The toxicological data
available on trimethylbenzenes is mainly rather
old. The acute inhalatory toxic dose for the 1,2,4-
and the 1,3,5-isomers is in the range 7000 –
9000 ppm [383, 384].
As a result of their occurrence in automo-
tive fuel and heating oil, low concentrations
of trimethylbenzenes can be demonstrated in
air and water. For example, measurements in
1974 in a tunnel in Rotterdam indicated a
level of 0.015 ppm pseudocumene; see [385] for
more details. An analysis of the drinking water
Hydrocarbons 49
of Cincinnati, Ohio, in 1980, revealed a con-
centration of 45 ng/L hemimellitene, 127 ng/L
pseudocumene, and 36 ng/L mesitylene [386].
Trimethylbenzenes are toxic to aquatic organ-
isms; LC50 values between 10 and 100 ppm have
been reported [387].
The TLV (TWA) level for trimethylbenzene is
25 ppm [388]. The MAK for Trimethylbenzenes
(all isomers (1,2,3–;1,2,4–;1,3,5–) has been set
at 20 ppm. [389]. The United States toxicity as-
sigment ranges from low to moderately toxic
for pseudocumene, to highly toxic for the iso-
mer mixture; disturbances in the central nervous
system and abnormal blood pictures have been
reported [387].
Tetramethylbenzenes. The oral toxicity of
tetramethylbenzenes is low; LD50 >5000 mg/kg
(rat, oral). Durene is additionally classified as in-
travenously highly toxic; LD50 180 mg/kg (i.v.,
mice). Tetramethylbenzenes cause mild skin re-
actions [387]. Two values are cited for the odor
threshold concentration of durene: 0.083 and
0.087 mg/m3 [385].
Hexamethylbenzene. Hexamethylbenzene
has low oral toxicity; LDLo 5000 mg/kg (rat).
It is, however, suspected of causing neoplastic
effects. Addition of nitromethane to hexam-
ethylbenzene can lead to an explosive reaction
[387].
Diethylbenzenes. Diethylbenzene is more
toxic than monoethylbenzene (→Ethylbenzene)
[390]. DEB is emitted into the environment from
engine fuel and heating oil [386]. Diethylben-
zenes are toxic to aquatic organisms; the LC50
is between 10 and 100 mg/L [387]. In the cell
multiplication inhibition test with protozoa, the
toxicity threshold is <10 mg/L [385]. In the Fed-
eral Republic of Germany DEB is classified as
an aquatic hazard; aquatic hazard class 2. DEB
causes mild to moderate irritation of the eye
and mucous membranes. The oral toxicity is
low: LDLo 5000 mg/kg (rat) [387]. The vapor
possesses an anesthetic action and may cause
headache, vertigo, or vomiting; the odor thresh-
old concentration is <10 ppm [391].
Triethylbenzene. The LC50 of triethylben-
zene (aquatic organisms) is between 100 and
1000 mg/L. The oral toxicity is low: LDLo
5000 mg/kg (rat). Implant experiments with hex-
aethylbenzene in mice have raised the suspicion
that the compound may be carcinogenic [387].
50 Hydrocarbons
Ethyltoluene. Ethyltoluene can, as with aro-
matic hydrocarbons, be detected in the envi-
ronment [385]. The oral toxicity of 2- and 4-
ethyltoluene is low: LDLo 5000 mg/kg (rat)
[387].
Cumene. Cumene appears to be orally as non-
toxic as propylbenzene, it is even less so by
inhalation [383, 392, 393]. Cumene has a pro-
longed depressant effect on the central nervous
system [394].
The MAK for cumene has been set at
250 mg/m3 , corresponding to a cumene vapor
concentration in air of 50 mL/m3 . The vapor
pressure of cumene at 20 ◦ C is 0.5 kPa. There
is a risk of cumene being absorbed through the
skin [395]. Inhalation of cumene vapor leads to
the delayed appearance of a long lasting narcotic
effect [396].
6.4. Biphenyls and Polyphenyls
Biphenyls. Biphenyl dust or vapor is irritat-
ing to the eye and mucous membrane at a con-
centration as low as 3 – 4 ppm [390], and to
the skin after extended exposure. A concen-
tration of >5 mg/m3 for long periods is con-
sidered a health hazard; systemic toxic effects
were elicited in humans by a concentration
with inhalatory exposure maxima of 128 ppm
[400, 401]. The olfactory threshold is ca. 0.06 –
0.3 mg/m3 . Some toxicity data are listed below
[397–399]:
LD50 (rat, oral) 3280 mg/kg
LD50 (rabbit, oral) 2400 mg/kg
LD50 (rabbit, dermal) 2500 mg/kg
TLm (fathead minnow, 96 h) 1.5 mg/L
Triphenyls are only minimally toxic [402].
Terphenyls. Toxicologically, terphenyl iso-
mers should be treated like biphenyl. The oral
LD50 is ca. 4.6 – 4.7 g/kg [403]. The toxicity of
partially (40 %) hydrogenated terphenyl (LD50
17.5 g/kg) is much less than that of the fully aro-
matic terphenyls [404–407].
6.5. Hydrocarbons from Coal Tar
The aromatic hydrocarbons under consideration
are usually found in complex mixtures of various
other polycyclic aromatic hydrocarbons (PAH),
which originate from combustion emissions and
industrial processes. For this reason toxic phe-
nomena possibly caused by exposure to humans
are often unlikely to be attributed to one sin-
gle aromatic compound. Incorporation of these
compounds into the body occurs predominantly
by way of the respiratory tract. Skin contact is
important with occupational exposure.
6.5.1. Biological Effects
6.5.1.1. Carcinogenicity and Mutagenicity
The aromatic compounds under discussion are
considered as noncarcinogens, but pyrene and
fluoranthene are able to enhance the carcino-
genic potential of benzo [a] pyrene [50-32-8]
when applied simultaneously to mouse skin (co-
carcinogenic effect) (→Carcinogenic Agents)
[408]. Phenanthrene resulted in just one positive
tumor-initiation test in a series of carcinogenic-
ity studies [408, 409]. Acenaphthene, known as
an inhibitor of mitosis in plant cells [410], in-
duced the initial phase of bronchocarcinoma in
rats after an inhalation period (4 months) of 4
hours per day with 0.5 – 1.25 mg/m3 [411], but
was negative in repeated topical application to
mouse skin [412]. With their low activity in in
vivo and in vitro assays, there is limited evi-
dence that phenanthrene and pyrene are muta-
genic [408, 409, 413–415]. As for fluoranthene,
several more recent positive results indicate a
mutagenic potential [408, 416–419].
So far, tests on fluorene have given negative
results [408].
6.5.1.2. Mammalian Toxicity and
Toxicokinetics
The acute toxicity of the aromatic compounds
under consideration is low: LD50 (oral and der-
mal) is 700 – 2700 mg/kg body weight in rodents
[408, 409, 411, 413]; LD50 (rat, oral) for ace-
naphthene is 10 000 mg/kg [421]. The LC50 (in-
halation) of pyrene is reported to be 170 mg/m3
[413]. No fatalities occurred in rats exposed
to indene vapor (800 – 900 ppm) for six in-
halation periods of 7.5 h each; however, sys-
temic pathological changes in the vascular sys-
tem and several organs were observed [412].

Longterm inhalation (4 months) of acenaph-
thene (0.5 –1.25 mg/m3 ) [411, 420] or pyrene
(0.3 – 3.6 mg/m3 ) [408, 409, 413] caused local
irritation and systemic pathological effects in
several organs and the blood of rodents. In direct
contact at high concentration, the compounds
under consideration are irritating to the skin
and mucous membranes; the tri- and polycyclic
aromatics especially are photosensitizing [408–
410,412, 413]. The compounds under consider-
ation are absorbed on inhalation or oral expo-
sure, and partly leave the body unchanged in the
feces and by exhalation [408, 409, 412, 413]. Af-
ter metabolic conversion to more water-soluble
intermediates (hydroxylation and conjugation),
they are excreted into the bile or the urine. The
half-life of pyrene is estimated to be 24 – 48 h in
rats and swine [413].
6.5.1.3. Ecotoxicology
Several studies have shown that the aromatic
compounds, in question are microbially decom-
posed [422–424], although the microbial con-
version of indene has not yet been described.
These compounds tend to be enriched in sedi-
ments, sludges, and aquatic organisms because
of their low water solubility. As a result of micro-
bial degradation, biotransformation, and pho-
tochemical decomposition, however, accumula-
tion is limited and checked. Nonetheless, a con-
centration in water of the order of their low sol-
ubility (<0.2 – 4.0 mg/L) may have adverse ef-
fects on aquatic life, as shown for acenaphthene
[425], phenanthrene, and pyrene [426, 427].
6.5.2. Safety Regulations
A threshold limit value (TLV) at the work
place exists only for indene: 10 ppm (45 mg/m3 )
(ACGIH, United States 1980). No other TLVs
have so far been established. The United States
standard for exposure to coal tar volatiles
(0.2 mg/m3 ) was recommended to be used in
case of pyrene-containing PAH mixtures, but a
maximum level of 0.1 mg/m3 should be applied
in the case of exposure to pure pyrene [413].
With the exception of indene all aromatics are
included in the recommended United States list
of priority pollutants (United States EPA, 1977)
[428].
Hydrocarbons 51
7. References
1. F. D. Rossini, B. J. Mair, Adv. Chem. Ser. 5
(1952) 341.
2. Selected values of properties of hydrocarbons
and related compounds, Am. Petrol. Inst.
Research Project 44, Supplementary vol.
A-72, 1975.
3. G. A. Olah et al., J. Am. Chem. Soc. 91 (1969)
3261.
4. G. A. Olah et al., J. Am. Chem. Soc. 97 (1975)
6807.
5. G. A. Olah et al., J. Am. Chem. Soc. 95 (1973)
7686.
6. G. A. Olah et al., J. Am. Chem. Soc. 95 (1973)
7680.
7. K. Nabert, G. Schön, Sicherheitstechnische
Kennzahlen brennbarer Gase und Dämpfe,
Deutscher Eichverlag, Berlin 1965.
8. R. C. Weast, ed., Handbook of Chemistry and
Physics, 55th ed., CRC Press, Cleveland 1974.
9. C. Zerbe, Mineralöle und Verwandte Produkte,
Springer, Berlin 1969.
10. Deutsche Shell AG, Erdöl-Nachrichten,
March 1976, p. 4.
11. F. Schmerling, H. Heneka, Erdöl Kohle Erdgas
Petrochem. 21 (1968) 705.
12. K. Hedden, J. Weitkamp, Chem. Ing. Tech. 47
(1975) 505.
13. J. W. Ward, Hydrocarbon Process. 54 (1975)
no. 9, 101.
14. R. Oberkobusch, Erdöl Kohle Erdgas
Petrochem. Brennst. Chem. 28 (1975) 558.
15. J. Schulze, Chem. Ing. Tech. 46 (1974) 925,
976.
16. H. Pichler, H. Schulz, Chem. Ing. Tech. 42
(1970) 1162.
17. C. D. Frohning, B. Cornils, Hydrocarbon
Process. 53 (1974) no. 11, 143.
18. J. C. Hoogendorn, Gas Wärme Int. 25 (1976)
283.
19. K. C. Hellwig, S. B. Alpert, E. S. Johanson,
R. H. Wolk, Brennst. Chem. 50 (1969) 263.
20. F. W. Richardson: Oil from Coal, Noyes Data
Corp., Park Ridge, N.J., 1975.
21. K. F. Schlupp, H. Wien, Angew. Chem. 88
(1976) 348.
22. G. Nashan, Erdöl Kohle Erdgas Petrochem.
Brennst. Chem. 28 (1975) 562.
23. G. Collin, Erdöl Kohle Erdgas Petrochem.
Brennst. Chem. 29 (1976) 159.
52 Hydrocarbons
24. H.-W. v. Gratkowski, Erdöl Kohle Erdgas
Petrochem. Brennst. Chem. 28 (1975) 81.
25. H. Hiller, Erdöl Kohle Erdgas Petrochem.
Brennst. Chem. 28 (1975) 74.
26. P. D. Bartlett, K. E. Schueller, J. Am. Chem.
Soc. 90 (1968) 6071.
27. Houben-Weyl, 4th ed., vol. IV, parts 3 and 4.
28. J. H. Lucas, F. Baardman, A. P. Kouwenhoven,
Angew. Chem. 87 (1975) 740, 88 (1976) 412.
29. W. C. van Ziill Langhout, Proc. World Pet.
Congr. 1975, vol. 5, 197.
30. W. L. Lafferty, Jr., W. R. Stokeld, Adv. Chem.
Ser. 103 (1971) no. 7, 130.
31. L. F. Albright, Chem. Eng. 73 (1966) no. 14,
119; 73 (1966) no. 19, 205; 73 (1966) no. 21,
209.
32. L. Alcock, R. Martin, K. H. Bourne, W. A.
Peet, J. Winsor, Oil Gas J. 72 (1974) no. 27,
102.
33. D. P. Thornton, D. J. Ward, R. A. Erickson,
Hydrocarbon Process. 51 (1972) no. 8, 81.
34. H. Jockel, B. E. Triebskorn, Hydrocarbon
Process. 52 (1973) no. 1, 93.
35. Hydrocarbon Process. 52 (1973) no. 11, 145.
36. P. R. Pujado, B. V. Vora, Energy Prog. 4
(1984) no. 3, 186.
37. M. S. Scurrell, Appl. Catal. 32 (1987) 1.
38. P. Leprince, A. Chauvel, L. Castex: Procédés
de Pétrochimie, Editions Technip, Paris 1971,
p. 269.
39. K. Ito, I. Dogane, K. Tanaka, Hydrocarbon
Process. 64 (1985) no. 10, 83.
40. S. C. Arnold, G. D. Sucui, L. Verde, A. Neri,
Hydrocarbon Process. 64 (1985) no. 9, 123.
41. Japan. Chemical Reporter, 1981, October, p. 2.
42. F. Solymosi, I. Tombácz, G. Kutsan, J. Chem.
Soc. Chem. Commun. 1985, 1455.
43. R. V. Porcelli, B. Juran, Hydrocarbon Process.
65 (1986) no. 3, 37.
44. W. Hinsen, M. Baerns, Chem. Zt. 107 (1982)
9.
45. Kh. E. Khcheyan, O. M. Revenko, A. N.
Shatalova: “Oxidative Methylation, a new
route to petrochemicals through alternative
feedstock,” Special Paper No. 16, World
Petroleum Congress 11th, 1983 (London).
46. G. de Pontanel: Proteins from Hydrocarbons,
Academic Press, London 1972.
47. S. Gutcho: Proteins from Hydrocarbons,
Noyes Data Corp., Park Ridge, N.J., 1973.
48. P. J. Rockwell: Single Cell Proteins from
Cellulose and Hydrocarbons, Noyes Data
Corp., Park Ridge, N.J., 1976.
49. G. Voss, Erdöl Kohle Erdgas Petrochem.
Brennst. Chem. 26 (1973) 249.
50. W. Dimmling, R. Seipenbusch, Hydrocarbon
Process. 54 (1975) no. 9, 169.
51. H. Riegel, H. D. Schindler, M. C. Sze, Erdöl
Kohle Erdgas Petrochem. Brennst. Chem. 26
(1973) 704.
52. R. P. Lowry, A. Aguilo, Hydrocarbon Process.
53 (1974) no. 11, 103.
53. B. V. Vora, T. Imai, Hydrocarbon Process. 61
(1982) no. 4, 171.
54. L. B. Phillips, Hydrocarbon Process. 42
(1963) no. 6, 159.
55. A. A. Di Giacomo, J. B. Maerker, J. W. Schall,
Chem. Eng. Prog. 57 (1961) no. 5, 35.
56. Universal Oil Products, US 3 649 177, 1972
(D. H. Rosback).
General References
57. F. Asinger: Die Petrolchemische Industrie,
vol. 1 (Gewinnung der Olefine), vol. 2
(Verwendung der Olefine) Akademie-Verlag,
Berlin 1971.
58. J. Falbe (ed.): Methodium Chimicum, vol. 4,
Thieme Verlag, Stuttgart 1980, pp. 35, 111.
59. Houben-Weyl, 4th ed., vol. 5/1 b, Thieme
Verlag, Stuttgart 1972.
60. S. Patai, J. Zabicky: The chemistry of alkenes,
vol. 1 (1964), vol. 2 (1970), Interscience
Publishers, New York.
61. E. S. Strauss, G. Hollis, O. Kamatari: Chemical
Economics Handbook Marketing Research
Report “Linear alpha-olefins,” Stanford
Research Institute, Menlo Park, CA, 1984.
62. B. Fell, Tenside Deterg. 12 (1975) 3.
63. H. Isa, CEER Chem. Econ. Eng. Rev., 9
(1977) 26.
64. A. H. Turner, J. Am. Oil Chem. Soc. 60 (1983)
623.
65. A. Behr, W. Keim, Arabian J. Sci. Eng. 10
(1985) 377.
Specific References
66. K. G. Mittal, J. Chem. Technol. Biotechnol. 36
(1986) 291 – 299.
67. D. B. Broughton, R. C. Berg, Chem. Eng. 77
(1970) 86.
68. B. V. Vora, P. R. Pujado, J. B. Spinner, T. Imai,
Proc. World Tenside Congr. 1984 (Munich)
vol. II, 16 – 26.
69. R. C. Berg, B. V. Vora in McKetta,
Cunningham (ed.): Encyclopedia of Chemical
Processing and Design, vol. 15, Marcel
Dekker, New York 1982, 266 – 284.
70. H. D. Wulff, Fette Seifen Anstrichm. 69 (1967)
32.

71. R. Stöbele, Chem. Ing. Tech. 36 (1964) 858.
72. K. Ziegler et al., Angew. Chem. 64 (1952) 323,
67 (1955) 542.
73. Gulf Res. Dev. Co., DE 1 443 927, 1961.
74. Ethyl Corp., US 3 906 053, 1973.
75. W. Keim, A. Behr, M. Röper in G. Wilkinson
(ed.): Comprehensive Organometallic
Chemistry, vol. 5, Pergamon Press, Oxford
1982, 371.
76. W. Keim, Chem. Ing. Tech. 56 (1984) 850.
77. Shell Development Comp., US 3 635 937,
1972; 3 637 636, 1972; 3 644 564, 1972;
3 644 563, 1972; 3 647 914, 1972; 3 647 915,
1972; 3 661 803, 1972; 3 686 159, 1972 (W.
Keim et al.).
78. E. R. Freitas, R. C. Gum, Chem. Eng. Progress.
75 (1979) no. 1, 73.
79. E. L. T. M. Spitzer, Seifen öle Fette Wachse
107 (1981) 141 – 144.
80. M. Sherwood, Chem. Ind. (London) (1982)
994 – 995.
81. W. Keim, F. H. Kowaldt, Erdöl Kohle Erdgas
Petrochem. Compendium (1978/79)
453 – 462.
82. Shell Oil Comp., US 3 647 906, 1972 (F. F.
Farley).
83. M. Peuckert, W. Keim, Organometallics 2
(1983) 594.
84. W. Keim, F. H. Kowaldt, R. Goddard, C.
Krüger, Angew. Chem. 90 (1978) 493.
85. W. Keim, A. Behr, B. Limbäcker, C. Krüger,
Angew. Chem. 95 (1983) 505.
86. W. Keim et al., J. Mol. Catal. 6 (1979) 79.
87. W. Keim et al., Organometallics 5 (1986)
2356.
88. Chem. Eng. News (1972), April 10, 16 – 17.
89. C. Commereuc, Y. Chauvin, J. Gaillard, J.
Léonard, J. Andrews, Hydrocarbon Process.
63 (1984) no. 11, 118 – 120.
90. W. J. Benedek, J.-L. Mauleon, Hydrocarbon
Process. 59 (1980) no. 5, 143.
91. J. Léonard, J. F. Gaillard, Hydrocarbon
Process. 60 (1981) no. 3, 99.
92. J. F. Boucher, G. Follain, D. Fulop, J. Gaillard,
Oil Gas J. 80 (1982) 84 – 86.
93. Y. Chauvin, J. Gaillard, J. Léonard, P.
Bonnifay, J. W. Andrews, Hydrocarbon
Process. 61 (1982) no. 5, 110 – 112.
94. D. Commereux, Y. Chauvin, G. Leger, J.
Gaillard, Rev. Inst. Fr. Pet. 37 (1982)
639 – 649.
95. J. Gaillard, Pet. Tech. 314 (1985) 20 – 27.
96. B. Bogdanović, Adv. Organomet. Chem. 17
(1979) 105.
Hydrocarbons 53
97. W. Keim, A. Behr, G. Kraus, J. Organomet.
Chem. 251 (1983) 377.
98. A. Behr, V. Falbe, U. Freudenberg, W. Keim,
Isr. J. Chem. 27 (1986) 277 – 279.
99. W. Keim, Ann. N.Y. Acad. Sci. 415 (1983)
191 – 200.
100. B. Bogdanović, B. Spliethoff, G. Wilke,
Angew. Chem. Int. Ed. Engl. 19 (1980) 622.
101. G. Speier, J. Organomet. Chem. 97 (1975)
109.
102. M. G. Barlow, M. J. Bryant, R. N. Haszeldine,
A. G. Mackie, J. Organomet. Chem. 21 (1970)
215.
103. S. Datta, M. B. Fisher, S. S. Wreford, J.
Organomet. Chem. 188 (1980) 353.
104. R. Schrock, S. McLain, J. Sancho, Pure Appl.
Chem. 52 (1980) 729.
105. J. A. Brennan, Ind. Eng. Chem. Prod. Res. Dev.
19 (1980) 2 – 6.
106. R. L. Shubkin, M. S. Baylerlan, A. R. Maler,
Ind. Eng. Chem. Prod. Res. Dev. 19 (1980)
15 – 19.
107. F. J. Karol, CHEMTECH 1983 (April) 222.
108. B. W. Werdelmann, Fette Seifen Anstrichm. 76
(1974) 1.
109. Chem. Eng. News, (1988) May 30, 10.
110. Eur. Chem. News, (1987) Dec. 21/28, 21.
111. Eur. Chem. News, (1981) July 6, 23.
112. Eur. Chem. News, (1987) Jan. 5/12, 19.
113. Eur. Chem. News, (1987) Jan. 19, 8.
114. Eur. Chem. News, (1981) Jan. 26, 27.
115. Eur. Chem. News, (1982) July 5, 26.
116. G. Wilke, Angew. Chem. Int. Ed. Engl. 2
(1963) 105.
117. A. C. L. Su, Adv. Organomet. Chem. 17
(1979) 269.
118. S. Warwel, Erdöl Erdgas Kohle 103 (1987)
238.
119. R. C. Weast, M. J. Astle (eds.): CRC Handbook
of Data on Organic Compounds, CRC Press,
Boca Raton, Florida, 1985.
120. S. Coffey (ed.): Rodd’s Chemistry of Carbon
Compounds, III A, Elsevier, Amsterdam 1971,
pp. 228 – 231.
121. A. D. Rudkovskii et al., Chem. Technol. Fuels
Oils (Engl. Transl.) 19 (1983) 387 – 391.
122. F. P. McCandless, Ind. Eng. Chem. Prod. Res.
Dev. 19 (1980) 612 – 616; 21 (1982)
483 – 488.
123. Idemitsu Kosan Co. Ltd., EP 0 104 607, 1983
(H. Sato).
124. Universal Oil Products, US 4 487 984, 1984 (T.
Imai).
125. O. A. Anunziata et al., Appl. Catal. 15 (1985)
235 – 245.
54 Hydrocarbons
126. J. Hanotier, M. Hanotier-Bridoux, J. Mol.
Catal. 12 (1981) 133 – 147.
127. L. I. Smith, Org. Synth., coll. vol. 2 (1948)
248 – 253.
128. Toyo Soda Manufacturing, DE-OS 3 330 912,
1983 (T. Sakamoto, T. Hironaka, K. Sekizawa,
Y. Tsutsumi).
129. Toa Nenryo, EP 0 101 651, 1983 (T. Maejima,
N. Tagaya, S. Sakurada).
130. J. Haggin, Chem. Eng. News, (1987) June 1,
22 – 28; (1987) June 22, 22 – 25.
131. International Flavors and Fragrances,
US 3 869 524, 1974 (K. K. Light, J. B.
Anglim).
132. Phillips Petroleum, US 4 568 784, 1985 (R. L.
Cobb).
133. S. A. Shama, C. C. Wamser, Org. Synth. 61
(1983) 62 – 64.
134. H. F. Davis, S. K. Chattopadhyay, P. K. Das, J.
Phys. Chem. 88 (1984) 2798–2803.
135. W. W. Kaeding, G. C. Barile, M. M. Wu, Catal.
Rev. Sci. Eng. 26 (1984) 597 – 612.
136. Dow Chemical, US 2 763 702, 1956 (J. L.
Amos, K. E. Coulter).
137. W. W. Kaeding, L. B. Young, C.-C. Chu, J.
Catal. 89 (1984) 267 – 273.
138. W. W. Kaeding, G. C. Barile, Polym. Sci.
Technol. (Plenum) 25 (1984) 223 – 241.
139. Eur. Chem. News, Oct. 17 (1988) p. 4.
140. Physical Constants of Hydrocarbons C1 to
C10 , “ASTM Spec. Techn. Publ. STP” 106
(1963).
141. Handbook of Chemistry and Physics, R. C.
Weast (ed.), 63rd ed., CRC Press, Boca Raton
1982 – 1983, p. D-213.
142. D. R. Stull, E. F. Westrum, G. C. Sinke: The
Chemical Thermodynamics of Organic
Compounds, J. Wiley & Sons, New York
1969, p. 370.
143. J. C. Butler, W. P. Webb, Chem. Eng. Data
Series 2 (1957) no. 1, 42 – 46.
144. G. Sorbe: Sicherheitstechnische Kenndaten
chemischer Stoffe, vol. 2,
ecomed-Verlagsgesellschaft mbH, Landsberg
1983, pp. 15 – 16.
145. P. R. Pujado, J. R. Salazar, C. V. Berger,
Hydrocarbon Process. 55 (1976) no. 3, 91.
146. R. C. Canfield, T. C. Unruh, Chem. Eng. NY
90 (1983) no. 3, 32.
147. Toyo Soda Manufacturing Co.,
DE-AS 2 418 038, 1974 (T. Yanagihera et al.).
148. Hitachi Chemical Co., JP-Kokai 77 139 030,
1974 (E. Kudo et al.); Chem. Abstr. 88 (1978)
120757 r.
149. BASF, DE-OS 2 141 491, 1971 (F. Neumayr).
150. Mitsui Petrochemical Ind., EP 0 046 678, 1981
(M. Hisaya).
151. J. P. Fortuin, M. J. Waale, R. P. Oosten, Pet.
Refiner 38 (1959) no. 6, 189.
152. A. H. Olzinger, Chem. Eng. NY 82 (1975)
no. 6, 50.
153. Universal Oil Products, US 3 763 259, 1973
(G. L. Hervert).
154. Mobil Oil Corp., EP-A 0 149 508, 1985 (W. W.
Kaeding).
155. J. Ewers, H. W. Voges, G. Maleck, Erdöl Kohle
Erdgas Petrochem. Compendium 74/75 (1975)
487.
156. K. Ito, I. Dogane, K. Tanaka, Hydrocarbon
Process. 65 (1986) no. 11, 89.
157. H. A. Colvin, J. Muse, CHEMTECH 1986,
no. 500.
158. K. Ito, Hydrocarbon Process. 52 (1973) no. 8,
89.
159. Phillips Petroleum Co., US 3 291 847, 1966
(P. F. Warner); US 3 316 315, 1967 (W. A.
Jones).
160. K. Weissermel, H.-J. Arpe: Industrielle
Organische Chemie, 3rd ed., VCH
Verlagsgesellschaft, Weinheim 1987,
pp. 72 – 79.
161. A. C. Olson, Ind. Eng. Chem. 52 (1960)
no. 10, 833.
162. R. Ströbele, Chem. Ing. Tech. 36 (1964) 858.
163. H.-D. Wulf, T. Böhm-Gößl, L. Rohrschneider,
Fette Seifen Anstrichm. 69 (1967) 32.
164. California Research Corp., US 2 477 382, 1949
(A. H. Lewis, A. C. Ettling); Chem. Abstr. 44
(1950) 1137 e.
165. H. Hartig, Chem. Ztg. 99 (1975) 175.
166. K. Ichikawa, H. Shingu, Kogyo Kagaku Zasshi
54 (1951) 183 – 184; Chem. Abstr. 47 (1953)
8662 b.
167. S. Murohashi, Nagaoka Kogyo Koto Senmon
Gakko Kenkyu Kiyo 9 (1973) 103 – 106;
Chem. Abstr. 80 (1974) 132917 r.
168. J. G. Matveev, D. A. Drapkina, R. L. Globus,
Tr. IREA 21 (1956) 83 – 89, Chem. Abstr. 52
(1958) 15474 e.
169. A. D. Harford, H. W. Vernon, US 2 897 112,
1956.
170. British Petroleum Co., Ltd., DE-AS 1 094 035,
1957 (A. Douglas, H. W. Vernon).
171. N. K. Moshinskaya, V. S. Olifer, L. I. Zhuoiev,
SU 137 993, 1960.
172. E. Siggel, US 2 710 849, 1952.
173. BASF, DAS 1 142 338, 1960 (W. Braun, E.
Hartwig, R. Krallmann).
174. FMC Corp., US 3 539 640, 1967 (M. I.
Stewart, O. K. Carlson).

175. Ashland Oil & Refining Co., US 3 529 944,
1967 (A. M. Leas).
176. Japan Petrochemicals Co., Ltd.,
DE-OS 2 210 133, 1972 (A. Sato, Y. Aida, I.
Shimizi).
177. H. Winkler: Der Steinkohlenteer, Verlag
Glückauf, Essen 1951, p. 143.
178. J. Chipman, S. B. Peltier, Ind. Eng. Chem. 21
(1929), 1106.
179. M. McEwen: Organic Coolant Handbook,
sponsored by U.S. Atomic Energy Comm.,
publ. by Monsanto Co., St. Louis, Mo., 1958.
180. J. M. Cork, H. Mandel, N. Ewbank (North
American Aviation Inc.), U.S. At. Energy
Comm. Sci. Rept. 5129 (1960) 1 – 28.
Recalculated by A. N. Syverud, The Dow
Chemical Co. Thermal Laboratory, Midland,
Mich.
181. M. McEwen, E. Wiederhold, Nucl. Eng. Part
IV Symp. Ser. 5 (1959) no. 22, 9 – 15.
182. M. J. Foole, Chem. Eng. Prog. 58 (1962) no. 4,
37 – 40.
183. Hydrocarbon Process. 46 (1967) no. 11, 187.
184. S. Feigelman, C. B. O’Connor, Hydrocarbon
Process. 45 (1966) no. 5, 140 – 144.
185. Hydrocarbon Process. Pet. Refin. 42 (1963)
no. 11, 204.
186. Bayer, BIOS Final Rep. Nr. 1787.
187. R. Egbert, Trans. AICHE vol. 34 (1938)
435 – 488.
188. R. Dasgupta, B. R. Maiti, Ind. Eng. Chem. 25
(1986) 381 – 386.
189. Otkrytiya/Itobret, SU 1 165 677, 1985.
190. BIOS Final Rep. Nr. 893.
191. Soc. Progil, US 3 227 525, 1964 (M. E.
Degeorges, M. Jaymond).
192. Soc. Progil, US 3 112 349, 1961 (M. E.
Degeorges, M. Jaymond).
193. Soc. Progil, DE-AS 1 249 244, 1961 (M.
Jaymond).
194. Dow Chemical Co., US 1 938 609, 1933 (J. H.
Reilly).
195. Monsanto Co., US 3 228 994, 1962 (W. D.
Robinson, W. Groves).
196. Hawker Siddeley Nuclear Power Co.,
FR 1 272 068, 1961.
197. Montecatini, BE 586 826, GB 865 302, IT
Anm. 1144/59, 1960, IT Prior. 1959 (C. Ferry,
B. Ponzini).
198. Monsanto Co., US 2 143 509, 1939 (C.
Conover, A. E. Huff).
199. Montecatini, IT 586 113, 1962.
200. Kambaror YU G; SU 6 79 565, 1977, Chem.
Abstr. 92 (1980) 58407 d.
Hydrocarbons 55
201. Monsanto Co., US 3 228 994, 1962 (W. D.
Robinson).
202. N. M. Indyukov, R. I. Gasanova, Khim.
prom-st. (Moscow) 1975, no. 7, 555.
203. Farbenfabr. Bayer AG, DE-AS 1 012 595, 1955
(H. Holzrichter, O. Tegtmeyer).
204. F. Ullmann, Justus Liebigs Ann. Chem. 332
(1904) 48.
205. E. I. El’bert, SU 215 911, 1968.
206. R. G. Ismailow, D. A. Aliev, Azerb. Neft. Khoz.
44 (1965) no. 1, 30.
207. L. Gattermann, R. Ehrhardt, Ber. Dtsch. Chem.
Ges. 23 (1890) 1226.
208. I. H. Gardner, P. Bergstrom, J. Am. Chem. Soc.
51 (1962) 3376.
209. Asahi Kasei Kogyo K. K., DE-AS 2 039 904,
1970.
210. Shell Oil Co., US 3 145 237, 1962 (R. van
Helden, B. Balder).
211. R. van Helden, G. Verberg, Rec. Trav. Chim.
Pays-Bas 84 (1965) 1263.
212. Gulf Research & Developm. Co.,
US 3 401 207, 1966 (C. M. Selwitz).
213. E. S. Rudakov, A. I. Lutsyk, R. I. Rudakova,
Kinet. Katal. 18 (1977) 525; Kinet. Catal.
(Engl. Transl.) 18 (1977) 441.
214. M. K. Starchevskii, M. N. Vargaftik, I. I.
Moiseev, Izv. Akad. Nauk SSSR 1979,
242;(Engl. Transl.) 227.
215. Teijin Ltd., US 3 963 787, 1973 (Y. Ichikawa,
T. Yamaji).
216. Ube Industries, JP 55 079 324, 1978.
217. L. N. Rachkovskaya et al., Kinet. Katal. 18
(1977) 792; Kinet. Catal. (Engl. Transl.) 18
(1977) 660.
218. Ube Industries, JP 80 141 417, 1979, Chem.
Abstr. 94 (1981) 156 561 k.
219. British Petroleum, DE-OS 1 909 045, 1969 (D.
Fiolka, C. R. Jentsch).
220. L. H. Slaugh, Tetrahedron 24 (1968) 4523.
221. L. H. Slaugh, J. A. Leonard, J. Catal. 13
(1969) 385.
222. G. N. Koshel et al., Dokl. Akad. Nauk SSSR
237 (1977) no. 1, 164 (Chem. Tech.); Chem.
Abstr. 88 (1978) 50 394 u.
223. Phillips Petroleum, US 4 123 470, 1977 (T. P.
Murths).
224. Farberov M I, SU 416 339, 1972; Chem. Abstr.
88 (1978) 62 124 u.
225. Standard Oil, US 4 218 572, 1978 (S. R.
Dolhyj, L. J. Velenyi).
226. Y. Yamazaki, T. Kawai, A. Yagishita, Sekiyu
Gakkaishi 21 (1978) no. 4, 229.
227. Koppers Co., US 3 059 987 1962 (H. P.
Baumann).
56 Hydrocarbons
228. N. Hendrich, Melliand Textilber. 43 (1962)
258.
229. W. M. Meylan, P. H. Howard: Chemical
Market Input/Output Analysis of Biphenyl and
Diphenyl Oxide to Assess Sources of
Environmental Contamination, Syracuse NY,
EPA Contract No. 68-01-3224, 1976.
230. W. E. Bachmann, H. T. Carke, J. Am. Chem.
Soc. 49 (1927) 2089 – 98.
231. L. W. Pickett, Proc. R. Soc. London A 142
(1933) 333 – 346.
232. H. G. Römer, E. Hertel, Z. Phys. Chem. Abt. B
21 (1933) 292 – 296.
233. R. J. Good, E. E. Hardy, A. M. Ellenburg, H. B.
Richards, Jr., J. Am. Chem. Soc. 75 (1953)
436.
234. J. A. Ellard, W. H. Yanko: Thermodynamic
Properties of Biphenyl and the Isomeric
Terphenyls, Final Report IDO-11,008,
Monsanto Research Corp., U.S. At. Energy
Comm. Contract AT (10-1)-1088,Oct. 31,
1963.
235. R. O. Bolt, R. O. Johnson, R. S. Kirk: Organic
Materials for Use as Reactor Coolant
Moderators, California Research Corp.,
June 13, 1956.
236. P. L. Geiringer: Handbook of Heat Transfer
Media, Reinhold Publ., New York 1962,
p. 173.
237. W. H. Hedley, M. V. Mines, J. Chem. Eng.
Data 15 (1970) 122.
238. H. Mandel, N. Eubank: Critical Constants of
Diphenyl and Terphenyl, Report No. NAA-SR
5129, Atomics International, Dec. 1960, p. 18.
239. N. Andrews, R. Ubbelohde, Proc. R. Soc.
London 228 A (1955) 435 – 37.
240. E. Pajda, J. Rusin, Chemik Wkladka 1972, 422.
241. R. Balent, Proc. Am. Power Conf. 21 (1959)
120.
242. Soc. Progil, DE-AS 1 272 281, 1962 (M.
Jaymond).
243. Farbenfabr. Bayer AG,
GB 957 065 = FR 1 297 408, 1961.
244. Hawker Siddeley Nuclear Power Co.,
FR 1 272 068, 1961.
245. D. B. Orechkin, SU 570 584, 1977.
246. Monsanto Co., DE-OS 2 306 454, 1973 (J. K.
Sears).
247. M. He, Huaxue Shiji 8 (1986) no. 1, 60 – 61;
Chem. Abstr. 105 (1986) 208 532.
General References
248. H.-G. Franck, G. Collin: Steinkohlenteer,
Springer Verlag, Berlin – Heidelberg – New
York 1968.
249. H.-G. Franck, J. W. Stadelhofer: Industrielle
Aromatenchemie, Springer Verlag,
Berlin – Heidelberg – New
York – London – Paris – Tokyo 1987.
250. Beilstein, 5, 586; 5 (1), 274; 5 (2), 494; 5 (3),
1776; 5 (4), 1834.
251. Beilstein, 5, 515; 5 (1), 248; 5 (2), 410; 5 (3),
1355; 5 (4), 1532.
252. Beilstein, 5, 625; 5 (1), 300; 5 (2), 531; 5 (3),
1936; 5 (4), 2142.
253. Beilstein, 5, 685; 5 (1), 344; 5 (2), 609; 5 (3),
2276; 5 (4), 2463.
254. Beilstein, 5, 667; 5 (1), 327; 5 (2), 579; 5 (3),
2136; 5 (4), 2293.
255. Beilstein, 5, 693; 5 (1), 343; 5 (2), 609; 5 (3),
2279; 5 (4), 2467.
Specific References
256. Rütgerswerke, DE 3 203 818, 1982 (K.
Stolzenberg, K. H. Koch, R. Marrett).
257. Nippon Steel, JP 82 98 219, 1980.
258. IG Farbenindustrie, DE 428 088, 1921.
259. Selden, GB 318 617, 1929.
260. Standard Oil, US 2 966 513, 1957 (E. K.
Fields).
261. Rütgerswerke, DE 1 262 268, 1959 (E. Marx,
H.-G. Franck, E. A. Dierk, K. Fiebach).
262. Ullmann, 4th ed., 18, 685.
263. Farbwerke Hoechst, DE 2 209 692, 1972 (K.
Eiglmeier); DE 2 343 964, 1973 (F.
Röhrscheid).
264. Institute of Chemical Sciences, Academy of
Sciences/Kazakh S.S.R., SU 1 084 275, 1982
(R. U. Umarova, A. K. Ergalieva, D. K.
Sembaev, B. V. Suborov)
265. BASF, DE 2 721 640, 1977 (A. Hettche, M.
Patsch); DE 2 653 346, 1976 (W. Eisfeld, W.
Disteldorf, A. Hettche).
266. BASF, DE 2 721 639, 1977 (A. Hettche, M.
Patsch).
267. Ullmann, 4th ed., 18, 693.
268. Farbwerke Hoechst, DE 2 304 873, 1973 (K.
Eiglmeier, H. Lübbers).
269. Toyo Soda Mfg., JP 85 94 928, JP 85 139 630,
1983 (M. Kubo, H. Satsuka, Y. Tsutsumi);
EP 217 316, 1986 (M. Kubo, K. Kawabata, Y.
Tsutsumi).
270. Coal Industry, EP 47 069, 1981 (J. G.
Robinson, S. A. Brain).
271. Nippon Steel, JP 83 176 210, 1982.
272. Institute of Heteroorganic Compounds,
Academy of Sciences/USSR, DE 3 526 155,
1985 (V. V. Korshak et al.).

273. Sun Research and Development, DE 1593 398,
1966 (A. Schneider, E. J. Janoski, R. N.
McGinnis).
274. G. Collin, M. Zander, Erdöl Kohle Erdgas
Petrochem. 33 (1980) 560.
275. BASF, DE 3 044 111, 1980 (H. Naarmann, K.
Penzien, D. Nägele, J. Schlag).
276. Ullmann, 4th ed., 21, 70.
277. R. Heering, P. Lobitz, K. Gehrke, Plaste
Kautsch. 31 (1984) 412 – 413; J. Stein, G.
Liebisch, S. Helling, Chem. Tech. (Leipzig) 31
(1979) 302 – 305.
278. Gulf Research and Development,
US 4 280 881, 1977 (J. C. Montagna, R. D.
Galli).
279. Dow Chemical, US 4 093 672, 1977 (Y. Ch.
Sun).
280. Rütgerswerke, DE 2 443 114, 1974 (K. Rühl,
G. Storch, H. Steinke, B. Charpey).
281. American Cyanamid, US 4 145 409, 1976
(N. R. Pasarela).
282. Manufacture de Produits Pharmaceutique A.
Christiaens, DE 2 060 721, 1970 (P. Vanhoff, P.
Clarebout).
283. Takeda Chemical Industries, JP 82 123 157,
1981.
284. Rütgerswerke, DE 2 208 253, 1972 (E.
Pastorek, H. Miels, W. Orth).
285. Pfizer, DE 2 244 556, 1973 (D. P. O’Shea, R. C.
Adams, S. Nakanishi).
286. Meiji Seika Kaisha, DE 2 911 492, 1979 (Y.
Sekizawa et al).
287. Wella, DE 3 215 341, 1982 (E. Konrad, H.
Mager).
288. Rütgerswerke, DE 2 635 401, 1976 (P.
Stäglich, D. Hausigk).
289. H. J. V. Winkler: Der Steinkohlenteer und
seine Bedeutung, Verlag Glückauf, Essen
1951, p. 156.
290. Nippon Steel, JP 82 98 220, 1980.
291. Rütgerswerke, DE 3 243 981, 1982 (W. Orth,
E. Pastorek, W. Fickert).
292. Celamerck, DE 2 816 388, 1979 (J. Gante et
al.).
293. Squibb, US 3 856 977, 1974 (E. T. Stiller, S. D.
Levine, P. A. Principe, P. A. Diassi).
294. Bayer, DE 3 530 819, 1985 (J. Hocker, L.
Rottmeier).
295. Ullmann, 4th ed., 23, 415.
296. Nippon Kayaku, DE 2 325 302, 1973 (D.
Matsunaga, M. Sumitani).
297. Chemische Fabrik Kalk, DE 2 725 857, 1977
(H. Jenkner).
298. Sandoz, DE 2 225 565, 1972 (H. P. Baumann).
299. Rütgerswerke, DE 2 704 648, 1977 (C. Finger).
Hydrocarbons 57
300. Agency of Industrial Sciences and Technology,
JP 81 82 430, 1979.
301. Magyar Tudomanyos Akademia, HU 36 764,
1983 (L. Simandi et al.).
302. Nippon Shokubai Kagaku Kogyo,
JP 85 233 028, 1984 (T. Yoneda, S. Nakahara).
303. Dow Chemical, US 4 297 514, 1980 (W. M.
King).
304. Hydrocarbon Research, DE 3 141 645, 1981.
(D. Huibers, R. Barclay, R. C. Shah).
305. Fuji-Xerox, JP 82 144 567, 1981; Ricoh,
JP 79 109 438, 1978 (E. Akutsu, A. Kojima);
Oce-van der Grinten, DE 2 525 381, 1975 (J.
Hagendorn, H. Knibbe); EP 24 762, 1981
(M. T. J. Peters, J. A. Verbunt).
306. Bayer, DE 2 430 476, 1974 (W. Draber, K. H.
Büchel, M. Plempel).
307. Röhm, DE 3 439 484, 1984 (A. Riemann, W.
Ude).
308. Isovolta, EP 41 496, 1981 (W. Rieder, M.
Fehrle).
309. Minnesota Mining and Mfg., EP 203 828, 1986
(W. J. Schultz, G. B. Portelli, J. P. Tane).
310. BASF, DE 3 209 426, 1982 (M. Patsch, H.
Flohr).
311. BASF, DE 3 536 259, 1985 (M. Patsch et al.).
312. Rütgerswerke, DE 2 114 219, 1971 (C. Finger,
H.-G. Franck, M. Zander).
313. Rütgerswerke, DE 2 166 862, 1971 (C. Finger,
H.-G. Franck, M. Zander).
314. Kureha Chemical Industry, JP 75 140 535,
1974 (M. Takahashi, A. Ito, A. Nobuo).
315. Kureha Chemical Industry, DE 2 447 894,
1973 (Y. Shisakaiso, A. Ito, M. Takahashi).
316. Kureha Chemical Industry, DE 2 446 591,
1974 (M. Takahashi et al.).
317. Rütgerswerke, DE 1 931 946, 1969 (C. Finger,
H.-G. Franck, M. Zander).
318. Wellcome Foundation, EP 125 702, 1984
(K. W. Bair).
319. Richardson Merell, US 3 907 791, 1975 (W. L.
Albrecht, R. W. Fleming).
320. Rütgerswerke, DE 1 126 372, 1958 (J.
Altpeter, H. Sauer).
321. BASF, DE 3 218 229, 1982 (H. Naarmann, J.
Schlag, G. Koehler).
322. VEB Kombinat Mikroelektronik,
DE 3 131 625, 1981 (H.-J. Deutscher et al.).
323. Fuji Standard Research, EP 196 169, 1986 (S.
Otani, Y. Nagai).
324. Coal Industry, EP 36 710, 1981 (J. G.
Robinson, D. I. Barnes).
325. Hercules, US 3 926 897, 1972 (C. L. Cessna).
326. Licentia, DE 2 406 540, 1974 (M. Saure, H.
Wagner, J. Wartusch).
58 Hydrocarbons
327. Union Carbide, US 3 834 166, 1974 (R. A.
Cupper, G. S. Somckh).
328. Bayer, DE 3 232 063, 1982 (H. Herzog, W.
Hohmann).
329. Burroughs Wellcome, US 4 511 582, 1983
(K. W. Bair).
330. Rütgerswerke, DE 1 240 065, 1965 (E. Egerer,
W. Fickert, L. Rappen, G. Grigoleit);
DE 3 305 528, 1983 (D. Tonnius, W. Orth, H.
Miele, E. Pastorek).
331. US Steel, US 2 956 065, 1958 (C. W. De Walt,
K. A. Schowalter).
332. Mitsubishi Metal, JP 60 238 432, 1984 (H.
Yoshida, M. Morikawa, T. Iwamura).
333. Nihon Joryu Kogyo, Nippon Shokubai Kagaku
Kogyo, JP 85 255 747, 1984 (Y. Saito et al.).
334. Nippon Shokubai Kagaku Kogyo,
JP 85 233 029, 1984 (I. Maeda, H. Mitsui, T.
Nakamura).
335. Nippon Shokubai Kagaku Kogyo,
JP 8 112 335, 1979.
336. Merck, US 3 506 434, 1970 (E. Jacobi et al.).
337. Ullmann, 4th ed. 12, 614.
338. Agfa-Gevaert, EP 2 546, 1979 (F. C.
Heugebaert, W. Janssens).
339. Fuji Photo Film, GB 2 009 955, 1979 (M.
Kitajima, H. Tachikawa, F. Shinozaki, T.
Ikeda).
340. PPG Industries, US 4 054 683, 1976 (G. W.
Gruber).
341. National Starch and Chemical, US 4 069 123,
1972 (M. M. Skoultchi, I. J. Davis).
342. Canadian Industries, US 4 012 280, 1976
(H. H. Holton).
343. Ciba-Geigy, CH 580 673, 1972 (W. Müller).
344. Eli Lilly, DE 2 631 888, 1977 (J. B. Campbell).
345. EGIS, DE 3 618 715, 1986 (P. Benko et al.).
346. H. J. V. Winkler: Der Steinkohlenteer und
seine Bedeutung, Verlag Glückauf, Essen
1951, p. 220.
347. Rütgerswerke, DE 2 914 670, 1979 (G.-P.
Blümer, M. Zander).
348. Hoechst, DE 3 532 882, 1985 (S. Schießler, E.
Spietschka).
349. Rütgerswerke, DE 2 709 015, 1977 (J. Omran,
M. Zander).
350. Hoechst, DE 3 134 131, 1981; DE 3 324 881,
1983 (S. Schießler, E. Spietschka).
351. Ullmann, 4th ed., 17, 471 – 472.
352. Bayer, EP 20 817, 1979 (H. Harnisch).
353. Ricoh, JP 7 908 527, 1977 (A. Kojima, M.
Hashimoto, M. Ohta).
354. Hoechst, DE 2 734 288, 1977 (W.
Wiedemann); DE 2 726 116, 1977 (H. J.
Schlosser, E. Lind).
355. British Petroleum, GB 1 399 814, 1971 (A. J.
Burn, T. A. Lighting).
356. Wellcome Foundation, EP 125 701, 1984
(K. W. Bair).
357. Pentel, JP 8 632 352, 1976 (K. Takemura).
358. W. R. von Oettingen: “Toxicity and Potential
Dangers of Aliphatic and Aromatic
Hydrocarbons,” Public Health Bull. (1940)
no. 255.
359. E. A. Wahrenbrock, E. I. Eger, R. G. Laravuso,
G. Maruschak, Anesthesiology 40 (1974)
no. 1, 19.
360. A. H. Nuckolls, Underwriters Laboratory
Report (Nov. 13, 1933) no. 2375.
361. Y. Henderson, H. W. Haggard: Noxious Gases,
2nd ed., Reinhold Publ. Co., New York 1943.
362. J. C. Krantz, Jr., C. J. Carr, J. F. Vitcha, J.
Pharmacol. Exp. Therap. 94 (1948) 315.
363. T. Ikoma, Nichidai Igaku Zasshi 31 (1972)
no. 2, 71.
364. D. M. Aviado, S. Zakhari, T. Watanabe:
Non-fluorinated Propellants and Solvents for
Aerosols, CRC Press, Cleveland, Ohio 1977,
pp. 49 – 81.
365. F. A. Patty, W. P. Yant: “Odor Intensity and
Symptoms Produced by Commercial Propane,
Butane, Pentane, Hexane and Heptane Vapor,”
U.S. Bur. Mines, Rep. Invest. 1929 no. 2979.
366. B. B. Shugaev, Arch. Environ. Health 18
(1969) 878.
367. Documentation of the Threshold Limit Values
for Substances in Workroom Air, 3rd ed.,
American Conference of Governmental
Industrial Hygiene, Cincinnati 1971.
368. M. G. Rumsby, J. B. Finean, J. Neurochem. 13
(1966) 1513.
369. D. A. Haydon, B. M. Hendry, S. R. Levinson,
J. Requena, Biochim. Biophys. Acta 470
(1977) 17.
370. N. Battistini et al., Riv. Patol. Nerv. Ment. 95
(1974) 871.
371. G. W. Paulson, G. W. Waylonis, Arch. Intern.
Med. 136 (1976) 880.
372. N. W. Lazarew, Arch. Exp. Pathol. Pharmakol.
143 (1929) 223.
373. E. T. Kimura, D. M. Elery, P. W. Dodge,
Toxicol. Appl. Pharmacol. 19 (1971) 699.
374. C. H. Hine, H. H. Zuidema, Ind. Med. 39
(1970) no. 5, 215.
375. H. H. Schaumburg, P. S. Spencer, Brain 99
(1976) 182.
376. K. Kurita, Jpn. J. Ind. Health Exp. 9 (1974)
672.
377. A. K. Ashbury, S. L. Nielsen, R. Telfer, J.
Neuropathol. Exp. Neurol. 33 (1974) 191.

378. E. G. Gonzales, J. A. Downey, Arch. Phys.
Med. Rehabil. 53 (1972) 333.
379. H. E. Swann, Jr., B. K. Kwon, G. H. Hogan,
W. M. Snellings, Am. Ind. Hyg. Assoc. J. 35
(1974) no. 9, 311.
380. W. R. F. Notten, P. W. Henderson, Biochem.
Pharm. 24 (1975) 1093.
381. C. F. Reinhardt et al., Arch. Environ. Health
22 (1971) 265.
382. A. Y. H. Lu, H. W. Strobel, M. J. Coon, Mol.
Pharamcol. 6 (1970) 213.
383. N. W. Lazarew, Arch. Exp. Pathol. Pharmakol.
143 (1929) 223.
384. Criteria for a Recommended Standard –
Occupational Exposure to Xylene, U.S. Dept.
Health, Education and Welfare, Natl. Inst.
Occup. Safety and Health, Washington, D.C.
(1975).
385. O. A. Anunziata et al., Appl. Catal. 15 (1985)
235 – 245. K. Verschueren: Handbook of
Environmental Data on Organic Chemicals,
2nd ed., Van Nostrand Reinhold, New York
1983.
386. W. E. Coleman et al., Arch. Environ. Contam.
Toxicol. 13 (1984) 171 – 178.
387. N. I. Sax: Dangerous Properties of Industrial
Materials, 6th ed., Van Nostrand Reinhold,
New York 1984.
388. Threshold Limit Values and Biological
Exposure Indices for 1987 – 1988, American
Conference of Governmental Industrial
Hygienists. Cincinnati, Ohio, 1987.
389. DFG, MAK- und BAT-Werte Liste 2001,
Mitteilung 37, Wiley-VCH, Weinheim, 2001.
390. H. W. Gerarde: Toxicology and Biochemistry
of Aromatic Hydrocarbons, Elsevier
Publishing Co., London 1960.
391. G. Hommel: Handbuch der gefährlichen
Güter, Springer Verlag, Berlin Heidelberg
1987.
392. H. W. Gerarde, Am. Med. Assoc. Arch. Ind.
Health 19 (1959) 403.
393. M. A. Wolf et al., Am. Med. Assoc. Arch. Ind.
Health 14 (1956) 387.
394. H. W. Werner, R. C. Dunn, W. F. von
Oettingen, Ind. Health Toxicol. 26 (1974) 264.
395. Maximale Arbeitsplatzkonzentrationen und
biologische Arbeitsstofftoleranzwerte, Dt.
Forschungsgemeinschaft, Verlag Chemie,
Weinheim 1984.
396. N. I. Sax: Dangerous Properties of Industrial
Materials, 6th ed., van Nostrand Reinhold.
New York 1984, p. 819.
397. H. E. Christensen, T. T. Luginbyhl: Registry of
Toxic Effects of Chemical Substances, U.S.
Hydrocarbons 59
Dept. of H. E. W., U.S. Government Printing
Office, Washington, D.C., 1975.
398. J. Opdyke, Food Cosmet. Toxicol. 12 (1974)
707.
399. J. M. Haas, H. W. Earhart, A. S. Todd: Book of
Papers, 1974 AATCC National Technical
Conference, American Association of Textile
Chemists and Colorists, Research Triangle
Park, N. C., p. 427.
400. I. Hakkinen et al., Arch. Environ. Health 26
(1973) 70.
401. A. M. Seppalainen, I. Hakkinen, J. Neurol.
Neurosurg. Psychiatry 38 (1975) 248.
402. Patty’s Industrial Hygiene and Toxicology,
vol. 2 B, Toxicology, 3rd ed.
403. Z. F. Khromenko, Gig. Tr. Prof. Zabol. 1976,
44.
404. Y. R. Adamson, J. L. Weeks, Arch. Environ.
Health 27 (1973) no. 8, 68.
405. Y. R. Adamson, J. M. Furlong, Arch. Environ.
Health 28 (1974) no. 3, 155.
406. R. J. Hawkins, J. L. Weeks, At. Energy Can.
Ltd. (Rep.) 4439 (1973); Chem. Abstr. 80
(1974) 65 963.
407. J. S. Henderson, J. L. Weeks, Ind. Med. Surg.
40 (1973) 10.
408. International Agency for Research on Cancer
(ed.): IARC Monographs on the Evaluation of
the Carcinogenic Risk of Chemicals to
Humans, vol. 32, IARC, Lyon 1983.
409. Phenanthrene, Dangerous Prop. Ind. Mat.
Rep. 6 (1986) 68 – 69.
410. E. E. Sandmeyer in G. D. Clayton, F. E.
Clayton (eds.): Patty’s Industrial Hygiene and
Toxicology, 3rd ed., vol. 2 B, J. Wiley & Sons,
New York 1981, p. 3353.
411. Y. S. Rotenberg, F. D. Mashbits, Gig. Tr. Prof.
Zabol. 9 (1965) 53 – 54.
412. H. W. Gerarde: Toxicology and Biochemistry of
Aromatic Hydrocarbons, Elsevier Publishing,
Amsterdam – London – New York 1960.
413. S. D. Keimig, R. W. Pattillo, Brookhaven Natl.
Lab., Rep. BNL-51885, Upton 1984.
414. Z. Matijasevic, E. Zeiger, Mutat. Res. 142
(1985) 149 – 152.
415. M. Sakai, D. Yoshida, S. Mizusaki, Mutat.
Res. 156 (1985) 61 – 63.
416. J. E. Rice, T. J. Hosted, E. J. Lavoie, Canc.
Lett. (Shannon Irel.) 24 (1984) 327 – 333.
417. F. Palitti et al., Mutat. Res. 174 (1986)
125 – 130.
418. J. R. Babson et al., Toxicol. Appl. Pharmacol.
85 (1986) 355 – 366.
419. R. P. Bos et al., Mutat. Res. 187 (1987)
119 – 125.
60 Hydrocarbons
420. A. L. Reshetyuk, E. J. Talakina, P. A.
En’yakova, Gig. Tr. Prof. Zabol. 14 (1970)
46 – 47.
421. K. Knobloch, S. Szendzikowski, A.
Slusarczyk-Zalobna, Med. Pr. 20 (1969)
210 – 222.
422. D. T. Gibson, V. Subramanian in D. T. Gibson
(ed.): Microbial Degradation of Organic
Compounds, Marcel Dekker, New
York – Basel 1984, pp. 181 – 252.
423. H. H. Tabak, S. A. Quave, C. J. Mashni, E. F.
Barth, J. Water Pollut. Control Fed. 53 (1981)
1503 –1518.
424. W. S. Hall, T. J. Leslie, K. L. Dickson, Environ.
Contam. Toxicol. 36 (1986) 286 – 293.
425. G. W. Holcombe, G. L. Phipps, J. T. Fiandt,
Ecotoxicol. Environ. Saf. 7 (1983) 400 – 409.
426. K. Verschueren: Handbook of Environmental
Data on Organic Chemicals, 2nd ed., Van
Nostrand, New York – London 1983.
427. B. R. Parkhurst in R. A. Conway (ed.):
Environmental Risk Analysis for Chemicals,
Van Nostrand, New
York – London,pp. 399 – 411.
428. M. Sittig: Priority Toxic Pollutants, Noyes
Data Corp., Park Ridge, N.J., 1980.
 Hydrochloric Acid 1

Hydrochloric Acid
Severin Austin, Hüls Aktiengesellschaft, Marl, Federal Republic of Germany
Arndt Glowacki, Hüls Aktiengesellschaft, Marl, Federal Republic of Germany

1. Introduction . . . . . . . . . . . . . . . 1 3.2.2. Reaction of Metal Chlorides with Sul-

2. Properties . . . . . . . . . . . . . . . . . 2 furic Acid . . . . . . . . . . . . . . . . . 8
2.1. Physical Properties . . . . . . . . . . . 2 3.2.3. Recovery as a Chlorination Byproduct 9
2.2. Chemical Properties . . . . . . . . . . 4 3.2.4. Recovery from Waste Incineration . . 11
3.3. Purification . . . . . . . . . . . . . . . . 12
2.3. The HCl–H2 O System . . . . . . . . . 5
4. Quality Specifications . . . . . . . . . 12
3. Production . . . . . . . . . . . . . . . . 6 5. Storage and Transportation . . . . . 13
3.1. Methods of Preparation . . . . . . . . 7 6. Uses and Economic Aspects . . . . . 13
3.2. Industrial Production . . . . . . . . . 7 7. Toxicology and Occupational Health 14
3.2.1. From Hydrogen and Chlorine . . . . . 7 8. References . . . . . . . . . . . . . . . . . 15

1. Introduction The industrial synthesis of hydrogen chlo-

Hydrochloric acid, also known as muriatic
acid, is a solution of hydrogen chloride (HCl)
[7647-01-0] in water. Both hydrochloric acid
and hydrogen chloride are discussed in this arti-
cle.
History. In the 15th century, the German al-
chemist Valentin heated so-called green vit-
riol [iron (II) sulfate, FeSO4 · 7 H2 O] with com-
mon salt. He thereby obtained what then was
called spirit of salt. In the 17th century, Glauber
prepared hydrochloric acid from common salt
and sulfuric acid. In 1790, Davy established the
composition of hydrogen chloride by synthesiz-
ing it from hydrogen and chlorine. In the same
year, Leblanc discovered the process named
after him for the production of soda, the first
stage of which is to react common salt with
sulfuric acid, liberating hydrogen chloride. This
was at first considered an undesirable byproduct
and simply released into the atmosphere in large
amounts. In 1863, English soda producers were
compelled by the Alkali Act to absorb the hy-
drogen chloride in water, which quickly led to
large-scale industrial use of the acid produced.
Excess hydrogen chloride that could not be used
as hydrochloric acid was oxidized to chlorine.
ride followed the development of the chlor-alkali
electrolytic process early in the 20th century.
This route was used in addition to that based
on the reaction between chlorides and sulfuric
acid or sodium hydrogen sulfate and resulted in
a product of higher purity.
These processes are becoming less important
because of the large amounts of hydrogen chlo-
ride arising as byproducts of chlorination pro-
cesses such as the production of vinyl chloride
from ethylene. Hydrochloric acid may also be
recovered from gases produced by the incin-
eration of chlorine-containing waste, a conse-
quence of the ever-increasing emphasis on envi-
ronmental protection.
2. Properties
2.1. Physical Properties
Under standard conditions, hydrogen chloride,
HCl, M r 36.461, is a colorless gas with a pun-
gent odor and an extremely mordant effect on the
mucous membranes of the respiratory system.
Some physical properties of hydrogen chloride
are given below.

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 283
2 Hydrochloric Acid

mp −114.4 ◦ C (158.8 K)
of liquid hydrogen chloride, in millibars, is given
bp (101.3 kPa) −85.0 ◦ C (188.1 K) by
Vapor density
(0 ◦ C, 1.639 g/L
101.3 kPa) logp= −8.555 × 102 T −1 +7.553
Liquid density
(−85.0 ◦ C) 1.187 g/cm3
Critical temperature T crit 51.6 ◦ C (324.7 K)
Critical pressure Pcrit 8.3 MPa (83.0 bar)
which is fairly accurate over the entire liquid
Critical density
crit 0.41 g/cm3 range.
Specific heat cp [1]
100 K 40 J mol−1 K−1
153 K 49 J mol−1 K−1
Thermal expansion coefficient
100 K 2.5×10−4 K−1
153 K 5.7×10−4 K−1
Isothermal compressibility κT
[2]
100 K 2×10−5 bar−1
153 K 1×10−4 bar−1
Enthalpy of fusion ∆H fus 1970 J/mol
54.0 J/g
Entropy of fusion ∆S fus 12.54 J mol−1 K−1
0.344 J g−1 K−1
Enthalpy of vaporization
∆H vap (−85.0 ◦ C) 16.150 kJ/mol
0.443 kJ/g
Figure 2. Specific heat cp of liquid hydrogen chloride

Figure 3. Surface tension (——) and dynamic viscosity (-

Figure 1. Density
of liquid hydrogen chloride - - -) of liquid hydrogen chloride

Solid hydrogen chloride undergoes a poly-

morphic transition at 98 K from an orthorhombic
face-centered low-temperature phase II to a cu-
bic face-centered high-temperature phase I: en-
thalpy of transformation ∆H = 1190 J/mol, en-
tropy of transformation ∆S = 12.1 J mol−1 K−1
[3]. The vapor pressure of solid hydrogen chlo-
ride, in millibars, is given [1] by

logp= −911.31T −1 +7.1875 + 4.313 × 10−3 T

On melting the volume increases by 14 %.

The density, specific heat, viscosity and sur-
face tension, thermal conductivity, and enthalpy
of vaporization of liquid hydrogen chloride, each
as a function of temperature, are shown in Fig-
ures 1, 2, 3, 4, 5, respectively. The vapor pressure Figure 4. Thermal conductivity λ of liquid hydrogen chlo-
ride
 Hydrochloric Acid 3

hydrogen chloride in various organic solvents is

listed in Table 1 [5–7].

Figure 6. Specific heat cp of gaseous hydrogen chloride

Figure 5. Enthalpy of vaporization ∆H vap of hydrogen
chloride

Table 1. Solubility of hydrogen chloride in various organic liquids

at an HCl partial pressure of 101.3 kPa (1.013 bar) ∗

Solvent Temperature, Dissolved HCl,

◦
C g/kg solvent

n-Hexane 7.1
n-Hexadecane 27 3.9
102 2.2
177 1.6
Benzene 20 20
40 12.5
Toluene 20 21.3
Methanol 0 1092
20 877
40 688 Figure 7. Viscosity η of gaseous hydrogen chloride at
Ethanol 0 838 101.3 kPa
10 756
20 681
30 610
Diisopropyl ether 10 349
Dibutyl ether 10 250
Dioxane 10 433
Tetrahydrofuran 10 584
Chloroform 15 8.5
25 6.9
Carbon tetrachloride 20 6.0
50 3.9

∗ Hydrogen chloride forms addition compounds with many

organic compounds; in such cases, its solubility often either is
unknown or cannot be defined exactly.

The vapor density of gaseous hydrogen

chloride shows that it consists of single HCl
molecules. The specific heat, viscosity, and ther-
mal conductivity of gaseous hydrogen chloride,
each as a function of temperature, are shown in Figure 8. Thermal conductivity λ of gaseous hydrogen
chloride at 101.3 kPa
Figures 678, respectively [4]. The solubility of
4 Hydrochloric Acid
The most important thermodynamic data are
as follows [8]:
Enthalpy of formation
∆H 0f (298.15 K) −92.31 kJ/mol
Entropy S 0 (298.15 K) 186.78 J K−1 mol−1
Enthalpy H 0 (298.15 K) – H 0 8.64 kJ/mol
(0)
Specific heat cp (298.15 K) 29.12 J mol−1 K−1
Free energy of formation
∆G 0 (293.15 K) −95.25 kJ/mol
Figure 9. Heats of solution (25 ◦ C) in the HCl–H2 O system
a) Total enthalpy of solution ∆sol H HCl , in kilojoules per
gram of hydrogen chloride; b) Total enthalpy of solution in
kilojoules per gram of solution; c) Partial molar enthalpy of
solution ∆sol H HCl ; d) Enthalpy of dilution ∆dil H HCl
2.2. Chemical Properties
Hydrogen chloride is thermally stable up to ap-
proximately 1500 ◦ C, but appreciable dissocia-
tion occurs above this temperature. Completely
dry hydrogen chloride is not very reactive: mild
steel is not measurably attacked. Moreover, re-
actions involving dry hydrogen chloride often
take place only in the presence of catalysts. In
contrast, a solution of hydrogen chloride in a
polar solvent is a strong acid and, therefore, an
aggressive reagent.
Hydrogen chloride dissolves readily in water,
liberating much heat. The enthalpy of solution,
(i.e., the total heat of solution ∆sol H HCl ), the
partial molar enthalpy of solution


∂∆sol HHCl
∆sol HHCl =
∂nHCl p,T ,nH2 O
and the partial molar enthalpy of dilution


∂∆sol HHCl
∆dil HHCl =
∂nH2 O p,T ,nHCl
are shown in Figure 9 [9]. The hydrochloric acid
formed is a strong and extremely aggressive
acid, attacking almost all industrially available
metals and alloys to some extent. Metals must,
therefore, be protected by an appropriate non-
metallic lining, or nonmetallic materials of con-
struction must be used. Tantalum or highly re-
sistant nickel alloys such as Hastelloy B may be
used under precisely defined conditions. Less
noble metals dissolve with liberation of hydro-
gen.
2.3. The HCl–H2 O System
In the HCl–H2 O system at low tempera-
ture, four congruently melting solid phases are
formed. The corresponding compositions are
given by the following formulas: HCl · H2 O
(mp −15.4 ◦ C), HCl · 2 H2 O (mp −18 ◦ C),
HCl · 3 H2 O (mp −25 ◦ C), and HCl · 6 H2 O (mp
−70 ◦ C). At higher temperature an important
physical property is the solubility of gaseous hy-
drogen chloride in water and the partial pressure
of hydrogen chloride and water over hydrochlo-
ric acid at various concentrations and tempera-
tures.
These relationships are shown in Figure 10;
they have great practical significance for the ab-
sorption and desorption of hydrogen chloride.
At lower temperature (below about 100 ◦ C) and
pressure (less than about 0.2 MPa), the curves
are based on data given in [10], [11]. After care-
ful verification of the original experimental data,
these remain the best values available; they are
consistent with the partial pressures of hydro-
gen chloride and water derived from e.m.f. mea-
surements [12]. Values for the high-pressure re-
gions have been determined more recently [13–
16] and these can be fitted smoothly to the older
values.

Figure 10. Vapor–liquid phase diagram in the HCl–H2 O
system
Temperature (◦ C) is shown for each isotherm. The two-
phase region is delimited from the gas phase by the broken
lines (- - - -) and from the liquid phase by the continuous lines
(——). Curve A–A connects the azeotropic points; B–B
is the critical segregation curve above the critical point of
water (point B: 374.1 ◦ C, 21.43 MPa)
The system has a vapor pressure minimum
(azeotrope) whose position shifts toward lower
hydrogen chloride concentration with increas-
ing pressure and temperature. At 101.3 kPa, the
azeotrope, bp 108.6 ◦ C, contains 20.4 wt % hy-
drogen chloride. When the hydrogen chloride
concentration is lower than that, the vapor has a
higher water content than the liquid with which
it is in equilibrium. In this case it is not possible
to produce pure anhydrous hydrogen chloride
by simple distillation. If the concentration of the
acid exceeds that of the azeotropic mixture, how-
ever, dry hydrogen chloride can be isolated by
distillation. The expense of distillation decreases
with increasing hydrogen chloride concentration
of the acid.
The hydrogen chloride gas phase coexisting
with concentrated hydrochloric acid of more
than 35 wt % has a very low water content. A
corollary of this is that the moisture content of
hydrogen chloride gas must be kept very low
to prevent condensation of aqueous hydrochlo-
ric acid, which would cause corrosion problems
in steel pipes and equipment. At higher pressure
and lower temperature, the permissible moisture
Hydrochloric Acid 5
content must be even lower. Limiting concentra-
tions are given in Table 2. For example, hydro-
gen chloride gas in steel pipes at 0.5 MPa and
30 ◦ C may contain up to 800 ppm (by volume)
of water, but at 0 ◦ C the tolerable water content
is < 100 ppm.
At higher pressure and below the critical tem-
perature of hydrogen chloride, the HCl–H2 O
system has a miscibility gap, i.e., highly con-
centrated hydrochloric acid coexists with liq-
uid hydrogen chloride. For example, at 0 ◦ C and
> 2.4 MPa, 64 % hydrochloric acid is in equilib-
rium with anhydrous liquid hydrogen chloride.
At lower pressure, hydrogen chloride becomes
a gas, and the equilibrium concentration of hy-
drochloric acid decreases with decreasing pres-
sure. With increasing temperature, the width of
the miscibility gap increases, and above the criti-
cal temperature of hydrogen chloride, the liquid
hydrogen chloride phase is no longer present.
Hydrochloric acid remains as the only homo-
geneous liquid phase, whose maximum concen-
tration decreases as the temperature rises: the
heterogeneous region simultaneously becomes
wider. As soon as the critical temperature of wa-
ter is exceeded, the liquid phase disappears. Nev-
ertheless, initially a heterogeneous region exists
that becomes smaller with increasing tempera-
ture. This region consists of two immiscible gas
phases in equilibrium: at 50 MPa and 400 ◦ C,
gaseous water containing approximately 13 %
hydrogen chloride is in equilibrium with hydro-
gen chloride containing approximately 35 % wa-
ter.
The density of aqueous hydrochloric acid for
the temperature range 0 – 100 ◦ C is given in Ta-
ble 3.
The specific heat cp decreases with increasing
concentration. Some values (at 20 ◦ C) follow:
HCl content,
wt % 16 20 24 28 32 36 40
cp , J g−1 K−1 3.11 2.93 2.78 2.65 2.55 2.47 2.40
The temperature dependence of cp for the
acid is considerably greater than that for wa-
ter. In the above concentration range and at
0 – 100 ◦ C, it is given by the following equation:


∆cp
=4.2 × 10−3 Jg−1 K−2
∆T
6 Hydrochloric Acid
Table 2. Mole fraction (×106 ) of water in hydrogen chloride gas in equilibrium with concentrated (saturated) hydrochloric acid (based on
the results of Kao [13] )

p, MPa Temperature, ◦ C

−10 0 10 20 30 40 50 60

0.1 60 440 870 2400 4500 9200 15 500 30 000

0.2 35 230 430 720 3200 6100 10 300 19 000
0.5 18 96 180 350 800 2000 3 900 7 500
1.0 10 60 120 240 430 700 2 200 4 400

Table 3. Density of hydrochloric acid, in grams per cubic centimeter, as a function of temperature

HCl concen- Temperature, ◦ C

tration, wt % 0 10 20 30 50 75 100

5 1.0266 1.0253 1.0230 1.0200 1.0120 0.9993 0.9841

10 1.0521 1.0501 1.0473 1.0440 1.0359 1.0234 1.0090
20 1.1065 1.1022 1.0979 1.0933 1.0839 1.0711 1.0575
30 1.1611 1.1551 1.1493 1.1433 1.1318 1.1175 1.1030
35 1.1875 1.1805 1.1738 1.1671 1.1540 1.1379 1.1220

The viscosity η is considerably higher than that
of pure water. For 10 wt % acid, the viscosity
varies with temperature as follows:
t, ◦ C 0 10
η, mPa · s 1.89 1.45
The variation of viscosity with acid concen-
tration is given below (20 ◦ C):
such as chlorine or chlorinated organic com-
pounds.
In nature, hydrogen chloride is liberated in
volcanic eruptions or other releases of subter-
20 ranean gas. The amount of hydrogen chloride
1.16 produced can be of the same order as that of
sulfur dioxide. Hydrochloric acid occurs in the
gastric juice of mammals; in the human stomach,
the concentration of acid is about 0.1 mol/L.

HCl, wt % 10 20 30
η, mPa · s 1.16 1.36 1.70 3.1. Methods of Preparation
Hydrogen chloride can be formed according to
The thermal conductivity λ is considerably the following reactions:
lower than that of water. For 30 % hydrochloric
acid, it is 4.6×10−5 W m−1 K−1 (30 ◦ C), com- 1) Synthesis from elements:
pared with 6.18×10−5 W m−1 K−1 for water.
H2 + Cl2 −→ 2 HCl (1)

3. Production 2) Reaction of metallic chlorides, particularly

sodium chloride, with sulfuric acid or a hy-
Free hydrogen chloride occurs naturally only in drogen sulfate:
low concentration; it has been detected in the
stratosphere at levels of ca. 10−11 vol % [17].
Inorganic chlorides, however, are widely dis-
tributed, chlorine being the eleventh chemical
element in order of abundance. Alkali-metal and
alkaline-earth chlorides, in particular rock salt, 3) As a byproduct of chlorination, e.g., in the
are the primary raw materials for all routes to hy- production of dichloromethane, trichloro-
drogen chloride or hydrochloric acid production ethylene, perchloroethylene, or vinyl chlo-
including those involving intermediate products ride:

C2 H4 + Cl2 −→ C2 H4 Cl2 (5)
C2 H4 Cl2 −→ C2 H3 Cl + HCl (6)
4) From spent pickle liquor in metal treatment,
by thermal decomposition of the hydrated
heavy-metal chlorides:
5) From incineration of chlorinated organic
waste:
C4 H6 Cl2 + 5 O2 −→ 4 CO2 + 2 H2 O + 2 HCl (8)
By far the greatest proportion of hydrogen
chloride is now obtained as a byproduct of chlo-
rination (method 3, Eqs. 5 and 6). The degree of
purification required depends on the end use. At
present, recovery from waste material according
to methods 4 and 5 is becoming increasingly im-
portant, whereas the amount of hydrogen chlo-
ride generated by method 2 is decreasing. Re-
covery of hydrochloric acid from pickle liquor
is described in [18]. Finally, hydrogen chloride
can be formed in undesirable reactions such as
the hydrolysis of Friedel – Crafts catalysts:
AlCl3 + 3 H2 O −→ Al(OH)3 + 3 HCl (9)
3.2. Industrial Production
3.2.1. From Hydrogen and Chlorine
The simplest method for producing hydrogen
chloride is direct synthesis from the elements,
which yields a very pure product. The reaction is
strongly exothermic; the standard enthalpy of re-
action of Equation (1) is ∆H 0R = −184 kJ. Mix-
tures of hydrogen and chlorine are extremely ex-
plosive over a wide range of composition, and
explosion can often be initiated by light. Care
must, therefore, be taken to avoid the occurrence
of reactive mixtures of chlorine and hydrogen in
industrial processes.
Hydrochloric Acid 7
Industrial production involves the use of a
burner; chlorine and hydrogen are fed by sepa-
rate concentric tubes into the combustion cham-
ber. After ignition, the chlorine burns in the hy-
drogen with a quiet, very hot flame at > 2000 ◦ C.
Silica has proved to be a good material of con-
struction for the burner. Cast iron or steel, some-
times water-cooled, or graphite has also been
used. The burner is usually located on the bot-
tom of a cylindrical combustion chamber, with
the flame directed vertically upward. The com-
bustion chamber is made of steel lined with re-
fractory brick. A cooling section of appropriate
size and shape is connected to the combustion
chamber.
If the reactants are very moist or otherwise
contaminated, corrosion-resistant construction
materials such as silica or graphite must be used.
Manufacturing plants still in use today, however,
are constructed of carbon steel and the hydro-
gen chloride produced is water-cooled only to a
temperature at which aqueous hydrochloric acid
cannot condense. The most probable contami-
nants are moisture in the hydrogen and oxygen
in the chlorine, but oxygen compounds such as
dichlorine oxide (Cl2 O) or carbon dioxide can
also cause problems by the formation of water.
The mixing ratio of the raw material is ad-
justed according to the intended use of the hy-
drogen chloride produced. A slight excess of
hydrogen or chlorine may be used; it is inad-
visable to work with exactly equimolar mixtures
because the reaction is not quite quantitative and
it is more difficult to avoid explosive mixtures.
Control deviations can easily lead to alternation
between excess hydrogen and excess chlorine,
with explosive mixtures appearing from time to
time. Hydrogen in hydrogen chloride is usually
less troublesome than chlorine, so that an ex-
cess of 1 – 2 % hydrogen is usually employed.
Depending on the purity of the raw materials,
extremely pure chlorine-free hydrogen chloride
can be obtained, with excess hydrogen being the
only impurity. This hydrogen chloride can be
used to produce very pure hydrochloric acid, or
it may be liquefied or supplied directly to the
user in steel pipes as a dry gas.
The use of steel for the synthesis furnaces
and coolers is a particularly attractive possibil-
ity. They can then be operated under elevated
pressure, the magnitude is determined only by
the pressure of chlorine and hydrogen and the
8 Hydrochloric Acid
temperature and pressure in the pipeline. Here
the pressure must be low enough, or the temper-
ature high enough, to prevent condensation of
aqueous hydrochloric acid (see Fig. 10 and Ta-
ble 2). The steel furnaces operated at Hüls can
each produce 30 – 40 t/d hydrogen chloride. The
hydrogen chloride gas is generated at a pressure
up to 0.7 MPa, the water content being < 50 ppm
(by volume); steel pipes are, therefore, used to
convey it from the plant.
Chemically related processes [19] include the
reaction of chlorine with carbon (in the form of
coke) and water vapor
C + 2 H2 O + 2 Cl2 −→ 4 HCl + CO2 (10)
or the reaction of chlorine with sulfur dioxide
and water
SO2 + 2 H2 O + Cl2 −→ 2 HCl + H2 SO4 (11)
3.2.2. Reaction of Metal Chlorides with
Sulfuric Acid
The sulfate process has a high energy consump-
tion. It is, therefore, gradually decreasing in
importance compared with other processes, es-
pecially organic chlorination, which inevitably
produces large amounts of hydrogen chloride
as a byproduct. Moreover, the old argument
that sulfate–hydrochloric acid provides a prod-
uct with higher purity no longer holds; byprod-
uct acid can now be purified sufficiently to sat-
isfy even requirements for high quality. Hence,
only the most important sulfate–hydrochloric
acid process is described here. Further details
are given in [20].
The reaction of common salt with con-
centrated sulfuric acid according to Equa-
tion (2) occurs at comparatively low temperature
(150 – 300 ◦ C); hydrogen chloride and sodium
hydrogen sulfate are formed. The latter reacts
with excess sodium chloride at a minimum tem-
perature of 550 – 600 ◦ C, according to Equa-
tion (3), forming neutral sodium sulfate. The use
of sodium hydrogen sulfate as a starting material
(Eq. 3) is quite feasible but has been discontin-
ued because the high temperature (600 – 800 ◦ C)
required and the high energy consumption ren-
der the process uneconomic.
Sulfuric acid is now the exclusive starting ma-
terial. In the Mannheim process, sodium sulfate
is produced; in the Berlin hydrochloric acid pro-
cess, sodium hydrogen sulfate.
In the Mannheim process, brick-lined exter-
nally fired muffle furnaces are used. These con-
tain a stirrer with a scraping action which pre-
vents agglomeration of the pasty mass and also
removes the sodium sulfate produced. Hydro-
gen chloride passes through the side of the muf-
fle and is propelled by a fan to the next stage,
the furnace being maintained at a slightly re-
duced pressure. The mixture of gases exhausted
from the furnace contains up to 85 % hydrogen
chloride, as well as air, sulfuric acid mist, and
fine particles of salt. As a rule, the gas must
undergo several purification stages such as fil-
tration through coke and activated carbon, wet
scrubbing, or chemical reaction.
The Berlin hydrochloric acid process yields
considerably purer hydrogen chloride gas than
the Mannheim process. The reaction of com-
mon salt with sulfuric acid takes place in molten
sodium hydrogen sulfate at about 300 ◦ C in cast-
iron retorts. Alternatively, potassium chloride
can be used, to give hydrogen chloride and potas-
sium hydrogen sulfate. This process has retained
some commercial importance because the prin-
cipal products are the potassium salts formed
(i.e., KHSO4 and K2 S2 O7 ) [21], whereas hy-
drogen chloride is considered as a byproduct.
Gas from the retorts is more concentrated be-
cause the equipment is more leak proof, and less
sulfuric acid mist is present as a result of the
lower working temperature.
3.2.3. Recovery as a Chlorination Byproduct
By far the largest amounts of hydrogen chlo-
ride and hydrochloric acid are produced as by-
products of chlorination. Equation (6) is merely
an example of the varied range of industrial
chlorination processes. Nevertheless, the ma-
jor source of hydrogen chloride is the crack-
ing of 1,2-dichloroethane to give vinyl chlo-
ride (→ Chlorinated Hydrocarbons). The quan-
tity of hydrogen chloride produced and con-
sumed can be extremely large, in many installa-
tions reaching the order of 10 000 m3 /h (16 t/h or
140 000 t/a) and more. Consequently, hydrogen
chloride is often fed directly to a chemical plant

which uses it as a raw material; installed stor-
age capacities for hydrogen chloride are usually
comparatively small. It is, therefore, essential to
operate the plants at a balanced rate of produc-
tion and consumption. If production problems
or market variations affect one of the plants or
products, the other plant can only continue to op-
erate if reserve production capacity (normally
not fully used) is installed. Buffer storage for
liquid hydrogen chloride is used to overcome
short-term inbalances in production versus con-
sumption.
Hydrogen chloride produced by organic chlo-
rination can be treated in various ways:
1) simple isolation of the gas by condensing the
chlorinated hydrocarbons;
2) isolation and purification of hydrogen chlo-
ride by distillation;
3) aqueous absorption of hydrogen chloride.
Condensation of the chlorinated hydrocar-
bons is used when the presence of residual chlo-
rinated hydrocarbons or other residues from the
chlorination process does not cause any prob-
lems in the hydrogen chloride, which is fed
as a gas. The degree of purification attainable
depends on condensation pressure and tem-
perature. One example is the chlorination of
chloromethane to dichloromethane; the hydro-
gen chloride formed is used only to convert
methanol to chloromethane, which in turn is
used to produce dichloromethane:
CH3 Cl + Cl2 −→ CH2 Cl2 + HCl (12)
CH3 OH + HCl −→ CH3 Cl + H2 O (13)
The chloromethane and dichloromethane con-
tent of hydrogen chloride produced according
to Equation (12) does not have any adverse ef-
fect on the reaction of methanol with hydrogen
chloride (Eq. 13).
Fractional distillation of the liquefied gas
may be employed if a purer product is required.
For example, the gaseous products from the
cracking of 1,2-dichloroethane are liquefied and
then separated by distillation into vinyl chloride
and hydrogen chloride. Liquefaction of these
gases by reducing the temperature to extremely
low values is uneconomical, so condensation
and distillation are carried out at elevated pres-
sure. An advantage of this procedure is that hy-
drogen chloride gas is produced at a high enough
Hydrochloric Acid 9
pressure (usually 1 – 2 MPa) for subsequent pro-
cessing, thereby avoiding the necessity for sub-
sequent mechanical compression of the gas.
Aqueous absorption of hydrogen chloride
gives hydrochloric acid as an intermediate prod-
uct and leads to inevitable problems with con-
struction materials. These problems have be-
come less severe since the introduction of flu-
oropolymers. In many installations, however,
aqueous treatment has been replaced by non-
aqueous methods (e.g., fractional distillation).
The absorption of hydrogen chloride can be ef-
fected with water or 20 wt % hydrochloric acid.
The use of water as the absorption medium is
appropriate if the primary aim is the produc-
tion of 30 – 35 wt % hydrochloric acid as the
end product. If demand for the latter is insuffi-
cient, the 20 % azeotrope is used for the absorp-
tion of hydrogen chloride, thereby concentrating
the acid to 30 – 35 wt %. The concentrated hy-
drochloric acid is first purified, if necessary. It is
then distilled to give gaseous hydrogen chloride
and azeotropic hydrochloric acid. The hydrogen
chloride is dried and delivered to the consumer;
the azeotrope is recycled to the absorption stage.
If the presence of chlorinated hydrocarbons in
hydrochloric acid is undesirable, they can be re-
moved by introducing a stripping step. A sim-
plified scheme is shown in Figure 11. A small
amount of water is fed to the absorption column
to render the gas stream free of hydrogen chlo-
ride. A corresponding amount of acid must be
separated for other use.
A number of process and equipment param-
eters are important for assessing the economics
of the aqueous absorption of hydrogen chloride
from a production gas stream:
1) The concentration of hydrogen chloride in
the gas stream determines the pressure and
temperature that must be maintained in the
absorption column.
2) The energy requirement for liberating pure
hydrogen chloride gas decreases with in-
creasing concentration of the concentrated
acid formed.
3) Production of concentrated hydrochloric
acid from a gas stream with a low hydro-
gen chloride concentration requires compar-
atively great effort because either the pres-
sure must be increased, the temperature re-
duced, or both.
10 Hydrochloric Acid
Figure 11. Flow diagram showing recovery of hydrogen chloride as a byproduct of chlorination in aqueous absorption
(CHC = chlorinated hydrocarbons)
a) Hydrogen chloride absorption column; b) Chlorinated hydrocarbon stripping column; c) Hydrogen chloride desorption
column
4) The presence of a high concentration of wa-
ter vapor in the production gas stream can,
under certain circumstances, render the pro-
duction of concentrated hydrochloric acid
impossible.
Industrial processes have been reviewed by
Schmidt [10] and Vieweg [22]. The latter pro-
poses the use of magnesium chloride solution to
absorb low levels of hydrogen chloride from gas
containing a high level of water vapor.
The choice of construction materials for hy-
drogen chloride absorption equipment is re-
stricted because of the possibility of attack both
by concentrated hydrochloric acid and by chlo-
rinated hydrocarbons. Graphite heat exchangers
are used to cool the acid. For a long time a special
ceramic was the only economically viable ma-
terial for the pumps. The distillation column is
lined with acid-proof bricks or graphite blocks,
and the bricks are cemented with phenolic resin.
Frequently, a special acid-resistant rubber lining
is applied between the steel casing and the brick
wall to prevent attack by acid that might pen-
etrate pores and cracks. Vessels and equipment
made entirely of phenolic resin are chemically
very resistant, but brittle and mechanically frag-
ile, and require support (e.g., by an outer shell of
glass-fiber-reinforced polyester resin). This type
of equipment is suitable only for operation at at-
mospheric or possibly slightly higher pressure.
The same is true for the chlorinated hydro-
carbon stripping column. If the concentration
of chlorinated hydrocarbons is low, or if they
have been effectively removed from the acid,
other materials such as poly(vinyl chloride) can
be used for storage vessels. The desorption col-
umn, however, is made of graphite because of
the high operating temperature. Alternatively,
rubber-lined columns with inner brick linings
can be used. Steel columns lined with perflu-
oropolymer, especially polytetrafluoroethylene
(PTFE), have been used with generally excel-
lent results, although differences exist among the
products from different manufacturers with re-
gard to microporosity, tear resistance, and flow
properties. Compared with graphite columns,
coated steel columns have the advantage that
they can operate at elevated pressure, but failure
of a PTFE lining would lead to rapid destruc-
tion of the steel casing. However, these columns
can be constructed of fairly small individual
elements that may be replaced quickly, reduc-
ing maintenance downtime. Heat exchangers are
made of graphite in all cases. Another possibil-
ity is to use tantalum for lining and evapora-
tors, although special precautions must be taken
to avoid embrittlement due to absorption of hy-

drogen. Furthermore, hydrogen fluoride must be
absent.
Completely dry hydrogen chloride will not
attack steel pipes or compressors. Hydrogen
chloride gas was formerly dried by using con-
centrated sulfuric acid, but another approach is
to make use of the properties of the HCl – H2 O
system (see Fig. 10 and Table 2). At low temper-
ature, the partial vapor pressure of water over
highly concentrated hydrochloric acid is very
low; hydrogen chloride can therefore be dried by
cooling to sufficiently low subzero temperature.
At slightly elevated pressure (pabs ≈ 0.3 MPa), it
is sufficient to attain −10 ◦ C to give satisfactory
results. Hydrochloric acid mist must, however,
be removed completely, or severe corrosion can
result.
3.2.4. Recovery from Waste Incineration
All chlorination processes produce undesirable
chlorinated hydrocarbons in addition to the
product desired. These unusable and often nox-
ious substances must be disposed of for environ-
mental reasons. One possibility is incineration
to hydrogen chloride, water, and carbon diox-
ide; the hydrogen chloride should be recovered
for both economic and environmental reasons.
Although the methods used vary in detail
[23–25], the general principle remains the same.
Combustion takes place in a refractory-lined fur-
nace at > 1000 ◦ C. Waste chlorinated hydro-
carbons containing up to 70 wt % chlorine nor-
mally possess good combustion properties, but
any further increase in chlorine content reduces
the calorific value to such an extent that fuel oil
or gas must be added to support combustion.
Optimum reaction conditions lie within the
following limits:
1) Excess oxygen must always be present to
guarantee complete decomposition of chlori-
nated hydrocarbons. The hydrogen chloride
gas formed is consequently able to react with
this oxygen to form chlorine and water in ac-
cordance with the Deacon equation:
−→
2 HCl + 1/2 O2 ←− Cl2 + H2 O (14)
The reaction has a negative enthalpy of reac-
tion (∆H 0r = − 57.1 kJ/mol) and entropy of
Hydrochloric Acid 11
reaction (∆S 0r = − 64.7 J K−1 mol−1 ). With
increasing temperature, the equilibrium
shifts from right to left, thus favoring forma-
tion of hydrogen chloride. The temperature
is maintained as high as construction mate-
rials permit, to prevent formation of chlo-
rine; even if the partial pressure of water is
kept high and the oxygen excess low, a tem-
perature > 1000 ◦ C is necessary to shift the
equilibrium to the left. This keeps the chlo-
rine at an acceptable level and also ensures
complete combustion.
2) Reaction temperature must not exceed
1200 ◦ C to prevent formation of nitrogen
oxides, which would contaminate the hy-
drochloric acid product and, which is worse,
cause waste gas purification problems. A
possible solution is to use pure oxygen in-
stead of air [16].
For these reasons, incineration of waste chlo-
rinated hydrocarbons is carried out between
1000 and 1200 ◦ C. Cooling the flue gases com-
bined with steam generation can cause prob-
lems with construction materials. The absence
of corrosive combustion products must, there-
fore, be assured. The treated wastes should have
a well-known composition (down to trace con-
taminants) for proper prediction of the expected
combustion products, their concentration, and
behavior. Slow cooling can cause chlorine to re-
form, so rapid cooling is preferred.
Hydrogen chloride is recovered from the
combustion gases by absorption, i.e., the process
is similar to that used in recovery from chlori-
nation processes, except for the following:
1) A dry process is impossible because water
vapor is always present.
2) The gases should be condensed in such a way
as to form the most concentrated hydrochlo-
ric acid possible so that gaseous hydrogen
chloride or concentrated hydrochloric acid
can be obtained. Dilute hydrochloric acid has
little commercial value.
3) Special corrosion problems must be over-
come at places of possible, but not neces-
sarily regular, acid condensation.
4) If the gaseous combustion products are used
directly, the operating temperature must al-
ways be kept above the condensation tem-
perature of hydrochloric acid.
12 Hydrochloric Acid
One of the oldest plants producing commer-
cial quantities of 33 wt % hydrochloric acid by
incinerating chlorinated hydrocarbons is that of
Atochem at St. Auban in the south of France
[25]. This plant opened in 1975 and consumes
approximately 16 000 t of chlorinated hydrocar-
bons material per year. Many additional chlo-
rinated hydrocarbon incineration plants have
since been built, and the hydrochloric acid pro-
duced is finding a wider range of uses (e.g., ad-
dition to ethylene) [27].
3.3. Purification
Many impurities in hydrogen chloride gas (e.g.,
SO2 , As, and Cl2 ) can be removed by adsorp-
tion on activated carbon [28]. As the importance
of the sulfate process decreases, the removal of
chlorinated hydrocarbons from hydrogen chlo-
ride gas or hydrochloric acid is of greater prac-
tical relevance. Gaseous hydrogen chloride can
be purified by low-temperature scrubbing with
a high-boiling solvent, which may be another
chlorinated hydrocarbon (e.g., hexachlorobuta-
diene or tetrachloroethane), or special oil frac-
tions. The use of activated carbon is often un-
necessary after such treatment. Chlorine may be
removed with carbon tetrachloride, in which it
is much more soluble than is hydrogen chloride
gas.
Hydrochloric acid, used as such or for gen-
eration of hydrogen chloride, contains mainly
volatile impurities such as chlorinated hydro-
carbons. In these cases the impurities may be
removed from the acid by stripping (see Fig. 11).
Stripping with an inert gas stream, which results
in lower energy consumption, is possible, al-
though in most cases heating of the gas is prefer-
able for environmental reasons. Inorganic im-
purities, in particular iron, are removed by ion
exchange.
4. Quality Specifications
The quality specifications for gaseous hydrogen
chloride depend on intended use. There are no
general quality standards. Only a trace (a few
parts per million) of water is usually permitted,
particularly if the gas is to be compressed and
liquefied. Dry hydrogen chloride does not attack
iron; thus, steel equipment and pipelines can be
used. Hydrogen chloride gas is regarded as dry
if no danger exists of hydrochloric acid conden-
sation at operating temperature and pressure.
Hydrochloric acid has no generally accepted
quality specification. In the Federal Repub-
lic of Germany, the only official standard is
DIN 19 610 (November 1975), which gives the
following specifications for hydrochloric acid
for boiler feed water treatment:
Hydrogen chloride content ≥ 30 wt %
Sulfate (SO2−
4 ) content ≤ 0.5 wt %
Iron (Fe3+ ) content ≤ 0.002 wt %
Content of organic chlorine compounds (ex-
pressed as
Cl− ) ≤ 0.02 g/L
Quality requirements for medicinal use are
given in DAB 9 [29] for both concentrated
(36 wt %) and dilute (10 wt %) hydrochloric
acid. Prescribed purity tests include free chlorine
(4 ppm), arsenic (2 ppm), heavy metals (2 ppm),
sulfate (20 ppm), and residue on evaporation
(0.01 %). Other countries have their own reg-
ulations [30].
5. Storage and Transportation
Wrought- and cast-iron, steel, and steel alloys
have good resistance toward dry hydrogen chlo-
ride gas or liquid. Steel is, therefore, suitable for
equipment and containers up to 280 ◦ C, includ-
ing pressure vessels used for storage and trans-
portation of dry hydrogen chloride.
Even traces of moisture significantly increase
the corrosion rate, and alloyed austenitic steels
become liable to stress crack corrosion. If the
working temperature cannot be raised above the
condensation point, either nonmetallic materi-
als or more exotic metals must be used, such
as nickel–molybdenum alloys with high molyb-
denum content, zirconium, niobium, tantalum,
or noble metals such as gold or the platinum-
group metals. In practice, it is imperative to spec-
ify precisely the operating limits for metals to
be used with hydrochloric acid or moist hydro-
gen chloride. In all cases, limits are given for
acid concentration or temperature. Contamina-
tion with oxidizing or complexing agents must
be avoided. Traces of fluoride ions in hydrochlo-
ric acid, for example, can lead to rapid corrosion

of zirconium or tantalum, although tantalum has
proved suitable for hydrochloric acid distillation
columns in the absence of damaging impurities.
A wide choice of nonmetallic construction
materials is available. For high temperature, ce-
ramics, graphite, or polytetrafluoroethylene may
be suitable. Phenolic resins are used as cement
for brick lining. At lower acid concentration (be-
low 25 %), enameled or glass apparatus can be
used. At lower temperature, (e.g., for storage and
transportation), vessels may be made of rubber-
lined steel, polyolefin, or poly(vinyl chloride),
the latter usually with glass-fiber reinforcement.
If the acid contains traces of chlorinated hydro-
carbons such as dichloromethane, vessels fabri-
cated in, or lined with, phenolic resin are pre-
ferred. Plastics such as poly(vinyl chloride) and
polyethylene can withstand use for several years.
For transportation of hydrochloric acid and
liquefied hydrogen chloride, special labels must
be used that comply with the international reg-
ulations for transportation of hazardous goods:
Hydrochloric acid: Hazard code number 80
UN no. 1789
Warning sign 8 (causes burns)
CFR 49: 172.101 Cor. M
RID/ADR, class 8, no. 5 b
Hydrogen chloride: Hazard code number 286
UN no. 1050
Warning signs 6.1 (poisonous) and
8 (causes burns)
6. Uses and Economic Aspects
Hydrochloric acid and hydrogen chloride are
two of the most important basic industrial chem-
icals.
The largest proportion of hydrogen chloride
is used immediately by the producer himself,
because the hydrogen chloride byproduct of or-
ganic chlorination processes has to be used as
completely as possible. This is done by means
of oxychlorination or hydrochlorination, or to a
small extent, recovery of chlorine (e.g., by the
KEL process, a modified Deacon process [31]
or by electrolysis of hydrochloric acid [32]).
Aqueous hydrochloric acid has many uses as
a strong inorganic acid: manufacture of chlo-
rides, dissolution of minerals, pickling and etch-
ing of metals, regeneration of ion-exchange
resins for water treatment, neutralization of al-
Hydrochloric Acid 13
kaline products or waste materials, acidification
of brine in chlor-alkali electrolysis, and many
others.
Hydrochloric acid and hydrogen chloride
production figures for various countries are
given in Table 4.
7. Toxicology and Occupational
Health
Hydrogen chloride causes severe irritation of
the eye and respiratory tract and, to a lesser
extent, the skin. The vapor can produce kera-
toconjunctivitis. Inhalation causes irritation and
damage to mucous membranes. Hydrogen chlo-
ride is usually detectable by odor at 1 – 5 ppm
and becomes objectionable at 5 – 10 ppm. Pro-
longed occupational exposure to excessive hy-
drogen chloride concentration causes increased
incidence of chronic bronchitis or stomach and
intestinal disorders due to abnormal acid levels.
Increased incidence of a characteristic dental de-
cay has also been observed with prolonged expo-
sure, even at apparently tolerable levels. A con-
centration > 10 ppm is strongly irritating, even
when some acclimatization has occurred [33].
Suitable protective goggles should be worn
by personnel working with hydrochloric acid;
the use of protective clothing, rubber gloves, and
rubber boots is recommended. Inhalation of the
vapor should be avoided. The respirator filter for
inorganic gases is suitable (in the Federal Repub-
lic of Germany, code letter B, code color gray),
as is the hydrogen chloride and sulfur dioxide
filter (code letter E, code color yellow [34] ).
In the Federal Republic of Germany,
hydrogen chloride and hydrochloric acid
(> 10 wt %) are controlled by the Verordnung
über gefährliche Stoffe (GefStoffV, regulations
for the control of dangerous substances), which
have been adjusted to the corresponding direc-
tives of the EEC [35]:
1) Anhydrous hydrogen chloride: hazard sym-
bol C (corrosive). Hazard information (“R”
notices): 35 (causes serious burns); 37
(causes irritation of respiratory tract). Safety
Instructions (“S” notices): 7/9 (keep con-
tainers tightly closed and store in a well-
ventilated place); 26 (on contact with the eye,
14 Hydrochloric Acid
Table 4. Production of hydrogen chloride and hydrochloric acid, 1979 – 1993 (based on 100 % HCl)

Production, Country
3
10 t Federal Republic of France United Kingdom United States Japan
Germany
1980 891 676 143 2551 570
1981 888 702 128 2269 565
1982 848 688 118 2252 551
1983 900 673 130 2468 511
1984 956 765 156 2732 521
1985 945 656 153 2807 549
1986 931 626 157 2413 549
1987 987 647 174 2495
1989 ∗ 958 238 167 3268
1991 ∗ 839 216 153 3381
1992 ∗ 878 199 148 3610 1786
1993 ∗ 820 190 3492
∗
[40]

thoroughly wash with water and seek medi-
cal advice).
2) Hydrochloric acid containing over 25 wt %
hydrogen chloride: hazard symbol C (corro-
sive, causes burns).“R” notices: 34 (causes
burns); 37 (see above).“S” notices: 2 (keep
away from children), 26 (see above).
3) Hydrochloric acid with 10 – 25 wt % hydro-
gen chloride: hazard symbol Xi , irritant.“R”
notices: 36/38 (causes eye and skin irrita-
tion)“S” notices: 2 (see above), 28 (on con-
tact with the skin, immediately wash with
copious amounts of water).
In the event of skin or eye contact, the af-
fected area should be rinsed immediately and
thoroughly with a large amount of water. Con-
taminated clothing should be treated similarly,
and alkaline cleansing solution may also be used.
For work with hydrogen chloride, especially at
elevated pressure, breathing apparatus with an
independent air supply and full protective cloth-
ing must be readily available in case of accident.
In the Federal Republic of Germany, the
MAK value for hydrogen chloride is 5 ppm by
volume, corresponding to 7 mg/m3 [36]. The
same values are valid as TLV in the United
States. In the Soviet Union 3 ppm (5 mg/m3 ) has
been established.
In clean air regulations (TA Luft, 1986), con-
centration limits for hydrogen chloride in the
atmosphere are 0.1 mg/m3 as an annual aver-
age and 0.2 mg/m3 for short periods. Phytotox-
icological investigation has shown that plants
have differing tolerances [37–39]. According
to guidelines issued by the VDI, 2310 sheet 4,
sensitive vegetation such as lettuce, strawber-
ries, spruce, birch, and apple trees should not
be exposed to a hydrogen chloride concentra-
tion higher than 0.8 mg/m3 as a daily average or
0.1 mg/m3 as a monthly average.
The concentration of hydrogen chloride in in-
dustrial waste gases must (according to regula-
tions in the Federal Republic of Germany) nor-
mally not exceed 30 mg/m3 , although the limit is
50 mg/m3 for waste incineration plants; the cost
of gas cleaning can, therefore, vary considerably.
A simple water scrubbing operation is often suf-
ficient, but more than one stage may be neces-
sary, sometimes including a final scrubbing with
an alkaline solution. In many cases it is not possi-
ble to produce a commercial hydrochloric acid.
The dilute acid is then neutralized before be-
ing discarded in the wastewater. Neutralization
with caustic soda occasionally results in a salt
solution that is pure enough to be recycled to a
chlorine plant.
8. References
1. H. Chihara, A. Inaba, J. Chem. Thermodyn. 8
(1976) 915 – 934.
2. S. Miskiewicz, K. Rieser, T. Dorfmüller, Ber.
Bunsenges. Phys. Chem. 80 (1976) 395 – 405.
3. K. Clusius, G. Wolf, Z. Naturforsch. A (1947)
495.
4. L. S. Adler, C. L. Yaws, Solid State Technol.
19 (1976) 35 – 38.

5. K. K. Tremper, J. M. Pransnitz, J. Chem. Eng.
Data 21 (1976) 295 – 299.
6. W. Gerrard, E. D. Macklen, Chem. Ind.
(London) 1959, 1070 – 1071.
7. S. M. Khodeeva, M. B. Rozovskii, Russ. J.
Phys. Chem. (Engl. Trans.) 49 (1975)
824 – 827.
8. Codata Recommended Key Values, J. Chem.
Thermodyn. 10 (1978) 904.
9. D. D. Wagman, W. H. Evans, I. Halow, V. B.
Parker, S. M. Bailey, R. H. Schumm, NBS
Tech. Note (U.S.) 270 (1965) no. 1, 25 – 27.
10. A. Schmidt, Chem. Ing. Tech. 25 (1953)
455 – 466.
11. Ullmann, 3rd ed., 15, 70.
12. J. J. Fritz, C. R. Fuget, Ind. Eng. Chem. Data 1
(1956) 110 – 112.
13. J. F. F. Kao, J. Chem. Eng. Data 15 (1970)
362 –367.
14. B. G. Staples, J. M. Procopio, G. J. Su, Chem.
Eng. (N.Y.) 77 (1970) 113 – 115.
15. W. Kindler, G. Wuster, H. Rau, Ber.
Bunsenges. Phys. Chem. 82 (1978) 543 – 545.
16. R. W. Bach, H. A. Friedrichs, H. Rau, High
Temp. High Pressures 9 (1977) 305 – 312.
17. O. F. Raper, C. B. Farmer, R. A. Toth, B. D.
Robbins, Geophys. Res. Lett. 4 (1977)
531 – 534.
18. Ullmann, 4th ed., 10, 431 – 433.
19. Ullmann, 3rd ed., 15, 75.
20. Ullmann, 3rd ed., 15, 71 – 73.
21. Ullmann, 4th ed., 13, 482.
22. R. Vieweg, Chem. Tech. (Leipzig) 12 (1960)
no. 4, 188 – 195.
23. J. C. Zimmer, R. Guaitella, Hydrocarbon
Process. (1976) 117 – 118.
Hydrochloric Acid 15
24. Hoechst, US 4 073 871, 1977 (W. Opitz, H.
Hennen); DT 2 611 671, 1976.
25. Rhône-Poulenc, Company brochure,
Combustion of Chlorinated Residues (1978).
26. Stauffer Chemical O., Hydrocarbon Process.
(1981) 170.
27. Uhde GmbH, EP 0 137 160, 1985, 1987 (U.
Rieker, G. Link).
28. Ullmann, 3rd ed., 15, 80.
29. DAB, vol. 9.
30. Europäisches Arzneibuch, vol. 2, 137 – 138,
Deutscher Apotheker-Verlag, Stuttgart 1975.
31. Ullmann, 4th ed., 9, 357.
32. Ullmann, 4th ed., 9, 355.
33. D. Henschler (ed.): Gesundheitsschädliche
Arbeitsstoffe, Verlag Chemie GmbH,
Weinheim 1972 – 1979.
34. Datenbank für wassergefährdende Stoffe
(DABAWAS) vom 23. 03. 1987, 2989.
35. Verordnung über gefährliche Arbeitsstoffe
vom 26. 08. 1986 (BGBL 1, S. 1470).
36. Mitteilung 23 der Senatskommission zur
Prüfung gesundheitsschädlicher Arbeitsstoffe,
VCH Verlagsgesellschaft mbH, Weinheim
1987.
37. H. J. Grecelius, W. Forwer, R. Zahn, Staub,
Reinhalt. Luft 36 (1976) 201 – 205.
38. H. van Haut, Schriftenr. LJB Landesanst.
Immissions Bodennutzungsschutz Landes
Nordrhein Westfalen 33 (1974) 39 – 43.
39. C. Holsenberg, Schriftenr. Ver. Wasser Boden
Lufthyg. 41 (1974) 267 – 284.
40. K. H. Büchel, H.-H. Moretto, P. Woditsch:
Industrial Inorganic Chemistry, Wiley-VCH,
Weinheim 2000, p. 163.
 Hydrocyclones 1

Hydrocyclones
Enrique Ortega-Rivas, Food and Chemical Engineering Program, Autonomous University of Chihuahua;
University Campus I, Chihuahua, Chih., 31170, Mexico

1. Introduction . . . . . . . . . . . . . . . 3 3.4.2. Capacity, Pressure Drop . . . . . . . . . 16

1.1. General Remarks . . . . . . . . . . . . 3 4. Effect of Variables on Hydrocyclone
1.2. Principle of Operation . . . . . . . . . 3 Performance . . . . . . . . . . . . . . . 16
2. Categories and Application . . . . . 4 4.1. Operation Variables . . . . . . . . . . 16
2.1. Thickening . . . . . . . . . . . . . . . . 4 4.2. Design Variables . . . . . . . . . . . . . 17
2.2. Clarification . . . . . . . . . . . . . . . 4 5. Selection and Design of Hydrocy-
2.3. Classification . . . . . . . . . . . . . . . 4 clone Systems . . . . . . . . . . . . . . . 17
2.4. Other Applications . . . . . . . . . . . 4 5.1. Analytical Solutions . . . . . . . . . . 18
3. Theoretical Background . . . . . . . 4 5.2. Graphical Solutions . . . . . . . . . . 18
3.1. Simultaneous Flow of Fluids and 5.3. Manufactures’ Choice . . . . . . . . . 18
Solids . . . . . . . . . . . . . . . . . . . . 4
5.4. Dimensionless Scale-Up of
3.1.1. Dynamics of Particles Submerged in
Hydrocyclones . . . . . . . . . . . . . . 19
Fluids . . . . . . . . . . . . . . . . . . . . 5
5.4.1. Introduction . . . . . . . . . . . . . . . . 19
3.1.2. Rheology of Suspensions . . . . . . . . 6
5.4.2. Definition and Derivation of
3.1.3. Flow of Suspensions . . . . . . . . . . . 7
Dimensionless Groups for Hydrocy-
3.2. Flow Patterns in Hydrocyclones . . 8
clones . . . . . . . . . . . . . . . . . . . . 19
3.3. Mechanisms of Particle Separation 9
5.4.3. Scale-Up at Low Concentrations . . . 20
3.3.1. Equilibrium-Orbit Theory . . . . . . . 10
3.3.2. Residence-Time Theory . . . . . . . . . 10 5.4.4. Scale-Up at High Concentrations . . . 21
3.3.3. Crowding Theory . . . . . . . . . . . . . 10 5.4.5. Considerations for non-Newtonian
3.3.4. Turbulent Two-phase Flow Theory . . 10 Behavior . . . . . . . . . . . . . . . . . . 21
3.4. Characteristics of Performance of 5.4.6. Empirical Models . . . . . . . . . . . . . 22
Hydrocyclones . . . . . . . . . . . . . . 11 5.4.7. Practical Applications . . . . . . . . . . 23
3.4.1. Separation Efficiency, Cut Size . . . . 11 6. References . . . . . . . . . . . . . . . . . 26

Notation Do overflow pipe diameter of hydrocyclone

(Dimensions given in terms of mass, M, (L)
length, L, time, T, and temperature,θ) Du underflow pipe diameter of hydrocy-
A arithmetic mean, acceleration (L/T2 ) clone (L)
A area (L 2 ) Ep partial efficiency
Ai area of inlet nozzle of hydrocyclone Et total efficiency
(L 2 ) E
t reduced total efficiency
A1n parameter dependent on the flow behav- Eu Euler number
ior index f fanning friction factor
C volume fraction of solids in suspension f pv purely viscous fanning friction factor
CD drag coefficient f (x) frequency
C
n parameter dependant on the flow behav- F c (x) cumulative percentage oversize of sep-
ior index arated solids
C y50 dimensionless cyclone number FD drag force (MLT−2 )
D pipe diameter (L) F f (x) cumulative percentage oversize of non-
Dc hydrocyclone diameter (L) separated solids
Di inlet diameter of hydrocyclone (L) F(x) cumulative percentage oversize of feed
solids

c 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.c13 c02.pub2
2 Hydrocyclones

g acceleration due to gravity (L/T2 ) Stk ∗50 (r) Stokes number for power-law fluids in-
G(x) grade efficiency cluding the reduced cut size
G
(x) reduced grade efficiency t time (T)
H height of liquid column or liquid head td detention time (T)
(L) ts settling time (T)
Hi height of rectangular inlet channel of T temperature (θ)
hydrocyclone (L) u fluid-particle relative velocity (L/T)
H 25/75 sharpness index ut terminal settling velocity under gravity
kp empirical constant for a family of geo- (L/T)
metrically similar hydrocyclones U volumetric flow rate of underflow sus-
k1, k2 constants pension (L3 /T)
K constant, correlation constant of the v linear velocity, hydrocyclone character-
power law istic velocity (L/T)
K
fluid consistency index (MTn /L2 ) vg terminal settling velocity under gravity
l vortex finder length (L) (L/T)
L hydrocyclone length (L) vi inlet tangential velocity (L/T)
L1 length of cylindrical part of hydrocy- vo maximum tangential velocity (L/T)
clone (L) vr radial settling velocity (L/T)
m exponent, mass (M) vt tangential velocity (L/T)
M mass flows rate of solids in suspension V volume (L3 )
(M/T) w vertical velocity at cyclone wall (1/T)
Mc mass flows rate of separated solids x particle size (L)
(M/T) xa arithmetic mean of particle size (L)
Mf mass flows rate of nonseparated solids xc cut size (L)
(M/T) xg geometric mean of particle distribution,
n exponent; slope of the curve repre- mass median size of particle (L)
senting the power law xm mode of particle size distribution
np empirical constant for a family of geo- xo maximum particle size in sample (L)
metrically similar hydrocyclones xr constant function of particles size range
n
flow behavior index (L)
N revolutions per minute or per second x0 particle size with zero centrifugal effi-
(1/T) ciency (L)
O volumetric flow rate of overflow sus- x 50 cut size (L)
pension (L3 /T) x
50 reduced cut size (L)
P pressure (M/LT2 ) x 98 approximate particle size fully sepa-
q volumetric flow rate per unit area (L/T) rated (L)
Q volumetric flow rate (L3 /T) Z acceleration factor
r parameter function of a log-ln plot in Zi width of rectangular inlet channel of
the coarse region hydrocyclone (L)
R radius of rotation (L) Greek
Rav average radius of rotation (L) α constant
Re Reynolds number γ viscosity coefficient for power-law flu-
Rep particle Reynolds number .
ids (M/LT2−n )
Re∗ Reynolds number for power-law fluids γ shear rate (1/T)
Rf underflow-to-throughput ratio ∆ difference
s parameter function of the slope of the θ hydrocyclone cone angle
log-ln plot in fine size regions µ liquid absolute viscosity (M/LT)
Stk 50 Stokes number µo viscosity of pure solvent (M/LT)
Stk
50 Stokes number including the reduced ρ liquid density (M/L3 )
cut size ρm slurry density (M/L3 )
Stk ∗50 Stokes number for power-law fluids ρs solids density (M/L3 )
σ standard deviation
 Hydrocyclones 3

τ shear stress (M/LT2 ) of a hydrocyclone is shown in Figure 1. As can

τw shear stress at tube wall (M/LT2 ) be concluded, separation is based on difference
φ volume fraction of spheres in suspen- in settling velocities, as in sedimentation. Thus,
sion size, density and shape of particles play a role
ω angular velocity (1/T) important in the operation.

1. Introduction

1.1. General Remarks

Hydrocyclones have been less employed than

centrifuges to develop high centrifugal forces
and perform separation of solids. Although
hydrocyclones were first patented more than 100
years ago, industrial use of them was practically
nonexisting until the late 1940s. Hydrocyclones
were early utilized in the paper and mineral in-
dustries, to become later popular in a larger num-
ber of factories from chemical, cement, and nu-
clear to the food, pharmaceutical, and oil indus-
tries. Although hydrocyclones continue finding
new applications, the dearth of specialized lit-
erature about them is surprising. With only two
known books written in English since the 1980s Figure 1. Diagram of a hydrocyclone
and some attention paid to them in the literature,
there is more to learn about hydrocyclones than
about other solid–liquid separation techniques.

1.2. Principle of Operation

A hydrocyclone is a sedimentation device,

which employs enhanced gravity force to sep-
arate solid from a carrying liquid. This men-
tioned force is in fact the centrifugal field pro-
duced when a fluid suspension is pumped tan-
gentially into a cone-cylindrical body. As a vor-
tex is created, coarse particles move towards the
wall and fines towards the axis. Flow is at first
down the inner wall of the cylindrical section and
the cone, as far as the stagnation point near the
cone apex. There, because the opening is small, Figure 2. Nomenclature and dimensions of hydrocyclone
the downward primary vortex is forced to turn Dc : cyclone body diameter, Do : overflow pipe diameter, Di :
upwards again, thus forming the secondary vor- inlet diameter, [(4HiZ i )/π]1/2 for rectangular inlet chan-
tex. Thus, at the bottom of the cone the flow splits nels, l: vortex finder length, L 1 : cylindrical section length,
L: hydrocyclone length, θ: cone angle, Du : underflow ori-
into two streams, the underflow containing most fice
of the coarse particles, and the overflow with
most of the fines in it. The overflow pipe gener-
Typical sizes of hydrocyclones range from
ally protrudes into the cylindrical body, to form
10 mm to 2.5 m in diameter, capacities of sin-
what is known as the vortex finder. A diagram
4 Hydrocyclones
gle units go from 0.1 to 7200 m3 /h, and pressure
drops vary from 0.34 to 6 bar. The efficiency,
reported as cut size, lies between 2 and 250 µm.
The main advantages of the hydrocyclone are
its simple, compact, and inexpensive construc-
tion, its ease of operation (no moving parts), its
short residence time, and its manageability of
sizes finer than those treated in mechanical clas-
sifiers. Some disadvantages are its high power
consumption, its high wear, its inflexibility and
its problem with blockage, especially in small
units. Figure 2 presents typical dimensions and
standard nomenclature of hydrocyclones.
2. Categories and Application
Hydrocyclones are very versatile pieces of
equipment. They can be employed as thicken-
ers, clarifiers, classifiers and even in liquid–liq-
uid separations. A quite useful and up-to-date
guide of applications and models of hydrocy-
clones is given in [1].
2.1. Thickening
For thickening proposes, hydrocyclones have
proved to be efficient, with underflow concen-
tration as high as 50 % obtained with some ma-
terial. Thus, they can easily compete in this field
with gravity thickeners, with advantages of less
space requirement. The main problem in the op-
eration, however, is the risk of blockage, but
hydrocyclones with variable underflow orifice
are employed to avoid this difficulty.
2.2. Clarification
When used in clarification, small diameter
hydrocyclones with parallel multicyclone ar-
rangement are preferred. Multiple cyclone units
with six to several hundred hydrocyclones in par-
allel are available and have been successfully
used.
2.3. Classification
Since the principle of operation of hydrocy-
clones is centrifugal, they can be used to separate
solids from solids in a suspension according to
particle size. Sometimes, either coarse or fine
particles may be removed from the product, and
hydrocyclones have performed such operations
known as degritting or refining, and desliming
or washing, respectively. This classifying char-
acteristic of hydrocyclones is also employed to
improve the performance of other filtration or
separation equipments. The coarse and fine par-
ticles separated from hydrocyclones normally
follow different paths in a processing plant, be-
ing sometimes fed into other units, to finish as a
different product.
2.4. Other Applications
As hydrocyclones have been used either as thick-
eners or as clarifiers, they have also performed
both functions in one operation. In order to do
so, two or more units are connected in series.
One of them functions as thickener and the oth-
ers as clarifiers. A typical arrangement, which
operates with the thickener at the first stage, fol-
lowed by two clarification stages, and the un-
derflow of them returned to the feed, has been
widely described in [2]. Other arrangements that
have also maximized the overall performance
have been analyzed in [1]. The centrifugal fea-
tures of hydrocyclones give them possibilities to
perform separation of two immiscible liquids.
Hydrocyclones have been applied in separation
of oil from water, in dewatering of light oils
and in producing highly concentrated samples
of lighter dispersed phases. Other applications
such as countercurrent washing of solids and
separation of gas bubbles from feed liquids have
as well been tried, giving satisfactory results.
3. Theoretical Background
3.1. Simultaneous Flow of Fluids and
Solids
As hydrocyclones are usually fed with a disper-
sion of solids in liquid, a study of the principal
characteristics of suspension is necessary. Since
concentration can be vary widely in hydrocy-
clone operations, particle–fluid interaction, as
well as proper rheology of the whole suspension,

are of the utmost importance. For a detailed de-
scription of fluid dynamics and rheology see →
Fluid Mechanics.
3.1.1. Dynamics of Particles Submerged in
Fluids
If a particle moves relative to the fluid in which
it is suspended, the force opposing the motion
is known as the drag force. Knowledge of the
magnitude of this force is essential if the parti-
cle motion is to be studied. Conventionally, the
drag force F D is expressed as:
ρu2
F D = CD A
2
where u is the particle–fluid relative velocity, ρ
is the fluid density, A is the area of the particle
projected in direction of the motion, and C D is a
coefficient of proportionality known as the drag
coefficient. Assuming the drag force is due to the
inertia of the fluid, C D would be constant and di-
mensional analysis shows that C D is generally a
function of the particle Reynolds number, i.e.:
uxρ
Rep =
µ
where x is the particle size and µ the medium
viscosity; the form of the function depends on
the regime of the flow. This relationship for rigid
spherical particles is shown in Figure 3. At low
Reynolds number under laminar flow conditions
when viscous forces prevail, C D can be deter-
mined theoretically from Navier–Stokes equa-
tions (see → Fluid Mechanics, Section 3.2) and
the solution is known as Stokes’ law and repre-
sented by:
FD = 3πµux
Figure 3. Drag coefficient versus particle Reynolds number
for spherical particles
Hydrocyclones 5
This is an approximation, which gives the
best results for Rep →0; the upper limit of its va-
lidity depends on the error that can be accepted.
The usually quoted limit for the Stokes region of
Rep = 0.2 is based on an error of about 2 % in
the terminal settling velocity. Equations 1, 2 and
3 combined give another form of Stokes’ law as
follows:
24
CD = (Rep <0.2) (4)
Rep
For Reynolds numbers > 1000 the flow is
fully turbulent with inertial forces prevailing,
and C D becomes constant and equal to 0.44 (the
(1) Newton region). The region in-between Rep =
0.2−1000 is known as the transition region, and
C D is either described in a graph or by one or
more empirical equations.
For a particle of mass m under the influence of
a field of acceleration a, the equation of motion
is:



du ρ
m = ma−ma −FD (5)
dt ρs
(2) where ρs is the density of the solids and t is the
time.
In applications of hydrocyclones concerned
with separation of fine particles, which are all
most difficult to separate, the Reynolds numbers
are low, often less than 0.2, due to the low val-
ues of x and u. Therefore, it is reasonable also
that the time necessary for the particle velocity
to reach values very close to the terminal set-
tling velocity is very short and being the field of
acceleration gravity, Equation (5) can be solved
to give:
(3)
x2 (ρs −ρ) g
ut = (6)
18µ
in which the acceleration due to gravity g has re-
placed a and ut is known as the terminal settling
velocity under gravity.
The radial settling velocity in a hydrocyclone
vr is due to the centrifugal acceleration, which
is proportional to the square of the tangential
velocity of the particle, and indirectly propor-
tional to the radius of the particle position. As
the tangential motion of particle is unopposed,
the tangential particle velocity can be taken as
equal to the tangential component of the fluid
6 Hydrocyclones
velocity at the same point. The inclusion of the
centrifugal terms in Equation (6) leads simply
to:
x2 −ρ) Rω 2
(ρs
vr = (7)
18µ
where R is the radius of the particle position and
ω is the angular velocity.
As the concentration of the suspension in-
creases, particles get closer together and inter-
fere with each other. If the particles are not
disturbed uniformly, the overall effect is a net
increase in settling velocity because the return
flow caused by volume displacement predom-
inates in particles-sparse regions. This is the
well-known effect of cluster formation, which
is significant only in nearly monosized suspen-
sions. With most practical widely dispersed sus-
pension, clusters do not survive long enough
to affect the settling behavior and, as the re-
turn flow is more uniformly distributed, the set-
tling rate steadily declines with increasing con-
centration. This phenomenon is referred to as
hindered settling and can be theoretically ap-
proached in three different manners: as a Stokes’
law correction by introduction of a multiply-
ing factor; by adopting effective fluid properties
for the suspension different from these of the
pure fluid; and by determination of bed expan-
sion with a modified version of the well know
Carman–Kozeny equation. All the approaches
can be shown to yield essentially identical re-
sults. Some important correlations accounting
for the hindered settling effect have been re-
viewed in [3]. It was shown that their differences
are minimal and, according to this, the Richard-
son and Zakii equation is an obvious choice in
practice. Such relation can be expressed as:
u
= (1−C)4.65 (8)
ut
where u is the settling velocity at concentration
C and ut is the settling velocity of a single par-
ticle.
The relationship above applies only to free,
particulate separation unaffected by coagulation
or flocculation, and where all particles are of uni-
form density.
3.1.2. Rheology of Suspensions
The rheological properties of suspension have
been studied since the beginning of the 20th
century. The first research was done by Ein-
stein[4] in his classical study of the viscosity
of dilute suspensions of rigid spheres. His ap-
proach is purely hydrodynamic, and his model
consists of an isolated sphere situated in a simple
shear flow field in an infinite fluid. A number of
workers have been extending Einstein’s analysis
[3, 5]. Three approaches have been presented:
the theoretical basis of viscosity computation,
the effect of particle texture and shape, and the
effect of concentration. Several relations have
been developed for suspensions in with nearest
neighbor interactions cannot be neglected. Guth
and Simha [6] considered the first order effect of
spheres interacting with one another. They found
that:
µ
= 1+2.5ø+14.1ø2 (9)
µo
where µ is the viscosity of the suspension, µo
is the viscosity of the pure solvent and φ is the
volume fraction of spheres in the suspension. A
later correction gave a value of 12.6 for the last
constant [7].
Several authors have made experimental
studies of the effects of concentration on the
viscosity of suspension of spheres. The data
they have proposed are considerably scattered.
Thomas [8] has tried to find the sources of scatter
in the data and found that scattering is caused
in large part by variations in particle size. For
small particle sizes (diameters < 1–10 µm) col-
loidal forces become important and the viscos-
ity begins to increase as particle size decreases.
The viscosity is also shear rate-dependent in this
case. For the particles larger than 1 to 10 µm,
the particle Reynolds number becomes signifi-
cant. The inertial effects results in an increase in
relative viscosity with increasing particle size.
Thomas has developed a unique curve, by elim-
inating diameter and shear rate effects. For low
concentration data, such a graph can be repre-
sented by:
µ
= 1+2.5ø+11.4ø2 (10)
µo
where all the components are already defined.
The behavior of some specific systems has
also been reported in the literature. Lin and
Brodkey [9] studied rheological properties of

slurry fuel, concluding that they are shear-
thinning under most conditions, but turn shear-
thickening at extreme conditions. Griskey et al.
[10] investigated the rheological behavior of
cornstarch dispersions and found that they go
from dilatant (shear-thickening) behavior at low
shear rates to Newtonian behaviors at higher
shear rates. The effect of particle aggregation
on suspension has also been studied. Devries
[11] presented a number of experimental results
concerning the particle aggregation in polymer
lattices induced or accelerated by simple shear
flow. No stable flow is possible under these con-
ditions, and the apparent viscosity is an increas-
ing function of the time of shearing.
3.1.3. Flow of Suspensions
Some flow patterns of relative interest will be re-
viewed. One of the most studied cases is that of
transport of suspensions such as slurries through
pipelines, due mainly to its occurrence in a wide
number of industrial processes. Most of the the-
ory developed for flow through tubes applies to
laminar flow in smooth tubes. There are some re-
lationships available for pressure-loss flow rate
in purely viscous and viscoelastics flows. Re-
lations for thixotropic and rheopectic systems
appear not to be available.
Under the assumptions that purely viscous
behavior prevails and that no slip occurs at the
tube wall, the power-law equation for laminar
flow is in the form [12]:

n

D∆P 8v
τw =
4L
= K
D
(11)
where τw is the shearing stress at the wall of the
tube, D is the tube diameter, ∆P is the pressure
drop, L is the length of the tube, v the mean ve-
locity, n’ is the slope of the line when the data
are plotted on logarithmic coordinates. For n’ =
1, the fluid is Newtonian; for n’ <1, pseudoplas-
tic, or Bingham plastic if the curve does not go
through the origin; and for n’ >1, dilatant. The
term K’ is the consistency index; as the name
suggest, the larger its value the thicker or more
viscous the fluid.
The constant K’ may be related to the anal-
ogous power-law constant K (see → Fluid Me-
chanics, Section 4.1.2) as follows [13]:

n
n−n

3n
+1 8v
K
= K (12)
4n
D
Hydrocyclones 7
if the fluid obeys the power-law, n = n’ and:

n
3n+1
K
= K (13)
4n
Since Equation (11) rigorously portrays the
laminar flow behavior of the fluid (provided n’
and K’ are evaluated at the correct shear stress),
one may use it to define a Reynolds number ap-
plicable to all purely viscous fluids under lam-
inar flow conditions. This dimensionless group
can be derived simply by the substitution of
(D∆P/4L) from Equation (11) into the usual
definition of the fanning friction factor, i.e.:
D∆P/4L
f = (14)
ρv 2 /2
such substitution leads to:
16γ
f = (15)
Dn
v 2−n
ρ


where γ = K
8n −1 and all the remaining com-
ponent as already defined.
By letting f = 16/Re as for Newtonian flu-
ids in laminar flow, the above mentioned gener-
alized Reynolds number can be obtained as:



Dn v 2−n ρ
Re∗ = (16)
γ
If the equation is desired in terms of K in-
stead of K’, Equation (12b) may be substituted
into Equation (15) and:
Dn v 2−n ρ
Re∗ =  n (17)
K8n−1 3n+1
4n
For Newtonian fluids n
= 1 and K
= µ,
so γ reduces to µ and Re∗ in Equation (16)
transforms to the familiar Dv/ρ/µ showing that
this traditional dimensionless group is merely a
special restricted form of the more general de-
scribed here.
Non-Newtonian fluids in turbulent flow gen-
erally show lower friction factors and, conse-
quently, lower pressure drops than Newtonian
fluids at corresponding Reynolds number. A de-
tailed analysis of non-Newtonian flow at tur-
bulent regime began with two papers published
in 1959 by Dodge and Metzner [14] and Shaver
and Merril [15]. Dodge and Metzner used solu-
tions of carboxymethyl cellulose and carbopal (a
B. F. Goodrich soluble thickener), as well as clay
8 Hydrocyclones
suspensions. They presented a viscous correla-
tion in terms of parameters from shear-stress–
shear-rate data and concluded that except for the
carboxymethyl cellulose, the behavior of these
materials could be correlated with the following
generalization of the Newtonian friction factor:
 case, the viscosity of the suspension has been
1  
= A1n log Re∗(f )1−0.5n
+Cn
∗
(18)
f tion factors.
where A1n and Cn
are parameters dependent
on the flow-behavior index n. While the choice
of power-law equation in the derivation requires
that the Reynolds number used in Equation (17)
be the power-law special case of Equation (16),
it has been shown [14] that any errors due to
this approximation are less important under tur-
bulent flow conditions than in the laminar region.
The utility of Equation (17) and of Reynolds
number outside the laminar region has been
shown diagrammatically in [14]. The authors
found excellent agreement between experimen-
tal and extrapolated data. Furthermore, devia-
tions from smooth curves were no greater for
those fluids, which did not obey the power law,
than for those which did. The parameters A1n
and Cn
in Equation (17) must be evaluated em-
pirically, just as in the simpler case of Newtonian
behavior.
By far, most of the study of the flow behav-
ior of non-Newtonian suspensions has been de-
voted to monosized and monoshaped systems,
mainly of spherical particle geometry. In real-
ity, however, solid–liquid mixtures usually con-
sist of dissolved solids with a variety of particle
sizes, shapes, and concentrations. The available
theory for the behavior of non-Newtonian sus-
pensions is, therefore, of a limited application,
and when dealing with real systems, empirical
corrections are necessary to fully explain the sus-
pension properties.
Studies of some real applications such as
slurry transport are also encountered in the liter-
ature. The slurry to be transported may have set-
tling or nonsettling characteristics. Sand in wa-
ter, salt in benzene, and clay in drilling mud are
typical examples of settling slurries, while pa-
per pulp is a good example of nonsettling slurry.
The law governing the settling rate of particles at
low Reynolds numbers is the well know Stokes
law. Settling phenomena of slurries in pipelines,
as well as modifications of original Stokes law
for flocculated suspensions and particles of odd
shapes, have been studied by Thomas [16]. The
turbulent regime is sometimes preferred to trans-
port slurries within pipe in order to prevent set-
tling, and to provide high throughputs. In this
normally used to correlate the turbulence fric-
The problem of distribution and mixing of
solids in turbulent slurry flow has been stud-
ied. The magnitude of flow velocity or Reynolds
number required for maintaining turbulent mix-
ing of solid–liquid slurries, may be computed
from the following equation:


ρv 2 ρm vD
= K 0.775 (19)
(ρs −ρ) gx µ
where ρ is the liquid density, ρs the solid density,
ρm the slurry density, v the minimum standard
velocity, g the acceleration due to the gravity, D
the mean particle diameter, µ the viscosity and
K’ a constant. When densities of solids and liq-
uids are known, along with the volume fraction
of solids in the slurry, the density of the mixture
may be calculated by:
ρm = Cρs + (1−C) ρ (20)
where C is the volume fraction of solids in the
slurry. Equation (18) is recommended for mean
diameters between 50 and 500 µm.
3.2. Flow Patterns in Hydrocyclones
The flow pattern in a hydrocyclone is com-
plex. However, it is normally accepted that the
hydrocyclone fluid flow consists of the three ve-
locity components: tangential, vertical, and ra-
dial. Figure 4 shows a diagrammatically repre-
sentation of these three velocity distributions
within a hydrocyclone. With respect to the tan-
gential component vt , the linear velocity of the
feed stream at the inlet point is given by:
Q
vt = (21)
Ai
where Q is the volumetric flow rate and Ai is the
area of the inlet nozzle.

Figure 4. Velocity distribution profiles within a hydrocy-
clone
As the flow develops in a spiral manner to-
wards the air core, the circular speed increases as
the diameter of rotation decreases. Under no fric-
tion loss conditions, the angular speed remains
constant and is given by:
vt R = constant
where R is any radius of rotation.
In practical systems frictional losses result in
a progressive decrease in tangential speed, thus
Equation [21] needs to be modified, e.g.:
vt Rn = constant
where: 0<n<1.
Experience has shown that the value of n de-
creases with increasing solids contents, with a
maximum value of about 0.3 for solids contents
of about 60 wt %. For normal operating con-
ditions of 15 to 25 % of solids approximately,
Trawinski[17] has suggested an average value
of 0.5.
The maximum tangential velocity will occur
in the secondary vortex immediately adjacent to
Hydrocyclones 9
the air core. Taking the proposed average value
of 0.5, this maximum velocity will be theoreti-
cally described by:

1/2
Di
vo = vt (24)
Do
Practically, the limit on the maximum tangen-
tial velocity is given by pressure drop across the
hydrocyclone. Under ideal conditions of no fric-
tion loss, the following relationship describes
this maximum velocity:

1/2
2∆P
vo = (25)
ρm
where ∆P is the pressure drop and ρm is the
density of the slurry. As can be seen from Figure
4b, the profile of the vertical velocity component
shows a well-defined position of zero velocity
outside of which there is a low speed down-
wards, and inside of which the flow is much
higher and upwards. This locus of zero verti-
cal velocity (LZVV) follows the profile of the
cyclone.
The radial component of velocity is inward,
and accurate measurement of it is difficult be-
cause its magnitude is small. Apparently, such
speed decreases proportionally to radius de-
creasing, from a value that can be expressed by:


θ
vr = wtan (26)
2
where w is the vertical velocity at the wall and
θ is the angle of the cone.
The radial position at which this velocity falls
(22)
to zero is not known precisely. Figure 4c shows
the radial component of velocity.
3.3. Mechanisms of Particle Separation
According to the flow patterns just described,
(23)
several models to explain the actual mecha-
nism of particle separation have been devel-
oped. Each model works well for particular
hydrocyclones only, the models are not of gen-
eral application, because every one relies on a
single dominant feature of the flow inside the
hydrocyclone. There have been at least four ap-
proaches to describe the behavior of particles
in a hydrocyclone: the equilibrium-orbit theory,
the residence-time theory, the crowding theory
and the turbulent two-phase theory.
10 Hydrocyclones

3.3.1. Equilibrium-Orbit Theory

9µL
x2 = (32)
An equilibrium orbit is proposed at a radius posi- N πv (ρs −ρ)
tion, at which the inward drag on the particle due This is a form similar to Equation 29 without
to the inward radial velocity of the fluid counter- the dependence on radial positions.
balances the outward force to the liquid rotation, Using direct analogy with gravity settling,
i.e.: Trawinski [22] proposed another simple ap-
πx3 v2 proach, which may be considered as a variation
(ρs −ρ) t = 3πµvr x (27)
6 R of the residence-time theory. He reported the fol-
or: lowing equation:

1/2
x2 (ρs −ρ) Rvr 18µq
= 2 (28) xc = (33)
18 µ vt g (ρs −ρ) Z

where x is the particle diameter or size, and the
densities and velocities are as previously de-
fined. If vt ∼ (v/Rn ) and vr ∼ (v/Rm ), where
v is the inlet velocity, then:



x2 ρs −ρ K

= R1+2n−m (29)
18 µ v
For given inlet velocity and liquid, the radius
of the orbit is proportional to the particle size. If
R is greater than the radial position for zero ver-
tical velocity, it is assumed that all particles will
be collected. If R is less than the radial position,
all particles will be carried away in the under-
flow. Thus, a particle will have a 50 % chance
of removal if the radius of its equilibrium orbit
is equal to that of the zero vertical velocity. Us-
ing this approach, Bradley [7] derived a relation
making assumptions to the position of the locus
of zero vertical velocity profile.
3.3.2. Residence-Time Theory
This theory advocates the importance of the time
spent in the hydrocyclone. Assuming the spiral
has a constant velocity equal to the average in-
let velocity, and supposing the liquid follows N
complete spirals at an average radius Rav , the
residence time will be:


2πRav
t= N (30)
v cording to this, it is proposed that separation
a particle of size x will travel a distance L in this
time t where:


L x2 (ρs −ρ) v2
=v= (31)
t 18µ R
Combining Equations 30 and 31 the following
relationship is obtained:
where x c is known as the cut point, q equals
(Q/A) and Z is the acceleration factor. Including
further considerations, e.g., for capacity and the
separating area (surface of the cylinder formed
by the air core), Trawinski [17] also proposed:
1/2
1/2
18µ Dc Di
xc = k 1 (H)−0.25 (34)
(ρs −ρ) g l
where k 1 is a constant, H is the liquid head, and
all the geometric characteristics are as defined
in Figure 2.
3.3.3. Crowding Theory
Due originally to Fahlstrom [18], this theory
proposes that the cut size is a function of the
capacity of the underflow orifice and the par-
ticle size distribution. The crowding effect can
swamp the primary interaction to the extent that
the cut size can be estimated from the mass re-
covery to the underflow. The support to this the-
ory has been given mainly by Bloor And Ing-
ham [19], by means of the mathematical mod-
els developed to describe the fluid mechanics of
hydrocyclones.
3.3.4. Turbulent Two-phase Flow Theory
This theory states the importance of the effect
caused by turbulent flow in hydrocyclones. Ac-
takes place under the influence of the centrifu-
gal field and the turbulent transport in the two-
phase flow. The influence of turbulence can be
described by an eddy diffusion coefficient equal
to, or smaller than the diffusion of the fluid parti-
cles. The assessment of the influence of such dif-
fusion on separation is very difficult in practice.

Bloor and Ingham [19] have also contributed to
the theoretical background of this theory. They
derived an analytical form of the radial velocity
and found an analytical solution to the equation
of motion for the tangential speed. They also
developed a more elaborated and physically re-
alistic model, for the eddy viscosity based on the
Prandtl mixing theory, which greatly simplified
the equation of motion.
Schubert and Neesse [20] proposed a set of
equations and derived two models for turbulent
cross-flow wet classification: the suspension-
partition model and the suspension-tapping
model. In the former the flow is divided into
overflow and underflow without changes in the
total cross-section, whereas in the latter over-
flow and underflow are tapped from the main
flow through small outlet openings. According
to them, the suspension-tapping model is analo-
gously applicable to the separation in hydrocy-
clones.
3.4. Characteristics of Performance of
Hydrocyclones
The most important features of performance of
hydrocyclones are the particle cut size, the pres-
sure drop and the volume throughput. Previous
work has established the cut size as the best way
to predict efficiency in hydrocyclones. In order
to understand and utilize the cut size concept, it
is necessary to deal with at least two topics, i.e.,
particle size analysis and efficiency of solid–liq-
uid separation (see → Particle Size Analysis and
Characterization of a Classification Process; →
Solid–Liquid Separation, Introduction).
3.4.1. Separation Efficiency, Cut Size
The solid phase of suspensions to be treated
in solid–liquid separation equipment generally
consists of an immense number of particles of
diverse sizes and shapes. All this population of
particles needs to be identified or characterized.
The frequency of occurrence of particles of ev-
ery size present, arranged and presented in a sta-
tistical manner, is known as the particle size dis-
tribution. The most common way of presenting
such distributions as well as the types of distri-
butions important to particles technology, can be
Hydrocyclones 11
found in the specialized literature [21, 22] (see
→ Particle Size Analysis and Characterization
of a Classification Process).
In solid–liquid separation it is important to
know the particle size distribution in order to
identify which part of it will be separated, and
transform this quantity in a measure of efficiency
(→ Particle Size Analysis and Characterization
of a Classification Process, Section 2.1). The
common way of presenting particle size data for
solid–liquid separation purposes is in form of
a plot. The most common ones are equivalent
sphere diameters, equivalent circle diameter, and
statistical diameters. The equivalent sphere di-
ameter is the diameter of a sphere which has
the same property as the particle itself. Such a
property could be the settling velocity. A diam-
eter derived from the settling velocity is known
as Stokes diameter and is a very useful quantity
for solid–liquid separation, especially to those
techniques in which the particle motion relative
to the fluid is the governing mechanism.
The amount of particle matter which belongs
to specified size classes on the particle axis may
be also represented in several ways. Number of
particles and mass of particles are the most com-
mons ones, but surface area and volume are used
as well. For solid–liquid separation techniques,
the most convenient form of expressing amount
of particle matter is by mass, because the bal-
ances necessary to define performance are nor-
mally mass balances.
In general, particle size distributions can be
presented as frequencies, f(x), or cumulative fre-
quencies, F(x), which are related to each other
by following equation:
dF (x)
f (x) = (35)
dx
the graphical representation of a particle size dis-
tribution is usually plotted in a cumulative form.
In a typical cumulative plot, points are entered
showing the amount of particulate material con-
tributed by particles below or above a specified
size. Hence, the curve presents a continuously
rising or deceasing character. These oversize and
undersize distributions, as illustrated in Figure
5, are simply related by:
F (x)oversize = 1−F (x) (36)
where F(x) is the cumulative fraction undersize.
12 Hydrocyclones
x

Figure 5. Relationship between frequency and cumulative
distributions
A cumulative plot will, therefore, include a
broad range of particle sizes. It is often conve-
nient, however, to refer to a single characteristic
size for system. Many characteristic sizes have
been proposed, most of them involving a math-
ematical formula. An important one, which can
be read off any cumulative plot of the particle
size data, is the median particle size. It is de-
fined as that particle size for which the particle
amount equals 50 % of the total. If the parti-
cle size is represented by number, such point is
called the number median size. If mass is used as
the measure of particle amount, this parameter
is known as the mass median size. The distribu-
tion between number and mass median is very
important, since they differ generally by a con-
siderable amount. Such difference means that
number and mass cumulative plots do not agree
for the same system of particles. The weight of a
particle, which varies as the cube of its diameter,
accounts for this mentioned disagreement.
For practical purposes, it is reasonable to fit
an analytical function to experimental particle
size distribution data, and then handle this func-
tion mathematically in further treatment. It is,
for example, very much easier to evaluate mean
sizes from analytical functions from experimen-
tal data. Several different distribution functions
can be found in the literature (→ Particle Size
Analysis and Characterization of a Classifica-
tion Process, Section 2.3). An exhaustive review
of some of them has been presented [23].
The normal distribution is widely used in
most fields of science for the representation of
populations. The normal distribution function is
given by:
where y is the probability density, x the diameter
of the particle, a the arithmetic mean, and σ the
standard deviation. From the normal function,
if the arithmetic mean a is zero, the probability
of occurrence within the interval form the mean
(zero) to the value x, i.e., F(x) is given by the
error function:
1 x2
F (x) = √ exp − 2 dx (38)
σ 2π 2σ
0
where σ and x have the same meaning as de-
scribed in the normal function.
Materials with a normal distribution of parti-
cle size are relatively rare and are found chiefly
among the particulates produced by chemical
processes like condensation or precipitation.
The importance of this function, however, is that
it provides an idealized error distribution built
upon the assumption that elementary errors or
small causes combine at random to produce the
observed effect.
The log-normal distribution is probably the
most widely used type of function; it is again a
two-parameter function, it is skewed to the right
and it gives equal probability to ratios of sizes
rather than to size differences, as in the normal
distribution. Furthermore, the log-normal func-
tion is the most useful one among the different
types of functions. It can be expressed in the
following form:

1 (lnx−lnxg )2
f (x) = √ exp − (39)
xlnσg 2π 2ln2 σg
where f(x) is the size distribution function for
particle size x, x g is the geometric mean of the
distribution, and σg the geometric standard de-
viation of lnx.
The log-normal distribution has the important
advantage of ease of conversion from one type
of size distribution to another. It can be shown
mathematically that all four particle size distri-
butions, by number, length, surface, and volume
(mass), when plotted on log-probability paper
are parallel lines with equal linear spacing.
The Rosin–Rammler distribution function is
another equation widely used in particle size
measurement. It is a two-parameter function,
usually given as cumulative percentage oversize:

 n 
1 (x−a)2 x
y= √ exp − (37) F (x) = exp − (40)
σ 2π 2σ 2 xr
 Hydrocyclones 13

where x r is a constant giving a measure of the with particles settling by gravitational forces, so
particle size range present and n is another con- techniques using this sort of forces to evaluate
stant characteristic of the material under analysis particle size distribution would be most appro-
and gives a measure of the steepness of the cu- priate.
mulative curve. The frequency distribution is ob- To evaluate the efficiency of separation it is
tained by differentiation of Equation 40, which necessary to take into account that solid–liq-
can be reduced to: uid separation is basically an imperfect process.

 Whereas the underflow is always wet slurry, the
1
log ln = nlogx−nlogxr (41) overflow can be considered as a turbid liquid.
F (x)
This general behavior is critic in hydrocyclones,
which gives a straight line if log (ln 1/F/ (x))
as they always, despite the presence of solids in
is plotted against lnx. This is the basis for the
the liquid, will split the flow into two streams.
Rosin–Rammler graph paper because it is the
Therefore, the aim of operation of a hydrocy-
size corresponding to 100/e = 36.8% and n
clone will be not only to obtain the above-
is the slope of the line. Rosin–Rammler charts
mentioned slurry, but also to deliver is as free
from German sources contain edge scales, which
from liquid as possible. Equally, for clarification
with the aid of a parallel rule allow direct estima-
purposes, the overflow should be drawn from the
tion of n. For accurate work, the use of large sizes
hydrocyclone with the least possible concentra-
(32.4 × 25.4 cm) charts designed by Harris [24]
tion of solids. As Tenberger and Rietema [25]
is recommended. The various mean sizes avail-
have stated, a single efficiency number can never
able may be easily calculated from the Rosin–
be capable of fully describe the result of separa-
Rammler equation; they all involve mathemati-
tion, except when it is ideal.
cal tables.
The imperfection of solid–liquid separation
The Harris model is a three-parameter equa-
has caused the need to express efficiency by dif-
tion, which is very versatile and fits many empir-
ferent means. A good review of these mentioned
ical size distributions. Harris [24] showed that
efficiencies is presented in [26].
most of the widely used two-parameter equa-
The first and most obvious definition of sep-
tions are in fact special cases of the equation:

 s r
x
F (x) = 1− (42)
xo
where F(x) is cumulative percentage oversize,
x o is the maximum size in the sample, s is a pa-
rameter concerned with the slope of the log–ln
plot in the fine region and r is a parameter con-
cerned with the shape of the log–ln plot in the
coarse region.
The ways for measuring particle size are di-
verse and have been largely improved in re-
cent years. Sieving and sedimentation are the
traditional ones, but microscopy, laser diffrac-
tion, and stream scanning techniques are cur-
rently spreading and finding applications. Sev-
eral methods for analytical determination of par-
ticle size analysis have been reviewed [23]. Al-
though the most novel techniques have the ad-
vantage of quick response, they have proved to
be very selective. Therefore, most of them are
of quite limited application. A rule of thumb
should be to use the analytical technique whose
indirect measurement agrees, as much as pos-
sible, with the specific application. For exam-
ple, in solid–liquid separations one is concerned
aration efficiency is simply the overall mass re-
covery as a fraction of the feed flow rate. Ac-
cording to Figure 6:
Mc
Et = (43)
M
where all the components are as defined in Fig-
ure 6.
If there is no accumulation of solids in the
separator then:
M = Mc +Mf (44)
and there is a choice of three possible combi-
nations of the material streams for the total ef-
ficiency testing. It can be shown [3] that if all
the operating conditions are equal, the most ac-
curate estimation of the local efficiency comes
from the two leaving streams.
As has been already mentioned, hydrocy-
clones act as flow dividers. Thus, they also split
the solids in at least the same ratio, as is the ratio
of underflow to throughput, Rf :
U
Rf = (45)
Q
14 Hydrocyclones


Mc
G(x) = (48)
M x

The grade efficiency has become a very use-

Figure 6. Schematic diagram of a separator. M: mass flow
rate of solids in the feed, M c : mass flow rate of separated
solids, M f : mass flow rate of nonseparated solids, F(x): cu-
mulative percentage oversize of feed solids, F c (x): cumula-
tive percentage oversize of separated solids, F f (x): cumula-
tive percentage oversize of nonseparated solids, Q: volumet-
ric flow rate of feed suspension, U: volumetric flow rate of
underflow suspension, O: volumetric flow rate of overflow
suspension
ful definition, because most industrial powders
consist of an infinite number of differently sized
particles. Thus, a single particle size really corre-
sponds to a range of particles having almost sim-
ilar sizes. Therefore, the grade efficiency of most
separation equipments is a continuous function
of x. This function is seldom expressed analyti-
cally but graphically. An S-shaped curve is usu-
ally obtained for separators like hydrocyclones,
in which inertial or gravity body forces perform
the separation.
As the value of the grade efficiency has the
character of probability, plotting the probability
for any given size fraction against particle size
would give a curve as known in Figure 7. This
sort of curve is normally called a grade efficiency
curve.

The total efficiency defined in Equation 43

will thus include the effect of this flow split-
ting known as dead flux. For this reason, a better
way of expressing efficiency of separators such
as hydrocyclones, is by means of a reduced ef-
ficiency which has been defined as:
Et −Rf
Et =
1−Rf
If complete separation is achieved, providing
that there is no accumulation of solids, the value
of the total efficiency from Equation 43 will be
the unit, and so will be the value of the reduced
total efficiency. If, on the other hand, there is no
separation at all, E t will equal Rf in order to sat-
isfy the requirements of an efficiency definition,
leading to a value of zero in Equation 46.
The total efficiency defined by Equation 43
includes all particle sizes present in the feed
solids. If only a narrow range of particle sizes
is of interest, another efficiency of separation
particular to that range can be defined. A math-
ematical expression of such partial efficiency is:


Mc
Ep =
M x1 /x2
where x 1 and x 2 represent the particle size limits
of a definite range.
If the particle size range in Equation 47 be-
comes infinitesimal, the obtained efficiency cor-
responds to a single particle size x and it is known
as the grade efficiency, defined by:
(46)
Figure 7. Grade efficiency and reduced grade efficiency
curve for hydrocyclones
A grade efficiency curve for hydrocyclone is
derived from screen analysis data on the feed,
overflow and underflow streams. As mentioned
before, it is practically impossible to fraction
common systems of particles in a finite num-
ber of sizes. It can be shown [27] that a practi-
cal grade efficiency curve is really derived from
a stepwise calculation, drawing a line through
the midpoints of size intervals. A curve thus de-
rived does not pass through the origin. This can
(47)
be explained bearing in mind that, as has been
previously emphasized, a hydrocyclone is a flow
divider, so the underflow always contains a cer-
tain quantity of very fine particles which simply
follow the flow, and are split in the same ratio
as the liquid. The apparently finite efficiency for

fine particles is, therefore, equal to the underflow
to throughput ratio Rf as shown in Figure 7. A
reduced grade efficiency, similar to the reduced
efficiency represented by Equation 46 can then
be obtained as:
G(x)−Rf
G
(x) = (49)
1−Rf
As can be seen in Figure 7, the correspondent
curve to this definition does cross the origin.
Hydrocyclones are among those equipments
whose separation performance is highly depen-
dent on particle size. Thus, according to this, an
overall efficiency of separation is not the best
way to evaluate their performance. For this rea-
son, it is advisable to express their efficiency for
each size of particle in the range measured. As
experience has shown, the only single number,
which in some way gives the separation power
of a hydrocyclone, is the cut size x 50 (for a def-
inition of cut size see → Particle Size Analy-
sis and Characterization of a Classification Pro-
cess, Section 3.1.4). All particles above the cut
size will be generally going to the underflow,
whereas those below the cut size will be nor-
mally reported to the overflow.
As can be concluded, the cut size is derived
from a grade efficiency curve. As the calculation
and plotting of a full grade efficiency curve is a
time-consuming task, alternative ways of deter-
mining the cut size have been proposed [28, 29].
The sharpness of separation can be related
to the shape of the curve in a number of man-
ners. Knowing the grade efficiency curve, com-
parisons of classification sharpness of different
equipments can be made by plotting G(x) against
the dimensionless size x/x 50 , and comparing the
resultant normalized curves. To define the curve
steepness, a ratio of two sizes corresponding to
two different percentages on the grade efficiency
curve on either side, and 50 % equidistant, can
be used. This parameter is called the sharpness
index and is represented by:


x25
H25/75 = (50)
x75
It was previously stated that the grade effi-
ciency curve does not go through the origin. The
cut size as has been defined is derived from this
curve. If, in order to practically assess the per-
formance of hydrocyclones, a reduced grade ef-
ficiency curve is used, the particle size which
Hydrocyclones 15
gives a 50 % efficiency in such a curve is called
the reduced cut size and is represented by x
50 .
The maximum obtainable efficiency related
to particle size would be that minimum particle
size with 100 % probability of being reported to
the underflow. Graphically, by extrapolating the
end part of the curve to the horizontal axis, such
size will be obtained. It has been proved that
in practice, the maximum of the efficiency is
around 98 %, and the minimum size correspon-
dent to this efficiency is represented by x 98 and
known as the approximate limit of separation.
Several correlations have been proposed in
literature to evaluate the cut size. They are ei-
ther empirical or theoretical, and are expressed
in the traditional British System of units, as well
as in SI units, even in both. Perhaps the most
important criterion for classifying the proposed
correlation for cut size, is by dividing them into
two groups considering whether the feed con-
centration has been taken into account. The main
correlations, which do not consider the feed con-
centration as a variable, will be reviewed as fol-
lows.
Bradley [2] proposed:

1/2
n

18πµ(1−Rf ) 2.3Do Di2
x50 = (51)
16LQ(ρs −ρ) Dc α
where Q is the flow rate, Do , Dc and Di are the
dimensions defined in Figure 2, and (and n are
empirical constants depending on the cyclone
design and the fluid properties.
Rietema [30] derived a dimensionless, char-
acteristic cyclone number C y50 represented by:
L∆P
Cy50 = x250 (ρs −ρ) (52)
µρQ
where ∆P is the pressure drop. Cy50 is deter-
mined experimentally and is not directly related
to the more important hydrocyclone variables.
However, Rietema has demonstrated the use-
fulness of the parameter and quotes a value of
3.50 for optimal separation, corresponding to
the conditions: L/Dc = 5, Di /Dc = 0.28,
Do /Dc = 0.34 and l/Dc = 0.4, where all the
dimensions are as defined in Figure 2.
Dahlstrom [31] proposed:


C(Do Di )0.68 1.73 1/2
x50 = 0.53
(53)
Q ρs −ρ
where C is a constant sensitive only to major
changes in the cyclone dimensions. This equa-
tion holds well for cyclones in which the length
16 Hydrocyclones
of the cylindrical section equals that of the cone,
and has been shown to hold for diameters of 3
to 14 inches (7.5 to 35 cm), for flows containing
up to 20 % solids by weight.
Yoshioka and Hotta [32] developed an equa-
tion based on the orbital concept and equilibrium
cone surface defined by the end of the vortex
finder and the cone apex. The equation is:

1/2
µ =
x50 = 0.2Dc0.1 Di0.6 Do0.8
Q(ρs −ρ)
which is expressed in SI units; the constant 0.2
was experimentally evaluated.
De Gelder [33] has derived an equation,
which does not include experimental values:



0.349Re(Dc −kDo ) ρs
x0 = Do −1
Di ρ 4. Effect of Variables on
where Re is the Reynolds number, x 0 is the di-
ameter of particle which is just not collected, i.e.,
with centrifugal efficiency of 0 % and k is a con-
stant, function of Di2 /Do Dc . Although Equation
55 was first developed for gas cyclones, it is said
to hold well for hydrocyclones.
Lynch and Rao [34] presented the following
empirical correlation:
Do Du ∆P Qlo
logx50 = − + − +K
2.6 3.5 10.7 52
where x∗50 is in micrometers, Do and Du are in
inches, ∆P is in pounds per square inch (psi),
Qlo (mass flow rate of the liquid in the overflow)
is in ton/h, and K is a function of the material
and the hydrocyclone (2.0 for silica and 2.5 for
copper ore).
3.4.2. Capacity, Pressure Drop
The capacity of a hydrocyclone can be evalu-
ated in terms of the volume flow rate delivered,
also known as throughput, which depends on
the available pressure drop. Thus, there is a loss
of pressure across the unit. The pressure drop
is an important variable and is more easily de-
fined than efficiency. Relationships between Q
and ∆P can be developed by means of estab-
lished theory for the flow of liquids in pipes.
There are a number of pressure drop correla-
tions reported in the literature, some of which
will be described.
Trawinski [35] proposed the correlation:

1/2
∆Pg
Q = KDi Do (57)
ρ
where K is a factor, which contains diameter ra-
tios, fraction loss, and cone angle variables. For
θ = 15◦ −30◦ , K equals 0.5.
Bradley’s approach [2] leads to the follow-
ing expression:
(∆P/ρ) α2  
(Dc /Do )2n −1 (58)
(54) 2
(vi /2g) n
where the left-hand side group is the pressure
difference in terms of inlet velocity heads and α
is the inlet velocity loss coefficient vc /vi whereas
n is the power from Equation 23.
1/2
(55)
Hydrocyclone Performance
A single conclusion can be made from the pre-
viously survey: the operation of hydrocyclones
is quite complex. There is not a unified theory to
fully explain the mechanisms of flow and parti-
cle separation inside a hydrocyclone. It is nor-
mally accepted that the overall behavior of the
particles of solids inside a unit will depend on
factors contributed by the liquid, by the solid,
(56)
and by the apparatus design. These factors may
be interrelated or independent of each other. In
order to study the effect of such factors upon the
performance of hydrocyclone units, they have
been usually divided into operational variables
and design variables.
4.1. Operation Variables
First of all, a pressure drop must exist in or-
der to drive the feed suspension across the unit.
An increment in this pressure drop will decrease
the cut size as well as increase the sharpness
of separation and flow rate. According to this,
great mass recoveries could be obtained raising
the pressure drop. Although such fact is literally
true, it is normally an uneconomical measure
because large pressure loss increases are nec-
essary. As has been mentioned earlier, practical
values of pressure drop range between 0.34 and
6 bar. However, for small cyclones, Trawinski
[27] has suggested an upper limit of 4 bar. As
in other separation techniques, a finite density

difference between the phases involved is essen-
tially necessary. Without it, no separation can be
achieved. As expected, experience has shown an
increase in separation efficiency as the density
difference increases.
The effect of feed solids concentration on
separation performance is well recognized in
practice. Generally, increasing input concentra-
tions reduces the total efficiency. Svarovsky and
Marasinghe [36] have proposed a semiempiri-
cal correlation, which allows scale-up prediction
of performance with feed concentrations higher
than 8 vol %. Neesse et al. [37] compared di-
lute flow (low solids concentration of the feed)
and dense flow (high solids concentration of the
feed) separations, focusing their study on the ef-
fect of turbulence.
Perhaps the least understood effect on per-
formance of a hydrocyclone is that of the vis-
cosity. Very little quantitative information exists
on the literature about it. It is normally accepted
that viscosity has little effect on separation ef-
ficiency. From simple theory it can be proved
that medium viscosity will only have an effect
at low Reynolds numbers (i.e., laminar flow),
but because turbulence is promoted by pumping,
the effect of viscosity will only be appreciable
when the flow is forced by static head or in very
large units. Nappier-Munn [38] found concor-
dance with theory about viscosity effects, while
Williamson et al. [39] studied the effect of vis-
cosity, along with pressure drop, in reducing the
minimum cut size of particles that can be sepa-
rated effectively. They found that cut size falls
as the pressure drop rises and as viscosity is re-
duced.
The effect of less common variables has also
been studied. Fontein et al. [40] examined the in-
fluence of some uncommon parameters such as
the roughness of the cyclone wall, as well as the
shape of the solid particles on suspension. They
found that rough-walled cyclones show lower ef-
ficiencies than smooth-walled ones. According
to them, this is mainly due to the fact the rotation
flow is greatly retarded by the rough wall, thus
decreasing the centrifugal force, which affects
separation.
Hydrocyclones 17
4.2. Design Variables
Some modifications and recommendations
about the geometry of hydrocyclones have ap-
peared in the literature, mainly from manufac-
tures. Despite many variations in design and sug-
gestions for varied configurations, the so-called
standard cyclones, such as the Rietema cyclone,
have proved the most efficient in separation and
capacity.
The most common materials of construc-
tion are steel, aluminum oxide, ceramics,
polyurethane, and other plastics. Sometimes the
inner body is lined with hard materials in order
to minimize wear caused by operation. Soft rub-
ber is popular and suitable where temperatures
do not exceed 60◦ C and where hydrocarbon oils
or solvents are absent. Ceramic liners can also
be used. Wear-resistant metals are useful when
particle sizes are above 100 µm or where tramp
material is suspected, which could damage rub-
ber linings. With temperatures above 60 ◦ C or in
presence of hydrocarbons, neoprene or nitrile, as
well as polyurethane rubbers have been success-
fully employed.
The effects of some geometry variables on
hydrocyclone performance are shown in Table
1.
Table 1. Effects of some design variables on hydrocyclone
performance
Variable Effect
Cyclone diameter increases cut size increases, pressure drop
usually decreases
Cyclone inlet diameter increases gravity force in cyclone
decreases, cut size increases,
capacity falls, pressure drop
decreases
Overflow diameter increases cut size increases, risk of coarse
sizes evident
Underflow diameter increases brings excess fines from liquid
phase into underflow
Cyclone shape becomes longer decreases cut size sharpness
separation
Cyclone wall become rougher capacity increases, efficiency
decreases
5. Selection and Design of
Hydrocyclone Systems
Hydrocyclones differ from other solid–liquid
separation equipments as they have a broad num-
ber of variables involved in their operation. The
influence of some variables is not completely
understood, and their effect may be completely
18 Hydrocyclones

different from one unit to another. Theses are described two methods for selection of units to
some of the reasons why the task of selecting be used in parallel for a given duty. Many other
a hydrocyclone system is difficult to deal with. authors have mentioned the advantage of using
Generalizations of performance characteristics multicyclone systems.
can seldom be made, and the choice necessar-
ily relies upon semiempirical relationships or
models, which are, inevitably, of limited appli- 5.2. Graphical Solutions
cation. In selecting or designing hydrocyclones
for a specific duty, the recommended way is ei- A series of charts or nomographs have been
ther using manufactures catalogues or to make also developed for the estimation of effects of
calculations to obtain cut point and throughputs. changes in operating variables on performance
of hydrocyclones. Zanker [43] presented a se-
ries of equations incorporated into some nomo-
5.1. Analytical Solutions graphs in order to make trial-and-error calcula-
tions faster. Rao and Nageswararao [44] sim-
The previously cited formulae for the computa- plified the use of charts to just one nomograph.
tion of cut points and capacities can be employed Again, combined with some calculations, they
as a first approximation. The actual choice can claimed such a graph to be a useful tool in select-
be mainly based on the function to be performed ing hydrocyclones for specific duties. Another
by the system. In classification applications, for proposed graphical method is given in [45]. The
instance, the cut size is usually fixed and the authors emphasized, however, that application
design is simply based upon achieving that cut of the technique should be for guidance purposes
size. On the other hand, in separation, a certain only.
degree of efficiency is normally required, and
the resulting design must be a compromise bet-
ween technical and economical considerations. 5.3. Manufactures’ Choice
The adaptability to industrial scale can represent
a real problem. Moir [41] pointed out that some Several equipment manufactures have published
of the correlations for calculating cut points and literature about selection of hydrocyclones in
throughputs for cyclones might not be suitable the form of brochures, technical reports, and
for industrial problems, since they were devel- so on. In many International Conferences held
oped using slurries with low solids contents and around the world some manufactures present pa-
small units. An alternative to high capacities is pers about selection, innovations, etc., as well
use of multiple hydrocyclone systems for indus- as table-top exhibitions of their available equip-
trial production. Gerrard and Liddle [42] have ment.
Table 2. Summary of some known hydrocyclone designs
Cyclone type and size of Di /Dc Do /Dc l/Dc L/Dc θ Stk 50 Eu** K p ∗∗ np ∗∗
hydrocyclone
Rietema’s design* Dc = 0.280 0.340 0.40 5.00 20◦ 0.0611 24.38 0.3748
0.075 m
◦
Bradley’s designDc = 0.133 0.200 0.33 6.85 9 0.1111 446.5 0.3230
0.038 m
◦
Mozley cyclone 1Dc = 0.022 0.154 0.214 0.57 7.43 6 0.1203 6381 0
m
Mozley cyclone 2Dc = 0.044 0.160 0.250 0.57 7.71 6◦ 0.1508 4451 0
m
Mozley cyclone 3Dc = 0.022 0.197 0.320 0.57 7.71 6◦ 0.2182 3441 0
m
Warman 3

model RDc = 0.290 0.200 0.31 4.00 15 ◦
0.1709 2.618 0.8000
0.076 m
◦
RW 2515 (AKW)Dc = 0.125 0.200 0.320 0.80 6.24 15 0.1642 2458 0
m
* Optimum separation.
** Optimum separation.

5.4. Dimensionless Scale-Up of
Hydrocyclones
5.4.1. Introduction
Because of the hydrocyclones’ inflexibility, the
methods of selection just reviewed can be only
used as an approximation. The correlations pro-
posed are often derived from studies of effects
of geometrical proportions on capacity and sep-
aration efficiency. In this sense, if literature on
hydrocyclones is rigidly followed when design-
ing a unit, it might lead to a unique hydrocyclone
choice, whose geometry could has been never
tested before, so reliable prediction of its per-
formance may be misleading. Based on this and
many other facts and factors, a scale-up proce-
dure to select hydrocyclones, attributed mainly
to Svarovsky [1] has been proposed. The method
consists of selecting a known design of hydrocy-
clone and, by using a series of dimensionless
relations, making the necessary predictions as
changes on fixed variables are introduced. A
summary of some known hydrocyclone designs
is presented in Table 2.
As the performance characteristics of
hydrocyclones involve a great number of vari-
ables, the use of dimensionless groups in this
method is an obvious advantage. The basic con-
cepts underlying dimensionless analysis and
scale-up procedures have been used for long
time in classical mechanics principally. Perhaps
the most common method of dimensional anal-
ysis is that attributed to Lord Rayleigh known
as the Pi theorem. The basic steps of dimen-
sional analysis can be found in (→ Scale-Up in
Chemical Engineering) [43].
5.4.2. Definition and Derivation of
Dimensionless Groups for Hydrocyclones
In the case of hydrocyclones, selection and op-
eration is based on the relationships between the
pressure drop and flow rate and the relationship
between separation efficiency and flow rate. The
pressure drop versus volumetric flow rate rela-
tionship is usually expressed as Eu = f (Re),
where Eu is the Euler number. These dimension-
less groups are graphically related as shown in
Figure 8. The Euler number is in fact a pres-
sure loss factor, easily defined as the limit on the
Hydrocyclones 19
maximum characteristic velocity v obtained by
a certain pressure drop ∆P across the hydrocy-
clone. It can be expressed as:
2∆P
Eu = (59)
ρv 2
where ρ is the density of the fluid.
Figure 8. A typical plot of Eu versus Re for hydrocyclones
The Reynolds number defines flow charac-
teristic of the system and in the case of hydrocy-
clones, the characteristic dimension may be
taken as the cyclone body diameter Dc , i.e.:
Dc vρ
Re = (60)
µ
where µ is the medium viscosity.
The relationship between separation effi-
ciency and flow rate is not significantly influ-
enced by operational variables, so it is com-
monly expressed in terms of cut size x 50 . The use
of the cut size to define efficiency of hydrocy-
clone is of the utmost importance since, as has
been previously emphasized their performance
is highly dependent on particle size. Since cut
size implies size of particles likely to be sepa-
rated, it follows that such particles must be in-
fluenced by forces exercised on the suspension.
The forces developed in a hydrocyclone can be
analyzed by sedimentation theory, and a dimen-
sionless group thus derived, the Stokes number,
will include cut size.
The Stokes number is a very useful theoret-
ical tool and, for the case of hydrocyclones, its
derivation may be carried out as follows. Con-
sidering a settling tank of length L and wide W,
it would have a settling area of A = LW, and if
the depth is H, then the retention volume is V =
20 Hydrocyclones

AH. With a flow through the tank of Q (m3 /h), 18µQDc

x2 = (68)
the detention time is given by: v 2 (ρs −ρ) A
V Considering the superficial velocity in the
td = (61)
Q hydrocyclone body as the characteristic veloc-


Particles to be removed must settle, under ity, the volumetric flow rate equals vπDc2 /4,
gravity influence, through the depth H requir- also A = πDc2 /4 for the cyclone body. Thus,
ing t s , where: substituting Q and A into Equation 68:
H 18µDc
ts = (62) x2 = (69)
ut v (ρs −ρ)
ut being the settling rate defined by Equation 6. Rearranging Equation 69 to express it in terms
Only particles for which t s is equal or less of inertial forces and hydrodynamic forces; the
than t d will be removed. Hence, by equating t s dimensionless group known as Stokes number
and t d a cut point or separation mesh is defined is obtained, i.e.:
as:
x2 (ρs −ρ) v
H V Stk = (70)
= (63) 18µDc
ut Q
Furthermore, if the dimension x is replaced
transposing for ut : by the specific cut size, x 50 :
HQ Q x250 (ρs −ρ) v
ut = = (64) Stk50 = (71)
V A 18µDc
If Equation 64 is then equated to Equation 6, or:
the limiting size of particles to be settled can be
x250 (ρs −ρ) v
derived giving: Stk50 (r) = (72)
18µDc
18µQ
x2 = (65) if the reduced cut size x
50 or x 50 (r) is incorpo-
g (ρs −ρ) A
rated.
For centrifugal systems the same replacement
done in Equation 6 leads to:
18µQ 5.4.3. Scale-Up at Low Concentrations
x2 = (66)
Rω 2 (ρs −ρ) A
At low concentrations of the feed (less than 1
Since ω = vt /R, where vt is the tangential ve- vol %), the flow pattern in a hydrocyclone is not
locity of the particle, Equation 66 can also be affected by the presence of particles in the flow
expressed as: and particle—particle interaction is negligible.
18µQR As only few particles report to the underflow, the
x2 = (67)
vt2 (ρs −ρ) A underflow-to-throughput ratio can be assumed
The radial settling velocity in a hydrocyclone to have no effect on the cut size. If this is the
is due to the centrifugal acceleration, which is case, the dimensional analysis gives two basic
proportional to the square of the tangential ve- relations between the three above described di-
locity of the particle and indirectly proportional mensionless groups:
to the radius of the particle position. As the tan- Stk50 Eu = constant (73)
gential motion of the particle is unopposed, the
tangential particle velocity can be taken as equal Eu = Kp Renp (74)
to the tangential component of the fluid veloc-
ity at the same point. For the same flow regime, where K p and np are empirical constants for a
the velocities anywhere in the flow in a hydrocy- family of geometrically similar cyclones. The
clone are proportional to a characteristic velocity product Stk50 Eu depends on the cyclone de-
v while the position radii are proportional to the sign, and on the underflow to throughput ratio.
cyclone diameter Dc . Under such assumptions Its value range from 0.06 to 0.33. The value of
Equation 67 can be approximated to: np is usually between 0.0 and 0.4 [1].
 Hydrocyclones 21

5.4.4. Scale-Up at High Concentrations flow behavior must, therefore, be evaluated by

At higher concentrations, the feed concentra-
tion as a fraction of volume C has to be in-
cluded as an additional dimensionless group.
The feed concentration of solids is a very impor-
tant variable, which affects the separation effi-
ciency and the pressure drop of a given hydrocy-
clone. It has been already mentioned the decreas-
ing in efficiency as feed concentration increases.
This means that the underflow-to-throughput ra-
tio has to be increased (to allow greater volume
of separated solids to be discharged). For this
reason, most hydrocyclones are equipped with
either a continuously variable underflow orifice,
or a series of replaceable nozzles.
Theoretical and empirical approaches of the
effect of feed solids concentration in hydrocy-
clones have appeared in the literature. Svarovsky
and Marasinghe [36] have developed the follow-
ing expression for the effect of high-feed solids
concentration:
Stk50 (r) = k1 (1−Rf ) exp(k2 C) (75)
where Stk 50 (r) includes the reduced cut size,
as already defined, while k 1 and k 2 are empir-
ical constants, which depend on the hydrocy-
24 1/2
clone configuration as well as on solids proper-
other means.
The point above mentioned, make neces-
sary to include parameters that characterize non-
Newtonian fluids in the previously developed
dimensionless groups, if they are going to be
used in scale-up of hydrocyclones dealing with
suspensions, which might be non-Newtonian. In
this case, an approach similar to that employed
in the derivation of Equation 17 may be carried
out. Firstly, in this type of fluid, the drag coeffi-
cient, C D , for steady-state creeping flow past a
sphere based on a macroscopic momentum and
energy balance of Wasserman and Slattery [47]
is:
24
CD = (76)
Re∗
where Re* is a generalized Reynolds number
analogous to that defined by Equation 17, be-
ing its characteristic dimension x, i.e,. the parti-
n
−1
cle diameter, and γ = K
(3) . Also, K
=




n
K[( 2n +1)/3n ] , for this case.
For Newtonian fluids, an approximation of
the drag coefficient in Stokes and the intermedi-
ate region [48] is:
 24
CD = +4.5 (Re<500) (77)
Re Re

ties. The correlation has proved to hold well for a modification of Equation 77 to better fit the
concentrations above 8 vol %, and the values of data is:
the constants k 1 and k 2 were found to be 9.05  24
24
1/2
CD = +α3 (Re∗<500) (78)
× 10−5 and 6.461, respectively, for limestone Reα1 Reα2

and a hydrocyclone manufactured by Amberger where the alpha constants are α1 = 1.1, α2 =
Kaolinwerke (AKW). 0.2, and α3 = 7.5.
By substitution of 24/Re* for 24/Re in Equa-
tion 78 and modification of the constants to en-
5.4.5. Considerations for non-Newtonian sure a friction factor of 0.44 at Re* = 500, the
Behavior resultant non-Newtonian drag coefficient corre-
lation is:
When fine powders are dispersed into liquids the  1/2
CD = 24 24
resultant suspension is usually non-Newtonian. Re∗α1 Re∗α2
+α3 (Re∗<500)(79)
The non-Newtonian behavior of suspensions is where the alpha constant values are as previously
normally a direct function of concentration. In noted.
this sense, the viscosity of a dilute suspension The terminal settling velocity for non-
nearly equals that of the suspending liquid. Thus, Newtonian fluids (Re<500) is found by equat-
when dealing with suspension of low solids ing Relation (77) with the generalized form of
contents either the apparent suspension viscos- the drag coefficient, i.e.:
ity or the suspending liquid viscosity can be

 1/2
4gx ρs −ρ 24 24
used for characterization purposes. However, for = +α3 (80)
3u2t ρ Re∗α1 Re∗α2
more concentrated suspensions non-Newtonian
behavior may be present and their viscosities Substituting Equation 17 and eliminating
cannot be considered constant anymore; their Reynolds number, Equation 80 becomes:
22 Hydrocyclones

5.4.6. Empirical Models



2
gxn +1
(ρs −ρ)
An exhaustive study for concentrations up to 10
18γ vol % was carried out by Medroho [49] in order
 α2
n
to prove the applicability of Equations 73, 74
2n
α3 ρx 2n
+α2 (2−n
)
= ut + ut (81)
24 γ and 75. He employed three geometrically sim-
ilar hydrocyclones of Rietema’s [30] optimum
For small Reynolds numbers (Re∗<0.1), the geometry and obtained the following relations:
second term of the right-hand side of Equation
81 is negligible and the simplification leads to: Stk50 (r)Eu = 0.047[ln(1/Rf )]0.74 exp(8.96C) (89)
 1/n

gxn
+1 (ρs −ρ)
ut = 18γ
(Re∗<0.1) (82)
Eu = 71(Re)−0.116 (Di /Dc )−1.3 exp(2.12C) (90)

For Newtonian fluids n
= 1, γ equals µ, so

Equation 82 becomes the familiar expression of Rf = 1218(Du /Dc )−4.75 (Eu)−0.30 (91)
Stokes law as described by Equation 6. where Rf is the underflow-to-throughput ratio,
Following a similar procedure as when ob- C is the volume fraction of solids, Di , Dc , and
taining Stokes’ number expression, Equation 82 Du are the inlet, body, and underflow diame-
can be equated to Equation 6, i.e.: ters of the hydrocyclone, respectively. Covering

1/n
also the low feed concentration range (i.e. up
Q gxn
+1 (ρs −ρ)
= (83) to 10 vol %), Antunes and Medroho [50] de-
A 18γ


veloped a set of equations bearing similarity to
transposing for xn +1 : those presented above, but for hydrocyclones of

Bradley’s [2] geometry.

18γQn
xn +1
= (84) Ortega-Rivas and Svarovsky [51] have pro-
g (ρs −ρ) An

posed a model for concentrated suspensions and
Again, considering centrifugal acceleration: accounting for non-Newtonian behavior of the

treated slurries. They covered a feed concentra-

18γRQn
xn +1
=
(85) tion range from 5 to 25 vol % and used geometri-
vt2 (ρs −ρ) An
cally similar hydrocyclones following Rietema’s
and, under the assumptions made before: [30] proportions. The relationships that they de-

rived are presented below:

18γQn Dc
xn +1
= (86)
v (ρs −ρ) An

2 ∗
Stk50 (r)Eu = 0.006[ln(1/Rf )]2.37 exp(6.84C) (92)
Substituting Q and A and rearranging
terms, the Stokes number expression for non- Eu = 1686(Re∗ )−0.035 exp(−3.39C) (93)
Newtonian systems is obtained, i.e.:


Rf = 32.8(Du /Dc )1.53 (Re∗ )−0.34 exp(3.70C) (94)
xn +1 (ρs −ρ) v 2−n
Stk∗ = (87)
18γDc where the dimensionless Reynolds and Stokes
It can be noticed that, again, that for Newto- numbers include the parameters of characteri-
nian fluids this generalized Stokes number re- zation of non-Newtonian suspension previously
duces to the common expression represented by described. Equations 92, 93 and 94 allow reli-
Equation 12. able hydrocyclones design and scale-up at con-
Finally, since concentration is an important centrations up to 25 vol % of shear-thickening
factor in this case, replacing the reduced cut size suspensions. They apply to Rietema’s geometry,
for the simple particle size, x: but there is some evidence suggesting that they



could equally be employed to other geometries,
n +1
∗ x50 (r) (ρs −ρ) v 2−n because reasonable agreement has been found
Stk50 (r) = (88)
18γDc with previous models of hydrocyclone design
in which x 50 (r) is preferred to x
50 in order to and performance. The authors also claim that
avoid confusion.

for the case of Newtonian suspensions, i.e., low
solids contents, the above mentioned equations
would reduce to the common definitions so they
conclude their model is of wide applicability.
All the above described relationships have
the advantage of presenting graphical functions,
such as the one shown in Figure 8. Those graphs
can be handled for design purposes by manipu-
lating variables and predict, in a practical man-
ner, operation effects of hydrocyclone technol-
ogy.
Figure 9. Hydrocyclone test rig
5.4.7. Practical Applications
In order to test the different models previously
described in this article, experimental runs in
small hydrocyclone rigs are advised. Standard
geometries of cyclones, such as the Rietema’s
hydrocyclone, can be used to run tests at dif-
ferent diameters and pressure drops. As previ-
ously mentioned, many different design graphs
describing the relationships between dimension-
less groups may be available for variable manip-
ulation and prediction of best performance vari-
ables. A hydrocyclone test rig is shown in Fig-
ure 9. The units are changeable and they may
be equipped with removable underflow nozzles
Hydrocyclones 23
(Fig. 10.), in order to verify the effect that varia-
tions in the underflow orifice diameter may have
in the cut size. Once a number of tests have been
performed, the possible scale-up to industrial
size may follow.
Figure 10. Rietema’s proportions hydrocyclone with re-
placeable underflow spigots
Since cut size is a direct function of hydrocy-
clone diameter, the best efficiency for a number
of useful applications would be given by small
units. Cut sizes down to 2 µm have been reported
in 10 mm diameter hydrocyclones [51]. On the
other hand, also capacity is a direct function of
hydrocyclone diameter, so larger units will de-
liver higher flow rates. Therefore, it would be
necessary to reach a compromise between ef-
ficiency and capacity for industrial operation.
In this sense, a basic small geometry may be
kept to cut as fine particles as possible, while
series arrangements may be required to render
industrial capacities. A number of arrangements
have been experienced, but probably the most
efficient is the radial (spider) arrangement with
the hydrocyclones axes horizontally placed like
spokes in a wheel lying down. As shown in Fig-
ure 11, one “wheel” or hydrocyclone deck ac-
commodates a number of “spokes” or hydrocy-
24 Hydrocyclones
Figure 11. Diagram of radial (spider) arrangement of hydrocyclones for industrial scale
clone bodies. For cyclone cavities with a cylin-
drical diameter of 10 mm the diameter of the
deck is 300 mm. The whole cyclone deck can be
molded by injection in one piece. The feed zone
is the annular space closest to the periphery. The
overflow is through peripheral screw-in-vortex
finders. The underflow leaves via the central cir-
cular space where the small apex openings enter.
The design feed flow per cyclone cavity is
about 250 L/h corresponding to a pressure drop
of 600 kPa, approximately. A deck with 20 cy-
clones in parallel consequently may have a ca-
pacity of 5 m3 /h. This is, however, not the max-
imum capacity. The cyclone cavities can be
closed off individually by insertion of a blinding
device in order to control the capacity. For feed
flows above 5 m3 /h, the decks can be stacked on
top of each other and held together by connecting
rods through the inner annular space. There is a
limit to the number of decks that can be stacked
on top of each other. If the stack of decks be-
comes too high, the pressure drop over the feed
zone will result in a lower inlet pressure to the cy-
clones furthest downstream and, consequently,
these cyclones will give lower throughputs and
lower efficiency. Some other arrangements have
been tested with different degrees of success.
Some of them have been reviewed in [1].
 Hydrocyclones 25

As has been discussed, hydrocyclone tech-  4.75

0.005m
nology has numerous applications in a wide va- Rf = 1218 (102.7)−0.3 = 0.1346 (97)
0.0254m
riety of industries. Hydrocyclones have been
particularly used for long time in mineral pro- Equation 89 can then be employed to compute
cessing, as well as in the oil and gas industry. the value of the Stk 50 (r) group, i.e.:
In these industries, many materials can be con- 0.047[ln (1/0.1346)]0.74 exp [8.96 (0.05)]
sidered inert powders, and thus the equations Stk50 (r) =
102.7
of the dimensionless scale-up model proposed = 0.0019 (98)
by Medronho and Svarovsky [49] can be em-
ployed to carry out experimental tests, as sug- Finally, transposing the cut size from the defini-
gested above. Hydrocyclones have been also tion of the Stokes number given by Equation 72,
successfully used in other industries, such as the the result is obtained as follows:
food and biochemical industry. Since most ma-
terials in these industries may behave as non-
Newtonian, the model suggested by Ortega-
Rivas and Svarovsky [51] would be more suit-
able for modeling purposes of processes involv-
ing biological systems.
Some examples of application of the dimen-
sionless scale-up model are given below.
Example 1. Predict the cut size obtained if
1 L/s of a suspension containing 5 vol % of a
2150 kg/m3 dolomite slurry in water, is going to
be separated in a 25.4 mm diameter hydrocy-
clone of Rietema’s geometry. The available
pressure is 2 × 105 Pa and the underflow orifice
is 5 mm.
The relatively low concentration of suspen-
sion would allow use of the density and viscos-
ity of water as suspending medium properties.
Also, since no temperature has been given, a
reasonable assumption for the water density is
1000 kg/m3 and for the water viscosity 0.001 kg
m−1 s−1 . Thus, bearing in mind to use the super-
ficial velocity in the cyclone body, [4Q/π(Dc )2 ]
as the characteristic one, the following results
are obtained:
The characteristic velocity is, firstly, ob-
tained, as defined above, i.e.:
4 (0.001) m3 /s
v= = 1.97 m/s (95)
3.1416(0.0254)2 m2
Then, substituting into Equation 59, the Euler
number is calculated as:
 ing streams discharge at atmosphere conditions.
2 2×105 kg/s2 ·m
Eu = = 102.7 (96)
(1000) kg/m3 (1.97) m/s
As stated before, due to the low suspension con-
centration, the set of equations suggested for
this case [49] can be considered, and thus, using
Equation 91 the underflow-to-throughput ratio
may be obtained as:
 
(0.0019) (18) (0.001) kg/m·s (0.0254) m
x
2
50 =
(2150−1000) kg/m3 (1.97) m/s
= 1.55×105 m (99)
The cut size calculated is approximately equal
to 15 µm.
Example 2. A slurry containing calcium car-
bonate at a concentration of 15 vol % solids,
is going to be clarified using hydrocyclones
of Rietema’s proportions. The slurry has been
characterized being non-Newtonian in behav-
ior, with values of the parameters n
and K

of 1.39 and 2 × 10−5 kg m−1 · s−(2−n) , respec-
tively. The density of the slurry has been deter-
mined as 1250 kg/m3 . It is required to clarify
8 L/s of slurry, counting on a maximum pump-
ing capacity of 315 kPa. Find out the optimum
cyclone diameter and the number of units, if
necessary, in order to obtain a cut size of 8 µm.
Small scale experimentation has determined val-
ues of some dimensionless relationships as fol-
lows: Stk 50 Eu = 0.0625, and Eu = 720. The cal-
cium carbonate density has been determined as
2800 kg/m3 .
The value of the parameter n corresponds to
a shear-thickening fluid, while the requirement
of using Rietema’s cyclones makes it possible
to use the model suggested by Ortega-Rivas
and Svarovsky [51]. The maximum pumping
capacity would produce a pressure drop of the
same value (315 000 N/m2 ), because the leav-
Substituting the definition of superficial veloc-
ity in the cyclone body, as used in the previous
example, into Equation 59, and considering the
density of the slurry due to its high concentra-
tion, the following relation is obtained:
26 Hydrocyclones

4 (Q2 ) (Eu) (ρm )
Dc =
(1.23) (∆P)
Substituting, also, the same characteristic veloc-
ity in Equation 88 and transposing for the re-
duced cut size:


xn
50
+1
(r) =



(Stk∗50 ) (18) (3)n −1 (K
) [(2n+1) / (3n)]n

Englewood Cliffs., NJ, USA 1965.
(1.27Q)2−n (ρs −ρ)
As previously stated, the cut size is directly pro-
portional to the hydrocyclone diameter, while
this diameter is inversely proportional to the flow
rate. It would be, therefore, necessary to verify
whether the required cut size could be obtained
in a single hydrocyclone, whose diameter would
match the capacity needed. If both variables do
not coincide (cut size and flow rate with a single
unit), iterations would be needed until a number
of units is found, each of them separating the
required size, while processing the appropriate
fraction of the total capacity. A first approxima-
tion would consist in calculating the hydrocy-
clone diameter, using the first equation derived
above, which will handle the 8 L/s of slurry to be
treated. Once having that diameter, with the sec-
ond equation, the cut size that the single unit sep-
arates, will be known. If such diameter is wider
than 8 µm, the volumetric flow rate will be frac-
tioned in equal parts to recalculate hydrocyclone
diameter until a match of number of units and
cut size is found. The first approximation leads
to the following values:
Dc = (102)
 16. D. G. Thomas, AIChE J. 8 (1962) 373.
2 6 2 3
4 (0.008) m /s (720) (1250) kg/m
= 0.11 m
(1.23) (315000) kg/s2 ·m
and:
x∗50 =



2.39 8.68×10−5 (18) (1.53) 2×10−5
(0.0608) (m3 /s)
x∗50 = 15 µm
So it is found that a 0.11 m diameter hydrocy-
clone will separate particles much coarser than
the required cut size. By doing iterations, when
fractioning the rate four times, i.e., when Q
= 0.002 m3 /s, a diameter of hydrocyclone Dc
= 0.055 m and a cut size of 7.8 × 10−6 m are
found. Therefore, the election would be the use
of four units, 5.5 cm in diameter, to handle the
capacity required and obtain the cut size needed.
6. References
(100)
1. L. Svarovsky: Hydrocyclones, Holt Saunders,
Eastbourne, UK 1984.
2. D. Bradley: The Hydrocyclone, Pergamon
Press, Oxford 1965.
(101) 3. J. Happel, H. Brenner: Low Reynolds Number
5−2n
Hydrodynamics, Prentice-Hall Inc.,
(Dc )
4. A. Einstein, Ann. Phys. 19 (1906) 289.
5. H. L. Frisch, R. Simha: “The Viscosity of
Colloidal Suspension and Macromolecular
Solutions”, in F. Eirich (ed.): Rheology, Vol. 1,
Academic Press, New York 1956.
6. E. Guth, R. Simha, Kolloid Z. 74 (1936) 226.
7. R. Simha, J. Appl. Phys. 23 (1952) 1020.
8. D. G. Thomas, Ind. Eng. Chem. 55 (1963) 18.
9. S. F. Lin, R. S. Brodkey,J. Rheol. 29 (1985)
147.
10. R. G. Griskey, D. G. Nechrebecki, P. J.
Notheis, R. T. Balmer, J. Rheol. 29 (1985) 349.
11. A. J. De Vries, Rheol. Abs. 5 (1985) 32.
12. A. B. Metzner: “non-Newtonian Technology:
Fluid Mechanics, Mixing, and Heat Transfer”,
in T. B. Drew, J. W. Hoopes Jr. (ed): Advances
in Chemical Engineering, vol. 1, Academic
Press, New York 1959.
13. A. B. Metzner: “Flow of non-Newtonian
Fluids”, in V. Streeter (ed.): Handbook of
Fluid Mechanics, McGraw-Hill Book Co., Inc,
New York 1961.
14. D. W. Dodge, A. B. Metzner, AIChE J. 5
(1959) 189.
15. R. G. Shaver, E. W. Merril, AIChE J. 5 (1959)
181.
17. H. F. Trawinski, Filtr. Sep. 6 (1969) 7.
18. R. R. Irani, C. F. Callis: Particle Size
Measurement: Interpretation and Applications,
John Wiley & Sons, London 1963.
(103)
kg/m·s2−n (0.872) (0.007445) m5−2n
2−n
(1800) kg/m3
(104)
19. P. H. Fahlstrom, Proc. Inst. Min. Metall. 69
(1960) 632.
20. M. I. G. Bloor, D. B. Ingham, Filt. Sep. 21
(1984) 266.
21. H. Schubert, T. Neesse: “A Hydrocyclone
Separation Model in Consideration of the
Turbulent Multi-phase Flow”, in BHRA
Conference on Hydrocyclones Proceedings,
Paper 3, Cambridge, UK 1980.

22. T. Allen: Particle Size Measurement, Chapman
& Hall, London 1981.
23. G. V. Barbosa-Canovas, E. Ortega-Rivas, P.
Juliano, H. Yan: Food Powders: Physical
Properties, Processing and Functionality,
Kluwer Academic/Plenum Publishers, New
York 2005.
24. C. C. Harris, Powder Technol. 5 (1971/1972)
39.
25. H. J. E. v. Tenbergen, K. Rietema: “Efficiency
of Phase Separations”, in K. Rietema, G. C.
Verver (eds.): Cyclones in Industry, Elsevier
Publ. Co., Amsterdam 1961.
26. L. Svarovsky: “The Efficiency in Separation
Processes”, in R. J. Wakeman (ed.): Progress
in Filtration and Separation, Elsevier Publ.
Co., Amsterdam 1979.
27. H. F. Trawinski: “Hydrocyclones”, in D. B.
Purchas (ed.): Solid/Liquid Separation
Equipment Scale-up, Uplands Press Ltd.,
Croydon, UK 1977.
28. H. F. Trawinski, Aufbereitungs-Technik. 17
(1976) 449.
29. M. A. Doheim, G. A. Ibraheim, A. A. Ahmed,
Int. J. Miner. Process. 14 (1985) 149.
30. K. Rietema, Chem. Eng. Sci. 15 (1961) 298.
31. D. A. Dahlstrom, Trans. Amer. Inst. Min.
(Metall.) Engrs. 184 (1949) 331.
32. N. Yoshioka, Y. Hotta, Chem. Eng. Japan. 19
(1955) 632.
33. A. L. De Gelder: “Model Tests with
Hydrocyclones”, in Proceedings of IChemE
Symposium on Scaling-up of Chemical Plants
and Processes, London, 1957.
34. A. J. Lynch, T. C. Rao, Indian J. Technol. 6
(1968) 106.
35. H. F. Trawinski, Chem. Ing. Tech. 30 (1958)
85.
36. L. Svarovsky, B. S. Marasinghe: “Performance
of Hydrocyclones at High Feed Solid
Concentrations”, in BHRA Conference on
Hydrocyclones Proceedings, Paper 10,
Cambridge, UK 1980.
37. T. Neesse, W. Dalman, D. Espig: “Effect of
Turbulence on the Efficiency of Separation in
Hydrocyclones at High Feed Solids
Hydrocyclones 27
Concentration”, in BHRA Conference on
Hydrocyclones Proceedings, Paper B3, Bath,
UK 1984.
38. T. J. Nappier-Munn: “Influence of Medium
Viscosity on the Density Separation of
Minerals in Cyclones” in BHRA Conference
on Hydrocyclones Proceedings, Paper 6,
Cambridge, UK 1980.
39. R. D. Williamson, T. R. Bott, I. H. Newson:
“Small Particle Separation in a
Hydrocyclone”, in BHRA Conference on
Hydrocyclones Proceedings, Bath, UK 1984.
40. F. J. Fontein, J. G. Van Kooy, H. A. Leniger,
Brit. Chem. Eng. 7 (1962) 410.
41. D. N. Moir, The Chemical Engineer 523
(1985) 20.
42. A. M. Gerrard, C. J. Liddle, The Chemical
Engineer 407 (1975) 295.
43. A. Zanker, Chem. Eng. 82 (1977) 122.
44. T. C. Rao, K. Nageswararao, Chem Eng. 82
(1975) 121.
45. R. W. Day, C. N. Grichar: “Hydrocyclone
Separation”, in P. A. Schweitzer (ed.):
Handbook of Separation Techniques for
Chemical Engineers, McGraw-Hill Book Co,
New York 1979.
46. S. J. Kline: Dimensional Analysis as a
Research Tool, McGraw-Hill Book Co, New
York 1962.
47. M. L. Wasserman, J. C. Slattery, AIChE J. 10
(1964) 383.
48. R. B. Bird, W. E. Stewart, E. N. Lightfoot:
Transport Phenomena, John Wiley & Sons,
New York 2001.
49. R. A. Medronho, L. Svarovsky: “Tests to
Verify Hydrocyclone Scale-up Procedure”, in
BHRA Conference on Hydrocyclones
Proceedings, Paper A1, Bath, UK 1984.
50. M. Antunes, R. A. Medronho: “Bradley
Hydrocyclones: Design and Performance
Analysis”, in L. Svarovsky, M. T. Thew (eds.):
Hydrocyclones: Analysis and Applications,
Kluwer Academic Publishers, Dordrecht,
Netherlands 1992.
51. E. Ortega-Rivas, L. Svarovsky, Fluid/Part.
Sep. J. 6 (1993) 104.
 Hydrogen Peroxide 1

Hydrogen Peroxide
Gustaaf Goor, Degussa AG, Hanau, Germany (Chap. 1, 2, 3, 4, 5, 6 and 7)
Jürgen Glenneberg, Degussa AG, Hanau, Germany (Chap. 1, 2, 3, 4, 5, 6 and 7)
Sylvia Jacobi, Degussa AG, Hanau, Germany (Chap. 8)

1. Introduction and Historical 4.4. Other Processes . . . . . . . . . . . . 20

Aspects . . . . . . . . . . . . . . . . . 1 4.4.1. Production from Peroxy Compounds 20
2. Physical Properties . . . . . . . . . . 3 4.4.2. Production from Hydrogen
3. Chemical Properties . . . . . . . . . 5 and Oxygen . . . . . . . . . . . . . . . 20
4. Production . . . . . . . . . . . . . . . 6 4.4.3. Production from Carbon Monoxide,
4.1. Anthraquinone Process Oxygen, and Water . . . . . . . . . . 22
(AO Process) . . . . . . . . . . . . . . 6 4.4.4. Production by Autoxidation of
4.1.1. Principles . . . . . . . . . . . . . . . . 6 Organic Compounds . . . . . . . . . 22
4.1.2. Process Description . . . . . . . . . . 9 4.4.5. Production by Cathodic Reduction
4.1.2.1. Hydrogenation . . . . . . . . . . . . . 10 of Oxygen . . . . . . . . . . . . . . . . 22
4.1.2.2. Oxidation . . . . . . . . . . . . . . . . 14 5. Storage and Transportation . . . 23
4.1.2.3. Extraction and Drying . . . . . . . . 15 6. Safety . . . . . . . . . . . . . . . . . . 24
4.1.2.4. Working Solution Purification and 7. Uses . . . . . . . . . . . . . . . . . . . 25
Regeneration . . . . . . . . . . . . . . 15 8. Toxicology and Occupational
4.1.2.5. Purification of Crude Hydrogen Health . . . . . . . . . . . . . . . . . . 26
Peroxide . . . . . . . . . . . . . . . . . 16 9. Environmental Effects . . . . . . . 28
4.1.2.6. Concentration of Hydrogen 9.1. Atmospheric Fate . . . . . . . . . . 28
Peroxide . . . . . . . . . . . . . . . . . 17 9.2. Aquatic Fate . . . . . . . . . . . . . . 29
4.2. 2-Propanol Process (Shell Process) 17 9.3. Biodegradation . . . . . . . . . . . . 30
4.3. Electrochemical Processes . . . . . 18 9.4. Environmental Releases . . . . . . 30
4.3.1. Degussa–Weissenstein Process . . . 19 9.5. Environmental Effects . . . . . . . 30
4.3.2. Münchner Process . . . . . . . . . . . 20 10. References . . . . . . . . . . . . . . . 31

This review deals with chemical and physical
properties of hydrogen peroxide and gives de-
tailed descriptions about the various production
processes. Today, almost all hydrogen peroxide
is produced by organic autoxidation processes,
primarily the anthraquinone process. Other as-
pects in this review are handling, safety and us-
age of hydrogen peroxide. Hydrogen peroxide
occurs naturally. Because of its chemical reac-
tivity, hydrogen peroxide also has toxic prop-
erties. The formation and degradation of atmo-
spheric hydrogen peroxide is photochemically
mediated. Degradation of aquatic hydrogen per-
oxide is catalyzed by biological material or by
transition metals. Toxicity data are collected for
various organisms.
1. Introduction and Historical
Aspects
The industrial production of hydrogen peroxide
passed through three phases, starting with wet
chemical processes, followed by electrochem-
ical processes, and then by organic autoxida-
tion processes. Almost all hydrogen peroxide
is now produced by organic autoxidation (AO)
processes, primarily the anthraquinone process.
Wet Chemical Processes. In 1818, L. J.
Thenard [9] obtained hydrogen peroxide for the
first time by treating barium peroxide [1302-29-
6] with nitric acid. This process was improved by
using hydrochloric acid to release hydrogen per-
oxide. The water-soluble barium chloride which

c 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 443.pub2
2 Hydrogen Peroxide
is formed simultaneously, was precipitated with
sulfuric acid.
BaO2 +2 HCl→BaCl2 +H2 O2
BaCl2 +H2 SO4 →BaSO4 +2 HCl
BaO2 +H2 SO4 →BaSO4 +H2 O2
With this reaction, Thenard established the
foundation for the commercial manufacture of
aqueous hydrogen peroxide solutions based on
wet chemical processes, which began around
1880.
Industrial plants using the barium peroxide
process were still operating until the middle of
the 20th century. Around 1900, approximately
10 000 t/a of barium peroxide was processed by
this technology, yielding ca. 2000 t/a of hydro-
gen peroxide. Sales opportunities for the coprod-
uct barium sulfate (BaSO4 , blanc fixe) had a de-
cisive effect on the profitability of the process.
The 3 % aqueous hydrogen peroxide solu-
tions manufactured by the barium peroxide pro-
cess had only a limited market because of their
high production cost, low hydrogen peroxide
content, and unsatisfactory stability due to im-
purities.
Electrochemical Processes. The introduc-
tion of electrochemical processes eliminated the
disadvantages of the barium peroxide process.
In 1853, Meidinger discovered the forma-
tion of hydrogen peroxide during electrolysis
of aqueous sulfuric acid [10]. In 1878, Berth-
elot showed that peroxodisulfuric acid [13445-
49-3] is formed during this electrolysis and then
hydrolyzed by water to give sulfuric acid and
hydrogen peroxide via peroxomonosulfuric acid
[12188-01-1] [11]:
2 H2 SO4 →H2 S2 O8 +H2
H2 S2 O8 +H2 O→H2 SO5 +H2 SO4
H2 SO5 +H2 O→H2 SO4 +H2 O2
2 H2 O→H2 O2 +H2
In 1905, following the work of Elbs and Schoen-
herr [12], G. Teichner developed an electrolytic
process at the Konsortium für Elektrochemis-
che Industrie. The first hydrogen peroxide plant
based on this new technology went on stream in
1908 at the Österreichische Chemische Werke
in Weissenstein. The Weissenstein process was
followed in 1910 by the Münchner process de-
veloped by A. Pietzsch and G. Adolph at the
Elektrochemische Werke, Munich, and then in
1924 by the Riedel–Loewenstein process de-
veloped by L. Loewenstein and first used by
Riedel-de-Haen. In these processes, an ammo-
nium sulfate solution was electrolyzed instead
of free sulfuric acid, and the resulting am-
monium peroxodisulfate (Riedel–Loewenstein)
or potassium peroxodisulfate derived therefrom
(Pietzsch–Adolph) was hydrolyzed. After intro-
duction of these processes, production (calcu-
lated as 100 % hydrogen peroxide) increased
steadily and, in 1950, reached approximately
(30–35) × 103 t/a [13].
Organic Autoxidation Processes. The deci-
sive breakthrough in industrial production of hy-
drogen peroxide, which enabled the construc-
tion of modern large-scale plants, came with the
development of organic autoxidation processes,
especially the anthraquinone process.
In 1901, Manchot discovered that hydro-
quinones (or hydrazobenzenes) react quantita-
tively with oxygen to form quinones (or azoben-
zenes) and hydrogen peroxide [14]. In 1932,
Walton and Filson in the United States went
back to this work and proposed a cyclic process
for the manufacture of hydrogen peroxide based
on azobenzene–hydrazobenzene [15].
Pfleiderer of I. G. Farbenindustrie AG, Lud-
wigshafen (BASF), subsequently developed a
process for the alkaline autoxidation of hydra-
zobenzene in which sodium peroxide is ob-
tained instead of free hydrogen peroxide [16].
This process was used on an industrial scale by
Kymmene AB in Kuusankoski (Finland) under
BASF licence (see Section 4.4.4.
The azobenzene process had two major tech-
nical drawbacks: hydrogenation of azobenzene
was performed with sodium amalgam, and ox-
idation of hydrazobenzene proceeded satisfac-
torily only in alkaline solution. Pfleiderer and
Riedel overcame these problems by using alky-
lated anthraquinones instead of azobenzene. Be-
tween 1935 and 1945, BASF developed the
Riedel–Pfleiderer process (generally referred to
as the anthraquinone or AO process) in a pi-
lot plant with a monthly output of 30 t. At the
 Hydrogen Peroxide 3

end of World War II, this work was stopped in wt % hydrogen peroxide is obtained by frac-
Germany because of Allied Control Commis- tional crystallization of highly concentrated (ca.
sion Regulations, but it formed the basis for all 90 wt %) aqueous solutions. Pure 100 wt % hy-
present production processes. In 1953, Du Pont drogen peroxide is stable at room temperature. It
of Memphis, Tennessee, commissioned the first is of scientific interest only and is not produced
hydrogen peroxide production plant based on the on an industrial scale.
AO process. Several other companies began to Data pertaining to molecular structure (ob-
produce hydrogen peroxide by this process, and tained by neutron diffraction on solid H2 O2 ) are
production capacity increased greatly. In 2002, as follows [18]:
the global production capacity excluding China
was estimated to be 2.8 × 106 t/a (calculated as Bond length, O−O 145.3 ± 0.7 pm
100 % H2 O2 ) [17]. Bond length, O−H 99.8 ± 0.5 pm
Bond angle, O–O–H 102.7 ± 0.3◦
Another organic autoxidation process, the Azimuthal angle 90.2 ± 0.6◦
2-propanol process, used by Shell in Norco
(United States) between 1957 and 1980, made
only a limited contribution to the increase of hy- Table 1. Physical properties of hydrogen peroxide and water
drogen peroxide production with an annual pro- Property Value
duction capacity of about 15 000 t.
H2 O2 H2 O
Various patents on direct combination of hy- mp, ◦ C −0.43 0
drogen and oxygen have been published but up bp (101.3 kPa), ◦ C 150.2 100
to now, the direct synthesis of hydrogen peroxide Heat of fusion, J/g 368 334
has not been implemented on a large industrial Heat of vaporization, J g−1 K−1
At 25 ◦ C 1519 2443
scale. At bp 1387 2258
Specific heat, J g−1 K−1
Liquid (25 ◦ C) 2.629 4.182
Gas (25 ◦ C) 1.269 1.865
2. Physical Properties Relative density, g/cm3
0 ◦C 1.4700 0.9998
Hydrogen peroxide [7722-84-1], M r 34.016, is 20 ◦ C 1.4500 0.9980
25 ◦ C 1.4425 0.9971
a clear, colorless liquid which is miscible with Viscosity, mPa·s
water in all proportions. Hydrogen peroxide and 0 ◦C 1.819 1.792
highly concentrated aqueous solutions (> 65 20 ◦ C 1.249 1.002
Critical temperature, ◦ C 457 374.2
wt %) are soluble in a variety of organic solvents
Critical pressure, MPa 20.99 21.44
such as carboxylic esters. Refractive index n20D 1.4084 1.3330
Hydrogen peroxide and water do not form an
azeotropic mixture and, in theory, can be sepa- A discussion of the molecular structure can
rated completely by distillation. In practice, 100 be found in [19].

Table 2. Physical properties of aqueous hydrogen peroxide solutions

Property H2 O2 concentration, wt %
35 50 70 90
Relative density
0 ◦C 1.1441 1.2110 1.3071 1.4136
20 ◦ C 1.1312 1.1953 1.2886 1.3920
25 ◦ C 1.1282 1.1914 1.2839 1.3867
Viscosity, mPa · s
0 ◦C 1.82 1.87 1.93 1.88
20 ◦ C 1.11 1.17 1.23 1.26
Refractive index
n20
D 1.3563 1.3672 1.3827 1.3995
mp, ◦ C −33 −52.2 −40.3 −11.9
bp (101.3 kPa), ◦ C 107.9 113.8 125.5 141.3
H2 O2 partial pressure
(30 ◦ C), kPa 0.05 0.11 0.17 0.29
4 Hydrogen Peroxide

Physical constants of hydrogen peroxide and

water are compared in Table 1. For a more de-
tailed comparison, see [20]. Physical constants
of some aqueous hydrogen peroxide solutions in
commercially available concentrations are listed
in Table 2.
The vapor pressure and partial pressure of
aqueous hydrogen peroxide solutions are shown
as a function of temperature in Figures 1 and
2, respectively. Figure 3 shows the vapor–liquid
equilibrium curve for aqueous hydrogen perox-
ide solutions [21].

Figure 1. Vapor pressure of aqueous hydrogen peroxide

solutions

Figure 3. Vapor–liquid equilibrium curve for water–hy-

drogen peroxide

Figure 4. Freezing point curve for water–hydrogen peroxide

The solid–liquid phase diagram (Fig.

Figure 2. Partial pressure of hydrogen peroxide over
aqueous hydrogen peroxide solutions 4) shows eutectic points for the mixtures
 Hydrogen Peroxide 5

ice–H2 O2 · 2 H2 O at 45.2 wt % hydrogen per- Table 4. Hydrogen peroxide as a reducing agent [20]

oxide and for solid H2 O2 –H2 O2 · 2 H2 O at Redox reaction Standard

potential
61.2 wt % hydrogen peroxide with a congruent E0 , V *
melting point for the compound H2 O2 · 2 H2 O pH 0
between them. HOOH ⇒ 2 H+ + O2 + 2 e− − 0.66
Heat of formation and of decomposition of 5 e− +MnO− +
4 +8 H →Mn
2+
+4 HOH + 1.51
1 e− + Ce4+ ⇒ Ce3+ + 1.61
hydrogen peroxide are: pH 14
HOOH + 2 OH− ⇒ 2 HOH + O2 + 2 e− + 0.08
H2 (g) + O2 (g) ⇒ H2 O2 (g) −136.2 kJ/mol 1 e− +ClO2 →ClO− 2 + 1.16
H2 (g) + O2 (g) ⇒ H2 O2 (l) −187.9 kJ/mol 2 e− + ClO− + HOH ⇒ Cl− + 2 OH− + 0.89
H2 O2 (g) ⇒ H2 O (g) + 1/2 O2 (g) −105.8 kJ/mol * Standard potential of redox reactions measured against a
H2 O2 (l) ⇒ H2 O (l) + 1/2 O2 (g) −98.3 kJ/mol hydrogen electrode (25 ◦ C, 100 kPa).

Decomposition of hydrogen peroxide occurs

with disproportionation
3. Chemical Properties
H2 O2 →H2 O+1/2 O2
Dissociation. Hydrogen peroxide is weakly and is extremely important in handling hydrogen
acidic in aqueous solution, with a dissociation peroxide during storage and in the laboratory.
constant of 1.78 × 10−12 (pK 11.75) at 20 ◦ C. This reaction is highly exothermic (see Chap-
As a weak acid, hydrogen peroxide forms salts ter 2) and takes place in the presence of small
with various metals. amounts of catalyst even in aqueous solution. In
the absence of catalyst, it occurs only in the gas
Oxidation and Reduction. Hydrogen per- phase at high temperature.
oxide can behave both as an oxidizing and as Decomposition can be catalyzed both homo-
a reducing agent (Tables 3 and 4). Systems with geneously by dissolved ions (especially of the
a redox potential E 0 < − 1.80 V at pH 0 can- metals iron, copper, manganese, and chromium)
not be oxidized by hydrogen peroxide; systems and heterogeneously by suspended oxides and
with a redox potential E 0 > − 0.66 V at this pH hydroxides (e.g., those of manganese, iron, cop-
cannot be reduced by hydrogen peroxide [6, 22]. per, palladium, or mercury) and by metals such
as platinum, osmium, and silver.
Table 3. Hydrogen peroxide as an oxidizing agent [20]
Redox reaction Standard Substitution. The hydrogen atoms of hydro-
potential
E0 , V *
gen peroxide can be substituted by alkyl and acyl
pH 0 groups, leading to the formation of
HOOH + 2 H+ + 2 e− ⇒ 2 HOH +1.80
HSO− 2−
3 +HOH→SO4 +3 H +2 e
+ −
−0.17 H−O−O−R alkyl hydroperoxides
NO− 2 +HOH→NO −
3 +2 H +
+2 e −
−0.94 R−O−O−R dialkyl peroxides
2 Cl− ⇒ Cl2 + 2 e− −1.36 H−O−O−Ac percarboxylic acids
− −
2 Br ⇒ Br2 + 2 e −1.07 Ac−O−O−Ac diacyl peroxides
2 I− ⇒ I2 + 2 e− −0.54
pH 14 Hydrogen peroxide forms peroxohydrates
HOOH + 2 e− ⇒ 2 OH− +0.87
Mn(OH)2 + 2 OH− ⇒
with a number of compounds. Addition com-
MnO(OH)2 + HOH + 2 e− +0.05 pounds with sodium carbonate (sodium carbon-
* Standard potential of redox reactions measured against a ate peroxohydrate [15630-89-4]) and with urea
hydrogen electrode (25 ◦ C, 100 kPa). (urea peroxohydrate [124-43-6]) are industri-
ally important. For a detailed description of per-
The reactions of hydrogen peroxide oxo compounds see → Peroxo Compounds, In-
with potassium permanganate [23] or organic; → Peroxy Compounds, Organic.
cerium(IV)sulfate are used for quantitative de-
termination (volumetric titration) of hydrogen
peroxide content.
6 Hydrogen Peroxide
4. Production
4.1. Anthraquinone Process (AO
Process)
4.1.1. Principles
In the AO process, 2-alkyl-9,10-anthraquinones
react with hydrogen in the presence of a catalyst
to form the corresponding hydroquinones.
After the catalyst is removed (otherwise, the hy-
drogen peroxide would decompose), the hydro-
quinones are oxidized to quinones with oxygen
(usually air) with simultaneous quantitative for-
mation of hydrogen peroxide:
Hydrogen peroxide is extracted with water, and
the quinones are returned to the hydrogenator to
complete the loop.
The AO process, therefore, leads to the net
formation of hydrogen peroxide from gaseous
hydrogen and oxygen.
Solvents. Anthraquinones must be dissolved
in a suitable solvent for hydrogenation, oxida-
tion, and extraction (the so-called working so-
lution). Although a number of individual sol-
vents have been proposed (e.g., aromatic al-
cohols or their esters [24]), solvent mixtures
are almost always used because the quinones
and hydroquinones have different solubilities.
Quinones dissolve readily in nonpolar, aromatic
solvents (quinone solvents). Hydroquinones dis-
solve well in polar solvents, especially alcohols
and esters (hydroquinone solvents). Proposed
solvents or solvent mixtures are
Quinone solvents
– Benzene, xylene [25, 28]
– tert-Butylbenzene [26]
– Trimethylbenzene [28]
– Tetramethylbenzene (isodurene)[29]
– Mixtures of polyalkylated benzenes, e.g., C9-
C11 aromatic solvents [30, 31]
– Methylnaphthalene [32]
Hydroquinone solvents
– Alkyl phosphates, e.g., tris-(2-ethylhexyl)-
phosphate [33]
– Nonyl alcohols, e.g., diisobutylcarbinol [32]
– Alkylcyclohexanol esters, e.g., 2-methyl-cy-
clohexylacetate [34]
– N,N-Dialkyl carbonamides, e.g., N,N-dibutyl
propionamide [35]
– N-Alkyl N-aryl carbonamides e.g. N-ethyl-
N-phenyl benzamid [36]
– N,N-Dialkyl carbamates, e.g., 2-ethylhexyl-
N-butylcarbamate [37]
– Tetraalkyl ureas, e.g., tetra-n-butylurea [42]
– Cycloalkyl ureas, e.g., dihexyl propyleneurea
[38]
– Phenylalkyl ureas, e.g., N,N-dibutyl-N  -
methyl-N  -phenylurea [39]
– N-Alkyl-2-pyrrolidones, e.g., octyl pyrroli-
done [40]
– N-Alkyl caprolactams, e.g., n-octyl capro-
lactam [41]
Solvent mixtures [43, 44]:
– Polyalkylated benzenes and alkyl phosphates
[30]
– Polyalkylated benzenes and tetraalkyl ureas
[42]
– Trimethylbenzenes and alkylcyclohexanol es-
ters [28]
– Methylnaphthalene and nonyl alcohols [32]
The following criteria must be fulfilled when
choosing solvents and preparing solvent mix-
tures:
1) Good solubility of both quinone and hydro-
quinone
2) Good stability in the hydrogenator and oxi-
dizer
3) Low solubility in water and aqueous hydrogen
peroxide solutions
4) Sufficiently lower density than water to ensure
separation of the two phases during extraction
5) Low volatility, i.e., high boiling point and
flash point
6) High distribution coefficient for hydrogen
peroxide in the solvent–water system
7) Low toxicity

Quinones. Criteria similar to those used for
selecting solvents also apply to the quinone or
quinone mixture which is used as the working
compound:
1) Good solubility of the quinone form
2) Good solubility of the hydroquinone form
3) Good resistance to oxidation
4) Good availability
The choice of quinone depends mainly on the
properties, especially the solubility, of the prod-
ucts formed in the AO process. Examples of
working compounds are 2-ethylanthraquinone
[84-51-5] [45], 2-tert-butylanthraquinone [84-
47-9] [46], 2-neopentylanthraquinone [123257-
96-6] [47], 2-isohexenylanthraquinone [71308-
16-2] [48], diethylanthraquinones [49], mixtures
of 2-amylanthraquinones [54], and mixtures of
different alkylanthraquinones [50–53].
The formation of degradation products and
their ability to be regenerated to active quinones
also play a role in the decision. In addition to
hydroquinone formation, a number of secondary
reactions occur during the hydrogenation step;
of these, hydrogenation of the anthraquinone
ring system is particularly important. One ring
of the anthraquinone molecule, preferably the
one that is not substituted by an alkyl group, is
hydrogenated. In the case of R=ethyl, this is ac-
complished with selectivities higher than 90 %.
2-Alkyl-5,6,7,8-tetrahydro-9,10-dihydroxy-
anthracene is formed, which is also oxidized by
oxygen to regenerate the 2-alkyl-5,6,7,8-tetra-
hydroanthraquinone (known as “tetra”), with
simultaneous quantitative formation of hydro-
gen peroxide.
Hydrogen Peroxide 7
Although “tetra” is more readily hydro-
genated than a 2-alkylanthraquinone, the result-
ing “tetra” hydroquinone is much more difficult
to oxidize than the readily oxidizable anthrahy-
droquinone. The kinetics and basic physico-
chemical parameters of tetrahydroquinone ox-
idation have been studied [55–57]. Small
amounts of amines accelerate hydroquinone ox-
idation [58].
The formation of “tetra” in the synthesis loop
depends on process conditions and has led to two
methods for carrying out this process: the “an-
thra” system and the “all-tetra” system.
Anthra System. The slower oxidation rate
of the “tetra” hydroquinone causes difficulties
in the oxidizer. Actions have therefore been pro-
posed which suppress the formation of “tetra” or
dehydrogenate it to anthraquinone.
When the “tetra” content of the working solu-
tion is kept low (i.e., if 2-alkylanthrahydroqui-
none is formed almost exclusively during hy-
drogenation), the process is referred to as the
“anthra” system.
“Tetra” formation is suppressed by using se-
lective catalysts [59], special solvents [60], spe-
cial working compounds [61], and mild hydro-
genation conditions [62] (e.g., the use of olefins
has been recommend for the steady dehydroge-
nation of “tetra” [63]).
“Tetra” can be dehydrogenated in the pres-
ence of activated aluminum oxide:
When the “anthra” system is used, tautomeric
2-alkyl-10-hydroxy-9-anthrone (oxanthrone) is
also formed during hydrogenation [13, 64]:
8 Hydrogen Peroxide

Oxanthrone is not oxidized by oxygen to

form hydrogen peroxide and therefore leads
to loss of active quinone. Further hydrogen-
ation of oxanthrone leads to anthrone and tetra-
hydroanthrones [65]. Although oxanthrone and
anthrone can be regenerated to active quinones,
subsequent oxidative dimerization of anthrone
may lead to dianthrones which represent a loss
of quinone [13].
This means that the hydroquinone leaving the
hydrogenation step is always the “tetra” hydro-
quinone. Dianthrone formation does not occur
in the “all-tetra” system, but oxidation of “tetra”
hydroquinone leads to a byproduct, the so-called
epoxide:

The epoxide does not participate in the for-

mation of hydrogen peroxide and leads to a loss
of active quinone. The formation of epoxides
seems to be minimized when using high oxygen
The buildup of anthrones and oxanthrones content [69], or mixing hydrogenated working
can be minimized by adding an aromatic ter- solution with a portion of oxidized working so-
tiary amine to the working solution [66] or by lution containing hydrogen peroxide and then
pretreatment of the hydrogenation catalyst with feeding to the oxidizer [70]. Different measures
halides [67]. have been suggested for regenerating “tetra”
from the epoxide [71, 73] (see Chapter 4.1.2.4)
“All-Tetra” System. If “tetra” formation is
not suppressed during hydrogenation or “tetra”
is not dehydrogenated, an equilibrium is reached
in which the hydroquinone charged to the ox-
idizer consists exclusively of 2-alkyl-5,6,7,8-
tetrahydroanthrahydroquinone. Such a system is
called an “all-tetra” system.
If anthraquinone is hydrogenated in addition
to “tetra” during the hydrogenation step, hydro-
gen transfer occurs [68]:
 Hydrogen Peroxide 9

If the “tetra” content of the “all-tetra” 4.1.2. Process Description

system is high, another byproduct can be
formed, namely, 2-alkyl-1,2,3,4,5,6,7,8-octa-
hydroanthraquinone [13]:
The cyclic Riedel–Pfleiderer or BASF process
(Chap. 1) forms the technological basis for all
modern AO processes. Developments include
improvement of the individual process steps, use
of stable working solutions, and use of selective
hydrogenation catalysts. The basic principles of
the process are illustrated in Figure 5.

Hydrogenation. From the storage tank or

Although this “octa” hydroquinone is oxi-
dized by oxygen to quinone with the formation
of hydrogen peroxide, the reaction is too slow to
be important in the formation of hydrogen per-
oxide. “Octa” thus represents a decomposition
product from which quinone cannot be regener-
ated.
Octa formation depends on hydrogenation
conditions and catalyst type. Other degradation
products may be formed besides octa [65, 74,
75].
hydrogenation feed tank (a), the working solu-
tion enters the hydrogenator (b) where it is hy-
drogenated in the presence of a suspended, sup-
ported, or fixed-bed catalyst.
If a suspended catalyst (e.g., palladium black
or Raney nickel) or a supported catalyst (e.g.,
palladium) is used, the hydrogenation step in-
cludes a main filtration stage which retains the
catalyst and allows it to be returned to the hy-
drogenator. The heat of reaction released during
hydrogenation can be removed (1) before hydro-
genation (by cooling the oxidized working solu-

Figure 5. Production of hydrogen peroxide by the anthraquinone (AO) process

a) Storage tank for working solution or hydrogenator feed tank; b) Hydrogenator; c) Safety filtration; d) Oxidizer; e) Separator;
f) Activated carbon adsorber; g) Extraction; h) Drying; i) Prepurification; j) Crude product storage tank; k) Hydrogen perox-
ide concentration; l) Hydrogen peroxide storage tank; m) Demineralized water feed tank; n) Regeneration and purification;
o) Solvent storage tank; p) Working solution make up tank; q) Catalyst regeneration
10 Hydrogen Peroxide
tion), (2) during hydrogenation (by cooling the
reactor), or (3) after hydrogenation (by cooling
the hydrogenated working solution).
Oxidation. Before the hydrogenated work-
ing solution that contains hydroquinone can be
fed to the oxidation step, it must pass through
a safety filtration stage (c). This is particularly
important because the hydrogenation catalysts
used in the AO process (palladium and Raney
nickel) also catalyze the decomposition of hy-
drogen peroxide. Even a small amount of these
catalysts in the oxidation and extraction steps
would lead to considerable loss of hydrogen per-
oxide and serious disturbances.
During the oxidation step (d), the hydro-
genated working solution is gassed with air. Dis-
solved hydroquinones are oxidized to quinones,
and hydrogen peroxide is formed.
After the working solution has been separated
(e1 ), air from the oxidation step passes over ac-
tivated carbon adsorbers (f), and adsorbed sol-
vent is recovered from them. Several adsorbers
are usually loaded and regenerated alternately,
for example, with steam.
Extraction and Recovery of the Working
Solution. The oxidized working solution is then
treated with water to extract hydrogen peroxide
(g).
The working solution leaving the extraction
unit must be adjusted to a specific water con-
tent before being returned to the hydrogenation
step. Free water taken up by the working solu-
tion during extraction is separated (e2 ) and the
water content is adjusted to the desired level in
the drier (h).
To purify the working solution and regenerate
the quinone decomposition products into active
quinones, some or all of the working solution is
passed through a regeneration step (n).
Hydrogen Peroxide Concentration. Crude
aqueous hydrogen peroxide from the extraction
stage (H2 O2 concentration 15–40 wt %) is fed
into the crude product storage tank (j) via a pre-
purification unit (i).
From the crude product storage tank, aqueous
hydrogen peroxide goes to the concentration unit
(k) where it is distilled. Here, hydrogen peroxide
is freed from most of its impurities and concen-
trated to the commercial concentration of 50–70
wt %; it is then collected in a storage tank (l).
Water vapor produced during distillation is
condensed and fed into the water storage tank
(m).
Auxiliary Processes. A number of addi-
tional processes are required to maintain the AO
operation. For example, to maintain hydrogen-
ation activity, part of the catalyst is removed,
regenerated in the catalyst regeneration area (q),
and returned to the hydrogenator. To compensate
for quinone and solvent losses, working solution
is periodically made up with anthraquinone and
solvent (o, p).
4.1.2.1. Hydrogenation
The hydrogenation step is the most important
step of modern AO processes. Quinone decom-
position products that cannot be regenerated into
active quinone are formed during this step (cf.
Section 4.1.1). New hydrogenation catalysts and
hydrogenation reactors have been developed that
deviate totally from the BASF principle. Here,
design of the hydrogenator depends largely on
the type of catalyst used.
Four typical reactors for the three usual cat-
alyst systems (suspended, supported, and fixed-
bed catalysts) are discussed in this section.
BASF Hydrogenation Step. The hydrogen-
ation step in the BASF plant (Fig. 6) uses a Raney
nickel catalyst at a slight excess pressure of ap-
proximately 0.2 MPa and at 30–36 ◦ C. Because
Raney nickel is sensitive to oxygen, the work-
ing solution from the extraction or drying and
purification steps cannot be fed directly into the
hydrogenator. This working solution still con-
tains residual hydrogen peroxide and must be
passed over a decomposition catalyst (e.g., sup-
ported Ni–Ag), together with a fraction of the
hydrogenated working solution (which also con-
tains hydroquinone), to remove hydrogen perox-
ide completely:

Figure 6. BASF hydrogenation step
a) Precontact column; b) Hydrogenator feed tank; c) Stirred tank reactor with internal filters; d) Catalyst feed tank; e) Safety
filtration; f) Catalyst separator; g) Pump; h) Oxidizer feed tank
The solutions are passed through the precon-
tact column (Fig. 6, a) and collected in the hy-
drogenator feed tank (b). The working solution
is then pumped into the stirred vessel reactor
(c) and is gassed with hydrogen in the pres-
ence of Raney nickel. Periodic addition of small
amounts of hydrogenation catalyst from the cat-
alyst feed tank (d) allows a constant rate of hy-
drogen conversion in the hydrogenator. Hydro-
genated working solution is collected in the ox-
idizer feed tank (h) through the internal filters
in the stirred vessel, thus exploiting the excess
pressure in the reactor. The solution is then led
into the oxidation step via the safety filter (e).
A sidestream of hydrogenated working solution
is withdrawn and recycled to the precontact col-
umn (a).
When the concentration of Raney nickel in
the hydrogenation reactor reaches a certain limit,
the content of the reactor is drained into the cat-
alyst separator (f). Raney nickel settles to the
bottom, and catalyst-free supernatant is pumped
back to the hydrogenator.
A significant disadvantage of Raney nickel
as catalyst is its limited selectivity, i.e., the ratio
of hydroquinone formation to “tetra” formation.
BASF largely eliminated this by pretreating the
catalyst with ammonium formate [76].
Alternatives were subsequently suggested for
pretreating Raney nickel (e.g., nitriles [77],
amines [78], and aldehyde solutions [79]). In re-
Hydrogen Peroxide 11
cent years, more selective nickel catalysts, based
on Ni-B or Ni-Cr-B alloys were developed [80].
The pyrophoric properties of Raney nickel
also require more stringent safety procedures
when handling the material. Raney nickel is still
used today in some AO plants, but palladium
catalysts are preferred because of their higher
selectivity and simpler handling.
Degussa Hydrogenation Step. Degussa
proposed the use of palladium black as the
hydrogenation catalyst. This exploits the advan-
tages offered by a suspended catalyst and avoids
the disadvantages of Raney nickel. Equipment
that allows good conversion of hydrogen with
very finely distributed palladium black is shown
in Figure 7 [81].
The most important feature of the loop reac-
tor (c) is the connection in series of pipes with
different diameters. The working solution flows
downward in the large pipes at a rate of 0.7–1.5
m/s and flows upward in the narrower pipes at
1.5–3 m/s. Efficiency can be improved by in-
stallation of internal plate-stacks with a channel
width of 10-50 mm [82] or external premixing
of gas and liquid [83].
Degussa proposed a carbon filter [84]. Ce-
ramic [85] and cross-flow [86] filters can also be
used. A decline in filter performance can be over-
come by periodic back flush with hydrogenated
working solution through the filter into the hy-
drogenator.
12 Hydrogen Peroxide

Figure 7. Degussa hydrogenation step

a) Reactor circulating pump; b) Filter; c) Loop reactor; d) Separator; e) Oxidizer feed tank; f) Back flush pump

Advantages of this hydrogenation system are

1) Almost complete conversion of hydrogen
2) Nonpyrophoric catalyst
3) Easy exchange of palladium black
4) Easy regeneration of the catalyst

Laporte Hydrogenation Step. Laporte

Chemicals and other companies proposed the
use of supported palladium catalysts [87]. Sup-
ports based on alumina [88], silica [89], and
sodium aluminum silicate [90] are used. These
catalysts have the advantage that their particle
diameter in the range of 0.02–0.2 mm makes
their filtration and recirculation to the reactor
simpler than those of palladium black.
Laporte proposed the apparatus shown in Fig-
Figure 8. Laporte hydrogenation step
ure 8 for industrial hydrogenation [91]. The re- a) Hydrogenator; b) Reactor tubes; c) H2 compressor;
actor contains a series of tubes whose lower ends d) Internal filter
lie just above the bottom of the reactor and whose
top ends are just below the liquid surface. Hy-
drogen is fed into the bottom of each tube, and Fixed-Bed Hydrogenation Step. Fixed-
very small gas bubbles are formed by distribu- bed hydrogenation represents a simple solu-
tors. Upward flow occurs in the tube due to the tion for the hydrogenation step; it involves a
density difference between the solutions in the palladium catalyst and avoids the problem of
tube and in the reactor. The catalyst suspension filtration and recirculation of catalyst into the
is drawn into the pipe by the continuous flow of reactor. The first industrial fixed-bed hydrogen-
working solution. To obtain a sufficiently high ation unit for the AO process was commissioned
airlift effect in the tube, hydrogen must be cir- by FMC (Fig. 9) [94].
culated continuously. Continuously stirred tank
reactors (CSTR) [92] and bubble-column reac-
tors are also in use.
 Hydrogen Peroxide 13

2) A long working life because replacing a fixed-

3) Good productivity
4) Easy regeneration of the catalyst
Figure 9. FMC hydrogenation step
a) Fixed-bed reactor; b) Catalyst; c) Reactor circulating
pump; d) Heat exchanger; e) Hydrogenator feed pump
Most fixed-bed catalysts have a diameter of
0.2–5 mm and a palladium load of 0.1–0.5 %.
The working solution is pumped to the top of
the reactor. A side stream of the hydrogenated
working solution is also fed into the fresh work-
ing solution after the heat of reaction has been
removed in a heat exchanger (d). The catalyst
must fulfill a number of requirements such as
1) High abrasion resistance to allow simplifica-
tion of the filtration step
bed catalyst is more complicated than replac-
ing a suspended catalyst
External [98] or internal [99] static mixers
may be used to ensure good liquid/gas disper-
sion. To enhance performance, operation in the
foam regime with a liquid/gas ratio not higher
than 1.5/10, temperatures of 45–75 ◦ C, and a
pressure of 0.18–0.5 MPa [93] is proposed.
A fixed-bed reactor with honeycomb mono-
liths is described by EKA [95]. The monoliths
contain parallel channels with a diameter of 1–3
mm. Applied to the walls is a thin layer of porous
silica gel, coated with palladium.
Desactivation plays an important role when
operating fixed-bed catalysts. Impact of factors
like type of support [96] or water [97] on side
reactions was studied. Several measures, e.g.,
treatment with steam and hydrogen peroxide
[100] or acid wash [101] are described for cata-
lyst regeneration.

Figure 10. BASF oxidation step

a) Separator; b) Oxidation columns; c) Siphon; d) Extractor feed tank; e) Compressor
14 Hydrogen Peroxide
Figure 11. Degussa oxidation step
a) Concurrent-flow oxidation reactors; b) Separator; c) Air compressor
In recent years, especially AO plants built in
China were equipped with fixed-bed hydrogen-
ation.
4.1.2.2. Oxidation
In industrial plants, the catalyst-free hydro-
genated working solution is usually oxidized
with slightly pressurized air (up to 0.5 MPa).
The oxidizer off-gas is then directed to a series
of activated carbon adsorbers to purify it and re-
cover the solvents escaping with it.
For economic reasons, the following objec-
tives are sought in designing industrial oxidation
reactors:
1) Efficient utilization of atmospheric oxygen to
reduce the volume of the off-gas and the size
of activated carbon adsorbers
2) Low compressor pressure to decrease energy
costs
3) Small reactor volume or reduced holdup of
working solution to lower investment costs
for equipment and working solution
BASF Oxidation Step (Fig. 10). A working
solution containing benzene as the quinone sol-
vent has been used at BASF. For safety reasons
and to minimize the volume of off-gas, the BASF
oxidation step uses a nitrogen–oxygen mixture.
The mixture is cycled in a closed circuit, and
its composition is kept constant by adding pure
oxygen to compensate for the amount consumed
in the formation of hydrogen peroxide.
Hydrogenated working solution enters a sep-
arator (a) and then flows through four oxidation
columns (b) arranged in series as a cascade. Ox-
idized working solution flows into the extractor
feed tank (d) via a siphon (c). The nitrogen–
oxygen mixture is compressed and fed into each
of the four reactors.
Off-gas from the reactors is collected in a pipe
and separated from entrained droplets of work-
ing solution. After compensation for consumed
oxygen, the gas mixture is returned to the com-
pressor (e).
Degussa Oxidation Step (Fig. 11) [102].
The hydrogenated working solution flows cocur-
rently upward with the oxidizing air through the
first reactor. The working solution and air are
separated in the upper part of the reactor. Off-
gas from the reactor is fed to the activated car-
bon adsorption unit, and working solution flows
into the second and third reactors, which are in-
stalled in series. The air needed for oxidation is
pressurized with a compressor (c) and fed to the
third reactor. The working solution and air flow

countercurrently, but both streams are divided
into cocurrent flow segments.
Laporte Oxidation Step (Fig. 12). In La-
porte’s Warrington plant, oxidation was carried
out in a packed cocurrent flow column [103,
104]. The entire volume of the oxidation reac-
tor (a) was used for air gassing. The air and the
working solution leaving the top of the column
were fed together into a separator (b). The air
then reached the alternately operated two-stage
activated carbon adsorbers (c), and the work-
ing solution passed to the extraction stage. To
enhance gas-liquid mass transfer, premixing of
working solution and air by a venturi tube is pro-
posed [105]. The use of a tubular reactor with
static mixing elements is described by Kemira
in [106].
Figure 12. Laporte oxidation step
a) Oxidation reactor; b) Separator; c) Activated carbon ad-
sorber; d) Air compressor
Allied Chemical Oxidation Step. To
shorten the residence time of the working so-
lution in the oxidation step, Allied Chemical
suggested a counterflow reactor in which res-
idence times are less than 2.5 min at a partial
oxygen pressure of 70–90 kPa [107]. Space–
time yield can be improved by using perforated
trays with small hole size (cross-sectional area
of holes below 3 mm2 ) [108].
Hydrogen Peroxide 15
4.1.2.3. Extraction and Drying
In the BASF plant, a sieve-tray extraction col-
umn extracts hydrogen peroxide from the work-
ing solution. A number of other extractors, such
as packed columns, pulsed packed columns, and
Podbielniak extractors, have been proposed.
Besides extraction column design, the ex-
traction efficiency is strongly influenced by the
distribution coefficient, which is depending on
working solution composition (e.g., type of hy-
droquinone solvent).
It has been found, that by injection of a
small amount of air into the extraction column
(gas-agitated sieve plate column) extraction effi-
ciency is increased by 30–40 % [109]. Processes
which integrate oxidation and extraction in a sin-
gle apparatus are also known [112, 113].
Working solution leaving the extraction unit
contains dispersed water droplets and is led into
coagulators and separators to separate free wa-
ter. The working solution is then dried to a spe-
cific water content.
Because the solubility of water in the work-
ing solution depends on temperature, its mois-
ture content can be adjusted by carrying out the
extraction at low temperature, separating the dis-
persed water, and then increasing the tempera-
ture of the working solution by about 30 ◦ C be-
fore it reaches the hydrogenation step [110]. The
working solution may be alternatively dried by
using the off-gas from the oxidation step [111].
In extraction and drying according to BASF
(Fig. 13), working solution leaving the head of
the extraction column (a) is initially freed from
entrained water in a water separator (b). The
working solution then passes through aqueous
potassium carbonate solution for drying (c).
4.1.2.4. Working Solution Purification and
Regeneration
During the constant circulation of working solu-
tion, degradation products are formed not only
from the working compound, but also from the
solvents. These degradation products must be
removed from the working solution to prevent
(1) deterioration of the crude hydrogen peroxide
(color, smell, dissolved organic compounds), (2)
16 Hydrogen Peroxide

Figure 13. BASF extraction and drying step

a) Sieve–tray extraction column; b) Water separator; c) Drier; d) Potassium carbonate solution separator; e) Potassium car-
bonate solution make-up and feed tank; f) Potassium carbonate concentration

an increase in density and viscosity of the work-
ing solution, and (3) a decrease in interfacial ten-
sion of the working solution, which promotes
the formation of an emulsion during extraction.
In addition, degradation products decrease the
activity and lifetime of the hydrogenation cata-
lysts. Regeneration and purification can be per-
formed with hydrogenated working solution or
with working solution after extraction.
Numerous methods have been suggested for
purifying the working solution and regenerat-
ing active quinone from the quinone degradation
products. Examples include the following:
1) Treatment of the hydrogenated working solu-
tion with alkaline substances [71] or aqueous
alkali hydroxide [72]
2) Treatment of the working solution with
sodium aluminum silicates at 50–200 ◦ C
[114]
3) Treatment of the working solution after hy-
drogenation or extraction with active alu-
minum oxide at 20–150 ◦ C [115–118]
4) Treatment with alkali hydroxide, calcium hy-
droxide, ammonia, or amines in the presence
of oxygen or hydrogen peroxide [119]
5) Treatment of the working solution with aque-
ous sodium hydroxide or potassium hydrox-
ide solution (8–17 mol/L) in the presence of
oxidants [120]
6) Treatment with solvent and liquid carbon
dioxide [121]
7) Treatment with a noncyclic hydrocarbon
[122]
4.1.2.5. Purification of Crude Hydrogen
Peroxide
The aqueous hydrogen peroxide leaving the ex-
traction is impure. Separators are usually in-
stalled after extraction (Section 4.1.2.3) [123].
Measures for reducing the amount of dissolved
organic compounds include treatment of the
crude product with polyethylene [124], ion ex-
changers [125, 126], or hydrocarbons [127,
128]. Water-soluble organic compounds can
also be oxidized by heating and then extracted
with suitable solvents (e.g., the quinone solvent)
[129]. The effect of purification on carbon con-
tent of crude hydrogen peroxide is illustrated in
Table 5.
 Hydrogen Peroxide 17
Table 5. Processes for purifying crude hydrogen peroxide
Process Carbon content of H2 O2 , mg/L Reference
Before After
purifi- purifi-
cation cation
Extraction with xylene 580 450 [130]
Extraction with methylcyclo-
hexanol acetate 580 390 [130]
Extraction with xylene and
methylcyclohexanol acetate 580 370 [130]
Brief treatment with 0.5 wt %
activated carbon 134 11 [131]
Treatment with adsorber resin 320 140 [132]
1000 100 [133]
Reverse-osmosis membrane 130 18 [134]
Supercritical carbon dioxide 145 84 [135]

Purified crude product is then usually fed to
a distillation unit where it is purified further and
concentrated to the usual commercial concen-
tration of 50–70 wt % hydrogen peroxide. This
product can be used directly in a number of ap-
plications. Reverse osmosis membranes can also
be used to purify crude, distilled or concentrated
(up to 70 %) hydrogen peroxide [134, 137]. A
disadvantage of this procedure is that only part
of the feedstream leaves the unit as purified per-
meate, while the impurities are accumulated in
the retentate.
For use in the electronics industry, hydrogen
peroxide with a very low content (ppt level) of
metal impurities is needed. Many patents deal
with further purification of crude or prepuri-
fied hydrogen peroxide with a combination of
cation- and anion- exchange resins [136–140].
Some methods make use of of reverse osmosis
in combination with other methods [134, 141,
142], adsorption of metal impurities on activated
carbon [143], stannic oxide [144], or zirconium
phosphate [145].
4.1.2.6. Concentration of Hydrogen
Peroxide
Water and hydrogen peroxide do not form an
azeotrope; their boiling point difference at at-
mospheric pressure is 50.2 ◦ C (Table 1). Dis-
tillation of water from aqueous hydrogen per-
oxide can therefore be used to concentrate the
solution or to completely separate hydrogen per-
oxide from water. A number of technical safety
requirements must be observed. Distillation of
dilute solutions has been used industrially to pro-
duce commercial 50–70 wt % hydrogen perox-
ide solution; the technology is described in [3, 5,
146–152]. The use of structered packings [150]
in distillation columns enhances perfomance.
Concentrations higher than 90 % are preferably
produced by crystallization [153, 154].
4.2. 2-Propanol Process (Shell Process)
In 1945 Harris discovered that primary and sec-
ondary alcohols react with oxygen to form hy-
drogen peroxide and an aldehyde or ketone, re-
spectively [155]:
RCH2 OH+O2 →RCO+H2 O2
R2 CHOH+O2 →R2 C=O+H2 O2
Oxidation of the alcohol can be carried out in
the liquid or the gas phase. Because aldehydes
formed in the reaction with primary alcohols are
oxidized easily, only the oxidation of secondary
alcohols, especially 2-propanol [67-63-0], has
industrial importance.
A process for the simultaneous production
of hydrogen peroxide and acetone [67-64-1] by
atmospheric oxidation of 2-propanol was first
suggested by Rust [156, 157].
Reaction of 2-propanol with oxygen in the
liquid phase does not require a special catalyst
because it is catalyzed by the hydrogen peroxide
product. However, a small amount (0.5–1 wt %)
of hydrogen peroxide is added to 2-propanol to
shorten the induction period. The reaction pro-
ceeds as follows:
18 Hydrogen Peroxide
HOOH→2 HO•
Secondary reactions also take place but are
not discussed here. To reduce the formation of
byproducts, particularly acetic acid, which af-
fects the quality of hydrogen peroxide, only part
of the 2-propanol is oxidized, and oxidation is
carried out in several consecutive steps at de-
creasing temperature [158].
The 2-propanol process is used only in
two plants in the former Soviet Union, which
have been operating since 1968 and 1972. The
Shell Oil Company decommissioned its plant in
Norco, Louisiana in 1980 [159, 160]; that plant
had been in operation since 1957. The technol-
ogy of the Shell 2-propanol process is described
in [161]. The hydrogen peroxide yield of the
Russian plants based on 2-propanol or atmo-
spheric oxygen is 87–98 %; the acetone yield
is 92–94 % [162, 163].
An alternative process, which makes use of
the liquid phase oxidation of methylbenzyl al-
cohol, coproduct in the production of propylene
oxide by epoxidation of propylene with ethyl-
benzene hydroperoxide, was developed by Arco
but has not been commercialized [164–166].
Similar to the AO process the operations include
hydrogenation, oxidation, extraction, and purifi-
cation steps. Oxidation of methylbenzyl alcohol
with air at elevated temperature yields hydrogen
peroxide and acetophenone. Methylbenzyl alco-
hol conversion in the reactor is 30 % with H2 O2
selectivity about 80 %. The hydrogen peroxide
content in the organic phase at the exit of the
oxidizer is approximately 5 wt %. After addi-
tion of ethylbenzene hydrogen peroxide is ex-
tracted with water. The organic phase overflow-
ing at the top of the extractor contains methyl-
benzyl alcohol, ethylbenzene and acetophenone.
Whereas ethylbenzene is distilled off and re-
cycled to the extractor acetophenone is hydro-
genated to methylbenzyl alcohol. Advantages of
the ARCO process compared with the AO pro-
cess are: cheap working solution, methylbenzyl
alcohol is used both as solvent and as working
compound and high hydrogen peroxide concen-
tration in the working solution after oxidation.
4.3. Electrochemical Processes
Electrochemical production of hydrogen perox-
ide is of minor industrial importance compared
to the AO process.
Because electrochemical processes played an
important part in the development of hydrogen
peroxide and their technology reached a high
standard, essential features of the three known
cyclic processes are described below. These pro-
cesses are described in detail in [3, 5].
The basis of electrochemical processes is an-
odic oxidative coupling of sulfate ions to per-
oxodisulfate ions. Sulfuric acid is oxidized in the
Degussa–Weissenstein process, whereas ammo-
nium hydrogensulfate solutions are used in the
Münchner process.
2 H2 SO4 →H2 S2 O8 +H2
2 (NH4 ) HSO4 →(NH4 )2 S2 O8 +H2
Hydrogen peroxide is formed by hydrolysis
of the peroxodisulfate ion via the peroxomono-
sulfate anion:
S2 O2− − −
8 +H2 O→HSO4 +SO3 (OOH)
SO3 (OOH)− +H2 O→HSO−
4 +HOOH
In the Degussa–Weissenstein and Riedel–
Loewenstein processes, peroxodisulfuric acid
or ammonium peroxodisulfate derived from
 Hydrogen Peroxide 19
Table 6. Yields and energy consumption of electrochemical processes
Yield or consumption* Degussa–Weissen-stein Münchner process Riedel–Loewenstein
process process
Current yield, % 73 83 83
Hydrolysis and distilla-
tion yield, % 93–96 87 82–85
Total yield, % 68–70 72 68–70.5
Electrical energy, kW · h 1280 1760 1700
Steam, t 1.85 2.5 4.5–4.8
Cooling water, m3 200 240 250
* Parameters are expressed per 100 kg of hydrogen peroxide (calculated as 100 wt %) in the form of a distilled 35 wt % solution.

the electrolysis is hydrolyzed directly. In the 4.3.1. Degussa–Weissenstein Process

Münchner process, ammonium peroxodisulfate
is converted to potassium peroxodisulfate,
which is then hydrolyzed. The Münchner pro-
cess, therefore, consists of several coupled cyclic
processes.
All three processes have a very high energy
consumption and an unsatisfactory yield of ca.
70 % (Table 6).
Huron developed a modified Loewenstein
process, which makes use of platinum–titanium
anodes and achieves a yield of 85 % [167]
In the Degussa–Weissenstein process (Fig. 14),
sulfuric acid is adjusted to a concentration of
550–570 g/L (a) before flowing through the
catholyte chamber (b) where residual peroxide is
destroyed and heavy-metal ions are precipitated.
After addition of chemicals to raise the electro-
chemical potential (e.g., ammonium rhodanide
or hydrochloric acid), sulfuric acid passes to the
anolyte chamber where peroxodisulfuric acid is
formed. Hydrolysis of peroxodisulfuric acid and

Figure 14. Degussa–Weissenstein process

a) Electrolyzer feed tank; b) Electrolyzer; c) Evaporation and hydrolysis; d) Hydrogen peroxide expulsion and hydrolysis;
e) Acid separator; f) Water–hydrogen peroxide fractionating column; g) Acid purification
20 Hydrogen Peroxide
distillation of hydrogen peroxide are carried out
in two steps (c, d). Hydrolysis is 80–90 % com-
pleted in (c), and distillation is achieved in ap-
proximately equal parts in two steps (c, d). In
the expulsion column (d), hydrogen peroxide is
expelled from the concentrated sulfuric acid so-
lution by countercurrent steam. Entrained acid is
separated in an acid separator (e), and hydrogen
peroxide–water vapors are separated in a frac-
tionating column (f). An aqueous solution with
a maximum content of 45 wt % hydrogen per-
oxide is obtained. To purify the circulating so-
lution, part of the stream is distilled constantly
(g).
4.3.2. Münchner Process
In the Münchner process (Fig. 15), acid ammoni-
um sulfate is electrolyzed in electrolyzers with-
out diaphragms.
The circulating solution contains 300 g of
(NH4 )2 SO4 , 40 g of K2 SO4 , 100 g of H2 SO4 ,
and 57.6 g of S2 O2−8 per liter before entering the
electrolyzer. Its peroxodisulfate concentration is
increased to 115.2 g/L by the electrolytic pro-
cess. After electrolysis, the circulating solution
is cooled and concentrated in a vacuum cooler
(b) by evaporation of water and then fed into
a crystallizer (c). Half of the peroxodisulfate is
precipitated as potassium peroxodisulfate by ad-
dition of potassium hydrogen sulfate and sepa-
rated in a centrifuge (d). The supernatant mother
liquor diluted with washing water is returned
to the electrolyzer. Solid potassium peroxodi-
sulfate is hydrolyzed (e) and hydrogen peroxide
is steam distilled.
The remaining potassium hydrogensulfate–
sulfuric acid residue is separated in a centrifuge
(f). Sulfuric acid is recycled to the hydrolysis
and distillation stage (e) via a purification stage
(g). Potassium hydrogensulfate is returned to
the crystallizer (c) for precipitation of peroxo-
disulfate.
To purify the circulating solution, a side
stream is mixed with ammonia in the purifica-
tion step (h) to precipitate iron. The hydrogen
peroxide–water vapors formed during hydroly-
sis and distillation are led through an acid sep-
arator (i) and separated in the rectification step
(j). Because hydrolysis is carried out batchwise,
the initial vapor mixture is rich in hydrogen per-
oxide, giving a distillate of 60 wt % which then
drops to 20 wt %.
4.4. Other Processes
A number of reactions and processes for the pro-
duction of hydrogen peroxide are described in
the literature. Some of them are no longer eco-
nomical (e.g., the barium peroxide, azobenzene.
Only a few examples can be described here;
detailed information may be found in the litera-
ture [1, 3, 5].
4.4.1. Production from Peroxy Compounds
Peroxy compounds can be decomposed to form
hydrogen peroxide. Examples of such reac-
tions, which include the historically important
Thenard reaction of barium peroxide with nitric
or sulfuric acid, are listed in Table 7.
Table 7. Production of hydrogen peroxide from peroxy compounds
Peroxide Reaction Refer-
ence
BaO2 reaction of barium peroxide with [90]
H2 SO4 in the presence of a small
amount of HCl allows quantitative
conversion; ca 6 % aqueous H2 O2
solutions are obtained
BaO2 as above, but in the presence of HCl [91]
and H3 PO4
Na2 O2 sodium peroxide solutions are passed [92]
through acidic ion exchangers to
exchange Na+ for H+
(CH3 )3 CO–OH tert-butyl hydroperoxide is [93]
hydrolized to tert-butanol and H2 O2
in the presence of H2 SO4 in a
fractionating column
(CH3 )3 CO–OH H2 O2 and isobutene are formed from[94]
tert-butyl hydroperoxide in an
aqueous medium in the presence of
an acidic ion-exchange resin
peracetic acid is hydrolyzed to [95]
CH3 COOH and H2 O2 in the
presence of H2 SO4 ; to shift the
equilibrium to H2 O2 , CH3 COOH is
reacted with an alcohol and the ester
is distilled
peracetic acid is hydrolyzed with [96]
acidic ion exchangers, and the
equilibrium mixture is separated by
distillation
4.4.2. Production from Hydrogen and
Oxygen
Hydrogen peroxide can be formed by direct re-
action of oxygen and hydrogen. However, the
 Hydrogen Peroxide 21

Figure 15. Münchner process

a) Electrolyzer; b) Vacuum cooler; c) Conversion (crystallization); d) Centrifuge; e) Hydrolysis–distillation; f) Centrifuge;
g) Sulfuric acid purification; h) Purification of cycle solution; i) Acid separator; j) Fractionating column

most important hydrogen peroxide manufactur- ing an explosive oxygen/hydrogen gas mixture.
ing process, the AO process, also allows produc- The process was intensively investigated during
tion of hydrogen peroxide from the elements. the last two decades, a demonstration plant with
In 1935 [175] Pietzsch reported that silent a capacity of several kilotons of hydrogen perox-
electrical discharge into a gas containing 95 % ide in methanol is scheduled to be started up in
hydrogen and 5 % oxygen in the presence of the middle of 2006 [179]. For this direct combi-
water vapor at 150–160 ◦ C and 100 kPa, and nation of hydrogen and oxygen various catalytic
subsequent condensation result in a 10 % aque- systems, mostly with precious metals, were pro-
ous solution of hydrogen peroxide. The power posed. The process was especially investigated
requirements are high (40 kWh per kilogram during the last decade, but has reached only pilot
H2 O2 ). Silent electrical discharge was proposed plant status. The ideal direct combination pro-
for producing ultrapure hydrogen peroxide for cess should have:
the semiconductor industry [176].
1) High hydrogen peroxide production rate and
In 1985, DuPont developed and patented a
low decomposition
new liquid-phase hydrogen peroxide process
2) A catalyst with long lifetime and minimal pre-
that directly combines hydrogen and oxygen and
cious metal losses
that allows a 50 % reduction in investment costs
3) Use of a non explosive gas mixture
[177]. The DuPont patent [178] described the
4) Synthesis of hydrogen peroxide at low pres-
production of up to 25 % hydrogen peroxide so-
sure and ambient temperature
lutions with a palladium on carbon catalyst in
5) Noncorrosive reaction medium
acidic aqueous solution containing halide ions
6) High hydrogen selectivity and hydrogen per-
as promoter at 10–25 ◦ C and up to 17 MPa us-
oxide concentrations of at least 5 %
22 Hydrogen Peroxide
Heterogenous catalysts such as palladium or
platinum or mixtures thereof on silica, alumina,
charcoal or hydrophobic supports [180] or fixed-
bed catalysts [181] are preferred. Fixing the cat-
alyst on a halogenated resin eliminates dissolu-
tion of catalyst and corrosion of the equipment
but results in low selectivity and product con-
centration [182, 183] Extremely high hydrogen
selectivities (better than 90 %) using nonexplo-
sive feedstreams can be reached with palladium
on carbon black nanocatalysts [184].
To overcome the concerns over process safety
while using explosive gas mixtures, special reac-
tor designs are proposed: e.g., separately feeding
hydrogen and oxygen into a completely liquid-
filled tubular reactor [185], injecting dispersed
minute bubbles of hydrogen and oxygen into a
rapidly flowing liquid medium [186], or sepa-
rately feeding very small bubbles of hydrogen
and oxygen in the lower part of a mechanically
agitated reactor in such a way that the molar
ratio of hydrogen to oxygen falls within the ex-
plosive range but in the gas phase the ratio of the
reactants lies outside this range [187]. Hydrogen
peroxide concentrations as high as 20 % at hy-
drogen selectivities of 80–83 % can be reached.
Use of methanol instead of water as reaction
medium enables to lift the catalyst productivity
from 207 to 879 gram H2 O2 per gram catalyst
per hour [188]. A vapour phase technique using
a fluidized bed catalyst consisting of a thin layer
(50 nm) of palladium on polytetrafluoroethylene
is described [189, 190].
4.4.3. Production from Carbon Monoxide,
Oxygen, and Water
Hydrogen peroxide can be formed from carbon
monoxide, oxygen and water.
CO+O2 +H2 O→H2 O2 +CO2
The reaction was first described by Traube in
1884. Later palladium [191] or palladium phos-
phine and arsine complexes [192, 193] were
used as catalysts, but the latter lacked oxidation
stability. A palladium bathocuproine complex
in a two-phase acidic solution should overcome
these difficulties [194], but reaction pressure (5–
7 MPa) and excess oxygen (ca. 10–20-fold) were
high. The advantage of the process is that oxida-
tion and reduction takes place in a single reac-
tor. Improved catalyst stability is reported when
using naphthoquinone as mediator [195]. The
process is not used commercially.
4.4.4. Production by Autoxidation of
Organic Compounds
Autoxidation processes for the production of hy-
drogen peroxide include the AO process and
the 2-propanol process discussed in Sections
Anthraquinone Process (AO Process) and 2-
Propanol Process (Shell Process). A variation
of the AO process is the use of H2 S instead of
H2 for the reduction of quinones [196, 197]. The
azobenzene process has occasionally been used
on a commercial scale. Analogous to the AO
process, azobenzene is reduced to hydrazoben-
zene which is then oxidized to azobenzene. The
reaction proceeds at alkaline pH and results in
formation of Na2 O2 · 8 H2 O [198–200].
Other autoxidation processes have been sug-
gested, such as the oxidation of hydrocarbons
and alcohols, but they have no commercial sig-
nificance. They are usually carried out in the gas
phase at high temperature and give a multicom-
ponent reaction mixture due to subsequent sec-
ondary reactions. Hydrogen peroxide can be also
generated by the oxidation of hydrazine or hy-
droxylamine [201].
4.4.5. Production by Cathodic Reduction of
Oxygen
The formation of hydrogen peroxide (max. 1 %
solution) by cathodic reduction of oxygen was
reported by Traube in 1882 [202]. The process
was improved by Fischer and Priess who in-
troduced pressurized electrolysis (ca. 10 MPa),
which gave 1.3–2.7 % hydrogen peroxide solu-
tions with current efficiencies of 83–90 % [203].
The cathodic reduction of oxygen can be de-
scribed by the Equation (1).
O2 +H2 O+2 e− →HOO− +HO− (1)
Hydrogen peroxide is formed according to
Equation (2).
HOO− +H2 O→H2 O2 +HO− (2)
Side reactions are

HOO− +H2 O+2 e− →3 HO−
and
H2 O2 −2 e− →O2 +2 H+
Reaction (3) occurs at high current densities.
To increase the space–time yield at low current
densities, electrodes with high specific surface
areas are used. Catalytic decomposition of hy-
drogen peroxide is avoided by using nonmetal-
lic construction materials and limiting pH and
temperature. The anodic decompostion (Eq. 4)
is limited by using diaphragms or by controlling
the anolyte flow. Oxygen evolved at the anode
can be reused for cathodic reduction.
The first attempts to bring this process to tech-
nological maturity were made by Berl, followed
by Kastening [204], and more recently by Huron
and Dow (H-D Tech) [205–207] and others [208,
209]. An overview is given in [210, 211].
The Huron–Dow process seems to be the only
one used on an industrial scale. It is intended
for small on-site units, especially for the pulp
industry. A unit with ca. 3000 t/a capacity has
been operated since 1992 at Georgia Pacific in
Muskogee/USA. The process uses a trickle-bed
cathode and yields an alkaline solution with ca.
4 % hydrogen peroxide. The H2 O2 /alkali ratio is
1/1.7. The process is operated with an excess of
> 90 vol % oxygen near atmospheric pressure
(ca. 0.11 MPa). A graphite/carbon black/teflon
composite cathode with a lifetime of over one
year is used, and the current yield is 80–95 %.
The solution must have a minimum alkali con-
centration of 4 %; otherwise current yields are
low.
For lowering the basicity of the hydrogen per-
oxide, combinations with dialysis cells are pro-
posed [212]. Newer developments make use of
fuel cell designs [213, 214]. While oxygen is re-
duced at the cathode, hydrogen at a pressure of
4.8–6.2 MPa is oxidized at the anode. No exter-
nal power supply is needed, and alkaline-free hy-
Table 8. Transport regulations for hydrogen peroxide
H2 O2 concentration, wt %
8–20
RID/ADR 5.1, No. 1 c
IMDG Code 5.1
UN No. 2984
packing
group III
Hydrogen Peroxide 23
(3)
drogen peroxide solutions with concentrations
of up to 15 % can be produced.
(4)
5. Storage and Transportation [4]
The decomposition of hydrogen peroxide
caused by catalytic impurities and the associated
release of heat have been described in Chap-
ters Physical Properties and Chemical Proper-
ties. Great care must be taken in production,
storage, and transportation to prevent these im-
purities from entering the hydrogen peroxide so-
lution and to ensure that hydrogen peroxide is
put only into perfectly clean containers. Because
commercial hydrogen peroxide always contains
small quantities of catalytic impurities, the sta-
bilization of peroxide solutions is extremely im-
portant. The stabilizing effects of inorganic and
organic compounds are described in detail in [3].
Sodium pyrophosphate [7722-88-5] and
sodium stannate [12058-66-1] are the preferred
stabilizers and are added separately or together
[215, 216]. Phosphonic- or aminophosphonic
acids may also be used [217, 226].
Aluminum (99.5 %), aluminum–magnesium
alloys, or stainless steels are good construction
materials [218]. Because of their corrosion resis-
tance, polyethylene containers and storage tanks
are preferred for hydrogen peroxide concentra-
tions up to 50 wt %. Before metallic tanks and
containers can be used, their surfaces must be
passivated. Iron particles may become attached
to the surface during the rolling of aluminum,
and they must be removed. Aluminum is there-
fore treated with dilute sodium hydroxide and
then passivated with dilute nitric acid. Because
hydrogen forms during the caustic treatment of
aluminum with sodium hydroxide, safety pre-
cautions must be taken to avoid a detonating gas
atmosphere. It is extremely important to ensure
that no hydrogen peroxide is trapped, e.g. bet-
ween closed valves. If decomposition occurs, ex-
20–60 > 60
5.1, No. 1 b 5.1, No. 1a
5.1 5.1
2014 2015
packing packing
group II group I
24 Hydrogen Peroxide

tremely high pressures result which lead to very

serious explosions.
Hydrogen peroxide is admitted for air trans-
port only up to a concentration of 40 wt % (IATA
regulations).
No special regulations apply to the transport
of solutions of up to 8 wt % hydrogen perox-
ide. Table 8 lists international regulations for rail
(RID), road (ADR), and sea (IMDG code) trans-
port of more highly concentrated solutions.

6. Safety
Hydrogen peroxide has a high energy content:
H2 O2 (l) →H2 O (l) +1/2 O2 (g) ∆H
= −98.30 kJ/mol Figure 16. Dependence of the explosive range of the
vapor phase (> 26 mol % H2 O2 ) on the temperature and
Furthermore, its high oxygen content (47 wt %) the hydrogen peroxide concentration of liquid phase (at
is available for combustion reactions and the ox- atmospheric pressure)
idation of organic compounds. This means that
special safety conditions apply in handling hy-
drogen peroxide [219, 232]. Safety studies can
be divided into three groups:
1) Examination of the pure vapor phase,
2) Examination of the pure liquid (aqueous)
phase, and
3) Examination of mixtures or solutions of or-
ganic compounds with or in aqueous hydro-
gen peroxide.
Results are summarized below, but the lit-
erature must be consulted for details. In spe-
cific cases, especially when hydrogen peroxide
is used with organic compounds, the intended
work area must be checked for hazards if it has
not already been certified as safe by a safety ex-
amination.

Pure Hydrogen Peroxide — Vapor Phase

[5, 220–223]. Explosive vapor mixtures are
formed at atmospheric pressure when the hydro-
gen peroxide concentration in the vapor phase
exceeds 26 mol %. Figure 16 shows the explo-
sive range as a function of temperature and hy-
drogen peroxide concentration.
For the industrial concentration of hydrogen
peroxide solutions, it is important to note that at
reduced pressure the critical hydrogen peroxide Figure 17. Dependence of the explosive range of vapor
concentration of the vapor phase may exceed 26 phase on pressure
mol % (Fig. 17).

Pure Hydrogen Peroxide — Liquid (Aque-
ous) Phase [224–226]. The explosive risk of
pure, highly concentrated hydrogen peroxide so-
lutions has been examined by various methods,
but the sometimes contradictory findings are dif-
ficult to interpret consistently.
Results are influenced primarily by test con-
ditions and not by hydrogen peroxide concentra-
tion. For example, 90.7 wt % hydrogen peroxide
in a tube with an internal diameter of 2.67 cm
(1.05 inches) could not be exploded by the deto-
nation shock of a tetryl booster, whereas explo-
sion occurred in a tube with an internal diam-
eter of 4.09 cm (1.61 inches). Similarly, 86 %
hydrogen peroxide in a tube of 4.09 cm inter-
nal diameter was exploded by raising the tem-
perature above 50 ◦ C [224, 225]. The decisive
factors for explosion of pure aqueous hydrogen
peroxide seem to be the degree of occlusion, the
insulation, and the energy of detonation.
Mixtures or Solutions of Aqueous Hydro-
gen Peroxide with Organic Compounds [219,
227–234]. The hazards associated with mix-
tures of hydrogen peroxide and organic chem-
icals are important for its industrial production
and, especially, for its uses. As an example,
Figure 18 shows results obtained with the hy-
drogen peroxide–acetone–water system [227].
Many other organic compounds give similar re-
sults. The size of the explosive region depends
on the organic compound and the test conditions.
Figure 18. Explosive range (hatched area) of hydrogen
peroxide–acetone–water mixtures
Dashed line, indicates the stoichiometry
8 H2 O2 +CH3 COCH3 →3 CO2 +11 H2 O
Hydrogen Peroxide 25
For practical use, safety tests of the intended
working area are very important. When systems
such as hydrogen peroxide–formic acid–water
are used, other products may be formed (in this
case performic acid), that are more dangerous
than the original mixture [228].
7. Uses
Development of the AO process permitted large-
scale production of hydrogen peroxide, which
is now used widely in almost all industrial ar-
eas. Its main use is in bleaching (→ Bleaching,
Chap. 2.2.3). Uses in the chemical industry and
in environmental protection are increasing be-
cause its great advantage is that the degradation
product is water.
Hydrogen peroxide is used in the textile in-
dustry for bleaching cotton, linen, bast fibers,
wool, silk, polyester fiber, and polyurethane
fiber. In the pulp and paper industry, it is used to
bleach sulfate and sulfite cellulose, wood pulp,
and wastepaper, and to brighten wood veneers
and wooden structures.
The chemical industry employs hydrogen
peroxide for the production of peroxy com-
pounds such as sodium perborate, sodium per-
carbonate, metallic peroxides, or percarboxyl-
ic acids. Hydrogen peroxide is very important
in organic chemistry [229] for epoxidation and
hydroxylation (manufacture of plasticizers and
stabilizers for the plastics industry), oxidation
(manufacture of amine oxides as surfactants for
detergents), oxohalogenation, and initiation of
polymerization. BASF and Dow Chemical have
developed a cost-effective process for produc-
tion of propylene oxide using hydrogen perox-
ide; the first commercial plant with a capacity
of 300 000 t/a of propylene oxide using this new
technology is scheduled to go on stream early
2008 [234].
Sulfuric acid solutions of hydrogen peroxide
are used for the pickling and chemical polish-
ing of copper, brass, and other copper alloys,
as well as for etching and cleaning printed cir-
cuit boards. Highly purified hydrogen peroxide
is used in wet chemical processes during the
manufacture of silicon semiconductor devices
to clean silicon wafers and to remove photore-
sist layers. It is also used for in situ leaching in
underground uranium mining.
26 Hydrogen Peroxide
Hydrogen peroxide is increasingly used in
environmental protection to detoxify effluents
containing formaldehyde, phenols, or cyanide
(e.g., wastewater from mines and tempering
works, galvanizer concentrate, photochemical
effluents), and to deodorize sulfur-containing ef-
fluents. Smoke and exhaust gases containing sul-
fur dioxide and even dioxins can be completely
detoxified with hydrogen peroxide.
Hydrogen peroxide is a highly efficient dis-
infectant (especially for packaging materials); it
is also used as a bleaching agent in hair prepa-
rations and as a propellant [226] in space tech-
nology.
8. Toxicology and Occupational
Health
Hydrogen peroxide plays a major role in physio-
logical oxidation processes in living organisms.
Hydrogen peroxide is produced during normal
aerobic cell metabolism, which involves a num-
ber of enzymes, in particular superoxide dismu-
tase. Humans may exhale up to 1 mg/m3 of hy-
drogen peroxide resulting from endogenous for-
mation [236].
Reactive oxygen species, including hydrogen
peroxide, may also lead to oxidative damage to
cells. Therefore a number of defense systems
can be found in biological systems which lead
to detoxification of reactive oxygen species. The
most important detoxification reactions for hy-
drogen peroxide are its conversion to oxygen
and water by enzymes such as catalase, perox-
idase, and selenium-dependent glutathione per-
oxidase. The activity of these enzymes varies
between tissues and between different animal
species or strains [236]. Additionally, a variety
of nonenzymatic antioxidants, such as Vitamin
E and A, ubiquinols, ascorbic acid, glutathione,
complete an efficient intra- and extracellular net-
work of defenses [237].
There is limited information concerning the
absorption of exogenously applied hydrogen
peroxide. The permeability of biological mem-
branes to hydrogen peroxide is comparable to
that of water, but less than that of oxygen [236].
Although significant amounts of topically ap-
plied hydrogen peroxide can penetrate the epi-
dermis or mucous membranes, the substance un-
dergoes rapid spontaneous or enzyme-catalyzed
decomposition to water and oxygen in the un-
derlying tissue. This may lead to the formation
of small gas bubbles and reversible blanching
of the exposed tissue. Formation of larger vol-
umes of gaseous oxygen can lead to detachment
of cell layers and the rupture of tissues and or-
gans. Locally formed oxygen is removed by the
blood. However, the increase in oxygen content
in the blood results in a hyperbaric response.
There is no evidence that hydrogen peroxide it-
self is absorbed in amounts sufficient to produce
systemic effects [236]. In in vitro studies with
rat blood plasma additions of up to 1 mM of hy-
drogen peroxide were degraded to nondectable
concentrations within less than 10 min [238].
This experimental result supports the assump-
tion that hydrogen peroxide is not systemically
available.
The major adverse effects of hydrogen per-
oxide are related to its irritating properties.
The acute oral toxicity of hydrogen perox-
ide in animals depends on the strength of its so-
lution. In the rat LD50 values ranged from >
5000 mg/kg for a 10 % solution to 1200 mg/kg
for 35 % solution. In humans, toxic effects ob-
served after accidental ingestion of concentrated
solutions are generally related to the irritant or
corrosive action and the rapid generation of large
volumes of oxygen in the gastrointestinal tract.
Single cases of deaths due to gas embolism fol-
lowing accidental swallowing of hydrogen per-
oxide solutions have been reported [239], how-
ever, in other case reports complete recovery was
observed within 2-3 weeks [236].
Complete recovery has occurred within two
to three weeks even in near fatal cases [236].
Effects of exposure to hydrogen peroxide by
inhalation are limited to irritation of the respi-
ratory tract and the eyes. Exposure to hydrogen
peroxide vapor (338–427 mg/m3 ) induced signs
of irritation in rodents. In humans exposed for
4 h the irritation threshold for the respiratory
tract was 10 mg/m3 and for the skin 20 mg/m3 .
At these concentrations, eye and throat irritation
were reported [236].
Slight nasal irritation was reported in fac-
tory workers for short-term exposure to ca. 3.5
mg/m3 [236]. Gradual bleaching of hair was re-
ported for exposure levels of 0.7–1.4 mg/m3 , but
it was uncertain whether bleaching was due to
vapor exposure or to transfer of hydrogen per-
oxide from the hands (gloves) to the hair [236].

The dermal toxicity of hydrogen peroxide so-
lutions of 35–75 % in rabbits was low (LD50 ,
dermal, rabbit: > 2000 mg/kg). LD50 values of
a 90 % solution ranged from ca. 690 mg/kg in
rabbits to > 2000 mg/kg in rats, pigs, and cats
[236].
The irritating properties of hydrogen perox-
ide are dependent on the concentration of the
solution. Hydrogen peroxide solutions of 10 %
or less were nonirritating, and solutions of 35 %
slightly irritating to rabbit skin. Higher concen-
trations (≥ 50 %) are corrosive to rabbit skin.
Hydrogen peroxide (50 %) was corrosive to rab-
bit skin after 1 h of exposure, whereas 70 % hy-
drogen peroxide showed corrosive effects after
only 3 min of exposure [236]. Dermal exposure
of humans to dilute solutions has been reported
to result in transient symptoms without perma-
nent effects, such as paresthesia, blanching, and
blistering [236].
With regard to eye irritation solutions of 5 %
or less are not irritant to rabbit eyes; a concen-
tration of 6 % caused slight reversible eye irrita-
tion in rabbits. Concentrations of 10 % or higher
caused irreversible corneal damage. In humans,
1–3 % solutions have been used for eye treat-
ment without causing injury. However, solutions
containing more than 200 mg/L have been re-
ported to cause hyperemia, and concentrations
of greater 100 mg/L transient pain and stinging
[236]. Hydrogen peroxide (6 %) was not a skin
sensitizer in guinea pigs [236]. Human experi-
ence from dermatological hospital departments
confirms that hydrogen peroxide is no skin sen-
sitizer [239].
After repeated inhalation exposure (6 h/d, 2
to 3 d per week, 3 weeks), mortality in mice was
observed at 80 mg/m3 hydrogen peroxide. Af-
ter prolonged exposure, in rats and dogs (5 or 6
h/d, 5 d per week for 16 or 26 weeks, respec-
tively), respiratory irritation and transient skin
thickening were observed at 10 mg/m3 hydro-
gen peroxide [236]. In a 28-day inhalation study
in rats the NOAEL for respiratory tract irritation
was 2.9 mg/m3 . Clinical signs of irritation were
observed from 14.6 mg/m3 [240].
In humans, pulmonary function monitoring
showed no evidence of adverse effects from oc-
cupational exposure for several years at levels
ranging from not detectable to 0.79 mg/m3 (8-h
TWA) [236]. Further surveys of workers occu-
pationally exposed to hydrogen peroxide in dif-
Hydrogen Peroxide 27
ferent production plants for up to 40 years with
recent exposures below 1.4 mg/m3 (8-h TWA),
but occasional peak exposures up to 5 mg/m3
confirmed the absence of effects on lung func-
tion or any other clinical effects apart from sin-
gle observations of skin or throat irritation, skin
or hair whitening after acute incidents of high
short-term exposures [239]. In a population ex-
posed to levels of 2 - 3 mg/m3 (8-h TWA) and
frequent peak exposure levels up to 11 mg/m3
symptoms including eye and respiratory irrita-
tion, headaches and temporary loss of olfaction
as well as skin and hair blanching and three cases
of recurring bronchitis–sinusitis were reported.
All persons recovered completely after the ex-
posure was reduced to below the occupational
exposure limit of 1.4 mg/m3 and the repeated
peak exposures were avoided. The authors as-
sume that the recurrent peak exposure levels of
up to 11 mg/m3 may have contributed to the ob-
served symptoms [241].
More detailed studies of the effects of hydro-
gen peroxide after prolonged exposure were per-
formed by administration in drinking water. Af-
ter prolonged exposure (146 d), decreased body-
weight gain was observed in rats at concentra-
tions of 0.25 % and above [242] (other studies
observed first effects at higher levels [236]).
In a 90-d drinking water study in catalase-
deficient mice receiving 100–3000 ppm of hy-
drogen peroxide, irritation of the mucous mem-
branes of the gastrointestinal tract was observed.
The no observed adverse effect level (NOAEL)
was 100 ppm, corresponding to 26 and 37 mg
kg−1 d−1 for male and female mice, respec-
tively. After cessation of treatment these effects
were fully reversible. No systemic effects were
observed [243]. In studies on the same mouse
strain with a 6–7 month exposure period, signs
of gastrointestinal irritation were observed at ap-
proximately the same dose level as in the 90-d
study. Therefore it can be assumed that the effect
is not cumulative and the NOAEL does not de-
crease with increasing exposure time. A NOAEL
of 56.2 mg kg−1 d−1 was reported in a 12-week
gavage study in rats [236].
Results of mutagenicity studies indicate that
hydrogen peroxide induces gene mutations in
vitro without metabolic activation in bacteria,
yeast, and mammalian cells. Hydrogen perox-
ide also induced chromosome aberrations and
micronuclei in mammalian cell cultures without
28 Hydrogen Peroxide
a metabolic activation system. However, addi-
tion of exogenous metabolic activators or cata-
lase lowered or abolished the genotoxic response
[236]. In a mechanistic study, it has be shown
that hydrogen peroxide induced controlled cell
death (apoptosis) rather than genotoxic lesions
in cell cultures of human lymphocytes and did
not induce 8-hydoxyguanosin formation [244].
In in vivo tests, no chromosomal aberrations
or micronuclei were observed in the bone mar-
row of rats and mice after repeated oral or sin-
gle intraperitoneal administration [236]. An in
vivo/ex vivo study for unscheduled DNA syn-
thesis (UDS) after intravenous administration of
the maximum tolerated dose to rats did not show
an increase in UDS compared to the solvent con-
trol group [245]. In a study of local genotoxic-
ity to mouse skin, 70 % hydrogen peroxide did
not cause any mutations, hydroxylation of DNA
bases, or gave an indication of cellular prolifer-
ation, thus suggesting that the substance has no
local genotoxic potential in vivo [239]. In con-
clusion, the available studies indicate that hy-
drogen peroxide has no genotoxic or mutagenic
potential under in vivo conditions.
The carcinogenic effects of hydrogen perox-
ide were studied in rats and mice. Hydrogen per-
oxide induced an increase of duodenal tumors
in mice chronically exposed to 0.1–0.4 % of hy-
drogen peroxide in drinking water. The tumor
incidence correlated with a specific inflamma-
tory response in this tissue and was more pro-
nounced in those mice strains which had a low
catalase activity [236]. The absence of preneo-
plastic changes and the reversibility of the ef-
fects in a 90-d drinking water study in mice of
the same strain [243] (with 0.01–0.3 % hydro-
gen peroxide in drinking water), suggest that the
tumors observed in the two-year study seem to
be a secondary response to a prolonged inflam-
matory process in this specific strain of mice.
Rats exposed to near lethal concentrations in
their drinking water developed forestomach pa-
pilloma but no tumors in the glandular stomach
and duodenum [236]. Chronic hydrogen perox-
ide exposure at concentrations which induce sig-
nificant inflammation of the exposed tissue has
been shown to give a weak tumor-promoting ef-
fect. Repeated topical application of 15 % hy-
drogen peroxide to Sencar mice did not induce
skin tumors; this demonstrates the absence of
initiating properties [236].
On the basis of these studies ACGIH has
grouped hydrogen peroxide into category A3 for
carcinogenicity: “The agent is carcinogenic in
experimental animals at relatively high doses,
by routes of administration, at sites of histologi-
cal types, or by mechanisms that are not consid-
ered relevant to worker exposure. Available epi-
demiological studies do not confirm an increased
risk of cancer in exposed humans. Available ev-
idence suggests that the agent is not likely to
cause cancer in humans except under uncom-
mon or unlikely routes of exposure” [246]. Sim-
ilar conclusions were drawn in the EU risk as-
sessment on hydrogen peroxide [239].
Occupational exposure limits have been es-
tablished for hydrogen peroxide in different
countries, for example by ACGIH [246], the
Dutch Social Economic Council [247], the UK
HSE [251], Belgium [248], the French Ministry
of Labour [249], and the German AGS (Auss-
chuss für Gefahrstoffe) [250]: 8-h TWA: 1.4
mg/m3 or 1.5 mg/m3 . A 15-min STEL of 2.8
mg/m3 was established by HSE (max. 1 h in a
24-h period) [251], and a 15-min STEL of 1.4
mg/m3 by AGS [250].
9. Environmental Effects
Hydrogen peroxide occurs naturally. Its forma-
tion, use, and degradation are strongly related to
the oxygen-containing atmosphere of our planet
and adaptation of nature to oxygen. The concen-
tration of hydrogen peroxide in environmental
compartments results from a dynamic equilib-
rium between its natural production and degra-
dation in photochemical, chemical, and biolog-
ical processes.
Because of its chemical reactivity, hydrogen
peroxide also has toxic properties that are some-
times used as a natural defense mechanism. The
toxicity of hydrogen peroxide to organisms and
cells is dependent on the balance of intoxication
and detoxification. However, detoxification re-
actions can in turn lead to toxic effects due to
the release of oxygen gas [252].
9.1. Atmospheric Fate
The atmospheric fate of hydrogen peroxide is
governed by generation and degradation reac-
tions. The formation of atmospheric hydrogen

peroxide is photochemically mediated. Con-
sequently, natural concentrations are latitude-
dependent. High solar irradiation leads to high
hydrogen peroxide concentrations. Hydrogen
peroxide in atmosphere is produced through rad-
ical reactions. Light, oxygen, hydrocarbons, and
free radicals in the atmosphere mediate hydro-
gen peroxide formation.
In an unpolluted atmosphere, the formation
of H2 O2 is ultimately due to the photolysis of
ozone (O3 ) at wavelengths < 310 nm to give
oxygen atoms. The latter are converted to free
radical species, including • OH, RO• , HO•2 , and
RO•2 , which initiate the breakdown of organic
species and give rise to HO•2 radicals which are
precursors of H2 O2 [253]. The superoxide rad-
ical O2−• is the direct precursor of H2 O2 :
2 O2−• +2 H+ →H2 O2 +O2
The rate of H2 O2 formation from O2−• is
of the second order and is also temperature-
and humidity-dependent. Levels of H2 O2 were
raised by 70 % when air temperature increased
from 10 to 30 ◦ C, other factors remaining con-
stant [254].
Water vapor is involved in another process of
H2 O2 formation, namely, the production of hy-
drated hydroperoxyl radicals, which rapidly re-
act with HO•2 radicals to generate H2 O2 . About
twice as much H2 O2 is formed at 100 % rela-
tive humidity than at 50 % humidity at ambient
temperature (25 ◦ C) [255].
Degradation in the atmosphere is due to direct
and indirect photolysis reactions and to chemical
reaction with organic substances. Degradation
by indirect photolysis is expected to be the main
degradation mechanism in air. Indirect photoly-
sis is due to sensitization by secondary reactions
with OH and O2 radicals and organic substances.
The most important degradation reaction is re-
action with the hydroxyl radical:
H2 O2 +• OH→H2 O+HO2 •
Direct hydrogen peroxide photolysis is mediated
by radiation at wavelengths of 280–380 nm:
2 H2 O+hv→H2 O+• OH+HO2 •
•
OH+HO2 • →H2 O+O2
The rate for the direct photolysis of hydrogen
peroxide for moderate light intensities is directly
Hydrogen Peroxide 29
proportional to the hydrogen peroxide concen-
tration and the square root of the light intensity
[256]. Kleinman estimated the direct photolysis
half-life to be 2.14 days [257].
The formation of organic hydroperoxides
(R−OOH) appears to be another pathway for
the decomposition of H2 O2 in the atmosphere.
They are formed by hydroxy radical induced ox-
idation of hydrocarbons or direct oxidation of
aldehydes by hydrogen peroxide:
R−CHO+H2 O2 →R−OOH
However, the measured levels of R−OOH have
been reported to be only 1 to 10 % of the ambient
air concentrations of hydrogen peroxide [258].
The overall tropospheric half-life of hydro-
gen peroxide was reported to be 10 to 20 h
[257]. Atmospheric levels of hydrogen peroxide
ranged from 0.02 ppb to 4 ppb in unpolluted air
and up to 180 ppb in smog situations. Concen-
trations in rainwater ranged from 0 to approxi-
mately 2000 µg/L, cloud water contained up to
8000 µg/L [252].
9.2. Aquatic Fate
Hydrogen peroxide levels in natural waters
range from 0.14 to 58 µg/L in seawater and from
0.34 to 109 µg/L in freshwater [252]. The con-
centrations are dependent on the daytime and
sunlight intensity and are decreasing with in-
creasing depth. In surface water a photochemi-
cal process involving dissolved, light-absorbing
organic matter (e.g., humic acids) and molec-
ular oxygen produces hydrogen peroxide. The
wavelengths needed are between 290 and 385
nm [256, 259, 260].
Organic matter +hν→ Organic radical•
Organic radical• +O2 →Organic radical+• +O−
2
2 O− • +
2 +2 H →H2 O2 +O2
In seawater, the Weiss mechanism also con-
tributes to the formation of hydrogen peroxide
[261].
Fe2+ +O2 →Fe3+ +O−
2
•
Fe2+ +O− • +
2 +2 H →Fe
3+
+H2 O2
Hydrogen peroxide concentrations in surface
water can also originate from precipitation of
rain [259].
30 Hydrogen Peroxide
9.3. Biodegradation
The main source of degradation of hydrogen per-
oxide in natural waters is the degradation by bi-
ological material, e.g., plankton. The half-life in
water decreases with increasing size of the mi-
crobial population in water [252].
Consequently, the half-life in natural waters
was reported to range from a few hours to sev-
eral days [262]. Degradation is also catalyzed by
transition metals. This may in particular play a
role for the more rapid degradation in seawater.
Direct photolysis does not play a major role in
aquatic decomposition.
In municipal wastewater, hydrogen peroxide
degrades rapidly because of the content of mi-
croorganisms and organic material. Half-lives
were reported to be below 15 min [263]. Half-
life was reported to be below 2 min in an acti-
vated sludge respiration test according to OECD
guideline 209 [252].
9.4. Environmental Releases
Environmental releases of hydrogen peroxide,
in particular to surface water, can occur during
production and use. Where there is no waste-
water treatment, higher concentrations may be
released that may lead to toxicity to aquatic or-
ganisms. A review of the environmental releases
from different production and use scenarios is
given in the EU risk assessment of hydrogen per-
oxide [239].
9.5. Environmental Effects
Hydrogen peroxide is moderately toxic to
aquatic organisms. For hydrogen peroxide, acute
ecotoxicity data are available for three aquatic
tropic levels (fish, daphnia, and algae). The tox-
icity of hydrogen peroxide to salt water species
was of the same order of magnitude as that to
freshwater species. The 96-h LC50 values for
freshwater fish range from 16.4 to 37.4 mg/L
H2 O2 [252]. The 48-h EC50 value for Daph-
nia pulex was 2.4 mg/L and the 24-h EC50 for
Daphnia magna was 2.3 mg/L [239].
For algae, the initial algae concentration in-
fluences the outcome of the study. The most ap-
propriate studies are those with the OECD stan-
dard of 10 000 cells/mL at the moment of addi-
tion of hydrogen peroxide. Under these condi-
tions, the 72-h EC50 (biomass) in the green al-
gae Chlorella vulgaris was 2.5 mg/L, based on
growth rate the 72-h EC50 was 4.3 mg/L [265].
Other algae species were less sensitive, for ex-
ample the 96-h IC50 for Scenedesmus quadri-
cauda was 27.5 mg/L [266].
For the bacterium Pseudomonas putida, a 16-
h EC10 of 11 mg/l was determined [252]. At
concentrations from 30 mg/mL (no observed ef-
fect concentration in an activated sludge respi-
ration inhibition test), some toxicity to degrad-
ing microorganisms has been observed, the EC50
was 466 mg/L in this study [264]. However, at
concentrations up to 200 mg/L, H2 O2 does not
affect the performance of biological treatment
works (activated sludge process). It has even
been shown to stimulate the biodegradation in
wastewater treatment works [252].
Chronic toxicity studies on standard fish and
daphnia species are not available, but studies
on different life stages of fish [267] and pro-
longed (56-d) exposure of zebra mussels [268]
do not indicate effects of higher sensitivity than
in acute studies. The lowest long-term aquatic
no-effect concentration (NOEC) was 0.1 mg/L
in Chlorella vulgaris [265]. In a field study, the
influence of hydrogen exposure on phytoplank-
ton and bacterioplankton in a small humic lake
in Canada has been examined. At H2 O2 con-
centrations above 3.4 mg/L, both phytoplankton
and bacterioplankton were inhibited. Bacteria
were inhibited at concentrations between 0.12
and 3.4 µg/L. Photosynthetic activity of phyto-
plankton was increased at concentrations bet-
ween 3.4 and 34 µg/L. Interestingly, the toxicity
to phytoplankton was dependent on the time of
the day. During the period of lowest background
H2 O2 concentrations at sunrise, no effect was
observed even when concentrations as high as
170 µg/L were added. At sunset, a concentra-
tion of 1.7 µg/L resulted in a decrease of primary
production by algae suggesting that the detox-
ification capacity of the cells might have been
exceeded because of the higher natural concen-
trations [269].
The effect of hydrogen peroxide to plants has
been tested on growth, seed germination, and
early seed growth of rice, corn, soybean, pig-
weed, tomatoes, and barnyard grass. Concentra-
tions up to 218 mg/L in irrigation water have no

significant effect on any of the above parameters
[252].
When assessing the risk of hydrogen perox-
ide to environmental organisms, the background
concentrations in the environment and the ca-
pacity to adapt to increased hydrogen perox-
ide concentrations have to be considered. The
medium used in laboratory tests is quite different
from the natural conditions. Depending on the
environmental conditions, many materials (e.g.,
dissolved transition metals, solid metals, solid
metal oxides, hydroxides, activated carbon, en-
zymes) may contribute to rapidly reduced hy-
drogen peroxide levels. It is thus assumed that
under laboratory conditions the test organisms
are more exposed and thus more sensitive than
under natural conditions to the prolonged expo-
sure of H2 O2 . In field conditions, the more sen-
sitive species may also be protected from a direct
effect of hydrogen peroxide by the presence of
the total biomass and the suspended solids.
However, it is obvious that accidental re-
leases, in particular of highly concentrated hy-
drogen peroxide, can immediately be lethal to
aquatic organisms. With regard to long-term ef-
fects of lower levels, it can be expected that adap-
tation of ecosystems may well be possible.
10. References
General Reference
1. Gmelin, System no. 3, Sauerstoff.
2. Kirk-Othmer, 5th ed., 14, Wiley, 2005.
3. W. Machu: Das Wasserstoffperoxid und die
Perverbindungen, 2nd ed., Springer, Wien
1951.
4. Wasserstoffperoxid, Berufsgenossenschaft der
chemischen Industrie, Merkblatt M 009, BGI
792, Jedermann-Verlag, Heidelberg 2000.
5. W. C. Schumb, C. N. Satterfield, R. L.
Wentworth: Hydrogen Peroxide, Reinhold
Publ. Co., New York 1955.
6. Ullmann, 5th ed., A13, 443–465.
7. W. M. Weigert: Wasserstoffperoxid und seine
Derivate, Hüthig, Heidelberg 1978.
8. Winnacker-Küchler, 5th ed., 3, 682–760,
Wiley-VCH, Weinheim, Germany 2005.
Specific Reference
9. L. J. Thenard, Ann. Chim. Phys. 8 (1818) 306.
10. H. Meidinger, Ann. Chem. Pharm. 88 (1853)
57.
Hydrogen Peroxide 31
11. H. Berthelot, C. R. Hebd. Seances Acad. Sci.
86 (1878) 71.
12. K. Elbs, O. Schönherr, Z. Elektrochem. 1
(1895) 417, 468.
13. W. M. Weigert, H. Delle, G. Käbisch, Chem.
Ztg. 99 (1975) 101.
14. W. Manchot, Liebigs Ann. Chem. 314 (1901)
177;
316 (1901) 318, 331.
15. J. H. Walton, G. W. Filson, J. Am. Chem. Soc.
54 (1932) 3228.
16. O. v. Schickh, Chem. Ing. Tech. 32 (1960) 462.
17. J. Lacson, S. Schlag and G. Toki: “CEH
Product Review Hydrogen Peroxide” in
Chemical Economics Handbook, SRI
International, Menlo Park 2002, 741. 5000A.
18. W. R. Busing, H. A. Levy, J. Chem. Phys. 42
(1965) 3054.
19. P. A. Giguere, J. Chem. Ed. 60 (1983) 399.
20. C. L. Yaws, H. S. N. Setty, Chem. Eng. (1974)
67.
21. G. Scatchard, G. M. Karanagh, L. B. Ticknor,
J. Am. Chem. Soc. 74 (1952) 3715.
22. W. M. Latimar: The Oxidation States of the
Elements and their Potentials in Aqueous
Solutions, 2nd ed., Prentice Hall, New York
1952.
23. Hydrogen Peroxide for Industrial Use –
Determination of Content, International
Organization for Standardization, ISO/DIS
7157.
24. Degussa, DE 1 039 500, 1956 (G. Käbisch, E.
Richter, A. Gross).
25. I.G. Farbenindustrie AG, GB 508 081, 1939;
US 2 215 883, 1939.
26. Degussa, DE-AS 1 945 750, 1969 (G.
Käbisch, R. Trübe).
27. EDOGAWA, DE 1 112 051, 1957 (K.
Hiratsuka, S. Matsumura, M. Matsuda, K.
Takagi, T. Abe).
28. Laporte Chemicals Ltd., GB 747190, 1953 (C.
W. Lefeuvre).
29. Akzo Nobel, WO 2001098204, 2000 (M.
Nystroem, C. Jaernvik).
30. Degussa, DE 1 261 838, 1963 (G. Käbisch).
31. RO 108956, 1991 (M. Mihalcescu, C. Andrei,
C. Dragoi).
32. E. I. du Pont de Nemours & Co., DE 888 840,
1950 (C. R. Harris, J. W. Sprauer).
33. BECCO, US 2 537 655, 1950 (L. H. Dawsey,
C. K. Muehlhauser, R. R. Umhoefer).
34. Laporte Chemicals Ltd., DE 933 088, 1953 (C.
W. Lefeuvre).
35. E. I. du Pont de Nemours & Co., US 4046868,
1973 (L. G. Vaughan).
36. Oxysnthese, EP 287 421, 1988 (C. Pralus).
32 Hydrogen Peroxide
37. Kemira, US 5 114 701, 1990 (R. Aksela, E.
Suokas).
38. EKA, EP 284 580, 1988 (U. M. Andersson, E.
A. Bengtsson).
39. Degussa, DE 3 027 253, 1980 (G. Goor, W.
Kunkel).
40. FMC Corp., US 4 394 369, 1983 (R. Wayne).
41. EKA, EP 286 610, 1988 (U. M. Andersson, E.
A. Bengtsson).
42. Degussa, DE 2 018 686, 1970 (G.
Giesselmann, G. Schreyer, W. Weigert).
43. J. Fang, Huagong Keji 6 (1998) 18–23.
44. I. G. Farbenindustrie AG, US 2 215 883, 1937
(H.-J. Riedl, G. Pfleiderer).
45. I.G. Farbenindustrie AG, US 2 158 525, 1936
(H.-J. Riedl, G. Pfleiderer).
46. E. I. du Pont de Nemours & Co., US
2 689 169, 1949 (W. S. Hinegardner).
47. BASF, EP 348464, 1990 (M. Eggersdorfer, J.
Henkelmann, W. Grosch).
48. Kawasaki Kasei, JP 58180452, 1983 (S.
Numata).
49. Montecatini, IT 601 435, 1960 (C. Ferri, A.
Niglierina, B. Ponzini).
BASF, DE 43 39 649, 1993 (H. Hiller).
50. Degussa, DE 1 063 581, 1956 (L. R. Darbee,
D. F. Kreuz).
51. Kvaerner, EP 958 238, 1997 (E. Bengtsson, A.
Casen, A. Unum).
52. Degussa, US 6 153 169, 1998 (J. Glenneberg,
G. Goor, E. Staab, H. Angert).
53. Akzo Nobel, EP 1101733, 1999 (M.
Nystroem, C. Jaernvik, H. Thor, S. Saari).
54. EDOGAWA, DE-AS 1 085 861, 1957 (K.
Hiratsuka et al.).
55. E. Santacesaria, R. Ferro, S. Ricci, S. Carra,
Ind. Eng. Chem. Res. 26 (1987) 155;
E. Santacesaria, M. Di Serio, A. Russo, U.
Leone, R. Velotti, Chem. Eng. Sci. 54 (1999)
2799–2806.
56. P. Oinas, G. Wild, N. Midoux, H. Haario,
Chem. Eng. Proc. 34 (1995) 503.
57. A. Golja, M. Krajnc, S. Flajs, M. Grcar:
Zbornik Referatov s Posvetovanja Slovenski
Kemijski Dnevi, Maribor, Slovenia, Sept.
20-21, 2001, Issue Part 2, 723–729.
58. L. Kurc, V. Vopravil, L. Cerveny, Res. Chem.
Intermed. 27 (2001) 249–258.
Degussa, DE 10114 982, 2001 (T. Haas, J.
Glenneberg, R. Wagner, R. Vanheertum).
59. Laporte Chemicals Ltd., GB 741 444, 1953
(W. R. Holmes, C. W. Lefeuvre, R. Lait, W. S.
Wood).
60. Solvay & Cie., DE 1 041 928, 1955 (M.
Gonze, E. Leblon).
61. Degussa, DE 1 051 257, 1957 (G. Käbisch, E.
Richter).
62. E. I. du Pont de Nemours & Co., US
2 673 140, 1951 (J. W. Sprauer).
63. EDOGAWA, FR 1 319 025, 1962.
64. EDOGAWA, DE-OS 2 141 234, 1971 (H. Shin
et al.).
65. J. Petr, L. Kurc, Z. Belohlav, L. Cerveny,
Chem. Eng. Proc. 43 (2004) 887–894.
66. E. I. du Pont de Nemours & Co., US
4 668 499, 1986 (J. D. Rushmere).
67. Solvay Interox, EP 546 616, 1992 (V.
Deremince, C. Vogels).
68. T. Berglin, N. H. Schöön, Ind. Eng. Chem.
Process Des. Rev. 22 (1983) 150–153.
69. EKA, DE 2 419 534, 1973 (B. G. Franzen).
70. Degussa, EP 1274651, 2000 (T. Haas, J.
Glenneberg, R. Wagner, R. Vanheertum).
71. Degussa, DE 1 273 499, 1964 (G. Schreyer, O.
Weiberg, G. Giesselmann).
72. Degussa, WO 2005 014471, 2004 (G. Goor, E.
Staab, J. Glenneberg).
73. F. Sandelin, P. Oinas, T. Salmi, H. Haario,
Stud. Surf. Sci. Catal. 88 (1994) 635.
74. A. Drelinkiewicz, J. Mol. Catal. A 101 (1995)
61;
Bull. Pol. Acad. Sci. Chem. 43 (1995) 51.;
A. Drelinkiewicz, R. Laitinen, R. Kangas, J.
Pursiainen, Appl. Catal. A: General 284
(2005), 59–67.
75. Solvay Interox, WO 98 15 350, 1997 (V.
Mathieu, P. Pennetreau, N. Vanlautem).
76. BASF, DE 801 840, 1951 (G. Pfleiderer, H.-J.
Riedl, W. Deuschel).
77. FMC Corp., US 2 720 532, 1952 (R. R.
Umhoefer).
78. FMC Corp., US 2 730 533, 1952 (R. R.
Umhoefer).
79. Kali-Chemie, DE-AS 1 137 719, 1960 (V.
Hauschild, M. Stahl).
80. J. Fang et al., J. Catal. 229 (2005) 97–204.
81. Degussa, DE 1 542 089, 1965 (G. Käbisch, H.
Herzog).
Degussa, US 4 428 923, 1982 (W. Kunkel, J.
Kemnade, D. Schneider).
82. Degussa, WO 2000 015550, 1999 (H.
Eickhoff, R. Schütte).
83. Degussa, EP 1 334 062, 2001 (B. Maurer, P.
Korl).
84. Degussa, GB 1 096 260, 1965.
85. Degussa, EP 687 649, 1995 (H. Birkenbeil, U.
Brand, G. Goor, W. Kunkel).
86. Degussa, EP-A 571 762, 1993 (H. Eickhoff, R.
Goedecke, H. Huber, K. Mangartz, R.
Schirrmacher).

87. Laporte Chemicals Ltd., GB 718 306, 1953 (C.
A. Morgan).
88. Du Pont, US 5 772 977, 1995 (C.L. Jenkins, F.
B. Kirby, T. A. Koch).
89. Mitsubishi, DE 19713376, 1997 (K.
Ogasawara, T. Kato, N. Okuda, T. Konishi, K.
Kato).
90. Laporte Chemicals Ltd., GB 746 385, 1953 (C.
W. Lefeuvre, C. A. Morgan);
GB 922 022, 1959 (R. Lait, D.R.
Lloyd-Owen).
91. Laporte Chemicals Ltd., GB 718 307, 1952 (J.
A. Williams, C. W. Lefeuvre).
92. Du Pont, US 4 374 820, 1981 (T. E. Guenter).
93. Solvay, WO 99 040024, 1998 (D. Vandenberg,
J. P. Ganhy, N. Vanlautem).
94. FMC Corp., US 3 009 782, 1958 (D. H.
Porter).
95. EKA, US 5 063 043, 1990 (E. A. Bengtsson).
EKA, US 4 552 748, 1983 (T. Berglin, W.
Herrmann).
96. A. Drelinkiewicz, A. Pukkinen, R. Kangas, R.
Laitinen, Catal. Lett. 94 (2004) 157–170.
97. A. Drelinkiewicz, R. Laitinen, R. Kangas, J.
Pursiainen, Appl. Catal. A: General 284
(2005) 59–67.
98. FMC Corp., US 4 428 922, 1982 (Q. G.
Hopkins).
99. Kemira, US 5 071 634, 1990 (T. Maunula,
E.-L. Mustonen, I. Turunen, P. Virta);
Kemira, WO 93 02 960, 1992 (E. Maekiniemi,
T. Maunula, A. Pukkinen, I. Turunen).
100. Q. Wang, L. Wang, Y. Wang, F. Hei, Z. Li, Z.
Mi, React. Kinet. Catal. Lett. 81 (2004)
297–304.
101. EKA, EP 778 085, 1996 (M. Nystroem, M.
Silverstroem).
102. Degussa, DE 2 003 268, 1970 (M. Liebert, H.
Delle, G. Käbisch).
103. Chem. Process Eng. (London) 40 (1959) no. 1,
5.
104. Br. Chem. Eng. 4 (1959) 88.
105. Österr. Chemische Werke, WO 86 06710,
1986 (J. Kemnade, B. Maurer).
106. Kemira, DE-OS 4 029 784, 1990 (I. Turunen,
E. L. Mustonen).
107. Allied Chemical Corp., US 2 902 347, 1954 (J.
N. Cosby, R. L. Harris, E. H. Rissmiller).
108. Degussa, WO 2000 017098, 1999 (H.
Eickhoff, R. Schütte).
109. S. Lue, Z. Mi, Y. Wang, L. Wang, Chem. Eng.
Technol. 27 (2004) 903–908.
110. Pittsburgh Plate Glass Co., DE 977 130, 1965
(J. R. Cox, T. E. Brown).
Hydrogen Peroxide 33
111. EKA, EP 078 780, 1982 (B. G. Franzen, W.
Herrmann).
112. Kemira, EP-A 726 227, 1996 (I. Turunen).
113. Research Disclosure 40735, March 1998
(Derwent No. 98-205225).
114. Pittsburgh Plate Glass Co., US 3 055 838,
1957 (J. W. Moore).
115. E. I. du Pont de Nemours & Co., US 2 739 875,
1951 (J. W. Sprauer, T. V. Williams).
116. Foret, US 4 566 998, 1981 (V. M. Ochoa, R.
Pueyo).
117. Mitsubishi Gas, DE-OS 19 715 034, 1997 (K.
Ogasawara, T. Kato, K. Minato, K. Kato, M.
Sumitani).
118. FMC, US 6 180 804, 1998 (E. F. Dougherty).
119. Laporte Chemicals Ltd., DE-AS 1 138 743,
1959 (B. K. Howe, S. Andrus, G. N. Davis).
120. Laporte Chemicals Ltd., GB 991 413, 1961 (B.
K. Howe).
121. FMC Corp., US 4 824 609, 1988 (D. S. Sethi).
122. FMC Corp., US 4 668 436, 1985 (D. S. Sethi).
123. Du Pont Australia, WO 95 04 702, 1994 (P. J.
Collins, P. B. Humphries, G. J. J. Prendergast).
124. Columbia-Southern Chemical Corp., GB
795 433, 1955 (J. R. Cox, M. Korach).
125. Pittsburgh Plate Glass Co., GB 924 625, 1958
(M. Q. Freeland, P. R. Johnston).
126. Mitsubishi Gas, US 4 999 179, 1989 (Y.
Sugihara, S. Shimokawa).
127. E. I. du Pont de Nemours & Co., DE-AS
1 036 225, 1956 (J. W. Sprauer).
128. Laporte Chemicals Ltd., GB 841 323, 1956
(H. B. Hopkins).
129. FMC Corp., US 3 043 666, 1960 (G. W.
Siwinski).
130. Laporte Chemicals Ltd., DE-AS 1 467 091,
1964 (S. D. Williams).
131. E. I. du Pont de Nemours & Co., US 3 016 290,
1956 (A. A. Elston, W. S. Hinegardner).
132. Oxysynthèse, FR 1 539 843, 1967 (P. Thirion).
133. Foret, ES 547 571, 1985 (R. P. Gracia, V. O.
Bendicho, J. M. Lopez Martinez).
134. E. I. du Pont de Nemours & Co., US
4 879 043, 1988 (J. H. Boughton, R. A. Butz,
H. C. T. Cheng, J. R. Dennis, B. T. Hannon, J.
H. Weigel).
135. Kemira, EP-A 703 189, 1995 (I. Turunen).
136. Mitsubishi Gas, EP 774 442, 1996 (M. Hattori,
Y. Minamikawa, S. Murakami).
137. Ausimont, EP 930 269, 1999 (U. P. Bianchi,
U. Leone, M. Lucci).
138. Mitsubishi Gas, US 5 534 238, 1995 (S.
Kajiwara, H. Serizawa).
139. Solvay Interox, DE-OS 4 222 109, 1992 (H.
Honig).
34 Hydrogen Peroxide
140. Solvay Interox, WO 92 06 918, 1991 (J. R.
Crampton, F. R. F. Hardy, M. H. Millar, G. W.
Morris).
141. Sumitomo, US 5 906 738, 1994 (Y. Matsuda,
A. Morisaki, Y. Sawaguri).
142. Solvay, EP 1520839, 2003 (R. Owen, J. Bosse,
M. Sell).
143. E. I. du Pont de Nemours & Co., US
4 985 228, 1990 (K. Kirksey).
144. Solvay Interox, WO 90 11 967, 1990 (N. D.
Feasey, G. W. Morris).
145. FMC Corp., US 5 266 298, 1992 (T.
Frianeza-Kullberg, D. Gibilisco, J. L.
Manganaro, J. R. Reed).
146. Laporte Chemicals Ltd., US 3 073 755, 1963
(R. H. Banfield, A. M. Hildon).
147. Columbia-Southern Chemical Corp., US
2 960 448, 1960 (W. S. Straub, C. J.
Sindlinger).
148. Degussa, US 3 060 105, 1962 (J. Müller);
Degussa, US 3 714 342, 1970 (G. Käbisch).
149. FMC Corp., US 3 152 052, 1964 (T. M.
Jenney, D. H. Porter, R. H. Rice).
150. Sulzer, EP 419 406, 1990 (A. Meili).
Sulzer, EP 835 680, 1996 (A. Meili).
151. Kemira, EP-A 727 388, 1996 (P. Johnsson, T.
Mattila, K. Saari).
152. Solvay Interox, EP 529 722, 1992 (L.
Signorini, K. Glinos).
Solvay, EP 1 090663, 1999 (S. Bloomfield, G.
E. Williams).
153. FMC, US 2 724640, 1947 (G.G. Crewson, J.
R. Ryan).
154. Degussa, WO 2002 044083, 2001, (M. Creutz,
S. Gross, S. Nordhoff, R. Wagner).
155. E. I. du Pont de Nemours & Co., US
2 479 111, 1949 (C. R. Harris).
156. Shell Development Co., US 2 871 104, 1951
(F. F. Rust).
157. N. V. De Bataafsche Petroleum Maatschappij,
GB 708 339, 1954.
158. N. V. De Bataafsche Petroleum Maatschappij,
BE 1 002 295, 1953 (J. R. Skinner, S. E.
Steinle).
159. Chem. Eng. News 35 (1957) no. 48, 26.
160. Chem. Week 125 (1979) no. 18, 13.
161. Ullmann, 4th ed., 17, 694.
162. E. J. Pneva, Chimitscheskoja Promyschlenost
9 (1976) 19.
163. Khimprom, SU 1 685 865, 1991 (V. V.
Chikurov, I. A. Militsin, L. K. Zharikov).
164. ARCO, US 4 897 252, 1989 (R. N. Cochran,
L. M. Candela).
165. ARCO, US 4 975 266, 1989 (R. S. Albal, R.
N. Cochran, L. M. Candela).
166. ARCO, US 5 039 508, 1990 (R. N. Cochran,
L. M. Candela).
167. D. F. Dong, J. R. Jackson, Int. Forum
Electrolysis Chem. Ind., 9th Electrosynthesis,
Lancaster, NY 1995, p. 215–228.
168. P. Askenasy, R. Rose, Z. Anorg. Allg. Chem.
189 (1930) 1, 10.
169. J. Lawrence, R. B. Peacock, C. W. Lefeuvre,
Bios Final Report 1432, (1946), no22.
170. D. J. Campbell, E. F. Lefson, US 2 497 812,
1950.
171. Atlantic Richfield Co., US 3 891 748, 1973.
172. Atochem, Oxysynthese, EP 237 402, 1987
(J.-P. Schirmann, M. Pralus).
173. Wacker-Chemie, US 312 421, 1964 (W.
Lohringer, J. Sixt).
174. Celanese Corp., US 3 341 297, 1967 (A. F.
MacLean, A. L. Stautzenberger).
175. Elektrochem. Werke München, US 2 015 040,
1935 (A. Pietzsch).
176. Mitsubishi Denki K. K., DE-OS 4 106 980,
1991 (S. Endoh, K. Namba, S. Yagi, K.
Maeda).
177. Chemical Week 1987, December 9, 20.
178. E. I. du Pont de Nemours & Co., US
4 681 751, 1985 (L. W. Gosser).
179. Chem. Eng. News, 17 Oct 2005, 83 (42), p 13.
180. Atomic Energy of Canada Ltd., EP 366 419,
1989 (K. T. Chuang).
181. J. R. Kosak, Chem. Ind. (Dekker) 62 (1995)
115–124.
182. Mitsubishi Gas Chemical Company Inc., EP
492 064 (Y. Hiramatsu, Y. Ishiuchi, H.
Nagashima).
183. Mitsubishi Gas Chemical Company Inc., EP
498 166 (Y. Hiramatsu, Y. Ishiuchi, H.
Nagashima).
184. Hydrocarbon Technologies Inc., US 6 168 775
(B. Zhou, L-K. Lee).
185. E. I. du Pont de Nemours & Co., US 5 194 242
(M. A. Paoli).
186. Advanced Peroxide Technology Inc., WO
98/31629 (H. Huckins).
187. Atofina, WO 01/05498 (M. Devic).
188. Hydrocarbon Technologies Inc., US
2004/0018143 A1 (B. Zhou, M. Rueter).
189. Eka Chemicals AB, WO 99/32398 (M.
Nyström, J. Wanngard, W. Herrmann).
190. Eka Chemicals AB, WO 99/32399 (M.
Nyström, J. Wanngard, W. Herrmann)
191. Halcon, US 4 462 978, 1982 (W. F. Brill).
192. BOC, US 4 711 772, 1986 (S. E. Jacobson).
193. Enichem, EP-A 788 998, 1997 (R. Bortolo, R.
D’Aloisio, D. Bianchi).

194. Enichem, EP-A 808 796, 1997 (D. Bianchi, R.
Bortolo, R. D’Aloisio, M. Ricci, S. Soattini).
D. Bianchi, R. Bortolo, R. D’Aloisio, M.
Ricci, Angew. Chem. 111 (1999) 734;
J. Mol. Catal. A: Chem 150 (1999) 87.
195. C. Querci, R. D’Aloisio, R. Bortolo, M. Ricci,
D. Bianchi, J. Mol. Catal. A: Chem. 176
(2001) 95–100.
196. Solvay Interox, US 5 545 391, 1992 (P. A.
Clapp, D. F. Evans, A. F. E. McKenna).
197. Canadian Industries, BE 660 719, 1965 (D.
Hudson, D. R. Lusby, F. Van Zeggeren).
198. Kymmene AB, DE 1 054 970, 1954 (R. M. R.
Örnhjelm).
199. Kymmene AB, DE 1 036 829, 1955 (R. M. R.
Örnhjelm).
200. Kymmene AB, DE 1 063 131, 1956 (R. M.
Rafaelson).
201. T. S. Sheriff, J. Chem. Soc. Dalton Trans.
1992, 1051.
202. M. Traube, Ber. Dtsch. Chem. Ges. 15 (1882)
2434.
203. F. Fischer, O. Priess, Ber. Dtsch. Chem. Ges.
46 (1913) 698.
204. B. Kastening, W. Faul, Chem. Ing. Tech. 49
(1977) no. 11, 911.
205. J. A. McIntyre, R. F. Philips in T. R. Beck et
al. (eds.): Electrochemical Process and Plant
Design, Pennington, NJ, 1983, pp. 79–97.
206. Dow Chemical, EP 86 896, 1982 (J. A.
McIntyre, R. F. Philips).
207. H-D Tech, EP 216 428, 1986 (A. L. Clifford,
D. F. Dong).
208. Metallgesellschaft, WO 94 24 336, 1994 (E.
Hillrichs, M. Kienberger, U. Sander).
209. Occidental, US 4 384 931, 1981 (R. J.
Jasinski, C. G. Kuehn).
210. P. C. Foller, R. T. Bombard, J. Appl.
Elektrochem. 25 (1995) 613–627.
211. C. Oloman: Electrochemical Processing for
the Pulp and Paper Industry, The
Electrochemical Consultancy Ltd., Romsey,
Hants, England, 1996, pp. 143–152.
212. H-D-Tech, US 5 106 464, 1990 (D. J. Rogers,
A. L. Clifford).
213. S. P. Webb, J. A. McIntyre: “Generation of
Hydrogen Peroxide in a Shorted Fuel Cell”,
Int. Forum Electrolysis Chem. Ind., 10th
Electrosynthesis, Lancaster, NY 1996, p.
213–225.
(1996), p. 213–225.
214. Dow Chemical, WO 95 30 474, 1995 (J. A.
McIntyre, S. P. Webb);
Dow Chemical, WO 97 13 006, 1996 (J. A.
McIntyre, S. P. Webb).
Hydrogen Peroxide 35
215. Laporte Chemicals Ltd., GB 893 069, 1959 (P.
V. Rutledge, D. S. Bell).
216. Columbia-Southern Chemical Corp., US
2 904 517, 1959 (P. H. Baker).
217. S. Croft, B. C. Gilbert, J. R. L. Smith, J. K.
Stell, W. R. Sanderson, J. Chem. Soc. Perkin
Trans. 2 1992, 153.
218. E. Rabald, D. Behrens (ed.),
DECHEMA-Werkstoff-Tabelle
Wasserstoffperoxid, DECHEMA, Frankfurt
1971; 43.
Ergänzungslieferung, DECHEMA, Frankfurt,
2000;
R. Bloom, L.E. Weeks, Ch. W. Raleigh,
Corrosion 16 (1960) 164t–170t.
219. J. Mackenzie, Chem. Eng. 1990, 84–90.
220. C. N. Satterfield, G. M. Kavanagh, H. Resnick,
Ind. Eng. Chem. 43 (1951) no. 11, 2507.
221. J. M. Monger, H. J. Baumgartner, G. C. Hood,
C. E. Sanborn, Journ. Chem. Eng. Data 9
(1964) no. 1, 119.
222. E. S. Shanley, P. F. Strong, L. R. Darbee, N. D.
Lee, Ind. Eng. Chem. Fundam. 9 (1970) no. 4,
574.
223. C. N. Satterfield, P. J. Cocotti, A. H. R.
Feldbrugge, Ind. Eng. Chem. 47 (1955) 1040.
224. Bur. Mines Inf. Circ. 8387 (1968).
225. E. Haeuseler, Explosionsstoffe 6/7 (1953) 64.
226. a) N. Nimmerfroh, D. Pauls, S. McMahon,
Propulse HTP Manufaxture: Quality,
Transportation and Handling, 36 th AIAA /
ASME / SAE / ASEE Joint Propulsion
Conference, Huntsville / AL, 2000; b) N.
Nimmerfroh, E. Walzer, C. Brossmer,
European Space Agency, Special Publication
SP-484 (2001) 65–73.
227. E. S. Shanley, F. P. Greenspan, Ind. Eng.
Chem. 39 (1947) no. 12, 1536.
228. J. M. Monger, H. Selbo, D. C. Lehwalder, J.
Chem. Eng. Data 6 (1961) no. 1, 23.
229. C. W. Jones: Applications of Hydrogen
Peroxide and Derivatives, RSC Clean
Technology Monographs, The Royal Society
of Chemistry, Cambridge 1999.
230. Bur. Mines Rep. Invest. 5877 (1961).
231. O. Klais, Thermochimica Acta 225 (1993)
213–222.
232. P. Urban, L. Bretherick: Handbook of Reactive
Chemical Hazards, 6th ed.,
Butterworth-Heinemann, Oxford 1999.
233. Shell Chemical Company, Concentrated
Hydrogen Peroxide, Summary of Research
Data on Safety Limitations, New York 1961.
36 Hydrogen Peroxide
234. A. Schreck, A. Knorr, K. D Wehrstadt, P. A.
Wandrey, T. Gmeinwieser, J. Steinbach,
Journal of Hazardous Materials A108 (2004)
1–7.
235. PetroChem. News 43 (2005) no. 44, 1.
236. European Center of Ecotoxicology and
Toxicology of Chemicals (ECETOC), Special
Report No. 10, Hydrogen Peroxide OEL
Criteria Document, ECETOC, Brussels 1996.
237. S. A. Kelly, C. M. Havrilla, T. C. Brady, K. H.
Abramo, E. D. Levin, Env. Health Perspect.
106 (1998) 375–384.
238. CEFIC Peroxygens hydrogen peroxide risk
assessment task force, Preliminary
Experiments for the Determination of
Hydrogen Peroxide in Blood, unpublished
report, (1999).
239. European Union, Institute for Health and
Consumer Protection, European Chemicals
Bureau, Existing Substances 2nd priority list
38 (2003) Risk Assessment Report Hydrogen
peroxide, European Commission Joint
Research Centre, EUR 20844 EN.
240. CEFIC Peroxygen Sector Group, Hydrogen
Peroxide: 28-day Inhalation Study in Rats
unpublished report (2002).
241. V. Riihimäki, A. Toppila, P. Piirilä, E.
Kuosoma, P. Pfäffli, P. Tuomela, J. Occup.
Health 44 (2002) 433–438.
242. A. Romanovski, J. R. Murray, M. J. Huston,
Pharm. Acta Helv. 35 (1960) 354–357.
243. CEFIC Peroxygen Sector Group: Hydrogen
Peroxide 13-Week Drinking Water Study with
6-Week Recovery Period in C57BL/6NCrlBR
Mice, unpublished report, (1997).
244. M. Fenech, J. Crott, J. Turner, S. Brown,
Mutagenesis, 14 (6) (1999), 605–612.
245. CEFIC peroxygen sector group, Coming
Hazelton Report No. 514/24, 1997
(unpublished).
246. Threshold Limit Values for Chemical
Substances and Biological Exposure Indices,
ACGIH, Cincinnati 2005.
247. Sociaal Economische Raad (NL), MAC
Waarden, Waterstoffperoxide 90 % (2005).
248. Koninklijk besluit van 11 maart 2002
betreffende de bescherming van de gezondheid
en de veiligheid van de werknemers tegen de
risico’s van chemische agentia op het werk
(B.S. 14.3.2002, ed. 2; erratum B.S. 26.6.2002,
ed. 2; erratum B.S. 4.12.2002) BIJLAGE I
Grenswaarden voor beroepsmatige
blootstelling.
249. Réferences françaises, INRS, ND2098,
Valeurs limites d’exposition professionnelle
aux agents chimiques en France (2005).
250. TRGS 900, Technische Regeln für
Gefahrstoffe, Grenzwerte in der Luft am
Arbeitsplatz, Bekanntmachung des BMA
Ausschuß für Gefahrstoffe, BArbBl 7/8
(2004).
251. Health and Safety Executive, EH40/ 2005
Occupational Exposure Limits (2005).
252. European Center of Ecotoxicology and
Toxicology of chemicals (ECETOC), Joint
Assessments of Commodity Chemicals No. 22,
Hydrogen Peroxide, ECETOC, Brussels 1993.
253. R. P. Wayne: The photochemistry of ozone. In:
O. Hutzinger (ed.): The handbook of
environmental chemistry 2E, reactions and
processes, Springer, Berlin, 10 (1988).
254. H. Sakugawa, I. R. Kaplan, W. Tsai, and Y.
Cohen, Envir Sci. Technol. 24 (1990)
1452–1462.
255. J. G. Calvert and W. R. Stockwell, Envir. Sci.
Technol. 17 (1983) 28A–443A.
256. W. C. Schumb, C. N. Satterfield, R. L.
Wentworth, (eds): Hydrogen Peroxide.
Reinhold Publishing, New York, 1955, 64–68,
181, 458–464, 535–547.
257. L. Kleinman, J. Geophys. Res. 91 (10) (1986)
889–904.
258. E. Hellpointner, S. Gaeb, Nature 337 (1989)
631–634.
259. W. J. Cooper, R. G. Zika, Science 220 (1983)
711–712.
260. R. G. Zika, J. W. Moffett, R. G. Pestane, W. J.
Cooper, E. S. Saltzman, Geochim. Cosmochim.
Acta 49 (1985) 1173–1184.
261. J. M. Moffett, and R. G. Zika, Environ. Sci.
Technol. 21 (1987) 804–810.
262. W. J. Cooper and D. R. S. Lean, Environ. Sci.
Technol. 23 (11) (1989) 1425–1428.
263. W. Guhl, A. G. Berends, Surfactants 2 (2000)
98–102.
264. A. H. C. Groeneveld, W. A. de Groot:
Activated sludge, respiration test with
hydrogen peroxide, Unpublished report
(1999), Solvay pharmaceuticals, report
A.SOL.S.003.
265. Degussa AG, US-IT No. 91-0114-DKO, 1991
(unpublished).
266. C. A. Clark: Ph. D. Thesis, University of Bath,
1991.
267. M. Fitzpatrick, C. B. Schreck, R. L. Chitwood,
The Progressive Fish Culturist, 57 (1995)
153–155.
268. P. Klerks, P. C. Fraleigh, J. AWWA 1991,
92–100.
269. M. A. Xenopoulos, D. F. Bird, Photochem.
Photobiol. (1997) 471–478.
 Hydrogen Sulfide 1

Hydrogen Sulfide
Francois Pouliquen, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 1)
Claude Blanc, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 2)
Emmanuel Arretz, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chaps. 3 and 8)
Ives Labat, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 3)
Jacques Tournier-Lasserve, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 4)
Alain Ladousse, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 5)
Jean Nougayrede, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chaps. 5 and 8)
Gérard Savin, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chaps. 6 and 8)
Raoul Ivaldi, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 7)
Monique Nicolas, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 9)
Jean Fialaire, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 10)
René Millischer, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 10)
Charles Azema, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 11)
Lucien Espagno, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 11)
Henri Hemmer, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 11)
Jacques Perrot, Elf Aquitaine, Tour Elf, Paris La Defense, France(Chap. 11)

1. Introduction . . . . . . . . . . . . . . . 2 4.2. Recovery from Gas . . . . . . . . . . . 7

2. Physical Properties . . . . . . . . . . . 2 5. Environmental Protection . . . . . . 9
3. Chemical Properties . . . . . . . . . . 4 6. Analysis . . . . . . . . . . . . . . . . . . 11
3.1. Acid Properties . . . . . . . . . . . . . 4 7. Storage, Handling, and Transporta-
3.2. Reducing Properties . . . . . . . . . . 4 tion . . . . . . . . . . . . . . . . . . . . . 12
3.3. Reactions with Organic Compounds 4 7.1. Storage and Handling . . . . . . . . . 12
3.3.1. Hydrogen Sulfide as a Nucleophilic 7.2. Transportation . . . . . . . . . . . . . . 12
Reagent . . . . . . . . . . . . . . . . . . . 4
8. Uses . . . . . . . . . . . . . . . . . . . . . 12
3.3.2. Hydrogen Sulfide as a Radical Reagent 6
3.3.3. Reaction of Hydrogen Sulfide in the 9. Economic Aspects . . . . . . . . . . . . 13
Form of Hydrosulfide . . . . . . . . . . 6 10. Toxicology . . . . . . . . . . . . . . . . . 14
4. Production . . . . . . . . . . . . . . . . 7 11. Occupational Health and Safety . . 15
4.1. Production by Chemical Reaction . 7 12. References . . . . . . . . . . . . . . . . . 18

1. Introduction natural gas that had to be eliminated. However,

Hydrogen sulfide was first described as a natu-
ral compound by alchemists in the Middle Ages.
The successful development of the sour gas field
at Lacq (France) by the Société Nationale des
Pétroles d’Aquitaine in 1957 modified the econ-
omy of hydrogen sulfide and allowed worldwide
exploitation of natural gas with a high hydrogen
sulfide content. Before this, hydrogen sulfide
was regarded as a toxic, corrosive component of
it has now become an important raw material for
producing sulfur and thioorganic compounds.
Some of the data and indications given are
the result of Elf Aquitaine’s experience, either
from the Lacq field or from various gas and oil
desulfurization plants that the company has built
around the world.
Natural Occurrence The quantity of hydro-
gen sulfide naturally released into the atmo-

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 467
2 Hydrogen Sulfide

sphere is thought to vary annually from 100×106 encountered in the development of very sour oil
to 324×106 t [1–3], with half of this coming or gas fields are also discussed in the literature:
from volcanoes [4, 5], flooded ground [6–9], or avoidance of formation of SO2− 4 in edgewater
hydrogeological sources [10, 11] and the other [44]; tools for studying hydrogen sulfide wells
half from oceans [12,13]. [45,46]; neutralization of hydrogen sulfide dur-
Hydrogen sulfide has been identified in the ing drilling [47–52]; and reserves located at great
atmosphere of stars by spectroscopic methods. depth [53] or under high pressure [19].
Calculations have been made of the thermal
equilibrium of chemical species found in the at-
mosphere of several planets [14]. 2. Physical Properties
Hydrogen sulfide is present as an occluded
gas in minerals such as topaz, tourmaline, water Hydrogen sulfide [7783-06-4], H2 S, is also
chrysolite, obsidian, and granite. The hydrogen known as monosulfane, M r 34.08. The gas is
sulfide found in ore-bearing deposits may orig- colorless, heavier than air, and has a character-
inate from bacteria: its reducing properties ac- istic rotten-egg odor.
count for the presence of uranium in the form of Physical properties are listed below:
nonoxidized mineral deposits.
Salt mines also contain hydrogen sulfide, bp (101.3 kPa) −60.28 ◦ C
Triple point (23.2 kPa) −85.49 ◦ C
sometimes mixed with other gases. Volcanoes
Critical temperature 100.25 ◦ C
release large quantities of hydrogen sulfide in Critical pressure 8.96 MPa
the fumaroles, and in gases and vapors produced Critical density 347.62 kg/m3
from solfataras. Density of solid [54] at
Flooded ground (swamps, rice fields) also −195 ◦ C 1.217 kg/m3
−170 ◦ C 1.166 kg/m3
contains hydrogen sulfide which is generally Density of liquid [55] at
formed in the soil by bacterial reduction of sul- bp 949.0 kg/m3
fates. However, its occurrence may also depend −30 ◦ C 893.8 kg/m3
on the type of soil, the presence of organic ma- 9.7 ◦ C 813.8 kg/m3
90.4 ◦ C 532.6 kg/m3
terial, and the degree of aeration. Density of vapor [55] at
Water is another natural medium that contains −80 ◦ C 0.71 kg/m3
hydrogen sulfide. The dissolved or gaseous hy- bp 1.99 kg/m3
drogen sulfide found in lakes, saltwater ponds, −30 ◦ C 6.82 kg/m3
56.9 ◦ C 71.65 kg/m3
and marine sediments originates from sulfate-
Vapor pressure of [55] at
reducing bacteria. Mineral waters from ther- −30 ◦ C 0.381 MPa
mal springs are studied either in vitro, to ex- 0 ◦C 1.032 MPa
amine sulfate reduction by bacteria, or in situ, 30 ◦ C 2.275 MPa
where the hydrodynamic effects (water velocity 60 ◦ C 4.335 MPa
Viscosity of vapor (101.3 kPa) [56] at
in soil, mineralization, and temperature) are cou- 0 ◦C 0.01179 mPa · s
pled with the biological production of hydrogen 15 ◦ C 0.01241 mPa · s
sulfide. Wastewater of industrial [15] or other 50 ◦ C 0.01398 mPa · s
origin [16,17] often contains hydrogen sulfide. 100 ◦ C 0.01608 mPa · s
Viscosity of liquid (101.3 kPa) [57] at
Finally, hydrogen sulfide is found in hydro- −11.5 ◦ C 0.00169 mPa · s
carbon reserves: natural gas, associated gas, and 15 ◦ C 0.00133 mPa · s
oil. An extensive bibliography exists on this sub- 30 ◦ C 0.00116 mPa · s
ject. Further information on particular reservoirs 50 ◦ C 0.00094 mPa · s
Surface tension at
can be found in the following references: Federal
−60.2 ◦ C [58] 25.43 kN/m
Republic of Germany [18, 19]; Lacq (France) 25 ◦ C [59] 11.3 kN/m
[20,21]; Astrakhan (former Soviet Union) [22– 40 ◦ C [59] 8.7 kN/m
32]; Orenbourg (former Soviet Union) [33–41]; Refractive index (n0D ) of gas at 0.1 MPa
United States [42, ; gives the composition of 349 and 0 ◦ C [60] 1.000644
Dipole moment [61] 3.269×10−30 C · m
reservoirs]; and People’s Republic of China [43] (0.98 debye)
(reservoir containing the highest known percent- Latent heat of fusion 2.378 kJ/mol
age, i.e., 90 %, of hydrogen sulfide). Problems
 Hydrogen Sulfide 3
Flammability limits in air (20 ◦ C and
101.3 kPa) [63] Table 4. Solubility of hydrogen sulfide in water and organic
Lower 4 vol % solvents
Upper 44 vol %
Autoignition temperature Solvent Tempera- Partial pressure Mole fraction
ture, of H2 S, MPa of H2 S in
at 101.3 kPa 290 ◦ C ◦
C liquid phase
Thermal conductivity of gas (101.3 kPa)
[64] at Water [64] 30 1.718 0.0238
0 ◦C 1.340 kJ m−1 s−1 K−1 50 1.708 0.0170
80 ◦ C 1.784 kJ m−1 s−1 K−1 Methanol [65] 0 0.0533 0.0286
120 ◦ C 2.010 kJ m−1 s−1 K−1 −25.6 0.0533 0.0655
200 ◦ C 2.433 kJ m−1 s−1 K−1 Propylene 40 2.378 0.7246
carbonate [66] 100 4.96 0.5459
Compressibility factors of hydrogen sulfide N-Methyl-2- 23.5 0.0415 0.0580
gas and liquid are given in Table 1. The latent pyrrolidone [67] 35 0.1013 0.0971
n-Decane [68] 15 0.1013 0.0541
heat of vaporization and specific heat of the liq- 25 0.1013 0.0465
uid are listed in Table 2. Table 3 gives specific
heats of hydrogen sulfide gas. The solubilities
of hydrogen sulfide in water and some organic
solvents are listed in Table 4. 3. Chemical Properties
Table 1. Compressibility factor z of gaseous and liquid hydrogen
sulfide [55] 3.1. Acid Properties
Temperature, K Pressure, MPa z (gas) z (liquid)
Hydrogen sulfide in aqueous solution is a weak
190.00 0.0271 0.99321 0.00059 acid with two dissociation constants:
230.00 0.2247 0.96164 0.00436
270.00 0.94015 0.8888 0.01697 H2 S
HS− + H+ K 1 = 10−7 (18 ◦ C)
310.00 2.6642 0.77786 0.04705 HS−
S2− + H+ K 2 = 10−13 – 10−14
350.00 5.9719 0.60792 0.11345
370.00 8.4622 0.42049 0.19604 Saturation of the first acid function with
sodium hydroxide solution is an exothermic re-
Table 2. Heat of vaporization (Qvap ) and specific heats (cp , cv ) of action:
liquid hydrogen sulfide [55]
NaOH + H2 S −→ NaSH + H2 O ∆H = − 36.7 kJ/mol
Temperature, K Qvap , kJ/mol cp , cv ,
J mol−1 K−1 J mol−1 K−1 Hydrogen sulfide can precipitate salts of
190 19.52 67.53 43.89
heavy metals (Hg, Pb, Ag, Bi, Cu, Cd, As, Sb, Sn,
230 17.96 69.64 42.78 Co, Ni, Fe, Zn, Mn, Al) as sulfides. It dissolves
270 15.89 71.67 39.37 and reacts with amines and can be released by
300 13.95 75.20 35.45 heating. Numerous processes for removing hy-
330 11.50 86.83 31.04
360 7.37 163.40 30.80 drogen sulfide from natural gas (i.e., gas sweet-
ening processes) are based on its absorption by
amines (see Chap. 5).
Table 3. Specific heats of gaseous hydrogen sulfide [55]

Pressure, Temperature, cp , cv ,
MPa K J mol−1 K−1 J mol−1 K−1 3.2. Reducing Properties
0.05 200 34.09 25.38 The standard redox potentials of hydrogen sul-
300 34.39 25.98
400 35.70 27.34 fide at 25 ◦ C follow [69]:
0.1013 220 34.54 25.59
300 34.55 26.02 S + 2 H+ + 2 e−
H2 S E 0 = + 0.14 V (pH = 0)
400 35.76 27.35 S + 2 e−
S2− E 0 = − 0.48 V (pH = 14)
1 300 39.13 27.41
400 37.13 27.65 At pH 0, hydrogen sulfide is barely dissoci-
500 38.02 29.08 ated; at a pH of ca. 12, partial dismutation mod-
5 400 47.82 30.20 ification of the oxidation level of sulfur can oc-
500 41.62 29.86
600 41.38 31.15 cur and the sulfide is oxidized to form oxygen –
sulfur compounds.
4 Hydrogen Sulfide
Hydrogen sulfide reacts with numerous ox-
idizing agents to form sulfur compounds with
varying degrees of oxidation. Depending on op-
erating conditions and the oxidizing species con-
cerned, sulfur oxides (SO2 , SO3 ), salts (S2 O2−
3 ,
SO2−3 , SO 2−
4 ), polythionic compounds, sulfur,
and sulfanes are obtained.
1) Aqueous solutions of hydrogen sulfide can
be oxidized in air.
2) In air, solutions of ammonium sulfide form
orange polysulfides and even oxygen – sulfur
compounds.
3) In an acid medium, iodine quantitatively oxi-
dizes hydrogen sulfide to sulfur. This reaction
is used to analyze hydrogen sulfide.
4) Strong oxidizing agents (e.g., MnO− −
4 , ClO ,
Cl2 , and Br2 ) react with hydrogen sulfide to
give sulfate ions.
5) Hydrogen sulfide can react with sulfur to
form sulfanes. The sulfanes can be converted
back to hydrogen sulfide by base catalysis
[70].
6) Hydrogen sulfide or alkaline hydrosulfides
can reduce polysulfides to thiols [71]:
R−S−S−R + 2 NaSH −→ 2 RSH + Na2 S2
This reaction is used for the preparation of
thiokol polymers.
7) Hydrogen sulfide can reduce sulfoxides to
sulfides [72]:
R−SO−R + H2 S −→ R−S−R + H2 O + S
8) The reaction of hydrogen sulfide and sulfur
dioxide (Claus reaction) is used for the pro-
duction of sulfur (→ Sulfur):
2 H2 S + SO2 −→ 3/8 S8 + 2 H2 O
3.3. Reactions with Organic Compounds
Hydrogen sulfide reacts with numerous organic
compounds directly, as an intermediate (e.g.,
sodium, potassium, or ammonium hydrosul-
fide), or as an alkaline sulfide.
Hydrogen sulfide itself can behave as ei-
ther a nucleophilic reagent or a radical reagent,
whereas hydrosulfides or sulfides behave solely
as nucleophilic reagents.
3.3.1. Hydrogen Sulfide as a Nucleophilic
Reagent
Hydrogen sulfide reacts as a nucleophilic
reagent in the presence of acidic, basic, or metal-
lic oxide catalysts.
Compounds such as alcohols or olefins react
with hydrogen sulfide to form thiols (RSH). Thi-
ols (with the exception of tertiary thiols) react
like hydrogen sulfide to give the corresponding
organic sulfides (RSR).
Substitution Reactions. Because of its
widespread industrial applications, the best
known substitution reaction is sulfhydrolysis
(thiolation) of alcohols. This reaction occurs cat-
alytically in the gas phase at ca. 300 ◦ C. Thiols
were first obtained by reacting hydrogen sul-
fide with alcohols on thorin or on pumice im-
pregnated with thorin [73]. Subsequently, more
selective, more active catalysts were developed
based on activated alumina containing promot-
ers such as soda or potash [74], metallic oxides,
alkaline salts of transition-metal oxo acids [75],
and heteropolyacids or their alkaline salts [76].
Primary alcohols are converted very effi-
ciently into the corresponding thiols:
H2 S + ROH −→ RSH + H2 O
However, with secondary and especially ter-
tiary alcohols, dehydration reactions are more
prevalent.
Under catalytic reaction conditions simi-
lar to those used for alcohols, hydrogen sul-
fide reacts with aliphatic ethers to produce thi-
ols. Cyclic ethers are converted to the cor-
responding sulfides by catalysis in the gas
phase. For example, reaction of hydrogen sulfide
with tetrahydrofuran yields tetrahydrothiophene
[110-01-0] [77]:
A similar reaction with furan gives thiophene
[110-02-1] [78]:
 Hydrogen Sulfide 5

Hydrogen sulfide reacts with acetic anhy-

dride in the presence of basic resins to form
thioacetic acid [507-09-5] [79]:
An industrially important example of this re-
H2 S + CH3 COOCOCH3 −→ CH3 COSH + CH3 COOH action is the addition of hydrogen sulfide to
cyanamide to form thiourea [62-56-6]:
Addition Reactions. Catalytic addition of
hydrogen sulfide to olefins follows Markovni-
kov’s rule (i.e., the sulfhydryl group adds pref-
erentially to the more highly substituted carbon
of the double bond).
With aldehydes and ketones, hydrogen sul-
Linear olefins react with hydrogen sulfide in
fide yields dithiols or cyclic products, depending
the presence of acid catalysts (e.g., acid zeo-
on conditions [92]:
lites, cation-exchange resins, alumina – silicas)
to yield secondary thiols [80]. With branched
olefins and acidic catalysts such as aluminum
chloride, alumina – silica [81], zeolites [82], or
cation-exchange resins [83], tertiary thiols are
formed. Ethylene and cyclohexene are converted
into thiols on metal oxide catalysts [84].
Hydrogen sulfide adds to the double bond of
acrylic derivatives (nitrile, acids, esters) in the
presence of basic catalysts such as amines [85],
anion-exchange resins [86], or magnesium ox-
ide [86]. Thus, with formaldehyde, methanedithiol
Nucleophilic addition of the thiol function (HSCH2 SH) [74-93-1] is obtained [93]. Ace-
occurs at the carbon in the β-position with re- tone reacts with sulfur and hydrogen sulfide in
spect to the substituent group: an ammoniacal medium to form a cyclic poly-
sulfide [4475-72-3] [94]:
H2 S + CH2 =CHX −→ HSCH2 CH2 X
X =−CN, −COOH, −COOR

Base-catalyzed cleavage of epoxides with

hydrogen sulfide gives mercaptoalcohols
(hydroxythiols):
Basic catalysts may be mineral bases,
amines, quaternary ammonium salts [87], anion-
exchange resins [88], zeolites [89], or guani-
dines [90].
With ethylene oxide, the main product
is mercaptoethanol [60-24-2]. Thiodiglycol or
bis(2-hydroxyethyl) sulfide [111-48-8] is ob-
tained as byproduct. The reaction can be cat-
alyzed by thiodiglycol without a base [91].
The addition of hydrogen sulfide to nitriles
is catalyzed by bases such as amines and gives
thioamides:
3.3.2. Hydrogen Sulfide as a Radical
Reagent
Hydrogen sulfide can react as a free radical:
HS−H −→ HS · + H ·
Generation of these two species by thermal
dissociation is difficult. Use of chemical initia-
tors or exposure to UV light is much more effi-
cient.
Radical addition of hydrogen sulfide to
olefins yields thiols. In practice, the reaction can
be initiated as follows:
1) Chemical initiators can be used at tempera-
tures corresponding to their thermal decom-
position. Azo derivatives, in particular azo-
bisisobutyronitrile, have been used [95] with
6 Hydrogen Sulfide

or without catalyst promoters such as organic

phosphines and phosphites [96].
2) Ultraviolet radiation can be employed. Re-
action proceeds either by direct photolysis
(λ = 253.7 nm), or via intermediate radical
initiators (λ = 300 – 400 nm). The initiators This method is used to synthesize thiogly-
are either chemical initiators that are decom- colic acid [68-11-1] and mercaptopropionic acid
posed at a lower temperature by photons [107-96-0].
(e.g., azo derivatives [97], phosphites [98], Other mercaptans are obtained by similar
acetophenone derivatives [99]) or aromatic techniques [103, 104]:
carbonyl derivatives such as benzophenone
[100] or thioxanthone [101], preferably as- NaSH + ClCH2 CH2 OR −→
sociated with organic phosphites [100, 101]. HSCH2 CH2 OR + NaClNaSH + Cl(CH2 )3 Si(OCH3 )3
Addition proceeds by a free-radical chain mech- −→ HS(CH2 )3 Si(OCH3 )3 + NaCl
anism and is of the anti-Markovnikov type, i.e., NaSH + C6 H5 CH2 Cl −→ C6 H5 CH2 SH + NaCl
the HS · radical adds to the least substituted car-
bon of the double bond. Acid chlorides react particularly well with
With a terminal olefin, the corresponding pri- alkaline hydrosulfides [105]. For example,
mary thiol is selectively formed. However, the thiobenzoic acid [98-91-9] is synthesized by re-
thiol may decompose into a thiyl radical that action of benzoyl chloride with sodium hydro-
can also add to the olefin to give the correspond- sulfide:
ing organic sulfide. The percentage of sulfide
formed can be minimized by using an excess of
hydrogen sulfide.
Photochemical addition of hydrogen sulfide
is also possible with olefins containing func-
tional groups (e.g., allyl chloride, acrylic acid,
and acrylates).

3.3.3. Reaction of Hydrogen Sulfide in the

Form of Hydrosulfide Some active esters may be subjected to
sulfhydrolysis with hydrosulfides to form the
For certain ionic reactions, hydrogen sulfide corresponding thiols [106, 107]:
must be supplied in the form of alkaline hydro-
sulfides, or ammonium hydrosulfide in an aque-
ous or aqueous – alcoholic solution.

Nucleophilic Substitution. Nucleophilic

substitution of halogen compounds is facili-
tated if the halogen is activated by its electronic
environment. Thus, nucleophilic substitution
cannot be used to obtain thiol compounds from
chlorinated hydrocarbons. However, if the chlo-
rine is activated by an α, β, or γ heteroatom
(O, Si, S), or by an electron-attracting group
(C=O, C=N), substitution proceeds easily, often
with total conversion of chlorine and excellent
selectivity [102]:
Nucleophilic Addition. In an aqueous –
alcoholic medium and in the presence of car-
bon disulfide, sodium hydrosulfide can react
selectively with activated olefins to give thi-
ols. Methyl mercaptopropionate [2935-90-2]
has been prepared by this technique [108]:

4. Production
Hydrogen sulfide can be produced by (1) chem-
ical reaction of sulfur with hydrogen or a hy-
drocarbon, or (2) hydrogen reduction or acid
decomposition of a sulfide. It can also be reco-
vered from natural gas, associated gases, and re-
finery gases.
4.1. Production by Chemical Reaction
Reaction of Sulfur with Hydrogen. Hydro-
gen sulfide is formed when hydrogen is fed into
sulfur at 350 ◦ C with or without a catalyst (e.g.,
bauxite, aluminum silicate, cobalt molybdate)
[109, 110]. High pressure also aids the reaction.
H2 + 1/2 S2
H2 S
In the laboratory, hydrogen is fed into molten
sulfur. Reaction occurs in a tube filled with
pumice [111–113]. In the industrial production
of hydrogen sulfide (Fig. 1), the reaction occurs
at 450 ◦ C and 0.7 MPa. Hydrogen sulfide is sep-
arated from sulfur vapor by cooling and solidi-
fying the sulfur on the walls of a heat exchanger.
Figure 1. Industrial production of hydrogen sulfide from hydrogen and sulfur
a) Sulfur makeup pump; b) Sulfur cooler; c) Sulfur recycle pump; d) Reactor; e) Electric heater; f) Quench tower; g) Gas
cooler
Hydrogen Sulfide 7
The exchanger is alternately cooled and heated
to remelt and eliminate the sulfur deposit. Oper-
ating data for a typical plant follow [114]:
Production capacity: 1 – 60 t/d
Consumption (per metric tonne of hydrogen sulfide produced):
Hydrogen (purity 98 %) 67 kg
Sulfur (purity 99 %) 1066 kg
Electrical power 266 kW · h
Typical analysis of product 95 vol % H2 S, 3 vol % H2 , 2 vol %
impurities
Reaction of Sulfur with Hydrocarbons.
Sulfur reacts with methane to form a mixture of
hydrogen sulfide and carbon disulfide. Carbon
disulfide can be hydrolyzed, giving additional
hydrogen sulfide [110] , [115,116]. With higher
alkanes (>C10 ) having a low vapor pressure,
reaction takes place at 200 – 300 ◦ C [115, 117].
Examples of laboratory preparation are given in
[118, 119].
Reduction of Sulfides. Hydrogen sulfide is
formed by reduction of sulfides (e.g., FeS2 , CuS)
with hydrogen at > 500 ◦ C.
Reaction of Sulfides with Dilute Acid. The
most convenient method of preparing hydrogen
8 Hydrogen Sulfide
Table 5. Industrial absorption processes for hydrogen sulfide
sulfide in the laboratory is to react sodium sul- recovery from gases
fide with dilute hydrochloric [120, 121] or sul-
Solvent Trade name Lincensing
furic [122] acid. Other sulfides (CaS, ZnS, FeS, company
Sb2 S3 ) can also be used. The purest hydrogen
sulfide is obtained from Sb2 S3 . Chemical absorption with
aqueous solutions
N-Methyldiethanolamine (5 N Ucarsol-MDEA Union
solution) Carbide
Diethanolamine (4 N solution) SNPA-DEA Elf
4.2. Recovery from Gas (→ Gas Aquitaine
Production, Chap. 5.4.2.;→ Gas Production, Diisopropanolamine (2 N solution) ADIP Shell
N-Methyldiethanolamine (4 N MDEA Elf
Chap. 5.4.3.; → Natural Gas) solution) Aquitaine
Diglycolamine (6 N solution) Econamine Fluor
Natural gas or gas associated with crude oil is Hindered amine solution Flexsorb Exxon
Chemical absorption with alkaline
sour, i.e., it contains varying quantities of hydro- salt solutions
gen sulfide, from traces to 70 – 80 vol %. This K2 CO3 , catalyst (amine borates) Catacarb Eickmeyer
hydrogen sulfide is nearly always associated
K2 CO3 , catalyst (diethanolamine) Benfield Union
with varying quantities of carbon dioxide. Re- Carbide
finery gases produced in hydrodesulfurization K2 CO3 , catalyst (arsenic salts) Giammarco- Giammarco
or cracking units have a very high ratio of hy- Vetrocoke
Methylaminopropionate Alkacid-M BASF
drogen sulfide to carbon dioxide. When the hy- Dimethylaminoacetate Alkacid-DIK BASF
drogen sulfide level of the gas reaches a certain Physical absorption
limit, partial or total sweetening (i.e., removal Anhydrous propylene carbonate Fluor Fluor
of hydrogen sulfide) is necessary to comply with Dimethyl ether of poly(ethylene Selexol Norton
glycol)
transportation, distribution, and antipollution re- Cold methanol Rectisol Lurgi
quirements. Sweetening gives an acid gas mix- N-Methylpyrrolidone Purisol Lurgi
ture of hydrogen sulfide and carbon dioxide that Aqueous solution of
diisopropanolamine (D) or
is generally used for sulfur recovery by the Claus methyldiethanol-
process (→ Sulfur) but also serves as a source of amine (M), sulfolane Sulfinol D (or M) Shell
hydrogen sulfide. Hydrogen sulfide is obtained Mixture of monoethanolamine or Amisol Lurgi
diethanolamine and methanol
from the gas Methyl isopropyl ethers of Sepasolv BASF
poly(ethylene glycol)
1) directly if the hydrogen sulfide – carbon
dioxide ratio is high enough to give an acid
gas with the desired hydrogen sulfide content Figures 2 and 3 show typical process schemes
(80 – 90 vol %) or if a selective sweetening for absorption with a chemical and a physical
process is used; solvent, respectively. The sour gas is washed in
2) after additional treatment, consisting of se- a countercurrent absorber. The rich solvent is
lective reprocessing (desired purity, 90 – regenerated by flashing and stripping. Flashing
95 %) or compression and subsequent dis- (physical solvent) and reboiling (chemical sol-
tillation (minimum purity 99.5 %). vent) are the main regeneration factors.
The production process used depends on the
Absorption. Hydrogen sulfide can be reco- characteristics of the gas to be treated and the
vered from gas containing sufficiently high desired hydrogen sulfide specifications.
quantities of it by (selectively) sweetening the If maximum selectivity with regard to hy-
gas with respect to carbon dioxide by chemical drogen sulfide is required, a process based on
or physical absorption [123–126]. absorption with a selective amine is chosen (a
Absorption by a chemical solvent (amines or tertiary alkanolamine such as N-methyldietha-
potassium carbonate) is a function of pressure nolamine, MDEA) [127].
and temperature. Absorption by a physical sol- Hydrogen sulfide reacts readily and directly
vent is primarily a function of pressure. with tertiary amines. Carbon dioxide, however,
Industrial absorption processes for gas sweet- can only react with tertiary amines indirectly in
ening are summarized in Table 5. the form of carbonic acid to produce amine car-
bonates. Carbonic acid first has to be formed by
 Hydrogen Sulfide 9

Figure 2. Recovery of hydrogen sulfide with a chemical solvent

a) Absorber; b) Flash drum; c) Amine – amine exchanger; d) Condenser; e) Reflux drum; f) Reflux pump; g) Regenerator;
h) Reboiler; i) Cooler; j) Lean amine high-pressure pump

Figure 3. Recovery of hydrogen sulfide with a physical solvent

a) Absorber; b) – d) First, second, and third flash drums; e) Semilean amine high-pressure pump; f) Amine – amine exchanger;
g) Reheater; h) Condenser; i) Reflux drum; j) Reflux pump; k) Hydrogen sulfide stripper; l) Reboiler; m) Lean amine high-
pressure pump; n) Power recovery; o) Recycle compressor
10 Hydrogen Sulfide
reaction of carbon dioxide with water and this
step is much slower than direct reaction. An im-
portant principle of selective absorption is there-
fore the selection of a gas – liquid contact time
that is long enough to eliminate hydrogen sul-
fide, but short enough to minimize the coabsorp-
tion of carbon dioxide.
If the partial pressure of hydrogen sulfide
in the sour gas is sufficiently high (≥ 0.3 MPa)
and significant quantities of hydrocarbons (a few
percent) can be tolerated, a physical solvent pro-
cess may also be chosen.
Distillation. The industrial distillation of an
acid gas in which the purity of hydrogen sulfide
is insufficient was developed by Elf Aquitaine
at its Lacq plant [128]. Hydrogen sulfide is ob-
tained either as a liquid or as a gas with a purity
> 99.5 %. The acid gas is first treated with ac-
tivated carbon to eliminate heavy hydrocarbons
and dried with a solid adsorbent (alumina, sil-
ica beads, or molecular sieves). It is then com-
pressed to 2.5 – 3.5 MPa and distilled in a tray
column or packed column. Liquid hydrogen sul-
fide is obtained at the bottom of the column and
easily vaporized for more convenient distribu-
tion.
If the acid gas has a very low hydrogen sulfide
content (H2 S/CO2 <0.2), separation of hydro-
gen sulfide and carbon dioxide can be improved
by introducing a C3 – C6 alkane above the col-
umn feed [129]. The hydrogen sulfide obtained
is slightly less pure.
5. Environmental Protection
Hydrogen sulfide is not only produced industri-
ally, it also originates from natural sources (see
also Chap. 1):
1) diffusion into underground water from oil
and natural gas wells,
2) formation in lakes by aerobic bacterial fer-
mentation,
3) production during biological wastewater
treatment, and
4) formation as a degradation product in the
agricultural food industry, slaughterhouses,
and the paper industry.
In the environment, the toxicity of hydrogen sul-
fide is due to the hydrogen sulfide molecule itself
rather than to hydrosulfide or sulfur ions. For ex-
ample, bacterial metabolism is inhibited when
the hydrogen sulfide content of the surround-
ings reaches 70 – 200 mg/L [130]. Because hy-
drogen sulfide does not accumulate in living or-
ganisms, it cannot contaminate the food chain
[131]. Natural elimination of hydrogen sulfide
occurs mainly by diffusion into the atmosphere,
but it is also subjected to biological or chemi-
cal degradation to form sulfur oxides and sulfur.
In water, hydrogen sulfide is converted into ele-
mental sulfur or insoluble heavy-metal sulfides.
Accidental leaks of hydrogen sulfide gas can
be absorbed directly with aqueous solutions of
sodium carbonate, sodium hydroxide, sodium
hypochlorite, or iron(III) chloride. When liquid
hydrogen sulfide is spilled on the ground, adsor-
bents such as activated carbon should be used to
prevent the spread of liquid and reduce the emis-
sion of vapor [132]. Water polluted with hydro-
gen sulfide can be stripped and neutralized with
lime, sodium carbonate, or sodium bicarbonate.
Sulfur ions are then eliminated by precipitation.
Industrial effluents containing hydrogen sul-
fide must be treated or incinerated if their hy-
drogen sulfide content exceeds 0.001 vol %. The
final procedure depends on the sulfur content
and local discharge regulations. For example,
for residual gases in sulfur plants, treatment can
consist of
1) hydrogenation of residual sulfur compounds,
followed by washing with amines that selec-
tively absorb hydrogen sulfide and recycling
of the effluent from the amine regenerator
to the inlet of the sulfur plant: Scot (Shell),
Elfapur (Elf Aquitaine) processes;
2) incorporation of an additional treatment
stage based on the catalytic Claus reaction
at low temperature (135 –150 ◦ C). The sul-
fur obtained remains trapped in the cata-
lyst and must be desorbed during a subse-
quent regeneration cycle: Sulfreen process
(Elf Aquitaine, Lurgi).
When the sulfur concentration is not too high,
washing the residual gas with chemically regen-
erable substances (as used in the LO-CAT pro-
cess) or with nonregenerable substances (e.g.,
sodium hypochlorite or sodium hydroxide) may
be more economical. Alternatively, the gases
may be treated by conventional adsorption (ac-
tivated carbon, metal oxides) or by adsorption

combined with biological regeneration [133].
The latter involves immobilizing bacteria on an
adsorbent to incite autopurification.
6. Analysis
There is no universal method of analysis for the
determination of hydrogen sulfide. Each method
depends not only on the medium in which the
quantity of hydrogen sulfide is to be deter-
mined (gaseous mixtures, liquid effluents) and
on the concentration of hydrogen sulfide in the
medium, but also on the nature of the desired
results.
The following examples are by no means ex-
haustive:
1) Determination in a sour gas by iodometry
[134], gas-liquid chromatography [135], and
argentimetry [136].
2) Determination in a commercial purified gas
by iodometry [137], polarography [138].
3) Determination in a fuel gas with lead acetate
[139].
4) Determination in the atmosphere [140].
7. Storage, Handling, and
Transportation
7.1. Storage and Handling
The most important problem in storing hydro-
gen sulfide is the material used for the storage
vessel.
Figure 4. Storage tank for liquid hydrogen sulfide equipped with safety devices
a) Storage tank; b) Safety valve; c) Relief valve; d) Blowdown tank; e) Block valve
Hydrogen Sulfide 11
Anhydrous hydrogen sulfide does not react
with steel at ambient temperature. However, in
an aqueous medium, it attacks iron to form iron
sulfide and hydrogen. The hydrogen can diffuse
into the metal where it is trapped in cavities,
causing blistering and embrittlement. The de-
gree of corrosion depends on the grade of steel
and is very limited with stainless steel. Fer-
ritic steels are much more sensitive to hydro-
gen embrittlement than austenitic steels. Embrit-
tlement can be important with highly resistant
steel. High-strength steel exposed to hydrogen
sulfide is likely to crack under stress. Stress-
relieving treatment after welding reduces resid-
ual stresses. Thus, the steel used for hydrogen
sulfide storage vessels and pipes must be chosen
carefully.
Because of the problems associated with
transportation of hydrogen sulfide and its toxic-
ity, storage capacities are restricted; the product
must be used close to its production site when-
ever possible. Nevertheless, hydrogen sulfide is
transported by pipeline over several kilometers
in Europe and the United States. Liquid hydro-
gen sulfide is also transported in specially de-
signed containers. Particular attention should be
given to emptying devices (valves) and their pro-
tection during transportation.
Storage tanks for liquid hydrogen sulfide
should be equipped with a safety device system
as shown in Figure 4. The blowdown tank is a
reserve storage capacity that enables rapid emp-
tying of liquid hydrogen sulfide in case of acci-
dent. It can be operated by an “emergency stop”
12 Hydrogen Sulfide
switch that automatically actuates the safety
valves. The gas phase in the storage tank is con-
nected directly to a flare by means of a relief
valve that can be bypassed via a safety valve.
7.2. Transportation
Transportation of liquid hydrogen sulfide by
air (IATA) is strictly prohibited. Transporta-
tion in bulk containers by sea is also prohibited
(IMDG). Small containers of < 1 t or small bot-
tles are generally used for road and rail trans-
portation. The transport of hydrogen sulfide in
Europe by rail has been abandoned; it is used
instead at the site of production. In the United
States, liquid hydrogen sulfide is transported in
specially designed tankers. Hazard classifica-
tions for hydrogen sulfide are as follows:
IMDG Code, class 2.1, label 2.1 + 6
RID/ADR, class 2, label 3 + 6.1
United States, CFR 49 : 172.02, flammable
UN no. 1053
The sulfur content of crude oil can result in the
production of hydrogen sulfide, representing a
risk for field personnel and for the transporter.
In sea transport, maintenance of storage tanks
under a blanket of inert gas does not eliminate
the risk of explosion. The risk of sulfur forma-
tion in the tankers should also be kept in mind.
Wearing gas masks in the pump room of tankers
is compulsory when the hydrogen sulfide con-
tent exceeds 10 ppm. The same measures should
be taken for personnel connecting up the loading
arms.
8. Uses
Production of Sulfur. Natural gas con-
taining hydrogen sulfide is currently one of
the most important sources of elemental sulfur.
For commercial use, sour natural gas must first
be sweetened to remove the acid compounds,
mainly hydrogen sulfide and carbon dioxide.
The resulting acid gas is used for the produc-
tion of sulfur by the Claus process [141, 142]:
2 H2 S + SO2 −→ 3/n Sn + 2 H2 O
For further details, see → Sulfur. Sulfur is
marketed in liquid or solid form. In the liquid
state, itmust first be degassed to remove all traces
of hydrogen sulfide. The sales specification of
sulfur in Europe requires a hydrogen sulfide con-
tent of < 10 ppm. This value is obtained by de-
gassing liquid sulfur into sulfur pits with ammo-
nia before dispatch.
Production of Metal Sulfides. The reaction
of hydrogen sulfide with soda or lime solutions
results in formation of the corresponding sul-
fides. Sodium sulfide is generally marketed in
a solid hydrated form. Acid sulfides or hydro-
sulfides [e.g., NaSH or Ca(SH)2 ] are marketed
either as crystals or as concentrated aqueous so-
lutions.
Sodium sulfide is used as a lignin demethyla-
tion agent in the Kraft process for manufacturing
paper pulp. Mixtures of sodium and calcium sul-
fides are powerful reducing agents of disulfide
bonds. They attack the cystine molecules of ker-
atin and are used in the leather industry for chem-
ical depilation of animal hides (→ Leather).
Flotation. Preliminary treatment with hy-
drogen sulfide is sometimes required for the pu-
rification of ore by selective flotation. Certain
mineral species in the ore are sulfurized, which
considerably increases the selectivity of flotation
collectors and facilitates separation of valuable
metallic species. This process is used in the pu-
rification of nickel and manganese.
Catalyst Activation and Poisoning. Certain
catalysts reach their maximum activity only
when their constituent metals are in the form of
sulfides (e.g., the hydrotreating catalysts used in
oil refining). Preliminary treatment with hydro-
gen sulfide is sometimes necessary.
However, hydrogen sulfide also poisons some
hydrogenation and reforming catalysts with a
platinum or palladium base. This can be used to
advantage to reduce initial catalyst activity and
limit coking (i.e., formation of carbonaceous de-
posits). Gasoline reforming catalysts, for exam-
ple, are always selectively poisoned with a small
quantity of hydrogen sulfide prior to use.
Treatment of Metallic Surfaces. Superfi-
cial sulfurization of metals with hydrogen sul-
fide allows modification of their physical or
chemical properties. Hydrogen sulfide is used
 Hydrogen Sulfide 13

for passivation of the walls of reactors operating
at high temperature in petrochemical operations
such as steam cracking and hydrodealkylation.
This treatment prevents desirable secondary re-
actions.
Another use of hydrogen sulfide is the forma-
tion of a layer of sulfide on the surface of steel
wires or plates that are to be coated with paint
or plastic. This treatment protects the coating or
improves the action of adhesive primers.
Analytical Chemistry. Hydrogen sulfide
has long been used to selectively precipitate
group 1, 2, and 3 metal cations. The metals are
qualitatively identified by the color of the sul-
fides formed.
Production of Thioorganic Compounds.
The most common industrial application of hy-
drogen sulfide is the production of organic com-
pounds with a thiol function (i.e., RSH, where
R is a hydrocarbon group) or a thiol group and
another function (e.g., alcohol, acid).
Table 6 shows the most important thioor-
ganic compounds (in terms of tonnage) pro-
duced from hydrogen sulfide in the United States
and Europe. Other important thioorganic com-
pounds derived from hydrogen sulfide are tetra-
hydrothiophene and mixtures of thiols or sul-
fides (often containing tert-butanethiol or di-
ethylsulfide). These are used as strong-smelling
additives for gas odorization.
Long-chain thiols also represent a range of
useful products, especially n-dodecanethiol and
Table 6. Thioorganic compounds produced from hydrogen sulfide
n-octanethiol. These are used as intermediates
in the synthesis of plant protection agents and as
polymerization agents.
9. Economic Aspects
Total world sulfur production was 56.6×106 t in
1986, compared with 51.7×106 t in 1977. Sul-
fur recovered from the treatment of sour gas (i.e.,
from hydrogen sulfide) accounts for 25 % of this
(ca. 14×106 t in 1986, 11.8×106 t in 1977).
The primary producing countries are listed
below, with their 1986 production figures in 106
metric tonnes per year:
Western Canada 5.2
United States 2.2
former Soviet Union 1.8
Saudi Arabia 1.2
Federal Republic of Germany 1.0
France 0.9
Recovery of sulfur from hydrogen sulfide will
continue to expand worldwide. The former So-
viet Union is expected to become the main pro-
ducer of sulfur from sour gas due to development
of the Astrakhan gas field (hydrogen sulfide con-
tent: 26 %). Astrakhan I started producing early
in 1987 and Astrakhan II is due to be put on
stream early in 1989.
Canada should continue to be a major pro-
ducer because of the development of the Car-
oline gas field in Alberta. The Middle East is
also expected to play a major role, with substan-
tial tonnage from Iran and Qatar. Production in

Product Formula Raw materials Annual world Producers Uses

production, t

Methanethiol (methyl CH3 SH methanol + H2 S 70 000 Pennwalt Rhone Poulenc base product for
mercaptan) [74-93-1] Elf Aquitaine manufacture of methionine,
dimethyl disulfide, and plant
protection products
Thioglycolic acid HSCH2 COOH chloroacetic 15 000 Elf Aquitaine Bock Merck manufacture of esters,
(mercapto acetic acid) acid + NaSH poly(vinyl chloride)
[68-11-1] stabilizers, and cosmetics
tert-Dodecanethiol C12 H25 SH tetrapropylene or 13 000 Pennwalt Phillips Elf chain-transfer agent in
[25103-58-6] (mercaptan mixture) triisobutylene Aquitaine Bayer radical polymerization;
+ H2 S manufacture of polysulfides
Ethanethiol (ethyl CH3 CH2 SH ethylene + H2 S 12 000 Phillips Pennwalt Elf manufacture of plant
mercaptan) [75-08-1] Aquitaine protection products
Mercaptoethanol HSCH2 CH2 OH ethylene 8000 Phillips Morton Thiokol soluble chain-transfer agent
[60-24-2] oxide + H2 S Alcolac BASF Elf for polymerization in
Aquitaine aqueous media; manufacture
of plant protection products;
poly(vinyl chloride)
stabilizers
14 Hydrogen Sulfide
the United States and the Federal Republic of
Germany is likely to stabilize at (2.3 – 2.5)×106
and 1.0×106 t, respectively. In France, produc-
tion will decline gradually due to depletion of the
Lacq gas field (hydrogen sulfide content: 15 %).
Sulfur recovered from hydrogen sulfide has been
the major source of accumulated stocks, which
are used to swing production to balance the mar-
ket when needed. At the end of 1987, the stock
of sulfur from sour gas, in tonnes, was
Western Canada 6.7×106
France 2.2×106
Saudi Arabia 1.5×106
10. Toxicology
Hydrogen sulfide is a toxic gas. Human poison-
ing is related to its acute toxicity. Chronic toxic-
ity has not been demonstrated in people exposed
to concentrations lower than those that induce
acute or subacute poisoning [143, 144].
Animal Data. Much toxicological data has
been published on animals exposed to massive
dosages of hydrogen sulfide. These data confirm
observations made on humans, but information
on chronic or long-term effects remains limited.
The acute toxicity of hydrogen sulfide is com-
parable in animals and humans.
Chronic toxicity studies have been carried out
on rabbits, rats, and mice that inhaled daily doses
of hydrogen sulfide at concentrations in air up
to 80 ppm over three to five consecutive months
[143–145]. Despite numerous detailed exami-
nations, including neurological tests, no specific
chronic effects were found. Mutagenicity testing
has shown a lack of genotoxic effects on bacteria
in vitro [146] and on rats in vivo [147].
Metabolism. Lack of a chronic toxic effect
of hydrogen sulfide in animals and most prob-
ably in humans may be explained by the fact
that it does not build up in the body. Hydro-
gen sulfide is metabolized rapidly, mainly into
less toxic, oxidized metabolites that are quickly
eliminated via the kidneys.
Hydrogen sulfide is readily absorbed by the
pulmonary tract, which at the same time repre-
sents an excretory pathway for untransformed
hydrogen sulfide. Hydrogen sulfide that is not
eliminated by the respiratory tract binds to cir-
culating proteins in the blood and is distributed
rapidly to all tissues, including the nervous sys-
tem. It is subsequently metabolized and detoxi-
fied as shown in Figure 5 [144].
Hydrogen sulfide is detoxified by oxidation
to sulfate (the primary reaction) or by methy-
lation (not yet confirmed). The toxicity of hy-
drogen sulfide is due to its reaction with protein
constituents of essential enzymes or metallopro-
teins, especially the iron-containing metallopro-
teins of the respiratory cell chain.
During chronic exposure, the amount of hy-
drogen sulfide entering the organism is so small
that it is neutralized by the above detoxification
reactions. However, during massive acute expo-
sure, the detoxification capacity is exceeded; the
reaction of hydrogen sulfide with proteins can
thus lead to poisoning.
Human Risk and Pathology. The acute
toxic risk of hydrogen sulfide and the pathology
observed in human poisoning are summarized
in Tables 7 and 8 (see next pages) [148, 149].
11. Occupational Health and Safety
The main safety precautions listed here result
from thirty years’ experience at Lacq [150] and
from the literature. Hydrogen sulfide has several
properties that make it a highly dangerous gas:
1) high toxicity;
2) flammability: it ignites spontaneously at
260 ◦ C (compared to 537 ◦ C for methane)
and forms an explosive mixture with air when
the hydrogen sulfide content is 4 – 44 vol %
(compared to 5 – 15 % for methane) [151];
3) density: being heavier than air, it accumulates
in low-lying areas of installations;
4) ability to embrittle steel (Section 7.1); and
5) insidiousness: persons intoxicated with hy-
drogen sulfide no longer discern its char-
acteristic smell of rotten eggs before the
dangerous threshold concentration (50 –
100 ppm) is reached.
General Precautions. Areas in which hy-
drogen sulfide is produced, utilized, and stored
must be isolated and delimited.
 Hydrogen Sulfide 15

Figure 5. Metabolism of hydrogen sulfide in humans and animals

Abbreviations: GSH, reduced glutathione; GSSG, glutathione; NADP+ , oxidized nicotinamide – adenine dinucleotide phos-
phate; NADPH, reduced nicotinamide – adenine dinucleotide phosphate
Table 7. Acute toxic effects of hydrogen sulfide in humans

Concentration Duration of exposure

of H2 S, ppm 15 min 15 min – 1 h 1–4h 4–8h Comments

10 eye irritation maximum tolerable

concentration for
prolonged exposure
50 – 100 loss of olfactory eye irritation eye and bronchial danger in case of working conditions
perception irritation continuous exposure necessitate protective
measures
150 – 250 loss of olfactory eye and bronchial serious respiratory distress and asthenia working conditions
perception irritation necessitate protective
measures
300 – 400 loss of olfactory severe respiratory pulmonary edema and risk of death risk of death if no
perception, eye and distress, acute asthenia appropriate measures
bronchial irritation, taken
asthenia
500 – 1000 loss of consciousness, risk of pulmonary risk of death if no
respiratory distress edema and death appropriate measures
taken
>1000 immediate loss of consciousness and respiratory distress risk of death if no
appropriate measures
taken
16 Hydrogen Sulfide
Table 8. Human pathology in hydrogen sulfide poisoning

Degree of poisoning

Weak Intermediate Severe

Early signs equilibrium problems
Immediate effects loss of consciousness for a
few seconds; rapid regain
of consciousness
Long-term effects recovery
Emergency treatment evacuation from danger
area; administration of
oxygen
warning signs of short duration
sudden loss of consciousness lasting a
few minutes, followed by headache,
respiratory distress, gastric pain,
diarrhea (following day)
recovery or bronchial fragility; asthenia
evacuation from danger area;
administration of oxygen; respiratory
assistance if needed; intravenous
vitamin C; medical observation for
several hours
immediate coma
prolonged deep coma with cyanosis,
respiratory distress, occasionally cardiac
distress; regain of consciousness is difficult
and agitated
bronchopulmonary sequelae
evacuation from danger area; administration
of oxygen; respiratory assistance if needed;
intravenous vitamin C; cardiotonics and
neurosedatives if needed; hospitalization

Only authorized personnel should have ac-
cess to these areas, i.e., personnel having un-
dergone medical tests to detect ocular, respira-
tory, or cardiovascular disease, and signs of al-
coholism. All authorized personnel should un-
dergo special training to deal with the inherent
risks of working in a hydrogen sulfide environ-
ment. They should be equipped with gas masks
and undergo regular training in their use.
The usual precautions should be taken in all
areas where an explosive atmosphere is likely to
be found (sources of flame, electric equipment,
ect.). “No smoking” signs should be displayed
and strictly observed.
Only piping, apparatus, and containers made
from special steel or alloys should be used; the
grade should be suitable for the conditions of
use. The use of copper and its alloys should be
prohibited.
Storage and Transportation (see also
Chap. 7). The storage area should be well ven-
tilated and, if possible, equipped with a water
cooling system in case of fire.
Hydrogen sulfide containers should be stored
separately from incompatible products (oxi-
dizing agents, inflammable products, powerful
acids, etc.) and away from direct sunlight. A spe-
cial parking area should be provided for vehicles
transporting hydrogen sulfide.
Each container should carry a label indicat-
ing that the contents are highly flammable and
toxic. The label should also list the special risks
and include cautionary advice.
– Modified EEC directive 67/548
– EEC No. 016-001-00-4
– Danger symbols: highly flammable and toxic
– Nature of special risks attaching to danger-
ous substances:
– R.13 Extremely flammable liquefied gas
– R.26 Very toxic by inhalation
– Safety advice concercing dangerous chemi-
cal substances:
– S7/9 Keep container tightly closed in a well-
ventilated place
– S.25 Avoid contact with eyes
– S.45 In case of accident or clinical symp-
toms medical advice should be sought im-
mediately.
Installation and Handling. Apparatus
should be properly closed and tightly sealed.
Waste gas from discharge and pressure valves
should be sent to a flare to avoid emission of
hydrogen sulfide into the atmosphere. This flare
can burn off hydrogen sulfide in dangerous sit-
uations, eliminate air from the discharge piping
before discharge, and purge hydrogen sulfide
after discharge.
Emissions produced under normal conditions
should be recovered at source, and low-lying ar-
eas should be ventilated thoroughly.
Personnel working on apparatus containing
(or having contained) hydrogen sulfide should
be equipped with insulated gas masks and wear
protective clothing, goggles, and long gloves.
No one should be allowed to work alone in a
zone where hydrogen sulfide is handled.
Frequent checks of the atmosphere should
be made. Hydrogen sulfide detection devices
should be installed (see below); smell should not
be relied on for detection.
Programmed preventive maintenance should
be carried out in all hydrogen sulfide installa-
tions, particularly on sealing devices. A fire per-

mit should be obtained to use tools producing
flame, heat, or sparks.
Pyrophoric sulfides should be kept constantly
humid. They should be extracted and incinerated
as soon as possible. After maintenance work, all
traces of humidity should be eliminated.
Emergency Measures. A plan should be
drawn up for evacuation of personnel if the con-
centration of hydrogen sulfide exceeds the norm.
Gas masks must be put on as soon as an alarm
is given. Appropriate measures to be taken in
the case of leakage, fire, or accident should be
displayed prominently in the installation.
In the event of leakage without fire, the leak
should be isolated and sprayed with water. Un-
controllable gas should be sent to a flare pit, and
all sources of heat should be removed or shut
down.
If the leaking material has caught fire, the fire
should be put out only when it presents an espe-
cially serious threat to the environment or when
the leak can be rapidly controlled.
In the event of accidental gassing, lifesaving
intervention should be undertaken only after the
rescuer has donned an insulating gas mask. The
victim should be removed rapidly from the dan-
ger zone and should receive necessary first aid
and medical treatment as soon as possible.
Detection of Hydrogen Sulfide in Air. Per-
manent detection devices are necessary both in
and around the hydrogen sulfide storage area.
They should set off both audible and visual
alarms as soon as the concentration of hydrogen
sulfide in air exceeds 10 ppm. Several models are
available. An electrochemical sensor functions
on the gas diffusion principle: hydrogen sulfide
in air reacts chemically with an electrolyte (Sul-
fytron 4010, Draeger, FRG). In a semiconduc-
tor detector, air containing hydrogen sulfide dif-
fuses into the specially adapted semiconductor
and modifies its electrical resistance (General
Monitors, United States).
Portable detectors can be used for random
measurement of hydrogen sulfide concentra-
tion. An electrically or manually operated pump
sucks air through a reactive tube divided into
zones that become colored on contact with the
polluting gas (Draeger model 21/31 with specific
tube for hydrogen sulfide).
Hydrogen Sulfide 17
A simple but effective method of detecting a
leak is to use lead acetate paper. The paper be-
comes black in the presence of hydrogen sulfide
but gives no indication of concentration.
Exposure Limit Values [152]. The average
permitted concentration of hydrogen sulfide in
air varies from country to country:
TLV (TWA) 5 – 10 ppm (vol)
TLV (STEL) 10 – 15 ppm
12. References
1. E. Eriksson, Tellus 12 (1960) 63. H.
Roffmann, M. C. Bender: “H2 S: new sources
and potential impacts,” 68th Annual Meeting
of the Air Pollution Control Association , June
1975.
2. B. Bonsang, La Recherche 13 (1982) no. 137,
1132 – 1142.
3. D. Möller, Atmos. Environ. 18 (1984) no. 1,
29 – 39.
4. T. P. Kudryautseva, Deposited document –
VINITI 160, 7 – 15; Chem. Abstr. 94 (1981)
no. 22, 177 614 v.
5. A. I. Kravtsov, G. I. Voitov, Izv. Vyssh. Uchebn.
Zaved., Geol. Razved. 19 (1976) no. 4, 18 – 25;
Chem. Abstr. 85 (1976) no. 24, 179 877 w.
6. H. G. Spratt, Jr., M. D. Morgan, R. E. Good,
Appl. Environ. Microbiol. 53 (1987) no. 7,
1406 – 1411; Chem. Abstr. 107 (1987) no. 11,
95 855 x.
7. J. P. Hollis, Jr., Acta Phytopathol. Acad. Sci.
Hung. 20 (1985) no. 3 – 4, 321 – 326; Chem.
Abstr. 105 (1985) no. 21, 190 028 x.
8. I. Benda, Arch. Mikrobiol. 27 (1957)
337 – 374; Chem. Abstr. 53 (1959) 22 208 g.
9. R. Vàmos, M. Ando, Acta Univ. Acta Biol.,
Szeged. 5 (1959) 61 – 69; Chem. Abstr. 54
(1960) 17 770 f.
10. V. G. Popov, Gazov. Prom-st Ser.: Geol.
Razved. Gazov. Gazokondens. Mestorozhd.
1981, no. 10, 10 – 18; Chem. Abstr. 96 (1982)
no. 20, 165 091 a.
11. E. E. London, Geol. Nefti Gaza 1975, no. 8,
24 – 30; Chem. Abstr. 84 (1976) no. 6,
33 375 p.
12. C. A. Brook, Geol. Surv. Prof. Pap. U.S. 1141
(1981) 193 – 203.
13. R. J. Ferek, R. B. Chatfield, M. O. Andrae,
Nature (London) 320 (1986) 514 – 516.
14. E. R. Lippincott, R. V. Eck, M. O. Dayhoff, C.
Sagan, Astrophys. J. 147 (1967) 753 – 764;
Chem. Abstr. 67 (1967) 58 795 n.
18 Hydrogen Sulfide
15. K. Pasternak, Acta Hydrobiol. 12 (1970) no. 4,
377 – 390; Chem. Abstr. 75 (1971) no. 6,
40 097 s.
16. N. F. Voznaya, Lb. Tr. Leningr. Inzh. Stroit.
Inst. 25 (1957) 123 – 133; Chem. Abstr. 53
(1959) 22 637 i.
17. W. J. Mueller, GWF Gas Wasserfach 102
(1961) 986 – 989; Chem. Abstr. 56 (1962)
1303 e.
18. F. W. Brinkmann, SPE/DDE Low Permeability
Symposium, Denver, CO, May 27 – 29, 1981.
19. H. Luebben, Erdoel Erdgas Z. 98 (1982) July,
225 – 231.
20. K. Le Tran, J. Connan, B. Van der Weide, Adv.
Org. Geochem. Proc. Int. Meet., 6th 1973,
761 – 789; Chem. Abstr. 85 (1976) no. 14,
96 664 y.
21. J. L. Rumeau, C. Sourisse, Adv. Org. Geochem.
Proc. Int. Meet. 7th 1975, 525 – 538; Chem.
Abstr. 89 (1978) no. 20, 165 749 b.
22. L. I. Rovnin, N. V. Mizinov, N. I. Voronin,
Geol. Nefti Gaza 1977, no. 10, 41 – 43; Chem.
Abstr. 88 (1978) no. 2, 9336 p.
23. N. V. Mezinov, A. S. Zinger, V. G. Grushevoi,
N. I. Voronin, Geol. Nefti Gaza 1979, no. 6,
1 – 5; Chem. Abstr. 91 (1979) no. 18,
142 832 a.
24. R. G. Pankina, V. L. Mekhtieva, S. P.
Maksimov, Geol. Nefti Gaza 1983, no. 4,
45 – 51; Chem. Abstr. 99 (1983) no. 4,
25 095 s.
25. R. S. Aliev, Geol. Stroenie i Neftegezonosn.
Nizhn. Povolzh’ya Saratov 1984, 86 – 93;
Chem. Abstr. 102 (1985) no. 12, 97 906 k.
26. I. A. Lagunova, A. Ya. Vvedenskaya, Geol.
Nefti Gaza 1986, no. 12, 20 – 23; Chem. Abstr.
106 (1987) no. 10, 69 776 g.
27. S. Kh. Aigistova et al., Gazov. Prom-st 1983,
no. 8, 30 – 31; Chem. Abstr. 99 (1983) no. 20,
160 948 n.
28. N. B. Valitov, Sov. Geol. 1986, no. 3, 93 – 100;
Chem. Abstr. 104 (1986) no. 24, 209 594 g.
29. N. B. Valitov, Geol Nefti Gaza 1986, no. 5,
23 – 28; Chem. Abstr. 105 (1986) no. 6,
45 921 d.
30. N. I. Voronin, E. I. Ben’ko, V. F. Laktynshina,
L. A. Anisimov, Geol. Nefti Gaza 1986, no. 7,
58 – 60, Chem. Abstr. 105 (1986) no. 18,
155 735 d.
31. V. F. Perepelichenko, S. V. Eremeeva, A. F.
Il’in, Gazov. Prom-st 1986, no. 12, 30 – 31;
Chem. Abstr. 106 (1987) no. 10, 69 839 e.
32. V. I. Lapshin, Yu. Yu. Kruglov, A. P. Zheltov,
Geol. Nefti Gaza 1987, no. 2, 41 – 43; Chem.
Abstr. 107 (1987) no. 4, 25 703 x.
33. V. I. Danilov, V. P. Pronchnk, A.
Ya. Litovchenko, Tr. Gos. Inst. Proekt. Issled.
Rab. Neft. Prom-sti 1973, no. 19, 154 – 157;
Chem. Abstr. 83 (1975) no. 26, 208 160 y.
34. Kh. I. Ismailova, Razrab. Gazov. Mestorozhd.
Transp. Gaza 1974, no. 3, 40 – 46; Chem.
Abstr. 84 (1976) no. 10, 62 087 b.
35. Yu. A. Zhurov, Gazov. Prom-st 1976, no. 9,
36 – 39; Chem. Abstr. 86 (1977) no. 20,
142 511 s.
36. P. M. Lomako, Gazov. Prom-st 1979, no. 11,
33 – 36; Chem. Abstr. 92 (1980) no. 18,
149 624 p.
37. V. N. Kortsenshtein, Geol. Nefti Gaza 1980,
no. 4, 56 – 59; Chem. Abstr. 93 (1980) no. 6,
49 924 p.
38. A. I. Polivanova, G. S. Akopova, O. F.
Tuzlukova, Geol. Priro. Gaza Polezn.
Kompon. Prir. Gazov Usloviya Form. Resur.
(1979) M 87 – 95; Chem. Abstr. 93 (1980)
no. 18, 170 614 r.
39. V. M. Kir’yashkim, V. S. Goncharov, I. A.
Borzenkov, A. K. Aleshina, Gazov. Prom-st,
Ser. Geol. Razved. Gazov. Gazokondens.
Mestorozhd. 1980, no. 6, 28 – 34; Chem. Abstr.
93 (1980) no. 16, 152 640 s.
40. G. P. Vdovykin, Gazov. Prom-st 1980, no. 10,
29 – 31; Chem. Abstr. 94 (1981) no. 14,
105 998 a.
41. A. G. Burmistrov, B. V. Surkov, Probl. Dobychi
Gaza (Primera Razrab. Orenb. Gazokondens.
Mestorozhd.) 1979, 159 – 163;O. F.
Khudyakov (ed.), Vses. Nauchno-Issled. Inst.
Prir. Gazov. (Moscow) USSR; Chem. Abstr. 95
(1981) no. 18, 153 270 z.
42. F. R. Hertweck, Jr., D. D. Fox, Inf. Circul. –
U.S. Bur. Mines, IC 8993, 1984, p. 130; Chem.
Abstr. 102 (1985) no. 10, 81 281 s.
43. J. B. Hyne, World Oil 194/195 (1982) 75 – 78.
44. A. V. Kormukhina, A. I. Mishukovskaya, N. V.
Verts, Korroz. Zashch. Neftegazov. Prom-sti
1976, no. 3, 5 – 7; Chem. Abstr. 85 (1976)
no. 14, 96 679 g.
45. J. G. Connell, R. M. Shapiro, Proc. Annu.
Southwest. Pet. Short Course 23 (76) 1 – 8;
Chem. Abstr. 86 (1977) no. 14, 92 779 n.
46. T. L. Pugh, Elastomerics 116 (1984) no. 6,
35 – 39; Chem. Abstr. 101 (1984) no. 16,
133 459 g.
47. K. A. Shishin et al., Neft. Khoz. 1977, no. 12,
51 – 52; Chem. Abstr. 88 (1978) no. 16,
107 600 q.
48. R. M. Khasaev, L. V. Konysheva, L. S.
Leon’teva, R. M. Abdullaev, SU 825 579;
Chem. Abstr. 95 (1981) no. 8, 64 940 a.

49. Yu. N. Sokolov, B. V. Yakubenko, T.
Ya. Deputatova, Geol. Str. Neftegazonosn.
Nizh. Povolzh’ya, Saratov 1981, no. 4,
106 – 109; Chem. Abstr. 97 (1982) no. 26,
219 164 t.
50. A. P. Krezub, V. V. Kainash, Burenie (Moskva)
1981, no. 12, 7 – 9; Chem. Abstr. 98 (1983)
no. 16, 128 805 g.
51. B. A. Andreson, G. P. Bochkarev, E. I.
Ogarkova, R. Kh. Khazipov, SU 1127893 A1,
1984; Chem. Abstr. 102 (1985) no. 14,
116 396 n.
52. J. Raczkowski, A. Dudek, A. Hejnar, Nafta
(Katowice Pol.) 42 (1986) no. 4, 106 – 110;
Chem. Abstr. 106 (1987) no. 6, 35 660 z.
53. Oil Gas J. 82 (1984) 76.
54. Ullman, 4th ed., 21, 169.
55. R. Goodwin: “Hydrogen sulfide provisional
thermophysical properties from 188 to 700 K
at pressures to 75 MPa,” NBS Report no.
NBSIR 83 – 1694 (1983) 175.
56. “L’Air liquide,” Encyclopédie des gaz,
Elsevier, Amsterdam 1976, p. 935.
57. J. W. Hennel, K. Kikrynicki, Acta Phys. Pol.
18 (1959) 523 – 526.
58. Gmelin, 4 a/b, p. 108.
59. P. Pascal, “Nouveau traité de chimie minérale,”
vol. XIII, Masson Ed., Paris 1956.
60. R. C. Weast (ed.): Handbook of Chemistry and
Physics, 53rd ed., CRC Press, Cleveland,
1972/73, p. 222.
61. C. Huiszoom, A. Dynanus, Physica 31 (1965)
1049 – 1052.
62. A. Barua, A. Manna, P. Mukhopadhyay, J.
Chem. Phys. 49 (1968) 2422 – 2425.
63. L’Air liquide: Encyclopédie des gaz, Elsevier,
Amsterdam 1976, p. 936.
64. M. Burgess, R. Germann. AIChE J 15 (1969)
272 –275.
65. J. D. Zel’venski, A. V. Strunina, Gazov.
Promyšlennost 6 (1960) no. 1, 42 – 47.
66. E. Isaacs, F. Otto, A. Mather, Can. J. Chem.
Eng. 55 (1977) 751 – 752.
67. A. Kohl, F. Riesenfeld: Gas Purification, 3rd
ed., Gulf Pub. Co., Houston 1976, p. 783.
68. M. King, H. Al Najjar, Chem. Eng. Sci. 32
(1977) 1241 – 1246.
69. G. Charlot: Les réactions chimiques en
solutions aqueuses, Masson, Paris 1983.
70. Société Nationale des Pétroles d’Aquitaine,
FR 1435 788, 1964 (J. Solinhac). Shell, FR
2 134 677, 1972 (R. S. Anderson). Exxon, EP
45 636, 1981 (L. T. Howard). Elf Aquitaine,
FR 2 601 350, 1986 (R. Voirin, A. Pepy).
Hydrogen Sulfide 19
71. J. C. Jorczak, E. M. Fettes, Ind. Eng. Chem. 43
(1951) 324.
72. D. Martin, H. G. Hauthal: Dimethylsulfoxide,
Van Nostrand Reinhold Ltd., Wokingham,
England 1971.
73. P. Sabatier, A. Mailhes, C. R. Acad. Sci. 150
(1910) 1217.
74. H. O. Folkins, E. L. Miller, Ind. Eng. Chem.
Prod. Res. Dev. 1 (1962) 271.
75. Pure Oil, US 2 820 062, 1958 (H. O. Folkins,
E. L. Miller).
76. Pennsalt Chemicals Corp., US 3 035 097, 1962
(T. E. Deger, B. Buchholz, R. H. Goshorn).
77. Y. K. Yurev, M. N. Prokina, J. Gen. Chem.
USSR (Engl. Transl.) 7 (1937) 1868. Pennsalt,
GB 913 584, 1960 (T. E. Deger, R. H. Goshorn,
B. Buchholz).
78. Pennsalt, US 3 197 483, 1961 (T. E. Deger,
R. H. Goshorn, B. Buchholz).
79. Société Nationale des Pétroles d’Aquitaine,
FR 1 417 154, 1964 (P. Vannel). Shell,
EP 122 654, 1980 (F. H. Elliott).
80. Société Nationale Elf Aquitaine (Production),
Demande brevet français 88 01879, 1988 (E.
Arretz).
81. Pennsalt, US 2 951 875, 1960 (B. Loev, R. H.
Goshorn).
82. Pennsalt, US 4 102 931, 1978 (B. Buchholz).
83. Société Nationale Elf Aquitaine (Production),
EP 101 356, 1982 (E. Arretz, A. Mirassou, C.
Landoussy, P. Augé).
84. Société Nationale Elf Aquitaine (Production),
FR 2 330 674, 1975 (E. Arretz, A. Pfister).
Phillips, US 3 963 785, 1975 (D. H. Kubicek).
Seitetsu, JP 79 132 551, 1978 (M. Kawamura,
M. Hatta, M. Nakamura, N. Koune, Y.
Iwasaki).
85. R. Dahlbom, Acta. Chem. Scand. 5 (1951)
690.
86. Toyo Chemical Industry, JP 81 147 763, 1980.
Phillips, EP 208 323, 1985 (J. S. Roberts).
87. Henkel, BE 731 879, 1968. Henkel, BE
731 881, 1968.
88. Société Nationale des Pétroles d’Aquitaine,
FR 1 359 678, 1963 (B. Pflugfelder, P. Vannel).
Toyo Kasei, JP 81 110 667, 1980.
89. Pennwalt, US 4 281 202, 1980 (B. Buchholz,
C. B. Welsh, H. C. Henry).
90. Société Nationale Elf Aquitaine (Production),
EP 274 934, 1988 (E. Arretz, P. Augé, A.
Mirassou, C. Landoussy).
91. GB 585 655, 1947 (F. N. Woodward). Röhm
and Haas, DE 1 221 217, 1965 (H. Walter, W.
Gruber). Société Nationale Elf Aquitaine
(Production), FR 2 483 917, 1980 (E. Arretz).
20 Hydrogen Sulfide
Société Nationale Elf Aquitaine (Production),
US 4507505, 1985 (E. Arretz).
92. T. L. Cairns et al., J. Am Chem. Soc. 74 (1952)
3982. R. Mayer, Angew. Chem. Int. Ed. Engl.
2 (1963) 277.
93. Société Nationale des Pétroles d’Aquitaine,
FR 1 578 891, 1968 (Y. Labat).
94. B. Magnusson, Acta Chem. Scand. 13 (1959)
1031.
95. Du Pont, US 2 551 813, 1949 (P. S. Pinkney).
96. Stauffer, US 3 689 568, 1969 (R. J. Eletto).
97. Phillips, US 3 223 738, 1965 (D. L. Crain).
98. Phillips, US 3 050 452, 1960 (R. P. Louthan).
99. Pennwalt, US 4 140 604, 1978 (D. A. Dimmig).
100. Société Nationale Elf Aquitaine (Production),
EP 5 400, 1979 (J. Ollivier, G. Souloumiac, J.
Suberlucq).
101. Société Nationale Elf Aquitaine (Production),
EP 60754, 1981 (E. Arretz, C. Landoussy, A.
Mirassou, J. Ollivier).
102. Akzo, US 3 927 085, 1974 (A. G. Zenkel, A.
Toth, H. Nagerlein, G. Meyer).
103. Dow Corning Corp., US 4 082 790, 1975 (J. L.
Speier). Stauffer, US 4 740 623, 1988 (J. B.
Heather). Société Nationale Elf Aquitaine
(Production) Demande brevet françois
8 804 693, 1988 (Y. Labat).
104. J. E. Bittel, J. L. Speier, J. Org. Chem. 43
(1978) 1687.
105. P. Noble, D. S. Tarbell, Org. Synth. 4 (1963)
924.
106. Société Nationale des Pétroles d’Aquitaine,
US 3 781 366, 1973 (E. Arretz, A. Donche, A.
Pfister).
107. Wakunaga, Yakuhin Kabushiki Kaisha, US
4 212 826, 1979 (H. Yamaguchi).
108. Sankyo Organic Chemicals Cy, US 3 840 586,
1974 (T. Chiba et al.). Degussa, US 4 490 307,
1984 (H. Kuenk).
109. S. Dushman, J. Am. Chem. Soc. 43 (1921) 427.
110. C. M. Thacker, E. Miller, Ind. Eng. Chem. 36
(1944) 182.
111. K. T. Zimmermann, Angew. Chem. 74 (1962)
151.
112. K. Pohl, G. Rebentish, Chem. Tech. (Leipzig)
19 (1967) 503.
113. L. T. Filimonova et al., Neftepererab.
Neftekhim. (Moscow) 1 (1976) 42; Chem.
Abstr. 85 (1976) no. 96, 155.
114. Girdler, Company publication, “Compact
Package H2S generators” Licensee: Chemical
and Industrial Corp., Girdler, Zurich,
Switzerland.
115. R. F. Bacon, E. S. Boe, Ind. Eng. Chem. 37
(1945) 469.
116. R. C. Forney, J. M. Smith, Ind. Eng. Chem. 43
(1951) 1841.
117. W. A. Bryce, J. Chem. Soc. 4 (1949) 3379.
118. F. Seel, Angew. Chem. 68 (1956) 789.
119. F. Gomez Vigide et al., Inf. Quim. Anal.
(Madrid) 13 (1959) 61.
120. M. A. Korshunov et al., Zavod. Lab. 34 (1968)
285; Chem. Abstr. 69 (1968) 7944.
121. E. N. Konovalov et al., Gasov. Prom. 6 (1972)
47; Chem. Abstr. 77 (1972) 77 435.
122. R. M. Levit, Zavod. Lab. 27 (1961) 352;
Chem. Abstr. 55 (1961) 25 387.
123. A. L. Kohl, F. C. Riesenfeld: Gas Purification,
4th ed., Gulf Publishing Co., Houston 1985.
124. J. I. Lee, F. D. Otto, A. E. Mather, J. Chem.
Eng. Data 17 (1972) 465.
125. R. N. Maddox: Gas and Liquid Sweetening,
3rd ed., Campbell Petroleum Series, Norman
(OK) 1982.
126. Proceedings of Gas Conditioning
Conferences, University of Oklahoma –
Norman (OK) (since 1981).
127. C. Blanc, J. Elgue, Hydrocarbon Process. 60
(1981) no. 8, 111 – 116.
128. Elf Aquitaine proprietary documents,
Licensee: Elf Aquitaine.
129. J. M. Ryan, A. S. Holmes, US 4 293 322, 1981;
Chem. Abstr. 96 (1982) P 37667.
130. Dangerous Properties of Industrial Materials,
N-I-SAX-1979. Van Nostrand, New York
1979.
131. Oil and Hazardous Materials. Technical
Assistance Data System U.S., Environmental
Protection Agency, Oil and special materials
Control Division, Washington 1980.
132. Sulfure d’hydrogène – Collection Enviroguide.
Environnement Canada-Ottawa 1984.
133. Société SAPS – Anticorrosion – Brevet
Français no. 2 591 122.
134. Determination of H2S in Sour Gas by
Iodometry, SNEA (P) Internal Method no. 52.
135. Determination of H2S in Sour Gas by
Gas-liquid Chromatography, SNEA (P)
Internal Method no. 80.
136. Determination of H2S in Sour Gas by
Argentimetry, SNEA (P) Internal Method
no. 53.
137. Determination of H2 S in Natural Gas by a
Cadmium-Sulfate-Iodometric Method, ASTM
D 2385.
138. Determination of H2S in a Purified
Commercial Natural Gas by a Polarographic
Method, SNEA (P) Internal Method no. 134.

139. Determination of H2 S in a Fuel Gas by a Lead
Acetate Reaction Rate Method, ASTM D
4085.
140. Determination of H2 S in the Atmosphere,
ASTM D 4323.
141. P. Grancher, Hydrocarbon Process. July 1978,
155 –160; Hydrocarbon Process. Sept. 1978,
257 – 262.
142. B. Gamson, R. Elkins, Chem. Eng. Prog. 49
(1953) 203 – 215.
143. Criteria for a Recommended Standard
Occupational Exposure to Hydrogen Sulfide,
U.S. Department of Health Education, and
Welfare Public Health Service – Center for
disease control NIOSH – May 1977.
144. R. O. Beauchamps et al.: “A critical review of
the literature on Hydrogen Sulfide toxicity,”
CRC Crit. Rev. Toxicol. 13 (1984) no. 1,
25 – 97.
145. Toxigenics, Inc.: “90 day vapor inhalation
toxicity study of hydrogen sulfide in Fischer
344 rats, CIIT docket no. 22063, in Sprague
Dawley rats, CIIT docket no. 32063, in
B6C3F1 mice, CIIT docket no. 42063.
Chemical Industry Institute of Toxicology,
Research Triangle Park, N.C., U.S. vol. 1, 2,
1983.
146. T. J. Hughes et al., “Validation of chemical and
biological techniques for evaluation of vapors
in ambient air/mutagenicity testing of twelve
Hydrometallurgy → Metallurgy
Hydrogen Sulfide 21
(12) vapor-phase compounds,” Report
EPA-600/1-84-005; order no. PB-84-164219,
p. 204. Gov. Rep. Announce. Index U.S. 84
(1984) no. 10, 65.
147. M. M. Derzhko, I. R. Barilyak, “Effect of
Sulfur and Hydrogen Sulfide on the
Chromosomal Apparatus of Somatic Cells,”
Dopov. Akad. Nank Ukr. RSR, Ser. B: Geol.
Khim. Biol. Nanki 1 (1986) 70 – 72.
148. D. Demaret, J. Fialaire, “Les intoxications par
hydrogène sulfuré dans une raffinerie de gaz
naturel,” J. Eur. Toxicol. 7 (1974) no. 1,
32 – 36.
149. D. Demaret, J. Fialaire, “Intoxications par
hydrogène sulfuré; urgence médicale dans
l’entreprise,” Arch. Mal. Prof. Méd. Trav.
Sécur. Soc. 39 (1978) 761 – 767.
150. C. Azema, “Règles generales de sécurité dans
l’emploi de l’hydrogène sulfuré,” Soc. Nat.
Elf-Aquitaine (Prod.), in press.
151. “Inflammabilité et explosivité des gaz, liquides
et solides usuees,” Institut National de
Recherche et de la Securité, Note
no. 740-63-71 2nd quarter 1971.
152. “Valeur limite pour les concentrations des
substances dangereuses dans l’air des lieux de
travail,” Institut National de Recherche et de la
Securité, Note ND 1653-129-87-4th quarter
1987.
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
My Profile L

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your Librarian BROWSE THIS TITLE
Save title to My Profile Article Titles A–Z
Hydrogen Email this page
Standard Article Topics
Print this page
Peter Häussinger1,ReinerLohmüller2, Allan M.
Watson3
1Linde AG Werksgruppe TVT München, Höllriegelskreuth, Federal Republic of Germany SEARCH THIS TITLE
2Fachhochschule Ostfriesland, Emden, Federal Republic of Germany
3Linde AG München, Höllriegelskreuth, Federal Republic of Germany
Advanced Product Search
Copyright © 2002 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights reserved.
DOI: 10.1002/14356007.a13_297 Search All Content
Article Online Posting Date: June 15, 2000 Acronym Finder

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. History
2. Properties
2.1. Physical Properties
2.1.1. General Physical Data
2.1.2. Solubilities, Diffusion Properties and Evaporation Equilibria
2.2. Chemical Properties
2.3. Ignition and Detonation Performance
2.4. Isotopes
3. Occurence
4. Production
4.1. Production from Coal and Hydrocarbons
4.1.1. Coke Oven Gas
4.1.2. Gasification of Coal and Hydrocarbons
4.1.2.1. Coal Gasification
4.1.2.2. Gasification of Liquid and Gaseous Hydrocarbons
4.1.2.3. Examples of Industrial Scale Hydrogen Production (by Partial Oxidation of Hydrocarbons)
4.1.3. Catalytic Reforming of Hydrocarbons
4.1.3.1. Feedstocks
4.1.3.2. Tubular Reformer
4.1.3.3. Steam Reforming Plant
4.1.3.4. Special Designs
4.1.3.5. Autothermal Reactors
4.1.4. Refinery Processes
4.1.5. Petrochemical Processes
4.2. Electrolysis

page 1 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
4.2.1. Principles
4.2.2. Conventional Water Electrolysis
4.2.3. New Developments in Electrolytic Processes
4.2.4. Hydrogen as Byproduct from other Electrochemical Processes
4.3. Thermochemical Water Cleavage
4.3.1. Thermodynamics of Closed Cycles
4.3.2. Further Criteria of Thermochemical Cycles
4.3.3. Thermochemical Cyclic Processes
4.3.4. Hybrid Cycle Processes
4.4. Other Methods for the Cleavage of Water
4.4.1. Thermolytic and Radiolytic Processes
4.4.2. Photochemical or Photoelectrical Water Cleavage
4.4.3. Hydrogen Formation in Biological Systems
4.5. Other Chemical Processes
4.5.1. Hydrogen from Conversion of Metals
4.5.2. Hydrogen from Ammonia
4.5.3. Hydrogen from Methanol
4.5.4. Hydrogen from Hydrogen Sulfide
4.5.5. Hydrogen as Byproduct
4.6. Economic Aspects
4.6.1. Analysis of the Cost Structure
4.6.2. Prognosis of the Hydrogen Costs
5. Purification of Hydrogen
5.1. Low-Temperature Processes
5.1.1. Condensation Processes
5.1.2. Condensation and Rectification
5.1.3. Low-Temperature Absorption Processes
5.2. Adsorption Processes
5.2.1. Principles
5.2.2. Thermally Regenerated Adsorbers
5.2.3. Pressure-Swing Adsorption (PSA)
5.3. Catalytic Gas Purification
5.3.1. Carbon Monoxide Removal
5.3.2. Methanation
5.3.3. Selective Oxidation
5.3.4. Catalytic Removal of Sulfur Compounds
5.3.5. Removal of Oxygen
5.3.6. Removal of Nitrogen Oxides
5.4. Scrubbing Processes
5.4.1. Principles
5.4.2. Physical Scrubbing (Rectisol Process)
5.4.3. Chemical Scrubbing Processes
5.5. Permeation Processes
5.5.1. Metal Membrane Techniques
5.5.2. Polymer Membrane Processes
6. Liquefaction of Hydrogen

page 2 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
6.1. Principles of Hydrogen Liquefaction
6.2. Liquefaction Plants
6.3. Further Developments
7. Handling of Hydrogen
7.1. Quality Specifications
7.2. Compression and Expansion
7.3. Transportation and Distribution
7.4. Storage
7.5. Materials
7.6. Removal of Hydrogen
7.7. Safety Aspects
7.8. Toxicology
7.9. Analysis
8. Conventional Uses
8.1. Hydrogen in the Chemical Industry
8.1.1. Ammonia Synthesis
8.1.2. Hydrogen in Refinery Processes
8.1.3. Hydrogen in Coal Refinement
8.1.4. Synthesis Gas
8.1.5. Hydrogen in Organic Synthesis
8.1.6. Hydrogen in Inorganic Synthesis
8.2. Hydrogen in Metallurgy
8.3. Other Uses
9. Hydrogen Energy
9.1. Aims of the Hydrogen Energy Economy
9.2. Production Systems for Hydrogen Energy
9.2.1. Energy Sources
9.2.2. Production Methods for the Energy Carrier Hydrogen
9.3. Hydrogen-Energy Conversion Systems
9.3.1. New Developments for Hydrogen Storage
9.3.1.1. Cryoadsorption
9.3.1.2. Hydride Technology
9.3.1.3. Liquid Organic Hydrogen Carriers
9.3.2. Combustion and Heating
9.3.3. Hydrogen Fuel Systems
9.3.3.1. Hydrogen Propulsion Systems for Space
9.3.3.2. Hydrogen as Aviation Fuel
9.3.3.3. Hydrogen for Automotive Vehicle Transport
9.3.4. Electricity Generation from Hydrogen
9.3.4.1. Heat - Power Processes
9.3.4.2. Electrochemical Energy Conversion
9.4. Future Developments
9.4.1. Production Schemes
9.4.2. Hydrogen Energy Economics
9.4.3. International Research

page 3 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
[Top of Page]

1. History
In the middle ages PARACELSUS is reported to have noted that a gas is yielded when iron is dissolved in the “spirit of vitriole”. TURQUET DE MAYERNE (1573 – 1655) noted that this gas was inflammable.
However, hydrogen was first separated and identified in the second half of the 18th century; BOYLE produced “facticious air” from diluted sulfuric acid and iron.

CAVENDISH (1731 – 1810) proved that there were different types of air, one of which was “inflammable air” and that a number of metals, when dissolved in acid, produced various amounts of this gas. In
1766, he published precise values for the specific weight and density. During the late 1770s, he made experiments with electrical discharges in a hydrogen – oxygen mixture thereby producing water.

In 1785, LAVOISIER, demonstrated the splitting of water into hydrogen and oxygen in a heated copper tube. He also gave the “inflammable air” the name hydrogen.

Chronology of Further Significant Events

1800. First production of hydrogen and oxygen by electrolysis (NICHOLSON and CARLYLE)
1898. Liquefaction of hydrogen using the Linde process (J. DEWAR)
1902. First commercial electrolysis installation by OERLIKON
1929. Production of pure para-hydrogen (K. F. BONHOEFFER, P. HARTECK)
1931. Discovery of the hydrogen isotope deuterium (H. C. UREY)
1935. Synthesis of “superheavy hydrogen”, tritium, through neutron bombardment of deuterated phosphoric acid (P. HARTECK, OLIFANT, E. RUTHERFORD)
1954. Ignition of the first hydrogen bomb on the Bikini Atoll (USA)
1955. Description of the utilization of hydrogen as energy carrier medium (JUSTI)
1969. Development of an overall hydrogen energy concept using the favorable properties of hydrogen and including non-conventional energy systems (J. O. M. BOCKRIS, D. P. GREGORY, C.
MARCHETTI, T. N. VEZIROGLU, and others)
1986. The total hydrogen production worldwide amounts to 500×106 m3 (STP)/a. U.S. hydrogen production and consumption is about 8×106 t.

Hydrogen in Transportation. The buoyancy of hydrogen quickly lead to the use of this gas in balloons (C. CHARLES, 1783), but the use for aeronautics was ended by the HINDENBURG disaster at
Lakehurst in 1937. Experiments using hydrogen as an engine fuel had already been carried out in 1820 by W. CECIL. These experiments were followed by investigations in car engines up until the Second
World War and partly afterwards (RICARDO, BURSTAL, ERREN). Various transport systems using hydrogen were tested, e.g., in the United States by R. BILLINGS or in the Federal Republic of Germany
(Daimler Benz, BMW) in the last decade.

The potential of hydrogen as an aviation fuel was established 1957 with the lift-off of a hydrogen-powered B-57 twin-engine jet bomber. From 1963 on liquid hydrogen –liquid oxygen propelled rockets
were launched. For the Apollo moon flights 12×106 L of liquid hydrogen were necessary to tank the Saturn carrier rockets.

[Top of Page]

2. Properties
Hydrogen (hydrogenium, ύ = the water, = to give birth) [12 385-13-6] H, Ar 1.00 797, atomic number 1, is the first element in the periodic table, having the electron configuration
1 s1. It has usually the oxidation state +1, but in salt-type hydrides –1 is also possible. Three isotopes with Ar 1, 2 and 3 are known; the isotope with Ar 3 is unstable. The differences in the relative isotopic
masses are large so that physical and kinetic properties can differ considerably. The isotope with the relative mass 2 is named deuterium [16873-17-9] (symbol D), that with the mass 3 tritium [
15 086-10-9] (symbol T). The atomic nuclei, all with a single positive charge, are given the names proton, deuteron, and triton.

With the expected increasing significance of hydrogen as a universal chemical and as an energy carrier, its physical and thermodynamic properties have undergone further extensive investigations. The
compilation given here is an excerpt; comprehensive data can be found in various summaries and other literature, e.g., in [1-8].

2.1. Physical Properties

2.1.1. General Physical Data
The atomic radius of the free hydrogen atom is 53 pm, the covalent atomic radius of hydrogen in crystal structures is between 30 and 35 pm. The length of the H – C bond in saturated hydrocarbons is ca.
112 pm, in water vapor the H – O bond has a length of 96 pm. The isotopes H, D and T form diatomic molecules, the combinations being H2, [1333-74-0], D2 [7782-39-0], T2 [10028-17-8], HD [
13983-20-5], HT [14 885-60-0] and DT [14885-61-1]. Atomic and molecular data are given in Table 1.

Table 1. Atomic and molecular data of hydrogen and hydrogen isotopes

page 4 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Protium Deuterium Tritium

Ar 1.007825 2.0140 3.01605

Natural abundance, % 99.985 0.015 <10–12
Half-life, a 12.26
Particle energies, MeV 0.01861 ( -emission)
Thermal neutron capture cross section, barn (10–24 cm2) 0.322 0.51 · 10–3 <6 · 10–6
Nuclear spin, h/2 1/2 1 1/2
Nuclear magnetic moment, nuclear magnetons +2.79278 +0.85742

n-H2 n-D2 n-T2 HD HT DT

Mr 2.016 4.029 6.034 3.022 4.025 5.022

Bond length, pm 74.6 74.2 74.1
Dissociation energy, kJ/mol 431.6 439.2 435.1
Triple point
temperature, K 13.96 18.73 20.62 16.60 17.63 19.71

pressure, kPa 7.3 17.1 21.6 12.8 17.7 19.4

Critical point
temperature, K 32.98 38.35 40.44 35.91 37.13 39.42

pressure, MPa 1.31 1.67 1.85 1.48 1.57 1.77

Normal boiling point, K (atmospheric pressure) 20.39 23.67 25.04 22.13 22.92 24.38
Density at normal boiling point
, kg/m3 70.811 162.50 260.17 114.80 158.62 211.54
L

3 1.316 2.230 3.136 1.802 2.310 2.694

V, kg/m
Heat of vaporisation (25 K), J/mol 825 1175 1400 1000 1100 1290

Ortho- and Para-Hydrogen. The atomic nuclei of hydrogen and its isotopes possess a spin, the value of which is 1/2 for H and T, and 1 for D. In the diatomic molecules both atomic nuclei can either
rotate in the same or in opposing direction, i.e., parallel or anti-parallel. The molecules H2, D2 and T2 can therefore exist in the form of two isomers with different nuclear spin, namely in the ortho-form
(parallel nuclear spin) and in the para-form (anti-parallel nuclear spin); the two forms are interconvertible and in equilibrium with each other. This effect can be observed in all homonuclear diatomic
molecules in which the nuclei have spins, but in the case of hydrogen the effects on properties are most pronounced. The following forms can be distinguished:

o-H2 ortho-hydrogen
p-H2 para-hydrogen
e-H2 equilibrium hydrogen
n-H2 normal hydrogen

page 5 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
e-H2, e-D2 and e-T2 all correspond to mixtures of the equilibrium concentration of definite temperatures. n-H2 (n-D2, n-T2) is a mixture of the equilibrium concentration at ambient temperature, i.e.,
25 mol % p-H2 or p-T2 and 75 % o-H2 or o-T2. n-D2 contains 33.33 % p-D2 and 66.67 % o-D2.

The conversions o-H2 to p-H2 and o-T2 to p-T2 are exothermic; both equilibrium composition and heat of conversion are a function of temperature, but independent of pressure (see Figs. 1 and 2).

Figure 1. Equilibrium composition for the conversion of o-hydrogen – p-hydrogen

Figure 2. Heat of conversion of o-hydrogen – p-hydrogen

For D2 the ortho : para relation is different to that of H2 and T2. The conversion p-D2 to o-D2 is exothermic so that at low temperatures the ortho – form is prevalent, e.g., at 20 K 99.52 %.

The ortho – para equilibrium is attained very slowly in the gaseous phase (see Fig. 3), but more rapidly in the condensed state (see Chap. Liquefaction of Hydrogen). The conversion of deuterium is slower
than that of hydrogen, but tritium reacts more rapidly than H2. Solid tritium, for example has a conversion rate 210 times faster than solid hydrogen. The conversion rate can be increased in the presence
of catalysts or magnetic fields. The hydroxides of Fe (III), Co (III), Ni (II), Cr (III), Mn (IV) or active charcoal are effective as catalysts [9], [10]. Apart from strong homogeneous magnetic fields, the
inhomogeneous magnetic fields of paramagnetic substances, electrical discharges as well as thermo and photochemical activation increase the conversion rate.

Figure 3. Self conversion of o-hydrogen to p-hydrogen in absence of catalysts as a function of time

Thermodynamic Properties. The differences in the masses of the isotopes are relatively large, so that thermodynamic properties differ considerably. Small differences are also found among the spin-spin
isomers so that for certain purposes attention must be paid to the actual form present. At low temperature p-H2, o-D2 or p-T2 can be present in a virtually pure state. The properties of n-H2 (o – p
isomerization prevented by rapid cooling) are also known at low temperatures.

Hydrogen gas is colorless, non-poisonous, odorless and tasteless. At 0 °C and 100 kPa 1 L of H2 has a mass of 0.0898 g. The most important properties of n-H2 and p-H2 are given in Tables 2-4, and
Figure 4.

Table 2. Selection of physical properties of n-hydrogen and p-hydrogen at fixed points

p-Hydrogen n-Hydrogen

Triple point
Temperature, K 13.803 13.957

Pressure, kPa 7.04 7.2

Density (solid), kg/m3 86.48 86.71

Density (liquid), kg/m3 77.03 77.21

page 6 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Density (vapor), kg/m3 0.126 0.130

Boiling point (101.3 kPa),

K 20.268 20.39

Heat of vaporization,
J/mol 898.30 899.1
Liquid phase
Density, kg/m3 70.78 70.96

Specific heat capacity

Cp, J mol–1 K–1 19.70 19.7

Cv, J mol–1 K–1 11.60 11.6

Enthalpy,* J/mol –516.6 548.3

Entropy, J mol–1 K–1 16.08 34.92

Viscosity, mPa s 13.2×10–3 13.3×10–3

Velocity of sound, m/s 1089 1101

Thermal conductivity,

W m–1 K–1 98.92×10–3 100×10–3

Compressibility factor 0.01712 0.01698

Vapor phase
Density, kg/m3 1.338 1.331

Specific heat capacity

Cp, J mol–1 K–1 24.49 24.60

Cv, J mol–1 K–1 13.10 13.2

Enthalpy,* J/mol 381.61 1447.4

Entropy, J mol–1 K–1 60.41 78.94

Viscosity, mPa s 1.13×10–3 1.11×10–3

Velocity of sound, m/s 355 357

Thermal conductivity,

W m–1 K–1 16.94×10–3 16.5×10–3

Compressibility factor 0.906 0.906

Critical point
Temperature, K 32.976 33.19

Pressure, MPa 1.29 1.325

Density, kg/m3 31.43 30.12

page 7 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Properties at STP
(273.15 K, 101.3 kPa)

Density, kg/m3 0.0899 0.0899

Specific heat capacity

Cp, J mol–1 K–1 30.35 28.59

Cv, J mol–1 K–1 21.87 20.3

Viscosity, mPa s 8.34×10–3 8.34×10–3

Velocity of sound, m/s 1246 1246

Thermal conductivity,

W m–1 K–1 182.6×10–3 173.9×10–3

Dielectric constant 1.00027 1.000271

Compressibility factor 1.0005 1.00042

Prandtl number 0.6873 0.680

* The reference state for enthalpy is zero for the ideal gas at 0 K.

Table 3. Some physical and thermodynamic properties of gaseous n-hydrogen

T, °C Density, kg/m3 Cp, J mol–1 K–1 Cv, J mol–1 K–1 Enthalpy, kJ/mol Entropy, J mol–1 K–1

–150 0.1994 23.97 15.65 3.7086 191.38

–150 0.1418 26.42 18.08 4.9434 126.98
– 50 0.1100 27.85 19.55 6.3342 133.87
0 0.0899 28.61 20.32 7.7492 139.59
+ 50 0.0760 29.03 20.68 9.1914 144.44
+100 0.0658 29.20 20.84 10.6461 148.62
+150 0.0580 29.20 20.91 12.1062 152.30

Table 4. Vapor pressure and density of p-hydrogen at low temperatures

Tempera- Vapor Density, kg/m3

ture, K pressure, kPa

S L G

1 11.0×10–37 89.024

5 4.76×10–3 88.965
10 2.556×102 88.136 0.006

page 8 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
12 1.837×103 87.532 0.037
13.803 a 7.0 86.503 77.019 0.126
20 93.5 71.086 1.247
20.268 b 101.3 70.779 1.338
30 822.5 53.930 10.887
32.976 c 1.293×103 31.43

a Triple point.
b 101.3 kPa.
c Critical point.

Figure 4. Vapor pressure of n-hydrogen and p-hydrogen

Contrary to most other gases, the inversion temperature of hydrogen lies below ambient temperature. The temperature, at which the negative Joule – Thomson coefficient (heating by expansion)
changes to a positive (cooling by expansion), can be seen from the Joule – Thomson inversion curve (Fig. 5). A cooling effect is obtained only when the decompression takes place from a state which is
characterized by a point below the inversion curve.

Figure 5. Joule – Thomson inversion curve for p-hydrogen with vapor pressure curve

Liquid hydrogen is a colorless, very mobile liquid with low viscosity and surface tension. Some properties are given in Tables 2 and 4.

Solid hydrogen is colorless and crystallizes in the hexagonal closest-packed structure. Data for p-H2 are: a = 375 pm, c/a = 1.633, molar volume 22.56 cm3/mol [11]. At higher pressure several phase
transitions occur [12]. Under extremely high pressure (ca. 2 – 3×1011 Pa) a metallic, electrically conductive hydrogen phase with a density of >1.000 kg/m3 has been reported [13-15]. Some properties of
solid p-H2 at the triple point are presented in Table 5.

Table 5. Properties of solid p-hydrogen at triple point (7.04 kPa, 13.803 K)

Density, kg/m3 86.50

Heat of fusion, J/mol 117.5
Heat of sublimation, J/mol 1022.9
Molar heat, J mol–1 K–1 5.73
Enthalpy, J/mol –740.2
Entropy, J mol–1 K–1 1.49
Thermal conductivity, W m–1 K–1 0.9

page 9 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Dielectric constant 1.286

2.1.2. Solubilities, Diffusion Properties and Evaporation Equilibria

Solubility of Hydrogen in Liquids. Hydrogen is only slightly soluble in liquids. In contrast to the highly-soluble gases, however, the solubility generally increases with increasing temperature. Examples
for the solubility in nonaqueous solvents are given in Table 6 and Figure 6.

Table 6. Solubility of hydrogen in nonaqueous solvents

Solvent t, °C p, a MPa Solubility, mol %

Hexane 0 0.1 5.82×10–2

Benzene 20 0.1 2.47×10–2
Diethylether 0 0.1 5.29×10–2
Ethanol 0 0.1 1.8×10–2
Carbon tetrachloride 0 0.1 2.60×10–2
Freon 113 25 0.1 6.55×10–2
Naphtha b 25 1 0.031 d
Naphtha 25 5 0.151 d
Naphtha 25 10 0.295 d
Gas oil c 25 1 0.027 d
Gas oil 25 5 0.137 d
Gas oil 25 10 0.274 d

a Partial pressure of hydrogen.

b Density 0.8003 g/cm3, vapor pressure 10.7 kPa (25 °C).
c Density 0.8319 g/cm3, vapor pressure 0.3 kPa (25 °C).
d Moles per kilogram solvent

Figure 6. Solubility of hydrogen in various liquids (hydrogen pressure 10 MPa)

a) N2; b) CO; c) Ar; d) O2; e) CH4; f) C2H4; g) C2H6; h) C3H6; i) C4H10; j) C6H14; k) CO2

The water solubility has a minimum at ca. 50 °C and then increases again with increasing temperatures (see Table 7). Electrolytes lower the water solubility because of the salt effect.

Table 7. Solubility of hydrogen in water at a hydrogen pressure of 101.3 kPa [17]

T, K Solubility, Ostwald –
103 mol % coefficient ×102

273.15 1.755 2.184

283.15 1.576 2.032

page 10 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
293.15 1.455 1.940
303.15 1.377 1.893
313.15 1.330 1.883
323.15 1.310 1.905
333.15 1.312 1.957
343.15 1.333 2.037
353.15 1.371 2.144

Because of the reduced volatility, deuterium, tritium, and their various interisotopic compounds have a somewhat greater solubility than the comparable hydrogen compounds. The relative differences in
solubility amount to a few percent only (close to the accuracy of measurement) and diminish with increasing temperature.

Solubility of Hydrogen in Metals. Metals, with not completely occupied inner electron levels, have a particularly high solubility potential for hydrogen. Hydrogen is dissolved in metals not in the molecular
but in the atomic form, whereby non-stoichiometrical (alloy-type) or stoichiometrical (partly hydride-type) compounds may be formed.

The solubility of hydrogen in metals at small hydrogen concentrations is proportional to the root of the hydrogen partial pressure (Sievert's law). The hydrogen solubility decreases with increasing
temperature. Hydrides can be decomposed thermally.

Palladium and palladium – silver alloys with a silver content <40 wt % have a particularly high solubility potential for hydrogen. The hydrogen atoms give their electrons to the electron gas of the palladium
metal and occupy interstitial sites of the metal crystal structure. Palladium can absorb 2800 times its own volume of hydrogen. Figure 7 shows isotherms for hydrogen in palladium. They are similar to the
Van der Waals isotherms of real gases. For an explanation of the so-called plateau dissociation pressure see Section Hydrogen-Energy Conversion Systems.

Figure 7. Pressure – composition isotherms for hydrogen in palladium [19]

The solubility of the hydrogen isotopes in palladium decreases in the order H > D > T; with group 5 metals (V, Nb, Ta), however, the solubility increases in this order. This effect is explained by differences
in the vibration energies of the metal structure (in palladium: octaeder holes are occupied; in group 5 metals: tetraeder holes are occupied).

Impurities in the metal and in the hydrogen can lower the solubility or cancel it completely. Especially carbon monoxide, hydrogen sulfide and oxygen are poisonous with this respect.

Dissolved hydrogen in metals influences their mechanical properties (hydrogen embrittlement, see Section Materials), magnetic properties (reduction of the paramagnetism of Cr and Ti) and can cause
superconductivity. For example, nonsuperconductive palladium becomes superconducting if hydrogen is added, to the extent of PdH0.8. The transition temperature of PdH1.0 is 8.8 K, that of PdD1.0
10.7 K, whereas Pd0.55Cu0.45H0.7 even has a transition temperature of 17 K.

The absorption of hydrogen in metals and metal alloys has become important for the storage of hydrogen in hydride storage systems. An overview of hydrogen in metals is given in [20] and [21] (
Hydrides, see Section Hydrogen-Energy Conversion Systems).

Solubility of Gases in Liquid Hydrogen. With the exception of helium all gases are solid at the temperature of liquid hydrogen. In liquid hydrogen they are soluble only in traces: oxygen (32 – 26 K)
<0.5 ppm; nitrogen (32 K) ca. 15 ppm, (20.4 K) <0.1 ppm.

The solubility of helium in liquid hydrogen at 1 MPa is shown in Figure 8: In the range of 16 – 26 K it increases from 0.5 to 3 mol % and then decreases at higher temperatures.

page 11 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 8. Solubility of helium in liquid hydrogen (pressure 1 MPa)

• liquid concentration He

Vapor – Liquid Equilibria. Various vapor – liquid equlibrium systems exist at low temperature. However, only the isotope mixtures, e.g., H2 – HD, or H2 – D2 show simple boiling point diagrams with only
little deviation from the ideal behavior. The vapor pressures of the mixture H2 – HD, e.g., deviate 4 % at the most from that of an ideal mixture [22], [23].

Two- and multicomponent mixtures with nitrogen, carbon monoxide and methane are important in low-temperature processes for separation or purification of hydrogen. In these mixtures the hydrogen in
the hydrogen-poor liquid phase is supercritical. Low hydrogen contents in the coexistent liquid can only be reached in the neighborhood of the melting point of the non-hydrogen components.

Figure 9 shows the hydrogen equilibrium concentrations in the liquid and the gas phase in the systems H2 – CO and H2 – N2 [17], [18].

Figure 9. Vapor – liquid equilibria of the systems H2 – N2 (A) and H2 – CO (B)

Isotherms show (clockwise from left to right) the composition of the liquid phase, the critical pressure at the maximum of the isotherm, and the composition of the gas phase as a function of pressure (concentrations in
mol %)

The system H2 – Ne shows a complicated behavior. Figure 10 shows a p-x-diagram for the composition range in which a miscibility gap in the liquid phase occurs: If a mixture with the composition x1, is
compressed isothermally, first two two-phase regions, each with a different liquid-phase composition are found. Then a one-phase region (a hydrogen-rich liquid phase) is reached. At the composition x2
only one two-phase liquid system exists. At the composition x3 an azeotropic mixture is formed, the liquid phase has the same composition as the gas phase. Equilibrium compositions of the system H2 –
Ne are given in [16].

Figure 10. Sketch of a typical isotherm in the range of existence of a miscibility gap of the system n-H2 and Ne [16]

(V = vapor phase, liq. 1 = hydrogen-rich liquid phase; liq. 2 = neon-rich liquid phase)

Diffusion Properties. Hydrogen is the element with the highest diffusion capacity. Some diffusion coefficients of hydrogen in gases and liquids are given in Table 8. For further values see [1], [4], [24],
[25]. In metals with a high hydrogen solubility usually a high hydrogen diffusion coefficient is also found, e.g., in palladium 5×10–7 cm2/s and in vanadium 5×10–5 cm2/s (at 25 °C). These values are in the
same range as the diffusion coefficient of hydrogen ions in water (ca. 10–4 cm2/s).

Table 8. Diffusion coefficients of hydrogen in gases (pressure 100 kPa) and liquids

Diffusion in t, °C D, cm2/s

page 12 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
N2 0 0.674
O2 0 0.701
H2 (selfdiffusion) 0 1.285
H2O, vapor 0 0.759
H2O, liquid 25 4.8×10–5
Iron, smelting 1600 5.64×10–3
Aluminum, smelting 960 1.28×10–5

2.2. Chemical Properties

The first electron shell can be filled with a maximum of two electrons. Therefore, the chemistry of hydrogen depends mainly on three processes, (1) loss of the valency electron to yield the hydrogen ion,
H+; (2) gain of an electron to form the hydride ion, H–, and (3) formation of an electron pair bond. The degree of covalent: ionic character of this bond depends on the electronegativity of the element to
which hydrogen is attached. The electronegativity of hydrogen according to PAULING is 2.1. Except in the hydrogen molecule itself, where the bond is homopolar, all H – X bonds will possess some polar
character.

Hydrogen is not exceptionally reactive, although hydrogen atoms react with one another and with all other elements with the exception of the noble gases (only with helium short-lived instable forms such
as HeH+ and are known). Hydrogen oxidizes less electronegative elements (e.g., alkali and alkaline earth metals), and reduces more electronegative ones (e.g., halogens, oxygen, nitrogen,
carbon).

The strength of the H – X bond in covalent hydrides depends on the electronegativity and size of the element X. The bond strength decreases in a group with increasing atomic number and generally
increases across any period. The most stable covalent bonds are those formed between two hydrogen atoms (or isotopes), or with halogens, oxygen, carbon, and nitrogen. The average values for the
bond energy are given in Table 9.

Table 9. Average bond-dissociation energies of representative covalent hydrogen bonds at 298 K

Bond kJ/mol Bond kJ/mol

H–H 428.2 H–N 314

H–D 440 H–N= 435.4
D–D 443.4 H–O 428.2
T–T 446 H–OH 498.1
H–CH3 435.3 H–S 344.5
H–C= 452 H–Cl 431.6
H–C≡ 536 H–F 568.9

Halogens react with hydrogen to yield hydrogen halides, with increasing reactivity in the sequence iodine, bromine, chlorine, and fluorine. At –210 °C liquid fluorine ignites immediately in hydrogen; solid
fluorine explodes violently in the presence of liquid hydrogen; similarly, combined hydrogen reacts more or less violently with fluorine. Chlorine and bromine also react explosively with hydrogen on
excitation by heat or ultraviolet radiation (chlorine hydrogen reaction). Fluorine, chlorine, and bromine react with hydrogen in a radical chain reaction:

initiation reaction

propagation reactions

page 13 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Oxygen reacts with hydrogen according to

Above 550 °C the reaction occurs with flame propagation, explosion or detonation (oxyhydrogen reaction). The flame temperature is limited by the thermal dissociation of water vapor and attains a
maximum of 2700 °C. For the explosion limits in air or oxygen see Section Ignition and Detonation Performance.

At elevated temperatures in presence of suitable catalysts, hydrogen will react with nitrogen to form ammonia. Industrially this is one of the most important reactions ( Ammonia).

Hydrogen reacts with carbon at high temperatures to yield methane. The equilibrium of the reaction is on the methane side at low temperatures, and on the hydrogen and carbon side at high temperatures.
The equilibrium constant, using graphite as a basis, is 2×105 bar at 150 °C and 9.4×10–3 bar at 1000 °C. Equilibria related to amorphous carbon and other values are given in [26], [27]. Carbon monoxide
and carbon dioxide can react with hydrogen in the presence of a catalyst. Depending on the reaction conditions, catalyst and CO : H2 ratio a variety of products can be formed ( Gas Production, see
Section Synthesis Gas).

Unsaturated hydrocarbons are converted to saturated or partially saturated hydrocarbons by hydrogenation.

Further reactions of hydrogen include reductions of functional organic groups (–NO2 NH2, –COOH CH2OH, etc). In the petroleum industry ( Oil Refining) hydrogenation and hydrocracking
reactions are of importance as well as desulfurization and selective hydrogenation (dienes, monoolefins). For details of reduction and hydrogenation with hydrogen of other compounds see the appropriate
keywords.

A great number of metals react with hydrogen to form hydrides. Depending on the nature of the hydrogen bond, hydrides are classified into three principal categories: saline, covalent and metallic (see
surveys in [20], [21]), ( Hydrides).

The saline hydrides or ionic hydrides are formed when the strongly electropositive alkali metals and the alkaline earth metals (calcium, strontium, barium), all usually in the elemental form, react with
hydrogen at high temperatures. Salt-like hydrides contain the hydride ion, H–; they are crystalline, have high heats of formation and high melting points. They are reactive, powerful reducing agents, all of
them reacting with water to liberate hydrogen.

Covalent hydrides may be either solid, liquid or gaseous; typical are the hydrides of boron, aluminum, silicon, germanium and tin. Beryllium and magnesium hydride seem to represent a transition between
ionic and covalent hydrides. The bond between hydrogen and the atom is of the nonpolar electron-sharing type. Molecular stability often is provided by means of three-centered two-electron bridge bonds
(e.g., in diborane, B2H6, an electron-deficient type molecule), by hydride anion formation ( , ), or formation of polymeric structures [polymeric (AlH3)x].

Metallic hydrides are formed by the transition metals. Transformation of a pure metal into a hydride occurs through continuous solution of hydrogen in the metal with subsequent abrupt phase transition at
defined stoichiometric hydride phases (ZrH2, PdH, VH, VH2).

The Hydrogen Bond. Substances containing hydrogen coupled to the most electronegative elements exhibit properties that are best explained by assuming that the hydrogen atom of an H – X bond still
has a small but significant affinity for other electronegative atoms although it remains strongly bonded to the original atom. This relatively weak secondary bond is called a hydrogen bond normally depicted
by the notation

Hydrogen bonds result in molecular association. NH3, H2O, and HF have much higher boiling points and heats of vaporization than the corresponding homologues PH3, H2S, and HCl. Carboxylic acids
such as acetic or benzoic acids are present as dimers with the configuration

The presence of hydrogen bonds also causes formation of crystalline hydrates such as NH3 · H2O, SO2 · H2O hydrocarbon hydrates (e.g., with natural gas components) as well as the zig-zag structure of
the (HF)n polymer.

The existence of the hydrogen bond can be shown by infrared spectroscopy. When the hydrogen of the X – H group forms a hydrogen bond with the atom Y, the X – H stretching frequency is lowered and
the absorption band characteristically broadened. In NMR spectroscopy a proton, which participates in a hydrogen bond, produces a definite paramagnetic shift.

page 14 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The energies of hydrogen bonds are in the range of 4 – 40 kJ/mol, which is small as compared to 200 – 400 kJ/mol for ordinary bonds. The average bond length is 200 – 300 pm (O – H . . . O in ice
276 pm).

Atomic Hydrogen. Atomic hydrogen is much more reactive than the molecular form. The dissociation reaction

is highly endothermic. Atomic hydrogen can be produced thermally by supplying sufficient energy, e.g., in electric arcs with a high current density, electrical discharges at low pressures, irradiation with UV
light as well as electron bombardment (10 – 20 eV). The hydrogen atom has a short half-life (ca. 0.3 s) and reacts even at room temperature with nonmetallic elements such as halogens, oxygen, sulfur,
and phosphor to form the corresponding hydrogen compounds. At room temperature it reduces many oxides to the elements. The heat of recombination of hydrogen molecules from hydrogen atoms leads
to very high temperatures (Langmuir flare, 4000 K).

Atomic hydrogen can be absorbed in the metallic structure of some elements of the Group 8 – 10 (Fe, Co, Ni) of the periodic system. The suitability of many catalysts for reactions involving hydrogen is
based on the dissociation and solubility of hydrogen in the atomic form.

Hydrogen formed upon dissolution of metals in acids is very reactive at the moment of formation (nascent hydrogen), much more than normal hydrogen. This strong reduction capability is often explained
by a shortterm formation of atomic hydrogen. Another explanation is the intermediate formation of a hydride-type compound which is very similar to Grignard compounds ( Magnesium Compounds).

The Hydrogen Ion. The ionization potential of the reaction

is 13.595 eV, which is even higher than the first ionization potential of the noble gas xenon.

The hydrogen ion can be formed only in a medium which solvates the protons. The solvation process provides the energy required for bond rupture, the order of magnitude can be seen in the solvation
reaction with water:

In aqueous systems H3O+ is the form implied by the customary designation hydrogen ion. The structure of the H3O+ ion is that of a rather flat triangular pyramid with a HOH bond angle of ca. 1100, O – H
bonds distances of 102 pm and a H – H distance of 172 pm. Normally, it is coordinated with four water molecules not all of which are equivalent, so that the aquated hydrogen ion is best represented as
H9O4 · H2O [28]. The lifetime of an individual H3O+ ion is exceedingly short, 10–13 s, because the protons are rapidly exchanged among the water molecules.

Hydrogen Electrode. In aqueous solutions the hydrogen ion activity is given in terms of the pH value, defined as log 1/[H+], where [H+] is the hydrogen ion activity. By definition, the potential of the redox
system

is zero (E = 0.000 V) at all temperatures, when an inert metallic electrode is immersed in a solution of unit activity (i.e., pH = 0) in equilibrium with hydrogen gas at 100 kPa pressure. The electrode material
is a small platinum foil or a wire, coated with finely dispersed platinum black and is placed partly in the solution and partly in the hydrogen atmosphere. This standard hydrogen electrode is the reference
electrode of all other oxidation-reduction systems.

For pH measurement the potential difference of a hydrogen electrode to any other suitable reference electrode can be used. The calomel electrode, or half-cell, is usually employed. The pH value is given
by

where: is the potential difference between the hydrogen and the reference electrode (in Volt), v the potential of the reference electrode (in Volt), and pH2 the hydrogen partial pressure (in bar) [29],
[30].

2.3. Ignition and Detonation Performance

Given sufficient thermal activation (ignition energy), hydrogen reacts more or less violently with oxidizing agents such as oxygen (air), fluorine or chlorine, and N2O. Depending on conditions, combustion,
deflagration or detonation may occur.

Ignition and detonation properties of hydrogen – air mixtures are particularly important from the safety aspect (see also Section Safety Aspects). The flammability limits (i.e., the difference between the rich
and lean limit concentration) are exeptionally wide for hydrogen. Table 10 shows a comparison of safety-relevant thermo-physical and combustion properties of hydrogen with those of methane, propane

page 15 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
and gasoline [31]. The flammability limits are affected by temperature as shown in Figure 11, so that a preheated mixture has considerably wider limits for coherent flames [32]. An increase in pressures
up to 10 kPa has only a small effect. Water vapor has a strongly inhibiting influence on the oxyhydrogen reaction (see ternary diagram, Figure 12). The flammability limits of hydrogen in pure oxygen are
4.65 – 93.9 vol % hydrogen, those of deuterium in pure oxygen 5.0 – 95 vol %. Ignition and detonation performance based on ignition range, ignition energy, flame propagation as well as transition from
deflagration to detonation, are discussed in detail in Section Safety Aspects.

Table 10. Combustion and explosion properties of hydrogen, methane, propane and gasoline

Hydrogen Methane Propane Gasoline

Density of gas at standard conditions, kg/m3 (STP) 0.084 0.65 2.42 4.4 a
Heat of vaporisation, J/g 445.6 509.9 250 – 400
Lower heating value, kJ/g 119.93 50.02 46.35 44.5
Higher heating value, kJ/g 141.8 55.3 50.41 48
Thermal conductivity of gas at standard conditions,
mW cm–1 K–1 1.897 0.33 0.18 0.112

Diffusion coefficient in air at standard conditions, cm2/s 0.61 0.16 0.12 0.05
Flammability limits in air, vol % 4.0 – 75
5.3 – 15 2.1 – 9.5 1 – 7.6
Detonability limits in air, vol % 18.3 – 59
6.3 – 13.5 1.1 – 3.3
Limiting oxygen index, vol % 5
12.1 11.6 b
Stoichiometric composition in air, vol % 29.53 9.48 4.03 1.76
Minimum energy for ignition in air, mJ 0.02 0.29 0.26 0.24
Autoignition temperature, K 858 813 760 500 – 744
Flame temperature in air, K 2318 2148 2385 2470
Maximum burning velocity in air at standard conditions, m/s 3.46 0.45 0.47 1.76
Detonation velocity in air at standard conditions, km/s 1.48 – 2.15 1.4 – 1.64 1.85 1.4 – 1.7 c
Energy d of explosion, mass-related, gTNT/g 24 11 10 10
Energy d of explosion, volume-related, gTNT/m3 (STP) 2.02 7.03 20.5 44.2

a 100kPa and 15.5 °C.

b Average value for a mixture of C1– C4 and higher hydrocarbons including benzene.
c Based on the properties of n-pentane and benzene.
d Theoretical explosive yields.

Figure 11. Effect of temperature on flammability limits of hydrogen in air (pressure 100 kPa)

Figure 12. Flammability and detonability limits of the three component system hydrogen – air – water vapor [33]

a) 42 °C, 100 kPa; b) 167 °C, 100 kPa, c) 167 °C, 800 kPa

page 16 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Hydrogen reacts with chlorine after excitation with radiation, heat or in the presence of catalysts. The ignition and explosion behavior of this reaction is of interest for the chlor-alkali electrolysis, because
chlorine is usually liquefied after formation. The residual gas contains hydrogen (among other components) and might reach the ignition or detonation range because of the decreasing chlorine
concentration in the gas phase during chlorine liquefaction.

The lower ignition limit of hydrogen-chlorine air mixtures is ca. 4 – 6.5 vol % H2, the upper ignition limit lies between 73 vol % H2 (Cl2 concentration = 0 vol %) and 89 vol % H2 (air
concentration = 0 vol %). The lower detonation limit is between ca. 20 vol % H2 (Cl2 concentration = 0) and ca. 25 vol % H2 (Cl2 concentration = 30 – 40 vol %). An overview of the ignition and detonation
properties of hydrogen – chlorine mixtures, and the influence of oxygen and inert gas is given in [34].

2.4. Isotopes
Hydrogen and deuterium are stable isotopes; tritium emits a low-energy -radiation (disintegration into ). The half-life is 12.26 a. The natural isotope distribution on earth is 99.985 atom % H and
0.0145 atom % D. Natural tritium occurs in the upper atmosphere and is caused by interaction of cosmic rays, mainly according to:

Other tritium sources are nuclear reactors, where tritium is formed from lithium or D2O by a neutron capture mechanism:

Tritium is also formed by fission processes, e.g., in nuclear experiments. Tritium disintegration and varying formation rates lead to changes in the tritium content. It is estimated to be ca. 10–13 – 10–12
atom % of the total hydrogen [35]. An overview is given in [36] and [37].

Physical Properties. The most important atomic and molecular data of hydrogen, deuterium, tritium and their diatomic combinations are given in Table 1. Further thermodynamic data are tabulated in [38]
for deuterium and [39] for tritium. Hydrogen, deuterium, and tritium show nuclear spin isomerism (see Section General Physical Data).

Figure 13 shows the vapor pressures of hydrogen molecules with a different isotopic composition. In [40] the system D2 – DT – T2 is described in the range 17 – 22 K. The compounds are totally miscible
in the liquid and in the solid state. A 50 : 50 mol % mixture of D2 – T2 has a T content of 42 % in the gas phase and of 52 % in the solid phase at 19.7 K.

Figure 13. Vapor pressure of hydrogen isotopes a) H2; b) HD; c) HT; d) D2; e) DT; f) T2

Various chemical and physical processes are used for the enrichment and separation of D2 or D2O from the natural isotopic mixture. All chemical processes are based on isotopic exchange reactions,
e.g., the dual temperature H2O – H2S process, according to

in the ammonia exchange process

page 17 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

and variations of this process such as the methylamine hydrogen process from Aecl-Sulzer according to

The processes are described in [36] as well as in Isotopes, Natural.

On a small scale the electrolysis of enriched H2O – D2O is used, because this gives an isotopic separation factor of 7 – 8.

Low-temperature destillation of H2 – D2 is an example of the physical separation of the isotopes [41]. The heat generated by the ortho-para conversion must be taken into consideration. Other processes
use pressure swing adsorption techniques [42]; the various permeation rates of H2 and D2 through metal membranes (Pd, Ta); or differences in the hydride or deuteride formation with metals [43], (see
Section Hydrogen-Energy Conversion Systems).

In general, the chemical properties of H, D and T are essentially identical. Small differences exist in the rates and the equilibrium constants of reactions because of the kinetic isotope and equilibrium
effects. A greater amount of energy is required for the dissociation of a D – X bond and the activation into the transition state than for the comparable H – X bond (for numerable investigations see [44]). In
biological systems substitution of deuterium for hydrogen atom can alter the delicately balanced processes substantially.

Deuterium has excellent properties as a moderator (i.e., a high ratio of neutron deceleration to absorption characteristics). The separation of the hydrogen isotopes is technically important for the
production of heavy water (D2O) as a moderator for nuclear reactors ( Nuclear Technology).

Tritium also is used to mark organic compounds and for dating specimens. A detailed description of tritium and its compounds is found in [45].

[Top of Page]

3. Occurence
Hydrogen is the most abundant element in the universe. 70 – 80 wt % of the observable universe consists of hydrogen, the rest is mostly helium. Especially the younger stars (the novas) inter-stellar dust,
and gas clouds contain hydrogen in large amounts. More than 50 wt % of the sun consists of hydrogen.

The neutral hydrogen atom which emits electromagnetic waves with a frequency of 1420 MHz, yields information on the structure of the universe. Large regions of interstellar space are filled with protons,
which emit a radio continuum [46], [47].

Of the elements in the earth's crust, hydrogen takes the 9th place with 0.88 wt %. Based on the atomic concentration of 15.4 atom % it is the third most common element after oxygen and silicon. In the
free state, hydrogen can be found in some vulcanic gases (up to ca. 30 vol %), in natural gas (from traces up to 40 vol % [48]); small quantities are trapped in some minerals, rocks and meterorites. In the
upper atmosphere, the troposphere and the stratosphere, hydrogen is present only in traces of ca. 0.5 ppm (mL/m3). Because of photolytic water cleavage the hydrogen concentration increases with
increasing height. The exosphere (2000 – 20 000 km) contains atomic hydrogen.

Chemically bound hydrogen is present all over the earth: as part of the Earth's water mass (11.2 wt %), as part of organic matter (including fossil substances such as natural gas, oil, and coal). About
10 wt % of the human body consists of hydrogen.

[Top of Page]

4. Production
Hydrocarbons are the main source for the production of hydrogen on an industrial scale. An even larger source, however is water. When hydrogen is used as fuel, water is produced again. Thus, a cycle
comparable to the carbon cycle (CO2 – biomass – CO2) —but much simpler—is involved, so that water can be regarded as an unlimited source of hydrogen. With the help of an external energy source,
hydrogen can be produced from either of these sources alone or from a mixture of both.

Figure 14 shows the production of hydrogen using different sources of energy. The many production routes for hydrogen, however, are by no means of equal economic importance. Most of the hydrogen
for industrial uses is produced from natural gas and oil, either as a main product or as a byproduct, a process involving a chemical conversion. A smaller percentage is produced electrolytically or occurs
as a byproduct of electrochemical processes. Here the energy applied is in the form of electricity, which itself may be produced using hydroelectric, nuclear or, still predominant, the energy stored in fossil
fuels.

page 18 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 14. Primary energy, raw materials, production methods, paths of production, transformation, and generation of and for hydrogen, adapted from [49]

Thermal, thermochemical, biochemical and photochemical processes have so far not found industrial applications. The preference of producing hydrogen from hydrocarbons is due to the fact that the
energy demand using hydrocarbons is much lower than that for electrolysis of water (see Table 11).

Table 11. Theoretical energy consumption for the production of hydrogen from various hydrocarbons, coal and water

Natural gas LPG as Naphtha as Heavy oil as Coal as Water

as CH4 CH2.6 CH2.2 CH1.4 CH0.7 H2O

Higher heating value, *

kJ/kg 50 050 46 130 44 300 41 300 38 100

kJ/mol 890 2480

Byproducts CO2 CO2 CO2 CO2/S CO2/S O2
Specific production
of byproducts,
kmol/kmol H2 0.25 0.30 0.32 0.37/0.003 – 0.015 0.43/0.002 – 0.01 0.5

Theoretical energy
consumption, *
kJ/kmol H2 41 280 37 500 38 350 50 300 57 150 242 000

* Calculated from the heat of formation, oxidizing carbon completely to carbon dioxide (at 298 K and 0.1 MPa) [73], [78].

Considering the strong correlation between energy prices and the prices for hydrocarbons, it is unlikely that the high percentage of the hydrocarbon processes for producing industrial hydrogen will change
substantially in the near future. This last statement does not take into consideration environmental pollution by emission of carbon dioxide, sulfur compounds, nitrogen oxides and heat, so that this
(environmental) aspect may reverse the situation one day.

4.1. Production from Coal and Hydrocarbons

The yield of hydrogen from hydrocarbons is summarized in Figure 15. Hydrocarbons and coal are characterized by their molar H : C ratio. Chemically pure compounds such as methane and its
homologues can be found on line (A) and give maximum hydrogen yield. Raw materials such as hard coal (a), lignite (b) and natural gas (h) give a lower hydrogen yield because they contain inert
components (sulfur, nitrogen, oxygen or minerals etc.). Other important sources for the production of hydrogen such as vacuum residue (c), naphtha (d) or hydrogen-rich off-gases (e.g. coke oven gas (j),
are also shown in Figure 15.

Figure 15. Hydrogen production from hydrocarbons and coal

A) Maximum theoretical yield from (a) hard coal, (b) bituminous coal, (c) vacuum residue, (d) naphtha, (e) butane, (f) propane, (g) ethane, (h) natural gas, (i) methane, (j) coke oven gas

B) Commercial production by steam-reforming, including fuel, no credit for export steamLower curve: small plants, ca. 500 m3 (STP)/hUpper curve: large plants

page 19 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

C) Commercial production by partial oxidationlarge plants only

Upper curve: without oxygen production

The yields actually achieved in industrial hydrogen production are quite different from the maximum possible theoretical yields. This, however, is not due to a low degree of conversion but to the fact, that
the energy needed to produce hydrogen is usually supplied by the raw material itself. Technical and economic aspects of the production are dealt with in this chapter. For the latter not only the yield but
also the prices of the feedstock and the investment costs for the plant are important. Because of the fact that (1) there are only small differences between the cost of hydrocarbons on the basis of heating
value, (2) the investment costs for the production of hydrogen from light hydrocarbons are lower, and (3) yields from light hydrocarbons are higher, as shown in Figure 15, hydrogen is produced preferably
from light hydrocarbons, if available.

4.1.1. Coke Oven Gas (see also Coal Pyrolysis)

Coke oven gas (COG) was formerly a major source of hydrogen. Coke oven gas provided, in conjunction with water gas, almost all of the hydrogen for ammonia synthesis and for coal liquefaction (
Coal Liquefaction) in the first half of this century. While the demand for hydrogen has risen sharply since 1950, the production of coke has been reduced. New production techniques for steel as well as
reduced steel production in most industrialized countries have led to a decreasing market for coke and COG. Producing iron by direct reduction of iron oxide pellets with synthesis gas, for example,
requires no coke at all. Nevertheless COG is, where available, an important and convenient source of hydrogen, but on the whole, production from coke oven gas has now been replaced by more effective
hydrocarbon-based processes and by coal gasification.

Modern coke oven plants produce ca. 350 m3 (STP) of coke oven gas per tonne of coal (15.6 mol/kg). 45 % of the raw COG produced is utilized to fire the coke oven batteries and supply other energy
requirements of the coke oven plant. The rest can be used as a source of hydrogen or exported as a fuel gas. It can be expected that further developments and better heat management within the coke
oven plant will raise the amount of excess gas up to 70 % of the total gas produced [50].

Raw Gas Treatment. Coke oven gas leaves the coke oven batteries at 750 – 850 °C and approximately atmospheric pressure. The main impurities which have to be eliminated are tar, several heavy to
medium oil fractions with a high aromatic content, hydrogen cyanide, ammonia, organic and inorganic sulfur compounds, naphthalene, nitrogen oxides and water. The concentrations of the impurities vary
with the kind of coal and the type and duration of the coking process. Several proprietary techniques for cleaning of raw COG are available [51], [52]. Typical cleaning steps are:

1. Quench cooling to ambient temperature (15 – 30 °C) and separation of the oil and naphthalene fractions,
2. First raw gas compression,
3. Absorption of hydrogen cyanide, ammonia, and sulfur compounds,
4. Second raw gas compression

Pretreated COG contains 50 – 60 % hydrogen (see Table 12).

Table 12. Range of composition for precleaned coke oven gases and production of hydrogen from typical gas by pressure swing adsorption

Components Range of composition, vol % Typical composition, vol % Hydrogen product, vol % Fuel gas, vol %

H2 50 – 64 57 99.9* 22
N2 3 – 12 3 0.09 5.5
CO 4–8 6.5 <10 ppm 12
CO2 1–5 3 < 1 ppm 5.5
CH4 24 – 27 26 0.01 47
C2H4 1–3 2 4
C2H6 0.7 – 2 1 2
C3H6 0.3 – 0.6 0.5 1

page 20 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
C6H6 0.5 – 1 0.8 1
O2 0.1 – 0.5 0.1 <10 ppm
H2S + COS 0.1 – 0.8 0.1 <10 ppm
Organic sulfur, 0.5 – 10 2
g/kmol
Total 100 100 100
LHV, kJ/mol 370 – 460 440 245 588

HHV, kJ/mol 420 – 520 490 286 645

Typical flows with
pressure swing
adsorption,

m3 (STP)/h 100 45 55

* Possible up to 99.999 %.

Separation of Hydrogen from Pretreated COG. The separation of hydrogen from pretreated COG has been accomplished in two different ways.

Low-Temperature Condensation. Conventionally, hydrogen is separated by a low-temperature condensation process. Condensation of methane, ethylene and other compounds with a higher boiling point
is carried out in regenerators (large vessels filled with ceramic materials) at elevated pressures (0.5 – 1 MPa) and low temperatures (60 – 70 K).

The hydrogen purity attainable is 96.5 %. Utilizing the low temperatures, hydrogen can be further cleaned in a liquid nitrogen scrubber (see Section Low-Temperature Absorption Processes), where the
residual methane, carbon monoxide and oxygen content can be reduced down to low levels. In the nitrogen scrubber, the gas can be conditioned for ammonia synthesis gas production (25 % N2, 75 % H
2).

By fractionating the primary condensate from the regenerators ethane, ethylene, propane, and fuel gas can be obtained as byproducts along with hydrogen or ammonia synthesis gas.

Plants with capacities up to 3×105 m3 (STP)/h have been built using this type of technology.

Pressure Swing Adsorption. Pure hydrogen can be recovered at much lower investment costs in modern pressure-swing adsorption (PSA) plants ( Adsorption – Regeneration by Pressure Swing). A
first industrial-scale plant with 10 000 m3 (STP)/h capacity was put into operation in the German Ruhrgebiet in 1982. Figure 16 shows the principal steps of the plant.

Figure 16. Hydrogen production from coke oven gas by pressure swing adsorption (Linde, Federal Republic of Germany)

a) Raw gas compressor; b) Cooler; c) Preadsorption (regeneration not shown); d) Pressure swing adsorption; e) Oxygen removal; f) Water removal; g) Rest gas equalizer; h) Fuel gas compressor

After compression, higher hydrocarbons and sulfur compounds are separated from the raw gas by a preadsorption stage. Separation of hydrogen from the other gas components occurs in an adsorption
process with four or more beds (see Section Adsorption Processes).

Special precautions must be taken to avoid accumulation of oxygen on the adsorbent. Part of the oxygen remains in the product gas and is converted to water in a special reactor (see Section Selective
Oxidation for Deoxo reaction). Hydrogen is then dried in another adsorption process yielding a purity >99.9 %, before it is delivered to a hydrogen pipeline system (see Chap. Handling of Hydrogen).

About 80 % of the hydrogen is recovered by this process. Thus, depending on the initial composition, the gas volume is reduced by 40 – 50 % on producing pure hydrogen. However, the residual gas,
which is usually used for firing the coke oven batteries or otherwise as a fuel gas, has a lower heating value (LHV) of ca. 587 kJ/kmol so that recovery of the valuable hydrogen product stream leads only
to a 20 – 30 % drop in the total heat content of the remaining fuel gas (see Table 12).

Proposed Technology for Production of Hydrogen from COG by Chemical Processes. A higher hydrogen yield could be obtained, if the hydrocarbons in the raw or pretreated COG were converted to
hydrogen.

page 21 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Catalytic processes for this purpose are steam-reforming and autothermic processes (see Section Catalytic Reforming of Hydrocarbons). Both processes run best when the sulfur content is low and
unsaturated and cyclic hydrocarbons are absent. This requires multi-stage precleaning processes, which are technically feasible and often used for other applications ( Gas Production). For COG,
however, they have not yet been commercially applied so far though they have been frequently proposed [53], [54].

Partial oxidation of COG (see Section Gasification of Coal and Hydrocarbons) has also been proposed [52]. But, because of difficulties with the low initial pressure (approx. atmospheric) and the periodical
changes in amount and composition of COG, no commercial application has been reported.

The above refers to COG from hard coal only. Similar aspects apply when gas from coking operations in refineries (see Section Refinery Processes) or from modern pyrolysis processes of hard coal,
lignite, organic materials (straw, refuse, etc.) is considered as a source of hydrogen.

4.1.2. Gasification of Coal and Hydrocarbons

4.1.2.1. Coal Gasification
For the production of hydrogen or hydrogen-rich gas streams coal gasification is usually performed with oxygen (purity > 95 %) at high temperature and elevated pressure. Coal gasification is described in
detail elsewhere ( Gas Production). Here, only aspects relevant to the production of hydrogen and hydrogen-rich gases are dealt with.

Coal gasification on an industrial scale produces:

1. Synthesis gases, rich in carbon monoxide and hydrogen, e.g., gas for Fischer-Tropsch synthesis of gasoline at SASOL, South Africa ( Coal Liquefaction), [55];
2. Synthesis gas for the production of acetic anhydride at Tennessee Eastman, United States [56];
3. Ammonia synthesis gas: 3 H2 + N2, e.g., Modderfontein, South Africa [57];
4. Substitute natural gas >90 % CH4, e.g., North Dakota, United States [58];
5. Medium- and low-Btu fuel gas for combined-cycle processes [59] consisting of hydrogen, carbon monoxide, methane, carbon dioxide. The gas may also contain considerable amounts of inerts
(nitrogen, argon), if gasification is performed with air.
6. Pure hydrogen (hydrogen content > 95 %) from a side stream of a synthesis gas unit [60].

All coal gasification processes can be used to manufacture hydrogen. The highest yield in the primary gasification step, however, is achieved, if pulverized coal is gasified with oxygen at high
temperatures. This is done in the entrained-bed gasification reactor, commercialized especially in the Koppers – Totzek and the Texaco process.

The most important process steps are shown in Figure 17 for the Texaco coal gasification process.

Figure 17. Texaco coal gasification process; direct quench mode of operation

A) Coal grinding and slurry preparation; B) Gasification and gas cooling; C) Gas scrubbing

a) Coal grinding mill, b) Slurry tank; c) Slurry pump; d) Texaco gasifier; e) Quench chamber; f) Lockhopper; g) Particulate scrubber; h) Clarifier; i) Slag separator; j) Slag sump

Gasification temperatures > 1300 °C result in a synthesis gas which contains only minor amounts of methane and other hydrocarbons. The chemical equilibrium at these temperatures leads to a gas
containing mainly hydrogen and carbon monoxide (see Table 13). The latter can be shift converted to hydrogen (see below).

Table 13. Hydrogen production by coal gasification. Effect of gasification pressure, temperature and operation mode

Low temperature gasification Medium temperature gasification High temperature gasification

Feedstock
Coal type North Dakota lignite Bituminous coal Bituminous coal
Volatile matter, wt % 26.9 34.3 33.5

Fixed carbon, wt % 30.6 50.2 41.7

Moisture, wt % 36.2 5.0 13.0

Ash, wt % 6.3 10.5 11.8

page 22 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Carbon, wt % (waf*) 73.9 83.6 76.7

Hydrogen, wt % 5.2 5.7 5.9

Oxygen, wt % 18.9 6.8 12.6

Nitrogen, wt % 1.0 1.6 1.1

Sulfur, wt % 1.0 2.0 3.7

Gasification conditions
Operating pressure, 3.1 3.0 2.0 – 8.0
MPa
Operating temperature, ° 700 – 800 950 1250 – 1400
C
Ratio: steam: oxygen, 6.9 1 1
kg/m3 (STP)
Oxygen consumption
per crude gas,

m3/m3 (STP) 0.13 0.2 0.32

Crude gas yield, m 2000 2200 2200

3 (STP)/t waf*

Raw gas composition

H2, vol % 39.3 28.0 35.8

CO, vol % 15.8 56.4 44.6

CO2, vol % 31.5 3.0 18.0

CH4, vol % 11.8 7.1 0.0

C2+, vol % 0.8 0.4 0.0

N2/Ar, vol % 0.4 4.3 0.5

H2S/COS, vol % 0.4 0.8 1.1

Other ammonia, benzene, phenolics, tar, organic

acids
Commercial processes fixed bed/slow moving bed gasification fixed bed/slow moving bed: British Gas – entrained bed gasification: Texaco,
processes: Lurgi, Wellmann – Galusha Lurgi, fluidized bed: Winkler Hygas, Shell, Koppers – Totzek

* waf = water and ash free.

Gasification pressure affects the hydrogen yield and the economics of the process. Whereas the yield of hydrogen is slightly reduced, the economics are improved at high pressure because hydrogen is
usually utilized at higher pressure. Modern gasification processes, therefore, are operated at least at 2 MPa. The negative effect on the hydrogen yield can be offset by a slight increase in temperature.

Table 13 shows the operating conditions of several commercial gasification processes. Since different coals have been gasified, the hydrogen yields cannot be compared directly. However, the influence
of gasification temperature on hydrogen yield and the formation of byproducts can be clearly seen.

Principally, all types of coal can be gasified to produce hydrogen, although certain constraints associated with the gasification process or the characteristics of the coal have to be considered. The
gasification of petroleum coke in a fluidized bed gasifier has been proposed [61] and the gasification of coke from tar sands has been successfully tested in an entrained bed gasifier [62].

Gasification of other solid fuels such as peat, wood, refuse, or sludge from wastewater treatment plants leads to a low-Btu gas, which is less suitable for the production of hydrogen. Because of the high

page 23 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
moisture and oxygen content of the raw material, these gases are generally used on the plant site for heat or electricity production. In Finland one plant is in operation since 1988 which produces ammonia
synthesis gas from peat, employing a high temperature Winkler gasifier ( Gas Production) [63].

Research and Development. Commercial conditions, namely the low crude oil prices in the 1980s, have prevented further development of new coal gasification processes developed in the 1970s from
bench-scale experiments to commercial-scale processes. Instead research has aimed at making the established processes more reliable. Further progress has been made in improving the purification
steps from the raw synthesis gas to pure hydrogen (see Chap. Purification of Hydrogen).

4.1.2.2. Gasification of Liquid and Gaseous Hydrocarbons

Partial oxidation of hydrocarbons in general is described in detail elsewhere ( Gas Production – Partial Oxidation of Hydrocarbons). For the production of hydrogen, the Texaco and the Shell partial
oxidation (PO) processes are applied commercially [64-67].

The principal steps of an oil gasification process (Texaco process) are similar to the coal gasification depicted in Figure 17. The Shell process differs in details such as the burner construction, the reactor
cooling, the waste heat boiler and the soot recycle. Some more details of both processes are given below:

Feedstock Problems
High boiling point fractions Cracking and carbon formation on heating, high pressure drops at high viscosity
Solid residues Slurrying hydrophobic residues, high demands on the burner

High metal content Corrosion and metal deposition in down-stream equipment

Ash and slag content Mechanical equipment for solid removal, melting point of the ash, eutectic formation with brick work, wear by erosion
Treatment of carbon deposits No recycle: carbon loss, disposal problems Recycle: metal content build-up
Heat recovery Quench: reliability, direct steam production for shift conversion applied with: H2 production
Waste-heat boiler: high steam production rate; applied with: synthesis gas production

Heavy residues from petrochemical processes are the preferred feedstocks for the production of hydrogen and carbon monoxide-rich gases. Oil fractions with high sulfur and/or high heavy metal contents
are difficult to handle in upgrading processes such as hydrogenation or coking. For environmental reasons they should not be used as fuels. In principle, hydrogen can be produced by PO processes from
all hydrocarbons. The production, however, is only economically viable if heavy residues are utilized. (see Section Economic Aspects for an economic comparison of hydrogen-producing processes).

Gaseous hydrocarbons are also occasionally converted to hydrogen and carbon monoxide in PO processes. This is done, if the gases cannot be processed by catalytic steam reforming (see Section
Catalytic Reforming of Hydrocarbons) because of high concentrations of certain impurities such as olefins and H2S or if the process must handle various feedstocks from natural gas up to fuel oils.
Another incentive to use PO for the conversion of gaseous fuel is the possibility to obtain synthesis gas with a high carbon monoxide content.

The overall reactions during partial oxidation are as follows (heat of reaction see Table 14):

Oxidation reactions

(1)

(2)

Crack reaction

(3)

Water gas reaction

(4)

Shift reaction

(5)

Methane reforming

page 24 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
(6)

Table 14. Equations and reaction enthalpies for the most frequent reactions occurring during catalytic steam reforming and their technical conditions [78]

H at Technical reaction conditions

101.3 kPa,

298 K, kJ/mol Start of reaction at t, End of reaction at t, ° Reaction pressure, Heat of reaction H,
°C C MPa kJ/mol

Reforming reactions
CH4+ H2O CO + 3 H2 (7) 206.36 500 870 2.0 234.66

CH4+ 2 H2O CO2+ 4 H2 (8) 165.13

C2H6+ H2O CO + CH4+ 2 H2 (9) 141.22 500 870 2.0 166.56

C2H6+ 2 H2O 2 CO + 5H2 (10) 347.50 500 870 2.0 131.68

C3H8+ H2O C2H6+ CO + 2 H2 (11) 150.68 500 870 2.0 179.24

C3H8+ 2 H2O CH4+ 2 CO + 4 H2 291.99 500 870 2.0 335.10

(12)
C3H8+ 3 H2O 3CO + 7H2 (13) 498.36 500 870 2.0 563.64

Shift reaction
CO + H2O CO2+ H2 (14) –41.16 220 250 2.0 –37.19

350 420 2.0 –36.48

Oxidation reaction
C + O2 CO2 (15) –196.89

2 CO + O2 2 CO2 (16) –566.37 800 1000 2.0 –542.58

2 H2+ O2 2 H2O (17) –484.17 800 1000 2.0 –480.06

CH4+ 2 O2 2 H2O + CO2 (18) –803.04 300 1400 0.1 –646.83

2 C2H6+ 7 O2 6 H2O + 4 CO2 (19) –2857.6 300 1400 0.1 –2329.2

C3H8+ 5 O2 4 H2O + 3 CO2 (20) –2045.4 300 1400 0.1 –1673.9

Soot formation reactions

CH4 C + 2 H2 (21) 74.9

2 CO C + CO2 (22) –172.54

Carbonyl formation
Ni + 4 CO Ni(CO)4 (23) –159.50

Fe + 5 CO Fe(CO)5 (24) –233.20

page 25 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The reactions of the hydrocarbon feedstock with substoichiometric amounts of pure oxygen (oxygen content > 95 %) follow Equation (1). Other reactions, which take place simultaneously or immediately
following, are the water gas reaction (Eq. 4), the CO shift (Eq. 5) and the methane formation or decomposition reaction (Eq. 6).

The reactions take place primarily in the flame and in the empty brick-lined reactor. The residence time is usually <5 s, the temperature 1250 – 1500 °C. Pressure is often adjusted to accomodate
subsequent processes, 3 – 12 MPa being the usual range.

The subsequent formation of hydrogen by the water – gas shift reaction (Eq. 5, see also Section Catalytic Gas Purification), is of equal importance for hydrogen production as is the immediate formation of
hydrogen (Eq. 1). Carbon monoxide is thereby shifted to hydrogen to the highest degree possible. The shift reaction occurs simultaneously to the oxidation-gasification reactions at temperatures > 1200 °C
without the presence of catalysts. Higher hydrogen yields, however, can be achieved by attaining equilibrium at 200 – 500 °C in the presence of steam and suitable catalysts.

Two process routes have been applied to perform the cooling down, gas purification and shift reaction:

1. The classical sequence of the three aforementioned process steps calls for a two-stage physical scrubbing process (e.g., the Rectisol wash, see Section Scrubbing Processes). The process gas
from the partial oxidation reactor is cooled in a waste-heat boiler thus producing high pressure steam. Soot and ash particles, as well as water-soluble components, are removed with a water
quench at temperatures between 300 °C and ambient. Hydrogen sulfide is removed (e.g., at – 50 °C) in the first stage of a Rectisol absorption unit. The synthesis gas is then heated again to 300 °C
and steam is added. The shift reaction occurs in a series of adiabatic reactors with intermediate cooling and decreasing inlet temperatures at each stage. After subsequent carbon dioxide removal in
the second stage of the absorber the carbon monoxide content can be as low as 0.5 vol %. The catalyst, especially the one used at the lowest temperature, is extermely susceptible to sulfur
poisoning (see Section Adsorption Processes).
2. Since sulfur-tolerant shift catalysts have been developed (see Section Adsorption Processes), an alternative process has become possible. Here the high water content required in the gas for the
CO-shift reaction is achieved by direct quenching of water into the raw gas. The shift reaction takes place in a series of adiabatic reactors without prior removal of sulfur. It has been proposed to
replace them by one temperature-controlled, quasi-isothermic reactor [68]. The sulfur compounds hydrogen sulfide, carbonyl sulfide, and carbon dioxide are removed in a single stage Rectisol
wash, which in its regeneration section can produce two separate fractions, namely a carbon dioxide and a hydrogen sulfide – carbon dioxide mixture. The CO content of the purified gas is
<0.5 vol %.

4.1.2.3. Examples of Industrial Scale Hydrogen Production (by Partial Oxidation of Hydrocarbons)
Production of Hydrogen-Rich Ammonia Synthesis Gas. One example is the production of hydrogen-rich ammonia synthesis gas by partial oxidation of high-sulfur-content vacuum residue (
Ammonia).

Starting from 44 t/h vacuum residue with a sulfur content of up to 4.5 wt %, 116 000 m3 (STP)/h of 96 vol % pure hydrogen are generated. Hydrogen is further purified using a liquid nitrogen scrubber (
Gas Production – Liquid Nitrogen Wash Process), whereby 148 900 m3 (STP)/h of ammonia synthesis gas with a composition H2 = 75 mol % and N2 = 25 mol %, is obtained. 56.25 t/h of ammonia can
be produced from this synthesis gas.

Hydrogen from Residues of Upgrading Operations in Refineries [69], [70]. After optimization of many technical, environmental and economic parameters the process sequence in Figure 18 consists of
the following steps:

1) Air Separation Molecular sieve unit (because of the high air pollution at the site)
External compression of oxygen to 7.5 MPa
2) Gasification Texaco combined reactor in two parallel trains (special arrangements to cope with high ash concentrations in the feedstock)
3) Soot extraction 100 % soot recycling, using the Texaco naphtha soot extraction system No special extraction oil required
Vertical decanter
4) Carbon monoxide 3 beds with cobalt – molybdenum sulfide catalyst with a residual carbon monoxide concentration of 0.5 vol %. Waste-heat recovery
converter for steam generation and other uses within the plant
5) Sour gas absorber Single-stage selective Rectisol absorber, two parallel absorbing columns for converted and carbon monoxide-rich feed gas
Purified streams of hydrogen, carbon monoxide, methanol synthesis gas and Claus fraction containing 70 % hydrogen sulfide
6) Methanation Standard version with nickel catalyst and integrated zinc oxide bed.
7) Carbon monoxide cold- Single column apparatus with internal carbon monoxide recycle (cold-box = containment for low-temperature process)
box

Figure 18. Gasification of refinery residue for the production of hydrogen, methanol synthesis gas, and carbon monoxide (Linde, Federal Republic of Germany)

page 26 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Material balance: The numbers denote the process streams in the above Figure

1 Feedstock : Bitumen, m = 33.3 t/h, p = 0.4 MPa, t = 150 °C, density (25 °C) = 1.056, viscosity (150 °C) = 8 000 Pa sComposition: 83.02 wt % carbon, 10.19 wt % hydrogen, 0.41 wt % nitrogen, 6.31 wt % sulfur, 45 ppm
nickel (max 75), 150 ppm vanadium (max 250), 1000 ppm ash (max 1500)

Gas balance:

Mol, % 2 3 4 5 6 7 8 9

H2 19.52 38.40 0.02 48.22 98.43 67.86 0.03

N2 0.1 0.05 0.10 4.26 0.12 0.12 0.10 0.20
CO 21.61 41.61 51.23 26.91 98.74
Ar 0.4 0.06 0.11 0.14 0.14 0.11 0.30
CH4 0.13 0.25 0.29 1.17 0.22 0.73
CO2 2.12 3.98 21.12 4.80
H2 S 0.64 1.25 73.30
COS 0.03 0.06 1.3
O2 99.5
H2 O 55.84 14.24 0.14

total m3(STP)/h 24 740 159 900 29 420 1 954 11 756 59 300 26 500 4 530
p, MPa 7.3 6.0 5.9 3.0 5.7 5.0 5.0 3.4
t, °C 120 240 170 40 –50 40 40 38

Figure 18 shows the overall process layout.

4.1.3. Catalytic Reforming of Hydrocarbons

A comprehensive treatment of catalytic reforming is given elsewhere (see Gas Production – Steam Reforming of Natural Gas and Other Hydrocarbons). Here a brief outline of the most important
industrial routes for hydrogen production is given.

4.1.3.1. Feedstocks
Hydrocarbons react with steam in an endothermic reaction to form carbon monoxide and hydrogen (Eq. 7, Table 14). The most important feedstock for the catalytic steam reforming process is natural gas.
Other feedstocks are associated gas, propane, butane, liquefied petroleum gas and some naphtha fractions. The choice is usually made on the availability and the price of the raw material. Preference is
given to lighter feedstock, if the price on a heating value basis is comparable. The reason for this is the better operability of plants using lighter feedstocks as well as the easier pretreatment operations,
and thus the lower investment costs for comparable plants.

The chemical process is described by

(25)

The shift reaction, where carbon monoxide, formed according to (25), reacts with excess water, occurs concurrently to this reaction. Additional hydrogen is formed by this exothermic process:

(14)

Standard enthalpies of some important reactions occurring during steam reforming are given in Table 14. The heat of reaction of the overall endothermic process is provided by indirect heating. In
autothermal reactors (e.g., secondary reformers) the oxidation reaction (see Table 14) supplies the necessary energy. This reaction takes place simultaneously or in advance of the reforming reaction.

4.1.3.2. Tubular Reformer

page 27 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The equilibrium of the hydrocarbon steam reaction (7) and the water-gas shift reaction (14) dictates the conditions that must be met to achieve maximum hydrogen yields:

high temperature at the end of the reactor

high excess of steam
low pressure

These conditions can be realized best, if the endothermic reforming reaction occurs in a fired tubular reactor at high temperature (800 – 900 °C) and low pressure. The process pressure varies between
1.5 and 3.0 MPa. The exact choice of the process conditions depends on reaction kinetics, mechanical and economical terms.

The amount of steam added to the hydrocarbon feedstock is quantified by the molar steam-to-carbon ratio (S/C = ratio of S moles of H2O to C moles of carbon contained in the hydrocarbon feedstock. If
present, carbon monoxide and sometimes also carbon dioxide, is calculated with the C moles). The S/C ratio has values from 2.5 to 6, depending on the feedstock and subject to process optimization
considerations. The principals of the optimization of a steam reformer plant for the production of hydrogen are discussed in more detail in Section Steam Reforming Plant.

The next step is the shift reaction at ca. 340 – 450 °C (high-temperature shift). In certain cases this is followed by a second shift reaction at 200 – 250 °C (low-temperature shift).

Catalyst. In commercial plants steam reforming and the concurrent shift reaction are catalyzed by supported nickel catalysts. The catalyst contains 15 – 25 wt % nickel oxide on a mineral carrier. Carrier
materials are -alumina, aluminosilicates, cement, and magnesia. Before start-up, nickel oxide must be reduced to metallic nickel. This is done preferably with hydrogen but also with natural gas or even
with the feedgas itself.

With certain types of catalyst uranium oxide and chromium oxide are used as a promoter. This is reported to give a higher resistance to catalyst poisoning by sulfur components and a lower tendency to
form carbon deposits (according to Eq. 21 in Table 14). For special applications such as low-temperature reforming, noble metal catalysts have been developed [71]. They have not been used
commercially so far.

For the feedstocks LPG and naphtha, nickel catalysts with alkaline carriers [72], [73] or alkaline-free catalysts with magnesium oxide as additive [74] can be used. Both types of catalyst are less active than
the conventional nickel catalyst. Therefore, a less rapid decomposition of the hydrocarbons is achieved. At the same time, the reaction of water with any carbon formed (Eq. 4) is catalyzed. Catalysts with
alkaline carriers are mostly used for the upper part of the reactor tube only. The second part of the reactor is filled with more active nickel catalyst. Whereas the first catalyst gradually decomposes the
higher hydrocarbons, the active catalyst is more suitable for the methane-rich fractions formed in the upper part. The aforementioned combination of catalyst can also be employed with natural gas
feedstock, thereby offering the opportunity to switch to LPG or naphtha in case of natural gas shortages.

Required properties of the catalyst carriers are high specific area, low pressure drop and high mechanical resistance at temperatures up to 1000 °C. The catalysts are usually in the form of rings (e.g.,
outer diameter 16 mm, height 16 mm, inner diameter 8 mm). Other forms, such as saddles, stars, and spoked wheels are also commercially available.

The main catalyst poison in steam reforming plants is sulfur, which is present in most feedstocks. Concentrations as low as 0.1 ppm form a deactivating layer on the catalyst [74]. To some extent activity
loss of a poisoned catalyst can be offset by raising the reaction temperature. This helps to reconvert the inactive nickel sulfide to active nickel sites.

Worse than small constant sulfur concentrations are sulfur peaks in the feedstock. They lead to permanent deactivation of the catalyst and to mechanical problems because of carbon deposits.

Other catalyst poisons are chlorine and other halogen compounds as well as heavy metals, especially lead, arsenic, and vanadium.

Nickel-free catalysts have been proposed for feedstocks heavier than naphtha [75], [76]. These catalysts consist mostly of strontium, aluminum and calcium oxides. They seem to be satisfactorily resistant
to coking and work with high-sulfur feedstocks. However, with these catalysts even at reaction temperatures as high as 1000 °C the methane concentration in the exit gas is so high, that a secondary
reformer has to be installed [77].

Equilibrium Conditions and Kinetics. The reactions in commercial reformers lead to a product composition which is usually very close to the equilibrium of the shift reaction (Eq. 14 in Table 14) and of
the methane reforming reaction (Eq. 7) for design purposes. The resulting gas composition corresponds to an equilibrium temperature, which is usually only 5 – 20 K below the measured actual outlet
temperature.

Higher hydrocarbons than methane are not usually found in the outlet products of a steam reformer. The equilibrium composition can be calculated from equilibrium data given in [73] or [78] as a function
of the temperature, pressure and S/C ratio. For the dependence of the equilibrium composition on pressure and temperature at one (undisclosed) S/C ratio see Gas Production.

A review of the many aspects of intrinsic kinetics of methane steam reforming is presented [79]. For most practical conditions, ranges of S/C and pressures, first-order kinetics can be employed for
methane reforming. For low conversions of methane, which is not typical for industrial use, the concurrent carbon monoxide shift reaction deviates from equilibrium.

For the design of the reforming reactor not only the reaction kinetics within the reformer tube are important, but also the pattern of heat input required for the endothermic hydrocarbon decomposition
reactions.

Heat distribution in conventional reformers is influenced (1) by the geometry of burner arrangements, (2) the pitch of the tubes, (3) the type and length of the flame, (4) the radiation of the flames, flue

page 28 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
gases and refractory walls, (5) size, shape and thickness of the tubes, (6) materials, and (7) whether the tube surfaces are machined. The calculation procedure for a standard type of hydrogen steam
reformer is indicated in Figure 19.

Figure 19. Tube wall temperature and heat flux in a reformer tube

A) Outer tube wall temperature vs. heat flux profile divided into, e.g., eight reformer sections

B) Hatched area: inclined to the right = heat requirements of the reaction; inclined to the left = heat provided by firing

Iteration for illustration purposes not completed

The radiation conditions in the fire box depend on the temperature of the flue gas, the reformer tubes and the refractory walls, as well as their emissivities. A zonal computer program calculates the
interaction among these three factors for the desired heat duty at a given tube skin temperature and results in the firing requirements. The effects of combustion, air or fuel preheat are also considered in
this calculation.

The given tube wall temperature is shown in Figure 19 A. The heat duty on the fire box side is shown for eight reformer section zones as one of the lines in Figure 19 B. The program which calculates the
composition inside the reformer tube allows the heat requirements per length of tube to be calculated at the same time. This heat is supplied through the reformer tube wall and the given tube wall
temperature determines the heat transferred to the inside of the tube. This is shown by the second line in the section zones of Figure 19 B. The aim is thus to find a tube skin temperature profile which will
supply the heat to the process as required, or, in other words, to match the heat flux profiles. When this is accomplished, the temperature profile along the tube and the firing rate have been determined.
The former is important for the mechanical design of the tube, the latter for the utility consumption.

Other factors, such as heat losses, flame lengths, staged burning, flue gas recirculation, position of the flue gas outlet duct, circumferential and other tube temperature asymmetries, also play a role.
Therefore it is essential that results be corroborated by field data and the programs adjusted accordingly to reflect the influence of the above mentioned factors.

The behavior of the radiant section can thus be predicted by the results of these calculations with sufficient accuracy. Literature has been published in this field [80-82]. However, some problems in heat
and mass transfer, influence of pressure and the decomposition of higher hydrocarbons are still not completely resolved.

The utilization of the flue gas leaving the reformer is also of prime importance. An exact prediction of the temperature of the flue gas leaving the radiant zone (i.e., bridgewall temperature) is essential
because it affects the firing duty (i.e. the amount of heat introduced to the reformer by firing), and thus overall performance of the reformer.

Numerous variations in the order of the heat exachanger coils are possible. Normally the excess heat in the flue gas is utilized to produce export steam. Whatever arrangement chosen, the overall thermal
efficiency of reformer and convection zone can normally be expected to be ca. 92 %.

Carbon Deposits. Formation of carbon deposits occurs predominantly because of the decomposition of methane according to Equation (21) in Table 14.

Other reactions leading to soot formation or carbon-coke deposits are cracking of higher hydrocarbons, especially aromatic compounds, and of olefins.

Another reaction leading to carbon deposition, the Boudouard reaction (22) in Table 14

is favored by low temperature and high pressure typically found at the inlet portion of radiant tubes in a steam reformer.

Theoretically at least, carbon laydown can be predicted from the equilibria of Equations (21) and (22) if other carbon monoxide forming reactions such as the steam reforming (4) and the shift reaction (14),
are also taken into account (see Table 14).

Since the early investigations of DENT [83] much has been published on the kinetics and conditions of carbon laydown on commercial and experimental catalysts in plants and in the laboratory [84-86]. As
a result, several factors influencing soot formation in commercial steam reforming processes have been identified:

Choice of Catalyst. Acid sites on the catalyst and its support, e.g., nickel on silica, catalyze carbon formation. Alkaline supports and a spinel structure of the catalyst permit operation close to the soot limits
of equilibrium conditions.

Feedstock Selection and Pretreatment. Olefins and aromatics lead to carbon deposits and catalyst deactivation even in regions which are carbon-free according to equilibrium conditions. Therefore, if the

page 29 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
presence of these compounds cannot be avoided, a recycle of product hydrogen is often used to hydrogenate the unsaturated hydrocarbons. The hydrogen recycle ratio depends on the amount of
unsaturated hydrocarbons expected and will often be ca. 3 – 5 mol % of the carbon content of the feedstock. Hydrogenation is usually performed simultaneously to the hydrogenation of sulfur containing
compounds and the same type of catalyst (a sulfided cobalt or molybdenum catalyst) can be employed.

Kinetic Effects. High space velocities in critical regions also help to avoid soot formation.

Steam : Carbon (S/C) Ratio. Proper choice of the S/C ratio in steam reforming is imporant. Carbon formation is avoided if high S/C ratios are chosen because the equilibrium of the feed gas is shifted from
the carbon regions. The S/C ratio also influences the hydrogen : carbon monoxide ratio of the product gas. This is important, if synthesis gas, not hydrogen, is to be produced. The shift reaction, which
follows the reforming stage in hydrogen production plants and in many other applications, is also influenced by the S/C ratio. A high S/C ratio favors hydrogen production. At low S/C ratios, the danger of
hydrocarbon formation exists with subsequent carbon deposits in the high-temperature shift reactor. In the worst case reduction of the iron oxide catalyst may even occur.

Finally, economic evaluations indicate that the optimum S/C ratio tends to be low. Figure 20 shows that the feedstock consumption is slightly less at low S/C ratios. Because investment cost also declines
with smaller gas flows, lower values of S/C are thus favored, although the aforementioned restrictions must be taken into account.

Figure 20. Consumption of natural gas in a modern hydrogen plant with PSA unit as a function of steam – carbon ratio

a) Total consumption of natural gas; b) Portion of natural gas used as process gas; c) Portion of natural gas used as fuel gas in addition to PSA off-gas [89]

Common values for natural gas operation are S : C 2.5 – 3, and for LPG and naphtha feedstocks S : C 3 – 5. If the desired product is not hydrogen, but carbon monoxide, the S/C ratio should be
substantially lower.

Reformer Types. The aim of a reformer design for the production of hydrogen is to obtain a maximum temperature at the exit of the reformer tube and thus the maximum possible degree of reforming,
using a minimum amount of fuel.

The fuel utilization is optimized, if the heat delivered to each section of the reformer tube corresponds to the heat requirements of the reaction in the tube. Various types of furnace and the geometry of the
tubes and the burners in the furnace are shown in Figure 21 [87], [88].

Figure 21. Types of steam reformer for hydrogen production; furnace and convection zone configurations

Cylindrical Furnaces: A) Small plants, pilot plants; B) Standard configuration; C) Tube/burner configuration for plants with higher capacity (up to 5 000 m3 (STP)/h = 60 mol/s H2); Process gas and flames (flue gas) usually
cocurrent

Box Furnaces: D) Standard configuration; E) Bottom fired reformer; F) Side fired reformer; G) Terrace wall reformer; H) Combined configuration;

Convection Zone: I) Standard configuration, horizontal at grade; usually cocurrent flow of process gas and flames; J) Vertical configuration, downflow; K) Vertical configuration, upflow; L) Asymmetrical configuration; M)
Central configuration;

a) Catalyst filled reformer tube; b) Burner; c) Radiation zone; d) Convection section; e) Fluegas duct; f) Inlet gas distribution system; g) Outlet gas tube (pig tail); h) Collector; i) Flue gas blower; j) Stack

For small reformer plants with a hydrogen capacity of up to 1000 m3 (STP)/h (12.5 mol/s) cylindrical furnaces are often used. These furnaces require the least expenditure with regard to steel construction
and insulation materials and can be prefabricated in the workshop. The plant size is then limited by the maximum transport dimensions.

Figure 21 C shows an arrangement, which accomodates a large number of tubes and allows a hydrogen production of up to 5000 m3 (STP)/h (60 mol/s). Because of the varying distances of the tubes to
wall and burner (see A and B) and because of the varying circumferential heat flux of each tube (C) an even heat flux distribution cannot be attained. This leads to designs with lower overall heat flux and
shorter tube lengths for this type of cylindrical furnace.

For larger hydrogen plants box furnaces are well-proven. Often they can be prefabricated, so that a certain degree of preassembly is possible. The various types of box furnaces mainly differ in the
arrangement of the burners. For hydrogen plants the usual configuration is the top-fired furnace (D), but bottom-fired (E), sidewall-fired (F) and combined types (G, H) are also in use. The box furnace
allows a more even heat load distribution to each tube. The heat flux can therefore be increased.

Because the radiation is evenly distributed around the tube, the mechanical stress is minimized and larger tube diameters (up to 0.18 m outer diameter) can be used. The total number of tubes per

page 30 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
quantity of hydrogen produced can thus be decreased. This reduces furnace volume, piping and the number of burners.

In almost all designs the process gas flows through the tubes from the top to the bottom. The hot outlet gas is usually gathered in a manifold system, which may or may not be lined, and led to the process
gas boiler. This equipment is best placed at ground level. Three different arrangements are possible for the combination of process gas – flue gas flow.

Most modern reformers are top-fired reformers with cocurrent flow. Figure 22 A shows the temperature profile of the outer tube wall and the gas temperature in the tube interior. The high temperature
attained rapidly at the beginning of the tube allows a high reaction rate over a large portion of the tube.

Figure 22. Temperature profiles in different steam reformer designs

A) Cocurrent flow; B) Sidewall-fired; C) Countercurrent flow

The upper curve shows the approximate flue gas temperature, the lower curve the temperature of the process gas, with an adjusted inlet temperature of 550 °C and exit temperature of 800 °C. The curve between shows the
skin temperature of the tube wall, important for the mechanical design of the reformer tubes

Sidewall-fired furnaces exhibit a very even flue gas temperature profile and allow a certain amount of freedom to adjust the tube wall temperature (see Figure 22 B).

The flue gas exit temperature with a countercurrent system is particularly low (see Figure 22 C), which indicates that a large fraction of the heat content of the flue gas has been utilized. This type is used
preferentially whenever a furnace with the convection zone at the top, using natural draft, and with a low degree of further utilization of the heat content of the flue gas is envisaged.

Further critical items of steam reformers are:

1. The support of the tubes to minimize the load on the distribution system and to avoid bowing of the tubes.
2. The preheated gas mixture (ca. 500 °C) enters the reformer via the inlet distribution system. The distribution system must be designed to take up the vertical expansion of the reformer tubes (for
example, a 14 m tube may expand by as much as 0.25 m under operating conditions).
3. The horizontal thermal expansion of the outlet collector must be compensated for by the connecting pipework, the so-called pigtails.
4. Equal gas distribution over the tubes is attained by proper hydraulic design of the inlet and outlet collectors coupled with uniform loading of the catalyst in the reformer tubes.
5. The ratio of tubes : burners in large top-fired reformers is ca. 4. Smaller numbers of burners have the disadvantage of uneven heat distribution to the tubes but the advantage of lower investments
because of the reduction in piping and ducting.
6. Design of burner, insulation, flue gas duct and reformer geometry. For details see [90].

Only a fraction of the heat supplied to the reformer furnace is used for reforming. A large amount of the heat in the flue gas is carried into the convection zone, where it is used to preheat the reformer feed,
to produce and to superheat steam and, depending on circumstances at the site, maybe to preheat the combustion air.

The convection section can be installed as shown in Figure 21 (I – M). In most large hydrogen plants a flue gas fan must be installed, to induce the slight vacuum necessary for the draft burners and to
compensate for the pressure drop over the heat exchanger coils in the convection section. For small plants, the convection section is sometimes installed on top of the reformer. This reduces the power
requirements of the fan or allows design for natural draft.

The design of the convection section itself depends on the overall energy balance of the plant (see below) and the way the energy is to be utilized within the plant. This is influenced by the use of steam or
electric drives or of gas turbines, etc. [91].

Energy Balance. The energy required for reforming in the furnace is (1) utilized as reforming energy and (2) in the convection section for preheating and steam production. In addition, energy losses such
as insulation losses and the heat content of the flue gas lost through the stack must be considered. The efficiency factor for the steam reformer can be defined in several ways, for example:

Firebox Efficiency. The firebox efficiency is the ratio of energy necessary for reforming (QR) to the energy of fuel combustion (QLHV) (usually expressed in terms of the lower heating value of the fuel). If air
or fuel are preheated it is necessary to consider the amount of energy brought in by preheated air QA, and/or preheated fuel QF to give:

Firebox efficiencies lie in the range 0.45 – 0.55. This value only gives the relationship of the distribution of the energy between radiant and convection section, not, however, the degree of energy
utilization.

Furnace Efficiency. The energy utilization in firebox and convection section is characterized by the overall furnace efficiency. This is the ratio of the sum of the utilized energy for reforming QR, for steam
production QS and for any further energy utilization in the convection zone (e.g., feed preheating and superheating, export steam superheating, economizer QC) to the total energy supplied to the firebox

page 31 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
QLHV + QA + QF:

The efficiency thus defined usually lies between 0.85 and 0.93 for hydrogen plants. The insulation losses are ca. 300 – 400 W/m2 in the furnace section and ca. 600 – 700 W/m2 in the convection zone
which, depending on the size of the plant, amounts to ca. 1 – 5 % of the fired energy. The flue gas losses also vary with the size of the plant and the energy utilization and amount to 6 – 10 %.

Energy management plays a vital role in steam reformer design. The efficiency and the economy of the process are determined by the design variables, e.g., feedstock and fuel costs, bonus for export
steam, use of electric or steam drives, S/C ratio, product gas pressure, etc. Suggestions for optimizing steam reformer plants can be found in the literature [92-94]. One typical example is discussed below.

4.1.3.3. Steam Reforming Plant

Figure 23 shows a simplified process flow diagram of a typical hydrogen plant. As in the majority of such plants, the feedstock is natural gas. Table 15 summarizes the process data. For the production of
pure hydrogen, the following process stages are necessary:

Table 15. Mass balance of a typical steam reformer plant for the production of 5000 m3 (STP)/h hydrogen. Feedstock: natural gas; byproduct: steam (3.0 MPa, 255 °C) 4700 kg/h (the numbers
in Figure 23 denote the balance points)

Feed, 1 Before desulfurization, 2 Reformer inlet, 3 Reformer outlet, 4 Pressure swing adsorption, 5 H2 product, 6

H2, vol % 0 4.8 1.1 51.9 74.6 99.99

N2, vol % 1 1 0.3 0.2 0.25 0.01
CO, vol % 0 0 0 10.7 5.3 0
CO2, vol % 0.5 0.5 0.1 5.1 14.8 0
CH4, vol % 95 90.4 26.1 3.8 4.9 0
C2+, vol % 3.5 3.3 1.0 0 0 0
H2O, vol % 0 0 71.4 28.3 0.15 0

Quantity, m3 (STP/h) 2160 2070 7160 10450 8180 5000.5

Temperature, °C 20 390 520 850 20 20
Pressure, MPa 2.0 1.9 1.9 1.7 1.6 1.5

Figure 23. Hydrogen production by steam reforming, gas purification by pressure swing adsorption (Design: German Linde)

a) Desulfurization; b) Feed preheater/superheater; c) Reformer; d) Waste-heat boiler; e) CO shift reactor (HT shift); f) Cooling of raw gas; g) Pressure swing adsorption; h) Off-gas puffer for fuel; i) Convection zone with
steam production, steam superheating, and air preheating

Feed Preheating and Desulfurization. Usually, natural gas contains only small amounts of sulfur compounds, predominantly in the form of hydrogen sulfide. Because of the high sulfur sensitivity of the
catalyst in the reformer (and in the LT shift reactor, if installed) hydrogen sulfide and other sulfur compounds must be removed on a zinc oxide bed (see Section Catalytic Gas Purification). If higher organic
sulfur compounds (mercaptans, thiophene) or carbonyl sulfide in concentrations in the ppm-range are present, the zinc oxide bed alone is not sufficient and the sulfur compounds are converted to
hydrogen sulfide in a hydrogenation stage. The hydrogen necessary is taken from the product stream and amounts to ca. 5 % for natural gas (whereas for naphtha and LPG 10 % and 25 % H2,
respectively, is used). Hydrogen recycle and hydrogenation also lowers the amount of olefins in the feed. The excess hydrogen in the feed helps to keep the reforming catalyst at the top of tube in a
reduced state.

For hydrogenation and desulfurization the inlet gas is preheated to 350 – 400 °C. The gas leaves the purification section with a sulfur content of <1 ppm at essentially the same temperature.

page 32 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Steam Addition, Reaction and Cooling. Steam is added to the inlet feed to give a ratio of 2.5 moles of steam per mole of carbon. This mixture can be further preheated in the convection section to 520 °
C without cracking occurring. Reforming takes place in the reformer tube, the outlet temperature being 850 °C. After the reaction the gas is cooled quickly, thereby producing steam in the process gas
boiler.

Water Gas Shift Reaction and Gas Purification.

High-Temperature Shift Conversion. The gas cooled down to 350 °C in the process gas boiler still contains a fair amount of carbon monoxide which can be converted into hydrogen by the water-gas shift
reaction (Eq. 14, Table 14). An iron-oxide based catalyst is used for this. The shift reaction is usually carried out in a one-stage adiabatic reactor. Since the reaction is exothermic, the process gas
temperature rises to 400 – 450 °C so the conversion is incomplete.

Pressure-Swing Adsorption (PSA). After high-temperature shift conversion, the gas is cooled down to ambient temperature. Water is condensed and removed before the gas is fed to at least three (but up
to twelve) parallel adsorbers, which run in a predetermined cyclic operation. The various purge and decompression gases of the adsorbers are led into a large buffer vessel to compensate for variations in
pressure and gas concentrations and are subsequently fired in the reformer along with a small amount of fuel gas for trimming purposes.

Classical Process. In an alternative route after the HT shift, the so-called classical process, the gas is cooled to ca. 220 °C and fed into a low-temperature shift reactor; followed by a carbon dioxide
scrubber (monoethanolamine or hot potash, see Section Scrubbing Processes). The remaining impurities in the hydrogen, traces of carbon monoxide and carbon dioxide, are removed by a catalytic
methanation step. Using this process, a hydrogen purity of 97 – 99 % can be attained (not taking the inert gases into account). Both process routes have been compared in the literature [92], [95], [96]. For
pure hydrogen plants the PSA version has become standard. Plants are built according to the classical route only in special cases in which (1) the feedstock (natural gas, LPG, naphtha) is expensive, (2) a
cheap fuel is available to fire the reformer, (3) carbon dioxide is a desired product, or (4) a hydrogen purity of <99.9 % can be tolerated. For the coproduction of carbon dioxide combinations of both
processes have also been built.

Heating, Steam Production, Heat Management. In Figure 23 the fuel gas is essentially the purge gas of the PSA plant. In addition, small amounts of the feed gas are fired. The combustion air is
preheated in the convection section up to 500 °C. The process gas boiler and the flue gas boiler produce the required process steam. Depending on the design conditions of the plant, export steam is
available for other users. From the point of view of the heat balance, a hydrogen plant can be optimized to be self-supporting, for maximum production of export steam, for minimum hydrocarbon
consumption, etc.

4.1.3.4. Special Designs

The general design of tubular steam reformers has changed little in the past 20 years. In particular new materials for reformer tubes, heat exchangers and insulation as well as better and fewer burners
and a better heat management have led to a highly sophisticated type of plant with an excellent efficiency and reliability.

Nevertheless, many new designs have been developed although their economic importance is so far very limited. The main areas for improvement of the reformer are:

improvement of the heat transfer for the endothermic reaction,

reduction of the plant size,
reduction of the maximum temperature (at the burner and in the flue gas).

Pressurized Fireboxes. The pressure in the fireboxes of conventional tubular reformers is atmospheric or slightly below. The reasons for this are:

1. Simple construction and safe operation of the reformer furnace,

2. Simple sealing of large numbers of reformer tubes,
3. Purge gas (e.g., from the PSA) can be fired at the low pressure,
4. Induced draft burners or burners using combustion air at low pressures can be used.

The use of higher pressure in the firebox leads to higher heat transfer coefficients from the flue gas to the reformer tube by convection. Figure 24 shows the construction of such a reformer and a similar
heat exchanger concept [97], [98].

Figure 24. Pressurized and heat exchange type reformers

A) KTI (Kinetics Technology International BV), United Technology Corp. version [99]; B) Haldor Topsøe AS, Denmark [97], [98]

a) Gas inlet; b) Reaction zone (circular zone[s]); c) Reformed gas outlet; d) Heating gas inlet; e) Combustion air; f) Burner; g) Radiant section; h) Section with forced convection; i) Flue gas outlet; j) Insulation

The catalyst is contained in the concentric space between the tubes so that the thermal expansion caused by start-up and shut-down of the reformer tube does not lead to sealing problems. Pressure on

page 33 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
the flue gas side is 300 – 500 kPa [99]. One disadvantage is that both fuel gas and combustion air must be compressed. Plants with a hydrogen capacity of 120 – 5000 m3 (STP)/h (1.5 – 60 mol/s) have
been built.

Helium-Heated Reformers. To utilize the heat of the high-temperature nuclear reactor for the production of hydrogen from methane, a reformer with helium heating has been suggested. Since the helium
temperature at the present stage of development is 900 °C, only a small temperature difference is available for the heat transfer. The reformer is similar to a conventional tube bundle reactor [100],
however, the construction is very complex for safety and other considerations. Principally speaking, the reactor and heat exchanger could be used for other endothermic reactions. Up to now, only two
prototypes with a capacity of 0.8 and 10 MW have been in operation. The helium is heated electrically in these plants.

Electrically Heated Reformers. For the production of small quantities of hydrogen electrically heated reformers have been suggested [101]. The hot product gas is used to heat the feed gas and produce
part of the steam required. The energy requirement of ca. 300 kJ per mole of hydrogen is less than for a conventional steam reformer or for the electrolysis, whereby 80 kJ are brought in electrically and
220 kJ correspond to the lower heating value of the hydrocarbon used. However, it should be pointed out, that fossil and electrical energy can be compared on an equivalent basis only under special
conditions.

Steam Reforming with Integrated Hydrogen Separation. High-temperature-resistant membranes have been integrated in a tubular reformer. Thus, hydrogen is withdrawn from the reaction mixture as it
is formed. The chemical equilibrium is permanently shifted towards a higher feedstock conversion. For this reason either lower temperatures can be used or higher yields can be obtained. Palladium foil
(100 µm) on a porous cylinder inserted into the reformer tube has been suggested as a suitable membrane. A small laboratory-scale plant has been built and successfully tested [102].

4.1.3.5. Autothermal Reactors

In autothermal reactors (secondary reformers) the energy for the production of carbon monoxide and hydrogen is produced by catalytic partial oxidation of part of the feedstock.

This type of reactor is employed:

1. to achieve high equilibrium temperatures (up to 1000 °C) thus minimizing methane concentration,
2. to enable use of elevated pressures (up to 5 MPa) together with high equilibrium temperatures (900 – 1000 °C) not feasible in fired steam reformer tubes,
3. to introduce nitrogen for the production of a stoichiometric syngas mixture for the ammonia synthesis.

In the first two cases the reactor is operated with pure oxygen (and steam). In the third case air or turbine off-gases are used. Figure 25 depicts three different types of autothermal reactor.

Figure 25. Designs of secondary (autothermal) reformers

A) Water-cooled jacket reactor with burner for methane-rich gases; B) Secondary reformer with toric oxygen or air supply device [90]; C) Secondary reformer with water-cooled tip without water jacket [103]

a) Raw gas inlet; b) Reformed gas outlet; c) Air, oxygen, and steam inlet; d) Catalyst; e) High-temperature catalyst; f) Inert material; g) Internal insulation; h) Multiple layers insulation; i) Burner; j) Water jacket

With the reactors shown in Figure 25 A and B no reaction occurs before the gas enters the catalyst bed. Reactor C features a burner, thus thermal oxidation takes place in the empty space above the
catalyst bed. In all cases the highest temperatures occur at the top of the reactor because of the catalytic (A, B) or thermal oxidation (C) of hydrogen and carbon monoxide. The endothermic reforming of
methane and other hydrocarbons according to the steam – methane equilibrium (Eq. 7 Table 14) is carried out in the lower part of the reactors. Nickel catalysts are used, often in the form of impregnated
alumina lumps, which have a high resistance to fusion and thermal shock.

The waste-heat boiler is mounted immediately after the reactor, usually without a transfer line. Synthesis gas cools down rapidly in the boiler, thus avoiding recombination of carbon monoxide and
hydrogen.

The direct application of the catalytic autothermal reactor without preceding decomposition of at least part of the hydrocarbons in a primary reformer has been studied in experimental plants [104]. Several
layers of catalyst with increasing activities are used. Soot formation and high-peak temperatures are avoided by using a suitable catalyst at high space velocities, even though benzene feedstock, which is
particularly liable to soot formation at a low S/C ratio is used. The experimental results are depicted in Figure 26. Because of the high reaction temperature the secondary reformers can stand a higher
sulfur concentration in the feedstock. This fact has led to several tests with high-sulfur feedstock. However, commercial application has not yet been realized [76], [77].

Figure 26. Autothermal reforming of benzene on Ni-catalysts of increasing activity [104]

Pressure 101.3 kPa; inlet temperature 565 °C; molar air/carbon ratio 1.8; molar steam/carbon ratio 0.6

page 34 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

a) Space velocity = 5512 m3 (STP) h–1 m–3, carbon formation; b) Space velocity = 11 023 m3 (STP) h–1 m–3, carbon formation; c) Space velocity = 16 535 m3 (STP) h–1 m–3, carbon formation; d) Space
velocity = 22 047 m3 (STP) h–1 m–3, no carbon formation

4.1.4. Refinery Processes

In refineries some processes require hydrogen, while others are a potential source of hydrogen. With the increasing quality demands for refinery products and with the increasing use of bottom-of-the-
barrel products, the transfer of hydrogen within the refinery is increasing. Hydrogen is produced by steam reforming of hydrocarbons and by recovery from hydrogen-rich gases and refinery offgases,
formerly used as heating gases, (Chap. Purification of Hydrogen).

Production and recovery of hydrogen from the byproducts of the following types of refinery processes are possible:

cokers and visbreakers

(thermal cracking processes)
catalytic crackers
catalytic reformers

For a detailed description of the processes see Oil Refining. Here only facts relating to the hydrogen production are discussed.

Cokers and Visbreakers. In cokers and visbreakers heavy crude and residues are converted to petroleum coke, oil, light hydrocarbons (benzene, naphtha, LPG) and gas. Depending on the process,
hydrogen is present in a wide range of concentrations. Since (like the coal coking process) petroleum coking processes need gas for heating purposes, adsorption processes (PSA) are best suited to
recover the hydrogen because they feature a very clean hydrogen product and an off-gas suitable as fuel (see Section Coke Oven Gas ).

If more hydrogen is required in the refinery, it can be derived from the fuel gas and from the LPG fraction by steam reforming. In times of oil shortage processes have been suggested to convert the excess
coke into synthesis gas and hydrogen by coke gasification [105], [106].

Catalytic Crackers. Catalytic cracking is the most important process step for the production of light products from gas oil and increasingly from vacuum gas oil and residues.

In catalytic cracking the molecular mass of the main fraction of the feed is lowered, while another part is converted to coke. The coke is deposited on the hot catalyst. The catalyst is regenerated in one or
two stages by burning the coke off with air, which also provides the energy for the endothermic cracking process. The cracking processes occurring can be categorized as follows [107]:

paraffins and naphthenes are cracked to olefins and to alkanes with shorter chain length,
monoaromatic compounds are dealkylated without ring cleavage,
diaromatics and polyaromatics are dealkylated and converted to coke.

Hydrogen is formed only in the last type of reaction, whereas the first two reactions produce light hydrocarbons and therefore need a certain amount of hydrogen. Figure 27 shows for various feedstocks,
that an increase in coke production goes hand in hand with an increase in hydrogen production. Thus, a catalytic cracker can be operated in such a manner that enough hydrogen for subsequent
processes is formed.

Figure 27. Hydrogen production vs. catalytic coke yield using amorphous catalyst for different feedstocks [108]; West Texas Permian; Light Iranian; Kuwait; Residuum

page 35 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Figure 28 shows how the hydrogen production from catalytic cracking can be linked to other hydrogen consuming processes. After desulfurization by absorption processes (see Section Scrubbing
Processes) the off gas is usually purified and fractionated in a cryogenic unit (see Section Low-Temperature Processes).

Figure 28. Hydrogen production by catalytic cracking with subsequent use in hydro-desulfurization processes

——H2-rich gases; —— H2 purified

Catalytic Reformers. Naphtha fractions are reformed to improve the quality of gasoline. The most important reactions occurring during this process are the dehydrogenation of naphthenes to aromatics.
For the dehydrogenation of cyclohexane, e.g., the hydrogen formation is as follows:

This reaction is endothermic and is favored by low pressures. However, because hydrocracking occurs as a concurrent reaction the pressures are chosen between 1 and 3 MPa and the reaction
temperature lies in the range of 300 – 450 °C. The reaction is performed on platinum catalysts, with other metals, e.g., rhenium, as promoters. The feedstock for the reforming process must be carefully
purified from sulfur and nitrogen compounds (both <1 ppm). This uses up a great deal of the hydrogen produced. Table 16 summarizes the results of some reforming processes. The aromatic content of
the feedstock is substantially increased and the hydrogen set free raises the content in the off-gas to 85 %.

Table 16. Hydrogen as a byproduct of catalytic reforming processes. Range of production and composition of the H2-rich gas fraction

Range

Feedstock: naphtha
Boiling range, °C 80 – 200

Paraffins, vol % 75 – 50 – (25)

Naphthenes, vol % 30 – 50

Aromatics, vol % 5 – 20 – (40)

Liquid products
Octane number (95) – 100 – 105

Reformate, wt % 75 – 88

Paraffins, vol % 30 – 20

Naphthenes, vol % 0.5 – 2.5

Aromatics, vol % 65 – 85

Hydrocarbons C2– C4, wt % 10 – 15

Hydrogen, wt % 2–3
Composition of H2-rich gas typical

H2,vol % 75 – 85 82

CH4, vol % 5–9 7

C2, vol % 4–7 5

C3, vol % 2.5 – 5.5 4

C4, vol % 0.5 – 2.5 2

page 36 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

4.1.5. Petrochemical Processes

Steam Cracking for Olefin Production. During the production of ethylene ( Ethylene) and propene ( Propene) large amounts of hydrogen are formed as byproduct. The composition of some
crude gases from steam cracking of ethane, naphtha and hydrogenated vacuum gas oil is given in Table 17.

Table 17. Hydrogen as a byproduct in the production of ethylene in modern short residence time cracking furnaces. Cracking conditions standardized to an ethylene yield of 32 %

Raw gas Ethane Naphtha Gas oil

After cracker and dryer

H2, mol % 33 16 10

CH4, mol % 5 32 30

C2H4, mol % 32 32 32

C2H6, mol % 28 5 6

C3H6, mol % 1 11 16

C3H8, mol % 1 1

Other: C4+, CO,

H2S, mol % 1 3 5

H2-fraction of demethanizer

H2, mol % 80 – 95

CO, mol % 1 – 0.5

CH4, mol % 19 – 4.5

C2, mol % 0.1

Since propane, propene, ethane, and ethylene are fractionated at low temperature, hydrogen is usually separated from methane by condensation of the latter (see Section Low-Temperature Processes).
The hydrogen recovered has a purity of 90 – 96 %.

Acetylene Production. Hydrogen is also produced as byproduct in the production of acetylene by thermal cracking ( Acetylene, Acetylene – Production from Liquid Hydrocarbons (Plasma Arc
Process)). Because of the extreme cracking conditions the hydrogen content in the crack gas is particularly high. The production of 120 000 t/a acetylene and 50 000 t/a ethylene using the Hüls electric arc
process yields 360×106 m3 (STP)/a hydrogen as byproduct. In Table 18 the crude gases of several processes are compared. The acetylenes are recovered from the raw gas by absorption. Pure hydrogen
is recovered from the remaining gas by cryogenic processes, sometimes in combination with adsorption processes.

Table 18. Hydrogen as a byproduct in acetylene production (raw gas after quench and condensation, dry basis)

BASF Process based on Hüls Plasma arc processb

Methane a LPG Naphtha Methane-rich gas

page 37 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Gas composition
H2, vol % 56.5 46.4 42.7 57.6

C2H2, vol % 7.5 8.2 8.8 15.5

CH4, vol % 5.2 5.0 4.9 13.8

Other hydrocarbons, vol % 0.8 1.0 1.2 2.5

CO, vol % 25.8 35.0 37.9 0.6

CO2, O2 inerts, vol % 4.2 4.4 4.5 10.0

a 1460 m3 (STP) H2 per tonne of feed.

b 3300 m3 (STP) H2 per ton of C2H2, 1830 m3 (STP) H2 per tonne of feed

In a further development of the electric arc process for the production of hydrogen, reduction gas, acetylene and ethylene, a plasma process has been suggested. A great number of feedstocks can be
used ranging from natural gas to coal [109], [110]. The hydrocarbons are not cracked directly in the arc but a hydrogen plasma is used as energy carrier. In this way the acetylene yield can be increased
while the hydrogen yield (and the amount of carbon deposits) decrease.

4.2. Electrolysis
Electrolysis has been used for approximately 100 years for hydrogen production. The first large installation was by Norsk Hydro in 1927 in Norway. Further plants were erected by Cominco in Trail, British
Columbia, Canada in 1940 and, from 1945 onwards, some other plants with capacities up to 33 000 m3 (STP)/h of hydrogen. As explained in Section Economic Aspects, the erection of large electrolysis
plants depends on the availability of cheap electricity from hydropower stations. Plants for smaller amounts of hydrogen (50 – 500 m3 (STP)/h) are often used by the industry because they are simple to
operate. Moderate power costs are an additional incentive. A total of ca. 5 % of the world hydrogen production is by means of electrolysis.

4.2.1. Principles
For general principles see Electrochemistry.

If a potential is applied to the electrodes of an electrolysis cell filled with a suitable electrolyte, the following reaction occurs at the electrodes:

1. Cathode

(26)

2. Anode

(27)

3. Cell Reaction

(28)

where l is liquid, g is gaseous, and aq is dissolved in water.

Pure water is not suitable as an electrolyte because of its very low conductivity. Therefore, aqueous solutions of potassium or sodium hydroxide, sodium chloride, hydrochloric acid, etc. are used.
Depending on the electrolyte and the material of the electrodes the reaction at the anode may lead to other products, in particular to sodium hydroxide and chlorine (see Section Hydrogen as Byproduct
from other Electrochemical Processes).

The change in enthalpy for the electrochemical decomposition of water is

page 38 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

(29)

where H is the enthalpy, G is Gibbs free energy, S entropy, T temperature, n number of electrons transferred per formula conversion, F the Faraday constant, and E the electrochemical potential.

The thermodynamic standard data at p = 101.3 kPa and T = 298.15 K are:

Reaction enthalpy =
Reaction entropy =
Gibbs free energy of reaction =
The value of the Faraday constant is

Using these data the required electrical energy becomes:

(30)

where Erev is the ideal (reversible) decomposition potential. Under standard conditions it amounts to:

(31)

Because of irreversible processes on the electrodes and because of the resistance of the electrolyte the actual decomposition potential is always higher than the ideal value.

According to Equation (29) the free energy, G, is smaller than the reaction enthalpy, H, by the amount T S. In the ideal case operation of the cell requires an amount of electrical energy, W = G,
and the addition of heat, Q = T S.

If both amounts of energy are to be supplied in the electrical form, the potential under standard conditions is increased by the thermal potential EQ .

(31a)

Thus, the theoretical minimum decomposition potential at standard conditions is:

Figure 29 shows the temperature dependence of the energy or the decomposition potential, respectively at a constant pressure of 100 kPa. An increase in temperature lowers the reversible decomposition
potential, while at the same time the thermal fraction of the total energy required increases. This means, that at higher temperatures heat energy instead of electrical energy can be partially used.
Technical aspects of this behavior in regions above the boiling point of water (373 K) will also be considered in more detail in Section New Developments in Electrolytic Processes .

Figure 29. Temperature dependence of hydrogen formation by electrolysis at 101.3 kPa. The energy scale (kJ/mol) shows the standard free energy G0 and the enthalpy H0 of water cleavage. The voltage scale (V)
allows determination of the corresponding theoretical cell voltage

a) Formation with H2 heat output; b) H2 formation with heat input necessary; c) No H2 formation by electrolysis possible

The decomposition potential also depends on pressure. From the pressure dependence of the free energy

the following expression can be derived for the pressure dependence of the decomposition potential of an ideal gas, i.e., neglecting the activity coefficients:

page 39 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

For water at 298 K, taking the activity coefficients into account this equation becomes

so that, on increasing the pressure from 100 to 1000 kPa, an increase of 0.0435 (= 0.0189 · ln 10) V would be expected. However, the fact that the gas produced at the electrodes has a smaller effective
volume leads to a reduced overpotential at the electrodes, so that for electrolysis under increased pressure no increase in the total potential is in fact found.

The potential actually required is the sum of the following contributions:

where U is the cell voltage (V), E1 the theoretical cell voltage (Erev) (V), E2, E3 the anodic and cathodic overvoltage (V) at the phase boundary electrolyte-electrode, and E4 is the voltage drop (V), due to
electrical resistance of the electrolyte system.

The overvoltages E2 and E3 can be influenced by suitable choice of the electrode materials and their surface conditions. The voltage drop in the electrolyte, E4, depends (1) on the conductivity of the
electrolyte, (2) on the permeability of cell diaphragms, (3) on the distance of the electrodes from one another, and (4) on the current density. Figure 30 shows the influence of the current density on the cell
voltage during water electrolysis using potassium hydroxide electrolyte.

Figure 30. Relationships between current density and cell voltage for alkaline water electrolysis

—— conventional electrolysis, 30 wt % KOH, 90 °C; – – – prognosis for advanced conventional electrolysis according to [111], [112]

E1 theoretical reversible cell voltage; E2 anodic overvoltage, difficult to minimize; E3 cathodic overvoltage, can be reduced by suitable catalysts; E4 resistance of the electrolysis system, can be reduced by the use of thin-
walled diaphragms and zero-gap constructions

The efficiency of the electrolysis is related to the minimum voltage according to Equation (29):

Under standard conditions:

Principle Construction of a Water-Electrolysis Cell. Figure 31 shows the principal construction of a cell. Anode and cathode regions are separated by a diaphragm which allows the current to flow, but
is otherwise gas-tight. The cell is filled with electrolyte. Oxygen is formed during the electrolysis at the anode, hydrogen at the cathode. Depending on the arrangement of the electrodes and the
diaphragm, the cells may be designated as unipolar or bipolar cells (see Fig. 32).

Figure 31. Schematic of a water electrolysis cell

a) Anode; b) Diaphragm; c) Cathode; d) Oxygen outlet; e) Anodic electrolyte cycle; f) Anion; g) Cation; h) Cathodic electrolyte cycle; i) Hydrogen outlet; j) Cell wall

page 40 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 32. Different modes of water electrolysis cell according to [117]

a) Cell walls; b) Electrolyte; c) Cathode; d) Anode; e) Hydrogen outlet; f) Oxygen outlet; g) Gas collector; h) Diaphragm; i) Outer cathode; j) Outer anode; k) Bipolar electrode; l) Insulation

The unipolar cells are further differentiated in bell-type (without diaphragms), the usual diaphragm cells, and a bell-diaphragm type. A characteristic of the unipolar cell is that cathode and anode each have
their own separate cell region.

Bipolar cells are characterized by the fact that the metallic separator between two cells connected in series serves as cathode in one of the cells and as anode in the next cell. The arrangement of the
bipolar electrodes is similar to the construction of a filter press, so that they are often named accordingly.

Electrolyte. To minimize the resistance losses of the electrolytes usually 25 – 36 wt % potassium hydroxide solution is used. Other electrolytes are aqueous sodium hydroxide or, for the chlor-alkali
electrolysis, solutions of hydrochloric acid, sodium chloride, etc. Electrolytes which are immobilized in polymers can also be used.

Since the conductivity of conventional electrolytes increases with temperature, electrolysis units usually operate at 70 – 90 °C, just below the boiling point of the aqueous solution. The conductivity of
sodium hydroxide and potassium hydroxide as a function of temperature and electrolyte concentration is shown in Figure 33.

Figure 33. Conductivity of electrolytes used in conventional water electrolysis

a) NaOH, 40 °C; b) NaOH, 60 °C; c) NaOH, 80 °C; d) KOH, 60 °C; e) KOH, 80 °C

Materials. Electrochemical processes place particularly high requirements on corrosion protection. The conventional water electrolysis with potassium hydroxide as electrolyte can be carried out using
carbon steel as a material of construction. Areas particularly subject to attack are coated with plastics or ceramics or are nickel-plated.

The cathode is usually made of steel. To reduce the hydrogen overvoltage E3 the surface texture may be activated and coated with various catalysts. The anode and the electrodes of bipolar cells are
usually made of nickel or nickel-coated steel. Next to platinum, which cannot be used for economic reasons, this material has the lowest overvoltage.

The diaphragm in the classical electrolysis units was initially made of asbestos. The asbestos mats, which separate the electrodes from one another, were reinforced with nickel nets. Because of the
health hazards involved in the use of asbestos, substitute materials are being looked for. Ceramics and hydrophiliated polymers may be the solution to this problem.

4.2.2. Conventional Water Electrolysis

A conventional electrolysis system uses an alkali solution, usually potassium hydroxide, as electrolyte. The electrodes are in unipolar or bipolar arrangements. Hydrogen production takes place either at
ambient or at increased pressure (Commercial plants up to 3 MPa).

Normal-Pressure Electrolysis Units. The advantages of a process at almost ambient pressure are:

1. Low mechanical strains on piping fittings, seals, and instruments,

2. Simple operation, because feed water is not required under pressure, quick start-up, simple maintenance,
3. Safer operation because of avoidance of high-pressure hydrogen, small pressure differences between the evolved gases, and simple blanketing of the electrolysis units and the piping network with
nitrogen under low pressure

Unipolar Stuart cells are offered by a single producer (Electrolyser Inc., Toronto). The particular advantage of this type of cell is its robust nature. Since the cells are arranged in parallel they can be easily
maintained or renewed, if necessary, without interruption of the process.

All other producers use bipolar cells which are mounted in series (filter press). This allows the use of higher voltages (up to 600 V) per unit. A schematic of bipolar electrolysis is shown in Figure 34. Pre-
electrodes, which are placed close to the diaphragm to give the smallest possible resistance between the electrodes, can be clearly seen. The electrodes themselves are perforated or made from nets, so

page 41 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
that the gas bubbles evolve behind the electrodes and do not increase the resistance between the electrodes.

Figure 34. Cross section of a bipolar, atmospheric pressure electrolysis unit with alkaline electrolyte (Brown Boveri)

a) First anode with pre-electrode; b) Diaphragm; c) Pre-electrode of the cathode; d) Cathode for the first cell compartment and anode for the second cell compartment-bipolar middle electrode; e) Cell frame with insulation
for the middle electrode; f) Distribution pipe for electrolyte; g) Inlet pipe for electrolyte; h) Take-off system for electrolyte and dissolved gases; i) Overflow pipe and collecting vessel for electrolyte and H2 (an analogous
system is used for O2); j) Electrolyte filter; k) Circulation pump; l) End plate

The operating conditions for electrolysis units of various producers are given in Table 19 (see also Electrochemistry). Typical operating conditions are electrolyte temperatures of 70 – 90 °C, a cell
voltage of 1.85 – 2.05 V and a current density of 2 – 3 kA/m2. The specific power consumption is between 4 and 5 kWh/m3 (STP).

Table 19. Producer of medium and large sized commercial electrolysis units in alphabetical order. Information from producers and Achema publications in 1988

Producer, Type of Electrolyte Pressure, MPa Current density, A/m2 Cell Voltage, V Commercial status 1988 size of
country electrolysis conc., wt % Temperature, °C Power consumption, kWh/m3 (STP) H2 basic units, m3 (STP) H2/h a

BBC AG bipolar KOH ambient 2000 5 – 300; 4×300

Switzerland filterpress 25 80 2.05/4.9
Davy-Bamag. bipolar KOH ambient 3 – 330
Fed. Rep. of filterpress 4.2 – 4.5
Germany
Electrolyser unipolar KOH ambient 1340 0.5 – 100
Canada tank 28 70 1.8/4.3 – 5.0 c
2500 larger plants
1.9/4.4
Krebs-Kosmo bipolar KOH ambient 1000 – 9000 20 – 200
Fed. Rep. of filterpress 28 75 1.6 – 2/3.9 – 4.8
Germany
Lurgi (Zdanski- bipolar KOH 30 2000 110 – 750
Lonza)
Fed. Rep. of filterpress 25 90 1.86/4.3 – 4.6
Germany
Norsk Hydro b bipolar KOH ambient 1750
Norway filterpress 25 80 1.75/4.1
Oronzio de Nora bipolar KOH ambient 1500 5 – 1000
Italy filterpress 80 1.85 – 1.95/4.6
Teledyne bipolar KOH 0.7 3000 1 – 42
United States filterpress 25 80 2.1 – 2.2/5.5 – 6.1

a Much larger plants have been built by combining the appropriate number of units.
b No up-to-date information available.
c Higher value: standard electrode; lower value: activated.

It must be taken into account whether the energy consumption is only for the electrolysis itself, or whether further installed units, such as electrolyte pumps, water treatment plant, gas purification, are
included.

The gas purity after compression and water removal is in general > 99.8 % for hydrogen and > 99.6 % for oxygen. Higher purities are possible by means of catalytic conversion and/or adsorptive drying

page 42 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
units.

Pressure Electrolysis. Pressure electrolysis units in the range of 0.6 – 20 MPa are available. Since hydrogen is usually required at high pressure, and because the power consumption for the pressure
electrolysis does not increase significantly, these units save both equipment and energy for hydrogen compression. The construction and the sealing of the electrolysis cells, however, is complicated (see
Fig. 35). The operating data reported by producers of pressure electrolysis are given in Tables 19 and 20. To compare the data with those of low pressure electrolysis the energy consumption for hydrogen
compression to the end pressure has to be considered.

Table 20. Examples for small electrolysis units, projects, novelties, small series and advanced technologies

Producer, country Type of electroysis Electrolyte Pressure, MPa Current density, A/m2 Cell Voltage, V Power Development
(membrane) Temperature, °C consumption, kWh/m3 (STP) H2 status

Advanced Electrolysis bipolar KOH, 0.6 10 000 project

[113],
Federal Republic of filterpress conventional 140 1.8/4.3
Germany
Asea Brown Boveri, bipolar water 2.0 10 000 small plants
Switzerland (polysulfone) 80 1.74/4.2 1 – 20 m3
(STP)/h
High Temp. Electrolysis bipolar water 0.3 6000 1 3 (STP)/h
[127],
Federal Republic of Y2O3+ ZrO2 1000 1.32/3.2 pilot plant
Germany projected
Hydrogen Appliances, bipolar water 2.5 4.2 – 5 small plants
Belgium (polysulfone) 50 0.04 – 1 m3
(STP)/h
0.1 – 5 m3
(STP)/h
Sema Metra Conseil, with additional conventional 20.0 5.5 – 6 20 – 100 m3
France compressing fluid (STP)/h
Sunshine-Project, bipolar KOH (30 wt %) 2.0 4000 20 m3 (STP)/h
Japan [120], [126] filterpress conventional 120 1.65/4.0 pilot plant
projected
Vandenborre, bipolar KOH (30 wt %) 1.0 small plants
Belgium 100 /3.9

Figure 35. Cross section and exploded view of cells of a pressure electrolysis unit (Lurgi, Zdansky-Lonza pressure electrolysis)

a) Bipolar electrodes, dimple plate cell partititon; b) Pre-electrodes in the form of nets on both sides of the asbestos diaphragm (c); d) Cell frame; e) Sealing ring, sealing the electrolyte against the surroundings and
insulating the bipolar sections against each other; f) Electrolyte duct; g) Bore in the electrolyte duct for supplying the electrolyte solution to both sides of the electrolyte; h) Parallel H2 and O2 ducts; i) Bore in the gas duct; j)
Electrolyte compartment

Advanced Conventional Electrolysis. The conventional potassium hydroxide electrolysis according to [113], [114], can be improved considerably. Some of the suggestions made have been
incorporated in single commercial electrolysis units. Other components necessary are still in the development stage. Areas of development are:

page 43 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Zero-Gap Cell Geometry. By placing the electrodes directly on the surface of the diaphragm the space between the electrodes can be reduced and the voltage drop in the electrolyte, and thus the heat
loss caused by the current, can be minimized. Figure 36 shows perforated foil electrodes in a zero-gap construction.

Figure 36. Zero gap cell geometry. Perforated cathodes and anodes, similar to the nets in Figure 35, in direct contact with the diaphragm and the membrane, respectively. The conical apertures with different diameters for
O2 and H2, respectively, facilitate the detatchment of the developing gas bubbles [113]

In particular the construction with pre-electrodes has already been used in commercial plants. The disadvantage of this type is that in these electrolysis units, with high current density and high operating
temperatures, the corrosion tendency is greatly increased. Products of corrosion, which may be formed at the anode, can grow through the diaphragm and lead to dangerous short-circuits.

New Materials for Diaphragms. A replacement for asbestos diaphragms are polysulfones, with hydrophobic character decreased by means of additives (e.g., polyantimonic acid) [115].

Another line of development is the use of oxide-ceramic diaphragms based on barium or calcium titanate or nickel oxide [116]. On a metallic net a hydrophilic gas-tight oxide layer is deposited. The
thickness of the layer is <500 µm, which leads to a low resistance of ca. 2×10–5 Ω/m2. The stability against electrolytes at high temperatures is also favorable.

Electro-Catalysts. The kinetic inhibition of the electrode processes can be reduced by use of catalysts. This is already partly done in conventional plants. To assist the oxygen formation at the anode,
lanthanum-containing perowskites, nickel-cobalt oxides and Raney nickel has been suggested.

The overvoltage at the cathode during the evolution of hydrogen can be reduced by the use of noble metals (platinum black). Commercial catalysts are, however, free of noble metals. Sintered, foamed,
and alkali-leached nickel compounds, which are stabilized by the addition of TiO2, ZrO2 and MoO3, are used. The long-term stability and the corrosion resistance against hot alkali solutions expecially in
the depolarized state are the main problems encountered in the use of electro catalysts.

The increase in capacity of the new types of electrolysis cells makes further improvements necessary. The increased corrosion caused by the higher operating temperatures and greater current densities
requires special materials of construction. In large-capacity plants nickel-plated steel, in low-capacity plants (≤ 20 m3 [STP]) polysulfone, is employed. In particular, the gas collection system for the fine
hydrogen bubbles necessitates additional equipment.

If all the theoretical possibilities to improve the electrolysis were employed, the water electrolysis could be greatly improved. Region b in Figure 37 shows the boundaries of the improvement possible in
alkaline water electrolysis [113], [114].

Figure 37. Illustration of the cell voltage and current density of commercially available or projected electrolysis units

The cell voltage U (V) can be converted by the equation E = 2.392 · U to the specific energy in kWh/m3 (STP) H2.

a) Range of commercially available electrolysis units [113], [116]; b) Range of advanced and membrane electrolysis [111], [114]; c) Steam electrolysis

suppliers data (see Table 19); present results; pilot plant project

Electrolysis Plants (see Fig. 38). Apart from the electrolysis unit an electrolysis plant consists of a number of different installations, such as electricity supply, water treatment, electrolyte circulation, the
knock-out system (phase separator) to separate electrolyte from the product gases, gas purification, and the instrument and control system.

Figure 38. Process diagram of an electrolysis unit with the most important components

a) Rectifier unit; b) Process water demineralizer ion exchange unit; c) Electrolytic cell's water electrolyzer; d) Gas separator and cooler; e) Gas scrubber; f) Electrolyte tank and transfer pump; g) Gas holder; h) Filter; i)
Compressor; j) Gas purifier; k) Drying; l) High-pressure storage tank and cylinder filling station

The water consumed during electrolysis has to be continuously replaced; the consumption is 0.805 L per cubic meter of hydrogen plus evaporation losses. The water make-up must meet stringent purity

page 44 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
requirements to avoid a build-up of substances in the electrolyte solution which poison the electrodes or promote corrosion. The required water purity (conductivity 1 µS/cm) is obtained with the aid of ion
exchangers.

The power, which in large plants is supplied in the form of three-phase current at a high voltage, must be transformed and rectified down to the necessary d.c. voltage. In addition, power for the various
auxiliary equipment, such as the electrolyte circulation pump and the instrument and control system is necessary. Considerable amounts of potassium hydroxide are only necessary for the first filling.
Since the energy losses in the electrolyte are converted to heat, a large amount of cooling water is necessary to operate the plant. This is also true for the gas purification unit.

The gas purification depends on the purity requirements placed on the gas and may be a simple condensation to remove moisture present. For higher purity requirements the catalytic conversion of the
oxygen present in the hydrogen (see Section Removal of Oxygen) and, after that, an adsorptive drying stage (Section Thermally Regenerated Adsorbers) is used.

4.2.3. New Developments in Electrolytic Processes [117], [118]

Electrolysis with Proton-Conducting Ion-Exchange Membranes. In this type of electrolysis, better known as solid polymer electrolysis (SPE, General Electric), the diaphragm is replaced by a proton-
conducting ion-exchange membrane ( Electrochemistry, Fuel Cells – Proton Exchange Membrane Fuel Cells (PEMFC)). The water to be electrolyzed is introduced only to the anode side, and no
additives to alter the conductivity are necessary.

The solid electrolyte consists of a Nafion membrane with the approximate structure:

A cell consists of a ca. 0.25 mm thick film of this polymer. The following substances have been reported as electrode and catalyst material: platinum metal (cathode), and ruthenium oxide hydrate (anode).

The electrode layers have a high specific surface. The membrane becomes saturated with the water added at the anode side, which is electrolyzed and serves as a cooling medium at the same time. The
electrical charge is transported by means of the hydrated H+ ions of the sulfonic acid groups. Current densities of up to 20 000 A/m2 can be reached. The principal construction of the electrolysis cell is
shown in Figure 39 using the Membrel electrolysis (Asea Brown Boveri) as an example [119].

Figure 39. Principles and construction of an SPE electrolysis unit, Type: Membrel, Asea Brown Boveri [119].

At too high voltages ozone is produced at the anode

a) Porous cathodic current collector; b) Membrane, solid state electrolyte; c) Porous anodic current collector; d) Catalyst

The electrodes are in contact over the whole surface with a porous conductor; this is graphite on the cathode side and sintered nickel or titanium on the anodic side. The electrodes are in a bipolar
arrangement.

Electrolysis units with proton-conducting membranes are offered in a capacity up to about 5 m3/h of hydrogen from various producers. The characteristic operating data of such a module and of a
conventional electrolysis unit are compared in Table 21. The advantages of these SPE units are not so much in the savings in power, but especially in the reduced volume of the unit, due to the much
higher current density and the size reduction attainable because of the absence of the potassium hydroxide electrolyte system.

Table 21. Comparison of conventional potassium hydroxide electrolysis with electrolysis using polysulphone electrolyte and membranes

BBC KOH electrolysis ABB Membrel electrolysis Hydrogen appliances SPE electrolysis

Electrolyte 25 wt % KOH polysulfone Nafion polysulfone Nafion

Temperature, °C 80 80 50
System pressure, MPa 0.1 ≤2.0 2.5
Current density, A/m2 2000 10 000

page 45 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Cell voltage, V 2.05 1.74
Efficiency ( = 1.48/U), % 72 85
Energy consumption (3.54/ ), 4.9 4.2 4.2 (at 20 % efficiency)
kWh/m3 (STP) 5 kWh/m3 (STP) (at 100 % efficiency)
Block volume 30

Further developments are proceeding in this direction. For example, in test plant modules used in the Japanese Sunshine Project [120], membranes and catalysts have been operated at current densities
up to 20 000 A/m2 and temperatures up to 150 °C.

Solid polymer electrolytic cells can also be used advantageously, for the coproduction of ozone (O3) if the appropriate overvoltage is applied (see also Fig. 39).

Water Vapor Electrolysis. According to Equation (29) the total energy requirement for electrolysis is only slightly temperature-dependent. However, the reversible part of the energy i.e., G, which must
be supplied in the form of electrical energy, decreases with increasing temperature (see. Eq. 31 a and Fig. 29). This means that an increasing amount of the total energy can be supplied by thermal
energy. Overvoltages, which result from kinetic inhibition, also decrease with increasing temperature.

In the conventional alkali electrolysis the reaction temperature is limited because of the vapor pressure of the electrolyte and corrosion problems. Thus, for the asbestos-diaphragm cell the upper
temperature limit of operation is 120 °C. For the polysulfone membrane the upper temperature limit is ca. 150 °C.

In high-temperature electrolysis doped zirconium dioxide and some other rare earth oxides are used as electrolyte. At temperatures >800 °C ZrO2 and some other rare earth oxides are oxygen ion
conductors (lattice defect conductors). In Figure 40 the specific resistance of some mixed oxides is given. These mixed oxides are sintered into ring-shaped membranes, which are kept as thin as possible
(0.5 mm, goal is 20 – 30 µm [121]) to reduce the electrical resistance of the cell. The rings are built up into tubes and the tubes to bundles of cells (Fig. 41).

Figure 40. Specific resistance r (Ω cm) of various solid state electrolytes with oxygen-ion conductivity as a function of temperature [122]

a) (ThO2)0.93 (Y2O3)0.07; b) (ZrO2)0.88 (CaO)0.12; c) (ZrO2)0.90 (Y2O3)0.10; d) (ZrO2)0.92 (Yb2O3)0.08

Figure 41. Principles and construction of a bipolar high-temperature electrolysis cell with an oxygen-ion conductive ZrO2 electrolyte

a) Anode; b) Electrolyte (ZrO2); c) Cathode; d) Insulation; e) Conductive cell sections; f) O2 space; g) H2, H2O duct

Porous electrodes are positioned on the surface of the solid electrolyte. Mixed oxides with perovskite structure are used (LaNiO3, LaMnO3) for the anode. The cathode is a ZrO2/Ni cermet, which prevents
the active nickel metal of the cathode from agglomerating [123]. Water vapor is led to the cathode, whereby the cathode reaction for hydrogen formation takes place:

The oxygen ion is oxidized at the anode:

Typical operating conditions are:

page 46 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Potential 1.3 V
Current density 0.4 A/cm2
Operating temperature 900 °C
Electrical energy consumption ca. 3 kWh per cubic meter (STP) H2

The performance characteristics of the high-temperature electrolysis is also given in Figure 37.

The process has been tested on a laboratory scale [124-126]. A pilot plant for the production of 1 m3 (STP) hydrogen per hour is under construction. A 3.5 MW pilot plant is planned [127] but a number of
material problems must be solved before realization.

The reversal of this process is expected to have a potential in fuel cell applications. This applies to the aforementioned solid polymer electrolysis as well ( Fuel Cells – Proton Exchange Membrane Fuel
Cells (PEMFC)). The direct production of electric power from hydrogen, carbon monoxide or hydrocarbons with air or pure oxygen in fuel cells of this type is still in the first stages of development.

4.2.4. Hydrogen as Byproduct from other Electrochemical Processes

See also Electrochemistry; Chlorine.

Electrochemical processes for the production of sodium hydroxide, chlorine and chlorine compounds yield hydrogen as a byproduct. About 3 % of the world hydrogen production, but a higher percentage
of the merchant hydrogen, is produced in this way. The economically most important electrolysis is the chlor – alkali process. At present there are three different commercially important technologies to
produce sodium hydroxide and chlorine.

In the mercury process, hydrogen is generated by decomposition of sodium or potassium amalgam, which is itself formed in a concurrent reaction to the kinetically strongly inhibited direct hydrogen
evolution (32) on the mercury cathode (33).

(32)

(33)

(34)

The decomposition reaction (34) is carried out in either horizontal or vertical decomposers. Because of the hydrogen overvoltage, a catalytically active decomposition material (technically graphite) is used
and high temperatures are necessary. Hydrogen from the amalgam decomposers must be freed from mercury [to <0.002 mg Hg/m3 (STP)] by cooling, scrubbing with brine and use of iodine-activated
charcoal or a catalytic process.

In the diaphragm process, in which the overall reaction (35) occurs, hydrogen is formed on a hollow cathode made of iron, which is separated from the anode region by a diaphragm.

(35)

Both processes are, at this time, being replaced by the membrane process because of environmental and economic reasons.

The membrane process accounted for 10 % of the world capacity in 1987. New plants projected are based almost exclusively on the membrane process. Figure 42 shows the principal reactions and
products of a membrane cell.

Figure 42. Schematic construction of an electrolysis cell for the production of chloride, sodium hydroxide, and hydrogen using the membrane process (usually bipolar construction)

a) Cathode; b) Anode; c) Membrane; d) Insulation

In the membrane process perfluorinated organic polymers with sulfonyl or carboxyl groups are used exclusively as membranes. The anodes are made of titanium, the cathodes of stainless steel or nickel.

page 47 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Membrane cells can be in either unipolar or bipolar arrangement [128]. In addition to the main products chlorine and sodium hydroxide, the process yields high-purity hydrogen (> 99.9 %). Drying and
removal of oxygen traces can be carried out, if required.

Hydrochloric Acid Electrolysis. Apart from the chlor – alkali electrolysis the electrolysis of hydrochloric acid (in particular for treatment of weak aqueous solutions to enable their disposal) has some
economic importance. The cell products (chlorine and hydrogen) are saturated with water vapor and hydrogen chloride corresponding to the partial pressure of 20 % hydrochloric acid at 75 – 85 °C; both
product streams are cooled and the remaining chlorine and hydrogen chloride removed from the hydrogen stream by scrubbing with sodium hydroxide. Hydrogen purity is 99.9 %. For further electrolysis
processes of some economic importance, see Electrochemistry.

4.3. Thermochemical Water Cleavage

In the range T < 2000 K and p = 100 kPa the cleavage of water is described by:

(36)

The reaction is endothermic, the thermodynamic standard values are given below [129]:

Reaction
enthalpy
Reaction
entropy
Gibbs free
energy

The equation for the equilibrium constant is as follows:

The reaction enthalpy, H0, is fairly temperature independent, as is S0. As a result, the Gibbs free energy, G0, decreases substantially with increasing temperature (see Fig. 43). This behavior is
typical for highly endothermic reactions.

Figure 43. Gibb's energy, enthalpy, and entropy term of enthalpy (T S 0) for cleavage of water in a temperature range of 300 – 1200 K

The energy required for the production of hydrogen can be delivered in the form of chemical energy (see Section Production from Coal and Hydrocarbons), electrical energy (see Section Electrolysis),
heat, light or radiation (see Section Other Methods for the Cleavage of Water), or in some combination of these various energy forms. In many cases, the total process consists of a number of steps using
chemical, electrochemical, biochemical or photochemical systems. Purely thermal cleavage of water occurs only at >2500 K (see Fig. 48). A commercial realization of this reaction is presently not feasible.

In thermochemical cyclic processes reaction (36) is carried out in steps with the aid of chemical redox reactions utilizing energy at lower temperature levels. The substances involved in the reaction
sequences are recycled. The great advantage compared to electrolysis is, that the low efficiency, resulting from production of electricity, can be avoided and thermal energy is directly converted into
chemical energy. If one of the steps involved is an electrolysis step, the process is named a hybrid cyclic process.

The possibility of producing hydrogen via multistage cyclic processes was first pointed out by FUNK and REINSTROM [130]. In particular, the possibility of utilizing the heat of high-temperature reactors led
to intensive investigations, and a number of processes for laboratory and pilot plants were selected.

4.3.1. Thermodynamics of Closed Cycles

Efficiency. A thermochemical cyclic process is subject to many thermodynamical constraints. The maximum attainable efficiency can be calculated according to [130].

page 48 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

T1 is the upper process temperature of the heat source, T2 the lower process temperature of the heat sink. The factor H0 / G0 is the ratio of the enthalpy to Gibbs free energy for the water cleavage
reaction at 298 K and 100 kPa. The factor has a value of 1.2. Table 22 shows calculated efficiencies for T2 = 298 K.

Table 22. Theoretical thermal efficiency of the thermochemical water cleavage (T2= 298 K)

T1, K , % T1, K , %

400 30.7 1000 84.6

600 60.6 1200 90.5
800 75.6 1400 94.8

Thermodynamical considerations show that in the range 300 – 1000 K at least three process stages are necessary to dissociate water thermochemically. At the same time, calculations show that the
theoretical efficiency decreases as the number of stages increase [131].

If the heat losses and irreversible chemical reactions that are involved in practical processes are considered, the maximum efficiency attainable amounts to only 50 – 80 % of the theoretical value.
However, the overall efficiency reached should still be clearly better than the total overall efficiency of water electrolysis (ca. 30 %).

Entropy-Temperature Diagram. A thermochemical cycle can be displayed graphically in an entropy-temperature diagram. Figure 44 shows this for the MARK 9 process (see Table 23). For simplification
it is assumed that the reactions occur at equilibrium conditions ( G0 = 0), that heat losses can be avoided and that no energy is necessary for the transport of reactants. Since the enclosed area
corresponds to usable energy, in this case a constant equal to – G0 = 237.3 kJ/mol, the efficiency of the cyclic process can be determined using other thermodynamic data, e.g., the amount of heat
absorbed by the process [132]. The thermodynamics of this process are summarized in [133], [134].

Table 23. Selection of reaction sequences of thermochemical cycles for water cleavage

Designation, Reaction sequence Number of Maximum

reactions temperature, °C

Metal – bromine cycles

MARK 1, (de Beni, Euratom) 4 730

CaBr2+ 2 H2O Ca(OH)2+ 2 HBr

2 HBr + Hg HgBr2+ H2

HgBr2+ Ca(OH)2 CaBr2+ HgO + H2O

MARK-1 B, MARK-1 C, MARK-1 S, MARK 5 use also bromine

Iron – chlorine cycles
MARK 9 (C. Hardy, Euratom) 3 650

6 FeCl2+ 8 H2O 2 Fe3O4+ 12 HCl + 2 H2

2 Fe3O4+ 12 HCl + 3 Cl2 6 FeCl3+ 6 H2O + O2

6 FeCl3 6 FeCl2+ 3 Cl2

page 49 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
MARK 7, MARK 7 A/B, A 2 (Institute of Gas Technology),
AGNES (General Electric) are further cycles of the Fe – Cl family
Vanadium – chlorine cycles
MARK 3 (Euratom), Allison div. of General Motors,
RWTH Aachen (Fed. Rep of Germany) 4 800

2 VCl2+ 2 HCl 2 VCl3+ H2

2 VCl3 VCl2+ VCl4

2 VCl4 2 VCl3+ Cl2

Cl2+ H2O 2 HCl + H2

Metal – alkalimetal cycles
Wentorf “C” (Li, K, Ni, J2) 6 700

MARK-2 C (De Beni, Euratom, Na, Mn, C, Hg) 6 850

Methane cycles
Methane – Methanol – Arsenous Oxide Cycle 5 700
CH4+ H2O CO + 3 H2

CO + 2 H2 CH3OH

CH3OH + As2O4 CH4+ As2O5

½ As2O5 ½ As2O3+ ½ O2

½ As2O5+ ½ As2O3 As2O4

Figure 44. Entropy – temperature diagram for MARK 9 cycle [131]

a) Hydrolysis of iron(II) chloride and hydrogen formation at 1100 K; b) Decomposition of iron(III) chloride releasing chlorine; c) Fusion of iron(III) chloride at 497 K; d) Chlorination reaction of magnetite at 497 K

The entropy changes at 298 K and 375 K include entropy of formation and vaporization of water. Without those values is 0.477

4.3.2. Further Criteria of Thermochemical Cycles

Further criteria must be considered in order to judge the practicability of theoretical cycles. For example, the reactions should be in the temperature range of 300 – 1000 K; the reaction rate should be
sufficiently large; chemistry and kinetics should be known. Furthermore, internal heat compensation should be possible; separation of the reaction products should be feasable using a minimum of energy
with a minimum movement of solids. Finally, process and material know-how should be available. A further important condition for the development of a process is the nature of the high-temperature
source, e.g., solar concentrator or nuclear high-temperature reactor, and the way this energy is integrated [135], [136].

With the help of computer programs and taking the aforementioned criteria into consideration, several thousand cyclic processes have been suggested. A summary of more than 200 processes can be
found in the literature [137-139]. After the feasibility from a technical and process point of view has been investigated, 20 to 30 of these processes remain. These, finally, have been studied in greater
detail.

Thermochemical cyclic processes can be categorized according to various aspects, e.g., the number of reaction stages, or the maximum temperature. However, it has become accepted to group them in

page 50 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
families according to the chemicals involved, e.g., iron-chlorine family (Fe, Cl2, FeCl2, FeCl3), sulfur-halogen family (SO2, halogen, X2), copper-chlorine family (Cu, Cl2, CuCl, CuCl2). In addition, the basic
reaction types for the thermochemical cyclic processes can be classified as follows:

Hydrogen formation usually occurs at low or medium temperatures (ambient to 400 °C). Examples are the reactions of metals (Eq. 37) or metal halides (Eq. 38) with acids or the decomposition of hydrides
(Eq. 39).

(37)

(38)

(39)

Oxygen formation takes place at high temperatures (700 – 1000 °C) by oxygen elimination from oxides or oxygen-containing acids. The thermal cleavage of sulfuric acid (Eq. 40), for example, is one of the
best-investigated high-temperature process steps:

(40)

Water is introduced into the cycles and allowed to react at various temperature levels. For example, the Deacon reaction (41) is of great interest at high temperatures, while at low temperatures the
reaction of water with bromine or iodine in the presence of sulfur dioxide [Bunsen reaction (42) known since 1853] is particularly important:

(41)

(42)

4.3.3. Thermochemical Cyclic Processes

Table 23 shows a selection of the various “purely thermochemical” reaction systems described in the literature. The Mark I cycle (Mark cycles, nomenclature JRC, i.e., Joint Research Center, Ispra, Italy)
with a four-step mechanism was developed by DE BENI in 1969. At the moment the processes belonging to the halogen-sulfur family are invoking the most interest. As an example, the sulfur – iodine
process (S – I process) developed by General Atomic Technologies [140] will be described in more detail.

The main reactions are sulfuric acid cleavage (43) at high temperature and the Bunsen reaction (42), a low-temperature step, in which the water cleavage takes place. Hydrogen is converted to a
hydrogen halide and oxygen is incorporated in the sulfuric acid

Using the molar amounts suggested by General Atomic, the Bunsen reaction can be written as:

The excess of iodine leads to a separation of two liquid phases, a lighter phase of aqueous sulfuric acid (1) and a heavier one consisting of hydrogen iodide, water and iodine (2). The process is
subdivided into 4 stages as demanded by the necessary basic operations (Fig. 45 A) [140].

Figure 45. A) Material flows in the S – I process according to General Atomic

B) Section III within the S – I process according to General Atomic

page 51 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
In stage 1 utilizing the Bunsen reaction, the separation of acidic products H2SO4 and HI and oxygen occurs. The oxygen is obtained in a mixture with sulfur dioxide from the sulfuric acid decomposition.
This gas mixture is passed through the liquid mixture of water, sulfuric acid, iodine and hydrogen iodide. Here sulfur dioxide reacts according to the Bunsen reaction and oxygen remains as the sole
gaseous component. The exothermic Bunsen reaction is performed at 120 °C, to make certain, that the products remain in the liquid phase.

Stage 2 represents the drying of sulfuric acid and the sulfur trioxide cleavage. In this stage pure sulfuric acid is produced in a six-stage drying process at medium temperature. In an ensuing evaporation
the dry sulfuric acid decomposes into water and sulfur trioxide. After this, sulfur trioxide is catalytically decomposed to sulfur dioxide and oxygen by using high-temperature energy.

In stage 3 pure hydrogen iodide is extracted from the heavy product phase of the Bunsen reaction and can then be decomposed economically in stage 4 using low-temperature energy. To obtain pure
hydrogen iodide for the decomposition in stage 4, the heavy product phase of the Bunsen reaction is separated into three pure fractions. Figure 45 shows a simplified process scheme of this stage. The
addition of highly concentrated phosphoric acid to the system causes a new phase separation, whereby the heavy phase contains iodine and the light phase hydrogen iodide, water and phosphoric acid.
The light phase can be separated off at moderate temperatures. The gaseous product at the head of the separation column is almost pure hydrogen iodide. This is cooled, compressed to 5 MPa and fed to
stage 4 in the liquid state.

The bottom product of the column consists of aqueous phosphoric acid. To be able to recycle the phosphoric acid, it must be freed from the excess water. This drying is achieved in a multistage process
with vapor compression and consumes a large amount of electrical energy.

The following suggestions have been made to improve stages 2 and 3: power generation by decompression during sulfuric acid drying [141] and optimization of the hydrogen iodide separation with
integrated hydrogen formation [142]. This process has been operated on a laboratory scale with a thermal efficiency of 47 %; the aforementioned improvements and optimization of the heat system is said
to make an efficiency of 55 % possible.

Several variations of the S – I process are being investigated, e.g., the NIS process with the system Ni – NiSO4 [143], the Mg – S – I cycle with the system MgO – MgSO4 and the C – I – S process using
methanol as a reactant as suggested by Japan Atomic Research Institute [144].

4.3.4. Hybrid Cycle Processes

Thermochemical – electrochemical hybrid cycle processes are a special form of thermochemical cyclic processes. Here an electrochemical reaction is involved. Hybrid cycle processes can, therefore,
usually be designed as two-stage processes.

The electrochemical reaction stage is normally run at low temperature, and replaces either the hydrogen producing or the oxygen producing reaction. The necessary energy for the electrochemical
reaction must be less than that required for direct electrolysis of water. Table 24 summarizes the thermochemical – electrochemical hybrid cyclic processes.

Table 24. Thermochemical – electrochemical cycles (hybrid cycles)

Designation, reaction sequence Number of Temperature, °C

maximum
reactions

Sulfur – halogen hybrid cycles

MARK 13 (JRC Ispra) 3 650 – 850

Br2+ SO2+ 2 H2O 2 HBr + H2SO4

H2SO4 H2O + SO2+ ½ O2

2 HBr H2Br2 (electrolytic step)

Sulfur hybrid cycles

MARK 11 (Westinghouse Electric Corp., JRC Ispra, General Atomic Technologies) 2 900

H2SO4 SO2+ H2O + ½ O2

SO2+ H2O H2SO4+ H2 (electrolytic step)

Alkali metal hydride – hybrid cycles
Me/MeH – hybrid process (F. Behr) 2 700

page 52 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
H2O + 2 Me 2 MeH + ½ O2 (electrolytic step)

2 MeH 2 Me + H2

Hydrocarbon – hybrid cycles

Methane – methanol – methanal process (KFA Jülich) 4 950

CH4+ H2O 3 H2+ CO

CO + 2 H2 CH3OH

CH3OH CH2O + H2

CH2O + H2O CH4+ O2 (electrolytic step)

The most important hybrid cyclic process is the so-called Westinghouse cycle (MARK 11, according to the Ispra nomenclature). A large amount of work on this sulfur-family process and some of its steps
have been carried out by Los Alamos Scientific Laboratories, Westinghouse Electric Corp., JRC Ispra, KFA Jülich, General Electric and others.

The sulfur hybrid cycle is a simple two-stage cyclic process. The first step is the electrochemical low-temperature reaction consisting of the steps (1) cathodic hydrogen formation (Eq. 45) and (2) anodic
oxidation of sulfur dioxide to sulfuric acid (Eq. 46).

(45)

(46)

In the high-temperature reaction (Eq. 47) the electrically produced sulfuric acid is cleaved into oxygen, water, and sulfur dioxide, the latter is recycled to the anodic process.

(47)

This reaction can be accelerated in the temperature range of 500 – 1000 °C with catalysts (Fe2O3, CuO, Pt – TiO2, Pt – SiO2). Figure 46 depicts the flow sheet of the MARK 11 process. The stages of
developement for the key steps of the process are as follows:

Figure 46. Schematic of the Westinghouse (Mark 11) hybrid sulfuric acid cycle, simplified [145]

a) Electrolysis; b) H2SO4 surge tank; c) Acid separator; d) Acid concentrator; e) Acid vaporizer; f) SO3 decomposition reactor; g) Steam condensor; h) SO2 condensor

Electrolysis. The desired conditions to be attained for the anodic reactions are:

50 wt % H2SO4
SO2 pressure 100 – 200 kPa
temperature 80 °C

The high anodic overvoltage makes potentials of 680 – 800 mV necessary to obtain current densities of 200 mA/cm2. To reduce this overvoltage catalysts are being tested. Diffusion of sulfur dioxide from
the anode to the cathode and its reduction to sulfur or hydrogen sulfide there must be prevented. This is being attempted by use of H+-permeable separator membranes (SPE, diaphragms) [146], gas
diffusion electrodes as anodes [147], or three-compartment electrolysis cells [148].

Concentration and Cleavage of Sulfuric Acid. The ca. 50 wt % sulfuric acid produced at the anode is first concentrated to maximum 98 wt % in several concentration, temperature and pressure stages,
then evaporated at 450 °C and 20 MPa and subsequently decomposed (at 400 – 800 °C into SO3 – H2O, at > 800 °C into SO2 – O2). Material corrosion problems occur. Silicon carbide and silicon nitride

page 53 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
materials are suitable for the evaporation equipment. For the decomposition of sulfuric acid special alloys with aluminum oxide coating (SS 444, Incolloy 800, Hastelloy G) have been proposed as reactor
materials. The sulfur trioxide decomposition has been investigated on a pilot scale at the JRC Ispra in the so-called Christina process. This process has the following special features:

heat provision by hot air

adiabatically-operated cracking reactor with a
special construction (see Fig. 47)
downstream heat recovery section.

Figure 47. Design of the reactor for SO3 decomposition for the Christina process

a) Quartz wool; b) Refractory lining; c) Catalyst container (Incoloy 800); d) Catalyst (iron oxide); e) Thermocouples; f) Pressure vessel (500 °C)

Figure 47 shows the sulfur trioxide cracker which is constructed for a pressure of 2.5 MPa and, with 11 s residence time, gives 95 % conversion with a crack rating of 12 kW. The results of the
investigation show that sulfur trioxide cracking (a step of various cycles of the sulfur family), is technically feasible [149].

For the MARK 11 and related processes the theoretical thermal efficiency is 40 – 45 %. At present, an efficiency of 30 % is attainable.

4.4. Other Methods for the Cleavage of Water

All processes mentioned in the following are either still in the laboratory or in the development stage. A review is given by BOCKRIS in [150], where further methods such as plasmolysis, i.e., water splitting
by high-temperature plasmas, magnetolysis, i.e., electrolysis using a potential difference created by magnetic induction, and magmalysis, i.e., interaction of steam with fresh basaltic lava in volcanic tubes,
are presented.

4.4.1. Thermolytic and Radiolytic Processes

Water Thermolysis. The direct thermal cleavage of water, theoretically the simplest method and a priori an ideal process, requires very high temperatures. Figure 48 shows the equilibrium composition of
dissociating steam at atmospheric pressure as a function of temperature [151]. The six components present to a significant extent are: H2O, H2, O2, H, O and OH. The degree of water dissociation as a
function of temperature is:

2000 K 0.69 mol %

2300 K 2.64 mol %
2700 K 10.35 mol %
3000 K 22.4 mol %
3500 K 57.43 mol %

Figure 48. Molar fraction a of species for the dissociation of water at thermodynamic equilibrium as a function of temperature starting from pure steam at a pressure of 105 Pa

Appreciable quantities of hydrogen can only be obtained at temperatures significantly above 2000 K.

page 54 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
An economic utilization of this process is difficult, because of the recombination of the reaction products at the high temperatures. Other problems are associated with the recovery or separation of
hydrogen from the hot mixture, conservation and recovery of the high-temperature energy and unsolved material problems.

The three different approaches generally considered are:

1. Catalytic thermal decomposition of water on metallic wires at 1600 K, the small degree of decomposition being compensated for by high reaction rates [152].
2. Decomposition of water in the vicinity of a high-temperature-resistant membrane which allows the removal of one of the products with accompanying displacement of thermodynamic equilibrium.
Stabilized zirconia or porous ceramic membranes look promising [153].
3. Decomposition of water, followed by rapid quenching of product gases. A quenching speed of 105 – 106 K/s is required. Special quenching arrangements using cold steam or inert gas jets indicate
that up to 90 % of the hydrogen formed can be recovered [154].

The research in this area is aimed at the utilization of high-temperature solar or nuclear energy.

Water Radiolysis. Water can also be cleaved by high-energy radiation. The overall formula for the water radiolysis is:

The principles of water radiolysis and radiation chemistry in general are described in [155].

Energy from nuclear sources is released as high-quantum energy photons ( -radiation), high-speed particles, either charged ( -particles, protons, and -radiation), or uncharged (neutrons). The rate of
interaction of radiation with a medium, i.e., the rate of energy transfer, is termed linear energy transfer (LET). Pure -radiation sources have the smallest, uranium fission fragments the highest value (in
water). The radiochemical yield (G value) depends on the LET value of the radiation. The G value is defined as the number of molecules converted per 100 eV absorbed energy. The G(H2) value is greatly
influenced by the physical state and turbulence in the medium as well as the chemical composition.

The irradiation of water in the liquid state with -rays (60Co) gives a G(H2) yield of 0.45, in the vapor state at 110 °C G(H2) is 5.2 [156]. With 210Po a G(H2) value of 1.8 can be attained.

A well-known example of hydrogen formation is the undesired dissociation of the reactor coolant water in nuclear reactors (For safety measures see Section Removal of Hydrogen). Hydrogen production
coupled with heat production in nuclear reactors has often been proposed or is the subject of experimental research. The most suitable configuration is seen in the so-called homogeneous reactor [157].
Here normal water serves as a solvent for the nuclear fuel (94 % enriched uranyl sulfate) and as a moderator. Using a G(H2) value of 1.6 a hydrogen yield of 28.8 kg H2 per day and Megawatt can be
calculated. Only a fraction (ca. 35 %) of the energy is used for the formation of hydrogen, the rest is converted to heat.

Hydrogen, along with higher hydrocarbons, is formed during the radiolysis of methane with a G (H2) = 6.4. The radiolysis of higher aliphatic hydrocarbons yields, along with hydrogen (G (H2) ethane = 6.6,
G (H2) propane = 8.2), a large number of other products [158].

Another indirect process for the production of hydrogen is considered to be of interest. Carbon monoxide, derived from the radiolysis of carbon dioxide, is converted to hydrogen and carbon dioxide in a
subsequent shift reaction; carbon dioxide is then recycled. With a G (CO) value of 10 a higher hydrogen yield than obtainable with direct water radiolysis has been calculated [156].

4.4.2. Photochemical or Photoelectrical Water Cleavage

Direct photolytic cleavage of a water molecule leads initially to H and OH radicals

The energy necessary corresponds to 1 mol of light quanta with an energy of at least 4.73 eV ( ≤ 262 nm). This process, however, only occurs in the vacuum UV range at a wavelength of ≤ 123 nm,
corresponding to a quantum energy of 10 eV [159]. Because of the unavoidable recombination reactions only a small fraction of hydrogen is found in the reaction mixture.

The photolytic cleavage of water with solar radiation is thus not possible; the process must be carried out in stages using several light quanta and with separation of the reactive intermediates. For the
development of a process various possible routes can be followed.

Photochemical Systems. It has been known for a long time, that acidic water solutions of transition metal ions such as Cu2+, Fe2+, Eu2+, evolve hydrogen upon irradiation with UV light. The metal ions
provide an electron and are thereby oxidized. The quantum yield is relatively high. A cyclic water cleavage with recirculation of the oxidized metal ion back to the initial (reduced) stage is not possible
except with Fe3+ and Ce4+.

Efforts are presently concentrated on cyclic photo-redox processes with assistance of photosensitizers. The energy terms of such a system for the cleavage of water are shown in Figure 49. Photon
energy, after exciting the sensitizer S to S*, is converted into chemical potential by creating the product ions S+ and R– (R is referred to as electron relay compound). This light-induced redox reaction is
followed by catalytic steps causing water decomposition [160].

page 55 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 49. Energy term scheme of a light-induced redox reaction coupled with water cleavage [160]

S = sensitizer; R = electron relay compound; Cat1, Cat2 = catalysts

Examples for sensitizers are: , porphyrine derivatives and acridine dyes such as proflavine. As electron relay compounds a viologen derivative, e.g., N-methyl-N′-tetradecylbipyridine, MV2+,
salicylate complexes of Eu3+, V3+, and macrocyclic cobalt complexes are used. Suitable catalysts are finely dispersed particles of platinum for formation of hydrogen and colloidal or macrodispersed RuO
2, which assists in the oxygen generation. A generalized overall reaction scheme is given in Figure 50. Difficulties to be overcome are the prevention of side-reactions and attaining a sufficient rate
reaction. The former is achieved by using microheterogeneous systems (instead of homogeneous systems), the latter by using bifunctional catalysts, e.g., TiO2 particles doped with RuO2, which serve at
the same time as support materials for platinum.

Figure 50. Schematic for cyclic water decomposition in a combined catalytic system [161]

A solution containing 10–4 mol/L and 2×10–3 mol/L MV2+ gives a quantum yield in hydrogen (and oxygen) formation of 2 %, with hydrogen yields being 400 mL/h per liter of substrate. This
remains constant over at least 40 h [161].

Photoelectrochemical Systems. The photoelectrical water cleavage is based on the absorption of photons in a semiconductor electrode, which is in contact with an electrolyte. One part of the redox
process, either oxidation or reduction, takes place on it. The opposing process takes place at the second electrode, thus separating hydrogen and oxygen evolution locally. Figure 51 depicts the energy
term scheme of such a photoelectrochemical cell with an n-type semiconductor anode [162]. The steps of the cyclic process can be described as:

1. Absorption of photons creating charge separation by transportation of electrons from the valence band to the conduction band (electron – hole pairs);
2. Abstraction of electrons from the valence band of water, recombination with defect electrons (holes) and oxygen formation;
Anodic reaction

3. Migration of hydrated protons to the cathode;

4. Transfer of electrons from the Fermi niveau of the metal to the protons followed by hydrogen formation.
Cathodic reaction

The choice of the semiconductor is determined by the height of the band gap and its relative energy level. To cleave water, the energy difference between valence band and conduction band must be at
least 1.23 eV; in actual fact, it must be much larger. The edge of the conduction band must be sufficiently above the redox Fermi level for water oxidation.

Figure 51. Energy scheme of the photoelectrolysis of water with a SrTiO3 semiconductor electrode

a) Conduction band; b) Valence band

EF = Fermi energy; EG = energy gap between valence and conduction band; EB = “buckling” of the energy level at the boundary layer electrode/electrolyte

page 56 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
With SrTiO3 semiconductor anodes the band levels are at the right height. The band gap, however, is 3.2 eV, so that only the absorption of UV light is effective for charge separation. In case of the TiO2
semiconductor electrode the conduction band has to be raised by 0.5 V above the hydrogen potential by application of an external voltage. A further complication is the photolytic decomposition of
photoactive semiconductor surfaces (photocorrosion) of suitable electrodes [163].

In recent years the light-induced cleavage of water in a microheterogeneous semiconductor dispersion has been investigated (see [164]). As semiconductors titanium oxide, titanates, vanadium oxides
and cadmium sulfide can be used. The principal mechanism is shown in Figure 52 A. Titanium oxide particles with a diameter of 100 – 500 nm form a transparent dispersion. Locally separated catalyst
deposits of platinum and RuO2 promote the water splitting reactions by reducing protons and forming oxygen; in the case of TiO2, oxygen normally remains attached to the particle surface in the form of a
peroxide.

Figure 52. A) Catalytic water cleavage through band gap excitation of colloidal semiconductors

B) Complete water photolysis scheme through sensitization of a semiconductor particle

(vb = valence band; cb = conduction band, EF = Fermi energy)

For particles of semiconductors with a large band gap a sensitization mechanism is required. Through light excitation, the sensitizer acquires the chemical potential necessary to promote an electron to the
conducting band of the particle from where it is transferred to the platinum catalyst deposited on part of the surface. Following this, the redox reaction leading to water decomposition takes place as
described above.

Figure 52 B shows a schematic diagram of the complete process. The sensitization of titanium oxide by forming surface derivatives with transition metal complexes shifts the photochemical excitation
range down to 600 nm with a maximum at 480 nm, whereas pure titanium oxide exhibits a band-gap transition absorption only <400 nm [165]. In principle, it should be possible to use such an absorber
system and to attain conversion efficiencies of the solar water cleavage up to 15 %.

4.4.3. Hydrogen Formation in Biological Systems

The capability of producing molecular hydrogen by biological action is inherent to some microorganisms including chemotrophs, i.e., organisms obtaining energy from chemical substrates, and
phototrophs, i.e., those using light as a source of energy.

The essential step for the biochemical hydrogen formation is the reduction of two protons by the transfer of two electrons. This reaction is brought about by the interaction of appropriate electron-donating
enzyme systems. The energy is supplied by any suitable substrate or by a photosystem using visible radiation. The process can take place in free organisms or in immobilized biological systems.

Hydrogen Production from Organic Substrates by Fermentative Bacteria. Some bacteria can be caused to ferment carbohydrates with the development of hydrogen and byproducts such as carbon
dioxide. Obligate and facultative anaerobic microorganisms, e.g., various types of Clostridium, types of Ruminococcus, Serratia marcescens, and Bacillus polymyxa, are a means of carrying out this
reaction.

In most cases, hydrogen is evolved as the result of anaerobic oxidation of pyruvate, an intermediate in carbohydrate metabolism. Starch, cellulose, and glucose can serve as substrates. The growth
conditions of the microorganisms, the pH value of the substrate and the hydrogen partial pressure have a great influence on hydrogen production. Yields range from 0.1 – 2.5 mol H2 per mole substrate
([166], [167]). The fermentation of carbohydrates is not of any importance from the point of view of hydrogen production. The energy conversion corresponds to ca. 20 % of the energy of combustion of the
organic substrates.

Photobiological Production of Hydrogen. The ability to set hydrogen free with the aid of solar energy is found in a number of phototropic organisms such as purple bacteria, green bacteria,
cyanobacteria (procaryotes) and several algae (eucaryotes). The majority are aquatic species. An overview is presented in [167].

The mechanisms vary from organism to organism, the main steps, however, are similar to those of the photosynthesis using electron-transporting photosystems connected in series. For example, with
green algae, such as the Chlamydomonas types, the mechanism of the photobiological water decomposition follows the scheme depicted in Figure 53.

Figure 53. Photosynthetic electron transport system of hydrogen production [168]

PSI, PSII = photosystems I and II; Y, X = primary electron acceptors of photosystems I and II; Chl = chlorophyll; Q = plastoquinone; Cyt B, Cyt C = Cytochromes B and C; PC = plastocyanine; [Mn] = manganese carrying
water oxidizing enzyme; Fd = Ferredoxin

page 57 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

The active reaction centers, together with light-harvesting antenna pigments, are located in the so-called thylakoide membrane of chloroplasts. After irradiation, photosystem II (PS II) extracts electrons
from water at a redox potential of + 0.85 V with the assistance of a manganese-containing enzyme. The electron acceptor of PS II delivers electrons to photosystem I through a down-hill electron chain.
Photosystem I elevates the electrons to an energy level which allows the reduction of protons to hydrogen (E0 = – 0.42 V). For the normal photosynthetic carbon dioxide fixation the electrons are
transferred by means of the NADP – NADPH system at E0 = – 320 mV, and carbohydrates are generated from carbon dioxide. In the case of hydrogen formation, however, the electrons are transferred to
the protons by means of the low-molecular mass protein ferredoxin, a powerful reductant, and the enzyme hydrogenase.

Oxygen evolves on the inside and the ferredoxin reduction and hydrogen formation takes place on the outside of the thylakoide membrane (see Fig. 53). Since oxygen diffuses freely through the
membrane and deactivates the extremely sensitive hydrogenase, such a system is of little interest for the commercial biophotolysis of water.

Purple bacteria lead to photosynthesis with sulfide, thiosulfate, sulfur or organic compounds as electron donors. Some strains (like Rhodopseudomonas capsulata) produce hydrogen and carbon dioxide
during irradiation in the presence of organic compounds such as maleate, succinate or especially lactate [169]. Some mutants of Rh. sphaeroides may completely oxidize glucose to carbon dioxide and
hydrogen. The necessary conditions (low concentrations of N-containing compounds and the absence of molecular nitrogen) indicate that the hydrogen photoproduction is catalyzed by the enzyme
nitrogenase. The energy supply is most probably by photosystems I and II and the reduced electron carrier ferredoxin. At the moment, work is proceeding to develop an economical process for the
photoproduction of hydrogen from highly polluted waste water using purple bacteria [170].

Cyanobacteria (blue-green algae) represent the most numerous group of phototropic procaryontes. Some of them are able to produce hydrogen. Many filamentous cyanobacteria (e.g., Anabaena
cylindrica or Nostoc muscorum) have special types of cells (heterocysts) in their filaments of vegetative cells. These heterocysts contain the enzymes nitrogenase and hydrogenase, but are deficient in the
photosystem II, and are thus not able to cleave water and to produce oxygen. The exceedingly oxygen-sensitive nitrogenase remains active because evolution of oxygen by photosystems I and II is
confined to the vegetative cells. The carbohydrate of the vegetative cells provides the reductant for the nitrogen reduction to ammonia, respectively for hydrogen formation. Figure 54 shows schematically
the interaction of the metabolism of vegetative cells and heterocysts [171]. The rate of hydrogen production of cyanobacteria is low, yet the feasibility with such systems is considered to be favorable,
because hydrogen evolution takes place at ambient conditions and it is expected that suitable mutants can be selected.

Figure 54. Model of hydrogen metabolism by heterocystous blue-green algae

PSI, PSII = Photosystems I and II

Cell-free systems containing chloroplasts, hydrogenases and an electron carrier such as ferredoxin, sometimes immobilized, have been investigated. In a comparatively short period of time a considerable
increase in the rate of hydrogen evolution and stability of such systems has been achieved. Latest data show, that this process can go on for ca. 6 h with a rate of hydrogen production amounting to
100 mmol per hour and gram of chlorophyll [172]. Reasons for the stop of the reaction are inhibition of hydrogenases by oxygen, loss of activity of the chloroplasts and autooxidation of electron carriers.

Table 25 shows hydrogen evolution rates by phototropic microorganisms and cell-free systems [167]. The ultimate aim is the development of hydrogen-producing photoreactors containing artificial enzyme
systems, but results are not promising at present.

Table 25. Maximum rate of hydrogen evolution by biological systems [167]

Phototropic microorganisms H2 evolution per hour and gram dry biomass

mL mmol

page 58 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Purple bacteria
Rhodospirillum rubrum 146 6.5

Rhodopseudomonas palustris A 54 2.4

Rh. capsulata SL 130 – 150 5.8 – 6.7

Rh. capsulata B 10 300 – 500 13.4

Rh. capsulata LB 2 178 7.9

Thiocapsa roseopersicina 13 0.6

Cyanobacteria
Anabaena cylindrica B 629 5 – 40 0.2 – 2.0

An. variabilis 32 1.3

Spirulina platensis 9 0.4

Green algae
Chlorella vulgaris 4 0.2

Chlamydomonas reinhardii 137 C 45 2.0

Chl. moevussii ICC 97 8 0.4

Cell-free systems* H2 evolution per hour and gram chlorophyll Functioning

mL mmol time, h

Chp + Fd + H2-ase (C. pasteur) 200 – 350 9.0 – 15.5 3.0 – 6.0

Chp + Fd + H2-ase (Th. roseopersicina) 110 – 220 5.0 – 10 3.0 – 6.0

Chp + Fd + H2-ase (from various sources) 450 – 900 20 – 40 4.0

*Selected basic components: chloroplasts of spinach (Chp), ferredoxin (Fd), hydrogenases of various sources; electron donor H2O; the experiments were carried out in the presence of O2-
scavengers

4.5. Other Chemical Processes

4.5.1. Hydrogen from Conversion of Metals
Alkali and alkaline earth metals react with water or steam according to Equation (48) to yield hydrogen. For hydrogen production in the laboratory the reaction of metals with acids or bases is still used
(Eq. 49 and 50).

(48)

(49)

(50)

Aluminum, when the surface has been activated by amalgamation, reacts with hot water with hydrogen evolution [173]. Zinc, that has been activated with copper, can split water vapor.

Hydrogen can be obtained from metal hydrides by heating them as well as by subjecting them to the influence of water ( Hydrides, see Section Hydride Technology). For certain purposes, especially in
out-of-door duties (balloons) hydrogen can be produced from lithium hydride, which is light and easily transported (2.8 L H2 per gram of LiH), from calcium hydride (1.06 L per gram of CaH2) or from
complex hydrides, such as lithium alanate (2.36 L per gram of LiAlH4).

page 59 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The steam – iron process, which is one of the oldest ways of producing hydrogen, has been subject to reawakened interest. In principle, it is a cyclic process for water cleavage, whereby coal is
consumed. The principles of the steam – iron process is shown in Figure 55. In the first step, coal is gasified with steam and air to produce a reducing gas containing carbon monoxide and hydrogen (lean
reducing gas). This gas is reacted with iron oxides in the second stage to produce reduced forms of iron oxides. The reduced species are then reoxidized in the third step of the process with water to form
the original oxides at the same time producing hydrogen.

Figure 55. Principle of the steam – iron process

a) Iron oxide reduction reaction; b) Water cleavage reaction

Characteristic data for hydrogen production by the steam – iron process in generators, as they were previously used, are:

Table 26.
Periodical operation,
cycle time 15 – 30 min
reduction time : oxidation time 2:1
mean reaction temperature 700 – 900 °C
hydrogen production per generator 250 m3/h

Continuous production of hydrogen can be achieved if reduction and oxidation of the solids occur in two separate reaction zones and the solid is transported between them. New process suggestions [174]
incorporate separated fluidized beds, as can be seen from Figure 56. The reduction gas is produced externally in a coal – steam – air gasification stage (a). The main disadvantage of the steam – iron
process is that only 60 % of the reduction potential of the gas is utilized. Power production by both gas and steam turbines and production of process steam may be used to compensate for this.

Figure 56. Process diagram of the continuous hydrogen production by the steam – iron process

a) Reducing gas producer; b) Ash disposal; c) Steam – iron system; d) Reducter; e) Oxidizer; f) Waste heat recovery; g) Desulfurization; h) Dust removal; i) Combustor; j) Air compression; k) Gas turbine; l) Steam generation
with steam turbine

Other redox systems for water cleavage which, according to thermodynamics are more efficient than iron, have been suggested, e.g., Zn – ZnO dissolved in lead, 300 – 500 °C, 10 – 20 MPa [175], and
Sn – SnO2 dissolved in tin, ca. 800 °C, 2 MPa [176].

4.5.2. Hydrogen from Ammonia

Ammonia is easily thermally decomposed according to the reaction:

The endothermic reaction is favored by high temperature and low pressure and is accelerated by the presence of nickel or iron catalysts. From one kilogram of ammonia 1.97 m3 (STP) of hydrogen and
0.66 m3 (STP) of nitrogen are obtained with an energy requirement amounting to 2.7 MJ without consideration of preheating and evaporation. The residual ammonia concentration is usually lower than
500 ppm (mL/m3) depending on the equilibrium conditions. Figure 57 shows the corresponding residual ammonia concentrations at 100 kPa with the temperature as parameter. At higher pressure, a
higher ammonia content remains in the product gas (e.g., 1000 °C, 3 MPa: NH3 residual concentration 0.2 vol %).

page 60 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 57. Residual ammonia concentration corresponding to equilibrium of ammonia formation reaction at 100 kPa

The reaction is conducted commercially at 800 – 900 °C. Standard plants for 1 – 100 m3 (STP)/h are heated by a separate fuel gas and require water to cool down the product gas. The process can also
be conducted autothermally, either by burning a part of the product hydrogen or by preigniting ammonia before the catalyst reactor. This results, however, in a mixture containing less hydrogen (ca.
50 vol %) and containing the product of combustion, water vapor. Usually carbon-free combustion air is used to avoid traces of carbon dioxide and carbon monoxide in the product gas [177].

The hydrogen – nitrogen mixture produced by cracking can be used without further treatment in various areas, e.g., as a protective atmosphere in metal processing (see Section Hydrogen in Metallurgy),
or for catalytic oxygen removal from gases. For the recovery of pure hydrogen on a small scale, suitable process steps can be used, e.g., pressure-swing adsorption, diffusion cells (see Section
Permeation Processes) or hydride technique (see Section Hydride Technology).

The possibility of hydrogen generation from ammonia leads to consideration of the use of anhydrous liquid ammonia as medium for hydrogen distribution and storage (ammonia as energy carrier, see
Section Future Developments). Liquid ammonia contains 17.8 wt % hydrogen, which is 50 % more hydrogen per volume than liquid hydrogen [178].

4.5.3. Hydrogen from Methanol

Hydrogen and carbon monoxide can be obtained from the decomposition of methanol by reversal of the synthesis reaction. The endothermic decomposition occurs at temperatures >700 °C without a
catalyst, or at 300 – 450 °C on catalysts based on copper – nickel or zinc –chromium alloys. Higher hydrogen yields are obtained by catalytic steam reforming of methanol. The process proceeds via initial
methanol decomposition (Eq. 51) followed by a water-gas shift reaction (Eq. 52) as shown in the overall equation (Eq. 53).

Information on thermodynamic and kinetic investigations, suitable catalysts and a summary of literature, is given in [179-181].

The reaction is carried out technically at pressures of 0.7 – 3 MPa on CuO – ZnO or CuO –Cr2O5 catalysts. The molar ratio of water to methanol will vary from 0.67 to 1.5; excess steam lowers the risk of
carbon formation.

Hydrogen production from methanol is used in small to medium-sized plants (some of which are mobile) for the production of 1 – 2 000 m3 (STP)/h. Purification of the product gas is performed by carbon
dioxide scrubbing, pressure-swing adsorption or (especially for small units and high purity), by diffusion through a palladium-silver membrane. Figure 58 shows the schematic of a methanol cracker for
50 m3 (STP)/h of hydrogen at 1.4 MPa using a PSA purification unit to give 99.9999 % purity. 1.5 m3 (STP) of hydrogen are produced per kilogram of methanol.

Figure 58. Methanol reformer with pressure swing adsorption purification for the production of 50 m3 (STP) hydrogen per hour, purity 99.9999 vol % (German Linde)

a) Methanol/water vessel; b) Pump; c) Preheater; d) Reformer; e) Adsorber; f) Heater for heat transfer medium; g) Condensate feedback; h) Cooling water

In Berlin, peak requirements for town gas are covered by methanol cracking. At 1.5 – 20 MPa and ca. 500 °C endothermic methanol reforming reaction and exothermic methanation of the carbon
monoxide takes place and a mixture with town gas quality (ca. 50 % H2, 25 % CH4, 4 % CO, 20 % CO2) is obtained. The investment costs are ca. 40 % lower than for a naphtha reformer and the
operating costs are also favorable, so that the gas production costs are relatively low, for the small number of days with peak loads [182].

4.5.4. Hydrogen from Hydrogen Sulfide

A potential source of hydrogen is hydrogen sulfide which can be obtained in large amounts from scrubbing processes for synthesis and natural gas as well as from desulfurizing processes in refineries.
Normally, hydrogen sulfide is converted to elemental sulfur by means of the Claus process or by other oxidizing processes. The recovery of hydrogen and sulfur from waste hydrogen sulfide is an
interesting alternative to the manufacture of sulfur alone.

page 61 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The decomposition reaction is endothermic. At temperatures <1800 K the equilibrium position is thermodynamically unfavorable with respect to hydrogen formation (see Table 27) [183].

Table 27. Calculated equilibria concentration of hydrogen in the decomposition of hydrogen sulfide at 100 kPa [191]

Temperature, K Equilibrium concentration,

mol %

400 1.3×10–3
500 6.1×10–3
600 2.0×10–2
700 4.8×10–2
900 1.9
1073* 7.1
1200 13.1
1400 25.6
1600 37.7

* K = 1.45×10–2.

The reaction proceeds readily in the presence of catalysts, e.g., platinum – cobalt (at 1000 °C) [184], disulfides of molybdenum or tungsten (at 800 °C) [185], or other transition metal sulfides, preferably
supported on alumina (at 500 – 800 °C) [186].

Development work to increase hydrogen yield and thus to establish commercial use of hydrogen sulfide decomposition, suggests several approaches to avoid thermodynamic restrictions.

Cyclic Processes. Thermodynamically favorable cycles have been suggested, e.g.:

(54a)

(30.0 MPa, 300 K)

(54b)

(600 K, 70 MPa, activated carbon) [187]

(55a)

(470 K, 100 kPa, over COS, NiS)

(55b)

(1000 K, 100 kPa) [188]

(56a)

(56b)

(56c)

(suggested [189])

Equilibrium Displacement. The hydrogen yield can be considerably enhanced by removing either sulfur or hydrogen from the reaction system, thereby shifting the equilibrium. A variety of methods have
been considered for the separation of hydrogen from hydrogen sulfide in hot gas streams. Examples are the use of zeolitic, polymeric, metallic, glass and ceramic oxide membranes, pressure-swing

page 62 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
adsorption and thermal diffusion [190]. The hydrogen yields can be increased to twice the equilibrium value by diffusion through Vycor-type glass membranes or a microporous alumina membrane at
1000 °C without degradation of the latter [191]. Laboratory experiments using thermal diffusion columns at moderate to high temperature showed effective separation of hydrogen sulfide decomposition
mixtures yielding hydrogen concentration > 90 % [190].

Economic realization of hydrogen sulfide cleavage would provide commercial products and offer the possibility of replacing Claus processes and of recycling hydrogen in refinery hydrodesulfurization
processes. However, energy has to be provided at relatively high temperatures. Research is focused on (1) thermal decomposition processes using concentrated solar radiation or (2)
photoelectrochemical hydrogen sulfide splitting by irradiation with visible light using colloidal semiconductor electrodes (see Section Photochemical or Photoelectrical Water Cleavage, [192]).

4.5.5. Hydrogen as Byproduct

In industrial organic chemistry reasonable amounts of hydrogen are often generated. The amounts are particularly large in the petroleum refineries within the area of catalytic reforming. The most
important reactions here are thermal or catalytic cracking, dehydrogenation and oxidation (see Table 28 and Sections Refinery Processes and Petrochemical Processes).

Table 28. Reactions in industrial organic chemistry providing hydrogen

page 63 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
4.6. Economic Aspects
At present 77 % of the hydrogen produced is from petrochemicals, 18 % from coal, 4 % by electrolysis of aqueous solutions and at the most 1 % from other sources. A review of the cost structure of
hydrogen production will, therefore, pay particular attention to the production methods described in Section Production from Coal and Hydrocarbons. Electrolytic processes are also used under certain
economic circumstances and are, therefore, included here and in a prognosis regarding the future supply of industrial hydrogen. For predictions of hydrogen production costs for use as an energy carrier
see Chapter Hydrogen Energy.

4.6.1. Analysis of the Cost Structure

Hydrogen and hydrogen-rich synthesis gases (in particular H2 – CO and H2 – N2) are either produced directly in plants especially built for that purpose or are recovered as a byproduct. In areas with highly
concentrated industrial sites, such as mid-Europe, the ratio of directly to indirectly produced hydrogen is almost 1 : 1 [193]. In the developing countries and in countries with industrial sites widely separated
from one another, the fraction of directly produced hydrogen is much larger than the fraction recovered in combined plants.

According to the broad range of feedstocks and the highly developed production processes from each of these feedstocks, an analysis of the cost structure depends on the local conditions, i.e., on the
availability and price of the feedstocks. In many countries feedstock prices, such as those used for the direct production of hydrogen, are subject to intervention by the government. Heavy taxation is the
rule, but subsidies also occur. Further uncertainties in the cost comparison are the various safety and environmental standards in each country and region which to a lesser extent, however, lead to
different hydrogen costs.

For the conditions in the Federal Republic of Germany in 1983 costs for direct production of hydrogen have been calculated [194]. The feedstock prices (Fig. 59) include costs for the feedstock itself and,
because in the cases considered the feedstock is also used as fuel, the fuel costs. The rest are capital, material and other energy costs (without the feedstock), personnel and administration costs. They
lead to a typical situation, which is also found for other countries in which the feedstock is available at world market prices: Steam reforming of natural gas exhibits the lowest investment costs.

Figure 59. Feedstock and production costs for hydrogen in the Federal Republic of Germany in 1983 in DM/1000 m3 (STP) H2; numbers in columns converted to $ by the factor 1.8 (scale on the left)

plain area = costs of feedstock; hatched area = costs of production

The costs of hydrogen production from lignite are almost similarily low, whereby the low feedstock prices for surface-mined coal offsets the high cost of the plant. Whereas the much higher cost of
hydrogen from naphtha or heavy fuel oil is a situation found worldwide, the high cost for hydrogen from hard coal is a special case limited to the Federal Republic of Germany, and reflects the special
coalmining conditions there. On the world market the cost for coal is usually only slightly higher than the cost for lignite, so that the cost for hydrogen from hard coal should be much less.

The influence of the feedstock price is shown in Figure 60. The dashed lines correspond to the feedstock prices and the resulting hydrogen production costs in the Federal Republic of Germany in 1983
and correspond with the data given in Figure 59.

Figure 60. Production costs for H2 as a function of feedstock price (Federal Republic of Germany, 1983)

As long as local conditions do not lead to large changes in the capital-dependent costs (ordinate of Fig. 59) the given presentation can be used to give a rough estimate of hydrogen costs and
comparisons between various feedstocks by substituting local and actual prices.

The results in Figure 59 and 60 are for a capacity of 100 000 m3 (STP)/h, which is a very large plant, although not unusual. If the dependence of the hydrogen production cost on plant capacity is required,
then the simplified assumption may be made that the utility costs are independent of the capacity and that the investment costs become relatively less important as long as a large one-train plant can be
built. If a modular construction is used, the investment costs do not decrease with increasing plant size. The upper limits for one-train units are ca. 500 m3 (STP)/h for electrolysis units, ca. 80 000 m3
(STP)/h for partial oxidation and ca. 100 000 m3 (STP)/h for steam reforming plants.

Gas purification plants, in particular pressure-swing adsorption plants, are built in modules from ca. 50 000 m3 (STP)/h upwards, so that a reduction in costs on capacity increase for larger plants is not to

page 64 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
be expected.

For comparison of hydrogen plants with a capacity of ca. 10 m3 (STP)/h, it makes sense to choose plant types that would not be considered for very large plants, e.g., electrolysis plants or methanol
reformers. Figure 61 shows the hydrogen costs as a function of the capacity for Europe in 1983. In addition to these production processes solely for hydrogen, the costs of recovery of hydrogen by means
of pressure-swing adsorption from a hydrogen-rich refinery off gas are also given. In this case only the difference in heating value of the refinery gas before and after the hydrogen removal is used for
calculation of the feedstock price. For the investment cost the cost of a pressure-swing adsorption unit without any further prepurification can be taken. The range of costs is caused by the additional
expenditure for feedstock compression to adsorption pressure. Plants of this type up to a capacity of ca. 100 000 m3 (STP)/h feed gas have been built.

Figure 61. Production costs for H2 as a function of the production process and plant capacity [195]

Basis of the cost determination:Investment costs: 1983, Federal Republic of Germany

Natural gas: 8.5 DM/GJ, HHV

Fuel gas: 8.5 DM/GJ, LHV

Power: 0.09 – 0.15 DM/kWh

Naphtha: 650 DM/t

Electrolysis: boundary limits 0.15 – 0.09 DM/kWh

Methanol reforming: boundary limits 525 – 400 DM/t

Steam reforming: boundary limits naphtha – natural gas

Partial oxidation: boundary limits heavy fuel oil 300 DM/t,residues 100 DM/t

A similar curve for U.S. prices of hydrogen was developed in 1978 with forecasts until 1980 [197]. Figure 61 also contains average commercial prices for 5580 m3 (STP)/h merchant hydrogen which
exceed prices for electrolytically produced hydrogen from a plant size of 0.1×106 scf/d (see Section Future Developments) on up.

The consequences of part-load operation of the plant also determines the choice of the hydrogen-production process. For a low-capacity rating, processes with low investment costs are preferred,
because in this case the product cost depends more or less directly on the necessary feedstock costs and is not influenced by writing off an expensive high-capacity plant. For hydrogen plants in the lower
capacity range (ca. 100 – 500 m3 [STP]/h) electrolysis or methanol reforming is the best choice. For large plants this factor leads to the choice of steam reforming plants using natural gas.

4.6.2. Prognosis of the Hydrogen Costs

A short-term prognosis of the hydrogen cost development for industrial use (see Chap. Hydrogen Energy for hydrogen as energy carrier), should be oriented on the present stable supply situation and
technically well-proven processes. With the aid of Figure 60 such a prognosis can be made. It shows that in the future hydrogen will still be preferentially produced on a large scale from hydrocarbons. The
prognosis can therefore be reduced to investigating price and availability of hydrocarbons in the coming years.

Should it become possible to decouple the price of electrical energy from the price of fossil fuels by using nuclear energy or alternative energy sources and to make electrical energy relatively cheap, the
electrolysis process could play an increasingly important role. Figure 62 shows what price changes must occur to assure electrolysis a greater market fraction on a purely economic basis.

Figure 62. Cost prospects for electrolysis as compared to the main process for hydrogen production, steam reforming of natural gas [196]

a) Conventional electrolysis; b) Advanced electrolysis; c) Natural gas; d) Natural gas price (Mid Europe); e) Price range for electrical energy (Mid Europe)

[Top of Page]

5. Purification of Hydrogen
In many cases it is necessary to purify technical grade hydrogen by using a variety of processes. These must be chosen in accordance with the final use of the hydrogen to give an economically

page 65 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
satisfactory result. Chapter Purification of Hydrogen gives an overview of the technically important purification steps. They are classified according to the physical and chemical operations involved in the
separation processes.

Frequently in industrial-scale plants combinations of the various processes are used. In addition, several alternative processes may be available and may be used depending on the economic situation. In
Chapter Production and Chapter Handling of Hydrogen some examples of typical combinations of various gas production and purification processes are given.

5.1. Low-Temperature Processes

The large temperature difference between the boiling point of hydrogen and that of the most frequent impurities is the basis of the low-temperature process (LT process) for the purification of hydrogen.
Table 29 shows important properties of hydrogen and the most frequent impurities present in the hydrogen to be treated. The impurities are condensed or sublimated. Low-temperature rectification
processes are used, if the impurities are also to be separated from one another. The use of scrubbing and adsorption processes at low temperatures leads to both high purities and high yields.

Table 29. Boiling points at atmospheric pressure, triple point and critical data of hydrogen and some important components which can be present in raw hydrogen streams

Component Boiling point at 101.3 kPa Triple point data Critical data

Temperature, Temperature, Pressure, Temperature, Pressure,

K °C K kPa K kPa

He 4.23 –268.9 2.17b 5.1 5.3 230

H2a 20.4 –252.8 13.9 7.0 33.3 1 320

Ne 27.1 –246.1 24.6 43.3 44.5 2 760

Ar 87.3 –185.9 83.8 68.8 150.7 4 870
N2 77.3 –195.9 63.2 12.6 126.1 3 390
O2 90.2 –183.0 54.4 0.2 154.3 5 040
CO 81.6 –191.5 68.1 15.4 133.0 3 500
NO 121.2 –151.9 109.9 2.3 179.2 6 590
H2O 373.2 100.0 273.2 0.6 647.4 22 140
H2S 212.9 – 60.3 187.6 23.2 373.6 9 020
SO2 263.2 – 10.0 197.8 1.7 430.4 7 870
CO2 194.7c – 78.5 216.6 18.0 304.2 7 390
NH3 239.8 – 33.4 195.4 6.1 405.6 11 300
CH4 111.7 –161.5 90.7 11.7 190.7 4 620
C2H6 184.6 – 88.6 89.9 <0.1 305.3 4 890
C2H4 169.5 –103.7 104.0 0.1 282.0 6 280
C3H8 231.1 – 42.1 85.4 <0.1 370.0 4 260
C3H6 225.5 – 47.7 88.0 <0.1 365.3 4 570

a In more detail in Section Physical Properties.

b Upper point.
c Sublimation

An overview of the basic physical processes which are used for the liquefaction of gases as well as the most important parts of the plants, equipment and machines is given in [198] and [199]. Properties
of two- and multi-component mixtures with hydrogen, which are the basis for the calculation of the separation processes given in this chapter, can be seen in Figures 8, 9 and 10 (see Chap. Properties).

Such property data are available in computer data banks of engineering companies. Some data otherwise available is given in [200], [201].

page 66 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
5.1.1. Condensation Processes
Simple condensation processes are used, if the difference between the boiling point of hydrogen and the condensing phase, usually hydrocarbons, is sufficiently high. Hydrogen is frequently separated
from refinery and dealkylation off-gases, as well as ammonia and methanol synthesis purge gases by the condensation process.

The gas from which hydrogen is to be recovered, usually contains components which are solid at the temperature of the final condensation step. Therefore, a prepurification stage is usually required to
prevent blockage of heat exchangers, knock-out drums, and piping in the low-temperature section of the plant.

Prepurification and drying remove in particular impurities with a high boiling point such as water, carbon dioxide, and hydrocarbons such as benzene and toluene. These processes (often carried out in
stages) are scrubbing (e.g., glycol scrubbing for water removal), adsorption (e.g., on activated carbon or molecular sieve beds), or condensation and sublimation of the impurities on reversible regenerator
beds or reversing heat exchangers (Revex).

The residual gas mixtures, usually hydrogen – methane, hydrogen – carbon monoxide, or hydrogen – nitrogen mixtures, are cooled down below the condensation temperature. The lower the temperature
chosen, the purer is the hydrogen product. In the case of the mixture hydrogen – methane the temperature in the condensator can be lowered below the solidification point of pure methane by the addition
of ethane or propane and the hydrogen purity can thereby be increased.

The hydrogen yield depends on the solubility of the hydrogen in the condensate and on the amount of condensate. The yield of hydrogen increases, if the condensate is depressurized in a series of stages
and the hydrogen-rich depressurization gas is recycled to the first stage of the compressor. Figure 63 shows the most important ways of generating refrigeration for low-temperature processes in industrial
plants.

Figure 63. Principle ways of producing refrigeration for low-temperature hydrogen purification

A) Joule – Thomson expansion of residual gas; B) Hydrogen expansion in a turbine; C) Cold from external refrigeration sources (NH3, freon, ethylene, etc.); D) Refrigeration cycle with residual gas

If amount and pressure of the impurities, such as methane, carbon monoxide or nitrogen, are large enough (> 3 – 5 vol %), the refrigeration requirements of the plant can be covered by depressurizing the
condensate in a throttle (Joule – Thomson effect) (Fig. 63 A). Hydrogen is then recovered at high pressure. If the nitrogen or methane content is too low or variable, so that the refrigeration requirements
cannot be covered reliably, then hydrogen itself can be expanded, e.g., by producing work in a turbine (Fig. 63 B).

If hydrogen, however, is required under high pressure, external refrigeration must be used. This can be done by a conventional external refrigeration plant using ammonia or freon as refrigerant
(Fig. 63 C), or using other agents that are suitable for the required temperature level of condensation, in this case, e.g., nitrogen, carbon monoxide or methane (Fig. 63 D). The latter possibility requires the
most sophisticated equipment. However, it leads to the greatest plant flexibility with regard to throughput and concentration of the impurities, because the refrigeration load is controlled directly by means
of the external refrigeration cycle.

Some examples of processes using condensation alone are:

1. Refinery-gas purification plants, in which refrigeration is by a methane cycle or by decompressing the condensate (Fig. 63 A) with partial depressurization of the product [202].
2. Purification of the off-gas of the toluene dealkylation. Here a two-stage condensation is used for the recovery of pure hydrogen, a C2, and a methane fraction. The refrigeration is supplied by an
ammonia or freon cycle if necessary [203].
3. Recovery of hydrogen from the purge gas of ammonia plants. Here, a one-stage condensation process is sufficient. This process is particularly economic, when argon is recovered as an additional
product by fractionation. A nitrogen cycle is used as refrigerant [204], [205].

Further common variations of the condensation process are as follows:

1. Staged decompression of the condensate and recycle of the decompressed gas to increase the hydrogen yield.
2. Partial condensation to fractionate the condensate and, eventually, to avoid the precipitation of solids. In this case the refrigeration can be supplied at a higher temperature, thus saving energy, by
using refrigeration cycles which are optimized for the various condensation temperatures (see Fig. 63).

5.1.2. Condensation and Rectification

In large-scale plants higher yields and greater product purity are obtained by adding a rectification stage to the condensation step. This is also an advantage, if the impurities must be separated from one
another. For the refrigeration many tailored variations exist.

Ethylene Plants. Pure hydrogen is recovered as a byproduct after low-temperature separation of ethane and ethylene from the methane – hydrogen fraction. The demethanizer plays an important part in
the economics of the plant because the refrigeration required is at the lowest temperature of the process ( Ethylene – Ethylene Fractionation.).

page 67 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Coal Liquefaction. Hydrogen is recovered from the off-gas of the Fischer – Tropsch synthesis (operated on a large scale in South Africa by SASOL) by means of low-temperature condensation and
rectification processes.

Figure 64 shows the LT separation of the purification unit of the SASOL II plant. In this unit 8675 mol/s (700 000 m3 [STP]/h) crude gas is separated into 7 fractions, 5 of them in a LT plant. The hydrogen
fraction is ca. 3700 mol/s (300 000 m3 [STP]/h) of 80 % pure H2 [206]. The crude gas, freed from carbon dioxide, is cooled down close to the hydrate point (i.e., the temperature at which precipitation of
solid gas hydrates occurs) [207]. The water condensate is removed and separated into a water and a C3+ fraction. The gas phase is dried in an adsorber and cooled down to –35 °C. The condensate
produced in this stage is fed to the rectification column g and separated into a C2 and a C3+ fraction. In the following condensation stage a C2 fraction is recovered by cooling down to ca. –95 °C, in the
final stage at –150 °C liquid methane is produced. Methane, the C2 fraction and the C3 product are separated in two columns. The gaseous phase in the final knock-out drum e5, is ca. 80 % pure
hydrogen.

Figure 64. Production of H2 and hydrocarbon fractions from Fischer – Tropsch synthesis off-gas by staged condensation. Separation of the hydrocarbon fractions by low-temperature rectification [206]

a) Adsorber for drying and elimination of traces of long-chain hydrocarbons; b) Heat exchanger; c) Compressor; d) CH4 recycle compressor; e1 – e5) Knock-out drums; f) Pump; g) Column

This purity is suitable for the Fischer – Tropsch synthesis for which the main portion of the hydrogen-rich gas is required. A smaller part of the gas is used for hydrogenation purposes and is purified to
99.5 % using a pressure-swing adsorption process (see Section Pressure-Swing Adsorption (PSA)).

The refrigeration for the LT plant is supplied by an ethylene-propylene cascade and a methane cycle (see Figs. 63 C and 64, compressor d). The refrigeration must compensate for losses to the
surroundings, in the heat exchanger and in the liquid product.

The refrigeration is supplied in 7 stages (including methane cycle) between + 9 °C and – 150 °C. The ethylene-propylene cycle has a power requirement of ca. 23 MW.

Low-temperature processes have been proposed for purification of the flash gas which is recovered on the decompression of the coal oil in plants for coal liquefaction by means of hydrogenation.
Compared to other alternatives such as scrubbing at low temperature or diffusion separators [208], this concept has shown to be superior with regard to yield, energy requirements, economics and
feasibility.

Hydrogen – Carbon Monoxide Recovery. Figure 65 shows a low-temperature process, which simultaneously delivers hydrogen and carbon monoxide of high purity from converted gases and gases
from partial oxidation. After leaving a carbon dioxide scrubber, the converted and non-converted crude gas is (1) freed from traces of carbon dioxide and water in an adsorber, (2) cooled in two heat
exchangers and in the sump of the carbon monoxide column down to the condensation temperature of carbon monoxide, i.e., ca. 80 K.

Figure 65. H2 production from synthesis gas by low-temperature condensation and pressure-swing adsorption and production of pure carbon monoxide by rectification (German Linde)

a) Preadsorber; b) Heat exchanger; c) Condensation of CO and CH4; d) Expansion turbine; e) Joule – Thomson expansion for H2 recycle; f) CO/CH4 rectification; g) Compression of expanded gas and PSA gas

Energy and refrigeration is produced by expanding hydrogen in a turbine, thereby expanding down to the inlet pressure of a PSA plant in which hydrogen of 99.9 % end purity can be recovered. The
carbon monoxide content in the hydrogen is <0.5 ppm. Liquid carbon monoxide is decompressed stagewise, whereby the major part of the refrigeration requirements of the plant is covered. After warming,
the carbon monoxide is fed into a rectification column and purified to about 99 %, the maximum hydrogen impurity thereby being 0.5 %. A plant corresponding to Fig. 65 for a gas throughput of 833 mol/s
(67 250 m3 [STP]/h) synthesis gas has been constructed in two trains. The purge gas of the PSA plant can be recycled to obtain a higher yield of hydrogen and carbon monoxide [209].

Hydrogen – Nitrogen Rectification. Whereas in low-temperature hydrogen purification processes the rectification is employed principally to purify the byproducts, large-scale industrial processes also
exist in which hydrogen itself is directly involved in the rectification. In the Braun purifier (part of an Ammonia synthesis process, using steam reforming of light hydrocarbons) ( Ammonia) [210]
ammonia synthesis gas is produced by condensing excess nitrogen from the raw gas which contains hydrogen, carbon monoxide, methane and nitrogen. Figure 66 shows the rectification of the hydrogen-
nitrogen mixture. The refrigeration is provided by the decompression under production of energy in turbine (b) and by the Joule – Thomson decompression of the remaining gas in a throttle. A similar

page 68 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
process for the production of ammonia synthesis gas from nitrogen rich blast-furnace gas or pit gases has been described [211].

Figure 66. Rectification of an H2/N2 mixture and production of NH3 synthesis gas (molar ratio H2 : N2 = 3 : 1) in the Braun purifier [210]

a) Heat exchanger; b) Expansion turbine; c) Rectification column; d) Joule – Thomson expansion

Ammonia Purge Gas Separation. A low temperature process for ammonia plant purge gas purification features especially high yields in ammonia recovery and in recovery of hydrogen as a product and
for recycle. The application of this process is especially advantageous if the low temperatures occurring in the initial condensation step are utilized in further stages to recover other valuable components
such as noble gases and methane [212], [213]. Figure 67 shows the integration of the low-temperature process into the ammonia plant. Several plants have been built for the recovery of liquid argon. If
only hydrogen is to be recycled, adsorption is a competing process (see Section Adsorption Processes). Combinations of both processes have also been proposed [209], [214].

Figure 67. Integration of the low-temperature process into an ammonia plant [215]

a) Compressor; b) Recycle compressor; c) NH3 condensate separator; d) Expansion turbine

5.1.3. Low-Temperature Absorption Processes

Absorption processes can yield greater purities than the condensation processes. Absorption processes are operated at higher temperature and thus with lower refrigeration requirements. The low-
temperature processes described in this section are physical absorption processes, in which the impurities are condensed and either completely or partially utilized as scrubbing agent (for absorption
processes with other scrubbing agents see Section Scrubbing Processes).

Nitrogen Wash. A high-purity hydrogen – nitrogen mixture for ammonia synthesis is obtained by using a liquid nitrogen scrubber ( Ammonia) [216]. This scrubber can not only reduce the catalyst
poison carbon monoxide down to a few ppm but the inert gases such as methane and argon can also be removed. The methanation step can thus be avoided.

For modern ammonia processes, in which the synthesis gas is produced at high pressure, the nitrogen scrubber is particularly interesting (see Section Gasification of Coal and Hydrocarbons) At low
temperatures around 80 K the system hydrogen – nitrogen is still in the subcritical range, although pure nitrogen under these conditions is already supercritical. The necessary refrigeration for the nitrogen
scrubber is produced by decompression of the nitrogen into the hydrogen stream.

Methane Scrubber. For the recovery of pure hydrogen from hydrogen – carbon monoxide mixtures (e.g., from synthesis gas) the use of a methane scrubber is frequently an alternative to the
condensation of carbon monoxide. Methane scrubbers attain much higher hydrogen purities, especially with regard to the carbon monoxide content, than would be possible with the carbon monoxide
condensation process. Simultaneously, pure carbon monoxide can be produced. Figure 68 shows the most important process stages of a methane scrubber [217].

Figure 68. Plant for the recovery of hydrogen and carbon monoxide from H2/CO-rich gases using a CH4 scrubbing process

a) Adsorber for removal of H2O and CO2; b) Heat exchanger; c) Scrubbing column for CO with subcooled liquid methane; d) Knock-out drum; e) Separation of loaded wash methane to liquid lean methane (bottom) and pure
carbon monoxide (top); f) Pump; g) Expansion turbine; h) Compressor

Crude gas between 1.5 and 4.5 MPa is dried in adsorber (a) and cooled down to ca. 90 K in a heat exchanger (b). The carbon monoxide remaining in the gas phase after condensation is removed down to
a few ppm in the scrubber (c) with subcooled liquid methane. The liquid phase from the knock-out drum (d) and from the column bottom after decompression and partial evaporation is fed into the carbon
monoxide – methane separation column (e). Here the sump product is pure methane, which is used for scrubbing, whereas a carbon monoxide fraction is obtained as the head product. A carbon
monoxide cycle provides the heating and the reflux for the separation column as well as the cooling for the scrubbing column. By decompressing a fraction of the carbon monoxide cycle and generating
energy the refrigeration for the heat exchanger is obtained.

page 69 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
In a similar fashion, however without a carbon monoxide – methane separation column, methane scrubbers are used for the purification of refinery off-gas or for the production of pure hydrogen from the
purge gas of hydrogenation plants (coal hydrogenation, heavy oil hydrogenation). Similarly, as shown in Figure 68, methane scrubbers combined with adsorption processes can be used for the recovery of
pure hydrogen, oxo-synthesis gas and a methane-rich heating gas fraction [213].

Other Scrubbing Agents. Absorption processes at low temperatures are not confined to the aforementioned scrubbing agents. Higher hydrocarbons such as ethylene or propane have also been used.
Hydrogen itself can also be used as a scrubbing agent if the product desired is helium. Processes of this kind have been proposed and plants built for the recovery of helium and neon from air separation
units and from ammonia purge gas streams [218]. In both cases the noble gases are already preconcentrated by several orders of magnitude compared to their natural abundance.

5.2. Adsorption Processes

5.2.1. Principles
Because of the only slight interaction of hydrogen with the usual adsorbing agents in technically important pressure and temperature ranges, adsorption processes are particularly suitable not only for the
separation of hydrogen from other gases but also for the separation of impurity traces from hydrogen. For the principles of adsorption see Adsorption [219-221].

The adsorbents for hydrogen purification are chosen according to the impurities to be removed. The most common adsorbents are given in Table 30. The gas mixtures are separated by various effects.
Steric effects prevail if the adsorbent has a uniform and narrow micropore distribution, as is found particularly with the zeolites. Equilibrium effects result from different adsorption forces of the gases on the
solids, and kinetic effects are caused by the different diffusion mobilities of the gases into the pores of the adsorbing agents.

Table 30. Adsorbents for the purification of raw hydrogen

Type Main adsorption duty

Aluminum oxides water

Silica gel water, CO2, C4+
Activated carbon CO2, CH4, C2+, N2
Molecular sieve (zeolite) CH4, CO, N2
Carbon molecular sieve O2

Although each of these effects is present to a various degree in the adsorbent-gas combinations given in Table 30, the strength of the adsorption forces of the various impurities in the raw hydrogen
mixtures can be used to put them in the order shown in Table 31. The sequence shown there may change slightly depending on adsorption agent, partial pressure of the impurity and total gas pressure of
the system.

Table 31. Adsorption forces for various components in hydrogen purification

page 70 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

The static equilibrium loading capacity of the adsorbent for a particular component is depicted in the form of an adsorption isotherm (loading capacity depending on partial pressure at constant
temperature). Adsorption isotherms have typical values for the adsorbents and the components in question and must be determined experimentally. For a number of typical components and some
adsorbents these values have been compiled in handbooks and in specification sheets of the various producers.

Figure 69 shows the principal adsorption isotherms of various gases present during hydrogen purification on carbon molecular sieve (active charcoal obtained from coal). There is a characteristic affinity
between each particular gas and the solid.

Figure 69. Principle form of equilibrium isotherms of mixtures of hydrogen with nitrogen, carbon monoxide, and methane [222]

Adsorbent material: carbon molecular sieves (activated carbon obtained from coal)

Adsorption is an exothermic process. Desorption, subsequently, is endothermic. In thermally regenerated adsorbers the energy necessary for the desorption is obtained by indirect heating of the bed or by
using hot purge gas. In pressure-swing adsorption the desorption energy is obtained from the energy stored in the adsorption bed (thereby causing a temperature drop) if the pressure or the partial
pressure of the components to be desorbed is decreased.

Hydrogen purification takes place in adsorption vessels with fixed beds of adsorbents. Other types, such as trickle beds, adsorption columns with solid circulation, moving beds, or adsorption in ebullating
beds, are not used for hydrogen purification on a technical scale.

5.2.2. Thermally Regenerated Adsorbers

Thermally regenerated adsorbers are used to remove traces of impurities from hydrogen streams or for selective removal of certain substances which disturb ensuing process steps. The process is often
designated as temperature-swing adsorption (TSA).

The adsorbers are usually switched alternately. While one adsorber is in the adsorption stage, the other is regenerated or is held in readiness. Switching usually takes place in a defined sequence,
whereby the time for the adsorption step must be chosen shorter than the time calculated for the breakthrough of the impurities. Typical cycle times for these processes are in the range of some hours up
to several days.

The process is illustrated in Figures 70 and 71 for the adsorption of carbon dioxide in a hydrogen-rich synthesis gas on a molecular sieve (5 A). As shown in Figure 71, the adsorber becomes loaded to
give a local concentration distribution which is indicated by curve a. The maximum adsorbed amount of carbon dioxide in the zone of the equilibrium concentration EZ is given by the equilibrium isotherm
for 25 °C at point a in Figure 70.

page 71 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 70. Adsorption and regeneration of CO2 and molecular sieve 5 A, pellets

a) Adsorption loading; b) Desorption loading after decomposition and heating up; c) Theoretical desorption loading after decompression; d) Desorption loading after heating up without decompression

pA = adsorption pressure; pD = desorption pressure

Figure 71. Concentration profiles in an adsorber (MTZ = mass transfer zone; EZ = equilibrium loading zone)

a) End of adsorption; b) End of cocurrent expansion; c) End of regeneration (residual loading); d) Useful cyclic capacity

In most cases regeneration is initiated by depressurizing. This procedure causes a desorption along the 25 °C isotherm to point c. However, because the depressurization is not isothermal, a cooling effect
occurs, so that besides the kinetics of the desorption the adsorption conditions at the lower resulting temperature must also be considered. For a so-called cocurrent depressurization the distribution of the
concentration in the adsorber is given by curve b of Fig. 71. The equilibrium along the isotherm is attained very slowly. The desorption energy requirements caused by the depressurization lead to a strong
cooling of the adsorbent, which in the presence of water vapor can lead to formation of a liquid phase and thus to an irreversible damage of the adsorbent.

The main desorption takes place after the pressure has been lowered (point c in Fig. 70) mainly by increasing the temperature, in this case to 200 °C (point b or, if no previous depressurization took place,
point d). The required heat is provided either indirectly by means of inserted heat exchanger coils or by direct heating with a purge gas at higher temperature. After regeneration, the adsorber is cooled to
normal operating temperature and pressurized back to normal adsorption pressure.

Applications. Before low-temperature purification and separation can be carried out (see Section Low-Temperature Processes), gas components, which would at low temperatures form solid deposits in
the heat exchangers and knock-out drums, must be removed. In synthesis gases from partial oxidation and steam reforming these are in particular water and carbon dioxide, whereas in refinery gases
these are usually higher hydrocarbons. In dealkylation and coke oven gases benzene, toluene, and other aromatics must be removed, whereas in sulfur-containing raw gases hydrogen sulfide and
carbonyl sulfide are the disturbing components.

If these impurities are present in higher concentrations, the major portion may first be removed by condensation (water and hydrocarbons) and/or scrubbing (CO2, H2S and COS). Final purification is then
by TSA. A process description of the removal of water and carbon dioxide from a water-rich synthesis gas by TSA is given elsewhere ( Gas Production – The “Classic” Method).

Further applications for TSA in hydrogen purification processes are:

1. Removal of water and ammonia from ammonia purge gases before low-temperature separation.
2. Purification of hydrogen from electrolysis plants. The main impurities are water and oxygen. The latter is converted by means of catalytic oxidation (see Section Catalytic Gas Purification) to water
and then removed by condensation and TSA. Hydrogen from the chlor-alkali electrolysis contains also traces of mercury. This can be removed by TSA on a special molecular sieve [223].
3. Removal of various impurities from pure hydrogen before liquefaction. Here the adsorption is carried out at liquid nitrogen temperature. The regeneration is done with pure hydrogen from the recycle
gas at ambient temperature (see Chap. Liquefaction of Hydrogen).

5.2.3. Pressure-Swing Adsorption (PSA)

Adsorption Mechanisms. Contrary to adsorption with thermal regeneration, pressure-swing adsorption operates at almost constant temperature using partial-pressure differences. However, the small
temperature gradients occurring with ad- and desorption contribute considerably to the mechanism of the process. The components to be removed are adsorbed at high system pressure and high partial
pressure and are desorbed at low system pressure in accordance with the appropriate equilibrium isotherms of the various impurities (see Fig. 69).

Figure 72 shows, for the example of a hydrogen – carbon monoxide mixture, how carbon monoxide, which is easily adsorbed, reaches the equilibrium loading, a, on the molecular sieve, while pure
hydrogen exits the adsorber. When the adsorber capacity is exhausted, the adsorbent must be regenerated. The carbon monoxide loading on the absorbent is reduced by depressurization and hydrogen
purging. The difference c is the available working capacity of the adsorbent. However, as in the case of TSA, the actual working capacity of the total adsorbent is influenced by the presence of mass
transfer zones (MTZ) (see Fig. 71). The length of a MTZ is basically determined by (1) the concentration of the component to be removed in the raw gas, (2) by the required product purity, and (3) by
kinetic effects.

page 72 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 72. Adsorption isotherms for CO on molecular sieve

a) Adsorption; b) Desorption; c) Adsorption capacity

The PSA Process. Like the TSA process, the PSA process is cyclic, however, contrary to temperature swings, pressure changes can be carried out more rapidly and allow a shorter adsorption time.
Typical cycle times are in the range of 3 to 10 min. Thus, removal of impurities present in high concentrations can be accomplished using moderately sized adsorber vessels. The essential steps carried
out in a PSA unit are shown in Figure 73 [224]. Each adsorber is performing a different task, depending on which phase of the complete PSA cycle is regarded. The chronological sequence of all steps
constitutes the cycle of one adsorber:

1. Adsorption at the pressure of the raw gas (highest system pressure), hydrogen production (adsorber A);
2. Cocurrent depressurization, provision of gas for pressure equalization and purging (adsorbers F and E);
3. Countercurrent depressurization, production of part of the tail gas (adsorber C);
4. Countercurrent purging (lowest system pressure), production of another part of the tail gas (adsorber D);
5. Countercurrent repressurization (adsorber B).

Figure 73. Six-bed PSA unit. Main flows for adsorption, repressurization, counterflow expansion (dump), counterflow purge, etc. for H2 production by adsorber A. After termination of this adsorption step, next on-stream
adsorber is B, etc.

Adsorber A B C D E F
Function Adsorption Repres- Dump Purge Depres- Depres-
surization surization surization
Start pressure, (MPa) 2.8 1.6 0.4 0.2 0.8 2.8
Final pressure, (MPa) 2.8 2.2 0.2 0.2 0.4 2.2

Figure 74 [225] shows the different process stages with respect to the equilibrium loading on the adsorbent. The possibility of continuing the decompression stage into vacuum is also shown. This latter
mode of operation is indicated for the enrichment and recovery of the adsorbed component. However, production of concentrated byproducts during the production of pure hydrogen is usually carried out
by low-temperature processes and not by the adsorption process.

Figure 74. Theoretical pattern of the status of PSA adsorbers in respect to loading during one adsorption cycle [225]

a) Adsorption; b) Cocurrent depressurization; c) Countercurrent depressurization; d) Purge; e) Depressurization to vacuum; f) Repressurization; g) Equilibrium adsorption isothermal

pA = adsorption pressure, p = purge gas pressure; pv = final vacuum obtained

The mode of operation of a PSA can be summarized as follows:

1. Pure hydrogen is produced continuously at adsorption pressure.

2. Tail gas is produced continuously, the pressure, composition, and amount of which fluctuate periodically. These fluctuations are smoothed out by buffer vessels.
3. The efficiency of a PSA is defined by the amount of pure hydrogen produced in relation to the total hydrogen amount present in the crude gas. Depending on the crude gas composition and the
degree of technical sophistication utilized, the PSA efficiency ranges between 70 and 90 %.

Technical PSA Plants. The simplest PSA system requires three adsorbers to enable continuous operation. Hydrogen yields of > 80 % can only be attained in systems with more than 4 adsorbers.

Since the size of the adsorber is usually limited because of manufacturing or transport considerations, larger quantities of crude gas can only be treated if several adsorbers are used in the adsorption

page 73 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
phase simultaneously (multi-bed process) [226], [227]. With such schemes PSA plants using 3 to 12 adsorbers can process amounts of crude gas ranging from 100 to more than 100 000 m3 (STP)/h.

The maximum attainable hydrogen yield is influenced by a number of factors, the most important are:

1. Composition of the crude gas.

For the economical operation of a plant the hydrogen content of the crude gas should be at least 50 %. The maximum hydrogen content is 95 – 99 %. Depending on the total amount of gas, TSA
plants may be more suitable for such high-purity raw gases.
The concentration and nature of impurities with respect to their ad- and desorption behavior determines basically the size of the adsorber which in turn results in higher or lower loss of hydrogen.
2. Pressure ratio feed gas: purge gas.
This influences the size of the adsorber and the extent of the effort necessary for regeneration. At high pressure ratios the yields can be increased by using several pressure equalization steps
(plants with more than 4 adsorbers).
3. Purity requirements and nature of impurities.
The hydrogen purity in the PSA plants is usually >99.9 %. Higher purities with respect to those compounds which are poorly adsorbed, such as the inert gases nitrogen, argon, etc., are attainable at
the expense of the yield. In spite of this, purities of 99.9999 %, i.e., hydrogen with <1 ppm impurities, can be realized on an industrial scale.
4. Temperature of the feed gas.
High temperatures (>40 °C) result in a lower capacity of the plant, or in a poorer yield. Lower temperatures are therefore favored. Care must be taken, however, to avoid condensation of certain
impurities in the regeneration step. For example, liquid or freezing water can damage the adsorbent.

Important developments in the design of large PSA plants were first possible after the introduction of microprocessor control systems. The main problems facing the control system are caused by the
speed and the precision required by the short-time cycles. For this reason, modern plants are controlled by microprocessors using a 1-s sequence achieving “real time” control. Beyond this, it is now
possible, not only to control the sequence and valves, but also to have a real process control system. This allows a high degree of flexibility for the changing of various process parameters and the
indication of disturbances. In particular, the automatic initiation of disturbance strategies in the case of process, pressure and valve failures, e.g., shut-down of an upset adsorber without interrupting
hydrogen production, is possible. Very high requirements are placed on the vessels and valves with up to 50 000 cycles (i.e., valve closures) per year. The operational practice has shown, however, that
these plants are very reliable, so that they have become standard.

In the past years purification of hydrogen by means of PSA has increased more than any other gas purification process. Figure 75 gives an impression of the breakthrough of this new technology. As can
be seen from the number of plants corresponding to the plant capacity, not only the number of plants per year put on stream has increased, but also the amount of hydrogen purified per plant. Plants
> 100 000 m3 (STP)/h inlet gas are not uncommon nowadays.

Figure 75. Worldwide capacity of H2 gas purification plants with pressure-swing adsorption technology (put into operation from 1966 to 1984)

Applications [224].
Gas Purification in Steam-Reforming Plants (see Section Catalytic Reforming of Hydrocarbons). Other gas purification processes have been more or less completely replaced in this area [228]. For the
production of hydrogen and carbon monoxide by steam reforming, PSA in combination with a low-temperature plant can also be economically used. The amount of gas handled by the plants built ranges
from 500 – 87 000 m3 (STP)/h of product gas. These plants use the tail gas produced to fire the reformer. Specially designed control loops and buffer systems lead to a very homogeneous tail-gas flow
which is discharged with only small fluctuation in composition and flow, so that a reliable operation of the burners is possible.

Refinery Gases (see Section Refinery Processes). Because of the increasing demand for hydrogen for hydrogenation in refineries the PSA process is used there for hydrogen recovery and production.
Refinery gases have a hydrogen content of 60 – 80 % the remainder being higher boiling hydrocarbons (C6 – C10). These are adsorbed well, but are difficult to desorb which can cause damage to the
adsorbent. By choice of suitable adsorbents these difficulties can be avoided. Plants with feed gas capacities up to 100 000 m3 (STP)/h have already been built. Figure 76 shows the hydrogen purification
section in a large refinery in Kuwait.

Figure 76. Production of hydrogen from refinery off-gas. Plant with 12 adsorbers and 2 buffer drums (in the center) for flow and composition equalization

Plant capacity: 100 000 m3 (STP)/h refinery gas; H2 content: 80 %; product purity: >99 %; recovery rate: 89.5 %. Erected by German Linde in Kuwait in 1986

page 74 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Coke Oven Gas (see Section Coke Oven Gas ). Coke oven gas purification places the highest demands on the adsorption technology because of the inlet gas composition. Special preadsorbers have
been used to remove impurities such as tar, naphthalene and higher sulfur compounds, which have a poor desorption capacity.

Hydrogen-Rich Gases from Various Synthesis Processes. Apart from numerous small plants which operate for example with ammonia or methanol plant purge gas, the coal gasification at SASOL is to be
mentioned. 160 000 m3 (STP)/h of off-gas from the coal gasification, respectively the Fischer – Tropsch synthesis is purified in five PSA plants.

5.3. Catalytic Gas Purification

Four distinct cases for the use of catalytic processes for gas purification can be distinguished:

1. A component of the crude gas is to be catalytically converted to hydrogen, i.e., a secondary hydrogen production process.
2. Impurities in the crude hydrogen are to be catalytically converted into a form which allows them to be more easily removed from the gas by other purification processes (condensation, adsorption).
3. Gas components, which even in traces could be detrimental to the use of the hydrogen, are converted into harmless components with only minor changes in the gas purity.
4. Adsorption processes, in which certain impurities are chemically bound to the adsorbent so strongly that a regeneration during normal operation does not take place.

5.3.1. Carbon Monoxide Removal

Hydrogen from synthesis gas may contain up to 50 % carbon monoxide (partial oxidation). Carbon monoxide can be recovered for CO production by a methane scrubber (see Section Low-Temperature
Absorption Processes) and used for synthesis, or it can be reacted with water to produce hydrogen and carbon dioxide (water – gas reaction, shift conversion):

The reaction is exothermal. Because of the equilibrium conditions the carbon monoxide conversion is incomplete in adiabatic reactors. For increased carbon monoxide conversion a greater amount of
water (cH2O/cCO > 3) is necessary. At high carbon monoxide concentration a series of adiabatic reactors with intercooling is used. Isothermal reactors have also been suggested [229]. Depending on the
type of catalyst used, three types of conversion may be identified.

Conversion of Sulfur-Containing Gas. Crude gas from partial oxidation or coal gasification, which contains at least 50 ppm hydrogen sulfide, can be converted on metals of group 6, 8, 9, or 10 of the
periodic table (exceptions: Cr and Fe) or on mixtures of these metals. The carrier is aluminum oxide, sometimes alkaline compounds are present as a promoter. The metals are totally or partially sulfided,
the sulfided form being the most active form. The usual catalysts contain cobalt, molybdenum, and nickel as active components. They can be used at 10 MPa and 250 – 550 °C.

An industrial process with these catalysts was described in Section Gasification of Coal and Hydrocarbons. A detailed description of the kinetics of crude-gas conversion is found in [230], patents in [231].

High-Temperature Shift (HT Conversion). For essentially sulfur-free gases, e.g., the product gas of the steam reformer or the secondary reformer, the well-known carrier-free conversion catalysts on
iron oxide – chromium oxide basis can be used, e.g., 74 % Fe2O3, 10 % Cr2O3, 0.2 % MgO (rest: volatile components). Sulfur up to several hundred ppm can be tolerated although the activity then drops;
catalyst poisons are phosphorus, silicon and unsaturated hydrocarbons in the presence of NOx.

The shift reaction runs almost to equilibrium at space velocities between 3000 and 10 000 m3 (STP) h–1 m–3 and pressures up to 7 MPa. The reaction temperatures are in the range 300 – 530 °C [232].

In steam reforming plants usually only one HT conversion stage is installed. To decrease the carbon monoxide concentration further, low-temperature conversion catalysts are used. If the purified
synthesis gas in coal gasification plants is to be converted, a series of adiabatic reactors can be used. The last stage of this cascade usually consists of a low-temperature conversion.

Low-Temperature (LT) Conversion. The catalysts for LT conversion are supplied in their oxidized form, the major component of the unreduced catalyst is CuO, usually in a mixture with ZnO. Other
components may be Cr2O3 and Fe2O3 (stabilizers) and Al2O3 (carrier) [230], [233]. The temperature range used is 190 – 260 °C; the lower temperature is limited by the dew point of the gas, the upper
temperature because of the sensitivity of the copper catalyst to thermal deactivation.

These catalysts are extreme sensitive to sulfur, contents of 0.1 ppm leading to their deactivation, so that a suitably large reactor volume must be used to ensure that any slow deactivation of the catalyst
will not cause any premature shut-down of the unit. Very often guard beds are placed before the actual catalyst [234]. For more details on the use of HT and LT shift conversion in steam reforming plants
see Section Catalytic Reforming of Hydrocarbons and Gas Production.

page 75 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The thermal stability of copper catalysts can be raised considerably if they are used in a silica-supported form. These supported catalysts are prepared by precipitation from a homogeneous solution and
are found to be very active for the water – gas shift reaction [235]. The catalyst can be reduced at a temperature of 450 °C and can be operated between 275 and 400 °C with initial carbon monoxide
concentrations up to 10 %. A Fe2O3 – Cr2O3 catalyst has been prepared in a similar way and also shows improved activity at lower temperature, at the same time the resistance to traces of sulfur and
chlorine compounds [236] still being preserved.

Because of considerations of equilibrium, reaction temperature and efficiency of the catalyst, the shift reaction can remove carbon monoxide only down to a remaining concentration of ca. 0.2 – 0.5 vol %
in the dry gas. For refinery gases, however, purities <0.1 %, for hydrogenation of fats and oils <10 ppm (mL/m3) and for ammonia synthesis <1 ppm (mL/m3) are required (see Chap. Conventional Uses).
For these and other purposes a further purification stage is necessary. In many cases this is the methanation step.

5.3.2. Methanation
Here, methanation is described as a process for the removal of traces of carbon monoxide or carbon dioxide. For the conversion of large carbon monoxide concentrations to yield substitute natural gas
see Gas Production – Methanation of Rich Gas.

Especially for ammonia plants the carbon monoxide concentration in the gas must be decreased (see also Fig. 77). This is done by using the reaction

The methanation converts carbon monoxide, which is a poison for the ammonia synthesis catalyst, to the inert gas methane. However, losses in hydrogen content (3 mol H2 per mole of converted carbon
monoxide) must be accepted. The design of methanation reactions is described [237].

Figure 77. Removal of CO and CO2 from ammonia synthesis gas (3H2 + N2) by selective oxidation (SELECTOXO), CO2 scrubbing, and methanation

a) Selective oxidation; b) CO2 scrubber; c) Methanation; d) Regeneration of the scrubbing agent; e) Gas cooling; f) Gas heating

Under typical methanation conditions (high hydrogen partial pressure, low methane and water content) the equilibrium is shifted far to the methane side, so the resulting carbon monoxide concentrations
are usually below the detection limit. Nickel oxide, sometimes together with chromium oxide, is used as a catalyst. However, it must be reduced before the start-up. Aluminum oxide or diatomaceous earth
is used as carrier. The metallic content of the catalyst is very high (30 – 60 %) compared to the nickel catalyst used for steam reforming.

The following undesired side reactions may occur:

5.3.3. Selective Oxidation

The disadvantage of hydrogen losses and increased inert content in the aforementioned methanation process can be avoided, if carbon monoxide is oxidized to carbon dioxide, because the latter can be
readily removed from hydrogen. However, as a side reaction oxidation of hydrogen, which is present in much greater concentration, may occur:

In spite of this, selective oxidation of carbon monoxide can be achieved on a special noble metal catalyst [238-240]. The incorporation of this process stage into the synthesis gas production for ammonia
is shown in Figure 77.

The selective oxidation is placed behind the last shift reactor after removal of excess water by condensation. Air, which is taken from the compressed air for the secondary reformer, is used as oxidizing
agent. Oxidation takes place at 50 – 70 °C in an adiabatic bed reactor. The temperature increase resulting from the heat of reaction amounts to 10 K per 0.1 mol % carbon monoxide. The carbon
monoxide content can be reduced to ca. 100 ppm (mL/m3) by this method. If the gas is to be used for ammonia synthesis, an additional methanation stage must still be incorporated for final clean-up.

5.3.4. Catalytic Removal of Sulfur Compounds

page 76 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Sulfur compounds are catalyst poisons for hydrogen production and the various syntheses using hydrogen. Desulfurization is achieved in two stages: (1) hydrogenation and (2) hydrogen sulfide removal.
Depending on the feed gas, these stages may be carried out separately or simultaneously. In the first step, organic sulfur compounds and carbonyl sulfide are hydrogenated to hydrogen sulfide

a cobalt – molybdenum contact is normally used as hydrogenation catalyst. If the gas contains a great amount of carbon monoxide, e.g., coke oven gases, the cobalt – molybdenum catalyst must be
sulfided in order to avoid methanation of carbon monoxide (see Section Methanation).

Hydrogenation of sulfur compounds is favored by high hydrogen partial pressure and temperatures around 350 °C. After the hydrogenation of the sulfur compounds, the hydrogen sulfide formed can be
removed by chemical or physical scrubbing (see Section Scrubbing Processes), if its concentration is > 1 %

For lower concentrations, hydrogen sulfide is absorbed on zinc oxide.

Zinc oxide is available in the form of spheres, tablets or extrusions. The average sulfur pick-up until hydrogen sulfide breaks through the bed is ca. 25 % of the stoichiometric amount. The reaction
temperatures are 100 – 350 °C (max. 400 °C). Organic sulfur compounds and carbonyl sulfide are partially cracked and absorbed in the zinc oxide bed, so that for natural gases with a low sulfur content
the hydrogenation may not be necessary. The final purity after a zinc oxide bed is <1 ppm (mL/m3) sulfur. The zinc oxide cannot be regenerated in situ but must be replaced.

Other materials used for desulfurization are, e.g., iron hydroxide (a byproduct of aluminum production), iron ore, and other purification agents containing iron oxide or iron hydroxide activated with alkaline
compounds [241]. These absorption agents are mainly used for purification of coke oven gas before it is used as town gas. The achievable purity is <2 mg sulfur/m3 (STP) (ca. 10 ppm [mL/m3] S as H2S).
Iron- or copper-activated carbon can also be used. The regeneration of these agents with superheated steam is, however, only partially possible.

Chemisorption:

Regeneration:

5.3.5. Removal of Oxygen

Hydrogen from coke oven gas and from an electrolysis (contrary to the hydrogen produced from other reducing processes) very often contains oxygen in concentrations up to 1 %. This oxygen is removed
in deoxo reactors.

Small levels of platinum and palladium on typical carriers such as -Al2O3, active charcoal, or aluminosilicates provide highly active catalysts for the conversion (e.g., 0.3 % palladium on alumina). The
catalyst is active even at ambient temperature, although usually higher temperatures are used.

Because of the strongly exothermic reaction a heat exchanger is installed after the deoxo reactor. The reaction product water is removed by adsorption. An example for the use of this process with
prepurified coke oven gas is given in Figure 16.

Another method for oxygen removal is the use of catalytically active non-noble metals, usually copper or nickel, on carriers, e.g., silicon dioxide. These contacts operate at higher temperatures (300 °C)
and are less sensitive to impurities such as hydrogen sulfide than are the noble metal catalysts. Both catalytic processes for oxygen removal are very selective, yielding purities of 1 ppm (mL/m3) oxygen.

5.3.6. Removal of Nitrogen Oxides

In gases from coke ovens and coal gasification plants, in particular from entrained-bed gasifiers, traces of nitrogen oxides are present. These must be removed before gas separation in a cryogenic
process is possible. Principally, nitrogen oxides can be removed from hydrogen streams by either oxidation or reduction. In the reduction process the nitrogen oxides are hydrogenated on sulfided cobalt –
molybdenum catalysts, and water and nitrogen are produced. The reaction can be accelerated if small amounts of ammonia are present in the gas. Water and ammonia are removed in downstream
scrubbers. The nitrogen produced can usually be neglected for the hydrogen purity.

Oxidative removal of NO takes place with ozone or, particularly selective, with ClO2. The NO2 produced is removed by use of scrubbers with alkaline-reducing agents (reduction to nitrogen). Final purities
of ca. 0.01 ppm (mL/m3) NO can be attained.

page 77 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
5.4. Scrubbing Processes
Scrubbing processes are used to purify large quantities of hydrogen. The compounds to be removed are absorbed in the scrubbing liquid. Since this can be done by utilizing either their different solubility
in the scrubbing agent or by reaction with the scrubbing agent, the scrubbing processes are subdivided into physical and chemical processes.

5.4.1. Principles
For a detailed description of the physical and chemical principles of gas scrubbing see Absorption and [242-244].

For selection, design and economics of a physical absorption process the most important factor is the solubility coefficient of a component given as (m3 [STP] bar–1 t–1) or ′ (kmol bar–1 t–1) [200]. The
capacity of an absorbent can be represented by the solubility curve as a function of pressure and temperature. These are shown in Figure 78 for carbon dioxide.

Figure 78. Absorption equilibria of carbon dioxide in various solvents [244]

a) Methanol, –15 °C; b) MDEA (methyl diethanolamine), 75 °C, 3.5 molar; c) MEA (monoethanolamine) 75 °C, 3 molar; d) Hot potassium carbonate, 110 °C; e) TEA (triethanolamine), 75 °C, 2.5 molar; f) NMP (N-
methylpyrrolidone), 20 °C; g) Propylenecarbonate, 25 °C; h) Selexol (polyethyleneglycol dimethyl ethers)

For physical absorption the capacity at first follows Henry's law (i.e., a linear relationship, curves a, f, g, h), while for chemical absorbents typically a curve, which approaches a saturation value (curves b,
c, d, e) is found. Apart from solely physical or solely chemical scrubbing, combined process can be used. Here the scrubbing stages may be separate and operated in series (e.g., crude and fine
scrubbing) or a mixture of a physical and a chemical scrubbing agent may be used.

Chemical scrubbing is only suitable for acid gas components such as CO2, H2S, COS, and HCN. Physical scrubbing, on the other hand, can also be used for other components such as benzene, and
higher hydrocarbons.

The raw gases for hydrogen production, depending on their source, contain several impurities in various concentrations; apart from that, they are at various pressures. For the choice of the optimum
scrubbing process, the following factors are important:

1. Type, amount and partial pressure of the component to be removed;

2. Type, amount and total pressure of the gas to be purified;
3. Required final purity, utilization or subsequent process steps for the product gas and the impurities removed;
4. Consumption figures

Physical scrubbing is preferred for large gas quantities at high pressure with high concentrations of impurities to be removed; chemical scrubbing is particularly suitable when the absolute amount of the
component to be removed is small. It can be used both at atmospheric and at higher pressure. With chemical scrubbing it must also be certain that there is no component in the raw gas, that can react
irreversibly with the scrubbing agent. Carbonyl sulfide, for example reacts with MEA (monoethanolamine) to form stable salts. Here the absorbent cannot be regenerated by heat alone, and this leads to a
loss of scrubbing agent.

Figure 79 shows how the choice of a suitable scrubbing process can be made for certain limiting parameters (e.g., partial pressure of the substance to be removed in the inlet and the exit gas). An
overview of the processes used at present to remove acidic components from gases (synthesis and natural gas), is given in [246].

Figure 79. Examples for the selection of a suitable scrubbing process for CO2 or CO2 – H2S removal with respect to partial pressures in feedgas and product gas, [245]

A) Selection of solvent for CO2-removal, no H2S; a) Physical solvent plus amine; b) Physical solvent, physical solvent plus amine or activated hot potassium carbonate; c) Physical solvents, d) Physical solvents or activated
hot potassium carbonate; e) Activated hot potassium carbonate or concentrated amine; f) Activated hot potassium carbonate or amine; g) Amine

B) Selection of solvent for simultaneous removal of H2S and CO2; a) High loading amines (DEA), Selexol, Rectisol; b) Physical solvents; c) Physical solvents or diglycolamine; d) Activated hot potassium carbonate, Sulfinol,
amines; e) Amines, Sulfinol

The most important processes used for purification in hydrogen production are:

1. Carbon dioxide removal from converted steam reformer gases for the production of hydrogen or ammonia synthesis gas. Here the hot potash scrubbers and modifications thereof, often combined
with an amine wash, are used (see Absorption). In recent times activated MDEA (methyl diethanolamine) wash processes and certain mixed amine processes have been found to have some

page 78 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
economic advantages, however, there still is a tendency to use physical scrubbing processes (Selexol, Sepasolv, etc.) for crude gas purification.
2. For the simultaneous and separate removal of carbon dioxide and hydrogen sulfide from converted or unconverted raw gas from partial oxidation of heavy oil or coal, the Rectisol process has been
successful and is used almost exclusively for coal gasification plants.

5.4.2. Physical Scrubbing (Rectisol Process)

The Rectisol process ( Gas Production) (jointly developed by Linde and Lurgi) is a physical scrubbing process with methanol as the preferred scrubbing agent. Newer variations use other organic
solvents or solvent mixtures (e.g., Rectisol II, [248]). Since the absorption capability of organic solvents increases at lower temperatures and since, in addition, losses of scrubbing agent are smaller,
Rectisol processes are usually operated at 200 – 260 K. The solubility of various gases in methanol is given in [200] and [248].

Depending on the process chosen, carbon dioxide and hydrogen sulfide can be recovered together or separately. Usually the simpler plants have a common scrubbing stage. Proper configuration of the
main scrubbing and the purification section allows a very high attainable purity (20 – 50 ppm [mL/m3] CO2). For special purposes (e.g., methanol synthesis gas), a certain amount of CO2 can be left in the
gas and the fine scrubbing can be used only for the sulfur containing compounds. The disadvantage of the simple process is that a tail gas is obtained which, apart from carbon dioxide, also contains a
certain amount of sulfur compounds and other traces (HCN, CO). Such a gas must, therefore, be treated by special processes.

For the production of a sulfur-free carbon dioxide fraction and a hydrogen sulfide fraction suitable for the Claus process, selective Rectisol scrubbing processes are used ( Gas Production). Figure 80
shows a simplified process scheme with process description. Rectisol is very flexible with respect to the attainable purity of the product gas, the pressure range and any possible process variations. In
addition, there are no corrosion problems. For ammonia synthesis gas the Rectisol process can be combined advantageously with a low-temperature liquid nitrogen wash (see Section Low-Temperature
Processes). Other physical gas scrubbing processes are summarized in Table 32.

Table 32. Physical gas-scrubbing processes

Name Solvent Tempera- Licensor

ture, °C

Fluor- propylene carbonate ambient Fluor

Solvent
Purisol N-methyl pyrrolidone ambient Lurgi
Selexol polyethyleneglycol
dimethylether 0 – 40 Norton
Sepasolv polyethyleneglycol
diisopropylether 0 – 40 BASF

Figure 80. Selective Rectisol scrubbing process for production of a sulfur-free carbon dioxide fraction and a hydrogen sulfide fraction suitable as Claus plant feed [249]

5.4.3. Chemical Scrubbing Processes

An example of chemical scrubbing in hydrogen production is the combination Benfield – DEA scrubber (so-called High-Pure Benfield). For the SASOL II and III plants carbon dioxide is removed from the
recycle gas of the Fischer –Tropsch synthesis, which in total amounts to 7×105 m3 (STP). After the ensuing low-temperature separation, a gas with 80 % hydrogen is obtained which is fed either to the
Synthol synthesis or to a PSA plant. A process scheme is shown in Figure 81.

Figure 81. Flowsheet of a combined Benfield – DEA scrubber (High Pure Benfield) for CO2 removal from Fischer – Tropsch synthesis recycle gas (SASOL II and III) [250]

a) Absorption column; b) Regeneration column; c) Steam generator; d) Air cooler; e) Water cooler; f) DEA scrubber; g) Heat exchanger; h) Pump

page 79 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

High-Pure Benfield Process. Both scrubbers contain an absorption and a regeneration tower, the latter operating at almost atmospheric pressure. The temperature ranges between 70 and 108 °C in (a)
and between 50 and 60 °C in (f). The absorbents are regenerated by indirect heating with steam. The final cooling of the regenerated absorbents takes place in air coolers, the cooling of the purified gas in
water coolers. The hot potash scrubber (first stage) lowers the carbon dioxide level below 0.8 vol %, the final DEA scrubber below 80 ppm (mL/m3). The relatively small carbon dioxide amount from the
DEA regeneration column (b2) is fed to the hot potash regeneration column (b1) and there, together with the main amount of carbon dioxide, vented into the atmosphere or recovered and used for other
purposes.

Table 33 lists some other chemical scrubbers.

Table 33. Industrially important chemical scrubbing processes

Classifica- Scrubbing Additives Trade name,

tion agent licensor

Amine MEA
washes MEA not Aminguard
specified (UCC)
DEA
HDEA Linde
MDEA
DGA Fluor
DIPA Adip
activated Shell
MDEA, various BASF
t-amines various UCARSOL
(UCC)
Physical – sulfolane, DIPA Sulfinol
chemical (Shell)

washes methanol, amine Amisol

(Lurgi)
Hot potash K2CO3– H2O vanadium, Benfield

washes DEA
amineborate Catacarb
borax Lurgi
arsenate Giammarco
Caustic NaOH – H2O only used for fuel purifica-
scrubber tion, no regeneration

page 80 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Oxidative scrubbing processes are sometimes used to remove hydrogen sulfide, e.g., for the purification of coke oven gas or other gases from hydrogen purification plants. Hydrogen sulfide is absorbed
and oxidized in the scrubbing agent to form elemental sulfur ( Gas Production). Such processes are:

Stretford process (British Gas Corp.)

Locat process (ARI)
Sulfolin process (Linde)

An overview of the literature on hydrogen sulfide removal from coke oven gas, low-Btu gas, rest gases and natural gas is given in [251], on newer oxidative scrubbing processes in [252]. The scrubbing
processes described, however, are unable to remove the so-called “inert components” (noble gases, nitrogen, carbon monoxide, methane). In general, further purification steps on the basis of adsorption
and catalysis (see Section Catalytic Gas Purification) as well as low-temperature processes including cryogenic scrubbing with liquid methane (see Section Low-Temperature Processes) are necessary.

5.5. Permeation Processes

For purification of hydrogen and for separation of hydrogen from gas mixtures processes based on the principle of diffusion and permeation through membranes are also suitable. The physical
fundamentals of diffusion processes are given in [253] together with a detailed description of diffusion separation methods in [254].

The driving force for gas transport through membranes is the partial pressure of the gases. The permeating flux Qi (per unit area of surface and time) is given by Equation (57):

(57)

where Pi, the permeation coefficient, is the product of solubility and diffusion coefficients of a gas i, pi2 and pi1 are partial pressures on the high- and the low-pressure sides of the membrane and d is the
membrane thickness.

To separate various components it is essential to have different permeation coefficients. The separation capability of the membrane is defined by the selectivity ; this is the ratio of the permeation
coefficients of the two gases i and j

The fact that the permeation coefficient of hydrogen is much higher than that of other gases is utilized to separate hydrogen by diffusion through metal or polymer membranes.

5.5.1. Metal Membrane Techniques

Hydrogen diffuses well and with a high specific permeation rate through membranes consisting of palladium, palladium – silver or palladium – rare earths. Silver – palladium alloys with 23 wt % silver are
most suitable. The diffusion rate of hydrogen is highest at a composition close to the maximum solubility of silver in palladium (i.e., ca. 23 wt %); furthermore, material distortion by – phase transition
in the palladium-hydrogen system is substantially decreased. Optimum operating conditions, i.e., maximum hydrogen throughput and high durability are found at 350 °C and 2 MPa hydrogen pressure.
Unsaturated hydrocarbons, halogens and sulfur compounds can lead to rupture of the membrane. The hydrogen suffers a great pressure drop after permeation through the membrane, however, it is
particularly pure (impurities and moisture <0.5 ppm [mL/m3]). Palladium membrane diffusion is discussed in [255].

Portable small-sized hydrogen generators or containerized medium-sized units (the latter are constructed from modular units) can be used for the final purification of hydrogen or for separation from raw
gas mixtures, e.g., from methanol or ammonia reforming (see Section Other Chemical Processes). Main parts of a module are the diffusion elements which consist of a number of silver – palladium tubes
contained within a cylindrical cell.

Figure 82 shows the cell performance data of commercial pure-hydrogen plants on the basis of diffusion. The raw hydrogen must be fed to the unit at 2.1 MPa, the operating temperature is 300 °C and the
average specific energy consumption amounts to ca. 100 W per cubic meter of hydrogen [256]. The process is considered suitable for small- to medium-scale plants. Metal membrane processes have not
been adopted for industrial-scale hydrogen purification.

Figure 82. Operating data of hydrogen diffusion cells with Pd-Ag membranes

a) Yields (%, related to input hydrogen amount); b) Hydrogen content in off-gas (vol %); c) Cell output) (% of nominal capacity)

5.5.2. Polymer Membrane Processes

page 81 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
New developments in the area of polymer membrane materials also allow their utilization in gas separation technology. The main criteria for suitable separation membranes are high permeation rates
together with high selectivity, pressure-resistant membrane structures and chemical as well as thermal stability. State-of-the-art membranes have asymmetric composite layer structures on a woven
material support or a composite hollow-fiber membrane structure. The membrane separation stage is built as module; for gas separation module constructions with high packing densities are used, namely
spiral-wound modules (> 900 m2/m3) and hollowfiber modules (9000 m2/m3). An overview is given in [257][258][259]. From the process point of view various module arrangements such as cascades or
membrane columns [260] may be used.

Asymmetric composite membranes based on cellulose acetate or polysulfones have a particularly high specific permeability for hydrogen, as do aromatic polyimide membranes; they are used for the
industrial scale separation of hydrogen.

The product of Permea, a Monsanto company, is leading in this field. The Perma Prism process [261] uses asymmetrical hollow-fiber membranes of polysulfone with an active siloxane layer (300 –
500 µm outer diameter, 100 – 250 µm inner diameter, effective siloxane separating layer 0.1 µm). Several 10 000 of hollow fibers are connected together in a module, as shown in Figure 83. The fiber
bundle is sealed at one end (top) and enclosed in an epoxide mantle at the other end. The modules are installed vertically and have a diameter of 100 or 200 mm and a maximum length of 3 or 6 m. The
feed gas, which should be freed of disturbing components (NH3, H2S, CH3OH) by a suitable scrubbing system, is fed into the module near the base. The permeate is removed from the base of the module
while the remaining gas is removed from the top. The fibers can be operated with a pressure of up to 15 MPa, whereby the pressure difference may amount to 7 MPa. The operating temperature can be
up to 100 °C.

Figure 83. The Prism separator module of Permea

a) Feed stream of mixed gases; b) Permeate gas outlet; c) ASME-coded carbon steel shell; d) Hollow fiber membranes; e) Fiber bundle plug; f) Nonpermeate gas outlet

The stability of the membranes towards impurities is given in % of the corresponding saturation value at 20 °C:

Hydrogen sulfide up to 10 % of the saturation value

Ammonia up to 2 % of the saturation value
Water up to 100 % of the saturation value
Hydrocarbons up to 80 % of the saturation value
(C2 – C6)

The degree of purity and the hydrogen recovery rate are determined by the partial pressure difference between the inside and the outside of the hollow fiber, the remaining gas components (i.e., their
permeation coefficient) and their partial pressure.

Figure 84 A shows the rate of hydrogen recovery and the obtainable purity as a function of various hydrogen contents in the raw gas. Plots for two feeds (70 % and 40 % H2) in Figure 84 B show that the
ratio of total pressures dramatically affects permeate purity, whereas the effect of further increase in membrane selectivity leads to a plateau above a selectivity factor of 40.

Figure 84. A) Obtainable hydrogen purity vs. hydrogen recovery rate

a) Partial pressure difference 3.52 MPa, H2 in feed 63 vol %; b) Partial pressure difference 3.52 MPa, H2 in feed 86 vol %B) Permeate purity with two feeds at different pressure ratios [257] ------- 40 % H2, 60 % C1; ——
70 % H2, 30 % C1

p1 = permeate pressure, p2 = feed gas pressure

Hollow-fiber separators were adopted quickly for commercial applications, more than 160 membrane separation systems have been installed for hydrogen recovery since 1979. Preferred applications are
hydrogen recovery from purge gases of ammonia or methanol synthesis loops or from other hydrogen containing gases in the refinery or gas industries. Table 34 gives a survey. Figure 85 shows the

page 82 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
process scheme of a plant for hydrogen recovery from the purge gas of a 545 t/d ammonia plant. 3000 m3 (STP)/h gas is led through two rows of modules (containing 8 and 4 modules, respectively) at
various pressure levels, 90 % of the hydrogen thereby being recovered. This system is especially useful, because a part of the hydrogen can be led back to the synthesis stage at 7 MPa. In a 1000 t/d
ammonia plant the recovered hydrogen leads to an additional ammonia production of 40 t/d (ca. 5 %) [262]. However, other valuable components of the gas, such as argon, cannot be recovered in a pure
form.

Table 34. Hydrogen containing gases membrane modules can handle

Source of feedgas Components Pressure, Product Pretreatment, remarks

MPa composi-
% H2 others tion, % H2

Ammonia synthesis purge gas 60 N2, CH4, Ar, NH3 13.0 H2, 85 water wash, NH3– conc.

<200 ppm (mL/m3)

Methanol synthesis purge gas 58
Catalytic reformer off-gas
CO, CO2, N2, CH4, 5.0 H2, CO2, 80 water wash, CH3OH conc.
CH3OH, H2O <100 ppm (mL/m3)
70 – 80 CH4, alkanes, 3.0 – 5.0 H2, 95 oil separation by refrigera-
aromatics tion or oil scrubber,
C2– C6 <80 % satur.

aromatics <50 ppm (mL/m3)

Catalytic cracker purge gas 10 – 30 N2, O2, CH4, CO, 0.7 – 1.4 H2, 60 – 80 compression, water wash
CO2, H2S, alkenes,
higher and lower
olefins, aromatics
Hydrocracker purge gas 75 CH4, H2S, H2O, 7.0 – 11.0 H2, 95 aromatics <50 ppm (mL/m3)
NH4, alkenes,
aromatics
Hydrotreater purge gas 60 CH4, H2O, alkenes, 35 – 50 H2, 97 feed <30 mol % H2, uneco-
aromatics nomic, aromatics <50 ppm
(mL/m3)
Toluene hydrodealkylation CH4, H2O, alkenes, 30 H2 oil separation by refrigera-
aromatics tion or scrubber, aromatics
<50 ppm, (mL/m3)
Steam – methane reformer gas CH4, CO, CO2, H2O 30 – 80 H2, H2/CO2
Oxo-synthesis gas 70 CO variable H2/CO 1 : 1 H2/CO adjustment

Figure 85. Flowsheet of a hydrogen recovery unit with Prism separator modules in an ammonia plant

a) Water scrubber; b) First bank Prism separators; c) Second bank Prism separators; d) Syngas compressor first stage (2.4 MPa suction pressure); e) Syngas compressor second stage (6.8 MPa suction pressure); f) H2
recycle to second stage suction syngas compressor; g) Fuel gas to NOx abatement or primary reformer; h) H2 recycle to first stage suction syngas compressor

page 83 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
[Top of Page]

6. Liquefaction of Hydrogen
The boiling point of hydrogen is 20.39 K at 101.3 kPa (see Table 2) for n-H2 and 20.26 K for p-H2. For more details on the physical properties of n-H2 and p-H2 see Chapter Properties.

Like other gases, hydrogen is liquefied because storage and transportation is easier in the condensed state. Further applications of liquid hydrogen are rocket propulsion and scientific experiments, e.g.,
bubble chambers which are used to produce visible trails of various particles involved in reactions of nuclei and elementary particles. A wide application for liquid hydrogen is predicted in the future
hydrogen energy concept (see Chap. Hydrogen Energy).

6.1. Principles of Hydrogen Liquefaction

For the liquefication of gases [263] energy must be removed until the condensation point is reached. This is achieved either by cooling down to low temperatures with external means, such as liquid
nitrogen or the magneto-caloric effect, or internally by irreversible or isentropic decompression.

The decompression of hydrogen in a throttle valve can only lead to a temperature decrease in a certain range of the Joule – Thomson curve (see Fig. 5). The range which is used technically is from 77 K,
the boiling point of liquid nitrogen at 0.1 MPa, down to 25 K. In this range (see Fig. 86) the isenthalpic lines (H = constant) in the T – S diagram are particularly steep. The decompression in the throttle
leads to a decrease in temperature or, at the boiling point, to a high degree of liquefaction at a constant temperature. The whole range of the isenthalpic lines, isobars and isochores in the T – S diagram,
especially in the two-phase areas and around the triple point is given in [264].

Figure 86. Temperature entropy diagram of n-hydrogen. Illustrated are the liquefaction process and the cooling cycle for a modern liquefaction process, described in Figure 88

Isentropic decompression in turbines or other machines is, generally speaking, possible in all temperature ranges. However, control of the two-phase streams in the neighborhood of the condensation
point leads to technical problems. Apart from this, the efficiency of a decompression machine decreases for technical reasons at low temperatures. Therefore, this process is only used for hydrogen above
25 K.

Applying a magnetic field to a magnetic substance causes an alteration in the inner structure of the material which in turn leads to a temperature change. In particular, demagnetization of certain metals in
the neighborhood of the Curie temperature leads to a maximum temperature decrease. (The Curie temperature is the temperature above which ferro-magnetic material reverts to the paramagnetic state.)
Under favorable conditions an isentropic temperature decrease of 1 K/T flux density may be attained.

To obtain greater temperature differences respectively entropy changes, metals or alloys with different Curie temperatures of the components are necessary. For the liquefaction of hydrogen the rare
earths (in particular gadolinium), their compounds and alloys are suitable. For a detailed description of the magneto-caloric effect in cryogenic refrigeration processes see [265] and [266].

Hydrogen liquefaction is technically achieved by a series of cyclic processes. The theoretical lowest energy consumption (minimum work) in a reversible Carnot process for the production of liquid p-H2
(0.1 MPa) from gaseous n – H2 (25 °C, 0.1 MPa) amounts to 14.23 kJ/g = 3.95 kWh/kg.

In natural hydrogen (n-H2) at 25 °C the ratio of the nuclear spin isomers o-H2 and p-H2 is ca. 0.75 : 0.25. At the boiling point of hydrogen the equilibrium lies almost completely on the side of p-H2 (see
Section Physical Properties). The reaction enthalpy H of the exothermal o – p conversion is in the same range as the heat of evaporation

H o – p (25 K) = 1.448 kJ/mol

H n – p (25 K) = 1.086 kJ/mol
Hevap for p-H2 (25 K) = 0.811 kJ/mol
Hevap for n-H2 (25 K) = 0.828 kJ/mol

Thus, if the reaction rate of the o – p-conversion is large in comparison to the retention time in storage and transport vessels, a major part of stored liquid n-hydrogen is evaporated even after a few days

page 84 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
[267]. In addition, heat losses through the insulation occur. For this reason, technical hydrogen liquefaction units have o – p conversion reactors incorporated into the process. The o – p conversion takes
place on iron oxide gel, iron hydroxide and chromium oxide on alumina catalysts.

The heat of reaction of the o – p conversion must be compensated for by additional refrigeration capacity of the liquefaction unit. In the ideal case the reaction should take place along the equilibrium line
(see Fig. 1), because then the lowest reaction enthalpy is involved and the necessary refrigeration is thermodynamically favorable, because it can be delivered at the highest temperature level. In technical
plants this ideal case can be simulated by using a cascade of adiabatic or isothermal reactors.

Particular requirements are placed on the purity of the hydrogen to be liquefied, because at liquid hydrogen temperatures all impurities except helium become solid and this would lead to a blockage of the
refrigeration unit. Before liquefaction hydrogen of > 99.5 % purity from electrolysis or of 99.99 % purity from a PSA plant is subjected to the following purification steps:

1. Oxygen removal by catalytic conversion in an adiabatic reactor at ambient inlet temperature (electrolysis);
2. Adsorptive drying (electrolysis);
3. Adsorptive removal of remaining impurities nitrogen, methane, argon, carbon monoxide near their condensation temperatures, technically usually at LN2 temperature, (the adsorbers are
regenerated with a fraction of heated hydrogen product) (electrolysis and PSA)

The various purification processes are described in more detail in Chapter Purification of Hydrogen.

6.2. Liquefaction Plants

Up to now, there are only a few hydrogen liquefaction plants operational in the world. The largest, with capacities up to 60 t/d, are in the United States and distribute liquid hydrogen throughout the whole
country for general industry purposes and, in particular, as a rocket fuel to the space industry.

Some smaller plants are in operation in Europe, serving also space-transport related purposes, such as in Southern France, or scientific purposes as in the Federal Republic of Germany. The latest plant
with a capacity of 5 t/d was erected in the Netherlands in 1988. For the first time European gas customers are to be delivered with liquid hydrogen from this plant. Very large plants have been described.
Examples are those envisioned for the hydrogen energy concept with 10 modules, each of a capacity of 250 t/d [268].

Small plants use the Linde – Hampson process for liquefaction of hydrogen [269], [270]. The refrigeration is produced by expansion through a throttle at low temperatures. Precooling with liquid nitrogen
(LN2) at reduced pressure (bp 65 K) and relatively high hydrogen pressure (e.g., 12 MPa) is necessary. If liquid hydrogen is used directly, without storage, e.g., for bubble chambers, the o – p conversion
becomes unnecessary. The liquefaction unit becomes simpler, although the specific energy consumption of 100 kJ/g (including the LN2 consumption) still remains high.

The process can also be conducted in two steps. Thereby, the hydrogen is expanded to an intermediate pressure after LN2 cooling and part is recycled to the second stage of the compressor. In Figure 87
a two-stage Linde-Hampson process is shown.

Figure 87. Hydrogen liquefaction for a two-stage Linde – Hampson process

a) Prepurification, adsorption; b) Heat exchanger; c) Hydrogen compressor; d) Expansion valve

Large-scale plants for hydrogen liquefaction not only utilize Joule – Thomson decompression for the production of the necessary refrigeration but also the decompression of hydrogen in turbines.

Figure 88 shows the process scheme of a hydrogen liquefaction unit without nitrogen precooling [271]. Purified hydrogen is cooled down to ca. 85 K in three heat exchangers against decompressed
recycle hydrogen. A parallel stream of pressurized recycle hydrogen is also cooled down and, in the first turbine, provides part of the necessary refrigeration. In two further decompression stages the
refrigeration necessary to cool down hydrogen further and to compensate for the o-p conversion is produced. The liquefaction takes place to ca. 85 % by means of Joule – Thomson decompression of the
product hydrogen before entering the last reactor heat exchanger. Completion of the liquefaction and subcooling is accomplished by heat exchange with a small fraction of the recycle hydrogen cooled by
Joule – Thomson decompression to lower pressures. The recycle hydrogen quantity is ca. 12 times that of the liquefied hydrogen.

Figure 88. Process flow diagram of a commercial H2 liquefier without LN2 precooling (German Linde) [271]

page 85 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

a) Adsorptive and catalytic precleaning; b1 – b8) Heat exchangers; c) Adsorbers at LN2 temperatures; d) o-p conversion reactor; e1 – e3) Expansion turbines; f1 – f2) Expansion values; g1) Compressor; g2) Recycle
compressor for H2 coolant cycle

The T – S diagram (see Fig. 86) shows the thermodynamic cycle of this hydrogen refrigeration process. In particular, it illustrates the enthalpy differences obtained in the expansion turbines in three
temperature ranges. The efficiency and losses during the liquefaction of hydrogen are shown in the energy flux diagram of Figure 89. The specific energy consumption amounts to 60.6 kJ per gram of
liquid p-H2.

Figure 89. Sankey diagram of a commercial H2 liquefier, revealing the losses of the process (Process according to Figure 88)

b1 – b8) Heat exchangers; e1 – e3) Expansion turbines

Utilizing the energy produced in the turbines ( = 0.8) a thermodynamic efficiency of 23.7 %, corresponding to a specific consumption of 56.3 kJ per gram of p-H2 (98 % pure, rest o-H2), is obtained. The
power consumption of the hydrogen recycle compressor can be reduced considerably, if the cooling down to the temperature of liquid nitrogen is carried out with imported LN2, by a nitrogen refrigeration
cascade or by a multi-stage cascade with various refrigerants, e.g., propene, ethylene, methane and nitrogen (see Section Production from Coal and Hydrocarbons). The attainable saving in power must
be balanced against the higher cost of the equipment and utilities (LN2).

In large-scale plants [272] for LH2 production, the specific energy consumption can be reduced by the following means:

1. Use of LN2 for precooling,

2. Increased efficiency of compressors and turbines,
3. Decreased insulation losses, and
4. Decreased losses in heat exchangers and reactors.

6.3. Further Developments

Besides the Linde – Hampson process (refrigeration by isenthalpic decompression) and the Claude process (refrigeration by isentropic decompression) a number of other processes, such as the Stirling
process, have been suggested for the liquefaction of hydrogen. These have been partially tested in small pilot-scale plants.

In pilot plants for low temperatures between 1 and 4 K, magnetic refrigeration has already been experimentally verified [266], [273], [274]. Magneto-caloric refrigeration is achieved in a similar way to
thermodynamic processes, such as the isentropic expansion of gas. In the latter process the gas to be liquefied is used as a working medium, whereas in the magneto-caloric process the working medium
is a ferromagnetic material (e.g., in the form of a bed through which the hydrogen is flowing).

Carnot, Stirling, Ericson and the Brayton process, which operates between two isentropes and two curves of constant magnetic field strength, can be used for the technical realization of the magneto-
caloric process. The strong magnetic fields required (≥10 T) can be produced with the help of superconducting magnets.

The discovery of superconductors with a transition temperature around the boiling point of liquid nitrogen [275] is expected to lead to new developments in this area. Figure 90 shows the concept of a
magneto-caloric liquefier operating with helium as a cooling fluid. In the four regenerators (d1 – d4) helium is cooled by the magneto-caloric effect. After each stage a partial stream is drawn off for indirect
cooling of hydrogen. For the regenerators a common magnetic field coming from a superconducting magnet (8 – 10 T) is provided. The magnet here is still cooled with liquid helium, which must be
produced in an additional stage. The projected technical data of such a hydrogen liquefaction unit are listed below:

page 86 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Duration of magnetocaloric cycle, s ca. 10
Magnetic flux density, T ca. 10
Cooling power (at 20.4 K), W 20
Liquefaction rate, L/h 0.7
Liquefaction work, kWh/kg 7.26 (54 %
of Carnot)
Power consumption, W 557
Heat losses (at 300 K), W 621
Volume, L ca. 76
Mass, kg ca. 166

The low power consumption and the efficiency of 54 % of the Carnot cycle is remarkable [265], [270].

Figure 90. Principle flow diagram of a refrigerator using a magneto-caloric refrigeration cycle. The magnetic regenerators work in a cyclic mode

a) o-p Conversion reactors; b) Heat exchangers; c) Helium cycle compressor; d) Magneto-caloric regenerators

Hydrogen Slush. Liquid – solid hydrogen mixtures (slush) have an increased heat absorption capacity and a higher density than normal boiling liquids. This is an advantageous characteristic in space
application, where longer storage times and higher storage densities are important.

Hydrogen slush is produced by a freeze – thaw method, where vapor is continuously withdrawn from a stirred mixture of liquid and solid hydrogen, liquefied and reexpanded into the storage vessel. The
latent heat of vaporization is used to freeze the hydrogen. The mixture is then at its triple point (13.8 K, p-H2, 7 kPa) and has to be protected against air leakage into the system by keeping it under a
helium atmosphere.

Solid hydrogen can also be produced by freezing liquid hydrogen at higher pressure on a surface cooled by liquid or gaseous helium. In this case the higher pressure can only be maintained if a
temperature stratification can be generated at the surface of the liquid phase [276]. For properties of hydrogen slush see [277].

[Top of Page]

7. Handling of Hydrogen
7.1. Quality Specifications
The hydrogen content of technical hydrogen produced from hydrocarbons is between 97 and 99.5 vol %. The impurities are mainly methane and nitrogen; traces of oxygen, carbon monoxide, carbon
dioxide and, depending on the production and purification stages, hydrogen sulfide or ammonia may be present. Hydrogen produced electrolytically is typically >99.5 vol % pure. Impurities in this case are
mainly nitrogen, oxygen and water. The mercury content must be limited to max. 0.15×10–3 mg/m3 (for protection of the aluminum material in the downstream purification units [278]).

Hydrogen is available in high and very high purities. The impurities are removed by means of catalytic combustion (oxygen), drying systems (water) as well as by adsorption and diffusion systems.
Hydrogen of the utmost purity is needed, for example, in the semiconductor industry. It is purified by diffusion through palladium silver membranes or, more recently, by adsorption and purification on
hydride material (purity 7.0, i.e., total impurities <0.1 ppm [mL/m3]) [279]. Table 35 gives an overview of typical hydrogen purities commercially available.

Table 35. Commercial hydrogen qualities

Hydrogen “pure” Liquid

page 87 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
hydrogen
2.5 3.0 3.8 5.0 5.3 5.6 6.0 6.0

Purity, vol % 99.5 99.9 99.98 99.999 99.9993 99.9996 99.9999 99.9999

Impurities, ppm (mL/m3)

O2 * ≤ 50 ≤ 10 ≤2 ≤2 ≤1 ** ≤0.2

N2 ≤500 ≤200 ≤3 ≤3 ≤2 ≤0.2

H2O ≤100 ≤ 20 ≤5 ≤2 ≤1 ≤0.5

CO ≤0.1
CO2 ≤0.1

CnHm ≤0.5 ≤0.5 ≤0.1 ≤0.1

* Remainder N2, O2 and H2O.

** Total amount of impurities ≤1 ppm (mL/m3) thereof O2 ≤0.5 ppm (mL/m3) and CnHm ≤0.1 ppm (mL/m3).

7.2. Compression and Expansion

Compression. Hydrogen is often needed at high and very high pressures (e.g., for hydrogenations: 5 – 70 MPa, for ammonia synthesis: 8 – 30 MPa). Because of its low molecular mass, however, it is
difficult to compress in conventional machines. Special types of compressors have therefore been developed.

For very large amounts of hydrogen, (>50 000 m3 [STP]), such as are necessary, for example, in the ammonia synthesis, radial-type high-pressure turbo compressors have become standard. Because the
pressure ratio attainable with the hydrogen – nitrogen mixture is relatively low, multistage compressors are used. Leakage problems can be handled by suitable constructions. Extra stages of the synthesis
gas compressor are used to recycle hydrogen. In smallercapacity plants (<500 t/d of ammonia) the synthesis gas is compressed in so-called mole pumps. Compressor and motor are thereby enclosed in a
steel mantle.

For smaller quantities of gas and high end pressures piston compressors are chosen. Depending on the purity requirements placed on the compressed hydrogen three types may be used:

1. Oil-lubricated piston compressors for the compression of hydrogen for hydrogenation of oil products.
2. If the small quantities of oil in the hydrogen stream after the oil separator or adsorber cannot be tolerated, water-lubricated piston compressors can be used.
3. For the compression of high-purity hydrogen dry-running piston compressors are used. The pistons are sealed with specially pivoted and guided PTFE (polytetrafluoroethylene) piston rings or, to
avoid friction losses, with labyrinth seals. The leakage gas is fed back to the suction side of the compressor via an active carbon filter in both cases.

Smaller quantities of low-pressure gas requiring moderate compression ratios can be handled with screw compressors. This type of compressor is used especially in hydrogen liquefaction plants [280].

Apart from these possibilities, compression can also be achieved by pumping liquid hydrogen to the high pressure desired and then evaporating the liquid. This is used for the LH2 injection in some
hydrogen-fueled vehicle engines. Pressures up to 10 MPa may be obtained [281].

Because of the problems encountered in the compression of hydrogen, methods for the indirect compression have been considered.

In the Donor Solvent process ( Coal Liquefaction – Exxon Donor Solvent Process) [282] for coal hydrogenation, aromatic compounds at a low pressure of 5.0 MPa are first hydrogenated to naphthenic
compounds. When the hydrogenated donor liquid is subsequently subjected to the conditions prevailing in coal hydrogenation processes (400 °C, 25 MPa), the hydrogen is set free and used for coal
hydrogenation. In this way the amount of hydrogen to be compressed can be reduced considerably.

Decompression. The decompression of hydrogen-rich streams is used to produce refrigeration (see Chap. Liquefaction of Hydrogen) and to recover energy. For high pressure differences and small
amounts of gas, piston expansion machines are used; large amounts of gas are handled by turbines. For high pressure drops and large amounts of gas a multistage decompression in turbines is applied.
Depending on the power and the utilization of the energy the decompressors drive either generators, brake compressors, recycle compressors or oil brakes.

page 88 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The power thereby generated can be between 1 and 25×103 kW. Since the efficiency of turbines increases with increasing speed, high revolutions are desirable. Turbines with revolutions up to 1×105 rpm
are built. Depending on their power rating, these turbines may have oil, gas or magnetic bearings.

7.3. Transportation and Distribution

The consumers of hydrogen in the chemical industry (ammonia plants, synthesis gas plants, petrochemical plants) usually produce hydrogen on site. Small- and medium-size consumers can also be
supplied with gaseous or liquid hydrogen. The principal possibilities for hydrogen transport within the structure of regional hydrogen logistics are shown in Figure 91 [283]. In any future hydrogen energy
concept, transportation of large quantities of hydrogen between regions will be necessary and for this reason, transport in pipelines or liquid hydrogen transport in tankers is under consideration.

Figure 91. Possibilities for hydrogen transport

High-Pressure Hydrogen Transport. Small amounts of hydrogen can be transported in aluminum pressure cans (contents ca. 5 L H2) and in steel cylinders of various sizes (up to 70 L with H2 at
20 MPa). Greater amounts of hydrogen are transported in bundles of steel cylinders or in large sized pressure containers, transport being by road or rail. Modern high-pressure transports have a capacity
of >4000 m3 of hydrogen (see Fig. 92) and can be filled in less than 2 h. Oil-lubricated piston compressors are used to compress the gas to 20 MPa; oil and water adsorbers or, for higher purities, special
low-temperature adsorbers assure the necessary product purity.

Figure 92. Road transporter for hydrogen with 9 seamless pressure vessels, max. operating pressure 20 MPa (courtesy of Mannesmann Röhren-Werke, Düsseldorf)

Pipelines play a particular role in the transport of large quantities of hydrogen. Table 36 gives information on existing hydrogen pipeline nets. Such nets are run, e.g., in the USA by Air Products, in France
by L'Air Liquide and in the Federal Republic of Germany by Chemische Werke Hüls.

Table 36. Existing hydrogen pipelines

Length, km Diameter, mm Operating pressure, MPa Flow rate capacity

L'Air Liquide, France, Belgium, Netherlands GH2 total 300 100 <10
Sweden GH2 6 transfer 50 – 250 0.01 – 2.8 10 000 m3/h
lines, 6 km
Air Products, Houston, USA GH2 62 80 – 30 3.5 – 4.0 270×106 m3/a
Chemische Werke Hüls, Rhein-Ruhr-Gebiet, FRG GH2 208 100 – 300 2.2 250×106 m3/a
NASA, Florida, USA LH2 0.4 150 0.15

The last-named system was built in 1938. It has a length of 210 km and can transport 250×106 m3 (STP)/a; four suppliers and fourteen users are tied into the system. The experience gained with this
pipeline has been used to provide specifications for planning, constructing, operating and maintaining such a system [284]. If natural gas pipelines are to be converted for the transport of hydrogen, it is
necessary to consider the suitability of the steel material incorporated under the expected conditions in the presence of hydrogen (pressure level up to 10 MPa; cyclic loads; see Section Materials).

page 89 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
LH2 Transport. For laboratory uses liquid hydrogen is supplied in transport dewars. Larger quantities are transported in tanks ( 5000 L) or in tank trailers (3 – 5×104 L); in the United States liquid hydrogen
is also transported by rail (105 L containers). Oversea containers, containing up to 20 000 L of liquid hydrogen are used to supply the European space project Ariane in French Guayana. During the Apollo
program, barges with dewars containing up to 900 m3 of liquid hydrogen were used.

To keep evaporation losses small, the transport vessels have either perlite vacuum insulation or superinsulation (see Section Storage). Figure 93 shows a transport dewar with an additional liquid nitrogen
cooling system (evaporation loss <1 %/d) [285].

Figure 93. Laboratory transportation dewar with liquid nitrogen cooling [285]

a) Filling manifold and safety valve; b) Protective shell; c) Liquid hydrogen 21 K; d) Liquid nitrogen 77 K; e) High vacuum; f) Outer shell

Test rigs and transfer lines for liquid hydrogen in the space industry are vacuum-insulated. Piston pumps are suitable to transport small quantities of liquid hydrogen at high pressure, whereas larger
quantities require multi-stage radial or axial pumps.

Transportation Costs. The total costs of the transport chain from the hydrogen production to the user site have often been assessed, e.g., in [286], [287]. A number of parameters have to be considered,
e.g., to transport the same effective energy through transfer lines hydrogen requires 3.5 – 4 times the amount of compression energy compared with natural gas; the total costs are higher by a factor of
1.3 – 1.6. During the distribution of LH2 an evaporation loss of 7 – 11 % of the handled quantity must be calculated [288]. A summary of the transport and distribution costs for hydrogen is given [289].
Pipelines are by far the most cost-effective means of transporting gaseous hydrogen over both short as well as long distances. The transport of liquid hydrogen via rail or road is ca. 5 times more
expensive, the use of pressure gas cylinders is by far the most expensive (transportation costs are ca. 50 times higher, including the costs of idle equipment, such as gas cylinders and vehicles).

7.4. Storage
Conventional methods for storing hydrogen are well-proven and in service. Depending on the end use, (1) stationary, large-size storage systems, (2) stationary, small-size storage systems, and (3) mobile
storage systems can be defined. The latter may be used for transport, for distribution or as a fuel reservoir for prime movers. In all these types of storage systems, hydrogen can be handled in the gaseous
(GH2) or liquid (LH2) state.

Here methods, such as gas storage under pressure or storage in liquid form, are described. Newer concepts, such as absorption in metals (hydride storage systems) and cryoadsorber storage systems
are being developed with a view to future use and are described in Section Hydrogen-Energy Conversion Systems. Table 37 shows various concepts for the storage of hydrogen [289]. For stationary,
large-volume systems the optimum storage type is determined by the specific storage capacity T. This is the amount of energy storage capacity per unit of the rated charging power (kWh/kW). Specific
storage cost estimations including further parameters, e.g., utilization, storage capability, annual cost of facility and subsystems are presented in [290].

Table 37. Overview of processes and possibilities for hydrogen storage

Designation Compressed hydrogen gas storage Absorption by metal Cryo- Liquid hydrogen
hydrides adsorption storage

below low high e.g., FeTi e.g., FeTi large small

ground pressure pressure + MgNi volume volume
tank tank

Operating conditions
Pressure, MPa 10 1.2 15 0.3 – 5 0.3 – 5 4.2 0.15 0.15

Temperature ambient ambient ambient 0 – 80 °C 80 – 300 °C 77 K 21 K 21 K

Developmental status commercial commercial commercial development development development commercial commercial
Range of application site-dependent medium short term or low to medium mobile medium long term high low capacity
seasonal storage capacity mobile capacity storage capacity capacity

page 90 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Specific energy storage >1000 <30 <30 <3 30 – 45 >50
capacity T, kWh/kW
Problem areas availability, safety costs, energy capacity, refrigera-tion safety evaporation
diffusion losses transfer weight source losses, safety

Compressed Hydrogen Storage. Pressure storage systems of various sizes and pressure ranges are considered state of the art. The same applies for temporary storage in pipelines (“line packing”). For
hydrogen, steel cylinders (up to 70 L at 20 MPa), stationary high pressure vessels (20 m3 at 20 MPa) or low-pressure spherical vessels (15 000 m3 at 1.2 – 1.6 MPa) are available. Low-pressure and high-
pressure systems are suitable for the storage of day or week supplies.

Underground storage in caverns or aquifers is expected to be possible; storage pressure is up to 16 MPa. Gas losses up to 1 – 3 % of the storage capacity are foreseen [291]. Specific problems due to
hydrogen are not anticipated, however, the availability of suitable salt and aquifer formations is limited.

Liquid Hydrogen Storage. The technique for liquid hydrogen storage is well-developed; it is used for the storage, transport, distribution of hydrogen. The vessels are insulated with double-walls (dewars),
the space between the inside and outside wall is evacuated and filled with perlite (large vessels) or super-insulation. The rate of loss because of evaporation diminishes with increasing size of the vessel.
The losses are defined by the amount of hydrogen evaporated daily. Laboratory vessels have 2 – 3 %, mobile LH2 transport tanks 1 %, perlite-insulated large-volume vessels <0.1 % losses. For additional
reduction of heat leakage the refrigeration available in the boil-off gas can be utilized to cool the insulation. This is achieved by radiation shields thermally connected to the vent tube [292]. Liquid hydrogen
vessels are attainable from 100 L cans to large tanks with up to 5000 m3 volume. Tanks have a high specific storage capacity (ca. 100 kWh/kW) and are suitable as long-term storage systems. Figure 94
shows a tank with a capacity of 2000 m3 LH2 installed at the Kennedy Space Flight Center in Florida.

Figure 94. Liquid hydrogen tank with a capacity of 3800 m3 LH2 at the Kennedy Space Flight Center in Florida (Photo NASA)

Super-Insulation [293]. A large number of layers of insulation foil is wound around the inside vessel (10 – 60 layers per centimeter). A layer consists of a plastic film, both sides of which have been
sputtered with metal, or it is produced from a very fine aluminum foil, which is held apart by spacers of a fabric made from glass filaments or similar material. Energy transport by radiation and conduction
of the insulation material is more or less completely interrupted. Values of 5×10–6 Wm–1K–1 for the heat conductivity can be attained.

Liquid hydrogen storage costs have been estimated [286], [294] as has also been the cost of liquefaction [295]. The energy consumption to liquefy the hydrogen amounts to ca. 1/3 of its lower heating
value.

Cost Comparison of Large-Volume Hydrogen Storage. The energy stored in a certain volume of hydrogen, relative to that stored in the same volume of natural gas, is determined by the ratio of the
volumetric heating values. Therefore, for otherwise equal systems, the hydrogen storage is more expensive by the factor of ca. 3 than natural gas storage. Studies of the cost situation for various hydrogen
storage systems allow the following conclusions [289]:

1. Underground hydrogen storage is the most economic method for all areas of application.
2. Storage in pressure vessels is the most economic above-ground storage system when t <30 h. It is suitable for all short-term storage systems.
3. LH2 storage is the most economic storage system above ground for t > 30 h.
4. It is not clear, whether the recent concepts for hydrogen storage (metal hydrides or cryo-adsorber systems) (see Section Hydrogen-Energy Conversion Systems) are technically and economically
more favorable options.

7.5. Materials
Metallic materials are used during production, transport and storage of hydrogen. The choice of material and its stability in the presence of gaseous, liquid or electrolytically produced hydrogen is therefore
of great importance and essential for safe handling of hydrogen in a future hydrogen energy carrier concept [296]. Literature on hydrogen and corrosion is given in [297-299].

Mechanism of Hydrogen Corrosion. The destruction caused by hydrogen in various materials occurs via different mechanisms. The starting point is always the formation of diffusable atomic hydrogen.
This can be formed by: hydrogen ions, thermal dissociation, or chemisorption and dissociation on active metal surfaces.

The embrittlement induced by formation of atomic hydrogen is well-known in electrolytical processes with gaseous hydrogen. Such defects are mainly encountered if the material surface is activated by
plastic deformation, i.e., cyclic stress loading makes a metal more susceptible to hydrogen attack.

page 91 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Atomic hydrogen is dissolved in the material, whereby it either occupies interstitial sites or accumulates in so-called traps. The reactions which lead to damage are described.

Types of Reaction Leading to Damage. The dissolved diffusable atomic hydrogen can interact with lattice dislocations, lattice defects, internal surfaces, or fields of tensile strain caused by mechanical
distortion of the material. At these locations, segregation of the material occurs if the concentration of accumulated hydrogen atoms reaches a critical value. The formation of cracks is thereby initiated or, if
microcracks or intergranular fractures are present, crack growth is enhanced (HIC, hydrogen-induced cracking [300]).

Hydrogen can react with the material and form hydrides, e.g., with titanium, niobium, tantalum, molybdenum and tungsten. Because of the fact that enhanced hydride formation takes place in the stress
field of cracks, and that hydrides are extremely brittle materials, such elements in alloys can lead to hydrogen embrittlement.

A reaction of hydrogen with carbon contained in ferritic steels and alloys at higher temperature (t > 200 °C) is methane formation. This interaction is called decarburization.

Recombination of atomic hydrogen to form molecular hydrogen can also be observed. Both reactions, formation of molecular hydrogen and methane, cause high internal pressures because a gas is
trapped inside the material. This can lead to a weakening at the grain boundary, followed by formation of cavities, internal cracks and blistering (fish eyes). Macrocracks can form, if the material is placed
under strain.

Effects of Hydrogen Pressure Below Temperatures of 200 °C. At constant pressure no critical reaction is observed. Carbon steel equipment may be used up to a recommended hydrogen pressure of
90.0 MPa. At higher pressure austenitic stainless steel should be used.

In areas where the strain changes periodically, e.g., in transport and storage vessels, changes in the ductility and rate of crack formation have been observed. These changes are influenced by the
hydrogen pressure, the temperature, and the frequency of the cycling. The rate of the fissure formation was found to increase by a factor of 3 – 100, depending on the cycle frequency and hydrogen
pressure as compared to the same treatment in a hydrogen-free atmosphere [301].

Effects of Hydrogen Pressure at Temperatures Above 200 °C. Hydrogen attack with accompanying decarburization by methane formation occurs at >200 °C and increases rapidly with increasing
temperature. It depends on the hydrogen partial pressure and on the type of alloyed steel. Metals such as chromium, molybdenum, tungsten, vanadium, titanium, and niobium reduce the number of
nucleation sites by lowering the amount of carbon available for methane formation (carbide stabilizers). The behavior of steel at elevated temperature and hydrogen pressure is summarized in a diagram
(see Fig. 95). The original curves “Operating Limits for Steels in Hydrogen Service” were first published by G. A. NELSON in 1949 [302]. They were developed as a result of data from a variety of
commercial processes and laboratory experiments. Periodic revisions have been made over the years, the latest being published in [303].

Figure 95. Operating limits for steels in hydrogen service (Nelson diagram)

------- surface decarburization; —— internal decarburization (hydrogen attack)

Below and to the left of the curves for each alloy satisfactory service has been experienced at periods of exposure up to 35 a. The length of time before hydrogen attack can be detected is termed the
“incubation period”. It is maintained, that during the incubation period methane pressure builds up in submicroscopic voids; when a void reaches a critical size, microscopically visible fissures appear and
the growth rate thereof increases [304]. At temperatures >540 °C the curves are not verified. In this temperature range, high-alloy austenitic steels, which resist decarburization, must be used.

Importance of Foreign Substances. Impurities in hydrogen can inhibit hydrogen attack more or less. Carbon monoxide and water reduce the corrosion activity, oxygen and sulfur dioxide are particularly
inhibitive.

Ammonia causes the formation of nitride layers on steel which, when subject to deformation, are ruptured. For this reason, steels used for ammonia synthesis plants must be suitable for both hydrogen
and nitride attack. Above 600 °C carbon monoxide causes a carburization of the material. Below 450 °C carbonyl formation can occur which simply removes layers from the iron surface (maximum at PCO
12.5 MPa, 250 °C). Carbon dioxide and hydrogen sulfide are strongly oxidizing at elevated temperatures and can accelerate hydrogen attack. A knowledge of the various forms of hydrogen attack on
metallic materials together with the effect of other gas components is especially necessary for high-pressure processes. The choice of material is of particular importance. A summary is given in [305].

Other Hydrogen-Resistant Materials. Other hydrogen-resistant materials are:

Copper and copper alloys up to 400 °C,

Nickel up to 250 °C,
Monel up to 250 °C, and
Inconell up to 540 °C.

For temperatures > 1400 °C the following materials are suitable:

page 92 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Aluminum oxide
Zirconium oxide
Silicon carbide
Silicon nitride

In general, plastics are not subject to attack by hydrogen at ambient temperatures. At higher temperature, the following materials are suitable:

Poly(vinyl chloride) and polyethylene (up to ca. 60 °C)

Polychloroprene and polyisobutylene (up to ca. 100 °C)
Poly(tetrafluoroethylene) (up to ca. 250 °C)

Materials for Liquid Hydrogen (LH2). At the low temperatures of liquid hydrogen none of the corrosion inducing mechanisms can take place. Material problems are reduced to the effect on mechanical
properties.

Mechanical design with metallic and non-metallic materials at low temperatures as well as construction and assembly methods are compiled [306]. The alteration of material properties, such as elastic
moduli, fracture behavior, plasticity and fatigue behavior, is handled in detail.

Only a few metals and alloys commercially available are used in cryogenic engineering. Aluminum alloys are widely used, because of their relative cheapness, good weldability and high toughness.
Austenitic steels (300 series) are also widely used, because they retain high strength and good ductility down to low temperature. For cryogenic work, low carbon grades are preferred to prevent the
precipitation of chromium carbides during welding. Other materials used are nickel-iron, titanium, and copper alloys.

Selection and use of organic polymer materials at cryogenic temperatures is described [307]. Modification of epoxy or polyester resins by incorporation of inert fillers or the use of fiber-reinforced
composite materials and laminated structures using glass, Kevlar, carbon, etc., give materials satisfactory for cryogenic service.

7.6. Removal of Hydrogen

The removal of hydrogen may become necessary, if it develops in an uncontrolled manner and causes a safety risk or if a gas must be purified from traces of hydrogen. Normally, hydrogen, which cannot
be utilized, is burned, often together with other disturbing compounds and combustible waste gases. If the hydrogen content is not sufficient to support a flame, the so-called flameless combustion with
oxygen at 700 – 800 °C with external heating can be used. This technique is applied particularly often for fume incinerators.

More common is the catalytic combustion on noble metal catalysts (Deoxo) or mixed-oxide catalysts, e.g., CuO – MnO2 (Hopcalite or similar) [308]; catalytic recombinators are also used in maintenance-
free lead accumulators.

Further, hydrogen can be removed by absorption and oxidation, e.g., by aqueous solutions of silver or palladium salts or sodium chlorate. As a solid “absorbent” copper oxide or nickel oxide is suitable.

For the removal of 1 m3 (STP) of hydrogen ca. 3 kg of copper in the form of copper oxide are necessary; the reaction takes place at 200 °C, the catalyst can be regenerated with air. For production of high-
purity inert gases hydrogen traces can also be removed by means of hydride material (see Section New Developments for Hydrogen Storage).

Hydrogen Mitigation in the Nuclear Industry. During normal operation in water-moderated nuclear reactors small amounts of hydrogen are permanently formed by radiolytic splitting of water. In the
waste-gas system of nuclear plants recombinators are installed. These are usually commercially available catalytic or thermal units.

In case of a nuclear core uncovery accident, hydrogen may be produced by several mechanisms. (1) The exothermic zircalloy – steam reaction occurs once the fuel cladding reaches temperatures of
900 – 1200 °C. (2) A considerable degree of radiolytic water cleavage inevitably accompanies an emergency core cooling procedure (with a great amount of water) for a period of time after the loss of the
coolant has occurred. Hazard analysis studies provide values of 5000 m3 of hydrogen formed within 50 d [309]. It is believed that at Three Miles Island hydrogen amounting to 8 vol % (370 kg)
accumulated in the atmosphere of the containment building. (3) The interaction of a hypothetical core melt with concrete or other materials present within a containment system is considered to be an
additional source of hydrogen.

Various specific countermeasures to cope with the hydrogen problem have been studied; some of them have been proposed for installation or have been implemented [310]. First of all, hydrogen
distribution is to be prevented by subcompartmented containments, and devices to reduce hydrogen concentration should be provided (fans, recombinators, etc.). In case of massive hydrogen release in
small containments, blanketing with nitrogen is foreseen, whereas in intermediate containments the use of water fogs or Halon 1301 injection is recommended to mitigate hydrogen deflagration risk and
inhibit combustion. Finally, deliberate ignition and controlled burning systems are reported to be effective hydrogen controlling measures, whereby catalytic and spark hydrogen igniters can be used [311],
[312].

page 93 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
7.7. Safety Aspects
General. Hydrogen forms inflammable and explosive mixtures with oxygen and other elements. Basic ignition and explosion data are given in Section Ignition and Detonation Performance. For
comprehensive compilations of safety characteristics and hazards on handling hydrogen see [313-315].

The envisaged large-scale introduction of hydrogen as an energy carrier requires a comparative evaluation of safety aspects with regard to other fuels. Comparative data for methane (representing natural
gas), propane and gasoline are shown in Table 10. Volumetric leakage of hydrogen gas will be 1.3 – 2.8 times as large as gaseous methane leakage and approximately four times that of air under the
same conditions (rule: “air-proof is not hydrogen-proof”).

Released hydrogen disperses rapidly by turbulent convection, drift and buoyancy, thus shortening the hazard duration. The prompt dispersion, however, favors the formation of gas mixtures within the
wide flammability and detonability limits; the lower limit is the vital one in most applications and is comparable to that of other fuels.

The minimum energy for ignition of hydrogen – air mixtures is extremely low. However, the ignition energy inherent in virtually every source is more than sufficient for ignition of any other fuel – air mixture
as well. Ignition by catalytic action is also possible.

Fire Hazards. Hydrogen flames are nearly invisible in daylight. Fire-detection technology is summarized in [316]. Hydrogen fires last only one fifth to one tenth of the time of hydrocarbon fires, and the fire
damage is less severe because of several characteristics:

1. high burning rate resulting from rapid mixing and high propagation velocity,
2. high buoyant velocity,
3. high rate of vapor generation of liquid hydrogen

Although the maximum flame temperature is not much different from that of other fuels, the thermal energy radiated from the flame is only a part of that of a natural gas flame. Smoke inhalation, one of the
major causes of injury and, therefore, a main parameter of fire damage, is considered less serious in the case of hydrogen because the sole combustion product is water vapor.

Explosive Hazards. High laminar burning velocity as well as the high laminar flame speed of hydrogen makes the transition to turbulent flame speeds exceeding 800 m/s up to several km/s easy. Hence,
hydrogen is more sensitive to deflagration to detonation transition (DDT) than hydrocarbons. Studies on DDT mechanisms are referred to in [317]. Hydrogen has by far the widest limits of detonability;
detonation of stoichiometric hydrogen – air mixtures in confined spaces will produce a static pressure rise of ca. 15 : 1. If a DDT process precedes, i.e., unburnt hydrogen – air mixtures are
precompressed, a pressure rise ratio of 120 : 1 could result.

Explosions are rated in terms of the amount of energy released, commonly expressed as an equivalent quantity of TNT. The theoretical maximum values of explosive potentials for fuels (TNT equivalents)
are recorded in Table 20. Experimental data indicate, that real yield factors of 10 % are considered reasonable. Note that hydrogen is most potent on a mass basis and least potent on a volumetric basis. It
has also the least theoretical explosive potential, when equivalent energy storage is taken into account.

Preventive Measures. After analyzing accidents involving hydrogen and hydrogen-carrying equipment, preventive safety aspects are discussed in [318]. Risk avoidance by general preventive design
measures, as well as reactive fire and explosion measures are presented [315], [318], [319]. In safety concepts, distinctions are made between primary, secondary and tertiary measures. Primary safety
precautions aim at the exclusion of causative risks such as leakage, formation of explosive mixtures by proper conceptual design (inertization, open-air installation, flame arrestors, etc.). Secondary
measures consist mainly in the avoidance of ignition sources of any kind (electrostatically or mechanically generated sparks). Tertiary measures should minimize dangerous results in case fire or explosion
occurs. This is achieved by installation of explosion-proof or explosion relief systems, hydrogen process shut-down systems and suitable fire extinguishing systems.

Safety Regulations. Regulations (mandatory) and standards (mandatory or nonmandatory) apply for the safe production, storage and handling of hydrogen. They are mostly concerned with
transportation; other operations are covered by more general regulations. Excellent listings are given for the United States [320] and for Canada [315].

United States. Examples of nonmandatory standards are those issued by ASME, ANSI, NFPA, etc., that may be adopted by regulatory bodies. Current mandatory regulations are Title 49 of CFR (Code of
Federal Regulations) and the requirements of DOT (Department of Transport) and CGA (Compressor Gas Association). General industrial safety matters including the production and handling of
inflammable compressed and liquefied gases are regulated by OSHA [321] (Occupational Safety and Health Administration).

Federal Republic of Germany. Mandatory regulations are Druckbehälterverordnungen (TRB, TRG, i.e., regulations for installations and examination of pressure vessels and filling thereof),
Unfallverhütungsvorschrift Gase (UVV, i.e., instructions for the prevention of accidents) and GGVS and GGVE, (i.e., instructions for the transport of dangerous goods).

EEC Regulations. For the EEC the association of the European Gas Industry (IGC) is making efforts to introduce common and uniform safety standards.

Future Outlook. Considering the possible use of hydrogen in a future hydrogen energy concept (see Chap. Hydrogen Energy), no major safety problems are anticipated. Hydrogen has been handled
successfully for decades in the process industries. Handling as a liquid is also routine. In comparison with other inflammable gases or liquids, the safety risk of hydrogen as a fuel is not significantly
different.

7.8. Toxicology
Hydrogen does not show any physiological effect. It is nonpoisonous. Inhalation of the gas leads to sleepiness and a high-pitched voice. A danger of asphyxiation exists if the oxygen content sinks below

page 94 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
18 vol % because of hydrogen accumulation in the air. Direct skin contact with cold gaseous or liquid hydrogen leads to numbness and a whitish coloring of the skin and to frost bite. The risk is higher
than, e.g., with liquid nitrogen because of the greater temperature differences and the higher thermal conductivity of hydrogen.

7.9. Analysis
The physical properties of hydrogen, some chemical reactions, and its nuclear spin characteristics are used for hydrogen analysis.

Hydrogen in Gas Mixtures. For rapid determination or for continuous measurement in process gases the absolute determination (or the comparison to a reference gas) of properties such as gas density,
conductivity or sonar velocity can be used. The most frequently used method for analysis of hydrogen in all types of gas mixtures is gas chromatography. Apart from mixtures also containing helium,
hydrogen is eluated first. In the case of detection with a hot-wire detector using helium as carrier gas an unusual nonlinearity exists in the conductivity of hydrogen – helium mixtures, which leads to a
minimum at 8 % hydrogen. This must be taken into consideration, or alternately argon can be used as carrier gas. Hydrogen contents in the percent range can be determined by separation and integral
detection according to JANAK (CO2 as carrier gas, absorption of the carrier gas in KOH, volumetric determination of the gas in an azometer).

Gas-chromatographic determination of the six isotopic species and their nuclear spin isomers is possible. Alumina or silica gel at liquid nitrogen temperature [322] or complex polymer molecular sieves at
room temperature [323] are suitable separation materials. If these are impregnated with paramagnetic substances such as Fe2O3 or MnCl2, rapid o – p isomerization is achieved and the spin-isomer
mixture corresponding to the equilibrium at the separation temperature is eluated [324].

Trace impurities in hydrogen can be determined qualitatively by low-temperature condensation or by enrichment on silica gel at low temperatures and, after desorption, they can be separated by gas
chromatography. A method for the determination of traces of carbon monoxide or carbon dioxide is methanation and analyzed with hydrogen followed by methane detection with a sensitive flame
ionization detector.

Chemically Bound Hydrogen. In organic molecules hydrogen is determined using classical elemental analysis. After the sample is weighed (usually in the micro- or submicro-range, i.e., 0.2 – 10 mg and
0.02 – 0.2 mg, respectively) it is combusted and the amount of water produced, as determined by weighing, is related to the hydrogen content. Nowadays automatic C, H, N analyzers with simultaneous
detection systems are used. The content of hydrogen in metals, an important value for steel production, is determined firstly by vacuum extraction of the hydrogen from the heated or molten sample and,
finally, by volumetric (vacuum pipette) or heat conductivity measurement [325]. In a new method, a sample is taken by inserting a porous ceramic bell and purging out the hydrogen with a nitrogen stream
[326].

“Active hydrogen”, e.g., hydrogen in OH groups or in enolizable C-H-bonds, reacts with LiAlH4 or metal organic methyl compounds and can be determined as hydrogen or methane volumetrically.

Modern instrumental methods are mass spectrometry and nuclear spin resonance. By means of the first method hydrogen and its isotopes in complex gas mixtures can be analyzed over a large
concentration range. Nuclear magnetic resonance (1H-NMR) is an invaluable instrument for the analysis and structure determination in organic and inorganic chemistry. 1H-NMR spectroscopy is also
applied for quantitative determination of the hydrogen content in various materials such as coal or hydrocarbons, to measure oil and water contents, e.g., in the food industry, or to pursue intricate
problems in reaction kinetics [325].

[Top of Page]

8. Conventional Uses
8.1. Hydrogen in the Chemical Industry
Hydrogen is an important raw material for the chemical industry. It takes part in reactions either by addition (hydrogenation) or by means of its reduction potential. Most of the hydrogen is used for these
hydrogenation and reduction processes. An overview on production and consumption structures for the Federal Republic of Germany (1978) and the United States (1984) is given in Figure 96 [328].
Changes in the production and utilization sector have been relatively little since that time. On the production side the fraction of the hydrogen produced using natural gas has increased at the expense of
the LPG and naphtha-based plants. Some plants based on heavy oil have even been closed down and replaced by natural gas or lignite gasification plants. In the utilization sector, hydrogen production is
stagnating. Apart from ammonia synthesis, synthesis with hydrogen – carbon monoxide gas mixtures to produce methanol, hydrocarbons and oxo-synthesis products are of particular importance. Large
quantities of hydrogen are needed in refineries for hydrotreating, whereas its use in coal processing is just beginning. The use of hydrogen in metallurgy is based in particular on its reducing properties.

Figure 96. Hydrogen production and consumption structures

I Production, Federal Republic of Germany, 1978

II Consumption, Federal Republic of Germany, 1978

III Consumption, United States, 1984

At present, pure hydrogen is not used directly to produce energy (apart from space technology). Mixtures with hydrogen, however, have been used for a long time for combustion purposes. Even today the

page 95 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
amount of hydrogen distributed in urban or in industrial heating gas systems is very large.

In comparing the hydrogen production or usage data, a distinction must be made between the so-called captive hydrogen and the merchant hydrogen: Captive hydrogen uses, such as for ammonia and
methanol synthesis or in various hydrotreating processes, relate to onsite hydrogen production and immediate consumption of the hydrogen. The designation merchant hydrogen is given to those
quantities which are sold on the market. Merchant sales of hydrogen account for 10 % or less of the total hydrogen production. Import and export amounts are normally negligible. Table 38 gives a
summary of a literature survey on hydrogen quantities produced as well as some forecasts for the future.

Table 38. Consumption of hydrogen and forecast consumption amounts [329-338]

Consumption of hydrogen, m3×109

Federal Republic of Germany 1978: 17 [329] 1984: 19.1 [330] 2000: 23 [331]
Japan 1985: 20 [335]
United States 1978: 80 [332] 1985: 90 [333] 2000: 146 – 207 [334]
Worldwide 1980: 300 [336] 1986: 500 [337] 2025: 1500 – 2000 [338]

8.1.1. Ammonia Synthesis

Hydrogen reacts with nitrogen to yield ammonia. The reaction is exothermic.

For details on the properties, production and usage of ammonia see Ammonia.

Technical ammonia syntheses are operated at 15 – 25 MPa. On a catalyst the reaction starts at ca. 350 °C, maximum reaction rates are reached in the range of 470 – 530 °C.

1 965 m3 (STP) of hydrogen and 668 m3 (STP) of nitrogen are necessary for each tonne of ammonia produced. For hydrogen production, the usual processes are suitable (see Section Production from
Coal and Hydrocarbons). The production of the ammonia synthesis gas is usually followed directly by the ammonia synthesis. The necessary nitrogen content in the synthesis gas is introduced by the
addition of air in a secondary reformer or via a liquid nitrogen scrubber. The conversion achieved per pass is ca. 25 – 35 %, so that a recycle loop is installed. Condensed liquid ammonia is removed
continuously. Impurities containing oxygen, such as carbon monoxide, carbon dioxide, oxygen and water, are catalyst poisons and must be completely removed from the synthesis gas.

The make-up gas to the synthesis loop contains certain amounts of inert gases such as methane (from methanation), argon (from the process air) or helium (from natural gas). To prevent accumulation of
the inert gases in the synthesis reactor, some recycle gas must be continuously removed from the loop, thus causing a loss of hydrogen. Most plants nowadays have a recovery unit:

cryogenic hydrogen recovery units, sometimes combined with argon production (see Section Low-Temperature Processes)
PSA hydrogen recovery units (see Section Adsorption Processes)
membrane separation hydrogen recovery units (see Section Permeation Processes).

Hydrogen recovery by reversible hydride formation has also been suggested [339].

Modern synthesis plants have a capacity of 1000 – 2000 t/d; this corresponds to an hourly hydrogen requirement of 80 000 – 160 000 m3 (STP). At present 75 – 80 % of the synthesis hydrogen is
produced from natural gas, naphtha or LPG. 10 – 15 % are made from heavy fuel oil by partial oxidation and ca. 5 % from coal. Electrolysis hydrogen is used for ammonia production only in special cases,
e.g., if cheap electric power is available. Small units are situated in Norway, Peru and Egypt [340], producing ca. 5 % of the world production of hydrogen.

An ammonia molecule contains three hydrogen atoms. For hydrogen utilization in future energy systems (see Chap. Hydrogen Energy), ammonia has often been suggested as a carrier. Hydrogen could
be stored and transported more economically in the form of ammonia [341], but the latter is untenable as a direct fuel because of its toxicity and the formation of NOx. However, on reconverting the
ammonia to hydrogen the stored energy can be made available at any desired site. An overview of the possibilities of an ammonia-hydrogen concept in energy technology is given in [342].

8.1.2. Hydrogen in Refinery Processes ( Oil Refining)

In refineries hydrogenation processes are included to increase the hydrogen content of the heavy crude oil fractions and to produce lighter fractions by reduction of the molecular mass. At the same time,
undesirable elements, such as sulfur, nitrogen and metals, can be removed. For this group of related processes the designation hydroprocessing is used.

Process Fundamentals. Hydrogenation processes are characterized according to the feedstock, the partial pressure of the hydrogen used and the reaction temperature (see Fig. 97) [343]. In

page 96 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
hydrotreating processes addition reactions and hydrogenating cleavage (hydrogenolysis) occur. In hydrocracking processes mainly dehydrogenation, cracking, isomerization and dealkylation reactions
take place. For the catalysts used in the individual processes, promoter and carrier materials, selectivity, life-times, coke and metal deposition see [344]; a compilation of available hydroprocessing
catalysts is given in [345]. Newly developed bi- or multifunctional catalyst systems based on zeolites combine cracking functions on acid centers of the carrier materials with hydrogenation and
dehydrogenation functions of the active metal components.

Figure 97. Process temperatures and pressures of refinery hydrogenating processes

a) Hydrocracking; b) Residue desulfurization; c) Heavy gas oil catalytic cracker feed hydrofining; d) Middle distillate; e) Naphtha hydrofining

The processes are carried out in one or two stages in the gaseous or liquid phase. Table 39 shows some examples of the most important hydroprocessing routes, typical process conditions and product
ranges. Overviews are given in [346] and [347].

Table 39. Feed and product characteristics and process parameters of various hydrogenation processes during hydroprocessing

Unionfining (UNOCAL) Go-Fining (Exxon) Hydroconverting (Linde) Hydrocracking (BASF/IFP)

Feed virgin kerosine arabian heavy VGO a VGO a arabian light VGO a
Density, g/cm3 0.806 0.933 0.910 0.909

Boiling range, °C 160 – 240 370 – 560 340 – 550 370 – 475

Sulfur content, wt % 0.4 2.96 2.7 2.55

Processing conditions
Reactor stages 1 1 1 2 (for max. gasoline
production)
Pressure, MPa 2.5 – 6.0 5.8 moderate

Temperature, °C 340 395 400°/400 °C

Hydrogen consumption, m3/t 10 87 150 390

Products
Yields, wt % naphtha, 1 % LPG, 0.4 % LPG + gasoline, C4-, 15.6 %
39 % light gasoline, 23.8 %
destillate, 99 % gasoline, 0.8 % middle distillate, heavy gasoline,
28 % 61.3 %
middle distillate, 100.6 % residues, 33 % H2S + NH3, 2.8 %
Main product
Density, g/cm3 0.890 0.813 0.664/0.758

Boiling range, °C 163/240 >200

Sulfur content, wt % 0.01 0.1 0.2

RON b of gasoline fraction >88 86

a VGO = vacuum gas oil.

b RON = research octane number.

page 97 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The use of hydrogen has also been suggested in other refinery areas; addition of hydrogen to the crude oil distillation is reported to make this more economic (hydrostripping, [348]). Subjecting residual
oils or heavy crudes to hydrogen for short periods at high temperatures and pressures of up to 50 MPa leads to feed upgrading (hydropyrolysis, [349]).

Hydrogen Requirements. The hydrogen necessary for the hydroprocessing is produced in the refinery. The main source is the catalytic reforming unit where ca. 45 m3 of hydrogen per ton of processed
crude oil is produced. Additional hydrogen is produced by steam reforming of methane, LPG, and naphtha, and by partial oxidation. In the United States 60 – 70 % is produced by catalytic reforming, 30 %
by steam reforming, and 2 % by partial oxidation.

The connections between processes utilizing and processes producing hydrogen in oil refining operations are shown schematically in Figure 98.

Figure 98. Integration of H2-producing and H2-consuming process units into crude oil processing

H2 units from H2-producing process

⇒ H2 units to H2-consuming process

The carbon:hydrogen ratio in the feedstocks (heavy and residual crudes, tar sands, oil shale) as well as their sulfur, nitrogen and metal components will become increasingly unfavorable in the future. To
cover the expected increasing hydrogen demand it is necessary to improve the hydrogen recovery in refineries and, eventually, to utilize other potential sources of byproduct hydrogen.

8.1.3. Hydrogen in Coal Refinement

Coal refinement processes which are performed in the presence of hydrogen are coal hydrogenation, hydropyrolysis and hydrogenating coal gasification.

Coal Hydrogenation. ( Coal Liquefaction – Direct Coal Liquefaction, the first three Sections). By thermal or thermocatalytical treatment in the presence of hydrogen under pressure, coal is almost
completely converted into liquid and gaseous products (direct coal liquefaction). Details of the chemical fundamentals, types of process and reaction techniques are given in [350]; the use of catalysts is
described in [351]. The reaction is catalyzed by iron compounds (sulfides or oxides), molybdenum, cobalt and tin or even mineral clays. Catalysts which can be cheaply disposed of are especially favored,
but also highly-active cobalt – molybdenum carrier catalysts are being investigated.

Hydrogen can either react directly (in its molecular form) with coal (direct hydrogenation) or it can be transferred by means of hydrogen-carrying solvents (donor solvent coal liquefaction or hydrogenating
extration) ( Coal Liquefaction – Exxon Donor Solvent Process). This type of process with chemically bound hydrogen can be carried out in the presence of an additional amount of molecular hydrogen
and is known as short hydrogenation, the product being primarily bitumen or solvent-refined coal (SRC).

Hydrogen is required for the important process stages, such as hydrogenative cracking in the slurry phase (feedstock slurries of hard coal, lignite or coal extract), cracking and refining hydrogenation in
mixed phases (feedstock hydrogenation residues, tars) and for cracking and refining hydrogenation in the gas phase (feedstock middle distillates and bottom cuts). Donor solvent processes employ
external, catalytical or noncatalytical hydrogenation of the recycled solvent.

Hydrogen Consumption and Product Range. The products obtained range from small quantities of coal residue over heavy and middle distillates and gasoline to gases. The boiling range of the
hydrocarbon products depends on the quantity of hydrogen. This is shown in Figure 99 for hydrogenation in the slurry phase. Hydrogen consumption per tonne of dry coal for the catalytic direct coal
hydrogenation amounts to ca. 1000 m3.

Figure 99. Relation of hydrogen consumption and product distribution during slurry-phase hydrogenation of coal

Hydrogen Production, Recycle Gas Separation. In all coal hydrogenation processes nonreacted hydrogen must be recovered and recycled back to the reaction together with fresh make-up hydrogen.
The economics of the projected large-scale processes for direct hydrogenation depend not only on the further improvement of the hydrogenation stage (milder hydrogenating conditions, improved liquid
product yields) but in particular on:

improved separation of the gaseous hydrogenation products with complete recycle of hydrogen with an acceptable purity (inert components decrease the partial pressure of the hydrogen) and
low-cost production of the make-up hydrogen using improved and integrated processes for steam reforming and partial oxidation of byproducts of the coal hydrogenation, which are otherwise

page 98 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
difficult to utilize.

Process Developments. After the oil embargo of 1973, many developments were revived. The developments being furthered in the Federal Republic of Germany are based to a great extent on the
modified IG process. Two plants on a pilot scale (6 t/d and 200 t/d) have been erected. In the United States, on the other hand, the process concepts being pursued involve new process steps, e.g.,
SRC 1, SRC 2 (Gulf Oil), Exxon Donor Solvent Process, H-coal process (Hydrocarbon Research). An overview is given in [352]. Improved two-stage processes are finding increasing attention, for
example in Japan [353].

Hydropyrolysis. In this process coal is subjected to high temperatures (700 – 1000 °C) for short residence times in the presence of hydrogen and thereby thermally cleaved and partially hydrogenated
(flash pyrolysis). The hydrogen consumption of 200 – 300 m3 per tonne of coal is relatively low and 25 – 45 % liquid hydrocarbons are obtained; about half of the feedstock coal is recovered in the form of
a low-sulfur residue coke [354].

Hydrogasification of Coal. In hydrogasification processes, coal is converted in the presence of hydrogen and/or steam in one stage or in a series of stages directly to methane. The aim of many process
developments is the production of SNG.

The basis is the strongly exothermic reaction of hydrogen with carbon according to (greatly simplified):

Equilibrium data are given in [350]. The equilibrium yield of methane is increased at high pressure and low temperature. Table 40 shows calculated methane concentrations at equilibrium. The reaction
sequence is a combination of the rapid hydrogenating gasification of the organic components in the coal and the slow heterogeneous gas – solid reaction between semicoke or coke residue and molecular
hydrogen. The reaction occurs relatively slowly in the temperature range of 700 – 1300 °C; the technically feasible operating conditions are nowadays considered to be between 800 and 900 °C and 4.0 –
10.0 MPa. The multiphase reaction must be conducted in a fluidized bed and, because of the short residence times resulting, gas recycle processes must be used.

Table 40. Methane concentrations of the carbon hydrogenation reaction at equilibrium

Temperature, °C Pressure, MPa CH4 content,

vol %

800 1 26
10 63
100 87
1000 1 8
10 37
100 72

Most important for the economics are the hydrogen production costs. The overall methane yield per kilogram of carbon depends in particular on the process used to produce hydrogen:

hydrogen from the coal feedstock, autothermal 0.6 m3 (STP) CH4

per kilogram of carbon
hydrogen from recycle methane, allothermal 1 m3 (STP) CH4
per kilogram of carbon
hydrogen from external sources, e.g., water electrolysis 1.8 m3 (STP) CH
4
per kilogram of carbon

Examples. The process developed by Union Rheinische Braunkohlenkraftstoff AG operates in a fluidized bed at 800 – 1000 °C and 5.5 – 9.5 MPa (data from the pilot plant operation). The crude gas
obtained contains up to 40 % methane; the degree of gasification is ca. 82 %. For hydrogen production the residue coke is gasified in a High-Temperature Winkler (HTW) gasifier [355].

Hygas process (Institute of Gas Technology, United States): Hydrogen is produced from the residue coke in the lower part of a multistage reactor.

page 99 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The Hydrane process (Bureau of Mines, United States) uses a two-stage gasifier with external hydrogen production from residue cokes, oxygen and steam.

8.1.4. Synthesis Gas

The process for the production of hydrogen from hydrocarbons yields gas mixtures containing mainly hydrogen and carbon monoxide (see Section Production from Coal and Hydrocarbons). Such gases
are named synthesis gases since they are used for the synthesis of special chemical products. Some hydrogen – carbon monoxide combinations (because of the production method or the final use) have
the designation water gas, cracked gas, methanol synthesis gas or oxo-synthesis gas. The fundamental reactions of synthesis gas chemistry are methanol synthesis, Fischer – Tropsch synthesis, oxo
synthesis (hydroformylation), and methane synthesis; further synthesis gas reactions are being developed. Excellent overviews are given in [356][357][358].

Methanol Synthesis. Methanol can be produced from hydrogen – carbon oxide mixtures by means of the catalytic reaction of carbon monoxide and some carbon dioxide with hydrogen. Methanol may be
considered virtually as synthesis gas at maximum compression.

The presence of a certain amount of carbon dioxide in the percentage range is necessary to optimize the reaction. Side reactions, also strongly exothermic, can lead to formation of byproducts such as
methane, higher alcohols, or dimethylether.

2520 m3 (STP) of synthesis gas (70 % H2, 21 % CO, 7 % CO2) are necessary to produce a tonne of methanol. The purity of the latter depends on the catalyst used. Copper – zinc catalysts used in the
low-pressure process require a sulfur-free gas (H2S <1 ppm [mL/m3]). Catalysts for the high- and medium-pressure processes (ZnO activated with chromic acid) can accept 30 ppm (mL/m3) of hydrogen
sulfide. For details on the process schemes, catalysts, process techniques, economics, etc. see, for example, [360].

Fischer – Tropsch Synthesis. Hydrogen – carbon monoxide mixtures react to hydrocarbons according to reaction equations (58) – (62), (60) is the main secondary reaction. An example for one of the
general equations for alkane synthesis is (63), for olefin synthesis (64). The overall reaction balance including synthesis gas formation corresponds to an indirect coal hydrogenation.

(58)

(59)

(60)

(61)

(62)

(63)

(64)

The Fischer – Tropsch synthesis is strongly exothermic. The total heat of reaction amounts to ca. 25 % of the heat of combustion of the synthesis gas. Undesirable side reactions include methanation, the
Boudouard reaction, coke deposition, oxidation of the catalyst, or carbide formation. Other competing reactions such as alcohol, aldehyde, ketone, carbonic acid, and ester formation, also occur.

For development of the process, kinetics, thermodynamics and mechanism of the reactions, distribution of the products, etc., see Coal Liquefaction and [359-362]. Up to 1954 4 000 publications and
4 020 patents regarding Fischer – Tropsch catalysts had already been reported [363], but virtually only those based on cobalt and iron have been successful.

The yield of the Fischer – Tropsch synthesis is determined by the initial hydrogen – carbon monoxide ratio and the stoichiometric requirements of the desired reaction products. The maximum yield is
obtained when the aforementioned are nearly equal and thus yield 208.5 g hydrocarbon from 1 m3 (STP) of synthesis gas; technically ratios of 1.7 : 1 to 3 : 1 are used.

The Fischer – Tropsch synthesis is used on a technical scale nowadays only in South Africa at SASOL. At SASOL I, five units of the ARGE reactor of the Ruhrchemie – Lurgi (gas phase, fixed bed
reactors, 2.3 – 2.5 MPa, 220 – 250 °C, 40 m3 Fe – Cu – K2O – SiO2 catalyst, 9 – 12 months running time) and three reactors using the Synthol circulating fluidized bed of Kellogg – SASOL (gas phase,
2.0 – 2.3 MPa, 300 –340 °C, 30 – 140 t Fe catalyst produced by reduction of magnetite; running time 45 d) are in operation. At SASOL II (start-up 1980) and SASOL III (start-up 1982) only the more
flexible Synthol process is used. The enormous gas amounts (2×1 650 000 m3 [STP]/h raw gas, 2×1 000 000 m3 [STP]/h feed gas) make the recovery of byproducts (ethylene, alcohols esters, carbonic
acids, etc.) economically worthwhile [364].

Variations of the Fischer – Tropsch synthesis with other selective catalysts can lead to the following products:

page 100 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
low-molecular-mass olefins (C2 – C4, mostly ethylene) [365], [366]
medium- and long-chain olefins [366]
polyethylene [367]

Polyethylene is obtained at high pressures (100 – 200 MPa), low temperature (100 – 120 °C), and using a ruthenium-based catalyst.

Methane Synthesis. The catalytic hydrogenation of carbon monoxide to methane is strongly exothermic

Consecutive and side reactions (shift conversion, Boudouard equilibrium, hydrogenation of carbon) make the calculation of equilibrium conditions very complex. The influence of the feed gas composition
and operating parameters on the equilibrium composition of the product gas and the kinetic data, which are somewhat contradictory, are summarized in [368].

Classical methanation catalysts are magnesium-promoted nickel catalysts with diatomaceous earth as carrier. Catalyst developments are aimed at increasing the thermal and mechanical stability and
yielding catalysts which are simultaneously suitable for the cracking of naphtha with subsequent methanation. Methanation is technically important as a catalytic purification step for the removal of carbon
monoxide from gases (see Section Catalytic Gas Purification).

Hydroformylation of Olefins. Hydroformylation, also called oxo synthesis or Roelen reaction (O. Roelen, 1938), is a commercial-scale process for the production of aldehydes. By catalytic addition of
hydrogen and carbon monoxide to an olefin an aldehyde is obtained under chain elongation

Hydroformylation is generally an exothermic, homogeneously catalyzed liquid-phase reaction of the olefin with hydrogen and carbon monoxide. As catalyst cobalt carbonyl hydride or, in some variations of
the oxo-synthesis, cobalt- or rhodium-phosphin complexes are used.

Propen is the olefin mostly used. The oxoproducts are converted to alcohols, carboxylic acids, aldol-condensation products, and primary amines. About 20 commercial processes are state of the art. An
excellent review is given in [369] (see Oxo Synthesis).

Homologation. Under the reaction conditions of the hydroformylation alcohols and aldehydes react with carbon monoxide – hydrogen under elongation of the chain by one CH2- unit

Homologation has been performed with a number of alcohols, the production of ethanol from methanol has been most intensively investigated. The homologation is not used industrially because of the
many side reactions which take place [370].

Synthesis Gas as Chemical Feedstock. Hydrogen – carbon monoxide mixtures, hydrogen alone, and their primary product methanol are important feedstocks for the chemical industry. Nowadays,
ethylene which is produced from propane, ethane, naphtha, or gas oil is the most important feedstock for the production of industrial organic chemicals in the chemical industry. Bascially, it is, however,
possible to obtain these compounds from synthesis gas thus changing the feedstock basis to coal (see Fig. 100).

Figure 100. Synthesis gas as feedstock in the chemical industry

8.1.5. Hydrogen in Organic Synthesis

Hydrogen is required for the production of chemicals and intermediates in organic chemistry. A large number of hydrogenations or reductions are carried out on a technical scale ( Hydrogenation and

page 101 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Dehydrogenation).

Activated and nonactivated double and triple bonds in olefins and acetylenes can be easily partially or totally hydrogenated, whereas the hydrogenation of aromatic and heterocyclic bonds requires more
energetic conditions. Functional groups, such as carbonyl, nitro, nitroso, and nitrile groups, can also be hydrogenated.

The reaction conditions are dictated by equilibrium (65). The reactions are exothermal and run in the presence of a catalyst.

(65)

Directions for carrying out catalytic hydrogenations on a laboratory or industrial scale are given in [371]. Summaries of hydrogenation reactions are given in [372] and [373]. Hydrogenation catalysts are
metals of groups 8 – 10 of the periodic system (see front matter of this volume), e.g., Raney nickel, as well as copper and molybdenum. In particular the noble metals (Pt, Pd), are highly-active catalysts
[374]. Homogeneous systems with molecularly dispersed catalysts in the solution, can be used for special synthesis problems (selective hydrogenation, asymmetric synthesis) but are at present of no
great importance in commercial areas because of the frequently encountered difficulty to remove the catalyst from the reaction mixture. Table 41 gives an overview of commercially used hydrogenation
reactions.

Table 41. Selection of important industrial hydrogenation reactions

page 102 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

High-purity hydrogen is necessary for the partial or total hydrogenation of fats and oils (for the production of edible fats or for technical purposes). In fat hydrogenation the polyene, triene, and diene fatty
acids in their glyceride form are selectively hydrogenated to the corresponding monoene acids.

The industrial production of sugar alcohols, such as sorbitol, xylitol or mannitol from the corresponding sugars is carried out by catalytic hydrogenation. Batch suspension processes using Raney nickel
catalysts are mainly employed under reaction conditions of 120 – 150 °C and 3 – 7 MPa [375].

8.1.6. Hydrogen in Inorganic Synthesis

The catalytic hydrogenation of anthraquinone and its derivatives followed by their auto-oxidation to yield hydrogen peroxide is the basis of the commercially important process for hydrogen peroxide
production ( Hydrogen Peroxide – Anthraquinone Process (AO Process)). Further important reactions in inorganic chemistry are the production of hydrochloric acid from hydrogen and chlorine ( )
and the hydroxylamine synthesis ( Hydroxylamine).

8.2. Hydrogen in Metallurgy

Iron Metallurgy. To reduce iron ore, apart from coke (classical blast furnace process), other reducing agents can be used. For reduction a gas containing hydrogen, carbon monoxide, or mixtures of these
is suitable. The reduction gas is produced by steam reforming or partial oxidation of fossil fuels. These “direct reduction” processes ( Iron – Production) yield sponge iron, which can be melted to give
crude iron which is further processed to steel.

The leading direct reduction technologies are the Midrex, the HyL I, and the HyL III process with 90 % of the total capacity [376]. The hydrogen content of the reducing gas is ca. 40 – 65 vol % (Midrex,
shaft furnace) and 75 vol % (HyL III, retorts). To fully utilize the reduction potential of the gas, carbon dioxide and water vapor are removed and the gas is recycled.

The use of pure hydrogen has advantages with respect to the reaction time, the degree of reduction and the texture of the reduced pellets [377], but the carbidizing reaction necessary for steel production
cannot take place, so that reduction with pure hydrogen has not been able to establish itself.

Nonferrous Metallurgy. Hydrogen is employed as reducing agent and as utility in some powder metallurgy production processes. Table 42 shows the use of hydrogen during the production and handling

page 103 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
of various nonferrous metals.

Table 42. Use of hydrogen in the nonferrous metallurgy

Metal Unit operation Product

Copper reduction of copper salt solutions under pressure Cu powder

Nickel selective reduction during cobalt production Ni powder
Cobalt reduction of aqueous cobalt salt solutions under pressure (4 MPa, 175 °C) Co powder
Molybdenum, tungsten reduction of the oxides or molybdates and tungstenates Mo, W powder
Tantalum reduction of tantalum chloride, TaCl5, in hydrogen plasma Ta hydride, Ta powder
Germanium reduction of germanium tetroxide, GeO4, at 650 °C Ge powder for further processing in zone melting
Uranium reduction of the higher uranium oxides at 650 °C UO2

For recovery of copper from its sulfidic ores reduction with hydrogen in the presence of calcium oxide has been suggested [378]. The thermodynamically unfavorable position of the hydrogen reduction
reaction on metal sulfides is improved by the removal of the developing hydrogen sulfide (as CaS) from the equilibrium mixture.

8.3. Other Uses

Use of the High Temperature of the Oxyhydrogen Flame. The combustion of a stoichiometrical hydrogen – oxygen mixture leads to flame temperatures in the range of 3000 – 3500 K. Such flames can
be used for:

Cutting and autogenous welding in the metal industry.

Growing synthetic crystals (saphires, rubies) by melting aluminum oxide doped with the appropiate metal oxides.
Production and handling of quartz glass (melting of rock crystals).

A process for the production of olefins from hydrocarbons is described [379]. The hydrocarbon to be cracked is atomized and fed into a 1000 – 2000 °C hot gas stream of hydrogen and steam, which is
produced in a prereactor by combustion of excess hydrogen in oxygen. High olefin yields are reported to be possible even from heavy hydrocarbons.

Hydrogen Plasma as a Heating Agent. A further temperature increase by using the free energy of the hydrogen combustion reaction is hindered by strongly endothermic processes, such as the onset of
molecule dissociation and ionization of the atoms. Higher temperatures can be reached in a plasma gas produced by means of an electric arc or a high-frequency field. Hydrogen dissociates around
5000 K into hydrogen-atoms, which start to ionize to H+ at 9000 K; the recombination of ions to atoms or of atoms to molecules produces a great amount of heat. The heat content of hydrogen plasma is
used in some processes for the production of acetylene from hydrocarbons (see Section Production from Coal and Hydrocarbons; Acetylene – Production from Coal (Arc Coal Process)). Hydrogen
plasma has advantages compared to ordinary electric arc pyrolysis, for example, much lower carbon formation.

Hydrogen in Metal Processing. The high temperature of the plasma stream is used for plasma welding and cutting. The addition of hydrogen to plasma argon (2 – 15 vol %) increases the temperature
and the cutting velocity. This process is used principally for cutting or welding high-alloy steels. Forming gas, a gas mixture of 8 – 20 % hydrogen and nitrogen, prevents the oxidation at the far side of the
weld (root cut), e.g., when pipes are being welded.

In the thermal atomization technique for the production of surface layers, a plasma jet is used to spray high-temperature melting metals, their oxides, borides, etc. As plasma gas hydrogen, hydrogen –
nitrogen or hydrogen – argon mixtures can be used.

Heat treatment of metals (annealing, sintering, melting) requires the presence of an inert atmosphere. A mixture of nitrogen with hydrogen is usually used. Hydrogen alone can decrease the annealing time
by 30 % because of its higher heat conductivity and the shorter reaction times for the reduction of oxide layers [380]. Mixtures of nitrogen, hydrocarbons, and hydrogen are used for nitriding and
carbonizing annealing.

During the production of plate glass by the float glass process, a protective nitrogen – hydrogen atmosphere is employed to keep the metal melt surface free of oxides.

Semiconductor Technology. For the production of doped semiconductors, traces of the elements in the form of the corresponding hydrides (SiH4, AsH3, GeH4) are mixed with a high-purity hydrogen
stream in the required concentration and deposited by thermal decomposition on a silicon carrier.

Water Treatment. In potable water, unduly high nitrate contents can be lowered by denitrification with hydrogen-oxidizing bacteria in bioreactors. Hydrogen, dissolved at 0.4 – 0.6 MPa, is used as an

page 104 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
energy supply for the biomass [381]. For the treatment of boiler feed and process water, dissolved oxygen can be removed by conversion with hydrogen in the presence of palladium-doped exchange
resins.

Other Uses. Hydrogen is also used for the following purposes: as alloying element in metals to produce required chemical properties, and to increase the transition temperature of superconducting alloys,
as well as carrier and fuel gas in gas chromatography (flame ionization detector). Liquid hydrogen is used as a refrigerant, e.g., in bubble chambers [382] and for cooling of superconducting metals below
the transition temperatures.

[Top of Page]

9. Hydrogen Energy
Rev. WILLIAM CECIL of Cambridge, England, is the first person reported to have suggested using hydrogen in an engine in 1820. In 1874, JULES VERNE described in one of his last books “the fuel to be
used, once coal, ran out . . . – Water . . . but decomposed into its primitive elements . . . and decomposed, doubtless, by electricity.” Unfortunately, he did not disclose any details on the nature of the
primary energy source to split the water.

In the 1920s and the 1930s, a great number of scientists, especially J.B.S. HALDANE in Scotland and R. ERREN in Germany, described virtually every method of production, storage and application
mentioned so far in today's hydrogen energy concept, e.g., electricity generation using wind turbines, electrolysis, hydrogen storage in liquid form, and utilization in fuel cells and aircrafts. More on the story
of hydrogen energy in general can be found in [383].

9.1. Aims of the Hydrogen Energy Economy

Hydrogen is an energy carrier, an indirect source of energy, not a resource itself. In spite of this, hydrogen was at first foreseen as a substitute energy form, in particular to substitute coal, natural gas, oil,
and any products derived from them.

The reserves of fossil energy, in particular oil, are limited. This led to the belief that hydrogen could soon be used economically as a substitute energy form. The economic viability of the energy carrier
hydrogen is enhanced by price advantage in transport over large distances (see Section Hydrogen Energy Economics) and the possibility of energy storage (see Section Hydrogen-Energy Conversion
Systems).

According to Figure 14, hydrogen for use as a universal energy carrier can be generated by using all possible primary energy sources. Coal, nuclear energy including fusion (although still undeveloped),
and solar energy have been named.

In addition to economic reasons for the introduction of new energy sources and new energy carriers, ecological reasons are becoming increasingly important, for example:

The increasing carbon dioxide content of the atmosphere, caused by the use of fossil energy sources (Fig. 101) and other human activities such as destruction of the tropical rain forests are predicted to
lead to extensive climatic changes, e.g., a slow warming up of the atmosphere and the oceans, and melting of pole caps [384], [386]. Because of the natural fluctuations in the climate this theory has not
been proven beyond doubt. In particular, little is known about compensation effects.

Figure 101. Actual and projected increase of the CO2 content of the atmosphere, caused by human action, especially burning of hydrocarbons, coal, and wood

A) Measurements in Hawaii [385] from 1959 to 1983; B) Some projections extrapolating two scenarios of energy use

a) Maximum use; b) Optimum conservation of natural resources [384]

Well-known scientists have been demanding for years (without success) that a “low-risk strategy” for the use of fossil resources should be followed [387]. The utilization of the energy carriers coal, oil (in
the form of gasoline, diesel, kerosene, etc.) and natural gas leads to emissions of sulfur dioxide, nitrogen oxides and hydrocarbons. In Europe, this has already led to the destruction of extensive wooded
areas and to damage of the rest of the forests. The use of hydrogen as an energy carrier coupled with suitable utilization techniques could become particularly important for environmental protection.

Hydrogen competes with the conventional energy carriers hydrocarbons (methane, LPG, gasoline, etc.), coal, and electric power.

Further energy carriers are those recovered from regenerable sources and from refuse: biogas, alcohols, and vegetable oils.

For the suitability of an energy carrier, the following aspects are important:

conversion efficiency of the primary energy carrier to the final use,

page 105 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
availability,
ease of storage,
safety, and
ecological and economic evaluation

The evaluation of the energy carriers for specific applications is given in Section Hydrogen-Energy Conversion Systems, the safety aspect is treated in Section Safety Aspects.

9.2. Production Systems for Hydrogen Energy

For use as a chemical reactant hydrogen is produced almost exclusively from hydrocarbons (see Table 11), because integration into the production processes and resulting overall plant economics are
determining. A future hydrogen energy concept should be based on forms of energy which influence the environment as little as possible. It should offer overall economic advantages and should be
socially acceptable. This is the case with solar energy, used (1) directly in photovoltaic or solarthermic plants or in direct hydrogen production, (2) climatically induced energy gradients, e.g., wind, waves,
heat, (3) indirectly via production of biomass (carbohydrates, plant oil, ethanol, etc.). As an intermediate solution for a limited transition period a combined utilization of fossil primary energy, nuclear energy
and regenerable energy forms have been suggested.

Figure 14 gives a review of the principal possibilities of generating hydrogen from the primary energy sources solar energy, fossil fuels, nuclear and geothermal energy. In hydrogen energy scenarios, the
economical and ecological aspects of the conversion of primary energy to the required energy form heat, power, light, etc., using hydrogen as energy carrier is investigated from the point of view of
present technologies.

9.2.1. Energy Sources

Direct Solar Energy. Solar radiation can be directly used to produce hydrogen (see Section Other Methods for the Cleavage of Water) by means of photochemical, photobiological or
photoelectrochemical effects.

These methods, however, have up to now been only able to produce small amounts on a laboratory scale. The catalysts and the biomass tested have a short lifespan. Developments, especially in the
area of basic research, are being followed up.

Thus, the short-term realization of projected hydrogen production systems is based on the indirect utilization of solar energy using photovoltaic cells ( Photovoltaic Cells) with which electricity or, more
exactly, electric voltage is produced. In addition, solar-thermal power stations, in which utilizable energy in the form of heat is developed, will become operable. In particular, the photovoltaic technology is
already intensively utilized. Not only plants for power production, but also pilot hydrogen producing systems are already in operation or under construction.

Utilization of solar energy for photovoltaic power production is possible all over the world. The energy utilization factor in countries with high sunshine intensity (e.g., Sahara) is ca. 2.5 times higher than in
the temperate climate zones (mid Europe) [388].

Those areas which are considered to be very good sites have a radiation energy > 2300 kWh/m2 per year. If not only the high sunshine intensity, but also the geological suitability for setting up solar cell
modules is considered, then ca. 6×105 km2 are available worldwide; this is less than 0.5 % of the surface area of the globe. This area alone would be sufficient at a hydrogen production equivalent to
7000 tce/km2 (tce = tons of coal equivalent, 1 tce 29.31×109 J) per year, to cover the world energy consumption [389]. But also other concepts which do not concentrate as much on the areas with high
radiation intensity, but on installation of solar cells close to the user on free surfaces, e.g., the roofs of houses, are said to be sufficient for hydrogen production [390].

Indirect Solar Energy. Solar radiation leads indirectly to energy potentials on the earth; these have been used by mankind for thousands of years to produce energy.

Water power is utilized at sites distant from private industrial consumers to produce hydrogen for ammonia production (Norway, Egypt, India, etc.). Hydrogen is produced by means of alkaline electrolysis
(see Section Principles). An overview of the worldwide utilization of water power is given in [391] and, with respect to the utilization for hydrogen production, in [392]. According to this, the potential
hydroenergy of 14 700 TWh/a worldwide is utilized at present to ca. 15 %. However, the greatest unused potential lies in zones far away from population and industry, such as the Arctic, Greenland,
Alaska, Sibiria, or in the equator zones (Congo) [393]. About 30 % of the additional potential could be used for hydrogen production [392]. On the other hand, too great a utilization of the hydroenergy
potential meets with ecological disapproval because of the destruction of the landscape by dam-building, flooding of large areas, destruction of forrests, etc.

Wind energy is used on islands to satisfy local energy requirements and to feed power into the grid. Pilot plants for the production of hydrogen already exist (see Section Future Developments). The
coastal regions are particularly suitable for the generation of wind energy. Areas with particularly high wind speeds and low energy requirements which could be used for overregional production of
hydrogen energy are Patagonia (South America), Somalia (Africa), and the Southwest Australian coast.

Biomass, Refuse, Wastewater. Hydrogen can be recovered from these substances by various methods.

The direct production of hydrogen by means of bacteria is possible; yields of up to 20 % have been attained (see Section Hydrogen Formation in Biological Systems). Other possibilities are the biological
production of alcohols or methane from biomass or refuse by fermentation or other anaerobic processes. These substances can be converted by the methods described in Section Production from Coal
and Hydrocarbons into hydrogen. Synthesis gases or hydrogen can be produced from biomass and refuse by pyrolysis or gasification processes analogous to those described for fossil feedstocks (see
Section Production from Coal and Hydrocarbons).

page 106 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The hydrogen production potential from biomass, refuse, or wastewater is not considered to be very high for various reasons. Firstly, other valuable products and energy carriers, such as edible products,
alcohol and methane, can be produced directly from the substances. Secondly, these sources are usually widely distributed, so that for a central industrial hydrogen economy the costs for transport and
logistics would be very high.

Nuclear Energy and Other Energy Forms. Nuclear energy has also been suggested for the production of hydrogen [394], [395]. Various types of conventional and breeding reactors can supply power
for hydrogen production, whereas the high-temperature reactor supplies heat for chemical reactions. The direct production of hydrogen by means of radiolysis (see Section Thermolytic and Radiolytic
Processes) is an undesired accompanying phenomenon in pressure and boiling water reactors.

Other energy forms, such as geothermal or tidal energy, play a marginal role in the energy production. They need not be considered for the hydrogen energy concept.

A combined utilization of nuclear energy and fossil, nonconventional resources has been sugggested for the tar sands in Canada, for oil shale deposits in the United States, and for the heavy oil found in
Venezuela. Hydrogen and steam are thereby cogenerated. Steam is used for the recovery of hydrocarbons and for processing, hydrogen is used for the hydrogenation of heavy oil.

9.2.2. Production Methods for the Energy Carrier Hydrogen

The conventional methods described in Sections Production from Coal and Hydrocarbons and Electrolysis as well as Chapter Purification of Hydrogen are available for the purification and production of
hydrogen for the hydrogen energy scenario.

Purification Methods. The purification requirements for the hydrogen energy economy depend not so much on the planned utilization as on the hydrogen storage technique and transport (vectorization).
The highest purity is necessary if hydrogen is liquefied, because all impurities, apart from helium, are deposited as solids and could block the liquefication equipment.

Storage and transportation in metal hydrides also place high requirements on the hydrogen purity, because most impurities lead to irreversible destruction of the storage material. Purity requirements are
less for the direct combustion of hydrogen and the reaction with oxygen or air in fuel cells. All purity requirements can be met using conventional technologies, particularly adsorption (see Section
Adsorption Processes).

Electrolysis. The most important hydrogen production method for the hydrogen energy scenario is electrolysis (see Section Electrolysis). Here, the energy carrier electricity is converted to the energy
carrier hydrogen. Conventional electrolyses have a conversion efficiency of 80 – 85 %. Improved electrolyses are expected to have an efficiency of 88 – 94 % [396], which can hardly be increased further.
In the high-temperature electrolyses at present under development, however, part of the energy necessary for the electrolysis can be supplied as thermal energy and it can therefore produce hydrogen
with a higher efficiency than with solely an electrical supply.

The combination of regenerable energy forms, such as photovoltaic or wind energy with electrolysis units has undergone a long-term test in pilot projects (see Section Future Developments). In addition,
detailed system analyses have been made [397]:

electrolysis and thermal solar power plants (power generator can be driven by steam or a Stirling motor),
electrolysis and thermal solar power plants, coupled with high-temperature electrolysis [398],
electrolysis and wind energy,
electrolysis and photovoltaic plants.

Particular problems with the electrolysis are caused by the fluctuating operation of the solar plant cycle at various times of the day as well as weather-dependent fluctuations of solar and wind energy.

Typical problems caused by rapid load changes:

degradation and destruction of diaphragm because of local overload,

corrosion caused by current leakages, and
deactivation of the electrode catalysts

Problems caused by slower (daily) changes:

cooling of the electrolysis module,

mixing of gas and electrolyte,
pressure equalization between cathode and anode, and
start-up and shut-down procedures including eventual prepolarization or blanketing.

Additional Methods of Hydrogen Production. Additional methods are:

1. Hydrogen production by means of thermochemical cycle processes using thermal energy (solar or nuclear energy)
2. Hydrogen production by gasification of refuse or biomass (for process principles see coal or heavy oil gasification, Section Gasification of Coal and Hydrocarbons)

page 107 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3. Hydrogen production by pyrolysis of refuse or biomass (for gas purification see coke oven gas treatment, Section Coke Oven Gas ).

For technical and economic reasons the aforementioned methods play only a small part in the production of hydrogen for energy systems.

9.3. Hydrogen-Energy Conversion Systems

For utilizing the energy contained in the secondary energy carrier hydrogen, numerous techniques are available, in the area of energy technology in industry and household, in power stations and in the
transportation sector. Most of these are based on principles which have been known for a long time. Section Hydrogen-Energy Conversion Systems gives an overview of the state of the art and of present
developments world-wide.

9.3.1. New Developments for Hydrogen Storage

The possibility to store hydrogen amounts to the possibility to store energy; this is the particular advantage of a secondary energy carrier. In addition to storage in the gaseous or liquid state (Section
Storage) there are numerous investigations or experimental applications of new processes for hydrogen storage in solids.

9.3.1.1. Cryoadsorption
Great quantities of gaseous hydrogen can be adsorbed on suitable adsorbents at medium pressure and low temperature. If the adsorber is cooled to 60 – 100 K and pressurized to ca. 4 MPa, the average
hydrogen density on the active surface increases to values comparable to the density of liquid hydrogen. The effective storage density is, however, much smaller because part of the volume is taken up by
the adsorbent and the pressure vessel; however, liquefaction can be avoided.

The process has been investigated experimentally [399], [400]. Only materials which are light and very porous, with a high specific surface or with a specifically activated surface can be used as adsorbing
agents. Operating data for a cryoadsorption storage system on the basis of activated carbon bulk density 0.35 g/cm3 are listed below:

Temperature range for isothermal 65 – 70 K

operation
Charge-discharge pressures 0.2 – 4.2 MPa
Storage density of hydrogen 68 g H2 per

(75 K and 4.2 MPa) kilogram adsorbent

Maximum hydrogen storage 82 g H2 per
density (65 K and 4.2 MPa) kilogram adsorbent
Average mass related storage 50 g H2/kg
capacity
Average volume related storage 15 g H2/L
capacity

The economic range for the technical application will presumably be limited to medium-scale storage with a specific storage capacity of 20 – 30 kWh/kW.

9.3.1.2. Hydride Technology

The capability of metals or metal alloys to reversibly store hydrogen in the form of hydrides has lead to a great number of investigations with respect to technical applications. The main areas are seen in
the storage of hydrogen but also in the separation of hydrogen from gas mixtures, the purification of gases, and in storage and separation of hydrogen isotopes. An overview is given [401][402][403][404].

Physical Fundamentals. Ionic and metallic hydrides are formed (exothermic process) and decomposed (endothermic process) according to

The hydrogen absorption and hydride formation behavior can be described by pressure – composition isotherms (see Fig. 102 A) as follows. After catalytic dissociation of the hydrogen molecule some of
the atoms lose their electron and occupy an interstitial position in the crystal structure of the host metal (solid solution, -phase). As the hydrogen concentration is increased, local H – H interaction causes
nucleation which induces growth of the hydride phase ( -phase). The two phases ( + ) coexist until, through further hydrogen uptake, the stoichiometric limit is reached. Here, the isotherms reach a
plateau level (constant pressure) which is given the name ‘dissociation plateau' or ‘plateau pressure'. After the pure -phase is reached, any further increase of the hydrogen : metal ratio requires a steep
increase in hydrogen pressure. At higher hydrogen pressure further plateaus may be found because of the existence of other hydride phases ( , , . . .).

page 108 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 102. A) Ideal pressure-composition isotherms for a hydrogen-metal system

B) Absorption and dissociation isotherms of LaNi5 hydride with hysteresis

C) Hysteresis loops of test cycles of FeTi at 25 °C. The plateau pressure of the monohydride stabilizes after ca. 20 cycles, whereas that of the dihydride formation is still rising

D) Pressure-composition isotherms of hydrogen and deuterium in zone-refined vanadium at 40 °C

On reducing the hydrogen pressure the process is run through in the reverse direction (desorption isotherm). Along the desorption plateau, the hydride phase decomposes into hydrogen-saturated metal
and hydrogen gas. The decomposition process is endothermic. Absorption and desorption isotherms show a hysteresis effect. This is shown in Figure 102 B and C for LaNi5 and FeTi [405], [406].

The plateau pressure p, which varies strongly with temperature, is connected with the formation or decomposition enthalpy by the Van't Hoff equation. The entropy term, S 0, consists mainly of the
standard entropy of gaseous hydrogen (S 0[ ] = 130.6 J mol–1 K–1), because the entropy of the hydrogen dissolved in the metal can be neglected.

This relationship can be illustrated graphically for all hydrides and for the corresponding hydride transition steps (hydride isochores) (see Fig. 103). Applications can be found in a temperature range from –
30 – 300 °C and at pressures up to 10 MPa.

Figure 103. Hydride isochores for various metals and intermetallics. H2-plateau dissociation pressure vs. 1/T

Mm denotes mischmetal, an inexpensive mixture of rare earths

The heat evolved by the hydrogen – metal reaction depends on the type of hydride. It can amount to ca. 30 % of the heat of combustion of the absorbed hydrogen. To liberate hydrogen, at least the same
amount of energy is required (see Table 43).

Table 43. Comparison of the hydrogen absorption data of some metal – metal hydride systems

Hydride system Heat of reaction, kJ per mole H2 Plateau region a Hydrogen content Energy density,c

pH2, MPa T, °C wt % g/mL b kJ/g

Elemental metals
Mg – MgH2 77.4 0.1 286 7.7 0.107 9.9

Ti – TiH2 124 0.1 650 4.0 0.15 5.7

V – VH2 0.4 40 2.1 0.19 2.95

Alloys
Mg2Ni – Mg2NiH4 64.5 1.0 350 3.16 0.081 4.5

FeTi – FeTiH1.95 31.4 0.5 30 1.72 0.096 2.5

page 109 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
LaNi5– LaNi2H5.9 30.2 0.5 60 1.37 0.089 1.95

Ti0.98Zr0.02V0.43Fe0.09Cr0.05Mn1.5 1.0 24 1.75 0.09 2.5

Liquid hydrogen for comparison
LH2 100 0.077 142

a At selected conditions, dissociation pressure at 50 % of H/Me ratio.

b Refers to volume of compact hydride material without voids.
c Refers to HHV of absorbed hydrogen

Intrinsic Properties of the Hydride-Forming Materials. To analyze the utilization possibilities of hydride-forming metals, some further properties, some of which are crucial, must be known. The overall
kinetics of hydride formation, e.g., is governed by the intrinsic kinetics (dissociation, surface penetration, diffusion, phase transformation), heat flow, and mass flow. In particular, surfaces of intermetallic
compounds containing transition metals accelerate the dissociation reaction, whereas thin oxide layers on the surfaces of elemental metals often block this. Easy hydrogen sorption is facilitated by surface
segregation with continuous formation of reactive subsurfaces (self-activation). Addition of small quantities of transition elements, e.g., nickel, will always favor activation. Impurities in the hydrogen, such
as carbon monoxide, hydrogen sulfide and sulfur dioxide slow down the sorption rate because of surface deactivation, whereas oxidizing impurities, such as oxygen and water, gradually decrease the
capacity by reaction with the host compound. Such “poisoning” of the hydride material can in many cases by reversed by hydrogenation – dehydrogenation cycles with pure hydrogen at higher
temperature.

Diffusion is normally not the rate-determining step because most of the hydride-forming materials disintegrate into powder with a grain size down to 1 µm.

Incorporation of the hydrogen into the lattice and the / -phase transition causes a deformation of the metal structure. The accompanying increase in volume can amount to 25 % and may lead to
undesired distortion of the containing vessel.

Detailed analyses of the kinetics of hydride formation of LaNi5 and MgNi are available [407]. For most hydride-forming systems using intermetallic compounds, heat transfer is rate-controlling. Sufficient
heat transfer can be provided by adding inert metals (Cu, Al) or inert ceramic beads or by pressing the material into pellets to fit into heat exchanger tubes.

Hydride-Forming Materials. The criteria for selection of a material for a specific application are:

high adsorption capacity

suitable temperature and pressure range
sufficient exchange rate
long-term stability
long cycle life

In the temperature range from –20 – 400 °C the following groups of alloys are known:

– 20 – 100 °C
AB2 alloys, e.g., ZrMn2
AB5 alloys, e.g., LaNi5
FeTi-based alloys
300 – 400 °C
Mg and Mg-based alloys

Some metal – metal hydride systems and their important properties are summarized in Table 44. In recent years development has concentrated on cheaper multicomponent alloys and modeling materials
to optimize uptake and storage parameters. About 10 different storage alloys are commercially available from producers in Europe, Japan and the United States at prices between $ 5 and 30 per kilogram.

Table 44. Hydrogen content of fuel gases and corresponding energy content

Fuel gases H2 content, % HHV, kJ/m3 (STP) LHV, kJ/m3 (STP)

Natural gas, rich gas up to 12.5 25 000 – 39 800 35 800 (CH4)

page 110 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Town gas 40 – 60 16 700 – 20 000
Coke oven gas 50 – 60 20 000 – 23 000 16 500 – 20 500
Water gas 50 10 500 – 12 200 10 700 – 11 700
Producer gas, blast furnace gas 2 – 5 4 300 – 4600 3 900 – 4700
Carbon monoxide 12 650 12 650
Hydrogen 100 12 760 10 800

Stationary Hydride Storage. For the stationary storage of hydrogen by using hydrides, FeTi or TiVFeMn alloys can be employed. A representative example for the many developments in this field is the
hydride storage vessel of Gesellschaft für Hydrid- und Wasserstofftechnik (HWT). Figure 104 shows a module which consists of a bundle of pressure vessels containing the powdered metal alloy between
fins. A sintered metal filter prevents the loss of powder. The gas bundle is surrounded by the heat transfer medium, water. The essential operating data are:

hydrogen capacity 1700 – 2000 m3 (STP)

active storage material 10 t
operating pressure 5 MPa

operating temperature 80 °C

Figure 104. Hydride storage tank

a) Storage tube; b) H2-guide tube; c) Lamella; d) Hydride; e) Filter (sintered metal); f) Gas-collecting tube; g) H2 connection; h) Water inlet (module); i) Water outlet (module); j) Filling body; k) Shell; l) Water inlet (centr.); m)
H2 inlet/outlet (centr.); n) Container

Hydrogen used to fill the vessel should have at least 3.0 (99.9 %) purity. Hydrogen with a purity of 6.0 (max 1 ppm [mL/m3] impurity) can be drawn off [408].

Mobile Hydride Storage. Possible applications for the hydride storage technology are mobile storage vessels for the laboratory or for hydrogen-fueled cars (see Section Hydrogen Fuel Systems). Figure
105 shows a hydrogen laboratory storage vessel and the corresponding ab- and desorption curves. These vessels are available for hydrogen quantities of 1 – 5 m3 [409]. In addition to the small space
requirements, they offer further advantages such as safe handling, purification effect and availability at higher pressure levels.

Figure 105. Hydrogen laboratory storage vessel and the pertaining adsorption and desorption curves

A) Schematic of a storage vessel; B) Loading at 0.5 MPa, water cooling at 20 °C; C) Unloading against 0.1 MPa, heat exchange medium is water at 80 °C

The energy density of hydride storage systems, including the vessel material, is in the range of 300 Wh/kg for the so-called low-temperature hydrides.

Separation of Hydrogen Isotopes via Hydride Formation. Some metals and alloys (TiNi, Ti2Ni, V) show considerable differences in the plateau pressure of their hydrides and deuterides (see
Fig. 102 D). This can be used for the enrichment of deuterium in deuterium-containing hydrogen streams [410]. The recombination of H and D atoms can lead to the formation of HD. The separation of
tritium over vanadium monohydride in a temperature-swing process has been described [411].

Purification of Gases. Hydride-forming materials can be used to purify hydrogen as well as noble gases. Various mechanisms are thereby involved: oxygen, carbon dioxide, and water are removed by

page 111 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
the hydride material because of its getter properties. Carbon monoxide is converted to methane, which is more easily removed, whereas inert gas components, such as the noble gases, can be removed
by a short purging step. The hydrogen, which can then be recovered from the hydride bed, has a purity of >7.0, i.e., impurities content <0.1 ppm (mL/m3) (see Section Quality Specifications).

Hydrogen Recovery. The selective removal of hydrogen by means of hydride-forming metals can be used to separate hydrogen from a gas mixture. The feasibility of this method for recovery of hydrogen
from the purge gas stream of an ammonia synthesis loop was demonstrated in a pilot plant. The unit operated with three absorber columns and produced hydrogen with a purity of 99 % and a recovery
rate of 90 – 93 % of the hydrogen contained in the purge gas [412]. The process is adiabatic, the heat evolved during absorption is stored in the LaNi5 hydride pellets, which contain a certain amount of
ballast material. This stored heat is used for hydrogen desorption. The hydrogen uptake and liberation is pressure-controlled (adiabatic pressure-swing process). The complex dependency of the hydride
bed lifetime on alloy type, trace components in the gas, poisons, and various combinations thereof has prevented commercial utilization until now [413].

Thermochemical Machines. Metal hydrides are promising materials for use in thermochemical machines such as heat pumps, heat transformers, refrigerators or hydrogen compressors. The principle is
based on thermodynamic cycles in a closed system; a heat pump cycle is shown as an example in Figure 106. The equilibrium lines in the ln p – 1/T diagram are based on measurements of pressure –
composition isotherms. The required heat output is assumed to be at 40 °C. Taking into account the different material imperfections (i.e., hysteresis, plateau slope) and the pressure drop necessary to
drive the reaction, the level of the low-temperature heat source changes from 2.5 °C (ideal case) to 16.5 °C, while the level of the high-temperature source rises from 82 °C to 103 °C. An advantage is that
hydrogen is handled in a closed loop without the risk of contamination. More than 10 developments are being followed up worldwide (United States, Federal Republic of Germany, Japan). The planned
applications are in household heating, air-conditioning, refrigeration and power generation. The investigations concern reaction bed designs, flow of the heating or cooling media, heat transfer, etc., using
fast-reacting alloys of LaNi5, LaNi4.9Al0.1, FeTi for the low-temperature range and CaNi5, Mg2Ni, LaNi4.7Al0.3 for the high-temperature range [414].

Figure 106. Thermodynamic cycle of a heat pump

a, b) Mean values of equilibrium lines for absorption and desorption

Dashed lines: pressures at end of plateau

Vertical arrows: inflow or outflow of heat

Horizontal arrows: transfer of hydrogen

(LmNi4.8 is a lanthanum-rich mischmetal)

Compression of Hydrogen. Hydrogen can be compressed thermally by absorption at low and desorption at higher temperature. A machine working on this principle as an open system with several
stages is commercially available [415].

9.3.1.3. Liquid Organic Hydrogen Carriers

With the help of hydrogenation – dehydrogenation reactions it is possible to store hydrogen reversibly in a chemical compound. The combinations benzene – cyclohexane and toluene –methylcyclohexane
have been investigated using noble metal – aluminum oxide carrier catalysts for the reversible reaction. Transportable hydrogenation – dehydrogenation systems operating at 380 – 500 °C and 1.1 MPa
have been designed for a truck using hydrogen fuel; the necessary energy for the exothermic dehydrogenation reaction is taken from the exhaust gas of the engine. A specific energy density of 56 g
hydrogen per liter liquid hydrocarbon is attained, but the weight of the necessary equipment must be taken into account [416].

The reversible uptake of hydrogen in liquid organic substances has also been suggested for transcontinental hydrogen transport. Existing facilities for gasoline could be used [417].

9.3.2. Combustion and Heating

The energy contained in the hydrogen can be released most simply by combustion with oxygen or with air. Characteristic combustion data are given in Chapter Properties.

Hydrogen Content of Fuel Gases. Hydrogen mixtures have long been used in the form of town gas for heating purposes. Table 44 shows the hydrogen content of some fuel gases. The difference
between LHV and HHV (condensation enthalpy of water formed during combustion) is greatest for pure hydrogen, because water is formed as the only product.

Emissions. Since only water is obtained as a product of the combustion, emissions, such as carbon monoxide, sulfur dioxide, incompletely combusted hydrocarbons, particles, soot, ash, which are
obtained by combustion of fossil fuels, are avoided (see Section Production Systems for Hydrogen Energy for a discussion of the carbon dioxide problem). If hydrogen is burned in the presence of air, the
high temperatures favor the formation of nitrogen oxides. The NOx formation depends on the flame temperature and can be influenced by the value of excess air, . Figure 107 shows typical NO
concentrations for a hydrogen – air flame. Maximum NO formation occurs in the range of the stoichiometric hydrogen – oxygen ratio, whereas under lean conditions (excess air) low nitric oxide
concentrations may be attained [418].

page 112 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 107. NO concentration resulting from hydrogen – air flames as a function of excess air value

([NO]e = equilibrium concentration; [NO]5 = concentration behind flame front within 5 ms; p = 3.2 MPa)

Hydrogen Burners. Differences in combustion properties, i.e., flame speed and flame stability as well as density and heating value, make it necessary to use a different burner geometry when hydrogen
or hydrogen-rich gases are to be burned. The flame speed of hydrogen – air mixtures is seven times higher than that of methane; the particular problem is therefore the flash-back over a wide range of
hydrogen – primary air mixtures.

The most important criteria for household burners are summarized, e.g., in [419]. The ratio of secondary to primary air must be increased, if hydrogen is used. Forced draft burners for heating systems can
be retrofitted by removal of the baffle plate and by shortening the ignition electrodes [420]. Industrial hydrogen burners are used for heat, steam, and electrical power production. The state of the art are
premix burners or nozzle-mix burners with high mixture velocities. For supplying heat in certain industries (semiconductors, glass) the simultaneous presence of a reducing atmosphere is advantageous.

Catalytic Combustion. The thermal energy of hydrogen can also be released by means of flameless catalytic combustion. Suitable catalysts, apart from platinum and palladium which initiate reaction at
room temperatures, are metal oxides, such as MnO2 – CuO – Ag2O (Hopcalite) or Co3O4. The surface can be enlarged accordingly by a suitable carrier material in the form of porous plates or
honeycombs. Catalytic burners operate either on the diffusion principle or by premixing hydrogen with air. Figure 108 shows a catalytic diffusion burner. An overview of Japanese investigations is given
[421].

Figure 108. Catalytic diffusion burner

a) H2 entrance port; b) Electric ignition wire; c) Distribution manifold; d) Catalytic combustion zone; e) Oxygen from air; f) Heat radiation

Catalytic heaters react very flexibly; the temperature can be adjusted and limited by means of the throughput; the heat is produced in the infrared range. The combustion is free of emissions. Since no
stacks are required, almost 100 % of the heat set free can be utilized. In addition, the moisture content of the air can be regulated by the reaction product, water.

Transport to Small Consumers. Hydrogen could be distributed via present nets to industrial and household consumers and could substitute natural gas. The energy transport capacity of such a low-
pressure network (ca. 10 kPa) leads to about the same heat flux for hydrogen as for natural gas.

9.3.3. Hydrogen Fuel Systems

In all sectors of transportation (space and aircraft propulsion, earthbound traffic systems including individual transport) the fuel systems hydrogen – air or hydrogen – oxygen can be used for the prime
mover.

9.3.3.1. Hydrogen Propulsion Systems for Space

Because of the large specific enthalpy of combustion LH2, combined with liquid oxygen (LOX), produces (apart from the system LH2 – LF2) the greatest specific impulse. Jet velocities up to 4500 m/s can
be reached using combustion chamber pressures of 5 to 22 MPa.

The hydrogen – oxygen motor NASA Centaur (Pratt & Whitney) was used for the third stage of the Atlas Centaur and the fourth stage of the Titan III rocket. For the Apollo project Rocketdyne motors were
used for the second and third stage of the Saturn V [422]. The peak of the US development is the Space Shuttle main engine, a high-pressure motor with a combustion chamber pressure of 21.5 MPa.

The European Ariane carrier system uses LH2 – LOX in its third stage. Using 10 t of this fuel combination, the HM7 motor develops a thrust of 60 kN over a period of 700 s. A further development for
800 kN thrust and a fuel requirement of 120 t with the designation HM60 is being developed [423].

9.3.3.2. Hydrogen as Aviation Fuel

In the future the following alternative aviation fuels will be available:

jet fuel “aviation grade” from coal (Synjet)

page 113 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
liquid methane (LCH4)
liquid hydrogen (LH2)

Liquid hydrogen has already been tested in 1957 in a B 57 Canberra by the NACA Lewis Research Center (Cleveland, Ohio) and its suitability as an aviation fuel was proven. Since 1973, especially in the
United States, investigation of the use of LH2 as an aviation fuel has been follwed up (overviews [424-426]). In 1988, test flights were made in the Soviet Union with a civil aircraft using LH2 as fuel.

Hydrogen is suitable for turbines in the sub- and supersonic regions. Engine designs for turbines converted to hydrogen fuel have certain advantages: They are smaller, easier to control and have a longer
lifetime than normal turbine fuels. The cryogenic fuel is used to cool engine oil and turbine cooling air, whereas the turbine core exhaust heats up the fuel before injection into the combustor. The savings
in the specific fuel consumption amount to 8 %. Ecological benefits are the lower noise level, smaller emission of NOx, and absence of CO, CO2, and hydrocarbons.

The energy content related to volume and mass of LH2 is the determining factor for the construction of a hydrogen aircraft. The four times greater fuel volume as compared to hydrocarbons for the same
energy content and the cryogenic properties require a completely new concept for the aircraft fuselage and tanks. To minimize surface – volume ratio, the insulated tanks are placed both fore and aft of the
passenger compartment. Weight advantages are gained by tank configurations which are integrated in the aircraft fuselage structure. For the complicated aircraft fuel system three absolutely equivalent
fuel booster pumps per tank compartment are foreseen [427].

The 2.8 times higher energy content of LH2 compared to kerosene (related to mass) makes it possible to have a much smaller lift-off weight for the same range and capacity. A comparison of design data
and various performance characteristics of subsonic aircrafts for the fuels listed is shown in Table 45.

Table 45. Comparison of design and performance data of alternate fueled subsonic aircraft (basis: 400 passengers, 44 000 kg payload, 5500 NMi range, 0.85 Mach cruising speed)

Liquid hydrogen (LH2) Liquid methane (LCH4) Synjet

Gross mass, kg 168 830 225 570 232 060

Block fuel, kg 21 620 58 980 72 530
Operating empty
mass, kg 103 300 116 170 107 370

Wing area, m2 296.8 385 380.3

Span, m 51.8 58.9 58.5
Fuselage length, m 65.7 61.4 60
Gross mass, kg 156 130
*
Block fuel weight , kg 15 270
Specific fuel con-
sumption (cruise),

kg h–1 N–1 0.206 0.504 0.615

Lift/drag (cruise) 17.4 19.21 19.13

Thrust per engine, N 135 000 177 000 185 000
Energy utilization,
kJ seat–1 km–1 636 723 759

* with respect to laminar flow control.

Similar studies have been presented for LH2-fueled supersonic aircraft. Here the immense fuel requirements are even more in favor of LH2. A comparison shows a reduction in the lift-off weight by ca.
50 %.

Laminar Flow Control (LFC). The cryo-fuel LH2 makes it possible to use the so-called cryogenic wall cooling technology. By cooling major aerodynamic surfaces of the aircraft to ca. 150 K, laminar flow
in boundary layers is stabilized with a consequent reduction of skin-friction drag. The necessary refrigeration can be taken from the evaporating LH2 directly or via a secondary refrigeration system (e.g.,

page 114 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
LN2). Overall drag reduction in the order of 20 % leads to a fuel saving of about the same amount at cruising speeds [428].

Airport requirements for LH2 as fuel have been investigated in different studies. The facilities include the liquefaction plant, LH2 storage capacity and LH2 fuel handling systems [429], [430]. Because of
planning and development issues, the use of LH2 as an alternative fuel for aviation will not occur before the next two aircraft generations, i.e., about 2030.

9.3.3.3. Hydrogen for Automotive Vehicle Transport

Mobile Otto and Diesel internal-combustion engines for vehicles can be operated on hydrogen or hydrogen – gasoline mixtures. RICARDO (1924) and BURSTALL (1927) carried out some of the first
investigations. In 1930, ERREN made intensive studies on hydrogen – air and hydrogen – oxygen motors. The first internal-combustion engine car in the United States fueled by hydrogen was presented
by BILLINGS in 1966 in Utah. At the moment further developments are in the Federal Republic of Germany, Japan, Australia and the Soviet Union. The literature on this topic is abundant. Overviews are
given in [431-433].

Modification of the Engine. Some properties of hydrogen make engine modifications necessary; low ignition energy and high flame speeds can cause self-ignition during the mixture preparation or flame
flash-back. The octane number of hydrogen is much lower than that of gasoline. However, the wide ignition range allows burning of lean mixtures and gives a large control range.

Various techniques are used to prepare the feed mixture to the combustion chamber. With external mixing a homogeneous hydrogen – air mixture is introduced. Uncontrolled preignition or flash-back into
the intake manifold is prevented by adding ballast, preferentially water, and by timed individual port injection of hydrogen close to the cylinder intake. Internal mixture formation is achieved by direct
injection into the combustion chamber. The necessary supply pressure must be at least 1.5 MPa, or even, depending on the time of injection, more than 10.0 MPa. This is only possible by using piston
pumps for liquid hydrogen. Figure 109 shows the principle layout of internal and external mixture formation.

Figure 109. Internal mixture formation (A) using cryogenic hydrogen (direct injection) and external mixture formation (B) using ambient-temperature hydrogen

Attainable specific charge heating values (energy content per cylinder loading) are compared in Table 46 under various operating conditions. Revolutions per minute (rpm) and effective efficiencies are
comparable to the gasoline motor. Because of the lower air quantities possible and the lower heating value of the mixture, external mixture formation with water injection gives power ratings comparable to
that of Diesel motors, whereas with internal mixtures ca. 40 kW/L cylinder volumes can be attained [435].

Table 46. Specific charge heating values of hydrogen – air mixtures in combustion engines compared to gasoline

Operating conditions Maximum of spec. Maximum of Maximum of

charge HV efficiency excess air
Hspec. max e max min

kJ/L kJ/L kJ/L

Gasoline 1 37.7 1.1 34.4 0.83 31.6

Hydrogen (external mixture formation) 1 31.9 0.4 15.5 0.2 8.3
Hydrogen (internal mixture formation) 1 45.3 0.4 18.1 0.2 9

Hydrogen-powered engines emit only water and NOx. Numerous investigations concerning the calculation and measurement of the nitric oxide emission have been carried out, e.g., [436], [437]. In Figure
110 typical NOx values for various modes of operation with hydrogen and gasoline are shown as a function of the equivalent ratio ( = l/ ). Especially in lean modes of operation, i.e., with excess air,
hydrogen motors operate more or less emission-free.

Figure 110. Typical NOx values for various modes of operation of hydrogen (a – d) and gasoline engines

page 115 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

a) External mixture formation, load governing; b) External mixture formation, quality governing; c) External mixture formation, optimum attainable values; d) Internal mixture formation; e) Typical values for Otto and Diesel
engines

On-Board Hydrogen Storage. At present three storage possibilities can be considered:

in the gaseous state in pressure vessels,

in the liquid state in vacuum-insulated tanks, or
chemically bound in metal hydride storage tanks.

Table 47 compares various mobile storage methods for vehicles. It is clear that none of the hydrogen storage systems have reached the undoubted advantages of the fossil energy carrier, gasoline.

Table 47. Comparison of mobile storage systems for hydrogen (basis: 45 m3 H2 corresponding to 15 L gasoline)

Liquid H2 Gaseous H2 Metal hydride-H2

Hydrogen content 57 L 45 m3 4 kg
Gross storage mass, kg 20 2×60 320
Hydride material, kg 220
Maximum operating pressure, MPa 0.4 30.0 5.0
Operating temperature, °C –253 –20 – +50 –20 – 100
Energy densities,
kWh/kg* 6–7 1.1 0.4

kWh/L* 0.95 0.55 0.8

* Energy density of a gasoline tank: 8 – 8 kWh/kg (8 – 9 kWh/L).

Pressure gas storage vessels are disadvantageous because of their large weight and volume. The high storage pressure of ≥ 20.0 MPa raises safety and supply questions.

Liquid hydrogen tanks for vehicles are larger than gasoline tanks and require a complicated construction because of the low storage temperature. Characteristic data for a tank in the boot of a car (BMW-
DFVLR experimental vehicle) are [434]:

double-walled cryo-vessel for 45 L of hydrogen,

vacuum superinsulation cooled by means of evaporating hydrogen,
mass (filled) of 45 kg,
operating pressure 0.2 MPa, lowest operation temperature –253 °C,
evaporation losses <2 %,
hydrogen delivery pressure via LH2 piston pumps 1.5 – 2.0 MPa.

Problems exist concerning the evaporation rate, safety questions and tanking techniques. Tanking of the test vehicles occurred at a computer-controlled, half-automatic LH2 station, so that maloperation
was avoided. The tanking time was ca. 5 min.

page 116 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Mobile hydrogen hydride storage vessels are used for example in Daimler Benz test vehicles. For details of the hydride technology see Section Hydride Technology. Suitable hydride-forming metals are
those, in which reversible hydrogen absorption occurs in the temperature range of –25 – 100 °C. These are low-temperature hydride materials, such as FeTi. High-temperature hydride materials, such as
Mg2Ni, have a higher capacity, but more energy at 350 °C is required to drive off the stored hydrogen. Figure 111 shows a hydride tank using a titanium – vanadium – manganese alloy for a vehicle.
Addition or removal of heat is performed by a heat carrier, such as a water – defrosting agent mixture connected externally during tanking or internally during operation of the vehicle. With a full storage
vessel the minimum pressure of 0.2 MPa required to start the motor even at –20 °C is available [438]. High-purity hydrogen (99.999 %) must be used to avoid a passivation of the active material. A quick
tanking operation with gaseous hydrogen at 5.0 MPa takes ca. 10 min.

Figure 111. Hydride storage tank for motor vehicle drives of HWT

a) Storage tube; b) H2-guide tube; c) Lamella; d) Hydride; e) Filter (sintered metal); f) Gas-collecting tube; g) H2 connection; h) Water inlet; i) Water outlet; j) Filling body; k) Vent valve connection; l) Shell

The motor concept chosen for a vehicle depends on the method of storing the hydrogen on board. Low-pressure gaseous hydrogen allows only external mixture formation, whereas liquid hydrogen,
because of its cryogenic properties, is more flexible (liquid compression, cooling, injection).

State of the Art. Representative for the many investigations being carried out in several countries, two development projects in the Federal Republic of Germany are described:

Between 1984 and 1986 Daimler Benz tested a fleet of 10 hydrogen vehicles under normal operating conditions in Berlin. Delivery vans (Fig. 112) were equipped with a 2.3 liter spark-ignition
engine with external mixture formation and water injection and a power of ca. 77 kW. Hydrogen storage was by four low-temperature, liquid-heated hydride storage units with a total mass of 560 kg
and a hydrogen capacity of 66 m3. This was sufficient for a range of 120 km in city traffic. The total mileage for the entire fleet amounted to 160 000 km [439].
Deutsche Forschungsanstalt für Luft- und Raumfahrt (DFVLR) has been working with BMW since 1979 on vehicles with LH2 tanks and different approaches for the mixture formation. The present
state of development is the 745i Sedan with a 3.5 liter spark ignition engine with a so-called “late internal mixture formation,” i.e., hydrogen injection during the compression stroke. The power output
is claimed to be ca. 150 kW [440].

Figure 112. Daimler-Benz delivery van with modified Otto engine for hydrogen as fuel and onboard hydride hydrogen storage

a) Refueling connectors; b) Hydride hydrogen storage; c) Filter; d) Pressure-reducing device; e) Magnetic shut-off valve; f) Hydrogen metering device; g) Differential pressure controller; h) Hydrogen distributor; i) Inlet
manifold; j) Hydrogen injection nozzles; k) Recirculating pump; l) Exhaust gas – water heat exchanger; m) Exhaust flaps; n) Storage for engine-water injection system; o) Water feed pump; p) Filter; q) Water metering
device; r) Water injection nozzles; s) Electronic control unit

9.3.4. Electricity Generation from Hydrogen

Apart from the direct utilization of hydrogen for heating purposes (see Section Combustion and Heating) the generation of electricity is important in the hydrogen energy concept. This can be done
indirectly via combined heat – power processes or directly by electrochemical means.

9.3.4.1. Heat – Power Processes

Suitable heat – power processes for power generation from hydrogen are [441]:

1. conventional steam turbine processes with hydrogen-fueled boilers,

2. open hydrogen – air gas turbine processes, and
3. hydrogen – oxygen steam turbine processes

The thermal efficiency of various hydrogen-fired power station processes with up to 100 MW power is given in Figure 113 as a function of the range of turbine-entrance temperature tE. A hydrogen –
oxygen power station with condensation attains an efficiency of 52 % at full load. This type of power station (line a) is most suitable for peak-load power production. For immediate increase of power
output of fossil-fired ground-load or medium-load power stations a hydrogen – oxygen steam production unit can be integrated (see Fig. 113, line b instantaneous reserve plant).

page 117 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 113. Thermal efficiency of various hydrogen-fired power station processes

a) Hydrogen-oxygen power station; b) Conventional hydrogen power station; c) Hydrogen power station with gas turbine; d) Hydrogen power station with gas turbine, improved design

A new component suggested for power stations is the hydrogen – oxygen steam producer. An experimental version for a thermal capacity of 15 – 40 MW exists at the DVFLR (Fed. Rep. of Germany). A
schematic is shown in Figure 114. Hydrogen and oxygen are introduced through a suitable lance and equally distributed in the combustion chamber before burning. The combustion product, the
temperature of which is 3000 °C, is cooled down to the required steam temperature by water quenching and fed to the turbine. It is particularly important to maintain the stoichiometric ratio of the fuel
mixture in order to avoid noncondensable gas components (excess oxygen or hydrogen). This is achieved by means of a sophisticated control strategy. Starting from zero, full load can be reached within
1 s (here 40 MW); steam with a pressure of 4.0 – 9.0 MPa and a temperature of 560 – 950 °C can be produced [442].

Figure 114. Schematic of the DFVLR H2 – O2 steam producer

a) Injection head; b) Gas distribution; c, d) Injection manifold; e) Combustion chamber; f) Injection rings; g) Radial water injection

9.3.4.2. Electrochemical Energy Conversion

In a reversal of the electrolysis process, the chemical energy of hydrogen can be converted directly into electrical energy. This takes place in fuel cells by allowing the “cold” combustion of hydrogen and
oxygen to occur electrochemically ( Fuel Cells) [443].

At room temperature the maximum possible voltage of the hydrogen – oxygen system is 1.23 V. In practice, cell voltages of 0.7 V are reached. Since heat – power conversions are avoided, efficiencies of
40 – 60 % can be attained. For details of the mode of operation, electrodes, catalyst materials and construction see [444], [445]. The most developed versions are the low-temperature fuel cells, which are
usually identified by the electrolyte employed, as alkaline fuel cells (AFC), phosphoric acid fuel cells (PAFC), and solid polymer electrolyte fuel cells (SPEFC).

Alkaline fuel cells are used for power production in space as well as for special applications (relay stations, emergency power supply, military purposes). Without exception, they use pure hydrogen and
carbon dioxide-free oxygen and have a low specific mass (7.2 kg/kW) and a long lifetime. They are, however, very expensive (1500 $/kW) [446].

Phosophoric acid fuel cells operate at 200 °C and use hydrogen or hydrogen – carbon monoxide mixtures produced from natural gas and air. The complete unit contains not only the cell modules but also
various components, such as utility supplies, systems for temperature control, means to remove product water, and controls for the electrolyte pumping.

High-temperature fuel cells are still in the development stage; molten carbonate fuel cells (MCFC) operate at 600 °C and use an alkali-metal carbonate melt as electrolyte. Solid-oxide fuel cells (SOFC)
contain ceramic ZrO2 membranes, (solid electrolytes); on these electrolytes hydrogen reacts with air at 800 – 1000 °C. The hydrogen production from methane and steam can be integrated into the cell
[447].

Fuel Cell Power Plants. Hydrogen, hydrogen-rich gases, or gases which can be converted to hydrogen can serve as fuel for fuel cell power plants. As depicted in Figure 115, such a facility for district
electricity and heat generation comprises three subsystems; a fuel processor, a power section made up of a number of modules and a power conditioner. The suitability of the phosphoric acid fuel cell for
this purpose has been proven in the United States in a field test (46 40-kW PAFC units) [449]. In Japan two 1 MW blocks have been built as an alternative energy production to thermal power generation
[450].

Figure 115. Fuel cell power plant [448]

a) Fuel processor; b) Fuel cell power section; c) Power conditioner

9.4. Future Developments

Several books and monographs dealing with the perspectives and planning of the utilization of hydrogen as a major universal energy carrier have been published [386], [389], [451-453].

page 118 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
9.4.1. Production Schemes
Possible Schedule. The introduction of hydrogen as an energy carrier can be assumed to be governed mainly by economic aspects. The following developments can be expected [454]:

1. First stage of utilization. Future hydrogen utilization will be based on current uses of hydrogen, which are mainly nonenergetic. Hydrogen produced from non-hydrocarbon sources (normally from
electrolysis) will be introduced to serve similar purposes.
Primary energywill be supplied by nuclear or hydropower stations with energy reserves available during off-peak periods. The economic conditions for electrolytic production of hydrogen are
especially favorable in France (nuclear power) [455] and Canada (hydropower) [456].
2. Special Applications. The combined production and utilization of hydrogen and oxygen by electrolysis could provide an additional economic impetus. Possible fields of utilization are:
heavy oil upgrading (hydrogen for hydrogenation, oxygen for residue gasification) [457], [458]
wood gasification combined with methanol production [459]
all other kinds of coal, residue, or waste gasification with H2, CH4, CH3OH, etc. as products
biological process such as aerobic wastewater treatment (oxygen required). Hydrogen together with any methane produced in anaerobic stages, could be used as fuel for power production
with gas motors or for a denitrification process.
3. Introduction as a Direct Energy Supply. Use of hydrogen will start as a supplement to actual energy markets, addition of up to 10 % hydrogen in natural gas nets or use as a fuel in transportation
(city buses, underground/underwater devices, aircraft).
The primary energy supply for the uses in both last sections would be the same as for the utilization as energy off peak power. In addition, wind, water, and solar units, which use electrolytically
produced hydrogen for long-term storage would also be used.
4. Advanced Stage of Utilization. In countries with high solar radiation intensity but little industry, hydrogen could be used as a means of storing solar energy [460]. The solar radiation is converted into
electric power using solar collectors, photovoltaic units, or Stirling motors driven by solar heat.
First decentralized applications for hydrogen as an energy carrier could be for refrigerating large-scale units and for supplying heat in agricultural and industrial operations.
Schemes for supplying the total energy requirements to remote areas by hydrogen have also been suggested, e.g., energy requirements of an island in Alaska supplied via wind energy and
hydrogen [461].
5. Hydrogen-Energy World. The final goal of a world selfsufficient in energy (that does not depend on fossil resources) is the erection of large solar farms in the sunflooded desert areas of Africa, Asia,
and Australia. Countries in these regions could export hydrogen (or electric power) to industrial regions with less sun. For more details see [454, pp. 297 – 304].
According to this study, an area of 0.14× 106 km2 (0.1 % of the land surface of the globe), will be required for the production of 1000×106 tce/a. The total surface of suitable areas in desert regions
is 0.6×106 km2 which is more than sufficient.
The material requirements for the construction of a solar farm are also within the capabilities of an industrialized country. This is shown in Table 48 [454, p. 321] for the Federal Republic of
Germany. Nickel and silicon are the exceptions, the first because of the limited resources available and the latter because appropriate production technologies still have to be developed. Supply of
other key items of solar farms such as solar cell panels, heliostat mirrors, Stirling motors, power generators, and electrolysis units, is also adequate (see Fig. 116) [454, p. 329].

Table 48. Comparison of material requirements for hydrogen plants with the consumption of primary raw materials in the Federal Republic of Germany (1983). Projected annual production of
hydrogen plants ca. 155 PJ/a (equivalent to the energy produced by 100 100-MW-units per year, usable ca. 4000 h/a) [389]

Raw material Annual con- Consumption for production of solar plants, % of annual consumption (1983)

sumption 1983*,
10 3t/a Photovoltaic solar cell Solar tower Paraboloid mirrors, solarthermic Wind converter

Steel 26 600 12.7 32.0 24.0 11.8

Aluminum 1085 11.5 16.6 18.4 1.8
Nickel 63 95.2 206.3 135.0 47.6
Glass 1240 126.6 209.0 178.2 6.5
Concrete 210 000 10.6 18.3 21.0 5.0
Silicon 0.36 16 667

* Annual primary energy use in 1980: World 300×103 PJ/a; Western Europe 76×103 PJ/a ; Fed. Rep. of Germany 7.6×103 PJ/a ; North America 90×103 PJ/a.

Figure 116. Comparison of annual production of certain industrial products in the Federal Republic of Germany and the required comparable items for hydrogen plants for the hydrogen energy concept [460]

page 119 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Projects. Already existing hydrogen projects serve and will serve the following purposes:

1. Demonstration of feasability
2. Study of technical problems occurring during operation under realistic conditions
3. Optimization and coordination of the various systems involved (e.g., mechanical – photovoltaic systems, electronic – photovoltaic systems)
4. Moving from the “tailor-made” stage of the components to a prototype production stage
5. Demonstration of self-sufficiency in terms of energy supply without fossil fuel or nuclear power

Although some individual components of a future hydrogen economy have already been tested (see Section Hydrogen-Energy Conversion Systems), few projects have realized the concept in its entirety
from hydrogen production to its utilization. In 1979, R. E. BILLINGS built a house in Independence, Missouri, in which the energy supply, including that for car and tractor, was based on hydrogen. The
primary energy was by hydropower, storage by means of metal hydrides and solar energy was used as subsidiary energy for heating purposes.

In Harnøsand, Sweden, O. TEGSTRØM has converted the energy supply of his car and his house, to hydrogen. Wind is used to supply the primary energy via and turbines with 55 kW. Metal hydrides are
used for hydrogen storage.

Hysolar [462]. Founded in 1985, Hysolar (a joint venture between the Federal Republic of Germany and Saudi Arabia) foresees the installation of three photovoltaic electrolysis plants: a 100-kW (electric
power) plant is to be built in Saudi-Arabia; two plants with 10 and 2 kW at a German and a Saudi Arabian university respectively serve research and demonstration purposes [462].

Solar Hydrogen Bavaria (SWB) is scheduled to go on stream in the Federal Republic of Germany in 1989 [463]. Figure 117 shows a diagram of the plant. Project costs amount to more than 50×106 DM
(33×106 DM for the plant, 12×106 DM for personnel, 5×106 DM for infrastructure).

Figure 117. Block diagram of the Bavaria Solar Hydrogen Plant (planning status)

a) Photovoltaic cell; b) d.c. Converter; c) Power net; d) a.c./d.c. Converter; e) Electrolysis power supply; f) Electrolysis unit; g) Oxygen purification; h) Hydrogen purification; i) Alkaline fuel cell; j) Phosphoric acid fuel cell; k)
Gas heater; l) Gas heater; m) Heat storage

Plant data:

Solar generator – 500 kWel from various modules

Photovoltaic cells: total surface area 7000 m2
Electrolysis unit:
100 kWel conventional KOH electrolysis
100 kWel zero gap KOH/NaOH electrolysis
80 kWel solid polymer membrane electrolysis
Gas storage:
5000 m3 (STP) of H2 at 3 MPa
500 m3 (STP) of O2 at 3 MPa
Fuel cells:

page 120 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
6 – 8 kWel alkaline type
80 kWel phosphoric acid type
Catalytic combustion: 4 kWth
Gas boiler: 20 kWth

Euro-Quebec Project [464]. In this study initiated in the Federal Republic of Germany, the transport of hydroelectric power from Canada to Europe by means of hydrogen is being investigated. The
Canadian hydropower is converted into hydrogen by electrolysis. The hydrogen is transported in one of three forms: (1) as liquid hydrogen shipped in special tankers (or eventually by aircraft); (2) as liquid
ammonia shipped in special tankers; or (3) as methylcyclohexane shipped in oil tankers. The hydrogen is used in Europe for electricity, heat generation, vehicle and aviation propulsion, and hydrogen as
an additive to natural gas for domestic or industrial use.

Figure 118 A shows the mass flow diagram developed for Phase I of the study; the energy flow diagram is shown in Fig. 18 B.

Figure 118. Flow diagrams for the transportation of hydrogen from Canada to Europe (Euro – Quebec project). It is planned to transport hydrogen in liquid form or chemically bound as methylcyclohexane (MCH) or
ammonia by sea or air

A) Mass flow diagrams (transportation energy is neglected)

B) Energy diagram (in the energy diagram for chemically bound hydrogen only ammonia is shown because MCH is very similar and slightly less effective, = 0.5 instead of 0.514)

Other Projects. Apart from these direct hydrogen projects, a great number of solar energy stations and wind farms (particularly common in California) can be seen as precursor of the hydrogen energy
economy.

All hydrogen projects to date convert the universal but expensive energy carrier electricity with notable losses to hydrogen. Sometimes the cycle is completed and the hydrogen is converted back to
electricity.

Direct production of hydrogen from solar radiation (e.g., by photochemical or biological methods) is still under research.

9.4.2. Hydrogen Energy Economics

Various factors can accelerate the introduction of the hydrogen technology, but do not exclude other solutions for the energy supply such as drastic savings or introduction of new energy forms. For time
schedules for the introduction of hydrogen technology and other prognoses see [451], [454], [465]. The political decision to use hydrogen as an energy carrier reflects the following:

1. Technological and Ecological Aspects. The utilization of fossil and nuclear energy can be reduced in favor of energy savings, renewable resources, and solar energy. This could be forced by climate
changes and environmental pollution.
2. Economic Aspects. Increased energy costs and storage costs for conventional primary energy and energy carriers could accelerate the introduction of the hydrogen economy.
3. Political Aspects. National hydrogen energy economics could be favored by the effect of primary energy import on the balance of trade, the creation of jobs by national energy programs, and the
improvement of the structure of the economy.
4. Security of Supply. Energy produced from a local primary energy source can be stored in the form of hydrogen.
5. Social Compatibility. Energy production is coupled with an environmentally acceptable energy carrier.

Price estimates for hydrogen based on present-day technologies (see Fig. 119) [454, p. 341] depend on many factors. Comparisons of hydrogen production by means of solar cell plants, solar collector
plants, parabolic mirror plants, or wind farms cannot consider the innovative potential of each technology over the projected time span.

Figure 119. Hydrogen production costs for four hydrogen plants and three different start-up dates (interest rate: 8 %, depreciation: 30 a, construction: 4 a, inflation: 5 %). Comparison with real escalating hydrogen (– – –)
and natural gas (——) costs (J = annual real rate of escalation, cost of electricity production = E, of hydrogen production = H, of piping = P, of operation = O, and T of transport for 3300 km)

The plants are: PV = photovoltaic solar cell plants, ST = solar thermic, solar power plants, PM = parabolic mirror solar thermic plants, WE = wind energy farms [454]

Another method to assess the role of hydrogen technology in the future energy supply is to compare the efficiency of energy supply systems. In Figure 120 the efficiencies of room heating using electric
power produced by photovoltaic means are compared in the Federal Republic of Germany. The power is generated inside the country and then distributed either in a hydrogen pipeline (A) or as such (B)
to the user. The much more efficient power production in the Sahara is also considered. Energy can be transported directly as current (D) or as hydrogen (C).

page 121 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 120. Comparison of photovoltaic energy generation in the Federal Republic of Germany (A, B) and in arid regions in North Africa (Sahara) (C, D), transportation in the form of hydrogen (A, C) or in the electric grid (B,
D). End use: room heating in the Federal Republic of Germany [466]

9.4.3. International Research

A particular advantage of hydrogen technology is that all necessary systems and components are in an advanced state of development. Some technologies, such as hydrogen transport and storage, have
been used industrially for decades. In spite of this, intensive research is still required in important areas. These areas are being investigated by several countries which cooperate under the auspices of the
International Energy Agency, IEA (Table 49) [467].

Table 49. Hydrogen research activities of some member countries of the International Energy Agency (IEA) [467]

Subtask Subject Federal Republic of Germany Japan Sweden Switzerland United States

Storage Metal hydride storage × × × × ×

Liquefaction of hydrogen ×
Hydrogen storage on carbon ×
Conversion Engine with inert gas recirculation ×
LH2-Pump for internal combustion engines ×
Reversible hydrogen storage electrode ×
Safety Alternation of material properties by
hydrogen embrittlement × ×
Safety aspects of hydrogen combustion ×
Man years 1986: 13 10.5 1.5 3.2 4.5
Man years 1987: 11.8 10.5 1.5 4 4.5
Man years 1988: 11 10.5 1.5 4 4.5

Canada. Canada plans to generate hydrogen by hydropower and to test applications in pilot projects. Some activities are listed in Table 49. Systems analysis studies regarding special applications for
hydrogen have also been carried out [468]. Canadian activities involve basic research in universities, industry, (especially in the field of electrolysis), and national research centers (sponsored by the
National Research Council, NRC).

Federal Republic of Germany [452], [466]. Industry, research institutions, and universities are continuing the development of hydrogen technology with the help of government funds. Pilot projects are
described in Section Hydrogen Energy Economics. In the 1990s, work will concentrate on following areas:

1. Energy generation (photovoltaic and solar thermal techniques in cooperation with countries with high solar radiation intensities)
2. Hydrogen production (electrolysis, advanced alkaline electrolysis, high-temperature electrolysis, basic research on biological hydrogen generation)
3. Applications (fuel cells, stationary power applications, peak power, power – heat combined cycle)
4. Materials and safety

The Federal Republic of Germany takes the leading role in the international funding of hydrogen technology. State funding in the energy and renewable resources areas amounts to 200×106 DM/a;
100×106 DM of this sum are to be spent in the 1990s on hydrogen technology.

Japan. Japanese activities are concentrated in the Sunshine Project of the Ministry of International Trade and Industry (MITI). According to annual reports on this topic [469], the following areas are being
investigated (1986):

page 122 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
1. Hydrogen production by SPE water electrolysis (see Section Electrolysis)
2. Storage and transportation of hydrogen by using metal hydrides
3. Utilization of hydrogen (combustion, reciprocating engines, cars, etc.)
4. Safety (explosion prevention).
5. Photochemical processes, thermochemical cycles (Br– system cycle), high-temperature – high-pressure water electrolysis, and hydrogen from biomass

Funding has decreased continuously from a maximum of 950×106 Yen in 1980 to 190×106 Yen in 1987. However, funds for solar energy (which amounted to 7560×106 Yen in 1987) and overhead costs
are not included in these figures. Japan is also a member of the IEA (see Table 49).

United States. Research in the United States deals principally with energy generation via solar radiation and thus only indirectly with hydrogen technology, although a remarkable amount of solar energy
is generated, in the South and the West of the United States in particular. This energy is fed directly into the grid. Hydrogen is at present not under consideration as an energy carrier.

Basic research at various universities takes place within the IEA framework and is sponsored by industry and government.

Soviet Union. Proof of the high standard of research in hydrogen technologies was the 1988 report of the flight of a Tupolev aircraft powered by liquid hydrogen (see Section Hydrogen-Energy Conversion
Systems).

[Top of Page]

References
1. Gmelin System no. 2, 8.
2. Ullmann, 3rd ed., 18, 501.
3. Ullmann, 4th ed., 24, 243.
4. R. D. McCarthy: Hydrogen Technology Survey – Thermophysical Properties, NASA SP-3089, Washington D.C. 1975.
5. JANAF: Thermochemical Tables, NSRD NBS 37, Washington D.C. 1975.
6. H. M. Roder in R. D. McCarthy, J. Hord (eds.): Selected Properties of Hydrogen, Dept. of Commerce NBS Report, Boulder 1981.
7. K. E. Cox, K. D. Williamson, Jr. (eds.): Hydrogen: Its Technology and Implications, 2nd vol. 1 – 5, Boca Raton 1979.
8. L'Air Liquide: Encyclopedie des Gaz, Elsevier, Amsterdam 1976.
9. C. L. Newton, Chem. Process. Eng. (London) 48 (1967), no. 12, 51 – 58.
10. European Space Agency Contract Report, Contract 5977/84/NL/PB(SC), Chilton, United Kingdom, Jan. 1986.
11. K. F. Mucker, D. White et al., Phys. Rev. Lett. 15 (1965) 586. Links
12. N. Sullivan et al., Recherche 11 (111) (1980) 587, see also Bull. Inst. Int. Froid 2 (1984) 484.
13. G. A. Neece et al., J. Comput. Phys. 7 (1971) 621. Links
14. F. V. Grigorev, S. B. Kormer, O. L. Mikhailov, JETP Lett. (Engl. Transl.) 16 (1972) no. 5, 201.
15. R. Hawke, Phys. Rev. Lett. 41 (1978) 994. Links
16. Landolt-Börnstein IV 2c, Springer Verlag, Berlin-Heidelberg-New York 1980.
17. C. L. Young (ed.): IUPAC Solubility Data Series, vol. 5/6; Hydrogen and Deuterium, Pergamon Press, Oxford 1981.
18. D. Behrens, R. Eckermann (eds.): DECHEMA Chemistry Data Series vol. 6; “Vapor-Liquid Equilibria for Mixtures of Low Boiling Substances,” Dechema, Frankfurt 1982.
19. J. R. Lacher, Proc. R. Soc. London, Ser. A 161 (1937) 526.
20. W. Mueller, J. Blackledge, G. Libowitz (eds.): Metal Hydrides, Academic Press, New York – London 1968.
21. G. Alefeld, J. Völkl (eds.): Hydrogen in Metals, vol 1, 2, Springer – Berlin – Heidelberg – New York 1978.
22. H. J. Hoge, R. D. Arnold, J. Res. Natl. Bur. Stand. (U.S.) 47 (1951) 63.
23. H. U. Karwat, Chem. Ing. Tech. 32 (1960) 605.
24. P. Protopapas et al., Hydrogene Met. Congr. Int. 1972, 414.
25. G. Holleck, J. Phys. Chem. 74 (1970) 503.
26. D. R. Stull et al. in JANAF Thermochemical Tables, 2nd ed., NSRD NBS, Washington D.C. 1971.
27. J. W. Colton, Hydrocarbon Process. 60 (1981) 177.
28. P. A. Giguere, J. Chem. Educ. 56 (1979) 571. Links
29. D. J. G. Ives, G. J. Janz: Reference Electrodes, "Chap. 2", Academic Press, New York 1961.

page 123 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
30. N. N., Pure Appl. Chem. 57 (1985) 531. Links
31. J. Hord, Int. J. Hydrogen Energy 3 (1978) 157.
32. I. L. Drell, F. E. Belles: Survey of Hydrogen Combustion Properties, NACA Report 1383, Washington D.C. 1952.
33. M. Fischer, H. Eichert, DFVLR-Nachr. 34 (1981) 33.
34. W. C. Eichelberger et al., Chem. Eng. Prog. 57 (1961) no. 8, 94.
35. Tritium in the Physical and Biological Sciences, vol. 1, Int. Atomic Energy Agency, Vienna 1962, p. 56.
36. Kirk Othmer, 7, 539.
37. G. Vasaru: Deuterium and Heavy Water – A Selected Bibliography, Elsevier, Amsterdam – Oxford – New York 1975.
38. R. Prydz: The Thermodynamic Properties of Deuterium, NBS Report 9276, Boulder (Col.) 1966.
39. D. H. Liebenberg, R. L. Mills, J. C. Bronson, LA-7071-MS Informal Report, Los Alamos 1978.
40. P. C. Souers et al.: Three-Phase Region of D2-DT-T2, Lawrence Livermore Laboratory Report UCRL – 79 036, 1977.
41. H. Gutowski, Kerntechnik 12 (1970) no. 5/6, 233.
42. Y. W. Wong et al., Sep. Sci. Technol 15 (1980) 423.
43. G. D. Sandrock, E. L. Huston, Chem. Ing. Tech. 11 (1981) 754.
44. K. B. Wiberg, Chem. Rev. 55 (1955) 713. Links
45. E. A. Evans: Tritium and its Compounds, 2nd ed., J. Wiley & Sons, New York 1974.
46. A. Roberts, Sci. Am. 206 (1963) no. 6, 94.
47. G. Winnewisser: Cosmochemistry, Springer Verlag, Berlin 1973.
48. E. Goebel et al., Oil Gas J. 81 (1984, May 7) 215.
49. N. Getoff: Wasserstoff als Energieträger, Springer, Wien – New York 1977, p. 5. NASA/ASEE Systems Design Study “Hydrogen Energy Carrier,” I, II (1973)
50. H. Bertling, G. Nashan, Erdöl Kohle Erdgas Petrochem. 34 (1981) 397 – 401.
51. G. Dalluge, P. Polster, Chem. Tech. (Leipzig) 33 (1981) 636 – 640.
52. G. Bernard, J. Vidal, Pétrole et techniques 307 (May 1984) 28 – 38.
53. J. F. Meckel, C. Flockenhaus, Chem. Anlagen + Verfahren 1981, May, 122 – 138.
54. Y. Osumi, CEER Chem. Econ. Eng. Rev. 17 (June 1985) 5 – 20.
55. J. G. Kronseder, M. J. P. Bogard, Encycl. Chem. Process 9 (1979) 299 – 328.
56. G. G. Mayfield, V. H. Agreda, Energy Prog. 6 (1986) no. 4, 214 – 218.
57. J. F. Clausen, C. A. Zee: TRW-Systems and Energy, EPA contract 68–02–2635.
58. A. K. Kuhn, Chem. Eng. Prog. (1982) no. 4, 64 – 68.
59. W. N. Clark, R. L. Litzinger, I. R. Straughan, Energy Prog. 7 (1987) no. 2, 99 – 104.
60. B. Brunke et al., Erdöl, Erdgas 103 (1987) 289 – 293.
61. D. H. Cortez, C. J. La Delfa, Hydrocarbon Process. 61 (1981, Feb.) 111.
62. F. A. Clauson, C. P. Marion, AIChE – IMIQ Energy Meeting, Acapulco, Oct. 1980.
63. Uhde GmbH, Dortmund, private communication (1988).
64. H. Jungfer, R. Lohmüller, Erdöl Kohle Erdgas Petrochem. 39 (1986) no. 4, 188 – 1892.
65. C. P. Marion, J. R. Muenger, AiChE, (Houston, Texas, USA, 1981) 4.
66. H.-J. Madsack, Hydrocarbon Process. 61 (1982) no. 7, 169 – 172.
67. Y. Shimizu, CEER Chem. Econ. Eng. Rev. 10 (1982) no. 7, 9 – 14.
68. Linde AG, DE 34 14 716.0, 1984 (R. Lohmüller, M. Heisel, P. Hesse, U. Lahne).
69. H. Jungfer, Linde Reports on Science and Technology 40 (1985) 14 – 20.
70. R. M. Eccles, A. M. Gray, W. B. Livingston, Oil Gas J. 79 (1982, Apr. 12) 123 – 129.
71. Osaka Gas Corp. JP 108 1726, 30, 34.
72. G. W. Bridger, Chem. Process Eng. (London) 53 (1972) 38.
73. Catalyst Handbook ICI, Wolfe Scientific Books, London 1972.
74. J. Rostrup-Nielsen, Chem. Eng. Prog. 73 (1977) no. 9, 87 – 92. Links
75. Toyo Engineering, US 4 025 457, 1977 (T. Tomita, K. Kikuchi, T. Sakamoto).

page 124 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
76. R. G. Minet et al., EPRI-Report EM 1487 (1980), EM 2096 (1981)
77. Nitrogen, 144 (1983) 33.
78. D. R. Stull, E. F. Westrum, G. C. Sinke: The Chemical Thermodynamics of Organic Compounds, Wiley, New York 1969, p. 631 ff. W. D. Seider, R. Gautam, C. W. White: “Computation of Phase and
Chemical Equilibrium. A Review,” ACS-Symp. Ser. 124 (1980) 115 – 134.
79. J. P. van Hook, Catal. Rev. Sci. Eng. 21 (1980) no. 1, 1 – 51.
80. J. C. de Deken, E. F. Devos, G. F. Froment, ACS Symp. Ser. 196 (1982) 181 – 197.
81. A. P. Murray, T. S. Snyder, Ind. Eng. Chem. Process Des. Dev. 24 (1985) 286 – 294.
82. C. P. P. Singh, D. N. Saraf, Ind. Eng. Chem. Process Des. Dev. 18 (1979) no. 1, 1 – 7.
83. F. J. Dent et al., Trans. Inst. Gas Eng. 95 (1945/46) 602 – 709.
84. S. Paloumbis, E. E. Petersen: “Chemical Reaction Engineering,” ACS Symp. Ser. 196 (1982) 489 – 497.
85. J. W. Colton, Hydrocarbon Process. 60 (1981) no. 1, 177 – 184.
86. D. L. Trimm, Catal. Rev. Sci. Eng. 16 (1977) no. 2, 155 – 190.
87. K. H. Brachthäuser, R. R. J. Jungerhans, H. Tielen, Gas Wärme Int. 22 (1973) no. 6, 209 – 214.
88. M. R. Agarwal, T. Bajpai, Chem. Age India (1977) no. 11, 995 – 1010.
89. A. Watson, Hydrocarbon Process. 62 (1983) no. 3 91 – 95.
90. H. D. Marsch, J. J. Herbort, Hydrocarbon Process. 61 (1982) no. 6, 101 – 105.
91. Ann., Nitrogen 166 (1987) no. 2, 24 – 31.
92. W. Balthasar, D. J. Hambleton, Int. J. Hydrogen Energy 5 (1980) 21 – 31.
93. W. F. van Weenen, J. Tielrooy, Proc., 4th World Hydrogen Energy Conf., Pasadena, California, USA, 1982.
94. S. P. Goff, S. I. Wang, Chem. Eng. Prog. (1987) no. 8, 46 – 53.
95. J. L. Heck, T. Johansen, Hydrocarbon Process. 57 (1978) no. 1, 175 – 177.
96. S.-J. Wang, Oil Gas J. 80 (1982) no. 11, 55 – 62. Links
97. J. R. Rostrup-Nielsen, H. Stahl, N. R. Udengaard, Fuel Cell Seminar, Tucson, Arizona, USA, 1985.
98. Endurance Testing of a High-Efficiency Steam Reformer for Fuel Cell Power Plants, EPRI AP-6071, Project 2192–1 Final Report, October 1988.
99. United Technologies Corp., US 4 071 330, 1978 (R. A. Sederquist).
100. J. Singh, H. F. Nielsen, R. Streicher, Brennst. Wärme Kraft 33 (1981) 412 – 416.
101. L. Silberring, Proc. of British Sulphur Corp., 4th Conf., paper 5, London, 1981, pp. 93 – 112.
102. M. Oertel et al. Chem. Eng. Technol. 10 (1987) 248 – 255.
103. E. J. Nobles in A. V. Sluck, G. Russel James (eds.): Ammonia, vol. I, Marcel Dekker, New York 1973, pp. 275 – 290.
104. M. Fytzani-Stephanopoulos, G. E. Voecks, Energy Progress 1 (1981) nos. 1 – 4, 52 – 58.
105. Alsands Projekt, Oil Gas J. 78 (1980) no. 24, 61.
106. G. R. Charamonte, R. Sharma, Hydrocarbon Process. 62 (1982) no. 9, 255 – 257.
107. R. R. Dean, J. L. Mauleon, W. S. Letzsch, Oil Gas J. 80 (1982) no. 40, 75 – 80; no. 41, 168 – 176.
108. Atlantic Richfield Comp., G. P. Masologites, L. H. Beckberger, R. R. Edison, company leaflet, 1974, Harvey, Illinois, United States.
109. R. Müller, Proc. World Hydrogen Energy Conf., 4th, Pasadena, California USA, 1982.
110. J. G. Wragg, M. A. Kaleel, C. S. Kim, Coal Processing Technology VI, Am. Inst. Chem. Eng. (1980) 186 – 192.
111. H. Wendt in C. J. Winter, J. Nitsch (eds.): Wasserstoff als Energieträger, Springer Verlag, Berlin 1986, p. 179.
112. C. Carpetis, Dtsch. Forsch. Versuchsanst. Luft Raumfahrt Nachr. 34 (1981) 15.
113. H. Wendt, Chem. Ing. Tech. 56 (1984) no. 4, 265 – 272.
114. J. Divisek, Brennst. Wärme Kraft 38 (1986) no. 11, 512 – 515.
115. H. Vandenborre, R. Leysen, Electrochim. Acta 23 (1978) 803 – 804. Links
116. J. Divisek, P. Malinowski, J. Electrochem. Soc. 133 (1986) no. 5, 915 – 920. Links
117. M. S. Casper, Chemical Technol, Review no. 102, Noyes Data Corp. (Park Ridge, N.J., 1978).
118. G. Sandstede, Chem. Ing. Tech. 61 (1989) 349 – 361.
119. Brown Boveri, R. Oberlin, M. Fischer, Technik 8 – 86, 1986, pp. 445 – 450.
120. Japan's Sunshine Project, Annual Summary, 1983 – 1986 of Hydrogen Energy, R&D, Pub. Japan Ind. Tech. Ass., Tokyo 1987.

page 125 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
121. G. Imarisio, A. A. Strub (eds.): Hydrogen as an Energy Carrier Proc., 3rd Int. Seminar, Lyon 1983, Dordrecht, Netherlands 1983, pp. 237 – 243.
122. W. Baukal, Chem. Ing. Tech. 41 (1969) 791.
123. W. Dönitz, R. Streicher, Chem. Ing. Tech. 52 (1980) 436.
124. E. Erdle, W. Dönitz, VDI-Ber. 602 (1987) 51 – 51.
125. A. O. Isenber, Solid State Ionics (1981) no. 3/4, 431 – 437.
126. K. Iwahara, Kagaku Sochi 28 (1986) no. 1, 90 – 91.
127. K. H. Quandt, R. Streicher, Int. J. Hydrogen Energy 11 (1986) no. 5, 309 – 315.
128. D. Bergner, M. Hartmann, R. Staab, Chem. Ing. Tech. 59 (1987) no. 4, 271 – 280.
129. I. Barin, O. Knacke: Thermochemical Properties of Inorganic Substances, Springer Verlag, Berlin – Heidelberg – New York – Düsseldorf 1973.
130. J. E. Funk, R. M. Reinstrom, Ind. Eng. Chem. Proc. Res. Dev. 5 (1966) 366.
131. B. M. Abraham, F. Schreiner, Ind. Eng. Chem. Fundam. 13 (1974) 30.
132. H. Hofmann, Chem. Ing. Tech. 48 (1976) 87.
133. G. E. Beghi, Int. J. Hydrogen Energy 10 (1985) 431.
134. H. Engels, J. E. Funk, K. Hesselmann, K. F. Knoche, Int. J. Hydrogen Energy 12 (1987) 291.
135. R. Schulten: Nuclear Energy as a Primary Energy Source for Hydrogen Production, 2nd World Hydrogen Conference, Zürich 1978.
136. M. G. Bowman: Interfacing Primary Heat Sources and Cycles for Thermochemical Hydrogen Production, 3rd World Hydrogen Energy Conference, Tokyo 1980.
137. C. E. Bamberger, D. M. Richardson, Cryogenics 16 (1976) 197. Links
138. C. E. Bamberger, D. M. Richardson, Cryogenics 18 (1978) 170. Links
139. G. DeBeni, J. Electrochem. Soc. 128 (1982) 1.
140. G. E. Besenbruch et al.: Hydrogen Production by GENERAL ATOMIC Sulfur Iodine Process. A Progress Report, 3rd World Hydrogen Energy Conference, Tokyo 1980.
141. K. F. Knoche, H. Schepers, K. Hesselmann: Second Law and Cost Analysis of the Oxygen Generation Step of the GENERAL ATOMIC Sulfur-Iodine Cycle. Proceedings of the 5th World Hydrogen
Energy Conference, Toronto 1984
142. H. Engels, K. F. Knoche, M. Roth: Direct Dissociation of Hydrogen Iodide—An Alternative to the GENERAL ATOMIC Process. Proceedings of the 6th World Hydrogen Energy Conference, Vienna
1986.
143. S. Sato et al. in : T. N. Veziroglu, J. B. Taylor (eds.): “Studies on the Nickel-Iodine-Process for Hydrogen Production III,” Hydrogen Energy Progress V, vol 2, Pergamon Press, New York 1984,
pp. 457 – 465.
144. S. Shimizu, Int. J. Hydrogen Energy 12 (1987) 687.
145. K. F. Knoche, J. E. Funk, Int. J. Hydrogen Energy 2 (1977) 377, 387.
146. J. M. Sedlak et al., Hydrogen production using solid-polymer-electrolyte technology for water electrolysis and hybrid sulfur cycle, General Electric Company, Wilmington, Mass. and Electric Power
Research Institute, Palo Alto, CA. (ed.), EM-1185, Research Project 1086–3, 1979, pp. 6/1–8-1.
147. K. Petrov, I. Nikilov, T. Vitanov, Int. J. Hydrogen Energy 10 (1985) 245.
148. B. D. Struck, et al. in G. Imarisio, A. S. Strub (eds.): Hydrogen as an energy carrier, "Development and scaling up of an electrolytic cell for the anodic oxidation of sulphur dioxide and cathodic
production of hydrogen for the sulphuric acid hybrid cycle: EUR 8651, EUR 8300" D. Reidel Publishing Company, Dordrecht, 1983, pp. 35 – 45.
149. E. Bilgen, C. Bilgen, Int. J. Hydrogen Energy 11 (1986) 241.
150. O. M. Bockris, B. Danapani, D. Cocke, J. Ghoro-ghchian, Int. J. Hydrogen Energy 10 (1985) 179.
151. J. Lede, F. Lapique, J. Villermaux, Int. J. Hydrogen Energy 8 (1983) 675.
152. H. H. G. Jellinek, H. Kachi, Int. J. Hydrogen Energy 9 (1984) 677.
153. J. Lede et al., Int. Chem. Engineering 26 (1986) 647.
154. J. Lede, J. Villermaux, R. Ouzane, M. A. Hossain, R. Ouahes, Int. J. Hydrogen Energy 12 (1987) 3.
155. A. Farhataziz, M. A. Rodgers: Radiation Chemistry, VCH Verlagsgesellschaft, Weinheim, Germany 1987.
156. A. Danno, CEER Chem, Econ. Eng. Rev. (1974) 7.
157. W. Kerr, D. P. Majumdar in T. N. Veziroglu (ed.): Hydrogen Energy, Part A, Plenum Press, New York 1975, pp. 167.
158. J. W. T. Spinks, R. J. Woods: An Introduction to Radiochemistry, J. Wiley & Sons, New York, 1964.
159. I. Tanaka, T. Carrington, H. P. Broida, J. Chem. Phys. 35 (1961) Links
160. K. Kalyanasundaram, M. Grätzel, Angew. Chem. Int. Ed. Engl. 18 (1979) 70.
161. M. Grätzel, Ber. Bunsenges. Phys. Chem. 84 (1980) 981.
162. M. Tomkiewicz, H. Fay, Appl. Phys. 18 (1979) 1.

page 126 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
163. H. Gerischer, W. Mindt, Electrochim. Acta 13 (1968) 1329. Links
164. K. Kalyanasundaram, E. Pelizetti, M. Grätzel, Coord. Chem. Rev. (1989) in print.
165. D. Duhong, N. Serpone, M. Grätzel, Helv. Chim. Acta 67 (1984) 1012.
166. P. Candon, Nature (London) 262 (1976) 715.
167. E. N. Kondratieva, I. N. Gogotov in A. Fiechetr (ed.) Adv. Biochem. Eng. Biotechnol. 28 (1983) 139 – 191.
168. J. M. Olson, J. D. Bernstein, Ind. Eng. Chem. Prod. Res. Dev. 21 (1982) 640.
169. H. Gest, M. D. Kamen, Science (Washington, D.C.) 109 (1949) 558.
170. R. Bollinger, H. Zürrer, R. Bachofen, Appl. Microbiol. Biotechnol. 23 (1985) 147.
171. A. Daday, A. H. Mac Kerras, G. D. Smith, J. Gen. Microbiol 131 (1985) 231.
172. K. K. Rao, L. Rosa, D. O. Hall, Biochem. Biophys. Res. Commun. 68 (1976) 21. Links
173. Techno Japan 19 (1986) no. 7, 104.
174. P. B. Tarman, R. Biljetina: The Steam-Iron-process for Hydrogen Production, 85th National Meeting of the AIChE, Philadelphia 1978.
175. W. H. Seitzer, US 3 928 550, 1975.
176. D. C. Erickson: Hydrogen from Coal via Tin Redox Process, Final Report; Energy Concepts Company, Annapolis, USA 1981.
177. G. Kurz, Proc, Eng. (1976, Oct.) 97.
178. G. Strickland, Int. J. Hydrogen Energy 9 (1984) 759.
179. T. Yamamoto, CEER Chem. Econ. Eng. Rev. 13 (1981) no. 1 – 2, 11.
180. J. C. Amphlett et al., Can. J. Chem. Eng. 59 (1981) 720. Links
181. J. C. Amphlett, M. J. Evans, R. F. Mann, R. D. Weir, Can. J. Chem. Eng. 63 (1985) 605. Links
182. K. Restin, P. Jurkat, H. Hiller, GWF Gas Wasserfach Gas Erdgas 123 (1982) 105.
183. R. W. Bartlett: Preliminary Evaluation of Processes for Recovering Hydrogen from Hydrogen Sulfide, "File No. 79–447", SRI International, July 1979, p. 55.
184. M. E. D. Raymond, Can. Sulfur Symp. Calgary, Canada 1974.
185. Y. Kotera, N. Todo, K. Fukuda, US 3 962 409, 1976.
186. T. Chivers, J. B. Hyne, C. Lau, Int. J. Hydrogen Energy 5 (1980) 499.
187. W. C. Chen, US 4 066 739, 1978.
188. M. Dokiya, H. Fukuda, T. Kameyama, Bull. Chem. Soc. Jpn. 51 (1978) 150. Links
189. I. G. Nixon, Chem. Br. May 5, (1974) 179.
190. T. Chivers, C. Lau, Int. J. Hydrogen Energy 12 (1987) 561.
191. T. Kameyama, Int. J. Hydrogen Energy 8 (1983) 5.
192. S. A. Naman, S. M. Aliwi, K. Al-Emara, Int. J. Hydrogen Energy 11 (1986) 33.
193. H. Kalenda, G. Ruckelhaus, BMFT-Forschungsbericht T 81–1012 (1981).
194. J. Schulze, H. Gaensslen, Chem. Ind. 36 (1984) no. 3, 135 – 140;no. 4, 202 – 208.
195. R. Lohmüller, Chem. Ind. 36 (1984) no. 3, 147 – 151.
196. G. Bauer in [126], p. 195
197. C. R. Baker, NASA Contractor Report 159 163 (1979).
198. A. L. Kohl, F. C. Riesenfeld: Gas Purification, 4th ed., Gulf Publishing, Houston 1985.
199. H. Baldus, K. Baumgärtner, H. Knapp, M. Streich, “Verflüssigung von Gasen,” in Winnacker, Küchler, 4th ed., vol. 2, Hanser Verlag, Wien – München 1983.
200. Landolt-Börnstein, IV, 2 c , Springer Verlag, Berlin – Heidelberg – New York 1980.
201. M. G. Zellner, J. M. Claitor, J. M. Prausnitz, Ind. Eng. Chem. Fundam. (1970) no. 9, 549 – 564.
202. H. C. Rowles, D. M. Nicholas, H. L. Vines, M. F. Hilton, Energy Prog. 61 (1986) no. 3, 25 – 32.
203. W. Förg, Chem. Process Eng. (London) 52 (1971) no. 2, 57 – 68.
204. A. Haslam, P. Brook, H. Isalski, Hydrocarbon Process. 55 (1976) no. 1, 103 – 106.
205. T. Matsuoka, Chem. Age India 30 (1979) no. 3, 119 – 128.
206. W. Scholz, Linde-Ber. Tech. Wiss. 45 (1979) 9 – 17 (LINDE Reports on Science and Technology).
207. Technical Data Book of Petroleum Refining, vol. 2, American Petr. Inst. 1982, pp. 9 – 17.
208. H. Becker et al., Ger. Chem. Eng. 7 (1984) no. 2, 91 – 101.

page 127 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
209. Petrocarbon, GB 10 189.71, 1971 (K. S. Doherty).
210. B. J. Grotz, Hydrocarbon Process. 46 (1967) no. 4, 197 – 202.
211. E. Karwat, Erdöl Kohle Erdgas Petrochem. 17 (1964) 439 – 447.
212. R. Fabian, D. Tillmann, CPP Ed. Europe (1988) no. 3, 35 – 37.
213. J. S. Davis, J. R. Martin, Chem. Eng. Prog. 76 (1980) , no. 2, 72 – 79.
214. S. I. Wang, D. M. Nicholas, S. P. DiMartino, Oil Gas J. 82 (1984) no. 5, 111 – 117. Links
215. With permission of W. Schmid, LINDE AG Haus der Technik Vortrag 1985.
216. W. Scholz, DECHEMA Monogr. 58 (1968) 31.
217. W. Förg, GWF Gas Wasserfach: Gas/Erdgas 113 (1972) 538 – 540.
218. W. Förg, Chem. Anlagen + Verfahren 1970, no. 3, 33 – 44.
219. D. W. Breck: Zeolite, Molecular Sieves, J. Wiley & Sons, New York 1974.
220. D. M. Ruthven: Principles of Adsorption and Adsorption Processes, J. Wiley & Sons, New York 1974.
221. A. Mersmann, U. Münstermann, J. Schadl, Chem. Ing. Tech. 55 (1983) 446 – 458.
222. K. Knoblauch, H. Jüntgen, W. Peters, Erdöl Kohle Erdgas Petrochem. 32 (1979) no. 12, 551 – 556.
223. J. H. Collins, W. C. Miller, J. E. Philcox, Union Carbide Corp., Air Poll. Contr. Assoc., 65th Meeting, Miami Beach 1972.
224. F. G. Wießner, Gas Separation and Purification 2 (1980) no. 3, 115 – 119.
225. E. Richter, K.-B. Harder, K. Knoblauch, H. Jüntgen, Chem. Ing. Tech. 56 (1984) no. 9, 684 – 691.
226. Union Carbide, US 3 986 849, 1975 (A. Fuderer, E. Rudelstorfer).
227. D. H. Werner, G. A. Schlichtharle, Ammonia Plant Saf. 24 (1980) 12 – 18
228. A. Watson, Hydrocarbon Process. 62 (1983) no. 3, 91 – 95.
229. F. J. Brykowski (ed.): Ammonia and Synthesis Gas, Recent Energy-Saving Processes, Noyes Data Corp., Park Ridge, New Jersey 1981, pp. 130 – 147.
230. D. S. Newsome, Catal. Rev. Sci. Eng. 21 (1980) no. 2, 275 – 318.
231. F. J. Brykowski (ed.), [229], pp. 147 – 156.
232. J. M. Moe, Chem. Eng. Prog. 58 (1962) no. 3, 33 – 36 Links
233. F. J. Brykowski (ed.), [229], 156 – 162.
234. P. N. Hawker, Hydrocarbon Process. 61 (1982) no. 4, 183 – 187.
235. E. Kuijpers, Appl. Catal. 25 (1986) 139 – 147.
236. W. van der Wal, E. Kuijpers, P. Berben, J. W. Gens, Proc. Int. Gas Res. Conf., 1986, Toronto, pp. 1134 – 1143.
237. D. W. Allen, W. H. Yen, Chem. Eng. Prog. 69 (1973) no. 1, 75 – 79. Links
238. Engelhard Min. & Chem., US 4 238 468, 1980.
239. J. H. Cramer, L. E. Thompson, T. H. Hsiumg, W. T. McShea, 2nd Latin American Petrochemical Congress, Mexico City, Nov. 1978.
240. J. C. Bonacci, T. G. Otchy, AIChE Symp. (Denver, Co., Aug. 1977).
241. R. Assmann, GWF Gas Wasserfach: Gas/Erdgas 121 (1980) no. 1, 23 – 27.
242. A. L. Kohl, F. C. Riesenfeld (eds.): Gas Purification, 4th ed., Gulf Publishing Company Book Division, Houston – London – Paris – Tokyo 1985.
243. S. A. Newman (ed.): Acid and Sour Gas Treating Processes, Gulf Publishing Company Book Division, Houston – London – Paris – Tokyo 1985.
244. H. W. Schmidt, H. J. Henrici, Chem. Ztg. 96 (1972) 154.
245. R. N. Tennyson, R. P. Schaaf, Oil Gas J. 75 (1978) 78.
246. Proc. Symp. Ammonia from Coal, TVA Muscle Shoals, May 1979.
247. LINDE AG, US 4 332 596, 1982 (G. Ranke).
248. K. Takeuchi, K. Matsumura, K. Yaginuma, Fluid Phase Equilib. 14 (1983) 255.
249. G. Ranke, Linde Reports on Science and Technology 18 (1973) 7.
250. W. Scholz, Linde Reports on Science and Technology 30 (1979) 9.
251. G. E. Pack: Investigation of Sulfur Removal Using Direct Oxidation Processes, 78th Annual Meeting of the Air Pollution Control Association, Detroit June 1985. F. Rosendahl, Erdöl Kohle Erdgas
Petrochem. 34 (1981) 132.
252. M. Heisel, F. J. Marold, Hydrocarbon Process. 66 (1987) 35.
253. Ullmann, 4th ed., 1, 134.

page 128 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
254. Kirk Othmer, 7, 639.
255. G. J. Grashoff, C. E. Pilkington, C. W. Corti, Platinum Met. Rev. 27 (1983) 157.
256. Matthey Metals Ltd, Company Brochure, Diffusion Units.
257. “Special Report: Permeation – Membrane Separation,” Hydrocarbon Process. 62 (1983) no. 8, 43 – 62.
258. S. Egli, A. Ruf, A. Buck, Swiss Chem. 6 (1984) 90.
259. A. K. Fritsche, R. S. Narayan, CEER Chem. Econ. Eng. Rev. 19 (1987) nos. 1, 2, 3 (19).
260. R. Rautenbach, W. Dahm, Chem. Eng. Process. 21 (1987) 141. Links
261. Hydrogen Separation Systems, PRISM Separators, Monsanto, Company Brochure, 1981.
262. D. L. Mac Lean, C. E. Prince, Y. C. Chae, Chem. Eng. Prog. 76 (1980) March, no. 98.
263. R. Barron: Cryogenic Systems, McGraw-Hill, New York 1966.
264. H. W. Wooley, R. B. Scott, F. G. Brickwedde, Journal Res. NBS 41 (1948) 379, RP 1932.
265. J. A. Barclay, W. A. Steyert, Magnetic Refrigeration for Space Applications, Report on a Design Study, Los Alamos Sci. Lab. Rep. LA-8134 (1980).
266. T. Hasimoto, T. Numasawa, M. Shino, T. Okada, Cryogenics 21 (1981) 647.
267. H. Blechschmidt, Kälte 9 (1976) 357.
268. C. L. Newton, Chem. Process. Eng. (London) 48 (1967) no. 12, 51 – 58.
269. G. G. Haselden: “Refrigeration and Liquefaction Cycles,” in G. G. Haselden (ed.): Cryogenic Fundamentals, Academic Press, London 1971, pp. 17 – 89.
270. W. Peschka: Flüssiger Wasserstoff als Energieträger, Springer Verlag, Wien – New York 1984, pp. 33 – 37.
271. H. Gutowski, M. Wanner, Haus Tech. Vortragsveröff. 448 (1982) 60 – 64.
272. C. R. Baker, R. L. Shanner: “A Study of the Efficiency of Hydrogen Liquefaction,” Int. J. Hydrogen Energy 3, (1978) 321 – 334.
273. J. A. Barclay, ASME Paper 81-HT-82, 20th Joint ASME/AiChE National Heat Transfer Conf., Milwaukee, Wisc., Aug 1981.
274. W. A. Steyert, J. Appl. Phys. 49 (1978) 1227 – 1230. Links
275. J. G. Bednorz, K. A. Müller, Z. Phys. B: Condens. Matter Quanta 64 (1986) 189.
276. R. O. Voth, Cryogenics 25 (1985) 511 – 517. Links
277. C. Sindt, Cryogenics 10 (1970) 372. Links
278. W. Tunn, Erdöl Kohle Erdgas Petrochem. 26 (1973) 498.
279. O. Bernauer, C. Halene, J. Less Common Met. 131 (1987) 213.
280. E. Turnwald, LINDE Ber. Techn. Wiss. 36 (1975) 15.
281. W. Peschka, Int. J Hydrogen Energy 7 (1982) 661 – 670.
282. D. T. Wade, L. L. Ansell, W. R. Epperly, Chem. Tech. (Heidelberg) 12 (1982) 242.
283. K. Baumgärtner, R. Ewald, Chem. Ing. Tech. 59 (1987) 296.
284. G. Kaske, H. G. Fasold, P. Schmidt in D. Behrens (ed.): Transport- und Verteilungsleitungen in Wasserstofftechnologie, DECHEMA, Frankfurt/Main 1986, p. 462.
285. W. Diehl, Gas Aktuell 16 (1978) 20.
286. G. G. Leth: Transmission of Gaseous Hydrogen, Messer Griesheim, Company Brochure. F. J. Edeskuty, K. O. Williamson “Liquid Hydrogen Storage and Transmission,” in E. Cox, K. D. Williams Jr.
(eds.): Hydrogen, Its Technology and Implications, vol. 2, CRC Press, Boca Raton 1979.
287. G. Beghi et al.: Hydrogen, Oxygen and Natural Gas by Pipelines, Comparative Transport Costs. Rep. EUR-5103 e, Euratom-C.C.R., Ispra (Italy) 1973.
288. G. D. Brewer (ed): LH2 Airport Requirements Study, NASA CR 2700, Oct. 1976.
289. C. Carpetis in C. J. Winter, J. Nitsch (eds.): Wasserstoff als Energieträger, “Speicherung, Transport und Verteilung von Wasserstoff,” Springer Verlag, Berlin – Heidelberg – New York – Tokyo 1986,
p. 245.
290. C. Carpetis, Int. J. Hydrogen Energy 7 (1982) 191.
291. L. Paterson, Int. J. Hydrogen Energy 8 (1983) 53.
292. J. A. Paivanas, O. P. Roberts, D. I. J. Wang, Adv. Cryog. Eng. 10 (1965) 197.
293. W. Peschka (ed.): Flüssiger Wasserstoff als Energieträger, Springer, Wien – New York 1984.
294. J. Hord, W. R. Parrish in: K. E. Cox, K. D. Williamson (eds.): Hydrogen—Its Technology and Implications, “Economics of Hydrogen,” vol 5, CRC Press, Boca Raton 1979.
295. N. Hallet: Study, Cost and System Analysis of Liquid Hydrogen Production, NASA—Rep. CR 73–226, 1968.
296. H. R. Gray et al.: Potential Structural Materials Problems in a Hydrogen Energy System, NASA TM-X71 752, June 1975.
297. R. A. Oriani, J. P. Hirth, M. Smialowski (eds.): Hydrogen Degradation of Ferrous Alloys, Noyes Publications, Park Ridge, USA 1985.
298. W. Thomson, I. M. Bernstein (eds.): Effect of Hydrogen on Behaviour of Metals, Metal Soc. AIME, New York 1976.

page 129 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
299. C. G. Interrante, G. M. Pressouvre (eds.): Current Solutions to Hydrogen Problems in Steel, ASM Metals Park, Ohio 1982.
300. J. Eastman et al. in M. Bernstein, A. W. Thomson (eds.): Hydrogen Effects in Metals, AIME, New York 1981, p. 397.
301. C. S. Carter: ARPA Handbook on Stress Corrosion and Corrosion Fatigue, “Stress Corrosion Cracking and Corrosion Fatigue of Medium-Strength and High-Strength Steels,” 1977.
302. G. A. Nelson: Hydrogen Plant Steels, Proc. API 29 M (III) (1949) 163.
303. W. A. Bonner, Oil Gas J. 77 (1977, July 25) 172.
304. F. H. Vitovec: The Groth Rate of Fissures During Hydrogen Attack of Steels, Proc. API 44 III 179 (1964).
305. W. Gayk: “Werkstoffauswahl bei Heißgasbeanspruchung,” DECHEMA Monogr. 87 (1980) 73.
306. B. A. Hands (ed.): Cryogenic Engineering, Academic Press, London 1986.
307. D. Ervans, J. T. Morgan in [306] p. 271.
308. R. V. Homsy, C. A. Glaton, Review of Literature on Catalytic Recombination of Hydrogen Oxygen, ORNTL-MT 55, May 1968.
309. M. P. Sherman: The Behaviour of Hydrogen During Accidents in Light Water Reactors, Sandia Laboratories, NUREG/CR-1561, August 1980.
310. H. Karwat: Hydrogen Mitigation, Report at CEC Brussel, Belgium 1982.
311. L. B. Thompson, J. J. Haugh, R. G. Zalosh, Nucl. Saf. 25 (1984) 350.
312. R. Heck, Atomkernenerg. Kerntech. 53 (1988) no. 1, 56.
313. NASA Lewis Research Center, Hydrogen Safety Manual, NASA Tech. Memo TM-X 52 454 (1968).
314. J. Hord, Int. J. Hydrogen Energy 3 (1978) 157.
315. K. M. Kalyanam, D. R. Hay: Safety Guide for Hydrogen, National Research Council Hydrogen Safety Committee, Ottawa 1987.
316. R. L. Custer, R. G. Bright: Fire Detection, State of the Art, NASA CR 134 642, NBS TN 839 June 1974.
317. H. Eichert, M. Fischer, Int. J. Hydrogen Energy 11 (1986) 117.
318. W. Balthasar, J. W. M. van Rijnsoever: Hydrogen is Safe, KTI Publication, Zoetermeer, NL 1981.
319. K. Baumgärtner, R. Ewald, Chem. Ing. Tech. 59 (1987) 296.
320. J. Hord, Int. J. Hydrogen Energy 5 (1980) 579.
321. Occupational Safety and Health Act of 1970, PL 91–596, in United States Statutes at Large, 91st Congress, Second Session 84, Part 2, US Government Printing Office, Washington, DC 1971,
p. 1590.
322. C. Gentry, R. Schott, Anal Chem. 42 (1970) 7.
323. M. Riederer, W. Sawodny, J. Chromatogr. 179 (1979) 337. Links
324. K. Fujita, T. Kwan, Int. Chem. Eng. 8 (1968) 552.
325. C. Laycock, B. Foxall, Iron and Steel Int. (1978, Aug.) 223.
326. J. Plessers, R. Maes, E. Vangelooven, Stahl Eisen 108 (1988) 85.
327. Bruker Analytische Messtechnik, Company Brochure, Bruker Minispec Application Notes, Karlsruhe 1987.
328. J. Schulze, H. Gaenssten, Chem. Ind. 36 (1984) no. 3, 135 – 140; no. 4, 202 – 208. M. A. S. Guth, Int. J. Hydrogen Energy 10, (1985) no. 1, 1 – 10.
329. H. Kalenda, G. Ruckelshauss, (Bundesministerium für Forschung und Technologie) – Forschungsbericht T 81–012 (1981).
330. G. Kaske: Wasserstofftechnologie, DECHEMA Studie, Frankfurt/Main 1986, p. 488.
331. H. G. Eickhoff, D. Martinsen, M. Walbeck, Int. J. Hydrogen Energy 9 (1984) 232.
332. S. Kasper, Int. J. Hydrogen Energy 8 (1983) 183.
333. Chem. Bus. Suppl. of Chem. Market. Reporter 11 (1985) 13.
334. M. A. S. Guth, Int. J. Hydrogen Energy 10 (1985) 1.
335. Y. Osumi, CEER Chem. Econ. Eng. Rev. (1985) June no. 5.
336. E. Wicke, Chem. Ing. Tech. 54 (1982) 41.
337. M. Fischer, G. Kreysa, G. Sandstede, Int. J Hydrogen Energy 12 (1987) 39.
338. W. Van Deelen: Proc. 3rd World Energy Conference, vol. 4, Tokyo 1980, p. 2323.
339. J. G. Santangelo, G. T. Chen: “Hydrogen Recovery with Metal Hydrides,” ACS Symposium on Advances in Hydrogen Manufacture, Los Angeles 1982.
340. T. Grundt, K. Christiansen, Int. J. Hydrogen Energy 7 (1982) 247.
341. L. Green Jr. Int. J. Hydrogen Energy 7 (1981) 335.
342. H. L. Blomelburg, Plant Oper. Prog. 1 (1982) 175.
343. H. F. Mason, R. S. Tolberg, AIChE Symp. Ser. 148 (1975) no. 71, 220.

page 130 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
344. H. J. Satrina (ed.): Hydroprocessing Catalyst for Heavy Oil and Coal, Noyes Data Corp., Park Ridge, NJ, 1982.
345. International Refining Catalyst Compilation 1987, Oil Gas J. 85 (1987, Oct. 5) 41.
346. 1986 Refining Process Handbook, Hydrocarbon Process. 65 (1986) no. 9, 83.
347. R. A. Meyers (ed.): Handbook of Petroleum Refining Processes, McGraw Hill, New York 1986.
348. M. Kuno, Hydrocarbon Process. 60 (1981) 124.
349. B. Schütze, H. Hofmann, Erdöl Kohle Erdgas Petrochem. 36 (1983) 457.
350. W. Krönig, Kohlehydrierung in J. Falbe (ed.): Chemierohstoffe aus Kohle, Thieme Verlag, Stuttgart 1977.
351. E. E. Donath, History of Catalysis in Coal Liquefaction, in J. R. Anderson, M. Bondart (eds.): Catalysis, vol. 3, Springer Verlag, Berlin – Heidelberg – New York 1982, p. 1.
352. P. Nowacki (ed.): Coal Liquefaction Processes, Noyes Data Corp., New Jersey 1979, pp. 76 – 159.
353. S. Ueda, Y. Miki, M. Naka, M. Wada, Status of Synfuel Development in Japan, 19th IECEC, San Francisco, CA, 1984.
354. K. H. van Heek, W. Wanzl, Erdöl Erdgas 105 (1989) 30.
355. R. Fladerer, L. Schrader, Chem. Ing. Tech. 54 (1982) 884.
356. J. Falbe (ed.): Chemierohstoffe aus Kohle, Thieme Verlag, Stuttgart 1977.
357. J. Fahey, “Industrial Chemicals via C1-Processes,” ACS Symp. Ser. 328 (1986) .
358. J. Falbe (ed.): New Syntheses with Carbon Monoxide, Springer Verlag, Berlin – Heidelberg – New York 1980.
359. F. Asinger (ed.): Methanol, Chemie- und Energierohstoff, Springer Verlag, Berlin – Heidelberg – New York – London – Paris – Tokyo 1986.
360. C. D. Frohning et al. in [357], Chap. 8, pp. 219 – 299.
361. Ch. Masters, Adv. Organomet. Chem. 17 (1979) 61 – 103. Links
362. M. A. Vanice, Catal. Rev. Sci. Eng. 14 (1976) 153.
363. H. C. Anderson, J. L. Wiley, A. Newall, Bibliography of the Fischer – Tropsch – Syntheses, Bull. 544, Bureau of Mines, United States Goverment Printing Office, Washington 1954.
364. M. E. Dry, CHEMTECH 12 (1982) 744.
365. V. U. S. Rad, R. J. Germly, Hydrocarbon Process. 59 (1980) no. 11, 139.
366. L. König, J. Gaube, Chem. Ing. Tech. 55 (1983) 14.
367. C. D. Frohning, Polymethylene Synthesis in J. Falbe (ed.): New Syntheses with Carbon Monoxide, Springer Verlag, Berlin – Heidelberg – New York 1980.
368. F. W. Koeller, H. Roberts, B. Britz, Hydrocarbon Process. 53 (1974) no. 4, 69.
369. B. Cornils: “Hydroformylation, Oxo Synthesis, Roelen Reaction” in J. Falbe (ed.): New Syntheses with Carbon Monoxide, Springer Verlag, Berlin – Heidelberg – New York 1980.
370. H. Bahrmann, B. Cornils, Chem. Ztg. 104 (1980) 39.
371. R. L. Augustine, “Organic Functional Group Hydrogenation,” Catal. Rev. Sci. Eng. 13 (1976) 285.
372. M. Freifelder (ed.): Practical Catalytic Hydrogenation, Wiley Interscience, New York 1971.
373. M. Freifelder (ed.): Catalytic Hydrogenation in Organic Synthesis, Procedures and Commentary, Wiley Intercience, New York 1978.
374. R. J. Peterson (ed.): Hydrogenation Catalysts, Noyes Data Corporation, Park Ridge, NJ 1977.
375. R. Albrecht, A. Strätz, G. Vollheim, Chem. Ing. Tech. 52 (1980) 582.
376. R. Steffen, H. B. Lüngen, Stahl Eisen 108 (1988) 339.
377. E. Mazank, A. Brachany, C. Bryk, Stahl Eisen 102 (1982) 3.
378. F. Habashi, R. Dugdale, N. Nagmori, Metall (Berlin) 28 (1974) 1051.
379. Rockwell Int. Corp., DE-OS 30 428 878, 1981 (J. Friedmann et al.).
380. W. Pötke, R. Jeschdar, G. Kehse, Stahl Eisen 108 (1988) 581.
381. U. J. Gährs, P. Rutten, G. Schnorr, Gas Aktuell, Messer Griesheim brochure, Frankfurt 1987.
382. U. Gieger et al., Cryogenics 11 (1976) 451.
383. P. Hoffmann: The Forever Fuel, Westview Press, Boulder, CO, 1981.
384. G. M. Woodwell, Sci. Am. 238 (1978) no. 1, 34 – 44. Links
385. C. D. Keeling et al.: “Atmospheric Carbon Dioxide Variations. Hawaii,” Tellus 28 (1976) no. 6, 538 – 551.
386. J. O. M. Bockris: Energy: The Solar Hydrogen Alternative, Wiley & Sons, New York 1980.
387. W. Bach et al. (eds.): Renewable Energy Prospects, Pergamon Press, Oxford 1980.
388. H. E. Landsberg, Sol. Energy 5 (1961) 95 – 98.
389. C. Voigt in C. J. Winter, J. Nitsch (eds.): Wasserstoff als Energieträger, Springer Verlag, Berlin 1986, p. 301.

page 131 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
390. H. Tributsch in H. Scheer (ed.): Die gespeicherte Sonne, Piper, München 1987, pp. 119 – 155.
391. J. Cotillon, Houille Blanche 37 (1982) 381 – 390.
392. C. Voigt in [389], p. 305.
393. M. Sappa, Proc. 3rd World Hydrogen Energy Conf., vol. 3, Tokyo 1980, pp. 1373 – 1406.
394. R. L. Le Roy: The Metallurgical Society of CIM, annu. vols. 1 – 10, 1978.
395. F. Conrad, Energie 40 (1988) no. 6, 46 – 53.
396. H. Wendt, Chem. Ing. Tech. 56 (1984) 265 – 272.
397. W. Schnurnberger, W. Seeger, H. Steeb in [389], pp. 205 – 243.
398. W. Doenitz, R. Schmidberger, Int. J. Hydrogen Energy 7 (1982) 321 – 330.
399. C. Carpetis, W. Peschka, Int. J. Hydrogen Energy 5 (1980) 539.
400. J. S. Noh, R. K. Agarwal, J. A. Schwarz, Int. J. Hydrogen Energy 12 (1987) 693.
401. G. Alefeld, J. Völkl (eds.): Hydrogen in Metals, vol. I and vol. II. Topics and Applied Physics, vols. 28, 29, Springer Verlag, Berlin – Heidelberg – New York 1978.
402. G. G. Libowitz, G. D. Sandrock (eds.): Proceedings and Applications of Metal Hydrides, vols. 1 and 2, Elsevier, Amsterdam 1980.
403. H. Wenzl, Int. Met. Rev. 27 (1982) 140.
404. G. D. Sandrock, S. Suda, *** Applications in L. Schlapbach (ed.): Hydrogen in Intermetallic Compounds, Springer Verlag, Berlin – Heidelberg – New York 1986.
405. F. A. Kuijpers, H. H. VanMal, J. Less-Common Met. 23 (1971) 395.
406. P. D. Goodell, G. D. Sandrock, E. L. Huston, J. Less-Common Met. 73 (1980) 135.
407. M. J. Song, M. Pezat, B. Darrier, P. Hagenmüller, J. Solid State Chem. 56 (1985) 191. Links
408. O. Bernauer, C. Halene, J. Less-Common Met. 131 (1987) 213.
409. HWT Gesellschaft für Hydrid- und Wasserstofftechnik mbH, Mülheim/Ruhr, company brochure 1986.
410. H. Buchner, Int J. Hydrogen Energy 3 (1981) 259.
411. F. B. Hill, Y. W. Wong, Y. N. I. Chan, AIChe J. 28 (1982) 1. Links
412. J. G. Santangelo, G. T. Chen, Hydrogen Recovery with Metal Hydrides, ACS Symposium on Advances in Hydrogen Manufacture, Las Vegas 1982.
413. E. L. Houston, Development of a Commercial Metal Hydride Process for Hydrogen Recovery, Final Report vols. 1 and 2, MPD Technology Corp., Wyckoff, New Jersey, 1984.
414. L. Schlapbach, M. Groll, Metal Hydrides for Energy Technologies, IEA Report Task VII, Storage, Conversion and Safety, March 1986.
415. Ergenics Inc., USA; company brochure.
416. M. Taube et al., Int. J. Hydrogen Energy 10 (1985) 595.
417. G. Giacomazzi, CEC Joint Research Centre, Report EUR-11 330 – EN, Ispra (Italy) 1986.
418. K. Hoyermann, H. Janaer, H. G. Wagner, Ber. Bunsenges. Phys. Chem. 79 (1975) 858.
419. J. Pangborn, M. C. Scott, Domestic Uses of Hydrogen in K. E. Cox, K. D. Williamson, Jr. (eds.): Hydrogen: Its Technology and Implications, CRC-Press, Boca Raton, Florida 1979, p. 151.
420. Einsatzmöglichkeiten neuer Energiesysteme, Programmstudie des BMFT, part III: Wasserstoff, Bonn 1975.
421. M. Haruta, H. Sano, Int. J. Hydrogen Energy 7 (1982) 729, 737, 801.
422. J. L. Sloop, Liquid Hydrogen as a Propulsion Fuel, The NASA History Series, BASA-SP-4404, TL 785 S 58, 1978.
423. H. Dederra, E. Kirner, MBB: Vulcain HM60-Thrust Chamber Critical Development Items, 15th ISTS-Congress, Tokyo, May 1986.
424. D. Brewer, Astronaut. Aeronaut. 12 (1974) 40.
425. P. Hofmann The forever Fuel. Hydrogen as Aircraft Fuel, p. 139, Westview Press, Boulder, CO, 1982.
426. International Symposium (DGLR mit DFVLR) Hydrogen in Air Transportation, Sept. 11th to 14th, 1979, Stuttgart.
427. G. D. Brewer, Int. J. Hydrogen Energy 7 (1982) 21.
428. E. Reshoto, J. Aircr. 16 (1979) 584.
429. G. D. Brewer (ed.): NASA Contractor Report CR 2700, Washington, D.C., 1976.
430. H. P. Alder, Int. J. Hydrogen Energy 12 (1987) 571.
431. W. Peschka, Flüssiger Wasserstoff als Energieträger, Springer Verlag, Wien – New York 1984, pp. 129 – 184.
432. H. Buchner, Int. J. Hydrogen Energy 9 (1984) 501.
433. P. Hofmann: The Forever Fuel: The Story of Hydrogen, Westview Press, Boulder, CO, 1982, pp. 105 – 138.
434. BMW AG, Hydrogen Vehicle, company brochure, Munich 1986.
435. BMFT Studie, Neuen Kraftstoffen auf der Spur, Bonn 1974.

page 132 of 133

Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
436. L. A. Sinclair, J. S. Wallace, Int. J. Hydrogen Energy 12 (1984) 123.
437. H. S. Homan, P. C. T. De Boer, W. J. McLean, Int. J. Hydrogen Energy 8 (1983) 131.
438. K. Feucht, R. Povel, W. Gelse, “Wasserstoff-Energietechnik,” VDI Ber. 602 (1987) 185.
439. G. Withalm, W. Gelse, The Mercedes Benz Hydrogen Engine for Application in a Vehicle Fleet, VI World Hydrogen Energy Conference, Vienna 1986.
440. W. Strobl, W. Peschka, “Wasserstoff-Energietechnik”, VDI-Ber. 602 (1987) 203.
441. H. J. Sternfed, H. Wojkowsky, W. Schnurrenberger, Brennst. Wärme Kraft 34 (1982) 351, 401.
442. H. J. Sternfeld, K. Wolfmüller, VAB Kraftwerkstech. 66 (1986) 793.
443. H. Liebhafsky, E. J. Cairns: Fuel Cells and Fuel Cell Batteries, Wiley, New York 1968.
444. K. K. Ushiba, CHEMTECH 14 (1984) 300.
445. H. Wendt, W. Jenseit, Chem. Ing. Tech. 60 (1988) 180.
446. K. Kordesch, J. C. T. Olivera, Int. J. Hydrogen Energy 13 (1988) 411.
447. W. Dönitz, State of the Art and Future Developments of Solid Oxide Electrolyte Fuel Cells, Proceedings of the UNESCO Workshop on Fuel Cells, June 1985.
448. A. P. Fickett, Int. J. Hydrogen Energy 8 (1983) 617.
449. S. Knable in E. A. Gillis, (ed.): Fuel Cell Seminar 1986, Program and Abstracts, Tucson Ariz., 1986, pp. 27.
450. N. Itoh, Science & Technology in Japan 5 (1986) no. 18, 16.
451. N. Getoff: Wasserstoff als Energieträger, Springer Verlag, Wien – New York 1977.
452. D. Behrens (ed.): Wasserstofftechnologie. Perspektiven für Forschung und Entwicklung, DECHEMA, Frankfurt 1986.
453. K. E. Cox, K. D. Williamson (eds.): Hydrogen, its Technology and Implications, vols. I – V, CRS Press, 1979.
454. J. Nitsch, C. Voigt in [389], pp. 288 – 310
455. R. Aureille, J. Pottier, Proc. 5th World Hydrogen Energy Conf., vol. 1, Toronto 1984, pp. 279 – 288.
456. A. G. Barnstaple, A. J. Petrella, Report no. 81 105 and 82 452, Ontario Hydro, Toronto 1981, 1982.
457. G. C. Crawford, A. K. Stuart, Int. J. Hydrogen Energy 9 (1984) no. 7, 619 – 625.
458. R. H. Shannon, R. D. Richardson, Int. J. Hydrogen Energy 8 (1983) no. 10, 783 – 792.
459. A. J. Petrella, D. Chisholm, A. G. Barnstaple, Hydrogen study, vol. 4, Report no. 82 645, Ontario Hydro, Toronto 1982.
460. J. Nitsch, C. Voigt in [389], p. 358.
461. T. Donakowski, C. Blazek, M. Novil, E. Daniels, Int. J. Hydrogen Energy 9 (1984) no. 7, 613 – 618.
462. Hysolar—Solar Hydrogen Production and Utilization, Projektvereinbarung Riad/Bonn, March 1986.
463. R. Meggle, DECHEMA Monogr. 112 (1988) 325 – 534.
464. J. Gretz, J. P. Baselt, O. Ullmann, H. Wendt: “The 100 MW Euro-Quebec Hydro-Hydrogen Pilot Project,” VDI-Rep. 725 (1989) 81 – 91.
465. C. Marchetti, Int. J. Hydrogen Energy 10 (1985) no. 4, 215 – 219.
466. Solare Wasserstoffenergiewirtschaft, BMFT-Ausschuß, Bonn, Feb. 1988.
467. International Energy Agency, Program of Work 1986 – 1989, Deutsche Forschungs- und Versuchsanstalt für Luft- und Raumfahrt, Köln 1986.
468. New Uses for H2in Canadian Energy Systems, International Energy Agency, Task VII reports, Nov. 1985.
469. The Sunshine Project, Agenca of Ind. Sci. Tech. of MITI, Tokyo 1987. 1986 Annual Summary of Hydrogen Energy R + D, Japan Industrial Technology Association, Tokyo 1987.
[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 133 of 133


Hydrogenation and Dehydrogenation
Heterogeneous Catalysis and Solid Catalysts is a separate keyword.
Paul N. Rylander, (Engelhard Corp.), Iselin, New Jersey 08830, United States
1. Hydrogenation . . . . . . . . . . . . . .
1.1. Hydrogen Availability and Mass
Transport . . . . . . . . . . . . . . . . .
1.2. Catalysts . . . . . . . . . . . . . . . . . .
1.3. Techniques . . . . . . . . . . . . . . . .
1.3.1. Slurry Reactors . . . . . . . . . . . . . .
1.3.2. Trickle-Bed Reactors . . . . . . . . . .
1.3.3. Bubble Column and Ebulliated Bed
Reactors . . . . . . . . . . . . . . . . . .
1. Hydrogenation
Catalytic hydrogenation refers to the addition of
hydrogen to an organic molecule in the presence
of a catalyst. If the molecule undergoes cleav-
age, the reaction is referred to as hydrogenolysis.
These reactions are used to produce a variety of
both bulk and fine organic chemicals. In addi-
tion, hydrogenation is often used in various pu-
rification processes (e.g., selective reduction of
trace amounts of acetylene in an ethylene stream
or traces of butadiene in a butene stream, and re-
moval of unwanted oxygen, hydrogen, or carbon
oxides from a variety of systems).
A large number of hydrogenation catalysts
are now available, and the scope of hydrogen-
ation has become very large. Most functional
groups can be reduced readily, often under mild
conditions, and frequently with high chemo-,
regio-, and stereoselectivity. Monographs detail-
ing the scope and use of hydrogenation have
been written by Augustine [1], Freifelder
[2], Rylander [3–5], and Zymalkowski [6].
Hydrogenation is an exothermic reaction and
the equilibrium usually lies far toward the hy-
drogenated product under most operating tem-
peratures:
As the temperature increases, the equilibrium
shifts toward the left, and the reverse reaction,
dehydrogenation, can itself be a useful synthetic
process (see Chap. 2). Industrial hydrogenation

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a13 487
Hydrogenation and Dehydrogenation 1
1 1.4. Hydrogenation of Individual Com-
pounds . . . . . . . . . . . . . . . . . . . 3
2 2. Dehydrogenation . . . . . . . . . . . . 9
2 2.1. Styrene . . . . . . . . . . . . . . . . . . . 9
3 2.2. Dehydrogenation of n-Alkanes . . . 10
3 2.3. Catalytic Reforming of Hydrocar-
3 bons . . . . . . . . . . . . . . . . . . . . . 11
2.4. Liquid-Phase Dehydrogenation . . . 11
3 3. References . . . . . . . . . . . . . . . . . 12
requires a means of limiting the temperature in-
crease arising from the exothermic reaction; a
higher temperature usually has a favorable ef-
fect on reaction rate, but may affect selectivity
and catalyst life adversely.
Hydrogenation is influenced by a number
of factors, including catalyst, solvent, substrate
purity, and operating conditions. Temperature,
pressure, agitation, and catalyst loading can each
influence both the rate and the selectivity of hy-
drogenation. In general, as temperature, pres-
sure, agitation, and catalyst loading increase, the
rate of hydrogenation increases until a limit is
reached. The influence of operating variables on
selectivity is less straightforward but can often
be linked to the effect of these and other variables
on hydrogen availability at the catalyst surface.
1.1. Hydrogen Availability and Mass
Transport
Hydrogenation is carried out in both the liquid
and the gas phases, but liquid-phase reactions are
much more common. In liquid-phase systems,
transport of hydrogen to the catalyst surface is
frequently the rate-limiting step. To react with
the substrate, hydrogen must move from the gas
phase into the liquid and then to the solid cata-
lyst and into its porous structure. This net move-
ment of hydrogen to the catalyst is the result
of concentration gradients, which develop when
hydrogen is consumed by the catalytic reaction.
2 Hydrogenation and Dehydrogenation
The concentration of hydrogen at the catalyst
surface can vary widely, depending on the ratio
of the rate of hydrogen consumption to the rate
of hydrogen supply [7].
This rate ratio can have an important influ-
ence on selectivity. It is, for instance, an im-
portant factor in controlling selective removal
of multiple unsaturation and the extent of cis –
trans isomerization during hydrogenation of nat-
ural oils to make edible products.
1.2. Catalysts
Hydrogenation catalysts are of two types, het-
erogeneous and homogeneous. Heterogeneous
catalysts are solids that form a distinct phase
in the gas or liquid environment. Homogeneous
catalysts dissolve in the liquid medium, forming
only a single phase; a well-known catalyst of this
type is the Wilkinson catalyst, (Ph3 P)3 RhCl.
Homogeneous Catalysts. Homogeneous
hydrogenation catalysts are used infrequently
in industry. They are difficult to separate from
the reaction media and are not employed if
a suitable heterogeneous catalyst is available.
Nonetheless, they are useful in a variety of
hydrogenation processes that are not possible
with heterogeneous catalysts [8], [9]. Expanded
use of homogeneously catalyzed hydrogenation
awaits only industrial demand for its products.
A noteworthy industrial use of a homoge-
neous catalyst is the asymmetric hydrogenation
of an achiral olefinic precusor of l-Dopa (3,4-di-
hydroxyphenylalanine, used in the treatment of
Parkinson’s disease; see → Parkinsonism Treat-
ment) over a complex rhodium catalyst carrying
chiral ligands. Optical yields are nearly quanti-
tative.
Heterogeneous Catalysts. Heterogeneous
catalysts can be divided into two types: those
used in fixed-bed processing, in which the cat-
alyst is stationary and the reactants pass over
it, and those used in fluidized-bed or slurry
processing. The former have a relatively large
particle size, 0.79 – 6.35 mm, in various forms.
Slurry or fluidized-bed catalysts are fine pow-
ders that can be suspended readily in a liquid or
gas.
Many metals and metal oxides have general
hydrogenation activity, but marked differences
exist among them. Nickel, copper, cobalt, chro-
mium, zinc, iron, and the platinum group (plat-
inum, palladium, rhodium, and ruthenium) are
among the elements frequently found in com-
mercial hydrogenation catalysts. Combinations
of these or other elements are sometimes used
to confer additional activity, selectivity, stability,
or life.
Palladium, platinum, rhodium, and ruthe-
nium make exceptionally active hydrogenation
catalysts and are used frequently when equip-
ment or stability limitations permit operation
only under mild conditions. They can effectively
reduce most functional groups. Nickel is used to
reduce a variety of functional groups, usually
under more vigorous conditions than the plat-
inum group. Copper chromite is used mainly to
reduce esters to alcohols. Cobalt or iron, in the
presence of ammonia, is an effective catalyst for
the reduction of nitriles to primary amines.
Catalyst Selectivity. In series reactions of
the type A → B → C, both mechanistic and ther-
modynamic selectivity are involved in the for-
mation of B. These selectivities can vary widely
with the catalyst used. Thermodynamic selec-
tivity, which is related to the relative free en-
ergies of adsorption of A and B, refers to the
competition of A and B for catalyst sites. The
tendency for B to occupy a catalyst site rela-
tive to A increases as B accumulates and as A
is depleted. High thermodynamic selectivity de-
mands that A be adsorbed much more strongly
than B and that the rate of displacement of B
by A be high relative to the rate of hydrogen-
ation of B to C. Thermodynamic selectivity is
affected by depletion of A in the vicinity of the
catalyst because of intra- and interparticle dif-
fusion resistance. Mechanistic selectivity refers
to the extent to which C is formed from A with-
out the formation of desorbed B; it is related to
the relative rates of A → B and A → C in the
equation B ← A → C. Either thermodynamic or
mechanistic selectivity may be more important,
but both must be high if B is to be obtained in
good yield.
 Hydrogenation and Dehydrogenation
1.3. Techniques
All hydrogenation reactors serve to bring hy-
drogen, catalyst, and substrate together in the
absence of oxygen. Liquid-phase substrates may
be hydrogenated neat or in the presence of a suit-
able solvent; use of a solvent is one of the best
available means for markedly altering selectiv-
ity. Solvents also help to moderate the heat of
reaction, aid in catalyst handling and recovery,
and permit the reduction of solid substrates. A
convenient solvent may be the product itself or
the solvent used in a previous or subsequent step.
Militating against the use of solvent is dimin-
ished space – time yield and the cost of solvent
and of its removal.
1.3.1. Slurry Reactors
Hydrogenation is exothermic, and some means
of removing the reaction heat must be provided
to avoid excessive temperature. In liquid-phase
slurry reactions, cooling is usually accomplished
by coils within the reactor. The reaction mass can
also be cooled by circulating the mixture over
external cooling elements. The latter save reac-
tor space and permit the reactor to be cleaned
more easily.
Slurry reactors must be provided with some
sort of agitation to aid in the rapid transport
of hydrogen from the gas phase to the cata-
lyst suspended in the liquid medium. Agitation
also helps in the removal of heat, and agitators
can be designed to favor either hydrogen trans-
port or heat removal, as required. Externally
cooled systems can be engineered to return the
reaction slurry to the reactor by a venturi tube
through which hydrogen is drawn into the pass-
ing stream. Such a system gives efficient mixing,
and much of the hydrogenation is thought to take
place within the nozzle itself. Alternatively, the
slurry can be returned to the reactor through a
shower-type arrangement in which substrate and
catalyst particles are sprayed into hydrogen in
the head space.
1.3.2. Trickle-Bed Reactors
Continuous hydrogenation is carried out com-
monly in trickle-bed reactors (columns packed
3
with catalyst over which substrate and hydrogen
pass concurrently downward). This type of reac-
tor is used extensively in the petroleum industry
for various reactions such as hydrocracking and
hydrodesulfurization. The residence time distri-
bution is close to that of plug flow, which per-
mits high conversion in a single vessel. Oper-
ation is continuous, handling is minimal, and
catalyst separation from product is trivial. Heat
removal, however, is a major problem unless
small-diameter tubes are used. At times, reac-
tors are placed in series, with hydrogen injected
between beds for temperature control. Beds are
usually of uneven height, becoming longer as
the reactant is depleted [10].
1.3.3. Bubble Column and Ebulliated Bed
Reactors
Bubble columns resemble trickle-bed reactors
except for the direction of flow. In bubble col-
umn reactors, flow is upward; liquid fills the
void space, and the catalyst is completely wet-
ted. Ebulliated bed reactors are similar except
that the catalyst is fluidized by high gas and liq-
uid flow rates. This type of reactor is used in
H-oil and H-coal processes (see also → Coal
Liquefaction, Chap. 3.9.5.).
1.4. Hydrogenation of Individual
Compounds
Alkynes. Alkynes are adsorbed on a cata-
lyst surface much more strongly than the cor-
responding alkenes, thus permitting selective
hydrogenation of traces of alkynes in massive
amounts of alkenes. Hydrogenation of trace
alkynes, usually over a low percentage of palla-
dium on an alumina support with restricted hy-
drogen supply, is used widely to purify various
alkene-containing streams.
The strong adsorption of alkynes also per-
mits selective reduction to the cis alkenes in high
yield, a reaction used in the production of a va-
riety of fine chemicals, such as vitamins A and
K, prostaglandins, and pheromones. The Lind-
lar catalyst (Pd – Pb on CaCO3 ), with or without
additional modifiers, is often used for this selec-
tive reduction. Ideally, the reaction comes to a
4 Hydrogenation and Dehydrogenation
virtual halt after absorption of 1 mol of hydro-
gen, but terminating the reaction after consump-
tion of the stoichiometric amount of hydrogen
may be necessary. The largest volume hydrogen-
ations are of alkynols and alkynediols. Butane-
1,4-diol [110-63-4], derived from acetylene, has
long been made by hydrogenation of butyne-
diol over a base-metal catalyst such as nickel –
copper – manganese on silica (→ Butanediols,
Butenediol, and Butynediol, Chap. 1.3.). Nickel,
the major component, is an effective catalyst for
hydrogenation of both acetylenic and olefinic
bonds. At the butenediol stage, some aldehyde is
formed by rearrangement; this is removed effec-
tively by copper. A small amount of manganese
stabilizes the structure. The active binary cata-
lyst palladium – ruthenium on carbon behaves in
similar fashion.
Alkenes. Hydrogenation of olefins has ex-
tensive industrial application. Dripolene, or
pyrolysis gasoline, is stabilized against oxida-
tion by selective reduction of dienes in a com-
plex mixture of olefins and aromatics. The pre-
ferred catalyst consists of pellets of palladium
on alumina. If the aim of reduction is to produce
gasoline, olefins are preserved to boost the oc-
tane rating, but if the product is to be used as a
source of pure aromatics, olefins are removed in
a second stage over a base-metal catalyst, such
as nickel or cobalt – molybdenum. Two steps are
used for complete olefin hydrogenation, because
the more vigorous conditions required for com-
plete reduction cause severe fouling of the cata-
lyst by polymerization of the diene.
The largest single application of olefin hy-
drogenation is the partial hydrogenation of
natural oils to produce margarine, shortening,
toppings, salad oil, and other edible products
(→ Margarines and Shortenings). A huge mar-
ket also exists for products derived by more com-
plete hydrogenation; the problem of selectivity
described below is not present.
Edible products are derived by partial hydro-
genation of various oils (e.g., cottonseed, soy-
bean, sunflower, and rapeseed oil) all of which
are glycerides of long-chain saturated and cis-
unsaturated fatty acids (→ Fats and Fatty Oils,
Chap. 4.3.). Partial hydrogenation of these prod-
ucts improves oxidative stability and alters plas-
tic properties suitably. Oxidative instability is
caused primarily by the presence of various ho-
moconjugated dienes and trienes, and an aim of
hydrogenation is preferential removal of multi-
ple unsaturation. This task is facilitated by the
tendency of hydrogenation catalysts to isomer-
ize homoconjugated into conjugated olefins; the
latter, relative to monoenes, are strongly ad-
sorbed on the catalyst. Catalysts differ widely
in their ability to promote isomerization, and
a close parallel exists between this ability and
the selective removal of polyunsaturation. As a
consequence of double-bond migration, isomer-
ization of the naturally occurring all-cis olefins
into trans olefins may or may not occur, depend-
ing on the conformation of the chain at the time
of migration. Many catalysts have been stud-
ied for these reactions, but industrial practice
overwhelmingly favors some form of nickel. Hy-
drogenation of edible oils is carried out almost
entirely in batchwise slurry operation at 140 –
190 ◦ C and 0.2 – 0.3 MPa.
Notwithstanding the huge volume of prod-
uct, processing is largely carried out in a slurry
batch mode. Continuous fixed-bed operation is
made difficult by intermittent production, vari-
able feedstocks, and the problem of adjusting
conditions to compensate for declining catalyst
activity.
Aromatic and Heteroaromatic Rings.
Compounds with aromatic or heterocyclic rings
are hydrogenated readily and usually without
difficulty if other reducible groups are absent.
Cyclohexane [110-82-7] is obtained mainly
by hydrogenation of sulfur-free benzene over
supported fixed-bed nickel or platinum cata-
lysts (170 – 230 ◦ C, 2.5 MPa, → Cyclohexane,
Chap. 4.1.). The reaction is reversed at higher
temperature, which sets an upper limit to a per-
missible operating temperature of ca. 230 ◦ C.
With catalysts of controlled acidity, the prod-
uct can be obtained virtually free of benzene
and methylcyclopentane. Tetralin [119-64-2] is
produced by hydrogenation of naphthalene over
nickel sulfide or nickel – molybdenum at 400 ◦ C
and 2 – 6 MPa; a higher pressure gives decalin.
Hexahydrobenzoic acid [98-89-5], an inter-
mediate for caprolactam (nylon 6 monomer), is
made by hydrogenation of benzoic acid over pal-
ladium on carbon powder by using the product
as solvent to help prevent excessive temperature
(→ Caprolactam, Chap. 4.3.). When only base-
metal catalysts were available, this was consid-
 Hydrogenation and Dehydrogenation
ered a very difficult reduction; base-metal cat-
alysts are attacked by the carboxylic acid func-
tion.
An important monomer 1,4-bis(hydroxyme-
thyl)cyclohexane is made by two-stage hydro-
genation of dimethyl terephthalate. The first
stage (160 – 180 ◦ C, 30 MPa), carried out in a
fixed-bed reactor over supported palladium, pro-
duces dimethyl hexahydroterephthalate. Part of
this product is recycled to moderate the temper-
ature increase resulting from the highly exother-
mic hydrogenation (197.6 kJ/mol). The second
stage uses a copper chromite catalyst under sim-
ilar conditions.
Perhydrogenated rosins are made by hydro-
genation of rosin. Rosins contain dehydroabietic
acid and similar compounds, which are difficult
to reduce. The preferred catalyst is palladium,
because it causes the least amount of unwanted
decarbonylation and decarboxylation, reactions
that lead to catalyst inhibition through compet-
ing adsorption with carbon monoxide for cata-
lyst sites [11].
Cyclohexanone [108-94-1] is produced by
hydrogenation of phenol either in the vapor
phase or in a liquid slurry over supported palla-
dium catalysts (→ Cyclohexanol and Cyclohex-
anone, Chap. 3.1.). In the latter, purified phenol
and catalyst slurry enter the first of several re-
actors and are exposed to synthesis gas (75 %
N2 , 25 % H2 ). A controlled pressure drop bet-
ween the reactors ensures forward flow without
the need for pumps. The product with the low-
est boiling point, cyclohexanone, is removed as
vapor in the gas flow. In the last reactor, conver-
sion of phenol is about 95 % complete. Slurry
emerging from the last reactor is centrifuged,
and recovered catalyst is returned for recycling.
5
Reactor temperature is held between 155 and
170 ◦ C, pressure at 0.5 – 1.4 MPa [12].
Carbonyl Compounds. Aldehydes and ke-
tones behave similarly on hydrogenation, and
the same order of catalyst activity usually ap-
plies to both functional groups. The difference
between aromatic and aliphatic carbonyls is
marked, however, and the preferred catalysts dif-
fer.
Aliphatic carbonyl compounds are reduced
to the corresponding alcohols with little danger
of overhydrogenation. Alcohols are in fact an
excellent solvent for hydrogenation, even under
vigorous conditions.
Hydrogenation of mesityl oxide [141-79-7]
is a good example. Formed by aldol condensa-
tion of acetone, mesityl oxide can be quantita-
tively reduced to methyl isobutyl ketone (MIBK)
over palladium; if desired, MIBK can be reduced
further to 4-methylpentan-2-ol over base metals
such as nickel or copper.
The first two steps can be combined by em-
ploying a dual-functional catalyst, i.e., one capa-
ble of catalyzing both condensation and hydro-
genation; examples include palladium on zirco-
nium phosphate or on a zeolite.
2-Ethylhexanol [104-76-7] an important
chemical widely used in plasticizers, is also de-
rived from aldol condensation, followed by hy-
drogenation of the resulting 2-ethyl-2-hexenal.
Reduction is usually carried out in two stages
over supported nickel, with the carbon – carbon
double bond reduced first, followed by the alde-
hyde. One-stage reduction is also performed, but
6 Hydrogenation and Dehydrogenation
the two-stage process makes temperature control
easier and yields better product quality (see also
→ 2-Ethylhexanol, Chap. 3.).
Sorbitol [50-70-4] is produced on a large
scale by hydrogenation of d-glucose over a
nickel or ruthenium catalyst. Sorbitol is used in
a variety of foods, pharmaceuticals, and cosmet-
ics, as well as in the production of polyurethanes,
resins, and vitamin C.
This hydrogenation is not without complica-
tions. The temperature must be high enough to
give acceptable rates, but not so high as to pro-
duce various byproducts resulting mainly from
cyclization. The process must be operated within
a narrow pH range. Nickel catalysts are damaged
if the solution becomes too acidic. If the solu-
tion becomes too alkaline, gluconic acid is pro-
duced by the Cannizzaro reaction, and mannitol
is formed through epimerization at the α-carbon
[13].
Aromatic carbonyl compounds are reduced
easily, and either reduction to the carbinol or
loss of the hydroxyl group can be achieved at
will.
Both reactions provide a rich source of
products and intermediates in the synthesis of
fine chemicals. Palladium is often the catalyst
of choice for either hydrogenation or hydro-
genolysis. The same catalyst can be used in both
steps; in neutral media, both the mechanistic and
the thermodynamic selectivities toward the al-
cohol are very high. The reactions are usually
conducted in the presence of small amounts of
strong acid when the hydrogenolysis product is
desired.
Nitro Compounds. Most nitro compounds
are reduced in slurry reactors, although nitro-
benzene itself is converted to aniline [62-53-3]
in vapor-phase fixed-bed reactors as well as in
fluidized beds. Catalysts for vapor-phase hydro-
genation vary with the producer and include sul-
fided nickel chromates, nickel – aluminum, sil-
ver – manganese oxide, and copper – silica. The
reaction temperature is high (300 – 475 ◦ C), and
the pressure is 0.5 – 1.5 MPa [14]. Aniline is
also produced in continuous slurry reactors by
using palladium on carbon black promoted by
platinum and iron, palladium on activated car-
bon or nickel on silica [14] (see also → Aniline,
Chap. 3.2.).
The course of hydrogenation of nitroaromatic
compounds is complex, and a variety of products
can be obtained. Partial hydrogenation yields
phenylhydroxylamines. Reaction is best under
mild conditions in neutral media with platinum
on carbon, inhibited by addition of dimethyl
sulfoxide to prevent overhydrogenation [15].
Phenylhydroxylamines are key intermedi-
ates in several important processes. In aqueous
acid solution, phenylhydroxylamine rearranges
to yield 4-aminophenol. If the hydrogenation
of nitrobenzene is carried out in sulfuric acid,
the intermediate phenylhydroxylamine can be
made to rearrange as it forms. This process has
become the preferred route to 4-aminophenol
[123-30-8] (see also → Aminophenols). The hy-
droxylamine is susceptible to further hydrogen-
ation, a reaction whose rate must be made min-
imal relative to the rate of rearrangement.
The preferred catalyst for this process is plat-
inum on carbon. Carbon is inert to the hydrolytic
environment, and platinum is resistant to attack
by acid, as well as being the most effective cata-
lyst for converting nitrobenzene to the hydroxyl-
amine.
Reduction of nitrobenzene in aqueous
sodium hydroxide solution over platinum on
carbon as catalyst is used to produce hydra-
zobenzenes. Sodium hydroxide promotes cou-
pling between the intermediates, nitrosobenzene
and phenylhydroxylamine. The coupling reac-
tion with subsequent benzidine rearrangement
is a useful source of diaminobiphenyls [16].
 Hydrogenation and Dehydrogenation
Diaminotoluene is produced by hydrogen-
ation of dinitrotoluene in the liquid phase by
using supported palladium or nickel catalyst.
Depending on the producer, conditions vary
widely: 50 – 130 ◦ C and 0.2 – 20 MPa. The prod-
uct is converted mostly to toluene diisocyanate
for use in polyurethanes.
Hydrogenation of Nitrate Ion. The hydro-
genation of nitrate ion in phosphoric acid
or of nitrogen monoxide in sulfuric acid
provides large-scale routes to hydroxylamine
[7803-49-8] an intermediate for caprolactam
(see also → Hydroxylamine, Chap. 4.1.). The
latter process produces 0.9 t of ammonium sul-
fate (a sometimes unwanted byproduct) per
ton of caprolactam; the former, the hydroxyl-
amine – phosphate – oxime (HPO) process, pro-
duces none, and employs novel chemistry and
engineering (→ Caprolactam). Reduction is car-
ried out in a slurry with an aqueous solution of
20 % phosphoric acid and 20 % ammonium ni-
trate over palladium on activated carbon, pro-
moted by germanium dioxide, an unusual pro-
moter. The reduced solution is filtered through a
bank of filter sticks mounted within the hydro-
genator. The filtrate then passes in countercur-
rent fashion into a cyclohexanone –toluene mix-
ture to form cyclohexanone oxime in toluene;
the depleted aqueous solution is returned to the
hydrogenator and enriched with more nitrate for
further reaction [17].
Nitriles. Catalytic hydrogenation of nitriles
yields a number of products, including primary,
secondary, and tertiary amines, aldehydes, and
alcohols. All of these products arise from inter-
mediate imines.
The primary amine formed in this sequence can
give an addition product with the imine, which is
converted to a secondary amine, either by elim-
ination of ammonia and subsequent hydrogen-
ation, or by direct hydrogenolysis.
7
Tertiary amines are formed similarly from the
addition of a secondary amine to an imine. Alde-
hydes are obtained in good yield by hydrogen-
ation of a nitrile in aqueous media.
Reduction of a nitrile in the presence of an
added amine permits the formation of unsym-
metrical amines.
The major product obtained in the hydrogen-
ation of nitriles depends strongly on the catalyst.
For instance, hydrogenation of low molecular
mass aliphatic nitriles over rhodium gives sec-
ondary amines in high yield, whereas excellent
yields of tertiary amines are obtained by reduc-
tion over platinum or palladium.
Primary amines are formed from nitriles most
easily by hydrogenation carried out in the pres-
ence of ammonia. The latter prevents secondary
amine formation by competitively reacting with
the intermediate imine.
Cobalt, nickel, and iron are useful catalysts
for the hydrogenation of nitriles; reduction over
these catalysts is usually carried out under vig-
orous conditions. Platinum-group catalysts op-
erate under mild conditions and are used in the
production of fine chemicals.
Adiponitrile. A much studied and used re-
duction of nitriles is the synthesis of hexa-
methylenediamine [124-09-4] from adiponitrile
(see also → Hexamethylenediamine, Chap. 4.).
Hexamethylenediamine is a monomer of nylon
66, as well as an intermediate in the production
of diisocyanates used for foams and resins. Re-
duction of dinitriles presents special problems:
secondary amine formation can, in principle,
8 Hydrogenation and Dehydrogenation
continue indefinitely, resulting in higher molec-
ular mass compounds that cause catalyst fouling
and a loss in yield. Additionally, intramolecu-
lar coupling can occur to produce hexameth-
yleneimine. Both of these reactions are sup-
pressed by ammonia. Hydrogenation of adiponi-
trile also produces small amounts of 1,2-diami-
nocyclohexane, which is unusual in that carbon –
carbon rather than carbon– nitrogen bond forma-
tion takes place. Cobalt, nickel, or iron catalysts
are used at high temperature and pressure (100 –
180 ◦ C, 30 – 60 MPa). Smaller quantities of the
higher homologues of adiponitrile are produced
in similar fashion.
Reductive Alkylation and Amination. Re-
ductive alkylation is the term applied to the
process of attaching alkyl groups derived from
an aldehyde or ketone to ammonia or to primary
and secondary amines. The term is also applied
to those compounds that are not amines or car-
bonyls but are converted to them in the course
of reduction. Reductive alkylation is a two-step
process in which the amine and the carbonyl
first react to give an intermediate that is con-
verted to the final product either by elimination
of water and subsequent hydrogenation of the
resulting imine, or by direct hydrogenolysis of
the addition product.
Nickel, rhodium, palladium, platinum, and
Raney cobalt have all been used for industrial
alkylation. Palladium and platinum are equally
effective when reduction is carried out with
small molecules (e.g., formaldehyde or ace-
tone), but as the molecule increases in size,
platinum becomes the better catalyst. Sulfided
nickel, platinum, or rhodium catalysts are very
effective for reductive alkylation. These cata-
lysts can be used with lower quality feedstocks
because they are more resistant to poisoning than
nonsulfided catalysts; furthermore, they have a
diminished tendency to reduce the carbonyl to
an alcohol and are useful in reductive alkylation
when dehydrohalogenation must be avoided, as
in the conversion of halonitroaromatics to N-
alkylated haloanilines [18].
Reductive alkylation is used widely for
the synthesis of alkylated anilines. Com-
pounds of this type are important antioxidants
(→ Antioxidants).
Either the aniline or the nitro compound makes
a suitable reactant; both are used industrially.
Carboxylic Acids and Esters. Carboxylic
acids and their esters are reduced only with dif-
ficulty; both require vigorous conditions, unless
activated. Rhenium and ruthenium have been
used to reduce carboxylic acids at 20 – 100 MPa
and up to 300 ◦ C. Decarboxylation occurs as a
side reaction, the extent increasing with increas-
ing temperature. Carboxylic esters are reduced
more easily than free acids; the latter tend to
attack both the equipment and the catalyst. Cop-
per chromite catalysts are often used for the
hydrogenolysis of esters.
An industrial route to the important monomer
butane-1,4-diol [110-63-4] involves esterifica-
tion of maleic anhydride with ethanol to obtain
diethyl maleate, and hydrogenolysis of the di-
ester over copper chromite; butyrolactone and
tetrahydrofuran are byproducts. The relative
amounts of these products can be controlled
by reaction conditions. The equilibrium ratio
of butyrolactone and butane-1,4-diol depends
markedly on pressure: at 0.1 MPa and 270 ◦ C,
butyrolactone is highly favored; at 30 MPa,
the diol is favored in the ratio 94 : 6 (see
→ Butyrolactone).
Fatty alcohols are produced in large vol-
ume by hydrogenolysis of fatty acids, or prefer-
ably their esters, under vigorous conditions
(275 –300 ◦ C, 20 – 26 MPa). Reduction of acids
requires the use of corrosion-resistant equip-
ment; ester reduction is carried out in carbon
steel equipment. Natural triglycerides are suit-
able substrates, but methyl esters are prefer-
able. The latter allows easy separation of the
product alcohols and avoids problems connected
with high-temperature decomposition of glyc-
erol. Catalysts are usually a type of copper
chromite. These reduce double bonds as well as
ester groups so that saturated alcohols are pro-
duced regardless of starting material. Unsatu-
rated fatty alcohols are produced by hydrogen-
ation of unsaturated esters over cadmium-based
 Hydrogenation and Dehydrogenation
catalysts at a temperature lower than that nor-
mally used in ester reduction (→ Fatty Alcohols,
Chap. 2.3.3.2., → Fatty Alcohols, Chap. 3.).
2. Dehydrogenation
Dehydrogenation is an important industrial re-
action, although the numbers of applications are
few compared to hydrogenation. The lesser use
of dehydrogenation stems from several aspects
of its chemistry.
Dehydrogenation is an endothermic reaction,
and the high temperature necessary to obtain a
useful conversion to the dehydrogenated prod-
uct restricts the reaction to those substrates and
products with suitable thermal stability. In gen-
eral, industrial dehydrogenation falls into one
of the following categories: (1) the acceptable
product is a complex mixture, as in the mix-
ture of olefins obtained from alkane dehydro-
genation, used for detergent alkylate; (2) an un-
saturated system is extended, as in the conver-
sion of benzene to biphenyl or ethylbenzene to
styrene; (3) aromatization occurs, a reaction that
provides a driving force as well as a convenient
stopping place; (4) a point of attack exists, as in
the conversion of methanol or ethanol to form-
aldehyde or acetaldehyde, respectively; (5) the
choice of product is limited, as in the conversion
of butane or butene to butadiene [9].
2.1. Styrene
Styrene [100-42-5], an important monomer, is
obtained largely by dehydrogenation of ethyl-
benzene in the presence of superheated steam
(→ Styrene).
The catalyst for this reaction usually contains
some major active ingredient, a stabilizer, and
an alkali-metal promoter. Common commercial
catalysts contain iron oxide promoted by potas-
sium and a structural stabilizer. Potassium is a
very effective promoter and increases the reac-
tion rate by an order of magnitude. Potassium
is volatile under reaction conditions and moves
9
toward the reactor exit; redistribution of potas-
sium occurs within the catalyst particles as well.
Styrene formation is endothermic, and as a con-
sequence, the center of a catalyst pellet in use
becomes cooler than the edges; this temperature
gradient causes inward migration of potassium,
resulting in catalyst deactivation near both the
center and the edges due to excess and deficient
potassium, respectively.
Superheated steam plays an important role in
styrene production and serves several purposes:
it is introduced directly into the reactor in large
molar excess to provide part of the necessary
heat to drive the endothermic reaction; it serves
as an oxidant to maintain the iron oxide cata-
lyst in the appropriate oxidation state and re-
moves carbonaceous deposits from the catalyst
through the water – gas shift reaction; it reduces
the partial pressure of hydrocarbons, thereby in-
creasing conversion to styrene. Optimum con-
version is determined by the balance between the
advantage of increased conversion and the cost
of increased steam. Over the last two decades,
improved processing has permitted a continual
decrease in the molar ratio of steam to hydro-
carbon. The ratio is between 6 and 12, with
65 – 70 % conversion and 90 – 95 % selectivity,
at 600 – 650 ◦ C and 51 – 101 kPa. Some reactors
operate in an isothermal mode, heat being sup-
plied by a heat-transfer medium on the outside
of the reactor tubes, which offers the advan-
tage of reduced steam requirements; however,
this advantage is offset by higher capital costs.
Most processes operate in the adiabatic mode.
In practice, conversion is limited by the equi-
librium between styrene, ethylbenzene, and hy-
drogen. As styrene approaches equilibrium con-
centration, its rate of formation decreases but
the rate of byproduct formation does not; the
yield, therefore, falls with increased conversion
[19], [20]. The limitation on conversion can be
cirumvented if the reaction is run as an oxida-
tive dehydrogenation, (i.e., with oxygen present
to combine with liberated hydrogen and remove
it from the equilibrium). Despite much effort, no
large-scale plant operates in this mode.
The technology of ethylbenzene dehydrogen-
ation has been extended to produce both α-
methylstyrene and divinylbenzene by dehydro-
genation of cumene and diethylbenzene, respec-
tively.
10 Hydrogenation and Dehydrogenation

2.2. Dehydrogenation of n-Alkanes

Olefins are hydrogenated easily and quantita-
tively to alkanes under ambient conditions, but is related to pressure by the equation
the reverse reaction is not achieved as readily.
In hydrogenation, the olefinic function provides 4C 3 / (1−C) (1 + 2C)2 = K/p2
a point of attack and the stability of the alkane
Conversion is increased by operation at lower
product provides automatic termination. Alka-
pressure or by use of a diluent.
nes, on the other hand, have no unique point
Butenes and butadiene are produced by
of attack, and the alkene resulting from dehy-
dehydrogenation of butane, although the pro-
drogenation is more easily dehydrogenated than
cess finds increasing competition from oxida-
the starting material itself. Moreover, at ambi-
tive dehydrogenation and from thermal cracking
ent conditions, the equilibrium between alkane
(→ Butadiene, Chap. 4.; → Butenes, Chap. 4.).
and alkene lies completely in favor of the alkane
Operating conditions vary: the temperature is
[21]. Alkenes begin to appear in traces only at
550 – 650 ◦ C and the pressure 30 – 300 kPa, con-
300 ◦ C; at least 400 ◦ C is required for more sub-
tact times are a few seconds. Catalysts coke
stantial conversion (ca. 10 %). The high temper-
rapidly and are oxidatively regenerated every
ature necessary for alkene formation also favors
10 – 100 min. By mixing the catalyst with inert
side reactions such as polymerization and crack-
material, much of the heat liberated during ox-
ing.
idation can be captured and used to drive the
Dehydrogenation of alkanes is greatly helped
endothermic dehydrogenation. Most catalysts
by a suitable catalyst where it is thermodynami-
in this service are based on chromia – alumina.
cally possible at a temperature below that of ther-
These catalysts are strongly inhibited by water
mal decomposition. Propane, butane, and alka-
vapor; the butane feed must, therefore, be dried
nes in the detergent alkylate range (ca. C10 –
carefully.
C14 ) have all been catalytically dehydrogenated
Other catalysts, such as nickel – calcium
on a large scale. Little advantage is gained by
phosphate stabilized with chromia, iron oxide
catalysis in the dehydrogenation of C2 hydro-
promoted by potassium, and chromium oxides,
carbons, because the equilibrium is unfavorable
require the presence of steam to help keep the
at a temperature below that required for thermal
catalyst in the proper oxidation state. Excessive
decomposition.
coking occurs if these catalysts become reduced.
Cracking (cleavage of the carbon – carbon
Steam also acts as a source of heat and shifts the
bond) is energetically favored over cleavage of
conversion toward the dehydrogenated product
the carbon – hydrogen bond, and catalysts such
by lowering the partial pressure of the latter.
as nickel, which are excellent for hydrogen-
ation of alkenes, lead to extensive formation of
methane when applied to dehydrogenation.
The reaction expressing the general alkane – 2.3. Catalytic Reforming of
alkene equilibrium is Hydrocarbons

The objective of catalytic reforming is the con-

Alkenes are favored by lower pressure, as ex-
pressed by the equation


C 2 / 1 −C 2 = K/p
where C is the fraction dehydrogenated, K is
the equilibrium constant, and p the pressure in
atmospheres. This equation holds for the conver-
sion of alkanes to monoenes and of monoenes
to dienes. Conversion of alkanes to dienes (e.g.,
butane to butadiene)
version of saturated hydrocarbons to aromatics
as selectively as possible. This process is the
chief source of high-octane fuel and of aromatic
hydrocarbons. Catalytic reforming involves a
complex series of reactions in which dehydro-
genation figures prominently. Catalysts are dual-
functional, having both a metallic element to cat-
alyze dehydrogenation – hydrogenation and an
acidic function to catalyze hydrocarbon rear-
rangements. Catalysts typically contain a small
amount of highly dispersed platinum (<1 %)
 Hydrogenation and Dehydrogenation
supported on alumina (150 – 300 m2 /g), as well
as a second metal such as rhenium. Polymetal-
lic catalysts permit operation under more severe
conditions; they also possess improved stability
and activity.
Typical operating conditions are 430 –
530 ◦ C and 1.0 – 3.5 MPa. A unit consists of
several, frequently four, fixed-bed reactors in
series. Heated naphtha enters the first reactor,
where the major reaction is endothermic dehy-
drogenation of cycloalkanes accompanied by a
decrease in temperature. The effluent is reheated
and the process continued for all reactors. The
liquid product contains 60 – 70 wt % aromatics;
the gaseous product contains light hydrocar-
bons admixed with 60 – 90 mol % hydrogen. The
gaseous portion is recycled to the first reactor in
a ratio of 5 – 10 mol of recycle gas per mole of
naphtha to maintain a high partial pressure of hy-
drogen in the system. The presence of hydrogen
diminishes the yield of aromatics, but hydrogen
is necessary to maintain catalyst activity by re-
tarding formation of hydrocarbon residues that
poison the catalyst. An aim of processing is op-
eration at as low a partial pressure of hydrogen
as possible, commensurate with good catalyst
life.
Dehydrogenation of cyclohexanes is the most
facile of all reactions that occur in reforming.
On the other hand, only traces of olefins can be
formed by dehydrogenation of alkanes under re-
forming conditions, because of an unfavorable
equilibrium. Nonetheless, dehydrogenation is of
considerable importance because olefins appear
to be intermediates in the overall process [22].
2.4. Liquid-Phase Dehydrogenation
The constraints imposed on dehydrogenation by
equilibrium considerations can be loosened if
hydrogen is removed from the products as they
are formed. Dehydrogenation in the liquid phase
in an open vessel fulfills this goal. Hydrogen es-
capes from the system, whereas products and
reactants remain in the vessel or are returned af-
ter condensation. The lower temperature limit is
then set not by the equilibrium but by a conve-
nient reaction rate.
In practice, the substrate, neat or dissolved in
an inert solvent of convenient boiling point, is
refluxed in the presence of catalyst. Hydrogen
11
escapes while other volatile products are con-
densed and returned, with or without removal
of the product. If the product is removed, the
operation can be made continuous.
All the considerations involving the influence
of mass transport on the rate of reaction (see Sec-
tion 1.1) apply in reverse to dehydrogenation.
Good stirring helps to remove liberated hydro-
gen from the vicinity of the catalyst, and reflux
facilitates hydrogen removal from the liquid. If
the substrate or product is thermally unstable at
normal reflux temperature, the temperature can
be lowered by reducing the pressure. Liquid-
phase dehydrogenation is used in the production
of a number of fine chemicals of limited thermal
stability.
Dehydrogenation may be facilitated by the
use of a hydrogen acceptor to remove hydrogen
from the system as it is formed [23].
The hydrogen transfer reaction above may be
written generally as follows:
These hydrogenation – dehydrogenation reac-
tions are useful for carrying out hydrogenations
in places where hydrogen is unavailable or its
use is not permitted; only simple equipment is
required [24]. Hydrogen transfer reactions often
give better yields of hydrogenated products than
are obtained by direct hydrogenation.
3. References
1. R. L. Augustine: Catalytic Hydrogenation,
Marcel Dekker, New York 1985.
2. M. Freifelder: Practical Catalytic
Hydrogenation, Wiley-Interscience, New
York 1971.
3. P. N. Rylander: Catalytic Hydrogenation over
Platinum Metals, Academic Press, New York
1967.
4. P. N. Rylander: Catalytic Hydrogenation in
Organic Syntheses, Academic Press, New
York 1979.
12 Hydrogenation and Dehydrogenation
5. P. N. Rylander: Hydrogenation Methods,
Academic Press, London 1985.
6. F. Zymalkowski: Katalytische Hydrierungen,
Enke Verlag, Stuttgart 1965.
7. G. W. Roberts in P. N. Rylander, H. Greenfield
(eds.): Catalysis in Organic Syntheses,
Academic Press, New York 1976, p. 1.
8. P. A. Chaloner: Handbook of Coordination
Catalysis in Organic Chemistry, Butterworths,
London 1986.
9. P. N. Rylander, Organic Syntheses with Noble
Metal Catalysts, Academic Press, New York
1973.
10. C. N. Satterfield, Am. Inst. Chem. Eng. 21
(1975) no. 2, 209.
11. J. B. Montgomery et al., Ind. Eng. Chem. 50
(1958) 313.
12. J. F. VanPeppen et al. in R. L. Augustine (ed.)
Catalysis of Organic Reactions, Marcel
Dekker, New York 1985.
13. L. W. Wright, Chemtech. 1974, 42.
14. A. M. Stratz in J. R. Kosak (ed.): Catalysis of
Organic Reactions, Marcel Dekker, New York
1984.
Hydrogen Bromide → Bromine Compounds
Hydrogen Chloride → Hydrochloric Acid
Hydrogen Cyanide → Cyano Compounds, Inorganic
Hydrogen Economy → Hydrogen
Hydrogen Fluoride → Fluorine Compounds, Inorganic
Hydrogen Iodide → Iodine and Iodine Compounds
15. P. N. Rylander et al., Ann. N.Y. Acad. Sci. 172
(1970) 266.
16. C. C. Coe, J. W. Brockington in P. N. Rylander,
H. Greenfield (eds.): Catalysis of Organic
Reactions, Marcel Dekker, New York 1987.
17. G. M. Cor, M. VandeMoesdijk in J. R. Kosak
(ed.): Catalysis of Organic Reactions, Marcel
Dekker, New York 1984.
18. F. S. Dovell, H. Greenfield, J. Am. Chem. Soc.
87 (1965) 2767.
19. C. L. Thomas: Catalytic Processes and Proven
Catalysts, Academic Press, New York 1970.
20. E. H. Lee, Catal. Rev.-Sci. Eng. 8 (1973) 284.
21. K. K. Kearby in P. H. Emmett (ed.): Catalysis,
vol. 3, Reinhold Publishing Corp., New York
1955.
22. J. H. Sinfelt in J. R. Anderson, M. Boudart
(eds.): Catalysis Science and Technology,
vol. 1, Springer-Verlag, New York 1981.
23. J. F. Marschik, P. N. Rylander, US 3 517 021,
1970.
24. G. Brieger, T. J. Nestrick, Chem. Rev. 74
(1974) 567.
 Hydroquinone 1

Hydroquinone
Phillip M. Hudnall, Tennessee Eastman Company, Kingsport, Tennessee 37662, United States

1. Introduction . . . . . . . . . . . . . . . . . 1 4.3. Oxidation of Aniline . . . . . . . . . . . 4

2. Physical Properties . . . . . . . . . . . . 1 4.4. Other Processes . . . . . . . . . . . . . . 4
3. Chemical Properties . . . . . . . . . . . 2 5. Environmental Protection . . . . . . . . 5
4. Production . . . . . . . . . . . . . . . . . . 3 6. Quality Specifications and Analysis . 5
4.1. Hydroperoxidation of p-Diisopropyl- 7. Uses . . . . . . . . . . . . . . . . . . . . . . 5
benzene . . . . . . . . . . . . . . . . . . . . 3 8. Toxicology and Occupational Health . 6
4.2. Hydroxylation of Phenol . . . . . . . . . 4 9. References . . . . . . . . . . . . . . . . . . 7

1. Introduction Commercial grades of hydroquinone are typ-

Hydroquinone (1) [123-31-9] (1,4-benzenediol,
1,4-dihydroxybenzene, p-dihydroxybenzene,
hydroquinol, quinol), C6 H6 O2 , M r 110.11, first
described in 1844 by Woehler, was obtained
by the addition of hydrogen to 1,4-benzoqui-
none.
ically white to off-white crystalline materials;
solutions of hydroquinone are discolored by ox-
idation on contact with air. Physical constants
and solubility data are listed in Table 1. Various
spectra of hydroquinone are readily available:
IR (prism) [2, a], IR (grating) [2, b], 13 C NMR
[2, c], proton NMR [2, d], UV [2, e], and FTIR
[3].
Table 1. Physical properties of hydroquinone

mp, ◦ C 172
bp (101.3 kPa), ◦ C 285
d 15
4 1.332
Flash point (closed cup), ◦ C 165
Hydroquinone occurs naturally as hydro- Autoignition temperature (closed cup), ◦ C 516
quinone β-d-glucopyranoside (arbutin) in the Flammability limits, vol %
Critical temperature, ◦ C
1.6 – 15.3
549
leaves of several plants, including bearberry, Critical pressure, MPa 7.45
cranberry, cowberry, and some varieties of pear Critical volume, m3 /mol 3.0×10−4
[1], usually accompanied by its methyl ether, Liquid molar volume, m3 /mol 1.226×10−4
Heat of formation, kJ/mol −261.71
methylarbutin. Arbutin is easily hydrolyzed to Gibbs free energy of formation, kJ/mol −176.13
hydroquinone and glucose in hot, dilute aque- Absolute entropy, J mol−1 K−1 344.17
ous acid; hydrolysis of methylarbutin produces Heat of fusion at melting point, kJ/mol 27.1
hydroquinone monomethyl ether and glucose. Standard net heat of combustion, kJ/mol −2.74×103
Dipole moment, D 1.40
Van der Waals volume, m3 /mol 5.998×10−5
Van der Waals area, m2 /mol 8.46×105
Dissociation constant (30 ◦ C),
2. Physical Properties K1 1.22×10−10
K2 9.28×10−13
Hydroquinone is a colorless crystalline solid Vapor pressure (132.4 ◦ C), kPa 0.133
when pure. Three crystalline modifications ex- Solubility parameter, (J/m3 )0.5 2.3753×104
Solubility ∗, g per 100 g solvent (30 ◦ C)
ist: the stable α-form, mp 173.8 – 174.8 ◦ C, is Ethanol 46.4
obtained as hexagonal needles by crystalliza- Acetone 28.4
tion from water; the labile γ-form, mp 169 ◦ C, is Water 8.3
obtained as monoclinic prisms by sublimation; Benzene 0.06
Tetrachloromethane 0.01
the labile β-modification is obtained as needles
or prisms by crystallization from methanol or ∗ Reference [4].
isopropyl alcohol.

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 499
2 Hydroquinone

3. Chemical Properties
Hydroquinone is easily converted by most ox-
idizing agents to p-benzoquinone [106-51-4]
(→ Benzoquinone); even neutral aqueous solu-
tions of hydroquinone darken on exposure to air.
The rate of oxidation of hydroquinone by air is
accelerated in alkaline solution. The oxidation
product can add water to give 1,2,4-benzenetriol;
subsequent oxidation may result in the formation
of humic acids. The reactivity of hydroquinone is generally
The redox potential E 0 of hydroquinone is similar to that of phenol (→ Phenol). One or both
699 mV [5]; the half-wave potential E 1/2 is hydroxyl groups may be converted to an ether or
560 mV at pH 0 [5] and 234 mV at pH 5 – 6 [6]. ester. Commercially important oxygen deriva-
Oxidation of hydroquinone involves loss of an tives include hydroquinone monomethyl ether
electron and elimination of a proton to give the (5) [150-76-5], hydroquinone dimethyl ether (6)
relatively stable semibenzoquinone radical (2). [150-78-7], and hydroquinone bis(2-hydroxy-
The ability of hydroquinone to act as an an- ethyl) ether (7) [104-38-1].
tioxidant is a consequence of the formation and
stability of this radical. A second one-electron
transfer and proton elimination results in the for-
mation of p-benzoquinone (3).

Hydroquinone esters undergo the Fries re-

arrangement to give acyl-substituted hydro-
quinones.
Hydroquinone and its ethers may be C-
alkylated under Friedel – Crafts conditions to
Hydroquinone and p-benzoquinone form produce a variety of mono- and disub-
the equimolar charge-transfer complex quinhy- stituted products. Among the commercially
drone. The quinhydrone complex is typically important alkylated derivatives are 2- tert-
dark greenish-black, mp 171 ◦ C. butylhydroquinone (8) [1948-33-0], 2,5-bis[2-
Hydroquinone is a reducing agent, whose re- (2-methylbutyl)]hydroquinone (9), and 2-tert-
duction potential is suitable for reducing silver butyl-4-methoxyphenol (BHA) (10) [121-00-6].
halide. Grains of silver halide that have been ex-
posed to light are reduced much faster than unex-
posed or underexposed grains. This property is
the basis for the use of hydroquinone as a devel-
oping agent in photography (→ Photography).
The selectivity is a result of changes in the ex-
posed silver halide crystals, which reduce the
energy barrier for reaction with hydroquinone
to form elemental silver and p-benzoquinone.
Sodium sulfite is added to the developing so-
lution to prevent formation of quinhydrone by
converting the p-benzoquinone to sodium 2,5-
dihydroxybenzenesulfonate (4), a weaker devel-
oping agent than hydroquinone.
 Hydroquinone 3

Hydroquinone reacts with alkylamines or ities, and major producers are listed in Table 2.
arylamines to form substituted arylamines. Although other processes are known, they are
Commercially important products include p- not industrially important.
N-methylaminophenol (11) [150-75-4], usually
marketed as the sulfate salt [55-55-0], and N,N -
diphenyl-p-phenylenediamine (12) [74-31-7]. 4.1. Hydroperoxidation of
p-Diisopropylbenzene

p-Diisopropylbenzene (p-DIPB) is produced

by Friedel–Crafts alkylation of benzene with
propene (→ Acylation and Alkylation). Purified
p-DIPB (13) is subsequently converted to the
dihydroperoxide (DHP) by air oxidation under
Hydroquinone may be chlorinated with chlo- slightly alkaline conditions at 80 – 90 ◦ C. The
rine or sulfuryl chloride to provide deriva- DHP (14) is separated from the reaction mix-
tives that range from mono- to tetrachlorinated. ture
Kolbe–Schmitt carboxylation of hydroquinone
with carbon dioxide leads to 2,5-dihydroxyben-
zoic acid ( gentisic acid). Quinizarin [81-64-1]
(1,4-dihydroxyanthraquinone) can be prepared
by the condensation of hydroquinone with ph-
thalic anhydride (see also → Anthraquinone
Dyes and Intermediates, Chap. 2.6.1.1.). Cat-
alytic hydrogenation of hydroquinone gives 1,4-
cyclohexanediol ( quinitol). Sulfonation of hy-
droquinone results in the formation of hydro- either by extraction or by crystallization, and
quinone mono- or disulfonic acid, usually iso- is then cleaved to hydroquinone (1) and ace-
lated as the corresponding potassium salt. tone by an acid-catalyzed Hock rearrangement;
the DHP solution is treated with sulfuric acid
catalyst (0.2 – 1.0 %) at 60 – 80 ◦ C. The hydro-
4. Production quinone is crystallized and isolated. The overall
yield of (1) (based on p-DIPB) is ca. 80 %.
Table 2. Production of hydroquinone

Process Capacity, Major producers/location

t/a 4.2. Hydroxylation of Phenol
Oxidation of 25 000 Eastman Chemical
p-diisopropyl- Products/United States The catalyzed hydroxylation of phenol at ca.
benzene Goodyear/United States 80 ◦ C with 70 % aqueous hydrogen peroxide
Mitsui/Japan
produces a mixture of hydroquinone and cate-
Hydroxylation 14 000 Rhône-Poulenc/France
of phenol Enichem/Italy chol (15). The catalyst may be a strong mineral
Ube/Japan acid, iron(II), or a cobalt(II) salt. Depending on
Oxidation 4 000 People’s Republic of China catalyst selection, the ratio of catechol to hydro-
of aniline
Total world 43 000
quinone can be varied from 3 : 1 to 0.1 : 1; in
capacity practice, the ratio is typically 1.5 : 1, i.e., cate-
chol is the major product. The reaction proceeds
by an ionic mechanism in which hydrogen per-
Three processes are used for the industrial oxide is polarized by the strong acid catalyst and
production of hydroquinone: hydroperoxidation phenol is subsequently hydroxylated the result-
of p-diisopropylbenzene [100-18-5], hydroxyla- ing isomers are separated by a series of extrac-
tion of phenol [108-95-2], and oxidation of ani- tions and solvent-stripping operations.
line [62-52-3]. The processes, production capac-
4 Hydroquinone
The ratio of the products hydroquinone and
catechol may be influenced by the presence of
superacids or shape-selective zeolites. Thus, use
of a vanadium–modified Nafion perfluorosul-
fonate polymer for the oxidation of phenol gave
a hydroquinone–catechol ratio of 12.5 : 1 [7].
The use of a shape–selective zeolite in the hy-
droxylation of phenol has yielded hydroquinone
with 99 % selectivity [8].
4.3. Oxidation of Aniline
Oxidation of aniline [62-53-3] is the oldest pro-
cess used for the production of hydroquinone.
Aniline is oxidized with manganese dioxide
(15 – 20 % excess) in aqueous sulfuric acid at
0 – 5 ◦ C to produce p-benzoquinone (3). This in-
termediate is removed from the reaction mixture
by steam stripping and collected. A byproduct,
manganese sulfate, may be retrieved from the
depleted reaction mixture and sold for agricul-
tural applications.
Hydroquinone is obtained from the interme-
diate p-benzoquinone by reduction with iron at
55 – 65 ◦ C or by catalytic hydrogenation. The
product (usually technical grade) is crystallized,
isolated from the typically aqueous stream by
centrifugation, and dried in a vacuum dryer. The
overall yield of hydroquinone from aniline is ca.
85 %.
Aniline oxidation is a batchwise process and,
therefore, relatively labor intensive. The use of
ground manganese ore and finely divided iron
leads to extremely abrasive processing condi-
tions; a great deal of maintenance is usually
required. Disposal of the inorganic coproducts
(amounting to ca. 85 % of the total weight of
products) is of environmental concern.
4.4. Other Processes
Carboxylate esters of aromatic diols can be
prepared by Baeyer–Villiger oxidation of a 4-
hydroxy-substituted aromatic ketone with 97 %
conversion and 97 % efficiency [9]. Microbio-
logical oxidation of either benzene or phenol
produces hydroquinone with very high selec-
tivity [10]. Copper-catalyzed air oxidation of
phenol provides p-benzoquinone in greater than
90 % selectivity [11]. Benzene has been oxi-
dized to hydroquinone in the presence of cop-
per(I) chloride [12] or titanium [13]. Oxidation
of benzene in aqueous solution with ozone gives
p-benzoquinone and hydroquinone [14].
Other preparative methods include the reac-
tion of p-isopropenylphenol with 30 % aque-
ous hydrogen peroxide under acidic condi-
tions [15]; p-isopropenylphenol may be ob-
tained from bisphenol A [80-05-7] by alkaline
cracking. Hydroquinone can also be produced
by the carbonylation of acetylene [16], [17]; cat-
alytic hydrogenation of nitrobenzene in acidic
solution [18]; acid hydrolysis of nitrobenzene
[19] or p-nitrosophenol [20]; and electrochemi-
cal oxidation of benzene or phenol in dilute sul-
furic acid, followed by reduction [21], [22].
5. Environmental Protection
An estimate of environmental effects can be ob-
tained from the data available [23–29]. Hydro-
quinone has a high BOD and can cause oxy-
gen depletion in aqueous systems. It has high
potential for affecting some aquatic organisms
and moderate potential for affecting microor-
ganisms used in secondary waste treatment. Hy-
droquinone has a moderate potential to affect
the germination or early growth of some plants.
 Hydroquinone 5

It is readily biodegraded and has a low ten- 7. Uses

dency to concentrate in biological systems. Di-
rect, instantaneous discharge of hydroquinone to
a large body of water at a final concentration of
0.005 mg/L or less, followed by secondary waste
treatment, is not expected to cause adverse en-
vironmental effects.
6. Quality Specifications and
Analysis
Five grades of hydroquinone are available:
photographic, technical, U.S.P., inhibitor, and
polyester. The most important grades (in terms
of volume) are photographic and technical.
The quality requirements for photographic-
grade hydroquinone are defined in American
National Standards Institute (ANSI) Specifica-
tion PH 4.126-1972. Manufacturers specifica-
tions for photographic and technical grades are
given in Table 3. Specifications for U.S.P.-grade
hydroquinone, used in pharmaceutical applica-
tions, are described in U.S.P.-XXI [30]. Inhibitor
and polyester grades are used in special applica-
tions, and the specifications for each grade are
provided by the producer. Hydroquinone may
be analyzed by a variety of techniques, includ-
ing spectroscopic, chromatographic, titrimetric,
and electrochemical methods.
Hydroquinone and its derivatives are used in
photographic applications, the rubber industry,
monomer inhibitors, dyes and pigments, antiox-
idants, agricultural chemicals, and other diverse
and special applications. Demand for hydro-
quinone and its derivatives is shown in Table 4,
according to market segment. Consumption (in
tonnes) of hydroquinone and derivatives by ge-
ographical area is as follows:
United States 13 100
Europe 9 600
Japan 3 800
Other 4 800
Total 31 300
The largest demand for hydroquinone is as
a photographic developer, principally for black-
and-white film, lithography, photochemical ma-
chining, microfilm, and X-ray film (→ Imaging
Technology, → Photography). Many derivatives
of hydroquinone are used in photographic
applications, e.g., the sulfate salt of p-N-
methylaminophenol (11) and potassium 2,5-di-
hydroxybenzenesulfonate.
Table 4. Worldwide demand for hydroquinone and its derivatives
by market segment (1987)
Market segment Demand

t % of total

Photography 11 600 37
Rubber industry 8 100 26
Table 3. Specifications of hydroquinone
Monomer inhibitors 3 100 10
Dyes and pigments 2 800 9
Property Specifications
Antioxidants 650 2
Photographic Technical Agricultural chemicals 650 2
[32] [31] Others 4 400 14
Total 31 300 100
Appearance free-flowing white to
white crystals offwhite
crystals
mp, ◦ C (min.) 171 169 The second largest consumer of hydro-
Assay, wt % (min.) 99.4 99.0 quinone is the rubber industry, which requires
Ash content, wt % (max.) 0.04
hydroquinone for the production of antioxi-
pH (5 % aqueous solution) 4.1 – 4.7
Iron, ppm (max.) 10 dants and antiozonants (→ Antioxidants). Hy-
Water, % (max.) 0.6 1.0 droquinone derivatives used in this area include
Resorcinol, ppm (max.) 50 N,N -diaryl-p-phenylenediamines (e.g., N,N -
Solubility in water complete
Molar absorptivity ∗ (max). 0.07 (700)
diphenyl-p-phenylenediamine), dialkylated hy-
mL g−1 cm−1 (λmax , nm) 0.12 (520) droquinones, N-alkyl-p-aminophenols, dialkyl-
0.19 (455) 1.0 (455) p-phenylenediamines, and aralkyl-p-phenylene-
∗ In 0.1 M aqueous hydrochloric acid. diamines.
6 Hydroquinone
Hydroquinone, hydroquinone monomethyl
ether, and p-benzoquinone are used exten-
sively in the vinyl monomer industry to inhibit
free-radical polymerization during both pro-
cessing and storage. Hydroquinone, p-benzo-
quinone, 2-methylhydroquinone (toluhydro-
quinone), 2-tert-butylhydroquinone, and 2,5-
di-tert-butylhydroquinone are used as stabiliz-
ers for unsaturated polyester resins. Food-grade
antioxidants include 2-tert-butylhydroquinone
(8) and 2-tert-butyl-4-methoxyphenol (10)
(→ Antioxidants, Chap. 5.1.).
Hydroquinone dimethyl ether (6) is used as
a starting material for a family of dyes and pig-
ments based on the 2-amino- and 2-amino-5-
chloro derivatives. Quinizarin is an intermedi-
ate for textile dyes (→ Anthraquinone Dyes and
Intermediates). The fungicide Chloroneb is pro-
duced from (6) as raw material. The herbicide
ethofumesate is produced from p-benzoquinone.
Fluazifop-butyl [33] is representative of a new
family of herbicides based on the o-alkylation of
hydroquinone with 2-halopropionic acid deriva-
tives [34–37].
Hydroquinone bis(2-hydroxyethyl) ether (7),
the product of the reaction of hydroquinone with
ethylene oxide, is used as a chain extender in
thermosetting urethane polymers [38].
Hydroquinone (U.S.P. grade) and several
derivatives are used in topical formulations as
skin bleaching and depigmenting agents.
Several new applications for hydroquinone
and its derivatives are emerging. The oxygen-
scavenging properties of hydroquinone are be-
ing exploited for use in boiler water treatment
[39]. Hydroquinone and certain C-alkylated
or Carylated derivatives are useful monomers
for the preparation of a variety of polymers,
including liquid crystal polyesters for high-
performance plastics, composites, and fibers. In
addition to high tensile and impact strengths,
these materials exhibit good weatherability, sol-
vent resistance, flame retardance, transparency
to microwave radiation, and retention of strength
at elevated temperature [40], [41].
8. Toxicology and Occupational
Health
Acute toxicity data are shown in Table 5. In
a two-year study of rats receiving a diet con-
taining up to 1 % hydroquinone, no effects were
observed on final body weight, hematology, or
pathology; at 5 % in the diet, a 46 % weight
loss together with toxic effects on the bone mar-
row and liver was found within nine weeks [43].
When applied to the skin of mice, hydroquinone
was not carcinogenic and did not promote the
carcinogenic activity of benzo[a]pyrene [44]. In
a skin-painting study in rats, croton oil failed
to promote any carcinogenicity from hydro-
quinone [45]. Hydroquinone produced a nega-
tive response in the Salmonella bacterial mu-
tagenesis assay [46]. Positive results were ob-
served in the micronucleus test [47] and in an
in vitro test for sister chromosome exchanges in
human lymphocytes [48].
Table 5. Toxicity data for hydroquinone [23]
Test Species Result
Acute oral LD50 rat 400 mg/kg
Acute oral LD50 mouse 100 – 200 mg/kg
Dermal LD50 guinea pig > 1000 mg/kg
Skin irritation guinea pig slight
Eye irritation rabbit moderate erythema,
clearing by day 14
Sensitization (guinea pig)
Freund’s adjuvant 3/10 slight
Modified Buhler 7/15 low,
2/10 moderate
Maximization [42] 70 % sensitized,
strong
Maternal toxicity and fetotoxicity were ob-
served in pregnant rats given 300 mg/kg of hy-
droquinone daily, but no compound-related ter-
atogenic effects occurred; no adverse effects
were observed at doses up to 100 mg/kg per day
[23]. Rats given a single dose of 200 mg/kg of
(14 C)hydroquinone excreted 91.9 % in the urine
within 2 – 4 d, 3.8 % in the feces, and 0.39 % in
expired air, with 1.25 % remaining in the body;
radiolabeled compounds identified in the urine
included hydroquinone monoglucuronide (56 –
66 %), hydroquinone monosulfate (25 – 42 %),
and hydroquinone (1.1 – 8.6 %) [49]. The per-
cutaneous absorption rate of (14 C)hydroquinone
in beagles was ca. 1.1 µg cm−2 h−1 , indicating
that hydroquinone is poorly absorbed through
the skin [23].
The OSHA Permitted Exposure Limit (PEL)
and ACGIH Threshold Limit Value (TLV) for
hydroquinone is 2 mg/m3 TWA. Effects of hu-

man exposure are well documented (see below).
Direct contact of hydroquinone dust and quinone
vapor with the eye may cause irritation. Chronic
exposure at high levels has caused brownish dis-
coloration of the cornea and conjunctiva, as well
as distortion of the cornea, leading in some cases
to decreased visual acuity and even blindness.
Exposure to levels below the TLV of 2 mg/m3
has not produced these effects [50].
A low incidence of allergic contact dermatitis
has been reported from handling large amounts
of hydroquinone, where exposure to dust oc-
curred, and from repeated contact with alkaline
photographic developer solutions [23]. Skin sen-
sitization was noted in 2 members of a group of
48 middle-aged males who used a depigmen-
tation ointment containing 5 % hydroquinone;
38 subjects treated with a 2 % ointment and 26
subjects treated with a 3 % ointment showed no
sensitization [51], [52].
In reported poisonings, the lethal dosage
for adults is 5 – 12 g (70 – 170 mg/kg). Lower
doses cause no adverse effects: two men in-
gested 500 mg/d for five months and 17 men
and women ingested 300 mg/d for three to five
months without abnormal symptoms or any ab-
normalities in blood and urine [43].
9. References
1. Beilstein, VI3 , 4374.
2. Sadtler Standard Spectra, Sadtler Research
Laboratories, Division of Bio-Rad
Laboratories, a. Infrared Prism Spectrum
nos. 153, 13206; b. Infrared Grating Spectrum
no. 47; c. Carbon-13 NMR Spectrum no. 38; d.
Proton NMR Spectrum no. 10350; e.
Ultraviolet Spectrum no. 60.
3. C. J. Pouchert (ed.), The Aldrich Library of
FTIR Spectra, Aldrich Chemical Company,
Milwaukee, Wisconsin 1984, vol. 1,
no. 1107D.
4. A. Seidell: Solubilities of Organic
Compounds, 3rd ed., vol. 2, Van Nostrand
Publishing, New York 1941, p. 396.
5. H. Musso, H. Doepp, Chem. Ber. 100 (1967)
3267.
6. P. J. Elving, A. F. Krivis, Anal. Chem. 30
(1958) 1645.
7. FMC, EP-A 132 783, 1985 (R. A. Bull).
8. Mobil Oil, US 4 578 521, 1986 (C. D. Chang,
S. D. Hellring).
Hydroquinone 7
9. Celanese, EP-A 178 929, 1986; US-A 661 552,
1984 (H. R. Gerberich).
10. Agency of Industrial Sciences & Technology,
JP-Kokai 8 678 394, 1986 (A. Yoshikawa).
11. J. E. Lyons, C-Y. Hsu, Biological and Inorganic
Copper Chemistry, in K. D. Karlin, J. Zubieta
(eds.): Proc. Conf. Copper Coord. Chem. (July
23 – 27, 1984), vol. 2, Adenine Press,
Guilderland, New York 1986, pp. 57 – 76.
12. J. Van Gent, A. A. Wismeijer, A. W. P. G.
Peters-Rit, H. Van Bekkum, Tetrahedron Lett.
27 (1986) 1059 – 1062.
13. Mitsui Petrochemicals Industries, JP-Kokai
859 889, 1985.
14. C. H. Kuo, H. S. Soong, Chem. Eng. J.
(Lausanne) 28 (1984) 163 – 171.
15. Upjohn, DE 2 214 971, 1972, and
GB 1 344 602, 1974 (P. S. Charleton).
16. Ajinomoto, US 3 420 895, 1969 (H.
Wakamatsu).
17. Lonza, GB 1 215 568, 1970 (P. Pino).
18. Koppers, US 3 953 509, 1974 (N. P. Greco).
19. Mitsui Toatsu, JP-Kokai 77 102 235, 1976 (F.
Matsuda).
20. Upjohn, US 3 676 503, 1972 (W. v. E. Doering,
W. J. Farrisey, Jr.).
21. M. Ya. Fioshin, Elektrokhimiya 13 (1977) 381.
22. M. Fremery, H. Hoever, G. Schwarzlose,
Chem. Ing. Technol. 46 (1974) 635.
23. Unpublished results, Health & Environment
Laboratories, Eastman Kodak, Rochester, New
York.
24. K. Verschueren: Handbook of Environmental
Data on Organic Chemicals, 2nd ed., Van
Nostrand Reinhold, New York 1983,
pp. 746 – 748.
25. Environmental Effects of Photoprocessing
Chemicals, vols. 1 and 2, National
Association of Photographic Manufacturers,
and Hydroscience, Harrison, New York 1974.
26. G. M. De Graeve, D. L. Geiger, J. S. Meyer,
H. L. Bergman, Arch. Environ. Contam.
Toxicol. 9 (1980) 557.
27. J. E. McKee, H. W. Wolf (eds.): Water Quality
Criteria, State of California, 1963,
Publication no. 3-A.
28. I. Juhnke, D. Luedemann, Z. Wasser Abwasser
Forsch. 11 (1978) 161.
29. A. J. Leo, C. Hansch (eds.): Chemical
Parameter Data Base, Medicinal Chemistry
Project, Pomona College, Seaver Chemistry
Laboratory, Claremont, Calif. 1985.
30. The United States Pharmacopeia XXI, The
United States Pharmacopeial Convention,
Rockville, Maryland, 1985.
8 Hydroquinone
31. Hydroquinone, Technical Grade, Sales
Specification no. 3503-9, Tennessee Eastman
Company, Kingsport, Tenn.
32. Hydroquinone, Photographic Grade, Sales
Specification no. 4305-1, Tennessee Eastman
Company, Kingsport, Tenn.
33. Ishihara Sangyo Kaisha, DE 2 812 571, 1979
(R. Nishiyama et al.).
34. Dow Chemical Co., DD 217 694, 1985 (H.
Johnson, L. H. Troxell).
35. Dow Chemical Co., EP-A 97 460, 1984 (H.
Johnson, L. H. Toxell).
36. Bayer AG, DE 3 219 789, 1983 (H. Foerster et
al.).
37. Zoecon Corp., US 4 529 438, 1985 (S. Lee).
38. Farbenfabriken Bayer Aktiengesellschaft and
Mobay Chemical Company, US 3 016 364,
1962 (E. Muller).
39. S. Romaine, I. J. Cotton, Proceedings
American Power Conference 48 (1986) 1066.
40. D. G. Baird, Fiber Producer (October1984)
66.
41. A. S. Wood, Modern Plastics (April1985) 78.
42. B. F. J. Goodwin, R. W. R. Crevel, A. W.
Johnson, Contact Dermatitis 7 (1981) 248.
43. A. J. Carlson, N. R. Brewer, Proc. Soc. Exp.
Biol. Med. 84 (1953) 684.
44. B. L. van Duuren, B. M. Goldschmidt, J. Nat.
Cancer Inst. 56 (1976) 1237.
Hydroxyanthraquinones → Anthraquinone Dyes and Intermediates
Hydroxybenzaldehydes → Benzaldehyde
45. IARC Monographs on the Evaluation of the
Carcinogenic Risk of Chemicals to Humans,
vol. 15, International Agency for Research on
Cancer (IARC), Lyon, France 1977,
pp. 155 – 175.
46. I. Florin, L. Rutberg, M. Curvall, C. R. Enzell,
Toxicology 15 (1980) 219.
47. E. Gocke, M.-T. King, K. Eckhardt, D. Wild,
Mutat. Res. 90 (1981) 91.
48. K. Morimoto, S. Wolff, A. Koizumi, Mutat.
Res. 119 (1983) 355.
49. G. D. Divincenzo, M. L. Hamilton, R. C.
Reynolds, D. A. Ziegler, Toxicology 33 (1984)
9.
50. F. L. Oglesby, R. L. Raleigh: Eye Injury
Associated with Exposure to Hydroquinone
and Quinone, 2nd ed., Laboratory of
Industrial Medicine, Tennessee Eastman Co.,
Kingsport, Tenn. 1973, p. 2.
51. NIOSH Criteria for a Recommended Standard
for Hydroquinone, National Institute for
Occupational Safety & Health (NIOSH), U.S.
Government Printing Office, Wahington, D.C.,
1978.
52. M. C. Spencer, J. Am. Med. Assoc. 194 (1965)
962.

Hydroxycarboxylic Acids, Aliphatic
Butyrolactone, Citric Acid, Lactic Acid, and Tartaric Acid are separate keywords.
Karlheinz Miltenberger, Hoechst Aktiengesellschaft, Gersthofen, Federal Republic of Germany
1. Introduction . . . . . . . . . . . . . . .
2. General Characteristics . . . . . . . .
2.1. Physical Properties . . . . . . . . . . .
2.2. Chemical Properties . . . . . . . . . .
3. Preparation . . . . . . . . . . . . . . . .
4. Analysis . . . . . . . . . . . . . . . . . .
5. Specific Aliphatic Hydroxycarbox-
ylic Acids of Commercial Signifi-
cance . . . . . . . . . . . . . . . . . . . .
5.1. Glycolic Acid (Hydroxyacetic Acid,
2-Hydroxyethanoic Acid) . . . . . . .
1. Introduction
Hydroxycarboxylic acids are widely distributed
in nature, playing an important role as metabolic
intermediates in both plants and animals. Exam-
ples can be found in the tricarboxylic acid cycle
leading to the degradation (β-oxidation) of fatty
acids, as well as in the fermentation of sugars.
Hydroxycarboxylic acids are also excreted in the
urine, especially in metabolic disorders. Thus,
diabetics are incapable of further metabolizing
the β-hydroxybutyric acid and acetoacetic acid
derived from butyric acid, so these materials are
simply eliminated.
One of the first reports of a naturally occur-
ring hydroxycarboxylic acid occurred in 1780
when Scheele described the isolation of (R)-
(−)-lactic acid from soured milk. The sub-
stance is formed during the fermentation of lac-
tose by lactobacilli. Scheele also discovered S-
(−)-malic acid (in 1785), S-(+)-tartaric acid (in
1769), and citric acid (in 1784) isolating each as
its calcium salt. Another member of the group,
glycolic acid, is a constituent of unripe fruit, es-
pecially grapes. Related natural substances in-
clude tartronic, hydroxystearic, mandelic (see
→ Hydroxycarboxylic Acids, Aromatic), and
numerous other fruit- or sugar-derived acids.
Definition In contrast to the purely aro-
matic hydroxycarboxylic acids (phenolic car-

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a13 507
Hydroxycarboxylic Acids, Aliphatic 1
1 5.1.1. Preparation . . . . . . . . . . . . . . . . . 7
2 5.1.2. Uses . . . . . . . . . . . . . . . . . . . . . 7
2 5.1.3. Derivatives . . . . . . . . . . . . . . . . . 8
2 5.2. Hydroxypropionic Acids and
4 β-Propiolactone . . . . . . . . . . . 8
6 5.3. Hydroxybutyric Acids . . . . . . . . . 9
5.4. (R,S)-Malic Acid
[(R,S)-Hydroxysuccinic Acid,
7 (R,S)-Hydroxybutanedioic Acid] . . 10
6. Toxicology . . . . . . . . . . . . . . . . . 10
7 7. References . . . . . . . . . . . . . . . . . 11
boxylic acids), which contain phenolic hydroxyl
groups on the aromatic nucleus and thus display
enhanced acidity, aliphatic hydroxycarboxylic
acids are characterized by purely aliphatic al-
cohol hydroxyl groups.
Aliphatic hydroxycarboxylic acids may have
one or several hydroxyl groups, together with
one or more carboxyl groups. In the simplest
case, the monohydroxymonocarboxylic acids,
the compounds can be conveniently subdivided
into three categories:
1) 2- or α-hydroxycarboxylic acids,
2) 3- or β-hydroxycarboxylic acids, and
3) hydroxycarboxylic acids with the hydroxyl
group located in the 4- (i.e., γ-) position, or
even more remote from the carboxyl group.
Table 1 lists several commercially impor-
tant hydroxycarboxylic acids and related com-
pounds, grouped in accordance with the above
scheme.
2. General Characteristics
The overall characteristics of a hydroxycarbox-
ylic acid are a function of the number of hy-
droxyl and carboxylic acid groups present as
well as their relative placement.
2 Hydroxycarboxylic Acids, Aliphatic
Table 1. Commercially important hydroxycarboxylic acids and derivatives
2.1. Physical Properties
Most naturally occurring hydroxycarboxylic
acids are chiral and thus exhibit optical activ-
ity. In pure form, they are usually crystalliz-
able. Hydroxycarboxylic acids have exception-
ally high boiling points relative to corresponding
unsubstituted acids with the same chain length.
This is a consequence of association among the
hydroxyl groups. For this reason, many of these
compounds cannot be distilled without decom-
position, even under vacuum.
Table 2 summarizes the physical properties of
the commercially important hydroxycarboxylic
acids and their derivatives listed in Table 1.
2.2. Chemical Properties
Hydroxycarboxylic acids readily form the ex-
pected salts at the carboxyl group. The acid-
ity of α- and β-hydroxycarboxylic acids is en-
hanced relative to the corresponding unsubsti-
tuted acids because of the proximity of the po-
lar hydroxyl group. Compounds in this category
may be esterified at either the carboxyl or the hy-
droxyl function by reaction with other alcohols
or acids, respectively. The actual course of such
reactions depends once again on the placement
of the hydroxyl group relative to the carboxylic
acid group.
Both α- and β-hydroxycarboxylic acids read-
ily form esters with other alcohol groups but
the latter also have a tendency to eliminate wa-
ter under the reaction conditions. The γ- and
 Hydroxycarboxylic Acids, Aliphatic 3
Table 2. Physical properties of the hydroxycarboxylic acids listed in Table .

Hydroxycarboxylic acid mp, ◦ C bp, ◦ C Density, Refractive pK 25

A1 ,
d 20
4 index, n20
D (25 ◦ C)

Glycolic acid 78 – 80 100 (decomp.) 1.49 (at 25 ◦ C) 3.81

n-Butyl glycolate 178 – 186 or 82 – 97 (at 2.7 kPa) 1.015 – 1.023 1.423 – 1.4246
(R,S)-Mandelic acid 118 – 121 1.300 3.37
(R,S)-Lactic acid 23 – 33 119 – 123 (at 1.6 – 2 kPa) 1.206 (25 ◦ C) 1.4392 3.86
(R,S)-α-Hydroxybutyric acid 43 – 44 138 (at 1.9 kPa) or 260 (decomp.) 1.125
Hydracrylic acid syrup decomp. 1.0474 1.4489 4.51
β-Propiolactone −33.4 51 (at 1.4 kPa) or 162 (decomp.) 1.1460 1.4135
(R,S)-Tropic acid 118 160 (decomp.)
(R,S)-β-Hydroxybutyric acid 48 – 50 or 94 – 96 (at 0.01 kPa) or 130 1.4424 4.39
syrup (at 1.5 – 1.9 kPa)
γ-Hydroxybutyric acid −17 decomp. 4.71
γ-Butyrolactone −43.5 206 1.13 1.4352
(R,S)-Ricinoleic acid 5 230 – 235 (at 1.2 kPa) 0.94 (at 15 ◦ C) 1.46393
(at 45 ◦ C)
(R,S)-Glyceric acid syrup
Tartronic acid 156 – 158 2.31
(decomp.)
(R,S)-Malic acid 131 1.601 3.4
(R,S)-Tartaric acid 205 3.03
meso-Tartaric acid 159 – 160 3.11
Citric acid 153 1.665 3.13

δ-hydroxycarboxylic acids undergo intramolec-

ular esterification to form lactones, this reaction
being faster than intermolecular esterification.
Compounds in which the hydroxyl group is more
remote from the acid moiety are again subject to
normal esterification.
α-Hydroxycarboxylic acids dimerize with
elimination of water to produce six-membered
cyclic diesters called 1,4-dioxane-2,5-diones ac-
cording to IUPAC. These are commonly referred
to as lactides, because the phenomenon was first
encountered with lactic acid.
Note that the latter type of esterification, known
as estolide formation, is not restricted to α-hy-
droxycarboxylic acids. Another important char-
acteristic of both α- and β-hydroxycarboxylic
acids is that the proximity of the two functional
groups weakens the intervening C−C bond(s).
Thus, treatment of such compounds with sul-
furic acid results in the elimination of formic
acid, which decomposes in the presence of con-
centrated sulfuric acid to carbon monoxide and
water.

The rest of the chain is converted into aldehydes

The process may be accompanied by inter- (in the case of α-hydroxycarboxylic acids) or ke-
molecular esterification, which produces linear tones (in the case of β-hydroxycarboxylic acids)
polyesters. with one fewer carbon atom.
4 Hydroxycarboxylic Acids, Aliphatic
This elimination reaction proceeds so
smoothly that it is used for quantitative determi-
nation of certain hydroxycarboxylic acids. Thus,
lactic acid is converted easily to acetaldehyde,
and β-hydroxybutyric acid to acetone. Oxida-
tive cleavage is also possible. For example, treat-
ment of α-hydroxycarboxylic acids with hydro-
gen peroxide in the presence of iron(II) ions
leads to easy elimination of carbon dioxide to
produce again aldehydes containing one fewer
carbon atom in the chain.
β-Hydroxycarboxylic acids readily undergo
intramolecular elimination of water to produce
α,β-unsaturated carboxylic acids; e.g.,
Thus, β-hydroxypropionic acid (hydracrylic
acid) can be dehydrated to acrylic acid, and β-
hydroxybutyric acid to crotonic acid.
The γ- and δ-hydroxycarboxylic acids un-
dergo dehydration even at room temperature,
yielding γ- and δ-lactones. The corresponding
free acid may be regenerated in aqueous alka-
line medium, but often it is stable only in the
form of its salts. Lactones are internal esters of
hydroxycarboxylic acids.
Treatment of a lactone with ammonia produces
the corresponding lactam.
The related derivatives of α-hydroxycar-
boxylic acids, the three-membered α-lactones,
are unstable and are considered only as reac-
tion intermediates. Four-membered lactones (β-
lactones from β-hydroxycarboxylic acids) are
known, but special methods are required to pre-
pare them [1], [2]. The exceptional reactivity
of these compounds is a consequence of their
highly strained rings. For example, they react
with ammonia to form amides of the correspond-
ing hydroxycarboxylic acids, which are subject
to further conversion to amino acids.
Because of the properties described above,
the hydroxyl group of a hydroxycarboxylic acid
can be converted into various derivatives. Thus,
treatment of lactic acid with hydrogen bromide
results in α-bromopropionic acid. Hydrogen io-
dide can be used to reduce a hydroxycarboxylic
acid to the corresponding unsubstituted acid,
which is particularly easy with acids from the
sugar family.
Saccharic acid lactones are converted easily
to polyhydroxyaldehydes (aldoses).
Phosphorus pentachloride leads to replace-
ment of the hydroxyl group in a hydroxycarbox-
ylic acid by chlorine, with the carboxyl group
simultaneously converted into an acid chloride
moiety. Treatment with thionyl chloride results
in an unstable sulfurous ester of the acid chlo-
ride, which will react, with an alcohol to give
a hydroxyester or with ammonia to yield a hy-
droxyamide. If the hydroxyl group is protected
by acetylation, thionyl chloride treatment leads
only to the corresponding acid chloride.
3. Preparation
In addition to their isolation from natural
sources, hydroxycarboxylic acids can also be
obtained by a number of synthetic routes.
One general approach involves hydrolysis of a
halocarboxylic acid:
X−R−COOH + OH− −→ HO−R−COOH + X−
This route is in most cases quite straightfor-
ward for the preparation of α-hydroxycarbox-
ylic acids (e.g., glycolic acid), but β- and γ-
halocarboxylic acids lead instead to the corre-
sponding lactones.
Certain hydroxycarboxylic acids may also be
prepared by acid- or base-catalyzed hydration of
an unsaturated carboxylic acid.
For example, either maleic or fumaric acid
can serve as a source of racemic tartaric acid,
and acrylic acid can be converted in this way to
β-hydroxypropionic acid (hydracrylic acid).
α-Hydroxycarboxylic acids may be prepared
by the action of zinc and an alkyl iodide on an
oxalate ester. Cyanohydrin synthesis is another
synthetic route to α-hydroxycarboxylic acids,
one example being the synthesis of (R,S)-lactic
acid from acetaldehyde via lactonitrile:
 Hydroxycarboxylic Acids, Aliphatic
Another interesting preparative method in-
volves diazotization as a means of deaminating
an amino acid:
α-Olefins react with liquid dinitrogen tetrox-
ide at 15 – 20 ◦ C to produce nitrate esters of
α-hydroxycarboxylic acids, compounds whose
electrochemical reactivity is described in [3].
β-Hydroxycarboxylic acids can be prepared
from epoxides by treatment with hydro-
gencyanide, followed by hydrolysis of the in-
termediate nitriles:
An alternative is the Kolbe nitrile synthesis in-
volving addition of hypochlorous acid to an
olefin. This results initially in a chlorohydrin,
which can be converted with cyanide ion to a
hydroxynitrile, which is then hydrolyzed:
β-Hydroxycarboxylic acids can also be pre-
pared from α-halocarboxylic acid esters by Re-
formatsky reaction with an aldehyde or ketone
in the presence of activated zinc [4]. Another ap-
proach utilizes β-lactones, which are readily ac-
cessible through ketene treatment of aldehydes
or ketones [5]. Thus, acetaldehyde reacts with
ketene in the presence of zinc salts to give β-
butyrolactone. Finally, esters of β-ketoacids are
subject to catalytic hydrogenation, a method that
is also applicable to the synthesis of γ- and δ-hy-
5
droxycarboxylic acids from the corresponding
γ- and δ-ketoesters.
For example, catalytic hydrogenation of ace-
toacetic esters give the esters of (R,S)-β-hy-
droxybutyric acid.
Both γ- and δ-hydroxycarboxylic acids can
be prepared by alkaline hydrolysis of the corre-
sponding halo acids. In this case, the products
are normally isolated in the form of alkali-metal
salts.
Dihydroxycarboxylic acids may be made by
reaction of unsaturated fatty acids (e.g., oleic
acid) with hydrogen peroxide or peracid, in
which epoxides are formed as intermediates.
Polyhydroxy acids are also commonly prepared
from unsaturated carboxylic acids, usually by
permanganate oxidation.
Many of the methods described above lend
themselves equally well to the synthesis of poly-
basic hydroxycarboxylic acids. For instance,
tartronic acid has been prepared both by hydro-
lysis of monobromomalonic acid or by sodium
amalgam reduction of mesoxalic acid (ketoma-
lonic acid):
Another approach starts with glycerol, which
upon oxidation with potassium permanganate
yields tartronic acid, whereas treatment with ni-
tric acid yields glyceric acid.
As in this case, careful selection of oxidizing
agent often enables various hydroxycarboxylic
acids to be obtained by starting with the same
precursor, a point well illustrated by the oxida-
tion of sugars (aldoses and ketoses). Mild oxi-
dizing agents such as aqueous bromine or dilute
nitric acid attack only the aldehyde functions of
6 Hydroxycarboxylic Acids, Aliphatic
aldoses (→ Carbohydrates), producing glyconic
acids (e.g., gluconic acid from glucose). Such
glyconic acids (also called “onic acids”) are sta-
ble only in alkaline solution, where they exist as
alkali-metal salts. The free acids rapidly cyclize
to γ- or δ-lactones, with the γ-lactones preferred.
More powerful oxidizing agents such as con-
centrated nitric acid result in polyhydroxy diba-
sic acids ( saccharic acids), which immediately
cyclize to lactones. Thus, glucose leads to glu-
caric acid; mannose to mannaric acid; and galac-
tose to mucic acid (galactaric acid). Even more
vigorous oxidation causes cleavage of the car-
bon chain, yielding lower hydroxydicarboxylic
acids.
Numerous branched and unbranched hy-
droxycarboxylic acids may be prepared by
cleavage or oxidation (with alkaline perman-
ganate, chromic acid, dilute nitric acid, or con-
centrated nitric acid) of a variety of starting ma-
terials [6–9]. Enzymatic production by fermen-
tation is becoming increasingly important.
4. Analysis
The literature describes an array of color re-
actions for qualitative determination of hy-
droxycarboxylic acids. These include the yel-
low color obtained in the presence of iron(II)
chloride, characteristic complexation phenom-
ena with copper salts, and the distinctive color
produced with ammonium vanadate. Neverthe-
less, none of these reactions can be regarded as
truly specific.
The most satisfactory approach to quanti-
tative analysis involves acidimetric titration of
carboxyl groups. Because of the nearly univer-
sal tendency to lactone formation in equilibrium
with the free acid, free and total acids must be
determined separately. Thus, lactones are treated
in the same way as other esters: Hydrolysis with
excess NaOH is followed by back titration with
acid. Other recommended methods for quanti-
tative analysis rely on acetylation of hydroxyl
groups or esterification of carboxyl groups, with
subsequent determination of the resulting water.
Instrumental methods of analysis are play-
ing an increasingly important role, including
IR, NMR and MS spectroscopy and, especially,
chromatography (e.g., thin-layer chromatogra-
phy and HPLC). Nevertheless, such methods
prove reliable only if they are carefully adapted
to suit the specific problem. A typical procedure
involves methylation or silylation, followed by
identification and quantification with GC – MS
[7].
5. Specific Aliphatic
Hydroxycarboxylic Acids of
Commercial Significance
5.1. Glycolic Acid (Hydroxyacetic Acid,
2-Hydroxyethanoic Acid) [10]
The chemical structure, molecular mass, and
CAS registry number of glycolic acid are given
in Table 1. Its physical properties are listed in
Table 2 and in the following material:
mp:
α-Modification 78 – 80 ◦ C
β-Modification, metastable 63 ◦ C
Heat of combustion −697.23 kJ/mol
Heat of solution in water
(infinite dilution) −11.71 kJ/mol
Dissociation constant (at 25 ◦ C) 1.54×10−4
pH (1 M solution) 2.4
Solid glycolic acid forms colorless, mono-
clinic, prismatic crystals. The acid is freely sol-
uble in water, methanol, ethanol, acetone, and
 Hydroxycarboxylic Acids, Aliphatic
ethyl acetate. Glycolic acid is only slightly steam
volatile and cannot be destilled under vacuum;
attempts result in self-esterification with loss of
water, leading to di- and polyglycolides. How-
ever, newer results show that glycolic acid can
be vaporized without decomposition. In the va-
por phase it can be dehydrogenated catalytically
to glyoxylic acid [11].
5.1.1. Preparation
Glycolic acid is usually produced by hydroly-
sis of molten monochloroacetic acid with 50 %
aqueous sodium hydroxide at 90 – 130 ◦ C. The
resulting glycolic acid solution has a concentra-
tion of ca. 60 % and contains 12 – 14 % sodium
chloride. The salt may be removed by evapo-
rative concentration, followed by extraction of
the acid with acetone [12]. Attempts have also
been made to conduct the hydrolysis with acid
catalysts at 150 – 200 ◦ C with water or steam un-
der pressure [13]. In this case, the byproduct is
hydrogen chloride, rather than sodium chloride,
which can be removed by distillation. The prin-
cipal disadvantage of the method is the need for
relatively large volumes of water.
Glycolic acid is produced commercially in
the United States (Du Pont) by treating formal-
dehyde or trioxymethylene with carbon monox-
ide and water in the presence of acid catalysts at
>30 MPa [14]. Another process, previously uti-
lized by Degussa, involves electrolytic reduction
of oxalic acid [15]. The compound can also be
prepared in 90 % yield by hydrolysis of the cor-
responding nitrile, obtained by reaction of form-
aldehyde with hydrocyanic acid [16].
5.1.2. Uses
Glycolic acid is available commercially as either
a 57 % (Hoechst) or a 70 % (Du Pont) aqueous
solution. Total annual consumption worldwide
is ca. 2000 – 3000 t of solution.
Glycolic acid is used in textile dyeing, print-
ing, and creaseproofing. The fact that it can
form a chelate with calcium(II) ions makes it
well-suited to hide deliming in the leather in-
dustry, as well as to inclusion in alum and
chrome mordants and in fur-processing opera-
tions. The compound has little tendency to cause
7
corrosion, and this characteristic, coupled with
its bactericidal properties, makes it suitable for
incorporation into acidic cleansing agents. It
is especially well-adapted to cleansing opera-
tions involving milk containers, milk-processing
equipment, and drinking fountains, as well as
rust and scale removal in heat exchangers and
pipelines. Glycolic acid inhibits the growth of
iron-oxidizing bacteria. The use of glycolic acid
eliminates the need for simultaneous addition of
chelating agents and bactericides. The effective-
ness of glycolic acid as a complexing agent also
contributes to its use in copper polishes, as an
etching agent for lithographic plates, and in the
preparation of electropolishing and galvanizing
baths.
Safety Considerations. According to section
1.4 of the hazardous materials regulation,
Gef Stoff V (Federal Republic of Germany), and
to the EEC guidelines, glycolic acid solutions are
“corrosive.”
Analysis. Glycolic acid may be detected
qualitatively by the violet color formed with
2,7-dihydroxynaphthalene [17]. The preferred
method of quantitative analysis (in the absence
of other acidic or hydrolyzable substances) is
acidimetric titration. Because of the tendency of
lactide formation free and total acid must be de-
termined separately.
5.1.3. Derivatives
The glycolic acid esters methyl glycolate
[96-35-5], bp 147 – 149 ◦ C, and ethyl glycolate
[623-50-7], bp 158 – 159 ◦ C, both serve as start-
ing materials for the laboratory preparation of
pure glycolic acid [18] and were formerly em-
ployed as solvents for resins and for nitro- or
acetyl cellulose. Apart from these two materi-
als, only carboxymethyl cellulose [9004-32-4],
(→ Cellulose Ethers, Chap. 6. and n-butyl gly-
colate have commercial significance.
n-Butyl Glycolate (see Table 1).
Physical Properties. n-Butyl glycolate is a
colorless liquid, which is miscible with most or-
ganic solvents. Its physical properties are sum-
marized in Table 2. The solubility in water is lim-
ited to 8 wt % (20 ◦ C), although the compound
itself may contain up to 25 wt % water.
8 Hydroxycarboxylic Acids, Aliphatic
Production. n-Butyl glycolate is produced
by treatment of sodium chloroacetate with n-
butyl alcohol at 125 – 160 ◦ C, followed by vac-
uum distillation [19].
Applications. n-Butyl glycolate is used pri-
marily as a varnish additive, valued because of its
low volatility. Trade names include Polysolvan-
O (Hoechst) and GB-Ester (Wacker). It confers
smooth spreading properties and high gloss on
nitrocellulose varnish. For acetyl cellulose, it is
effective as a blush inhibitor under conditions of
high humidity. Due to its good blending proper-
ties, n-butyl glycolate is also used as an additive
in alkyd resins and oil-based paints.
Safety Considerations. n-Butyl glycolate is
not regarded as hazardous.
5.2. Hydroxypropionic Acids and
β -Propiolactone
Of the two isomeric hydroxypropionic acids,
only α-hydroxypropionic acid (→ Lactic Acid)
is optically active.
β-Hydroxypropionic Acid (Hydracrylic
Acid, 3-Hydroxypropanoic Acid) [20] (see Ta-
ble 1).
Physical Properties. The most important
physical properties of β-hydroxypropionic acid
are listed in Table 2); its dissociation constant at
25 ◦ C is 3.11×10−5 . β-Hydroxypropionic acid
forms a highly acidic syrup, which loses water
upon heating.
Preparation. β-Hydroxypropionic acid can
be prepared by alkaline hydration of acrylic acid
or by treatment of ethylene chlorohydrin with
sodium cyanide, followed by hydrolysis of the
resulting β-hydroxypropionitrile. It is also read-
ily obtained by hydrolysis of the commercially
important and easily accessible compound β-
propiolactone.
β-Propiolactone (Oxetan-2-One) [21] (see
Table 1).
Physical Properties [1]. The physical prop-
erties of β-propiolactone are summarized in Ta-
ble 2 and in the following material:
Standard heat of formation −330.0 kJ/mol
Standard heat of combustion −1422.0 kJ/mol
β-Propiolactone is a colorless, highly reac-
tive liquid, that is soluble in water, alcohol,
acetone, and chloroform (solubility in water at
25 ◦ C, 37 vol %).
Chemical Properties. β-Propiolactone re-
acts with alcohols, acid chlorides, ammonia,
and water to yield β-substituted propionic acid
derivatives. The most important characteristic
of the substance is its ability to polymerize.
This highly exothermic process occurs simply
by warming, although it is also catalyzed by
both acid and base.
Preparation. β-Propiolactone is synthe-
sized by passing equimolar amounts of ketene
and formaldehyde into either acetone or β-
propiolactone itself. The reaction is carried out
at low temperature (<20 ◦ C) with a yield of
ca. 90 % [5]. Both aluminum chloride and zinc
chloride have been employed as catalysts, and
the use of methyl borate has also been suggested.
Applications. As late as 1974, β-
propiolactone was used in the United States
in the preparation of acrylic acid and acrylate
esters [22]. Today, its principal significance is
as a reactive intermediate in organic syntheses;
a small amount is treated with ammonia to pro-
vide β-alanine. β-Propiolactone was also used
as a disinfectant. It appeared to be an attractive
replacement for formaldehyde due to its 25-fold
greater disinfecting power, but it has since been
abandoned because of its carcinogenic proper-
ties.
Safety Considerations. In the EEC haz-
ardous materials guideline and in the Gef Stoff V
(Federal Republic of Germany) β-propiolactone
is classed as a strong carcinogen and is con-
sidered extremely dangerous at concentrations
≥1 wt %.
5.3. Hydroxybutyric Acids
Of the three structural isomers α-, β-, and γ-hy-
droxybutyric acid, only the α- and β-acids are
optically active. Racemic mixtures of each can
be separated into optical antipodes by way of
the corresponding strychnine or brucine salts.

All three acids are soluble in water, etha-
nol, diethyl ether, and numerous organic sol-
vents, and decompose readily on dry distilla-
tion. β-Hydroxybutyric acid easily loses water
to give crotonic acid, and γ-hydroxybutyric acid
is transformed even at low temperature to γ-
butyrolactone.
(R,S-)-α-Hydroxybutyric Acid [(R,S-
)-2-Hydroxybutanoic Acid] [23], (see Ta-
ble 1). The physical properties of (R,S)-α-hy-
droxybutyric acid are listed in Table 2.
Preparation. (R,S)-α-Hydroxybutyric acid
results as main product from treating butyr-
aldehyde with alkaline sodium hypochlorite
solution or from heating α-bromobutyric acid
with formamide at 150 ◦ C.
(R,S-)-β-Hydroxybutyric Acid [(R,S-)-3-
Hydroxybutanoic Acid] [24], (see Table 1).
Physical Properties. The mp of racemic β-hy-
droxybutyric acid is 48 – 50 ◦ C; it is commonly
described in the literature as a syrupy liquid (see
Table 2). Pure (R)-(−)-β-hydroxybutyric acid
[625-72-9] is a solid with mp 49 – 50 ◦ C.
Preparation. (R,S)-β-Hydroxybutyric acid
results from hydrolysis of β-butyrolactone,
which in turn can be produced by the addi-
tion of ketene to acetaldehyde in the presence of
boron trifluoride or zinc salts [5]. The compound
is also prepared by reduction of acetoacetic acid
with sodium amalgam in alkaline solution.
γ-Hydroxybutyric Acid (4-Hydroxybutanoic
Acid) [25] (see Table 1).
Physical Properties. The most important
physical properties of γ-hydroxybutyric acid
are listed in Table 2. The dissociation constant
at 25 ◦ C is 1.93×10−5 . At ambient temperature,
the acid is a liquid and consists in part of lactone.
Preparation. γ-Hydroxybutyric acid is ob-
tained from γ-butyrolactone by alkaline hydro-
lysis and isolated as the alkali-metal salt. γ-
Butyrolactone is produced by dehydrocycliza-
tion of 1,4-butanediol.
Applications. Hydroxybutyric acids were
utilized in the manufacture of plasticizers, but
their current applications are quite limited. The
related lactones are some times used as starting
materials or intermediates in organic synthesis.
Optically active β-hydroxybutyric acid
is an important pharmacological agent
Hydroxycarboxylic Acids, Aliphatic 9
[26]. Biodegradable poly(−(R)-(−)-β-hy-
droxybutyric acids (named PHB) have been
used in the context of controlled long-term
parenteral application of drugs through im-
plantation. These materials offer a number
of advantages relative to other implantation
polymers [e.g., poly(methacrylic acid) deriva-
tives, poly(glycolic acid), poly(lactic acid)] [27].
Poly(β-hydroxybutyric acid) has also been con-
sidered as a potential textile material.
γ-Hydroxybutyric acid has been used as an
anesthetic and pain killer. It has a stabilizing ef-
fect on blood pressure and induces a state of un-
consciousness, characteristics that have encour-
aged its application especially in geriatrics [28].
5.4. (R,S)-Malic Acid
[(R,S)-Hydroxysuccinic Acid,
(R,S)-Hydroxybutanedioic Acid] [29],
[30](see Table 1.)
Physical Properties (see Table 2). Malic
acid is a colorless, odorless, crystalline sub-
stance. It is highly soluble in water (100 g
of water dissolves 126.3 g of malic acid at
20 ◦ C), methanol, ethanol, acetone, ether, and
other polar solvents. Because it is a dibasic
acid, it has two dissociation constants: at 20 ◦ C,
K A1 = 3.9×10−4 and K A2 = 1.4×10−5 .
Preparation. (R,S)-Malic acid is prepared
commercially in the United States and Canada
by hydration of maleic anhydride (→ Maleic and
Fumaric Acids) [31]. The sole manufacturer in
the United States is Alberta Gas, with an annual
capacity of ca. 5000 t. In this process maleic acid
is heated at ca. 180 ◦ C (under a pressure of ca.
1 MPa), malic acid is yielded as the main prod-
uct. Byproducts are maleic and fumaric acids.
The latter can be separated by filtration and re-
turned to the process stream because of its low
water solubility.
The filtrate is then concentrated; this causes sep-
aration of the malic acid, which is purified by
multiple washings, evaporation, and recrystal-
lization until the contents of fumaric and maleic
10 Hydroxycarboxylic Acids, Aliphatic
acids are reduced to 7.5 and <500 ppm, respec-
tively. Additional purification is required to pre-
pare pharmacological-grade material [32], [33].
Applications. (R,S)-Malic acid closely re-
sembles both citric and tartaric acids in its phys-
ical and chemical properties. However, the com-
pound has a more neutral flavor, so it is of-
ten preferred when the flavor of citric acid is
considered objectionable. Examples include the
impregnation of packing material for foodstuffs
such as cheese or the acidification required dur-
ing preparation of baked goods. It is widely used
in the food industry as an acidulant and, to a
lesser degree, as an acidity regulator. Relative
to citric or tartaric acid, the flavor imparted by
malic acid is not only more neutral but also
of longer duration. Thus, citric acid imparts a
strongly acidic taste to the tongue almost imme-
diately, but the effect is brief. With malic acid,
the sensation is less intense, but it lasts longer
and is more consistent. Soft-drink manufacturers
have discovered that this property helps mask
the aftertaste associated with certain nonnutri-
tive sweeteners and leads to a more balanced
flavor profile. Moreover, synergistic effects are
observed between these sweeteners and malic
acid, which permit reductions of up to 20 % in
the amount of sweetener required and 10 % in
malic acid. Consequently, significant economic
benefits may be realized. The economic picture
is also influenced favorably by the fact that malic
acid is available as an anhydrous powder.
Malic acid is used primarily as an ingredient
in hard candys and other sweets, jams, jellies,
and various canned fruits and vegetables. Most
countries authorize its use as an additive in food-
stuffs.
(R)-(+) -Malic acid [636-61-3] and (S)-(−)-
malic acid [97-67-6] are isolated from natural
sources by resolution of (R,S)-malic acid with
the aid of an optically active base. (S)-(−)-Malic
acid is also available through microbiological
fermentation of fumaric acid [33]. The levorota-
tory S-enantiomer is widely distributed in fruit
and displays the following characteristics that
differ from those of the racemate: mp, 100 ◦ C;
d 20 18 ◦
4 , 1.595; αD , −2.3 (7 wt % in H2 O). The
compound is soluble to the extent of 36.4 g in
100 g of water at 20 ◦ C.
6. Toxicology
As previously noted, most important aliphatic
hydroxycarboxylic acids occur as natural prod-
ucts. Because they serve as intermediates in
plant and animal metabolic pathways, they do
not exhibit many exceptionally toxic properties.
Thus, only a few compounds listed in Table 1 are
included in the Register of Toxic Effects of Chem-
ical Substances prepared by NIOSH [35]. Cer-
tain hydroxycarboxylic acids (e.g., lactic, malic,
tartaric, and citric) are even accepted as food ad-
ditives or preservatives in most countries.
The following toxicity data for citric acid are
taken from the NIOSH list [35, p. 945]:
LD50 (oral, rat) 11 700 mg/kg
LD50 (oral, mouse) 5 040 mg/kg
LD50 (oral, rabbit) 7 000 mg/kg
Glycolic acid was also permitted as a food
preservative but later came to be considered sus-
pect [36]. Today, it is described as moderately
toxic, with the following data [35, p. 386],
LD50 (oral, rat) 950 mg/kg
LD50 (oral, guinea pig) 920 mg/kg
A derivative of glycolic acid n-butyl glycolate
is described as displaying narcotic properties. It
is also an irritant to both skin and mucous mem-
branes. Nevertheless, it is markedly less toxic
than glycolic acid, having an LD50 (oral, rat) of
4595 mg/kg.
The lowest reported lethal dose for β-hy-
droxypropionic acid ( hydracrylic acid) derives
from intravenous injection into rats: LD50 (i.v.,
rat) = 50 mg/kg [35, p. 447].
A derivative of β-hydroxypropionic acid β-
propiolactone is of considerable interest because
of its viricidal and bactericidal properties, and
was used medically in sterilization and disinfec-
tion. Nevertheless, it has since been found to
be a mutagen [37]. Animal studies have pro-
vided incontrovertible evidence that the com-
pound displays carcinogenic properties irrespec-
tive of whether it is delivered orally, topically
on the skin, or by intravenous, intraperitoneal,
or subcutaneous injection [38], [35, p. 1326]. β-
Propiolactone is a powerful irritant to the skin
and eyes. The following toxicity data have been
reported:
 Hydroxycarboxylic Acids, Aliphatic
LC50 (inhalation, rat) 25 ppm, 6 h
LD50 (i.v., mouse) 345 mg/kg
Because of its cytotoxic properties, β-
propiolactone has been added to Category A 2
of the official list of carcinogenic materials [39].
By contrast, γ-butyrolactone is completely
free of these cytotoxic characteristics. Despite
numerous animal studies, no evidence has been
obtained to suggest carcinogenicity. Nonethe-
less, the compound is moderately toxic, hav-
ing an LD50 (oral, rat) of 1800 mg/kg [35,
p. 946]. The narcotic effects observed when γ-
butyrolactone is administered orally or intraperi-
toneally have been ascribed to γ-hydroxybutyric
acid, from which it is derived.
From animal studies, the sodium salt of γ-hy-
droxybutyric acid is known to possess narcotic
properties [40]. A dose of ca. 40 mg/kg leads to
nearly natural sleep in humans, with essentially
no side effects.
7. References
1. H. E. Zaugg: “β-Lactones,” Org. React. (N.Y.)
8 (1954) 305 ff.
2. A. Griesbeck, D. Seebach, Helv. Chim. Acta
70 (1987) 1320, 1326.
3. Jaroslawl. Technological Institute,
DE-OS 2 326 297, 1974 (M. S. Rusakowa, et
al.).
4. R. L. Shriner: “The Reformatsky Reaction,”
Org. React. (N.Y.) 1 (1942) 1 ff.
5. Goodrich Co., US 2 356 459, 1941 (F. E.
Kung); US 2 424 589, 2 424 590, 1944 (T. R.
Steadman).
6. Escambia Chem. Corp., US 2 811 545, 1956
(T. R. Steadman).
7. K. Niemela, E. Sjostrom, Carbohydr. Res. 144
(1985) no. 1, 87 – 92.
8. K, Niemela, E. Sjostrom, Holzforschung 40
(1986) no. 1, 9 – 14.
9. K. Garves in J. F. Kennedy, (ed.): Cellulose Its
Derivatives, Ellis Horwood, Chichester 1985,
pp. 487 – 494.
10. Beilstein, 3, 228; 3 (1), 88; 3 (2), 167; 3 (3),
367; 3 (4), 571.
11. Hoechst, DE-OS 2 904 754, 1980 (H. Baltes,
E. I. Leupold, F. Wunder).
12. Hoechst, DE-OS 2 812 682, 1979 (E. I.
Leupold, B. Wojtech, H. J. Arpe).
11
13. Hoechst, DE-OS 2 810 975, 1979 (H. Klug, H.
Woppert, H. Korbanka).
14. DuPont, US 2 152 852, 1939 (D. J. Loder);
US 2 153 064, 1939 (A. T. Larson);
US 2 443 482, 1948 (M. T. Shattuck).
15. Degussa, DE 194 038, 1903.
16. I. G. Farbenind., DE 459 602, 1925 (O.
Schmidt).
17. E. J. E. Eegriwe, Z. Anal. Chem. 89 (1932)
124.
18. DuPont, US 2 331 094, 1940 (D. J. Loder).
19. I. G. Farbenind., DE 522 786, 1929 (W.
Bülow).
20. Beilstein, 3, 295; 3 (1), 122; 3 (2), 212; 3 (3),
522; 3 (4), 689.
21. Beilstein, 17 (1), 130; 17/5 (3/4), 4157; 17/9
(5), 3.
22. Ullmann, 4th ed., 19, p. 452.
23. Beilstein, 3, 302; 3 (1), 114; 3 (2), 216; 3 (3),
561; 3 (4), 755.
24. Beilstein, 3, 308; 3 (1), 116; 3 (2), 220; 3 (3),
571; 3 (4), 761.
25. Beilstein, 3, 311; 3 (2), 222; 3 (3), 581; 3 (4),
774.
26. D. Seebach, M. Züger, Helv. Chim. Acta 65
(1982) 495 – 503.
27. W. Korsatko, B. Wabnegg, G. Braunegg, H. M.
Tillian et al., Pharm. Ind. 45 (1983) no. 5,
525 – 527; no. 10, 1004 – 1007; Pharm. Ind. 46
(1984) no. 9, 952 – 954.
28. W. Bushart, P. Rittmeyer: Anästhesie mit
γ-Hydroxybuttersäure, Springer-Verlag,
Berlin 1973.
29. Beilstein, 3, 435; 3 (1), 154; 3 (2), 289; 3 (3),
918; 3 (4), 1124.
30. H. Rudy: Fruchtsäuren, Wissenschaft und
Technik, Hüthig-Verlag, Heidelberg 1967.
31. T. Yoshida, O. Hibino, S. Morita, Nippon
Kagaku Kaishi (1921 – 47) 53 (1950),
399 – 401; Nippon Chem. Ind. JP 4 360, 1950.
32. Chemie Linz, FR 2 193 810, 1973.
33. Allied Chem. Corp., DE-AS 1 518 522, 1965
(Winstrom, O. Leon, Ingleman, Miltin Russel).
34. Hüls, DE-OS 3 434 880 A1, 1986; Supplement
to DE-OS 3 310 849.
35. Registry of Toxic Effects of Chemical
Substances, 1983 – 84. Cumulative
Supplement to the 1981 – 82 Edition NIOSH.
U.S. Department of Health and Human
Services Public Health Service, Centers for
Disease Control. National Institute for
Occupational Safety and Health.
36. W. F. Riker, G. Gold, J. Am. Pharm. Assoc.
Sci. Ed. 31 (1942) 306.
12 Hydroxycarboxylic Acids, Aliphatic
37. H. H. Smith, A. M. Srb, “Induction of
mutations with β-propiolacton,” Science 114
(1951) 490.
38. D. J. Brusick, Mutat. Res. 39 (1977) no. 3 – 4,
241 –255.
39. D. Henschler, Vors. d. Arbeitsst.-Komm. d.
Dtsch. Forschungsgemeinsch.:
Gesundheitsschädliche Arbeitsstoffe,
Toxikologisch-arbeitsmedizinische
Begründung von MAK-Werten, vol. IV, 14.
Lieferung, VCH-Verlagsgesellschaft
Weinheim, 1988.
40. H. Laborit et al., JAMA J. Am. med. Assoc.
175 (1961) 520.
 Hydroxycarboxylic Acids, Aromatic 1

Hydroxycarboxylic Acids, Aromatic

Salicylic Acid is a separate keyword; Naphtholcarboxylic Acids see → Naphthalene Derivatives
Edwin Ritzer, Bayer AG, Leverkusen, Federal Republic of Germany
Rudolf Sundermann, Bayer AG, Leverkusen, Federal Republic of Germany

1. Introduction . . . . . . . . . . . . . . . 1 4.1.2. 4-Hydroxybenzoic Acid . . . . . . . . . 4

2. General Properties . . . . . . . . . . . 2 4.1.3. Cresotic Acids . . . . . . . . . . . . . . . 5
3. Production . . . . . . . . . . . . . . . . 2 4.2. Dihydroxybenzoic Acids . . . . . . . 6
4. Individual Aromatic Hydroxycar- 4.3. Trihydroxybenzoic Acids . . . . . . . 6
boxylic Acids . . . . . . . . . . . . . . . 3 4.4. Mandelic Acid . . . . . . . . . . . . . . 7
4.1. Monohydroxycarboxylic Acids . . . 3 5. Toxicology . . . . . . . . . . . . . . . . . 8
4.1.1. 3-Hydroxybenzoic Acid . . . . . . . . . 3 6. References . . . . . . . . . . . . . . . . . 8

1. Introduction in the production of dyes and plastics is increas-

Aromatic hydroxycarboxylic acids (i.e., hy-
droxycarboxylic acids with phenolic hydroxyl
groups) occur widely in nature in the form
of their esters, ethers, intermolecular esters (
depsides), or cyclic ether – esters (depsidones).
Structural formulas of the most important dep-
sides and of depsidone are given below:
ingly important, especially in color developing,
and in liquid crystal polymers (→Liquid Crys-
tals).
2. General Properties
Monohydroxybenzoic acids are dibasic acids.
Thus, one mole of alkali forms the carboxyl-
ate and a second mole converts the phenolic
hydroxyl group into a phenolate. The acidity
of each group depends on its relative position
(see Table 1). The carboxyl and hydroxyl group
acidities of 3- and 4-hydroxybenzoic acids are
comparable to those of benzoic acid and phe-
nol. However, the corresponding acidities of 2-
hydroxybenzoic acid (salicylic acid [69-72-7])
are different. The higher acidity of the carboxyl-
ate group in this case is the result of intramolec-
ular chelate formation.

Table 1. Dependence of dissociation constants for aromatic

hydroxycarboxylic acids on relative position of hydroxyl and
carboxyl groups

K 1 (−COOH), K 2 (−OH),
×10−5 ×10−10
Hydrolyzable tanning agents such as tannin
(ester of gallic acid, tannic acid) [5424-20] are Benzoic acid [65-85-0] 6.3
esters of aromatic hydroxycarboxylic acids with Phenol [108-95-2] 9.9
2-Hydroxy-
polyhydric alcohols and sugars. Aromatic hy- benzoic acid [69-72-7] 105 0.0004
droxycarboxylic acids are used mainly for phar- 3-Hydroxy-
maceutical and cosmetic purposes: as disinfec- benzoic acid [99-06-9] 8.3 1.17
tants, preservatives, emulsifiers, analgesics, and 4-Hydroxy-
benzoic acid [99-96-7] 2.9 4.8
antirheumatic and antipyretic drugs. Their use

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 519
2 Hydroxycarboxylic Acids, Aromatic
This accounts for its high steam volatility
and ease of sublimation. 2-Hydroxy- and 4-
hydroxybenzoic acids decarboxylate above their
melting point; 2,4- and 2,6-dihydroxybenzoic
acids decarboxylate even when heated in an
aqueous solution. This process is catalyzed by
strong acid. The carboxyl group is esterified
with alcohols under acid catalysis without de-
tectable etherization of the phenolic hydroxyl
group. Ether – esters can be obtained easily from
the dialkali salts by using alkyl halides or dialkyl
sulfates. Alkaline hydrolysis of ether – esters
produces alkoxycarboxylic acids. Electrophilic
substitution (e.g., nitration, sulfonation, chlori-
nation) leads to attack of the aromatic nucleus
at the 2- or 4-position, with the 4-position pre-
ferred.
Analysis. Methods of analyzing carboxyl-
ic acids and phenols are also applicable to hy-
droxycarboxylic acids. In practice, alkalimetric
titration (in special cases, potentiometric titra-
tion) is used. With potentiometric determination
of the equivalence point, the hydroxyl group can
also be determined. Trace analyses are carried
out by colorimetric methods with iron(III) chlo-
ride [4]. Thin-layer chromatography is a suit-
able method for checking purity. Here, indi-
vidual acids are made visible by means of al-
kaline (Na2 CO3 ) coupling with diazotized 4-
nitroaniline, by spraying with iron(III) chloride,
or by fluorescence in UV light. For quantitative
analysis, HPLC methods are suitable. Alterna-
tively, the components can be derivatized and
capillary gas chromatographic methods used to
separate them [5].
3. Production
Kolbe – Schmitt or Marassé Carboxyla-
tion. Yields according to the Kolbe reaction of
solid anhydrous alkali phenolates in a stream
of carbon dioxide were improved by Schmitt
who carried out the reaction under pressure [6].
The Marassé method involves the conversion of
phenol with potassium carbonate and does not
work with less expensive sodium carbonate [7].
The Kolbe – Schmitt synthesis is a gas-solid re-
action; thus the phenolate must be present as a
fine powder rather than a melt to achieve better
contact. For the same reason, inert high boil-
ers, such as higher alkanes, are used for thor-
ough mixing [8]. The alkali phenolates are con-
verted into alkali-metal salts of the correspond-
ing hydroxycarboxylic acids, generally by stir-
ring, in pressurized vessels at 120 – 180 ◦ C and
a carbon dioxide pressure of 0.5 MPa. Condi-
tions up to 260 ◦ C and 120 MPa are required
in the presence of electron-withdrawing groups
or in the case of sterically hindered phenols. 3-
Nitrophenol [555-84-7] is carboxylated with a
yield of only 19 % [9]. Carboxylation occurs eas-
ily when the aromatic compounds are activated
by electron-donating substituents, as in the case
of aminophenols or di- and trihydroxyaromatic
compounds. Reaction then occurs even in an
aqueous alkali-metal carbonate solution.
The yield of dicarboxylic acid increases
with increasing temperature and pressure above
1.0 MPa. At 10.0 MPa and 220 ◦ C the equilib-
rium is shifted mainly toward the product [10].
The type of alkali metal is important for the
course of the reaction. Sodium and lithium phe-
nolates are carboxylated at medium tempera-
ture (150 – 180 ◦ C) to form mainly 2-hydroxy-
carboxylic acids. Potassium phenolates give 2 –
4 mixtures at medium temperature, but above
220 ◦ C the 4-hydroxycarboxylic acid is favored.
Exchange of a Sulfonic Acid, a Halide, or
a Diazonium Group for the Hydroxyl Group.
The most common method for producing 3-
hydroxybenzoic acid [99-06-9] involves the al-
kali melt of 3-sulfobenzoic acid [121-53-9] [11].
In some cases, if the carboxyl group is acti-
vated by electron-donating substituents, decar-
boxylation takes place during melting. Nu-
cleophilic substitution of halogen by the hy-
droxyl group is particularly effective if the halo-
gen atom is activated by electron-withdrawing
groups in the ortho or para position [12]. The
conversion of aminobenzoic acids into hydroxy-
carboxylic acids is possible via the diazonium
salt [13].
Hydroxycarboxylic Acids via Oxidation.
Oxidation of alkylphenols is another method of
producing aromatic hydroxycarboxylic acids. In
general the phenolic hydroxyl group must be
 Hydroxycarboxylic Acids, Aromatic
protected (e.g., by esterification with sulfuric
acid). The alkyl moiety of the sulfuric acid es-
ter is then oxidized easily by using potassium
permanganate or chromic acid [14]. Aromatic
hydroxycarboxylic acids can also be produced
by oxidative alkali melt of appropriately substi-
tuted phenols. Alkyl groups ortho to the phenolic
hydroxyl group are oxidized most easily [15].
The oxidative alkali melt of aromatic hydroxy-
aldehydes gives the corresponding hydroxycar-
boxylic acids in good yield [16]. In the presence
of silver catalysts, oxidation is possible even in
aqueous alkali solution [17]. Hydroxyacetophe-
nones can be converted into hydroxycarboxylic
acids by using iodine in pyridine (haloform re-
action), followed by alkaline hydrolysis of the
intermediate pyridinium iodide [18].
Direct introduction of a hydroxyl group into
the aromatic nucleus by oxidation is possible in
the case of ammonium benzoates by using hy-
drogen peroxide [19]. Thus, the first stage of the
Dow process (production of phenols from ben-
zoic acids) involves hydroxylation by air of the
metal benzoate ortho to the carboxyl group [20].
Introduction of a carboxyl group is achieved
by reaction of phenols with tetrachloromethane
in the presence of alkali and powdered copper
or copper salts [21]. Alkyl ethers of hydroxy-
carboxylic acids can be obtained directly by car-
boxylation of alkoxyphenyl metal halides (e.g.,
magnesium or lithium compounds) with carbon
dioxide. Amides and thioamides of alkoxycar-
boxylic acids are formed by conversion of phe-
nol ethers with carbamoyl chlorides, isocya-
nates, or isothiocyanates in the presence of alu-
minum chloride [22]. Rearrangement of arylur-
ethanes in the presence of Lewis acids (e.g., alu-
minum chloride) may give good yields of the
amides of the corresponding hydroxycarboxylic
acids [23].
4. Individual Aromatic
Hydroxycarboxylic Acids
4.1. Monohydroxycarboxylic Acids
4.1.1. 3-Hydroxybenzoic Acid
Physical Properties. 3-Hydroxybenzoic
acid [99-06-9], C7 H6 O3 , M r 138.12, crystal-
lizes as white needles (from water) or rhombic
3
prisms (from alcohol). Important physical prop-
erties follow: mp 203 ◦ C; d 25 1.473; dissocia-
tion constants K1 8.71×10−5 , K 2 1.18×10−10
(at 19 ◦ C); solubility (in 100 g of solution): wa-
ter 6.11 g (69 ◦ C), 98 % ethanol 39.6 g (65 ◦ C),
and n-butanol 20.7 g (36.5 ◦ C).
Chemical Properties. 3-Hydroxybenzoic
acid does not undergo decarboxylation at ele-
vated temperature and is stable up to 300 ◦ C.
This is in contrast to 2- and 4-hydroxybenzoic
acids. Electrophilic substitution (e.g., nitration,
halogenation, sulfonation) occurs ortho or para
to the hydroxyl group. Thus, nitration with 62 %
aqueous nitric acid gives 2-nitro-3-hydroxyben-
zoic acid as the main product, along with 4-nitro-
and 6-nitro-3-hydroxybenzoic acids [24].
Production. 3-Hydroxybenzoic acid can be
obtained by alkali melt of the sodium salt of
3-sulfobenzoic acid at 210 – 220 ◦ C. The crude
acid precipitates on acidification. It can be puri-
fied by recrystallization from water with the ad-
dition of activated carbon [11]. 3-Hydroxyben-
zoic acid can also be obtained from 3-nitroben-
zoic acid esters. This occurs via catalytic hydro-
genation followed by diazotization of the ami-
nobenzoic acid ester formed. The diazonium salt
is then heated [25] with aqueous sulfuric acid.
Another production method is oxidation of 3-
hydroxybenzaldehyde [100-83-4] by bubbling
air through a hot aqueous sodium hydroxide sus-
pension [26].
Uses. 3-Hydroxybenzoic acid is used as
an intermediate in the manufacture of pharma-
ceuticals and pesticides. The 3-hydroxybenzo-
ic acid ester of tropine is used in ophthalmol-
ogy to dilate the pupils. The sodium salt of 3-
hydroxybenzoic acid is a cholepoietic agent. Es-
ters and metal salts have germicidal and preser-
vative properties, and are used in food preser-
vation. Ether – esters are suitable as plasticizers
for vinyl and cellulose resins.
4.1.2. 4-Hydroxybenzoic Acid
Physical Properties. 4-Hydroxybenzoic
acid [99-96-7] C7 H6 O3 , M r 138.12, is a white
granular crystalline powder consisting of mono-
4 Hydroxycarboxylic Acids, Aromatic
Table 2. Properties of 4-hydroxybenzoic acid esters

Methyl Ethyl n-Propyl n-Butyl Benzyl

[99-76-3] [120-47-8] [94-13-3] [94-26-8] [94-18-8]

Physical properties
Mr 152.12 166.17 180.2 194.22 228.24
mp, ◦ C 131 118 97 70 111
Solubility at 25 ◦ C in 100 g of
Water 0.25 0.17 0.05 0.02 0.006
Water (80 ◦ C) 2.0 0.86 0.30 0.15 0.09
Methanol 59 115 124 220 79
Ethanol 52 70 95 210 72
Acetone 64 84 105 240 102
Benzene 0.7 1.65 3.0 40 2.6
Diethyl ether 23 43 50 150 42
Tetrachloromethane 0.1 0.9 0.8 1.0 0.08
Bacteriostatic properties
Phenol = 1.0 3.8 8.0 17.0 32.0 109.0

clinic prisms. The physical properties are sum-
marized here: mp 216.3 ◦ C; d 20 1.497; dissoci-
ation constants K 1 3.3×10−5 , K 2 4.8×10−10
(19 ◦ C); solubility (in 100 g of solution): wa-
ter 0.49 g (20 ◦ C), 33.5 g (100 ◦ C); 99 % ethanol
38.75 g (67 ◦ C); and n-butanol 19.5 g (32.5 ◦ C).
4-Hydroxybenzoic acid is sparingly soluble in
chloroform.
Table 2 lists properties of the most commonly
used esters.
Chemical Properties. Above 200 ◦ C, 4-
hydroxybenzoic acid undergoes decarbox-
ylation to produce phenol and carbon diox-
ide. Electrophilic substitution (e.g., nitration,
sulfonation, or halogenation) occurs ortho to
the hydroxyl group. With iron(III) chloride, no
color reaction occurs but a yellow precipitate is
formed.
Production. 4-Hydroxybenzoic acid is pro-
duced by carboxylation of potassium phenolate
above 220 ◦ C and with a carbon dioxide pres-
sure of 0.45 MPa. The yield of 4-hydroxyben-
zoic acid (partly present as the dipotassium salt)
is 80 % of the theoretical amount in which the
conversion of phenolate is 60 %. 4-Hydroxyben-
zoic acid can also be obtained by isomeriza-
tion of dipotassium salicylate [27] at temper-
atures up to 240 ◦ C and under carbon diox-
ide pressure (slightly above atmospheric pres-
sure) [28], or by oxidation of the alkali-metal
salt of p-cresol [106-44-5] on metal oxides
at 260 – 270 ◦ C [29]. Reaction of phenol with
tetrachloromethane [56-23-5] in the presence
of alkali produces salicylic acid [69-72-7] as
the main product and smaller amounts of 4-
hydroxybenzoic acid [99-96-7] [21].
Uses. 4-Hydroxybenzoic acid is an interme-
diate in the manufacture of dyes, pharmaceuti-
cals, pesticides [30], and plastics. It is used as
an emulsifier and also as a corrosion-protection
agent [31]. Its esters and corresponding sodium
salts are antimicrobial substances with a high
no-effect level (1200 mg per kilogram of body
weight) [32]. They are, therefore, used under
weakly acidic to weakly basic conditions as food
preservatives (e.g., in fruit juice), cosmetics, and
pharmaceuticals, as well as in technical products
(e.g., in lubricants, anti-freezing agents) [33].
Benzyl esters are relatively soluble in prepa-
rations containing sugar. Table 2 lists observed
bacteriostatic effects [34] compared to phenol.
4-Hydroxybenzoic acid is also used in poly-
mers as a component in the manufacture of
polyester and as a constituent of liquid crystal
polymers [35] (→ Liquid Crystals).
Esters. The food preservative “ PHB-Ester”
is a mixture of the ethyl and n-propyl es-
ters of 4-hydroxybenzoic acid. According to
the preservatives regulations in the different
countries [EEC guideline 64/52 date: 5. 11. 63,
24th amendment 85/585 date: 20. 12. 85; United
States: Food and Drug Administration, HHS,
21 CFR Ch.I, 4. 1. 86, § 172.145, § 184.1490,
§
184.1670; Federal Republic of Germany:
Verordnung über die Zulassung von Zusatzstof-
fen zu Lebensmitteln, date: 22. 12. 81 (BGBl.
 Hydroxycarboxylic Acids, Aromatic
I, p. 1625/33) 1. amendment date: 20. 12. 84
(BGBl. I, p. 1652)], PHB-Ester in the amount
of 0.1 – 10 g/kg is allowed in the preservation of
foodstuffs. In Europe, PHB-Esters are identified
as E 214 – E 219, in the United States as heptyl-
paraben, methylparaben and propylparaben. A
mixture of 60 % ethyl and 40 % propyl ester is
particularly effective. Use of such preservatives
is, however, restricted by law in selected food-
stuffs (e.g., wine) [36]. Some higher esters are
used for termite control [37].
The PHB-Esters that are solids as pure com-
ponents give eutectic mixtures which are liquid
at room temperature and are used as oil-in-water
emulsions (→ Emulsions) [38]. See Table 2.
4-Hydroxybenzoic acid esters, particularly
PHB-benzyl ester, are also employed as color
developers in pressure-sensitive [39] and ther-
mosensitive [40] writing systems. Here, purifi-
cation and decoloration of the PHB-Ester are
achieved through distillation with the addition
of sodium hydrogen sulfite [41].
4.1.3. Cresotic Acids
Of the ten isomeric cresotic acids, C8 H8 O3 , M r
152.15, o- and p-cresotic acids are of most in-
terest.
o-Cresotic acid [83-40-9] ( 2-hydroxy-
3-toluic acid, 2-hydroxy-3-methylbenzoic
acid), mp 165 ◦ C, dissociation constant K 1
6.76×10−4 , is produced by carboxylation of
the sodium salt of o-cresol [95-48-7] at 150 –
180 ◦ C. The acid is sparingly soluble in water
and has properties similar to salicylic acid. It
is used as an intermediate in the preparation of
pharmaceuticals and dyes. Its methyl ester is
used as an aid in textile dyeing.
m-Cresotic Acid [50-85-1] ( 2-hydroxy-4-
toluic acid, 2-hydroxy-4-methylbenzoic acid,
4-methylsalicylic acid), mp 177 ◦ C, dissocia-
tion constant K 1 6.84×10−4 , is produced by
carboxylation of the sodium salt of m-cresol
5
[108-39-4]. It is used as an additive in galvanic
baths [42].
p-Cresotic Acid [89-56-5] ( 2-hydroxy-5-
methylbenzoic acid, p-homosalicylic acid), mp
153 ◦ C, is produced by carboxylation of the
sodium salt of p-cresol [106-44-5] with carbon
dioxide at 150 – 180 ◦ C. The acid is used as an
intermediate in dye production.
4.2. Dihydroxybenzoic Acids
3,4-Dihydroxy- and 2,3-dihydroxybenzoic
acids are both obtained by heating catechol
[120-80-9] in aqueous ammonium carbonate
solution at 140 ◦ C.
3,4-Dihydroxybenzoic acid [99-50-3] (pro-
tocatechuic acid), C7 H6 O4 , M r 154.1, mp
203 ◦ C, undergoes decomposition to catechol
[120-80-9] and carbon dioxide. The acid crystal-
lizes as a monohydrate from aqueous solution.
The most important derivative of this
acid is the monomethyl ether, 4-hydroxy-3-
methoxybenzoic acid [121-34-6] (vanillic acid),
C8 H8 O4 , M r 168.14, mp 211 ◦ C (from water).
It is obtained by careful alkali melt of vanillin
[121-33-5] or oxidation of vanillin with silver
oxide [43]. Vanillic acid and its esters are con-
stituents of preservatives and disinfectants.
2,3-Dihydroxybenzoic acid [303-38-8],
C7 H6 O4 , M r 154.1, mp 206 ◦ C, undergoes
decomposition to catechol and carbon diox-
ide. The acid crystallizes from water as a dihy-
drate. Among other uses, it serves as an iron-
complexing agent in drugs [44].
2,4-Dihydroxybenzoic acid [89-86-1] (p-
hydroxysalicylic acid, β-resorcylic acid),
C7 H6 O4 , M r 154.1, mp 226 ◦ C (decomp.),
is sparingly soluble in water. The dissocia-
tion constant K 1 is 6.05×10−4 (30 ◦ C). The
acid is produced by conversion of resorcinol
[108-46-3] with aqueous potassium carbonate
solution at 100 ◦ C and a carbon dioxide pressure
of 0.45 MPa. In this case, 2,6-dihydroxybenzoic
6 Hydroxycarboxylic Acids, Aromatic
acid [303-07-1] (γ-resorcylic acid) is formed
as byproduct and can be separated by recrys-
tallization [45]. Pure 2,4-dihydroxybenzoic
acid can be produced by oxidation of 2,4-di-
hydroxyacetophenone [89-84-9] with iodine in
pyridine [18].
2,4-Dihydroxybenzoic acid is used as an in-
termediate in the manufacture of dyes, pharma-
ceuticals and of additives in photography, and
cosmetics. Dyes made from resorcylic acids dif-
fer from those made from salicylic acid in their
deeper and more intense colors.
The acid is used to complex titanium(IV) flu-
oride [46], in the passivation of aluminum [47],
and as an additive in drilling fluids [48].
3,5-Dihydroxybenzoic acid [99-10-5] (α-
resorcylic acid), C7 H6 O4 , M r 154.1, mp 237 ◦ C
(decomp.), is obtained by alkali melt of 3,5-
disulfobenzoic acid [49].
2,5-Dihydroxybenzoic acid [490-79-9] (gen-
tisic acid, 2,5-DHBA), C7 H6 O4 , M r 154.1, mp
205 ◦ C, is soluble in water and alcohols but in-
soluble in benzene and chloroform. The acid is
stable in boiling water and gives a deep blue
color with iron(III) chloride. Here, the acid de-
composes to form p-benzoquinone [106-51-4]
and carbon dioxide.
2,5-Dihydroxybenzoic acid is obtained by
carboxylation of hydroquinone [123-31-9] with
aqueous potassium hydrogen carbonate solution
at 130 ◦ C. It can also be made by oxidation
of salicylic acid with potassium peroxodisulfate
[7727-21-7] in the presence of iron sulfate [50].
2,5-Dihydroxybenzoic acid is not obtained by
reaction of carbon dioxide with the dialkali
salt of hydroquinone: instead hydroquinone-2,5-
dicarboxylic acid is formed.
2-5-Dihydroxybenzoic acid is used as an an-
tirheumatic drug and, more recently, as a color
developer for thermal paper [51].
4.3. Trihydroxybenzoic Acids
3,4,5-Trihydroxybenzoic Acid [149-91-7].
Physical Properties. 3,4,5-Trihydroxyben-
zoic acid (gallic acid), C7 H6 O5 , M r 170.12,
is a white to pale yellow crystalline powder,
which crystallizes from water in the form of
silky needles as the monohydrate (mp 258 –
263 ◦ C) and undergoes decomposition to pyro-
gallol [87-66-1] and carbon dioxde: d 25 1.694,
dissociation constants K 1 4.63×10−3 (at 30 ◦ C)
and K 2 1.41×10−9 . 3,4,5-Trihydroxybenzoic
acid is soluble in warm water, ethanol, diethyl
ether, and acetone but insoluble in benzene and
chloroform.
Chemical Properties. Solutions of gallic
acid, particularly of the alkali-metal salts, absorb
oxygen and turn brown in air like pyrogallol. The
acid is a strong reducing agent that can reduce
gold or silver salts to the metal. With iron(III)
salts, an intensive blue complex is formed that
is used for ink dyes. Gallnut ink, which consists
of gallic acid and iron(II) sulfate, produces the
blue iron(III) – gallic acid complex on paper in
air.
When gallic acid is heated with con-
centrated sulfuric acid, hexahydroxyanthra-
quinone [82-12-2] ( rufigallic acid) is produced
by condensation. Reaction with p-nitrosodi-
methylaniline hydrogen chloride produces ox-
azine derivatives (→ Azine Dyes, Chap. 3.2.,
last paragraph). Oxidation with arsenic acid, per-
manganate, persulfate, or iodine yields ellagic
acid [476-66-4], as does reaction of methyl gal-
late with iron(III) chloride. The carboxyl group
undergoes azeotropic esterification with alco-
hols, or, by the Fischer method, with alcoholic
hydrochloric acid. The ether – ester of gallic acid
is obtained by conversion of dialkyl sulfates or
alkyl halides in the presence of alkali, or by using
diazomethane [334-88-3]. The hydroxyl group
in the 4-position is the most reactive. 3,4,5-
Trimethoxybenzoic acid [118-41-2] undergoes
partial hydrolysis in strong acid to form 4-
hydroxy-3,5-dimethoxybenzoic acid [530-57-4]
( syringic acid).
Production. Gallic acid, a component of
many tanning agents is an endogenous product
in plants. The acid occurs free or bound to tannin
[1401-55-4] (e.g., in divi-divi, oak bark, gall-
nuts, pomegranate roots, sumac, and tea). The
acid is produced from tannin-rich aqueous gall-
nut extracts by acidic or alkaline hydrolysis. It
 Hydroxycarboxylic Acids, Aromatic
is also obtained by using the enzyme tannase
[9025-71-2] or molds (Penicillium glaucum, As-
pergillus niger) to cleave tannin by fermenta-
tion. Both the metabolism of gallic acid and its
impact on plant growth enzymes have been stud-
ied [52].
Uses. Gallic acid is used for the manu-
facture of iron gallnut ink (see above) and
as an intermediate in the production of dyes
(e.g., anthragallol [620-64-2], gallocyanine
[1562-85-2], galloflavin [568-80-9], and rufigal-
lic acid [82-12-2]). Gallate esters, particularly
methyl gallate [99-24-1] ( gallicin), mp 157 ◦ C,
and propyl gallate [121-79-9], mp 150 ◦ C, are
used as antioxidants and food preservatives (e.g.,
for fats) [53]. Gallic acid is also employed as
a reducing agent for pharmaceutical purposes (
Dermatol, Airol, bismuth salt of gallic acid) [54]
and for the production of photographic develop-
ers and pyrogallol.
4.4. Mandelic Acid
Racemic (R,S)-Mandelic Acid [611-72-3].
Physical Properties. (R,S)- Mandelic acid
(α-hydroxyphenylacetic acid, phenylglycolic
acid), C6 H5 −CH(OH)−COOH, C8 H8 O3 , M r
152.15, mp 118 – 121 ◦ C, is a colorless crys-
talline powder. It is readily soluble in ethanol
and diethyl ether, less soluble in chloroform, and
insoluble in petroleum ether. Its water solubility
is 15 g per 100 mL of water. In alkaline solution,
mandelic acid dissolves to form salts.
Chemical Properties. Mandelic acid turns
brown when exposed to light. It can be esterified
easily with alcohols in the presence of hydrogen
chloride.
Production. (R,S)-mandelic acid is pro-
duced by hydrolysis of mandelic acid nitrile
[532-28-5] (mp 10 ◦ C) with hydrochloric acid.
Mandelic acid nitrile is obtained by the conver-
sion of benzaldehyde [100-52-7] and hydrogen
cyanide in the nascent state (NaCN + HCl) un-
der refrigeration. Another method of produc-
ing mandelic acid is the hydrolysis of α,α-di-
chloroacetophenone with alkali.
Uses. Mandelic acid esters have analgesic,
antirheumatic, and spasmolytic effects. For ex-
amples, cis-3,5,5-trimethylcyclohexyl mande-
late [456-59-7] ( cyclandelate, Spasmocyclon,
Clan-dilon), C17 H24 O3 , M r 276.38, mp 54 –
7
58 ◦ C, is used as a spasmolytic and to stimulate
blood circulation. Mandelic acid tropine ester
[51-56-9] (homatropine – HBr), C16 H22 BrNO3 ,
M r 356.26, mp 212 – 215 ◦ C, is readily soluble in
water and ethanol, and is used in ophthalmology
to dilate the pupils. Hexamethylenetetramine
mandelate [587-23-5] ( mandelamine, Man-
dropine, diuramine, hexydaline) C14 H20 N4 O3 ,
M r 292.34, mp 127 – 130 ◦ C, forms colorless
crystals that are readily soluble in diethyl ether.
This salt is used to treat urinary tract infections.
Levorotatory (R)(−)-Mandelic Acid
[611-71-2]. (R)-Mandelic acid forms color-
less crystals, mp 132 – 135 ◦ C, specific rotation
[α]20 ◦
D (c = 5) −154 to 157 C. It is produced
by resolution of racemic (R,S)-mandelic acid
with cinchonine [118-10-5] in aqueous solu-
tion, whereby the salt of (S)-(+)-mandelic acid
crystallizes. If resolution is performed with S-
(+)-aminodiol [3306-06-7], (R)-(−)-mandelic
acid crystallizes, while (S)-(+)-mandelic acid
[17199-29-0] remains in solution. (R)-(−)- and
(S)-(+)-mandelic acid can again be subjected to
racemic resolution after isomerization.
5. Toxicology
Toxicological and physiological properties of
the different aromatic hydroxycarboxylic acids
are summarized in the following material.
3-Hydroxybenzoic Acid. LD50 3700 mg/kg
(rat, i.p.), TDLo 400 mg/kg (rat, s.c., 11 d preg-
nant); classification: drug, teratogen [55, ; date:
1/84].
4-Hydroxybenzoic Acid. LD50 2200 mg/kg
(mouse, oral), LD50 210 mg/kg (mouse, i.p.),
LD50 1050 mg/kg (mouse, s.c.); classification:
drug, mutagen [55, ; date: 8/83].
o-Cresotic Acid. LD50 1000 mg/kg (mouse,
oral). LD50 345 mg/kg (mouse, i.v.); classifica-
tion: drug [55, ; date: 6/85].
m-Cresotic Acid. LD50 1800 mg/kg (mouse,
oral); classification: drug [55, ; date: 6/85].
p-Cresotic Acid. LD50 1000 mg/kg (mouse,
oral); classification: drug [55, ; date: 6/85].
4-Hydroxy-3-methoxybenzoic Acid. Classi-
fication: mutagen [55, ; date: 8/82].
2,4-Dihydroxybenzoic Acid. LDLo 642 mg/kg
(rat, s.c., 11 d pregnant); classification: teratogen
[55, ; date: 3/84].
 8 Hydroxycarboxylic Acids, Aromatic
3,5-Dihydroxybenzoic Acid. LD50 2000 mg/kg
(mouse, i.v.) [55, ; date: 6/84].
2,5-Dihydroxybenzoic Acid. 2,5 Dihydroxyben-
zoic acid is an intermediate in the metabolism of
salicyclic acid [56]: LD50 800 mg/kg (rat, oral)
LD50 3000 mg/kg (rat, i.p.), LD50 374 mg/kg
(mouse, i.v.), LDLo 380 mg/kg (rat, s.c., 9 d
pregnant), LDLo 642 mg/kg (rat, s.c., 11 d pre-
gant); classification: teratogen, natural product
[55, ; date: 5/84].
3,4,5-Trihydroxybenzoic Acid. LD50
320 mg/kg (mouse, i.v.), LD50 5000 mg/kg (rab-
bit, oral), LDLo 800 mg/kg (mouse, i.p.), LDLo
5 mg/kg (mouse, s.c., 1 d pregnant); classifica-
tion: agricultural chemical, mutagen, teratogen
[55, ; date: 3/85].
6. References
General References
1. Beilstein 10, 43, 134, 149, 192, 214, 375, 464,
10 (1), 20, 63, 68, 83, 95, 173, 232, 10 (2), 25,
79, 88, 112, 248, 331, 10 (3), 87, 422, 1363,
2057, 10 (4), 125, 315, 345, 1414, 1971.
2. E. Rodd: Chemistry of Carbon-Compounds,
vol. 3, Elsevier, Amsterdam-London-New
York 1956, pp. 756 – 815.
3. A. S. Lindey, H. Jeskey, Chem. Rev. 57 (1957)
583 – 620. Houben Weyl 8,372 – 377.
Specific References
4. J. E. Heestermann, Chem. Weekbl. 32 (1935)
463.
5. C. G. Barroso, R. C. Torrijos, J. A.
Perez-Bustamante, Chromatographia 17
(1983) no. 5, 249 – 252. R. Shinohara,
Suishitsu Odaku Kenkyu 7 (1984) no. 9,
549 – 552; Chem. Abstr. 102 (1985) no. 26,
225 717 p.
6. H. Kolbe, Justus Liebigs Ann. Chem. 113
(1860) 125 –127. H. Kolbe, E. Lautermann,
Justus Liebigs Ann. Chem. 115 (1860)
157 – 206. H. Kolbe, J. Prakt. Chem. 10
(1874) no. 2, 89 – 111. O. Hartmann, J. Prakt.
Chem. 16 (1877) no. 2, 35 – 59. H. Kolbe, DE
426, 1877. F. W. v. Heyden, DE 29 939, 1884;
31 240, 1884(R. Schmitt). F. W. v. Heyden, DE
33 635, 1885 (R. Schmitt). F. W. v. Heyden,
DE 38 742, 1886 (R. Seifert). S. Marassé, DE
73 279, 1893.
7. D. Cameron, H. Jeskey, O. Baine, J. Org.
Chem. 15 (1950) 233 – 236. O. Baine, J. Org.
Chem. 19 (1954) 510 – 514.
8. Rhône-Poulenc, DE-OS 27 54 239, 1976 (J. C.
Choulet, J. Nouvel).
9. F. Wessely, Monatsh. Chem. 81 (1950)
1071 – 1091.
10. S. E. Hunt, J. I. Jones, A. S. Lindsey, Chem.
Ind. (London) 1955, 417 – 418.
11. M. F. Clarke, L. N. Owen, J. Chem. Soc. 1950,
2108 – 2115.
12. K. W. Rosenmund, H. Harms, Ber. Dtsch.
Chem. Ges. 53 (1920) 2226 – 2240.
13. F. F. Blicke, F. D. Smith, J. L. Powers, J. Am.
Chem. Soc. 54 (1932) 1460 – 1471.
14. B. Heymann, W. König, Ber. Dtsch. Chem.
Ges. 19 (1886) 704 – 706.
15. O. Jakobson, Ber. Dtsch. Chem. Ges. 11
(1878) 376, 570.
16. G. Lock, Ber. Dtsch. Chem. Ges. 62 (1929)
1177 – 1188.
17. I. A. Pearl, J. Am. Chem. Soc. 67 (1945)
1628 – 1629. I. A. Pearl, J. Am. Chem. Soc. 68
(1946) 429 – 432. I. A. Pearl, J. Org. Chem. 12
(1947) 79 – 84.
18. L. C. King, M. McWhirter, D. M. Barton, J.
Am. Chem. Soc. 67 (1945) 2089 – 2095.
19. H. D. Dakin, M. D. Herter, J. Biol. Chem. 3
(1907) 419 – 434.
20. G. Navazio, A. Scipioni, Chim. Ind. (Milan)
50 (1968) 1086 – 1090.
21. J. Zelter, M. Landau, DE 258 887, 1912.
22. L. Gattermann, Justus Liebigs Ann. Chem.
244 (1888) 29 – 76.
23. G. J. Gershzon, J. Gen. Chem. USSR (Engl.
Transl.) 13 (1943) 82.
24. H. Schäfer, Memorial des Poudres 27 (1937)
153, 156, 157.
25. H. E. Ungnade, A. S. Henick, J. Am. Chem.
Soc. 64 (1942) 1737 – 1738.
26. I. G. Farben, DE 506 438, 1926 (F. Hans);
Friedländer 17, 472.
27. H. Kupferberg, J. Prakt. Chem. 16 (1877) no.
2, 424 – 447.
28. C. A. Buehler, W. E. Cate, Org. Synth. Coll. 2
(1943) 341 – 343. Dow Chemical, US
1 937 477, 1932 (L. E. Milk, W. W. Allen).
S. E. Hunt, J. Chem. Soc. 1958, 3152 – 3160.
A. J. Rostron, A. M. Spivey, J. Chem. Soc.
1964, 3092 – 3096.
29. P. Friedländer, O. Löw-Beer, DT 170 230,
1905.
30. G. W. K. Cavill, J. M. Vincent, J. Chem. Soc.
Ind. 66 (1947) 175 – 182.
31. BASF Wyandotte, US 4 234 440, 1979 (St. T.
Hirozawa, E. F. O’Brien, J. C. Wilson).
32. G. M. Cramer et al., Food, Cosmet. Toxicol. 16
(1978) 255.
 Hydroxycarboxylic Acids, Aromatic
33. A. J. Maga, K. Lorenz, Cereala Sci. Today 18
(1973) no. 10, 326 – 328, 350; Chem. Abstr. 84
(1976) no. 23, 163 164 e. H. P. Fiedler: Lexikon
der Hilfsstoffe für Pharmazie, Kosmetik und
angrenzende Gebiete, 2nd ed., Editio Cantor
Aulendorf 1981, pp. 480 ff.
34. F. J. Bandelin, Drug Cosmet. Ind. 64 (1949)
430.
35. F. Hoffmann-La Roche, DE-OS 2 447 098,
1974 (A. Boller, H. Scherrer). Hitachi Ltd., EP
110 299, 1983 (S. Hattori, K. Iwasaki, T.
Kitamura, A. Muko et al.). Chisso Corp.,
DE-AS 2 751 403, 1977 (T. Inukai, H. Sato, K.
Yokohama).
36. C. S. Ough, Chem. Eng. News 65 (1987)
19 – 28.
37. Yoshitomi Pharmaceutical Industries, JP
62 76 651, 1987 (T. Inoue, M. Horie, T. Sato).
38. Mallinckrodt Inc., US 4 309 564, 1980 (D. F.
Loncrini, J. J. Taylor).
39. Mitsubishi Paper Mills Ltd., JP 58-138 689, JP
58-151 292, JP 58-8 686, EP-AP 86638, 1983
(H. Tsukahara, T. Torii).
40. Ueno Pharma Co., JP 58-158 289, 1983 (R.
Keno, H. Tsuchiya, S. Itou, T. Tsuchida).
Honshupaper Co., JP 58-158 290, 1983 (M.
Jsoda, S. Nishimatsu, K. Kuniyama). Fuji
Photo Film Co., JP 58-87 089, 1983 (M.
Watanabe). Konzaki Paper Mfg Co. Ltd., JP-P
210 057-82, 1982 (H. A. T. Suzuki, R. O.
Hayashi, N. I. O. Arai, Y. I. N. Oeda).
41. Mitsui Toatsu Kagaku KK, JP 49-19 275, 1974
(R. Fujihara, K. Nitta).
42. Siemens, DE-OS 3 049 417, 1980 (H.
Januschkowetz, H. Laub).
43. I. A. Pearl, J. Am. Chem. Soc. 68 (1946)
2180 – 2181. I. A. Pearl, J. Am. Chem. Soc. 68
(1946) 1100 – 1101.
Hydroxyhydroquinone → Phenol Derivatives
9
44. A. Cerami, J. Graziano, R. W. Grady, C. M.
Peterson, Proc. Symp.: Dev. Iron Chelaters
Clin. Use 1975, 261 – 273; Chem. Abstr. 87
(1977) no. 7, 47 789 a.
45. K. Brunner, Justus Liebigs Ann. Chem. 351
(1907) 313 – 331.
46. Beecham Inc., US 4 601 898, US 4 601 899,
1985 (R. E. Stier, W. H. Dunn, J. D. Vidra).
47. Kaiser Aluminium & Chemical Corp., US
3 714 000, 1971 (G. A. Dorsey, Jr.).
48. Texaco Inc., US 3 537 992, 1967 (J. H.
Kolaian).
49. L. Barth, K. Senhofer, Justus Liebigs Ann.
Chem. 159 (1871) 216 – 230.
50. Chem. Fabr. AG vorm. E. Schering, DE
81 297, 1894. F. Mauthner, J. Prakt. Chem.
156 (1940) 150 – 153.
51. Fuji Photo Film Co., GB 1 360 581, 1971 (E.
Katsusuke, I. Sadao).
52. E. A. Haddock, Phytochemistry 21 (1982)
no. 5, 1049 – 1062. E. Hasalam, Fortschr.
Chem. Org. Naturst. 41 (1982) 1 – 46.
S. Kobayashi, M. Ariga, T. Ozawa, H.
Imagawa, Agric. Biol. Chem. 48 (1984) no. 2,
389 – 395.
53. J. W. Thompson, E. R. Sherwin, J. Am. Oil
Chem. Soc. 43 (1966) no. 12, 683 – 686.
54. P. Pfeiffer, E. Schmitz, Pharmazie 5 (1950)
517 – 520.
55. Cis. Inc., 7215 York Road, Baltimore, MD
21212, CIS-Data bank (RTECS and
OHMTADS).
56. D. R. Boreham, B. K. Martin, Br. J.
Pharmacol. 37 (1969) no. 1, 294 – 300; Chem.
Abstr. 71 (1969) 121 980 u.

Hypnotics
Hartmund Wollweber, Bayer AG, Wuppertal, Federal Republic of Germany
1. Introduction . . . . . . . . . . . . . . . . .
1.1. Amino Acids and Hormones Influenc-
ing Sleep . . . . . . . . . . . . . . . . . . .
1.2. Influence of Hypnotics on Sleep . . . .
1.3. Requirements of Hypnotics . . . . . . .
1.4. Applications of Hypnotics . . . . . . . .
1.5. Consumption of Hypnotics . . . . . . .
1.6. Classification and Evaluation of Hyp-
notics . . . . . . . . . . . . . . . . . . . . .
2. Sedatives, Antihistamines, and Tran-
quilizers . . . . . . . . . . . . . . . . . . . .
3. Alcohols and Aldehydes . . . . . . . . .
1. Introduction
Hypnotics (Greek υπνoζ, sleep) are drugs that
promote the onset of sleep and induce a state
similar to natural sleep by depressing the cen-
tral nervous system [1–28]. Hypnotics are not
a sharply defined group of drugs because their
sleep-inducing effect greatly depends on the
dose. Small doses of a hypnotic have a sedative
action; increasing doses cause, in succession, a
hypnotic effect, a narcotic effect, and finally se-
vere respiratory depression and death.
Sleep that is properly induced by drugs ap-
pears to resemble physiological sleep closely.
Both states share the following symptoms:
– reduction of heart rate
– reduction of metabolism
– decrease in concentration of calcium ions in
blood
– decrease in muscle tone and blood pressure
– contraction of bronchi and
– contraction of pupils
In contrast to narcotized persons, sleeping
persons can be awakened.
Physiology of Sleep. Marked differences
between drug-induced sleep and physiologi-
cal sleep can, however, be detected by using
polygraphic methods, such as measurement of
pH changes and eye movements (electrooculo-
gram), electromyogram, and electroencephalo-
gram (EEG). On the basis of data obtained from

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a13 533
Hypnotics 1
1 4. Urethanes, Acyclic Urea Derivatives,
and Amides . . . . . . . . . . . . . . . . . 8
2 5. Cyclic Carboxamides, Carboximides,
3 Cyclic Urea Derivatives, and Related
4 Compounds . . . . . . . . . . . . . . . . . 9
4 5.1. Barbituric Acids . . . . . . . . . . . . . . 10
4 5.2. Piperidinediones . . . . . . . . . . . . . . 15
5.3. Quinazolinones . . . . . . . . . . . . . . . 17
4
5.4. Benzodiazepines . . . . . . . . . . . . . . 18
5 6. Other Hypnotics . . . . . . . . . . . . . . 27
6 7. References . . . . . . . . . . . . . . . . . . 28
these methods, two types of sleep of importance
for the restoration of the body have been de-
scribed:
1) Orthodox, synchronized sleep or nonrapid
eye movement (NREM, nonREM) sleep can
be distinguished by δ waves corresponding
to periods of reduced activity on the EEG (δ
sleep). Since these waves are slow, this phase
is also referred to as slow wave sleep (SWS).
2) Paradox sleep or rapid eye movement (REM)
sleep has high EEG activity.
The two types of sleep form a cycle of 110 –
140 min; four to five cycles are observed in
adults each night (Fig. 1). Sleep is extremely im-
portant for the renewal of strength. The main
phases of natural sleep are an initial deep, dream-
less sleep, during which the body temperature
falls, and a light, dream-filled, restless sleep be-
fore waking. The depth of orthodox sleep has
four levels, as seen on the EEG (Fig. 1A), which
can partly be correlated with certain body func-
tions [10].
The areas of the brain that regulate body tem-
perature and the sleep – wake cycle are close
to each other and are functionally related. The
modulations of sleep are correlated with the cir-
cadian rhythm of body temperature. The on-
set of sleep at night coincides with a lowering
of body temperature, characteristic of NREM
sleep. People who display a temperature differ-
ence of less than 1 ◦ C are short sleepers and those
2 Hypnotics
Figure 1. Different phases of normal adult sleep (ca. 8 h)
A) Electroencephalogram; B) Depth of sleep; C) Body temperature; D) Body movement
who display larger differences (>1 ◦ C) are long
sleepers. Body tissues are regenerated during or-
thodox sleep, and cerebral functions during the
REM phase.
The REM phases account for ca. 20 – 25 %
of sleep and appear to be periods of relatively
high cerebral activity in which dreams occur.
They are detected by measuring the spontaneous
movements of the closed eye. In the absence of
these dream phases sleep is not restful. Dreams
are of great importance for learning and mem-
ory; they assimilate the experiences of the pre-
ceding days. Generally, dreams can be remem-
bered slightly or not at all. With increasing age,
the duration of REM sleep steadily decreases
to about 20 – 30 % of its original duration. The
REM phase is especially activated when a per-
son is confronted with an emotionally arousing,
intellectual task calling for creative effort, i.e.,
when alterations in thinking or behavior are nec-
essary. Confrontation with problems also inten-
sifies REM sleep, e.g., an astronaut experiences
greatly increased REM sleep in his first night
in space. Continuous suppression of REM sleep
leads to psychological disturbances and psycho-
somatic symptoms.
1.1. Amino Acids and Hormones
Influencing Sleep
Three neurohormones are involved in control-
ling sleep: serotonin, noradrenaline, and acetyl-
choline. Serotonin regulates orthodox sleep,
whereas acetylcholine and noradrenaline influ-
ence the REM phase. The rates of cell division
are the highest during sleep, and large amounts
of growth hormone (GH), which stimulates pro-

tein synthesis, are secreted during the third and
fourth levels of orthodox sleep. The EEG waves
at this stage are also large and slow.
Melatonin is rhythmically secreted by the
pineal gland. It stimulates specific receptors in
the body’s biological clock, the suprachiasmatic
nuclei of the hypothalamus and may thus regu-
late the human circadian rhythm by increasing
serotonin levels. The human biological clock can
be “adjusted” with exogenous melatonin. Conti-
nous exposure of humans to light leads to sleep
disorders by inhibiting melatonin synthesis [16],
[29], [30].
A sleep inducing nonapeptide, delta-sleep-
inducing peptide [62568-57-4] (DSIP), has been
isolated from rabbit serum. This peptide, which
also occurs in humans, has been synthesized
and has been prepared by genetic engineering
techniques. DSIP coordinates the rhythmic cir-
cadian sleep – wake cycles by promoting sleep
functions (without having a sedative effect), and
enhances mental performance when awake. At
the same time, DSIP invokes a higher tolerance
to stress. The action of this nonapeptide is at-
tributed to modulation of the adrenergic recep-
tors [31].
l-Tryptophan, a common amino acid found
in the human body is a precursor of serotonin. In
daily doses of 4 – 7 g it causes sedation during
the day and a shortening of latency of sleep at
night. It is registered in Switzerland as a “bio-
logical sleep-inducing drug.” However, studies
have questioned the sleep-modulating effect of
this amino acid [32].
1.2. Influence of Hypnotics on Sleep
Most hypnotic drugs shorten both phases of
sleep and suppress the paradoxical (REM)
phase. Thus, the sleep curve is levelled out, and
the EEG resembles that of a narcotized per-
son. Hence, drug-induced sleep does not have
the restorative quality of natural sleep and is
no substitute for the treatment of insomnia. A
research report of the Max Planck Institut für
Psychiatrie (München, 1980) states that conven-
tional sleep-inducing drugs do not produce rest-
ful sleep; since drug-induced sleep is not iden-
tical to natural sleep, taking sleeping pills is not
advised. Pharmacologists also recommend the
use of hypnotic drugs only for a few days, if at
Hypnotics 3
all, because of their many adverse effects. Alco-
hol should not be taken with sedatives or hyp-
notics because the reactions of the patient are
then adversely affected.
Prolonged use of sedatives and hypnotics re-
duces their effect on the cells of the central ner-
vous system (see also Section 5.1). Most hyp-
notics, particularly highly active preparations,
can create psychological and physical depen-
dence, resulting in withdrawal symptoms when
treatment is terminated. In addition, many hyp-
notic drugs that suppress and shorten the REM
phase produce a rebound effect, i.e., a sudden
withdrawal after extended periods of use causing
a prolongation of the REM phase. As in alcohol
withdrawal, this effect can even lead to delirium
tremens. In this way, the body tries to make up
for lost REM sleep.
1.3. Requirements of Hypnotics
There is still no completely satisfactory hypnotic
drug. The following requirements can be defined
for an ideal hypnotic:
– reliable effect
– wide therapeutic range
– limited duration of action, no longer than 8 h
– physiological sleep
– no disturbance of deep sleep or REM sleep
– low toxicity
– no enzyme induction
– no respiratory depression
– no additive effect with alcohol
– no amnesia, confusion, or ataxia after arousal
in the night
– no excessive daytime sedation
– no loss of potency on repeated use
– no risk of dependency, e.g., insomnia after
termination of treatment
– no after effects (e.g., hangover)
1.4. Applications of Hypnotics
Sleep-inducing drugs are used by the following
groups:
1) healthy people with demanding jobs or irreg-
ular working hours, who often take sleeping
pills after stimulants such as coffee;
4 Hypnotics
2) sick people with symptoms such as fever,
sore throat, or cold;
3) old people with insomnia due to cerebral
sclerosis usually have an inverted day – night
rhythm; they are awake and mobile at night,
and tired and peaceful in the morning.
4) sick people with uncomfortable and sleep-
less nights due to pain, who have difficulty
in breathing, heart trouble, etc.; and
5) bedridden people with serious illnesses, who
are deprived of physical exercise.
People belonging to groups 1 and 2 can be
treated with mild relaxants by reducing the sus-
ceptibility of the reticular formation (site of the
center of wakefulness in the brain) to external
stimuli. These substances include plant compo-
nents with sedative properties, such as valerian
and hop, and tranquilizers or sedatives which,
even in high doses, do not have narcotic prop-
erties. People belonging to groups 4 and 5 may
have to be treated for extended periods with hyp-
notics and even with antipsychotic and antide-
pressant drugs.
1.5. Consumption of Hypnotics
The importance of hypnotic drugs is reflected in
the fact that 21 % of the population of France
(February 1975) and 32 % of the inhabitants of
Los Angeles (1973) suffered from insomnia. In
France, 2.5×106 tablets, capsules, and dragées
are taken every evening to promote sleep [33–
38]. The dramatic increase in the consumption
of hypnotics with age can only be partly justi-
fied because older people require less sleep at
night. Instead, they have a more pronounced en-
dogenous sleep profile with increased cerebral
readiness to sleep at 9.00, 13.00, and 17.00 h.
1.6. Classification and Evaluation of
Hypnotics
Hypnotics may be classified as follows [34]:
1) benzodiazepine hypnotics;
2) nonbenzodiazepine hypnotics, e.g., chloral
hydrate, barbiturates, piperidinediones, and
methaqualone;
3) other drugs with sedative effects, (e.g.,
tranquilizers, antidepressants, and antihis-
tamines); and
4) natural preparations, e.g., valerian, ethanol,
and DSIP.
There is no sharp distinction between seda-
tives and hypnotic drugs; differences in activity
are quantitative rather than qualitative. Both in-
hibit the activation system in the reticular for-
mation. Drugs belonging to groups 1 and 2 are
true hypnotics. Drugs belonging to group 3 are
only occasionally used as hypnotics.
Der Bewertende Arzneimittelindex (The Val-
uating Drug Index) assesses “benzodiazepine
sleep-inducing drugs” positively, especially be-
cause of their wide therapeutic range [37]. Other
recommended drugs are chloral and its deriva-
tives (see Sections page 7page 7), diphenhy-
dramine, and doxylamine (see Sectionpage 5).
Derivatives of barbituric acid, piperidinediones,
and quinazolinones are not recommended. In
comparison with the benzodiazepines, these
cause more undesirable effects (respiratory de-
pression, negative influence on REM sleep, en-
zyme induction, and problems of dependence).
Their therapeutic range is narrow. Compounds
containing bromine are also to be avoided be-
cause they may cause bromism (see Chap. 2).
Valerian preparations (e.g., Valepotriate) and
their components are not favored because of
their in vitro cytotoxic and alkylating action. The
therapeutic importance of these drugs cannot be
assessed at present [37].
The lists of trade names cited for individual
hypnotics are not complete.
2. Sedatives, Antihistamines, and
Tranquilizers
Sedatives and Tranquilizers. There is no
significant distinction between sedatives and
hypnotics; the same drugs are often used for both
purposes. Sedatives (→ Sedatives) and tranquil-
izers (→ Psychopharmacological Agents) are
used to sedate and relieve anxiety during the day
and, in addition, to induce sleep at night. Thera-
peutic doses of tranquilizers have a calming, re-
laxing effect and alleviate psychological symp-
toms such as fear, anxiety, restlessness, and ten-
sion. Ideally, they should not produce sedative

or hypnotic effects. The first therapeutically used
hypnotics were plant extracts with either seda-
tive (valerian root and hop) or analgesic action
(opium alkaloids). In spite of the sedative effect
of opium alkaloids [39], they should not be used
for obvious reasons (addiction, euphoria, poor
dose adjustment, and legality). Of all the plant
extracts available, only one substance, valmane,
is exactly defined from a chemical standpoint.
Valepotriate is a mixture of valtratum
[18296-44-1] (1), acevaltratum [25161-41-5]
(2), and didrovaltratum [18296-45-2] (3):
These three compounds, which occur in na-
ture in the root of Valeriana officinalis, have a
sedative, mildly psychotropic action. They act as
mild hypnotic agents because they restore physi-
ological equilibrium in vegetative disorders. The
activity of valmane has been questioned [38].
Trade Names. Valmane (Kali-Chemie, FRG),
Nevripan (Medopharm, FRG).
Simple bromine-containing sedatives
(sodium, potassium, or calcium bromide, or
the hydrobromides of magnesium glutamate
or magnesium aspartate) should not be used
under any circumstances. If given for long pe-
riods, they cause a mental condition known as
bromism, which is characterized by symptoms
such as loss of concentration and memory, in-
somnia, rash, and psychoses.
Antihistamines. The following antihis-
tamines have sedative properties and are some-
times used as hypnotics:
Meclozine or Meclizine [569-65-3], 1-
[4-chlorobenzhydryl]-4-(3-tolyl)piperazine
dihydrochloride (→Antiallergic Agents,
Chap. 2.1.4.)
Hypnotics 5
Trade Names: Calmonal (Heiden, FRG), Bonine
(Pfizer, USA).
Diphenhydramine hydrochloride [147-24-0],
2-(benzhydryloxy)-N,N-dimethylethylamine
(→Antiallergic Agents)
Trade Names: Benadryl (Parke-Davis, USA),
Halbmond (Much, FRG).
Doxylamine succinate [562-10-7], 2-di-
methylaminoethoxyphenylmethyl-2-picoline
Trade Names. Decapryn, Meteprine (Merrell,
USA), Unisom (Pfizer, USA).
Psychopharmacological Agents. Of the
large number of tricyclic antipsychotropic drugs
available, two are used as hypnotics:
Perimetazine [13093-88-4], perimet-
hazine, 1- [3-(2-methoxyphenothiazin-10-yl)-
2-methylpropyl]-4-piperidinol, C22 H28 N2 O2 S,
M r 384.54 (→Phenothiazines and Derivatives).
Trade Name. Leptryl (Roger Bellon, USA).
Perlapine [1977-11-3], 6-(4-methyl-1-piper-
azinyl)-11 H-dibenz [b,e] azepine, C19 H21 N2 ,
M r 291.40, mp 136 – 138 ◦ C.
This compound, in the form of yellow prisms,
is prepared by chlorinating 11 H-dibenz [b,e]
azepin-6-one with phosphoryl chloride to give
the imide chloride, which is subsequently treated
with N-methylpiperazine [40]. The compound
thus obtained has hypnotic properties [41]; LD50
(oral) 415 mg/kg (mouse).
Trade Names. Hypnodine (Wander, FRG;
Takeda, Japan), Pipnodine (Takeda, Japan).
3. Alcohols and Aldehydes
The discovery that after producing an initial ex-
citation phase, ethanol has a general sedative ac-
tion, led to the development of tertiary and halo-
genated alcohols as hypnotics. These substances
are rarely used today. Chloral hydrate has long-
lasting sedative and hypnotic effects because it
is metabolized in the body to 2,2,2-trichloroeth-
anol, which also has a sedative action.
6 Hypnotics
Methylpentynol [77-75-8], 3-methyl-1-
pentyn-3-ol, C6 H10 O, M r 98.14, bp 121 –
122 ◦ C, is a highly mobile liquid with a pungent
smell and burning taste. It solidifies at −30.6 ◦ C
and is soluble in almost all organic solvents. It is
produced by reacting 2-butanone with sodium
acetylide.
As a result of its short duration of action (0.5 –
1 g act for 1 – 2 h), this preparation is especially
useful in promoting sleep in cases of mild in-
somnia.
Trade Names. Atemorin (Scherer, USA), Dal-
gol, Dorison, Dormison (Schering Corp., USA),
Pentadorm (Ruef, Austria), Somnesin (British
Drug Houses, UK), Util (Leo, Denmark).
Ethchlorvynol [113-18-8], 3-
(β-chlorovinyl)-3-hydroxy-pent-1-ine,
C7 H9 OCl, M r 144.61, bp 173 – 174 ◦ C, is a liq-
uid with a pungent, aromatic smell. It darkens
slowly when exposed to air and light. The com-
pound is immiscible with water but miscible
with most organic solvents.
It is prepared by reacting 1-chloropent-1-en-3-
one with acetylene [42]. This preparation is a
useful short-term hypnotic, but abuse can lead
to habituation and addiction.
Trade Names. Roeridorm (Roerig, FRG),
Arvynol (Pfizer, USA), Placidyl, Serenesil (Ab-
bott, USA).
Paraldehyde [123-63-7], paracetaldehyde,
2,4,6-trimethyl-1,3,5-trioxane, C6 H12 O3 , M r
132.16, bp 124 ◦ C, is a colorless liquid with an
aromatic smell and unpleasant taste. One part
of paraldehyde is soluble in eight parts of water
at 25 ◦ C. For other properties and synthesis, see
→Acetaldehyde, Chap. 8.1. This substance has
an unpleasant taste and imparts a bad odor to
the breath; addiction is possible. Paraldehyde is
only used in clinics and is no longer prescribed.
Trade Names. Paraldehyde Thilo (Thilo,
FRG; Fellows, USA), Paral (O. Neal,
Jones & Feldmann, USA), Dannetène (Bottu,
France).
Chlorethate [5634-37-7], bis(2,2,2-tri-
chloroethyl)carbonate, C5 H4 Cl6 O3 , M r 324.82
is prepared by the reaction of phosgene with
2,2,2-trichloroethanol in the presence of a base
[43], [44]. The insolubility of this substance
prevents gastric irritation, which is produced
by the hypnotically active parent compound,
2,2,2-trichloroethanol [44].
Trade Name. Clorets (Smith, Kline & French,
USA).
Chloral hydrate [302-17-0], 2,2,2-
trichloro-1,1-ethanediol, C2 H3 Cl3 O2 ,
M r 165.42, mp 57 ◦ C, bp 98 ◦ C (decomp. into
chloral and water), is a crystalline substance
with a pungent odor and taste. It is readily
soluble in water, ethanol, diethyl ether, and
chloroform. For production see →Chloroacet-
aldehyde, Chap. 3.3.
Doses of 0.5 – 0.75 g rapidly produce deep
sleep, which lasts 4 – 8 h, without morning hang-
over. Standard doses of 0.5 g do not produce
REM sleep, but doses of more than 0.8 g gen-
erally have a negative influence. Adverse ef-
fects are irritation of the mucous membranes and
hepato- and nephrotoxicity. Prolonged adminis-
tration can cause habituation, and physical and
psychological addiction.
Trade Names. Ansopal (Ferrosan, Denmark),
Aquachloral (Webcon, USA), Chloradorm
(Knoll Lab., Australia), Chloraldurat (Pohl-
Boskamp, FRG), Cohidrate (Coastal, USA),
Escre (SS Pharmaceutical, Japan), H. S. Need
(Hanlon, USA), Lanchloral (Lancet, Aus-
tralia), Nervifene (Interdelta, Switzerland),
Novochlorhydrate (Novopharm, Canada),
Oradrate (Coast, USA), Somnos (Merck,
Sharp & Dohme, USA), Suppojuvent Sedante
(Juventus, Spain).
Chloral Derivatives.
Chloral betaine [2218-68-0], C7 H14 Cl3 NO4 ,
M r 282.57, mp 122.5 – 124.5 ◦ C, is prepared
from betaine hydrate and chloral hydrate [45].
This adduct has the same hypnotic properties
and contraindications as chloral hydrate, but it
does not have the unpleasant physical properties.

Trade Names. Beta-Chlor (Mead Johnson,
USA) Somilan (British Drug Houses, UK),
Somnalchlor (Schiaparelli, Italy).
Triclofos [306-52-5], 2,2,2-trichloro-
ethyl dihydrogen phosphate, C2 H4 Cl3 O4 P,
M r 229.39.
The sodium salt, triclofos sodium [7246-20-0],
C2 H3 Cl3 NaO4 P, is soluble in water. This prepa-
ration has the same physiological properties as
chloral hydrate, but it does not have the unpleas-
ant physical characteristics. The compound is
metabolized in the body mainly to the hypnot-
ically active 2,2,2-trichloroethanol. A dose of
1.5 g is equivalent to 1 g of chloral hydrate.
Trade Names. Triclos (Lakeside, USA), Triclo-
syl (Glaxo, UK), Tricloran (C.T.S. Israel).
Petrichloral [78-12-6], pentaerythritchloral,
C13 H16 Cl12 O8 , M r 725.76, mp 52 – 54 ◦ C [46].
Chloralodol [3563-58-4], chlorhexadol,
2-methyl-(2 ,2 ,2 -trichloro-1 -hydroxy-4-
ethoxy)-pentan-2-ol, C8 H15 Cl3 O3 , M r 265.58,
mp 102 – 124 ◦ C [47].
Trade Names. Lora (Wallace, USA), Mechloral,
Mecloral (Dumex, Denmark).
Chloralose [15879-93-3], α-O-(2,2,2-tri-
chloroethylidene)-1,2-α-d-glucofuranose-(R),
C8 H11 Cl3 O6 , M r 309.54, mp 187 ◦ C, [α]22
D + 19
(5 % solution in 98 % ethanol) [48].
Hypnotics 7
Trade Names. Chloralosan (Kuhlmann, France),
Somio (Sidel, France).
Carbocloral [541-79-7], N-(2,2,2-
trichloro-1-hydroxyethyl)- ethyl carbamate,
C5 H8 Cl3 NO3 , M r 236.49, mp 103 ◦ C (de-
comp.) [49].
Trade Name. Prodorm (Parke-Davis, USA).
4. Urethanes, Acyclic Urea
Derivatives, and Amides
These compounds are mild hypnotic agents.
They are generally used to promote the onset
of sleep or as daytime sedatives, often in com-
bination with weak analgesics. They are well-
absorbed and, depending on the structure, have
few side effects. Dependence on these com-
pounds occurs less frequently than on barbitu-
rates and benzodiazepines.
Urethane [51-79-6], ethyl carbamate, a mild
hypnotic agent, is no longer prescribed because
its cytostatic properties can cause agranulocyto-
sis on prolonged administration.
Ethinamate is the most important carbamate.
Some other derivatives of carbamic acid are
used as central muscle relaxants (→Skeletal
Muscle Relaxants) and as psychotropic drugs
(→Psychopharmacological Agents).
Ureides are still widely prescribed as mild
sedatives and hypnotic agents. They are used
alone and in combinations. However, bromine-
containing derivatives should be given with great
care because of the danger of bromism.
Most amides, especially those of simple
branched fatty acids, have only a mild sedative
action, but some are frequently used to promote
sleep.
Ethinamate [126-52-3], 1-ethynylcyclo-
hexylcarbamate, C9 H13 NO2 , M r 167.21,
mp 96 – 98 ◦ C, solubility in water 0.25 %, in
ethanol 35 %, and
in hexane 2 %, is produced by treating 1-
ethynylcyclohexan-1-ol with (1) carbamic acid
chloride in the presence of sodium amide [50]
or (2) with phosgene and ammonia [51].
8 Hypnotics
Trade Names. Valamin (Schering, FRG),
Valmid, Valmidate (Lilly, USA).
Hexapropymate [358-52-1], 1-(2-
propynyl)cyclohexanol carbamate, C10 H15 NO2 ,
M r 181.23, mp 99 ◦ C, is synthesized as de-
scribed for ethinamate [52].
Trade Names. Merinax (Labaz, France), Mekos
(Helsingborg, Sweden).
Carfimate [3567-38-2], 1-phenyl-2-
propynyl carbamate [53], C10 H9 NO2 ,
M r 175.18, mp 86 – 87 ◦ C, is a hypnosedative.
Trade Names. Equilium (Delagrange, France),
Nirvetil, Nirvotin (Erba, Italy).
Ectylurea [95-04-5], cis-(2-ethylcrotonoyl)
urea, C7 H12 N2 O2 , M r 156.18, mp 191 – 193 ◦ C
(also mp 158 ◦ C), is prepared by treating (2-
bromo-2-ethylbutyryl)urea with sodium hy-
droxide or by reacting carbromal with silver ox-
ide [54]. It is a sedative with a mild hypnotic
action.
Trade Names. Cronil (Farmigea, Italy), Dis-
tesol (Locatelli, Italy), Ectyl (ACO, Sweden),
Levanil (Upjohn, USA), Neuroprocin (Minerva-
Chemie, Netherlands).
Carbromal [77-65-6], 2-bromo-2,2-
diethylacetylurea, C7 H13 BrN2 O2 , M r 237.11,
mp 116 – 119 ◦ C; 1 g of this compound dissolves
in 3000 mL of water, 18 mL of ethanol, 3 mL of
chloroform, and 14 mL of ether. It is soluble
in alkaline solution and in concentrated sulfu-
ric, hydrochloric, and nitric acids. The reaction
of α-bromo-α-ethylbutyric acid bromide with
urea at 50 ◦ C produces carbromal [55]. The ef-
fects are attributed to the unchanged molecule,
even though all the bromine in the organism is
broken down to bromide. The same applies to
acecarbromal and bromisoval.
Trade Names. Adalin (Bayer, FRG), Bromadyl
(Vicario, France), Diacid (Rieswerke, Austria),
Fydalex (Boots, UK), Hyperysin (Hommel,
Switzerland).
Acecarbromal [77-66-7], acetylcarbromal,
N-acetyl-N  -(2-bromo-2-diethylacetyl)urea,
C9 H15 BrN2 O3 , M r 279.14, mp 109 ◦ C, is
slightly soluble in ethanol and ethyl acetate.
It is prepared by the acetylation of carbromal
with acetic anhydride in the presence of zinc
chloride [56]. Abuse can cause habituation and
addiction.
Trade Names. Abasin (Bayer, FRG; Winthrop,
USA), Adityl, Carbased (Mallard, USA).
Bromoisoval [496-67-3], α-bromoisoval-
erylurea, C6 H11 BrN2 O2 , M r 223.08, mp 147 –
149 ◦ C, is soluble in hot water, diethyl ether,
ethanol, and alkali. The compound is pre-
pared by the acylation of urea with α-
bromoisovalerylbromide [57].
Trade Name. Bromuvan (Treibacher Chem.
Werke, Austria), Bromovaleryl, Bromyl (ACO,
Sweden), Dagrabromyl (Dagra, Netherlands),
Dormigene (Pharmacobel, Belgium), Isobromyl
(Clin-Comar-Byla, France), Isoval (Mission,
USA), Melosan (Allied Lab., USA).
Apronalide [528-92-7], 2-allyl-2-isopro-
pylacetylurea, 1-(2-isopropyl-4-pentenoyl)-
urea, C9 H19 N2 O2 , M r 184.23, mp 194 ◦ C [58].
This compound is no longer produced in the
Federal Republic of Germany because of its
severe adverse effects (purpura hemorrhagica).
Trade Names. Dormid (Rohto, Finland), Neu-
rokin (Leerbeck & Holm, Denmark), Sedormid
(Roche, Switzerland).
Pheneturide [90-49-3], ethylphenacemide,
2-phenylbutyrylurea, C11 H14 N2 O2 , M r 206.24,
mp 149 – 150 ◦ C, is prepared by acylating urea
with the α-phenylbutyric chloride [59]. It is not
only used as a hypnotic agent, but also in the
treatment of epilepsy.
 Hypnotics 9

Trade Names. Benuride (Bengue, UK; Vinas,
Spain), Deturid (Polfa, Poland; Sapos, Switzer-
land).
Capuride [5579-13-5], 2-ethyl-3-
methylvalerylurea, C9 H18 N2 O2 , M r 186.25,
mp 172 ◦ C [60].
Trade Name. Pacinox (McNeil, USA).
Valdettamil, Novonal [512-48-1], 2,2-
diethyl-4-pentenamide, C9 H17 NO, M r 155.23,
mp 75 – 76 ◦ C, is soluble in 120 parts of water
and freely soluble in alcohol and ether [61].
Trade Names. Insomnia (ICN, USA), and
as combinations: Arantil (Hoechst, FRG),
Dentigoa (Scheurich, Switzerland).
diacylated cyclic ureas: barbiturates (8) and
thiobarbiturates (9)
N-(β-carbonyl)ethyl derivatives: 2,4-piperi-
dinediones (6)
N-iminoacyl derivatives: quinazolinones (7)
phenyl N-iminoacyl derivatives: benzodiaze-
pines (11)
All groups, except 5, 7, and 11, have a doubly
substituted carbon atom adjacent to the carbonyl
group of the amide structure. These substituents
greatly influence the action and the distribution
of the drug in the body.

Butoctamidesemisuccinate [32838-28-1],
N-(2-ethylhexyl)-3-hydroxybutyramide hy-
drogen succinate, C16 H29 NO5 , M r 315.41,
mp 36.5 ◦ C, is prepared by the ami-
nolysis of ethyl 3-hydroxybutyrate with 2-
ethylhexylamine and the subsequent formation
of the semiester of succinic acid [62].

Butoctamide is related to (1-methyl)heptyl

γ-bromoacetate, which is found in human
cerebrospinal liquid. It increases the brain 5-
hydroxytryptamine level and has an antitumor
effect. It is given in doses of 200 mg (approxi-
mately as effective as 400 mg of bromoacetyl-
urea [63].

Figure 2. Cyclic carboxamide, carboximide, and urea struc-

5. Cyclic Carboxamides,
Carboximides, Cyclic Urea
Derivatives, and Related Compounds
Most of the synthetic sleep-inducing drugs used
today have a carboxylic acid amide group (4 in
Fig. 2) incorporated into a ring structure. The
ring nitrogen atom is linked to many different
structural elements (Fig. 2).
The following groups of compounds belong
to this category:
N-acyl derivatives: cyclic imides, e.g., (5) and
2,6-piperidinediones (10)
tures used in hypnotics.
5.1. Barbituric Acids [64–66]
The first representative of this group of sub-
stances, barbituric acid, was discovered in 1845
[67]. A synthesis of the first barbiturate with
hypnotic properties, 5,5-diethylbarbituric acid
(barbital), was completed in 1882; its effects
were described in 1903. Since then about 60
different barbiturates have been introduced, of
which 20 – 25 are still used in the industrial
10 Hypnotics
Table 1. Duration of action dose, biological half-life, and metabolism of barbiturates

Compound Type ∗ Mean hypnotic Duration of Onset of Plasma Biological Proportion pK value
dose, g hypnotic action, min protein half-life, h metabolized, %
effect, h binding,
% ∗∗
Barbital L 0.25 – 0.5 10 – 24 30 – 60 2 72 – 120 < 5 7.88
Phenobarbital L 0.1 – 0.2 4 – 12 30 – 60 48 – 96 80 7.29
Butobarbital L 0.05 – 0.1 6 – 12 30 – 60 95 8.12
Probarbital L–M 0.12 – 0.25 4 – 12 30 – 60
Amobarbital M–L 0.1 – 0.3 2–8 15 – 30 15 >95
Cyclobarbital M 0.1 – 0.2 2–8 15 – 30 >99 7.80
(in 3 d 42 %)
Aprobarbital M 0.065 – 0.13 2–8 15 – 30 80 8.15
Allobarbital M 0.1 – 0.3 2–8 15 – 30 70 7.92
Vinylbital M 0.15 6–7 15 – 30 35 in 4 d
Heptabarb S–M 0.1 – 0.4 2–4 30 98 7.78
Secbutabarbital S–M 0.1 2–4 30
Pentobarbital S–M 0.1 – 0.2 2–4 30 40 15 99 7.88
(in 1 d 25 %)
Butallyonal S–M 0.2 2–4 30 >90 8.00
Talbutal S 0.12 – 0.15 2–4 30
Cyclopentobarbital S 0.12 – 0.25 2–4 30 8.10
Propallyonal S 0.1 – 0.3 2–4 30 >99 8.05
Secobarbital VS – S 0.1 – 0.2 1–3 15 45 90 7.80
Hexobarbital VS – S 0.25 – 0.4 1–4 15 2 3.5 95 8.30

∗ Duration of action: L = long; M = medium; S = short; VS = very short.

∗∗ Percentage of the barbiturate that is bound to the protein.

countries. Important properties are listed in Ta-
ble 1. As hypnotics, barbituric acids all pro-
duce almost identical pharmacodynamic ef-
fects. However, they show wide differences
in pharmacokinetics and metabolism. Differ-
ences in onset and duration of activity are re-
lated to lipid and water solubility, and rate
of metabolism. Lipophilic barbiturates, e.g.,
the narcotically active, lipid-soluble thiobarbitu-
rates, and the N-monoalkylbarbiturates such as
hexobarbital rapidly disappear in adipose tissue.
As a result of their redistribution in lipophilic
compartments of the body, these compounds
have a very short duration of action and are
rapidly broken down (short-acting narcotics,
→Anesthetics, General, Chap. 3.1.).
From a therapeutic standpoint, barbiturates
used as hypnotic agents are classified according
to their duration of action (see also Table 1):
1) long-acting compounds (barbital, phenobar-
bital),
2) medium-acting compounds (amobarbital, cy-
clobarbital),
3) short-acting compounds (hexobarbital), and
4) compounds whose effects begin after 15 –
20 min (secobarbital).
The duration of action is primarily deter-
mined by the rate of enzymatic degradation in
the liver, which depends on the chemical struc-
ture of the barbiturate. Continual use causes en-
zyme induction in the liver and thus a gradual
increase in the rate of barbiturate degradation.
Increasing doses are required with prolonged
use (barbiturate habituation). In some patients
barbiturates not only have a sedative and hyp-
notic action, but also produce excitement and
euphoria. This sense of well-being may easily
lead to barbiturate addiction, a physical depen-
dence with withdrawal symptoms. Barbiturate
addicts take the hypnotic because of its euphoric
properties and try to compensate for the central
depression by using psychoanaleptic drugs.
Cases of accidental and suicidal barbiturate
poisoning are fairly common. They are charac-
terized by unconsciousness (which can precede
a delirious stage), central respiratory depression
accompanied by oxygen deficiency symptoms,
and in severe poisoning circulatory collapse with
death in 12 h to 4 d. The prognosis for barbitu-
rate poisoning is favorable if treatment is imme-
diate. Essential measures include artificial res-
piration, irrigation of the stomach, circulatory

support, forced osmotic diuresis using manni-
tol, exchange transfusion, and hemodialysis.
Barbiturates produce a condition superfi-
cially resembling natural sleep. However, the
REM phase, which is very important for rest
and recuperation, is shortened. Continued use
of barbiturates leads to a very low REM value,
and sudden withdrawal increases the REM phase
(rebound effect). The adverse effects of barbitu-
rates are well-documented after more than 50
years of experience. This is an important advan-
tage over newly developed hypnotics [68].
The pK values of the hypnotically active bar-
bituric acids, range between 7.78 and 8.30. Dis-
ubstituted barbituric acids are present, to ca.
50 %, in the undissociated form under physio-
logical conditions (pH 7.4). Only this form of the
drug can pass through biological membranes.
The degree of dissociation is strongly influenced
by slight alterations in pH.
Production. The 5,5-disubstituted barbituric
acids are still synthesized by the classical con-
densation of substituted malonic esters (12),
malonic amides, or cyanoacetic esters (13) with
urea (14), thiourea (15), guanidine (16), or di-
cyanodiamide (17), followed by the hydrolysis
of the resulting imino- or cyanobarbiturate [69]
(Fig. 3).
Under acidic or neutral conditions, substi-
tuted malonic acids or malonic chlorides can
also be used for the condensation with urea
or thiourea. It is preferable for the substituents
R1 and R2 to be present in the malonic ester
or cyanoacetic ester molecule; disubstitution of
barbituric acids at the 5-position is only possi-
ble with highly reactive halides, such as allyl
halides.
Barbital [57-44-3], barbitone, 5,5-
diethylbarbituric acid, C8 H12 N2 O3 , M r 184.19,
mp 188 ◦ C, (sodium salt [144-02-5],
C8 H11 N2 NaO3 , Mr 206.18), has a bitter taste.
One part of the acid dissolves in 170 parts of
water at 25 ◦ C or 17 parts of water at 100 ◦ C.
It is freely soluble in acetone, ammonia, and
lye. Barbital is prepared by the condensation
of diethyl malonate with urea in the presence
of sodium ethoxide [70]. Both the acid and its
sodium salt are used as long-acting hypnotics;
abuse can cause habituation or addiction.
Hypnotics 11
Trade Names. Dormileno (Faes, Spain), Hyp-
nox (Püschel, Austria), Veronal (Bayer, FRG;
Merck, FRG), Veroletten (EAS-Labor, Austria).
Sodium salt: Barbinetten (Hormosani, Aus-
tria).
Phenobarbital [50-06-6], phenemalum,
phenobarbitone, 5-ethyl-5-phenylbarbituric
acid, C12 H12 N2 O3 , M r 232.23, mp 174 –
178 ◦ C. The acid crystallizes from water in
the form of flakes. It is odorless, with a slightly
bitter taste; an aqueous solution reddens litmus
paper. One part dissolves in 1100 parts of water
at 20 ◦ C, 40 parts of boiling water, 10 parts of
ethanol, 18 parts of diethyl ether, and 60 parts
of chloroform. It is synthesized by the reaction
of urea with ethyl phenyl malonate in sodium
ethoxide solution [71].
The sodium salt of phenobarbital [57-30-7],
C12 H11 N2 NaO3 , M r 254.23, is a colorless, crys-
talline, hygroscopic, freely water-soluble pow-
der. It is slightly soluble in ethanol, and insolu-
ble in diethyl ether and chloroform. The aqueous
solution is stable for only a short time.
Phenobarbital is a long-acting hypnotic with
anticonvulsant properties. Long-term use in
higher than therapeutic doses can cause phys-
ical addiction. An overdose resulting in blood
levels of 8 – 12 mg/100 mL can cause death.
Trade Names. Agrypnal (Eggochemia, Austria),
Aphenylbarbit (Streuli, Switzerland), Calminal
(Wolfs, Belgium), Epsylone (Powell, Canada),
Eskabarb Span (Smith, Kline & French, USA),
Fenemal (DAK, Denmark), Gardenal (Spe-
cia, Canada), Gardenale (Farmitalia Carlo
Erba, Italy), Hypnolone (Hartz, Canada),
Pepinal, Lepinaletten (Arzneimittelwerk Dres-
den, GDR), Luminal (Bayer, FRG; E. Merck,
FRG; Winthrop, USA), Mediphen (Medic,
Canada), Phen Bar (Saunders, Canada), Phe-
nobarbyl (Synochem, FRG), Seda-Tablinen
(Beiersdorf, FRG), Solfoton (Poythress, USA),
Teolaxin (Paul Maney, Canada); Phenobarbi-
tal diethylamine salt: Gratusminal (Farmasimes,
Spain); Phenobarbital calcium salt: Fenileal
(Turon, Spain), Lumcalcio (Abello, Spain).
Butobarbital [77-28-1], butobarbitone, 5-
butyl-5-ethylbarbituric acid, C10 H16 N2 O3 ,
M r 212.24, mp 124 – 127 ◦ C, forms slightly
bitter crystals; 1 g dissolves in 5 mL ethanol;
preparation is described in [72]. It is a long-
12 Hypnotics
Figure 3. Synthesis of barbiturates.
acting hypnotic; abuse can cause habituation or
addiction.
Trade Names. Butenil, Butynoct (Interpharm,
Austria), Etoval, Longanoct, Sonabarb (Protea,
Australia), Soneryl (Specia, Italy; May & Baker,
USA).
Probarbital [76-76-6], 5-ethyl-5-isopro-
pylbarbituric acid, C9 H14 N2 O3 , M r 198.22,
mp 197 – 198 ◦ C, is slightly soluble in cold
water, but more soluble in hot water and eth-
anol. Sodium salt [143-82-8], C9 H13 N2 NaO3 ,
M r 220.20, is a hygroscopic powder that is solu-
ble in water. The calcium trihydrate [545-74-4],
C18 H26 CaN4 O6 · 3 H2 O, M r 488.55, forms
slightly bitter crystals; 1 g dissolves in 40 g wa-
ter; preparation is described in [73]. Probarbital
is a long- to medium-acting hypnotic; abuse can
cause habituation or addiction.
Trade Names. Ipral Sodium, Ipral Calcium
(Squibb, USA).
Amobarbital [57-43-2], 5-ethyl-5-isopen-
tylbarbituric acid, C11 H18 N2 O3 , M r 226.28, mp
156 – 158 ◦ C, forms slightly bitter crystals; 1 g
dissolves in 1300 mL water and in 5 mL ethanol.
The sodium salt, C11 H17 N2 NaO3 [64-43-7],
M r 248.25, is hygroscopic and readily soluble
in water; preparation is described in [74], and
metabolism in [75].
Amobarbitol is a widely used medium- to
long-acting hypnotic. It resembles secobarbi-
tal and pentobarbital in action with a some-
what longer duration. A blood level of 3 –
6 mg/100 mL of blood is fatal; abuse can cause
habituation or addiction.
Trade Names. Amsal (Adams, Australia), Amy-
cal (AFI, Sweden), Amydorm, Amytal (Lilly,
USA), Isobec (Pharbec, Canada), Isonal (ICN,
Canada), Neur-Amyl (Fawns & McAllen, Aus-
tralia), Placidel (Miquel, Spain). Stadadorm
(Stada, FRG).
Amobarbital sodium: Altinal, Amsal
(Adams, Australia), Amsebarb (Paul Maney,
Canada), Amylbarb Sodium (Protea, Aus-
tralia), Amylobeta (Bramble, Australia), Amy-
tal Sodium (Lilly, USA), Barbamyl (Teva,
Israel), Dorminal (Boots, UK), Neur-Amyl
(Fawns & McAllen, Australia), Novamobarb
(Novopharm, Canada).
Cyclobarbital [52-31-3], cyclobarbitone,
5-(1-cyclohexen-1-yl)-5-ethylbarbituric
acid,C12 H16 N2 O3 , M r 236,26, mp 171 –
174 ◦ C, forms bitter crystals and is moderately
soluble in water. The calcium salt [143-76-0],
C24 H30 CaN4 O6 , M r 510.59, is prepared by heat-

ing ethyl(1-cyclohexen-1-yl) cyanoacetate with
guanidine sulfate in sodium ethoxide, followed
by hydrolysis with dilute sulfuric acid [76]. One
part dissolves in 70 parts of water. This com-
pound is a medium-acting hypnotic; abuse can
cause habituation or addiction.
Trade Names. Fabadorm (Bayer, FRG), Phan-
odorm (Bayer, FRG; Merck, FRG; Winthrop,
USA), Phanotal (Kwizda, Austria).
Aprobarbital [77-02-1], 5-allyl-5-isopro-
pylbarbituric acid, C10 H14 N2 O3 , M r 210.23,
mp 140 – 141.5 ◦ C, forms slightly bitter crystals,
and is insoluble in water, but soluble in ethanol.
The sodium salt [125-88-2], C10 H13 N2 NaO3 ,
M r 232.21, a hygroscopic, bitter powder, is sol-
uble in water. Preparation is described in [77].
This compound is a medium-acting hypnotic;
abuse can cause habituation or addiction.
Trade Names. Alurate (Hoffmann-La Roche,
Switzerland), Numal (United Chemicals, USA).
Allobarbital [52-43-7], allobarbitone,
5,5-diallylbarbituric acid, C10 H12 N2 O3 ,
M r 208.21, mp 171 – 173 ◦ C, is slightly bitter.
One part dissolves in 300 parts of cold water
and in 50 parts of boiling water. It is prepared
by heating diallyl malonate with urea in sodium
ethoxide. It can also be synthesized by the re-
action of barbituric acid with allyl bromide in
the presence of sodium acetate or copper sulfate
[78], [79]. It is a mediumacting hypnotic; abuse
can lead to habituation or addiction.
Trade Names. Diadol (Durst, USA), Dial (Ciba-
Geigy, Switzerland); it is contained in Spasmo-
Cibalgine (Ciba, Switzerland).
Vinylbital [2430-49-1], vinylbarbi-
tal, 5-(1-methylbutyl)-5-vinylbarbituric
acid,C11 H16 N2 O3 , M r 224.27, mp 90 –
91.5 ◦ C, is prepared by the reaction of guanidine
with the appropriate malonic ester and saponifi-
cation. It can also be synthesized by the addition
of acetylene to 5-(1-methylbutyl)barbituric acid
in the presence of zinc stearate [80].
Hypnotics 13
Vinylbital is a medium-acting hypnotic;
abuse can cause addiction or habituation.
Trade Names. Bykonox (Byk Belga, Belgium),
Optanox, Suppoptanox (Valpan, France), Speda
(Byk-Gulden, FRG).
Heptabarb [509-86-4], heptabarbital, hept-
abarbitone, 5-ethyl-5-(1-cycloheptenyl)barbitu-
ric acid, C13 H18 N2 O3 , M r 250.29, mp 174 ◦ C;
one part dissolves in 13 parts of methanol,
20 parts of ethanol, 13 parts of acetone, 40
parts of diethyl ether, and 5000 parts of wa-
ter. It is synthesized by the condensation of
ethyl(cyclohepten-1-yl)-cyanoacetic ester with
urea [81].
This compound is a short- to medium-acting
hypnotic; abuse can cause habituation or addic-
tion.
Trade Names. Medomin, Medapan (Ciba-
Geigy, Switzerland).
Secbutabarbital [125-40-6], secbutobarbi-
tal, butabarbital, secbutobarbitone, 5-ethyl-5-(2-
butyl)barbituric acid, C10 H16 N2 O3 , M r 212.24,
mp 165 – 168 ◦ C; the sodium salt [143-81-7],
C10 H15 N2 NaO3 , M r 234.23, is a bitter powder;
1 g dissolves in 2 mL of water, and the pH of a
1 % aqueous solution is 9.0 – 10.2. Preparation is
described in [82]. Secbutabarbital is a medium-
to long-acting hypnotic. It is widely used in the
United States as a sedative. Blood levels of 3.5 –
6 mg/100 mL are fatal; abuse can cause habitu-
ation or addiction.
Trade Names. BBS (Reid-Provident, USA),
Butabarbital Sodium (Parke-Davis, USA),
Butabarpal (Philadelphia-Labs., USA), Bu-
tamide (Zenith, USA), Buticaps (Carter-
Wallace, USA), Butisol (McNeil, USA), Day-
Barb (Anca, Canada), Merisyl (Meriot, Canada),
Neo-Barb (Neo, Canada), Sarisol (Halsey Drug,
USA).
14 Hypnotics

Pentobarbital [76-74-4], pentobarbi- Trade Names. Pernocton (Riedel De Haen,

tone, 5-ethyl-5-(1-methylbutyl)-barbituric FRG) Pernoston, Sonbutal (Tutag Pharm, USA).
acid,C11 H18 N2 O3 , M r 226.28, mp 130 ◦ C;
the sodium salt [57-33-0], C11 H17 N2 NaO3 , Vinbarbital [125-42-8], 5-ethyl-5-(1-
M r 248.26, is unstable in aqueous solu- methyl-1-butenyl)barbituric acid, C11 H16 N2 O3 ,
tion. It is prepared by boiling ethyl-1- M r 224.3, mp 161 – 163 ◦ C, forms slightly bitter
methylbutylcyanoacetic ester with guanidine crystals. Preparation is described in [85]; abuse
in sodium ethoxide solution and saponifying the can cause addiction.
product with dilute sulfuric acid [83]. This com-
pound is a short- to medium-acting hypnotic
and is more often used as a hypnotic than as
a sedative. A dose of more than 3 g can cause
death; blood levels of 1 – 2.5 mg/100 mL are
fatal; abuse can lead to habituation or addiction. Trade Names. Delvinal (Merck, Sharp & Dohme,
Trade Names. Neodorm (Minden, FRG). USA), Diminal (Astra, Sweden).
Pentobarbital sodium: Butylone (Hartz,
Canada), Dorminal (Alfasan, Netherlands), Talbutal [115-44-6], 5-allyl-5-sec-butylbar-
Embutal (Abbott, USA), Hypnol (Stickley, bituric acid, C11 H16 N2 O3 , M r 224.25, mp 108 –
Canada), Isoamytal, Isobarb (U.S. Standard, 110 ◦ C, forms slightly bitter crystals. Prepara-
USA), Iturate, Lethobarb (Loveridge, UK), tion is described in [86]. This compound is a
Napental (Beecham, UK), Narcoren (Veteri- short-acting hypnotic; abuse can cause habitua-
naria, Switzerland), Nembutal (Abbott, USA), tion or addiction.
Nova-Rectal (Nova, Canada), Novopentobarb Trade Names. Lotusate (Winthrop, USA), Pro-
(Novopharm, Canada), Pembule (Novocol, fundol (Promonta, FRG).
USA), Penbon (Adams, Australia), Pental (Van-
Pelt & Brown, USA; Saunders, Canada), Pen-
Cyclopentobarbital [76-68-6], 5-allyl-
tone (Faulding, Australia), Pentosol (Chroma-
5-(2-cyclopenten-1-yl(barbituric acid,
lloy, USA), Prodormol (Teva, Israel), Sedanox
C12 H14 N2 O3 , M r 234.25, mp 139 – 140 ◦ C,
(Troncin, France), Sombutol (Farmos Group,
forms slightly bitter crystals and is moderately
Finland), Somnopentyl (Pitman-Moore, USA),
soluble in hot water. This acid is synthesized by
Vetanarcol (Veterinaria, Switzerland).
the condensation of urea with the appropriately
Pentobarbital calcium: Insom Rapido (An-
disubstituted malonic ester [87]. It is a short-
dreu, Spain), Repocal (Desitin, FRG).
acting hypnotic; abuse can lead to habituation
and addiction.
Butallyonal [1142-70-7], 5-(2-bromoallyl)-
5-sec-butylbarbituric acid, C11 H15 BrN2 O3 , M r
303.16, mp 130 – 133 ◦ C, slightly bitter crys-
tals, is almost insoluble in water, but solu-
ble in ethanol. The sodium salt [3486-86-0],
C11 H14 BrN2 NaO3 , M r 325.15, forms a bitter,
crystalline powder and is soluble in water; the Trade Names. Cyclopal (Siegfried, Switzer-
pH of a 10 % aqueous solution is ca. 9.5. Prepa- land), Cyclopental (Pharmacia, Sweden).
ration is described in [84]. This compound is
a short- to medium-acting hypnotic; abuse can Propallyonal [545-93-7], 5-(2-bromoallyl)-
cause habituation or addiction. 5-isopropylbarbituric acid, C10 H13 BrN2 O3 , M r
289.13, mp 177 – 179 ◦ C, forms slightly bitter
crystals and is moderately soluble in water.
Preparation is described in [88]. This drug is a
short-acting hypnotic; abuse can cause habitua-
tion or addiction.

Trade Name. Noctal (Union Chimique, Bel-
gium).
Secobarbital, 5-allyl-5-(2-pentyl)barbituric
acid, C12 H18 N2 O3 , M r 238.28, mp 100 ◦ C,
is slightly bitter. The sodium salt [309-43-3],
C12 H17 N2 NaO3 , M r 260.27, is a hygroscopic,
bitter powder; preparation is described in [89].
It resembles pentobarbital but has a shorter du-
ration of action and is especially suitable for
injection. Blood levels of 1 – 2.5 mg/100 mL are
fatal; abuse can cause habituation or addiction.
Trade Names. Seconal (Lilly, USA).
Secobarbital sodium (secobarbital solubile):
Dormona (Wiedenmann, Switzerland), Im-
menoctal (Houde-ISH, France), Novosecobarb
(Novopharm, Canada), Proquinal (Protea, Aus-
tralia), S.C.B.Tal (Novo, Canada), Sebar (Van-
gard, USA), Secaps (Saunders, Canada), Sec-
ocaps (M.T.C. Canada), Secogen (Paul Maney,
Canada), Seconal, Seotal (Lilly, USA), Quinbar
(Adams, Australia).
Hexobarbital [56-29-1], 1,5-dimethyl-5-
(cyclohexen-1-yl)barbituric acid, C12 H16 N2 O3 ,
M r 236.26, mp 146 ◦ C. For preparation and prop-
erties, see →Anesthetics, General. It is a short-
acting hypnotic, which rapidly induces sleep.
Hexobarbital is frequently used as a short-acting
narcotic in the form of its sodium salt [50-09-9],
Evipan sodium, C12 H15 N2 NaO3 , M r 258.25.
The short duration of action and the short bi-
ological half-life are related to the metabolism
by cleavage of the N-methyl group and oxida-
tion of the cyclohexenyl group at the α position.
Evipan sodium is used as a narcotic.
Trade Names. Citopan (Nyco, Norway), Cy-
clopan (Interpharm, Austria), Sombucaps, Som-
bulex (Riker, USA), Tobinal (Siegfried, Switzer-
land), Toleran (Kwizda, Austria).
Hypnotics 15
5.2. Piperidinediones
As already mentioned, barbituric acids are cy-
clic urea derivatives. Another large class of hyp-
notics is derived from the lactams of 5-ami-
noglutaric acid; they are divided into two groups:
2,4-piperidinediones and 2,6-piperidinediones
(glutaric imides). They all have a disubstituted
carbon atom located α to the carbonyl group of
the lactam residue.
The hypnotic action of the piperidinediones
qualitatively resembles that of the barbiturates,
but they are less active and are used in larger
doses. Adverse effects and toxic reactions are
less frequent than with barbiturates. The piperi-
dinediones are rapidly broken down in the
body; the symptoms of poisoning resemble those
caused by barbiturates.
Taglutimide (see page 16) and thalido-
mide [50-35-1] have a monosubstituted car-
bon atom at position 3. The notorious thalido-
mide (3-phthalimido-2,6-dioxopiperidine, Con-
tergan) was removed from the market because it
caused severe adverse effects (prolonged neuri-
tis and malformation of the fetus).
Pyrithyldione [77-04-3], 3,3-diethyl-2,4-
dioxo-1,2,3,4-tetrahydropyridine, C9 H13 NO2 ,
M r 167.20 is structurally related to the piperi-
dinediones and exists in three crystalline modi-
fications: (1) mp 92 – 93 ◦ C, (2) mp 97 – 98 ◦ C,
(3) mp 81 – 86 ◦ C.
Trade Names. Persedon (Hoffmann-La Roche,
Switzerland), Benedorm (Arzneimittelwerke,
GDR).
This drug is no longer used because of its pro-
found adverse effects (agranulocytosis). It has
been replaced by the homologous methyprylon
which does not cause blood dyscrasia.
Methyprylon [125-64-4], 3,3,diethyl-5-
methyl-2,4-piperidinedione, C10 H17 NO2 , M r
183.26, mp 74 – 77 ◦ C, is soluble in water, etha-
nol, and chloroform. It is prepared by formylat-
16 Hypnotics
ing 3,3-diethylpiperidine-2,4-dione (dihypry-
lone) and reducing the resulting 3,3-diethyl-
5-hydroxymethylpiperidine-2,4-dione with hy-
drogen and Raney nickel [90]. The drug acts
rapidly but longer than pyrithyldione; it is less
toxic than phenobarbital. Prolonged use causes
habituation and addiction.
Trade Names. Noludar, Nolurate (Hoffmann-La
Roche, Switzerland).
Glutethimide [77-21-4], 3-ethyl-3-
phenylpiperidine-2,6-dione, C13 H15 NO2 ,
M r 217.26, mp 84 ◦ C, is soluble in ethyl ac-
etate, acetone, diethyl ether, chloroform, eth-
anol, and methanol, and almost insoluble in
water. It forms a hydrate and a water-soluble
hydrochloride. Glutethimide is prepared by the
addition of 2-phenylbutyronitrile to acrylic ester
to give 2-ethyl-2-phenylpentane-1,5-dicarbox-
ylic acid mononitrile, followed by cyclization
with sulfuric acid and acetic acid [91]. This
compound is effective in inducing sleep that
lasts through the night and is also suitable as
a daytime sedative. Prolonged administration
causes habituation and addiction.
Trade Names. Doriden (Ciba-Geigy, Switzer-
land; Takeda, Japan), Elrodorm (Dtsch. Hydrier-
werke, FRG), Glimide (Polfa, Poland), Gludorm
(Knoll, FRG), Regenox (Richter, Hungary).
Alonimid [2897-83-8], 1,2,3,4-tetrahydro-
4-oxonaphthalene-1-spiro-3 -piperidine-2 ,6 -
dione, C14 H13 NO3 , M r 243.25, mp 197 –
199 ◦ C,
is prepared by the Michael addition of ethyl
acrylate to phenylacetonitrile, followed by cy-
clization with polyphosphoric acid [92].
Trade Name. Alonimide (Merrell, USA).
Taglutimide [14166-26-8], biglumide,
2-(bicyclo [2,2,1] heptane-2-endo-3-endo-
dicarboximido)-glutarimide, C14 H16 N2 O4 ,
M r 276.29, mp 237 ◦ C, is unstable in alkaline
solution, and is slightly soluble in water and in
organic solvents. It is prepared by the reaction
of endo-bicyclo [2,2,2] -heptane-1,2-dicarbox-
ylic acid anhydride with l(+)-glutamic acid,
followed by cyclization with urea in dimethyl
sulfoxide at 180 ◦ C [93].
Unlike thalidomide, taglutimide does not
produce any teratogenic effects [94]. It is a use-
ful hypnotic because it causes little change in the
electroencephalogram of normal sleep [95].
Trade Name. Synval (Kwizda, Austria).
5.3. Quinazolinones
Quinazolinones are cyclization products of an-
thranilamide. Some have analgesic and an-
tipyretic properties. Only those derivatives in
which the nitrogen atom at position 3 has an 2-
substituted phenyl group are hypnotically active
[96]. The effect is very fast in onset and lasts
3 – 4 h. These compounds are similar in hyp-
notic action to the barbiturates but are reported
not to suppress REM sleep. They inhibit the
biosynthesis of serotonin and catecholamines,
which influence the physiology of sleep [97].
The quinazolinones are a group of hypnotics

with codeinelike, cough-suppressant, analgesic
properties; they are useful in patients who do not
tolerate or respond to barbiturates. Prolonged
use may cause habituation and psychological de-
pendence. No specific antagonists are available
for quinazolinone poisoning.
Methaqualone [72-44-6], 2-methyl-3-
o-tolyl-4-(3 H)quinazolinone, C16 H14 N2 O,
M r 250.30, mp 120 ◦ C (also 114 – 116 ◦ C), is
soluble in ethanol, diethyl ether, and chloro-
form, but almost insoluble in water. The hydro-
chloride [340-56-7], C16 H15 ClN2 O, M r 286.85,
mp 255 – 265 ◦ C, is prepared by the condensa-
tion of N-acetylanthranilic acid with o-toluidine
in the presence of phosphoryl chloride [98], [76].
Methaquolone is present in numerous combina-
tions.
Prolonged administration of methaqualone
has caused paresthesia of the extremities; daily
doses of 0.3 g should not be taken for more than
four weeks (in exceptional cases, eight weeks).
Trade Names. Bon-Sonnil (Diselen, Spain),
Dormogen (Leciva, Czechoslovakia), Dormu-
til (Chem.-Pharm. Werk Oranienburg, GDR),
Mequin (Lemmon, USA), Mozambin (Gerot,
Austria), Pro Dorm (Schürholz, FRG), Som-
notropon (Tropon, FRG), Torinal (Medicamen-
tos y Productos Quimic, Spain), Tuazolona
(Nessa, Spain).
Methaqualone hydrochloride: Cateudyl
(Covor, Belgium), Dormir (Langley, Aus-
tralia), Hyptor (Toraude, Canada), Melsed,
Melsedin (Boots, UK), Mequelon (Merck-
Frosst, Canada), Methasedil (Cooper, Switzer-
land), Nobadorm (Streuli, Switzerland), Nor-
morest (Merck, FRG), Noxybel (Probel, Bel-
gium), Optimil (Wallace, USA), Optinoxan
(Robisch, FRG), Parest (Parke-Davis, USA),
Parmilene (Chiesi, Italy), Pexaqualone (Ther-
apex, Canada), Revonal (Merck, FRG), Ri-
porest (Farmitalia, Italy), Sedalone (Phar-
bec, Canada), Somberol (Boots, UK), Som-
nafac (Smith, Miller & Patch, USA), Somnium
(Fargal, Italy), Sovelin, Soverin (Weifa, Nor-
Hypnotics 17
way), Sovinal (ND & K, Denmark), Toraflon
(Toraude, Canada), Tualone (ICN, Canada), Tu-
azol (Strasenburgh, USA).
Etaqualone [7432-25-9], 3-(2-ethylphenyl)-
2-methyl-4(3 H)-quinazolinone, C17 H16 N2 O,
Mr 264.31, mp 81 ◦ C. Hydrochloride
[97979-65-2], C17 H17 ClN2 O, mp 247 ◦ C.
Etaqualone is prepared by the condensation
of N-acetylanthranilic acid with 2-ethylaniline
in the presence of phosphoryl chloride [99].
Trade Name. Aolan (Beiersdorf, FRG).
Nitromethaqualone [340-52-3], 3-(2 -

methoxy-4 -nitrophenyl)-2-methyl-4(3 H-)
quinazolinone, C16 H13 N3 O4 , M r 311.29,
mp 135 – 137 ◦ C, and the hydrochloride
[3946-23-4], mp 246 – 248 ◦ C, are prepared as
the preceding compounds [100]. For action, see
[101].
Trade Name. Parnox (Santos, Spain).
5.4. Benzodiazepines
Many benzodiazepine derivatives used as
psychotropic agents (→Psychopharmacological
Agents) promote sleep by acting as mental re-
laxants and also by having a hypnotic action of
their own [102], [103]. The diverse action of this
group of drugs clearly illustrates the difficulty
of distinguishing hypnotics from sedatives and
tranquilizers if the sleep-inducing effect is the
only feature of interest. As with the barbiturates,
the sedative and tranquilizing effects of benzodi-
azepines induce “better sleep.” After an initially
steep increase in the use of benzodiazepines as
hypnotics, a drop in consumption is slowly be-
coming apparent in Europe and the United States
[104], [105]. Even though benzodiazepines are
18 Hypnotics
Table 2. Duration of action, biological half-life, and metabolism of benzodiazepines

Compound Onset of Half-life, h Main metabolic route ∗ Main active metabolite and Adult hypnotic
action, min half-life, h dose, mg
Nitrazepam 16 – 30 20 – 50 nitroreduction, acetylation none 5 – 10
Flunitrazepam 9 – 20 10 – 25 demethylation, nitroreduction dealkylflunitrazepam, 20 – 30 1–2
Flurazepam 15 – 45 2.3 oxidation desalkylfluorazepam, 50 – 100 15 – 30
Oxazepam 45 – 90 5 – 15 conjugation none 15 – 30
Lorazepam 15 – 45 10 – 15 (20) conjugation none 2–4
Lormetazepam 15 – 45 9 – 15 conjugation none 0.5 – 2
Temazepam 45 – 60 9.5 – 12 (17) conjugation none 15 – 30
Doxefazepam 15 – 40 7–8 dealkylation, conjugation none 10 – 20
Cinolazepam 15 – 45 3.8 conjugation none 15 – 30
Quazepam 60 39 oxidation, demethylation,
conjugation
2-oxoquazepam, 39; demethyl- 7.5 – 15
2-oxoquazepam, 73
Haloxazolam 45 – 60 2–4 cleavage of oxazoline residue 7-bromobenzodiazepin-2-one 5 – 10
(18, R = Br, R1 = H, R2 = F)
Triazolam 18 – 30 1.5 – 3 oxidation none 0.25 – 0.5
Estazolam 15 – 40 8 – 31 oxidation none 2–4
Midazolam 16 – 30 1–3 oxidation, hydroxylation,
conjugation none 10 – 30
Loprazolam 15 – 30 10 – 16 oxidation “loprazolam-N-oxide”, 5 – 8 1–2
Brotizolam 10 – 20 4.5 – 9 oxidation “hydroxymethyl brotizolam”,
4 – 8; 3-hydroxy-brotizolam 0.25 – 1

∗ Plasma protein binding rates are high (usually 90 – 95 %). In most cases, conjugation indicates glucuronization, and oxidation
denotes the introduction of a hydroxyl group at the 3-position or in the methyl group of a heterocycle condensed with the
benzodiazepine moiety.

not ideal sleep-inducing agents, new products
are constantly being marketed. Properties of the
most important benzodiazepines are listed in Ta-
ble 2.
Mode of Action. The benzodiazepines have
the same hypnotic action as the barbiturates,
with differences in selectivity, metabolism, and
adverse effects. Large doses do not possess the
narcotic properties exhibited by most barbitu-
rates. Benzodiazepines do not solely have a
sleep-inducing action; they exert a combination
of effects: sedation, relief of anxiety (produc-
tion of a better state of mind), muscle relax-
ation, and anticonvulsant action. These effects
shorten both the intermittent waking times and
deep sleep (reduction of δ sleep). In addition,
a hangover always occurs on the next day. The
effects of barbiturates and benzodiazepines are
compared in Table 3. From a pharmacological
standpoint, the hypnotically active benzodiaze-
pines have a depressant action on the electrical
discharges in the areas controlled by the limbic
system and block the function of the reticular
formation in the brain stem. They are similar to
the barbiturates in this respect; however, quanti-
tative differences could be of importance. Stud-
ies point to an interaction between benzodiaze-
pines and their receptors. Thus, they activate the
postsynaptic receptors for γ-aminobutyric acid
(GABA, an inhibitory neurotransmitter) and in-
crease the frequency of the opening of the chlo-
ride ion canals [106].
In other words, the benzodiazepines act
in the central nervous system by interfering
with GABA-mediated neurotransmission [107],
[108]. Traces of benzodiazepine derivatives
(e.g., diazepam and lorazepam) have been found
in wheat and potato and are identical to the syn-
thetic substances. Benzodiazepines also occur
in untreated rats, as well as in human serum and
brain. However, the concentration is very low
(several nanograms per gram of potato) and a
direct pharmacological effect on the central ner-
vous system has not been demonstrated [108],
[109].
Metabolism. All benzodiazepines are me-
tabolized in the body [110], but in some cases
the retention times are very long. In addition,
the pharmacologically active metabolites of the
benzodiazepines have long residence times in
 Hypnotics 19

the body, and the rate of metabolic degradation ataxia, disorientation, and nightmares. Regu-
decreases considerably with age. A plasma half- lar administration of most benzodiazepines in-
life of 1.5 – 40 h, and in extreme cases 47 – 100 h, creases the dose necessary to evoke the same
is observed [111]. The main metabolic transfor- response.
mations of benzodiazepines are (Fig. 4): Indications for the use of benzodiazepine
hypnotics are given in Table 4.
1) Hydroxylation at position 3 with formation
of more active products, which are, however, Table 3. Comparison of the effects and properties of barbiturates
and benzodiazepines
usually rapidly eliminated after glucuroniza-
tion Effect or property Barbiturates Benzodiaze-
pines
2) Demethylation at N-1 with formation of
longer acting demethylbenzodiazepines
3) Ring opening at C-3, which again frequently Narcotic yes no
Sedative yes yes
produces active metabolites that can be recy- Muscle relaxation almost none yes
clized in the organism Anticonvulsant yes yes
4) Heterocyclic moiety of benzodiazepines Shortening of REM sleep and yes yes
increase in REM phase on withdrawal
containing fused heterocyclic structures, (rebound effect)
(e.g., triazolam) rapidly broken down to form Retrograde amnesia yes almost none
inactive metabolites Accumulation yes yes
Hangover yes (clearly yes (also
sedated positive
Therapy. The expectation that therapeutic waking influence on
problems could be solved with benzodiazepines phase) nervousness,
apprehen-
of lower half-life has proved unfounded. Al- sion,
though the faster onset of effect, swift relief of tension,
listlessness
anxiety, and rapid elimination of these drugs pre- in the
vent a hangover, rebound insomnia occurs as a waking
phase)
severe adverse effect to produce delirium and
Tolerance yes yes
exogenous psychosis, (a type of premature with- Abuse yes yes
drawal symptom) during the first night of use. Enzyme induction yes no
A temporary shortening of REM sleep follows Therapeutic range narrow wide
Toxicity relatively relatively
with subsequent paradoxical alertness, fear, ag- high low
itation, and anterograde amnesia during the sec-
ond half of the night [104]. In addition, abuse of
the rapid onset of activity benzodiazepines (e.g.,
lorazepam and flunitrazepam) has been reported, Synthesis. The hypnotically active benzo-
especially among young people. diazepines can be classified as benzo-1,4-
From a therapeutic standpoint, the benzo- diazepinones (18), or benzodiazepines con-
diazepine hypnotics are classified according to taining fused heterocyclic moieties, e.g., triazo-
their duration of action: lobenzodiazepines (21a–21c). The synthesis of
these compounds is shown in Figures 5 and 6,
1) long-acting compounds, e.g., flurazepam, respectively [102].
quazepam, nitrazepam, and flunitrazepam;
2) medium-acting compounds, e.g., Nitrazepam [146-22-5], 1,3-dihydro-7-
temazepam, lorazepam, loprazolam, ox- nitro-5-phenyl-2H-1,4-benzodiazepin-2-one,
azepam, and brotizolam; and (18a),C15 H11 N3 O3 , M r 281.26, mp 224 –
3) short-acting compounds, e.g., triazolam, mi- 226 ◦ C, is prepared by the cyclization of 2-
dazolam, and cinolazepam. (aminoacetamino)-5-nitrobenzophenone or by
Long-acting benzodiazepines should not be the nitration of benzodiazepinone with an un-
used to treat insomnia because of accumulation, substituted phenyl substituent [112], [113].
occurrence of undesired effects before sleep in- Indications: problems in falling asleep, fre-
duction, daytime drowsiness, lowered activity, quent awakening during the night, and early
awakening in the morning. The duration of in-
20 Hypnotics
Figure 4. Metabolism of triazolam.
duced sleep is 7 – 8 h; a strong hangover occurs
next day [114].
Trade Names. Alodorm (Alphapharm, Aus-
tralia), Apodorm (Apoth. Lab., Norway), Arem
(Lennon, Republic of South Africa), Atem-
pol (Norgine, UK), Benzalin (Shionogi, Japan),
Calsmin (Upjohn, USA), Dormo-Puren (Klinge,
FRG), Dumolid (Dumex, Denmark), Eatan N
(Desitin, FRG), Hipsal (Salvat, Spain), Ime-
son (Desitin, FRG), Insomin (Orion, Fin-
land), Ipersed (Sidus, Italy), Lagazepam (La-
gap, Switzerland), Megadon, Mogadan, Mo-
gadon (Roche, Switzerland), Mitidin (Savorma,
Italy), Nelbon (Sankyo, Japan), Neuchlonic
(Taiyo, Japan), Nitepam (USV, USA), Nitra-
dos (Berk, UK), Nitrempax (Lafi, Brazil), No-
vanox (Pfleger, FRG), Numbon (Ikapharm,
Israel), Ormodon (Ormed, South Africa),
Pacisyn, Paxisyn (Synthetic, Finland), Pel-
son (Infale, Spain), Radedorm (Berlin-Chemie,
GDR), Relact (Lemonier, Argentina), Remnos
(DDSA, UK), Somitran (Farmos Group, Fin-
land), Somnased (Duncan Flockhart, UK), Som-
nite (Norgine, UK), Somnolin (Dima, Italy),
Unisomnia (Unigreg, UK).
Flunitrazepam [1622-62-4], 5-(2-
fluorophenyl)- 1, 3- dihydro-1-methyl-7-
nitro-2H-1,4-benzodiazepin-2-one, (18b),
C16 H12 FN3 O3 , M r 313.30, mp 166 – 167 ◦ C
(also 170 – 172 ◦ C). The synthesis of this com-
pound is analogous to that of nitrazepam [113].
Pharmacologically, flunitrazepam is one of the
most active benzodiazepinones; the hypnotic ef-
fects begin with doses as low as 1 – 2 mg [115].
It is used in anesthesiology in solution form to
induce sleep.
Trade Names. Flunipam (A. L., Norway), Hypn-
odorm (Teva, Israel; Alphapharm, Australia),
Hipnosedon, Narcozep, Rohpinol, Rohypnol,
Roipnol (Hoffmann-La Roche, Switzerland).
Flurazepam [17617-23-1], 7-chloro-1-
(2-diethylaminoethyl)-5-(2-fluorophenyl)-
1,3-dihydro-2H-1,4-benzodiazepin-2-one,
(18c),C21 H23 ClFN3 O, M r 387.89, is oily. The
dihydrochloride [1172-18-5], C21 H25 Cl3 FN3 O,
M r 460.83, mp 215.5 – 217.5 ◦ C (decomp.),
forms light yellow crystals. Flurazepam is
synthesized by the alkylation of 18 (R = Cl,

Table 4. Indications for the use of benzodiazepine hypnotics ∗
R1 = H, R2 = F) with diethylaminoethyl chlo-
ride [116]. Another method uses 2-(diethyl-
aminoacetylamino)-5-chloro-2 -fluorobenzo-
phenone as the starting material [117]. Flu-
razepam can also be synthesized by the oxidative
ring extension of 2-aminomethyl-5-chloro-1-(2-
diethylaminoethyl)-3-(2 -fluorophenyl)indole
with chromium trioxide [118].
Hypnotics 21
Figure 5. Synthesis of benzo-1,4-diazepinones (18), and
their 2-thiono-(19) and 3-hydroxy derivatives (20)
Nitrazepam (18a): R = NO2 ; R1 = H; R2 = H
Flunitrazepam (18b): R = NO2 ; R1 = CH3 ; R2 = F
Flurazepam (18c): R = Cl; R1 = CH2 −CH2 N (C2 H5 )2 ;
R2 = F
Nimetazepam (18d): R = NO2 ; R1 = CH3 ; R2 = H
Quazepam (19): R = Cl; R1 = CH2 −CF3 ; R2 = F
Oxazepam (20a): R = Cl; R1 = H; R2 = H
Lorazepam (20b): R = Cl; R1 = H; R2 = Cl
Lormetazepam (20c): R = Cl; R1 = CH3 ; R2 = Cl
Temazepam (20d): R = Cl; R1 = CH3 ; R2 = H
Doxefazepam (20e): R = Cl; R1 = CH2 −CH2 OH; R2 = F
Cinolazepam (20f): R = Cl; R1 = CH2 −CH2 −CN; R2 = F
22 Hypnotics
The full hypnotic response of flurazepam is
evoked even after prolonged use. However, it has
an extremely long plasma half-life (47 – 100 h),
with long-lasting metabolites. It promotes the
onset of sleep, reduces the frequency of interrup-
tions at night, and increases the total duration of
sleep. The effects of the drug begin within 20 –
30 min, and sleep usually lasts for 7 – 8 h. Flu-
razepam can reduce REM sleep, but does not
produce a REM rebound and associated sleep
disturbance on termination of treatment. This
drug strongly suppresses “level 4” of orthodox
sleep (see Chap. 1). It is given in doses of 30 –
60 mg [119].
Trade Names. Noctosom (Ikapharm Kfar-Sava,
Israel), Staurodorm (Dolorgiet, FRG).
Flurazepam hydrochloride: Benzozil (Ky-
owa, Japan), Dalmadorm, Dalmane, Dalmate,
Dalmene, Dormador, Dormodor (Hoffmann-La
Roche, Switzerland), Felison (Sigurta, Italy),
Insumin (Kyorin, Japan), Natam (Unifa, Ar-
gentina), Novoflupam (Novopharm, Canada),
Remdue (Biomedica Foscama, Italy), Somnol
(Horner, Canada), Som-Pam (ICN, Canada),
Valdorm (Valeas, Italy).
Nimetazepam [2011-67-8], 1,3-di-
hydro-1-methyl-7-nitro-5-phenyl-2H-1,4-
benzodiazepin-2-one (1-methylnitrazepam),
(18d), C16 H13 N3 O3 , M r 295.30, mp 156.5 –
157.5 ◦ C, forms light yellow crystals and is pre-
pared analogously to nitrazepam [113], [120]. It
has a long half-life and resembles nitrazepam in
activity, but also has anticonvulsant and muscle
relaxant properties [121].
Trade Name. Erimin (Sumitomo, Japan).
Figure 6. Synthesis of triazolobenzodiazepines (21)
Estazolam (21a): R = H; R1 = H
Triazolam (21b): R = CH3 ; R1 = Cl
Midazolam (21c): R = CH3 ; R1 = F
Oxazepam [604-75-1], 7-chloro-1,3-di-
hydro-3-hydroxy-5-phenyl-2H-1,4-
benzodiazepin-2-one,(20a), C15 H11 ClN2 O2 ,
M r 286.72, mp 205 –206 ◦ C, is prepared by the
oxidation of 18 (R = Cl, R1 = H, R2 = H) with
peracetic acid, followed by the reactions de-
scribed in Figure 5. Oxazepam is primarily a
sedative, but also has sleep-inducing properties
[122].
Trade Names. Adumbran (Thomae, FRG;
Boehringer Ingelheim, FRG), Alepam (Al-
phapharm, Australia), Alopam (A. L., Norway),
Anxiolit (Gerot, Austria; Medichemie, Switzer-
land), Aslapax (Asla, Spain), Benzotran (Pro-
tea, Australia), Durazepam (Durachemie, FRG),
Enidrel (Syncro, Argentina), Expidet (Wyeth,
UK), Isodin (Tosi, Italy), Murelax (Ayerst, Aus-
tralia), Noctazepam (Brenner, Austria), Oxepam
(Wyeth, UK), Oxpam (ICN, Canada), Praxiten
(Wyeth, UK), Propax (Cipan, Portugal), Psico-
pax (Bama-Geve, Spain), Sedokin (Geymonat
Sud, Italy), Senepax, Serax, Serenid (Wyeth,
UK), Serepax (Wyeth, UK; Ferrosan, Den-
mark), Seresta (Wyeth, Netherlands), Sigacalm
(Siegfried, Switzerland), Sobril (KabiVitrum,

Sweden), Uskan (Desitin, FRG), Vaben (Rafa,
Israel), Zapex (Riva, Canada).
Lorazepam [846-49-1], 7-chloro-5-(2-
chlorophenyl)- 1, 3- dihydro- 3- hydroxy- 2H-
1,4-benzodiazepin-2-one, (20b),
C15 H10 Cl2 N2 O2 , M r 321.17, mp 166 – 168 ◦ C,
is prepared from 18 (R = Cl, R1 = H, R2 = Cl),
analogously to oxazepam. Lorazepam is pri-
marily an anti-anxiety drug; it also has sleep-
inducing properties and is used in anesthesiol-
ogy [123].
Trade Names. Ativan (Wyeth, USA), Bonton
(Unipharma, Israel), Control (Sigurta, Italy),
Donix (Llorens, Spain), Emotion (Alpes, Ar-
gentina), Emotival (Armstrong, Argentina), Lar-
pose (Cipla, India), Laubeel (Desitin, FRG),
Loram (Lek, Yugoslavia), Lorans (Schiapar-
elli, Italy), Lorax (Wyeth, UK), Lorivan (Disco,
Israel), Merlit (Ebewe, Austria), Nervistop L
(Ingram, Argentina), Orfidal (Orfi, Spain), Pro
Dorm (Schürholz, FRG), Securit (Perrel, Italy),
Sedatival (Raffo, Argentina), Tavor (Wyeth,
FRG), Temesta (Ferrosan, Denmark; Wyeth,
Netherlands), Tolid (Dologiet, FRG), Tra-
pax (Wyeth, Argentina), Wypax (Yamanouchi,
Japan).
Trimethylacetate: Drupal (Novag, Spain), Pi-
ralone (Ferrer, Spain), Placinor (Robert, Spain).
Lormetazepam [848-75-9], 7-chloro-
5-(2-chlorophenyl)- 1, 3-dihydro-3-hydroxy-
1-methyl-2H-1,4-benzodiazepin-2-one (N-
methyllorazepam), (20c), C16 H12 Cl2 N2 O2 ,
M r 335.19, mp 192 – 194 ◦ C, is prepared analo-
gously to oxazepam [124]. The hypnotic effect
of 1 – 2 mg is similar to that of nitrazepam [125].
Trade Names. Lembrol (Vinas, Spain), Loramet
(Wyeth, Belgium, Netherlands), Minias (Far-
mades, Italy), Noctamid, Pronoctan (Schering,
FRG), Nocton (Eurolab, Chile).
Temazepam [846-50-4], 7-chloro-1,3-di-
hydro-3- hydroxy- 1- methyl- 5- phenyl-2H-
1,4-benzodiazepin-2-one (3-hydroxydiazepam,
N-methyloxazepam, oxydiazepam), (20d),
C16 H13 ClN2 O2 , M r 300.74, mp 119 – 121 ◦ C,
prepared analogously to oxazepam [126]. It is
used as a hypnotic drug and should be taken
1 – 2 h before going to sleep [127].
Trade Names. Euhypnos (Farmitalia Carlo Erba,
Italy; Sigma, Australia), Levanxene (Farmitalia
Hypnotics 23
Carlo Erba, Italy; Aesca, Austria), Mabertin
(Sidus, Argentina), Normison (Wyeth, UK),
Planum (Farmitalia Carlo Erba, Italy), Remestan
(Wyeth, FRG), Restoril (Sandoz, Switzerland;
Anca, Canada), Somaz (Quantum, USA), Tenso
(Castejon, Spain), Veroqual (Lek, Yugoslavia).
Doxefazepam [40762-15-0], 7-chloro-5-(2-
fluorophenyl)- 1, 3- dihydro- 3- hydroxy- 1- (2-
hydroxyethyl)-2H-1,4-benzodiazepin-2-one,
(20e), C17 H14 ClFN2 O3 , M r 348.57, mp 138 –
140 ◦ C, is prepared by the alkylation of 7-
chloro-5-(2-fluorophenyl)-1,3-dihydro-2H-1,4-
benzodiazepin-2-one N-oxide with 2-bromoeth-
anol or 2-bromoethyl acetate, followed by
Polonowski rearrangement and saponification
[128], [129]. Doxefazepam is a hypnotic drug
with two to four times the activity and half the
toxicity of flurazepam; it causes significantly
less hangover [130].
Trade Name. Doxans (Schiaparelli, Italy).
Cinolazepam [75696-02-5], 7-chloro-5-
(2-fluorophenyl)- 2, 3- dihydro- 3- hydroxy-2-
oxo-1H-1,4-benzodiazepine-1-propionitrile,
(20f),C18 H13 ClFN3 O2 , M r 356.84, mp 190 –
193 ◦ C, is prepared by the cyanoethylation
of compound 20 (R = Cl, R1 = H, R2 = F) in
the presence of trimethylammonium hydrox-
ide [131]. This drug resembles doxefazepam in
action [132].
Quazepam [36735-22-5], 7-chloro-5-(2-
fluorophenyl)-1-(trifluoro-2,2,2-ethyl)-1, 3-di-
hydro-2H-1,4-benzodiazepine-2-thione, (19),
C17 H11 ClF4 N2 S, M r 386.6, mp 138 – 139 ◦ C,
is synthesized by the reaction of compound
18 (R = Cl, R1 = CH2 CF3 , R2 = F) with phos-
phorus pentasulfide [133]. Quazepam is a sleep-
inducing drug; it has a long duration of action
and causes a hangover. It also produces long-
acting metabolites in the body [134].
Trade Names. Dormalin (Schering Corp., USA),
Prosedar, Selepam, Cetrane (Schering Plough,
USA).
Haloxazolam [59128-97-1], 10-bromo-
11b-(2- fluorophenyl)- 2, 3, 7- 11b- tetra-
hydrooxazolo-[3,2-d][1,4]benzodiazepin-
6(5H)-one,C17 H14 BrFN2 O2 , M r 377.22,
mp 185 ◦ C, is synthesized by the reaction of
bromo- (or tosyloxy-) acetamidobenzophenone
24 Hypnotics
with ethanolamine, followed by cyclization in
the presence of acetic acid [135].
Haloxazolam, oxazolam, and cloxazolam be-
long to a group of derivatives with an oxazoline
ring at the 4,5-position of the benzodiazepine
ring, which causes a change in some biologi-
cal parameters. Haloxazolam is primarily used
as a hypnotic drug in the treatment of insomnia
[136].
Trade Name. Somelin (Sankyo, Japan).
Estazolam [29975-16-4], 8-chloro-6-
phenyl-4H- [1,2,4]triazolo - [4,3- a][1,4]benzo-
diazepine, (21a), C16 H11 ClN4 , M r 294.75,
mp 228 – 229 ◦ C, is prepared as shown in Fig-
ure 6 [137], [138]. Indications are problems in
falling asleep, insomnia, and sleep induction in
anesthesiology [139]. The metabolism of the
triazolobenzodiazepines differs from that of the
benzodiazepinones in that the triazole ring re-
mains intact in most of the metabolites. One
of the first transformations in the metabolism
of benzodiazepines is the cleavage of the sub-
stituent at position 1.
Trade Names. Domnamid (Lundbeck, Den-
mark), Esilgan (Amer. Cyanamid, USA; Takeda,
Japan), Eurodin (Takeda, Japan), Nuetalon
(Casenne, France), Tasedan (Takeda, Mex-
ico), Somnatrol (Abbott, Peru), Noctal (Abbott,
Brazil).
Triazolam [28911-01-5], 8-chloro-6-
(2-chlorophenyl)- 1- methyl- 4H- [1,2,4]tri-
azolo[4,3-a][1,4]-benzodiazepine, (21b),
C17 H12 Cl2 N4 , M r 343.22, mp 223 – 225 ◦ C, is
prepared as shown in Figure 6 [138], [140]. Tri-
azolam is an extremely strong, but short-acting
hypnotic drug, the hypnotic dose being 0.5 mg
(equivalent to 30 mg of flurazepam) [141].
Doses of 0.125 – 1 mg are effective in induc-
ing sleep [142]. The maximum effect is pro-
duced 1 – 2 h after administration and lasts up
to 8 h; it is totally eliminated after 24 h. As with
nitrazepam and flunitrazepam, withdrawal after
prolonged use causes rebound insomnia [143].
Studies question the occurrence of the rebound
effect when standard doses (0.125 – 0.25 mg) are
taken [144]. For tolerance studies, see [145] and
for a review [146].
Trade Names. Halcion (Upjohn, USA), Novi-
dorm (Syntial, Argentina), Novodorm (Rubio,
Spain), Nuctane (Bago, Argentina).
Midazolam [59467-70-8], 8-chloro-6-(2-
fluorophenyl)- 1- methyl- 4H- imidazo[1,5-a]
[1,4]benzodiazepine,(21c), C18 H13 ClFN3 ,
M r 325.77, mp 158 – 160 ◦ C (maleate, mp 114 –
117 ◦ C), is synthesized as shown in Figure 6
[147]. The presence of the imidazole ring en-
sures a stable solution, a short duration of ac-
tion, and the formation of water-soluble in-
jectable salts. Below pH 4, the diazepine ring
is opened between positions 4 and 5 to give a
water-soluble salt of the primary amine. Ring
closure takes place above pH 4 with a half-life
of 10 min (22 → 21c). Indications are premedi-
cation before operations and anesthesia [148].
Trade Names. Dormicum, Dormonil, Hypnovel,
Sorenor (Hoffmann-La Roche, Switzerland),
Flormidal (Galenika, Yugoslavia).
 Hypnotics 25

Brotizolam [54801-81-7], 2-bromo-4-(2-

chlorophenyl)-9-methyl-6H-thieno[3,2-f ]
[1,2,4]triazolo[4,3-a][1,4]diazepine, (24),
C15 H10 BrClN4 S, M r 393.7, mp 212 – 214 ◦ C,
is synthesized by the condensation of 2-ami-
no-3-(2-chlorobenzoyl)-thiophene (23), with
aminoacetic ester, bromination, and fusion of
the triazole ring [149].

Brotizolam is very effective in prolonging

sleep and is an especially useful hypnotic for
light to moderate insomnia [150].

Trade Names. Ladormin, Lendorm, Lendormin

(Boehringer, FRG).

Loprazolam [61197-93-1], 6-(2-chloro-

phenyl)-2,4-dihydro-2-[(4-methyl-1-piper-
azinyl)methylene]- 8- nitro- 1H- imidazo[1,2-
a][1,4]benzodiazepin-1-one, (25),
C23 H21 ClN6 O3 , M r 464.91, mp 214 – 215 ◦ C,
(methanesulfonate, M r 560.71, mp 205 –
210 ◦ C), is synthesized by the condensation
of the appropriate benzodiazepinethione with
glycine, followed by cyclization. Condensation
with dimethylformamide acetal and subsequent
aminolysis with N-methylpiperazine gives lo-
prazolam (Fig. 7) [151].
This drug is an effective, medium-acting hyp-
notic and is especially suitable for the short-term
treatment of insomnia. It is also prescribed to
promote the onset of sleep and prevent frequent
awakening during the night [152].
Figure 7. Synthesis of loprazolam (25)
Trade Names. Dormonoct (Roussel Lab., UK),
Havlane (Diamant, France).
Rilmafazone [99593-25-6], 5-[(2-ami-
noacetamido)methyl]- 1- [4- chloro-2(2-chloro-
benzoyl)-phenyl]- N, N- dimethyl- 1H-1,2,4-tri-
azole-3-carboxamide, (26), C21 H20 Cl2 N6 O3
· HCl · 2 H2 O, M r 547.82, mp 107 ◦ C, is pre-
pared from 2 ,5-dichloro-2-aminobenzophe-
none as described in [153] (Fig. 8). Rilmafa-
zone belongs to a group of open-ring benzodi-
azepinone analogs called peptide aminobenzo-
phenones. It appears to be a precursor of the
triazolobenzodiazepinones. It has both sedative
26 Hypnotics

Figure 8. Synthesis of rilmafazone (26).

and antianxiety effects. Doses of 1 – 2 mg pro- subsequently esterified with 4-methyl-1-piper-

mote the onset of sleep. Indication is neurotic azinyl chloride [155].
insomnia. This drug was introduced in 1988
(Shionogi, Japan) [154].

6. Other Hypnotics
Other substances interact with the benzodi-
azepine receptors in vitro and in vivo. They
resemble benzodiazepines in general activity,
with similar sedative, anti-anxiety, and hypnotic
properties, but differences in adverse effects ex-
ist. The practical value of these drugs, which
include zopiclone, suriclone (anxiety-relieving),
zolpidem, and alpidem, will become apparent on
extended use.
A dose of 7.5 mg is required for the treat-
Zopiclone [43200-80-2], 6(5-chloro-2- ment of insomnia, which is equivalent to the ef-
pyridyl)-5-(4-methylpiperazin-1-yl)carbon- fect produced by 5 – 10 mg of nitrazepam. The
yloxy-7- oxo- 6,7- dihydro- 5H - pyrrolo[3,4- adverse effects are similar to those produced
b]pyrazine, (29), C17 H17 ClN6 O3 , M r 388.82, by short-acting benzodiazepines, but zopiclone
mp 178 ◦ C, is synthesized by the condensation of causes less hangover [156]. Zopiclone is, for the
2,3-piperazinedicarboxylic acid anhydride with most part, converted to inactive metabolites. The
2-amino-5-chloropyridine to give compound 27, elimination halftime for zopiclone and its active
which is partially reduced to compound 28 and

N-oxido metabolite (11 % formation) is 3.5 – 6 h
[157].
Trade Names. Imovane, Imovance, Theraplix
(Rhône-Poulenc, France), Amoban, Amovane
(Rhône-Poulenc, France).
Zolpidem [82626-48-0], 2-(4-methylphe-
nyl)-6-N,N- trimethylimidazo- [1,2-a]pyridine-
3-acetamide-l-(+)-tartrate (2 : 1) [99294-93-6],
(30a), (C19 H21 N3 O)2 · C4 H6 O6 , M r 764.88, is
prepared as described in [158]. This short-acting
hypnotic is rapid in onset; it reacts specifically
with the benzodiazepine type 1 receptor [159].
It is registered as Stilnox (Synthelabo, France)
and was put on the market in 1988 (licenced by
Searle).
Sulfones. The first synthetic hypnotics to be
introduced were sulfones (sulfonal and trional).
These sedatives and hypnotics are no longer used
because of their high toxicity and severe adverse
effects.
Antiepileptic Agents. Chlormethiazole
[53-45-9], 5-(2-chloroethyl)-4-methylthiazole,
is an antiepileptic drug occasionally used as a
hypnotic.
Trade Name. Distraneurin (Pharma Stern,
FRG).
Fenadiazole [1008-65-7], 2-(2-hydroxy-
phenyl)-1,3,4-oxadiazole, C8 H6 N2 O2 , M r
140.14, mp 111 – 112 ◦ C, is prepared by the con-
densation of salicylhydrazide with ethyl ortho-
formate [161]. The LD50 is 940 mg/kg (mouse,
i.p.).
Trade Name. Viodor (Violani-Farmavigor,
Italy).
Hypnotics 27
7. References
General References
1. AMA Drug Evaluations: Drugs Used for
Anxiety and Sleep Disorders, 6th ed., Chicago
1986, pp. 81 – 110.
2. J. A. Vida in W. O. Foye (ed.): “Central
Nervous System Depressants:
Sedative-Hypnotics,” Principles of Medicinal
Chemistry, Lea u. Febinger, Philadelphia 1976,
pp. 155 – 183.
3. J. A. Vida in A. Burger (ed.):
“Sedative-Hypnotics,” Medicinal Chemistry,
4th ed., Part III, New York, Chichester,
Brisbane, Toronto, 1980, pp. 787 – 828.
4. K. W. Wheeler: “Non-barbiturate Hypnotics,”
Med. Chem. (Academic) 6 (1963) 1 – 245.
5. W. J. Doran: “Barbituric Acid Hypnotics,”
Med. Chem. (Academic) 4 (1959) 1 – 334.
6. J. Hoffmann, B. Mencke, in G. Ehrhardt, H.
Ruschig (eds.): “Hypnotika,” Arzneimittel, 2nd
ed., vol. 1, Verlag Chemie, Weinheim 1966
pp. 37 – 66.
7. S. Ebel: “Hypnotika, Sedativa,” in
Synthetische Arzneimittel, Verlag Chemie,
Weinheim 1979, pp. 109 – 159.
8. E. Mutschler: Arzneimittel-Nebenwirkungen,
Hypnotika, Wissenschaftl. Verlags GmbH,
Stuttgart 1972, pp. 68 – 77.
9. S. C. Harvey in L. S. Goodman, A. Gilman,
A. G. Gilman, G. B. Koelle (eds.): “Hypnotics
and Sedatives,” The Pharmacological Basis of
Therapeutics, 7th ed.,Macmillan Publ., New
York, Collier Macmillan, Toronto, Ballière
Tindall, London 1985, pp. 339 – 371.
10. J. Oswald in E. Jucker (ed.): “Drug Research
and Human Sleep,” Arzneimittelforschung,
vol. 22, Birkhäuser, Basel 1978, pp. 355 – 372.
11. M. Monnier et al., “Biology of Sleep. An
Interdisciplinary Survey,” Experientia 36
(1980) 1 – 142.
12. J. P. Olié, S. Lepastier, G. Dana, J. M. Birée,
“Medicaments hypnotiques,” Med. Actuelle 5
(1978) 53 – 61; Chem. Abstr. 88 (1978)
145 812.
13. M. D. Allen, D. J. Greenblatt, Meyler’s Side
Eff. Drugs 4 (1980) 29 – 36; 3 (1979) 29 – 38;
2 (1978) 36 –44; 1 (1977) 21 – 29; VIII
1972 – 1975, 64 – 83, Excerpta Medica,
Amsterdam – Oxford.
14. H. P. Büch: “Medikamentöse Behandlung von
Schlafstörungen und Erregungszuständen,
Pharmakologie der Sedativa/Hypnotika,”
Dtsch. Apoth. Ztg. 121 (1981) 1951 – 1959.
28 Hypnotics
15. A. A. Borbély: “Schlafmittel und Schlaf:
übersicht und therapeutische Richtlinien,”
Ther. Umschau 43 (1986) 509 – 516.
16. E. Soos: “Schlaf und Schlafmittel,” OAZ
Österr. Apoth. Ztg. 28 (1974) 93 – 101.
17. A. N. Nicholson: “Hypnotics Their Place in
Therapeutics,” Drugs 31 (1986) 164 – 171.
18. E. A. Swinyard: Sedatives und Hypnotics in
Remington Pharmaceutical Sciences, 7th ed.,
Mack Publ., Easton, Pa., 1985,
pp. 1059 – 1074.
19. D. Schneider-Helmert: Schlafmittel:
“Symptomatisch oder kurativ?” Schweiz.
ärzteztg. 67 (1986) 1477 – 1483.
20. W. P. Koella: “The Organization and
Regulation of Sleep,” Experientia 40 (1984)
309 – 338.
21. W. Keup: “Mißbrauch und Abhängigkeit von
Sedativa und Hypnotika,” Internist 27 (1986)
746 – 756.
22. K. Laemmel: “Der Schlaf-Seismograph von
Körper und Seele,” Schweiz. Rundsch. Med.
Prax. 74 (1974) 1305 – 1310.
23. A. Borbély: Das Geheimnis des Schlafs,
Deutsche Verlagsanstalt GmbH, Stuttgart
1984.
24. T. Zilker, M. Clarmann:
“Schlafmittelvergiftungen,” MMW Münch.
Med. Wochenschr. 126 (1984) 297 – 300.
25. V. B. Meyer-Rochow: “Vom Winterschlaf
lernen,” Selecta 1984, 1116 – 1120.
26. Drug, Facts and Comparisons, Facts and
Comparisons Div., St. Louis 1982, 976 – 1016
(a division of J. B. Lippincott Co.,
Philadelphia-Toronto).
27. Martindale, The Extra Pharmacopoeia,
Hypnotic and Sedatives, 28th ed.,
Pharmaceutical Press, London 1982,
pp. 792 – 817.
28. E. A. Swinyard in Remington’s
Pharmaceutical Sciences, 17th ed., Mack
Publ. Easton, Pa., 1985, 1059 –1074.
Specific References
29. A. A. Borbély, J. Neural. Trans. Suppl. 21
(1986) 243 – 254.
30. S. H. Reppert et al., Science (Washington,
D.C.) 242 (1988) 78 – 81.
31. M. Monnier et al., Experientia 33 (1977) 548.
A. E. Kornmüller, H. B. Lux, K. Wink, M.
Klee, Naturwissenschaften 48 (1961)
503 – 505. D. Schneider-Helmert: “Die
körpereigene Schlafsubstanz DSIP,” Schweiz.
Rundsch. Med. 74 (1985) 1311 – 1315. M. V.
Graf, A. J. Kastin: “Delta-Sleep-Inducing
Peptide (DSIP): An Update,” Peptides (N.Y.) 7
(1986) 1165 – 1187; Neurosci. Biobehav. Rev.
8 (1984) 83 – 93.
32. A. A. Borbély, G. Youmbi-Balderer: “Ist
Tryptophan ein biologisches Schlafmittel,”
Schweiz. Rundsch. Med. Prax. 75 (1986)
785 – 790.
33. J. P. Olié, S. Lepastier, S. Dana, J. M. Birée,
Med. Actuelle 5 (1978) 53.
34. A. A. Borbély, Ther. Umsch. 43 (1986) 509.
35. M. B. Scharf, E. O. Bixler, A. Kales, C. R.
Soldatos, Pharmacology 19 (1979) 173 – 181.
A. Kales et al., Pharmacology 26 (1983)
138 – 149.
36. K. Laemmel, Schweiz. Rundsch. Med. Prax.
74 (1985) 1305 – 1310.
37. E. Greiser: Bewertender Arzneimittelindex:
Hypnotika, vol. 2, Med. Pharm. Verlag,
Wiesbaden 1983, pp. 1 – 25.
38. J. Krieglstein, D. Grusla, Dtsch. Apoth. Ztg.
128 (1988) 2041.
39. E. F. Domino, Annu. Rev. Pharmacol. 2
(1962) 215.
40. F. Hunzicker, F. Künzle, J. Schmitz, Helv.
Chim. Acta 49 (1966) 1 433. Wander,
GB 1 006 156, 1965; US 3 389 139, 1968.
41. G. Stille, A. Sayers, H. Lauener, E.
Eichenberger, Psychopharmakologia 28
(1973) 325; Chem. Abstr. 78 (1973) 131 982 c.
Y. Take et al., Takeda Kenkyushoho 29 (1970)
416; Chem. Abstr. 74 (1971) 21 815 n. H.
Yokotani et al., Takeda Kenkyushoho 29
(1970) 441; Chem. Abstr. 74 (1971) 21 816 p.
42. W. M. McLamre, S. Y. P. An, A. Barley, J. Org.
Chem. 20 (1955) 109. Pfizer, US 2 746 900,
1953.
43. Smith, Kline and French, DE-AS 1 183 333,
1963.
44. H. C. Caldwell, H. J. Adams, D. E. Rivard, J. V.
Swintovsky, J. Pharm. Sci. 56 (1967) 920.
45. British Drug Houses, GB 874 246, 1959.
46. Am. Home Prod., US 2 784 237, 1957.
47. Danske Medicinal, US 2 931 838, 1960.
48. A. Pictet, F. H. Reichel, Helv. Chim. Acta 6
(1923) 621. W. Freudenberg, A. M. Vajda, J.
Am. Chem. Soc. 59 (1937) 1955.
49. C. Bischoff, Ber. Dtsch. Chem. Ges. 7 (1874)
628.
50. Rheinpreussen, DE-AS 1 021 843, 1951.
51. Schering, US 2 816 910, 1953, Dtsch. Prior.
1952.
52. P. Läuger, M. Prost, R. Charlier, Helv. Chim.
Acta 42 (1959) 2379. Labaz, US 2 931 828,
1960.

53. Carlo Erba, GB 736 340, 1955.
54. G. Newbery, J. Chem. Soc. 127 (1925) 295.
Miles, US 2 854 379, 1958; US 2 931 832,
1960.
55. Bayer, DE 225 710, 1909. Friedländer 10,
1 160.
56. Bayer, DE 286 760, 1913. Friedländer 12,
704. Bayer, DE 327 129, 1917. Friedländer
13, 809.
57. Knoll, DE 185 962, 1906. Friedländer 8, 1217;
US 914 518, 1909.
58. Merck Index 10 (1983) 773. E. Schröder, C.
Rufer, R. Schmiechen: Pharmazeutische
Chemie, Thieme Verlag, Stuttgart – New York
1982, pp. 251 – 252.
59. Bayer, DE 249 241, 1912. Friedländer 10,
1 164. Labs. Aub., CH 374 644, 1964.
60. Millmaster Onyx, US 3 282 998, 1966. E. H.
Volwiler, D. L. Tabern, J. Am. Chem. Soc. 58
(1936) 1 352.
61. M. Bockmühl, O. Schaumann, Dtsch. Med.
Wochenschr. 54 (1928) 270.
62. Lion Dentifrice (Tokio), DE-OS 1 768 445,
1967; US 3 639 457, 1967; DE-OS 2 753 792,
1977.
63. N. Okudeira, S. Torii, S. Endo,
Psychopharmacology 70 (1980) 117. K.
Hillier, Drugs Future 6 (1981) 536 – 537.
64. W. J. Doran in F. F. Blicke, R. H. Cox (eds.):
“Barbituric Acid Hypnotics,” Med. Chem. Ser.
Monogr. 4 (1959).
65. R. Y. Levina, F. K. Velichko: “Advances in the
Chemistry of Barbituric Acids,” Russ. Chem.
Rev. (Engl. Transl.) 29 (1960) 437 – 459.
66. Kirk-Othmer, 2nd ed., 3, 60 – 77.
67. A. Schlieper, Justus Liebigs Ann. Chem. 55
(1845) 251; 56 (1845) 1.
68. R. Gross, Internist 17 (1976) 229.
69. R. J. Levina, F. K. Velichko: “Advances in the
Chemistry of Barbituric Acids,” Russ. Chem.
Rev. (Engl. Transl.) 29 (1960) 437.
70. E. Fischer, A. Dilthey, Justus Liebigs Ann.
Chem. 335 (1904) 338. Merck, DE 146 496,
1903.Friedländer 7, 651.
71. Bayer, DE 247 952, 1911, Friedländer 11,
926.
72. Poulenc Frères, US 1 609 520, 1926.
73. Lambert Thorpe Co., US 1 255 951, 1917; US
1 576 014,1926.
74. E. Leyraud, GB 191 008, 1922. Eli Lilly,
US 1 514 573, 1926; US 1 856 792, 1932.
75. H. H. Frey, M. P. Magnussen, Arzneim. Forsch.
16 (1966) 612.
76. Bayer, DE 442 655, 1924; DE 231 150, 1924.
77. Hoffmann-La Roche, US 1 444 802, 1923.
Hypnotics 29
78. Ciba, DE 268 158, 1911.
79. Hoffmann-La Roche, DE 526 854, 1930.
80. M. Seefelder, Justus Liebigs Ann. Chem. 652
(1962) 107. BASF, Ludwigshafen, Festschr.
Carl Wurster 60. Geburtstag 1960, 71; Chem.
Abstr. 56 (1962) 9 927 i; BASF FP 869 Méd.
1961.
81. Geigy, FR 870 714, 1941; DE 756 489, 1941.
82. Eli Lilly, US 1 856 792, 1932.
83. Bayer, DE 293 163, 1915. Geigy, GB 650 354,
1948.
84. J. D. Riedel AG, US 1 739 662, 1929.
85. A. C. Cope, E. M. Hancock, J. Am. Chem. Soc.
61 (1939) 776. Sharp & Dohme, US 2 187 703,
1940.
86. E. H. Volwiler, J. Am. Chem. Soc. 47 (1925)
2236.
87. Compagnie de Bethune, DE 589 947, 1930.
Friedländer 19, 1 198.
88. J. D. Riedel AG, US 1 622 129, 1927;
DE 481 733, 1923.
89. Eli Lilly, US 1 954 429, 1934.
90. O. Schneider, H. Frick, A. H. Lutz, Experientia
10 (1954) 135. Festschrift F. Reinhardt AG,
Basel 1936, p. 195. Hoffmann-La Roche,
US 2 151 047, 1938; US 2 680 116, 1954.
91. Ciba-Geigy, US 2 673 205, 1954,
Schweiz.-Prior. 1951; DE 950 193, 1952. L.
Hoffmann, E. Tagmann, Helv. Chim. Acta 35
(1952) 1 541.
92. Richardson-Merrell Inc. BE 753 846, 1971;
DE-AS 2 035 636, 1969.
93. H. Koch, J. Kotlan, Monatsh. Chem 97 (1966)
1 648;Kwizda, DE-AS 1 695 346, 1966.
94. L. Stockinger, H. Koch, Arzneim. Forsch. 19
(1969) 167.
95. G. Argyropoulos, G. Lanner, E. Pasquali,
Wien. Med. Wochenschr. 126 (1976) 75; Med.
Actual. 16 (1980) 388.
96. K. H. Boltze et al., Arzneim. Forsch. 13 (1963)
688.
97. B. Dubnick, C. A. Towne in A. Cerletti, F. J.
Bove (eds.): “The Present Status of
Psychotropic Drugs,” Proc. of the VI. Internat.
Congr. of the Coll. Internat.
Neuro-Psychopharmacologicum, Tarragona
Apr. 1968 pp. 24 – 27; Excerpta Med. 1969,
77 – 83.
98. I. K. Kacker, S. H. Zaheer, J. Indian Chem.
Soc. 28 (1951) 344; Chem. Abstr. 46 (1952)
11 210. General Aniline u. Film Corp.,
US 3 213 094, 1963. Lab. Toraude,
GB 843 073, 1960.
30 Hypnotics
99. Beiersdorf AG, BE 615 282, 1962; Chem.
Abstr. 58 (1963) 13 971 ef. S. S. Parmar, K.
Kishok, P. K. Seth, R. C. Arona, J. Med. Chem.
12 (1969) 138.
100. J. Klosa, J. Prakt. Chem. 14 (1961) 84; 20
(1963) 283.
101. E. Szirmai, Ther. Umsch. 20 (1963) 542.
102. L. H. Sternbach: “The Benzodiazepine Story,”
Progr. Drug Res. 22 (1978) 229 – 266. W.
Schallek, W. D. Horst, W. Schlosser:
“Mechanism of Action of Benzodiazepines,”
Adv. Pharmacol. Chemother. 16 (1979)
45 – 87. L. O. Randall, W. Schallek, L. H.
Sternbach, R. Y. Ning in M. Gordon (ed.):
“Chemistry and Pharmacology of the
1,4-Benzodiazepines,”
Psychopharmacological Agents, A Series of
Monographs in Medicinal Chemistry, vol. 3,
Academic Press, New York – San
Francisco – London 1974, p. 175. V. C.
Williams, G. C. Varnado, P. U. Nwangwu:
“Perspectives on the Clinical Pharmacology of
Benzodiazepines,” Med. Actual. 21 (1985)
75 – 96. T. A. Hamor, J. L. Martin in G. P. Ellis,
G. B. West (eds.): “The Benzodiazepines,”
Progress in Medicinal Chemistry 20 (1985)
158 – 223. A. Vela-Bueno, A. Kales:
“Benzodiazepine Hypnotics in the
Multidimensional Treatment of Insomnia,”
Med. Actual. 22 (1986) 271 – 281. C. Pina, A.
Catedra: “Newer Benzodiazepines, Med.
Actual. 11 (1975) 176; “Chemical Synthesis of
Benzodiazepines (nur Literatur),” Med.
Actual. 11 (1975) 211 – 212. H. Schütz:
Benzodiazepines: A Handbook, Springer
Verlag, Berlin 1982. C. Bellantuono, V. Reggi,
G. Tognoni, S. Garattini: “Benzodiazepines –
Clinical Pharmacology and Therapeutic Use,”
Drugs 19 (1980) 195 – 219. S. Garattini: The
Benzdiazepines, Raven Press, New York 1973.
103. M. Lader (ed.): “New Perspectives in
Benzodiazepine Therapy,” Arzneim. Forsch.
30 (1980) no. 5 a, 851 – 913. U. Klotz:
“Clinical Pharmacology of Benzodiazepines,”
in Progress in Clinical Biochemistry and
Medicine, vol. 1, Springer,
Berlin – Heidelberg – New York – Tokyo 1984,
pp. 117 – 166.
104. H. Schrum: “Benzodiazepine als Schlafmittel,”
Schweiz. Rundsch. Med. Prax. 74 (1985)
1301 – 1304.
105. R. Weber, Dtsch. Apoth. Ztg. 128 (1988)
2643 – 2652.
106. L. L. Iversen, Nature (London) 275 (1978)
447. R. E. Study, J. L. Barker, JAMA J. Am.
Med. Assoc. 247 (1982) 2147 – 2151. U.
Klotz: Progress in Clinical Biochemistry and
Medicin, Springer, Berlin – Heidelberg – New
York – Tokyo 1984, pp. 127 – 129. T. A.
Hamor, J. L. Martin in G. P. Ellis, G. B. West
(eds.): Progress in Medicinal Chemistry 20
(1985) p. 183.
107. W. E. Müller, MMW Münch. Med. Wochenschr.
124 (1982) 879 – 884; Dtsch. Apoth. Ztg. 126
(1986) 1036 – 1044.
108. W. E. Müller, Med. Actual. 24 (1988)
649 – 663.
109. E. Müller, Dtsch. Apoth. Ztg. 128 (1988)
672 – 674. Roche Magazin 1988, no. 31, 3 – 13
(Hoffmann-La Roche).
110. S. Ebel, H. Schütz: “Biotransformation von
Benzodiazepin-Derivaten,” Pharm. Unserer
Zeit 8 (1979) 87. U. Klotz, L. Kangas, J.
Kanto: “Clinical Pharmacocineties of
Benzodiazepines,” Prog. Pharmacol. 3 (1980)
1 – 72. W. E. Müller: “über die Bedeutung der
Plasmaproteinbindung für Verteilung und
Wirkung von Tranquillantien vom
Benzodiazepintyp,” Klin. Wochenschr. 55
(1977) 105 – 110.
111. R. G. Borland, A. N. Nicholson, Br. J. Clin.
Pharmacol. 2 (1975) 9. S. C. Harvey in
Goodman Gilman, The Pharmacological Basis
of Therapeutics, 7th ed., Macmillan Publ.,
New York, Toronto, London 1985, p. 352; see
also [102] p. 82.
112. Hoffmann-La Roche, DE-AS 1 136 709, 1959;
DE-AS 1 145 626, 1959; US 3 109 843, 1961;
US 3 116 203, 1962; US 3 123 529, 1964.
Delmar Chem., DE-AS 1 811 785, 1968.
113. L. H. Sternbach et al., J. Med. Chem. 6 (1963)
261. Hoffmann-La Roche, US 3 116 203,
1963; US 3 123 529, 1964; US 3 203 900,
1964; GB 1 039 372, 1966.
114. M. W. Johns, Drugs 9 (1975) 448. J. Rieder,
Arzneim. Forsch. 23 (1973) 212. Martindale:
The Extrapharmacopoea, 28th ed.,
Pharmaceutical Press, London 1982,
pp. 807 –809. Med. Actual. 1 (1964)
132 – 138; 9 (1973) 237.
115. J. M. Monti, H. Altier, Psychopharmacologia
32 (1973) 343. W. Schallek, J. Kovacs, A.
Kuehn, J. Thomas, Arch. Int. Pharmacodyn.
Ther. 206 (1973) 161. S. A. Kaplan et al., J.
Pharm. Sci. 63 (1974) 527. R. M. Joye, A. J.
Hance, K. F. Killam, Neuropharmacology 10
(1971) 483. J. Freuchen, J. Østergaard, B. O.
Mikkelsen, Curr. Ther. Res. 23 (1978) 90.
R. T. Owen, Med. Actual 19 (1983) 163 – 167;

11 (1975) 176 – 178. M. A. K. Mattila, H. M.
Larni, Drugs 20 (1980) 353 – 374.
116. L. H. Sternbach et al., J. Med. Chem. 8 (1965)
815. Hoffmann-La Roche, BE 629 005, 1963;
US Prior. 1962; Chem. Abstr. 60 (1963)
1 589 f; NL 6 401 335, US Prior. 1963; Chem.
Abstr. 62 (1965) 5288.
117. Hoffmann-La Roche, US 3 567 710, 1971.
118. S. Inaba, K. Ishizumi, H. Yamamoto, Chem.
Pharm. Bull. 19 (1971) 263.
119. L. O. Randall et al., Arch. Int. Pharmacodyn.
Ther. 178 (1969) 216. H. Jick, Curr. Therap.
Res 9 (1967) 355. J. d’Armagnac et al.,
Therapie 26 (1971) 439. A. Schwartz, F. M.
Vane, E. Postma, J. Med. Chem. 11 (1968)
770. Med. Actual. 6 (1970) 217. USP XXI,
447 – 449, Rockville, Md. 1985. AMA: Drug
Evaluations, 6th ed., Am. Med. Assoc.,
Chicago 1986, pp. 99 – 100.
120. S. Inaba, K. Ishizumi, K. Mori, H. Yamamoto,
Chem. Pharm. Bull. 19 (1971) 722.
Hoffmann-La Roche, US 3 144 439, 1964; US
3 144 890, 1964; US 3 243 427, 1966. Dumex,
ZA 6 706 791, 1968; Chem. Abstr. 70 (1969)
57 916 c. Sumitomo, DE-OS 1 816 046, 1970.
121. S. Sakai, S. Kitajawa, H. Yamamoto, Arzneim.
Forsch. 22 (1972) 534. H. Yamamoto et al.,
Sumitomo Kagaku, Tokushugo 1977, 13 – 23;
Chem. Abstr. 90 (1977) 26 w.
122. Martindale: The Extra Pharmacopoeia, 21st
ed., The Pharmaceutical Press, London 1982,
p. 1549. USP XXI, 762, Rockville, Md. (1985).
123. M. J. Bluckman et al., Arzneim. Forsch. 21
(1971) 1049 – 1102. Martindale: The Extra
Pharmacopoeia, 21st ed., The Pharmaceutical
Press, London 1982, p. 1543.
124. S. C. Bell et al., J. Med. Chem. 11 (1968) 457.
A. Nudelman, J. Pharm. Sci. 63 (1974) 1886.
Amer. Home Prod., BE 621 819, 1963; Chem.
Abstr. 60 (1964) 2993 b; DE-OS 1 445 412,
1962; DE-OS 1 645 904, 1962;
DE-OS 1 795 553, 1962; DE-OS 1 795 556,
1962.
125. A. Doenicke et al., Anaesthesist 28 (1979)
578. H. Ott et al., Anaesthesist 28 (1979) 29.
R. Girkin et al., Xenobiotica 10 (1980) 401.
M. Hümpel et al., Eur. J. Drug. Metab.
Pharmacokinet. 4 (1979) 237. J. Oswald et
al., Br. Med. J. 1982, I, 860. R. T. Owen, Drugs
Future 5 (1980) 495 – 496. Pharmcast 87-10,
6-II-136 – 6-II-138, Technomic Inf. Service,
Tokyo.
126. S. C. Bell, S. J. Childress, J. Org. Chem. 27
(1962) 1691. Am. Home Prod., US 3 197 467,
Hypnotics 31
1965. Hoffmann-La Roche, US 3 340 253,
1967; US 3 374 225, 1968.
127. L. O. Randall, C. L. Scheckel, W. Pool, Arch.
Int. Pharmacodyn. Ther. 185 (1970) 135. P.
Sarteschi et al., Arzneim. Forsch. 22 (1972)
93. H. J. Schwandt, W. Sadee, K.-H. Beyer,
Xenobiotica 4 (1974) 733. S. H. Curry, A. N.
Nicholson, R. Whelpton, Br. J. Pharmacol. 57
(1976) 427. AMA: Drug Evaluations, 6th ed.,
Am. Med. Assoc. Chicago 1986,
pp. 102 – 103. Drug, Facts and Comparisons,
Facts and Comparisons Div., St. Louis 1982,
pp. 991 – 993 (a division of J. B. Lipincott Co.,
Philadelphia –Toronto).
128. G. F. Tamagnone, R. De Maria, F. De Marchi,
Arzneim. Forsch. 25 (1975) 720. Schiapparelli,
DE-OS 2 338 058, 1972. Japk 7480 086,
GB 1 431 282, 1972. Kleemann-Engel:
Pharmazeutische Wirkstoffe, Suppl.
1982 – 1987, Thieme-Verlag, Stuttgart – New
York 1987, pp. 1151 – 1152.
129. Hoffmann-La Roche, ZA 6 802 239, 1967.
130. G. F. Tamagnone, M. V. Torrielli, F. De Mardi,
J . Pharm. Pharmacol. 26 (1974) 566. V.
Monaco, A. Bontante, Curr. Ther. Res. 37
(1985) 626. L. De Angelis, Drugs Future 3
(1978) 145 – 148. S. Mardene, C. Biechie,
G. M. Nano, Ther. Drug Monit. 3 (1981)
351 – 356. M. Babbini, M. V. Torielli, E.
Strumia et al., Arzneim. Forsch. 25 (1975)
1294.
131. Gerot-Pharmazeutika, DE-OS 2 950 235, 1980;
AT 379 391, 1984; Chem. Abstr. 104 (1986)
168 499 j.
132. H. Koch, Drugs Future 7 (1982) 99.
133. M. Steinmann, J. G. Topliss, R. Alekel et al., J.
Med. Chem. 16 (1973) 1354. Schering,
US 3 920 818, 1974. Scherico Ltd.,
DE-OS 2 138 733, 1971.
134. H. Winsauer, D. E. O’Hair, Curr. Ther. Res. 35
(1984) 228. A. Barnett, L. C. Jorio, E. Ongini,
Arzneim. Forsch. 32 (1982) 1452. J. Castaner,
Drugs Future 3 (1978) 139; 8 (1983) 165; 9
(1984) 152; 10 (1985) 171. S. J. Ankier, K. A.
Goa, Drugs 35 (1988) 42 – 62. AMA: Drug
Evaluations, 6th ed., Am. Med. Assoc.,
Chicago 1986, p. 102. Annu. Drug Data
Report 8 (1986) 119; 2 (1979 – 1980) 192.
135. T. Miyadere et al., J. Med. Chem. 14 (1971)
520. Sankyo, DE-OS 1 812 252, 1969; J-Kokai
1967; Chem. Abstr. 71 (1969) 124516 h.
136. T. Kamioka, J. Nakayama, T. Hara, H. Takaji,
Arzneim. Forsch. 28 (1978) 838. Y. Sakai,
Jpn. J. Pharmacol. 33 (1983) 1017 – 1025. M.
Yoshimoto et al., Chem. Pharm. Bull. 25
32 Hypnotics
(1977) 1378. Med. Actual. 17 (1981)
283 – 286. P. J. Thorpe, Drugs Future 6 (1981)
414 – 415.
137. J. B. Hester, D. J. Duchamp, G. C. Chichester,
Tetrahedron Lett. 1971, 1609. J. B. Hester,
A. D. Rudzick, B. V. Kamdar, J. Med. Chem.
14 (1971) 1078. K. Meduro, H. Tawada, Y.
Kuwada, Chem. Pharm. Bull. 21 (1973) 1619,
2382. K. Kamiya, Y. Wada, M. Nishikawa,
Chem. Pharma. Bull. 21 (1973) 1520. K.
Megruo, Y. Kuwada, Tetrahedron Lett. 1970,
4039. D. L. Loffen, R. J. Fryer, D. A. Katonak,
F. Wong, J. Org. Chem. 40 (1975) 894.
138. Upjohn, DE-OS 2 012 190, 1969;
US 3 701 782, 1969; DE-OS 2 302 525, 1973;
US 4 018 788, 1977; US 3 987 052, 1969.
Toyama, DE-OS 1 955 349, 1971. Takeda,
DE-OS 2 114 441, 1971; DE-OS 1 965 894,
1969; US 4 116 956, 1976.
139. S. Tanayama, Y. Shirakawa, Y. Kanai, Z.
Suzuoki, Xenobiotika 4 (1974) 33. Y. Kanai,
Xenobiotika 4 (1974) 441. F. S. Eberts,
Pharmacologist 16 (1974) 196; Drug Metab.
Dispos. 5 (1977) 547. H. Isozaki, M. Tanaka,
K. Inanaga, Curr. Ther. Res. 20 (1976) 493. T.
Momose, S. Shii, T. Kuge, Curr. Ther. Res. 19
(1976) 277. W. Ehrenstein, M.
Müller-Limmroth, M. Opfermann, F. Weber,
Arzneim. Forsch. 32 (1982) 280. J. K. Walsh
et al., Curr. Ther. Res. 36 (1984) 866. Med.
Actual. 12 (1976) 355 – 357. Annual Drug
Data Report 2 (1979/1980) 85 – 86.
140. J. B. Hester, A. D. Rudzick, B. V. Kamdar, J.
Med. Chem. 14 (1971) 1078.
141. A. Kales, M. B. Scharf, J. D. Kales, Science
(Washington, D.C.) 201 (1978) 1039.
142. T. Roth, M. Kramer, T. Lutz, Drugs Exp. Clin.
Res. 1 (1977) 279. D. A. C. Hegelbach-Feller,
J. M. Tschopp, S. Christeller, J. Fabre,
Arzneim. Forsch. 38 (1988) 387.
143. M. Leibowitz, A. Sunshine, J. Clin.
Pharmacol. 18 (1978) 302. K. K. Okawa,
G. S. Allen, J. Int. Med. Res. 6 (1978) 343.
B. B. Knap, E. L. Boyd, G. Linsenmeyer, O. J.
Linet, Curr. Ther. Res. 23 (1978) 230.
144. A. N. Sing, Curr. Ther. Res. Clin. Exp. 27
(1980) 627. A. N. Nicholson, Br. J. Clin.
Pharmacol. 9 (1980) 223.
145. L. Lasagnon, Lancet 1 (1980), 815.
146. E. A. Swinyard in Remingtons Pharmaceutical
Sciences, 17th ed., Mack Publ., Easton, Pa.
1985, p. 1064. AMA: Drug Evaluation, 6th
ed., Am. Med. Assoc., Chicago 1986, p. 103. J.
Castaner, Drugs Future 1 (1976) 393 – 396; 2
(1977) 558.
147. A. Walser et al., J. Org. Chem. 43 (1978) 936.
Hoffmann-La Roche, DE-OS 2 540 522, 1976.
148. M. Gemper et al., Arzneim. Forsch. 31 (1981)
2177 – 2288. D. A. C. Hegelbach-Feller, J. M.
Tschopp, S. Christeller, J. Fabre, Arzneim.
Forsch. 38 (1988) 387. J. Kanto, Med. Actual,
19 (1983) 376 – 381.
149. K. H. Weber, W. Sirrenberg, O. Spohn, H.
Daniel, Arzneim. Forsch. 36 (1986) 518.
Boehringer, DE-OS 2 410 030,
1974 = US 4 094 984, 1975; DE-OS 2 503 235,
1975; DE-OS 2 533 924, 1975. Hoffmann-La
Roche, DE-OS 2 221 623, 1971. K. Gewald,
Chem. Ber. 98 (1965) 3571.
150. N. Ferrara et al., Curr. Ther. Res. 37 (1985)
295. J. Grünberger et al., Curr. Ther. Res. 24
(1978) 427. B. Saletu, J. Grünberger, J.
Volavka, P. Berner, Arzneim. Forsch. 29
(1979) 700. M. S. Langley, S. P. Clissold,
Drugs 35 (1988) 104 – 122. Pharmcast 87-10,
Technomic Information Service, Tokyo 1987,
pp. 6-II-151 – 6-II-152.
151. R. Agar et al., J. Med. Chem. 20 (1977) 1035.
Russel-UCLAF, DE-OS 2 605 652, 1976; US
4 044 142, 1977.
152. T. G. Johns, D. C. Piper, G. W. L. James, Arch.
Int. Pharmacodyn. Ther. 240 (1979) 53. J.
Hindmarch, C. A. Clyde, Drugs Exp. Clin.
Res. 2 (1980) 61. R. T. Owen, Med. Actual. 20
(1984) 59 – 63. Loprazolam Clinical
Workshop, The Medicine Publishing
Foundation, Oxford, 1983.
153. K. Hirai et al., Heterocycl. Chem. 19 (1982)
1363; J. Med. Chem. 23 (1980) 764; 24 (1981)
20. Shionogi, DE-OS 2 725 164, 1977;
US 4 159 374, 1977. Hoffmann-La Roche,
DE-OS 2 537 069, 1975.
154. K. Hirose, A. Matsushita, M. Eigyo, Arzneim.
Forsch. 31 (1981) 63. Drugs Future 9 (1984)
522 – 524; 11 (1986) 626 – 628. Pharmcast
87-1, Technomic Information Service, Tokyo
1987, pp. 5-III-139 – 5-III-141.
155. Rhône-Poulenc, DE-OS 2 300 491, 1973; US
3 862 149, 1975. C. Jeanmart, C. Cotrel, C. R.
Ser. C 287 (1978) 377.
156. R. Duriez, Therapie 34 (1979) 317. E.
Wickström, K. E. Giercksky, Eur. J. Clin.
Pharmacol. 17 (1980) 93; Clin. Therap. 3
(1978) 59. . J. C. Blanchard, A. Boireau, C.
Garrett, L. Julou, Life Sci. 25 (1979) 2417.
P. H. Wu, J. W. Phillis, A. S. Bender, Life Sci.
28 (1981) 1023. J. Castaner, Drugs Future 4
(1979) 59 – 60; 5 (1980) 56; 6 (1981) 62 – 63;
7 (1982) 69 – 70; 8 (1983) 79; 9 (1984) 81; 10
 Hypnotics 33

(1985) 88 – 89; 11 (1986) 76 – 78; 12 (1987)
95; 13 (1988) 88 – 89. Pharmacast 87-7,
Technomic Inf. Service, Tokyo 1987, 1989,
6-I-87 – 6-I-89. Int. Pharmacopsychiatry 17
(1982) (Suppl. 2), 1 – 252. K. A. Goa, R. C.
Heel, Drugs 32 (1986) 48 – 65.
157. G. Houghton et al., Int. J. Clin. Pharmacol.
Biopharm. 23 (1985) 97. R. Duriez et al.,
Therapie 34 (1979) 317. J. Gaillot et al.,
Pharmacology 27 (1983) (Suppl. 2) 76 – 91.
158. Synthelabor, EP 50 563, 1981. P. George, H.
Depoortere, J. Pharm-Sci. (in Press).
159. S. Arbilla et al., Naunyn-Schmiedebergs Arch.
Pharmacol. 330 (1985) 248. M. Dimsdale et
al., Drugs Future 12 (1987) 777 – 780.
160. M. Dimsdale et al., Drugs Future 13 (1988)
106 – 109.
161. J. Maillard, M. Vincent, M. Bernard, Bull. Soc.
Chim. Fr. 1961, 529; 1962, 1580.
Logeais-Labs., GB 902 388, 1962;Fr. M. 379,
1962; Chem. Abstr. 57 (1962) 15251 g.

Hypochlorous Acid → Chlorine

Hypochlorous Acid → Chlorine Oxides and Chlorine Oxygen Acids
Hypoglycemic Drugs → Antidiabetic Drugs
Hypophosphorus Acid → Phosphorus Compounds, Inorganic

Hypnotics
Hartmund Wollweber, Bayer AG, Wuppertal, Federal Republic of Germany
1. Introduction . . . . . . . . . . . . . . . . .
1.1. Amino Acids and Hormones Influenc-
ing Sleep . . . . . . . . . . . . . . . . . . .
1.2. Influence of Hypnotics on Sleep . . . .
1.3. Requirements of Hypnotics . . . . . . .
1.4. Applications of Hypnotics . . . . . . . .
1.5. Consumption of Hypnotics . . . . . . .
1.6. Classification and Evaluation of Hyp-
notics . . . . . . . . . . . . . . . . . . . . .
2. Sedatives, Antihistamines, and Tran-
quilizers . . . . . . . . . . . . . . . . . . . .
3. Alcohols and Aldehydes . . . . . . . . .
1. Introduction
Hypnotics (Greek υπνoζ, sleep) are drugs that
promote the onset of sleep and induce a state
similar to natural sleep by depressing the cen-
tral nervous system [1–28]. Hypnotics are not
a sharply defined group of drugs because their
sleep-inducing effect greatly depends on the
dose. Small doses of a hypnotic have a sedative
action; increasing doses cause, in succession, a
hypnotic effect, a narcotic effect, and finally se-
vere respiratory depression and death.
Sleep that is properly induced by drugs ap-
pears to resemble physiological sleep closely.
Both states share the following symptoms:
– reduction of heart rate
– reduction of metabolism
– decrease in concentration of calcium ions in
blood
– decrease in muscle tone and blood pressure
– contraction of bronchi and
– contraction of pupils
In contrast to narcotized persons, sleeping
persons can be awakened.
Physiology of Sleep. Marked differences
between drug-induced sleep and physiologi-
cal sleep can, however, be detected by using
polygraphic methods, such as measurement of
pH changes and eye movements (electrooculo-
gram), electromyogram, and electroencephalo-
gram (EEG). On the basis of data obtained from

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a13 533
Hypnotics 1
1 4. Urethanes, Acyclic Urea Derivatives,
and Amides . . . . . . . . . . . . . . . . . 8
2 5. Cyclic Carboxamides, Carboximides,
3 Cyclic Urea Derivatives, and Related
4 Compounds . . . . . . . . . . . . . . . . . 9
4 5.1. Barbituric Acids . . . . . . . . . . . . . . 10
4 5.2. Piperidinediones . . . . . . . . . . . . . . 15
5.3. Quinazolinones . . . . . . . . . . . . . . . 17
4
5.4. Benzodiazepines . . . . . . . . . . . . . . 18
5 6. Other Hypnotics . . . . . . . . . . . . . . 27
6 7. References . . . . . . . . . . . . . . . . . . 28
these methods, two types of sleep of importance
for the restoration of the body have been de-
scribed:
1) Orthodox, synchronized sleep or nonrapid
eye movement (NREM, nonREM) sleep can
be distinguished by δ waves corresponding
to periods of reduced activity on the EEG (δ
sleep). Since these waves are slow, this phase
is also referred to as slow wave sleep (SWS).
2) Paradox sleep or rapid eye movement (REM)
sleep has high EEG activity.
The two types of sleep form a cycle of 110 –
140 min; four to five cycles are observed in
adults each night (Fig. 1). Sleep is extremely im-
portant for the renewal of strength. The main
phases of natural sleep are an initial deep, dream-
less sleep, during which the body temperature
falls, and a light, dream-filled, restless sleep be-
fore waking. The depth of orthodox sleep has
four levels, as seen on the EEG (Fig. 1A), which
can partly be correlated with certain body func-
tions [10].
The areas of the brain that regulate body tem-
perature and the sleep – wake cycle are close
to each other and are functionally related. The
modulations of sleep are correlated with the cir-
cadian rhythm of body temperature. The on-
set of sleep at night coincides with a lowering
of body temperature, characteristic of NREM
sleep. People who display a temperature differ-
ence of less than 1 ◦ C are short sleepers and those
2 Hypnotics
Figure 1. Different phases of normal adult sleep (ca. 8 h)
A) Electroencephalogram; B) Depth of sleep; C) Body temperature; D) Body movement
who display larger differences (>1 ◦ C) are long
sleepers. Body tissues are regenerated during or-
thodox sleep, and cerebral functions during the
REM phase.
The REM phases account for ca. 20 – 25 %
of sleep and appear to be periods of relatively
high cerebral activity in which dreams occur.
They are detected by measuring the spontaneous
movements of the closed eye. In the absence of
these dream phases sleep is not restful. Dreams
are of great importance for learning and mem-
ory; they assimilate the experiences of the pre-
ceding days. Generally, dreams can be remem-
bered slightly or not at all. With increasing age,
the duration of REM sleep steadily decreases
to about 20 – 30 % of its original duration. The
REM phase is especially activated when a per-
son is confronted with an emotionally arousing,
intellectual task calling for creative effort, i.e.,
when alterations in thinking or behavior are nec-
essary. Confrontation with problems also inten-
sifies REM sleep, e.g., an astronaut experiences
greatly increased REM sleep in his first night
in space. Continuous suppression of REM sleep
leads to psychological disturbances and psycho-
somatic symptoms.
1.1. Amino Acids and Hormones
Influencing Sleep
Three neurohormones are involved in control-
ling sleep: serotonin, noradrenaline, and acetyl-
choline. Serotonin regulates orthodox sleep,
whereas acetylcholine and noradrenaline influ-
ence the REM phase. The rates of cell division
are the highest during sleep, and large amounts
of growth hormone (GH), which stimulates pro-

tein synthesis, are secreted during the third and
fourth levels of orthodox sleep. The EEG waves
at this stage are also large and slow.
Melatonin is rhythmically secreted by the
pineal gland. It stimulates specific receptors in
the body’s biological clock, the suprachiasmatic
nuclei of the hypothalamus and may thus regu-
late the human circadian rhythm by increasing
serotonin levels. The human biological clock can
be “adjusted” with exogenous melatonin. Conti-
nous exposure of humans to light leads to sleep
disorders by inhibiting melatonin synthesis [16],
[29], [30].
A sleep inducing nonapeptide, delta-sleep-
inducing peptide [62568-57-4] (DSIP), has been
isolated from rabbit serum. This peptide, which
also occurs in humans, has been synthesized
and has been prepared by genetic engineering
techniques. DSIP coordinates the rhythmic cir-
cadian sleep – wake cycles by promoting sleep
functions (without having a sedative effect), and
enhances mental performance when awake. At
the same time, DSIP invokes a higher tolerance
to stress. The action of this nonapeptide is at-
tributed to modulation of the adrenergic recep-
tors [31].
l-Tryptophan, a common amino acid found
in the human body is a precursor of serotonin. In
daily doses of 4 – 7 g it causes sedation during
the day and a shortening of latency of sleep at
night. It is registered in Switzerland as a “bio-
logical sleep-inducing drug.” However, studies
have questioned the sleep-modulating effect of
this amino acid [32].
1.2. Influence of Hypnotics on Sleep
Most hypnotic drugs shorten both phases of
sleep and suppress the paradoxical (REM)
phase. Thus, the sleep curve is levelled out, and
the EEG resembles that of a narcotized per-
son. Hence, drug-induced sleep does not have
the restorative quality of natural sleep and is
no substitute for the treatment of insomnia. A
research report of the Max Planck Institut für
Psychiatrie (München, 1980) states that conven-
tional sleep-inducing drugs do not produce rest-
ful sleep; since drug-induced sleep is not iden-
tical to natural sleep, taking sleeping pills is not
advised. Pharmacologists also recommend the
use of hypnotic drugs only for a few days, if at
Hypnotics 3
all, because of their many adverse effects. Alco-
hol should not be taken with sedatives or hyp-
notics because the reactions of the patient are
then adversely affected.
Prolonged use of sedatives and hypnotics re-
duces their effect on the cells of the central ner-
vous system (see also Section 5.1). Most hyp-
notics, particularly highly active preparations,
can create psychological and physical depen-
dence, resulting in withdrawal symptoms when
treatment is terminated. In addition, many hyp-
notic drugs that suppress and shorten the REM
phase produce a rebound effect, i.e., a sudden
withdrawal after extended periods of use causing
a prolongation of the REM phase. As in alcohol
withdrawal, this effect can even lead to delirium
tremens. In this way, the body tries to make up
for lost REM sleep.
1.3. Requirements of Hypnotics
There is still no completely satisfactory hypnotic
drug. The following requirements can be defined
for an ideal hypnotic:
– reliable effect
– wide therapeutic range
– limited duration of action, no longer than 8 h
– physiological sleep
– no disturbance of deep sleep or REM sleep
– low toxicity
– no enzyme induction
– no respiratory depression
– no additive effect with alcohol
– no amnesia, confusion, or ataxia after arousal
in the night
– no excessive daytime sedation
– no loss of potency on repeated use
– no risk of dependency, e.g., insomnia after
termination of treatment
– no after effects (e.g., hangover)
1.4. Applications of Hypnotics
Sleep-inducing drugs are used by the following
groups:
1) healthy people with demanding jobs or irreg-
ular working hours, who often take sleeping
pills after stimulants such as coffee;
4 Hypnotics
2) sick people with symptoms such as fever,
sore throat, or cold;
3) old people with insomnia due to cerebral
sclerosis usually have an inverted day – night
rhythm; they are awake and mobile at night,
and tired and peaceful in the morning.
4) sick people with uncomfortable and sleep-
less nights due to pain, who have difficulty
in breathing, heart trouble, etc.; and
5) bedridden people with serious illnesses, who
are deprived of physical exercise.
People belonging to groups 1 and 2 can be
treated with mild relaxants by reducing the sus-
ceptibility of the reticular formation (site of the
center of wakefulness in the brain) to external
stimuli. These substances include plant compo-
nents with sedative properties, such as valerian
and hop, and tranquilizers or sedatives which,
even in high doses, do not have narcotic prop-
erties. People belonging to groups 4 and 5 may
have to be treated for extended periods with hyp-
notics and even with antipsychotic and antide-
pressant drugs.
1.5. Consumption of Hypnotics
The importance of hypnotic drugs is reflected in
the fact that 21 % of the population of France
(February 1975) and 32 % of the inhabitants of
Los Angeles (1973) suffered from insomnia. In
France, 2.5×106 tablets, capsules, and dragées
are taken every evening to promote sleep [33–
38]. The dramatic increase in the consumption
of hypnotics with age can only be partly justi-
fied because older people require less sleep at
night. Instead, they have a more pronounced en-
dogenous sleep profile with increased cerebral
readiness to sleep at 9.00, 13.00, and 17.00 h.
1.6. Classification and Evaluation of
Hypnotics
Hypnotics may be classified as follows [34]:
1) benzodiazepine hypnotics;
2) nonbenzodiazepine hypnotics, e.g., chloral
hydrate, barbiturates, piperidinediones, and
methaqualone;
3) other drugs with sedative effects, (e.g.,
tranquilizers, antidepressants, and antihis-
tamines); and
4) natural preparations, e.g., valerian, ethanol,
and DSIP.
There is no sharp distinction between seda-
tives and hypnotic drugs; differences in activity
are quantitative rather than qualitative. Both in-
hibit the activation system in the reticular for-
mation. Drugs belonging to groups 1 and 2 are
true hypnotics. Drugs belonging to group 3 are
only occasionally used as hypnotics.
Der Bewertende Arzneimittelindex (The Val-
uating Drug Index) assesses “benzodiazepine
sleep-inducing drugs” positively, especially be-
cause of their wide therapeutic range [37]. Other
recommended drugs are chloral and its deriva-
tives (see Sections page 7page 7), diphenhy-
dramine, and doxylamine (see Sectionpage 5).
Derivatives of barbituric acid, piperidinediones,
and quinazolinones are not recommended. In
comparison with the benzodiazepines, these
cause more undesirable effects (respiratory de-
pression, negative influence on REM sleep, en-
zyme induction, and problems of dependence).
Their therapeutic range is narrow. Compounds
containing bromine are also to be avoided be-
cause they may cause bromism (see Chap. 2).
Valerian preparations (e.g., Valepotriate) and
their components are not favored because of
their in vitro cytotoxic and alkylating action. The
therapeutic importance of these drugs cannot be
assessed at present [37].
The lists of trade names cited for individual
hypnotics are not complete.
2. Sedatives, Antihistamines, and
Tranquilizers
Sedatives and Tranquilizers. There is no
significant distinction between sedatives and
hypnotics; the same drugs are often used for both
purposes. Sedatives (→ Sedatives) and tranquil-
izers (→ Psychopharmacological Agents) are
used to sedate and relieve anxiety during the day
and, in addition, to induce sleep at night. Thera-
peutic doses of tranquilizers have a calming, re-
laxing effect and alleviate psychological symp-
toms such as fear, anxiety, restlessness, and ten-
sion. Ideally, they should not produce sedative

or hypnotic effects. The first therapeutically used
hypnotics were plant extracts with either seda-
tive (valerian root and hop) or analgesic action
(opium alkaloids). In spite of the sedative effect
of opium alkaloids [39], they should not be used
for obvious reasons (addiction, euphoria, poor
dose adjustment, and legality). Of all the plant
extracts available, only one substance, valmane,
is exactly defined from a chemical standpoint.
Valepotriate is a mixture of valtratum
[18296-44-1] (1), acevaltratum [25161-41-5]
(2), and didrovaltratum [18296-45-2] (3):
These three compounds, which occur in na-
ture in the root of Valeriana officinalis, have a
sedative, mildly psychotropic action. They act as
mild hypnotic agents because they restore physi-
ological equilibrium in vegetative disorders. The
activity of valmane has been questioned [38].
Trade Names. Valmane (Kali-Chemie, FRG),
Nevripan (Medopharm, FRG).
Simple bromine-containing sedatives
(sodium, potassium, or calcium bromide, or
the hydrobromides of magnesium glutamate
or magnesium aspartate) should not be used
under any circumstances. If given for long pe-
riods, they cause a mental condition known as
bromism, which is characterized by symptoms
such as loss of concentration and memory, in-
somnia, rash, and psychoses.
Antihistamines. The following antihis-
tamines have sedative properties and are some-
times used as hypnotics:
Meclozine or Meclizine [569-65-3], 1-
[4-chlorobenzhydryl]-4-(3-tolyl)piperazine
dihydrochloride (→Antiallergic Agents,
Chap. 2.1.4.)
Hypnotics 5
Trade Names: Calmonal (Heiden, FRG), Bonine
(Pfizer, USA).
Diphenhydramine hydrochloride [147-24-0],
2-(benzhydryloxy)-N,N-dimethylethylamine
(→Antiallergic Agents)
Trade Names: Benadryl (Parke-Davis, USA),
Halbmond (Much, FRG).
Doxylamine succinate [562-10-7], 2-di-
methylaminoethoxyphenylmethyl-2-picoline
Trade Names. Decapryn, Meteprine (Merrell,
USA), Unisom (Pfizer, USA).
Psychopharmacological Agents. Of the
large number of tricyclic antipsychotropic drugs
available, two are used as hypnotics:
Perimetazine [13093-88-4], perimet-
hazine, 1- [3-(2-methoxyphenothiazin-10-yl)-
2-methylpropyl]-4-piperidinol, C22 H28 N2 O2 S,
M r 384.54 (→Phenothiazines and Derivatives).
Trade Name. Leptryl (Roger Bellon, USA).
Perlapine [1977-11-3], 6-(4-methyl-1-piper-
azinyl)-11 H-dibenz [b,e] azepine, C19 H21 N2 ,
M r 291.40, mp 136 – 138 ◦ C.
This compound, in the form of yellow prisms,
is prepared by chlorinating 11 H-dibenz [b,e]
azepin-6-one with phosphoryl chloride to give
the imide chloride, which is subsequently treated
with N-methylpiperazine [40]. The compound
thus obtained has hypnotic properties [41]; LD50
(oral) 415 mg/kg (mouse).
Trade Names. Hypnodine (Wander, FRG;
Takeda, Japan), Pipnodine (Takeda, Japan).
3. Alcohols and Aldehydes
The discovery that after producing an initial ex-
citation phase, ethanol has a general sedative ac-
tion, led to the development of tertiary and halo-
genated alcohols as hypnotics. These substances
are rarely used today. Chloral hydrate has long-
lasting sedative and hypnotic effects because it
is metabolized in the body to 2,2,2-trichloroeth-
anol, which also has a sedative action.
6 Hypnotics
Methylpentynol [77-75-8], 3-methyl-1-
pentyn-3-ol, C6 H10 O, M r 98.14, bp 121 –
122 ◦ C, is a highly mobile liquid with a pungent
smell and burning taste. It solidifies at −30.6 ◦ C
and is soluble in almost all organic solvents. It is
produced by reacting 2-butanone with sodium
acetylide.
As a result of its short duration of action (0.5 –
1 g act for 1 – 2 h), this preparation is especially
useful in promoting sleep in cases of mild in-
somnia.
Trade Names. Atemorin (Scherer, USA), Dal-
gol, Dorison, Dormison (Schering Corp., USA),
Pentadorm (Ruef, Austria), Somnesin (British
Drug Houses, UK), Util (Leo, Denmark).
Ethchlorvynol [113-18-8], 3-
(β-chlorovinyl)-3-hydroxy-pent-1-ine,
C7 H9 OCl, M r 144.61, bp 173 – 174 ◦ C, is a liq-
uid with a pungent, aromatic smell. It darkens
slowly when exposed to air and light. The com-
pound is immiscible with water but miscible
with most organic solvents.
It is prepared by reacting 1-chloropent-1-en-3-
one with acetylene [42]. This preparation is a
useful short-term hypnotic, but abuse can lead
to habituation and addiction.
Trade Names. Roeridorm (Roerig, FRG),
Arvynol (Pfizer, USA), Placidyl, Serenesil (Ab-
bott, USA).
Paraldehyde [123-63-7], paracetaldehyde,
2,4,6-trimethyl-1,3,5-trioxane, C6 H12 O3 , M r
132.16, bp 124 ◦ C, is a colorless liquid with an
aromatic smell and unpleasant taste. One part
of paraldehyde is soluble in eight parts of water
at 25 ◦ C. For other properties and synthesis, see
→Acetaldehyde, Chap. 8.1. This substance has
an unpleasant taste and imparts a bad odor to
the breath; addiction is possible. Paraldehyde is
only used in clinics and is no longer prescribed.
Trade Names. Paraldehyde Thilo (Thilo,
FRG; Fellows, USA), Paral (O. Neal,
Jones & Feldmann, USA), Dannetène (Bottu,
France).
Chlorethate [5634-37-7], bis(2,2,2-tri-
chloroethyl)carbonate, C5 H4 Cl6 O3 , M r 324.82
is prepared by the reaction of phosgene with
2,2,2-trichloroethanol in the presence of a base
[43], [44]. The insolubility of this substance
prevents gastric irritation, which is produced
by the hypnotically active parent compound,
2,2,2-trichloroethanol [44].
Trade Name. Clorets (Smith, Kline & French,
USA).
Chloral hydrate [302-17-0], 2,2,2-
trichloro-1,1-ethanediol, C2 H3 Cl3 O2 ,
M r 165.42, mp 57 ◦ C, bp 98 ◦ C (decomp. into
chloral and water), is a crystalline substance
with a pungent odor and taste. It is readily
soluble in water, ethanol, diethyl ether, and
chloroform. For production see →Chloroacet-
aldehyde, Chap. 3.3.
Doses of 0.5 – 0.75 g rapidly produce deep
sleep, which lasts 4 – 8 h, without morning hang-
over. Standard doses of 0.5 g do not produce
REM sleep, but doses of more than 0.8 g gen-
erally have a negative influence. Adverse ef-
fects are irritation of the mucous membranes and
hepato- and nephrotoxicity. Prolonged adminis-
tration can cause habituation, and physical and
psychological addiction.
Trade Names. Ansopal (Ferrosan, Denmark),
Aquachloral (Webcon, USA), Chloradorm
(Knoll Lab., Australia), Chloraldurat (Pohl-
Boskamp, FRG), Cohidrate (Coastal, USA),
Escre (SS Pharmaceutical, Japan), H. S. Need
(Hanlon, USA), Lanchloral (Lancet, Aus-
tralia), Nervifene (Interdelta, Switzerland),
Novochlorhydrate (Novopharm, Canada),
Oradrate (Coast, USA), Somnos (Merck,
Sharp & Dohme, USA), Suppojuvent Sedante
(Juventus, Spain).
Chloral Derivatives.
Chloral betaine [2218-68-0], C7 H14 Cl3 NO4 ,
M r 282.57, mp 122.5 – 124.5 ◦ C, is prepared
from betaine hydrate and chloral hydrate [45].
This adduct has the same hypnotic properties
and contraindications as chloral hydrate, but it
does not have the unpleasant physical properties.

Trade Names. Beta-Chlor (Mead Johnson,
USA) Somilan (British Drug Houses, UK),
Somnalchlor (Schiaparelli, Italy).
Triclofos [306-52-5], 2,2,2-trichloro-
ethyl dihydrogen phosphate, C2 H4 Cl3 O4 P,
M r 229.39.
The sodium salt, triclofos sodium [7246-20-0],
C2 H3 Cl3 NaO4 P, is soluble in water. This prepa-
ration has the same physiological properties as
chloral hydrate, but it does not have the unpleas-
ant physical characteristics. The compound is
metabolized in the body mainly to the hypnot-
ically active 2,2,2-trichloroethanol. A dose of
1.5 g is equivalent to 1 g of chloral hydrate.
Trade Names. Triclos (Lakeside, USA), Triclo-
syl (Glaxo, UK), Tricloran (C.T.S. Israel).
Petrichloral [78-12-6], pentaerythritchloral,
C13 H16 Cl12 O8 , M r 725.76, mp 52 – 54 ◦ C [46].
Chloralodol [3563-58-4], chlorhexadol,
2-methyl-(2 ,2 ,2 -trichloro-1 -hydroxy-4-
ethoxy)-pentan-2-ol, C8 H15 Cl3 O3 , M r 265.58,
mp 102 – 124 ◦ C [47].
Trade Names. Lora (Wallace, USA), Mechloral,
Mecloral (Dumex, Denmark).
Chloralose [15879-93-3], α-O-(2,2,2-tri-
chloroethylidene)-1,2-α-d-glucofuranose-(R),
C8 H11 Cl3 O6 , M r 309.54, mp 187 ◦ C, [α]22
D + 19
(5 % solution in 98 % ethanol) [48].
Hypnotics 7
Trade Names. Chloralosan (Kuhlmann, France),
Somio (Sidel, France).
Carbocloral [541-79-7], N-(2,2,2-
trichloro-1-hydroxyethyl)- ethyl carbamate,
C5 H8 Cl3 NO3 , M r 236.49, mp 103 ◦ C (de-
comp.) [49].
Trade Name. Prodorm (Parke-Davis, USA).
4. Urethanes, Acyclic Urea
Derivatives, and Amides
These compounds are mild hypnotic agents.
They are generally used to promote the onset
of sleep or as daytime sedatives, often in com-
bination with weak analgesics. They are well-
absorbed and, depending on the structure, have
few side effects. Dependence on these com-
pounds occurs less frequently than on barbitu-
rates and benzodiazepines.
Urethane [51-79-6], ethyl carbamate, a mild
hypnotic agent, is no longer prescribed because
its cytostatic properties can cause agranulocyto-
sis on prolonged administration.
Ethinamate is the most important carbamate.
Some other derivatives of carbamic acid are
used as central muscle relaxants (→Skeletal
Muscle Relaxants) and as psychotropic drugs
(→Psychopharmacological Agents).
Ureides are still widely prescribed as mild
sedatives and hypnotic agents. They are used
alone and in combinations. However, bromine-
containing derivatives should be given with great
care because of the danger of bromism.
Most amides, especially those of simple
branched fatty acids, have only a mild sedative
action, but some are frequently used to promote
sleep.
Ethinamate [126-52-3], 1-ethynylcyclo-
hexylcarbamate, C9 H13 NO2 , M r 167.21,
mp 96 – 98 ◦ C, solubility in water 0.25 %, in
ethanol 35 %, and
in hexane 2 %, is produced by treating 1-
ethynylcyclohexan-1-ol with (1) carbamic acid
chloride in the presence of sodium amide [50]
or (2) with phosgene and ammonia [51].
8 Hypnotics
Trade Names. Valamin (Schering, FRG),
Valmid, Valmidate (Lilly, USA).
Hexapropymate [358-52-1], 1-(2-
propynyl)cyclohexanol carbamate, C10 H15 NO2 ,
M r 181.23, mp 99 ◦ C, is synthesized as de-
scribed for ethinamate [52].
Trade Names. Merinax (Labaz, France), Mekos
(Helsingborg, Sweden).
Carfimate [3567-38-2], 1-phenyl-2-
propynyl carbamate [53], C10 H9 NO2 ,
M r 175.18, mp 86 – 87 ◦ C, is a hypnosedative.
Trade Names. Equilium (Delagrange, France),
Nirvetil, Nirvotin (Erba, Italy).
Ectylurea [95-04-5], cis-(2-ethylcrotonoyl)
urea, C7 H12 N2 O2 , M r 156.18, mp 191 – 193 ◦ C
(also mp 158 ◦ C), is prepared by treating (2-
bromo-2-ethylbutyryl)urea with sodium hy-
droxide or by reacting carbromal with silver ox-
ide [54]. It is a sedative with a mild hypnotic
action.
Trade Names. Cronil (Farmigea, Italy), Dis-
tesol (Locatelli, Italy), Ectyl (ACO, Sweden),
Levanil (Upjohn, USA), Neuroprocin (Minerva-
Chemie, Netherlands).
Carbromal [77-65-6], 2-bromo-2,2-
diethylacetylurea, C7 H13 BrN2 O2 , M r 237.11,
mp 116 – 119 ◦ C; 1 g of this compound dissolves
in 3000 mL of water, 18 mL of ethanol, 3 mL of
chloroform, and 14 mL of ether. It is soluble
in alkaline solution and in concentrated sulfu-
ric, hydrochloric, and nitric acids. The reaction
of α-bromo-α-ethylbutyric acid bromide with
urea at 50 ◦ C produces carbromal [55]. The ef-
fects are attributed to the unchanged molecule,
even though all the bromine in the organism is
broken down to bromide. The same applies to
acecarbromal and bromisoval.
Trade Names. Adalin (Bayer, FRG), Bromadyl
(Vicario, France), Diacid (Rieswerke, Austria),
Fydalex (Boots, UK), Hyperysin (Hommel,
Switzerland).
Acecarbromal [77-66-7], acetylcarbromal,
N-acetyl-N  -(2-bromo-2-diethylacetyl)urea,
C9 H15 BrN2 O3 , M r 279.14, mp 109 ◦ C, is
slightly soluble in ethanol and ethyl acetate.
It is prepared by the acetylation of carbromal
with acetic anhydride in the presence of zinc
chloride [56]. Abuse can cause habituation and
addiction.
Trade Names. Abasin (Bayer, FRG; Winthrop,
USA), Adityl, Carbased (Mallard, USA).
Bromoisoval [496-67-3], α-bromoisoval-
erylurea, C6 H11 BrN2 O2 , M r 223.08, mp 147 –
149 ◦ C, is soluble in hot water, diethyl ether,
ethanol, and alkali. The compound is pre-
pared by the acylation of urea with α-
bromoisovalerylbromide [57].
Trade Name. Bromuvan (Treibacher Chem.
Werke, Austria), Bromovaleryl, Bromyl (ACO,
Sweden), Dagrabromyl (Dagra, Netherlands),
Dormigene (Pharmacobel, Belgium), Isobromyl
(Clin-Comar-Byla, France), Isoval (Mission,
USA), Melosan (Allied Lab., USA).
Apronalide [528-92-7], 2-allyl-2-isopro-
pylacetylurea, 1-(2-isopropyl-4-pentenoyl)-
urea, C9 H19 N2 O2 , M r 184.23, mp 194 ◦ C [58].
This compound is no longer produced in the
Federal Republic of Germany because of its
severe adverse effects (purpura hemorrhagica).
Trade Names. Dormid (Rohto, Finland), Neu-
rokin (Leerbeck & Holm, Denmark), Sedormid
(Roche, Switzerland).
Pheneturide [90-49-3], ethylphenacemide,
2-phenylbutyrylurea, C11 H14 N2 O2 , M r 206.24,
mp 149 – 150 ◦ C, is prepared by acylating urea
with the α-phenylbutyric chloride [59]. It is not
only used as a hypnotic agent, but also in the
treatment of epilepsy.
 Hypnotics 9

Trade Names. Benuride (Bengue, UK; Vinas,
Spain), Deturid (Polfa, Poland; Sapos, Switzer-
land).
Capuride [5579-13-5], 2-ethyl-3-
methylvalerylurea, C9 H18 N2 O2 , M r 186.25,
mp 172 ◦ C [60].
Trade Name. Pacinox (McNeil, USA).
Valdettamil, Novonal [512-48-1], 2,2-
diethyl-4-pentenamide, C9 H17 NO, M r 155.23,
mp 75 – 76 ◦ C, is soluble in 120 parts of water
and freely soluble in alcohol and ether [61].
Trade Names. Insomnia (ICN, USA), and
as combinations: Arantil (Hoechst, FRG),
Dentigoa (Scheurich, Switzerland).
diacylated cyclic ureas: barbiturates (8) and
thiobarbiturates (9)
N-(β-carbonyl)ethyl derivatives: 2,4-piperi-
dinediones (6)
N-iminoacyl derivatives: quinazolinones (7)
phenyl N-iminoacyl derivatives: benzodiaze-
pines (11)
All groups, except 5, 7, and 11, have a doubly
substituted carbon atom adjacent to the carbonyl
group of the amide structure. These substituents
greatly influence the action and the distribution
of the drug in the body.

Butoctamidesemisuccinate [32838-28-1],
N-(2-ethylhexyl)-3-hydroxybutyramide hy-
drogen succinate, C16 H29 NO5 , M r 315.41,
mp 36.5 ◦ C, is prepared by the ami-
nolysis of ethyl 3-hydroxybutyrate with 2-
ethylhexylamine and the subsequent formation
of the semiester of succinic acid [62].

Butoctamide is related to (1-methyl)heptyl

γ-bromoacetate, which is found in human
cerebrospinal liquid. It increases the brain 5-
hydroxytryptamine level and has an antitumor
effect. It is given in doses of 200 mg (approxi-
mately as effective as 400 mg of bromoacetyl-
urea [63].

Figure 2. Cyclic carboxamide, carboximide, and urea struc-

5. Cyclic Carboxamides,
Carboximides, Cyclic Urea
Derivatives, and Related Compounds
Most of the synthetic sleep-inducing drugs used
today have a carboxylic acid amide group (4 in
Fig. 2) incorporated into a ring structure. The
ring nitrogen atom is linked to many different
structural elements (Fig. 2).
The following groups of compounds belong
to this category:
N-acyl derivatives: cyclic imides, e.g., (5) and
2,6-piperidinediones (10)
tures used in hypnotics.
5.1. Barbituric Acids [64–66]
The first representative of this group of sub-
stances, barbituric acid, was discovered in 1845
[67]. A synthesis of the first barbiturate with
hypnotic properties, 5,5-diethylbarbituric acid
(barbital), was completed in 1882; its effects
were described in 1903. Since then about 60
different barbiturates have been introduced, of
which 20 – 25 are still used in the industrial
10 Hypnotics
Table 1. Duration of action dose, biological half-life, and metabolism of barbiturates

Compound Type ∗ Mean hypnotic Duration of Onset of Plasma Biological Proportion pK value
dose, g hypnotic action, min protein half-life, h metabolized, %
effect, h binding,
% ∗∗
Barbital L 0.25 – 0.5 10 – 24 30 – 60 2 72 – 120 < 5 7.88
Phenobarbital L 0.1 – 0.2 4 – 12 30 – 60 48 – 96 80 7.29
Butobarbital L 0.05 – 0.1 6 – 12 30 – 60 95 8.12
Probarbital L–M 0.12 – 0.25 4 – 12 30 – 60
Amobarbital M–L 0.1 – 0.3 2–8 15 – 30 15 >95
Cyclobarbital M 0.1 – 0.2 2–8 15 – 30 >99 7.80
(in 3 d 42 %)
Aprobarbital M 0.065 – 0.13 2–8 15 – 30 80 8.15
Allobarbital M 0.1 – 0.3 2–8 15 – 30 70 7.92
Vinylbital M 0.15 6–7 15 – 30 35 in 4 d
Heptabarb S–M 0.1 – 0.4 2–4 30 98 7.78
Secbutabarbital S–M 0.1 2–4 30
Pentobarbital S–M 0.1 – 0.2 2–4 30 40 15 99 7.88
(in 1 d 25 %)
Butallyonal S–M 0.2 2–4 30 >90 8.00
Talbutal S 0.12 – 0.15 2–4 30
Cyclopentobarbital S 0.12 – 0.25 2–4 30 8.10
Propallyonal S 0.1 – 0.3 2–4 30 >99 8.05
Secobarbital VS – S 0.1 – 0.2 1–3 15 45 90 7.80
Hexobarbital VS – S 0.25 – 0.4 1–4 15 2 3.5 95 8.30

∗ Duration of action: L = long; M = medium; S = short; VS = very short.

∗∗ Percentage of the barbiturate that is bound to the protein.

countries. Important properties are listed in Ta-
ble 1. As hypnotics, barbituric acids all pro-
duce almost identical pharmacodynamic ef-
fects. However, they show wide differences
in pharmacokinetics and metabolism. Differ-
ences in onset and duration of activity are re-
lated to lipid and water solubility, and rate
of metabolism. Lipophilic barbiturates, e.g.,
the narcotically active, lipid-soluble thiobarbitu-
rates, and the N-monoalkylbarbiturates such as
hexobarbital rapidly disappear in adipose tissue.
As a result of their redistribution in lipophilic
compartments of the body, these compounds
have a very short duration of action and are
rapidly broken down (short-acting narcotics,
→Anesthetics, General, Chap. 3.1.).
From a therapeutic standpoint, barbiturates
used as hypnotic agents are classified according
to their duration of action (see also Table 1):
1) long-acting compounds (barbital, phenobar-
bital),
2) medium-acting compounds (amobarbital, cy-
clobarbital),
3) short-acting compounds (hexobarbital), and
4) compounds whose effects begin after 15 –
20 min (secobarbital).
The duration of action is primarily deter-
mined by the rate of enzymatic degradation in
the liver, which depends on the chemical struc-
ture of the barbiturate. Continual use causes en-
zyme induction in the liver and thus a gradual
increase in the rate of barbiturate degradation.
Increasing doses are required with prolonged
use (barbiturate habituation). In some patients
barbiturates not only have a sedative and hyp-
notic action, but also produce excitement and
euphoria. This sense of well-being may easily
lead to barbiturate addiction, a physical depen-
dence with withdrawal symptoms. Barbiturate
addicts take the hypnotic because of its euphoric
properties and try to compensate for the central
depression by using psychoanaleptic drugs.
Cases of accidental and suicidal barbiturate
poisoning are fairly common. They are charac-
terized by unconsciousness (which can precede
a delirious stage), central respiratory depression
accompanied by oxygen deficiency symptoms,
and in severe poisoning circulatory collapse with
death in 12 h to 4 d. The prognosis for barbitu-
rate poisoning is favorable if treatment is imme-
diate. Essential measures include artificial res-
piration, irrigation of the stomach, circulatory

support, forced osmotic diuresis using manni-
tol, exchange transfusion, and hemodialysis.
Barbiturates produce a condition superfi-
cially resembling natural sleep. However, the
REM phase, which is very important for rest
and recuperation, is shortened. Continued use
of barbiturates leads to a very low REM value,
and sudden withdrawal increases the REM phase
(rebound effect). The adverse effects of barbitu-
rates are well-documented after more than 50
years of experience. This is an important advan-
tage over newly developed hypnotics [68].
The pK values of the hypnotically active bar-
bituric acids, range between 7.78 and 8.30. Dis-
ubstituted barbituric acids are present, to ca.
50 %, in the undissociated form under physio-
logical conditions (pH 7.4). Only this form of the
drug can pass through biological membranes.
The degree of dissociation is strongly influenced
by slight alterations in pH.
Production. The 5,5-disubstituted barbituric
acids are still synthesized by the classical con-
densation of substituted malonic esters (12),
malonic amides, or cyanoacetic esters (13) with
urea (14), thiourea (15), guanidine (16), or di-
cyanodiamide (17), followed by the hydrolysis
of the resulting imino- or cyanobarbiturate [69]
(Fig. 3).
Under acidic or neutral conditions, substi-
tuted malonic acids or malonic chlorides can
also be used for the condensation with urea
or thiourea. It is preferable for the substituents
R1 and R2 to be present in the malonic ester
or cyanoacetic ester molecule; disubstitution of
barbituric acids at the 5-position is only possi-
ble with highly reactive halides, such as allyl
halides.
Barbital [57-44-3], barbitone, 5,5-
diethylbarbituric acid, C8 H12 N2 O3 , M r 184.19,
mp 188 ◦ C, (sodium salt [144-02-5],
C8 H11 N2 NaO3 , Mr 206.18), has a bitter taste.
One part of the acid dissolves in 170 parts of
water at 25 ◦ C or 17 parts of water at 100 ◦ C.
It is freely soluble in acetone, ammonia, and
lye. Barbital is prepared by the condensation
of diethyl malonate with urea in the presence
of sodium ethoxide [70]. Both the acid and its
sodium salt are used as long-acting hypnotics;
abuse can cause habituation or addiction.
Hypnotics 11
Trade Names. Dormileno (Faes, Spain), Hyp-
nox (Püschel, Austria), Veronal (Bayer, FRG;
Merck, FRG), Veroletten (EAS-Labor, Austria).
Sodium salt: Barbinetten (Hormosani, Aus-
tria).
Phenobarbital [50-06-6], phenemalum,
phenobarbitone, 5-ethyl-5-phenylbarbituric
acid, C12 H12 N2 O3 , M r 232.23, mp 174 –
178 ◦ C. The acid crystallizes from water in
the form of flakes. It is odorless, with a slightly
bitter taste; an aqueous solution reddens litmus
paper. One part dissolves in 1100 parts of water
at 20 ◦ C, 40 parts of boiling water, 10 parts of
ethanol, 18 parts of diethyl ether, and 60 parts
of chloroform. It is synthesized by the reaction
of urea with ethyl phenyl malonate in sodium
ethoxide solution [71].
The sodium salt of phenobarbital [57-30-7],
C12 H11 N2 NaO3 , M r 254.23, is a colorless, crys-
talline, hygroscopic, freely water-soluble pow-
der. It is slightly soluble in ethanol, and insolu-
ble in diethyl ether and chloroform. The aqueous
solution is stable for only a short time.
Phenobarbital is a long-acting hypnotic with
anticonvulsant properties. Long-term use in
higher than therapeutic doses can cause phys-
ical addiction. An overdose resulting in blood
levels of 8 – 12 mg/100 mL can cause death.
Trade Names. Agrypnal (Eggochemia, Austria),
Aphenylbarbit (Streuli, Switzerland), Calminal
(Wolfs, Belgium), Epsylone (Powell, Canada),
Eskabarb Span (Smith, Kline & French, USA),
Fenemal (DAK, Denmark), Gardenal (Spe-
cia, Canada), Gardenale (Farmitalia Carlo
Erba, Italy), Hypnolone (Hartz, Canada),
Pepinal, Lepinaletten (Arzneimittelwerk Dres-
den, GDR), Luminal (Bayer, FRG; E. Merck,
FRG; Winthrop, USA), Mediphen (Medic,
Canada), Phen Bar (Saunders, Canada), Phe-
nobarbyl (Synochem, FRG), Seda-Tablinen
(Beiersdorf, FRG), Solfoton (Poythress, USA),
Teolaxin (Paul Maney, Canada); Phenobarbi-
tal diethylamine salt: Gratusminal (Farmasimes,
Spain); Phenobarbital calcium salt: Fenileal
(Turon, Spain), Lumcalcio (Abello, Spain).
Butobarbital [77-28-1], butobarbitone, 5-
butyl-5-ethylbarbituric acid, C10 H16 N2 O3 ,
M r 212.24, mp 124 – 127 ◦ C, forms slightly
bitter crystals; 1 g dissolves in 5 mL ethanol;
preparation is described in [72]. It is a long-
12 Hypnotics
Figure 3. Synthesis of barbiturates.
acting hypnotic; abuse can cause habituation or
addiction.
Trade Names. Butenil, Butynoct (Interpharm,
Austria), Etoval, Longanoct, Sonabarb (Protea,
Australia), Soneryl (Specia, Italy; May & Baker,
USA).
Probarbital [76-76-6], 5-ethyl-5-isopro-
pylbarbituric acid, C9 H14 N2 O3 , M r 198.22,
mp 197 – 198 ◦ C, is slightly soluble in cold
water, but more soluble in hot water and eth-
anol. Sodium salt [143-82-8], C9 H13 N2 NaO3 ,
M r 220.20, is a hygroscopic powder that is solu-
ble in water. The calcium trihydrate [545-74-4],
C18 H26 CaN4 O6 · 3 H2 O, M r 488.55, forms
slightly bitter crystals; 1 g dissolves in 40 g wa-
ter; preparation is described in [73]. Probarbital
is a long- to medium-acting hypnotic; abuse can
cause habituation or addiction.
Trade Names. Ipral Sodium, Ipral Calcium
(Squibb, USA).
Amobarbital [57-43-2], 5-ethyl-5-isopen-
tylbarbituric acid, C11 H18 N2 O3 , M r 226.28, mp
156 – 158 ◦ C, forms slightly bitter crystals; 1 g
dissolves in 1300 mL water and in 5 mL ethanol.
The sodium salt, C11 H17 N2 NaO3 [64-43-7],
M r 248.25, is hygroscopic and readily soluble
in water; preparation is described in [74], and
metabolism in [75].
Amobarbitol is a widely used medium- to
long-acting hypnotic. It resembles secobarbi-
tal and pentobarbital in action with a some-
what longer duration. A blood level of 3 –
6 mg/100 mL of blood is fatal; abuse can cause
habituation or addiction.
Trade Names. Amsal (Adams, Australia), Amy-
cal (AFI, Sweden), Amydorm, Amytal (Lilly,
USA), Isobec (Pharbec, Canada), Isonal (ICN,
Canada), Neur-Amyl (Fawns & McAllen, Aus-
tralia), Placidel (Miquel, Spain). Stadadorm
(Stada, FRG).
Amobarbital sodium: Altinal, Amsal
(Adams, Australia), Amsebarb (Paul Maney,
Canada), Amylbarb Sodium (Protea, Aus-
tralia), Amylobeta (Bramble, Australia), Amy-
tal Sodium (Lilly, USA), Barbamyl (Teva,
Israel), Dorminal (Boots, UK), Neur-Amyl
(Fawns & McAllen, Australia), Novamobarb
(Novopharm, Canada).
Cyclobarbital [52-31-3], cyclobarbitone,
5-(1-cyclohexen-1-yl)-5-ethylbarbituric
acid,C12 H16 N2 O3 , M r 236,26, mp 171 –
174 ◦ C, forms bitter crystals and is moderately
soluble in water. The calcium salt [143-76-0],
C24 H30 CaN4 O6 , M r 510.59, is prepared by heat-

ing ethyl(1-cyclohexen-1-yl) cyanoacetate with
guanidine sulfate in sodium ethoxide, followed
by hydrolysis with dilute sulfuric acid [76]. One
part dissolves in 70 parts of water. This com-
pound is a medium-acting hypnotic; abuse can
cause habituation or addiction.
Trade Names. Fabadorm (Bayer, FRG), Phan-
odorm (Bayer, FRG; Merck, FRG; Winthrop,
USA), Phanotal (Kwizda, Austria).
Aprobarbital [77-02-1], 5-allyl-5-isopro-
pylbarbituric acid, C10 H14 N2 O3 , M r 210.23,
mp 140 – 141.5 ◦ C, forms slightly bitter crystals,
and is insoluble in water, but soluble in ethanol.
The sodium salt [125-88-2], C10 H13 N2 NaO3 ,
M r 232.21, a hygroscopic, bitter powder, is sol-
uble in water. Preparation is described in [77].
This compound is a medium-acting hypnotic;
abuse can cause habituation or addiction.
Trade Names. Alurate (Hoffmann-La Roche,
Switzerland), Numal (United Chemicals, USA).
Allobarbital [52-43-7], allobarbitone,
5,5-diallylbarbituric acid, C10 H12 N2 O3 ,
M r 208.21, mp 171 – 173 ◦ C, is slightly bitter.
One part dissolves in 300 parts of cold water
and in 50 parts of boiling water. It is prepared
by heating diallyl malonate with urea in sodium
ethoxide. It can also be synthesized by the re-
action of barbituric acid with allyl bromide in
the presence of sodium acetate or copper sulfate
[78], [79]. It is a mediumacting hypnotic; abuse
can lead to habituation or addiction.
Trade Names. Diadol (Durst, USA), Dial (Ciba-
Geigy, Switzerland); it is contained in Spasmo-
Cibalgine (Ciba, Switzerland).
Vinylbital [2430-49-1], vinylbarbi-
tal, 5-(1-methylbutyl)-5-vinylbarbituric
acid,C11 H16 N2 O3 , M r 224.27, mp 90 –
91.5 ◦ C, is prepared by the reaction of guanidine
with the appropriate malonic ester and saponifi-
cation. It can also be synthesized by the addition
of acetylene to 5-(1-methylbutyl)barbituric acid
in the presence of zinc stearate [80].
Hypnotics 13
Vinylbital is a medium-acting hypnotic;
abuse can cause addiction or habituation.
Trade Names. Bykonox (Byk Belga, Belgium),
Optanox, Suppoptanox (Valpan, France), Speda
(Byk-Gulden, FRG).
Heptabarb [509-86-4], heptabarbital, hept-
abarbitone, 5-ethyl-5-(1-cycloheptenyl)barbitu-
ric acid, C13 H18 N2 O3 , M r 250.29, mp 174 ◦ C;
one part dissolves in 13 parts of methanol,
20 parts of ethanol, 13 parts of acetone, 40
parts of diethyl ether, and 5000 parts of wa-
ter. It is synthesized by the condensation of
ethyl(cyclohepten-1-yl)-cyanoacetic ester with
urea [81].
This compound is a short- to medium-acting
hypnotic; abuse can cause habituation or addic-
tion.
Trade Names. Medomin, Medapan (Ciba-
Geigy, Switzerland).
Secbutabarbital [125-40-6], secbutobarbi-
tal, butabarbital, secbutobarbitone, 5-ethyl-5-(2-
butyl)barbituric acid, C10 H16 N2 O3 , M r 212.24,
mp 165 – 168 ◦ C; the sodium salt [143-81-7],
C10 H15 N2 NaO3 , M r 234.23, is a bitter powder;
1 g dissolves in 2 mL of water, and the pH of a
1 % aqueous solution is 9.0 – 10.2. Preparation is
described in [82]. Secbutabarbital is a medium-
to long-acting hypnotic. It is widely used in the
United States as a sedative. Blood levels of 3.5 –
6 mg/100 mL are fatal; abuse can cause habitu-
ation or addiction.
Trade Names. BBS (Reid-Provident, USA),
Butabarbital Sodium (Parke-Davis, USA),
Butabarpal (Philadelphia-Labs., USA), Bu-
tamide (Zenith, USA), Buticaps (Carter-
Wallace, USA), Butisol (McNeil, USA), Day-
Barb (Anca, Canada), Merisyl (Meriot, Canada),
Neo-Barb (Neo, Canada), Sarisol (Halsey Drug,
USA).
14 Hypnotics

Pentobarbital [76-74-4], pentobarbi- Trade Names. Pernocton (Riedel De Haen,

tone, 5-ethyl-5-(1-methylbutyl)-barbituric FRG) Pernoston, Sonbutal (Tutag Pharm, USA).
acid,C11 H18 N2 O3 , M r 226.28, mp 130 ◦ C;
the sodium salt [57-33-0], C11 H17 N2 NaO3 , Vinbarbital [125-42-8], 5-ethyl-5-(1-
M r 248.26, is unstable in aqueous solu- methyl-1-butenyl)barbituric acid, C11 H16 N2 O3 ,
tion. It is prepared by boiling ethyl-1- M r 224.3, mp 161 – 163 ◦ C, forms slightly bitter
methylbutylcyanoacetic ester with guanidine crystals. Preparation is described in [85]; abuse
in sodium ethoxide solution and saponifying the can cause addiction.
product with dilute sulfuric acid [83]. This com-
pound is a short- to medium-acting hypnotic
and is more often used as a hypnotic than as
a sedative. A dose of more than 3 g can cause
death; blood levels of 1 – 2.5 mg/100 mL are
fatal; abuse can lead to habituation or addiction. Trade Names. Delvinal (Merck, Sharp & Dohme,
Trade Names. Neodorm (Minden, FRG). USA), Diminal (Astra, Sweden).
Pentobarbital sodium: Butylone (Hartz,
Canada), Dorminal (Alfasan, Netherlands), Talbutal [115-44-6], 5-allyl-5-sec-butylbar-
Embutal (Abbott, USA), Hypnol (Stickley, bituric acid, C11 H16 N2 O3 , M r 224.25, mp 108 –
Canada), Isoamytal, Isobarb (U.S. Standard, 110 ◦ C, forms slightly bitter crystals. Prepara-
USA), Iturate, Lethobarb (Loveridge, UK), tion is described in [86]. This compound is a
Napental (Beecham, UK), Narcoren (Veteri- short-acting hypnotic; abuse can cause habitua-
naria, Switzerland), Nembutal (Abbott, USA), tion or addiction.
Nova-Rectal (Nova, Canada), Novopentobarb Trade Names. Lotusate (Winthrop, USA), Pro-
(Novopharm, Canada), Pembule (Novocol, fundol (Promonta, FRG).
USA), Penbon (Adams, Australia), Pental (Van-
Pelt & Brown, USA; Saunders, Canada), Pen-
Cyclopentobarbital [76-68-6], 5-allyl-
tone (Faulding, Australia), Pentosol (Chroma-
5-(2-cyclopenten-1-yl(barbituric acid,
lloy, USA), Prodormol (Teva, Israel), Sedanox
C12 H14 N2 O3 , M r 234.25, mp 139 – 140 ◦ C,
(Troncin, France), Sombutol (Farmos Group,
forms slightly bitter crystals and is moderately
Finland), Somnopentyl (Pitman-Moore, USA),
soluble in hot water. This acid is synthesized by
Vetanarcol (Veterinaria, Switzerland).
the condensation of urea with the appropriately
Pentobarbital calcium: Insom Rapido (An-
disubstituted malonic ester [87]. It is a short-
dreu, Spain), Repocal (Desitin, FRG).
acting hypnotic; abuse can lead to habituation
and addiction.
Butallyonal [1142-70-7], 5-(2-bromoallyl)-
5-sec-butylbarbituric acid, C11 H15 BrN2 O3 , M r
303.16, mp 130 – 133 ◦ C, slightly bitter crys-
tals, is almost insoluble in water, but solu-
ble in ethanol. The sodium salt [3486-86-0],
C11 H14 BrN2 NaO3 , M r 325.15, forms a bitter,
crystalline powder and is soluble in water; the Trade Names. Cyclopal (Siegfried, Switzer-
pH of a 10 % aqueous solution is ca. 9.5. Prepa- land), Cyclopental (Pharmacia, Sweden).
ration is described in [84]. This compound is
a short- to medium-acting hypnotic; abuse can Propallyonal [545-93-7], 5-(2-bromoallyl)-
cause habituation or addiction. 5-isopropylbarbituric acid, C10 H13 BrN2 O3 , M r
289.13, mp 177 – 179 ◦ C, forms slightly bitter
crystals and is moderately soluble in water.
Preparation is described in [88]. This drug is a
short-acting hypnotic; abuse can cause habitua-
tion or addiction.

Trade Name. Noctal (Union Chimique, Bel-
gium).
Secobarbital, 5-allyl-5-(2-pentyl)barbituric
acid, C12 H18 N2 O3 , M r 238.28, mp 100 ◦ C,
is slightly bitter. The sodium salt [309-43-3],
C12 H17 N2 NaO3 , M r 260.27, is a hygroscopic,
bitter powder; preparation is described in [89].
It resembles pentobarbital but has a shorter du-
ration of action and is especially suitable for
injection. Blood levels of 1 – 2.5 mg/100 mL are
fatal; abuse can cause habituation or addiction.
Trade Names. Seconal (Lilly, USA).
Secobarbital sodium (secobarbital solubile):
Dormona (Wiedenmann, Switzerland), Im-
menoctal (Houde-ISH, France), Novosecobarb
(Novopharm, Canada), Proquinal (Protea, Aus-
tralia), S.C.B.Tal (Novo, Canada), Sebar (Van-
gard, USA), Secaps (Saunders, Canada), Sec-
ocaps (M.T.C. Canada), Secogen (Paul Maney,
Canada), Seconal, Seotal (Lilly, USA), Quinbar
(Adams, Australia).
Hexobarbital [56-29-1], 1,5-dimethyl-5-
(cyclohexen-1-yl)barbituric acid, C12 H16 N2 O3 ,
M r 236.26, mp 146 ◦ C. For preparation and prop-
erties, see →Anesthetics, General. It is a short-
acting hypnotic, which rapidly induces sleep.
Hexobarbital is frequently used as a short-acting
narcotic in the form of its sodium salt [50-09-9],
Evipan sodium, C12 H15 N2 NaO3 , M r 258.25.
The short duration of action and the short bi-
ological half-life are related to the metabolism
by cleavage of the N-methyl group and oxida-
tion of the cyclohexenyl group at the α position.
Evipan sodium is used as a narcotic.
Trade Names. Citopan (Nyco, Norway), Cy-
clopan (Interpharm, Austria), Sombucaps, Som-
bulex (Riker, USA), Tobinal (Siegfried, Switzer-
land), Toleran (Kwizda, Austria).
Hypnotics 15
5.2. Piperidinediones
As already mentioned, barbituric acids are cy-
clic urea derivatives. Another large class of hyp-
notics is derived from the lactams of 5-ami-
noglutaric acid; they are divided into two groups:
2,4-piperidinediones and 2,6-piperidinediones
(glutaric imides). They all have a disubstituted
carbon atom located α to the carbonyl group of
the lactam residue.
The hypnotic action of the piperidinediones
qualitatively resembles that of the barbiturates,
but they are less active and are used in larger
doses. Adverse effects and toxic reactions are
less frequent than with barbiturates. The piperi-
dinediones are rapidly broken down in the
body; the symptoms of poisoning resemble those
caused by barbiturates.
Taglutimide (see page 16) and thalido-
mide [50-35-1] have a monosubstituted car-
bon atom at position 3. The notorious thalido-
mide (3-phthalimido-2,6-dioxopiperidine, Con-
tergan) was removed from the market because it
caused severe adverse effects (prolonged neuri-
tis and malformation of the fetus).
Pyrithyldione [77-04-3], 3,3-diethyl-2,4-
dioxo-1,2,3,4-tetrahydropyridine, C9 H13 NO2 ,
M r 167.20 is structurally related to the piperi-
dinediones and exists in three crystalline modi-
fications: (1) mp 92 – 93 ◦ C, (2) mp 97 – 98 ◦ C,
(3) mp 81 – 86 ◦ C.
Trade Names. Persedon (Hoffmann-La Roche,
Switzerland), Benedorm (Arzneimittelwerke,
GDR).
This drug is no longer used because of its pro-
found adverse effects (agranulocytosis). It has
been replaced by the homologous methyprylon
which does not cause blood dyscrasia.
Methyprylon [125-64-4], 3,3,diethyl-5-
methyl-2,4-piperidinedione, C10 H17 NO2 , M r
183.26, mp 74 – 77 ◦ C, is soluble in water, etha-
nol, and chloroform. It is prepared by formylat-
16 Hypnotics
ing 3,3-diethylpiperidine-2,4-dione (dihypry-
lone) and reducing the resulting 3,3-diethyl-
5-hydroxymethylpiperidine-2,4-dione with hy-
drogen and Raney nickel [90]. The drug acts
rapidly but longer than pyrithyldione; it is less
toxic than phenobarbital. Prolonged use causes
habituation and addiction.
Trade Names. Noludar, Nolurate (Hoffmann-La
Roche, Switzerland).
Glutethimide [77-21-4], 3-ethyl-3-
phenylpiperidine-2,6-dione, C13 H15 NO2 ,
M r 217.26, mp 84 ◦ C, is soluble in ethyl ac-
etate, acetone, diethyl ether, chloroform, eth-
anol, and methanol, and almost insoluble in
water. It forms a hydrate and a water-soluble
hydrochloride. Glutethimide is prepared by the
addition of 2-phenylbutyronitrile to acrylic ester
to give 2-ethyl-2-phenylpentane-1,5-dicarbox-
ylic acid mononitrile, followed by cyclization
with sulfuric acid and acetic acid [91]. This
compound is effective in inducing sleep that
lasts through the night and is also suitable as
a daytime sedative. Prolonged administration
causes habituation and addiction.
Trade Names. Doriden (Ciba-Geigy, Switzer-
land; Takeda, Japan), Elrodorm (Dtsch. Hydrier-
werke, FRG), Glimide (Polfa, Poland), Gludorm
(Knoll, FRG), Regenox (Richter, Hungary).
Alonimid [2897-83-8], 1,2,3,4-tetrahydro-
4-oxonaphthalene-1-spiro-3 -piperidine-2 ,6 -
dione, C14 H13 NO3 , M r 243.25, mp 197 –
199 ◦ C,
is prepared by the Michael addition of ethyl
acrylate to phenylacetonitrile, followed by cy-
clization with polyphosphoric acid [92].
Trade Name. Alonimide (Merrell, USA).
Taglutimide [14166-26-8], biglumide,
2-(bicyclo [2,2,1] heptane-2-endo-3-endo-
dicarboximido)-glutarimide, C14 H16 N2 O4 ,
M r 276.29, mp 237 ◦ C, is unstable in alkaline
solution, and is slightly soluble in water and in
organic solvents. It is prepared by the reaction
of endo-bicyclo [2,2,2] -heptane-1,2-dicarbox-
ylic acid anhydride with l(+)-glutamic acid,
followed by cyclization with urea in dimethyl
sulfoxide at 180 ◦ C [93].
Unlike thalidomide, taglutimide does not
produce any teratogenic effects [94]. It is a use-
ful hypnotic because it causes little change in the
electroencephalogram of normal sleep [95].
Trade Name. Synval (Kwizda, Austria).
5.3. Quinazolinones
Quinazolinones are cyclization products of an-
thranilamide. Some have analgesic and an-
tipyretic properties. Only those derivatives in
which the nitrogen atom at position 3 has an 2-
substituted phenyl group are hypnotically active
[96]. The effect is very fast in onset and lasts
3 – 4 h. These compounds are similar in hyp-
notic action to the barbiturates but are reported
not to suppress REM sleep. They inhibit the
biosynthesis of serotonin and catecholamines,
which influence the physiology of sleep [97].
The quinazolinones are a group of hypnotics

with codeinelike, cough-suppressant, analgesic
properties; they are useful in patients who do not
tolerate or respond to barbiturates. Prolonged
use may cause habituation and psychological de-
pendence. No specific antagonists are available
for quinazolinone poisoning.
Methaqualone [72-44-6], 2-methyl-3-
o-tolyl-4-(3 H)quinazolinone, C16 H14 N2 O,
M r 250.30, mp 120 ◦ C (also 114 – 116 ◦ C), is
soluble in ethanol, diethyl ether, and chloro-
form, but almost insoluble in water. The hydro-
chloride [340-56-7], C16 H15 ClN2 O, M r 286.85,
mp 255 – 265 ◦ C, is prepared by the condensa-
tion of N-acetylanthranilic acid with o-toluidine
in the presence of phosphoryl chloride [98], [76].
Methaquolone is present in numerous combina-
tions.
Prolonged administration of methaqualone
has caused paresthesia of the extremities; daily
doses of 0.3 g should not be taken for more than
four weeks (in exceptional cases, eight weeks).
Trade Names. Bon-Sonnil (Diselen, Spain),
Dormogen (Leciva, Czechoslovakia), Dormu-
til (Chem.-Pharm. Werk Oranienburg, GDR),
Mequin (Lemmon, USA), Mozambin (Gerot,
Austria), Pro Dorm (Schürholz, FRG), Som-
notropon (Tropon, FRG), Torinal (Medicamen-
tos y Productos Quimic, Spain), Tuazolona
(Nessa, Spain).
Methaqualone hydrochloride: Cateudyl
(Covor, Belgium), Dormir (Langley, Aus-
tralia), Hyptor (Toraude, Canada), Melsed,
Melsedin (Boots, UK), Mequelon (Merck-
Frosst, Canada), Methasedil (Cooper, Switzer-
land), Nobadorm (Streuli, Switzerland), Nor-
morest (Merck, FRG), Noxybel (Probel, Bel-
gium), Optimil (Wallace, USA), Optinoxan
(Robisch, FRG), Parest (Parke-Davis, USA),
Parmilene (Chiesi, Italy), Pexaqualone (Ther-
apex, Canada), Revonal (Merck, FRG), Ri-
porest (Farmitalia, Italy), Sedalone (Phar-
bec, Canada), Somberol (Boots, UK), Som-
nafac (Smith, Miller & Patch, USA), Somnium
(Fargal, Italy), Sovelin, Soverin (Weifa, Nor-
Hypnotics 17
way), Sovinal (ND & K, Denmark), Toraflon
(Toraude, Canada), Tualone (ICN, Canada), Tu-
azol (Strasenburgh, USA).
Etaqualone [7432-25-9], 3-(2-ethylphenyl)-
2-methyl-4(3 H)-quinazolinone, C17 H16 N2 O,
Mr 264.31, mp 81 ◦ C. Hydrochloride
[97979-65-2], C17 H17 ClN2 O, mp 247 ◦ C.
Etaqualone is prepared by the condensation
of N-acetylanthranilic acid with 2-ethylaniline
in the presence of phosphoryl chloride [99].
Trade Name. Aolan (Beiersdorf, FRG).
Nitromethaqualone [340-52-3], 3-(2 -

methoxy-4 -nitrophenyl)-2-methyl-4(3 H-)
quinazolinone, C16 H13 N3 O4 , M r 311.29,
mp 135 – 137 ◦ C, and the hydrochloride
[3946-23-4], mp 246 – 248 ◦ C, are prepared as
the preceding compounds [100]. For action, see
[101].
Trade Name. Parnox (Santos, Spain).
5.4. Benzodiazepines
Many benzodiazepine derivatives used as
psychotropic agents (→Psychopharmacological
Agents) promote sleep by acting as mental re-
laxants and also by having a hypnotic action of
their own [102], [103]. The diverse action of this
group of drugs clearly illustrates the difficulty
of distinguishing hypnotics from sedatives and
tranquilizers if the sleep-inducing effect is the
only feature of interest. As with the barbiturates,
the sedative and tranquilizing effects of benzodi-
azepines induce “better sleep.” After an initially
steep increase in the use of benzodiazepines as
hypnotics, a drop in consumption is slowly be-
coming apparent in Europe and the United States
[104], [105]. Even though benzodiazepines are
18 Hypnotics
Table 2. Duration of action, biological half-life, and metabolism of benzodiazepines

Compound Onset of Half-life, h Main metabolic route ∗ Main active metabolite and Adult hypnotic
action, min half-life, h dose, mg
Nitrazepam 16 – 30 20 – 50 nitroreduction, acetylation none 5 – 10
Flunitrazepam 9 – 20 10 – 25 demethylation, nitroreduction dealkylflunitrazepam, 20 – 30 1–2
Flurazepam 15 – 45 2.3 oxidation desalkylfluorazepam, 50 – 100 15 – 30
Oxazepam 45 – 90 5 – 15 conjugation none 15 – 30
Lorazepam 15 – 45 10 – 15 (20) conjugation none 2–4
Lormetazepam 15 – 45 9 – 15 conjugation none 0.5 – 2
Temazepam 45 – 60 9.5 – 12 (17) conjugation none 15 – 30
Doxefazepam 15 – 40 7–8 dealkylation, conjugation none 10 – 20
Cinolazepam 15 – 45 3.8 conjugation none 15 – 30
Quazepam 60 39 oxidation, demethylation,
conjugation
2-oxoquazepam, 39; demethyl- 7.5 – 15
2-oxoquazepam, 73
Haloxazolam 45 – 60 2–4 cleavage of oxazoline residue 7-bromobenzodiazepin-2-one 5 – 10
(18, R = Br, R1 = H, R2 = F)
Triazolam 18 – 30 1.5 – 3 oxidation none 0.25 – 0.5
Estazolam 15 – 40 8 – 31 oxidation none 2–4
Midazolam 16 – 30 1–3 oxidation, hydroxylation,
conjugation none 10 – 30
Loprazolam 15 – 30 10 – 16 oxidation “loprazolam-N-oxide”, 5 – 8 1–2
Brotizolam 10 – 20 4.5 – 9 oxidation “hydroxymethyl brotizolam”,
4 – 8; 3-hydroxy-brotizolam 0.25 – 1

∗ Plasma protein binding rates are high (usually 90 – 95 %). In most cases, conjugation indicates glucuronization, and oxidation
denotes the introduction of a hydroxyl group at the 3-position or in the methyl group of a heterocycle condensed with the
benzodiazepine moiety.

not ideal sleep-inducing agents, new products
are constantly being marketed. Properties of the
most important benzodiazepines are listed in Ta-
ble 2.
Mode of Action. The benzodiazepines have
the same hypnotic action as the barbiturates,
with differences in selectivity, metabolism, and
adverse effects. Large doses do not possess the
narcotic properties exhibited by most barbitu-
rates. Benzodiazepines do not solely have a
sleep-inducing action; they exert a combination
of effects: sedation, relief of anxiety (produc-
tion of a better state of mind), muscle relax-
ation, and anticonvulsant action. These effects
shorten both the intermittent waking times and
deep sleep (reduction of δ sleep). In addition,
a hangover always occurs on the next day. The
effects of barbiturates and benzodiazepines are
compared in Table 3. From a pharmacological
standpoint, the hypnotically active benzodiaze-
pines have a depressant action on the electrical
discharges in the areas controlled by the limbic
system and block the function of the reticular
formation in the brain stem. They are similar to
the barbiturates in this respect; however, quanti-
tative differences could be of importance. Stud-
ies point to an interaction between benzodiaze-
pines and their receptors. Thus, they activate the
postsynaptic receptors for γ-aminobutyric acid
(GABA, an inhibitory neurotransmitter) and in-
crease the frequency of the opening of the chlo-
ride ion canals [106].
In other words, the benzodiazepines act
in the central nervous system by interfering
with GABA-mediated neurotransmission [107],
[108]. Traces of benzodiazepine derivatives
(e.g., diazepam and lorazepam) have been found
in wheat and potato and are identical to the syn-
thetic substances. Benzodiazepines also occur
in untreated rats, as well as in human serum and
brain. However, the concentration is very low
(several nanograms per gram of potato) and a
direct pharmacological effect on the central ner-
vous system has not been demonstrated [108],
[109].
Metabolism. All benzodiazepines are me-
tabolized in the body [110], but in some cases
the retention times are very long. In addition,
the pharmacologically active metabolites of the
benzodiazepines have long residence times in
 Hypnotics 19

the body, and the rate of metabolic degradation ataxia, disorientation, and nightmares. Regu-
decreases considerably with age. A plasma half- lar administration of most benzodiazepines in-
life of 1.5 – 40 h, and in extreme cases 47 – 100 h, creases the dose necessary to evoke the same
is observed [111]. The main metabolic transfor- response.
mations of benzodiazepines are (Fig. 4): Indications for the use of benzodiazepine
hypnotics are given in Table 4.
1) Hydroxylation at position 3 with formation
of more active products, which are, however, Table 3. Comparison of the effects and properties of barbiturates
and benzodiazepines
usually rapidly eliminated after glucuroniza-
tion Effect or property Barbiturates Benzodiaze-
pines
2) Demethylation at N-1 with formation of
longer acting demethylbenzodiazepines
3) Ring opening at C-3, which again frequently Narcotic yes no
Sedative yes yes
produces active metabolites that can be recy- Muscle relaxation almost none yes
clized in the organism Anticonvulsant yes yes
4) Heterocyclic moiety of benzodiazepines Shortening of REM sleep and yes yes
increase in REM phase on withdrawal
containing fused heterocyclic structures, (rebound effect)
(e.g., triazolam) rapidly broken down to form Retrograde amnesia yes almost none
inactive metabolites Accumulation yes yes
Hangover yes (clearly yes (also
sedated positive
Therapy. The expectation that therapeutic waking influence on
problems could be solved with benzodiazepines phase) nervousness,
apprehen-
of lower half-life has proved unfounded. Al- sion,
though the faster onset of effect, swift relief of tension,
listlessness
anxiety, and rapid elimination of these drugs pre- in the
vent a hangover, rebound insomnia occurs as a waking
phase)
severe adverse effect to produce delirium and
Tolerance yes yes
exogenous psychosis, (a type of premature with- Abuse yes yes
drawal symptom) during the first night of use. Enzyme induction yes no
A temporary shortening of REM sleep follows Therapeutic range narrow wide
Toxicity relatively relatively
with subsequent paradoxical alertness, fear, ag- high low
itation, and anterograde amnesia during the sec-
ond half of the night [104]. In addition, abuse of
the rapid onset of activity benzodiazepines (e.g.,
lorazepam and flunitrazepam) has been reported, Synthesis. The hypnotically active benzo-
especially among young people. diazepines can be classified as benzo-1,4-
From a therapeutic standpoint, the benzo- diazepinones (18), or benzodiazepines con-
diazepine hypnotics are classified according to taining fused heterocyclic moieties, e.g., triazo-
their duration of action: lobenzodiazepines (21a–21c). The synthesis of
these compounds is shown in Figures 5 and 6,
1) long-acting compounds, e.g., flurazepam, respectively [102].
quazepam, nitrazepam, and flunitrazepam;
2) medium-acting compounds, e.g., Nitrazepam [146-22-5], 1,3-dihydro-7-
temazepam, lorazepam, loprazolam, ox- nitro-5-phenyl-2H-1,4-benzodiazepin-2-one,
azepam, and brotizolam; and (18a),C15 H11 N3 O3 , M r 281.26, mp 224 –
3) short-acting compounds, e.g., triazolam, mi- 226 ◦ C, is prepared by the cyclization of 2-
dazolam, and cinolazepam. (aminoacetamino)-5-nitrobenzophenone or by
Long-acting benzodiazepines should not be the nitration of benzodiazepinone with an un-
used to treat insomnia because of accumulation, substituted phenyl substituent [112], [113].
occurrence of undesired effects before sleep in- Indications: problems in falling asleep, fre-
duction, daytime drowsiness, lowered activity, quent awakening during the night, and early
awakening in the morning. The duration of in-
20 Hypnotics
Figure 4. Metabolism of triazolam.
duced sleep is 7 – 8 h; a strong hangover occurs
next day [114].
Trade Names. Alodorm (Alphapharm, Aus-
tralia), Apodorm (Apoth. Lab., Norway), Arem
(Lennon, Republic of South Africa), Atem-
pol (Norgine, UK), Benzalin (Shionogi, Japan),
Calsmin (Upjohn, USA), Dormo-Puren (Klinge,
FRG), Dumolid (Dumex, Denmark), Eatan N
(Desitin, FRG), Hipsal (Salvat, Spain), Ime-
son (Desitin, FRG), Insomin (Orion, Fin-
land), Ipersed (Sidus, Italy), Lagazepam (La-
gap, Switzerland), Megadon, Mogadan, Mo-
gadon (Roche, Switzerland), Mitidin (Savorma,
Italy), Nelbon (Sankyo, Japan), Neuchlonic
(Taiyo, Japan), Nitepam (USV, USA), Nitra-
dos (Berk, UK), Nitrempax (Lafi, Brazil), No-
vanox (Pfleger, FRG), Numbon (Ikapharm,
Israel), Ormodon (Ormed, South Africa),
Pacisyn, Paxisyn (Synthetic, Finland), Pel-
son (Infale, Spain), Radedorm (Berlin-Chemie,
GDR), Relact (Lemonier, Argentina), Remnos
(DDSA, UK), Somitran (Farmos Group, Fin-
land), Somnased (Duncan Flockhart, UK), Som-
nite (Norgine, UK), Somnolin (Dima, Italy),
Unisomnia (Unigreg, UK).
Flunitrazepam [1622-62-4], 5-(2-
fluorophenyl)- 1, 3- dihydro-1-methyl-7-
nitro-2H-1,4-benzodiazepin-2-one, (18b),
C16 H12 FN3 O3 , M r 313.30, mp 166 – 167 ◦ C
(also 170 – 172 ◦ C). The synthesis of this com-
pound is analogous to that of nitrazepam [113].
Pharmacologically, flunitrazepam is one of the
most active benzodiazepinones; the hypnotic ef-
fects begin with doses as low as 1 – 2 mg [115].
It is used in anesthesiology in solution form to
induce sleep.
Trade Names. Flunipam (A. L., Norway), Hypn-
odorm (Teva, Israel; Alphapharm, Australia),
Hipnosedon, Narcozep, Rohpinol, Rohypnol,
Roipnol (Hoffmann-La Roche, Switzerland).
Flurazepam [17617-23-1], 7-chloro-1-
(2-diethylaminoethyl)-5-(2-fluorophenyl)-
1,3-dihydro-2H-1,4-benzodiazepin-2-one,
(18c),C21 H23 ClFN3 O, M r 387.89, is oily. The
dihydrochloride [1172-18-5], C21 H25 Cl3 FN3 O,
M r 460.83, mp 215.5 – 217.5 ◦ C (decomp.),
forms light yellow crystals. Flurazepam is
synthesized by the alkylation of 18 (R = Cl,

Table 4. Indications for the use of benzodiazepine hypnotics ∗
R1 = H, R2 = F) with diethylaminoethyl chlo-
ride [116]. Another method uses 2-(diethyl-
aminoacetylamino)-5-chloro-2 -fluorobenzo-
phenone as the starting material [117]. Flu-
razepam can also be synthesized by the oxidative
ring extension of 2-aminomethyl-5-chloro-1-(2-
diethylaminoethyl)-3-(2 -fluorophenyl)indole
with chromium trioxide [118].
Hypnotics 21
Figure 5. Synthesis of benzo-1,4-diazepinones (18), and
their 2-thiono-(19) and 3-hydroxy derivatives (20)
Nitrazepam (18a): R = NO2 ; R1 = H; R2 = H
Flunitrazepam (18b): R = NO2 ; R1 = CH3 ; R2 = F
Flurazepam (18c): R = Cl; R1 = CH2 −CH2 N (C2 H5 )2 ;
R2 = F
Nimetazepam (18d): R = NO2 ; R1 = CH3 ; R2 = H
Quazepam (19): R = Cl; R1 = CH2 −CF3 ; R2 = F
Oxazepam (20a): R = Cl; R1 = H; R2 = H
Lorazepam (20b): R = Cl; R1 = H; R2 = Cl
Lormetazepam (20c): R = Cl; R1 = CH3 ; R2 = Cl
Temazepam (20d): R = Cl; R1 = CH3 ; R2 = H
Doxefazepam (20e): R = Cl; R1 = CH2 −CH2 OH; R2 = F
Cinolazepam (20f): R = Cl; R1 = CH2 −CH2 −CN; R2 = F
22 Hypnotics
The full hypnotic response of flurazepam is
evoked even after prolonged use. However, it has
an extremely long plasma half-life (47 – 100 h),
with long-lasting metabolites. It promotes the
onset of sleep, reduces the frequency of interrup-
tions at night, and increases the total duration of
sleep. The effects of the drug begin within 20 –
30 min, and sleep usually lasts for 7 – 8 h. Flu-
razepam can reduce REM sleep, but does not
produce a REM rebound and associated sleep
disturbance on termination of treatment. This
drug strongly suppresses “level 4” of orthodox
sleep (see Chap. 1). It is given in doses of 30 –
60 mg [119].
Trade Names. Noctosom (Ikapharm Kfar-Sava,
Israel), Staurodorm (Dolorgiet, FRG).
Flurazepam hydrochloride: Benzozil (Ky-
owa, Japan), Dalmadorm, Dalmane, Dalmate,
Dalmene, Dormador, Dormodor (Hoffmann-La
Roche, Switzerland), Felison (Sigurta, Italy),
Insumin (Kyorin, Japan), Natam (Unifa, Ar-
gentina), Novoflupam (Novopharm, Canada),
Remdue (Biomedica Foscama, Italy), Somnol
(Horner, Canada), Som-Pam (ICN, Canada),
Valdorm (Valeas, Italy).
Nimetazepam [2011-67-8], 1,3-di-
hydro-1-methyl-7-nitro-5-phenyl-2H-1,4-
benzodiazepin-2-one (1-methylnitrazepam),
(18d), C16 H13 N3 O3 , M r 295.30, mp 156.5 –
157.5 ◦ C, forms light yellow crystals and is pre-
pared analogously to nitrazepam [113], [120]. It
has a long half-life and resembles nitrazepam in
activity, but also has anticonvulsant and muscle
relaxant properties [121].
Trade Name. Erimin (Sumitomo, Japan).
Figure 6. Synthesis of triazolobenzodiazepines (21)
Estazolam (21a): R = H; R1 = H
Triazolam (21b): R = CH3 ; R1 = Cl
Midazolam (21c): R = CH3 ; R1 = F
Oxazepam [604-75-1], 7-chloro-1,3-di-
hydro-3-hydroxy-5-phenyl-2H-1,4-
benzodiazepin-2-one,(20a), C15 H11 ClN2 O2 ,
M r 286.72, mp 205 –206 ◦ C, is prepared by the
oxidation of 18 (R = Cl, R1 = H, R2 = H) with
peracetic acid, followed by the reactions de-
scribed in Figure 5. Oxazepam is primarily a
sedative, but also has sleep-inducing properties
[122].
Trade Names. Adumbran (Thomae, FRG;
Boehringer Ingelheim, FRG), Alepam (Al-
phapharm, Australia), Alopam (A. L., Norway),
Anxiolit (Gerot, Austria; Medichemie, Switzer-
land), Aslapax (Asla, Spain), Benzotran (Pro-
tea, Australia), Durazepam (Durachemie, FRG),
Enidrel (Syncro, Argentina), Expidet (Wyeth,
UK), Isodin (Tosi, Italy), Murelax (Ayerst, Aus-
tralia), Noctazepam (Brenner, Austria), Oxepam
(Wyeth, UK), Oxpam (ICN, Canada), Praxiten
(Wyeth, UK), Propax (Cipan, Portugal), Psico-
pax (Bama-Geve, Spain), Sedokin (Geymonat
Sud, Italy), Senepax, Serax, Serenid (Wyeth,
UK), Serepax (Wyeth, UK; Ferrosan, Den-
mark), Seresta (Wyeth, Netherlands), Sigacalm
(Siegfried, Switzerland), Sobril (KabiVitrum,

Sweden), Uskan (Desitin, FRG), Vaben (Rafa,
Israel), Zapex (Riva, Canada).
Lorazepam [846-49-1], 7-chloro-5-(2-
chlorophenyl)- 1, 3- dihydro- 3- hydroxy- 2H-
1,4-benzodiazepin-2-one, (20b),
C15 H10 Cl2 N2 O2 , M r 321.17, mp 166 – 168 ◦ C,
is prepared from 18 (R = Cl, R1 = H, R2 = Cl),
analogously to oxazepam. Lorazepam is pri-
marily an anti-anxiety drug; it also has sleep-
inducing properties and is used in anesthesiol-
ogy [123].
Trade Names. Ativan (Wyeth, USA), Bonton
(Unipharma, Israel), Control (Sigurta, Italy),
Donix (Llorens, Spain), Emotion (Alpes, Ar-
gentina), Emotival (Armstrong, Argentina), Lar-
pose (Cipla, India), Laubeel (Desitin, FRG),
Loram (Lek, Yugoslavia), Lorans (Schiapar-
elli, Italy), Lorax (Wyeth, UK), Lorivan (Disco,
Israel), Merlit (Ebewe, Austria), Nervistop L
(Ingram, Argentina), Orfidal (Orfi, Spain), Pro
Dorm (Schürholz, FRG), Securit (Perrel, Italy),
Sedatival (Raffo, Argentina), Tavor (Wyeth,
FRG), Temesta (Ferrosan, Denmark; Wyeth,
Netherlands), Tolid (Dologiet, FRG), Tra-
pax (Wyeth, Argentina), Wypax (Yamanouchi,
Japan).
Trimethylacetate: Drupal (Novag, Spain), Pi-
ralone (Ferrer, Spain), Placinor (Robert, Spain).
Lormetazepam [848-75-9], 7-chloro-
5-(2-chlorophenyl)- 1, 3-dihydro-3-hydroxy-
1-methyl-2H-1,4-benzodiazepin-2-one (N-
methyllorazepam), (20c), C16 H12 Cl2 N2 O2 ,
M r 335.19, mp 192 – 194 ◦ C, is prepared analo-
gously to oxazepam [124]. The hypnotic effect
of 1 – 2 mg is similar to that of nitrazepam [125].
Trade Names. Lembrol (Vinas, Spain), Loramet
(Wyeth, Belgium, Netherlands), Minias (Far-
mades, Italy), Noctamid, Pronoctan (Schering,
FRG), Nocton (Eurolab, Chile).
Temazepam [846-50-4], 7-chloro-1,3-di-
hydro-3- hydroxy- 1- methyl- 5- phenyl-2H-
1,4-benzodiazepin-2-one (3-hydroxydiazepam,
N-methyloxazepam, oxydiazepam), (20d),
C16 H13 ClN2 O2 , M r 300.74, mp 119 – 121 ◦ C,
prepared analogously to oxazepam [126]. It is
used as a hypnotic drug and should be taken
1 – 2 h before going to sleep [127].
Trade Names. Euhypnos (Farmitalia Carlo Erba,
Italy; Sigma, Australia), Levanxene (Farmitalia
Hypnotics 23
Carlo Erba, Italy; Aesca, Austria), Mabertin
(Sidus, Argentina), Normison (Wyeth, UK),
Planum (Farmitalia Carlo Erba, Italy), Remestan
(Wyeth, FRG), Restoril (Sandoz, Switzerland;
Anca, Canada), Somaz (Quantum, USA), Tenso
(Castejon, Spain), Veroqual (Lek, Yugoslavia).
Doxefazepam [40762-15-0], 7-chloro-5-(2-
fluorophenyl)- 1, 3- dihydro- 3- hydroxy- 1- (2-
hydroxyethyl)-2H-1,4-benzodiazepin-2-one,
(20e), C17 H14 ClFN2 O3 , M r 348.57, mp 138 –
140 ◦ C, is prepared by the alkylation of 7-
chloro-5-(2-fluorophenyl)-1,3-dihydro-2H-1,4-
benzodiazepin-2-one N-oxide with 2-bromoeth-
anol or 2-bromoethyl acetate, followed by
Polonowski rearrangement and saponification
[128], [129]. Doxefazepam is a hypnotic drug
with two to four times the activity and half the
toxicity of flurazepam; it causes significantly
less hangover [130].
Trade Name. Doxans (Schiaparelli, Italy).
Cinolazepam [75696-02-5], 7-chloro-5-
(2-fluorophenyl)- 2, 3- dihydro- 3- hydroxy-2-
oxo-1H-1,4-benzodiazepine-1-propionitrile,
(20f),C18 H13 ClFN3 O2 , M r 356.84, mp 190 –
193 ◦ C, is prepared by the cyanoethylation
of compound 20 (R = Cl, R1 = H, R2 = F) in
the presence of trimethylammonium hydrox-
ide [131]. This drug resembles doxefazepam in
action [132].
Quazepam [36735-22-5], 7-chloro-5-(2-
fluorophenyl)-1-(trifluoro-2,2,2-ethyl)-1, 3-di-
hydro-2H-1,4-benzodiazepine-2-thione, (19),
C17 H11 ClF4 N2 S, M r 386.6, mp 138 – 139 ◦ C,
is synthesized by the reaction of compound
18 (R = Cl, R1 = CH2 CF3 , R2 = F) with phos-
phorus pentasulfide [133]. Quazepam is a sleep-
inducing drug; it has a long duration of action
and causes a hangover. It also produces long-
acting metabolites in the body [134].
Trade Names. Dormalin (Schering Corp., USA),
Prosedar, Selepam, Cetrane (Schering Plough,
USA).
Haloxazolam [59128-97-1], 10-bromo-
11b-(2- fluorophenyl)- 2, 3, 7- 11b- tetra-
hydrooxazolo-[3,2-d][1,4]benzodiazepin-
6(5H)-one,C17 H14 BrFN2 O2 , M r 377.22,
mp 185 ◦ C, is synthesized by the reaction of
bromo- (or tosyloxy-) acetamidobenzophenone
24 Hypnotics
with ethanolamine, followed by cyclization in
the presence of acetic acid [135].
Haloxazolam, oxazolam, and cloxazolam be-
long to a group of derivatives with an oxazoline
ring at the 4,5-position of the benzodiazepine
ring, which causes a change in some biologi-
cal parameters. Haloxazolam is primarily used
as a hypnotic drug in the treatment of insomnia
[136].
Trade Name. Somelin (Sankyo, Japan).
Estazolam [29975-16-4], 8-chloro-6-
phenyl-4H- [1,2,4]triazolo - [4,3- a][1,4]benzo-
diazepine, (21a), C16 H11 ClN4 , M r 294.75,
mp 228 – 229 ◦ C, is prepared as shown in Fig-
ure 6 [137], [138]. Indications are problems in
falling asleep, insomnia, and sleep induction in
anesthesiology [139]. The metabolism of the
triazolobenzodiazepines differs from that of the
benzodiazepinones in that the triazole ring re-
mains intact in most of the metabolites. One
of the first transformations in the metabolism
of benzodiazepines is the cleavage of the sub-
stituent at position 1.
Trade Names. Domnamid (Lundbeck, Den-
mark), Esilgan (Amer. Cyanamid, USA; Takeda,
Japan), Eurodin (Takeda, Japan), Nuetalon
(Casenne, France), Tasedan (Takeda, Mex-
ico), Somnatrol (Abbott, Peru), Noctal (Abbott,
Brazil).
Triazolam [28911-01-5], 8-chloro-6-
(2-chlorophenyl)- 1- methyl- 4H- [1,2,4]tri-
azolo[4,3-a][1,4]-benzodiazepine, (21b),
C17 H12 Cl2 N4 , M r 343.22, mp 223 – 225 ◦ C, is
prepared as shown in Figure 6 [138], [140]. Tri-
azolam is an extremely strong, but short-acting
hypnotic drug, the hypnotic dose being 0.5 mg
(equivalent to 30 mg of flurazepam) [141].
Doses of 0.125 – 1 mg are effective in induc-
ing sleep [142]. The maximum effect is pro-
duced 1 – 2 h after administration and lasts up
to 8 h; it is totally eliminated after 24 h. As with
nitrazepam and flunitrazepam, withdrawal after
prolonged use causes rebound insomnia [143].
Studies question the occurrence of the rebound
effect when standard doses (0.125 – 0.25 mg) are
taken [144]. For tolerance studies, see [145] and
for a review [146].
Trade Names. Halcion (Upjohn, USA), Novi-
dorm (Syntial, Argentina), Novodorm (Rubio,
Spain), Nuctane (Bago, Argentina).
Midazolam [59467-70-8], 8-chloro-6-(2-
fluorophenyl)- 1- methyl- 4H- imidazo[1,5-a]
[1,4]benzodiazepine,(21c), C18 H13 ClFN3 ,
M r 325.77, mp 158 – 160 ◦ C (maleate, mp 114 –
117 ◦ C), is synthesized as shown in Figure 6
[147]. The presence of the imidazole ring en-
sures a stable solution, a short duration of ac-
tion, and the formation of water-soluble in-
jectable salts. Below pH 4, the diazepine ring
is opened between positions 4 and 5 to give a
water-soluble salt of the primary amine. Ring
closure takes place above pH 4 with a half-life
of 10 min (22 → 21c). Indications are premedi-
cation before operations and anesthesia [148].
Trade Names. Dormicum, Dormonil, Hypnovel,
Sorenor (Hoffmann-La Roche, Switzerland),
Flormidal (Galenika, Yugoslavia).
 Hypnotics 25

Brotizolam [54801-81-7], 2-bromo-4-(2-

chlorophenyl)-9-methyl-6H-thieno[3,2-f ]
[1,2,4]triazolo[4,3-a][1,4]diazepine, (24),
C15 H10 BrClN4 S, M r 393.7, mp 212 – 214 ◦ C,
is synthesized by the condensation of 2-ami-
no-3-(2-chlorobenzoyl)-thiophene (23), with
aminoacetic ester, bromination, and fusion of
the triazole ring [149].

Brotizolam is very effective in prolonging

sleep and is an especially useful hypnotic for
light to moderate insomnia [150].

Trade Names. Ladormin, Lendorm, Lendormin

(Boehringer, FRG).

Loprazolam [61197-93-1], 6-(2-chloro-

phenyl)-2,4-dihydro-2-[(4-methyl-1-piper-
azinyl)methylene]- 8- nitro- 1H- imidazo[1,2-
a][1,4]benzodiazepin-1-one, (25),
C23 H21 ClN6 O3 , M r 464.91, mp 214 – 215 ◦ C,
(methanesulfonate, M r 560.71, mp 205 –
210 ◦ C), is synthesized by the condensation
of the appropriate benzodiazepinethione with
glycine, followed by cyclization. Condensation
with dimethylformamide acetal and subsequent
aminolysis with N-methylpiperazine gives lo-
prazolam (Fig. 7) [151].
This drug is an effective, medium-acting hyp-
notic and is especially suitable for the short-term
treatment of insomnia. It is also prescribed to
promote the onset of sleep and prevent frequent
awakening during the night [152].
Figure 7. Synthesis of loprazolam (25)
Trade Names. Dormonoct (Roussel Lab., UK),
Havlane (Diamant, France).
Rilmafazone [99593-25-6], 5-[(2-ami-
noacetamido)methyl]- 1- [4- chloro-2(2-chloro-
benzoyl)-phenyl]- N, N- dimethyl- 1H-1,2,4-tri-
azole-3-carboxamide, (26), C21 H20 Cl2 N6 O3
· HCl · 2 H2 O, M r 547.82, mp 107 ◦ C, is pre-
pared from 2 ,5-dichloro-2-aminobenzophe-
none as described in [153] (Fig. 8). Rilmafa-
zone belongs to a group of open-ring benzodi-
azepinone analogs called peptide aminobenzo-
phenones. It appears to be a precursor of the
triazolobenzodiazepinones. It has both sedative
26 Hypnotics

Figure 8. Synthesis of rilmafazone (26).

and antianxiety effects. Doses of 1 – 2 mg pro- subsequently esterified with 4-methyl-1-piper-

mote the onset of sleep. Indication is neurotic azinyl chloride [155].
insomnia. This drug was introduced in 1988
(Shionogi, Japan) [154].

6. Other Hypnotics
Other substances interact with the benzodi-
azepine receptors in vitro and in vivo. They
resemble benzodiazepines in general activity,
with similar sedative, anti-anxiety, and hypnotic
properties, but differences in adverse effects ex-
ist. The practical value of these drugs, which
include zopiclone, suriclone (anxiety-relieving),
zolpidem, and alpidem, will become apparent on
extended use.
A dose of 7.5 mg is required for the treat-
Zopiclone [43200-80-2], 6(5-chloro-2- ment of insomnia, which is equivalent to the ef-
pyridyl)-5-(4-methylpiperazin-1-yl)carbon- fect produced by 5 – 10 mg of nitrazepam. The
yloxy-7- oxo- 6,7- dihydro- 5H - pyrrolo[3,4- adverse effects are similar to those produced
b]pyrazine, (29), C17 H17 ClN6 O3 , M r 388.82, by short-acting benzodiazepines, but zopiclone
mp 178 ◦ C, is synthesized by the condensation of causes less hangover [156]. Zopiclone is, for the
2,3-piperazinedicarboxylic acid anhydride with most part, converted to inactive metabolites. The
2-amino-5-chloropyridine to give compound 27, elimination halftime for zopiclone and its active
which is partially reduced to compound 28 and

N-oxido metabolite (11 % formation) is 3.5 – 6 h
[157].
Trade Names. Imovane, Imovance, Theraplix
(Rhône-Poulenc, France), Amoban, Amovane
(Rhône-Poulenc, France).
Zolpidem [82626-48-0], 2-(4-methylphe-
nyl)-6-N,N- trimethylimidazo- [1,2-a]pyridine-
3-acetamide-l-(+)-tartrate (2 : 1) [99294-93-6],
(30a), (C19 H21 N3 O)2 · C4 H6 O6 , M r 764.88, is
prepared as described in [158]. This short-acting
hypnotic is rapid in onset; it reacts specifically
with the benzodiazepine type 1 receptor [159].
It is registered as Stilnox (Synthelabo, France)
and was put on the market in 1988 (licenced by
Searle).
Sulfones. The first synthetic hypnotics to be
introduced were sulfones (sulfonal and trional).
These sedatives and hypnotics are no longer used
because of their high toxicity and severe adverse
effects.
Antiepileptic Agents. Chlormethiazole
[53-45-9], 5-(2-chloroethyl)-4-methylthiazole,
is an antiepileptic drug occasionally used as a
hypnotic.
Trade Name. Distraneurin (Pharma Stern,
FRG).
Fenadiazole [1008-65-7], 2-(2-hydroxy-
phenyl)-1,3,4-oxadiazole, C8 H6 N2 O2 , M r
140.14, mp 111 – 112 ◦ C, is prepared by the con-
densation of salicylhydrazide with ethyl ortho-
formate [161]. The LD50 is 940 mg/kg (mouse,
i.p.).
Trade Name. Viodor (Violani-Farmavigor,
Italy).
Hypnotics 27
7. References
General References
1. AMA Drug Evaluations: Drugs Used for
Anxiety and Sleep Disorders, 6th ed., Chicago
1986, pp. 81 – 110.
2. J. A. Vida in W. O. Foye (ed.): “Central
Nervous System Depressants:
Sedative-Hypnotics,” Principles of Medicinal
Chemistry, Lea u. Febinger, Philadelphia 1976,
pp. 155 – 183.
3. J. A. Vida in A. Burger (ed.):
“Sedative-Hypnotics,” Medicinal Chemistry,
4th ed., Part III, New York, Chichester,
Brisbane, Toronto, 1980, pp. 787 – 828.
4. K. W. Wheeler: “Non-barbiturate Hypnotics,”
Med. Chem. (Academic) 6 (1963) 1 – 245.
5. W. J. Doran: “Barbituric Acid Hypnotics,”
Med. Chem. (Academic) 4 (1959) 1 – 334.
6. J. Hoffmann, B. Mencke, in G. Ehrhardt, H.
Ruschig (eds.): “Hypnotika,” Arzneimittel, 2nd
ed., vol. 1, Verlag Chemie, Weinheim 1966
pp. 37 – 66.
7. S. Ebel: “Hypnotika, Sedativa,” in
Synthetische Arzneimittel, Verlag Chemie,
Weinheim 1979, pp. 109 – 159.
8. E. Mutschler: Arzneimittel-Nebenwirkungen,
Hypnotika, Wissenschaftl. Verlags GmbH,
Stuttgart 1972, pp. 68 – 77.
9. S. C. Harvey in L. S. Goodman, A. Gilman,
A. G. Gilman, G. B. Koelle (eds.): “Hypnotics
and Sedatives,” The Pharmacological Basis of
Therapeutics, 7th ed.,Macmillan Publ., New
York, Collier Macmillan, Toronto, Ballière
Tindall, London 1985, pp. 339 – 371.
10. J. Oswald in E. Jucker (ed.): “Drug Research
and Human Sleep,” Arzneimittelforschung,
vol. 22, Birkhäuser, Basel 1978, pp. 355 – 372.
11. M. Monnier et al., “Biology of Sleep. An
Interdisciplinary Survey,” Experientia 36
(1980) 1 – 142.
12. J. P. Olié, S. Lepastier, G. Dana, J. M. Birée,
“Medicaments hypnotiques,” Med. Actuelle 5
(1978) 53 – 61; Chem. Abstr. 88 (1978)
145 812.
13. M. D. Allen, D. J. Greenblatt, Meyler’s Side
Eff. Drugs 4 (1980) 29 – 36; 3 (1979) 29 – 38;
2 (1978) 36 –44; 1 (1977) 21 – 29; VIII
1972 – 1975, 64 – 83, Excerpta Medica,
Amsterdam – Oxford.
14. H. P. Büch: “Medikamentöse Behandlung von
Schlafstörungen und Erregungszuständen,
Pharmakologie der Sedativa/Hypnotika,”
Dtsch. Apoth. Ztg. 121 (1981) 1951 – 1959.
28 Hypnotics
15. A. A. Borbély: “Schlafmittel und Schlaf:
übersicht und therapeutische Richtlinien,”
Ther. Umschau 43 (1986) 509 – 516.
16. E. Soos: “Schlaf und Schlafmittel,” OAZ
Österr. Apoth. Ztg. 28 (1974) 93 – 101.
17. A. N. Nicholson: “Hypnotics Their Place in
Therapeutics,” Drugs 31 (1986) 164 – 171.
18. E. A. Swinyard: Sedatives und Hypnotics in
Remington Pharmaceutical Sciences, 7th ed.,
Mack Publ., Easton, Pa., 1985,
pp. 1059 – 1074.
19. D. Schneider-Helmert: Schlafmittel:
“Symptomatisch oder kurativ?” Schweiz.
ärzteztg. 67 (1986) 1477 – 1483.
20. W. P. Koella: “The Organization and
Regulation of Sleep,” Experientia 40 (1984)
309 – 338.
21. W. Keup: “Mißbrauch und Abhängigkeit von
Sedativa und Hypnotika,” Internist 27 (1986)
746 – 756.
22. K. Laemmel: “Der Schlaf-Seismograph von
Körper und Seele,” Schweiz. Rundsch. Med.
Prax. 74 (1974) 1305 – 1310.
23. A. Borbély: Das Geheimnis des Schlafs,
Deutsche Verlagsanstalt GmbH, Stuttgart
1984.
24. T. Zilker, M. Clarmann:
“Schlafmittelvergiftungen,” MMW Münch.
Med. Wochenschr. 126 (1984) 297 – 300.
25. V. B. Meyer-Rochow: “Vom Winterschlaf
lernen,” Selecta 1984, 1116 – 1120.
26. Drug, Facts and Comparisons, Facts and
Comparisons Div., St. Louis 1982, 976 – 1016
(a division of J. B. Lippincott Co.,
Philadelphia-Toronto).
27. Martindale, The Extra Pharmacopoeia,
Hypnotic and Sedatives, 28th ed.,
Pharmaceutical Press, London 1982,
pp. 792 – 817.
28. E. A. Swinyard in Remington’s
Pharmaceutical Sciences, 17th ed., Mack
Publ. Easton, Pa., 1985, 1059 –1074.
Specific References
29. A. A. Borbély, J. Neural. Trans. Suppl. 21
(1986) 243 – 254.
30. S. H. Reppert et al., Science (Washington,
D.C.) 242 (1988) 78 – 81.
31. M. Monnier et al., Experientia 33 (1977) 548.
A. E. Kornmüller, H. B. Lux, K. Wink, M.
Klee, Naturwissenschaften 48 (1961)
503 – 505. D. Schneider-Helmert: “Die
körpereigene Schlafsubstanz DSIP,” Schweiz.
Rundsch. Med. 74 (1985) 1311 – 1315. M. V.
Graf, A. J. Kastin: “Delta-Sleep-Inducing
Peptide (DSIP): An Update,” Peptides (N.Y.) 7
(1986) 1165 – 1187; Neurosci. Biobehav. Rev.
8 (1984) 83 – 93.
32. A. A. Borbély, G. Youmbi-Balderer: “Ist
Tryptophan ein biologisches Schlafmittel,”
Schweiz. Rundsch. Med. Prax. 75 (1986)
785 – 790.
33. J. P. Olié, S. Lepastier, S. Dana, J. M. Birée,
Med. Actuelle 5 (1978) 53.
34. A. A. Borbély, Ther. Umsch. 43 (1986) 509.
35. M. B. Scharf, E. O. Bixler, A. Kales, C. R.
Soldatos, Pharmacology 19 (1979) 173 – 181.
A. Kales et al., Pharmacology 26 (1983)
138 – 149.
36. K. Laemmel, Schweiz. Rundsch. Med. Prax.
74 (1985) 1305 – 1310.
37. E. Greiser: Bewertender Arzneimittelindex:
Hypnotika, vol. 2, Med. Pharm. Verlag,
Wiesbaden 1983, pp. 1 – 25.
38. J. Krieglstein, D. Grusla, Dtsch. Apoth. Ztg.
128 (1988) 2041.
39. E. F. Domino, Annu. Rev. Pharmacol. 2
(1962) 215.
40. F. Hunzicker, F. Künzle, J. Schmitz, Helv.
Chim. Acta 49 (1966) 1 433. Wander,
GB 1 006 156, 1965; US 3 389 139, 1968.
41. G. Stille, A. Sayers, H. Lauener, E.
Eichenberger, Psychopharmakologia 28
(1973) 325; Chem. Abstr. 78 (1973) 131 982 c.
Y. Take et al., Takeda Kenkyushoho 29 (1970)
416; Chem. Abstr. 74 (1971) 21 815 n. H.
Yokotani et al., Takeda Kenkyushoho 29
(1970) 441; Chem. Abstr. 74 (1971) 21 816 p.
42. W. M. McLamre, S. Y. P. An, A. Barley, J. Org.
Chem. 20 (1955) 109. Pfizer, US 2 746 900,
1953.
43. Smith, Kline and French, DE-AS 1 183 333,
1963.
44. H. C. Caldwell, H. J. Adams, D. E. Rivard, J. V.
Swintovsky, J. Pharm. Sci. 56 (1967) 920.
45. British Drug Houses, GB 874 246, 1959.
46. Am. Home Prod., US 2 784 237, 1957.
47. Danske Medicinal, US 2 931 838, 1960.
48. A. Pictet, F. H. Reichel, Helv. Chim. Acta 6
(1923) 621. W. Freudenberg, A. M. Vajda, J.
Am. Chem. Soc. 59 (1937) 1955.
49. C. Bischoff, Ber. Dtsch. Chem. Ges. 7 (1874)
628.
50. Rheinpreussen, DE-AS 1 021 843, 1951.
51. Schering, US 2 816 910, 1953, Dtsch. Prior.
1952.
52. P. Läuger, M. Prost, R. Charlier, Helv. Chim.
Acta 42 (1959) 2379. Labaz, US 2 931 828,
1960.

53. Carlo Erba, GB 736 340, 1955.
54. G. Newbery, J. Chem. Soc. 127 (1925) 295.
Miles, US 2 854 379, 1958; US 2 931 832,
1960.
55. Bayer, DE 225 710, 1909. Friedländer 10,
1 160.
56. Bayer, DE 286 760, 1913. Friedländer 12,
704. Bayer, DE 327 129, 1917. Friedländer
13, 809.
57. Knoll, DE 185 962, 1906. Friedländer 8, 1217;
US 914 518, 1909.
58. Merck Index 10 (1983) 773. E. Schröder, C.
Rufer, R. Schmiechen: Pharmazeutische
Chemie, Thieme Verlag, Stuttgart – New York
1982, pp. 251 – 252.
59. Bayer, DE 249 241, 1912. Friedländer 10,
1 164. Labs. Aub., CH 374 644, 1964.
60. Millmaster Onyx, US 3 282 998, 1966. E. H.
Volwiler, D. L. Tabern, J. Am. Chem. Soc. 58
(1936) 1 352.
61. M. Bockmühl, O. Schaumann, Dtsch. Med.
Wochenschr. 54 (1928) 270.
62. Lion Dentifrice (Tokio), DE-OS 1 768 445,
1967; US 3 639 457, 1967; DE-OS 2 753 792,
1977.
63. N. Okudeira, S. Torii, S. Endo,
Psychopharmacology 70 (1980) 117. K.
Hillier, Drugs Future 6 (1981) 536 – 537.
64. W. J. Doran in F. F. Blicke, R. H. Cox (eds.):
“Barbituric Acid Hypnotics,” Med. Chem. Ser.
Monogr. 4 (1959).
65. R. Y. Levina, F. K. Velichko: “Advances in the
Chemistry of Barbituric Acids,” Russ. Chem.
Rev. (Engl. Transl.) 29 (1960) 437 – 459.
66. Kirk-Othmer, 2nd ed., 3, 60 – 77.
67. A. Schlieper, Justus Liebigs Ann. Chem. 55
(1845) 251; 56 (1845) 1.
68. R. Gross, Internist 17 (1976) 229.
69. R. J. Levina, F. K. Velichko: “Advances in the
Chemistry of Barbituric Acids,” Russ. Chem.
Rev. (Engl. Transl.) 29 (1960) 437.
70. E. Fischer, A. Dilthey, Justus Liebigs Ann.
Chem. 335 (1904) 338. Merck, DE 146 496,
1903.Friedländer 7, 651.
71. Bayer, DE 247 952, 1911, Friedländer 11,
926.
72. Poulenc Frères, US 1 609 520, 1926.
73. Lambert Thorpe Co., US 1 255 951, 1917; US
1 576 014,1926.
74. E. Leyraud, GB 191 008, 1922. Eli Lilly,
US 1 514 573, 1926; US 1 856 792, 1932.
75. H. H. Frey, M. P. Magnussen, Arzneim. Forsch.
16 (1966) 612.
76. Bayer, DE 442 655, 1924; DE 231 150, 1924.
77. Hoffmann-La Roche, US 1 444 802, 1923.
Hypnotics 29
78. Ciba, DE 268 158, 1911.
79. Hoffmann-La Roche, DE 526 854, 1930.
80. M. Seefelder, Justus Liebigs Ann. Chem. 652
(1962) 107. BASF, Ludwigshafen, Festschr.
Carl Wurster 60. Geburtstag 1960, 71; Chem.
Abstr. 56 (1962) 9 927 i; BASF FP 869 Méd.
1961.
81. Geigy, FR 870 714, 1941; DE 756 489, 1941.
82. Eli Lilly, US 1 856 792, 1932.
83. Bayer, DE 293 163, 1915. Geigy, GB 650 354,
1948.
84. J. D. Riedel AG, US 1 739 662, 1929.
85. A. C. Cope, E. M. Hancock, J. Am. Chem. Soc.
61 (1939) 776. Sharp & Dohme, US 2 187 703,
1940.
86. E. H. Volwiler, J. Am. Chem. Soc. 47 (1925)
2236.
87. Compagnie de Bethune, DE 589 947, 1930.
Friedländer 19, 1 198.
88. J. D. Riedel AG, US 1 622 129, 1927;
DE 481 733, 1923.
89. Eli Lilly, US 1 954 429, 1934.
90. O. Schneider, H. Frick, A. H. Lutz, Experientia
10 (1954) 135. Festschrift F. Reinhardt AG,
Basel 1936, p. 195. Hoffmann-La Roche,
US 2 151 047, 1938; US 2 680 116, 1954.
91. Ciba-Geigy, US 2 673 205, 1954,
Schweiz.-Prior. 1951; DE 950 193, 1952. L.
Hoffmann, E. Tagmann, Helv. Chim. Acta 35
(1952) 1 541.
92. Richardson-Merrell Inc. BE 753 846, 1971;
DE-AS 2 035 636, 1969.
93. H. Koch, J. Kotlan, Monatsh. Chem 97 (1966)
1 648;Kwizda, DE-AS 1 695 346, 1966.
94. L. Stockinger, H. Koch, Arzneim. Forsch. 19
(1969) 167.
95. G. Argyropoulos, G. Lanner, E. Pasquali,
Wien. Med. Wochenschr. 126 (1976) 75; Med.
Actual. 16 (1980) 388.
96. K. H. Boltze et al., Arzneim. Forsch. 13 (1963)
688.
97. B. Dubnick, C. A. Towne in A. Cerletti, F. J.
Bove (eds.): “The Present Status of
Psychotropic Drugs,” Proc. of the VI. Internat.
Congr. of the Coll. Internat.
Neuro-Psychopharmacologicum, Tarragona
Apr. 1968 pp. 24 – 27; Excerpta Med. 1969,
77 – 83.
98. I. K. Kacker, S. H. Zaheer, J. Indian Chem.
Soc. 28 (1951) 344; Chem. Abstr. 46 (1952)
11 210. General Aniline u. Film Corp.,
US 3 213 094, 1963. Lab. Toraude,
GB 843 073, 1960.
30 Hypnotics
99. Beiersdorf AG, BE 615 282, 1962; Chem.
Abstr. 58 (1963) 13 971 ef. S. S. Parmar, K.
Kishok, P. K. Seth, R. C. Arona, J. Med. Chem.
12 (1969) 138.
100. J. Klosa, J. Prakt. Chem. 14 (1961) 84; 20
(1963) 283.
101. E. Szirmai, Ther. Umsch. 20 (1963) 542.
102. L. H. Sternbach: “The Benzodiazepine Story,”
Progr. Drug Res. 22 (1978) 229 – 266. W.
Schallek, W. D. Horst, W. Schlosser:
“Mechanism of Action of Benzodiazepines,”
Adv. Pharmacol. Chemother. 16 (1979)
45 – 87. L. O. Randall, W. Schallek, L. H.
Sternbach, R. Y. Ning in M. Gordon (ed.):
“Chemistry and Pharmacology of the
1,4-Benzodiazepines,”
Psychopharmacological Agents, A Series of
Monographs in Medicinal Chemistry, vol. 3,
Academic Press, New York – San
Francisco – London 1974, p. 175. V. C.
Williams, G. C. Varnado, P. U. Nwangwu:
“Perspectives on the Clinical Pharmacology of
Benzodiazepines,” Med. Actual. 21 (1985)
75 – 96. T. A. Hamor, J. L. Martin in G. P. Ellis,
G. B. West (eds.): “The Benzodiazepines,”
Progress in Medicinal Chemistry 20 (1985)
158 – 223. A. Vela-Bueno, A. Kales:
“Benzodiazepine Hypnotics in the
Multidimensional Treatment of Insomnia,”
Med. Actual. 22 (1986) 271 – 281. C. Pina, A.
Catedra: “Newer Benzodiazepines, Med.
Actual. 11 (1975) 176; “Chemical Synthesis of
Benzodiazepines (nur Literatur),” Med.
Actual. 11 (1975) 211 – 212. H. Schütz:
Benzodiazepines: A Handbook, Springer
Verlag, Berlin 1982. C. Bellantuono, V. Reggi,
G. Tognoni, S. Garattini: “Benzodiazepines –
Clinical Pharmacology and Therapeutic Use,”
Drugs 19 (1980) 195 – 219. S. Garattini: The
Benzdiazepines, Raven Press, New York 1973.
103. M. Lader (ed.): “New Perspectives in
Benzodiazepine Therapy,” Arzneim. Forsch.
30 (1980) no. 5 a, 851 – 913. U. Klotz:
“Clinical Pharmacology of Benzodiazepines,”
in Progress in Clinical Biochemistry and
Medicine, vol. 1, Springer,
Berlin – Heidelberg – New York – Tokyo 1984,
pp. 117 – 166.
104. H. Schrum: “Benzodiazepine als Schlafmittel,”
Schweiz. Rundsch. Med. Prax. 74 (1985)
1301 – 1304.
105. R. Weber, Dtsch. Apoth. Ztg. 128 (1988)
2643 – 2652.
106. L. L. Iversen, Nature (London) 275 (1978)
447. R. E. Study, J. L. Barker, JAMA J. Am.
Med. Assoc. 247 (1982) 2147 – 2151. U.
Klotz: Progress in Clinical Biochemistry and
Medicin, Springer, Berlin – Heidelberg – New
York – Tokyo 1984, pp. 127 – 129. T. A.
Hamor, J. L. Martin in G. P. Ellis, G. B. West
(eds.): Progress in Medicinal Chemistry 20
(1985) p. 183.
107. W. E. Müller, MMW Münch. Med. Wochenschr.
124 (1982) 879 – 884; Dtsch. Apoth. Ztg. 126
(1986) 1036 – 1044.
108. W. E. Müller, Med. Actual. 24 (1988)
649 – 663.
109. E. Müller, Dtsch. Apoth. Ztg. 128 (1988)
672 – 674. Roche Magazin 1988, no. 31, 3 – 13
(Hoffmann-La Roche).
110. S. Ebel, H. Schütz: “Biotransformation von
Benzodiazepin-Derivaten,” Pharm. Unserer
Zeit 8 (1979) 87. U. Klotz, L. Kangas, J.
Kanto: “Clinical Pharmacocineties of
Benzodiazepines,” Prog. Pharmacol. 3 (1980)
1 – 72. W. E. Müller: “über die Bedeutung der
Plasmaproteinbindung für Verteilung und
Wirkung von Tranquillantien vom
Benzodiazepintyp,” Klin. Wochenschr. 55
(1977) 105 – 110.
111. R. G. Borland, A. N. Nicholson, Br. J. Clin.
Pharmacol. 2 (1975) 9. S. C. Harvey in
Goodman Gilman, The Pharmacological Basis
of Therapeutics, 7th ed., Macmillan Publ.,
New York, Toronto, London 1985, p. 352; see
also [102] p. 82.
112. Hoffmann-La Roche, DE-AS 1 136 709, 1959;
DE-AS 1 145 626, 1959; US 3 109 843, 1961;
US 3 116 203, 1962; US 3 123 529, 1964.
Delmar Chem., DE-AS 1 811 785, 1968.
113. L. H. Sternbach et al., J. Med. Chem. 6 (1963)
261. Hoffmann-La Roche, US 3 116 203,
1963; US 3 123 529, 1964; US 3 203 900,
1964; GB 1 039 372, 1966.
114. M. W. Johns, Drugs 9 (1975) 448. J. Rieder,
Arzneim. Forsch. 23 (1973) 212. Martindale:
The Extrapharmacopoea, 28th ed.,
Pharmaceutical Press, London 1982,
pp. 807 –809. Med. Actual. 1 (1964)
132 – 138; 9 (1973) 237.
115. J. M. Monti, H. Altier, Psychopharmacologia
32 (1973) 343. W. Schallek, J. Kovacs, A.
Kuehn, J. Thomas, Arch. Int. Pharmacodyn.
Ther. 206 (1973) 161. S. A. Kaplan et al., J.
Pharm. Sci. 63 (1974) 527. R. M. Joye, A. J.
Hance, K. F. Killam, Neuropharmacology 10
(1971) 483. J. Freuchen, J. Østergaard, B. O.
Mikkelsen, Curr. Ther. Res. 23 (1978) 90.
R. T. Owen, Med. Actual 19 (1983) 163 – 167;

11 (1975) 176 – 178. M. A. K. Mattila, H. M.
Larni, Drugs 20 (1980) 353 – 374.
116. L. H. Sternbach et al., J. Med. Chem. 8 (1965)
815. Hoffmann-La Roche, BE 629 005, 1963;
US Prior. 1962; Chem. Abstr. 60 (1963)
1 589 f; NL 6 401 335, US Prior. 1963; Chem.
Abstr. 62 (1965) 5288.
117. Hoffmann-La Roche, US 3 567 710, 1971.
118. S. Inaba, K. Ishizumi, H. Yamamoto, Chem.
Pharm. Bull. 19 (1971) 263.
119. L. O. Randall et al., Arch. Int. Pharmacodyn.
Ther. 178 (1969) 216. H. Jick, Curr. Therap.
Res 9 (1967) 355. J. d’Armagnac et al.,
Therapie 26 (1971) 439. A. Schwartz, F. M.
Vane, E. Postma, J. Med. Chem. 11 (1968)
770. Med. Actual. 6 (1970) 217. USP XXI,
447 – 449, Rockville, Md. 1985. AMA: Drug
Evaluations, 6th ed., Am. Med. Assoc.,
Chicago 1986, pp. 99 – 100.
120. S. Inaba, K. Ishizumi, K. Mori, H. Yamamoto,
Chem. Pharm. Bull. 19 (1971) 722.
Hoffmann-La Roche, US 3 144 439, 1964; US
3 144 890, 1964; US 3 243 427, 1966. Dumex,
ZA 6 706 791, 1968; Chem. Abstr. 70 (1969)
57 916 c. Sumitomo, DE-OS 1 816 046, 1970.
121. S. Sakai, S. Kitajawa, H. Yamamoto, Arzneim.
Forsch. 22 (1972) 534. H. Yamamoto et al.,
Sumitomo Kagaku, Tokushugo 1977, 13 – 23;
Chem. Abstr. 90 (1977) 26 w.
122. Martindale: The Extra Pharmacopoeia, 21st
ed., The Pharmaceutical Press, London 1982,
p. 1549. USP XXI, 762, Rockville, Md. (1985).
123. M. J. Bluckman et al., Arzneim. Forsch. 21
(1971) 1049 – 1102. Martindale: The Extra
Pharmacopoeia, 21st ed., The Pharmaceutical
Press, London 1982, p. 1543.
124. S. C. Bell et al., J. Med. Chem. 11 (1968) 457.
A. Nudelman, J. Pharm. Sci. 63 (1974) 1886.
Amer. Home Prod., BE 621 819, 1963; Chem.
Abstr. 60 (1964) 2993 b; DE-OS 1 445 412,
1962; DE-OS 1 645 904, 1962;
DE-OS 1 795 553, 1962; DE-OS 1 795 556,
1962.
125. A. Doenicke et al., Anaesthesist 28 (1979)
578. H. Ott et al., Anaesthesist 28 (1979) 29.
R. Girkin et al., Xenobiotica 10 (1980) 401.
M. Hümpel et al., Eur. J. Drug. Metab.
Pharmacokinet. 4 (1979) 237. J. Oswald et
al., Br. Med. J. 1982, I, 860. R. T. Owen, Drugs
Future 5 (1980) 495 – 496. Pharmcast 87-10,
6-II-136 – 6-II-138, Technomic Inf. Service,
Tokyo.
126. S. C. Bell, S. J. Childress, J. Org. Chem. 27
(1962) 1691. Am. Home Prod., US 3 197 467,
Hypnotics 31
1965. Hoffmann-La Roche, US 3 340 253,
1967; US 3 374 225, 1968.
127. L. O. Randall, C. L. Scheckel, W. Pool, Arch.
Int. Pharmacodyn. Ther. 185 (1970) 135. P.
Sarteschi et al., Arzneim. Forsch. 22 (1972)
93. H. J. Schwandt, W. Sadee, K.-H. Beyer,
Xenobiotica 4 (1974) 733. S. H. Curry, A. N.
Nicholson, R. Whelpton, Br. J. Pharmacol. 57
(1976) 427. AMA: Drug Evaluations, 6th ed.,
Am. Med. Assoc. Chicago 1986,
pp. 102 – 103. Drug, Facts and Comparisons,
Facts and Comparisons Div., St. Louis 1982,
pp. 991 – 993 (a division of J. B. Lipincott Co.,
Philadelphia –Toronto).
128. G. F. Tamagnone, R. De Maria, F. De Marchi,
Arzneim. Forsch. 25 (1975) 720. Schiapparelli,
DE-OS 2 338 058, 1972. Japk 7480 086,
GB 1 431 282, 1972. Kleemann-Engel:
Pharmazeutische Wirkstoffe, Suppl.
1982 – 1987, Thieme-Verlag, Stuttgart – New
York 1987, pp. 1151 – 1152.
129. Hoffmann-La Roche, ZA 6 802 239, 1967.
130. G. F. Tamagnone, M. V. Torrielli, F. De Mardi,
J . Pharm. Pharmacol. 26 (1974) 566. V.
Monaco, A. Bontante, Curr. Ther. Res. 37
(1985) 626. L. De Angelis, Drugs Future 3
(1978) 145 – 148. S. Mardene, C. Biechie,
G. M. Nano, Ther. Drug Monit. 3 (1981)
351 – 356. M. Babbini, M. V. Torielli, E.
Strumia et al., Arzneim. Forsch. 25 (1975)
1294.
131. Gerot-Pharmazeutika, DE-OS 2 950 235, 1980;
AT 379 391, 1984; Chem. Abstr. 104 (1986)
168 499 j.
132. H. Koch, Drugs Future 7 (1982) 99.
133. M. Steinmann, J. G. Topliss, R. Alekel et al., J.
Med. Chem. 16 (1973) 1354. Schering,
US 3 920 818, 1974. Scherico Ltd.,
DE-OS 2 138 733, 1971.
134. H. Winsauer, D. E. O’Hair, Curr. Ther. Res. 35
(1984) 228. A. Barnett, L. C. Jorio, E. Ongini,
Arzneim. Forsch. 32 (1982) 1452. J. Castaner,
Drugs Future 3 (1978) 139; 8 (1983) 165; 9
(1984) 152; 10 (1985) 171. S. J. Ankier, K. A.
Goa, Drugs 35 (1988) 42 – 62. AMA: Drug
Evaluations, 6th ed., Am. Med. Assoc.,
Chicago 1986, p. 102. Annu. Drug Data
Report 8 (1986) 119; 2 (1979 – 1980) 192.
135. T. Miyadere et al., J. Med. Chem. 14 (1971)
520. Sankyo, DE-OS 1 812 252, 1969; J-Kokai
1967; Chem. Abstr. 71 (1969) 124516 h.
136. T. Kamioka, J. Nakayama, T. Hara, H. Takaji,
Arzneim. Forsch. 28 (1978) 838. Y. Sakai,
Jpn. J. Pharmacol. 33 (1983) 1017 – 1025. M.
Yoshimoto et al., Chem. Pharm. Bull. 25
32 Hypnotics
(1977) 1378. Med. Actual. 17 (1981)
283 – 286. P. J. Thorpe, Drugs Future 6 (1981)
414 – 415.
137. J. B. Hester, D. J. Duchamp, G. C. Chichester,
Tetrahedron Lett. 1971, 1609. J. B. Hester,
A. D. Rudzick, B. V. Kamdar, J. Med. Chem.
14 (1971) 1078. K. Meduro, H. Tawada, Y.
Kuwada, Chem. Pharm. Bull. 21 (1973) 1619,
2382. K. Kamiya, Y. Wada, M. Nishikawa,
Chem. Pharma. Bull. 21 (1973) 1520. K.
Megruo, Y. Kuwada, Tetrahedron Lett. 1970,
4039. D. L. Loffen, R. J. Fryer, D. A. Katonak,
F. Wong, J. Org. Chem. 40 (1975) 894.
138. Upjohn, DE-OS 2 012 190, 1969;
US 3 701 782, 1969; DE-OS 2 302 525, 1973;
US 4 018 788, 1977; US 3 987 052, 1969.
Toyama, DE-OS 1 955 349, 1971. Takeda,
DE-OS 2 114 441, 1971; DE-OS 1 965 894,
1969; US 4 116 956, 1976.
139. S. Tanayama, Y. Shirakawa, Y. Kanai, Z.
Suzuoki, Xenobiotika 4 (1974) 33. Y. Kanai,
Xenobiotika 4 (1974) 441. F. S. Eberts,
Pharmacologist 16 (1974) 196; Drug Metab.
Dispos. 5 (1977) 547. H. Isozaki, M. Tanaka,
K. Inanaga, Curr. Ther. Res. 20 (1976) 493. T.
Momose, S. Shii, T. Kuge, Curr. Ther. Res. 19
(1976) 277. W. Ehrenstein, M.
Müller-Limmroth, M. Opfermann, F. Weber,
Arzneim. Forsch. 32 (1982) 280. J. K. Walsh
et al., Curr. Ther. Res. 36 (1984) 866. Med.
Actual. 12 (1976) 355 – 357. Annual Drug
Data Report 2 (1979/1980) 85 – 86.
140. J. B. Hester, A. D. Rudzick, B. V. Kamdar, J.
Med. Chem. 14 (1971) 1078.
141. A. Kales, M. B. Scharf, J. D. Kales, Science
(Washington, D.C.) 201 (1978) 1039.
142. T. Roth, M. Kramer, T. Lutz, Drugs Exp. Clin.
Res. 1 (1977) 279. D. A. C. Hegelbach-Feller,
J. M. Tschopp, S. Christeller, J. Fabre,
Arzneim. Forsch. 38 (1988) 387.
143. M. Leibowitz, A. Sunshine, J. Clin.
Pharmacol. 18 (1978) 302. K. K. Okawa,
G. S. Allen, J. Int. Med. Res. 6 (1978) 343.
B. B. Knap, E. L. Boyd, G. Linsenmeyer, O. J.
Linet, Curr. Ther. Res. 23 (1978) 230.
144. A. N. Sing, Curr. Ther. Res. Clin. Exp. 27
(1980) 627. A. N. Nicholson, Br. J. Clin.
Pharmacol. 9 (1980) 223.
145. L. Lasagnon, Lancet 1 (1980), 815.
146. E. A. Swinyard in Remingtons Pharmaceutical
Sciences, 17th ed., Mack Publ., Easton, Pa.
1985, p. 1064. AMA: Drug Evaluation, 6th
ed., Am. Med. Assoc., Chicago 1986, p. 103. J.
Castaner, Drugs Future 1 (1976) 393 – 396; 2
(1977) 558.
147. A. Walser et al., J. Org. Chem. 43 (1978) 936.
Hoffmann-La Roche, DE-OS 2 540 522, 1976.
148. M. Gemper et al., Arzneim. Forsch. 31 (1981)
2177 – 2288. D. A. C. Hegelbach-Feller, J. M.
Tschopp, S. Christeller, J. Fabre, Arzneim.
Forsch. 38 (1988) 387. J. Kanto, Med. Actual,
19 (1983) 376 – 381.
149. K. H. Weber, W. Sirrenberg, O. Spohn, H.
Daniel, Arzneim. Forsch. 36 (1986) 518.
Boehringer, DE-OS 2 410 030,
1974 = US 4 094 984, 1975; DE-OS 2 503 235,
1975; DE-OS 2 533 924, 1975. Hoffmann-La
Roche, DE-OS 2 221 623, 1971. K. Gewald,
Chem. Ber. 98 (1965) 3571.
150. N. Ferrara et al., Curr. Ther. Res. 37 (1985)
295. J. Grünberger et al., Curr. Ther. Res. 24
(1978) 427. B. Saletu, J. Grünberger, J.
Volavka, P. Berner, Arzneim. Forsch. 29
(1979) 700. M. S. Langley, S. P. Clissold,
Drugs 35 (1988) 104 – 122. Pharmcast 87-10,
Technomic Information Service, Tokyo 1987,
pp. 6-II-151 – 6-II-152.
151. R. Agar et al., J. Med. Chem. 20 (1977) 1035.
Russel-UCLAF, DE-OS 2 605 652, 1976; US
4 044 142, 1977.
152. T. G. Johns, D. C. Piper, G. W. L. James, Arch.
Int. Pharmacodyn. Ther. 240 (1979) 53. J.
Hindmarch, C. A. Clyde, Drugs Exp. Clin.
Res. 2 (1980) 61. R. T. Owen, Med. Actual. 20
(1984) 59 – 63. Loprazolam Clinical
Workshop, The Medicine Publishing
Foundation, Oxford, 1983.
153. K. Hirai et al., Heterocycl. Chem. 19 (1982)
1363; J. Med. Chem. 23 (1980) 764; 24 (1981)
20. Shionogi, DE-OS 2 725 164, 1977;
US 4 159 374, 1977. Hoffmann-La Roche,
DE-OS 2 537 069, 1975.
154. K. Hirose, A. Matsushita, M. Eigyo, Arzneim.
Forsch. 31 (1981) 63. Drugs Future 9 (1984)
522 – 524; 11 (1986) 626 – 628. Pharmcast
87-1, Technomic Information Service, Tokyo
1987, pp. 5-III-139 – 5-III-141.
155. Rhône-Poulenc, DE-OS 2 300 491, 1973; US
3 862 149, 1975. C. Jeanmart, C. Cotrel, C. R.
Ser. C 287 (1978) 377.
156. R. Duriez, Therapie 34 (1979) 317. E.
Wickström, K. E. Giercksky, Eur. J. Clin.
Pharmacol. 17 (1980) 93; Clin. Therap. 3
(1978) 59. . J. C. Blanchard, A. Boireau, C.
Garrett, L. Julou, Life Sci. 25 (1979) 2417.
P. H. Wu, J. W. Phillis, A. S. Bender, Life Sci.
28 (1981) 1023. J. Castaner, Drugs Future 4
(1979) 59 – 60; 5 (1980) 56; 6 (1981) 62 – 63;
7 (1982) 69 – 70; 8 (1983) 79; 9 (1984) 81; 10
 Hypnotics 33

(1985) 88 – 89; 11 (1986) 76 – 78; 12 (1987)
95; 13 (1988) 88 – 89. Pharmacast 87-7,
Technomic Inf. Service, Tokyo 1987, 1989,
6-I-87 – 6-I-89. Int. Pharmacopsychiatry 17
(1982) (Suppl. 2), 1 – 252. K. A. Goa, R. C.
Heel, Drugs 32 (1986) 48 – 65.
157. G. Houghton et al., Int. J. Clin. Pharmacol.
Biopharm. 23 (1985) 97. R. Duriez et al.,
Therapie 34 (1979) 317. J. Gaillot et al.,
Pharmacology 27 (1983) (Suppl. 2) 76 – 91.
158. Synthelabor, EP 50 563, 1981. P. George, H.
Depoortere, J. Pharm-Sci. (in Press).
159. S. Arbilla et al., Naunyn-Schmiedebergs Arch.
Pharmacol. 330 (1985) 248. M. Dimsdale et
al., Drugs Future 12 (1987) 777 – 780.
160. M. Dimsdale et al., Drugs Future 13 (1988)
106 – 109.
161. J. Maillard, M. Vincent, M. Bernard, Bull. Soc.
Chim. Fr. 1961, 529; 1962, 1580.
Logeais-Labs., GB 902 388, 1962;Fr. M. 379,
1962; Chem. Abstr. 57 (1962) 15251 g.

Hypochlorous Acid → Chlorine

Hypochlorous Acid → Chlorine Oxides and Chlorine Oxygen Acids
Hypoglycemic Drugs → Antidiabetic Drugs
Hypophosphorus Acid → Phosphorus Compounds, Inorganic

Ice Cream and Frozen Desserts
Philip G. Keeney, Pennsylvania State University, University Park, Pennsylvania 16803, United States
1. Introduction . . . . . . . . . . . . . . . .
2. Classification of Ice Cream Products
3. Legal Aspects . . . . . . . . . . . . . . .
4. Economic Aspects . . . . . . . . . . . .
5. Physical Structure . . . . . . . . . . . .
6. Mix Composition . . . . . . . . . . . . .
6.1. Fats . . . . . . . . . . . . . . . . . . . . .
6.2. Nonfat Milk Solids . . . . . . . . . . . .
1. Introduction
The term ice cream is generally used to cover
several types of desserts that are consumed in
a frozen or partially frozen state. Ice cream
is a frozen foam of an oil-in-water emulsion,
which contains more than 55 wt % water and
ice. Carbohydrate solutes contribute sweetness,
but also control freezing and melting charac-
teristics. The properties of the foam matrix are
largely determined by lipid and protein; small
amounts of emulsifier and hydrocolloid function
as manufacturing aids and help to stabilize phys-
ical properties during storage and distribution.
Flavor ingredients, mostly of botanical origin,
(e.g., fruit, fruit products, cocoa, coffee, ginger,
vanilla), allow great variations in aroma, taste,
and texture impressions.
2. Classification of Ice Cream
Products
The manufacturer has several variables to con-
sider in the design and production of products to
meet market needs. Common desirable traits are
sweet and cool tastes that satisfy the expected re-
freshing qualities. Superimposed on these basic
requirements are the influences of the character-
istic flavor ingredients and the effects of fat, pro-
tein, other mix components, air, and ice crystals
on creaminess, smoothness, and other physical
qualities perceived upon consumption.
The “Codex Standard on Edible Ices and Ice
Mixes” gives a comprehensive survey of types

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a13 563
Ice Cream and Frozen Desserts 1
. 1 6.3. Sweetener Solids . . . . . . . . . . . . . . 5
. 1 6.4. Stabilizers and Emulsifiers . . . . . . . 6
. 4 7. Manufacture of Mix . . . . . . . . . . . . 6
. 4
8. Freezing Mix to Produce Ice Cream . 7
. 4
8.1. Freezing as Mix Agitates . . . . . . . . . 7
. 5
. 5 8.2. Quiescent Hardening . . . . . . . . . . . 8
. 5 9. References . . . . . . . . . . . . . . . . . . 8
of ice cream and a detailed account of food addi-
tives (colors, emulsifers, stabilizers, thickening
agents, flavors, etc.) including maximum lev-
els and microbiological requirements [1]. The
Codex Standards are merely recommendations
for UN countries but the importance of the Edi-
ble Ice Standards has not declined since the EEC
commission abandoned its plan to introduce Ed-
ible Ice Guidelines a few years ago.
Most countries have regulatory definitions
consistent with local customs and needs. Almost
always, one product class reflects the dominance
of milk ingredients in determining consumption
characteristics. In the United States these are ice
cream and ice milk, which are similar except that
the latter contains less fat. In the United King-
dom, ice cream made with milk fat is termed
dairy ice cream. In Europe an increasing propor-
tion of ice cream is sold as premium ice cream,
which has a higher fat content than normal ice
cream and is not as highly whipped.
In some countries (e.g. Japan, Sweden, Eng-
land), vegetable fats may be used in place of milk
fat. In fact ca. 80 % of ice cream consumed in
these countries contains vegetable fat. Mellorine
is a product in which all or a portion of the milk
fat is replaced by another fat of plant or animal
origin.
A second class of products is dominated by
the soluble carbohydrate fraction and contains
little or no fat and protein. These are the sher-
bets, sorbets, and ices. A third category com-
prises products with a dietary attraction. Exam-
ples are artificially sweetened products for peo-
ple who desire low-calorie foods or who must
control dietary sugar intake. In these products
2 Ice Cream and Frozen Desserts
Table 1. International compositional requirements in ice cream and related products a

Table 1. Continued.
sucrose is replaced by sugar substitutes such as
fructose or sorbitol to ensure that the ice cream
retains desirable freezing and melting character-
istics.
Ice cream is sold in a variety of forms, shapes,
and flavors. Forms of supply include:
1) single portions (sticks, cups, tubs, cones,
sandwiches, puppet ice),
2) family-sized portion (bricks, bombes, multi-
packs), and
3) catering packs.
A further differentiation among products is
based on whether freezing occurs quiescently,
and/or under agitation. Most stick confections
are subjected to quiescent freezing by pouring
liquid mix into molds immersed in a brine so-
lution at −35 ◦ C. The finished product with in-
serted stick has the geometry of the mold, which
allows for a great variety of shapes and sizes.
Freezing under agitation refers to products in
which 35 – 55 % of the water content is frozen
as the temperature is lowered to between −4 and
−6 ◦ C in a swept-surface heat-exchange cylin-
der, which also distributes air as overrun. (Over-
Ice Cream and Frozen Desserts 3
run refers to volume expansion caused by air in-
corporation, it being 80 % when 1 L of liquid mix
yields 1.8 L of frozen product.) Product extruded
from the freezer can be dispensed into cones or
cups to be consumed in soft form which cannot
be stored or transported. Alternatively, it can be
packaged and placed in a colder environment (at
below −18 ◦ C) to allow quiescent freezing of
most of the remaining water to yield ice cream
in its typically hardened, packaged form. The
latter process is a combination of freezing quies-
cently and under agitation. Hardened ice cream
(including varieties that are easily scoopable at
−20 ◦ C) can be stored and transported.
3. Legal Aspects
International compositional requirements for ice
cream and related products are summarized in
Table 1.
In the United States ice cream, ice milk, sher-
bet, fruit ice, and mellorine are defined under the
standards of the federal Food and Drug Admin-
istration [2]. Guidelines for other products are
specified by individual states. Key provisions for
4 Ice Cream and Frozen Desserts

ice cream are a minimum of 10 % milk fat and
20 % total milk solids, and a product weight of at
least 4.5 lb per gallon (0.54 kg/L). The weight re-
quirement prevents more than a doubling of mix
volume through air incorporation during manu-
facture. Ice milk is low-fat ice cream (with a fat
content of 2 – 7 wt %). In mellorine some or part
of the milk fat is replaced by another plant or an-
imal fat. Sherbet must contain 1 – 2 % milk fat
with not more than 5 % total milk solids and a
minimum weight of 6 lb per gallon (0.71 kg/L).
Most sherbets have a tart taste due to the addi-
tion of an edible acid, usually citric acid. Fruit
ice requirements are similar to those of sherbet,
except that milk solids are not allowed.
air cell [5]. Evidence exists for accumulation of
protein and fat near the interface between air
and the liquid, unfrozen continuous phase (also
called serum) [5]. Emulsions are thermodynam-
ically unstable [6] and changes occurring are the
result of the shearing of the mix during freezing
and whipping in the swept-surface cylinder. Par-
ticle size distribution for fat tends towards larger
globules and formation of agglomerates is evi-
dent. Depending on the amount of change, the
physical characteristics of the frozen emulsion
matrix are profoundly influenced.

4. Economic Aspects
In the United States, 65 % of the 1.4×109 gal-
lon (5.3×106 m3 ) annual production of frozen
desserts is ice cream, 22 % is ice milk, 7 % sher-
bet, 1 % mellorine, and 5 % other products [3].
Soft-serve products account for 21 % of the to-
tal production, most being ice milk. At 22 L
per capita, production is highest in the United
States, followed closely by Finland at 21 L, New
Zealand and Australia at 19 L, then Canada and
Figure 1. Illustration of ice cream structure
Sweden with 14 – 19 L. a) Ice crystal; b) Fat globules; c) Air cell

5. Physical Structure
Ice cream being a foam of an oil-in-water emul-
sion is a composite of liquid, solid, and gaseous
phases (Fig. 1). The numbers and sizes of air
cells and ice crystals are influenced by several
manufacturing and storage variables. A repre-
sentative diameter for an ice crystal is 30 µm
with air cells averaging 60 µm. The proportion
of the mass which is present as a continuous liq-
uid phase depends on water – ice equilibria as
determined by temperature.
Changes in the nonfrozen phase during freez-
Figure 2. Wall of air cell in ice cream revealed by freeze-
ing are considerably more complex than a mere etching and transmission electron microscopy [4].
concentration of soluble constituents as water
crystallizes. The beating and scraping action
which facilitates heat transfer and air incorpo-
ration in the freezing cylinder can profoundly 6. Mix Composition
alter colloidal and emulsion properties. An elec-
tron micrograph of ice cream (Fig. 2) shows that Table 2 shows typical mix compositions for two
the fat particles are located at the surface of the types of ice cream, an ice milk, and sherbet [7].
 Ice Cream and Frozen Desserts 5

Mix for fruit ice is similar to that for sherbet, the emulsion in homogenized mix, it also allows
except for the absence of milk solids. Although controlled colloid and emulsion changes during
each mix constituent is important for determin- freezing. Protein is essential for the encapsula-
ing the qualities perceived upon consumption of tion of air as overrun, and its hydrophilic prop-
ice cream, the amount and type of fat is espe- erties enhance desirable texture characteristics.
cially critical. Thus, the fat level of the mix is Because the nonfat solids content of milk is
decided first and then the proportions of the other only 8.6 wt %, a concentrate must be employed
constituents are selected to complement it. to reach the levels specified in Table 2. These
solids are best provided by whole or skim milk
Table 2. Representative compositions for ice cream, ice milk, and
sherbet (wt %) concentrated to 25 – 35 wt % solids, or by skim
milk powder (→Milk and Dairy Products). In the
Ingredient Regular Super Ice milk Sherbet
ice premium
United States, 25 % of the nonfat milk solids in
cream ice cream ice cream can be supplied by whey, a cheese in-
dustry byproduct (→Cheese, Processed Cheese,
Milk fat 10.0 16.5 4.0 1.5
Nonfat milk and Whey, Chap. 7.2.). Since whey solids con-
solids 11.5 9.0 13.0 3.0 tain only 13 % protein compared to 35 % in skim
Sucrose 10.2 15.5 9.0 23.0 milk solids, ice cream composition is altered
Corn sweetener
solids 6.8 9.0 7.0
somewhat when whey is used as a partial sub-
Stabilizer/ stitute for skim milk. However, if whey is com-
emulsifier 0.3 0.2 0.4 0.3 bined with commercially available casein, the
Total solids 38.8 41.2 35.4 34.8 composition of skim milk is duplicated quite
closely.

6.1. Fats 6.3. Sweetener Solids

Fat is primarily responsible for the creaminess
and mellowing qualities of ice cream products.
It also depresses the intensity of cold sensations
experienced upon consuming a partially frozen
product. An important role for fat in establish-
ing the physical matrix has been referred to in
Chapter 5.
Fresh cream (milk and milk products) pro-
vides most of the fat for ice cream in the United
States. It is readily available and, being a liquid,
has handling advantages. Cream is recognized
best from the standpoint of flavor. For economic
reasons, butter is more commonly used in Eu-
rope; where allowed, a vegetable fat can also be
used, either alone or in combination with butter
or cream. Preferred are partially hydrogenated
fats of the lauric type from coconut and palm
kernel.
6.2. Nonfat Milk Solids
The nonfat solids of milk consist of 35 % pro-
tein, 55 % lactose, and 10 % mineral ash. The
protein fraction is particularly important for ice
cream. Protein coats fat particles and stabilizes
Ice cream is a sweet-tasting food, but the pre-
ferred sweetness intensity varies among markets
around the world. In the United States, the sweet-
ness equivalency of a 15 % sucrose solution is
preferred. Sweetness in ice cream is provided by
soluble carbohydrates which themselves vary in
sweetness. Lactose has a fifth and 36 DE corn
sweetener solids half the sweetness intensity of
sucrose, but solids containing fructose may be
even sweeter than sucrose.
Of importance equal to sweetness control is
the dominant role of the soluble carbohydrates
in determining the colligative freezing proper-
ties of the mix. The water – ice equilibrium at
any temperature mostly reflects the combined
concentrations of these solutes.
Depending on the combination of carbohy-
drate ingredients, total mix solids may be in-
creased or decreased to influence texture and
body characteristics. In this regard, the higher
saccharides of low-conversion corn syrups have
a favorable impact [8].
Sucrose from sugar cane and sugar beet has
been the sweetener of choice for ice cream.
Primarily for economic reasons, syrups of hy-
drolyzed corn starch now partially replace su-
6 Ice Cream and Frozen Desserts
crose in most frozen dessert products. Corn
syrups of the low conversion 36 DE type com-
pensate to a degree for the gradual reduction over
many years in the milk solids content of econ-
omy and regular ice cream brands. Since their
introduction in the 1970s, fructose syrups ob-
tained by hydrolysis of starch syrup to glucose
and enzyme isomerization have gained a strong
foothold. Conventional corn syrups are not as
sweet as sucrose, but fructose syrups are equally
sweet or sweeter. Although they are the most
economical source of sweetness, fructose syrups
increase the molar concentration of solutes, thus
requiring lower freezing and storage tempera-
tures to maintain proper hardness in packaged
products. For a detailed description of syrups,
see →Glucose and Glucose-Containing Syrups.
6.4. Stabilizers and Emulsifiers
In most countries the addition of stabilizers and
emulsifiers to ice cream is legally restricted.
Most ice cream products contain 0.1 – 0.2 wt %
hydrocolloid stabilizer which functions as a wa-
ter control agent and increases the viscosity of
the unfrozen aqueous phase. By mechanisms
which are poorly understood, the hydrocolloids
delay growth of large ice crystals and prevent de-
terioration of quality during storage and distri-
bution. Emulsifiers, used at the same or slightly
higher levels than hydrocolloids, serve as a man-
ufacturing aid by helping to control emulsion
changes and lead to good extrusion properties
for ice cream from the freezing cylinder. For a
description of emulsifiers, see →Foods, 3. Food
Additives, Chap. 3.6.
A large number of hydrocolloids function in
ice cream systems; guar gum, locust bean gum,
cellulose gum, and carrageenan are used most
frequently. Most are used in combination as sta-
bilizer blends, selection being based on the rhe-
ological properties desired by a particular ice
cream manufacturer. Generally, carrageenan is
a component of stabilizer blends, e.g., with guar
or locust bean gum at a ratio of 1 : 10 [9] or at
a concentration of 0.018 wt % [10]. It functions
as a manufacturing aid to prevent separation of
mix into layers during quiescent storage prior to
freezing into ice cream.
Monoglycerides in mixture with diglyc-
erides, and polysorbate-80 have a long history
of use as emulsifiers for ice cream. The polysor-
bates are especially effective in hastening con-
trolled emulsion destabilization leading to desir-
able extrusion properties for ice cream [11].
7. Manufacture of Mix
Preparation of a mix for freezing into ice cream
is a multistep process involving the assembly of
the mix ingredients, pasteurization of the result-
ing fluid mix, homogenization to stabilize the
emulsion, and finally the cooling and storage of
the finished mix to await freezing. Equipment
can be sized to meet production needs. Mix out-
put varies from 2000 L/h for a small ice cream
factory to 20 000 L/h for a high-volume oper-
ation. Hygiene plays an important role in the
manufacture and treatment of ice cream. The hy-
gienic preparation of ice cream requires not only
pasteurization, but also regular cleaning and dis-
infection of equipment, proper packing and stor-
age, i.e., Good Manufacturing Practice. Pasteur-
ization is not compulsory in all countries, e.g., in
the Federal Republic of Germany it is only nec-
essary for the types “Eiskrem” (the main type
of industrially produced edible ice) and “Ein-
facheiskrem”. The microbiological state of ice
cream is subject to very stringent regulations in
most countries, e.g., total microbial count, co-
liforms, and pathogenic microorganisms (espe-
cially Salmonella).
Pasteurization. The primary goal of pas-
teurization is the guaranteed destruction of
pathogenic microorganisms. In the process, al-
most all nonpathogens are simultaneously de-
stroyed, but the mix is not sterile and must be
stored at 4 ◦ C to prevent multiplication of sur-
viving microorganisms. Pasteurization regimes
are defined and strictly controlled by federal and
state regulatory agencies.
In batch pasteurization, all ingredients in
the desired proportions are fed into a jacketed
vat, heated for 10 – 30 min under agitation to
68.3 ◦ C, and held at this minimum pasteuriza-
tion temperature for at least 30 min. The vat con-
tents must then be pumped through the homog-
enizer before a new batch can be started.
Continuous high temperature – short time
(HTST) pasteurization is more energy efficient
than batch pasteurization and better suited to

high-volume plants. Mix ingredients are assem-
bled in a vat similar to the batch process. How-
ever, the mix is pasteurized by pumping it from
the vat as a thin film through a plate heat ex-
changer. Thin film heating is almost instanta-
neous and continuous HTST pasteurization is
the preferred approach to pasteurization in the
ice cream industry. Temperature – time condi-
tions for HTST pasteurization are 78 – 80 ◦ C
for 20 – 40 s (US minimum requirements are
79.4 ◦ C for 25 s).
The times and temperatures for batch and
HTST pasteurization are acceptable minimum
combinations. Often, a higher temperature or
longer hold is employed to provide added pro-
tection relative to microbial kill. Furthermore, a
significant number of manufacturers believe that
an extra measure of heat energy, especially at
elevated temperature, improves ice cream body
and texture.
Homogenization. Following batch pasteur-
ization or as an integral part of the HTST
process, mix passes through a homogenizer.
Homogenizers are high-pressure pumps which
force mix through a restricted valve under de-
fined conditions of turbulence and shear. During
homogenization, fat globules are broken up to
less than 2 µm in diameter; the resulting small
globules become coated with milk protein, emul-
sifier, and other surface-active constituents. The
temperature of the mix should be at least 50 ◦ C
to ensure that the fat is completely melted.
Storage of Mix. The prepared mix is held
at 4 ◦ C or less in a storage tank until it can be
frozen into ice cream. In the United States, freez-
ing is usually carried out the same day the mix is
made, with only a 2 – 3 h hold in the storage tank
to allow complete crystallization of fat globules.
The former practice of holding the cold mix for
at least 6 h (usually overnight) before freezing
commenced has been been superseded by the
use of hydrocolloid stabilizers.
8. Freezing Mix to Produce Ice
Cream
Mix water begins to freeze at −2.5 ◦ C, the exact
temperature being a function of the solution con-
Ice Cream and Frozen Desserts 7
centration. Dissolved components in the aque-
ous phase that control this colligative property
are the lactose and soluble salts of milk, as well
as the carbohydrates supplied by sweetening and
flavoring ingredients. Measured and calculated
values for the freezing point depression are in
close agreement, at least in regard to the initial
freezing temperature [12]. As water freezes, the
solutes become more concentrated, thus requir-
ing even lower temperatures for the continuing
change of state of water to ice.
Consider a mix containing 15 wt % sucrose,
7 wt % lactose and milk salts (i.e., 22 wt % solu-
ble solids), and 38 wt % total solids (62 wt % wa-
ter). The ratio of soluble solids to unfrozen wa-
ter is 35 : 100 at the start of freezing. When soft
ice cream at −5.5 ◦ C leaves the swept-surface
cylinder of the freezer to be packaged, the ra-
tio will be 70 : 100 since 50 % of the mix water
will have crystallized as ice. At a typical stor-
age temperature near −20 ◦ C the ratio will be
275 : 100. Obviously, as temperature decreases
and water freezes, ice cream stiffens progres-
sively to a hardened mass.
8.1. Freezing as Mix Agitates
Ice cream freezers are of two types, batch and
continuous. In batch freezers, the mix is frozen
and whipped in single batches of 10 – 20 L de-
pending on machine size. The refrigerant (ei-
ther ammonia or Freon) in the jacketed rim of
the cylinder, absorbs heat from the mix causing
it to freeze as a thin film. The film is contin-
uously scraped off the cylinder wall by revolv-
ing scraper blades. Depending on conditions, 4 –
10 min is required to reach −4.4 ◦ C, which is a
typical discharge temperature for batch-frozen
ice cream. The proportion of mix water frozen
(ca. 40 %) at this temperature results in a vis-
cosity or thickness for the whipping mass that
is ideal for incorporation and retention of air as
overrun. If the mix is either too highly frozen or
not frozen enough, air cannot be retained within
the matrix.
With continuous freezers, mix is pumped
through the freezing cylinder and exits
semifrozen in a continuous stream. Because the
cylinder barrel is pressurized, product stiffness
is not nearly as important for air incorporation as
in batch freezing. Consequently, ice cream flow-
8 Ice Cream and Frozen Desserts
ing from a continuous freezer can be at −6 ◦ C
or even lower, which means more extensive nu-
cleation during the fast freezing stage of ice
cream manufacture. This produces smaller ice
crystals and a smoother texture. Because con-
tinuous freezers use a pumping system, they can
be adapted for molding ice cream by extrusion,
for incorporating variegating syrups and a great
variety of particulate flavor ingredients, and the
automatic filling of packages. Several sizes of
continuous freezers allow mix pumping rates
between 200 and 2000 L/h over an almost infi-
nite range of overruns. Because of their many
advantages, continuous freezers are preferred
over batch machines, except in the smallest ice
cream factories.
8.2. Quiescent Hardening
The final stage of ice cream manufacture in-
volves further freezing to convert packages of
soft ice cream to a hardened form. Hardening
takes place in either a hardening room or tun-
nel. Ideally in a hardening room, currents of air
at −30 ◦ C sweep across all sides of a carton or
package. For this to occur, the packages must
be kept separated on shelves until core temper-
atures are well below −20 ◦ C, this may require
several hours depending on package size. Only
then can the packages be bundled and placed
on pallets for storage until shipment from the
factory. In tunnel hardening, packages move
through the tunnel on a conveyor in −50 ◦ C high
velocity air. Hardening is much faster than in
a hardening room, and upon exiting the tunnel
packages are ready to be bundled and placed on
Illites → Clays
pallets for storage. The main advantages of tun-
nel hardening are efficiencies of time, labor, and
space utilization.
9. References
1. Codex Standard for Edible Ices and Ice Mixes
137, 1981; IDF Bulletin 172 (1984) Appendix
II.
2. Code of Federal Regulations 21, Food and
Drugs Parts 110 – 199, U.S. Government
Printing Office, Washington, D.C., 1987,
pp. 228 – 236.
3. The Latest Scoop, International Ice Cream
Association, Washington, D.C., 1987.
4. K. C. Berger in S. Friberg (ed.): Food
Emulsions, vol. 5, Dekker, New York 1976,
pp. 141 – 213.
5. P. G. Keeney, J. A. Maga, J. Dairy Sci. 48
(1965) 1591 – 1596.
6. S. Friberg in S. Friberg (ed.): Food Emulsions,
vol. 5, Dekker, New York 1976, pp. 1 – 37.
7. P. G. Keeney: Commercial Ice Cream and
Other Frozen Desserts, Extension Circular
553, The Pennsylvania State University,
University Park, Pa., 1972, pp. 1 – 50.
8. P. G. Keeney, D. V. Josephson, Ice Cream
Trade J. 57 (1961) no. 5, 28 – 34.
9. K. A. Hyde, J. Rothwell: Ice Cream, Churchill
Livingstone, Edinburgh 1973.
10. P. G. Keeney, M. Kroger in B. H. Webb, A. H.
Johnson, J. A. Alford (eds): Fundamentals of
Dairy Chemistry, AVI Publishing 6;
Washington 1974.
11. P. G. Keeney, Food Technol. (Chicago) 36
(1982) no. 11, 65 – 70.
12. W. Arbuckle: Ice Cream, 4th ed., AVI
Publishing Co., Westport, CN, 1986, p. 46.
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Imaging Technology Save title to My Profile
Standard Article Topics
Email this page
Detlef Winkelmann1, Manfred Lutz2, Damodar M. Pai3, Andrew R. Melnyk4, Print this page
Richard Hann5, Walter Crooks6, Keith S. Pennington7, Francis C. Lee8, C. SEARCH THIS TITLE
Wayne Jaeger9, Don R. Titterington10, Walter Lutz11, Arno Bräuninger12,
Luc De Brabandere13, Frans Claes14, Rene De Keyzer15,
Wilhelmus Janssens16, Johan Verelst17, Werner Frass18, Thomas Telser19, Advanced Product Search
Horst Hoffmann20, Bernd Bronstert21, Karl-August Springstein22, Rod Potts23, Search All Content
Hartmut Steppan24, Donald C. Mammato25, Thomas Stoudt26, Michael C. Acronym Finder
P. Watts27, David Allen28
1Hoechst Aktiengesellschaft, Werk Kalle-Albert, Wiesbaden,
Federal Republic of Germany
2AEG Elektrofotografie GmbH, Warstein, Federal Republic of
Germany
3Xerox Corporation, Joseph C. Wilson Center for Technology,
Rochester, New York 14644, United States
4Xerox Corporation, Joseph C. Wilson Center for Technology,
Rochester, New York 14644, United States
5ICI Imagedata, Brantham Industrial Estate, Manningtree, United
Kingdom
6IBM, San Jose, California 95193, United States
7IBM, San Jose, California 95193, United States
8IBM, San Jose, California 95193, United States
9Xerox, Wilsonville, Oregon 97070, United States
10Xerox, Wilsonville, Oregon 97070, United States
11Kalle ReproMedia, Werk Kalle-Albert, Wiesbaden, Federal
Republic of Germany
12Hoechst Aktiengesellschaft, Werk Kalle-Albert, Wiesbaden,
Federal Republic of Germany
13Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
14Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
15Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
16Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
17Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
18Hoechst Aktiengesellschaft, Werk Kalle-Albert, Wiesbaden,
Federal Republic of Germany
19BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of
Germany
20BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of
Germany
21BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of
Germany
22Hamburg, Federal Republic of Germany
23Agfa-Gevaert N.V., Mortsel (Antwerpen), Belgium
24Hoechst Aktiengesellschaft, Werk Kalle-Albert, Wiesbaden,
Federal Republic of Germany
25Hoechst Celanese Corporation, Somerville, New Jersey 08876,
United States
26Hoechst Celanese Corporation, Somerville, New Jersey 08876,
United States
27Hoechst Celanese Corporation, Somerville, New Jersey 08876,
United States
28College of Manufacturing, Cranfield Institute of Technology,
Cranfield, United Kingdom

Copyright © 2003 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved.
DOI: 10.1002/14356007.a13_571.pub2
Article Online Posting Date: March 15, 2003

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Introduction

page 1 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2. Copying and Nonimpact Printing Processes
2.1. Office Copying and Printing
2.1.1. Electrophotography
2.1.1.1. Photoreceptor
2.1.1.2. Corotron Charging
2.1.1.3. Light Exposure
2.1.1.4. Image Development
2.1.1.5. Image Transfer and Fusing
2.1.1.6. Photoreceptor Cleaning and Erase
2.1.1.7. Other Electrophotographic Imaging Systems
2.1.2. Thermographic Printing
2.1.2.1. Thermal Print Transducers
2.1.2.2. Resistive Ribbon Technology
2.1.2.3. Imaging Materials and Mechanisms
2.1.2.4. Comparison of Thermal Technologies
2.1.3. Ink-Jet Printing
2.1.3.1. Continuous Ink Jet
2.1.3.2. Impulse (Drop-on-Demand) Ink Jet
2.1.3.3. Ink-Jet Nozzle Orifice
2.1.3.4. Ink-Jet Inks
2.1.3.5. Colorants Used in Ink Jet Inks
2.1.3.6. Color Printing with Ink Jets
2.2. Technical Copying
2.2.1. Diazotyping
2.2.1.1. Photolysis
2.2.1.2. Coupling Reaction
2.2.1.3. Production of Diazotyping Material
2.2.1.4. Processing
2.2.2. Other Photochemical Systems
2.2.3. Silver Process
2.2.3.1. Black and White Copying Materials Based on Silver Diffusion Transfer Technology
2.2.3.2. Modern Lithographic Printing Applications
2.2.3.3. Color-Copying Materials Based on Dye Diffusion Chemistry
2.3. Microfilms and Microfiches
2.3.1. Diazotyping
2.3.2. Vesicular Film
3. Imaging in Graphic Arts
3.1. Graphic Arts Photography
3.1.1. Classes of Photomaterials Used in Prepress Production
3.1.2. Flow Scheme of Graphic Prepress Process
3.2. Basics of the Use of Light-Sensitive Nonsilver Materials in the Graphic Arts
3.2.1. Application
3.2.2. Chemical Fundamentals
3.2.2.1. Photosolubilising Systems
3.2.2.2. Photoinsolubilizing (Photocuring) Systems
3.2.2.3. Photophysical Principles of Image Generation
3.3. Color Proofing
3.3.1. Basics
3.3.2. Applications
3.3.3. Overlay Systems
3.3.4. Single-Sheet Systems
3.3.4.1. Precolored Systems
3.3.4.2. Systems With Colors Generated During Processing
3.3.5. Digital Proofing
3.4. Platemaking
3.4.1. Lithography (Planographic Printing, Offset)
3.4.1.1. Substrates
3.4.1.2. Coating and Further Processing
3.4.1.3. Processing of Presensitized Plates

page 2 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.4.2. Letterpress and Flexography
3.4.2.1. Structure of Photopolymeric Letterpress and Flexographic Plates
3.4.2.2. Production and Requirements of Photopolymer Letterpress and Flexographic Plates
3.4.2.3. Chemistry of Photopolymer Relief Printing Plates
3.4.2.4. Future Developments
3.4.3. Gravure Printing
3.4.3.1. Conventional Method of Gravure Platemaking
3.4.3.2. Electromechanical Gravure Platemaking
3.4.3.3. Electron-Beam and Laser Gravure
3.4.4. Screen Printing
3.4.4.1. The Screen
3.4.4.2. The Stencil
4. Imaging for Electronics
4.1. Photoresists
4.1.1. Industrial Applications of Photoresists
4.1.2. Function and Chemistry of Photoresists
4.1.2.1. Positive Photoresists
4.1.2.2. Negative Photoresists
4.1.3. Testing Methods
4.1.4. Economic Aspects and Suppliers of Photoresists
4.1.5. Occupational Health and Environmental Protection
4.2. Printed Circuits (Printed Circuit Boards)
4.2.1. Methods of Producing Printed Circuit Boards
4.2.2. Masking Techniques
4.2.2.1. Exposure Masks
4.2.2.2. Screen Printing
4.2.2.3. Photoprinting
4.2.2.4. Other Techniques
4.2.3. Economic Aspects
4.2.4. Occupational Health and Environmental Protection
4.3. Microelectronic Devices
4.3.1. Integrated Circuit Manufacturing
4.3.2. Photoprocess
4.3.3. Techniques for Submicron Lithography
5. Photochemical Machining
5.1. Introduction
5.2. Artwork Generation and Phototool Production
5.3. Materials
5.4. Photoresist Systems
5.5. Etching Technology
5.6. Process Capability
5.7. Economic Aspects
5.8. Products

[Top of Page]

1. Introduction
Detlef Winkelmann and Manfred Lutz

The roots of modern imaging go back to the invention of photography in the middle of the 19th century. In those early days
photography was used primarily in portraiture, illustration, and documentary imaging. In the first half of the 20th century new
applications of imaging technology emerged, which altogether had the main purpose to store information for later use or for
distribution. Examples include microfilm recording, prepress printing processes, reprography, or office copying.

The emergence of solid state electronics in the middle of the 20th century had an impact on imaging in two ways. The first
one was that now images could be recorded, transmitted, or distributed electronically. Fax machines and wireless
transmitting of photographs became the first applications of this new technology.

Secondly, imaging techniques were used to produce for example integrated circuits, and micromechanical or microelectronic
components. In consequence imaging techniques created the basis for today's imaging and communication technology which
may probably end up in the course of the 21st century in the cross-media concept. This will convert information such as text,
images, photographs, in combination with animation or interactive components into a “document” that can be produced as

page 3 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
either a printed report, digital hardware in form of a CD-ROM or as an information platform in the internet.

In spite of this vision still a wide variety of imaging techniques exists and will stay for a while in application in analog, hybrid,
or digital systems. But mere analog imaging systems will not see any more growth in the market.

In principle, two main fields for imaging exist today:

1. The production of permanent, visually perceptible two-dimensional information,

2. The production of goods or devices as printed circuits boards, microelectronic, and micromechanical devices as well
as photochemical machining.

For both fields sophisticated materials and processes have been developed in the past for analog image taking, processing,
and output. With the new digital systems image taking and processing will be handled by computer systems and only the
processing software and the output hardware remain unique for the different fields of imaging. This means for example that
output on paper will still require another technique than the writing process for a mask to be used for the manufacturing of
microelectronic devices.

Copying became very popular and widespread in the office environment with the invention of the electrophotographic
process which easily could duplicate a document for further distribution. Now with the availability of highly integrated circuits,
documents are digitized by scanners and a hard copy is prepared by a printer using one of the nonimpact output
technologies as electrophotography, ink jet, thermography, and electro- or magnetostatics. Such a system could work either
as a digital copier, a fax machine, or a computer printer.

Furthermore the availability of high resolution scanners and printers both in black and white and in color enables the
replacement of analog processes in other areas than office communication. Examples are imaging in graphic arts, wide
format copying and reprography, color proofing, plate making, and offset printing.

It is only a matter of time that analog diazotyping — still used in wide format copying and reprography — will be replaced
completely by digital copiers or plotters using electrophotograhic, electrostatic, or ink jet processes, since drawings and
layouts are nowadays almost all designed on CAD systems and are digitized anyway.

Print shops are still using to a very large extent lithographic materials, processes, and machinery. But printing industry talks
about computer to plate, computer to press, and computer to paper.

All three of them include digital processing of information during the prepress steps but require special output technologies.

Computer to plate needs a photopolymer sensitive to the light of a raster image systems, mostly laser exposure, to change
the physicochemical behavior necessary for the printing process.

Computer to press means there is no printing plate. The print master is directly generated on a sleeve or a cylinder in the
printing press by converting light or temperature sensitive materials into durable print masters and washing away the
unexposed areas. When the print job is finished the sleeve or the cylinder are cleaned and can be prepared for a new job.
Besides the savings in materials and set up time, there are no plates to be stored. For a reprint the digital file has to be
reloaded only.

Finally, computer to paper is the fastest and most convenient way to print small editions. Such systems employ one of the
nonimpact printing techniques as electrophotography, ink jet, electrostatics, or magnetostatics, depending one the market
segment they target.

In the field of imaging methods for the manufacture of integrated circuits, and micromechanical or microelectronic
components again digital image processing is used to design the masks for the following process steps.

In the course of time efficient photoresists, exposure systems, and process units have been developed for the various
applications.

To achieve smaller size of e.g., electronic devices and therefore a higher integration is the driving force for the submicron
lithography. Two lithographic systems have to be distinguished: One generates the mask — today mostly an electron beam
system — and the other one structures the wafer. Current 256 MBit DRAMs (dynamic random-access memory) with 250 nm
structures are exposed with a KrF laser ( = 248 nm). Future 4 and 16 GBit DRAMs require shorter wavelength lasers or
other technologies as UV, X-Ray, electron beam, or ion lithography. These are expected to be capable to create structures
smaller than 100 nm corresponding to 64 GBit-DRAMS and even more within the next decade.

If this target can be reached it would have again an impact on imaging technologies since more memory and faster
processors will be available for image processing.

[Top of Page]

2. Copying and Nonimpact Printing Processes

Manfred Lutz, Damodar M. Pai, Andrew R. Melnyk, Richard Hann, Walter Crooks, Keith S. Pennington, Francis C. Lee, C. Wayne Jaeger, Don
R. Titterington, Walter Lutz, Arno Bräuninger, Luc De Brabandere, Frans Claes, Rene De Keyzer, Wilhelmus Janssens, Rod Potts

2.1. Office Copying and Printing

Since the invention of xerography by CHESTER CARLSON in 1938, a variety of imaging systems have been developed into

page 4 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
commercial products [1], [2] for the purpose of duplicating analog information — like in office copiers, or plain paper fax
machines — or for the purpose of distribution and remote printing of originally analog, or digital information.

As systems components analog copiers are provided with optics to image the original onto the process unit in order to
produce the copy (Fig. 1). Printers, however, need a character generator and raster or scanning optics and sufficient memory
for the intermediate data storage (Fig. 2). For digital copiers and plain paper fax machines a scanner unit has to be included
into the system to digitize the initial analog original (Fig. 3).

Figure 1. Components of modern electrophotographic systems — Copier

Figure 2. Components of modern electrophotographic systems — Printer

Figure 3. Components of modern electrophotographic systems — Fax machine or digital copier

Together they all need a process unit to produce the copy or printout finally. Electrophotography is the technology used in the
process unit in virtually all copiers commercially available today and is also the most prevalent printer technology.

2.1.1. Electrophotography
In electrophotography, a charge pattern replicating the light image is formed on a photoconducting film.

Charged, pigmented, thermoplastic particles, called toner, are selectively attracted to the charge pattern, thereby making it
visible. The toner image is then transferred to the paper, and fixed by softening and fusing the toner to the paper. This
imaging process is called the Carlson or xerographic process.

Other imaging principles in electrophotography include Electrofax, thermoplastic and overcoated xerography, ionography,
and magnetography [1-7]. The major portion of the following section is devoted to xerography; the other forms of
electrophotographic imaging are only briefly described.

Since electronic imaging is developing very fast it is recommended to consult the web sites of IS&T (The Society for Imaging
Science and Technology) and SPIE (International Society for Optical Engineering) for updates on technical trends [4].

The six steps of xerographic imaging are (Fig. 4):

1. Charging of a high-resistivity photoreceptor in the dark.

2. Imagewise exposure of the charged photoreceptor; resulting in a charge pattern, the latent image, on the surface.
3. Developing the latent image by applying charged toner particles that come into contact with the charge patterns.
4. Transferring the toner image onto the paper or other media.
5. Fusing the toner to the media.
6. Cleaning and conditioning the photoreceptor for the next imaging cycle.

2.1.1.1. Photoreceptor
Description. The photoreceptor is the device or transducer (Fig. 4) that converts the light pattern into an electrostatic
pattern. At minimum, it consists of a photoconductive layer on a conductive substrate (Fig. 5). The essential steps of
photoreceptor operation—charging, charge generation, and charge transport—are also indicated in Figure 5. The first step is
the production of a uniform electric field across the photoconductor layer by corona charging. In the second step, exposure,
light photons incident on the photoreceptor in the pattern of the image are absorbed and generate conducting charge
(electron-hole pairs). Under the influence of the electric field produced by the surface charge, these drift across the
photoconductor, neutralize the surface charge, and produce the electrostatic image. The photogeneration and charge
transport steps may be performed in separate material layers (Fig. 5 B) that are optimized for their respective functions. In
addition to the charge generation and transport layers, other layers for adhesion or charge blocking (to minimize charge
injection into the photoconductor) may also be present.

page 5 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 4. Schematic of a xerographic reproduction machine, indicating the important process steps

a) Corona charging; b) Light exposure; c) Laser; d) Rotating polygon; e) Development; f) Transfer; g) Fixing; h) Cleaning;
i) Erase

Figure 5. Photoreceptor structures

A) Single-layer photoreceptor; B) Dual-layer photoreceptor

In the single-layer photoreceptor, photogeneration can take place at the top surface or in the bulk. In two-layer
photoreceptors, charge photogeneration takes place in the thin photogeneration layer which is located either at the bottom or
at the top.

The earliest photoreceptor (used in the Xerox model D) was a flat plate similar to a photographic plate, which required
manual handling. Drums allowed automatic machines of the type shown in Figure 4. Flexible photoreceptors were fabricated
initially as sheets of film (Electrofax ZnO) [5], later as scrolls, and more recently as belts. Belts enable full frame flash
exposure that increases the throughput speed of duplicators.

The essential characteristic of the photoreceptor is the photo-discharge curve, which relates light exposure to the voltage on
the photoreceptor (Fig. 6) [6]. The dependence of photogeneration efficiency on electric field generally determines the shape
of the photo-discharge curve [7], [8], which relates to the gray (density) scale reproduction characteristics of the xerographic
system [9].

Figure 6. Photoinduced discharge characteristics of typical photoconductors showing the potentials at the end of charging
step, dark decay and discharge curves

Photoreceptor Requirements. To be useful in xerography, a photoreceptor should meet the following criteria [10]:

1. The photoreceptor must hold the corona charge image in the dark. This requires trapping the surface charge and, in
some cases, a blocking contact between the conductive substrate and the photoconductor. Additionally, the rate of
charge loss in the dark (dark discharge) must be minimized.
2. In a copier application, the photoreceptor should discharge efficiently at all wavelengths of the visible spectrum. In
printer applications, the wavelength of the laser or exposure system is fundamental, for example, 800 nm for gallium
arsenide lasers. The figure of merit is the efficiency of photogeneration, which is generally less than unity and
decreases with decreasing electric field.
3. The photogenerated charge carriers must traverse the photoreceptor in times that are short compared to the image
development time. This sets a lower limit to the charge carrier mobility (velocity per unit electric field).
4. The charged, partially discharged, and residual voltages must remain stable with repetitive cycling during a multiple
copy run. A condition termed cycle-up results from a buildup of residual voltage caused by accumulation of trapped
charge through repetitive use. A condition termed cycle-down results from increased dark discharge with repetitive
use. Charge trapping is caused by certain impurities, which necessitates stringent purity requirements (<1013 impurity
atoms per cubic centimeter).
5. Photoreceptor materials must be capable of being fabricated into large-area, defect-free films. Materials must be
sufficiently stable to perform in a corona environment containing ozone, nitrogen oxides, and other effluents, as well as
to withstand wear by the development and cleaning processes.

Brief Survey of Photoreceptors.

Inorganic Photoreceptors. Amorphous selenium (a-Se) photoreceptors, consisting of vacuum-deposited films of amorphous
selenium 20 – 100 µm thick, on aluminum drums, were the most widely used photoreceptors in xerography during the 1970s
and 1980s. Amorphous selenium meets most of the photoreceptor requirements, except sensitivity throughout the visible
spectrum; it is highly sensitive to blue light only (Fig. 7). Addition of tellurium to a-Se increases the red sensitivity [11], and
multilayer devices consisting of a thick a-Se charge transport layer overcoated with a thin, red-sensitive, selenium-tellurium
layer are used commonly as photoreceptors [12-15]. Another weakness of a-Se is that it crystallizes readily at elevated
temperature, and crystallized Se does not accept charge. Arsenic is added to a-Se to retard the rate of crystallization and
increase red sensitivity [16]. Photoreceptors containing 40 atom % arsenic and 60 atom % selenium have high sensitivity in
the entire visible spectrum and are employed in high-speed printers with helium – neon lasers or LED image bars (Fig. 7).

page 6 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 7. Relative spectral photosensitivities of inorganic photoconductors

a) a-Selenium; b) a-Selenium with 10 wt % Te; c) a-Arsenic triselenide; d) a-Silicon

A more recent addition to the family of inorganic photoreceptors is hydrogenated amorphous silicon (a-Si: H) [17-20] which is
sensitive in the entire visible spectrum and has sufficient 780 nm sensitivity to be used with gallium arsenide lasers (Fig. 7). It
is vacuum-coated by chemical vapor decomposition of silane (SiH4). Like crystalline silicon, a-Si: H can be made positively or
negatively conducting by doping with boron or phosphorus. Other inorganic photoreceptor materials applied in xerography
are solvent-coated layers of zinc oxide [5], [21] and cadmium sulfide [22], [23] in an organic polymer binder.

Organic Photoreceptors. When CARLSON invented xerography, he employed sulfur and anthracene as photoreceptors, and
the initial commercialization of his idea relied on inorganic photoreceptors [1]. The current trend is toward the use of organic
photoreceptors because of their material variety, economy, and flexibility. High-speed copying machines use belts coated
with organic photoreceptors, whereas personal copiers employ aluminum drums dip-coated with multilayered organic
photoreceptors. The first organic photoreceptor consisted of a single-layer charge-transfer complex of an electron-donor
polymer (polyvinylcarbazole) with an electron-acceptor (2,4,7-trinitrofluorenone) [24-26]. This charge-transfer complex
absorbs visible light and enables charge transport of both electrons and holes. The photogeneration efficiency of this
photoreceptor is relatively low and highly field dependent [27].

To enhance photosensitivity of organic photoreceptors dual-layer structures are employed, as with selenium – tellurium
photoreceptors [28], [29]. Hole transporting polyvinylcarbazole is replaced by other electron-donor molecules in solid
solutions of mechanically durable polymers such as the polycarbonates of bisphenol A or Z. Typical electron-donor
molecules include triphenylamines and diamines [30-32], pyrazolines [33], hydrazones [34], [35], oxadiazoles [36], [37], and
stilbenes [38]. Transport layers generally consist of a 40 – 50 wt % solution of the electron-donor compound, 10 – 25 µm
thick, and have hole mobilities of 10–6 – 10–5 cm2 V–1 s–1. Because these organic charge transport layers are transparent,
the photogeneration layer is overcoated with the transport layer and the photoreceptor is charged negatively. The
photogeneration layers consist of pigments such as phthalocyanines [39], [40], bisazocompounds [33], [41], thiapyrylium
salts [42], [43], perylenes [36], [44], and many other pigments. The relative spectral sensitivity of several of these is shown in
Figures 8 and 9. A very comprehensive survey on photoconductivity and carrier transport phenomena of both inorganic and
organic materials can be found in [36], [46]. More recent studies and papers are summarized in [47] and [48].

Figure 8. Relative spectral photosensitivities of organic photoconductors for printers

a) Metal-free phthalocyanine; b) Titanyl phthalocyanine

Figure 9. Relative spectral photosensitivities of organic photoconductors for copiers

a) Dibromoanthanthrone; b) and c) Bisazo pigments

2.1.1.2. Corotron Charging

The first step in the operation of a photoreceptor is to charge it to a (uniform) voltage by depositing charge on the free
surface of the photoreceptor with a device called a corotron (Fig. 10 A). The simplest and most common corotron design
consists of a thin wire (50 – 150-µm diameter) stretched between two insulating supports and electrically connected to a
high-voltage supply [1], [2]. The wire is surrounded on three sides by a conducting shield spaced ca. 1 cm from the wire, as
shown in Figure 10 A. A voltage of 3 – 8 kV on the wire produces a high electric field that ionizes the surrounding air, and
ions are deposited on the photoreceptor surface, thereby charging it.

Figure 10. Schematic of a simple corotron (A) and a scorotron (B)

With a positively biased wire, electrons and negative ions in the surrounding air are drawn to the wire, causing secondary
ionization consisting of electrons and positive and negative ions in the vicinity of the wire. Positive ions, primarily hydrated

page 7 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
protons, (H2O)nH+ [49], are repelled from the wire and driven by the voltage difference to the shield and photoreceptor. With
a negatively biased wire, electrons are emitted from the wire and ionize the air. The electrons attach to molecules so that
anions are formed, predominantly carbonate ions, which are deposited on the photoreceptor [49].

The charging current flowing to the photoreceptor surface depends on the electric field and, hence, on photoreceptor voltage.
The charging current is inversely proportional to the voltage. At some high voltage, all current flow to the photoreceptor is cut
off. This current-voltage characteristic minimizes the voltage variation on the photoreceptor surface. It is exploited in a device
called a scorotron (from screened corotron; Fig. 10 B), in which a wire screen or metal mesh is placed between the corotron
and the photoreceptor. The cutoff voltage becomes the voltage applied to the screen, which provides a convenient way of
controlling the voltage on the photoreceptor.

Primary charge rollers (PCR) (Fig. 11) are used in many desk top printers and copiers to charge the organic photoreceptor.
The mechanism of the charging process is generally understood as ionization of air in the gap between the PCR and the
photoreceptor. The advantages are that contact charging realizes extremely low ozone emission, compactness, and low
charging voltage (below 1 kV) as compared with conventional charging devices, such as corotrons and scorotrons [50], [51].
There are two types of contact charge rollers. One consists of a steel sleeve with a conducting or semiconducting rubber
coating leaving a small gap between the roller and the photoreceptor [52]. Charging is controlled by the electrical discharge
in the vicinity of the gap. Stable operation and uniformity of surface potential across the photoreceptor is achieved by
superimposing an a.c. or pulsed voltage over the d.c. voltage.

Figure 11. Contact roller charging subsystem

The other one is a softer, more brush type roller, which has a closer contact to the photoreceptor. Not only the electrical
discharge, but also a charge injection at the contact area, controls the charging. The amount of charge injection depends on
the effective size of the contact area and increases with the humidity in the ambient air.

2.1.1.3. Light Exposure

Light Exposure in Copiers. The model D, a manually operated copier, was the first commercial copier introduced by Xerox
and employed standard camera projection optics to image and photodischarge the photoreceptor. This type of light exposure
optics is still used in flash exposure systems of high-speed duplicators. The full image is projected on a flat area of a belt
photoreceptor with a xenon flash lamp to freeze the image on the moving belt.

Copiers with drum photoreceptors require scanning optics that project a moving narrow strip of the original image through a
slit on, and synchronously with, the moving photoreceptor surface. Many techniques have been used to accomplish this. One
is to project the original image through a lens on a rotating mirror that scans the image across a slit synchronously with
photoreceptor motion. Another involves scanning the original image with moving lamps and mirrors, thereby projecting the
image through folded optics that maintain the focus of the original on the photoreceptor through a slit. In small, personal
copiers, another approach is to move the original with rollers or a translating platen while the optics remain stationary. One
advantage of this system is the use of very low-cost fiber optics in a bar, called selfoc lenses, in place of lenses and mirrors.
Early copiers, with a-Se photoreceptors, had fluorescent tube lamps in scanning optics and xenon lamps in full frame
exposures. Current copiers with scanning optics tend to have high-temperature incandescent tube lamps in place of
fluorescent lamps.

Light Exposure in Printers. Printers discharge the photoreceptor digitally with light spots that are on or off. Digital imaging is
performed in several ways. The majority of xerographic printers use a polygon laser flying spotlight exposure. The laser light
beam is projected onto a rotating polygon mirror which causes the beam to be swept repetitively across the photoreceptor.
By timing the sweeps to repeat after the photoreceptor has advanced the width of the beam spot, the surface of the
photoreceptor is raster scanned much like the electron beam in a television tube. The beam is turned on or off by an
optoacoustic modulator or by directly modulating a gallium arsenide diode laser. Other approaches used are image bar
arrays of individually addressable light-emitting diodes, or liquid crystal and magneto-optic light valves [53]. The two most
commonly used lasers, helium – neon and gallium – aluminum – arsenic, emit at 630 and 780 – 820 nm, respectively,
whereas the light-emitting diodes used in image bars emit at 660 – 720 nm. Hence, photoreceptors for the xerographic
printers require photosensitivities in the red to near-infrared spectrum.

2.1.1.4. Image Development

Electric Field of the Charge Image. After the charge and exposure steps, the image consists of a charge pattern across the
photoconductor. This produces an electric field pattern with field lines flowing from the surface charge to the countercharge
on the conductive substrate [54], [55]. Unless an electrode is closer than the thickness of the photoreceptor, the electric field
is very weak above a large charge area. With a distant electrode, field lines extend into the space outside the photoconductor
only at the edges of lines or solid patterns (Fig. 12) [56]. Because development occurs by attraction of charged toner particles
along the field lines, the toner is attracted only on the fringe of wide (solid) image areas. To develop solid image areas, a
development electrode must be introduced, as illustrated in Figure 12. Generally, the electrode is biased to a voltage higher
than the white image background potential to prevent toner development in the white areas [57].

page 8 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 12. Effect of a development electrode on electric fields and toner development

The original image, electrostatic image in cross section, electrostatic field strength, and developed image are illustrated

Development Materials. The charge image is rendered visible by the development step. A typical developer consists of a
black or colored powder called the toner which contains thermoplastic particles, typically 5 – 25 µm in size, with 5 – 10 %
pigment to give the desired color [1], [2], [58]. The function of the thermoplastic is to fuse and fix the image to the paper by
heat, pressure, or a combination of the two. Commonly employed thermoplastics are random copolymers of styrene with
methacrylates or acrylates. Black toner contains a pigment dispersion of carbon black, about 1 µm in size; color toners
contain mixtures of dyes and pigments of the required color.

The widely used, two-component development employs a mixture of a few weight percent of toner with larger size (ca. 100 –
700 µm) carrier beads. Carrier beads serve two functions; they provide (1) a method for mechanically transporting the fine
toner powders and (2) a means of charging the toner. Because the fine toner powders are difficult to control and their
leakage contaminates xerographic hardware, they are triboelectrically attached to the much larger and heavier carrier beads.
With agitation, charge is exchanged between the toner and the carrier bead, and the toner is charged by triboelectrification.
Typically, as many as 103 toner particles may be attached to a carrier bead.

In charged area development which is always employed in photocopying, the toner is charged to a polarity opposite that of
the photoreceptor, and the high-potential areas are developed black whereas partially discharged areas produce gray
densities. Printers can also employ discharge area development, in which the polarity of the toner is the same as that of the
photoreceptor and only low-voltage or discharged areas are developed.

Magnetic Brush Development. Magnetic brush development is the most widely used toner development process. The
development device consists of a series of magnets inside a nonmagnetic shell [1], [2], [56]. As the photoreceptor moves into
the development zone, toner-covered carrier beads are transported by the shell (Fig. 13) counterrotating to the
photoreceptor. The magnetic carrier beads tend to form brushlike filaments along the magnetic lines of force, as shown in
Figure 14, hence the name magnetic brush development. The toner is stripped from the carrier beads by the electric field of
the charge pattern, and toner-depleted carrier beads are returned to the mixing region where they are covered with toner
particles. The carrier beads are coated with low surface energy materials, such as fluorinated polymers to optimize toner
charging, minimize permanent adhesion of the toner to the carrier, and control the conductivity of the toner – carrier mixture.
The effective developer – electrode spacing, which controls the field in large solid area development as discussed above,
can be reduced from the roll-to-photoreceptor distance by the conductivity of carrier beads. By varying the degree of polymer
coverage, the toner – carrier mixture can be made insulating (called insulating magnetic brush) or conducting [59]. The
former enhances fringe-field development which is useful for fine lines, while the latter enhances solid area development.

Figure 13. Schematic of a magnetic brush development system. The inserts schematically illustrate a developer brush fiber
and a developer bead with toner attached.

a) Toner dispenser; b) Rotating shell; c) Photoreceptor; d) Paddle mixer; e) Developer; f) Toner; g) Carrier

Figure 14. Magnetic brush developer unit. Toner is mixed and charged with carrier beads. Magnet arrays are stationary
while the cylinder around them rotates. The cylinder is biased to assist in developing large exposed areas.

Print quality, as determined by line resolution, gray density range and uniformity, and white background density (or
cleanliness), is controlled by several factors [6]. These include electrostatics (the shape of the photodischarge curve and the
image and cleaning voltage difference or contrast), the size and charge distribution of toner particles, the toner concentration,
and the developer flow rate. The copy darker – lighter buttons on office copiers adjust the electrostatics, and some modern
copiers maintain copy quality by sensing and adjusting the electrostatics. Toner must be added constantly to the developer
mixture to maintain a constant toner concentration. Too much leads to white background printout (dirt) and dark copies, and
too little leads to light copies. Most copiers have sensors that measure toner concentration or image density and control the
addition of toner. Although carrier beads are not used up, they must be replaced over regular intervals because constant use
causes surface damage.

Other Development Systems.

Cascade development is an older two-component development technology not in use today [56], [60], [61]. The mixture of

page 9 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
carrier beads and toner is tumbled (cascaded) over the photoreceptor surface by gravity. Because the carriers used for this
system are large, heavy beads of glass or steel, their size limits the developer electrode spacing to distances >500 µm,
making this system unsuitable for solid area densities.

Mixing the thermoplastic polymer with magnetic iron oxide makes the toner itself magnetic, eliminating the need for carrier
bead transport in magnetic brush development systems [62]. In this single-component development, the toner is charged
either by induction or by triboelectric charging against a donor roll. Single-component development enables smaller and
lighter developer transport hardware, used in personal copiers, because the carrier beads and mixing hardware take up most
of the volume and require heavy magnets. Because magnetic iron oxide is brownish red, it limits the ability to produce bright
color toners.

Aerosol development, also called powder cloud development, is one of the earliest forms of single-component development
[63], [64]. Toner particles are charged by blowing them through fine-bore tubes constructed from metal or ceramic. A
shortcoming of this process is that the toner charge distribution is broad and not easily controlled. This process is especially
suited to develop X-ray images in xeroradiography or mammography where continuous gray scale, high-resolution images
are required to reproduce slight differences in contrast potentials.

Electrophoretic or liquid development relies on charge exchange between micrometer-size toner particles and a liquid
medium (in which they are suspended) to charge the toner [1], [2], [65], [66]. The liquid may be an alkane such as kerosene
mineral oil, or isopar (a mixture of branched C8 – C14 isoparaffins) to which charge control agents are added. Liquid
development produces high-resolution images because the particle size is small and can approach the image very closely
[67]. A disadvantage is the delivery of an objectionable residual solvent onto the copy.

2.1.1.5. Image Transfer and Fusing

The developed image is transferred by placing a paper on the photoreceptor and applying a corona charge of the same
polarity as the photoreceptor charge on the back side of the paper [1]. The reversed electric field breaks the adhesion
between the toner and photoreceptor surface, transferring the toner particles to the paper. A biased roll can also be used for
transfer.

To fix the transferred toner image to the paper, four types of fusers are employed commercially: hot roll, cold pressure roll,
radiant, and flash fusers. Because rheological requirements for toners, such as melting point and melt viscosity, differ for the
four types of fusers, toner design must take into account the fuser type.

Hot roll fusing is the most common type of fusing. Here the paper with the toner image passes between two rollers, one of
which is heated, for example by an internal quartz lamp. To penetrate the paper fibers, the toner image must be heated to a
temperature substantially higher than the glass transition temperature of the toner polymer. Since the paper spends only 5 –
10 ms in the nip between the rollers, it is desirable that the glass transition temperature of the polymer be low to minimize
the requirements on the fuser heating power.

To prevent melted toner from sticking, the rollers are either permanently coated with a low surface energy material such as
Teflon or repeatedly coated with a release agent such as silicone oil [68]. Alternately, the release agent, usually a polyolefin
wax such as polyethylene or polypropylene, can be incorporated into the toner, eliminating the need for external application.
Image gloss, an important consideration for copy appearance, may be controlled by the fuser coating. A noncompliant
coating such as Teflon produces a high-gloss image, whereas a compliant coating such as Viton produces a low-gloss
image.

In cold roll fusing, toned paper passes between two polished steel rolls at room temperature, and the image fuses by
pressure alone. The main advantage of cold roll fusing is low power consumption, with no need for standby power. In radiant
fusing, the toned paper is passed under a heated coil or quartz lamp. The residence time required for radiant fusing is 200 –
500 ms longer than for roll fusing. Flash fusing is similar to radiant fusing except that the toner is fused by a high-intensity
light flash of approximately 5 ms duration. Finally, the toner can also be fused to the paper by exposing it to solvent vapors.

2.1.1.6. Photoreceptor Cleaning and Erase

Before resuming the cycle, the photoreceptor must be cleaned of residual charge and toner. Residual toner is removed by a
brush, wiper blade, or magnetic roll (similar to that used in development). Prior to cleaning, the toner may be charge-
neutralized with an a.c. corotron. The cleaning systems can have an auger or vacuum system that collects the removed
toner, and in some cases circulates it back into the developer system. Residual charge is usually removed by exposure to a
high-intensity lamp.

2.1.1.7. Other Electrophotographic Imaging Systems

Electrofax. Electrofax is the trade name of a xerographic system that uses a photosensitive paper on which the final image
appears, eliminating the transfer and cleaning steps [5], [20]. The photoconductor layer consists of a layer of silicone resin
binder, 5 – 15 µm thick, containing 50 wt % zinc oxide particles which give it a white color. Because zinc oxide absorbs only
UV light, it is dye-sensitized with, e.g., rose bengal, fluorescein, or methylene blue, and the appearance of the paper is made
as close as possible to plain paper by selecting an appropriate mixture of dyes to produce a light gray color.

To produce transparencies, direct formation of the final image on the photoreceptor has also been employed with transparent
organic photoconductors. By using liquid development and omitting the transfer step, very high image resolution can be
achieved. Another variant involves peeling the thin photoconductor layer, with the fused image, from the substrate and
laminating it to the paper [69].

Thermoplastic and Overcoated Xerography.

Thermoplastic xerography produces images by surface deformation of a thermoplastic. The photoreceptor can consist of an

page 10 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
organic thermoplastic photoconductor or a photoconductive layer coated with a thin layer of thermoplastic material [70], [71].
In the latter, the charge pattern, formed by conventional means (Fig. 15 A), remains across both layers in the dark regions
and across the thermoplastic layer in light-exposed areas (Fig. 15 B). By charging with an a.c. corotron to zero potential, the
charge is removed in the unexposed region but remains in the exposed region (Fig. 15 C). Upon heating, the electric field in
the exposed region causes the thermoplastic to deform into a wrinkling pattern (Fig. 15 D), and differential light scattering
renders the image visible.

Figure 15. Process steps of overcoated thermoplastic xerography

A) Thermoplastic layer; B) Transparent photoconducting layer; C) Conductive layer; D) Supporting layer

Overcoated Xerography. A variant of the xerographic process is to form the image on a photoconductor overcoated with an
insulating layer as described above, develop it with toner, and transfer it to paper by conventional means [2]. Several
variations of this charge image formation have been proposed [23], [72], and one, the Canon NP process, has been used
extensively in copiers throughout the 1970s and early 1980s. This process utilizes a photosentive element with a conductive
base, a photoconductive layer, e.g., cadmium sulfide in an organic polymer binder, and an insulating top layer. The steps
involved in forming an electrostatic image are shown in Figure 16. The difference to the Carlson process is that an additional
a.c. corona discharge and the imagewise exposure take place simultaneously. After a final overall exposure the electrostatic
image is formed on the insulator surface and can be processed by conventional xerographic methods.

Figure 16. Electrostatic image formation by the Canon NP process

Another related process, called TESI (transfer of electrostatic image), forms or transfers the electrostatic image on an
insulating layer that is separate from the photoconductor [2].

Ionography is a xerographic process in which the light exposure step is eliminated and the charged image pattern is created
directly on an insulating dielectric by ion deposition. An array of addressable charging pins or ion apertures is used to
electronically create the charge pattern (Fig. 17), which is then developed by liquid or dry toner.

Figure 17. Schematic cross section of an ion generation site in the print cartridge. The three electrodes, the drive or RF line,
the finger electrode, and the screen electrode, are separated by insulating layers.

TonerJet is a direct printing process where charged toner particles are deposited directly onto a plain paper surface to form
a visible image pattern. An array of electrodes creates dot-sized electrostatic fields to draw charged toner particles through
openings in a mesh, and deposit them in an image pattern on a plain paper surface.

A scheme of a TonerJet print zone is shown in Figure 18. Monocomponent magnetic toner is transported on a developer
sleeve that carries the toner into the print zone while charging each particle by contact charge exchange with the developer
sleeve material.

Figure 18. Simplified TonerJet print zone; inset: Etched circuit aperture array

A uniform electric field is created between a high potential on the back electrode below the paper and a low potential on the
developer sleeve. That uniform field is modified by control potentials on individual ring electrodes in a two-dimensional array
placed in the print zone. The dot-sized electrostatic fields draw the toner from the developer sleeve through apertures in the
ring electrodes, and deposit them on the paper surface where they are fused by pressure or heat [74].

page 11 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Magnetography is the magnetic analog of electrophotography. In this printing process, a magnetic latent image is created
on a film of ferromagnetic material, using an array of magnetic inductive heads. The latent image is then developed with
magnetic toner. A vertical magnetization, perpendicular to the ferromagnetic film, can be created by a pin-shaped magnetic
pole aligned perpendicular to the film (Fig. 19). The return path for the flux is through a magnetic sublayer under the film and
a large counter pole. The force attracting the toner containing soft magnetic material is proportional to the product of the
magnetic field and its gradient and is by an order of magnitude smaller than in the forces in an electrophotographic developer
unit [75], [76].

Figure 19. Energizing the recording head's coil creates a magnetic flux in the circuit composed of the head and drum cores.
Because the field under the flux-closing pole is below the coercivity of the recording layer, a dot is recorded only under the
higher-intensity recording pole

2.1.2. Thermographic Printing

In thermal transfer printing, heat generated from an electrically resistive source (the thermal transducer) works on a thermally
sensitive material (the dye or ink) to create a permanent image. The image must be created before heat is lost to a thermal
reservoir (the paper). The thermal transfer printing process is shown schematically in Figure 20.

Figure 20. Thermal transfer printing process

a) Thermal transducer; b) Thermally reactive dye or ink; c) Paper

Basically, two thermal transducer technologies exist. One uses a transducer of electrically resistive elements contained in
multilayer rigid head structures. In the more recently developed technology, the electrically resistive elements are part of a
ribbon structure [85]. The ribbon structure normally contains a thermoplastic image-forming material on one side. Heat
generated in the head or the ribbon transducer either melts a thermoplastic ink and transfers it to a receiving substrate. In
direct thermal transfer printing, heat generated in the head causes thermally sensitive materials on the paper to react,
forming a colored image. The technology using thermally sensitive paper is known as direct thermal transfer printing. Print
heads can consist of a small array of heating elements that print a matrix character (e.g., 5 horizontal × 7 vertical (Fig. 21 A));
or a simple vertical linear array (Fig. 21 B). The head may also consist of a page width of elements printing up to a line of
dots at a time (Fig. 21 C).

Figure 21. Thermal print head configurations

A) Matrix; B) Serial; C) Line

THE = thermal heat element

2.1.2.1. Thermal Print Transducers

Rigid thermal heads are generally subdivided into three types: (1) the transistor – resistor silicon mesa (i.e., plateau), (2) thin-
film resistive heads, and (3) thick-film resistive heads [86].

Silicon Mesa Technology. The basic structure of the silicon mesa (patented by Texas Instrument) is shown in Figure 22
[87].

Figure 22. Silicon mesa technology

a) Semiconductor chip; b) Mesas (operating temperature 175 – 200 °C)

The printing elements have a driver transistor – resistor pair deposited on the base of a silicon chip. The silicon chip is part of
a heater element array of air-isolated mesas. Each chip is epoxy-bonded to a ceramic substrate, which in turn is epoxy-
bonded to a metal heat sink. A small base current through the transistor controls a larger current in the resistor. Joule heating
of the resistor heats the silicon mesa. Mesa elements, when in contact with thermally sensitive paper, produce an image
corresponding to the elements addressed. Silicon mesas are coated with silicon dioxide, which provides thermal isolation and
prevents wear of the silicon. Operating temperature of the mesa elements is ca. 200 °C.

Thin-Film Technology. Thin-film thermal heads consist of a thin-film resistive heating element sandwiched between a
protective layer and a heat-insulating layer. The heat-insulating layer is bonded to a heat sink. A typical structure is shown in
Figure 23.

page 12 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 23. Structure of a thin-film thermal head [115], operating temperature ≈ 400 °C

a) Ta2O5-protective layer (5 µm); b) SiO2-protective layer (2 µm); c) Aluminum lead (2 µm); d) Ta2N resistor; e) Aluminum
substrate; f) Glass glaze (40 µm)

The actual material used in a thermal head varies with the manufacturer. Protective layers are chosen from materials such as
SiO2, Ta2O5 [88], [89], or SiC [89]. For reliability, the wear life of a protective layer must exceed the cycle life of the resistor
and other head parts.

The resistor elements are chosen from materials such as Ta2N [88], [94], Ta – SiC [90], Zr – N [91], Ta – SiO2, Cr – Si – O,
Si–Ta [93], Ni – Cr, and Ta – Al. Thermal insulating layers can be glass or Al2O3 [93], and conductors are usually NiCr – Au
[88] or Ni – Au [86].

All materials must be selected judiciously; similar coefficients of expansion in the various layers help prevent crack formation
at high operating temperature. When employed for direct thermal printing, protective layers must be resistant to abrasion by
the paper. The resistive layer must have a low temperature coefficient of resistance and good electrical stability under
operating conditions.

The Meander heat element is an alternate thin-film heater design [95] (Fig. 24).

Figure 24. Plane and sectional view of the meander heat element

a) NiCr – Au lead; b) Exothermic elements; c) Ta2N lead; d) Protective layer (Ta2O5); e) Protective layer (SiO2); f) Heat-
insulating layer (glaze layer); g) Alumina board

Meander elements are thicker than conventional rectangular thin-film elements and are considered more resistant to
oxidative degradation. They provide more uniform heating of the protective layers, resulting in more uniform image density.

Thin-film elements can provide up to 16 dots per millimeter for high-resolution printing. A print speed of 20 lines per second
can be achieved on thermal paper with a linear thermal head.

Head temperature varies from 250 to 550 °C depending on the duration of heating. Both parameters are determined by the
total energy required to create the printed image.

Thick-Film Technology. In thick-film resistive technology [86], a resistor paste (borosilicate glass, lanthanum glass –
ruthenium oxide) (DuPont 1211/1221, 1411/1421, 9145, Mathey-Bishop RRIOOR/RRIKOO [92]) is silk-screened onto a
ceramic substrate (Tektronix alumina, Tektronix fosterite [96]) to produce the individual print elements shown in Figure 25.
Another way to make the resistor is by laser machining a continuous bar of printed resistor paste [96]. Either way, the
elements are approximately 25 µm (1 mil) thick. The element conductors (Owen Illinois 99+ Au [96]) must be thin enough to
permit high resistor relief. However, their resistance must be low enough to accommodate the return path current, particularly
when the route is through a common current return path.

Figure 25. Print head for thick-film technology

a) Conductor; b) Resistor; c) Ceramic substrate

Current passing through thick-film resistors causes Joule heating while the head is in contact with the thermally sensitive
paper. This heat creates a permanent image. Resistance drift of the resistor elements represents greater threat of failure
than mechanical wear.

The required print head temperature is determined by the sensitivity of the thermal paper and the duration of heating. A peak
temperature of 450 °C can be reached.

2.1.2.2. Resistive Ribbon Technology

In resistive ribbon technology, the heating function is transferred from the head to a ribbon structure. A 40-electrode flexible
head selectively injects electrical current into a resistive ribbon [97]. The resistive ribbon is a 16-µm composite film of
polycarbonate imbedded with electrically conductive carbon black. It has a sheet resistance in the range of 500 – 900 Ω per
square [98]. On the side opposite the electrodes, the composite film is overcoated with a thin layer of aluminum (100 –
200 nm) applied by vacuum deposition. The electrically conductive aluminum assures that current flows anisotropically
rather than randomly through the carbon-filled polycarbonate layer. During the printing process, current flows from the
activated electrodes in the 40-stylus head through the conductive polycarbonate layer to the aluminum layer. Tunneling
conduction has been described as the mechanism responsible for electrical conduction in the carbon black doped composite

page 13 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
material [99]. Localized heat is generated by (1) contact resistance between the electrode and the carbon-filled polymer,
(2) bulk resistance of the carbon-filled polymer, and (3) interface resistance between the carbon-filled polycarbonate and the
aluminum layer. The natural formation of a 4 nm oxide layer on the aluminum increases this interface resistance [100] and
contributes to the overall Joule heating that occurs during a current pulse. A large return electrode minimizes Joule heating
as the current flows to ground. The resistive ribbon configuration is shown in Figure 26. As can be seen from the figure, the
100-nm aluminum layer is overcoated with a thermoplastic transfer layer. Heat generated in the resistive substrate melts the
thermoplastic layer, transferring it to the paper.

Figure 26. Resistive ribbon printing technology

a) Pulse generator; b) Print electrode; c) Carbon-loaded polycarbonate, conductive substrate, 16 µm; d) Grounded plate;
e) Aluminum layer, 100 nm; f) Transfer coating; g) Paper

The operating temperature in organic ribbon structures must be similar to that attained in rigid head structures but must not
affect the integrity of the resistive substrate. A modeling of ribbon temperatures with the 5796-PBH Advanced Statistical
Analysis Program (ASTAP) is shown in Figure 27 [101], [107].

Figure 27. Simulated temperature distribution in resistive ribbon technology

The predicted temperatures within the structure after a 12.5 ms, 25-mA current pulse are highest directly beneath the
electrode, and a maximum peak temperature (420 °C) occurs at the interface between the carbon-filled polycarbonate and
the 100 nm aluminum layer. Ink temperature is predicted to be 325 °C.

Figure 27 predicts that the composite temperature should decrease to 100 °C in 10 ms; this is considerably lower than the
glass transition temperature of the polycarbonate (Tg = 145 °C). The short cycle time permits the organic structure to retain
its integrity during printing. Additionally, relatively slow lateral thermal spread allows the resistive ribbon to produce high-
resolution printing.

2.1.2.3. Imaging Materials and Mechanisms

Thermal Paper. Thermal paper consists of a 5 to 10 µm-thick coating on a smooth base paper [105]. The coating contains a
leuco dye and a phenol developer. The developer can be modified to improve the thermal stability of the coating [106]. Both
components are shown on the page.

The color-forming reactants are dispersed separately from the binder (e.g., styrene – maleic anhydride copolymer) and other
coating constituents, and then combined in the correct ratio prior to application. After application to the smooth base, the
paper is gently dried below the activation temperature for the reaction between leuco dye and phenol.

When heated by a thermal head, both compounds react to form a color.

page 14 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

The higher the temperature (up to the point where all reactants are consumed), the higher is the optical density. Optical
density as a function of temperature in a static 5 s test is shown in Figure 28 [105].

Figure 28. Optical density vs. temperature for three papers of different sensitivities

a) High; b) Medium; c) Low

Heating duration with a thermal head depends on print velocity and is on the order of 2 – 8 ms. Thin- and thick-film head
temperatures greatly exceed the reaction temperature of the chemicals and compensate for the short print residence times.
Temperatures as high as 450 °C can be reached with both thin-film and thick-film heads.

Print images are also developed on thermal paper if the paper is placed adjacent to the aluminum-coated side of a resistive
ribbon transducer (without thermoplastic ink). Pulsing the head in the normal manner then results in an image.

Inks for Thermal Ribbons. In thermal ink transfer printing, imaging is accomplished by selectively pulsing a thermal head in
contact with a polyester (Mylar, DuPont) or similarly stable base. The base contains, on the side opposite that contacted by
the head, an ink layer which is in contact with paper (Fig. 29) during printing [109].

Figure 29. Thermal ink transfer printing

a) Transparent film; b) Ink layer; c) Transferred ink; d) Plain paper; e) Head

Ink transfer occurs as a result of the ink melting, wetting the paper, separating from the substrate, and solidifying on the
paper. The melting point of the ink, heat capacity, ink thickness, melt viscosity, ink open time (i.e., the time the ink remains in
a molten stage), paper roughness, and head pressure determine the amount of energy required for printing. Because in the
thermal ink transfer process, heat generated in the head must pass through Mylar or a similar thermal insulator which adds to
heat loss, inks with the lowest permissible melting point are chosen. The composition of 60 °C melting point ink is shown
below, in weight percent [109]:

Carnauba wax 20
Ester wax 40

Pigment 20

page 15 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Oil 10

Additives 10

Wax-based inks tend to smear and, for this reason, low-melting thermoplastic polymer-based inks are currently being
developed to replace them.

Inks for Resistive Ribbons. Ink formulations for resistive ribbon printing can be relatively simple if they are for printers only.
Ink composition is much more complex in typewriter applications where lift-off correction is desired.

For printer applications, polyamide resins with a carbon loading of ca. 8 wt % are very satisfactory [110]. Polyamide inks
based on Versamid 940 or 950 have the correct melting-temperature viscosity and open time for high-quality images. Poly
(vinyl acetate) (Aldrich Chemical), an amorphous polymer with a glass transition temperature of 30 °C and a melting point of
65 – 70 °C, can also be used as the thermoplastic binder for resistive ribbon inks. Addition of ca. 10 wt % Regal 330 carbon
black and a trace of Methyl Violet produces an ink with good transfer characteristics and good optical density.

These inks are modified for use with the resistive ribbon-saving feature. Here, ribbon is fed at a rate slower than the
movement of the head with respect to the paper. For example, if the rate of printing is 80 cps (characters per second)
(8 inches/s, 10 pitch), then the ribbon is fed at 4 inches/s for a 2 : 1 ribbon saving. This feature improves ribbon mileage and
provides a cost-per-character saving at increased print velocities. Inks used with the ribbon-saving feature must have good
flow properties. They must print two or more characters from an amount of ink normally required for one character, with
minimum loss of optical density of the print; thus, formulation adjustments are necessary. A low melting point (70 °C), low
melt viscosity resin is blended with a medium melting point (90 – 100 °C), medium melt viscosity resin. These blends provide
engraved character print quality at paper: ribbon velocity ratios of 2 : 1.

Ink formulations for typewriter ribbons must allow for lift-off correction; i.e., the image on the typewritten page must be
removable. Resistive ribbon printing occurs by activating the electrodes at an energy level sufficient to transfer ink from the
aluminum side of the polycarbonate substrate to the paper. The electrodes in the 40-electrode linear array are activated
according to the desired print pattern, and electrical current passing through the ribbon to the aluminum ground causes Joule
heating. The heat melts the ink, which, at the print pressure, transfers to the paper (Fig. 30).

Figure 30. Basic components of resistive ribbon technology

A) Printing process: a) Paper; b) Print head; c) Ribbon supply; d) Ribbon take-up; e) Pinch rollers

B) Correction process: a) Electrodes; b) Correction roller; c) Correction block

In the correction mode, all 40 electrodes are turned on at a reduced energy level for the entire block matrix of the character to
be corrected [111]. Energy is reduced for character correction by modulating the pulse width of the current to the electrodes
(typically a 50 % duty cycle). The head pressure (2 N) is the same for both correcting and printing. The correction roller force
is 2.5 N [111]. The block technique heats the ink to a temperature at which it permanently bonds to the character to be
corrected, but not to the paper. The unwanted character is removed from the paper when the ribbon is lifted away. The block
technique can correct any character with a single 40-electrode pattern.

An ink formulation that allows both printing and correction to be performed is relatively complex. When the ribbon is
manufactured, the aluminum layer is first coated with a low melting point release layer [112]. This intermediate release layer
is overcoated with a water-based pigmented ink layer consisting of ca. 11 % carbon black, and a mixture of ethylene – vinyl
acetate copolymer and a poly(ethyl acrylate) – polyacrylonitrile latex.

The two-layer structure produces high-quality printing [250 printing elements per inch (100 µm) vertical ×360 printing
elements per inch (75 µm) horizontal] [111] when sufficient thermal energy is used to give complete ink transfer. It also
permits lift-off correction when the thermal energy is reduced appropriately (see Fig. 30).

2.1.2.4. Comparison of Thermal Technologies

A comparison of the various thermal print technologies based on the literature is given in Table 1, which is intended to serve
as a guide. Data were obtained (or calculated) from a number of sources.

Table 1. Comparison of various thermal print technologies

Technology Application Resolution, Energy to print, Peak temperature,

dots/mm J/cm2 °C

Silicon mesa direct thermal 200 [87]

paper
Thin film direct thermal 8 [113] 4 [113] 400 [104]

page 16 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
paper
thermal ink 3.6 [109] 6 [109]
Thin film direct 8 – 16 [102] 4 [102] 550 [109]
meander
Thick film direct 1 – 3 [103] 8 [96] 450 [114]
Resistive ribbon thermal ink 9.4 (vertical) [111] 3 420 [101]
14.2 (horizontal)

Table 1 shows that all technologies require approximately the same order of energy for printing. They probably also have
approximately the same level of thermal efficiency. The efficiency of resistive ribbon thermal transfer printing is ca. 6 %; that
is, ca. 94 % of the heat generated in the ribbon is lost to the head, ribbon, ink, paper, and environment. In direct thermal
transfer printing, 85 % of the thermal energy is dissipated in the head and 15 % goes to the paper [114]. What fraction of the
15 % is used to develop the image and what fraction is lost to the paper is not known, but thermal efficiency could be very
similar to that of resistive ribbon printing [108].

Although thermal transfer printing is relatively inefficient, the energy required to print is still considerably less than that
required for impact printers (the wire matrix, for example, is ca. 100 J/cm2). On the other hand, the energy required for
bubble or thermal ink jet (see Section Ink-Jet Printing) is on the order of 0.35 J/cm2.

Selection of Appropriate Thermal Transfer Printing Technology. The type of thermal technology used depends on the
application. Calculators, which require neither permanence nor high resolution, use direct thermal transfer printing with thick-
or thin-film heads [115]. Direct technology is also used in typewriters and computer printers, with thin-film heads being
employed when good-quality printing is required. If high print speed is required, a full-page linear array of thermal elements is
employed in the printer. This type of head lends itself well to graphics.

If high-quality printing with good archival properties is required, thermal ribbon printing is employed; a high-resolution thin-film
head or the resistive ribbon printing process is used. Both technologies have APA (all points addressable) and graphic
capabilities, but with resistive ribbon printing a greater variety of papers can be used because of the head flexibility. It also
has the potential for higher print speeds than thermal head printing because no duty cycle is required during operation.
However, because the heating function is in the ribbon and not in the head, resistive ribbon printing supply costs are higher
than thermal head – thermal transfer supply costs. The dual print and correction function offsets this higher cost somewhat
for typewriters. The process of reducing the rate of relative movement of the ribbon past the paper lowers ribbon costs for
printers considerably.

2.1.3. Ink-Jet Printing

In ink-jet technology, tiny drops of ink fluid are projected directly onto a surface for printing without physical contact between
the printing device and the surface. The placement of each drop on the printing substrate is controlled electronically; and ink
jet printing has become one of the preferred methods for printing data that has been generated or manipulated by a
computer. Printing is accomplished by moving the print head across the substrate or vice versa. This direct printing process
has been incorporated into many different low-cost and relatively simple printing mechanisms [116]. In addition, the
noncontact feature of the technology allows for a variety of printing applications in which high printing speed, quietness, and
substrate insensitivity are required. High-quality printing can be achieved by manipulating the drop size (with print resolution
up to at least 40 dots per millimeter), by using inks containing multiple levels of dyes, or both [117]. Color printing is achieved
by simply placing the proper colorant in the ink vehicles.

Ink-jet technology can be traced back more than 200 years. In 1754, L'ABBÉ NOLLE, Master of Physics for the French
Dauphin, carried out research on the effects of static electricity on the flow of liquid drops from capillary tubes. In the late
1800s, Lord RAYLEIGH discovered that a fluid jet emerging from a nozzle breaks up into droplets and that the size of the
droplets can be controlled by applying a uniform disturbance [118]. The first practical ink-jet apparatus was disclosed by
THOMSON (Lord KELVIN) in his patent filed in 1867 [119]. Since then, patent activity has increased worldwide for various types
of ink-jet marking devices and has accelerated in the last 10 – 15 years to the point that hundreds of ink jet related patents
are issued worldwide each year [120-129]. A very large number of companies and many universities and research institutes
have been involved in the development of ink-jet printing, and many will continue to invest heavily for the forseeable future
[130-134].

The principle of drop generation varies among different ink-jet technologies. Electrostatic, magnetic, piezoelectric (acoustic
pressure), electrothermal (Joule heating), mechanical microvalve, and spark discharge are among those that have been
demonstrated. Regardless of the method employed ink jets can be classified into two basic categories: continuous ink jet and
impulse (drop-on-demand) ink jet.

Continuous ink jet is characterized by pressurized ink which is emitted through a nozzle to generate drops of ink directed to
the marking substrate in a continuous stream. A method of steering the drops to the desired printing location and disposing of
unwanted ink drops is part of the print mechanism. This technology has been widely used for industrial marking, coding and
labeling. It has also been adapted for very high quality color printing and computer-based high speed commercial printing.
Continuous ink jet systems were the first to mature into well accepted products. Although this technology is capable of very
high performance in both print speed and print quality, it has been somewhat limited because of its complexity.

Impulse ink jet differs from continuous ink jet in that the ink supply is maintained at or near atmospheric pressure. An ink drop
is ejected from a nozzle only on demand when a controlled excitation is applied to the drop-generating transducer. A huge
amount of effort has gone into the devlopment of drop on demand systems in the 1980s and 1990s, particularly in the

page 17 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
integration of the technology into small, relatively cheap printing devices that integrate hundreds of nozzles.

Applications of ink-jet printing appear endless. The printing process is termed nonimpact because no high mechanical impact
pressure is involved in transferring ink to the receiving surface as in letterpress or wire matrix printing. Hence, ink-jet printing
can be carried out on fragile objects such as eggs. Additionally, because it is nonimpact and no platen is required to support
the receiving substrate, the process can be used to print on cans, bottles, textiles, and many other objects. Ink-jet printing is
particularly suited to three- or four-color subtractive printing by having a set of nozzles for each of the cyan, magenta, and
yellow (also black for four-color printing) primary colors. The process is also well-suited to more conventional printing such as
word processing, facsimile printing, high-speed computer printout, black and white or color copiers (if attached to an optical
scanner), typewriters, bar code, label printers, and graphics images. However, the technology has found its widest
acceptance in the area of color computer-based printing, for which it is particularly suited.

Many products have been developed to satisfy a variety of these applications. Continuous ink-jet technology is used, for
instance, in labeling, coding, and marking systems available from Videojet, Domino Printing Systems, Imaje, and others. Iris
Graphics manufactures a color graphic drum printer specifically intended for graphic arts proofing. Scitex Digital Printing
Systems is developing a relatively high speed, full color, computer driven commercial printing system using a version of this
approach.

Impulse ink-jet technology is used in color printers developed by Tektronix, Hewlett-Packard, Canon, Lexmark, and many
others [135].

2.1.3.1. Continuous Ink Jet

Three types of continuous ink jets are available: the Hertz type, the Sweet type [122] (or the raster-scan type), and the binary
type. They differ from each other by the way the drops break off and the way they are steered.

Hertz-Type Continuous Ink Jet [123]. The Hertz continuous ink-jet device consists of a nozzle connected to a high-
pressure ink supply, a charge electrode, and an aperture plate (see Fig. 31). During operation, an unperturbed stream of ink,
emanating from the nozzle orifice, is broken into drops of irregular size because of Rayleigh instability. Turning the drop
stream on and off to the print medium is accomplished by turning the charge electrode on and off. Only uncharged ink drops
pass straight through the small aperture and reach the printing substrate. When ink drops are not wanted on the substrate
during printing, a sufficiently high-voltage pulse is applied to the charge electrode located at the drop break-off point, and ink
drops are inductively charged beyond their stability limit. At this point, the Coulombic force within each drop overcomes the
surface tension, causing the drop to explode into a fine mist. The receiving substrate is shielded from this unwanted mist by
the aperture plate. Unwanted ink can be discarded or recirculated to the ink reservoir through proper filtration because only
2 – 5 % of the expelled ink arrives on the print substrate.

Figure 31. The Hertz continuous ink jet

a) Pressurized ink supply; b) Nozzle; c) Charge electrode; d) Aperture plate spray catcher; e) Print substrate

In this printing method, more than one drop is used to print a single spot on the print medium. Controlling the number of tiny
drops per printed spot allows multiple gray level printing with very high-quality results. The nozzle size is typically ca. 10 µm.
Pump pressure can be up to ca. 4 MPa to achieve a jet velocity of 25 m/s. Charge electrode voltage on the order of several
hundred volts has been used. Drop frequencies up to 1 MHz have been reported. One shortcoming of this technology is that
the aperture plate cannot effectively block 100 % of the unwanted mist. Some of this fine mist ends up on the printing
medium as background and impairs print quality. This approach is the basis for the very high quality proofing printers made
by Iris Graphics.[136].

Raster-Scan or Sweet-Type Continuous Ink Jet [137]. The Sweet-type continuous ink-jet device is analogous to an
electron gun, except that the electron beam is replaced by a drop stream of conductive ink. The Rayleigh method of creating
uniform-size drops is employed. As illustrated in Figure 32, a piezoelectric crystal is used to perturb the pressurized ink
stream expelled from the nozzle orifice by the action of the pump. Under this regular excitation by the piezoelectric crystal,
the jet stream breaks into uniform drops at a short distance from the nozzle orifice. A charge electrode is placed at the point
of drop break-off, which selectively charges the ink drops according to the printing signal. The charged drop is steered away
from the normal trajectory as it passes through the pair of deflection electrodes which provide a constant, high electric field.
The larger the electrical charge on the drop, the more it is deflected. Hence, by controlling the sequence and level of drop
charging, information can be printed by raster scanning of the charged ink drops. Uncharged drops (unwanted drops) are
collected in the gutter and either discarded or recirculated.

Figure 32. The raster-scan continuous ink jet

a) Piezo crystal; b) Nozzle; c) Crystal driver; d) Pump; e) Ink supply; f) Charging control; g) Paper

This technique projects one drop to one print spot on the receiving substrate. Various nozzle sizes have been used for

page 18 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
different applications. Typical drop size is about twice the nozzle diameter. The pump pressure is lower than that in the Hertz
device, typically < 34 kPa. Drop-generating frequency can be up to several hundred kHz; drop velocity is ca. 20 m/s. Drop
charging can be achieved with electrical pulses <1 00 V. The drops are deflected with the aid of an electric field of ca. 10 –
20 kV/cm. Most of the commercial marking and labeling devices use this technology

Binary Continuous Ink Jet [138] is similar to the Sweet-type except that uncharged drops are used for printing (Fig. 33).
Charged drops are deflected into the gutter. The advantage of this approach, compared to the raster scan, is that charging of
drops need not be very precise because the relatively large gutter can capture deflected drops with moderate deflection
error. However, because printing is done with one nozzle per print spot, a multiple nozzle print head is normally employed.
This technology is being used in a page-wide array of jets by Scitex Digital Printing for their digital printing press.

Figure 33. The binary continuous ink jet

a) Piezo crystal; b) Nozzle; c) Crystal driver; d) Pump; e) Ink supply; f) Charging control; g) Gutter; h) Paper

Technology Issues in Continuous Ink Jet. Although utlization of a continuous ink jet appears to be a simple direct printing
process, the operational tolerance of each component of the printing device requires rather stringent control. This creates
complicated engineering problems. For example, breaking up the jet stream is not always neat and straightforward. Smaller
drops, called satellite drops, traveling at a different velocity can occur between two consecutive main drops. These satellite
drops are charged differently and, therefore, can be deflected onto the printing substrate at undesirable locations. Print
quality is degraded by such background spots. To some extent, satellite drops can be controlled by proper selection of ink
properties, excitation methods, nozzle geometry, and jet velocity.

The break-off point of the ink stream, a short distance from the nozzle exit, depends strongly on ink properties that are a
function of temperature. Proper charging of the drop can be achieved only when the break-off point resides within the charge
electrode which is at a fixed distance from the nozzle. Hence, ink properties must be controlled precisely to ensure proper
ink-jet operation.

When consecutive drops are charged, electrostatic repulsion exists among these drops, which leads to deviation from the
normal flight path (Fig. 34). One solution is to maintain a large spacing between drops by either increasing drop velocity or
reducing drop frequency. However, drops with too high a velocity cause splashing on the print substrate, and the throughput
decreases when jets are operated at too low a frequency. Another approach is to charge only every other drop (the
alternative uncharged drops are called guard drops).

Figure 34. Charge repulsion of ink drops

Aerodynamic factors also affect drop motion. In a continuous stream of traveling ink drops, the trailing drop is pulled along by
the influence of the wake of the leading drop. As the drops move toward the receiving medium, the very first drop
experiences a higher drag than the trailing drops. Therefore, the first drop slows down and tends to merge with the following
drops, causing misplacement of printing information.

When ink is recirculated in the continuous ink-jet device, the loss of ink solvent through evaporation causes a shift of ink
properties. Furthermore, stopping and starting of the ink jet can also lead to crusting of ink in the nozzle orifice and splashing
of ink on the charge and deflection electrodes. Contamination of the charge electrode and deflection plates not only produces
corrosion on hardware but also affects the uniformity of the electric field, resulting in inconsistent drop projection. In multiple-
nozzle ink-jet arrays, problems are compounded by compliance with channel uniformity requirements. An excellent
description of the physics of continuous ink-jet technology can be found in [144-151].

Although, all of these problems can be managed through careful engineering, the increase in cost and complexity has
undoubtedly minimized some of the attractive attributes of continuous ink-jet technology.

2.1.3.2. Impulse (Drop-on-Demand) Ink Jet

The three typical impulse ink-jet devices are the piezoelectric, the thermally driven “bubble-jet”, and the mechanical
microvalve. This latter method generates huge drops and has been widely used for box labeling where large, crude
characters are appropriate and for mural or advertising printing where the image produced is large and is viewed from a large
distance.

Piezoelectric Impulse Ink Jet [125-127], [152]. Piezoelectric materials such as barium titanate and lead zirconate titanate (
Ceramics, Electronic), once properly polarized, change their physical dimensions when exposed to an electric field
through their surface electrodes. This is essentially the principle used by all piezoelectric impulse ink-jet drop generators.
When an electric pulse is applied to the transducer, it induces motion on the piezoelectric element which generates a
pressure wave inside the ink (Fig. 35). A portion of the pressure wave travels to the nozzle exit where it interacts with the ink
meniscus, causing a drop of ink to be ejected from the nozzle. Ink drops are expelled only when the transducer is excited.
Refill of the nozzle is accomplished by capillary action and the effect of inertia.

page 19 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 35. Operational principle of impulse ink jet

a) Transducer; b) Nozzle; c) Paper; d) Ink supply at atmospheric pressure; e) Data pulse train

One of the driving forces in the practice of this technology has been the continued miniaturization and integration of the drive
elements and fluid cavities of the devices to pack more and more nozzles into less and less space. An analogy to the
advances in integrated circuit technology is a good one and in fact many of the strategies employed and materials used for
this integration come from that industry. One of the advantages of smaller devices is that they are inherently amenable to
high drop generation frquencies. Drop frequencies of well over 20 kHz has been demonstrated and are continuing to climb.
Energy requirements are extremely low, on the order of only a few microjoules per drop. Discussions of the physics of
piezoelectric impulse ink-jet technology can be found in [153-158].

Until the 1990s, integration of piezoelectric ink jet devices into large arrays that were cost competitive with bubble jet
technology had lagged significantly behind. However, efforts at Epson, Xaar, Tektronix and Spectra, among others, are
rapidly closing that gap and are eliminating this as an inherent disadvantage of this method. One of the inherent advantages
of piezo-driven devices is that they can accommodate a wide variety of inks, including phase change or hot melt ink
technology.

Thermal-Bubble Jet [128]. Thermal-bubble jets differ from the piezoelectric type by their use of drop propulsion. A tiny
resistive heater produced by thin-film vacuum deposition is used. Figure 36 A shows a typical thin-film heater structure on a
silicon substrate (g). The heat delay layer (f), underneath the resistive film, improves thermal efficiency. The passivation layer
(b, d) on top of the resistive film protects the heater from mechanically enhanced corrosion. Two typical designs have been
used in the industry, the top-shooter (Fig. 36 A) and the end-shooter (Fig. 36 B).

Figure 36. Structure and design of thermal-bubble ink jets

A) Top-shooter: a) Nozzle plate; b) Inorganic overcoats; c) Resistor; d) Polyimide overcoat; e) Conductor; f) Thermal barrier;
g) Substrate

B) End-shooter: a) Glass body; b) Ink supply inlet; c) Ceramic substrate; d) Resistive element; e) Conductive lead

The principle of bubble jet operation is shown in Figure 37. When an electric pulse is applied to the heating element, Joule
heating for a few microseconds causes the ink adjacent to the heater to be superheated and eventually exceed its nucleation
threshold. A vapor bubble forms over the heater and expands rapidly because of the energy supplied from the superheated
ink layer. The growing bubble acts like a piston pump to expel a drop of ink from the nozzle. As usable energy in the
superheated layer is depleted, the vapor bubble starts to collapse and ultimately disappears. The nozzle is then refilled by
capillary action and an inertial effect.

Figure 37. Operational principle of thermal-bubble ink jet

—— Voltage; –·–·– Temperature; –··–··– Bubble volume

As mentioned above, a significant advantage of this technology is the easy manufacturability of the device. Because batch
fabrication techniques are used to produce the bubble jet head, hundreds or thousands of nozzles can be produced with
precision and at low cost. A good review is given in [159].

An inherent limitation of bubble jet technology is the need for the ink to generate a vapor and to be able to survive the very
rapid and very large temperature transients that it sees. This has been overcome quite well in many commercial systems that
use aqueous ink, but does limit the selection of ink materials and types.

Mechanical Microvalve Printing. This technology is based on electrically actuated microvalves which cut off a drop of ink
delivered by a pressurized supply. The orifice sizes for this technology are very large and the repetition rates very low.
However, it is very well suited for the applications in which it is used and amenable to a very wide variety of inks.

Other Drop on Demand Technologies. Several other ways of generating drops have been demonstrated in the laboratory
or have been adapted into pilot products that did not achieve market acceptance. These methods include jetting devices
driven by spark discharge, electrostatic forces, magnetic forces, and acoustic lenses. They are described in the references
listed and will not be mentioned further.

Technology Issues in Impulse Ink Jet. The advantages of impulse ink-jet technology, its low energy requirements and
ease of integration into arrays, can also be pitfalls. Small amounts of air bubbles, foreign particles, and contamination at the
nozzle exit can have devastating effects on the performance of the print head. Variations in ink properties, whether caused

page 20 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
by a change in environmental conditions (e.g., temperature and humidity), evaporative loss, or time-dependent chemical
reaction in the ink, can render the print head nonfunctional. Channel to channel variability due to the above effects and
manufacturing variability is also a huge issue for practitioners. The key to a successful impulse ink jet relies on the
development of a unique ink, a simple automated maintenance procedure, and careful control of the manufacturing process
for both inks and print heads. The requirements of ink-jet inks are discussed in Section Ink-Jet Inks.

As briefly mentiond above, material selection is particularly important in thermal-bubble jets because operating conditions are
quite harsh. Repetitive high thermal stresses are imposed on the thin-film structure. Cavitational impact during bubble
collapse erodes the top layers of the heater. Aggressive chemical corrosion also occurs at the elevated operating
temperature. Discussions of the materials and failure mechanisms in thermal-bubble ink-jet devices appear in [161-163].

2.1.3.3. Ink-Jet Nozzle Orifice

In addition to the motion transducers, the nozzle orifice plays a crucial role in the operation of an ink-jet print head because
jet and drop formation depend greatly on its quality and uniformity. These requirements are even more stringent for ink-jet
arrays. Accurate positioning of the nozzles, uniform stream directionality, and identical drop break-off must be maintained
across the array. The surface on the front side of the nozzle plate should not be wetted by the ink, yet wettable surfaces are
needed for the inside of the nozzle cavity. These requirements place a significant burden on the manufacturing of the nozzle
plate.

Drawn-glass nozzles as well as mechanically punched or laser-drilled metal nozzle plates were used formerly. The current
techniques for nozzle plate fabrication are batch-manufacturing methods. Photolithographic etching on silicon, electroforming
(see Chap. Photochemical Machining) with nickel, or laser ablation using a photomask and excimer laser technology are
three of the most popular methods. Many nozzle plates and thousands of nozzles can be manufactured at one time by these
processes. Details of these methods have been described [159], [164].

2.1.3.4. Ink-Jet Inks

The requirements for ink-jet inks are different and in some ways much more stringent than those for inks for traditional impact
printing methods (see Section Thermographic Printing). The patent literature on ink jet ink technology has grown
exponentially over the last 20 years and a wide variety of novel ink chemistries and applications have been proposed. For the
practitioner of this technology the field is still wide open and the increasing demands placed on ink jet printing systems
require increasing sophistication and complexity in the ink systems.

Physical Properties. The physical properties of ink-jet inks should be fully compatible with the ink-jet technology used
because ink is an integral part of the printing mechanism during operation. Inks used in continuous ink jet are generally
electrically conductive (resistance < 1000 Ω/ cm) and low in viscosity (< 4 mm2/s). Impulse ink-jet inks need not be
conductive but can require higher viscosity (< 20 mm2/s) for better damping of pressure wave reverberation after drop
ejection. Thermal-bubble jets require, in addition, ink components that are thermally stable above 300 °C. In many
applications, purified versions of dyes and colorants may be required to avoid ionic corrosion of print head hardware and first-
drop delay because of undesirable absorption phenomena or chemical reactions at the nozzle meniscus or elsewhere in the
system.

Printing Property Requirements. No long-term problems such as crusting, corrosion, precipitation, or first-drop delay should
occur in the ink-jet print head in which ink resides for the life of the printer. Ideally, printing on a variety of print substrates
should produce the same high-quality (sharp, dense, and precise) results with rapid dry time, waterfastness, and
lightfastness. Finally, ink-jet inks should be nontoxic, have low flammability, and offer a good and stable shelf life. The ink
should show good chemical, smear, and rub resistance and high optical density.

Three types of ink-jet inks are available, (1) water-base inks, (2) oil – nonaqueous-base inks, and (3) the hot-melt system
[165].

Water-Based Inks. The dilemma with aqueous ink jets is the desire to never have the inks dry out while in the print head, but
have the inks instantly dry when they are printed on paper or a transparency. Water-based inks usually contain > 70 % water.
Small amounts of humectants such as glycols are added to reduce the rate of evaporation and prevent precipitation of dyes
when evaporation occurs at the orifice. Biocides are added to prevent the growth of microorganisms. For continuous ink jets,
some salt must be included in the formulation to obtain the necessary electrical conductivity. However, because high
conductivity generally leads to electrochemical corrosion, basic corrosion inhibitors may be needed. For impulse ink jets,
corrosion is a smaller problem because conductivity is not required. Thickeners such as water-soluble polymers may be used
to increase the viscosity of inks formulated for piezoelectric impulse jets. These polymers can also enhance print quality on
the substrate by limiting drop spread by various mechanisms. However, these thickeners can cause maintenance problems
because restarting the jet after it has been out of use can be difficult if the wrong materials are chosen. To shorten dry time,
surfactants or penetrants may be added. However, print quality degradation must be avoided because feathering usually
occurs on the print substrates with these additives. Hundreds of patent applications representing a huge amount of work
have been devoted to overcoming these problems and achieving high print quality on so-called “plain papers”. The quality of
prints obtained from aqueous ink jet printers on typical office papers has radically improved in the past decade. However, for
color printing applications the best image quality is still achieved on specially coated papers designed as part of the printing
system that uses aqueous inks.

Oil – Nonaqueous-Base Inks. This type of ink can contain more than 50 % nonvolatile vehicles such as glycols and fatty
acids or their esters. They have low toxicity, minimal corrosive effects, and low volatility. First-drop delay and long-term
maintenance are normally not problems. The viscosity of these inks is higher than that of water-based inks, which makes
them most suitable for application in piezoelectric impulse ink jets. However, because drying of these inks on paper (or other
print substrates) relies on absorption into the paper fibers (or the substrate matrix), less than optimal print quality usually
occurs on printing paper because the ink bleeds along certain directions (e.g., long fibers). However, this property has
recently been greatly minimized by the use of ink systems which contain pigments that precipitate from the ink upon printing

page 21 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
and do not tend to follow the solvent as it wicks out into the paper. Oil – nonaqueous-base inks are not used in continuous
ink-jet applications because of the difficulty of dissolving high concentrations of salts in the ink to make it conductive.
However, many solvent based inks are used in continuous jets for industrial marking applications. These systems are
particularly good for dissolving high polymers that give the inks good durability on substrates such as glass and metals.
Because the solvents are usually polar organic solvents such as methyl ethyl ketone, methanol, ethanol, and the like, they
can dissolve certain salts in high enough quantities to provide the appropriate conductivity. Another interesting development
in this area is the adaptation of UV curable ink systems to ink jet marking and coding technology. Conceptually, this
eliminates any volatile solvents from the process, and in fact, these ink systems are now available in commercial marking
applications.

Hot-Melt Inks [165] are solid at room temperature. They have also been referred to as phase-change or solid inks. Print
heads using these inks are kept at elevated temperature so that the ink contained in the print head system is in the liquid
state. Molten ink is ejected in discrete drops and flies through the air as a liquid. The very low mass of the ink drops
compared to the substrate enables extremely rapid ink solidification upon reaching the receiving surface so that very fast dry
times and little feathering are obtained. Print quality is usually outstanding, and print substrate sensitivity is minimal. The
initial implementation of this technology was in devices that jetted the ink directly on to the final substrate. However, the most
widely accepted commercial use of the technology is in color printers that use the system in an offset printing fashion. In this
implementation, the ink is initially jetted onto an intermediate drum, on which it solidifies but still remains soft and malleable at
the drum's elevated temperature. After the complete image has been assembled on the drum, it is transferred in one step to
the final substrate by means of a pressure nip. The transfer process is very analogous to the final transfer step in offset
lithographic printing or to toner transfer from the imaging drum or belt to the substrate in photocopiers or laser printers. Unlike
those technologies, offset phase-change ink jet printing requires complete, clean transfer of the image, which is achieved by
the use of a sacrificial oil layer on the drum. This process provides for high dot spread while maintaining tight control of the
final dot size and shape. It also provides (in combination with the ink) for high image durability and a minimization of the final
ink film thickness, which were two large limitations of the technology as originally implemented. The process also is
amenable to high speed printing and a simple, realizable paper path.

The ink vehicles are typically waxes with melting points of ca. 80 – 95 °C. The selection of colorants for hot melt inks is very
stringent because their solubility in the waxy carrier is usually low. The waxes, as well as the colorants, must be nontoxic and
environmentally safe. The colorants must have the right shade of cyan, magenta, yellow, or black, and be soluble in the
waxes that are usually not particularly good solvents for colorants. The colorants and inks must be thermally stable for
extended periods while the inks are molten when the printer is held in the ready-to-print state. The resulting prints are
expected to have excellent fastness to light and good resistance to abrasion and scratching.

Advantages of hot melt inks over aqueous inks are their lack of drying at the ink jet nozzle. Hot melt ink's rapid solidifying of
the ink drop gives precise control of dot gain and excellent image quality on almost any media. Prints made with hot melt inks
are not marred by water. The loading of the solid ink sticks into the printer is a simple, clean operation.

Printer start-up is slower from a complete power off because time is required to liquefy a block of solid ink. A power on – off
cycle causes volume changes during an ink phase change in the print head, usually requiring a maintenance purge to
remove any ingested air bubbles. These problems are mitigated by leaving the power on continuously. The printer goes into
a standby mode when not being used for several hours. In the standby mode the temperature is dropped to just above the
freezing point of the ink. In this manner, the printer can be brought back to the ready-to-print state very quickly. Hot melt ink
jet printers have found commercial success particularly in local area networks of approximately a dozen personal computers,
where the printer is used often and the speed of the printer is more than sufficient to keep up with the demand.

2.1.3.5. Colorants Used in Ink Jet Inks [166]

Colorants for ink jet inks originally came from the textile trade inventory and, more commonly, from a limited list of food dyes.
Because the ink jet orifice is about the width of a human hair, ink developers quickly discovered the dye purity requirements
for ink jets are much more stringent than for coloring fabric or even food. Both textile and food dyes require significant
purification before inks made from them will perform reliably in ink jet printers. After the first generation of ink jet inks, ink
formulators discovered that purification of commercially available dyes alone was insufficient. Even after purification, the dyes
lacked some desired properties, such as wetfastness or lightfastness. Color chemists are developing a second generation of
colorants specifically designed for ink jet applications.

Blacks. C.I. Food Black 2 was one of the first black dyes used in aqueous ink-jet inks [167]. Printing with ink made with Food
Black 2 gives an image easily marred by water. To overcome the lack of wetfastness, ICI chemists eliminated or replaced
some of the hydrophilic sulfonic acid groups with carboxyl groups in Food Black 2 [168], [169]. The modified dye has much
improved wetfastness while retaining the desirable characteristics of the original dye.

page 22 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

There is significant patent activity in the field of pigmented aqueous inks and major improvements have been made in
pigment dispersions [170]. Dye-based inks tend to have brighter colors and pigment-based inks generally have duller colors
but better lightfastness. The main reasons why color pigments have not yet come into general use include the difficulty in
producing good dispersions and getting reliable performance in ink-jet print heads. To achieve better lightfastness than that
provided by dye-based inks, Du Pont has pushed the development of aqueous pigment dispersions for ink-jet applications
[171]. The only commercial application of aqueous ink jet inks using colored pigments has been in wide format ink-jet
printers. Encad utilizes disposable print heads printing at 300 dots per inch. Mutoh uses piezoelectric driven “permanent”
print heads with much larger than normal apertures for the printing of outdoor signs about 100 dots per inch. In the printing of
signs, lightfastness is more important than a large color gamut. In the wide format printer application, the ink-jet print heads
are used almost continuously. It is yet to be determined in typical office and home use, whether ink-jet printers will be used
often enough to prevent clogged nozzles due to the settling and agglomeration of the color pigments. Improved dispersion
formulations appear to have mitigated some of the reliability concerns of pigmented inks, but not yet solved the lack of bright
primary colors. Within Hewlett Packard's ink jet color set, only the dye-based black ink has been changed to the Du Pont
carbon black aqueous dispersion. Hewlett Packard's cyan, magenta, and yellow inks are still dye-based [172].

Black dyes consisting of a 2 : 1 dye – metal complex, such as C.I. Solvent Blacks 27, 28, 29, 35, 45, etc., afford excellent
thermal stability, solubility, and lightfastness in hot-melt ink applications.

Cyans. The greenish blue color required for a good cyan in an aqueous ink was originally accomplished with C.I. Acid Blue 9
(also known as Food Blue 2):

page 23 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Unfortunately, this triphenylmethane dye has relatively poor lightfastness. Most inks for current ink-jet applications use a cyan
dye based on the copper phthalocyanine chromogen. The type and number of substituents on the chromogen determines the
ink system for which the dye is suitable. Aqueous inks have used C.I. Direct Blue 86 (see below, W = X = SO3Na, and
Y = Z = H) and C.I. Direct Blue 199 (W = X = SO2NH2, Y = SO3Na, and Z = H) [173].

Oil and hot melt inks require more oleophilic substituents. such as those found in C.I. Solvent Blues 25, 44, 48, 67, 70, etc.
These cyan dyes normally have two to four sulfonamide substituents. In C.I. Solvent Blue 25, for instance,
W = X = Y = Z = SO2NHR where R is a six carbon aliphatic tail. The other phthalocyanine cyan dyes differ only by the
number of sulfonamide groups present and the aliphatic chain length [174]. In the manufacture of these cyan dyes the
sulfonic acid ammonium salt is formed as byproduct instead of the desired sulfonamide. Increasing amounts of the polar
ammonium salt decrease the dye solubility in nonpolar oils and waxes. In pigment applications the material most commonly
used is C.I. Pigment Blue 15:3 where W = X = Y = Z = H.

Magentas. Bright magentas are commonly produced with the xanthene chromogen. Xanthene dyes are tinctorially strong
(generating a lot of color per unit weight of dye), thermally stable, and produce clean, bright, almost ideal magenta shades
that are difficult to reproduce at the same cost with other chromogens. However, xanthene dyes are notoriously poor in
lightfastness.

C.I. Acid Red 52 (A = B = ) produces a strong, bright magenta shade in aqueous inks.

In some applications, the less bright, but more lightfast C.I. Reactive Red 180 has been used in combination with the poor
lightfast Acid Red 52 [175].

page 24 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

When the water-soluble, acid form of C.I. Acid Red 52 reacts with a tertiary alkyl primary amine, it produces +NH3CR3 (where
R3 represents a total of 12 to 14 carbons), which is soluble in oil and wax-based inks. The dye – amine complex is quite
stable to oxidation [176]. A magenta dye successfully used in Tektronix' hot melt inks is also based on the xanthene
chromogen. It is C.I. Solvent Red 49 (A = COO– and B = H), which requires an acid medium to develop color.

If hot melt ink prints using Solvent Red 49 are laminated and the adhesive comes in direct contact with the ink, some
laminate adhesives will cause a significant loss in color by enabling formation of the colorless lactone (leuco form). It is best
to use the laminates recommended by the printer manufacturer.

The anthraquinone chromogen has been used to produce acceptable lightfast magentas. C.I. Disperse Red 60 (see below,
X = H, Y = OC6H6, and Z = OH) and C.I. Solvent Red 172 (X = C6Br2H2CH3, Y = H, and Z = OH) are soluble in hot melt inks:

There are several drawbacks with anthraquinones. They are not nearly as tinctorially strong as the xanthenes and therefore
require more dye to produce an equivalent color level. The amount of anthraquinones that can be added to an ink is limited
by their tendency to crystallize and migrate particularly under the influence of fingerprint oils. Crystallization and migration of
the dye causes an objectionable shift in the original color of the print. In pigment applications the material most commonly
used is C.I. Pigment Red 122.

page 25 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Yellows are easily obtained from many azo derivatives and azo dyes can be used in aqueous, oil, or hot melt inks. C.I. Acid
Yellow 23 and C.I. Direct Yellow 86 have each been used as a process shade yellow in aqueous inks.

In general, yellows from azo derivatives are not as thermally stable as anthraquinones. Yellows that can form the hydrazone
tautomer such as C.I. Acid Yellow 23, C.I. Disperse Yellow 119 (see below, where A = H, B = NO2, R = Et), and C.I. Solvent
Yellow 162 (see below, where A = SO2NHCH2CH(Et)(CH2)3CH3, B = H and R = n-butyl), appear to be more thermally stable
than azo dyes that can not readily form the hydrazone tautomer.

Solvent Yellow 146, as Orasol Yellow 4GN from Ciba, is structurally similar to C.I. Solvent Yellow 162 and has been
successfully used in hot-melt inks [177].

In pigment applications the material most commonly used for a yellow shade is C.I. Pigment Yellow 17 (see below, where
X = OCH3) or the redder yellow, Pigment Yellow 12 (X = H).

page 26 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

2.1.3.6. Color Printing with Ink Jets

Ink-jet printers are excellent for color printing because the amount of colorant required to create the image can be metered
precisely with each drop of ink. The potential applications of ink-jet color printers are many. They range from limited color
business graphics to full color prints from conventional silver halide color negatives or still video cameras (solid-state
cameras which take still pictures like conventional cameras) [178].

In color ink-jet printing, the substractive system (see Photography) is used to present a variety of colors to the eye [179].
Typically, amounts of red, green, and blue are subtracted from a white reflecting (paper) substrate. This is accomplished by
applying proportions of cyan, magenta, and yellow dyes that absorb red, green, and blue, respectively. By overlaying
saturated yellow, magenta, and cyan primary inks in appropriate combinations, saturated patches of red, green, blue, and
black colors can be obtained.

Creation of more complex colors requires the addition of primary colors with varying degrees of saturation (white
component). Total saturation denotes the absence of a white component; pink is a desaturated red. In ink-jet printing,
saturation adjustment is affected by placing the transparent ink on only a portion of a small white picture element. This
element should be small enough so that the eye cannot resolve the spots of primary ink themselves but integrates over the
element and reacts to the composite spectrum.

In color printing, hue relates to the dominant absorption wavelength and determines whether the color appears blue, green,
or orange. Lightness refers to the relative amount of reflected light. Adding black to red decreases lightness [180-182]. These
principles are illustrated in Figure 38.

Figure 38. Elements determining chromatic expression

Ink and Paper for High-Quality Color Ink-Jet Printing. The complex list of requirements for ink-jet inks described in
Section Ink-Jet Inks applies to inks for color printing but with additional requirements for the dye or colorants. To reproduce a
large gamut of colors, dyes must be chosen in which the peak wavelength corresponds to cyan, magenta, and yellow.

In addition to providing a wide color range, the inks must be colorfast and pH stable, and they should not aggregate at the
concentrations used [179], [183]. To obtain high-quality images from water-base inks applied from an ink-jet nozzle, the
properties of the surface on which printing occurs must allow controlled spreading and penetration of the ink. In this case, ink
drops of specific dimensions and color can be printed. A photographic-grade base paper, coated with barytes, to which a thin
film of poly(vinyl alcohol) has been applied, is an excellent substrate for ink-jet printing if ultrahigh-quality color images are
desired.

Creating Colors by Using Ink-Jet Matrix Printing. Consideration is given here to full color ink-jet printing. The straight
forward seven-color printing used for business graphics and charts is only a small subset of full color printing. Color print data
can be obtained from stored data or from a color scanner, as shown schematically in Figure 39; the latter describes ink-jet
matrix printing. Additionally, the matrix printing method is used for continuous ink-jet printing as well as impulse printing [178],
[179].

Figure 39. Substractive color reproduction

To demonstrate how a variety of composite colors can be achieved, a small white matrix composed of four elements (shown
in Fig. 40) should be considered. The matrix is small enough so that the eye does not resolve the individual elements but

page 27 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
integrates over the whole matrix. By covering one element of the matrix with yellow ink, one-fourth of the reflected blue
intensity is absorbed. The visual effect is a yellow-hued, very light, desaturated color (Fig. 40 B). If yellow coverage is
increased, the hue remains the same, but saturation increases and lightness decreases (Fig. 40 C). A black component can
be included by adding an equal amount of cyan and magenta as demonstrated in Figure 41 A, and a hue shift can be
attained by adding unequal amounts of the primary inks as shown in Figure 41 B.

Figure 40. Varying saturation

A) White paper: hue neutral, lightness maximum

B) Light yellow: hue yellow, lightness high, saturation low

C) Saturated yellow: hue unchanged, lightness reduced, saturation maximum

Figure 41. A) Varying lightness: a) Saturated yellow; b) Hue yellow saturation unchanged, lightness decreased

B) Varying hue: hue shifted to green, saturation full, lightness decreased

One means of determining the amount of ink to be printed in ink-jet printing is to compare the analog level of the scanner to a
preset threshold value. If the scanner falls below the threshold at any print element, a drop of ink is placed in that element.
This basic print or no-print access is limited with respect to quality. To copy continuous tone prints, such as photographs,
which contain varying levels of color, a means of reproducing these levels such as the threshold matrix technique must be
employed.

Threshold Matrix Technique. Associated with the print matrix is a multilevel threshold matrix, with threshold levels selected so
that the color scale is reproduced correctly. An example is shown in Figure 42.

Figure 42. Threshold printing

Here, 50 % of the original blue is reflected. Thus, to satisfy reproduction, 50 % of the blue must be absorbed from the four-
element print matrix of the reproduction. This is accomplished by printing two elements with blue-absorbing yellow ink. The
decision to print is made by comparing the scanner and threshold values, and printing occurs on elements where the scanner
is below the threshold setting. The same logic is applied to green – magenta printing and red – cyan printing to obtain the
same red, green, and blue reflectance in the copy as in the original. In this example, a four-element gray scale matrix is
described.

Techniques Required for High-Quality Color Reproduction. By using the multilevel threshold matrix technique, good-
quality color reproduction is obtained. The limitations of this method relate to drop size, drop placement errors, and nonideal
dyes used in inks.

Drop size has been adjusted experimentally in black and white printing to produce a resolution of 57 pels/mm (i.e., printing
elements = no. of drops per millimeter). This resolution is excellent; however, problems of reliability of the printing device are
severe. To obtain a resolution of 57 pels/mm, nozzle diameters of ca. 12 µm are required. With such nozzles, clogging
occurs frequently, even with sophisticated ink filtration procedures.

Drop placement errors, which are inevitable when the multilevel threshold matrix technique is employed, can be alleviated by
using error carry or error diffusion techniques. The multiple error correction algorithm (MECCA) is a method available for this
purpose [184]. In the MECCA method, drop printing is controlled by a single-level, simple threshold, but the error incurred at
each scan – print element is carried to the next element. To be more specific, the analog level of the scanner indicates some
level of gray (color lightness), and the print result after thresholding is either a color or white, depending on whether or not a
drop of ink is printed. The error then is the difference between the desired gray (color lightness) and the resultant color or
white. This error is determined and added to the scan value of the succeeding element. Further print enhancement is
obtained by weighing surrounding elements.

This method not only reproduces gray and color levels, but also reduces some of the objectionable texture and patterns of
the threshold matrix technique. Edge definition is sharper and moiré patterns are decreased significantly. However,

page 28 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
considerably more data processing is required.

The problem of reproducing color with inks containing real dyes is that the yellow, magenta, and cyan inks have secondary
absorptions in unwanted regions of the visible spectrum. If ideal inks could be used, the amount of ink required to reproduce
a color could be calculated in a straightforward way based on the principal absorption bands of the primary inks. For
example, if the green scanner measures some amount of green absorbed in the original, then by knowing the green
absorbance of magenta ink, the amount of magenta ink required to obtain a similar green absorption can be determined.
However, cyan and yellow inks also have some secondary absorption bands in the green region, and if some of these inks
must be printed with magenta, an excess amount of absorbed green occurs in the reproduction. Other secondary additions
compound the color-matching error.

If secondary absorption can be calculated quantitatively, it can be compensated in a number of ways. One simple way is to
hold back a preset amount of ink of low equivalent density while maintaining full density reproduction of the saturated primary
colors. A more accurate method of reproducing color is to include secondary absorption in color-matching equations. Hence,
when a certain density of red is required, the secondary red absorption characteristics of the yellow and magenta inks are
taken into consideration in a quantitative way. Algorithms have been derived to do this for red, blue, and green densities
[185]. Application of these algorithms to the scanned data permits extremely accurate color reproduction. An ink-jet color-
copying robot is shown in Figure 43.

Figure 43. Ink-jet color-copying robot

a) Electronic interface; b) Red separation scanner; c) Photomultiplier tube (PMT); d) Original; e) Cyan ink-jet printer; f) Green
scanner; g) Blue scanner; h) Magenta printer; i) Yellow printer; j) Ink supply

Undercolor Removal. The black component of dark colors can be produced by applying all three primary inks to a print
element. With undercolor removal, the three primary inks required to create black are replaced by a single amount of black
ink. This requires the addition of a black print head to the cyan, magenta, and yellow print heads.

Such a system has several advantages. One is that no print element has more than two layers of ink. This saves ink and
reduces printing problems caused by large amounts of wet ink on the printing surface. Another advantage is that more
freedom is allowed in formulating the three primary color inks because the need to produce a good three-color black is
eliminated. One disadvantage is that drop placement must be accurate because the black drop must not fall on the same
element as the primary colors, which would negate their effect and degrade color quality.

2.2. Technical Copying

Technical copying refers to the copying of technical documents: technical drawings, construction designs, sketches, drafts,
patterns, graphics, parts lists, and so on. Unlike office copying, the documents produced in technical settings are often
characterized by their large size.

Indeed, special photochemical methods have always been used to copy large-size originals, and the blueprinting technique
remained in use until the 1920s. Blueprinting is based on the reduction of iron(III) salts and the formation of Berlin blue. It
was finally replaced by diazotyping. Other photochemical procedures for special applications, such as the Dylux process,
have come into use more recently.

Documents of smaller size can, of course, be copied by using electrophotographic office copiers, which are available even for
sizes as large as DIN A 0. However, diazotyping is still a method used for copying technical papers because of its simplicity,
versatility, and cost effectiveness.

The change has taken place in the preparation of technical drawings [computer-assisted design (CAD) technique] is also of
considerable importance. These documents are no longer filed as drawings on paper or film, but are stored in digital form.
When required, the file can be used to produce drawings with the aid of a plotter.

2.2.1. Diazotyping
The diazotyping process is based on two fundamental properties of certain aromatic diazonium salts:

1. They can react with aromatic coupling components or those with enolizable hydrogen to yield dyes (coupling reaction).
This reaction shows a strong pH dependence.
2. On exposure to light, they lose nitrogen and, consequently, their coupling ability (photolysis) [186-188].

2.2.1.1. Photolysis
Photolysis proceeds as follows:

Heterolytic cleavage of the C–N bond occurs from the excited singlet state and is the rate-limiting step. Subsequent reaction

page 29 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
of the carbonium ion with the nucleophilic partner proceeds very rapidly.

In a nonpolar environment (e.g., in a polymer matrix), homolytic photolysis is also considered:

The quantum yield from the photolysis of commercially useful diazonium salts is, without exception, < 1. Until now,
sensitization has been achieved only in exceptional cases (e.g., in solution or in a special polymer matrix) [189]. Hence,
suitable diazonium salts must absorb in the emission range of strong light sources (e.g., mercury lamps) and, in addition,
should be sufficiently stable and produce light-decomposition products that are colorless.

Examples of commonly used diazonium salts are:

These diazonium salts with electron-donor substituents in the para position absorb in the “actinic range”, i.e., in the near
ultraviolet (380 – 420 nm) (Fig. 44). The light sources used for diazotyping are mercury high-pressure lamps, as well as
special fluorescent lamps (superactinic or ultraactinic). Because an irradiation energy of 50 – 100 mJ/cm2 is required,
diazotyping materials are used exclusively for contact copying. They can be processed in daylight and do not require a
darkroom.

Figure 44. Light absorption of diazonium salts I and III and line spectrum of the mercury lamp

2.2.1.2. Coupling Reaction

A barely visible yellow diazo picture is discernible on the diazotype material after it has been exposed to light under a
transparent original. This image can be destroyed by further exposure and has much too little contrast for practical use. In the
second step, development, the coupling reaction converts the diazo image to a high-contrast, light-stable color picture.
Hence, diazotyping is a positive process: i.e., dark parts of the original are reproduced as such.

The coupling reaction proceeds as follows [190]:

A base is required to convert the phenol or naphthol to the corresponding phenolate or naphtholate. This base must be
added during development. Conversely, acids are added to the diazotype materials to prevent early coupling and stabilize
the diazonium salts. The coupling component is more important than the diazonium salt for the color of the resulting dyes.
For example, blue shades are obtained with naphthols; compounds with enolizable hydrogen or monohydric phenols give
yellow coupling products; resorcinol and its derivatives produce brown to red shades; and very dark dyes are obtained from
phloroglucinol. Black printing diazotype material represents a special case because it is produced by mixing various couplers,
usually blue, yellow, and brown. Examples of commonly used couplers are:

page 30 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

2.2.1.3. Production of Diazotyping Material

Opaque or transparent paper, transparentized paper, lacquered paper, and polyester films are employed as base materials.

Special wood-free, sized papers are used. They are usually precoated with an aqueous dispersion of fine filling material
(white pigments; Pigments, Inorganic) and binders before the light-sensitive coating is applied. The precoat homogenizes
the surface of the paper, avoids penetration of light sensitive solution into the paper and improves contrast. Transparent
papers are frequently prelacquered (e.g., with cellulose esters) to improve transparency and mechanical strength. The most
extensively used synthetic films are polyesters because they are dimensionally and mechanically stable. They are usually
chemically pretreated to achieve the adhesion of the lacquer layer. The functions of the layers are to accept the diazo coating
and to allow writing with both pencil and ink.

After pretreatment of the base material on one or both sides, it is coated with the solution of diazotype chemicals.
Pretreatment and coating are carried out in one working process in large coating machines that have several coating and
drying stations. The running speed varies from a few meters per minute to >100 m/min, depending on material. Aqueous
coating solutions can be used on opaque paper of 40 – 210 g/m2. However, solutions in organic solvents are necessary for
lacquered paper and for film. The film thickness usually varies from 36 to 125 µm.

2.2.1.4. Processing
Diazo printing machines with different degrees of efficiency are available for copying. High-performance printers are largely
automated and capable of releasing folded copies. They achieve processing speeds >20 m/min [191]. Two different
development techniques are available:

1. In one-component materials, the light-sensitive layer contains only the diazonium salt and stabilizers. The coupler is
added during development as a buffered aqueous solution. Two different qualities of paper are used. One quality
needs to be dried in a high-performance developping unit immediately after development (semiwet method). For the
other quality 3 g/m2 of developer is sufficient for application on one side so that drying can be dispensed with
altogether [192].
2. In two-component materials, the light-sensitive layer contains not only the diazonium salt and stabilizers, but also the
couplers. Development is carried out by passing the material through an ammonia atmosphere in which coupling
occurs (dry method). The ammonia technique is more versatile and simpler in the choice of materials: black, blue, red,
and brown printing types can be employed.

Two-component material can also be developed by applying a small amount (approximately 2 g/m2) of an alkaline organic
solution [193]. However, the copying quality of ammonia development is not achieved in this process.

Annual consumption of diazo printing material in the Western world is estimated at 1.1×109 m2.

Some important producers of diazotype material are:

Azon Corporation, Johnson City, New York.

OCE Imaging, Charleston, Illinois.
Dietzgen Corporation, North Brunswick, New Jersey.
Oce van der Grinten NV, Venlo, The Netherlands.
Ozalid Company, Loughton, United Kingdom.
Regma SA, Arques-la-Bataille, France.
Ricoh Company, Tokyo, Japan.

2.2.2. Other Photochemical Systems

The Dylux method, [194], [195], developed by DuPont, has attained some importance, especially in the graphics industry as
proof material. This method requires only light and no chemicals for processing. In normal processing, selective exposure to
UV light ( 1) produces the corresponding dyed picture (negative method). Upon subsequent overall exposure to visible light
( 2), the UV-sensitive substances are deactivated and the copy is fixed. Conversely, positive copies are obtained if the
selective exposure to visible light is followed by overall exposure to UV light.

Dye formation occurs as a result of the photochemical oxidation of a leuco dye (1) by an oxidizing agent:

page 31 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

The energy of irradiation required to attain a color density of 1.0 is 15 – 50 mJ/cm2.

A second photoredox system consisting of a quinone (3) and a hydrogen donor (4) is required for inactivation. The quinone is
reduced photochemically to hydroquinone, which is a radical trap and stops further dye formation.

(3) may be pyrene quinone and phenanthraquinone; (4) can be a polyether, a nitrilotriacetic acid derivative, or a
triethanolamine ester.

Inactivation requires 0.5 – 2 J/cm2 of light energy. Dylux material is available on paper as well as on film base. It is suitable
for the reproduction of continuous tone, line, and screen pictures.

Another interesting technique has been developed by the Mead Corporation [196]. This method is based on the joint
microencapsulation of a monomer, an initiator for photopolymerization, and a dye precursor; sheets are coated with these
microcapsules. During exposure, the monomer polymerizes in the exposed areas and the microcapsules harden.

If the exposed sheet and a receptor sheet are subject to pressure, the capsules that were not hardened burst and their
contents produce a dye on those areas of the receptor sheet that were unexposed. Hence, a positive copy of the original is
obtained.

This system can form the basis for a color-copying technique (cycolor) if three different types of microcapsules are employed,
each containing a dye precursor for one of the three primary colors in combination with an appropriate selectively absorbing
initiator. No chemicals are required for processing. However, this system has not yet been adopted commercially.

2.2.3. Silver Process

The silver halide process has been used for about 150 years to reproduce images and documents of all kinds and is
described extensively elsewhere (see Photography).

The silver halide process is capable of producing negative and positive copies with excellent archival permanence and
remarkable resolving power. Further useful properties are the wide range of speeds (i.e., sensitivity) and contrast values
obtainable.

Silver halide materials are not used for office copying because of the availability of electrostatic photocopiers using plain
paper (see Section Electrophotography). However, the silver process remains for special and professional copying
applications. For example, pre-press copy operations such as page make-up demand high quality halftone reproduction and
high-density images. Until the advent of electronic page make-up, diffusion transfer materials were extensively employed in
pre-press work.

Before the introduction of electrophotography, the silver diffusion transfer process was the most important copying technique
(1950–1960). The black and white silver diffusion transfer system (Copyrapid) was introduced in the 1950's. The process was
independently invented by ROTT at Gevaert and WEYDE at Agfa [197] in 1938 ( Photography – History ). In principle, two

page 32 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
coated papers are employed which in one processing step provides a negative and a positive image within ca. 1 min (see
Section Black and White Copying Materials Based on Silver Diffusion Transfer Technology) by diffusion of the unexposed
silver in complexed form from the negative photo-sensitive paper to the positive paper.

Compared to classical negative processing, in which printing the positive is followed by a second exposure and processing,
this meant a great improvement in speed and convenience.

The silver halide diffusion transfer process has recently gained widespread use as the light sensitive media for Computer-to
plate (CtP) lithographic printing plates (see Section Color-Copying Materials Based on Dye Diffusion Chemistry).

Related to black and white copying systems is the color diffusion transfer process. This was first designed by Polaroid for
instant imaging and then developed for professional purposes by Agfa-Gevaert (Copycolor) [198]. Dyes generated during
color development of the negative migrate to a positive receiving layer (see Section Color-Copying Materials Based on Dye
Diffusion Chemistry).

2.2.3.1. Black and White Copying Materials Based on Silver Diffusion Transfer Technology
Diffusion Transfer. The diffusion transfer process consists of a sequence of physical and chemical processes coupled by
diffusion. The process starts with an exposure step. In a camera, a light pattern is exposed, which produces a latent image in
the emulsion layer, as shown in Figure 45.

Figure 45. Diffusion transfer mechanism

a) Support; b) Negative; c) Emulsion layer; d) Light absorption; e) Light reflection

= AgX grain; with dot = AgX grain with latent image

Then the exposed negative (b) and a receiving material (a) are passed through the developing bath while being pressed
together (c), as shown in Figure 46 [201].

Figure 46. Developing bath

The exposed negative and the receiving material are passed through the developing bath while being pressed together

a) Receiving material; b) Exposed material; c) Squeegee rollers

Before both materials reach the squeegee rollers (c), the emulsion is fully developed according to the scheme in Figure 47.

Figure 47. Schematic of diffusion transfer development. Before the exposed negative and the receiving material reach the
squeegee rollers the emulsion is fully developed

a) Support; b) Emulsion layer = AgX grain; with dot = AgX grain with latent image; = Fully developed AgX grain

Silver ions in the silver halide grains which bear a latent image are reduced by reaction with the developer (hydroquinone).

Parallel to this reaction silver halide is dissolved by an aqueous solution of the complexing agent (sodium thiosulfate) [
7772-98-7], with the formation of water-soluble silver (I) complexes.

From the squeegee rollers, these complexes are transferred by diffusion to the receiving layer where a positive image forms
on the nuclei by physical development (Fig. 48). The nuclei in the receiving layer are (heavy) metal sulfide clusters, which act
as a nucleating catalyst for silver deposition. During this physical development, the silver complexes deposit on the nuclei
where silver is reduced. The electrons are supplied by a superadditive pair of developers (usually hydroquinone [123-31-9],
and methyl-Phenidone) ( Photography).

Figure 48. Schematic of a positive image formed on the nuclei by physical development

page 33 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

a) Support; b) Emulsion layer; c) Negative; d) Positive; e) Nuclei-containing layer

= AgX grain; = Fully developed AgX grain; •••• = Ag in the nuclei-containing layer; Dotted area = Nuclei

After the transfer phase, the positive is separated from the negative, rinsed (optional), and dried [202].

System Elements. A diffusion transfer system consists of a light-sensitive material (negative), a receiving material (positive),
and a processing solution [197].

Most of the negatives are composed of two, three, or more layers. The top layer shields the emulsion layer from the
environment and is called the antistress layer. Transport properties and contact rheology in this polymer layer are controlled
by the addition of wetting agents, adhesion control agents, and matting agents.

The emulsion layer below the antistress layer consists of AgCl, AgClBr, or AgClBrI microcrystals between 0.1 and 1.0 µm in
size. This emulsion is chemically and spectrally sensitized, and can be negative- as well as positive-working (direct-positive
emulsion). After transfer, the former yields a positive legible image; the latter, a negative one.

Often the emulsion layer is undercoated with a third layer which is used as an antihalation layer and contains antihalation
dyes. In some materials, this layer also contains developer substances so that the system becomes processible by activation
with an alkaline solution ( Photography).

Figure 49 shows a transmission electron micrograph (TEM) of the layer composition of a diffusion transfer negative. This
layer assembly is coated on polyethylene-coated paper or a subbed polyester film. The film version can be used for obtaining
a reverse-reading image by exposure through the back [203].

Figure 49. TEM micrograph of the layer composition of a negative

a) Antistress layer; b) Emulsion layer; c) Anti-halation layer; d) Polyethylene – paper

Receiving Material (Positive). Receiving materials are also made of several layers and are applied to polyethylene-coated
paper or polyester. The protective layer (antistress layer) functions as in the light-sensitive material and, at the same time,
avoids the formation of superficial metal deposits (bronzing).

Under this layer, the nuclei-containing layer is located. The nuclei catalyze the physical development. Mixtures of metal
sulfides are generally used as nuclei, but single sulfides or colloidal metal sols are also possible (Ag, Au, PdS, Ag2S, etc.).
The radius of these colloidal particles ranges from 3 to 7 nm. The nuclei are dispersed in a polymer phase. In addition to
gelatin, cellulose derivatives, poly(vinyl alcohol) [9002-89-5], polyacrylamide, and colloidal silica [7631-86-9] are used.
Swelling of these polymer layers controls the diffusion of reactants through the gel and is decisive for transfer rate and
sensitometry. A hardening system regulates the final swelling.

A third layer is sometimes located under the nuclei-containing layer [204]; dyes, toning agents, stabilizers, developing agent
systems, and silver ligands can be incorporated into it. In Figure 50, the composition of a two-layer positive after silver
transfer is illustrated by means of a TEM micrograph.

Figure 50. TEM micrograph of positive layer composition after silver transfer

a) Antistress layer; b) Nuclei layer; c) Adhesion layer; d) Polyethylene terephthalate

In the nuclei-containing layer, the image is formed on the nuclei and results in a finely divided silver deposit, whose particles
are a maximum of 20 nm in size. This is considerably smaller than that of the chemically developed AgX grains of the
emulsion layer (300 – 600 nm).

Processing Solution. Processing of diffusion transfer systems occurs in developers as well as in activators, depending on
whether or not developing substances have been incorporated into the negative material.

Processing occurs at pH values >10. Therefore, inorganic bases (NaOH, Na3PO4, Na2CO3, etc.), organic bases (secondary
and tertiary alkanolamines), and mixtures of both are used.

The diffusion transfer step is made possible by the formation of a soluble and reducible silver complex. Thiosulfates and

page 34 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
thiocyanates are used as inorganic silver complexing agents. They are incorporated in the form of their ammonium,
potassium, and sodium salts. For special purposes, organic silver ligand systems are also used (hydantoin and pyrimidine
derivatives).

These chemicals are protected from air oxidation by the addition of sodium sulfite. To control the structure and transfer speed
on the silver deposition process, inorganic potassium bromide and organic (mercaptotetrazoles, benzimidazoles, etc.)
inhibitors that inhibit the physical development process are used. Viscosity and surface tension of the bath are regulated with
polymeric thickening agents (cellulose derivatives) and low molecular mass surfactants.

In addition, a superadditive system of developing substances consisting of a hydroquinone derivative and an alkyl- or
dialkylpyrazolidinone is added to the developing liquid [205].

2.2.3.2. Modern Lithographic Printing Applications

The Copyproof black-and-white repro products from Agfa remain the most important diffusion transfer copy materials. The
Copyproof system consists of negatives and a range of positive receiving films and papers.

However, this application of diffusion transfer technology is under pressure from electronic and digital reproduction methods.
Of growing importance today is the use of the silver diffusion transfer process to make lithographic printing plates. Silver,
when deposited by a diffusion transfer process, is in the form of densely packed clusters of metallic silver, unlike the
filamentary morphology that is formed with usual photographic development. In this densely packed form, silver is capable of
accepting lithographic ink. Thus, if an image of silver is formed on a lithographic surface by the diffusion transfer process and
ink and water are applied, the silver accepts the ink whilst the water lies in the non-image background. On press, the ink on
the silver image is transferred via an impression cylinder to the paper.

In so-called analogue litho plate-making, a photographic film is generated which carries the image to be printed. This
image is placed as a mask in contact with a printing plate and flood exposed to UV light. In such a process, the printing plate
is a low light sensitive resin-based material. Silver halides on the other hand are highly sensitive to a wide range of visible
wavelengths of light. Consequently, the diffusion transfer process has been incorporated into lithoplates that can be imaged
directly without the need for the film intermediate. For many years, plates based on the diffusion transfer process have been
used for direct exposure in a camera to a paste-up copy or original. In the Agfa Copyrapid Plus system (Fig. 51 A) a paper
donor sheet (i.e., the light sensitive negative) is exposed to the original or paste-up on a standard process camera and then
processed (see Fig. 46) in contact with a grained and anodised aluminium receiving sheet (i.e., the positive). After further
treatment to improve the oleophilicity of the silver image, the plate is loaded onto the press.

Figure 51. Simplified schematics of lithographic printing plates based on the diffusion transfer process.

a is the two-sheet Copyrapid transfer system, where a) is aluminium, b) is a grained and anodised alumina film containing
the nucleation particles, c) is the diffusion transfer developer, d) is the exposed silver halide emulsion coating, e) is the paper
support.

b is the mono-sheet Setprint/Supermaster or Digiplate/Silver Master system, where f) is the lithographic layer containing
nucleating particles, and g) is the paper or PET support.

c is the mono-sheet Silverlith, Lithostar, or Digiplate aluminium system.

All systems are positive-working lithographic plates. h) Developer diffusion into emulsion and nucleation layers.

Copyrapid Plus is a two-sheet system. A printing plate, which combines the donor and receiver functions, is classed as a
mono-sheet system (Figs. 51 B, 52). Mitsubishi Paper Mills and Agfa manufacture flexible plates in large volumes in which
the substrate can be either PET or resin-coated paper (Silver Digiplate and Silver Master from Mitsubishi Paper Mills, and
Setprint Plus and Supermaster from Agfa respectively for imagesetter and camera exposures). Mono-sheet products with a
transparent PET base can be exposed on a standard process camera through the base from a right-reading original. For
paper based materials, a camera fitted with a reversing mirror is required to give a right-reading image on exposure.

Figure 52. Cross-section of a Silverlith plate before and after processing.

Both the two-sheet and mono-sheet plate systems described above are better suited to small offset, single or 2-color work.
With the advent of digital page make-up and the development of stable lasers, computer-to-plate (CtP) became possible in
the late 1980's. CtP lithoplates capable of commercial and newspaper 4-color work became the major theme of the 1990's
pre-press market. Silverlith from the lithoplate manufacturer Howson-Algraphy was the first CtP fully grained and anodised
aluminium printing plate (Fig. 51 C, 52). It used the diffusion transfer process and combined the mono-sheet idea with a true
lithographic substrate. Thus, the grained and anodised substrate is coated with a colloidal silver nucleating agent followed by
a chlorobromide/gelatin photo-emulsion. After exposure, the plate is developed in a diffusion transfer developer where the
page 35 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
exposed areas are rapidly converted into filamentary silver that stays in the gelatin medium. The unexposed areas dissolve
at a slower rate to form soluble silver complexes that diffuse to the nuclei lying on the anodic film where they are reduced to
metallic silver. Because of the rough grain structure of the substrate, the silver becomes mechanically attached. When the
plate is washed with water after development, a positive image of silver remains adhered to the anodised substrate. Only
about 0.8 gm/m2 of silver is deposited in the image areas at a thickness of around 10 to 50 nm but despite this print runs of
many 100,000's of impressions are possible.

Continued developments in this field at DuPont-Howson and at Agfa-Gevaert have led to the present range of Lithostar Plus
(LAP) aluminum plates from Agfa. These printing plates can now be exposed on platesetters using 488 nm argon-ion lasers
(LAP-B), 532 nm frequency-doubled YAG lasers (LAP-O), 633 nm HeNe and 670 nm diode lasers (LAP-R) and, in 2000/01,
405 nm solid state lasers (LAP-V). Mitsubishi Paper Mills have also entered the market with a similar family of aluminium
diffusion transfer plates (Silver Digiplate SDP- A and SDP- R for argon-ion and HeNe or red diode lasers respectively).

No other CtP consumable has the breadth of application as these plates. The diffusion transfer process finds itself, once
more, at the center of a revolution in applied imaging technology.

2.2.3.3. Color-Copying Materials Based on Dye Diffusion Chemistry

Although the Ektaflex and Agfachrome Speed systems described below are no longer available they are included for their
technical interest.

In the first half of the 1980s, color-copying materials based on dye diffusion chemistry came on the market; until then, these
had been used only in the manufacture of instant color materials for amateur cameras (Kodak PR 10, Polaroid SX 70, and
Polacolor).

Kodak Ektaflex System. Kodak elaborated the Ektaflex system with the PR 10 chemistry ( Photography). After some
years, this was withdrawn from the market because of patent litigation with Polaroid. In this system, a two-sheet material was
used, consisting of a light-sensitive sheet and a receptor sheet on which the color image formed by diffusion transfer.
Positive-working and negative-working versions were available.

In the negative-working version, common silver halide emulsions were used. The positive character of the second material
was obtained with special direct-positive emulsions such as those in the PR 10 instant material.

For both types, the same receptor sheet was used. A special processing apparatus was developed that moistened only the
light-sensitive sheet for 20 s. Then the film was laminated onto the receiving paper by a pair of pressure rollers. The
sandwich could be separated after a contact time of 8 – 15 min, depending on the temperature and type of paper. Rinsing
was not necessary because the light-sensitive sheet contained a neutralization layer and a timing layer, which lowered the
pH to neutral (a timing layer is a binder layer which delays the penetration of OH- ions to the neutralization layer). The copy
was dry after a few minutes. An overhead film for making transparencies was also available.

Agfa – Gevaert Color Diffusion System. In Agfa-Gevaert diffusion materials, a completely new chemical concept is used. This
is a system of dye release, which enables the replacement of direct positive emulsions with grains containing fog centers by
negative emulsions that are simpler to prepare. In this way, direct-positive color images are obtained.

The principle is deduced from a reaction described in the chemical literature: 4-hydroxybenzylphenylsulfones are cleaved in
alkaline medium to form a sulfinate and a quinone methide, although other sulfones are highly alkali-stable [207].

The reaction can be redox-controlled by introducing an additional hydroxyl group in the phenol nucleus. In this way, o- and p-
hydroquinone derivatives are obtained, which convert to quinones on oxidation. These quinones become more alkali-stable
with increasing number of substituents attached to the ring. Azo dye sulfinic acids are used as sulfinate components.

The following molecule has its bulky substituent on the hydroquinone residue:

page 36 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
This hydroquinone derivative can basically be oxidized by exposed silver halide in the presence of an auxiliary developing
substance (electron transfer agent, e.g., Phenidone) to a stable p-quinone that no longer releases its dye moiety [208].

For practical purposes, however, dye releasing appears to occur too quickly being only partly controllable by exposed silver
halide. Therefore, the chemical system is adapted further as follows [209].

Starting materials are two compounds, the alkali-stable quinone and an electron donor. At unexposed areas, the electron
donor is capable of reducing the quinone to the corresponding hydroquinone, which releases the dye – sulphinic acid moiety.
At exposed image areas, the electron donor is oxidized by exposed silver halide via the electron transfer agent so that the
stable quinone continues to exist there. The following compound is used as the electron donor:

Stable nondiffusing quinone derivatives that supply the dye – sulfinic acid moiety in the material include the following:

At the exposed image areas, two competing reactions take place:

1. oxidation of the electron donor by exposed silver halide via the electron transfer agent, and
2. oxidation of the electron donor by dye – quinone after which the dye – sulfinic acid moiety is released. This reaction is
undesirable and must be prevented. For this reason, the most rapidly developing silver halide emulsions (mainly AgCl)
are used.

The released, diffusing dyes are very pure, spectral-grade azo dyes without developer residue, so that very brilliant color
images can be obtained by this mechanism.

Without further sophistication the absorption of the incorporated dyes would interfere with the exposure of the silver halide
grains, resulting in a loss of sensitivity. Because of the pH-dependence of the absorption spectra this can be prevented by
keeping the pH of the material sufficiently low. Only during and after the alkali activation are the absorption spectra shifted to
a longer wavelength and the correct color is obtained. The dye-releasing compounds can be incorporated directly into the
silver halide emulsion layers without any loss of sensitivity. Thus, the material can be structured in a considerably simpler
way.

On the basis of this chemical principle, Agfa – Gevaert put two photographic materials on the market in the early 1980s: a
two-sheet version (Copycolor) and a mono-sheet version (Agfachrome Speed).

The Copycolor material consists of a light-sensitive and a receiving sheet to which the dyes diffuse during processing and on
which they are immobilized by the mordanting agent. The composition of both sheets, as well as the operation, are
represented schematically in Figure 53.

page 37 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 53. Schematic of a light sensitive (A) and a receiving sheet (B)

a) Backing layer; b) Polyester support; c) Red-sensitive layer with cyan-releasing compound; d) Intermediate layer with
sodium light-absorbing filter; e) Green-sensitive layer with magenta-releasing compound; f) Intermediate layer with yellow-
filter; g) Blue-sensitive layer with yellow-releasing compound; h) Antistress layer; i) Antistress layer (receiving sheet);
j) Mordanting layer; k) Support

The receiving sheets are coated symmetrically, which allows diffusion images to be generated on both sides. Different
versions exist: mat paper, glossy paper, and transparent film for the production of transparencies for overhead projection.

The light-sensitive sheet has a special absorption filter for sodium vapor light in the intermediate part of the red-sensitive
layer, which allows work in the darkroom under sodium vapor lamps.

Gradation of the material is rather steep (g = ca. 2.6) and can be regulated as desired by use of the contrast-reducing
screens.

Thus, a universal copying material is available for copying various originals, such as normal color photographs, color prints,
colored line originals and graphs, aquarelles, transparencies, and graphic selections. It can also be used for making
transparencies for overhead projection.

Light sources may be color enlargers, repro cameras, contact exposure units, etc.

The system does not need any special processors as in the case of Kodak Ektaflex; processing can take place in diffusion
transfer processors according to the Eisbein principle (see Fig. 46). The processing liquid is a highly alkaline activator.

On the basis of the same chemical principle, Agfa-Gevaert also developed a mono-sheet material: Agfachrome Speed [210].
The light-sensitive silver halide layers are covered with a white titanium oxide layer coated with the mordanting layer.

Exposure is through the support layer. The structure of the material is represented schematically in Figure 54.

Figure 54. Schematic of a monosheet material (Agfachrome speed)

a) Backing layer; b) Transparent support; c) Blue-sensitive layer with yellow-releasing compound; d) Intermediate layer;
e) Green-sensitive layer with magenta-releasing compound; f) Red-sensitive layer with cyan-releasing compound; g) Carbon
black layer; h) TiO2 layer; i) Mordanting layer; j) Antistress layer

2.3. Microfilms and Microfiches

2.3.1. Diazotyping
Although silver halide film is preferred for microfilm archives, working copies (“active microfilm”) are often made with diazo
film, especially when several copies are required [211].

The diazotyping process is simple and economical, and the diazo material is especially suitable because it is grain-free and,
consequently, gives very high resolution.

In the 1980s, polyester film has been used almost exclusively as the base material. The film is chemically pretreated to
increase adhesiveness and then coated with a laquer that usually contains the chemicals for diazotyping. In general, the
binders employed are cellulose esters, and the diazo compounds used are, in principle, the same as those described in
Section Diazotyping. The diazonium tetrafluoroborates and diazonium sulfosalicylates are preferred to zinc chloride double
salts because of their greater solubility in organic solvents and their higher thermal stability.

Diazo duplicating films, which give blue-lined copies, are employed only as reading film; black-lined duplicates can be used
to make further copies (generations). Diazo films are suited primarily for the duplication of writing because of their steep
gradation. Other types of film with specially adjusted, flatter gradation are employed for continuous tone pictures. Diazo
duplicating films are available in the same forms as photographic films (i.e., as roll film, flat film (microfiche), and aperture
cards). Processing of these films is done exclusively with the dry (ammonia) method (see Section Processing). Duplicating
machines with different classes of efficiency are available for roll films, microfiche, and aperture cards. High-performance
duplicators produce up to 2000 duplicates of size DIN A 6 per hour.

2.3.2. Vesicular Film

Like diazotyping, the vesicular method is also based on the photolysis of diazonium salts by UV light. In practice, the same
diazonium salts are employed in this procedure. They are applied in a thin layer of special thermoplastic material to a
transparent base material which is usually a polyester film.

During development, the picture is produced not with the aid of the diazo compound present on the areas covered by the
picture, but with the aid of nitrogen formed by photolysis on the nonpicture areas. Heating for a short time (≤1 s) to 100 –

page 38 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
150 °C softens the thermoplastic material; the nitrogen, which is at first dissolved, expands and microscopically small
bubbles (0.5 – 5 µm) are formed. Light is scattered at the interfaces between bubble and polymer, and these areas appear
dark when examined in transmitted light. This is normally a negative process and thus represents a welcome addition to the
positive diazotype duplicating films.

During fixation, the diazonium salt still present is destroyed by overall exposure to light. Without subsequent heating, the
nitrogen produced escapes slowly from the layer with no bubble formation.

The light sensitivity of this system is somewhat higher than diazotyping because the diazo compound need not be entirely
decomposed for picture formation, and development exerts a certain intensification effect. Both methods are used only for
contact exposure. Gradation with the vesicular film process is even steeper than with diazotyping.

Although this method was developed a long time ago [212], it was not used commercially until the 1950s [213]. The
duplication of computer output microfilm (COM) represents the most important application of this technique. At present, all
important producers of diazo duplicating film also supply vesicular film. Duplicating machines can often be used for the
processing of either diazo duplicating film or vesicular film.

Worldwide consumption of duplicating film (diazo and vesicular film) was estimated at 60 × 106 m2 for 1996. Some producers
(and trade names) of diazo and vesicular film are:

Rexam, Runcorn, Cheshire, United Kingdom.

Messerli AG, Glattbrugg/Zürich, Switzerland.
Rexam, South Hadley, Massachusetts.
Somar Manufacturing Company, Tokyo, Japan (Somic-D).
SKC, Sunnyvale, California.

[Top of Page]

3. Imaging in Graphic Arts

Johan Verelst, Werner Frass, Thomas Telser, Horst Hoffmann, Bernd Bronstert, Karl-August Springstein, Rod Potts

3.1. Graphic Arts Photography

3.1.1. Classes of Photomaterials Used in Prepress Production
Graphic photomaterials are characterized by their gradation, maximum density, energy requirement (i.e., the inverse of their
photographic speed), spectral sensitivity (see Photography), and recommended processing.

Phototypesetting Paper [217]. Most phototypesetting papers consist of a waterproof, resin-coated paper base covered with
a photographic emulsion having a gradation of 2.5 – 5, a maximum density of > 1.5, and requiring an exposure of 0.1 –
10 mJ/m2, depending on the type of light source in the phototypesetter. This also dictates the spectral sensitization: cathode-
ray tube (CRT) recorders necessitate orthochromatic materials; helium – neon lasers require peak sensitivity at 633 nm; and
the newest diode – laser devices need spectral sensitization to peak at 780 – 820 nm. In most phototypesetters, the
exposure times of a single image element range from 100 to 500 ns; for these very short exposures, the photoemulsion
should be optimized.

Phototypesetting Film [217]. Phototypesetting films are manufactured on a transparent film base consisting of cellulose
triacetate or poly(ethylene terephthalate) (PETP). The emulsion properties are similar to those of phototypesetting paper.
However, the silver halide content is higher, allowing for a maximum density > 4 and a gradient in the range of 3 – 6.
Phototypesetting materials are usually processed in fast Phenidone – hydroquinone developers (rapid access developers).

Scanner Recording Films [218]. Recording films for color-separation scanners and for output recorders of computer-based
color image assembly systems are always manufactured on a stable PETP base. The emulsion coatings are quite similar to
those of phototypesetting films; some have even higher gradient and higher maxi-mum density. The spectral sensitivity can
have a peak at 488 nm for argon ion lasers or at 633 nm for helium – neon lasers; both are commonly used light sources in
these applications. Scanner recording films are usually processed in rapid access developers; when extreme edge acuteness
of the recorded halftone image is required, lith developers are used.

Lith and Line Films [219]. Lith films produce very high gradients (> 10) and very high maximum density (> 5) when
processed in lith developers. Lith developers contain only hydroquinone as the developing substance and almost no sulfite.
The most common lith developers are orthochromatically sensitized and have their speed adapted for use in reprographic
cameras. Lith films are typically used for conversion of continuous tone pictures into halftones.

Line films are quite similar in density and spectral sensitivity, but they are usually processed in rapid access developers. The
gradients obtained are typically in the range 5 – 10. Their major application is the reproduction of line art.

Silver Halide Diffusion Transfer Materials [220]. Making a direct positive reproduction in one step is possible with silver
halide diffusion transfer materials (e.g., Copyproof of Agfa; see Section Silver Process). A silver halide-containing donor
sheet is exposed imagewise. By means of a developer containing a silver complexant, unexposed silver is transferred to a
receptor sheet consisting of paper or film. This process yields particularly high gradients in the transferred image, which
allows for excellent halftone images and very clean line art reproductions.

Contact Film [221]. Contact films are low-speed, nonorthochromatic line films. The most popular can be handled in bright
room light without fogging; the exposure is done with intense UV light, and processing relies on Rapid Access Developers.

page 39 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.1.2. Flow Scheme of Graphic Prepress Process.
In general, a publisher's job file contains four kinds of information: (1) text copy; (2) color pictures (most often continuous tone
color transparencies); (3) miscellaneous graphic elements and monochrome pictures; and (4) layout, which describes how all
the information is to be arranged on the pages.

In most printing processes (with the exception of gravure printing), ink transfer onto the paper is a bilevel mechanism [222],
[223] (i.e., an image element is either printed at full ink strength or with no ink at all). To incorporate continuous tone pictures
on the printed page, they must first be converted into a screened, bilevel halftone format [224].

Color-Separation Images. In color printing, pictures are reproduced by making a color-separation image for each primary
printing color (i.e., yellow, magenta, cyan, and the supporting black). The desired color print is obtained by printing the four
color-separation images on top of each other, by using the corresponding printing inks [214], [225].

Photo-Optical Color Separation Process. Color separations can be made by using a reprographic camera or an enlarger, and
suitable photographic films. The process of making color separations typically involves successive exposure of panchromatic
continuous tone film to the orginal picture through blue, green, red, and amber filters. The resulting continuous tone
negatives contain the yellow (Y), magenta (M), cyan (C), and black (K = key) information of the original, respectively. The
separation negatives are then copied onto orthochromatic lith film through a contact screen, to produce four halftone
separation positives that can be used to produce the printing plates.

Color Separation with Electronic Separation Scanners. Since the early 1970s the photo-optical process of making color
separations has largely been replaced by electronic separation scanners (see below).

Prepress Work Flow for Producing Color Pages (Fig. 55). Layout data pass through the central branch and dictate how all
informational elements should be positioned on the page (in many cases, the trade shop which makes the color separations
must verify and clean up the layout for inaccuracies).

Figure 55. Typical prepress workflow with stand-alone color separation scanner

Photomaterials: a) Typesetting paper; b) Typesetting film; c) Scanner film; d) Lith and line; e) Contact film; f) Color proofing;
g) Printing plates

In the left branch, text is composed according to the layout and set on phototypesetting paper for proofreading and correction
[226]. Correction proofs are often produced on plain paper by electronic page printers [227], [228].

After approval, the typesetter produces a final take, again in phototypesetting paper. Then a film negative is shot in a camera
with lith or line film. This negative is mounted onto the final page by the stripping department. For high-quality publications,
the final take is often made on phototypesetting film. (In many cases, text composition is done in a separate typesetting trade
shop.)

Bringing line art and monochrome continuous tone pictures to the final size, and converting the latter to halftones, are most
frequently done with a reprographic camera by using film or paper having very high contrast (i.e., line or lith materials as well
as silver halide diffusion transfer materials) [229].

On the right of the flow chart, color pictures are analyzed by the color-separation scanner. In real time, image data are
converted into Y,M,C,K data, which take into account the spectral characteristics of the printing inks [214], [230]. Moreover, a
scaling transformation is executed according to the dimensional instructions of the layout. Four color-separation halftone
films are obtained, one for each of the subtractive primary printing colors (yellow, magenta, and cyan) and for the supporting
black. During the conversion of continuous tone data into raster halftones, the nonlinear transfer characteristics of the printing
are compensated. Color reproduction can be corrected by retouching, often under guidance of a color proof [231]. Both wet
(dot etching) and dry retouching techniques are currently used in most shops [232], [233], with the latter being preferred.
After approval, color separations of each picture go to the stripping department where they are assembled into pages.

The stripping department is the nodal point at which all page elements come together in their final form. After manual
assembly, a clean copy should be made on contact film. Then, a color proof can be made of the page for the client's
approval.

A number of individual page films are usually mounted together to accommodate the size of the printing plates. The
imposition of the pages assures the proper page order after folding and binding.

Computer-Based Color Image Assembly Systems. Manual stripping and dot etching are the most labor-intensive nodes in the
work flow. Investments in image-processing technology are made in these areas to reduce production costs. The functionality
of modern computer-based color image assembly systems (Fig. 56) goes far beyond the automation of stripping images onto
a page. All pictures for a page are scanned, digitized, and stored on hard disks. Image manipulation and page assembly are
performed at an interactive color work station [234], [235]. The color display allows the operator to decide whether interacting
with the system will produce the desired effect on printing. (This soft proof at the work station can also be transmitted to a
distant client for approval; remote proofing and the hard copy proof by means of digital proof recorders are still in their
pioneering stage.)

page 40 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 56. Typical prepress workflow with color image assembly system

Photomaterials: a) Typesetting paper; b) Typesetting film; c) Scanner film; d) Lith and line; e) Contact film; f) Color proofing;
g) Printing plates

In most color system trade shops, the assembled images are recorded on scan film. Next, the text parts are combined with
the images by making copies on contact film. Color proofs of these finished pages are usually submitted to the client for
approval. Whenever the client requires any color changes, either a new run must be made through the system, or the work
already done can be saved in part by retouching. In many cases, the latter proves to be more economical. This explains the
interest of most shops in computer-assisted retouching systems.

3.2. Basics of the Use of Light-Sensitive Nonsilver Materials in the Graphic Arts
To reproduce information stored in the photographic original in a printing process, this information must be transferred to a
printing plate, which is generally coated with a light-sensitive material. This process is known as platemaking (in a narrower
sense “plate” also means the unexposed, light-sensitive plate. In German it is differentiated by using the term
“Druckformherstellung” for platemaking).

However, prior to platemaking the appearance of the final printed result must be determined. This is especially important in
four-color printing because (1) the human eye is extremely sensitive to color deviations; (2) many color originals are intended
to be reproduced in a slightly different shade of color and it is of interest to determine in advance whether the color variation
is correct; and (3) the printed result has economic consequences (e.g., in the case of a mail-order catalogue).

Plates can be prepared to obtain a preprint on a small printing press so that the printed result can be viewed in advance.
Preprints of this type are called press proofs. Although this process is still found in several regions around the world, it is less
common these days. It is very useful when several copies of the four-color prints are needed, the process is very time
consuming and it has now largely been superseded by digital proofing techniques.

An alternative way to press proofs to obtain an advance impression of the printed result is to use photochemical methods
similar to those employed in platemaking to prepare four-color copies of the photographic original. This technique is
sometimes called color-proofprinting, color proofing, or off-press proofing and can be generated either by analogue or digital
means. Because the application and chemical fundamentals of both color-proof materials and light-sensitive printing plates
are to a great extent identical, these principles are described in a separate section.

3.2.1. Application
The recording materials carry on a substrate (e.g., metal, paper, or plastic film) a more or less thin layer (0.1 – 7000 µm) of
light-sensitive coating. These light-sensitive coatings undergo photochemical or photophysical change [236] on exposure to
light from a suitable source. Usually, the light reproduces an image of the photographic original with its transparent and
covered picture elements on the light-sensitive recording material. The light sources used are xenon lamps and metal-halide
doped mercury lamps that emit in the spectral range 350 – 450 nm. Modulated lasers, (e.g., of 488 nm wavelength) are used
for line-by-line scanning exposure. A more recent development is that of thermally sensitive plates which are also exposed by
modulated lasers typically at 830 nm.

The processing of light-sensitive recording materials must occur in the total absence of daylight. Yellow light (light without the
blue component) is generally used for lighting the workplace. Thermal plates on the other hand may be safely handled under
daylight conditions.

After imagewise exposure of the recording material, the unexposed areas must be protected against the influence of light,
which would result in the destruction of recorded information. This process is generally known as development. It can be
achieved by chemical reaction of the exposed or unexposed areas of the coating, by dissolving the exposed or unexposed
areas from the carrier (subtractive development), or by selective deposition of a developing substance on the exposed or
unexposed areas of the coating (additive development). The photomechanical method is presented in Figure 57 as an
example of subtractive development.

Figure 57. Photochemical platemaking

a) Light-sensitive layer, negative working; b) Carrier material; c) Film original; d) Printing plate

The process is defined as either negative or positive working, depending on whether the transparent or the opaque areas of
the photographic original appear as the image in print. These terms cannot be related to particular chemical processes but
depend equally on the composition of the lightsensitive recording material and the read-out process (i.e., type of

page 41 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
development and intended use).

As far as chemistry is concerned, two basic types of systems exist: coatings that become soluble after exposure to light
(photosolubilising coatings) and those that become insoluble after exposure (photo cross-linking, photopolymerising or
photoinsolubilising coatings). A photo cross-linking formulation can be used to produce a negative lithographic plate or a
positive screen-printing stencil because the mechanism of ink transfer in lithography is different from that in screen printing.
The third system applied in information recording in graphic arts is a photophysical one: photoconductors can be used as
active agents that are several orders of magnitude more sensitive than photochemical materials.

3.2.2. Chemical Fundamentals

3.2.2.1. Photosolubilising Systems
Derivatives of 1,2-Naphthoquinone Diazide (2-Diazo-1-oxo-1,2-dihydronaphthalenes). In most cases in which
photosolubilising systems are used, a derivative of 1,2-naphthoquinone diazide is the active substance. Photolysis of this
class of substances was elucidated by SÜS in 1944 [237]. The mechanism corresponds to the Wolff rearrangement of a 1,2-
diazoketone. The first step is the light-induced elimination of nitrogen from 1 with the intermediate formation of a carbene (2),
which in turn rearranges to give a ketene (3). Layers of this type usually contain traces of water, which are sufficient to
convert this ketene to the indenecarboxylic acid (4). Substance 4 probably rearranges to give the isomeric acid (5).

The exposed layer is brought into contact with an aqueous alkaline salt solution (typically Na2SiO3) for development. In this
process, the indenecarboxylic acid formed in the exposed areas dissolves; in the unexposed regions, the naphthoquinone
diazide remains unchanged on the carrier.

For practical reasons, the 4- or 5-sulfonic or carboxylic acid esters of naphthoquinone diazides with phenol derivatives are
usually employed instead of the parent substances. These esters are synthesised by reaction of the sulphonyl chlorides or
carboxylic acid chlorides of naphthoquinone diazide with the corresponding mono- or polyfunctional phenols (e.g., 4-
cumylphenol [238] or the condensation product of pyrogallol and acetone [239].

Naphthoquinone diazide derivatives are not the only substances present in light-sensitive layers because, despite their
relatively high molecular mass, they are not good film formers. Instead, they are usually employed in combination with a
polymeric binder having good film-forming properties. Novolaks have proved especially useful. The solubility of the light-
sensitive layer is greatly reduced by interactions between the naphthoquinone diazide and the novolak [240]. These
interactions cease after exposure of the naphthoquinone diazide. In this way, the difference in solubility between exposed
and unexposed areas is greater than it would be if the naphthoquinone diazide were used alone.

Other Systems. Two related systems are described briefly here because they have attained a certain practical status.

Photolysis of o-Nitrophenylcarbinol Esters. The carboxylic acid esters of 2-nitrobenzylalcohol, for example, undergo a light-
induced rearrangement and cleavage to 2-nitrosobenzaldehyde and free carboxylic acid when irradiated with shortwave UV
light of 300 – 400 nm. On treatment of the exposed recording material with aqueous alkali, the free carboxylic acid and the
products resulting from 2-nitrosobenzaldehyde dissolve from the exposed areas [241], [242].

Photoinitiated Hydrolysis of Acetals and Orthoesters. A whole series of substances is known which eliminates acids upon
photolysis. Examples are diazonium salts of halogen complex anions [243], certain naphthoquinone diazide derivatives [244],
and trichloromethyl-s-triazines [245] or -oxadiazoles [246] with chromophoric substituents. If these photolytic acid donors are
combined with easily hydrolysable compounds (e.g., acetals or orthoesters), coatings are obtained that are highly sensitive to
light because of the catalytic action of the acid formed [247-251].

3.2.2.2. Photoinsolubilizing (Photocuring) Systems

Many more photochemical systems are known that become insoluble after exposure to light. The first photochemical
nonsilver materials used to record information belonged to this type of system (asphalt, gelatin-dichromate).

page 42 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Several different mechanisms may be utilised in the formulation of such systems [252]. However, only a few are of
commercial relevance today:

1. photoreduction of chromium(VI) compounds,

2. photolysis of aromatic diazonium salts,
3. photolysis of p-diazoiminoquinones,
4. photolysis of aromatic azides,
5. photodimerization of activated double bonds,
6. photoaddition of mercapto groups to double bonds, and
7. photoinitiated polymerisation.

Photoreduction of Chromium(VI) Compounds. is of historical interest only [253]. By the middle of the nineteenth century,
alkaline or ammoniacal mixtures of water-soluble organic colloids, such as gelatin, isinglass, starch, gum arabic, albumin,
casein [and later, poly(vinyl alcohol)], and water-soluble dichromates were known to be sensitive to light. The colloids lose
their water solubility in the process and retain only a more or less pronounced swelling capacity; i.e., they are cured. The use
of this process is seldom if ever encountered today because of the environmental issues associated with the use of
chromium (VI) compounds.

Photolysis of Aromatic Diazonium Salts. It was found as early as 1931 that high molecular mass polyfunctional aromatic
diazonium salts can be used advantageously in the formulation of light-sensitive printing plates and other photomechanically
processible materials [254]. Due in part to the ease of development in aqueous based developers, it remains today the most
commonly utilized process in commercial negative working plate coatings.

Diazonium salts of this type can be produced by the condensation of p-diazodiphenylamine derivatives with lower aldehydes
in an acidic medium and precipitated as sulfonates. Activated diethers and bismethylol compounds can also be used instead
of aldehydes [255].

The photochemical reactions occurring during exposure have not been elucidated fully. However, the higher molecular mass
diazonium compounds can be assumed to form similar reaction products on exposure as lower molecular mass compounds,
which can be investigated more easily. Nitrogen is known to be eliminated from diazonium compounds on exposure to UV
light. The aryl cation produced can react with nucleophilic substances in the layer to form substitution products such as
phenols or phenol ethers. These products have a much lower solubility in polar solvents. However, it is unlikely that this
process occurs exclusively. Diazonium compounds in the binder matrix can also be photolyzed to give radicals, as in thermal
decomposition. The resulting radicals can cause cross-linkages by hydrogen abstraction and similar reactions, which also
lower the solubility of the matrix. Both mechanisms result in a product that is less soluble than the starting material.
Unexposed areas are removed from the carrier during development, and the exposed portions are retained.

Photolysis of p-Diazoiminoquinones. p-Diazoiminoquinones (6), with R1, R2, R3 = alkyl, undergo ultraviolet-induced
nitrogen elimination to form carbenes, which can react further to

produce higher molecular mass substances [256]. These substances are less soluble in organic solvents and dilute alkali or
acid than the parent diazo compounds. Synthesis of p-diazoiminoquinones can be achieved, for example, by starting with 1-
chloro-4-nitrobenzene-2-sulfonyl chloride [257].

Photolysis of Aromatic Azides. Low molecular mass aromatic azides (e.g., 4,4′-diazidostilbene-2,2′-disulfonic acid) can
also be used in the photochemical curing of organic colloids. On exposure to UV light, azides eliminate nitrogen to form
highly reactive nitrenes, which can react further by hydrogen abstraction, insertion in C–H bonds, reaction with olefinic double
bonds in the matrix, or dimerization. Cyclorubber ( ) may be used as the binding agent.

Polyfunctional azides are the active substances of choice because they lead to a higher cross-linking density and can also be
sensitised [258]. During development, which is often carried out with organic solvents, unreacted azides and the binding
agent are dissolved from the carrier. Lithographic plates made with such materials are characterised by long run length under
very harsh press conditions. However the increasing trend towards aqueous developers in recent years means that the use
of such plates has diminished rapidly.

Photodimerisation of Activated Double Bonds. , -Unsaturated carbonyl compounds, such as chalcones (phenyl styryl
ketone derivatives) and cinnamic acid derivatives are accessible to photochemical dimerisation. Eastman Kodak made a very
detailed study of the photochemical behaviour of cinnamic acid esters in the 1950s [259]. The dimerisation reaction
apparently starts from trans-cinnamic acid and proceeds via the excited triplet state of the cinnamate molecule and radical
intermediates to give derivatives of -truxillic acid (diphenyleyclobutanedicarboxylic acid). Radical side reactions complicate
the course of the overall reaction. This reaction mechanism offers numerous possibilities for sensitisation with triplet
sensitizers. Thus, the light-sensitive region of cinnamates, which lies below 350 nm, can be extended to > 400 nm [260].

Polymeric cinnamates, preferentially poly(vinyl cinnamate), are used widely in industry to increase the difference in solubility
between exposed and unexposed portions of the recording material. Poly(vinyl cinnamate) can be produced from poly(vinyl

page 43 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
alcohol) and cinnamoyl chloride. Another possibility is the use of polyesters from p-phenylenebisacrylic acid and dihydric
alcohols. A single cross-linking step between two polymer chains results in an approximate doubling of the molar mass and
hence, a dramatic decrease in solubility. Development is usually carried out with organic solvents and again their use today
is diminishing as the demand for solvent-free developer increases.

“Thiol – Ene” Addition. Photochemical addition of mercapto groups to terminal double bonds of allyl or vinyl ethers also
leads to an increase in molar mass and a decrease in solubility. This effect is especially pronounced when polyfunctional
mercaptans react with polyfunctional olefins because a high cross-linking density is generated in the process. Initiators play
an important part in starting the reaction. They absorb light and lead to the formation of mercapto radicals, which then add to
the double bond. In general, light-sensitive materials of this type are viscous liquids which are applied to the desired carrier
by the user [261].

Radical Photoinitiated Polymerization [262]. Radical photoinitiated polymerisation was first suggested by GATES [263] in
1945 and then investigated in detail by Du Pont. The reaction principle is based on the photochemical generation of radicals
that initiate polymerisation of a reactive monomer [264]. Only the initial reaction is dependent on light; the chain reaction
represents an intensification mechanism, and thus the quantum yield of the total reaction > 1. Hence, a system of this type
should be very sensitive to light. However, unlike the photoreaction of cinnamates in which a single reaction step causes a
doubling in molar mass, each step in the chain reaction increases the molar mass only slightly. Consequently, many chain
propagation steps are required before the difference in solubility between image and nonimage areas is sufficient for
development.

Nonvolatile low molecular mass polymerisable compounds, as well as high molecular mass compounds, can be used as
monomers. Polyfunctional compounds (e.g., esters of acrylic or methacrylic acid, such as pentaerythritol triacrylate or
trimethylolpropane triacrylate) are preferred for yielding extensive three-dimensional cross-linked networks [265]. Various
substances and combinations have been proposed as initiator systems, e.g., aromatic carbonyl compounds, heterocycles,
azo or diazo compounds, mercaptans or disulphides, and photoredox systems [266].

A light-sensitive coating that consists only of monomer and initiator is soft, sticky and flowable. Addition of a sufficient amount
of suitable, polymeric binder yields a nonslip, stable coating.

These properties can be exploited directly for imaging. A coating of this type can lose its surface stickiness after exposure.
The sticky image can then be made visible by application of a colored powder [266].

Unhindered propagation of the photoinitiated chain reaction in the exposed areas can be maintained only if atmospheric
oxygen is largely excluded from the reaction site. This may be achieved by using a covering layer of poly(vinyl alcohol) or a
covering film (e.g., of polyester) [266].

3.2.2.3. Photophysical Principles of Image Generation

Electrophotography [267], [268] (see Section Electrophotography). The basic principle of electrophotography is the
photoconductor effect. Photoconductors are compounds that are insulators in the dark (conductivity 10–13 S/m), but whose
conductivity increases by a factor of ca. 108 after exposure to light. Lithographic plate systems based on this principle were
very popular during the 1980's and early ‘90's, but their use has diminished significantly of late.

For imaging, the photoconductor layer is charged electrostatically in the dark under a high-voltage corona. Thereafter, it
behaves like the dielectric material of a charged condenser. During exposure, pairs of charge carriers are generated in the
photoconducting layer. These pairs are separated under the influence of the electric field and migrate to the two poles of the
condenser, and the voltage in the exposed areas collapses. A latent charge image remains in the unexposed areas, which
can be developed in different ways. Resin particles with an electrically opposite charge can be applied to the latent charge
image in powder form (dry toner) or as a dispersion (liquid toner). These particles are deposited only in the charged image
areas. The toner image can, in principle, be processed further in different ways (Fig. 58).

Figure 58. Further processing of a toner image generated by electrophotography

a) Charged plate; b) Exposed plate; c) Developed plate; d) Molten toner; e) Hydrophilized plate; f) Uncoated plate

The light sensitivity of electrophotographic recording materials is 103 – 104 times higher than that of photochemically acting
materials described above [236]. As with silver halide layers, the spectral sensitivity to light of electrophotographic materials
can be extended to cover a wide range by the use of sensitiser dyes. Thus, the sensitivity of the recording materials can be
adjusted very specifically to various light sources (argon-ion laser 488 nm, helium – neon laser 632 nm, and tungsten lamps).

Different types of photoconductors are known: Zinc oxide is the most common inorganic photoconductor. It is applied as a
pigment dispersion in a binder matrix. In the past, layers on paper were used in office copying; today, they are applied in
small-size offset printing [269-271]. Sputtered cadmium sulphide layers are also used, e.g., as electrophotographic film on
polyester film and as light-sensitive material on specially prepared steel plates for the production of color-proof prints [272].

Furthermore, a large number of organic photoconductors are available, e.g., nitrogen heterocycles such as poly-N-
vinylcarbazole and derivatives of oxadiazole [273], oxazole, thiazole, thiadiazole, or imidazole with suitable electron donor
and electron acceptor substituents.

Like zinc oxide, organic pigments can also be used in binder-containing dispersions for the production of photoconductive

page 44 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
coatings. Copper phthalocyanine is especially suited to this application. The different crystal modifications of copper
phthalocyanine vary somewhat in their electrophotographic effectiveness [274].

The properties of photoconductors, i.e., the formation of pairs of charge carriers in light and the migration of these charges in
an electric field, are not necessarily combined in the same chemical substance. Cases are known in which two substances or
coatings are combined to achieve the photoconductor effect. The charges are generated in one of the substances, whereas
the other is responsible for charge transport (e.g., perylene pigments and low molecular mass photoconductors) [275].

Photolysis of Silver Halides. The principle of silver halide photographic film is also applied in different forms to record
information in the graphic arts [276].

On the one hand, the corresponding films can be used in reprophotography; on the other hand, silver halide coatings can be
applied to layers of lower light sensitivity. In this case, the higher sensitivity of the silver halide coating is exploited for image
production, so an image-carrying mask is obtained. Subsequent overall exposure transfers the information in the mask to the
layer below, which better fulfills the requirements of printing.

Furthermore, the principle of silver salt diffusion (see Section Silver Process) can be used for the production of proofs and
printing plates (see Photography).

The high sensitivity of the silver salt diffusion transfer system means that printing plates can be produced directly from hard
copy (via a camera) or, more usually nowadays, from digital copy (via a laser platesetter) without the requirement for a film
intermediate. Lithographic printing plates based on the silver diffusion transfer process are described in Section Modern
Lithographic Printing Applications. They hold a significant share in both Direct-to-Plate and Computer-to-Plate markets.
There are two types of assembly, two-sheet systems where the negative light-sensitive donor is separate from the positive
receiver sheet, and single-sheet systems where the donor and receiver layers are incorporated onto the same substrate
carrier. Products based on single-sheet assemblies are dominant. The substrate can be reinforced paper, polyethylene
terephthalate or lithographic grained and anodised aluminium.

3.3. Color Proofing [277]

3.3.1. Basics
Four-color prints are produced by the subsequent printing of the three subtractive primary colors (cyan, yellow, and
magenta). All color nuances, including black, can be reproduced in this way. However, black is usually added as the fourth
color to increase contrast or, as an independent color, to reproduce deep shadows [278].

It is necessary, at different points in the production process of a four-color print, to check the image quality prior to printing or
give the customer a reliable impression of the final printed result. Press proofs are still used today when the highest fidelity of
reproduction is demanded or when a medium-to-large number of proofs is required that must be shown to people who are
involved in decision making at different places.

In many other cases, the same purpose can be achieved by making a single four-color copy from the screened color
separation films, the color proof. Thus, the technician and the customer can be given an advance visual impression of the
colors before production of the final printing plates and before printing. In principle, this can be achieved on a color monitor
as well (soft proof). In this case, screened subtractive primary colors must be reproduced by using additive primaries. The
latter are emitted by luminescent substances placed at the monitor in a raster arrangement. This poses a problem.

In contrast to soft proofs, a color copy on paper or other materials is known as a hard proof. The appropriate materials are
exposed in contact with the color separation films. A negative color-proofing material must be used for negative films and a
positive color-proofing material for positive films.

The color pigments or dyes applied in the production of color proofs must correspond as closely as possible to the pigments
in the printing inks to be used later in printing. Because standard colors vary in definition in different countries, different color-
proofing systems must be employed with regard to color reproduction. Even the light source used to judge color proofs is
standardized.

3.3.2. Applications
Color-proofing materials are used [277] in various applications: (1) for quality control checks in the production of color
separation films and as indicators for necessary corrections, (2) as internal quality proofs during production, (3) to check
pictures from different sources for comparable quality of reproduction, (4) as layout proofs, (5) as color proofs for customer
approval, or (6) as guideline for the printer at the printing press.

In commercial printing color proofs may be more economical than pressprinting if only a very limited number of large-size
four-color images is required as drafts for advertising agencies, and in the production of animated films.

The quality of reproduction required for various applications differs. For example, a layout proof need not be as exact as the
proof shown to a customer for approval. However, for simulation of the highest quality, the characteristics of the printing
process to be used must be taken into account in producing the color-proof separations. This means that other colors and,
therefore, other color-proofing results are required for offset printing, for flexography, or for gravure proofing with offset
means, but even then the characteristics of a special offset press must be simulated in the color proof. Thus, quality levels
differ even among color-proofing processes.

On the one hand, the (older) overlay systems result in four subtractive color images on transparent plastic film which when
placed one upon another in register give a four-color picture. On the other hand, in the (more recent) single-sheet systems
the four-color picture appears on an opaque white background (ideally, on the final printing paper).

page 45 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.3.3. Overlay Systems
Overlay materials generally consist of a transparent (film) substrate which is coated in the factory with one or more dyed
and/or light-sensitive layers. The light sensitivity is caused by photosolubilizing (in the case of positive transparent originals)
or photocuring (in the case of negative transparent originals) compounds in the coating. After respective exposure through
the four corresponding screened color separation films and after development, four imagewise colored films are produced
which, when accurately registered, give a colored impression of the expected result.

The advantage of this procedure is that it also allows examination of a combination of single films. So the printed result can
be compared with the set point at all stages in production. The disadvantage of overlay systems is that four separate films
with their imperfect transparency of the film substrate cause a marked graying of the picture and background.

Trade Names. Overlay systems available on the market differ in structure and processing procedures. The overlay proof
Color Key, produced by 3 M, is available only for negative copies [279]. It has a light-sensitive and a colored coating on the
carrier film. The more soluble (in this case, the unexposed) areas of the coating are removed during development in the
MR 424 processor.

Hoechst – Celanese offers an overlay system called Colorlink which is available in both a negative and a positive version
[280], [281]. In these products, the light-sensitive diazo layer contains the dye as well. Here, too, a developing liquid and a
processing device (CF 262) are available. The developing liquid is an aqueous solution.

DuPont has developed a product that does not require a liquid developer. This product, called Cromacheck, is available only
for negative copies and has a sandwich structure: (1) opaque film, (2) light-sensitive colored photopolymer layer, and
(3) clear film. After exposure, the exposed areas adhere more strongly to the clear than to the opaque film. Conditions are
exactly reversed for the unexposed portions. Image areas are then separated from nonpicture areas by physical separation
of the two films (peeling apart). The color image is present on the clear film. An advantage of proofing with Cromacheck is its
quickness [282].

3.3.4. Single-Sheet Systems

In comparison with overlay systems, the advantage of single-sheet products is that the proof not only is made of one piece
like a press print, but can also be handled like a print.

A distinction must be made between systems that are produced as colored layers and those in which the color or pigment is
applied during processing. In the former, the color density (and the color hue) is predetermined by the production process. In
the latter, the color density can be influenced within certain limits during processing.

3.3.4.1. Precolored Systems

In the simplest precolored systems, the user applies to a substrate a standard pigmented or colored, light-sensitive
dispersion or solution that hardens on exposure to light. The material is then dried, exposed, and developed (usually with tap
water). The procedure is repeated with the four process colors to give a four-color proof.

This system has many advantages including low cost of materials and miscibility of colored liquids plus the fact that the color
density can be varied to some extent by varying coating conditions.

Trade names are Watercote, Kwik-Proof, and dr Color from the Direct Reproduction Corp. These systems differ mainly in the
method of coating. Two other products of this kind, Aqua Color Proof from Graph Mark and Quadracolor from Castcraft
Industries, also use porous carriers, such as paper, after the pores have been impregnated with a special covering layer to
prevent penetration of the light-sensitive formulations into the paper.

Another approach involves presensitized products. These offer not only a guaranteed constant coating quality but also
greater ease of processing. Here, the color density of the final product cannot be manipulated.

Trade names are 3 M Transfer Key, 3 M Matchprint II [283], Hoechst – Celanese Colorlink Pressmatch [284], [285], and Fuji
Photo Film Color Art. Apart from Transfer Key, which is offered only as a negative system, all other types are offered in both
negative and positive versions.

The light-sensitive principles used are the photolysis of quinone diazides, photohardening of diazonium salts, or
photoinitiated polymerization.

Another single-sheet system that is processed similarly, Agfa – Gevaert Agfaproof, was introduced in 1986. It is based on the
silver halide technology and is used for the production of four-color proofs with negative transparent films.

Despite the somewhat important differences in application, the processing principle of these systems can be simplified as
follows: The light-sensitive layers, which are often embedded between two foils for transport and storage, are first laminated
onto a receiver material. This is achieved via a special adhesive layer that is activated by heat or pressure, the latent
adhesiveness of the light-sensitive layer, or wet transfer (Agfaproof). The layer is subsequently exposed in contact with the
corresponding color separation film and then developed. This process is repeated four times to give a four-color proof.

Only in the case of Color Art from Fuji Photo Film is the light-sensitive layer exposed and developed before being transferred
to a receiver sheet. The advantage of this system is that all four process colors can be exposed simultaneously side by side.
However, accurate transfer of the single color pictures causes certain difficulties.

In some of the products mentioned above, the four-layer structure can be transferred from the receiving sheet to printing
paper. This gives a particularly good simulation of the four-color print.

page 46 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.3.4.2. Systems With Colors Generated During Processing
Ektaflex PCT from Eastman Kodak has a silver salt diffusion color film structure in which the colors are generated during the
development process (see Section Color-Copying Materials Based on Dye Diffusion Chemistry; Photography).

Photochemical Imaging. As described in Section Photoinsolubilizing (Photocuring) Systems, not only can radical
photoinitiated polymerization be used to prepare materials that are developed by dissolving the nonpolymerized areas off the
carrier, but can also be used to formulate the starting material so that it has a sticky surface which imagewise loses its
stickiness during exposure. In an additive development process, the layer is then treated with a colored powder that adheres
permanently only to the sticky areas of the layer [266]. This principle has been used by DuPont in the concept of its positive
acting Cromalin system [286].

The toner can be applied manually with pads. However, more uniform results are obtained when the toner is applied in a
toning device (e.g., ATM III). The application of the color powder can be totally avoided by coating the toner on a polyolefin
film which is brought in contact with the sticky image. In this way, the toner layer is transferred imagewise to the exposed
proof material without soiling the machine with toner powder [287].

A negative version of the Cromalin process has been developed as well. Here, removal of the cover film from the exposed
material leaves the exposed areas sticky. This material is processed similarly to positive material.

Additive development is also used in another negative proofing process called Spectra, which was developed by KEUFFEL
and ESSER [288] and is marketed by Polaroid. The light-sensitive material consists of a polyester film coated with a sticky
polymer layer, a light-sensitive naphthoquinone diazide layer, and an overcoat. On imagewise exposure and subsequent
development, the sticky layer is uncovered in the exposed areas. The quinone diazide layer is completely solubilized in a
second exposure step, which is integrated in the processor. After additive development with color powder, the four color
pictures are developed again and finally successively transferred in register to a receiving material (e.g., the printing stock).

Electrophotographic Imaging. The advantage of electrophotographic imaging lies in the extreme light sensitivity of
electrophotographic materials. An added advantage in proofing is that color density and screen dot diameter can be
controlled independently.

The disadvantage of this system is that imaging and reproduction quality depend to a great extent on the developer, usually a
dispersion of fine, dyed resin particles in an aliphatic hydrocarbon (isopar). The key issue is to maintain the stability of this
dispersion over a long period under varying climatic conditions.

The first process of this type, the Remak method, was developed by Research Laboratories of Australia. However, the
method was not adopted commercially because of the extreme sensitivity of the employed zinc oxide materials to
atmospheric humidity.

Kimoto of Japan has put a similar proofing system, Kimofax, on the market. This method uses an organic photoconductor
and four colored dry toners. However, it can be employed only to proof simple line work (e.g., in map printing).

In the late 1970s, a high-resolution electrophotographic recording material was made available that used photoconducting
layers of cadmium sulfide [243] which were sputtered on steel at high frequency (KC-crystalplate). This material, together
with the high-resolution color toners of the Remak type, is ideally suited to the production of an electrophotographic proofing
system. Considerable development in the area of electrophotographic proofing occurred as a result of collaboration between
Coulter Systems Corporation of the United States and Stork of the Netherlands. The final version of the imaging cycle,
conducted in a largely automated device, consists of the following steps: (1) charging (ca. 35 V) of the KC plate, (2) contact
exposure, (3) toner application, (4) transfer of the toner image to an intermediate, and then (5) transfer to the final receiving
carrier (e.g., the printing paper). The image quality obtained is comparable to that of photochemically produced proofs.

In 1986 Eastman Kodak introduced another electrophotographic proofing system called Signature. Instead of working with a
double toner transfer, this system uses an organic photoconductor applied to a transparent, conducting carrier which is
exposed from the back through a color separation film. All four subtractive color pictures are collected on the photoconductor
and transferred to coated paper in a laminator outside the imaging device.

The equipment requirements of all electrophotographic proofing systems are relatively high. Because the light sensitivity of
the materials used is very high, they must be processed in a closed device to avoid working in a darkroom.

3.3.5. Digital Proofing

Color separation is carried out at present in color separation scanners. Color data are stored in an electronic memory and
can be recalled and inspected on a high-resolution color screen (soft proof). However, an additional proof on paper or other
substrate (hard proof) is often needed. A satisfying technical solution to this problem is not yet marketed. However, a number
of possibilities exist, and various announcements have been made that new systems will be introduced in the near future. In
principle, the approaches being considered are silver halide materials, photochemical systems, electrophotography (see
Section Electrophotography), and other digitally controlled imaging methods (thermotransfer, ink jet, see Sections
Thermographic Printing and Ink-Jet Printing).

The use of silver halide (i.e., color film materials) as continuous tone materials for the simulation of screened printed images
is possible only to a limited extent. Polaroid is developing a silver – halide based system, which can be directly exposed in a
scanner. The 3 M Company has proposed the use of photochemical systems for the production of digital color proofs
(Dataproof). Ultraviolet lasers are employed for imaging.

Digital Electrophotographic Color Proofing. In 1986 DuPont announced a new digital electrophotographic color proofing
method DDP (direct digital proofing). Digital data from the scanner are stored in an optical memory and recalled into the DDP

page 47 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
system. The charge images are developed with liquid toner and transferred to the printing paper.

Eastman Kodak is working on the adaptation of the Signature method to digital data input. Stork is developing a scanner that
can be employed for the digital exposure of the inorganic photoconductor in use.

Digital Ink-Jet Color Proofing. The ink-jet method can be used for the multicolored digital imaging of plain paper. This method
has the advantages of low cost and high speed. On the other hand, color matching, uniformity of larger color areas, and
reproduction of conventional screened structures have not yet been perfected, although improving rapidly. This method can
be employed for position and layout proofing. It is manufactured by Iris Graphics.

Digital Thermotransfer Color Proofing. In the thermotransfer (see Section Thermographic Printing) method, the color
substance is transferred thermally point by point from a coated paper to a receiver sheet. The digitally controlled transfer is
repeated four times to give a four-color image. Transfer is achieved by sublimation of the dye or by contact transfer of a
thermoplastic substance under the influence of a thermo head. The result resembles a continuous-tone picture and has a dot
density of 120 dots per centimeter. The imaging device is manufactured by Dainippon Printing Co., Tokyo.

3.4. Platemaking
Printing is defined, according to the DIN standard [289], as “the reproduction in any quantity of a presentation in text and/or
picture form by transfer of printing ink or coloring substances to the material to be printed using a printing form”. Each of the
four printing processes uses a different kind of plate [214].

In lithography, image and nonimage areas are essentially in the same level, but they differ in their physicochemical behavior.
The image areas are oleophilic; the nonimage areas, hydrophilic.

In letterpress printing, the image areas are raised above the surface level of the plate, and the nonprinting areas are lowered
to different extents.

In gravure printing, the image areas consist of engraved or etched wells from which ink is transferred. The surface of the
gravure cylinder acts as nonprinting area.

In screen printing, nonprinting areas (i.e., ink-impermeable areas) are protected by a stencil. Areas not covered by the stencil
represent image areas.

In the Federal Republic of Germany these processes were used in 1986 for printing as follows: 57 % lithography, ca. 26 %
letterpress, and ca. 17 % gravure printing. Screen printing is used mainly for textiles and in the production of printed circuits
(see Section Printed Circuits (Printed Circuit Boards)) [290].

3.4.1. Lithography (Planographic Printing, Offset)

Lithography was invented by SENEFELDER in Munich in 1797. Because he used limestone from Solnhofen as the carrier, this
method is called “writing on stone” or lithography. The American W. RUBELS made lithography an industrially useful printing
process, by transferring the printing ink from the plate to an intermediate carrier, the rubber blanket, and then from the
blanket to printing paper (offset).

Because of its technological characteristics, offset printing is widely employed worldwide in the production of high quality
one- and multicolor prints for medium runs in advertising material, calendars, art prints, books, brochures, newspapers, and
packaging. Runs ranging from 10000 to 250000 are typical, but under certain press conditions, much higher runs are
achievable.

An important characteristic of offset printing is that platemaking is simple, quick, and economical. On a suitable hydrophilic
carrier (e.g., a plastic film coated with silicates in a binder) oleophilic image areas can be made by hand or with a typewriter
for simple black and white line work.

For more demanding work plates with a light-sensitive coating are used. Presensitized (PS) plates of this type are produced
by various manufacturers and are available commercially. World sales in 2001 were ca. 360 × 106 m2.

The photomechanical method, which uses subtractive development, is described in Section Application. Once the information
for printing exits, either in digital (computer) or analogue (film) form, it can be transferred to the plate by exposure in a matter
of minutes. The exposed plate becomes a printing plate on development. A lithographic plate can be prepared in ca. 5–8 min.

Wipe-on plates are still being used in some countries, mainly the United States and East-bloc countries. Worldwide
consumption in 2001 amounts to ca. 40 × 106 m2. These plates are not precoated.

Printers buy the uncoated carrier, and then coat it with a light-sensitive solution shortly before imaging and then dry it. Wipe-
on solutions are supplied only for negative plates. The resulting coated plates usually have a storage life of a few days before
they are no longer developable. Once exposed, these plates are developed with an emulsion lacquer. The lacquer dissolves
the unexposed light-sensitive coating (<1 µm thick) from the substrate, and deposits resins and pigments on exposed areas
to improve the stability of these image areas during the printing process.

All light-sensitive lithographic plates, whether wipe-on or presensitized, have a simple structure: (1) substrate with sufficiently
high mechanical strength and a hydrophilic or potentially hydrophilic surface, and (2) a light-sensitive layer.

3.4.1.1. Substrates
Demands made on substrates in lithography arise from the stresses to which the plate is exposed during both platemaking

page 48 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
and printing [291]. During development, the exposed plate usually comes into contact with acidic and alkaline salt solutions,
and occasionally with organic solvents. The surface of the plate is often brushed or sponged. The finished plate is punched,
bent and finally mounted on the metal cylinder of the press. During each printing cycle, the surface of the plate comes in
contact with an acidic (pH ≈ 5) dampening solution and then with solvent-containing ink, which is transferred from the plate to
paper via the rubber blanket.

For these reasons, the substrates must be resistant to kink marks, creases, and scratching during use. The material must be
totally flat and finally should have a high dimensional stability, as well as high abrasion resistance in the press.

Great demands are also made on the chemical resistance of the surface to acidic and alkaline agents. Not only should the
surface be stable to corrosion during use, it should also have a well defined microroughness and a high hydrophilicity for
spreading the dampening agent. The hydrophilicity should not be lost on repeated and prolonged contact with printing ink.
Microroughness also improves adhesion of the printing layer.

The presslife expected of an offset printing plate varies between some ten and several hundred thousand runs depending on
the substrate and the printing layers.

Special water-resistant paper with a hydrophilic coating can be employed as substrate for runs of a few hundred or thousand.

Composite materials made of paper and aluminium foil or plastic film (usually polyester film with a hydrophilic coating) are
suitable for runs up to ca. 10 000. These materials have better wet strength and dimensional stability.

Highly pure (99.5 %) or low-alloy (ca. 98.5 %) aluminum is generally used as the substrate for lithographic plates. The
surface quality and material properties of this litho-type aluminium, used in the production of offset plates, are stipulated by
extremely narrow specifications. At first, the raw litho-type aluminium was processed in sheets only; however, the processing
of coils is the method of choice today. After this treatment, the final product is generally an aluminium web that is finished on
one side and often coated in line. However, there also exists a market for plates that are grained and coated on both sides
(Mainly in the USA), this market is declining rapidly. These plates are used for shorter runs (up to 20000–30000) and
average printing quality.

Surface Graining. Mechanical graining is one of the oldest methods used for surface finishing. If the aluminium surface is
treated with rotating wire brushes, it acquires a silky luster. The surface structure then shows a depth of roughness of R
Z = 3 – 6 µm, with a marked anisotropic roughness in the direction of brushing. When abrasives such as suspensions of
quartz, corundum, or pumice powder in water are used, the carrier surface acquires an isotropic roughness, with RZ values of
2 – 10 µm. The abrasive suspension is either pressed onto the surface of the aluminium sheet, with the help of porcelain
balls, or applied with rotating nylon brushes, or sprayed on under high pressure [292].

Electrochemical graining has been used increasingly since the mid-1960s. The surface quality of the aluminum plays an
important part in this method. The raw aluminum surface is first degreased and then uniformly corroded by using an a.c.
treatment in dilute nitric acid, hydrochloric acid, or mixtures of these with other acids. This process of graining is a kind of
pitting corrosion which attacks the entire surface of the aluminum leaving a very uniform rough surface [293], [294].

The depth of roughness can be controlled within wide limits (2.5 – 9 µm) by varying the electrical parameters. Both
processing speed and the uniformity of the resulting substrates depend on temperature, flow rate of the electrolyte, geometry
of the tank, and other parameters. Combination processes have also been used in graining. In these processes, mechanical
graining is followed by electrochemical treatment [295].

Graining with wire brushes produces the lowest increase in surface area of the plate, and electrochemical graining the
highest. This influences not only the quality of reproduction and the optical resolution but also the printing properties (water
spreading and length of run). The finer the graining of an aluminum surface, the lower is the mechanical resistance. With
wire-brushed aluminum, a second hardening step is not required. However, an anodic oxidation step is essential for
electrochemically grained aluminum.

Anodic Oxidation. Anodic oxidation involves conversion of the uppermost layer of the aluminium workpiece to aluminium
oxide, without affecting the surface topography too much [296]. Two procedures are primarily used for this electrochemical
reaction in which aluminum is connected as the anode: anodic oxidation in sulphuric acid [297] or in a phosphoric acid
electrolyte [298]. The former leads to smaller pores and thicker oxide layers than the latter. Mixtures of the above-mentioned
acids and multistep processes are described in the patent literature [299].

In general, not more than 5 g of oxide is produced per square meter; depending on the roughness of the surface, this
corresponds to a layer thickness of ca. 1 µm. The oxide formed during anodising is X-ray amorphous and an extremely
strong adsorbent. The properties of the substrate are greatly improved during this step, i.e., the adhesion to the light-
sensitive coating; resistance of the surface to chemicals; its hydrophilicity, hardness and abrasion resistance, and thus the
potential of running length. However, the extent to which the aforementioned aspects can be effective depends on process
parameters and on the process itself.

Posttreatment. Posttreatment of the oxide with certain hydrophilising solutions has proved advantageous in the production
of PS-plates. The main purpose of this treatment is to improve the hydrophilicity of the aluminum oxide surface. However, the
adhesive properties, developability, and shelf life of the plates are influenced at the same time. Solutions of the following
substances have been proposed for this purpose: alkali silicates [300], phosphoric acids [301], hexafluorometalates of group
4 elements (titanium, zirconium, hafnium) [302], heteropolyacids, oxo acids of pentavalent phosphorus [303], and hydrophilic
colloids [304]. All of these compounds coat the surface of the aluminium oxide to generate a very polar covering layer, which
improves the hydrophilicity of the surface.

page 49 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.4.1.2. Coating and Further Processing [305]
Preliminary treatment of the carrier is followed by in-line coating of the plate in a fully continuous process. A solution of the
components of the light-sensitive coating in a more or less volatile solvent mixture is applied to the carrier in a defined wet
film thickness. The coatings are generally applied with a fishtail die, with a roller coater, or by spraying. The thickness of dry
coatings is 0.3 – 5 µm.

The line speed in modern plants is typically in the range 40–80 m/min. The length of individual steps is adjusted to this
process speed. After drying, the web is automatically inspected for coating defects. It is cut apart, and defective plates are
removed. Cutting of the plates requires special attention. On the one hand, the plates must be cut to size with an accuracy of
some tenth of a millimeter. Plate measurements must also be able to be changed quickly and accurately, which is not an
easy task with these high process speeds. On the other hand, the plates must be cut at right angles, and finally, no burrs
should be produced during cutting because these would cause problems for the customer. Presensitized lithographic plates
are available in many sizes up to ca. 3 m2 (1500 × 2000 mm). The finished faultless plates are stacked by using a suitable
interleaving paper and packed in lightproof wrapping paper (if necessary, after shrink-film packaging).

The shelf life required for these presensitised lithographic plates depends on the extent of their distribution. Also, the more
diverse the climatic conditions under which the plates are to be stored and used, the more diverse are the requirements.

3.4.1.3. Processing of Presensitized Plates

The light-sensitive layer applied in the coating step is a mixture of numerous components, which is formulated to meet the
requirements of the techniques used in plate manufacturing and to fulfil the demands of the customers in using the plates.
For various applications, different types of plate are available.

In principle, similar processing procedures are applicable to all presensitised plates. In the first step, light exposure, certain
selected areas of the plate are photochemically or photophysically changed, leaving the remaining portions unexposed. How
this step is undertaken depends very much on whether the information to be printed is available in digital (computer) or
analogue (film or paper) format.

In analogue format, the method of transfer depends very much on the format of the film or paper original. This step is usually
conducted in a copying frame if an original-sized transparent film is used as the original. The copying frame is evacuated in
the process to make sure that the film original is in close contact with the light-sensitive coating on the plate during exposure.
If an enlargement or a reduction in size of the original is required, plate exposure is carried out through the film-original in a
projection camera. Exposures of especially sensitive presensitised plates can also be made in a projection camera with
opaque originals and reflected light. However, with the advent of Computer-to-Plate technology and digital workflows, the use
of this particular system has diminished significantly in recent years.

Where the information is in digital format, plates are exposed with a platesetter. Here, all areas of the plate are not exposed
at the same time (parallel exposure), but one area is exposed after another (sequential exposure). A focused beam of light,
generally a laser beam, is used to scan the light-sensitive material line by line while being modulated imagewise as shown in
Figure 59. A complete picture is created by the close lining up and partial overlapping of many modulated lines of light.

Figure 59. Image composition at line-by-line exposure

In the case of photosensitive plates, the actinic spectral region is limited at the lower end by the absorption of UV light by
glass and the physiological dangers of shortwave UV radiation, and at the upper end by the sensitivity of the light-sensitive
material to yellow light (room lighting).

In the next step, development, the exposed plate is stabilised in a specific way to make sure that the image information is not
erased by further action of light. This can be achieved by a chemical reaction, by washout development (subtractive), or by
the deposition of a covering substance (additive; see Section Application). Suitable developers are offered and
recommended by plate producers. These preparations are designed specifically for the development of a particular plate and
can usually be employed for that plate only. The plate and the developer form a “processing system.” Development is usually
carried out in a processor in which the developer is applied and allowed to act on the plate in a very specific manner. The
system dependence between plate and developer is in general stronger than that between plate, chemical and processor.

The third step that generally follows development directly is gumming. It serves the purpose of protecting the hydrophilic
nonprinting areas from contamination by fatty substances in the air. An aqueous solution of film-forming hydrophilic colloid
(dextrin, gum arabic) is applied to the plate and dried.

Despite the fact that plates of various makes have these processing steps in common, a number of different ways of
converting a presensitised plate to a finished printing plate still exist. These methods differ in the starting material, on the type
of original copy available (e.g., positive film, negative film, microfilm, paper, or a data carrier); in the time available for
platemaking (e.g., in newspaper printing or packaging); and in the use of the plate (e.g., to print an art calendar, a business
form, or a metal sheet). The chemical systems applied in several process variations for the preparation of conventional
(analogue) plates will be described.

page 50 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Negative Transparent Film. Platemaking for newspaper printing is a typical example of the use of a negative film in full plate
size as the original. Most of the newspapers produced in Europe by the offset principle are printed with the help of
presensitized plates. In the United States however wipe-on plates are still being used widely although this market is also
declining as newspapers move to CtP.

To transfer information on the negative film to the plate so that the positive image is obtained, the coating on the plate
corresponding to the transparent areas of the film original must be converted to the printing areas. This is achieved with a
photoinsolubilising light-sensitive layer. Plates of this type are called negative plates. Negative plates are used in more than
66 % of the printing done in the United States. However, this is declining rapidly as the USA is moving more quickly to CtP
solutions than most markets.

Presensitized negative plates suitable for newspaper printing must be able to be exposed and developed quickly because
many printing plates have to be made just before the start of the presses. Also, these plates must have good wear
characteristics in the press because newsprint has a rough, dusty surface and the web offset presses employed in
newspaper printing have very high paper speed (ca. 10 m/s). The plates should produce up to 150 000 good prints. They
must be anodized.

Various light-sensitive systems are used by individual producers, which may be broadly categorised as either
photoinsolubilizing, photopolymerising, or photocrosslinking systems:

1. Oligomeric diazonium salts are widely employed. In some cases, the coating also contains a special binder, e.g., an
acetal resin. The ease of development, as well as the printing properties of the plate, depend on the modification of the
resin.
2. Photoinitiated polymerization is also used in the production of negative plates. These layers can be built up with
binders containing carboxyl groups and then developed with aqueous alkaline salt solutions. Plates based on this
principle are extremely robust and abrasion-resistant.
3. The reaction products of azide-substituted aromatic carboxylic acid derivatives with polymers having hydroxyl groups
can also be used in negative plate coatings.
4. The same applies to plates with unsaturated polyesters that can undergo light-induced cross-linking. These polyesters
are made from phenylenebis(acrylic acid) and 1,4-bis(hydroxymethylcyclohexane).

All these coating formulations contain dyes or pigments, in addition to the actual active substance, and change color on
exposure. Thus, the user can distinguish between exposed and unexposed plates. The plates also show a good color
contrast between image and nonimage areas after development. The increasing trend on environmental grounds towards
aqueous-based developers in recent years mean that plates of types 1) and 2) tend to dominate in today's market.

Aqueous diazonium salt solutions are used for wipe-on plates. These plates are not dyed and have very thin coatings.
Because their shelf life is limited, they are always prepared shortly before use. Wipe-on plates are developed with an
emulsion lacquer that has a double function: On the one hand it dissolves the unexposed areas. On the other hand, a resin is
deposited from the lacquer on the exposed images to strengthen the thin exposed layer.

Presensitised negative plates are produced by Agfa-Gevaert, Fuji Photo Film Co., KPG, Lastra, and others. Wipe-on plates
are produced by American Litho, Anocoil, Citiplate, and Western Lithotech.

Positive Transparent Film. In vast parts of Europe and Asia, the use of positive transparent film as the original copy is the
method of choice for platemaking—to print advertising, pamphlets, brochures, and art prints, in most cases in four-color print.
If the transparent areas of the film original are to become the nonprinting areas after plate development, a presensitised
offset plate must have a photosolubilising coating. Plates of this type are called positive plates. Positive plates are usually
used for art prints (e.g., exhibition catalogues). For a four-color print (see Section Basics), four plates are produced by
exposing positive plates through the positive color separation films.

Whereas the photocuring layer of negative plates capable of sufficiently long runs requires a definite minimum amount of light
to cure the image areas, the number of prints made with a positive plate is independent of exposure time. The light merely
serves to solubilise the nonprinting areas of the coating. For this reason, image reproduction can be controlled by varying the
exposure time of positive plates without the risk of insufficient run length. This is the starting point for the standardizing
system introduced to the European market, which provides the prerequisites for a highly uniform quality of reproduction. A
further advantage of the positive copy is that the original, especially for pictorial depictions, can be checked more easily for
mistakes, and the color separations (i.e., the film originals for color printing plates) can be mounted more easily in register on
film than with negative originals.

Positive plates use grained, anodized (1 – 4 g of oxide per square meter) aluminum as a substrate. The light-sensitive
formulations contain naphthoquinone diazides, binders, and additives such as dyes, indicators, and pigments. These
additives bring about good handling properties, i.e., they improve color contrast after exposure, colour contrast after
development, and developer resistance.

After exposure, positive plates are developed with aqueous alkaline solutions (pH 12.5 – 13.5) of alkali phosphates or alkali
silicates. Plates made this way can be subjected to heat treatment (6 min at 230 °C) to make them capable of longer press
runs. The diazonaphthalinone derivatives decompose at this temperature. They subsequently undergo a complicated cross-
linking reaction which improves the mechanical properties of the layer to such an extent that these baked plates are capable
of press runs two to three times longer than those obtained with untreated plates [306]. In practice, the plate is covered with a
protective layer of heat-resistant, hydrophilic, film-forming material before being baked to make sure that nonimage areas are
not soiled with oleophilic substances [307].

Diazonaphthalinone on exposure forms derivatives of indenecarboxylic acid, which can easily undergo decarboxylation under

page 51 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
certain conditions. This reaction helps to utilize these plates not only in photosolubilisation of quinone diazide layers, but also
in exposure of the same plate through a negative film, followed by the decarboxylation of the indenecarboxylic acid in the
exposed areas (2 min at 100 – 140 °C) [308] to give image areas insoluble in the developer. A subsequent overall or
imagewise exposure solubilises areas of the plate that were previously unexposed. In this way, both positive and negative
originals can be economically used to expose the same plate, thereby saving the expense of additional film.

Producers of presensitised positive plates are Agfa-Gevaert, Fuji Photo Film, KPG, Lastra, and others.

The waterless plate represents another modern variation of platemaking with positive transparent originals for special
application. Whereas in conventional offset printing, the nonimage areas are wetted with water (the image areas are
hydrophobic) to prevent the ink from wetting these areas, the nonprinting areas of a waterless plate consist of a silicone
rubber, which repels ink even without water [309]. These plates are also available in a negative version [310]. They are
produced by Toray.

Projection Platemaking using Negative Film or Microfilm Originals. Processing microfilm was very popular during the
1980s as it involved the saving of film costs. This method was used advantageously in printing books. Today this is very
much a niche market mainly for poster printing. Annual sales are < 1,000,000 m2. The equipment used includes a source of
UV light and an optical system for enlargement and projection of the microfilm onto the plate.

Highly light-sensitive special products must be employed as presensitised printing plates. The coating of these plates is
based on various photocrosslinking mechanisms: (1) photolysis of aromatic diazonium salts, (2) photodimerisation of
cinnamic acid derivatives, (3) photoinitiated polymerization, and (4) silver halide photolysis. The light sensitivity of the
materials increases in this order. Although the energy requirement of type 1 can be reduced to approximately 10 mJ/cm2 at
best, values of 0.1 – 1 mJ/cm2 have been reached with types 2 and 3. In comparison, silver halide materials require still less
energy (<1 µJ/cm2) for imaging. Of the individual highly light-sensitive plates, types 2 and, especially, 3 are capable of the
longest press runs. Producers of highly light-sensitive negative plates are Agfa-Gevaert (types 1 and 3), Fuji Photo Film Co.
(type 4), and KPG (types 1 and 2).

Positive Paper Originals. Positive paper originals can be used as input in two ways: parallel exposure in a camera or
sequential exposure in a scanner [311].

Camera Imaging. Camera imaging is increasingly important in small-size offset printing for short runs and in newspaper
printing with different material–equipment combinations.

1. In the Opti-Copy camera or in the Sixt RHI, the image of a positive paper original is projected, by means of an optical
system, onto a silver halide-sensitized plate which uses the silver salt diffusion principle for imaging. In this case, a
transparent polyester film is employed as substrate. Because these cameras do not contain deflecting mirrors,
exposure is made through the back of the plate to make sure that a right reading printing image (required in offset
printing) is obtained. Prints in quantities up to 50 000 can be produced. Producers are 3M, Agfa-Gevaert, and
Mitsubishi Paper Mills. A xenon discharge tube (with filter) is used as the light source in both cases.
2. Paper printing plates coated with zinc oxide (inorganic photoconductor) have found wide application for shorter runs in
small-size offset printing. This is an economical method for printing several hundred A4 or A3 copies. Special cameras
that take a roll of zinc oxide-coated material and are equipped with high-voltage corona, table for the original,
exposure platen, illuminating device, optics, and development device are produced by different firms (Eskofot, Itek,
Iwatsu Electric Co., Ricoh, and Fuji Photo Film).
The photoconducting zinc oxide pigment is embedded in a resin and sensitised with sensitizer dyes. An electrically
conductive paper serves as the substrate. Processing is carried out as described in Section Photophysical Principles
of Image Generation. A suitable salt solution (e.g., K, [Fe (CN),] or, more recently, other compounds such as phytic
acid) is finally used to convert the zinc oxide surface that is not covered with toner to the corresponding zinc salts,
which function as hydrophilic nonimage areas in offset printing. The energy requirement for imaging a zinc oxide plate
is 1–10 µJ/cm2.
Producers are AM international, Oji Paper, Ricoh, Scott Graphics, and Tomoegawa.
3. Another variation of electrophotography is used extensively by newspaper printers for platemaking [312]. In this
process, organic photoconductors are employed as the light-sensitive substance. They are embedded together with
sensitiser dyes in an alkali-soluble binder matrix. Imaging occurs as described in Section Photophysical Principles of
Image Generation. Finally, the areas of lithographic carrier not covered by toner are uncovered with the help of an
alkaline decoating solution. Only the toned areas still contain organic ink receptive substances. This plate can then be
used in a web offset press in the same way as a negative or positive plate. An automatic camera of the type employed
for processing such electrophotographic plates is shown in Figure 60.
The advantage of this method of platemaking, in addition to lower material cost, is a reduction of labor and time
required for the last (and latest) pages of the newspaper.

Figure 60. Automatic camera for processing electrophotographic plates

Producers of presensitised electrophotographic plates of this type are Agfa-Gevaert and KPG.
A variation of electrophotographic platemaking is a process that utilizes an effect called persistent conductivity. The
material is exposed without first electrically charging the surface. The electrons released on exposure to light lead to a
certain conductivity of the material which persists even after exposure [313].

page 52 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Scanner Imaging. The sequential imaging of radiation-sensitive plates may be utilized for the facsimile transmission of
imaging signals and the transfer of data from computer by direct modulation of the light source.

A variety of light sources are employed for sequential (scanner) imaging: water-cooled argon-ion lasers in UV mode (351 and
364 nm), helium – cadmium lasers (442 nm), water-cooled argon-ion lasers in visible mode (488 – 514 nm), air-cooled argon-
ion lasers (488 – 514 nm), helium-neon lasers (633 nm), LED (light-emitting diodes), laser diodes, normal incandescent
lamps, and YAG lasers (yttrium-aluminum-garnet) (1060 nm).

A correspondingly large number of radiation-sensitive materials are used for the production of lithographic plates. The light
source must be adjusted to the recording material in such a way that the following requirements are fulfilled:

1. The radiation employed must correspond to an absorption and sensitivity band of the recording material.
2. The energy density of the radiation must be high enough to induce a change of the recording layer within a time span
that is justifiable for practical use and which is sufficient for printing purposes. In other words, a weaker light source
would require a more sensitive recording material.

The closer the arrangement of the exposure lines, the better is the quality of reproduction achieved in sequential, line-by-line
exposure. However, the sequential exposure of a presensitised plate becomes more time-consuming for larger size plates. A
simple estimation shows that with an area of 600 µm2 per image element (which corresponds to a screen dot of 2 % area
coverage in the 60 metric screen) and with a plate size of just 400 × 600 mm2, only 0.15 µs is available for the exposure of a
single element if exposure of the entire plate is to be accomplished in 1 min.

Lasers beams that satisfy the first condition for a given recording material, after passing through the optical system and
reaching the plate with a light energy of 0.4 W still require a radiation-sensitive coating, capable of being imaged with a light
energy of not more than 10 mJ/cm2 for imaging. Highly light-sensitive coatings must be used [311]. Coatings based on both
diazonium salts or on photodimerising polymers can be formulated to give a light sensitivity of ca. 10 mJ/cm2. In particular,
further development of photoinitiated polymerisation has led to the production of highly efficient plates that require only 50–
150 µJ/cm2 for imaging. The use of such plates and systems is particularly popular in newspapers where fast make-ready
and durability of run are important factors. Producers: Agfa-Gevaert and Mitsubishi Chemical Corporation.

The flatbed scanner shown in Figure 61 illustrates the principle of image production.

Figure 61. Flatbed scanner for reading (——) and writing (----) an original. Fiber optics of the reading beam, photomultiplier,
and signal input to modulator are not shown

a) Argon-ion laser; b) Pyramidal mirror (up to 8000 rpm); c) Modulator; d) Helium – neon laser; e) Original; f) Plate

Although conventional UV-sensitive plates require several times this density of energy and are not suitable for imaging in the
manner described above, a system has been recently developed by BasysPrint that enables conventional UV-sensitive
plates to be integrated into the digital workflow. The use of such a UV platesetter offers significant cost advantages over
other fully integrated digital systems.

Electrophotographic materials can be sensitized for different wavelengths of the visible spectrum, including the region around
500 nm. The energy requirement for imaging is ca. 40 J/cm2. Hence, although principally developed for use with projection
cameras, plates of this type can be imaged with small, air-cooled argon-ion lasers that have a power of only a few milliwatts
of light energy (on the plate). Flatbed scanners were used during the 1980s for the direct recording of facsimile transferred
data on electrophotographic plates [311] (producers: Agfa-Gevaert, KPG).

The above describes early attempts at direct lithographic plate making. From the early 1990s computer memory and laser
optical design were able to store and output digital image information with increasing efficiency and cost effectiveness.
Computer-to-Plate (CtP) products then became the objective of lithographic plate manufacturers. Early entrants in this race
were electrophotographic plates described above but adapted for laser exposure [311] (producers Agfa-Gevaert, KPG).
However because of image quality problems and processing complexity they were soon replaced with plates using the high
light-sensitive silver diffusion transfer process (see Section Modern Lithographic Printing Applications) or plates based on
photoinitiated polymerisation.

The first platesetters all used visible laser; air cooled Argon-ion (488 nm), Helium-Neon (633 nm), semi-conductor diode
(670 nm), and frequency-doubled YAG (532 nm) were important development milestones. In the mid-1990s infra-red laser
diodes (830 nm) started a trend for thermally imaged printing plates. Development of thermal platesetters by CREO Inc. and
the launch of thermally sensitive printing plates by Kodak (later Kodak Polychrome Graphics) started a counter-revolution to
the marketing of visible light CtP products. Based on Novolac resins (phenol or cresol formaldehyde polymers) traditionally
used in positive analogue plates, the thermal plates function by having a cationic dye (typically tri-phenyl methane structures)
that insolubilises the Novolac resin but which, on thermal (laser) irradiation, causes a structural disruption so making the
resin soluble in alkali developer. Thermal CtP plates are characterised by having sharp reproduction and can be handled in
normal room lighting conditions.

Further important developments can be expected in the field of thermally exposed CtP whilst visible-light sensitive CtP are
moving into a mature phase of development. Thermal exposure also opens up the possibility of designing plates that only
require exposure, that is they can be produced without chemical processing (producers: Agfa-Gevaert, KPG, Fuji Photo Film,
Lastra).

page 53 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Imaging of Plates from Digital Data. Steadily increasing data storage capacities permit the fast processing of ever growing
quantities of information and text. Not only the composition of text columns, but also page makeup of type and electronically
screened four-colour illustrations, can be carried out by computer and checked on the screen. The completely integrated
digital system from text preparation to platemaking is described in Section Graphic Arts Photography).

Because of the digital structure of the type and image information in the computer, printing plates can be exposed directly in
output scanners without any intermediate imagewise recording of information on a carrier material. These systems are
largely independent of an original copy and, consequently, of silver film as well. The steps involved from the information to
the plate are presented schematically in Figure 62.

Figure 62. Comparison of different pathways from original to printing plate

3.4.2. Letterpress and Flexography

Letterpress is the oldest method of printing. In it, printing areas of the printing plate are raised above nonprinting areas. Ink
is applied to the raised areas with the help of ink rollers and then transferred to the printing material. In the past, letterpress
printing plates were made of metal (lead, zinc, or magnesium). The appearance of photo setting (i.e., setting information
directly on film) necessitated the development of a direct path from film to plate. This led to the emergence of photopolymer
relief plates in the 1960s; today, letterpress almost exclusively uses plates of light-sensitive plastics. The polymeric binders
used mainly in letterpress printing plates are polyamide or poly(vinyl alcohol).

Flexography is a special form of relief printing (called aniline or rubber printing in the past). In flexography the printing plate
is a soft and flexible material. Before 1970 rubber printing forms were used. Production of a rubber printing form involved the
following steps. First a hard letterpress printing form is produced, which is used to mold a duroplastic flong. Then the rubbery
raw material is filled into the flong and vulcanized to yield the rubber printing form. Today most flexographic applications use
light-sensitive photopolymer printing plates, which contain flexible thermoplastic elastomers (e.g., styrene – isoprene –
styrene block copolymers) as polymeric binders. These flexographic printing plates can be processed more easily than
rubber and show better print results.

Letterpress printing (platen and rotary printing procedures) ranges from job printing (e.g., visiting cards, programs, and
business papers) to printing of adhesive labels, forms, tubes, cups, newspapers, and paperback books. The market share of
letterpress newspaper printing is decreasing due to the dominance of lithography, whereas the market share or letterpress
label printing is increasing. Flexography is especially suited for printing paper, cellophane, aluminum, and plastic films, and is
also used to print uneven materials (e.g., corrugated board). Nearly all products printed by flexography are packaging
materials.

3.4.2.1. Structure of Photopolymeric Letterpress and Flexographic Plates

A solid photopolymer printing plate consists of a light-sensitive photopolymer layer, which forms the relief layer after
processing (see Fig. 63). Depending on the application the thickness of this layer varies from 700 to 7000 µm (flexographic
printing plates) or 30 to 2000 µm (letterpress printing plates). The very sensitive photopolymer layer is covered by a
protective film to prevent soiling. The plates are provided with an additional thin layer (release layer) between the protective
film and the photopolymer layer. This release layer is usually made of poly(vinyl alcohol), polyamide, or cellulose derivatives
[315], [316]. The function of this layer is to ensure a tack-free surface and to provide a good contact to the film negative
during exposure. The adhesive layer mechanically anchors the photopolymer layer to the base material. Steel, aluminum, or
polyester are used as base materials. High demands have to be fulfilled by the adhesive layer, because the relief is
subjected to considerable mechanical and chemical stress during printing. Solvent-resistant two-component lacquers based
on the cross-linking reaction between multifunctional polyols and polyisocyanates are often employed for the adhesive layer
[317].

Figure 63. Structure of photopolymer printing plates

3.4.2.2. Production and Requirements of Photopolymer Letterpress and Flexographic Plates

Processing of letterpress and flexographic printing plates involves the following steps:

Exposure;
Washout;
Drying;
Posttreatment (if necessary).

Exposure (see Fig. 64). First the protective film is peeled off, then a film negative is placed on the plate. During exposure

page 54 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
with UV-A light (wavelength range 350 – 400 nm, UV tubes or mercury lamps) the polymer layer polymerizes at the areas
corresponding to the transparent areas of the film negative. The exposed areas of the photopolymer layer cross-link and
become insoluble. The nonexposed areas of the photopolymer layer remain soluble. The exposure is performed under
vacuum to ensure a good contact between plate and film negative.

Figure 64. Exposure

Washout (see Fig. 65). After exposure, the soluble parts of the plate are washed out in special devices (spraying or rubbing
system). Letterpress plates are washed with alcohol – water mixtures or with pure water. Most flexographic printing plates
have to be washed with organic solvents (see Table 2). Before 1990 a mixture of tetrachloroethylene with butanol was used
as washout solution. Today less harmful solvents, such as paraffinic and naphthenic hydrocarbons, aromatic hydrocarbons,
or alkylesters have been substituted for tetrachloroethylene in most countries [318]. Strong efforts are made to develop
flexographic printing plates, which can be washed out with water or aqueous solution (see Section Gravure Printing).

Table 2. Trade names and manufacturers of relief printing plates

Trade name Company Washout

Flexographic AFP Ashai organic solvent

printing plates nyloflex BASF organic solvent
Cyrel DuPont organic solvent
Flex-Light Polyfibron organic solvent
Flexceed Supratech water
Okha-Flex Tokyo Okha organic solvent
Cosmolight Toyobo aqueous solution

Letterpress nyloprint BASF water or water – alcohol

printing plates Napp plate Napp Systems water
Miraclon Tokyo Okha water or water – alcohol
Rigilon Tokyo Okha water
Torelief Toray water
Printight Toyobo water

Figure 65. Washout

Drying and Posttreatment (see Fig. 66). Once washed, the plates are dried and subsequently reexposed for complete
polymerization. In this way, they attain their final mechanical properties for the printing process. Flexographic plates are
sticky after production and must be subjected to an additional posttreatment process. By exposure of the plates with UV-C
light (wavelength 250 nm) the surface of the relief is hardened and detackified [319].

Figure 66. Drying

Both letterpress and flexography have a wide spectrum of applications. The demands made on the plate in different
applications are so divers that they cannot be met by a single plate. These requirements depend on the ink, the material to
be printed, and the type of press. In flexography printing inks of low viscosity are used. Jobs on paper, napkin, or corrugated

page 55 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
board are performed with water-based inks. Plastics and metal foils are printed with alcohol-based inks or UV inks. The
uneveness of the substrate determines the thickness of the flexographic printing plates. Corrugated board is printed with
plates with a relief height of about 3.0 mm. Plastics is printed with thin plates with a relief height of about 0.7 mm. The trend
is to use plates as thin as possible to reduce thickness tolerances, to increase print quality, and to reduce cost [320]. In
letterpress applications UV inks or oil-based inks are used. Due to the higher viscosity of the printing inks the relief heights of
letterpress plates are lower (0.3 – 0.7 mm).

3.4.2.3. Chemistry of Photopolymer Relief Printing Plates

The principle of all photopolymer relief printing plates is based on the UV-induced photopolymerization of olefinic unsaturated
compounds [321], [322]. Other processes used for reprographic products (e.g., photolysis of o-quinone diazides or
dimerization of cinnamic acid; see Section Lithography (Planographic Printing, Offset)) are of little importance for letterpress
because of the low utilizable photochemical quantum efficiency. Relief printing plates are up to 1000-fold thicker than
lithographic printing plates. Hence, in keeping with the latest findings, only a chain reaction such as radical polymerization
can reach the required low exposure times of 10 – 1000 s.

The development of commercially available photopolymer relief printing plates, based on polymeric binder, vinyl monomer,
and photoinitiator was accelerated considerably by the work of PLAMBECK et al. in the mid 1950s [323]. This was based on
the work of GATES on photopolymerization of methyl methacrylate through a film negative, which paved the way for
photomechanically produced relief printing plates [324].

In the absence of a sensitizer, olefinic compounds absorb light only in the shortwave UV range below 300 nm. However, this
radiation cannot be used for cross-linking because light sources with sufficient energy output at this wavelength are not
available and common film negative materials are not transparent enough below 300 nm. To initiate photopolymerization in
relief printing plates with long-wave UV-A light (wavelength range 350 – 400 nm), the photopolymer formulation must contain
substances that are capable to absorb UV-A light and generate polymerization-initiating radicals. Usually compounds are
added to the photopolymer formulation that either undergo a homolytic bond cleavage to yield a polymerization-initiating
radical pair or form suitable initiating radicals by abstraction of a hydrogen radical from another compound [325].

Examples of hydrogen abstraction initiators are derivatives of anthraquinone, thioxanthone, benzophenone, and
acetophenone. Some alpha splitters are derivatives of benzoin, benzil, benzoylphosphines, and benzoylphosphine oxides
[326].

Examples of cross-linkable vinyl compounds (polymeric, oligomeric, or monomeric) are:

1. mono- or polyfunctional acrylates and methacrylates [327], [328];

2. mono- or polyfunctional acrylamide or methacrylamide derivatives [329];
3. allyl ethers (e.g., polyurethane oligomers with allylic functionality that cross-link easily with polyfunctional mercaptans
[330]);
4. mono- or polyfunctional esters of maleic or fumaric acid [331].

Apart from the initiator and the polymerizable vinyl compound, many relief printing plate formulations contain a polymeric
binder, which controls the physical chemical properties, e.g., elasticity, hardness, resistance to solvents. The polymeric
binder is, as a rule, the main component of the photopolymer formulation and is the matrix that takes up the generally liquid
monomers and makes them manageable. Often the polymer takes part in the UV-induced cross-linking reaction and,
together with the monomers, forms a three-dimensional cross-linked network. Some binders in commercially available relief
printing plates are:

1. soluble polyamides [329];

2. partly saponified poly(vinyl acetate) [e.g., poly(vinyl alcohol)] [328];
3. polyacrylates, polyesters, or polyurethanes;
4. thermoplastic elastomers, such as styrene – isoprene – styrene rubbers or styrene – butadiene – styrene block
copolymers, ethylene – propylene – diene rubbers, or nitrile rubbers [327], [332].

As a rule, polar polymers, such as polyamide or poly(vinyl alcohol), are used in letterpress plates. Because of the high
polarity of the polymers the plates can be washed out in polar solvents such as water or alcohol and can be printed with
nonpolar printing inks such as UV or oil-based inks.

In contrast, nonpolar polymers such as thermoplastic elastomers are used in flexographic printing plates. Due to the nonpolar
nature of these polymers, flexographic printing plates have to be washed out in nonpolar organic solvents such as
hydrocarbons or chlorinated hydrocarbons and can be printed with polar printing inks based on water or alcohol.

Because of ecological reasons strong efforts are made to develop flexographic printing plates which, in the unexposed state,
can be washed with water and, in the exposed state, do not swell in water or alcohol-based printing inks. Today nearly all
manufacturers of flexographic printing plates have developmental plates which can be washed in water or aqueous solution.
But only some of these are commercially available (see Table 2). The photopolymer layer of these plates often contains a
mixture of polymers of different polarity [333]. Another possibility is to incorporate monomers with different polarity into one
polymer [334-336].

Further auxiliary agents that may be added to photopolymer formulations are:

1. Polymerization inhibitors;
2. Plasticizers;
3. Dyes and pigments which improve the relief form and the resolution of the printing relief [337].

page 56 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
3.4.2.4. Future Developments
Photopolymeric letterpress and flexographic plates have connected relief printing to modern computer technology
(phototypesetting and reproduction technology). The print quality of letterpress and flexography improved substantially by the
use of photopolymer plates and reprographic techniques which adjust the film negative to an optimum print result.

The next generation of photopolymer plates will not even need a film. Computer-to-plate (CTP) techniques which are state of
the art in lithography will be introduced in letterpress and flexography. To overcome the difficulties arising from the plate
thickness of relief printing plates, multilayer plates are being developed. Two filmless exposable flexographic printing plates
were launched by BASF and DuPont in 1996. Both plates have a thin black layer of high optical density on top of the
photopolymer layer. This black layer is imagewise laser-ablated by exposure with a yttrium – aluminum – garnet (YAG) laser
(1064 nm). By this process the film negative is created in situ on the plate. Then the plate is processed like a conventional
flexographic printing plate by UV exposure, washout, drying, and posttreatment [338], [339]. The processing of conventional
flexographic printing plates is compared with the CTP-process in Figure 67.

Figure 67. Conventional processing of photopolymer plates and CTP process

Further developments tend strongly toward improved environmental protection and occupational health. Skin-sensitizing
monomers will be replaced in photopolymer plates. Further progress is expected with respect to water-washable flexographic
printing plates. In flexographic printing water-based inks and UV inks will be substituted for alcohol-based inks. Photopolymer
plates with a better resistance against UV inks will be developed. The print quality of letterpress and flexography will increase
further.

3.4.3. Gravure Printing

Rotogravure is the printing process for highest printing quality. Printing can be performed on nearly any kind of printing
material (also on inexpensive paper). Two types of rotogravure exist: rotary printing and sheet-fed printing.

Rotary printing is the most important process for mass production of publications, mainly magazines and catalogues. The
printing forms for gravure printing are cylinders. These cylinders have a steel jacket which, in turn, has either a copper or a
plastic coating depending on the process used to make the gravure cells. In the majority of practical applications, a base
copper coating is electrolytically applied to the steel jacket. A plastic coating is applied to the steel jacket only when lasers
are used in gravure platemaking (see Section Electron-Beam and Laser Gravure). The layer thickness of the base copper is
approximately 0.5 – 3 mm, depending on whether the gravure cells are placed directly in the copper base or in a detachable
copper skin (=Ballard Shell) on the base copper (etched or engraved). If a Ballard Shell is used, the thickness of the base
copper is only 0.5 – 1 mm. After application of the base copper, the cylinder must be turned on a lathe very precisely to the
desired diameter. Then the cylinder surface must be polished.

A copper-coated cylinder of this type serves as the carrier for the Ballard Shell used for printing. The thickness of a Ballard
Shell ranges from 0.1 to 0.15 mm. This shell should not combine with the copper base but should remain easily detachable.
For this reason, a separating layer must be applied between the base copper and the shell. A solution of silver nitrate (or
silver cyanide) and calcium cyanide is generally used for this purpose. The advantage of this process is that the cylinder
always has a uniform diameter due to the application of the copper.

The next step involves placing the printing elements — the gravure cells — in the gravure cylinder (raw cylinder) prepared as
described above. Until the early 1960s, only chemical methods were used for this purpose. These methods have been
improved and refined continuously, especially by using electronic controlling and standardizing devices. This type of gravure
platemaking is known as the conventional process to distinguish it from other processes.

With increased industrialization of many process steps in the printing industry, especially in gravure printing,
electromechanical gravure is becoming better established in gravure platemaking. This process represents application of the
technology developed by Dr.-Ing. Rudolf Hell (Kiel) for letterpress platemaking. The appliances required for this purpose
were called Klischographs. In letterpress printing, they were used to remove nonprinting areas from the plate material. In
gravure platemaking, exactly the opposite occurs: the printing elements or gravure cells are engraved into the plate material.
The appliances required for gravure platemaking are called Helio-Klischographs, analogous to the trade name Klischograph,
and are used worldwide for publication rotogravure (magazines and catalogues).

3.4.3.1. Conventional Method of Gravure Platemaking

Because of the present status of gravure platemaking and the numerous variations that exist, only the fundamentals of the
conventional process are discussed here. The problems connected with this method are easily discernible and have
determined the trend toward methods that are more easily controllable and reproducible. Specific details of the conventional
method are described in the literature [340-342].

For the production of printing cylinders, the information to be printed (text or pictures) must be available as complete pages,
in the form of a page makeup or stored data. In the conventional process, the text is present as a phototypesetting film
(positive, black images on a light background), and the picture material is available as a continuous-tone diapositive. For
multicolor gravure, positive color separation diapositives are prepared with the printing colors yellow, magenta, cyan, and
black, corresponding to the page layout. These originals cannot be transferred directly to the raw cylinder. An intermediate
carrier, carbon tissue, is required. This paper is available in rolls and is coated with gelatin, which gives an etchable washout

page 57 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
relief. Shortly before use, the gelatin coating is made sensitive to light with a chromium salt solution. Use of carbon tissue
requires considerable care. Workrooms must be air-conditioned and dust free. The sheets of paper can be cut to the required
size only after they have become acclimatized. Prepared sheets are usually dried in an oven and then stored in a deep
freeze until used.

In gravure printing the range of tone values is realized by varing the volume of the printing elements (cells) on the printing
form (cylinder). As in other traditional printing processes (e.g., letterpress printing, lithography) the printing ink must be
transferred to, e.g., paper in the form of very small portions (80 – 200 lines per inch). These portions are a result of the so-
called screening process. Each original (image), may it be text or picture, must be screened. This process takes place in the
pre-press department for making the films that are to be copied on, e.g., pigment paper (carbon tissue) [342], [343]. This
results in the formation of area elements of uniform size. Finally, these areas must be etched to a depth corresponding in
size to the required tone value. These are the gravure cells, and the entire net is called the gravure screen [342]. It serves
two purposes simultaneously. On the one hand, the quantity of ink in each individual cell is defined by the volume of that cell.

On the other hand, the doctor blade, which wipes the ink off the surface of the cylinder, before printing, is provided with a
surface of uniform height to ride on.

To produce the cells and partitions, a screen is copied on the carbon tissue. The variable-depth process normally uses a
cross screen with a 45° rotation with respect to the side lines. It consists of a transparent net and squares (Fig. 68). The ratio
of partition width to length of a square (i.e., screen – partition ratio) is usually 1 : 3. The screen ruling (i.e., the number of lines
per centimeter) is normally 70 for printing publications. Larger screens are usually used to printing packaging and textiles;
finer screens are used for stamps.

Figure 68. Gravure screen for production of cells and partitions

Carbon tissue is sensitive to light ca. 300 – 450 nm in wavelength. However, the maximum sensitivity is at 360 – 380 nm.
Consequently, only light sources emitting predominantly shortwave radiation are suitable for gravure. To obtain an especially
sharp halftone copy of the screen partitions, a point source of light projected through a lens is employed. The areas on the
carbon tissue that are exposed to light (the areas under the transparent net) lose their water solubility. The chromium –
gelatin layer is thus hardened to varying degrees depending on the intensity of light.

After the screen has been copied on the carbon paper, the picture (i.e., the fully stripped page) is copied in a second step.
Depending on the tone value (degree of density) of the different image areas, light penetrates more or less into the
chromium – gelatin layer. Thus, the curing of the gelatin layer is directly proportional to the density of the original to be
copied. The gelatin remains water-soluble in those areas where no light passes through. The screen copied in the previous
step is not affected by picture copying. It is maintained in lines on all the printing areas, both in the text and in the smooth
dark areas as well.

The finished sheets of carbon tissue are then aligned on the copper cylinder with a special transfer machine. The highly
polished copper cylinder must be degreased carefully. The copied sheets of carbon tissue are briefly wetted and then
pressed onto the cylinder. This must be a highly precise step because carbon tissue is not dimensionally stable.

Despite standardization and very careful operating methods, inaccuracies in positioning cannot always be avoided. As a
result, other materials have been developed as substitutes for carbon tissue. These materials use a stabilizing foil or film
instead of a paper base but undergo similar processing steps.

The copies transferred onto the cylinder are developed with warm water (40 °C) in a developing machine. After a few
minutes, the paper backing is wet enough to be removed carefully from the rotating cylinder. Development continues (ca.
10 – 15 min) until the soluble gelatin is completely washed out, leaving the exposed (cured) gelatin relief clearly visible. The
completely blank areas give the deepest wells after etching, and a layer of hardened gelatin of varying thickness remains on
the other wells. The cylinder is then cooled to room temperature and dried with alcohol. These steps are carried out in fully
automatic and programmable developing machines.

The etching process is the most difficult step in gravure platemaking. All areas that are not to be etched are covered with an
acid-resistant asphalt lacquer before etching. The etching process is now largely automated in specially constructed etching
machines. Solutions of iron(III) chlorid, commonly called perchloride of iron, of varying densities are used for etching. The
heavier solutions are able to soften and penetrate the thin gelatin layers (equivalent to the shaded areas in the film positive).
Lighter solutions are used progressively as it becomes necessary to penetrate the thicker gelatin layers (equivalent to the
middle tone to highlight range) [340-342]. Etching is done in several stages, with four to six solutions at different Baumé
density levels (Bé). The most concentrated solution is applied first, and less dense solutions are used stepwise. A variation of
the etching process uses one etching bath only with an etching solution of a single Baumé level. The numerous problems
associated with manual and mechanical etching are described [340], [341].

3.4.3.2. Electromechanical Gravure Platemaking

Modern electronics, together with precision mechanics, guarantees the highest standards of operational safety and
reproducibility. Gravure cylinders with accurately programmed printing characteristics can be produced with the Helio-
klischographs, which were introduced to the public in 1962. This process is generally called “cylinder gravure.”

An important advantage of cylinder gravure is that it involves only one operation which starts with the original (stripped page)

page 58 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
and leads directly to the ready-for-press cylinder. Furthermore, it eliminates all the difficult-to-control problems that can occur
during chemical processes, during the use of intermediate carriers (e.g., carbon paper), and during conventional etching. In
addition, all problems associated with the halftone copy are eliminated because the halftone process is carried out
electronically.

Cylinder gravure can be considered a procedure for the transfer of information. The originals to be engraved are mounted in
a ribbon-by-ribbon manner as black and white originals, right side up, on the image cylinder. Each line of the original is
scanned with a scanning system. This is the same process used for the transfer of press photos or for electronic color
correction in color scanners. The optical signals scanned from the picture cylinder are converted to electrical signals, which
are fed into a computer. The desired gravure gradation is adjusted in this data-processing unit, which can be switched to
scan either positive or negative originals. Electrical signals obtained this way are then conveyed to the engraving heads,
which engrave the printing cylinder. Based on the principle of transmission of information, the unit with the image cylinder can
be located at any distance from the engraving unit (long-range transmission). Electrical signals from an image cylinder can
also be used to control several engraving units at the same time. In general, however, the scanning unit and the engraving
unit have a common machine bed (Fig. 69).

Figure 69. Helio-Klischograph type K 201 (Dr.-Ing. R. Hell, Kiel). The scanning cylinder is in the middle of the figure; the
gravure cylinder to be engraved on the right.

Electrical signals transmitted to the engraving unit operate an engraving system for each string. These engraving systems
are moved with great precision synchronously with the scanning systems. Each engraving system has a sharply cut diamond
stylus that moves up and down ca. 4000 times each second at right angles to the surface of the cylinder. Every time the
stylus penetrates the cylinder, a pyramid-shaped cell is engraved into the surface. The depth of penetration of the stylus and,
thus, the volume of the cell correspond to a particular tone value on the scanning side. In the surface of the cylinder, the cells
have a rhombic shape and are arranged in the gravure tracks so that the cells of neighboring tracks are staggered [340],
[341].

An engraving efficiency of 0.27 m2/h is attained with a cylinder of 1000-mm circumference and a screen of 70 lines per
centimeter. It is customary to use as many scanning or engraving heads as ribbons are provided for printing. For this reason,
the engraving of a cylinder requires as much time as the engraving of a string (i.e., one hour) independent of the number of
strings.

Because cylinder gravure basically processes electrical signals, this method can be used for purposes that are difficult or
impossible to achieve with conventional gravure (e.g., repetition in the circumferential direction, repetition in the direction of
the cylinder axis, and postgravure [340], [344]).

3.4.3.3. Electron-Beam and Laser Gravure

Since the beginning of the 1960s, repeated attempts have been made to develop a method that uses an electron beam for
cylinder gravure. This idea was initiated by the European Rotogravure Association (ERA). Development of an electron-beam
gravure (EBG) machine was undertaken by the company Dr.-Ing. Rudolf Hell (Kiel) [345] and was continued after the
company changed its name to Linotype-Hell following a takeover by Linotype. However, the project was discontinued in
March 1993 after detailed cost analyses showed that the likely running costs of an EBG machine would have made it
uneconomic to operate.

In contrast to EBG (electron-beam gravure), the application of lasers does not require the creation of a vacuum. Research on
the use of a laser engraving machine was begun in the mid 1970s by Crosfield Electronics (London). This work finally
resulted in 1982 – 1983 in an entire system for the production of gravure plates. The Laser Gravure System 700 was
presented to the public in 1983 in a form that had been tested at the same time by Odham-Sun Printers Ltd. (Watford,
England) under practical gravure printing conditions (Fig. 70) [346]. However, this development also ultimately proved not to
be economically viable and was itself abandoned.

Figure 70. Engraving machine of the system Lasergravure 700 (Crosfield Electronics, United Kingdom)

The laser gravure process differs considerably from all other gravure platemaking processes. Because laser beams are
strongly reflected from metallic surfaces. The problems which this causes were solved after many years of development work
by the company MDC Max Dätwyler AG in Bleienbach, Switzerland, whose decisive breakthrough came with the
development of a special alloy which could be applied to the gravure cylinder. The “Laserstar” laser gravure system, based
on this idea, gives an engraving frequency of 35 kHz with up to 240 lines per centimeter. Printing results so far obtained with
these cylinders confirm the precision and high state of technical development of the “Laserstar” system. The first practical
printing press of this type is expected to be operating in 1997.

Development work is aimed at increasing the engraving frequency to 70 – 140 kHz and at producing a second generation of
alloys for the process.
page 59 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Doubling of the engraving frequency incurs a cost increase of ca. 25 %.

3.4.4. Screen Printing [347], [348]

Screen printing was the successor of stencil printing, a method used in China and Japan in very early times. This printing
process was reportedly invented by MIYASAKI (1664 – 1736) in Japan. Not only is screen printing suited to flat materials such
as stickers, road signs, large advertisements, pictorial depictions on automobiles, and decals, but also items such as glass,
tile, T-shirts, plastic bottles, and printed circuits can be printed on by this process.

The stencil in screen printing consists of a screen mounted on a frame and the parts of the stencil that are fixed to the
screen. Printing occurs under the screen by forcing the ink through the screen openings with a squeegee onto the printing
material. Not only the frame, but also the screen material, fabric thickness, mesh size, yarn diameter, type of weave, and
structure of the stencil can vary from job to job. Thus, screens are not prefabricated or presensitized; rather, the individual
materials required for stencil making are available commercially and the final stencil must be made at the printer's from these
materials.

3.4.4.1. The Screen

Materials used for screens are textile fabrics such as silk, polyamide, and polyester fibers, but metals such as steel, nickel,
and phosphor bronze may also be employed. The screens are pretreated before coating to improve adhesion of the stencil
and printing quality. The stencil is removed from the screen after use so that the screen can be reused for other jobs.

3.4.4.2. The Stencil

Like other printing processes, the printing form in screen printing can be produced manually or photomechanically.

Manual Methods. For instance, a stencil can be made by sticking pieces of film or firm paper on the nonimage areas of the
screen. In another method, a suitably coated film or paper is used to produce a stencil. The image is engraved into the
coating, the coating is removed from the image areas, and the remaining layer is transferred to the screen with heat or
solvent.

Still another method employs a lacquer, which is applied to the nonimage areas on the screen, or a covering agent, which is
applied to the entire surface over a resistant image that is subsequently washed out with a suitable solvent to produce the
printing screen.

Photomechanical Platemaking. A light-sensitive coating is employed in both photomechanical methods: the direct and the
indirect method. In the first case, the coating is applied to the screen before exposure and development; in the other, it is
exposed on a carrier, developed, and then transferred to the screen.

In the direct method, the screen mounted on the frame is coated with a light-sensitive solution called an emulsion. The
carefully dried coating can then be exposed to UV light through a positive original in a copying frame. After exposure the
soluble areas are removed by a water rinse. Various formulations are being used as the emulsion:

1. Colloid – dichromate solutions (see Section Photoinsolubilizing (Photocuring) Systems) are highly sensitive to light,
can be developed easily with water, are very resistant to solvent-containing ink, and allow easy regeneration of the
screen. Substances such as gelatin are used as colloids. The emulsion is made at the print shop from its individual
components because the light-sensitive mixture is stable for only a limited time at room temperature.
2. Colloid – diazonium salt solutions (see Section Photoinsolubilizing (Photocuring) Systems) have been available for
some time as substitutes for chromium(VI)-containing emulsions. Poly(vinyl alcohol) or poly(vinyl alcohol) – poly(vinyl
acetate) is used as the colloid. The processing procedure is comparable to that of colloid – dichromate solutions.
Colloid – diazonium emulsions are sometimes said to have a lower sensitivity to light than the corresponding
dichromate solutions. The reason for this may be that the spectral sensitivity of dichromate coatings exceeds 500 nm,
whereas that of aromatic diazonium compounds is limited to the range 360 – 420 nm.
3. In more recent times, photopolymer formulations have also been offered for screen printing [349]. These fomulations
may be epoxy resins with unsaturated side groups or acetals of 4-(4′-ethenylpyridinium)benzaldehyde (4-stilbazolium
aldehyde) with poly(vinyl alcohol), which are susceptible to cross-linking dimerization. The water solubility of the
emulsion is guaranteed by the very polar structure of the pyridinium units. In some preparations, these polymers are
used in combination with acrylate monomers.

In the indirect method, a binder film, which may be either sensitive or insensitive to light, is used on a carrier as the starting
material for stencil making. In the print shop the insensitive film is wetted with a solution of a light-sensitive agent (sensitized)
and dried before exposure. From this point on, their processing is basically comparable: exposure, development, transfer of
the stencil to the screen, and removal of the film carrier. The advantages of the indirect method are accurate reproduction,
even with mesh-crossing contours, as well as constant layer thickness which, in turn, is advantageous for print quality. The
precoated materials used consist of poly(vinyl alcohol) [or poly(vinyl alcohol) with water-soluble diazonium salts] and are
dyed.

The combination stencil is a photo stencil which is produced commercially as a light-sensitive film on a carrier. This film is
first transferred to the fabric with water or emulsion. After the emulsion has been dried, the carrier is removed and the light-
sensitive stencil is exposed. Development of the exposed stencil is carried out with a jet of tap water, as in the direct method.

Some advantages of the combination stencil are accurate positioning of the image because of direct exposure and sharp
reproduction favored by the smooth surface of the screen. The light-sensitive formulations that can be used for these
coatings are colloids which are sensitized with diazonium salts or with styrylpyridinium salts.

Producers of light-sensitive products for making stencils are Hoechst, Kissel und Wolf, Ulano, Autotype, Sati, Sericol,

page 60 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
NazDar, Aicello, and Murakami.

[Top of Page]

4. Imaging for Electronics

Hartmut Steppan, Donald C. Mammato, Thomas Stoudt, Michael C. P. Watts

This chapter occupies a special position because “imaging” in the classical sense is discussed here only in exceptional
cases. Here, imaging refers to the selective coating of substrate surfaces to protect coated areas from the effects of etching
reagents, metallization baths, etc. Because of the controlling function of selective coating, this technique allows the overall,
undifferentiated action of working media (chemical tools) on the entire workpiece, which is a prerequisite for mass production.

Photoresists play a very important part in the production of selective coatings (Section Photoresists). Other important aspects
of imaging for electronics are the production of printed circuits (see Section Printed Circuits (Printed Circuit Boards)) and the
production of integrated circuits (see Section Microelectronic Devices).

4.1. Photoresists
The term resist is derived from resistant. It refers to a material that is used to coat areas of a substrate surface to protect
them from the reagent in question during chemical or physical treatment (e.g., acidic or alkaline etching, etching with plasma,
galvanic metallization, diffusion of foreign substances, to dope the substrate etc.).

Generally, two methods are used for the production of selective resist coatings:

1. selective application only to those areas of the substrate surface that are to be protected: e.g., application by screen
printing (see Section Screen Printing ) [350], magnetography [351], or other printing processes; and
2. covering the entire substrate surface with a layer of resist, which is subsequently removed from the areas to be
processed.

The latter includes the photoresist technique [352-354], which plays an important part in the production of certain printing
plates (multi-metal plates, zinc and magnesium stereotype plates), but is of paramount importance in electronics for the
production of printed circuit boards (PCBs) and integrated circuits (ICs).

Photoresists are defined as radiation-sensitive or, in a narrower sense, light-sensitive materials whose solubility is altered
when exposed to radiation. The selective resist coating is produced by exposing the photoresist layer through a mask. In this
way, only certain areas of the resist layer are exposed to radiation. Subsequent development with a suitable liquid, the
developer, results in removal of the more soluble coating areas, leaving these areas of the substrate uncovered.

In the field of integrated circuits, however, tendency to avoid wet processes is growing, not only in etching and stripping, but
also in developing [355-360].

Classification of Photoresists. A number of different photoresists are available on the market today. The characteristic
features are mode of operation (negative, positive), physical state (liquid, solid), differentiating radiation (e.g., short- and long-
wave UV, X-rays, or electron beams), development (with organic solvents, aqueous solutions, etc.), application (e.g, resists
for the production of solder masks), number of components (e.g., one-, two-, or multicomponent systems), and chemical
composition [352], [353], [361-364].

If the solubility of a photoresist increases during exposure to radiation, it is termed a positive resist. If the solubility decreases,
it is a negative resist (see Fig. 71). Whether a photoresist is positive or negative depends not only on its composition, but
also on the method of processing.

Figure 71. A) Application of a resist mask using a photoresist;

B) Development by dissolving the more soluble areas of the layer a) Positive photoresist; b) Negative photoresist;
c) Exposure mask; d) Substrate covered with photoresist; e) Substrate with selective resist coating

Liquid resists are available in two versions:

1. solutions of photoresist components in solvents in which the wet film dries on evaporation of the solvent, to give a
uniformly smooth layer; and
2. solvent-free liquid systems (without exception negative systems) with reactive diluents, e.g., liquid acrylates or
methacrylates, which polymerize or cross-link during radiation and thus become a component of the solid resist
coating [365].

Dry film resists are solid photoresists (thickness ca. 15 – 100 µm) coated and dried on temporary carriers (usually polyester)
[366]. In general, the layers are also covered with a protective foil (polyolefin). After removal of the protective foil, these
photoresist materials adhere to (largely flat) substrates when they are heated under pressure.

For many years, a spectral light sensitivity for photoresists in the range 350 – 450 nm was customary and adequate. Today,
photoresists sensitive to more energetic radiation (e.g., UV light of 249 or 313 nm, as well as electron beam, ion beam and X-
rays) [361-364], [367] are becoming increasingly important for microlithography because they allow higher resolution. Despite
the different nature of these rays, the term “photoresist” is retained. On the other hand, photoresists sensitive to lasers of

page 61 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
488 nm are now fairly important in the field of printed circuit boards.

In general, the photochemistry of photoresists sensitive to electromagnetic radiation of wavelengths >300 nm is quite
comprehensible. Selective reactions cause predictable changes in the solubility of the resist material. However, more
energetic radiation excites not only conjugated bond systems but also isolated double bonds, non-bonding electron pairs,
and finally, -bonds. Ionizing radiation and electron beams, ion beams or X-rays produce very reactive intermediates, which
in turn take part in successive nonselective reactions. Fragmentation and depolymerization often occur simultaneously with
grafting and cross-linking processes, and can be influenced by adjusting the radiation dose.

4.1.1. Industrial Applications of Photoresists

A series of rules, not entirely without exceptions, applies to the use of photoresists; these are described briefly below. In
general, thin photoresist layers ( Langmuir – Blodgett layers represent the extreme case [368]) resolve better than thick
layers; positive-acting layers are better than negative layers; and layers that can be developed in aqueous alkaline solutions
(less swelling) resolve better than those developable with organic solvents.

A list of important properties of industrial photoresists follows:

Solubility
Purity (particle-free)
Sensitivity to differentiating radiation
Resolving power and accuracy of the structure transfer
Processing properties (development, stripping, etc.)
Processing latitude
Adhesion to substrate
Resistance to etching
Resistance to electroplating
Thermal stability of the resist mask
Product constancy from batch to batch
Shelf life
Toxicological safety

Photoresists are applied in the production or implementation of certain products and processes. A list of these, along with
some photoresist characteristics and their importance in individual cases, appears in Table 3. No single photoresist can fulfill
all these requirements which, in fact, must be extended for special techniques. A large number of photoresists are necessary,
each one functioning optimally in a particular area only.

Table 3. Application and importance of some properties of photoresists

Property Product or process *

Printed Photochemical Electroforming Integrated Micromechanics

circuit machining [370] [371], [372] circuits [376]
boards [373][374]
[352], [375]
[353],
[369]

Resolving (+) + (+) + +

power
Resistance to + + – + +
etching
Resistance to + – + (+) +
electroplating
Heat + ** – – + –
resistance of
resist mask

* symbols: + important; (+) less important; – unimportant.

** only for solder masks.

4.1.2. Function and Chemistry of Photoresists

The photoresist technique dates back to the early nineteenth century [352], [377]. Until now, an exceptionally large number of
chemical systems and compositions have been tested for their suitability as photoresists for many different applications [352],
[353], [361], [362], [378-381]. Many of these systems have been used in practice, but only a small selection is presented
here.

4.1.2.1. Positive Photoresists

Diazonaphthalinone Systems. All positive photoresists used on a large scale in industry for the production of integrated
circuits are liquid systems. These systems are based on the light-induced Wolff rearrangement [382-384] of
diazonaphthalinone derivatives (2-diazo-1-naphthalinones) (1) which give 3-indenecarboxylic acid in the last step.
page 62 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Apart from compounds of type 1, derivatives of diazonaphthalinone-4-sulfonic acid are also used for positive photoresists (R
represents primarily aryloxy groups, but occasionally also substituted amine residues).

The differentiation principle is also clearly demonstrated: namely, the conversion of oleophilic compound 1, which is insoluble
in aqueous systems, to alkali-soluble compound 2. However, the situation is more complicated in the case of industrial resists
because they contain a novolac binder (3) (a non-self-curing phenolic resin) in addition to the light-sensitive component (1)
[353].

The fact that these resist masks do not swell during development in buffered aqueous alkali-metal hydroxide or
tetramethylammonium hydroxide solutions contributes to their high resolution potential. They are used in the long- and
middle-wavelength UV range. Their sensitivity to more energetic radiation is generally too low.

Systems That Are Fragmented or Depolymerized by Radiation. High molecular mass poly(methyl methacrylate) (4; see
Fig. 72) is fragmented by energy-rich radiation (shortwavelength UV light of 230 – 280 nm; electron beams, ion beams, and
X-rays) to give fragments that are more readily soluble (5) [385]. Mixtures of alcohols and ketones are used as developer.
This high-resolution positive resist is used primarily as an electron beam resist and has low radiation sensitivity.

Figure 72. Fragmentation of high-molecular mass poly(methyl methacrylate) by energy-rich radiation

A) High-molecular, slightly soluble; B) Low-molecular, readily soluble

A more sensitive electron beam resist is polybutene sulfone (6). This compound can be partially depolymerized to its volatile
comonomers by electron beams. Here, too, mixtures of alcohols and ketones are used as developer.

Both basic types of resists have been modified in many ways [381], [386].

Photocatalytic Systems. As a rule, these systems contain only small amounts of a compound that produces a catalyst
(usually an acid) on exposure to radiation. This, in turn, catalyzes the conversion of other components of the layer that are
not affected by radiation, causing an increase in the solubility of the layer.

Examples of such systems are

1. the radiation-induced, acid-catalyzed hydrolysis of orthocarboxylic acid esters (7) in combination with novolacs, to give
alcohols and carboxylic acid esters [353] (Fig. 73); and
2. the radiation-induced, acid-catalyzed fragmentation of poly(p-tert-butyloxycarbonyloxystyrene) (8), which is insoluble in
aqueous alkali, to give alkali-soluble poly(vinyl phenol) (9), as well as isobutene and carbon dioxide (see Fig. 74) ,
[388].

page 63 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 73. Photocatalytic system, photolytic acid donor novolac orthoester

R1, R2, R3 = alkyl groups

Figure 74. Photocatalytic system, photolytic acid donor poly(p-tert-butyloxycarbonyloxystyrene) [352]

Photoresists for all common types of radiation can be formulated by suitable selection of compounds that yield acids on
exposure to this type radiation. Thick layers are possible.

4.1.2.2. Negative Photoresists

Negative-working photoresists quantitatively represent the largest part of all photoresists used commercially.

Azide Systems. The negative system most often used in the production of integrated circuits consists of partially cyclized
polyisoprene (10) and an organic azide (e.g., a bisazide such as 11; see Fig. 75) as the light-sensitive component. On
exposure to radiation, nitrogen is eliminated from the azide groups to yield highly reactive nitrenes, which in turn react with
the cyclorubber and reduce its solubility [353]. The resist mask swells considerably during development with aromatic
hydrocarbons.

Figure 75. Azide-based photoresist

A) Partially cyclized polyisoprene (important structural elements)

B) Organic azide sensitive to radiation

Negative photoresists based on poly(vinyl phenol) – organic azide, which can be developed with aqueous alkaline solutions,
react more favorably in this respect [389]. Azide-based photoresist systems are applied preferentially in the short- to long-
wave UV range.

Photodimerization Systems [390]. The oldest commercially employed negative photoresists based on (sensitized)
polymeric cinnamic acid esters [391], e.g., poly(vinyl cinnamate) [392], belong to this group. On exposure to light, the
(sensitized) photodimerization of neighboring cinnamate groups to cyclobutane structures results in the reduction of solubility
[393].

Another system used is photodimerization of chalcone (phenyl styryl ketone) groups (e.g., in epoxy resins based on 4,4′-
dihydroxychalcones). Once exposed and developed, systems of this type permit the production of a resist mask that can be
thermally cured (solder masks) [394].

The spectral sensitivity to light of the system (a part of which is sensitized) lies in the short- to long-wave UV range. Most of
these systems must be developed with organic solvents.

Photoinitiated Systems. Analogous to photocatalytic systems, photoinitiated systems contain only small amounts of a
photoinitiator which is converted to an active species on exposure to radiation. This species, in turn, initiates a chain reaction
that results in a reduction of solubility (application in thick layers is possible).

Photoinitiated Polymerization. A number of review articles describing the principles and versatile commercial aspects of
these systems are available [354], [395-397]. Only radical polymerization has attained real importance. Negative dry film
resists, which are generally formulated as described below, represent the most important application.

Main components
1. Photoinitiator system that yields radicals on irradiation [397]
2. Radically polymerizable monomers (e.g., esters and amides of acrylic and methacrylic acids)
3. Polymeric binders (e.g., polymethacrylates)
Modifying components (among others)
1. Thermal polymerization inhibitors
2. Adhesives
3. Plasticizers
4. Dyes

As far as quantity is concerned, components 2 and 3 account for the largest part of the photoresist composition. Depending

page 64 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
on the polymeric binder, either an organic solvent or aqueous alkali is required for development. In fact, the latter is
becoming increasingly important. Examples of layer formulations are given in [398-400]. Photopolymerizable polyamide –
carboxylic acids are used in the production of high-temperature-resistant masks for the manufacture of integrated circuits
[401].

Thiol – Ene Systems. Thiols add to alkenes in photoinitiated radical chain reactions to give thioethers [402]. This principle is
exploited in solvent-free liquid systems [365].

Other Systems. Copolymers of the glycidyl ester of methacrylic acid and ethyl acrylate are of some importance as negative
electron beam resists.

4.1.3. Testing Methods

Apart from controlling purity and identity by standard chemical and physical techniques, the testing of photoresists includes a
series of function tests, e.g., determination of light sensitivity, resist adhesion, coating properties, and particle content. The
methods used are naturally more numerous and extensive for photoresists intended for microapplication [403] than for those
used to produce relatively coarser structures [404].

4.1.4. Economic Aspects and Suppliers of Photoresists

Liquid photoresists are applied mainly in the field of microlithography. However, liquid systems are again gaining importance
in the printed circuit sector because of economic factors and the higher resolution attainable. An assessment of market
development is given below [405].

World market, ($×106)

1984 1989

Liquid resists (microapplications)

Negative 73 195

Positive 73 265
Dry film resists
Negative 325 875

Some suppliers of liquid and dry films resists are listed in Tables 4 and 5, respectively.

The following companies have a significant share of the market: Hoechst, Shipley, and Tokyo Ohka in the liquid resist
(positive) sector; DuPont and Dynachem in the dry film resist sector.

Table 4. Suppliers of dry film resists

Company Trade name Mode of operation

Positive Negative

Du Pont Riston x
Morton Thiocol
(Dynachem) Laminar x
Hercules Aquamer x
Hitachi Chemical Photec x
Hoechst Ozatec x x

Table 5. Suppliers of liquid resists for microapplications

Company Trade Name Mode of operation

Positive Negative

Hoechst AZ x
Olin Hunt Waycoat x x
Specialty Prod.
Shipley Microposit x

page 65 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Tokyo Ohka OFPR, x x
Kogyo TSMR, etc.
Union Carbide x x
Corp. (KTI)

4.1.5. Occupational Health and Environmental Protection

Safety regulations and instructions for use available from manufacturers must be followed carefully when handling
photoresists. Some comments on the physiological effects and safe handling of a series of components of UV-hardenable
systems are given in [406] and [407].

Careful attention must be paid to avoid possible hazards from light sources and UV radiation [408].

Injunctions provided by law on the limitation of solvent release into the environment and the prevention of water pollution
must be met when handling photoresists (coating with liquid resists; wash-out development of imagewise exposed resist
layers with organic solvents or aqueous solutions). Advanced techniques are available for the removal of solvents from outlet
air (e.g., adsorption by active charcoal, followed by recovery) and the separation of dissolved photoresists from aqueous
developer or stripper [409-412].

4.2. Printed Circuits (Printed Circuit Boards)

The term printed circuits implies that an electrical circuit is produced by printing. In a strict sense, only the thick-film circuit
printed with the aid of the screen printing technique satisfies this description. In this case, conductors, resistors, dielectrics,
etc., are produced by imprinting: e.g., conductivity pastes, resistance pastes, and dielectric pastes on ceramic substrates
(usually aluminum oxide) and subsequent burning-in [413].

Today, the term printed circuit includes all variations of the printed circuit board. The printed circuit board can be made from a
rigid or a flexible insulant carrier. Wiring paths are arranged on one or both sides of the board, with conducting connections
between the two levels if necessary.

To achieve the high wiring path density demanded by today's sophisticated applications and to prevent exceedingly fine
conducting tracks, these can also be distributed among several parallel levels connected to each other through the insulating
material (multilayer technique) [414]. The printed circuit board is the base for the assembly of functional components; it
contains all the conductive connections required among components. In case of high circuit density, wiring paths are usually
covered with a solder mask, which leaves only the solder pads exposed. The printed circuit board frequently carries
information for the assembly as well (lettering print).

4.2.1. Methods of Producing Printed Circuit Boards

The many different and versatile techniques available for the production of printed circuit boards are described in detail in
[369], [415-420].

The actual production of wiring paths and plated-through holes can be achieved by using real printing processes (screen
printing, magnetography, and others) or photoresist techniques, also called photoprinting. In each case, the substrate (e.g.,
an insulating plate covered with a copper sheet) is first coated with a temporary selective resist coating. Wiring paths are
then produced by etching (substractive step), electroplating, or chemical metal deposition (additive step), possibly in
combination with etching. After this step, the resist coating is removed (stripped).

A special technique (fully additive process) for making wiring paths is available, which does not require a selective resist
coating [421], [422].

Solder masks can also be applied via both screen printing and photoprinting. These masks usually remain as the permanent
coatings part of the printed circuit board.

4.2.2. Masking Techniques

4.2.2.1. Exposure Masks [423], [424]
Production of a printed circuit board begins with a layout, which takes into account the requirements of the wiring diagram. In
fact, complicated printed circuit boards require the use of CAD (computer-assisted design) techniques to draw up the layout.
The layout is then transferred to a silver film of the required size (polyester or glass carrier) by using a computer-controlled
plotter (photoplotter or laser plotter). The resulting film is called an exposure mask (photomask). An exposure mask of this
type is required for each conductive level and for each solder mask. Exposure masks are used both in the production of
screen printing forms (see Section Screen Printing ) and in the exposure of photoresist layers in photoprinting.

4.2.2.2. Screen Printing

The screen printing process is the oldest technique for the application of resist coatings stable to etching, electroplating, and
soldering, as well as lettering in the production of printed circuit boards. The large number of special screen inks can be
simplified as follows:

Drying or hardening
One-component systems physical

page 66 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
oxidative
by heating
by UV light
Two-component systems by chemical reactions
(reactive hardening)

Two-component systems (e.g., mixtures of epoxide resin and hardening agent) and UV-hardening systems, in particular,
fulfill the requirements for a modern system because they yield resist masks of higher quality (resistance to chemicals,
physical properties).

Monofil (if necessary, yellow colored) polyester and steel screens are used preferentially in the production of screen printing
forms (see also Section Screen Printing ) for circuit printing. Factors responsible for achieving the highest possible resolution
are screen selection (number of filaments per centimeter and filament thickness), sharpness of edges of the stencil as well as
its size, resistance to deformation, dimensional stability, properties of the printing ink, and finally, execution of the printing
process [425].

As a rule, screen printing is used to make printed circuit boards on which the pattern width of the printing is at least 0.15 mm.
However, it is applied mainly in the production of much coarser structures. For more detailed information, see [426-429].

4.2.2.3. Photoprinting
Photoprinting gives considerably better resolution than screen printing and is now used widely in the production of printed
circuit boards. This masking technique makes use of photoresists (see Section Photoresists) to produce coatings that are
resistant to etching, plating baths, and soldering.

The entire surface of the substrate is coated with photoresist (dipping, curtain coating, roller coating, electrostatic spraying of
liquid resists, and lamination of dry film resists) and, in the case of solvent-containing photoresists, dried. Exposure is
achieved through an exposure mask, generating differences in the solubility of the coating. Proximity exposure is used in the
case of solvent-free liquid systems [365]. Also, laser exposure without the use of an exposure mask is being discussed at
present [422], [430].

Once exposed, the areas that have become soluble (positive process) or that have remained soluble (negative process) are
washed out by development, thereby selectively uncovering the substrate surface to be processed.

Photoprinting of printed circuit boards requires the application of negative photoresists, predominantly dry film resists, almost
exclusively [366]. In certain applications (e.g., the production of solder masks), liquid resists are being used increasingly for
technical or economic reasons [431].

4.2.2.4. Other Techniques

Magnetographic [351] and electrophotographic [432] techniques are also of some interest. Special techniques are required to
produce wiring paths on three-dimensional printed circuit boards [433].

4.2.3. Economic Aspects

Worldwide total sale of printed circuit boards in 1984 was estimated at $ 10×109. Materials used in the production of printed
circuit boards accounted for $ 9.2×109, with an annual growth rate of ca. 12 % [405].

4.2.4. Occupational Health and Environmental Protection

Questions connected with the use of photoresists are dealt with in Section Occupational Health and Environmental
Protection. Safe handling of strongly acidic or alkaline solutions and of heavy-metal compounds is the subject of a number of
special publications and of laws specific to each country. For further details, see [410] and [434].

4.3. Microelectronic Devices

Imaging technology for modern microelectronic applications must meet some unique performance requirements, such as

1. resolving features only two times larger than the exposure wavelength;
2. “unbelievably perfect” patterns with defect densities of fewer than one defect >0.2 µm per 10 cm2 (defects are defined
as flaws ≥0.2 µm);
3. the resist must mask a number of aggressive wet and dry etch environments;
4. imaging of features on substrates with widely varying reflectivity and steps equal in height to the final line width.

The characteristics follow directly from the manufacturing requirements of silicon-based integrated circuits and the available
exposure technology. Integrated circuits are the dominant microelectronic commercial devices produced by imaging.

4.3.1. Integrated Circuit Manufacturing [435-437]

An integrated circuit consists of a threedimensional structure of electrically conductive regions in the surface of a silicon
wafer (Fig. 76). The regions vary in physical properties as evidenced by the resistivity changing from 10–6 Ω/cm2 (metal)
through 10–1 – 104 Ω/cm2 (doped silicon, Semiconductors) to infinity (silicon dioxide). The conductive regions are
assembled into transistors, capacitors, etc., and linked together to process either analog or digital signals. Digital signal
processing leads to computer applications with either metal – oxide – silicon (MOS) field effect transistors or bipolar

page 67 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
transistors ( Semiconductors); MOS is the simplest technology and is used in high-density, cheap memories and
microprocessors. The bipolar technology is more complex but gives faster switching speeds and is used in high-speed
applications.

Figure 76. Perspective schematic of a 1 µm transistor [435]

A typical integrated circuit is ca. 15 mm on one side and is manufactured on a silicon wafer 100 – 200 mm in diameter, so
that >100 circuits can exist on an individual wafer. After manufacture, the circuits are tested, and the wafer is diced to remove
the good die. Yields can vary from 1 – 2 % for an experimental product to >90 % for a proven one. The yield determines the
commercial viability of a production circuit.

The processes used to form the three-dimensional structures in silicon can be divided into vertical and horizontal operations.
Vertical operations generate layers in the silicon by

1. growth (e.g., oxidation of silicon to silicon dioxide);

2. deposition (e.g., chemical vapor deposition of silane and oxygen to form silicon dioxide) or processes that transfer
material from one solid phase to another via a gas (e.g., evaporation or sputtering); and
3. doping, either by diffusion of gas at high temperature or by scanning an accelerated beam of ions across the wafer
(ion implantation).

Horizontal structures are produced by transferring a pattern from a mask into the photoresist, developing the photoresist, and
then transferring the pattern into the underlying device by etching (Fig. 77). The mask usually consists of a pattern etched in
chrome on a glass substrate; the manufacturing process is discussed in the following section. Transfer from the photoresist
into the substrate proceeds via wet or dry etching. A wet etch reacts with the surface to give water-soluble products. In dry
etching, either a plasma is used which reacts with the surface to yield volatile products, or the surface is sputtered away in a
solid – gas – solid transfer. Reactive ion etching is a combination of plasma and sputter etch designed to give anisotropic
(depth etches faster than width) etching. In ion implantation (vertical operation), a resist or etched silicon dioxide is often
used as an implantation mask so that a vertical layer is formed and the pattern is transferred simultaneously. A typical device
requires 10 – 15 layers and, hence, a similiar number of mask levels and etching operations.

Figure 77. Process sequence for a “mask level”. The sequence is repeated about ten times to produce the final device.

Defect Density Requirements. The scale of integrated circuits is most difficult to comprehend; 1 Mbit random access memory
is taken as a practical reference scale. The minimum line width of 1 µm is imaged by using light with a wavelength of 436 nm.
Transistor performance must be controlled to ± 10 %, so line width must be controlled to the same tolerance (i.e., ± 1/4 of the
exposure wavelength). These lines must be controlled across a 150-mm wafer, with a wafer width – line width ratio of >105.
The critical defect size is 0.2 µm, which follows from line width control requirements. The defect density can be derived from
the circuit size ( 1 cm2) and the number of mask levels ( 10). If a defect exists on any level, the device will fail; therefore,
defect density should be <1 (defect) per 10 cm2.

4.3.2. Photoprocess [438-441]

The high degree of control and low defect density described in the previous section place some unique constraints on the
photoprocess. All processing must be done under clean-room conditions, on automated equipment that minimizes particles
generated by handling. The schematic in Figure 77 can be used to follow the discussion.

Substrate. One of the most significant differences between microelectronic lithography and most other lithographic
applications arises because the optical properties of the substrate change from mask level to mask level, and across
individual wafers. The optical properties are a function of the properties of the top layer of the wafer at each mask level. In
most cases, this top layer is uniform and unpatterned, but the refractive index varies from mask level to mask level, thus
causing reflectivity changes. If the top layer is a dielectric, it will be transparent and the patterned underlying layers will cause
local reflectivity variations across the wafer.

Steps also occur in the substrate that are caused by the underlying patterned layers which can be of the same height as the
line width. These steps cause variations in resist thickness which, in turn, lead to variations in line width. If the steps are
covered by highly reflective metal or polysilicon, scattered light from the steps also causes line width variation. Finally, grain
in materials is also a source of light scattering.

Surface Preparation. Surface preparation is essential to ensure adhesion of many photoresists. Freshly prepared surfaces
without adsorbed water give the best adhesion. Adhesion of substrates is often improved by removing the surface water by
dehydration baking. Aqueous liquids with a high affinity for hydrophilic surfaces (metals or silicon dioxide) exhibit the worst
adhesion. Adhesion promoters are often used, especially with optical positive resists that employ aqueous developers. The
most common adhesion promoter is hexamethyldisilazane, which makes surfaces hydrophobic before the resist is applied.

page 68 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Selection of Photoresists. The selection of a photoresist is the result of cost – performance analysis. Two classes of
material are in common use: (1) negative cross-linking systems based on cyclized butadiene, and (2) positive materials
based on novolac resins, or naphthoquinone diazides which improve the dissolution rate in aqueous alkali on exposure.

The performance criteria and relative performance of the materials are summerized in Table 6. The dominant factor is
resolution in terms of the minimum resolvable feature and the profile quality of the resolved lines. Cross-linking negative
resists swell in the developer and limit the resolution to 2 – 3 µm. Positive resists developed in aqueous alkali do not appear
to swell at all, and standing waves <0.1 µm in size can be resolved easily. Swelling also produces rounded profiles in
negative resists in contrast to almost vertical walls in positive ones.

Table 6. Comparison of positive and negative resists

Negative Positive

Resolution <2 – 3 µm <0.1 µm

Profile round vertical
Process control good good
Adhesion excellent good
(adhesion promoter)
Dry etch resistance good excellent
Sensitivity, mJ/cm2 20 200 *

* Modern equipment can expose at these doses.

Positive resists are three to five times more expensive than negative ones. If features of <4 µm must be resolved, positive
resists should be chosen.

Coating. The objective of coating is to produce layers with a uniformity >20.0 nm to ensure uniform optical properties (no
“thin-film interference”): 2 – 3 mL of resist is placed in the middle of the wafer, and the wafer is spun at 3000 – 4000 rpm,
causing rapid solvent evaporation in the presence of high shear forces on the photoresist. The puddle that is initially
0.5 cm thick is spun down to a thin film of 1×10–4 cm with the desired uniformity. The formation of convection cells, which
produce surface irregularities (striations), can be prevented by modifying the surface tension of the resist [442].

Baking. The objective of baking is to remove the solvent without degrading the photosensitizer. The three commercial
options are

1. convection ovens in which the wafers are placed in a hot circulating airstream;
2. hot plates in which the wafers are held in contact with a large thermal mass (a large mass, thermally heated); and
3. infrared ovens that use radiant heating of the wafer.

Hot plate baking gives better thermal coupling, can be automated easily, and is the current standard technique. In many
resists, a significant amount of solvent remains in the layer at a temperature that can degrade the photosensitizer [443].
Because the solvent affects the dissolution rate, control of the baking operation is critical.

Exposure. The objective of exposure is to transfer the image from the mask to the wafer and convert the photosensitizer
locally to form a “latent image.” All commercial exposure equipment uses high-pressure mercury lamps, with the 436-, 408-,
and 365-nm lines overlapping the sensitizer absorption curve (see Section Diazotyping).

A variety of commercial exposure equipment is available. From the point of view of imaging alone, the differences can be
summarized as (1) image transfer process, (2) image field shape, (3) mask size, and (4) illumination spectrum.

Two image transfer processes exist: (1) contact – proximity process in which the mask is placed in contact with, or close to,
the wafer; and (2) projection in which the mask is imaged through a lens. In the contact – proximity process, image quality is
determined by near field diffraction from the mask, which results in excellent images in contact printing. The major problems
in contact printing are the defects caused by forcing mask and wafer together. Projection imaging is controlled by far field
diffraction effects associated with the lens.

Projection technologies are further classified by the shape of the image field. The whole wafer cannot be imaged directly, and
different optical configurations produce differently shaped image fields that must be moved to cover the whole wafer.
Scanning cameras have optics that produce a narrow slit field as wide as a wafer, which is scanned to cover the whole wafer.
Step and repeat cameras produce a rectangular field that is stepped across the wafer.

The mask size relative to the final image size is also a function of optics. In contact printers and projection machines with
reflective optics, mask and image are of the same size “1 x.” By contrast, machines with transmissive optics use reduction so
that the mask is five to ten times larger than the final pattern, which facilitates mask making.

The illumination spectrum is also a function of optics. Machines with reflective optics use multiple mercury lines, whereas
machines with transmissive optics use a single wavelength (usually 436 nm) because a chromatic correction must be made.

Exposure systems are selected according to resolution of the lens, throughput, alignment, and focus accuracy. Step and

page 69 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
repeat cameras dominate high-resolution applications.

Postexposure Operations. Process operations after exposure are designed to modify the latent image in some way. The
best known physical modification is postexposure baking, which induces diffusion of the sensitizer in positive resists to
eliminate standing waves. Standing waves are generated by exposure machines that use a single wavelength. Chemical
modifications include image reversal of positive resists by baking in an ammonia atmosphere or simply by baking one
particular resist (AZ 5200) [444].

Developing. The objective of developing is to remove the exposed photoresist uniformly. The two commercial methods are
(1) batch developing in which a boat of 20 wafers is immersed in a tank, and (2) spray developing in which the developer is
sprayed on each wafer as it rotates. Spray development can be automated easily and is the dominant technology in modern
facilities.

Hard Baking. The objective of hard baking is to remove the solvent and cross-link the resist to improve thermal stability and
adhesion. The commercial implementation options are the same as in any other baking operations.

Stripping. After all processing steps have been completed, including etching and ion implantation, the photoresist must be
removed without residues. The two options are wet or dry stripping. Wet stripping agents can be subdivided further into
solutions that oxidize the resist (sulfuric acid – hydrogen peroxide mixtures) and solutions that dissolve the resist (polar
solvent – water mixtures). Oxidizing systems remove all water-soluble oxides and produce the cleanest surface. Of course,
they cannot be used if the wafer surface is oxidizable (e.g., a metal). Solutions that dissolve the resist remove all soluble
material, so strongly cross-linked resist may be left behind.

Dry stripping agents are either oxygen-plasma or ozone based, and they remove all compounds that form volatile oxides.
Unfortunately, metal contaminants in the resist, such as sodium and iron, form nonvolatile oxides so they are always left after
dry stripping. Some form of combined wet – dry stripping is used for the most severe cases.

Mask Making. Many types of mask are available, including silver halide and bright or antireflective chrome masks, generated
by electron-beam or optical equipment. For large-scale integrated circuits, mask making is nearly always done by electron-
beam lithography on antireflective chrome. The mask making process is a very simple integrated process using photoresists
on a chrome layer on a glass plate. The chrome is wet-etched to generate the mask. Antireflecting chrome is chrome that
has been surface oxidized. The mask will be 1, 5, or 10 times the size of the final pattern, depending on the optics of the
production lithography machine. The various lithography machines are discussed in more detail in the following sections.

4.3.3. Techniques for Submicron Lithography [440], [445]

Because of the increased integration of devices, techniques must be developed to integrate smaller features.

These can be obtained by

1. resolving smaller features with existing lenses,

2. improving the resolution of lenses by using either shorter wavelengths or lenses with a larger numerical aperture (NA),
or
3. higher resolution imaging technology such as electron beam or X-ray.

Resolving Smaller Features with Existing Lenses. To resolve smaller features, line width control must be maintained
although the quality of the image delivered by the lens becomes poorer. The image quality, as measured by the modulation
transfer function (MTF) for sets or equal lines and spaces, decreases rapidly with decreasing feature size (Fig. 78 B).

Figure 78. Schematic showing the definition of modulation transfer function (MTF) (A) and the variation of MTF with
resolution (B). A partial coherence of 0.7 is the value for most imaging systems

Experience has shown that single-layer lithography can be produced down to an MTF of 0.8 [445]. A 436-nm lens with an NA
of 0.35 will print 1-µm lines and spaces with an MTF of 0.8. Multilayer techniques have been used to make devices at an
MTF of 0.6 [446]. Below an MTF of 0.6, the image becomes so blurred that light from neighboring features overlaps and
produces line widths that vary depending on the number and size of these neighbors (“optical proximity effect”) [446]. These
effects cause very complex line width variations that are probably uncorrectable; hence, an MTF of 0.6 represents the
practical resolution limit of a lens [445].

Practical lithography at modulation transfer functions between 0.8 and 0.6 must use processes with exceptional control
because the image is so poor. These process variations that determine the line width can be grouped as follows:

1. Variations of incident image (e.g., dose, focus)

2. Variations of material dissolution rate (e.g., reflectivity, steps)

The classical solution to the first two problems is to use thin, very transparent resists. Unfortunately, thin resists give poor line

page 70 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
width control across steps, and transparent resists are very sensitive to reflections.

Many advanced resist processes have been described that attempt to reduce the effects of the three different process
variations on line width. Examples are image reversed positive resists [444], contrast enhancement materials [447], and
multilayer resist systems [446] including the “built on mask” (BOM, a process for producing a chemical mask in situ) system
[448]. In general, all these systems give excellent wall profiles and, hence, good line width control over steps. Most of them
improve sensitivity relative to image variations. Substrate reflection remains the most difficult problem to eliminate. Any
processes in which light from the initial or any pattern transfer exposure hits the underlying substrate can cause problems.
Only processes in which etching is used for pattern transfer are completely free of reflectivity problems. All these systems
attempt to gain process control at the expense of additional process complexity. This complexity results in higher costs and
additional defects.

Improved Lenses. The resolution of a lens is proportional to the ratio of wavelength to numerical aperture; i.e., as the
wavelength decreases, diffraction is reduced, and as the NA (“size”) of the lens increases, more of the diffracted light is
captured and resolved back onto the wafer (Fig. 79 A). Reducing the wavelength becomes progressively more difficult
because of decreasing glass transmission, and it becomes extremely difficult at ≤310 nm where quartz is the only
conventional lens material available. Consequently, chromatic abberations at short wavelengths are very difficult to correct,
and very narrow band excimer laser light sources are being investigated.

Figure 79. Advanced resist techniques

A) Optical lithographic system; B) X-ray system; C) Electron-beam system

Conventional positive resist based on diazo compounds works well down to 310 nm, although special diazo sensitizers give
the best results [444]. At ≤310 nm, the aromatic polymers that provide high dry etching resistance are very absorbing, so
alternative resins (e.g., copolymers of styrene) are needed. Very intense laser light sources can also result in different
(potentially beneficial) photochemistry.

Resolution may also be increased by increasing the NA, with the disadvantage of a significantly reduced depth of field. The
depth of field is inversely proportional to the square of NA and increases linearly with wavelength. The sensitivity of the
process to focus changes can be reduced by using a thin imaging layer and by planarizing the substrate as much as
possible. A multilayer system may be essential for use with lenses of very high NA (>0.4).

X-Ray Lithography. X-ray lithography involves proximity printing by using very short-wavelength photons to reduce
diffraction effects (Fig. 79 B). The mask is the crucial point in X-ray lithography because the mask base must transmit soft X-
rays. The base is typically a 5-µm membrane of silicon carbide or boron nitride, which provides mechanical strength and
stability, with a gold absorber pattern. The mask must be of the same size as the final pattern (“IX“). The technological
problems associated with masks of low defect density represent the greatest unknowns in X-ray lithography.

Three X-ray sources are possible: generation by (1) electron impact, (2) plasma, and (3) synchrotron rings. Only synchrotron
rings have sufficient intensity for submicron step and repeat lithography. The synchrotron ring forces high-energy electrons
around corners to generate X-rays. The principal problem with the synchrotron ring is its cost, which is estimated to be $ 5 –
10×106.

Even with a synchrotron source, photoresist must have a quantum efficiency one to two orders of magnitude greater than
conventional materials. Catalytic resist systems in which a photogenerated species catalyzes degradation of 50 – 100
dissolution inhibitor molecules appear to be the most satisfactory.

Electron-Beam Lithography. Electron-beam lithography differs from other technologies in that it is a primary pattern
generation technique. A focused electron beam is scanned across a wafer and switched on and off at each location to
delineate the pattern. Throughput is a function of the number of locations that must be visited and the length of time that the
beam must reside at each location to expose the resist. The residence time is a function of beam intensity and resist
sensitivity. To meet a throughput of 30 wafers an hour requires sophisticated high-speed electronics and very fast resists.
The major drawback for its use in submicron lithography is that the number of locations varies with the square of the
minimum feature sizes, which makes throughput very sensitive to minimum feature size.

The resist requirements are similar to X-ray, and similar catalytic systems are being evaluated. The exposure of patterns by
electrons backscattered from the substrate (proximity effect) causes line width variation.

[Top of Page]

5. Photochemical Machining
David Allen

5.1. Introduction
Using techniques similar to those employed for the production of printed circuit boards (see Section Printed Circuits (Printed
Circuit Boards)) and integrated circuits (see Section Microelectronic Devices), the photochemical machining (PCM) industry
plays a valuable role worldwide in the production of precision parts and decorative items. Examples are color television
shadow masks; integrated circuit lead frames; magnetic recording head laminations; heat ladders, plates, and sinks;
attenuators, light choppers, and encoder disks; grills, grids, sieves, mesh, and screens; washers, shims, and gaskets;
jewelry; and decorative signs, plaques, and nameplates (Fig. 80).

page 71 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 80. A selection of parts made by PCM

In the multistage PCM process shown in Figure 81, the use of photoresists (see Section Photoresists) ensures higher
resolution than that obtained by traditional metal-cutting techniques. Details are given in [449].

Figure 81. The modern PCM process

Terminology. Photochemical machining is also known as photoetching, photochemical milling, photomilling,

photofabrication, and photochemical etching. In the United States, chemical blanking and photoforming (a term used in
Europe for a very different technique, see Section Economic Aspects) are also synonyms for this technique.

5.2. Artwork Generation and Phototool Production

Various methods exist for the production of phototools. Some employ the manufacture of enlarged artwork (ink drawings;
scraperboard art; adhesive colored tapes or lettering on card; or, most common today, computer-aided artwork generated
from scribed “cut and peel” film) which is subsequently photoreduced to the correct size.

Other methods produce a phototool of the correct size by using a photoplotter or a laser pattern generator to expose the
photographic emulsion selectively according to computer-aided design data.

Once a master image has been made, it may be processed photographically by “step and repeat” and contact printing to
form a multiple-image, registered, double-sided phototool.

Ultimately, the future may see the demise of phototooling. At present, photoresists can be imaged directly from CAD data by
using UV lasers. To achieve economical, rapid, high-quality imaging, precision control of a high-resolution, high-powered,
low-cost laser is essential.

5.3. Materials
Virtually all materials can be etched, although some are etched more easily than others. Etching is basically rapid, controlled
corrosion. Thus, corrosion-resistant materials are difficult to etch and require extremely corrosive etchants. Some metals
(tantalum, gold, titanium) [450] and alloys (Elgiloy and Duratherm 600, containing 43 wt % and 41 wt % cobalt, resp.) [451]
fall into this category, together with ceramics [452], glass, and plastics. However, many materials used commonly in
manufacturing can be etched readily by using aqueous metal salt solutions or dilute acid or alkali (see Section Etching
Technology).

5.4. Photoresist Systems

Photoresists are described in Section Photoresists.

Due to the nature of PCM, which involves the production of parts from sheet metal and foil with substantial thickness,
photoresist resolution is not usually a problem.

Adequate resolution can be obtained by using a relatively thin photoresist coating und UV light as an exposure source. This
coating can be applied to the metal as a liquid or solid film. Photoresists are commonly applied by

1. dipcoating (withdrawal of the metal sheet through a meniscus of liquid);

2. flowing (used, for example, in the mass production of television shadow masks from continuous coils of mild steel);
and
3. laminating dry films of photoresist onto metal sheets by a combination of heat and pressure.

In addition to the photoresists mentioned in Section Photoresists, other negative-working photoresists used in PCM include
formulations based on fish gelatin and casein sensitized with ammonium dichromate.

The photoresist is exposed with UV light through a phototool and forms a stencil after development.

The developer may be an organic solvent or an aqueous solution. For environmental reasons, the use of organic solvent
developers is decreasing and that of aqueous developers is increasing, especially for dry film photoresists.

page 72 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
After development and rinsing, liquid resists are baked to increase their resistance to the etchant. Because etching usually
produces appreciable undercut (see Section Etching Technology), most of the resists used in PCM are negative-working as
their stencil edges are less prone to fracturing (cf. positive-working resists).

After etching, the resist stencil is usually stripped from the substrate. Again, environmental considerations favor the use of
aqueous strippers such as potassium hydroxide solution for dry film photoresists.

5.5. Etching Technology

Material not covered by the photoresist stencil is attacked chemically by the etching solution to form soluble products. For
etching to be economical, the etchants must be relatively innocuous, fast-acting on a wide variety of materials, and easily
disposed of when spent.

For these reasons, etching is usually performed with either iron(III) chloride solutions (suitable for copper or nickel and their
alloys, ferrous materials including stainless steel, and aluminum) or copper(II) chloride solutions which etch copper and its
alloys.

As material dissolves into the etchant, the concentration of impurities increases and the activity of etchant decreases. To
maintain a constant etchant activity and to lower costs, the technique of etchant regeneration is now being used widely. In
the ideal case of etching iron with iron(III) chloride and regenerating the etchant, excess iron(III) chloride is produced

The situation is more complicated when etching of alloys is considered. Regeneration methods involve (1) chemical
treatment with gases such as chlorine and ozone, or oxidizing agents such as sodium chlorate; and (2) electrolytic treatment
based on anodic oxidation of iron(II) to iron(III) ions.

Similarly, economic advantages accrue if copper(II) chloride solution, used as an etchant for copper, is regenerated:

The economics of each process depends on plant and chemical availability and energy costs, but cost reduction factors of
ten have been quoted in the United States [453]. High carbon content in alloys results in a lower cost reduction factor [454].

When a metal is being etched downward, a sideways etching under the stencil also results, leading to undercut (see Fig. 82).
This is responsible for some of the limitations of PCM (see Section Process Capability), such as the dependence of minimum
hole size on material thickness.

Figure 82. Development of etched edge profiles and undercut profile (a) (b) (c) (d) as etch time increases

5.6. Process Capability

The main advantages of PCM are as follows:

1. The method is a low-temperature technique that does not affect the physical or chemical properties of the starting
material.
2. The saying “If you can draw it, PCM can make it” is basically true, provided the limitations discussed below are
considered. The PCM process is ideal for creating complex metal parts involving arrays of apertures and shapes that
are difficult to obtain by its traditional rival, stamping (see Section Economic Aspects).
3. Although usually associated with the production of planar components, PCM may be used to etch the surfaces of
three-dimensional components based on cylindrical, conical, or spherical geometry [455].
4. Lead times are short so that the field test of a new product can start very quickly. If modifications are required, they are
inexpensive and quick to make.
5. No burrs or distortions are produced during manufacture.
6. Components may be produced separately or tabbed into a sheet for ease of handling.

In general, the main limitations of PCM, in material of thickness T, are

1. minimum hole diameter min = 1.1 T,

2. upper thickness limitation of 1.5 mm (although 6 mm is possible for low-resolution work),
3. some deviation from a classical straight profile at the edges of parts, usually controlled to <0.25 T.

5.7. Economic Aspects

Every manufacturing method has its advantages and disadvantages, and should be selected accordingly. If the part to be

page 73 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
produced is at the limit of the process capability of the method, the yield of acceptable products will be poor, which is
reflected in higher production costs.

Other processes, which may be competitive with PCM, are, therefore,

1. photoforming, (i.e., electroforming through a photoresist stencil);

2. wire electro-discharge machining, i.e., spark erosion using a thin wire as the cathode, a dielectric fluid such as water,
and sheet metal as the anode and workpiece;
3. laser beam machining, i.e., laser cutting with an intense, focused heat source, as obtained, e.g., with a carbon dioxide
laser;
4. stamping or fine blanking and piercing.

In deciding which method to use for part production, the technical aspects of the methods should be checked to ensure the
selection of appropriate techniques for the material used and its thickness as well as part specifications in terms of edge
profile, flatness, and dimensions. A study of model parts capable of being manufactured by all five methods determined that
the factors influencing the most economical method of production were part complexity, batch size, and part thickness [456].
Generally, large-volume production favors stamping; very low volumes favor wire electro discharge machining; high
complexity favors photochemical machining; thicker components favor laser beam machining; thinner components favor
photochemical machining, and extremely thin (<0.025 mm thick), high-resolution components in metals capable of being
electroplated favor photoforming.

5.8. Products
The range of products is described in the introduction, but the number of products made worldwide by PCM is difficult to
estimate. However, 1988 output figures from Japan include 14×109 integrated circuit lead frames, 110×106 vacuum
fluorescent display meshes, and 36×106 color television shadow masks (including 10×106 high-resolution masks) made by
this technique [457].

[Top of Page]

References
1. J. H. Dessauer, H. E. Clark: Xerography and Related Processes, Focal Press, London 1965.
2. R. M. Schaffert: Electrophotography, 2nd ed., Focal Press, London 1975.
3. W. E. Haas, in J. Sturge, V. Walworth, A. Shepp (eds.): Imaging Processes and Materials, Neblette's Eighth Edition,
Van Nostrand Reinhold, New York 1989, pp. 375 – 405.
4. IS&T - The Society for Imaging Science and Technology http://www.imaging.org SPIE - International Society for Optical
Engineering http://www.spie.org
5. C. J. Young, H. G. Grieg, RCA Rev. 15 (1954) 441 – 484.
6. M. E. Scharfe, F. W. Schmidlin in L. Morton (ed.): Advances in Electronics and Electron Physics, Academic Press, New
York 1975, pp. 83 – 146.
7. D. M. Pai, R. C. Enck, Phys. Rev. B 11 (1975) 5163 – 5174.
8. I. Chen. J. Mort, J. Appl. Phys. 43 (1972) 1165 – 1170.
9. M. E. Scharfe: Electrophotography, Principles and Optimization, Research Studies Press, Letchworth, England 1984.
10. D. M. Pai, A. Melnyk, Proc. SPIE Int. Soc. Opt. Eng. 81 (1986) 83 – 95.
11. J. H. Neyhart, Photogr. Sci. Eng. 10 (1966) 126 – 132.
12. L. Cheung, G. M. T. Foley, P. Fournia, B. E. Springett, Photogr. Sci. Eng. 26 (1982) 245 – 249.
13. K. Kiyota, A. Teshima, M. Tanaka, Photogr. Sci. Eng. 24 (1980) 289 – 292.
14. M. Lutz, SPIE Proceedings 759 (1987) 35 – 39.
15. H. Pinsler, H. Walsdorfer, M. Lutz, J. Imaging. Technol. 16 (1990) 55 – 59.
16. E. J. Felty in O. Helwich (ed.): Proceedings of the Second International Congress on Reprography, Cologne, Germany
1967, pp. 40 – 46.
17. W. E. Spear, P. G. LeComber, Solid State Commun. 17 (1976) 1193 – 1195.
18. I. Shimizu, J. Non Cryst. Solids 35, 36 (1983) 773 – 778.
19. N. Yamamoto, Jpn. J. Appl. Phys. 20 (1981) 305 – 310.
20. D. M. Pai in J. Adler, H. Fritzche, S. R. Ovshinsky (eds.): Physics of Disordered Materials, Plenum Press, New York
1985, pp. 579 – 592.
21. J. A. Amick, RCA Rev. 20 (1959) 770 – 784.
22. P. Mark, Photogr. Sci. Eng. 18 (1974) 254 – 261.
23. K. Yoshida, K. Kinoshita, T. Kawamura, Appl. Opt. Suppl. 3 (1969) 170 – 175.
24. M. Lardon, E. Lell-Doller, J. W. Weigl, Mol. Cryst. 2 (1967) 241 – 246.
25. W. D. Gill, J. Appl. Phys. 43 (1972) 5033 – 5040.
26. G. Weiser, J. Appl. Phys. 43 (1972) 5028 – 5033.
27. P. J. Melz, J. Chem. Phys. 57 (1972) 1694 – 1699.
28. P. J. Regensburger, Photochem. Photobiol. 8 (1968) 429 – 440. Links
29. D. M. Pai, J. Yanus, Photogr. Sci. Eng. 27 (1983) 14 – 19.
30. G. Pfister, Phys. Rev. B 16 (1977) 3676 – 3687.

page 74 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
31. P. M. Borsenberger, P. M. Mey, A. Chowdry, J. Appl. Phys. 49 (1978) 273 – 279. Links
32. M. Stolka, J. F. Yanus, D. M. Pai, J. Phys. Chem. 88 (1984) 4707 – 4714.
33. P. J. Melz et al., Photogr. Sci. Eng. 21 (1977) 73 – 78.
34. A. Kakuta, T. Mori, H. Morishita, IEEE Trans. Ind. Appl. IA-17 (1981) 4 – 10.
35. L. B. Schein, A. Rosenberg, S. L. Rice, J. Appl. Phys. 60 (1986) 4287 – 4292. Links
36. E. G. Schlosser, J. Appl. Photogr. Eng. 4 (1978) 118 – 122.
37. E. S. Baltazzi, J. Appl. Photogr. Eng. 6 (1980) 147 – 152.
38. D. Winkelmann, J. Appl. Photogr. Eng. 4 (1978) 187 – 194.
39. R. O. Loutfy, A. M. Hor, A. Rucklidge, J. Imaging Sci. 31 (1987) 31 – 37.
40. N. C. Khe et al., Photogr. Sci. Eng. 28 (1984) 196 – 203.
41. K. Arishima, H. Hiratsuka, A. Tate, T. Okada, Appl. Phys. Lett. 40 (1982) 279 – 281. Links
42. W. J. Dullmage et al., J. Appl. Phys. 49 (1978) 5543 – 5554.
43. P. Borsenberger, A. Chowdry, D. C. Hoesterey, W. Mey, J. Appl. Phys. 4 (1978) 5555 – 5564.
44. D. Winkelmann, K. Reuter in J. Gaynor (ed.): Advances in Non-Impact Printing Technologies for Computer and Office
Applications, Van Nostrand, New York 1982, pp. 446 – 462.
45. D. Adam, H.-J. Humpert, S. Dreihöfer, H. Pinsler, M. Lutz, Proc. 13. Int. Conf. on Digital Printing Technologies, IS&T,
Seattle 1997, 245 – 247
46. J. Mort, D. M. Pai: Photoconductivity and Related Phenomena, Elsevier, New York 1976.
47. A. S. Diamond (ed.): Handbook of Imaging Materials, Marcel Dekker, New York 1991.
48. P. M. Borsenberger, D. S. Weiss: Organic Photoreceptors for Xerography, Marcel Dekker, New York 1998.
49. M. M. Shahin, Appl. Opt. Suppl. 3 (1969) 106 – 110.
50. Canon, US Patent 4 851 960, 1989 (S. Nakamura, H. Hirabayashi, J. Araya, N. Koitabashi).
51. S. Matsumoto et al., J. Imaging Sci. Technol. 39 (1995) 132.
52. H. Kawamoto, H. Satoh, J. Imaging Sci. Technol. 38 (1994) 383.
53. M. Lutz, J. Imaging Sci. Technol. 11 (1985) 256.
54. R. M. Schaffert, Photogr. Sci. Eng. 6 (1962) 197 – 215.
55. H. E. J. Neugebauer, J. Appl. Opt. 3 (1964) 385 – 393.
56. T. L. Thourson, IEEE Trans. Electron Devices, ED 19 (1972) 495 – 511.
57. L. B. Schein: Electrophotography and Development Physics, Springer Verlag, Berlin 1987.
58. J. W. Weigl in M. Hair, M. D. Croucher (eds.): Colloids and Surfaces in Reprographic Technology, American Chemical
Society, Washington, DC 1982, pp. 139 – 182.
59. L. B. Schein, Photogr. Sci. Eng. 19 (1975) 255 – 265.
60. W. A. Sullivan, T. L. Thourson, Photogr. Sci. Eng. 11 (1967) 115 – 120.
61. J. T. Bickmore, K. W. Gunther, J. F. Knapp, W. A. Sullivan, Photogr. Sci. Eng. 14 (1970) 42 – 51.
62. T. Takahasi, N. Hosomo, J. Kanbe, T. Toymono, Photogr. Sci. Eng. 26 (1982) 254 – 261.
63. R. B. Lewis, H. M. Stark in W. F. Berg, K. Hauffe (eds.): Current Problems in Electrophotography, "Chapter X", de
Gruyter, Berlin 1972, pp. 322 – 335.
64. J. T. Bickmore, M. Levy, J. Hall, Photogr. Sci. Eng. 4 (1960) 37 – 43.
65. K. A. Metcalfe, J. Sci. Instrum. 32 (1955) 74 – 75.
66. J. Dahlquist, I. Brodie, J. Appl. Phys. 40 (1969) 3020 – 3027. Links
67. C. E. Case, T. M. Lawson, 14. Int. Conf. on Digital Printing Technologies, IS&T, Toronto, 1998, pp. 226 – 230.
68. D. A. Seanor, Photogr. Sci. Eng. 22 (1978) 240 – 243.
69. T. Nylund, C. Cowan, J. Spence, L. Steele, 3rd Int. Cong. on Advances in Non-Impact Printing Technologies, SPSE,
San Francisco, USA, 1986, pp. 74 – 80.
70. R. W. Gundlach, C. J. Klaus, Photogr. Sci. Eng. 7 (1963) 14 – 19.
71. P. Cressman, J. Appl. Phys. 34 (1963) 2327 – 2330.
72. M. J. Mitsui, IEEE Trans. ED 19 (1972) 397.
73. J. R. Rumsey, D. Bennewitz, J. Imaging Technol. 12 (1986) 144.
74. J. Johnson, 10. Int. Congr. on Advances in Non-Impact Printing Technologies, IS&T, New Orleans, 1994, 311 – 313.
75. H. A. Hermanson, R. E. Drews, D. G. Parker, F. Tomak, S. Swackhamer, in S. W. Ing, M. D. Tabak, W. E. Haas (eds.):
Electrophotography, Fourth International Conference, SPSE, Springfield, VA 1983, pp. 541 – 570.
76. J. J. Eltgen, in S. W. Ing, M. D. Tabak, W. E. Haas (eds.): Electrophotography, Fourth International Conference, SPSE,
Springfield, VA 1983, p. 519.
77. V. Tulagin, J. Opt. Soc. Am. 59 (1969) 328 – 331.
78. P. Warter, V. Tulagin, L. Carreira, 3rd Int. Cong. on Advances in Non-Impact Printing Technologies, SPSE, San
Francisco, USA, 1986, pp. 228–332.
79. P. Cressman, G. C. Hartman, J. Chem. Phys. 61 (1974) 2740 – 2746. Links
80. G. Hartmann, F. Schmidlin, J. Appl. Phys. 46 (1975) 226 – 278.
81. W. L. Goffe, Photogr. Sci. Eng. 15 (1971) 304 – 308.
82. A. L. Pundsak, Photogr. Sci. Eng. 18 (1974) 642 – 651.
83. P. Selenyi, J. Appl. Phys. 9 (1938) 637 – 641.

page 75 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
84. G. L. Pressman, 2nd International Conf. on Electrophotography, SPSE, 1974, pp. 38 – 45.
85. K. S. Pennington, W. Crooks, “Resistive Ribbon Thermal Transfer Printing (R 2 T 2): A New Printing Technology”,
Technical Abstracts Second International Congress on Non-Impact Printing, Society for Photo-Optical Scientists and
Engineers, Springfield, Virginia 1984, p. 236.
86. W. Boles, Thermal Printhead Technology and Design, Hewlett-Packard Company, Desktop Computer Division, Fort
Collins, Colorado, Computer Design October, 1979.
87. Texas Instruments, Inc., US 3 982 093, 1976 (W. S. Henrion).
88. S. Shibata, K Murasugi, K. Kaminishi: “New Type Thermal Printing Head Using Thin Film”, IEEE Trans. Parts Hybrids
Packag. 12 (1976) no. 13, 223 – 230.
89. K. Nunomura, N. Koyama, T. Matsubara: “Ta-SiC Thin Film Resistors of Highly Reliable Thermal Printing Heads”, IEEE
Trans. Components Hybrids Manuf. Technol. CHMT-6 (1983) no. 4, 260 – 263.
90. R. Shaw, J. appl. Photogr. Eng. 2 (1976) 17 – 27
91. J. W. Weigl, Angew. Chem. Int. Ed. Engl. 16 (1977) 374 – 392
92. L. B. Schein: Electrophotography and Development Physics, Springer Verlag, Berlin 1987.
93. T. Nozawa, T. Suzuki, “Zirconium Nitride Thin Film Resistors for Thermal Print Heads”, Int. J. Hybrid Microelectron., 7
(1984) no. 3, 63 – 67.
94. W. Butler, “Technological Aspects of the Development of a Thin Film Thermal Printhead”, Int. J. Hybrid Microelectron.,
5 (1982) no. 1, 1981 European Hybrid Microelectronics Conference, Avignon, France.
95. S. Shibata, M. Ito, “Evaluation of Characteristics of Meander Heat Elements for a Thermal Printing Head”, Electronics
and Communications in Japan, 66-C, (1983) no. 4, 321 – 327.
96. P. A. McLaughlin, E. W. Stapleton, J. E. Tumbaugh: “Development of a Thick Film Thermal Print Head”, Tektronix, Inc.,
Beaverton, Oregon, International Microelectron Symposium, San Francisco, California October 22 – 24, 1973, Sess.
Paper 1 A-4. Available from International Society for Hybrid Microelectron, Inc. Park Ridge, Illinois.
97. W. Crooks, C. Hafer, W. Weiche, US 4 456 915, 1984.
98. G. F. Brooks, W. Crooks, W. Weiche, US 4 103 066, 1978.
99. E. K. Sichel, P. Scheng, J. I. Gittleman, Phys. Rev. Lett. 40 (1978) 1197; Phys. Rev. B Condens. Matter 18 (1978)
5712; Phys. Rev. B. Condens. Matter 24 (1981) 6131.
100. K. K. Shih, D. B. Dove: “Electrical Properties of Resistive Ribbon”, IBM J. Res. Dev. 29 (1985) no. 5, 519 – 526.
101. T. G. Twardeck, “Characterization of a Resistive Ribbon Thermal Transfer Printing Process”, IBM J. Res. Dev. 29
(1985) no. 5, 470 – 477.
102. S. Shibata, K. Hihei: “Development of 16 dots/mm Thermal Printing Head”, Int. J. Hybrid Microelectron. 6 (1983) no. 1,
International Microelectronics Symp. Philadelpia, PA, October 31 to November 2, 1983, pp. 531 – 536.
103. P. Barnwell: “Thick Film Thermal Print Heads”, Electrocomponent Sci. Technol. 6 (1980) no. 3 – 4, 135 – 139.
104. Y. Tokunaga, M. Yoshida, M. Zazawa: “Life Estimation of Thermal Print Heads Consisting of Ta2N Thin-Film
Resistors”, IEEE Trans. Components, Hybrids Manuf. Technol. CHMT 4 (1981) no. 1, 148 – 153.
105. K. Mashio: “Current Developments of Thermal Papers”, TAPPI Print Reprog. Test Conf. Proc., Rochester, New York
TAPPI, Atlanta, Georgia November 11 – 14,1979, pp. 133 – 139.
106. Graphic Controls Corp., US 4 500 354, 1982 (B. J. Katsen et al.).
107. IBM branch offices, IBM Program Product NO. 5796-PBH, Advanced Statistical Analysis Program (ASTAP) Reference
Manual, No. SH 20–1180–0.
108. K. J. Fowler, C. H. Hafer: “Thermal Paper Printer”, IBM Technical Disclosure Bulletin, December, 1983, pp. 3781 –
3782.
109. Y. Tokunaga, K. Sugiyama: “Thermal Ink-Transfer Imaging”, IEEE Trans Electron Devices ED-27 (1980) no. 1, 218 –
222.
110. W. J. Weiche, US 4 308 318, 1981.
111. S. Applegate et al.: “Implementation of the Resistive Ribbon Technology in a Printer and Correcting Typewriter”, IBM J.
Res. Dev. 29 (1985) no. 5, 459 – 469.
112. H. Findlay, K. Jones, US 4 453 839, 1984.
113. T. Saito et al.: “A Large Scale Hybrid Thermal Print Head”, Electro component Sc. Technol. 10 (1983) 261 – 267.
114. J. C. Hurt, W. B. Baril: “Thick Film Thermal Printer Temperature Studies”, Int. Microelectron Conf., Proc. of the Tech.
Prog., Anaheim, CA, February 11 – 13, 1975, New York, June 17 – 19, 1975, Industrial and Scientific Conference
Management, Inc., Chicago, Illinois 1975, pp. 129 – 134.
115. L. F. Valenti: “Thermal Printhead Technology”, Int. IGC Conference for Continuing Education Program, Amsterdam,
Holland April 1985.
116. H. Taub: “Thermal Ink Jet Printing”, IGC Conference, Key Biscayne, Florida October 27 – 29, 1985.
117. T. Yamada et al.: “High-Resolution Full Color Printer by Microdot Ink-Jet Printing Method”, SID Digest, 14 (1983) 102.
118. Lord Rayleigh: “Capillarity”, The Theory of Sound, 2nd ed. vol. 2, "CLXX", Dover Publications Inc., New York 1945.
119. W. Thomson, GB 156 897, 1867.
120. C. W. Hansell, US 1 941 001, 1929.
121. R. Elmquist, US 2 566 443, 1950.
122. R. G. Sweet, US 3 596 275, 1964.
123. C. H. Hertz, S. I. Simonsson, US 3 416 153, 1968.
124. C. W. Hansell, US 2 512 743, 1946.
125. E. L. Kyser, S. B. Sears, US 3 946 398, 1970.

page 76 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
126. S. J. Zoltan, US 3 683 212, 1970.
127. N. G. E. Stemme, US 3 747 120, 1971.
128. T. Hara et al., US 4 296 421, 1979.
129. M. Bovio et al., US 4 238 807, 1978.
130. C. H. Hertz, G. Rydgren: “Ink Jet Marking by Ultrasonic Jet Control”, The Second International Congress on Advances
in Non-Impact Printing Technologies, SPSE Arlington, USA Nov. 4 – 8, 1984, p. 99.
131. N. Kiyohiro et al.: “Measurement of Droplet Velocity Using High Speed Motion Pictures”, The Second International
Congress on Advances in Non-Impact Printing Technologies, SPSE, Arlington, USA Nov. 4 – 8, 1984, p. 20.
132. N. Bugdayci et al.: “Axisymmetric Motion of Radially Polarized Piezoelectric Cylinders Used in Ink Jet Printing”, IBM J.
Res. Dev. 27 (1983) 171 – 180.
133. E. P. Hofer: “Spot-Size Modulation in Drop-on-Demand Ink Jet Technology”, SID Digest, New Orleans, USA 16 (1985)
321.
134. W. Runge: “One-Dimensional Simulation of a Phase-Change Cycle in Thermal Ink Jet Printers”, SID Digest, New
Orleans, USA 18 (1987) 189.
135. M. Ikeda: “The Bubble Jet Printing Technology for BJ-80 Printer”, IGC 18th Annual Conference on Ink Jet Printing,
Amsterdam, March 27 – 29, 1985.
136. C. H. Hertz, B. Samuelsson: “High Quality Ink Jet Printing of Color Images”, The Third International Congress on
Advances in Non-Impact Printing Technologies, SPSE, San Francisco, California, USA August 24 – 28, 1986.
137. F. Schmidlin, IEEE Trans. Electron Devices 19 (1972) 448 – 457.
138. R. N. Mills, Display, 1987 International Symposium, New Orleans, Louisiana, USA May 11 – 15, 1987.
139. G. S. Minto: “Ink Jet Printing”, IGC Lecture, Carmel, California, USA Dec. 10 – 12, 1978.
140. B. Kazan, US 3 287 734, 1966.
141. P. F. Kurz, L. A. Nash, US 3 715 219, 1969.
142. B. Hermanrud: “The Compound Jet and Its Application”, The Second International Congress on Advances in Non-
Impact Printing Technologies, SPSE, Arlington, Virginia, USA Nov. 4 – 8, 1984, p. 35.
143. F. J. Kamphoefner: “Ink Jet Printing”, IEEE Trans. Electron Devices ED-19 (1972) no. 4, 584.
144. W. L. Buehner et al.: “Application of Ink Jet Technology to a Word Processing Output Printer”, IBM J. Res Dev. 21
(1977) no. 1, 2 – 9.
145. S. A. Curry et al.: “Scale Model of an Ink Jet”, IBM J. Res Dev. 21 (1977) no. 1, 10 – 20.
146. W. T. Pimbley et al.: “Satellite Drop Formation in a Liquid Jet”, IBM J. Res Dev. 21 (1977) no. 1, 21 – 30.
147. T. G. Twardeck et al.: “Effect of Parameter Variations on Drop Placement in an Electrostatic Ink Jet Printer”, IBM J.
Res. Dev. 21 (1977) no. 1, 31 – 36.
148. G. L. Fillmore et al.: “Drop Charging and Deflection in an Electrostatic Ink Jet Printer”, IBM J. Res. Dev. 21 (1977)
no. 1, 37 – 47.
149. H. C. Lee et al.: “Boundary Layer around a Liquid Jet”, IBM J. Res. Dev. 21 (1977) no. 1, 48 – 51.
150. J. M. Carmichael et al.: “Controlling Print Height in an Ink Jet Printer”, IBM J. Res. Dev. 21 (1977) no. 1, 52 – 55.
151. M. Levanoni et al.: “Study of Fluid Flow through Scaled-Up Ink Jet Nozzles”, IBM J. Res. Dev. 21 (1977) no. 1, 56 – 68.
152. S. D. Hawkins, US 4 459 601, 1984.
153. F. C. Lee et al.: “The Application of Drop-on-Demand Ink Jet Technology to Color Printing”, IBM J. Res. Dev. 28 (1984)
no. 3, 307.
154. D. B. Bogy et al.: “Experimental and Theoretical Study of Wave Propagation Phenomena in Drop-on-Demand Ink Jet
Devices”, IBM J. Res. Dev. 28 (1984) no. 3, 314.
155. J. E. Fromm: “Numerical Calculation of the Fluid Dynamics of Drop-on-Demand Jets”, IBM J. Res. Dev. 28 (1984)
no. 3, 322.
156. F. C. Lee: “Pressure-Wave Generation in Impulse Ink-Jet Head Using Elongated Fluid-Filled Piezoelectric Transducer”,
SID Digest, 18 (1987) 186.
157. H. Gerhauser et al.: “The Effect of Pulse Shape on the Drop Volume and the Frequency Response of Drop-on-Demand
Ink Jet Transducer”, SID Digest, 14 (1983) 110.
158. F. C. Lee: “Drop Break-Off Dynamics of Drop-on-Demand Ink Jet”, The Second International Congress on Advances in
Non-Impact Printing Technologies, SPSE, Springfield, USA Nov. 4 – 8, 1984, 9.
159. Hewlett-Packard Journal 36, May, 1985.
160. E. Manini et al.: “The Liquid Spark Ink Jet Technology”, The Third International Congress on Advances in Non-Impact
Printing Technologies, August, 1986, p. 145.
161. G. Olive et al.: “Materials and Processing Studies for Thermal Ink-Jet Devices”, SID Digest 17 (1986) 105.
162. L. Chang, G. Olive: “Factors Influencing the Lifetime of Thermal Ink-Jet Heaters”, SID Digest 18 (1987) 192.
163. A. R. Forouhi, J. Eldridge: “Prediction of Thermal Ink-Jet Lifetime from Chemical Dissolution Rate of Passivation
Layers”, SID Digest 18 (1987) 196.
164. E. Bassous et al.: “Ink Jet Printing Nozzle Arrays Etched in Silicon”, Appl. Phys. Lett. 31 (1977) 135. Links
165. A. R. Lin: “Advances in Ink for Ink Jet Technology”, Annual Guide to Ribbons & Toners, 1985, p. 51.
166. P. Gregory: “High-Technology Applications of Organic Colorants, Topics in Applied Chemistry”, Plenum Press, New
York 1991.
167. Canon, US 4 585 484, 1986 (M. Haruta, T. Nakousai, K. Iwata, K. Nakousai).
168. J. S. Campbell, P. Gregory, P. Mistry, Proceedings of the IS&T's Seventh International Congress on Advances in Non-
Impact Technologies. Springfield, Virginia, IS&T 1991, vol. 2, pp. 142 – 145.

page 77 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
169. US 5 203 912, 1993 (D. Greenwood, N. Hughes, R. W. Kenyon, S. L. Hindagolla).
170. Lexmark, US 5 719 204, 1998 (B. L. Beach, A. M. Pearson, J. X. Sun, R. B. Watkins).
171. Du Pont, US 5 712 338, 1998 (D. H. Donovan, G. A. Lane, C. C. Tang).
172. Hewlett-Packard, US 5 745 140, 1998 (J. L. Stoffel et al.).
173. Hewlett-Packard, US 5 534 051, 1996 (H. P. Lauw).
174. Seiko Epson JP HEISEI4–164975, 1992; Chem. Abstr. 117 (1992) no. 18, 173524m.
175. Hewlett-Packard, US 5 536 306, 1996 (L. E. Johnson et al.).
176. Tektronix, US 5 084 099, 1992 (C. W. Jaeger, C. F. Sheley).
177. Tektronix, US 5 372 852, 1994 (D. R. Titterington, L. V. Bui, L. M. Hirschy, H. R. Frame).
178. J. Maed: “Color Printer Performance Is Improved, But Demand Had Yet to Blossom”, Sanyo Electric Company, OEP,
May, 1985.
179. W. Crooks, E. W. Luttman, A. B. Jaffe: “High Quality Color Printing with Continuous Ink-Jet”, SPSE 1st International
Conference on Advances in Non-Impact Printing Technologies, Venice, Italy June, 1981.
180. C. Rainwater: Light and Color: Golden Press, Western Publishing Company 1971.
181. F. W. Billmeyer, Max Saltzman: Principles of Color Technology, Wiley, New York 1966.
182. D. B. Judd, G. Wyszecki: Color in Business, Science, and Industry, Wiley, New York 1975.
183. E. Coates: “Aggregation of Dyos in Aqueous Solutions”, ISDC, 1969, 355 – 368.
184. IBM Report, P. Stucki, IBM Zurich Research Laboratory: “Mecca—A Multiple Error Correction Compalation Algorithm
for Bi-Level Image Hardcopy Reproduction”, Switzerland.
185. IBM internal report, E. W. Luttman, ISD Advanced Technology: “Color Reproduction with Ink Jet”, San Jose, California.
186. J. Kosar: Light Sensitive Systems, Wiley Interscience, New York – London – Sidney 1965.
187. M. S. Dinaburg: Photosensitive Diazo Compounds, The Focal Press, London – New York 1964.
188. J. Munder: “Plenarvorträge”, 4th Intern. Kongress für Reprografie und Information, Hannover 1975, p. 120; AWV
Fachbericht 349, Frankfurt/Main 1975. Ausschuß für wirtschaftliche Verwaltung in Wirtschaft und öffentlicher Hand.
189. J. Epperlein et al., J. Signalaufzeichnungsmaterialien 11 (1983) no. 6, 403 – 424.
190. H. Zollinger: Azo and Diazo Chemistry, Aliphatic and Aromatic Compounds, Chap. 10, 11, Interscience, New York
1961.
191. O. Frank: Die Lichtpaustechnik, Dorotheen-Verlag, Stuttgart 1959.
192. Océ-van der Grinten, DE 23 25 579, 1973.
193. Adressograph-Multigraph Corp., DE 15 22 449, 1966.
194. R. Dessauer, Chimia 29 (1975) no. 11, 467 – 469.
195. E. Brinkman, G. Delzenne, A. Poot, J. Willems: Unconventional Imaging Processes, The Focal Press, London – New
York 1978, pp. 68 – 73.
196. The Mead Corporation, US 4 399 209, 1983 (F. W. Sanders, G. F. Hillenbrandt); US 4 440 846, 1984 (J. S. Arney, R.
F. Wright) R. F. Wright).
197. E. A. Rott: Photographic Silver Halide Diffusion Transfer Processes, Focal Press London – New York 1972.
198. E. H. Land, H. G. Rogers, V. K. Walworth in J. M. Sturge (ed.): Neblette's Handbook of Photography and Reprography,
7th ed., Van Nostrand, New York – London 1977.
199. L. W. Tregilius et al., US 3 179 517, 1965; R. Tarkinson, Photogramm. Eng. 31 (1965) no. 1, 126 – 130.
200. W. R. Hawken, Copying Methods Manual, LTP Publications no. 11, American Library Association, Chicago 1966.
201. W. Eisbein, US 2 657 618, 1953.
202. G. P. Levenson in T. H. James (ed.): Theory of the Photographic Process, “Diffusion transfer & mono-baths”,
Macmillan Publ. Co., New York 1977.
203. H. Vermeulen, A. Poot, J. W. Van den Houte, US 435 7407, 1982.
204. R. M. De Keyzer EP-A 87 2017 009, 1989.
205. Y. Tsubai, K. Okazaki, S. Horii, A. Yoshida, GB 2 159 968 A, 1985.
206. D. H. Schultze: “Graphic arts applications of silver diffusion transfer systems”, J. Appl. Photogr. Eng. 5 (1979) no. 3,
163.
207. J. Velek et al., Collect. Czech. Chem. Commun. 41 (1976) 1419. Links
208. Agfa-Gevaert, DE-OS 2 654 213, 1977 (E. Meier, W. Laessig).
209. Agfa-Gevaert US 4 232 107, 1980 (W. Janssen).
210. M. Peters, Lecture given on SPSE congress in Boston, MA, USA, on May 20 – 24, 1984. J. Imag. Technol. 11 (1985)
101.
211. G. A. Plunka (ed.): Micrographic Film Technology, The National Micrographics Association, Silver Spring, Maryland,
USA 1983.
212. Kalle, DE 559 795, 1932 (W. Krieger).
213. H. H. Mc Gregor, J. Microgr. 9 (1975) 13 – 20.
General References
214. J. Yule: Principles of Color Reproduction, J. Wiley & Sons, New York 1967.
215. M. H. Bruno: Pocket Pal—A Graphic Arts Production Handbook, International Paper Company, New York 1983.
216. H.-J. Teichmann, et al.: Neue technische Kommunikation in der Druckindustrie, Bundesverband Druck E.V.,
Wiesbaden 1982.

page 78 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Specific References
217. H. Philippaerts, R. Vanesch, E. Hofman, J. Verelst, Dtsch. Drucker 19 (1983) no. 28, 38 – 42.
218. E. Dutoit, Druckspiegel 42 (1987) no. 8, 801 – 802.
219. H. Zwicky, Polygraph 38 (1985) no. 18, 1348, 1350, 1352 – 1354.
220. A. Rott, E. Weyde: Photographic Silver Halide Diffusion Processes, Focal Press, London 1972.
221. J. Verelst, Chimia 34 (1980) no. 4, 190 – 199. Links
222. J. Heller: Printmaking Today, Holt, Rinehart & Winston, New York 1972.
223. W. Stupp, MFM Mod. Fototech. 27 (1979) no. 6, 290 – 292.
224. E. Born: Handbuch der Rasterfotografie, Verlag Ambipress, Basel 1983.
225. in [215] pp. 79 – 80.
226. P. Müller, Druckspiegel 40 (1985) no. 7, 682, 684 –685.
227. E. Fritz, Polygraph 38 (1985) no. 22, 1740, 1742, 1747, 1750, 1752.
228. D. Spencer, Seybold Report on Publishing Systems 16 (1987) no. 11, 3 – 20.
229. E. Seubert, Druckspiegel 39 (1984) no. 2, 152 – 158.
230. J. Van der Schans et al.: Working with Color, Elsevier, Amsterdam 1984.
231. M. H. Bruno: Principles of Color Proofing, Gamma Communications, Salem 1986.
232. H. Zwicky, Fachh. Chemigr. Lithogr. Tiefdruck (1983) no. 4, 183 – 186.
233. N. N., Druckspiegel 39 (1984) no. 5, 486 – 489.
234. J. Birkenshaw: “Trends in Text and Image Processing”, J. Photogr. Sci. 32 (1984) 219 – 220.
235. P. Dyson, J. Seybold, W. Solimeno, D. Stivison, Seybold Report on Publishing Systems 14 (1985) no. 19, 3 – 26.
236. W. Fraß, Chem. Unsere Zeit 17 (1983) 10 – 20, 33 – 40.
237. O. Süs, Justus Liebigs Ann. Chem. 556 (1944) 65 – 84.
238. Kalle AG, DE 1 543 721, 1966 (O. Süs, F. Uhlig, H. Schäfer).
239. Polychrome Corp., DE 1 522 478, 1965; US 3 635 709, 1969 (K. Kobayashi).
240. M. J. Bowden, L. F. Thompson, S. R. Fahrenholtz, E. M. Doerries, J. Electrochem. Soc. 128 (1981) 1304 – 1313. Links
241. BASF, DE 2 150 691, 1973 (H. Barzynski, M.-J. Jun, D. Sänger, O. Volkert).
242. H. Barzynski, D. Sänger, Angew. Makromol. Chem. 93 (1981) 131 – 141. Links
243. American Can Co., DE-OS 2 329 407, 1972 (Sh. I. Schlesinger, V. Cochran).
244. Kalle AG, DE 2 331 377, 1973 (P. Stahlhofen, R. Beutel).
245. Hoechst AG, DE 2 718 259, 1977 (G. Buhr).
246. Fuji Photo Film Co. Ltd., DE 2 851 471, 1977 (M. Iwashaki); DE 2 851 472, 1977 (M. Iwashaki, S. Sato, N. Yasuo).
247. Hoechst AG, DE 2 610 842, 1975 (G. Buhr, H. Ruckert, W. Fraß).
248. Hoechst AG, DE 2 718 254, 1977 (J. Sander, G. Buhr, H. Ruckert).
249. Hoechst AG, EP 0 006 626, 1978 (J. Sander, G. Buhr, H. Ruckert).
250. Hoechst AG, EP 0 006 627, 1978 (J. Sander, D. Skaletz, G. Buhr, G. Lohaus).
251. Minnesota Mining and Manuf. Co., US 3 779 778, 1972 (G. H. Smith, J. A. Bonham).
252. J. Kosar, Light-Sensitive Systems, Chemistry and Application of Nonsilver Halide, Photographic Processes, J.
Wiley & Sons, New York – London – Sydney 1965.
253. B. Duncalf, A. S. Dunn, J. Appl. Polym. Sci. 8 (1964) 1763 – 1776. Links
254. Kalle AG, DE 581 697, 1931 (R. Zahn).
255. Kalle AG, DE 2 024 242, 1969 (H. Steppan).
256. O. Süs, K. Möller, H. Heiss, Liebigs Ann. Chem. 598 (1956) 123 – 138.
257. O. Süs, J. Munder, H. Steppan, Angew. Chem. 74 (1962) 985 – 988.
258. Howson-Algraphy Ltd., DE 2 251 828, 1971 (A. P. Gates).
259. Eastman Kodak Co., GB 695 262, 1953; US 2 725 372, 1955 (L. M. Minsk).
260. Eastman Kodak Co., DE-AS 2 717 778, 1976 (J. A. Vanallen, M. P. Cunningham, D. P. Specht, S. Y. Farid).
261. W. R. Grace u. Co., US 3 661 744, 1972; US 3 697 395, 1972; US 3 697 402, 1972; US 3 700 574, 1972 (C. L. Kehr,
W. R. Wszolek).
262. H. Steppan: “Photopolymerisierbare und photovernetzbare Systeme”, Plenarvorträge auf dem 4. Internationalen
Kongreß für Reprographie und Information 1975, 13. – 17. 04. 1975, Hannover, 147 – 162.
263. Imperial Chemical Industries Ltd., GB 566 795, 1945 (W. E. F. Gates).
264. D. W. Woodward, V. C. Chambers, A. B. Cohen, Photographic Science and Engineering 7 (1963) 360 – 368.
265. DuPont, US 3 261 686, 1966 (J. R. Celeste, S. Bauer).
266. P. Walker, V. J. Webers, G. A. Thommes, J. Photogr. Sci. 18 (1970) 150 – 158.
267. R. M. Schaffert: Electrophotography, Focal Press. London, 5th ed. 1980.
268. J. W. Weigl, Angew. Chem. 89 (1977) 386 – 406.
269. H. J. Gerritsen, W. Ruppel, A. Rose, Helv. Phys. Acta 30 (1957) 504 – 512.
270. K. Hauffe, J. Photogr. Sci. 10 (1962) 321 – 330.
271. R. B. Comizzoli, D. A. Ross, Photogr. Sci. Eng. 13 (1969) 265 – 270.
272. M. R. Kuehnle, J. Appl. Photogr. Eng. 4 (1978) 155 – 159.

page 79 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
273. Kalle & Co. AG, DE 1 058 836, 1956 (W. Neugebauer, M. Tomanek, H. Behmenburg).
274. Xerox Corp., DE 1 497 205, 1969 (J. F. Byrne).
275. W. Wiedemann: “Organische Photoleiter – Ein überblick”, Chem. Ztg. 106 (1982) 3 – 14; Chem. Ztg. 106 (1982) 15 –
29.
276. Winnacker-Küchler, 4th ed., 7, Organische Technologie III, 537 – 562.
277. M. H. Bruno: Principles of Color Proofing, Gama Communications, Salem N. H. 1986.
278. G. Werner: Standardisierter Mehrfarben-Offsetdruck, Polygraph Verlag, Frankfurt/M. 1982.
279. Minnesota Mining and Manuf. Co., US 3 136 637, 1958 (G. W. Larson); DE 1 912 801, 1968 (P. C. van Beusekom).
280. Azoplate Corp., DE 2 032 947, 1969 (H. G. Fernekess, T. N. Gillich).
281. Kalle AG, DE 1 291 197, 1963 (F. Endermann, F. Uhlig).
282. DuPont, US 4 489 153, 1983 (R. W. Ashcraft, J. G. Moehlmann).
283. Minnesota Mining and Manufacturing Co., EP-A 0 115 899, 1983 (P. M. Koelsch, R. I. Krech).
284. American Hoechst Corp., US 4 650 738, 1984 (St. J. W. Platzer, G. I. Koletar).
285. American Hoechst Corp., US 4 659 642, 1984 (St. J. W. Platzer, G. I. Koletar, R. L. Shadrach).
286. DuPont, US 3 649 268, 1969 (V. F. H. Chu, A. B. Cohen).
287. DuPont (Deutschland) GmbH, DE 3 625 014, 1986 (H. Fröhlich).
288. Keuffel and Esser, US 4 376 158, 1977 (D. S. Speckler).
289. DIN 16 544, Begriffe der Reproduktionstechnik im graphischen Gewerbe, Jan. 1966.
290. Jahresbericht 1986 des Bundesverbandes Druck e.V., Wiesbaden, p. 12.
291. Winnacker-Küchler, 4th ed., 7, Org. Technologie III, 524 – 526.
292. Polychrome Corp., US 3 891 516, 1967 (S. L. Chu).
293. Hoechst AG, US 4 666 576, 1985; US 4 661 219, 1985 (E. Pliefke).
294. Vickers Ltd., DE 2 816 307, 1977 (M. Ould, G. N. Stevens).
295. Davidson Manufacturing Corp., US 2 344 510, 1944 (W. T. Hagelin). Polychrome Corp., GB 1 582 620, 1977 (S. L.
Chu, E. Golda, J.-Ch. Huang, A. L. Wilkes).
296. A. W. Brace, P. G. Sheasby: The Technology of Anodizing Aluminium, Technicopy Limited, Stonehouse, Glos. GL 10
2 NA, GB 1968.
297. Hoechst AG, DE 2 811 396, 1978 (G. Usbeck).
298. W. L. Howson Ltd., CA 875 681, 1968 (L. Watkinson, B. Moore).
299. Hoechst AG, DE 2 836 803, 1978 (G. Usbeck).
300. Minnesota Mining and Manufacturing Co., US 2 714 066, 1954 (C. L. Jewett, J. M. Case).
301. Kalle AG, DE 1 621 478, 1967 (G. Berghäuser, F. Uhlig).
302. Polychrome Corp., US 3 440 050, 1965.
303. Hoechst, DE 3 206 470, 1982 (D. Mohr).
304. Minnesota Mining and Manuf. Co., DE 1 091 433, 1954/1955 (C. L. Jewett, J. M. Case, J. F. Dowdall).
305. in [291] pp. 527 – 532.
306. Kalle AG, DE 1 447 963, 1965 (O. Velten, F. Uhlig).
307. Vickers Ltd., GB 1 513 368, 1974 (H. Wielinga).
308. Fuji Photo Film Co. Ltd., DE-OS 2 461 912, 1974 (M. Watanabe); Hoechst, EP 0 131 238, 1983 (P. Stahlhofen, G.
Mack); EP 0 133 216, 1983 (P. Stahlhofen).
309. Toray Industries Inc., US 347 303, 1978 (M. Asano, Y. Mori, Y. Takayama).
310. Toray Industries Inc., US 4 342 820, 1979 (T. Kinashi, T. Fujita, N. Kawabe).
311. H. W. Vollmann, Angew. Chem. 92 (1980) 95 –106.
312. Kalle AG, DE 1 117 391, 1959 (F. Uhlig).
313. Minnesota Mining and Manuf. Co., US 3 563 734, 1971 (B. L. Shely).
314. James River Graphics Inc., US 4 245 003, 1979 (R. L. Oransky, D. G. Savage).
315. BASF, DE 3 226 779, 1982 (J. Lynch, R. Vyvial, M. Zuerger, K. Borho).
316. BASF, DE 3 908 763, 1990 (T. Telser, W. Hümmer, K. Kurtz).
317. BASF, DE 2 202 357, 1990 (C. Srna et al.); DE 1 597 515, 1967 (H. Henkler, H. Hartmann, K. Gulbins, H. Wilhelm).
318. BASF, EP 0 470 071, 1992 (T. Telser et al.). DuPont, DE 3 828 551, 1988 (M. Schober et al.).
319. Asahi, EP 0 017 927, 1984 (S. Nakamura et al.).
320. D. Finna, V. Jansen, R. Michels, Flexo + Tiefdruck 2 (1995).
321. J. Kosar: Light Sensitive Systems, J. Wiley & Sons, New York 1965.
322. H. Barzynski, K. Penzien, O. Volkert, Chem. Ztg. 96 (1972) no. 10, 545 – 51.
323. DuPont, US 2 760 863, 1952 (L. Plambeck).
324. W. E. F. Gates, CA 441 538 (1947).
325. H. J. Timpe, H. Baumann: Photopolymere, Dt. Verlag für Grundstoffindustrie, Leipzig 1988.
326. BASF, DE-OS 2 232 497 1972 (M.-J. Jun, M. Fischer, O. Volkert, E. Hickmann); DE-OS 2 232 498, 1972 (E.
Hickmann); DE-OS 2 909 992, 1979 (P. Lechtken et al.).
327. DuPont, DE 2 215 090, 1981 (G. Chen).

page 80 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

328. Nippon Paint, US 3 801 328 1974 (Y. Takimoto et al.).

329. BASF, DE-OS 1 522 463, 1966 (G. Faulhaber et al.).
330. W. R. Grace, US 3 700 574, 1970 (C. L. Kehr, S. Spring, W. R. Wszolek);
331. Asahi, DE 2 815 678, 1980 (S. Nakamura et al.).
332. BASF, EP 0 027 612, 1982 (G. Heinz, P. Richter, M. Jun); DE 3 803 457, 1988 (K. Kurtz, H. Koch, T. Telser, M.
Zürger).
333. Toyo Boseki, EP 0 273 393, 1988 (A. Tomita et al.).
334. BASF, DE 3 540 950, 1987 (H. Koch, W. Ziegler).
335. Nippon Paint, EP 0 261 910, 1994 (K. Ishikawa, K. Konishi, H. Kusuda).
336. DuPont, EP 0 356 953, 1990 (M. Fryd, E. Leberzammer).
337. BASF, DE-OS 2 722 421, 1977 (B. Bronstert, W. Kuesters, M. Jun, M.-J. Jun, G. Wallbillich).
338. DuPont, US 5 262 275, 1993 (R. N. Fan).
339. BASF, DE-OS 19 53 68 05.3, 1995 (T. Loerzer et al.).
340. R. Golpon: Lehrbuch der Druckindustrie, “Tiefdruck”, vol. 5, Polygraph Verlag, Frankfurt/M. 1974.
341. O. Leidung (ed.): Technical Guide for the Gravure Industry, Gravure Technical Association, Inc., New York, N.Y. 1975.
342. E. D. Stiebner: Bruckmann's Handbuch der Drucktechnik, 4. ed.. Bruckmann, München 1986.
343. F. Bauer: Lexikon der Reproduktionstechnik, Verlag Beruf + Schule, Itzehoe 1986.
344. K.-A. Springstein: Elektronische Bildverarbeitung von A – Z, Verlag Beruf + Schule, Itzehoe 1982.
345. W. Boppel: Die Elektronenstrahlgravur von Hell zur Herstellung von Tiefdruckzylindern, Dtsch. Drucker 16 (1986) 44 –
49.
346. K.-A. Springstein: “Erfolgreicher Praxiseinsatz des ersten ‘Lasergravure-700'-Systems in England.” Dtsch. Drucker 25
(1984) 48 – 60.
347. J. V. Duppen: Handbuch für den Siebdruck, Verlag Der Siebdruck, D-2400 Lübeck 1986.
348. O. Zeman, IS + L, Fachzeitschrift für industriellen Siebdruck + Leiterplattentechnik 5 (1984) 24 – 31.
349. Agency of Industrial Science and Technology, MITI, DE 2 932 376, 1978 (Ichimura, Konihiro).
350. see [347-349]
351. DuPont, EP-A 4383, 1979 (G. R. Nacci, D. G. Pye).
352. W. S. De Forest: Photoresist, Materials and Processes, McGraw-Hill, New York 1975.
353. H. Steppan, G. Buhr, H. Vollmann, Angew. Chem. 94 (1982) 471 – 485.
354. C. G. Roffey: Photopolymerization of Surface Coatings, Chap. 6, John Wiley & Sons, Chichester, Reprint 1985,
pp. 263 – 304.
355. S. J. Moss, Polym. Degrad. Stab. 17 (1987) 205 – 222.
356. G. N. Taylor, T. M. Wolf, L. Stillwagon, Solid State Technol. (1984, Feb.) 145 – 155.
357. N. G. Einspruch, G. B. Larrabee (eds.): “Materials and Process Characterization”, VLSI Electronics Microstructure
Science, vol. 6, "Chap. 4", Academic Press, New York 1983, pp. 217 – 253.
358. M. Tsuda, J. Imaging Technol. 11 (1985) 158 – 163.
359. F. Coopmans, B. Roland, Solid State Technol. (1987, June) 93 – 99.
360. R. Srinivasan, Polym. Degrad. Stab. 17 (1987) 193 – 203.
361. L. F. Thompson, C. G. Willson, J. M. J. Fréchet (eds.): “Materials for Microlithography-Radiation-Sensitive Polymers”,
ACS Symp. Ser. 266 (1984) 39 – 117.
362. L. F. Thompson, C. G. Willson, M. J. Bowden (eds.): “Introduction to Microlithography—Theory, Materials, and
Processing”, ACS Symp. Ser. 219 (1983) 87 – 159.
363. J. Bargon (ed.): Methods and Materials in Microelectronic Technology, Plenum Press, New York 1984, pp. 181 – 241.
364. E. R. Howells (ed.): Technology of Chemicals and Materials for Electronics, Chap. 8, Ellis Horwood Ltd., Chichester
1984, pp. 128 – 148.
365. Grace, GB 2 091 493, 1982 (F. J. Rendulic, R. K. Trasavage, P. A. Boduch).
366. R. H. Wopschall, Solid State Technol. (1986, June) 153 – 157.
367. J. E. Jensen, Solid State Technol. (1984, June) 145 – 150.
368. S. T. Kowel et al., Thin Solid Films 152 (1987) 377 – 403.
369. M. Hummel: Einführung in die Leiterplattentechnologie, E. G. Leuze Verlag, Saulgau/Württ. 1985.
370. G. W. W. Stevens, J. Appl. Photogr. Eng. 8 (1982) p. 90 – 100.
371. “Galvanoformung, Berichtsband über das 8. Ulmer Gespräch der Deutschen Gesellschaft für Galvano- und
Oberflächentechnik”, 24./25. April 1986 in Neu-Ulm, E. G. Leuze Verlag, Saulgau/Württ., 1986.
372. W. Becht, W. Ehrfeld, A. Maner, D. Schmidt, Galvanotechnik 77 (1986) 2695 – 2705.
373. D. J. Elliott: Integrated Circuit Fabrication Technology, McGraw-Hill, New York 1982.
374. L. K. White, RCA Rev. 47 (1986) 345 – 379.
375. W. R. Brunsvold, D. M. Crockatt, G. J. Hefferon, C. F. Lyons, Opt. Eng. 26 (1987) 330–336. Links
376. J. B. Angell, S. C. Terry, P. W. Barth, Sci. Am. 248 (1983) no. 4, 36 – 47. Links
377. M. Hepher, J. Photogr. Sci. 12 (1964) 181 – 190.
378. R. Phillips: Sources and Applications of Ultraviolet Radiation, Academic Press, London 1983, pp. 33 – 37, 56 – 111.
379. W. Schnabel, H. Sotobayashi, Prog. Polym. Sci. 9 (1983) 297 – 365. Links

page 81 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

380. T. Davidson (ed.): “Polymers in Electronics”, ACS Symp. Ser. 242 (1984) Beiträge 1 – 21.
381. Landolt-Börnstein: Zahlenwerte und Funktionen aus Naturwissenschaften und Technik, Neue Serie, Gruppe III:
“Kristall- und Festkörperphysik”, vol. 17, Teilband c, Springer-Verlag, Berlin – Heidelberg pp. 250 – 280.
382. H. Meier, K.-P. Zeller, Angew. Chem. 87 (1975) 52 – 63.
383. O. Süs: Justus Liebigs. Ann. Chem. 556 (1944) 65 – 84.
384. J. Pacansky, J. R. Lyerla, IBM J. Res. Dev. 23 (1979) 42 – 55. Links
385. in [362] pp. 122 ff.
386. A. S. Gozdz, Solid State Technol. (1987, June) 75 – 79.
387. IBM, US 4 491 628, 1982 (H. Ito, C. G. Willson, J. M. J. Fréchet).
388. S. P. Pappas, J. Imaging Technol. 11 (1985) 146 – 157.
389. M. Hashimoto, T. Iwayanagi, H. Shiraishi, S. Nonogaki, Polym. Eng. Sci. 26 (1986) 1090 – 1095. Links
390. W. L. Dilling, Chem. Rev. 83 (1983) 1–47. Links
391. Kodak, US 1 973 493, 1932 (C. J. Malm, C. R. Fordyce).
392. Kodak, GB 695 197, 1949 (L. M. Minsk, W. P. Vandeusen).
393. A. Reiser, P. L. Egerton, Photogr. Sci. Eng. 23 (1979) 144 – 150.
394. Ciba-Geigy, DE 2 342 407, 1973 (S. A. C. Zahir, H. Rembold, E. Losert).
395. D. W. Woodward, V. C. Chambers, A. B. Cohen, Photogr. Sci. Eng. 7 (1963) 360; J. Kosar: Light-Sensitive Systems:
Chemistry and Application of Nonsilver Halide Photographic Processes, Wiley, New York 1965 – 1970, "Chap. 3, 4 and
5"; P. Walker, V. J. Webers, G. A. Thommes, J. Photogr. Sci. 18 (1970) 150; H. Barzynski, K. Penzien, O. Volkert,
Chem.-Ztg. 96 (1972) 545; H. Steppan, 4. Ing. Kongr. Reprogr. Inf. Hannover 1975, p. 147; A. Ledwith, Pure Appl.
Chem. 49 (1977) 431; Links G. A. Delzenne, Adv. Photochem. 11 (1979) 1; Makromol. Chem. Suppl. 2 (1979) 169; J.
L. R. Williams, R. C. Daly, S. Y. Farid, Euchem. 80, Conf. Photopolymer. Photocrosslink. Org. Coat., Stockholm 1980;
H. Böttcher, J. Signalaufzeichnungsmaterialien 8 (1980) 405.
396. in [378] pp. 56 – 111.
397. H. J. Hageman, Prog. Org. Coat. 13 (1985) 123 – 150.
398. DuPont, BE 685 011, 1966 (J. R. Celeste).
399. Dynachem Corp., US 4 610 951, 1985 (M. A. Lipson, G. A. Derrico, S. Y. Tark, T. Yamazaki).
400. Fuji Photo Film Co., DE-OS 3 601 997, 1986 (K. Maemoto et al.).
401. R. Rubner, H. Ahne, E. Kühn, G. Kolodziej, Photogr. Sci. Eng. 23 (1979) 303; H. J. Boos, G. Kolodziej, R. Rubner,
Galvanotechnik 68 (1977) 984; Siemens, DAS 2 462 105, 1979 (R. Rubner, W. Kleeberg, E. Kühn).
402. G. A. Delzenne, Adv. Photochem. 11 (1979) 1; C. L. Kehr, W. R. Wszolek, Am. Chem. Soc. Div. Org. Coat. Plast.
Chem. Pap. 33 (1973) 295. C. R. Morgan, A. D. Ketley, ibid. 33 (1973) 281.
403. in [373] Chapter 13, pp. 311 – 337.
404. R. H. Wopschall: Proceedings of the Technical Program National Electronic Packaging and Production Conference,
Anaheim, California March 1 – 3, 1983, pp. 11 – 15.
405. C. Tiby, Chem. Ind. (Düsseldorf) 109 (1986) 1044 – 1050.
406. in [354] Chap. 7, pp. 305 – 338.
407. S. G. Wentink, S. D. Koch (eds.): UV Curing in Screen Printing for Printed Circuits and the Graphic Arts, Chap. 9,
Technology Marketing Corp. Norwalk, Conn. USA 1981, pp. 221 – 242.
408. in [378] pp. 406 – 425.
409. J. C. Knaff, W. Liebsch, Galvanotechnik 75 (1984) 115 – 120.
410. J. Eason, R. DeBisschop, Galvanotechnik 75 (1984) 936 – 940.
411. U. v. Mylius, J. Leudolph, Galvanotechnik 77 (1986) 402 – 406.
412. H. Dengler, W. Röcker, Galvanotechnik 77 (1986) 2237 – 2240.
413. K. G. Faas, J. Swozil (eds.): Verdrahtungen und Verbindungen in der Nachrichtentechnik, Akade-mische
Verlagsgesellschaft Frankfurt am Main 1974, pp. 196 – 250.
414. J. A. Scarlett (ed.): The Multilayer Printed Circuit Handbook, Electrical Publications Ltd. 1985; cf. Galvanotechnik 77
(1976) 2196.
415. in [413] pp. 66 – 195.
416. J. A. Scarlett: Printed Circuit Boards for Microelectronics, 2nd ed., Chap. 2, Electrochemical Publications Ltd., Ayr
(Scotland) 1980, pp. 19 – 80.
417. H. Müller: Konstruktive Gestaltung und Fertigung in der Elektronik, vol. 1, Chap. 3.2, “Bildübertragung”, pp. 46 – 67,
Chap. 3.3, “Verfahren zur Metallstrukturierung”, pp. 68 – 96, Vieweg & Sohn, Braunschweig/Wiesbaden 1981.
418. J. A. Scarlett: An Introduction to Printed Circuit Technology, Electrochemical Publications Ltd., Ayr (Scotland) 1984.
419. G. Herrmann et al.: Handbuch der Leiterplattentechnik, “Laminate – Manufacturing – Assembly – Test, “2nd ed.,
Chap. 1 – 23, E. G. Leuze Verlag, Saulgau (Württ.) 1982.
420. R. H. Clark: Handbook of Printed Circuit Manufacturing, Van Nostrand, New York 1985.
421. in [419] Chap. 15, pp. 274 – 283.
422. A. Gemmler, J. L. Jostan, Galvanotechnik 77 (1986) 51 – 60.
423. P. Lund: Präzisionsvorlagen für Leiterplatten, Hanser Verlag, München – Wien 1979.
424. in [417] pp. 20 – 46.
425. S. A. Partridge, Circuit World 13 (1987) no. 2, 4 – 13.
426 in [419] Chap 6 pp 90 – 124 Chap 7 pp 125 – 151
page 82 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

427. in [407] Chap. 6, pp. 133 – 156.

428. W. Peters, Galvanotechnik 74 (1983) 1395 – 99.
429. C. W. Ruff, Metalloberfläche 41 (1987) 135 – 139.
430. P. Britton, Circuit World 13 (1986) no. 1, 4 – 7.
431. J. Tuck, Circuits Manufacturing (1987) 24 – 33.
432. A. Gemmler, J. L. Jostan, Galvanotechnik 77 (1986) 1182 – 88.
433. R. Germany, Circuit World 13 (1987) no. 3, 68 – 71.
434. in [419] Chap. 29, pp. 535 – 54.
435. C. Mead, L. Conway: Introduction to VLSI Systems, Addison-Wesley Publishing Reading 1980.
436. S. Wolf, R. Tumbler: Silicon Processing for the VLSI Era, Process Technology, Lattice Press, Sunset Beach 1987.
437. R. Muller, T. Kamins: Device Electronics for Integrated Circuits, J. Wiley, New York 1977.
438. W. S. DeForest: Photoresist Materials and Processes, McGraw Hill, New York 1975.
439. D. J. Elliott: Integrated Circuit Fabrication Technology, McGraw Hill, New York 1982.
440. L. F. Thompson, C. Willson, M. Bowden: “Introduction to Microlithography. Theory, Materials and Processing” ACS
Symp. Ser. 219 (1983) 16.
441. L. White, D. Meyerhofer, RCA Rev. 44 (1983) 110.
442. B. Daniels, C. Szmanda, M. Templeton, P. Trefonas, Proc. SPIE Int. Soc. Opt. Eng. 631 (1986) 192.
443. J. Shaw. M. Frisch, F. Dill, IBM J. Res. Dev. 26 (1976) 220.
444. V. Marriott, C. Garza, M. Spak, Proc. SPIE Int. Soc. Opt. Eng. 771 (1987) 221.
445. A. Broers, Solid State Technol. 28 (1985) no. 6, 119. Links
446. K. Bartlett et al., Proc. SPIE Int. Soc. Opt. Eng. 394 (1983) 49.
447. P. West, B. Griffing, Proc. SPIE Int. Soc. Opt. Eng. 394 (1983) 33.
448. F. A. Vollenbroeck et al., Microcircuit Eng., 245 (1985).
449. D. M. Allen: The Principles and Practice of Photochemical Machining and Photoetching, Adam Hilger, Bristol 1986.
450. D. M. Allen, P. J. Gillbanks, Proc. Tech. Program Nat. Electron. Packag. Prod. Conf. 1986, 487 – 494.
451. D. M. Allen, Proc. Tech. Program Nat. Electron. Packag. Prod. Conf. 1987, 583 – 589.
452. E. Makino, T. Sato, Y. Yamada, Precis. Eng. 9 (1987) 153 – 157.
453. P. M. Wible, Photo Chem. Mach. Inst. J. 6 (1981) 7 – 11.
454. R. Markle, Photo Chem. Mach. Inst. J. 23 (1986) 10 – 12.
455. D. M. Allen, J. R. C. Brightman, P. J. Gillbanks, P. M. Smith, Annals of the CIRP 36/1 (1987) 91 –94.
456. D. M. Allen, P. J. Gillbanks, A. J. Palmer, Proc. Int. Sym. Electric. Mach., ISEM-9, (1989) 246 –249.
457. Photofabrication Guide Book published by the Japan Photofabrication Association, Sapporo 1989.
[Top of Page]

Related Articles
Photography is a separate keyword.

[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 83 of 83
 Imidazole and Derivatives 1

Imidazole and Derivatives

Klaus Ebel, BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of Germany (Chaps. 2– 6)
Hermann Koehler, BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of Germany (Chaps. 2– 6)
Armin O. Gamer, BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of Germany (Section 7.1)
Rudolf Jäckh, BASF Aktiengesellschaft, Ludwigshafen, Federal Republic of Germany (Section 7.2)

1. Introduction . . . . . . . . . . . . . . . . . 1 6. Uses . . . . . . . . . . . . . . . . . . . . . . 4
2. Physical Properties . . . . . . . . . . . . 1 6.1. Intermediates for Herbicides and
3. Chemical Properties . . . . . . . . . . . 1 Pharmaceuticals . . . . . . . . . . . . . . 4
4. Production . . . . . . . . . . . . . . . . . . 2 6.2. Other Uses . . . . . . . . . . . . . . . . . . 4
4.1. The Radziszewski Reaction . . . . . . . 2
7. Toxicology and Occupational Health . 4
4.2. Dehydrogenation of ∆2 -Imidazolines 2
7.1. Imidazoles without Nitro Substituents 4
4.3. Functionalized Imidazoles by Modifi-
cation of Simple Imidazoles . . . . . . . 3 7.2. Nitroimidazoles . . . . . . . . . . . . . . . 7
5. Quality Specifications and Analysis . 4 8. References . . . . . . . . . . . . . . . . . . 8

1. Introduction Table 1. Physical properties of imidazole

Imidazole [288-32-4], C3 H4 N2 , M r 68.10, was Flash point 154 ◦ C

first synthesized in 1858 by Debus from ammo- d 110
4 1.0257
Bulk density 0.55 – 0.63 g/cm3
nia and glyoxal; it was originally named gly- Viscosity (100 ◦ C) 2.696 mPa · s
oxalin. The name imidazole was introduced by Solubility in water∗ 241
Hantzsch. Industrial production of imidazole Enthalpy of fusion 12.96 kJ/mol
began in the 1950s; a wide range of derivatives Enthalpy of vaporization (220 – 265 ◦ C) 57.12 kJ/mol

is now available in industrial quantities. ∗ In grams per 100 g of solvent at 20 ◦ C.

2. Physical Properties
Some important physical properties of imida-
zole are listed in Table 1. Molecular masses,
boiling points, melting points, and CAS registry
numbers of the most important imidazole deriva-
tives are listed in Table 2. Substitution at nitro-
gen generally causes a reduction in the melting
and boiling points, because the N−H group is
no longer available for intermolecular hydrogen
bonding. Distillation of nitroimidazoles is ac-
companied by the risk of spontaneous decom-
position.
3. Chemical Properties
Imidazole is a moderately strong base (pK b = 7.0
[1]), and a weak acid (pK a = 14.9 [1]). Imi-
dazoles substituted with electron-withdrawing
groups are stronger acids than imidazole itself;
e.g., 4(5)-nitroimidazole has a pK a of 9.3 [1].
Imidazole is stable at 400 ◦ C, possesses con-
siderable aromatic character, and undergoes the
usual electrophilic aromatic substitution reac-
tions. Nitration and sulfonation require, how-
ever, far more drastic conditions than the corre-
sponding reactions with benzene. Other substi-
tution reactions of imidazole include halogena-
tion, hydroxymethylation, coupling with aro-
matic diazonium salts, and carboxylation. Cat-
alytic hydrogenation has not been described [1],
[2].

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a13 661
2 Imidazole and Derivatives
Table 2. Physical properties of imidazoles

CAS registry no. Mr mp, ◦ C bp, ◦ C

Imidazole [288-32-4] 68.1 90 256

1-Methylimidazole [616-47-7] 82.1 −1 198.5
2-Methylimidazole [693-98-1] 82.1 145 270
4(5)-Methylimidazole [822-36-6] 82.1 46 271
2-Ethylimidazole [1072-62-4] 96.1 77 268
2-Isopropylimidazole [36947-68-9] 110.1 134 278
2-Ethyl-4(5)-methylimidazole [931-36-2] 110.1 45 260
2-Phenylimidazole [670-96-2] 144.2 145 336
1-Vinylimidazole [1072-63-5] 94.1 80∗
4(5)-Methyl-5(4)-hydroxymethylimidazole [29636-87-1] 112.1 138
4(5)-Nitroimidazole [3034-38-6] 113.1 305∗∗
2-Methyl-4(5)-nitroimidazole [696-23-1] 127.1 244∗∗
2-Isopropyl-4(5)-nitroimidazole [13373-32-5] 155.2 184
1-Methyl-5-nitroimidazole [3034-42-2] 127.1 59
1,2-Dimethyl-5-nitroimidazole [551-92-8] 141.1 139
1-Methyl-2-isopropyl-5-nitroimidazole [14885-29-1] 169.2 56
1-(2-Hydroxyethyl)-2-methyl-5-nitroimidazole [443-48-1] 171.2 155

∗ At 1.3 kPa.
∗∗ Decomposes.

4. Production be prepared from the starting materials listed in

Of the many known methods for producing imi-
dazoles, only the following are of industrial im-
portance.
4.1. The Radziszewski Reaction
In the generally applicable Radziszewski reac-
tion, a 1,2-dicarbonyl compound is condensed
with an aldehyde and ammonia (R1 = H) in a
molar ratio of 1 : 1 : 2, respectively. Replace-
ment of a molar equivalent of ammonia with a
primary amine (R1 = alkyl or aryl) leads to the
corresponding 1-substituted imidazoles.
The reaction is usually carried out in water or
a water – alcohol mixture at 50 – 100 ◦ C. Work-
up may involve the usual processes (e.g., distil-
lation, extraction, and crystallization). Distilla-
tion leads to imidazole with a purity 99 %. The
yield is generally 60 – 85 % [3]. This process al-
lows the most important simple imidazoles to
Table 3.
4.2. Dehydrogenation of
∆2 -Imidazolines
∆2 - Imidazolines can be obtained by several
routes from 1,2-diamino compounds and car-
boxylic acid derivatives:
1) by reaction of a diamine with a carboxylic
acid over an acidic heterogeneous catalyst
(e.g., alumina – phosphoric acid) in the gas
phase [4];
2) by reaction of a diamine with a carboxylic
acid nitrile in the presence of sulfur [5] or
copper salts [6] in the liquid phase; or
3) by preparation of diformyl derivatives from a
diamine and formic acid esters, followed by
conversion to imidazoline in the gas phase at
200 – 350 ◦ C over a heterogeneous catalyst
(e.g., zinc oxide – alumina) [7].
The ∆2 -imidazoline is then dehydrogenated
in the gas phase over a precious metal at ca.
300 ◦ C [8] or on alumina – zinc oxide at 400 –
500 ◦ C [9]. In some cases this synthesis gives
even better yields than the Radziszewski reac-
tion (e.g., for 2-aryl-substituted imidazoles).
 Imidazole and Derivatives 3
Table 3. Compounds used to prepare simple imidazoles by the Radziszewski reaction

Imidazole Dicarbonyl compound Aldehyde Amine

Imidazole glyoxal formaldehyde ammonia

1-Methylimidazole glyoxal formaldehyde ammonia, methylamine
2-Methylimidazole glyoxal acetaldehyde ammonia
2-Ethylimidazole glyoxal propionaldehyde ammonia
2-Isopropylimidazole glyoxal isobutyraldehyde ammonia
4(5)-Methylimidazole methylglyoxal formaldehyde ammonia
2-Ethyl-4(5)-methylimidazole methylglyoxal propionaldehyde ammonia

1-Vinylimidazole is produced by base-

4.3. Functionalized Imidazoles by
Modification of Simple Imidazoles
Nitroimidazoles. Reaction of imidazole
with 60 – 80 % nitric acid in the presence of
concentrated sulfuric acid at 120 – 150 ◦ C leads
to 4(5)-nitroimidazoles [10], [11]; nitration at
C-2 has not been observed. 4(5)-Nitroimidazole,
2-methyl-4(5)-nitroimidazole, and 2-isopropyl-
4(5)-nitroimidazole have been prepared by this
method.
catalyzed addition of acetylene to imidazole in
the liquid phase in an autoclave [16].
1-(2-Hydroxyethyl)-2-methyl-5-nitroimida-
zole ( metronidazole) is obtained by addition of
ethylene oxide to 2-methyl-4(5)-nitroimidazole
in formic acid at 25 – 40 ◦ C [17].
Michael addition of imidazole to acry-
lonitrile followed by catalytic hydrogenation
of the intermediate 1-(2-cyanoethyl)imidazole
[23996-53-4] leads to 1-(3-aminopropyl)imida-
zole [5036-48-6] [18].
Hydroxymethylimidazoles. 4(5)-Methyl-
5(4)-hydroxymethylimidazole can be obtained
by treatment of 4(5)-methylimidazole with
aqueous formaldehyde under acidic [19] or basic
[20] conditions.
Quaternary Salts. Treatment of 1-alkylimi-
dazoles with benzyl chloride at 100 ◦ C in
solvent-free melt results in formation of 3-
benzyl-1-alkylimidazolium chloride [21].

1-Substituted Imidazoles. Imidazole and

alkylimidazoles can be alkylated with alcohols 5. Quality Specifications and
in the gas phase at 300 ◦ C over acidic catalysts Analysis
such as alumina – phosphoric acid or silica –
phosphoric acid [12], [13]. Simple imidazoles are generally available in
In the liquid phase, imidazoles may be alky- 99 % purity, which is particularly important for
lated with dialkyl sulfates or alkyl halides in po- their use in pharmaceutical production. Purity
lar solvents such as alcohols and in the presence is usually determined by reverse-phase HPLC;
of a base, e.g., alkali-metal hydroxides or alkox- volatile derivatives may be analyzed by capillary
ides [14]. column gas chromatography. A typical quality
Of particular importance for the prepara- specification for imidazole is shown below:
tion of 1-alkyl-5-nitroimidazoles is the treat- Imidazole (min.) 99.5 %
ment of 4(5)-nitroimidazoles with dialkyl sul- 2-Methylimidazole (max.) 0.2 %
fates in formic acid as solvent [15]. This method Water (max.) 0.2 %
has been used for 1-methyl-, 1,2-dimethyl-, and Other compounds (max.) 0.1 %
1-methyl-2-isopropyl-5-nitroimidazoles.
4 Imidazole and Derivatives
6. Uses
6.1. Intermediates for Herbicides and
Pharmaceuticals
The imidazole ring is a constituent of several im-
portant natural products, including purine, his-
tamine, histidine, and nucleic acids. Therefore,
the bulk of imidazole produced is used in the
preparation of biologically active compounds.
1-Substituted imidazoles such as N-propyl-
N-[2-(2,4,6-trichlorophenoxy)ethyl]imidazo-
le-1-carboxamide (prochloraz) [67747-09-5],
1-[2-(2,4-dichlorophenyl)-2-(2-propenyl-
oxy)ethyl]-1H-imidazole ( imazalil)
[35554-44-0], and 1-(4-chlorophenoxy)-1-
(1H-imidazolyl-1-yl)-3,3-dimethylbutanone (
climbazole) [38083-17-9] are used as fungi-
cides in crop protection [22]. Clotrimazole
[23593-75-1] and bifonazole [60628-96-8] are
used as antimycotics (→ Antimycotics).
4(5)-Methylimidazole and 4(5)-methyl-5(4)-
hydroxymethylimidazole are used as intermedi-
ates for the H2 -blocker cimetidine [51481-61-9]
(→ Antiulcer Drugs).
Nitroimidazoles are important chemother-
apeutic agents [23]: e.g., 1,2-dimethyl-5-
nitroimidazole ( dimetridazole), 1-methyl-2-
isopropyl-5-nitroimidazole ( Ipronidazole), and
1-methyl-2-carbamoyloxymethyl-5-nitroimida-
zole ( ronidazole) [7681-76-7] are used as
veterinary products. Human chemotherapeu-
tics based on 4(5)-nitroimidazole or 2-methyl-
4(5)-nitroimidazole include metronidazole
[443-48-1], tinidazole [19387-91-8], nimora-
zole [6506-37-2], ornidazole [16773-42-5],
secnidazole [3366-95-8], and carnidazole
[42116-76-7] (→ Chemotherapeutics).
Quaternary imidazolium derivatives possess
algicidal and bactericidal properties [21].
6.2. Other Uses
1-Vinylimidazole is employed as a copolymer in
the production of cationic polymers for various
uses. Alkylimidazoles are used as hardeners for
epoxy resins and for polyurethanes [24]. Less
important uses for alkyl- and arylimidazoles in-
clude photography and dyes.
7. Toxicology and Occupational
Health
7.1. Imidazoles without Nitro
Substituents
The prominent feature in imidazole toxicity is
the corrosive nature of the compounds. This
property is eliminated by 2-phenyl-substitution.
In terms of the labelling directives of the EEC
the acute systemic toxicity ranges from toxic (T)
to harmful to health (Xn ).
Acute poisoning results in excitement of the
central nervous system and convulsions; there-
fore, imidazole has been tested as an antide-
pressant [25]. Interaction of imidazole with the
dopaminergic and noradrenergic systems in the
brain was postulated [25]. Repeated application
of some of the compounds revealed no spe-
cific toxic effects. The liver appears to be af-
fected after treatment with high doses. None of
the available in vitro short-term tests revealed a
mutagenic potential for nonnitroimidazole com-
pounds [26–29].
Imidazole. Acute toxicity data (see Table 4)
show only slight toxic effects on animals ex-
posed via oral or parenteral routes. The main
symptoms of intoxication at high doses are
seizures, tremors, loss of balance, opisthotonus
(spasm of muscles in the back), and restlessness.
Single oral exposure of pregnant rats to a dose
of 240 mg/kg of imidazoles on the 12th or 13th
day of gestation produced no harmful effect in
the progeny [30]. Topically, imidazole is corro-
sive to the skin and eye, resulting in necrosis and
irreversible damage to these organs (Table 4).
Repeated administration to rats for 28 days by
gavage (through a stomach tube) revealed no ob-
servable effect at a level up to 62.5 mg kg−1 d−1 .
Higher doses up to 500 mg/kg led to swelling of
the liver and other unspecific toxic effects with-
out mortality. The alteration of the liver may be
due to the enhancement of metabolic activity in
this organ [26].
1-Alkyl-Substituted Imidazoles. These
compounds become more noxious with increas-
ing length of the side chain: 1-n-butylimidazole
is fairly toxic, regardless of exposure route. Fur-
ther increase in chain length reduces the toxicity
to that of the parent compound (see Table 4).
 Imidazole and Derivatives 5
Table 4. Toxicity of imidazoles containing saturated substituents

Substance Acute toxicity, LD50 , mg/kg Effect on skin References

and eye

Imidazole Rat, oral 970 Corrosive [25], [26],

Mouse, oral 880 [31], [35]
Mouse, oral (male) 1880
Rat, s.c. 626
Mouse, s.c. 560; 817
Rat, i.p. 620
Mouse, i.p. 520
Mouse, i.v. 475
1-Methylimidazole Rat, oral 1139 Corrosive [26], [31]
Mouse, oral (male) 1400
Rabbit, dermal 400 < 640
Mouse, i.p. 445
1-Ethylimidazole Rat, oral 900 ∗ Corrosive [26]
Mouse, i.p. 300 ∗
1-Propylimidazole Rat, oral 400 ∗ Corrosive [26]
Rabbit, dermal 200 ∗
Mouse, i.p. 90 ∗
1-n-Butylimidazole Rat, oral 80 Strongly irritant [26]
Rabbit, dermal <200
Rat, dermal 50 <200
Rat, inhalation 4 h (aerosol) 0.11 mg/L
Mouse, i.p. 40
1-Decylimidazole Rat, oral 470 Corrosive [26]
Mouse, i.p. 150
1-Dodecylimidazole Rat, oral 1100∗ Corrosive [26]
Rabbit, dermal 400
Mouse, i.p. 125∗
2-Methylimidazole Rat, oral 1300 Strongly irritant [26], [31]
Mouse, oral (male) 1400 to corrosive
Rat, dermal ca. 2000
Rabbit, dermal 200
Mouse, i.p. 350
2-Ethylimidazole Rat, oral 1400 Corrosive [26]
Rabbit, dermal 200
Mouse, i.p. 400
2-Isopropylimidazole Rat, oral >1000 <2000 Irritant [26]
Rabbit, dermal 400
Mouse, i.p.
50 <200
4-Methylimidazole Rat, oral 350 Strongly irritant [26], [31],
[36]
Mouse, oral (male) 370 to corrosive
Rabbit, dermal 440
Mouse, i.p. (male) 165
1,2-Dimethylimidazole Rat, oral 1300 Strongly irritant [26]
Rabbit, dermal 200 to corrosive
Rat, inhalation 4 h (aerosol
of 50 % solution)
3 mg/L
Mouse, i.p. 300
2-Ethyl-4(5)-methylimidazole Rat, oral 731; 1000 Strongly irritant [26]
Rabbit, dermal 400 to corrosive
Mouse, i.p. 200
1-Vinylimidazole Rat, oral 1100 Strongly irritant to corrosive [26]
Mouse, s.c. 650

∗ LD50 , µL/kg.
6 Imidazole and Derivatives
2-Methylimidazole behaves like 1-
methylimidazole and imidazole (Table 4). Its
irritant properties seem to be slightly less than
those of the other two compounds but are nev-
ertheless pronounced.
A 28-day gavage study in rats revealed at the
lowest dose of 100 mg kg−1 d−1 only a slight de-
crease in blood protein content. Higher doses (up
to 800 mg/kg) led to unspecific clinical symp-
toms as well as changes in clinicochemical pa-
rameters and organ weights without mortality
[26].
4-Methylimidazole is the most toxic of the
methyl-substituted imidazoles and is thought
to be the toxic agent in ammoniated molasses
and hay, which may be responsible for postna-
tal losses in ruminants fed with these products
[31], [32]. It is fairly toxic via the oral and der-
mal routes. 4-Methylimidazole is the most po-
tent convulsive within this group of compounds.
In addition to the data in Table 4, further stud-
ies in rabbits [31] and calves [32] demonstrate
that single oral doses of 120 – 150 or 200 mg/kg
(rabbit) and 400 mg/kg (calf) lead to convul-
sion and death. No clinical, clinicochemical, or
pathological effects were observed in the study
with calves after four repeated doses of 25 or
50 mg/kg within 96 h. At a level of 100 mg/kg,
mild behavioral changes appeared.
Intraperitoneal dosage of 43 mg/kg to rats for
up to 10 weeks led to enlargement of the liver in
some animals, but with no effect on the blood or
clinicochemical parameters [33].
1-Vinylimidazole. Resorption of 1-
vinylimidazole through the skin in fatal amounts
is possible in a relatively short exposure time.
The symptoms of intoxication differ with dose
and species but excitation and convulsion are a
common feature at high doses.
After repeated application (rat, rabbit, cat,
oral) for up to 7 weeks, unspecific clinical
symptoms (weight loss, changes in blood count)
and pathological findings (sporadic incidence of
liver damage) were reported. The level of no ob-
servable effect was between 10 and 100 µL/kg
[26].
No toxic effects were observed (mouse, rat,
rabbit, cat) after repeated inhalation (five 6-h
exposures) of an atmosphere saturated with
vinylimidazole vapor at room temperature [26],
[34].
N-Methyl-2-vinylimidazole and N-butyl-2-
vinylimidazole showed a sensitizing effect on
the skin (guinea pig) in addition to the corrosive
action [34].
Phenylimidazoles. 2-Phenyl-substituted
imidazoles display reduced irritancy but no
change in acute systemic toxicity compared
with imidazole. The acute oral LD50 for 2-
phenylimidazole [26], 1-methyl-2-phenylimi-
dazole [26], and 1-vinyl-2-phenylimidazole [26]
is ca. 1000 mg/kg; the LD50 (mouse, i.p.) is ca.
200 mg/kg.
The symptoms of intoxication indicate de-
pression of the central nervous system by 2-
phenylimidazole and a convulsive action for the
other two compounds.
Inhalation Hazard Tests. With the excep-
tion of imidazole, 2-isopropylimidazole, and 1-
vinylimidazole, all of the compounds listed in
Table 4 have also been subjected to seven- or
eight-hour practice-oriented inhalation hazard
tests in which rats were exposed to an atmo-
sphere highly enriched in or saturated with the
vapor or the volatile components of the test sub-
stance [26]. Of the compounds tested, only 1-n-
butylimidazole resulted in mortality; three out
of six rats died after exposure for 8 h [26].
7.2. Nitroimidazoles
Nitroimidazoles are a well-established group of
antiprotozoal and bactericidal agents. Due to
their aromatic nitro group, they may undergo
metabolization to give electrophilic intermedi-
ates under certain conditions [37], [38] and thus
entail bacteriotoxic, but also mutagenic and po-
tentially carcinogenic properties. Metaboliza-
tion of the nitro group may be achieved under
anaerobic conditions by nitroreductases of in-
testinal bacteria or of the liver [39]. Bone mar-
row cells also have the potential to reduce the
nitro group even under aerobic conditions [40].
Acute Toxicity. The acute toxicity data for
nitroimidazoles indicate a moderate to consid-
erable acute toxicity depending on the sub-
stituents. In some cases the pH may also play a

role. The materials so far investigated are mostly
eye irritants, but not skin irritants.
The acute toxicities (LD50 , mg/kg) for some
nitroimidazoles are as follows:
4(5)-Nitroimidazole [26] 1660 (rat, oral)
1,2-Dimethyl-5-nitroimidazole 3000 (rat, oral)
hydrochloride (dimetridazole
hydrochloride) [26] 200 – 2000 (cat,
oral)
1-Methyl-5-nitroimidazole [26] ca. 200 (rat, oral)
1-(2-Hydroxymethyl)-2-methyl-5-nitroimida- 3800 (mouse, oral)
zole
(metronidazole) [41]
α-Methoxymethyl-2-nitroimidazole-1-etha- 2130 (rat, oral)
nol
(misonidazole) [42]
2-Methyl-4(5)-nitroimidazole [26] 1540 (rat, oral)
Four-hour dust inhalation tests on rats with
4(5)-nitroimidazole and dimetridazole gave
LC50 values of 6.56 and 5.3 mg/L, respec-
tively. Dimetridazole and 1-methyl-5-nitroimi-
dazole have an irritant action on the eyes of rab-
bits, while 4(5)-nitroimidazole is non-irritating
[26]. None of the three compounds exhibits irri-
tancy towards the skin of rabbits [26].
Mutagenicity. Several nitroimidazoles have
been subjected to bacterial point mutation as-
says (such as the Ames test), and, in most cases,
were found to exert mutagenicity [37], [43–51].
Among these were the following:
1-Methyl-5-nitroimidazole
2-Methyl-4(5)-nitroimidazole
1,2-Dimethyl-5-nitroimidazole (dimetridazole)
1-(2-Hydroxymethyl)-2-methyl-5-nitroimida-
zole (metronidazole)
2-Isopropyl-4-nitroimidazole
1-Methyl-2-isopropyl-5-nitroimidazole
(ipronidazole)
1-(3-Chloro-2-hydroxypropyl)-2-methyl-5-
nitroimidazole (ornidazole)
(1-Methyl-5-nitroimidazole-2-yl)methyl carba-
mate (ronidazole)
α-Methoxymethyl-2-nitroimidazole-1-ethanol
(misonidazole)
Nitroimidazoles also induce streptomycin-
resistant mutants (forward mutations) in Kleb-
siella pneumoniae [44], [45].
Conflicting results were earlier reported with
4(5)-nitroimidazole, 2-methyl-4(5)-nitroimida-
zole, and 2-methyl-4(5)-nitroimidazole-1-yl-
acetate [43], [46], [47]. Negative results may
Imidazole and Derivatives 7
be due to the inadequate metabolic activation
achieved in the test systems. Materials bearing a
carboxylate or sulfate group may in fact be non-
mutagenic [43], [45], but the data are insufficient
for a firm conclusion.
Urine and other body fluids of patients treated
with metronidazole in doses of 750 mg/d showed
mutagenic activity in the Ames test [52], [53].
Patients who received 200 – 1200 mg/d for sev-
eral months were found to have two- to four-fold
increases in chromosomal abnormalities in pe-
ripheral leukocytes [54].
Efforts have been undertaken to lower the
mutagenicity of nitroimidazoles by derivatisa-
tion and substitution [47], [48], [55]. Note, how-
ever, that mutagenicity in vitro is not quantita-
tively extrapolable to in vivo situations. Further-
more, a lower mutagenicity does not necessarily
indicate a decrease in the carcinogenic potential
as well.
Carcinogenicity. The carcinogenic potential
has been investigated for some nitroimidazoles
used or foreseen as drugs. However, the study de-
signs were not fully in accordance with present
standards. The carcinogenic responses obtained
so far were moderate or low. Human experi-
ence with nitroimidazole drugs has not revealed
cancer cases due to therapeutic usage. Metron-
idazole is listed as a carcinogen by IARC [56].
Mice receiving 0.06, 0.15, 0.3, or 0.5 % (ca. 90,
225, 450, or 750 mg/kg) of metronidazole in the
diet throughout their life-span showed a dose-
dependent increase in the incidence of lung tu-
mors and lymphomas. The lowest dose did not
significantly increase the incidence of tumors
[57].
No carcinogenic effects were found in rats in
a feeding experiment with 0.135 % (ca. 200 mg
kg −1 d−1 ) of metronidazole in the diet for 66
weeks and a post-observation period of 10 weeks
[58]. This administration period is, however, too
short to fully rule out carcinogenic effects.
More recent investigations with metron-
idazole in the diet of mice (75, 150, or
600 mg kg−1 d−1 for 92 weeks) and rats
(0.06, 0.3 or 0.6 %, equivalent to 30, 150 or
300 mg kg−1 d−1 for 150 weeks) showed dose-
related increases in tumor incidence (liver and
mammary tumors, Leydig cell and pituitary tu-
mors) [59], [60].
8 Imidazole and Derivatives
After gavage administration of metronida-
zole to mice during pregnancy and lactation
(2 mg per animal per day for 5 d per week) an
increased tumor incidence was observed in the
offspring [61].
Metronidazole (0.15 or 0.3 %) in the diet of
hamsters did not increase the incidence of tu-
mors; however, published details of the study
are insufficient for a reliable assessment of the
carcinogenic effect [62].
A group of patients who had been treated with
metronidazole during the period 1960 –1969 did
not show increased tumor rates in a follow-up
study published in 1988 [63].
Dimetridazole was found to have tumori-
genic activity in rats receiving 0.2 % in the diet
for 46 weeks. Of 35 Sprague – Dawley rats, 25
developed a benign, but possibly premalignant
mammary tumor (fibroadenoma) compared to 4
from 35 rats in the control group [58]. The ad-
ministration period was, however, too short to
fully determine the tumorigenic potential.
Ornidazole (25, 100, or 400 mg kg−1 d−1 )
was reported to cause no tumors in a two-year
feeding experiment in rats [64].
The overall conclusion from all experiments
is that the carcinogenic potential of nitroimida-
zoles has to be taken into account, but the amount
and quality of the available data do not allow firm
conclusions.
Neurotoxic effects. There are several case
reports on peripheral neuropathy due to Metron-
idazole treatment. The symptoms appear to be
reversible upon ceasing treatment [65].
8. References
General References
1. M. R. Grimmett in K. T. Potts: Comprehensive
Heterocyclic Chemistry, vol. 5, Pergamon
Press, 1984, pp. 345 – 456.
2. K. Hofmann: The Chemistry of Heterocyclic
Compounds, Imidazole and Derivatives, Part
I, Intersci. Publ. Inc., New York 1953.
Specific References
3. BASF, DE-OS 2 360 175, 1979 (A. Frank, H.
Karn, H. Spänig).
4. BASF, EP 0 012 371, 1982 (T. Dockner, U.
Kempe, H. Krug, P. Magnussen, W. Prätorius,
H. Szymanski).
5. BASF, DE 2 512 513, 1976 (A. Frank, T.
Dockner).
6. BASF, EP 0 111 073, 1987 (A. Frank, T.
Dockner).
7. BASF, EP 0 036 521, 1983 (T. Dockner, U.
Kempe, H. Krug, P. Magnussen, W. Prätorius).
8. Houdry Process Corp., US 2 847 417, 1958
(W. E. Erner).
9. BASF, EP 0 000 208, 1979 (T. Dockner, A.
Frank).
10. C. Corsar et al., Arzneim. Forsch. 16 (1966)
23.
11. BASF, DE 2 645 172, 1982 (M. Wetzler, T.
Dockner).
12. BASF, DE 1 670 293, 1976 (H. Spänig, A.
Steimmig, J. Sand).
13. BASF, DE-AS 2 233 908, 1977 (T. Dockner,
H. Krug).
14. M. Häring, Helv. Chim. Acta 42 (1959)
1845 – 1846.
15. A. Grimison, J. H. Ridd, B. V. Smith, J. Chem.
Soc. 1960, 1357 – 1362.
16. W. Reppe, Justus Liebigs Ann. Chem. 601
(1956) 81 – 138.
17. Rhône-Poulenc, DE-OS 1 470 200, 1970 (C.
Podesva, K. Vagi).
18. Y. Masashige, M. Masaichiro, Chem. Pharm.
Bull. 24 (1976) 1480 – 1484.
19. BASF, EP 0 015 516, 1982 (U. Kempe, T.
Dockner, F. Frank, H. Karn).
20. Delalande S.A., DE-OS 2 934 925, 1978 (Y. A.
Hubert-Brierre).
21. E. R. Shepard, H. A. Shonle, J. Am. Chem.
Soc. 69 (1947) 2269 – 2270.
22. W. Perkow: Wirksubstanzen der
Pflanzenschutz- und
Schädlingsbekämpfungsmittel, 2nd ed., Verlag
Paul Parey, Berlin – Hamburg 1985.
23. M. D. Nair, K. Nagarajan, Prog. Drug Res. 27
(1983) 163 – 252.
24. Shell Oil Co., US 3 538 039, 1970 (W. L.
Lantz, J. P. Manasia).
25. F. Ferrari, Arch. Int. Pharmacodyn. Ther. 227
(1985) 303 – 312.
26. BASF Aktiengesellschaft, unpublished results.
27. L. E. Voogd et al., Mutat. Res. 66 (1979)
207 – 221.
28. A. Momii et al., Iyakuhin Kenkyu 10 (1979)
351 – 357.
29. E. Grzybowska et al., Acta Microbiol. Pol. 34
(1985) 111 – 120.
30. J. A. Ruddick et al., Teratology 13 (1976)
263 – 266.

31. K. Nishie et al., Toxicol. Appl. Pharmacol. 14
(1969) 301 – 307.
32. T. E. Fairbrother et al., Vet. Hum. Toxicol. 29
(1987) 312 – 315.
33. M. H. Idaka, Okayama Igakkai Zasshi 88
(1976) 673 – 680.
34. Ullmann, 4th ed., 23, 617.
35. Niosh: Registry of Toxic Effects of Chemicals.
36. M. Hidaka, Okayama Igakkai Zasshi 88
(1976) 653 – 657.
37. D. G. Lindmark et al., Antimicrob. Agents
Chemother. 10 (1976) 476 – 482.
38. D. I. Edwards, M. Dye, H. Carne, J. Gen.
Microbiol. 76 (1973) 135.
39. J. L. Blumer et al., Cancer Res. 40 (1980)
4599 – 4605.
40. M. Isildar et al., Toxicol. Appl. Pharmacol. 94
(1988) 305 – 310.
41. B. Cavalleri et al., J. Med. Chem. 20 (1977)
1522.
42. National Technical Information Service USA,
PB 81-121212.
43. C. E. Voogd et al., Mutat. Res. 86 (1981)
243 – 277.
44. C. E. Voogd et al., Mutat. Res. 26 (1974)
483 – 490.
45. C. E. Voogd et al., Mutat. Res. 66 (1979)
207 – 221.
46. H. S. Rosenkranz et al., Mutat. Res. 38 (1976)
203 – 206.
47. J. B. Chin et al., Mutat. Res. 58 (1978) 1 – 10.
48. J. S. Walsh et al., J. Med. Chem. 30 (1987)
150 – 156.
49. M. Suzangar et al., Biochem. Pharmacol. 36
(1987) no. 21, 3743 – 3749.
Imidazole and Derivatives 9
50. P. G. Wislocki et al., Chem. Biol. Interactions
49 (1984) 27 – 38.
51. H. S. Rosenkranz et al., Biochem. Biophys.
Res. Commun. 66 (1975) 520 – 525.
52. M. S. Legator et al., Science 188 (1975)
1118 – 1119.
53. E. Mohtashamipur et al., Mutagenesis 1
(1986) 371 – 374.
54. F. Mitelman et al., Lancet, 1976, 802.
55. G. Cantelli-Forti et al., Teratog. Carcinog.
Mutag. 3 (1983) 51 – 63.
56. IARC Monograph 13, 113 – 121.
57. M. Rustia, P. Shubik, J. Nat. Cancer Inst. 48
(1972) 721 – 729.
58. S. M. Cohen et al., J. Nat. Cancer Inst. 51
(1973) 403 – 417.
59. J. H. Rust, An assessment of metronidazole
tumorigenicity studies in the mouse and the
rat, in S. M. Finegold (ed.): Metronidazole,
Proc. Intern. Metronidazole Conf. Montreal,
Excerpta Medica, Amsterdam (1977)
138 – 144.
60. M. Rustia, P. Shubik, J. Nat. Cancer Inst. 63
(1979) 863 – 868.
61. M. Chacko et al., Cancer Res. Clin. Oncol.
112 (1986) 135 – 140.
62. F. C. Roe, Proc. Intern. Metronidazole Conf.
Montreal, Excerpta Medica, Amsterdam
(1977) 132 – 137.
63. C. M. Beard et al., Mayo Clin. Proc. 63 (1988)
147 – 153.
64. R. Richle et al., Arzneim. Forsch. 28 (1978)
612 – 625.
65. W. G. Bradley et al., Br. Med. J. 2 (1977)
610 – 611.
 Immobilized Biocatalysts 1

Immobilized Biocatalysts
Jiří E. Přenosil, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland
Ömer M. Kut, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland
Irving J. Dunn, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland
Elmar Heinzle, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland

1. Introduction . . . . . . . . . . . . . . 3 4.3. Internal Mass Transfer . . . . . . . 24

2. Immobilization Techniques . . . . 5 5. Design of Immobilized Biocatalyst
2.1. Introduction . . . . . . . . . . . . . . 5 Reactors . . . . . . . . . . . . . . . . . 28
2.2. Methods of Enzyme Immobiliza- 5.1. Batch Stirred-Tank Reactor . . . . 28
tion . . . . . . . . . . . . . . . . . . . . 5 5.2. Plug-Flow Tubular Reactor . . . . 30
2.2.1. Carriers for Enzyme Immobilization 5 5.3. Continuous Stirred-Tank Reactor 30
2.2.2. Methods for Insoluble Enzymes . . 6 5.4. Packed- or Fixed-Bed Reactors . 30
2.2.2.1. Carrier Binding . . . . . . . . . . . . . 6 5.5. Fluidized-Bed Reactors . . . . . . . 31
2.2.2.2. Cross-Linking . . . . . . . . . . . . . 8 5.6. Membrane Reactor Systems . . . 32
2.2.2.3. Enzyme Copolymerization . . . . . 11 5.7. Laboratory Reactors for Kinetic
2.2.2.4. Entrapment . . . . . . . . . . . . . . . 11 Measurements . . . . . . . . . . . . . 32
2.2.3. Methods for Soluble Enzymes . . . 12 5.7.1. Batch and Continuous Tank Reactors 32
2.3. Methods of Cell Immobilization . 12 5.7.2. Experimental Differential Recycle
2.3.1. Introduction . . . . . . . . . . . . . . . 12 Reactor . . . . . . . . . . . . . . . . . . 33
2.3.2. Cell Supports . . . . . . . . . . . . . . 13 6. Immobilization Media . . . . . . . 34
2.3.3. Immobilization Techniques . . . . . 14 6.1. Characteristics of Solid Supports 34
2.3.3.1. Cell Immobilization without a Sup-
6.2. Characteristics of Semipermeable
port . . . . . . . . . . . . . . . . . . . . 14
Separators . . . . . . . . . . . . . . . 34
2.3.3.2. Binding of Cells to a Carrier . . . . 14
7. Application of Immobilized Bio-
2.3.3.3. Immobilization of Cells by Entrap-
catalysts . . . . . . . . . . . . . . . . . 34
ment . . . . . . . . . . . . . . . . . . . 15
7.1. Analysis . . . . . . . . . . . . . . . . . 34
2.4. Methods of Organelle Immobiliza-
7.2. Immobilized Enzymes in Thera-
tion . . . . . . . . . . . . . . . . . . . . 17
peutic Medicine . . . . . . . . . . . . 36
2.5. Co-immobilization of Biocatalysts 17
7.3. Ethanol Production Using Immo-
3. Activity and Kinetics of Immobi-
bilized Cells . . . . . . . . . . . . . . 37
lized Biocatalysts . . . . . . . . . . . 18
3.1. Effects of Immobilization on En- 7.4. Industrial Applications of Immo-
zyme Activity . . . . . . . . . . . . . 18 bilized Biocatalysts . . . . . . . . . . 37
3.2. Kinetics of Immobilized Biocata- 7.4.1. Production of High-Fructose Corn
lysts . . . . . . . . . . . . . . . . . . . . 20 Syrup . . . . . . . . . . . . . . . . . . . 38
3.3. Assay of Immobilized Biocatalysts 21 7.4.2. Amino Acid Production . . . . . . . 39
4. Mass Transfer in Immobilized 7.4.3. Environmental Applications . . . . . 41
Biocatalyst Systems . . . . . . . . . 21 8. Economic Aspects . . . . . . . . . . 41
4.1. External Mass Transfer . . . . . . 21 9. Safety, Environmental, and Legal
4.2. Prediction and Correlation of Ex- Aspects . . . . . . . . . . . . . . . . . 43
ternal Mass-Transfer Coefficients 22 9.1. Immobilized Enzymes . . . . . . . . 43
4.2.1. Catalyst Beds . . . . . . . . . . . . . . 22 9.2. Immobilized Cells . . . . . . . . . . 44
4.2.2. Particles in an Agitated Tank . . . . 23 10. References . . . . . . . . . . . . . . . 46

Symbols: D diffusivity, m2 /s
A area, m2 F Faraday constant; volumetric flow rate,
C concentration, g/m3 m3 /s
d diameter, m j diffusion flux, g m−2 s−1

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 001
2 Immobilized Biocatalysts

jD factor in mass-transfer correlation R recycle stream

k Boltzmann constant; mass-transfer con- Re Reynolds number
stant; rate constant s surface; system
K M Michaelis – Menten constant s surface
L total length, depth, or thickness, m S solid; substrate
m mass Sc Schmidt number
n exponent or reaction order Sh Sherwood number
N dimensionless number SL liquid diffusion film at the solid – liquid
P power, g m−2 s−3 interface
PD−R parameter for diffusion – reaction pro- SLt liquid diffusion film at the solid – liquid
cess interface using terminal velocity
r reaction rate, g m−3 s−1 or g m−2 s−1 t terminal velocity
R gas constant T total system
S substrate concentration, g/m3 or mol/m3 0 bulk; initial value; inlet; zero order

S dimensionless substrate S/S0 1 first order; outlet
t time, s 2 second order
t dimensionless time t/(L2 /D)
T temperature, K
v flow velocity, m/s; reaction velocity,
g m−3 s−1 1. Introduction
V volume, m3 or m3 /mol
z valence Catalytic activity is an important property of
Z length, m many biological materials termed biocatalysts,
Z dimensionless length Z/L which include whole cells, parts of cells, and en-
ε void fraction; volume fraction zymes. Biocatalysts catalyze chemical reactions
η effectiveness factor that usually occur in living organisms.
µ viscosity, g m−1 s−1 The catalytic power of enzymes has been ex-
µ/ kinematic viscosity, m2 /s ploited by humans for thousands of years. The
 density, g/m3 final product is often formed by a number of re-
τ residence time, s action steps catalyzed by a system of enzymes
Φ Thiele modulus acting in ordered fashion. The enzymatic system
ψ electrostatic potential, V is usually localized within a living cell (microor-
ganism, animal, or plant cell). The catalytic ac-
tivity of the whole cell or parts of it can be used
in the same manner as isolated enzymes.
Indices The use of single purified enzymes may be
A location in bulk, substance A desirable for reaction modeling or process man-
app apparent agement, but in many cases this approach is as-
avg average value sociated with practical difficulties (primarily en-
B location on surface, substance B zyme stability and purification cost). Therefore,
bulk bulk solution the use of whole cells or parts of cells containing
c catalyst an appropriate enzyme system may be advanta-
i component i geous, especially when complex reactions are
I impeller involved. Furthermore, the problems of resid-
L liquid ual biomass and wastewater disposal involved
m mass in large-scale enzyme production, as well as the
max maximum costs of enzyme isolation and purification, may
M matrix become prohibitive. The lower specific activity
n nth position or number of whole cells and a potentially lower selectivity
p particle are often balanced by lower cost and simplicity
Pe Péclet number of use.
r reactor
 Immobilized Biocatalysts 3

Many processes utilized by humans mimic lysts consisting of either enzymes or whole cells,
nature, and enzyme immobilization is no excep- which occur in suspended form or are immobi-
tion. All native intracellular enzymes seem to be lized in solids, gels, or biofilms.
bound to the cell membrane or localized in the
gel-like surroundings of a cellular organelle [1]. Historical Aspects. The use of fixed biocat-
Cellular enzyme systems vary greatly from alysts is closely connected with the history of
two or more sequentially acting independent en- human civilization. Vinegar manufacturing was
zymes to multienzyme complexes consisting of one of the first industrial processes to use im-
tightly aggregated enzymatic or regulatory sub- mobilized living cells. Alcoholic solutions were
units which lose their activity upon separation trickled through vats packed with wood shav-
[2], [3]. ings. Microbial films developed on the shavings
Under natural conditions, microorganisms and converted the ethanol to acetic acid (“Quick”
tend to grow in aggregates attached to solid sur- process, invented in 1823). Attempts to immo-
faces or to form microbial films (biofilms) on a bilize enzymes date back as far as 1916, when
variety of materials. Studies on microbial popu- Nelson and Griffin showed that immobilized
lations in rivers show that very few microorgan- invertase was active [5].
isms exist free in suspension; instead, they are An overview of the historical development of
associated with solid surfaces (e.g., silt particles) immobilized biocatalysts is given in Table 1.
[4].
Biologically active materials are immobi-
lized mainly to allow repeated or continuous use
of their activity in a controlled environment (e.g., Table 1. Historical development of immobilized biocatalysts
(partly based on [6])
a bioreactor). Anchorage of an enzyme molecule
to a carrier can also stabilize its tertiary structure Year Biocatalyst, process, or event Comments
and prolong its useful life. 1823 vinegar manufacture from
Enzymes can be retained in a reactor by cova- alcoholic solutions by using biofilm
lent, ionic, or adsorptive attachment to surfaces, on wood chips (“Quick” process)
1916 invertase adsorbed on charcoal [5]
cross-linking, entrapment in gels and polymers, 1948 urease insolubilization [7]
inclusion in micelles, or retention by membrane 1949 physiologically active protein
systems. The methods used for retaining whole immobilization [8]
1951 albumin bound to diazonium
cells in a reactor are immobilization in gels and derivative of 4-aminobenzyl
polymers, entrapment in porous solids or beds cellulose [9]
of particulates, formation of biofilms and flocs, 1953 immobilization of
carboxypeptidase, diastase, pepsin,
and the utilization of membrane systems. In con- and ribonuclease [10]
tinuous processes, attention must be paid to the 1969 continuous optical resolution of first industrial
d,l-amino acids by immobilized process employing
problem of enzyme loss, especially when the cell aminoacylase [11] an immobilized
wall has been made permeable to facilitate sub- biocatalyst
strate and product transport and the carrier ma- 1971 First Enzyme Engineering definition of
Conference at Henniker, New immobilized
trix does not retain the enzyme. Hampshire [12] enzymes
Most of the techniques and principles applied 1972 isomerization of glucose to
to immobilized enzymes are applicable to im- fructose by immobilized glucose
isomerase [13]
mobilized whole cells. In fact, from the stand- 1973 production of l-aspartic acid by first industrial use of
point of engineering, little reason exists for dis- immobilized Escherichia coli immobilized
tinguishing between these two types of biocata- (aspartase) [14] microbial cells
1975 adsorption of rat liver mitochondria first immobilization
lyst. The application of biofilms or microorgan- on alkylsilanized glass beads [15] of subcellular
isms that are naturally immobilized by their abil- particles
ity to form flocs has no counterpart in enzyme 1979 adsorption of algae (Anabaena
cylindrica) on glass beads for the
technology, but the same principles of diffusion evolution of hydrogen [16]
and reaction are also involved here.
Similarly, a dissolved enzyme differs little
from a suspended organism in a membrane re-
actor. Therefore, this article deals with biocata-
4 Immobilized Biocatalysts
Table 1. Continued
Year Biocatalyst, process, or event Comments
1979 immobilization of plant cells first immobilization
(Morinda, Catharanthus, Digitalis) of plant cells
by the calcium alginate method
[17]
1979 slice of porcine kidney first immobilization
immobilized in the membrane of of animal cells
an ammonia electrode for amino
acid detection [18]
1987 Corynebacterium spp. immobilized first reported
in polyacrylamide gel for industrial production
acrylamide production [19] of a commodity
chemical by
immobilized cells
2. Immobilization Techniques
2.1. Introduction
Since the late 1960s a variety of techniques have
been developed for immobilizing biocatalysts.
Complete coverage of the vast number of pa-
pers, books, and patents on this topic is almost
impossible. The reader is referred to comprehen-
sive reviews and monographs [20–38].
The definition of biocatalysts used in this
chapter is that drawn up by the Working Party on
Applied Biocatalysts within the European Fed-
eration of Biotechnology: “Immobilized biocat-
alysts are enzymes, cells, or organelles (or com-
binations of them) which are in a state that per-
mits their reuse” [39].
Biocatalysts dissolved in aqueous buffer so-
lutions are usually referred to as soluble or na-
tive enzymes, cells, cell parts, or organelles. The
terms immobilized, fixed, or insolubilized en-
zymes, cells, etc., denote biocatalysts that are
bound to a support. The material to which en-
zymes or cells are bound is called the carrier,
support, or matrix. The agent enabling enzyme –
support binding is referred to as a cross-linking
agent, bifunctional agent, or carrier activator.
For clarity, techniques for immobilization of en-
zymes and cells are discussed separately (Sec-
tions 2.2 and 2.3, respectively).
2.2. Methods of Enzyme Immobilization
(see also → Enzymes, Chap. 3.3.)
The definition of immobilized biocatalysts given
in Section 2.1 includes enzymes that are
1) modified into a water-insoluble form,
2) retained by an ultrafiltration membrane in-
side a reactor, or
3) bound to another macromolecule to restrict
their mobility.
Several schemes have been suggested for
classifying immobilization techniques [24],
[30], [40], [41]. The classification used here is
based on that of Kennedy and Cabral [41]
and attempts to combine the nature of the inter-
action responsible for immobilization with that
of the support (Fig. 1).
A perfect, universal immobilization method
does not exist. Each end use requires evaluation
of the individual steps according to criteria such
as the purpose of immobilization, activity, sta-
bility, simplicity, and economic feasibility [42],
[43].
2.2.1. Carriers for Enzyme Immobilization
The major components of an immobilized en-
zyme system are the enzyme, the support, and
the mode of interaction of the enzyme with the
support. The support may be a membrane, a
water-insoluble solid, or a polymer matrix. The
support (carrier) should have the following prop-
erties [44], [45]:
1) Large surface area and high permeability
2) Sufficient functional groups for enzyme at-
tachment under nondenaturing conditions
3) Hydrophilic character
4) Water insolubility
5) Chemical and thermal stability
6) Mechanical strength
7) High rigidity and suitable particle form
8) Resistance to microbial attack
9) Regenerability
10) Toxicological safety
11) Low or justifiable price
Carriers can be classified according to their
morphology (porous, nonporous, or gel type) or
their chemical composition. The most widely
used supports are listed in Table 2. The charac-
teristics of these materials are presented in [44],
[45]. Sources of commercial carriers are listed
in [46].
 Immobilized Biocatalysts 5

Figure 1. Classification of enzyme immobilization methods

Table 2. Chemical classification of matrixes used for enzyme

immobilization [45]
2.2.2. Methods for Insoluble Enzymes
Organic matrixes Inorganic matrixes
2.2.2.1. Carrier Binding
Natural polymers Minerals
Polysaccharides Attapulgite clays Enzymes are large protein molecules with chem-
Cellulose Bentonite
Starch Kieselghur
ically reactive groups, ionic groups, and hy-
Dextran Pumice stone drophilic domains that can all participate in
Agar and agarose the immobilization process through physical ad-
Alginate sorption, ionic binding, or covalent linkage.
Carrageenan
Chitin and chitosan
Proteins Physical Adsorption. In physical adsorp-
Collagen tion, the enzyme adheres to the surface of a
Gelatin
Albumin
support which has not been specially function-
Silk alized for covalent binding [47]. Attachment is
Carbon materials achieved by physical interactions such as van
(activated carbon)
der Waals forces, hydrogen bonding, or hy-
Synthetic polymers Synthetic materials
Polystyrene Nonporous glass drophilic – hydrophobic effects. In the absence
Polyacrylates and poly- Controlled pore glass of chemical reactions, practically no conforma-
methacrylates Controlled pore metal tional changes occur in the enzyme and associ-
Polyacrylamide oxides
Hydroxyalkyl methacrylates Metals
ated loss of activity is minimal. However, bind-
Glycidyl methacrylates ing forces between the enzyme and carrier are
Maleic anhydride polymers generally weak, and leakage of enzyme by des-
Vinyl and allyl polymers orption may occur in flow systems. Changes
Polyamides
of operational parameters such as pH, ionic
strength, or temperature can increase leakage
dramatically. Commonly used absorbents are
6 Immobilized Biocatalysts
alumina, activated carbon, clay, diatomaceous
earth, glass, and hydroxyapatite. The problems
of enzyme immobilization by physical adsorp-
tion are discussed in [47], [48], and examples
are given in [49], [50]. In some cases such prob-
lems can be eliminated by cross-linking the ad-
sorbed enzyme using multifunctional reagents
(Section 2.2.2.2).
Ionic Binding. Ionic binding is an estab-
lished, simple way of immobilizing enzymes.
The binding forces are ion – ion interactions and
are stronger than those in simple physical ad-
sorption. Again, however, binding stability is af-
fected by changes of pH or ionic strength. Both
commercial anion and cation exchangers can be
used as carriers. Examples are given in [24],
[49].
Chelate Binding. The chelating properties
of a transition metal such as titanium or zirco-
nium are employed to couple enzymes to an or-
ganic material (e.g., cellulose, sawdust, alginic
acid) or an inorganic support such as celite or
glass [26], [41], [51]. This method was initally
used by Novais in 1971 [52] and was developed
further by Kennedy and co-workers [53]. The
chemistry of metal – enzyme coupling, proper-
ties of the immobilized system, and detailed
working procedures are given in [54].
Biospecific Binding. Biospecific interac-
tions between enzymes and other molecular
species (e.g., lectins or antibodies) are also used
for binding. Lectins such as concanavalin A
[11028-71-0] are glycoproteins, mainly of plant
origin, that bind strongly to specific carbohy-
drate residues. For example, a carbohydrate-
containing enzyme (e.g., l-ascorbate oxidase)
can be bound specifically to commercial Con-
canavalin A-Sepharose [55]. The same method
has been used to immobilize invertase for con-
tinuous conversion of sucrose [56].
A novel method for preparing highly ac-
tive immobilized enzymes is to bind them
with specific monoclonal antibodies attached
to carriers [57–59] (Fig. 2). The enzyme is
bound at a specific, well-defined site without
affecting its catalytic activity. Transglutami-
nase (glutaminylpeptide γ-glutamyltransferase
[80146-85-6] E.C. 2.3.2.13) has been immobi-
lized in this way on cross-linked agarose gel
beads ( AffiGel 10) to produce a very spe-
cific, highly active biocatalyst starting from a
crude enzyme preparation [57]. In the immo-
bilization of carboxypeptidase A [11075-17-5]
(E.C. 3.4.17.1), the highest enzyme activity was
shown with a mouse monoclonal antibody –
enzyme complex on Eupergit C (Röhm Pharma)
or on a Protein A-Sepharose CL 4B (Pharmacia)
[58].
Figure 2. Schematic representation of biospecific immobi-
lization
Covalent Binding. A water-insoluble car-
rier can be covalently bound to the enzyme via
the reactive side groups of amino acid residues
(e.g., amino, hydroxyl, thiol, or phenolic groups)
that are not associated with the active site or the
substrate binding site (see Table 3). This tech-
nique includes the greatest number and variety of
published immobilization methods [43]. Under
normal conditions, covalent binding is stable and
no enzyme leakage occurs if substrate concen-
tration or ionic strength changes during the reac-
tion. Coupling reactions are often complicated
and are carried out under relatively severe con-
ditions in comparison with physical adsorption
methods. Covalent binding can alter the confor-
mation and active site of the enzyme, resulting in
significant loss of activity or selectivity. The ad-
vantage of this approach is the wide range of pos-
sible binding methods and carriers, which allows
great flexibility in designing biocatalysts with
specific chemical and physical properties [55].
For example, if a hydrophilic environment is re-
quired, a coupling agent containing a number of
hydroxyl groups may be utilized. An aromatic
coupling agent can improve the compatibility of
a substrate – enzyme reaction when the substrate
is also aromatic [48]. Glycoenzymes can be cou-
pled selectively via their carbohydrate moieties
[60], [61].

Table 3. Reactive residues of proteins [45]
A survey of coupling methods and sources
of commercial matrix materials can be found in
[46].
Support materials in their original form do not
usually possess reactive groups for direct cou-
pling with enzymes. Their existing surface hy-
droxyl, amino, amido, or carboxyl groups must
be chemically activated. The resulting reactive
groups on the carrier surface and the coupling
reactions are given in Table 4. The chemical re-
actions employed for coupling can be classified
according to the chemical characteristics of the
carrier [43], [48] or the coupling reaction [24],
[41].
Diazo Coupling is based on coupling of the en-
zyme (e.g., the imidazole side chain of histidine
or the phenolic group of tyrosine) to the aryldia-
zonium groups of the carrier. The aromatic ami-
no groups on the surface of the carrier can be di-
azotized easily with nitrite in an acidic medium.
Typical carrier structures with aromatic amines
are given in [50]. Examples of applications of
the method are given in [24], [49].
Peptide Bond Formation. The nucleophilic
(amino, hydroxyl, thiol) groups of the enzyme
can attack the activated functional groups to
form a covalent (amide, peptide, or other) bond.
Immobilized Biocatalysts 7
Two different surface activation processes can
be used [24], [44]:
1) Supports containing carboxyl groups are
converted to reactive derivatives (e.g., acyl
azide, acid chloride, acid anhydride, isocy-
anate, isothiocyanate, cyclic carbonate, or
cyclic imidocarbonate), which can react di-
rectly with free amino functions of the pro-
teins. The well-known activation of polysac-
charides with cyanogen bromide [506-68-3]
is an example of this method [62]. The vic-
inal hydroxyl groups of the polysaccharide
matrix (e.g., dextran, cellulose, or agarose)
react with cyanogen bromide at pH 10 – 11.5
to give the reactive imidocarbonate deriva-
tive. At pH 9 – 10 the free amino groups react
to give essentially substituted isoureas [63].
2) Condensing reagents such as carbodi-
imides or Woodward’s reagent K (N-
ethyl-5-phenylisoxazolium-3 -sulfonate)
[4156-16-5] can form amide linkages bet-
ween the free amino and carboxyl groups of
the carrier surface and the enzyme.
Examples of enzyme immobilization by
amide binding are presented in [24], [49].
Alkylation or Arylation. Free amino, phe-
nolic, or thiol groups of an enzyme are alkylated
or arylated with an activated functional group of
the support such as halide, epoxy, or vinylsulfo-
nyl groups. Supports with hydroxyl groups (e.g.,
polysaccharides or minerals) can be activated
by treatment with cyanuric chloride derivatives.
Another widely used support contains reactive
epoxy groups. Details of modifications of the
method and practical examples are given in [24],
[41], [49], [50].
Miscellaneous Methods. Several other cova-
lent binding methods have been described for
the immobilization of enzymes. These include
Schiff base formation, Ugi reaction [64], thiol –
disulfide interchange, amidination, mercury –
enzyme interaction, and γ-radiation induced
coupling. These methods have limited applica-
bility and are reviewed in [20], [24], [41], [45],
[49], [50].
2.2.2.2. Cross-Linking
The enzyme can be immobilized by cross-
linking it to other enzyme molecules or to
8 Immobilized Biocatalysts
Table 4. Carrier and enzyme functional groups used for the immobilization of enzymes by covalent binding [45]
 Immobilized Biocatalysts 9
Table 4. Continued
10 Immobilized Biocatalysts
an inert protein such as albumin (co-cross-
linking) and precipitating the resulting aggre-
gate [24], [25], [49]. This method can also be
used in combination with a carrier such as a
membrane, where the physically adsorbed en-
zymes are cross-linked on the membrane sur-
face [65]. Bi- or multifunctional reagents such
as glutaraldehyde [111-30-8], toluene diisocya-
nate [1321-38-6], or bisdiazobenzidine deriva-
tives can be used as cross-linking agents. The
primary disadvantages of this method are the
relatively severe reaction conditions and diffi-
culties in controlling the reaction. Hence, activ-
ities of such biocatalysts are generally low [24],
[25], [40], [41], [66].
2.2.2.3. Enzyme Copolymerization
The enzyme can be immobilized by copolymer-
izing it with the polymer matrix. Although the
enzyme is then a covalently bound part of the
matrix, this method can be regarded as a modifi-
cation of gel entrapment (see Section 2.2.2.4).
In the original method, proteins were viny-
lated with acylating or alkylating monomers and
copolymerized with other monomers. The ad-
vantages of this method over matrix entrapment
are the higher activity and higher stability due
to the lack of enzyme bleeding [67], [68].
In enzyme graft copolymerization, enzyme
copolymer is grafted onto a matrix such as a
polysaccharide carrier [69], [70]. High mechan-
ical strength can be attained, depending on the
carrier chosen. High intra particle diffusion re-
sistances can be circumvented if the enzyme
copolymer is grafted to the surface of a carrier
with low porosity [55]. For example, horseradish
peroxidase (E.C. 1.11.1.7) was immobilized un-
der strictly controlled conditions to give a bio-
catalyst with kinetic constants close to those of
the native enzyme but with higher thermal sta-
bility [71].
2.2.2.4. Entrapment
Entrapment may be regarded as the physical con-
finement of an enzyme in a semipermeable ma-
trix. The matrix lattice must be tight enough for
the enzyme to be retained but must allow perme-
ation of substrate and product(s). The enzyme is
generally not chemically bound to the matrix as
in enzyme copolymerization. Entrapment meth-
ods can be subdivided into gel entrapment and
microencapsulation and reverse micelles.
Gel Entrapment. In this procedure, free en-
zyme is entrapped within the interstitial spaces
of a cross-linked, water-insoluble polymeric gel.
Gel-forming materials such as polysaccharides,
proteins, or synthetic polymers can be em-
ployed. Among the natural polysaccharides, al-
ginate, agar, and κ-carrageenan are widely used
in preparing immobilized biocatalysts.
Polyacrylamide gel was the first material em-
ployed for entrapping enzymes [e.g., trypsin
(E.C. 3.4.21.4), papain (E.C. 3.4.22.2), β-
amylase (E.C. 3.2.1.2), and d-fructose 1,6-
diphosphate aldolase (E.C. 4.1.2.13)] [72]. The
polymerization of acrylamide can be initiated
with potassium persulfate and by X- or γ-
radiation [73]. Another method uses prepoly-
mers that can be cross-linked by UV radi-
ation; photo-cross-linkable resin prepolymers
with varying hydrophilic – hydrophobic ratios,
chain lengths, and anionic or cationic func-
tional groups are available [74]. Such biocata-
lysts can also be used for bioconversion of highly
lipophilic or slightly water-soluble substrates in
reaction media consisting of water and a water-
miscible organic cosolvent or water-containing
organic solvents. Hydrophilic gel matrixes and
organic solvents of low polarity are generally
preferred for bioconversion of lipophilic sub-
strates [75]. For further details of this method,
the literature should be consulted [20], [24],
[41], [43], [45], [55].
A special modification of this method in-
volves enzyme entrapment during the spinning
of synthetic fibers such as cellulose triacetate
[76], [77]. High surface areas can be achieved by
using very fine fibers, which also show chemical
resistance to weak acid, alkali, and some organic
solvents. Fiber-entrapped enzymes are used in
the pharmaceutical industry (penicillin acylase
E.C. 3.5.1.11 is used in the synthesis of anti-
biotics with a β-lactam ring) and food industry
(β-D-galactosidase, E.C. 3.2.1.23, is used for the
hydrolysis of lactose). The risk of microbial con-
tamination is one disadvantage of this method;
a reactor densely packed with fibrous material
cannot be sterilized easily.

Microencapsulation. In microencapsula-
tion, enzymes are immobilized by enclosing
them in membranes that are permeable to the
substrate and the product. The resulting mi-
crocapsules generally have a diameter of 1 –
100 µm. Chang, who pioneered this approach,
describes the procedures in detail [78]; see also
→ Microencapsulation. An emulsion is usually
prepared from an organic phase and an aque-
ous enzyme-containing phase in the presence of
a surfactant. The membrane-forming polymer,
dissolved in an organic solvent, is then added.
On standing, polymer precipitates on the micro-
droplet interface to form stable microcapsules
[55], [79].
Unmodified coenzymes or multienzyme sys-
tems can also be immobilized by microencapsu-
lation with lipid – polyamide membranes [80].
Microencapsulation gives the highest enzyme
concentrations per unit volume of immobilized
preparation [48]. However, as in all semiperme-
able membrane systems, the substrate diffusion
rate limits the applications of microencapsulated
biocatalysts.
Reverse Micelles. Amphiphilic surfactant
molecules can form reverse micelles in hy-
drocarbon solvents. The micelles contain wa-
ter (water pools) and can be used as hosts for
biopolymers. The solubilized biopolymers (en-
zymes or even whole cells) retain their biologi-
cal activity and are protected against the organic
solvent by the surfactant envelope. Such sys-
tems may be applied for biocatalytic conversion
of water-insoluble substrates or other reactions
requiring the presence of an organic phase. The
micelles must be kept in an appropriate reactor
which allows physical separation of enzyme and
reagents; use of membrane reactors seems to of-
fer substantial advantages [81]. A review of the
use of enzymes and nucleic acids in hydrocarbon
micellar solutions is presented in [82].
2.2.3. Methods for Soluble Enzymes
In the immobilization methods described above,
modification of the native enzyme or its mi-
croenvironment often decreases enzyme activity
or selectivity. To keep the enzymes in their native
state, an enzyme solution can be separated from
the substrate and product by a semipermeable
Immobilized Biocatalysts 11
membrane. The membrane pores allow substrate
and product diffusion and physically confine the
larger enzyme molecules. Flat sheet ultrafiltra-
tion or microfiltration membranes or hollow-
fiber membranes can be used (Section 5.6). Co-
factors, which are normally small molecules that
could diffuse through the membrane, can be re-
tained in the reaction zone by coupling them to
larger molecules. For example, nicotinamide –
adenine dinucleotide (NAD) is completely re-
tained by a hollow fiber cell when it is attached to
a soluble poly(ethylene glycol) chain [83]. The
enzyme solution is also protected by the mem-
brane from microbial attack. Process selectivity
can be controlled by proper design of membrane
porosity and hydrophilicity. The main problems
associated with membrane reactors are mem-
brane fouling caused by the adsorption of diffu-
sion species and mass-transfer resistances [45].
A review of membrane immobilization is given
in [84]; see also → Membranes and Membrane
Separations.
Limitation of diffusion through membranes
can be circumvented by using soluble –
insoluble immobilized enzymes. Such prepara-
tions are very useful for conversion of water-
insoluble substrates such as cellulose. Cellulase
from Aspergillus niger has been immobilized
on poly(l-glutamic acid), which is soluble in
neutral and alkaline solution but can be pre-
cipitated by lowering the pH without loss of
enzyme activity [85]. Preparation of soluble –
insoluble immobilized proteases such as papain
(E.C. 3.4.22.2) or chymotrypsin (E.C. 3.4.21.1),
their properties, and possible applications are
discussed in [86].
2.3. Methods of Cell Immobilization
2.3.1. Introduction
Although self-immobilization of cells on sur-
faces commonly occurs in nature, the intentional
attachment of cells to a support is a recent devel-
opment. Immobilized whole-cell systems ( mi-
crobial, yeast, plant, or mammalian cells) are
much more versatile than the corresponding im-
mobilized enzymes. Such biocatalysts can con-
sist of intact or disintegrated dead cells that con-
tain active enzymes. However, viable resting or
12 Immobilized Biocatalysts
growing cells can also be immobilized; this tech-
nique is used especially with eukaryotic cells
where the whole metabolic machinery is often
required for their specific application.
The state of the art in immobilized cell tech-
nology is described in detail in [21], [22], [25],
[40], [43], [66], [87–102]. As in enzyme immo-
bilization, no “ideal” general method exists. This
chapter merely gives an overview of the advan-
tages and disadvantages of immobilized whole-
cell systems and the most generally applied im-
mobilization techniques.
The primary advantage of immobilized cell
cultures as opposed to suspension cultures is
the reduced cost of bioprocessing as a result
of higher cell densities, easy separation, and re-
peated or continuous use of biocatalyst. Immo-
bilization may also increase the shear resistance
of shear-sensitive eukaryotic cells. The enzymes
retain their native geometry and microenviron-
ment, and are protected against changes in the
bulk system (temperature, pH, ionic strength,
mechanical stress). Support materials generally
stabilize biocatalyst activity and storage stabil-
ity, but this phenomenon is not fully understood.
Correct choice of the hydrophobicity – hy-
drophilicity balance of the support can alter the
microenvironment by increasing the concentra-
tion of limiting substrate or by decreasing the
concentration of inhibiting product in the vicin-
ity of the biocatalyst [92], [103]. Poor diffu-
sion and permeability into the carrier matrix and
through the cell membrane limit the applicabil-
ity of such biocatalyst systems to low molecular
mass substrates or products. This transport bar-
rier can protect the cells against microbial or
viral infection if the pore width of the matrix is
sufficiently small [92].
The reactions catalyzed by immobilized
whole-cell biocatalysts can be classified as fol-
lows:
1) Reactions involving single enzymes (biocon-
versions)
2) Reactions involving multienzyme systems
with or without cofactors
3) Reactions involving a complete metabolic
pathway yielding primary or secondary
metabolites
The use of immobilized microbial cells is ad-
vantageous in the following areas [96]:
1) when the desired enzymes are intracellular
and the extracted, purified enzymes become
unstable after immobilization;
2) when the microorganism does not contain in-
terfering enzymes or when such enzymes can
be inactivated without loss of desired cat-
alytic activity; and
3) when substrates and products do not have a
high molecular mass and can diffuse through
the cell membrane.
The use of whole cells instead of isolated
enzymes obviates costly, tedious purification
procedures. In multienzyme reactions, the cell
structure provides the proper geometric en-
zyme arrangement which is extremely difficult
to achieve by enzyme co-immobilization. For re-
actions requiring cofactor regeneration or for the
production of primary or secondary metabolites,
the immobilized cells must be viable. If cell vi-
ability is not a prerequisite, diffusion problems
can be solved by chemical or electrical perme-
abilization of the cell membrane.
The major advantage of such biocatalytic sys-
tems is that the operational stability can often
be markedly enhanced by reactivating the en-
zymes in the immobilized cells [104]. Such re-
activation can be achieved either by brief incuba-
tion with growth medium or by continuous con-
trolled growth of the immobilized cells [105].
2.3.2. Cell Supports
A variety of inorganic and organic matrices are
available for cell immobilization; they interact
with the cell surface by physical or chemical
bonds [106], [107]. A comprehensive list of
commercial support materials is given in [46].
A number of factors must be considered in the
selection of a suitable carrier for large-scale use
[98], [100], [103]:
1) The material should be available in sufficient
quantities and at low price.
2) The material should have a large surface area
accessible to cells and reactants.
3) The material must be mechanically, chemi-
cally, and thermally stable under process and
storage conditions.
4) The matrix should contain a sufficient num-
ber of functional groups to bind the cells.

5) The material should not reduce cell activity
or initiate cell lysis.
6) The material should be easy to handle in the
immobilization procedure.
7) The material should be capable of recycling
or safe disposal.
8) In the case of viable growing cells, the matrix
should have a sufficiently large void volume
or be elastic enough to accommodate new
cells.
2.3.3. Immobilization Techniques
Most of the common cell immobilization tech-
niques are extensions or modifications of those
discussed in Section 2.2.2 for enzymes. These
methods can be grouped according to the way
in which cells bind to the matrix (Fig. 3). An al-
ternative way of classifying these methods is to
use the route by which the biocatalyst is prepared
[103]:
1) Carrier-free immobilization
2) Immobilization of a given biomass onto a
preformed carrier surface
3) Immobilization of a given biomass during the
course of carrier formation (e.g., by polymer-
ization)
4) Immobilization by controlled growth of an
inoculum or by germination of immobilized
spores
2.3.3.1. Cell Immobilization without a
Support
The simplest way to immobilize biocatalysts is
to use their intrinsic tendency to aggregate or
flocculate at high cell densities. Yeast, mold, and
plant cells aggregate [90], [95], [99], [104] and
are relatively stable to shear fields in fluidized-
bed reactors [99], [108]. Flocculation can also
be induced by polyelectrolytes such as chitosan
[93], [99], [104].
Cell aggregation can be induced by low
molecular mass bi- or multifunctional reagents
such as glutaraldehyde, diazotized diamines, or
toluene diisocyanate, but the toxicity and the
high reactivity of these substances limit their ap-
plicability in immobilizing viable cells. Exam-
ples of immobilization by intercellular aggrega-
tion are given in [93], [100], [104].
Immobilized Biocatalysts 13
2.3.3.2. Binding of Cells to a Carrier
Physical Adsorption. Immobilization of
microorganisms or cells by adhesion or adsorp-
tion to a preformed matrix is based on their in-
herent affinity for solid surfaces during growth.
Such immobilization can generally be achieved
by keeping the inert carrier and the actively
growing cells in contact for a defined period.
Additional chemicals are usually unnecessary,
and fixation is carried out under growth con-
ditions; viable cell preparations can, therefore,
be obtained. Examples of cells immobilized
by adhesion or adsorption to various support
materials are listed in [100], [104].
Although the method is very simple, mild,
and versatile, its use is limited by adhesion ca-
pacity and adhesive strength, which usually de-
pend on environmental conditions (pH, ionic
strength) [93]. Use of a high flow rate as a
method of selection in a fluidized bed yields
good results. Such surface-attached cell systems
( biofilms) have been used extensively in waste-
water treatment [95], [109], [110].
A promising application of this method is the
use of mammalian cells bound to preformed sur-
faces to produce therapeutic biochemicals. The
cells are immobilized on microcarriers (small-
diameter beads, 100 – 200 µm) manufactured
from different synthetic polymers (e.g., polysty-
rene, gelatin, dextran, polyacrylamide, or glass)
that offer a large specific surface area for cell
growth [111–113]. The primary weakness of this
method is the exposure of surface-grown cells
to mechanical stress resulting from bead – bead
collisions and impeller impact. The number of
cell divisions is also restricted by the limited
outer surface area. Both problems have been
partially solved by the development of macrop-
orous microcarriers prepared from gelatin, dex-
tran, polyacrylamide, or agarose [113], [114].
Cells grow on the much larger interior surface
of the porous structure and are also protected
against mechanical stress.
A monolithic ceramic matrix has been de-
veloped for large-scale cell cultures. It consists
of a ceramic cylinder with uniform longitudinal
channels of square cross section. The matrix has
a surface area of 32 cm2 per cubic centimeter of
volume. A matrix with a very smooth surface
14 Immobilized Biocatalysts
Figure 3. Classification of cell immobilization methods
is used for the growth of anchorage-dependent
cells. Nonadherent cell types (e.g., hybridoma
cells) are cultured in a matrix with a very rough
surface and are physically entrapped in the pores
of the channel walls. Unlike microcarrier cell
cultures, no shear problems occur in the ceramic
matrix [113], [115].
For adherent cell growth, the scalability is
almost linear and depends on the available sur-
face area. For immobilized hybridoma cultures,
scaleup is based on the volume of the porous
matrix [116].
Ionic Binding. Ionic binding is a special
case of physical adsorption where charged mi-
crobial cells can electrostatically interact with
the ions on a carrier surface to form stable com-
plexes. Synthetic ion-exchange resins, modified
cellulose derivatives, or inorganic materials can
be used as carriers. Cell adsorption is mainly
affected by factors such as pH, ionic strength,
surface charge, cell age, or composition of the
carrier surface. Examples for the application of
this method are given in [93], [98].
Covalent Binding. Cells can also bind to the
functional groups of the carrier surface by cova-
lent bonds. This technique is frequently used for
enzyme immobilization but has only limited use
in whole-cell immobilization because the toxic-
ity of the coupling agents often results in loss of
cell viability or enzyme activity [66], [104].
The main advantages of this method are re-
duced or eliminated cell washout and improved
mechanical strength of the biocatalyst without
increasing diffusion barriers. The problems as-
sociated with various organic and inorganic sup-
ports, along with coupling techniques, are dis-
cussed in [104]; applications are reviewed in
[100]. The preparation, properties, and exper-
imental applications of eukaryotic cells cova-
lently bound to solid supports are described in
[117].
2.3.3.3. Immobilization of Cells by
Entrapment
Cells may be immobilized by entrapment with a
polymer lattice, a membrane, or microcapsules.
Lattice Entrapment. The widely used
method of lattice entrapment involves retention
of cells within the network of a polymer matrix.
The lattice is tight enough to prevent release of
cells, while allowing diffusion of substrates and
products. The cell does not bind to the matrix,
and the polymer network is generally formed
in the presence of the cells to be entrapped.
Hydrophilic porous matrices are formed under
mild enough conditions to allow cell entrapment

with minimum loss of viability and activity. The
prepared biocatalyst forms pellets which should
have good mechanical and chemical stability, as
well as high cell loading capacity. The high vi-
able cell density results in an enhanced demand
for nutrients and oxygen, which must overcome
the diffusion barriers of the gel structure.
In this section, only the most extensively used
matrices (e.g., polyacrylamide gel, alginate gel,
κ-carrageenan, and photo-cross-linkable resins)
are discussed. More detailed information on
other systems can be found in [21], [90].
Polyacrylamide Gel. Polyacrylamide gel
was used to immobilize whole cells in the
first successful application of lattice entrapment
used for the biotechnological production of bio-
chemicals [118]. The same technique was also
employed in the production of l-aspartic acid
[119]. The method is based on free-radical poly-
merization in an aqueous acrylamide monomer
solution containing the cells at low tempera-
ture. The degree of cross-linking determines the
porosity and mechanical properties of the pel-
lets. The major disadvantage of the method is the
toxicity of the acrylamide monomer, the cross-
linking agent (e.g., N,N -methylenebisacrylam-
ide [110-26-9]), and the polymerization initia-
tor (e.g., N,N,N ,N -tetramethylethylenediamine
[110-18-9]) which can decrease cell viability
and enzyme activity [96], [120]. Hydroxyethyl
methacrylate [868-77-9] can be used as a less
toxic alternative to acrylamide [43], [121].
With acrylamide monomers, optimal results
are achieved in isotonic buffered solution at
low temperature (4 – 12 ◦ C) by keeping expo-
sure time to a minimum (3 – 4 min). Examples of
successful applications of this method are given
in [96], [98].
Alginate [9005-38-3] is extracted from sea-
weed and is a linear copolymer of β-d-
mannuronic acid and α-l-guluronic acid linked
by 1,4-glycosidic bonds. It forms a gel in the
presence of multivalent ions, usually calcium or
aluminum. The controlled entrapment of cells
is simple and generally nontoxic. Various cell
types can be immobilized with negligible loss
of viability. However, the matrix can be solubi-
lized in the presence of Ca2+ -chelating agents
such as phosphate, citrate, or ethylenediamine-
tetraacetic acid. The alginate matrix is mechan-
ically weak so that the growing cells (especially
plant cells) can be released from, or even dis-
Immobilized Biocatalysts 15
integrate, the beads. Another drawback of the
alginate method for viable animal cells is the
difficulty of producing sufficiently small beads
to overcome oxygen limitation in their interior
[113].
Strengths and weaknesses of the calcium al-
ginate method are discussed in [122]; examples
of applications are given in [96], [98], [100].
κ-Carrageenan [11114-20-8] is a readily
available, nontoxic, high molecular mass, sul-
fonated polysaccharide extracted from seaweed.
It consists of β-d-galactose 4-sulfate and 3,6-
anhydro-d-galactose units. Only the sodium salt
of κ-carrageenan is soluble in cold water. The
gelation characteristics of κ-carrageenan are dis-
cussed in [123]. Induction of gelation with potas-
sium ions can be performed under very mild
conditions without the use of chemicals that de-
crease enzyme activity. Another advantage of
this technique is that the immobilized cell bio-
catalysts can be tailor-made into pellets, sheets,
beads, etc. [96].
Two entrapment procedures are available. In
the one-step procedure, a large number of cells
are entrapped in the gel. The cell suspension
in physiological saline is mixed with soluble κ-
carrageenan gel under sterile conditions, and the
gel beads are incubated in a nutrient medium
[98], [122]. Another very elegant method in-
volves the use of photo-cross-linkable resins
[74], [124], [125]. Active groups (e.g., vinyl
groups) are coupled to polyethylene or poly-
(propylene glycol) oligomers of controlled chain
length to prepare prepolymers. The prepolymers
are then mixed with the cell culture, and cross–
linking is initiated with UV radiation. The ad-
vantages of this method are:
1) entrapment is simple and proceeds under
very mild conditions;
2) prepolymers do not contain toxic monomers;
3) the network structure of the gels can be
adapted as required; and
4) optimal physicochemical gel properties can
be achieved by selecting suitable prepoly-
mers [125].
Membrane Entrapment. In membrane en-
trapment, the cells are not held in a porous ma-
trix (as in lattice entrapment), but are retained by
a semipermeable membrane barrier. The mem-
brane allows the diffusion of soluble material
16 Immobilized Biocatalysts
to and from the immobilized cells while retain-
ing and protecting enclosed organisms. Hollow-
fiber bioreactors, in which the organism is con-
fined on one side of the porous fiber and the sol-
uble substrate and products on the other, have
found the widest application. Higher cell densi-
ties can be achieved than with lattice-entrapped
biocatalysts [126], but growth must be con-
trolled to prevent excessive buildup of biomass
and rupture of the membrane [99]. The supply
of oxygen and nutrients to the cells and the re-
moval of carbon dioxide from the cells occur by
diffusion and may be insufficient. Model equa-
tions for mass transfer in hollow-fiber reactors
are discussed in [127], [128].
The state of the art of membrane reactors is
described in [84]. Developments in hollow-fiber
reactors are discussed in [129], and for flat-bed
hollow-fiber reactors in [130]. A flat membrane
reactor with different compartments for cells,
medium, and product for large-scale cultivation
of mammalian cells is described in [131].
Microencapsulation. In microencapsula-
tion, cells are placed inside an alginate bead
that is coated with a semipermeable membrane.
The membrane is usually composed of calcium
alginate and poly-l-lysine. After being coated
the alginate gel can be solubilized with sodium
citrate, leaving the cells entrapped in hollow
spheres where they can be cultured. Different
capsule diameters (from 20 nm to 2 mm) and
controlled membrane porosity can be achieved
[127]. In larger microcapsules, however, radial
concentration gradients of nutrients or oxygen
can result in a necrotic core [127] .
This well-known enzyme immobilization
technique (Section 2.2.2.3) was first adapted for
the whole-cell immobilization of Micrococcus
denitrificans [132] and of viable mammalian
cells [133]. Applications of this type of immo-
bilized biocatalyst [134] include the production
of monoclonal antiboides from hybridoma cells
[135].
2.4. Methods of Organelle
Immobilization
Enzymes in living cells may be associated in
supramolecular structures, which are assembled
into organelles such as mitochondria, chloro-
plasts, peroxisomes (microbodies), vacuoles,
and nuclei. Such organelles have their own dis-
tinct metabolic functions and, therefore, pro-
vide excellent immobilized biocatalysts that can
carry out multistep, multifunctional reactions.
Immobilization generally increases organelle
longevity.
Examples of the uses of immobilized or-
ganelles are adenosine 5 -triphosphate (ATP) re-
generation, metabolite oxidation, and photopro-
duction of hydrogen [136]. Most research has
been devoted to the isolation and immobilization
of chloroplasts as potential solar energy photo-
converters [136], [137]. Older work on organelle
immobilization is reviewed in [137], [138]. En-
trapment techniques using photo-cross-linkable
prepolymers are described in [125].
Immobilization of rat liver mitochondria on
a solid support consisting of alkysilanized glass
beads was first reported in 1975 [140]. Rat liver
mitochondria have also been entrapped in poly-
acrylamide gel [141]; yeast mitochondria have
been entrapped in a polyurethane matrix [142].
2.5. Co-immobilization of Biocatalysts
Co-immobilization refers to the simultaneous
immobilization of specific combinations of
cells, organelles, or enzymes that show com-
plementary biocatalytic activity. The synergis-
tic advantages of co-immobilization are, how-
ever, limited by the biological compatibility of
the components [143].
Co-immobilization can be achieved by coen-
trapment of preimmobilized enzymes and whole
cells in a single matrix or by direct coupling of
enzymes to the microorganisms [25], [143]. The
former method was used to immobilize galac-
tosidase and Saccharomyces cerevisiae for the
conversion of cellobiose to ethanol [144]. Yeast
cells or dead mycelia have been employed for
direct coupling of enzymes to cell walls in food-
processing applications [25], [143]. Another in-
teresting application is the bioproduction of hy-
drogen gas with hydrogenase and chloroplasts
[145]. Co-immobilization results in a fivefold
increase in hydrogen production compared with
the system employing two separately immobi-
lized catalytic species. Recent results on integral
 Immobilized Biocatalysts 17

hydrogen-producing biocatalysts are presented

in [139].
Further information on the state of the art of
co-immobilization, its limitations and prospects
can be found in [25], [90], [143], [146], [147].

3. Activity and Kinetics of

Immobilized Biocatalysts

3.1. Effects of Immobilization on

Enzyme Activity

Immobilization can be considered a random

binding process which may, therefore, nega-
tively influence the active site of an enzyme to
varying degrees, even in the same immobiliza-
tion procedure. A lower net specific activity of
the immobilized enzyme is observed. The degree
of activity retention depends on the immobiliza-
tion method used. Major reasons for the loss of
enzymatic activity in immobilization are listed
below and illustrated in Figure 4:
1) Enzyme may be denatured by reactants or
products involved in the immobilization pro-
cedure (entrapment matrix formation, cova-
lent binding reaction, or cross-linking reac-
tion).
2) Immobilization conditions may lead to de-
naturation or deactivation.
3) Enzyme denaturation may be caused by a re-
active group in the active site participating in Figure 4. Effects of immobilization on enzyme activity (see
the binding reaction. text for further explanation)
4) Binding forces may hold the enzyme
molecule in an inactive configuration (con- Coupling yield gives no information about
formational changes). the amount of active protein immobilized. In
5) Orientation of the bound enzyme molecule practice, however, the amount of activity re-
relative to the support may prevent substrate tained after immobilization must be known. In
access to the active site or product release the presence of mass-transfer effects, the amount
(steric hindrance). of immobilized active protein can only be de-
termined accurately by active site titration; this
Measurement of activity retention after im- method is based on stoichiometric and reversible
mobilization is difficult because of mass-trans- binding of a specific titrant. Unfortunately, it can
fer effects. It seems to decrease with increas- be applied to very few enzymes (e.g., trypsin
ing amount of enzyme bound to the matrix and [148]). Other methods aiming at the separation
increasing immobilization time. It also varies of mass transfer and kinetics are usually less
greatly with different supports. The coupling precise and require substantial experimental and
yield represents the binding capacity of a car- computational effort [149] (see also Chap. 4).
rier and is defined as
Stabilization. Immobilized enzymes and na-
tive enzymes are largely deactivated by the
18 Immobilized Biocatalysts

same effects. However, immobilization often in- Similarly, transfer of a charged substrate
creases enzyme stability, probably due to stabi- through the external liquid film to an enzyme
lization of the tertiary protein structure. located on the surface of the support may influ-
The elimination of autolytic denaturation of ence enzyme kinetics. Several expressions for
hydrolases is another beneficial effect of immo- K app
M are given in [154].
bilization, especially by covalent binding meth-
ods [150]. Some physical methods (e.g., the use
of membranes or microencapsulation) in which
the enzyme remains in the soluble form, do not
have a protective effect.
An effective application of immobilized bio-
catalysts is their use in continuous reactors. This
requires high biocatalyst stability, which can be
obtained by using appropriate immobilization
methods and modifications of the biocatalyst mi-
croenvironment. Various additives such as an-
tioxidants, salts, polyols, sugars, and other poly-
mers may be especially useful for this purpose.
Biocatalyst stabilization by immobilization or
solute addition is discussed in [151]. The effect
of cross-linking is treated in [152]. Figure 5. Electrostatic effects on pH optimum of enzyme
activity
Electrostatic Effects. An apparent shift in a) Positively charged matrix; b) Native enzyme; c) Nega-
the pH optimum of a biocatalyst immobilized tively charged matrix
on an ionic support may be observed due to mi-
croenvironmental effects caused by the charge of
the matrix. The local pH of a negatively charged Polarity. The polarity of a support is one of
matrix (in which the enzyme is located) is lower the parameters determining its affinity for the en-
than that of the bulk solution, which causes the zyme substrate. The attractive or repulsive forces
pH optimum observed in the bulk solution to alter the substrate concentration in the microen-
increase (Fig. 5, curve c). A positively charged vironment of the support (i.e., partition of sub-
matrix has a higher local pH than the bulk solu- strate), thus influencing the enzymatic reaction
tion, thus shifting the apparent pH optimum to a rate.
lower value (Fig. 5, curve a). The reaction rate of a polar substrate can be
A thermodynamic expression has been de- increased by using polar supports for immobi-
rived for the shift of the pH activity profile for an lization; hydrophobic supports increase the re-
enzyme immobilized to a charged support [153]: action rates of apolar substrates.

zF ψ Nonuniform Distribution of Enzyme in the

∆pH = pHlocal −pHbulk = 0.43
kT Matrix. Biocatalysts with nonuniform activity
where z is the valence of the substrate, F is the are of two main types:
Faraday constant, ψ is the electrostatic poten- 1) Composite catalysts consisting of uniformly
tial, k is the Boltzmann constant, and T is the active and inactive layers
absolute temperature. 2) Biocatalysts with a continuous activity pro-
Analogously, for the change in the Michaelis file. The distribution can be convex (obtained
constant K M (for definition of K M , see Sec- from incomplete adsorption in a support) or
tion 3.2): concave (obtained by incomplete back wash-
ing of a support of uniform initial profile).
app KM zF ψ
∆pKM = pKM −pKM = log app = 0.43
KM kT Numerical simulation shows that a biocat-
alyst with activity concentrated in the shell
where the superscript app denotes apparent. (convex distribution) is always more effective

for enzymatic reactions of positive order such
as Michaelis – Menten kinetics or competitive
product inhibition. A catalyst whose activity
increases toward its core (concave distribu-
tion) can give greater utilization efficiency for
substrate-inhibited reactions [155].
3.2. Kinetics of Immobilized
Biocatalysts
The rate of reaction catalyzed by a soluble en-
zyme can usually be described by a Michaelis –
Menten type of kinetic equation (→ Enzymes,
Chap. 2.). The simplest mechanism for such a re-
action assumes the reversible formation of an en-
zyme – substrate (ES) complex, which decom-
poses irreversibly into product (P) and free en-
zyme (E):
S + E
ES −→ E + P
The well-known Michaelis – Menten equa-
tion is
vmax CS
rS =
KM +CS
where r s is the reaction rate of the substrate,
vmax is the maximum reaction rate, K M is
the Michaelis constant, and C S is the substrate
concentration. The reaction rate is a function
of the substrate and enzyme concentrations,
pH, temperature, solvent composition, and ionic
strength.
The corresponding reaction mechanism for
an immobilized enzyme is more complicated be-
cause (1) the enzyme activity may be changed
by immobilization and (2) the substrate must be
transported to the support and the product must
be transported from the reaction site back into
the bulk solution:
The reaction rate is governed by additional
parameters resulting from immobilization, i.e.,
conformational changes, chemical changes, par-
titioning effects (electrostatic, hydrophobic –
hydrophilic interactions), and diffusional effects
Immobilized Biocatalysts 19
(i.e., external and internal mass transfer; see
Chap. 4).
Mass-transfer resistance must be considered
in kinetic measurements or when modeling
processes with immobilized biocatalysts. The
familiar Michaelis – Menten equation can be
adapted by using the effectiveness factor η:
vmax CSA
rS =η
KM +CSA
or in its linearized form
1 KM 1 1
= +
rS ηvmax CSA ηvmax
where C SA is the substrate concentration at the
surface. Since both the substrate concentration
at the surface C SA and the effectiveness fac-
tor η can vary with the bulk substrate concen-
tration C S , the double reciprocal Lineweaver –
Burk diagram and other linearized forms of the
Michaelis – Menten equation are no longer lin-
ear. In Figure 6 it can be seen that the influence of
mass transfer decreases at high substrate concen-
tration and disappears totally when this concen-
tration becomes infinite. The effectiveness factor
η changes with the bulk substrate concentration,
reaching its lowest value η (0) when C SA = 0.
Figure 6. Influence of mass-transfer resistance on the kinet-
ics of immobilized enzymes in a Lineweaver – Burk plot
At low substrate concentrations, the values of the appar-
ent kinetic constants are determined. At high substrate con-
centrations, the effectiveness factor η approaches unity and,
at least in theory, intrinsic kinetic parameters can be deter-
mined.
20 Immobilized Biocatalysts

3.3. Assay of Immobilized Biocatalysts phase or flocculant cell mass. Thus, mass trans-
fer through a liquid diffusion film from the bulk
Immobilized biocatalysts are usually particu- liquid to the surface of the biocatalyst may have
late or filamentous; therefore, conventional as- a significant effect on biocatalyst activity; it is
say methods for soluble enzymes can rarely be referred to as external mass transfer. Because
applied without modification. Activity can be the reaction site may be located within a gel,
measured in a stationary catalyst bed, a homoge- porous solid, biofilm, or biofloc, transfer of sub-
neous biocatalyst suspension, or a fluidized-bed strate from the exterior surface to a point within
reactor. the carrier is usually necessary. This is called in-
The stationary catalyst bed (column reac- ternal mass transfer or intra particle transport.
tor) has the disadvantages of axial substrate and Reviews on transfer effects with immobilized
product concentration gradients and the danger biocatalysts are found in [158–160].
of channeling. These can be overcome by using
a very low catalyst bed height or a high flow rate.
Homogeneous biocatalyst suspension 4.1. External Mass Transfer
(stirred-tank reactor) avoids concentration gra-
dients. However, to obtain a homogeneous sus- To reach the reaction site, the substrate is trans-
pension, intensive mixing is required and this ported from the bulk liquid by convection to the
can lead to mechanical destruction of the bio- stagnant diffusion film (point A in Fig. 7) and by
catalyst. diffusion through the film (from A to B) to the
The fluidized-bed reactor may be considered reaction site, which may be on the surface of the
a compromise between the two preceding meth- carrier (point B) or within the carrier (from B to
ods. To achieve sufficiently dense fluidized beds, C).
high-density particles and columns with internal
baffles may be needed. Advantages are a low
pressure drop across the bed, minimum particle
attrition, no channeling, and no bed clogging.
Analytical methods are discussed in [149],
[156]. They may be discrete or continuous; in
the latter case, conventional methods cannot be
applied without some modification. Most impor-
tant is to work under well-defined, reproducible
conditions.
Summarizing, immobilization is one of the
major considerations in the design of bioreac-
tors. Beside the advantages of continuous or
repetitive processes, high stability, fast reaction,
and easy separation from the reaction medium,
immobilized biocatalysts also have disadvan-
tages such as mass-transfer resistance, necessity
of immobilization procedures, and additional
costs of immobilization reagents. A comprehen-
sive review of these engineering aspects with re-
spect to reactor design is given in [157].

4. Mass Transfer in Immobilized

Biocatalyst Systems
Figure 7. Concentration profiles around and in a porous ma-
The retention of immobilized enzymes and cells trix symmetrically exposed to bulk liquid
usually requires the presence of solid carrier a) Substrate; b) Product

Consider the reaction of a substrate which is
transferred from the bulk liquid to an enzyme
immobilized on a nonporous surface (point B in
Fig. 7). For a first-order reaction (or Michaelis –
Menten kinetics with S < K M ) at steady state
(i.e., diffusion equal to surface reaction rate),
the apparent reaction rate per unit surface area
(r app ) is
rapp =kSL (SA −SB ) =ks SB (1)
where S A is the substrate concentration at point
A, S B the substrate concentration at point B, k SL
the mass-transfer coefficient for the liquid film
at the solid – liquid interface, and k s the surface
reaction rate constant. Solving for S B and rear-
ranging give
ks
rapp = ·SA (2)
(ks /kSL ) +1
Two extremes can be identified. For k s /k SL
 1, S B approaches zero (Eq. 1), and the reac-
tion is completely mass transfer controlled with
an apparent rate of k SL S A . For k s /k SL  1, S B
approaches S A , and the reaction is kinetically
controlled with an apparent rate of k s S A . In the
intermediate situation given by Equation (2), the
apparent reaction rate is influenced by both the
rate constant (k s ) and the mass-transfer coeffi-
cient (k SL ).
For a zero-order reaction,
kSL (SA −SB ) =kS (3)
where k s is a zero-order rate constant. The sub-
strate concentration at the reaction surface (S B )
is thus
SB =SA − (ks /kSL ) (4)
The ratio of the rate constant to the mass-transfer
coefficient determines S B , but does not influence
the overall rate.
For Michaelis – Menten kinetics between
zero and first order, Equation (1) becomes
rapp =kSL (SA −SB ) = vmax SB / (KM +SB ) (5)
Solving Equation (5) for S B shows that all co-
efficients influence the overall reaction rate and
also that external transfer changes the overall re-
action in such a way that it does not follow the
Michaelis – Menten form with respect to bulk
Immobilized Biocatalysts 21
concentration. Thus, the apparent kinetics of an
immobilized enzyme differ from the intrinsic ki-
netics of the native enzyme. This can be seen
most easily from Equation (5) when S B is low. In
this case, the surface rate reduces to vmax S B /K M
and the apparent rate becomes
(vmax /KM ) SA
rapp = (6)
(vmax /KM kSL ) +1
i.e., the apparent rate depends on the transfer
coefficient k SL .
Generally, true kinetics can be determined
only by operating in the kinetic regime with-
out diffusion. This can be achieved by reducing
the ratio of vmax /k SL sufficiently as seen from
Equation (6). Thus, transfer-promoting param-
eters (flow rate in packed and fluidized beds,
stirring speed in tanks) should be increased, and
those that promote reaction rate should be de-
creased (enzyme loading, temperature).
4.2. Prediction and Correlation of
External Mass-Transfer Coefficients
4.2.1. Catalyst Beds
The approximate value of k SL for a bed of
packed or fluidized particles can be predicted by
using data from the literature, which are usually
presented as dimensionless correlations having
the form

n1
n2
kSL dp dp vs  µ
∼ (7)
D µ D
where  is the liquid density, µ the viscosity, vs
the superficial linear velocity, and n1 and n2 are
empirical constants.
NSh ∼ (NRe )n1 (NSc )n2 (8)
These are referred to as Sherwood – Reynolds –
Schmidt number (N Sh , N Re , N Sc ) correlations.
Experimentally, k SL has been found to vary with
D2/3 , which on rearrangement of Equations (7)
and (8) gives the jD factor or Chilton – Colburn
correlation:

2/3
kSL µ
jD = =f (NRe ) (9)
vs D
22 Immobilized Biocatalysts

Correlations of the above form can be found
in [161–163]. For the region of low relative ve-
locity, theory and experiment show that when
the Péclet number N Pe (= d/D) is greater than
500 [165], the following holds:
being used to calculate a terminal velocity mass-
transfer coefficient k SLt [164]. Much of the data
can be approximated by
kSL = 2kSLt

kSL dp The power input of the stirrer per unit vol-

=0.99 (NRe )1/3 (NSc )1/3 =
D ume of fluid (P/V ) has been used to replace the
0.99 (NPe )1/3 (10) velocity in the term N Re , which results in the
following nondimensionless correlation [165]:
For high flow rates (N Re > 1), theory and exper-

1/4
2/3
iment give kSL dp (P/V ) µ µ
∼
D  D
NSh = 0.5 (NRe )1/2 (NSc )1/3 (11)
The individual dependencies are
Fluidized beds are more complex because the
void fraction ε also varies with velocity. This ef- kSL ∼ (P/V )1/4 , d0p , µ−5/12 , 5/12 , D2/3
fect has been taken into account by replacing vs
by vs /(1 − ε) [164]. Other data give
Dimensionless groups are thus useful in cor-
kSL ∼ (P/V )0.2 , d0p , µ−0.4 , D0.7
relating experimental data and predicting mass-
transfer coefficients. However, during experi- Another dimensionless correlation of the
mentation, measurement of the dependency of form [161]
individual operating parameters and fluid prop-
erties on the mass-transfer coefficient or on the  1/3
1/3
kSL dp (P/V ) d 4p 2 µ
overall reaction rate is of interest. Their indepen- ∼
D µ3 D
dent influence on k SL can be deduced from di-
mensionless correlations given in the literature. has been used, which gives the individual depen-
From Equation (7), dencies
kSL = 1/3
kSL ∼ (P/V )1/3 , dp , µ−2/3 , 1/3 , D2/3
 
f dn
p
1 −1 , v n1 , µn2 −n1 , D 1−n2 , n1 −n2 (12) Because P/V cannot be measured easily, the
Reynolds number of the impeller (N ReI ) is also
used, which for a given geometry in the fully
In Table 5, the values of these exponents for turbulent regime gives
stagnant flow (N Sh = 2), low velocities (Eq. 10),
and high velocities (Eq. 11) are summarized.  3/4
kSL dp N d2I 
∼
D µ

4.2.2. Particles in an Agitated Tank or

NSh ∼ (NReI )3/4

The prediction and correlation of stirred-tank
mass-transfer data are complex because of the
difficulty in defining a fluid velocity and the fact
that the particle – fluid relative velocity depends
on particle size and density. These effects have
been characterized by the terminal velocity of
a falling particle as calculated by Stoke’s law,
with
This correlation suggests that k SL is proportional
to N 3/4 , as is found for glass – enzyme particles,
whereas small gel – enzyme particles, which
have a low density difference, exhibit a much
lower sensitivity to stirrer speed, k SL ∼ N 1/5
[165].

NSh = 2.0 + 0.6 (NRe )1/2 (NSc )1/3


Table 5. Values of exponents for parameter dependency of external mass-transfer coefficients in packed columns
Variable
Stagnant flow: v = 0, N Sh = 2,
k SL = 2 D/d p
Particle diameter d p −1
Flow velocity v 0 +1/3
L 0 0
µ 0
Diffusivity D +1
4.3. Internal Mass Transfer [161], [166],
[167]
The uptake of substrates and products within
solid matrices (internal mass transfer) occurs by
diffusional transport, whose driving force is a
concentration gradient. The reaction rate within
the matrix normally depends on the substrate and
product concentrations. A mathematical model
is extremely useful for understanding and pre-
dicting the competing effects of diffusion and
reaction. The fundamentals of modeling meth-
ods are discussed below.
Consider a solid matrix, such as a membrane,
and a substrate A with an initial concentration
C A = C A0 on the left of the matrix and zero on
the right (C A = 0). This concentration difference
causes a flux of substrate A from left to right,
and the flux rate jA at any point in the matrix is
given by Fick’s law
dCA
jA = −DA−M
dZ
where DA−M is the diffusivity of A in the matrix
material. If no reaction occurs in the matrix, a
steady-state concentration gradient with a con-
stant slope is established (solid line, a, in Fig. 8).
Because the gradient is constant, the amount of
A entering any region from the left is equal to
the amount of A leaving at the right and no ac-
cumulation occurs within the matrix.
If cells or enzymes are present within the ma-
trix which convert substrate A to product, the
substrate profile is curved (curve b in Fig. 8).
This occurs because the substrate is consumed
in the reaction and the flux of substrate from the
left exceeds that leaving at the right. A mass bal-
ance can be written over a segment of the matrix.
Under steady-state conditions, the accumula-
tion is zero, and in the absence of reaction the
Immobilized Biocatalysts 23
Exponent
Low flow: N Re < 1, High flow: N Re > 1,
k SL = (2 – 5)×10−3 cm/s k SL = (10 – 40)×10−3 cm/s
−2/3 −1/2
+1/2
+1/6
0 −1/6
+2/3 +2/3
diffusion rates in and out must be equal. If reac-
tion occurs, the difference between the diffusion
rates is equal to the reaction rate. Balances of this
form can be written for any component; a com-
plete description usually requires balances for
every component participating in the reaction.
Figure 8. Flux of substrate A (initial bulk concentration
C A0 ) across a solid matrix
a) No reaction; b) Reaction
The reaction rate at each position in the im-
mobilized biocatalyst matrix is generally a func-
tion of local substrate and product concentra-
tions and thus varies from position to position.
These local rates cannot normally be measured;
only the net rate can be determined by monitor-
ing concentration changes in the bulk liquid.
Consider a situation in which the substrate
diffuses into the matrix from one side and does
not penetrate completely because it is consumed
by reaction with a biocatalyst (Fig. 9). Under
steady-state conditions, the diffusion rate of sub-
strate into the matrix equals the net reaction rate.
Thus, the flux J A multiplied by the matrix sur-
face area A can be equated to the average volu-
metric reaction rate r avg multiplied by the vol-
ume of the matrix V M
 
jA,Z=0 A=ravg VM
In both batch reactors and well-mixed contin-
uous reactors, the diffusion flux at the interface
24 Immobilized Biocatalysts

and hence the substrate profile within the bio-

catalyst are determined by the substrate concen-
tration in the reactor C A1 . In the batch reactor,
concentrations and rates vary with time.

Figure 9. Unilateral flux of substrate A and concentration

gradient in a matrix containing a biocatalyst

The inner region of a thin matrix is more ac-

cessible to substrate than that of a thick matrix
with its longer diffusion path. Thus, the effec-
tiveness of the biocatalyst is generally greatest
when the matrix thickness (or particle size) is
smallest. The effective overall reaction rate per
unit volume increases with decreasing matrix
thickness. The overall influence of diffusion is
Figure 10. Diffusion of substrate A from both sides of a
described by the effectiveness factor η: matrix
a) Complete penetration; b) Incomplete penetration
Overall reaction rate
η=
Reaction rate achieved at bulk liquid conditions

Consider the matrix again but this time with Derivation of a Finite Difference Model
substrate diffusing in from both sides (Fig. 10). for Diffusion – Reaction Phenomena. Con-
The concentration profiles are the result of com- sider the case shown in Figure 11A in which sub-
petition between reaction and diffusion. The ra- strate concentration varies from S 0 in the bulk
tio of maximum diffusion rate to maximum reac- liquid to a zero gradient at position L (a wall
tion rate gives a dimensionless parameter PD−R or the center of a symmetrical particle). The
matrix is divided into a suitably large number
jA D (CA0 /L) A of sections such that the change in concentra-
PD−R = =
rmax AL r (CA0 ) AL tion between sections is not too large. Diffusion
fluxes between the sections are portrayed in Fig-
where D is the same as DA−M and L is the matrix ure 11B.
thickness. A component mass balance is written for each
For a first-order reaction, r=kCA0 ; segment and for each component:
therefore,
D
PD−R =
kL2
For a zero-order reaction, r = k and
DCA0
PD−R = dSn
kL2 A∆Z =jn−1 A − jn A+rS A∆Z
dt
Therefore, L plays an important role. The
higher PD−R , the higher is the diffusion reac- where r S is the production rate of S by reaction
tion ratio and the smaller the substrate gradient. per unit volume of matrix.
By using Fick’s law in the form
 Immobilized Biocatalysts 25

or
(Sn−1 −Sn )
jn = DS ∂S  ∂ 2 S   
∆Z = −k1 L2 /DS S 
∂t ∂Z 2
the differential difference equation form is ob-
tained: Thus, the solution depends only on k 1 L 2 /DS ,
a dimensionless reaction – diffusion parameter
dSn (Sn−1 −2Sn +Sn+1 ) for a first-order reaction (PD−R1 ), whose square
=DS +rSn
dt ∆Z 2 is known in the literature as the Thiele modulus.
Thus, N equations are obtained, one for each For a zero-order reaction, the equation be-
segment of the catalyst matrix. This is the finite comes
difference form, which is especially suitable for ∂S  ∂ 2 S   
simulation programming. The equivalent partial = −k0 L2 /DS S0
∂t ∂Z 2
differential equation is
where k 0 L 2 /DS S 0 is the governing parameter for
∂S ∂2S a zero-order reaction (PD−R0 ).
=DS +rS
∂t ∂Z 2
The boundary conditions must be described Complete and Incomplete Substrate Pen-
separately. For the above case, dS/dZ = 0 at etration. Complete penetration of the substrate
Z = L, and S = S 0 at Z = 0. The equations for the results in a zero concentration gradient either be-
first and last elements must be written accord- cause of wall conditions at a solid support or be-
ingly. cause of the symmetry of diffusion at the center.
Incomplete penetration results in the substrate
profile shown in Figure 12. The shape of the
substrate profile is obtained by integrating the
diffusion – reaction equations.

Figure 12. Substrate profile after incomplete penetration of

Figure 11. Diffusion of a substrate from a bulk liquid in a the matrix
catalyst matrix
A) Substrate concentration; B) Diffusion fluxes between sec- Integration for a single substrate with incom-
tions shown in A plete penetration due to a zero-order reaction
gives
Dimensionless Model. The governing di- AM 
mensionless model is obtained by defining the rapp = − 2DS kS0
VM
variables to range between 0 and 1. Thus,
S 
= S/S 0 , Z = Z/L, and for time, t > = where AM and V M are the area and volume of
t/ L2 /D . the matrix, respectively. Thus, when diffusional
Substituting into the differential equation for effects lead to incomplete penetration, the zero-
the first-order case r S = − kS gives order reaction has an apparent order of 1/2, a
value midway between the first-order depen-
dS  d2 S  dency of diffusion and the zero order of the re-
S0 =DS S0 −k1 S0 S 
(L2 /D S) dt L2 dZ 2 action.
26 Immobilized Biocatalysts
Effectiveness Factor. Analytical solutions
of the general diffusion – reaction equation for
simple reaction orders are given in [161], [166].
In Figure 13 the values of η for zero, first, and
second reaction order are plotted against the
square root of the Thiele modulus ϕ where
 creased to 1 if diffusion-free enzyme or cell ki-
ϕ= L2 kS0n−1 /D
Figure 13. Dependence of effectiveness factor η on Thiele
modulus for reactions in a flat film with diffusion from one
side (adapted from [161])
a) Zero order; b) First order; c) Second order
As seen from the foregoing, the diffusion –
reaction phenomena for single reactions with
simple kinetics are governed by a single param-
eter. For a reaction of order n, the parameter is
(L 2 k S 0 n−1 /D) and is the ratio of the reaction
rate to the diffusion rate. The magnitude of the
parameter can be used to predict the influence of
diffusion on overall kinetics. A value much less
than 1 means that diffusion has no influence and
the concentration profile is flat. A value much
above 1 indicates that diffusion is important and
that the concentration gradients in the matrix are
large. For a zero-order reaction where the pa-
rameter is L 2 k/DS 0 , the overall rate would be
influenced only if the substrate were exhausted
within the porous support. A first-order reaction
(parameter L 2 k/D) is influenced by the shape of
the profile. This difference is reflected in the ef-
fectiveness factor plot of Figure 13.
For zero-, first-, and second-order, and
Michaelis – Menten kinetics, η is always less
than 1, but for substrate inhibition, the decreased
concentration inside the matrix can cause η to be
greater than 1. For zero-order kinetics, the lim-
ited penetration of substrate causes η to become
less than unity.
Experimental Strategies for Overcoming
Internal Diffusion. In practice, η must be in-
netics are to be measured. This can be achieved
by
1) decreasing the particle size d p ,
2) decreasing the enzyme or cell loading k,
3) increasing the diffusivity by immobilizing on
a support with a more open pore structure
(impossible in gels or biofilms), and
4) decreasing the temperature to decrease k.
Experimental strategy is given in Table 6,
which summarizes the dependency of the ob-
served kinetics on the experimental variables for
a reaction influenced by internal diffusion.
Table 6. Effect of internal mass transfer on the apparent kinetics of
a reaction influenced by internal diffusion
Experimental variable Observation
change
Decrease in particle r app increases; K M,app decreases;
diameter d p apparent enzyme concentration increases
Increase in initial in low S 0 range, r app increases with S 0 ;
substrate in high S 0 range, r app increases with
concentration S 0 1/2
S 0 and reaches a maximum,
K M,app > K M
Increase in flow rate no effect on r app above the point where
or mixing rate external mass transfer is eliminated
Decrease in enzyme K M,app decreases; value of d p at which
loading r app is maximum increases
Decrease in
temperature
apparent enzyme concentration increases
Modeling Complex Diffusion – Reaction
Systems. The diffusion – reaction phenomena
for complex, multireaction systems within a
biocatalyst matrix can be adequately described
by the quasi-homogeneous diffusion – reaction
model presented previously. A batch or contin-
uous biofilm nitrification reactor can be simu-
lated by combining the above model for the solid
phase with the appropriate one for the batch or
continuous liquid phases as shown in Figure 14.
Thus mass balances can be written for each
component in the bulk liquid and coupled to the
solid matrix by a mass transfer flux, js
js = −DS−M (S1 −S0 ) /∆Z

where DS−M is the diffusivity of substrate in the
matrix and S 0 is the concentration in the liquid.
The final model for numerical solution then
consists of N equations for each component in
the solid matrix plus equations for each compo-
nent in the liquid phase. Solution can easily be
made using a simulation language on a digital
computer [166].
Figure 14. Coupling of bulk liquid by the diffusion flux to
the matrix
5. Design of Immobilized Biocatalyst
Reactors
In immobilized biocatalyst reactors, the biocat-
alysts may be immobilized in or on a carrier,
immobilized by linkage among one another to
form larger particles, or confined within mem-
brane barriers [168–170].
Various reactor configurations are shown in
Figure 15. Their advantages and disadvantages
are summarized in Table 7. For purposes of mod-
eling, the tank and tubular models, together with
the external and internal transfer concepts, and
the intrinsic kinetics, provide adequate flexibil-
ity to describe all reactor types. The choice of
a particular reactor configuration depends on
many factors including
1) batch or continuous operation required;
2) immobilized biocatalyst characteristics (par-
ticle size, catalyst density, reaction rates, sta-
bility, etc.);
3) mass-transfer requirements (nutrients, oxy-
gen etc.);
4) downstream processing requirements (prod-
uct stability, undesirable byproducts);
5) shear imposed on the biocatalyst (plant and
animal cell systems);
6) energy requirements;
Immobilized Biocatalysts 27
7) control of environmental conditions (temper-
ature, pH, substrates, inhibitory and toxic
products);
8) substrate availability and waste treatment;
and
9) aseptic operation.
Most of the reactors can be run in a batch, fed-
batch, or continuous mode (fed-batch denotes
semicontinuous operation with unequal feed and
effluent streams). For batch operation, packed
and fluidized-bed systems, and membrane sys-
tems, must be run in a recycle mode.
To describe reactor systems quantitatively,
mass balances for total mass and for individual
components, as well as energy balances, must be
set up. The general form of a component balance
for component i is
For a well-mixed tank, this gives
d (V Ci )1
= (F Ci )0 − (F Ci )1 + (ri V )1 (13)
dt
where V is the volume in cubic meters, Ci is the
concentration of component i in moles per cu-
bic meter, F is the flow rate in cubic meters per
second, and ri is the reaction rate in moles per
cubic meter per second. Subscript 0 character-
izes the stream entering the reactor; subscript 1,
the stream leaving the reactor and in the reaction
because the liquid in the reactor is assumed to
be well mixed. Equation (13) may be used for a
batch, fed-batch, or continuous stirred-tank re-
actor.
5.1. Batch Stirred-Tank Reactor
For a constant-volume batch stirred-tank reactor
(Fig. 15A), the general balance given in Equa-
tion (13) becomes
dCi
=+ri (14)
dt
The parameter ri is defined as the rate per
unit volume; for substrates ri < 0 and for prod-
ucts ri > 0. In catalyst systems, the reaction rate
28 Immobilized Biocatalysts
Table 7. Advantages and disadvantages of batch and continuous reactors

Mode of Reactor type Advantages Disadvantages

operation

Batch stirred tank simple equipment for small-scale production; time-consuming downtime for loading and
aseptic operation easier cleaning; reaction conditions vary with time;
difficult to control
tubular reactors (e.g., recycle mode for kinetic studies not used for production in this mode
packed bed)
Continuous constant reaction conditions; better product requires flow-regulating equipment (pumps,
quality valves)
stirred tank good external mass transfer; easy control of usually low conversion; shear problems
environmental conditions; easy recharge or
withdrawal of biocatalyst
tubular reactors high conversion external mass transfer may be insufficient;
environmental control difficult
packed bed simple construction; high biocatalyst density high pressure drop; channeling; mechanical
strength of biocatalyst required
fluidized bed good external mass transfer; easy recharge or high flow rates; low biocatalyst density
removal of catalyst
membrane reactor separation of catalyst from substrates and mass-transfer limitations; membrane fouling;
products pressure drop often high

Figure 15. Immobilized biocatalyst reactor types

A) Batch stirred-tank reactor; B) Continuous stirred-tank reactor; C) Continuous stirred-tank membrane reactor with recycle;
D) Packed-bed column reactor; E) Fluidized-bed reactor with recycle; F) Tubular membrane reactor
The hatched area indicates the immobilized biocatalyst.
 Immobilized Biocatalysts 29

(r m,i , mol kg−1 s−1 ) is often related to the cata- The necessary residence time τ can be deter-
lyst mass mc : mined as shown in Figure 17 (hatched area).
Comparison of these results with those of Fig-
dCi rm,i mc
=+ (15) ure 16 shows that a CSTR is favorable for sub-
dt V
strate inhibition (B) and autocatalytic growth ki-
The residence time τ necessary for a cer- netics (C) at low and medium conversion (large
tain conversion can be obtained by integration values of Ci ). A tank reactor can be run in a
of Equation (14) fed-batch mode to keep the concentration of in-
hibitory substrate at a desirable level and, there-
Ci fore, obtain a performance similar to that of a
dCi
τ= (16)
ri CSTR. Batch and plug-flow reactors are favor-
Ci0 able with reaction kinetics having a reaction or-
A plot of 1/ri versus concentration gives the der n > 0 and with product inhibition kinetics.
reactor residence time necessary for the required In the case of zero-order kinetics, both reactor
conversion. Three typical cases (Michaelis – types show the same performance.
Menten kinetics, substrate inhibition, and auto- Mass transfer in CSTRs may be very high
catalytic reaction) are shown in Figure 16. The and control is simple. They are, therefore, used
hatched area represents the necessary residence when high mass transfer is required (e.g., aer-
time. ation) and control of environmental conditions
A number of applications including the pro- (e.g., pH) is critical.
duction of high-fructose corn syrup, lactase-free
milk, and 6-aminopenicillanic acid are listed in
[171]. 5.4. Packed- or Fixed-Bed Reactors

Table 8. Literature on fixed-bed biocatalyst applications

5.2. Plug-Flow Tubular Reactor
Application Biocatalyst Ref-
er-
Under steady-state conditions, an idealized ence
plug-flow tubular reactor with no axial mixing
Production of l-amino acids adsorbed aminoacylase [178]
but complete radial mixing shows the same fluid Aspartate production aspartase in immobilized [178]
effluent residence time behavior as a batch tank cells
reactor. Therefore, Equation (14) and Figure 16 Diverse applications, enzymes and cells [172]
laboratory and large-scale immobilized in gels
are also valid for a steady-state tubular reactor. Anaerobic wastewater biofilm on porous ceramic [179]
Immobilized biocatalysts are usually em- treatment
ployed in packed-bed, fluidized-bed, or mem- Drinking water denitrification whole cells in alginate [180]
Ethanol production cells on wood chips [181]
brane reactors. In some analytical applications, High-fructose corn syrup, enzyme immobilized and [192]
the biocatalyst is immobilized on the wall of the packed-bed enzyme reactor product adsorbed on zeolite
tube. These reactor types are discussed in Sec- and moving-bed adsorption
Inulin hydrolysis cells immobilized by [183]
tion 7.1. adsorption on wood
shavings
Commercial ethanol yeast cells in alginate gel [174]
production
5.3. Continuous Stirred-Tank Reactor Continuous production of enzyme on corn grits [184]
sucrose
Acid whey hydrolysis enzyme on ion-exchange [185]
In an ideal continuous stirred-tank reactor resin
(CSTR, Fig. 15B), each entering fluid element Ethanol production, modeling immobilized yeast [186]
is mixed instantaneously with the entire reactor Hydrolysis of starch, immobilized enzyme [187]
modeling
volume. The reactor operates at reactor outlet Mammalian cell culture various cell types [188],
conditions. For steady-state conditions and con- [189]
stant volume, Equation (14) gives
Ci0 −Ci1
τ= − (17)
ri1
30 Immobilized Biocatalysts

Figure 16. Graphical determination of residence time (hatched area) for given conversion and different kinetic characteristics
in a batch reactor
A) Michaelis – Menten kinetics; B) Substrate inhibition kinetics; C) Autocatalytic reaction (e.g., growth)

Figure 17. Graphical determination of residence time (hatched area) for given conversion and different kinetic characteristics
in a continuous stirred-tank reactor
A) Michaelis – Menten kinetics; B) Substrate inhibition kinetics; C) Autocatalytic reaction (e.g., growth)

Most continuous laboratory and large-scale
production units are of the packed- or fixed-
bed type (Fig. 15D) [172–174]; they are usually
tubular plug-flow reactors. In addition to axial
gradients, internal and external diffusion pro-
cesses may cause gradients at or within the bio-
catalyst. In large-scale reactors, the mechanical
strength of biocatalyst particles, pressure drop,
and control are important. Deviations from ideal
plug flow may be observed, especially if chan-
neling occurs in the bed. Applications of fixed-
bed reactors are listed in Table 8.
5.5. Fluidized-Bed Reactors
Fluidized-bed reactors (Fig. 15E) have found in-
creased application because they provide high
external mass transfer to small reactive parti-
cles consisting of immobilized cells, enzymes,
or biofilms. This is important because grow-
ing organisms would otherwise block fixed- or
packed-bed columns. Fluidized-bed reactors are
of the tubular type and models are, therefore,
similar to those for an ideal plug-flow reac-
tor. Figure 18 shows a large-scale anaerobic
fluidized-bed reactor for wastewater treatment
[175]. A minimum flow rate is required because
fluidization occurs at a minimum superficial liq-
uid velocity which depends on the density differ-
ence and the particle size. To ensure a minimum
flow rate, a certain amount of recycle is required.
The recycle ratio decreases with increasing re-
actor size. Small laboratory-scale fluidized-bed
biocatalyst reactors are, therefore, often run as
 Immobilized Biocatalysts 31

differential recycle reactors with low conversion crofiltration membranes may be used in a recy-
per passage. cle stirred-tank reactor (Fig. 15C) to retain en-
zymes, cells, or cofactors. A tubular membrane
reactor is shown in Figure 15F.
Membrane reactors may be classified accord-
ing to several criteria:
1) Membrane type (ultrafiltration, microfiltra-
tion, dialysis, liquid membrane, pervapora-
tion)
2) Biocatalyst type (enzyme, whole cells)
3) Number of liquid phases
4) Number of different membranes
5) Permeation (product or substrate)
6) Location of biocatalyst (in liquid phase or
membrane)
Figure 18. Large-scale anaerobic fluidized-bed reactor for 7) Membrane geometry (flat, tubular, spiral,
wastewater treatment etc.)
Courtesy of Gist-brocades, Delft, Netherlands.
8) Driving force for membrane transport (pres-
Literature on the application of fluidized-bed sure or concentration gradient)
biocatalyst reactors is listed in Table 9. 9) Flow of individual liquid phases (batch or
continuous)
Table 9. Literature on fluidized-bed biocatalyst applications 10) Sterility
Application Biocatalyst Refer- Applications of membrane biocatalyst are
ence
listed in Table 10.
Review on the application various types [190]
of fluidization Table 10. Literature on membrane bioreactor applications
Biological application biofilms on various [191]
supports Application Biocatalyst Refer-
Ethanol production cells in alginate [192] ence
Selection of particles flocculant biomass on [193] Mammalian cell hollow fiber reactor [204–
support particles culture 209]
2-Propanol – butanol Clostridium spp. on [194] Large-scale animal various [195]
production calcium alginate cell culture
Anaerobic wastewater biofilms [175], Hybridoma culture membrane gas supply and [210]
treatment [195], perfusion in animal cell culture
[196]
Microbial cultures hollow fiber reactor [211],
Design of fluidized-bed various types [197] [212]
reactors
Membranes and various [213]
Wastewater treatment biofilms [198] membrane processes
Animal cell culture special particles [199] Scaleup, engineering various [214],
Plant cell culture spontaneously growing [210] aspects [215]
pellets
Three-phase fluidized beds various types [201–
203]

5.7. Laboratory Reactors for Kinetic

5.6. Membrane Reactor Systems
(→ Membranes and Membrane Separations)
Membrane systems provide a means of retaining
enzymes and cells within a biological reactor.
Progress in membrane technology has produced
membrane reactor systems that can operate in
a variety of modes [176]. Ultrafiltration or mi-
Measurements
5.7.1. Batch and Continuous Tank Reactors
Both batch and continuous tank reactors lend
themselves to the determination of kinetic pa-
rameters on a laboratory scale. According to
Equation (14), the reaction rate in a batch reac-
tor is calculated from the slope of the concentra-
tion – time curve. The CSTR, on the other hand,
32 Immobilized Biocatalysts
requires measuring the steady-state difference
between outlet and inlet concentrations and the
mean residence time (Eq. 17). The CSTR usu-
ally yields better results, except when reaction
rates are very low.
In a batch experiment, substrate and product
concentrations vary with time and may be dif-
ficult to monitor and interpret. Suitable design
of experiments (e.g., variation of initial concen-
trations, supply of an excess of one component)
is, therefore, required to establish reaction rate
kinetics and determine rate constants.
At steady state the CSTR can be used to
measure rates. New steady states can be estab-
lished by changing the flow rate, inlet concen-
tration,etc.
Evaluation of experimental data may be per-
formed in serveral ways:
1) Differential equations can be integrated al-
gebraically and rearranged to give a lin-
ear function for kinetic parameters. In this
way, model applicability and model pa-
rameters may be estimated in one plot
(e.g., Lineweaver – Burk plot for Michaelis –
Menten kinetics [169].
2) If accumulation terms are equal to reaction
rates, a plot of terms versus concentration
may result in useful concentration – rate rela-
tionships and, after rearrangement, yield re-
action rate parameter values.
3) A third method uses numerical integra-
tion and linear or nonlinear parameter esti-
mation methods to check model equations
and to estimate model parameter values. A
user-friendly, numerically powerful tool is
available (SIMUSOLV [177]). Modern pro-
gram packages including excellent simula-
tion methods may help greatly in optimal de-
sign of kinetic experiments.
A short summary of experimental strategies
to determine rate parameters for complex bio-
chemical kinetics is given in [179].
5.7.2. Experimental Differential Recycle
Reactor
Reactions occurring in and on particulate solids
(e.g., in beads of immobilized enzymes or whole
cells) are generally sensitive to external mass
transfer (Section 4.1). Study of the influence of
flow rate on the overall reaction rate is, there-
fore, of interest; a convenient reactor configu-
ration is shown in Figure 19. The system has a
reactor that holds the biocatalyst particles, either
as a packed or fluidized bed. Variable flow rates
through the reactor are achieved by recycle. The
reactor is usually operated at low residence time
(high flow, small volume) so that the conversion
per pass is small. The axial gradients are then
sufficiently low for concentrations in the reactor
to be approximated by the average between inlet
and outlet. Thus, steady-state balances take the
form
0 =FR (Ci0 −Ci1 ) ±ri,avg Vr
where F R is the flow rate of the recycle stream,
ri, avg is the reaction rate at the mean concentra-
tion Ci, avg , and V r is the volume of the reactor.
The difference Ci0 − Ci1 must be large enough
for accurate measurement but small enough for
average concentration to be a valid approxima-
tion.
Figure 19. Differential recycle loop reactor for kinetic stud-
ies
a) Sampling sensor; b) Biocatalyst bed; c) Sensor and control
(pH, temperature, etc.); d) Flow meter
The recycle system shown in Figure 18 can
be operated batchwise, with no flow entering or
leaving the recycle loop and a small amount of
conversion occurring per pass. The balances be-
come
dCi
VT = ± ri Vr
dt
where V T is the volume of the total system. Cor-
responding balances for a continuously fed re-
cycle system can be set up [169].

6. Immobilization Media
Biocatalysts must be immobilized in a different
phase or physically separated from the phase that
leaves the reaction vessel and contains the prod-
ucts.
6.1. Characteristics of Solid Supports
A solid-state carrier or support is generally used
for biocatalyst immobilization in a different
phase. Because the catalytic activity of a biocat-
alyst reactor is proportional to the active enzyme
(or whole cell) concentration, carriers with high
surface area are commonly employed. Porous
solids provide areas ranging from one to sev-
eral hundred square meters per gram. However,
a high-area support is characterized by small-
diameter pores which cause diffusional limita-
tions. Very small pores prohibit the entrance of
large protein molecules, rendering at least part of
the support area inactive. This phenomenon can
be used for selective immobilization of enzymes
of different molecular mass [216]. The support
should not always be assumed to be entirely in-
ert because it may modify enzymatic properties
(e.g., conformation) and may itself exhibit par-
tition effects (adsorption, pH shift, hydropho-
bicity, etc.). Adsorption of inhibitors from the
reaction mixture may also be detrimental to bio-
catalyst activity.
Many other carrier properties, although
equally important for the function of an en-
zyme reactor, are beyond the scope of this chap-
ter. For nonporous particles, these include size,
size distribution, shape, abrasion in stirred ves-
sels or fluidized beds, apparent viscosity when
suspended in liquid, density, voidage, swelling
characteristics, and osmotic characteristics. Im-
portant properties of porous media are porosity,
permeability, pore size distribution, accessibil-
ity of inner surface for the binding of enzymes
and cells, wetting characteristics, and tortuosity.
Most of these properties and their influence on
enzyme reactor performance are treated in [217],
[218]. Physical characterization of entrapment
networks for immobilization of whole cells is
discussed in [219].
Immobilized Biocatalysts 33
6.2. Characteristics of Semipermeable
Separators
Membranes with suitable pore size can be used
as physical barriers to retain a biocatalyst in a
reactor and allow free passage of substrate and
product. The biocatalyst can be used in either a
soluble or an immobilized form. Possible con-
figurations of biocatalyst and membrane with re-
spect to flow regime are given in Figure 20. A
large number of combinations and operational
modes are conceivable based on
1) membrane type (ultrafiltration, dialysis, liq-
uid membrane),
2) biocatalyst type,
3) number of liquid phases,
4) number of different membranes,
5) permeation of product or substrate,
6) location of biocatalyst (in liquid phase, in
membrane),
7) membrane geometry (flat, tubular, spiral,
etc.), and
8) driving force for membrane transport.
7. Application of Immobilized
Biocatalysts
7.1. Analysis
Most of the applications of immobilized bio-
catalysts are analytical. The specific molecu-
lar recognition and molecular transformation of
biocatalysts are combined with a suitable trans-
ducer to allow a given property (e.g., optical)
to be monitored. Specific analysis in complex
mixtures and detection of reaction products are
thus possible. Bioanalytical devices with inti-
mate contact of biological material and trans-
ducer are called biosensors (→ Chemical and
Biochemical Sensors).
A number of analytical systems have been
designed:
1) Flow systems: biocatalyst column with sub-
sequent detection and segmented flow or flow
injection systems
2) Dipsticks and multilayer films: biocatalyst
within a thin layer with colorimetric or pho-
tometric detection
34 Immobilized Biocatalysts

Figure 20. Location of a soluble or suspended biocatalyst (A) and an immobilized biocatalyst (B) and possible transport fluxes
in a membrane reactor
Abbreviations: E = enzyme (biocatalyst); P = product; S = substrate

3) Electrodes: amperometric or potentiometric
detection
4) Optodes: optical detection by using light ab-
sorption, fluorometry, or luminometry
5) Thermistor probes: calorimetric measure-
ment of heat of reaction
6) Transistors: similar to electrodes, but semi-
conductor devices
An extensive literature exists on analysis via
immobilized biocatalysts (see Table 11).
Table 11. Reviews on analytical application of immobilized
biocatalysts
gen ions (pH), oxygen, hydrogen peroxide, and
ammonia]. Because such compounds and ions
are often involved in enzymatic reactions as re-
actants or products, these electrodes can be em-
ployed in conjunction with a suitable enzyme to
monitor substrate concentration. Important re-
views on enzymes and microbial sensors are
given in [232–237].

Title Comments Refer-

ence

Biosensors: a survey of what is commercial sensors in [220]

done and used commercially medicine, defense,
environmental analysis
Immobilized enzymes in analysis optical thin-film enzyme [221]
assays and immunoassays
Electrochemical sensors in the developments in on-line [222]
analysis and control of monitoring by biocatalyst
bioprocesses electrodes
Whole-cell biosensors whole-cell electrodes [223]
Immunosensors: antibody-based immunosystems and [224]
biosensors transducers
Analytical uses of immobilized monograph [225]
enzymes
Biosensors [226]
Applications of immobilized various systems, [227]
biocatalysts in chemical analysis diffusion effects
Immobilized whole-cell monitoring of various [228]
biocatalysts chemicals Figure 21. Enzyme electrode used in clinical analysis
Biosensors international workshop [229] a) Electrode; b) Enzyme membrane; c) Stirrer
Immobilized enzymes and cells section on analytical [230]
applications A typical setup for this procedure is shown
Chemical sensors book on chemical and [231]
biological sensors
in Figure 21. It can be used, for example, to
measure glucose concentration. A membrane-
covered oxygen electrode is employed together
with a membrane containing immobilized glu-
Enzyme Sensors for Clinical Analysis. cose oxidase [9001-37-0] (E.C. 1.1.1.49). Glu-
Electrodes are available for the measurement of cose and dissolved oxygen react to form glu-
a number of important parameters [e.g., hydro- conic acid and hydrogen peroxide. The oxygen

electrode detects oxygen consumption which is
proportional to reaction rate and thus to glucose
concentration. At low glucose levels, the result-
ing electrode signal is linearly related to this con-
centration. Alternatively, a hydrogen peroxide
electrode can be used.
Enzyme electrodes have been constructed for
at least 40 different substances, but very few
have been commercialized (Table 12). The most
important factor is enzyme stability, which is
critical for the lifetime of the enzyme electrode,
and depends largely on the immobilization pro-
cedure. Stabilities of three to four months have
often been attained.
Table 12. Examples of enzyme electrodes
Analysis Enzyme system Electrode
Alcohols alcohol oxidase H 2 O2
Amino acids l-amino acid oxidase H 2 O2
Cholesterol cholesterol oxidase O 2 , H 2 O2
Creatinine creatinase NH+ 4
Glucose glucose oxidase on poly(vinyl O 2 , H 2 O2
chloride) membrane
Sucrose invertase, mutarotase, and
glucose oxidase on collagen O2
Urea urease on Teflon NH+4
Microorganisms can also be immobilized on
the outside of the electrode. Thus, for exam-
ple, phenol can be detected by using an oxy-
gen electrode together with phenol-oxidizing
microorganisms. Stabilities of 5 – 30 d have been
achieved for over 20 compounds by using immo-
bilized organisms [237].
For certain analyses, the immobilized en-
zyme is better employed in a reactor. The sample
is passed through the reactor, and the concentra-
tion of the desired substance in the reactor efflu-
ent is then measured downstream by an appro-
priate electrode sensor. This arrangement allows
adjustment of the enzyme quantity.
7.2. Immobilized Enzymes in
Therapeutic Medicine
The ability of enzymes to carry out specific
biological reactions is being exploited as a
therapeutic method in medicine (→ Enzymes,
Chap. 5.4.). A number of applications have been
tested on animals, but very few on humans.
The enzymes require immobilization and con-
tact with body fluids and organs, either inside or
Immobilized Biocatalysts 35
outside the body (intra- or extracorporeal). Re-
view articles in this rapidly developing field are
given in [238–240].
Intracorporeal methods require immobiliza-
tion in or on microparticles or blood cell frag-
ments, or with macromolecules [238]. Distinc-
tion is made between enzymes circulated with
the blood and those used for local treatment of
specific organs. Soluble polymers and micropar-
ticles can be used for immobilization in the blood
stream. Slow release by diffusion or biodegrada-
tion from particles is employed for local use.
Enzymes introduced into the body may have
the advantage of targeting specific sites but in-
crease the danger of toxic or immune response.
The carrier and the immobilized enzyme must
exhibit a minimum of allergic and immuno-
genic effects. Fortunately, immobilization often
reduces immunogenicity. A partial list of intra-
corporeally employed enzymes, their therapeu-
tic use, and the type of carrier employed is given
in Table 13. Microencapsulation methods are
important for a variety of enzyme drug applica-
tions and have the following advantages [238],
[240]:
1) ultrathin (20 nm) synthetic membrane cap-
sules;
2) small particle size (100 µm);
3) high transfer rates;
4) biological particles (normal cells, cell ghosts,
synthetic liposomes);
5) wide range of immobilization applications
(single enzymes, enzyme mixtures, whole
cells, absorbents);
6) intra- and extracorporeal use;
7) possibility of biodegradable materials (pro-
teins, lipids);
8) wide range of administration methods (oral,
intravenous, local injection); and
9) minimal immune reactions.
Extracorporeal application requires access
to the blood stream and anticoagulant treatment
[239]. The equipment is usually a small immo-
bilized enzyme reactor connected to the patient
(much like a hemodialysis unit) and, therefore, is
not suitable for continuous treatment. To obtain
the necessary activity, a large surface area of par-
ticles (packed and fluidized beds) or membranes
(hollow-fiber cartridges) is necessary, which in-
creases the danger of blood clotting. Extracor-
36 Immobilized Biocatalysts
Table 13. Methods of immobilizing intracorporeal enzymes

Carrier Enzyme Therapy Reference

Artificial cells asparaginase cancer [238–240]

or microcapsules urease model system [238], [240]
Erythrocyte ghosts urokinase thrombosis [238]
asparaginase cancer [238–240]
Liposomes asparaginase cancer [238–240]
Soluble polymers arginase hyperargininemia [238], [239]
Poly(ethylene glycol) uricase gout [239]
Albumin, dextran glutaminase cancer [238], [239]

poreal applications, including the reactor type, spp.) and a bacterium (Hymomonas mobilis)
are given in Table 14. have been investigated.
An alternative way to obtain high cell con-
Table 14. Applications of extracorporeal enzyme therapy [239]
centrations and thus high conversion rates is
Reactor device Enzyme Therapy to immobilize the cells by adhesion or ad-
sorption on porous surfaces and entrapment in
Fluidized bed heparinase anticoagulant
and packed removal gels. Adhesion has been achieved with organic
beds of seph- bilirubin oxidase liver failure materials (wood chips, cotton fibers, sawdust,
arose beads urea cycle enzymes liver failure anion-exchange resins, and gelatin-coated glass
Hollow-fiber arginase hyperargininemia
cartridges asparaginase cancer
beads) and with porous inorganics (ceramics,
Packed bed of UDP – glucuronyl liver failure glass fibers, silica gel, and vermiculite). Produc-
agarose beads transferase tion rates in the range of 20 – 60 g L−1 h−1 or
higher have been reported. Entrapment in poly-
mer gel matrices is the most frequently inves-
tigated method. Natural polymers such as algi-
nate and carrageenan give the best results. Sta-
7.3. Ethanol Production Using ble operations over 100 d have been reported
Immobilized Cells with packed and fluidized beds, 1 – 3-h residence
times, and production rates of 15 – 50 g L−1 h−1 .
Catalyzed by the oil crisis, research activity has Flocculating cells can also be retained in
been particularly intense in the field of ethanol a vertical column with an upper sedimen-
production for fuel. Review articles on ethanol tation section and slow upflow. Operations
production using fermentation of glucose with over 30 d have given production rates of 14 –
immobilized cells are given in [241–245]; see 100 g L−1 h−1 .
also → Ethanol. Closest to industrial application is the pi-
Because raw materials constitute two-thirds lot plant of the Kyowa Hakko Kogyo Co. in
of the production costs, yield is particularly im- Japan. With a total reactor volume of 4 m3 in
portant. Another major cost is separation of the five fluidized-bed columns, 8.5 % ethanol is pro-
alcohol, usually by distillation, from water. Be- duced at a rate of 20 g L−1 h−1 . Immobilization
cause ethanol is toxic to microorganisms, its fi- is with alginate.
nal concentration is usually well below 100 g/L.
In conventional suspended-culture batch fer-
mentation, ethanol production rates are 1.8 – 7.4. Industrial Applications of
2.5 g L−1 h−1 ; rates in continuous tanks are 6 – Immobilized Biocatalysts
8 g L−1 h−1 . Suspended cell systems can be im-
proved by cell separation and recycle systems The more important industrial applications of
(production rate 30 – 40 g L−1 h−1 ), but this en- immobilized biocatalysts are summarized in Ta-
tails considerable capital and operational ex- ble 15.
pense [243]. Immobilized yeast (Saccharomyces

Table 15. Industrial applications of immobilized cells [245]
Product Cell system
Acetic acid Acetobacter on wood shavings 1880
in a trickle-filter reactor
l-Aspartic acid Escherichia coli in various gels 1973 –
High-fructose corn various systems
syrup
l-Malic acid Brevibacterium in carrageenan 1974 –
gel
Penicillin gel-entrapped Escherichia coli 1969 –
Raffinose hydrolysis pellets of Mortierella fungus
Prednisolone Curvularia and Arthrobacter
(steroid coentrapped in polyacrylamide
transformation) gel
7.4.1. Production of High-Fructose Corn
Syrup
High-fructose (42 %) corn syrup (HFCS) from
cornstarch is by far the most important prod-
uct of immobilized biocatalyst technology; the
estimate annual worldwide production for 1982
was 6×106 t on a dry basis (5.5×106 t in the
United States by 11 companies and 0.4×106 t
in Japan by 16 companies). Glucose from corn-
starch is converted to fructose with glucose iso-
merase [9055-00-9] (E.C. 5.3.1.18) in the form
of a variety of immobilized whole cells. Im-
mobilization methods include gel entrapment
with cross-linking, binding on ion-exchange
resin, flocculation with electrolytes, extrusion
of cells and cross-linking, and heat treat-
ment (→ Enzymes, Chap. 4.2.1., → Enzymes,
Chap. 4.3.; → Glucose and Glucose-Containing
Syrups, Chap. 6.2.). Developed in Japan in 1966,
the rights to the process were acquired by a U.S.
company which by 1978 had improved the glu-
cose isomerase enzyme production of Strepto-
myces by mutation by a factor of 200 [245].
The process starts with a relatively inexpen-
sive raw material, corn. Cornstarch is converted
to glucose syrup by soluble amyloglucosidase
from Aspergillus niger in a batch process. The
glucose-containing syrup is then purified and
treated by a continuous process using immo-
bilized glucose isomerase. The resulting fruc-
tose composition is typically 42 %, which is
Immobilized Biocatalysts 37
close to the theoretical maximum for the equilib-
Year in- rium (57 %). This syrup has essentially the same
troduced sweetness as sucrose sugar.
on a Mutant Streptomyces murinus cells, that are
commer-
cial scale disrupted mechanically, cross-linked with glu-
taraldehyde, flocculated, filtered, extruded to
pellets (0.3 – 1 mm), and dried [246]. Activity
decreases to 10 % in 127 d.
1976 One of the largest enzyme suppliers for this
1966 –
1974 process, the Novo Company in Denmark, gives
the following information (Table 16) [246]:
1977
Table 16. Operational details of the Novo Sweetzyme reactors
1978 [246]
1968 –
1974 Parameter Comments
1978
Biocatalyst mutant Streptomyces murinus cells, disrupted
mechanically, cross-linked with glutaraldehyde,
flocculated, filtered, extruded to pellets
(0.3 – 1 mm), and dried
Reactor Capacity 100 t/d (dry material) fixed bed with
downflow (diameter 0.95 m, height 3.1 m, total
volume 17.4 m3 )
two lines of four reactors in series (with recycle)
or parallel
5240 kg biocatalyst in all reactors, 30 kg replaced
per day, 16 day replacement
flow rate of syrup (dry basis) 100 t/d
Feed 40 % solids (95 % glucose)
composition
Product 42 % fructose, 53 % glucose, 5 %
composition oligosaccharides, 70 – 75 % total solids after
evaporation
Control cycle flow adjustment necessary due to
decreasing catalyst activity to keep fructose
product concentration constant. Flow varies
linearly from 10 to 8 m3 /h over 400 h
Catalyst activity decreases to 10 % in 127 days, catalyst
lifetime replaced alternately in the eight reactors every 16
days
Conditions reactor outlet pH 7 – 7.5, feed pH 7.5 – 8.2,
Mg : Ca ratio 15, 60 ◦ C, oxygen-free
Yield 0.44 t of fructose per tonne of glucose
Productivity 3.3 t HFCS (dry) per kilogram of catalyst
1.4 t fructose (dry) per kilogram of catalyst
Daily 100 t HFCS (dry)
production
rates
43 t fructrose
16.4 t fructose per kilogram of catalyst
5.7 t/m3 HFCS (dry)
The syrup is purified by filtration, concen-
trated from 30 % to 40 % by evaporation, treated
with activated carbon to remove soluble impuri-
ties (peptides and amino acids), and by ion ex-
change to remove calcium ions. Oxygen must
finally be removed before the syrup enters the
fixed-bed isomerization reactors. Purification is
critical for obtaining high product quality and
maximum enzyme stability.
38 Immobilized Biocatalysts
The feed stream is brought to pH 7.5 – 8 with
sodium carbonate. Magnesium, which activates
and stabilizes the enzyme, is continuously added
in the form of magnesium sulfate so that the
Mg : Ca ratio is above 15. Before entering the
reactors the syrup is passed through a heat ex-
changer and brought to 60 ◦ C.
The isomerization reactors are columns filled
with immobilized enzyme (“ Sweetzyme”) con-
sisting of microbial cells which have been cross-
linked and extruded to small pellets (see Ta-
ble 16). Since the enzyme has a limited life-
time, the reactor columns must be refilled al-
ternately with fresh biocatalyst when the activ-
ity reaches 10 % of the maximum. The contin-
ual fall in activity requires that the flow rate be
reduced to maintain the desired conversion to
fructose. After enzyme replacement the flow can
be increased. Temperature variation cannot be
used to compensate for decreased enzyme activ-
ity because higher temperatures increase activity
but lead to poorer stability and product quality.
Lower temperatures give longer catalyst life but
increase the risk of microbial infection.
Post-isomerization often involves contacting
with activated carbon and ion-exchange resin
to remove color-forming material. The syrup is
then evaporated to 70 – 75 % dry substance and
stored at 30 ◦ C.
Separation of the fructose (usually by ion ex-
change) allows unreacted glucose to be recycled
to the reactors and results in syrups containing
55 % fructose (enriched fructose corn syrup,
EFCS) and 90 % fructose (very enriched fruc-
tose corn syrups, VEFCS). The new higher fruc-
tose syrups were widely accepted by the U.S.
soft drink industry when introduced in 1976;
more than half of the sugar used in soft drinks is
HFCS.
7.4.2. Amino Acid Production
Chemically synthesized d, l-amino acids re-
quire optical resolution to obtain the biologi-
cally active l-form. For many years the Tan-
abe Seiyaku Co. processed acyl-d,l-amino acid
mixtures by stereospecific hydrolysis with sol-
uble l-amino acid acylases according to the re-
action
The d-amino acid was racemized by heating to
the d,l-form and recycled. Removal of the re-
maining enzyme required complicated purifica-
tion (→ Enzymes, Chap. 4.9.1.).
In 1969 the company introduced an immo-
bilized enzyme process based on mold amino-
acylase attached by ionic binding to DEAE-
Sephadex in a packed-bed column reactor
(Fig. 22). The l-amino acid can be separated
easily by selective crystallization from the con-
centrated product stream. The d-amino acid can
then be recycled after racemization. The produc-
tion cost of the immobilized enzyme process is
40 % lower than with the soluble enzyme pro-
cess. A number of l-amino acids are reported to
have been synthesized this way: phenylalanine,
methionine, tryptophan, and valine.
Reviews of the production of amino acids
by immobilized enzymes and whole cells are
given in [245], [247–250]. Industrial processes
are listed in Table 17. Other processes, not yet
industrialized, are listed in Table 18.
l-Aspartic Acid. In 1973 a Tanabe Seiyaku
Co. process went onstream to produce l-aspartic
acid [56-84-8] from fumaric acid and ammo-
nia by using Escherichia coli cells immobilized
in polyacrylamide gel; this was the first large-
scale industrial process using immobilized mi-
croorganisms. The catalyst had a half-life of four
months, and the activity was sufficient to pro-
duce 1800 kg of l-aspartic acid per day in a
1000-L reactor. In 1979, immobilization in car-
rageenan and treatment with glutaraldehyde in-
creased the biocatalyst half-life to 680 d and the
production to 3400 kg/d. In 1982, a strain of
E. coli with sevenfold higher aspartase activity,
specially treated to eliminate fumarase activity,
was introduced into the process.
l-Alanine. l-Alanine [56-41-7] is important
in pharmaceuticals and as a taste additive in
foods. It can be produced by decarboxylation
of l-aspartic acid with l-aspartate 4-decarbox-
ylase [9024-57-1] (E.C. 4.1.1.12).
This process was operated batchwise with
soluble enzyme by the Tanabe Seiyaku Co.
In 1982 the company industrialized a continu-
 Immobilized Biocatalysts 39

Figure 22. Flow diagram for the continuous production of l-amino acids by immobilized aminoacylase (Tanabe Seiyaku,
Japan)
a) Pump; b) Filter; c) Hot-water tank; d) Heat exchanger; e) Control panel; f) Enzyme column; g) Continuous evaporator;
h) Crystallizer; i) Separator; j) Tank for racemization; k) Flow meter

Table 17. Industrial processes for amino acids using immobilized Table 18. Amino acid syntheses with biocatalysts not yet used
biocatalysts commercially
Product Substrates Catalyst Reactor Product Substrates Biocatalyst Carrier
l-Alanine d,l-amino acid aminoacylase packed bed l-Glutamic glucose Brevibacterium collagen
racemic on (Chibata) acid flavum polyacryl-
mixtures DEAESephadex Corynebac- amide
l-Methionine terium
l-Phenylala- l-Lysine d,l-α-amino-ε- caprolactam Sephadex
nine caprolactam hydrolase and
l-Tryptophan caprolactam
l-Valine racemase
l-Tryptophan indole, pyruvate, trytophanase Sepharose
l-Aspartic fumaric acid Escherichia coli packed and ammonia
acid and ammonia in agar and column l-Tryptophan indole and serine Escherichia coli polyacryl-
carrageenan amide
l-Tyrosine phenol, pyruvate, β-tyrosinase Sepharose
l-Alanine l-aspartic acid E. coli in two packed and ammonia
carrageenan and columns in
Pseudomonas series
dacunhae in
carrageenan
40 Immobilized Biocatalysts
ous process based on carrageenan-immobilized
Pseudomonas dacunhae in a pressurized
packed-bed column. The pressure keeps the
carbon dioxide in solution, which results in a
smaller pH change and less axial mixing. The l-
aspartic acid substrate is produced by immobi-
lized E. coli cells as described previously, mak-
ing this method the first use of two different im-
mobilized cells in series (Fig. 23).
7.4.3. Environmental Applications
Mixed microbial cultures can be adapted by con-
tinuous techniques under anaerobic and aero-
bic conditions to attach and adhere to almost
any surface and to degrade a variety of wastes.
The organisms develop into a thin film termed a
biofilm, which is typically 0.1 – 0.5 mm thick.
If a large surface area is available in the reac-
tor for biofilm growth, then a high concentration
of organisms can be retained within a small re-
actor volume to give a high activity per unit vol-
ume. Retention of the biomass as a biofilm is par-
ticularly important for slow-growing mixed bac-
terial cultures, such as those involved in anaer-
obic treatment of food industry wastes, aerobic
nitrification for ammonia removal, and removal
of toxic organic wastes, such as phenol. Immobi-
lization allows the process to be operated at high
flow rates, which exceed the rates that would
otherwise wash suspended organisms out of the
reactor. Retention of the active biomass in large
quantities allows the reactor to operate at rates
up to 50 times higher than those for nonimmo-
bilized systems.
Biofilm reactors can be designed in a vari-
ety of ways, depending on whether the process
is anaerobic or aerobic. Column reactors lend
themselves to both processes. Packed beds of
solid particles (e.g., bricks, stones, or plastic me-
dia) retain organisms by entrapment and film
growth, and are known as aerobic trickling fil-
ters or anaerobic filters. Expanded or fluidized
beds use small particles (e.g., sand) which are
lifted by the upflowing liquid. The organisms
must form a tight biofilm, otherwise they would
be swept out of the reactor. A variety of other ma-
terials (plastic media, honeycomb-type packing,
wound steel wire particles, activated charcoal)
have been used for biofilm development. Impor-
tant requirements are (1) maximal surface area,
(2) absence of clogging and thus low-pressure
drop flow properties for packed beds, and (3)
sufficiently high-density particles for fluidized
beds.
Plastic foam and other light porous material
can be added to conventional activated sludge
reactors to entrap biofilms for biomass reten-
tion. Another type of reactor, the upflow anaero-
bic sludge blanket reactor, uses the tendency of
anaerobic cultures to form granules by natural
flocculation and to thus be retained in a column
reactor with slow upward flow.
Most of the above reactors are classified as
high-rate processes and have been tested on var-
ious scales. They have found limited large-scale
application in the biological treatment of indus-
trial waste. The field is still in the developmental
stage. Review articles provide more detailed in-
formation [251–255].
8. Economic Aspects
The world biotechnology market has a very high
growth rate: 25 % of the world pharmaceuti-
cal market is already covered by biotechno-
logically manufactured products. In the chem-
ical industry, this fraction is approaching 10 %.
According to market growth projections, U.S.
biotechnology industry sales increased from
about $ 400×106 in 1987 to $ 2.5×109 by 1990,
and more than $ 25×109 by 2000 [256]. The
world biotechnology market in the year 2000
was predicted to be $ 65×109 [257].
An algorithm has been developed for prelim-
inary analysis of the economic viability of an
immobilized biocatalyst production process for
high-value, intermediate-value, and commodity
products [258]. A detailed list of market values,
in dollars per kilograms, and U.S. market sizes,
in kilograms per year, of potential biotechnolog-
ical products is given in that article. Manufactur-
ing costs of immobilized cell systems are dis-
cussed along with the effects of feedstock costs
and biocatalyst characteristics (e.g., selectivity,
longevity, inherent productivity, and yield) on
process economics [258].
Immobilized Enzymes. The total market for
industrial enzymes was about $ 400×106 in
1987 [259] and is growing by about 15 % each
year. Enzyme immobilization has not initiated
 Immobilized Biocatalysts 41

Figure 23. Flow diagram for continuous production of l-aspartic acid and l-alanine by immobilized cells (Tanabe Seiyaku,
Japan)
a) Substrate solution tank; b) Pump; c) Immobilized Escherichia coli column (1000 L); d) Creptallinger; e) pH control tank;
f) Immobilized Pseudomonas dacunhae closed column (2000 L); g) Safety valve; h) Pressure control valve; i) Evaporator;
j) Flow meter

the expected revolution in the enzyme industry
[260]. Soluble use-and-discard enzymes have,
by far, the most dominant market share. Only
one immobilized enzyme product, immobilized
glucose isomerase, is used in amounts > 50 t
a year; 1500 – 1750 t are used worldwide an-
nually [261]. The three other major immobi-
lized enzymes are aminoacylase (about 5 t/a),
lactase (< 5 t/a), and penicillin G acylase (3 –
4 t/a) [261]. Other immobilized enzymes are
also available but are used in even smaller
amounts: glucoamylase, hydantoinase, inver-
tase, nitrilase, RNAse, and penicillin V acylase
[260].
Many immobilized enzyme biocatalysts are
developed and manufactured by the company
that uses them. Most economic data for such
processes are confidential; published data are ex-
tremely sparse. A theoretical economic analysis
of the advantages and disadvantages of a contin-
uous immobilized enzyme process, compared to
a batchwise soluble enzyme system, is presented
in [262].
Immobilized microorganisms are pre-
ferred biocatalysts for processes with multiple
enzyme reactions especially where cofactor re-
generation is essential. However, ascertaining
which processes have been industrialized is dif-
ficult; in many cases, whether immobilized cells
or enzymes are used is unclear [263].
Recombinant Escherichia coli can be used
for the production of relatively simple pep-
tide hormones with pharmaceutical applications
such as human insulin, human growth factor,
or human interferons (→ Genetic Engineering,
Chap. 5.1., → Genetic Engineering, Chap. 5.2.).
42 Immobilized Biocatalysts
An immobilized biocatalyst is reportedly used
for the conversion [264]. A hepatitis B surface
antigen vaccine prepared in recombinant yeast
was approved for marketing, and many other
products from genetically engineered microor-
ganisms and yeast are in the clinical test stage
(interferons, lymphokines, hormones, enzymes,
and vaccines) [265].
Immobilized Mammalian Cells. Many
complex proteins of pharmacological interest
have precise folding and glycosylation require-
ments that cannot be fulfilled in simple microor-
ganisms. By using hybridoma or recombinant
DNA techniques, mammalian cells can often be
persuaded to produce and secrete useful quanti-
ties of these compounds [265].
Tissue plasminogen activator (TPA) is one
of the latest products of this technology. It dis-
solves blood clots that cause heart attacks and
can be produced in recombinant E. coli or yeast
or in glycosylated form in mammalian cells. The
first three months’ sales in the United States in
1987 were almost $ 60×106 . Worldwide sales
are projected to reach about $ 500×106 in the
next few years [266].
Monoclonal antibodies represent another
major group of genetically engineered mam-
malian cell products. More than 100 mono-
clonal antibodies are offered, with annual sales
of nearly $ 500×106 [266].
Generally, the economics of microbial sys-
tems are more favorable than those of mam-
malian cell culture systems for recombinant pro-
teins. When the use of recombinant microbial
systems is impossible or impractical, cell cul-
ture systems will gain a market position if the
products are high-value biologicals (> $ 100/g)
[265].
Whether immobilized cell processes will play
an important role in classical fermentation is still
an open question. Large-scale cell culture tech-
nologies are discussed in [267], [268]. Immobi-
lized cell cultures using hollow fibers, ceramic
matrices, or microcarriers and microencapsu-
lated cell systems for industrial processes are be-
ing developed. β-Interferon was produced from
human diploid fibroblasts grown on microcarri-
ers at a scale > 1000 L. Anchorage-independent
hybridoma cells can be immobilized in perfusion
systems [269]. Calculation of reactor productiv-
ity and process economics for large-scale cell
cultures is described in [268].
Immobilized Plant Cells. Plant cell cul-
tures can also be used for the production of
metabolites such as pharmaceuticals, chemi-
cals, flavors, and fragrances. The first prod-
uct obtained from mass plant cell cultures was
shikonin [517-89-5], a red pigment composed
of eight naphthoquinone molecules. Shikonin is
produced by a two-stage fermentation process
and is a high-value chemical ($ 4000/kg) with
a limited annual market capacity of ca. 15 kg
[270].
Immobilized plant cell systems will be used
mainly for products of cells in the station-
ary growth phase. The release of intracellularly
stored products by intermittent permeabilization
of immobilized cells can be a great economic
advantage, allowing reutilization of the biomass
[271]. The continuous immobilized plant cell
process in combination with strain selection and
improved product leakage allows production of
plant-derived chemicals in the range of $ 20 –
25/kg [272]. However, in the present industrial
state of technology for plant cell cultures, a
relatively small number of products have both
high value per weight and sufficient market size
[273].
9. Safety, Environmental, and Legal
Aspects
9.1. Immobilized Enzymes
Although enzymes are found throughout nature,
they cannot always be regarded as harmless. En-
zymes do not show systemic toxicity in animal
tests and have not been reported to be muta-
genic, carcinogenic, or teratogenic [274]. How-
ever, an enzyme can cause allergic reactions in
humans; proteolytic enzymes such as papain can
be deleterious to skin or mucous membranes.
Futhermore, commercial enzyme preparations
contain foreign materials (e.g., additives, stabi-
lizers). In the safety evaluation of commercial
enzyme preparations three major areas should
be considered:
1) catalytic activity of the pure enzyme,

2) allergenic reactions produced by the enzyme
itself or by other proteins in the preparation,
and
3) presence of toxic metabolites such as myco-
toxins or antibiotics.
The preparation and use of industrial en-
zymes are subject to a variety of national laws
and regulations. Generally, regulations concern-
ing biocatalysts are determined primarly by the
end use of the enzyme, such as food-processing,
agricultural, or medical applications.
Safety aspects of enzyme preparations in
food processing were considered by the Joint
FAO/WHO Expert Committee on Food Addi-
tives (JECFA) in several sessions. The commit-
tee laid down criteria for evaluating enzymes
according to the source material [275]. En-
zymes obtained from animal, plant, or micro-
bial sources commonly used as food, or used in
the preparation of food are accepted as foods if
satisfactory chemical and microbiological spec-
ifications can be established. No toxicological
studies are required for such enzymes. For en-
zymes obtained from nonpathogenic microor-
ganisms commonly found as contaminants in
food, JECFA recommended short-term toxic-
ity experiments. More extensive toxicological
studies, as well as chemical and microbiologi-
cal specifications, are required for enzymes ob-
tained from less well-known microorganisms
[276].
In the United States and Canada, enzymes
from microbial sources can be marketed only if
they are produced according to the guidelines for
“good manufacturing practices” (GMP) [277]
and the U.S. Food Chemicals Codex (FCC)
[278]. In many other countries, regulations are
based on these guidelines or on the JECFA rec-
ommendations.
In the United States, enzymes used in food,
animal feed, medical devices, and diagnostic
systems, as well as enzyme-based drugs and bio-
engineered food products, are regulated by the
Food and Drug Administration (FDA). Gener-
ally, any product that is considered food can be
marketed without undergoing premarket clear-
ance if it is generally recognized as safe (GRAS)
[279]. Some enzymes are assigned to this cate-
gory. Enzymes that are not GRAS are subject
to approval as food additives and must undergo
comprehensive safety testing before being sub-
Immobilized Biocatalysts 43
mitted to the FDA. Lists of enzymes permitted in
food manufacture in the United States, Canada,
and the United Kingdom can be found in [276].
The carrier and immobilization techniques
used for immobilized enzymes must also be
studied in safety evaluations [275]. The U.S.
regulations require that the fixing agents con-
sist of GRAS substances or that the immobi-
lized enzyme preparation be washed thoroughly
to remove fixing-agent residues. The only two
fixing agents approved for the preparation of
immobilized glucose isomerase in high-fructose
corn syrup production are diethylaminoethyl
cellulose (DEAE cellulose) and glutaraldehyde
[280]. Examples of pending or unsuccessful pe-
titions for GRAS affirmation are given in [280],
[281]. Cross-linking agents are highly reactive
and often toxic. Many of them lack sufficient
toxicity data and are not approved for use in
food. On the other hand, the use of nontoxic
natural matrix materials such as κ-carrageenan,
alginate, or monomer-free prepolymers for en-
zyme entrapment seems to be a feasible way of
producing safe bioctalyst preparations for food
processing.
The safety program required to obtain af-
firmation of GRAS status for immobilized
Lipozyme through scientific procedures is pre-
sented in [281]. Lipozyme is a lipase from Mu-
cor miehei that is immobilized on a macroporous
anion-exchange resin; it is used for modification
of edible fats and oils. The information required
to obtain official approval of immobilized en-
zymes should be presented by enzyme producers
to the authorities as discussed in [282].
As long as the materials used for immobiliz-
ing enzymes are toxicologically acceptable or
do not to leak measurably into the substrate or
reactor effluent streams, the use of immobilized
enzymes represents a very safe technique [274].
9.2. Immobilized Cells
A major area of risk assessment in biotechnol-
ogy is associated with the use of living cells,
especially if they are genetically engineered.
Extensive public discussion has focused on the
possible danger to humans and the environment
from the intended or accidental release of re-
combinant microorganisms (→ Genetic Engi-
neering, Chap. 6.).
44 Immobilized Biocatalysts
The safety of mammalian cell products used
in medicine is discussed in [283], [284]. Bio-
chemical purity and the hypothetical risk of
transmitting biologically active agents to prod-
uct recipients should be considered. The risk of
contamination by adventitious agents such as
fungi, bacteria, and viruses, and the putative risk
of tumorigenicity can be eliminated by effec-
tive process design, characterization studies, and
routine biochemical and microbiological testing
[284].
Substantial experimental evidence indicates
that even if mammalian tumor cells are intro-
duced accidentally into a human, tumor forma-
tion does not occur because the foreign cells
are rejected and destroyed. Only large inoc-
ula (> 106 cells) can initiate tumor formation
in immunosuppressed rodents [283]. Scientific
methods are available to assure that recombinant
products derived from continuous mammalian
cell lines are safe for use as human therapeutics
[284].
Because the highly selected mammalian or
plant cell lines used in fermentation cannot sur-
vive in the environment without optimized cul-
ture media, accidental release of such cells does
not present a serious problem. However, unin-
tended release can become a central problem
with genetically engineered microorganisms.
In industrial processes employing microorgan-
isms, nonpathogens are preferred. If the use of
pathogens is unavoidable, they should be phys-
ically and biologically contained to prevent or
limit their release. Safety limits depend on the
risk class of microorganism involved; a risk clas-
sification list has been drawn up by the Safety in
Biotechnology Working Party of the European
Federation of Biotechnology (EFB) [285]. Con-
tainment is achieved by employing good labora-
tory practice, appropriate equipment, and spe-
cial design. Immobilization should contribute
positively to this.
For recombinant DNA work, large-scale in-
dustrial processes are arbitrarily considered to
involve culture volumes > 10 L and are sub-
ject to special regulations. Guidelines have been
established by the U.S. National Institutes of
Health (NIH) [286], [287] and by the Organi-
zation for Economic Cooperation and Develop-
ment (OECD) [284], [288], [289].
These guidelines conclude that the hazards
associated with recombinant DNA microorgan-
isms can be assessed and managed in a way
similar to that used for other organisms. For
EFB Class 1 microorganisms (nonpathogens)
free of adventitious agents and having limited
capacity for survival without adverse conse-
quences in the environment, the use of good in-
dustrial largescale practice (GILSP) is allowed.
For Classes 2 and 3 with higher risks, more strin-
gent rules must be obeyed [287], [289].
With immobilized living cell biocatalysts,
safety problems exist only in the case of leakage
or escape. To prevent such escape, any waste gas
should be sterilized by filtering and heat treat-
ment if necessary. Contaminated effluents must
be inactivated by chemical or physical means,
and accidental spills must be collected. Biotech-
nological operations can be performed with the
same level of safety as microbiological labora-
tory work if good microbiological techniques
(GMT) are applied [289].
The history of the development of govern-
ment regulation of recombinant DNA research
and manufacturing processes in the United
States is reviewed in [290], [291]. Efforts to co-
ordinate scientific review procedures, approval
policies, and rulings have resulted in a “coordi-
nated framework” [292] that defines the opera-
tion of a flexible case-by-case risk assessment
[293].
The fundamental question in work with re-
combinant DNA is whether genetically engi-
neered organisms pose environmental risks that
require entirely new, specific regulations. Over-
all consensus has been reached that the nature
of the organism to be released is what matters,
not the method by which it was produced [294],
[295].
In Europe, an EEC commission is planning
to announce a series of biotechnology directives
that define the levels of physical and biological
containment and accident control, as well as pro-
vide the basis for authorizing the planned release
of genetically engineered organisms into the en-
vironment [295]. The European Committee on
Regulatory Aspects of Biotechnology (ECRAB)
has suggested a stepwise framework for risk
analysis in biotechnological processes from ini-
tiation to commercialization [296].

10. References
1. E. S. Kempner, J. H. Miller, Exp. Cell Res. 51
(1968) 141.
2. K. Mosbach, B. Matthiasson in K. Mosbach
(ed.): Methods of Enzymology, vol. 44,
Academic Press, New York 1976, p. 453.
3. “Multistep Enzyme Systems and Coenzymes,”
K. Mosbach (ed.): Methods of Enzymology,
Sect. 1, vol. 136, Academic Press, New York
1987.
4. H. Heukelekian, N. C. Dondero: Principles
and Applications in Aquatic Microbiology,
Wiley-Interscience, New York 1964.
5. J. M. Nelson, E. G. Griffin: “Adsorption of
Invertase,” J. Am. Chem. Soc. 38 (1916)
1109 – 1115.
6. I. Chibata, T. Tossa in “Immobilized Microbial
Cells,” I. Chibata, L. B. Wingard, Jr. (eds):
Applied Biochemistry and Bioengineering,
vol. 4, Academic Press, New York 1983,
pp. 1 – 9.
7. J. B. Sumner, Science 108 (1948) 410.
8. F. Micheel, J. Ewers, Macromol. Chem. 3
(1949) 200.
9. D. H. Campbell, E. Luescher, L. S. Lermann,
Proc. Natl. Acad. Sci. USA 37 (1951) 575.
10. N. Grubhofer, L. Schleith,
Naturwissenschaften 40 (1953) 508.
11. I. Chibata et al., Ferment. Technol. Today Proc.
Int. Ferment. Symp., 4th, Kyoto 1972, p. 383.
12. L. B. Wingard Jr. (ed): Enzyme Engineering,
vol. 1, Wiley-Interscience, New York 1972.
13. K. L. Smiley, G. V. Strandberg, Adv. Appl.
Microbiol. 15 (1972) 13 – 38.
14. I. Chibata, T. Tosa, T. Sato, Appl. Microbiol.
27 (1974) 878.
15. B. Arkles, W. S. Brinigar, J. Biol. Chem. 250
(1975) 8856.
16. G. R. Lambert, A. Daday, G. D. Smith FEBS
Lett. 101 (1979) 125.
17. P. Brodelius, B. Deus, K. Mosbach, M. H.
Zenk, FEBS Lett. 103 (1979) 93.
18. G. A. Rechnitz, M. A. Arnold, M. E.
Meyerhoff, Nature (London) 278 (1979) 466.
19. S. Nishikawa, Y. Ashina, I. Watanabe Kagaku
Kogaku 51 (1987) 423.
20. K. Mosbach (ed.): Immobilized Enzymes,
vol. 44, Academic Press, New York 1976.
21. K. Mosbach (ed): Immobilized Enzymes and
Cells, Part B, Methods in Enzymology,
vol. 135, Academic Press, Orlando 1987.
22. K. Mosbach (ed.): Immobilized Enzymes and
Cells, Part C, Methods in Enzymology,
vol. 136, Academic Press, Orlando 1987.
Immobilized Biocatalysts 45
23. L. B. Wingard Jr., E. Katchalski-Katzir, L.
Goldstein (eds.): Immobilized Enzyme
Principles, Applied Biochemistry and
Bioengineering, vol. 1, Academic Press, New
York 1976.
24. I. Chibata: Immobilized Enzymes, Research
and Development, Kodansha, Tokyo 1978.
25. W. Hartmeier: Immobilisierte
Biokatalysatoren, Springer Verlag, Berlin
1986.
26. R. A. Messing (ed.): Immobilized Enzymes for
Industrial Reactors, Academic Press, New
York 1975.
27. M. D. Trevan: Immobilized Enzymes:
Introduction and Applications in
Biotechnology, J. Wiley, New York 1980.
28. A. Wiseman (ed.): Handbook of Enzyme
Biotechnology, 2nd ed., Ellis Horwood,
Chichester UK 1985.
29. J. Woodward (ed.): Immobilized Cells and
Enzymes – A Practical Approach, IRL Press,
Oxford UK 1985.
30. O. R. Zaborsky: Immobilized Enzymes, CRC
Press, Cleveland, Ohio 1973.
31. L. B. Wingard Jr. (ed.): Enzyme Engineering,
vol. 1, Wiley-Interscience, New York 1972.
32. E. K. Pye, L. B. Wingard Jr. (eds.): Enzyme
Engineering, vol. 2, Plenum Press, New York
1974.
33. E. K. Pye, H. H. Weetall (eds.): Enzyme
Engineering, vol. 3, Plenum Press, New York
1978.
34. G. B. Broun, G. Manecke, L. B. Wingard Jr.
(eds.): Enzyme Engineering, vol. 4, Plenum
Press, New York 1978.
35. H. H. Weetall, G. P. Royer (eds.): Enzyme
Engineering, vol. 5, Plenum Press, New York
1980.
36. I. Chibata, S. Fukui, L. B. Wingard Jr. (eds.):
Enzyme Engineering, vol. 6, Plenum Press,
New York 1982.
37. A. I. Laskin, G. T. Tsao, L. B. Wingard Jr.
(eds.): Enzyme Engineering, vol. 7, Ann. N.Y.
Acad. Sci. No. 434, New York 1984.
38. A. I. Laskin, K. Mosbach, D. Thomas, L. B.
Wingard Jr. (eds.): Enzyme Engineering 8,
Ann. N.Y. Acad. Sci. No. 501, New York 1987.
39. P. B. Poulsen, Enzyme Microb. Technol. 5
(1983) 304 – 307.
40. A. M. Klibanov, Science (Washington, D.C.)
219 (1983) 722 – 727.
41. J. F. Kennedy, J. M. S. Cabral in W. H. Scouten
(ed.): Solid Phase Biochemistry, J. Wiley &
Sons, New York 1983, pp. 253 – 391.
46 Immobilized Biocatalysts
42. Y. Y. Linko. P. Linko in J. Tramper, H. C. van
der Plas, P. Linko (eds.): Biocatalysis in
Organic Synthesis, Elsevier, Amsterdam
1985, pp. 159 – 178.
43. A. Rosevear, J. Chem. Technol. Biotechnol.
Biotechnol. 34 B (1984) 127 – 150.
44. J. Tramper in W. H. Scouten (ed.): Solid Phase
Biochemistry, J. Wiley & Sons, New York
1983, pp. 393 – 477.
45. J. F. Kennedy, J. M. S. Cabral in H.-J. Rehm,
G. Reed (eds.): Biotechnology, A
Comprehensive Treatise in 8 Volumes, vol. 7 a,
VCH Verlagsgesellschaft, Weinheim 1987,
pp. 347 – 404.
46. W. H. Scouten in K. Mosbach (ed.): Methods
in Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 30 – 65.
47. R. A. Messing (ed.): Immobilized Enzymes for
Industrial Reactors, Academic Press, New
York 1975.
48. R. A. Messing in M. Moo-Young (ed.):
Comprehensive Biotechnology, vol. 2,
Pergamon Press, Oxford 1985, pp. 191 – 201.
49. J. F. Kennedy, C. A. White in A. Wiseman
(ed.): Handbook of Enzyme Biotechnology,
2nd ed., Ellis Horwood, Chichester UK 1985,
pp. 147 – 207, 1985 a.
50. J. F. Kennedy, C. A. White in A. Wiseman
(ed.): Handbook of Enzyme Biotechnology,
2nd ed., Ellis Horwood, Chichester UK 1985,
pp. 380 – 420, 1985 b.
51. A. N. Emery, J. S. Hough, J. M. Novais, T. P.
Lyons, Chem. Eng. (London) 259 (1972)
71 – 76.
52. J. M. Novais: Ph. D. Thesis, University of
Birmingham 1971.
53. J. F. Kennedy, Chem. Soc. Rev. 8 (1979)
221 – 257.
54. J. F. Kennedy, J. M. S. Cabral in J. Woodward
(ed.): Immobilized Cells and Enzymes – A
Practical Approach, IRL Press, Oxford UK
1985, pp. 19 –37.
55. F. R. Bernath, K. Venkatasubramanian in A. L.
Demain, N. A. Solomon (eds.): Manual of
Industrial Microbiology and Biotechnology,
American Society for Microbiology,
Washington D.C. 1986, pp. 230 – 247.
56. J. Iqbal, M. Saleemuddin, Enzyme Microb.
Technol. 4 (1982) 175 – 178.
57. K. Ikura et al., J. Appl. Biochem. 6 (1984)
222 – 231.
58. B. Solomon, R. Koppel, E. Katchalski-Katzir,
Biotechnol. Bioeng. 28 (1986) 1213 – 1221.
59. B. Solomon, Z. Hollander, R. Koppel, E.
Katchalski-Katzir in K. Mosbach (ed.):
Methods in Enzymology, vol. 135, Academic
Press, Orlando 1987, pp. 160 – 170.
60. M. Marek, O. Valentova, J. Kas, Biotechnol.
Bioeng. 26 (1984) 1223 – 1226.
61. G. P. Royer in K. Mosbach (ed.): Methods in
Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 141 – 146.
62. R. Axen, J. Porath, S. Ernbaeck, Nature
(London) 214 (1967) 1302 – 1304.
63. J. Kohn, M. Wilchek in W. H. Scouten (ed.):
Solid Phase Biochemistry, J. Wiley & Sons,
New York 1983, pp. 599 – 630.
64. I. Ugi, Angew. Chem. Int. Ed. Engl. 1 (1962)
8 – 21.
65. J. E. Prenosil, T. Hediger, Biotechnol. Bioeng.
31 (1988) 913 – 921.
66. I. Chibata, T. Tosa, T. Sato, J. Mol. Catal. 37
(1986) 1 – 24.
67. D. Jaworek in E. K. Pye, L. B. Wingard Jr.
(eds.): Enzyme Engineering, vol. 2, Plenum
Press, New York 1974, pp. 105 – 114.
68. D. Jaworek, H. Botsch, J. Maier in K.
Mosbach (ed.): Methods in Enzymology,
vol. 44, Academic Press, New York 1976,
pp. 195 – 201.
69. L. D’Angiuro et al., Biotechnol. Bioeng. 22
(1980) 2251 – 2272.
70. P. Cremonesi, L. D’Angiuro, Biotechnol.
Bioeng. 25 (1983) 735 – 744.
71. L. D’Angiuro, C. de Lalla, P. Cremonesi,
Biotechnol. Bioeng. 27 (1985) 1548 – 1553.
72. P. Bernfeld, J. Wan, Science (Washington,
D.C.) 142 (1963) 678 – 679.
73. K. Kawashima in K. Mosbach (ed.): Methods
in Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 146 – 153.
74. S. Fukui, A. Tanaka, Adv. Biochem. Eng.
Biotechnol. 29 (1984) 1 – 33.
75. S. Fukui, A. Tanaka, T. Iida in C. Laane, J.
Tramper, M. D. Lilly (eds.): Biocatalysis in
Organic Media, Elsevier, Amsterdam 1987,
pp. 21 – 41.
76. D. Dinelli, Process Biochem. 7 (1972) 9 – 12.
77. D. Dinelli et al. in E. K. Pye, H. H. Weetall
(eds.): Enzyme Engineering, vol. 3, Plenum
Press, New York 1978, pp. 477 – 481.
78. T. M. S. Chang in K. Mosbach (ed.): Methods
in Enzymology, vol. 44, Academic Press, New
York 1976, pp. 201 – 218.
79. H. E. Klein, D. W. Sundstrom, D. Shim in J.
Woodward (ed.): Immobilized Cells and
Enzymes – A Practical Approach, IRL Press,
Oxford UK 1985, pp. 49 – 54.
80. Y. T. Tu, T. M. S. Chang, Enzyme Microb.
Technol. 4 (1982) 327 – 331.

81. P. L. Luisi et al. in [37] , p. 549.
82. P. L. Luisi, L. J. Magid, CRC Crit. Rev.
Biochem. 20 (1986) no. 4, 409 – 474.
83. R. Wichmann, C. Wandrey, A. F. Buchmann,
M. R. Kula, Biotechnol. Bioeng. 23 (1981)
2789 – 2802.
84. V. G. Gekas, Enzyme Microb. Technol. 8
(1986) 450 – 460.
85. T. Takeuchi, K. Makino, Biotechnol. Bioeng.
29 (1987) 160 – 164.
86. M. Fujimura, T. Mori, T. Tosa, Biotechnol.
Bioeng. 29 (1987) 747 – 752.
87. I. Chibata, L. B. Wingard Jr. (eds.):
Immobilized Microbial Cells, Applied
Biochemistry and Bioengineering, vol. 4,
Academic Press, New York 1983.
88. B. Mattiasson (ed.): Immobilized Cells and
Organelles, vol. 1, 2, CRC Press, Boca Raton,
Florida 1983.
89. C. Webb, G. M. Black, B. Atkinson (eds.):
Process Engineering Aspects of Immobilized
Cell Systems, The Institution of Chemical
Engineers, Rugby UK 1986.
90. J. Tampion, M. D. Tampion: Immobilized
Cells: Principles and Applications,
Cambridge University Press, Cambridge 1987.
91. M. Moo-Young (ed.): Bioreactor Immobilized
Enzymes and Cells, Fundamentals and
Applications, Elsevier Applied Science,
London 1988.
92. S. Birnbaum, P.-O. Larsson, K. Mosbach in
W. H. Scouten (ed.): Solid Phase Biochemistry,
J. Wiley & Sons, New York 1983,
pp. 679 – 762.
93. J. Klein, K. D. Vorlop in M. Moo-Young (ed.):
Comprehensive Biotechnology, vol. 2,
Pergamon Press, Oxford 1985, pp. 203 – 224.
94. J. Tramper, Trends Biotechnol. 3 (1985)
45 – 50.
95. S. F. Karel, S. B. Libicki, C. R. Robertson,
Chem. Eng. Sci. 40 (1985) 1321 – 1354.
96. I. Chibata, T. Tosa, T. Sato in A. L. Demain,
N. A. Solomon (eds.): Manual of Industrial
Microbiology and Biotechnology, American
Society for Microbiology, Washington D.C.
1986, pp. 217 – 229.
97. P. K. Robinson, A. L. Mak, M. D. Trevan,
Process Biochem. 21 (1986) 122 – 127.
98. P. Brodelius, E. J. Vandamme in H.-J. Rehm,
G. Reed (eds.): Biotechnology, A
Comprehensive Treatise in 8 Volumes, vol. 7 a,
VCH Verlagsgesellschaft, Weinheim 1987,
pp. 405 – 464.
99. C. D. Scott, Enzyme Microb. Technol. 9 (1987)
66 – 73.
Immobilized Biocatalysts 47
100. C. Akin in G. E. Russell (ed.): Biotechnology
and Genetic Engineering Reviews, vol. 5,
Intercept, Wimborne UK 1987, pp. 319 – 367.
101. P. Brodelius in M. Moo-Young (ed.):
Bioreactor Immobilized Enzymes and Cells,
Fundamentals and Applications, Elsevier
Applied Science, London 1988, pp. 167 – 196.
102. J. Klein in M. Moo-Young (ed.): Bioreactor
Immobilized Enzymes and Cells,
Fundamentals and Applications, Elsevier
Applied Science, London 1988, pp. 1 – 8.
103. J. Klein, F. Wagner in I. Chibata, L. B. Wingard
Jr. (eds.): Immobilized Microbial Cells,
Academic Press, New York 1983, pp. 11 – 51.
104. J. F. Kennedy, J. M. S. Cabral in I. Chibata,
L. B. Wingard Jr. (eds.): Immobilized
Microbial Cells, Academic Press, New York
1983, pp. 189 – 280.
105. B. Mattiasson in B. Mattiasson (ed.):
Immobilized Cells and Organelles, vol. 2,
CRC Press, Boca Raton, Florida 1983,
pp. 23 – 40.
106. F. B. Kolot, Process Biochem. 16 (1981) no. 5,
2 – 9.
107. F. B. Kolot, Process Biochem. 16 (1981) no. 6,
30 – 33.
108. M. E. Hegglin: Ph. D. Thesis no. 8628, Swiss
Federal Institute of Technology, Zürich 1988.
109. H. Tanaka, S. Uzman, I. J. Dunn, Biotechnol.
Bioeng. 23 (1981) 1683 – 1702.
110. I. J. Dunn in K. Mosbach (ed.): Immobilized
Enzymes and Cells, Part B, Methods in
Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 300 – 307.
111. A. L. van Wezel, Nature (London) 216 (1967)
64 – 65.
112. J. M. Clark, M. D. Hirtenstein, J. Interferon
Res. 1 (1981) 391 – 400.
113. K. Nilsson, Trends Biotechnol. 5 (1987)
73 – 78.
114. K. Nilsson, F. Buzaky, K. Mosbach,
Bio/Technology 4 (1986) 989 – 990.
115. B. K. Lydersen et al. in J. Feder, W. R. Tolbert
(eds.): Large-Scale Mammalian Cell Culture,
Academic Press, Orlando 1985, pp. 39 – 58.
116. G. G. Pugh, G. J. Berg, C. H. J. Sear in G. W.
Moody, P. B. Baker (eds.): Bioreactors and
Biotransformations, Elsevier Applied
Science, London 1987, pp. 121 – 131.
117. V. Jirku, J. Turková in K. Mosbach (ed.):
Immobilized Enzymes and Cells, Part B,
Methods in Enzymology, vol. 135, Academic
Press, Orlando 1987, pp. 341 – 357.
118. K. Mosbach, R. Mosbach, Acta Chem. Scand.
20 (1966) 2807 – 2810.
48 Immobilized Biocatalysts
119. I. Chibata, T. Tosa, T. Sato, Appl. Microbiol.
27 (1974) 878 – 885.
120. G. K. Skryabin, K. A. Koshcheenko in K.
Mosbach (ed.): Immobilized Enzymes and
Cells, Part B, Methods in Enzymology,
vol. 135, Academic Press, Orlando 1987,
pp. 198 – 222.
121. M. Kumakura, M. Yoshida, I. Kaetsu,
Biotechnol. Bioeng. 21 (1979) 679 – 688.
122. C. Bucke in K. Mosbach (ed.): Immobilized
Enzymes and Cells, Part B, Methods in
Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 175 – 189.
123. I. Chibata, T. Tosa, T. Sato, I. Takata in K.
Mosbach (ed.): Immobilized Enzymes and
Cells, Part B, Methods in Enzymology,
vol. 135, Academic Press, Orlando 1987,
pp. 189 – 198.
124. T. Omata, A. Tanaka, T. Yamane, S. Fukui,
Eur. J. Appl. Microbiol. Biotechnol. 6 (1979)
207 – 215.
125. S. Fukui, K. Sonomoto, A. Tanaka in K.
Mosbach (ed.): Immobilized Enzymes and
Cells, Part B, Methods in Enzymology,
vol. 135, Academic Press, Orlando 1987,
pp. 230 – 252.
126. M. L. Shuler in K. Mosbach (ed.): Immobilized
Enzymes and Cells, Part B, Methods in
Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 372 – 387.
127. C. Heath, G. Belfort, Adv. Biochem. Eng.
Biotechnol. 34 (1987) 1 – 31.
128. K. Ku et al, Biotechnol. Bioeng. 23 (1981)
79 – 95.
129. J. Van Brunt, Bio/Technology 4 (1986) 45 – 50.
130. J. Feder in K. Mosbach (ed.): Immobilized
Enzymes and Cells, Part B, Methods in
Enzymology, vol. 135, Academic Press,
Orlando 1987, pp. 393 – 399.
131. G. Klement, W. Schreier, H. W. D. Katinger in
M. Moo-Young (ed.): Bioreactor Immobilized
Enzymes and Cells, Fundamentals and
Applications, Elsevier Applied Science,
London 1988, pp. 53 – 58.
132. R. R. Mohan, N. N. Li, Biotechnol. Bioeng. 17
(1975) 1137 – 1156.
133. F. Lim, A. Sun, Science (Washington, D.C.)
210 (1980) 908 – 910.
134. A. Jarvis, T. Gridina, Bio Techniques 1 (1983)
22 –27.
135. R. G. Rupp in J. Feder, W. K. Tolbert (eds.):
Large-Scale Mammalian Cell Culture,
Academic Press, Orlando 1985, pp. 19 – 36.
136. J. N. Barbotin et al. in K. Mosbach (ed.):
Immobilized Enzymes and Cells, Part B,
Methods in Enzymology, vol. 135, Academic
Press, Orlando 1987, pp. 454 – 472.
137. A. Tanaka, S. Fukui in B. Mattiasson (ed.):
Immobilized Cells and Organelles, vol. 1,
CRC Press, Boca Raton, Florida 1983,
pp. 101 – 119.
138. H. Ochiai, A. Tanaka, S. Fukui in I. Chibata,
L. B. Wingard Jr. (eds.): Immobilized
Microbial Cells, Academic Press, New York
1983, pp. 153 – 187.
139. P. E. Gispy, K. K. Rao, D. O. Hall in K.
Mosbach (ed.): Immobilized Enzymes and
Cells, Part B, Methods in Enzymology,
vol. 135, Academic Press, Orlando 1987,
pp. 440 – 454.
140. B. Arkles, W. S. Brinigar, J. Biol. Chem. 250
(1975) 8856 – 8862.
141. S. S. Godbole, R. Kaul, S. F. D’Souza, G. B.
Nadkarni, Biotechnol. Bioeng. 25 (1983)
217 – 224.
142. A. Tanaka, N. Hagi, G. Gellf, S. Fukui, Agric.
Biol. Chem. 44 (1980) 2399 – 2405.
143. W. Hartmeier, Trends Biotechnol. 3 (1985)
149 – 153.
144. B. Hahn-Hägerdal, Biotechnol. Bioeng. 26
(1984) 771 – 774.
145. M. F. Cocquempot et al., Biotechnol Lett. 4
(1982) 313 – 318.
146. B. Hahn-Hägerdal in B. Mattiasson (ed.):
Immobilized Cells and Organelles, vol. 2,
CRC Press, Boca Raton, Florida 1983,
pp. 79 – 94.
147. M. Charles, J. Phillips in M. Moo-Young (ed.):
Comprehensive Biotechnology, vol. 2,
Pergamon Press, Oxford 1985, pp. 225 – 229.
148. K. Buchholz (ed.): “Characterization of
Immobilized Biocatalysts,” DECHEMA
Monogr. 84 (1979) 198 – 200.
149. J. E. Prenosil, I. J. Dunn, E. Heinzle in H.-J.
Rehm, G. Reed (eds.): Biotechnology, vol. 7 a,
Chap. 10, VCH Verlagsgesellschaft, Weinheim
1987.
150. A. Bar-Eli, E. Katschalski, Nature (London)
188 (1960) 856.
151. P. Monsan, D. Combes in K. Mosbach (ed.):
Methods in Enzymology, vol. 137, Academic
Press, New York 1988, pp. 584.
152. K. Martinek, V. P. Torchilin in K. Mosbach
(ed.): Methods in Enzymology, vol. 137,
Academic Press, New York 1988, pp. 615.
153. L. Goldstein, Y. Levin, E. Katchalski,
Biochemistry 3 (1964) 1913 – 1919.
154. D. I. C. Wang et al., Fermentation and Enzyme
Technology, J. Wiley & Sons, New York 1979.

155. Hor-Da Juang, Hung-Shan Weng, Biotech.
Bioeng. 26 (1984) 623 – 626.
156. K. Mosbach (ed.): Methods in Enzymology,
vol. 44, Acad. Press, New York 1976.
157. S. Furusaki, J. Chem. Eng. Japan 21 (1988) ,
219 –230.
158. K. Buchholz: “Reaction engineering
parameters for immobilized biocatalysts,”
Adv. Biochem. Eng. 24 (1982) 39.
159. S. F. Karel, S. B. Libicki, C. R. Robertson,
Chem. Eng. Sci. 40 (1985) 1321 – 1354.
160. J. M. Radovich, Enzyme Microb. Technol. 7
(1985) 2 – 10.
161. C. N. Satterfield, T. K. Sherwood: The Role of
Diffusion in Catalysis, Addision Wesley,
Reading, Mass. 1963.
162. C. N. Satterfield, Mass Transfer in
Heterogeneous Catalysis, MIT Press,
Cambridge, Mass. 1970.
163. P. H. Calderbank in V. W. Uhl, J. B. Gray
(eds.): Mixing, vol. 2, Academic Press, London
1967.
164. T. K. Sherwood, R. L. Pigford, C. Wilke: Mass
Transfer, McGraw-Hill, New York 1975.
165. K. Buchholz: “Mass transfer influence on
effectiveness,” DECHEMA-Monogr. 84
(1979) 212.
166. S. F. Karel, S. B. Libicki, C. R. Robertson,
Chem. Eng. Sci. 40 (1985) 1321 – 1354.
167. H. Tanaka, I. J. Dunn: “Kinetics of Biofilm
Nitrification,” Biotechnol. Bioeng. 24 (1982)
669.
168. D. I. C. Wang et al.: Fermentation and Enzyme
Technology, J. Wiley & Sons, New York 1979.
169. J. E. Prenosil, I. J. Dunn, E. Heinzle in H.-J.
Rehm, R. Reed (eds.): Biotechnology, vol. 7 a,
VCH Verlagsgesellschaft, Weinheim 1987,
p. 489.
170. R. E. Spier, J. B. Griffiths: Animal Cell
Biotechnology, vol. 3, Academic Press,
London 1988.
171. C. L. Cooney, Science (Washington, D.C.) 219
(1983) 728.
172. I. Chibata in P. Linko, J. Larinkari (eds.): Food
Process Engineering, vol. 2, Applied Science
Publ., London 1979, p. 1.
173. K. Venkatasubramanian in P. Linko, J.
Larinkari (eds.): Food Process Engineering,
vol. 2, Applied Science Publ., London 1979,
p. 162.
174. H. Samejima et al., Ann. N.Y. Acad. Sci. 434
(1984) 394.
175. J. J. Heijnen et al., GWF Gas Wasserfach:
Wasser/Abwasser 126 (1985) 81.
Immobilized Biocatalysts 49
176. J. Haggin, Chem. Eng. News 66 (1988) no. 23,
7.
177. E. C. Steiner, G. E. Blau, G. L. Agin: Simusolv,
Modeling and Simulation Software, Dow
Chemical Company, Midland, Michigan 1986.
178. I. Chibata, T. Tosa, T. Sato in J. F. Kennedy
(ed.): Biotechnology, vol. 7 a, VCH
Verlagsgesellschaft, Weinheim 1987, p. 653.
179. R. A. Messing, Biotechnol. Bioeng. 24 (1982)
1115.
180. I. Nilsson, S. Ohlson, Eur. J. Appl. Microbiol.
Biotechnol. 14 (1982) 86.
181. M. A. Gencer, R. Mutharasan, Biotechnol.
Bioeng. 25 (1983) 2243.
182. K. Hashimoto, S. Adachi, H. Noujima, Y.
Ueda, Biotechnol. Bioeng. 25 (1983) 2371.
183. J. P. Guiraud, A. M. Bajon, P. Chautard, P.
Galzy, Enzyme Microb. Technol. 5 (1983) 185.
184. P. Monsan, D. Combes, Biotechnol. Bioeng.
26 (1984) 347.
185. J. E. Prenosil, E. Stuker, T. Hediger, J. R.
Bourne, Ann. N.Y. Acad. Sci. 434 (1984) 136.
186. F. Godia, C. Casas, C. Sola, Biotechnol.
Bioeng. 30 (1987) 836.
187. J. M. S. Cabral, J. M. Novais, J. P. Cardoso,
J. F. Kennedy, J. Chem. Tech. Biotechnol. 36
(1986) 247.
188. J. B. Griffiths in R. E. Spier, J. B. Griffiths
(eds.): Animal Cell Biotechnology, vol. 3,
Academic Press, London 1988, p. 179.
189. O.-W. Merten, Trends Biotechnol. 5 (1987)
230.
190. C. G. J. Baker, A. Margaritis, M. A.
Bergougnou in M. Moo-Young (ed.):
Advances in Biotechnology, vol. 1, Pergamon
Press, Toronto 1981, p. 635.
191. P. F. Cooper, B. Atkinson: Biological Fluidized
Bed Treatment of Water and Wastewater, Ellis
Horwood, Chichester UK 1981.
192. M. Nagashima et al., Biotechnol. Bioeng. 26
(1984) 992.
193. G. F. Andrews, Biotechnol. Prog. 2 (1986) 16.
194. G. H. Schoutens et al., J. Chem. Technol.
Biotechnol. 36 (1986) 335.
195. S. J. Chen, C. T. Li, W. K. Shieh, J. Chem.
Technol. Biotechnol. 35 B (1985) 101.
196. M. Denac, I. J. Dunn, Biotechnol. Bioeng. 32
(1988) 159.
197. G. F. Andrews, J. Przezdziecki, Biotechnol.
Bioeng. 28 (1986) 802.
198. W. K. Shieh, J. D. Keenan, Adv. Biochem. Eng.
Biotechnol. 33 (1986) 131.
199. P. W. Runstadler Jr., S. R. Cernek in R. E.
Spier, J. B. Griffiths (eds.): Animal Cell
Biotechnology, vol. 3, Academic Press,
London 1988, p. 305.
50 Immobilized Biocatalysts
200. M. Hegglin, J. E. Prenosil in O. M. Neijssel,
R. R. van der Meer, K. Ch. A. M. Luyben
(eds.): Proceedings 4th European Congress on
Biotechnology, 1987, vol. 2, Elsevier,
Amsterdam 1987, p. 400.
201. S. B. Karkare, R. C. Dean, Jr., K.
Venkatasubramanian, BioTechnology 3 (1985)
247.
202. L.-S. Fan, K. Fujie, T.-R. Long, W.-T. Tang,
Biotechnol. Bioeng. 30 (1987) 498.
203. N. Bernet et al., J. Ind. Microbiol. 2 (1987)
129.
204. E. Adema, A. J. Sinskey, Biotechnol. Prog. 3
(1987) 74.
205. J. A. Schonberg, G. Belfort, Biotechnol. Prog.
3 (1987) 80.
206. S. L. Gallagher, J. P. Tharakan, P. C. Chau,
Biotechnol. Tech. 1 (1987) 91.
207. J. P. Tharakan, P. C. Chau, Biotechnol. Bioeng.
28 (1986) 329.
208. O. T. Schönherr, P. J. T. A. van Gelder in R. E.
Spier, J. B. Griffiths (eds.): Animal Cell
Biotechnology, vol. 3, Academic Press,
London 1988, p. 337.
209. M. A. Tyo, B. J. Bulbulian, B. Z. Menken, T. J.
Murphy in R. E. Spier, J. B. Griffiths (eds.):
Animal Cell Biotechnology, vol. 3, Academic
Press, London 1988, p. 357.
210. J. Lehmann, J. Vorlop, H. Büntemeyer in R. E.
Spier, J. B. Griffiths (eds.): Animal Cell
Biotechnology, vol. 3, Academic Press,
London 1988, p. 222.
211. J. S. Hwang, H. N. Chang, Biotechnol. Lett. 9
(1987) 237.
212. T. B. V. Roy, H. W. Blanch, C. R. Wilke,
Trends Biotechnol. 1 (1983) 135.
213. H. Strathmann, Trends Biotechnol. 3 (1985)
112.
214. R. S. Tutunjian, BioTechnology 3 (1985) 615.
215. H. N. Chang, Biotechnol. Adv. 5 (1987) 129.
216. R. A. Messing, Biotechnol. Bioeng. 16 (1974)
897 – 908.
217. K. Buchholz: “Mass transfer influence on
effectiveness,” DECHEMA-Monogr. 84
(1979) 212.
218. K. Buchholz, J. Klein in K. Mosbach (ed.):
Methods in Enzymology, vol. 135, Acad.
Press, Orlando 1987.
219. J. Klein, F. Wagner: “Immobilized whole
cells,” DECHEMA-Monogr. 84 (1979)
265 – 311.
220. V. M. Oven, Ann. N.Y. Acad. Sci. 501 (1987)
233.
221. L. J. Kricka, G. H. G. Thorpe, Trends
Biotechnol. 4 (1986) 253.
222. J. V. Twork, A. M. Yacynych, Biotechnol.
Progress 2 (1986) 67.
223. D. M. Rawson, Intern. Ind. Biotechnol. 8
(1988) no. 2, 18.
224. J. N. North, Trends Biotechnol. 3 (1985) 180.
225. G. G. Guilbault, Analytical Uses of
Immobilized Enzymes, Marcel Dekker, New
York 1984.
226. C. R. Lowe, Trends Biotechnol. 2 (1984) 59.
227. L. D. Bowers, Anal. Chem. 58 (1986) 513 A.
228. K. P. Flint in Biotechnology of Waste
Treatment and Exploitation, Ellis Horwood,
Chichester 1987, p. 32.
229. R. D. Schmid, Biosensors: International
Workshop 1987, VCH Verlagsgesellschaft,
Weinheim 1987.
230. K. Mosbach, “Immobilized Enzymes and
Cells,” in: Methods in Enzymology, Part D,
vol. 137, Academic Press, San Diego 1988.
231. T. E. Edmonds, Chemical Sensors, Blackie,
Glasgow 1988.
232. S. W. May, S. R. Padgette, Bio/Technol. 10
(1983) 677 – 686.
233. G. G. Guilbault, Enzyme Eng. 6 (1982)
395 – 404.
234. S. Suzuki, I. Karube, Appl. Biochem. Bioeng.
3 (1981) 145 – 174.
235. I. Karube, S. Suzuki in A. I. Laskin (ed.):
Enzymes and Immobilized Cells in
Biotechnology, Benjamin/Cummings, Menlo
Park 1985, pp. 209 – 227.
236. G. G. Guilbault, G. de Olivera Neto in A. I.
Laskin (ed.): Enzymes and Immobilized Cells
in Biotechnology, Benjamin/Cummings,
Menlo Park California 1985, pp. 227 – 261.
237. I. Karube in J. F. Kennedy (ed.):
Biotechnology, vol. 7 a, VCH
Verlagsgesellschaft, Weinheim 1987,
pp. 685 – 708.
238. V. P. Torchilin in Adv. Drug Delivery Rev. 1
(1987) 41 – 86.
239. M. D. Klein, R. Langer, Trends Biotechnol. 4
(1986) 179 – 186.
240. T. M. S. Chang in A. I. Laskin (ed.): Enzymes
and Immobilized Cells in Biotechnology,
Benjamin/Cummings, Menlo Park, California
1985, pp. 263 –288.
241. A. Margaritis, F. J. A. Merchant, CRC Crit.
Rev. Biotechnol. 1 (1984) 339 – 393.
242. F. Godia, C. Casas, C. Sola, Proc. Biochem.
22 (1987) 43 – 48.
243. M. J. Nunez, J. M. Lema, Enzyme, Microb.
Technol. 9 (1987) 642 – 651.
244. C. Akin, Biotechnol. and Genet. Eng. Rev. 5
(1987) 320 – 367.

245. P. Linko, Y.-Y. Linko, CRC Crit Rev.
Biotechnol. 1 (1984) 289 – 337.
246. Novo Technical Report, 175d-GB, Novo Allé,
Bagsvaerd, Denmark 1985.
247. I. Chibata, T. Tosa, T. Sato in A. I. Laskin
(ed.): Enzymes and Immobilized Cells in
Biotechnology, Benjamin/Cummings, Menlo
Park, California 1985, pp. 37 – 70.
248. I. Chibata, T. Tosa, T. Sato in J. F. Kennedy
(ed.): Biotechnology, vol. 7 a, VCH
Verlagsgesellschaft, Weinheim 1987,
pp. 653 – 684.
249. I. Chibata, T. Tosa, T. Sato, J. Mol. Cat. 37
(1986) 1 – 24.
250. Y.-Y. Linko, P. Linko in J. Tramper, H. C. van
der Plas, P. Linko (eds.): Biocatalysts in
Organic Synthesis, Elsevier, Amsterdam
1985, pp. 159 – 178.
251. M. S. Switzenbaum, Enzyme Microb. Technol.
5 (1983) 242 – 250.
252. P. F. Cooper, B. Atkinson (eds.): Fluidized Bed
Treatment of Water and Wastewater, Ellis
Horwood, Chichester 1981.
253. C. Webb, B. M. Black, B. Atkinson (eds.):
Process Engineering Aspects of Immobilized
Cell Systems, Inst. Chem. Engineers, Rugby
1986.
254. C. Webb in C. F. Forster, D. A. J. Wase (eds.):
Environmental Biotechnology, Ellis Horwood,
Chichester 1987, pp. 347 – 376.
255. D. Barnes, P. A. Fitzgerald in C. F. Forster,
D. A. J. Wase (eds.): Environmental
Biotechnology, Ellis Horwood, Chichester
1987, pp. 57 – 113.
256. R. E. Shamel, J. J. Chow, Chem. Eng. Prog. 83
(1987) no. 10, 41 – 44.
257. C. G. Rosén, CHEMTECH 17 (1987) no. 10,
612 – 618.
258. W. E. Goldstein, Biotechnol. Bioeng. Symp. 17
(1986) 763 – 776.
259. P. L. Layman, Chem. Eng. News 64 (1986)
no. 37, Sept. 15, 11 – 14.
260. P. B. Poulsen in [38] , pp. 413 – 419.
261. P. B. Poulsen: Third European Congress on
Biotechnology, München 1984, Proceedings,
vol. 4, VCH Verlagsgesellschaft, Weinheim
1985, pp. 339 – 344.
262. M. M. Domach, T. J. Gricus, Biotechnol.
Bioeng. 27 (1985) 1458 – 1465.
263. P. Linko: Third European Congress on
Biotechnology, München 1984, Proceedings,
vol. 4, VCH Verlagsgesellschaft, Weinheim
1985, pp. 357 – 363.
264. K. Kieslich: Third European Congress on
Biotechnology, München 1984, Proceedings,
Immobilized Biocatalysts 51
vol. 4, VCH Verlagsgesellschaft, Weinheim
1985, pp. 39 – 72.
265. A. S. Lubiniecki, Dev. Ind. Microbiol. 28
(1987) 133 – 138.
266. L. H. Smithson, CHEMTECH 18 (1988) no. 3,
168 – 173.
267. W. R. Arathoon, J. R. Birch, Science
(Washington, D.C.) 232 (1986) 1390 – 1395.
268. R. W. Swartz in M. Phillips, S. P. Shoemaker,
R. D. Middlekauff, R. M. Ottenbrite (eds.):
“The Impact of Chemistry on Biotechnology,”
ACS Symp. Ser. 362 (1988) 102 – 120.
269. H. Katinger in O. M. Neijssel, R. R. van der
Meer, K. Ch. A. M. Luyben (eds.): Proceedings
4th European Congress on Biotechnology,
vol. 4, Elsevier, Amsterdam 1987, pp. 21 – 42.
270. V. Petiard, P. Steck in J. M. Cardoso et al.
(eds.): Perspect. in Biotechnol. NATO ASI Ser.
A 128 (1987) 139 – 153.
271. P. Brodelius in A. I. Laskin (ed.): Enzymes and
Immobilized Cells in Biotechnology,
Benjamin/ Cummings, Menlo Park, California
1985, pp.109 – 147.
272. O. Sahai, M. Knuth, Biotechnol. Prog. 1
(1985) 1 – 9.
273. V. Petiard et al. in O. M. Neijssel, R. R. van der
Meer, K. Ch. A. M. Luyben (eds.):
Proceedings 4th European Congress on
Biotechnology, 1987, vol. 4, Elsevier,
Amsterdam 1987, pp. 157 – 170.
274. P. F. Noordervliet, D. A. Toet in H.-J. Rehm,
G. Reed (eds.) Biotechnology, A
Comprehensive Treatise in 8 Volumes, vol. 7 a,
VCH Verlagsgesellschaft, Weinheim 1987,
pp. 711 – 741.
275. “Evaluation of Certain Food Additives,” 21st
Report of the Joint FAO/WHO Expert
Committee on Food Additives (JECFA),
W.H.O. Tech. Rep. Ser. 617 (1978).
276. W. H. B. Denner in T. Godfrey, J. R. Reichelt
(eds.): Industrial Enzymology, The Nature
Press, New York 1983, pp. 111 – 137.
277. Code of Federal Regulations, 21 CFR Sec.
110.1, Office of Federal Register, National
Archives and Records Service, Washington
D.C. 1981.
278. Food Chemicals Codex, 3rd ed., National
Academy of Science/National Research
Council, Food and Nutrition Board,
Committee on Codex Specifications, National
Academy Press, Washington D.C. 1981.
279. J. S. Kahan, J. N. Gibbs, Appl. Biochem.
Biotechnol. 11 (1985) 489 – 506.
280. H. H. Weetall, W. H. Pitcher Jr., Science
(Washington D.C.) 232 (1986) 1396 – 1403.
52 Immobilized Biocatalysts
281. J. R. Fordham, N. H. Block, Dev. Ind.
Microbiol. 28 (1987) 25 – 31.
282. J. R. Reichelt in T. Godfrey, J. R. Reichelt
(eds.): Industrial Enzymology, The Nature
Press, New York 1983, pp. 138 – 156.
283. J. C. Petricciani, Swiss Biotech. 5(2a) (1987)
19 – 20.
284. A. S. Lubiniecki, Dev. Ind. Microbiol. 28
(1987) 133 – 138.
285. “Safe Biotechnology – General
Considerations,” report prepared by the Safety
in Biotechnology Working Party of the
European Federation of Biotechnology, Third
European Congress on Biotechnology, vol. 3,
VCH Verlagsgesellschaft, Weinheim 1984,
pp. 647 – 662.
286. “Guidelines for Research Involving
Recombinant DNA Molecules,” National
Institute of Health (NIH) revised version, US
Fed. Reg. 49 (Nov. 23, 1984) 46 266.
287. C. Sekhar, Chem. Eng. (N.Y.) 63 (1985)
no. 17, 57 – 59.
288. Recombinant DNA Safety Considerations,
OECD, Paris 1986.
Immunodiagnostic Reagents → Diagnostic Reagents
289. S. P. Vranch in O. M. Neijssel, R. R. van der
Meer, K. Ch. A. M. Luyben (eds.):
Proceedings 4th European Congress on
Biotechnology, vol. 4, Elsevier, Amsterdam
1987, pp. 355 – 363.
290. R. L. Greenstein, Appl. Biochem. Biotechnol.
11 (1985) 489 – 506.
291. T. O. McGarity, K. O. Bayer in M. Moo-Young
(ed.): Comprehensive Biotechnology, vol. 4,
Pergamon Press, Oxford 1985, pp. 587 – 608.
292. “Coordinated Framework for the Regulation of
Biotechnology,” U.S. Fed. Reg. 51 (June 26,
1986) 23 302 – 23 393.
293. D. T. Kingsbury, Trends Biotechnol. 6 (1988)
39 – 42.
294. Introduction of Genetically Engineered
Organisms into the Environment: Key Issues,
US National Academy of Sciences,
Washington D.C. 1987.
295. B. P. Ager, Trends Biotechnol. 6 (1988)
42 – 44.
296. European Committee on Regulatory Aspects
of Biotechnology (ECRAB), Swiss Biotech.
4(5) (1986) 15 – 20.
 Immunotherapy and Vaccines 1

Immunotherapy and Vaccines

Stanley J. Cryz, Jr., Swiss Serum and Vaccine Institute Berne, Switzerland (Chaps. 1 and2)
Marta Granstrom, Departments of Clinical Microbiology and of Vaccine Production, National
Bacteriological Laboratory, Karolinska Hospital, Stockholm, Sweden (Chap. 3)
Bruno Gottstein, Institut für Parasitologie, Universität Zürich, Zürich, Switzerland (Section 4.1)
Luc Perrin, Division d’Hematologie, Hôpital Cantonal Universitaire, Genève, Switzerland (Section 4.2)
Alan Cross, Department of Bacterial Diseases, Walter Reed Army Institute of Research, Washington D.C.
20307-5100, United States (Chap. 5)
James Larrick, Genelabs Incorporated, Redwood City, California, United States (Chap. 6)

1. Introduction . . . . . . . . . . . . . . 3 3.11. Tick-Borne Encephalitis Vaccine 37

1.1. Historical Aspects . . . . . . . . . . 3 3.12. Japanese Encephalitis Vaccine . . 38
1.2. Principles and Definitions . . . . . 4 3.13. Smallpox Vaccine . . . . . . . . . . . 39
1.2.1. Antigens . . . . . . . . . . . . . . . . . 4 3.14. Rift Valley Fever Vaccine . . . . . 39
1.2.2. Antibodies . . . . . . . . . . . . . . . . 5 4. Vaccines against Parasites . . . . . 39
1.2.3. Immune Response . . . . . . . . . . . 7 4.1. Vaccines against Helminths . . . . 39
1.2.4. Active Immunization . . . . . . . . . 8 4.1.1. Vaccines against Schistosoma . . . . 40
1.2.5. Passive Immunization . . . . . . . . . 9 4.1.2. Vaccines against Nematodes . . . . 41
1.2.6. Genetic Engineering . . . . . . . . . 9 4.1.2.1. Gastrointestinal Nematodes . . . . . 41
2. Bacterial Vaccines . . . . . . . . . . 10 4.1.2.2. Tissue-Invading Nematodes (Filari-
2.1. Diphtheria Vaccine . . . . . . . . . . 10 idae) . . . . . . . . . . . . . . . . . . . 42
2.2. Tetanus Vaccine . . . . . . . . . . . . 11 4.1.3. Vaccines against Cestodes . . . . . . 43
2.3. Pertussis Vaccine . . . . . . . . . . . 12
4.2. Malaria Vaccine . . . . . . . . . . . 44
2.4. Typhoid Fever Vaccine . . . . . . . 14
4.2.1. Strategy for Malaria Vaccine Devel-
2.5. Streptococcus pneumoniae Vaccine 15
opment . . . . . . . . . . . . . . . . . . 45
2.6. Shigella Vaccines . . . . . . . . . . . 17
4.2.2. Sporozoite Vaccines . . . . . . . . . . 45
2.7. Cholera Vaccine . . . . . . . . . . . 17
4.2.3. Asexual Blood Stage Vaccine . . . . 47
2.8. Vaccines Against Nosocomial
Pathogenes . . . . . . . . . . . . . . . 19 4.2.3.1. Merozoite Surface Antigens . . . . . 48
2.9. Meningococcal Meningitis Vac- 4.2.3.2. Rhoptry Antigens . . . . . . . . . . . 48
cine . . . . . . . . . . . . . . . . . . . . 20 4.2.3.3. Antigens Associated with the Mem-
2.10. Tuberculosis Vaccine . . . . . . . . 21 brane of Infected Erythrocytes . . . 49
2.11. Escherichia coli Vaccines . . . . . . 23 4.2.3.4. Other Proteins and Synthetic Pep-
2.12. Neisseria gonorrhoeae Vaccine . . 24 tides . . . . . . . . . . . . . . . . . . . . 49
2.13. Hemophilus influenzae Type b 4.2.4. Sexual Stages–Transmission Block-
Vaccines . . . . . . . . . . . . . . . . . 25 ing Immunity . . . . . . . . . . . . . . 50
3. Viral Vaccines . . . . . . . . . . . . . 26 5. Immunotherapy . . . . . . . . . . . 50
3.1. Measles Vaccine . . . . . . . . . . . . 26 5.1. Gamma Globulin Preparations . . 51
3.2. Mumps Vaccine . . . . . . . . . . . . 27 5.1.1. Standard Immune Serum Globulin . 51
3.3. Rubella Vaccine . . . . . . . . . . . . 28 5.1.2. Immunoglobulin for Intravenous
3.4. Combined Measles – Mumps – Use . . . . . . . . . . . . . . . . . . . . 52
Rubella Vaccine . . . . . . . . . . . . 29 5.1.3. Hyperimmune Globulins and Anti-
3.5. Polio Vaccine . . . . . . . . . . . . . . 30 toxins . . . . . . . . . . . . . . . . . . . 53
3.6. Hepatitis B Vaccine . . . . . . . . . 31 5.1.4. Production Requirements . . . . . . 53
3.7. Rabies Vaccine . . . . . . . . . . . . 33 5.2. Prophylaxis with Immune Serum
3.8. Influenza Vaccine . . . . . . . . . . . 34 Globulin . . . . . . . . . . . . . . . . . 54
3.9. Varicella Vaccine . . . . . . . . . . . 35 5.3. Prophylaxis with Hyperimmune
3.10. Yellow Fever Vaccine . . . . . . . . 36 Globulins . . . . . . . . . . . . . . . . 55

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 049
2 Immunotherapy and Vaccines

5.4. Therapy with Immune Serum 5.7. Adverse Effects of Gamma Glob-
Globulin . . . . . . . . . . . . . . . . . 56 ulin Preparations . . . . . . . . . . . 61
5.5. Prophylaxis and Therapy with 5.8. Future Prospects . . . . . . . . . . . 63
Intravenous Immunoglobulin 6. Immunotherapeutic Uses of
(IVIG) . . . . . . . . . . . . . . . . . . 56 Monoclonal Antibodies . . . . . . . 63
5.5.1. Viral Infection . . . . . . . . . . . . . 57 6.1. Introduction . . . . . . . . . . . . . . 63
5.5.2. Bacterial Infection . . . . . . . . . . . 57 6.2. Bacterial Targets . . . . . . . . . . . 64
5.5.3. Noninfectious Diseases . . . . . . . . 59 6.3. Viral and Chlamydial Targets . . 65
5.5.3.1. Therapeutic Effect of IVIG . . . . . 59 6.4. Parasite Targets . . . . . . . . . . . . 66
5.5.3.2. Mechanism of Action . . . . . . . . . 59 7. References . . . . . . . . . . . . . . . 66
5.6. Prophylaxis and Therapy with
Plasma and Other Blood Products 60

Abbreviations used in this article: RNA ribonucleic acid

TCID tissue culture infectious dose
AIDS acquired immune deficiency syndrome
BCG acillus Calmette-Guerin
CMV Cytomegalovirus
CPS capsular polysaccharide 1. Introduction [1–9]
CS circumsporozoite
Da dalton 1.1. Historical Aspects
DNA deoxyribonucleic acid
Immunization is the most efficient, cost-
DPT diphtheria–pertussis–tetanus
effective means of preventing infectious dis-
DPT-Pol. diphtheria–pertussis–tetanus–polio
eases. The concept of preventing disease by vac-
DT diphtheria–tetanus
cination is old: in China and India, the prac-
FHA filamentous hemagglutinin
tice of “variolation,” whereby small quantities
HBIG human anti-HBV immune globulin
of material from disease pustules were used to
HBsAg hepatitis B surface antigen
immunize people against smallpox, was prac-
HBV hepatitis B virus
ticed before 1000 b.c. The first “rational” ap-
HIV human immunodeficiency virus
proach to vaccination was taken by Jenner in
HRIG human rabies immune globulin
1798, who used naturally attenuated cowpox
humab human monoclonal antibodies
to immunize against smallpox. About 100 years
Ig immunoglobulin
later, Pasteur introduced vaccines against an-
IPV inactivated polio vaccine
thrax and rabies based upon attenuated viru-
ISG immune serum globulin
lent organisms. The discovery by von Behring
ITP idiopathic thrombocytopenic
in 1890, that serum antibodies could neutralize
purpura
diphtheria toxin opened the door for a new av-
IU international units
enue of vaccine development and passive ther-
IVIG intravenous immunoglobulin
apy, whereby preformed antibodies were trans-
Lf limit of flocculation
ferred to at-risk patients. By the beginning of the
LPS lipopolysaccharide
20th century, serum obtained from immunized
MMR measles–mumps–rubella
animals was used to treat a variety of diseases
NANP asparagine–alanine–asparagine–
including diphtheria and tetanus. The use of hu-
proline
man serum followed shortly in 1907.
NVDP asparagine–valine–aspartic acid–
Since the turn of the century, a wide range
proline
of vaccines and antisera have been introduced
OPV oral polio vaccine
to manage infectious and noninfectious dis-
PFU plaque forming units
eases (Table 1). The development of such agents
PT pertussis toxin
has required expertise from a variety of dis-
RESA ring-infected erythrocyte surface
ciplines including microbiology, biochemistry,
antigen
immunology, and molecular biology.
 Immunotherapy and Vaccines 3

Table 1. Vaccines and immunoglobulins

foreign or “non-self” molecules are immuno-
genic. Usually the larger and more complex a
Type of vaccine or Disease molecule is, the more immunogenic it will be.
immunoglobulin For example, the gram-negative bacterial cell
Bacterial cholera envelope shown in Figure 1 contains many dif-
vaccines diphtheria
Hemophilus influenzae b
ferent somatic (cell-associated) antigens, such
meningococcal meningitis as lipopolysaccharide (LPS), outer-membrane
pertussis proteins, and phospholipids. Numerous factors
Streptococcus pneumoniae determine the immunogenicity of a purified
tetanus
tuberculosis
molecule. Size is of critical importance: pro-
typhoid fever teins with a molecular mass of ≤ 20 000 are
Viral vaccines hepatitis B poorly immunogenic. Similarly, simple polysac-
influenza
charides composed of a limited number of re-
Japanese encephalitis
measles peating monosaccharides are not immunogenic
mumps unless their molecular mass exceeds 500 000.
polio Small molecules can be rendered immunogenic
rabies
Rift Valley fever
by covalently coupling them to larger molecules,
smallpox forming conjugates.
tick-borne encephalitis A single antigen may contain many epitopes,
varicella which are specific areas of the molecule with a
yellow fever
Immunoglobulins diphtheria three-dimensional configuration that induces an
against Hemophilus influenzae, meningo- immune response. Complex molecules, such as
bacterial coccal meningitis, Streptococcus large proteins composed of many different ami-
diseases pneumoniae (polyvalent preparation)
pertussis
no acids, contain more epitopes than a compara-
tetanus tively simple polysaccharide composed of two or
Immunoglobulins cytomegalovirus three monosaccharide repeats. The immune re-
against viral hepatitis A sponse to a given antigen can vary greatly among
diseases hepatitis B
human immunodeficiency virus (HIV)
species and individuals within a species due to
(normal intravenous immunoglobu- immune regulation genes.
lin preparation administered to HIV-
positive infants)
measles
mumps 1.2.2. Antibodies
rabies
rubella Antibodies are proteins found primarily in the
vaccinia
varicella
serum which are produced by B cells in response
Immunoglobulins hypogammaglobulinemia to contact with a foreign antigen. There are sev-
against nonin- rhesus factor eral different classes of antibodies with charac-
fectious idiotypic thrombocytopenia purpura teristic functions (see Chap. 5, Table 3). A sche-
diseases
matic of an immunoglobulin G (IgG) molecule
is shown in Figure 2. Immunoglobulins are com-
posed of light and heavy chains held together by
disulfide bonds. Each chain has a variable and
1.2. Principles and Definitions a constant region. The tertiary structure of the
variable region accounts for the specificity of
1.2.1. Antigens antibody binding. An antibody produced from
a given clone of B cells (Section 1.2.3) recog-
An antigen is a molecule that can elicit an im- nizes and binds to a given epitope or closely
mune response (either humoral, i.e., antibody- related epitopes. Antibodies which recognize
mediated, or cellular, i.e., cell-mediated) or an more than one epitope are termed cross-reactive.
immune reaction, such as an allergic reaction. The strength with which an antibody binds to an
An antigen that evokes an immune response is antigen is termed affinity and is determined by
commonly referred to as an immunogen. Only
4 Immunotherapy and Vaccines
Figure 1. Schematic representation of the gram-negative bacterial cell envelope
the “fit” between the immunoglobulin binding
site and the epitope.
Figure 2. Schematic representation of an immunoglobulin
G (IgG) molecule
Antibodies produced by a single clone of
B cells are termed monoclonal antibodies and
recognize only a single epitope (see Chap. 6).
Polyclonal antibodies are produced by several
B cell clones which recognize the same antigen
but bind to different epitopes.
Immunoglobulins are divided into five
classes termed IgG, IgM, IgA, IgD, and IgE
based upon physical and structural differences;
they are all glycoproteins. The vast majority of
immunoglobulins circulate in the plasma frac-
tion, but certain cells of the immune system can
express immunoglobulin on their surface.
Immunoglobulin G (IgG) has a molecular
mass of 150 000 and has two antigen binding
sites; it represents approximately 80 % of all im-
munoglobulins in normal serum (8 – 16 mg/mL).
Four IgG subclasses (IgG 1 – IgG 4) are found
which account for ca. 70, 19, 8, and 3 % of
total IgG, respectively; they differ with re-
spect to their antigenic properties in the con-
stant region of the heavy chains. The func-
tional attributes of these subclasses are detailed
elsewhere (see Section 5.1). Immunoglobulin G
readily crosses the placenta and provides pro-
tection against a variety of infectious diseases
in the neonate. For example, immunization of
the mother against tetanus shortly before deliv-
ery ensures a protective level of antibody for
the infant. Immunoglobulin G also diffuses into
the extravascular tissue more readily than other
immunoglobulin classes. The IgG antibody is
thought to be responsible for neutralizing the
majority of bacterial toxins (e.g., tetanus and
diphtheria toxin) formed during an infection.
Upon binding with invading bacteria, IgG ac-
tivates the complement system (a group of inter-
acting serum proteins) which attracts phagocytic
cells. The binding of complement components
to the Fc region of IgG permits the uptake and
killing of the bacteria by phagocytes.
Immunoglobulin M (IgM) is a 900 000 dal-
ton pentamer whose five IgG-like units are held
together by intramolecular disulfide bonds and
stabilized by a polypeptide “anchor” termed the
“J-chain.” IgM comprises 5 – 10 % of normal cir-
culating immunoglobulins (1 – 2 mg/mL). Due
to its large size, IgM is confined to the in-
travascular space. It binds complement but, un-
like IgG, does not bind directly to phagocytic
cells. Immunoglobulin M is usually the first an-
tibody class formed in response to foreign anti-

gen exposure. Due to its multivalency, IgM is ex-
tremely efficient at agglutinating bacteria. This
phenomenon is important in the control of bac-
teremia and the neutralization of LPS released
by gram-negative bacteria.
Immunoglobulin A (IgA) occurs as 160 000
dalton monomers or 320 000 dalton dimers.
Monomers are found primarily in the intravascu-
lar space and comprise approximately 10 – 15 %
(1.4 – 4 mg/mL) of total serum immunoglobu-
lins. There are two subclasses, IgA 1 and IgA 2.
Dimeric IgA, termed secretory IgA, is formed
by noncovalent interaction with the “secretory
piece,” a glycoprotein of about 60 000 daltons.
Secretory IgA is the predominant immunoglob-
ulin found in mucous secretions, such as tears,
colostrum, pulmonary, intestinal, and urinogen-
ital fluids; IgA can bind complement and react
with phagocytic cells.
Although parenteral immunization stimu-
lates a vigorous serum IgA response, secretory
IgA is considered to be of greater importance
due to its “first line” defensive role in body se-
cretions. Secretory IgA plays a critical role in
providing protection against respiratory, intesti-
nal, and urinogenital tract bacterial pathogens.
Immunoglobulin D (IgD) is a monomer of ca.
180 000 daltons; concentrations vary widely in
humans ranging from 0 to ca. 0.4 mg/mL. Com-
pared to other immunoglobulin classes, IgD is
very susceptible to proteolysis and possesses a
short half-life (about 3 days). Since IgD does not
fix complement or bind to phagocytic cells, it is
not considered to be a “protective” immunoglob-
ulin as are IgA, IgG, and IgM. However, IgD is
abundant on the surface of B lymphocytes and
may play a central role in the activation pro-
cess which leads to the development of antibody-
secreting plasma cells.
Immunoglobulin E (IgE) is a monomer of
200 000 daltons; its average serum concentra-
tion (ca. 250 ng/mL) is the lowest of any im-
munoglobulin. Although IgE does not fix com-
plement or react directly with phagocytic cells,
it has a very high affinity for most cells and ba-
sophils. Immunoglobulin E appears to mediate
both a protective and a detrimental immune re-
sponse. At the mucoid surface, pathogens bind-
ing to IgE stimulate an acute inflammatory re-
sponse by triggering the release of potent medi-
ators from mast cells and basophils. A protective
role for IgE is indicated in several chronic par-
Immunotherapy and Vaccines 5
asitic infections, most notably schistosomiasis
(see Chap. 4) where high levels (> 100 µg/mL)
of serum IgE have been noted. Immunoglobu-
lin E induces an inflammatory response and “re-
cruits” effector cells to the area. In contrast, IgE-
mediated degranulation of effector cells leads to
the release of vasoreactive molecules. Individu-
als suffering from certain allergies can also have
elevated serum IgE levels.
1.2.3. Immune Response
The chain of events leading to an immune
response is extremely complicated and not
yet completely understood. A simple model
is shown in Figure 3. The human is capable
of forming an immune response to thousands
of foreign antigens. The humoral (antibody-
mediated) response depends on the B cells
which are lymphocytes derived from bone mar-
row stem cells. Upon maturation, B cells form
antibody-secreting plasma cells.
Each clone of B cells has an immunoglob-
ulin molecule with a recognition site for a spe-
cific antigen on its surface. Binding of the appro-
priate antigen to the B cell causes proliferation
of the clone whereby the progeny cells secrete
antibody whose specificity is identical to that
of the cell-surface immunoglobulin. A foreign
antigen can be taken up by macrophages, pro-
cessed, and presented upon the cell surface, or
may remain in a soluble state. All antigens can be
termed T-dependent or T-independent depend-
ing on whether they require or do not require
the interaction of T cells for antibody synthe-
sis (T cells are lymphocytes derived from the
thymus). In the case of a T-independent anti-
gen (usually polymers, such as bacterial capsular
polysaccharides), the antigen can cross-link the
B cell surface antibody molecules of the B cell;
this initiates proliferation and antibody synthe-
sis. A T-dependent antigen requires, in addition
to binding to B-cell surface immunoglobulin, the
release of T cell factors which act upon the B cell
to initiate proliferation.
A human exposed to an antigen in this way
is considered primed. This is a critical factor
in immunization. After initial exposure to a T-
dependent antigen by immunization, the induced
serum antibody concentration is low and re-
turns to near-basal levels comparatively quickly
6 Immunotherapy and Vaccines
Figure 3. Immunological response cascade to a foreign antigen or vaccine, leading to antibody response or cell-mediated
response
(the average half-life of IgG is ca. 22 days,
whereas IgM is ca. 5 days). Upon revaccination
(boosting) or natural exposure to the pathogen
or toxin, a vigorous antibody response occurs
in which high antibody levels are synthesized
over a longer period of time. This is termed an
anamnestic response. Therefore, high levels of
serum antibodies do not have to be present for
the individual to be protected by prior vaccina-
tion. A T-independent antigen does not evoke an
anamnestic response.
The induction of a cell-mediated immune
(CMI) response is rather more complicated and
less well defined. The CMI response plays a crit-
ical role in immunity to pathogens able to live
and proliferate within host cells. The first step
is “activation” of T cells. Like B cells, T cells
also recognize specific antigens. Recognition of
an antigen on the surface of a macrophage by
T cells results in the release of interleukin 1 (IL-
1), a lymphokine which activates them. One sub-
population of T cells synthesizes interleukin 2
(IL-2), also known as T-cell growth factor,
which causes a second subpopulation of acti-
vated T cells expressing the IL-2 receptor to pro-
liferate and become cytotoxic T cells. Cytotoxic
T cells recognize the specific antigen they are
activated against when expressed on the surface
of infected cells; they thus kill these infected
cells by lysis. A third subpopulation of T cells
evolve to be “primed” memory cells which un-
dergo rapid proliferation upon reexposure to the
same antigen.
Immunization with a given vaccine may in-
duce a CMI, a humoral antibody response, or
both depending on the infecting pathogen. Both
types of immunity are usually desirable and
may act synergistically. For example, vaccine-
induced immunity to many viral diseases, such
as rabies, measles, mumps, and rubella, is con-
firmed by measuring serum antibody levels.
However, infected cells can only be destroyed

by cytotoxic T cells. Circulating antibody prob-
ably prevents the spread of the virus, while cy-
totoxic T cells eliminate already infected cells.
Tuberculosis vaccines stimulate a good antibody
response but are less proctective due to the sup-
pression of the critical CMI response.
1.2.4. Active Immunization
Active immunization entails the administration
of one or more antigens to a host in the form
of a vaccine in an attempt to elicit a protec-
tive immune response. In this way an individ-
ual can be rendered immune to a variety of dis-
eases. Most vaccines are administered to infants
or young children. The trend is to combine sev-
eral monovalent vaccines (vaccines containing
a single antigen or vaccine strain) to form mul-
tivalent vaccines capable of simultaneously in-
ducing immunity to several diseases. Infants are
routinely immunized against diphtheria, tetanus,
pertussis, and, in some countries, polio by ad-
ministering the vaccines in a single injection.
Similarly, young children are immunized simul-
taneously with a multivalent measles, mumps,
and rubella vaccine. Administration of multi-
valent vaccines reduces the number of visits to
healthcare centers, a critical point in developing
countries.
Historically, parenteral vaccines have usu-
ally been used (i.e., injected with a syringe and
needle). This approach is extremely success-
ful for systemic diseases such as diphtheria or
measles. However, in localized diseases (e.g.,
intestinal infections such as cholera), parenteral
vaccines are of limited use since the immune
system needs to be stimulated at the site of in-
fection (e.g., local immune system). Therefore,
effort is now being directed at controlling intesti-
nal infectious diseases with orally administered
vaccines. Oral vaccines can consist of
1) killed intact bacteria,
2) toxoids (i.e., toxins that are still immuno-
genic but are rendered biologically inactive
by treatment with a chemical, heat, or muta-
tion)
3) subunit vaccines, in which only the nontoxic
portion of a molecule is used, or
4) live-attenuated vaccines, in which a viral or
bacterial strain is rendered nonpathogenic
Immunotherapy and Vaccines 7
(e.g., by passaging the virus in cell culture or
deletion of bacterial genes), but is still able to
multiply to a limited degree, thereby eliciting
a protective immune response in the absence
of disease symptoms.
One major problem associated with highly
purified antigens or subunit vaccines is re-
duced immunogenicity. Unfortunately, as pro-
tective antigens are purified from either bacte-
ria, viruses, or parasites in order to free them
from nontoxic substances such as LPS, they lose
their ability to evoke an immune response. Their
immunogenicity can, however, be improved by
using an adjuvant. Adjuvants function either to
present the antigen to the immune system for a
prolonged period of time, or to nonspecifically
stimulate the immune system by releasing im-
mune modulators such as lymphokines. To date,
only aluminum gels have been licensed as adju-
vants for human use. Tetanus or diphtheria tox-
oid is absorbed onto the gel which allows the
toxoid to persist longer at the injection site. Bio-
logically active peptides that can nonspecifically
stimulate the immune system via lymphokine
release or downregulate the suppression of the
immune system, are being evaluated.
1.2.5. Passive Immunization
Passive immunization entails the transfer of pre-
formed immunoglobulins to a host. Passive im-
munization is usually employed after known or
suspected exposure to a given pathogen (for ex-
ample, after being bitten by a rabid animal). It
is used in cases were disease progress is rapid
if the pathogen or toxin is not neutralized. The
time needed for the host to mount a protective
immune response following active immuniza-
tion (7 – 14 days) would be too long. In some
instances, globulin may be passively adminis-
tered as a prophylactic measure, as in the case
of travellers entering an area where hepatitis A
is endemic.
Immunoglobulin preparations are assayed to
confirm that they contain a high titer of neu-
tralizing antibodies against the disease in ques-
tion. Such hyperimmune globulins must contain
at least five-fold higher levels than normal glob-
ulin and are obtained by screening plasma units
(human) for a given antibody. Alternatively, vol-
8 Immunotherapy and Vaccines
unteers may be vaccinated to enrich their plasma
for a given antibody.
Immunoglobulin is administered either intra-
muscularly or intravenously. Only a limited vol-
ume of globulin can be given intramuscularly.
Intravenous administration allows for compara-
tively large quantities (400 mg of immunoglob-
ulin per kilogram body weight) of immunoglob-
ulin to be administered quickly.
Several problems are associated with the
preparation of immunoglobulin for passive ther-
apy. Firstly, the identification of plasma donors
who are “hyperimmune” to a given antigen
requires expensive and tedious screening pro-
grams. Secondly, the administration of im-
munoglobulins entails the possibility of trans-
mitting viral diseases, such as hepatitis B and
non-A, non-B hepatitis. Additional screening
procedures are therefore required.
An alternative to antibody obtained from
donors is the use of human monoclonal anti-
bodies (humabs, see Chap. 6). Humabs can be
synthesized by hybridoma cell lines that are pro-
duced by fusing a human B cell secreting a de-
sired antibody to a nonsecreting cell line. Hy-
bridomas can be grown in fermentors containing
up to 1000 L of serum-free medium. Antibody
yields can reach 100 mg per liter of medium. The
use of humabs can circumvent the need to obtain
plasma from many donors and the risk of viral
disease transmission, all at a lower cost. Humabs
have been produced against a variety of infec-
tious agents (e.g., cytomegalovirus, diphtheria
and tetanus toxins) and will be tested in the near
future.
1.2.6. Genetic Engineering
The application of recombinant DNA technol-
ogy to the field of vaccine development has led
to remarkable progress since the late 1970s. Crit-
ical protective antigenic determinants can be de-
fined, cloned, produced on a large scale by us-
ing an appropriate expression system (vector and
host), and purified. The first vaccine for human
use to be produced by genetic engineering is
that against hepatitis B; the surface antigen was
cloned, then expressed in yeast, and purified to
homogeneity. These vaccines have virtually re-
placed the first-generation, plasma-derived anti-
gen vaccines due to increased safety and lower
costs.
Numerous licensed, live-attenuated vaccines
(measles, polio, mumps, rubella, varicella, and
typhoid fever) have been obtained by rather em-
pirical techniques, such as passage on tissue cul-
ture or chemical mutagenesis. Surprisingly, the
precise genetic alterations responsible for aviru-
lence are unknown. Several live-attenuated bac-
terial vaccine strains have been developed by
using genetic engineering. The most advanced,
in regards to clinical testing, are for cholera
and typhoid fever. These strains have been at-
tenuated by inactivating a gene or genes es-
sential for virulence. A related approach has
been to clone a given protective antigen and ex-
press it in a suitable, nonvirulent “carrier” strain.
For example, surface glycoprotein from HIV-
1 and rabies virus have been introduced into
vaccinia virus (smallpox vaccine). In addition,
antigens from several enteric pathogens such
as Shigella sonnei, S. dysenteriae, and Vibrio
cholerae have been introduced into the licensed
live-oral typhoid vaccine strain, Ty21a. Such re-
combinant strains may prove useful as “biva-
lent” vaccines, conferring protection against two
enteric pathogens.
2. Bacterial Vaccines
The chemotherapy of bacterial dis-
ease is treated elsewhere (→ Antibiotics ;
→ Chemotherapeutics, Chap. 2.).
2.1. Diphtheria Vaccine
Etiological Agent and Pathogenesis. The
causative agent of diphtheria is Corynebac-
terium diphtheriae, first isolated by Loeffler
in 1884. The disease is spread by inhalation of
infected droplets and is characterized by the for-
mation of a pseudo-membrane in the nasophar-
ynx caused by localized bacterial replication.
Disease symptoms are caused exclusively by
the production of a lethal toxin which spreads
from the site of infection via the bloodstream.
Death is due to the inhibition of protein synthesis
by the toxin in vital organs [10].

History of Immunization. Studies by Roux
and Yersin in 1888 demonstrated that diphthe-
ria is a “toxicosis.” Two years later, Behring
and Kitasato established that the disease could
be prevented in animals by immunization with
a crude diphtheria toxoid. Attempts to extend
these findings to humans used a “toxoid” pre-
pared by combining diphtheria toxin and anti-
toxin. The first large-scale immunization pro-
gram with such a product was performed on New
York school children in 1922. Soon after, crude
toxoids prepared by formaldehyde treatment of
C. diphtheriae filtrates replaced toxin-antitoxin
toxoids. Routine mass vaccination against diph-
theria was initiated in Western Europe and North
America shortly after World War II.
Production and Properties of Diphthe-
ria Vaccine. Diphtheria toxoid is prepared by
detoxification of diphtheria toxin with formalde-
hyde. Derivatives of the hypertoxinogenic Park
Williams 8 strain are used by most manufac-
turers. Large quantities of toxin are produced
in fermentors (ca. 200 – 5000 L) and yields ap-
proach 500 mg/L. Although manufacturing pro-
cesses vary, most employ the following steps.
Formaldehyde [50-00-0] (0.4 – 0.6 %) is added
to cell-free culture supernatants which are then
stored at 35 – 37 ◦ C for 3 – 5 weeks. Detoxifica-
tion is accomplished first by reaction of form-
aldehyde with the ε-amino groups of lysine
and then by formation of irreversible methylene
bridges between aromatic amino acids. Upon
confirmation of detoxification by animal test-
ing, the toxoid is purified by diafiltration, ammo-
nium sulfate or ethanol fractionation, or anion-
exchange chromatography.
Immunization Recommendations. Diph-
theria toxoid is rarely used as a monovalent
vaccine, but is usually combined with tetanus
toxoid (DT, see Section 2.2), pertussis vaccine
(DPT, see Section 2.3), or inactivated polio vac-
cine (DPT-Pol, see Section 3.5) absorbed onto
an aluminum salt adjuvant. Toxoid content is
usually expressed in limits of flocculation, Lf
(1 Lf ≈ 2 µg of toxoid). For primary immuniza-
tion of infants, ca. 25 Lf of toxoid is adminis-
tered intramuscularly starting at 6 – 12 weeks
of age. Three doses are usually administered at
4 – 8 week intervals with a fourth dose 4 – 12
months later. To maintain life-long immunity,
Immunotherapy and Vaccines 9
booster doses, together with tetanus toxoid, are
recommended every 7 – 10 years.
Adverse Reactions. The absolute reacto-
genicity of diphtheria toxoid is difficult to deter-
mine since it is usually administered with tetanus
toxoid or pertussis vaccines. Overall, the vaccine
is considered to be safe; it primarily evokes mild,
transient, local reactions.
Vaccine Efficacy. No large-scale field trials
have been performed to determine vaccine ef-
ficacy. This is because it would be unethical to
withhold the vaccine from control subjects in
view of the dramatic decrease in disease fol-
lowing immunization in the 1920s. Furthermore,
overwhelming data indicate that proper vacci-
nation can provide absolute protection. For ex-
ample, mass vaccination with diphtheria toxoid
in Romania in 1958 showed that within seven
years morbidity and mortality due to diphtheria
declined by more then 99 %. In Western Europe
and North America, where vaccination is univer-
sal, diphtheria has been virtually eradicated [11].
Rare cases occur almost exclusively in individ-
uals lacking a proper history of vaccination.
Future Prospects. Given the combination of
vaccine safety, efficacy, and cost, there has been
little impetus to develop and test a “second gen-
eration” diphtheria toxoid. Any such vaccine
would have to completely eliminate the possibil-
ity of toxic reversion and be more economical to
produce. In another approach, synthetic peptides
that express key epitopes of diphtheria toxin are
conjugated to carrier proteins; these conjugates
can elicit neutralizing antibodies [12]. The draw-
back with this approach is the cost of purifying
the carrier protein.
The most promising approach to toxoid de-
velopment is the use of nontoxic mutant proteins
[13], obtained by recombinant DNA technology.
Such proteins can be constructed by deletion
of specific regions responsible for toxicity [14].
In addition, these “toxoids” can be synthesized
in yields equal to those obtained by production
strains of C. diphtheriae. The efficacy of such
proteins is being evaluated.
10 Immunotherapy and Vaccines
2.2. Tetanus Vaccine
Etiological Agent and Pathogenesis. The
causative agent of tetanus is the spore-forming
bacillus, Clostridium tetani. The disease, first
described by Hippocrates, is spread via con-
tamination of wounds or abrasions with soil
containing C. tetani spores. Bacteria multiply-
ing at the wound site release a potent neurotoxin
which enters the bloodstream and acts upon
the central nervous system evoking the “spas-
tic paralysis” characteristic of tetanus [15]. The
majority of tetanus cases occur in neonates due
to the nonsterile severing of the umbilical cord.
History of Immunization. The identifica-
tion of tetanus as a toxicosis by Faber in 1890
led to the development of crude tetanus toxoids
as early as 1893. Initial attempts to vaccinate
against tetanus were carried out during World
War I. By 1926, parenteral immunization of hu-
mans with a safe, effective formalin toxoid was
achieved. Until the mid-1960s, only such crude
toxoids were available.
Production and Properties of Tetanus Vac-
cine. The presently used tetanus toxoid is a
partially purified preparation. The hypertoxino-
genic Harvard strain of C. tetani is most widely
used for production of tetanus toxin. The bacte-
ria are grown in fermentors until the cells autol-
yse, thereby releasing the toxin into the medium.
Intact cells are removed by filtration. Formalde-
hyde is added to the filtrate to a final concentra-
tion of 0.4 – 0.6 %, the pH adjusted to 7.4 – 7.6,
and the mixture held at 35 – 37 ◦ C for about four
weeks. After confirmation of detoxification by
animal testing, tetanus toxoid is purified by di-
afiltration and ammonium sulfate fractionation.
Immunization Recommendations.
Tetanus toxoid is most frequently adsorbed to
an aluminum salt adjuvant and administered to-
gether with diphtheria toxoid (DT, Section 2.1),
pertussis vaccine (DPT, Section 2.3), or inac-
tivated polio vaccine (DPT-Pol, Section 3.5).
For primary immunization of infants, each dose
of vaccine contains 20 – 30 Lf’s of tetanus tox-
oid as DPT or DPT-Pol. Immunization starts at
6 – 12 weeks of age and consists of three doses
given at 4 – 8 week intervals with a fourth dose
4 – 12 months later. To maintain life-long im-
munity (≥ 0.01 IU/mL of serum), booster doses,
often combined with diphtheria toxoid, are rec-
ommended every 7 – 10 years.
Tetanus vaccine is often administered as a
routine prophylactic measure following punc-
ture wound trauma. This practice can lead to
hyperimmunization and attendant reactions. Al-
ternatively, an individual may be “tolerized” by
repeated vaccination and, therefore, unable to
mount a protective immune response. Vaccina-
tion of pregnant women is recommended with
the final dose given about three weeks before
expected delivery. The immune response of hu-
mans to tetanus vaccine varies widely [16]. The
immunization schedule used appears to influ-
ence the magnitude of the immune response
[17].
Adverse Reactions. Serious reactions fol-
lowing vaccination with tetanus toxoid are infre-
quent: swelling, pain, and/or redness at the injec-
tion site are the most common. However, imme-
diate and delayed hypersensitivity, an Arthus-
like reaction, and in rare instances (< 1 case in
2×106 ), neurological sequelae have been noted
[15]. The majority of severe reactions are associ-
ated with high levels of pre-existing antitetanus
antibody. Therefore, attention should be paid to
the interval between booster doses to avoid these
occurrences.
Vaccine Efficacy. Data supporting the effi-
cacy of tetanus toxoid vaccine comes primarily
from retrospective analysis of attack rates in im-
mune versus nonimmune populations. The first
such study was the evaluation of soldiers during
World War II, where tetanus occurred primar-
ily in nonimmune individuals. Immunization of
pregnant women completely prevented neona-
tal tetanus [18]. Recently, cases of tetanus ob-
served in the United States occurred exclusively
in nonimmune individuals. Numerous studies
have also shown that abbreviated one- or two-
dose immunization regimens are able to induce
long-lasting immunity [15]. This is of partic-
ular importance in developing countries where
a multidose immunization regimen may not be
feasible due to an inadequate health care system
or economic constraints.

Future Prospects. As with diphtheria tox-
oid, tetanus vaccine is an extremely economic,
effective, safe vaccine. Efforts directed to a “new
generation” of tetanus vaccine are aimed at pro-
viding a toxoid more amenable to mass vacci-
nation. The main thrust has been to produce a
safer, more immunogenic toxoid capable of in-
ducing long-lasting immunity after one or two
administrations without reactogenic adjuvants.
The substitution of glutaraldehyde for formal-
dehyde yields a safe, immunogenic vaccine not
requiring adjuvants [19]. Production of toxoids
by recombinant DNA techniques similar to those
described for diphtheria toxin is also feasible
[14].
2.3. Pertussis Vaccine
Etiological Agent and Pathogenesis. Bor-
detella pertussis, the cause of pertussis or
“whooping cough,” is a gram-negative rod that
was first isolated by Bordet and Gengou in
1906. Humans are the only known host for
B. pertussis. The disease is spread by infectious
droplets with the vast majority of cases occurring
in young children. B. pertussis shows a marked
trophism for the ciliated epithelial cells of the up-
per respiratory tract. Initial symptoms are sim-
ilar to those of the common cold, but worsen
within 10 – 20 days. The paroxysmal stage, last-
ing an average of 15 – 20 days, is characterized
by severe bouts of coughing. The convalescent
stage can last for many months when secondary
infections can present a major problem. B. per-
tussis remains localized within the initial dis-
ease stages which implies that the severe symp-
toms are toxin-induced [20]. This is supported
by the fact that B. pertussis can synthesize a large
number of toxic extracellular factors including
pertussis toxin (also referred to as lymphocy-
tosis promoting factor), adenylate cyclase, der-
monecrotic toxin, and tracheal cytotoxin [21].
History of Immunization. Attempts at im-
munization against pertussis were initiated in
the 1920s using killed whole-cell vaccines [22].
These studies demonstrated that vaccination
could not only prevent a substantial proportion
of disease, but that symptoms in vaccinated indi-
viduals who became infected were milder than in
Immunotherapy and Vaccines 11
unvaccinated controls. Routine large-scale im-
munization against pertussis began shortly after
World War II using standardized whole-cell vac-
cines of known potency. Even though the safety
of these vaccines has been under attack, they
are still used in most areas of the world except
Japan, where acellular vaccines have been used
since 1980 (see below).
Production and Properties of Pertussis
Vaccine. At present, there is no standardized
method, culture medium, or bacterial strain for
producing pertussis vaccine. A given manufac-
turer’s production procedure is designed to yield
a vaccine which will meet the minimal require-
ments of the appropriate regulatory agency. B.
pertussis is usually grown in fermentors on a
synthetic or semisynthetic medium. The cells
are harvested by centrifugation and resuspended
to a given opacity using a reference standard.
Inactivation is accomplished by heating and/or
adding formaldehyde or thimerosal. In view
of the diversity of production techniques em-
ployed, toxic components which escape inacti-
vation can vary considerably from manufacturer
to manufacturer.
Immunization Recommendations. Pertus-
sis vaccine is used almost exclusively in combi-
nation with diphtheria and tetanus vaccines (see
Sections 2.1 and 2.2). The vaccine is adsorbed
onto an aluminum salt adjuvant. Immunization
usually commences at 6 – 12 weeks of age and
consists of three doses of vaccine given intra-
muscularly at 4 – 8 week intervals with a fourth
dose given 4 – 12 months later. Vaccination with
the acellular vaccine used in Japan consists of
2 – 3 doses of vaccine given at 4 – 8 week inter-
vals starting at two years of age.
Adverse Reactions. The majority of chil-
dren receiving pertussis vaccine experience an
adverse reaction. Approximately 40 – 50 % of
children have a local reaction and 30 – 40 %
a systemic reaction (anorexia, vomiting, fret-
fulness, fever, or persistent crying) [23]. Of
greater concern is the infrequent occurrence
of convulsions and hypotonia. As a result of
concern regarding vaccine safety, a large-scale
study was initiated in England to determine
the incidence of serious reactions in children
aged 2 – 36 months [24]. The risk of vaccine-
12 Immunotherapy and Vaccines
induced neurological illness was estimated to
be 1/110 000 vaccinations and that for en-
cephalopathy, 1/310 000 vaccinations. It is ex-
tremely difficult to establish a causal relation-
ship in individual cases in such a study; inci-
dence rates were evaluated on a temporal basis.
Furthermore, neurological symptoms believed
to be vaccine-related constitute only a small pro-
portion of similar cases seen in this age group.
Review of these data by health authorities has
led to the conclusion that the benefit of vacci-
nation outweighs the attendant risk in view of
the fact that clinical pertussis can often result in
neurological sequela.
Vaccine Efficacy. The first indication of vac-
cine efficacy came from a trial conducted in 1929
during an epidemic on the Faroe Islands; vac-
cination afforded 73 % protection against clini-
cal disease [22]. Extensive information has been
obtained in Japan, England, and Sweden where
disease incidence correlates inversely with the
overall immune status of the general population.
Routine immunization against pertussis in
Japan was initiated in 1947 – 1949 with an ac-
ceptance rate of roughly 90 %. The number of
pertussis cases declined from 152 072 in 1947
to less than 400 by 1971. In 1975, controversy
concerning vaccine safety briefly halted vaccine
usage. When it was resumed, the acceptance rate
declined to ca. 25 – 30 %. Pertussis returned to
epidemic proportions by 1979 with more than
13 000 cases reported nationally. At the urging
of federal health authorities, vaccine acceptance
rates increased to 65 – 70 % by 1982 with a con-
comitant decline in cases.
Similarly, concern about vaccine safety in
England resulted in a dramatic decline in vaccine
acceptance from ca. 80 % in 1973 to ca. 30 % in
1978. As in Japan, the disease became epidemic
with more than 100 000 cases reported between
1977 and 1980. In Sweden, routine vaccination
against pertussis was initiated in the early 1950s.
One decade later, > 90 % of children were con-
sidered to be immune. Due to changes in the
manufacturing technique, the pertussis vaccine
used in the mid-1970s was not potent. Shortly
after, a marked increase of pertussis cases was
seen. When an effective vaccine was reintro-
duced, disease incidence was dramatically re-
duced.
Case-contact studies have shown the whole-
cell vaccine to be 63 – 95 % effective at prevent-
ing overt disease. Efficacy of acellular vaccines
is discussed below.
Future Prospects. Efforts to develop a “sec-
ond generation” pertussis vaccine have centered
on using purified detoxified antigen preparations
containing a minimum amount of lipopolysac-
charide. Several such acellular vaccines have
been developed and clinically evaluated. They
are composed of detoxified pertussis toxin
(PT) alone or in combination with filamentous
hemagglutinin (FHA), two key protective anti-
gens [25]. The first generation acellular vaccines
were produced in Japan using supernatant from
static cultures as a source of antigen. Cell-free
supernatants were subjected to ammonium sul-
fate precipitation followed by sucrose density
gradient ultracentrifugation to simultaneously
enrich PT and FHA and eliminate lipopolysac-
charide; PT was inactivated by formaldehyde
treatment. Preliminary testing in Japan showed
the vaccine to be far better tolerated than whole-
cell vaccine; case-contact studies revealed that
it was ca. 90 % effective.
Second generation acellular vaccines of a
greater purity have been produced on a large
scale using fermentor-grown cultures. A mono-
valent formalin PT toxoid and a bivalent PT
toxoid – FHA vaccine have been evaluated in a
placebo-controlled trial in 5 – 11 month old chil-
dren in Sweden [26]. Children received 2 doses
of vaccine 8 – 12 weeks apart. The small num-
ber of children vaccinated (ca. 3000) does not
allow evaluation of vaccine safety regarding the
rare neurological reactions. However, the vac-
cine evoked far fewer local reactions in compar-
ison to the whole-cell vaccine. After 15 months
of observation, the PT toxoid vaccine was 54 %
effective whereas the PT toxoid-FHA vaccine
was 69 % effective against all forms of clinical
pertussis. Both vaccines were equally effective
(ca. 80 %) against severe disease.
2.4. Typhoid Fever Vaccine
Etiological Agent and Pathogenesis.
Salmonella typhi, the causative agent of typhoid
fever, is a gram-negative bacillus first isolated

by Gaffky in 1884. Infection is due to inges-
tion of the organisms in contaminated food or
water. The bacteria penetrate the epithelium of
the small intestine and are ingested by reticu-
loendothelial cells. Unlike most bacteria, the ty-
phoid bacillus can survive and multiply within
phagocytic cells and then spreads to the spleen,
liver, lymph nodes, and gallbladder. Symptoms
appear 1 – 2 weeks after exposure when the bac-
teria enter the bloodstream. Despite appropriate
treatment, 1 – 2 % of those infected will become
chronic asymptomatic carriers serving as an in-
fectious reservoir. Typhoid primarily occurs in
developing countries or in areas of poor sani-
tation. In endemic areas, the disease is primar-
ily contracted by school-aged children, but trav-
ellers of all ages entering an endemic area are at
risk.
History of Immunization. Attempts at im-
munizing against typhoid fever began in 1896.
The first mass vaccination was conducted by the
British Army during World War I using a killed
whole-cell vaccine. The attack rate among vac-
cinated soldiers was far less than for nonimmu-
nized soldiers. Consequently, the use of whole-
cell vaccines became common practice in the
armed forces.
In the 1960s and 1970s, attempts were made
to use killed oral immunization. Killed bacte-
ria were administered to volunteers at 1×1011
per dose; twelve doses resulted in a modest, but
significant protection rate of 30 % [27]. Field tri-
als in India showed that such killed preparations
were ineffective at preventing disease.
An alternative approach to killed oral vac-
cines against typhoid was to use attenuated live-
oral vaccine strains. The first candidate was
a streptomycin-dependent mutant developed in
1967 [28]. When orally administered to volun-
teers in doses of up to 1011 bacteria, this strain
did not evoke adverse reactions. Good levels
of protection (66 – 78 %) were achieved with
freshly harvested cultures. However, lyophilized
preparations afforded little or no protection [28],
which diminished interest in this strain.
The second live-oral vaccine candidate was a
gal ε mutant termed Ty21a which lacked glucose
1,4-epimerase activity and therefore could not
synthesize complete lipopolysaccharide [29].
Freshly harvested cultures of this strain were
safe and effective in volunteer challenge studies
Immunotherapy and Vaccines 13
[30]. A lyophilized Ty21a preparation afforded
> 90 % protection for three years in a field trial
in Egypt [31].
Production and Properties of Typhoid
Vaccines. Killed whole-cell vaccines are usu-
ally produced from the Ty2 strain of S. typhi.
Fermentor-grown cultures are inactivated either
by heating in combination with phenol or form-
aldehyde or by acetone drying. Although the lat-
ter method yields a more efficacious vaccine (see
below), only the vaccine produced by the former
method is readily available. Each 0.5 mL dose
of vaccine contains (1 – 3)×109 killed organ-
isms. Vaccine potency is estimated by a mouse-
protection test.
The attenuated live-oral Ty21a vaccine strain
is produced from fermentor-grown cultures. The
harvested cells are suspended in a cryoprotec-
tive medium consisting of sugar and amino acids
and lyophilized. The lyophilizate is placed in
gelatin capsules which are then coated with
an acid-resistant layer. Each capsule contains
(1 – 5)×109 viable bacteria. Vaccine potency is
based upon the number of viable organisms per
capsule [32].
Immunization Recommendations. People
living in or travelling to an endemic area are
vaccinated with two doses of whole-cell vac-
cine administered subcutaneously or intramus-
cularly 14 days apart. A single booster dose is
advised for individuals travelling to an endemic
area if 2 – 3 years have elapsed since primary
immunization.
Three doses of attenuated live-oral vaccine
are ingested, each dose on an alternate day. An
additional complete immunization course is ad-
vised when entering an endemic area if more
than 1 – 2 years have elapsed since primary im-
munization. The vaccine is for use in children
six years of age or older and in all ages of non-
immunocompromised adults.
Adverse Reactions. Reactions following
immunization with the parenteral whole-cell
vaccine are frequent and consist of local pain,
swelling, and redness, often accompanied by
fever, chills, and malaise. Transient debilitating
reactions occur in ca. 10 % of vaccine recipi-
ents. Due to the high rate of adverse reactions,
parenteral typhoid vaccine is not widely used.
14 Immunotherapy and Vaccines
Reactions following ingestion of the live oral
Ty21a vaccine are rare (< 1 per 100 000 doses).
Reactions are usually mild, transient, and re-
solve of their own accord. Gastrointestinal dis-
turbances and rash are most frequently reported.
Vaccine Efficacy. To accurately assess the
efficacy of parenteral typhoid vaccine, several
field trials were conducted in the 1960s [33].
In most instances, two doses of either heat–
phenol or acetone-dried vaccine were admin-
istered. Protection rates ranged from 51 – 77 %
and 79 –93 %, respectively, for up to seven years.
A further trial was conducted in Tonga in 1966 –
1973 where school-aged children received one
or two doses of acetone-dried vaccine. Surpris-
ingly, one dose gave no protection whereas two
doses afforded only 40 % protection.
Efficacy of the live oral Ty21a vaccine ad-
ministered in enteric-coated capsules has been
evaluated in field trials in school-aged children
in Chile [34]. Three doses of vaccine provided
about 70 % protection over a three-year period.
Future Prospects. Two additional typhoid
vaccines are undergoing clinical evaluation. The
first is a double auxotrophic mutant of S. typhi re-
quiring 4-aminobenzoate and adenine [35]. Al-
though ingestion of up to 1010 organisms only
evoked a poor humoral immune response to
S. typhi cellular antigens, most vaccines mani-
fested a specific cell-mediated immune response
to S. typhi lipopolysaccharide.
The second candidate vaccine is a capsular
polysaccharide termed Vi . Purified Vi antigen
is safe upon parenteral immunization. A single
dose of vaccine provided ca. 70 % protection for
ca. one year in Nepal [36].
2.5. Streptococcus pneumoniae Vaccine
Etiological Agent and Pathogenesis. Strep-
tococcus pneumoniae, often referred to as
“pneumococcus,” is a gram-positive diplococ-
cus first isolated by Sternberg and Pasteur
in 1881. The pneumococcus usually causes dis-
ease subsequent to a viral infection or in a
compromised individual, i.e., is an opportunistic
pathogen. Pneumonia is the most common dis-
ease syndrome caused by S. pneumoniae; bac-
teremia and meningitis are often sequelae to
pneumococcal pneumonia. Streptococcus pneu-
moniae is also a leading cause of otitis media in
young children. The pneumococcus often colo-
nizes the nasopharynxes of healthy individuals
which then serve as infectious foci. The organ-
isms can then spread to another person by in-
fected droplets or translocate to the inner ear or
lungs following a viral infection.
History of Immunization. Early attempts
at immunization against pneumococcal pneu-
monia employed heat-killed, whole-cell vac-
cines (reviewed in [37] ). These trials and re-
lated studies led to the conclusion that immunity
is mediated by serospecific antibody produced
against the polysaccharide capsule surround-
ing the pneumococcus. A large-scale trial per-
formed in the mid-1930s with a bivalent capsular
polysaccharide (CPS) vaccine showed a modest
decrease in disease. Conclusive proof of vaccine
efficacy came from trials conducted with polyva-
lent vaccines; protection was serospecific. Fur-
ther studies showed that vaccines comprising
up to six distinct capsular antigens could be
effectively used [38]. The first pneumococcal
polysaccharide vaccines were hexavalent and
were licensed in the early 1950s. However, ow-
ing to the belief that antibiotics could effectively
control pneumococcal disease, vaccine accep-
tance was low, leading to a halt in production.
Production and Properties of Streptococ-
cus pneumoniae Vaccine. This vaccine is com-
posed of 23 types of purified CPS that have
been selected on the basis of seroepidemiolog-
ical surveillance studies of bacteremic isolates
[39].
Pneumococcal CPSs are purified by a va-
riety of techniques, depending upon serotype,
which include precipitation with organic sol-
vents, treatment with detergent, digestion with
DNAse, RNAse, and protease, and ultracentrifu-
gation. The vaccine contains only traces of pro-
tein and nucleic acids (≤ 2 wt %) and is analyzed
for its CPS constituents, serological purity, and
pyrogenicity. Vaccine potency is based upon the
determination of the molecular mass of the CPS,
because a minimum antigen size is required for
immunogenicity in humans. One human dose
consists of 25 µg of each antigen in 0.5 mL ad-
ministered intramuscularly or subcutaneously.

Immunization Recommendations. Pneu-
mococcal vaccine is recommended for immuno-
competent individuals with underlying clinical
conditions that increase their risk of acquiring
pneumococcal bacteremia or pneumonia. These
conditions include alcoholism, asplenia, sickle
cell anemia, reduced pulmonary function, con-
gestive heart failure, diabetes, reduced renal
function, and cirrhosis [40]. Routine vaccina-
tion of the elderly is also warranted due to the
increased incidence of disease in this group. Pa-
tients with underlying conditions (leukemia,
Hodgkin’s disease, myeloma, and treatment
with immunosuppressive drugs) that render
them susceptible to pneumococcal disease may
be considered candidates for vaccination even
though a proportion may not respond to immu-
nization [40]. Revaccination is not advised due
to severe reactions.
Adverse Reactions. Pneumococcal vaccine
is very safe [40]. Mild local reactions occur in
ca. 30 – 40 % of vaccine recipients, fever and se-
vere local reactions in ca. 1 %. Anaphylactic re-
actions are extremely rare with a frequency of
5/1×106 doses administered. Severe local and
systemic reactions are more likely to occur upon
revaccination.
Vaccine Efficacy. Considerable controversy
exists concerning the degree of protection pro-
vided by pneumococcal vaccine. The vaccine
was highly effective in reducing the incidence of
pneumococcal pneumonia among South African
gold miners [41]. However, this population dif-
fers substantially in age and in general health
from those patients for whom the vaccine is now
recommended. Several studies have attempted
to determine vaccine efficacy among debilitated
and/or elderly patients. The distribution of S.
pneumoniae serotypes among vaccinated and
nonvaccinated individuals with invasive pneu-
mococcal disease has been compared [42]; effi-
cacy was estimated to be 61 – 65 %. In contrast,
a vaccine trial on high-risk patients, aged > 55,
did not demonstrate any significant benefit [43],
probably due to the inability of these patients
to mount or maintain a protective antibody re-
sponse. Similar findings were reported in a study
on patients suffering from chronic pulmonary
disease [44]. In spite of these divergent findings,
Immunotherapy and Vaccines 15
health authorities still strongly recommend im-
munization based upon risk–benefit analysis.
Future Prospects. The current pneumococ-
cal vaccine is poorly immunogenic in children
under two years of age and in certain debili-
tated patient populations [40]. Attempts have
been made to improve immunogenicity by co-
valently coupling pneumococcal CPS to carrier
proteins [45]. Such conjugates evoked a more
vigorous response in mice and rhesus monkeys
when compared to native CPS [46]. However, a
type 6A-tetanus toxoid conjugate administered
to young adult volunteers was only slightly more
immunogenic than native 6A-CPS, and a booster
dose did not significantly increase anti CPS anti-
body levels [47]. However, these volunteers had
been previously immunized with tetanus toxoid
and possessed low, but detectable, levels of an-
tibody to 6A-CPS.
2.6. Shigella Vaccines
Etiological Agent and Pathogenesis.
Shigellosis or bacillary dysentery is caused by
Shigellae, most notably S. sonnei, S. flexneri 2 a
and 3, and S. dysenteriae 1. Disease results from
ingestion of as few as 10 bacteria, making it
the most infective bacterial enteric pathogen. In
developing countries, where the attack rate can
exceed 100 000/1×106 population, shigellosis
is primarily a disease of young children aged six
months to six years.
Once ingested, the organisms penetrate the
epithelium of the colon and rapidly multiply.
This leads to localized inflammation and ulcer-
ation, possibly due to synthesis of a potent cyto-
toxin [48]. Symptoms include fever, bloody di-
arrhea, cramps, tenesmus, and shock. The mor-
tality rate for untreated disease can exceed 10 %.
History of Immunization. Immunization of
humans with killed whole-cell vaccines admin-
istered parenterally does not provide signifi-
cant protection against shigellosis [49]. Sub-
sequent attempts at immunization have there-
fore centered on use of attenuated live-oral vac-
cine strains. Vaccines based on streptomycin-
dependent derivatives of S. flexneri and S. sonnei
are well tolerated and provide significant immu-
nity for 6 – 12 months [50], [51]. However, these
16 Immunotherapy and Vaccines
vaccines were not pursued due to the necessity
of administering four doses for primary immu-
nization, yearly boosting to maintain efficacy,
and genetic instability.
Attempts have been made to produce live-
oral shigella vaccines by conjugal transfer of
genes coding for S. flexneri surface antigens into
E. coli. Multiple doses evoked no significant ad-
verse reactions when fed to volunteers, but failed
to protect against subsequent challenge [52].
The identification of critical Shigellae-
protective antigens has allowed their expression
in the attenuated live-oral typhoid vaccine strain,
S. typhi Ty21a [53]. One strain, S. typhi Ty21a-
5076-1C, carries a plasmid coding for the form I
(O-polysaccharide) antigen of S. sonnei. This
strain is safe and has afforded significant pro-
tection in volunteer studies [54].
Future Prospects. At present, no vaccines
are available for use against shigellosis. As noted
above, S. typhi 5076-1C that expresses S. sonnei
form I antigen has shown promise in volunteer
studies. However, efficacy varies from lot to lot
and probably depends on the degree of flagella-
tion of the bacteria. The construction of a S. typhi
Ty21a strain expressing S. flexneri 2a type and
group antigens has been described [55], but this
strain has not yet been clinically evaluated.
Genes coding for the production of the O-
antigen of S. dysenteriae 1 have been cloned and
inserted in a plasmid [56]. The S. dysenteriae
O-antigen can be expressed in E. coli K12 and
S. typhi Ty21a by introduction of the recom-
binant plasmid. Volunteer studies to determine
their safety are scheduled.
2.7. Cholera Vaccine
Etiological Agent and Pathogenesis. The
causative agent of cholera is Vibrio cholerae,
a gram-negative motile bacillus which is trans-
mitted via the fecal–oral route. Upon passage
through the stomach acid barrier, V. cholerae
colonizes the ileum where it rapidly multiplies.
The release of a potent enterotoxin (cholera
toxin) causes massive watery diarrhea [57]. In
most cases, oral or intravenous fluid replacement
is sufficient treatment.
History of Immunization. Parenteral
whole-cell cholera vaccines were employed by
Fenan as early as 1885. Starting in the 1960s,
several field trials were performed using inac-
tivated whole-cell vaccines or cholera antigens
(see “Vaccine Efficacy”).
Numerous attempts have been made to orally
vaccinate against cholera. Multiple oral doses
of heat-killed V. cholerae evoked both a local
and humoral antibody response [58]. Ten daily
oral doses of 1.6×1010 killed vibrios provided
protection against a homologous challenge [59].
However, parenteral vaccine gave slightly bet-
ter protection. Although promising, the massive
quantities of organisms used and the need for
multiple doses made the oral approach expen-
sive and cumbersome.
Several environmental isolates of V. cholerae
with reduced virulence have been evaluated as
live-oral vaccines. These strains failed to pro-
vide significant protection due to their inability
to colonize the small intestine and evoke a pro-
tective immune response [60]. A hypotoxino-
genic mutant of V. cholerae isolated by chemi-
cal mutagenesis afforded significant protection
against diarrhea. However, due to its propen-
sity to revert, it was not studied further [61].
Chemical mutagenesis has provided a strain of
V. cholerae 3083 (Texas Star-SR), that produces
only the B subunit of cholera toxin responsible
for binding to cells and lacks the enzymatically
active A subunit. Although vaccination with this
strain afforded good levels of protection against
cholera, about 25 % of vaccinees experienced
mild to moderate diarrhea [62].
Production and Properties of Cholera Vac-
cine. The licensed cholera vaccine is composed
of inactivated whole cells of V. cholerae and
is administered parenterally. Fermentor-grown
cultures are inactivated by a combination of
heat and phenol or formaldehyde. Usually,
two cultures of Ogawa-classical and Inaba-El
Tor serotypes-biotypes are produced and then
mixed. Cell concentration is adjusted to (5 –
8)×109 cells/dose (0.5 mL). Vaccine potency is
estimated by an intraperitoneal mouse challenge
test. Two doses of vaccine given at two-week in-
tervals are recommended. A booster dose is rec-
ommended whenever entering an endemic area.

Adverse Reactions. Approximately 20 –
30 % of vaccinees have a mild to moderate
local reaction consisting of pain, redness, and/or
swelling. Systemic reactions such as fever,
headache, or malaise occur in ca. 5 % of vacci-
nees.
Vaccine Efficacy. Various parenteral vac-
cines, including inactivated whole-cell, purified
lipopolysaccharide, cell-free Inaba antigen, and
a whole-cell vaccine in adjuvant have been eval-
uated in field trials [58]. The results can be sum-
marized as follows:
1) two doses of vaccine are superior to one;
2) protection afforded by vaccination is greater
in adults than children;
3) vaccination is only moderately effective in
nonendemic areas; and
4) protection lasts for only 3 – 6 months.
Therefore, the currently available cholera
vaccine is not an effective tool to control en-
demic cholera, it is best employed by travellers
entering an endemic area for a short time.
Future Prospects. Due to advances in under-
standing the pathogenesis and molecular genet-
ics of V. cholerae, one or more safe and effective
cholera vaccines will probably be introduced in
the near future.
Several attenuated live-oral vaccine strains
have been developed by deleting the cholera
toxin gene or its enzymatically active A subunit
[63], [64]. A derivative of V. cholerae 16961, in
which the cholera toxin gene was deleted af-
forded 89 % protection against clinical disease
although 50 % of the volunteers had mild to
moderate diarrhea. A second vaccine strain was
produced by deletion of the A subunit of cholera
toxin from strain 395. Again, good protection
was afforded by a single dose, but mild diarrhea
was observed in 60 % of the subjects. The dis-
covery, that strains of V. cholerae can produce
a Shiga-like cytotoxin suggested that at least
part of the diarrhea could be due to this toxin
[65]. Therefore, Kaper and coworkers (personal
communication) deleted the A subunit of cholera
toxin from strain 569B which is naturally cyto-
toxin negative. This strain, called CVD-103, is
far less reactinogenic than previously evaluated
strains, eliciting mild diarrhea in ∼ 10 % of sub-
jects. Vaccination with CVD-103 afforded ex-
Immunotherapy and Vaccines 17
cellent protection (87 %) against a homologous
challenge. Furthermore, significant protection
(67 – 78 %) was seen even when the challenge
was of a different serotype or biotype.
Orally administered inactivated vaccines
consisting of 1×1011 killed vibrios and 1 mg of
the B subunit of cholera toxin or 1×1011 killed
vibrios alone per dose have been evaluated [66].
After six months of surveillance, three doses of
the combined vaccine afforded 85 % protection,
whereas the cells alone were 58 % effective. Af-
ter one year, protection had declined to 62 %
for the combined vaccine and 53 % for the cells
alone.
2.8. Vaccines Against Nosocomial
Pathogenes
The routine use of broad-spectrum antibiotics,
more frequent invasive surgery, increasing use of
immunosuppressive agents in the management
of cancer, and the ability to prolong the lives
of critically ill patients have led to a dramatic
increase of hospital-acquired (nosocomial) bac-
terial infections [67]. In the United States alone,
nosocomial infections are a contributing fac-
tor in ca. 70 000 deaths per year; their treat-
ment costs > 1×109 $ [68], [69]. The mortality
rate for nosocomial bacteremia and pneumonia
has remained unacceptably high (≥ 25 %) de-
spite the introduction of numerous antibacterial
agents [68]. Therefore, much effort has been de-
voted to developing immunological agents for
controlling these infections [70].
Etiological agents and pathogenesis. Pseu-
domonas aeruginosa, Escherichia coli, Kleb-
siella spp., and Staphylococcus aureus account
for the majority of serious nosocomial infec-
tions, such as bacteremia and pneumonia [71].
Infection with these agents usually arises due to
the translocation of bacteria from the skin, intes-
tine, or nasopharyngeal cavity following disrup-
tion of the host’s defense systems by trauma, in-
vasive surgical procedures, treatment with anti-
biotics, or immunosuppressive agents. Foci of
infection are usually formed from which the bac-
teria enter the bloodstream. Death is most likely
caused by shock.
18 Immunotherapy and Vaccines
Future Prospects. No vaccines are licensed
against the four pathogens listed above. At-
tempts to control bacterial nosocomial infec-
tions by immunological means are complicated
because of the diverse patient populations who
are at high risk and the substantial number
of bacterial pathogens of varying serotypes in-
volved. Furthermore, vaccination of at-risk pa-
tients may not be feasible because of the time
needed to mount a protective immune response
and because a substantial portion of these pa-
tients are unable to mount a significant anti-
body response. A more effective approach is
to passively immunize patients by using a hy-
perimmune globulin for intravenous use (see
Chap. 5): vaccines developed against nosoco-
mial pathogens could then be used to vacci-
nate healthy donors whose plasma would be pro-
cessed into such a globulin [70].
Two different approaches are being taken
to develop appropriate hyperimmune products.
The first is based upon developing vaccines
against the relevant serospecific bacterial anti-
gens. The second is to evoke an antibody re-
sponse to a common epitope expressed by the
lipopolysaccharide of gram-negative bacteria.
Serospecific Vaccine against Pseudomonas
aeruginosa. Human immunity to P. aeruginosa
depends on the presence of humoral anti-
body directed against serospecific lipopolysac-
charide (LPS) determinants and toxin A [72].
Several lipopolysaccharide-containing vaccines
have been developed and clinically evaluated
[73]. These vaccines have not gained wide ac-
ceptance due to their toxicity, poor immuno-
genicity, and poorly characterized antigenic con-
tent. Two serotypes of P. aeruginosa high molec-
ular mass polysaccharides have been purified
and found to be safe and immunogenic in hu-
mans [74]. In an attempt to produce both an anti-
toxin A and an antiLPS antibody response, toxin
A – O-polysaccharide conjugate vaccines were
synthesized. Such conjugates are nontoxic, non-
pyrogenic, safe when administered to humans,
and evoke an immune response to both vac-
cine moieties [75]. A polyvalent vaccine based
upon O-polysaccharide serotypes is undergoing
clinical evaluation; this vaccine would “cover”
>90 % of P. aeruginosa bacteremic isolates.
Serospecific Vaccine against Klebsiella spp.
Antibody to Klebsiella capsular polysaccharide
(CPS) is highly protective against experimental
infections [76]. Several experimental vaccines
composed of purified Klebsiella CPS were found
to be safe and immunogenic in human volun-
teers [70]. Based upon the seroepidemiology of
Klebsiella blood isolates, a 24-valent CPS vac-
cine has been developed which is safe and im-
munogenic in humans. It also elicits production
of antibody to 11 “cross-reactive” CPS serotypes
not included in the vaccine and therefore, of-
fers potential “coverage” against ca. 75 % of all
Klebsiella bacteremic isolates.
Serospecific Vaccine against Escherichia
coli. Antibodies to both O (LPS) and K (cap-
sular) antigens can provide protection against
experimental E. coli infections [77], [78]. How-
ever, ca. 40 % of E. coli bacteremic isolates
do not have a polysaccharide capsule [79]. In
addition, the two most frequently encountered
K serotypes among blood isolates, K 1 and
K 5, are nonimmunogenic in humans. Approxi-
mately 90 % of E. coli bacteremic isolates can be
typed according to their O-antigen and most of
them can be grouped within 11 serotypes, mak-
ing a polyvalent formulation feasible [79]. The
toxicity of native LPS precludes its use as a vac-
cine. However, nontoxic, serologically reactive
O-polysaccharide can be isolated from E. coli
LPS and O-polysaccharide – protein conjugates
are being constructed in a manner similar to that
described for P. aeruginosa (see above).
Serospecific Staphylococcus aureus Vac-
cine. Staphylococcus aureus produces numer-
ous somatic antigens and extracellular toxins
which have been implicated as virulence fac-
tors. These include teichoic acid, lipoteichoic
acid, exopolysaccharide, capsular polysaccha-
ride, α-toxin, and β-toxin. Patients with deep-
seated S. aureus disease usually mount an anti-
body response to teichoic acid, α-toxin, and β-
toxin [80]. However, the relative protective ca-
pacities of antibodies against these antigens are
unknown.
A serotyping scheme has been described
for S. aureus based upon cell-surface capsular
polysaccharides [81]. Approximately 80 – 90 %
of S. aureus clinical isolates belong to one of
these serotypes [82]. Serospecific CPS is pro-

duced during experimental S. aureus infections
and can be isolated from blood [83]. The protec-
tive capacity of antibody to CPS is being evalu-
ated.
Crossreactive Anticore Glycolipid Vac-
cine. Antibody produced against the “core” gly-
colipid of E. coli strain J 5 crossreacts with the
LPS of virtually all gram-negative bacteria [84].
Polyclonal or monoclonal antiJ 5 antibody (see
also Section 6.1) can afford significant protec-
tion against several gram-negative pathogens
[85]. Protective antibody is directed against the
lipid A moiety of LPS and neutralizes the toxic
activities of LPS. When used prophylactically
or therapeutically, human plasma enriched with
antiJ 5 antibody provides significant protection
against death due to “gram-negative shock” [86],
[87].
2.9. Meningococcal Meningitis Vaccine
Etiological Agent and Pathogenesis. The
causative agent of meningococcal disease is
Neisseria meningitidis, a gram-negative diplo-
coccus. Humans are the only known reservoir
for N. meningitidis. The meningococcus col-
onizes the nasopharynx; infrequently, it en-
ters the bloodstream and infects the meninges.
Meningococcal disease is most often seen in
children less than 18 months of age and in
settings such as military recruit camps where
individuals from different geographical areas
come into constant close contact [88]. The dis-
ease is endemic worldwide with an annual in-
cidence of about 3/100 000 population. Numer-
ous epidemics occurred in the 1970s and 1980s
with attack rates exceeding 500/100 000. Eleven
meningococcal capsular serotypes are known,
with the majority of disease (ca. 95 %) caused
by groups A, B, C, W 135, and Y [88], [89].
History of Immunization. The first attempts
at immunizing against meningococcal disease
used vaccines made of killed bacteria and were
unsuccessful [90]. Attention then centered on
using purified capsular antigens. In 1945 Kabat
demonstrated that up to 1 mg of purified antigen
could be administered to humans with no unto-
ward reactions. However, these studies were not
continued because the prophylactic treatment of
Immunotherapy and Vaccines 19
meningococcal meningitis with antibiotics ap-
peared to be highly effective. However, the ap-
pearance of resistant strains in the early 1960s
revived interest in vaccine development. In the
late 1960s, highly purified group A and group C
capsular polysaccharides with a high molecular
mass were shown to be safe and immunogenic
in humans [91]. Human immune sera contained
elevated levels of antibody capable of lysing
meningococci in the presence of complement.
Production and Properties of Meningo-
coccal Vaccines. Available vaccines contain
serogroup A, C, W 135, and Y capsular anti-
gens. The most widely used formulations are
the tetravalent A, C, W 135, and Y, and the bi-
valent A and C vaccines. Capsular antigens are
purified from fermentor-grown cultures by co-
precipitation with detergents followed by eth-
anol fractionation, extraction with organic sol-
vents, and ultracentrifugation. Preservation of
the high molecular mass of the capsular anti-
gens is of critical importance since this governs
the immune response. Each vaccine dose con-
tains 50 µg of capsular antigen per serotype in
lyophilized form. Potency of the vaccine is based
upon various physicochemical properties (size
and purity). No animal test is required.
Immunization Recommendations.
Meningococcal vaccine can be administered
to any immunocompetent individual over 18
months of age. A single dose is given either
subcutaneously or intramuscularly. The group A
antigen is immunogenic in children ≥6 months
of age, but two doses must be administered
within a 2 – 3 month interval. Due to the low in-
cidence of the disease in the general population,
vaccine acceptance has been poor, it is usually
used to immunize military recruits or travellers.
Adverse Reactions. The meningococcal
vaccines are very safe–about 250×106 doses
of vaccine have been administered worldwide
with no vaccine-associated fatalities reported.
Mild, transient local reactions occur in about
25 % of vaccinees.
Vaccine Efficacy. Conclusive proof that
group C meningococcal vaccine prevents dis-
ease was presented in 1970 [92]. Subsequently,
the group A vaccine was found to be effective
20 Immunotherapy and Vaccines
in controlling endemic and epidemic disease in
adults and children [93], [94]. Group C vaccine
elicited a protective antibody response only in
children ≥18 months of age, whereas the two
doses of group A vaccine were immunogenic
in children ≥6 months of age [95]. Efficacy for
the W 135 and Y capsular antigens has not been
clinically demonstrated, but is assumed because
they engender appropriate levels of relevant
functional antibodies in humans.
Future Prospects. Current meningococcal
vaccines have two major limitations. First, they
do not provide protection against group B or-
ganisms which may account for up to 50 % of
endemic cases. Secondly, the group C vaccine
is nonimmunogenic in children ≤18 months of
age. The major difficulty in producing a group
B meningococcal vaccine is that the group B
capsular antigen, a homopolymer of α (2→8)-
linked scialic acid, is poorly immunogenic in
humans, probably because similar structures are
found in human gangliosides and fetal pro-
teins. Even when coupled to carrier proteins
[96] or mixed with group B outer membrane
proteins [97], the group B antigen is a poor
immunogen. An alternative approach is to use
serotype-specific outer membrane proteins, 18
such serotypes have been identified. Outer mem-
brane protein vaccines containing group B or C
capsular antigens were found to be safe and im-
munogenic in humans [98], [99]. A large-scale
field trial to determine the efficacy of these vac-
cines has been conducted.
In an attempt to construct vaccines capable
of engendering a protective immune response
in young children, capsular antigens have been
coupled to carrier proteins to form conjugate
vaccines [96]. Group A and C antigens coupled
to tetanus toxoid are more immunogenic in an-
imals than native capsular antigens. Studies are
in progress to determine the safety and immuno-
genicity of group A and C conjugates in humans.
2.10. Tuberculosis Vaccine
Etiological Agents and Pathogenesis. Hu-
man pulmonary tuberculosis is caused by
the bacilli, Mycobacterium tuberculosis and
M. bovis, the latter causing disease primarily in
children. M. tuberculosis is spread by infected
droplets, whereas M. bovis is disseminated by
ingestion of contaminated milk. The vast ma-
jority of clinical disease occurs in middle-aged
individuals in developed countries and in chil-
dren under 10 years of age in underdevel-
oped countries. The incidence of tuberculosis
ranges from ca. 10/100 000 population in Eu-
rope and North America to 500/100 000 in parts
of Africa. Although 95 – 98 % of people ex-
posed to M. tuberculosis become infected, i.e.,
become tuberculin-positive, they do not mani-
fest signs of clinical disease and are immune
[100]. Only a small percentage of infections
progress to a disease state. Once the inhaled
bacilli reach the lower respiratory tract, they are
ingested by alveolar macrophages. Rapid intra-
cellular growth with inflammatory cell recruit-
ment leads to granuloma formation. In the ma-
jority of cases, this is as far as the disease pro-
gresses. Active disease presents clinical symp-
toms ranging from self-limiting pulmonary in-
volvement to acute disseminated disease which
rapidly leads to death.
History of Immunization. Efforts to protect
humans against tuberculosis have almost ex-
clusively employed Bacillus Calmette – Guérin
(BCG) vaccine (see below). Oral BCG vaccine
was first evaluated in the 1920s in newborns.
Subsequent studies showed BCG vaccine to be
safe when given orally or parenterally. In the
early 1930s, large-scale trials with BCG vaccine
were started. Subsequent trials involving Amer-
ican Indians and Eskimos showed that intrader-
mally administered BCG conferred protection
against disease in infants. Additional proof of
vaccine efficacy was obtained in a small-scale
trial with Canadian Indians [101]. Even in the
face of a high disease transmittance rate and a
mortality rate of 800/100 000, the vaccine was
more than 80 % effective over a 9-year observa-
tion period.
Production and Properties of Tuberculo-
sis Vaccine. The original BCG vaccine strain
was developed at the Pasteur Institute over 50
years ago by in vitro cultivation of M. bovis for
10 years. Numerous substrains are now used to
produce commercial products. The bacteria are
grown on Sauton’s liquid media either in bottles
with a high surface area or in submerged cul-
ture. They are then collected, suspended in a sta-

bilizer solution, homogenized, and lyophilized.
The bulk material is standardized by opacity,
dry mass, and number of viable bacteria per unit
mass. The vaccine is usually filled into glass am-
pules which are sealed under vacuum.
Controls of BCG vaccine are stringent. First,
the culture is examined for purity, viability, and
identity. Innocuity is tested by injection of vac-
cine into guinea pigs. The number of live or-
ganisms are determined by viability counting.
General safety tests to confirm absence of unex-
pected toxicity are also performed. Resistance of
the bacteria to elevated temperature is also docu-
mented. Vaccine potency is based upon the total
viable bacteria per dose. However, this is not
totally satisfactory because (1) different BCG
strains may vary in their ability to evoke a protec-
tive cell-mediated immune response, and (2) the
ratio of viable to nonviable bacteria can influ-
ence the magnitude of both the humoral and
cell-mediated immune response to vaccination
[102].
Immunization Recommendations. A sin-
gle intradermal dose containing (4 – 8)×105 vi-
able organisms in 0.1 mL is normally adminis-
tered. The preferred use of BCG vaccine de-
pends on the disease incidence in the area in
question. In developed areas where disease is
infrequent, vaccination should be limited to in-
dividuals at high risk to exposure, i.e., family
members of tuberculosis patients, health care
workers expected to come into contact with tu-
berculosis patients or infected clinical speci-
mens, or travellers entering an area of high en-
demicity for a prolonged period of time. In de-
veloping areas with a high rate of disease inci-
dence, routine immunization of infants is a cost-
effective method of controlling tuberculosis. In
addition, all tuberculin-negative children upon
entry into school and tuberculin-negative adults
should be vaccinated.
Adverse Reactions. Correctly administered
BCG vaccine is considered to be safe; one vac-
cination in 100 000 will result in a noticeable
reaction [103]. A significant proportion of these
reactions are ulcers caused by inadvertent subcu-
taneous administration of the vaccine. Of greater
concern are the reports of mycobacterial disease
following immunization [104]. Great care must
be taken to insure that a potential vaccinee is
Immunotherapy and Vaccines 21
not immunocompromised as the vaccine organ-
isms can then rapidly multiply and lead to death
[105].
Vaccine Efficacy. In numerous field trials
the efficacy of BCG vaccine ranged from 0 –
80 % [100]. Although three trials failed to doc-
ument significant vaccine-induced protection,
considerable support still exists for large-scale
vaccination in developing countries [106].
Future Prospects. At present, no “second-
generation” tuberculosis vaccines are being clin-
ically evaluated. Although the tuberculin skin
test offers a practical method to measure “im-
munity” to disease, the molecular mechanism re-
sponsible for inducing a tuberculin-positive state
is not known.
Several M. tuberculosis cell surface anti-
gens have been identified which can induce
a cell-mediated response in experimental ani-
mals [107]. The cloning of such antigens and
their high-level expression in bacteria or yeast
would permit economical large-scale purifica-
tion. Alternatively, such cloned genes could be
inserted into a suitable vector such as vaccinia
virus [108]. Since most tuberculosis vaccine are
used in underdeveloped nations, any new vac-
cine must not only be safe and effective, but also
economical to produce and control.
2.11. Escherichia coli Vaccines
Etiological Agent and Pathogenesis. Es-
cherichia coli, a gram-negative motile rod, is a
major cause of diarrheal disease, urinary tract
infections, neonatal meningitis, and nosocomial
infections (see Section 2.8). Escherichia coli
produces a variety of virulence factors including
a polysaccharide capsule antigen (K-antigen),
lipopolysaccharide antigen (O-antigen), fim-
briae, enterotoxins, and cytotoxins. The major-
ity of E. coli strains causing a given disease fall
within a limited number of K or O serotypes.
Escherichia coli is the major cause of diar-
rhea worldwide. Three “categories” are recog-
nized, i.e., enterotoxigenic, enteroinvasive, and
enteropathogenic E. coli strains.
1) Enterotoxigenic E. coli invariably causes wa-
tery diarrhea which may be accompanied
22 Immunotherapy and Vaccines
by cramps, fever, or vomiting. The dis-
ease is acquired by ingesting contaminated
food and affects infants, young children, and
travellers to underdeveloped countries. The
strains produce either a heat-stable entero-
toxin and/or a heat-labile enterotoxin; they
also possess fimbriae which mediate attach-
ment to the ileal epithelium. Three distinct
fimbriae (CFA/I, CFA/II, and E 8775) have
been identified on enterotoxigenic E. coli
strains isolated from human disease.
2) Enteroinvasive E. coli causes a dysentery-
like disease characterized by fever, diarrhea,
cramps, and stools containing blood and mu-
cus. The disease is acquired by ingestion of
contaminated food. The bacteria readily pen-
etrate the ileal and colonic epithelia where
they rapidly multiply resulting in tissue de-
struction. They can then infect adjacent cells
or penetrate the underlying lamina propria.
3) Enteropathogenic E. coli affects primarily
infants and young children after ingestion of
contaminated food causing watery diarrhea
often accompanied by fever and/or vomit-
ing. These E. coli strains rarely produce a
heat-labile or heat-stable enterotoxin and are
unable to invade eukaryotic cells. They col-
onize the ileal epithelium with destruction
of the surrounding brush border in the ab-
sence of invasion. Enteropathogenic E. coli
strains can synthesize a cytotoxin similar to
that produced by Shigella dysenteriae 1. This
so-called Shiga-like toxin is thought to elicit
the clinical symptoms.
Approximately 50 % of neonatal meningi-
tis is due to E. coli transmitted from mother
to infant. The bacteria apparently penetrate the
intestinal epithelium of the neonate, enter the
bloodstream, and are then translocated to the
meninges. The majority of strains (>80 %) caus-
ing meningitis express the K 1 capsular antigen.
However, the O-antigen also appears to play a
role in virulence. Escherichia coli meningitis
has a high attendant fatality rate.
Escherichia coli is the causative agent of ca.
90 % of nonobstructive urinary tract infections
which are much more common among women
than men. The clinical syndromes can include
bacteriuria, cystitis, and pyelonephritis. Renal
scarring associated with pyelonephritis occurs
most often in children ≤5 years of age. The
infecting organisms are thought to be acquired
from the flora colonizing the vagina and peri-
urethral areas. P and type I fimbriae are believed
to play an important role by mediating tissue
colonization [109].
History of Immunization. Initial attempts
to vaccinate against E. coli diarrhea used crude
cellular extracts derived from O types 111, 55,
and 86 [110]. In one study, hospitalized infants
up to one year of age received multiple doses
of oral vaccine. Overall efficacy was 41 %. Ef-
forts to develop vaccines against E. coli diar-
rhea center around the use of purified antigens
or live-oral attenuated strains to elicit an anti-
toxic or antiadhesion response. Calves or neona-
tal piglets reared by mothers immunized with
purified K 88, K 99, or 987 P fimbriae, proc-
holeragnoid (heat-aggregated cholera toxin with
reduced toxicity) or heat-labile enterotoxin are
protected against E. coli bacillosis [111–113].
Graded doses (45 – 1800 µg) of purified type I
fimbriae have been parenterally administered to
human volunteers [114]. The vaccine was well
tolerated and elicited both a humoral IgG and
an intestinal secretory IgA antibody response.
A moderate level of serospecific protein was ob-
tained in volunteer challenge studies.
Oral administration of 1 mg of purified CFA/I
evoked a significant secretory IgA response in
about 50 % of vaccinated humans [115]. Re-
sponders were protected against challenge with
an St+ , LT+ , CFA/I+ E. coli strain.
Future Prospects. No vaccines are available
to prevent diarrhea urinary tract infections, or
meningitis caused by E. coli. At present, both
attenuated live-oral and inactivated vaccines are
being evaluated as to their ability to prevent
E. coli diarrhea. The production of a killed, inac-
tivated, oral, whole-cell vaccine from a CFA/I+
strain has been described [115]. The majority of
vaccinees responded with a good secretory IgA
antibody response to CFA/I. The diarrhea attack
rate for volunteers challenged with a toxinogenic
CFA/I+ E. coli strain was 89 % for those in the
placebo group and 20 % for those in the vaccine
group.
A spontaneous mutant of E. coli has been iso-
lated which has lost the genes for production
of the heat-labile and heat-stable enterotoxins
but still expresses CFA/II. This strain has been

proposed as a likely carrier for cloned genes ex-
pressing relevant protective antigens such as col-
onization factors or “enterotoxoids” [116]. Sev-
eral purified antigens show promise as oral vac-
cine candidates including conjugates of the two
enterotoxins [117], the B subunit of the heat-
labile toxin [118], and toxoid termed “procol-
igenoid” produced by heating the heat-labile en-
terotoxin [119].
Since the majority of E. coli strains associ-
ated with neonatal meningitis possess the K 1
capsule antigen, this would appear to be an ex-
cellent vaccine candidate. However, the K 1 anti-
gen (which is identical to the group B meningo-
coccal capsule) is not immunogenic in humans
(see Section 2.9).
The most promising vaccines to prevent seri-
ous E. coli urinary tract infections (pyelonephri-
tis) contain type I fimbrial antigens. Two fimbrial
amino acid sequences have been shown to pre-
vent colonization by a homologous E. coli strain
in a murine pyelonephritis model [120]. Anti-
body to one sequence also afforded protection
against an E. coli strain expressing a distinct fim-
brial serotype.
2.12. Neisseria gonorrhoeae Vaccine
Etiological Agent and Pathogenesis.
Neisseria gonorrhoeae was identified as the
causative agent of gonorrhea by Bumm in 1885.
It is a gram-negative diplococcus with hu-
mans serving as its only known natural host.
Gonorrhea is a sexually transmitted disease of
epidemic proportions worldwide. More than
500 000 cases are reported annually in the
United States. The infection is usually restricted
to the entry site, i.e., the genitalia, pharynx,
and/or rectum. More than 90 % of infected males
display symptoms characterized by pain upon
urination and an urethral exudate [121]. How-
ever, up to 50 % of infected women remain
asymptomatic and can serve as carriers. In ap-
proximately 10 % of women with genital gon-
orrhea, the bacteria invade the bloodstream and
result in pelvic inflammatory disease. Dissemi-
nated gonococcal infections are far less frequent
in male patients.
When cultured on solid medium, four colo-
nial variants (T 1 – T 4) can be discerned [122] ;
however, only the piliated T 1 and T 2 variants
Immunotherapy and Vaccines 23
can cause disease [123]. The first step in the
infectious process is pili-mediated attachment
of the gonococcus to mucus-secreting epithelial
cells. Shortly thereafter, the bacteria are internal-
ized and enter the submucosa [121]. Treatment
of gonorrhea has been complicated by the oc-
currence of multiple antibiotic-resistant clinical
isolates.
History of Immunization. The first attempt
to prevent gonorrhea by immunization used
a killed whole-cell vaccine administered par-
enterally [124], but did not provide protection
in Eskimos. Subsequent human vaccine trials
have used purified intact pili: antipili antibody
blocks attachment of the gonococcus to epithe-
lial cells, and, an antipili secretory IgA anti-
body response is mounted during natural in-
fection [125]. Vaccination of human volunteers
with purified N. gonorrhoeae pili elicited a hu-
moral and a local vaginal antibody response.
These antibodies blocked the attachment of
N. gonorrhoeae to epithelial cells in vitro [125].
Two 2 mg doses of a monovalent pili vaccine
resulted in significant protection against chal-
lenge with the homologous strain [126]. How-
ever, no protection was observed when the chal-
lenge strain possessed a pilus serotype modestly
cross-reactive with the vaccine [121]. A subse-
quent field trial with this monovalent vaccine
failed to show any protection [125].
Future Prospects. No vaccine has been li-
censed for the prevention of gonorrhea. The
above human volunteer studies showed that pro-
tection against gonorrhea mediated by antipili
antibody is serospecific [125]. Two different ap-
proaches are being taken to circumvent the prob-
lem of antigenic variation among pili:
1) Use of a polyvalent vaccine containing those
types of pili which predominate in a given
area [126].
2) Use of conserved antigenic domains from
the antigenic pili molecule. A cyanogen bro-
mide cleavage fragment can induce produc-
tion of an antibody that can recognize var-
ious pili types [127]. Peptides correspond-
ing to this conserved region block attach-
ment of the gonococcus to epithelial cells
in vitro [128]. If such peptides can induce
a broadly protective immune response in
24 Immunotherapy and Vaccines
humans, synthetic peptides containing the
proper sequence could be coupled with car-
rier proteins to produce a conjugate vaccine.
A possible alternative to pili as vaccine
antigens are the outer membrane proteins of
N. gonorrhoeae, the most promising being the
PI protein. The PI protein has a sufficiently re-
stricted antigenic variation to make construction
of a polyvalent vaccine feasible. Furthermore,
antiPI antibody is bactericidal and may block en-
try of the gonococcus into epithelial cells [129].
A parenterally administered, outer membrane
preparation enriched with PI protein elicits an
antibody response in humans [130].
2.13. Hemophilus influenzae Type b
Vaccines
Etiological Agent and Pathogenesis.
Hemophilus influenzae is a gram-negative bacil-
lus first isolated by Pfeiffer in 1892. The vast
majority (>90 %) of human disease is caused by
type b capsular serotype organisms in children
between the ages of three months to two years
[131]. Hemophilus influenzae type b can cause
several serious disease syndromes predominated
by meningitis which is endemic worldwide and
is one of the three leading causes of bacte-
rial meningitis in developed countries. Infec-
tion is by inhalation of contaminated aerosols.
After localized multiplication, the organisms
enter the bloodstream and translocate to the
meninges. Even though effective antibiotic treat-
ment against H. influenzae type b has been avail-
able for many years, meningitis due to this or-
ganism still has a high attendant morbidity and
mortality rate.
History of Immunization. Intensive efforts
to understand human immunity to H. influenzae
type b disease have largely replaced clinical
trials with crude, poorly characterized vac-
cines. The capsular polysaccharide is a criti-
cal virulence factor because it confers resistance
to lysis by serum [132]. Antibody-dependent,
complement-mediated bacteriolysis is the basis
for immunity against H. influenzae type b [132].
Antibodies directed against the type b capsular
polysaccharide are bactericidal in the presence
of complement and are believed to be the pre-
dominant protective antibody [131].
Production and Properties of H. influenzae
Type b Vaccines. Two vaccines are licensed
for use against H. influenzae type b: a purified
capsular polysaccharide and a conjugate vac-
cine composed of capsular polysaccharide co-
valently coupled to diphtheria toxoid. The cap-
sular antigen is purified by coprecipitation of the
capsule with detergent, digestion with nucleases
and protease, extraction in phenol, and ultracen-
trifugation. In the conjugate vaccine, the cap-
sule is covalently linked to diphtheria toxoid by
a bifunctional spacer molecule. The potency of
H. influenzae type b vaccines is not determined
in animals but by analysis of physicochemical
characteristics which confirm that the capsular
antigen is in an immunogenic form.
Immunization Recommendations. The
capsular polysaccharide vaccine was first li-
censed for use in the United States in 1985. A
single 25 µg dose, administerd parenterally is
recommended for children 2 – 6 years of age; no
booster dose is needed.
For the conjugate vaccine licensed in 1987,
a single dose is recommended for children 18
months to 6 years of age; no booster dose is
needed.
Vaccine Efficacy. The purified capsular
polysaccharide vaccine was evaluated in a trial
on Finnish children [133]. Vaccine efficacy was
approximately 90 % for children who were vac-
cinated at 2 – 5 years of age or older. Vacci-
nation provided no protection in children 3 –
18 months of age due to poor immunogenicity
[132]. However, retrospective studies performed
in the United States have found vaccine effi-
cacy to range from 55 – 88 % [133]. In contrast,
a capsular polysaccharide – diphtheria toxoid
conjugate vaccine evoked protective levels of
antibody in a majority of 18-month old children
[135].
Future Prospects. A H. influenzae type b
vaccine is clearly needed which is immunogenic
in children <18 months of age, where approxi-
mately 75 % of serious disease occurs. Several
candidate vaccines including the diphtheria tox-
oid conjugate mentioned above and a Neisseria
meningitides outer membrane protein – capsular

polysaccharide conjugate are being evaluated in
clinical trials [135], [136]. A large-scale Finnish
trial is being performed to evaluate the diphthe-
ria toxoid conjugate vaccine. Vaccine efficacy
was 83 % in children vaccinated at 3, 4, 6, and
14 months of age [137]. In contrast, this vaccine
afforded insignificant protection when used to
immunize Alaskan Eskimo children at 2, 3, and
4 months of age.
3. Viral Vaccines
3.1. Measles Vaccine
Etiological Agent and Pathogenesis.
Measles (rubeola) is caused by a paramyxovirus
first isolated by Enders and Peebles in 1954.
This highly contagious disease is transmitted
in an aerosol of virus-infected droplets from
the respiratory tract. During the prevaccination
era, children in industrialized countries gener-
ally had a self-limiting disease characterized
by conjunctivitis, bronchitis, fever, and a rash;
complications included pneumonia (mostly bac-
terial) and encephalitis. A late complication of
measles in early childhood is the development of
a progressive, fatal neurological disease, termed
subacute sclerosing panencephalitis. Death due
to complications following measles infection
remains one of the leading causes of child mor-
tality in developing countries.
History of Immunization. Two approaches
to vaccine development were adopted, one was
based on killed (inactivated) strains and the other
on live, attenuated strains [138]. The formalde-
hyde-inactivated, alum-precipitated vaccine was
licensed in the United States in 1963 and with-
drawn in 1967 because its protective effect was
insufficient. In addition, the vaccinees exposed
to the wild virus developed atypical measles with
fever and rash but also pneumonitis.
The attenuated live vaccine with the Edmon-
ston B strain was also licensed in 1963 in the
United States. A large-scale field trial in the
United Kingdom showed high seroconversion
rates, a long antibody response, and a 84 – 94 %
decrease in attack rate among the 36 000 im-
munized children. Further attenuation of the
vaccine strain in chick embryo cells yielded
the Schwartz and the Moraten strains. Both
Immunotherapy and Vaccines 25
strains are included in current measles (and com-
bined measles – mumps – rubella) vaccines. At-
tenuation of the original Edmonston strain in
human diploid cells yielded the Edmonston-
Zagreb strain, also in current use [139]. With the
introduction of general immunization against
measles, the disease has almost disappeared in
many countries (see below).
Production and Properties of Measles Vac-
cines. The attenuated measles vaccine strain is
usually grown in primary culture using chick
embryo cells or human diploid cells. The WHO
test requirements for the seed virus and cell sub-
strates are described in[140, pp. 52 – 73],[141,
p. 179 ]. Tests for monitoring measles vaccine
grown in chick embryo cells include tests for
nonadsorbing viruses and avian leukosis viruses.
For vaccines grown in human diploid cells, chro-
mosomal monitoring requirements have been
formulated. Potency is checked by titration of
the virus in tissue culture, (minimum require-
ment, 1000 TCID50 per human dose). To test
for stability, a sample of the final freeze-dried
vaccine is incubated at 37 ◦ C for seven days: the
sample must then contain at least 1000 TCID50
in each human dose. The vaccine is lyophilized;
reconstituted vaccine should be used immedi-
ately or stored at 0 – 10 ◦ C for not more than
8 h.
Immunization Recommendations. The
vaccine is given at the age of 15 – 24 months
[142] or even younger in developing coun-
tries [139]. In the United States and many
European countries, a combined measles –
mumps – rubella (MMR) is usually given (see
Section 3.4). As for all live vaccines, immu-
nization of immuncompromised individuals is
not generally recommended. Immunization of
pregnant women should be avoided although
no increased risks have been documented for
measles vaccine.
Adverse Reactions. Current attenuated
measles strains give a few mild reactions, mostly
consisting of fever and rash 7 – 10 days after
immunization. The neurologic reaction of the
Guillian-Barré syndrome has been reported but
is extremely rare.
26 Immunotherapy and Vaccines
Vaccine Efficacy. The rapid decrease of
measles morbidity in the United States after in-
troduction of general immunization is a good
example of vaccine efficacy [142], [143]. Over
a period of 20 years the rate of notified cases fell
by more than 99 %. Subacute sclerosing panen-
cephalitis also decreased. The overall protective
efficacy of the current strains is 90 % but is lower
in children of 12 months of age due to the pres-
ence of maternal antibody. Life-long immunity
after one dose of vaccine has not been proven.
Increased disease incidence in the United States
has raised the question of the possible need for
a two-dose schedule [142].
Future Prospects. Given the efficacy and
acceptability of the currently used attenuated
measles strains, there has been little incentive
for vaccine development. However, current vac-
cines are not very stable at high temperature,
which is a concern in many developing coun-
tries. Furthermore, a large proportion of mor-
bidity and mortality in children in developing
countries occurs before the recommended age
for immunization. Research is centered on solv-
ing these two problems. More immunogenic
strains and/or higher doses of attenuated vac-
cines are being investigated [139]. Another ap-
proach would be to renew work on the devel-
opment of a killed (inactivated) vaccine. Inacti-
vated vaccines have so far failed, this is probably
the result of a lack of neutralizing antibodies to
the viral fusion surface protein that is destroyed
in the inactivation process [144]. Subunit vac-
cines based on the hemagglutinin and the fusion
proteins of measles virus have successfully been
tested in animals and may provide a vaccine es-
pecially suited for developing countries [145].
3.2. Mumps Vaccine
Etiological Agent and Pathogenesis.
Mumps is caused by a member of the paramyx-
ovirus group and is also known as (epidemic)
parotitis due to its predominant symptom–
infection of the salivary (parotid) gland. Trans-
mission occurs via infected aerosol droplets.
The disease is generally characterized by mod-
erate fever and swelling of the salivary glands.
Subclinical infection occurs in ca. one third of
infected individuals. Most cases occur in chil-
dren 5 – 10 years of age. The most frequent com-
plication is meningoencephalitis, giving clinical
symptoms in >10 % of patients. Orchitis, a less
common but feared complication in postpuber-
tal males, results in impairment of fertility in
10 – 20 % of cases but absolute sterility is rare.
History of Immunization. Isolation of the
causative agent for mumps by Habel in 1945
from chick embryo was followed by attenuation
studies for vaccine development. Development
of an inactivated (killed) vaccine, was also ini-
tiated; a killed vaccine was licensed in the US
in 1950 – 1978 but had low long-term protective
efficacy [146]. An inactivated mumps vaccine
was also produced in Finland for immunizing
military recruits; it decreased disease incidence
by 94 % [147]. Immunization with an attenu-
ated mumps vaccine (Jeryl Lynn strain cultured
in chick embryonic cells), licensed in 1967 in
the United States resulted in a 97 % decline of
disease incidence in the general population by
1981 [146], [148], [149]. Chick embryonic cells
are used to propagate another highly attenuated
strain, Urabe Am 9 developed in Japan [150].
The Rubini strain, developed in Switzerland, is
used in a vaccine in which the virus is grown in
human diploid cells [151]. Vaccines based on the
Leningrad-3 strain have been used in the former
Soviet Union and other countries since 1974.
Production and Characterization of
Mumps Vaccines. The WHO requirements for
the above-mentioned live mumps vaccine have
been formulated [152, pp. 139 – 164]. Controls,
thermostability tests, and requirements are as
for measles vaccines, i.e., the minimum potency
should be retained after inoculation for one
week at 37 ◦ C (see Section 3.1). No minimum
requirement of potency has been established
by the WHO but a commonly used minimum
dosage is 5000 TCID50 per human dose. The
vaccine is lyophilized, reconstituted vaccine
should be used without delay.
Immunization Recommendations.
Mumps vaccine is recommended to be given
to all susceptible individuals over the age of
12 months [153]. It is usually given in com-
bination with measles and rubella vaccines in
industrialized countries (see Section 3.4).

Vaccine Efficacy. Clinical efficacy is 75 –
90 % for the live vaccine containing the Jeryl
Lynn strain [149]. Similar results have been
shown for the other strains or inferred from sero-
logic comparisons. The long-term protective ef-
ficacy of the inactivated mumps vaccine used in
Finland has not been established but antibody re-
sponses indicate that it may give a shorter-term
immunity than a live vaccine due to a lack of an-
tibody response to the fusion protein [154] as in
the case of the inactivated measles vaccine (see
Section 3.1).
Adverse Reactions. Mumps vaccine is one
of the least reactogenic attenuated vaccines.
Side effects are rare and mild. No cases of
atypical mumps were found after immunization
of Finnish recruits with killed mumps vaccine
[154].
Future Prospects. The present vaccine is
highly satisfactory and little effort has been de-
voted to further development. Subunit vaccines
based on the hemagglutinin neuraminidase and
the fusion surface proteins of mumps virus are
protective in animal models [155].
3.3. Rubella Vaccine
Etiological Agent and Pathogenesis.
Rubella, also known as German measles, is
caused by a member of the togavirus group. The
virus was isolated in 1962 by two independent
American groups. In 1941 Gregg reported that
this mild disease could cause severe congenital
cataracts in children born to mothers who were
infected during the first trimester of pregnancy.
Transmission of the virus occurs by the respira-
tory route. The clinical picture is a discrete rash
and low-grade fever, 25 – 50 % of cases remain
subclinical. Arthritis is a common complication;
encephalitis is less common than in measles or
varicellae (1/6000 cases). The congenital rubella
syndrome is usually characterized by hearing
impairment, ocular lesions, cardiac malforma-
tion, microcephaly, and mental retardation. The
severity of defects is related to fetal age at the
time of maternal infection, with the most se-
vere damage seen after infection during the first
month of pregnancy.
Immunotherapy and Vaccines 27
History of Immunization. Live attenuated
rubella vaccines were licensed in several coun-
tries in 1969 – 1970. One of the first vaccines
contained the Cendehill strain, grown in rabbit
kidney cells [156]. Other vaccines were based on
the HPV-77 strain, grown in either duck embryo
cultures or in dog kidney cultures. The latter vac-
cine produced arthritis symptoms; 50 – 60 % of
adult female vaccinees as compared to 10 – 20 %
for the other two vaccines. A vaccine based on
the RA 27/3 strain, isolated and propagated in
human diploid cells, was licensed in the mid-
1970s in Europe and in 1979 in the USA [157].
This vaccine gave a better immune response and
had less side effects than the other vaccines; it
has replaced other rubella strains in the current
vaccines.
Production and Properties of Rubella Vac-
cines. The attenuated RA 27/3 strain is grown in
human diploid cells. The WHO requirements for
rubella vaccine include control of normal karol-
ogy of the human diploid cells[158, pp. 54 – 84],
[159, pp. 313 – 316]. The minimal potency re-
quirement, determined by titration in tissue cul-
ture, is 1000 TCID50 per human dose. The vac-
cine is lyophilized; reconstituted vaccine should
be used immediately or stored at 2 – 8 ◦ C for not
more than 8 h.
Immunization Recommendations. Vacci-
nation strategies range from protection of the
individual as in the U.K. to indirect protection
as in the US [160], [161]. Protection is achieved
by general immunization of adolescent girls,
usually combined with post partum vaccination
of seronegative women. Indirect protection of
adult women by immunization of young children
is often combined with vaccination of seroneg-
ative women.
A combination of these strategies is used in
some European countries, e.g., Sweden [162].
This program includes routine screening of
all pregnant women, post partum vaccination
of seronegatives, and two-dose immunization
of children with combined measles – mumps –
rubella vaccine. Although the risks for the fetus
seem small, the vaccine should not be given to
pregnant women and contraceptive measures are
recommended for three months after immuniza-
tion.
28 Immunotherapy and Vaccines
Adverse Reactions. The RA 27/3 strain has
few and mild side effects. The joint manifesta-
tions, mainly arthralgia, are more common in
adults. Intrauterine infections during pregnancy
have been documented for all three strains (Cen-
dehill, HPV-77, and RA 27/3) [163]. No abnor-
malities related to congenital rubella infections
have been documented for any of the strains.
Vaccine Efficacy. The efficacy of the current
strain is ca. 90 %. The incidence of congeni-
tal rubella virus infection decreased in countries
with a routine immunization program against the
disease. The duration of immunity has not yet
been determined. The risks to the fetus upon a
maternal reinfection are also unknown and may
be very small.
Future Prospects. The current vaccine is
considered safe and efficient and no efforts are
devoted to further development.
3.4. Combined Measles – Mumps –
Rubella Vaccine
History of Immunization. The first com-
bined measles – mumps – rubella (MMR) vac-
cine was licensed in the US in 1971. It con-
tains the Moraten measles strain (Section 3.1),
the Jeryl Lynn mumps strain (Section 3.2), both
grown in chick embryo cultures, and the RA
27/3 rubella strain (Section 3.3), grown in hu-
man diploid cells. Three other combined vac-
cines were licensed later. The triple vaccine has
replaced the monovalent vaccines used in gen-
eral immunization programs for children in the
United States and in many European countries.
Production and Properties of Combined
Vaccines. The minimum potencies for the at-
tenuated strains of measles, mumps, and rubella
virus, are the same as in the monovalent vac-
cines, i.e., 1000 TCID50 for measles, >5000
TCID50 for mumps, and >1000 TCID50 for
rubella. The vaccine is lyophilized and should
be stored at 2 – 8 ◦ C.
Immunization Recommendations. In
many countries, one dose of vaccine is ad-
ministered to children at the age of 15 – 24
months. Two doses are given in some Euro-
pean countries [162]. In Sweden, a first dose is
administered at 18 months and a second at 12
years of age, replacing the monovalent programs
for measles and rubella immunization respec-
tively. In the United States, the combination of
MMR with oral polio vaccine and diphteria –
tetanus – pertussis (DTP) at 15 months has been
recommended [164], [165].
Adverse Reactions. The side effects are the
same as for the monovalent vaccines, the
measles component being the most reactogenic.
The rate of adverse reactions is <0.5 – 4 % [166].
Vaccine Efficacy and Future Prospects.
The efficacy of the combined preparation is the
same as for the monovalent vaccines with an
overall efficacy of 90 – 95 % [143]. The life-long
protective efficacy of a single injection has not
been proven. Further combinations with an at-
tenuated varicella component were investigated
[167].
3.5. Polio Vaccine
Etiological Agent and Pathogenesis. Polio
(infantile paralysis) is caused by a picornavirus
of the genus enteroviruses. A poliovirus strain
was first isolated in cell culture by Enders,
Weller, and Robbins in 1949. In 1951, polio
virus isolates were officially grouped into three
serotypes, type 1 (Brunhilde), type 2 (Lansing),
and type 3 (Leon). Transmission is mainly via
the oral – fecal route and the virus reaches the
central nervous system by way of the blood
stream. The vast majority of infections are sub-
clinical. The nonparalytic disease has a mild or
minor form with fever and general malaise and
a more severe or major form with additional
symptoms of meningitis/meningoencephalitis.
Paralytic polio, with its most severe bulbar form,
is estimated to represent 5 – 10 % of clinical
cases. The mortality rate is 5 – 10 % of clini-
cal cases, i.e., 1 – 2 % of all infections. Severe
paralysis is seen in 10 – 20 % of clinical cases,
and mild or moderate paralysis in 30 %.
With improved sanitation and standards of
living, a shift towards infection at a higher
age was observed in industrialized countries in
the prevaccination era. During the first half of

this century, peak incidence was noted in the
age group 5 – 14 years. A large proportion of
cases occurred in young adults. In the develop-
ing countries, polio has maintained its character
of infantile paralysis with the majority of chil-
dren being infected during the first few years of
life. Cases of paralytic polio occur only in the
youngest age groups.
History of Immunization. The first large-
scale field trial of an inactivated (killed) polio
vaccine was launched in the US in the early
1950s, only a few years after successful prop-
agation of the virus in tissue culture. Inactivated
polio vaccine (IPV) was licensed for general use
in the United States in 1955 [168], [169]. In
1955, cases of atypical paralytic polio were re-
ported, most of them were associated with two
lots of vaccine from Cutter [170]. Although
live poliovirus was recovered from vaccine sup-
plied by other manufacturers, this failure of in-
activation is known as the Cutter incident. The
total toll was 269 cases, of which 192 were par-
alytic (with ten deaths). Clinical trials with a
live attenuated oral polio vaccine (OPV) were
started in 1958 and the vaccine was licensed in
the United States in 1962 [171], [172]. Immu-
nization with OPV alone has been used from the
late 1950s – early 1960s in Europe. One of the
largest immunization campaigns was launched
in the former Soviet Union in 1960 when OPV
was given to ca. 77×106 people. A few Euro-
pean countries have used only IPV [173–175].
In Sweden, clinical trials with IPV were started
in 1955 and general immunization was intro-
duced in 1957 [173]. All three strategies were
effective in reducing the incidence of paralytic
polio. With IPV alone, an 80 % reduction rate
was achieved in the US with less than 50 % of
the population immunized [168], [169], [171].
The decrease of polio continued at the same rate
after introduction of OPV. Introduction of the
massive IPV campaign in Sweden resulted in a
93 % decrease of paralytic disease after six years
of immunization and elimination of the disease
by 1964 [173].
Production and Properties of Polio Vac-
cines. Oral polio vaccines are manufactured
with the attenuated Sabin strains. Inactivated po-
lio vaccines are mostly produced with the type 1
Mahoney strain, type 2 MEF-1 strain, and type 3
Immunotherapy and Vaccines 29
Saukett strain. Primary and secondary monkey
(Cynomolgus) kidney cultures or human diploid
cells are most commonly used for culture. The
WHO requirements for both OPV and IPV pro-
duced in primary monkey kidney cells include
tests
1) for the absence of cytopathogenic viruses
with the exception of some foamy
viruses[141, pp. 40 – 84] , [159, pp. 157 –
173], [176],[182, pp. 108 – 110]
2) the absence of simian B virus in rabbits, and
3) the absence of SV 40 in sensitive cells.
(usually primary green monkey Cercopite-
cus kidney cultures). Minimal potency require-
ments for OPV were formulated by the WHO in
1987[152, pp. 165 – 166]. A single human dose
of trivalent oral vaccine should contain approx-
imately 105.5 – 106.5 infectious units of type 1,
104.5 – 105.5 of type 2, and 105.0 – 106.0 of type
3. For IPV, recommendations of 40 : 8 : 32 D-
antigen units per human dose for types 1, 2, and
3 respectively were issued in 1981 [141]. The
potency tests for OPV (and IPV prior to inactiva-
tion) are performed by titration in tissue culture.
In vivo potency tests are also required for IPV
but neither the animal species nor the number of
injections is specified. Inactivation of polio vac-
cine by formaldehyde is controlled by titration
of polio-sensitive cells (usually Cynomolgus or
Cercopitecus) in tissue culture. Oral polio vac-
cines are supplemented with a stabilizer, usually
magnesium chloride or sorbitol. Both OPV and
IPV contain antibiotics, usually neomycin; OPV
is best stored at −20 ◦ C but can be stored at 2 –
4 ◦ C for a variable length of time; IPV is stored
at 2 – 8 ◦ C.
Immunization Recommendations. Rec-
ommendations vary but at least three doses of
OPV are given. In the US, five doses are rec-
ommended, in Sweden four, and in Finland six.
The WHO recommends four doses for infants
in developing countries.
Adverse Reactions. Serious adverse reac-
tions of paralytic polio have only been reported
for OPV with a predominance for type 3 [169],
[177] (incidence = 1/1×106 vaccine recipients,
including both nonimmune and immune indi-
viduals). No cases of IPV-induced paralytic po-
lio have been reported after 30 years of use (with
30 Immunotherapy and Vaccines
the exception of the Cutter incident). IPV is con-
sidered to be the least reactogenic of all the vac-
cines used for childhood immunization.
Vaccine Efficacy. Both OPV and IPV have
proven highly efficient in eliminating polio from
industrialized countries with general immuniza-
tion programs. The protective efficacy is close to
100 % after at least three doses of vaccine with
adequate immunogenicity. Serologic data re-
ported for OPV from developing countries have
been less than encouraging. Both vaccines have
eliminated the circulation of wild type strains in
industrialized countries with high immunization
rates. Outbreaks among groups refusing immu-
nization have occurred [168]. In 1984, nine cases
of paralytic polio caused by type 3 occurred
in Finland [177]. The low immunogenicity of
the IPV used in Finland has been known since
the late 1960s [178]. The new IPV developed
in the Netherlands with its high antigen content
is given in European countries in at least three
doses [179]. Two doses or even one dose of the
new IPV preparations have been claimed to be
highly protective. A two dose-schedule has been
used in Africa and protective efficiency was ca.
89 % [180]. Inactivated polio vaccine is more
heat-stable and can be combined with diphte-
ria – tetanus – pertussis vaccines.
Future Prospects. Further development of
OPV is aimed at reducing production costs by
use of microcarrier cultures (cell cultures on
beads etc.) in fermentors [179]. Use of contin-
uous cell lines is being investigated. An IPV
produced in a continuous cell line is currently
licensed in France [181]. The WHO require-
ments for production of vaccines in continuous
cell lines are formulated in [182, pp. 93 – 107].
3.6. Hepatitis B Vaccine
Etiological Agent and Pathogenesis. Hep-
atitis B is caused by the hepatitis B virus (HBV),
a member of the hepadna virus group. The pres-
ence of a new antigen called the Australia (Au)
antigen or hepatitis B surface antigen (HBsAg)
in the blood of patients with leukemia, Down’s
syndrome, and hepatitis was shown by Blum-
berg in 1964 – 67. Other antigens (HBeAg and
HBcAg) were described later and correlated
with infectivity; they are found in the 42-nm
Dane particle, i.e., the infectious virion.
The HBV is a pathogen only for humans, but
certain monkeys, in particular chimpanzees, are
also susceptible to infection. Transmission in hu-
mans occurs mainly by inoculation with infected
blood. The virus is excreted in body fluids, caus-
ing infection through saliva and sexual contact.
A chronic carrier state is established in 5 – 10 %
of adult cases; chronic active hepatitis develop in
25 – 30 % of the carriers, often leading to cirrho-
sis of the liver: 90 % of infants infected at birth
become carriers. Hepatitis B occurs throughout
the world; the incidence varies from <0.5 % in
Western Europe and the US to 5 – 15 % in South-
east Asia and Southern Africa [183], [184].
Hepatitis B virus plays an important role
in hepatocellular carcinoma although the exact
mechanism has not yet been determined [183].
History of Immunization. Heat-inactivated
HBsAg-positive serum was used by Krugman
in 1971 to immunize children. A 70 % protec-
tion rate was found upon challenge with active
virus. The high protective efficacy of hepatitis B
vaccine derived from purified, inactivated HB-
sAg from positive plasma was demonstrated in
1980 [185]. The vaccine, developed in the US,
was licensed for general use in 1981. Studies
in 1980 – 1983 also showed that the HBV vac-
cine in combination with human antiHBV im-
mune globulin (HBIG) prevented development
of the carrier state in infants born to carrier moth-
ers. Hepatitis B vaccine was the first vaccine
to be routinely produced by DNA recombinant
technology in yeast (→ Genetic Engineering,
Chap. 5.2.). Serologic studies indicate that the
protective efficacy of the recombinant vaccines
is the same as that of the plasma-derived vaccine.
Genetically engineered vaccines were licensed
in the US in 1986 and subsequently in several
European countries.
Production and Properties of Hepatitis B
Vaccines. The WHO requirements for plasma-
derived HBV specify guidelines for selection of
donors of HBsAg-positive plasma[186, pp. 70 –
101]. The plasma pool must be subjected to ex-
tensive tests in animals, fertile eggs and cell cul-
tures for extraneous viruses and Mycobacterium
tuberculosis. The purified HBsAg (>95 % pure)
is inactivated by treatment with pepsin followed

by urea (8 mol/L), and formaldehyde or by heat
with or without formaldehyde. After controls for
purity and antigen content, an alum adjuvant is
added; the vaccine must be stored at 5 ± 3 ◦ C.
No minimal requirement of antigen content has
been formulated but a commonly used vaccine
contains 20 µg HBsAg per human dose.
The WHO requirements for HBV made by
recombinant DNA technology state that such
vaccines may contain the S gene products or
the S/pre-S combination [152, pp. 106 – 138]; li-
censed recombinant vaccines contain only the S
gene product. A full description of the host cell
(Saccharomyces cerevisiae) and the expression
vector is required.
The HBsAg is commonly purified by precipi-
tation, ultrafiltration, and chromatography. Tests
for HBsAg and residual cell or plasmid DNA
are required before addition of adjuvant. No
minimal antigen requirements have been formu-
lated but a potency assay in mice has been out-
lined. Commonly used recombinant yeast vac-
cines contain 10 or 20 µg HBsAg per human
dose.
Immunization Recommendations. Most
Western European countries and the US have
issued recommendations for vaccination of risk
groups [187–189]. In general, pre-exposure pro-
phylaxis is recommended for medical and other
staff in frequent contact with high-risk groups
or with blood from such groups. Immunization
of patient groups frequently receiving blood or
blood products, as well as patients in certain
institutions is also recommended.
For pre-exposure prophylaxis, three 20 µg
doses of plasma-derived vaccine is given at day
0, 30 and 6 months intramuscularly in the del-
toid region. Children <11 years of age should
receive 5 µg at each injection. Postexposure pro-
phylaxis for infants to HBsAg, HBeAg-positive
mothers is 0.5 mL intramuscular injection of hu-
man antiHBV immune globulin (HBIG, see also
Section 5.3). Vaccine (10 µg) should be given at
birth and then at one and six months of age. For
postexposure prophylaxis of adults, 1 mL HBIG
should be given immediately together with three
doses of vaccine administrated as for preexpo-
sure prophylaxis.
Adverse Reactions. Side effects are mild
and mainly local at the site of injection.
Immunotherapy and Vaccines 31
Vaccine Efficacy. The overall vaccine effi-
cacy is 90 % with no response upon immuniza-
tion in 5 – 10 % of the vaccines. In individuals
with seroconversion to HBsAg, protective effi-
cacy is almost 100 %. Studies on the duration
of protection beyond 5 – 10 years are not yet
available. The protective efficacy of HBIG and
vaccine administered to newborns born to car-
rier mothers has been estimated to be more than
90 % [183].
Future Prospects. Current HBV vaccines
are safe and have a high efficacy but also high
production costs. Research is aimed at the de-
velopment of cheaper, more immunogenic vac-
cines. The inclusion of the pre-S (P 31) re-
gion of the viral genome in addition to the
present S (P 25) region is being investigated
in yeast-derived recombinant vaccines [190],
[191]. Other antigens such as HBcAg are also
being studied. Polypeptide vaccines, based on
the major determinants of HBsAg, are undergo-
ing clinical trials. Synthetic peptides may pro-
vide the ultimate solution to the problem of im-
munogenic, readily available HBV vaccines.
3.7. Rabies Vaccine
Etiological Agent and Pathogenesis. Ra-
bies is a lethal disease caused by a neu-
tropic rhabdovirus affecting humans and warm-
blooded animals. The most common route of
transmission to humans is by infected saliva
through the bite from a rabid animal. The virus
ascends along peripheral nerves to the central
nervous system. The incubation period varies
from ten days to several months. Symptoms in-
clude a nonspecific prodromal stage followed by
an acute neurological phase ending in coma and
death. The disease is endemic in animals in most
parts of the world with some exceptions such as
the United Kingdom and parts of Scandinavia
[192]. In the urban form, stray dogs and cats act
as vectors, whereas the sylvatic form involves
wild animals. Urban rabies is well-controlled
in most countries, but progressive spread of the
sylvatic form by foxes is a major problem in
Europe. In the US, other animal species are in-
volved including bats. In Europe and Africa,
bats are the vector for an antigenic variant of the
virus [193].
32 Immunotherapy and Vaccines
History of Immunization. In 1885 Pasteur
administered the first ever vaccination to a young
boy who had been bitten by a rabid dog [194].
Serial injections containing a virus strain that
had been propagated and attenuated in the spinal
cord of rabbits (“fixed” virus) were given in an
increasingly virulent form. The treatment was
rapidly adopted as standard postexposure pro-
phylaxis. The attenuated vaccine was later re-
placed by an inactivated (killed) vaccine also
produced in neural tissue [194]. However, in
some patients the myelin content of the vac-
cine caused sensitization resulting in neurolog-
ical disease. An inactivated vaccine produced
in duck embryos decreased the risk of neural
tissue sensitization. The most widely used ra-
bies vaccine in Europe and in the United States
was developed in the 1960s and is produced in
human diploid cells. A highly purified, concen-
trated duck embryo vaccine is also available. In
Europe, other vaccines are produced in primary
animal cells and continuous cell lines; non hu-
man primate diploid cells are used in the US.
Production and Properties of Rabies Vac-
cines. Attenuated strains of rabies virus are
grown in tissue culture or embryonated duck
eggs (see above). The virus suspension is usu-
ally concentrated and in some cases purified.
Only inactivated vaccines are allowed for human
use. Inactivation is mostly achieved by treat-
ment with β-propiolactone, but phenol, formal-
dehyde, or ultraviolet irradiation are also used.
The WHO has established requirements for con-
trols of the cells used for virus propagation
[152, pp. 167 – 194 ],[159, pp. 54 – 95]. The po-
tency of the vaccine is determined in a mouse
challenge assay (minimal requirement 2.5 IU
per human dose). The vaccine is then usually
lyophilized although adjuvanted preparations
are also in use.
Immunization Recommendations. The
recommendations vary by vaccine and coun-
try. As pre-exposure prophylaxis, three 1 mL
doses are usually recommended on days 0, 7,
and 28 or on days 0, 28, and 90 (or 365) as in-
tramuscular injections in the deltoid or supras-
capular region [192], [195]. As postexposure
prophylaxis, recommended measures include
local wound care, administration of vaccine,
and administration of human rabies immune
globulin (HRIG, 20 IU per kilogram of body
weight; see also Section 5.3). Purified equine
rabies immune globulin (40 IU/kg) is used in
many developing countries with few side ef-
fects [196]. Postexposure vaccine prophylaxis
is given in five 1 mL doses on days 0, 3, 7, 14,
and 28 by intramuscular injection in previously
unimmunized individuals. In individuals who
have received pre-exposure prophylaxis, two
doses are given on days 0 and 3. Children under
four years of age receive 0.5 mL injections.
Vaccine Efficacy. The protective efficacy of
the human diploid cell vaccine given (with
HRIG) as postexposure prophylaxis was first
documented in field trials conducted in Iran in
1974 – 75; the survival rate was 100 %. Other
studies have confirmed the protective effect of
pre- and postexposure immunization with this
vaccine. The general opinion that postexposure
vaccine prophylaxis in combination with im-
mune globulin conveys 100 % protection has
been challenged by two cases of vaccine failure
[197]. In other vaccines, a high protective effi-
cacy has been proven in field trials or inferred
from comparative serologic studies.
Adverse Reactions. The most common side
effects are local redness and induration at the in-
jection site and fever. Desquiting episodes of a
serum sickness-like reaction upon repeated im-
munization have been reported [198] and are
possibly due to a sensitizing complex formation
between β-propiolactone and human serum al-
bumin [199] or to the high content of bovine
serum residues in the vaccine [200].
Future Prospects. Work is centered on the
development of large-scale vaccine production
techniques to decrease the high costs and thereby
increase availability in poorer countries. One
such approach is the use of microcarrier cul-
ture systems in fermentors [179]. A vaccine pro-
duced in this manner in a continuous cell line has
been evaluated. A subunit vaccine based on the
surface glycoprotein of the virus is a further pos-
sibility. The peptide segment of the glycoprotein
that induces the production of neutralizing an-
tibodies has been identified and synthesized. A
synthetic peptide vaccine may therefore be the
most attractive future alternative [201].

3.8. Influenza Vaccine
Etiological Agent and Pathogenesis. In-
fluenza is caused by two antigenically dis-
tinct members of the orthomyxovirus group–
influenza viruses A and B. In 1934 Andrewes
managed to transfer the influenza A virus from
human material to ferrets and later to mice.
In 1940, Francis and Magill independently
isolated influenza B by transmission to ferrets.
The virus was subsequently propagated in em-
bryonated hen’s eggs and tissue culture. The
disease is transmitted by droplet infection from
the respiratory tract; it is characterized by high
fever, muscle pain, and a dry cough. Pneumo-
nia is the main cause of mortality in elderly
people and persons with chronic underlying dis-
eases. Influenza occurs worldwide with regular
epidemics during the winter months. The re-
current epidemics are caused by small changes
(antigenic drift) in the main pathogenic determi-
nants of the virus, i.e., hemagglutinin and neu-
raminidase. Large pandemics occur at 10 – 20
year intervals and are due to substantial changes
(antigenic shift) in the pathogenic determinants.
Antigenic drift is seen in both influenza A and
B whereas antigenic shift has only been noted
in influenza A [202].
History of Immunization. The first in-
fluenza vaccines produced in hens’ eggs were
tested in humans in the early 1940s. These inac-
tivated whole virus vaccines had a low content
of viral antigens and a high content of contam-
inating egg protein. Consequently, the protec-
tive effect was low and the rate of adverse re-
actions high. Whole virus vaccines purified by
ultracentrifugation showed clearly improved im-
munogenicity and a decreased rate of side reac-
tions. Zonal centrifugation further improved the
vaccine for adults but still caused adverse reac-
tions in children. Disruption of influenza vaccine
with detergent was used to develop split vaccines
in the mid 1960s. Subunit vaccine introduced
in the mid 1970s contain concentrated, purified
hemagglutinin and neuraminidase [202].
The development of live, attenuated virus
strains was pursued in the former Soviet Union
by serial passages in hens’ eggs [202]. However,
doubt was cast on their stability. More stable at-
tenuated vaccines were obtained by isolation of
temperature-sensitive and cold-adopted mutants
Immunotherapy and Vaccines 33
[203]. These strains were successfully tested in
humans in the late 1970s but are not in gen-
eral use. Attenuated vaccines, based on avian –
human recombinant strains, are being evaluated
in clinical trials.
Production and Properties of Influenza
Vaccines. The most commonly used influenza
vaccines are of the whole or split virus type.
The strains used are specified in annual WHO
recommendations. The virus is grown in embry-
onated hens’ eggs (usually pathogen-free) and
the allantoic fluid is harvested. According to
the WHO requirements, the inactivation method
used (usually addition of formaldehyde or β-
propiolactone) should inactivate avian leuko-
sis viruses and mycoplasma[204, pp. 148 – 170].
The virus is concentrated and purified by high-
speed centrifugation either before or after inac-
tivation. Effective inactivation is controlled by
inoculation of embryonated hens’ eggs. Hemag-
glutinin content is usually checked by single
radial immunodiffusion against a WHO stan-
dard. Whole virus and split virus vaccines usu-
ally contain 10 – 15 µg hemagglutinin per hu-
man dose. The inactivated vaccines are usually
supplemented with a preservative and stored at
2 – 8 ◦ C.
The WHO requirements for attenuated (live)
vaccines stipulate that pathogen-free eggs must
be used; the absence of other pathogens must be
controlled in tissue cultures and animals[204,
pp. 171 – 194]. Vaccine potency is determined
by titration in embryonated eggs. No minimal
requirements of infective dose have been for-
mulated.
Immunization Recommendations. Most
countries have recommendations for yearly im-
munization of high-risk groups. Some countries
and the WHO recommend vaccination of all in-
dividuals over 65 years. In the United States, im-
munization of children with chronic pulmonary
or cardiac disorders or with other chronic dis-
eases residing in institutional care is recom-
mended; 0.25 and 0.5 mL of split virus vaccine
are given in the age groups 6 – 35 months and
3 – 12 years, respectively [205]. For individu-
als older than 12 years, 0.5 mL of vaccine is
recommended.
34 Immunotherapy and Vaccines
Adverse Reactions. The incidence of sys-
temic (febrile) and local reactions is <10 % and
<20 % for the whole virus vaccines and split
vaccines, respectively. An increased rate of the
neurological reaction of Guillain – Barre’s syn-
drome was reported in the United States af-
ter large-scale immunization against swine in-
fluenza in 1976.
Protective Efficacy. The protective efficacy
of influenza vaccines is controversial. Discrep-
ancies are probably due to different vaccines,
number of injections given, the age groups un-
der study, and occurrence of antigenic drift dur-
ing the study period. The newer purified whole
virus and split virus vaccines are considered to
be 70 – 90 % effective in healthy adults. The du-
ration of protection is largely dependent on the
degree of antigenic drift. In the case of antigenic
shift, little or no protection can be expected. The
immune response and protective efficacy in chil-
dren receiving chemotherapy and in debilitated
elderly persons are lower [206]. The protective
efficacy in preventing death has been estimated
at 74 % [207].
Future Prospects. Research on influenza
vaccines is directed toward the synthesis of the
hemagglutinin and the neuraminidase antigens
and to improvement of their immunogenicity
[201]. Another line of development is the atten-
uated live vaccine approach [203].
3.9. Varicella Vaccine
Etiological Agent and Pathogenesis. The
varicella-zoster virus, a member of the herpes
virus group, causes two distinct clinical manifes-
tations [208]–varicella and herpes zoster. Vari-
cella, also known as chickenpox, is the primary
infection. Herpes zoster is caused by the reacti-
vation of the latent varicella virus in ganglion tis-
sue. Varicella is usually a mild infection charac-
terized by a vesicular rash. Transmission occurs
by droplet infection from the respiratory tract
and by direct contact with vesicle fluid. Compli-
cations are encephalitis (1/1000 – 1/3000 cases)
and pneumonia. The disease occurs worldwide,
with 90 – 95 % of cases occurring before the age
of 15 years. Suspected cases of congenital vari-
cella have been described mainly after varicella
during the first trimester of pregnancy. Neonatal
varicella with a 30 % mortality rate occurs with
maternal varicella within one week prior to term
without prophylaxis. Varicella in the compro-
mised host is a severe disease with a 10 – 20 %
mortality rate. Herpes zoster, a localized, often
painful vesicular rash is most common in the
elderly; it occurs in 10 % of the population.
History of Immunization. The varicella-
zoster virus was first isolated and propagated in
tissue culture by Weller and Stoddard
in 1952. Attenuated, live varicella vaccine was
developed in Japan in 1970 using the Oka strain
[209]. After serial passages, the attenuated strain
was adapted to human diploid cell cultures. The
immunogenicity and protective efficacy of the
vaccine in healthy children has been demon-
strated [209], [210]. Efficacy of the vaccine in
children with malignant disease, in particular
leukemia, has also been shown [211]. The Oka
strain is used for vaccine production in Japan,
Europe, and the United States.
Production and Properties of Varicella
Vaccines. The attenuated, live Oka strain is
propagated in human diploid cell cultures. The
WHO requirements include control for the ab-
sence of adventitious agents and the usual con-
ditions for culture in human diploid cells [186,
pp. 102 –133]. No minimal potency require-
ments have yet been formulated. Varying doses
have been used in clinical trials. The most com-
mon formulation is ≥2000 plaque-forming units
per human dose. The vaccine is lyophilized; re-
constituted vaccine should be used without de-
lay.
Immunization Recommendations. Vari-
cella vaccine is not yet recommended for general
immunization in Europe or the United States.
It is given to children with leukemia during re-
mission or when chemotherapy is withheld for
one week prior to and after vaccination [211].
A second dose is often given to children who
remain seronegative after the first injection.
Adverse Reactions. In healthy children and
adults, local swelling and pain occurred in 1 –
5 % of healthy children and 20 % of adults. Sys-
temic reactions with fever and a rash are reported
in 5 – 10 % of both healthy children and adults.

Fever and rash occurred in 40 % of vaccine re-
cipients with chemotherapy suspended for two
weeks.
Vaccine Efficacy. Protective efficacy in
healthy children is 95 – 100 % with persistence
of immunity over a 5 – 10 year period. In adults,
protective efficacy is 60 – 80 % and possibly of
shorter duration. In children with malignancies,
protective efficacy is 60 – 92 %. Spread from
vaccine-induced vesicular rash has been docu-
mented in household contacts.
Future Prospects. The currently investi-
gated, attenuated Oka strain vaccine is effective
but long-term immunity and zoster incidence re-
main to be established. The problems of latency
and of vaccine-induced rash could be overcome
by the development of a subunit vaccine pro-
duced either by a recombinant DNA technique
or by using synthetic peptides.
3.10. Yellow Fever Vaccine
Etiological Agent and Pathogenesis. Yel-
low fever is a hemorrhagic fever caused by a
member of the toga virus group [212], [213].
The virus was transmitted to rhesus monkeys
by Mathis and coworkers in 1927. The virus
strain was then propagated by serial passages
of intracerebral inoculations in white mice. Yel-
low fever is transmitted by mosquitoes of the
genus Haemagogus in South America and of
the genus Aedes in Africa. The animal reser-
voir is mainly monkeys. Subclinical infections
are common. The incubation period in clini-
cal cases is 3 – 6 days; symptoms range from
transient fever and headache to high fever with
meningoencephalitis, followed by jaundice and
hemorrhagic manifestations. In the malignant
form all these symptoms are present and death
occurs within one week. Mortality rates are
40 – 50 % in the severe forms of the disease.
History of Immunization. Two types of at-
tenuated vaccine were developed in the early
1930s cultured in neural tissue (Dakar vaccine)
and in chick embryo [214]. Initially, both types
were given together with human immune serum.
In the late 1930s, the vaccine cultured in neutral
Immunotherapy and Vaccines 35
tissue was used for mass immunization in Sene-
gal and the chick embryo vaccine was tested in
Brazil. The vaccines were administered by scar-
ification using normal human serum as stabi-
lizer. The chick embryo cultured vaccine (first
the 17 E vaccine and later the 17 D vaccine)
has been most widely used. Numerous prob-
lems and accidents were associated with both
vaccines. The human serum used as stabilizer
caused hepatitis affecting many vaccinees in the
armed forces during World War II. Systemic re-
actions were common. Severe postvaccination
encephalitis with a high mortality rate occurred
mainly with the neural tissue cultured vaccine.
This reaction (mostly in children < one year
of age) was also reported with some lots of the
17 D vaccine with increased neurotropism; fur-
thermore, loss of protective efficacy was noted
in tropical climates due to low thermostability.
The 17 D vaccine is the result of developments
aimed at careful definition of the properties of
the seed virus and at improved thermostability
[215].
Production and Properties of Yellow Fever
Vaccines. Only certain institutes are approved
by the WHO for production of yellow fever
vaccine[159, pp. 34 – 53]. The seed lot virus,
usually a substrain of 17 D-204, has to be
shown to be free from neurotropism by test-
ing in monkeys[159, pp. 34 – 53],[182, pp. 113 –
141],[216] . Most producers use seed lots that are
free of leukosis virus for production but their
use is not mandatory. Virus-infected embryos
are harvested, homogenized, and the supernatant
is used as vaccine. Several tests for adventitious
agents are performed. Virus titrations are per-
formed in a mouse assay with a minimal potency
requirement of 1000 LD50 per human dose. The
vaccines are lyophilized in the presence of stabi-
lizer. Most current vaccines retain the minimal
requirement for two weeks at 22 ◦ C. A vaccine
stable for two weeks at 37 ◦ C is requested by
WHO for use in tropical areas.
Immunization Recommendations. Vacci-
nation of visitors to endemic areas in equatorial
Africa and northern parts of South America is
recommended. Immunization is mandatory in
several countries for visitors from endemic ar-
eas. One 0.5 mL injection is given to both adults
and children for both primary and booster immu-
36 Immunotherapy and Vaccines
nization. Booster injections are recommended
every ten years. The vaccine is not recommended
for children under one year of age or for pregnant
women.
Adverse Reactions. Yellow fever vaccines
are safe and induce only minor local reactions.
Transient headache can occur.
Protective Efficacy. Current vaccines are es-
timated to be 90 – 95 % protective. Immuniza-
tion has decreased or eliminated the disease in
many endemic areas.
3.11. Tick-Borne Encephalitis Vaccine
Etiological Agent and Pathogenesis. The
tick-borne encephalitis virus, a member of the
toga virus group, was first isolated in the former
Soviet Union in 1937 [212]. Two antigenically
distinct forms of the virus cause the disease in
Europe and in the eastern former Soviet Union.
The main vector for the European form is Ixodes
ricinus and for the Eastern form I. perulcatus.
Many wild and domestic animals can be in-
fected; the main animal reservoirs are small
mammals such as field mice. Transmission is
usually by tick bite but infection can occur by
drinking untreated cow milk. The incubation
period is 7 – 14 days before onset of fever and
malaise. After a 1 – 2 week recovery period, a
second stage with fever and neurological symp-
toms can follow: meningitis (40 % of cases),
meningoencephalitis (40 %), and severe menin-
goencephalomyelitis (20 %). Mortality rates are
1 – 2 % in the European form and 20 – 25 % in
the Eastern form; neurological sequelae are seen
in 15 – 40 % of cases. The disease is subclinical
or abortive with only the first stage in 75 % of
infections. Tick-borne encephalitis is endemic
in Central Europe, in the Balkan countries, and
in Finland and Sweden.
History of Immunization. The first inacti-
vated (killed) vaccine was developed and used
in humans in the former Soviet Union in 1939
followed by an attenuated, live vaccine in the
1960s [217]. An inactivated vaccine, developed
in Europe with a virus strain isolated from a tick
in Austria, was subjected to clinical trials in 1973
[218]. The vaccine used in Western Europe is a
purified, concentrated version of this.
Production and Properties of Tick-Borne
Encephalitis Vaccine. The inactivated vaccine
is produced by propagation of the virus in
hens’ eggs, followed by purification by continu-
ous flow zonal ultracentrifugation, and inactiva-
tion with formaldehyde. No WHO requirements
have been formulated. The vaccine contains not
less than 25 protective doses per human dose
assayed in a mouse protection test. Human al-
bumin is used as stabilizer and aluminum hy-
droxide as adjuvant.
Immunization Recommendations. In Aus-
tria and Bavaria, immunization of children older
than one year is recommended. Other endemic
countries recommend immunization of forest
workers and other high-risk groups. Primary im-
munization consists of two 0.5 mL intramuscu-
lar injections at 1 – 3 month interval, followed by
a third 0.5 mL dose 9 – 12 months after the sec-
ond. Booster injections are recommended every
three years.
Adverse Reactions. Local and systemic re-
actions are rare and mild. Low-grade fever is
occasionally seen, mainly in children after the
first injection.
Protective Efficacy. The vaccine is at least
95 % protective against all European virus
strains in children and young adults. Seroconver-
sion rates of about 90 % are reported for persons
over 65 years of age.
3.12. Japanese Encephalitis Vaccine
Etiological Agent and Pathogenesis. The
Japanese encephalitis virus, belonging to the
toga virus group, was first isolated in 1935 in
Japan [212]. The disease is transmitted by the
mosquito Culex tritaeniorhynchus; the main ani-
mal reservoirs are pigs and birds. The incubation
period is 5–15 days. The disease is subclinical
in at least 95 % of infections. In clinical cases
symptoms vary from mild febrile disease with
headache to severe encephalitis. Paralytic forms
more commonly affect the upper extremities.
In endemic areas the disease affects mainly

younger children but also elderly people with
mortality rates of 50 %. Neurological sequelae
have been reported in 30 – 40 % of survivors of
the severe clinical forms.
History of Immunization. A formaldehyde-
killed vaccine, consisting of a 5 % suspension
of infected mouse brain tissue, was introduced
for human use in Japan in 1954 [219]. The vac-
cine (Nakayama strain) has been purified by pro-
tamine sulfate precipitation since the late 1950s
and by absorption with charcoal or kaolin since
the early 1960s. A protective efficacy of 80 –
90 % for the vaccine was shown. The highly pu-
rified, mouse brain cultured vaccine produced
in Japan is also used for immunizing travellers
to endemic areas. In China, a vaccine produced
in primary hamster kidney cells has been exten-
sively used since the late 1950s.
Production and Properties of Japanese
Encephalitis Vaccines. Several Japanese
manufacturers produce the purified vaccine from
culture of mouse neural tissue by similar meth-
ods. No WHO requirements have been formu-
lated. A vaccine used for immunizing travellers
to endemic areas is prepared by infecting mouse
brain with the Nakayama strain. The brain ho-
mogenate is purified by protamine sulfate treat-
ment and then inactivated with formaldehyde.
Further purification involves ultracentrifugation
on a sucrose density gradient. The vaccine is
lyophilized; the reconstituted vaccine must be
used immediately.
Immunization Recommendations. Pri-
mary immunization consists of subcutaneous
injection of two 1 mL doses at a 1 – 2 week
interval. A third 1 mL dose is recommended
one month later as is a regular booster injection
every 1 – 3 years. Extensive immunization in en-
demic areas has been considered by the WHO.
Children less than three years of age should re-
ceive 0.5 mL doses. Immunization is generally
recommended for health care workers and other
people with extended stay in endemic areas.
Adverse Reactions. Only a few mild, local
and systemic reactions have been reported.
Protective Efficacy. Immunization has dras-
tically reduced disease incidence in Japan. The
Immunotherapy and Vaccines 37
protective efficacy is estimated to be 90 – 95 %
from serologic studies.
3.13. Smallpox Vaccine
Smallpox (variola) was caused by a member of
the pox virus group, which also includes the
vaccinia virus used for immunization. The dis-
ease was one of the most devastating infections
in human history. Eradication of smallpox is
the success story of immunization. In 1967 the
WHO launched a massive eradication program–
smallpox was still reported from 42 countries. In
May 1980, WHO officially declared the world
free from smallpox. No proven cases of variola
have occurred in the past decade.
Two types of vaccine were manufactured, calf
lymph vaccine and egg vaccine [220]. Both liq-
uid and lyophilized forms were used. Require-
ments for manufacturing, control, and potency
were formulated by the WHO. Immunization by
multiple puncture with a bifurcated needle was
most commonly used. Severe adverse reactions
included postvaccination encephalitis and dis-
seminated vaccinia.
General immunization against smallpox was
withdrawn in most European countries in the
mid 1970s. Requirement for smallpox vaccina-
tion was abandoned for international travel in
1982.
Recommendations for civilian immunization
in the United States include only laboratory
workers handling variola virus or other closely
related orthopox viruses. Military personnel in
the United States and the former Soviet Union
are routinely vaccinated against smallpox [221].
3.14. Rift Valley Fever Vaccine
Rift Valley fever is an arthropod-borne disease
known only in Africa [212], [213]. It is caused
by a member of the Bunya virus group. The ma-
jor vectors are mosquitoes (Culex theileri and
Aedes cabbalus). The main natural hosts are
sheep, cattle, and goats. The incubation period
is 2 – 6 days. Rift Valley fever is a febrile dis-
ease with headache and abdominal pain lasting
less than one week. Hemorrhagic fever with liver
38 Immunotherapy and Vaccines
necrosis and encephalitis are the severe mani-
festations causing mortality and sequelae. Out-
breaks occurred in Africa during the 1970s, with
the largest outbreak in Egypt in 1977 – 1978.
An inactivated (killed) vaccine (NDBR-103)
was produced by the US army in 1967 [222]. The
Entebbe strain of the virus was grown in primary
monkey kidney cells, inactivated with formalde-
hyde, and lyophilized. The adverse reactions are
few and mild. One case of Guillain – Barré oc-
curred in a Swedish military vaccinee [223]. Se-
roconversion rates after subcutaneous injection
of three 1 mL doses given at 1 – 2 week inter-
vals were over 95 %. A newer vaccine (GSD-
200) is based on a cloned version of the orig-
inal seed virus (Entebbe strain) and grown in
diploid rhesus monkey cells. The WHO has for-
mulated requirements for inactivated Rift Val-
ley fever vaccines produced in primary monkey
kidney cells and in human or non-human pri-
mate diploid cells[141, pp. 104 – 143]. No mini-
mal potency requirements have been formulated.
4. Vaccines against Parasites
4.1. Vaccines against Helminths
Helminths represent one of the major causes
of infectious diseases affecting humans and do-
mestic animals. This results not only in a dele-
terious effect to the health of the hosts, but
also in great economic losses. Although ma-
jor advances have been made in the chemother-
apy and epidemiology of diseases caused by
helminths (→Anthelmintics) immunotherapy
has produced only minor breakthroughs in the
field of veterinary parasitology.
As a result of a long evolutionary develop-
ment and a close parasite – host relationship,
helminths have evolved strategies for circum-
venting complete elimination by the host im-
mune response. Nevertheless, the immune re-
sponse usually exerts deleterious effect upon
their growth and proliferation. Thus, the use of
a vaccine resulting in partial or complete pro-
tection might be one of the most cost-effective
means of controlling helminth diseases.
Only a few reliable vaccines are available
on a commercial scale for immunoprophylaxis
of helminthoses in livestock. A vaccine against
lungworms in cattle is a commercial success.
Another vaccine against hookworms in dogs,
although immunologically efficient, has failed
commercially.
Important in the development of vaccines is
the identification, isolation, and testing of puta-
tive protective antigens. Successful vaccination
against helminth infections requires the prim-
ing of those responses which may subsequently
be triggered during natural infection. Since anti-
gen presentation plays a central role in the ac-
quired immune response, the development of
accessory cells and the activation of T and B
lymphocytes has to be taken into account [224].
Developments have focussed on vaccines pro-
duced by recombinant DNA techniques. In vitro
cultivation methods have been used for produc-
ing helminth antigens for immunoprophylaxis
and combined with recombinant DNA technol-
ogy [225].
The helminthoses described in this chapter
were chosen on the basis of the parasites’ im-
portance to human health or because of their in-
teresting biology.
4.1.1. Vaccines against Schistosoma
Etiologic Agents, Pathogenesis, and Epi-
demiology. Human schistosomiasis occurs pri-
marily in tropical countries where it is one of
the most threatening diseases. The infection af-
fects about (200 – 300) ×106 people throughout
the world; more than 600×106 people live in
Schistosoma-endemic areas. The importance of
schistosomiasis has also increased in industrial-
ized countries with intensive tourism and influx
of high numbers of refugees from endemic areas
[226].
The main causative agents of schistosomiasis
are helminths of the genus Schistosoma which
use aquatic or amphibious snails as intermediate
hosts:
1) Schistosoma haematobium uses aquatic
snails of the genus Bulinus as intermediate
hosts and occurs mainly in Africa and some
middle-eastern countries.
2) S. mansoni uses aquatic snails of the genus
Biomphalaria as intermediate hosts; it oc-
curs in Africa, parts of Arabia, northern and
eastern parts of South America, and some
Carribean islands.
 Immunotherapy and Vaccines 39
Table 2. Main cells and antibodies active against Schistosoma in
3) S. japonicum uses amphibious snails of the experimental animal models or in vitro
genus Oncomelania as intermediate hosts; it
Develop- Cells and anti- Mechanisms
occurs in Japan, the Philippines, and parts of
mental stage bodies involved
China, Thailand, and Indonesia. of parasite

The life cycle of all three Schistosoma species Egg effector T lympho- initiation of granu-
cytes loma formation
is similar. The fully developed miracidium (lymphokines)
hatches from the egg in water and infects the Schistosomula neutrophils, IgG tegumental damage
intermediate host (i.e., the snail) where it multi- and killing of young
plies asexually to produce numerous cercariae. larvae under certain
conditions
The cercariae are released into the water and Schistosomula eosinophils, IgG killing of young
infect humans by penetration through the skin. larvae in the pres-
They develop into immature worms (schistoso- ence of comple-
mula) which migrate to the lungs and liver. The ment
Schistosomula eosinophils, IgE killing of young
adult, sexually mature worms mate and migrate larvae
to their final destination which varies accord- Schistosomula macrophages, IgE IgE-dependent
ing to the species: S. haematobium migrates to cytotoxicity
the veins of the vesical plexus, S. mansoni and
S. japonicum to the mesenteric veins. The adult
The immune response may result in resis-
paired worms produce eggs (300 – 3000 per pair
tance to reinfection but not to simultaneous ex-
per day) which pass through the vessels and tis-
pulsion of an established parasite population
sues into the lumen of the gut and bladder. The
from primary exposure [235]. In this way, the
eggs escape from the host in the feces and urine
parasites evade the immune response. An al-
and the cycle is repeated.
ternative escape mechanism is that the worms
Eggs from S. haematobium are mainly found
may be covered by bound host IgG [236]. The
in the bladder and urogenital tract causing hema-
immunoglobulin is partially cleaved by a par-
turia and fibrosis of the bladder. In severe
asite protease to produce peptides which may
cases malignancies may develop. Eggs from
inhibit macrophage activation and thus depress
S. mansoni and S. japonicum are trapped in the
macrophage-mediated, IgE-dependent destruc-
liver and bowels causing hepatomegaly. Sub-
tion (cytotoxicity) of the schistosomula.
acute disease is probably due to the passage
In the development of schistosomiasis vac-
of worms through lungs leading to cough, pul-
cines, the findings regarding age-dependent host
monary infiltrates, and fever [227]. Mainly in
resistance will require application and efficacy
S. japonicum, acute schistosomiasis occurs 5 –
at a very early age, before the child is exposed to
7 weeks after heavy primary infection. The ill-
natural infections [237]. The search for antigens
ness if often associated with diarrhea, fever, hep-
that mediate protective immunity has concen-
atosplenomegaly, resulting in liver fibrosis and
trated on the exposed outer surface of the young
ascites.
schistosomula [238]. About 90 % of exposed
epitopes consist of carbohydrates that crossre-
Immunity and Vaccine Design. An age-
act with Schistosoma egg antigen. Antibodies
dependent resistance to reinfection af-
to surface polypeptide antigen do not generally
ter chemotherapy was demonstrated with
crossreact with egg antigen but are present on
S. haematobium and S. mansoni [228], [229].
the surface membrane of adult worms. Mono-
This is associated with an increased lympho-
clonal antibodies to some of these molecules
cyte proliferation in response to egg, cercar-
seem to confer partial resistance when passively
ial, and adult worm antigen [230], [231]. Lack
administered to animals and could thus be poten-
of reinfection is also related to the eosinophil
tially protective antigens [239–241]. A range of
count [228], [232]; human eosinophils mediate
candidate vaccine antigens of S. mansoni have
antibody-dependent damage to the schistoso-
been identified, several have been cloned and
mula of S. mansoni [233]. The main immune
expressed in Escherichia coli, yeast, or vac-
mechanisms involved in schistosomiasis are
cinia virus [238]. This important achievement
summarized in Table 2 and reviewed in [234].
40 Immunotherapy and Vaccines
will facilitate the production of large amounts
of polypeptides for vaccination trials.
One of the most promising candidates is
a schistosomula surface polypeptide with a
molecular mass of 28 000. The gene coding
for this polypeptide has been cloned [242] and
expressed in E. coli. Immunization with this
recombinant antigen induced a high level of
serum cytotoxicity towards schistosomulas in
the rat, hamster, and monkey. Significant pro-
tection against a natural challenge infection with
live cercariae was obtained in rats and hamsters
[242]. These results can be viewed with opti-
mism for the development of a schistosomiasis
vaccine.
4.1.2. Vaccines against Nematodes
4.1.2.1. Gastrointestinal Nematodes
The most prevalent and pathogenic gastroin-
testinal nematode parasites of humans belong to
the genera Ascaris, Strongyloides, Trichinella,
Trichuris, Ancylostoma, and Necator. Although
many species parasitize deeper tissues of the
body, the majority are intestinal. The intestine
has been maintained as a site for adult stage de-
velopment, whereas larval stages often invade
other host tissues. The intestine consists of a se-
ries of distinct parasite habitats (gut sections,
lumen, mucosa, etc.) each having its own char-
acteristics. Large worms such as Ascaris must
live within the lumen, smaller species such as
hookworms are associated with the mucosa. For
a long time research on immunity against in-
testinal worms was given low priority due to
the lack of knowledge about local immune re-
sponses in the gut. Studies have shown that in-
testinal worms are indeed subject to protective
immune responses, although these responses
differ somewhat from classical immunity in the
body because the worms live in the gut lumen.
Special features are (1) macromolecular antigen
uptake across the intact mucosal epithelium or
by specialized epithelial cells overlying Peyer’s
patches and (2) complexation of the antigen by
dimeric IgA secreted from the mucosa or in-
testinal IgG Fab fragments. Cells from the un-
derlying lamina propria participate in cytotoxic-
ity and hypersensitivity reactions and thus affect
mucosal structure and function. Furthermore, a
large variety of nonlymphoid effector cells oc-
cur within the intestine, including natural killer
cells, macrophages, neutrophils, eosinophils and
basophils; their numbers increase during para-
site infections.
No vaccine against gastrointestinal nema-
todes is presently available for human applica-
tion. However, promising, successful trials in
veterinary parasitology have initiated interesting
work in human gastrointestinal nematode infec-
tions and will probably result in vaccine supply
in the near future.
Hookworm Disease. Various species of the
family Ancylostomatidae are responsible for
hookworm disease in humans, the most impor-
tant being Ancylostoma duodenale and Necator
americanus. Over a fifth of the world’s popula-
tion is afflicted by this disease, mainly in trop-
ical and subtropical regions. Adult hookworms
have a length of about 0.7 – 1.8 cm and a hook-
like anterior end with a distinct mouth area. Fe-
males release eggs in the small intestine which
are excreted in the feces of the host. In humid
surroundings first-stage larvae hatch from the
eggs and develop into infective larvae, which
penetrate through the skin into new hosts. After
migration through lymph or blood vessels, the
larvae finally develop into mature adult worms
in the small intestine, completing the parasite’s
life cycle. Hookworms suck blood from mi-
crolesions in the host’s small intestine, thereby
causing chronic gastrointestinal blood loss, ane-
mia, and hypoalbuminemia. With a large worm
burden death may occur. Patients usually suffer
from extreme weakness, pallor, secondary res-
piratory tract infections, skin irritations, heart
palpitations, and gastrointestinal distress.
Immunity and Vaccine Design. Nematode
parasites present special problems for the host’s
protective immune response, because they pos-
sess a tough, protective, external cuticle. The
cuticle is both antigenic and immunogenic, but
it is doubtful whether responses directed against
its surface play a major role in immunity against
intestinal species. Protective responses are more
likely initiated by antigens released through the
orifices of the parasite [243]. One such antigen
with a potential vaccine function is a secreted
proteolytic enzyme [244]. Hookworms attached
to the mucosa secrete the enzyme from glands
in their mouths. The enzyme degrades the host

proteins and inhibits blood coagulation which
permits the hookworms to feed for an indefi-
nite period of time. Immunizing the host against
this enzyme would result in inhibition of any
enzyme secreted by worms after a subsequent
challenge infection. The worms would then be
unable to feed. The gene coding for the enzyme
in question has been cloned. Future investiga-
tions will have to demonstrate the applicability
of this kind of vaccine.
A vaccine has been developed for the control
of hookworm infections in dogs [245]. It was
based upon irradiation-attenuated infective lar-
vae and was administered parenterally. Although
the vaccine was highly effective in preventing
hookworm disease, it was withdrawn because it
did not completely prevent infection, and effec-
tive anthelminthic chemotherapy was available.
This example suggests that protective immunity
may also occur in human hookworm infections.
4.1.2.2. Tissue-Invading Nematodes
(Filariidae)
Many nematode species which live as adults
in the intestine, (e.g., Ascaris, hookworms, and
Trichinella) undergo development in parenteral
tissues. Other species are wholly confined to
these tissues and have no contact with the in-
testine (tissue-invading nematodes). This closed
habitation site requires special conditions in or-
der to obtain biological contact with the outside
world, especially for reproduction. In the major
group of tissue-invading nematodes, the Filari-
idae, the worms overcome this problem by using
bloodfeeding arthropods as intermediate hosts.
The female worms release microfilariae larvae
which circulate in the blood or accumulate in
the skin of the host. The arthropod feeds on the
blood and takes up the microfilariae which de-
velop into infective larvae. At a following blood
meal, infective larvae are reinoculated into new
human hosts.
Filariasis. The human disease filariasis com-
prises an extremely heterogeneous group of
diseases. The main filiarial parasite species
in humans are Wuchereria bancrofti and Bru-
gia malayi, which are both transmitted by
mosquitoes. The disease is widespread in tropi-
cal and subtropical regions affecting ca. 90×106
Immunotherapy and Vaccines 41
persons. Its early symptoms are fever, lymphan-
gitis, and lymphadenitis. A following chronic
stage is frequently characterized by more se-
rious clinical manifestation including elephan-
tiasis, hydrocele, and pulmonary eosinophilia.
Adult worms are found in the lymphatic system,
microfilariae may be found in blood.
Filiarial infections caused by Onchocerca
volvulus occur in Africa and South and Central
America, affects millions of patients, many of
whom become blind. Adult worms are generally
located under the skin, forming typical nodules;
less often they penetrate deeply into the tissues.
Pathology in onchocerciasis is due entirely to
the microfilarial stage. Microfilariae are present
in the skin and may penetrate into the eye, thus
leading to severe eye lesions and blindness.
Loa loa is prevalent in the forest areas of
West Africa. Adult worms penetrate into the tis-
sues, provoking transient edema. Severe clin-
ical symptoms are rare, loiasis being gen-
erally regarded as a benign infection [246].
Dipetalonema perstans, D. streptocerca, and
Mansonella ozzardi infect humans (Africa,
South America) but usually asymptomatically
and rarely cause significant diseases [247].
Many of the changes associated with filarial
infection are immunopathological in origin and
hypersensitivity reactions are important in their
development.
Immunity and Vaccine Design. Although fi-
larial parasites provoke a strong immune re-
sponse in the human host, the chronicity of these
infections implies the absence of a protective re-
sponse or the evasion of such responses by the
worms. There is also no direct evidence that fi-
larial infections in nature confer resistance to
reinfection with the same parasite species.
No vaccines against filarial parasites are
available. The development of new strategies for
immunological control depends on a thorough
understanding of immunological host – parasite
relationship. Many studies on protective im-
munity in animals have concentrated upon re-
sponses directed against larval stages. A vaccine
against microfilariae would inhibit transmission
of the disease, a vaccine against infective larvae
would provide protective immunity against pri-
mary infection of hosts. Inoculation with infec-
tive larvae attenuated by irradiation [248], [249],
only confers partial protection to a subsequent
challenge infection. Filarial vaccines based on
42 Immunotherapy and Vaccines
irradiated larvae cannot be used in humans with-
out first determining whether these attenuated
larvae induce pathological changes [250].
The target antigens of antimicrofilarial im-
mune responses are probably located on the sur-
face of the microfilariae; they are currently being
characterized [251]. Antibodies to surface anti-
gens mediate adherence of host cells to the mi-
crofilariae; this can result in worm killing [243].
Using Dipetalonema vitae as a model, IgE was
found to be the primary immunoglobulin in-
volved in cell adherence to the worm cuticle.
The first cell type to adhere is the eosinophil.
Subsequent, adherence of macrophages is fol-
lowed by release of lysosomal enzymes which
degrade the cuticle. In humans, a comparatively
long time is needed to establish such immunity.
This may be due to pronounced immunosup-
pression or because the microfilariae cover their
surface with host components and thus render
their antigenic surface epitopes less accessible
to the host’s immune system. The search for fi-
larial antigens that can safely and successfully
be used in humans is still continuing; it is still
not known how restricted series of antigens can
be used to protect natural hosts against first or
persistent infections with a complex, adaptable,
genetically diverse parasite. Recombinant DNA
technology and hybridoma technology may pro-
vide potential candidates for successful filarial
vaccines.
4.1.3. Vaccines against Cestodes
Etiologic Agents, Pathogenesis, and Epi-
demiology. This section deals with the hy-
datidosis and cysticercosis disease complexes,
which are caused by the larval stages of tape-
worms belonging to the family Taeniidae. The
most important causative agents of hydatido-
sis (echinococcosis) are Echinococcus granu-
losus and E. multilocularis, whose life cycles
involve a definitive and an intermediate mam-
malian host. The definitive hosts are carni-
vores (mainly dogs for E. granulosus and foxes
for E. multilocularis) in whose intestines the
adult stage worms occur. Intermediate hosts are
herbivorous and omnivorous species in which
the larvae (metacestodes) develop. Humans and
other intermediate hosts become infected by
ingesting eggs passed in the feces of defini-
tive hosts. The diseases caused by the metaces-
todes are referred to as cystic echinococco-
sis for E. granulosus and alveolar echinococ-
cosis for E. multilocularis. Echinococcus gran-
ulosus is prevalent throughout the world and
is a public health and economic problem in
many areas. Echinococcus multilocularis only
exists in the northern hemisphere and is rela-
tively frequently seen in the former Soviet Union
(Siberia), central Europe, northern China, Japan,
and Alaska. The fully developed metacestode
of E. granulosus is a typically unilocular, fluid-
filled cyst, which is located in the liver, lungs,
and other organs. Echinococcus multilocularis
metacestode conforms a vesiculated parasitic
mass in the liver of the host, it proliferates by
continuous exogenous budding with possible
metastasis formation in other organs.
The causative agent of cysticercosis is the
tapeworm Taenia solium. Humans are the oblig-
atory definitive hosts, pigs act as intermediate
hosts. Metacestode stage infection can also oc-
cur in humans and may result in infection of the
central nervous system by parasite larvae (neu-
rocysticercosis). Morbidity includes intracranial
hypertension, basal arachnoiditis, focal neuro-
logical deficits, and dementia. Hyperendemic ar-
eas are found in Latin America, Africa, and Asia;
areas of lower endemicity occur in southern and
eastern Europe.
Immunity and Vaccine Design. Host pro-
tective immunity is a striking feature of re-
peated infection with cestodes in mammalian
intermediate hosts [225]. It plays a major role
in regulating natural transmission of these para-
sites, and substantial research efforts have been
undertaken towards development of vaccines
against cestodes of veterinary importance [252].
For human cestode infections, it is debatable
whether there is sufficient importance to war-
rant the research required to develop a vaccine.
In areas of high egg contamination (e.g., the
Turkana district (Kenya) for E. granulosus, St
Lawrence Island for E. multilocularis, or Mex-
ico for T. solium), vaccination should be given to
very young persons, as patients usually become
infected at very young age. An ideal human vac-
cine requires complete, long-lasting protection.
Experiments in veterinary parasitology, how-
ever, demonstrated exactly opposite results. Vac-

cination of animals against Taenia species re-
sulted only in marked reduction of cyst numbers
which persisted only for a maximum of one year.
In addition, strong adjuvants had to be employed
which are not tolerated by humans. Other cri-
teria may be important; a review of protective
immune mechanisms is given in [253].
Little attention has been paid to the vacci-
nation of definitive hosts (dogs and foxes for
Echinococcus, and humans for T. solium). Ex-
periments [254], [255] demonstrated an im-
mune response after infection as well as highly
significant suppression of egg production by
E. granulosus after immunization of dogs with
secretory antigens derived from adult tape-
worms. This approach seems to be the most
likely control measure for reducing infection
risk in humans.
4.2. Malaria Vaccine
Malaria remains a major health problem in many
tropical and subtropical countries and affects
hundreds of million of people each year. A ma-
jor effort was made to control malaria from
1950 – 1970 with insecticides (→ Insect Con-
trol) and antimalarial drugs (→ Chemotherapeu-
tics, Chap. 3.3.). The initial remarkable results
have been difficult to maintain. The advent of
drug-resistant parasite strains and of insecticide-
resistant mosquito vectors are major obstacles
in the effort to control malaria. Since the early
1970s new approaches have been explored such
as vector control through biological agents and
control of malaria infection through vaccines.
The development of malaria vaccine has re-
ceived considerable impetus: first, because im-
munization of animals with whole parasites can
induce a degree of protection equal or superior
to that induced following natural infection [256–
258]; second, because in vitro culture systems
have been developed for the blood and hepatic
stages of the malarial parasite Plasmodium fal-
ciparum [259], [260]; and third, because mono-
clonal antibody and recombinant DNA tech-
niques have been used to identify and produce
the parasite polypeptides possibly involved in
the development of protective immunity.
Immunotherapy and Vaccines 43
4.2.1. Strategy for Malaria Vaccine
Development
More than 100 species of malarial plasmodia
are known, but only four infect humans: Plas-
modium falciparum, which is responsible for the
majority of human deaths, P. vivax, P. malariae,
and P. ovale.
The life cycle of the plasmodia is complex
(Fig. 4). The female anopheline mosquitoes in-
oculate sporozoites into the blood of the verte-
brate host. Within minutes the sporozoites in-
vade the liver parenchymal cells (hepatocytes)
where they divide asexually and develop into
merozoites which rupture the hepatocytes and
reenter the blood. In the subsequent erythrocytic
cycle, the merozoites invade the red blood cells
and mature into schizonts within 48 – 72 h de-
pending on the species. The mature schizonts re-
lease merozoites which invade new erythrocytes.
The erythrocytic cycle is responsible for the
clinical manifestations of malaria. Some mero-
zoites differentiate into sexual stages called ga-
metocytes which are ingested by the mosquito.
Fertilization of the gametes occurs solely in
the midgut of mosquito. The resulting zygotes
develop into ookinetes and then into oocysts.
Sporozoites are released from mature oocysts
and migrate to the mosquito salivary glands. The
cycle is then repeated.
This complex life cycle involves continuous
morphologic, enzymatic, and antigenic changes
which are linked to the parasite’s environmen-
tal adaptation to the host. The invasive stages
of the parasite (sporozoites, merozoites, ga-
metes) have unique, stage-specific surface de-
terminants. Furthermore, immunologic crossre-
activity exists between plasmodia species and
between the developmental stages of a given
species. However, immunization experiments in
animal models have demonstrated that the anti-
genic determinants involved in protective re-
sponses are species- and stage-specific. Three
types of malaria vaccine can therefore be de-
vised, based on:
1) sporozoites,
2) asexual stages (merozoites, schizonts), and
3) sexual stages (gametes).
In addition, stagespecific parasitic antigens
are expressed on liver cells containing mero-
zoites [261] and may also be candidates for vac-
44 Immunotherapy and Vaccines
Figure 4. Plasmodium falciparum life cycle showing targets for vaccine development
cine development. A favored approach is the de-
velopment of a multivalent vaccine containing
components of several malaria stages. The se-
lection of defined parasite components versus
whole parasites (sporozoites, merozoites, ga-
metes) for vaccine development is indicated by
the following reasons:
1) Large-scale production and purification of
whole parasites is not feasible.
2) Parasites cannot be obtained free of host
components (e.g., mosquito salivary glands,
erythrocyte membranes) which may induce
adverse autoimmune reactions.
3) Most of the parasite components are irrele-
vant to the induction of protective responses
and their inclusion may impair truly protec-
tive responses or induce immunopathologic
lesions in the host.
The strategy for the development of malaria
vaccine is as follows:
1) Identification and selection of malaria anti-
gens that are the target of protective immune
responses.
2) Functional, biochemical, and immunological
characterization of these antigens, including
identification of B and T cell epitopes.
3) Cloning of the genes coding for protective
antigens, determination of DNA and amino
acid sequences–what is the level of antigenic
diversity?
4) Production of candidate protective antigens
or epitopes by genetic engineering or chem-
ical synthesis.
5) Evaluation of candidate protective antigens
in terms of production of antigens, safety
tests, adjuvants, carriers, etc.
6) Immunization trials in monkeys and human
volunteers.
4.2.2. Sporozoite Vaccines
Following invasion of the hepatocytes, the
sporozoites develop into thousands of mero-

zoites, each of which may invade an erythrocyte.
Obviously a vaccine which neutralizes sporo-
zoites before their entry into liver cells or within
liver cells would optimally prevent malaria in-
fection. Vaccination with attenuated, irradiated
sporozoites in rodents, monkeys, and humans
induced complete protection against malaria
[262], [263]. The induced immunity is species-
and stage-specific but not strain-specific. It is
at least partially mediated by antibodies as is
shown by the protection against sporozoite chal-
lenge afforded by the passive transfer of anti-
sporozoite monoclonal antibody [268] and by
neutralization of sporozoite infectivity follow-
ing incubation with the serum of protected ani-
mals. Deposition of specific antibodies on the
sporozoite surface results in the formation of
a tail-like precipitate; this is called the circum-
sporozoite (CS) reaction. Indirect evidence sug-
gests that antisporozoite antibodies may play a
role in human malaria [264], [265].
The control of sporozoite-induced infection
also involves T cell-dependent mechanisms.
In animal models, effector T cells can me-
diate antisporozoite immunity via antibody-
independent mechanisms; for example, immu-
nization with irradiated sporozoites can protect
B cell-deficient mice against subsequent chal-
lenge infection with viable sporozoites [266].
Activation of malaria-specific T cells by malaria
antigens induces the secretion of lymphokines
which may act directly on the malaria parasite
or indirectly by activating host effector systems
[267].
The CS protein has been identified in several
plasmodia species. Passive transfer of mono-
clonal antibodies directed against the CS protein
protects mice from sporozoite challenge infec-
tion [268]. Indirect evidence suggests that CS
protein is involved in the binding and penetra-
tion of sporozoites into liver cells. Similarly,
fragments of monoclonal antibodies against CS
prevent the attachment of sporozoites to hepato-
cytes in vitro [269].
The gene coding for the CS protein has
been cloned in several plasmodial species [270],
[271]. The protein contains a central block
of tandemly repeated amino acids which vary
in number and sequence among the different
malaria species [272], [273]. For example, in
P. falciparum the central area consists of 37
copies of the tetrapeptide asparagine – alanine –
Immunotherapy and Vaccines 45
asparagine – proline (NANP) and four copies
of the tetrapeptide asparagine – valine – aspartic
acid – proline (NVDP). The regions flanking
the repeats are more highly conserved between
species than the repeats. Within a species, lim-
ited variations also occur outside the repeats
[274].
The repeats cover the surface of mature
sporozoites; for P. falciparum ca. 108 molecules
of NANP are expressed on the membrane of
mature sporozoites. The NANP repeats are the
target of protective monoclonal antibodies pas-
sively transferred in vivo. The antibody response
against sporozoites in humans is also mainly di-
rected against the NANP repeats [275], [276].
In view of these findings, and because the
NANP repeats are present on all the isolates
of P. falciparum tested, two malaria vaccines
based on NANP repeats have been prepared and
tested on human volunteers. The first, produced
by DNA recombinant technology, consisted of
32 repeats of NANP and NVDP fused to a 32
amino acid tail. The second was composed of
three NANP repeats (NANP 3) conjugated to
tetanus toxoid. Both formulations use aluminum
hydroxide as adjuvant [277], [278]. The vac-
cines were safe and did not induce adverse reac-
tions. Volunteers with high antisporozoite anti-
body titers were challenged with sporozoites and
some were protected or presented a delay in ap-
pearance of parasitemia. Protection was shown
in individuals with the highest antibody titers.
An antisporozoite vaccine must induce high
antibody titers for a prolonged period and ideally
a boosting effect should occur following expo-
sure to sporozoites. Higher antibody responses
can be obtained by changing the formulation
and concentration of the immunogens and by
using other adjuvants or live-attenuated vectors
(vaccinia virus, salmonella) that carry and ex-
press the gene coding for the CS protein. Opti-
mal antibody formation is dependent on the col-
laboration of T helper cells and B cells. In the
two sporozoite vaccines tested, T cell help was
provided by foreign protein (tetanus toxoid) but
was unable to boost the antiNANP antibody re-
sponse following exposure to sporozoites. More
efficient sporozoite vaccines should contain T
cell epitopes derived from sporozoites and ide-
ally from the CS protein; in this context, proper
help is provided by sporozoite-specific, primed
T cells and can lead to optimal antiNANP anti-
46 Immunotherapy and Vaccines
body production by B cells. In a mouse model
the response to some of the T cell epitopes of the
CS protein is restricted by antigens of the ma-
jor histocompatibility complex (MHC class II)
[279–282]. In human populations, only three im-
munodominant epitopes are located in polymor-
phic regions of the CS protein outside the repet-
itive area. Since T cells have exquisitely specific
reactivity, it follows that the polymorphism of T
cell determinants may be responsible for a lack
of proper help following exposure to sporozoites
with T cell areas on CS that are different from
those present on sporozoites responsible for pre-
vious infections in the same individual.
Therefore, it seems that more efficient vac-
cines should be based on either native malaria
polypeptide(s) or cocktails of synthetic polypep-
tides containing multiple B and T cell epitopes.
These epitopes should be selected in relation to
constant and variant parasite components and
in relation to epitope binding and recognition
by components of the major histocompatibility
complex of the human host.
4.2.3. Asexual Blood Stage Vaccine
The multiplication of asexual blood stages (
merozoites and schizonts) is responsible for the
morbidity and mortality associated with malaria.
The level of parasitemia usually correlates with
the severity of malaria infection. Immunity to
malaria is mostly acquired but natural immunity
also plays a role. Several single-gene disorders
affecting erythrocytes, (e.g., sickle cell anemia,
the thalassemias, and glucose phosphate defi-
ciency) reduce the severity of malaria infection.
Another genetic characteristic, the lack of the
Duffy blood group antigens, is associated with
complete resistance to P. vivax infection [283];
the Duffy blood group antigen or a closely as-
sociated antigen may be the receptor for P. vivax
merozoites at the surface of erythrocytes.
The development of acquired resistance to
malaria depends on the frequency and duration
of the exposure to the parasite [284]. In endemic
areas, babies born to immune mothers are re-
sistant to malaria during the first three months
of life due to the presence of maternal antibod-
ies transferred during gestation. Later they suf-
fer from severe, recurrent attacks; most deaths
due to malaria occur in young children. From
adolescence to adulthood there is a decrease in
the severity and frequency of malaria attacks but
sterile immunity is probably never achieved. In
this context two types of vaccine based on asex-
ual blood stages can be envisaged; (1) a vaccine
which is more efficient than nature and leads
to sterile immunity (i.e., infection no longer de-
tectable) or (2) a vaccine capable of attenuating
the parasite load by transforming the immune
system of a non immune individual into that of
an adult living in an endemic area.
The immune response to blood stages is
complex and is directed against several anti-
gens. Both antibody-mediated responses and
cell-mediated, antibody-independent responses
control asexual blood-stage infection. In hu-
mans, passive transfer of immunoglobulins puri-
fied from the sera of immune adults abort malaria
infection in nonimmune infected children [285].
The antibodies may react with the surface of the
merozoites and provoke their lysis upon addi-
tion of complement, or enhance their phagocy-
tosis by mononuclear cells, or simply inhibit the
binding of merozoites to erythrocytes. Other tar-
gets for antibodies are antigens on the surface
of erythrocytes containing schizonts; binding of
antibodies to schizonts may also lead to their
destruction by phagocytosis [286] or induce the
endothelial release of schizonts which may be
later destroyed in the spleen [287]. Immunity
to asexual blood stages also operates through
a variety of antibody-independent mechanisms:
T cell-dependent release of lymphokines, induc-
tion of oxidizing radicals leading to intracellular
death of the malaria parasites, and activation of
mononuclear cells in the spleen.
Immunization with merozoites and/or sch-
izonts in a variety of plasmodia – host systems
resulted in partial to almost complete protection
[288]. Subsequent investigations were aimed at
the characterization of components capable of
inducing immunity. Several hundreds of asex-
ual blood stage components can raise an im-
mune response but only very few of the evoked
responses are helpful to the host. Characteris-
tics of some of the candidate antigens for asex-
ual bloodstage vaccines are discussed in Sec-
tions 4.2.3.1, 4.2.3.2, 4.2.3.3, 4.2.3.4.

4.2.3.1. Merozoite Surface Antigens
A protein with a molecular mass of 190 –
200 kDa has been identified at the surface of
P. falciparum schizonts and merozoites [289],
[290]. During maturation of schizonts this
polypeptide is processed into several compo-
nents, one of them (M r 83 000) being the main
surface component of the merozoites [290]. An
important feature of the 190 – 200 kDa protein
is its genetic polymorphism [291–294]. The
gene coding for the protein can be divided into
blocks ranging in homology among different
P. falciparum isolates from 10 – 87 % at the ami-
no acid level. A relatively short region of vari-
able tripeptide repeats is found close to the N
terminus. The blocks encoding for the N and C
terminal sequences are highly conserved.
Immunization with the 190 – 200 kDa protein
derived from P. falciparum in monkeys [295–
297] and with an analogous protein from P. yoelii
[298] can induce at least partial protection. Im-
munization with synthetic polypeptides corre-
sponding to defined parts of the molecule (for
example, the N terminus and amino acids 277 –
287) also confer partial protection in monkeys
[299], [300].
An antigen with a molecular mass of 51 kDa
is also expressed at the surface of P. falciparum
merozoites and is the target of inhibitory mono-
clonal antibodies. It contains variant and con-
stant epitopes for various P. falciparum isolates.
There is considerable antigenic diversity
among P. vivax isolates as regards the compo-
nents exposed at the surface of merozoites.
4.2.3.2. Rhoptry Antigens
Rhoptries are apical organelles of the mero-
zoites which release their contents onto the ery-
throcyte membrane during invasion. A mono-
clonal antibody directed against a rhoptry pro-
tein of a rodent malaria, P. yoelii, reduced the
virulence of the infection and a monoclonal an-
tibody directed against 82 and 41 kDa compo-
nents of P. falciparum inhibited the growth of
P. falciparum in vitro [301–303]. The 82 kDa
component is processed into 82 and 65 kDa com-
ponents [304].
The 82 kDa polypeptide is membrane-bound
through a glycosyl – phosphatide – inositol an-
Immunotherapy and Vaccines 47
chor; hydrolysis of its anchor activates the pro-
teolytic activity of the 76 kDa polypeptide and
may play a role in the invasion of erythrocytes
by merozoites [305]. The 41 kDa polypeptide
displays aldolase activity [306]. Interestingly,
both the 76 kDa and the 41 kDa components
can induce at least partial protection against
P. falciparum infection in monkeys [307], [308].
The gene coding for the 41 kDa protein has been
cloned and presents two interesting characteris-
tics in terms of vaccine development; absence
of variable amino acid repeats and almost com-
plete conservation of the amino acid sequences
among isolates from P. falciparum [306].
Another rhoptry antigen of P. falciparum
with a molecular mass of 225 kDa has been iden-
tified in the peduncle of the rhoptries. It is syn-
thesized as a 240 kDa polypeptide which is pro-
cessed into a 225 kDa protein during schizogony
and is quantitatively recovered in the culture su-
pernatant following merozoite invasion [309].
A third set of rhoptry-associated proteins
is the 105 – 130 – 140 kDa complex composed
of three coprecipitating but unrelated proteins
[310]. The 225 kDa proteins and the 105 – 130 –
140 kDa complex have not been evaluated in im-
munization trials.
4.2.3.3. Antigens Associated with the
Membrane of Infected Erythrocytes
The ring-infected erythrocyte surface antigen
(RESA) is a P. falciparum antigen with a molec-
ular mass of 155 kDa. It is synthesized in tropho-
zoites, accumulates in the merozoite, and fol-
lowing invasion becomes associated with the
membrane of erythrocytes containing ring forms
of the parasite [311], [312] but is not directly ac-
cessible on the external erythrocyte surface. An-
tiRESA antibodies inhibit the multiplication of
asexual blood stages in vitro and may interfere
with the invasion process [313]. The gene cod-
ing for RESA has been cloned and sequenced
[314]. It contains two blocks of repetitive ami-
no acid sequences which are the immunodomi-
nant regions of the molecules in terms of anti-
body response. Antibodies directed against the
RESA repeats crossreact with at least six other
asexual blood stage components. Aotus mon-
keys have been immunized with fusion proteins
corresponding to various areas of the RESA and
48 Immunotherapy and Vaccines
with synthetic polypeptides corresponding to
the repetitive sequences [315]. Partial protection
was observed in some groups of animals; work
is in progress to optimize the efficacy of immu-
nization based on RESA-derived molecules.
Erythrocytes containing mature asexual
blood stages of P. falciparum attach to endothe-
lial cells lining the venules of deep tissues.
This cytoadherence of mature parasites pre-
vents their passage through the spleen and thus
their exposure to localized destructive mech-
anisms. Electron-dense protuberances (knobs)
on the plasma membranes of infected erythro-
cytes are implicated in cytoadherence [316].
The genes coding for two knob components
(knob-associated histidine-rich protein M r 85 –
105 kDa) and mature parasite-infected erythro-
cyte surface antigen (M r 240 – 300 kDa) have
been cloned [317], [318]. The two proteins dif-
fer antigenically among isolates and contain re-
peated amino acid sequences. Cytoadherence
can be inhibited by antisera in a strain-specific
manner [319]. However, an antigenically invari-
ant epitope has also been identified on the sur-
face of infected erythrocyte isolates and may be
an important antigen for vaccine development
[320].
4.2.3.4. Other Proteins and Synthetic
Peptides
A number of other antigens are also candidates
for vaccine development. P. falciparum requires
exogenous iron in the form of ferrotransferrin. A
malaria transferrin receptor at the surface of in-
fected erythrocytes transports bound ferrotrans-
ferrin to the parasite and may be used as a target
for the vaccine [321]. Glycophorins exposed at
the erythrocyte surface may act as ligands for
P. falciparum merozoites and P. falciparum pro-
teins have been identified which either bind to
glycophorins or to human erythrocytes [322],
[323]. A prominent antigen of P. falciparum with
an apparent molecular mass of 126 – 140 kDa
is associated with merozoite release. The gene
coding for this protein has been cloned and con-
tains at least two stretches of amino acid repeats,
one being composed of polyserine repeats [324].
Monkeys immunized with this protein are pro-
tected from a lethal challenge infection [295].
Several immunization trials have been con-
ducted in monkeys using synthetic peptides de-
rived from asexual blood stages of P. falciparum
coupled to carrier proteins [298], [299], [315].
A partial protective response was observed with
peptides corresponding to various areas of the
190 – 200 kDa protein, to RESA, and to frag-
ments of parasite components identified by their
molecular mass of 55 and 35 kDa [299]. Syn-
thetic hybrid polymer – proteins containing sev-
eral peptides corresponding to epitopes of 195 –
200 kDa, RESA, 55 kDa, 35 kDa, and CS pro-
tein have been used for immunization of human
volunteers [324]. The vaccine was well toler-
ated and no adverse effects were observed. All
the immunized and control volunteers had patent
parasitemia but the majority of the immunized
volunteers were able to control their parasitemia
in the absence of drug therapy. The immune re-
sponse was low in terms of specific antibody
production, and cell mediated responses as mea-
sured by proliferation assays was undetectable.
4.2.4. Sexual Stages–Transmission Blocking
Immunity
The transmission of malaria from the vertebrate
host to the mosquito vector is effected by sexual
parasite stages–the gametocytes–which develop
from merozoites. Within the vertebrate host the
gametocytes are surrounded by the erythrocyte
membrane; following ingestion by the mosquito
vector, the gametes become extracellular. The
female gametes are fertilized by the male ga-
metes in the midgut of the mosquito to pro-
duce zygotes which develop into ookinetes. The
ookinetes penetrate the midgut wall where they
remain to form oocysts in which the sporozoites
develop.
In the vertebrate host, the sexual stages do not
produce illness, and their intracellular localiza-
tion prevents direct attack by host effector mech-
anisms. Within the midgut of the mosquito the
extracellular gametes are exposed to antigamete
and/or antizygote antibodies from the vertebrate
host that are ingested with the mosquito’s blood
meal. The antibodies partially or completely pre-
vent the development of sexual stages and subse-
quent production of sporozoites. Transmission
of the parasite is therefore blocked. This phe-

nomenon is termed transmission blocking im-
munity.
The development of vaccines based on sexual
blood stages could have an important impact on
the epidemiology of malaria in endemic areas
by reducing the level of malaria transmission.
Ideally, transmission blocking vaccines have to
be used in combination with vaccines based on
sporozoite and/or asexual blood stages (see Sec-
tions 4.2.2 and 4.2.3).
Transmission blocking immunity has been
induced by immunization with extracellular ga-
metes in several animals [325–328]. The in-
duced antigamete response is long lasting and
in some cases is boosted by malaria infection,
probably due to the gametocyte antigens in the
circulation of the vertebrate host [328], [329].
There is also evidence that in P. vivax malaria in
humans the antigamete response is boosted dur-
ing natural infection [330]. Addition of sera of
previously infected individuals to gametes can
prevent fertilization and development of oocysts
in mosquitos.
Specific targets for antigamete immunity
have been identified using monoclonal antibod-
ies in species including the human parasites
P. falciparum and P. vivax [331], [332]. The an-
tibodies act against the gametes by prevent-
ing fertilization and against the zygotes and
ookinetes by preventing further development. In
P. falciparum the target antigens for inhibition of
fertilization are polypeptides with a molecular
mass of 230 kDa and 45 – 48 kDa [331], [332].
Some of the epitopes on the 45 – 48 kDa anti-
gen have been defined and are the targets of in-
hibitory monoclonal antibodies [333]. New anti-
gens are expressed at the surface of the zygote
and ookinete and one of them (M r = 25 kDa) is
a probable target of inhibitory monoclonal anti-
bodies [331].
5. Immunotherapy
Since the late nineteenth century considerable
progress has been made in our concepts of pas-
sive immunotherapy and in the development of
preparations safe for human use; however, the
proper role for such therapy in clinical medicine
still needs to be defined.
By 1900 immune serum from various ani-
mal species had been used to treat pneumonia,
Immunotherapy and Vaccines 49
tetanus, diphtheria, and rabies. Human serum
was first used in 1907 by Cenci for the modifi-
cation of measles and later for mumps and per-
tussis [334]. Placental extracts prepared by am-
monium sulfate precipitation [335], [336] were
also employed and may be considered as the first
immunoglobulins prepared for human therapy
[336]. Placental material is still used as a source
of immunoglobulin.
The serious hypersensitivity reactions as-
sociated with animal serum proteins and the
risk of viral hepatitis with convalescent human
serum limited the use of serum therapy to life-
threatening infections. In the 1920s attempts
were made to separate the immune substances
from animal serum by alcohol or acetone treat-
ment. One such preparation, Huntoon’s antibody
solution, was administered intravenously to over
400 patients without the occurrence of ana-
phylaxis or serum sickness; however, pyrexia,
cyanosis, and dyspnea did occur and were im-
plicated in the deaths of three patients [337].
The introduction of antibiotics in the 1930s de-
creased the demand for serum therapy [338].
During this period, however, the experimental
basis for combination therapy with antimicro-
bials and hyperimmune animal serum was es-
tablished [339], [340].
A wide variety of biological products is
now available for immunotherapy, the most im-
portant being purified gamma globulin for in-
tramuscular injection (standard immune serum
globulin, ISG) and globulin for intravenous in-
jection (standard intravenous immunoglobulin,
IVIG). Hyperimmune globulins with a high an-
tibody titer against specific pathogens are also
used. Additional preparations include antitox-
ins, plasma, and other blood products.
5.1. Gamma Globulin Preparations
5.1.1. Standard Immune Serum Globulin
Historical Aspects. In 1936 Arne Tiselius
separated serum proteins into four major frac-
tions by electrophoresis; subsequently he and
Kabat found that immunoglobulin occurred
predominantly in the gamma electrophoretic
fraction [341]. Cohn and colleagues devised
a procedure for recovering immunoglobulins
(gamma globulins) from serum on a large scale.
50 Immunotherapy and Vaccines

The serum proteins were fractionated by precip- cytes (mononuclear cells and neutrophils) and
itation with ethanol under carefully controlled basophils, respectively [349]. The IgG 3 has a
conditions of pH, temperature, protein concen- short half-life (nine days); IgG 4 is unable to bind
tration, and ionic strength [336], [342]. This pro- complement.
cess enriched and stabilized the gamma glob-
ulin from plasma at a relatively uniform anti-
body content while also denaturing most viruses
[334]. Such gamma globulin prepared from large
pools (>500 donors) of donor plasma were used
in the treatment of infectious diseases during
World War II [343].
Shortly after World War II gamma globulin
was shown to contain antibody titers adequate
for the prevention or attenuation of measles
[344], infectious hepatitis [345], and polio [346].
Since the report of hypogammaglobulinemia in
1952, and the demonstration that gamma globu-
lin administered on a monthly basis decreased
the incidence of infection [347], antibody re-
placement of this deficiency has been routine.

Properties of Gamma Globulins. Gamma

globulins occur at a serum concentration of
600 – 1200 mg/100 mL in adults, they represent
approximately 11 – 14 % of total serum proteins
[341] and 80 % of serum antibody [348]. Im-
munoglobulin G has a half-life in the normal cir-
culation of approximately 25 days (35 – 40 days
in patients with agammaglobulinemia) and is
synthesized by adults at a daily rate of 35 mg/kg
body weight [348]. Its rate of synthesis is regu-
lated by serum IgG levels. Immunoglobulin M
and IgA have lower serum concentrations and
shorter half-lifes than IgG. The molecular mass Figure 5. Preparation of immunoglobulin by Cohn – Oncley
of IgG has been estimated to be 145 000 [341] fractionation (“6/9 method”)
with 2.5 wt % being carbohydrate that is asso- Cohn fractionation of plasma by cold – alcohol procedures
ciated with the heavy chain (see also Fig. 5). yields a fraction II precipitate or “paste”. This is followed
by applying the Oncley procedure to Cohn fraction II ma-
terial to give immunoglobulin. The procedure is based on
The IgG isotype can be divided into four sub- solubility differences that depend on ethanol concentration,
classes (Table 3). The most abundant is IgG 1 ionic strength, pH, temperature, and protein concentration.
(60 – 70 % of total IgG) which binds to the Fc Precipitates are usually collected by centrifugation.
receptors of neutrophils and mononuclear cells
and to the first component of complement [349].
Preparation of Immune Serum Globu-
The IgG 2 subclass (20 – 30 % of total IgG) acti-
lin. Standard immune serum globulin (ISG,
vates the classical complement pathway poorly
gamma globulin) is prepared from the plasma of
but can activate complement by the alternate
pools of >1000 donors (see Fig. 5); → Blood,
pathway. It is more resistant to proteolysis than
Chap. 2.2.5.). This minimizes differences bet-
the other subtypes [348]. The antibody response
ween individual antibody levels to specific anti-
to pneumococcal and hemophilus polysaccha-
gens and ensures a broad range of antibody
rides may be related to the preimmune levels
specificities.
of this subclass [350]. The subclasses IgG 3
and IgG 4 bind to the Fc receptors of phago-

Table 3. Immunoglobulin G subclasses ∗
Property
Percentage of total IgG (adult)
Molecular mass (daltons)
No. of interchain disulfide bonds
No. of amino acids in hinge region
Half-life (days)
Fc receptor binding
mononuclear cells ∗∗
neutrophils
∗ Adapted from [239].
∗∗ ++ strong binding; + detectable binding; ± binding detected in some, but not all studies.
Standard immune serum globulin is usually
prepared as a 16.5 % injectable solution (165 mg
per mL) and represents a ca. 20 – 25-fold con-
centration of plasma IgG [336], [343]. This pro-
vides an effective dose of antibody in a relatively
small volume [343]. Because of the viscosity of
the preparation, it can only be given intramuscu-
larly or subcutaneously [343], usually through a
large gauge needle (16 – 18 ga). It has been esti-
mated that 1 g of ISG contains 4×1018 antibody
molecules with >107 specificities [334]. The
product, although highly stable at 4 ◦ C, can still
undergo proteolysis, presumably due to plasmin
contamination [336]. The ISG may also contain
blood group substances, IgA, and IgG dimers.
The presence of IgA can result in anaphylactic
reactions in patients who lack IgA [336].
5.1.2. Immunoglobulin for Intravenous Use
Intravenously administered gamma globulin is
needed to allow greater patient comfort, increase
acceptability, and provide larger quantities of an-
tibody. This is particularly important for patients
who have a small muscle mass (children) or in-
sufficient skin surface (burns), who are at risk
from uncontrollable bleeds (Wiskott-Aldrich or
other bleeding diatheses), or who need either
large repeated doses (immunodeficient) or rapid
onset of peak levels (intoxicated patients). Fur-
thermore, intravenous administration avoids the
local degradation of ISG antibody that may oc-
cur at the injection site [351].
In 1948 Cohn administered 25 – 50 mL of
ISG intravenously and saw no adverse effects
[336], [352]. Janeway and Cohn repeated the
experiment with a new preparation and induced
severe reactions after only 2 mL. This reaction
Immunotherapy and Vaccines 51
IgG 1 IgG 2 IgG 3 IgG 4
60 – 70 20 – 30 5 – 10 <5
146 000 146 000 170 000 146 000
2 4 11 2
15 12 62 12
21 – 23 20 – 23 7–8 21 – 23
++ + ++ ±
++ ± ++ +
was attributed to contamination of the prepara-
tion by staphylococcal enterotoxin [352]. The
intravenous administration of ISG was repeated
by Janeway in 1970 [336], and induced se-
vere cardiovascular (tachycardia, arrhythmias,
hypotension, severe chest pain), tachypneic,
and pyrogenic (fever, chills, malaise) reactions
[343]. These adverse effects were attributed to
the presence of immunoglobulin aggregates that
activated complement [353], [354].
To avoid the above-mentioned severe reac-
tions, methods were introduced to rid the prepa-
rations of the high molecular mass immunoglob-
ulin aggregates. Although they could be re-
moved by centrifugation, reaggregation usually
occurred and this method was not practical on
a large scale [355]. Consequently, attention was
turned to chemical modification.
Various methods of chemical modification
have been tried. Degradation with the enzymes
pepsin and plasmin resulted in the formation of
antibody fragments with immunologic activity
[336], [356]. In addition, since the proteolytic
enzyme was not removed, it may have contin-
ued to be active in the absence of lyophiliza-
tion; porcine pepsin also induced antibody for-
mation [354]. Reaggregation was also prevented
by acidification with hydrochloric acid [357]
and reduction – oxidation with dithiothreitol,
iodoacetamide, sulfite, or tetrathionite. Alkyla-
tion and acylation of immunoglobulin has also
been accomplished with β-propiolactone.
The IVIG preparations that have been re-
duced and alkylated or otherwise modified may
have impaired complement binding and altered
subclass distribution [358], impaired opsonic ac-
tivity in vitro [359], shortened serum half-life,
and decreased protective efficacy [360]. Con-
52 Immunotherapy and Vaccines
sequently, a new generation of native or in-
tact IVIG preparations has been produced with-
out modification by methods which include ad-
justment to pH 4, use of poly(ethylene glycol),
ethanol precipitation, ultra- or diafiltration, and
ion-exchange chromatography [353]. Four un-
modified products are licensed in the United
States: one prepared by adjustment to pH 4 in
the presence of traces of pepsin, one prepared
by ion-exchange chromatography and ultrafil-
tration, and a third by diafiltration, ultrafiltra-
tion, and adjustment of pH to 4 – 4.5. A fourth
product (Venoglobulin-1 from Alpha Therapeu-
tic Corporation) licensed in the United States but
produced in Japan is prepared by poly(ethylene
glycol) fractionation and ion-exchange adsorp-
tion.
Commercially prepared IVIG is usually sta-
bilized with a mono- or disaccharide, such as
10 % maltose [338] or glucose. This consider-
ably decreases the incidence of side effects that
accompany the infusion of IVIG [336], [357],
[361], perhaps by minimizing precipitation or
aggregation. Maltose can cause a mild diuretic
effect [362].
5.1.3. Hyperimmune Globulins and
Antitoxins
Hyperimmune globulins are high-titered prepa-
rations of ISG against a specific antibody. They
are prepared from the plasma of patients who
have recently recovered from the disease or who
have a high titer as a result of a previous vaccina-
tion or natural infection. Antitoxins for use in in-
toxication with botulinus or diphtheria toxin are
prepared from snake venom or in horses [363].
5.1.4. Production Requirements
The WHO requirements stipulate that current
immunoglobulin products
1) are prepared from pools of >1000 donors;
2) are free of kinins, plasmin, and prekallikrein
activity;
3) have a low IgA content;
4) are as free as possible from immunoglobulin
aggregates;
5) contain at least 90 % intact IgG without frag-
ments;
6) are as unmodified as possible so that opsonic,
complement, and other biologic activities are
maintained;
7) contain all IgG subclasses;
8) have high levels of antibody to at least two
bacterial species or toxins and two viruses (to
be ascertained by neutralization tests), and
9) contain at least 0.1 International Units (IU)
of antibody to hepatitis B and have a 1 : 1000
titer to hepatitis A [364].
In addition, manufacturers should specify the
diluent and any chemical modifications used. Fi-
nally a product can only be classified as a hyper-
immune immunoglobulin if the antibody level is
five times that of standard ISG preparations.
There is no consensus on which laboratory
tests should be used to predict the safety of IgG
(contact activation, anticomplementary activ-
ity) [336]. Safety requirements should, however,
consider the IgG half-life and virus transmission
(particularly of human immunodeficiency virus
HIV, but also of hepatitis B and non-A, non-B
hepatitis viruses) [353].
5.2. Prophylaxis with Immune Serum
Globulin
Passive immunization with ISG has been rec-
ommended for the short-term prevention of dis-
ease when vaccines for active immunization are
unavailable or when active immunization was
not given before disease exposure; it should be
given before expected contact or early in the dis-
ease incubation. In these situations active immu-
nization is always preferable because the anti-
bodies subsequently formed by the vaccine pro-
vide longterm protection. Immune serum glob-
ulin is also used for antibody replacement in pa-
tients who lack adequate serum levels of im-
munoglobulin (i.e., hypogammaglobulinemia)
[335].
The efficacy of ISG for the short-term pro-
phylaxis of specific infections was established
shortly after its widespread availability in the
1940s. These include:
Hepatitis. Stokes and Neefe showed that
ISG could prevent or modify the course of hep-
atitis A [335], [336], [345], but Krugman found

that ISG modified, but did not prevent this dis-
ease [365]. The ISG dose for treating hepatitis
A is 0.02 mL/kg body weight [336]. A hyper-
immune antibody preparation is also available
commercially. In the case of hepatitis B, ISG is
not recommended. Instead, hyperimmune glob-
ulin is given (see Section 5.3). Although ISG can
decrease the incidence of post-transfusion hep-
atitis [366], its use for non-A, non-B hepatitis is
considered optional [364].
Measles. The ability of convalescent serum
to modify the course of measles was demon-
strated in 1907 [335]. The ISG has also been
shown to prevent or attenuate the disease [343].
The introduction of active immunization with a
measles vaccine in 1963 (see Section 3.1) sig-
nificantly decreased the incidence of this disease
in the United States. Standard ISG is now rec-
ommended for infants under one year of age or
for immunodeficient patients within six days of
acute exposure to a case of measles [364].
Polio. If given early, ISG can modify the par-
alytic complications of polio [335], [346]. An
unexposed individual should receive 0.15 mL/kg
body weight.
Prevention of Infection in Patients with
Hypogammaglobulinemia. An intramuscular
ISG dose of 100 mg/kg given every 3 – 4 weeks
is recommended for patients with hypogam-
maglobulinemia. This maintains a serum level
of circulating IgG above 200 mg/mL [336] and
confers protection against a wide variety of in-
fections [342]. This minimum recommended
dose is based on a study by the British Medical
Research Council conducted between 1956 – 66
on 176 patients [367]. Although serum IgG lev-
els at this dose rarely rise to the normal range
[368], total replacement of IgG does not appear
necessary for preventing infection [369]. Al-
though a monthly dose of 200 mg/kg was found
superior to 100 mg/kg, most patients did not tol-
erate more than the 100 mg/kg [361]. Two other
studies also showed that higher doses of ISG de-
creased the incidence of acute infections [334].
Individualization of doses and their frequency
has also been suggested [368].
Rubella. Standard ISG is unreliable in the
modification of rubella [335], [337], [365]. High
Immunotherapy and Vaccines 53
titers of antibody are needed [365]. Neverthe-
less, the use of ISG is recommended for women
exposed to the disease during early pregnancy
[364].
Clinical Studies. On the basis of data avail-
able in 1980, a committee of the WHO observed
that it was “inappropriate” to use standard im-
munoglobulin for the prevention of infection in
premature infants, during the physiologic hy-
pogammaglobulinemia of infancy, or for mal-
nutrition. Its use is contraindicated in patients
with selective IgA deficiency [364]. Previous at-
tempts in the 1960s showed that ISG did not
prevent infection in a variety of clinical situa-
tions [370], [371] including multiple myeloma
[372]. Although ISG was unable to prevent up-
per or lower respiratory tract infections in chil-
dren or institutionalized, elderly adults, a sig-
nificant decrease was seen in the incidence of
mumps in children and fevers of unknown origin
in adults [373]. Its use in the prevention of in-
fection in burned patients has yielded conflicting
data [335]. In a study on Peruvian children with
burns over at least 10 % of their body surface
area, there was a 41 % incidence of septicemia
and 15 % mortality in control patients compared
to a 21 % incidence of septicemia and 6 % mor-
tality among those who received either plasma
or ISG [374]. However, in another study in the
United States administration of ISG did not ben-
eficially affect either the rate of septicemia or
mortality in burned patients [375]. More re-
cently, patients admitted to a burn unit in India
were randomized into groups that received ac-
tive immunization against Pseudomonas aerug-
inosa, passive immunization with a cold etha-
nol precipitate of plasma from normal immu-
nized volunteers, both immunologic treatments,
or neither. In children, but not adults, there was
a significant reduction in mortality following a
daily dose of just over 20 mg protein for three
days. There was a decrease in the incidence
of not only P. aeruginosa, but also other gram-
negative bacilli among those treated immuno-
logically [376].
The ability of ISG to prevent infections has
been most extensively studied in high-risk, pre-
mature infants. This population has a hypogam-
maglobulinemia that exposes them to a high inci-
dence of bacterial infection [377]. The more pre-
mature the infant the lower the IgG, since most of
54 Immunotherapy and Vaccines
the transplacental transfer of IgG occurs during
the last six weeks of gestation [370]. Children
do not attain adult levels of immunoglobulin un-
til two years of age. In two studies using 0.5 –
3 mL/kg, no prophylactic effect was observed
[371], [378], although there was a suggestion
that the high dose regimen may have had a ben-
eficial effect.
5.3. Prophylaxis with Hyperimmune
Globulins
Several hyperimmune globulins are commer-
cially available or undergoing development. Hy-
perimmune IVIG preparations are discussed in
Section 5.5.
Diphtheria [342]. Diphtheria antitoxin is
obtained from the blood of horses immunized
against diphtheria toxin. Individuals should
be tested for sensitivity to horse serum be-
fore being given the product. For prophylaxis,
1000 – 5000 IU of antitoxin is administered to
Schick-positive individuals exposed to diphthe-
ria. Higher doses (20 000 – 80 000 IU) are given
for treatment.
Hepatitis. A hyperimmune globulin is avail-
able for hepatitis A. In the case of hepatitis B,
a hyperimmune globulin is administered after
mucosal or percutaneous exposure (including
sexual contact) to an antigen-positive individual
[336]. This is also recommended for newborns
of antigen-positive mothers [365].
Mumps. Immune serum globulin has no ef-
ficacy in the modification of this disease, but
hyperimmune globulin does [343].
Pertussis. Hyperimmune globulin modifies
the disease. For example, 2.5 mL of hyperim-
mune antipertussis gamma globulin with a fol-
lowup dose at 5 – 7 days led to a 75 % reduction
in disease among exposed, nonimmune individ-
uals [343]. However, use of hyperimmune glob-
ulin has been superseded by antibiotics [379].
Rabies. Hyperimmune globulin is recom-
mended in addition to active immunization fol-
lowing exposure to a possible or proven case of
rabies [364] (see Section 3.7). Combined active
and passive immunization against rabies have
been shown to be superior to active immuniza-
tion alone [336].
Rho(D) immune globulin. This hyperim-
mune globulin is recommended for rhesus-
negative mothers who deliver rhesus-positive
infants [364].
Tetanus. The efficacy of tetanus immune
globulin in either the prophylaxis or treatment
of tetanus has not been clearly shown in clinical
study. Current recommendations are for 250 IU
injected intramuscularly for individuals whose
tetanus immunization is not known and whose
wound is of a sufficiently serious nature. For
treatment of clinical tetanus, doses of 3000 –
6000 units are recommended.
Vaccinia. Hyperimmune globulin is used for
prophylaxis against smallpox and for treat-
ment of the dermal complications of vaccination
[343].
Varicella. The efficacy of standard ISG in
the prophylaxis of this disease is not well-
established [335], [365]. Use of the hyperim-
mune product, however, is indicated in individ-
uals who have never had chicken pox, who are
exposed to acute cases and belong to a high risk
group (newborns or immunocompromized pa-
tients), and women who are pregnant [364].
Hyperimmune Globulins under Devel-
opment. Globulins with high titers to cy-
tomegalovirus, and Pseudomonas aeruginosa
have been tested. A product (bacterial polysac-
charide immune globulin) has been tested that
is prepared from the sera of donors immunized
with licensed vaccines against Hemophilus in-
fluenzae (type b), pneumococci, and meningo-
cocci.
Certain subpopulations of children, such as
native American Indians and Eskimos, are at
particularly high risk of acquiring serious infec-
tion with encapsulated bacteria. Standard ISG
preparations have failed to show a beneficial
prophylactic effect for many types of infection.
Immune serum globulin made hyperimmune to
polysaccharide antigens by immunizing volun-
teers with licensed vaccines against pneumo-
cocci, H. influenzae, and meningococci has been

used to immunize Apache infants. There was
a significant reduction in the incidence of sys-
temic disease caused by H. influenzae and pneu-
mococci during the first six followup months as
well as a significant decrease in the incidence of
bacteremia [380].
5.4. Therapy with Immune Serum
Globulin
Data suggest ISG is active in bacterial infections
in animal models [379], [381] and may be syner-
gistic with antibiotics [339], [340], [382]. How-
ever in 1968 Schless and Harrell [379] and
others [336] observed that there was little evi-
dence to support its therapeutic efficacy in es-
tablished infection in humans; they suggested
the need for a controlled clinical trial of ISG
in the treatment of systemic infection in patients
that were not deficient in antibody [379]. Indeed,
since the most functionally-active antibody to
gram-negative bacteria is IgM and not the IgG
found in ISG, little benefit was to be expected
[342]. The ISG is active against a wide vari-
ety of human pathogens in animal models of
infection [379]. When used with antimicrobial
agents, ISG has possible benefits in experimen-
tal infection with a wide variety of organisms
[382], [383] and in clinical infections in humans
[339], [340], [384], [385]. Large daily doses of
ISG administered intravenously for 10 days to
patients with leukemia and fever was well toler-
ated [386]. However, no benefit was found for
patients who received the ISG in addition to anti-
biotics (compared to antibiotics alone). The ad-
ministration of ISG to shorten the course of in-
fection in children under two years of age was
ineffective [387].
5.5. Prophylaxis and Therapy with
Intravenous Immunoglobulin (IVIG)
In studies on the prophylactic or therapeutic ef-
ficacy of ISG, the possibility that higher doses
of standard immunoglobulin might improve ef-
ficacy was a recurring theme. With the availabil-
ity of IVIG, larger volumes of immunoglobulin
could be administered directly into the blood-
stream (see Section 5.1.2). In addition, with the
Immunotherapy and Vaccines 55
ability to screen large numbers of samples for
antibody levels or to immunize volunteers with
an increasing number of vaccines, an increas-
ing number of publications have examined the
efficacy of passive immunotherapy, particularly
with hyperimmune IVIG preparations, in the
treatment and prophylaxis of infectious diseases.
Intravenous immunoglobulin G is effective in
the prevention of infection in patients with hy-
pogammaglobulinemia [358]. It has also been
shown to be effective for treating chronic in-
fection in such patients who developed sinopul-
monary infection despite ISG maintenance ther-
apy [351].
5.5.1. Viral Infection
Since cytomegalovirus (CMV) is a frequent
cause of infection in patients undergoing organ
transplantation and effective antiviral therapy is
lacking, there has been considerable interest in
the use of IVIG that is hyperimmune in CMV
antibody for both the prophylaxis and treatment
of CMV infections.
Prophylaxis. Hyperimmune CMV-IVIG (to-
tal dose 550 mg/kg) was shown to decrease
the attack rate of symptomatic CMV infec-
tion among CMV-negative patients who re-
ceived kidneys from CMV antibody-positive
donors from 60 – 21 % [388]. This treatment
also decreased the incidence of fungal and
parasitic infection. Hyperimmune CMV-IVIG
(3×200 mg/kg doses) also prevented mortality
and interstitial pneumonia from CMV for 120
days in leukemic patients who underwent bone
marrow transplantation [389]; however in an-
other study it did not prevent acquisition of in-
fection [390]. In another trial, nonimmune IVIG
with a high antiCMV titer did not decrease the
incidence of CMV seroconversion in bone mar-
row transplantation patients [391]; however, the
incidence of symptoms and interstitial pneumo-
nia decreased. In contrast, CMV-IVIG given to
CMV-negative patients undergoing bone mar-
row transplantation had no effect on either the
prevention of new disease or the amelioration
of established disease [392]. Therapy of bacte-
rial infections following infection with human
immunodeficiency virus are discussed in Sec-
tion 5.5.2.
56 Immunotherapy and Vaccines
Therapy. Immunoglobulin has been used
sporadically in the treatment of viral disease.
High-titered CMV-IVIG from screened donors
did not show efficacy in the treatment of bone
marrow transplant patients with documented
CMV infection [393]. IVIG has been shown to
alter the course of echovirus encephalitis infec-
tion in three patients with hypogammaglobuline-
mia. Although no benefit occurred from giving
the IVIG intravenously in two of these patients,
intraventricular administration resulted in clin-
ical cures [394], [395]. The use of IVIG was
unable to alter the lethal course of polymyositis
secondary to echovirus in another patient with
hypogammaglobulinemia [396]. IVIG has also
been used experimentally in the successful treat-
ment of herpes infection in mice [397].
5.5.2. Bacterial Infection
Numerous studies have demonstrated the ef-
ficacy of hyperimmune IVIG in the prophy-
laxis and, if used early after infection, the treat-
ment of bacterial infection. In 1943 Alexan-
der showed that the combination of a sulfa drug
and animal hyperimmune sera was more effec-
tive in reducing mortality than either agent alone
[339], [340]. A number of studies have demon-
strated the efficacy of IVIG in both the pre-
vention and treatment of experimental infection
with H. influenzae in neonatal rat models [398];
with E. coli [399] and Klebsiella [400] in mouse
models; and with P. aeruginosa in neutropenic
and burned rodent models [401], [402]. Intra-
venous immunoglobulin hyperimmune to group
B streptococcal surface antigens can prevent and
treat serious bacteremia in experimental infec-
tion in monkeys [403].
Prophylaxis. Earlier data with ISG in both
experimental bacterial infection in animals and
in clinical infection in humans indicated that in
some situations (e.g., patients with hypogam-
maglobulinemia), exogenous standard gamma
globulin could prevent the acquisition of serious
bacterial infection. Similar studies with IVIG
have established the efficacy of these prepara-
tions [358].
Since earlier investigators believed that larger
doses of ISG might be effective in the preven-
tion of bacterial infection in high-risk neonates,
it is not surprising that similar studies have now
been reported with IVIG. In one nursery with
a high rate of infection, a single dose of IVIG
(120 mg/kg within 2 h of birth) decreased ac-
quisition of infection and mortality in preterm,
lowbirthweight neonates [404]. A second dose
at eight days, conferred no further advantage.
Prophylaxis with IVIG (0.5 mg/kg/week for four
weeks) was significantly better in preventing in-
fection and death in neonates, but only in the
subpopulation that weighed less than 1500 g and
had a gestational age of less than 34 weeks [370].
In another study, antibiotics were given either
alone or with IVIG to women 27 – 36 weeks
pregnant who had chorioamnionitis. Only high
doses (24 g/d for five days) of IVIG given after
the 32nd week of pregnancy prevented infection
in the delivered babies [405]. The investigators
concluded that little transplacental transfer of
IgG occurred before the 32nd week of gestation.
Data on the use of standard IVIG for the pre-
vention of bacterial infections in patients with-
out hypogammaglobulinemia is limited. High
levels of specific antibody are needed; these may
not be found in standard, nonimmune IVIG, de-
spite the possibility of delivering larger amounts
of IVIG than was the case with ISG. Hyper-
immune products have been shown to prevent
specific infections in experimental models (see
Sections 5.4, 5.5.2, and 5.6). The administration
of nonimmune IVIG (1000 mg/kg) before bone
marrow transplantation and weekly for 17 weeks
thereafter had no effect on the acquisition of ei-
ther bacterial or fungal infection [406]. Clinical
trials are being performed to test the efficacy of
IVIG in the prophylaxis of bacterial infection
in adults with chronic lymphocytic leukemia,
a condition which may be complicated by hy-
pogammaglobulinemia [407].
Therapy. Patients with human immunodefi-
ciency virus (HIV) infection (i.e., acquired im-
mune deficiency syndrome, AIDS) have a dys-
function in their humoral immune system which
includes both a decreased antibody response to
bacterial antigens and an altered distribution of
IgG subclasses [408]. Unlike adults, who tend
to acquire opportunistic infections (infections
in which immunodeficient individuals are in-
fected by organisms that are normally with-
stood by immunocompetent individuals), young
children and particularly infants infected with

HIV often resemble patients with primary hu-
moral immunodeficiency and tend to suffer from
bacterial infections. Consequently, prophylaxis
with monthly doses of IVIG has been used in
the management of these patients. In one pi-
lot study, a decreased incidence in episodes of
fever and bacteremia was noted. This was ac-
companied by clinical improvement, prolonga-
tion of life, and improvements in other immuno-
logic parameters [408]. In a 37 month old child
with AIDS, monthly doses of IVIG produced
increases in IgG 2 antibody and antibody to 12
pneumococcal serotypes, and prevented subse-
quent episodes of bacteremia [409]. On the basis
of these preliminary data, IVIG prophylaxis has
been advocated in the treatment of childhood
AIDS [408]. Periodic administration of IVIG
decreased lactic dehydrogenase activity (a pro-
posed indicator of pulmonary interstitial inflam-
mation) in adults and children with HIV infec-
tion [410].
Attempts to demonstrate a significant effect
in the therapy of bacterial infections with IVIG
have been much less successful than with its use
in prophylaxis. This may be attributable to the
shorter half-life of IVIG in the blood during in-
fection [368], [381]. The addition of IVIG to
standard regimens of antibiotics decreased the
mortality from documented bacteremia, particu-
larly among preterm infants. The number of sub-
jects studied was too small for statistical analy-
sis, however [411].
In adults, little data is available to support
the use of IVIG in the treatment of bacterial in-
fections. This may be due to the need for high
levels of antibody specific for the invading or-
ganism (rather than simply high levels of non-
specific antibody) as well as the need for prompt
initiation of therapy. In most experimental stud-
ies with hyperimmune IVIG, little benefit can
be shown if the exogenous immunoglobulin is
given more than 8 h after infection (see [381],
[383]). In clinical medicine, however, identifi-
cation of the time of onset of infection is often
difficult.
5.5.3. Noninfectious Diseases
5.5.3.1. Therapeutic Effect of IVIG
In 1981 it was reported that a patient who re-
ceived IVIG for hypogammaglobulinemia also
Immunotherapy and Vaccines 57
recovered from bleeding secondary to a co-
incidental platelet deficiency (thrombocytope-
nia) following the infusion [412]. Following
this fortuitous observation, experiments showed
that a high dose of IVIG could indeed re-
verse thrombocytopenia in the absence of hy-
pogammaglobulinemia; furthermore, a regimen
of 400 mg/kg/d for five days was effective treat-
ment for idiopathic thrombocytopenic purpura
(ITP) in children [413]. The use of IVIG and oral
steroids was also compared. Among those who
responded rapidly to treatment (62 %), IVIG was
as efficacious as steroids; however the slower
responders responded better to the IVIG treat-
ment. Administration of IVIG was accompanied
by a doubling of serum IgG; IgM levels rose
under both treatment regimens. The IVIG may
possibly reduce clearance of antibody-coated
platelets via an Fc-mediated mechanism (see
Section 5.5.3.2) [414]. The dosage regimen used
in the above study (400 mg/kg/d for five days)
has been used in many other studies on the non-
infectious uses of IVIG. However, in another
study 800 – 1000 mg/kg was given as a single in-
fusion to children with ITP with similar results
and no reported untoward effects [415]. Experi-
ence with IVIG for ITP has not been uniformly
successful however [416].
IVIG has been used in the treatment of
immunologically-mediated blood disorders in
both pediatric and adult populations. These
include autoimmune hemolytic anemia [417],
[418], autoimmune neutropenia [419], post-
transfusion purpura [420], ITP in adults [421],
[422], chronic ITP in adults and children
[423], [424], thrombocytopenia secondary to al-
loimmunization in leukemic patients receiving
platelet transfusions who became refractory to
subsequent platelet transfusions [425], thrombo-
cytopenia secondary to transplacental passage of
antiplatelet antibodies [426], ITP of pregnancy
[427], antibody-mediated red cell aplasia [428],
during pregnancy in women with severe rhesus
immunization [429], and in conjunction with cy-
clophosphamide used to treat antibody to factor
VIII in hemophilia [430].
In addition to these hematologic disorders,
two large trials conducted in patients with
Kawasaki’s disease have compared the use
of aspirin alone to that of aspirin and IVIG
(400 mg/kg/d for 4 and 3 days) [431], [432].
58 Immunotherapy and Vaccines
The IVIG reduced the fever and the incidence
of coronary artery disease [431], [432]. It sup-
pressed the T and B cell activation characteristic
of patients with this disease and decreased the
levels of spontaneous immunoglobulin synthe-
sis in vitro [433].
Placentally derived immunoglobulin has
been administered to patients with severe
rheumatoid arthritis because the symptoms of
some patients improved during pregnancy [434].
None of five patients given IVIG derived from
control plasma improved whereas 3/5 improved
under the placentally derived globulin. An an-
tihuman leukocyte antigen – DR surface antigen
(anti HLA-DR) antibody in the placental prepa-
ration was presumed to be a possible mechanism
for this improvement.
High dose immunoglobulin treatments have
also been used in patients with Felty’s syndrome
[435], myasthenia gravis [436], epilepsy [437],
and multiple sclerosis [438]. In the latter study
one-third of patients worsened with IVIG ther-
apy.
5.5.3.2. Mechanism of Action
Many immunologic mechanisms have been in-
voked to explain the beneficial effects of IVIG
in noninfectious illnesses. IVIG is a potent im-
mune modulator; the role of the Fc portion of
the immunoglobulin molecule in this respect is
still a subject of active investigation.
Antibody-coated material is removed by
phagocytes in a process known as Fc-receptor-
mediated clearance. The Fc portion of the an-
tibody coating the bacterium cell, protein etc.
binds to the Fc receptor of the phagocyte. The
material is then taken up by the phagocyte and
digested. Initially, the Fc receptors in the retic-
uloendothelial system were thought to become
saturated by the high dose of immunoglobulins
in IVIG resulting in a decreased clearance of
IgG-coated particles due to competitive inhibi-
tion [417], [439]. This is consistent with studies
in which radiolabelled, autologous erythrocytes
were cleared more slowly after IVIG infusion
than before [414], [423]; this treatment altered
Fc receptor affinity, not receptor number [440].
Other suggested mechanisms include decreased
synthesis of autoantibodies [441], the clearance
of persistent, occult viral infections [396], [441],
the protection of platelets or megakaryocytes
against antiplatelet antibodies [441], antiidio-
typic suppression of antibody synthesis [436],
[442], blocking of the Fc receptor by antibod-
ies, and the production of antilymphocyte an-
tibody [443]. However, patients with ITP have
responded to IVIG without demonstrable alter-
ation in Fc-receptor mediated clearance [441]. In
addition, preincubation of erythrocytes in IVIG
failed to inhibit the phagocytosis of sensitized
erythrocytes by cultured macrophages [444].
Thus, IVIG may lead to improvement in such
patients by multiple mechanisms.
IVIG is a potent modulator of antibody pro-
duction both in vitro and in vivo. It inhibits
lectin-driven B cell differentiation in vitro [445],
[446] and immunoglobulin production by pe-
ripheral blood mononuclear cells stimulated
with pokeweed mitogen [446]. For these ef-
fects the Fc portion of immunoglobulin must
be present: the Fc portion alone was 100 fold
more effective than the intact IgG. The observa-
tion that IgM antibody rises after IVIG infusion
has raised the possibility that it may stimulate
some immunoglobulin-producing cell popula-
tions [423]. Interestingly, the monthly admin-
istration of ISG to premature infants during the
first year of life resulted in a significantly lower
level of immunoglobulin compared to the con-
trol group [371], [447].
In adults with ITP, the infusion of IVIG
(400 mg/kg) led to a decrease in T 4 (helper)
lymphocytes and elevation in T 8 (suppressor)
lymphocytes (and decrease in T 4/T 8 ratio)
[421]. Similarly, the administration of IVIG to
patients with hypogammaglobulinemia led to in-
creased suppressor T cell activity, decreased to-
tal T cells, a significant decrease in the T 4/T 8
ratio, and a decrease in lectin-induced im-
munoglobulin syntheses in vitro [448].
The ability of IVIG to decrease antibody pro-
duction may be desirable in patients producing
autoantibody, but dangerous in those with infec-
tion [445]. In patients with acute otitis, repeated
monthly [379] infusions of IVIG resulted in
higher IgG levels if the initial antipneumococcal
antibodies were low. However, in the presence of
high initial levels of antibody, specific antipneu-
mococcal antibody levels decreased [449]. The
suggestion has also been made that high levels
of nonspecific antibody could lead to a reduc-
tion in the survival of specific antibody [450],

[451]. Evidence suggests that IVIG may even
exacerbate infections. A patient with neutrope-
nia died due to acceleration of the yeast infec-
tion shortly after infusion of IVIG; the IVIG may
have blocked the normal Fc-mediated clearance
mechanism [452].
Immunoglobulin G can bind both native C 3
and C 3 b during complement activation by sol-
uble immune complexes and by bacteria [453],
[454]. Finally since immunoglobulin has unique
antigenic determinants, antibodies to the im-
munoglobulin may develop [455], [456]. These
antibodies may play a role in the anti-idiotype
network.
In summary, exogenous immunoglobulin has
a diverse, potent effect on a wide variety of
immune regulatory mechanisms. These interac-
tions may often work to the patient’s benefit, but
may also result in previously unsuspected ad-
verse effects.
5.6. Prophylaxis and Therapy with
Plasma and Other Blood Products
Although the regular, large-scale, clinical use of
plasma and other blood products is considered
impractical (both for production and safety rea-
sons), passive immunotherapy with these agents
may provide some efficacy in the prophylaxis
and treatment of infections in humans.
Plasma has been used to treat infections, par-
ticularly those caused by Pseudomonas [374],
[457]. Plasma was superior to ISG in the pre-
vention of infection among patients who suf-
fered >30 % burns [457]. Similar observations
were made during experimental infection [458].
One possible explanation for the higher efficacy
of plasma is that it contains IgG, IgA, and IgM
(ISG contains only IgG). This may improve the
distribution of antibody at different anatomic
sites. Since the isotypes have functional differ-
ences, the antibacterial activity is also increased
(natural antibody against gram-negative bacilli
is thought to be predominantly of the IgM iso-
type). However, plasma infusions carry the risk
of transmitting hepatitis (there is no manufactur-
ing process for plasma that inactivates the virus
as is the case with ISG).
Further efforts in the use of passive im-
munotherapy have also focused on septic
Immunotherapy and Vaccines 59
shock. The rapid infusion of large volumes
of lyophilized human plasma from blood
with >40 µg/mL of antibody to a mixture of
lipopolysaccaride serotypes resulted in nearly
7-fold decrease in mortality in South African
women treated in an obstetrical–gynecological
ward for septic shock [459]. Equine plasma simi-
larly screened for antilipopolysaccaride antibod-
ies has also been used routinely in veterinary
practice in South Africa [460].
Opsonins. The defective neonatal antibody
(opsonic) response that mediates the uptake and
killing of bacteria by phagocytes can be par-
tially corrected with gamma globulin [461]. The
ability passively administered opsonins to pre-
vent infection was first suggested in one study
in which fresh whole blood was administered to
infants to prevent group B streptococcal sepsis.
All nine infants transfused with blood having
antibody to group B streptococci lived versus
3/6 transfused with blood having undetectable
antibody titers. Protection was correlated with
opsonic antibody titers in the infants’ blood and
with having >40 % of their blood volume re-
placed [462].
Postimmune Serum. Donors were immu-
nized with a J 5 mutant of E. coli to elicit
antibodies to widely shared core epitopes in
the lipopolysaccharides of gram-negative bacte-
ria. The passive administration of postimmune
serum from these individuals to patients in bac-
terial endotoxic shock decreased the incidence
of death as well as the need for pressor ther-
apy in patients with profound shock. However,
since this protection did not correlate with the
presence of antibody to the core epitopes, it is
not clear whether the protection was conferred
by the antibody [463]. In a follow-up study,
the prophylactic administration of postimmune
plasma appeared to significantly prevent shock
and death from gram-negative infections in pa-
tients admitted to a surgical intensive care unit.
Such treatment had no effect on infection rate,
however [464]. In both studies the protective
moiety in the plasma and serum was presumed
to be antibody to the endotoxin. In any event,
for this to be a practical therapy, the protec-
tive portion of the postimmune product must be
mass produced in the form of a safe, standard-
ized preparation (e.g., made into an IVIG from
60 Immunotherapy and Vaccines
a large pool of donors immunized with the J 5
vaccine or into a monoclonal antibody prepara-
tion). Attempts to show a protective effect from
an IVIG prepared from immunized donors have
been unsuccessful to date [465]. Nevertheless,
monoclonal antibody preparations directed to-
ward a J 5 epitope were evaluated in human vol-
unteers (see also Section 6.1).
5.7. Adverse Effects of Gamma Globulin
Preparations
Immune serum globulin G is one of the safest
biological products available with an overall in-
cidence of reactions of 3 – 12 % [334]. Rare sys-
temic (anaphylactic) reactions may occur during
or within minutes of administration of ISG (ca.
1/500 – 1/1000 injections). Late-occurring sys-
temic reactions (within hours or days) include
arthralgias, pyrexia, and diarrhea and are not un-
common in immunodeficient patients [364]. Lo-
cal reactions occur with intramuscular adminis-
tration of ISG and are related primarily to the
large volume administered. Local reactions to
ISG can be prevented by using small volumes or
by pretreatment with analgesics. Systemic re-
actions may be modified by the use of aspirin,
hydrocortisone, or antihistamines.
Systemic reactions with IVIG are more com-
mon than with ISG and often depend on the
rate of infusion and the type of preparation.
Patients receiving IVIG for hypogammaglobu-
linemia have a reported incidence of 2.5 % re-
actions which are usually mild [358]; the inci-
dence of these reactions decreases after the first
two months of therapy. The cause of reactions to
IVIG are not clear but may depend on aggregate
formation, IgA contamination, and activated en-
zymes that initiate the release of inflammatory
mediators [364].
The experimental work discussed above and
clinical experience give reason for caution in the
use of immunoglobulin preparations. Clinical
experience primarily involves concern about vi-
ral transmission, anaphylactic reactions among
patients with IgA deficiency, and isolated case
reports of unusual occurrences. The indications
for the use of gamma globulin should be well
founded [363].
Viral Transmission. Immune serum globu-
lin G has had an outstanding safety record with
regard to lack of transmission of virus to recip-
ients [466]. For example, ISG is free of serum
hepatitis, even if the virus occurs in the origi-
nal plasma. Ten children who received plasma
for prophylaxis against measles in England be-
came jaundiced and three died. In contrast, when
ISG prepared from the same plasma was given
to 56 children, only one child became jaun-
diced [343]. In the United States, 4 out of 15
volunteers who were inoculated subcutaneously
with hepatitis-positive plasma developed hepati-
tis. Five further volunteers received large doses
of ISG and none developed hepatitis [343]. The
ability of the Cohn – Oncley fractionation (see
Fig. 5) to inactivate virus may in part be due
to the high concentration of antivirus antibody
[353]. Non-A, non-B hepatitis has been reported
in pastes produced in this procedure: in one, the
lyophilization step was omitted, and in two oth-
ers an ion-exchange step was added to lower IgA
but it lowered IgG 4 levels as well [353].
In contrast, non-A, non-B (NANB) hepati-
tis has been reported following IVIG infusion
[467–469]. In one such episode 16 out of 77
patients who received IVIG for immunodefi-
ciency acquired hepatitis with death occurring
in 5 cases. Interestingly, no hepatitis was ob-
served following administration of ISG prepared
from the same serum pool [467]. Since neither
of the IVIG products was subjected to a recog-
nized virucidal finishing step, such as treatment
with acid, β-propiolactone, or ultraviolet radia-
tion, the virus may still have been viable. Thus
ethanol fractionation by itself might not fully
inactivate putative virus [470], though it is viru-
cidal for enveloped viruses [471]. Supplemen-
tary virucidal procedure could make IVIG safe
from NANB hepatitis transmission [471]. Un-
like plasma, IVIG has not been known to trans-
mit hepatitis B [472].
There have been no confirmed cases of trans-
mission of HIV virus in IVIG but antibody to
HIV has been isolated from two patients with
primary hypogammaglobulinemia [471]. When
HIV was added to plasma and then processed
into IVIG, >104 PFU/mL (PFU denotes plaque-
forming units) of HIV were inactivated dur-
ing alcohol fractionation and poly-(ethylene gly-
col) fractionation to IVIG [473]. Antibody to
HIV was detected by an enzyme-linked im-

munoabsorbent assay (ELISA) in all 10 lots
of a reduced, alkylated IVIG and in 4/8 lots
of a pH 4/pepsin IVIG; 8/10 and 3/8 lots were
also positive by Western blot analysis. The HIV
antibody-positive lots, which were negative for
HIV by culture and reverse transcriptase activity,
were infused into patients. Comparison of pre-
and post-infusion sera demonstrated that anti-
body to HIV was detectable for up to one month
before converting to a negative antibody status
[472].
One woman who had received hyperim-
mune Rho (D) immunoglobulin subsequently
was found to be infected with HIV but belonged
to a group with an increased risk of acquiring
infection with HIV [474].
Anaphylaxis. Anaphylaxis has been docu-
mented following the administration of IVIG
to some patients with hypogammaglobulinemia.
Indeed, patients with antibody deficiency have
an increased incidence of adverse reactions to
IVIG [342]. Low levels of IgA are associated
with an increased risk of such anaphylactic re-
actions [475]; IVIG products differ markedly in
their IgA content [476]. In two patients with
common variable immunodeficiency, IgE anti-
body to IgA in the infused IVIG was felt to ac-
count for these reactions [475]. In addition, IgE
has been detected in commercial preparations of
IVIG [477].
Other clinical problems have been de-
scribed which involve the use of IVIG. IVIG
prophylaxis precipitated cryoglobulinemic
nephropathy in a patient with hypogammaglob-
ulinemia secondary to B cell neoplasm [478].
The passive transfer of antibodies against
specific blood types in IVIG may also be associ-
ated with hemolytic anemia or cause problems
in the crossmatching of blood prior to surgery
[479]. Immunologically normal patients may re-
ceive antibody to immunoglobulin that is of a
different genotype.
5.8. Future Prospects
Large amounts of immunoglobulin can now be
safely delivered directly into the bloodstream
by using IVIG. This need was recognized by
Cohn in the 1940s. Since immunoglobulin is
Immunotherapy and Vaccines 61
now known to be a potent modulator of the im-
mune system, IVIG has been used as a form of
therapy for many noninfectious diseases. How-
ever, the impact of such treatment on the abil-
ity to respond to infectious diseases often com-
plicates these conditions and still has to be as-
sessed. Studies to date do not clearly show that
the indications for the use of IVIG in infectious
diseases should be expanded beyond those al-
ready shown for ISG. In some conditions involv-
ing functional hypogammaglobulinemia (e.g.,
HIV infection in young children, chronic lym-
phocytic leukemia), prophylaxis with standard
IVIG may ultimately demonstrate efficacy. In
certain subpopulations that are at risk of infec-
tion with specific organisms over a finite period
of time, immunoglobulin preparations enriched
in antibody to those organisms (i.e., hyperim-
mune preparations) may prove useful. The most
promising use appears to be in the prophylaxis
of specific infections; this usually involves hy-
perimmune serum and, as shown by studies in
neonates, the substantial replacement of blood
volume.
The use of IVIG to treat established infections
with pharmacological (not replacement) doses
of immunoglobulin, (with the possible excep-
tion of antitoxin treatment of septic shock), still
has no proven role. This may be related to the
need for prompt intervention and a large amount
of a specific antibody, requiring rapid identifi-
cation of the microorganisms involved. Prepa-
rations that are known to be immunotherapeu-
tically effective are less efficacious when given
8 h or more after the onset of infection. Substan-
tial evidence suggests that passive immunother-
apy may provide an enhanced therapeutic bene-
fit; however, this still remains to be proved in a
prospective, controlled therapeutic trial.
Finally, more recent studies that show some
benefit from passive immunotherapy have not
been performed with IVIG preparations, e.g.,
[376], [463], [464], [459]. Thus despite the sub-
stantial progress made in the development of
IVIG preparations for human use, their firm rec-
ommendation for widespread use in infectious
disease requires further well-designed studies.
62 Immunotherapy and Vaccines

6. Immunotherapeutic Uses of 3) Viral antigens: hepatitis A and B, rabies,

Monoclonal Antibodies CMV, HSV, and varicella zoster
4) Bacterial antigens: tetanus, endotoxins, and
When a vertebrate is vaccinated with a foreign pneumococcus
antigen, it produces a mixture of antibodies that 5) Elimination of circulating drugs (overdoses)
can bind to the antigen. Each antibody binds to a 6) Antisnake venom
different epitope on the immunogen and is made 7) Fertility control (e.g., anti-β-human chori-
by an individual clone of B lymphocytes. Indus- onic gonadotropin)
trial quantities of a single (i.e., monoclonal) anti-
Humabs that recognize many of these anti-
body recognizing a specific epitope can be made
gens have been produced, and will probably
by fusing a B lymphocyte producing the anti-
augment or replace pooled antisera in the near
body of interest with an immortal tumor cell line
future. Concerns, which include the possible
to give hybridomas. For a detailed description of
contamination of pooled globulins by infec-
monoclonal antibodies, see → Monoclonal An-
tious agents (e.g., human immunodeficiency
tibodies.
virus HIV, various hepatitis viruses, and cy-
Infectious diseases continue to cause much
tomegalovirus) and the diminished availability
morbidity and mortality despite the discovery of
of serum donors, as well as the relative ease of
a wide variety of novel antimicrobial agents and
reproducible humab manufacture, will acceler-
improved methods for immunizing at-risk pop-
ate this trend.
ulations. Therapy with human or animal serum
Humabs will be preferred for the diagnosis
is an effective adjunct to conventional drug ther-
and treatment of a variety of diseases because
apy and, prior to the development of antimicro-
they minimize the problems encountered when
bials, was the only effective therapy for many
a foreign animal monoclonal antibody is admin-
infectious diseases (see Chap. 5). The efficacy
istered (e.g., anaphylaxis, clinical manifesta-
of serum treatment for various infectious dis-
tions of immune complex formation, and reduc-
eases suggests that this approach could be a use-
tion of efficacy by anti-antibodies). In well over
ful adjunct to drug therapy if safety, efficacy, and
half of the patients treated to date with murine
production problems could be overcome. Hu-
monoclonal antibodies, the human antimouse
man monoclonal antibodies (humabs) may solve
immunoglobulin response has limited their use-
many of the problems associated with serum
fulness [480]. Only a fraction of the antimouse
therapy. Therapeutic humabs have been devel-
immune response is directed to the variable re-
oped against many target organisms and are be-
gion (idiotype) of the rodent immunoglobulins.
ing tested in clinical trials.
This suggests that humabs will be more effective
than murine monoclonal antibodies. Further-
more, humabs are more likely to have species-
6.1. Introduction specific carbohydrates which may be impor-
tant in a number of effector functions, such as
Despite the fact that the first reports describing
Fc receptor-mediated, antibody-dependent, cel-
the generation of antigen-specific murine mono-
lular cytotoxicity, complement activation, and
clonal antibodies were published in the mid
phagocytosis [481]. Serum half-life and effec-
1970s, routine production of humabs has been
tor functions of immunoglobulin subclasses are
more difficult. Methods to generate and produce
very important for designing the optimal anti-
these molecules have improved and serious ef-
infectious agent therapeutic. There are a number
forts are underway to develop them as therapeu-
of reasons why passively administered humabs
tic products for which passively administered
will be preferred for therapeutic use. A good
human antisera against specific target antigens
summary and technical background are to be
are in use in a number of clinical settings:
found in [482], [483].
1) Red cell antigens: Rh (hemolytic disease of
the newborn)
2) White cell antigens:
antilymphocyte/thymocyte globulin

6.2. Bacterial Targets
Various bacterial targets are recognized by hum-
abs:
Tetanus toxoid [484–488]
Diphtheria toxoid [489]
Gram-negative endotoxins [491–494]
Pseudomonas aeruginosa lipopolysaccha-
ride [496] and exotoxin A [497]
Hemophilus influenzae [498], [499]
Mycobacterium leprae [500]
Neisseria meningitides [501]
Pneumococcus [502], [503]
Tetanus-neutralizing humabs have been fre-
quently generated because of the ease of ob-
taining immune B cells from vaccinated persons
[484–488].
Because immunization against tetanus in the
United States is universal and most individuals
have relatively high titers, it is unlikely that any
of these monoclonals will be scaled up and used
in this country. The same is true of antidiphtheria
monoclonal antibodies [489]. In other countries,
vaccination against tetanus and diphtheria is not
widespread and administration of immune hu-
man or animal sera is still practiced. Concern
about HIV-contaminated serum has stimulated
efforts to generate humabs against tetanus for
use in India, South America, and southern Eu-
rope.
Gram-negative bacterial infections account
for 1 – 2 % of hospital admissions and up to
100 000 deaths each year in the USA. Despite
therapies including antibiotics and various sup-
port measures, mortality rates remain as high as
50 – 70 %. Much interest has been generated by
the idea that lipopolysaccharide, the lethal com-
ponent of gram-negative bacteria, might have
antigenic determinants that are shared by many
bacterial species. The outer carbohydrate do-
main of lipopolysaccharide varies highly from
one species to the next whereas the inner car-
bohydrate core and lipid A regions appear to
be much more conserved. In a clinical trial an-
tisera were obtained from firemen vaccinated
with a rough mutant E. coli strain called J 5
that lacks the outer carbohydrate. This anti-
serum gave significant protection to patients
with gram-negative sepsis [490]. Several labo-
ratories have reported the generation of hum-
abs that are highly cross-reactive with bacterial
Immunotherapy and Vaccines 63
lipopolysaccharide [491–494]. They are all of
the IgM class and protect mice against infec-
tion when given at high doses. Two of these
and a murine monoclonal of the same speci-
ficity are in clinical trials for the treatment of
gram-negative sepsis. At present it is too early
to determine whether they are successful, how-
ever several points should be made regarding this
approach:
Firstly, it is unclear whether antibody was
the active factor in the antisera in the original
study [490]. Secondly, no documentation on the
binding of the antisera to the bacteria causing
a specific case of sepsis was made, i.e., no cor-
relation could be made between response and
nonresponse. Thirdly, immune serum is known
to contain many other factors that might have
benefited the treated patients. Fourthly, although
cross-reactive antisera and monoclonal antibod-
ies appear to bind to a wide number of bac-
teria and blotted lipopolysaccharide, questions
still exist regarding the accessibility of the core
and lipid A region for antibody binding. Fifthly,
several trials with slightly different design have
given inconsistent results [495]. Thus for a num-
ber of reasons the so-called core concept is ques-
tionable.
Type-specific horse antibodies recognizing
typhoid and pneumococcus immunotypes were
used from the turn of the century up to the begin-
ning of the antibiotic era in the 1940s. The gener-
ation of humabs recognizing type-specific deter-
minants on gram-negative bacteria is a reason-
able alternative to the above-mentioned highly
cross-reactive monoclonals. Although the diver-
sity of possible immunotypes of gram-negative
bacteria is large, a particular subset is proba-
bly responsible for most of the invasive bac-
teremias. Several groups are trying to generate
a number of type-specific humabs and adminis-
ter a cocktail of antibodies. This approach has
already been tried for Pseudomonas aeruginosa
[496]. In model systems, murine monoclonal an-
tibodies recognizing a single immunotype were
much more effective than those recognizing a
core determinant that was shared between Pseu-
domonas species. If the cocktail approach is to
prove effective, it may be necessary to add hum-
abs that neutralize various virulence factors. In
the case of Pseudomonas infections, humabs
have also been generated to exotoxin A [497].
64 Immunotherapy and Vaccines
Humabs have been reported that recognize
Hemophilus influenza type B capsular polysac-
charide [498]. The successful introduction of
an H. influenza type B capsular polysaccharide
vaccine [499] (see Section 2.13) has made the
clinical use of these humabs less attractive than
originally anticipated. However, a safe humab
for this life-threatening disease may still be de-
veloped for clinical use in conjunction with the
vaccines.
Anti-mycobacterium leprae humabs have
been reported [500]. In view of the central role of
cell-mediated immunity in resistance and recov-
ery from this disease, these reagents are unlikely
to find clinical application. Nevertheless, these
antibodies and those made by patients suffering
from other infectious diseases can be used to
probe the human humoral immune response, to
clone antigens recognized by the humabs, and
to investigate antibody-mediated autoimmunity
initiated by microorganisms.
6.3. Viral and Chlamydial Targets
Antiviral humabs are especially attractive ther-
apeutic targets. Immunoprophylaxis of cy-
tomegalovirus (CMV) infections in immuno-
suppressed patients (e.g., transplantation) is of
major clinical interest. Several groups have gen-
erated CMV-neutralizing humabs [504], [505]
and clinical trials were scheduled to begin in
1989.
Humabs that neutralize hepatitis B virus
[506–508] will replace the antisera currently
used after acute exposure to this agent. These
humabs may also be useful for “active – passive”
immunization of at-risk infants. A very large
population of infants, particularly in less devel-
oped countries, become chronically infected in
the perinatal period. Evidence suggests [509]
that administration of immune sera will prevent
lifelong infection and could have an impact on
the prevalence of the most common human can-
cer in the world, hepatitis-B-positive hepatoma
[509]. The widespread use of anti-hepatitis A
antisera suggests that the humabs generated to
this target may also find clinical application.
A minor yet important target for humabs is
the treatment of varicella zoster infections [510].
These humabs may have to be developed as an
orphan drug given the small numbers of patients.
(The development of drugs for rare diseases is
supported by the US government; these pharma-
ceuticals are called orphan drugs.)
The acquired immune deficiency syndrome
( AIDS) is caused by the human immunodefi-
ciency viruses (HIV) and is currently the major
infectious viral disease problem. Many groups
are generating humabs that recognize one or
more HIV strains. Many of these humabs rec-
ognize gene product gp 160 that is found in the
viral envelope [511], [512]. At present, there is
no evidence that a humoral immune response can
prevent or change the pattern of HIV infection
in people or animals. The viruses undergo rapid
mutation (five times the rate of influenza). Fur-
thermore, they can spread by cell to cell contact
and reside permanently out of reach of humoral
immunity inside the mononuclear phagocytes
scattered throughout the body. Trials are under-
way to test the efficacy of a high-titer anti-HIV
antiserum in the prevention of neonatal AIDS.
The dismal failure of all attempted vaccines and
passive serum to delay the spread of the virus
in subhuman primate models is very discourag-
ing. Antiviral humabs attached to toxins or engi-
neered to bring cytotoxic cells into contact with
infected cells might improve the potency of the
antibody approach.
Several humabs have been generated against
chlamydia [513] and other viruses:
Epstein Barr Virus [514]
Herpes simplex [515–517]
Human T cell leukemia virus I [518]
Measles (SSPE) virus [519], [520]
Rabies [521]
Rubella [522]
X 31 influenza virus [523]
However, clinical interest in the therapeutic
purposes and use of these antibodies is limited.
6.4. Parasite Targets
The major therapeutic target of interest among
parasitic diseases is malaria (see Section 4.2).
Much less is known about the humoral immune
response of humans suffering from other par-
asitic diseases (Section 4.1). Passively admin-
istered humabs may have therapeutic potential
in acute Falciparum malaria. Immunoglobulin

M and IgG humabs have been generated [524],
[525]. Monoclonal antibodies are also useful re-
search tools in the development of malaria vac-
cines (Chap. 4).
7. References
1. I. Roitt: Essential Immunology, 5th ed.,
Blackwell Scientific Publications, Oxford,
United Kingdom 1984.
2. B. D. Davis, R. Dubecco, H. N. Eisen, H. S.
Ginsberg: Microbiology: Including
Immunology and Molecular Genetics, 3rd ed.,
Harper and Row, Hagerstown, MD, 1980.
3. R. H. Schwartz, A. Yano, W. E. Paul:
“Interaction between antigen-presenting cells
and primed T-lymphocytes,” Immunol. Rev. 40
(1978) 153.
4. W. Stroler, L. A. Hanson, K. W. Sell (eds.):
Recent Advances in Mucosal Immunity, Raven
Press, New York 1982.
5. T. Yamamura, T. Tada: Progress in
Immunology V, Academic Press, Tokyo 1984.
6. J. B. Robbins, J. C. Hill, J. C. Sadoff (eds.):
Bacterial Vaccines, Seminars in Infectious
Disease, vol. IV, Thieme-Stratton, New York
1982.
7. U. E. Nydegger (ed.): Immunotherapy, A
Guide to Immunoglobulin Prophylaxis and
Therapy, Academic Press, London 1981.
8. R. E. Black, M. M. Levine, M. L. Clements, G.
Losonsky et al.: “Prevention of shigellosis by a
Salmonella typhi – Shigella sonnei bivalent
vaccine,” J. Infect. Dis. 155 (1987)
1260 – 1265.
9. M. M. Levine, J. B. Kaper, D. Herrington, J.
Ketley et al.: “Safety, immunogenicity, and
efficacy of recombinant live oral cholera
vaccines, CVD 103 and CVD 103-HgR,”
Lancet 2 (1988) 467 – 470.
10. R. J. Collier: “Diphtheria toxin: mode of action
and structure,” Bacteriol. Rev. 39 (1975)
54 – 85.
11. A. M. Pappenheimer, Jr. in R. Germanier (ed.):
“Diphtheria,” Bacterial Vaccines, Academic
Press, Orlando, FL 1984, pp. 1 – 36.
12. F. Audibert et al.: “Active antitoxic
immunization by a diphtheria toxin synthetic
oligopeptide,” Nature (London) 289 (1981)
593 – 594.
13. T. Uchida, D. M. Gill, A. M. Pappenheimer,
Jr.: “Mutation in the structural gene for
diphtheria toxin carried by temperate phage
β,” Nature (London) 283 (1971) 8 – 11.
Immunotherapy and Vaccines 65
14. L. Greenfield, H. L. Dovey, F. C. Lawyer, D. H.
Gelfand: “High level expression of diphtheria
toxin peptides in Escherichia coli,”
Bio/Technology 4 (1986) 1006 – 1011.
15. B. Bizzinin in R. Germanier (ed.): “Tetanus,”
Bacterial Vaccines, Academic Press, Orlando,
FL 1984, pp. 37 – 68.
16. M. C. Hardegree et al.: “Titration of tetanus
toxoids in international units; relationship to
antitoxin responses of rhesus monkeys,” 6th
Proc. Int. Conf. Tetanus 1981, 1982,
pp. 409 – 423.
17. M. Huet: “La standardisation des vaccins
tetaniques,” 6th Proc. Int. Conf. Tetanus 1981,
1982 pp. 425 – 433.
18. K. W. Newell, A. Dueñas-Lehmann, D. R.
Leblanc, N. Garces-Osorio: “The use of toxoid
for the prevention of tetanus neonatorum; final
report of a double-blind controlled field trial,”
Bull. W.H.O. 35 (1966) 863 – 871.
19. N. Guerin, C. Fillastre: “Vaccin D.T.C. Polan,”
6th Proc. Int. Conf. Tetanus 1981, 1982,
pp. 477 – 479.
20. M. Pittman: “Pertussis toxin; the cause of the
harmful effects and prolonged immunity of
whooping cough,“ Rev. Infect. Dis. 1 (1979)
401 – 412.
21. C. R. Manclark, J. L. Cowell in R. Germanier
(ed.): “Pertussis,” Bacterial Vaccines,
Academic Press, Orlando, FL 1984,
pp. 69 – 106.
22. T. Madsen: “Vaccination against whooping
cough,” JAMA J. Am. Med. Assoc. 101 (1933)
187 – 188.
23. C. L. Cody et al.: “Nature and rates of adverse
reactions associated with DTP and DT
immunizations in infants and children,”
Pediatrics 68 (1981) 650 – 660.
24. R. Aldersdale et al.: Whooping Cough. Reports
from the Committee on Safety of Medicines
and the Joint Committee on Vaccination and
Immunization, vol. 4, H. M. Stationary Office,
London, England 1981, pp. 79 – 169.
25. M. Oda, J. L. Cowell, D. G. Burstyn, R.
Manclark: “Protective activities of the
filamentous hemagglutinin and the
lymphocytosis-promoting factor of Bordetella
pertussis in mice,” J. Infect. Dis. 150 (1984)
823 – 833.
26. Ad Hoc Group for the study of Pertussis
Vaccines: “Placebo controlled trial of two
acellular pertussis vaccines in
Sweden–protective efficacy and adresse
events.” Lancet 1 (1988) 955 – 960.
66 Immunotherapy and Vaccines
27. C. S. Chuttani et al.: “Controlled field trial of a
high-dose oral killed typhoid vaccine in India,”
Bull. W.H.O. 55 (1977) 643 – 644.
28. M. Reitman: “Infectivity and antigenicity of
streptomycin-dependent Salmonella typhosa,”
J. Infect. Dis. 117 (1967) 101 – 107.
29. R. Germanier, E. Fürer: “Isolation and
characterization of gal ε mutant Ty 21 a of
Salmonella typhi; a candidate strain for a live,
oral typhoid vaccine,” J. Infect. Dis. 131
(1975) 553 – 558.
30. R. H. Gilman, R. B. Hornick, W. E. Woodward,
H. L. Dupont et al.: “Evaluation of a
UDP-glucose-4-epimeraseless mutant of
Salmonella typhi as a live oral vaccine,” J.
Infect. Dis. 136 (1977) 717 – 723.
31. M. H. Wahdan, C. Sérié, Y. Cerisier, S. Sallam
et al.: “A controlled field trial of live
Salmonella typhi Ty 21 a oral vaccine against
typhoid; three year results,” J. Infect. Dis. 145
(1982) 292 – 295.
32. World Health Organization Committee on
Biological Standardization: Requirement for
typhoid vaccine (live-attenuated, Ty21 a, oral).
World Health Organization Technical Report,
series 700; Geneva, Switzerland: 34th report,
1984, pp. 48 – 68.
33. R. Germanier in R. Germanier (ed.): “Typhoid
fever,” Bacterial Vaccines, Academic Press,
Orlando, FL 1984, pp. 137 – 165.
34. M. M. Levine, C. Ferreccio, R. E. Black, R.
Germanier, Chilean Typhoid Committee:
“Large-scale field trial of Ty21 a live oral
typhoid vaccine in enteric-coated capsule
formulation,” Lancet ii (1987) 1049 – 1052.
35. M. M. Levine, D. Herrington, J. R. Murphy,
J. G. Morris et al.: Safety, infectivity,
immunogenicity, and in vivo stability of two
auxotrophic mutant strains of Samonella typhi,
541 Ty and 543 Ty, as live oral vaccine in
humans, J. Clin. Invest. 79 (1987) 888 – 902.
36. I. L. Acharya, C. V. Lowe, R. Thapa, V. L.
Gurubacharya et al.: “Prevention of typhoid
fever in Nepal with the Vi capsular
polysaccharide of Salmonella typhi,” N. Engl.
J. Med. 317 (1987) 1101 –1104.
37. R. Austrian: “Of gold and pneumococci; a
history of pneumococcal vaccines in South
Africa,” Trans. Am. Clin. Climatol. Assoc. 89
(1977) 141 – 161.
38. C. M. MacLeod, M. R. Krauss: “Stepwise
intratype transformation of pneumococcus
from R to S by way of a variant intermediate in
capsular polysaccharide production,” J. Exp.
Med. 86 (1947) 439 – 453.
39. J. B. Robbins, R. Austrian, C.-J. Lee, S. C.
Rastogi et al.: “Considerations for formulating
the second-generation pneumococcal capsular
polysaccharide vaccine with emphasis on the
cross-reactive types,” J. Infect. Dis. 148
(1983) 1136 – 1159.
40. Health and Public Policy Committee,
American College of Physicians:
“Pneumococcal vaccine,” Ann. Intern. Med.
104 (1986) 118 – 120.
41. P. Smit, D. Oberholzer, S. Hayden-Smith, H. J.
Koornhof et al.: “Protective efficacy of
pneumonococcal polysaccharide vaccines,”
JAMA J. Am. Med. Assoc. 238 (1977)
2613 – 2616.
42. G. Bolan, C. V. Broome, R. R. Facklam, B. D.
Plikaytis et al.: “Pneumococcal vaccine
efficacy in selected populations in the United
States,” Ann. Intern. Med. 104 (1986) 1 – 6.
43. M. S. Simberkoff, A. P. Cross, M. Al-Ibrahim,
A. L. Baltch et al.: “Efficacy of pneumococcal
vaccine in high-risk patients,” N. Engl. J. Med.
315 (1986) 1318 – 1327.
44. J. A. Leech, A. Gervais, F. L. Ruben: “Efficacy
of pneumococcal vaccine in severe chronic
obstructive pulmonary disease,” Can. Med.
Assoc. J. 136 (1987) 361 – 365.
45. C. Chu, R. Schneerson, J. B. Robbins, S. C.
Rastogi: “Further studies on the
immunogenicity of Haemophilus influenzae
type b and pneumococcal type 6 A
polysaccharide-protein conjugates,” Infect.
Immun. 40 (1983) 245 – 256.
46. R. Schneerson, J. R. Robbins, C. Chu, A.
Sutton et al.: “Serum antibody response of
juvenile and infant rhesus monkeys injected
with Haemophilus influenzae type b and
pneumococcus type 6 A capsular
polysaccharide-protein conjugates,” Infect.
Immun. 45 (1984) 582 – 591.
47. R. Schneerson, J. B. Robbins, J. C. Parke, Jr.,
C. Bell et al.: “Quantitative and qualitative
analyses of serum antibodies elicited in adults
by Haemophilus influenzae type b and
pneumococcus type 6 A capsular
polysaccharide-tetanus toxoid conjugates,”
Infect. Immun. 52 (1986) 519 – 528.
48. A. D. O’Brien, M. R. Thompson, P. Gemski,
B. P. Doctor et al.: “Biological properties of
Shigella flexneri 2 A toxin and its serological
relationship to Shigella dysenteriae 1 toxin,”
Infect. Immun. 15 (1977) 796 – 798.
49. A. R. Higgins, T. M. Floyd, M. A. Kader:
“Studies in shigellosis. III. A controlled

evaluation of a monovalent Shigella vaccine in
a highly endemic environment,” Am. J. Trop.
Med. Hyg. 4 (1955) 281 – 288.
50. D. M. Mel, E. J. Gangarosa, M. L.
Radovanović, B. L. Arsić et al.: “Studies on
vaccination against bacillary dysentery. 6.
Protection of children by oral immunization
with streptomycin-dependent Shigella strains,”
Bull. W.H.O. 45 (1971) 457 – 464.
51. M. M. Levine, P. A. Rice, E. J. Gangarosa,
G. K. Morris et al.: “An outbreak of Sonnei
shigellosis in a population receiving oral
attenuated shigella vaccines,” Am. J.
Epidemiol. 99 (1973) 30 – 36.
52. S. B. Formal, M. M. Levine in R. Germanier
(ed.): “Shigellosis,” Bacterial Vaccines,
Academic Press, Orlando, FL 1984,
pp. 167 – 186.
53. S. B. Formal, L. S. Baron, D. J. Kopecko, O.
Washington et al.: “Construction of a potential
bivalent vaccine strain: introduction of
Shigella sonnei Form I antigen genes into the
gal ε Salmonella typhi Ty 21 a typhoid vaccine
strain,” Infect. Immun. 34 (1981) 746 – 750.
54. R. E. Black, M. M. Levine, M. L. Clements, G.
Losonsky et al.: “Prevention of shigellosis by a
Salmonella typhi – Shigella sonnei bivalent
vaccine,” J. Infect. Dis. 155 (1987)
1260 – 1265.
55. L. S. Baron, D. J. Kopecko, S. B. Formal, R.
Seid et al.: “Introduction of Shigella flexneri
2 a type and group antigen genes into oral
typhoid vaccine strain Salmonella typhi
Ty 21 a,” Infect. Immun. 55 (1987)
2797 – 2801.
56. S. Sturm, K. Timmis: “Cloning of the rfb gene
region of Shigella dysenteriae 1 and
construction of an rfb-rfp gene cassette for the
development of lipopolysaccharide-based live
anti-dysentery vaccines,” Microb. Pathog. 1
(1986) 289 – 297.
57. R. A. Finkelstein: “Immunology of cholera,”
Immunol. 69 (1975) 137 – 195.
58. R. A. Finkelstein: “Cholera,” CRC Crit. Rev.
Microbiol. 2 (1973) 553 – 623.
59. R. A. Cash, S. I. Music, J. P. Libonati, J. P.
Craig et al.: “Response of man to infection
with Vibrio cholerae. II. Protection from
illness afforded by previous disease and
vaccine,” J. Infect. Dis. 130 (1974) 325 – 333.
60. R. A. Cash, S. I. Music, J. P. Libonati, A. R.
Schwartz et al.: “Live oral cholera vaccine;
evaluation of the clinical effectiveness of two
strains in humans,” Infect. Immun. 10 (1974)
762 – 764.
Immunotherapy and Vaccines 67
61. W. E. Woodward, R. H. Gilman, R. B. Hornick,
J. P. Libonati et al.: “Efficacy of a live oral
cholera vaccine in human volunteers,” Dev.
Biol. Stand. 33 (1976) 108 – 112.
62. M. M. Levine, R. E. Black, M. L. Clements, C.
Lanata et al.: “Evaluation in humans of
attenuated Vibrio cholerae El Tor Ogawa strain
Texas Star-SR as a live oral vaccine, Infect.
Immun. 43 (1984) 515 – 522.
63. J. B. Kaper, M. M. Levine: “Cloned cholera
enterotoxin genes in study and prevention of
cholera,” Lancet ii (1981) 1162 – 1163.
64. J. J. Mekalanos, D. J. Suartz, G. D. N. Pearson,
N. Harford et al.: “Cholera toxin genes:
nucleotide sequence, deletion analysis, and
vaccine development,” Nature (London) 306
(1983) 551 – 557.
65. A. D. O’Brien, M. E. Chen, R. K. Holmes et
al.: Environmental and human isolates of
Vibrio cholerae and Vibrio parahaemolyticus
produce a Shigella dysenteriae 1 (Shiga)-like
cytotoxin,” Lancet i (1984) 77 – 78.
66. J. D. Clements, J. R. Harris, M. R. Khan, B. A.
Kay et al.: “Field trial of oral cholera vaccine
in Bangladesh,” Lancet ii (1986) 124 – 127.
67. J. E. McGowan, M. W. Barnes, M. Finland:
“Bacteremia at Boston City Hospital;
occurrence and mortality during 12 selected
years (1935 – 1972), with special reference to
hospital-acquired cases,” J. Infect. Dis. 132
(1975) 316 – 335.
68. C. S. Bryan, K. L. Reynolds, E. R. Brenner:
“Analysis of 1186 episodes of gram-negative
bacteremia in non-university hospitals; the
effects of antimicrobial therapy,” Rev. Infect.
Dis. 5 (1983) 629 – 638.
69. R. W. Haley: Preliminary cost-benefit analysis
of hospital infection control programs (The
SENIC Project), Proc. Internat. Workshop,
Baiersbromm, Germany (1977) pp. 93 – 95.
70. S. J. Cryz, Jr.: “Prospects for the prevention
and control of gram-negative nosocomial
infections,” Vaccine 5 (1987) 261 – 265.
71. W. R. Jarvis, J. W. White, V. P. Munn, J. L.
Mosser et al.: “Nosocomial infection
surveillance,” 1983. Centers for Disease
Control Surveillance Summaries 33, no. 299,
pp. 955 – 2155 (1985).
72. M. S. Pollack, L. S. Young: “Protective activity
of antibodies to exotoxin A and
lipopolysaccharide at the onset of
Pseudomonas aeruginosa septicemia in man,”
J. Clin. Invest. 63 (1979) 276 – 286.
68 Immunotherapy and Vaccines
73. S. J. Cryz, Jr. in R. Germanier (ed.):
“Pseudomonas aeruginosa infections,”
Bacterial Vaccines, Academic Press, Orlando,
FL 1984, pp. 317 – 351.
74. G. B. Pier, S. E. Bennett: “Structural analysis
and immunogenicity of Pseudomonas
aeruginosa immunotype 2 high molecular
weight polysaccharide,” J. Clin. Invest. 77
(1986) 491 – 495.
75. S. J. Cryz, Jr., E. Fürer, A. S. Cross, A.
Wegmann et al.: “Safety and immunogenicity
of a Pseudomonas aeruginosa
O-polysaccharide-toxin A conjugate vaccine
in humans,” J. Clin. Invest. 80 (1987) 51 – 56.
76. S. J. Cryz, Jr., A. S. Cross, E. Fürer, N.
Chariatte, et al.: “Activity of intravenous
immune globulin against Klebsiella,” J. Lab.
Clin. Med. 108 (1986) 182 – 189.
77. W. D. Welch, W. J. Martin, P. Stevens, L. S.
Young: “Relative opsonic and protective
activities of antibodies against Kl, O and lipid
A antigens of Escherichia coli,” Scand. J.
Infect. Dis. 11 (1979) 291 – 298.
78. B. Kaijser, S. Ahlstedt: “Protective capacity of
antibodies to Escherichia coli O and K
antigens,” Infect. Immun. 17 (1977) 286 – 289.
79. A. S. Cross, P. Gemski, J. C. Sadoff, F. Ørskov
et al.: “The importance of the Kl capsule in
invasive infections caused by Escherichia
coli,” J. Infect. Dis. 149 (1984) 184 – 193.
80. M. Granström, I. Julander, R. Möllby:
“Serological diagnosis of deep Staphylococcus
aureus infections by enzyme-linked
immunosorbent assay (ELISA) for
staphylococcal hemolysin and teichoic acid,”
Scand. J. Infect. Dis. Suppl. 41 (1983)
132 – 139.
81. W. W. Karakawa, J. M. Fournier, W. F. Vann,
R. Arbeit et al.: “Method for the serological
typing of the capsular polysaccharides of
Staphylococcus aureus,” J. Clin. Microbiol. 22
(1985) 445 – 447.
82. H. K. Hochkeppel, D. G. Braun, W. Visher, A.
Imm et al.: “Serotyping and electron
microscopy studies of Staphylococcus aureus
clinical isolates with monoclonal antibodies to
capsular polysaccharide types 5 and 8,” J.
Clin. Microbiol. 25 (1987) 526 – 530.
83. R. D. Arbeit, J. M. Nelles: “Capsular
polysaccharide antigenemia in rats with
experimental endocarditis due to
Staphylococcus aureus,” J. Infect. Dis. 155
(1987) 242 – 246.
84. L. M. Mutharia, G. Crockford, W. C. Bogard,
Jr., R. E. W. Hancock: “Monoclonal antibodies
specific for Escherichia coli J 5
lipolysaccharide: cross-reaction with other
gram-negative bacterial species,” Infect.
Immun. 45 (1984) 631 – 636.
85. E. J. Ziegler, J. A. McCutchan, H. Douglas,
A. I. Braude: “Treatment of E. coli and
Klebsiella bacteremia in agranulocytopenic
animals with antiserum to a UDP-GAL
epimerase-deficient mutant,” J. Immunol. 111
(1973) 433 – 438.
86. E. J. Ziegler, J. A. McCutchan, J. Fierer, M. P.
Glauser et al.: “Treatment of gramnegative
bacteremia and shock with human antiserum
to a mutant of Escherichia coli,” N. Engl. J.
Med. 307 (1982) 1225 – 1230.
87. J. D. Baumgartner, M. P. Glauser, J. A.
McCutchan, E. J. Ziegler et al.: “Prevention of
gramnegative shock and death in surgical
patients by antibody to endotoxin core
glycolipid,” Lancet ii (1985) 59 – 63.
88. J. D. Band, M. E. Chamberland, T. Platt, R. E.
Weaver et al.: “Trends in meningococcal
disease in the United States, 1975 – 1980,” J.
Infect. Dis. 148 (1983) 754 – 758.
89. E. C. Gotschlich in R. Germanier (ed.):
“Meningococcal meningitis,” Bacterial
Vaccines, Academic Press, Orlando, FL 1984,
pp. 237 – 255.
90. L. Lapeyssonnie: “La meningite
cerebro-spinale en Afrique,” Bull. W.H.O. 28
(1963) 1 – 114.
91. E. C. Gotschlich, I. Goldschneider, M. S.
Artenstein: “Human immunity to the
meningococcus. IV. Immunogenicity of group
A and group C meningococcal
polysaccharides in human volunteers; J. Exp.
Med. 129 (1969) 1367 – 1384.
92. M. S. Artenstein, R. Gold, J. G. Zimmerly,
F. A. Wyle et al.: “Prevention of
meningococcal disease by group C
polysaccharide vaccine,” N. Engl. J. Med. 282
(1970) 417 – 420.
93. P. H. Mäkelä et al.: “Effect of group-A
meningococcal vaccine in army recruits in
Finland,” Lancet ii (1975) 883 – 886.
94. M. H. Wahdan, S. A. Sallam, M. N. Hassan,
A. A. Gawad et al.: “A second controlled field
trial of a serogroup A meningococcal vaccine
in Alexandria,” Bull. W.H.O. 55 (1977)
645 – 651.
95. H. Peltola, P. H. Mäkelä, H. M. Käyhty, H.
Jousimies et al.: “Clinical efficacy of
meningococcus group A capsular
polysaccharide vaccine in children three

months to five years of age,” N. Engl. J. Med.
297 (1977) 686 – 691.
96. H. J. Jennings, C. Lugowski:
“Immunochemistry of groups A, B and C
meningococcal polysaccharide-tetanus toxoid
conjugates,” J. Immunol. 127 (1981)
1011 – 1018.
97. C. E. Frasch: “Prospects for the prevention of
meningococcal disease: special reference to
group B,” Vaccine 5 (1987) 3 – 4.
98. W. D. Zollinger, R. E. Mandrell, J. M. Griffis in
J. B. Robbins, J. C. Hill, J. C. Sadoff (eds.):
“Enhancement of immunological activity by
monovalent complexing of meningococcal
group B polysaccharide and outer membrane
protein,” Seminars in Infectious Disease,
vol. 4, Thieme-Stratton, New York 1982,
pp. 254 – 262.
99. C. E. Frasch, G. Coetzee, D. M. Zahradrik,
L. Y. Wang in G. K. Schoolnik (ed.): “New
developments in meningococcal vaccines,”
Pathogenic Neisseria, American Society for
Microbiology, Washington, D.C. 1985,
pp. 633 – 640.
100. F. M. Collins in R. Germanier (ed.):
“Tuberculosis,” Bacterial Vaccines, Academic
Press, Orlando, FL 1984, pp. 373 – 418.
101. R. G. Ferguson, A. B. Simes: “BCG
vaccination of Indian infants in
Saskatchewan,” Tubercle 30 (1949) 5 – 11.
102. S. V. Boyden: “The effect of previous
injections of tuberculoprotein on the
development of tuberculin sensitivity
following B.C.G. vaccination in guinea pigs,”
Br. J. Exp. Pathol. 38 (1957) 611 – 617.
103. A. Lotte, O. Wasz-Höckert, F. Lert, N.
Dumitrescu et al.: “Balance of risks and
balance of cost regarding tuberculosis and
antituberculosis vaccination,” Dev. Biol. Stand.
43 (1979) 111 – 119.
104. R. Torriani, A. Zimmermann, A. Morell: “Die
BCG-Sepsis als letale Komplikation der
BCG-Impfung,” Schweiz. Med. Wochenschr.
109 (1979) 708 –713.
105. W. Hennessen, H. Freudenstein, H.
Engelhardt: “Observations on a BCG vaccine
causing adverse reactions in newborns,” J.
Biol. Stand. 5 (1977) 139 – 146.
106. D. B. Travers: “BCG vaccination,” Lancet i
(1981) 1001 – 1002.
107. M. C. Lu, M. H. Lien, R. E. Becker, H. C.
Heine et al.: “Genes for immunodominant
protein antigens are highly homologous in
Mycobacterium tuberculosis, Mycobacterium
africanum, and the vaccine strain
Immunotherapy and Vaccines 69
Mycobacterium bovis BCG,” Infect. Immun.
55 (1987) 2378 – 2382.
108. B. Moss, C. Flexner: “Vaccinia virus
expression vectors,” Ann. Rev. Immunol. 5
(1987) 305 – 324.
109. H. Leffler, C. Svanborg-Eden: “Glycolipid
receptors for uropathogenic E. coli attaching to
human urinary tract epithelial cells and
agglutinating human erythrocytes,” FEMS
Microbiol. Lett. 8 (1980) 127 – 134.
110. K. Rauss, I. Kétyi, E. Matusovits, L. Szendrei
et al.: “Specific oral prevention of infantile
enteritis. III. Experiments with corpuscular
vaccine,” Acta. Microbiol. Acad. Sci. Hung.
19 (1972) 19 – 28.
111. E. Fürer et al.: “Protection against
colibacillosis in neonatal piglets by
immunization of dams with procholeragenoid,”
Infect. Immun. 35 (1982) 887 – 894.
112. L. Doberescu, C. Huygelen: “Protection of
piglets against neonatal E. coli enteritis by
immunization of the sow with a vaccine
containing heat-labile enterotoxin (LT),
Zentralbl. Veterinärmed. Reihe A 23 (1976)
79 – 88.
113. R. L. Morgan, R. E. Isaacson, H. W. Micon,
C. C. Brinton et al.: “Immunization of suckling
pigs against enterotoxigenic Escherichia
coli-induced diarrheal disease by vaccinating
dams with purified 987 or K 99 pili; protection
correlates with pilus homology of vaccine and
challenge,” Infect. Immun. 22 (1978)
771 – 777.
114. M. M. Levine, R. E. Black, C. C. Brinton, Jr.,
M. L. Clements et al.: “Reactogenicity,
immunogenicity and efficacy studies of
Escherichia coli type 1 somatic pili parenteral
vaccine in man, Scand. J. Infect. Dis. Suppl. 33
(1982) 83 – 95.
115. D. G. Evans, D. J. Evans, Jr., A. R. Opekun,
D. Y. Graham in A. Tagliabue, R. Rappuoli,
S. E. Piazzi (eds.): “Oral whole cell vaccine
protective against enterotoxigenic Escherichia
coli diarrhea,” Bacterial Vaccines and Local
Immunity, Edita Da Sclavo Sp.A., Siena, Italy
1986, pp. 155 – 156.
116. M. M. Levine in R. Germanier (ed.):
Escherichia coli infections,” Bacterial
Vaccines, Academic Press, Orlando, FL 1984
pp. 187 – 235.
117. F. A. Klipstein, R. F. Engert, R. A. Houghton:
“Protection in rabbits immunized with a
vaccine of Escherichia coli heat-stable toxin
cross-linked to the heat-labile toxin B
subunit,” Infect. Immun. 40 (1983) 888 – 893.
70 Immunotherapy and Vaccines
118. J. Holmgren: “Actions of cholera toxin and the
prevention and treatment of cholera,” Nature
(London) 292 (1981) 413 – 417.
119. R. A. Finkelstein, C. V. Sciortino, L. C. Rieke,
M. F. Burks et al.: “Preparation of
procholigenoids from Escherichia coli
heat-labile enterotoxins,” Infect. Immun. 45
(1984) 518 – 521.
120. M. A. Schmidt, P. O’Hanley, G. K. Schoolinik
in A. Tagliabue, R. Rappuoli, S. E. Piazzi
(eds.): Synthetic peptides protect against
infection in a model of experimental
pyelonephritis in mice,” Bacterial Vaccines
and Local Immunity, Edita Da Sclavo Sp.A.,
Siena, Italy 1986, pp. 389 – 398.
121. E. C. Gotschlich in R. Germanier (ed.):
‘’Gonorrhea,” Bacterial Vaccines, Academic
Press, Orlando, FL 1984, pp. 353 – 371.
122. D. S. Kellogg, Jr., I. R. Cohen, L. C. Norins,
A. L. Schroeter et al.: “Neisseria gonorrhoeae.
II. Colonial variation and pathogenicity during
35 months in vitro,” J. Bacteriol. 96 (1968)
596 – 605.
123. J. Swanson, S. J. Kraus, E. C. Gotschlich:
“Studies on gonococcus infection. I. Pili and
zones of adhesion; their relation to gonococcal
growth patterns,” J. Exp. Med. 134 (1971)
886 – 906.
124. L. Greenberg, B. B. Diena, F. A. Ashton, R.
Wallace et al.: “Gonococcal vaccine studies in
Inuvik,” Can. J. Public Health 65 (1974)
29 – 33.
125. E. C. Tramont, J. W. Boslego: “Pilus vaccines,”
Vaccine 3 (1985) 3 – 10.
126. C. C. Brinton, S. W. Wood, A. Brown, A. M.
Labik et al. in : J. B. Robbins, J. Hill, J. C.
Sadoff (eds.): “The development of a Neisseria
pilus vaccine for gonorrhea and meningococcal
meningitis,” Seminars in Infectious Disease,
vol. 4, Bacterial Vaccines, Thieme-Stratton,
Inc., New York 1982, pp. 140 – 159.
127. G. K. Schoolinik, J. Y. Tai, E. C. Gotschlich:
“A pilus peptide vaccine for prevention of
gonorrhea,” Prog. Allergy 33 (1983)
314 – 331.
128. J. B. Rothbard, R. Fernandez, L. Wang,
N. N. H. Teng et al.: “Antibodies to peptides
corresponding to conserved sequence of
gonococcal pilins block bacterial adhesion,”
Proc. Nat. Acad. Sci. U.S.A (2) (1985)
915 – 919.
129. M. S. Blake in G. G. Jackson, H. Thomas (ed.):
“Functions of the outer membrane proteins of
Neisseria gonorrhoeae,” The Pathogenesis of
Bacterial Infections, Springer-Verlag, Berlin
1985, pp. 51 – 66.
130. T. M. Buchanan, M. S. Siegel, K. C. S. Chen,
W. A. Pearce in J. B. Robbins, J. Hill, J. C.
Sadoff (ed.): “Development of a vaccine to
prevent gonorrhea,” Seminars in Infectious
Disease, vol. 4, Bacterial Vaccines,
Thieme-Stratton, Inc., New York 1982,
pp. 160 – 164.
131. J. B. Robbins, R. Schneerson, M. Pittman in R.
Germanier (ed.): Haemophilus influenzae type
b infections,” Bacterial Vaccines, Academic
Press, Orlando, FL 1984, pp. 289 – 316.
132. A. Sutton, R. Schneerson, S. Kendall-Morris,
J. B. Robbins: “Differential complement
resistance mediates virulence of Haemophilus
influenzae type b,” Infect. Immun. 35 (1982)
95 – 104.
133. H. Peltola, H. Käyhty, A. Sivonen, P. H.
Mäkelä: “Haemophilus influenzae type b
capsular polysaccharide vaccine in children; a
double-blind field study of 100 000 vaccines 3
months to 5 years of age in Finland,”
Pediatrics 60 (1977) 730 – 737.
134. E. A. Mortimer, “Efficacy of Haemophilus b
Polysaccharide Vaccine: an Enigma,” J. Am.
Med. Assoc. 260 (1988) 1454 – 1455.
135. H. Käyhty, J. Eskola, H. Peltola, M. G. Stout et
al.: “Immunogenicity in infants of a vaccine
composed of Haemophilus influenzae type b
capsular polysaccharide mixed with DPT or
conjugated to diphtheria toxoid,” J. Infect. Dis.
155 (1987) 100 – 106.
136. A. A. Lenoir, P. D. Granoff, D. M. Granoff:
“Immunogenicity of Haemophilus influenzae
type b polysaccharide-Neisseria meningitidis
outer membrane protein conjugate vaccine in
2- to 6-month-old infants,” Pediatrics 80
(1987) 283 – 287.
137. J. Eskola, H. Peltola, A. K. Takala, H. Käyhty
et al.: “Efficacy of Haemophilus influenzae
type b polysaccharide-diphtheria toxoid
conjugate vaccine in infancy,” N. Engl. J. Med.
317 (1987) 717 –722.
138. C. D. Mitchell, H. H. Balfour, Jr.: “Measles
control: So near and yet so far,” Prog. Med.
Virol. 31 (1985) 1 – 42.
139. L. E. Markowitz, R. H. Bernier:
“Immunization of young infants with
Edmonston-Zagreb measles vaccine,” Pediatr.
Infect. Dis. J. 6 (1987) 809 – 812.
140. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 329
(1966) 52 – 73.

141. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 673
(1982).
142. Centers for Disease Control: “Measles –
United States, 1986,” Morb. Mort. Wkly. Rep.
36 (1987) 301 – 305.
143. K. J. Bart, W. A. Orenstein, A. R. Hinman:
“The virtual elimination of rubella and mumps
from the United States and the use of
combined measles, mumps and rubella
vaccines (MMR) to eliminate measles,” Dev.
Biol. Stand. 65 (1986) 45 – 52.
144. E. Norrby, G. Enders-Ruckle, V. ter Meulen:
“Differences in the appearance of antibodies to
structural components of measles virus after
immunization with inactivated and live virus,”
J. Infect. Dis. 132 (1975) 262 – 269.
145. T. M. Varsanyi, B. Morein, A. Löve, E.
Norrby: “Protection against lethal measles
virus infection in mice by immune-stimulating
complexes containing the hemagglutinin or
fusion protein,” J. Virol. 61(1987)
3896 – 3901.
146. M. R. Hilleman: “Mumps vaccination,” in:
Modern Trends in Medical Virology, vol. 2,
Butterworths, London (1970) pp. 241 – 261.
147. K. Penttinen, K. Cantell, P. Somer, A.
Poikolainen: “Mumps vaccination in the
Finnish defense forces,” Am. J. Epidemiol. 88
(1968) 234 – 244.
148. M. R. Hilleman, R. E. Weibel, E. B. Buynak, J.
Stokes, Jr. et al.: “Live, attenuated
mumps-virus vaccine; 4. Protective efficacy as
measured in a field evaluation,” N. Engl. J.
Med. 276 (1967) 252 – 258.
149. Centers for Disease Control: “Efficacy of
mumps vaccine – Ohio,” Morb. Mort. Wkly.
Rep. 32 (1983) 391–398.
150. F. E. Andre, J. Peetermans: “Effect of
simultaneous administration of live measles
vaccine on the “take rate” of live mumps
vaccine,” Dev. Biol. Stand. 65 (1986)
101 – 107.
151. R. Glück, J. M. Hoskins, A. Wegmann, M. Just
et al.: “Rubini, a new live attenuated mumps
vaccine virus strain for human diploid cells,”
Dev. Biol. Stand. 65 (1986) 29 – 35.
152. WHO Expert Committee on biological
standardization: WHO Tech. Rep. Ser. 760
(1987).
153. Recommendation of the Immunization
Practices Advisory Committee (ACIP):
“Mumps vaccine,” Morb. Mort. Wkly. Rep. 31
(1982) 617 – 625.
Immunotherapy and Vaccines 71
154. K. Penttinen, E.-P. Helle, E. Norrby:
“Differences in antibody response induced by
formaldehyde inactivated and live mumps
vaccines,” Dev. Biol. Stand. 43 (1979)
265 – 268.
155. A. Löve, R. Rydbeck, G. Utter, C. örvell et al.:
‘’Monoclonal antibodies against fusion protein
are protective in necrotizing mumps
meningoencephalitis,” J. Virol. 58 (1986)
220 – 222.
156. F. T. Perkins: “Licensed vaccines,” Rev. Infect.
Dis. 7 (1985) Suppl. 1, 73 – 76.
157. S. A. Plotkin, F. Buser: “History of RA 27/3
rubella vaccine,” Rev. Infect. Dis. 7 (1985)
Suppl. 1, 77 – 78.
158. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 610
(1977) 54 – 84.
159. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 658
(1981).
160. J. A. Dudgeon: “Selective immunization;
Protection of the individual,” Rev. Infect. Dis.
7 (1985) Suppl. 1, 185 – 190.
161. Recommendation of the Immunization
Practices Advisory Committee (ACIP):
“Rubella prevention,” Morb. Mort. Wkly. Rep.
33 (1984) 301 – 318.
162. B. Christenson, M. Böttiger, L. Heller: “Mass
vaccination program aimed at eradicating
measles, mumps, and rubella in Sweden; first
experience,” Br. Med. J. 287 (1983) 389 – 391.
163. Centers for Disease Control: “Rubella
vaccination during pregnancy – United States,
1971 – 1986,” Morb. Mort. Wkly. Rep. 36
(1987) 457 – 461.
164. Recommendation of the Immunization
Practices Advisory Committee (ACIP): “New
recommended schedule for active
immunization of normal infants and children,”
Morb. Mort. Wkly. Rep 35 (1986) 577 – 579.
165. A. Deforest, S. S. Long, H. W. Lischner,
J. A. C. Girone et al.: “Simultaneous
administration of measles-mumps-rubella
vaccine with booster doses of
diphtheria-tetanus-pertussis and poliovirus
vaccines,” Pediatrics 81 (1988) 237 – 246.
166. H. Peltola, O. P. Heinonen: “Frequency of true
adverse reactions to measles-mumps-rubella
vaccine,” Lancet i (1986) 939 – 942.
167. A. M. Arbeter, L. Baker, S. E. Starr, S. A.
Plotkin: “The combination measles, mumps,
rubella and varicella vaccine in healthy
children,” Dev. Biol. Stand. 65 (1986) 89 – 93.
72 Immunotherapy and Vaccines
168. L. B. Schonberger, J. Kaplan, R. Kim-Farley,
M. Moore et al.: “Control of paralytic
poliomyelitis in the United States,” Rev. Infect.
Dis. 6 (1984) Suppl. 2, 424 – 426.
169. J. Salk: “Commentary: Poliomyelitis
vaccination – choosing a wise policy,” Pediatr.
Infect. Dis. J. 6 (1987) 889 – 893.
170. G. S. Wilson: The hazards of immunization,
University of London, The Athlone Press
1967.
171. A. B. Sabin: “Oral poliovirus vaccine: History
of its development and use and current
challenge to eliminate poliomyelitis from the
world,” J. Infect. Dis. 151 (1985) 420 – 436.
172. A. B. Sabin: “Commentary: Is there a need for
a change in poliomyelitis immunization
policy?” Pediatr. Infect. Dis. J. 6 (1987)
887 – 889.
173. M. Böttiger: “Experiences of vaccination with
inactivated poliovirus vaccine in Sweden,”
Dev. Biol. Stand. 47 (1981) 227 – 232.
174. K. Lapinleimu, M. Stenvik: “Experiences with
polio vaccination and herd immunity in
Finland,” Dev. Biol. Stand. 47 (1981)
241 – 246.
175. H. Bijkerk: “Poliomyelitis in the Netherlands,”
Dev. Biol. Stand. 47 (1981) 233 – 240.
176. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 687
(1983) 107 – 174.
177. A. M. Mcbean, J. F. Modlin: “Rationale for the
sequential use of inactivated poliovirus
vaccine and live attenuated poliovirus vaccine
for routine poliomyelitis immunization in the
United States,” Pediatr. Infect. Dis. J. 6 (1987)
881 – 887.
178. K. Lapinleimu: “Sero-immune pattern of
poliomyelitis type I and III in a population
vaccinated with inactivated polio vaccine,” in:
Proceedings of the 11th Symposium of the
European Association Against Poliomyelitis
and Allied Diseases, Rome 1966, Brussels;
European Association Against Poliomyelitis,
1967, pp. 119 – 25.
179. A. L. van Wezel, G. van Steenis, Ch. A.
Hannik, H. Cohen: “New approach to the
production of concentrated and purified
inactivated polio and rabies tissue culture
vaccines,” Dev. Biol. Stand. 41 (1978)
159 – 168.
180. S. E. Robertson, H. P. Traverso, J. A. Drucker,
E. Z. Rovira et al.: “Clinical efficacy of a new,
enhanced-potency, inactivated poliovirus
vaccine,” Lancet i (1988) 897 – 899.
181. B. J. Montagnon, B. Fanget, A. J. Nicolas:
“The large-scale cultivation of vero cells in
micro-carrier culture for virus vaccine
production. Preliminary results for killed
poliovirus vaccine,” Dev. Biol. Stand. 47
(1981) 55 – 64.
182. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 745
(1987).
183. Prevention of liver cancer: WHO Tech. Rep.
Ser. 691 (1983).
184. A. A. McLean: “Development of vaccines
against Hepatitis A and Hepatitis B,” Rev.
Infect. Dis. 8 (1986) 591 – 598.
185. W. Szmuness, C. E. Stevens, E. J. Harley, E. A.
Zang et al.: “Hepatitis B vaccine;
Demonstration of efficacy in a controlled
clinical trial in a high-risk population in the
United States,” N. Engl. J. Med. 303 (1980)
833 – 841.
186. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 725
(1985) 102 – 133.
187. Recommendation of the Immunization
Practices Advisory Committee (ACIP):
“Recommendations for protection against viral
hepatitis,” Morb. Mort. Wkly. Rep. 34 (1985)
313 – 335.
188. Recommendation of the Immunization
Practices Advisory Committee (ACIP):
“Update on Hepatitis B prevention,” Morb.
Mort. Wkly. Rep. 36 (1987) 353 – 366.
189. G. Eder, J. L. McDonel, F. Dorner: “Hepatitis
B vaccine,” in: Progress in Liver Diseases,
vol. 8, Grune and Stratton, New York 1986,
pp. 367 – 394.
190. A. Zuckerman: “Tomorrow’s hepatitis B
vaccines,” Vaccine 5 (1987) 165 – 167.
191. D. N. Standring, W. J. Rutter: “The molecular
analysis of hepatitis B virus,” in: Progress in
Liver Diseases, vol. 8, Grune and Stratton,
New York 1986, pp. 311 – 333.
192. WHO Expert Committee on Rabies: WHO
Tech. Rep. Ser. 709 (1984).
193. Centers for Disease Control: “Bat-rabies –
Europe,” Morb. Mort. Wkly. Rep. 35 (1986)
430 – 432.
194. P. Sureau: “Rabies vaccine production in
animal cell cultures,” Adv. Biochem. Eng.
Biotechnol. 34 (1987) 111 – 128.
195. B. D. Perry: “Rabies,” Vet. Clin. North. Am.
Small Anim. Pract. 17 (1987) 73 – 89.
196. H. Wilde, P. Chomchey, S. Prakongsri, P.
Punyaratabandhu: “Safety of equine rabies
immune globulin,” Lancet ii (1987) 1275.

197. Centers for Disease Control: “Human rabies
despite treatment with rabies immune globulin
and human diploid cell rabies vaccine –
Thailand,” Morb. Mort. Wkly. Rep. 36 (1987)
759 – 765.
198. Centers for Disease Control: “Systemic
allergic reactions following immunization with
human diploid cell rabies vaccine,” Morb.
Mort. Wkly. Rep. 33 (1984) 185 – 187.
199. M. C. Swanson, E. Rosanoff, M. Gurwith, M.
Deitch et al.: “IgE and IgG antibodies to
β-propiolactone and human serum albumin
associated with urticarial reactions to rabies
vaccine,” J. Infect. Dis. 155 (1987) 909 – 913.
200. M. Granström, M. Eriksson, G. Edevåg: “A
sandwich ELISA for bovine serum in viral
vaccines,” J. Biol. Stand. 15 (1987) 193 – 197.
201. B. Morein, K. Simons: “Subunit vaccines
against enveloped viruses; virosomes, micelles
and other protein complexes,” Vaccine 3
(1985) 83 – 93.
202. C. H. Stuart-Harris, G. C. Schild: “Influenza,”
The Viruses and Disease, E. Arnoldt Ltd,
London 1976.
203. P. F. Wright, D. T. Karzon: “Live Attenuated
Influenza Vaccines,” Prog. Med. Virol. 34
(1987) 70 – 88.
204. WHO Expert Committee on Biological
Standardization: WHO Techn. Rep. Ser. 638
(1979).
205. Recommendations of the Immunization
Practices Advisory Committee (ACIP):
“Prevention and Control of Influenza,” Morb.
Mort. Wkly. Rep. 36 (1987) 373 – 387.
206. P. A. Gross, A. L. Gould, A. E. Brown: “Effect
of cancer chemotherapy on the immune
response to influenza virus vaccine: Review of
published studies,” Rev. Infect. Dis. 7 (1985)
613 – 618.
207. M. A. Strassburg, S. Greenland, F. J. Sorvillo,
L. E. Lieb et al.: “Influenza in the elderly:
report of an outbreak and a review of vaccine
effectiveness reports,” Vaccine 4 (1986)
38 – 44.
208. J. M. Ostrove, G. Inchauspe: “The biology of
varicella-zoster virus” in: S. E. Straus
“Varicella-zoster virus infections: biology,
natural history, treatment, and prevention,”
Ann. Intern. Med. 108 (1988) 221 –237.
209. M. Takahashi: “Clinical overview of varicella
vaccine; Development and early studies,”
Pediatrics 78 (1986) Suppl., 736 – 741.
210. R. E. Weibel, B. J. Neff, B. J. Kuter, H. A.
Guess et al.: “Live attenuated varicella virus
Immunotherapy and Vaccines 73
vaccine. Efficacy trial in healthy children,” N.
Engl. J. Med. 310 (1984) 1409 – 1415.
211. A. A. Gershon, S. P. Steinberg, L. Gelb, and
The National Institute of Allergy and
Infectious Diseases Varicella Vaccine
Collaborative Study Group: “Live attenuated
varicella vaccine use in immunocompromised
children and adults,” Pediatrics 78 (1986)
Suppl., 757 – 762.
212. Arthropod-borne and rodent-borne viral
diseases: WHO Tech. Rep. Ser. 719 (1985).
213. Viral haemorrhagic fevers: WHO Tech. Rep.
Ser. 721 (1985).
214. K. C. Smithburn, C. Durieux, R. Koerber,
H. A. Penna et al.: Yellow Fever Vaccination,
World Health Organization, Geneva 1956.
215. J. C. Roche, A. Jouan, B. Brisou, R. Rodhain
et al.: “Comparative clinical study of a new
17 D thermostable yellow fever vaccine,”
Vaccine 4 (1986) Suppl., 163 – 165.
216. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 594
(1976) 23 – 49.
217. A. A. Smorodintsev, A. V. Dubov, V. I. Ilyenko,
V. G. Platonov: “A new approach to
development of live vaccine against tick-borne
encephalitis,” J. Hyg. 67 (1969) 13 – 20.
218. C. Kuntz, F. X. Heinz, H. Hofmann:
“Immunogenicity and reactogenicity of a
highly purified vaccine against tick-borne
encephalitis,” J. Med. Virol. 6 (1980)
103 – 109.
219. A. Oya in H. Fukumi (ed.): “Japanese
encephalitis vaccine,” The vaccination. Theory
and practice, International Medical
Foundation of Japan (1975) pp. 69 – 82.
220. WHO Expert Committee on Biological
Standardization: WHO Tech. Rep. Ser. 323
(1966) 56 – 71.
221. Recommendation of the Immunization
Practices Advisory Committee (ACIP):
Smallpox Vaccine Morb. Mort. Wkly. Rep. 34
(1985) 341 – 342.
222. R. Randall, L. N. Binn, V. R. Harrison:
“Immunization against Rift Valley fever
virus,” J. Immunol. 93 (1964) 293 – 299.
223. B. Niklasson: “Rift Valley fever virus vaccine
trial; Study of side effects in humans,” Scand.
J. Infect. Dis. 14 (1982) 105 – 109.
224. P. M. Kaye, Parasitol. Today 3 (1987)
293 – 299.
225. M. D. Rickard, M. J. Howell in E. E. R. Taylor,
J. R. Baker (ed.): In Vitro Methods for Parasite
Cultivation, Academic Press, London 1987,
pp. 407 – 451.
74 Immunotherapy and Vaccines
226. G. Piekarski: Medizinische Parasitologie,
Springer Verlag, Berlin 1987.
227. V. Houba in V. Houba (ed.): Immunological
Investigation of Tropical Parasitic Diseases,
Churchill Livingstone, Edinburgh 1980,
pp. 130 – 147.
228. P. Hagan, P. J. Moore, A. B. Adjukiewicz,
B. M. Greenwood et al.: Parasite Immunol. 7
(1985) 617 – 624.
229. A. E. Butterworth, M. Capron, J. S.
Cordingley, P. R. Dalton, Trans. R. Soc. Trop.
Med. Hyg. 79 (1985) 393 – 408.
230. G. Gazzinelli, J. R. Lambertucci, N. Katz, R. S.
Rocha, J. Immunol. 135 (1985) 2121 – 2127.
231. C. W. Todd, R. W. Goodgame, D. Colley, J.
Immunol. 122 (1979) 1440 – 1446.
232. R. F. Sturrock, R. Kimani, B. I. Cottrell, A. E.
Butterworth, Trans. R. Soc. Trop. Med. Hyg.
77 (1983) 363 – 371.
233. A. E. Butterworth, R. F. Sturrock, V. Houba,
A. A. F. Mahmoud, Nature (London) 256
(1975) 727 – 729.
234. A. E. Butterworth, Adv. Parasitol. 23 (1984)
143 – 235.
235. S. R. Smithers, R. J. Terry, Adv. Parasitol. 14
(1976) 399 – 422.
236. A. Capron, Fortschr. Zool. 27 (1982)
259 – 264.
237. A. E. Butterworth, P. Hagan, Parasit. Today 3
(1987) 11 – 16.
238. A. J. G. Simson, D. Cioli, Parasit. Today 3
(1987) 26 – 28.
239. C. Dissous, I. M. Crzych, A. Capron, J.
Immunol. 129 (1982) 2232 – 2234.
240. Q. D. Bickle, B. I. Andrews, M. G. Taylor,
Parasit Immunol. 8 (1986) 95 – 107.
241. C. Kelly, A. I. G. Simpson, E. Fox, S. M.
Phillips et al.: Parasit Immunol. 8 (1986)
193 – 198.
242. J. M. Balloul, P. Sondermeyer, D. Dreyer, M.
Capron, Nature (London) 326 (1987)
149 – 153.
243. D. Wakelin: Immunity to Parasites, Edward
Arnold Ltd., London 1984.
244. S. Friedmann, Gen Eng. News (1985) 24 – 25.
245. V. K. Vinayak, N. K. Gupta, A. K. Chopra,
G. L. Sharma et al., Parasitology 82 (1981)
375 – 382.
246. P. Ambroise-Thomas in V. Houba (ed.):
Immunological Investigation of Tropical
Parasitic Diseases, Churchill Livingstone,
Edinburgh 1980, pp. 84 – 103.
247. B. M. Ogilvie, M. J. Worms in S. Cohen, E. H.
Sadun (eds.): Immunology of Parasitic
Infections, Blackwell Scientific Publ., Oxford
1976, 397 – 402.
248. M. M. Wong, H. I. Fredericks, C. P.
Ramachandran, Bull. W.H.O. 40 (1969)
493 – 501.
249. C. P. Ramachandran, Southeast Asian J. Trop.
Med. Public Health 1 (1970) 78 – 92.
250. A. Haque, A. Capron in T. W. Pearson (ed.):
Parasite Antigens, Marcel Dekker Inc., New
York 1986, pp. 317 – 402.
251. R. M. Maizels, F. Partono, Sri Oemijati, D. A.
Denham et al., Parasitology 87 (1983)
249 – 263.
252. M. D. Rickard in C. Arme, P. W. Pappas (eds.):
Biology of the Eucestoda, Academic Press,
London 1983, pp. 539 – 579.
253. M. D. Rickard, F. F. Williams, Adv. Parasitol.
21 (1982) 229 – 296.
254. R. P. Herd, R. I. Chappel, D. Biddell, Int. J.
Parasitol. 5 (1975) 395 – 399.
255. D. J. Jenkins, M. D. Rickard, Aust. Vet. J. 63
(1986) 40 – 42.
256. R. S. Nussenzweig, J. Vanderberg, H. Most, C.
Orton, Nature (London) 216 (1967) 160 – 162.
257. W. A. Siddiqui, Science 197 (1977) 388 – 389.
258. G. H. Mitchell, W. H. G. Richards, G. A.
Butcher, S. Cohen, Lancet (i) (1977)
1335 – 1358.
259. W. Trager, J. B. Jensen, Science 193 (1976)
673 – 675.
260. D. Mazier, R. L. Beaudoin, S. Mellouk, P.
Druilhe et al., Science 227 (1985) 440 – 442.
261. C. Guerin-Marchand, P. Druilhe, B. Galey, A.
Londono et al., Nature (London) 329 (1987)
164 – 167.
262. A. H. Cochrane in J. P. Kreier (ed.): Malaria,
Immunology and Immunization vol. 3, Chap. 4,
Academic-Press, New-York, 1980.
263. D. F. Clyde, Am. J. Trop. Med. Hyg. 24 (1975)
397 – 401.
264. E. H Nardin, R. S. Nussenzweig, I. A.
McGregor, J. H. Bryan, Science 206 (1979)
597 – 599.
265. G. Del Giudice, H. D. Engers, C. Tougne, S. S.
Biro et al., Am. J. Trop. Med. Hyg. 36 (1987)
203 – 212.
266. D. H. Chen, R. E. Tigelaar, F. I. Weinbaum, J.
Immunol. 118 (1977) 1322 – 1327.
267. A. Ferreira, L. Schofield, V. Enea, H.
Schellekins et al., Science 232 (1986)
881 – 884.
268. P. Potocnjak, N. Yoshida, R. S. Nussenzweig,
V. Nussenzweig, J. Exp. Med. 151 (1980)
1504 – 1513.
269. D. Mazier, S. Mellouk, R. L. Beaudoin, B.
Texier et al., Science 231 (1986) 156 – 159.

270. J. B. Dame, J. L. Williams, T. F. McCutchan,
J. L. Weber et al., Science 225 (1984)
593 – 599.
271. V. Enea, J. Ellis, F. Zavala, D. E. Arnot et al.,
Science 225 (1984) 628 – 630.
272. D. J. Kemp, R. L. Coppel, R. F. Anders, Ann.
Rev. Microbiol. 41 (1987) 181 – 208.
273. V. F. De la Cruz, A. A. Lal, T. F. McCutchan, J.
Biol. Chem. 262 (1987) 11935 – 11941.
274. M. J. Lockyer, R. T. Schwarz, Mol. Biochem.
Parasitol. 22 (1987) 101 – 107.
275. G. Del Giudice, A. S. Verdini, M. Pinori, J. P.
Verhave et al., J. Clin. Microbiol. 25 (1987)
91 – 96.
276. S. L. Hoffman, R. Wistar, Jr., W. R. Ballou,
M. R. Hollingdale et al., New-Engl. J. Med.
315 (1986) 601 – 603.
277. D. A. Herrington, D. F. Clyde, G. Losonsky,
M. Cortesia et al., Nature (London) 328
(1987) 257 – 259.
278. W. R. Ballou, S. L. Hoffman, J. A. Sherwood,
M. R. Hollingdale et al., Lancet I (1987)
1277 – 1279.
279. G. Del Giudice, J. A. Cooper, J. Merino, A. S.
Verdini et al., J. Immunol. 137 (1986)
2952 – 2955.
280. M. F. Good, J. A. Berzofsky, W. L. Maloy, Y.
Hayashi et al., J. Exp. Med. 164 (1986)
655 –660.
281. A. R. Togna, G. Del Giudice, A. S. Verdini, F.
Bonetti et al., J. Immunol. 137 (1986)
2956 – 2960.
282. M. F. Good, D. Pombo, I. A. Quakyi, E. M.
Riley et al., Proc. Natl. Acad. Sci. USA 85
(1988) 1199 – 1203.
283. L. H. Miller, S. J. Mason, D. F. Clyde, M. H.
McGinis, New-Eng. J. Med. 295 (1976)
302 – 304.
284. M. J. Miller, Trans. Roy. Soc. Trop. Med. Hyg.
52 (1958) 152 – 158.
285. S. Cohen, I. A. McGregor, S. C. Carrington,
Nature (London) 192 (1961) 733 – 735.
286. A. Celada, A. Cruchau, L. H. Perrin, Clin. Exp.
Immunol. 47 (1982) 635 – 641.
287. P. H. David, Proc. Natl. Acad. Sci. USA 80
(1983) 5075 – 5081.
288. L. H. Miller, R. J. Howard, R. Carter, M. F.
Good et al., Science 234 (1986) 1349 – 1356.
289. L. H. Perrin, E. Ramirez, L. Er-Hsiang, P.-H.
Lambert; Clin. Exp. Immunol. 41 (1980)
91 – 106.
290. A. A. Holder, R. R. Freeman, J. Exp. Med. 156
(1982) 1528 – 1538.
291. J. S. McBride, D. Walliker, G. Morgan,
Science 217 (1982) 254 – 256.
Immunotherapy and Vaccines 75
292. A. A. Holder, M. J. Lockyer, K. G. Odink, J. S.
Sandhu et al., Nature (London) 317 (1985)
270 – 273.
293. K. G. Oding, M. J. Lockyer, S. C. Nicholls, Y.
Hillman et al., FEBS Lett. 173 (1984)
108 – 112.
294. M. Mackay, M. Goman, N. Bone, J. E. Hyde et
al., EMBO J. 4 (1985) 3823 – 3829.
295. L. H. Perrin, B. Merkli, M. Loche, C.
Chizzolini et al., J. Exp. Med. 160 (1984)
441 – 447.
296. R. Hall, J. E. Hyde, M. Goman, D. L. Simmons
et al., Nature (London) 311 (1984) 379 – 382.
297. W. A. Siddiqui, L. Q. Tam, K. J. Kramer,
G. S. N Hui et al., Proc. Natl. Acad. Sci. USA
84 (1987) 3014 – 3018.
298. A. A. Holder, R. R. Freeman, Nature (London)
294 (1981) 361 – 363.
299. M. E. Patarroyo, P. Romero, M. L. Torres, P.
Clavijo et al., Nature (London) 328 (1985)
629 – 631.
300. A. Cheung, J. Leban, A. R. Shaw, B. Merkli et
al., Proc. Natl. Acad. Sci. USA 83 (1986)
8323 – 8332.
301. L. H. Perrin, R. Dayal, Immunol. Rev. 61
(1982) 245 – 268.
302. R. R. Freeman, A. J. Trejosiewicz, G. A. Cross,
Nature (London) 284 (1980) 366 – 369.
303. L. H. Perrin, E. Ramirez, P.-H. Lambert, P. A.
Miescher, Nature (London) 289 (1981)
301 – 303.
304. C. Braun-Breton, M. Jendoubi, E. Brunet, L.
Perrin et al., Mol. Bioch. Parasitol. 20 (1986)
33 – 38.
305. C. Braun-Breton, T. L. Rosenberg, L. Pereira
Da Silva, Nature (London) 332 (1988)
457 – 459.
306. U. Certa, P. Ghersa, H. Dobeli, H. Matile et al.,
Science 240 (1988) 1036 – 1038.
307. P. Dubois, J. P. Dedet, T. Fandeur, C.
Roussilhon et al., Proc. Natl. Acad. Sci. USA
81 (1984) 229 – 335.
308. L. H. Perrin, B. Merkli, M. S. Gabra, J. Stocker
et al., J. Clin. Invest. 75 (1985) 1718 – 1725.
309. J. F. Dubremetz, P. Delplace, B. Fortier, G.
Tronchin et al., Mol. Biochem. Parasitol. 27
(1988) 135 – 142.
310. R. L. Coppel, J. G. Culvenor, A. E. Bianco,
P. E. Crewther et al., Mol. Biochem. Parasitol.
25 (1987) 73 – 81.
311. H. Perlamnn, K. Berzins, M. Wahlgren, J.
Carlsson et al., J. Exp. Med. 159 (1984)
1686 – 1693.
76 Immunotherapy and Vaccines
312. G. V. Brown, I. G. Culvenor, P. E. Crewther,
A. E. Bianco et al., J. Exp. Med. 162 (1985)
774 – 779.
313. B. Wahlin, M. Wahlgren, H. Perlamnn, H.
Berzins et al., Proc. Natl. Acad. Sci. USA 81
(1984) 7912 – 7916.
314. J. M. Favarolo, R. L. Coppel, L. M. Corcoran,
S. J. Foote et al., Nucl. Acid. Res. 14 (1986)
8265 –8277.
315. W. E. Collins, R. F. Anders, M. Pappaioanou,
G. H. Campbell et al., Nature (London) 323
(1986) 259 – 262.
316. S. A. Luse, L. H. Miller, Am. J. Trop. Med.
Hyg. 20 (1971) 655 – 670.
317. T. Triglia, H. D. Stahl, P. E. Crewther, D.
Scanlon et al., EMBO J. 6 (1987) 1413 – 1419.
318. M. Koenen, A. Scherf, O. Mercereau, G.
Langsley et al., Nature (London) 311 (1984)
382 – 385.
319. I. J. Udeinya, L. H. Miller, I. A. McGregor,
J. B. Jensen, Nature (London) 231 (1983)
429 – 431.
320. K. Marsh, R. J. Howard, Science 231 (1986)
150 – 153.
321. S. Pollack, J. Fleming, Brit. J. Haematol. 58
(1984) 289 – 294.
322. M. E. Perkins, J. Exp. Med. 160 (1984)
788 – 797.
323. D. Camus, T. J. Hadley, Science 230 (1985)
553 – 557.
324. P. Delplace, B. Fortier, G. Tronchin, J. F.
Dubremetz et al., Mol. Biochem. Parasitol. 23
(1987) 193 – 202.
325. M. E. Patarroyo, R. Amador, P. Lavijo, A.
Moreno et al., Nature (London) 332 (1988)
158 – 161.
326. R. W. Gwadz, Science 193 (1976)
1150 – 1153.
327. K. N. Mendis, G. A. T. Targett, Nature
(London) 277 (1979) 289 – 391.
328. R. W. Gwadz, L. C. Koontz, Infect. Immun. 44
(1984) 137 – 145
329. N. Kumar, R. Carter, Mol. Biochem. Parasitol.
13 (1984) 333 – 341.
330. K. Mendis, P. Udayama, R. Carter, P. H. David,
J. Cell Biochem. Suppl. 10 A (1986) 149 – 156.
331. A. N. Vermuelen, T. Ponnudurai, P. J. A.
Beckers, J. P. Verhave et al., J. Exp. Med. 162
(1985) 1460 – 1476.
332. J. Rener, P. M. Graves, R. Carter, J. L. Williams
et al., J. Exp. Med. 158 (1983) 976 – 980.
333. P. M. Graves, R. Carter, T. R. Burkot, J. Rener
et al., Infect. Immun. 48 (1985) 611 – 619.
334. E. R. Stiehm, E. Ashida, K. S. Kim, D. J.
Winston et al., Ann. Intern. Med. 107 (1987)
367 – 82.
335. E. R. Stiehm, Pediatrics 63 (1979) 301 – 19.
336. J. A. Finlayson, in C. S. F. Easmon, J.
Jeljaszewics (eds.): Medical Microbiology,
vol. 1, Academic Press, London 1982,
pp. 129 – 82.
337. R. Heffron: Pneumonia, With Special
Reference to Pneumococcus Lobar
Pneumonia, The Commonwealth Fund, New
York 1939.
338. R. H. Rousell, M. S. Collins, M. B. Dobkin,
R. E. Louie et al., Am. J. Med. Suppl. 76
(1984) 40 – 145.
339. H. E. Alexander, Am. J. Dis. Child. 66 (1943)
172 – 187.
340. H. E. Alexander, Am. J. Dis. Child. 66 (1943)
160 – 171.
341. E. Merler, F. S. Rosen, N. Engl. J. Med. 275
(1966) 480 – 86; 536 – 542.
342. C. A. Janeway, F. S. Rosen, N. Engl. J. med.
275 (1966) 826 – 831.
343. P. A. M. Gross, D. Gitlin, C. A. Janeway, N.
Engl. J. Med. 260 (1959) 170 – 178.
344. J. Stokes, E. P. Maris, S. S. Gellis, J. Clin.
Invest. 23 (1944) 531 – 540.
345. J. Stokes, J. R. Neefe, JAMA J. Am. Med.
Assoc. 127 (1945) 531 – 540.
346. W. M. Hammon, L. L. Coriell, J. Stokes et al.,
JAMA J. Am. Med. Assoc. 150 (1950) 739 – 49;
750 – 756.
347. C. A. Janeway, L. Apt, D. Gitlin, Trans. Assoc.
Am. Physician 66 (1953) 200 – 202.
348. P. A. M. Gross, D. Gitlin, C. A. Janeway, N.
Engl. J. Med. 260 (1959) 121 – 125.
349. R. G. Hamilton, Clin. Chem. (Winston Salem
N.C.) 33 (1987) 1707 – 1725.
350. G. R. Siber, P. H. Schur, A. C. Weitzman, G.
Schiffman, N. Engl. J. Med. 303 (1980)
178 – 182.
351. C. M. Roifman, H. M. Lederman, S. Lavi, L. D.
Stein et al., Am. J. Med. 79 (1985) 171 – 174.
352. S. Barandun, H. Isliker, Vox Sang. 51 (1986)
157 – 160.
353. A. Hassig, Vox Sang. 51 (1986) 10 – 17.
354. D. D. Schroeder, M. L. Dumas, Am. J. Med.
Suppl. 76 (1984) 33 – 39.
355. C. A. Janeway, E. Merler, F. S. Rosen, S.
Salmon et al., N. Engl. J. Med. 278 (1968)
919 – 923.
356. J. Passwell, F. S. Rosen, E. Merler in B.
Alving, J. Finlayson (eds.): Immunoglobulins;
characteristics and uses of intravenous
preparations, U.S.Dept. H.H.S., Washington
1979, pp. 139 – 142.

357. H. D. Ochs, S. H. Fischer, R. J. Wedgwood, J.
Clin. Immunol Suppl. 2 (1982) 22 S – 29 S.
358. C. Cunningham-Rundles, F. P. Siegal, E. M.
Smithwick, A. Lion-Boule et al., Ann. Intern.
Med. 101 (1984) 435 – 439.
359. M. S. Collins, J. H. Dorsey, J. Infect. Dis. 151
(1985) 1171 – 1173.
360. J. R. Schreiber, V. A. Barrus, G. R. Siber,
Infect. Immun. 47 (1985) 142 – 148.
361. H. D. Ochs et al., Lancet 2 (1980)
1158 – 1159.
362. Medical Letter 24 (1982) 81 – 82.
363. R. H. Johnson, R. J. Ellis, Ann. Intern. Med.
81 (1974) 61 – 67.
364. C. Cunningham-Rundles, L. A. Hanson, W. H.
Hitzig, W. Knapp et al., Bull. W.H.O. 60
(1982) 43 – 47.
365. S. Krugman, N. Engl. J. Med. 269 (1963)
195 – 201.
366. E. B. Grosman, S. G. Stewart, J. Stokes, JAMA
J. Am. Med. Assoc. 129 (1945) 991 – 994.
367. Medical Research Council Workingparty,
Lancet 1 (1969) 163 – 168.
368. C. L. S. Leen, P L. Yap, D. B. L. McClelland,
Vox Sang. 51 (1986) 278 – 286.
369. Lancet 1 (1983) 105 – 106.
370. G. Chirico, G. Rondini, A. Plebani, A. Chiara
et al., J. Pediatr. 110 (1987) 437 – 442.
371. J. Amer, E. Ott, F. A. Ibbott, D. O’Brien et al.,
Pediatrics 32 (1963) 4 – 9.
372. A. A. Hertler, S. C. Ross, J. Clin. Lab.
Immunol. 21 (1986) 177 – 181.
373. S. Baron, E. V. Barnet, R. S. Goldsmith, S.
Silbergeld et al., Am. J. Hyg. 79 (1964)
186 – 195.
374. N. A. Kefalides, J. A. Arana, A. Bazan, M.
Bocanegra et al., N. Engl. J. Med. 267 (1962)
317 – 324.
375. H. H. Stone, C. D. Graber, J. D. Martin, L.
Kolb, Surgery St. Louis 58 (1965) 810 – 814.
376. R. J. Jones, E. A. Roe, J. L. Gupta, Lancet 2
(1980) 1263 – 1265.
377. R. L. Wasserman, Pediatric Infectious
Diseases 5 (1986) 620 – 621.
378. J. A. Steen, Arch. Pediatr. 77 (1960) 291 – 304.
379. A. P. Schless, G. S. Harell, Amer. J. Med. 44
(1968) 325 – 329.
380. M. Santosham, R. Reid, D. M. Ambrusino,
M. C. Wolff et al., Engl. J. Med. 317 (1987)
923 – 9.
381. S. M. Rosenthal, R. C. Millican, J. Rust, Proc.
Soc. Exp. Biol. Med. 94 (1957) 214 – 217.
382. M. W. Fisher, Antibiot. Chemother.
(Washington D.C.) 7 (1957) 315 – 321.
Immunotherapy and Vaccines 77
383. M. W. Fisher, M. C. Manning, Antibiot.
Chemother. (Washington D.C.) 8 (1958)
29 – 31.
384. B. A. Waisbren, Antibiot. Chemother.
(Washington D.C.) 7 (1957) 322 – 333.
385. B. A. Waisbren, D. Lepley, Arch. Intern. Med.
109 (1962) 712 – 716.
386. G. P. Bodey, B. A. Nies, N. R. Mohberg, E. J.
Freireich, JAMA J. Am. Med. Assoc. 190
(1964) 1099 – 1102.
387. K. C. Finkel, J. C. Haworth, Pediatrics 25
(1960) 798 – 806.
388. D. R. Snydman, B. G. Werner, B.
Heinzer-Lacey, V. P. Berardi et al., N. Engl. J.
Med. 317 (1987) 1049 – 1054.
389. R. M. Condie, R. J. O’Reilly, Am. J. Med. 76
(Suppl. 30 March 1984) 134 – 141.
390. A. Hagenbeek, H. G. J. Brummelhuis, A.
Donkers, A. M. Dumas et al., J. Infect. Dis.
155 (1987) 897 – 902.
391. D. J. Winston, W. G. Ho, C.-H. Lin, M. D.
Budinger et al., Am. J. Med. 76 (Suppl.
30 March 1984) 128 – 133.
392. R. A. Bowden, M. Sayers, N. Flournoy, B.
Newton et al., N. Engl. J. Med. 314 (1986)
1006 – 1010.
393. E. C. Reed et al., J. Infect. Dis. 156 (1987)
641 –644.
394. K. Erlendsson, T. Swartz, J. M. Dwyer, N.
Engl. J. Med. 312 (1985) 351 – 353.
395. P. J. Mease, H. D. Ochs, R. J. Wedgood, N.
Engl. J. Med. 304 (1981) 1278 – 1281.
396. J. M. Crennan, R. E. Van Scoy, C. H. McKenna,
T. F. Smith, Am. J. Med. 81 (1986) 35 – 42.
397. K. S. Erlich, J. Mills, Rev. Infect. Dis. 8 (1986)
S 439 – 445.
398. D. Ambrosino, J. R. Schreiber, R. S. Daum,
G. R. Siber, Infect. Immun. 39 (1983)
709 – 714.
399. T. E. Harper, R. D. Christensen, G. Rothstein,
Pediatr. Res. 22 (1987) 455 – 460.
400. S. J. Cryz, A. S. Cross, E. Furer, N. Chariatte et
al., J. Lab. Clin. Med. 108 (1986) 182 – 189.
401. I. A. Holder, J. G. Naglich, Am. J. Med. 76
(1984) 161 – 167.
402. J. E. Pennington, G. J. Small, J. Infect. Dis.
155 (1987) 973 – 978.
403. V. G. Hemming, W. T. London, G. W. Fischer,
B. L. Curfman et al., J. Infect. Dis. 156 (1987)
655 – 658.
404. K. N. Haque, M. H. Zaidi, S. K. Haque, H.
Bahakim et al., Pediatric Infectious Diseases
5 (1986) 622 – 625.
405. D. Sidiropoulos, U. Herrmann, A. Morell, G.
von Muralt et al., J. Pediatr. St. Louis 109
(1986) 505 – 508.
78 Immunotherapy and Vaccines
406. W. G. Ho, D. J. Winston, K. Bartoni, R. E.
Champlin et al., Abstr. 456, 1983 Interscience
Confer. Infect. Dis. Chemother., Las Vegas,
Nevada.
407. C. Bunch, H. Chapel, K. R. Rai, R. P. Gale,
Blood 70 (1987) 224 a (Abstr); N. Engl. J.
Med. 319 (1988) 902 – 907.
408. H. D. Ochs, Pediatr. Infect. Dis. 6 (1987)
509 – 511.
409. C. C. Wood, J. G. McNamara, D. F. Schwarz,
W. W. Merrill et al., Pediatr. Infect. Dis. 6
(1987) 564 – 566.
410. B. A. Silverman, A. Rubinstein, Am. J. Med.
78 (1985) 728 – 736.
411. D. Sidiropoulos, U. Boehme, G. von Muralt,
A. Morell et al., Pediatr. Infect. Dis. 5 (1986)
193 –194.
412. S. Barandun, P. Imbach, A. Morrell, H. P.
Wagner in U. E. Nydegger (ed.):
Immunohemotherapy. A guide to
immunoglobulin prophylaxis and therapy,
Academic Press, New York 1981,
pp. 275 – 282.
413. P. Imbach, H. P. Wagner, W. Berchtold, G.
Gaedicke et al., Lancet 2 (1985) 464 – 468.
414. J. Fehr, V. Hoffmann, U. Kappeler, N. Engl. J.
Med. 306 (1982) 1254 – 1258.
415. S. Rosthoj, G. K. Steffensen, T. K. Guld, Acta
Paediatr. Scand. 76 (1987) 631 – 635.
416. R. Bohm, C. Hofstaetter, R. C. Briel, Blut 48
(1984) 469 – 470.
417. F. E. Leickly, R. H. Buckley, Am. J. Med. 82
(1987) 159 – 162.
418. G. W. Richmond, I. Ray, A. Korenblitt, J.
Pediatr. (St. Louis) 110 (1987) 917 – 919.
419. M. W. Hilgartner, J. Bussel, Am. J. Med. 83
(1987) 35 – 39.
420. T. Becker, S. Panzer, D. Maas, V. Kiefel et al.,
Br. J. Haematol. 61 (1985) 149 – 155.
421. F. Dammacco, G. Iodice, N. Campobasso, Br.
J. Haematol. 62 (1986) 125 – 135.
422. A. C. Newland, J. G. Treleaven, R. M.
Minchinton, A. H. Waters, Lancet 1 (1983)
84 – 87.
423. J. B. Bussel, R. P. Kimberly, R. D. Inman, I.
Schulmann et al., Blood 62 (1983) 480 – 486.
424. J. R. Duran-Suarez, A. Martin, M. C. Botella,
S. de la Torre et al., Haematologica 68 (1983)
564 – 566.
425. C. A. Schiffer et al., Blood 64 (1984)
937 – 940.
426. G. Chirico, M. Duse, A. G. Ugazio, G.
Rondini, J. Pediatr. (St. Louis) 103 (1983)
654 – 655.
427. E. C. Besa, M. W. McNab, A. J. Solan, M. J.
Lapes et al., Am. J. Hematol. 18 (1985)
373 – 379.
428. W. A. McGuire, H. H. Yang, E. Bruno, J.
Brandt et al., N. Engl. J. Med. 317 (1987)
1004 – 1008.
429. C. De la Camara, R. Arrieta, A. Gonzalez, E.
Iglesias et al., N. Engl. J. Med. 318 (1988)
519 – 520.
430. I. M. Nilsson, E. Berntorp, O. Zettervall, N.
Engl. J. Med. 318 (1988) 947 – 950.
431. J. W. Newburger, M. Takahashe, J. C. Burns,
A. S. Beiser et al., N. Engl. J. Med. 315 (1986)
341 – 347.
432. M. Nagashima, M. Matsushima, H. Matsuoka,
A. Ogawa et al., J. Pediatr. (St. Louis) 110
(1987) 710 – 712.
433. D. Y. M. Leung, J. C. Burns, J. W. Newburger,
R. S. Geha, J. Clin. Invest. 79 (1987)
468 – 472.
434. B. Combe, B. Cosso, J. Clot, M. Bonneau et
al., Am. J. Med. 78 (1985) 920 – 928.
435. F. C. Breedveld, A. Brand, W. G. van Aken, J.
Rheumatol. 12 (1985) 700 – 702.
436. E. L. Arsura, A. S. Bick, N. G. Brunner, T.
Namba et al., Arch. Intern. Med. 146 (1986)
1365 – 1368.
437. K. Kawada, P. I. Terasaki, Exp. Hematol (N.Y.)
15 (1987) 133 – 136.
438. E. Schuller, A. Govaerts, Eur. Neurol. 22
(1983) 205 – 212.
439. A. Salama, C. Mueller-Eckhardt, V. Kiefel,
Lancet 2 (1983) 193 – 195.
440. R. P. Kimberly, J. E. Salmon, J. B. Bussel,
M. K. Crow et al., J. Immunol. 132 (1984)
745 – 750.
441. U. Budde et al., Scand. J. Haematol. 37
(1986) 125 – 129.
442. A. Etzioni, S. Pollack, A. Benderly, N. Engl. J.
Med. 318 (1988) 994.
443. G. P. Sandilands, H. I. Atrah, G. Templeton,
J. E. Cocker et al., J. Clin. Lab. Immunol. 23
(1987) 109 – 115.
444. T. W. Jungi, S. Barandun, Vox Sang. 49 (1985)
9 – 19.
445. W. Stohl, Clin. Exp. Immunol. 62 (1985)
200 – 207.
446. F. Hashimoto, Y. Sakiyama, S. Matsumoto,
Clin. Exp. Immunol. 65 (1986) 409 – 415.
447. H. H. Hodes, Pediatrics. 32 (1963) 1 – 3.
448. W. B. White, C. R. Desbonnet, M. Ballow, Am.
J. Med. 83 (1987) 431 – 434.
449. K. Prellner, P. Christensen, O. Kalm, K.
Offenbartl, Acta Pathol. Microbiol. Scand.
Sect. 94 C (1986) 207 – 211.

450. K. K. Christensen, P. Christensen, Pediatric
Infectious Diseases 5 (1986) S 189 – 192.
451. J. G. Kelton, C. J. Carter, C. Rodger, G.
Bebenek et al., Blood 63 (1984) 1434 – 1438.
452. A. S. Cross, B. M. Alving, J. C. Sadoff, P.
Baldwin et al., Lancet 1 (1984) 912.
453. M. Berger, P. Rosencranz, C. Y. Brown, Clin.
Immunol. Immunopathol. 34 (1985) 227 – 236.
454. J. Kulics, E. Rajnavolgyi, G. Fust, J. Gergely,
Mol. Immunol. 20 (1983) 805 – 810.
455. E. F. Ellis, C. S. Henney, J. Allergy 43 (1969)
45 – 54.
456. R. C. Williams, O. J. Mellbye, G. Kronvall,
Infect. Immun. 6 (1972) 316 – 323.
457. D. S. Feingold, F. Oski, Arch. Intern. Med.
116 (1965) 226 – 228.
458. R. C. Millican, J. D. Rust, J. Infect. Dis. 107
(1960) 389 – 394.
459. E. Lachman, S. B. Pitsoe, S. L. Gaffin, Lancet
1 (1984) 981 – 983.
460. M. A. Thomson, J. S. Afr. Vet. Assoc. 54
(1983) 279 – 281.
461. M. L. Forman, R. Stiehm, N. Engl. J. Med.
281 (1969) 926 – 931.
462. A. S. Shigeoka, R. T. Hall, H. R. Hill, Lancet 1
(1978) 636 – 638.
463. E. J. Ziegler, J. A. McCutchan, J. Fierer, M. P.
Glauser et al., N. Engl. J. Med. 307 (1982)
1225 – 1230.
464. J.-D. Baumgartner, M. P. Glauser, J. A.
McCutchan, E. J. Ziegler et al., Lancet 2
(1985) 59 –63.
465. T. Calandra, J. Schellekens, J. Verhoef, M. P.
Glauser, Program and Abstracts for the 4th
International Symposium on Infections in the
Immunocompromised Host, Ronneby Brunn,
Sweden, 1986. Abstract No. 128.
466. A. H. Levy, J. Chronic Dis. 15 (1962)
589 – 598.
467. J. Bjorkander, C. Cunningham-Rundles, P.
Lundin, R. Olsson et al., Am. J. Med. 84
(1988) 107 – 111.
468. R. S. Lane, Lancet 2 (1983) 974 – 975.
469. H. D. Ochs, S. H. Fischer, F. S. Virant, M. L.
Lee et al., Lancet 1 (1985) 404 – 405.
470. C. L. S. Leen, P. L. Yap, G. Neill, D. B. L.
McClelland et al., Vox Sang. 50 (1986)
26 – 32.
471. B. Cuthbertson, R. J. Perry, P. R. Foster, K. G.
Reid, J. Infect. 15 (1987) 125 – 133.
472. C. C. Wood, A. E. Williams, J. G. McNamara,
J. A. Annunziata et al., Ann. Intern. Med. 105
(1986) 536 – 538.
473. Y. Hamamoto, S. Harada, N. Yamamoto, Y.
Uemura et al., Vox Sang. 53 (1987) 65 – 69.
Immunotherapy and Vaccines 79
474. Morb. Mort. Wkly. Rep. 36 (1987) 728 – 729.
475. A. W. Burks, H. A. Sampson, R. H. Buckley,
N. Engl. J. Med. 314 (1986) 560 – 564.
476. R. Apfelzweig, D. Piskiewicz, J. A. Hooper, J.
Clin. Immunol. 7 (1987) 46 – 50.
477. R. Paganelli, I. Quinti, G. P. D’Offizi, C.
Papetti et al., Vox Sang. 51 (1986) 87 – 91.
478. J. C. Barton, G. A. Herrera, J. H. Galla, L. F.
Bertoli et al., Am. J. Med. 82 (1987) 624 – 629.
479. A. G. Brox, D. Cournoyer, M. Sternbach, G.
Spurll, Am. J. Med. 82 (1987) 633 – 635.
480. R. Schroff, K. Foon, S. Beatty, R. Oldham et
al.: “Human anti-murine immunoglobulin
responses in patients receiving monoclonal
antibody therapy,” Can. Res. 45 (1985)
879 – 885.
481. M. Nose, H. Wigzell: “Biological significance
of carbohydrate chains on monoclonal
antibodies,” Proc. Natl. Acad. Sci. U.S.A. 80
(1983) 6632 – 6636.
482. E. Englemann, S. Foung, J. Larrick, A
Raubitschek (eds.): Human Hybridomas and
Monoclonal Antibodies, Plenum Press, New
York 1985.
483. A. Strelkelkaus (ed.): Human Hybridomas;
Diagnostic and Therapeutic Applications,
Marcel Dekker, New York 1986.
484. D. Kozbor, J. Roder, T. Chang, Z. Stplewski et
al.: “Human antitetanus toxoid monoclonal
antibody secreted by EBV-transformed human
B cells fused with murine myeloma,”
Hybridoma 1 (1982) 323 – 328.
485. J. Larrick, K. Truitt, A. Raubitschek, G. Senyk
et al., “Characterization of human hybridomas
secreting antibody in tetanus toxoid,” Proc.
Natl. Acad. Sci. U.S.A. 80 (1983) 6376 – 6380.
486. N. Chiorazzi, R. Wasserman, H. Kunkel: “Use
of Epstein-Barr virus-transformed B cell lines
for the generation of
immunoglobulin-producing human B cell
hybridomas,” J. Exp. Med. 156 (1982)
930 – 935.
487. F. Gigliotti, R. Insel: “Protective human
hybridoma antibody to tetanus toxin,” J. Clin.
Invest. 70 (1982) 1306 – 1309.
488. L. Olsson, T. Mazauric, J. Vincent-Falquet, J.
Armand: “A human monoclonal antibody
specific for tetanus toxoid,” Dev. Biol. Stand.
57(1984) 87 – 91.
489. S. Tsuchiya, S. Yokoyama, O. Yoshie, Y. Ono:
“Production fo diphtheria antitoxin antibody in
Epstein-Barr virus induced lymphoblastoid
cell lines.” J. Immunol. 124 (1980)
1970 – 1976.
80 Immunotherapy and Vaccines
490. E. Ziegler, J. McCutchan, J. Fierer et al.:
“Treatment of Gram-negative bacteremia and
shock with human anti-serum to a mutant
Escherichia coli,” N. Engl. J. Med. 307 (1982)
1225 – 1230.
491. W. Bogard, E. Hornberger, P. Kung in E.
Englemann, S. Foung, J. Larrick, A.
Raubitscheck (eds.): “Production and
characterization of human monoclonal
antibodies against Gram-negative bacteria,”
Human Hybridomas and Monoclonal
Antibodies, Plenum Press, New York 1985,
pp. 95 – 112.
492. N. Teng, H. Kaplan, J. Hebert et al.:
“Protection against Gram-negative bacteremia
and endotoxemia with human monoclonal IgM
antibodies,” Proc. Natl. Acad. Sci. U.S.A. 82
(1985) 179 – 184.
493. J. Larrick, M. Jahnsen, G. Senyk, S. Weiss et
al. in H. Friedman (ed.): Production of human
monoclonal antibodies recognizing
cross-reactive determinants on
lipopolysaccharides,” The Immunology and
Immunopharmacology of Endotoxins, 1986,
pp. 75 – 81.
494. M. Pollack, A. Raubitschek, J. Larrick:
“Cross-reactive human monoclonal antibodies
that recognize conserved epitopes in the
core-lipid A region of endotoxin,” J. Clin.
Invest. 79 (1987) 1421 – 1430.
495. J. Baumgartner, M. Glauser, J. McCutchan, E.
Zeigler et al.: “Prevention of gramnegative
shock and death in surgical patients by
antibody to endotoxin core glycolipid.” Lancet
2 (1985) 59 – 63.
496. J. Larrick, S. Hart, D. Lippman, M.
Glembourtt et al. in A. Strelkelkaus (ed.):
“Generation and characterization of human
monoclonal anti-Pseudomonas aeruginosa
antibodies,” Human Hybridomas; Diagnostic
and Therapeutic Applications, Marcel Dekker,
New York 1986, pp. 65 – 80.
497. J. Larrick, B. Dyer, G. Senyk et al. in E.
Engleman, S. Foung, J. Larrick, A.
Raubitschek (eds.): “In vitro expression of
human B cells for the production of human
monoclonal antibodies,” Human Hybridomas
and Monoclonal Antibodies, Plenum Press,
New York 1985, pp. 149 – 165.
498. K. Hunter, Jr., G. Fischer, V. Hemming, S.
Wilson et al.: “Antibacterial activity of a
human monoclonal antibody to Haemophilus
influenzae type B capsular polysaccharide,”
Lancet 2 (1982) 789 – 799.
499. H. Peltola, H. Kayhty, M. Virtanen, P. Makela:
“Prevention of Hemophilus influenza type b
bacteremic infections with the capsular
polysaccharide vaccine,” N. Engl. J. Med. 310
(1984) 1561 –1565.
500. T. Atlaw, D. Kozbor, J. Roder: “Human
monoclonal antibodies against Mycobacterium
Leprae,” Infect. Immun. 49 (1985) 104 – 110.
501. B. Brodeur, L. Lagace, Y. Larose, M. Martin et
al. in B. Schook (ed.): “Mouse-human
myeloma partner for the production of
heterohybridomas,” Monoclonal Antibodies ,
Marcel Dekker, 1986, p. 51.
502. M. Steinitz, S. Tamir, A. Goldfarb: “Human
anti-pneumococci antibody produced by an
Epstein-Barr virus (EBV)-immortalized cell
line,” J. Immunol. 132 (1984) 877 – 882.
503. J. Schwaber, M. Posner, S. Schlossman, H.
Lazarus: “Human-human hybrids secreting
pneumococcal antibodies,” Hum. Immunol. 9
(1984) 137 – 142.
504. D. Emanuel, J. Gold, J. Colacino, C. Lopez et
al.: “A human monoclonal antibody to
cytomegalovirus (CMV),” J. Immunol. 133
(1984) 2202 – 2205.
505. C. Amadei, S. Michelson, J. Frot, M. Fruchart
et al.: “Human anticytomegalovirus (CMV)
immunoglobulins secreted by
EBV-transformed B-lymphocytes cell lines,”
Dev. Biol. Stand. 57 (1984) 283 – 286.
506. K. Burnett, J. Leung, J. Marinis in E.
Engleman, S. Foung, J. Larrick, A.
Raubitschek (eds.): “Human monoclonal
antibodies to defined antigens,” Human
Hybridomas and Monoclonal Antibodies,
Plenum Press, New York 1985, pp. 113 – 133.
507. Y. Ichimori, K. Sasano, H. Itoh, S.
Hitosumachi et al.: “Establishment of
hybridomas secreting human monoclonal
antibodies against tetanus toxin and hepatitis B
virus surface antigen,” Biochem. Biophys. Res.
Commun. 129 (1985) 26 – 38.
508. E. Stricker, R. Tiebout, P. Lelie, W.
Zeijlemaker: “A human monoclonal IgG
antihepatitis B surface antibody; production,
properties and applications,” Scand. J.
Immunol. 22 (1985) 337 – 345.
509. R. Beasley et al.: “Hepatitis B immune
globulin (HBIG) efficacy in the interruption of
perinatal transmission of hepatitis B virus
carrier state,” Lancet 2 (1981) 388 – 393.
510. S. Foung et al.: “Human monoclonal
antibodies neutralizing Varicella Zoster virus,”
J. Infect. Dis. 52 (1985) 280 – 285.

511. L. Evans, J. Homsy, W. Morrow, I. Gaston et
al.: “Human monoclonal antibody directed
against gag gene products of the human
immunodeficiency virus,” J. Immunol. 140
(1988) 941 – 943.
512. B. Banapour et al.: “Characterization and
epitope mapping of a human monoclonal
antibody reactive with the envelope
glycoprotein of human immunodeficiency
virus,” J. Immunol. 138 (1987) 4027 – 33.
513. A. Rosen, K. Persson, G. Klein: “Human
monoclonal antibodies to a genus-specific
chlamydial antigen, produced by
EBV-transformed B cells,” J. Immunol. 130
(1983) 2899 – 2902.
514. S. Koizumi, S. Fujiwara, H. Kikuta et al.:
“Production of human monoclonal antibodies
against Epstein-Barr virus specific antigens by
the virus immortalized lymphoblastoid cell
lines,” Virology 150 (1986) 161 – 170.
515. J. Seigneurin et al.: “Herpes simplex virus
glycoprotein D; human monoclonal antibody
produced by bone marrow cell line,” Science
221 (1983) 173 – 175.
516. Y. Masuho et al.: “Generation of hybridomas
producing human monoclonal antibodies
against herpes simplex virus after in vitro
stimulation,” Biochem. Biophys. Res.
Commun. 135 (1986) 495 – 505.
517. L. Evans, C. Maragos, J. May: “Human
lymphoblastoid cell lines established from
peripheral blood lymphocytes secreting
immunoglobulins directed against herpes
simplex virus,” Immunol. Lett. 8 (1984)
39 – 50.
518. S. Matsushita, M. Robert-Gurhoff, J. Trepel, J.
Cossman et al.: “Human monoclonal
Immunotherapy and Vaccines 81
antibodies directed against an envelope
glycoprotein of human T-cell leukemia virus
type 1,” Proc. Natl. Acad. Sci. U.S.A. 83
(1986) 2672 – 2677.
519. C. Croce, A. Linnenbach, W. Hall, Z.
Steplewski et al.: “Production of human
hybridomas secreting antibodies to measles
virus,” Nature (London) 288 (1980) 488 – 489.
520. R. Ritts, Jr., A. Ruiz-Arguelles, K. Weyl et al.:
“Establishment and characterization of a
human non-secreting plasmacytoid cell line
and its hybridization with human B cells,” Int.
J. Cancer 31 (1983) 133 – 151.
521. J. Hilfenhaus, E. Kanzy, R. Kohler, W.
Willems: “Generation of human anti-rubella
monoclonal antibodies from human
hybridomas constructed with antigen-specific
Epstein-Barr virus transformed cell lines,”
Behring Inst. Mitt. 80 (1986) 31 – 40.
522. F. van Meel, P. Steenbakkers, J. Oomen:
“Human and chimpanzee monoclonal
antibodies,” J. Immunol. Methods 80 (1985)
267 – 280.
523. D. Crawford et al.: “Production of human
monoclonal antibody to X 31 influenze virus
nucleoprotein,” J. Gen. Virol. 64 (1983)
697 – 700.
524. R. Schmidt-Ullrich, J. Brown, R. Whittle, P.
Lin: “Human-human hybridomas secreting
monoclonal antibodies to the M.W. 195 000
Plasmodium falciparum blood stage antigen,”
J. Exp. Med. 163 (1986) 179 – 189.
525. R. Udomsangpetch et al.: “Human monoclonal
antibodies to Pf-155, a major antigen of
malaria parasite Plasmodium falciparum,”
Science 231 (1986) 55 – 59.
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Inclusion Compounds Save title to My Profile
Standard Article Topics
Email this page
Jerry L. Atwood1 Print this page
1Department of Chemistry, University of Alabama, Tuscaloosa,
SEARCH THIS TITLE
Alabama, United States

Copyright © 2002 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved. Advanced Product Search
DOI: 10.1002/14356007.a14_119 Search All Content
Article Online Posting Date: June 15, 2000 Acronym Finder

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Introduction
2. Classification
2.1. Inorganic Hosts
2.2. Organic Hosts
2.3. Biochemical Inclusion
3. Uses
3.1. Zeolites
3.2. Cyclodextrins

[Top of Page]

1. Introduction
Inclusion compounds are those formed between two substances that are intimately linked, but not with covalent bonds. One
substance is regarded as the host and the other as the guest. The term clathrate was originally used to describe the situation
in which the guest is completely encapsulated by the host. The first clathrate structures were elucidated in the 1940s and
current understanding dates from that time. Other aspects of the chemistry of inclusion compounds include molecular
recognition, supramolecular chemistry, and the chemistry of non-covalent bonds.

Although the field has developed rapidly since the fundamental ideas were developed in the 1940s, it is possible to find much
earlier reports of various types of inclusion compounds. For example, the chlorine clathrate hydrate was described by
FARADAY in 1923, graphite intercalates date from 1841, and cyclodextrins were reported in 1891. The early results were not
always well understood, because inclusion compounds are often non stoichiometric.

[Top of Page]

2. Classification
2.1. Inorganic Hosts
Werner Compounds. Werner compounds are named after the coordination chemist who made major contributions to the
field around 1900. The host molecules have the formula MX2A4, where M is a divalent transition metal cation, X is an anionic
ligand, and A is a neutral ligand, usually a substituted pyridine. A typical example is Ni (NCS)2(4-methylpyridine)4. The
structure shows octahedral coordination about the metal with a trans arrangement of the NCS– ligands. The pyridine groups
are arranged in a propeller fashion with an angle of twist of about 50° from a coplanar arrangement. In the solid state the
molecules pack in layers with the guest molecules included between the layers in a space bounded by the NCS– ligands. The
cavities occupied by the guests are connected by two-dimensional windows in a fashion similar to that found in zeolites.

Werner compounds were used to separate xylene isomers in the late 1950s, but the problems associated with handling
crystalline solids prevented commercialization of this process [1].

Hofmann-Type Compounds. A typical Hofmann compound has the formula Ni (NH3)2· Ni (CN)4· 2 C6H6. In the structure of
this compound there are two types of nickel atom, one that exhibits square planar four-coordination and one that has an
octahedral arrangement of nitrogen atoms around the nickel atom. The host possesses a layered structure with the ammonia

page 1 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
ligands protruding into the layers. The benzene guest molecule is completely enclosed, as is shown in Figure 1.

Figure 1. Hofmann's benzene clathrate [1]

⊕ = Ni(D4n); = Ni(On); = C; • = N; =H

Because of the limitations of space, only molecules of very specific shape and size are included. There are many ways to
modify Hofmann compounds to allow the inclusion of different guests.

Gas Clathrate Hydrates. These substances are based on the structure of ice. Small molecules, usually gases, are included
in voids in the hydrogen-bonded framework of water molecules. The clathrate hydrates have a geometrical similarity to
zeolites in that both are based on three-dimensional four-connected networks. The guests range in size from argon to
dioxane. Gas clathrate hydrates have been used for the separation of gases.

This type of host lattice is also formed by quaternary ammonium salt hydrates and alkylamine hydrates. Many of the gas
clathrate hydrates possess structures which are thermodynamically more stable than that of ice. Several have melting points
in the range 0 – 15 °C and one is reported to melt at 31.5 °C. Applications in heat storage have been sought but not realized.

With few exceptions, gas clathrate hydrates crystallize in one of two structures, I and II. The smaller guests belong to
structure I whereas the larger prefer structure II. The ideal unit cell content for structure I is 6 X · 2 Y · 46 H2O and for
structure II, 8 X · 16 Y · 136 H2O, where X refers to guests in 14-hedra and higher, and Y to those in 12-hedra. The
polyhedra are illustrated in Figure 2. Both X and Y type spaces can be filled by a small molecule, such as methane, which
forms the clathrate 8 CH4· 46 H2O. This clathrate hydrate, which has a melting point well above that of ice, is believed to be
responsible for blocking natural gas pipelines at temperatures above 0 °C in some cases.

Figure 2. Illustration of the void spaces in clathrate hydrates

a) 12-hedra; b) 14-hedra; c) 15-hedra; d) 16-hedra

Zeolites ( Zeolites) are porous tectosilicates with formulas such as Ca4(Al8Si4O12) · 24 H2O or (Na2,Ca,Mg)29(Al58Si
134O384) · 240 H2O. There are about sixty naturally occurring framework structures and a large number have also been
synthesized. Each structure has a unique system of channels and cavities such as that shown in Figure 3, which places the
total emphasis on the void system, rather than the covalent network of the host substance. The cations are usually
exchangeable and share cavities along with the guests.

Figure 3. Channel structure of zeolite ZSM-5

The cavities are rigid and adsorption is controlled by the size of the channels and by the guests that inhabit them.

The channel system may be classified as one-, two-, or three-dimensional. A one-dimensional system arises when diffusion
between channels is not possible because of size restrictions. The most important windows between channels are composed
of 8-, 10-, and 12-rings. A typical window opening for an 8-ring is 420 pm. A representation of the window concept is
presented in Figure 4, which emphasizes the 10-ring opening of the zeolite ZSM-5. This opening allows the entrance of
normal alkanes and simple aromatic hydrocarbons. The opening is so narrow, however, that p-xylene (of critical dimension
630 pm) may be differentiated from o-xylene (critical dimension 690 pm), while molecules such as pentamethylbenzene
(critical dimension 780 pm) are totally excluded. This aspect will be emphasized in Section Uses.

Figure 4. Representation of the openings between cavities in zeolites

page 2 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Applications in the area of electronic devices such as semiconductors will probably be realized, but have not reached
commercialization as yet [2].

Intercalates. Intercalation compounds are formed by the insertion of guest species into a host lattice. Graphite is the most
common substance which possesses this property ( Carbon – Graphite Compounds). The process of insertion is one in
which the host matrix remains essentially unchanged with regard to structure and composition, even though the layers may
often be forced far apart.

There are three types of hosts from a structural standpoint, one-, two-, and three-dimensional. Graphite represents the two-
dimensional type.

Hosts with special conductivity properties are finding applications in energy storage, and new analytical techniques based on
intercalation are being developed.

Other Inorganic Hosts. Of the many new types of inorganic hosts, two have potential applications which are under
investigation. Cyclophosphazenes are paddle-wheel shaped molecules which can entrap long-chain guest molecules. The
specific orientation of molecules in the channels has led to interesting solid-state polymerization results.

The liquid inclusion complexes referred to as liquid clathrates are based on salts which can entrap aromatic hydrocarbons.
Applications in the separation of aromatic from aliphatic hydrocarbons, and for the separation of the para and meta isomers
of aromatic compounds are under investigation.

2.2. Organic Hosts

Crown Ethers. Crown ethers are cyclic polyethers which were first recognized as a new class of macrocyclic compounds in
1967 ( ). Crown ethers have been typically used for the complexation of cations, but it is also possible to complex neutral
molecules and anions with many members of the series.

The crown ethers are essentially two-dimensional complexing agents. Three-dimensional, oligocyclic analogues of crown
ethers, the cryptands, are also known [1].

The crown ethers represent an important class of synthetic macrocycles. Applications are being sought in pharmaceuticals,
metal ion removal and recovery, and in chemical sensors [2].

Cyclodextrins. Cyclodextrins are cyclic oligosaccharides which are naturally formed by the enzymatic degradation of starch.
The three types most commonly obtained are alpha- (6 glucose units), beta- (7 glucose units), and gamma-cyclodextrin (8
glucose units). A schematic representation of these structures is shown in Figure 5. Since the subunit of the structures
(glucose) is rigid, the cyclodextrin molecule possesses a cavity, even in the absence of a guest. The approximate dimensions
of the cavities, given in Figure 5, are comparable to the molecular dimensions of many simple organic molecules. The fit of
guest in host is illustrated in Figure 6 for 4-iodoaniline and alpha-cyclodextrin.

Figure 5. Representations of the structures of alpha-, beta-, and gamma-cyclodextrins (values are given in picometers)

Figure 6. Structure of the inclusion complex of alpha-cyclodextrin with 4-iodoaniline using space-filling models

The cyclodextrins have hydrophobic interiors and hydrophilic exteriors, which leads to good water solubility. The exterior is
easily modified chemically.

Other Organic Hosts. Urea was found to form inclusion complexes in 1940. Here the host is the lattice of urea (or thiourea)
and the guest is typically a long-chain hydrocarbon. Since urea is such an inexpensive host, attempts have been made to
exploit its inclusion properties. In particular, its use in the separation of benzene and cyclohexane from n-heptane has been

page 3 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
investigated, but not commercialized.

Some phenolic compounds, such as hydroquinone, phenol, and Dianin's compound have interesting inclusion properties.

Considerable attention is now directed towards the calixarenes [3]. These molecules have cavities of molecular size for many
simple organic molecules and they can be readily and economically synthesized. A schematic view of the calixarene
structure and the crystal structure of an inclusion complex are shown in Figure 7A and B.

Figure 7. A) Schematic view of the structure of a calix(n)-arene (for n = 4, 6, or 8, the molecules are readily accessible)

B) The methyl ether of calix(4) arene with included toluene and complexed sodium ion

Other promising organic hosts under active investigation include deoxycholic acid, perhydrotriphenylene, cyclotriveratrylene,
and tri-o-thymotide.

2.3. Biochemical Inclusion

Since the processes of biochemical molecular recognition and catalysis are so clearly tied to the principles of inclusion
chemistry, much endeavor is aimed at duplication or improvement of certain aspects of biochemistry. In particular, enzyme
mimics are being sought. These should catalyze reactions specifically and under mild conditions. The most studied enzyme
mimics are the cyclodextrins. An example of a reaction of interest is the chlorination of anisole.

In this reaction the ratio of ortho- to para-isomer is 40 : 60. In the presence of alpha-cyclodextrin the ratio changes to 4 : 96.
The blocking of the ortho position is illustrated in Figure 8. Cyclodextrin also accelerates the reaction by providing a new
reaction pathway.

Figure 8. Schematic view of the structure of the complex of anisole and cyclodextrin

Crown ethers and cryptands have been studied for their action in cation transport. In biochemical systems it is important to
solubilize cations in media of low dielectric constant. Possible applications of this research in the pharmaceutical industry
may be anticipated.

[Top of Page]

3. Uses
3.1. Zeolites
Worldwide sales of naturally occurring zeolites exceeded 250 000 t in 1979, and the market is growing yearly. The major
applications can be divided into the four categories addressed below.

Adsorptive Separation of Hydrocarbons. By far the largest industrial application of inclusion compounds to date involves
the use of zeolites in adsorptive separations ( Adsorption – Oxidic Hydrophilic Adsorbents). The zeolites contain rigid
channels and cavities which are only accessible to molecules of appropriate size and shape. The first industrial application
came in 1962 with the Molex process for the separation of normal alkanes from branched-chain alkanes [4]. The process
uses 5 Å (500 pm) molecular sieves, and only straight-chain molecules can enter the zeolite channels. Branched-chain

page 4 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
alkanes are completely excluded.

The ideas which came to fruition with the Molex process have given rise to the Sorbex family of processes. In these large-
scale separations, a simulated moving-bed technique is used. Sorbex processes are now used in 60 plants worldwide
(Table 1) and produce a total of 4.5×109 t/a of purified products.

Table 1. Licensed sorbex plants [4]

Process Separation Number of plants

Parex p-xylene/C8 hydrocarbons 34

Olex alkenes/alkanes 4
Molex n-alkanes/branched and cyclic
hydrocarbons 16
Sarex fructose/dextrose and
polysaccharides 4
Cresex p- or m-cresol/cresol isomers 1
Cymex p-cymene or m-cymene/cymene
isomers 1

60

Several features of these adsorptive processes are of note. First, in most instances one component is included, but not to the
total exclusion of the others. The xylene isomers can all gain access to the channels, but some are adsorbed more strongly
than others. This might be as a result of their differing basicities. The base strength of aromatic hydrocarbons relative to
hydrogen fluoride [5] (relative basicity) at 0.1 molal are as follows:

toluene 0.63
p-xylene 1.0
o-xylene 1.1
m-xylene 26
mesitylene 13 000

In zeolites, the acidity of the cavities or channels, and thus the strength of adsorption, can be controlled by varying the Si : Al
ratio.

It is also worthwhile to compare adsorptive separations using zeolites to those based on liquid – liquid extraction. Figure 9
shows a typical plot of selectivity versus chain length for an alkane – alkene separation using liquid – liquid extraction. The
solubilities of alkenes and alkanes are similar and separation of their mixtures is ineffective. In Figure 10 an equivalent plot is
shown for an adsorptive separation with zeolites. Separation of the alkanes and alkenes is effected.

Figure 9. Liquid – liquid extraction selectivity of polar solvents in alkane – alkene systems [4]

Figure 10. Adsorption selectivity of Olex adsorbent in alkane – alkene systems [4]

Purifications of Gases and Liquids. Zeolites (molecular sieves) have long been used to remove water from organic
solvents. The molecular sieve is selected with a pore size that excludes the organic molecule. Regeneration is accomplished
by heating the molecular sieve under vacuum to remove the guest water molecules. This same technology can be used for
the purification of solvents from a range of contaminants.

Zeolites have been used to remove water from natural gas and air. They also find application in the sweetening of LPG
(liquified petroleum gas) where they both dehydrate and remove sulfur-containing compounds. Zeolites are also used to

page 5 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
remove traces of water which may cause problems in sealed systems, such as closed cycle refrigerants and the space
between dual-pane windows.

A recent application of molecular sieves is in exhaust systems for automobiles. Cooling of the system, particularly the
catalytic converter, results in condensation of moisture which ultimately leads to rusting. Appropriate placement of a packet of
zeolitic material in the system affords adsorption of the water, and when the engine is working the heat of the exhaust gases
removes the water molecules regenerating the zeolite.

Catalytic Cracking of Hydrocarbons. The use of zeolites in catalysis represents a major share of the total zeolite market.
The largest industrial process of this type is the catalytic cracking of hydrocarbons. The catalysts contain up to 40 % zeolite,
often lanthanide-ion exchanged, on a clay matrix. Such catalysts improve the yield and octane number of gasoline and
produce less coke than other methods. Zeolites also find extensive use in hydrocracking and in alkane isomerization. Their
application in shape-selective catalysis can be expected to grow [3].

Ion Exchange. As has been noted in previous sections, it is possible to exchange the cations in zeolites for others. In
general, more highly charged cations replace those having a lower charge. The major application of this is in the production
of detergents. The primary function of phosphates in detergents is the complexation of Mg2+ and Ca2+ ions (water softening).
Since the trend is now away from the use of phosphates, zeolites such as Type A, (Na2) (Al2SiO7) · 4.5 H2O, are being used
as replacements. However, Type A zeolite cannot remove highly hydrated Mg2+ ions since its pore size is too small.
Moreover, zeolites have low solubility so that their use is expected to decline as the demand for liquid detergents increases.
A secondary application is in point-of-use water filters. Not only is water softening accomplished, but also heavy metal ions
may be removed.

3.2. Cyclodextrins
Pharmaceuticals. Extensive studies have shown that beta-cyclodextrin is nontoxic. The first studies on humans were
performed in the early 1970s. Since cyclodextrins are oligosaccharides, it is believed that their structures decompose readily
under physiological conditions.

The numerous patents for pharmaceutical applications of cyclodextrins attest to the level of interest in this area [6].
Cyclodextrins are being used as drug complexation agents and as drug additives. Complexation of oral drugs provides many
potential benefits: (1) liquid drugs can be transferred into crystalline materials which can then be manufactured as tablets, (2)
compounds which react with one another may be formulated together, (3) smell and taste problems may be overcome, and
(4) cyclodextrin – drug complexes are not hygroscopic. More generally, both oral and liquid drugs show enhanced chemical
and physical stability with respect to oxidation by air, sensitivity to light, rate of disproportionation or polymerization, and
losses due to evaporation. Furthermore, the solubility of drugs may be greatly enhanced. A general scheme for drug
complexation and absorption is given in Figure 11.

Figure 11. Representation of the dissolution – dissociation – absorption process for a drug – cyclodextrin complex, F.CD
[6]

Food Additives. Complexes of cyclodextrins with food flavor molecules show the same enhanced physical and chemical
properties as do those of drugs. In addition, significant improvements in the formulation of the additives make it easier to
protect against contamination.

Beta-cyclodextrin has a sweet taste, comparable to that of sucrose. An interesting application is in the complexation of
aspartame. The complex is marketed in Japan under the name PalSweet and is said to have no bitter aftertaste.

Although it is well established that the cavity of beta-cyclodextrin can complex molecules such as aspartame, vanillin, and
menthol, the complexation of molecules such as naringin, a bitter component of grapefruit juice, is doubtful.

page 6 of 7
Inclusion Compounds : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Pesticides. The advantages of the use of cyclodextrin complexes in the pesticide industry are like those for drugs and food
additives, but with two additional features. First, the pesticide is enclosed in a hydrophilic cover (the cyclodextrin). This
means that absorption into the hydrophilic environment in the intestine of a pest is increased, whereas contact absorption
through the hydrophobic surface of the insect exoskeleton is decreased. This in turn means that these complexes can be
used to selectively attack herbivorous pests, while other insects remain largely unaffected by contact absorption. The second
feature of cyclodextrin – pesticide inclusion compounds concerns the aspect of timed-release activity. A complex may be
applied to dry soil and only be released when it rains.

[Top of Page]

References
1. J. L. Atwood, J. E. D. Davies, D. D. MacNicol (eds.): Inclusion Compounds, vols. 1 – 3, Academic Press, Orlando Florida
1983, 1984.
2. J. L. Atwood (ed.): Perspectives on Inclusion Phenomena and Molecular Recognition, Plenum, New York 1989.
3. J. L. Atwood, J. E. D. Davies, D. D. MacNicol (eds.): Inclusion Compounds, vols. 4 – 5, Oxford University Press, Oxford
1990.
4. D. B. Broughton, Sep. Sci. Technol. 19 (1984, 1985) 723.
5. H. A. Zinnen, S. H. Hobbs, S. A. Gembicki, Sep. Sci. Technol. 19 (1984 – 85) 737.
6. J. Szejtli: Cyclodextrin Technology, Kluwer, Dordrecht 1988.
[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 7 of 7
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Indicator Reagents Save title to My Profile
Standard Article Topics
Email this page
Erwin Ross1, Jutta Köthe2, Renate Naumann3, Wolfgang Fischer4, Wilhelm- Print this page
Dietrich Mayer5, Gerhard Wieland6, Ernest J. Newman7, Charles M. Wilson8 SEARCH THIS TITLE
1Merck KGaA, Darmstadt, Federal Republic of Germany
2Merck KGaA, Darmstadt, Federal Republic of Germany
3Merck KGaA, Darmstadt, Federal Republic of Germany Advanced Product Search
4Merck KGaA, Darmstadt, Federal Republic of Germany
Search All Content
5Merck KGaA, Darmstadt, Federal Republic of Germany
Acronym Finder
6Merck KGaA, Darmstadt, Federal Republic of Germany
7Merck Ltd. (formerly British Drug Houses Ltd.), Poole, United
Kingdom
8EM Science, Cincinnati, Ohio, United States

Copyright © 2002 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved.
DOI: 10.1002/14356007.a14_127
Article Online Posting Date: June 15, 2000

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Introduction
2. pH Indicators
2.1. Theory
2.2. Errors
2.3. Dye Classes and Individual pH Indicators
2.4. pH Indicator Papers
2.5. pH Indicators for Nonaqueous Solutions
3. Adsorption Indicators
4. Fluorescent Indicators
5. Chemiluminescent Indicators
6. Redox Indicators
6.1. Standard Potential and rH Value
6.2. Classes of Redox Indicators
7. Metal Indicators
7.1. Volumetry (Complexometry)
7.2. Colorimetry and Photometry
7.3. Structure
7.4. Ligand Classes
7.4.1. Ligands Forming Five-Membered Chelate Rings
7.4.2. Ligands Forming Six-Membered Chelate Rings
8. Analytical Test Kits

[Top of Page]

1. Introduction
Indicator reagents are substances that enable the course of a chemical reaction to be followed or the state of a chemical
system to be characterized. Their names are derived from the areas in which they are used, e.g., pH indicators, metal
indicators, or redox indicators.

Indicator reagents are often used to detect equivalence points during titration. They are also utilized for qualitative,
semiquantitative, or quantitative analysis, particularly of inorganic ions. In this case, indicator reagents can be employed in
the form of individual reagents, as components of a test kit, or as test strips or indicator papers.

page 1 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Indicators that are added directly to a solution are called internal indicators. The concentration of an internal indicator must
be very low compared to that of the substance being examined. Indicators that are not used within the observed system are
called external indicators. These include indicator papers and test strips.

The pH (acid – base) indicators are the oldest known indicator reagents. In 1660, BOYLE recognized that certain plant
extracts could be used as indicators because they changed color on treatment with acid or base. The indicator reagent in
longest use is the dye known as litmus, which is obtained from the orseille lichen (Roccella tinctoria). At an early stage, litmus
was coated onto cellulose to form litmus paper, which is still used widely today.

Indicator reactions that incorporate enzyme-catalyzed stages are not discussed in this article ( ; Enzymes – Amino
Acid Analysis and Protein Sequence Analysis, Enzymes – Site-Specific RNases in RNA Sequence Analysis,
Enzymes – Enzymes in Diagnosis, Enzymes – Enzymes for Food Analysis, [6], [30]).

[Top of Page]

2. pH Indicators
The term “indicator” is often employed without further explanation to refer to pH (acid – base) indicators.

2.1. Theory
The pH indicators are substances (usually organic dyes) that change color within defined pH ranges. According to OSTWALD,
the change in spectral properties as a function of pH is due to dissociation of the indicator molecule on accepting or donating
protons:

(1)

where InA is the indicator acid and InB the indicator base. The absorption spectra of InA and InB differ. If the laws of mass
balance are applied to Equation (1) and the activity coefficients are neglected, then

(2)

(3)

(4)

where KA is the dissociation constant of the indicator and pKA its negative logarithm to the base 10. According to
Equation (4), the ratio [InB]/[InA] is a function of pH. Consequently, the color of an indicator solution depends upon the pH.

If both forms of the indicator exist in equal concentrations, Equation (4) simplifies to

(5)

Indicators that are able to donate or accept several protons have several corresponding pKA values, and the above
equations must to be extended appropriately. Sometimes, for example with phenolphthalein, two pKA values can be used to
derive the following equation:

(6)

The dissociation equilibrium is not always sufficient to describe indicator properties. If, for example, a reaction takes place
prior to dissociation (as is the case with nitrophenols), a new equilibrium constant Ki must be defined:

(7)

where KU represents the equilibrium constant of the reaction product.

The pH value corresponding to the negative logarithm to the base 10 of Ki is defined as the transition point:

(8)

The pKA or pKi value is the pH at which the color change occurs. The color change of a given indicator generally extends
over a pH range. This range is called the transition region or transition interval and is defined by pKi± 1. In this region, colors
of various forms of the indicator coexist. If one form has an absorbance maximum at the wavelength 1 and the other at 2,
then the plot shown in Figure 1 is obtained when the pH is varied. Curve (a) represents the spectrum of the acidic form of the
indicator; curve (b), that of the alkaline form; and curve (c), a mixture of the two. All of the curves intersect at the isosbestic
point (P).

page 2 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 1. Absorption spectrum of a pH indicator as a function of pH (for explanation, see text)

a) Acidic form of indicator; b) Alkaline form of indicator; c) Mixture of acidic and alkaline forms

2.2. Errors
Errors occurring during the use of indicators include the following.

Salt Errors [13]. Neutral salts affect the activity coefficients of both indicator structures to varying extents. At salt
concentrations > 0.2 mol/L, changes in pKi values are generally encountered.

Protein Errors. The presence of charged macromolecules can disrupt the equilibrium because they bind to indicator
molecules of the opposite charge. With some indicators (e.g., Bromophenol Blue), the resulting color shifts can be so
pronounced that they are used for the analytical determination of protein.

Indicator Errors. Because the indicator can accept or donate protons, it may itself cause a pH change (e.g., in weakly
buffered solutions). During titration, the consumption of reagents by the indicator must also be considered.

Alcohol Errors. The addition of alcohols and other solvents to aqueous solutions can alter the dissociation constants of
indicators and, consequently, change the transition point.

2.3. Dye Classes and Individual pH Indicators [14]

The most commonly used pH indicators include azo dyes, nitrophenols, phthaleins, and sulfonphthaleins. Some other acid –
base indicators belong to the classes aniline – sulfonphthaleins, triphenylmethane dyes, or benzaurins.

Variation of the substituents of the basic chromophoric indicator structures yields pH indicators with transition points that
cover the complete pH range from 0 to 14. The indicators are generally used as 0.1 % solutions in water, ethanol, or water –
ethanol mixtures. Typical volumes range between 0.05 and 0.5 mL.

The most commonly used pH indicators are listed in Table 1 in order of increasing pH of their transition points. A graphic
overview of the pH ranges and color changes is given in Figure 2.

Table 1. Commonly used pH indicators

Transition Indicator CAS registry Color change Composition of

range, pH no. indicator solution,
%

0.0 – 2.0 Malachite Green (C.I. [2437-29-8] yellow – green 0.1 (in water)
42 000) blue
0.0 – 2.6 Brilliant Green (C.I. 42 040) [633-03-4] yellow – green 0.1 (in water)

0.1 – 2.3 Methyl Green (C.I. 42 590) [82-94-0] yellow – blue 0.1 (in water)

0.2 – 1.0 Picric acid (C.I. 10 305) [88-89-1] colorless – 0.1 (in 70 % ethanol)
yellow
0.5 – 2.5 Cresol Red [1733-12-6] red – yellow 0.1 (in 20 % ethanol)

0.8 – 2.6 Crystal Violet (C.I. 42 555) [548-62-9] yellow – blue 0.1 (in 70 % ethanol)
violet
1.2 – 2.3 Metanil Yellow (C.I. 13 065) [587-98-4] red – yellow 0.1 (in 20 % ethanol)

1.2 – 2.8 m-Cresol Purple [2303-01-7] red – yellow 0.04 (in 20 %

ethanol)
1.2 – 2.8 Thymol Blue [76-61-9] red – yellow 0.1 (in 96 % ethanol)

1.2 – 2.8 p-Xylenol Blue [125-31-5] red – yellow 0.1 (in 50 % ethanol)

1.2 – 2.8 Thymol Blue sodium salt [62625-21-2] red – yellow 0.05 (in water)

1.4 – 2.2 Quinaldine Red [117-92-0] colorless – pink 0.1 (in 60 % ethanol)

1.4 – 2.6 Tropaeolin OO [3012-37-1] red – yellow 0.1 (in water)

2.0 – 4.4 2,6-Dinitrophenol [573-56-8] colorless – 0.1 (in 70 % ethanol)

yellow
2.2 – 3.2 Phloxine B (C.I. 45 410) [18472-87-2] colorless – 0.1 (in water)
purple

page 3 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2.8 – 4.0 2,4-Dinitrophenol [51-28-5] colorless – 0.1 (in 70 % ethanol)
yellow
2.9 – 4.0 4-
Dimethylaminoazobenzene
(Methyl Yellow, C.I. [60-11-7] red – orange 0.1 – 0.5 (in 90 %
11 020) yellow ethanol)
3.0 – 4.5 Bromochlorophenol Blue [2553-71-1] yellow – blue 0.1 (in 20 % ethanol)
violet
3.0 – 4.6 Bromophenol Blue [115-39-9] yellow – blue 0.1 (in 20 % ethanol)
violet
3.0 – 4.6 Bromophenol Blue sodium [34725-61-6] green yellow – 0.05 (in water)
salt blue violet
3.0 – 5.2 Congo Red [573-58-0] blue – orange 0.2 (in water)
red
3.1 – 4.4 Methyl Orange (C.I. 13 025) [547-58-0] red – yellow 0.04 (in 20 %
orange ethanol)
3.5 – 5.8 2,5-Dinitrophenol [329-71-5] colorless – 0.1 (in 70 % ethanol)
yellow
3.7 – 5.0 1-Naphthyl Red [131-22-6] red – yellow 0.1 (in 70 % ethanol)

3.8 – 5.4 Bromocresol Green [76-60-8] yellow – blue 0.1 (in 20 % ethanol)

3.8 – 5.4 Bromocresol Green sodium [62625-32-5] yellow – blue 0.05 (in water)
salt
4.3 – 6.3 Alizarin (Red) S (C.I. [130-22-3] yellow – pink 0.1 (in water)
58 005)
4.4 – 6.2 Methyl Red (C.I. 13 020) [493-52-7] red – yellow – 0.1 (in 96 % ethanol)
orange
4.5 – 6.2 Methyl Red sodium salt [845-10-3] red – yellow – 0.05 (in water)
orange
4.7 – 6.3 Bromophenol Red [2800-80-8] yellow – purple 0.1 (in 20 % ethanol)

4.8 – 6.4 Chlorophenol Red [4430-20-0] yellow – purple 0.1 (in 20 % ethanol)

5.0 – 7.2 Hematoxylin (C.I. 75 290) [517-28-2] yellow – violet 0.05 (in 96 %
ethanol)
5.0 – 8.0 Litmus (C.I. 1242) [573-56-8] red – blue 4 (in water)

5.2 – 6.8 Bromocresol Purple [115-40-2] yellow – purple 0.1 (in 20 % ethanol)

5.4 – 7.5 4-Nitrophenol [100-02-7] colorless – 0.2 (in 96 % ethanol)

yellow
5.7 – 7.5 Bromoxylenol Blue [40070-59-5] yellow – blue 0.1 (in 96 % ethanol)

5.8 – 7.2 Alizarin (C.I. 58 000) [72-48-0] pale yellow – 0.5 (in 96 % ethanol)
violet red
5.8 – 7.6 Bromothymol Blue [76-59-5] yellow – blue 0.1 (in 20 % ethanol)

5.8 – 7.6 Bromothymol Blue sodium [34722-90-2] yellow – blue 0.05 (in water)
salt
6.0 – 7.0 Nitrazine Yellow (C.I. [5423-07-4] brownish 0.05 (in 70 %
14 890) yellow – blue ethanol)
violet
6.4 – 8.2 Phenol Red [143-74-8] yellow – red 0.1 (in 20 % ethanol)
violet
6.5 – 8.0 Phenol Red sodium salt [34487-61-1] yellow – red 0.05 (in water)
violet
6.5 – 8.5 Cresol Red [1733-12-6] orange – purple 0.1 (in 20 % ethanol)

6.6 – 8.6 3-Nitrophenol [554-84-7] colorless – 0.3 (in 96 % ethanol)

yellow
6.8 – 8.0 Neutral Red [553-24-2] blue red – 0.1 (in 70 % ethanol)
orange yellow
7.1 – 8.3 1-Naphtholphthalein [1301-55-9] pink brown – 0.1 in (96 % ethanol)
blue
7.4 – 9.0 m-Cresol Purple [2303-01-7] yellow – purple 0.04 (in 20 %
ethanol)
7.8 – 9.5 Thymol Blue sodium salt [62625-21-1] green – blue 0.05 (in water)

8.0 – 9.6 Thymol Blue [76-61-9] yellow – blue 0.1 (in 96 % ethanol)
p-Xylenol Blue [125-31-5] yellow – blue 0.1 (in 50 % ethanol)

page 4 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
8.0 – 9.6
8.2 – 9.5 o-Cresolphthalein [596-27-0] colorless – red 0.02 (in 50 %
violet ethanol)
8.2 – 9.8 Phenolphthalein [81-90-3] colorless – pink 0.1 (in 96 % ethanol)

9.3 – 10.5 Thymolphthalein [125-20-2] colorless – blue 0.1 (in 50 % ethanol)

9.4 – 12.0 Alizarin (Red) S (C.I. [130-22-3] brown orange – 0.1 (in water)
58 005) violet
10.2 – 11.8 Alizarin (C.I. 58 000) [72-48-0] brownish red – 0.5 (in 96 % ethanol)
violet
10.2 – 12.1 Alizarin Yellow GG (C.I. [584-42-9] pale yellow – 0.1 (in water)
14 025) brownish yellow
11.5 – 13.0 Alkali Blue (C.I. 42 765) [568-02-5] blue – pink 0.1 (in 90 % ethanol)
violet
11.6 – 13.0 Epsilon Blue [73904-21-9] orange – violet 0.1 (in water)
11.7 – 13.2 Malachite Green (C.I. [569-64-2] blue – colorless 0.1 (in water)
42 000)
11.7 – 14.0 Indigo Carmine (C.I. 73 015) [860-22-0] blue – yellow 0.25 (in 50 %
ethanol)
12.0 – 14.0 Acid Fuchsin (C.I. 42 685) [3244-88-0] red – colorless 0.1 (in water)

Figure 2. Color changes of commonly used pH indicators (courtesy of Merck)The pH ranges and color shades shown are
approximations.

Azo Dyes. By using 4-dimethylaminoazobenzene as an example, the chromophoric –N=N– group can be seen to display
antiauxochrome (electron acceptor) properties. In combination with auxochromic (electron-donating) groups and the -
electron system of the benzene ring, mesomers can be formed.

The yellow – red color transition observed on addition of acid is caused by stabilization of the mesomer system on
transformation to the dye ion.

Water-soluble azo indicators can be obtained by introducing the sodium salts of carboxylic or sulfonic acids as substituents.

page 5 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Azo dyes with specified transition ranges and solubility properties can be produced by changing the functionality and position
of these groups on the aromatic ring, as well as by replacing the N-methyl group with other N-alkyl residues. An example is:

4-Dimethylaminoazobenzene (Methyl Yellow, Butter Yellow) is soluble in ethanol and employed along with Methylene Blue
for the titration of carbonates (see Mixed Indicators.).

Methyl Orange (sodium-4-dimethylaminoazobenzene 4′-sulfonate) is soluble in water and is one of the most frequently
employed acid – base indicators. It is generally used for titration of strong acids and weak bases. The drawback of this
indicator is its weak color transition (red orange orange yellow), which makes detection of the transition point
difficult.

4-Dimethylaminoazobenzene-4′-sulfonamide has a direct transition from yellow to red.

Methyl Red (4-dimethylaminoazobenzene-2'-carboxylic acid) is also frequently used for titrating strong acids and weak
bases. Its color transition is more easily detected than that of Methyl Orange.

Nitrophenols. All of the nitrophenols used as pH indicators show a transition from colorless to yellow, which occurs when the
colorless indicator acid is converted into the yellow aci form:

The bathochromic effect here is also attributed to improved resonance stabilization of the resulting dye ion.

The transition range of nitrophenols is governed by the number and position of nitro groups. Mononitrophenols show a
transition in the pH range 5.0 – 8.6; dinitrophenols between pH 2.0 and 5.8; and picric acid between pH 0.2 and 1.0 (see
Table 1).

Phthaleins. Perhaps the best known phthalein indicator is phenolphthalein. The molecular structure of phenolphthalein and
its pH dependence are representative of all phthaleins:

page 6 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

I Phenolphthalein (colorless lactone, acidic)

II Resonance-stabilized anion (red, alkaline)
III Carbinol base (colorless, strongly alkaline)

In acidic media, phthalein derivatives are colorless because the central tetrahedral carbon atom is unable to participate in
any resonance structure. In alkaline media, a planar, resonance-stabilized anion is formed that corresponds to a
triphenylmethane dye. At extremely high pH, nucleophilic substitution of a hydroxyl group on the central carbon atom occurs;
this eliminates the possibility of resonance stabilization and leads to formation of the colorless triphenylmethanol base.

The transition range of phthaleins can also be influenced by the number and properties (electronic and steric) of one or both
phenyl residues. Phenolphthalein and thymolphthalein can be used to titrate weak acids.

Thymolphthalein

Sulfonphthaleins. Sulfonphthaleins are closely related to phthaleins: the carboxylic acid function is merely replaced by a
sulfonic acid group. The various pH-dependent structures are illustrated here with m-Cresol Purple, which has two transition
ranges:

page 7 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

I Red (strongly acidic)

II Yellow (two mesomeric forms, acidic)
III Purple (three mesomeric forms, alkaline)

The alkaline form of sulfonphthaleins is more stable than that of phthaleins; i.e., formation of a triphenylmethanol base and,
hence, loss of color under strongly alkaline conditions does not take place. The sulfonphthaleins have a sharp color transition
in both acidic and alkaline media. Table 2 lists other commonly used sulfonphthalein pH indicators.

Table 2. Important pH indicators from the sulfonphthalein dye series

Phenol Red Cresol Red

Chlorophenol Red Bromocresol Purple

page 8 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Bromophenol Blue Bromocresol Green

Thymol Blue

Bromothymol Blue

Other pH Indicators. If the hydroxyl groups in the 4′- and 4′′-positions of Phenol Red (see Table 2) are replaced by amino
groups, anilinesulfonphthalein [4538-11-8] is obtained:

Anilinesulfonphthalein shows two transition regions: yellow – violet at pH 1.3 – 1.9 and violet – yellow at pH 11.7 – 12.5.
Other anilinesulfonphthaleins are derived from this compound.

Benzaurin compounds are very similar to phthaleins in their basic structure and their tendency to be stabilized via
intermediate structures. Because these indicators lack hydrophilic groups, they are not water soluble and hence are used
mostly in nonaqueous systems (see Section pH Indicators for Nonaqueous Solutions).

Benzaurin [569-60-8]

Triphenylmethane dyes are structurally related to benzaurins. Malachite Green and Crystal Violet, the most important of
these dyes, are obtained by the introduction of dimethylamino groups in the 4′- and 4′′-positions of the benzaurin structure.

page 9 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

R = H: Malachite Green
R = N(CH3)2: Crystal Violet

Neutral Red is an azine dye ( Azine Dyes – Diaminophenazine Dyes).

Neutral Red
Triphenylmethane and azine dyes are also used as redox indicators (see Chap. Redox Indicators).

Mixed Indicators. An important criterion for the selection of an indicator is that the transition interval be clear, sharp, and
visually detectable. Ideally, the acidic and alkaline forms should possess complementary colors, e.g., red and green.
Unfortunately, no indicator fulfills this requirement. Furthermore, a mixture of colors exists in the transition range during which
both indicator structures occur simultaneously.

This situation can be improved in two ways. First, a pH-neutral dye can be added to the system to provide a screened
indicator (Table 3). Second, several indicators with similar transition ranges can be combined to form a mixed indicator
(Table 4).

Table 3. Screened pH indicators

Components of screened indicator Parts Transition Color change

mixture

by point, pH Acid Alkali

volume

Dimethylaminoazobenzene (0.1 % in 1 3.3 blue green

ethanol) violet
Methylene Blue (0.1 % in ethanol) 1

Methyl Orange (0.1 % in water) 1 4.1 violet green

Indigo Carmine (0.25 % in water) 1

Methyl Orange (0.2 % in water) 1 4.5 – 4.8 red green

fluorescence
Fluorescein (0.2 % in ethanol) 1

Methyl Red (0.2 % in ethanol) 1 5.4 red violet green

Methylene Blue (0.1 % in ethanol) 1

Neutral Red (0.1 % in ethanol) 1 7.0 violet green

blue
Methylene Blue (0.1 % in ethanol) 1

Phenol Red (0.1 % in ethanol) 1 7.3 green violet

Methylene Blue (0.02 % in water) 1

Phenolphthalein (0.1 % in ethanol) 1 8.9 green violet

page 10 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Methyl Green (0.1 % in ethanol) 2

Phenolphthalein (0.1 % in ethanol) 1 10.0 blue red

Nile Blue (0.2 % in ethanol) 2

Nile Blue (0.2 % in water) 2 10.8 green red brown

Alizarin Yellow (0.1 % in ethanol) 1

Table 4. Mixed pH indicators

Components of mixed indicator Parts Transition Color change

by volume point, pH Acid Alkali

Bromocresol Green (0.1 % in ethanol) 1 4.3 orange green

Methyl Orange (0.02 % in water) 1

Bromocresol Green (0.1 % in ethanol) 3 5.1 wine red green

Methyl Red (0.2 % in ethanol) 1

Bromocresol Green-Na (0.1 % in water) 1 6.1 yellow green blue violet

Chlorphenol Red-Na (0.1 % in water) 1

Bromocresol Purple-Na (0.1 % in water) 1 6.7 yellow violet blue

Bromothymol Blue-Na (0.1 % in water) 1

Bromothymol Blue-Na (0.1 % in water) 1 7.5 yellow violet

Phenol Red-Na (0.1 % in water) 1

Cresol Red-Na (0.1 % in water) 1 8.3 yellow violet

Thymol Blue-Na (0.1 % in water) 3

Thymol Blue (0.1 % in 50 % ethanol) 1 9.0 yellow violet

Phenolphthalein (0.1 % in 50 % ethanol) 3

Phenolphthalein (0.1 % in 50 % ethanol) 2 9.6 slightly pink violet

1-Naphtholphthalein (0.1 % in 50 % ethanol) 1

Phenolphthalein (0.1 % in ethanol) 1 9.9 colorless violet

Thymolphthalein (0.1 % in ethanol) 1

Thymolphthalein (0.1 % in ethanol) 2 10.2 yellow violet

Alizarin Yellow (0.1 % in ethanol) 1

For example, detection of the red – yellow color transition of Methyl Red, Methyl Orange, or 4-dimethylaminoazobenzene is
facilitated by adding Methylene Blue, a pH-independent blue dye. Under acidic conditions, a violet color is formed; on
addition of alkali, a green mixed color is produced. If these azo dye indicators are mixed instead with sulfonphthaleins with
similar transition ranges (e.g., addition of Bromocresol Green to Methyl Red), easily recognizable, sharp color changes result
(yellow – red to blue – green).

Universal Indicators. If several indicators of differing transition regions are combined, mixed indicators are obtained that can
be used over an extremely wide pH range. These “universal indicators” are typically employed in the form of test papers or
test strips for rapid pH determinations (see Section pH Indicator Papers). Defined values of pH are indicated within a large
pH range (usually in about five stages) by a variety of colors. The pH can be quickly determined to within 0.5 pH unit by
comparing the color of the indicator in solution with a reference color scale. Table 5 gives examples of indicators with
extended application ranges.

Table 5. Some pH indicators with extended measuring ranges

page 11 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Name Measuring Color change Number of colors on Accuracy
range, pH evaluation scale

Indicator fluid, 0–5 orange red, ocher, lime 11 ± 0.25 pH

green, light units
pH 0 – 5, Merck green, dark green,
brown green, blue

Universal 4 – 10 pink, yellow, yellow 13 ± 0.25 pH

indicator fluid, green, gray green, units
pH 4 – 10, gray, bluish gray, gray
Merck violet, violet

Indicator fluid, 9 – 13 yellow green, lime 5 ± 0.5 pH

green, gray green, units
pH 9 – 13, gray violet, violet
Merck

2.4. pH Indicator Papers

“Bleeding” Indicator Papers. Absorbent, uncoated, unfilled paper is impregnated with solutions of acid – base indicators
(see Section Dye Classes and Individual pH Indicators ) and dried in an inert atmosphere under constant conditions. One or
more indicator dyes can be applied, depending on the intended end use.

Three types of pH indicator papers are produced, which differ with regard to measuring range and accuracy. Simple pH
indicator papers (e.g., congo, litmus, or phenolphthalein papers) can be used to determine only whether a solution is acidic,
neutral, or alkaline. Universal indicator papers permit measurements over the total pH range 0 – 14 with an accuracy of ca. 1
pH unit. Special indicator papers cover ranges of 2 – 5 pH units; under ideal conditions, the pH can be determined with an
accuracy of 0.5 or even 0.2 – 0.3 units. In both universal and special indicator papers, the color obtained during contact with
the test solution is compared with a reference color scale.

The pH indicator papers produced by impregnation with dye solutions are prone to “bleeding”; i.e., a portion of the indicator
dye dissolves in the test solution during immersion of the paper or strip. This occurs particularly in an alkaline medium
(because the solubility of the dye reaches a maximum) and in weakly buffered solutions (because indicator papers must
remain in the solution longer). Dissolution of the indicator dye leads to areas of varying dye concentration and color on the
test strip, which makes analysis difficult and inaccurate. In addition, the test solution becomes contaminated. To avoid these
drawbacks, manufacturing techniques and the composition of the test paper have been modified. However, progress has
been limited, and fundamental difficulties associated with the impregnation technique have not been eliminated.

“Nonbleeding” Indicator Papers. Nonbleeding papers can be prepared by using direct [15] or reactive [16] dyes. In the
latter case, pure linters (raw material for paper) are suspended in water, and an aqueous solution of a reactive dye is then
added. An example of such a dye is

Yellow – violet, pH 5.5 – 8.0

which reacts via the sulfonic acid group of the side chain. The mixture is finally made alkaline, and the dye reacts with the
linters. After completion of the reaction, the fiber pulp is centrifuged, washed free of electrolytes, and processed on a paper
machine to form pH paper. The pH papers produced in this way are mostly bonded onto a plastic base material and used as
pH test strips.

Those pH papers with covalently bonded indicator dyes have the advantage of being able to remain indefinitely in the test
liquid, without the indicator dyes mixing with each other or dissolving. This improves measuring accuracy significantly,
particularly in alkaline solution, weakly buffered media, and hot solutions (up to ca. 70 °C). Furthermore, pH determination
can be carried out on turbid solutions or suspensions, because the strip can be rinsed briefly with distilled water without
changing its color. Another advantage is that covalently bonded indicator dyes are not prone to protein errors (see Section
Errors); hence, pH measurements can also be performed on biological specimens.

The introduction of mobile small reflectometers (cf. Chap. Analytical Test Kits) which allow the quantitative reading of test
strips also led to pH strips which can be numerically evaluated. It is to be mentioned that these products (commercial name
Reflectoquant) enable to perform pH measurements even in very complex and difficult matrices like cooling lubricants.

Recently, a new type of indicators was coupled to paper by the above-mentioned direct dye method. The resulting indicator
strips allow quick visual pH determination of concentrated acids [17] and bases [18], respectively. These products

page 12 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
(commercial names Aci-Test and Alka-Test) allow to substitute acid – base titration in process analysis [19].

Manufacturers of pH papers and test strips are:

Hydrion (United States)

Kloz (FRG)
Kyoritsu Chemical (Japan)
Macherey-Nagel (FRG)
Merck (FRG)
Riedel-de Haën (FRG)
Toyo Roshi (Japan)
Whatman (United Kingdom)

2.5. pH Indicators for Nonaqueous Solutions

The most widely used indicator dyes for nonaqueous solutions are listed in Table 6. Transition regions are not specified
because of inadequate correlation between the pH range in nonaqueous solutions and that in the aqueous phase. The
indicators are employed almost exclusively for titrations. Before routine use of a pH indicator in nonaqueous solutions, the
coincidence of the transition and equivalence points under the chosen conditions should be checked by potentiometric
titration. Detailed information is available in [20-22].

Table 6. Commonly used pH indicators for titrations in nonaqueous solvents

Substance to be titrated Indicator CAS Color change

registry
no.

Weak bases
In glacial acetic acid with perchloric Crystal Violet [548-62-9] violet – blue green –
acid green – yellow
Methyl Violet [8004-87-3] violet – blue – blue
green – yellow
Malachite Green [2437-29-8] brown green – green –
yellow
Nile Blue A [3625-57-8] blue – colorless
In acetic anhydride with perchloric Crystal Violet [548-62-9] blue – green yellow –
acid yellow
- [6948-88-5] yellow – green
naphtholbenzein
Weak acids
In dimethylformamide
With Thymol Blue [76-61-9] red – yellow – blue
tetraalkylammoniumhydroxide
With n-butylamine Azo Violet [74-39-5] red – blue

With pyridine 2-nitroaniline [88-74-4] yellow – red

[Top of Page]

3. Adsorption Indicators
Adsorption indicators provide information about a specific chemical condition by means of color or fluorescence reactions
that take place on the surface of a precipitate. In titrimetry, they are employed to indicate the end points of precipitation
reactions. The distinct change in color of the precipitate is generally easier to perceive than corresponding changes in the
color of the solution. This is because the first ions that bind to the precipitate are those present in excess in the solution to be
titrated. At the neutralization point, this excess is replaced by an excess of ions of opposite charge, and the charge on the
surface of the precipitate is reversed. This reversal is made visible by the simultaneous binding of anionic or cationic dyes to
the surface.

The best known titration – precipitation reactions include the argentometric determination of halogens and pseudohalogens
with such anionic dyes as fluorescein, eosin, phloxine, and Rose Bengal. A comparable but reversed titration reaction
employs the cationic dye rhodamine 6G for the determination of silver.

page 13 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Fluorescein: R1 = H, R2 = H
Eosin: R1 = Br, R2 = H
Phloxine: R1 = Br, R2 = Cl
Rose Bengal: R1 = I, R2 = Cl

Rhodamin 6G

The uses and color transitions of the adsorption indicators are given in Table 7. Further indicators, along with more detailed
information, can be found in [5, vol. 2, pp. 267 – 315], [23-26].

Table 7. Adsorption indicators

Name CAS Use Color change (solu-

registry no. tion precipitate)

Fluorescein (C.I. 45 350) [2321-07-5] titration of Cl–, Br–, I–, and SCN– pale green
with Ag+ orange
Eosin (C.I. 45 380) [ titration of Br– and I– with Ag+ orange red violet
17372-87-1]
Phloxine (C.I. 45 405) [6441-77-6] titration of Br– and I– with Ag+ orange red violet
Rose Bengal (C.I. [632-68-8] titration of I– with Ag+ red blue red
45 440)
Rhodamine 6G (C.I. [989-38-8] titration of Ag+ with Br– orange red violet
45 160)

[Top of Page]

4. Fluorescent Indicators
Fluorescent indicators are organic compounds whose changes in fluorescence allow determination of chemical parameters.
Fluorescence may be influenced by factors such as pH, the partial pressure of oxygen, or the presence of metal or halide
ions. The intensity of fluorescence can be increased by addition of surfactants [27]. Examples of fluorescent indicators are
listed in Table 8. Applications of fluorescent indicators include the titration of solutions in which conventional indicators yield
an indistinct color change at the end point (e.g., pH titration of beverages, synthetic resins, oils, or detergents). Titration is
normally carried out in diffuse (ideally red) light or in a darkened room, and the solution is irradiated with UV light (wavelength
300 – 400 nm).

Table 8. Fluorescent indicators

Name CAS Structural formula Comments

registry no.

Fluorescent pH indicators

page 14 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Erythrosin B [16423-68-0] transition range:
(C.I. 45 430)

pH 2.5 (no fluorescence) to

pH 4.0 (pale green
fluorescence)

2-Naphthionic [81-16-3] transition range:

acid
pH 12.0 (blue fluorescence)
to pH 13.0 (violet
fluorescence)
Umbelliferone [93-35-6] transition range:
pH 6.5 (no fluorescence) to
pH 8.5 (blue fluorescence)

Fluorescent metal indicators

Morin (C.I. [480-16-0] for titration of gallium (pH
75 660) 3.8);
transition: green
fluorescence to no
fluorescence for titration of
indium (pH 5.0);
transition: green
fluorescence to no
fluorescence
3-Hydroxy-2- [92-70-6] for titration of aluminum (pH
naphthoic acid 3.0);
transition: blue fluorescence
to green fluorescence

Fluorescent indicators are used in the construction of fluorescent optical sensors (optodes or optrodes). The indicator is fixed
to a substrate such as glass, an ion-exchange membrane, or cellulose. Alternatively, an indicator solution may be separated
from the test solution by a membrane that is permeable to the substance to be analyzed. In both cases, transmission of light
occurs via optical fibers [28]. Such sensors have been used to determine oxygen content [29] or pH [29], [30]. Combination
with optical fibers, however, has also increased interest in the use of fluorescent indicators for the titration of turbid or colored
solutions. This applies to titration of pH [31], metals [32], and precipitates [33]. New fluorescent pH indicators with a wide
transition range suitable for rapid pH determination are described in [34].

Special fluorescent indicators have been employed for the intracellular determination of pH [35], [36] and calcium [37], [38].
Indicators are introduced into the cell in the form of membrane-permeable hydrophobic esters. Inside the cell, the esters are
enzymatically cleaved to form the hydrophilic fluorescent species. The use of fluorescent metal indicators for fluorometric
elemental analysis is described in [39].

[Top of Page]

5. Chemiluminescent Indicators
Chemiluminescent indicators are organic substances whose changes in luminescence enable quantitative analysis of various
compounds. The energy for chemiluminescence is produced by an oxidation – reduction process, particularly by the
formation of transannular peroxides. Very low-intensity light is emitted without the evolution of thermal energy. This light is
usually detected by using equipment operated with the exclusion of external light. However, titration end points determined
with the assistance of chemiluminescent indicators can also be detected visually. Both increases and decreases in the
intensity of luminescence have been used for analytical purposes [40-42].

Chemiluminescent indicators have the advantage that determination can be carried out in highly colored samples. A
disadvantage is that metal contaminants can lead to nonluminescent reactions and, therefore, interfere with the
determination. Chemiluminescent indicators can also be used as redox indicators and, because the redox potential depends
on pH, pH determinations can be made.

An example of this application is the determination of the acid number in highly colored fats and oils by using lucigenin [
2315-97-1], which is oxidized in alkaline media:

page 15 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Other chemiluminescent indicators include luminol [521-31-3] and lophine [484-47-9], both of which are similarly oxidized in
alkaline media; their peroxides have blue and pale yellow luminescence, respectively.

Luminol

Lophine

Other indicators, along with a review of the applications of chemiluminescence in analytical chemistry and biochemistry, are
discussed in [43-45] (see also Luminescent Materials).

[Top of Page]

6. Redox Indicators
Redox indicators are organic compounds whose oxidized and reduced forms display different colors. The two levels of
oxidation form a redox system that can be described by its standard potential or rH value.

6.1. Standard Potential and rH Value

The standard potential U 0 of a redox indicator can be derived from the Nernst equation:

where U is the potential of the redox system; R is the gas constant; T, the absolute temperature; z, the number of electrons
involved in the reaction; F, the Faraday constant; and cox and cred are the molar concentrations of the oxidized and the
reduced species, respectively.

page 16 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The standard potential U 0 results when the activities of the oxidized and reduced species are both assumed to be 1. For the
electrochemical determination of U 0, a platinum electrode is immersed in the test solution and the potential is measured
relative to a standard hydrogen electrode.

Generally, the redox potential depends strongly on pH because hydrogen ions are directly involved in the redox process and,
therefore, must also be considered in the Nernst equation. The redox system of phenolindophenolate for example

can thus be described by

Another way of describing the oxidizing and reducing properties of a redox system is to use the rH value, which was
introduced as an analogue of pH. A platinum electrode immersed in a solution of a redox system is regarded as a hydrogen
electrode. The rH value is defined as the negative logarithm of the hydrogen pressure of this electrode. Measurement is once
again carried out relative to a standard hydrogen electrode. The redox potential is given by

where is the activity of the hydrogen ion and p the pressure of hydrogen. By incorporating pH and rH, the following
relationship is obtained:

From this equation, redox potential and rH value can be interconverted, provided the pH is known. Increasing rH values
indicate increasing oxidation.

To compare redox indicators, their respective redox potentials must be measured under standard conditions, i.e., in neutral
solution and with the redox system present in a semioxidized state.

The color of redox indicators does not change abruptly at a defined potential but rather over a certain range. The midpoint of
this range should be regarded as the rH value. Frequently, however, the rH transition range is also specified.

Practical Aspects. Redox indicators are usually added to the test solution. A minimal amount of indicator should be used
because the indicator also represents a redox system, which could influence the system under investigation.

The redox potential of the indicator must be greater than that of the titrant for an oxidizing titrant (i.e., when an oxidizing
solution is being added) and lower than that of the titrant for a reducing titrant. Furthermore, redox reactions often proceed
slowly and, therefore, require some time to reach equilibrium.

6.2. Classes of Redox Indicators

Table 9 lists a selection of the most commonly employed redox indicators. In this section, the most important classes of
redox indicators are discussed. Redox indicators are dealt with comprehensively in [1, pp. 469 – 678]. Detailed examples of
the use of redox indicators are described in [11], [12], [46], [47].

Table 9. Redox indicators

Indicator CAS registry Standard Approximate Color (reduced –

no. potential U 0, V rH oxidized)

Neutral Red (C.I. 50 040) [553-24-2] – 0.32 a 2 – 4.5 colorless – red

Safranine T or O (C.I. [477-73-6] – 0.29 a 4 – 7.5 colorless – red
50 240)
Indigo Carmine (C.I. 73 015) [860-22-0] – 0.11 a 8.5 – 10.5 yellow tinge – blue
Methylene Blue (C.I. 52 015) [61-73-4] + 0.01 a 13.5 – 15.5 colorless – blue
Thionin (C.I. 52 000) [581-64-6] + 0.06 a 15 – 17 colorless – violet
Thymolindophenol [2667-28-9] + 0.18 a 17.5 – 20 colorless – blue
(red at pH 9)

page 17 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2,6-Dichlorophenol [620-45-1] + 0.23 a 20 – 22.5 colorless – blue
indophenol
Gallocyanine (C.I. 51 030) [1562-85-2] + 0.02 b colorless – violet
blue
Safranine T or O (C.I. [477-73-6] + 0.24 b colorless – blue
50 240) violet
Indigo Carmine (C.I. 73 015) [860-22-0] + 0.29 b colorless – blue
Nile Blue (C.I. 51 180) [2381-85-3] + 0.41 b colorless – blue
red
Methylene Blue (C.I. 52 015) [61-73-4] + 0.53 b colorless – blue
Thionin (C.I. 52 000) [581-64-6] + 0.56 b colorless – violet
2,6-Dichlorophenol [620-45-1] + 0.67 b colorless – red
indophenol
Variamine Blue (C.I. 37 255) [3566-44-7] + 0.71 c colorless – blue
Diphenylamine [122-39-4] + 0.76 b colorless – violet
Diphenylamine-4-sulfonic [6211-24-1] + 0.85 d colorless – red
acid, barium salt violet
Tris(2,2-dipyridyl)iron(II) + 1.03 d red – blue
sulfate
+ 1.02 e
N-Phenylanthranilic acid [91-40-7] + 0.89 d colorless – purple
red
Ferroin [14708-99-7] + 1.06 d orange red – blue
+ 1.14 e
Nitroferroin [4199-88-6] + 1.25 d pink – colorless
+ 1.26 e

a pH 7.
b In strongly acidic solution.
c Standard potential is strongly pH dependent.
d In 1 mol/L H SO .
2 4
e In 0.5 mol/L H2SO4 .

Indophenols. The reaction equation describing the redox process with phenolindophenol [500-85-6] has already been given
in Section Standard Potential and rH Value. Thymolindophenol and 2,6-dichlorophenolindophenol are two redox indicators
with a similar basic structure. Applications include the titration of copper(I) ions with iron(III) sulfate [48] as well as the
determination of free halogen, hypohalogenites, and halogenates with ascorbic acid [49].

Azines, Oxazines, Thiazines. Azines used as redox indicators include safranine T and Neutral Red. Applications include the
titration of iron(III) salts with titanium(II) chloride [50] or chromium(II) salts [51]. Oxazine redox indicators include Nile Blue
and gallocyanine. The best known thiazines are Methylene Blue and thionin. Applications include the titration of iron(III) salts
with vanadium(II) compounds [52]. Many of these dyes are described in detail elsewhere ( Azine Dyes).

Indigo Derivatives ( ). An example of an indigo dye is the frequently employed Indigo Carmine. The redox process can
be expressed as follows:

Applications include the determination of rhenium with tin(II) solutions [53].

Diphenyl and Benzidine Derivatives. Diphenylamine-4-sulfonic acid and Variamine Blue are derivatives of diphenylamine.
The redox process for Variamine Blue can be described as follows:

page 18 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

A multitude of applications have been discussed in [46].

The benzidine derivatives formerly employed as redox indicators include o-tolidine, o-dianisidine, and 3,3-
dimethylnaphthidine. However, because benzidine and most of its derivatives are carcinogenic, they are not included in
Table 9.

1,10-Phenanthroline Complexes. The best known 1,10-phenanthroline complex is ferroin, an iron(II) complex with three
1,10-phenanthroline molecules (see Section Ligands Forming Five-Membered Chelate Rings). Utilization of these
compounds as redox indicators with particularly brilliant color transitions is discussed comprehensively in [46], [54]. An
example is the titration of chromium(VI) with iron(II) in determination of the chemical oxygen demand (COD) according to DIN
38 409, Parts 41 and 43, ISO 6060–1986, or Norme Française NF T 90–101.

[Top of Page]

7. Metal Indicators
Metal indicators are organic compounds that form specifically colored metal ion complexes in aqueous media. These dye
reactions can be employed for detection of the equivalence point in volumetry (complexometry), as well as for concentration
determinations in colorimetry or photometry. The demands placed on metal indicators differ slightly depending upon the area
of application.

The most important metal indicators are listed in Table 10. Detailed information is available on their applications in volumetry
(complexometry) [55-59]; colorimetry and photometry [60-65]; and masking [66], [67].

Table 10. Metal indicators

Indicator CAS For complexometric For photometric

registry no. determination of determination of

Alizarin Complexone
Alizarin S
Arsenazo III
Aurintricarboxylic acid
(Aluminon, C.I. 43 810)
2,2′-Bipyridine
Bromopyrogallol Red
Calcon (Eriochrome Blue
Black R)
Calconcarboxylic acid
Chrome Azurol S (C.I.
43 825)
Chromotropic acid, disodium
salt
Cuprizone
5-(4-
Dimethylaminobenzylidene)
rhodanine
[3952-78-1] Ba, Ca, Cd, Co, Hg, In, fluoride
Pb, Sr, Zn
[130-22-3] Al, lanthanides, Mo, Sc, Al, lanthanides, Sc, Th, Y,
Th, Y, Zr Zr
[62337-00-2] lanthanides, Th, Y lanthanides, Sc, Th, U, Y,
Zr
[4431-00-9] Al, Ca, Fe, Mg Al, Be, Fe, Ga,
lanthanides, Pd, Y
[366-18-7] Ag, Cd, Cu, Fe, Mo
[16574-43-9] Bi, Cd, Co, lanthanides, Ag, Bi, Co, Cu,
Mg, Mn, Ni, Pb lanthanides, Pb, Sb, Sn,
Th, Ti, U, W, Y, Zr
[2538-85-4] Ca, Cd, Mg, Mn, Zn
[3737-95-9] Ca
[1667-99-8] Al, Ca, Cu, Fe, Al, Be, Bi, Cr, Cu, Fe, Ga,
lanthanides, Mg, Ni, Th, In, lanthanides, Mg, Mn,
U, V, Y, Zr Ni, Pb, Pd, Sc, Sn, Th, Y,
Zr
[5808-22-0] Ti
[370-81-0] Cu
[536-17-4] Ag, Au, Cu, Hg, Pd

page 19 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Dimethylglyoxime
1,5-Diphenylcarbazide
Dithizone
Eriochrome Black T
Eriochrome Blue SE
Eriochrome Blue Black B
Eriochrome Cyanine R
(C.I. 43 810)
Fluorescein Complexone
(Calcein)
Glyoxalbis(2-hydroxyanil)
Hematoxylin
8-Hydroxyquinoline
2-Mercaptobenzothiazole
Methylthymol Blue
Murexide
1-Nitroso-2-naphthol
2-Nitroso-1-naphthol
Nitroso-R-salt
1,10-Phenanthroline
Phenylfluorone
Phthalein Purple
1-(2-Pyridylazo)-2-naphthol
(PAN)
4-(2-Pyridylazo)resorcinol
(PAR)
Pyrocatechol Violet
Pyrogallol Red
Sulfonazo III
5-Sulfosalicylic acid
4-(2-Thiazolylazo)resorcinol
(TAR)
Thorin
Thymolphthalexon
Tiron
Toluene-3,4-dithiol
Xylenol Orange
Zincon
7.1. Volumetry (Complexometry) [55-59]
A small amount of indicator is added to the solution to be titrated. The color of the metal-free indicator must differ from that of
[95-45-4] Co, Fe, Ni, Pd
[140-22-7] Hg, Ni, Pb, V Cr, Cu, Hg
[60-10-6] Ag, Au, Bi, Cd, Co, Cu, Ag, Au, Bi, Cd, Co, Cu,
Fe, Hg, In, Mn, Ni, Pb, Sn, Hg, Ni, Pb, Pd, Zn
Zn
[1787-61-7] Ba, Ca, Cd, Hg, In,
lanthanides, Mg, Mn, Pb,
Zn, Zr
[1058-92-0] Ca, Cd, Mg, Mn, Ni, Pb,
Zn
[3564-14-5] Ca, Cd, Mg, Mn, U, Zn, Zr
[3564-18-9] Al, Ca, Cu, Fe, Mg, Th, Zr Al, Be, Cr, Ga, Mg, Th, Tl,
Zr
[1461-15-0] Ba, Ca, Cr, Cu, Mg, Sr Ca
[1149-16-2] Ca, Sc, U
[517-28-2] Al, Bi, Cu, Th, Zr Sn
[148-24-3] Al, Bi, Ca, Cd, Cr, Cu, Fe,
Ga, Ge, In, La,
lanthanides, Mg, Nb, Sc,
Th, Ti, U, V, Y, Zn, Zr
[149-30-4] Au, Cd, Cu, Pb, Pd
[1945-77-3] Ba, Bi, Ca, Cd, Cu, Fe, Al, Be, Bi, Ca, Co, Cr, Cu,
Hg, In, lanthanides, Mg, Fe, Ga, In, lanthanides,
Mn, Pb, Sc, Sn, Th, Tl, Mg, Mn, Ni, Pb, Pd, Sc,
Zn, Zr Th, U, V, Y, Zn, Zr
[3051-09-0] Ag, Ca, Co, Cu, Mn, Ni, Ca, Cu, Sc
Sc, Th, Zn
[131-91-9] Co, Fe, Mo, Pd
[132-53-6] Co, Pd
[525-05-3] Cu, Ni Co, Fe, Pd
[66-71-7] Ag, Cu, Fe, Pd, V, Zn
[975-17-7] Ge, Mo, Nb, Sb, Sn, Ta, U
Ba, Ca, Cd, Mg, Sr Ba, Ca, Mg, Mo
[85-85-8] Al, Bi, Cd, Co, Cu, Fe, Ag, Bi, Cd, Co, Fe, Ga,
Ga, Hg, In, Mn, Ni, Pb, lanthanides, Pd, Sc, U, V,
Th, Tl, U, V, Zn Y, Zn, Zr
[1141-59-9] Al, Bi, Cd, Cu, Fe, Ga, Ag, Bi, Cd, Co, Cr, Cu, Fe,
Hg, In, lanthanides, Mn, Ga, Hg, La, Mn, Nb, Ni,
Ni, Pb, Sr, Th, Tl, Zn Pb, Pd, Sc, U, V, Y, Zn, Zr
[115-41-3] Bi, Cd, Co, Cu, Fe, Ga, In, Al, Bi, Cr, Cu, Fe, Ga, In,
Mg, Mn, Ni, Pb, Th, Zn Mg, Mn, Ni, Pb, Sc, Sn,
Th, Y, Zn, Zr
[32638-88-3] Bi, Co, Ni, Pb Ag, Mo
[1738-02-9] Ba, Ba
[5965-83-3] Fe, Tl Fe, Ti
[2246-46-0] Co, Cu, lanthanides, Ni, Tl as for PAR
[3688-92-4] Bi, Sc, Th, U, Y lanthanides
[1913-93-5] Ag, Ba, Ca, Cr, Mn, Sr Ba, Ca, lanthanides, Mg,
Mo, Sr
[149-45-1] Fe Fe, lanthanides, Mo, Ti, U,
W
[496-74-2] Ge, Mo, Sn, W
[1611-35-4] Bi, Ca, Cd, Co, Cu, Fe, Ag, Bi, Cd, Cr, Fe, Ga, Hg,
Hg, In, lanthanides, Mg, In, La, lanthanides, Mn,
Mn, Pb, Sc, Th, Tl, U, V, Mo, Nb, Ni, Pb, Sc, Sn,
Y, Zn, Zr Th, Ti, Tl, U, V, Y, Zn, Zr
[135-52-4] Cd, Hg, Pb, Zn Cu, Ga, Ni, Zn
page 20 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
the metal –indicator complex. Another complexing agent, usually an aminopolycarboxylic acid such as
ethylenediaminetetraacetic acid (Titriplex) is used as titrating agent, which binds the metal ions as a colorless complex. The
equivalence point is indicated by the color change between the metal – indicator complex and the free indicator.

Metal indicators for complexometry must satisfy the following requirements:

1. Complex Stability. The metal – indicator complex must be less stable than the metal – aminopolycarboxylic acid
complex.
2. Detection. The color difference between the metal – indicator complex and the free indicator must be large and clearly
visible.
3. Selectivity. Ideally, the indicator should be specific for only one element. Although this condition is rarely met,
selectivity can often be increased by correct choice of reaction conditions.
4. Shelf Life and Solubility. Both indicator and metal – indicator complex should be water soluble. The indicator should
have a long shelf life in aqueous or water-miscible mixtures.

7.2. Colorimetry and Photometry [60-65]

In visual colorimetry (color comparison) and photometry (measurement of light absorption), the formation of colored
complexes is used to determine metal ion concentration. The intensity of the resulting color is a function of the metal ion
concentration. The actual concentration is determined by comparing the color with standards or by measurement of the
absorbed light.

Metal indicators for colorimetry or photometry must satisfy the following requirements:

1. Complex Stability. Metal – indicator complexes should be stable with respect to time. Contrary to the situation with
complexometry, thermodynamic stability is less important. By selecting the conditions (indicator excess), even weak
complexes can be used to determine concentration.
2. Sensitivity. Sensitivity is governed primarily by the magnitude of the molar extinction coefficient. Metal – indicator
complexes typically used for colorimetric and photometric determinations have extinction coefficients ( ) in the region
of 5 × 103 to 1 × 105 L mol–1 cm–1.
3. Selectivity. Maximum selectivity is desired to minimize the number of processing steps required to remove interfering
contaminant ions. Selective or specific reagents are, however, still very rare. Selectivity can sometimes be enhanced
by modifying the indicator molecule; however, selectivity is usually improved by optimizing the reaction conditions and
the use of additives (e.g., masking agents) [66], [67].

7.3. Structure
The metal indicator and metal ions form complexes consisting of an electron donor and an electron acceptor. If the donor
molecule (ligand) contains two or more atoms with donor properties, the complex is known as a chelate. The number of
ligand donor atoms bound to the central ion determines the ligand functionality; for example, ligands with a single donor atom
are referred to as monodentate; those with two donor atoms, as bidentate.

Electron-donating atoms of the “functional group” of the ligand molecule include oxygen, nitrogen, and sulfur. The
fundamental properties of a complexing agent are determined by these functional groups, whereas other molecular
substituents influence sensitivity, selectivity, and color contrast. Incorporation of sterically hindered groups in an indicator
molecule may increase its selectivity. Increasing the size of the resonance system enhances sensitivity in all cases. The
latter is particularly advantageous if the change in resonance shifts the absorption maximum of the complex to longer
wavelengths because the human eye is more optically receptive to these wavelengths (important for visual techniques).

Most organic complexing agents can be regarded as bidentate and form five- or six-membered chelate rings. According to
the type of ligand, chelating reagents can be classified into ligands with O,O-, O,N-, N,N-, and N,S-donor atoms.

7.4. Ligand Classes

7.4.1. Ligands Forming Five-Membered Chelate Rings
O,O Donor Atoms. The O–C=C–O group is often present in chelating agents. The resulting complexes have the following
structure:

Important examples of this group include 1,2-dihydroxybenzenes and 1,2-dihydroxynaphthalenes, e.g., Tiron, Pyrocatechol
Violet, Pyrogallol Red, and Bromopyrogallol Red, as well as the fluorones.

Tiron

page 21 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Pyrocatechol Violet

R = H: Pyrogallol Red
R = Br: Bromopyrogallol Red

R = CH3, C6H5: Fluorones

A range of triphenylmethane dyes derived from Pyrogallol Red and Pyrocatechol Violet are frequently employed as indicators
in complexometry.

O,N Donor Atoms. The functional group in one of the oldest and best known complexing agents, 8-hydroxyquinoline,
consists of O–C=C–N. Because this functional group has only slight selectivity, the compound is a group reagent and forms
yellow complexes of varying intensity with more than 40 ions.

8-Hydroxyquinoline

The O–C–N=N group present in diphenylcarbazone [538-62-5] also forms five-membered chelates, with a structure
analogous to the 8-hydroxyquinoline complex. Chelates can be derived from the tautomeric hydroxy form:

Diphenylcarbazone
Diphenylcarbazone and the related diphenylcarbazide are used in the photometric determination of chromium and mercury.

N,N Donor Atoms. Many of the more important metal indicators contain N,N donor atoms. The most common groups are
N=C–C=N and N=C–N=N. The former is present in -dioximes, which are highly selective reagents for nickel and palladium.
The best known compound is the classic nickel reagent diacetyldioxime (dimethylglyoxime).

page 22 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Diacetyldioxime

Chelate

If the functional complexing group N=C–C=N is incorporated in a heterocyclic system, then a class of indicators is obtained
that is specific for iron and copper, e.g., 2,2′-dipyridyl (1) and 1,10-phenanthroline (2):

1 Copper indicators: 2 Iron indicators:

R = CH3, C6H5 R=H

The indicators form intensely colored 1 : 2- and 1 : 3-chelates with iron(II) and copper(I) compounds. In these chelates,
however, the central ion can easily be oxidized to iron(III) or copper(II), which give weakly colored complexes. This reaction
allows the chelates to be used as redox indicators (see Section Classes of Redox Indicators). Due to steric reasons, only the
copper compounds can form the tetrahedral 1 : 2-chelate; this explains their highly specific nature.

The N=C–N=N group occurs in the pyridylazoresorcinols (PAR) and pyridylazonaphthols (PAN).

4-(2-pyridylazo)resorcinol (PAR)

1-(2-pyridylazo)-2naphthol (PAN)
Chelation is thought to occur by formation of an additional five-membered ring with the N,O group:

This does not increase selectivity, but significantly improves the stability of the complex.

The pyridylazoresorcinols and pyridylazonaphthols are excellent indicator reagents for heavy-metal ions. They are frequently
employed in complexometry and photometry.

N,S Donor Atoms. The N=C–C–S group is present in the thio analogue of 8-hydroxyquinoline, i.e., 8-mercaptoquinoline [
491-33-8]. Due to the different reactivity of sulfur, this indicator is more selective than 8-hydroxyquinoline; it reacts only with
transition metals of pronounced sulfur affinity, as well as molybdenum, tungsten, and vanadium.

The N=N–C=S group in the sulfur analogue of diphenylcarbazone, diphenylthiocarbazone (dithizone), is of greater
importance. Dithizone forms complexes with a number of transition metals and thus is a group reagent:

page 23 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Nevertheless, dithizone can be employed for very selective and also extremely sensitive determinations by maintaining a
constant pH, employing masking reagents, and replacement (substitution) reactions [65].

7.4.2. Ligands Forming Six-Membered Chelate Rings

Six-membered chelate rings occur with a frequency similar to five-membered rings in important analytical complexes.
However, only O and N are donor atoms. The most important reagents are anthraquinones, 4,5-dihydroxynaphthalenes, 2-
hydroxybenzoic acids, 2-nitrosonaphthols, and methylaminodiacetic acid derivatives.

Anthraquinones. Anthraquinone indicators include the commonly used indicator, Alizarin S (1,2-dihydroxyanthraquinone-3-
sulfonic acid, sodium salt):

Chelate formation is limited mainly to cations from the second, third, and fourth groups of the periodic table.

4,5-Dihydroxynaphthalenes. Chromotropic acid [148-25-4] (4,5-dihydroxy-2,7-naphthalenedisulfonic acid) represents a

typical member of this group:

A multitude of compounds containing additional azo functions can be derived from the basic structure. An example is
phenylazochromotropic acid [4197-07-3], which is able to function in two ways. Normally, complex formation takes place via
the O,O ligand atoms, whereas the azo group forms a second ring as a result of hydrogen bonding:

Cations with a pronounced affinity toward nitrogen, however, can also form complexes via the N,O group.

page 24 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Other important indicators from this group include arsenazo I [520-10-5], arsenazo II [3861-75-4], arsenazo III [1668-00-4],
sulfonazo III [1738-02-9], chlorophosphonazo III [1914-99-4], and thorin [3688-92-4].

2-Hydroxybenzoic Acids. The simplest 2-hydroxybenzoic acids are salicylic [69-72-7] and 5-sulfosalicylic [5965-83-3] acids,
but they are of limited analytical importance.

Salicylic acid

5-Sulfosalicylic acid
Frequently used complexometric indicators and photometric reagents include Chrome Azurol S, aurintricarboxylic acid, and
Eriochrome Cyanine R:

Chrome Azurol S

Aurintricarboxylic acid

Eriochrome Cyanine R

2-Nitrosonaphthols. 2-Nitrosonaphthol indicators are employed preferentially in the analysis of cobalt; however, they can
also be used as group reagents. Typical examples include 1-nitroso-2-naphthol (3), 2-nitroso-1-naphthol (4), and the water-
soluble nitroso-R-salt (5).

3 4

page 25 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Methylaminodiacetic Acid Derivatives. Compounds containing a methylaminoacetic acid group and a triphenylmethane or
anthraquinone group can also form six-membered ring chelate complexes. Dyes of this type are preferred as indicators in
complexometry, but they are also suitable photometric reagents provided selectivity is not critical. The most important
examples include:

R1 R2 R3

Xylenol Orange CH3 H SO3H

Methylthymol Blue CH(CH3)2 CH3 SO3H
Thymolphthalexone CH(CH3)2 H COOH
Phthalein Purple CH3 H COOH

Alizarin Complexone

Fluorescein Complexone

[Top of Page]

8. Analytical Test Kits

Test kits are combinations of indicator reagents, miscellaneous reagents, and other auxiliaries that allow simple, usually rapid
analyses. They can be reasonably well classified according to their applications:

1. Determination of chemical parameters (especially ions) in nonbiological media

2. Determination of chemical and biological parameters in biological media such as body fluids (i.e., diagnostic test kits,
)
3. Determination of chemical parameters in gases (e.g., analytical tubes in work safety applications)

In this article, only the first group of analytical test kits is discussed. The pH papers and strips discussed in Section pH
Indicator Papers represent the first examples of such products.

Properties and Limitations. Most analytical test kits are designed for use without additional laboratory equipment, which

page 26 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
means that

1. reagents and auxiliary chemicals are provided ready for use;

2. correct dosage is possible with very simple equipment generally included in the kit;
3. complex operations such as weighing and thermostating are generally not necessary;
4. test kits usually contain a buffer system which, to a certain extent, can counteract possible pH deviations and prevent
associated interference;
5. the interferences of contaminating ions and other compounds are minimized by optimizing reaction conditions and
reagent composition (e.g., incorporation of masking agents);
6. manufacturers may also provide information pertaining to the tolerance levels of contaminant ion concentrations;
7. auxiliary materials necessary for evaluation of results are included in the kits; and
8. batch-to-batch reproducibility of kit properties is guaranteed by the manufacturer.

Test kits are, of course, also prone to certain limitations:

1. Like classical test procedures, test kits can only detect free or easily released ions. To detect complexed or bound
ions, a preliminary digestion stage is required.
2. Samples with extreme pH values must be adjusted prior to the determination.
3. When extreme conditions with regard to contaminant ions are involved, the sample must be appropriately treated.

Classification. Ideally, a test kit should combine the following properties:

1. high detection sensitivity,

2. maximum measuring range,
3. highest possible differentiation between divisions on the indicator scale,
4. absolute selectivity, and
5. convenient use.

Because a combination of all these requirements is not possible in practice, a compromise must be made. One or two of the
above requirements are satisfied, and the others are met to an extent acceptable for the problem in question. Hence, a
number of quite different types of test kits have been developed:

1. Test papers providing a yes – no result

Manufacturers. Macherey-Nagel, Merck (FRG); Kyoritsu Chemical, Sibata, Toyo Roshi (Japan).
2. Test papers or strips giving semiquantitative results (generally, visual differentiation between four to eight
concentrations)
Manufacturers. As for (1) but also ETS and Serim (both USA).
3. Test strips providing quantitative results that require instrumental evaluation [68-72]
Manufacturers. Merck (FRG).
4. Test kits for semiquantitative determinations (with up to ten, visually discernible concentrations)
Manufacturers. Macherey-Nagel, Merck (FRG); Kyoritsu Chemical, Sibata, Yoshitomi (Japan); Lovibond (UK);
CHEMetrics, Hach, Hellige, LaMotte (USA).
5. Test kits providing quantitative results that require instrumental evaluation (For example, manufacturers offer
photometers that have been optimized for specific test kits and, hence, simplify the test applications considerably.
However, the kits can also be used with conventional photometers.)
Manufacturers. Hoelzle & Chelius, Lange, Macherey-Nagel, Merck, Riedel-de Haen (FRG); Kyoritsu (Japan);
Lovibond, WPA (UK); CHEMetrics, Hach, LaMotte (USA).

Figure 3 shows nitrite test kits (Merck) that illustrate the possible options for sensitivity, measurement range, and
measurement intervals.

Figure 3. Measuring range and measuring intervals of nitrite test kits (Merck)

a) Test strip; b) – e) Test kits with visual evaluation in reflected light; f) Test kit with visual evaluation in transmitted light; g)
Test kit with instrumental evaluation

Horizontal bars indicate divisions of the color scale used to read the concentration.

Parameters that can be determined by using currently available tests are listed below:

Acid binding capacity

Acidity
Acids, organic
Alkalinity
Aluminum
Amines
Ammonium
Amphetamines
Antimony

page 27 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Adsorbable organic halogens (AOX)
Arsenic
Ascorbic acid
Bacteria, coliform
Barbiturates
Bismuth
Biological oxygen demand (BOD)
Boron
Bromide
Bromine
Cadmium
Calcium
Cannabinoids
Carbohydrates
Carbon dioxide
Carbon tetrachloride
Carbonate hardness
Chalk requirement (soil)
Chloride
Chlorine
Chlorine dioxide
Chromium
Cobalt
Cocaine
Chemical oxygen demand (COD)
Complexing agents
Copper
Cyanide
Cyanuric acid
Detergents
Diethylhydroxylamine
Ethanol
Ethylenediaminetetraacetic acid
Formaldehyde
Glycols
Gold
Heroin
Hydrazine
Hydrogen peroxide
Hypochlorite
Iodine
Iron
Lead
Lysergic acid diethylamide (LSD)
Magnesium
Manganese
Molybdenum
Morphine
Nickel
Nicotine
Nitrate
Nitrite
Nitrobenzene
Phenols
Oil
Opium
Oxidizing agents
Oxygen
Ozone
Penicillin
Pentachlorophenol
Peracetic acid
Permanganate value
Peroxidase
Peroxides
pH
Phenols
Phosgene
Phosphatases
Phosphates
Potassium
Pyridine
Quaternary ammonium compounds
Silicates
Silver

page 28 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Sodium
Styrene
Sugar
Sulfate
Sulfide
Sulfite
Tannin – lignin
Tartaric acid
Thiocyanate
Thioglycolic acid
Thiophene
Tin
Titanium
Total organic carbon (TOC)
Triazole
Trichloroacetic acid
Urea
Vanadium
Water
Water hardness (total,
carbonate, residual)
Water, inherent color
Zinc
Zirconium

Applications. The ranges of applications for test kits differ widely; the following are some important examples.

1. Sample screening: determination on site to establish whether laboratory analysis is necessary; determination of
approximate concentrations in the laboratory
2. Process control: sewage plants, fermentation equipment, chemical production plants, drinking water treatment plants,
and agriculture
3. Rapid analysis of wastewater: determination of substances that cannot withstand transportation, screening, control of
threshold values
4. Rapid analysis of surface water: emergency analysis following water pollution, determination of special parameters
(see above), testing the suitability of water in the construction industry, monitoring water quality in fish farming

Test kits involve a great deal of extremely innovative chemistry; however, the basic principles of analytical chemistry (e.g.,
correct sampling, validating) must not be neglected when using them.

[Top of Page]

References
General References
1. E. Bishop (ed.): Indicators, “International Series of Monographs in Analytical Chemistry no. 51,” Pergamon Press,
Oxford 1972.
2. G. Jander, K. F. Jahr: Maßanalyse, Theorie und Praxis der Titration mit chemischen und physikalischen Indikatoren,
15th ed., De Gruyter, Berlin 1989.
3. G. Kraft, J. Fischer: Indikation von Titrationen, De Gruyter, Berlin 1972.
4. I. M. Kolthoff, E. B. Sandell, E. J. Meehan, S. Bruckenstein: Quantitative Chemical Analysis, 4th ed., Collier-Macmillan,
London 1969, pp. 691 – 696.
5. V. Schmidt, W. D. Mayer in R. Bock, W. Fresenius,H. Günzler, W. Huber, G. Tölg (eds.): Analytiker-Taschenbuch,
vol. 2,pp. 267 – 315 (1981); vol. 3, pp. 37 – 86 (1983) Springer Verlag, Berlin-Heidelberg-New York.
6. H. U. Bergmeyer (ed.): Methods of Enzymatic Analysis, 3rd ed., vol. 1, VCH Verlagsgesellschaft, Weinheim, Germany
1983, pp. 197 – 232.
7. O. Sonntag: Trockenchemie, Analytik mit trägergebundenen Reagenzien, Thieme Verlag, Stuttgart-New York 1988.
8. F. W. Küster, A. Thiel, A. Ruland (eds.): Rechentafeln für die chemische Analytik, 104th ed., De Gruyter, Berlin 1993.
9. O.-A. Neumüller (ed.): Römpps Chemie-Lexikon, 10th ed., vol. 3, Georg Thieme Verlag, Stuttgart 1997, pp. 1906 –
1908.
10. E. B. Sandell, T. S. West, Pure Appl. Chem. 19 (1969) 427 – 436.
11. I. M. Kolthoff, V. A. Stenger: Volumetric Analysis, Interscience, New York vol. 1, 1942; vol. 2, 1947.
12. I. M. Kolthoff, R. Belcher: Volumetric Analysis, vol. 3, Interscience, New York 1957.
Specific References
13. G. D. Christian: “The Effect of Salts on Titration,” CRC Crit. Rev. Anal. Chem. 5 (1975) 119 – 163.
14. G. Ackermann, L. Sommer, W. I. Stephen, Pure Appl. Chem. 57 (1985) 845 – 848.
15. Behringwerke, DE 2 436 257, 1974 (E. Pfeil).
16. Merck, DE 1 698 247, 1968 (K. Neisius, W. Bäumer).
17. Merck, DE 4 139 302, 1991 (W. Fischer, R. Klink).
18. Merck, DE 4 333 696, 1993 (W. Fischer, S. Baum, T. Hartig, M. Schleehahn).

page 29 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

19. Merck, company information.

20. I. Gynenes: Titrationen in nichtwäßrigen Medien, 3rd ed., Enke Verlag, Stuttgart 1970.
21. K. Stammbach: Titrationen in nichtwäßrigen Lösungsmitteln, Schweiz. Laborantenzeitung, Separatdruck 1970.
22. J. S. Fritz: Acid-Base Titrations in Nonaqueous Media, Allyn and Bacon, Boston 1973.
23. E. Singh, Fresenius Z. Anal. Chem. 282 (1977) 299 – 300.
24. S. K. Garg, D. C. Agarwal, G. K. Chaturverdi, J. Indian Chem. Soc. 58 (1981) 662 – 663. Links
25. J. F. Coetzee in I. M. Kolthoff, P. J. Elving (eds.): Treatise on Analytical Chemistry, 2nd ed., Part I, vol. 3, Wiley-
Interscience, New York-ChichesterBrisbane-Toronto-Singapore 1983, pp. 362 – 378.
26. R. Kumar, D. C. Agarwal, G. K. Chaturverdi, Indian J. Chem. 24 A (1985) 631 – 632.
27. W. L. Hinze, H. N. Singh, Y. Baba, N. G. Harvey, Trends Anal. Chem. 3 (1984) 193 – 199.
28. O. S. Wolfbeis, Trends Anal. Chem. 4 (1985) 184 – 188.
29. N. Opitz, Chem. Ind. (Düsseldorf) 36 (1984) 742 – 744.
30. O. S. Wolfbeis, E. Fürlinger, H. Kroneis, H. Marsonner, Fresenius Z. Anal. Chem. 314 (1983) 119 – 124.
31. O. S. Wolfbeis, Fresenius Z. Anal. Chem. 320 (1985) 271 – 273.
32. O. S. Wolfbeis, B. P. H. Schaffar, Talanta 33 (1986) 867 – 870. Links
33. O. S. Wolfbeis, P. Hochmuth, Mikrochim. Acta 1984, 129 – 148. Links
34. O. S. Wolfbeis, H. Marhold, Fresenius Z. Anal. Chem. 327 (1987) 347 – 350.
35. T. J. Rink, R. Y. Tsien, T. Pozzan, J. Cell. Biol. 95 (1982) 189; Links Chem. Abstr. 97 (1982) 179 455 e.
36. J. A. Thomas, Soc. Gen. Physiol. Ser. 40 (1986) 311; Chem. Abstr. 106 (1987) 152 380 u.
37. G. Grynkiewicz, M. Poenie, R. Y. Tsien, J. Biol. Chem. 260 (1985) 3440. Links
38. G. H. R. Rao, Biochem. Biophys. Res. Commun. 132 (1985) 652. Links
39. G. Schwedt: Fluorimetrische Analyse, Verlag Chemie, Weinheim, Germany 1981.
40. O. V. Zui, A. V. Terletskaya, A. T. Pilipenko, Zh. Anal. Khim. 40 (1985) 120 – 124; Chem. Abstr. 102 (1985) 124 774 z.
41. I. E. Kalinichenko, T. M. Tkachuk, A. T. Pilipenko, Zh. Anal. Khim. 40 (1985) 1581 – 1585; Chem. Abstr. 104 (1986)
61 172 v.
42. L. L. Klopf, T. A. Niemann, Anal. Chem. 57 (1985) 46 – 51. Links
43. W. R. Seitz: “Chemiluminescence and Bioluminescence Analysis: Fundamentals and Biomedical Applications,” CRC
Crit. Rev. Anal. Chem. 13 (1981) 1 – 58.
44. L. J. Kricka, G. H. G. Thorpe, Analyst (London) 108 (1983) 1274 – 1296.
45. J. Schölmerich, R. Anderson, A. Kapp, M. Ernst, W. G. Woods (eds.): Bioluminescence and Chemiluminescence—New
Perspectives, Wiley-Interscience, New York – Chichester – Brisbane – Toronto – Singapore 1987.
46. A. Hulanicki, S. G ab, Pure Appl. Chem. 50 (1978) 463 – 498.
47. J. Barek, E. Berka: “Redox Titrations in Nonaqueous Media,” CRC Crit. Rev. Anal. Chem. 15 (1984) 163 – 222.
48. G. J. Misra, J. P. Tandon, Fresenius Z. Anal. Chem. 214 (1965) 94.
49. L. Erdey, G. Svekla, Fresenius Z. Anal. Chem. 167 (1959) 164.
50. P. V. K. Rao, G. G. Rao, Anal. Chim. Acta 29 (1962) 410.
51. J. P. Tandon, R. C. Mehrotra, Fresenius Z. Anal. Chem. 187 (1962) 40.
52. K. L. Chawla, J. P. Tandon, Talanta 12 (1965) 665. Links
53. G. Henze, R. Geyer, Fresenius Z. Anal. Chem. 200 (1964) 434.
54. A. A. Schilt: Analytical Applications of 1,10-Phenanthroline and Related Compounds, Pergamon Press, Oxford 1969.
55. E. Merck, company in formation, Komplexometrische Bestimmungsmethoden mit Titriplex, Darmstadt 1987.
56. A. Hulanicki, S. G ab, G. Ackermann, Pure Appl. Chem. 55 (1983) 1137 – 1230. Links
57. H. Freiser in I. M. Kolthoff, P. J. Elving (eds.): Treatise on Analytical Chemistry, Part I, vol. 3, Wiley-Interscience, New
York-Chichester-Brisbane-Toronto-Singapore 1983, pp. 395 – 505.
58. G. Schwarzenbach, H. Flaschka: Die komplexometrische Titration, 5th ed., Enke Verlag, Stuttgart 1965.
59. R. P ibil: Applied Complexometry, Pergamon Press, Oxford 1982.
60. O. G. Koch, G. A. Koch-Dedic: Handbuch der Spurenanalyse, 2nd ed., Springer Verlag, Berlin-Heidelberg-New York
1974.
61. J. Fries, H. Getrost: Organische Reagenzien für die Spurenanalyse, Merck, Darmstadt 1975.
62. S. B. Savrin: “Organic Reagents in Photometric Analysis,” CRC Crit. Rev. Anal. Chem. 8 (1979) 55 – 109.
63. F. Umland: Theorie und praktische Anwendung von Komplexbildnern in der Analyse, Akadem. Verlagsgesellschaft,
Frankfurt 1971.
64. Z. Marczenko: “Spectrophotometric Detection of Trace Elements,” CRC Crit. Rev. Anal. Chem. 15 (1984) 163 – 222.
65. H. M. N. H. Irving: “The Analytical Applications of Dithizone,” CRC Crit. Rev. Anal. Chem. 8 (1980) 321 – 366.
66. D. D. Perrin: Masking and Demasking of Chemical Reactions, Wiley-Interscience, New York 1970.
67. D. D. Perrin: “The Selection of Masking Agents for Use in Analytical Chemistry,” CRC Crit. Rev. Anal. Chem. 5 (1975)
85 – 115.
68. G. Schwedt, UmweltMagazin 22 (1994) 134 – 135.
69. L. C. Waters, R. A. Jenkins, R. W. Counts, DOE Methods, Method Number MS 100.
70. L. C. Waters, R. W. Counts, A. Palausky, R. A. Jenkins, J. Hazard. Mater. 43 (1995) 1 – 12. Links
71. H. Rehbein, T. Schmidt, Fischwirtschaft 43 (1996) 37 – 39.
page 30 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

72. R. Gromes, D. Tanzer, UmweltMagazin 25 (1997) 84 – 85.

[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 31 of 31
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Indigo and Indigo Colorants Save title to My Profile
Standard Article Topics
Email this page
Elmar Steingruber1 Print this page
1Frankenthal, Germany
SEARCH THIS TITLE
Copyright © 2004 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved.
DOI: 10.1002/14356007.a14_149.pub2 Advanced Product Search
Article Online Posting Date: August 15, 2004 Search All Content
Acronym Finder
Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Physical Properties
2. Chemical Properties
3. Occurrence and Raw Materials
4. Production
5. Environmental Protection
6. Quality Specifications
7. Chemical Analysis
8. Storage and Transportation
9. Uses
10. Commercial Forms
11. Economic Aspects
12. Toxicology and Occupational Health
13. Indigo Derivatives

[Top of Page]

Indigo (1) is a natural dye obtained from the plants Indigofera tinctoria and Isatis tintoria, which have been used to color
textiles in India, China, and Egypt for 4000 years. Since its discovery, the cultivation, extraction, and use of indigo have
spread throughout the world, and trade in indigo became a significant economic factor, reaching a volume of about 8000 t in
1900.

With the development and technical realization of chemical syntheses at the end of the nineteenth century, indigo became an
important product of the growing chemical industry. After the proof of its structure by ADOLF VON BAEYER in 1883,
experiments to prepare indigo synthetically culminated in 1887 in the commercialization and large-scale production of
synthetic indigo by the Badische Anilin- und Soda-Fabrik (now BASF Aktiengesellschaft) in Ludwigshafen.

Today, more than 100 years later, synthetic indigo is still an important product of the dye manufacturing industry and has
almost completely replaced the natural indigo from plants. Synthetic indigo is the largest single textile dye in the world,
supplied by only a small number of producers to over 500 dye houses, whose main product is blue denim (i.e., cotton twill
weave where the warp is dyed with indigo). Indigo is used for the manufacture of approximately 109 pairs of blue jeans
annually. The apparently unending life cycle of indigo is, in view of its moderate fixation on the fiber, a phenomenon that is
difficult to understand.

Worldwide acceptance of easily washed out, faded, indigo blue garments, not only as work clothes but also as fashionable
sport and leisure articles, opened up new markets for indigo and ensured increasing demand for dyers and outfitters.

page 1 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Apart from indigo itself, only few derivatives and related compounds, primarily halogenated indigo and thioindigo, have
reached a position of steady economic importance (see Chap. Indigo Derivatives).

The considerable list of publications concerning indigo began nearly 170 years ago [1] and is indicative of the unusual
cultural, as well as chemical, position occupied by this compound up until recent times [2-10].

[Top of Page]

1. Physical Properties

Indigo [482-89-3], indigotin, ( 2,2′-biindoline)-3,3′-dione, 2-(1,3-dihydro-3-oxo-2, H-indol-2-ylidene)-1,2-dihydro-3, H-indol-3-

one, C16H10N2O2, Mr 262.27, C.I. Vat Blue 1, C.I. 73000 exists at ambient temperature and pressure as dark blue–violet
needles or leaves with a reddish bronze metallic appearance. It sublimes above 170 °C (mp 390 – 392 °C). Indigo is
practically insoluble in water, aqueous acid, aqueous base, and nonpolar solvents, but slightly soluble in high-boiling polar
solvents, (e.g., aniline, nitrobenzene, phenol, phthalic anhydride, dimethyl formamide, and dimethyl sulfoxide); it crystallizes
in pure form from hot solutions. The color of indigo is unusually intense, compared to conjugated molecules of similar size
[11], and is influenced strongly by solvents. Despite the possibility of cis – trans isomerism about the central C C double
bond, only the trans isomer, stabilized by intra- and intermolecular hydrogen bonding, is observed [12, 13]. This fact also
explains the extremely low solubility, high melting point, and vibrational and electronic absorption spectra of indigo. Bond
orders and charge densities of the indigo chromophore have been investigated and calculated [12, 14, 15] as has the X-ray
crystal structure [16, 17].

[Top of Page]

2. Chemical Properties
Indigo is very resistant to light and elevated temperature, even in the presence of air. Substitution by electrophiles or
nucleophiles is difficult; only sulfonation in concentrated sulfuric acid, to give the di- and tetrasulfonic acid, and halogenation
in nitrobenzene, to introduce up to six halogen atoms, are successful (See Chap. Indigo Derivatives). With strong mineral
acids, indigo forms salts that decompose in water.

Reducing agents (e.g., sodium dithionite, hydroxyacetone, zinc, hydrogen or biological material in fermentation, recently also
electric current [18, 19]) convert indigo to the leuco form indigo white [6537-68-4](2), which dissolves in sodium hydroxide
solution to form the disodium salt [894-86-0] (3). This is the effective dyeing agent (vat indigo) which penetrates into the fiber
and, after reoxidation adheres rather superficially to the surface of cotton fibers; in contrast, bonding to the polypeptide fibers
of wool or silk is more saltlike.

Leuco indigo can be di- and tetraacetylated as well as methylated at the nitrogen and oxygen atoms.

The action of oxidizing agents leads to dehydroindigo (4); oxidation with permanganate or chromate cleaves the molecule at
the central double bond, forming isatin [91-56-5] (5).

Concentrated alkali at elevated temperature leads to decomposition of indigo with formation of aniline, N-methylaniline, and
anthranilic acid.

page 2 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
In biological systems, indigo is extremely stable. For example, in a biological wastewater treatment plant, indigo is not
degraded, but simply adsorbed to the activated sludge, probably as a result of its low solubility; the elimination is 90–100 %
[37]. This fact accounts for the low acute and chronic toxicity of indigo (see Chap. Toxicology and Occupational Health).

[Top of Page]

3. Occurrence and Raw Materials

Natural indigo occurs worldwide in numerous types of Indigofera, particularly in Indigofera tinctoria (in Asia and America),
and in Cruciferae such as Isatis tinctoria (dyer's woad, in Europe). The concentration of indican [487-60-5] (6), a glucoside of
indoxyl found in the stems and leaves, amounts to only 0.2 – 0.8 %. After indoxyl is released with enzymes, it oxidizes rapidly
to indigo.

Before the introduction of synthetic indigo in 1897 by BASF, the cultivation of indigo plants, especially in India, on an area of
more than 7000 km2 was a huge business with a reported output of 19 000 t of indigo in 1897. The natural indigo production
collapsed to 1000 t per year in 1914.

For the industrial synthesis of indigo, only two starting materials are now important. The first is N-phenylglycine [103-01-5] (7)
produced from aniline and chloroacetic acid, or by reaction of aniline with formaldehyde and sodium cyanide or hydrogen
cyanide followed by saponification of the nitrile (8). In an alternate reaction, the addition product 9 of aniline, formaldehyde,
and sodium bisulfite is treated with sodium cyanide to give the N-phenylglycine nitrile 8. Another route to the nitrile 8 is the
reaction of aniline with hydroxyacetonitrile (10) [20-22].

All four synthetic routes were used industrially [23]: the aniline-formaldehyde-hydrogen cyanide route by DyStar (which took
over the textile dye business from BASF, Germany), the aniline-formaldehyde-sodium cyanide variation by Buffalo Color
Corp. (USA; the plant has been closed in the meantime), the bisulfite version by Bann Química (Brazil), and the chloroacetic
acid route by Chinese manufacturers. ICI and Mitsui have dicontinued their indigo manufacture.

The second raw material still used for the industrial synthesis of indigo is anthranilic acid [118-92-3] (11), which is either
condensed with chloroacetic acid and alkali to give the salt of N-phenylglycine-o-carboxylic acid, or converted to the nitrile
(12) with formaldehyde and hydrogen cyanide; subsequent saponification yields (13). The N-phenylglycine-o-carboxylic acid
process was applied by BASF in 1897 to produce the first synthetic indigo on an industrial scale and employed in
Ludwigshafen until 1924. The ring closure used in this process to form indoxyl is easier because no sodium amide is
required, but the overall cost is considerably higher than the process based on aniline. The anthranilic acid route to
synthesize indigo is therefore no longer of any commercial importance.

Other intermediates used in the pioneering phase of synthetic indigo were N-(2-hydroxyethyl)aniline and N-ethyl-N-
phenylglycine; the use of these starting materials ceased in 1914 and 1904, respectively. The numerous other synthetic
routes to indigo have no commercial importance. The Baeyer – Drewsen process of 1882 is no longer in use; in this method,

page 3 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2-nitrobenzaldehyde was used to synthesize 4,4′-dichloroindigo, which is inaccessible by direct chlorination.

During the first 20 years of synthetic indigo manufacture, the decision as to which intermediate to use depended on the
availability of the petrochemical raw materials benzene and naphthalene. These two chemicals, then accessible only from
coal tar, were converted to aniline via nitrobenzene and anthranilic acid via phthalic anhydride, respectively. Today, the
decision is influenced predominantly by the process technology available and the production costs; in this respect, the aniline
process requires a higher expenditure for process equipment, but consumes cheaper raw materials.

Since about 1980, attempts to manufacture indigo by biosynthetic methods from glucose have been successful, using
recombinant microorganisms [24-26]. The disadvantage of handling dilute solutions in huge bioreactors, the difficulties of
separating indigo from biomass, and the treatment of large volumes of organic byproducts prevented a commercialization of
this process.

[Top of Page]

4. Production
The extraction and isolation of natural indigo are described in detail in [27], [28]. The indigo plants are cut and fermented in
water for 15 h. The glycoside indican (6) is enzymatically cleaved to glucose and indoxyl. Oxidation of indoxyl with air yields
indigo. The most important producer countries were India (particularly Bengal), Java, and Guatemala. Purification, based on
reduction and reoxidation, or fractional precipitation from sulfuric acid, was usually necessary because of the low purity of the
natural indigo of only 15–70 %.

The rapid success of synthetic indigo at the beginning of the 20th century had a devastating effect on the plantations of
natural indigo; by 1925 the production in India had dropped to only a few percent.

For the industrial synthesis of indigo the most important and also the most economic production process currently in use is
ring closure of N-phenylglycine with sodium amide [7782-92-5] in anhydrous sodium hydroxide and potassium hydroxide
solvent. The original process used by KARL HEUMANN in 1890 did not employ sodium amide and gave an indigo yield of only
ca. 10 % [3].

After the observation by J. PFLEGER in 1901 that the addition of sodium amide resulted in a high yield of indigo, this process
was introduced industrially by Farbwerke Hoechst. The process was also used by BASF in Ludwigshafen in 1905 with
cheaper calcium oxide as desiccant. When the price of sodium dropped sharply, BASF changed to sodium amide in 1926
[29].

The Heumann – Pfleger process provides indigo of adequate purity in yields up to 90 %. Investigation of reaction conditions
and raw material ratios is described in [2]. Ring closure of alkali salts of N-phenylglycine is carried out with about 2 mol of
sodium amide at ca. 200 °C and releases ammonia. The indoxyl formed in the anhydrous melt exists as the stable dialkali-
metal salt (14).

After ring closure, the melt is dissolved in water; the unstable indoxyl monoalkali salt is converted to indigo by oxidation with
air. Some decomposition occurs, leading to aniline, N-methylaniline, and anthranilic acid as byproducts. Further details
concerning the indoxyl melt are described extensively in [3, 23, 27-34].

The crystalline indigo formed by oxidation is separated from the aqueous alkali by filtration and washed with water. The usual
commercial forms are indigo powder, prepared by drying the aqueous filter cake, and reduced indigo in the form of indigo vat
40 % solution and 60 % grains; some manufacturers offer a paste or liquid containing 20 % indigo. The filtrate of the
oxidation mixture can be concentrated and the alkali used again for the indoxyl melt. The alkali solution concentrated to 50%
is filtered to remove precipitated anthranilic acid, sodium carbonate, and other impurities. Recovery of the alkali makes the
process more cost attractive because only a small amount of alkali is lost from the production cycle.

The production of indigo based on the more expensive anthranilic acid via N-phenylglycine-o-carboxylic acid (13) or its salts
is simpler because no anhydrous alkali or dehydrating agent like sodium amide is required. Nevertheless, the reaction
mixture is more viscous and therefore requires considerably more alkali to carry out the fusion [27, 28, 35].

page 4 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

In this process, ring closure to the indoxyl carboxylic acid also occurs at ca. 200 °C. The yield is 70 – 90 %, depending on
reaction conditions and the quality of the dicarboxylic acid (13). The fusion mass is dissolved in water, decarboxylation and
oxidation by air yields indigo. Isolation of indigo is similar to that in the N-phenylglycine process; the product quality is also
comparable.

The indoxyl fusion of N-phenylglycine (7) under anhydrous conditions can be carried out in carbon steel vessels, which suffer
practically no corrosion. Stainless steel is preferred for the considerably more aggressive water-containing melt from N-
phenylglycine-o-carboxylic acid (13). Both indoxyl oxidation and isolation of indigo use normal steel, which is sufficiently
resistant to dilute caustic alkali.

In addition to the two processes described above, another synthetic route starts from N-(2-hydroxyethyl)aniline. Although an
easily accessible raw material is involved and the process was used by BASF from 1907 to 1914, it is no longer of industrial
importance. A considerably higher temperature is required, at which hydrogen is released spontaneously; the yield and purity
of the indigo are lower. Further process improvements by Farbwerke Hoechst [23] could not revive this route.

Indigo is also marketed as the leuco indigo disodium salt 3. Prepared originally by reduction of indigo with zinc or iron, it is
now obtained by catalytic hydrogenation of indigo. Until 1980 indigo white (2) was marketed as a 50 % paste and used as a
syrupy product, predominantly in the Far East; it is no longer in use. Indigo vat 60 % is a granular product obtained by drying
a mixture of the sodium salt 3 and molasses, adjusted to an indigo content of 60 %. Since 2000, also a 40 % solution of
indigo vat is manufactured and sold; the prereduced indigo gained interest in the dye houses by saving costs and reducing
the salt content in the effluent.

[Top of Page]

5. Environmental Protection
Because of their chemical properties and toxicity, the chemicals used in the production of N-phenylglycine and indigo
(particularly aniline, formaldehyde, hydrogen cyanide, potassium or sodium hydroxide, sodium, ammonia, and sodium
amide), demand high standards for safe handling and processing. These are achieved by extensive use of closed systems.

Indigo possesses very low mammalian toxicity [36], [37] (see Chap. Toxicology and Occupational Health) and does not have
an adverse effect on adapted activated sludge systems (e.g., in rivers or biological effluent treatment plants). The rate of
biological degradation is quite low because of its low solubility, but indigo is adsorbed on activated sludge and thereby
eliminated from wastewater.

Byproducts from indigo synthesis may be present in wastewater resulting from filtrates, wash water, and flue-gas scrubbers;
the wastewater must then be purified. The byproducts aniline and anthranilic acid are easily degraded in biological water
treatment plants, but N-methylaniline is degraded more slowly. The ammonia released in the sodium amide process and in
the saponification of the nitriles 8 or 12 is preferably recovered but can also be removed from waste streams by nitrification.
Any alkali reaching the wastewater must be neutralized with acid.

The extraction of natural indigo also involves ecological problems, because 98 % of the dry plant material do not consist of
indigo and have to be degraded in the fermentation process or deposited in the indigo plantations.

[Top of Page]

6. Quality Specifications
Indigo is available in the following forms: technically pure powder, microgranules, or lumps considered to contain 100 %
indigo; liquid dispersions in water with an indigo content of 20–30 %; indigo vat 60 % grains and indigo vat 40 % solution.

The powder and microgranules still contain traces of byproducts from the synthesis, together with small amounts of additives
that influence the dispersibility. The following parameters serve as quality criteria: indigo content, dusting and wetting
properties, reduction rate, properties of the vat solution, and dyeing characteristics.

With the liquid dispersions, which usually contain 20 – 30 % indigo and are free-flowing liquids, quality criteria include indigo
content, alkali content (0 – 6 %), viscosity, specific gravity, storage stability, sedimentation behavior, and those parameters
used for the powder (i.e., reduction behavior, the quality of the reduced solution, and dyeing properties).

Indigo vat 60 % grains and indigo vat 40 % solution are tested for indigo, alkali, and water content, and for the amount of

page 5 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
reoxidized material.

For natural indigo, the most important quality criteria are its purity, solubility after reduction, and dyeing behavior.

[Top of Page]

7. Chemical Analysis
Numerous methods have been suggested for the determination of indigo content, some of which have been developed in the
era of natural indigo [27], [32]. The preferred method consists of sulfonation of an indigo sample with concentrated sulfuric
acid, followed by titration of the diluted sulfonic acid with sodium dithionite solution [38], [39]. Indigo sulfonic acid can also be
determined spectrophotometrically [40]; highly purified indigo, obtained by extraction with nitrobenzene or phenol, or by
recrystallization, is used as a standard. Other quality criteria are discussed in Chapter Quality Specifications. The effective
indigo content of indigo vat 40 % and 60 % is obtained after air oxidation of the vat solution and filtration, and of liquid paste
indigo products, after extraction with dilute mineral acids and filtration.

[Top of Page]

8. Storage and Transportation

No specific limitations have been placed on indigo because of the low toxicity. The danger of dust explosions must be
considered in storing indigo powder (dust explosion class 2). With a particle size of (20 – 60)×10–6 m, ignition may occur at
concentrations exceeding 50 g/m3. Indigo powder, indigo grains, indigo vat, and neutral liquid products are similarly regarded
as nonhazardous for transportation [37]. Alkaline liquid forms with 1 – 5 % alkali are regarded as corrosive for transportation
because of the action of alkali on aluminum [41].

[Top of Page]

9. Uses
Only small quantities of indigo are used for dyeing woven fabrics of cotton, wool, and silk (e.g., wax batik for fashionable
hand-crafted objects). The greater part of the indigo produced worldwide is used in continuous dyeing of cotton yarn. The
dyed cotton warps are woven with undyed weft in various twill weaves to give the typical blue denim cloth. The predominantly
heavy, sailcloth-like material is used by about 400 producers to manufacture mainly blue jeans. The jeans “phenomenon”
was examined in detail in 1981 from historical, economic, fashion, artistic, and business perspectives [42, 43]. Thanks to the
popularity of cotton clothes dyed with indigo (of which ca. 90 % are blue jeans and only ca. 10 % other fashion articles
including printed textiles), the consumption of indigo has attained a continuous growth rate worldwide since about 1975,
which is above average of the textile industry.

In dye houses specializing in indigo dyeing, continuously operated slasher or warp dyeing machines are used in which the
cotton yarn is immersed successively in three to six dye baths of vatted indigo. Indigo is reduced to the soluble leuco form
with alkali and hydrosulfite, or pre-reduced indigo vat is used, or electric current is used in specialized equipment to reduce
the indigo [18, 19]. After each immersion into the indigo vat, the yarn is squeezed and the adsorbed indigo oxidized by air
[44]. Worldwide, more than 100 dye houses run about 200 dyeing machines, with about 70 % being in the Far East,
particularly the Hong Kong area, followed by the United States and Europe. In 1980, a “Loop Dye 1 for 6” dyeing machine
was introduced, in which cotton yarn passes through the same vat up to six times [45].

[Top of Page]

10. Commercial Forms

Commercial products based on indigo include the following:

DyStar Indigo gran.

Indigo Bann 20 (paste)
Indigo Bann 30 (paste)
DyStar Indigo Vat 60 % Grains
DyStar Indigo Vat 40 % Solution
DyStar Indigo Black 5006 (mixture of indigo with a yellow dye)

[Top of Page]

11. Economic Aspects

From 1960 to 1980, the growing demand for blue jeans resulted in a steady increase in the indigo production capacity
worldwide. By 1988, the world capacity had reached 14 000 t per year of synthetic indigo and was still distributed over more
than ten countries, 12 000 t were supplied by the five major manufactureres BASF (Germany), Buffalo Color Corp. (USA), ICI
(UK and Brazil), Mitsui Toatsu (Japan) and Taisung (Taiwan). Smaller quantities were produced in China, the Soviet Union,
Mexico, Korea, and India.

Fierce price competition and globalization of world trade started to cause major changes, beginning about 1990. Local indigo
production was started in Brazil by Bann Química, and low price indigo from China began to compete on all markets. In 1996,
the Zeneca indigo plant in the UK (formerly ICI), was sold to BASF, which then reached a total capacity of 7000 t of indigo

page 6 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
per year. In 2000, BASF included its textile dye business in the joint venture DyStar, which consisted of the textile dye
activities of Bayer and former Hoechst. In the meantime, the UK plant was closed, Mitsui terminated their indigo production in
Japan, and in 2004 also Buffalo Color Corp. in the USA ended their operations. The world indigo consumption had reached
about 30 000 t in 2002.

The remaining indigo suppliers in the world are DyStar with a 7000 t/a capacity, Bann Química in Brazil with about 4000 t/a,
and manufacturers in China who constructed huge new capacities, an estimated 40 000 t/a or more. The indigo production
coming from China into the world market was expected to amount to 14 000 – 18 000 t in 2003.

The consumption of indigo for dyeing cotton yarn amounts to only 1.5 – 3 % of the weight of the cotton. The weight of blue
denim varies from ca. 150 to 450 g/m2. With 1.5 m2 of material per pair of trousers, only 3 – 12 g of indigo are used for each
pair of blue jeans. The world consumption of 30 000 t/a of indigo in 2003 allows the production of 2–4 × 109 pairs of blue
jeans with an enormous sales value, of which indigo itself represents only a small fraction.

As a result of the enormous indigo production capacities in China and the export pressure onto the world market, indigo
market prices have suffered a steep drop from $ 15 – 20 per kilogram in 1988 to $ 7 – 8 per kilogram in 2003 in Europe;
Chinese indigo is even offered at less than $ 4 per kilogram.

The estimated distribution of the indigo consumption in 2002 is shown in Table 1. Since 1988, the demand has increased
particularly in Turkey, India, and China. The Far East continues to be the largest growing indigo market.

Table 1. Indigo consumption in 2002 (t)

Europe 3300
USA, Canada 3300
Central America 2700
South America 2700
Turkey, Near East, Middle East 2000
Africa 480
India, Pakistan 1950
China 16 000
Rest of Asia 6300

Total 38 780

[Top of Page]

12. Toxicology and Occupational Health

Despite the long use of natural and synthetic indigo, only few toxicological data were known until a few decades ago. Since
1962, however, the toxicological properties of indigo have been increasingly examined and reported, in connection with the
application for registration of purified indigo as D & C Blue No. 6 for use in pharmaceuticals and cosmetics [36].

The low acute oral toxicity (LD50> 5000 mg/kg) has been confirmed in mammals [36] fish toxicity LC50 is > 1000 mg/L (96 h,
Leuciscus idus) [37]. Experiments to determine the acute dermal and inhalation toxicity demonstrated no toxic effects.
Subchronic and chronic feeding experiments on rats and dogs with purified indigo at up to 3 % of feed weight showed, after a
certain adaptation time, no serious negative effects other than coloring of the organs. Similarly, no negative effects were
found in rats over three generations in studies of reproduction or in teratogenic investigations. Repeated application of
purified indigo to the skin yielded no negative effects. With alkaline forms of commercial indigo, the corrosive action of the
alkali can be observed. No indication of sensitization in humans was found after repeated skin application of neutral indigo.

The Ames test for mutagenicity towards bacteria was applied to synthetic indigo in 1979 and 1982 and revealed a minimal
positive mutagenic effect, which was later found to be due to impurities [46-48].

[Top of Page]

13. Indigo Derivatives

In comparison with the use of indigo as a special product for dyeing blue jeans, other indigoid dyes (e.g., thioindigo and
substituted indigo derivatives) have considerably less economic importance.

Thioindigo [522-75-8], C.I. Vat Red 41, C.I. 73 300, C16H8O2S2, Mr 296.37, available from thiophenol-o-carboxylic acid and
chloroacetic acid via 2-hydroxythionaphthene (15), is not used for cotton dyeing because of its poor color fastness and
limited affinity for cotton, but is a brilliant red disperse dye for polyester (C.I. Disperse Red 364), marketed by DyStar.

4,7,4′,7′-Tetrachlorothioindigo, obtained from the dichloro compound 16 by cyclization or by direct chlorination of thioindigo,
is an industrially important pigment (see Pigments, Organic – Thioindigo Pigments).

page 7 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Among the halogen derivatives, Tyrian purple was obtained from historic times to the Middle Ages from various sea snails.
Despite its dull red-violet shade, this dye was much sought after. In 1909 the compound was identified as 6,6′-dibromoindigo
[19201-53-7], but it cannot be prepared by direct bromination of the parent indigo.

The halogenated indigo dyes still in use today are specialty products; their sales volume amounts to much less than 1 % of
the unsubstituted indigo products.

5,7,5′,7′-Tetrabromoindigo [2475-31-2] , C.I. Vat Blue 5, C.I. 73 065, is synthesized by bromination of indigo in acetic acid.
On account of its interesting neutral blue color the product is still sold as DyStar Brilliant Indigo 4 B (for dyeing) and DyStar
Indigo 4B sfx. (for printing). The commercial forms consist of a mixture of tri- and tetrabromoindigo with a low chlorine
content.

5,5′-Dibromo-4,4′-dichloroindigo, C.I. Vat Blue 2, C.I. 73045, obtained by bromination of 4,4′-dichloroindigo and sold as
Brilliant Indigo 4 G, had been a desirable dye for printing because of its interesting greenish blue shade. It is no longer
commercially available.

Among the sulfonation products of indigo the disodium salt of 5,5′-indigodisulfonic acid C16H8N2Na208S2 is known as
indigotine I or indigo carmine [860-22-0], C.I. Acid Blue 74, C.I. 73015, Mr 466.36.

page 8 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Registered also as C.I. Food Blue 1 and in the USA as FD&C Blue 2, it has found an application as colorant for food,
cosmetics, drugs, and animals feeds. It is marketed by BASF as Sicovit Indigotin 85 E 132.

[Top of Page]

References
1. Dumas, Annalen der Pharmacie 21 (1837) 72 – 77. Links
2. F. Henesey, J. Soc. Dyers Colour. 54 (1938) 105 – 115. Links
3. K. Holzach, Angew. Chem. A/60 (1948) 200 – 204. Links
4. H. Cassebaum, Melliand Textilber 48 (1967) 207. Links
5. M. Seefelder: Indigo: Kultur, Wissenschaft und Technik, ecomed, Landsberg, 1994.
6. H. Schmidt, Chemie in unserer Zeit 31 (1997), 121 – 128. Links
7. A. Dutly: Geschichte des Indigos, ETH Zürich, 2002, http://www.dutly.ch/indigohtml/indigo 1.html.
8. P. Miederer in: K. Hunger (ed.), Industrial Dyes: Chemistry, Properties, Applications. Wiley-VCH, Weinheim, Germany,
2003, p. 40 – 43, 204 – 214.
9. H. Zollinger: Color Chemistry, 3rd. rev. ed. VHCH Verlag Helvetica Chimica Acta, Zürich, and Wiley-VCH, Weinheim,
Germany, 2003, p. 259 – 268.
10. M. Seefelder: Indigo in Kultur, Wissenschaft und Technik, BASF Aktiengesellschaft, Ludwigshafen am Rhein 1982.
11. W. Lüttke, M. Klessinger, Chem. Ber. 97 (1964) 2342 – 2357. Links
12. M. Klessinger, W. Lüttke, Tetrahedron 19 (1963) Suppl. 2, no. 315 – 335. Links
13. M. Klessinger, W. Lüttke, Chem. Ber. 99 (1966) 2136 – 2145. Links
14. E. Wille, W. Lüttke, Angew. Chem. 83 (1971) 853; Links Angew. Chem. Int. Ed. Engl. 10 (1971) 803. Links
15. E. Wille, W. Lüttke, Chem Ber. 106 (1973) 3240. Links
16. H. v. Eller, Bull. Soc. Chim. Fr. 1955, 1426 – 1444. Links
17. M. Klessinger, Dyes and Pigments 3, (1982) 235. Links
18. W. Schrott T. Bechtold, Textilveredlung 2003 (9/10), 19 – 22. Links
19. T. Bechtold E. Burtscher O. Bobleter, Recent Res. Dev. in Electrochem. 1998, 245 – 264. Links
20. E. Knoevenagel, Ber. Dtsch. Chem. Ges. 37 (1904) 4080. Links H. Bucherer, Ber. Dtsch. Chem. Ges. 39 (1906) 2800.
Links
21. H. Bucherer, Ber. Dtsch. Chem. Ges. 39 (1906) 2800. Links
22. BASF, GB 1 119 256, 1966 (E. Tolksdorf).
23. J. G. Kern, H. Stenerson, Chem. Eng. News 24 (1946) 3164 – 3181. Links
24. D. Murdock, B. D. Ensley, C. Serdar, M. Thalen, Biotechnology 1993 11, 381. Links
25. J. W. Frost, J. Lievense, New J. Chem. 1994 18, 341. Links
26. H. Bialy, Nature Biotechnol. 1997 15, 110. Links
27. Ullmann, 1st ed. 6, 505 – 508.
28. Ullmann, 2nd ed. Urban & Schwarzenberg, Berlin, Wien 1930, 6, 244 – 246.
29. Ullmann, 3rd ed., Urban & Schwarzenberg, München, Berlin 7, 1956, 113 – 114. 8 (1957) 748 – 763.
30. Kirk-Othmer, 2nd ed., 7, (1965) 621 – 625; 11, (1966) 562 – 580, Interscience Publishers, New York.
31. Kirk-Othmer, 3nd ed., 8, 363 – 368;
32. K. Venkataraman: Chemistry of Synthetic Dyes, vol. 2, Academic Press, New York 1952, pp. 1003 – 1022.
33. Ullmann, 4th ed, 13, 177 – 181.
34. Ullmann's, 5th ed., A 14, 149 – 156.
35. M. Phillipps, J. Ind. Eng. Chem. 13 (1921) 759 – 762. Links
36. K. H. Ferber, J. Environ. Pathol. Toxicol. Oncol. 7 (1987) 73 – 83. Links
37. Sicherheitsdatenblatt DyStar Indigo gran, DyStar Textilfarben GmbH & Co., Leverkusen, 01.08.2002.
38. BASF Indigo rein, BASF Aktiengesellschaft Ludwigshafen am Rhein 1907, p. 14.
39. Badische Anilin- und Soda-Fabrik Z. Farben Ind. 8 (1909) 121 – 127, 137 – 141. Links
40. Lunge-Berl: Chemisch-technische Untersuchungsmethoden, 7th ed., Springer Verlag, Berlin 1924, pp. 1033 – 1059.
41. Sicherheitsdatenblatt DyStar Indigo Vat 40% Lsg., DyStar Textilfarben GmbH & Co., Leverkusen, 03.09.2002.
42. B. Wehdemeier-Pusch, A. Pusch, Das Phänomen “Jeans”, Universität Bielefeld 1981.
43. Blaue Wunder für individuelle Jeansmode, BASF-Gruppe, 09.07.2003, http://www.pressi.com/de/Mitteilung/69301.html.

page 9 of 10
Indigo and Indigo Colorants : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

44. Process Data Sheet: Continuous Dyeing of Cotton Warps with Indigo, BASF Aktiengesellschaft, Ludwigshafen am Rhein
1976 (German 1983).
45. B. Kramrisch, Am. Dyest. Rep. (Nov. 1980) 34 – 38.
46. B. N. Ames, J. McCann, E. Yamasaki, Mutat. Res. 31 (1975) 347 – 364. Links
47. J. M. Muzzall, W. L. Cook, Mutat. Res. 67 (1979) 1 – 8. Links
48. W. F. Jongen, G. M. Alink, Fd. Chem. Toxic. 20 (1982) 917 – 920. Links
[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 10 of 10
 Indium and Indium Compounds 1

Indium and Indium Compounds

Noël Felix, Metallurgie Hoboken-Overpelt, Hoboken, Belgium

1. Introduction . . . . . . . . . . . . . . . . . 1 4.5. Refining . . . . . . . . . . . . . . . . . . . . 7

2. Properties . . . . . . . . . . . . . . . . . . 1 5. Quality Specifications . . . . . . . . . . . 7
3. Resources and Raw Materials . . . . . 2 6. Analysis . . . . . . . . . . . . . . . . . . . . 7
4. Production . . . . . . . . . . . . . . . . . . 2 7. Indium Compounds . . . . . . . . . . . . 7
4.1. Production from Zinc Circuits . . . . . 3 8. Uses . . . . . . . . . . . . . . . . . . . . . . 8
4.2. Production from Lead Circuits . . . . 5 9. Economic Aspects . . . . . . . . . . . . . 10
4.3. Production from Tin Circuits . . . . . 6 10. Toxicology and Occupational Health . 11
4.4. Other Production Developments . . . 6 11. References . . . . . . . . . . . . . . . . . . 11

1. Introduction Some physical properties of indium are as

follows [3], [5], [7]:
Indium [7440-74-6], In, atomic number 49, Ar
Electronic configuration [Kr] 4d 10 5s2 5p1
114.82, is a metallic element of group 13 of the Ar 114.82
periodic table. Thermal neutron cross section
Indium has two natural isotopes: 115 In at 2200 m/s
(95.72 %) and 113 In (4.28 %). The abundance of absorption (194 ± 2)
×10−28 m2 /atom
indium in the earth crust is comparable to that scattering (2.2 ± 0.5)
of silver, i.e., 0.1 ppm [1]. ×10−28 m2 /atom
Indium was discovered in 1863 by F. Reich Crystal structure tetragonal A 6
and T. H. Richter during spectrometric analy- ao = 0.458 nm
co = 0.494 nm
sis of sphalerite ores at the Freiberg School of Number of atoms per unit cell 4
Mines in Germany [2]. Indium is named after Atomic radius (coordination
the indigo blue spectral lines that led to its iden- number 12) 0.162 nm
Ionic radius (In3+ ,
tification.
coordination number 6) 0.081 nm
Atomic volume 15.73×10−6 m3 /mol
Density at 293.15 K 7.310 g/cm3
2. Properties [3–8] mp
bp
429.75 K
2353.15 K
Enthalpy of fusion 3.26 kJ/mol
Physical Properties. Indium is a crystalline, Enthalpy of vaporization 231.2 kJ/mol
silvery white metal, which is very soft (softer Specific heat at 298.15 K 26.70 kJ mol−1 K−1
than lead), ductile, and malleable. Indium re- Entropy at 298.15 K 57.7 kJ mol−1 K−1
Entropy of fusion 7.58 kJ mol−1 K−1
tains its highly plastic properties at cryogenic Entropy of vaporization 98.56 kJ mol−1 K−1
temperatures. Indium can be deformed almost Coefficient of linear expansion
indefinitely under compression, is easily cold- (273.15 – 373.15 K) 3×10−5 K−1
welded, and, like tin, emits a high-pitched “cry” Vapor pressure p, kPa, at logp= 9.835
temperature T , K, between mp − 12860
T −0.7logT
on bending. and bp
Indium generally increases the strength, cor- Thermal conductivity
rosion resistance, and hardness of an alloy sys- (273.15 – 373.15 K) 71.1 W m−1 K−1
Surface tension at temperature
tem to which it is added. Even small concentra- T , N/m 0.602 − 10−4 T
tions of indium can have a considerable influ- Electrical resistivity
ence. at 3.38 K superconducting
at 273.15 K 8.4×10−8 Ωm
Molten indium wets clean glass. Indium has a
temperature coefficient
low melting point (429.75 K) but a high boiling (273.15 – 373.15 K) 4.9×10−3 K−1
point (2353.15 K). Indium becomes supercon-
ducting at 3.37 K.

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 157
2 Indium and Indium Compounds
Standard electrode potentials fides. The amount of indium in sphalerite may
In3+ + 3e
In −0.338 V
In3+ + 2e
In+ −0.40 V
vary widely even within a deposit. The indium
In2+ + e
In+ −0.40 V concentration is typically 10 – 20 ppm, but may
In3+ + e
In2+ −0.49 V be as high as 1 % (10 000 ppm). A sphalerite
Brinell hardness 0.9 deposit with a very high indium content (0.2 –
Tensile strength 2.645 MPa
Modulus of elasticity 10.8 GPa 0.3 %) is the Huari Huari zinc deposit in Bolivia.
Tin deposits with high indium contents (up to
0.21 wt %) are found in Cornwall (United King-
dom) and New Brunswick (Canada).
Chemical Properties. Metallic indium is
not oxidized by air or oxygen at room tempera-
ture. It reacts directly with arsenic, antimony, the
halogens, oxygen, phosphorus, sulfur, selenium, 4. Production
and tellurium when heated.
During smelting processes for the recovery of
Indium dissolves only slowly in cold dilute
base metals, indium concentrates in byproducts
mineral acids, but more readily in hot dilute
such as residues, flue dusts, slags, and metal-
or concentrated acids. Alkalis do not attack the
lic intermediates. The processes for the recov-
massive metal.
ery and production of indium are often complex
Indium forms alloys with most other metals.
and sophisticated; they are characterized by a
Extensive solid solutions are formed with lead,
low direct extraction efficiency. Therefore, the
thallium, and mercury.
recycling of the indium contained in intermedi-
ate products is important.
3. Resources and Raw Materials [9]
Indium does not occur in the native state. It is 4.1. Production from Zinc Circuits
widely spread in nature, generally in very low A process for the recovery of indium from sec-
concentrations. The content of indium in the ondary zinc oxide is described in [10], [11]. The
earth crust is estimated to be 0.1 ppm. Indium process is illustrated in Figure 1. In the first step,
is found as a trace element in many minerals most of the zinc is removed by leaching with di-
(see Table 1). lute sulfuric acid. The residue is leached with di-
Table 1. Indium contents of some minerals [9] ∗ lute hydrochloric acid to dissolve the indium. Tin
is removed from the indium solution by neutral-
Mineral Composition Indium content, ization to pH 1. Further neutralization then pre-
ppm
cipitates indium. The indium residue is leached
with sodium hydroxide to give crude indium hy-
Sphalerite ZnS 0.5 – 10 000
Galena PbS 0.5 – 100
droxide as an intermediate product. Dissolution
Chalcopyrite CuFeS2 0 – 1500 of this indium hydroxide in dilute hydrochloric
Enargite Cu3 AsS4 0 – 100 acid gives an indium solution which is purified
Bornite Cu5 FeS4 1 – 1000 by cementation of copper and arsenic with iron,
Tetrahedrite (Cu,Fe)12 Sb4 S13 0.1 – 160
Covellite CuS 0 – 500 followed by cementation of tin and lead with
Chalcocite Cu2 S 0 – 100 indium. Indium is recovered from the purified
Pyrite FeS2 0 – 50 solution by cementation with aluminum.
Stannite Cu2 FeSnS4 0 – 1500
Cassiterite SnO2 0.5 – 13 500
In the process developed by the Ana-
Wolframite (Fe,Mn)WO4 0 – 16 conda Copper Mining Company (Fig. 2), indium
Arsenopyrite FeAsS 0.3 – 20 is recovered from zinc calcine (roasted zinc
∗ Reproduced from Roskill: The Economics of Indium 4th ed., blende) or zinc oxide fume [12]. Zinc calcine or
1987, by permission of the publisher, Roskill Information zinc oxide fume is first leached in sulfuric acid at
Services Ltd. moderate pH, dissolving the bulk of the zinc and
leaving indium in the residue. In the second step,
Sphalerite is the most important indium-con- indium is dissolved from the residue by leaching
taining mineral followed by lead and copper sul- with sulfuric acid (20 – 25 g/L H2 SO4 ).
 Indium and Indium Compounds 3

Indium is precipitated from the resulting solu-

tion by adding zinc oxide and sodium hydro-
gen sulfite. The indium precipitate is purified by
leaching with concentrated sodium hydroxide,
washing with water and with dilute sulfuric acid
to remove zinc. The purified residue is dissolved
in sulfuric acid, heavy metals are precipitated
with hydrogen sulfide, and indium is recovered
as a sponge by adding zinc. The indium sponge
may be melted with sodium hydroxide and cast
into bars. Alternatively, the sponge may be pu-
rified by dissolving it in hydrochloric acid and
adding barium chloride to remove sulfate, and
hydrogen sulfide to remove heavy metals. After
filtration, indium is recovered from the solution
by electrolysis.
The extraction of indium from sphalerite of
the Rammelsberg and Bad Grund deposits in
the Federal Republic of Germany is described in
[13]. On refining the retort zinc by distillation,
indium remains with the lead. After oxidation
of the lead residue, indium is recovered from
the litharge, which contains 1 – 4 % indium. The
litharge is leached with sulfuric acid, which dis-
solves 88 – 97 % of the indium. Indium is reco-
vered by addition of zinc slabs to give a crude
indium sponge containing ca. 95 % indium. The
crude indium is redissolved in sulfuric acid. The
resulting solution is purified first by cementa-
tion with indium strips to remove tin and then
with slabs of a zinc alloy, containing 1 % in-
dium and 0.3 % cadmium, to remove thallium
and cadmium. Indium with a purity of greater
than 99.99 % is recovered from the purified so-
lution by precipitation with aluminum strips fol-
lowed by melting under sodium hydroxide and
sodium cyanide.
Chlorination of the lead retort residue under a
molten salt cover for the extraction of indium has
been patented by Asarco [14]. The chloride slag
is dissolved in water and indium is recovered by
cementation on zinc or aluminum.
At the zinc refinery of Mitsubishi in Akita
[15], indium is recovered from the leach residue
(400 ppm In) of the zinc concentrate (120 ppm
In). The residue is treated by a sulfation roasting
process. Leaching of the sulfated product in sul-
furic acid dissolves about 60 % of the indium to
give a leach solution containing 50 – 70 mg/L of
indium. This solution is concentrated to an in-
Figure 1. Recovery of indium from secondary zinc oxide
[10], [11]
4 Indium and Indium Compounds
Figure 2. Indium recovery by Anaconda Copper Mining Company [12]
dium content of 2 g/L by double neutralization
with calcium carbonate. The enriched solution
is then subjected to solvent extraction followed
by back extraction with hydrochloric acid. The
aqueous solution is concentrated to 20 – 30 g/L
indium. Indium is recovered by immersing alu-
minum plates in the solution. Indium sponge ad-
hering to the aluminum plates is removed for
melting and casting into anodes. Indium metal
of 99.99 % purity is then obtained by electrore-
fining.
4.2. Production from Lead Circuits
At Cominco’s integrated lead – zinc smelter at
Trail in Canada, tin and indium accumulate in
the fumes of the continuous drossing furnace
[16]. The flow sheet for the process is illus-
trated in Figure 3. A mixture of fumes, coke,
and antimony-containing slag is melted in a
direct-fired rotary furnace. Arsenic is removed
by adding scrap iron to form iron speiss. The
residual tin – indium – antimony – lead alloy is
cast into anodes and electrolyzed in a fluorosil-
icate solution, giving a solder cathode (Sn – Pb
alloy) and indium – antimony anode slimes. The
slimes are roasted with sulfuric acid. Indium is
dissolved from the sulfated product by leaching
with water. Copper is removed from the indium
solution by cementation on indium sheets. In-
dium is then recovered from the solution as a
sponge by addition of aluminum or zinc. The
indium sponge is melted and electrorefined to
give indium of 99.99 % purity.
 Indium and Indium Compounds 5

ing coke, a mixture of lead and zinc chlorides is

added, and indium is recovered as indium chlo-
ride in the slag.
The chloride slag of indium, lead, zinc, and
tin is leached by wet grinding. The solution is
purified by cementation of lead and tin on in-
dium sheet. Indium is recovered as a sponge by
adding zinc.

Figure 3. Indium recovery by Cominco [16]

Previously published in the Proceedings of the Symposium
on Quality Control in Non-Ferrous Pyrometallurgical Pro-
cesses from the 24th Annual Conference of Metallurgists,
August, 1985. Reprinted with permission of the Canadian Figure 4. Indium recovery by Capper Pass [21]
Institute of Mining and Metallurgy. Reproduced by permission of the Institution of Mining and
Metallurgy, from P. Halsall: Indium–Extraction from Lead,
Indium is recovered from lead bullion by Zinc and Tin circuits (Fig. 9). Trans. Instn. Min. Metall.
(Sect. C: Mineral Process. Extr. Metall.) 97, 1988, C 98.
Centromin of Peru at the La Oroya smelter, and
by Metallurgie Hoboken-Overpelt (MHO) at the
Hoboken smelter. At MHO, indium concentrates
in a complex residue during refining of lead 4.3. Production from Tin Circuits
by the Harris process (→Lead) [17]. At the La
Oroya refinery, lead bullion is first drossed to At Capper Pass, chlorination of electrorefined
remove copper (→Lead), then tin and indium tin gives a tin chloride slag which contains 2.7 %
are removed at a higher temperature [18], [19]. indium [20]. This slag is treated as shown in Fig-
The tin – indium dross is reduced to metal us- ure 4 [21]. Most of the tin is precipitated in the
6 Indium and Indium Compounds
first neutralization stage. The precipitate from
the second neutralization stage, which contains
some indium, is recycled for indium recovery.
Indium is recovered from the filtrate as a sponge
by cementation with zinc. The indium sponge of
95 % purity is electrorefined to give a 99.5 % in-
dium cathode, which is recast as the anode for the
second electrorefining stage, in which indium of
99.97 % purity is produced.
4.4. Other Production Developments
Solvent extraction systems have been investi-
gated to extract indium from various solutions.
The use of a mixture of hydroxyquinolines
(Kelex 100 and LIX 26) to extract indium from
acidic or alkaline aqueous solutions has been re-
ported [22].
Di-(2-ethylhexyl) phosphoric acid
(D2EHPA) in kerosene, and tributyl phosphate
(TBP) in kerosene are used to extract indium
from the solution obtained by leaching lead-
containing residues with sulfuric acid [23]. The
concentration of arsenic, cadmium, zinc, cop-
per, antimony, and iron in the extract is greatly
reduced. For good separation, iron should be
present in the ferrous state.
Solutions of D2EHPA and TBP in kerosene
are also reported to extract indium from sulfu-
ric acid solutions that contain tin. The addition
of soluble fluorides to the acid solution prior to
solvent extraction inhibits the extraction of tin
[24].
A process for the separation of indium from
gallium and germanium in sulfuric acid solution
using solvent extraction is described in [25].
Supported liquid membranes are used to ex-
tract indium from a copper-dross leach solu-
tion [26], [27]. The feed solution contains in-
dium (0.5 g/L), copper (60 g/L), and sulfuric
acid (30 g/L). The extractant is a 40 % solution
of D2EHPA in Escaid 100 (a mixture of 80 %
aliphatic and 20 % aromatic long-chain hydro-
carbons). A microporous polypropylenic sup-
port is used.
The recovery of indium and gallium from
spent semiconductor material has also been re-
ported [28].
4.5. Refining
Sponge indium with a purity up to 99.5 % re-
quires upgrading for most uses. For the semi-
conductor industry, a purity of at least 99.9999 %
is required. Electrorefining is a commonly used
technique.
In electrorefining, cotton-bagged indium an-
odes and pure indium cathodes are used. Metal-
lic impurities collect as anode slime in the cot-
ton bag, which thus prevents their entering the
bulk electrolyte and contaminating the indium
deposit. The electrolyte is generally a chloride
solution, but the use of solutions of sulfate [29],
cyanide [30], fluoroborate [31], and sulfamate
[32] has also been reported.
A typical composition of the electrolyte is
20 – 75 g/L In, 80 – 100 g/L NaCl, 1 g/L glue,
pH 2.2 – 2.5. The cathode current density is 1 –
2 A/dm2 . Apart from the dendritic nature of the
indium deposit, the difficulty with electrorefin-
ing lies in the fact that impurities which gener-
ally accompany indium have standard potentials
close to that of indium (see Table 2).
Table 2. Standard potentials, in volts
Cd2+ /Cd −0.402
In3+ /In −0.338
Tl+ /Tl −0.336
Sn2+ /Sn −0.136
Pb2+ /Pb −0.126
For purities up to 99.999 %, multistage elec-
trorefining is required [33], or electrorefining
must be supplemented by vacuum distillation or
zone refining or both of them.
The difference between the melting point and
boiling point of indium is greater than for most
other elements. This property allows the elim-
ination of impurities such as cadmium, sulfur,
selenium, tellurium, zinc, lead, and thallium by
vacuum distillation [34].
Elimination of cadmium, copper, zinc, gold,
silver, and nickel by zone refining is described
in [35]. Intensive cooling is necessary because
of the low melting point of indium.
A method for refining indium to semicon-
ductor grade, in which a chelating agent is used
to absorb indium selectively from solution, has
been reported [36].

5. Quality Specifications
Qualities ranging from 99.97 % to 99.99999 %
indium are commercially available. Impurities
may be metallic (Ag, Cu, Pb, Sb, Sn, Cd, Tl,
Zn, Fe, Al, Ga) and nonmetallic (S, Cl, O, Se,
N). Their concentrations depend on the origin of
the indium-containing feed material and on the
production and refining process.
Commercial grades are 99.97 % (3N7),
99.99 % (4N), 99.999 % (5N), 99.9999 % (6N),
and 99.99999 % (7N).
6. Analysis
Spectroscopic analysis enables the detection of
indium down to 10 ppm due to its characteristic
lines in the indigo blue region at wavelengths
of 4511.36, 4101.76, 3256.09, and 3039.39 nm.
The quantitative analysis of indium in ores,
compounds, and alloys is described in [37].
Spark source mass spectrometry (SSMS) and
graphite furnace atomic absorption spectrome-
try (GFAAS) are used in the analysis of high pu-
rity indium (≥99.9999 %). About 40 elements
can be determined with detection limits of 10 –
30 ppb by spark source mass spectrometry [38].
7. Indium Compounds
The most common valence of indium is three.
Monovalent and bivalent compounds of in-
dium with oxygen, sulfur, and halogens are also
known. The trivalent indium compounds are the
most stable. The lower valence compounds tend
to disproportionate to give the corresponding
trivalent compounds and indium metal.
The properties of some indium compounds
are given in Table 3 [7], [39].
Indium(III) oxide, In2 O3 , is formed by ther-
mal decomposition of indium hydroxide, car-
bonate, nitrate, or sulfate in air, or by burn-
ing indium in air. It is a pale yellow, amor-
phous solid when formed at low temperature, but
forms red-brown crystals when it is prepared at
high temperature. The amorphous form is sol-
uble in acids, but not in alkalies. Indium triox-
ide is reduced to indium by hydrogen, ammo-
nia, sodium, aluminum, and carbon. Indium(II)
oxide [12136-26-4], InO, and indium(I) oxide
Indium and Indium Compounds 7
[12030-22-7], In2 O, are formed by reduction
of indium-(III) oxide under carefully controlled
conditions.
Indium hydroxide [56108-30-6], In(OH)3 , is
precipitated from solutions of indium salts by
adjusting the pH in the range 4 – 10.
Indium trichloride, InCl3 , is obtained by
burning indium in excess chlorine. It is a white,
crystalline substance which is easily sublimed,
and is very soluble in water. Indium dichloride,
InCl2 , is obtained by the reaction of indium with
hydrogen chloride at 200 ◦ C. Indium monochlo-
ride, InCl, is formed by passing indium dichlo-
ride vapor over heated indium. The lower chlo-
rides, InCl and InCl2 , disproportionate in water
to give indium trichloride and indium metal.
Indium orthoborate [13709-93-8], InBO3 , is
prepared by melting a mixture of boric oxide and
indium hydroxide.
Indium forms alkyl and aryl compounds such
as trimethyl indium [3385-78-2], In(CH3 )3 ; tri-
ethyl indium [923-34-2], In(C2 H5 )3 ; and triph-
enyl indium [3958-47-2], In(C6 H5 )3 . They are
made by treating indium with the corresponding
mercury alkyls or mercury aryls, or by reaction
of indium trichloride with a Grignard reagent.
The semiconducting compounds indium an-
timonide, InSb, indium arsenide, InAs, and in-
dium phosphide, InP, are prepared by direct com-
bination of the high-purity elements at high tem-
perature. Indium phosphide is also obtained by
thermal decomposition of a mixture of a trialkyl
indium compound and phosphine (PH3 ).
8. Uses
Table 4 summarizes the western-world con-
sumption of indium by end use [9], [40], [41].
Low-melting alloys represent a major use of
indium. Indium is added to solder alloys, mostly
composed of tin and lead, to increase thermal
fatigue resistance, and to improve malleability
at low temperature and corrosion resistance. In-
dium solders are used in electronics for fix-
ing semiconductor chips to a base, for assem-
bling semiconductor devices and hybrid inte-
grated circuits, and to seal glass to metal in vac-
uum tubes. These solders are mainly used in the
form of a paste, whose main components are Pb,
Ag, Cd, In, Sn, and an organic binder.
8 Indium and Indium Compounds
Table 3. Properties of some indium compounds [7], [39]

Compound Mr Density, g/cm3 mp, K bp, K Enthalpy of for- CAS Registry

mation, kJ/mol Number

In2 O3 277.63 7.180 2183 −925 [1312-43-2]

InCl 150.27 4.190 498 881 −186 [13465-10-6]
InCl2 185.73 3.620 509 796 −362.4 [13465-11-7]
InCl3 221.18 3.450 856 771 ∗ −535.0 [10025-82-8]
InAs 189.47 1215 − 57.7 [1303-11-3]
InSb 236.57 798 − 31.1 [1312-41-0]
InP 145.79 1328 − 75.2 [22398-80-7]

∗ Sublimes

Fusible indium alloys are usually based on
bismuth in combination with lead, cadmium, tin,
and indium.
Fusible indium alloys are used to bend thin-
walled tubes without wrinkling the wall or
changing the original cross section. The alloy
is cast in the tube, rendering it effectively solid.
After bending, the alloy can be melted out using
hot water.
Table 4. Estimated western-world consumption of indium in 1985
[9]
Use
Low-melting alloys
Bearings
Dental alloys
Nuclear reactor control rods
Low-pressure sodium lamps
Electrical contacts
Alkaline dry batteries
Phosphors
Semiconductors
Lasers, photodetectors, integrated
circuits (Ga – As In – P, In – P)
CCDs in infra-red video cameras
(image orthicons, In – Sb)
Liquid crystal displays
∗ 1987 data [40].
∗∗ 1986 data for Japan [41].
These alloys are also used in fire-control sys-
tems. Restraining links that hold alarm, water
valve, and door operating mechanisms in place
are soldered with a low-melting alloy. A rise in
temperature sufficient to melt the alloy results in
the system coming into operation.
Fusible alloys are used for making disposable
patterns and dies for the founding of ferrous and
non-ferrous metals.
They are also used as temperature indicators
in situations where other methods of tempera-
ture measurement are impracticable. Small rods
of alloys of known melting point are inserted
into equipment such as aircraft header tanks, test
bearings, and experimental rigs. The melting of
the alloy thus indicates the temperature attained
in particular parts of the test system.
Addition of indium to lead – tin bearings for
heavy-duty and high-speed applications pro-
vides particularly high resistance to fatigue and
Consump-
seizure. These bearings are used in piston-type
tion, t aircraft engines, high-performance automobile
engines (formula 1) and in turbo-diesel truck en-
12
9
gines.
7 The addition of indium to gold dental al-
3.3 loys improves their mechanical characteristics
4–5
and increases resistance to discoloring. Small
8
6.8 ∗ amounts of indium are used to improve the
6 ∗∗ machinability of gold alloys for jewelry.
Indium is used in nuclear reactor control-rod
5
alloys (80 % Ag, 15 % In, and 5 % Cd). The con-
sumption of indium in this field is estimated at
3.5 3.3 t/a for the replacement of control rods in ex-
7 isting nuclear power plants [9].
In low pressure sodium lamps ( SOX lamps),
indium is applied as a thin layer of indium – tin
oxide on the inside of the outer glass envelope.
The indium – tin oxide film increases the operat-
ing temperature and the efficiency of the lamp.
Low pressure sodium lamps, characterized by
their orange color, are mainly used in the United
Kingdom, Belgium, and the Netherlands. The
production of SOX lamps accounts for 4 – 5 t/a
indium [9].
Indium is now used as a corrosion inhibitor in
alkaline dry batteries as a substitute for mercury.
This allows the mercury content of the amalga-
 Indium and Indium Compounds 9

mated zinc powder used in their production to application as windshield coatings in automo-
be reduced from 7 % to 1 – 3 % [42]. A target of biles is in development, particularly in Japan.
very low mercury content, or even complete re- Indium antimonide is used in Hall generators
moval, is being pursued. The low-mercury zinc to measure magnetic fields. Solar cells based on
powders have corrosion and degassing charac- copper – indium diselenide are being developed
teristics which are similar to those of the 7 % [43]. Small amounts of indium compounds are
mercury powders. used as additives for lubricating oils, and to pro-
Indium orthoborate is used as a phosphor in duce yellow glass.
cathode-ray tubes, where it is applied as a thin
layer. Indium phosphors emit light for a longer
time than zinc sulfide phosphors but are more ex- 9. Economic Aspects
pensive. They are used particularly in computer
monitors where image stability is of great im- The world production of indium in 1986 has
portance. In Japan, consumption of indium for been estimated at 80 t. Table 5 lists the estimated
this application was 6 t in 1986 and 8 t in 1987. indium production by country and company for
Intermetallic compounds of indium find ap- 1986. The main producers are located in Bel-
plication as semiconductors. The development gium, Japan, France, and Italy. Important indium
of the second generation of optical fiber com- refiners are located in the United States (Indium
munication systems in the early 1980s with sub- Corporation of America, 25 t in 1986 [9]) and the
stantially lower levels of attenuation at wave- United Kingdom (Mining and Chemical Prod-
lengths near the infrared region of 1.3 µm and ucts, 10 t in 1986 [9]). The U.S: Bureau of Mines
1.55 µm has led to the development of new estimates world production in 1992 140 metric
semiconductor lasers and photodetectors. Laser tons. [45]
diodes based on indium – gallium arsenide –
phosphide emit in the 1.27 – 1.6 µm region. Table 5. Estimated indium production in 1986 [9]
Indium – gallium arsenide photodetectors have
high responsivity at 1.3 and 1.55 µm. Country Company Production,
t
The development of integrated circuits and Belgium Metallurgie Hoboken-
opto-electronic integrated circuits based on in- Overpelt 15
dium is still in the research phase. France Peñarroya 19
Italy Pertusola 19
High-purity indium is used for laser diodes, United Kingdom Capper Pass 3
photodetectors, and integrated circuits. Johnson Matthey 2
Charge coupled devices (CCDs) operating in Netherlands Billiton 1
the infrared region are based on indium anti- German Democratic
Republic VEB 1
monide. Infrared detectors and infrared video Federal Republic
cameras, such as orthicons, use indium anti- of Germany Preussag 1
monide. Japan Nippon Mining 16
China 6
Thin films of indium – tin oxide are able to
Former Soviet Union 6
transfer electricity to light emitting or light re- Peru Centromin-Peru 4
flecting elements. They find application in liquid Canada Cominco 6
crystal displays and to a lesser extent in electro- World production 80
luminescent displays and touch panels. Reproduced from Roskill: The Economics of Indium 4th ed.,
Minor uses of indium include thin plastic 1987, by permission of the publisher, Roskill Information
Services Ltd.
films coated with indium – tin oxide that are be-
ing developed for use in touch panels (e.g., for
computer screens), liquid crystal devices, and
electroluminescent lamps. Indium oxide is used Figure 5 shows the evolution of the producer
as a coating for glass. Window glass coated with price for standard grade indium (99.97 %) from
an indium – tin oxide film to reflect infrared ra- the Indium Corporation of America. For indium
diation is being investigated. Indium oxide films of 99.99 % purity, a premium of 2 – 5 $/kg is
doped with tin oxide are used to prevent ic- charged. Very pure indium (7N or 99.99999 %)
ing and fogging on aircraft windshields. Their costs ca. 1500 $/kg [9].
10 Indium and Indium Compounds
Figure 5. Evolution of the producer price for standard grade
indium (99.97 %) (Indium Corporation of America)
10. Toxicology and Occupational
Health
There is no evidence of any health hazard from
industrial use of indium. No systemic effects in
humans exposed to indium have been reported
[44]. The TLV for indium is 0.1 mg/m3 .
11. References
1. H. E. Suess, H. C. Urey, Rev. Mod. Phys. 28
(1956) 56.
2. F. Reich, T. H. Richter, J. Prakt. Chem. 89
(1863) 441.
3. Kirk Othmer, 3rd ed., 13, p. 207.
4. J. R. Mills, R. A. King, C. E. T. White: Rare
Metals Handbook, 2nd ed., Chap. 13,
Reinhold Publishing Corporation, London
1961, p. 220.
5. Ullmann, 4th ed., 13, 197.
6. Handbook of Chemistry and Physics, 62nd
ed., The Chemical Rubber Co., Cleveland,
Ohio, 1981/82.
7. J. C. Bailar et al.: Comprehensive Inorganic
Chemistry, 1st ed., Pergamon Press, Oxford
1973, pp. 983 – 1117.
8. J. W. Mellor, A Comprehensive Treatise on
Inorganic and Theoretical Chemistry, vol. 5.
Chap. XXXV, Longmans, Green and Co.,
London 1946.
9. Roskill: The Economics of Indium 1987, 4th
ed., Roskill Information Services Ltd., London
1987.
10. E. Theurich, Freiberg. Forschungsh. B 90
(1963) 93.
11. L. Müller, Freiberg. Forschungsh. B 90 (1963)
105.
12. Anaconda Copper Mining Comp.,
US 2 384 610, 1940 (H. M. Doran et al.).
13. R. Kleinert, “Indium from the Rammelsberg
Ores,” Min. Mag. 83 (1950) 146.
14. Asarco, US 2 433 770, 1947 (Y. C. Lebedeff).
15. S. Takayanagi, K. Yajima, paper presented at
the 22nd Annual Conference of Metallurgists
of the Metallurgical Society of CIM,
Edmonton, August 1983, paper no. 19.4.
16. E. F. Milner, W. A. Van Beek, “Rotary furnace
reduction of lead drossing fume for recovery
of tin and indium,” paper presented at the 24th
Annual Conference of Metallurgists of the
Metallurgical Society of CIM, Vancouver,
August 1985.
17. J. De Keyser, W. Jespers, “The Harris Refinery
of Metallurgie Hoboken-Overpelt,” paper
presented at the 110th AIME meeting,
Chicago, Feb. 1981, AIME-TMS paper
no. A-81-10.
18. J. Coyle, Trans. Electrochem. Soc. 85 (1944)
223.
19. T. Quarm, Trans. Inst. Min. Metall. 60 (1950)
77.
20. Capper Pass Limited, GB 2 109 008 A, 1982
(P. Halsall).
21. P. Halsall, Trans. Inst. Min. Metall. Sect. C 97
(1988) 93.
22. Preussag Aktiengesellschaft Metall,
US 4 666 686, 1987 (W. Krajewski, K.
Hanusch).
23. Hazen Research, Inc. USA, CA 1 188 105,
1985 (E. Reynolds Jones, A. R. Williams).
24. Cominco Ltd., CA 1 218 237, 1987 (R. Perri et
al.).
25. Tian Run-cang Zou Jia-yun, Zhou Ling-Zhi in
M. J. Jones, P. Gill (eds.): Mineral Processing
and Extractive Metallurgy, London IMM,
1984, pp. 615 – 624.

26. Samim Societa Azionarva Minero
Metallurgica S. P. A., BE 902 890, 1986 (R.
Guerziero et al.).
27. R. Guerziero, L. Meregalli, X. Zhang, ISEC
Int. Solvent Extr. Conf., Dechema, Frankfurt
1986, pp. 585 – 589.
28. Nippon Mining, JP-KK 88 016 338, 1988.
29. H. B. Linford, Trans. Electrochem. Soc. 79
(1941) 443.
30. Oneida Community Ltd., US 1 965 251, 1934
(M. S. Murray, D. Gray).
31. General Motors, US 2 409 983, 1946 (W. M.
Martz).
32. Indium Corporation of America, US 2 458 839,
1949 (J. R. Dyer, T. J. Rowan).
33. G. S. Rao, G. H. Phatak, K. Gangadharan,
Indian J. Technol. 12 (1974) 256.
34. A. Arkady et al., US 4 287 030, 1981.
35. U. Wiese, Erzmetall 34 (1981) no. 4, 190.
36. Sumitomo Chem. Ind., JP-KK 204 830, 1986.
37. C. E. T. White, Metal. Bull. 47 (1977).
Indium and Indium Compounds 11
38. Y. D. Liu, J. Verlinden, F. A. Lams, E.
Adriaenssens, Bull. Soc. Chim. Belg. 95
(1986) 309.
39. I. Barin, O. Knacke, O. Kubaschewski:
Thermochemical Properties of Inorganic
Substances, Springer Verlag,
Berlin-Heidelberg-New York, 1977.
40. D. B. Winter, Met. Bull., August 1988.
41. Jpn. Ind. Rare Met. Ann. Rep., April 1988.
42. M. Meeus, Y. Strauven, L. Groothaert, “Power
Sources 11,” Proc. Int. Power Sources Symp.,
15th, (1987) 281 – 299.
43. C. E. T. White: Advanced
Materials & Processes Inc., Metal Progress,
December 1986, 69 – 72.
44. L. Friberg, G. F. Nordberg, V. Vouk: Handbook
on the Toxicology of Metals,
Elsevier/North-Holland Biomedical Press,
Amsterdam 1979, pp. 429 – 436.
45. Kirk-Othmer, 14 1992, p. 157.
 Indole 1

Indole
Gerd Collin, DECHEMA e.V., Frankfurt/Main, Federal Republic of Germany (Chaps. 2– 5)
Hartmut Höke, Weinheim, Federal Republic of Germany (Chap. 6)

1. Introduction . . . . . . . . . . . . . . . . . . 1 5. Uses . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Properties . . . . . . . . . . . . . . . . . . . . 1
6. Toxicology ................... 3
3. Production . . . . . . . . . . . . . . . . . . . 1
4. Derivatives . . . . . . . . . . . . . . . . . . . 1 7. References . . . . . . . . . . . . . . . . . . . 4

1. Introduction the indole is concentrated in a biphenyl – indole

fraction which boils between 245 and 255 ◦ C.
Indole [120-72-9], 1-benzo[b]pyrrole, C8 H7 N, Following extraction of phenols and bases, the
was discovered by A. v. Baeyer and C. A. Knop indole is isolated from this fraction and sepa-
in 1866 as the basic structure of the natural dye, rated from the other major component biphe-
indigo, from which it was obtained. In 1910, R. nyl, whose boiling point is only 0.3 ◦ C higher.
Weissgerber found indole in coal tar. This is achieved by melting with potassium hy-
droxide to give the potassium salt of indole, by
azeotropic distillation with diethylene glycol, or
by extraction with selective solvents (e.g., gly-
cols, aqueous dimethyl sulfoxide, or monoetha-
nolamine). The crude indole can then be puri-
fied by crystallization from aliphatic hydrocar-
2. Properties bon solvents.
In addition to its recovery from coal tar, in-
Physical Properties. Indole, M r 117.15, dole is also synthesized in technical quantities.
mp 52.5 ◦ C, bp 254.7 ◦ C (101.3 kPa), Methods used include Madelung synthesis of
d 1.22 g/cm3 (18 ◦ C), forms colorless, shiny formyltoluidine (from o-toluidine and formic
flakes with a slight jasmine aroma. It is readily acid), followed by cyclization [8–10], dehydro-
soluble in ethanol, diethyl ether, and benzene, genating cyclization of 2-ethylaniline [11–20],
soluble in hot water, slightly soluble in cold cyclocondensation of aniline and ethylene gly-
water, and volatile in steam. The heat of com- col in the liquid or gas phase [21–23], and cy-
bustion is 3.650 kJ/kg (25 ◦ C), the enthalpy of clization of 2-(2-nitrophenyl)ethanol [24].
vaporization 597.5 kJ/kg (10.3 – 27.4 ◦ C), and
the dipole moment 2.11 D (benzene).
4. Derivatives
Chemical Properties. As a secondary
amine, indole has a hydrogen atom which can be 3-Methylindole [83-34-1], skatole, C9 H9 N,
substituted by alkali metals. Oxidation leads to M r 131.18, mp 95 ◦ C, bp 266.4 ◦ C (101.3 kPa),
indigo, mild hydrogenation to 2,3-dihydroindole forms colorless flakes which are soluble in etha-
(indoline). Diindole, triindole, and resinification nol and slightly soluble in water. It can be reco-
products are obtained upon treatment with acids. vered from coal tar or synthesized by Fischer
reaction from the phenylhydrazone of propion-
aldehyde or, together with indole, by catalytic
3. Production dehydration of 2-(2-aminophenyl)-1,3-propane-
diol [25].
High-temperature coal tar contains on average
just under 0.2 % indole. In coal tar distillation,

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 167
2 Indole
3-(Dimethylaminomethyl)indole [87-52-5],
gramine, C11 H14 N2 , M r 174.25, mp 138 –
139 ◦ C, forms colorless needles. The com-
pound is soluble in ethanol, diethyl ether, and
chloroform and insoluble in water. It can be
synthesized from indole by the Mannich reac-
tion with dimethyl-amine and formaldehyde.
Gramine is used to synthesize d,l-tryptophan,
the plant growth regulators indole-3-acetic acid
and indole-3-butanoic acid as well as the inter-
mediate tryptamine.
Indole-3-aldehyde [487-89-8], C9 H7 NO,
M r 145.16, mp 194 ◦ C, forms colorless crystals.
It is readily soluble in ethanol and slightly solu-
ble in cold water. It can be produced by Vilsmeier
reaction of indole with formylmethylaniline or
with dimethylformamide and phosphorus oxy-
chloride. Indole-3-aldehyde can be used to make
d,l-tryptophan and cyanine dyes.
2,3-Dihydroindole [495-15-1], indo-
line C8 H9 N, M r 119.16, bp 229 – 230 ◦ C
(101.3 kPa), is a colorless liquid, which is
volatile in steam and soluble in diethyl ether,
acetone, and benzene, but only slightly soluble
in water. Indoline is obtained by hydrogenation
of indole or by catalytic cyclodehydration of
2-(2-aminophenyl)ethanol. A range of pharma-
ceuticals, as well as fungicides and bactericides,
can be produced from indoline.
3-(2-Aminoethyl)indole [61-54-1], trypta-
mine, C10 H12 N2 , M r 160.22, mp 118 ◦ C, forms
colorless needles. It is readily soluble in eth-
anol and acetone, and slightly soluble in wa-
ter, diethyl ether, chloroform, and benzene.
It can be produced by reacting indole with
ethylene imine, by catalytic hydrogenation of
indole-3-acetonitrile, or by reacting phenylhy-
drazine with 4-aminobutyraldehyde diethyl ac-
etal. Tryptamine is used to synthesize the va-
sodilator and antihypertensive, Vincamine.
3-(2-Hydroxyethyl)indole [526-55-6],
tryptophol, C10 H11 NO, M r 161.20, mp 59 ◦ C,
bp 174 ◦ C, (2 kPa) forms colorless prisms or
flakes that are readily soluble in methanol,
ethanol, diethyl ether, chloroform, and ace-
tone, soluble in benzene, and slightly soluble
in water. It is produced by reacting methyl-
1H-indole-3-acetate with lithium aluminum
hydride, or through reduction of methyl-1H-
indole-3-glyoxylate with sodium borohydride.
The antihypertensive Indoramin is produced
from tryptophol.
5. Uses
In line with its natural occurrence as a com-
ponent of jasmine oil and orange-blossom oil,
indole has been used for many years to fix fra-
grances and is therefore found in many perfumes
[26].
A further important application is the pro-
duction of the essential amino acid tryptophan
(→ Amino Acids, Chap. 3.2.17. ). Tryptophan
can be synthesized from indole by chemical re-
actions as a racemate (via the indole derivatives
gramine and indole-3-aldehyde). It is also pro-
duced by biotechnology in the enantiomerically
pure l-form (e.g., from indole and l-serine).
Indole is also used as a feedstock in the syn-
thesis of plant growth regulators, such as indole-
3-acetic acid [6250-86-8] ( heteroauxin) and

indole-3-butanoic acid [3127-44-4] (→ Plant
Growth Regulators). Fungicidal and bactericidal
plant protectives are synthesized from indoline
[27].
Derivatives of indole-3-acetic acid find ap-
plication as mild analgesics, e.g., indomethacin,
(→ Anti-inflammatory – Antirheumatic Drugs).
Indole and its derivatives are basic building
blocks for many other pharmaceuticals. Thus,
for example, indoline is used to synthesize
analgesics [28] and antihypertensives [29],
and tryptamine [61-54-1] [3-(2-aminoethyl)-
indole] for the synthesis of the vasodilator and
antihypertensive, Vincamine [30]. Tryptophol
[526-55-6] [3-(2-hydroxyethyl)indole] is used
to prepare the antihypertensive Indoramin [31]
(→ Antihypertensives, Chap. 4.2.2.2.). (Pyridy-
lalkyl)indoles can be produced from indole
and can be used as antidepressants, antihis-
tamines, and antihypertensives [32–34]. 3-
Piperidinylindoles can be obtained via the
Grignard reaction of 4-chloroindole with 3-
piperidinone, and can be used in the preven-
tion of anoxia [35]. Indole-2-carboxylic acid
derivatives are useful, e.g., as antihypertensives
[36–38]. 5-Chloroindole [17422-32-1] can be
synthesized from indoline via the intermedi-
ates 1-acylindoline, 1-acyl-5-chloroindoline,
and 5-chloroindoline [39]. It can be used
as an intermediate in the production of 5-
chlorotryptamine (for the production of tran-
quilizers and blood pressure lowering agents),
as well as in the production of antidepressives,
antiemetics, and drugs for the treatment of
Parkinson’s disease [40–42]. 2-Methylindole
[95-20-5] (as the heterocyclic coupling com-
ponent) is a starting material for commer-
cially important cationic diazo dyes. Cyanine
dyes are synthesized from 2-methylindole,
1-methyl-2-phenylindole, 2-methylindole-3-
aldehyde, and indole-3-aldehyde (→ Methine
Dyes and Pigments). In addition, indolyl-
methane dyes are produced from 1-methyl-2-
phenylindole (→ Triarylmethane and Diaryl-
methane Dyes). Reacting indole with 3,6-bis-
(2-carboxybenzoyl)carbazole affords carbazole
phthalein dyes, which are used as optical filter
agents in photographic emulsions [43].
Indole 3
6. Toxicology
Apart from contact during production and manu-
facture of indole, human exposure results from
smoking and intestinal degradation of nutri-
tional l-tryptophan. Indole is readily absorbed
from the gastrointestinal tract and rapidly con-
verted by the liver, e.g., to indican and oxindole,
which are excreted mainly into the urine [44],
[45]. The acute oral LD50 is 1000 – 1100 mg/kg
in rats [46], [47] and about 500 mg/kg in mice
[48]. Autopsy after acute exposure revealed
hemorrhages and hyperemia of inner organs and
tissues. Upon application to the skin and eyes
of rabbits, slight temporary irritation was ob-
served [49], [50]. Unlike 3-methylindole (ska-
tole), indole does not induce pulmonary injury
in cattle after repeated oral application [51].
At high chronic dosage (20 – 200 mg kg−1 d−1
orally and subcutaneously in mice, rats, rabbits,
and dogs for more than three months), indole
resulted in a shift in the blood cell differen-
tial picture (anemia, leukocytosis, or leukope-
nia) ; with persistently high intoxication, the
development of leukemia cannot be ruled out
[52–54]. In cattle, hemolysis and hemoglobin-
uric nephrosis were observed after several oral
doses of indole (100 or 200 mg/kg) [51]. Al-
though there is no other experimental evi-
dence for its cancerogenic potential [55], in-
dole may influence the formation of tumors
caused by other agents. Thus, very high oral
doses (800 mg kg−1 d−1 ) have been reported
to accelerate the development of tar-induced
skin tumors in mice [53] and 2-acetylami-
nofluorene-induced bladder carcinomas in ham-
sters and rats [56]. On the other hand, it si-
multaneously seemed to suppress 2-acetylami-
nofluorene-induced hepato-carcinogenesis [56].
A genotoxic effect of indole was confirmed in
the Ames test in combination with other mu-
tagenic agents [57]. However, indole itself was
neither mutagenic in the Ames test [57], [58],
nor in a cell transformation test [59], although
in a DNA repair assay with Bacillus subtilis, in-
dole caused reparable DNA damage [59].
4 Indole
7. References
General References
1. Beilstein, 20 304, 20(1) 121, 20(2) 196,
20(3) 3176 – 3180.
2. H. J. V. Winkler: Der Steinkohlenteer und
seine Aufarbeitung, Verlag Glückauf, Essen
1951, pp. 146 – 148, 151 – 152.
3. Rütgerswerke, Erzeugnisse aus
Steinkohlenteer, 1958, Frankfurt, Federal
Republic of Germany.
4. H.-G. Franck, G. Collin: Steinkohlenteer,
Springer Verlag, Berlin 1968, pp. 60 – 61,
180 – 181.
5. H.-G. Franck, J. W. Stadelhofer: Industrial
Aromatic Chemistry, Springer Verlag, Berlin
1988, pp. 417 – 418.
6. R. J. Sundberg: The Chemistry of Indoles,
Academic Press, New York 1970.
7. W. J. Houlihan: Indoles, John Wiley & Sons,
New York 1972.
Specific References
8. Mitsui Toatsu, JP 74 20 587, 1970 (R.
Fujiwars, Y. Yamada, N. Kuroda).
9. Kyowa Fermentation Industry, JP 73 76 864,
1972 (T. Yamauchi, S. Yada, S. Kudo).
10. Ube Industries, JP 78 95 965, 1977 (M. Nakai,
T. Komata, M. Oda).
11. Rütgerswerke, DE 2 224 556, 1972;
DE 2 401 017, 1974 (G. Grigoleit, R.
Oberkobusch, G. Collin).
12. Houdry, US 2 891 965, 1956 (S. E. Voltz, J. H.
Krause, W. E. Erner).
13. Snam Progetti, DE 2 148 961, 1971 (M. M.
Mauri, P. Moggi).
14. Teijin, JP 73 76 862, 1972 (T. Fukada, K.
Tanaka).
15. E. I. du Pont de Nemours, US 2 409 676, 1944
(W. F. Grasham, W. M. Brunner).
16. Socony Mobil Oil, NL 6 505 911, 1965.
17. Lonza, DE 2 441 439, 1974 (C. O’Murchu).
18. Société Nationale Elf Aquitaine,
DE 2 328 284, 1973 (M. Petinaux, J. Metzger,
J.-P. Aune, H. Knoche).
19. Sumitomo Chemical, JP 7 831 660, 1976 (M.
Tamura).
20. Mitsui Toatsu, JP 76 113 868, 1975 (K.
Yamamoto, K. Nitta, S. Ichikawa).
21. Mitsui Toatsu, EP 86 239, 1982 (T. Honda, K.
Terada).
22. Ube Industries, JP 83 128 371, 1982 (N.N.).
23. Keishitsu Ruibun Shinyoto Kaihatsu Gijutsu
Kenkyu Kumiai, JP 86 44 862, 1984 (M.
Imanari, T. Seto).
24. Bofors, DE 2 052 678, 1970 (J. M. Bakke,
H. E. Heikmann).
25. Rütgerswerke, DE 2 328 330, 1973 (G.
Grigoleit, R. Oberkobusch, G. Collin).
26. G. T. Walker, Seifen öle Fette Wachse 100
(1974) 375 – 377.
27. Pfizer, GB 1 394 373, 1975; GB 1 394 374,
1975 (R. J. Bass, R. C. Koch, H. C. Richards,
J. E. Thorpe).
28. Hoechst-Roussel Pharmaceuticals,
US 4 448 784 784, 1982 (E. J. Glamkowski,
J. M. Fortunato).
29. Science Union, FR 2 003 311, 1968 (L. Beregi,
P. Hugon, M. Laubie).
30. Roussel-UCLAF, DE 2 115 718, 1970 (J.
Warnant, A. Farcilli, E. Toromnoff).
31. Wyeth, GB 1 399 608, 1971 (J. L. Archibald).
32. Pharmindustrie, EP 65 907, 1982 (F. Audiau,
G. R. Le Fur).
33. Boehringer Ingelheim, EP 58 975, 1982 (K.
Freter, V. Fuchs, J. T. Oliver).
34. Wyeth, EP 2 886, 1979 (J. L. Archibald, T. J.
Ward).
35. Roussel-UCLAF, EP 21 924, 1981 (J.
Guillaume, L. Nedelec, C. Dumont, R.
Fournex).
36. Hoechst, DE 3 151 690, 1981 (H. Urbach, R.
Henning, V. Teetz, R. Geiger, R. Becker).
37. McNeilab, US 4 529 724, 1983 (C. Y. Ho).
38. American Cyanamid, US 4 619 941, 1984
(W. B. Wright, J. B. Press).
39. Rütgerswerke, DE 3 035 403, 1980 (M.
Maurer, W. Orth, W. Fickert).
40. MARPHA, DE 2 618 152, 1977 (A.
Champseix, C. Gueremy, G. Le Fur).
41. Roussel-UCLAF, DE 2 719 294, 1977 (F.
Clemence, D. Humbert).
42. Roussel-UCLAF, DE 2 738 646, 1978 (C.
Dumont, J. Guillaume, L. Nedelec).
43. Polaroid, US 3 932 455, 1976 (R. C. Bilofsky,
R. D. Cramer).
44. A. H. Beckett, D. M. Morton, Biochem.
Pharmacol. 15 (1966) 937 – 946.
45. L. J. King, D. V. Parke, R. T. Williams,
Biochem. J. 98 (1966) 266 – 277.
46. National Institute for Occupational Safety and
Health (NIOSH): Registry of Toxic Effects of
Chemical Substances, US Government
Printing Office, Washington D.C. 1987,
p. 12992.
47. Huntingdon Research Centre, Rütgerswerke
Order No. 338 a, 1979, Münster, Federal
Republic of Germany.

48. G. J. Martin, E. H. Rennebaum, M. R.
Thompson, Ann. Intern. Med. 18 (1943)
57 – 71.
49. Huntingdon Research Centre, Rütgerswerke
Order No. 338 b, 1979, Münster, Federal
Republic of Germany.
50. Huntingdon Research Centre, Rütgerswerke
Order No. 338 c, 1979, Münster, Federal
Republic of Germany.
51. A. C. Hammond, J. R. Carlson, Vet. Rec. 107
(1980) 344 – 346.
52. W. T. Taylor, Med. Ann. D. C. 8 (1939)
362 – 363.
Industrial Hygiene → Occupational Health and Safety
Indole 5
53. K. Kaiser, Z. Krebsforsch. 59 (1953)
448 – 495.
54. H. Ehrhart, W. Stich, Klin. Wochenschr. 35
(1957) 504 – 511.
55. M. J. Shear, J. Leiter, J. Natl. Cancer. Inst.
(US) 2 (1941) 241 – 258.
56. M. Matsumoto, M. L. Hopp. R. Oyasu, Invest.
Urol. 14 (1976) 206 – 209.
57. T. Matsumoto, D. Yoshida, S. Mizusaki,
Mutat. Res. 56 (1977) 85 – 88.
58. M. Curvall, J. Florin, T. Jansson, Toxicology
23 (1982) 1 – 10.
59. T. Osawa, M. Namiki, K. Suzuki, T. Mitsuoka,
Mutat. Res. 122 (1983) 299 – 304.
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Information Storage Materials Save title to My Profile
Standard Article Topics
Email this page
Roman E. Wochele1, Barry K. Middleton2, Henk van Houten3, Johan P. W. Print this page
B. Duchateau3, Hans J. G. Kloosterboer3, Jan A. Th. Verhoeven4, René van SEARCH THIS TITLE
Vlimmeren5, Peter E. J. Legierse6, Dirk J. Gravesteijn3, C. David Wright7,
Herman J. Borg3, Heinrich Heitmann8, Jacques Heemskerk9
1Gramm Oberflächentechnik, Ditzingen, Germany Advanced Product Search
2Manchester University, Manchester, United Kingdom
Search All Content
3Philips Research Laboratories, Eindhoven, The Netherlands
Acronym Finder
4Toolex International N.V., Veldhoven, The Netherlands
5Toolex International N.V., Veldhoven, The Netherlands
6Optical Disc Mastering (ODM), Eindhoven, The Netherlands
7Manchester School of Engineering, University of Manchester,
United Kingdom
8Philips Research Laboratories, Hamburg, Federal Republic of
Germany
9Philips, Eindhoven, The Netherlands

Copyright © 2004 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved.
DOI: 10.1002/14356007.a14_171.pub2
Article Online Posting Date: August 15, 2004

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Magnetic Recording
1.1. Introduction
1.1.1. History
1.1.2. Magnetism and Magnetic Materials: Definitions
1.2. Principles
1.2.1. Digital Record and Replay
1.2.2. Media Requirements for Digital Recording
1.2.3. a.c. Bias Audio Recording
1.2.4. Video Recording
1.2.5. Transports
1.2.6. Electronic Systems
1.3. Magnetic Recording Media
1.3.1. -Fe2O3
1.3.2. Chromium Dioxide
1.3.3. Cobalt-Modified -Fe2O3
1.3.4. Metal Particles
1.3.5. Metal-Evaporated Tape (MET)
1.3.6. Metal Film Disk Media
1.3.7. Perpendicular Recording
1.4. Recording Heads
1.5. Economic Aspects
2. Optical Recording
2.1. General Introduction
2.1.1. Optical Storage Media
2.1.2. The Optical Channel
2.1.2.1. Focusing and Tracking (Servo Control)
2.1.2.2. Data Channel
2.1.3. Compact Disc Family

page 1 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2.1.3.1. Read-Only, Replicated Discs
2.1.3.2. Recordable and Rewritable Discs
2.1.4. Digital Versatile Disc (DVD)
2.1.5. HD-DVD and Blue Laser Discs
2.1.6. Magneto-Optical Storage
2.1.7. Future Developments
2.1.7.1. Evolutionary Progress in Optical Disc Storage
2.1.7.2. Three-Dimensional Optical Storage
2.2. Optical Disc Mastering and Stamper Production
2.2.1. General
2.2.2. Requirements
2.2.3. Mastering Methods
2.2.4. Electroforming of Stampers
2.3. Read-Only Optical Recording Materials
2.3.1. Production Processes
2.3.2. Material Specifications
2.3.3. Molding Processes
2.3.4. Photoreplication Process
2.3.5. Metallization
2.3.6. Protective Coating
2.3.7. Nonstandard Processes
2.4. Write-Once Optical Recording Materials
2.4.1. Material Specifications
2.4.2. Deposition Techniques
2.4.3. Tellurium Alloys
2.4.4. Organic Dyes
2.4.5. Metal Particle Colloids
2.4.6. Metal-Polymer Bilayers
2.4.7. Alloying Bilayers
2.4.8. Agglomeration
2.4.9. Surface Texture
2.5. Reversible Optical Recording Materials
2.5.1. Magneto-Optic Materials
2.5.2. Phase-Change Recording Media
2.5.3. Liquid Crystals
2.5.4. Photochromics
2.5.5. Photochemical Hole Burning
2.6. Economic Aspects of Optical Recording
2.6.1. Consumer Electronics (CE)
2.6.2. Data Applications
3. Acknowledgment

[Top of Page]

1. Magnetic Recording
Barry K. Middleton

1.1. Introduction
Magnetic recording technology is widespread in its adoption and its impact on our lives. Be it in portable audio recorders, hi-fi
audio systems, video recorders, or in floppy or rigid disk systems in computers, the products of the technology are widely
used and appreciated today. Novel data storage systems are being prepared and queuing up for possible launch into the
market place. All these are propelled by mighty industrial concerns in various parts of the world primarily in America and
Japan, but also in Europe.

The majority of audio and video recording has traditionally taken place using analogue techniques although in recent times
products operating in the digital domain have been introduced. The future will see an accelerated shift to digital techniques
and this means that there is a continued convergence of most recording technologies into the data storage arena where all
information from whatever source can be recorded and manipulated using digital methods.

1.1.1. History
Magnetic recording [1] was first demonstrated by the Danish inventor VALDEMAR POULSEN in 1898. He developed a machine
for recording voice messages onto steel wire and was able to demonstrate recording and replay in a recognizable form. This

page 2 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
achievement quickly received acclaim. Despite the low signal levels and high levels of distortion in the recordings,
developments of this equipment were made and a telephone message recording machine called a Telegraphone was
eventually put on the market. However the Telegraphone was not widely adopted, partly through its lack of clarity, and the full
exploitation of the technology had to await the later developments of electronic technology and the means to reduce
distortions.

A major step forward came in 1935 with the development in Germany by BASF and AEG of a machine called the
Magnetophone. This machine, and its later variants, brought into play two new developments which remain today as the core
of many magnetic recording systems. The first was the introduction of magnetic particles as the recording medium and the
second was the use of a.c. bias to linearize the characteristics of the recording medium. With these developments record and
replay with high clarity was achieved and it led eventually, with further progress, to the widespread adoption of a.c. biased
recording systems. The most popular current manifestation of the technology is in the compact cassette audio system
introduced by Philips in 1963.

Digital recording emerged during the late 1940s and 1950s in response to the need for back-up storage for computer
devices. By these times magnetic tape recording for audio was in widespread use and the technology was adapted for the
recording of digital waveforms without the use of a.c. bias. In the early machines bit packing densities were very low but this
increased rapidly with successive generations of equipment. Record and replay took place at data rates which were slow
compared with the processing rates of computers and access times to information became a critical issue. This led to the
development of faster access mechanical systems such as the magnetic drum and later the very successful magnetic disk.
The latter is now the mainstay of storage for computers and disk systems exist in a variety of forms and formats employing
rigid and flexible disk media.

Video recording first took place in “fixed head” systems similar to those used in audio and data recording. However, the
wide bandwidths required by video signals meant that very high tape speeds had to be used and this required very long
tapes for program length recordings. To alleviate this problem scanning head technology was developed in which tracks were
recorded very close together in directions across the tape length to give very high area packing densities of information. First
came the transverse scan technology introduced by Ampex in 1956 and later the helical scan technologies were introduced
for domestic use and are now most widely manifested in VHS recorders [2]. Helical scan technology has now been adapted
for use in data storage systems especially for computer back-up applications.

Recording technology continues to move forward and new developments invariably break fresh ground in terms of the
amount of data stored per unit area of recording surface. Here comes the pressure for new and advanced recording media.

1.1.2. Magnetism and Magnetic Materials: Definitions ( Magnetic Materials)

For a more detailed description of the terms used in the specification of magnetic materials, particularly those used in
recording media and in magnetic heads, see, e.g., [3].

The magnetic properties of materials are often characterized by the magnetic induction or flux density within them. This
quantity is represented by the symbol B and measured in units of tesla (T). In free space B is given by

(1)

where 0 = 4 · 10–7 H/m is the permeability of free space and H is the magnetic field strength measured in A/m. The
magnetic field is usually generated by a coil carrying a current and the magnitude of the field depends on that current and the
geometry of the coil hence leading to a field strength which is expressed in A/m.

When the induction is measured inside a magnetic material it may be represented by

(2)

where H is the applied field strength and M is an additional contribution, measured in A/m, arising from the magnetic material
and which is usually referred to as its magnetization. Sometimes Equation (2) is written in a modified form

(3)

where J (sometimes the symbol I is used) is also referred to as the magnetization but it is now in units of Tesla. Both
representations of magnetization appear in the literature but in this article both M and H will be quoted in the same units of
A/m. This is the approach adopted in most publications although there is still a widespread use in many parts of industry and
academia of cgs and some imperial units.

In ferromagnetic and ferrimagnetic materials the relationship between the magnetization produced in a material and the
applied magnetic field is usually nonlinear. A typical relationship is shown in Figure 1. Starting with a demagnetized sample,
that is one with zero net magnetization, the application of a magnetic field increasing from zero takes the magnetization along
the curve a. First of all the magnetization rises very slowly, but as the field increases the rate of rise increases. Eventually the
magnetization approaches a value known as the saturation magnetization Ms. This curve is known as the initial
magnetization curve. On removal of the field the magnetization drops back to the value Mr, known as the remanent
magnetization. On increasing the field in a negative direction the magnetization eventually starts to reduce and is brought to
zero at a field labeled –Hc which is the negative value of the coercivity. Further increase of this field in the negative direction
eventually takes the magnetization through to its negative saturation value of –Ms. Bringing the field back to zero and then
increasing it again in a positive direction traces out a curve with the positive going arrows, which is a mirror image of that
produced by negative going fields, and brings the material back to the positively saturated state. The curve just traversed is
known as the major hysteresis loop and the quantities coercivity Hc, remanent magnetization Mr, saturation magnetization M

page 3 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
s,and hysteresis and loop squareness Mr/Ms are all properties very important in characterizing materials and especially
those used in magnetic recording.

Figure 1. Initial magnetization curve (a) and major hysteresis loop of a typical magnetic material.

Sometimes magnetic materials are quoted as being magnetically hard or magnetically soft. Magnetically hard materials (
Magnetic Materials – Parameters of Hard Magnetic Materials) are those in which very large magnetic fields have to be
applied to change their states of magnetization. These types of materials have high coercivities and are permanent magnet
materials. Magnetic recording media must be made of hard magnetic materials so they retain their recorded magnetizations
permanently until they are purposely overwritten with new information or are erased. Soft magnetic materials ( Magnetic
Materials – Parameters of Soft Magnetic Materials) are those with extremely low coercivities. In this case the application of
only small magnetic fields is sufficient to change their magnetizations. Such materials are used in magnetic recording heads
where small currents flowing in the coils of the heads are sufficient to change their states of magnetization and therefore
change the fields produced by them.

1.2. Principles
1.2.1. Digital Record and Replay
Figure 2 shows, in diagrammatic form, a simple record and replay system. It consists of a magnetic recording medium
moving from left to right firstly under a record head and then under a replay head. The medium is made of permanent magnet
material so that once a pattern is recorded into it it remains permanently magnetized, while the record and replay heads are
made of soft magnetic material in which the magnetizations are easily alterable. The heads are of the ring type (see
Section Recording Heads), so named because of their open shape, and they are designed to produce a high magnetic field
in the recording medium sufficient to alter its magnetization. The replay head senses the magnetic fields arising from the
magnetized medium as it passes underneath.

Figure 2. The digital record and replay processes. See text for description.

Recording takes place as follows. A current I flowing in the record head in the direction shown produces a field in the coil of
the head which magnetizes the core of the head as shown. This produces a very high magnetic field in the gap of the head
while below the head gap the field leaks out and penetrates the recording medium. With a good design of the pole region of
the head this leakage or fringe field is sufficiently high to magnetize the recording medium. As shown in Figure 2 the field
produced under the head is directed to the left and while this field remains in place the medium that is transported through it
is magnetized to the left. A reversal of the record current to the opposite sense (direction) produces a magnetic field under
the record head which is directed to the right. This then leads to the magnetization of the medium also to the right. Hence an
alternating polarity of record current leads to an alternating polarity of magnetization along the tape.

While it is possible to change the sense of the record current extremely quickly with respect to time, the magnetization needs
some distance along the tape over which its sense can be changed. This imposes limitations on the number of reversals or
changes of magnetization that can be recorded per unit length of medium. These changes of magnetization, or transitions,
are used in recording to carry the bits of information and so the number of these transitions per unit length is often referred to
as the bit packing density. Obviously an advanced and good recording tape will have transitions which are extremely narrow
so that a large number of them can be recorded per unit length of medium. Much of modern recording technology centers on
developing media which can sustain narrow transitions and heads which are able to record them.

The replay process takes place as follows. The replay head is sensitive to the magnetization in the recording medium as it
passes under its gap. The head produces in its magnetic circuit a flux which is related to the flux in the recording medium.
This flux passes through the coil of the head and its rate of change gives the output voltage. The latter is therefore related to
the rate of change of flux in the medium below the head gap and so only changes of magnetization in the medium produce
an output. The input current and the corresponding output voltage are shown in Figure 2. Figure 3 top, again shows the
output voltage as a function of time for transitions widely spaced on the medium. The narrower the recorded transitions the
higher and narrower are the replayed pulses. The converse is true in that the wider the transitions the smaller and wider are
the pulses.

Figure 3. Output waveforms at different bit packing densities.

page 4 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Obviously it is desirable to record as much information as possible per unit length of recording medium. When this happens
the replay pulses are moved closer together and eventually they overlap. This is illustrated in Figure 3 where in subsequent
diagrams the packing density of information increases. It can easily be seen that as pulses overlap they detract from one
another and the output amplitude is reduced. Thus, the price of recording at high densities is a loss of signal amplitude, and,
as with all electronic systems, ultimately performance is limited by the signal-to-noise ratio and so the reducing signal level
will determine the performance of the recording system.

Figure 4 shows a typical curve of output voltage as a function of packing density which in this case is for an advanced thin-
film disk recording medium. The recording densities involved are high and the transition widths narrow. A high performing
recording medium will have narrow transitions recorded in it and output voltage will fall off slowly with increasing packing
density. A lower performing recording medium will have wider transitions recorded in it and its output will fall off quickly as
packing density increases. More detailed discussions of this are in the literature (see, e.g., [4]) but more will also be said
about this in later sections.

Figure 4. Output voltage amplitude (after amplification) as a function of packing density for a thin film disk recording medium
of thickness 30 nm, coercivity 175 kA/m, head gap 0.25 µm, head to medium spacing 0.06 µm, and estimated transition
width parameter of 0.095 µm.

1.2.2. Media Requirements for Digital Recording

There are two aspects of recording with which the recording media must be adequately designed to cope [4]. The first is their
ability to have recorded within them narrow transitions to facilitate small bit sizes, and thus high bit packing densities, and the
second is to retain those transitions once they have been recorded.

A transition is created in a recording medium when the current in the record head and therefore the field under it is reversed
(see Fig. 2). This field is very large immediately under the center of the gap and falls away with distance to the right because
this side of the head produces the recorded transition. The new magnetization under the gap likewise gives way to the
previously recorded magnetization to the right of the head gap. The head field is therefore decaying rapidly with distance
within the transition and the magnetization gradient is given by

(4)

where (dH/dx) is the head field gradient and (dM/dH) is the variation of magnetization with applied field. The center of a
recorded transition is at M = 0 and according to the hysteresis loop of Figure 1 this is where H = Hc and the slope of the
hysteresis loop is highest. For narrowest transitions high hysteresis loop squarenesses are required of the media and high
field gradients are required of the record heads.

However the treatment involving only head fields is a simplification. Changes of magnetization in a magnetic medium
produce stray (or demagnetizing) fields which depend on the magnitude of the magnetization recorded in the media and also
on the thicknesses of the media. These add to the head fields and together they shape the recording process as indicated by
Equation (4) (see, e.g., [4-6]). Taking these into account and assuming the magnetization distribution along the medium
through a transition is of the form [7]

(5)

where a is a parameter proportional to the transition width, as shown in Figure 5. It can be shown that in a thin recording
medium of thickness D and with a rectangular hysteresis loop [4]

(6)

where d is the record head to medium spacing. However, when the recording medium is unable to retain such narrow
recorded transitions as predicted by Equation (6) because the self demagnetizing fields in the transitions are high, there is a
minimum value for transition width determined by the fact that the demagnetizing fields should not exceed the coercivity of
the medium [8]. Then [9]

(7)

page 5 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Thus, the write process in a thin film medium will produce a transition of width given by Equation (6) provided the width is
larger than the minimum which self demagnetizing fields allow according to Equation (7). Thus the resultant transition width
takes the larger of the values given by Equations (6) or (7). Either way the controlling factor is the ratio (Mr D/Hc) and a small
value is required for narrow transitions and high density recording.

Figure 5. Assumed magnetization distribution as a function of distance through a transition.

The tangent at the origin shows the overall transition width to be a.

Media suited for narrow transitions and high density (short wavelength) recording therefore need to be thin and have a high
coercivity. These requirements are put alongside that of high hysteresis loop squareness and together they have persisted as
the guiding requirements for advanced new media.

1.2.3. a.c. Bias Audio Recording

In audio recording the magnetization level which is induced in the tape must be linearly related to the recording signal. This is
achieved by using a.c. bias recording [4], [10]. The principle of a.c. bias recording is illustrated in Figure 6 which shows
graphs of magnetization remaining in a tape after the application and removal of magnetic fields H (remanence curves). Take
the curve labeled with a 0 where the field has been applied to a previously demagnetized specimen of tape and the
relationship between the magnetization and field is clearly not linear. Therefore a signal field applied by the recording head
would not induce in the medium a magnetization directly proportional to it. However if an a.c. bias field Hac is added to the
applied field H and taken away in step with the applied field, the relationship between magnetization and field is altered
according to the amount of a.c. bias added. Figure 6 shows curves for a number of values of a.c. bias field and it is clear that
increasing the a.c. bias removes the non-linearities of the magnetization curve eventually producing, at low d.c. fields, the
required linear relationship. The use of an a.c. bias field in conjunction with the applied field leads to a set of curves which
are known as anhysteretic remanence curves. Where the field and bias reduce in unison, as in the recording process, they
are known as modified remanence curves, whereas the “standard” anhysteretic remanence curves are obtained by reducing
the bias field to zero before the d.c. field is removed [11], [12].

Figure 6. Modified remanence curves for -Fe2O3 recording tapes. The numbers alongside each curve indicate the amount
of a.c. bias.

In the recording situation the applied field represents the signal which is at low frequency and to this is added a much higher
frequency a.c. bias field. For example for audio signals of up to 20 kHz bandwidth the a.c. bias added must have a frequency
at the very least of 65 kHz. The sum of the bias and the signal is passed through the record head and this then creates a
magnetization in the medium which is linearly related to the record signal. The role of the bias field is purely to linearize the
relationship between magnetization and signal field and is itself not recorded on the recording medium.

The requirements for the recording process are therefore for high linearity of the relationship between applied field and
magnetization, but for retention of the recorded pattern at short wavelengths the tape must be thin to keep demagnetizing
fields low and the coercivity must be high to resist demagnetization. Thus the requirements for good short wave length audio
recording mirror those for high bit density digital recording.

The output voltage generated in the replay head is, as before, related to the magnetization gradient recorded in the medium.
For a record signal which takes the form of a sine wave the differential is a cosine whose amplitude increases linearly with
frequency. The output from an ideal recorder therefore has an amplitude which rises linearly with frequency but has suffered
a 90 ° phase shift from the record signal. This linear curve is something of an idealization because at high frequencies losses
in the replay process come into play [12-14]. These are caused by the finite spacing between the head and the medium, the
finite thickness of the medium, and the finite width of the gap. They conspire to reduce the output at high signal frequencies
as shown in Figure 7. Therefore the overall characteristic of a recording system is of a bandpass nature with no output at
d.c., very little signal at low frequencies, and also little output at high frequencies due to the replay losses.

Figure 7. Output as a function of frequency for an audio recording tape.

1.2.4. Video Recording

Video recording takes place as an analogue process but does not use a.c. bias. It achieves its linearity by the use of

page 6 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
frequency modulation in which the video signal is applied to a high frequency carrier wave so that the instantaneous
frequency of the modulated waveform is linear with the instantaneous magnitude of the video signal [2], [13]. It is therefore
the variations in frequency, and thus the spacings between the crossovers in the waveform, which carry the video
information. In the recording machine the frequency modulated waveform is applied directly to the record head. Despite the
distortions in unbiased recording the signal, represented by the spacings between the crossovers, is faithfully recorded and
so after replay and demodulation the video information can be recovered [2], [13].

The overall characteristics of the recorder in terms of transfer characteristics are much the same as described in Section a.c.
Bias Audio Recording, and shown in Figure 7, except that the bandwidths are now much wider and reach out to several MHz.
The requirements of the recording medium are likewise similar, i.e., thinness and high coercivity.

1.2.5. Transports
Audio recording systems use heads which are fixed in position and tapes which move past and in contact with them. With a
tape speed of 4.76 cm/s a length of tape of approximately 172 m would be needed for 1 h playing time. Such lengths of
tape are accommodated in a spool of small diameter which can be carried in a cassette. The same tape with a recording
bandwidth of 15 kHz would at the highest frequency have recorded a wavelength on the tape which can be shown (using
= f ) to be 3.2 µm.

In a video recording system [2] where, say, a 5 MHz bandwidth is required and where it is assumed that recording can take
place down to 1 µm wavelength a head to medium speed of 5 m/s would be required. With a fixed head recording system this
would then mean that 18 000 m of tape would be needed for 1 h playing time and this is clearly impractical. A different type of
transport system is necessary and the first approach to a practical alternative was delivered in 1956 by the introduction of the
transverse scan system developed by Ampex Corporation in the USA [2], [13]. In this system a tape moves at very low speed
past and in contact with the circumference of a spinning wheel on which are mounted four heads equally spaced (see Fig. 8).
Each of those heads in turn passes transversely across the tape at high speed, sufficient to record at the highest frequency
and to create tracks of recorded information across the tape. With the tape moving at a low speed it moves only a short
distance before the next head starts its traverse across the tape. In that way tracks are recorded extremely close together
and the amount of information per unit area of tape is very high. By packing the information this way a low usage of tape
results, because of the low tape speed, but a high bandwidth operation, because of the high head to medium speed.

Figure 8. Simplified picture of a transverse scan tape transport.

Transverse scan technology has today been superseded by helical scan technology as depicted in Figure 9. This is
employed commonly in a wide range of video recorders and in the VHS domestic system. Here a rotating drum, which
replaces the wheel of the transverse scan system, has two heads on its circumference and spins at high speed. The axis of
the rotation of the drum is no longer parallel to the tape direction but has been rotated through an angle such that the
recorded track is now diagonally across the tape. The tracks are now several centimeters long and a single head scan along
its track is sufficient to record or reproduce one frame of a television picture. Again there is a high head to medium speed and
a high track density to enable a high area density of information, a low tape speed and low tape consumption. These helical
scan systems work extremely efficiently and are now in widespread usage for video, digital audio, and digital data storage
systems.

Figure 9. Simplified picture of a helical scan tape transport.

For computer memory applications floppy disk and rigid disk systems are the norm. A transport consists of a disk which is
rotated at high speed on a suitable spindle and a mechanism for moving heads above appropriate tracks. In rigid disk
systems a head cantilever is used which is pivoted, like a gramophone pickup arm, and sweeps in an arc across the disk
surface to the appropriate track. In floppy disk systems the head is positioned on a cantilever which is moved radially across
the disk to the appropriate track.

1.2.6. Electronic Systems

As shown in Section a.c. Bias Audio Recording a sine wave record current of constant amplitude gives rise to an output
current which varies very strongly as a function of frequency and which also has a phase shift of 90 °. Thus, the magnetic
recording channel introduced distortions which, though they are linear, give an output which is not recognizable immediately
as a replica of the input. To correct for this distortion, equalizers are employed which take the output from the replay head
and filter it to a form which is a replica of the input. In an audio recorder, and usually in a video recorder, the equalizer has
characteristics which are the exact inverse of the recording characteristics and, since the decrease of output at high
frequencies depends on the tape and recording characteristics, they compensate for the limitations introduced by them. In
addition to this amplitude correction also a phase correction of 90 ° is needed. Practical equalizers are associated with
circuits having particular time constants which are specified in the literature [15]. These types of equalizers are referred to as
flat equalizers since they make the overall gain of the system flat, i.e., independent of frequency.

In digital recorders flat equalizers have also been widely used but they are just one of the equalization strategies which are

page 7 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
available [13]. In digital recording the requirement is not necessarily that the output signal from the equalizer is a replica of
the input, but that from the output from the equalizer makes it possible to detect whether a digit 0 or digit 1 had been
recorded. Thus the role of the equalizer is to render visible the recorded pattern in such a form that the data originally
recorded can be easily recovered. A number of equalizer strategies are available for use in digital recording systems. These
include peak detection, Nyquist filtering, and class IV partial response equalization. In all cases the equalizer compensates in
some way for the distortions introduced to the signal by the recording properties of the recording medium.

1.3. Magnetic Recording Media

A typical recording tape consists of a dispersion of magnetic fine particles in a nonmagnetic binder which is spread onto a
plastic tape backing, as shown very diagrammatically in Figure 10. The fine particles are usually less than 0.5 µm in length
which is an essential requirement for creating the high coercivity that is needed for magnetic recording media. The dispersion
of the particles is necessary to avoid their close proximities and the resultant effect of interparticle interactions and to achieve
a uniform coating. Figure 11 shows dispersions of three different types of particle and demonstrates how variable the sizes,
shapes, and orientations of particles can be. Despite these variations, tapes with high uniformity and with the required
magnetic hysteresis properties and recording performances can be obtained.

Figure 10. Simplified diagram of a magnetic recording tape.

Figure 11. Transmission electron micrographs of magnetic pigment particles

A) Cobalt-modified -Fe2O3; B) Passivated metallic iron; C) CrO2

Fine particles are used because their dimensions are sufficiently small not to have room for domain walls and so they contain
a single domain. Thus, domain wall motion, which takes place at low applied fields, is avoided and so are low coercivities.
But, if the particles have either shape anisotropy or a high uniaxial crystal anisotropy [16] (Fig. 12), the magnetizations are in
stable orientations when they are oriented along either senses of the easy axes. Consequently only the application of very
high magnetic fields in the opposite directions to those of the magnetizations will cause the magnetizations to change their
polarities from one sense to the other [16-18]. A magnetization which exists in only one of two stable states gives rise to a
rectangular hysteresis loop of high coercivity and when particles are packed in a binder with their easy axes aligned the
resultant tape has high coercivity and high hysteresis loop squareness. A detailed description of the properties of magnetic
tapes is given in [16].

Figure 12. Magnetic fine particles with A) shape anisotropy, B) uniaxial crystal anisotropy

In the production of magnetic tape the particles are mixed into a slurry and then milled for an extended period to separate the
particles and break down any agglomerates. The slurry is provided with various agents and chemicals for wetting and
lubrication, and with solvents. The slurry is then applied in thin layers to a plastic backing using coating machines with either
gravure roles, doctor blades, extrusion slips, or reversed role systems. While the coating is still wet it is passed through a
magnet assembly which aligns the easy axes of the particles along the coating direction, and it then passes through an oven
to drive off the solvents and solidify the tape. After removal of the solvents the coating on the tape contains close to 50 vol %
magnetic particles (Fig. 10). The tape may also have a coating on its reverse side which contains carbon to reduce static
electricity and/or friction, and sometimes carbon may be included in the magnetic coating itself to reduce friction between
tape and the record head.

The final coated tape often has a rough surface with particles protruding from it. To improve the surface finish the tapes are
passed between rollers to flatten their surfaces, a process called calendering. When a coating has been produced for use in
floppy disks their surfaces may well be further improved by polishing using a ceramic polisher.

1.3.1. -Fe2O3
-Ferric oxide recording tape is easily recognized by its red brown appearance and is widely used as a standard grade audio
tape. In addition, it has also been used in digital tapes for computer and instrumentation applications, and also in floppy and
rigid disks. As a recording material it dates back to the late 1930s and for most, if not all, of that period it has been the
mainstay of tape recording. Nowadays it is no longer considered an advanced medium. It is giving way to other materials in
the latest applications which demand higher density recording performances.

-Ferric oxide particles can be produced by a number of methods. Usually they involve chemical precipitation processes to
produce, first of all, needle shaped particles of -FeOOH which are then dehydrated to -Fe2O3, reduced to Fe3O4 and
finally reoxidized to -Fe2O3. The process steps require careful control to ensure the appropriate morphology of the
particles, and that they have the required needle shaped uniaxial anisotropy for high coercivity. This is particularly important
since -Fe2O3 has cubic anisotropy which would result in coercivities which are too low. Particle sizes are typically less than
0.5 µm.

page 8 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Properties of -Fe2O3 are shown in Table 1 where typical coercivities are seen to be in the range of approximately 25 –
30 kA/m.

Table 1. Magnetic properties of tapes and particles

Material Coercivity Tape coating Remanent Remanent Particle

Hc, kA/m thickness, magnetization magnetization bulk lengths,
µm coating, kA/m material, kA/m µm

-Fe2O3 25 – 35 5–8 80 – 130 375 0.25 – 1

CrO2 40 – 60 3–5 100 – 140 480 0.2 – 0.4
Co- -Fe2O3 50 – 75 3–6 100 – 140 0.3 – 0.5
Fe, FeCo 100 – 200 0.2 – 3 200 – 300 1700 – 1900 0.2 – 0.5
MET * 80 – 160 0.1 – 0.15 400 – 500
CoP 60 – 120 0.5 – 0.1 300 – 800
Co, CoNi, 60 – 200 0.02 – 0.05 300 – 800
CoNiCr
CoCr 60 – 200 –1 300 – 800
(perpendicular)

* MET = Metal-evaporated tape.

1.3.2. Chromium Dioxide

Chromium dioxide particles were introduced as recording media in the late 1960s. The fabrication processes produce
particles which are almost perfect crystalline whiskers of submicron dimensions and high axial ratios. They have higher
coercivities than -Fe2O3 particles and because of their smoothness can easily be oriented within the tapes to produce
media with high coercivities and high hysteresis loop squarenesses. Table 1 gives some properties.

CrO2 is produced by reacting CrO3 and Cr2O3 in an aqueous medium and at high temperature and pressure. The CrO2
particles are formed in parallel bundles and their shapes and sizes are influenced by the presence of additives during the
preparation process. Some passivation of the particles is needed because of their chemical reactivity. Typical particles are
less than 0.5 µm long, have axial ratios of 10 or more, and have coercivities in the range 40 – 60 kA/m. Figure 11 C is a
micrograph of typical CrO2 particles.

Since tapes produced using chromium dioxide particles have higher coercivities than tapes using -ferric oxide they are
capable of superior recording performances. They are used in higher grade audio tapes, video tapes (VHS) and some
computer tape applications. They are more expensive than -ferric oxide particles and to justify their premium price their
higher performances are a requisite.

1.3.3. Cobalt-Modified -Fe2O3

The coercivity of -Fe2O3 is too low for modern high density recording needs and increases in its coercivity have been
obtained by modification of -Fe2O3 with cobalt. In cobalt-substituted iron oxide some of the iron atoms are replaced by
cobalt atoms on appropriate atomic sites. This is achieved by introducing cobalt in a suitable form during the preparation
process for -Fe2O3 and allowing the various heating processes to diffuse the cobalt into the -Fe2O3. The presence of
cobalt results in higher crystalline anisotropy and ultimately a higher coercivity. However, the coercivities of these particles
depend on temperature and on time. These dependencies are undesirable. In cobalt-adsorbed -Fe2O3 layers of cobalt are
produced on the surfaces of the -Fe2O3 and without heat treatment the cobalt remains on the surfaces as a coating. This
increases the anisotropies of the particles to produce increased coercivities. The cobalt-adsorbed particles are found to have
a higher thermal stability than the cobalt-substituted ones and have been most widely used. Typical properties for tapes
prepared using these particles are shown in Table 1. A micrograph of some particles is shown in Figure 11 A. These tapes
have found widespread application in video recording systems (VHS) and also in audio tapes, digital tapes, and floppy disks.

1.3.4. Metal Particles

Particles of metallic iron are formed by the reduction of -Fe2O3 with hydrogen. This yields particles of pure iron which are
extremely reactive and prone to spontaneous ignition and which need some passivation before they can be used. The
passivation is usually achieved through a surface oxidation to produce very thin protective layers of oxides of iron. Inclusion
of the particles into tape occurs by a similar process to that used with -Fe2O3 although some special problems exist with
the dispersion of the metallic particles because of their tendency to cluster together due to high magnetostatic interactions.
Nevertheless very high coercivity tapes have been produced with good uniformity and very high recording performances.

Many different particle tapes have been produced which have different proportions of nickel or cobalt included in addition to
the iron. Only a few percent of nickel but more than 25 % of cobalt may be included. Some properties are shown in Table 1.

Besides benefiting from very high coercivities metal particle tapes also benefit from reductions in their thicknesses. The most

page 9 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
spectacular improvement has come via the dual layer coating process in which a thin layer of magnetic particles is coated on
the backing at the same time as an underlayer of TiO2 particles [19]. Layers of less than 0.5 µm thickness have been
achieved and this in combination with the high coercivity results in very high density recording performances which are
comparable with those achieved by metal film media.

Metal particle tapes are intended for use in the latest high performance digital recording systems. Digital audio, digital video,
and digital data recording are all relevant application areas where high density performance is a requisite.

1.3.5. Metal-Evaporated Tape (MET)

Thin metal film recording media have been produced by evaporation of cobalt or cobalt nickel alloy (typically Co80Ni20) in an
oxygen atmosphere and subsequent deposition onto plastic substrates. Deposition is usually at an angle of incidence other
than normal to the planes of the substrates (see Fig. 13). The films produced by this method are granular and the grains
have been shown to be columnar and inclined to the normal to the film planes [20]. The grain boundaries, which contain
metallic oxides, appear to create discontinuities between the grains and disrupt exchange coupling but also introduce
significant shape anisotropy in the grains. The crystalline easy axes are usually along the directions of the grains and these,
along with the shape anisotropies, cause the magnetizations to tip out of the tape planes and towards the axes of the grains.

Figure 13. Diagrammatic view of the vacuum evaporation process for producing metal evaporated tape.

The tipping of the magnetization out of the media planes results in recording properties which are slightly different from
typical longitudinal recording media [21]. The recording properties also depend on which way the media are moved passed
the recording heads. For example the granular structure, illustrated in a rather idealized way in Figure 14, moving from left to
right through a head field would give rise to slightly different recording properties if the tape were reversed and moved
through the same head field. Improved performances are observed with the orientation shown in the right-hand diagram.
However, the recording properties experienced as a result of the movement of the tapes in the two ways are only slightly
different and they remain essentially longitudinal in character [21]. Some properties of the tapes are given in Table 1.

Figure 14. The passage of a metal evaporated tape past a record head

The two pictures show the tape being moved in such a way as to produce different recording geometries.

The different recording characteristics for different directions of tape motion have been circumvented by the production of
two-layer and three-layer tapes as illustrated diagrammatically in Figure 15. These have been produced by sequential
depositions of the metals onto substrates moving in different directions for each subsequent layer. The increased production
difficulties are obviously undesirable and the three-layer tape is no longer produced as the two-layer tape gives rise to similar
recording properties for both directions of tape motion.

Figure 15. Representation of two-layer and three-layer metal evaporated tapes.

Metal films have suffered from a number of mechanical problems especially crazing of the coated films and low wear
resistance [22]. Wear resistance has been vastly improved by growth induced surface features (asperities/nodules) and by
the use of either diamond-like or graphite coatings on their wearing surfaces.

1.3.6. Metal Film Disk Media

Early disk recording machines employed -Fe2O3 media but these have now been exclusively superseded by thin-film
media. The latter are used partly because of their thinness but also because of the very high coercivities that are available.

The first thin films to come into use were produced by the electroless chemical deposition of cobalt phosphorus onto nickel
phosphorus-coated aluminum alloy substrates [20], [23]. The small percentage of phosphorus, up to about 5 %, is almost
insoluble in the depositing cobalt and so is pushed into the grain boundaries. This nonmagnetic layer between the grains
prevents exchange coupling across the boundaries and isolates them from one another (see Fig. 16). The result is that the
films behave like planar arrays of closely packed single domain particles and this leads to high coercivities. Film thicknesses
were of around 50 nm and coercivities were up to 100 kA/m and more.

page 10 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 16. Grain structure of a CoP film of thickness 90 nm and coercivity 60 kA/m. The grains are around 40 nm in
diameter.

The chemical deposition process has now given way to film deposition by sputtering. Cobalt films and cobalt – nickel films
have been prepared this way by deposition onto aluminum alloy or glass substrates which have been previously coated with
a chromium underlayer. The chromium of the underlayer permeates the magnetic coating and provides nonmagnetic
separating layers between the grains [24]. This again ensures discontinuity of the magnetic structure causing the grains to be
effectively separate particles and this results in high coercivity. This process has been supported by the inclusion of small
quantities of metals such as platinum or tantalum [25]. Film thicknesses are now down to less than 30 nm and coercivities
are up to around 200 kA/m. These properties are sufficient to allow extremely high densities of recording where linear
densities of over 6000 bits/mm and area densities of over 15 Mbits/mm2 have been demonstrated.

1.3.7. Perpendicular Recording

In perpendicular recording the recording medium is magnetized alternately in positive and negative directions normal to the
medium plane [26] as shown in Figure 17, contrary to the situation in longitudinal recording (see Fig. 2). To achieve this
orientation of magnetization a recording medium is needed which has an easy axis pointing in a direction normal to its plane
and there need to be heads which produce large magnetic field components in the same direction. Such an arrangement has
been achieved in practice with the use of high anisotropy media and it has been shown that perpendicular recording can
sustain very narrow recorded transitions and extremely high linear packing densities [26]. Despite this success it has proved
difficult to bring the technology into widespread and common usage.

Figure 17. Perpendicular magnetic recording on a medium with an easy axis directed normal to its plane.

The materials used for perpendicular recording are invariably cobalt – chromium alloys with cobalt : chromium ratios around
80 : 20 [27]. The recording layers are prepared by sputtering or evaporation, have thicknesses in the region 0.1 to 1 µm and
coercivities in the perpendicular direction of up to 200 kA/m.

The chromium is included with the cobalt to produce a lower magnetization so as to avoid demagnetizing effects arising from
the perpendicular orientation of the magnetization and thus make it easier for the magnetization to point out of the film media
planes. A typical crystal structure for a cobalt – chromium medium is shown in Figure 18 where it can be seen that in
comparison to longitudinal recording the grains have grown to become columnar in nature and have boundaries which are
relatively easily discernible. These boundaries are chromium rich and help to produce magnetic discontinuity between the
grains. It has been regularly proposed that these films behave as arrays of interacting fine particles, with magnetic
anisotropies due to shape and crystalline c axes being parallel to the long axes of the grains.

Figure 18. Transmission electron micrograph of a cobalt chromium thin film with its easy axis normal to the film plane.

Two-layer recording media have also been produced for perpendicular recording in which the cobalt – chromium layer has
been deposited onto a previously deposited soft magnetic nickel – iron coating, whose function is twofold. Firstly the coating
reduces the demagnetizing effects within the medium during the recording process, and secondly influences the nature of
fields produced by the record heads and thus also the record process (Fig. 19) [26].

Figure 19. Diagram of a two-layer perpendicular recording medium.

A high coercivity cobalt chromium thin film with its easy axis directed normal to its plane sits on a soft NiFe thin film.

Perpendicular recording has required the development of some new head structures, which provide large fields in the
directions perpendicular to the recording media planes. Nevertheless, ring heads have been used and these have performed
sufficiently well to facilitate high density recording. However, they are not the optimum and the range of new heads has also
been produced [26]. Figure 20 shows a single pole head situated above a recording medium. This consists of a single sheet
of material whose thickness corresponds to the gap in the ring head. The head is driven by a coil which is often mounted on
an auxiliary pole as shown in Figure 20 A and which is situated below the recording medium. When a current is passed into
this coil the auxiliary pole is magnetized and it produces a low magnetic field which magnetizes the main pole. The latter
produces a high localized field near its tip which penetrates into the recording medium and magnetizes it to achieve
recording. This type of structure has been shown to work but it is clumsy in that it requires poles either side of the medium
and the spacing between the two causes many difficulties. These difficulties have been alleviated by using a two-layer

page 11 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
medium shown in Figure 20 B, where the single pole above the medium carries its own coil and when driven this head
induces an image of itself in the soft magnetic underlayer below the recording medium. Therefore there is effectively a pole
above and below the medium and together they form a suitable record head structure. With these pole heads extremely high
density recording has been achieved and they represent the highest densities so far to appear in the literature.

Figure 20. Pole heads for perpendicular recording

A) A single pole and an auxiliary pole; B) A single pole and a recording medium with underlayer.

Particulate media have also been prepared for perpendicular recording although, as a generalization, high quality media
employing the same particles as used for longitudinal recording have not been successfully produced. However, barium
ferrite particles have been used [26] which tend to form hexagonal platelets with their magnetic easy axes normal to their
planes (see Fig. 21). When coated onto tapes the platelets usually lie with their planes parallel to the tape planes and so
have their easy axis in the perpendicular direction. Thus media have been produced with high anisotropy and this has given
rise to stable perpendicular magnetizations and high coercivities in that direction.

Figure 21. Recording tape containing barium ferrite particles.

Perpendicular media have been used for high density recording and have also been included in a commercial perpendicular
recording floppy disk drive. However, exploitation of this medium and its corresponding technology has been rather limited to
date mostly due to the continued success of the competitive longitudinal recording. In spite of that perpendicular recording
remains as having potential for future application in very high density disk storage systems.

1.4. Recording Heads

Ring Heads. The most widely used type of recording head is the so-called ring head [28]. Ring heads have been the
mainstay of magnetic recording throughout its existence and are expected to remain so for some time ahead although other
and more varied types may appear. A ring head is illustrated in a simplistic way in Figure 2, which gives a good idea of the
shape of the heads. They are often formed by bringing together two C cores (i.e., circular cores in the shape of the letter C)
on each of which a coil is wound and which are joined together. Gaps are formed at both the front and back of the head, the
one at the rear is simply a result of the fabrication process of the head and is not required, while the one at the front is
specifically designed along with pole shaping to produce a large field just below the gap. Variations on this design abound
and another common shape is made by bringing together a C and an I core. Here one half of the head is essentially straight
(I core) while the other is circular. This latter head was often used when two heads were to be placed side by side, and back
to back, to effect record and replay with one package.

Metal heads, typically of molybdenum permalloy (Ni – Fe – Mo alloy), Alfenol (Al – Fe) or Sendust (a Fe – Si – Al alloy), have
been widely used in audio and data recording and are fabricated from stacks of thin laminations. The higher the bandwidth
required the thinner the laminations. Although metal heads have high relative permeabilities (4000 – 11 000) and high
saturation magnetizations, (up to 800 kA/m), they are very much susceptible to eddy currents. For very high frequency
operation ferrite heads, typically Ni – Zn or Mn – Zn ferrites, are used as these are of very low electrical conductivity and very
much less susceptible to eddy currents. They suffer, however, from low magnetization (up to 400 kA/m) and, often but not in
all cases, low relative permeability (up to 10 000) and so they may have difficulty in recording on very high coercivity media.
Some alleviation from this problem can be obtained by using very high saturation materials to make the pole tip region of the
heads.

Metal in gap heads [29] are also used for a number of applications and in these various metals are included in the gap
region of the head. Their function is either to divert the flux from passing through the gap between the poles of the head and
out into the recording medium below, or to provide high magnetization in the pole tips to avoid saturation during recording.
Such heads are widely used in video recording systems.

Thin-film heads are now gaining wider acceptance. They are widely used on disk machines, in some digital data tape
systems, and also in some video and audio applications. A thin-film head is essentially a ring head but it is made using thin-
film technology and lithographic processes to produce a very small magnetic circuit of a ring type. Such a head made of
80 : 20 NiFe is shown in Figure 22. Overall their efficiencies are rather similar to those of the monolithic heads but they have
the advantage that they are more amenable to mass production.

Figure 22. Cross-section of a thin-film recording head.

page 12 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Thin-Film Flux-Sensitive Heads. For replay thin-film flux-sensitive heads have been produced and in diagrammatic form
are shown in Figure 23. The one in Figure 23 A consists of an element of soft magnetic material which is held in close
proximity to the tape and therefore it experiences magnetic fields emanating from the tape as it moves by. In
magnetoresistive elements any change of magnetization causes a change of resistance and so as the recording medium
moves by and its magnetic fields alter the magnetization in the element there is a change in resistance. This can be detected
using electronic circuitry. There are many different designs of flux-sensitive heads which use a number of different
magnetoresistive effects in single or multi-layer NiFe films. Such heads are very sensitive to magnetic fields and produce
high output voltages. This is good because of sensitivity reasons but it has a special benefit in that being particularly sensitive
it allows thinner recording media to be used and thinner recording media allow narrower transitions and higher recording
densities. Hence for special applications such as high density recording on rigid disks the use of magnetoresistive heads is a
major step forward.

The head shown in Figure 23 B is a shielded magnetoresistive head. Here the shields “absorb” the flux emanating from the
tape and shield the sensitive element except when the recorded transition is directly below. In this way the head only sees
the transition when it is directly below the element and the influence of nearby transitions is reduced. Hence the resolution of
the read back process is improved.

Figure 23. Idealized picture of (A) unshielded and (B) shielded magnetoresistive (MR) replay heads.

1.5. Economic Aspects

Magnetic recording is a technology which pervades all aspects of our lives. Be it audio, video or data recording its products
are in every household and in every company. Each of the categories of business is at the $ 109 per annum level and drives
huge industrial concerns to compete for the corresponding lucrative businesses. Both America and Japan have excelled in
these technologies although their research and development, and production facilities are dispersed and spread into many
parts of the world particularly in the Far East. In this way many other countries have benefited from the distribution of the
production facilities to parts of the world where appropriate labor at highly economic prices is available.

The total values of media delivered worldwide during 1997 are given below. Figures are in $ × 109.

Rigid disks 4.4

VHS video tapes 3.5
Audio cassettes 2.0
Data tapes 1.4
Floppy disks 1.4
Other audio visual tapes 0.9
Other disks 0.3

The total media business amounted to around $ 14.4 × 109 but when the media are incorporated into products and systems
the turnover of this enlarged industry is at least ten times or more of the figure quoted.

The audio market has been dominated by demands for cassette, particularly the compact cassette introduced by Philips, and
open reel tapes for a.c. bias recording systems. The cassettes are universal, and remain one of the most popular vehicles for
audio entertainment. The digital variants are DAT (Digital Audio Tape) and DCC (Digital Compact Cassette), which both
delivered very high quality audio, comparable in quality to compact discs, but which have not achieved such wide acceptance
in the domestic market. Both have, however, found their places in the professional world where their qualities are fully
recognized.

Analogue video recording systems are dominated in the domestic market by the helical scan VHS system which sustains its
leading position despite the fact that technology can now provide better performances on smaller cassettes. 8 mm wide tape
systems are available and are used in portable video camera systems. Cassette systems are also in use for professional
video applications. For the future on the domestic market Data VHS and digital video cassettes are likely to appear soon.
Many new systems, some based on fixed head technology, are likely future products and will have the ability to interface into
computer systems to increasingly enable multimedia applications. At the professional level digital video recorders and high
definition television (HDTV) systems have been developed and will find increasing application. Overall there is a massive
demand for recording media for video applications and this situation looks likely to continue for the foreseeable future
although there is expected to be a rising demand and competition from various compact and other optical disc systems.

Data recording for instrumentation and computer applications may involve tape, cassette, cartridge, floppy disk, and rigid
disk. All have there place in the system of storage hierarchy and have corresponding importance. Disk technology is central
to computer storage and it is especially in the area of rigid disk technology where the pioneering advances in recording
performance take place leading to extremely rapid improvements in terms of information storage densities and data rates.
3½ inch and 5¼ technologies predominate although 2½ inch and 1.8 inch disks have achieved positions in the market.
Smaller disk sizes have also been introduced for special applications.

[Top of Page]

page 13 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2. Optical Recording
Roman E. Wochele, Henk van Houten, Johan P. W. B. Duchateau, Hans J. G. Kloosterboer, Jan A. Th. Verhoeven, René van Vlimmeren, Peter E.
J. Legierse, Dirk J. Gravesteijn, C. David Wright, Herman J. Borg, Heinrich Heitmann, Jacques Heemskerk

2.1. General Introduction

2.1.1. Optical Storage Media
Optical storage started on a small scale with the analogue video disc, commercially launched in 1978. Since then, it has
become a mainstream technology due to the success of the Compact Disc digital audio system (CD-A) in the 1980s, the
Compact Disc Read-Only Memory (CD-ROM) in the 1990s and the Digital Versatile Disc (DVD) in the late 1990s. In all
cases, the digital information is encoded on a 12 cm diameter disc in the form of a relief structure, consisting of pits with
varying lengths. This relief structure can be replicated at low cost and in high volume by injection molding of polycarbonate
using a stamper carrying the desired pattern (see Section Optical Disc Mastering and Stamper Production).

The information layer in a Compact Disc is coated with a metallic mirror and covered by a protective lacquer. The layer is
addressed by a converging laser beam, incident on the 1.2 mm thick transparent polycarbonate substrate, and focused to a
diffraction-limited light spot [30], [31] (see Fig. 24). Reading through the substrate has the key advantage that the detrimental
effects of dust and small scratches on the read-out signal are minimized, since these are out of focus. As a result, optical
discs present a rugged, highly reliable, removable storage medium.

Figure 24. Substrate incident read-out of the information layer. The objective lens, corrected for the 1.2 mm polycarbonate
substrate, guarantees an aberration-free focused spot. Dust particles on the substrate surface are out of focus and therefore
they hardly deteriorate the read-out signal.

Cheap replicability, removability, and a long data life up to 100 years and more are the key technical reasons for the success
of the Compact Disc as the medium of choice for the distribution of prerecorded digital information. It should be realized that
removability also implies interchangeability of discs and players made by different manufacturers. This explains why
worldwide standardization is so important in optical disc storage.

In addition to read-only optical storage systems, write-once/read many (WORM)-type compact disc systems (CD-R, DVD+/–
R) as well as rewritable compact disc systems (CD-RW, DVD+/–RW, DVD–RAM) have been standardized [32], [33], [34].

Again, compatibility with their read-only counterparts is an important condition for widespread market acceptance. Such
compatibility has been achieved in the case of write-once discs by employing laser-induced pit formation in organic dye films
as a recording mechanism. An alternative solution is the laser-induced amorphous-to-crystalline phase change in thin films
made of inorganic alloys (see Section Write-Once Optical Recording Materials). Phase-change recording with appropriately
optimized materials is used for rewritable optical discs. Yet another possibility is magneto-optical recording, based on
magnetic domain reversal (see Section Reversible Optical Recording Materials), which is used, for example, in the 64 mm
diameter MiniDisc. However, magneto-optical discs are not compatible with CD/DVD, because they cannot be read by the
nonpolarizing optics of a standard CD/DVD player. The bit pattern in the various types of information layers is illustrated in
Figure 25.

Figure 25. Microphotographs of the information layer for various types of discs

A) Replicated CD; B) Ablative write-once system, based on a tellurium alloy; C) CD-RW (phase-change); D) MiniDisc (MO)

A compact disc offers 650 up to 900 Megabytes (MB) of storage capacity for ROM, WORM, and rewritable media. The DVD–
ROM provides 4.7 Gigabytes (GB) for a single information layer, 8.5 GB for a dual layer disc, and 17 GB for a double
sided/dual layer disc. Recordable and rewritable versions of the DVD offering 4.7 GB in a single-sided/single-layer disc
(DVD+/–R, DVD+/–RW, DVD–RAM) and 8.5 GB for a single-sided/double-layer disc (DVD+R9) are close to standardization.
The new high-density versions of optical storage media (HD-DVD) using a blue-violet laser for writing/reading provide 27 GB
(BD = Blu-ray Disc) [35], [36] and 30 GB (UDO = Ultra Density Optical) [37].

Future generations of optical storage should address the capacity domain over 100 GB. The roadmap for UDO shows for the
third generation a capacity of 120 GB. Innovation will be required on many fronts: the optical head, short-wavelength
semiconductor lasers, improved signal processing, modulation codes and error correction, single and multi-layer media, and
shorter wavelength mastering equipment. Some of the evolutionary and revolutionary options are briefly mentioned in Future
Developments. In Figure 26 a comparison of the bit pattern density of CD, DVD, and BD is shown.

Figure 26. Comparison of the bit pattern density of optical storage media

page 14 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

2.1.2. The Optical Channel

When it is reflected by the information layer, the diffraction-limited focused laser beam is modulated in amplitude, phase, or
polarization [30]. The relief structure in the replicated Compact Disc predominantly modulates the phase of the reflected light
beam. Phase modulation is also the main effect in write-once ablative dye systems, but phase-change recording results in a
strong amplitude effect. Magneto-optical storage exploits the polarization modulation on reflection of perpendicularly
magnetized domains caused by the magneto-optic Kerr effect. Discussions of magneto-optical data storage can be found in
[38], [39]. In the following, some of the generic physical characteristics of the optical channel, including tracking and focusing,
as well as the data path are described. For convenience, the Compact Disc system will be taken as a point of reference.

2.1.2.1. Focusing and Tracking (Servo Control)

The basic optical read-out principle is shown in Figure 27. A laser beam is focused through the substrate by an objective lens
onto the information layer on the spinning disc. Reflected light returns through the lens and is directed to a photodiode
detector. The diameter (full width at half maximum) of a diffraction-limited focused spot is given by /2NA, with NA the
numerical aperture of the objective lens. In Compact Disc players, NA is usually chosen in the range 0.40 – 0.52 (DVD: 0.6;
UDO/BD: 0.7/0.85), giving a spot diameter of about 1 µm. The high brightness of a laser is required to achieve an adequate
signal-to-noise ratio [30].

Figure 27. The light path in an optical recording system consists of a

a) Semiconductor laser; b) Half-silvered mirror acting as a beam splitter; c) Collimator lens; d) Objective lens for focusing the
beam. The lens is mounted on an actuator for the adjustment of the disc-lens separation; e) Optical disc; f) A double wedge
structure for the Foucault focus measurement; g) Detector for the beam reflected from the disc. The detector is a multiple
sensing device, and information about focusing and positioning of the beam with respect to the track on the disc is also
obtained during the read out.

The Compact Disc system uses a laser diode of group III-V elements (e.g., gallium arsenide) emitting at a wavelength in the
near-IR (780 nm). The CD-R system employs the same wavelength. The DVD standard is based on a red laser (650 nm), the
BD/UDO standard on a blue-violet laser (405 nm), using gallium nitride.

The diameter of the objective lens is chosen so that a free working distance (the separation between the lens and the nearest
surface of the disc) of more than 1 mm is achieved. The diameter of the converging beam as it enters the substrate is
typically 0.7 mm, resulting in reduced sensitivity to dust and scratches on the disc. As the substrate is part of the optical path,
its thickness, variations in thickness and angular orientation, optical homogeneity, birefringence, and surface flatness must
be included in the list of specifications of the optical system (see Sections Optical Disc Mastering and Stamper Production
and Read-Only Optical Recording Materials).

In order to be able to scan the information layer on the disc, two functions must be provided by the optical system of the
recorder: focusing and tracking.

Focusing. The depth of focus z of the laser beam is given in a good approximation by the equation:

This is of the order of a few micrometers for CD. It is completely impractical to specify the flatness and alignment of the
information surface to this degree. The lens – disc separation is therefore continuously adjusted during read-out. The focus
adjustment is made using servo control of an electromechanical actuator for the lens movement. A dynamic measurement of
the amount and direction of de-focus is required as input to the servo system. Several optical strategies are used, a brief
description of the two most popular methods will illustrate the general principles.

Figure 28. Schematic drawing of the Foucault double wedge principle for generation of the focus error signal. The incident
laser beam (arrow) is deflected by a beam splitter, and is focused onto the disc. The reflected light is transmitted by the
beam splitter, and reaches a second focus close to the apex of a double wedge. With four detectors the focus error signal is
extracted by coupling the detectors according to the scheme (top + bottom detector) minus (sum of two middle detectors).
The focus error signal changes from a negative value (bottom) to a positive value (middle). At the focal point (upper) the
error signal is exactly zero.

page 15 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Foucault Method. In the Foucault method, depicted in Figure 28, the apex of a prism (or double wedge) is placed on the
optical axis at the ideal focal point of the returning beam. Four photodetectors, arranged in a line perpendicular to both the
optical axis and the apex, are placed behind the focal point. The difference signal from the photodetectors yields an S-
shaped focus error signal over a range of ± 4 µm with respect to the ideal focal position. The acquisition range, which is of
importance during the initial focusing action, is much larger (up to 25 µm). This signal can be used to position the objective
lens.

Figure 29. Astigmatic focusing method. In this picture a plan-parallel plate is used for generating perpendicularly-oriented
astigmatic lines on either side of the best focus position.

Astigmatic Focusing Method. In the astigmatic focusing method (Fig. 29) a cylindrical lens is placed in the returning beam so
that perpendicularly oriented astigmatic lines are generated on either side of the best focus position. The relative intensity of
these lines can be measured using an appropriately oriented quadrant photodetector. The output is again used to give the
desired S-shaped focus error signal.

Tracking. In addition to focusing, a dynamic measurement is required in the optical system to detect and follow the
information track on the disc. In the Compact Disc, the track pitch is 1.6 µm, and the laser spot should be centered on the
track to within 0.1 µm. Several methods are used, as illustrated in Figure 30.

Figure 30. Different radial tracking methods

A) Three-spot tracking: the two satellite spots, generated by a grating, have a radial offset of 1/4 of the track pitch and an
offset in the track direction of typically 20 µm. The d.c. content of the satellite spots is measured on separate detectors and
will vary as a function of the radial position of the scanning spot.

B) Radial wobble method: an extra movement is given to the scanning spot relative to the tracks. From the intensity
fluctuations on the detector it is possible to extract a radial error signal.

C) Radial push-pull method: the light flux through the pupil of the objective is separated in two parts and from the difference
signal a tracking error signal is derived. Variations in light intensity over the pupil are due to the track structure, acting as a
diffraction grating.

D) Differential phase detection: this method uses a quadrant detector over the cross-section of the returning beam and
extracts a radial error signal by comparing the phases of the high frequency signals on the four quadrants.

Three-Spot Tracking Method. The three-spot tracking method (Fig. 30, first mechanism) employs a grating structure in the
light path to generate two satellite spots from the main beam. The grating is oriented so that the radial offsets of the satellite
spots are 1/4 of the track pitch, with a considerable (ca. 10 to 20 µm) tangential offset. If the main spot is on the track, the
satellite spots ride partly on the land between the tracks, both preceding and following the central read-out spot. By
measuring the d.c. content of the satellites on separate detectors and by establishing their difference signal, a tracking error
signal can be generated.

Radial Wobble Method. In the radial wobble method (Fig. 30, second mechanism), the scanning spot is given an oscillatory
movement in the radial direction. A radial error signal is deducted from the phase and intensity of the modulated reflected
signal at the oscillation frequency.

Radial Push-Pull Method. The radial push-pull tracking method (Fig. 30, third mechanism) exploits the fact that the track
structure of the disc acts as a diffraction grating, with for CD a period of 1.6 µm. Optical interference occurs at the detector
between the zeroth and two first-order beams diffracted from the track structure. Using a detector split along the plane of
symmetry of the far-field radial diffraction pattern, a tracking error signal is derived from the difference signal. The same split
detector can be used for the read-out signal by summing the two contributions. This is the central aperture signal.

For prerecorded discs, the information itself is sufficient to form the required grating structure. In this case, the phase depth
of the relief structure is chosen to be near to, but not equal to, the value /4n, The exact value /4n leads to a maximum in
the central aperture read-out signal, but to a minimum push-pull signal. For recordable discs, no information is initially
present. In this case, it is customary to manufacture a continuous pregroove (a groove structure replicated to the disc) for
tracking purposes. The depth of the pregroove is usually chosen to be close to /8n. The exact value leads to a maximum in
the push-pull tracking signal, but the central aperture read-out signal is significantly reduced. A variant of the push-pull
tracking method is the three-spot push-pull method, frequently used in CD drives.

Sampled Servo Method. The sampled servo tracking method (not shown in Fig. 30), used in some magneto-optical drives,
requires special patterns of optical effects, present on the disc at predetermined intervals. The read-out of the pattern is used
to adjust tracking and time-base errors; the servo system is then disabled until the next special pattern is encountered.

Differential Phase Detection. Finally, differential phase detection (DPD) (Fig. 30, fourth mechanism) is a method which
derives a tracking error signal from the timing difference of signals detected by different parts of a four-quadrant detector.
Such a phase difference occurs when the diffraction limited light spot impinges with a radial offset on a pit or written mark on
a disc.

page 16 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Lens Adjustment. The focus and tracking error signals serve to adjust the objective lens, using actuators. The objective lens
is commonly mounted on an electromechanical actuator driven by the optical focus error signal. The maximum vertical
displacement of the lens for focusing is of the order of 1 mm. For tracking, the entire optical pick-up unit (OPU) can be
mounted in a linear or rotating carriage assembly. This low-bandwidth coarse displacement actuator is commonly
supplemented with a high bandwidth, fine displacement actuator which drives the objective in the radial direction, to
compensate for the inevitable track eccentricity.

In an alternative approach, most of the components of the optical pick-up unit (such as laser, detectors, beam splitters, and
collimating lens) are mounted rigidly into the optical recorder. The collimated laser beam is sent to a unit that normally
includes an actuated deflecting mirror for fine radial tracking and the objective lens, both mounted on a movable carriage
[30], [31].

These servo systems are designed to relax the tolerances that must be placed on the geometrical parameters of the optical
disc. Nevertheless, the performance of the servo systems is limited in terms of the magnitude and bandwidth of the
corrections that can be accomplished (see, e.g., [30] ).

2.1.2.2. Data Channel

For analogue video disc systems, the standard composite video signals are directly frequency modulated, and recorded on
the disc (at carrier frequencies of 7.6 or 8.6 MHz). The signal path of a digital optical storage system is analogous to that of a
digital communications system, as depicted in Figure 31 for a rewritable optical disc system.

Figure 31. Block diagram of the read–write channel of an optical disc recorder.

In the recording process (write channel), an analogue input signal is digitized (A/D converter), and an error correction code
(ECC) is added. For the CD digital audio system, the audio signal is sampled at a rate of 44.1 kHz, which according to
Nyquist's sampling criterion is sufficient to reproduce audio signals up to 20 kHz [30]. The number of bits is 32 per sample
(16 per stereo channel), resulting in a bit stream of 1.41 Mbit/s. The audio bits are grouped into frames containing 6 samples
each. For each 3 original audio bits, 1 bit is added for error correction, leading to a bit stream of 1.83 Mbit/s. Even though
error correction is used, the raw bit error rate must be low (below 10-4). The final data stream is interleaved or dispersed over
several code words on the disc in order to circumvent the effects of optical disc defects such as pinholes [cross-interleaved
Reed-Solomon code (CIRC)]. One more byte is added to each audio frame for playback and display functions.

To adapt the binary data pattern to the modulation transfer characteristics of the optical channel, a modulation code is
applied, before the data are stored on the disc. A run-length limited sequence of bits is chosen, with a minimum run length of
3 and a maximum of 11 channel bits. This choice has been made to achieve an optimal match to the spatial frequency
characteristics of the optical channel, and to realize an optimal data density. The chosen modulation scheme is known as
Eight-to-Fourteen Modulation (EFM), because each byte of audio data (including CIRC and sub code data) is transposed
through a look-up table into 14 channel bits [41]. Three further merging bits are added. The channel bit rate is 4.3 Mbit/s.

Both the encoding and the error correction are treated in detail elsewhere [41]. Recordable and rewritable discs require
additional signal processing to generate the appropriate laser pulses needed for the chosen write strategy, as discussed in
Section Compact Disc Family. The general aspects of encoding and signal processing of magneto-optical recording systems
are similar, although the details differ [38], [39]. In the case of a read-only, replicated disc, the write process is replaced by
the disc mastering process.

The read channel is similar, but is followed in reverse order (see Fig. 31). In the read-out process, extra steps are needed to
determine the proper timing information (clock recovery), and to compensate for the particular frequency response of the
channel (equalization). The read-out of such a disc is discussed in some detail in what follows. Modulation of the reflected
light intensity from the optical disc is caused by the interaction of a diffraction-limited spot with optical effects on the disc
having characteristic dimensions of the order of the spot size. In such cases the modulation can be modeled by considering
the effects of light diffraction by supposedly periodic structures (amplitude or phase gratings) on the disc [30], [42]. A rigorous
treatment of the coupling requires a complete vector diffraction theory solution of Maxwell's equations; the complexity of this
approach has so far prohibited its widespread application. Fortunately, scalar diffraction theory can give a reasonable
prediction of the relationship between optical amplitude or phase structures on the disc and the read-out signal
characteristics. The optical response to the spatially modulated phase structures on the disc is described by the modulation
transfer function (MTF). For central aperture detection of the optical signal, the MTF decreases monotonically from its
maximum value at low spatial frequencies on the disc, to zero at a spatial cut-off frequency given by 2NA/ . The modulation
approaches zero at the cut-off frequency because the length of a pit plus the space between pits then equals the diameter of
the diffraction limited laser spot. The cut-off frequency determines the maximum information density that can be stored on the
disc.

In good approximation, the amplitude of the detected signal is proportional to the modulation depth of the effects on the disc,
modified by the frequency dependent MTF. Translated into the signal domain, the maximum signal frequency that can be
extracted from an optical disc is 2vNA/ , with v the linear velocity of the track. The frequency dependence of the MTF results
in timing jitter due to inter-symbol interference (ISI). An additional source of jitter is cross talk from adjacent tracks, but this is
not yet significant in the compact disc. Electronic filtering may be used to reduce ISI by boosting the high frequency
response, and thereby compensating for the MTF decay (equalization [43]). The practical frequency range of the recorded
signal is typically limited to about 75 % of the cut-off frequency. The noise sources in the optical detection system also
include electronic effects such as the dark current noise of the photodetector and the current amplifiers, the ultimate limit

page 17 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
being photon shot noise. From the media the important noise sources are the microroughness of the information-carrying
surface and optical inhomogeneities and anisotropy of the substrate (e.g., birefringence). All these sources produce
amplitude noise in the detected signal [44]. In addition, irregularities in the definition of the pits or recording marks on the disc
are a cause of noise. Especially in write-once and rewritable media timing jitter associated with the microstructure of the
information layer and the recorded mark edge definition are critical factors. The geometry of the optical marking effects in
these media is difficult to predict, as it is only indirectly related to the intensity profile of the laser beam, because of the
nonlinear and nonuniform thermal processes involved.

The EFM modulation code, jointly with the frequency dependence of the modulation transfer function of the optical channel,
and the non-linearities in the photodiode transfer characteristic give rise to the eye pattern of a CD bit detector, as illustrated
in Figure 32 bottom. (The term eye pattern derives from the similarity of the lozenge-shaped pattern formed by adjacent zero-
crossings of the oscilloscope traces). The EFM choice yields a read-out signal with a small low frequency content of the
information on the disc, so that the focus and tracking servo systems are not influenced by the low-frequency components of
the data stream.

Figure 32. The information is present on the disc in the form of pits in a reflective surface. When the laser light hits a pit, the
intensity of the reflected light is much lower than of the light reflected back from the reflective surface between the pits.
These differences in intensity are translated into electrical signals. The eye pattern is the result of triggering the digital signal
to the rising edges of the signal. The clearly opened eye pattern is the result of a virtually perfect combination of optical disc
and read-out light path.

In the ensuing sections on optical recording materials the concept of carrier-to-noise ratio (CNR) is used as a figure of merit.
This CNR is the signal-to-noise ratio of a single frequency written on the disc (the carrier). The signal and noise are
measured in narrow band widths centered at and adjacent to that frequency, respectively. Although precise predictions on
the performance of optical recording media cannot be made on the basis of CNR, it gives a quick impression about the
potential usefulness of the optical recording material.

2.1.3. Compact Disc Family

2.1.3.1. Read-Only, Replicated Discs
The Compact Disc (CD) digital audio system appeared in a coordinated international introduction in 1983. Since then,
derivatives of the CD system have emerged that use extensions of the original CD standard prepared jointly by Philips and
Sony in 1979. The major derivatives are CD-ROM and CD-Recordable. Their technical specifications are described in the so-
called “Red Book”, “Yellow Book” and “Orange Book”, respectively [45], [46]. Table 2 shows an overview of the CD books.

Table 2. Overview of the CD books, standards, and specifications

Book Description, specifications Dates (of latest version), comments

Red Book defines the physical properties of the
(CD Audio) compact disc and the digital audio encoding:
z audio specification for 16-bit PCM.
z disc specification, including physical
parameters.
z optical stylus and parameters.
z deviations and block error rate.
z modulation system and error
correction.
z control and display system and CD
CD Audio (Aug 1995, Red Book);
subcode/control and display system
(Nov 1991, extension to Red Book for
CD Graphics); CD Text (Sep 1996,
extension to Red Book)
the Red Book is also available as the
international standard ISO/IEC 60908

graphics
allows for up to 74 min of digital sound.

Sample rate of 44.1 kHz, or 44 100 samples

per second. Transfer rate of 150 kilobytes
per second. Also known as "single-speed" or
1X. can contain up to 99 tracks.
Yellow defines the CD-ROM specification plus an
Book (CD- extension for CD-ROM XA:
ROM & z disc specification, optical stylus
CD-ROM parameters, modulation and error
XA) correction and control and display
system (from Red Book)
z digital data structure (the sector
structure and the ECC and EDC for a
CD-ROM disc)
z disc format including Q channel and
sector structure using Mode 2 sectors
(CD-ROM XA)
CD-ROM (Nov 1988, original Yellow
Book); CD-ROM XA (May 1991, CD-
ROM XA system description)
The Yellow Book is also available as
the international standard, ISO/IEC
10149
Yellow Book, Mode 1, ISO 9660 Level 1
is the most widely utilized standard
since it will allow virtually all PCs to
access its data. Mode 2 will permit the

page 18 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
z data retrieval structure based on ISO "combining" of Mode1 and Mode2 data
9660 including file interleaving whichon a single disc. Mode 2 is primarily the
is not available for Mode 1 data (CD- entertainment portion of the Yellow
ROM XA) Book standard. CD-ROM/XA, Bridge
z audio and graphics encoding discs (including Photo CD, Karaoke
CD, and Video CD), andGreen Book, or
CD-i, fall under the Mode2 standard of
The Yellow Book standard defines two data Yellow Book. Mode2 discs are intended
types. Mode 1 describes CD-ROM data with to be played on dedicated consumer
Logical Error Correction Code (LECC), which electronics platforms.
provides room for 2048 bytes of user data
and is the mode used to store data that is
unforgiving of error, like computer programs
or databases. Mode 2 describes sector data
stored without LECC, which provides
increased room for 2336 bytes of user data.
This mode is typically used for data that is
more tolerant of error, like audio, video, or
graphics.
White Book defines the Video CD specification: Karaoke CD (1993, Original VCD for
(Video CD) z disc format including use of tracks, Karaoke); Video CD (Apr 1995, Current
Video CD information area, segment Video CD specification); Internet (Apr
play item area, audio/video tracks and 1997, extension for linking to websites);
CD-DA tracks Super Video CD (Nov 1998, Higher
z data retrieval structure, compatible quality video using MPEG-2)
with ISO 9660
z MPEG audio/video track encoding the White Book specification for Video
z segment play item encoding for video CD was announced by JVC, Philips,

sequences, video stills and CD-DA SONY and Matsushita in July 1993. A
tracks special CD-ROM/XA subset designed
z play sequence descriptors for pre- to hold MPEG-1 video. White Book
programmed sequences defines Mode 2, Form 2 data standard
z user data fields for scan data and which creates a disc that can contain
closed captions up to 74 min of full-motion video.
z Examples of play sequences and Limited with a slow playback rate and
playback control only 680 MB of storage space, the DVD
standard is designed to replace the
troubled Video CD. MPEG stands for
Motion Picture Experts Group.
Blue Book defines the (also known as CD Extra) Enhanced Music CD (Dec 1995)
(enhanced specification for multisession pressed disc:
CD) z disc specification and data format a subset of the Orange Book
including the two sessions (audio and specification, the Blue Book standard
data) was designed expressly for stamped
z directory structure (to ISO 9660) multisession discs limited to two
including the directories for CD Extra sessions, one music and one data. Also
information, pictures and data known as CD Plus, Blue Book is a
z format of the CD Plus information defined, licensed standard supported
files, picture file formats and other by Philips, Sony, Microsoft, and Apple.
codes and file formats Discs play on CD-Audio players with no
z MPEG still picture data format possibility of producing static and on
computers with newer CD-ROM drives.
Photo CD defines the specification for Photo CD: specified by Kodak and Philips based
Book z general disc format including example on the CD-i Bridge specification

of program area layout, index table,

volume descriptor, data area, Last updated: Dec 1994
subcode Q-channel skew, CD-DA
clips and microcontroller readable
sectors
z data retrieval structures including

directory structure
z image data coding including a

description of image coding and

image packs
z ADPCM files for simultaneous

playback of audio and images by

interleaving
z playback program system including

playlist files
Orange defines CD-Recordable discs with Part I: CD-MO (Nov 1990) Magneto
Book (CD- multisession capability: Optical disks
R and CD- z disc specification for unrecorded and

RW) recorded discs Part II: CD-R (Dec 1998) CD-WO -

page 19 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
z pre-groove modulation Write Once
z data organization including linking
z multisession and hybrid discs Part III: CD-RW (May 1998) Rewritable
z recommendations for measurement of

reflectivity, power control etc.

Purple defines the specification of double-density the informal name for specification
Book compact disk (DDCD): produced by Philips and Sony in 2000
(DDCD) z narrowing the track pitch to 1.1 µm to describe their double-density
from 1.6 µm compact disc (DDCD) format, which
z Shortening the minimum pit length to increases the storage capacity of the
0.623 µm from 0.833 µm disc to 1.3 GB through means such as
z new type of error correction (CIRC7) increasing its number of tracks and pits.
z adaptation of the ISO 9660 file format Purple Book-defined products were
z scanning velocity of 0.9 m/s released in recordable and rewritable -
rather than read-only - formats, and
require special CD-RW drives to handle
them.
CD-i defines specification for CD-i Bridge CD-I Bridge is a Philips/Sony
Bridge z disc format defining CD-i Bridge discs specification, for discs intended to play

as conforming to the CD-ROM XA on CD-i players and other platforms

specification such as the PC. Confusion reigned
z data retrieval structure as per ISO temporarily when Philips put MPEG-1
9660. A CD-i application program is video on Green Book discs, which
mandatory and stored in the CDI could only be played on CD-i machines.
directory However, all new MPEG-1 film titles
z audio data coding which includes conform to the White Book standard for
ADPCM and MPEG Video CDs, which may be read on any
z video data coding for compatibility White Book-compatible machine
with CD-i and CD-ROM XA including CD-i and suitable CD-ROM
z multisession disc structure including drives. The video and sound are
sector addressing and volume space compressed together using the MPEG
z CD-i related data since all CD-i 1 standard, and recorded onto a CD
players must be able to read CD-i Bridge disc.
Bridge data.
Green defines the CD-i disc format: CD-i is a real interactive system
Book (CD- z Volume Compact Disc Interactive Full delivering various qualities of video,

i) Functional Specification various qualities of audio together with

z Microware OS/9 Technical Manual full user interactivity in real time. CD-i
was created by Sony and Philips mainly
to provide entertainment by interfacing
with televisions and stereo systems
(May 1994). Compact Disc Interactive
is the only specification which not only
defines the disc and the data, but also
the entire hardware and operating
platform.
ISO 9660 ISO 9660 (International Standards Organization) is the file and directory naming
standard that refined High Sierra to be readable on any platform, regardless of the
content format. ISO-9660 specifies each file name consist of three components:
name, extension and version. A name or extension may consists of zero or more
characters of the set [A through Z], [0 through 9] and the _ (underscore). The version
number ranges from 1 to 32767. There are three levels of this standard defined by
ISO-Files are listed in alphabetical order. A directory name consists of one or more
characters of the set. The Volume Name of the disc can contain up to 11 of the set.
Directories are sorted alphabetically. ISO 9660 Level 1 - Known as the "DOS"
Character set, the number of characters is restricted to 8 and the number of
characters of the extension is restricted to 3. Also referred to as 8.3 (eight-dot-three).
ISO 9660 Level 3 - The total length of the name and extension is restricted to 30
characters (excluding the point). ISO 9660 (Joliet) - The total length of the name and
extension is restricted to 64 characters. ISO 9660 Rock Ridge Interchange Protocol -
Permits UNIX and variants to use the naming standards of the UNIX platform and
performs like a native volume. Hybrid - Permits "Partitioning of the disc" to allow the
disc to perform natively in two or more platforms, operating systems, or environments.

Further members of the CD family, such as CD-i (interactive CD), Photo-CD, Video-CD, and the rewritable CD-RW, have
also been standardized. It is generally recognized that a key factor behind the success of the Compact Disc system is the
high level of interchangeability between the different members of the family. As discussed in [46], this interchangeability is a
result of standardization at various levels. The standardizations related to the media hardware include test conditions of the
optical system, characteristics of the disc, the read, write and erase characteristics (i.e., the interaction of the laser with the
information layer), and the data and servo formats. Software related standardization deals with file labels and formats. To
allow true interchangeability, the CD-related standards are well protected internationally. Aside from the system and media
components, the CD (audio as well as ROM) has had a very significant effect on the recording (software/content) industry.

page 20 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
This industry, therefore, plays an important role in setting future standards for optical disc systems.

The CD-ROM, containing up to 650 MB of user storage capacity, has been designed for the large scale distribution of
computer software or large-scale data bases. After a slow start (it was first introduced on the market as early as 1985), it has
now become a standard attachment for a PC. Since computer applications typically have a random bit stream, the CD-ROM
cannot rely on some of the interpolation techniques used in CD-Digital Audio to handle detection errors, whereas also the
addressing information needs to be bit-accurate to individual data frames. For this reason the CD-ROM requires some
additional error detection and correction encoding compared to the basic CD-Digital Audio disc [43]. The last years have
witnessed a rapid succession of CD-ROM drive generations, with ever increasing data rates (from the original CD-Audio data
rate up to 52 ×). The drive spins at 52 × at more than 10 000 rpm. To reach these speeds several writing/reading strategies
such as CLV, CAV, Z-CLV and P-CAV have been developed (Fig. 33.)

Figure 33. Overview of CD writing/reading strategies

Constant Linear Velocity (CLV). The disc is read/written at a constant speed. The data transfer rate is kept at a fixed level
by changing the rotation speed.

Constant Angular Velocity (CAV) The disc is read/written at a constantly increasing speed. The drive maintains a
consistent rotation speed which will result in a variable transfer rate.

Partial-Constant Angular Velocity (P-CAV): The disc is being read/written at an increasing speed until a certain point
(speed). After this point the speed will not increase anymore and remain at this level. P-CAV is a combination of CLV and
CAV. The transfer rate increases until the drive reaches its maximum speed (CAV). From that point the drive will slow down
and the transfer rate will be constant (CLV). Because P-CAV drives reach their maximum speed much sooner than CAV
drives, the average transfer rate should be higher.

Zoned-Constant Linear Velocity (Z-CLV). The disc is divided into several CLV zones. After each zone the write speed is
increased (note that the Z-CLV mode is not used when reading a disc). Most high-speed recorders are using this technology.

Regarding the actual trend CDs will most likely continue to be popular for music recording and playback and for smaller
quantities of computer data storage. Newer technologies like the digital versatile disc (DVD) store much more data in the
same space and have higher transfer rates.

2.1.3.2. Recordable and Rewritable Discs

Various recordable versions of the compact disc have been proposed. An early magneto-optic version of the CD (CD-MO,
described in the orange book, part 1) was not a success, because of incompatibility with the CD. A 12 inch write-once/read
many (WORM) disc was introduced as early as 1984, based on organic dyes in the information layer. It was intended as a
replacement for magnetic (Winchester) discs in computers. Again success was rather limited. A 12 cm disc with 650 MB
capacity, compatible with CD, was defined based on the WORM format (orange book, part 2). Known as a recordable
compact disc (CD-R) it is now rapidly penetrating the market. It can be used in a multi-session mode, by allowing consecutive
recording sessions for a single disc (at the expense of capacity loss due to the need to define lead-in and lead-out sections
for each session). The CD-R can be used for applications such as photo-archiving, back-up, or the copying of read-only discs
for personal use [45].

A rewritable CD based on phase-change recording technology has been announced in 1995, and shown at industry briefings
in fall 1996 under the name of CD-RW. In phase-change recording, laser pulse sequences are used to write a mark (see
Fig. 34)

Figure 34. Principles of phase-change recording. Amorphous marks are written by using a short, high-power laser pulse to
melt the recording material and subsequently quench it to below the crystallization temperature, so that nucleation and
growth of the crystalline phase is avoided. Information is erased by heating the amorphous state to a temperature between
the crystallization temperature and the melting point long enough to regain the crystalline state.

In between the pulses, the laser power is turned to a low level, so that after each laser pulse the molten material is quenched
to below the crystallization temperature, resulting in an amorphous dot. A series of overlapping amorphous dots forms a
recording mark. The reflection coefficient in the amorphous state is low, so that a recording mark corresponds to a pit in a
CD. Direct overwrite can be accomplished by switching the laser power to the erase level in the regions between two marks
(see Fig. 35). For an optimized erasing sequence the material is heated to a lower temperature (below the melting point), for
a time long enough to regain the high-reflectance polycrystalline state (similar to the high reflectance of the CD “lands”,
formed by the regions in between the pits replicated in the polycarbonate substrate).

page 21 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 35. Direct overwrite strategy for phase-change recording. Each recording mark is formed by a train of short laser
pulses which result in overlapping amorphous dots. Direct overwrite can easily be accomplished by switching the laser to the
erase level between the writing of successive information marks.

The drives for recordable and rewritable CDs are similar to their read-only counter parts, although a higher-power 780 nm
semiconductor laser is required, and some additional electronics to generate the write strategies.

Almost every CD-R can hold between 660 and 690 MB data but usually the total space of a CD-R is physically set to 650 –
659 MB. The extra space is normally used for the mandatory 90-s lead-out area. By shortening this lead-out area, extra data
can be written on the CD-R. This method of writing a CD-R is called oversizing or overburning. The reason that the total
space of a CD-R is set to maximum 659 MB is because most CD-writers are not able to write more than this amount of data
to a CD-R (because of mechanical or firmware shortcomings). Thus bigger CD-Rs are useless for these CD-writers. A CD-R
is in fact recordable up to the end of the physical pre-groove. To oversize CD-R's the Disc-At-Once (DAO) mode of the CD-
writer is utilized. In the DAO mode, instead of TAO (Track-At-Once), the writer must provide its own lead-out (which now can
be moved to the very end of the CD-R).

In 1998 the 80 min/700 MB CD-R's came on the market. When they were released they were quite expensive ($ 25) but
within a year the price dropped considerably ($ 2). The secret behind the CD-R 700 (80 min), 800 (90 min) and 900 (100 min)
is that every CD-R has a spiral (pregroove), which starts on the inside and ends on the outer part of the CD, where the data
is written. The laser head will always follow this spiral from beginning to end to read (or write) data. The distance between the
start- and end-position of the spiral is defined in the CD Red-Book and is the same for all 63, 74, 80, 90, and 100 min CD-R
media. So the spiral has to be wound tighter to fit 80 min (or more) data in the same space where normally only 74 min data
resides. 80, 90 and 100 min CD-R's can be written by almost any CD-writer. However, a few CD-writer manufacturers (such
as HP and Sony) do not support “big CD-R’s” as these do not comply to the CD Red-Book standard.

2.1.4. Digital Versatile Disc (DVD)

The success of the compact disc family is based on the availability of replicated read-only discs, which are a removable
medium ideally suited to the low-cost distribution and preservation of digital information. Recordable (and rewritable) versions
became successful only after compatibility with read-only discs had been achieved. Removability and compatibility require
standardization, and this hampers incremental increases in the storage capacity, in contrast to non-removable media such as
hard discs. At first, such a capacity increase was not needed, because a 650 MB CD-ROM offered far more capacity than
could be handled by early PCs. However, due to the rapid progress in processing power typical file sizes have increased
dramatically (from 1 MB in 1982 to 500 MB in 1995 [45]). In addition, the digitalization of audio, graphical, and video
information has led to the emergence of multimedia PCs. Thus, a high density successor to the CD was needed.

After considerable debate about the respective merits of various proposed solutions [45], a new standard has been laid down
in the blue book for the digital versatile disc, or DVD-ROM, in fall 1996. A DVD-ROM [47] can store 4.7 GB data on a single
information layer — a 7-fold capacity increase over a CD-ROM. The required increase in areal data density was made
possible by a number of changes in physical parameters, listed in Table 3. The basic idea is to reduce the diffraction-limited
spot size used for read-out, given by /2NA, through the choice of a shorter wavelength diode laser (650 nm instead of
780 nm), and a larger numerical aperture of the objective lens (NA = 0.6 instead of 0.45). This enables a reduction of the
track pitch and of the channel bit length. The physical bit density was thereby increased by a factor of 4.7 (see Table 3 and
Fig. 37). This comes at the expense of a severe reduction in operating margins, such as disc tilt and focus error. These
margins were kept at an acceptable level by reducing the substrate thickness from 1.2 mm to 0.6 mm. Because of the large
installed base of CD-Audio and CD-ROM media, the same physical format was required (12 cm diameter, 1.2 mm thickness).
A DVD therefore consists of two 0.6 mm substrates bonded back-to-back. A double-sided version of the DVD was thus made
possible, allowing a total capacity of 9.4 GB. In addition, a double-layer version of the DVD has been standardized, with a
capacity of 8.5 GB per side (see Fig. 36). Fig. 38 shows an overview of the DVD read only formats.

Figure 36. DVD double layer structure. Information layer 0 is covered with a semi-reflective layer and is separated by a
spacing layer from information layer 1. The latter is covered with a full reflective layer. Read-out of both information layers is
assured by switching focus.

Figure 37. Comparison of CD and DVD

Figure 38. Overview of the DVD read only formats, a) DVD-5 (single-sided, single-layer, 4.7 GB), b) DVD-9 (single-sided,
double-layer, 8.5 GB), c) DVD-10 (double-sided, single-layer, 9.4 GB), d) DVD-18 (double-sided, double-layer, 17.0 GB)

page 22 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

In addition to the changes in physical bit density, changes were also made in the data format and the channel code (from
EFM to EFM+, an eight-to-sixteen code). The DVD is more sensitive to disc defects due the increased density, and thus
requires a more powerful error correction scheme. The cross-interleaved Reed–Solomon code (CIRC) of the compact disc
was replaced by the Reed–Solomon Product Code (RSPC), which is so effective that it takes up a smaller fraction of the
capacity (13 % instead of 30 % [46]).

To enable multimedia applications in both consumer electronics appliances and as a computer peripheral, an open file format
was adopted for the DVD [Universal Disc Format (UDF) an extension of the CD-ROM ISO 9660 file format]. The DVD
standard accommodates high-quality digital compressed video according to the MPEG-2 variable bit rate standard. When
used for video applications, a DVD can also accommodate 8 channels of audio and up to 32 channels of sub-titles. The
format is backwards compatible with the older MPEG-1 standard used in the Video-CD.

Table 3. From CD to DVD

Parameter CD DVD

Wavelength , nm 780 650

Numerical aperture NA 0.45 0.60
Track pitch, µm 1.6 0.74
Channel bit length, nm 277 133
Shortest pit/land length, nm 831 399
Modulation code* EFM EFM+
Physical bit density, Mbit/cm2 106 508
Reference velocity CLV, m/s 1.2 4.0
Substrate thickness D, mm 1.2 0.6
Spot size /2NA, µm 0.9 0.55
Capacity, GB 0.65 4.7
Depth of focus ( /2NA2), mm ±2 ± 0.9
Tilt margin ( /NA3D), ° ± 0.6 ± 0.3
Thickness margin ( /NA4), mm ± 0.1 ± 0.02

* For EFM one has 17 channels bits (14 modulation bits and 3 merging bits) for 8 data bits. Each channel bit
corresponds to of the minimum mark length. The physical bit density equals 1/(track pitch × channel bit
length × 17/8). For EFM+ the factor 17/8 is replaced by 16/8.

The coexistence of DVD and CD calls for the development of players that can accommodate both types of disc. This is
nontrivial, because of the differences in physical parameters listed in Table 3. For example, the difference in substrate
thickness requires optical pick-up units that can read out information layers through substrates of 1.2 and 0.6 mm thickness,
while the objective cannot correct the different amounts of spherical aberration simultaneously. Acceptable and creative
solutions have nevertheless been found [48].

Similar to the rainbow books for the Compact Disc, there are also specifications for the DVD (see Table 4).

Table 4. Overview of the DVD books, standards, and specifications

Part 1 Part 2 Part 3

Book name Physicala File systemb Applicationc

A DVD–ROM read-only ISO 9660 + UDF undefined
B DVD–Video read-only UDF MPEG-2 video
C DVD–Audio read-only UDF high quality audio
D DVD–R write once UDF not defined

page 23 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
DVD–RAM and DVD–RW rewritable UDF
E RW not defined

aPart 1 defines the physical format of a DVD–ROM disc. Various categories of information are defined:
z Physical - Disc size, curvature, reflectivity, etc.
z Sectors - Data format of the sector headers and ECC blocks.
z Layout - Lead-in/out areas, data areas. Differences between physical and logical sector numbering.
z Basic data-format and location of sectors that describe the disc type and layout (e.g. layer count, physical, and
logical sector address mapping, etc.)
bPart 2 defines the file system DVD uses a UDF-Bridge file system. That is, both an ISO-9660 and a UDF file system
structure are present on the disc. This is to allow backward compatibility for file access.
cPart 3 defines the DVD–Video The DVD–Video specification defines the format of all files stored in the VIDEO_TS
directory on a DVD–Video disc (IFO, BUP, VOB). Unlike the physical and file system specifications, this information is
not officially available.

In contrast to recordable and rewritable CDs, two main competing standards and three recordable DVD formats exist on the
DVD writing side, i. e., DVD–RAM, DVD–R/RW and DVD+R/RW. DVD–R/RW (this is a combination of DVD–R and DVD–
RW-note the minus in the name), mainly developed and promoted by Pioneer, has taken an early lead. The other competing
standard, DVD+RW (note the "plus" in place of the dash), mainly developed and promoted by Philips, making a strong bid.
All in all there are six recordable versions of DVD: DVD–R for general, DVD–R for authoring, DVD–RAM, DVD–RW,
DVD+RW, and DVD+R. DVD–R and DVD+R can record data once, like CD-R, whereas DVD–RAM, DVD–RW, and
DVD+RW can be rewritten thousands of times, like CD-RW. DVD–R was first available in fall 1997. DVD–RAM followed in
summer 1998. DVD–RW came out in Japan in December 1999, but was not available in the USA until spring 2001. DVD+RW
became available in fall 2001. DVD+R was released in mid-2002.The situation is similar to that of the competing VHS and
Beta video standards decades ago. But unlike the video tapes of old, the two main competing DVD standards are designed
to create discs that may be read by most consumer DVD movie players. Drives of both standards can read commercial DVD
movies and work as CD readers and burners. Current DVD–recorders are multiple-format writers, that support both formats,
DVD+R/RW and DVD–R/RW.

The third DVD rewritable standard, DVD–RAM, runs in the PC-world a distant third. A DVD–RAM is initially more a
removable storage device for computers than a video-recording format, although it has become widely used in DVD video
recorders, because of the flexibility the system provides in editing a recording. This technology is incompatible with
commercial DVD players — data stored on a DVD–RAM disc cannot be read by DVD movie players and many other DVD
drives. DVD–RAM discs are housed in cartridges and hold between 2.6 and 9.4 GB.

The double-layer format for recordable DVDs, announced in 2003 by Pioneer (DVD–R9) and Philips (DVD+R9) right after
each other, has once again shown that the format fight is not over. Double-layer rewritable discs are still under development.
The double-layer DVD+R system uses a thin organic dye film as data storage layer [34]. The recording principle is the same
as used in single-layer recordable DVDs and is based on irreversibly modifying the dye's physical and chemical structure,
induced by heating with a focused laser beam. The recorded marks have optical properties different from their unmodified
surroundings, giving the same read-out signals as read-only discs. The digital information is contained in the length of the
recorded marks and the unmodified spaces between them. Adding an additional storage layer, that can be accessed from the
same side of the disc, provides challenging issues in organic dye-based optical recording. The transmission of the upper
recording layer is > 50% to allow for read-out and recording of the lower recording layer. In the upper layer of DVD+R9 a thin
silver-alloy layer is used as reflector material. The reflection of the upper layer should be at least 18% to ensure compatibility
with the double-layer DVD–ROM standard. The lower recording layer has a high reflection (> 50%) and high power
sensitivity, because the upper recording layer absorbs and reflects part of the incoming light. Effectively, at least 18%
reflection of the lower layer is required. The transparent spacer layer is about 55 µm thick and serves to physically separate
the two recording layers. The laser beam can easily be focused on either of the two layers by adjusting the position of the
objective lens. So the double-layer recordable DVD will probably be readable in all DVD players, the point is, that the existing
DVD recorders have to be updated or exchanged. The future will tell if this will be as a big success as the current single-layer
DVD–Recordable. In Table 5 an overview of the currently available recordable DVDs (April 2004) and their specifications is
given.

Table 5. Overview of the specifications for recordable DVDs

Media DVD–RAM DVD–RW DVD+RW DVD–R DVD+R DVD+R9

Published by yes yes no yes no no

DVD–Forum
Hardware Panasonic / Pioneer / Philips / Pioneer / Philips / Philips /
maker Toshiba / Sharp / Ricoh / Panasonic Ricoh / MKM
Hitachi / Sony Sony Sony
Samsung
Promotion Recordable RWPPI DVD +RW RWPPI / DVD +RW DVD +RW
organization DVD Council Alliance Recordable Alliance Alliance
DVD Council
Book version Ver 2.1 Ver 1.1 Ver 1.1 Ver 2.0 / for Ver 1.0 n/a
general
Capacity/ 4.7 GB 8.5 GB
sided (double

page 24 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
layer)
Wavelength, 650 (0.6) 650 (0.6) 650 (0.6) 650 (0.6) 650 (0.6) 650 (0.6)
nm (NA)
Channel bit 0.14~0.146 0.133 0.133 0.133 0.133 0.133
length, µm
Track pitch , 0.615 0.74 0.74 0.74 0.74 0.74
µm
Depth, nm ~70 ~30 ~30 ~180 ~180 n/a
Media type Phase change Dye
Active layer GeSbTe AgInSbTe AgInSbTe cyanine / azo dyes
material
Reflectivity, % 15~25 18~30% 18~30% 45~85% 45~85% > 18 %
Recording 2700 1350-2700 1350- 1350-10800 1350- 3240-n/a
speed, kB/s 10800 21600
Tracking land/groove groove groove groove groove groove
method
Addressing CAPA LPP wobble LPP Wobble Wobble
Format ZCLV CLV CLV/CAV CLV CLV/CAV CLV/CAV
Tracking P-P DPD DPD DPD DPD DPD
Recording random random / random / sequential sequential sequential
sequential sequential
Finalization without with with with with with
Defect yes no no no no no
management
Copy CPRM CPRM CPRM - - n/a
protection
Cartridge selectable no no no no no
Cyclability, 100 000 1000 1000 once once once
times

2.1.5. HD-DVD and Blue Laser Discs

Blue laser systems are able to store more data than DVDs because of the shorter wavelength of blue light. This means that
the laser, which is used to record data on the disc, produces a smaller spot on the recording layer, which in turn means that
the space needed for each bit of data is smaller. HD-DVD (HD stands for both high-density and high-definition) was under
development before DVD came out and finally emerged in 2003. Some high-definition versions of HD-DVD use the original
DVD physical format but depend on new video encoding technology such as H.264 to fit high-definition video in the space,
that used to hold only standard-definition video. High-density formats use blue or violet lasers to read smaller pits, increasing
data capacity to around 15 to 30 GB per layer. High-density formats use high-definition MPEG-2 video and may also use
advanced encoding formats. Early in 2004 are six systems are under development, announced or released (see Table 6).

Table 6. Summary of specifications for HD-DVD (under development, announced, or released)

Format Backers Data Laser Video Capacity (single Data

depth layer/dual layer) rate

HD-DVD DVD Forum 0.6 mm blue HD MPEG-2, 15 GB / 30 GB (ROM)20 36

(AOD) (405 H.264, VC9,MPEG- GB / 40 GB (recordable) Mbit/s
nm) 4
Blu-ray Blu-Ray Disc 0.1 mm blue HD MPEG-2 27 GB / 50GB 36
Founders (405 Mbit/s
nm)
UDO Sony/ 0.1 mm blue n/a 30 GB / 60 GB 64
Plasmon (405 Mbit/s
nm)
EVD eWorld (Govt. 0.6 mm red (650 HD MPEG-2 (later n/a / 8.5 GB (ROM) 22
of China) nm) AVC) Mbit/s
FVD 1 ITRI (Taiwan) 0.6 mm blue AVC 17GB / n/a 25.05
(405 Mbit/s
nm)
FVD 2 ITRI (Taiwan) 0.1 mm blue AVC 17GB / n/a 31.59
(405 Mbit/s
nm)

page 25 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
HD-DVD (AOD). The DVD Forum's next-generation DVD, once called Advanced Optical Disc (AOD), currently being called
HD-DVD, is a modification of the existing DVD physical format to enable storage of about 15 GB per layer using a blue-
ultraviolet readout laser. The same 0.6-mm data depth as in DVD is used. AOD is designed to improve data capacity while
theoretically being able to use existing replication equipment. It is primarily supported by Toshiba and NEC.

Blu-Ray Disc (BD). The first HD-DVD format announced was the Blu-ray disc. This format was developed outside of the
DVD Forum, that supports the successful DVD–Video format. Blu-ray is a new high-density physical format that will hold 23
to 27 GB per layer using a blue-ultraviolet laser and a 0.1-mm data depth. Because of the 0.1-mm cover layer it will require
significant changes to production equipment. Blu-ray is initially intended for home recording, professional recording, and data
recording. Mass-market distribution of prerecorded movies will come later, after the read-only format, called BD-ROM, is
developed and the details of video, audio, interactivity, and copy protection are hammered out. Blue-ray supporters are LG,
Panasonic, Philips, Pioneer, Hitachi, Mitsubishi, Samsung, Sharp, Sony, and Thomson. Sony released the first BD recorder
in Japan in April 2003. In Figure 39 the development in data density from the CD to the Blu-ray disc is shown.

Figure 39. Comparison of CD, DVD and Blu-ray technology

The most important technical details of the Blu-ray disc are up to 27 GB per layer using 0.1-mm recording depth (to reduce
aberration from disc tilt), a 405-nm blue-violet semiconductor laser with 0.85 NA lens design to provide 0.32 µm track pitch
(half that of DVD) and a pit length as small as 0.138 µm. Variations include 23.3 GB capacity with 0.160-µm minimum pit
length (used by Sony’s Professional Disc system) and 25 GB capacity with 0.149-µm minimum pit length. The physical discs
use phase-change groove recording on a 12-cm diameter, 1.2-mm thick disc, similar to DVD–RW and DVD+RW. The data
transfer rate is 36 Mbit/s. Recording capacity on a single layer is about 2 h of HD video (at 28 Mbit/s) or about 10 h of
standard-definition video (at 4.5 Mbit/s). The cartridge size is 129 ×131 × 7 mm. There are plans to produce dual-layer
recordable discs, holding about 50 GB per side, but such discs will take a few additional years to appear.

UDO. UDO (Ultra Density Optical) is a 5.25-inch format with a recording density of 7.4 Gb/in2, mainly developed by Plasmon
and Sony. The target market for UDO is professional optical data storage and secure long-term archival storage for valuable
business information. The system is technically similar to the Blu-ray disc format, that was developed by a consortium of nine
companies led by Sony, although both systems are incompatible. The main difference between the two formats is in the data
transfer rate. Blu-ray, which is aimed at consumer recording of high-definition television, can record data onto the discs at a
rate of up to 36 Mbit/s to match the data rate of digital television. However, UDO can record data in the first version at 64
Mbit/s or double the rate of Blu-ray. The first version of the UDO format is a single-sided, single-layer optical disc with a
capacity of 30 GB. The roadmap extends to a 60 GB capacity version by 2006 at 96 Mbit/s and a 120 GB capacity version by
2008 at 144 Mbit/s. UDO uses a patented phase-change recording process, that permanently alters the molecular structure
of true write-once media, ensuring data integrity at the most fundamental level. UDO is also available with rewritable media
for archive environments, where data needs to be deleted or media capacity reused. Unlike true write-once media, rewritable
media allow the phase change recording process to be reversed. Enhanced error correction algorithms (Reed-Solomon) and
read-ahead defect management complement UDO's phase change recording process to guarantee high data integrity.

EVD. A government-backed consortium of companies in China, called eWorld, has developed a domestic version of DVD
called EVD (Enhanced Versatile Disc). EVD uses red laser discs but with tighter tolerances than DVD to hold more data.
Video is encoded in HD MPEG-2, although a future version will use the China-developed AVC compression format. EVD
players first appeared in China at the beginning of 2004.

FVD 1 and FVD 2. The Advanced Optical Storage Research Alliance (AOSRA), formed by Taiwan's Industrial Technology
Research Institute (ITRI) has its own variations of blue-laser formats. FVD 1 uses a 0.6-mm data depth similar to AOD, and
FVD 2 uses a 0.1-mm data depth similar to Blu-ray.

2.1.6. Magneto-Optical Storage

Magneto-optic recording is intrinsically a rewritable optical storage technology [38], [39]. The medium is a disc consisting of a
transparent substrate, covered with a thin magnetic film with perpendicular magnetic anisotropy (so that the magnetization
vector is oriented perpendicular to the plane of the film). A typical material is the rare earth – transition metal alloy TbFeCo.
Initially, the disc is magnetized uniformly. Writing occurs by a thermomagnetic process (see Fig. 40). A focused laser spot is
used to heat the thin permanent magnetic layer locally above its Curie point (typically 200 °C). A small coil is positioned at
the other side of the disc, opposite to the laser spot. It creates a weak magnetic field (a few hundred Gauss) that causes
reversal of the magnetization vector in the heated region of the film. The size of the written domain can be less than the spot
diameter. The written domain is quite stable at room temperature, because of the large coercive field of the materials used in
magneto-optic recording. In comparison with hard discs or floppy discs, MO discs are much less sensitive to stray magnetic
fields, so that they are an attractive removable storage medium.

Figure 40. Principle of magneto-optical recording. A focused laser beam L heats the magnetic layer R to above the Curie
point, so that local magnetization reversal because of the applied external field H can occur. The magneto-optical effect
rotates the plane of polarization of the linearly polarized read-out spot. The sense of the rotation inverts when the direction of
the magnetization is reversed.

page 26 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Read-out of an MO disc can be done with the same laser, used at a lower power. The read-out mechanism is based on the
magneto-optic Kerr effect: the direction of polarization of the laser light is rotated on reflection off the magnetized domain.
This is detected by means of polarizing optics.

A disadvantage of magneto-optic recording in its simplest form (i.e., writing with a constant external magnetic field, and
modulated laser power) is that it does not allow for direct overwrite (erasure requires an external field of opposite sign). This
limitation may be overcome by keeping the laser power constant, and switching the direction of the external magnetic field
(magnetic field modulation, or MFM). This technique requires a very small electromagnet, flying at a very short distance from
the disc surface. An alternative way to achieve a direct overwrite capability is to make use of exchange-coupled thin film
stacks, in combination with laser-power modulation [50], [51].

To circumvent the recording density limit imposed by the size of the focused laser spot, the magnetic amplification technique
has been proposed. A second magnetic layer with a lower Curie point, for example GdFeCo, is sputtered on top of the
recording layer. When heated above about 100 °C by means of the focused laser spot, the material covered by the spot
copies — through exchange coupling — the magnetization of the underlying information layer. In the presence of an external
magnetic field of the proper orientation, the size of the magnetized area is comparable to that of the laser spot, which can be
considerably larger than the size of the bit mark written in the recording layer. In reverse external field, no magnetic field
amplification occurs, and the read-out signal vanishes [50], [51].

MO recording drives have not been very successful as a PC peripheral, or on the consumer electronics market. This may be
due in part to the fact that no standard file format exists for magneto-optic discs (comparable to the ISO 9660 format for CD-
ROM). The exchange-ability of MO discs is therefore limited. MO recording has thus mainly addressed professional and
niche markets.

An exception is the recordable version of the 64 mm diameter MiniDisc system, which is aimed at the consumer electronics
market. It is gaining acceptance, especially in Japan. Its general specifications are very similar to those of the CD: it is based
on a laser with 780 nm wavelength, an objective with a numerical aperture of 0.45, the modulation code is EFM, and error
correction is done by an advanced form of the cross-interleaved Reed-Solomon code. The MiniDisc can be used to record
74 min of compressed digital audio. For portable applications the shock-sensitivity is improved by using a semiconductor
buffer memory.

2.1.7. Future Developments

2.1.7.1. Evolutionary Progress in Optical Disc Storage
The introduction of the DVD and the new blue laser discs is following the increase in file size, which itself follows the increase
in capacity of hard disk drives. The storage capacity of magnetic hard disks is currently increasing at a rate of about 50–60%
per year. The areal density of current high-end hard disks is more than 60 Gbits/in2. As shown in Figure 41, the physical bit
density of state-of-the art magnetic hard disks exceeds that of an optical medium. There is currently much debate on the
fundamental limits of both optical and magnetic storage. Magnetic recording areal density has increased approximately
2.5 × 106 fold since 1957, but researchers are now debating the ultimate limits imposed by superparamagnetic effects and
media limitations such as grain size. On the optical side, new advances in recording films such as “superresolution” and more
sophisticated encoding/decoding schemes are extending fundamental limits imposed by the available laser technologies. It is
expected that this rapid progress will continue unabated in the next decade. A similar rate of progress is anticipated for
semiconductor memories (DRAM). Also the processing power of microprocessors is rapidly increasing. Typical file sizes will
also keep growing, allowing new applications in domains such as multi-media, or 3D interactive graphics. It is, therefore, to
be expected that in the future a higher-capacity successor of current optical media will be required.

Figure 41. Areal density development of physical density of optical and magnetic storage media and DRAM

A possible path to an even higher capacity replicated optical disc, as already used in the blue laser discs, is offered by the
combination of shorter wavelengths (GaN lasers will enable 410 nm or even 360 nm recording), thinner substrates, and
higher numerical apertures. A high NA may be achieved by a dual objective [52]. The extra lens could either be mounted on
a second actuator (requiring additional error signals for servo control), or on a slider, as in a hard disk drive. In its ultimate
form, the extra lens is a solid immersion lens (SIL) in near-contact with the optical disc. An NA > 1 is then possible [53].
Points of concern are contamination and damage of the disc surface. This is a problem shared with other forms of near-field
(or scanned probe) storage [54].

In addition, the possibility exist to use optical superresolution. This technique uses an active layer or additional coatings
which change their optical properties under intense laser illumination. The net effect is to spatially filter the central part of the
laser spot. In one form (called rear-aperture detection) the coating replaces the aluminum mirror in a compact disc. If a
chalcogenide material such as Ge2Sb2Te5 is used, it melts under strong laser illumination, and becomes highly reflective. In
a second form (called front aperture detection) a complementary effect is used, and only the central part of the focused laser
spot can reach the information layer. In both cases it is possible to read out premastered marks with a size below the
diffraction limit [55].

Finally, as in magnetic storage, more sophisticated detection techniques, such as partial response maximum likelihood (or
Viterbi) detection, in combination with new formats (such as land/groove recording) may be considered.

page 27 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
In view of the nature of optical disc storage as a standardized, removable medium, starting in the market with mass-
distributed replicated prerecorded media, it should be expected that the capacity increase will proceed by big steps, as in the
transition from CD to DVD, rather than incrementally as in the hard disk drive industry.

2.1.7.2. Three-Dimensional Optical Storage

It is quite conceivable that even higher storage capacities, 100 GB and beyond, will be required for the consumer market in
the future. One can then resort to juke boxes of optical discs, or to magnetic or optical tape recording [56]. Alternatively,
forms of three-dimensional optical storage are under development. The third dimension can be spatial (holography, multi-
layered bit recording [57], two-photon addressing of thick media), spectral (spectral hole burning [58], or a combination of
both (holography combined with spectral coding [59]). Early in 2004 InPhase Technologies presented the first generation of a
200 GB holographic drive, that uses the blue laser technology (405 nm). The system will arrive at the market in 2006. In
holographic data storage, light from a coherent laser source is split into two beams, signal (data-carrying) and reference
beams. Digital data to be stored are encoded onto the signal beam via a spatial light modulator. The data or strings of bits
are first arranged into pages or large arrays. The 0's and 1's of the data pages are translated into pixels of the spatial light
modulator that either block or transmit light. The light of the signal beam traverses through the modulator and is therefore
encoded with the “checkerboard” pattern of the data page. This encoded beam then interferes with the reference beam
through the volume of a photosensitive recording medium, storing the digital data pages. The interference pattern induces
modulations in the refractive index of the recording material yielding diffractive volume gratings. The reference beam is used
during readout to diffract off the recorded gratings, reconstructing the stored array of bits. The reconstructed array is
projected onto a pixelated detector that reads the data in parallel. This parallel readout of data provides holography with its
data transfer rates up to 800 of Mbit/s. The readout of data depends sensitively on the characteristics of the reference beam.
By varying the reference beam, for example by changing its angle of incidence or wavelength, many different data pages can
be recorded in the same volume of material and read out by applying a reference beam identical to that used during writing.
This process of multiplexing data exploits the enormous storage capacity of holography.

2.2. Optical Disc Mastering and Stamper Production

2.2.1. General
Commercial production of optical media requires large scale and reliable replication of the optical media substrates with the
relevant physical structures. These submicron structures have to provide the well specified optoelectronic signals. Based on
technical and commercial criteria, molding or embossing technologies are presently the method of choice. Both these
technologies require a “mother disc” which is called “stamper”. This is generally a nickel-based disc which can be mounted in
a given substrate replication system.

Mastering and stamper production refer to a family of processes for making this “mother disc”. The first step in the creation of
the relevant mother disc is the mastering process. Master manufacturing is followed by making a one-to-one replica from that
master by a well-established electroforming procedure which yields the stamper [30], [60-62]. A typical production chain of
optical discs is shown in Figure 42 [63]. The information to be recorded is brought into a standardized format by “data
formatting” (including error correction methods).

Figure 42. Flow diagram for the production of optical media.

The formatted data is written by a laser beam recorder (LBR) on the prepared master disc, which is subsequently developed
and a thin metal layer is applied before a nickel copy can be made from the master by electroforming. This nickel stamper is
then used for mass replication of optical media substrates. The substrates are covered with a reflective metal layer, a
protective coating and a printed label in the disc finishing processes.

The media requirements for playability (staying in track and being synchronized, mark formation during recording of user
data) and for manufacturing strongly determine the shape and the tolerances of the physical structures. Therefore first a
review on these requirements will be presented.

2.2.2. Requirements
Playability Requirements. For writable media (write-once, rewritable media), the physical structures have to provide reliable
(virtually error-free) tracking and allocation or timing. As depicted in Figure 43 tracking is normally done by a pre-embossed
groove (often called pregroove), by a “pseudo pregroove” consisting of a continuous sequence of pits, or by a structure of
regularly positioned pits (sampled servo tracking). The tracking technologies based on these concepts are well proven to
provide reliable tracking of the laser beam during writing, erasing and/or reading (see Section General Introduction).

Figure 43. Examples of tracking provisions. Continuous groove, pseudo groove, and sampled servo.

Addressing a package of data on the disc can be done by allocation procedures using (1) sectors with pre-embossed sector
headings; (2) by using discrete time flags along the tracks; (3) by timing procedures using embedded time marking along the
pregrooves either by depth and/or width modulation; or (4) by a radial wobbling around the central track line (Fig. 44).

page 28 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 44. Typical provisions for addressing data on optical media.

For MO both sampled servo and continuous groove are shown.

The CD-R concept embodies a pregroove for tracking and a wobbling feature of the track for timing and/or addressing. The
ISO-MO standards use a combination of a pregroove with pits on the land between the grooves for sectorization to provide
data allocation. For DVD-RAM a combination of on-land and in-groove recording is chosen with sector headers positioned
half way between the track center lines.

Read-only media have to provide the data too. A common principle is to use the sequence of pits, which represents the
encoded user data (audio, video, or data and their error detection and correction information) also for tracking and timing.
The sequence of pits acts as a pseudo-track, and the data is grouped in specific frames in such a way that a number of these
groups can be used to provide addressing (and timing) information.

Table 7 gives an overview of specific functionalities for the physical structures and their implementations. The dimensions of
the physical structures depend strongly on the vacuum wavelength of the read-out laser and the refractive index n of the
substrate. In going to shorter wavelengths for read-out, smaller physical structures have to be made by mastering.

Table 7. Physical structures of various optical media and their functionalities

Optical medium Tracking Timing location Data Track pitch Typical pit
provisions recording sizes*

Laser Disc pseudo groove time flags analog pits 1.7 µm depth: 0.13 µm
width: 0.5 µm
Compact Disc pseudo groove frame no. digital pits 1.6 µm depth: 0.13 –
0.15 µm
width: 0.5 µm
CD-Recordable grooves radial wobble in groove 1.6 µm
CD-Rewritable grooves radial wobble in groove 1.6 µm
Mini Disc grooves radial wobble in groove 1.6 µm
(recordable)
ISO-MO grooves pre-recorded on land 1.6 µm –
sector headings 0.9 µm
DVD-ROM pseudo groove frame no. digital pits 0.74 µm depth: 0.11 –
0.12 µm
width: 0.25 µm
DVD-Recordable grooves 0.74 µm
DVD-RAM

* Only for pre-recorded media. Writable media show a large variation in sensitive layers and therefore in depths and
widths of the physical structures.

In addition, the recording performance of writable media depends on the groove shape too. Rewritable systems struggle in
general with signal-to-noise problems. The presence of a groove decreases the reflected signal and thereby the signal-to-
noise figure. In order to avoid this, the signal-to-noise ratio can be improved by either recording the data pits on the flat area
between the grooves (“on-land” recording) or by making the grooves rather wide and flat (“wide groove” recording). This is
illustrated in Figure 45.

Figure 45. Examples of groove recording

a) In-groove (small groove); b) On-land; c) Wide-groove

The marks represent the user written data.

The “on-land” recording format is standardized for the ISO-MO media. Figure 46 shows an atomic force microscopy (AFM)
picture of the pre-recorded header of “on-land” recording.

page 29 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 46. AFM picture of MO format showing grooves and sector headers.

Specific microscopic structures have to secure a good recording performance. Hence, optoelectronic tracking and
synchronization specification as well as recording performance determine the optimal physical structure.

Each deviation from the intended structure will be reproduced in the final product. In addition, statistical phenomena as well
as reproduction imperfections contribute cumulatively with the subsequent production steps. It is the final disc which has to
provide virtually error-free tracking, clocking and data recovery. This puts high requirements on the mastering technology,
both for its resolution as well as for its quality level and secure large number replication. A first parameter to characterize the
quality of the formed pits during mastering is jitter; being the statistical deviation from the pit/land and land/pit transitions with
respect to the clock.

Disc Manufacturing Requirements. The playability imposes, in general, specific requirements on the microscopic structure
of the physical structures. For example, data pits give optimal optoelectronic signals if their microstructure shows a flat
bottom, steep edges, and a depth of one quarter of the effective wavelength of the play back laser. For ease of replication,
however, a stamper should have no sharp cornered pits. The slopes involved would be as shallow as possible. This makes
the replication or embossing process easier. Control of the parameters which determine the physical microstructures enables
an optimal compromise between the requirements involved.

2.2.3. Mastering Methods

Several technologies have been explored to produce a master disc. Mechanical pressing of pits into a relatively soft and
ductile metal layer by actuation of a piezo driven diamond-stylus impression tool was one of them [64]. Technologies based
on ion and electron beams have been investigated too.

On the basis of commercial criteria and on technical requirements for resolution, reliability, and microstructuring of the
physical structures, optical laser mastering is presently the preferred industrial choice.

Optomechanical System for Laser Beam Cutting. The production of a spiral or concentric track is achieved by rotation of
the master disc and horizontal shifting of the laser spot. The light beam's intensity is modulated in the optical modulator in
order to generate the required physical structures (see Fig. 47).

Figure 47. Schematic of a laser beam recorder writing on a master disc.

Laser beam recorder systems for advanced formats, such as ISO-MO, use dual beam systems to generate simultaneously
both a groove and pit structures. Given the submicron precision of the location and the size tolerances of these structures, a
variety of technologies are used in order to secure stability in the optical, mechanical, and electrical subassemblies of the
optomechanical system for mastering.

In the laser beam mastering process, the limits of the recording process are determined by the size of the focused spot of the
laser. The most ideal spot formation can be obtained by focusing an optical beam free of aberrations (i.e., from an aberration-
free laser source) and homogeneously distributed over the opening of the diaphragm. Figure 48 A presents the intensity
distribution of the aberration-free and diffraction-limited spot. Even this ideal spot consists in reality of a central spot with full
width at half maximum (FWHM) diameter d:

The numerical aperture NA is given by the effective opening of the lens system (see Fig. 48 B). Around this central spot are,
even for the most optimal aberration-free situation, annuli of decreasing brightness. Any deviation from the ideal situation will
lead to larger spot diameters as well as to more intensities in the annuli. In practice, the resolution depends not merely on the
basic parameter /NA, but also on the quality of the laser, on the optical system — including homogeneity of the distribution
of the laser light over the limiting aperture, on lens quality, on alignment, and on system stability.

Figure 48. Intensity distribution of an ideal laser spot

A) Intensity distribution; B) Effective opening of the lens system and numerical aperture

page 30 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

In the conventional optical-disc mastering process, many disc manufacturers have been using laser beam recorders (LBR),
which utilize ultraviolet (UV) or deep-ultraviolet (deep-UV) lasers as light source. In April 2004 Pioneer presented a high-
precision electron beam recorder (EBR), that employs an electron beam as a recording beam to sharply narrow the beam
diameter. This can realize even finer pattern processing in the mastering process than LBRs. The high-precision EBR also
achieves high levels of record-positioning accuracy, thanks to the high-precision recording-position control technology.
Research and development for the EBR started with the production of laser discs in 1993. With EBR master discs for high-
density optical discs including Blu-ray discs can be manufactured, as well as discrete track media and patterned media -
higher-density hard disks regarded as highly promising future technologies.

The diameter of the spot size is the first parameter which controls the resolution of the final master recording (see Section d-
2.1.4-d-2.1.7).

Further optical requirements and characteristics of the mastering process have been described in detail [30], [65], [67], [44],
[66], [68-71]. An atomic force microscopic (AFM) picture of a DVD master is shown in Figure 49.

Figure 49. AFM picture of a DVD master.

Photoresist Master and Stamper Making. The used master disc is a circular, flat, polished glass substrate that has been
spin-coated with a layer (130 – 150 nm thick) of a conventional positive working photoresist. Before spincoating, the
substrate is usually treated with an adhesion promoting layer. The information is written into the resist layer in the air-incident
mode using a laser beam master recorder. After writing is complete, the exposed areas of the photoresist layer are
selectively removed by developing the disc in an aqueous alkali solution. In the optically exposed areas, a photochemical
reaction occurs such that the rate of dissolution in the developer is increased by several orders of magnitude. A description of
the novolac – o-naphthoquinone diazide photoresist system used for mastering is given in [60]. This type of photoresist is
also widely used in making integrated circuits ( Imaging Technology – Printed Circuits (Printed Circuit Boards)).

Development results in selective removal of the photoresist material from the exposed areas. The master disc is then rinsed
with water, dried, and coated with a metal layer of silver or nickel by means of wet chemical or vacuum technologies. The
master disc can now be read in reflection, so that its optical quality can be assessed. Next this master disc is used for
stamper production (see Section Electroforming of Stampers).

Direct Stamper Recording. In 1997 a modified photoresist technology was introduced for direct stamper making. This
process uses a negatively working photoresist in which the laser irradiated areas remain after the development process. The
negative photoresist is spin-coated on a metal substrate with the thickness of a conventional nickel stamper. After recording
and development the photoresist stamper is treated to harden the soft photoresist structures to withstand the high
temperatures in the molding. The process enables shorter cycle times to produce a stamper by skipping electroforming.

Ablative Dye Mastering. Instead of using conventional, chemically developable, photoresist materials, the mastering
process can also be carried out by direct laser writing in a thin layer of a degradable polymer such as nitrocellulose or
acetocellulose. The polymer is again deposited by spin-coating from a solution that also contains a dye. The laser light is
absorbed by the dye, converted into heat, and the polymer decomposes locally into volatile products.

This approach does not require chemical development. The ablative writing mode can, however, lead to debris and a pit rim
that increase the noise levels of the master and the replica. An additional feature and requirement for this approach is the
possibility of direct-read-after-write, which offers immediate optical quality assessment by read-out during writing and thereby
write-power control.

Photolithography with Dry Ion Etching. A modified photoresist mastering technique uses the exposed photoresist layer to
serve as a dry ion etch mask. lndium oxide, silicon oxide, or metals such as chromium are sputtered or otherwise deposited
on a glass substrate to the correct thickness (near /4n at read-out). The photoresist layer is spin-coated onto this sputtered
layer, exposed, and then chemically developed and dissolved. Reactive ion or reactive ion beam dry etching is used to
selectively etch only the oxide or metal layer. The remaining unexposed photoresist layer is then removed by dissolution,
leaving the desired relief structure of oxide or metal. This technique requires three more steps than normal photoresist
mastering, so that the latter is generally the preferred mastering technique.

page 31 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Direct Metal Mastering. Instead of laser writing in a photoresist layer, direct writing in a very thin (20 – 30 nm) metal coating
on a glass substrate is also possible. The heat generated by the laser beam evaporates the very thin metal layer. This
procedure also offers the advantages of no chemical development and immediate optical quality assessment by direct read-
out during writing. This approach has the disadvantage of creating debris and possibly a rim on the pit or hole during laser
writing leading to an increased noise in the master. Thus, these masters cannot be used to provide electroformed stampers
for molding and embossing based replication without complex steps to overcome this problem.

2.2.4. Electroforming of Stampers

Replicas and stampers are produced from the master by electroforming, also called electroplating, plating, or galvanics
(artref Electrochemical and Chemical Deposition a09_127). The photoactive layer of the glass master is first covered with a
thin (50 – 100 nm) metallization layer of nickel or nickel-vanadium. The end product leaving mastering is most widely referred
to as metallized glass master. From the metallized glass master one nickel metal copy (a 300 µm stamper father) is
generated by electrolysis. Cleaning, coating, polishing, and punching finish the stamper to fit the mold in the molding
process.

Basically electroforming has remained the same process since the old vinyl discs were produced, with certain refinements,
but it is still a highly manual process. The example of a typical stamper specification below gives a rough idea as to which
extent the refinements took place.

Figure 50. Stamper specification

1. Thickness T = 300 µm (+0/–10 µm)

2. Inner diameter Di = 34 mm (+0.010/–0 mm)
3. Outer diameter DO = 138 mm (+0.5/–0 mm)
4. Absence of burrs
5. Absence of pinholes and nodules
6. Back roughness RZ / RY = 0.5 – 2.5 µm
7. Low material stress /deformation

The specifications in Fig 50, together with information on how they are mainly influenced by electroforming parameters are
discussed in detail in the following sections.

Basic Model of Electroforming. The basic functionality of the electroforming process (shown in Fig. 51) can be summarized
as the transportation of nickel ions from the anode basket to the object at the cathode, on which they are deposited. The
medium for this transportation is the aqueous solution of nickel sulfamate. Controlling the electrical field and the current over
time allows to control thickness, respectively thickness variation, of the galvanoform, that is to be copied in the family process
or reinforced in the father process, respectively.

Figure 51. Basic electrolyte / electroforming model

Following the principle, that all ions in the solution are free motion-independent charge carriers, it is easy to picture the force
of the electrical field driving the ions. If the solution is well balanced in which respect? , then practically the only
conversions taking place are the two desired main reactions at the electrodes:

Anode reaction Cathode reaction

Ni Ni2+ + 2e– Ni2+ + 2e– Ni

At the cathode the positively charged nickel ions are discharged and the Ni atoms are built into the crystal lattice of the
already deposited nickel, building up the metal layer that forms the stamper. Simultaneously the negatively charged ions
(SO3NH2–, H2BO3–, Cl–) of the solution assist in the process of dissolving the solid nickel pellets. Thus the nickel metal of
the anode is brought to the state of freely moving ions, replenishing the nickel ions consumed at the cathode. The amount of
nickel precipitated on the cathode is governed by Faraday's law, that can be summarized in the electrochemical equivalent of
nickel: 1.095 g of nickel is precipitated per ampere hour (Ah) of current. For practical application a rule of thumb can be used:
For a 240 mm diameter glass master 1 Ah ≈ 3 µm of average thickness. A typical absolute working amperage for a 240 mm
glass master may be 80 – 120 A d.c. at 15 – 20 V.

Electrolyte Recipe and Process Parameters. Typically most electrolyte recipes in practical CD electroforming can be found
in the range of parameters listed in Table 8.

Table 8. Typical process parameters

page 32 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Nickel sulfamate c[Ni(SO3NH2)2] 300–450 g/L

Nickel chloride c [NiCl26H2O] 0 – 20 g/L

Boric acid c[H3BO3] 30 – 70 g/L
pH -Value 3.5 – 4.5
Temperature 45 – 65 °C
Anode material Sulfur-activated pellets (Ni-S-pellets)
Additives: generally as little as possible
Wetting agents
Brightening agents
Anode activators
Current densities 15 – 30 A/dm2

As the complexity of the recipe already indicates, the desired reactions are not the only reactions to take place, especially at
the desired deposition rates. One reason for this is that water also influences the process. Additionally all negatively charged
ions move towards the anode and all positively charged ions move to the cathode. This leads to undesired codepositions and
secondary reactions. By using a well-balanced nickel sulfamate electrolyte, these side effects can be reduced to a level, that
fulfills the specifications of the stamper. This requires some consideration with regard to the recipe as well as to the design of
the unit used for electroforming.

Recipe Considerations. Concentration of nickel sulfamate (Ni2+): Generally a higher concentration of nickel leads to a higher
availability of charge carriers that increase the conductivity of the solution, leading to a reduction of the anode cathode
voltage at a given working amperage. At concentrations tending toward the top limit of the range specified above, a
saturation peak is reached where no further gain in conductivity is possible. Similar to the effect on conductivity, the internal
material tensions (material stress) pass through a minimum at a certain, relatively high, concentration.

Concentration of nickel chloride (Cl–): The contribution of nickel chloride to the nickel concentration can generally be
neglected. The active component is the chloride content. The function of Cl– is to ensure good and constant anode corrosion
and reduce the risk of anode passivation at high current densities. One effect of anode passivation is depletion of (nickel)
ions in the electrolyte. Anode passivation must also be avoided, in order not to generate low pH values which would support
the hydrolysis of sulfamate ions to sulfate and ammonia. Both sulfate and ammonia lead to elevated levels of material stress,
even at low concentrations (≤ 1 g/L). It has become common knowledge that by using sulfur-activated nickel pellets the use
of nickel chloride can be avoided, provided certain precautions are taken. This is of considerable interest, because an
increased concentration of chloride ions in the electrolyte inevitably leads to proportionally increased levels of material stress.

Concentration of boric acid: Boric acid mainly serves as a pH buffer in the electrolyte solution. A minimum concentration is
required to ensure high-speed deposition of nickel at the cathode. Without boric acid during the process a film with an
extremely high value pH could be formed on the nickel surface, binding nickel as its hydroxide. This film would lead to
increased anode cathode voltages and in the worst case to blackened (burnt) nickel depositions.

pH Value: The pH value of the electrolyte solution has to be kept between 3,5 – 4.5. A pH value in this range has only little
effect on the quality of the metalwork. However, it is good manufacturing practice to keep the pH in rather tighter tolerances
than specified above. Extremely low pH values of 2.0 – 2.5 have to be avoided because they support the hydrolysis of the
sulfamate ion. As the H3O+-ions are mainly responsible for codeposition of hydrogen at the cathode, a well balanced pH
value is essential to avoid increased pitting or pinholes.

Temperature: Usually a somewhat higher temperature is desirable, as it leads to an increased mobility of the ions in the
solution. This increases the conductivity of the solution, reducing the anode - cathode voltage at a given working amperage.
Unfortunately, extremely high temperatures will also support the hydrolysis (decomposition) of the sulfamate ions. The
process of electroforming generates a considerable amount of heat. The actual temperature of the electrolyte between anode
and cathode will generally be higher than in the main tank and makes an appropriate cooling system imperative in the
system. Most modern electroforming systems use a titanium tube solenoid for this purpose, externally supplied with cooling
water. Tank temperature in nickel sulfamate in 60°C (± 2°C). Temperatures higher than 70°C lead to bad results.

Anode material: In CD/DVD electroforming so-called Ni-S -pellets are most widely used. Basically these pellets consist of
pure nickel metal with a small addition (≈ 0.2%) of sulfur. The sulfur content reduces the electrochemical potential of nickel,
supporting a stable anode corrosion, even at very low chloride concentrations or with no chloride at all in the electrolyte. As a
side effect of this sulfur addition a small amount of anode sludge (small black granules) is left over when the pellets are
consumed. This sludge is one of the major reasons for the necessity of good filtration to avoid nodules.

Additives: In CD/DVD electroforming the general rule applies that as few additives as possible should be used. Nearly all
additives are organic chemicals influencing the nickel electroforming process with respect to pitting (wetting agents), back
roughness (brightening agents) and anode activity. Unfortunately, most of the additives have unavoidable side effects
although a range of modern agents exist, that have only minor negative effects. But like all organic chemicals used as
additives, their utilization leaves organic degradation products, that contaminate the solution. The contamination increases
the need for active carbon treatment. Generally organic contamination may be a major source for process problems such as
increased material stress.

page 33 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Current densities: In electroforming the key figure for the rate of nickel deposition is the current density applied on the surface
of the cathode area. A working amperage of 100 A on a 240 mm diameter glass master for example results in a theoretical
current density of approximately 22.1 A/dm2. Apart from the limitations imposed on the process from the material side (see
above), the current density cannot be increased at will in order to reduce cycle times. Material stress is increased
proportionally to increasing current densities. As a result a compromise has to be found between acceptable material stress
and shortening of cycle times in electroforming. Unfortunately, no standardized method has yet been implemented to
measure material stress quantitatively at current densities and deposition thicknesses typical for CD electroforming. The fact
that it is often impossible to distinguish mechanical deformation and material stress in a stamper is the main reason for the
absence of such a measuring standard. Typical current programs feature a starting ramp (especially necessary for a glass
master's thin initial metal layer), leveling off to the working current. Modern control systems allow easy programming and
control of the ramping as well as direct ampere-hour setting to program the average thickness.

Schematic Layout of an Electroforming System. In Figure 52. the schematic layout of an electroforming system is
depicted. The reproducibility of metalwork with tightly controlled thickness variations requires a stable geometry of the
electrical field between anode and cathode. As mentioned before, the average thickness of an electroformed piece of
metalwork can be controlled by achieving well controlled ampere-hour values. Thickness variations over the surface of the
working piece require a suitably designed geometry of the electrical field between anode and cathode.

Figure 52. Schematic electroforming system (Source: technotrans AG)

Rotating spindles and control of the electrical field: Achieving a perfectly homogeneous electrical field should result in the
required thickness tolerances. But as the simple model of a plate condenser (Fig. 53) easily visualizes, in practice this is
virtually impossible to achieve.

Figure 53. Plate condenser model of a homogeneous electrical field

Not only is the field inhomogeneous at the edges of the conductive areas, but also the necessity of having anode and
cathode both perfectly parallel and flat can hardly be achieved in reality. The need of being able to install and remove anode
basket, mounting plate, and for example the glass master makes certain tolerances unavoidable. The consequence for the
electroforming process is that a relatively strong electrical field will lead to relatively higher deposition rates of nickel. In most
systems used in CD/DVD electroforming a rotating spindle at the cathode helps to generate a concentrically homogeneous
deposition of nickel. The evenness of nickel deposition can then be relatively easily controlled by adequate shielding. A
positive side effect of rotation is, that codeposited hydrogen will be washed away mechanically, reducing pitting. A total
thickness variation of ≤± 3 µm for a stamper father will be guaranteed by manufacturers of good quality electroforming
systems.

Flow control of electrolyte: For thermal reasons the electrolyte in a process cell must be exchanged at relatively high rates, to
ensure stable working temperatures of the electrolyte. Injecting conditioned electrolyte between anode and cathode is
typically the solution chosen in most electroforming systems. The design of the flow needs considerable care. Unwanted
thickness variations may be the result if the flow rate or its course are not designed and controlled adequately. A good design
typically allows to monitor and adjust flow rates as needed. Additionally a good flow design supports mechanical removal of
codeposited hydrogen (reducing the need for wetting agents as used for antipitting) and does not affect thickness variations
when the flow rate is changed.

Anode basket and diaphragm: Anode baskets are typically made from titanium, ensuring a flat anode surface for the
electrical field, while facilitating good and stable anode corrosion. Titanium is used, as its oxide layer is chemically inert and
will not allow corrosion of metallic titanium for anode - cathode voltages up to 20 – 22 V. To avoid the codeposition of anode
sludge on the cathode (nodules) the anode has to be clearly separated from the cathode. Preferably a diaphragm should be
used for this purpose combined with a well designed electrolyte flow. This minimizes the risk of sludge reaching the cathode.
Historically so-called anode bags made from fabric have been used to contain the anode sludge around the anode basket.

Filtration: High-quality pump filtration systems, made from similar materials as the process cells, are required to avoid
carrying over particles from the tank into the process cells. These particles could be nuclei that trigger the formation of
nodules. Preferably a pre- and a fine filtering stage, based on filter cartridge technology, should be included in the system.
This also facilitates active carbon treatment with cartridges, whenever an organic contamination is be suspected.

A schematic representation of the process is depicted in Figure 54

Figure 54. Schematic representation of the production of replicas and stampers for optical discs

A) Build-up of master

page 34 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Glass disc (G); Adhesive layer (A); Photoresist (P); Silver (S)

B) Master plated with nickel (N)

C) Nickel negative or father stamper after removal of photoresist, adhesive layer and silver

D) Nickel positive or mother stamper

E) Nickel negative or son production stamper

In the process of separating the father stamper from the master, the relatively soft photoresist master is destroyed. When a
silver layer is used as a metal coating, the silver has to be removed too. The nickel father stamper is replicated several times
by electroplating to obtain positive mother nickel stampers (for small series of discs the father stamper can be used directly
for disc replication). From the mother stampers, negative son nickel production stampers are replicated by electroplating. In
each of the nickel-to-nickel replication steps, a separation layer is required. Rinsing with a dilute aqueous solution of
potassium dichromate or an anodic treatment to form an oxide layer are typically used. Many negative production son nickel
stampers can be made by this family method from a single photoresist master disc.

The electroplating process and nickel solution used are quite similar to those used for gramophone record production. The
tolerance demands on the thickness of the nickel metal layer, the flatness of the metal backing, and the cleanliness of the
nickel solution are much more stringent for stampers used to produce optical discs.

Polymer Stampers. Polymer production stampers can also be made by photoreplication from a mother nickel stamper or
from an original photoresist master disc. A liquid, photopolymerizable coating is squeezed between a flexible plastic
substrate and the stamper or master, followed by photopolymerization of the coating. This process is similar to a disc
replication production process (2p or photopolymerization). For use with the typical elevated temperatures occurring in
production molding processes, the photopolymerized stamper is coated with a metal such as chromium or aluminum.

2.3. Read-Only Optical Recording Materials

2.3.1. Production Processes
Numerous processes have been investigated for the replication of optical discs. At present photopolymerization (for digital
optical recording, pregroove structures), injection molding (for video discs such as Laser Vision (LV), Compact Disc audio,
and DOR pregroove structures), and injection – compression molding (for CD audio) are among the most important. These
processes are shown schematically in Figure 55 and will be discussed in detail in Sections Molding Processes and
Photoreplication Process.

Figure 55. Production of optical discs

A) Photopolymerization

B) Injection molding

C) Injection-compression molding

a) Mold with stamper; b) Photopolymerizable coating; c) Transparent plastic substrate; d) Injection – compression molding
material; e) Mirror block; f) Injection molding material [72]

The choice of a particular production process and disc material depends mainly on economic factors, i.e., the choice will
depend on the state-of-the-art of the processes considered and on the availability of suitable materials.

The conventional production process for gramophone records is compression molding of a vinyl chloride – vinyl acetate
copolymer. This process is unsuitable for making optical discs, because of the high birefringence of the discs, the local
noncircularity of the tracks introduced on cooling, and problems with stamper contamination. The absence of birefringence is
of particular importance because anisotropy of the refractive index of the substrate will change the plane (linearly) polarized
laser beam used for read-out into an elliptically polarized one, which may cause severe disturbance of the read-out process.

Historically, photopolymerization (Fig. 55 A) was the first process used with LV discs because neither injection molding nor
compression molding could be carried out with sufficient accuracy in the 1970s for making 300 mm diameter discs. In the late
1970s and early 1980s, injection molding of poly(methyl methacrylate) (PMMA) became feasible. PMMA injection molding
has replaced photopolymerization, notably in mass production, for reasons of economy and simplicity (Fig. 55 B).

In the meantime, the 120 mm diameter CD was developed. CDs can be produced by injection molding and by injection –
compression molding of polycarbonate (PC), because of the development of special grades of PC that allow the
birefringence problem to be overcome (Fig. 55 C).

Compared to PMMA, PC has definite advantages with respect to moisture uptake and therefore dimensional stability. This is
of special importance with present Compact Discs because they are single-sided. Uptake of water through the nonmetallized
side of a PMMA disc leads to severe warpage.

With the double-sided video disc, water uptake is a smaller problem because of the symmetrical, sandwich-like structure of
the disc. These discs can also be made from a blend of PMMA and poly(vinyl chloride) (PVC) [73]. DOR discs can have a
glass or polymer substrate. For glass substrates, the preformatting (i.e., the provision of a pregroove and optionally clocking

page 35 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
and address information) is performed by means of the photopolymerization process (see Section Photoreplication Process).
For thermoplastic substrates, injection molding is used to make the substrate and the preformatting information in one step.
So-called mini-DOR substrates with 130 mm diameter are routinely made by injection molding of PC. For thermosetting
materials, similar single-step processes have been described [74]. A possible alternative is the use of photosetting materials
such as acrylates [75], [76].

2.3.2. Material Specifications

The specifications for discs and players are matched. The disc must therefore satisfy a number of requirements relating to
optics, geometry, and stability. This will be illustrated for video (LV) and audio discs (CD).

Optical Properties. At least 70 % optical reflection of the substrate-incident laser beam is required for proper read-out of the
stored information. Absorption and scattering losses that occur as the laser beam passes through the substrate twice depend
strongly on the type and purity of the material; the same applies to the reflection at the metal – polymer interface. Only a
limited number of combinations of substrate and mirror or reflective coating (and replication coating in the case of
photopolymerization) materials satisfy the requirements for high reflectance. The substrate used with photopolymerization
must also be transparent in the near UV (320 – 380 nm). Turbidity, microvoids, and dust particles must be absent, and the
substrate must be optically homogeneous.

Another optical requirement is connected with the use of linearly polarized laser light in the optical system of the player for
separating the incident from the reflected beam. Any change in polarization makes this separation less effective and causes
a drop in detection efficiency and an undesired feedback of modulated light to the laser. Birefringence invariably changes the
plane polarization of the incident laser beam into elliptical polarization.

For the video disc the birefringence must be < 20 nm for a single pass of a parallel beam of light. This imposes a serious
restriction on the choice of substrate material and the method of production. For CD discs, the requirements are less severe:
the birefringence must not exceed 50 nm for a single pass of a parallel beam of light.

Geometrical Properties. Specifications for discs and players are such as to ensure that any disc can be played on any
player.

The thickness of the video disc and the CD have been set to 1.25 ± 0.10 mm and 1.20 ± 0.10 mm, respectively. Much thinner
substrates are difficult to handle and to produce and are usually not stiff enough. Moreover, at a thickness of ca. 1 mm,
scratches or dust on the outer surface of the disc are so far outside the depth of focus of the objective lens of the player that
only tolerable signal degradation results.

The tolerances for flatness and parallelism are more severe than those for thickness. Any departure from planarity of the
inner surface of the disc causes vertical movement of the information track. The high rotational speed of the video disc (25 or
30 revolutions per second for PAL or NTSC formats respectively) and the restricted acceleration, velocity, and amplitude of
the focus control system impose restrictions on the tolerable local thickness variations. Tolerances depend on the radial
position as well as on the wavelength of the disturbance. If a sinusoidal disturbance is assumed along the track of a video
disc, it can be shown that, depending on the radial position on the disc, disturbances with a wavelength of 8 – 22 mm may
cause vertical amplitude deviations no greater than 2 µm and still be tolerated. With increasing wavelength of the
disturbance, the vertical tolerance also increases up to ca. 1000 µm. The tolerances with respect to radial displacements are
smaller. Radial amplitude deviations of 0.1 µm can be tolerated for wavelengths of 4 mm at the inner radius of the active area
of the disc and for wavelengths of 10 mm at the outer radius. If the wavelength exceeds 100 mm at the inner radius or
280 mm at the outer radius, radial amplitude deviations as large as 80 µm can be tolerated. The influence of radial and
tangential variations and tolerances have been discussed in detail [77].

Stability. The optical and geometrical characteristics should remain unaltered under climatic conditions likely to be
encountered during transportation, storage, and use. Desirable properties are high scratch resistance, toughness, and
fracture strength. Dimensional stability under wide variations in temperature and relative humidity, a certain inertness with
respect to cleaning agents, and the absence of metal mirror corrosion are also prerequisites.

Service life is tested in cyclic humidity tests in which discs are subjected to periodic alterations of temperature (usually from
20 – 45 °C) and relative humidity (usually from 50 – 96 %) [77].

Materials. Few substrate materials satisfy all the requirements mentioned above. In Figure 56, five polymers and an
inorganic glass are compared qualitatively for those material properties that are most relevant to optical disc production and
performance [78]. The larger the radius in the polar diagrams, the more favorable the respective property of the substrate
material.

Figure 56. Qualitative comparison of substrate materials for optical discs. The larger the radius, the more favorable the
indicated property of the substrate material. The properties shown are:

n, birefringence; W, water uptake; O, permeability of oxygen and water; H, heat distortion temperature; C, cost; P, ease of
processing; E, tensile modulus; T, toughness [78]

For PC in particular, but also for other polymers, considerable work has been carried out to develop a suitable grade for
optical disc manufacturing. Reduction of birefringence has required adaptation of the material and the processing used. The

page 36 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
properties of various substrate materials have been reviewed [78-81]. At present, PC and PMMA are the most widely used
polymers. Properties of discs made from these materials are listed in Table 9 [79], [80].

Table 9. Material requirements and properties for video discs (PMMA) and Compact Discs (PC) [79]

Property Video disc CD, DVD

Optical
Light transmission, %
630 nm ≥90 ≥90
300 – 380 nm ≥80 a

Birefringence (optical path difference single pass b), nm ≤20 ≤50

Optical defects >200 µm none none

Mechanical
Thickness (nominal), mm 1.25 ± 0.1 1.20 ± 0.1

Thickness tolerance, µm ± 20 c ± 50 d
Distortion, degrees ± 0.4 ± 0.6

Surface roughness, nm ≤10 ≤15

Rockwell hardness 105 86

Impact strength, kJ/m2 12 no failure

Thermal
Vicat softening temperature, °C 96 – 107 145

Melt index, g/10 min 10 – 22 (230 °C) 30 (260 °C)

Physical-chemical
Water absorption (100 % R.H.), (wt %) 2.1 0.4

Water diffusion coefficient (23 °C), mm2/s 5 × 10–7 48×10–7

a For photopolymerization.
b Parallel light beam.
c Disc diameter 305 mm.
d Disc diameter 120 mm.

2.3.3. Molding Processes

For injection molding of optical discs, hot molten polymer is injected at high pressure into a thin cold mold cavity at a high
flow rate through the gate located at the center of the cavity (see Fig. 55 B). A clean room environment is required. The
molding process and the molding material have been optimized to attain the required birefringence for CDs of <50 nm (single
pass for a parallel beam of light).

The birefringence of a disc is caused by two effects [82], frozen-in thermal stress and frozen-in orientation. Thermal stress
occurs because of inhomogeneous cooling combined with a change of mechanical properties upon solidification. The amount
of thermal stress is determined mainly by the difference between the mold temperature and the glass transition temperature
Tg of the polymer.

The orientation of the macromolecules, which is caused by the shear and elongational stresses in the melt during mold filling,
can be frozen-in by rapid cooling and vitrification. The amount of frozen-in orientation depends mainly on the injection
temperature of the melt.

A special problem with PC is the high stress optical coefficient (SOC), defined as the ratio of birefringence to stress. The
SOC of PC is more than ten times larger than that of PMMA. This causes the birefringence of PC substrates to be much
more sensitive to internal stress than the birefringence of PMMA substrates. Much effort is therefore spent in structural
modification of PC in order to reduce its intrinsic optical anisotropy [83], [84] and in blending materials with low stress optical
coefficients of opposite sign [85].

Careful analysis of the birefringence distribution in CDs made of PC has revealed that both the radial and the tangential
refractive indices vary in a complicated fashion throughout the thickness of the disc. The birefringence of a normally incident
beam is equal to the difference of these two components. Therefore, the birefringence also varies strongly through the
thickness of the disc (see Fig. 57) [86]. At the outer surface (z = 0.6), the sample is almost uniaxial; the birefringence nearly
vanishes. Just below the surface, the birefringence becomes strongly negative, then increases to a positive value and next
decreases to almost zero near the center plane of the disc. The observed overall birefringence is, therefore, determined by
the compensation of two large contributions.

page 37 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 57. Birefringence of a CD in the direction normal to the disc (Z), determined at a distance of 35 mm from the center
axis [86].

This explains why the observed birefringence is so sensitive to small variations in processing conditions. The art of
processing lies in finding the right balance between the injection or filling stage and the compression stage of the injection
molding process [86].

Important properties of the disc such as birefringence, residual stress (which causes birefringence), variation in thickness,
and replication of the pit pattern from the stamper are critically dependent on the flow behavior of the molten polymer. The
non-Newtonian dependence of the melt viscosity on the shear rate is therefore an important parameter. At the shear rates
commonly used in injection molding, the dependence of the viscosity of PC on shear-rate is much smaller than that, for
example, of PMMA. To obtain good moldability with PC, a relatively high temperature is required, but polymer degradation
sets an upper limit to the molding temperature [87].

Special low molecular mass grades of PC have been developed to extend the processing window to lower temperatures and
pressures while maintaining good flow properties [88-90]. The granulate polymer should contain neither gel nor dust
particles.

In addition to processing conditions and material properties, the molding equipment has also been optimized with respect to
product properties and cycle time [91], [92]. Without the increased precision, the injection molding of compact discs using PC
would not have been possible. The injection molding process has been described in detail [92], [93].

Injection – compression molding (see Fig. 55 C) is distinguished from injection molding in that a heated mold is filled by
injection of even hotter fluid polymer before complete closure of the mold occurs. Excess polymer is expelled from the mold
(flash). This is removed after cooling. Unlike injection molding, polymer dosage is not critical.

The cycle time for CD injection molding is ca. 10 s. For injection – compression molding, the cycle time is somewhat longer,
because the mold itself has to be heated up and then cooled before opening and removal of the disc. Since a heated mold is
used, however, considerably less polymer orientation is frozen-in. Process conditions and mold tolerance are, therefore, less
critical for obtaining acceptable birefringence than with injection molding. This is also reflected in the higher yield of the
injection – compression molding process. Metallization and protective coating of the discs are described in Sections
Metallization and Protective Coating.

2.3.4. Photoreplication Process

The photoreplication of LV video discs is shown in Figure 58 [77]. In the first step (Fig. 58 A), a liquid photosensitive coating
C is deposited on the metallic stamper or mold Mo. The slightly curved substrate S (PMMA, 1.25 mm thick) is pressed onto
the coated mold (Fig. 58 B). The latter forms a thin layer C (20 – 30 µm) between the substrate and the mold. The coating is
then irradiated with an array of fluorescent lamps ( = 350 nm) for 6 s. The substrate and the coating are then released from
the mold (Fig. 58 C), inverted, and a metallic reflective layer M is deposited on top of the organic coating. The reflective
mirror layer is covered in turn with an organic protective coating P (Fig. 58 D). Finally, the P-sides of two such discs are glued
together to yield a double-sided video disc. The principles of photopolymerization and the formation of densely cross-linked,
glassy networks have been described in detail [94].

Figure 58. The 2 p photopolymerization process [77]

C = Replication coating; Mo = mold; S = substrate; M = Mirror; P = protective coating

The photopolymerizable coating used for copying the information pattern must have a high curing rate, a low viscosity, and it
must wet the substrate surface. When cured, it must have high dimensional stability, durability, and lack of odor, easy release
from the mold, and good adhesion to the substrate and to the reflective layer.

Coatings that meet all these requirements are not easy to find, mainly because some of the requirements are conflicting, e.g.,
easy release from the metallic mold but good adhesion to the metallic reflective coating.

Acrylates are preferred monomers because of their high rate of polymerization. The most efficient photoinitiator is 2,2-
dimethoxy-2-phenyl acetophenone (DMPA). The absorption spectrum of DMPA strongly overlaps the emission spectrum of
readily available fluorescent lamps [95].

page 38 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The monomer coating's components were chosen with a view of desirable ultimate properties such as behavior upon
metallization and adhesion to the metallic reflective coating.

Correlating the monomer components' chemical functionality versus their content of inactive, saturated hydrocarbon groups
along with hardness, adhesive and release properties [95], [96] has made it possible to delineate a region on the overlapping
maps where potentially useful monomers and mixture compositions can be found. One single monomer coating and a few
mixtures have been selected for further investigation in this way (see Table 10).

Table 10. Video disc coating properties [95], [96]

Coating composition, Viscosity Curing time, s

wt % * 23 °C,
mPa·s 23 °C 80 °C

1 TPGDA NVP
57 % 29 % 7.8 1.5 1.0
TMPTA DMPA
10 % 4%

2 TPGDA NVP
61 % 17.5 % 12.3 1.7 1.0
TMPTA DMPA
17.5 % 4%

HDDA HDDA DMPA

96 % 4% 6.7 3.0 1.4

* DMPA = 2,2-dimethoxy-2-phenylacetophenone [24650-42-8]; HDDA = 1,6-hexanediol diacrylate [13048-33-4];

BDDA = 1,4-butanediol diacrylate; EHA = 2-ethylhexyl acrylate [103-11-7]; TPGDA = tripropyleneglycol diacrylate [
42978-66-5]; TMPTA = trimethylolpropane triacrylate [3290-92-4]; NVP = N-vinyl-2-pyrrolidone [88-12-0].

An important parameter of the coating is its hardness. The influence of chemical composition on hardness and dimensional
stability is shown in Figure 59 [96]. Measurement of coating hardness and its influence on signal-to-noise ratio have been
described [97]. Hardness is related to the cross-link density. Cross-link density is related to the extent of double bond
conversion. Unexpectedly, it turned out that at room temperature the highest cross-link densities can be obtained with the
diacrylates [95].

Figure 59. Scanning electron micrographs of the LaserVision video disc pit pattern after metal deposition on a coating of a
mixture of EHA and BDDA. The composition of coatings in pictures (A) and (B) is 80 % EHA – 20 % BDDA. After
metallization, the soft coating appears as a wave pattern, due to shear deformation caused by stresses in the aluminum
layer. Picture (C) has the composition 20 % EHA – 80 % BDDA. There is no wave pattern because the cured coating is more
densely cross-linked [95], [96]

After comparison of various parameters, 1,6-hexanediol diacrylate (HDDA) was selected as the single monomer coating. It
has been further studied together with the mixtures given in Table 10. Parameters studied included curing time under
process conditions, extractability of low molecular mass material, and the energy required for separation from the mold [95],
[96] (see Table 11).

Table 11. Separation of video discs from a mold (constant pulling rate of 8.33 mm/s)

Coating * Total Substrate Coating Separation Maximum

separation, deformation separation time, pulling
energy, energy, energy, force,
J J J s N

1 1.27 0.39 0.88 4.2 51

2 0.75 0.27 0.48 3.6 48
HDDA 0.30 0.08 0.22 2.4 31

* see Table 10.

During release of the substrate with the cured information layer from the stamper or mold, the disc can be warped in a

page 39 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
complicated way. Part of the total separation energy is required for the actual separation (release of adhesion), the remaining
part for the (in)elastic deformation of the substrate with the information-containing layer. Mold release is easiest for the
single-monomer system, although satisfactory results were obtained for all three mixtures (Table 11) [95], [96]. Video discs of
excellent quality have been made with both the single-monomer coating and with a number of mixtures.

2.3.5. Metallization
The metallic reflective coating should have a reflectance of at least 80 – 85 % (ca. 8 – 10 % is lost by scattering and by
reflection at the air – polymer interface) because at least 70 % of the laser light must be reflected by the disc. For this reason,
together with material cost, stability, and reactivity, the choice of material is limited to aluminum, silver, copper, gold, or
selected alloys of these metals [98]. The reflectances of the pure metals in the near-IR, where the discs are read, are 85 %,
99 %, 95 %, and 97 % respectively. The thickness of the reflective layers is about one hundred nanometers.

Aluminum can be deposited easily by thermal or electron-beam vacuum evaporation and, in contrast to silver, it adheres well
to the polymerized coatings, as well as to PC and PMMA. A disadvantage of this deposition technique is that it must be
carried out in batches, which makes it difficult to integrate into the otherwise continuous production process. Each batch
requires a relatively long time before the desired vacuum deposition pressure <10–3 Pa is reached, notably when PMMA is
used as a substrate. PMMA may contain 1 – 2 wt % water under ambient conditions [78], and this must be largely removed
before vacuum deposition may be started. If the deposition is started at an insufficiently low pressure, a mirror coating with
reduced reflectance is formed, owing to reaction of the growing layer with water vapor. The use of PMMA requires the design
of special evaporation equipment with large liquid nitrogen-cooled surfaces. The problems are much less severe with PC,
which absorbs much less water under ambient conditions.

For PMMA and PC, alternative techniques such as electroless silver deposition and magnetron sputtering have been
developed [77]. Electroless silver deposition is quite similar to the process used for many years in the production of glass
mirrors. The cured coating is simultaneously sprayed with an ammoniacal silver salt solution and a reducing agent. Adhesion
to the organic surface is poor, as with an evaporated silver coating. A special pretreatment is therefore necessary [99].

Magnetron sputtering uses a cold, solid target that is bombarded with ions, usually argon(I) ions. At an argon pressure of ca.
0.1 Pa and a sufficiently high d.c. voltage a discharge is formed between the target, acting as cathode, and an anode. With a
magnetic field, maintained by permanent magnets behind the cathode, a concentrated discharge plasma is produced
immediately above the target surface. Argon(I) ions with an energy of 300 – 500 eV can be extracted from the plasma; these
ions bombard the target surface and sputter its material.

The disc to be coated is situated directly opposite the target and outside the plasma region. The production rate is
determined mainly by the sputtering time, not by the time taken for loading and unloading, so that d.c. magnetron sputtering
of aluminum is also widely used in CD production. Sputtering also allows more freedom in the choice of the reflective mirror
material because binary and even ternary alloys can be deposited in this way. Some of these alloys have desirable
properties for optical disc applications, including high stability and low internal stress [98]. The layout of a magnetron sputter
is shown in Figure 60.

Figure 60. Schematic of a magnetron sputter

2.3.6. Protective Coating

The metallized side of the video disc or CD is spin-coated with a solution of a protective organic coating such as
nitrocellulose or with a UV-curable material. For Compact Discs, a label is applied by tampon printing to this side of the disc.
The discs are inspected, tested, and packed for shipment.

For video discs, the cured protective coating is sprayed with an acrylic-based contact adhesive. After being dried, two discs
are pressed together, care being taken that no inclusion of air occurs: this might introduce local variations in thickness, which
could produce unacceptably large accelerations that are difficult for the focus-servo system of the video disc player to handle
during playback. Moreover, deformation will also induce local stress, which may initiate degradation of the reflective metal
coating. If necessary, the video disc is mechanically balanced by removing some material from the perimeter of the
sandwiched double-sided disc. An adhesive and label are applied to each side and the discs are inspected, tested, and
packed for shipment.

2.3.7. Nonstandard Processes

Photopolymerization of Complete Discs. Instead of using a thin photopolymerizable coating, it is also possible to cast the
complete disc by photopolymerization of liquid monomers without using a substrate. In this case, provisions must be made to
cope with shrinkage during polymerization. In one design, the liquid monomer is confined between a stamper or mold and a
glass plate and irradiated through the glass. The glass plate is spring-loaded so that it can move towards the mold during
polymerization and attendant shrinkage [75]. Another invention used a construction that allows for replenishment of liquid
monomer during the polymerization and shrinkage [76]. The use of thermosetting polymers has also been proposed [74].

Embossing or Printing. A heated mold is pressed against the surface of a preformed polymer sheet, held in a press, for
long enough to transfer the encoded information without deforming the sheet. Nonplasticized calendered PVC sheet and
extruded PC film have been used [100-103]. A continuous web manufacturing process is possible. In addition to high
throughput, this allows the use of a relatively thin and soft sheet or foil, which is then metallized and laminated to a stiff thick
substrate such as PC. A further variation is the use of metallized polymer films [104-106], even directly in an injection-
molding machine [107].

page 40 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Photographic Reproduction.The master disc is converted into a photographic exposure or imaging mask by deposition of
chromium on the developed areas, followed by removal by dissolution of the photoresist. A contact copy can then be made
using a polymer substrate disc coated with a high-resolution photographic emulsion. Systems based on physical
development (PD) are particularly suitable [108]. The disc with the contact image can be read in transmission or,
alternatively, in reflection after coating with a reflective metal layer [109]. In another method, a reflective image is made
directly during the PD process [110]. Although good discs have been obtained in this last way, the whole process of contact-
copying suffers from sensitivity to dust and contamination of the imaging mask. It is also very complicated to create the
imaging mask and the process requires the use of extremely smooth and flat polymer substrates. A noncontact holographic
replication of CDs has been shown on photosensitive layers [111].

Photopolymer Reproduction.A photopolymer process, which also requires that the master disc be converted into an
imaging mask, with the contact copy steps and difficulties as in photographic reproduction, is also possible. The contact copy
is made using a polymer substrate disc coated with a high resolution photopolymer. The optical read-out signal is due to a
spatial variation in the refractive index.

Photopolymers are polymers in which optical density and refractive index change (typically n < 10–4) under the influence of
light, mainly as a result of photopolymerization of dissolved monomers [112], [113], [114].

The most suitable materials are polyacrylates, polyacrylamides, polystryrenes, polycarbonates and their copolymers, mixed
with various monomers, photoinitiators and photosensitizers [80]. PMMA is the preferred matrix polymer. Pattern-wise
irradiation causes a local depletion of unreacted monomer, which is replenished by diffusion from the unexposed areas. Next,
a flood exposure is applied in order to fix the refractive index pattern induced by the diffusion of the monomer [112], [113].
Especially large local changes in refractive index ( n > 10–2) have been obtained in styrene-doped PMMA, exposed to UV
radiation ( > 280 nm) [115], [116]. Green light ( = 514 nm) can also be used for optical recording in a layer where
titanocene chloride is the photoinitiator [116]. In addition to refractive index changes, photomechanical materials exist in
which relief structures can be attained with a spatial resolution <1 µm. This is possible by the local photopolymerization of
styrene, dissolved in PMMA, followed by removal of the unreacted monomer [115-117]. As mentioned earlier, PMMA is not
an ideal substrate for single-sided optical discs, primarily due to its high water uptake [78].

2.4. Write-Once Optical Recording Materials

Write-once recording mechanisms can be divided into those that involve macroscopic material transport and those that do
not. These mechanisms are illustrated in Figure 61. All these mechanisms are of the heat-mode category, that is, being
induced by the heat generated from the optical absorption spectra of the recording materials.

Figure 61. Write-once optical recording mechanisms

A) Hole burning (ablation); B) Bubble formation; C) Texture change; D) Alloying of bilayers; E) Island agglomeration;
F) Phase change is this figure still relevant?

2.4.1. Material Specifications

Write-once optical discs are generally used for the archival storage of information. A prime prerequisite, therefore, is
chemical and physical stability of the recording media and the written information. It is not sufficient that the substrate and the
films themselves are stable because interfacial phenomena often have a very detrimental effect on media stability.
Furthermore flat, homogeneous discs are required, without pinholes, cracks, or blisters. The disc must also have sufficient
reflectivity at the wavelength of the laser used, in order to focus and track the pregroove structure and to read out the stored
information.

For optical recording the laser must produce permanent optically-detectable changes in the sensitive layer of the discs.
These changes should be formed with the pulse powers and pulse durations available from a semiconductor diode laser.
Typically, a pulse duration of 20 – 100 ns and a power up to 10 mW are used. A temperature of the order of 1000 – 1500 °C
can thereby be attained [118-123].

A large difference in reflectivity is required between marked (written) and unmarked (unwritten) areas in order to obtain a
large read-out signal. A high signal-to-noise ratio is desired which implies both a large reflection change and a low noise
level. Contributions to the noise include irregularity in the size and position of the bits (write noise) and the microstructure of
the film (disc noise). The bit error rate or probability is the ratio of bits erroneously detected to the number of written bits. For
a practical storage system, the uncorrected or raw bit error rate is more important than the noise. Usually, a value of <10–4 is
desired [124].

Read-out is usually performed with the same laser as used for writing, but at a much lower power level. The heating caused
by the reading laser should not induce changes in the film or data, implying that mark formation must have a threshold
temperature.

To attain high sensitivity, very thin, strongly absorbing films are primarily used as active layers. The complex refractive index
of a material is given by the equation:

page 41 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
where n is the refractive index and k is the absorption index [125]. Then, the absorption length, , is related to k by:

The generation of heat will usually be most effective for a layer with a thickness less than . With thicker layers, deeper
portions receive less light and are hardly heated.

In the optical design of a disc, interference effects must be taken into account to calculate optimum layer thicknesses [125],
[126]. A compromise often has to be found between a high reflectivity and a low absorption. The dependence of reflectivity
and transmission on layer thickness for an organic dye is shown in Figure 62.

Figure 62. Dependence of reflectance (substrate-incident) and transmittance of a near-IR dye ( = 800 nm) on layer
thickness [119].

The reflectivity, absorption, or contrast can be optically tuned by using multilayer configurations; this is more complex, and
therefore expensive [126]. This concept is illustrated in Figure 63. A disadvantage of such constructions is that they are
relatively difficult to produce because of the often very narrow tolerances on the thicknesses of the layers. Extra interfacial
problems can also arise. In some applications, however, (see Section Magneto-Optic Materials), such structures must be
used to enhance the read-out signals to the necessary minimum level.

Figure 63. Principle of the anti-reflective trilayer structure [126].

2.4.2. Deposition Techniques

The thin films used in optical recording are generally metals, semiconductors, dielectrics, and organic materials. Inorganic
films are deposited by vacuum techniques such as thermal, flash- and electron-beam evaporation [127], and d.c. or r.f.
sputtering [128]. The evaporation techniques are all performed under high vacuum. The differences are mainly the methods
used to vaporize the materials.

Alloys are difficult to deposit by thermal evaporation because of the different vapor pressures of the components; sputtering
is the preferred technique. A further advantage of sputtering over thermal evaporation is the ability to deposit many materials
with a high accuracy in composition. Moreover, this technique is well suited for mass-production.

Organic layers are usually applied by means of spin solvent coating [129]. A solution of the organic material is deposited on a
slowly rotating disc, which is then rotated at a high speed. The solvent evaporates and the film is thereby formed.

2.4.3. Tellurium Alloys

The conceptually simplest optical recording mechanism is heating the film with the focused laser beam and melting the film
locally, followed by ablation of the material. The advantages of this mechanism are that the contrast is large and the process
is self-regulating. Ablation is the most widely used write-once optical recording mechanism. The most important write-once
optical recording materials are alloys consisting mainly of tellurium.

Tellurium has high absorption and reflection at 800 nm, a low melting point [124], and an acceptable chemical stability, at
least when it is alloyed with other elements such as selenium. Materials used include Te – Se – Sb – S [121], [124], [130],
[131] and Pb – Se – Te alloys [132]. Some indications of the selection procedure for write-once materials will be given by
treating the tellurium alloys in detail.

If laser heating is to be efficient, the layer must be as thin as possible [124]. At a layer thickness <20 nm, however, the films
become inhomogeneous, prone to oxidation, and the reflectivity becomes very low [127]. Pure tellurium films are easily
oxidized, resulting in a rapid decrease in absorption and reflection. The oxidative stability of tellurium, as extrapolated from a
thermal cycling test in a high R.H. environment, the so-called Z/AD test, can be increased by at least four orders of
magnitude by adding selenium (see Fig. 64) [130]. In this test, the films are cycled between 25 and 65 °C at 93 % relative
humidity [133]. Adding more than 15 atom % selenium results in decreasing absorption at wavelengths in the near-IR with
hardly any increase in corrosion resistance.

Figure 64. Transmittance (750 nm) of Te – Se alloys as a function of time in the Z/AD test [130].

page 42 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
The films are usually amorphous when deposited and crystallize rapidly at ambient or slightly elevated temperature [130].
The accompanying contraction is appreciable, resulting in stress and sometimes even cracks in the film. Adding antimony
has the advantage of inhibiting the crystallization (which is still rapid at ca. 90 °C) and simultaneously reducing the volume
contraction. The addition of 3 atom % antimony appears to be sufficient to prevent the above problems.

The advantage of antimony over other elements such as indium, thallium, germanium, tin, lead, arsenic, or bismuth, which
also reduce stress, is that antimony inhibits corrosion, gives more well-defined pits, and does not form volatile products at
90 °C [130]. If more than 5 atom % antimony is added, then the crystallization temperature becomes excessively high. With
3 – 5 atom % antimony and preferably 1 – 2 atom % sulfur, the activation energy for the amorphous-to-polycrystalline
transition is such that a reasonably rapid transition occurs at 90 °C (typically within 15 min). The optimum composition
appears to be TexSeySbqSz, with 0.81 < x < 0.85, 0.01 < y < 0.13, 0.03 < q < 0.04, and 0.01 < z < 0.02.

Pit formation is rather complicated [121], [122], [134], [135]. It depends on alloy composition, film thickness, and also on the
substrate surface [134], [135]. The pits usually consist of a hole surrounded by a rim containing most of the removed material
with some pit material ejected as debris. A prerequisite (though not a sufficient condition) for pit formation is that the layer is
melted locally by the laser pulse [122], [124]. For long laser pulses (1 s), the temperature required for pit formation appears
to be only slightly above the melting point. For short laser pulses, a temperature close to the boiling point is required [122].
The influence of the substrate is also extremely important. Apart from the thermal properties of the substrate, the interfacial
energy between the substrate and the film is important. Lowering the interfacial energy by using fluorinated polymers in the
substrate lowers the threshold energy, but the stability of the film also decreases appreciably [134], [135].

Optimization leads to discs with high carrier-to-noise ratio (CNR) values, for pulse energies of the order of 0.6 nJ. The
stability (as determined from the Z/AD test) of hermetically sealed, air-sandwich glass disc constructions is at least 20 years.
The air-sandwich prevents moisture from entering the 2 p lacquer, which could cause leaching of sodium from the glass into
the tellurium alloy active layer. Such leaching of sodium would result in an undesirable chemical degradation [130].

2.4.4. Organic Dyes

Organic dye films used as optical recording media in the ablative writing mode have advantages over tellurium alloy films.
Organic dye films combine high absorption coefficients with a low decomposition or melting temperature and low thermal
conductivities [119], [120], [136-142]. Moreover, films of organic dyes are usually applied by means of solvent spin-coating
[129], which is simpler than the vacuum techniques used for inorganic films. These potential advantages were recognized
early. Problems with organic dyes are their limited thermal and ambient light (daylight) stability, and the relatively narrow
absorption bands compared with inorganic materials.

In early work, before the advent of reliable, economical, high-power semiconductor lasers operating in the near-IR region,
dyes absorbing in the visible region were chosen [136], [137]. Visible dyes are used in high data density applications.
Usually, bi- or trilayer structures were used to obtain sufficient reflection [137], [138] (see Fig. 63). The dye (up to 30 wt %) is
dissolved in a polymer matrix. At higher concentrations, phase separation and dye crystallization occur. Polymeric dyes can
also be used. For both types of dye system, stable, sensitive media can be designed [139].

The threshold energy for laser marking appears to be determined only by the thermal efficiency of the organic dye layer. The
viscosity does not appear to be important [142]. The CNR, however, depends on the polymer. The higher the viscosity, the
shallower the pits and hence the lower the contrast and CNR [140], [142]. Furthermore, diffusion of the dye into the substrate
sometimes plays a role in the bit formation process [136].

The most important pit formation mechanism is probably melting of the film, followed by partial decomposition or evaporation.
The gaseous components exert a recoil pressure on the molten material and push it into the rim [138], [143], with some pit
material ejected as debris. The resulting pit shape is somewhat different from that observed with tellurium alloys (see Fig. 25,
Section General Introduction).

For semiconductor laser wavelengths, dyes absorbing in the near-IR are needed. They usually consist of conjugated
polymethine chains [143], [144]. These dyes are often susceptible to oxidation. Certain tetraarylpentamethines have sufficient
thermal and ambient light stability. The choice of anion is also important, both for stability and for solubility. The solubility in
the spin-coating solvent should be at least 1 – 2 wt % to obtain a sufficient layer thickness. The solubility in water should be
very low to ensure long-term integrity of the dye layer. For example, chlorides dissolve readily in alcohols and water, but they
are often unstable.

Another type of dye that can be used is the squarylium class [119], [120]. The absorption maximum is usually near 750 nm.
Substitution of the end groups by thiopyrilium units shifts the absorption maximum to 820 nm. With bulky substituents such
as tert-butyl groups, the solubility can be increased to 10 – 20 wt % in n-propanol.

For semiconductor laser write-once recording, pure dyes are coated in a single layer on a pregrooved substrate. In this way,
maximum dye loading is obtained. The reflectivity of these films, substrate-incident, is 15 – 20 %, sufficient for tracking,
focusing, and reading. The near-IR dye films appear to be up to 50 % more sensitive than tellurium alloys [119], [145]. The
organic dye discs can be made using a hermetically sealed air-sandwich construction or with an in-contact protective
overcoat.

Their high CNR values are, however, much more important. The well-defined, reliable, and reproducible pit shape and
position are also important. These properties make organic dye media well suited for high density data and video recording
[119], [138]. Written data can be read at least 105 times without deterioration at normal reading power [119]. Furthermore,
long term stability of up to 50 years is claimed [145]. A metal-overcoated organic dye system which combines a high initial
reflectivity (> 70 %) with a high contrast such that a written disc could be read by a standard CD player [146].

For DVD–R and DVD+R dyes absorbing in the red (650 nm) are used. Because of the high demands for dyes in high-speed

page 43 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
DVD–media every manufacturer uses his own special dye. In Table 12. an overview of some actually used dyes is given.

Table 12. Overview of some actually used dyes for CD-R and DVD+/–R

Phthalocyanine (780 nm, CD-R)

Cyanine (780 nm, CD-R)

Cyanine (650 nm, DVD+/–R)

Azo metal chelate (650 nm, DVD+/–R)

Dipyrromethene metal chelate (650 nm,

DVD+/–R)

page 44 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
2.4.5. Metal Particle Colloids
Small metal particles dispersed in a gelatin matrix can also give a highly absorbing film with low thermal conductivity [147],
[148]. The gelatin is loaded with filamentary silver grains (to absorb the laser light strongly) and with circular silver grains (to
enhance the reflectivity). A reflectivity of 50 % can be attained with this design [148]. The writing mechanism is based on the
absorption of the laser light by the silver particles, which heats and melts the gelatin film, resulting in a surface deformation
that can be read optically. The maximum CNR of the media, however, is not very high.

The colloidal film, on the other hand, is very sensitive; maximum CNR is obtained for a very low energy density (0.7 nJ/µm2).
This film can be overcoated with an in-contact protective lacquer. An interesting application is their use as an optically
recordable flexible card [148].

2.4.6. Metal-Polymer Bilayers

In metal deformation optical recording, a thin absorbing metal layer is plastically deformed by the gaseous products formed
by heating an underlying polymer layer [149]. The laser heats the metal film, and, by thermal conduction, the polymer bottom
layer also heats up and starts to decompose. When the gas pressure exceeds the elastic limit of the metal film a bubble is
formed that remains intact after the bilayer film is cooled. A similar marking mechanism has been found in films which are
composed of tellurium crystals dispersed in an organic polymer matrix [150], [151].

Material flow is not desired, only decomposition of the polymer. High molecular mass polymers are therefore used to obtain a
high softening temperature. A practical molecular mass is in excess of 80 000. Moreover, it is possible to spincoat flat, high
molecular mass polymer films. The polymers used are poly(methyl methacrylate) and polystyrene.

The metal films are required to withstand the large deformation without breaking at the speed of bubble growth (up to
10 m/s). This demands that the metal has high ductility, break resistance, and resilience. The metal layer should be
continuous, adhere to the polymer film, and be resistant to corrosion. Typical metals used are gold or platinum-based alloys
with <25 % additional elements.

Reproducible, well-defined bubbles can be written with reasonable laser power leading to a relatively high CNR and low bit
error rate. Furthermore, the long term stability is claimed to be high [149]. One problem of this recording mechanism is that it
is not self-limiting. Only a rather narrow range of laser power gives high CNR values. If the energy is increased too much, the
metal film breaks and irregular pits result. These metal – polymer bilayer films are used in hermetically-sealed air-sandwich
disc designs.

2.4.7. Alloying Bilayers

Another writing mechanism involving no macroscopic material transport uses the alloying or reaction of thin films, resulting in
a layer with different optical properties [157], [158]. In the proposed system, a complicated quadrilayer construction is
needed. An amorphous Sb2Se3 layer, followed by an amorphous Bi2Te3 layer, is deposited on the substrate. These two
layers react by laser heating to form a four element alloy. The third layer consists again of a second amorphous Sb2Se3
layer. This layer is used as a thermal barrier and to tune the optical properties of the quadrilayer to give a high signal. The
fourth thin film is an aluminum reflective layer. The reflectivity, substrate-incident, is chosen to be as low as possible
considering the tracking and focusing needs (10 – 14 %).

The recording sensitivity appears to be high: 6 mW at 5 m/s and the CNR is reasonably high. The long term stability is
expected to be excellent. Diffusion between the two thin film materials is negligible at ambient temperature.

2.4.8. Agglomeration
A third example of nonablative optical recording is agglomeration of very thin films of noble metals such as gold [118], [159],
[160], silver [161] or tellurium alloys such as Te – Cu [162], [163]. The noble metal films are typically 3 nm thick, whereas the
films of Te – Cu are typically 10 nm thick. The noble metal films, as deposited, are not uniform. They form very small islands
with radii of 5 – 10 nm. The Te – Cu films are uniform.

Laser heating close to or above the melting point induces coalescence of the noble metal islands into larger islands up to four
times their original size. In the Te – Cu uniform films, laser heating induces coalescence thereby producing a discontinuous
film with very small islands within the marks. These coalesence or agglomeration mechanisms induce a change in the optical
properties, which can be detected. To optimize the read-out contrast, an absorption trilayer configuration is used in the media
design.

The advantage of the films is that they are approximately twice as sensitive as ablative tellurium or tellurium alloys on
identical substrates. They are useful for high data rate applications. The media can be overcoated with in-contact protective
organic lacquers.

2.4.9. Surface Texture

When a textured surface or interface is locally heated by a laser above the melting point, the molten area will resolidify under
the influence of surface tension as a flat domain. The textured surface absorbs light strongly because its average periodicity
is much smaller than the wavelength of the incident laser light. The smooth domain area reflects strongly when the material is
a good reflector (see Fig. 61 C). The optical contrast can be high, and the disc noise may be often high.

The most investigated material for producing surface textured media is germanium [164]. The texture periodicity is ca. 40 nm.
High reflection contrast is obtained with ca. 3 nJ of input energy.

Another material is a textured polymer layer (moth-eye) coated with a metal reflector [165]. In the textured polymer layer, the
mark is formed not by melting, but by plastic deformation (blistering) of the polymer at high temperature. Moderately high

page 45 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
CNRs and good bit error rates are obtained.

2.5. Reversible Optical Recording Materials

In reversible optical recording materials, the laser beam used for recording must produce reversible, optically detectable
changes in the sensitive layer of the disc. These can be changes (1) in the direction of polarization, (2) in the intensity, (3) in
the phase of the reflected laser light, or (4) a combination of these. At present, the technologically most advanced erasable
optical storage technologies are magneto-optic (MO) and phase-change (PC) recording.

Read-out in magneto-optic recording is based on the magneto-optic Kerr effect, which causes a rotation of the direction of
polarization of the laser light when it is reflected at the perpendicularly magnetized recording layer. Magneto-optic materials
have been investigated since the late 1950s; generally, amorphous rare-earth transition-metal alloys (RE-TM) such as Tb –
Fe – Co and Gd – Fe – Co are used.

Favorable aspects of MO recording are its high recording density, which can be increased beyond the diffraction limit by
using magnetic super-resolution techniques [166-168], and its high data rate, which is not intrinsically limited by the recording
mechanism. Additionally, the overwrite repeatability (i.e., the number of overwrite cycles) of MO media is almost unlimited.
Disadvantages of MO recording are its rather complex media (complex thin-film stacks, tight optical tolerances), and the
requirement of polarization optics and (in some cases) a magnetic head in the drive. The compatibility with read-only drives
and discs is therefore rather limited.

Phase-change recording is based on the change in the intensity reflection coefficient when the recording material undergoes
a crystalline-to-amorphous phase transition. Phase-change materials for optical recording have been studied since 1970.
Typically, chalcogenide alloys such as Ge – Sb – Te or Ag – In – Sb – Te are used.

The signal amplitudes in phase-change recording are much higher than in MO, due to the large reflectivity difference
between the amorphous and crystalline state. The detection of nonpolarized, intensity-modulated light allows for compatibility
with read-only discs and simplifies the optics in a phase-change optical head, leading to a low cost drive. Direct overwrite is
easily realized by modulation of laser power. Disadvantages of phase-change recording are its inherently poor write power
sensitivity, due to the high melting temperature of the phase-change material, and the limited overwrite repeatability, due to
degradation of the media. The data rate is closely related to the crystallization rate of the phase-change material, which is
limited by atomic mobility.

The aforementioned recording mechanisms are induced by the heat generated from the absorption of optical energy by the
recording material. A second category is formed by photon mode recording mechanisms, which involve electronic transitions
in the optical absorption spectra. Examples of such recording technologies are photochromic and photochemical (spectral)
hole burning. Photon mode materials for optical recording have been investigated since the late 1970s. However, the
progress in this field has been relatively slow and is still far from applications for the mass consumer market.

2.5.1. Magneto-Optic Materials

History. Magneto-optic (MO) recording has been commercially available longer than any other reversible optical storage
format. The first generation of the ISO 5¼ inch (130 mm) MO drive was introduced in 1988 and offered 650 MB capacity on a
double-sided disc (i.e., 325 MB per side). The 3½ inch (90 mm) MO drive arrived on the market in 1991 with 128 MB
capacity. Since then considerable progress has been made in improvement of storage capacity, data transfer rates, and
cost : performance ratio. Current (1998) 130 mm drives, the so-called 4X MO, have a capacity of 2.6 GB (1.3 GB per side),
while the 90 mm format offers 640 MB per disc [169]. The other major MO storage product family is the Sony MiniDisc
system, introduced in 1992 and designed primarily for digital audio applications [170]. In 1993 a version of this for data
storage, MD DATA, entered the market with a capacity of 140 MB on a 2½ inch (64 mm) disc. MD DATA2, with a capacity of
650 MB, identical to that of a standard CD-ROM, has now been announced [171]. Looking to the immediate future, a number
of industrial companies have formed the Advanced Storage Magneto-Optical (ASMO) consortium with the aim of producing a
MO system with 6 GB capacity on a 120 mm single-sided disc, alongside compatibility with the popular DVD and CD formats
[172]. One fundamental problem with such high storage densities is that the size of the bit that is recorded in the MO disc is
actually too small (typically less than 0.2 µm in length) to be read back using conventional optical techniques. To overcome
this limitation the ASMO format proposes to read the data using a form of magnetically induced super-resolution (MSR) [172],
[173].

Paralleling these increases in storage capacities, MO recording has shown similar improvements in data transfer rates (i.e.,
how quickly data is transferred between the host system and the disc drive). However, unlike for conventional magnetic
recording, it is not straightforward to implement a direct overwrite (DOW) scheme in MO recording. Rather, old data is first
erased in one pass of the optical head before new data is written on a second pass. This obviously slows down data transfer,
and much effort has been put into finding reliable MO – DOW methods. The scheme adopted by Sony for their MiniDisc
system is essentially a form of thermally assisted magnetic recording known as Magnetic Field Modulation (MFM) [170],
[171]. Another approach, realized in the laboratory over a decade ago but only recently becoming a commercially reality, is
Light Intensity Modulation Direct Overwrite (LIMDOW) [174], [175]. Ways of combining these direct overwrite techniques with
super-resolution read-out are being developed [176].

All of the above developments place special materials requirements on the storage medium. The materials that have proved
consistently capable of meeting these requirements are based on alloys of rare earth (RE) and transition metal (TM)
elements, RE – TM films for short. All MO systems currently available commercially, and those under commercial
development, use RE – TM films as the storage medium. This does not mean that other materials are not suitable for MO
storage applications. Indeed, much research has been carried out into alternative MO media, including platinum – cobalt
multilayers and alloys for use with short-wavelength lasers [177], [178], garnet films for enhanced stability [179], [180],
manganese-based materials for high read-out signal levels [181], and, perhaps most exotic of all, europium and uranium
chalcogenides with enormous Kerr rotations [182]. One reason for the continued dominance of RE – TM alloy films is that
their magnetic, thermal, and optical properties can be varied in a controlled fashion by varying the alloy composition, the

page 46 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
deposition conditions, or both. Thus film properties can be tailor-made to suit a particular recording technique, be it
conventional MO recording or more advanced formats such as MFM, LIMDOW, or MSR [183].

Recording Principles. A more accurate name for MO recording is thermomagnetic-magneto-optic (TMO) recording. This
reflects the fact that the writing of data to the disc is effected by thermal and magnetic effects, whereas the read-out of data
from the disc relies on magneto-optic effects. The principle is as follows. A small region of a uniaxial magnetic film with the
easy axis perpendicular to the film plane is heated to a critical temperature, invariably the Curie temperature, at which the
coercivity falls to a very low value. The film then cools to ambient temperature in an external magnetic field. Upon cooling,
the heated region acquires the magnetization direction of the external field. The state of magnetization of the neighboring
unheated regions is not affected by this reversal. Provided that the coercivity at ambient temperature is sufficiently large, the
written magnetic domains, or bits, are stable even when the applied field is removed. This process is reversible and
repeatable; hence, MO recording is both erasable and rewritable [184]. The source of heat is a high-power, solid-state, laser
focused to a diffraction-limited spot of about 1 µm diameter and modulated by the information signal. The magnetic field is
supplied by a small electromagnet or rotatable permanent magnet.

Reading of the stored information is performed using the polar magneto-optic Kerr effect. The laser is again focused on the
MO film, this time at low power. The state of polarization of the beam reflected from the disc is dependent on the
magnetization direction, which in turn depends on whether a binary '1' or '0' has been written. A bipolar signal representing
the two states of magnetization is extracted by analyzing the polarization direction of the reflected beam using polarization
sensitive optics in the read-out head [185]. A typical MO drive is shown in Figure 65.

Figure 65. A typical magneto-optical disc drive with astigmatic focus and push – pull tracking.

Magneto-Optic Materials Requirements. To be a candidate for MO recording, a material must satisfy certain magnetic,
thermal, and optical requirements. For conventional MO recording the fundamental requirements include: (1) perpendicular
magnetic anisotropy, (2) a large coercivity at ambient temperature (to provide stability against stray magnetic fields and allow
for high recording densities), (3) a temperature variation of coercivity that allows writing and erasing at the moderate
temperatures achieved by heating with short duration pulses from solid state lasers, and (4) sufficiently large magneto-optic
effects (Kerr rotation) to generate a usable read-out signal. Additionally the recording medium should be environmentally and
chemically stable, homogeneous, low-noise and be suited to commercial production. It has not proved possible to meet all
these requirements using a single layer recording medium, and so the usual approach is to combine active, protection,
reflector, and enhancement films in a quadrilayer disc structure, as shown in Figure 66. Recording and read-out is achieved
by focusing through the substrate. This provides immunity to dirt and scratches on the disc surface and helps to ensure that
the disc is removable, one of the main assets of optical recording.

Figure 66. A typical quadrilayer disc structure based on RE – TM alloy as the active MO layer.

RE – TM Materials. The first reported studies of MO recording took place in 1957. This pre-dates the invention of the laser
and a thermal pen was used to write magnetic domains in a Mn – Bi film, with reading achieved by white light in a polarizing
microscope. Mn – Bi had suitable magnetic properties and large magneto-optic effects. However, Mn – Bi is thermally
unstable as it exists in several crystalline phases and its Curie temperature is close to its melting point. It also has a
polycrystalline morphology with relatively large grain size, resulting in high media noise levels that in turn results in a low
signal-to-noise ratio. In the beginning of the 1990s Mn – Bi films have been revisited, in particular by studying the effects of
additives — notably aluminum to form Mn – Bi – Al — for the purpose of reducing grain size, eliminating temperature-induced
phase transitions and enhancing MO effects [181], [186].

The major breakthrough in the development of MO recording materials came in 1973 when IBM researchers discovered a
new class of MO materials, the amorphous rare earth-transition metal (RE – TM) alloys [187]. These overcame many of the
shortcomings of early materials, and are nowadays the only media used in commercial products. A host of binary, ternary,
and quaternary RE – TM alloy compositions have been investigated. In cases where the RE constituents are from the heavy
RE metals (Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu), the magnetic moments of the TM and RE subnetworks tend to couple
antiferromagnetically, resulting in ferrimagnetic behavior. Where light RE elements (La, Ce, Pr, Nd, Pm, Sm, and Eu) are
involved, the subnetwork coupling is generally ferromagnetic. For commercial applications the TM elements of choice are Co
and Fe, whilst the RE elements are Tb and Gd, sometimes also Dy. In these cases, ferrimagnetic behavior is observed and
the magnetic and magneto-optic properties vary as a function of temperature as shown schematically in Figure 67.

page 47 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 67. Variation of the magnetic and magneto-optical properties of antiferromagnetically coupled amorphous RE – TM
materials as a function of temperature.

The RE and TM subnetwork moments, MRE and MTM, are equal in magnitude at the compensation temperature TCOMP
where, since MRE and MTM are coupled antiferromagnetically and MNET = |MTM + MRE|, the net magnetization is zero.
Above TCOMP the net magnetization initially increases and then goes to zero again at the Curie temperature, TC, where the
material becomes paramagnetic [188].

Straightforward binary RE – TM alloys are not very suitable for MO recording. For instance, the Curie temperature of Tb – Fe
is too low, whereas Gd – Fe has a very high TC in addition to an in-plane anisotropy. Consequently, practical MO RE – TM
materials are invariably ternary or quaternary alloys. Adding Tb to Gd – Fe, for example, increases the perpendicular
magnetic anisotropy and the coercivity, whilst at the same time decreasing the Curie temperature. Adding Co to Tb – Fe
increases both the Curie temperature and the Kerr rotation [189], [190]. The ability to vary magnetic properties over a wide
range by changing the RE – TM alloy composition provides enormous flexibility in material design, and is one of the main
reasons for the continuing predominance of RE – TM media. The compensation temperature is determined primarily by the
ratio of RE to TM in the film, with changes of a few percent shifting TCOMP by more than 100 °C [183], [191]. Curie
temperature depends mainly on the Co : Fe ratio; for example TC varies from 135 °C for Tb – Fe to over 400 °C for Tb – Co
[192]. Coercivity (HC) and anisotropy (KU) are strongly affected by the percentage RE content and the deposition conditions.
A materials design strategy can be summarized as [183]:

1. Determine TC by selecting Co : Fe ratio;

2. Determine TCOMP by selecting RE : TM ratio;
3. Determine KU and HC by deposition conditions and selecting Tb : Gd ratio (if appropriate).

Typical useful compositions in conventional MO recording applications are Gd20Tb5Fe75 and Tb27Fe65Co8 [193]. Additional
elements, such as Al, Cr, Ta or Ti, may be included to improve the oxidation, corrosion and pitting resistance of the RE – TM
film [194], [195]. However, the primary environmental protection for the RE – TM film is provided by the two dielectric layers
between which it is sandwiched (see Fig. 66). These are usually aluminum or silicon nitride (AlN or SiN) and without them the
lifetime of the RE – TM film can be measured in hours or even minutes [196].

The magneto-optic effects in RE – TM films are primarily brought about by the TM subnetwork, and only indirectly dependent
upon the RE content. Incoming photons excite the 3d valence electrons of the transition elements altering their orbital
angular momenta. Through the spin-orbit interaction, photons that are re-emitted (reflected) when the excited electrons relax
are polarized. For perpendicularly magnetized material and linearly polarized incoming light, this process manifests itself as
the polar Kerr effect. In general, the Kerr rotation decreases with increasing RE content. Also, as increasingly heavier RE
elements in the series Gd, Tb, Dy, Ho, are used, a monotonic decrease in Kerr rotation is observed [183]. The intrinsic Kerr
rotation angle for all RE – TM compositions of practical interest is small, having absolute values at visible wavelengths of less
than 0.5° [197]. Furthermore, the rotation decreases with decreasing wavelength. Since future optical recording systems are
expected to move to shorter, blue, wavelengths where the size of the focused laser spot is smaller and thus the storage
densities are larger, this decrease of Kerr rotation is of some concern and has led to much research into alternative short-
wavelength materials.

Platinum – Cobalt Multilayer Films. Of the materials considered as replacements for the RE – TM media, the most
promising are those of platinum-cobalt multi-layered thin films made of a stack of alternated Co and Pt layers, as shown in
Figure 68. These can be produced with perpendicular anisotropy and exhibit a Kerr rotation that increases considerably as
the wavelength is reduced. Although extremely thin, Pt – Co multilayer films are remarkably stable on exposure to air and
provide an inherently stable and simple disc structure. The films exhibit ferromagnetic behavior and are nanocrystalline.

Figure 68. Structure of a Pt – Co multilayer, N × (x Co + y Pt) + z Pt where N is typically 15, x is ≈ 0.4 nm, y is ≈ 1 nm and z
is ≈ 1 to 5 nm

The development of Pt – Co followed studies in the mid-1980s of the use of Pd – Co multilayers for perpendicular magnetic
recording applications [177]. Perpendicular anisotropy occurs for cobalt layer thicknesses < 0.8 nm, where the interface
anisotropy dominates over volume anisotropy [198], [199]. Optimum magnetic properties are found for Co layer thicknesses
in the range 0.3 to 0.6 nm and Pt layer thicknesses in the range 0.8 to 2 nm. The number of bilayers are typically in the range
10 to 15, giving an overall film thickness of around 10 to 40 nm [200], [201]. Use of Pt – Co multilayers as MO recording
media was demonstrated for the first time in 1989 [202], [203], [204]. Curie temperatures are typically around 350 °C to 400 °
C, somewhat higher than for RE – TM films and necessitating higher laser record powers. TC can be controlled to some

page 48 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
extent by variation of the Pt and Co layer thicknesses, by the addition to the Co layer of a few percent of Os or Re or by using
CoX – Pt type multilayers, where X includes a large number of TM, RE, or semiconductor elements [205]. The coercivity of
the Pt – Co multilayers depends very strongly on a number of extrinsic factors such as deposition conditions, substrate type
and treatment, and underlayer type and thickness. Films prepared by vacuum evaporation have room temperature
coercivities comparable to RE – TM films. However, films prepared by standard argon sputtering techniques, the preferred
commercial approach, have relatively low coercivities. HC can be increased by sputtering in krypton and xenon instead of
argon, and by using Pt or other underlayers [206], [207]. Comparative studies of the recording properties of Pt – Co
multilayers versus RE – TM films at a number of different wavelengths show that at 820 nm the RE – TM film provided the
best SNR, whereas at 458 nm the Pt – Co system had the highest SNR [208]. In spite of such promising results, Pt – Co
multilayer films show no sign of being adopted commercially. This in part is due to the established order, but also to the
versatility provided by RE – TM materials for the design of advanced media formats.

Garnets and Oxide-Based Materials. Garnets are oxides of RE metals and iron, expressed as R3Fe5O12 or 3 R2O3 · 5 Fe
2O3. They have outstanding structural and chemical stability and very favorable MO properties in the visible and soft-UV
region [209]. Their characteristics are deposition dependent and can also be tailored by chemical substitution. The
perpendicular anisotropy of sputtered garnets is primarily stress-induced and can be increased by substitution of ions with
high magnetostrictive effects such as dysprosium. Coercivity depends on composition and grain size, and thus on
preparation, with typical values 40 kA/m to 240 kA/m [209], [210]. Garnets of compositions (Bi,Dy)3(Fe,Al or Ga)5O12 [209]
Ce1Dy2Ga0.4Fe4.6O12 [211], and Bi3Fe5O12 [212], have good potential as candidates for MO recording. However, a major
drawback of all garnets is their highly polycrystalline nature. This, as in the case of the first MO media, Mn – Bi, increases
media noise to an extent that SNR levels are too low to be of practical use [213]. Research efforts are directed towards
reducing the grain size [214], [215]. One possible future application of garnets is in so-called volumetric recording, where
information is stored in more than one layer on each side of a disc [216]. The semitransparency of garnet films is an
attractive feature in such applications. Other iron oxide-based materials, such as Co-ferrite CoFe2O4 and hexagonal Ba-
ferrites, have a high MO figure of merit and an unsurpassed chemical stability and have been proposed for MO recording.
However, low SNR values are again obtained due to grain-induced noise.

Advanced Media. Advances in MO media development have enabled the technology to overcome two substantial
drawbacks of conventional MO recording: firstly its inability to perform direct overwrite; secondly the limit on storage density
imposed by optical diffraction which states that it is not possible to “see” features smaller than the wavelength used for
observation. Two types of DOW exist, Magnetic Field Modulation (MFM) and Light Intensity Modulation (LIMDOW). MFM
media are virtually identical to conventional RE – TM MO media. LIMDOW media are substantially more complicated,
although they are also made from amorphous RE – TM alloys.

LIMDOW was first demonstrated in 1987 and relies on exchange-coupling between various RE – TM layers of different
compositions [174]. In its most basic form the LIMDOW medium has two RE – TM layers; a memory layer and a reference
layer. The memory layer has a high coercivity HCM and a low Curie temperature TCM, whereas the reference layer has a
lower coercivity HCR and a higher Curie temperature TCR. The recorded information is always read from the memory layer.
Two magnets, an initializing magnet and a bias magnet, are needed for this scheme to operate. The initializing magnet
generates a magnetic field oriented in the direction corresponding to the magnetization state of an erased disc, i.e., logic '0'.
The field is large enough to saturate the reference layer at room temperature, but small enough so as not to affect the
memory layer. The bias field points to the opposite direction, i.e., logic '1', and is too small to alter the magnetization in either
layer at ambient temperature. The laser is operated at two power levels, PLOW and PHIGH, during writing. At PLOW the
temperature of the film exceeds the Curie temperature of the memory layer, but not that of the reference layer whose
magnetization remains unaffected by the bias field. On cooling, a copy temperature, TCOPY, below TCM is reached at which
the magnetization of the reference layer is copied to the memory layer by exchange coupling [217]. Thus, a '0' has been
written. To record a '1', the laser is pulsed at high power, PHIGH, to heat the entire film above TCR. This time on cooling the
reference layer's magnetization assumes the orientation of the bias field. When the TCOPY is reached the memory layer also
acquires the magnetization direction of the bias field via the exchange coupling with the reference layer. Finally the reference
layer is reinitialized as it passes above the initializing magnet (see Fig. 69). The need for two permanent magnets with one of
them, the initializing magnet, generating a high field complicates disc drive design. A solution to this problem is the
quadrilayer exchange-coupled structure that has two additional layers, an initializing layer and a switching layer, that remove
the requirement for an initializing magnet [217]. Commercial systems employing such materials have been launched on the
market by several manufacturers.

Figure 69. The LIMDOW direct overwrite mechanism in exchange-coupled double-layer films showing the disc structure
(top) and the overwrite mechanism (bottom).

MSR technology improves the read-out resolution by incorporating a read-out layer that copies information from the memory
layer only within a limited temperature range. Thus, data from the memory layer is read though what is effectively a very
small, substantially submicron, aperture that is placed directly above the memory layer [173]. The optical resolution is thus
determined by the size of the aperture, rather than the wavelength used for read-out.

page 49 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Several MSR schemes are possible, the commonest being Rear-Aperture-Detection (RAD), Front-Aperture-Detection (FAD)
and Central-Aperture-Detection (CAD); the name arising from the location of the effective aperture in the focused laser spot
that reads the data and induces the necessary temperature rise for aperture formation [169]. MSR techniques require at least
two active MO layers, one for the memory layer and one for read-out and aperture formation. In reality extra layers are
needed to control the interlayer coupling and switching. A typical CAD – MSR system is shown in Figure 70 [172], [218].
Here, a read-out layer of RE – TM material, typically Gd – Fe – Co, has an in-plane magnetization at room temperature. As
the temperature of the film is increased slightly by the read-out beam, a small aperture forms in which the magnetization, due
to magneto-static coupling with the memory layer, becomes perpendicular. The nonmagnetic intermediate layer, usually a
dielectric, moderates the level of magneto-static coupling. The memory layer is typically Tb – Fe – Co, as in conventional MO
films. Using such a scheme the bit size in both circumferential and radial directions can be substantially smaller than when
using a conventional recording approach. Over 6 Gbytes of storage on a single-sided 120 mm disc has been reported [172],
[218].

Figure 70. Central aperture detection system in which the readout layer is typically Gd – Fe – Co (≈ 30 nm thick) and the
memory layer is Tb – Fe – Co (≈ 40 nm thick).

Magnetically Amplified Magneto-Optic System. Yet another MO media format that is the topic of much research is the
Magnetically Amplified Magneto-Optic System (MAMOS). Here the opposite approach to that of MSR is adopted to achieve
high density read-out in so much as instead of reading out data through a small effective aperture, small bits are copied from
a memory layer to a read-out layer and then expanded in size before being read in a conventional way [219]. Again the
materials of choice for this scheme are RE – TM films due to the design flexibility that they offer.

2.5.2. Phase-Change Recording Media

Principles. Phase-change recording is based on the formation of (sub)micron-sized amorphous marks in a crystalline thin
film by using a focused laser beam. The amorphous and crystalline phase have different refractive indices, leading to
amplitude and/or phase modulation of the reflected laser beam during readout. Since the transition between the crystalline
and amorphous phase is reversible, information can be erased and/or overwritten (see Fig. 71).

Figure 71. Principle of phase-change rewritable optical recording. The write process involves the application of a short high-
power laser pulse to locally melt the phase-change layer. After turning the laser power to a lower level, the molten material is
quenched and an amorphous dot is formed. Recrystallization is prevented because the time interval between the melting
temperature Tm and the glass transition temperature Tg (the gray bar in the figure) is much shorter than the crystallization
time (the black bar).

The erase process involves the application of an intermediate d.c. erase power level. The phase-change layer is heated to a
temperature between the glass transition temperature Tg and the melting temperature Tm sufficiently long to recrystallize the
material.

An amorphous mark is written by using a short, high-power laser pulse that melts the recording layer locally. Immediately
after the pulse the laser is switched to a low level, so that the molten material is rapidly quenched; if the cooling rate is
sufficiently large, recrystallization can be avoided and an amorphous dot is formed. A recorded mark is erased by operating
the laser at an intermediate d.c. erase level, thus heating the recording layer to temperatures between the glass transition
temperature Tg and the melting point Tm. In this temperature range the atoms have sufficient mobility to enable
recrystallization of the material. For direct overwrite applications, the crystallization rate of the phase-change material should
be sufficiently fast to erase the recorded marks completely within the dwell time of the laser spot. The crystallization rate of
the material is thus of key importance to the data transfer rates that can be achieved.

Material Specifications. In order to understand the properties that are relevant for rewritable phase-change materials, it is
necessary to examine the crystallization process in some detail [220], [221]. Crystallization of the amorphous phase involves
nucleation and growth processes, similar to the crystallization of a liquid during undercooling. In fact, both cases can be
considered equivalent, because during heating the amorphous phase usually undergoes a glass transition (at the glass
transition temperature Tg) to become a supercooled liquid prior to crystallization.

Crystal Growth. The driving force for crystallization is the free energy difference G between the supercooled liquid (L) and
crystalline state (C), given by the equation:

where HLC and SLC are the enthalpy and entropy difference between the two states, respectively. This relation can, to
first order, be approximated by:

where Hf is the latent heat of fusion and Tm the melting temperature. The equation shows that the driving force for
crystallization increases with the undercooling Tm – T. In contrast, the mobility of the atoms in the supercooled liquid
decreases with increasing undercooling. The combination of both effects results in an initial increase of the crystal growth

page 50 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
rate when the material is cooled below Tm. At a certain undercooling the growth rate maximizes, and it decreases again
when the material is cooled further towards the glass transition temperature Tg. Below Tg, the mobility and consequently the
growth rate are negligible (Fig. 72).

Figure 72. Probability density of the nucleation and growth processes as a function of temperature T. Nucleation takes place
only at considerable undercooling Tm – T, with Tm the melting temperature. The growth rate maximizes at higher
temperatures. Below the glass transition temperature Tg the mobility of the atoms is too small to enable crystallization.

Nucleation. In order to start crystallization, nuclei have first to be formed. The energy required to create the solid-liquid
interface, given by the solid-liquid interfacial energy , forms a barrier towards nucleation. The work of formation Gn for a
cluster of n atoms is given by:

where G' is the free energy difference between an atom in the solid and an atom in the liquid, and A is a geometry factor.
Since G' is negative and positive, a critical nucleus size n* exists at which the heat of formation is maximal. Below this
size the nuclei will on average shrink, above this size the nuclei tend to grow further to crystallites. At small undercooling,
where the critical nucleus size n* is large, the nucleation rate is negligible. The onset of nucleation occurs only at substantial
undercooling. At temperatures just above Tg, the nucleation rate is small due to the low atomic mobility (Fig. 72).

The density of the nuclei determines the number and size of crystallites that are being formed. This is an important parameter
in optical recording, because the grain size of the crystalline phase determines the ultimate bit density that can be obtained.
However, the nature of the actual nucleation site is often unclear: Nuclei may form spontaneously within the material itself
(homogeneous nucleation), or at impurities or interfaces where the interfacial energy is decreased (heterogeneous
nucleation).

On the basis of crystallization theory, the following criteria for obtaining fast crystallization rates can be discerned:

1. The free energy difference between the supercooled liquid and the crystalline phase should be large. In an alloy
system, the free energy level of the liquid state varies slowly with composition, whereas the free energy of the
crystalline state drops steeply at stoichiometric compositions. Thus, the driving force for crystallization and
consequently the crystallization rate will be maximal for stoichiometric compounds.
2. The mobility of the atoms in the amorphous and supercooled liquid phase should be high. The mobility is affected by
the viscosity of the liquid phase, which depends on the binding energy of the atoms. When the binding forces between
atoms are strong, the mobility is decreased and crystallization is slow. In general, if compound materials do not have
strong covalent bonds the crystallization rate is relatively large.
3. The atomic diffusion distances between the amorphous and crystalline state should be small. These distances are
closely related to material composition and crystal structure. A nonstoichiometric compound generally crystallizes into
different phases. Phase separation requires atomic diffusion over longer distances and is time-consuming. Therefore,
crystallization of a stoichiometric compound into a single phase is generally faster. Crystal structure plays also a role.
The atomic distribution in the amorphous state is most probably isotropic as in the liquid state. Therefore, it is likely
that materials with a simple isotropic crystal structure crystallize more rapidly than materials with a more complex
anisotropic crystal structure.

A suitable phase-change optical recording material should have the following properties:

1. The optical constants should allow sufficient contrast between the amorphous and crystalline state at the recording
wavelength. This suggests the use of semiconductors instead of metals or insulators. Semiconductors generally
exhibit a larger difference between optical constants of the amorphous and crystalline phase, due to the fact that the
absorption edge shifts in the visible range during the phase transition.
2. The sensitivity of the material should be appropriate: The melting point must be low enough to melt the material with
the available laser power. On the other hand, the amorphous phase must be sufficiently stable against self-
crystallization at ambient and storage temperatures up to 50 °C. In practice, this means that the melting point Tm
should be between 500 and 1000 °C, while the glass transition temperature (typically 50 – 70 % of Tm) should be well
above 100 °C.
3. The amorphous-to-crystalline (a – c) and crystalline-to-amorphous (c – a) transitions should both be rapid. For high-
speed direct overwrite applications, the transitions should proceed within the dwell time of the laser spot, typically
100 ns or less. This criterion sets an upper limit on the complete erasure time, i.e., the time required to erase
(crystallize) an amorphous mark completely. On the other hand, the complete erasure time should not be too small to
prevent recrystallization of the molten material during the write (melt – quench) process.
4. The amorphous-to-crystalline (a – c) and crystalline-to-amorphous (c – a) transitions should be possible many times.
Depending on the application of the disc, a repeatability of up to 105 write – erase cycles is required. It appears
beneficial to use single phase materials, in order to prevent progressive phase separation during the many write and
erase cycles.

page 51 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Material Compositions. Since the first observation of rapid reversible laser-induced amorphous – crystalline transitions by
OVSHINSKY in 1970 [152], [257], a variety of materials for reversible phase-change recording has been investigated. Mainly,
alloys based upon elements from the chalcogen family (sulfur, selenium, and especially tellurium) have been used, because
these materials can be molten by low-power laser irradiation, quenched into an amorphous state easily, and have
appropriate optical differences between the two structural states.

In the early years of phase-change materials research (1970 – mid-1980s), the media consisted of a phase-change layer
sandwiched between dielectric layers on a plastic substrate. The relatively low cooling rate of these stacks led to the search
for phase-change materials which could be easily vitrified. Eutectic (low melting point) compositions of Te-based alloys were
found to be promising candidates: These alloys have a relatively low mobility in the liquid phase, resulting in an easy
amorphization process. The low mobility is explained by the presence of long Te-chains, even in the liquid state, which are
extensively cross-linked by the alloying elements that are used, such as Ge, As, or Sn [258], [260]. The negligible mobility at
room temperature results in long data retention times. However, the inherent disadvantage of the small mobility is the large
crystallization times of these materials, typically on the order of several hundreds of nanoseconds or microseconds [152],
[261] which excludes their use in high data rate direct overwrite applications.

During the 1980s, it was gradually appreciated that virtually every material can be vitrified if the cooling rate is sufficiently
large. By using thin-film stacks as described below (with a so-called “rapid cooling structure”), quench rates on the order of
1010 K/s can easily be obtained. Therefore, the rate of the crystallization process is nowadays a more important issue than
the ease of vitrification of the material. In order to shorten the crystallization times, stoichiometric compositions have been
studied from the mid-1980s. In 1986, the feasibility of direct overwriting with one laser beam was demonstrated for In – Se
based films [262], [263]. Around the same time, it was shown that a stoichiometric GeTe alloy can be crystallized within 30 ns
[258]. However, practical application of GeTe appeared to be limited by the large decrease in crystallization rate for
compositions that are only slightly off-stoichiometric, due to the decrease of the nucleation rate [264] and/or the phase
separation of excess Ge or Te [259].

A major breakthrough in phase-change materials research was the discovery of Ge – Sb – Te ternary alloys as recording
material [265-267]. These alloys crystallize within 100 ns for a relatively wide range of compositions near the GeTe – Sb2Te3
pseudo-binary line (Fig. 73). Especially the stoichiometric compounds on this line (Ge2Sb2Te5, GeSb2Te4, and GeSb4Te7),
with crystallization times below 50 ns, have interesting recording characteristics: Phase segregation does not take place in
these compounds during repeated melting and recrystallization, resulting in a direct overwrite repeatability of more than 105
cycles.

Figure 73. The Ge – Sb – Te ternary alloy system. The indicated stoichiometric compounds on the pseudo-binary GeTe-Sb
2Te3 tie line have crystallization times on the order of 30 – 50 ns, and are the nowadays most frequently used materials for
high-density phase-change recording.

In their thermodynamically stable configuration, the structure of the stoichiometric GeTe – Sb2Te3 compounds is built up of
planar homo-elemental layers which are fcc-stacked to form hexagonal lattices [266], [268], [269]. The periodicity of these
lattices varies from 9 (Ge2Sb2Te5) up to 21 layers (GeSb2Te4), which makes these structures rather complex. However,
under laser irradiation a metastable NaCl-type structure is formed, which is built up of Te-planes and planes with lattice sites
randomly occupied by Ge and Sb [260]. This isotropic cubic structure resembles the isotropic atomic distribution of the
amorphous state, so that only small atomic movement and bond angle adjustment is required for crystallization, which
explains the fast crystallization rate of these materials. Transformation from the metastable cubic phase to the hexagonal
equilibrium phase is kinetically inhibited under short laser-pulse irradiation.

Regarding their short crystallization time, pseudo-binary GeTe – Sb2Te3 alloys are particularly suitable recording materials
for media that are used at high linear velocities (5 m/s or higher), such as the high-density rewritable DVD media. These
compositions appear less practical for use at lower linear velocities, because then the recording layer (partially) recrystallizes
during the melt – quench process, due to the decreased cooling rate of the disc. Therefore, the crystallization rate of the
recording materials for low velocity rewritable media (such as CD-Rewritable) should be decreased intentionally. A possible
way to decrease the crystallization rate of the Ge – Sb – Te alloy is the choice of compositions remote from the pseudo-
binary GeTe – Sb2Te3 line, e.g., by adding Sb to a composition on the GeTe – Sb2Te3 tie line [270], or by adding an
additional element, such as Co [271].

In CD-RW media, introduced in 1996, the phase-change layer generally consists of a quaternary Ag – In – Sb – Te alloy
[272-274]. The crystallization time of these materials is typically on the order of 100 – 200 ns, making them suitable for the
range of linear velocities that is currently covered in CD-RW (CD-1X: 1.2 – 1.4 m/s, CD-2X: 2.4 – 2.8 m/s). An advantage of
these materials relative to (doped) Ge – Sb – Te alloys is their high sensitivity. Unfortunately, the overwrite repeatability is
limited to 104 cycles, possibly due to phase segregation in this multiphase material [275].

Structure of Phase-Change Recording Media. A rewritable phase-change optical disc consists of a thin-film stack which is
sputter-deposited onto a pregrooved plastic substrate (Fig. 74). The stack comprises at least four layers, a dielectric
underlayer, the phase-change recording layer, a dielectric upper layer, and a reflective mirror layer. The disc is covered with
a protective lacquer layer, which is applied onto the quadrilayer stack by spincoating. As in the case of prerecorded (read-
only) media, the information in the recording layer is written and read through the substrate, making the system very robust

page 52 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
against dust and scratches. The pregrooves in the substrate form a grating structure which is used for tracking with the radial
push – pull method (see Section The Optical Channel).

Figure 74. Typical four-layer stack design of current phase-change rewritable discs. The phase-change layer is sandwiched
between protective dielectric layers, which are also used to optimize the optical and thermal properties of the disc. The metal
layer has a mirror and heat-sink function.

In current commercially available recording formats, such as CD-RW, information is only recorded in the grooves (groove
recording, see Fig. 75). Typically, groove depths of approximately /8n (with the wavelength of the laser and n the
refractive index of the substrate) are used to optimize the magnitude of the push – pull tracking signal. The distance between
the grooves (track pitch) is an important parameter in optimizing the storage capacity of a rewritable disc. In practice, the
minimum track pitch in groove recording is limited by optical cross-talk between adjacent data tracks and by the stability of
push – pull tracking.

A further reduction of the track pitch is possible by applying land/groove recording (Fig. 75), in which information is recorded
in both land and groove tracks. Although adjacent data tracks are closer together, optical cross-talk between them can now
be canceled to a large extent by optimizing the depth (approximately at /6n) and shape of the grooves. The track pitch that
can be achieved in land/groove recording is typically 75 % of that in groove recording. Radial tracking is not really a critical
factor in land/groove recording, because the groove pitch (twice the track pitch) is much larger than its critical value
approached in groove recording. The limiting factors in reducing the land/groove recording track pitch are optical cross-talk
and thermal cross-erase, which is the erasure (recrystallization) of amorphous marks in the neighboring tracks during the
write process [276]. A possible solution against thermal cross-erase is the use of steep and/or deep grooves, so that the land
and groove tracks become thermally isolated [277].

Figure 75. Principles of groove recording (left panel) and land/groove recording (right panel).

Thin-Film Stacks Design. Materials choice and design of the thin-film stack is a rather complex matter, because the recording
performance of the disc depends critically on the physical and chemical properties of the various layers in the stack.

The sensitivity of the disc is closely related to the optical and thermal characteristics of the stack: A large part of the incident
optical energy should be converted into heat within the recording layer, while at the same time the cooling rate of the molten
material in the recording layer must be sufficiently large (on the order of 1 – 10 K/ns) to prevent recrystallization during
quenching. The cooling rate of the stack is mainly determined by the thermal resistance of the upper dielectric, and the heat
capacity and thermal conductivity of the metal mirror heat sink. In practice, these requirements limit the thickness range of
the upper dielectric layer. The requirement that the incident optical energy has to be used to a large extent for heating the
phase-change layer sets a lower limit to the layer thickness of the upper dielectric. If this layer would be too thin, the heat
would almost instantly flow into the metal mirror, which has a much larger thermal conductivity than the other materials. On
the other hand, the requirement of a sufficiently large cooling rate sets an upper limit to the thickness of the poorly heat
conducting upper dielectric layer. The lower dielectric layer should be taken sufficiently thick, so that the plastic substrate
does not heat too much during recording.

The optical contrast between the amorphous and crystalline phase is optimized by tuning the thickness of the phase-change
and the lower dielectric layer. As for prerecorded media, the background (crystalline phase) is designed to have a high
reflectivity, while the recorded marks (amorphous phase) have a low reflectivity. For sensitivity reasons, the crystalline
reflectivity is typically between 20 – 25 %, and the amorphous reflectivity below 5 %. In groove recording, the phase
difference between the light reflected from the crystalline or amorphous phase can be used to further optimize the signal
modulation. In land/groove recording, cancellation of optical cross-talk between adjacent data tracks requires a very small
phase difference between light reflected from the amorphous and crystalline phase.

The choice of dielectric layer materials is particularly important for the overall recording performance. The chemical nature of
the dielectrics influences the crystallization kinetics of the phase-change material, because the initial stage of crystallization
(nucleation) is sensitive to the energy of the phase-change/dielectric interface (see above). The dielectric layers should act
as diffusion barriers to prevent the phase-change material against oxidation, contamination, or vaporization, at storage
temperatures but also during the write process, when the temperature in the phase-change layer increases up to 1000 °C.
Stability of the dielectrics themselves during repeated heating is a key issue. Mechanical properties play also a role: The
thermal expansion of the various layers in the stack while locally heating the phase-change layer induces large stresses. This
results in void formation and mass transport in the phase-change layer, the formation of microcracks in the dielectric layers,
and/or delamination of the phase-change layer and dielectric layers. These phenomena may seriously limit the number of
direct overwrite cycles (DOW repeatability). The choice of suitable dielectric materials can to some extent reduce these
degradation effects.

In current phase-change recording media, the most frequently used dielectric is a mixture of 80 % ZnS and 20 % SiO2. This
mixture appears to give good power sensitivity and overwrite repeatability. A disadvantage of this material is its low sputter
deposition rate and consequently its cost. Therefore, alternative dielectrics (oxides, nitrides, carbides) are currently under
investigation. Table 13. shows the layers of a DVD–RAM phase-change stack.

page 53 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Table 13. Phase-change stack of a 4.7-GB DVD–RAM

Layer Name Thickness Role in stack

Polycarbonate substrate 1.2 mm grooved substrate with address information

ZnS-SiO2 1st dielectric 140 nm thermal isolation of the substrate, also tunes/enhances
layer reflectivity
GeCrN interface 5 nm changes interface energy and nucleation rate; reduced
S diffusion
Ge6Sb2Te9 active layer 9 nm Phase-change layer
GeCrN interface 5 nm changes interface energy and nucleation rate; reduced
S diffusion
ZnS-SiO2 2nd dielectric 25 nm thermal control of cooling rate
layer
GeCr absorption 55 nm optical compensation (normalized absorption between
control amorphousand crystallization marks); also isolates ZnS
and Ag
Ag-alloy heat sink 120 nm heat sink
UV lacquer corrosion 6 µm
protection

Recording Strategy. In early optical storage systems, information was recorded by using the pulse position modulation (PPM)
method. A PPM recording strategy uses a single high-power laser pulse to write an amorphous dot. The central position of
this amorphous dot corresponds to “1” in the binary code. An inherent advantage of the PPM method is that small distortions
of the recording mark do not introduce bit errors during read-out.

In current high-density optical storage systems the so-called pulse-width modulation (PWM) or mark-edge recording
technique is applied. In this method, the lengths of the amorphous marks and of the crystalline spaces between these marks
are varied in a stepwise fashion. Now, the edges of the amorphous marks correspond to “1” in the binary code, leading to an
increased recording density with respect to PPM recording.

The PWM direct overwrite strategy of the CD-Rewritable system is given in Figure 76. Marks are written by using a train of
short, high-power laser pulses. Between these pulses the laser power is switched to a low level, so that after each pulse the
molten material is quenched into the amorphous state. The written mark thus consists of a series of overlapping amorphous
dots, its length being defined by the number of pulses in the sequence. The advantage of using a train of short pulses rather
than a single long pulse is the reduction of heat accumulation and recrystallization during writing, resulting in a constant mark
width irrespective of mark length. Direct overwrite is accomplished by switching the laser power to an intermediate d.c. erase
level between successive write pulse trains.

Figure 76. The (direct over)write strategy of the CD-Rewritable recording system. Amorphous marks are written by using a
sequence of high power laser pulses (Pwrite). The width of the first write pulse is equal to the channel clock period T,
subsequent pulses have a width of T/2. Between the write pulses the laser is switched to a low power level (Pbias) to reduce
the accumulation of heat and to quench the molten material into the amorphous state. Between the write pulse trains, an
intermediate d.c. erase power level (Perase) is used to recrystallize previously written amorphous marks.

Read-out of recorded information involves the detection of the edges of amorphous marks. One of the most critical sources
of noise is the irregularity in the definition of recorded marks, which leads to timing jitter [fluctuation of the (detected) mark
edge position relative to its nominal position]. Reduction of jitter is a rather complex issue, because the geometry of written
marks is only indirectly related to the intensity profile of the laser beam, due to the nonlinear and nonuniform thermal
processes involved. One of the sources of jitter is the accumulation of heat in the recording layer during the (over)write
process, resulting in thermal interference between an amorphous mark and its preceding and succeeding crystalline spaces.
Write strategies have been developed which correct for these thermal interference effects, e.g., by shifting the first and last
write pulse in the sequence [278] or by using a variable preheating pulse before the write pulse train and a variable cooling
gap after the last write pulse.

Interference between old and new data is the main reason for the significant increase of “overwrite jitter” relative to “write-
once” jitter. The differences in optical absorption and thermal parameters of the amorphous and crystalline phase lead to
mark shape variations in overwriting existing data. A typical four-layer stack has a low reflective amorphous phase and a high
reflective crystalline phase, resulting in a higher absorption of the amorphous phase (Aa) than that of the crystalline phase
(Ac). The reflectivity (and absorption) of the recording stack is the result of optical interference between light reflected at the
various interfaces in the stack. Absorption of light may take place in several layers, depending on the optical constants and
thickness of the employed materials. The amount of light absorbed in the phase-change layer depends on the actual
(amorphous or crystalline) state of the material at the position of the laser spot. “Overwrite” jitter can be reduced by designing
the thin-film stacks so that Ac is slightly higher than Aa, to compensate for the latent heat of melting of the crystalline phase.
Stacks with Ac > Aa can be achieved in various ways, e.g., by incorporating an absorbing Si-layer between the upper

page 54 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
dielectric and the aluminum mirror [279] or by using a semi-transparent Si-mirror instead of aluminum [280]. Another
possibility is a stack with a high reflective amorphous state and a low reflective crystalline state, so that Ac> Aa. This can be
realized by inserting a semi-transparent layer (e.g., a thin Au layer) between the substrate and the lower dielectric [281].

2.5.3. Liquid Crystals

A liquid crystal consists of rod-shaped molecules ( Liquid Crystals). Order-to-disorder transitions occur at low temperature
(250 °C). Certain types of highly ordered liquid crystals (smectic) have been developed as erasable optical storage media
[223-225]. The writing mechanism is the formation of disordered light-scattering centers in micron-sized areas by heating an
adjacent metal electrode with the laser beam. The erase mechanism for the disordered light-scattering centers is
reorientation by laser heating in the presence of an external electric field generated using adjacent a.c. electrodes. The liquid
crystal layers are ca. 8 µm thick.

The light scattering is due to a spatial variation in the refractive index. The write laser pulse times can be very short
(< 100 ns), but the formation (freezing-in) of the disordered scattering center takes much longer (ca. 5 s) compared to the
transition times for inorganic phase change alloys (see Section Phase-Change Recording Media). The reorienting erasure
also has a long relaxation time (ca. 5 s). For the nonscattering ordered texture to have few defects (low noise), the smectic
liquid crystal layers must be oriented parallel, and the liquid crystal molecules oriented perpendicular to the enclosing
substrates.

This critical alignment can be achieved using well-known methods from liquid crystal display technology. The liquid crystal
layers must be totally encapsulated because of their low viscosity (< 20 mPa · s). In spite of several difficulties, moderate
CNRs are obtained and liquid crystal media can store data at densities and data rates compatible with optical media
requirements. The kinetics of this mechanism have created difficulties with DRAW or DRAE (direct-read-after-write or erase)
capabilities and with direct overwriting of information.

All of the above difficulties (except for the kinetic aspect) have been addressed using liquid crystalline polymers [226-230]. In
these polymers, the ordered molecular groups are side groups covalently linked to a main polymeric backbone. The ordered
structures generated in such polymers are frozen-in below the Tg of the backbone polymer. Such ordered polymers are
readily deposited on substrates in micron thickness using spin coating.

These polymers also contain dye side groups covalently linked to the backbone polymer. The dye groups allow localized
laser heating, resulting in high sensitivity and contrast. Formation and reorientation of the light-scattering centers (write –
erase) takes a relatively long time (>10 s). This time is longer than for nonpolymeric liquid crystals because of backbone
rotation and steric hindrances. The use of an electric field during reordering helps to speed up erasure of the disordered
scattering center. Organic order – disorder transitions can be extended to the monomolecular regime [231]. The deposition
technique is the classical Langmuir – Blodgett method ( Thin Films).

2.5.4. Photochromics
Photochromism involves a reversible electronic level transformation induced in the writing direction by absorption of short
wavelength light and in the erasing direction by absorption of longer wavelength light (photon mode recording mechanism)
[232], [80], [233-240], [241]. The two chemical species in these bistable switches are related, but have distinct optical
properties such as absorption or refractive index. Many photochromic reactions are completed within a few nanoseconds so
that no cooling time effects occur as in the heat mode optical recording mechanisms considered so far throughout this entire
article [232]. A wide range of organic photochromics with absorption maxima in the visible region is available. The spectral
sensitivity of photochromics has not yet been extended to the technically important near-IR region.

The reverse transition (erasure) is accelerated by light or heat. This spontaneous reversal or erasure to the more stable initial
species (bleaching) or to a different species (fatigue) is the major technical obstacle to the use of organic photochromics for
optical recording. Considerable progress has been made in designing organic photochromics with improved bleaching and
fatigue properties but stability sufficient for optical recording has still to be fully demonstrated. Organic photochromics can be
extended to the monomolecular level using Langmuir – Blodgett deposition techniques [239], [240]. Other photochemically
triggered mechanisms such as induced circular dichroism and molecular changes (e.g., hydrophilic to lipophilic, trans – cis
isomerization) are also being investigated [232].

Photochromics are often very finely dispersed in polymeric binders. Reversible photochromism in polymers is rare. A soluble
organometallic polysilane derivative that exhibits thermal photochromism in the near-UV has been synthesized [242-244].
Photostimulation and luminescence of electron traps in phosphors is another type of photochromism [245].

Bistable optical organic switches can be made from other organometallics. Cu and Ag TCNQ (7,7,8,8-tetracyano-1,4-
quinodimethane) undergoes an electronic transformation (by electron transfer) to elemental copper or silver and TCNQ [246],
[247]. This redox reaction occurs in a few nanoseconds and produces high contrast marks. The reaction is driven in one
direction by absorption of light. The reaction is claimed to be driven in the reverse direction by laser heating. Reproducing the
erasure results is difficult. A composition with excess TCNQ may expedite the erasure kinetics.

2.5.5. Photochemical Hole Burning

Cryogenic (<20 K), frequency-selective, photochemical hole-burning (PHB) offers optical storage densities greater than the
diffraction limited ca. 100 Mbits/cm2 [248-256]. Various photoreactive organic molecules and inorganic color centers can be
dispersed as the carrier species in amorphous or crystalline host materials. These carrier species have inhomogeneously
broadened absorption lines, up to 103 times broader than the homogeneous absorption band of the species alone, due to the
random local environments encountered in the host matrix.

PHB writing consists of burning a hole in the broad absorption band by depleting the subset of species whose frequency-

page 55 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
shifted absorption matches the wavelength of a frequency-scanned laser. PHB read-out uses special frequency modulation
spectroscopy techniques. Erasure is performed by heating the material above 20 K. New materials are being sought that
exhibit this effect at higher temperature, e.g., near the boiling point of liquid nitrogen or even at room temperature.

Photochromics, PHB, and other photochemically- and photophysically-triggered optical recording mechanisms (photon mode
materials) have been reviewed [232], [61], [241].

2.6. Economic Aspects of Optical Recording

Today's worldwide optical disc device market has gone through drastic changes since 2000 until now (2004). In 2000 the CD-
R and CD-RW markets were expanding, DVD–ROM was starting up and now these products have already become
commodities. The Audio-CD has reached its peak and still has the biggest part of the market, followed by CD-ROM, but since
2001 the CD market has slightly decreases, due to the growth of DVD. Extensive research has occurred in the field of optical
storage and many new products have appeared on the market. The market of DVD Recordables and Rewritables (DVD+/–
R/RW and DVD–RAM), the different formats were released from 1997 to 2002, keeps on expanding without unified standard.
Furthermore the launching of blue laser products such as Blu-ray and UDO has begun. On the other hand, another Blue
Laser standard, which was introduced as AOD, is now entering the crucial stage of standardization. Today, both cutting-edge
type and commodity-type optical disc devices are making waves in the market. Various categories of devices, such as CD-
R/RW, Combo, DVD–ROM, are sold in particular markets for specific applications, and have already established strong
positions in the PC market. Therefore, the entire optical disk market has been uplifted and is constantly growing.

2.6.1. Consumer Electronics (CE)

The optical storage market for consumer applications can be divided into three types of applications: audio, video and
games.

Audio. The audio market has been the largest part of the optical storage market over the last years and has reached its
peak. Now the customers are starting to buy DVD's preferably. The decline affected virtually all major markets, with Western
Europe showing particularly sharp falls compared to recent years. Sales in Germany went down by 19% in 2003 and by more
than 30% in value since 1999. Denmark, France, Sweden, Belgium, Greece, Ireland, Portugal and Switzerland also
experienced double digit declines. Since 2000 the industry has suffered global losses of 20% year after year. In 1983, the
first year they were available, 800 000 compact discs were sold in the United States. By 1990 that number reached nearly
109 discs sold worldwide. It is estimated that more than 12 × 109 prerecorded discs were sold worldwide in 2003 (14 × 109 in
2002). The number of sold CD-R and CD-RW — also used for audio recording — is still increasing (8 × 109 in 2003) and the
peak is expected in 2005 at 9 × 109 discs. Apart from the 120 mm CD discs, also the market for recordable MD discs (a 64
mm disc in a cartridge of 72 × 68 × 5 mm3) has already reached its peak in 2002 at 220 × 106 units and is now declining.

Video. Until the DVD was released the video market has been served for a long time by the laser disc; a 300-mm analogue
video format and the VCR (video cassette recorder). The laser disc, the 12-inch optical disc format that had been available
since 1978, is obsolete since 2001. Pioneer Entertainment, the long-time champion of laser disc, abandoned laser disc
production in the U.S. in June 1999 but continued to release small runs in Japan until 2001. The laser disc still fills niches in
education, training, and video installations, but it is fading even there. Existing laser disc players and discs will be around for
a while, though essentially no new discs are being produced. There is a total of over 35 000 titles worldwide that could be
played on over 7 × 106 laser disc players. It took DVD several years to reach this level, and there are still rare titles available
on laser disc but not on DVD. Looking at the video tape: DVD player sales exceeded VCR sales in 2001 and sales of DVD
recorders are expected to reach similar levels of VCR decks in 2006. DVD recorders are expected to slowly take over for
DVD players, in stand-alone products as well as in combination products such as DVD+VCR and TV+DVD. In 2004, the first
DVD recorder combination products are entering the market. There are several DVD recorder + VCR and “Home Theater in a
Box”+ DVD recorder systems scheduled to be launched in 2004. In 2003 alone, DVD recorder unit shipments increased by
more than 200% and for 2007 the shipment of more than 50 × 106 DVD recorders is expected. DVD recorders will accelerate
the death of VCRs once the price difference is small enough. DVDs have many advantages over tapes, such as no
rewinding, quick access to any part of a recording, and fundamentally lower technology cost for hardware and disc
production. Some projections show DVD recorder sales passing VCR sales in 2005. By 2010 VHS may be as dead as vinyl
records were in 2000. DVD–Video is a huge success, in fact the most successful format ever introduced into the consumer
electronics industry. In 1997, the first year DVD–Video players became available in the USA, 349 000 players were sold,
together with 5.5 × 106 discs. DVD player unit shipments in 2003 were 98 × 106 units and will top 100 × 106 in 2004.
Worldwide production of DVD recordings exceeded 2.5 × 109 discs in 2003, nominally half the number of audio CDs sold.
Sales of DVD recorders are expected to reach 13.6 × 106 units in 2004 and 22.6 × 106 in 2005. Since these recorders can
also play back both DVD and CD recordings, there may be little need in the future for playback-only units of any kind in the
home. Automotive DVD systems and portable DVD players are expected to grow by over 30% annually for the next five
years. In addition, while households in Europe, North America, and Japan will convert to DVD recorders, markets like China
will continue to consume DVD players. The total market will then begin its decline in 2005, though the portable and
automotive segments will still continue to grow.

Games. A very fast growing, no longer niche market for Consumer Electronics is the market of CD- and DVD–based
computer games and interactive information and entertainment programs. A distinction must be made between PC games on
the one hand and console games on the other. The most popular consoles (sold units until 2003) are Sony PlayStation 2
(51.2 × 106), Nintendo GameCube (9.55 × 106) and Microsoft Xbox (9.4 × 106). According to one forecast, sales of games
will overtake Audio-CDs in Europe in 2005. The global video game market will expand from $21.2 × 109 in 2002 to
$35.8 × 109 in 2007, growing at an 11% compound annual rate.

2.6.2. Data Applications

The distinction between “Consumer Electronics” and “Data Applications” is very useful for analysis; at the same time it may

page 56 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
be very misleading. For instance: virtually any CD/DVD–ROM drive is able to play Audio-CDs or movies on DVDs. The
distinction becomes even more blurred when interactive applications are considered. The catchword "multimedia" points
directly to the observation that the PC world and the CE world, which have expanded along two unrelated tracks for a long
time, are now beginning to overlap. The currently available PCs for home use are called Mediacenter-PC and this not only
because of the multimedia extension of the Microsoft Windows operating system released in 2004.

CD-ROM. The first CD-ROM drive was already introduced in 1984, but the market was completely stagnant until advances in
CD technology reduced the drive cost sufficiently and until sufficient titles became available. At the time of writing (2004), the
CD-ROM market is as big as the CD-Audio market. The market for PC CD-ROM drives is declining because state-of-the-art
PCs are loaded with combi-drives, a combination of DVD–ROM/CD-ROM, DVD–ROM/CD-recorder or DVD–recorder/CD-
recorder. The recorders play as a matter of course the prerecorded version of the discs.

CD-R, CD-RW. In 2003 about 10 × 109 discs were sold and the market is still increasing, but the peak is expected for 2005.

5.25″ MO. The last version of this professional optical storage technology was introduced in 2000 with rewritable 9.1 GB
double-sided discs. There are many bases installed for backup purposes but it is expected that these will be replaced by
UDO solutions.

UDO is the next generation 5.25" professional optical storage technology introduced in 2003. It is a convergent technology
that delivers the performance of 5.25" MO, the longevity of 12-inch WORM, and the cost-effectiveness of DVD. UDO utilizes
blue-violet lasers and phase-change media recording technology from DVD consumer products to provide a quantum leap in
data storage densities. First generation UDO products have 30 GB capacity, future generations will increase capacity to 60
GB and 120 GB and will provide full backward-read compatibility. WORM and rewritable media are/will be available.
Forecasts expect 30 000 units sold in 2004 increasing to 60 000 units in 2006.

DVD–R(9), DVD+R(9), DVD–RW, DVD+RW, DVD–RAM. DVD–R and DVD+R are WORM media, DVD–RW and DVD+RW
rewritable media. The first recordable version of DVD was introduced in 1998. The data capacity of 4.7 GB is equivalent to
DVD–ROM. The first double-layer versions (8.5 GB) became available in 2004. The third DVD rewritable standard, DVD–
RAM, runs a distant third, but has the best defect management and reliability. This technology is incompatible with most
commercial DVD players. DVD–RAM discs are housed in cartridges and hold between 2.6 GB and 9.4 GB. The market for
recordable and rewritable DVDs is dramatically increasing. In 2003 about 1 × 109 discs were sold worldwide, for 2004 sales
of nearly 2 × 109 units are expected.

Blu-ray Disc. The Blu-ray Disc, presented in 2003, enables the recording, rewriting and play-back of up to 27 GB of data on
a single-sided single-layer 12 cm CD/DVD size disc using a 405 nm blue-violet laser. Because the Blu-ray Disc utilizes global
standard "MPEG-2 Transport Stream" compression technology, highly compatible with digital broadcasting for video
recording, a wide range of content can be recorded. In addition, the adoption of a unique ID written on a Blu-ray Disc realizes
high quality copyright protection functions. The initial market for Blu-ray Discs is expected to be for data storage. The forecast
predicts that approximately 167 400 Blu-ray devices will be sold worldwide for PC data storage in 2005, rising to 584 000 in
2006. In contrast, only 84 500 will be sold as set-top box devices in 2005, rising to 185 000 in 2006.

[Top of Page]

References
1. M. Camras: Magnetic Recording Handbook, Van Nostrand, New York 1988.
2. H. Sugaya, K. Yokoyama: “Video Recording,” in C. D. Mee, E. D. Daniel (eds.): Magnetic Recording, vol. III, Video,
Audio, and Instrumentation Recording, "Chap. 2", McGraw-Hill, New York 1988.
3. D. Jiles: Introduction to Magnetism and Magnetic Materials, Chapman and Hall, London 1991.
4. B. K. Middleton: “Recording and Reproducing Process,” in C. D. Mee, E. D. Daniel (eds.): Magnetic Recording
Technology, 2nd ed., Chap. 2, McGraw-Hill, New York 1996.
5. M. L. Williams, R. L. Comstock, AIP Conf Proc 5 (1972) 738.
6. V. A. J. Maller, B. K. Middleton, Radio Electron Eng. 44 (1974) 281.
7. J. J. Miyata, R. R. Hartel, IRE Trans. Elect. Comp. EC-8 (1959) 159.
8. D. W. Chapman, Proc. IEEE 51 (1963) 394.
9. P. I. Bonyhard, A. V. Davies, B. K. Middleton, IEEE Trans. Magn. MAG-2 (1966) 1.
10. E. D. Daniel, J. Watkinson: “Audio Recording,” in C. D. Mee, E. D. Daniel (eds.): Magnetic Recording, vol. III, Video,
Audio, and Instrumentation Recording, Chap. 3, McGraw-Hill, New York 1988.
11. E. D. Daniel, I. Levine, J. Acoust. Soc. Am. 32 (1960) 1.
12. C. D. Mee: The Physics of Magnetic Recording, North-Holland, Amsterdam 1964.
13. J. C. Mallinson: The Foundations of Magnetic Recording, 2nd ed., Academic Press, New York 1993.
14. H. Neal Bertram: Theory of Magnetic Recording, 2nd ed., Cambridge University Press, Cambridge 1994.
15. J. Earl: Cassette Tape Recorders, Fountain Press, Watford 1977.
16. E. Koster: “Particulate Media,” in C. D. Mee, E. D. Daniel (eds.): Magnetic Recording, 2nd ed., McGraw-Hill, New York
1995.
17. E. C. Stoner, E. P. Wohlfarth, Trans. Roy. Soc. A240 (1948) 599.
18. I. S. Jacobs, C. P. Bean, Phys. Rev. 100 (1955) 1060.
19. H. Inaba et al., IEEE Trans. Magn. 29 (1993) 3602.
20. J. C. Lodder: “Magnetic Properties and Preparation of Thin-Film Magnetic Recording Media,” in K. H. J. Bushow, G. J.

page 57 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Long, F. Grandjean (eds.): High Density Digital Recording, Chap. 3, NATO ASI Series, Dordrecht 1992.
21. S. R. Cumpson, B. K. Middleton, J. J. Miles, J. Magn. Mat. 154 (1996) 382.
22. B. Bhushan: “Tribology of Magnetic Head-Medium Interface,” in K. H. J. Bushow, G. J. Long, F. Grandjean (eds.): High
Density Digital Recording, Chap. 10, NATO ASI Series, Dordrecht 1992.
23. R. D. Fisher, W. H. Chilton, J. Electrochem. Soc. 109 (1962) 485. Links
24. Y. Maeda, K. Takei, IEEE Trans. Magn. 27 (1991) 4721.
25. G. Tsang, M. Chen, T. Yogi, Proc. IEEE 81 (1993) 1344.
26. S. Iwasaki, J. Hokkyo (eds.): Perpendicular Magnetic Recording, Ohmsha, Tokyo 1991.
27. J. C. Lodder: “Preparation, Microstructure and Magnetic Properties of Co – Cr Thin Films,” in K. H. J. Bushow, G. J.
Long, F. Grandjean (eds.): High-Density Digital Recording, NATO ASI Series, Dordrecht 1992.
28. R. E. Jones, C. D. Mee, C. Tsang: “Recording Heads,” in C. D. Mee, E. D. Daniel (eds.): Magnetic Recording
Technology, 2nd ed., Chap. 6, McGraw-Hill, New York 1996.
29. J. J. M. Ruigrok: Short Wavelength Magnetic Recording, Elsevier, Amsterdam 1990.
30. G. Bouwhuis et al.: Principles of Optical Disc Systems, Adam Hilger Ltd., Bristol 1985.
31. A. Marchant: Optical Recording Addison-Wesley Pub. Co., Reading MA 1990.
32. http://www.dvdforum.com (international association of hardware manufacturers, software firms and other users of
Digital Versatile Discs).
33. http://www.iecee.org/ (official website of IECEE: Worldwide System for Conformity Testing and Certification of
Electrical Equipment).
34. http://www.research.philips.com/ (Global information center of Koninklijke Philips Electronics N.V., NL).
35. http://www.blu-raydisc-official.org (official website of Blu-ray Disc founders).
36. http://www.blu-raydisc.info (Blu-ray Disc license website).
37. http://www.udo.com (official website of UDO).
38. P. Asthana, B. I. Finkelstein, A. A. Fennema, IBM J. Res. Dev. 40 (1996) 543; T. W. McDaniel, P. C. Arnett, IBM J.
Res. Dev. 40 (1996) 311.
39. M. Mansuripur: The Physical Principles of Magneto-optical Recording, Cambridge University Press, Cambridge 1995.
40. I. Akasaki et al., Electron. Lett. 32 (1996) 1105; Links S. Taniguchi et al., Electron. Lett. 32 (1996) 552. Links
41. K. Schouwhamer-Immink: Coding Techniques for Digital Recorders, Prentice Hall, Englewood Cliffs, NJ 1991; J. P. J.
Heemskerk, K. A. Schouwhamer-Immink, Philips Tech. Rev. 40 (1982) 157.
42. H. H. Hopkins, J. Opt. Soc. Am. 69 (1979) 4.
43. E. F. Stikvoort, J. A. C. van Rens, IEEE J. On Selected Areas in Communications 10 (1992) 191.
44. J. P. J. Heemskerk, Appl. Opt. 17 (1978) 2007.
45. J. A. Th. Verhoeven, Q. A. G. van Vlimmeren, B. P. Th. M. van Dijk: High density CD, an industrial approach, ODME,
Veldhoven 1995.
46. B. van Dijk, J. Nijboer: “The CD family,“ in R. K. Jörgen (ed.): Digital Consumer Electronics Handbook, McGraw-Hill,
New York in print.
47. J. Taylor: DVD Demystified, 2nd. Ed., McGraw-Hill, New York year?
48. J. Braat: On the influence of substrate thickness in optical disc read-out, in print.
49. J. P. W. B. Duchateau, B. A. J. Jacobs, Proc. Magneto-Optical Recording Int. Symp. 1996, J. Magn. Soc. Jpn. 20,
Suppl. no. S1 (1996) 193.
50. T. W. McDaniel, P. C. Arnett, “Optical Data Storage Media,” IBM J. Res. Develop 40 (1996) 311.
51. M. Kaneko, IEEE Trans. Magn. 28 (1992) 2494.
52. S. M. Mansfield, W. R. Studenmund, G. S. Kino, K. Osato, Opt. Lett. 18 (1993) 305.
53. H. J. Mamin et al. IBM. J. Res. Develop. 39 (1995) 681.
54. D. W. Pohl, IBM J. Res. Develop. 39 (1995) 701.
55. K. Yashuda et al. Jpn. J. Appl. Phys. 32, pt. 1 (1993) 5210; Y. Kasami et al. Jpn. J. Appl. Phys. 3 (1996) 423.
56. P L M. Put, G. E. van Rosmalen, J. A. H. Kahlman, Jpn. J. Appl. Phys. 36 (1997) 572.
57. T. Tanaka, S. Kawata, J. Opt. Soc. Am. A 13 (1996) 935.
58. W. E. Moerner, Jpn. J. Appl. Phys. 28 (1989) S-3, 221.
59. I. Lorgere et al., J. Opt. Soc. Am. B 13 (1996) 2229.
60. P. L. M. Put, P. A. Kraakman, R. D. Morton, J. Pasman, Proc. SPIE-Int. Soc. Opt. Eng. 1663 (1992) 336.
61. J. lsailovic: Videodisc and Optical Memory Systems, Prentice Hall Press, New York 1985.
62. K. C. Pohlmanh: The Compact Disc: A Handbook of Theory and Use, A-R Editions, Inc., Madison, Wisconsin 1989.
63. J. A. Th Verhoeven et al.: Corporate CD-ROM publishing, ODME, Veldhoven 1994.
64. H. Redlich, G. Joschko, J. Audio Eng. Soc. 35 (1987) 130.
65. J. S. Winslow, IEEE Transactions on Consumer Electronics 11 (1976) 318.
66. P. E. J. Legierse, J. H. T. Pasman, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 29 (1988) no. 2, 223.
67. J. G. Dil, C. A. Westdorp, Appl. Opt. 18 (1979) no. 18, 3198.
68. J. H. T. Pasman, J. Audio Eng. Soc. 41 (1993) no. 1/2, 19.
69. P. L. M. Put, R. D. Morton, J. J. Rush, H. P. Urbach, “The resolution limit of optical disc mastering,” Jpn. J. Apl. Phys.
36 (1997) 539.

page 58 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
70. Y. J. Huh, J. S. Kim, T. Y. Nam, S. C. Kim, Jpn. J. Apl. Phys. 36 (1997) 403.
71. Y. Mitsuhashi, Appl. Opt. 25 (1986) 4013.
72. P. E. J. Legierse, Trans. Inst. Met. Finish. 65 (1985) 13.
73. N. V. Philips, US 4 668 549, 1987 (G. J. M. Lippits, P. E. J. Legierse, J. P. Dujardin).
74. S. Ohsawa et al., Appl. Opt. 25 (1986) 4027.
75. N. V. Philips, DE 30 19 189, 1979 (E. Broeksema and A. Smeets).
76. Hitachi, EP 0 211 330 A 2, 1987 (R. Suko, H. Miwa, T. Tajima).
77. H. C. Haverkorn van Rijsewijk, P. E. Legierse, G. E. Thomas, Philips Tech. Rev. 40 (1982) 287.
78. J. Hennig, Jpn. J. Appl. Phys. 26 Suppl. 26–4 (1987) 9.
79. J. Hennig, Kunststoffe 75 (1985) 425.
80. G. Kaempf, H. Loewer, M. W. Witman, Polym. Eng. Sci. 27 (1987) 1421. Links
81. R. G. Gossink, Angew. Makromol. Chem. 145/146 (1986) 365.
82. J. Javornicki: Photoplasticity Chap. 5, Elsevier, Amsterdam 1974.
83. G. H. Werumeus-Buning, R. Wimberger-Friedl, H. Janeschitz-Kriegl, T. M. Ford in L. A. Kleintjens, P. J. Lemstra (eds.):
Integration of Fundamental Polymer Science and Technology II, Elsevier Applied Science Publishers, London 1988,
p. 405.
84. G. H. Werumeus – Buning, R. M. Gijsen, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 29 (1988) no. 2, 211.
85. H. Saito, T. Inoue, J. Polym. Sci. Polym. Phys. Ed. 25 (1987) 1629.
86. R. Wimberger-Friedl, H. Janeschitz-Kriegl in L. A. Kleintjens, P. J. Lemstra (eds.): Integration of Fundamental Polymer
Science and Technology II, Elsevier Applied Science Publishers, London 1988, p. 410.
87. S. Tokuhara, Y. Onisawa, S. Iyoshi, H. Toba, Proceedings of the Optical Society of America Meeting on Optical Data
Storage, paper FC-A 5–1, Monterey (California) 1984.
88. W. Siebourg, Kunststoffe 76 (1986) 917.
89. J. M. Newcome, Proceedings of the Society of Plastics Engineers Annual Technical Conference (ANTEC '85), (1985)
p. 773.
90. R. Riess, H. Loewer, Plast. Eng. (1985) no. 7, 27.
91. Meiki Seisakusho, DE 20 29 353, 1980 (M. Kunogi).
92. S. Anders, B. Hardt, Kunststoffe 77 (1987) 21.
93. M. Eichlseder, N. Eusemann, Kunststoffe 77 (1987) 150.
94. J. G. Kloosterboer, Adv. Polym. Sci. 84 (1988) 1. Links
95. J. G. Kloosterboer, G. J. M. Lippits, J. Radiat. Curing 11 (1984) no. 2, 10.
96. J. G. Kloosterboer, G. J. M. Lippits, H. C. Meinders, Philips Tech. Rev. 40 (1982) 298.
97. A. J. M. van den Broek, P. E. J. Legierse, Photogr. Sci. Eng. 28 (1984) 3.
98. A. G. Dirks, J. J. van den Broek, Thin Solid Films 96 (1982) 257. Links
99. J. J. Ponjée, P. N. T. van Velzen, Philips Tech. Rev. 44 (1988) 81.
100. P. Palermo, A. Korpel, G. Dickinson, W. Watson, Opt. Laser Technol. 8 (1977) 169.
101. Plast. Technol. 15 (1987) 13.
102. Hallmark, US 4 790 893, 1988 (J. B. Watkins).
103. P. de Lancle, Mix (1987) no. 12, 466. V. W. Schild, Funkschau (1987) no. 6, 71.
104. Dai Nippon Printing, JP 01 115 621 A, 1989 (N. Kazahaya), JP 01 152 044 A, 1989 (M. Tabita).
105. 3 M Company, US 4 363 844, 1982; US 4 519 065, 1985 (T. W Lewis, R. J. Anderson, D. J. Kerfeld).
106. Matsushita, JP-Kokai 63–273231, 1988 (T. Higashi).
107. Kurz L. GmbH, PT 087 372 A, 1989 (W. Reinhart).
108. L. K. H. van Beek, Photogr. Sci. Eng. 20 (1976) 88.
109. N. V. Philips, CA 1 119 451, 1982 (E. J. Spiertz, A. J. G. op het Veld).
110. N. V. Philips, US 4 125 401, 1978 (E. J. Spiertz, C. F. W. Flinsenberg, L. K. H. van Beek).
111. Thorn EMI EP 0 271 300, 1988 (A. Lang and P. Bloch).
112. W. S. Colburn, K. A. Haines, Appl. Opt. 10 (1971) 1636.
113. R. T. Ingwall, H. L. Fielding Opt. Eng. 24 (1985) 809.
114. W. J. Tomlinson, E. A. Chandross in J. N. Pitts et al. (eds.): Adv. Photochem. 12 (1982) 201.
115. H. Franke et al., Colloid Polym. Sci. 262 (1984) 213. Links
116. J. Marotz et al., Physikertagung 48 (1984) München; Physik Verhandl. 391 (1984) 1.
117. T. Luckemeyer, H. Franke, Polym. Prepr. 29 (Am. Chem. Soc. Div. Polym. Chem.) (1988) no. 2, 206.
118. V. B. Jipson, H. N. Lynt, J. F. Graczyk, Appl. Phys. Lett. 43 (1983) 27. Links
119. D. J. Gravesteijn, J. van der Veen, Philips Tech. Rev. 41 (1983/1984) 325.
120. V. B. Jipson, C. R. Jones, J. Vac. Sci. Technol. 18 (1981) 105.
121. M. Chen, V. Marello, J. Vac. Sci. Technol. 18 (1981) 75.
122. P. Kivits, B. Jacobs, P. Zalm, J. Vac. Sci. Technol. 18 (1981) 68. P. Kivits, R. de Bont, B. Jacobs, P. Zalm, Thin Solid
Films 87 (1982) 21.
123. A. H. M. Holtslag, J. Appl. Phys. 66 (1989) no. 4, 1530.

page 59 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
124. D. Y. Lou, G. M. Blom, and G. C. Kenney, J. Vac. Sci. Technol. 18 (1981) 781.
125. O. Heavens: Optical Properties of Thin Solid Films, Academic Press, New York 1955.
126. A. E. Bell, R. A. Bartolini, Appl. Phys. Lett. 34 (1979) 275. Links A. E. Bell, F. W. Spong, IEEE J. Quantum Electron.
QE 14 (1978) 487.
127. R. Glang in R. Glang and E. Massel (eds.): Handbook of Thin Film Technology, McGraw Hill, New York 1970.
128. J. J. Scheer, J. Visser, Philips Tech. Rev. 39 (1980) 246.
129. D. Meyerhofer, J. Appl. Phys. 49 (1978) 3993.
130. L. Vriens, B. A. J. Jacobs, Philips Tech. Rev. 41 (1983/1984) 313.
131. J. A. Markvoort et al., Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 134.
132. A. Saito, T. Maeda, H. Sugiyama, Y. Tsunoda, Topical Meeting on Optical Data Storage, Technical Digest Series 10,
1987, p. 64.
133. International Electrotechnical Commission Publication 68–2-38, Test Z/AD, Geneva (1974).
134. D. J. Broer, L. Vriens, Appl. Phys. 32 (1983) 107.
135. C. R. Davis, W. Y. Lee, M. Chen, H. Ito, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 260.
136. V. Novotny, L. Alexandru, J. Appl. Phys. 50 (1979) 1215. Links V. Novotny, L. Alexandru, J. Polym. Sci. 24 (1979)
1321.
137. K. Y. Law et al., Appl. Phys. Lett. 36 (1980) 884. Links
138. D. G. Howe, J. J. Wrobel, J. Vac. Sci. Technol. 18 (1981) 92. J. J. Wrobel, A. B. Marchant, D. G. Howe, Appl. Phys.
Lett. 40 (1982) 928. Links
139. J. W. Wheeler, M. C. Gupta, C. E. Kurtz, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 39.
140. K. Y. Law, G. E. Johnson, J. Appl. Phys. 54 (1983) 4799. Links
141. G. E. Johnsonm, K. Y. Law, Soc. Photo Opt.Instrum. Eng. 420 (1983) 336.
142. M. F. Molaire, Topical Meeting on Optical Data Storage, Technical Digest Series 10, Optical Society of America,
Washington 1987, p. 90.
143. D. A. Hill, D. S. Soong, J. Appl. Phys. 61 (1987) 2132. Links
144. J. Griffiths, Color and Constitution of Organic Molecules, Academic Press, London 1976.
145. M. Itoh, S. Esho, K. Nakagawa, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 332.
146. E. Hamada et al., Proc. Soc. Photo Opt. Instrum. Eng 1078 (1989) 80.
147. J. M. Drexler, J. Vac. Sci. Technol. 18 (1981) 87.
148. T. L. Lichtenstein, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 96.
149. J. Cornet, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 86.
150. M. Mashita, N. Yasuda, Proc. Soc. Photo Opt. Instrum. Eng. 329 (1981) 190.
151. N. Yasuda, M. Mori, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 319.
152. J. Feinleib, J. de Neufville, S. C. Moss, S. R. Ovshinsky, Appl. Phys. Lett. 18 (1971) 254. Links
153. T. Ohta et al., Proc. Soc. Photo Opt. Instrum. Eng. 695 (1986) 2.
154. D. J. Gravesteijn et al., Appl. Opt. 26 (1987) no. 22, 4722.
155. Y. D. Tyan et al., Topical Meeting on Optical Data Storage, Technical Digest Series 10, 1987 , p. 44.
156. R. Barton, C. R. Davis, K. Rubin, G. Lim, Appl. Phys. Lett. 48 (1987) 1255.
157. Y. Nakane, N. Sato, H. Makino, S. Miyaoka, Proc. Soc. Photo Opt. Instrum. Eng. 529 (1985) 76.
158. S. Y. Suh, Proc. Soc. Photo Opt. Instrum. Eng. 695 (1986) 16.
159. V. B. Jipson, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 344.
160. J. H. Apfel, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 104.
161. A. Werner, H. Hibst, Appl. Opt. 28 (1989) no. 7, 1422.
162. P. F. Carcia et al., J. Appl. Phys. 64 (1988) no. 4, 1671. Links
163. Du Pont, US 4 857 373, 1989 (P. F. Carcia, F. D. Kalk).
164. S. Y. Suh, Proc. Soc. Photo Opt. Instrum. Eng. 382 (1983) 196. S. Y. Suh, H. G. Craighead, R. E. Howard, L. M.
Schiavone, Proc. Soc. Photo Opt. Instrum. Eng. 382 (1983) 199.
165. K. Gardner, R. J. Longman, P. R. Helfet, Topical Meeting on Optical Data Storage, Technical Digest Series 9, Optical
Society of America, Washington 1985, papers WDD 4 and WDD 5.
166. K. Aratani et al., Proc. Soc. Photo Opt. Instrum. Eng. 1499 (1991) 209.
167. H. Awano et al. , J. Appl. Phys. 69 (1996) 4257.
168. T. Shiratori, E. Fujii, Y. Miyaoka, Y. Hozumi, J. Magn. Soc. Jpn. 22, Supp. S2 ,(1998) 47., Proc. MORIS/ISOM (1997),
to be published in Jpn. J. Appl. Phys.
169. T. Suzuki, MRS Bulletin 21 (1996) 42. Links
170. T. Yashida, Proc. IEEE 82 (1994) 1492.
171. M. Shinoda et al., J. Magn. Soc. Jpn. 22 S2 (1997) 173.
172. Y. Tanaka et al., Jpn. J. Appl. Phys. 37 (1998) 2150.
173. K. Aratani et al., Proc. SPIE 1499 (1991) 209.
174. J. Saito, M. Sato, H. Matsumoto, H. Akasaka, Jpn. J. Appl. Phys. 26 Suppl. 26–4 (1987) 155.
175. J. Saito et al., J. Magn. Soc. Jpn. 22 S2 (1997) 133.
176. T. Hosokawa, A. Okamuro, M. Hashimoto, K. Miyata, Proc. SPIE 3109 (1997) 74.

page 60 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

177. P. F. Carcia, A. D. Meinhaldt, A. Suna, Appl. Phys. Lett. 47 (1985) 178. Links
178. W. B. Zeper, F. J. M. Greidanus, P. F. Carcia, C. R. Fincher, J. Appl. Phys. 65 (1989) 4971. Links
179. J. P. Krumme et al., J. Appl. Phys. 66 (1989) 4393. Links
180. B. Bechevet et al., J. Appl. Phys. 69 (1991) 4767. Links
181. G. Q. Di, S. Iwata, S. Uchiyama, J. Magn. Mag. Mater. 131 (1994) 242.
182. J. Schoenes: “Magneto-Optical Properties of Metals, Alloys and Compounds,” in R. W. Cahn, P. Haasen, E. J. Kramer
(eds.): Materials Science and Technology, vol 3A VCH Publishers, Weinheim, Germany 1992, p. 147.
183. M. H. Kryder: “Magnetic Information Storage,” in R. Gerber, C. D. Wright and G. Asti (eds.): Applied Magnetism, Kluwer
Academic Publ., Doredrecht 1994.
184. M. H. Kryder, J. Magn. Mat. 83 (1990) 1.
185. C. D. Wright: “The Magneto-Optical Readout Process,” in T. W. McDaneil, R. H. Victora (eds.): Handbook of Magneto-
Optical Data Recording, Noyes Publications, Westwood 1997, p. 476.
186. Y. Wang et al., J. Magn. Soc. Jpn. 17 (1993) 294.
187. P. Chaudhari, J. J. Cuomo, R. J. Gambino, Appl. Phys. Lett. 22 (1973) 337. Links
188. M. H. Kryder, J. Appl. Phys. 57 (1985) 3913. Links
189. M. Hartmann et al., J. Magn. Soc. Jpn. 15 S1 (1991) 165.
190. P. Hansen: “Magnetic Amorphous Alloys,” in K. H. J. Buschow (ed.): Handbook of Magnetic Materials, vol. 6 North-
Holland, Amsterdam, 1991.
191. D. Raasch, IEEE Trans. Magn. 29 (1993) 34.
192. S. Takayama et al., J. Appl. Phys. 61 (1987) 2364.
193. D. Mergel, P. Hansen, S. Klahn, SPIE Proc. 1274 (1990) 270.
194. T. Niihara et al., IEEE Trans. Magn. 24 (1988) 2437.
195. T. K. Hatwar, J. Appl. Phys. 70 (1991) 6335. Links
196. F. E. Luborsky, IEEE Trans. Magn. 20 (1984) 1305.
197. P. Hansen, H. Heitmann, IEEE Trans. Magn. 25 (1989) 4390.
198. P. F. Carcia, J. Vac. Sci. Technol. A5 (1987) 1975.
199. P. F. Carcia, J. Appl. Phys. 63 (1988) 5066. Links
200. F. J. A. den Broeder, H. C. Donkersloot, H. J. G. Draaisma, W. J. M. de Jonge, J. Appl. Phys. 61 (1987) 4317. Links
201. H. J. G. Draaisma, W. J. M. de Jonge, F. J. A. den Broeder, J. Magn. Magn. Mat. 66 (1987) 351.
202. W. B. Zeper, F. J. A. M. Greidanus, P. F. Carcia, IEEE Trans. Magn. 25 (1989) 3764.
203. Y. Ochiai, S. Hashimoto, K. Aso, IEEE Trans. Magn. 25 (1989) 3755.
204. F. J. A. M. Greidanus et al., Jpn. J. Appl. Phys. 28 S28–3 (1989) 37.
205. H. W. van Kesteren, W. B. Zeper, J. Magn. Mag. Mater. 120 (1993) 271.
206. P. G. Pitcher, J. Miller, D. P. A. Pearson, P. D. Gurney, J. Magn. Soc. Jpn. 17 (1993) 91.
207. P. F. Carcia, M. Reilly, Appl. Phys. Lett. 58 (1991) 191. Links
208. W. B. Zeper et al., IEEE Trans. Magn. 28 (1992) 2503.
209. M. Gomi, M. Abe, Mat. Res. Soc. Symp. Proc. 150 (1989) 121.
210. K. Shono et al. MRS Conf., San Diego, paper F 3–2 (1989).
211. B. Bechevet et al., J. Appl. Phys. 69 (1991) 4767. Links
212. T. Okuda, T. Katayama, K. Satoh, H. Yamamato, J. Appl. Phys. 69 (1991) 4580. Links
213. M. H. Kryder, Thin Solid Films 216 (1992) 174. Links
214. A. Itoh, K. Nakagawa, Jpn. J. Appl. Phys. 31 (1992) L790.
215. T. Suzuki, J. Appl. Phys. 69 (1991) 4756. Links
216. K. A. Rubin et al., SPIE Proc. 2338 (1994) 247.
217. K. Tsutsumi, T. Fukami, J. Magn. Magn. Mat. 118 (1993) 213.
218. A. Takahahi et al., J. Magn. Soc. Jpn. 22 (1997) S2, 67.
219. H. Awano et al., SPIE Proc. 3109 (1997) 83.
220. D. Turnbull, Contemp. Phys. 10 (1969) 473. Links
221. D. R. Uhlmann, J. Non-Cryst. Solids 7 (1972) 337. Links
222. T. Minewura, H. Andoh, Y. Maeda, J. Appl. Phys. 63 (1988) no. 9, 4632. Hoechst, US 4 478 782, 1984 (A. J. East, J. E.
Kuder).
223. Hewlett Packard Company, US 4 405 993, 1983 (H. Birecki, R. A. Burmeister, F. J. Kahn).
224. H. Birecki, S. Naberhuis, F. J. Kahn, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 194.
225. J. Pereda, Proc. Soc. Photo Opt. Instrum. Eng. 369 (1982) 284.
226. Röhm GmbH, EP-A 231–857 A-858 A; DE 3603–267 A, 1987 (H. Eich, J. Wendorff, B. Reck).
227. H. W. Schmidt: Ph. D. Thesis, University of Mainz 1985.
228. W. D. McIntyre, D. S. Soane, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 29 (1988) no. 2, 197.
229. I. Cabrera et al., Polym . Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 29 (1988) no. 2, 200.
230. H. E. Ringsdorf, Makromol. Chem. 183 (1982) 1245. V. P Shibaev et al., Polym. Commun. 24 (1983) no. 12, 364.
231. Canon KK, JP-Kokai 62149–482 A, JP - 483 A-484 A, JP - 487 A, JP 62151–384 A, JP -385 A, 1987 (H. Matsuda et

page 61 of 62
Information Storage Materials : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

al.).
232. S, Tazuke, Jpn. J. Appl. Phys. 26, Suppl. 26–4 (1987) 3.
233. Plessey, Ltd., GB 1 464 603, 1977 (H. G. Heller);Traqson Ltd., WO 88/0 95 49, 1988 (S. C. Wallace).
234. Japan Agency of Science and Technology, JP-Kokai 62209–186 A, 1987 (Y. Suzuki, K. Ichimura, M. Sakuragi).
235. W. Tomlinson, Appl. Opt. 23 (1984) no. 22, 3990.
236. L. M. Ralston, Proc. Soc. Photo Opt. Instrum. Eng. 420 (1983) 186.
237. G. H. Brown (ed.): Techniques in Chemistry III: Photochromism, Wiley-Interscience, New York 1971.
238. M. Irie, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 29 (1988) no. 2, 215.
239. C. Ishimoto et al., Appl . Phys. Lett. 49 (1986) no. 25, 1677.
240. Sony, EP 0 218 263, 1987 (C. Ishimoto, J. Seto, H. Tomimuro).
241. J. M. Pearson, CRC Crit. Rev. Solid State Mater. Sci. 13 (1986) no. 1, 1.
242. R. West, J. Organomet. Chem. 300 (1986) 327. Links
243. R. Miller, J. Am. Chem. Soc. 107 (1985) 2172. Links
244. R. Miller, Macromolecules 19 (1986) 611.
245. Quantex Company, US-A 870 877, 1986 (J. A. Lindmayer). Laser Focus World 25 (1989) no. 11, 119
246. Johns Hopkins University, US 4 371 883, 1983 (D. O. Cowan, T. O. Poehler, R. S. Potember).
247. R. C. Hoffman, R. S. Potember, Appl. Opt. 28 (1989) no. 7, 1417.
248. IBM Corporation, US 4 101 976, 1978 (G. Castro, D. Haarer, R. M. McFarlane, H. P. Trommsdorff).
249. W. E. Moerner, T. P. Carter, C. Brauchle, Appl. Phys. Lett. 50 (1987) no. 8, 430. Links
250. W. E. Moerner, R. M. McFarlane, W. Lenth, Topical Meeting on Optical Data Storage, Technical Digest Series 10,
Optical Society of America, Washington 1987, p. 151.
251. M. Yoshimura et al., Proc. Soc. Photo Opt. Instrum. Eng. 1078 (1989) 326.
252. W. Lenth et al., Proc. Soc. Photo Opt. Instrum. Eng. 695 (1986) 216.
253. T. Tani et al., Jpn. J. Appl. Phys. 26, Suppl. 26–4, (1987) 77.
254. D. Haarer, Jpn. J. Appl. Phys. 26, Suppl. 26–4, (1987) 227.
255. U. P. Wild et al., Jpn. J. Appl. Phys. 26, Suppl. 26–4, (1987) 233.
256. U. Bogner, K. Beck, P. Schatz, M. Maler, Topical Meeting on Optical Data Storage, Technical Digest Series 9, Optical
Society of America, Washington (1985) paper WBB 2.
257. J. Feinleib, S. R. Ovshinsky, J. Non-Cryst. Solids 2 (1970) 564.
258. M. Chen, K. Rubin, R. W. Barton, Appl. Phys. Lett. 49 (1986) 502. Links
259. S. R. Ovshinsky, J. Non-Cryst. Solids 141 (1992) 200. Links
260. N. Yamada, Mater. Res. Soc. Bulletin, (1996) Sept., 48.
261. A. W. Smith, Appl. Opt. 13 (1974) 795.
262. M. Terao et al., Proc. Soc. Photo Opt. Instrum. Eng. 695 (1986) 106.
263. T. Nishida et al., Appl. Phys. Lett. 50 (1987) 667. Links
264. J. H. Coombs, A P. J. M. Jongenelis, W. van Es-Spiekman, B. A. J. Jacobs, J. Appl. Phys. 78 (1995) 4918. Links
265. N. Yamada et al., Jpn. J. Appl. Phys. 26 Suppl. 26–4 ,(1987) 61.
266. N. Yamada, E. Ohno, K. Nishiuchi, N. Akahira, J. Appl. Phys. 69 (1991) 2849. Links
267. J. Gonzales-Hernandez et al., Appl. Phys. Commun. 11 (1992) 557.
268. K. A. Agaev, A. G. Talybov, Kristallografiya 11 (1966) 454.
269. I. I. Petrov, R. M. Imamov, Z. G. Pinsker, Kristallografiya 13 (1968) 417.
270. T. Ohta et al., Jpn. J. Appl. Phys. 28 Suppl. 28–3, (1989) 123.
271. M. Terao et al. Proc. Soc. Photo Opt. Instrum. Eng. 1078 (1989) 2.
272. H. Iwasaki et al., Jpn. J. Appl. Phys. 31 (1992) 461.
273. H. Iwasaki et al., Jpn. J. Appl. Phys. 32 (1993) 5241.
274. Y. Kageyama, H. Iwasaki, M. Harigaya, Y. Ide, Jpn. J. Appl. Phys. 35 (1996) 500.
275. T. Handa, J. Tominaga, S. Haratani, S. Takayama, Jpn. J. Appl. Phys. 32 (1993) 5226.
276. H. J. Borg, J. P. W. B. Duchateau, Proc. Soc. Photo Opt. Instrum. Eng. 3109 (1997) 20.
277. S. Morita, M. Nishiyama, T. Ueda, Jpn. J. Appl. Phys. 36 (1997) 444.
278. T. Ishida et al., Proc. Soc. Photo Opt. Instrum. Eng. 2338 (1994) 121.
279. M. Tero et al., Proc. Soc. Photo Opt. Instrum. Eng. 3109 (1997) 60.
280. M. Okada et al., Proc. Soc. Photo Opt. Instrum. Eng. 2514 (1995) 329.
281. H. Kobori et al., Proc. Soc. Photo Opt. Instrum. Eng. 2514 (1995) 127.
[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 62 of 62
 Infrared and Raman Spectroscopy 1

Infrared and Raman Spectroscopy

Hans-Ulrich Gremlich, Novartis AG, Basel, Switzerland

1. Introduction . . . . . . . . . . . . . . 2 4.8.8. Carbonate Salts . . . . . . . . . . . . . 24

2. Techniques . . . . . . . . . . . . . . . 3 4.8.9. Amides . . . . . . . . . . . . . . . . . . 24
2.1. Fourier Transform Infrared Tech- 4.8.10. Lactams . . . . . . . . . . . . . . . . . . 24
nique . . . . . . . . . . . . . . . . . . . 3 4.9. Sulfur-Containing Compounds . . 25
2.2. Dispersive and Fourier Transform 4.9.1. Thiols . . . . . . . . . . . . . . . . . . . 25
Raman Techniques . . . . . . . . . . 5 4.9.2. Sulfides and Disulfides . . . . . . . . 25
3. Basic Principles of Vibrational 4.9.3. Sulfones . . . . . . . . . . . . . . . . . . 26
Spectroscopy . . . . . . . . . . . . . . 7 4.10. Computer-Aided Spectral Inter-
3.1. The Vibrational Spectrum . . . . . 7 pretation . . . . . . . . . . . . . . . . . 26
3.2. Quantitative Analysis . . . . . . . . 8 5. Applications of Vibrational Spec-
3.3. The Symmetry of Molecules and troscopy . . . . . . . . . . . . . . . . . 27
Molecular Vibrations; Selection 5.1. Infrared Spectroscopy . . . . . . . . 27
Rules . . . . . . . . . . . . . . . . . . . 10 5.1.1. Transmission Spectroscopy . . . . . . 27
4. Interpretation of Infrared and Ra- 5.1.2. External Reflection Spectroscopy . . 28
man Spectra of Organic 5.1.3. Internal Reflection Spectroscopy . . 30
Compounds . . . . . . . . . . . . . . . 11 5.1.4. Diffuse Reflection Spectroscopy . . 31
4.1. The Concept of Group Frequencies 11
5.1.5. Emission Spectroscopy . . . . . . . . 33
4.2. Methyl and Methylene Groups . . 12
5.1.6. Photoacoustic Spectroscopy . . . . . 33
4.3. Alkene Groups . . . . . . . . . . . . . 15
5.1.7. Chromatography/Fourier Transform
4.4. Aromatic Rings . . . . . . . . . . . . 15
Infrared Spectroscopy . . . . . . . . . 34
4.5. Triple Bonds and Cumulated Dou-
5.1.8. Thermogravimetry/Fourier Trans-
ble Bonds . . . . . . . . . . . . . . . . 16
form Infrared Spectroscopy . . . . . . 36
4.6. Ethers, Alcohols, and Phenols . . . 18
5.1.9. Vibrational Circular Dichroism . . . 36
4.6.1. Ethers . . . . . . . . . . . . . . . . . . . 18
4.6.2. Alcohols and Phenols . . . . . . . . . 19 5.2. Raman Spectroscopy . . . . . . . . . 37
4.7. Amines, Azo, and Nitro 5.3. Fourier Transform Infrared and
Compounds . . . . . . . . . . . . . . . 19 Raman Microspectroscopy . . . . . 37
4.7.1. Amines . . . . . . . . . . . . . . . . . . 19 5.4. Time-Resolved FT-IR and FT-
4.7.2. Azo Compounds . . . . . . . . . . . . 20 Raman Spectroscopy . . . . . . . . . 39
4.7.3. Nitro Compounds . . . . . . . . . . . . 20 5.5. Vibrational Spectroscopic Imaging 39
4.8. Carbonyl Compounds . . . . . . . . 20 5.6. Infrared and Raman Spectroscopy
4.8.1. Ketones . . . . . . . . . . . . . . . . . . 20 of Polymers . . . . . . . . . . . . . . . 40
4.8.2. Aldehydes . . . . . . . . . . . . . . . . 21 6. Near-Infrared Spectroscopy . . . . 40
4.8.3. Carboxylic Acids . . . . . . . . . . . . 21 6.1. Comparison of Mid-Infrared and
4.8.4. Carboxylate Salts . . . . . . . . . . . . 23 Near-Infrared Spectroscopy . . . . 40
4.8.5. Anhydrides . . . . . . . . . . . . . . . . 23 6.2. Applications of Near-Infrared
4.8.6. Esters . . . . . . . . . . . . . . . . . . . 23 Spectroscopy . . . . . . . . . . . . . . 41
4.8.7. Lactones . . . . . . . . . . . . . . . . . 24 7. References . . . . . . . . . . . . . . . . 42

1. Introduction tant analytical tool in research and in technical

The roots of infrared spectroscopy go back to
the year 1800, when William Herschel disco-
vered the infrared region of the electromag-
netic spectrum. Since 1905, when William W.
Coblentz ran the first infrared spectrum [1],
vibrational spectroscopy has become an impor-
fields. In the late 1960s, infrared spectrometry
was generally believed to be a dying instrumen-
tal technique that was being superseded by nu-
clear magnetic resonance and mass spectrome-
try for structural determination, and by gas and
liquid chromatography for quantitative analysis.

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.b05 429
2 Infrared and Raman Spectroscopy
However, the appearance of the first research-
grade Fourier transform infrared (FT-IR) spec-
trometers in the early 1970s initiated a renais-
sance of infrared spectrometry. After analytical
instruments (since the late 1970s) and routine
instruments (since the mid 1980s), dedicated in-
struments are now available at reasonable prices.
With its fundamental multiplex or Fellgett’s ad-
vantage and throughput or Jacquinot’s advan-
tage, FT-IR offers a versatility of approach to
measurement problems often superior to other
techniques. Furthermore, FT-IR is capable of ex-
tracting from samples information that is diffi-
cult to obtain or even inaccessible for nuclear
magnetic resonance and mass spectrometry. Ap-
plications of modern FT-IR spectrometry in-
clude simple, routine identity and purity exami-
nations (quality control) as well as quantitative
analysis, process measurements, the identifica-
tion of unknown compounds, and the investiga-
tion of biological materials.
Raman and infrared spectra give images of
molecular vibrations which complement each
other; i.e., the combined evaluation of both
spectra yields more information about molec-
ular structure than when they are evaluated
separately. The 1990s witnessed a tremendous
growth in the capabilities and utilization of
Raman scattering measurements. The parallel
growth of Fourier transform Raman (FT-Raman)
with charge-coupled device (CCD) based dis-
persive instrumentation has provided the spec-
troscopist with a wide range of choice in instru-
mentation. There are clear strengths and weak-
nesses that accompany both types of instrumen-
tation. While FT-Raman systems virtually elimi-
nate fluorescence by using near-IR laser sources,
this advantage is paid for with limited sensitiv-
ity. On the other hand, fluorescence interferences
always play a role in dispersive measurements
with higher sensitivity.
As an alternative to infrared spectroscopy,
Raman spectroscopy can be easier to use in some
cases; for example, whereas water and glass are
strong infrared absorbers they are weak Raman
scatterers, so that it is easy to produce a good-
quality Raman spectrum of an aqueous sample
in a glass container.
General References. Details of vibrational
theory are given in the books by Herzberg [2],
[3]; Wilson, Decius, and Cross [4]; Steele
[5]; Koningstein [6]; and Long [7]. The tech-
nique and applications of FT-IR spectroscopy
are discussed in detail by Griffiths and de
Haseth [8], and by Mackenzie [9]. An in-
troduction to infrared and Raman spectroscopy,
dealing with theoretical as well as experimen-
tal and interpretational aspects, is provided by
Colthup, Daly, and Wiberley [10] and by
Schrader [11]. Books about Raman spec-
troscopy in particular are those of Kohlrausch
[12]; Baranska, Labudzinska, and Terpinski
[13]; and Hendra, Jones, and Warnes [14].
Practical FT-IR and Raman spectroscopy is co-
vered by Ferraro and Krishnan [15]; Pel-
letier [16]; and Grasselli and Bulkin [17].
For the interpretation of IR and Raman spectra
of organic compounds, see Socrates [18]; Bel-
lamy [19], [20]; Dollish, Fateley, and Bent-
ley [21], and Lin-Vien, Colthup, Fateley,
and Grasselli [22]. Vibrational spectroscopy
of inorganic and coordination compounds is
described by Nakamoto [23]. Biological ap-
plications of IR and Raman spectroscopy are
covered by Gremlich and Yan [24]. Various
collections of vibrational spectra are available.
Both Raman and IR spectra of about 1000 or-
ganic compounds are offered as an atlas by
Schrader [25]. The most comprehensive refer-
ence-spectra libraries, digital and hardcopy, can
be obtained from Sadtler [26]. A book on mod-
ern techniques in applied optical spectroscopy
was edited by Mirabella [27].
2. Techniques
2.1. Fourier Transform Infrared
Technique
The basis of Fourier transform infrared (FT-
IR) spectroscopy is the two-beam interferom-
eter, designed by Michelson in 1891 [28], and
shown schematically in Figure 1A. Broadband
infrared radiation is emitted by a thermal source
(globar, metal strips, Nernst glower) and falls
onto a beam splitter which, in the ideal case,
transmits half the radiation and reflects the other
half. The reflected half, after traversing a dis-
tance L, falls onto a fixed mirror M1. The ra-
diation is reflected by M1 and, after travers-
ing back along distance L, falls onto the beam
 Infrared and Raman Spectroscopy 3

splitter again. The transmitted radiation follows the apodization function used in computing the
a similar path and also traverses distance L; spectrum.
however, the mirror M2 of the interferometer
can be moved very precisely along the optical
axis by an additional distance x. Hence, the to-
tal path length of the transmitted radiation is
2 (L + x). On recombination at the beam split-
ter the two beams possess an optical path dif-
ference of ∆ = 2 x. Since they are spatially co-
herent, the two beams interfere on recombina-
tion. The beam, modulated by movement of the
mirror, leaves the interferometer, passes through
the sample cell and is finally focussed on the
detector. The signal registered by the detector,
the interferogram, is thus the radiation intensity
I (x), measured as a function of the displacement
x of the moving mirror M2 from the distance
L (Fig. 1B). The mathematical transformation,
a Fourier transform [8], of the interferogram,
performed by computer, initially provides a so-
called single-beam spectrum. This is compared
with a reference spectrum measured without the
sample to obtain a spectrum analogous to that
measured by conventional dispersive methods.
This spectrum is stored digitally in the com-
puter, from which it can be retrieved for further
use (Fig. 1C). In addition to Michelson interfer-
ometers, which often have air bearings for the
moving mirror and thus require a source of dry Figure 1. A) Schematic diagram of a Michelson interferom-
eter; B) Signal registered by the detector D, the interfero-
compressed air, mechanical interferometer con- gram; C) Spectrum obtained by Fourier transform (FT) of
cepts such as frictionless electromagnetic drive the interferogram
mechanisms or refractively scanned interferom- S = Radiation source; Sa = Sample cell; D = Detector;
eters have been developed by instrument manu- A = Amplifier; M1 = Fixed mirror; M2 = Movable mirror;
BS = Beam splitter; x = Mirror displacement
facturers.
Two major kinds of detector are used in mid- The first point is derived from the Rayleigh
infrared FT-IR spectrometers, the deuterated criterion of resolution, which states that to re-
triglycine sulfate (DTGS) pyroelectric detector solve two spectral lines separated by a distance
and the mercury cadmium telluride (MCT) pho- d, the interferogram must be measured to an op-
todetector. The DTGS type, which operates at tical path length of at least 1/d [29].
room temperature and has a wide frequency The second factor is related to the fact that in
range, is the most popular detector for FT-IR real life the interferogram is truncated at finite
instruments. The MCT detector responds more optical path difference. In addition, in the fast
quickly and is more sensitive than the DTGS Fourier transform (FFT) algorithm, according to
detector, but it operates at liquid-nitrogen tem- Cooley and Tukey [30], which is used to per-
perature and has both a limited frequency range form the Fourier transform faster than the clas-
and a limited dynamic range; therefore, it is used sical method, certain assumptions and simplifi-
only for special applications [8]. cations are made. The result is that the FFT of
The resolution of an interferometer depends a monochromatic source is not an infinitely nar-
mainly on two factors: firstly, the distance x the row line, corresponding to infinite optical path
moving mirror travels or the maximum optical difference and hence to infinite resolution (see
path difference ∆ = 2 x (Fig. 1), and secondly, above), but a definable shape known as instru-
4 Infrared and Raman Spectroscopy
mental line shape (ILS) function [8]. The ILS
of a perfectly aligned Michelson interferome-
ter has the shape of a (sin x)/x or sinc x func-
tion (Fig. 2), exhibiting strong side lobes or feet,
which may even have negative intensities. To
attenuate these spurious feet, so-called apodiza-
tion functions are used as weighting functions
for the interferogram. Numerous such functions
exist (e.g., triangular, Happ – Genzel, or Black-
man – Harris) which produce instrumental line
shapes with lower intensity side lobes, but also
with broader main lobes than the sinc function.
Hence, side-lobe suppression is only possible at
the cost of resolution, because the full width at
half-height of the ILS defines the best resolu-
tion achievable with a given apodization func-
tion [31]. In contrast to dispersive spectroscopy,
where the triangular apodization function is de-
termined by the slit of the grating spectrom-
eter, in FT-IR spectroscopy, a choice between
highest resolution (no apodization) and best side
lobe suppression (Blackman – Harris apodiza-
tion function) is possible.
Figure 2. The sinc x function: instrumental line shape func-
tion of a perfectly aligned Michelson interferometer with no
apodization
ν̃ = wavenumber, cm−1 ; ∆ = maximum retardation, cm
In comparison with conventional spec-
troscopy, the FT-IR method possesses significant
advantages. In grating spectrometers, the spec-
trum is measured directly by registering the radi-
ation intensity as a function of the continuously
changing wavelength given by a monochromator
[10]. Depending on the spectral resolution cho-
sen, only a very small part (in practice less than
0.1 %) of the radiation present in the monochro-
mator reaches the IR detector. In the FT-IR
spectrometer, all frequencies emitted by the IR
source reach the detector simultaneously, which
results in considerable time saving and a large
signal-to-noise ratio advantage over dispersive
instruments; this is the most important advan-
tage, known as the multiplex or Fellgett’s ad-
vantage [8]. A further advantage stems from
the fact that the circular apertures employed in
FT-IR spectrometers have larger surface areas
than the linear slits of grating spectrometers and
hence permit radiation throughputs at least six
times greater; this is known as the throughput
or Jacquinot’s advantage [8]. In FT-IR spec-
troscopy, the measuring time is the time required
to move the mirror M2 the distance necessary
to give the desired resolution; since the mirror
can be moved very rapidly, complete spectra can
be obtained within fractions of a second. By
contrast, the measuring time for a conventional
spectrum is on the order of minutes. In an FT
spectrum, the accuracy of each wavenumber is
coupled to the accuracy with which the position
of the moving mirror is determined; by using
an auxiliary HeNe laser interferometer the posi-
tion of the mirror can be determined to an accu-
racy better than 0.005 µm. This means that the
wavenumbers of an FT-IR spectrum can be de-
termined with high accuracy (< 0.01 cm−1 ). In
other words, FT-IR spectrometers have a built-in
wavenumber calibration of high precision; this
advantage is known as accuracy or Connes’ ad-
vantage [29]. As a result, it is possible to carry
out highly precise spectral subtractions and thus
very reliably detect slight spectral differences
between samples.
2.2. Dispersive and Fourier Transform
Raman Techniques
In modern dispersive Raman spectrometers, ar-
rays of detector elements such as photodiode
arrays or charge-coupled devices (CCDs) are
arranged in a polychromator. Here, each ele-
ment records a different spectral band at the
same time, thus making use of the multichannel
advantage. In dispersive instruments, the very
strong radiation at and near the exciting line, the
Rayleigh radiation, may produce stray radiation
in the entire spectrum with an intensity much
higher than that of the Raman lines. In order
to remove this unwanted radiation, line filters
 Infrared and Raman Spectroscopy
in the form of so-called notch filters or holo-
graphic filters are used which specifically reflect
the Rayleigh line. In conventional, dispersive
Raman spectroscopy visible lasers are usually
used to excite the Raman effect (see Section 3.1),
while in FT-Raman spectroscopy near-IR lasers
are employed. Excitation with visible lasers of-
ten produces strongly interfering fluorescence
originating from the sample or impurities; how-
ever, this is eliminated by using a neodymium
yttrium aluminum garnet (Nd : YAG) laser pro-
viding a discrete line excitation at 1.064 µm, as
at this wavelength in the near-IR region no elec-
tronic transitions are induced.
An FT-Raman spectrometer is shown in sche-
matic form in Figure 3A. The beam of the laser
source passes to the sample compartment where
it is focused, typically to a spot of 100 µm di-
ameter, onto the sample, either in 90◦ or 180◦
geometry (Fig. 3B). The scattered Raman light
is collected by an aspherical lens and passed
through a filter module to remove the Rayleigh
line and Rayleigh wings [32]. It is then directed
through the infrared input port into the interfer-
ometer which is optimized for the near-IR region
as is the liquid-nitrogen-cooled germanium or
room-temperature indium gallium arsenide (In-
GaAs) detector [33]. The procedure to obtain the
FT-Raman spectrum from the interferogram, the
signal registered by the detector, is essentially
the same as that described above for an FT-IR
spectrum.
Before FT-Raman spectrometers could be
successfully used several difficulties had to be
overcome. As the intensity of Raman scattering
is proportional to the fourth power of the fre-
quency of excitation a loss in sensitivity by a
factor between 20 and 90 must be compensated
for when the exciting radiation is shifted from
the visible to the near-IR region [32]. Moreover,
the noise equivalent power (NEP) of near-IR de-
tectors is greater by about one to two orders of
magnitude than that of a photomultiplier tube,
the usual detector for conventional Raman spec-
troscopy. However, both disadvantages are com-
pensated for by the throughput (Jacquinot) and
multiplex (Fellgett) advantages of Fourier trans-
form spectroscopy (see Section 2.1). Another
major problem is the fact that Raman scatter-
ing is always accompanied by very strong ra-
diation, the Rayleigh line and Rayleigh wings,
occurring at and near the frequency of excitation
5
with six to ten orders of magnitude higher inten-
sity than the Raman signal. The shot noise of
the Rayleigh line and Rayleigh wings produces
a background noise which would completely
mask all Raman lines. Therefore, an effective fil-
ter that removes the Rayleigh line and Rayleigh
wings in the range 750 – 100 cm−1 is one of the
most important components of a powerful FT-
Raman instrument [32]. In addition, optimiza-
tion of the sampling technique is indispensable
in order to convert the flux of exciting radiation
into a maximum flux of Raman radiation to be
received by the optical system of the spectrom-
eter [34]. The 180◦ backscattering arrangement
(Fig. 3B), rather than the 90◦ arrangement used
in visible, dispersive Raman technique, is un-
doubtedly the most efficient way of collecting
the Raman signal and provides the best quality
data in the near-IR FT-Raman experiment.
3. Basic Principles of Vibrational
Spectroscopy
3.1. The Vibrational Spectrum
To explain the origins of a vibrational spectrum
the vibration of a diatomic molecule is consid-
ered first. This can be illustrated by a molecular
model in which the atomic nuclei are represented
by two point masses m1 and m2 , and the inter-
atomic bond by a massless spring. According
to Hooke’s law, the vibrational frequency ν (in
s−1 ) determined by classical methods is given
by


1 f
ν= (1)
2π µ
where f is the force constant of the spring in
N/m, and µ is the reduced mass in kg:
m1 m2
µ= (2)
m1 +m2
Thus, the vibrational frequency is higher
when the force constant is higher, i.e., when
the bonding of the two atoms is stronger. Con-
versely, the heavier the vibrating masses are, the
smaller is the frequency ν. The frequency of fun-
damental vibrations is on the order of magnitude
of 1013 s−1 . Quantum-mechanical methods give
6 Infrared and Raman Spectroscopy

Figure 3. A) Schematic diagram of an FT-Raman spectrometer (Bruker RFS 100); B) Schematic diagram of the sample
compartment
L = laser; RA = Raman; SA = Sample
(Reproduced by permission of Bruker Optik GmbH, D-76275 Ettlingen, Germany)

the following expression for the vibrational en- transitions are always accompanied by rota-
ergy of a Hooke’s oscillator in a state character- tional transitions. The corresponding energies
ized by the vibrational quantum number v: are given by:

 
    1
h f 1 1 Evib,rot =hν0 v+ +BhJ (J+1) (4)
Evib = v+ =hν0 v+ (3) 2
2π µ 2 2
where J is the rotational quantum number and B
where f and µ have the same definition as in is the rotational constant:
Equation (1) and ν 0 is the vibrational frequency
of the ground state. This relation is valid with h
B= (5)
good accuracy for vibrational transitions from 8π 2 I
the ground state (v = 0) to the first excited state where I is the moment of inertia. In the real
(v = 1). In free molecules (gas state), vibrational world, molecular oscillations are inharmonic.

Therefore, the transition energy decreases with
increasing vibrational quantum number until
the molecule finally dissociates. The quantum
mechanical energy of a diatomic inharmonic os-
cillator is given in good approximation by:
   2 
1 1
Evib =hν0 v+ −x0 v+
2 2
where x 0 is the inharmonicity constant.
A complex molecule is considered as a sys-
tem of coupled inharmonic oscillators. If there
are N atomic nuclei in the molecule, there will be
a total of 3 N degrees of freedom of motion for all
the nuclear masses in the molecule. Subtracting
the pure translations and rotations of the entire
molecule leaves 3 N − 6 internal degrees of free-
dom for a nonlinear molecule and 3 N − 5 inter-
nal degrees of freedom for a linear molecule. The
internal degrees of freedom correspond to the
number of independent normal modes of vibra-
tion; their forms and frequencies must be calcu-
lated mathematically [10]. In each normal mode
of vibration all the atoms of the molecule vibrate
with the same frequency and pass through their
equilibrium positions simultaneously. The true
internal motions of the molecule, which consti-
tute its vibrational spectrum, are composed of
the normal vibrations as a coupled system of
these independent inharmonic oscillators. Thus,
a molecule is unambiguously characterized by
its vibrational spectrum. Real-world molecules
are best described by the model of inharmonic
oscillators because here transitions are allowed
which are forbidden to the harmonic oscilla-
tor: transitions from the ground state (v = 0) to
states with v = 2, 3, . . . (overtones), transitions
from an excited level of a vibration (difference
bands), and transitions between states belonging
to different types of normal mode (combination
bands).
In vibrational spectroscopy, instead of the fre-
quency ν (in s−1 ) the so-called wavenumber ν̃
expressed in cm−1 (reciprocal centimeters) is
mostly used; this signifies the number of waves
in a 1 cm wavetrain:
ν 1
ν̃= = (7)
(c/n) λ
where c is the velocity of light in a vacuum
(2.997925 × 1010 cm/s), c/n is the velocity of
light in a medium with refractive index n in
Infrared and Raman Spectroscopy 7
which the wavenumber is measured (the refrac-
tive index of air is 1.0003), and λ is the wave-
length in centimeters. In the infrared region of
the electromagnetic spectrum the practical unit
of wavelength is 10−4 cm or 10−6 m (i.e., µm),
and wavenumber and wavelength are related as
follows:
(6)
104
ν̃/cm−1 = (8)
λ/µm
In infrared spectroscopy, both wavenumber
and wavelength are used, whereas in Raman
spectroscopy only wavenumber is employed.
The infrared and microwave regions of the
electromagnetic spectrum are shown in Table 1.
In the near-infrared region, overtones and com-
bination bands are observed. Since these transi-
tions are only possible because of inharmonicity,
their probability is diminished. A good rule of
thumb is that the intensity of the first overtone
is about one-tenth that of the corresponding fun-
damental vibration. The fundamentals occur in
the mid-infrared range, which is the most im-
portant infrared region. In the far-infrared re-
gion, the fundamentals of heavy, single-bonded
atoms and the absorptions of inorganic coordi-
nation compounds [23] are found.
Infrared spectroscopy is based on the interac-
tion of an oscillating electromagnetic field with
a molecule, and it is only possible if the dipole
moment of the molecule changes as a result of a
molecular vibration. While the absorption fre-
quency depends on the molecular vibrational
frequency, the absorption intensity depends on
how effectively the infrared photon energy is
transferred to the molecule. It can be shown [3]
that the intensity of infrared absorption is pro-
portional to the square of the change in the dipole
moment µ, with respect to the change in the nor-
mal coordinate q describing the corresponding
molecular vibration:
 2
IIR ∼
∂µ
(9)
∂q
A second way to induce molecular vibrations
is to irradiate a sample with an intense source
of monochromatic radiation, usually in the visi-
ble or near-infrared region, this leads to the Ra-
man effect, which can be regarded as an inelastic
collision between the incident photon and the
molecule. As a result of the collision, the vi-
brational or rotational energy of the molecule
8 Infrared and Raman Spectroscopy
Table 1. Infrared and microwave regions of the electromagnetic spectrum
Region Wavelength λ, µm
Infrared
near 7.8×10−1 to 2.5
mid 2.5 to 5.0×101
far 5.0×101 to 1.0×103
Microwave 1.0×103 to 1.0×106
is changed by an amount ∆E m . For energy to
be conserved, the energy of the scattered photon
h ν s must differ from that of the incident photon
h ν i by an amount equal to ∆E m :
hνi −hνs = ∆Em =hνm
If the molecule gains energy, then ∆E m is
positive and ν s is smaller than ν i , giving rise to
so-called Stokes lines in the Raman spectrum. If
the molecule loses energy, ∆E m is negative and
ν s is larger than ν i , giving rise to so-called anti-
Stokes lines in the Raman spectrum. The dipole
moment µ induced by the Raman effect can be
related to the electric field E of the incident elec-
tromagnetic radiation as follows:
µ=αE
where α is the polarizability of the molecule. It
can be derived from classical Raman theory [3]
that for a molecular vibration to be Raman ac-
tive it must be accompanied by a change in the
polarizability of the molecule. The intensity of
Raman absorption is proportional to the square
of the change in polarizability α, with respect to
the corresponding normal coordinate q:
 2
∂α
IRA ∼
∂q
Although it follows from classical theory that
Stokes and anti-Stokes lines should have the
same intensity, according to quantum theory
and in agreement with experiment (Fig. 4) anti-
Stokes lines are a much less intense since the
number of molecules in the initial state v = 1 of
anti-Stokes lines is only e−(hν m /kT ) times the
number of molecules in the initial state v = 0 of
the Stokes lines [3]. The Raman shift in cm−1 ,
i.e., the difference ∆ν̃ between the wavenumber
ν̃ e of the exciting laser and the wavenumber ν̃ s
of the scattered Raman light, is indicated on the
abscissa of each Raman spectrum.
Wavenumber ν̃, cm−1 Frequency ν, s−1
12 800 to 4000 3.8×1014 to 1.2×1014
1.2×1014 to 6.0×1012
4000 to 200
6.0×1012 to 3.0×1011
200 to 10
1.0×1012 to 3.0×108
10 to 0.01
3.2. Quantitative Analysis
Quantitative analysis is well established not only
in ultraviolet/visible and in near-infrared spec-
troscopy, but it is also very important in mid-
(10)
infrared measurements. The general prerequisite
for spectrometric quantitative analysis is defined
as follows [35]: information derived from the
spectrum of a sample is related in mathematical
terms to changes in the level(s) of an individ-
ual component, or several components within
the sample or series of samples, i.e., the spectral
response of an analyte can be related by a math-
ematical function to changes in concentration
of the analyte. In the ideal situation, the mea-
(11)
sured spectroscopic feature varies linearly with
concentration. In reality, however, true linearity
is not always obtained, but this is not impor-
tant provided the measured function is repro-
ducible. Most practical analyses are not absolute
measurements, and normally measurements are
made on a given instrument within a fixed work-
ing environment; under such set circumstances,
reproducibility and consistency of the measure-
ment are the most important factors.
If I 0 is the intensity of monochromatic ra-
(12)
diation entering a sample and I is the intensity
transmitted by the sample, then the ratio I/I 0 is
the transmittance T of the sample. The percent
transmittance ( % T ) is 100 T . If the sample cell
has thickness b and the absorbing component
has concentration c, then the fundamental rela-
tion governing the absorption of radiation as a
function of transmittance is:
I
T= = 10−abc (13)
I0
 Infrared and Raman Spectroscopy 9

Figure 4. FT-Raman spectrum of ascorbic acid recorded on a Bruker FT-Raman accessory FRA 106 directly interfaced to a
Bruker FT-IR spectrometer IFS 66
The accessory was equipped with a germanium diode cooled by liquid nitrogen. Resolution was 2 cm−1 and laser output
power was 300 mW. Stokes shift: 3600 to 50 cm−1 , anti-Stokes shift: − 100 to − 2000 cm−1 .

The constant a is called the absorptivity and
is characteristic of a specific sample at a spe-
cific wavelength. As the transmittance T does
not vary linearly with the concentration of an
absorbing species the equation above is usually
transformed by taking the logarithms of both
sides of the equation and replacing I/I 0 with I 0 /I
to eliminate the negative sign:
I0
log =abc (14)
I the transmission cell, it is advisable to control
The term log (I 0 /I ) is the absorbance A,
which changes linearly with changes in concen-
tration of an absorbing species:
A=abc (15)
This fundamental equation for spectromet-
ric quantitative analysis is known as Beer –
Lambert – Bouguer law, sometimes shortened to
Beer’s law or Beer – Lambert law.
In practice, several steps are involved in the
development of a quantitative method [35]. The
first and probably most crucial for the ultimate
success of the analytical method is to understand
the system, i.e., to obtain reference spectra of
the analyte and all other components. In the sec-
ond step, the best method of sampling should be
determined. Then, with standards having been
prepared, the system must be calibrated. The last
step before analyzing samples is to prepare val-
idation samples and to evaluate the method.
With modern sampling techniques, good
quantitative infrared analysis with virtually ev-
ery type of sample is practicable; however, liq-
uids are ideal for this purpose, being measured
in a liquid cell of fixed thickness, either as 100 %
sample or diluted with solvent. In this connec-
tion it should be taken into account that there are
no ideal solvents for infrared spectroscopy [35].
In addition, because absorption bands and path
length can be influenced by the temperature of
the temperature.
Polymers are usually analyzed as pressed
films, and solids are often difficult to exam-
ine quantitatively by infrared methods; in these
cases, absorption band ratios [10] give the best
results. In fact, the main application of multi-
component quantitative infrared analysis is gas
analysis. If the difficulty in handling the gases
is overcome, multicomponent analysis can be
readily performed.
A simple chemical system can consist of a
pure single component, or of a single component
or more than one component in a mixture with no
spectral interference; it is assumed that the radia-
tion absorption by one component is not affected
by the presence of other components. In simple
systems, absorbance peak height measurements,
directly or by using a selected baseline [35], are
often employed for calibration and analysis. Be-
cause of intrinsic instrumental errors the practi-
cal limit for usable absorbance values is about
10 Infrared and Raman Spectroscopy

three. Peak height measurements are also sen-
sitive to changes in instrumental resolution and
can vary considerably from instrument to instru-
ment. To circumvent these problems, an alterna-
tive method is the use of integrated absorbance
or peak area [10].
Complex chemical systems are composed of
one or more components in a mixture with a sig-
nificant degree of spectral interference, or of sev-
eral components with a large amount of mutual
physical and/or chemical interaction. In these
cases, quantitative analysis is best performed by
statistical methods such as principal component
regression (PCR) or partial least squares (PLS)
[36]; these are offered in the software packages
of instrument manufacturers and software sup-
pliers. Artificial neural networks (ANNs) should
be primarily used when a data set is nonlinear
[37].
chiefly in the band intensities that the two types
of spectra differ, sometimes markedly so. On
the other hand, when a center of symmetry is
present, i.e., for molecules with a high degree of
symmetry, bands that are infrared active are Ra-
man inactive and vice versa. By using laser ex-
citation (i.e., linearly polarized light) (see Sec-
tion 2.2), Raman spectroscopy also provides a
means of recognizing totally symmetrical vibra-
tions [3]. The intensity of a Raman line depends
on the direction of the exciting electric field in re-
lation to the orientation of the analyzer (Fig. 5).
The latter can be either perpendicular or parallel
to the direction of the electric field. The depolar-
ization ratio  is the ratio between the intensities
of scattered light measured at each of these two
positions:
I⊥
= (16)
I

3.3. The Symmetry of Molecules and

Molecular Vibrations; Selection Rules
All molecules show symmetry properties and
they all possess at least one (trivial) symmetry el-
ement, the identity. The symmetry of a molecule
is important in spectroscopy because changes in
symmetry during molecular vibration determine
whether a vibrational dipole moment µ occurs
or not. As mentioned, a vibration is infrared ac-
tive if µ changes, if not, it is infrared inactive;
analogously, this is true of the polarizability α
and Raman activity.
In infrared and Raman spectroscopy, the sym-
metry of molecules is usually discussed in terms
of five symmetry elements and their correspond-
ing five symmetry operations [38]. If a wide
variety of molecules is investigated, it will be
found, as can be proved by mathematical group
theory [39], that only certain combinations of
symmetry elements are possible. Such restricted
combinations of symmetry elements, known as
point groups, are used for the classification of
molecules. Each point group is associated with
a set of normal vibrations. These in turn are
classified according to the symmetry of vibra-
tion. From this it is possible to predict whether a Figure 5. Orientation of the analyzer in relation to the di-
normal vibration is infrared or Raman active or rection of the electrical field of the exciting laser
neither. These are the so-called selection rules.
Many vibrations of molecules with low symme- The depolarization ratio of a Raman line de-
try are both infrared and Raman active, and it is pends on the symmetry of the molecular vibra-
 Infrared and Raman Spectroscopy 11

tion involved (i.e., the change in molecular sym- Table 2. Commonly used symbols and descriptions or different
metry induced by the corresponding vibration); vibrational forms
the maximum value of depolarization observed
with linearly polarized light is max
l = 3/4. If a
Raman line shows this extent of depolarization,
it is said to be depolarized, whereas, if the degree
of depolarization is smaller, the line is polarized.
It can be shown [3] that only Raman lines cor-
responding to totally symmetric vibrations can
have a degree of depolarization smaller than the
maximum value, that is, can be polarized.

4. Interpretation of Infrared and

Raman Spectra of Organic
Compounds

4.1. The Concept of Group Frequencies

A complex molecule can be considered as a

system of coupled inharmonic oscillators. Em-
pirically it is found that vibrational coupling
is restricted to certain submolecular groups of
atoms. This coupling is relatively constant from
molecule to molecule, so that the submolecu-
lar groups produce bands in a characteristic fre-
quency region of the vibrational spectrum. These
bands – the characteristic group frequencies –
are readily predictable and so form the empirical
basis for the interpretation of vibrational spectra.
Their position, intensity, and width are decisive
for the correlation of an absorption band with
a certain submolecular group. For example, the
vibrational spectra of n-heptane, n-octane, and
n-nonane have a number of bands in common;
these are the group frequencies for normal alka-
nes. This concept of group frequencies corre-
sponds to the concept of chemical shift in nuclear
magnetic resonance spectroscopy. These spectra
also have a number of bands which are not in
common; these so-called fingerprint bands are
characteristic of the individual chemical com-
pound and are used to distinguish one com-
pound from another. In Table 2, commonly used
symbols and descriptions of different vibrational
forms are given, and Table 3 lists usual abbrevia-
tions for the classification of vibrational absorp-
tion bands. For practical applications, spectral
interpretation is mainly based on personal expe-
rience.
Table 3. Usual abbreviations for the classification of vibrational
absorption bands
Abbreviation Signification
vs very strong
s strong
m medium
w weak
vw very weak
sh shoulder
b broad
sr sharp
v variable
In the following sections, dealing with
spectrum – structure correlations, characteristic
group frequencies having diagnostic value are
given in the form of charts. These correlation
charts show the positions of the characteristic
group frequencies which are the same in in-
frared and Raman spectra. The two types of
spectra mainly differ in the band intensities,
which, however, are not indicated in the correla-
tion charts. To provide this information, exam-
ples of infrared and Raman spectra of most of the
functional groups discussed are shown. Fourier
transform Raman spectra (Fig. 4) are usually
12 Infrared and Raman Spectroscopy

recorded between 3600 and 50 cm−1 ( Stokes They are best seen in the solid-phase spectra
shift) and − 100 to − 2000 cm−1 (anti-Stokes of long straight-chain compounds such as fatty
shift). For comparison with mid-infrared spec- acids [10].
tra, in the following sections the Fourier trans-
form Raman spectra are presented in the range
between 4000 and 400 cm−1 . The relative in-
tensities of vibrational absorption bands, which
are evident in the model spectra presented, are
important for appropriate spectral interpretation.
Specific differences between infrared and Ra-
man spectra are given special mention.
In the figures in this chapter (Figs. 8, 9, 10,
11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36,
37, 38, 39, 40, 41, 42, 43, 44), the upper curve
is the infrared spectrum, with intensity increas- Figure 6. Characteristic group frequencies of alkanes
ing from the bottom to the top of the diagram.
The lower curve is the Raman spectrum with an
ordinate linear in relative intensity units increas-
ing from the bottom to the top of the diagram.
The Raman spectra have not been corrected for
fluctuations in the sensitivity of the spectrome-
ter. The infrared spectra were recorded with a
Bruker IFS 66 FT-IR spectrometer; for record-
ing of the Raman spectra a Bruker FRA 106 FT-
Raman accessory was used.

4.2. Methyl and Methylene Groups

Figure 6 shows characteristic group frequen-
cies, and Figure 7 illustrates typical vibrations of
alkanes. In Figure 8, infrared and Raman spec-
tra of 2-methylbutane (isopentane) are given as
an example of alkane spectra. The most intense
bands are those of the antisymmetric and sym-
metric CH3 or CH2 stretching vibrations bet-
ween 2960 and 2850 cm−1 . Typical of isopro-
pyl and tertiary butyl groups is that the symmet-
ric CH3 deformation band, usually observed at
1375 cm−1 , is split into two bands near 1390
and 1370 cm−1 . In the Raman spectra, the sym- Figure 7. Typical vibrations of alkanes
metric CH3 deformation band near 1375 cm−1
is generally relatively weak in alkanes, but when The CH2 twisting vibrations in CH2 chains
the CH3 group is next to a double or triple bond have frequencies dispersed over the same region
or an aromatic ring the Raman intensity is no- as the wagging bands, as can be seen in the spec-
ticeably enhanced [10]. This is demonstrated in tra of polyethylene [25], for example. The in-
Figure 9, which shows the spectra of 2-methyl- frared intensity of these bands is weak, whereas
2-butene. in the Raman spectrum at about 1300 cm−1 the
The CH2 wagging bands are spread over a in-phase CH2 twist vibration is a useful group
region between 1350 and 1180 cm−1 , occurring frequency.
as a characteristic progression of weak bands.
 Infrared and Raman Spectroscopy 13

Figure 8. FT-infrared (A) and FT-Raman (B) spectra of 2-

methylbutane
Figure 9. FT-infrared (A) and FT-Raman (B) spectra of 2-
methyl-2-butene

Figure 10. FT-infrared (A) and FT-Raman (B) spectra of cyclohexane

14 Infrared and Raman Spectroscopy
Figure 11. Characteristic group frequencies of alkenes
Figure 13. Characteristic group frequencies of aromatic
rings
The C−C stretching vibration of CH2 chains
is seen only in the Raman spectrum, at 1131
and 1061 cm−1 [40], while the CH2 rocking vi-
bration at 720 cm−1 , which is characteristic for
longer (CH2 )n chains with n ≥ 4, is observed
only in the infrared spectrum (e.g., polyethyl-
ene) [25].
Aliphatic ring compounds are best character-
ized by their Raman spectra, since in the infrared
only very weak characteristic ring frequencies
due to C−C ring stretching vibrations are ob-
served near 1000 cm−1 (Fig. 10). Raman spec-
tra, in contrast, show prominent lines due to ring
stretching vibrations: cyclopropane 1188 cm−1
[25], cyclobutane 1001 cm−1 [40], cyclopen-
tane 886 cm−1 [25], and cyclohexane 802 cm−1
(Fig. 10).
4.3. Alkene Groups
Characteristic group frequencies of alkenes are
shown in Figure 11. In alkenes, the =CH
Figure 12. FT-infrared (A) and FT-Raman (B) spectra of
2,3-dimethyl-2-butene
stretching vibration generally occurs above
3000 cm−1 . In symmetrical trans or symmet-
rical tetrasubstituted double-bond compounds,
the C=C stretching frequency near 1640 cm−1 ,
usually a medium intensity band, is infrared in-
active because, in this case, no change of dipole
moment occurs as a result of the vibration. How-
ever, the C=C stretching vibration gives rise to a
strong Raman signal in the region between 1680
and 1630 cm−1 in all types of alkenes (Fig. 12).
In vinyl and vinylidine groups, the =CH2 in-
plane deformation gives rise to a medium inten-
sity band in both the infrared and Raman spec-
trum near 1415 cm−1 . In the case of the vinyl
group, an additional infrared and Raman defor-
mation band is observed near 1300 cm−1 [10].
In cis-1,2-dialkyl ethylenes, the in-plane defor-
mation band appears near 1405 cm−1 in the in-
frared and at ca. 1265 cm−1 in the Raman spec-
trum. In the corresponding trans compounds, the
in-plane bending vibrations occur at 1305 cm−1
in the Raman and at 1295 cm−1 in the infrared
spectrum. In these trans compounds, the =CH
 Infrared and Raman Spectroscopy 15

out-of-plane bending vibration gives rise to a
strong infrared band, of high diagnostic value,
occurring between 980 and 965 cm−1 .
a very strong signal at 1000 cm−1 occurs for
mono and meta substitution, a strong line is
observed between 1055 and 1015 cm−1 for or-
tho substitution, while no signal is observed
around 1000 cm−1 for para substitution. In in-
frared spectra, the out-of-plane CH wagging vi-
brations give rise to relatively prominent sum-
mation bands in the 2000 – 1650 cm−1 region.
As the summation band patterns are approxi-
mately constant for a particular ring substitution,
they can also be used to determine the type of
substitution [10].

4.5. Triple Bonds and Cumulated

Double Bonds

Triple bonds (X≡Y) and cumulated double

bonds (X=Y=Z) absorb roughly in the same re-
gion, namely, 2300 – 2000 cm−1 (Table 4). The
bands are very diagnostic because in this region
of the spectrum no other strong bands occur.
Table 4. Characteristic group frequencies of triple bonds and
cumulated double bonds

Bond ν (≡C−H), ν (X≡Y) or

cm−1 ν as (X=Y=Z), cm−1
−C≡C−H 3300 2260 – 2100
−C≡N 2260 – 2200
−N≡C 2165 – 2110
Figure 14. FT-infrared (A) and FT-Raman (B) spectra of −C≡N→O 2300 – 2290
toluene −N+ ≡N 2300 – 2140
−S−C≡N 2180 – 2140
−N=C=S 2150 – 2000
−N=C=O 2275 – 2230
4.4. Aromatic Rings −N=N+ =N− 2250 – 2100
−N=C=N− 2150 – 2100
Figure 13 shows characteristic group frequen- −C=C=O 2155 – 2130
cies of aromatic rings. The bands between 1600 −C=N+ =N− 2100 – 2010
−CO−C=N+ =N−
and 1460 cm−1 are ring modes involving C–C 2100 – 2050

partial double bonds of the aromatic ring. The

band at 1500 cm−1 is usually the strongest of
these bands, which generally show medium to
strong infrared but weak Raman intensities. The
use of both infrared and Raman spectra gives
reliable information about the type of substitu-
tion of an aromatic ring: in infrared spectra, in-
tense bands in the 850 – 675 cm−1 region, due to
out-of-plane CH wagging and out-of-plane sex-
tant ring bending vibrations, are indicative of
the type of substitution [10]. In the Raman spec-
tra, there are very strong lines at ca. 1000 cm−1
involving ring stretching and ring bending vi-
brations. As shown in Figures 14, 15, 16, 17,
Monosubstituted acetylenes have a relatively
narrow ≡CH stretching band near 3300 cm−1
which is strong in the infrared and weaker in
the Raman. In monoalkyl acetylenes, the C≡C
stretching vibration shows a weak infrared ab-
sorption near 2120 cm−1 , while no infrared
C≡C absorption is observed in symmetrically
substituted acetylenes because of the center of
symmetry. In Raman spectra, this vibration al-
ways occurs as a strong line. Nitriles are charac-
terized by a strong C≡N stretching frequency at
16 Infrared and Raman Spectroscopy

Figure 15. FT-infrared (A) and FT-Raman (B) spectra of o-xylene

Figure 16. FT-infrared (A) and FT-Raman (B) spectra of m-xylene

Figure 17. FT-infrared (A) and FT-Raman (B) spectra of p-xylene

 Infrared and Raman Spectroscopy 17

Figure 18. FT-infrared (A) and FT-Raman (B) spectra of acetonitrile

4.6. Ethers, Alcohols, and Phenols

4.6.1. Ethers

Characteristic group frequencies of ethers are

shown in Figure 20. The key band for
noncyclic ethers is the strong antisymmetric
C−O−C stretching vibration between 1140
and 1085 cm−1 for aliphatic, between 1275
and 1200 cm−1 for aromatic, and between
1225 and 1200 cm−1 for vinyl ethers. With
the exception of aliphatic – aromatic ethers,
the symmetric C−O−C stretching band at ca.
1050 cm−1 is too weak to have diagnostic
value. In cyclic ethers, both the antisymmetric
(1180 – 800 cm−1 ) and the symmetric (1270 –
800 cm−1 ) C−O−C stretching bands are char-
acteristic. The antisymmetric C−O−C stretch is
usually a strong infrared band and a weak Ra-
man signal, the symmetric C−O−C stretching
vibration generally being a strong Raman line
and a weaker infrared band (Fig. 21).

Figure 19. FT-infrared (A) and FT-Raman (B) spectra of

methoxy-acetonitrile

2260 – 2200 cm−1 in the infrared and in the Ra-

man (Fig. 18). Upon electronegative substitution
at the α-carbon atom, the infrared intensity of the
C≡N stretching vibration is considerably weak- Figure 20. Characteristic group frequencies of ethers
ened so that only the Raman spectrum, where a
strong signal is always observed, can be used for
identification (Fig. 19).
18 Infrared and Raman Spectroscopy
4.6.2. Alcohols and Phenols
Characteristic group frequencies of alcohols and
phenols are shown in Figure 22. Because of the
strongly polarized OH group, alcohols generally
show a weak Raman effect (and therefore are ex-
cellent solvents for Raman spectroscopy). In the
liquid or solid state, alcohols and phenols exhibit
strong infrared bands due to O−H stretching vi-
brations. Due to hydrogen bonding of the OH
groups, these bands are very broad in the pure
liquid or solid, as well as in concentrated solu-
tions or mixtures. The position of the similarly
intense C−O stretching band permits distinction
between primary, secondary, and tertiary alco-
hols, as well as phenols.
Figure 21. FT-infrared (A) and FT-Raman (B) spectra of
tetrahydrofuran
4.7. Amines, Azo, and Nitro Compounds
4.7.1. Amines
Characteristic group frequencies of amines are
shown in Figure 23. In the infrared and Raman,
key bands are the NH stretching bands, although
they are often not very intense. Primary amines
exhibit two bands due to antisymmetric (3550 –
3330 cm−1 ) and symmetric (3450 – 3250 cm−1 )
stretching, while for secondary amines only one
NH stretching band is observed. The intensities
of NH and NH2 stretching bands are generally
weaker in aliphatic than in aromatic amines. The
C−N stretching vibration is not necessarily di-
agnostic, because in aliphatic amines it gives rise
to absorption bands of only weak to medium in-
tensity between 1240 and 1000 cm−1 . In aro-
matic amines, strong C−N stretching bands are
observed in the range 1380 – 1250 cm−1 .
Figure 22. Characteristic group frequencies of alcohols and
phenols
Figure 23. Characteristic group frequencies of amines
 Infrared and Raman Spectroscopy
4.7.2. Azo Compounds
Characteristic group frequencies of azo com-
pounds are shown in Figure 24. In the in-
frared spectrum, similarly to symmetrical trans-
substituted ethylenes, the N=N stretching vi-
bration of the trans symmetrically substituted
azo group does not occur, whereas all azo com-
pounds show a medium (aliphatic compounds)
or strong (aromatic compounds) N=N stretch-
ing Raman line. This band occurs between 1580
and 1520 cm−1 for purely aliphatic or aliphatic –
aromatic, and between 1460 and 1380 cm−1 for
purely aromatic azo compounds. In azoaryls, an
additional strong Raman line involving aryl – N
stretch is observed near 1140 cm−1 . As an ex-
ample, the spectra of azobenzene are shown in
Figure 25.
Figure 24. Characteristic group frequencies of azo com-
pounds
4.7.3. Nitro Compounds
Characteristic group frequencies of nitro com-
pounds are shown in Figure 26. The nitro group,
with its two identical N–O partial double bonds,
gives rise to two bands due to antisymmetric and
Figure 25. FT-infrared (A) and FT-Raman (B) spectra of azobenzene
19
symmetric stretching vibrations. In the infrared,
the antisymmetric vibration causes a strong ab-
sorption at 1590 – 1545 cm−1 in aliphatic, and
at 1545 – 1500 cm−1 in aromatic nitro com-
pounds. In comparison, the infrared intensities
of the symmetric stretching vibration between
1390 and 1355 cm−1 for aliphatic, and 1370 and
1330 cm−1 for aromatic nitro compounds are
somewhat weaker. However, a very strong Ra-
man line due to symmetric vibration is observed
in this region (see Fig. 27).
4.8. Carbonyl Compounds
Because of the stretching of the C=O bond, car-
bonyl compounds generally give rise to a very
strong infrared and a medium to strong Raman
line between 1900 and 1550 cm−1 . The position
of the carbonyl band is influenced by inductive,
mesomeric, mass, and bond-angle effects [10].
4.8.1. Ketones
Characteristic group frequencies of ketones are
shown in Figure 28. Ketones are characterized
by a strong C=O band near 1715 cm−1 , often ac-
companied by an overtone at ca. 3450 cm−1 . Ke-
tones in strained rings show considerably higher
C=O frequencies, whereas in di-tert-butyl ke-
tone, for example, the carbonyl frequency is ob-
served at 1687 cm−1 ; a steric increase in the
C−C−C angle causes this lower value.
20 Infrared and Raman Spectroscopy

Figure 26. Characteristic group frequencies of nitro com- Figure 28. Characteristic group frequencies of ketones
pounds

Figure 27. FT-infrared (A) and FT-Raman (B) spectra of 4-nitrobenzoic acid

two CH bands at 2900 – 2800 cm−1 and 2775 –

2680 cm−1 . These bands are due to Fermi res-
onance [3], i.e., interaction of the CH stretch
fundamental with the overtone of the O=C–H
bending vibration near 1390 cm−1 .

Figure 29. Characteristic group frequencies of aldehydes

Figure 30. Characteristic group frequencies of carboxylic
acids

4.8.2. Aldehydes

Characteristic group frequencies of aldehydes 4.8.3. Carboxylic Acids

are shown in Figure 29. Together with the car-
Characteristic group frequencies of carboxylic
bonyl band, the most intense band in the in-
acids are shown in Figure 30. Because carbox-
frared spectrum, aldehydes are characterized by
ylic acids have a pronounced tendency to form
 Infrared and Raman Spectroscopy 21

hydrogen-bonded dimers, in the condensed state

they are characterized by a very broad infrared
OH stretching band centered near 3000 cm−1 .
As the dimer has a center of symmetry, the an-
tisymmetric C=O stretch at 1720 – 1680 cm−1
is infrared active only, whereas the symmet-
ric C=O stretch at 1680 – 1640 cm−1 is Raman
active only; this is shown in Figure 31 using Figure 32. Characteristic group frequencies of carboxylate
acetic acid as an example. In the infrared, a salts
broad band at 960 – 880 cm−1 due to out-of-
plane OH · · · O hydrogen bending is diagnostic
of carboxyl dimers, as is the medium intense
C−O stretching band at 1315 – 1280 cm−1 . On
the other hand, monomeric acids have a weak,
sharp OH absorption band at 3580 – 3500 cm−1
in the infrared, and the C=O stretching band ap-
pears at 1800 – 1740 cm−1 in both infrared and
Raman.

Figure 33. Characteristic group frequencies of anhydrides

Figure 31. FT-infrared (A) and FT-Raman (B) spectra of

acetic acid
Figure 34. FT-infrared (A) and FT-Raman (B) spectra of
acetic anhydride
22 Infrared and Raman Spectroscopy

the Raman. In noncyclic anhydrides, the in-

frared C=O band at higher wavenumber is usu-
ally more intense than the C=O band at lower
wavenumber, whereas in cyclic anhydrides the
opposite is observed. In the Raman, the C=O
line at higher wavenumber is generally stronger.
Unconjugated straight-chain anhydrides have a
strong infrared absorption, due to the C−O−C
stretching vibration, at 1050 – 1040 cm−1 (ex-
cept for acetic anhydride: 1125 cm−1 ), while
cyclic anhydrides exhibit this infrared band at
955 –895 cm−1 . In Figure 34, the spectra of
acetic anhydride are shown as an example.

Figure 35. Characteristic group frequencies of esters

Figure 36. Characteristic group frequencies of carbonate

salts

4.8.4. Carboxylate Salts

Characteristic group frequencies of carboxyl-

ate salts are shown in Figure 32. In carboxyl-
ate salts, the two identical C–O partial double
bonds give rise to two bands: the antisymmetric
CO2 stretching vibration at 1650 – 1550 cm−1 is
very strong in the infrared and weak in the Ra- Figure 37. FT-infrared (A) and FT-Raman (B) spectra of
man, while the corresponding symmetric vibra- calcium carbonate
tion at 1450 – 1350 cm−1 is somewhat weaker
in the infrared but strong in the Raman.
4.8.6. Esters
4.8.5. Anhydrides Figure 35 shows the characteristic group fre-
quencies of esters. In esters, the C=O stretch-
Characteristic group frequencies of anhydrides
ing vibration gives rise to an absorption band
are shown in Figure 33. Anhydrides are charac-
at 1740 cm−1 which is very strong in the in-
terized by two C=O stretching vibration bands,
frared and medium in the Raman. Its position
one of which occurs above 1800 cm−1 . These
is strongly influenced by the groups adjacent to
bands are strong in the infrared but weak in
 Infrared and Raman Spectroscopy 23

the ester group. The C−O stretching band at appears near 3100 cm−1 due to an overtone of
1300 – 1000 cm−1 often shows an infrared in- the amide II band.
tensity similar to the C=O band.

4.8.7. Lactones

In lactones, the position of the C=O band is

strongly dependent on the size of the ring
(Table 5). In lactones with unstrained, six-
membered and larger rings the position is similar
to that of noncyclic esters.

4.8.8. Carbonate Salts Figure 38. Characteristic group frequencies of amides

Characteristic group frequencies of carbonate

salts are shown in Figure 36. In inorganic car-
bonate salts, the antisymmetric CO3 stretching
vibration gives rise to a very broad absorption
band at 1520 – 1320 cm−1 , which is very strong
in the infrared but only weak in the Raman.
The symmetric CO3 stretching vibration, on the
other hand, gives rise to a very strong Raman line
at 1100 – 1030 cm−1 , which is normally inactive
in the infrared. Another characteristic medium
sharp infrared band at 880 – 800 cm−1 is caused
by out-of-plane deformation of the CO2− 3 ion
(Fig. 37).
4.8.10. Lactams
Characteristic group frequencies of lactams are
shown in Figure 39. As in lactones, the size of
the ring influences the C=O frequency in lac-
tams. In six- or seven-membered rings the car-
bonyl vibration absorbs at 1680 – 1630 cm−1 ,
the same as the noncyclic trans case. Lac-
tams in five-membered rings absorb at 1750 –
1700 cm−1 , while β-lactams absorb at 1780 –
1730 cm−1 . These bands are strong in the in-
frared and medium in the Raman.

4.8.9. Amides

Figure 38 shows characteristic group frequen-

cies of amides. In amides, the C=O stretching
vibration gives rise to an absorption at 1680 –
1600 cm−1 known as the amide I band; it is
strong in the infrared and medium to strong in
the Raman. In primary amides, a doublet is usu-
ally observed in the amide I region, the sec-
ond band being caused by NH2 deformation. A
characteristic band for secondary amides, which
mainly exist with the NH and C=O in the trans
configuration, is the amide II band at 1570 –
1510 cm−1 , involving in-plane NH bending and
C – N stretching. This band is less intense than
the amide I band.
As in amines, the NH stretching vibration
gives rise to two bands at 3550 – 3180 cm−1 in
primary amides and one near 3300 cm−1 in sec-
ondary amides; these are strong both in infrared
and Raman. In secondary amides a weaker band
Figure 39. Characteristic group frequencies of lactams
Because of the cyclic structure the NH and
C=O groups are forced into the cis configuration
so that no band comparable to the amide II band
in trans secondary amides occurs in lactams.
A characteristically strong NH stretching ab-
sorption near 3200 cm−1 occurs in the infrared,
which is only weak to medium in the Raman. A
weaker infrared band near 3100 cm−1 is due to
a combination band of the C=O stretching and
NH bending vibrations.
24 Infrared and Raman Spectroscopy
Table 5. Characteristic group frequencies of lactones
Lactone
β-Lactones
γ-Lactones: saturated
α, β-unsaturated
β, γ-unsaturated
δ-Lactones: saturated
α, β-unsaturated
β, γ-unsaturated
4.9. Sulfur-Containing Compounds
Sulfur-containing functional groups generally
show a strong Raman effect, and groups such
as C–S and S–S, for example, are extremely
weak infrared absorbers. Hence, Raman spec-
tra of sulfur-containing compounds have a much
greater diagnostic value than their infrared spec-
tra.
4.9.1. Thiols
Characteristic group frequencies of thiols are
shown in Figure 40. The S−H stretching vibra-
tion of aliphatic and aromatic thiols gives rise
to a medium to strong Raman line at 2590 –
2530 cm−1 ; this band is weak or very weak in
the infrared.
Figure 40. Characteristic group frequencies of thiols
Structure ν (C = O), cm−1
1840
1795 – 1775
1785 – 1775 and 1765 – 1740
1820 – 1795
1750 – 1735
1760 – 1750 and 1740 – 1715
4.9.2. Sulfides and Disulfides
Figure 41 shows characteristic group frequen-
cies of sulfides and disulfides. The stretching
vibration of C–S bonds gives rise to a weak
infrared but a strong Raman signal at 730 –
570 cm−1 . Similarly, the S–S stretching vibra-
tion at 500 cm−1 is a very strong Raman line but
very weak infrared band (Fig. 42). Both Raman
signals are very diagnostic in the conformational
analysis of disulfide bridges in proteins.
Figure 41. Characteristic group frequencies of sulfides and
disulfides
 Infrared and Raman Spectroscopy
Figure 42. FT-infrared (A) and FT-Raman (B) spectra of
l-cystine
4.9.3. Sulfones
Similar to nitro groups and carboxyl salts, sul-
fones are characterized by two bands due to anti-
symmetric and symmetric stretching of the SO2
group, the former at 1350 – 1290 cm−1 and the
latter at 1160 – 1120 cm−1 (Fig. 43). Both bands
are strong in the infrared, while only the sym-
metric vibration results in a strong Raman line
(Fig. 44).
Figure 43. Characteristic group frequencies of sulfones
25
Figure 44. FT-infrared (A) and FT-Raman (B) spectra of
dimethyl sulfone
4.10. Computer-Aided Spectral
Interpretation
After his first experiences an interpreting vi-
brational spectra the spectroscopist soon real-
izes the limited capacity of the human brain to
store and selectively retrieve all required spec-
tral data. Powerful micro- and personal comput-
ers are now available together with fast and effi-
cient software systems to help the spectroscopist
to identify unknown compounds. IR Mentor
[41] is a program that resembles an interac-
tive book or chart of functional group frequen-
cies. Although the final interpretation remains
in human hands, this program saves the spectro-
scopist time by making tabular correlation infor-
mation available in computer form, and, more-
over, it is also an ideal teaching tool. To facil-
itate the automated identification of unknown
compounds by spectral comparison numerous
systems based on library search, e.g., SPEC-
TACLE [42], GRAMS/32 [43], or those offered
26 Infrared and Raman Spectroscopy

by spectrometer manufacturers, are used. These The sample is placed in the light beam of an
systems employ various algorithms for spectral infrared spectrometer, and the intensity of the
search [44], and the one that uses full spectra incident beam is compared with that transmitted
according to the criteria of Lowry and Hup- by the sample. According to the fundamental re-
pler [45] is the most popular. The central hy- lation governing the absorption of radiation as a
pothesis behind library search is that if spectra function of transmittance (see Section 3.2), pa-
are similar then chemical structures are simi- rameters that can be determined in transmission
lar [46]. In principle, library search is separated spectroscopy are the thickness or the concen-
into identity and similarity search systems. The tration of the sample. The thickness can range
identity search is expected to identify the sam- from micrometers for solids and liquids to even
ple with only one of the reference compounds kilometers for gas samples.
in the library. If the sample is not identical to
one of the reference compounds the similarity
search presents a set of model compounds simi-
lar to the unknown one and an estimate of struc-
tural similarity. The size and contents of the li-
brary are crucial for a successful library search
system, especially a similarity search system.
A smaller library with carefully chosen spectra
of high quality is more useful than a compre-
hensive library containing spectra of all known
chemical compounds or as many as are avail-
able because, in similarity search, the retrieval
of an excessive number of closely similar ref-
erences for a particular sample only increases
output volume without providing additional in-
formation [46]. A critical discussion of the per-
formance of library search systems is presented
by Clerc [46]. While library search systems are
well-established, invaluable tools in daily ana-
lytical work, expert systems (i.e., computer pro-
grams that can interpret spectral data) based on
artificial neural networks (ANNs) are more and
more emerging [47], [48].

5. Applications of Vibrational
Spectroscopy
5.1. Infrared Spectroscopy
5.1.1. Transmission Spectroscopy
Transmission spectroscopy is the simplest sam-
pling technique in infrared spectroscopy, and
is generally used for routine spectral measure-
ments on all kinds of samples. It requires only
simple accessories, such as infrared-transparent
windows [e.g., potassium bromide, sodium chlo-
ride, or thallium bromoiodide (KRS-5)] for
gases and liquids, or a sample holder for solids.
Figure 45. Optical diagram of a White gas cell (A mode)
showing eight passes, to change the path length only one
objective mirror is rotated, the rotation is nonlinear and is
very small at long path lengths. For B mode operation, the
light beam enters and exits on one and the same side of the
field mirror, the objective mirrors are coupled and are rotated
together linearly over the entire operational range
(Reproduced by permission of Harrick Scientific Corpora-
tion, Ossining, NY 10562)
Apart from general laboratory gas analysis,
gas-phase transmission spectroscopy is used for
the characterization of air pollutants as well as
the monitoring of stack gas and of air quality in
work places. As air quality has become a major
concern for environmentalists, industrial manu-
facturers and many others, spectroscopic mea-
 Infrared and Raman Spectroscopy 27

surement in the gas phase is receiving increasing as possible. The sample is ground separately and
attention. However, two major problems must then mixed with potassium bromide, after which
be overcome: absorptions in the gas phase are the mixture is ground only so far as is necessary
many times weaker than those in the liquid or to achieve good mixing.
solid phase, and the species to be measured usu-
ally makes up only a small fraction of the to-
tal gas volume. The most popular way of sur- 5.1.2. External Reflection Spectroscopy
mounting these problems is to increase the path
length by reflecting the beam several times from
precisely controlled mirrors. Such multiple-pass
gas cells are all based on the White design [49]
or variations thereof. Figure 45 shows an op-
tical diagram of a White cell aligned for eight
passes. The length of the gas cell and the number
of passes employed determines the optical path
length, which may vary from centimeters to kilo-
meters. The number of passes is adjusted in four Figure 46. Near-normal (θ1 ) and grazing angle (θ2 ) inci-
pass increments by rotating one of the objective dence
mirrors. In the A mode of operation, only one
objective mirror is rotated, while in the B mode
the objective mirrors are coupled and rotated to-
gether linearly over the entire operational range
to change the path length. To prepare a sample,
the gas cell is evacuated, the sample is bled into
the cell, and the cell is sealed.
Liquid samples are easily studied with the aid
of a wide variety of liquid cells, including heat-
able, flow through, and variable path length cells. Figure 47. Types of specular reflection
These cells are constructed of two infrared trans-
parent windows with a spacer between them, In external reflection spectroscopy, electro-
thereby forming a cavity for the sample. As magnetic radiation is reflected at the interface
the samples are often strong infrared absorbers, of two materials differing in refractive index
for this spectral range the liquid transmission or from a smooth surface, a conductor such as
cell must be constructed with short optical path metal, a semiconductor, a dielectric, or a liquid.
lengths (0.025 – 1 mm). The intensity of light reflected is measured to
Polycrystalline or powdered samples can be obtain information about the surface chemistry
prepared as a suspension in mineral oil (Nujol and structure. The angle of collection is the same
mull), as a potassium bromide disk (pellet), or as as the angle of incidence and is measured from
thin films deposited on infrared-transparent sub- the normal to the surface (Fig. 46). Small-angle
strates. The potassium bromide pellet is the most incidence (θ1 , usually 10◦ ) is called near-normal
common way of preparing powder samples; in incidence while that at large angles (θ2 , usually
this method a small amount, usually 1 mg, of 80◦ ) is known as grazing angle incidence. A non-
finely-ground solid sample is mixed with pow- destructive method requiring no sample prepara-
dered potassium bromide, usually 300 mg, and tion, external reflection spectroscopy comprises
then pressed in an evacuated die under high pres- two different types of measurements, both being
sure. The resulting disks are transparent and termed specular reflectance. In the first type, the
yield excellent spectra. The only infrared ab- reflection from a smooth mirrorlike surface, e.g.,
sorption in the potassium bromide matrix is due free-standing film, single crystal face, or sur-
to small amounts of adsorbed water, which can, face of an organic material, is referred to as true
however, be confused with OH- containing im- specular reflection or simply specular reflection
purities in the sample. To minimize adsorbed (Fig. 47). In the second, a film of sample is de-
water the potassium bromide is ground as little posited on a highly reflecting substrate, usually a
28 Infrared and Raman Spectroscopy

metal, from which the infrared beam is reflected, p-polarized light is parallel to it whereas that
the radiation is transmitted through the sample of s-polarized light is perpendicular to it. For s-
layer, reflected from the substrate surface, and fi- polarized light, the phase of the reflected beam
nally transmitted back through the sample. This is shifted 180◦ from the incident beam at all an-
technique is called reflection – absorption (RA), gles of incidence (Fig. 48), and the two beams
infrared reflection-absorption spectroscopy (IR- interfere at the surface to produce a node. Hence
RAS), or transflectance. little interaction occurs between the infrared ra-
A material’s object properties are character- diation and films thinner than one-quarter of the
ized by its complex refractive index, n = n + ik, wavelength. However, for p-polarized light, the
where n is the refractive index and k is the ab- phase shift is approximately 90◦ at large angles
sorption index, both of which are dependent on of incidence, and the combination of incident
the frequency of incident light. Unlike normal and reflected beams at the surface produces a
transmission spectra, true specular reflection standing wave, giving an intense electrical field
spectra often have distorted, derivative bands be- oriented normal to the surface. For metal sub-
cause specular reflectance from a surface fre- strates the greatest enhancement of the electrical
quently has a large refractive index component. field is produced by using p-polarization at large
However, by using a mathematical operation angles of incidence (e.g., 80◦ ).
known as the Kramers – Kronig transformation
[50] a transmission-like spectrum showing the
expected absorptivity information can be cal-
culated from a measured, distorted spectrum.
As true specular reflectance is often measured
at or near normal incidence, the reflected en-
ergy is small, usually between 0 and 12 % for
most organic materials. However, when this low
signal-to-noise ratio is overcome by using more Figure 48. Phase shift of the reflected beam occurring at
sensitive cooled MCT detectors the true spec- grazing angle incidence with perpendicular (180◦ shift) and
parallel (90◦ shift) polarized light
ular reflection technique has the important ad-
vantage of permitting the noncontact analysis of
bulk solids and liquids. Depending on the thick-
ness of the film of sample, two types of reflec-
tion – absorption measurement are used: near-
normal and grazing angle. In near-normal re-
flection – absorption, the thickness of the sam-
ple usually ranges between 0.5 and 20 µm and
the angle of incidence varies between 10 and
30◦ . Hence, the amount of energy reflected from
the substrate is much greater than the amount
reflected from the front surface of the sample
film, and a completely transmission-like reflec-
tion – absorption spectrum is obtained. Grazing
angle reflection – absorption measurements are
more sensitive than near-normal reflectance and
are used to analyze sample films less than 1 µm
thick, and even monolayers, on highly reflec-
tive materials. The improved sensitivity is due
to the enhanced electromagnetic field that can
Figure 49. Ray diagram of the wafer/disk checker attach-
be produced at the reflection surface by using ment for recording spectra of thin films on large samples,
polarized light at large angles of incidence [51], e.g., lubricants on hard magnetic disks;
[52]. As the plane of incidence contains the in- M1 – M6 = mirrors
cident and reflected rays, as well as the normal (Reproduced by permission of Harrick Scientific Corpora-
tion, Ossining, NY 10562)
to the reflection surface, the electrical field of
 Infrared and Raman Spectroscopy
Reflection – absorption measurements are
used in the analysis of a wide range of industrial
products, including polymer coatings on metals,
lubricant films on hard magnetic disks (Fig. 49),
semiconductor surfaces, or the inner coating of
food and beverage containers. In addition, graz-
ing angle reflectance is especially suited to the
study of molecular orientations at surfaces [53].
External reflectance at different angles of inci-
dence can be employed to determine both the
thickness and refractive index of a film of sample
by measuring the interference fringe separation
[54]; Figure 50 shows one of the various external
reflection attachments which are available.
Figure 50. Ray diagram of the versatile reflection attach-
ment (VRA) with single diamond polarizer (PSD)
(Reproduced by permission of Harrick Scientific Corpora-
tion, Ossining, NY 10562)
Figure 51. Schematic representation of total internal reflec-
tion with: A) Single reflection; B) Multiple reflection
IRE (internal reflection element)
n1 = Refractive index of the internal reflection element;
n2 = Refractive index of the sample with n2 < n1 ; θ = Angle
of incidence; d p = Depth of penetration
29
5.1.3. Internal Reflection Spectroscopy
Internal reflection spectroscopy, also known as
attenuated total reflectance (ATR) or multiple
internal reflectance (MIR), is a versatile, non-
destructive technique for obtaining the infrared
spectrum of the surface of a material or the spec-
trum of materials either too thick or too strongly
absorbing to be analyzed by standard transmis-
sion spectroscopy.
The technique goes back to Newton [55]
who, in studies of the total reflection of light
at the interface between two media of different
refractive indices, discovered that an evanescent
wave in the less dense medium extends beyond
the reflecting interface. Internal reflection spec-
troscopy has been developed since 1959, when
it was reported that optical absorption spectra
could conveniently be obtained by measuring
the interaction of the evanescent wave with the
external less dense medium [56], [57]. In this
technique, the sample is placed in contact with
the internal reflection element (IRE), the light
is totally reflected, generally several times, and
the sample interacts with the evanescent wave
(Fig. 51) resulting in the absorption of radia-
tion by the sample at each point of reflection.
The internal reflection element is made from a
material with a high refractive index; zinc se-
lenide (ZnSe), thallium iodide – thallium bro-
mide (KRS-5), and germanium (Ge) are the most
commonly used. To obtain total internal reflec-
tion the angle of the incident radiation θ must
exceed the critical angle θc [58]. The critical an-
gle is defined as:
n2
θc = sin−1 (17)
n1
where n1 is the refractive index of the internal
reflection element and n2 is the refractive in-
dex of the sample. What makes ATR a power-
ful technique is the fact that the intensity of the
evanescent wave decays exponentially with the
distance from the surface of the internal reflec-
tion element. As the effective penetration depth
is usually a fraction of a wavelength, total inter-
nal reflectance is generally insensitive to sam-
ple thickness and so permits thick or strongly
absorbing samples to be analyzed. The depth of
penetration d p , defined as the distance required
for the electrical field amplitude to fall to e−1 of
its value at the interface, is given by:
30 Infrared and Raman Spectroscopy
λ1
dp =  1/2 (18)
2π sin2 θ−n221
where λ1 = λ/n1 is the wavelength in the denser
medium, and n21 = n2 /n1 is the ratio of the re-
fractive index of the less dense medium divided
by that of the denser [56].
Although ATR and transmission spectra of
the same sample closely resemble each other,
differences are observed because of the depen-
dency of the penetration depth on wavelength:
longer wavelength radiation penetrates further
into the sample, so that in an ATR spectrum
bands at longer wavelengths are more intense
than those at shorter ones.
The depth of penetration also depends on the
angle of incidence; hence, an angle of 45◦ , which
allows a large penetration depth, is generally
used to analyze organic substances, rather than
an angle of 60◦ , which results in a substantially
weaker spectrum due to the decreased depth of
penetration.
The degree of physical contact between sam-
ple and internal reflection element determines
the sensitivity of an ATR spectrum. To achieve
this, a horizontal ATR accessory such as Fa-
stIR [58], in which the top plate is the sam-
pling surface, is used [59]; reproducible contact
is ensured by a special sample clamp or powder
press. Good quality spectra are thus obtained for
many materials that present problems of analy-
sis with routine transmission methods, e.g., pow-
ders, pastes, adhesives, coatings, rubbers, fibers,
thick films, textiles, papers, greases, foams, and
viscous liquids. Possible methods of obtaining a
spectrum from a variety of samples are discussed
in [60]; in situ ATR spectroscopy of membranes
is described in [61], [62]. Liquid samples are
also well suited to ATR analysis. Most liquids re-
quire a very short path length; aqueous samples,
for instance, are measured at path lengths of no
more than ca. 15 µm, which makes the design of
transmission cells difficult because flow of liq-
uids is hindered; they also exhibit interference
fringes because of the small spacing between the
high refractive index infrared windows. These
problems are eliminated by using liquid ATR
cells, a variant of solid ATR, in which the inter-
nal reflection element is surrounded by a vessel
into which the liquid is poured. Various such
liquid cells are available, e.g., the Circle [63],
the Prism Liquid Cell (PLC) [58], the Squarecol
[64] and the Tunnel Cell [65]. With optimized
optical design and fixed internal reflection ele-
ments these cells provide a highly reproducible
path length which permits the quantitative analy-
sis of liquids and aqueous solutions [66]. Liquid
ATR cells are uniquely suited to fully automat-
able, on-line process-monitoring of liquids and
viscous fluids. In addition, with suitable optical
transfer modules such as the Axiot system [67],
it is possible to operate liquid cells outside the
sample compartment of the FT-IR spectrome-
ter. This is important for on-line applications in
which it is not practicable to pipe the sample
to the spectrometer because the analysis can be
carried out at the most desirable location, even
in a fume hood or on a process line. With an
ATR sensing head, immersion probes [68] of-
fer the possibility of in situ FT-IR spectroscopy,
for monitoring batch process reactions (Fig. 52),
laboratory process development, the identifica-
tion of hazardous waste, the verification of the
contents of storage drums, and the inspection of
incoming raw materials.
Figure 52. Schematic drawing of an FT-IR measurement
system utilizing the Deep Immersion Probe Model DPR-
124 mounted in a batch reaction vessel [68]
a) FT-IR spectrometer; b) Optical transfer elements; c) De-
tector assembly; d) Reaction vessel; e) Mixing blade; f) ATR
sensing head
(Reproduced by permission of Axiom Analytical, Inc.,
Irvine, CA 92614)
5.1.4. Diffuse Reflection Spectroscopy
Diffuse reflection spectroscopy, also known as
diffuse reflectance infrared Fourier transform
 Infrared and Raman Spectroscopy
(DRIFT) spectroscopy, enables the analysis of
many samples for which traditional techniques
fail, to be made with little or no sample prepara-
tion. Many substances in their natural state, es-
pecially powders but also any solid with a rough
surface, such as dyed textiles and printed pa-
pers, exhibit diffuse reflection, i.e., incident light
is scattered in all directions, unlike specular re-
flection (see Section 5.1.2) where the angle of
incidence equals the angle of reflection. Diffuse
reflectance spectra are in practice complex, be-
cause of spectral distortions caused by the mix-
ing of both absorbance and reflectance features
owing to contributions from transmission, inter-
nal, and specular components in the measured
radiation. Hence, DRIFT spectra are affected by
particle size, packing, and dilution of the sample.
In diffuse reflectance spectra large glossy par-
ticle faces or crystal surfaces cause specular re-
flection which produces inverted bands, known
as reststrahlen bands; this effect is reduced by
fine grinding of the sample. Reststrahlen bands
may also occur in neat samples with a rough
surface, such as dyed textiles, printed papers,
and agricultural and botanical specimens [69],
measured by direct nondestructive, DRIFT spec-
troscopy. In this case, roughening the surface
with emery paper reduces the disturbing bands
because with smaller, thinner particles on the
roughened surface the transmission component
of the collected light increases, thus decreas-
ing the reststrahlen effect. In strongly absorbing
samples, the slight penetration of the incident
beam produces mainly specular reflection at the
surface, which also results in strong reststrahlen
bands. By diluting the sample with a nonabsorb-
ing powder such as KBr the effect is minimized
or even eliminated, as this ensures deeper pen-
etration of the incident beam and, therefore, an
increased contribution of transmission and in-
ternal reflection components to the DRIFT spec-
trum. The diffuse reflectance spectrum is usually
calculated as the ratio of sample and reference
reflectance, with the pure diluent being taken as
reference. Typically, a mixture of 90 – 95 % dilu-
ent and 10 – 5 % sample is used. Although the
diluted spectra and the bulk reflectance spectra
of weakly absorbing compounds have a simi-
lar appearance, they show enhanced intensity at
lower wavenumbers compared to transmission
spectra.
31
The interpretation of diffuse reflectance is
based on a theory developed by Kubelka and
Munk [70] and extended by Kortüm, Braun,
and Herzog [71] for diluted samples in non-
absorbing matrices. The diffuse reflectance R∞
of a diluted sample of infinite thickness (i.e., a
sample for which an increase in thickness does
not appreciably change the spectrum) is linearly
related to the concentration c of the sample, as
given by the Kubelka – Munk (K – M) equation:
(1−R∞ )2 2.303ac
f (R∞ ) = = (19)
2R∞ s
where a is the absorptivity and s is the scat-
tering coefficient. Because the scattering coeffi-
cient depends on both particle size and degree of
sample packing, the linearity of the K – M func-
tion can only be assured if particle size and pack-
ing method are strictly controlled. If these con-
ditions are fulfilled, accurate quantitative diffuse
reflectance measurements can be obtained [72],
[73].
Initiated by the work of Griffiths et al. [74],
diffuse reflectance accessories have been de-
veloped that pay special attention to the opti-
mal reduction of the disturbing specular compo-
nent. For example, the Praying Mantis diffuse
reflectance attachment (Fig. 53) uses an off-line
collection angle incorporating two 6 : 1 90◦ off-
axis ellipsoidal mirrors. One ellipsoidal mirror
focuses the incident beam on the sample, while
the radiation diffusely reflected by it is collected
by the other. Both ellipsoidal mirrors are tilted
forward; therefore, the specular component is
deflected behind the collecting ellipsoid, per-
mitting the diffusely reflected component to be
collected. As diffuse reflectance is a very sensi-
tive technique for the analysis of trace quantities
of materials, all available attachments allow for
microsampling [75]. The analysis of thin-layer
chromatograms by DRIFT spectroscopy, in situ
or after transfer, is described in [76], [77]. Cat-
alytic reactions can be monitored by using a con-
trolled reaction chamber attached to a DRIFT
accessory [78], [79]. Intractable samples, such
as paint on automobile panels, can be analyzed
by using silicon carbide abrasives [80]: a sili-
con carbide disk is rubbed against the sample, a
small amount of which adheres to the disk. The
DRIFT spectrum of the sample is then quickly
obtained by placing the disk in the diffuse re-
flectance attachment.
32 Infrared and Raman Spectroscopy
Figure 53. Ray diagram of the Praying Mantis diffuse re-
flectance attachment
EM = Ellipsoidal mirror; PM = Planar mirror; S = Sample
(Reproduced by permission of Harrick Scientific Corpora-
tion, Ossining, NY 10562)
5.1.5. Emission Spectroscopy
Infrared emission spectroscopy is a useful and
effective technique for studying the surface of
liquid and solid organic, inorganic, and poly-
meric materials, and for the in situ process-
monitoring of high-temperature reactions and
thermal transformations [81], [82]. In contrast
to the conventional transmission technique, in
emission spectroscopy the sample itself is the
infrared source. The temperature of the sample
is raised, which increases the Boltzmann popula-
tion of the vibrational energy states. Infrared ra-
diation is emitted on return to the ground state. In
theory, emission spectra can always be measured
provided the sample is at a different temperature
to the detector. Like transmission spectra, emis-
sion spectra are obtained by dividing the emis-
sion of the sample by that of the reference, which
is often a blackbody source at the same temper-
ature as the sample. The main reason for the
lack of any extensive work in this technique be-
fore 1970 was the low intensity of infrared emis-
sion, which results in spectra with poor signal-
to-noise ratio. However, the availability of FT-IR
spectrometers with the potential for low-energy
sampling and for the development of highly ef-
ficient FT-IR emission spectroscopy accessories
[83], [84] has led to a resurgence of interest in
infrared emission spectroscopy.
Because of the phenomenon of self-absorp-
tion the ideal sample for conventional emission
studies is a thin layer (e.g., a polymer film), on
both metal and semiconductor surfaces [81]. A
sample is usually heated from below the emit-
ting surface, the lower surface thus having a
higher temperature than the upper one. There-
fore, radiation emitted from below the upper
surface is absorbed before it reaches the sur-
face, and this self-absorption of previously emit-
ted light severely truncates and alters features in
the emission spectra of optically thick samples.
This problem is overcome by using a laser for
controlled heat generation within a thin surface
layer of the sample, self-absorption of radiation
thus being minimized. These methods, known
as laser-induced thermal emission (LITE) spec-
troscopy [85], [86] and transient infrared emis-
sion spectroscopy (TIRES) [87], [88] can pro-
duce analytically useful emission spectra from
optically thick samples. Quantitative applica-
tions of infrared emission spectroscopy are de-
scribed in [89–91].
5.1.6. Photoacoustic Spectroscopy
Photoacoustic spectroscopy (PAS) [92–95] is a
fast, nondestructive method for analyzing var-
ious materials in the gas, liquid, or solid state
with virtually no sample preparation. In cases
where samples are insoluble, difficult to grind
into a powder, or of irregular shape – such as
coal, carbon-filled and conducting polymers,
pharmaceutical preparations, spin coatings on
fibers, and coatings on irregular surfaces –
photoacoustic spectroscopy complements other
established techniques such as attenuated to-
tal reflectance (ATR) and diffuse reflectance
(DRIFT). Moreover, measurements on samples
that show the Christiansen effect [96] when pre-
pared as KBr pellets for standard transmission
spectroscopy can be better performed by the
photoacoustic method. As most photoacoustic
measurements are made with solid or semisolid
samples, the sample is sealed in a small-volume
cell containing a nonabsorbing gas such as he-
lium and having windows for optical transmis-
sion (Fig. 54). The sample is illuminated with
modulated radiation from the monochromator
of a dispersive instrument or the interferometer
of a Fourier transform infrared spectrometer. At
 Infrared and Raman Spectroscopy
wavelengths where the sample absorbs a frac-
tion of the incident radiation, a modulated tem-
perature fluctuation at the same frequency, but
not necessarily with the same phase, as that of
the incident radiation is generated in the sam-
ple, and the surrounding inert gas produces pe-
riodic pressure waves in the sealed cell. These
fluctuations in pressure can be detected by a mi-
crophone because the modulation frequency of
the incident beam usually lies in the acoustic
range, e.g., between 100 and 1600 Hz. The low
power available from dispersive instruments re-
sults in spectra with a rather poor signal-to-noise
ratio; hence, photoacoustic mid-infrared spectra
of condensed substances are preferably obtained
by using an FT-IR spectrometer with a pho-
toacoustic cell accessory. In a rapid-scanning
Michelson interferometer with a mirror veloc-
ity of v (in cm/s), radiation of wavenumber ν̃
(in cm−1 ) is modulated with a frequency of
2 v ν̃ Hz [8]. Therefore, FT-IR spectrometers are
ideal for photoacoustic measurements in the
mid-infrared, where mirror velocities on the or-
der of 0.05 – 0.2 cm s−1 provide modulation fre-
quencies in the acoustic range. As carbon black
absorbs all incident radiation, the sample single-
beam spectrum is usually ratioed against a car-
bon black single-beam spectrum to obtain an
absorbance-like spectrum.
Figure 54. Schematic of a photoacoustic cell
Before the first Fourier transform infrared
photoacoustic spectrum was reported in 1978
[97], most infrared photoacoustic experiments
were performed by using high-intensity tun-
able laser sources. However, as the source is
monochromatic, this type of experiment is only
useful for monitoring a single line in the spec-
trum. Highly-developed photoacoustic acces-
sories are now available from FT-IR manufac-
turers or other companies [98], which have a
33
high sensitivity also for microsamples. As the
signal is inversely proportional to the cell vol-
ume, cells optimized for the measurement of
condensed samples have a small volume, typ-
ically well under 1 cm3 , to enhance the signal
amplitude [99]. Water vapor or carbon dioxide
generate good photoacoustic signals, so even
traces of these impurities cause serious interfer-
ence in the photoacoustic spectroscopy signal;
therefore, the purity of the gas in the cell is es-
sential for good-quality photoacoustic spectra.
Important applications of photoacoustic
spectroscopy are near-surface analysis [100] and
depth profiling of polymers and polymer lay-
ers [101–103]. The theory of depth profiling
is described in detail in [92]. It is based on
the fact that thermal diffusion length (the dis-
tance a thermal wave travels in the sample be-
fore its intensity is reduced by 1/e in magni-
tude) and hence intensity of the photoacoustic
signal are increased by lowering the modula-
tion frequency. In a rapid-scanning interferome-
ter this is achieved by decreasing the mirror ve-
locity. To date, rapid-scanning interferometers
that provide modulation frequencies which con-
tinuously change with wavenumber have gener-
ally been used. However, step-scan interferom-
eters accommodating any desired modulation
frequency, constant over the entire wavelength
range, produce improved photoacoustic spectra
and are thus more effective in depth profiling
[104–107].
Saturation effects, which can occur for
stronger absorption bands, are the main limita-
tion of photoacoustic spectroscopy. These dis-
tortions can be overcome by using thin samples
[93] or by analyzing the phase of the photoacous-
tic signal [108]. The latter method is particularly
useful for quantitative analysis.
5.1.7. Chromatography/Fourier Transform
Infrared Spectroscopy
The combination of chromatographic separation
with Fourier transform infrared spectroscopy
has significantly improved the analysis of com-
plex mixtures [109]. In chromatography/FT-IR
systems, an infrared detector providing informa-
tion on the structure of separated species is used
instead of standard bulk chromatography detec-
tors such as thermal conductivity, flame ioniza-
34 Infrared and Raman Spectroscopy
tion, ultraviolet, or fluorescence; these detectors
can be used for quantitative analysis when the
identities of the mixture components are known.
In these systems the following factors must
be taken into consideration: the physical state
of the matter to be analyzed, the optimization
of the interface between chromatograph and
FT-IR, and the specific requirements and con-
straints of all three parts. In gas chromatography
(GC)/FT-IR spectroscopy [110], one type of in-
terface is composed of temperature-controlled,
long, thin flow-through gas cells or “light pipes”
coated with gold for high reflectivity. As the
light-pipe techniques have a detection limit in
the nanogram range, subnanogram limits can
be routinely attained with the matrix-isolation
technique (GC/MI/FT-IR) [109], and even bet-
ter detection limits are achieved with the sub-
ambient trapping GC/FT-IR technique [109],
[111]. Rapid-scanning interferometers able to
record the entire spectrum in less than one sec-
ond at low resolution (usually 4 or 8 cm−1 ),
together with highly sensitive detectors (e.g.,
cooled MCT detectors), are essential in all col-
umn chromatography/FT-IR spectroscopic tech-
niques, with capillary GC/FT-IR requiring the
most rapid data acquisition. In comparison to
vapor phase species, rotations of matrix-isolated
species are hindered, which results in sharper
and more intense bands. Hence, for accurate
spectra representation, GC/MI/FT-IR spectra are
usually measured at 1 cm−1 resolution.
The Gram – Schmidt vector orthogonaliza-
tion method [112] is the most sensitive and
computationally most efficient and, hence, the
most popular procedure for the reconstruction
of the total chromatographic trace from in-
terferometric infrared data. Selective detection
is obtained by functional-group-specific chro-
matograms (chemigrams [113]), which are pro-
duced by computing absorbance spectra dur-
ing chromatographic separation and plotting
integrated absorbance for prespecified wave-
length windows as a function of separation time,
representing the quantity of a specific func-
tional group that elutes as a function time. In
chromatography/FT-IR spectral evaluation, in
order to classify unknowns as to compound type,
library searching (see Section 4.10) has proved
to be very useful.
The addition of a complementary detector
significantly increases the power of GC/FT-
IR analysis. Mass spectrometry (MS) is an
ideal choice [114] because the disadvantages
of each method – infrared spectroscopy often
cannot distinguish between long-chain homo-
logues, and mass spectrometry frequently fails
to distinguish isomeric species – are offset by the
other.
The crucial point in liquid chromatography
(LC)/FT-IR methods, however, is the interfer-
ences caused by the liquid phase. Actually, two
viable LC-IR systems are commercially avail-
able: the LC-Transform [115] and the InfraRed
Chromatograph (IRC) [116], [117]. While the
LC-Transform is a solvent-elimination inter-
face, the IRC provides full FT-IR spectra of
the eluent stream from an LC column. The LC-
Transform system utilizes a high-performance
ultrasonic nebulizer or a pneumatic linear cap-
illary nozzle to remove the mobile phase from
samples as they elute from the chromatograph.
The capillary is surrounded by hot, flowing
sheath gas which provides sufficient thermal en-
ergy to evaporate the mobile phase and to contain
the nebulized spray in a tight, focused cone as
the spray emerges from the nozzle. The sample
is deposited on a germanium sample-collection
disk as spots with diameters between 0.5 and
1 mm. In a consecutive, uncoupled step, good-
quality infrared spectra can be obtained from
these spots, which contain microgram to sub-
microgram amounts, by the use of a specially
designed optics module, comprising a beam con-
denser. While this optics module can be used in
any FT-IR spectrometer uncoupled from the LC-
Transform interface, the IRC includes an FT-IR
spectrometer, a data station for driving the spec-
trometer and for data-processing, and a vacuum
interface for depositing and scanning the sam-
ple. Being a real-time direct deposition interface,
the IRC uses an ultrasonic nebulizer, a desolva-
tion tube and a vacuum chamber to evaporate the
solvent from the LC stream. The sample residue
is then deposited onto a moving zinc selenide
window. The data station collects a continuous
array of spectra as the dry sample track on the
window moves through the focused beam of the
spectrometer.
LC-IR provides powerful identification pos-
sibilities, based on molecular structure. A high
degree of confidence can be placed on substance
identification that simultaneously matches in-
frared spectral characteristics and chromato-
 Infrared and Raman Spectroscopy
graphic elution time. This is especially so when
molecules that have various conformational iso-
mers are to be distinguished.
Supercritical fluid chromatography
(SFC)/FT-IR spectroscopy, generally with car-
bon dioxide as the mobile phase, bridges the
gap between GC/FT-IR and LC/FT-IR, and is
particularly useful for separating nonvolatile or
thermally labile materials not amenable to gas
chromatographic separation [109]. Flow cells,
mobile phase elimination and matrix-isolation
techniques are used as SFC/FT-IR interfaces.
Applications of chromatography/FT-IR spec-
troscopy involve toxins and carcinogens, waste-
water constituents, sediment extracts, airborne
species, pesticides and their degradation prod-
ucts, fuels and fuel feedstocks, flavors and
fragrances, natural products, pharmaceuticals,
biomedicine, and polymers.
5.1.8. Thermogravimetry/Fourier
Transform Infrared Spectroscopy
Thermogravimetry (TG), also called thermo-
gravimetric analysis (TGA), measures changes
in sample mass as a function of increasing
temperature and has become an important tool
for the analysis and characterization of mate-
rials [118], [119]. To correlate weight changes
with chemical changes, it is necessary to iden-
tify gases evolved from the thermogravimet-
ric analyzer furnace. This is made possible by
coupling thermogravimetry with FT-IR spec-
troscopy, which provides the necessary speed
and sensitivity.
A system has been described in which the
infrared beam of the FT-IR spectrometer is led
directly into the TG equipment [120], but in
commercially available systems [121], [122] the
evolved gases are conducted from the TG sys-
tem to the spectrometer by carrier gas flow via a
transfer line. The interface is heated to prevent
recondensation and incorporates the transfer line
and a flow-through gas cell with nitrogen or he-
lium as carrier gas.
After infrared spectra at low resolution (usu-
ally 4 cm−1 ) have been recorded as a func-
tion of time and temperature, the further proce-
dure is similar to GC/FT-IR (see Section 5.1.7)
and, in fact, GC/FT-IR software, such as Gram –
Schmidt thermogram reconstruction and library
35
search with a vapor-phase spectra database is
commonly used. Helpful for the identification in
the vapor phase is the virtual absence of interac-
tion between molecules, and therefore the non-
destructive IR detection enables in many cases
an easy and fast interpretation of single and over-
lapping absorption bands.
Providing structural information for sample
components down to nanogram levels, TG/FT-
IR applications include decomposition studies
of polymers and laminates [123], the analysis
of coal, oil shales, and petroleum source rocks
[124], [125], and the determination of activa-
tion energies [126] and thermal decomposition
of energetic materials [127].
The IR detection of evolved gases opens
new possibilities in the quantification of ther-
mogravimetric analysis. Besides measuring the
mass loss of a TG step, the total amount can
also be calculated after calibration of a single
component and integration over time [128]. A
further improvement in quantitative TG/FT-IR
is achieved by PulseTA [129]. Here, a certain
amount of gas or liquid is injected into the sys-
tem and used as internal calibration standard
for the quantification of the unknown amount
of evolved gas.
5.1.9. Vibrational Circular Dichroism
Vibrational circular dichroism (VCD) extends
the functionality of electronic circular dichroism
(CD) into the infrared spectral region, with VCD
bands of enantiomeric pairs of chiral molecules
having opposite signs. Early VCD spectrome-
ters were dispersive instruments equipped with
a CaF2 photoelastic modulator for creating left
and right circularly polarized radiation and a liq-
uid-nitrogen-cooled InSb detector. FT-IR-VCD
measurements are carried out by using the so-
called double modulation technique [130] in
which the VCD interferogram is carried by the
polarization modulation frequency in the region
of 50 kHz, and the IR transmission interfero-
gram occurs in the usual range of Fourier fre-
quencies, typically below 5 kHz.
Combining the structural specificity of FT-
IR spectroscopy with the stereosensitivity of cir-
cular dichroism, VCD allows the determination
of optical purity without enantiomeric separa-
tion or derivatization, as well as of absolute
36 Infrared and Raman Spectroscopy
configuration without crystallization. Simulta-
neous monitoring of the optical purity of multi-
ple chiral species, such as reactant and product
molecules in a reaction process is also possible,
as is the determination of conformations of bio-
logical molecules such as proteins, peptides, and
DNA [131], [132].
5.2. Raman Spectroscopy
In mid-infrared spectroscopy, Fourier transform
instruments are used almost exclusively. How-
ever, in Raman spectroscopy both conventional
dispersive and Fourier transform techniques
have their applications, the choice being gov-
erned by several factors [133], [134]. Conse-
quently, a modern Raman laboratory is equipped
with both Fourier transform and CCD-based dis-
persive instruments. For a routine fingerprint
analysis, the FT system is generally used, be-
cause it requires less operator skill and is quicker
to set up; the FT system is also be tried first
if samples are highly fluorescent or light sen-
sitive. However, if the utmost sensitivity is re-
quired, or if Raman lines with a shift smaller
than 100 cm−1 are to be recorded, conventional
spectrometers are usually preferred.
Raman spectroscopy offers a high degree of
versatility, and sampling is easy because no spe-
cial preparation is required. As it is a scatter-
ing technique, samples are simply placed in the
laser beam and the backscattered radiation is an-
alyzed.
Glass gives only a weak Raman signal, so that
FT-Raman spectra from liquids are recorded by
using a capillary cell [34], a sapphire sphere [34],
or glass cuvettes with the back surface silvered
for the favored 180◦ backscattering arrangement
(see Section 2.2). Moreover, the study of aque-
ous samples in glass or quartz vessels, which
is difficult in the mid-infrared region owing to
strong infrared absorption bands of water and
quartz, is possible in Raman spectroscopy be-
cause there are no significant bands causing in-
terference in the solution spectrum. Powders are
either measured by using microspherical cells
[33] or are filled into cells consisting of a brass
cylinder with a 3 mm hole drilled through, into
which a reflective metal rod is placed to com-
press the powder [14]. The Raman intensity is
related to the sample thickness, but once the
thickness exceeds 2 mm the intensity does not
increase. Optical fibers are used for remote sam-
pling, and for FT-Raman spectroscopy this is es-
pecially advantageous as the transmission by the
fibers, excited by a near-infrared Nd : YAG laser
[33], [135], can be at its highest just in the range
of the Raman spectrum. Guidelines for optimiz-
ing FT-Raman recording conditions are given
in [34]. Conventional dispersive Raman spec-
troscopy has not been widely used for quantita-
tive measurements, mainly because of difficul-
ties with band reproducibility. However, by us-
ing an FT-Raman spectrometer, spectra with ex-
cellent reproducibility, both in band position and
intensity, are obtained, so that this technique has
great potential for quantitative analysis [136–
138].
Along with general reviews [139–142] and
surveys of major areas of Raman spectroscopy
[16], [17], various reviews covering special ap-
plications of Raman spectroscopy, such as po-
larization measurements [143], polymers [144],
proteins and nucleic acids [145], pharmaceu-
ticals and biomaterials [146–150], paint sys-
tems [151], [152], inorganic and organometallic
materials [153], and thin-layer chromatography
[154], have been published. For further appli-
cations and special Raman techniques, such as
resonance Raman (RR), surface-enhanced Ra-
man spectroscopy (SERS), coherent anti-Stokes
Raman scattering (CARS), and stimulated Ra-
man scattering (SRS), see [14], [155].
5.3. Fourier Transform Infrared and
Raman Microspectroscopy
In conventional infrared microsampling, beam
size is reduced four to six times by using beam
condensers. These accessories, which permit
submicrogram samples of between 2 and 0.5 mm
in diameter to be measured, have, however,
drawbacks such as heating effects, since the en-
tire power of the infrared beam is condensed on
the sample, and the fact that the sample can-
not be visibly located, so that a skilled operator
is needed for this time-consuming work. These
disadvantages are largely overcome by coupling
an optical microscope to an FT-IR spectrome-
ter. As infrared lenses are poor in performance,
infrared microscopes use reflecting optics such
as Cassegrain objectives instead of the glass or
 Infrared and Raman Spectroscopy 37

quartz lenses employed in conventional optical of a laser source, the flux of Raman radia-
microscopes. Figure 55 shows the beam path of a tion is inversely proportional to the diameter
typical infrared microscope with both transmis- of the laser-beam focus at the sample, i.e., an
sion and reflection modes. The infrared beam optimized Raman sample is a microsample.
from the FT-IR instrument is focussed onto the However, Raman microspectroscopy able to ob-
sample placed in a standard microscope, and tain spatially resolved vibrational spectra to ca.
the beam that passes through the sample is col- 1 µm spatial resolution and using a conventional
lected by a Cassegrain objective with 15- or 36- optical microscope system has only recently
fold magnification that produces an image of been more widely appreciated. For Raman mi-
the sample within the barrel of the microscope. crospectroscopy both conventional [162] and
The radiation passes through the variable aper- FT-Raman spectrometers [163], [164] are em-
ture placed in the image plane and is focussed ployed, the latter being coupled by near-infrared
by another Cassegrain objective on a small area fiber optics to the microscope.
(250 × 250 µm or less) of the specially matched Applications of Raman microspectroscopy
MCT detector. As the visible optical train is include the analysis of a wide variety of organic
collinear with the infrared light path, it is pos- and inorganic materials, e.g., semiconductors,
sible to position the sample and to isolate and polymers, single fibers, molecular crystals, and
aperture the area for analysis visually. This is minerals [165–168].
the main advantage of an FT-IR microscope over
a beam condenser, and it enables the measure-
ment of, for example, microcontaminants that
cannot be removed from the sample, or the in-
dividual layers of a polymer laminate. FT-IR
microscopy requires virtually no sample prepa-
ration. A further advantage of the technique is
its unsurpassed sensitivity, which permits sam-
ples in the picogram range to be analyzed, with
the diffraction limit of infrared radiation (10 –
20 µm) as the limiting condition for the size of
measurement spot. To minimize stray light due
to diffraction being collected by the microscope,
particularly when the size of the area investi-
gated is approximately equal to the wavelength
employed, a second aperture is introduced into
the optics that focus the radiation onto the sam-
ple [156].
Uses of FT-IR microspectroscopy include
general characterization of particulate matter,
dichroic measurements with polarized light,
polymer characterization, semiconductor mea-
surements, the identification of contaminants, as
well as forensic, biological, and pharmaceutical
applications [157–159].
By using diamond cells with the beam con-
Figure 55. Beam path of a typical infrared microscope
denser or microscope, the thickness of a sample I = Infrared transmittance beam; II = Infrared reflectance
(e.g., paint chips) can be adjusted by squeez- beam; M1 = Condensing mirror; M2 and M3 = Cassegrain
ing. An alternative to this is an ultrasmall sam- objectives; M4 = Movable, semi-transparent mirror;
ple analyzer [160], which allows nondestructive M5 = Movable mirror; M6 = Mirror; SS = sample stage;
A = Aperture; E = Eyepiece; D = MCT-Detector, liquid N2
internal reflectance studies of microgram and cooled
nanogram samples to be performed. (Reproduced by permission of Bruker Optik GmbH,
In principle, Raman spectroscopy is a mi- D-76275 Ettlingen, Germany)
crotechnique [161] since, for a given light flux
38 Infrared and Raman Spectroscopy

Figure 56. Conceptualization of a three-dimensional hyperspectral image data cube

(A) Images as a function of wavenumber; B) Selected spectra associated with pixels (i, j) in the image
(Reproduced by permission of Marcel Dekker, Inc., 270 Madison Avenue, New York, NY 10016-0602)

5.4. Time-Resolved FT-IR and cations of time-resolved FT-IR and FT-Raman

FT-Raman Spectroscopy spectroscopy are described in [169].

Time-resolved FT-IR and FT-Raman techniques

allow studies of molecular reaction mechanisms
with time resolutions up to nanoseconds [169],
which is a prerequisite for the understanding
of biological processes at the molecular level.
The major problem in measuring biological re-
actions consists in selecting small absorption
bands of the molecular groups undergoing re-
actions from the large background absorption of
water and the entire biological molecule. This
difficulty is met by obtaining difference spec-
tra of two reaction states of the molecules un-
der investigation. The intrinsic advantages of FT
spectroscopy, i.e., the multiplex, the through-
put, and the accuracy (see Section 2.1) enable
such small absorption changes to be reliably de-
tected. With operation of the FT spectrometer in
the step-scan mode [169], a time resolution of a
few nanoseconds is achieved. Biological appli-
5.5. Vibrational Spectroscopic Imaging
Although vibrational spectroscopic imaging
methods are relative newcomers to the panoply
of vibrational techniques, they inherit the ver-
satility of the traditional single-point infrared
and Raman approaches and, in combination with
digital imaging technology, allow for new per-
spectives in the interpretation of the spectra.
Imaging advantages are derived, in part, from the
ability of an investigator to process and discern
effectively a complex two-dimensional spatial
representation of a sample in terms of a variety of
spectrally related molecular parameters. In par-
ticular, a single spectroscopic image, or chemi-
cal map, is capable of summarizing and convey-
ing a wealth of spatial, chemical, and molecular
data of benefit to both specialists and nonspe-
 Infrared and Raman Spectroscopy
cialists. Regardless of the specific method used
to collect data during a vibrational spectroscopic
imaging experiment, the final representation is
typically expressed as a three-dimensional im-
age cube, or hypercube, consisting of two spa-
tial dimensions and one spectral dimension. The
image cube concept, shown in Figure 56, is in-
terpreted in one of two ways, either (1) a series
of images collected as a function of wavenum-
ber (Figure 56A), or (2) a vibrational spectrum
corresponding to each spatial location, or pixel,
within the image plane (Figure 56B). The abil-
ity to merge and recall, within a single analyti-
cal technique, information corresponding to the
spatial and spectral axes of the image cube pro-
vides extraordinary flexibility in probing and ex-
tracting structural and compositional informa-
tion from samples. Thus, a data set may be either
summarized as a single image or expressed as a
series of images derived from one or more spec-
troscopic parameters, such as spectral peak in-
tensity, band frequency, or linewidth. The chem-
ical and morphological interpretations of the re-
sulting images are then based on the spatial vari-
ations of the intrinsic molecular property being
highlighted and their relevance to the biologi-
cal questions being pursued. In contrast to im-
ages generated by conventional point mapping,
which generally suffer from low spatial resolu-
tion due to relatively few sampling points, direct
imaging approaches using the advantages inher-
ent in two-dimensional array detectors derive an
enormous benefit from the highly parallel na-
ture of the data collection. That is, a single data
set which typically contains many tens of thou-
sands of independent spectral channels may be
analyzed and interpreted by using a variety of
statistical algorithms, allowing subtle spatial and
spectral variations that are often overlooked or
misinterpreted in traditional spectroscopic stud-
ies to be tested and revealed.
Vibrational spectroscopic imaging has been
widely applied to diverse materials, includ-
ing polymers, semiconductors, pharmaceuticals,
cosmetics, consumer products, and biological
materials. The strengths and adaptability of vi-
brational spectroscopic imaging rest not only
on the ability to determine chemical composi-
tions and component distribution within a sam-
ple, but also on being able to extract localized
molecular information relevant to the sample ar-
chitecture. For example, vibrational spectra are
39
sensitive to alterations in protein secondary and
tertiary structures. Spectral shifts could there-
fore be used to image specifically the distribu-
tions of specific structural moieties within a bi-
ological sample as opposed to measuring sim-
ply an overall distribution of a general class of
molecules. Multivariate approaches may also be
used to generate a composite metric indicative of
either disease progression or biological function.
Although these metrics can not be readily in-
terpreted in terms of biochemical changes, they
may often be statistically correlated with dis-
ease. In addition, by simultaneously recording
data on a two-dimensional focal-plane array de-
tector from a myriad of spatial locations within a
biological matrix, pooled spectra may be treated
statistically. In this manner, analysis of variance
can be employed to more robustly test the signif-
icance of observed spectral changes appearing,
for example, because of a perturbed or diseased
state.
Raman and infrared imaging techniques as
well as their biological applications are de-
scribed in detail in [170].
5.6. Infrared and Raman Spectroscopy
of Polymers (see also → Plastics Analysis,
Chap. 4.)
As polymer analysis is a discipline of its own
beyond the scope of this article, the reader is re-
ferred to the comprehensive works by Hummel
[171], Koenig [172], and Zerbi et al. [173].
6. Near-Infrared Spectroscopy
6.1. Comparison of Mid-Infrared and
Near-Infrared Spectroscopy
Near-infrared spectroscopy as a valuable tool
for qualitative and quantitative analysis [174],
[175] has experienced a revival through the de-
velopment of chemometric methods and the in-
troduction of fiber optics during the 1980s, and
high-quality near-infrared spectra are provided
by improved conventional and by Fourier trans-
form instruments.
Various aspects differentiate a mid-infrared
spectrum from the corresponding near-infrared
40 Infrared and Raman Spectroscopy
spectrum (Table 6). In the mid-infrared region
fundamental molecular vibrations always oc-
cur between 4000 and 200 cm−1 , while in the
near-infrared region overtones and combina-
tion bands of these fundamentals are observed
between 12 800 and 4000 cm−1 . Near-infrared
bands are due primarily to hydrogenic stretches
of C−H, N−H, and O−H bonds whose funda-
mental vibrations give rise to strong bands bet-
ween 4000 and 2000 cm−1 , so that their over-
tones and combinations occur above 4000 cm−1
as the most intense absorption bands in the near-
infrared spectrum. Since the near-infrared ab-
sorptions of polyatomic molecules thus mainly
reflect vibrational contributions from very few
functional groups, near-infrared spectroscopy is
less suitable for detailed qualitative analysis than
mid-infrared spectroscopy, which shows all (ac-
tive) fundamentals and the overtones and com-
binations of low-energy vibrations. The near-
infrared overtone and combination bands de-
pend more on their environment than does the
fundamental of the same vibration; slight per-
turbation in the bonding scheme causes only
small changes in the fundamental but signifi-
cant frequency shifts and amplitude changes in
the near-infrared. In going from the fundamental
to the first overtone the intensity of an absorp-
tion band decreases by a factor of about 10 –
100, so that the sensitivity of near-infrared spec-
troscopy is reduced in comparison with the mid-
infrared. This is a disadvantage when gases are
to be measured, but for liquids, it is a consider-
able advantage as regards sample handling be-
cause cells with convenient path lengths between
1 mm and 10 cm can be used. Glass or quartz is
used as window material. Near-infrared bands
of liquids and solids have relatively large band-
widths between 30 and 60 cm−1 , so that they
strongly overlap and direct assignment of bands
is generally not possible for larger, complex
molecules. This is why, in the near-infrared, easy
quantitative analysis using isolated bands, as in
the mid-infrared, is not possible. Chemometrics
[37], [176–178], however, by the application of
mathematical and statistical procedures, gener-
ates correlations between experimental data and
the chemical composition or physical properties
of the sample, and can be used in a general man-
ner to solve quantitative and qualitative analyti-
cal problems.
Table 6. Comparison of mid-infrared and near-infrared
spectroscopy
Mid-Infrared Near-Infrared
4000 – 200 cm−1 12 800 – 4000 cm−1
Fundamentals
High intensity
overtones and combinations
low intensity
High sensitivity (trace analysis low sensitivity (not suited for
possible) trace analysis)
Sharp, separated bands strongly overlapping bands
Easy quantitation with isolated quantitation complex,
bands chemometrics necessary
FT-instruments grating, filter, FT and
acousto-optical tunable scanning
instruments
6.2. Applications of Near-Infrared
Spectroscopy
Applications of near-infrared spectroscopy in-
clude chemistry [179], [180], the oil industry
[181], clinical analysis [182], biological and
medical analysis [183], and the pharmaceutical
industry [184], [185].
Since the intensities of characteristic near-
infrared bands are independent of or only
slightly dependent on the state of the system,
near-infrared spectroscopy is widely applicable
for the quantitative study of liquid and com-
pressed gaseous systems, including fluid sys-
tems, up to high pressures and temperatures.
The application of near-infrared data to rou-
tine quantitative analysis was initiated by Nor-
ris [186], who used diffuse reflectance measure-
ments to quantitatively determine major com-
ponents, such as moisture, protein and fat, in
agricultural commodities. In comparison with
mid-infrared, near-infrared diffuse reflectance
[187] is able to measure powdered samples
with minimal sample preparation, and lends it-
self extremely well to quantitative analysis be-
cause the smaller absorptivities and larger scat-
tering coefficients at shorter wavelengths result
in a larger scattering/absorption ratio. Although
near-infrared reflectance analysis was initially
developed for agricultural and food applications
[188], in combination with chemometrics it is
now applied to many other areas, e.g., poly-
mers [189], pharmaceuticals [190], organic sub-
stances [191], geological samples [192], and
thin-layer chromatography [193], [194].
 Infrared and Raman Spectroscopy
The excellent transmittance of quartz in the
near-infrared range has led to a further en-
hancement of the potential of near-infrared spec-
troscopy by the introduction of fiber optics.
Fiber-optic waveguides are used to transfer light
from the spectrometer to the sample and, af-
ter transmission or reflection, back to the spec-
trometer. Most fiber optic cables consist of three
concentric components: the inner part, through
which the light propagates, is called the core,
the middle layer is the cladding, and the outer
protective layer is the jacket. Crucial to the light
throughput of the fiber optic cable are the op-
tical properties of the core and cladding. The
difference in refractive index of the two ma-
terials defines the highest angle at which the
core/cladding interface exhibits total internal re-
flection, and hence allows the light to propagate
through the cable. This angle is related to the
numerical aperture (N.A.) of the fiber by the fol-
lowing equation:
 1/2
sinα= N.A. = n2core −n2cladding (20)
where α is the half-angle of acceptance and n is
the refractive index of the material. The greater
the difference in refractive index the greater the
half-angle of acceptance. Any light entering at
an angle greater than the half-angle of accep-
tance is lost to the cladding through absorption.
The core and cladding are usually of quartz (of
different refractive index), while the outer pro-
tective layer is a polymer; other fiber optic ma-
terials are zirconium fluoride (ZrF) and chalco-
genide (AsGeSe) [195]. The obvious advantage
of the optical fiber technique is that the loca-
tion of measurement can be separated from the
spectrometer by between two and several hun-
dred meters, depending on the length of the
waveguide. Since almost all practically relevant
substances show characteristic near-infrared ab-
sorption bands, quantitative analysis by near-
infrared spectroscopy is generally applicable to
on-line concentration measurements in connec-
tion with chemical reactions, chemical equilib-
ria, and phase equilibria. Various probes can be
integrated into different reactors or bypasses and
offer numerous remote-sensing and on-line pro-
cess control applications [196]. In the analysis of
hazardous or toxic materials this has proved of
outstanding importance. In the refining, petro-
chemical, polymer, plastics, and chemicals in-
41
dustries, typical applications of near-infrared re-
mote spectroscopy are the determination of the
octane number and oxygenates in gasoline, aro-
matics in fuels, the composition of solvent mix-
tures [197], [198], the hydroxyl number of poly-
ols, low contents of water in solvents, and for
polymer analysis.
7. References
1. W. W. Coblentz: Investigations of Infrared
Spectra, Carnegie Institution, Washington
1905, re-published, The Coblentz Society,
Norwalk 1962.
2. G. Herzberg: Molecular Spectra and
Molecular Structure I, Van Nostrand,
Princeton 1953.
3. G. Herzberg: Molecular Spectra and
Molecular Structure II, Van Nostrand,
Princeton 1960.
4. E. B. Wilson, Jr., J. C. Decius, P. C. Cross:
Molecular Vibrations, McGraw-Hill, New
York 1955, re-published Dover, Mineola 1980.
5. D. Steele: Theory of Vibrational Spectroscopy,
Saunders, Philadelphia 1971.
6. J. A. Koningstein: Introduction to the Theory
of the Raman Effect, Reidel, Dordrecht 1972.
7. D. A. Long: Raman Spectroscopy,
McGraw-Hill, New York 1977.
8. P. R. Griffiths, J. A. de Haseth: Fourier
Transform Infrared Spectroscopy, J. Wiley &
Sons, New York 1986.
9. M. W. Mackenzie: Advances in Applied
Fourier Transform Infrared Spectroscopy, J.
Wiley & Sons, New York 1988.
10. N. B. Colthup, L. H. Daly, S. E. Wiberley:
Introduction to Infrared and Raman
Spectroscopy, 3rd ed., Academic Press, San
Diego 1990.
11. B. Schrader (ed.): Infrared and Raman
Spectroscopy, VCH Verlagsgesellschaft,
Weinheim 1995.
12. K. W. F. Kohlrausch: Ramanspektren,
Akademische Verlagsgesellschaft, Leipzig
1943, re-published Heyden & Son, London
1973.
13. H. Baranska, A. Labudzinska, T. Terpinski:
Laser Raman Spectroscopy, J. Wiley & Sons,
New York 1983.
14. P. Hendra, C. Jones, G. Warnes: Fourier
Transform Raman Spectroscopy, Ellis
Horwood, Chichester 1991.
42 Infrared and Raman Spectroscopy
15. J. R. Ferraro, K. Krishnan (eds.): Practical
Fourier Transform Infrared Spectroscopy,
Academic Press, San Diego 1990.
16. M. J. Pelletier (ed.): Analytical Applications of
Raman Spectroscopy, Blackwell Science,
Oxford 1999.
17. J. G. Grasselli, B. J. Bulkin (eds.): Analytical
Raman Spectroscopy, J. Wiley & Sons, New
York 1991.
18. G. Socrates: Infrared Characteristic Group
Frequencies, J. Wiley & Sons, New York
1980.
19. L. J. Bellamy: The Infrared Spectra of
Complex Molecules, 3rd ed., vol. 1, Chapman
& Hall, London 1975.
20. L. J. Bellamy: The Infrared Spectra of
Complex Molecules, 2nd ed., vol. 2, Chapman
& Hall, London 1980.
21. F. R. Dollish, W. G. Fateley, F. F. Bentley:
Characteristic Raman Frequencies of Organic
Compounds, Wiley Interscience, New York
1974.
22. D. Lin-Vien, N. B. Colthup, W. G. Fateley,
J. G. Grasselli: The Handbook of Infrared and
Raman Characteristic Frequencies of Organic
Molecules, Academic Press, Boston 1991.
23. K. Nakamoto: Infrared and Raman
Spectroscopy of Inorganic and Coordination
Compounds, 5th ed., J. Wiley & Sons,
Chichester 1997.
24. H.-U. Gremlich, B. Yan (eds.): Infrared and
Raman Spectroscopy of Biological Materials,
Marcel Dekker, New York 2000.
25. B. Schrader: Raman/Infrared Atlas of Organic
Compounds, 2nd ed., VCH
Verlagsgesellschaft, Weinheim 1989.
26. Sadtler IR Digital Spectra Libraries, Sadtler
Division, Bio-Rad Laboratories, Inc., 3316
Spring Garden Street, Philadelphia, PA 19104.
27. F. M. Mirabella (ed.): Modern Techniques in
Applied Molecular Spectroscopy, J. Wiley &
Sons, New York 1998.
28. A. A. Michelson, Philos. Mag. 31 (1891)
256 – 259.
29. W. Herres, J. Gronholz, Comput. Appl. Lab. 4
(1984) 216 – 220.
30. J. W. Cooley, J. W. Tukey, Math. Comput. 19
(1965) 297 – 301.
31. J. Gronholz, W. Herres, Instrum. Computers 3
(1985) 10 – 19.
32. B. Schrader, A. Hoffmann, A. Simon, J.
Sawatzki, Vib. Spectrosc. 1 (1991) 239 – 250.
33. B. Schrader in J. R. Ferraro, K. Krishnan
(eds.): Practical Fourier Transform Infrared
Spectroscopy, Academic Press, San Diego
1990, p. 167.
34. B. Schrader, A. Hoffmann, S. Keller,
Spectrochim. Acta 47 A (1991) 1135 – 1148.
35. J. Coates in W. O. George, H. A. Willis (eds.):
Computer Methods in UV, Visible and IR
Spectroscopy, The Royal Society of
Chemistry, Cambridge 1990, p. 95.
36. P. S. Wilson in W. O. George, H. A. Willis
(eds.): Computer Methods in UV, Visible and
IR Spectroscopy, The Royal Society of
Chemistry, Cambridge 1990, p. 143.
37. F. Despagne, D. L. Massart, Analyst 123
(1998) 157R – 178R.
38. G. M. Barrow: Introduction to Molecular
Spectroscopy, McGraw-Hill, New York 1962.
39. S. R. La Paglia: Introductory Quantum
Chemistry, Harper & Row, New York 1971.
40. Ullmann, 4th ed., 5, 303.
41. Available from Sadtler Division, Bio-Rad
Laboratories, Inc., 3316 Spring Garden Street,
Philadelphia, PA 19104.
42. Available from LabControl GmbH, D-50858
Köln 40, Max-Planck-Str. 17 a, Germany.
43. Available from Galactic Industries
Corporation, 395 Main Street, Salem NH
03079.
44. J. P. Coates, R. W. Hannah in T. Theophanides
(ed.): Fourier Transform Infrared
Spectroscopy, D. Reidel Publ., Dordrecht
1984, p. 167.
45. S. R. Lowry, D. A. Hupplar, Anal. Chem. 53
(1981) 889 – 893.
46. J. T. Clerc in W. O. George, H. A. Willis (eds.):
Computer Methods in UV, Visible and IR
Spectroscopy, The Royal Society of
Chemistry, Cambridge 1990, p. 13.
47. D. A. Cirovic, Trends Anal. Chem. 16 (1997)
148 – 155.
48. H. Yang, P. R. Griffiths, Anal. Chem. 71
(1999) 751 – 761.
49. J. U. White, J. Opt. Soc. Am. 32 (1942)
285 – 288.
50. R. T. Graf, J. L. Koenig, H. Ishida, Appl.
Spectrosc. 39 (1985) 405 – 412.
51. R. G. Greenler, J. Chem. Phys. 44 (1966)
310 – 315.
52. R. G. Greenler, J. Chem. Phys. 50 (1969)
1963 – 1968.
53. D. L. Allara, J. D. Swalen, J. Phys. Chem. 86
(1982) 2700 – 2704.
54. N. J. Harrick, Appl. Opt. 10 (1971)
2344 – 2349.
55. I. Newton: Opticks, Dover, New York 1952.
 Infrared and Raman Spectroscopy
56. N. J. Harrick: Internal Reflection Spectroscopy,
Harrick Scientific Corporation, Ossining, New
York 1979.
57. J. Fahrenfort in A. Mangini (ed.): Adv. Mol.
Spectrosc. Proc. Int. Meet. 4th 1959, vol. 2,
1962, p. 701.
58. Available from Harrick Scientific Corporation,
88 Broadway, Ossining, NY 10562.
59. N. J. Harrick, M. Milosevic, S. L. Berets, Appl.
Spectrosc. 45 (1991) 944 – 948.
60. S. Compton, P. Stout, FTS 7/IR Notes no. 1,
Bio-Rad, Digilab Division, Cambridge, MA
02139, 1991.
61. U. P. Fringeli in F. M. Mirabella (ed.): Internal
Reflection Spectroscopy: Theory and
Application, Marcel Dekker, New York 1992,
p. 255.
62. U. P. Fringeli, Chimia 46 (1992) 200 – 214.
63. Available from Spectra-Tech Inc., Stamford,
CT 06906.
64. Available from GRASEBY SPECAC Ltd.,
Lagoon Road, St. Mary Cray, Orpington, Kent
BR5 3QX, U.K.
65. W. M. Doyle, Appl. Spectrosc. 44 (1990)
50 – 69.
66. B. E. Miller, N. D. Danielson, J. E. Katon,
Appl. Spectrosc. 42 (1988) 401 – 405.
67. Available from Axiom Analytical, Inc.,
17751-C Sky Park Circle, Irvine, CA 92614.
68. W. M. Doyle, Process Control Quality 2
(1992) 11 –41.
69. N. L. Owen, D. W. Thomas, Appl. Spectrosc.
43 (1989) 451 – 455.
70. P. Kubelka, F. Munk, Z. Tech. Phys. 12 (1931)
593 –601.
71. P. Kortüm, W. Braun, G. Herzog, Angew.
Chem., Int. Ed. Engl. 2 (1963) 333 – 341.
72. P. H. Turner, W. Herres, BRUKER FT-IR
Application Note 23, Bruker Analytische
Meßtechnik GmbH, D-76153 Karlsruhe,
Germany.
73. D. Reinecke, A. Jansen, F. Fister, U. Schernau,
Anal. Chem. 60 (1988) 1221 – 1224.
74. M. P. Fuller, P. R. Griffiths, Anal. Chem. 50
(1978) 1906 – 1910.
75. M. Fuller in Nicolet FT-IR Spectral Lines,
Nicolet Analytical Instruments, Madison, WI
53711, Spring/Summer 1990, p. 15.
76. G. Glauninger, K.-A. Kovar, V. Hoffmann,
100. C. Q. Yang, Appl. Spectrosc. 45 (1991)
Fresenius J. Anal. Chem. 338 (1990)
710 – 716.
101. J. C. Donini, K. H. Michaelian, Infrared Phys.
77. H. Yamamoto, K. Wada, T. Tajima, K.
Ichimura, Appl. Spectrosc. 45 (1991)
102. M. W. Urban, J. L. Koenig, Appl. Spectrosc.
253 – 259.
43
78. S. A. Johnson, R. M. Rinkus, T. C. Diebold,
V. A. Maroni, Appl. Spectrosc. 42 (1988)
1369 – 1375.
79. K. A. Martin, R. F. Zabransky, Appl. Spectrosc.
45 (1991) 68 – 72.
80. Si-Carb Sampler, available from Spectra-Tech
Inc., Stamford, CT 06906.
81. F. J. DeBlase, S. Compton, Appl. Spectrosc.
45 (1991) 611 – 618.
82. A. M. Vassallo, P. A. Cole-Clarke, L. S. K.
Pang, A. J. Palmisano, Appl. Spectrosc. 46
(1992) 73 – 78.
83. M. Handke, N. J. Harrick, Appl. Spectrosc. 40
(1986) 401 – 405.
84. R. T. Rewick, R. G. Messerschmidt, Appl.
Spectrosc. 45 (1991) 297 – 301.
85. L. T. Lin, D. D. Archibald, D. E. Honigs, Appl.
Spectrosc. 42 (1988) 477 – 483.
86. A. Tsuge, Y. Uwamino, T. Ishizuka, Appl.
Spectrosc. 43 (1989) 1145 – 1149.
87. R. W. Jones, J. F. McClelland, Anal. Chem. 61
(1989) 650 – 656.
88. R. W. Jones, J. F. McClelland, Anal. Chem. 61
(1989) 1810 – 1815.
89. R. J. Pell, B. C. Erickson, R. W. Hannah, J. B.
Callis, B. R. Kowalski, Anal. Chem. 60 (1988)
2824 –2827.
90. J. A. McGuire, B. Wangmaneerat, T. M.
Niemczyk, D. M. Haaland, Appl. Spectrosc.
46 (1992) 178 – 180.
91. B. Wangmaneerat et al., Appl. Spectrosc. 46
(1992) 340 – 344.
92. A. Rosencwaig: Photoacoustics and
Photoacoustic Spectroscopy, J. Wiley & Sons,
New York 1980.
93. J. F. McClelland, Anal. Chem. 55 (1983)
89 A – 105 A.
94. J. A. Graham, W. M. Grim III, W. G. Fateley in
J. R. Ferraro, L. J. Basile (eds.): Fourier
Transform Infrared Spectroscopy, vol. 4,
Academic Press, Orlando 1985, p. 345.
95. H. Coufal, J. F. McClelland, J. Mol. Struct.
173 (1988) 129 – 140.
96. G. Laufer, J. T. Huneke, B. S. H. Royce, Y. C.
Teng, Appl. Phys. Lett. 37 (1980) 517 – 519.
97. G. Busse, B. Bullemer, Infrared Phys. 18
(1978) 631 –634.
98. MTEC Photoacoustics, Inc., Ames, IA 50010.
99. R. O. Carter III, S. L. Wright, Appl. Spectrosc.
45 (1991) 1101 – 1103.
102 – 108.
24 (1984) 157 – 163.
40 (1986) 994 – 998.
44 Infrared and Raman Spectroscopy
103. C. Q. Yang, R. R. Bresee, W. G. Fateley, Appl.
Spectrosc. 41 (1987) 889 – 896.
104. R. Rubinovitz, J. Seebode, A. Simon, Bruker
Report 1 (1988) 35 – 38.
105. M. J. Smith, C. J. Manning, R. A. Palmer, J. L.
Chao, Appl. Spectrosc. 42 (1988) 546 – 555.
106. R. A. Crocombe, S. V. Compton, FTS/IR Notes
no. 82, Bio-Rad, Digilab Division, Cambridge,
MA 02139, 1991.
107. K. H. Michaelian, Appl. Spectrosc. 45 (1991)
302 –304.
108. L. Bertrand, Appl. Spectrosc. 42 (1988)
134 – 138.
109. R. White: “Chromatography/Fourier
Transform Infrared Spectroscopy and Its
Applications,” Practical Spectroscopy Series,
vol. 10, Marcel Dekker, New York 1990.
110. W. Herres: HRGC-FTIR: Capillary Gas
Chromatography-Fourier Transform Infrared
Spectroscopy, Hüthig Verlag, Heidelberg
1987.
111. S. Bourne et al., FTS/IR Notes no. 75,
Bio-Rad, Digilab Division, Cambridge, MA
02139, 1990.
112. R. L. White, G. N. Giss, G. M. Brissey, C. L.
Wilkins, Anal. Chem. 53 (1981) 1778 – 1782.
113. P. Coffey, D. R. Mattson, J. C. Wright, Am.
Lab. 10 (1978) 126 – 132.
114. K. S. Chiu, K. Biemann, K. Krishnan, S. L.
Hill, Anal. Chem. 56 (1984) 1610 – 1615.
115. Available from Lab Connections, Inc., 201
Forest Street, Marlborough, MA 01752.
116. Available from Bourne Scientific, Inc., 65
Bridge Street, Lexington, MA 02421-7927.
117. S. Bourne, Am. Lab. 16 (1998) 17F – 17J.
118. C. M. Earnest, Anal. Chem. 56 (1984)
1471 A –1486 A.
119. D. Dollimore, Anal. Chem. 64 (1992)
147 R – 153 R.
120. J. A. J. Jansen, J. H. van der Maas, A.
Posthuma de Boer, Appl. Spectrosc. 46 (1992)
88 – 92.
121. K. Ichimura, H. Ohta, T. Tajima, T. Okino,
Mikrochim. Acta 1 (1988) 157 – 161.
122. R. C. Wieboldt, S. R. Lowry, R. J. Rosenthal,
Mikrochim. Acta 1 (1988) 179 – 182.
123. J. Mullens et al., Bull. Soc. Chim. Belg. 101
(1992) 267 – 277.
124. R. M. Carangelo, P. R. Solomon, D. J. Gerson,
Fuel 66 (1987) 960 – 967.
125. J. K. Whelan, Energy Fuels 2 (1988) 65 – 73.
126. L. A. Sanchez, M. Y. Keating, Appl. Spectrosc.
42 (1988) 1253 – 1258.
127. M. D. Timken, J. K. Chen, T. B. Brill, Appl.
Spectrosc. 44 (1990) 701 – 706.
128. E. Post, S. Rahner, H. Möhler, A. Rager,
Thermochim. Acta 263 (1995) 1 – 6.
129. M. Maciejewski, F. Eigenmann, A. Baiker, E.
Dreser, A. Rager, Thermochim. Acta (2000) in
press.
130. L. A. Nafie in M. W. Mackenzie (ed.):
Advances in Applied FTIR Spectroscopy, John
Wiley and Sons, New York 1988, p. 67.
131. L. A. Nafie, T. B. Freedman in H.-U. Gremlich,
B. Yang (eds.): Infrared and Raman
Spectroscopy of Biological Materials, Marcel
Dekker, New York 2000, p. 15.
132. T. A. Keiderling in H.-U. Gremlich, B. Yang
(eds.): Infrared and Raman Spectroscopy of
Biological Materials, Marcel Dekker, New
York 2000, p. 55.
133. B. Schrader, S. Keller, Proc. SPIE Int. Soc.
Opt. Eng. 1575 (Int. Conf. Fourier Transform
Spectrosc., 8th, 1991) (1992) 30 – 39.
134. N. J. Everall, J. Lumsdon, Spectrosc. Eur. 4
(1992) 10 – 21.
135. C. G. Zimba, J. F. Rabolt, Appl. Spectrosc. 45
(1991) 162 – 165.
136. T. Jawhari, P. J. Hedra, H. A. Willis, M.
Judkins, Spectrochim. Acta 46 A (1990)
161 – 170.
137. F. T. Walder, M. J. Smith, Spectrochim. Acta
47 A (1991) 1201 – 1216.
138. H. Sadeghi-Jorabchi, R. H. Wilson, P. S.
Belton, J. D. Edwards-Webb, D. T. Coxon,
Spectrochim. Acta 47 A (1991) 1449 – 1458.
139. L. A Lyon et al., Anal. Chem. 70 (1998)
341R – 361R.
140. B. Chase in A. M. C. Davies, C. S. Creaser
(eds.): Anal. Appl. Spectrosc. 2, The Royal
Society of Chemistry, Cambridge 1991, p. 13.
141. K. P. J. Williams, S. M. Mason, TRAC 9
(1990) 119 –127.
142. P. J. Hendra, J. Mol. Struct. 266 (1992)
97 –114.
143. A. Hoffmann et al., J. Raman Spectrosc. 22
(1991) 497 – 503.
144. P. J. Hendra, C. H. Jones, Makromol. Chem.
Macromol. Symp. 52 (1991) 41 – 56.
145. G. J. Thomas, Ann. Rev. Biophys. Biomol.
Struct. 28 (1999) 1 – 27.
146. A. M. Tudor et al., J. Pharm. Biomed. Anal. 8
(1990) 717 – 720.
147. P. R. Carey, J. Biol. Chem. 274 (1999)
26 625 – 26 628.
148. J. Góral, V. Zichy, Spectrochim. Acta 46 A
(1990) 253 – 275.
149. I. W. Levin, E. N. Lewis, Anal. Chem. 62
(1990) 1101 A – 1111 A.
 Infrared and Raman Spectroscopy
150. S. Nie et al., Spectroscopy 5 (1990) 24 – 32.
151. G. Ellis, M. Claybourn, S. E. Richards,
Spectrochim. Acta 46 A (1990) 227 – 241.
152. D. Bourgeois, S. P. Church, Spectrochim. Acta
46 A (1990) 295 – 301.
153. T. N. Day, P. J. Hendra, A. J. Rest, A. J.
Rowlands, Spectrochim. Acta 47 A (1991)
1251 – 1262.
154. N. J. Everall, J. M. Chalmers, I. D. Newton,
Appl. Spectrosc. 46 (1992) 597 – 601.
155. D. L. Gerrard, J. Birnie, Anal. Chem. 64
(1992) 502 R – 513 R.
156. B. Yan, Acc. Chem. Res. 31 (1998) 621 – 630.
157. R. G. Messerschmidt, M. A. Harthcock (eds.):
Infrared Microspectroscopy: Theory and
Applications, Marcel Dekker, New York 1988.
158. K. Krishnan, S. L. Hill in J. R. Ferraro, K.
Krishnan (eds.): Practical Fourier Transform
Spectroscopy, Academic Press, San Diego
1990, p. 103.
159. S. Compton, J. Powell, FTS 7/IR Notes no. 2,
Bio-Rad, Digilab Division, Cambridge, MA
02139, 1991.
160. N. J. Harrick, M. Milosevic, S. L. Berets, Appl.
Spectrosc. 45 (1991) 944 – 948.
161. B. Schrader, Fresenius J. Anal. Chem. 337
(1990) 824 – 829.
162. J. D. Louden in D. J. Gardiner, P. R. Graves
(eds.): Practical Raman Spectroscopy,
Springer Verlag, Berlin 1989, p. 119.
163. J. Sawatzki, Fresenius J. Anal. Chem. 339
(1991) 267 –270.
164. A. Hoffmann, B. Schrader, R. Podschadlowski,
A. Simon, Proc. SPIE-Int. Soc. Opt. Eng. 1145
(Int. Conf. Fourier Transform Spectrosc., 7th,
1989) (1989) 372 – 373.
165. F. Bergin, Proc. Annu. Conf. Microbeam Anal.
Soc. 25 (1990) 235 – 239.
166. B. W. Cook in A. M. C. Davies, C. S. Creaser
(eds.): Anal. Appl. Spectrosc. 2, The Royal
Society of Chemistry, Cambridge 1991, p. 61.
167. A. J. Sommer, J. E. Katon, Appl. Spectrosc. 45
(1991) 527 – 534.
168. D. Gernet, W. Kiefer, H. Schmidt, Appl.
Spectrosc. 46 (1992) 571 – 576.
169. K. Gerwert in H.-U. Gremlich, B. Yang (eds.):
Infrared and Raman Spectroscopy of
Biological Materials, Marcel Dekker, New
York 2000, p. 193.
170. M. D. Schaeberle, I. W. Levin, E. N. Lewis in
H.-U. Gremlich, B. Yang (eds.): Infrared and
Raman Spectroscopy of Biological Materials,
Marcel Dekker, New York 2000, p. 231.
171. D. O. Hummel: Atlas of Polymer and Plastics
Analysis, 3rd ed., VCH Verlagsgesellschaft,
Weinheim 1991.
45
172. J. L. Koenig: Spectroscopy of Polmers,
American Chemical Society, Washington,
D.C. 1992.
173. G. Zerbi et al.: Modern Polymer Spectroscopy,
Wiley-VCH, Weinheim 1998.
174. J. Workman, Jr., J. Near Infrared Spectrosc. 1
(1993) 221.
175. I. Murray, I. A. Cowe (eds.): Making Light
Work: Advances in Near Infrared
Spectroscopy, VCH Verlagsgesellschaft,
Weinheim 1992.
176. P. Geladi, E. Dabbak, J. Near Infrared
Spectrosc. 3 (1995) 119.
177. D. L. Massart, B. G. M. Vandeginste, L. M. C.
Buydens, S. de Jong, P. J. Lewi, J.
Smeyers-Verbecke: Handbook of
Chemometrics and Qualimetrics, Part A,
Elsevier, Amsterdam 1997.
178. D. L. Massart, B. G. M. Vandeginste, L. M. C.
Buydens, S. de Jong, P. J. Lewi, J.
Smeyers-Verbecke: Handbook of
Chemometrics and Qualimetrics, Part B,
Elsevier, Amsterdam 1998.
179. S. M. Donahue, C. W. Brown, B. Caputo, M. D.
Modell, Anal. Chem. 60 (1988) 1873 – 1878.
180. J. Lin, C. W. Brown, Vibr. Spectrosc. 7 (1994)
117 – 123.
181. S. R. Westbrook, SAE Tech. Paper Ser. 930734
(1993) 1 – 10.
182. H. M. Heise in H.-U. Gremlich, B. Yan (eds.):
Infrared and Raman Spectroscopy of
Biological Materials, Marcel Dekker, New
York 2000, p. 259.
183. R. J. Dempsey, D. G. Davis, R. G. Buice, Jr.,
R. A. Lodder, Appl. Spectrosc. 50 (1996)
18A – 34A.
184. J. D. Kirsch, J. K. Drennen, Appl. Spectrosc.
Rev. 30 (1995) 139 – 174.
185. W. Plugge, C. van der Vlies, J. Pharm.
Biomed. Anal. 14 (1996) 891 – 898.
186. I. Ben-Gera, K. H. Norris, Israel J. Agric. Res.
18 (1968) 125 – 132.
187. D. L. Wetzel, Anal. Chem. 55 (1983)
1165 A – 1176 A.
188. P. Williams, K. Norris (eds.): Near-Infrared
Technology in the Agricultural and Food
Industries, AACC, St. Paul 1987.
189. C. E. Miller, Appl. Spectrosc. Rev. 26 (1991)
277 –339.
190. B. F. MacDonald, K. A. Prebble, J. Pharm.
Biomed. Anal. 11 (1993) 1077 – 1085.
191. L. G. Weyer, Appl. Spectrosc. Rev. 21 (1985)
1 – 43.
46 Infrared and Raman Spectroscopy
192. D. E. Honigs, T. B. Hirschfeld, G. M. Hieftje,
Appl. Spectrosc. 39 (1985) 1062 – 1065.
193. E. W. Ciurczak, L. J. Cline-Love, D. M.
Mustillo, Spectrosc. Int. 3 (1991) no. 5,
39 – 42.
194. E. W. Ciurczak, W. R. Murphy, D. M. Mustillo,
Spectrosc. Int. 3 (1991) no. 7, 39 – 44.
Ink-Jet Printing → Imaging Technology
Inks → Drawing and Writing Materials
Inorganic Fibers → Fibers, 5. Synthetic Inorganic
195. M. J. Smith, T. E. May, Int. Lab. 22 (1992)
18 – 24.
196. L. P. McDermott, Ad. Instrum. Control. 45
(1990) 669 – 677.
197. Guided Wave, Inc., Application Note
no. A3-987, El Dorado Hills, CA 95630.
198. Guided Wave, Inc., Application Note
no. A4-188, El Dorado Hills, CA 95630.

Inorganic Polymers
Charles E. Carraher, Jr., College of Science, Florida Atlantic University, Boca Raton, Florida 33431,
United States Florida Center for Environmental Studies, Palm Beach Gardens, Florida 33410, United States
Charles U. Pittman, Jr., Department of Chemistry, Mississippi State University, Mississippi State,
Mississippi 39762, United States
1. Introduction . . . . . . . . . . . . . . . . .
2. Polysilanes, Polygermanes, and
Polystannanes . . . . . . . . . . . . . . . .
3. Poly(Carborane – Siloxanes), Polycar-
bosilanes and Polysilazanes . . . . . . .
4. Polyphosphazenes . . . . . . . . . . . . .
5. Poly(Boron Nitride) . . . . . . . . . . . .
6. Poly(Sulfur Nitride) . . . . . . . . . . . .
7. Polysulfur . . . . . . . . . . . . . . . . . . .
8. Organometallic Polymers . . . . . . . . .
9. Sol – Gel Inorganic Polymers . . . . . .
1. Introduction
Inorganic polymers are macromolecules linked
by covalent bonds, whereby there is an ab-
sence or near-absence of organic units within
the backbone. Inorganic polymers include many
of the most important “natural” materials as
well as many synthetic materials. Natural in-
organic polymers include many of the rocks
about us, diamond, graphite, sulfur, boric ox-
ide, silica, polyphosphates, quartz, and glass.
Sulfur, selenium, and tellurium form macro-
molecules. Chalcogenide glasses are three-
dimensional polymers, as are the asbestoses and
a number of silicates. Most of these are in-
dustrially important, and some exhibit proper-
ties that are utilized in high-technology appli-
cations. Thus, arsenic sulfide, a chalcogenide
glass, is used as an infrared-transparent win-
dow, and modified zeolites as selective chelating
agents and catalysts.
Many scientific societies, including the Na-
tional Research Council [1], have emphasized
that the major hindrance to progress in most
technological areas is the lack of suitable ma-
terials. The search for new materials in the
1970s and 1980s generally focused on rearrang-
ing, mixing, blending, and copolymerizing well-
known organic polymers. Although great suc-
cess has been achieved, many new applications

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10.1002/14356007.a14 241
Inorganic Polymers 1
1 10. Inorganic Fibers and Whiskers . . . . . 14
11. Biomedical and Antimicrobial Poly-
2 mers . . . . . . . . . . . . . . . . . . . . . . 15
12. Boron Polymers . . . . . . . . . . . . . . . 17
4 13. Aluminum Polymers . . . . . . . . . . . . 19
6 14. Polymers with −Si−O−M−O− Back-
7 bones . . . . . . . . . . . . . . . . . . . . . . 20
8 15. Poly(Carbon Disulfide), Poly(Carbon
9 Diselenide), and Polythiocyanogen . . . 21
10 16. Phosphorus-Containing Polymers . . . 23
11 17. References . . . . . . . . . . . . . . . . . . 24
require materials whose performance greatly ex-
ceeds that of currently available materials. Such
performance requirements can be achieved only
by materials containing other elements in ad-
dition to those found in the usual carbon-based
polymers. For example, although the first polysi-
lane, poly(dimethylsilylene), was prepared in
1924 by Kipping [2], the initial products were
infusible and insoluble. In fact, these properties
were so unfavorable that polysilanes were not
developed until more than five decades later. To-
day, polysilanes offer combinations of chemical,
mechanical, and electrical properties not com-
mon to any other commercially available class
of materials.
Major applications of inorganic polymers in-
clude their use as bulk and specialty building
materials, biologically inert and biologically ac-
tive materials, high-strength materials, catalysts,
and speciality coatings. The breadth of the field
in relation to more traditional polymer chem-
istry is exemplified by comparing the number
and variety of elements typically found in com-
mercial organic polymers (C, H, N, Cl, F, O)
with the number and variety of readily available
inorganic materials. The relative abundance of
elements in the earth’s crust also attests to the im-
portance of noncarbon-based materials. In fact,
the most abundant natural building materials are
2 Inorganic Polymers
based on silicon – oxygen polymers, including
the majority of rocks, soil, and sand.
The alternative of utilizing heteroatom back-
bones or noncarbon-based polymers is attractive
because of the great variety of available reac-
tants. Thus, a wide range of compounds can be
exploited to obtain materials with specific prop-
erties. Furthermore, these materials are of inter-
est because the resulting bonds may have bond
energies greater than that of the C−C single
bond in alkanes (ca. 343 kJ/mol). Approximate
bond energies for other elements, in kilojoules
per mole, are B−N 389, Si−N 439, Si−O 795,
P−N 615, and Be−O 448. Bonding in these ma-
terials is a combination of ionic (nondirectional)
and covalent (directional) components; the ionic
character increases as the electronegativity dif-
ference between adjacent atoms increases. Fur-
thermore, the tendency to form multiple bonds
is high for many of the elements involved. The
thermal stability of many inorganic and organo-
metallic materials has been described [15].
Inorganic and organometallic polymers re-
present a rapidly growing field of chemical re-
search and already have many applications [16–
26]. Any division between inorganic and organ-
ometallic polymers is necessarily somewhat ar-
bitrary. The classification used here is conve-
nient but far from perfect. The reader is there-
fore referred to several other articles which re-
view the scope of organometallic polymers and
discuss many other systems which could also be
regarded as inorganic polymers [20], [22–27].
2. Polysilanes, Polygermanes, and
Polystannanes
Carbon has an unique ability to form strong sin-
gle, double, and triple bonds to itself. Because
of the relatively high bond energies (C–C sin-
gle bonds are of the order of 340 kJ/mol) com-
pounds containing carbon in their backbone do
not readily cleave. By comparison, fellow mem-
bers of group 14 have only recently been synthe-
sized because of the low bond energies (gener-
ally on the order of 85 – 125 kJ/mol). Polysilanes
may have been prepared by Kipping in 1924 [2].
Only recently have soluble, well-characterized
polysilanes been synthesized [3–5]. (Topics re-
lated to polysilanes have been recently reviewed
[6–14].)
Production. Polysilanes have been synthe-
sized by four methods:
1) Reductive coupling
2) Dehydrogenative coupling
3) “Masked” dienes
4) Anionic ring opening
Attempts to prepare polysilanes in the late
1940s and 1950s gave insoluble, intractable ma-
terials; the major route employed was the reac-
tion of dichlorosilanes with sodium (Fig. 1). For
example, polydimethylsilane [28883-63-8] was
investigated at Union Carbide but found to be
insoluble [28]. The key developments leading
to soluble polysilanes were the introduction of
large substituent groups to decrease order, the
use of two different substituents, and the co-
polymerization of different monomers (Fig. 1)
[29], [30]. This resulted in high molecular mass
(> 400 000), tractable, soluble polymers, which
could be purified easily, cast into films, and
molded.
Figure 1. Synthesis of polysilanes
Improved methods for preparing high molec-
ular mass polysilanes are being sought, because
the reaction of dichlorosilanes with sodium fre-
quently gives low yields of high molecular mass
polymers and because binodal molecular mass
distributions are obtained [28], [29],[31–35].
Carrying out the coupling reaction at low tem-
perature by using ultrasonication gives monon-
odal polymers with low polydispersity [35].
Alternating copolymers are formed from the
coupling of dichlorodisilanes with dilithium
salts of 1,2-diethynyldisilanes [36] or dilithio

polythiophenes [36], [37]. These products ex-
hibit σ – π conjugation. Some of these materi-
als also undergo solid-state transitions to liquid-
crystalline mesophases [36].
Water-soluble products are formed by cou-
pling silanes with oligomeric poly(ethylene gly-
col) side chains [38].
Dehydrogenative coupling of dialkyl and
monoalkyl silanes in the presence of transi-
tion metal catalysts such as titanocene and zir-
conocene dialkyls has given polysilylenes [36–
48]. Functionalized polysilanes have also been
produced. When the side chains contain amine
functional groups, the products can be made
water-soluble by formation of an amine salt net-
work [49–51]. Dendrimeric [52–54] products
can also be formed by dehydrogenative cou-
pling. These products exhibit σ conjugation in
many directions and show good conductivity.
They have been used as negative-type photore-
sists because exposure to the air or light gives
insoluble products.
Polygermanes and polystannanes have also
been prepared by similar routes [55–58]. Many
of these materials show better σ conductivity
than the corresponding polysilanes and have
been used as light-emitting diodes [59].
Properties and Uses. Much of the interest
in the polysilanes, polygermanes, and polystan-
nanes involves their σ delocalization and their
σ – π delocalization when coupled with arenes
or acetylenes. This is not unexpected since sil-
icon exists as a covalent network similar to
diamond. In exhibiting electrical conductivity,
germanium and tin show more typical “metal-
lic” bonding. Some polystannanes have been re-
ferred to as “molecular metals” [60]. Conduc-
tivity is increased by doping [61], illumination
[62], and application of an electric field.
Because of a number of interesting electronic
and physical properties exhibited by polysilanes,
a number of potential uses have been suggested
and/or shown to exist, including precursors of
β-SiC fibers [63–71], impregnation of ceramics
[71], [72], polymerization initiators [73], [74],
photoconductors for electrophotography [75–
78], nonlinear optical materials [79–87], mid-
UV photolithography bilayer materials [88–91],
contrast enhancement layers in photolithogra-
phy [92], deep UV-sensitive photoresists [90],
[93–100], and self-developing by excimer laser
Inorganic Polymers 3
or deep UV exposure [99–101]. The unusual ab-
sorption spectra of polysilanes have indicated
potential uses in a number of conducting areas.
The copolymer of polysilastyrene, when doped
with arsenic pentafluoride, becomes a semicon-
ductor. Conducting films have been reported
[102].
Polysilanes display UV absorption due to sil-
icon – silicon bonds of the backbone. For exam-
ple, poly(cyclohexylmethylsilane) exhibits an
absorption band at 326 nm [30]. This absorption
is not found in saturated carbon chains. Polysi-
lanes with phenyl side chains show strong ab-
sorption near 330 nm, resulting from interaction
between the phenyl groups and the silicon back-
bone, which acts as a σ → σ∗ or σ → π∗ chro-
mophore.
Polysilanes have a high sensitivity to degra-
dation by UV light. After light is absorbed, chain
cleavage occurs with measured quantum yields
of 0.20 – 0.97. This has led IBM to examine
polysilanes as resists for bilayer UV lithogra-
phy [103–105]. They serve as excellent reactive-
ion etching barriers for bilevel resist applications
because a protective layer of silicon dioxide is
formed during exposure of the polymers to oxy-
gen plasmas [104], [105].
The wavelength at which maximum absorp-
tion occurs (λmax ) depends on the polysi-
lane chain length. During irradiation nonlinear
bleaching (decrease in UV absorption) occurs,
because the polymer undergoes chain cleavage
to give shorter fragments of lower absorptiv-
ity. As the irradiation continues, the polysilanes
become increasingly transparent. This unique
bleaching latency is useful in contrast-enhanced
lithography [104], [105].
The λmax for UV absorption also depends
on the conformation of the polysilane backbone
[107], [108]. Polysilanes having a planar zigzag
backbone conformation absorb at much longer
wavelengths than those in which the backbone
is disordered or helical. The solid materials are
strongly thermochromic [107]. For example, the
λmax of poly(di-n-hexylsilane) in solution or im-
mediately after baking at 100 ◦ C is 317 nm, but
after 3 h at 21 ◦ C films exhibit λmax = 371 nm.
Polysilanes are excellent photocatalysts for
making high molecular mass vinyl polymers.
On irradiation, chain cleavage occurs to give
free radicals [106], which initiate vinyl poly-
merization and curing. The 3M Company has
4 Inorganic Polymers
commercialized this process. Polysilane initia-
tors are relatively insensitive to termination of
polymerization by oxygen.
Pioneering work by Yajima, [109] showed
that poly(dimethylsilane) can be converted into
β-SiC fibers with very high tensile strength
(350 kg/mm2 ) by a series of pyrolysis steps.
Thermolysis of poly(dimethylsilane) at 400 –
450 ◦ C gives a poly(carbosilane) which is frac-
tionated and then melt-spun into fibers. The re-
sulting fibers are oxidized at the surface in air at
350 ◦ C (cross-linking occurs) to provide rigid-
ity, then pyrolyzed at 1200 ◦ C under nitrogen
to give crystalline β-SiC (see also → Fibers, 5.
Synthetic Inorganic, Chap. 3.2.2.).
Objects have been manufactured from sil-
icon carbide by Shin Nisso Kako using
poly(silastyrene):
These materials contain 20 – 30 vol% of mi-
crovoids, which lower the ultimate strength
slightly but also inhibit crack growth, so that
they can be machined to exact dimensions.
Delocalization in the Si−Si σ-bond frame-
work makes polysilanes potential electrical con-
ductors. Although neutral polysilanes are in-
sulators, they become semiconductors when
doped with AsF6 or SbF5 [31]. Conductivi-
ties up to 0.5 Ω−1 cm−1 have been measured.
Polysilanes are excellent photoconductors op-
erating through hole migration with high hole
mobilities (1014 cm2 V−1 s−1 at ambient tem-
perature) [110]. This has possible applications in
electrophotography and communications tech-
nology. The discovery that poly(methylphe-
nylsilane) has nonlinear optical properties sug-
gests eventual applications in laser technology
[111].
A variety of linear, two- and three-
dimensional polysilanes are under investiga-
tion for use as specialty caulks, adhesives, oils,
sealants, viscosity regulators [31], [112], [113],
biologically active materials [113], [114], as
well as in electrophotography [115] and reprog-
raphy [116]. Although these applications are
largely in the research stage, the unique all-
silicon backbone provides an alternative to car-
bon and siloxane polymers.
3. Poly(Carborane – Siloxanes),
Polycarbosilanes and Polysilazanes
Silicon forms bonds with itself to give polysi-
lanes, but it can also bond to other elements.
Here several of these families are considered.
(For polysiloxanes, see → Silicones).
Poly(Carborane – Siloxanes). The poly(carbo-
rane – siloxanes) have a linear structure (1) in
which R1 and R2 can be alkyl, fluoroalkyl, or
aryl groups. The main chain contains carborane
polyhedra. Although the polyhedral carborane
C2 B10 H10 is most commonly employed in the
preparation of poly(carborane –siloxanes), the
closo-carboranes C2 B5 H7 and C2 B10 H12 have
also been used as precursors [117], [118]. Major
interest in this class of polymers results from (1)
the need for enhanced flame resistance and (2)
their high thermal and oxidative stability.
Production. The hydrolysis – condensation
of carborane – siloxane monomers containing
terminal Si−Cl bonds is one synthetic route to
the poly(carborane – siloxanes) [119]. Cohydro-
lysis – condensation leads to other compositions
of 1 [119]:

Another route involves reacting a dihalosilane
with a carborane-containing dihydroxydisilane
[120]. Synthetic routes to the required
monomers have been reported [121], [122].
Properties and Uses. For polymers of series
1, where R = R = CH3 and n = 1 – 5, the glass
transition temperature (T g ) decreases with in-
creasing siloxane content from −42 to −88 ◦ C
as n increases from 1 to 5 [123]. Replacing R2
by phenyl groups gives a completely amorphous
polymer, but T g increases to −12 ◦ C. When R2
is a mixture of methyl and phenyl groups, T g
is lower. Trifluoropropyl-modified polymers are
amorphous and exhibit T g values that decrease
with increasing n.
The thermal stability of poly(carborane –
siloxanes) is excellent. In an inert atmosphere,
rapid weight loss occurs only upon heating
above 400 ◦ C [123]. Phenyl-substituted poly-
mers are more stable than their methyl analogues
[124]. The amorphous polymers are elastomers
and are formulated with fillers and vulcanized
by using standard silicon technology. The high-
temperature capabilities have resulted in the ap-
plication of poly(carborane – siloxanes) as sta-
tionary phases in gas chromatography and the
fabrication of these polymers into O-rings, gas-
kets, and wire coatings for operating tempera-
tures >300 ◦ C [117], [120].
Polycarbosilanes have generally been made
as precursors for silicon carbide. Recently,
polysilaethylene and related functionalized
products have been prepared [125–127].
Polysilaethylene has a low T g (about − 140 ◦ C)
and melts at about room temperature (ca. 25 ◦ C).
Ease of substitution of the chloro derivative by
alkoxide, amine, and other functional groups
Inorganic Polymers 5
may lead to a number of interesting products
with derivative chemistry like that of polyphos-
phazenes. Fluorinated polysilaethylenes have
also been made [128], [129]. It is suggested that
these materials might exhibit piezo and pyro-
electric properties similar to those of poly(vinyl
fluoride).
Polysilazenes and Poly(N-methylsilazenes).
Polysilazenes. Most research focuses on the use
of polysilazanes as precursors to ceramics [130–
133]. These products have not proved to be very
useful since the ceramic yield is low (about 50 %
at 1000 ◦ C) and the free carbon content is high
(about 30 % at 1000 ◦ C).
They are mainly synthesized by aminolysis
and ammonolysis of dichlorosilanes [134],
[135], deamination or redistribution reactions of
aminosiliazanes [136], [137], and more recently
by the anionic and cationic ring-opening poly-
merization of cyclosilazanes. The latter products
are soluble in typical organic liquids. The Si–N
bond energy is greater than the Si–O bond en-
ergy. The temperature of initial weight loss for
the products formed from the ring-opening re-
action is about 50 ◦ C higher than that of polydi-
methylsiloxane. Future applications will focus
on areas similar to those of polysiloxanes them-
selves.
4. Polyphosphazenes
Polyphosphazenes are a well developed versatile
class of inorganic polymers with over 700 dif-
ferent polymers formed in the past three decades
[138–142].
Polyphosphazenes are polymers containing
alternating nitrogen and phosphorus atoms in
the backbone and alternating single and dou-
ble bonds [143]. Phosphorus is pentavalent and
has two pendant substituents (X and Y) which
can be alkyl, aryl, alkoxy, aryloxy, arylamino,
alkylamino, halogen, or pseudohalogen. Varia-
tion of X and Y leads to a wide variety of mate-
rials with highly diverse properties and applica-
tions [143–147].
6 Inorganic Polymers
Production. Polyphosphazenes can be pre-
pared in several ways [143–147]. Reaction
of ammonium chloride with phosphorus pen-
tachloride yields a mixture of cyclic and linear
oligomeric phosphazenes with a degree of poly-
merization (DP) of up to 20:
The ring opening polymerization of
hexachlorocyclotriphosphazene [940-71-6],
[16422-79-0] (3) leads to a rubbery material,
but when a carefully purified starting material
is polymerized under controlled conditions, a
degree of polymerization of up to 15 000 can
be obtained and the polymer can be purified
readily.
Poly(dichlorophosphazene) [26085-02-9]
degrades slowly in a moist environment. How-
ever, the chlorine atoms are readily (and com-
pletely) displaced by many different nucle-
ophiles. This reaction leads to a variety of highly
stable polymers [143], [146–150].
The condensation polymerization of appro-
priate N-silylphosphoramines has resulted in the
successful preparation of the first polyphospha-
zene to have only alkyl substituents on phos-
phorus [151]:
Recently, polyphosphazenes have been pre-
pared by room-temperature cationic (living)
polymerization of phosphoranimines [152–
154]. Certain initiators allow the formation of
multiple initiating sites and the formation of
star-branched polyphosphazenes with a narrow
polydispersity. Block structures with polymers
such as poly(ethylene oxide) have been made
that contain reactive amine end groups [155].
Copolymers with polyurethanes have also been
produced [156].
A number of polythionylphosphazenes have
been produced with a variety of aryloxy, alkoxy,
and amino appendages [157].
Polyphosphazenes have also been coupled
with organosilicon compounds [158].
Properties. The poly(organophosphazenes),
unlike their precursors the poly(dichlorophos-
phazenes), are resistant to hydrolysis and can be
vulcanized to form elastomers. Copolymers with
trifluoroethoxy and heptafluorobutoxy (1 : 1)
groups are elastomeric with T g = − 77 ◦ C [148],
whereas T g values for poly(arylaminophospha-
zenes) are much higher (see Table 1). Poly-
(aryloxyphosphazenes) tend to be intermediate
between these two classes.
Table 1. Glass transition temperatures (T g ) and crystalline melting
points (T m ) for some polyphosphazenes, - [P(Y)2 = N--]n
Substituent (Y) T g, ◦C T m, ◦C
Cl −63 30
OCH3 −76
OCH2 CF3 −66 242
OC6 H5 6 390
N(H)C6 H5 105
Polyphosphazenes exhibit some of the low-
est known rotational barriers for skeletal back-
bone bonds in polymers (as low as 420 J/mol of
repeating units), which is consistent with their
low T g values. Two first-order transitions are
usually found, with a temperature interval of ca.
150 – 200 ◦ C [143], [146], [147]. The nature of
the lower temperature transition resembles the
change into a mesomorphic state similar to that
observed in nematic liquid crystals. In practice,

polyphosphazenes are usually soft just above
this lower transition temperature, which allows
compression molding of films to be carried out.
Table 1 lists representative data.
The thermal stability of some polyphos-
phazenes is outstanding. Several classes ex-
hibit onset of weight loss above 300 ◦ C, as
shown by thermogravimetric analysis (TGA),
but these studies overestimate the stability.
Poly(diphenoxyphosphazene) [28212-48-8], for
example, undergoes rapid depolymerization
above 150 ◦ C. Fluoroalkoxyphosphazene vul-
canizates have excellent resistance to solvents
(e.g., lubricants and fuels) and display low
volume swells after long immersion in lu-
bricants [159]. Polyphosphazenes also display
good flame-retardant properties, with oxygen in-
dex values in the range 20 – 70, and they give off
only moderate amounts of smoke [160], [161].
Uses. The favorable properties of polyphos-
phazenes, as well as the high elastomer elon-
gations, which are greater than those of poly-
siloxanes from −60 to 200 ◦ C, have resulted
in the use of various polyphosphazenes as gas-
kets, specialty damping materials, and petro-
leum piping for arctic applications. They have
also been used as specialty membrane mate-
rials [150]. Polyphosphazene – salt complexes,
which have the highest ionic conductivities at
ambient temperature of all polymer – salt elec-
trolytes, are reported to be semiconducting ma-
terials [150]. Other uses include catalytic tem-
plates [150], biomedical materials [160], and
their most prominent use as specialty elastomers
in O-rings, gaskets, and fuel hoses [143], [161–
165].
The first commercial polyphosphazene elas-
tomers were synthesized and subsequently de-
veloped by the Firestone Tire and Rubber Co.
[166] and represent the first new class of semi-
inorganic elastomers to be developed commer-
cially since silicone rubber. Ethyl Corporation
holds the rights to many commercial uses and
produces both commercial polymers and fin-
ished products from this interesting class of ma-
terials [167].
Thin films of amino-substituted polythionyl
phosphazenes exhibit an unusual permeabil-
ity to molecular oxygen [168]. When a phos-
phorescent dye that is quenched by oxygen is
dispersed in the polymer film, the product can
Inorganic Polymers 7
be used as a visually interpretable oxygen sen-
sor. One potential use is to follow differences in
pressure across a surface area such as a wing of
a plane.
5. Poly(Boron Nitride)
Boron nitride [10043-11-5] occurs in three crys-
talline forms – α, β, and γ (→ Boron Carbide,
Boron Nitride, and Metal Borides, Chap. 2.).
The α-form is a layered, hexagonal polymer
much like graphite. A crystalline, tetrahedral
network polymer similar to diamond also exists
[169–172]. Although boron nitride is isoelec-
tronic with the diamond and graphite forms of
carbon, it exhibits different properties. For ex-
ample, layered boron nitride is white and an elec-
trical insulator. Thus its electrons appear to be
highly localized in contrast to those of graphite.
Like graphite, layered boron nitride is soft and
easily machined. Articles are fabricated by sin-
tering or machining. These articles can be used
in air up to 800 ◦ C or in an inert atmosphere up
to 1600 ◦ C. Boron nitride has a Young’s modu-
lus : density ratio of 20 (versus 4 for iron) and
a tensile strength : density ratio of 8 (versus 1.7
for glass and 0.3 for iron). The tetrahedral form
of boron nitride is almost as hard as diamond
and is therefore employed as a substitute for di-
amond in both cutting and grinding and in jew-
elry. Industrial applications include its use as an
abrasive, insulator, and refractory material.
The hexagonal modification of boron nitride
has been prepared from combinations of cheap
boron and nitrogen compounds, such as B(OH)3 ,
(NH2 )2 CO, and N2 , or KBH4 and NH4 Cl [162–
167, 169–175]. More recently, chemical vapor
deposition (CVD) methods have used mixtures
of BCl3 and NH3 , BF3 and NH3 , or B2 H6 and
NH3 to produce hexagonal boron nitride [175,
178–180].
Polymeric precursors to boron nitride are
beginning to attract attention. The reaction
of difunctionalized borazines with bis(trime-
thylsilyl)-amines gives soluble oils that form
films which can be pyrolyzed at 1200 ◦ C to
yield a boron carbide – boron nitride ceramic,
but removal of all of the carbon has not yet
been achieved [173]. The use of trifunctional bo-
razines (shown below) leads to hexagonal boron
8 Inorganic Polymers
nitride in good yields, with good crystallinity via
intermediates such as 5:
The formation of boron nitride nanocompos-
ites has been reviewed [176].
6. Poly(Sulfur Nitride)
The topic of poly(sulfur nitride) has been re-
viewed [177].
Poly(sulfur nitride) [56422-03-8] was first
prepared in 1910 by Burt, who passed vapors
of cyclic tetrasulfur tetranitride [28950-34-7],
S4 N4 , over heated silver gauze or silica wool
[181]. However, it was not studied extensively
until the 1970s, when purer products were ob-
tained [182]. Poly(sulfur nitride) films have a
lustrous golden color and are very striking in
appearance. Polymeric sulfur nitride is a crys-
talline, fibrous material that is malleable under
mild pressure. Electron micrographs reveal that
the crystals are composed of fiber layers which
are stacked parallel to one another along the
crystal axis; X-ray studies show that the poly-
mer consists of almost planar chains of alter-
nating sulfur and nitrogen atoms. All the S−N
bond lengths are similar and correspond to a sul-
fur – nitrogen bond order intermediate between
a single and a double bond. The polymer decom-
poses on heating at ca. 140 ◦ C (see below).
The polymer exhibits highly anisotropic elec-
trical conductivity. Room-temperature conduc-
tivities as high as 3700 Ω−1 cm−1 along the
crystal axis have been measured [184], [185].
Conductivity increases markedly upon cooling
to 4.2 K, and the polymer becomes supercon-
ducting at 0.26 K [186]. Poly(sulfur nitride) was
the first example of a polymeric superconduc-
tor. The conductivity increases on doping with
bromine, and (SNBr0.4 )n at room temperature
has a conductivity of 3.8×104 Ω−1 cm−1 paral-
lel to the fibers, but only 8 Ω−1 cm−1 perpendic-
ular to the fiber axis. This strong directionality is
related to efficient conductivity along the molec-
ular chain axis, whereas a lower degree of order
in chain packing along the perpendicular axis
results in lower conductivity.
The large-scale production of poly(sulfur ni-
tride) does not appear to have attracted much
attention because of its rather low thermal sta-
bility and because S4 N4 is explosive.
Further, polymerization is slow (6 – 8 weeks).
Efforts have aimed at developing alternate syn-
thetic routes.
Poly(sulfur nitride) has been prepared from
solution by addition of trimethylsilyl azide to
(NSCl)3 and other similar routes [187–189]. In
these processes, a powdered product is obtained
rather than the larger crystalline grains. Micro-
crystalline layers of poly(sulfur nitride) have
been prepared on platinum electrodes by elec-
trolysis of [S5 N5 ]+ [190], [191]. Additional syn-
thetic routes have also been found [177]. Even
so, the classical route appears to be the method
of choice for the synthesis of pure poly(sulfur
nitride).
The fine structure of poly(sulfur nitride) con-
tinues to be elucidated [177]. Currently, it is
believed the polymer chains are planar and lie
along the (crystallographic) b axis that develops
along the a axis of S2 N2 . There are two chains
in each unit cell with two SN units per chain.
Disordered chains are present between the “or-
dered” chains. Secondary interaction between
the chains is significant.
Polymer crystals are stable in air and water
for several days. Stress and strain properties are
like those of an anisotropic metal. The Young’s
modulus parallel to the axis of the chain is on
the order of 1010 N m−2 .

Poly(sulfur nitride) becomes superconduc-
tive at 0.26 K [192]. This property has been
extensively studied [177]. The onset of super-
conductivity T c increases with increasing pres-
sure up to about 9000 bar, whereupon it de-
creases, but at higher pressures T c again in-
creases to about 3 K. It is possible that at
higher pressures small amounts are converted
to a –SNNS– polymer [193]. Poly(sulfur ni-
tride) exhibits the Meissner – Ochsenfeld ef-
fect (→ Superconductors, Chap. 2.2.) below
about 20 K. Below 4.2 K negative transverse
magnetoresistance occurs in the presence of
weak magnetic fields (< 3 T), though a normal
positive magnetoresistance is found above 77 K
or in stronger fields [177].
Thin films have been produced with proper-
ties that at times are similar to and under certain
conditions appears to differ from those of the
crystals themselves. The reason(s) for these dif-
ferences are not currently fully understood. The
ability to make large crystals has been reported
[194], [195].
Recent interest has focused on electrical ap-
plications because of its unusual magnetic and
electrical properties. Suggested uses include
light-emitting diodes, solar cells, and transis-
tors [196], [197]. A poly(sulfur nitride) – Ga –
As solar cell with an efficiency of 6.2 % has
been described [198]. Blue light-emitting diodes
have been made on ZnS with either gold or
poly(sulfur nitride) barrier electrodes [199]. The
polymer gives barriers about 0.75 eV higher with
a hundredfold increase in quantum efficiency
in comparison to the gold-based electrode. A
power output of 440 W h/kg at 200 mA was
found for a lithium cell with a poly(sulfur ni-
tride) electrode [192]. Patents for the use of
poly(sulfur nitride) have been issued for a va-
riety of uses including the cathode in batter-
ies [200], [201], coatings on an image record-
ing sheet [202], etching mask [203–205], ex-
plosive initiator [206], electroconductive resins
[207], multilayered wiring patterns [208], sub-
limable polishing grains for semiconductor de-
vices [209], and in light-emitting diodes [210].
Extensive work has been done on poly(sulfur
nitride) electrodes [177]. A ruthenium(II)-
modified poly(sulfur nitride) electrode gener-
ated hydrogen continuously in aqueous sulfu-
ric acid solution at − 0.1 V under visible light
[211]. Molybdenum(VI) electrodes using film
Inorganic Polymers 9
and crystalline poly(sulfur nitride) can convert
ethyne to ethene at rate 106 times that achieved
by chemical systems [212]. Paste electrodes
of poly(sulfur nitride) exhibit metallic electro-
chemical behavior in nonaqueous liquids [213].
7. Polysulfur
Because of its removal from industrial waste to
meet environmental standards, sulfur has been
stockpiled and is available in large quantities
at a modest price. Although the stable form at
room temperature is cyclooctasulfur, S8 , lin-
ear polysulfur is formed on heating. Unfortu-
nately, the thermodynamically stable form of
sulfur is the S8 monomer and the polymer under-
goes depolymerization after some time. Various
methods have been used to retard depolymer-
ization to the S8 monomer. Removal of minute
amounts of the monomer through judicious ex-
traction with carbon disulfide retards depoly-
merization to some extent. Probably the most
effective method involves the addition of olefins,
such as limonene, cyclopentadiene, and myrcene
[214]. Some of these result in polysulfur that is
stable for more than five years. Such stabilized
polysulfur has been incorporated into concrete
and asphalt mixes to strengthen them. Concrete
blocks, posts, highway pavement, and parking
tire restrainers that contain stabilized polysulfur
are produced.
8. Organometallic Polymers
A variety of organometallic polymers have been
synthesized for a number of applications. Ex-
amples are shown in Figure 2. The metals may
be an integral part of the polymer chain (6) or
side chain (7); they can be connected through
carbon (8) or other atoms (9); they may be
bonded by coordination (9), covalent (6–8), or
π – metallocene-like bonding (10); or they can
be part of a linear (6, 8–10) or cross-linked (7)
product. A classical coordination polymer is de-
scribed by 11, and stacked metal-chain com-
plexes are depicted by 12, where the circles re-
present organic macrocycles such as phthalocya-
nine.
10 Inorganic Polymers
Figure 2. Some organometallic polymers
Polymeric metal phosphinates may have
single-, double-, or triple-bridged structures,
such as 13, 14, and 15, respectively. Such
structures are known for aluminum, beryllium,
cobalt, chromium, nickel, titanium, and tin
[229]. They form films, with thermal stability up
to 450 ◦ C claimed in some instances. The chro-
mium(III) polyphosphinates have been used as
thickening agents for silicone grease to improve
its high-pressure physical properties.
The metal moiety of an organometallic poly-
mer imparts particular properties. First, the
metal atom is typically a site for reaction with
Lewis bases and redox reactants. These sites
have been exploited for catalysis and may be re-
sponsible for internal rearrangement and degra-
dation. Second, the metal generally contributes
to polymer stiffness through added volume oc-
cupation and π-bond formation. Thus, flexi-
bility must be derived from the organic moi-
ety. Third, the metal-containing portion typi-
cally contributes to the dipolar, ionic charac-
ter of the products. The high chemical reactiv-
ity of the metal site results in polymers with
poor thermal stability. For example, products
derived from bis(cyclopentadienyl)titanium di-
chloride [1271-19-8] typically begin to undergo
small weight losses between 60 and 150 ◦ C due
to the evolution of cyclopentadiene and forma-
tion of a more thermally stable material. Because
of the tendency of the metal site to undergo reac-
tion, care is needed in selecting solvents so that
“permanent” bonding does not occur between
the metal-containing moiety and the solvent.
The variety of structures and properties of-
fered by organometallic polymers is enormous
and cannot be summarized in a representative
fashion in this short chapter. The reader is re-
ferred to reviews on organometallic polymers
[23–27], [215–228]. Important uses of organo-
metallic polymers are as follows:
1) Catalysis [226], [214], [231]
2) Biological uses, general [225], [232], [233–
238], [340, 376–378, 385]
3) Additives: coatings, textiles, plastics, and pa-
per [225], [230], [231], [239–242]
4) Electrical conductors, semiconductors, and
piezoelectric devices [243–246], [247–251]
5) Electrocatalysis, photoelectrocatalysis, pho-
tovoltaic cells, specialized electrodes, sensor
devices, electrochromism, electrolumines-
cence, and multicolor displays [225], [226],
[246], [252–258]
6) Analytical reagents [259–263]
7) Photoprotective agents [230], [231], [239–
242]

Figure 3. Schematic of the sol – gel process
8) Thickening agents [229]
9) Energy-transfer agents [265–267], [247],
[248]
10) Permanent coloring agents [230], [239–242]
11) Composites [268], [269]
12) Metal deposition [270–274]
13) Solar energy conversion [275], [276]
14) Ceramic precursors [277–282]
15) Superconductivity [283]
16) Functionalized electrodes (electrocatalysis,
photoelectrocatalysis, photovoltaic cells)
[284–287]
17) Ferromagnetic materials [288]
18) Photoconductive materials [262], [263],
[287]
19) Poly(vinyl alcohol) heat stabilizers [225]
20) Electrochemical sensors [222]
21) Nonlinear optics [222]
22) One-dimensional conductivity [222]
23) Nanostructures and nanotechnology [226].
Inorganic Polymers 11
9. Sol – Gel Inorganic Polymers
In the 1980s, several symposia were devoted to
sol – gel inorganic polymer preparation, indicat-
ing the explosive growth in this field [289–291].
This field has continued explosive growth un-
der additional “titles” such as organic/inorganic
composite materials. Reviews are given in [292–
294].
In sol – gel processes, ceramic polymer pre-
cursors are formed in solution at ambient tem-
perature; shaped by casting, film formation, or
fiber drawing; and then consolidated to furnish
dense glasses or polycrystalline ceramics [295].
The most common sol – gel processes employ
alkoxides of elements such as silicon, boron, ti-
tanium, and aluminum. In alcohol – water solu-
tion, the alkoxide groups are removed stepwise
by hydrolysis under acidic or basic catalysis and
replaced by hydroxyl groups, which then form
−M−O−M− linkages. Thus, branched poly-
meric chains grow and interconnect, as illus-
trated below for a silicate sol. Gelation even-
tually occurs as the growing polymers link to-
12 Inorganic Polymers

gether to form a network that spans the entire

solution volume. At this point (the gel point),
both the viscosity and the elastic modulus in-
crease rapidly.

The chemistry of the sol – gel process can

thus be tailored to design new polymeric glasses
or ceramics. For example, two or more metal
alkoxides can be mixed in varying ratios to
form mixed sol – gel polymers. Mixing Zr(OEt)4
with Si(OEt)4 could lead to a gel with structure
19. Similarly, a zirconia – alumina – silicate gel
The gel can be viewed as a viscoelastic ma- polymer of structure 20 could be obtained from
terial composed of interpenetrating liquid and Zr(OR)4 , Al(OR)3 , and Si(OR)4 .
solid phases. The solid network retards the es-
cape of the liquid and prevents structural col-
lapse. Gelation is advantageous in processing
because the gel freezes in shapes. Thus, the
sudden increase in viscosity locks in place the
shapes formed by casting, drawing of fibers, or
film formation. Some sols and gels may be ori-
ented or modified by drawing or shearing. The
gel can then be dried by evaporation to form a xe-
rogel or by supercritical fluid extraction to give
an aerogel. Consolidation to dense glasses or ce-
ramics is finally carried out by thermal treatment
and sintering. Figure 3 gives some generalized
sol –gel processes for ceramic or glass prepara-
tion.
Both aerogels and xerogels have high sur- The development of organically modified sil-
face area (> 500 m2 /g) and small pore diameter icates (ceramers) has resulted in a variety of new
(< 20 nm). They have been used as ultrafiltra- materials [296]. Thus, vinyl groups, epoxides,
tion media, antireflective coatings, and catalyst and methacrylate functions have been polymer-
supports. Final densification is accomplished by ized to give interpenetrating networks with a
viscous sintering in which the viscosity exceeds wide variety of structures and properties [297–
1012 Pa · s. 299]:
The rate of silicate sol and gel formation is
exceptionally sensitive to pH and to the water –
alcohol molar ratio in the reaction medium, as
is the solubility of the amorphous silica that is
formed. Silica networks are based on [SiO4 ]4−
tetrahedra modified by [O3 Si−O− M+ ] units.
Addition of B2 O3 , Al2 O3 , TiO2 , ZrO2 , and re-
lated network-forming units can result in the
formation of mixed glasses and ceramics. The
sol – gel approach uses the corresponding metal
alkoxides, as illustrated below, for the formation
of borosilicate glasses:

Applications of ceramers include adhesives
for glass surfaces [297], protective coatings
for medieval stained glass [300], and scratch-
resistant coatings for plastic eyeglass lenses.
Sol – gel preparations of tetraethoxysilane –
water – alcohol – HCl can be spun into fibers
once the appropriate viscosity has been reached.
These fibers are only slightly weaker than silica-
glass fibers; ZrO2 – SiO2 fibers have also been
made in this way. Sol – gel processing of Ti(i-
C3 H7 O)4 – H2 O – ethanol – HCl gives thin film
coatings of controlled thickness which exhibit
interference colors that vary with film thickness
and they are n-type semiconductors [301].
Hybrid materials have been made by incorpo-
rating end-capped poly(tetramethylene oxide)
(PTMO) blocks of molecular mass 650 – 2000
into tetraethoxysilane sol – gel glasses [302].
These materials exhibit high extensibility.
They contain interdispersed organic polymer
and inorganic polymer regions. Although the
PTMO blocks are well dispersed, some degree
of local phase separation occurs. Titanium iso-
propoxide has been used to incorporate TiO2
into these block systems [302].
Thus, although sol – gel chemistry has been
extensively investigated since the early 1960s
[304], [305], only now is it rapidly becoming
important in material synthesis.
Uses. Sol – gels offer properties between
those of organic polymers and inorganic materi-
als. They are particularly useful in areas where
they can replace materials to give better prop-
erties, better processability, and/or lower cost.
Thin-film applications are promising, and the
production of numerous nonlinear optical ma-
terials has been reported [292]. Thin-film pro-
tective coatings are being considered in a num-
ber of areas. Hot-melt adhesives for glass con-
tainers have been developed [293]. Hard contact
Inorganic Polymers 13
lens materials have also been developed [293].
Another area of practical application is the re-
inforcement of plastics and elastomers [293].
Recent production of sol – gels with nanoscale
pores allows their potential use as catalysts,
porous supports, and as selective absorbents
[293].
10. Inorganic Fibers and Whiskers
Inorganic fibers have been important materials
for over 50 years. Asbestos, glass, and carbon
fibers are particularly well known. Many of the
materials described herein are discussed in more
detail under → Fibers, 5. Synthetic Inorganic.
Many inorganic fibers exhibit very high ten-
sile strength and thermal stability (> 1000 ◦ C)
but suffer from high density and, high cost. This
chapter discusses the preparation and properties
of some inorganic fibers which are of interest
in the production of high-performance materi-
als, particularly advanced resin-matrix, ceramic-
matrix, and metallic-matrix composites. The
patent literature on inorganic fibers and com-
posites is extensive and growing rapidly [302].
Asbestos, once widely used as insulation,
now has more limited applicability because of
its health hazard. Small fibers lodge in the lungs
and can cause lung cancer and other lung dam-
age. Asbestos-containing shingles, paper, weld-
ing rods, floor tiles, and brake linings are, how-
ever, still manufactured. These products avoid
fibers of length 5 – 20 µm, which are the most
dangerous. Glass fibers are used widely as in-
sulation and in fiber-reinforced plastics. Car-
bon fibers are of increasing importance in high-
performance composites. Several metallocene-
type condensation polymers (e.g., 21) exhibit
fiber formation directly upon polymerization or
upon application of stress [306].
Inorganic fibers composed of alumina, sil-
ica, titania, zirconia, aluminum silicate, boron,
boron carbide, boron nitride, silicon boride, sili-
con carbide, silicon nitride and others are known.
14 Inorganic Polymers
Polymeric alumina, titania, and zirconia fibers
are produced industrially [307–310]. Zirconia
textiles are produced by Union Carbide under
the tradename Zircar.
Alumina fibers with a high Al2 O3 content
have high temperature stabilities and high mod-
uli. They have been made by dispersing alu-
minum salts on rayon fiber, heating to burn
off the organic material, and sintering. Alterna-
tively, slurries of hydrated alumina can be ex-
truded, dried, and heated to give polycrystalline
filaments [311], [312]. Fibers may also be ex-
truded and drawn from a viscous organic solu-
tion, followed by heating to high temperature
[313]. Polyaluminoxanes are spun into fibers
and sintered to give Al2 O3 fibers, or mixed with
silicon compounds, spun, and calcined at 1000 –
1200 ◦ C to give aluminosilicate fibers.
High-purity silicon dioxide fibers can be used
up to 1100 ◦ C and are of interest for advanced
composites [314]. They are made by melting a
glass composed of 75 % SiO2 and 25 % Na2 O,
and then spinning and stretching the fiber simul-
taneously at 1100 ◦ C. The fibers are then acid-
leached to remove Na2 O and dried above 300 ◦ C
[315]. In another process, silicon alkoxide poly-
mers are spun by a sol – gel process to make
high-purity SiO2 fibers. The fibers are heated
to 1000 ◦ C to give a 99.999 % pure quartz fiber
which is stable at 1000 ◦ C [316].
A variety of metal oxide fibers are now
available. Du Pont has developed an alumina –
zirconia fiber with a tensile strength > 2100 MPa
and a modulus of 380 GPa for use in high-
performance composites [317]. Even after expo-
sure to 1400 ◦ C for 100 h, its tensile strength was
1400 MPa and its modulus remained unchanged.
Stabilized zirconia and alumina – silica – boria
fibers have also been produced [315]. The
patent literature reports fibers of such materi-
als as alumina – chromia, thoria (thorium ox-
ide), titania, titania – silica, thoria – silicate, zir-
conia – alumina, zirconium silicate, and zirco-
nia – yttria, as well as whiskers of these and
many other metal oxides [318].
Boron – tungsten fibers have densities of
2.4 – 2.6 g/cm3 and elastic moduli as high
as 390 GPa. However, they lose both tensile
strength and modulus at high temperature. When
coated with silicon carbide and boron carbide,
they retain about half their initial strength in
air up to 800 – 1000 ◦ C. Boron fibers are much
less dense and are, therefore, useful in the con-
struction of lightweight composites [318] found
in tennis racket frames and aircraft. However,
their cost is high because of production com-
plexities. Processes involve chemical vapor de-
position of boron onto carbon fiber and other
cores, followed by a variety of treatments [328].
Boron carbide fibers are made from mixtures of
methane and boron trichloride by using CVD
to create a boron carbide layer on tungsten core
filaments [302]. Boron nitride fibers are used
with boron nitride matrices to give composites
that are employed as electrical insulators which
are also good thermal conductors. Boron nitride
fiber mats are useful as electrtic cell separators
in lithium sulfide batteries [302]. The fibers are
made from boric oxide fibers that are nitrided
by ammonia at ca. 800 ◦ C followed by stretch-
ing at 2000 ◦ C. They have a tensile strength of
2100 MPa and a modulus of 345 GPa.
Silicon carbide can withstand 1800 ◦ C un-
der oxidizing conditions. Two methods of pro-
duction are employed. In the first, CVD is
used to deposit silicon carbide on tungsten fil-
aments. In the second method, polysilanes are
melt-spun to give a precursor fiber, which on
heat treatment yields fibers 10 – 30 µm in di-
ameter (see Chap. 2), [109], [319], [320]. Sili-
con carbide fibers are more resistant to oxida-
tion at high temperature than carbon fibers and
have greater compressive strength and electri-
cal resistance. Silicon carbide – carbon fibers,
can be spun from mixtures of pitch and a
poly(dimethylsilane), followed by vacuum heat-
ing at 1400 ◦ C [321]. In a similar fashion, melt-
spinning of a polycarbosilazane followed by
high-temperature treatment gives silicon car-
bide – nitride fibers with a tensile strength of
1.2 MPa and a modulus of 187 MPa, which resist
oxidation up to 1200 ◦ C [332].
Silicon nitride fibers are used in advanced
composites for aircraft, radomes, and electrical
components. They are prepared either by reduc-
tion of silica in the presence of nitrogen in an
electric furnace [315] or by nitriding polycar-
bosilane precursor fibers with ammonia [323].
Sol – gel polytitanocarbosilanes have been
prepared and used to melt-spin a precursor fiber
which, on pyrolysis in nitrogen at 800 – 1500 ◦ C,
gives Si – C – Ti – O fibers. At 1300 ◦ C, a tensile
strength of 3.0 GPa and a modulus of 220 GPa
were achieved [324]. Polycarbosilanes can be

treated with titanium tetrabutoxide to give a
polymer from which fibers may be spun. Af-
ter heat treatment, Si – C – Ti – C fibers with a
Young’s modulus of 75 – 150 GPa and tensile
strength of 0.1 – 11 GPa were obtained [335].
More recently, a number of inorganic materi-
als have also been produced as whiskers (see also
→ Whiskers). In general, the tensile strengths
of whiskers are greater than those of fibers.
For instance the approximate tensile strength
of graphite in the bulk form is 1000 MPa, for
fibers it is 2800 MPa, and for whiskers it is
15 000 MPa. Carbon whiskers are produced by
depositing a layer of pyrocarbon from the vapor
phase on a catalytically grown carbon filament
[325].
Inorganic fibers and whiskers exhibit some
of the highest tensile strengths recorded. They
are useful when light weight and high strength
are needed. Uses include turbine blades, heat-
resistant re-entry vessels, golf club shafts, au-
tomotive radio aerials, fishing poles, dental fill-
ings, and parts of aircraft, including stealth air-
craft [326]. Many of these inorganic fibers and
whiskers are used as reinforcing agents in com-
posites [326].
11. Biomedical and Antimicrobial
Polymers
Silicones. Silicones (polysiloxanes) are
polymers having a silicon – oxygen backbone.
Among the numerous materials having biomed-
ical applications, silicone polymers play an
important and versatile role. Medical devices
made from polysiloxanes include artificial wrist,
toe, and finger joints; hip implants; oviductal
plugs [350], [351]; brain membranes; mam-
mary implants; tracheotomy vents; and artificial
hearts. For a detailed description of polysilox-
anes, see → Silicones [328–331]. This chapter
deals with their biomedical applications [332].
Polysiloxanes are widely applied in medicine
for numerous reasons including their generally
extremely low toxicity, high gas permeabil-
ity, good mechanical properties, and the wide
range of fabrication processes that can be used
(e.g., coating, encapsulation, casting, molding,
sealing, and extrusion). Furthermore, they are
generally inert to the body and to the immune
Inorganic Polymers 15
system. The biomedical applications of poly-
siloxanes are reviewed in [232]. Monomeric
and oligomeric siloxanes are generally not sold
as medical-grade materials and may contain
toxic trace contaminants that are not important
in industrial applications but may be critical
in biomedical applications. Proper procedures
must be adopted to ensure quality control and
absence of contaminants that may be toxic and
may adversely affect long-term, repeated use of
the end product.
Polysiloxanes are extensively employed in
biomedical devices as specialty surface treat-
ments [332–334]. For example, octadecylpoly-
siloxane ( Siliclad, Glassclad 18) is applied to
hydrophilic surfaces to reduce blood – surface
interactions and protein adsorption [334], [335].
Other polysilane-derived surface treatments
are sold under tradenames such as Glassclad
HP, D.C. 5700, Glassclad 6C, Dri-Film, and
Glassclad IM [232, 340]. Silicone membranes
play critical roles in oxygenation and dialysis.
Polysiloxanes are about ten times more perme-
able to oxygen than low-density polyethylene
and a hundred times more permeable than ny-
lon or butyl rubber [336]. Therefore, poly(di-
methylsiloxane) [1646-73-7] soft lenses and
poly(dimethylsiloxane-co-methacrylate) hard
lenses constitute a major portion of the con-
tact lens market. Silicone-based interpenetrating
polymer networks give clean, resilient gels with
cohesive properties, which are used in breast
implants [337–339, 232]. Bilayer membranes
employing a top layer of cured silicone polymer
are used as artificial skin [348], [349].
Low-consistency, room-temperature vulcan-
izing silicones cure to give low modulus
mechanical plugs when injected into the fallop-
ian tubes [350], [351]. This blocks the ovum
from reaching the uterus and serves to prevent
conception. A retrieval ring at the uteral end
makes the procedure reversible. Silicone poly-
mers are used to make dental impression ma-
terials [342–345]. Vinyl addition cured systems
are used in biomedical encapsulation and mold
making [346] and in maxillofacial prostheses.
High-consistency elastomers are used in com-
pression molding of shunts, flexing joints, and
catheter tubing [347]. Silicone – urethanes (e.g.,
cardiothane 51) are used in blood pumps, in-
traaortic balloons, and artificial hearts because
of their good fatigue strength, flexibility, tough-
16 Inorganic Polymers
ness, and retarded interaction with plasma pro-
teins [348].
Silicone – polycarbonate block copolymers
can be injection-molded, extruded, and cast into
films. Their synthesis is described in [349–351].
They have high tear strength as well as high
oxygen and water permeability. Applications in-
clude blood oxygenation, dialysis, and micro-
electrode materials [352–354].
Metal-Containing Polymers. A few
metalcontaining polymers are used as thera-
peutic agents in the treatment of cancer [355–
359], juvenile diabetes [360–362], Cooley’s
anemia [363], [364], and in the control of bac-
teria [359], [365–367] and viruses [359–362].
The polymer derived from the reaction of tetra-
chloroplatinate and methotrexate can inhibit the
virus that causes juvenile diabetes-like symp-
toms [360], [361].
Tin-containing compounds are inhibitory to
certain bacteria. A variety of methyl-, ethyl-, and
propyl-substituted organotin polymers effec-
tively inhibit Pseudomonas aeruginosa strains
which are responsible for about one-half of the
deaths of burn patients. Some suggested uses
are as additives to paints, insulation, bandages,
and clothing [369]. Degradation resistance has
been obtained by in situ polymerization of tin-
containing esters in wood [366], [367]. Impreg-
nated poles and planks show total stability when
exposed to seaside environment or seawater for
more than five years [374], [375]; untreated
wood is destroyed within a year. Similar results
have been obtained for tin-containing cross-
linked coatings in the control of ship hull foul-
ing. Nevertheless, plants and sea creatures sur-
vive within centimeters of test objects, suggest-
ing that the use of tin-containing polymers might
be environmentally acceptable. Tin-containing
polymers have been examined in the United
States and Japan as antifouling coatings [370–
375] and mildew-resistant paints [376]. Com-
mercial products are available; however, envi-
ronmental concerns in harbor areas remain.
12. Boron Polymers
The topic of boron polymers has been reviewed
[377].
A variety of boron-containing polymers have
been made by heating trialkyl phosphates with
boron trichloride at 300 ◦ C [378], [379]. At
900 ◦ C this reaction produces boron phosphate
glass. The boron polymers form networks prior
to glass formation.
Reaction of benzene boronic acid with
diphenylsilane gives transparent, film-forming,
polymeric borosiloxanes which can be drawn
into fibers [380]. Similar materials resulting
from boronic esters and chlorosilanes form elec-
trically insulating films when baked on copper
plates [381]. A variety of polymers containing
Si−O−B units have been reported [382]:
Boric acid forms condensation polymers with
polyhydroxy compounds. With ethylene glycol
and phenol, a tacky polymer is formed that is
hydrolyzed readily by water [383]. Many such
compositions have been studied as adhesives,
binders, coatings, and resins.
Polymers with all-boron backbones [384]
or alternating boron – phosphorous backbones
[385] have been reported. Pyrolysis of di-
methylphosphineborane in the presence of a
base gives linear high molecular mass polymers
[386]. In many cases, ring systems form in-
stead of linear polymers, with cyclic trimers and
tetramers predominating. Use of triethylamine
as a chain end blocking group favors the forma-
tion of linear polymers.
The action of water on diboron tetrachloride

gives subboric acid, B2 (OH)4 , which when
heated at 220 ◦ C loses water to yield a white
boron monoxide polymer [387]. This remains
unchanged up to 500 ◦ C, but at 650 ◦ C is trans-
formed into brown boron monoxide which has
lower solubility in water.
Low molecular mass polymers result from the
reaction of 1,1 -dilithioferrocene with phenyl-
boron dichloride, followed by treatment with
water [388]:
A large number of polyborazines have been
prepared, and many are reviewed in [389]. The
condensation of bis(alkylthio)borazines with
diphenylsilicondiol gives borazine – siloxane
polymers [390]. Cross-linked materials result
from the corresponding reactions with tri-
alkylthioborazines:
Reaction of β-chloroborazines with sodium sul-
fide, sodium disulfide (Na2 S2 ), or sodium tetra-
sulfide (Na2 S4 ) leads to polyborazines with sul-
fur linkages between the rings and molecular
masses up to 53 000 [391]. Other polyborazines
with oxygen [392], nitrogen [393], and phos-
Inorganic Polymers 17
phite bridges [394] between rings have also been
prepared.
Carborane polymers with a variety of struc-
tures have been made from the C2 B10 H10 icosa-
hedral cage systems, in which the carbon atoms
are located ortho, meta, or para to one another.
The best known examples are the Dexsil poly-
mers of the Olin Corporation [395], [396]. This
series has siloxane groups attached to the carbon
atoms of ortho-carborane (1,2-C2 B10 H12 ) and
varying numbers of dimethylsiloxane groups
between adjacent cages:
Many other groups have been used to connect
carborane cages. Treatment of dilithiocarborane
with phosphorus trichloride and sodium azide
leads to the very unusual structure [397]:
Dilithiocarboranes are reacted with dialkyl
metal dihalides to introduce metal functions
between the cages. For example, reaction
of ortho- or para-dilithiocarborane with di-
methyldichlorostannane introduces dimethyltin
bridges between the cages. Dimethylgermanium
bridges are introduced in a similar manner [398],
[399].
18 Inorganic Polymers
The dimethyltin-bridged para-carborane poly-
mer softens only above 420 ◦ C and decomposes
at 425 ◦ C, whereas its ortho analogue melts at
250 – 255 ◦ C. Random mixtures of meta-carbo-
rane (1,7-C2 B10 H12 ) and para-carborane (1,12-
C2 B10 H12 ) have been used to prepare dialkyltin-
bridged polymers.
A polymer bridged by −Ge(CH3 )2 NH−
units was prepared from dilithiocarborane by us-
ing excess dimethyldichlorogermane and then
ammonia:
Borazine bridges have also been introduced
[400]:
Refluxing the bis(sulfinylchloride) of meta-
carborane in water gives a poly(carbo-
ranethiosulfinate), whereas treatment with
dilithiocarborane gives the corresponding
poly(carboranesulfide) (mp 219 – 222 ◦ C) [401]:
13. Aluminum Polymers
Polymers containing only aluminum atoms in
the backbone are unknown. Aluminum – oxygen
and aluminum – nitrogen backbones are known,
but they are hydrolyzed readily by aqueous
acid and alkali. Polymers with an aluminum –
oxygen backbone are usually obtained by re-
action of aluminum alkoxides with water, di-
ols, organic acids, or amides. They occur as
gums to brittle solids; applications include use
as cross-linking agents, drying agents, gelling
agents, fuel additives, water repellents, paints,
varnishes, linoleum, and inks [402–406]. An ide-
alized linear structure is
where X can be RO− [407], RCOO− [408],
R3 SiO− [407], [409], or R(RO)P(O)O− [410].
These polymers are highly cross-linked, and
the aluminum is actually four-coordinate; cyclic
structures often predominate.
To increase the hydrolytic stability of Al−O
bonds and permit the synthesis of linear poly-
mers, four-coordinate aluminum ethyl acetoac-
etate complexes have been made as chelate rings
along the chain [411]. The use of diols gives hy-
drolytically unstable polymers [412].
A variety of coordination polymers involving
aluminum chelates have been made, but these are
considered true organometallic polymers and
are not discussed here. Andrianov pioneered
the use of R3 SiO− blocking groups to prevent
formation of three-dimensional networks. Al-
though oligomers such as

have been prepared [409], useful higher poly-
mers remain elusive.
Poly(aluminosiloxanes) are polymers con-
taining an Si−O−Al−O backbone. A typical
example results from the reaction of the sodium
salt of a poly(dimethylsiloxane) with aluminum
chloride.
Polymers with silicon – aluminum ratios of 0.8 –
23 have been made [413], [414]. Polymers with
low silicon – aluminum ratios are brittle and in-
soluble, indicating a three-dimensional struc-
ture; those with silicon – aluminum ratios of 7 –
23 are soluble and stable in both polar and non-
polar solvents even at 150 ◦ C.
The poly(aluminophenylsiloxane) having a
silicon – aluminum ratio of 4 is completely in-
fusible but very soluble in organic solvents; it
can be plasticized. A ladder structure has been
assigned:
Regular poly(aluminosiloxane) structures are
prepared by polycondensation of dialkyldiace-
toxysilanes with aluminum alkoxides.
Inorganic Polymers 19
14. Polymers with −Si−O−M−O−
Backbones
The early driving force in preparing poly-
(metallosiloxanes) was to enhance thermal sta-
bility. The group 13 elements boron and alu-
minum have been incorporated into siloxane
chains (see Chaps. 12 and 13). The metal can
also be from group 14 (Ge, Sn, Pb), 15 (P, As,
Sb), or 16 (S), or a transition metal (Ti, V, Cr,
Fe, Ni, Zr, Hf). The thermal and hydrolytic sta-
bilities of these polymers are usually inferior
to those of the parent poly(organosiloxanes),
but a variety of property modifications have
been achieved. The hydrolytic stability (to HCl)
of poly(titanoxane –phenylsiloxane) was shown
to be greater than that of its aluminum ana-
logue poly(aluminoxane – phenylsiloxane), but
less than that of poly-(diphenylsiloxane). Thus,
the order of stability with respect to cleavage by
hydrochloric acid at 90 ◦ C is
− Si − O − Si − O − > − Si − O − Ti − O − >
− Si − O − Al − O −.
Random – Si – O – Ge – O – and – Si – O –
Sn – O – polymers can be made by the
cohydrolysis of dialkyldichlorosilanes with di-
alkyldihalogermanes or dialkyldihalostannanes
[415]. Low molecular mass components are
treated further with sulfuric acid to give rubber-
like materials.
The random incorporation of tin is achieved by
cohydrolysis of dihalostannanes with dihalosi-
lanes [416] or by reaction of dialkyltin oxides
with a silanediol [417].
The most stable polymers in this series have high
silicon: tin ratios. Regular alternating structures
20 Inorganic Polymers

are achieved by reaction of dialkyldiacetoxys-

tannanes with dialkyldialkoxysilanes [418]:

Poly(titanosiloxanes) can be prepared by

cohydrolysis of dichlorosilanes with titanium
alkoxides to give random incorporation of
−Si(R)2 O− and −Ti(R)2 O− units. Alterna-
tively, diphenylsilanediol can react with titanium
alkoxides [419]. The resulting polymers are hard
coatings, which are opaque to UV light and suit-
able for use as refrigeration enamels [419].
Partial hydrolysis of alkylalkoxysilanes with ti-
tanium alkoxides, acylates, or alkylamides gives
copolymers that can be used to help catalyze
the drying of printing inks [420], [421]. Similar
silicon – zirconium (−Si−O−Zr−O−) prepara-
tions have also been reported [422]. A dif-
ferent type of titanium unit is represented by
the titanocene fragment. Condensation of di-
methylsilanediol [423] or its disodium salt [424]
with titanocene dichloride results in the incorpo-
ration of titanocene units into the polymers.
Reactive silicones and titanium alkoxides give
polymers with −Si−O−Ti−O− backbones
when heated in petroleum solvents. These are
said to be suitable for making waterproof leather
and textiles [426], [427].
Polymers with titanium – oxygen backbones,
related to siloxanes, are prepared by heating tri-
alkoxytitanium acylates and dialkoxytitanium
diacylates at 75 ◦ C. These materials have low
molecular mass, and whether they are linear
polymers or cyclic oligomers is as yet unre-
solved. Titanium alkoxides polymerize on treat-
ment with acetic acid, but problems with cross-
linking and three-dimensional network forma-
tion complicate this method.
Incorporation of a group 15 element is illus-
trated by the preparation of −Si−O−As−O−
chain polymers. Condensation of dimethyldi-
chlorosilane with CH3 AsO(OH)2 gives a trans-
parent, rubbery polymer that is used as an im-
pregnation rubber additive [428]. The arsenic
content inhibits the growth of molds.

A related polymer containing −O−Si−S−

Ti−S−Si− units in the backbone is made
by polycondensation of dimercaptotitanocene
[425]:

15. Poly(Carbon Disulfide),
Poly(Carbon Diselenide), and
Polythiocyanogen
The delocalized chalcogen-based polymers are
of special interest due to the observation of su-
perconductivity in [SN]n [429] and [Se]n [430].
Bridgman first showed that carbon disulfide can
be polymerized under high pressure (> 4.5 GPa)
at 175 ◦ C to give a black semiconducting solid
[431]. The (CS2 )n prepared under high pressure
was later assigned a head-to-tail linear structure
[432]. The head-to-head structure is 16.3 kJ/mol
lower in energy and, by analogy to (CSe2 )n ,
might be the more likely structure [433].
Reaction products vary with experimental con-
ditions. Chan and Jonsher isolated a low-
conductivity material (< 10−13 S/cm) and a
semiconducting material (10−3 S/cm) which
consisted of free sulfur in a (CS2−x ) matrix
[434]. They also proposed a cross-linked struc-
ture.
Anionic polymerization of carbon disulfide,
initiated by sodium dispersed in dipolar apro-
tic solvents, results in polymers with less than
two sulfur atoms per carbon [433], [435].
Poly(carbon disulfide) has also been prepared by
photolysis [436], irradiation [437], and plasma
deposition [438] to give structures described as
- [C3 S2 -]n , - [CS(S)m -]n , or mixtures [439].
Unlike carbon disulfide, carbon diselenide
can be polymerized both at high pressure and un-
der ambient conditions. The products obtained
range from insulators to metals and supercon-
ductors. Carbon diselenide polymerizes at a rate
of about 1 wt % per month at ambient tempera-
ture and pressure [440], [441]. When dissolved
Inorganic Polymers 21
in solvents such as methylene chloride and di-
oxane, it polymerizes readily under high pres-
sure (ca. 0.5 GPa) at 100 ◦ C [442]. This polymer
is an amorphous insulator. When the polymer is
heated to 130 – 160 ◦ C, a black semiconducting
(10−3 S/cm) material results. After high-pres-
sure polymerization and annealing at 130 ◦ C, a
crystalline product is obtained with a room-tem-
perature conductivity of 50 S/cm; it has been as-
signed a two-dimensional sheet structure [443–
445]. The polymer is thought to have a head-
to-head structure [440], [441], which is signifi-
cantly more stable than the head-to-tail structure
[433]:
It is typically contaminated with trigonal se-
lenium domains, depending on annealing and
polymerization conditions. The observed crys-
tallinity is due to both small and large trig-
onal selenium crystallites within the polymer
matrix, and this selenium is the reason for the
high conductivity [433]. As selenium crystallites
form, the polymer becomes highly disordered,
with the formation of −C−C−, −C=Se−, and
−C−Se−Se−C− units.
Carbon selenosulfide (CSeS) does not poly-
merize in solution under conditions similar to
those used to polymerize carbon diselenide
[433]. Polymerization of carbon diselenide in
the presence of excess carbon disulfide gives
only the selenium-containing product. The slow
polymerization of carbon diselenide at atmo-
spheric pressure and 25 ◦ C gives a black prod-
uct with an electrical conductivity < 10−6 S/cm,
consistent with a band gap of ca. 2 eV [433].
Polythiocyanogen. Polythiocyanogen or
parathiocyanogen is formed by the spontaneous
polymerization of thiocyanogen (SCN)2 [446].
While a number of structures are possible, a
linear structure with the SCN atoms connected
with a bond order of 1.5 is currently preferred
[447]. The polymer is a semiconductor that can
be doped with halogens [447].
22 Inorganic Polymers

16. Phosphorus-Containing
Polymers
Phosphorus-containing polymers are among the
three most important basic blocks of life. They
serve as the building blocks of the critical nucleic
acids, DNA, RNA, phosphosaccharides, phos-
phoproteins, and related biomaterials. Here, we
will focus only on synthetic phosphorus-con-
taining polymers. Reviews are given in [450–
455]. The reviews cover the synthesis and uses
of organic and inorganic phosphorus polymers
having a wide variety of compositions and struc-
tures. Polymer chains containing phosphorus
can be made by a variety of approaches, includ-
ing those summarized below:

Such reactions can be used to prepare artificial

glass, paint and lubricant additives, plasticizers,
and adhesives.
An interesting class of linear phosphorous –
boron backbone polymers has been reported
[449]. When the dimethylphosphine – borane
complex is pyrolyzed, a linear polymer with
DP = 80 is obtained which is soluble in hot ben-
zene:

A similar process yields borophane polymers

from tris(trialkylsilyl)phosphines [449]:

An interesting application of electrophilic

Polymers with a glasslike structure are
formed by cocondensation of metal alkoxides
or metal acetates with phosphorus oxychloride
or phosphoric acid:
substitution involves the incorporation of phos-
phorus bridges between ferrocene units [450],
[456]. However, this route only gives low molec-
ular mass materials.

Polyphosphazenes are discussed in Chap-
ter 4. However, a few carboranylphosphazene
polymers have been reported [457]. Ring-
opening polymerization of the carboranyl-
substituted phosphazene ring proceeds without
complications to yield a high molecular mass
polymer:
17. References
1. NRC: Polymer Science and Engineering,
National Academy Press, Washington, D.C.
1981.
2. F. S. Kipping, J. Chem. Soc. 125 (1924) 229.
3. R. West et al., J. Amer. Chem. Soc. 103 (1981)
7352.
4. J. Wesson, T. Williams, J. Polymer Sci., Polym.
Chem. Ed. 17 (1979) 2833.
5. R. Trujillo, J. Organometallic Chem. 198
(1980) C27.
6. S. Sawan, S. Ekhorutomwen, in J. Salamone
(ed.): Polymeric Materials Encyclopedia, vol.
9, CRC Press, Boca Raton, FL 1996, p. 6722.
7. S. Hayase, in J. Salamone (ed.): Polymeric
Materials Encyclopedia, vol. 9, CRC Press,
Boca Raton, FL 1996, p. 6734.
8. K. Matyjaszewski, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 9,
CRC Press, Boca Raton, FL 1996, p. 6741.
9. A. Soum et al., in J. Salamone (ed.): Polymeric
Materials Encyclopedia, vol. 9, CRC Press,
Boca Raton, FL 1996, p. 6747.
Inorganic Polymers 23
10. C. Tien, A. Savoca, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 9,
CRC Press, Boca Raton, FL 1996, p. 6758.
11. E. Hengge, J. Inorg. Organomet. Polym. 3
(1993) no. 4, 287.
12. S. Thames, K. Panjnani, J. Inorg.
Organometal. Polym. 6 (1996) no.2, 69.
13. A. Eckhardt, W. Schnabel, J. Inorg.
Organometal. Polym. 6 (1996) no.2, 95.
14. K. Matyjaszewski, J. Inorg. Organometal.
Polym. 1 (1991) no. 4, 463.
15. C. Carraher, J. Macromol. Sci. Chem. A17
(1982) no. 8, 1293.
16. F. G. A. Stone, W. A. G. Graham: Inorganic
Polymers, Academic Press, New York 1962.
17. K. A. Andrianov: Metalorganic Polymers,
Interscience, New York 1962.
18. S. N. Borisov, M. G. Voronkov, E. Lukevits:
Organosilicon Heteropolymers and
Heterocompounds, Plenum Press, New York
1970.
19. H. R. Allcock: Phosphorus-Nitrogen
Compounds, Academic Press, New York
1972.
20. H. R. Allcock: Heteroatom Ring Systems and
Polymers, Academic Press, New York 1967.
21. M. G. Voronkov, V. P. Mileshkevich, Yu. A.
Yuzhelevskii: The Siloxane Bond, Consultants
Bureau, New York 1978.
22. A. L. Rheingold: Homoatomic Rings, Chains
and Macromolecules of Main Group Elements,
Elsevier Scientific Publ. Co., New York 1977.
23. C. E. Carraher, J. E. Sheats, C. U. Pittman, Jr.,
(eds.): Organometallic Polymers, Academic
Press, New York 1978.
24. C. E. Carraher, J. E. Sheats, C. U. Pittman, Jr.,
(eds.): Advances in Organometallic and
Inorganic Polymers, Marcel Dekker, New
York 1982.
25. M. Zeldin, K. J. Wynne, H. R. Allcock (eds.):
“Inorganic and Organometallic Polymers,”
ACS Symp. Ser. 360 (1988).
26. C. U. Pittman, Jr., C. E. Carraher, Jr., J. R.
Reynolds in H. Mank, N. Bikales, C.
Overberger and G. Menges (eds.):
Encyclopedia of Polymer Science and
Engineering, “Organometallic Polymers,”
vol. 10, 2nd ed., J. Wiley, 1987, p. 541.
27. M. D. Rausch et al., “New Vinyl
Organometallic Monomers: Synthesis and
Polymerization Behavior” in B. M. Culbertson,
C. U. Pittmann, Jr. (eds.): New Monomers and
Polymers, Plenum Press, New York 1984; F. R.
Hartley: Supported Metal Complexes, Reidel
Publishing Co., Boston 1985; J. E. Sheats,
C. E. Carraher, Jr., C. U. Pittman, Jr.:
24 Inorganic Polymers
Metal-Containing Polymeric Systems, Plenum
Press, New York 1985; C. U. Pittman, Jr.,
“Vinyl Polymerization of Organic Monomers
Containing Transition Metals” in E. J. Becker,
M. Tsutsui (eds.): Organometallic Reactions,
vol. 6, Wiley-Interscience, New York 1977,
p. 1; C. U. Pittman, Jr. “Polymer Supported
Catalysts” in G. Wilkinson and F. G. A. Stone
(eds.): Comprehensive Organometallic
Chemistry, vol. 8, Chap. 55, Pergamon Press,
Oxford 1982.
28. J. P. Wesson, T. C. Williams, J. Polym. Sci.
Polym. Chem. Ed. 19 (1981) 65.
29. X. H. Zhang, R. West, J. Polym. Sci. Polym.
Chem. Ed. 22 (1984) 159, 225.
30. R. West, L. D. David, P. I. Djurovich, H. Yu, R.
Sinclair, Am. Ceram. Soc. Bull. 62 (1983)
no. 8, 899.
31. R. West, J. Organomet. Chem. 300 (1986)
327.
32. R. West et al., J. Am. Chem. Soc. 103 (1981)
7352.
33. L. A. Harrah, J. M. Zeigler, J. Polym. Sci.
Polym. Lett. Ed. 23 (1985) 209.
34. J. M. Zeigler, Polym. Prepr. Am. Chem. Soc.
Div. Polym. Chem. 28 (1987) no. 1, 424.
35. K. Matyjaszewski, Y. L. Chen, H. K. Kim:
“Inorganic and Organometallic Polymers,”
ACS Symp. Ser. 360 (1988).
36. R. West, S. Hayase, T. Iwahara, J. Inorg.
Organometal. Polym. 1 (1991) 545.
37. J. Wilderman, J. Herrema, G. Hadziioannou,
E. Schomaker, J. Inorg. Organometal. Polym.
1 (1991) 567.
38. C. Van Wakreem, T. Cleiji, J. Zwikker, L.
Jenneskens, Macromolecules 28 (1995) 8696.
39. J. Herrema et al., Macromolecules 28 (1995)
8102.
40. C. Aitken, J. Harrod, E. Samuel, Organomet.
Chem. 279 (1985) C11.
41. C. Aitken, J. Harrod, E. Samuel, J. Amer.
Chem. Soc. 108 (1986) 4059.
42. J. Harrod, ACS Symposium Ser. 360 (1988) 89.
43. H. Woo, T. Tilley, J. Amer. Chem. Soc. 111
(1989) 3757.
44. T. Tilley, Accounts Chemical Research 26
(1993) 22.
45. J. Banovetz, K. Stein, R. Waymouth,
Organometallics 10 (1991) 3430.
46. J. Corey, X. Zhu, T. Bedard, L. Lange,
Organometallics 10 (1991) 924.
47. H. Li, F. Gauvin, J. Harrod, Organometallics
12 (1993) 575.
48. C. Forsyth, S. Nolan, T. Marks,
Organometallics 10 (1991) 2543.
49. P. Bianconi, T. Weidman, J. Amer. Chem. Soc.
110 (1988) 2342.
50. W. Szymanski, G. Visscher, P. Bianconi,
Ultrasonics 28 (1990) 310.
51. W. Szymanski, G. Visscher, P. Bianconi,
Macromolecules 26 (1993) 869.
52. A. Sekignchi, M. Naugo, C. Kabuto, H.
Sakurai, J. Amer. Chem. Soc. 117 (1995) 4195.
53. H. Suzuki, Y. Kimata, S. Satoh, A. Kuriyama,
Chem. Lett. (1995) 293.
54. J. Lambert, J. Pfing, C. Stern, Angew. Chem.
Int. Ed. Engl. 34 (1995) 98.
55. T. Imori, T. Tilley, J. Chem. Soc. Commun.
(1993) 1607.
56. N. Devylder, M. Hill, K. Molloy, C. Price,
Chem. Commun. (1996) 711.
57. V. Lu, T. Tilley, Macromolecules 29 (1996)
5763.
58. J. Reichi, C. Popoff, L. Gallagher, E. Remsen,
D. Berry, J. Amer. Chem. Soc. 118 (1996)
9430.
59. M. Fujino, T. Hisaki, N. Matsumoto,
Macromolecules 28 (1995) 5017.
60. S. Adams, J. Drager, Angew. Chem. Int. Ed.
Engl. 26 (1987) 1255.
61. J. Mark, H. Allcock, R. West: Inorganic
Polymers, Prentice Hall, New York 1992.
62. H. Suzuki, H. Meyer, J. Simmerer, J. Yang, D.
Naarer, Adv. Mater. 5 (1993) 743.
63. R. West, L. David, P. Djurovick, H. Yu, Am.
Ceram. Soc. Bull. 62 (1983) 899.
64. M. Kumada, K. Tamao, Adv. Organomet.
Chem. 6 (1968) 19.
65. R. Barney, G. Chandra: Encyclopedia of
Polymer Science and Engineering, 2nd ed.,
vol. 13, John Wiley & Sons, New York 1988,
p. 312.
66. C. Schilling, Br. Polymer J. 18 (1986) 355.
67. J. Pilot et al., in A. Bassindale, P. Gaspar
(eds.): Frontiers in Organosilicon Chemistry,
Royal Society of Chemistry, Cambridge 1991.
68. W. Schmidt et al., Chem. Mater. 3 (1991) 257.
69. Y. Shieh, S. Sawan, Polymer Preprints 33
(1992) 1042.
70. D. Seyferth et al., Polymer Reprints 34 (1992)
no. 1, 223.
71. D. Seyferth, P. Czubarow, Chem. Mater. 6
(1994) no. 1, 10.
72. S. Yajima, Y. Hasegawa, L. Hayasi, M. Iimura,
J. Mater. Sci. 13 (1978) 2569.
73. R. West, A. Wolff, D. Peterson, J. Radiation
Curing 35 (1986) 40.
74. A. Wolf, R. West, Applied Organomet. Chem.
1 (1987) 7.

75. R. Miller, G. Wallraff, N. Clecak, R.
Sooriyakumaran, PMSE 60 (1989) 40.
76. A. Eckhardt, N. Yars, T. Wollny, W. Schnabel,
S. Nespurek, Ber. Bunsen-Ges. Phys. Chem.
98 (1994) 853.
77. G. VanderLaan et al., Macromolecules 27
(1994) no. 7, 853.
78. I. Kminek et al., Collect. Czech. Chem.
Commun. 58 (1993) no. 10, 2337.
79. F. Kajar, J. Messier, C. Rosilio, J. Appl. Phys.
60 (1986) 3040.
80. C. Callender, L. Robitaille, M. Lecleric, Opt.
Eng. (Bellingham, Wash.) 32 (1993) 2246.
81. S. Grigoras, T. Banton et al., Annu. Tech. Conf.
Soc. Plast. Eng. 50 (1992) no. 2, 2265.
82. S. Grigoras et al., Polymer Preprints 33
(1992) 655.
83. K. Matyjaszewski, J. Chrusciel, H. Kim, J.
Maxka, Polymer Preprints 34 (1993) 67.
84. R. Corriu, W. Douglas, Z. Yang, Y. Karakus,
G. Cross, D. Bloor, J. Organomet. Chem. 69
(1993) 455.
85. S. Grigoras et al., Synth. Met. 49 (1992) 293.
86. Y. Iwasa, T. Hasegawa, T. Koda, Y. Tokura, H.
Tachibana, Y. Kawabata, Synth. Met. 50
(1992) 415.
87. F. Kajzar, J. Messier, C. Rossilio, J. Appl.
Phys. 60 (1986) 3040.
88. D. Hofer, R. Miller, G. William, APIE
Advances in Resist Technology 16 (1984).
89. R. Miller et al., Polym. Eng. Sci. 29 (1989)
882.
90. T. Weidman, A. Joshi, Proc. SPIE-Int. Soc,
Opt. Eng.; Advances in Resist Technology and
Processing 1925 (1993) 145.
91. G. Wallraff et al., Polym. Mater. Sci. Eng. 66
(1992) 105.
92. D. Hofer, R. Miller, C. Wilson, A. Neurath,
SPIE-Advances in Resist Technology 469
(1984) 108.
93. C. Rosilio, A. Rosilio, B. Serre,
Microelectronic Eng. 6 (1987) 399.
94. P. Trefonas, R. West, R. Miller, D. Hofer, J.
Polym. Sci., Polymer Lett. Ed. 21 (1983) 823.
95. A. Gozdz, Polym. Adv. Technol. 5 (1994) 70.
96. A. Joslie, T. Weidman, A. Johnson, J. Miner,
D. Ibbotson, Proc. SPIE-Int. Soc. Opt. Eng.,
Adv. Resist Technology and Processing 1925
(1993) 709.
97. M. Sakata, T. Ito, Y. Yamashita, J.
Photopolym. Sci. Technol. 3 (1990).
98. R. Miller et al., “Polymer Microlithography,”
ACS Symp. Ser. (1989).
99. J. Su, T. Hsu, S. Sawan, J. Lavine, Polymer
Preprints 33 (1992) 1046.
Inorganic Polymers 25
100. C. G. Wilson, in L. Thompson, G. Wilson, M.
Bowden (eds.): Introduction to
Microlithography, American Chemical
Society, Washington DC 1983.
101. J. Zeigler, L. Harrah, A. Johnson,
SPIE-Advances in Resist Technology and
Processing 539 (1985) 166.
102. Chem. Abstr. 107 (1987) 218592u.
103. I.B.M. Corp., US 4 464 460, 1984 (H.
Hiraoka).
104. D. C. Hofer, R. D. Miller, C. G. Willson, J.
Soc. Photo Opt. Instrum. Eng. 469 (1984) 16.
105. D. C. Hofer, R. D. Miller, C. G. Willson, A. R.
Neureuther, J. Soc. Photo Opt. Instrum. Eng.
469 (1984) 108.
106. R. West, A. R. Wolff, D. J. Peterson, J. Radiat.
Curing 13 (1986) 35.
107. R. D. Miller et al.: “Inorganic and
Organometallic Polymers,” ACS Symp. Ser.
360 (1988).
108. J. Michl et al.: “Inorganic and Organometallic
Polymers,” ACS Symp. Ser. 360 (1988).
109. Y. Hasegawa, M. Iimura, S. Yajima, J. Mater.
Sci. 15 (1980) 1209.
110. R. G. Kepler, J. M. Zeigler, L. A. Harrah, S. R.
Kurtz, Phys. Rev. B 35 (1987) 2818.
111. F. Kazar, J. Messier, C. Rosilio, J. Appl. Phys.
60 (1986) 3040.
112. R. West “Organopolysilanes” in G. Wilkinson,
F. G. A. Stone, E. Abel (eds.): Comprehensive
Organometallic Chemistry, vol. 9, Pergamon
Press, Oxford, U.K., 1983.
113. R. Seymour, C. Carraher: Polymer Chemistry,
Marcel Dekker, New York 1987.
114. C. Gebelein, C. Carraher: Polymeric Materials
in Medication, Plenum Press, New York 1985.
115. R. West, J. Maxka, R. Sinclair, P. Cotts, Polym.
Prepr. Am. Chem. Soc. Div. Polym. Chem. 28
(1987) no. 1, 387.
116. J. F. Harrod, Polym. Prepr. Am. Chem. Soc.
Div. Polym. Chem. 28 (1987) no. 1, 403.
117. R. W. Williams, Pure Appl. Chem. 29 (1972)
569.
118. E. N. Peters, J. Macromol. Sci. Rev. Macromol.
Chem. C17 (1979) 173.
119. K. O. Knollmueller, R. N. Scott, H. Kawasnik,
J. R. Sieckhaus, J. Polym. Sci. Polym. Chem.
Ed. 9 (1971) 1071.
120. E. N. Peters et al., Rubber Chem. Technol. 48
(1979) 14.
121. S. Papetti, B. B. Schaeffer, A. P. Gray, T. L.
Heying, J. Polym. Sci. Polym. Chem. Ed. 4
(1966) 1623.
122. E. N. Peters, D. D. Stewart, J. Polym. Sci.
Polym. Lett. Ed. 17 (1979) 405.
26 Inorganic Polymers
123. M. B. Roller, J. K. Gillham, Polym. Eng. Sci. 8
(1974) 567.
124. E. N. Peters et al., J. Polym. Sci. Polym. Phys.
Ed. 15 (1977) 723.
125. L. Interrante et al., J. Amer. Chem. Soc. 116
(1994) 12025.
126. H. Wu. L. Interrante, Macromolecules 25
(1992) 1840.
127. I. Rushkin, L. Interrante, Macromolecules 28
(1995) 5160.
128. M. Lienhard et al., J. Amer. Chem. Soc. 116
(1994) 12020.
129. M. Lienhard, C. Weigand, T. Apple, B.
Farmer, L. Interrante, Polymer Preprints 39
(1998) no. 2, 808.
130. S. Yamjima, Y. Hasegawa et al., J. Mater. Sci.
12 (1978) 2569.
131. K. Wynne, R. Rice, Ann. Rev. Mater. Sci. 14
(1984) 297.
132. R. Laine, Y. Blum et al., ACS Symp. Ser. 360
(1988) 142.
133. J. Mark, H. Allcock, R. West, Inorganic
Polymers, Prentice Hall, Englewood Cliffs, NJ
1992, Chap. 6.
134. D. Seyferth, G. Wiseman, C. Prud’homme, J.
Am. Ceram. Soc. 66 (1983) C13.
135. Rhône-Poulence, EP 197 863 B1, 1988 (J.
Lebrun, H. Porte).
136. Bayer, US 3 853 567, 1974 (W. Verbeck).
137. Dow Corning, US 4 340 619, 1982 (J. Gaul,
Jr.).
138. H. Allcock, Adv. Mater. 6 (1994) 106.
139. H. Allcock, S. Kuharcik, J. Inorg.
Organometal. Polym. 6 (1996) no. 1, 1.
140. P. Bortolus, M. Gleria, J. Inorg. Organometal.
Polym. 4 (1994) no. 2, 95 and 4 (1994) no. 3,
205.
141. J. Inorg. Organometal. polym. 6 (1996) no. 3,
and 6 (1996) no. 4.
142. G. Bosscher, A. Jekel, J. van der Grampel, J.
Inorg. Organometal. Polym. 7 (1997) no. 1,
19.
143. H. R. Allcock: Phosphorus-Nitrogen
Compounds, Academic Press, New York
1972.
144. L. Hagnauer, C. Carraher, J. Sheats, C. U.
Pittman, Jr. (eds.): Advances in
Organometallic and Inorganic Polymer
Science, Chap. 17, Marcel Dekker, New York
1982.
145. R. Neilson, P. Wisian-Neilson in C. E.
Carraher, Jr., J. E. Sheats, C. U. Pittman, Jr.
(eds.): Advances in Organometallic and
Inorganic Polymer Sciences, Chap. 19, Marcel
Dekker, New York 1982, p. 425.
146. H. R. Allcock, CHEMTECH 5 (1975) 552.
147. H. R. Allcock, Angew. Chem. Int. Ed. Engl. 16
(1977) 147.
148. S. H. Rose, J. Polym. Sci. Part. B 6 (1968) 837.
149. C. Carraher, J. Macromol. Sci. Chem. A17
(1982) no. 8, 1293.
150. H. R. Allcock, Polymer Prepr. Am. Chem. Soc.
Div. Polym. Chem. 28 (1987) no. 1, 437.
151. R. H. Neilson et al., in M. Zeldin, K. J. Wynne
and H. R. Allcock (eds.): “Inorganic and
Organometallic Polymers,” Chap. 22, ACS
Symp. Ser. 360 (1988).
152. H. Allcock, J. Nelson, S. Reeves, C.
Honeyman, I. Manners, Macromolecules 29
(1996) 7740.
153. J. Nelson, H. Allcock, Macromolecules 30
(1997) 1854.
154. H. Allcock, S. Reeves, J. Nelson, C. Crane,
Macromolecules 30 (1997) 22B.
155. J. Nelson, H. Allcock, Polymer Reprints 39
(1998) no. 2, 631.
156. I. Dez, R. de Jaeger, J. Inorg. Organometal.
Polym. 6 (1996) no. 2, 111.
157. D. Gates, I. Manners, J. Chem. Soc., Dalton
Trans. (1997) 2525.
158. H. Allcock, S. Kuharcik, J. Inorg.
Organometal. Polym. 5 (1995) no. 4, 307.
159. J. C. Vicic, K. A. Reynard, J. Appl. Polym. Sci.
21 (1977) 3185.
160. R. Singler, G. L. Hagnauer in C. Carraher, J.
Sheats, C. U. Pittman, Jr. (eds.):
Organometallic Polymers, Chap. 26,
Academic Press, New York 1978.
161. R. E. Singler, N. S. Schneider, G. L. Hagnauer,
Polym. Eng. Sci. 15 (1975) 321.
162. H. R. Penton, Pol. Prepr. Am. Chem. Soc. Div.
Polym. Chem. 28 (1987) no. 1, 437.
163. C. Kim, H. R. Allcock, Polym. Prepr. Am.
Chem. Soc. Div. Polym. Chem. 28 (1987)
no. 1, 446.
164. R. Singler, R. Willingham, R. Lenz, A.
Furukana, H. Finkelmann, Polym. Prepr. Am.
Chem. Soc. Div. Polym. Chem. 28 (1987)
no. 1, 448.
165. G. S. Kyker, T. A. Antkowiak, Rubber Chem.
Technol. 47 (1974) 32.
166. D. F. Lohr, J. A. Beckman, Rubber and
Plastics News 16 (1982).
167. H. R. Penton in M. Zeldin, K. J. Wynne and
H. R. Allcock (eds.): “Inorganic and
Organometallic Polymers,” Chap. 21, ACS
Symp. Ser. 360 (1988).
168. Z. Pang et al., Adv. Mater. 8 (1996) 768.
169. W. Balmain, Phil. Mag. 21 (1842) 10.

170. R. Wentorf, J. Chem. Phys. 26 (1957) 956.
171. T. O’Connor, J. Amer. Chem. Soc. 84 (1962)
1753.
172. F. Bundy, R. Wentorf, J. Chem. Phys. 38
(1963) 1144.
173. Gmelin, Boron Compounds, 2nd Suppl. 1
(1983) 304.
174. N. J. Archer, Spec. Publ. Chem. Soc. No. 30
(1977) 167.
175. C. S. Kalyoncu, Ceram. Eng. Sci. Proc. (1985)
1356.
176. K. Gonslaves, X. Chen, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 5,
CRC Press, Boca Raton, FL 1996, p. 3259.
177. A. J. Banister, I. B. Gorrell, Adv. Mater. 10
(1998) no. 17, 1415.
178. T. Takahashi, H. Itoh, A. Takeuchi, J. Cryst.
Growth 47 (1979) 245.
179. M. Sano, M. Aoki, Thin Solid Films 83 (1981)
247.
180. K. L. Chopra et al., Thin Solid Films 126
(1985) 307.
181. F. P. Burt, J. Chem. Soc. 1910, 1171.
182. C. Mikulski et al., J. Am. Chem. Soc. 97
(1975) 6358.
183. M. Akhtar et al., in C. Carraher, J. Sheats, C. U.
Pittman, Jr. (eds.): Organometallic Polymers,
Chap. 30, Academic Press, New York 1978.
184. R. H. Baughman, P. A. Apgar, R. R. Chance,
A. G. MacDiarmid, A. F. Garito, J. Chem.
Phys. 66 (1977) 401.
185. A. G. MacDiarmid in R. B. King (ed.):
Inorganic Compounds with Unusual
Properties, Chap. 6, American Chemical
Society, Washington D.C. 1976.
186. R. L. Green et al., Lect. Notes Phys. 65 (1977)
603.
187. J. Passmore, M. Rao, J. Chem. Soc., Chem.
Commun, 1980 1268.
188. F. Kennett, G. MacLean, J. Passmore, M. Rao,
J. Chem. Soc., Dalton Trans 1982 851.
189. A. Banister, Z. Hauptam, J. Passmore, C.
Wong, P. White, J. Chem. Soc., Dalton Trans
1986 2371.
190. A. Banister, Z. Haupton, A. Kendrick, J.
Chem. Soc., Chem. Commun. 1983 1016.
191. A. Banister, Z. Hauptman, A. Kendrick, R.
Small, J. Chem. Soc., Dalton Trans 1987 915.
H. Fritz, R. Bruchhaus, Z. Naturforsch. B. 38
(1983) 1375.
192. R. L. Greene, G. B. Street, L. J. Suter, Phys.
Rev. Lett. 34 (1975) 577.
193. W. H. Jones, R. Bardo, J. Phy. Chem. 97
(1993) 4974.
194. I. Kakada, J. Cryst. Growth 55 (1981) 447.
Inorganic Polymers 27
195. H. Kahlert, B. Kundu, Mater. Res. Bull. 11
(1976) 967.
196. R. Scranton, J. Appl. Phys. 48 (1977) 3838.
197. R. Scranton, J. Best, J. McCaldin, J. Vac. Sci.
Technol. 14 (1977) 930.
198. M. Cohen, J. Harris, Appl. Phys. Lett. 33
(1978) 812.
199. A. Thomas, J. Woods, Z. Hauptman, J. Phys.
D: Appl. Phys. 16 (1983) 1123.
200. Asahi Chemical Industry Co. Ltd.,
JP 55 137 671, 1980; 58 93 164, 1983;
58 93 177, 1983; 59 49 155, 1984.
201. Electricité de France Service National,
EP 97 078 B1, 1986 (J. Badoz et al.).
202. Sony Corp., JP 55 108 944 A, 1980 (T.
Yoshioka, M. Ogawa, K. Hayashi).
203. Sony Corp., EP 517 165 A1, 1992; EP 517 165
B1, 1997; JP 07 230 984, 1955 (S. Kadomura).
204. Sony Corp., JP 07 263 420, 1995 (T.
Nagayama).
205. Sony Corp., JP 07 263 426, 1995 (N. Akiba).
206. The United States of America as represented
by the Secretary of the Army, US 4 206 705,
1980 (Z. Iqbal, H. Fair, D. Downs).
207. Nippon Denso Co. Ltd., JP 59 129 261, 1984
(H. Murametsu).
208. Sony Corp., JP 07 130 735, 1995 (T.
Nagayama).
209. Sony Corp., JP 09 148 285, 1997 (J. Sato).
210. Toshiba Corp., JP 05 136 458, 1993 (T.
Kamakura).
211. H. B. Mark, A. Voulgaropoulos, C. Meyer, J.
Chem. Soc., Chem. Commun. (1981) 1021.
212. J. Rubinson, T. Behymer, H. Mark, J. Amer.
Chem. Soc. 104 (1982) 1224.
213. R. Nowak, C. Joyal, D. Weber, J. Electroanal.
Chem. 143 (1983) 413.
214. B. R. Currell, A. J. Williams, A. J. Mooney and
B. J. Nash, Adv. Chem. Ser. 1 (1974) 1.
215. J. Sheats, C. Carraher, C. Pittman, M. Zeldin,
B. Currell: Inorganic and Metal-Containing
Polymeric Materials, Plenum, New York
1990.
216. C. Pittman, C. Carraher, M. Zeldin, B.
Culbertson, J. Sheats: Metal-Containing
Polymeric Materials, Plenum, New York
1996.
217. M. Zeldin, K. Wynne, H. Allcock: Inorganic
and Organometallic Polymers, ACS,
Washington, DC 1998.
218. Macromolecule-Metal Complexes III,
Seclected Papers from seminar of July 1989; J.
Macromol. Sci. Chem. A27 (1990) nos. 9 – 11,
1109.
28 Inorganic Polymers
219. I. Manners, Ann. Rep. Prog. Chem., Sect. A.
Inorg. Chem. 88 (1991) 77; 89 (1992) 93; 90
(1993) 103; 91 (1994) 131; 92 (1995) 127; 93 239.
(1996) 129.
220. J. Mark, H. Allcock, R. West, Inorganic 240.
Polymers, Prentice Hall, New York 1992.
221. J. Sheats, C. Carraher, C. Pittman, M. Zeldin, 241.
in C. Craver, C. Carraher (eds.): Applied
Polymer Science, Elsevier, New York 1999.
222. S. Takahashi, K. Sonogashira, in J. Salamone 242.
(ed.): Polymeric Materials Encyclopedia, vol.
6, CRC Press, Boca Raton, FL 1996, p. 4804.
223. A. D. Hunter, A. Guo, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 6, 243.
CRC Press, Boca Raton, FL 1996, p. 4813.
224. T. Endo, I. Tomita, in J. Salamone (ed.): 244.
Polymeric Materials Encyclopedia, vol. 6,
CRC Press, Boca Raton, FL 1996, p. 4822. 245.
225. R. Wei, L. Ya, W. Jinguo, X. Qifeng, in, J.
Salamone (ed.): Polymeric Materials 246.
Encyclopedia, vol. 6, CRC Press, Boca Raton,
FL 1996, p. 4826.
226. W. L. Driessen (ed.): Macromolecule-Metal 247.
Complexes, Huthig & Wepf, Oxford, CT 1998.
227. F. Ciardelli, E. Tsuchida, D. Wohrle (eds.): 248.
Macromolecule-Metal Complexes,
Springer-Verlag, Berlin 1996.
228. A. Pomogailo, V. Savost’yanov: Synthesis and
Polymerization of Metal-Containing
Monomers, CRC Press, Boca Raton, New 249.
York 1994.
229. B. P. Block, Inorg. Macromol. Rev. 1 (1970)
no. 2, 115.
230. C. Carraher, R. Schwarz, J. Schroeder, M. 250.
Schwarz, J. Macromol. Sci. Chem. A15
(1981) no. 5, 773. 251.
231. E. Neuse, H. Rosenberg: Metallocene
Polymers, Marcel Dekker, New York 1970. 252.
232. Dow Corning, GB 1 582 081, 1980.
233. C. Carraher, F. Li, D. Siegman-Louda, C.
253.
Butler, J. Ross, PMSE 77 (1997) 499.
234. C. Carraher, V. Saurino, C. Butler, D. Sterling,
PMSE 72 (1995) 192.
235. C. Carraher, D. Sterling, C. Butler, T.
254.
Ridgway, Biotechnical Polymers, Chap. 13,
Technomic, Lancaster, PA 1995.
255.
236. C. Carraher, M. Nagata, H. Stewart, S. Miao,
S. Carraher, A. Gaonkar, C. Highland, F. Li,
256.
PMSE 79 (1998) 52.
237. C. Carraher, A. Gaonkar, H. Stewart, S. Miao,
257.
D. Mitchell, C. Barosy, M. Colbert, R.
258.
Duffield, Polymer Reprints 38 (1997) no. 2,
572.
238. C. Carraher, H. Stewart, W. Soldani, J. Dela 259.
Torre, B. Pandya, L. Reckleben,
Metal-Containing Polymeric Materials,
Plenum, NY 1996.
C. Carraher et al., Polym. Mater. 55 (1986)
469.
C. Carraher, V. Foster, R. Linville, D. Stevison,
Polym. Mater. 56 (1987) 401.
C. Carraher, R. Schwarz, J. Schroeder, M.
Schwarz, H. M. Molloy, Org. Coat. Plast.
Chem. 43 (1981) 798.
C. Carraher, R. Schwarz, J. Schroeder, M.
Schwarz in C. Carraher, J. Preston (eds.):
Interfacial Synthesis, vol. 3, Chap. 6, Marcel
Dekker, New York 1982.
C. Carraher et al., Org. Coat. Plast. Chem. 44
(1982) 753.
J. Wrobleski, D. B. Brown, Inorg. Chem. 18
(1979) 498, 2738.
H. Matsuda, H. Nakaniski, M. Kato, J. Polym.
Sci. Polym. Lett. Ed. 22 (1984) 107.
J. E. Sheats, C. Carraher, C. Pittman (eds.):
Metal-Containing Polymeric Systems, Plenum
Press, New York 1985.
X. Zhang, R. West, J. Polym. Sci. Polym.
Chem. Ed. 22 (1984) 159.
R. Liepins, M. Timmons, N. Morosoff, J.
Surles in J. Sheats, C. Carraher, C. Pittman
(eds.): Metal-Containing Polymeric Systems,
Plenum Publishing, New York 1985,
pp. 225 – 236.
H. Yasuda et al., in J. Sheats, C. Carraher,
C. U. Pittman (eds.): Metal-Containing
Polymeric Systems, Plenum Publishing, New
York 1985, pp. 275 – 290.
C. Carraher et al., Organic Coatings Plastics
Chemistry 44 (1981) 753.
E. Savinova et al., J. Mol. Catal. 32 (1985)
149.
M. Kaneko, A. Yamada, Adv. Polym. Sci. 55
(1983).
M. S. Wrighton (ed.): Interfacial
Photoprocesses: Energy Conversion and
Synthesis, American Chemical Society,
Washington 1980.
U. T. Müller-Westerhoff, A. I. Nazzal, US
4 379 740, 1983.
F. C. Anson et al., J. Am. Chem. Soc. 106
(1984) 59.
E. R. Savinova et al., J. Mol. Catal. 32 (1985)
149, 159.
in [23] pp. 137 – 147.
B. J. Spalding, Chemical Week 135 (1986)
no. Oct. 8, 29.
C. Carraher, S. Tsuji, J. DiNunzio, W. Feld,
Polym. Mater. 55 (1986) 875.

260. C. Carraher, J. Schroeder, Polym. Prepr. 16
(1975) 659, J. Polym. Sci. Polym. Lett. Ed. 13
(1975) 215.
261. L. Donaruma, Polym. Prepr. Am. Chem. Soc.
Div. Polym. Chem. 22 (1981) 1.
262. B. Jiang et al., Coord. Chem. Rev. 171 (1998)
365.
263. W. E. Jones, L. Hermans, B. Jiang, in V.
Ramamurthy, K. Schanze (eds.): Molecular
and Supramolecular Photochemistry, vol. 2,
Marcel Dekker, New York 1999.
264. C. Pittman, J. Paint Technol. 39 (1967) 585.
265. N. E. Wolff, R. J. Pressley, Appl. Phys. Lett. 2
(1963) 152.
266. in [17] p. 425.
267. J. E. Sheats et al., in B. M. Culbertson, C. U.
Pittman, Jr. (eds.): New and Unusual
Monomers and Polymers, Plenum Press, New
York 1983, pp. 83 – 98; Polym. Mater. Sci. Eng.
49 (1983) no. 2, 363.
268. N. Bilow, H. Rosenberg, J. Polym. Sci. Part
A-1 7 (1969) 2689.
269. N. Bilow, H. Rosenberg, US 3 640 961, 1972.
270. C. U. Pittman, Jr., R. F. Felis, J. Organomet.
Chem. 72 (1974) 389.
271. C. U. Pittman, R. F. Felis, J. Organomet.
Chem. 72 (1974) 399.
272. R. Tannenbaum, E. P. Goldberg, C. L.
Flenniken in J. Sheats, C. Carraher, C. Pittman
(eds.): Metal-Containing Polymeric Systems,
Plenum Publishing, New York 1985,
pp. 303 – 340.
273. C. Carraher, X. Xu, PMSE 75 (1996).
274. C. Carraher, X. Xu: Modification of Polymers,
Plenum, New York 1997.
275. C. Carraher, V. Forster, R. Linville, D.
Steverson, R. Ventatachalam, Adhesives,
Sealants, and Coatings for Space and Harsh
Environments, Chap. 19, Plenum, NT 1988.
276. C. Carraher, A. Taylor-Murphy, PMSE 76
(1997) 409.
277. S. Yajima, H. Hayashi, M. Omori, Chem.
Letters. (1975) 931 and 1209.
278. R. West, Am. Ceram. Soc. Bull. 62 (1983) 825.
279. D. Seyferth, G. H. Wiseman, J. Amer. Ceram.
Soc. 67 (1984); Massachusetts Institute of
Technology, US 4 482 669, 1984 (D. Seyferth,
G. H. Wiseman).
280. R. Laine, Y. Blum, A. Chow, R. Hamlin, K.
Schwartz, D. Rowecliff, Polymer Reprints 28
(1987) no. 1, 393.
281. M. Arai, S. Sakurada, T. Isoda, T. Tomizawa,
Polymer Reprints 28 (1987) no. 1, 407.
282. C. Narula, R. Paine, R. Schaeffer, Polymer
Reprints 28 (1987) no. 1, 454.
Inorganic Polymers 29
283. C. Carraher, H. Zhuang, F. Medina, D. Baird,
R. Pennisi, B. Landreth, F. Nounou, PMSE 62
(1990) 633.
284. M. Kaneko, A. Yamada, Adv. Polym. Sci.
(1983) 55.
285. US 4 379 740, 1983 (U. Muller-Westerhoff, A.
Nazzal).
286. F. Anson, J. Amer. Chem. Soc. 106 (1984) 59.
287. B. J. Spaulding, Chemical Week, Oct 8
(1986)29 and C. M. Elliott, Abstracts, 193 ACS
Meeting, Denver CO, April, 1987, INOR 232.
288. M. MacLachlan, P. Aroca, N. Coombs, I.
Manners, G. Ozin, Adv. Mater. 10 (1998) 144.
289. Proceedings of the IIIrd Internat. Conference
on Ultrastructure Processing of Glasses,
Ceramics and Composites, Feb. 24 – 27, 1987.
290. C. J. Brinker, D. Clark, D. R. Ulrich (eds.):
Better Ceramics Through Chemistry II,
Materials Res. Soc., Pittsburgh, Pa., 1986.
291. L. L. Hench, D. R. Ulrich (eds.): Science of
Ceramic Chemical Processing,
Wiley-Interscience, New York 1986.
292. J. O’Reilly, B. Coltrain, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 6,
CRC Press, Boca Raton, FL 1996, p. 4772.
293. G. Wilkes, J. Wen, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 6,
CRC Press, Boca Raton, FL 1996, p. 4782.
294. Y. Chujo, in J. Salamone (ed.): Polymeric
Materials Encyclopedia, vol. 6, CRC Press,
Boca Raton, FL 1996, p. 4793.
295. C. J. Brinker et al., “Inorganic and
Organometallic Polymers,” ACS Symp. Ser.
360 (1988); D. R. Ulrich, CHEMTECH 18
(1988) 242.
296. H. K. Schmidt, “Inorganic and Organometallic
Polymers,” ACS Symp. Ser. 360 (1988).
297. H. Schmidt, H. Scholze, G. Tunker, J. Non
Cryst. Solids 80 (1986) 557.
298. H. Schmidt, H. Scholze, Springer Proceedings
in Physics, vol. 6, Aerogels, Springer
Publishing Co., 1986, p. 49.
299. G. Philipp, H. Schmidt, J. Non Cryst. Solids
63 (1984) 261.
300. G. Tunker, H. Patzelt, H. Schmidt, H. Scholze,
Glastechn. Ber. 59 (1986) 272.
301. S. Sakka, Polym. Prepr. Am. Chem. Soc. Div.
Polym. Chem. 28 (1987) no. 1, 430.
302. P. Brooke, H. Schurmans, J. Verhoest:
Inorganic Fibers and Composite Materials: A
Survey of Recent Developments, Pergamon
Press, Oxford, UK, 1984.
303. H. Huang, R. H. Glaser, G. L. Wilkes,
“Inorganic and Organometallic Polymers,”
ACS Symp. Ser. 360 (1988).
30 Inorganic Polymers
304. H. Dislich, Angew. Chem. Int. Ed. Engl. 106
(1970) 363.
305. B. E. Yoldas, J. Mater. Sci. 12 (1977) 203; 14
(1979) 1843.
306. C. Carraher, CHEMTECH 1972, 741.
307. V. C. Farmer, A. R. Fraser, J. M. Tait, J. Chem.
Soc. Chem. Commun. 1977, 462.
308. K. Wada et al., Clay Miner. 8 (1970) 487.
309. G. Alberti, U. Costantino, J. Chromatogr. 102
(1974) 5.
310. N. H. Ray: Inorganic Polymers, Chap. 7,
Academic Press, New York 1978.
311. J. D. Birchall, J. A. Bradbury, J. Dinwoodie in
W. Watt, B. V. Perov (eds.): Handbook of
Composites, North Holland, Amsterdam
1985, p. 115.
312. J. D. Birchall: “Inorganic Fibers” in M. B.
Bever (ed.): Encyclopedia of Material
Sciences and Engineering, Pergamon Press,
Oxford, UK 1986, p. 2333.
313. Sumitomo Chemical Co., US 4 101 615, 1973
(S. Horikiri, K. Tsuji, Y. Abe); 4 152 149, 1977
(S. Horikiri, K. Tsuji, Y. Abe).
314. J. Tretzel et al., “Production, Processing and
Application of Enka Silica Fibers,” 57th
Annual TRI Conference, Charlotte, N.C., April
21 – 22, 1987.
315. W. C. Miller in M. Grayson (ed.):
Encyclopedia of Textiles, Fibers and
Nonwoven Fabrics, J. Wiley and Sons, New
York 1984, p. 443.
316. Chem. Week 13 (1987) no. April 22.
317. J. C. Romine, Ceram. Eng. Sci. Proc. 8 (1987)
755.
318. J. O. Carlsson: “Boron Fibers” in M. B. Bever
(ed.): Encyclopedia of Material Science and
Engineering, Pergamon Press, Oxford, UK,
1986, p. 402.
319. J. O. Carlsson: “Silicon Carbide Fibers” in
M. B. Beuer (ed.): Encyclopedia of Material
Science and Engineering, Pergamon Press,
Oxford, UK, 1986, p. 4406.
320. S. Yajima, Y. Hasegawa, J. Hayashi, M. Imura,
J. Mater. Sci. 13 (1978) no. 12, 2569.
321. Y. Hasegawa, K. Okamura, Yogyo Kyokaishi
95 (1987) 99, 103.
322. W. Verbeck, DE-OS 2 218 960, 1973.
323. K. Okamura, M. Sato, Y. Hasegawa, Ceram.
Int. 13 (1987) 55.
324. T. Yamamura et al., “Proc. World Congress on
High Tech Ceramics, 6th Int. Meeting on
Modern Ceramics Technologies,” Mater. Sci.
Monogr. 38 (1987).
325. N. Tsubokawa, T. Yoshihara, H. Ueno, in J.
Salamone (ed.): Polymeric Materials
Encyclopedia, vol. 2, CRC Press, Boca Raton,
FL 1996, p. 973.
326. C. Carraher: Polymer Chemistry: An
Introduction, 4th ed., Dekker, New York 1996,
p. 122.
327. K. Okamura, Composites 18 (1987) 107.
328. W. Noll: Chemistry and Technology of
Silicones, Academic Press, New York 1968.
329. W. Lynch: Handbook of Silicon Rubber
Fabrication, Van Nostrand, New York 1978.
330. M. Ranney: Silicones, Noyes Data Corp., Park
Ridge, N.J., 1977.
331. B. Arkles, W. Peterson, R. Anderson:
Organosilicon Compounds: Register and
Review, Petrarch Systems, Bristol, Pa., 1982.
332. B. Arkles, CHEMTECH 13 (1983) 542.
333. A. Isquith, L. Abbot, P. Walters, Appl.
Microbiol. 24 (1972) 859.
334. B. Arkles, W. Brinigar, A. Miller in D.
Leyden, W. Collin (eds.): Silylated Surfaces,
Gordon Breach, New York 1980.
335. E. Sako, O. Kawamura, US 3 746 196, 1973.
336. W. Robb, Ann. N.Y. Acad. Sci. 146 (1968) 119.
337. M. Nelson, US 3 020 260, 1962.
338. E. Jeram, US 4 072 635, 1978.
339. T. Cronin, US 3 293 663, 1965.
340. I. V. Yannas, J. F. Burke, J. Biomed. Mater.
Res. 14 (1980) 65, 107.
341. N. Dagalakis et al., J. Biomed. Mater. Res. 14
(1980) 511.
342. R. A. Erb, US 4 245 623, 1981.
343. R. A. Erb, T. P. Reed, J. Reprod. Med. 23
(1979) 65.
344. A. Zaffaroni, US 3 905 360, 1975.
345. P. Hittmaier, W. Hechtel, H. Eckhart, US
4 035 453, 1977.
346. C. N. Raptis, R. Yu, J. G. Knapp, J. Prosthet.
Dent. 4414 (1980) 447.
347. A. B. Swanson, US 3 875 594, 1975.
348. T. Honda et al., J. Thorac. Cardiovasc. Surg.
69 (1975) 92.
349. D. Laurin, US 3 994 988, 1976.
350. D. G. LeGrand, J. Magila, Eng. Sci. 10 (1970)
349.
351. J. Riffle, R. Freelin, A. Banthia, J. McGrath, J.
Macromol. Sci. Chem. A15 (1981) 967.
352. E. Pierce, N. Dibelus, Trans. Am. Soc. Artif.
Intern. Organs 14 (1968) 220.
353. J. Montalvo, US 3 869 354, 1975.
354. J. Brown, G. Slusarczuk, O. LeBlanc, US
3 743 588, 1971.
355. H. Allcock in C. Carraher, J. E. Sheats, C. U.
Pittman, Jr. (eds.): Organometallic Polymers,
Academic Press, New York 1978, p. 283.

356. H. Allcock, R. Allen, J. O’Brien, J. Chem. Soc.
Chem. Commun. 1976, 717.
357. C. Carraher, W. J. Scott, J. Schroeder, D. J.
Giron, J. Macromol. Sci. Chem. A15 (1981)
no. 4, 625.
358. C. Carraher et al., in C. Gebelein, C. Carraher
(eds.): Polymeric Materials in Medication,
Chap. 15, Plenum Publishing, New York 1985.
359. C. Carraher in C. Gebelein, C. Carraher (eds.):
Bioactive Polymeric Systems, Chap. 22,
Plenum Publishing, New York 1985.
360. C. Carraher, I. Lopez, D. Giron, Polym. Mater.
53 (1985) 644.
361. C. Carraher, I. Lopez, D. Giron in C. Gebelein
(ed.): Advances in Biomedical Polymers,
Plenum Publishing, New York 1987, p. 311.
362. C. Carraher, N. Bigley, M. Trombley, D.
Giron, Polym. Mater. 57 (1987) 177.
363. A. Winston, in C. Gebelein, C. Carraher (eds.):
Bioactive Polymeric Systems, Chap. 21,
Plenum Publishing, New York 1985.
364. A. Winston, Polym. News 10 (1984) 6.
365. C. Carraher et al., in C. Carraher, C. Gebelein
(eds.): Biological Activities of Polymers,
Chap. 2, American Chemical Society,
Washington, D.C., 1982.
366. D. Anderson, J. Mendoza, B. Garg, R. V.
Subramanian in C. Carraher, C. Gebelein
(eds.): Biological Activities of Polymers,
Chap. 3, American Chemical Society,
Washington, D.C., 1982.
367. R. V. Subramanian, B. Garg, J. Corredor in C.
Carraher, J. Sheats, C. U. Pittman, Jr. (eds.):
Organometallic Polymers, Chap. 19,
Academic Press, New York 1978.
368. C. Carraher et al., Polym. Mater. 57 (1987)
186.
369. C. Carraher, D. Giron, J. Schroeder, C.
McNeely, US 4 312 981, 1982.
370. J. A. Montemarano, E. J. Dyckman, J. Paint
Technol. 47 (1975) 59.
371. A. T. Phillip, Prog. Org. Coat. 2 (1973/74)
159.
372. J. Leebrick, US 3 167 473, 1965.
373. D. Atherton, J. Verborgt, M. A. Winkeler, J.
Coat. Technol. 51 (1979) 88.
374. S. Matsuda, H. Kudara, US 4 174 339, 1979.
375. Shipcare International, 10 (1978) no. 11, 23.
376. R. W. Drisko, T. B. O’Neill, L. K. Schwab,
Technical Note no. N-1480, Naval Facilities
Engineering Command, Point Hueneme, CA,
May 1977.
377. Y. Chujo, in J. Salamone (ed.): Polymeric
Materials Encyclopedia, vol. 6, CRC Press,
Boca Raton, FL 1996, p. 4795.
Inorganic Polymers 31
378. W. Gerrard, J. Chem. Soc. 1960, 3170.
379. R. Bedell, M. J. Frazer, W. Gerrard, J. Chem.
Soc. 1960, 4037.
380. R. W. Upson, US 2 511 310, 1950.
381. K. Hizawa, E. Nojimoto, JP 4791, 1953.
382. J. M. Thompson: “Other Boron-Containing
Polymers” in M. F. Lappert, G. J. Leigh (eds.):
Developments in Inorganic Polymer
Chemistry, Elsevier, Amsterdam 1962, p. 57.
383. R. M. Washburn, E. Levens, C. F. Albright,
F. A. Billig: “Metal-Organic Compounds,”
Adv. Chem. Ser. 23 (1959) 139, 153.
384. W. Kuchen, R. Brinkmann, Angew. Chem. 72
(1960) 564.
385. A. B. Burg, Angew. Chem. 72 (1960) 183.
386. R. I. Wagner, F. F. Caserio, J. Inorg. Nucl.
Chem. 11 (1959) 259.
387. T. Wartik, E. F. Apple, J. Am. Chem. Soc. 77
(1955) 6400.
388. H. Rosenberg, F. L. Hedberg, Tech. Rept.
AFML-TR-69-68, AD 863, 817 (1969).
389. H. Steinberg, R. J. Brotherton: Organoboron
Chemistry, vol. 2, J. Wiley and Sons, New
York 1966, pp. 329 – 340, 363 – 364, 418.
390. B. M. Mikhailov, A. F. Galkin, Bull. Acad. Sci.
USSR, Div. Chem. Sci. (Engl. Transl.) 12
(1963) 575.
391. R. H. Toeniskoetter, R. Didchenko, CA
685 580, 1964 to Union Carbide.
392. R. H. Toeniskoetter, K. A. Killip, J. Am. Chem.
Soc. 86 (1964) 690.
393. V. Gutmann, A. Metter, R. Schlegel, Monatsh.
Chem. 94 (1963) 1071.
394. G. Chainani et al., J. Appl. Chem. 15 (1965)
372.
395. K. O. Knollmuller, R. N. Scott, H. Kwasnik,
J. F. Sieckhaus, J. Polym. Sci. Polym. Chem.
Ed. A1, 9 (1971) 1071.
396. S. Papetti et al., J. Polym. Sci. Polym. Chem.
Ed. 4 (1966) 1623.
397. R. P. Alexander, H. Schroeder, Angew. Chem.
76 (1964) 278.
398. H. A. Schroeder et al., Inorg. Chem. 8 (1969)
no. 11, 2444.
399. H. A. Schroeder, Inorg. Macromol. Rev. 1
(1970) 45.
400. I. B. Atkinson, D. B. Clapp, W. F. Flavell,
Polym. Prepr. Am. Chem. Soc. Div. Polym.
Chem. 13 (1972) 770.
401. N. Semenuk, S. Papetti, H. Schroeder, Inorg.
Chem. 8 (1969) no. 11, 2441.
402. J. Rinse, Ind. Eng. Chem. 56 (1964) no. 5, 42.
403. DuPont, US 2 744 071, 1956 (C. N. Theobald).
404. K. C. Pande, R. C. Mehrotra, Z. Anorg. Allg.
Chem. 286 (1956) 291.
32 Inorganic Polymers
405. Peter Lunt and Co., GB 573 083, 1945 (E. J.
Lush).
406. Hardman and Holden Ltd., GB 772 144, 1957
(J. Rinse).
407. A. L. McCloskey et al., WADC Tech. Rept.
59-761 (1959) pp. 117 – 200.
408. Koninklijke Industrieele Maatschappij
voorheen Noury, GB 783 679, 1957 (N. V. Van
der Lande).
409. K. A. Andrianov: Metalorganic Polymers,
Chap. 5, Interscience Publishers, New York
1965.
410. K. A. Andrianov, V. M. Novikov, Vysokomol.
Soedin. 1 (1959) 1390.
411. V. Kugler, J. Polym. Sci. 29 (1958) 637.
412. T. R. Patterson et al., J. Am. Chem. Soc. 81
(1959) 4213.
413. J. I. Jones: “Polymetallosiloxanes” in M. L.
Lappert, G. J. Leigh (eds.): Developments in
Inorganic Polymer Chemistry, Part I, Part II,
Chap. 7, 8, Elsevier, New York 1962.
414. D. C. Bradley: “Polymeric Metal Alkoxides,
Organometalloxanes and
Organometalloxanosiloxanes” in F. G. A.
Stone, W. A. G. Graham (eds.): Inorganic
Polymers, Chap. 7, Academic Press, New York
1962.
415. I. K. Stavitskii, Vysokomol. Soedin. 1 (1959)
1502.
416. S. N. Borisov, N. G. Sviridova, Vysokomol.
Soedin. 3 (1961) 50.
417. R. D. Crain, P. E. Koenig, WADC Tech. Rept.
69-427 (1960) p. 15.
418. F. A. Henglein, R. Lang, L. Schmack,
Makromol. Chem. 22 (1957) 103.
419. Du Pont, US 2 512 058, (H. C. Gulledge).
420. T. Boyd, GB 728 751, 1955.
421. T. Boyd, US 2 716 656, 1955.
422. A. Hancock, R. Sidlow, J. Oil Colour Chem.
Assoc. 35 (1952) 28.
423. W. H. Post, W. T. Schwartz, Jr., AD 255 545,
U.S. Dept. Commerce Office Tech. Serv.,
Washington D.C., 1961.
424. K. A. Andrianov, S. V. Pichkhadze,
Vysokomol. Soedin. 3 (1961) 577.
425. C. Ungurenasu, I. Haidec, Rev. Roumaine
Chim, 13 (1968) 957.
426. C. C. Currie, US 2 672 455, 1954.
427. Monsanto Inc., GB 723 989, 1955.
428. R. M. Kary, US 2 646 440, 1953.
429. W. P. Gill et al., Phys. Rev. Lett. 38 (1977)
1305.
430. F. B. Bundy, K. T. Dunn, J. Chem. Phys. 71
(1979) 1550.
431. P. W. Bridgman, Proc. Am. Acad. Arts Sci. 74
(1941) 399.
432. E. Whalley, Can. J. Chem. 38 (1960) 2105.
433. Y. Okamoto, Z. Iqbal, R. H. Baughman, J.
Macromol. Sci. Chem. A25 (1988) no. 5 – 7,
799.
434. W. S. Chan, A. K. Jonsher, Phys. Status Solidi
32 (1969) 749.
435. J. Tsukamoto, A. Takahashi, Jpn. J. Appl.
Phys. 25 (1986) L338.
436. M. Berthelot, Ann. Chim. Phys. 11 (1936) 15.
437. Y. Asano, Jpn. J. Appl. Phys. 22 (1983) 1618.
438. L. A. Wall, D. W. Brown, J. Polym. Sci. Part C
4 (1964) 1151.
439. Z. Iqbal et al., J. Chem. Phys. 85 (1986) 4019.
440. G. Grimm, H. Metzger, Ber. Dtsch. Chem.
Ges. 69 (1936) 1356.
441. A. J. Brown, E. Whalley, Inorg. Chem. 7
(1968) 1254.
442. Y. Okamoto, P. S. Wojciechowski, J. Chem.
Soc. Chem. Commun. 1982, 386.
443. H. Kobayashi et al., Chem. Lett. 1983, 1407.
444. H. Kobayashi, A. Kobayashi, Y. Sasaki, Mol.
Cryst. Liq. Cryst. 118 (1985) 427.
445. A. Kobayashi et al., Bull. Chem. Soc. Jpn. 59
(1986) 3821.
446. H. Emeleus, J. Anderson: Modern Aspects of
Inorganic Chemistry, 3rd ed.,
Routledge & Kegan, London 1960.
447. F. Cataldo, J. Inorg. Organomet. Polym. 7
(1997) no. 1, 35.
448. Y. L. Gefter: Organophosphorus Monomers
and Polymers, Pergamon Press, London 1962.
449. H. Noth, W. Schragle, Z. Naturforsch. B 166
(1961) 473.
450. C. U. Pittman, Jr., J. Polym. Sci. Polym. Chem.
Ed. 5 (1967) 371.
451. C. Shimasaki, H. Kitano, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 7,
CRC Press, Boca Raton, FL 1996.
452. Phosphorus, Sulfur, Silicon, Relat. Elem.,
ongoing journal.
453. D. Corbridge: “Studies in Inorganic Chemistry
10: Phosphorus, An Outline of Its Chemistry,”
Biochemistry and Technology, 4th ed.,
Elsevier, Amsterdam 1990.
454. G. Borissov, in J. Salamone (ed.): Polymeric
Materials Encyclopedia, vol. 7, CRC Press,
Boca Raton, FL 1996, p. 5092.
455. D. Liaw, D. Wang, in J. Salamone (ed.):
Polymeric Materials Encyclopedia, vol. 9,
CRC Press, Boca Raton, FL 1996, p. 6569.
456. E. W. Neuse, G. J. Chris, J. Macromol. Sci.
A-1 (1967) 371.
457. H. R. Allcock: “Rings, Clusters and Polymers
of the Main Group Elements,” ACS Symp. Ser.
232 (1983).
 Insect Control 1

Insect Control
Acaricides is a separate keyword

Robert L. Metcalf, Department of Entomology, University of Illinois, Urbana, Illinois 61801, United States

1. Insects as Enemies of Man . . . . . 3 6.7. Phosphorothioate Esters of

1.1. Development of Insecticides . . . . 4 S-Methyl Heterocycles . . . . . . . 34
1.2. Insect Resistance to Insecticides . 4 6.8. Miscellaneous Organophosphorus
1.2.1. Cross-Resistance and Multiple Re- Esters . . . . . . . . . . . . . . . . . . 35
sistance . . . . . . . . . . . . . . . . . . 5 6.9. Mode of Action and Toxicity . . . 36
1.2.2. Persistence of Genes for Insecticide 6.9.1. Inhibition of Acetylcholinesterase . 36
Resistance . . . . . . . . . . . . . . . . 6 6.9.2. Delayed Neuropathy . . . . . . . . . 38
1.3. Resurgence and Secondary Pests 7 6.9.3. Selective Toxicity . . . . . . . . . . . 38
1.4. Insecticides and Environmental 6.10. Environmental Toxicology . . . . . 39
Quality . . . . . . . . . . . . . . . . . 7 7. Carbamate Insecticides . . . . . . . 39
1.5. Integrated Pest Management . . . 8 7.1. Dimethylcarbamates of
2. Inorganic Insecticides . . . . . . . . 9 Heterocyclic Enols . . . . . . . . . . 40
2.1. Arsenicals . . . . . . . . . . . . . . . . 9 7.2. Methylcarbamates of Phenols . . 40
2.2. Fluorides . . . . . . . . . . . . . . . . 10 7.3. Methylcarbamates of Oximes . . . 43
3. Natural Product Insecticides . . . 10 7.4. Mode of Action . . . . . . . . . . . . 43
3.1. Nicotinoids . . . . . . . . . . . . . . . 11 7.5. Environmental Toxicology . . . . . 44
3.2. Rotenoids . . . . . . . . . . . . . . . . 11 8. Formamidines . . . . . . . . . . . . . 44
3.3. Veratrine Alkaloids . . . . . . . . . 12 9. Insect Growth Regulators . . . . . 44
3.4. Ryanodine . . . . . . . . . . . . . . . 12 9.1. Juvenoids . . . . . . . . . . . . . . . . 44
3.5. Pyrethroid Insecticides . . . . . . . 12 9.2. Chitin Synthesis Inhibitors . . . . 45
3.5.1. Synthetic Pyrethroids . . . . . . . . . 13 10. Other Types of Insecticides . . . . 45
3.5.1.1. Chrysanthemate Esters . . . . . . . . 13 11. Fumigants . . . . . . . . . . . . . . . 46
3.5.1.2. Chrysanthemate Analogues . . . . . 15 12. Microbial Insecticides . . . . . . . . 47
3.5.1.3. Phenylacetate Esters . . . . . . . . . 16 13. Genetic Control . . . . . . . . . . . . 48
3.5.2. Mode of Action . . . . . . . . . . . . 17 13.1. Sterile Male Release . . . . . . . . . 48
3.5.3. Environmental Toxicology . . . . . . 17 13.2. Chemosterilants . . . . . . . . . . . 49
3.5.4. Pyrethroid Synergists . . . . . . . . . 17 13.3. Host Plant Resistance . . . . . . . . 49
4. Dinitrophenols . . . . . . . . . . . . 18 14. Repellents . . . . . . . . . . . . . . . . 50
5. Organochlorine Insecticides . . . . 19 14.1. Repellents for Plant-Feeding In-
5.1. DDT and Derivatives . . . . . . . . 19 sects . . . . . . . . . . . . . . . . . . . 50
5.1.1. Mode of Action . . . . . . . . . . . . 20 14.2. Repellents for Blood-Feeding In-
5.1.2. Environmental Toxicology . . . . . . 20 sects . . . . . . . . . . . . . . . . . . . 50
5.2. Hexachlorocyclohexanes . . . . . . 21 14.3. Repellents for Wood-Feeding In-
5.3. Cyclodienes . . . . . . . . . . . . . . 21 sects . . . . . . . . . . . . . . . . . . . 52
5.4. Chlorinated Camphenes . . . . . . 25 14.4. Repellents for Fabric-Eating Pests 52
6. Organophosphorus Insecticides . 25 15. Attractants . . . . . . . . . . . . . . . 53
6.1. Phosphoric Acid Anhydrides . . . 26 15.1. Sex Pheromone Attractants . . . . 53
6.2. Vinyl Phosphates . . . . . . . . . . . 27 15.2. Aggregation Pheromones . . . . . 55
6.3. Aliphatic Phosphorothioate Es- 15.3. Kairomone Attractants . . . . . . . 55
ters . . . . . . . . . . . . . . . . . . . . 28 16. Insecticide Formulation . . . . . . 57
6.4. Phosphorothioate Esters of Phe- 17. Insecticide Residues in Foods . . . 57
nols . . . . . . . . . . . . . . . . . . . . 31 17.1. Residue Persistence . . . . . . . . . 57
6.5. Phosphonothioate Esters of Phe- 17.2. Environmental Chemistry of In-
nols . . . . . . . . . . . . . . . . . . . . 33 secticide Residues . . . . . . . . . . 58
6.6. Phosphorothioate Esters of Hete- 17.2.1. Inactivation . . . . . . . . . . . . . . . 58
rocyclic Enols . . . . . . . . . . . . . 33 17.2.2. Activation . . . . . . . . . . . . . . . . 62

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 263
2 Insect Control
17.3. Toxicology of Insecticide Residues
18. References . . . . . . . . . . . . . . .
1. Insects as Enemies of Man [1]
Insects attack humans and domestic animals; act
as vectors for human, animal, and plant diseases;
destroy structures (man-made buildings, etc.);
and compete for supplies of food and fiber. The
ca. 106 insect species that have been described
represent about 70 % of all animal species. At
least 600 insect species are important pests in
the United States; the world total exceeds 10 000
species [1].
Worldwide crop losses caused by insect pests
have been estimated as wheat 5.0 %, rice 26.7 %,
maize 12.4 %, potatoes 6.5 %, sugar beet and
cane 16.5 %, vegetables 8.7 %, fruit 5.8 %, oil
crops 11.5 %, and fiber and rubber crops 14.2 %
[2]; see also → Crop Protection. Total produc-
tion loss is estimated at 13.8 % or $ 29.7×109 .
In the United States, total losses from insect at-
tacks calculated in 1988 dollars are estimated at
$ 14×109 [1].
Losses resulting from the attacks of in-
sect vectors of human and animal disease
are enormous. Some 60 species of Anopheles
mosquitoes act as vectors of the plasmodia that
cause human malaria; worldwide, they are re-
sponsible for as many as 200 million clinical
cases and a million deaths annually. At least
250 million people suffer from lymphatic filari-
asis caused by the nematode parasites Wuchere-
ria bancrofti and Brugia malayi. This disease
is transmitted largely by the common house
mosquito Culex fatigans that breeds in pol-
luted water. Trachoma, a viral disease spread
by the housefly Musca domestica is the most
common cause of human blindness, infecting
an estimated 40 million victims. American try-
panosomiasis ( Chagas’ disease) is caused by
Trypanosoma cruzi and transmitted to as many
as 40 million humans by about 95 species of
blood-sucking Triatominae bugs, with a variety
of animal reservoirs. The dozen or so species
of tsetse flies (Glossina spp.) are vectors of
trypanosomiasis in Central Africa. Blackflies
of the Simulium damnosum complex are vec-
tors of the filarian Onchocerca volvulus that in-
fects about a million people in the Volta River
Basin of Africa. The parasite may attack the
62
63
eye; about 70 000 of its victims are considered
blind. Other important human diseases trans-
mitted by insects and related arthropod vec-
tors include plague (causative organism Yersinia
pestis) transmitted by the asiatic rat flea Xenop-
sylla cheopis and other fleas, endemic typhus
(Rickettsia prowazeki) spread by the human
body louse Pediculus humanus humanus, and
scrub typhus (R. orientalis) spread by Trombic-
ula mites [3].
Arthropods are particularly adept at transmit-
ting viral disease (arborviruses); insects, mites,
and ticks are vectors for at least 450 viruses.
The most feared are yellow fever, dengue,
and dengue hemorrhagic fevers (transmitted
largely by the domestic mosquito Aedes ae-
gypti), as well as the human encephalitides
such as St. Louis, western and eastern equine,
and Venezuelan encephalitis for which the
mosquitoes Culex tarsalis and C. pipiens are im-
portant vectors [4].
1.1. Development of Insecticides [5–13]
Humans have been striving to ameliorate the
enormous depredations caused by insects since
the dawn of history. Chinese writings of the
Shang Dynasty (1523 – 1027 b.c.) describe orga-
nized efforts to use fire to destroy plagues of the
migratory locust Locusta migratoria manilensis.
However, the causal agents of malaria, Plasmod-
ium spp., and the role of Anopheles mosquitoes
in their transmission have been recognized only
since 1871.
Plant-derived insecticides (e.g., rotenone, ve-
ratridines, pyrethrins, and nicotine) have been
used for insect control since antiquity. Major
developments in the use of chemicals for plant
protection began with Paris green or copper ace-
toarsenite in 1865, used as a stomach poison
to control the Colorado potato beetle Leptino-
tarsa decemlineata. Lead arsenate, introduced in
1892, and calcium arsenate, introduced in 1907,
became the most important insecticides during
the early twentieth century; production reached
a maximum of 41.1×106 kg of lead arsenate in

1944 and 38.1×106 kg of calcium arsenate in
1942 [11]. Much of this was applied to cotton
to control the boll weevil Anthonomus grandis
and the pink bollworm Pectinophora gossyp-
iella. Arsenicals had the pronounced disadvan-
tages of high toxicity to humans and domes-
tic animals, considerable phytotoxicity, and ex-
treme environmental persistence [6], [11]. Cry-
olite (sodium fluoroaluminate) was introduced
as a stomach poison insecticide in 1928 be-
cause of widespread concern about toxic ar-
senical residues on fruit and vegetables. These
stomach poison insecticides were effective only
against chewing insects and had little contact ac-
tion.
Although potassium dinitro-o-cresylate, the
first synthetic organic insecticide, was intro-
duced in Germany in 1892 as a dormant spray for
tree fruit insects, synthetic organic insecticides
had little success until the patenting of 1,1,1-
trichloro-2,2-bis(4-chlorophenyl)ethane in 1939
[14]. This insecticide was enormously success-
ful in vector control of malaria, typhus, and
other human diseases during World War II, and
widespread publicity attending its civilian de-
but in 1945 led to the synthesis and evalua-
tion of hundreds of thousands of synthetic or-
ganic insecticides. With its efficient contact ac-
tion, DDT became the mainstay of vector con-
trol programs and rapidly replaced the arsenicals
in crop protection. It became the standard rem-
edy for control of the codling moth Laspeyre-
sia pomonella, which attacked deciduous fruit,
and for the pink bollworm P. gossypiella, which
attacked cotton. For the first time, the use of
DDT offered prospects for chemical control
of the two great defoliators of North Ameri-
can forests, the gypsy moth Porthetria dispar
and the spruce budworm Choristoneura fumifer-
ana. The name DDT became a household word
for controlling insect pests in flower and veg-
etable gardens and for mothproofing woolens.
The success of DDT residual spray operations,
which eliminated malaria in the Latina province
of Italy and on Sardinia within four years, in-
augurated the malaria eradication campaign of
the World Health Organization in 1955. The
plan was to eliminate malaria by spraying, at
a level of 1 g/m2 , all human dwellings where
malaria transmission occurred [15]. The enor-
mous success of DDT in insect control stim-
ulated the research and development of other
Insect Control 3
synthetic insecticides. Large-scale use of other
organochlorine compounds followed (e.g., lin-
dane; toxaphene; and the cyclodienes chlordane,
heptachlor, aldrin, dieldrin, and endrin).
1.2. Insect Resistance to Insecticides
[16–18]
The ability of insects to rapidly develop resis-
tance to the lethal action of insecticides has be-
come a major factor in the control of insect pests.
Insecticide resistance depends on random muta-
tion that establishes an R-allele (a gene coding
for resistance) in the natural population of the
insect species. The mutant individual possesses
biochemical, physiological, or even behavioral
characteristics that confer some degree of immu-
nity to the insecticide. Widespread application of
the insecticide propagates the R-allele through
preferential survival, and it is dispersed through-
out the population. As the R-allele becomes suf-
ficiently common, insecticide effectiveness is re-
duced. If the R-allele is partially dominant, com-
pletely resistant RR homozygotes are selected
rapidly. Thus insecticide resistance has been de-
scribed as accelerated microevolution. These R-
alleles are thought to be present in typical insect
populations at frequencies of 10−4 – 10−2 , with
RR homozygotes present at 10−8 – 10−4 . With
the resistance of Anopheles gambiae mosquito to
dieldrin used for malaria eradication, R-alleles
were present in African populations at frequen-
cies of 0.04 – 6.0 %, and a single spray treatment
produced a population of 86 % RS and RR geno-
types, with virtual immunity [16].
The first description of insecticide resistance
in 1914 was that of the San Jose scale Quadras-
pidiotus perniciosus to lime sulfur spray. In
1916, resistance to hydrogen cyanide fumigant
was reported in the California red scale Aoni-
diella aurantii and the black scale Saissetia
oleae which attacked citrus. The number of sci-
entifically validated examples of insecticide re-
sistance increased gradually, and by 1946 re-
sistance was known in eleven species including
the codling moth Laspeyresia pomonella and the
peach twig borer Anarsia lineatella to lead arse-
nate, the citrus thrips Scirtothrips citri to potas-
sium antimonyl tartrate (tartar emetic), the wal-
nut husk fly Rhagoletis completa to cryolite, and
the cattle ticks Boophilus microplus and B. de-
4 Insect Control
Table 1. Rates of development of insect resistance to insecticides

Number of resistant species

Insecticide 1938 1948 1954 1970 1975 1980 1984 1989 Average
doubling time,
years

DDT, methoxychlor 0 3 13 98 203 229 233 263 6.3

Lindane, cyclodienes 0 1 5 140 225 269 276 291 5.0
Organophosphorus compounds 0 0 3 54 147 200 212 260 4.0
Carbamates 0 0 0 3 36 51 64 85 2.5
Pyrethroids 0 0 0 3 6 22 32 48 1–2
Total 7 14 25 224 364 428 447 505 8.3

coloratus to sodium arsenite dip. However, little
scientific attention was given to insecticide resis-
tance until the introduction of DDT in 1945, with
resistant strains of the housefly Musca domes-
tica appearing almost simultaneously in Swe-
den and Denmark in 1946, of the mosquitoes
Culex tarsalis in Italy and Aedes sollicitans in
Florida in 1947, of the bedbug Cimex lectular-
ius in Hawaii in 1947, and of the human body
louse Pediculus humanus humanus in Korea and
Japan in 1951 [16].
Burgeoning use of insecticides after World
War II resulted in an exponential increase in
the number of scientifically documented exam-
ples of resistance, which reached 224 species
in 1970, 364 in 1975, 447 in 1984, and 505 in
1988 (Table 1). Most of the early examples of
resistance were found in insect vectors of hu-
man disease because of the widespread use of
DDT, lindane, and dieldrin in vector control pro-
grams. By 1970, resistance was documented in
119 pests of crops, forests, and stored products,
compared to 166 pests of human and animal
health; in 1980 the respective figures were 283
agricultural pests compared to 198 pests of hu-
mans and animals [18].
As each new class of insecticides was intro-
duced and used widely, the rate of development
of resistant species followed an almost iden-
tical pattern of exponential growth until high-
level resistance or legal restrictions slowed fur-
ther usage (Table 1). Doubling time for the de-
velopment of resistant species has decreased
steadily with the introduction of each new class
of insecticide (i.e., DDT and methoxychlor, lin-
dane and cyclodienes, organophosphates, carba-
mates, and pyrethroids). This accelerated devel-
opment of multiple resistance is a result of the
persistence of R genes in the species genome
and of their interaction through a variety of re-
sistance mechanisms that affect both the detoxi-
fication of and the target-site insensitivity to var-
ious insecticides.
1.2.1. Cross-Resistance and Multiple
Resistance
The exponential growth in numbers of insect
species resistant to insecticides does not ade-
quately describe the impact of resistance on in-
sect control problems. Two general types of re-
sistance are common: cross-resistance and mul-
tiple resistance.
Cross-resistance enables resistant species to
survive exposure to chemically related insecti-
cides, such as DDT and methoxychlor, lindane
and heptachlor, parathion and diazinon, carbaryl
and carbofuran, or permethrin and fenvalerate.
Cross-resistance results from a common detoxi-
fication mechanism or a change in susceptibility
to a common biochemical lesion.
Multiple resistance enables resistant species
to survive exposure to several classes of in-
secticide with differing detoxification systems
and biochemical modes of action. Multiple re-
sistance results from interactions among sev-
eral types of evolutionarily developed survival
mechanisms so that entire categories of insec-
ticides become useless. Examples include the
altered acetylcholinesterase target site that pro-
duces multiple resistance to organophosphorus
and carbamate insecticides, which is found in
the housefly, the green rice leafhopper Nephotet-
tix cincticeps, the cattle tick Boophilus mi-
croplus, and the red spider mite Tetranychus ur-
ticae. Another type of multiple resistance is the

“knockdown” mechanism of nerve axon insen-
sitivity that promotes resistance to both DDT
and the pyrethroids, and has been demonstrated
in the housefly; a variety of mosquitoes; bed-
bugs of Cimex spp.; lepidopteran pests includ-
ing Spodoptera, Plutella, and Heliothis; and the
cattle tick B. microplus. Multiple resistance is
now present in at least 44 families of insects in
10 orders [19]. An increasing number of ma-
jor insect pests are resistant to all the principal
classes of insecticides (Table 1). These include
the housefly; the malaria vectors Anopheles albi-
manus, A. sacharovi, and A. stephensi; the Ger-
man cockroach Blattella germanica; the Col-
orado potato beetle Leptinotarsa decemlineata;
the green peach aphid Myzus persicae; the pear
psylla Psylla pyricola; the diamondback moth
Plutella xylostella; the cotton bollworms Helio-
this spp. and Spodoptera spp.; the red flour bee-
tle Tribolium castaneum; and the granary weevil
Sitophilus granarius.
In summary, cross-resistance limits the
choice of available insecticides, whereas mul-
tiple resistance reflects the past history of insec-
ticide use and precludes a return to those used
previously. Both types of insecticide resistance
severely deplete insecticide resources.
1.2.2. Persistence of Genes for Insecticide
Resistance
Gene frequency of a specific resistance allele
may decrease upon removal of insecticide pres-
sure; however, a changed background of residual
inheritance persists and causes resistant strains
to regain high-level resistance as soon as the in-
secticide is reapplied. Thus genes for DDT and
cyclodiene resistance in Danish houseflies have
persisted for more than 20 years, and the insecti-
cides became ineffective within two months af-
ter reapplication. The citrus thrips Scirtothrips
citri has retained its resistance to tartar emetic
bait for more than 45 years and to DDT for more
than 35 years. No signs of regression have ap-
peared in the multiple resistance patterns of the
Colorado potato beetle and the cotton bollworm
Heliothis virescens over as much as 30 years.
The combined effects of cross- and multi-
ple resistances and the persistence of resistant
genes have had catastrophic effects on many in-
sect control programs. At least 51 of the 60 im-
Insect Control 5
portant Anopheles mosquito vectors of malaria
are resistant to one or more of the few residual
insecticides that can be used safely in human
habitations. This resistance has so increased the
cost and decreased the effectiveness of malaria
control that the World Health Organization aban-
doned its widely heralded program for global
eradication of malaria in 1976 [20]. Important
insect pests of man, such as the housefly and
the German cockroach, can scarcely be con-
trolled by insecticides. Practical chemical con-
trol of agricultural pests such as the Colorado
potato beetle, the diamondback, and the He-
liothis cotton bollworms is in jeopardy. Resis-
tance has severely affected the economy of in-
sect control in agriculture and public health by
necessitating replacement of the relatively in-
expensive organochlorine insecticides with the
increasingly expensive organophosphates, car-
bamates, and pyrethroids. Adverse effects on
human health followed the withdrawal of rel-
atively safe organochlorines, such as DDT and
lindane, and their replacement by highly toxic
organophosphates, such as parathion and methyl
parathion. Insecticide resistance is a major factor
in efforts to replace programs for the chemical
control of insects by more ecologically oriented
programs for insect pest management.
1.3. Resurgence and Secondary Pests
[20], [21]
The overuse and injudicious use of broad-
spectrum insecticides has upset the biological
balance between insect pests and their natural
enemies so that dramatic population increases
(resurgence) follow. This resurgence is usually
the result of repeated applications of insecticides
that select resistant races of the pest and co-
incides with depletion or virtual elimination of
their important natural enemies. The associated
development of secondary pests results from re-
sistance and resurgence of nontarget insect her-
bivores which then become so numerous that
they attain pest status. Destructive resurgence
has become a major factor in the development
of ecologically oriented programs for integrated
pest management.
Proliferation of the use of the broad-spectrum
insecticides DDT, parathion, carbaryl, and re-
lated compounds in 1946 was responsible for the
6 Insect Control
enormous outbreak of redspider mites (Tetrany-
chidae) in deciduous and citrus fruit orchards
and on cotton and other row crops. These mites
were rarely abundant enough before 1946 to re-
quire pesticide application. The sudden appear-
ance in 1947 of the new words miticide and aca-
ricide for special pesticides to control these mite
pests provided dramatic proof of the impact of
broad-spectrum insecticides on the balance bet-
ween phytophagous mites and their predators.
The imbalances created were all the more dra-
matic because of the very short generation time
of the mites and the rapidity with which they
developed pesticide resistance.
Today, the use of broad-spectrum pesticides
has almost inevitably been followed by pest re-
sistance, pest resurgence, and the outbreak of
secondary pests. These occur most dramatically
on crops that are treated heavily with pesticides,
such as cotton, rice, and tree fruit. Invariably,
ecological imbalance has been related to the
suppression of key natural enemies. Cotton cul-
ture worldwide has been almost destroyed by
the resurgence of Heliothis spp. and Spodoptera
spp. bollworms that became devastating sec-
ondary pests when their predators Chrysopa,
Orius, Geocoris, and Nabis spp. were destroyed.
Attempts to control cotton bollworms with more
frequent applications of insecticides led to the
destruction of populations of the Encarsia and
Eretmocerus parasites of the whitefly Bemisia
tabaci. This secondary pest, which transmits a
variety of plant diseases, has become a critical
factor in profitable cotton production in several
parts of the world.
The green rice leafhopper Nephotettix cincti-
ceps and related leafhoppers are vectors of rice
diseases and have emerged as important sec-
ondary pests of rice as a result of the very heavy
use of benzene hexachloride (BHC), parathion,
and other broad-spectrum insecticides to con-
trol the rice stem borers Chilo suppressalis
and Tryporyza incertulas. Insecticide applica-
tion destroyed predatory spiders that suppressed
leafhopper populations.
1.4. Insecticides and Environmental
Quality [22–27]
Insecticides have been highly controversial be-
cause of their widespread and persistent effects
on the environment. This concern results from
their intrinsic toxicity to animals and humans,
as well as their purposeful application to fields,
orchards, forests, ponds and streams, and human
habitations. Insecticide application is generally
inefficent: much less than 0.1 % of the amount
applied is required to kill pests; the remainder
(> 99.9 %) is distributed throughout the ecosys-
tem.
With rapidly increasing use, environmen-
tal contamination from insecticides has be-
come massive. In 1976, ca. 73.5×106 kg of ac-
tive ingredients of insecticides was applied to
30.335×106 ha of United States cropland; av-
erage contamination was about 2.4 kg/ha [28].
For cotton, the most heavily treated crop,
29.5×106 kg of active ingredient was applied to
2.835×106 ha, an average of 10.3 kg/ha. Early
insecticides such as the arsenicals and nicotine
were highly toxic, and lead arsenate persisted
and accumulated in orchard soil to levels that
killed young trees.
The organochlorine insecticides had the
greatest impact on environmental quality be-
cause of their relatively long persistence, their
very high lipid – water partition coefficients, and
their relative resistance to biological degrada-
tion. Estimated average soil half-lifes in years
are for DDT 2.8, dieldrin 2.5, endrin 2.2, lindane
1.2, chlordane 1.0, and heptachlor 0.8; in some
soil types, persistence is much longer. These in-
secticides have very low water solubility (in the
ppb-ppm range) and have octanol – water parti-
tion values of 103 – 106 .
They accumulate in living organisms
(bioconcentration), and trace levels in water or
food increase by a factor of 103 – 106 in fish or
animals at the top level of the food chain. During
their maximum use in 1950 – 1970, organochlo-
rine insecticides became ubiquitous micropol-
lutants of the environment (see Chap. 17). Their
perturbing effects on wildlife and humans have
been major factors in their stringent regulation
or prohibition in the United State, Canada, and
both Eastern and Western Europe: e.g., U.S.
government withdrawal of registration for DDT
(1973), aldrin and dieldrin (1974), heptachlor
and chlordane (1976), chlordecone (1980),
endrin (1981), mirex (1982), and toxaphene
(1983). The average levels of organochlorine
insecticide residues in water, soil, and human
tissue have begun to decline, and threatened

populations of birds of prey have been restored.
However, worldwide use of these compounds
between 1946 and 1976 was so extensive (DDT
> 2.3×106 t; toxaphene 4.5×105 t; cyclodienes
including aldrin, dieldrin, heptachlor, chlordane,
and endrin > 2.7×105 t) that traces will remain
in the environment for many years.
Environmental pollution resulting from
overuse of organochlorine insecticides focused
attention on the importance of biodegradability.
Organophosphate insecticides react in the pres-
ence of light, air, and moisture, and are readily
degraded in living organisms by oxidation and
hydrolysis. Therefore, the use of these insecti-
cides has not led to serious problems of bioac-
cumulation and food chain transfer. Their soil
persistence is low, with half-lifes ranging from
1 – 2 weeks for malathion, to 3 – 6 months for
parathion, diazinon, azinphosmethyl, and chlor-
pyrifos.
However, organophosphates are highly toxic
to almost all animals including humans because
they irreversibly inhibit the enzyme acetyl-
cholinesterase. Acetylcholinesterase is required
to break down the neurotransmitter acetyl-
choline after it has been released from the
nerve ending. Unlike most organochlorine insec-
ticides, many organophosphate insecticides are
almost as toxic upon dermal exposure as upon in-
gestion. Furthermore, chronic exposure can pro-
gressively deplete vital acetylcholinesterase lev-
els to a point where slight additional exposure
can result in severe illness or death from cholin-
ergic crises. Human intoxication resulting from
the use of the organophosphates is intensified
by insecticide resistance, which necessitates in-
creased dosage and more frequent application.
The incidence of human intoxication after use
of organophosphorus insecticides has become
acute; the World Health Organization has esti-
mated that 500 000 hospitalizations and as many
as 20 000 deaths occur annually worldwide [29].
Introduction of carbamate insecticides pro-
vided some amelioration of both environmental
persistence and acute toxic hazards. Carbamates
are generally biodegradable and do not bioaccu-
mulate. Soil persistence is low, with half-lifes
of 1 – 2 weeks for carbaryl and 1 – 4 months for
carbofuran and aldicarb. Many carbamates are
highly toxic to humans and animals, but acetyl-
cholinesterase inhibition is readily reversible so
that severe intoxication is unlikely.
Insect Control 7
Pyrethroids have the advantage of extraor-
dinary effectiveness against insect pests, so
that the required dosages are only about one-
tenth to one-twentieth those of organochlo-
rines, organophosphates, and carbamates. How-
ever, the newer pyrethroids are more persis-
tent and lipophilic; they are extremely toxic to
beneficial insects and fish. Furthermore, unlike
organophosphates and carbamates, which have
atropine as an antidote, no effective antidote ex-
ists for pyrethroid poisoning.
1.5. Integrated Pest Management [30–36]
The discovery of DDT and subsequent de-
velopment of hundreds of organochlorine,
organophosphate, carbamate, and pyrethroid in-
secticides produced a euphoria about chemical
pest control that led to the pest eradication phi-
losophy prevalent during the 1950s and 1960s.
Widespread application of cheap, effective in-
secticides was supposed to provide a rapid solu-
tion to every insect pest control problem. The
World Health Organization began the global
eradication of malaria through annual resid-
ual (long-lasting) house spraying directed at
Anopheles mosquito vectors. The United States
undertook the insecticide eradication of pests
such as the gypsy moth Porthetria dispar, the
Japanese beetle Popillia japonica, the yellow
fever mosquito Aedes aegypti, and the red im-
ported fire ant Solenopsis invicta. As each erad-
ication program faltered because of insecticide
resistance, destruction of wildlife, and wide-
spread environmental pollution, the initial reac-
tion was to develop another insecticide of differ-
ent chemistry.
The publication of Rachel Carson’s Silent
Spring in 1962 [37] inaugurated public chal-
lenge to the notion that all insect control prob-
lems could be solved solely with insecticides.
Surveys demonstrated the continuing, unabated
damage resulting from insect pest attack despite
a greater than 10-fold increase in insecticide use
and a 20-fold proliferation in chemicals avail-
able for insect control (Table 2). Unbridled use
of broad-spectrum insecticides over the past 40
years had obviously not solved the manifold
problems of insect pest control. The benefits to
cost ratio of pest control activities to the farmer
and to public health have steadily decreased and
8 Insect Control
Table 2. Estimates of U.S. crop losses from insect attacks over 60 years ∗
Crop
1900 – 1904
Alfalfa 10
Apple 20
Cabbage 10
Corn 8 12
Cotton 10
Potato 10
∗ USDA data [20].
the nontarget effects of pesticides on environ-
mental quality have continued to multiply.
Pest control must be directed away from ex-
clusive reliance on toxic chemicals and toward
the optimization of pest control tactics in an eco-
nomically and ecologically sound manner, i.e.,
integrated pest management (IPM). A funda-
mental premise of IPM is to relegate pesticides
to emergency use and spray only when neces-
sary.
The primary goals of IPM are [33]
1) to determine how the life system of the pest
must be modified to reduce its numbers to
tolerable levels (i.e., below the economic in-
jury level);
2) to apply biological knowledge and current
technology to achieve the desired modifica-
tions (i.e., applied ecology); and
3) to devise procedures for pest control compat-
ible with economic and environmental con-
trol aspects (i.e., economic and social accep-
tance).
Successful IPM programs integrate a va-
riety of suppressive measures including both
density-dependent factors (natural enemy preda-
tors, parasitoids, and pathogens) and density-
independent factors (host plant resistance and
cultural practices that deter pest attack). The ba-
sics of IPM practice are
1) to establish sound economic injury levels
in which the cost of treatment balances the
value of pest damage;
2) to monitor regularly to determine pest popu-
lation levels; and
3) to implement control practices promptly
when the economic injury level is ap-
proached.
Loss of total crop, %
1910 – 1935 1942 – 1951 1951 – 1960
5 9 15
10 14 13
20 8 17
4 12
15 15 19
22 16 4
Insecticides are important components of
IPM programs; indeed, IPM is a dynamic strat-
egy for the judicious and selective use of in-
secticides. Insecticide management is concerned
with the safe, efficient, and economical handling
of insecticides during manufacture, utilization,
and disposal [38–40]. It involves selection of the
proper insecticide for the IPM program; selec-
tion of the mode, timing, and dosage of applica-
tion; and selection with regard to resistance and
resurgence. The nature of insecticide residues in
food and in the environment, and their impact
on humans, must be considered.
2. Inorganic Insecticides
2.1. Arsenicals [6], [11]
The first synthetic insecticides were arsenical
salts used as stomach poison insecticides to con-
trol chewing insects such as the Colorado potato
beetle Leptinotarsa decemlineata and the cotton
boll weevil Anthonomus grandis.
Paris green, copper acetoarsenite
[12002-03-8], [Cu (CH3 COO)] 2 · 3 Cu (AsO2 )2 ,
is a bright-green pigment, used originally as a
wall and shutter paint. Its insecticidal proper-
ties were discovered around 1865. Paris green
contains 44.3 % arsenic, with 2 – 3 % of it water-
soluble; the LD50 (rat, oral) is 100 mg/kg.
Lead arsenate was developed in 1892 to
combat the gypsy moth Porthetria dispar. It is a
safer arsenical for plant use because of its lower
content of water-soluble arsenic. Acid lead arse-
nate [7784-40-9], PbHAsO4 , contains 21.6 % ar-
senic with 0.25 % of it water-soluble. The LD50
(rat, oral) is 1050 mg/kg.

Basic lead arsenate [62648-18-4],
Pb4 (PbOH) (AsO4 )3 , with 14 % arsenic, was in-
troduced as a safer form to use on delicate fo-
liage in foggy or humid regions where acid lead
arsenate is hydrolyzed to liberate arsenic acid.
Calcium arsenate [7778-44-1], Ca3 (AsO4 )2 ,
was introduced in 1920 to control the
boll weevil on cotton. It contains 37.6 %
arsenic and has an LD50 (rat, oral) of
298 mg/kg. Basic calcium arsenate [74966-85-
1], [Ca3 (AsO4 )2 ] 3 · Ca(OH)2 , contains 25 %
arsenic with ca. 0.4 – 0.5 % of it water-soluble.
It is a safer form that does not hydrolyze readily.
Arsenic trioxide [1327-53-3], As2 O3 , with
75.7 % arsenic has an LD50 (rat, oral)
of 14.1 mg/kg. Sodium arsenite [7784-46-5],
NaAsO2 , with 55.7 % arsenic has an LD50 (rat,
oral) of 41 mg/kg. These two compounds have
been used as poisons in bran bait to kill migrat-
ing grasshoppers and in water dips to control
cattle ticks.
Mode of Action. The As3+ ion of arsenicals
specifically inhibits the sulfhydryl-containing
enzymes involved in cell respiration, e.g., pyru-
vate oxidase (E.C. 1.2.3.3.).
2.2. Fluorides [6]
Cryolite [15096-52-3], sodium fluoroalumi-
nate, Na3 AlF6 , 54.2 % fluorine, water-soluble
fluorine 0.035 – 0.061 %, was developed as a
safer stomach poison insecticide than the arseni-
cals, especially because of concern about per-
sistent lead arsenate residues on edible produce
such as apples. The LD50 (rat, oral) of cryolite
has been given as 200 – 13 500 mg/kg.
Sodium fluorosilicate [16893-85-9],
Na2 SiF6 , 60.6 % fluorine, water-soluble fluo-
rine 0.65 %, LD50 (rat, oral) 125 mg/kg, is used
for mothproofing.
Barium fluorosilicate [17125-80-3],
BaSiF6 , 40.8 % fluorine, water-soluble fluo-
rine 0.026 %, LD50 (rat, oral) 175 mg/kg, has
been used as a stomach poison for plant-feeding
insects.
Insect Control 9
Sodium fluoride [7681-49-4], NaF, 45.2 %
fluorine, water-soluble fluorine 4.3 %, LD50 (rat,
oral) 180 mg/kg, is used as a stomach poison
for cockroaches, for chewing lice, and in poison
baits.
Mode of Action. Inorganic fluorides com-
plex magnesium ions as magnesium fluorophos-
phate. They inhibit enzymes such as enolase
(E.C. 4.2.1.1) that require Mg2+ as a prosthetic
group. Thus, fluoride poisoning prevents phos-
phate transfer in oxidative metabolism.
3. Natural Product Insecticides
In this and the following chapters two values are
often cited for the LD50 . The first value refers to
male animals and the second to female animals.
3.1. Nicotinoids [8], [9]
Nicotine [54-11-5], l-1-methyl-2-(3 -
pyridyl)-pyrrolidine, bp 247 C, d 20
◦
4 1.009, one
of the earliest insecticides, was recommended in
1793 as a tea for the destruction of aphids. Nico-
tine is found in the leaves of Nicotiana tabacum
and N. rustica in amounts of 2 – 14 %, as well as
in Duboisia hopwoodii and Aesclepias syriaca.
It occurs as the principal alkaloid along with
small amounts of 12 other alkaloids, includ-
ing the water-miscible insecticidal compounds
nornicotine and anabasine. Nicotine is readily
volatile (vapor pressure 5.5 Pa at 25 ◦ C) and
dibasic (K b1 = 1×10−6 , K b2 = 1×10−11 ), easily
forming salts with acids and many metals. Its
LD50 (rat, oral) is 30 – 55 mg/kg.
Nicotine sulfate [65-30-5], (C10 H14 N2 )2 ·
H2 SO4 , has been used as an aphicide on fruit,
vegetables, and ornamentals because it is less
volatile than nicotine and safer to handle.
Nornicotine [494-97-3], 2-(3 -pyridyl)pyrro-
lidine, bp 270 ◦ C, d 20
4 1.07, occurs naturally as
both the d and the l isomer, the former in D.
hopwoodii and the latter in Nicotiana.
10 Insect Control
Anabasine [494-52-0], l-2-(3 -pyridyl)piperi-
dine, bp 281 ◦ C, d 20
4 1.047, is the principal al-
kaloid of Anabasis aphylla, present at 1 – 2 %
in the shoots, and is also found to about 1 % in
N. glauca.
Imidacloprid [105827–78–9], 1-[(6-
chloro-3-pyridinyl)-methyl]-N-nitro-2-
imidazolidinime, vapor pressure 0.2 µPa at
20 ◦ C, solubility in water 0.51 g/L, is a syn-
thetic nicotinoid with contact and systhemic
activity to aphids, leafhoppers, and white flies.
Its LD50 values are 424 – 495 (rat, oral) and >
5000 (rat, dermal).
Mode of Action. The nicotine alkaloids are
agonists of acetylcholine and affect the postsy-
naptic receptors of the insect nervous system,
facilitating synaptic transmission at low con-
centration and blocking transmission at higher
concentration. The nicotine molecule has di-
mensions similar to acetylcholine. Ionization of
the N-methylpyrrole group affects penetration
through the ionic barrier of the nerve sheath
to the site of action; the quaternary form has
an increased affinity for the negatively charged
acetylcholine receptor.
3.2. Rotenoids [8], [41]
Rotenoids are found in at least 68 plant species,
including Tephrosia, Derris, Lonchocarpus,
Milletia, and Mundulea. The important eco-
nomic species are, D. elliptica and D. malac-
censis from Malaysia and Indonesia, and cube
or timbo (L. utilis and L. urucu) from South
America. Insecticidal compounds are found in
the roots and seeds of these plants.
Rotenone [83-79-4], mp 165 ◦ C (a dimor-
phic form, mp 185 ◦ C) by far the most impor-
tant rotenoid, is five to ten times more effec-
tive than elliptone [478-10-4], mp 159 ◦ C (fu-
ran ring at E); sumatrol [82-10-0], 15-hy-
droxyrotenone, mp 188 ◦ C; malaccol [478-
07-9], 15-hydroxyelliptone, mp 244 ◦ C; l-α-
toxicarol [82-09-7], 15-hydroxyrotenone with
a 2,2-dimethyl-2H-pyranyl ring in place of ring
E, mp 101 ◦ C; and deguelin [522-17-8], with
a hydrogen on C-15 in place of the hydroxyl
group of toxicarol, mp 165 – 171 ◦ C. The con-
tent of these components in commercial sources
is variable: roots of D. elliptica average 5 – 9 %
rotenone and up to 31 % ether extractives (or-
ganic constituents extracted by diethyl ether),
whereas L. utilis averages 8 – 11 % rotenone and
25 % extractives. When exposed to light and air,
rotenone undergoes hydroxylation at C-7, fol-
lowed by dehydration to form dehydrorotenone
(C=C between rings B and C). These reactions
inactivate rotenone after 5 – 10 d of exposure to
normal sunlight. Rotenone is soluble in water
only to ca. 0.2 mg/L but is readily soluble in
petroleum solvents. The LD50 (rat, oral) is ca.
130 mg/kg.
Rotenone-containing insecticides have been
used as dusts for ground roots, dispersible pow-
ders, and emulsive extracts for application to ed-
ible products shortly before harvest, as well as
for the control of animal ectoparasites and cattle
grubs.
Mode of Action. Rotenone is a specific in-
hibitor of the respiratory enzyme NADH dehy-
drogenase (E.C. 1.6.99.3) and interferes with
the electron transport necessary for oxidation
of NADH (reduced nicotinamide adenine din-
ucleotide) to NAD by cytochrome b. Poison-
ing, therefore, inhibits the mitochondrial NAD-
dependent oxidation of the citric acid cycle in-
termediates.

3.3. Veratrine Alkaloids [41]
Sabadilla, obtained from the ground seeds
of Schoenocaulon officinale (Liliaceae) from
South and Central America, has been used in na-
tive louse powders for centuries. More recently,
it has been used in sugar bait sprays for thrips
control. The toxic principles of sabadilla are the
veratrine alkaloids which are present in the seeds
to 2 – 4 % and include ca. 13 % cevadine and
10 % veratridine, as well as lesser amounts of
cevadilline and sabadine.
Sabadilla preparations are activated by heat-
ing to 75 – 80 ◦ C for 4 h, moistening with 10 %
sodium carbonate solution, or milling with hy-
drated lime. Veratrine alkaloids are destroyed
rapidly by exposure to light; sabadilla, which
is both a contact and a stomach poison insec-
ticide, is safe for use on food crops within a
few days of harvest. The alkaloids most re-
sponsible for its insecticidal action are cevadine
[62-59-9], C32 H49 O9 N, mp 213 ◦ C, the angelic
acid ester of cevine; and veratridine [71-62-5],
C36 H51 O11 N, mp 180 ◦ C, the veratric acid es-
ter of cevine at C-3. These alkaloids are highly
poisonous; the intraperitoneal LD50 of cevadine
in mouse is 3.5 mg/kg, and it may cause severe
irritation of mucous membranes and the respira-
tory tract. Ground sabadilla seeds have an LD50
(rat, oral) of 5000 mg/kg.
3.4. Ryanodine [41]
Ryania, the ground roots and stems of Rya-
nia speciosa (Flacourtiaceae) native to South
America, contains 0.16 – 0.25 % of insecticidal
components of which ryanodine is the most im-
portant. Ryanodine [15662-33-6], C25 H33 O9 N,
mp 219 – 220 ◦ C is both a contact and a stomach
poison insecticide; it is more stable to the ac-
tion of light and air than rotenone or pyrethrum.
Insect Control 11
It is used as a residual insecticide for codling
moth control, particularly in home plantings.
Ryania is applied either as a wettable powder or
as a methanolic extract. The LD50 (rat, oral) is
750 mg/kg. Ryania uncouples the adenosine 5 -
triphosphate – adenosine 5 -diphosphate (ATP –
ADP) actomyosin cycle in the contraction of
striated muscle.
3.5. Pyrethroid Insecticides [41–43]
The insecticidal properties of pyrethrum from
the ground flowers of Chrysanthemum cinerari-
aefolium and C. coccineum (Compositae) were
recognized in the Transcaucasus region of Asia
around 1800. Manufacture of flea and louse
powders began about 1828, and when the na-
ture and source of the material was published
in 1851, it became used worldwide. Pyrethrum
flowers were the most important export of Dal-
matia until World War I, when Japan became
the leading producer. Pyrethrum production be-
gan in Kenya about 1932, and it was the princi-
pal supplier by 1940. World production of dried
pyrethrum flowers was about 23×103 t in 1975.
Kenya flowers contain ca. 1.3 % of the toxic con-
stituents, pyrethrins and cinerins; Japanese flow-
ers, ca. 1 %; and Dalmatian flowers, ca. 0.7 %.
Pyrethrum concentrates are prepared by extract-
ing the flowers with petroleum ether, ethylene
dichloride, methanol, acetone, or acetic acid.
The extracts are purified from waxes by reex-
traction with nitromethane or methanol, and ab-
sorption on activated charcoal, to produce a 25 %
pyrethrum concentrate containing about 10 %
pyrethrin I, 9 % pyrethrin II, 2 % cinerin I, 3 %
cinerin II, 1 % jasmolin I, and 1.1 % jasmolin II
(Fig. 1).
Pyrethrum is an especially valuable in-
secticide because of its rapid paralysis
(“ knockdown ”) of flying insects and because
its toxic principles are inactivated on exposure
to light. The main use of pyrethrum insecticide
is in pressurized aerosol “ bombs ” that contain
12 Insect Control
Figure 1. Structures of pyrethroids from Chrysanthemum cinerariaefolium
0.04 – 0.25 % active ingredient, together with
five to ten times this amount of piperonyl butox-
ide or other synergists to retard detoxification
in insects. Deodorized oil or water-based sprays
are also used. The principal areas of use for
pyrethrum products are
1) household insecticides for the control of
mosquitoes, flies, cockroaches, bedbugs,
ants, silverfish, and fabric pests;
2) livestock or cattle sprays to protect animals
from annoyance by biting and nonbiting flies;
3) protective sprays for grain in storage; and
4) limited use on fruit and vegetables to con-
trol insect pests prior to harvest when toxic
residues are very objectionable.
Toxic Constituents. Complete identifica-
tion of the active principles of pyrethrum re-
quired approximately 50 years, dating from the
work of Staudinger and Ruszicka in 1924.
There are six esters of the two cyclopropanecar-
boxylic acids, chrysanthemic acid and pyrethric
acid (Fig. 1), and three cyclic keto alcohols,
pyrethrolone, cinerolone, and jasmolone. The
esters are optically active with absolute config-
urations of 1R, 3R, and 4S; the double bond in
the acid side chain is cis and in the alcohols,
trans.
3.5.1. Synthetic Pyrethroids [44–46]
Elucidation of the chemical structure of the natu-
rally occurring pyrethroids stimulated the search
for effective synthetic derivatives of these costly
natural products. Since the early 1940s, struc-
tural optimization of pyrethroid molecules has
led to a series of broad-spectrum insecticides
with greatly enhanced activity and, in many
cases, extended residual action. Pyrethroids
have become one of the most important classes
of insecticides.
3.5.1.1. Chrysanthemate Esters
Allethrin [584-79-2], 2-methyl-4-oxo-3-
(2-propenyl)-2-cyclopenten-1-yl 2,2-dimethyl-
3-(2-methyl-1-propenyl)cyclopropanecarbox-
ylate, the first successful synthetic pyrethroid,
was developed in 1949 and is the allyl homo-
logue of cinerin I. The product is a brownish liq-
uid, d 20
4 1.005 – 1.015, containing 75 – 95 % of a
mixture of eight optical and geometric isomers
[70 % (±)-trans and 30 % (±)-cis acids] esteri-
fied with (±)-methylcyclopentenolone alcohol.
Allethrin has an LD50 (rat, oral) of 920 mg/kg;
it is as effective as natural pyrethrum against
flies and mosquitoes but has a narrow spec-
trum of activity. Bioallethrin is the (+)-trans-
chrysanthemate ester of the same alcohol with
enhanced “ knockdown ” activity.

Benzyl Chrysanthemates. Esterification of
(±)-cis,trans-chrysanthemic acid with substi-
tuted benzyl alcohols produced barthrin [70-
43-9], the 6-chloro-1,3-benzodioxol-5-yl es-
ter, and dimethrin [70-38-2], the 2,4-di-
methylbenzyl ester. These compounds have
very low mammalian toxicity (LD50 , rat, oral
> 20 g/kg) but a limited spectrum of insectici-
dal activity. Dimethrin has been used as a safe
mosquito larvicide for drinking water in cisterns.
Resmethrin [10453-86-8], 5-benzyl-3-
furanyl (±)-cis,trans-chrysanthemate, mp 43 –
48 ◦ C, bp 169 – 172 ◦ C (1.3 Pa), is a white waxy
solid. The compound contains 20 – 30 % (±)-
cis and 70 – 80 % (±)-trans isomers and has
an LD50 (rat, oral) of 2 000 mg/kg. The ben-
zylfuranyl moiety is readily decomposed by
light, and resmethrin is an outstanding house-
hold insecticide, 20 – 50 times more effective
than pyrethrum and of very short persistence.
Bioresmethrin [28434-01-7] is the cor-
responding (±)-trans-(1R,3R) ester, which is
about twice as toxic to insects; it has an LD50
(rat, oral) > 8 000 mg/kg.
Tetramethrin [7696-12-0], (1,3,4,5,6,7-
hexahydro-1,3-dioxo-2H-isoindol-2-yl)methyl
(±)-cis,trans-chrysanthemate, mp 65 – 80 ◦ C,
vapor pressure 6.2 µPa at 20 ◦ C, has an LD50
(rat, oral) of 1 000 mg/kg and produces rapid
“ knockdown ”. It is used in veterinary hygiene.
Insect Control 13
Phenothrin [26002-80-2], (3-phenoxyphe-
nyl)methyl (±)-cis,trans-chrysanthemate, a col-
orless liquid, d 25
4 1.058, is soluble in water to
0.2 mg/L. The LD50 (rat, oral) is 5000 mg/kg.
3.5.1.2. Chrysanthemate Analogues
More recent development in pyrethroid chem-
istry has involved substantial changes in the
structure of the acid moiety and esterification
with a variety of optimized alcohol moieties.
Permethrin [52645-53-1], (3-phenoxy-
phenyl)methyl 3-(2,2-dichloroethenyl)-2,2-di-
methyl cyclopropanecarboxylate is a mixture
of 70 % (±)-trans and 30 % (±)-cis configura-
tions, mp 35 ◦ C, water solubility 0.2 mg/L, va-
por pressure 45 µPa at 25 ◦ C. The LD50 (rat,
oral) is 1500, 2000 mg/kg. Substitution of chlo-
rine atoms for the dimethyvinyl group greatly
reduces photolability and increases persistence;
thus permethrin is a widely used broad-spectrum
insecticide for cotton and other field crops.
Cypermethrin [52315-07-8], α-
cyano(3-phenoxyphenyl)methyl 3-(2,2-di-
chloroethenyl)-2,2-dimethylcyclopropanecar-
boxylate, mp 60 – 80 ◦ C, water solubility
0.01 mg/L, is a mixture of 70 % (±)-trans and
30 % (±)-cis configurations; its LD50 (rat, oral)
14 Insect Control

is 310 mg/kg. Cypermethrin is widely used on 1250, 1070 mg/kg (rat, oral) and > 2000 mg/kg
agricultural crops and in cockroach control. (rabbit, dermal). Tralomethrin is used as a house-
hold, stored grain, foliar, and wood protectant
insecticide.

Cyhalothrin [68085-85-8], (RS)-α-cyano-

Cyfluthrin [68359–37–5], (R,S)-α-
cyano-(4-fluoro-3-phenoxyphenyl)-methyl-
(1R,S)-cis,trans-3-(2,2)-dichloroethenyl)-2,2–
dimethylcyclopropanecarboxylate, mp 60 ◦ C,
solubility in water 2 µg/L. Cyfluthrin is a mix-
ture of eight isomers; its LD50 values (rat) are
500, 800 (oral), and 600 (dermal). Cyfluthrin is
a wide-spectrum pyrethroid, especially effective
against greenhouse pests.
(3-phenoxyphenyl)methyl (±)-cis-(1R,S)-3-
(2-chloro-3,3,3-trifluoro-1-propenyl)-2,2-di-
methylcyclopropanecarboxylate, mp 49 ◦ C, va-
por pressure 0.2 µPa at 25 ◦ C soluble in water
(5 µg/L); its LD50 (rat, oral) is 79, 56 mg/kg.
Cyhalothrin is a persistent pyrethroid used on
cotton and other field crops, which has acaricidal
activity.

Kadethrin [58769-20-3], 5-(benzyl-

3-furanylmethyl) 3-[(dihydro-2-oxo-3(2H)-
thienylidene)-methyl]-cis-(1R,3S)-2,2-di-
methylcyclopropanecarboxylate, is a pyrethroid
of very rapid “ knockdown ”. The LD50 (rat,
oral) is 140, 1300 mg/kg.
Deltamethrin [52918-63-5], (S)-α-cyano-
3-phenoxybenzyl(+)-cis-(1R,3R)-2,2-dimethyl-
3-(2,2-dibromovinyl)cyclopropanecarboxylate,
mp 98 – 101 ◦ C, vapor pressure 2 µPa at 25 ◦ C,
the most active and persistent of all pyrethroid
insecticides, is effective against field crop in-
sects at doses of 100 g/ha. However, it is highly
toxic to mammals; its LD50 (rat, oral) is 25,
60 mg/kg.
Bioethanomethrin [22431-62-5],
5-(benzyl-3-furanylmethyl)3-(cyclopentyl-
idenemethyl)-(+)-trans-(1R,3S)-2,2-dimethyl-
cyclopropanecarboxylate. The LD50 (rat, oral)
is 100 mg/kg.

Tralomethrin [66841-25-6], (S)-α-

cyano-(3-phenoxyphenyl)-methyl-(1R,3S)-2,2-
diemthyl-3-(RS)-(1,2,2,2-tetrabromoethyl)cy-
clopropanecarboxylate, d 1.70, vapor pressure
17 mPa at 25 ◦ C, the tetrabromo analogue of
deltamethrin, is a mixture of two isomers. Sol-
ubility in water 70 mg/L. Its LD50 values are

Fenpropathrin [39515-41-8], α-cyano(3-
phenoxyphenyl)methyl 2,2,3,3-tetramethylcy-
clopropanecarboxylate, a white crystalline solid,
has an LD50 (rat, oral) of 25 mg/kg. It is a broad-
spectrum soil insecticide and acaricide.
Tefluthrin [79538-32-2], 2,3,5,6-tetra-
fluoro-4-methylphenylmethyl (2)-(1RS,3RS)-
3-(2-chloro-3,3,3-trifluoro-1-propenyl)-2,2-di-
methylcyclopropanecarboxylate, is a white
solid, mp 44 – 60 ◦ C, vapor pressure 1.0 µPa
at 25 ◦ C, water solubility 2 µg/L. The LD50
(rat, oral) is 23, 35 mg/kg. It is a persistent soil
insecticide.
3.5.1.3. Phenylacetate Esters
The phenylacetate esters of pyrethroids are
structurally optimized in both the acid and the
alcohol moieties.
Fenvalerate [51630-58-1], α-cyano(3-
phenoxyphenyl)methyl (±)-(2RS)-4-chloro-α-
(1-methylethyl)benzeneacetate, a clear liquid,
d 23 ◦
4 1.17, vapor pressure 1.4 µPa at 25 C, is
a widely used broad-spectrum insecticide. The
LD50 (rat, oral) is 450 mg/kg.
Esfenvalerate [66230-04-4] is the (+)-2-(S)
isomer of fenvalerate, mp 59 ◦ C.
Insect Control 15
Flucythrinate [70124-77-5], α-cyano-
(3-phenoxyphenyl)methyl(±)-(2S)-4-(difluoro-
methoxy)-α-(1-methylethyl)benzeneacetate, is
a broad-spectrum insecticide. The LD50 (rat,
oral) is 53, 87 mg/kg.
Fluvalinate [69409-94-5], α-cyano(3-
phenoxyphenyl)methyl N-[2-chloro-4-(tri-
fluoromethyl)phenyl]-dl-valine, vapor pressure
13 µPa at 25 ◦ C, solubility in water 2 µg/L, is a
yellowish liquid, and a general-purpose insecti-
cide.
3.5.2. Mode of Action [47]
Pyrethroids, particularly the optimized synthetic
derivatives, have very high lipid – water partition
coefficients and penetrate rapidly into the insect
nervous system. Their action is characterized by
excitation, uncoordination, and paralysis, pro-
ducing the characteristic “ knockdown effect ”.
The natural pyrethrins, resmethrin, perme-
thrin, and DDT, make the Na+ channel close
very slowly after each action potential, produc-
ing a prolonged afterpotential. Pyrethroids with
the α-cyano group, deltamethrin and cyperme-
thrin, delay closing of the Na+ channel so that
spontaneous generation of the action potential
is suppressed. These actions are a function of
the Ca2+ -dependent protein kinases associated
with the Na+ channels that phosphorylate neural
protein.
Extensive structure – activity studies of the
pyrethroids have defined the following essential
features for high insecticidal activity:
1) Acid Moiety. The 2,2-methyl groups of cy-
clopropanecarboxylic acid or its stereochem-
ical equivalent (the methyl groups of the α-
(1-methylethyl)benzeneacetic acid) are es-
16 Insect Control

sential. The C-3 substituent in cyclopropane
can be isobutenyl, 2,2-dihalovinyl, or tri-
fluoromethylchlorovinyl. The configuration
at C-1 and C-3 of the cyclopropane must be
either 1R, 3S or 1R, 3R.
2) Alcohol Moiety. The alcohol moiety should
contain a cyclopentenolone ring or stereo-
chemically similar furan or benzyl groups
and an unsaturated side chain or aromatic
group. The α-cyano group apparently acts as
a dimensional spacer to promote maximum
complementarity with the axonal receptor.
C=C double bonds of pyrethrins, the cinerin side
chains, and methyl groups of the isobutenyl moi-
ety. Synergists of methylenedioxyphenyl(1,3-
benzodioxole) react with the iron atom of cy-
tochrome P450 and prevent detoxification. Such
synergists are usually included with natural
pyrethrins at ratios of (10 – 20) : 1 in fly sprays
and aerosols. They may increase the activity of
pyrethrins as much as 30-fold. Extensive struc-
tural optimization of synthetic pyrethroids has
eliminated most of the detoxification sites. Syn-
ergists are generally ineffective with the persis-
tent pyrethroids.

3.5.3. Environmental Toxicology Piperonyl butoxide [51-03-6], 3-{[2-(2-

butoxyethoxy)ethoxy]methyl}-6-propyl-1,3-
The synthetic pyrethroids are insecticidally ac- benzodioxole, bp 180 ◦ C (0.13 kPa), d 25
4 1.04 –
tive at dosages of 20 – 100 g/ha; i.e., they are 1.07, LD50 (rat, oral) 7 500 – 6 150 mg/kg, is
about ten times more effective than other in- the synergist most commonly used with natural
secticides. Thus, their use lessens the environ- pyrethrins in pressurized aerosol sprays.
mental burden on nontarget organisms consider-
ably. However, pyrethroids are extremely toxic
to fish, with LC50 values in the 0.001-ppm range.
Persistence varies from 1 – 2 d for the natural
pyrethrins and resmethrin to 1 – 3 months for
deltamethrin and tefluthrin.
Acute toxicity to humans and higher ani- Other synergists with similar properties in-
mals ranges from very safe compounds such as clude the following:
resmethrin and permethrin to highly toxic com-
pounds such as deltamethrin, flucythrinate, and
tefluthrin. Many of the more active compounds Sulfoxide [120-62-7], 5-(2-octylsulfin-
are highly irritating to the skin and nasal mem- yl)propyl-1,3-benzodioxole, has an LD50 (rat,
branes. A major drawback of the highly toxic oral) of 2 000 mg/kg.
compounds is lack of an effective antidote for
acute poisoning.

3.5.4. Pyrethroid Synergists [5], [48]

Synergists are nontoxic compounds used in

conjunction with insecticides to produce con-
siderably more than additive toxicity. Strong Sesamex [51-14-9], 5-{1-[2-(2-
synergistic interactions have been found with ethoxyethoxy)-ethoxy]-ethoxy}-1,3-benzodi-
carbamates and with rotenone or ryanodine, oxole, has an LD50 (rat, oral) of 2 000 mg/kg.
but the use of synergists has practical impor-
tance only for pyrethroid insecticides. Natural
pyrethroids are rapidly detoxified in the insect
by mixed-function oxidases; thus, insects par-
alyzed with pyrethroids often recover. These
enzymes act through cytochrome P450 to pro-
duce hydroxylating (- OH) groups that attack the

MGK-264 [113-48-4], N-(2-ethylhexyl)-
[2.2.1]-5-heptene-2,3-dicarboximide, bp
158 ◦ C (0.26 kPa), d 18
4 1.05, LD50 (rat, oral)
2800 mg/kg, is an effective pyrethroid and car-
bamate synergist of totally different chemical
structure.
4. Dinitrophenols [8]
The first synthetic organic insecticide was the
potassium salt of 4,6-dinitro-2-methylphenol
patented as a dormant spray for orchard pests
in 1892. Dinitrophenols have a wide range of
biocidal activity and have been used as insecti-
cides, acaricides, herbicides, and fungicides.
Dinitro-o-cresol (DNOC) [534-52-1], 4,6-
dinitro-2-methylphenol, is a yellow crystalline
compound, mp 87 ◦ C, vapor pressure 6.8 mPa at
25 ◦ C, water solubility 0.013 wt %. It is a pseu-
doacid (K a 2.5×10−6 ) and readily forms water-
soluble ammonium, potassium or sodium salts.
The LD50 values (rat) are 30 mg/kg (oral) and
200 mg/kg (dermal).
Dinocap [39300-45-3], 2-(1-methylheptyl)-
4,6-dinitrophenyl 2-butenoate, bp 138 – 140 ◦ C
(6.7 Pa), LD50 (rat) 980, 1 190 mg/kg (oral) and
> 4 700 mg/kg (dermal), is an acaricide and
fungicide.
Insect Control 17
Dinoseb [88-85-7], 4,6-dinitro-2-(1-
methylpropyl)phenol, mp 42 ◦ C, has LD50
values (rat) of 136, 225 mg/kg (oral) and
> 1 000 mg/kg (dermal). Triethanolamine salts
have been used as a dormant spray for deciduous
fruit pests.
Binapacryl [485-31-4], the 3-methyl-2-
butenoate ester of dinoseb, mp 70 ◦ C, LD50 val-
ues (rat) 63, 58 mg/kg (oral) and 810, 720 mg/kg
(dermal), has been used as an acaricide for fo-
liage.
Mode of Action. The dinitrophenol insecti-
cides uncouple oxidative phosphorylation and
are highly toxic to plants, insects, higher ani-
mals, and humans. Poisoning increases (three- to
tenfold) cellular respiration. Death results from
metabolic exhaustion because of an inability to
utilize the energy provided by respiration and
glycolysis. This general biocidal activity has
rendered them obsolete as insecticides.
5. Organochlorine Insecticides [49],
[50]
5.1. DDT and Derivatives [5], [9], [51]
Although DDT was synthesized by O. Zeidler
in 1874, its insecticidal properties were disco-
vered by P. Müller in 1939 [51]. This discovery
marked the beginning of the era of synthetic or-
ganic insecticides. Because DDT appeared at the
onset of World War II, it rapidly became the stan-
dard remedy for control of typhus and malaria in
both military and civilian use. At its peak use in
1962, about 85×106 kg of DDT was produced
18 Insect Control
in the United States. The compound has an ap-
propriate combination of properties: (1) broad
spectrum of insecticidal activity, (2) long resid-
ual persistence, (3) low acute toxicity to mam-
mals, and (4) low cost. The enormous publicity
it received after World War II heightened its use
as a panacea for all insect pest control problems,
and DDT quickly became the standard remedy
for control of the codling moth on deciduous
fruit, the pink bollworm on cotton, the Colorado
potato beetle, and the European corn borer. Dur-
ing the 1960s, DDT was registered for use on 334
agricultural commodities in the United States.
It was highly effective against forest defoliators
such as the gypsy moth and the spruce budworm
and as a home mothproofing agent. The safety
and long residual action of DDT stimulated the
WHO program for global eradication of malaria,
and it still has major place in malaria control.
The compound known as DDT [50-29-
3], 1,1 -(2,2,2-trichloroethylidene)bis(4-chloro-
benzene), is a white crystalline solid, mp 109 ◦ C,
vapor pressure at 20 ◦ C 0.025 mPa. Technical
DDT is a white amorphous powder containing
65 – 80 % of the active 4,4 -DDT; 14 – 21 % of
the nearly inactive 2,4 -DDT [789-02-6], mp
73 – 74 ◦ C; up to 4 % of the insecticidal DDD
[72-54-8], 1,1 -(2,2-dichloroethylidene)bis(4-
chlorobenzene), mp 110 ◦ C; and traces of 2,2 -
DDT and bis(4-chlorophenyl)sulfone. Technical
DDT, mp 80 –94 ◦ C, is almost insoluble in water
(1.2 µg/L, readily soluble in xylene and tetralin
(600 g/L), and moderately soluble in mineral
oil and kerosene (50 – 80 g/L). In alkaline solu-
tion, DDT is dehydrochlorinated readily to the
noninsecticidal 1,1-dichloroethenylidenebis(4-
chlorobenzene) [72-55-9] (DDE), mp 85 ◦ C.
This reaction also occurs in biological systems
and accounts for the storage and accumulation
of DDE in lipid tissue of humans and higher
animals. The rat LD50 values of DDT are 113,
118 mg/kg (oral) and 2510 mg/kg (dermal).
Methoxychlor [72-43-5], 1,1 -(2,2,2-tri-
chloroethylidene)bis(4-methoxybenzene), mp
89 ◦ C, is less easily dehydrochlorinated than
DDT in alkaline solution or in biological sys-
tems. However, the methoxy groups are readily
dealkylated in vivo by microsomal oxidases to
produce phenols that are eliminated easily. Thus,
methoxychlor does not bioaccumulate like DDT
and is favored for general environmental use.
Methoxychlor is one of the safest of all insecti-
cides with an LD50 (rat, oral) of > 6 000 mg/kg.
Unfortunately, insects resistant to DDT are gen-
erally cross-resistant to methoxychlor.
The compound known as DDD [72-54-
8] (TDE), 1,1 -(2,2-dichloroethylidene)bis(4-
chlorobenzene), occurs as an impurity in tech-
nical DDT and has been marketed as an in-
secticide. Technical DDD contains mainly the
4,4 -isomer, mp 110 ◦ C, and 7 – 8 % of the 2,4 -
isomer, mp 76 ◦ C. The properties of DDD are
almost indistinguishable from DDT and the
product is even safer with LD50 values (rat)
of 3 400 mg/kg (oral) and > 10 000 mg/kg (der-
mal). Like DDT, DDD is also bioaccumulated
and its registration was also canceled in the
United States in 1974.
Perthane [72-56-0], 1,1 -(2,2-dichloroeth-
ylidene)bis(4-ethylbenzene), mp 60 – 61 ◦ C, is
the 4,4 -diethyl analogue of DDT. It is very
biodegradable and has low mammalian toxicity,
LD50 (rat, oral) 8 170 mg/kg. Perthane has been
used indoors, where very low acute and chronic
toxicities are desirable.
The compound DFDT [475-26-3],
1,1 -(2,2,2-trichloroethylidene)bis(4-fluoro-
benzene), 1,1 -(2,2,2-trichloroethylidene)bis-

(4-fluorobenzene), mp 44 ◦ C, has chemical and
insecticidal properties similar to DDT. It was
used as a sanitary insecticide by the German
army in World War II but is too phytotoxic
for agricultural use. The LD50 (rat, oral) is
900 mg/kg.
5.1.1. Mode of Action
Both DDT and its analogues specifically af-
fect the peripheral sensory organs of insects
to produce violent trains of afferent impulses
that cause hyperactivity and convulsions. Sub-
sequent paralysis and death are the result of
metabolic exhaustion or the production of an en-
dogenous neurotoxin. The insecticidal action of
DDT is facilitated by its highly lipophilic char-
acter, which promotes absorption through the
lipoprotein matrix of the chitinous insect cuticle.
The DDT molecule is absorbed into the sodium
gate mechanism of the insect nerve axon, pro-
longing the negative afterpotential and promot-
ing multiple, repetitive nerve impulse firing. The
critical site of action may be the inhibition of
Ca2+ ATPase.
5.1.2. Environmental Toxicology
Because DDT is the most durable and persis-
tent of all the commonly used contact insecti-
cides, the very properties that made it such an
effective insecticide–its low vapor pressure, sta-
bility to photooxidation, high fat solubility (ca.
100 g/kg), and low water solubility (1.2 µg/L) –
resulted in its becoming the prototype envi-
ronmental micropollutant. As a result of its
enormous use worldwide and the environmen-
tal stability of its primary degradation prod-
uct DDE, DDT – DDE residues are ubiquitous
in the global environment. By 1969, the av-
erage U.S. inhabitant had 2.3 – 4.0 mg/kg of
DDT and 4.3 –8.0 mg/kg of DDE stored in body
fat; levels in other countries were even higher.
This widespread persistence and bioaccumu-
lation were key factors in the withdrawal of
DDT registrations by the EPA, as of January
Insect Control 19
1974. The DDT residues in humans declined
slowly to about one-half the above values by
1978. In vivo DDT is slowly dechlorinated re-
ductively to DDD and dechlorinated further to
DDA or 4,4 -dichlorodiphenylacetic acid, the
predominant excretory metabolite. In anaero-
bic microorganisms, DDT may form 4,4 -di-
chlorodiphenylacetonitrile (DDCN). However,
most DDT that enters the environment is se-
questered as DDE. In humans, DDT is secreted
in milk as DDT, DDD, and DDE, with traces of
their 2,4 -isomers.
Whereas DDT is highly toxic to fish (LC50 for
trout and bluegill 0.002 – 0.008 mg/kg), it is only
moderately toxic to birds (oral LD50 for mallard
1 300 mg/kg and for pheasant > 2 240 mg/kg).
However, because of its high lipid – water par-
tition value, DDT bioaccumulates as much as
3×106 -fold from the waters of Lake Michigan
6 ng/L in the bodies of lake trout Salvelinus na-
maycush (> 20 mg/kg). In raptorial birds such as
eagle, osprey, and brown pelican, bioaccumula-
tion of DDE from fish has decreased eggshell
thickness by 10 – 15 %, resulting in widespread
reproductive failure. Because of the many envi-
ronmental problems resulting from massive use
of DDT, its registration for agricultural use was
canceled in the United States and Canada; Japan
and many European countries have enacted sim-
ilar prohibitions.
5.2. Hexachlorocyclohexanes [49]
Lindane [58-89-9], γ-1,2,3,4,5,6-
hexachlorocyclohexane, is the active component
of hexachlorocyclohexane (benzene hexachlo-
ride) insecticides. Benzene hexachloride was
first produced by Faraday in 1825 by chlorina-
tion of benzene in sunlight. Its insecticidal prop-
erties were discovered independently by Dupire
and Raucourt in France and by Slade et al.
in England around 1941. Crude benzene hex-
achloride is a grayish to brownish amorphous
solid, with a characteristic musty odor, which
begins to melt at 65 ◦ C. It consists of 10 – 18 %
of the active γ-isomer, mp 112 ◦ C, configuration
aaaeee (a is axial, e is equatorial), together with
at least four other nearly inactive stereoisomers:
α-isomer, mp 157 ◦ C (aaeeee or aeeeea), 55 –
70 %; β-isomer, mp 309 ◦ C (eeeeee), 5 – 14 %;
δ-isomer, mp 138 ◦ C (aeeeee), 6 – 8 %; ε-isomer,
20 Insect Control
mp 219 ◦ C (aeeaee), 3 – 4 %; and a trace of η-
isomer, mp 90 ◦ C (aeaaee). A heptachlorocy-
clohexane and an octachlorocyclohexane are
present as impurities and contribute to the un-
pleasant odor of benzene hexachloride.
Lindane is the only component with apprecia-
ble insecticidal activity; it is prepared by treat-
ing crude BHC with methanol or acetic acid in
which the α- and β-isomers are nearly insolu-
ble, followed by chromatographic adsorption or
fractional crystallization.
Lindane, vapor pressure 0.12 Pa at 20 ◦ C, is a
white crystalline product that is practically odor-
less, soluble in water to 7.3 mg/L, and soluble in
petroleum solvents. It is very stable to sunlight
and oxidation, and can be burned without ap-
preciable decomposition; however, it is decom-
posed by alkali to form mainly 1,2,4-trichloro-
benzene. The rat LD50 is 88, 91 mg/kg (oral) and
900, 1 000 mg/kg (dermal).
Lindane is used predominantly as a seed
dressing, as a soil insecticide, and for the con-
trol of locusts and grasshoppers. It is also used
to control ectoparasites of humans and domes-
tic animals and as a residual spray to control the
Anopheles vectors of malaria. Because of its rel-
atively high volatility, it is useful for controling
wood-boring insects of timber, fruit trees, and
ornamental plants.
The mode of action of lindane is not well un-
derstood, but like DDT and the cyclodienes it
causes distortion and consequent excitation of
ionic transmission in the arthropod nerve. The
mechanism is believed to be competitive block-
age of γ-aminobutyric acid (GABA) synaptic
transmission.
Crude benzene hexachloride causes severe
off-flavors in root crops grown in treated soil, but
the problem does not occur after the use of puri-
fied lindane. Lindane is highly toxic to fish, with
LD50 values to trout of 0.027 ppm and bluegill
0.068 ppm. It has low oral toxicity in birds, an
LD50 of > 2 000 mg/kg for mallards. The aver-
age soil half-life is about two years.
5.3. Cyclodienes [49]
The enormous commercial success of DDT
stimulated further investigation of organochlo-
rine insecticides. Julius Hyman developed
chlordane, the first cyclodiene insecticide, in
1944. The cyclodienes are polychlorinated cy-
clic hydrocarbons with endomethylene-bridged
structures. They are prepared by the Diels –
Alder reaction, with hexachlorocyclopentadiene
as the primary dienophile to form adducts with
a variety of dienes.
Chlordane [12789-03-6] is made by
chlorinating hexachlorodicyclopentadiene
(chlordene) to form two isomers of
1,2,4,5,6,7,8,8-octachloro-2,3,3a,4,7,7a-hexa-
hydro-4,7-methano-1H-indene, α-trans, mp
106.5 ◦ C, and β-cis, mp 104.5 ◦ C. The β-
isomer has significantly greater insecticidal ac-
tivity. Technical chlordane is an amber liquid of
d 25
4 1.61 which contains about 60 % of these iso-
mers, 10 – 20 % heptachlor, and lesser amounts
of chlordene and a nonachloro derivative. A to-
tal of 14 components have been identified by
gas – liquid chromatography.
Technical chlordane, vapor pressure 1.3 mPa
at 25 ◦ C, is soluble in petroleum solvents but
has very low water solubility (0.009 ppm). It
has rat LD50 values of 335, 430 mg/kg (oral)
and 840, 690 mg/kg (dermal). Chlordane has had
very wide use as a soil insecticide for termite
control and as a household insecticide for the
control of ants and cockroaches, together with
limited agricultural use as a soil insecticide.
Heptachlor [76-44-8], 1,4,5,6,7,8,8-
heptachloro-3a,4,7,7a-tetrahydro-4,7-methano-
indene, mp 95 ◦ C is three to five times as active
insecticidally as chlordane. The pure crystalline
material, vapor pressure 0.039 Pa at 20 ◦ C, is
soluble in petroleum solvents but very insolu-
ble in water to 0.056 ppm. The rat LD50 values

are 100, 162 mg/kg (oral) and 195, 250 mg/kg
(dermal). Heptachlor is oxidized readily to hep-
tachlor epoxide [1024-57-3], mp 159 – 160 ◦ C,
LD50 (rat, oral) 47 mg/kg, which is an important
and highly persistent environmental pollutant.
Endosulfan [115-29-7] is the adduct of
hexachlorocyclopentadiene and 1,4-dihydroxy-
2-butene reacted further with SOCl2 to produce
6,7,8,9,10,10-hexachloro-1,5,5a,6,9,9a-hexa-
hydro-6,9-methano-2,4,3-benzodioxathiepin-
3-oxide. The technical product is a brownish
solid, mp 70 – 100 ◦ C, vapor pressure 1.3 mPa
at 25 ◦ C, soluble in petroleum solvents but hav-
ing low solubility in water. The rat LD50 values
are 43, 18 mg/kg (oral) and 130, 74 mg/kg (der-
mal). Technical endosulfan consists of about
seven parts of the α-isomer, mp 108 – 109 ◦ C
(cis with regard to the sulfite groups) and three
parts of the β-trans isomer, mp 206 – 208 ◦ C.
The α-isomer has somewhat greater insectici-
dal activity and is slowly converted to the more
stable β-isomer at high temperature. Both iso-
mers oxidize slowly in air and in biological
systems to endosulfan sulfate [1031-07-8], mp
181 – 182 ◦ C. In acid media, both isomers form
endosulfan diol [2157-19-9], mp 203 – 205 ◦ C.
Endosulfan is a broad-spectrum insecticide
used to control pests of vegetables, fruit, field
crops, and ornamentals. Unlike other cyclodiene
insecticides, it is biodegradable by hydrolysis at
the sulfite ester bonds.
Isobenzan [297-78-9], 1,3,4,5,6,7,8,8-
octachloro-1,3,3a,4,7,7a-hexahydro-4,7-
methanoisobenzofuran is produced by pho-
tochlorination of the Diels – Alder adduct
Insect Control 21
of hexachlorocyclopentadiene and 2,5-di-
hydrofuran. It has a mp of 122 – 123 ◦ C and a
vapor pressure of 0.39 mPa at 20 ◦ C; it is soluble
in water to about 0.1 mg/L at 20 ◦ C. Isobenzan
had uses in Europe similar to heptachlor, but
production was discontinued in 1966 because
of its very high mammalian toxicity, LD50 (rat,
oral) 7, 8.7 mg/kg.
Aldrin [309-00-2], 1,2,3,4,10,10-
hexachloro-1,4,4a,5,8,8a-hexahydro-1,4-
endo,exo-5,8-dimethanonaphthalene, is the
Diels – Alder adduct of hexachlorocylopenta-
diene and bicyclo[2.2.1]-hepta-2,5-diene. It is
a colorless solid, mp 104 ◦ C, vapor pressure
5.2 mPa at 20 ◦ C. This compound is very slightly
soluble in water (to 0.027 mg/L) but soluble in
petroleum hydrocarbons. Aldrin has rat LD50
values of 39, 60 mg/kg (oral) and 98 mg/kg (der-
mal). It has been widely used as a seed treat-
ment and soil insecticide where it is gradually
oxidized to its epoxide dieldrin.
Dieldrin [60-57-1], 3,4,5,6,9,9-hexachloro-
1a,2,2a,3,6,6a,7,7a-octahydro-2,7 : 3,6-di-
methanonaphth[2,3-b]oxirene, is the 6,7-ep-
oxide of aldrin, mp 176 ◦ C, vapor pressure
0.023 mPa at 20 ◦ C. It is produced by oxidation
of aldrin with peracetic or perbenzoic acid, and is
very slightly soluble in water (to 0.05 ppm) and
soluble in petroleum solvents. Dieldrin has rat
LD50 values of 46 mg/kg (oral) and 90, 60 mg/kg
(dermal). It is a highly persistent insecticide and
was used extensively as a residual treatment
for control of the Anopheles mosquito vectors
of malaria and the tsetse fly vectors of African
22 Insect Control

trypanosomiasis. However, resistance to it de- United States in 1976. Mirex is environmentally

veloped very rapidly. In agriculture, it had more very stable; it is slowly oxidized to chlordecone
limited use as a seed treatment and soil insecti- and dechlorinated in sunlight to photomirex.
cide. Because of the environmental persistence
of dieldrin and its propensity for bioaccumu-
lation, the registrations of both dieldrin and
aldrin, which is readily epoxidized to dieldrin,
were canceled in the United States in 1974.

Chlordecone [143-50-0], decachloro-5-

oxo-pentacyclo[5.3.0.02,6 .03,9 .04,8 ]decane, mp
349 ◦ C (decomp.), is the 2-keto analogue of
mirex and is soluble in water to 4 g/L at 100 ◦ C.
Endrin [72-20-8], 3,4,5,6,9,9-hexachloro- The rat LD50 values are 95, 140 mg/kg (oral) and
1a,2,2a,3,6,6a,7,7a-octahydro-2,7 : 3,6-di- > 2 000 mg/kg (dermal). Like mirex, chlorde-
methanonaphth[2,3-b]oxirene, mp 245 ◦ C (de- cone is a stomach poison insecticide; it has been
comp.), vapor pressure 0.022 mPa at 25 ◦ C, is the used in bait for cockroaches and ants and for the
endo-endo-isomer of dieldrin and is produced by control of banana thrips. After widespread envi-
a Diels – Alder reaction of hexachloronorborna- ronmental contamination of the James River and
diene (formed from hexachlorocyclopentadiene Chesapeake Bay with chlordecone from manu-
and acetylene or vinyl chloride) with cyclo- facturing effluent, its registration was canceled
pentadiene, followed by epoxidation. Endrin in the United States in 1978.
is soluble in water to about 0.06 ppm and is
somewhat less soluble in petroleum solvents
than dieldrin. The rat LD50 values are 17.8,
7.5 mg/kg (oral) and 15 mg/kg (dermal). Endrin
has been used as a cotton insecticide, but because
of its high toxicity to fish and its propensity for
bioaccumulation, its uses have been restricted
substantially. It has also had very limited use as
a rodenticide for field mouse control.

Kelevan [4234-79-1], decachloro-5-(4-

ethoxycarbonyl-2-oxo-butyl)-5-hydroxypenta-
cyclo-[5.3.0.0.2,6 .03,9 .04,8 ]decone, mp 91 ◦ C,
is a chlordecone analogue that is produced by
reaction with ethyl levulinate. It has been used
in Europe as a stomach poison for the control of
Mirex [2385-85-5] is a dimer of hexa- the Colorado potato beetle.
chlorocyclopentadiene dodecachloropentacy-
clo-[5.3.0.02,6 .03,9 .04,8 ]decane, mp 485 ◦ C, rat
LD50 306, 600 mg/kg (oral) and > 2 000 mg/kg
(dermal). Mirex is extremely resistant to
biodegradation. It was once considered the per-
fect stomach poison for control of the imported
fire ant Solenopsis invicta but, even when used
in bait at a dosage of a few grams per hectare,
was found to bioaccumulate in birds and other
animals. Its registration was canceled in the

Bromodan [1715-40-8], 5-bromomethyl-
1,2,3,4,7,7-hexachlorobicyclo[2.2.1]hept-2-
ene, mp 75 – 79 ◦ C, is the Diels – Alder adduct
of hexachlorocyclopentadiene and allyl bro-
mide. It has very low mammalian toxicity, LD50
(rat, oral) 1.2×104 mg/kg, and has been used
in Europe to control ectoparasites of domestic
animals.
Mode of Action [52]. The cyclodienes ap-
pear to act by a common mechanism affect-
ing the nerve axon to produce hypersensitivity,
convulsions, prostration, and death. Spray crew-
men with lengthy exposure to dieldrin have suf-
fered from prolonged and repeated central ner-
vous system epileptiform convulsions; similar
nervous disturbances have occurred in workers
exposed to chlordecone. The lesion appears to
be blockage of GABA synaptic transmission by
competitive inhibition of the neurotransmitter
binding site. This mode of action is common
to the cyclodienes, toxaphene, and lindane. The
highest toxic action is shown by compounds con-
taining a polychlor center and an adjacent, prop-
erly oriented electronegative site (e.g., a double
bond adjacent to a Cl, O, S, or N atom).
Environmental Toxicology. Cyclodienes
add oxygen readily; e.g., aldrin and heptachlor
form the more stable epoxides dieldrin and hep-
tachlor epoxide. The epoxides are formed in the
tissues of animals that ingest or absorb the com-
pounds, and they are concentrated and stored
preferentially in fatty tissue. Epoxidation also
occurs in plants and soil. Because dieldrin and
heptachlor epoxide are more stable (e.g., the
half-life on alfalfa leaves is 7 – 8 d, compared
to < 1 d (for aldrin and heptachlor), in situ ox-
idation markedly affects the residual behavior
of these cyclodienes. Average soil half-lifes for
cyclodienes are estimated as aldrin 1 – 4 years,
dieldrin 1 – 7 years, chlordane 2 – 4 years, and
endrin 4 – 8 years. The high lipophilicity of
dieldrin and heptachlor epoxide has resulted in
severe environmental pollution. They bioaccu-
mulate in fish from water up to 105 -fold and are
Insect Control 23
present in human milk and fatty tissue. Dieldrin
residues in the body fat of U.S. inhabitants av-
eraged 0.22 – 0.35 ppm in 1970 and declined to
0.12 – 0.14 ppm by 1978, after cancellation of
the use of aldrin and dieldrin in 1974. Oxychlor-
dane [26880-48-8], mirex, and chlordecone are
also highly persistent and bioaccumulative.
The cyclodienes are extremely toxic to fish
with the following LC50 values (in milligrams
per kilogram) to trout and bluegill: endrin
0.001 – 0.002, endosulfan 0.001 – 0.003, aldrin
0.006 – 0.01, dieldrin 0.003 – 0.015, heptachlor
0.03 –0.026, and chlordane 0.022 – 0.095. The
LD50 values to pheasant and mallard, respec-
tively, are aldrin 16.8 and 520, dieldrin 79 and
381, and endrin 1.6 and 5.6 mg/kg. Lifetime ani-
mal feeding studies have shown that heptachlor,
aldrin, and dieldrin are carcinogens. The data
with endrin are equivocal.
5.4. Chlorinated Camphenes
Toxaphene, C10 H10 Cl8 , is a mixture of more
than 180 chlorinated cyclic compounds pro-
duced by photochlorination of the bicyclic ter-
pene camphene. The product is a yellowish,
semicrystalline gum, mp 65 – 90 ◦ C, d 254 1.64,
vapor pressure 0.13 mPa at 25 ◦ C. Toxaphene
is soluble in water to about 0.4 mg/L and very
soluble in petroleum solvents. It is unstable in
alkali, upon prolonged exposure to sunlight, and
above 155 ◦ C, liberating hydrogen chloride and
losing some of its insecticidal potency. The rat
LD50 values are 80, 90 mg/kg (oral) and 1 075,
780 mg/kg (dermal). The most active ingredi-
ents identified from technical toxaphene are
2,2,5- endo-6-exo-8,9,10-heptachlorobornane
[51775-36-1], with a mouse intraperitoneal
LD50 of 6.6 mg/kg, and 2,2,5-endo-6-exo-
8,9,9,10-octachlorobornane [58002-18-9], with
a mouse intraperitoneal LD50 of 3.1 mg/kg.
Each constitutes 2 – 6 % of the technical mix-
ture. Toxaphene is a broad-spectrum persistent
insecticide with major use on cotton and for the
control of animal ectoparasites.
24 Insect Control
Environmental Toxicology. Toxaphene is
extremely toxic to fish, with LC50 values to
trout and bluegill of 0.003 – 0.006 mg/kg. At a
water concentration as low as 50 ng/L, fish ex-
posed to toxaphene residues suffer broken-back
syndrome, a crippling collagen deformity. From
water, bioaccumulation occurs in fish at levels
up to 105 -fold. Toxaphene is also highly toxic
to birds: the oral LD50 values for pheasant mal-
lard are 40 and 71 mg/kg, respectively. Because
of the complexity of the toxaphene mixture, its
environmental persistence and fate are difficult
to assess. Substantial toxaphene residues have
been found throughout the Great Lakes, as a re-
sult of atmospheric distribution from heavy use
on cotton in the southern states. The estimated
soil half-life ranges from 2 months to 10 years.
Toxaphene is carcinogenic in rats and mice, and
its registration was canceled by the U.S. EPA in
1983.
6. Organophosphorus Insecticides
[53–57]
The organophosphorus (OP) insecticides are pri-
marily triesters of phosphoric acid P(O)(OH)3
and phosphorothioic acid P(S)(OH)3 .
Historical Aspects. Discovery of the biolog-
ical activity of organophosphate esters began
with the chance observation in 1932 that expo-
sure to the vapor of diethyl phosphorofluoridate,
(C2 H5 O)2 P(O)F, produced strong cholinergic
effects in humans [58]. Development of this
class of compounds as insecticides resulted from
the work of Gerhard Schrader in 1937, who
devised tetraethyl pyrophosphate ( TEPP) as
a substitute for nicotine. Schrader’s research
also led to the related, extremely toxic com-
pounds, the “ nerve gases ” tabun and sarin; be-
cause of wartime secrecy, the development of
insecticides did not become evident until 1946.
Because TEPP hydrolyzes rapidly in water, it
was ineffective in conjunction with Bordeaux
mixture used as a fungicide for potatoes. To
find a more stable compound to control the Col-
orado potato beetle Leptinotarsa decomlineata,
Schrader turned to phosphorothionate esters
and produced methyl and ethyl parathions, the
most widely used of the organophosphate insec-
ticides. Schrader’s development of schradan,
demeton (the first practical systemic insecti-
cide), azinphosmethyl (the standard remedy
for the codling moth Laspeyresia pomonella),
trichlorfon (a stomach poison), coumaphos (for
animal ectoparasites), and numerous other in-
secticidal molecules, followed from his elucida-
tion of the principles of organophosphorus tox-
icity based on the formula:
where R1 and R2 are short-chain alkoxy, alkyl,
or amino residues, X is a displaceable group
with a high-energy phosphorus bond, such as
fluorine or an acyl anhydride; and pentavalent
phosphorus is bonded directly to oxygen or sul-
fur. At present ca. 100 commercially successful
OP insecticides are marketed in estimated quan-
tities of (1.5 –2.5)×105 t. More than 105 differ-
ent organophosphorus esters appear to have been
synthesized and evaluated as pesticides. The po-
tential number of toxic structures that can be de-
vised from Schrader’s formula is in excess of
25×106 [59].
Organophosphorus insecticides are available
with a variety of properties. Some, such as TEPP
and mevinphos, have a very short residual ac-
tion; others, such as diazinon, azinphos-methyl,
and chlorpyrifos have a lengthy persistence.
Some have a broad spectrum of action (e.g.,
parathion and malathion), whereas others have a
very specific action (e.g., schradan and trichlor-
fon). The unique properties of plant absorption
and translocation found in dimethoate and deme-
ton have resulted in successful plant systemic
insecticides. Persistent compounds used in soil
and seed treatment (e.g., disulfoton, phorate, and
terbufos) protect newly developed seedlings.
Animal systemics such as cruformate can be fed
to cattle to kill the cattle grubs (Hypoderma spp.)
living in their bodies.
The organophosphorus insecticides are
biodegradable because of their chemical re-
activity, especially through the phosphate ester
bonds, and their use does not present serious
problems of bioaccumulation and food chain
transfer. Their major use as replacements for
organochlorine insecticides has had favorable
consequences on environmental quality.

However, organophosphorus compounds act
as specific inhibitors of the enzyme acetyl-
cholinesterase, which is an essential component
of nerve impulse transmission in both verte-
brates and invertebrates. Therefore, they gener-
ally lack selectivity. Careless use and improper
storage have caused hundreds of thousands of
cases of human poisoning and thousands of
deaths. By taking advantage of differences in
the detoxification processes and the structures
of the target enzyme in insects and mammals,
a few insecticides including malathion, ronnel,
fenitrothion, stirifos, and phoxim have been pro-
duced, which incorporate an effective degree of
insecticidal action with adequate safety to the
user and to domestic animals.
6.1. Phosphoric Acid Anhydrides
The aliphatic organophosphorus insecticides de-
veloped by Schrader hydrolyzed rapidly and
had very high mammalian toxicity. They are now
of largely historical interest.
Tetraethyl pyrophosphate [107-49-3] is a
clear, colorless liquid, bp 104 – 110 ◦ C (10.7 Pa),
d 25
4 1.1845. It is miscible with water and hy-
drolyzes rapidly (half-life 6.8 h at 25 ◦ C). The rat
LD50 values are 1.0 mg/kg (oral) and 2.4 mg/kg
(dermal).
Sulfotepp [3689-24-5], O,O,O ,O -tetra-
ethyl dithiopyrophosphate is a yellowish liq-
uid, bp 110 – 113 ◦ C (26.7 Pa), d 25
4 1.196. It
is soluble in water to 20 mg/L and hydrolyzes
only in alkaline solution. The LD50 (rat, oral) is
5 mg/kg. Sulfotepp is used in aerosols for green-
house pest control.
Dimefox [115-26-4], N,N,N  ,N  -tetra-
methylphosphorodiamidofluoridate, bp 67 ◦ C
Insect Control 25
(0.5 kPa), d 25
4 1.12, LD50 (rat, oral) 3.5 mg/kg,
and Schradan [152-16-9], octamethyl
pyrophosphoramide, bp 154 ◦ C (0.5 kPa), d 25 4
1.13, LD50 (rat, oral) 10 mg/kg, were the first
practical systemic insecticides for mites and
aphids.
6.2. Vinyl Phosphates
Dichlorvos [62-73-7], O,O-dimethyl
O-(2,2-dichlorovinyl) phosphate, bp 140 ◦ C
(2.7 kPa), vapor pressure 1.6 Pa at 20 ◦ C, d 20
4
1.314, is a colorless liquid that is soluble in
water to about 10 g/L. The half-life in water is
about 8 h. The LD50 values (rat, oral) are 80,
56 mg/kg. Dichlorvos is a rapid-acting insecti-
cide with strong fumigant action, used in bait
and aerosols for flies and mosquitoes. It has
been incorporated in slow-release plastic strips
and in pet collars against ectoparasites.
Naled [300-76-5], O,O-dimethyl O-(1,2-di-
bromo-2,2-dichloroethyl) phosphate, mp 27 ◦ C,
bp 110 ◦ C (66.7 Pa), vapor pressure 0.26 Pa at
20 ◦ C, has rat LD50 values of 250 mg/kg (oral)
and 800 mg/kg (dermal). Naled is a rapid-acting,
relatively nonpersistent insecticide.
Trichlorfon [52-68-6], O,O-dimethyl-1-
hydroxy-2,2,2-trichloroethyl phosphonate, mp
83 –84 ◦ C, d 20
4 1.73, vapor pressure 1 mPa at
20 ◦ C, is soluble in water to 15.4 % at 20 ◦ C. The
rat LD50 values are 630, 560 mg/kg (oral) and
26 Insect Control

> 2 000 mg/kg (dermal). Trichlorfon is a stom-
ach poison insecticide used on foliage for chew-
ing insects and in dry sugar bait for fly control.
Above pH 6, trichlorfon is converted rapidly to
dichlorvos, which accounts for its toxicity.
42 mg/kg (dermal). Dicrotophos is a contact and
systemic insecticide which has been injected
into trees to control bark and foliage pests.

Monocrotophos [6923-22-4], cis-O,O-di-

methyl-O-(1-methyl-2-methylcarbamoyl)vinyl
Mevinphos [7786-34-7], O,O-dimethyl O-
phosphate, mp 54 – 55 ◦ C, vapor pressure
(2-methoxycarbonyl-1-methylvinyl) phosphate,
0.9 mPa at 20 ◦ C, is miscible with water. The
bp 106 – 107 ◦ C (0.13 kPa), d 204 1.25, vapor rat LD50 values are 18, 20 mg/kg (oral) 126,
pressure 0.38 Pa at 21 ◦ C, is a mixture of one
112 mg/kg (dermal). Monocrotophos, a metabo-
part cis and two parts trans isomer, of which
lite of dicrotophos, is a systemic and contact
the former is about a ten times stronger insecti-
insecticide.
cide. Mevinphos is miscible with water; the rat
LD50 values are 6.1, 3.7 mg/kg (oral) and 4.7,
4.2 mg/kg (dermal). Mevinphos is a nonpersis-
tent systemic insecticide suitable for treatment
of edible produce close to harvest because the
residue dissipates rapidly by volatilization and
hydrolysis.
Crotoxyphos [7700-17-6], O,O-dimethyl
O-[1-methyl-2-(1-phenylcarbethoxy)vinyl]
phosphate, bp 135 ◦ C (4 Pa), d 254 1.19, va-
por pressure 1.8 mPa at 20 ◦ C, is soluble in
water to 1 g/L. The rat LD50 values are 110,
74 mg/kg (oral) and 375, 202 mg/kg (dermal).
Phosphamidon [13171-21-6], O,O-di- Crotoxyphos is used primarily to control animal
methyl O-(2-chloro-2-diethylcarbamoyl-1- ectoparasites.
methyl)vinyl phosphate, bp 160 ◦ C (0.2 kPa),
d 25 ◦
4 1.21, vapor pressure 3.2 mPa at 20 C, is
miscible with water. The rat LD50 values are
24 mg/kg (oral) and 143, 107 mg/kg (dermal).
Phosphamidon is absorbed rapidly by plant sur-
faces and decomposes quickly in the plant to
form a nonpersistent systemic insecticide.
Chlorfenvinphos [470-90-6], O,O-di-
ethyl O-[2-chloro-1-(2,4-dichlorophenyl)vinyl]
phosphate, bp 168 – 170 ◦ C (66.7 Pa), d 15
4 1.36,
vapor pressure 0.97 mPa at 25 ◦ C, is soluble in
water to 145 mg/L. The rat LD50 values are
15, 13 mg/kg (oral) and 31, 30 mg/kg (dermal).
Chlorfenvinphos is a contact and soil insecticide
Dicrotrophos [141-66-2], cis-O,O- for agricultural pests.
dimethyl O-(2-dimethylcarbamoyl)-1-
methylvinyl phosphate, bp 115 – 120 ◦ C
(0.13 Pa), d 23
4 1.19, is water miscible. The rat
LD50 values are 21, 16 mg/kg (oral) and 43,
 Insect Control 27

0.2 %. The technical isomer mixture (2 : 1 ra-

tio of Demeton S to Demeton O) has been used
widely as a systemic insecticide for ornamen-
tals and nursery stock. The LD50 (rat, oral) is
1.5 mg/kg. The rat LD50 values for the mixture
are 6.2, 2.5 mg/kg (oral) and 14, 8.2 mg/kg (der-
mal).

Tetrachlorvinphos [22248-79-9], Stir-

ifos, O,O-dimethyl O-[2-chloro-1-(2,4,5-tri-
chlorophenyl)vinyl] phosphate, mp 97 – 98 ◦ C,
vapor pressure 5 µPa at 20 ◦ C, is soluble in wa-
ter to 11 mg/L; its rat LD50 values are 1100,
1125 mg/kg (oral) and > 5 000 mg/kg (dermal).
Stirifos is used for protection against stored A series of related compounds are also marketed
product pests, against vegetable and fruit insects, as systemic insecticides but are less persistent
and in veterinary hygiene. and less toxic. The P=O ester is largely respon-
sible for systemic activity.

Demeton methyl [8022-00-2] is a mixture of

the O,O-dimethyl P=S and O,O-dimethyl P=O
esters analogous to demeton. The respective
LD50 values (rat, oral) are 180 and 40, 60 mg/kg.

Propetamphos [31218-83-4], O-[2-(iso-

propoxycarbonyl)-1-methylvinyl] O-methyl N-
ethyl phosphoramidothioate, bp 87 – 89 ◦ C
(0.7 Pa), d 204 1.129, is water-soluble to
Oxydemeton-methyl [301-12-2], O,O-
110 mg/L. The LD50 values (rat, oral) are 75,
dimethyl S-(2-ethylsulfinyl)ethyl phos-
82 mg/kg. Propetamphos is used as a household
phorothiolate has rat LD50 values 47, 52 mg/kg
insecticide for flies and cockroaches and to con-
(oral) and 173, 158 mg/kg (dermal).
trol insect pests of cotton.
Metasystox S. O,O-dimethyl S-(2-
ethylsulfinyl-1-methyl)ethyl phosphorothioate,
LD50 (rat, oral) 105 mg/kg, is also a commercial
insecticide.

6.3. Aliphatic Phosphorothioate Esters

Demeton O [8065-48-3], O,O-diethyl O-
(2-ethylthio)ethyl phosphorothioate (thiono-
isomer, demeton O), bp 123 ◦ C (0.13 kPa), d 21
4
1.119 is water-soluble to 66 mg/L. The LD50
(rat, oral) is 30 mg/kg. Demeton S [126-75-0], The P(S)S analogues of the demeton type are
O,O-diethyl S-(2-ethylthio)ethyl phosphoro- much less water-soluble, more persistent sys-
thioate (thiolo-isomer, demeton S), bp 128 ◦ C temics than P(O)S compounds. They have been
(0.13 kPa), d 21 21
4 4 1.132, is water-soluble to
28 Insect Control

very widely used in granular soil and seed treat- mp 51 – 52 ◦ C, d 65

4 1.277, is soluble in water to
ment for row crops and ornamentals. 70 g/L. The rat LD50 values are 215, 245 mg/kg
(oral) and 610 mg/kg (dermal). Dimethoate is a
Disulfoton [298-04-4], O,O-diethyl S-(2- widely used contact and systemic insecticide for
ethylthio)ethyl phosphorodithioate is a pale- vegetables, fruit, and ornamental plants.
yellow liquid, bp 62 ◦ C (1.3 Pa), d 21 20
4 4 1.144. It
is soluble in petroleum solvents and to 66 mg/L
in water. The rat LD50 values are 6.8, 2.3 mg/kg
(oral) and 25, 6 mg/kg (dermal).

Dimethoate Analogues. Omethoate

[1113-02-6], the P=O analogue of dimethoate,
is an activated metabolite of dimethoate and a
practical systemic insecticide that is especially
effective against aphids and red spider mites.
Thiometon [144-41-2], O,O-dimethyl S-(2-
The LD50 (rat, oral) is 50 mg/kg.
ethylthio)ethyl phosphorodithioate, has an LD50
(rat, oral) of 85 mg/kg.

Further dimethoate analogues include

formothion [2540-82-1], O,O-dimethyl S-
(N-formyl-N-methylcarbamoyl)methyl phos-
Phorate [298-02-2], O,O-diethyl S-(2-
phorodithioate, mp 25 ◦ C, LD50 (rat, oral)
ethylthio)methyl phosphorodithioate, bp 118 –
275 mg/kg; morphothion [144-41-2], O,O-
120 ◦ C (0.1 kPa), is a yellowish liquid, d 21 21 25
4 4 4
◦ dimethyl S-(morpholinocarbonyl)methyl
1.167, vapor pressure 0.11 Pa at 20 C, soluble
phosphorodithioate; and vamidothion
in petroleum solvents and in water to 85 mg/L.
[2275-23-2], O,O-dimethyl S-[2-(1-methyl-
The rat LD50 values are 2.3, 1.1 mg/kg (oral)
carbamoyl)ethylthio]ethyl phosphorothioate,
and 6.2, 2.5 mg/kg (dermal).
mp 46 – 48 ◦ C, soluble in water to 40 g/L, LD50
(rat, oral) 64, 105 mg/kg.

Terbufos [13071-79-9], O,O-diethyl S-(1,1-

dimethylethylthio)methyl phosphorodithioate is
a yellow liquid, bp 69 ◦ C (1.3 Pa), d 24
4 1.165,
soluble in water to 10 – 15 mg/L. The LD50 val-
ues (rat, oral) are 4.5, 9.0 mg/kg. Terbufos is
more persistent than phorate and is used as a
granular soil treatment for the control of corn
root worms.
Phenthoate [2597-03-7], O,O-dimethyl S-
(α-carboethoxy)phenylmethyl phosphorodi-
thioate, mp 17 – 18 ◦ C, d 204 1.226 has LD50
values (rat, oral) of 300, 400 mg/kg. Phenthoate
is used as a contact and systemic insecticide for
tree fruit pests.
Dimethoate [60-51-5], O,O-dimethyl S-(N-
methylcarbamoyl)methyl phosphorodithioate,

Carbophenothion [786-19-6], O,O,-diethyl
S-(4-chlorophenylthio)methyl phosphorodi-
thioate, bp 82 ◦ C (1.3 Pa), d 20
4 1.285, rat LD50
values 30, 10 mg/kg (oral) and 54, 27 mg/kg
(dermal), and its O,O-dimethyl analogue car-
bophenothion methyl [953-17-3], rat LD50
values 98, 120 mg/kg (oral) and 215, 190 mg/kg
(dermal), are persistent general-purpose con-
tact insecticides used on fruit, vegetables, and
forage.
Malathion [121-75-5], O,O-dimethyl S-
(1,2-dicarbethoxy)ethyl phosphorodithioate, bp
156 –157 ◦ C (93.3 Pa), is a yellow oil, d 25
4 1.23,
vapor pressure 5.2 mPa at 30 ◦ C, soluble in
water to 145 mg/L. Malathion hydrolyzes eas-
ily above pH 7.0 and below pH 5.0. The rat
LD50 values are 1375, 1000 mg/kg (oral) and
> 4 000 mg/kg (dermal). Malathion is a persis-
tent general-purpose insecticide, and its rela-
tively low toxicity makes it especially suitable
in household, home garden, vegetable, and fruit
insect control, as well as in control of insects for
public health (e.g., flies, mosquitoes, and lice).
Ethion [563-12-2], O,O,O ,O -tetraethyl

S,S -methylene diphosphorodithioate, bp
165 ◦ C (53.3 Pa), d 204 1.22, vapor pressure
0.2 mPa at 20 ◦ C, rat LD50 values 65, 27 mg/kg
(oral) and 245, 62 mg/kg (dermal), is an acari-
cide and insecticide used on fruit and vegetables.
Insect Control 29
Dioxathion [78-34-2], O,O,O ,O -tetraethyl
S,S  -(1,4-dioxane-2,3-diyl) diphosphorodi-
thioate, d 26
4 1.26, rat LD50 43, 23 mg/kg (oral)
and 235, 63 mg/kg (dermal), is a general-
purpose insecticide and acaricide for deciduous
fruit.
6.4. Phosphorothioate Esters of Phenols
The phosphorothioate esters of phenols contain
a high-energy phosphate bond–the pseudoan-
hydride bond of phosphoric acid and the phe-
nol. The 4-nitrophenyl phosphate esters, the
parathions, are the cheapest and most widely
used OP insecticides, but they are also among
the most toxic to nontarget organisms and hu-
mans. The hazard of human poisoning from the
use of parathions has stimulated the develop-
ment of safer analogues.
Parathion [56-38-2], O,O-diethyl O-4-
nitrophenyl phosphorothioate is a dark-brown
liquid with a garlic odor, bp 375 ◦ C, d 25
4 1.265,
and vapor pressure 4.9 mPa at 20 ◦ C. Parathion
is resistant to aqueous hydrolysis, the half-life
is 120 d in water and 8 h in lime water. Above
130 ◦ C, parathion isomerizes slowly to form
O,S-diethyl O-4-nitrophenyl phosphorothiolate
[597-88-6], which is a much less stable and
less effective insecticide. Parathion is reduced
readily to the nontoxic O,O-diethyl O-4-ami-
nophenyl phosphorothioate and oxidized with
difficulty to the very highly toxic paraoxon
O,O-diethyl O-4-nitrophenyl phosphorothiolate
[311-45-5]. Parathion is very soluble in aro-
matic petroleum solvents, very slightly soluble
in kerosene and mineral oil, and soluble in wa-
ter to 0.2 mg/L. The rat LD50 values are 13,
3.6 mg/kg (oral) and 21, 6.8 mg/kg (dermal).
30 Insect Control
Methyl parathion [298-00-0], O,O-di-
methyl O-4-nitrophenyl phosphorothioate, mp
36 ◦ C, d 20
4 1.358, has very similar properties to
parathion but hydrolyzes and isomerizes more
readily. The rat LD50 values are 14, 24 mg/kg
(oral) and 67 mg/kg (dermal). Because of its
lower acute toxicity, methyl parathion is used
more widely than parathion.
Fenitrothion [122-14-5], O,O-dimethyl O-
(3-methyl-4-nitrophenyl) phosphorothioate, a
brownish oil, bp 140 – 145 ◦ C (13.3 Pa), d 25 4
1.322, is soluble in aromatic hydrocarbons and
in water to 20 mg/L. The rat LD50 values are
740, 570 mg/kg (oral) and 300, 400 mg/kg (der-
mal). Fenitrothion is a general-purpose insecti-
cide with about the same spectrum of activity
as methyl parathion. It is also safe for residual
house spraying in malaria control.
Dicapthon [2463-84-5], O,O-dimethyl O-
(2-chloro-4-nitrophenyl) phosphorothioate, mp
51 ◦ C, rat LD50 400, 330 mg/kg (oral) and
790, 1250 mg/kg (dermal), and chlorthion
[500-28-7], O,O-dimethyl O-(3-chloro-4-nitro-
phenyl) phosphorothioate, mp 21 ◦ C, d 20
4 1.437,
rat LD50 880, 980 mg/kg (oral) and 4 500,
4 100 mg/kg (dermal), are two safer analogues of
methyl parathion that have been used as house-
hold insecticides.
Ronnel [299-84-3], fenchlorphos, O,O-di-
methyl O-2,4,5-trichlorophenyl phosphoro-
thioate, mp 35 – 37 ◦ C, is soluble in petroleum
hydrocarbons and in water to 80 mg/L. The
LD50 values (rat, oral) are 1 250, 2 630 mg/kg.
Ronnel is a safe, persistent insecticide that is
used in veterinary hygiene especially to control
cattle grubs. It is too phytotoxic for agricultural
use.
Bromophos [2104-96-3], O,O-dimethyl
O-(2,5-dichloro-4-bromophenyl) phosphoro-
thioate, mp 54 ◦ C, vapor pressure 17 mPa at
20 ◦ C, water solubility 40 mg/L, rat LD50
1 600, 1 730 mg/kg (oral) and > 5 000 mg/kg
(dermal), and iodofenphos [18181-70-9],
O,O-dimethyl O-(2,5-dichloro-4-iodophenyl)
phosphorothioate, mp 76 ◦ C, water solubility
20 mg/L, LD50 (rat, oral) 2100 mg/kg, are very
closely related to ronnel.
They are relatively safe used in veterinary hy-
giene and as household insecticides.
Fenthion [55-38-9], O,O-dimethylO-(3-
methyl-4-methylthiophenyl) phosphorothioate,
is a colorless oil, bp 105 ◦ C (1.3 Pa), d 20
4 1.250,
vapor pressure 3.9 mPa at 20 ◦ C, which is sol-
uble in water to 55 mg/L. The rat LD50 values
are 215, 245 mg/kg (oral) and 330 mg/kg (der-
mal). Fenthion is a persistent insecticide used in
veterinary hygiene and as a mosquito larvicide;
it is especially effective in polluted water. It has
also been used for bird pest control.
 Insect Control 31

Fensulfothion [115-90-2], O,O-diethyl O-

(4-methylsulfinylphenyl) phosphorothioate, bp
138 – 141 ◦ C (1.3 Pa), rat LD50 4.1, 1.8 mg/kg
(oral) and 19, 4.1 mg/kg (dermal), is used as a
soil insecticide and nematocide.
Sulprofos [35400-43-2], O-ethyl O-[(4-
methylthio)phenyl] S-propyl phosphorodi-
thioate, bp 155 – 158 ◦ C (13.3 Pa), d 20
4 1.20,
LD50 (rat, oral) variously given as 22.7, 60, and
227 mg/kg, is used for control of lepidopterous
larvae that attack cotton.

Temephos [3383-96-8], O,O,O ,O -tetra-

methyl O,O-(thiodi-4,1-phenylene) phosphoro-
dithioate, mp 30 ◦ C, d 20
4 1.32, water solubility
0.025 mg/L, rat LD50 8 600, 13 000 mg/kg (oral)
and > 4 000 mg/kg (dermal), is used as a larvi-
cide for blackflies and mosquitoes.
6.5. Phosphonothioate Esters of Phenols
A number of relatively persistent O-alkyl O-
aryl phosphorothioates have strong insecticidal
properties. Some produce delayed neuropathy in
hens and in humans, and have been withdrawn
from use.

EPN [2104-64-5], O-ethyl O-(4-nitro-

phenyl) phenylphosphonothioate, mp 41 ◦ C,
d 20
4 1.268, rat LD50 36, 7.7 mg/kg (oral) and
Coumaphos [56-72-4], O,O-diethyl-O-3-
230, 25 mg/kg (dermal), is used as a cotton
chloro-4-methyl-2-oxo-2H-1-benzopyran-7-yl
insecticide, but is a delayed neurotoxin.
phosphorothioate, mp 91 ◦ C, vapor pressure
0.013 mPa at 20 ◦ C, rat LD50 41, 16 mg/kg
(oral) and 1000 mg/kg (dermal), is applied as
a 0.25 – 0.5 % spray to control animal parasites
and cattle grubs.

Cyanofenphos [13067-93-1], O-ethyl O-

(4-cyanophenyl) phenylphosphonothioate, mp
83 ◦ C, vapor pressure 1.7 mPa at 25 ◦ C, water
solubility 6 mg/L, has an LD50 (rat, oral) of
44 mg/kg but is a delayed neurotoxin.
Isofenphos [25311-71-1], O-ethyl O-[2-
(isopropoxycarbonyl)phenyl] N-isopropyl
phosphoroamidothioate, bp 120 ◦ C (1.3 Pa), d 20
4
1.13, vapor pressure 0.52 mPa at 20 ◦ C, water
solubility 24 mg/L, LD50 (rat, oral) 28 mg/kg, is
a soil insecticide.
32 Insect Control

Leptophos [21609-90-5], O-methyl O-(4-

bromo-2,5-dichlorophenyl) phenylphospho-
nothioate, mp 71 – 72 ◦ C, d 25 4 1.53, soluble
in water to 0.03 mg/L, LD50 (rat, oral) 53,
45 mg/kg, is a persistent insecticide that is espe-
cially effective against lepidopterous larvae, but
it is a delayed neurotoxin.
Chlorpyrifos [2921-88-2], O,O-diethylO-
(3,5,6-trichloro-2-pyridinyl) phosphorothioate,
mp 42 – 43 ◦ C, vapor pressure 2.5 mPa at
25 ◦ C, water solubility 2 mg/L, rat LD50 135,
165 mg/kg (oral) and 2000 mg/kg (dermal), is
used widely as a soil insecticide, in veterinary
hygiene, and as a household insecticide for cock-
roach and termite control.
Fonofos [944-22-9], O-ethyl S-phenyl
ethylphosphonodithioate, bp 130 ◦ C (13.3 Pa),
d 25
4 1.16, water solubility 13 mg/L, LD50 (rat,
oral) 8, 17 mg/kg, is a persistent soil insecticide.

Chlorpyrifos-methyl [5598-13-0], the O,O-

Trichlornate [327-98-0], O-ethyl O-2,4,5-
trichlorophenyl ethylphosphonothioate, LD50
(rat, oral) 37.5 mg/kg, has been used as a soil
insecticide but is a delayed neurotoxin.
dimethyl analogue of chlorpyrifos, LD50 (rat,
oral) 941, 2140 mg/kg, is much safer and is used
to control blackfly and mosquito larvae.
Fospirate [5598-52-7], dimethyl 3,5,6-tri-
chloro-2-pyridinyl phosphate, mp 87 ◦ C, LD50
(rat, oral) 869 mg/kg, is the P = O analogue of
chlorpyrifos and is used as a veterinary parasiti-
cide.

Pirimiphos-ethyl [23505-41-1], pirim-

ithate, O,O-diethylO-(2-diethylamino-6-
methyl-4-pyrimidinyl) phosphorothioate, bp
128 – 132 ◦ C (5.3 Pa), d 1.165, LD50 (rat, oral)
140, 200 mg/kg, is used as an agricultural insec-
6.6. Phosphorothioate Esters of ticide and acaricide.
Heterocyclic Enols

Diazinon [333-41-5], O,O-diethyl O-[(2-

isopropyl-4-methyl)pyrimidin-6-yl] phosphoro-
thioate, is one of the oldest broad-spectrum
organophosphorus insecticides. Diazinon is a
brownish oil, bp 83 – 84 ◦ C (0.3 Pa), d 20
4 1.116,
vapor pressure 18.2 mPa at 20 ◦ C, soluble in wa- Pirimiphos-methyl [29232-93-7], O,O-
ter to 40 mg/L. The rat LD50 values are 250, dimethyl O-(2-diethylamino-6-methyl-4-
285 mg/kg (oral), and the rabbit LD50 (dermal) pyrimidinyl) phosphorothioate, d 30
4 1.157 rat
is > 1 000 mg/kg. Diazinon is used as a soil in- LD50 2 050 mg/kg (oral) and > 2 000 mg/kg
secticide and to control agricultural pests. (dermal), the O,O-dimethyl analogue of
pirimiphos-ethyl is a much safer insecticide
 Insect Control 33

used to control aphids, insects in stored prod-

ucts and for veterinary hygiene.

6.7. Phosphorothioate Esters of

S-Methyl Heterocycles

Azinphos-methyl [86-50-0], O,O-dimethyl Phosalone [2310-17-0], O,O-diethyl S-

S-(3,4-dihydro-4-oxobenzo-[d]-(1,2,3)-triazin- (6-chloro-2,3-dihydro-2-oxobenzoxazol-3-
3-yl methyl) phosphorodithioate, mp 73 – 74 ◦ C, yl)methyl phosphorodithioate, mp 48 ◦ C, LD50
d 20
4 1.44, water solubility 30 mg/L, has rat LD50
(rat, oral) 120, 175 mg/kg, is an insecticide and
values of 13, 11 mg/kg (oral) and 220 mg/kg acaricide for tree fruit pests.
(dermal). The compound is a persistent insec-
ticide used extensively to control the codling
moth and other pests of deciduous fruit, as well
as those of vegetables, ornamentals, and cotton.

Methidathion [950-37-8], O,O-dimethyl

S-[5-methoxy-2-oxo-1,3,4-thiadiazol-3(2H)-
yl]methyl phosphorodithioate, mp 39 – 40 ◦ C,
LD50 (rat, oral) 25, 48 mg/kg, is used as an in-
Azinphos-ethyl [2642-71-9], mp 53 ◦ C, d 20 4 secticide and acaricide for fruit and vegetables.
1.284, vapor pressure 0.029 mPa at 20 ◦ C, LD50
(rat, oral) 17.5 mg/kg, is the diethyl analogue of
azinphos-methyl and is also a commercial insec-
ticide.

Phosmet [732-11-6], O,O-dimethyl S-(N-

phthalimidomethyl) phosphorodithioate, mp
72 ◦ C, vapor pressure 0.13 Pa at 50 ◦ C, water
Azamethiphos [35575-96-3], O,O-dimethyl
solubility 24 mg/L, has rat LD50 values of 113,
S-[(6-chloro-2-oxooxazolo[4,5-b]pyridin-
160 mg/kg (oral) and > 1 500 mg/kg (dermal). It
3(2H)-3-yl)methyl] phosphorodithioate, mp
is used to control pests of deciduous fruit.
89 – 90 ◦ C, LD50 (rat, oral) 1 100, 1 750 mg/kg,
is used for household pest control.

Dialifor [10311-84-9], O,O-diethyl S-

(2-chloro-1-phthalimidoethyl) phosphorodi-
thioate, mp 67 – 69 ◦ C, rat oral LD50 43,
71 mg/kg and rabbit dermal LD50 145 mg/kg, is
used to control pests of deciduous fruit.
Menazon [78-57-9], O,O-dimethyl S-(4,6-
diamino-1,3,5-triazin-2-yl)methyl phosphoro-
dithioate, mp 160 – 162 ◦ C, vapor pressure
0.13 mPa at 25 ◦ C, water solubility 2.4 g/L,
LD50 (rat, oral) 1 950 mg/kg, is a systemic aphi-
cide.
34 Insect Control
Endothion [2778-04-3], O,O-dimethyl S-
[(5-methoxy-oxo-4H-pyran-2-yl)methyl]phos-
phorothioate, mp 96 ◦ C, LD50 (rat, oral) 30,
50 mg/kg, is a systemic insecticide.
6.8. Miscellaneous Organophosphorus
Esters
Crufomate [299-86-5], O-methyl O-
[2-chloro-4-(1,1-dimethylethyl)phenyl]
methylphosphoramidate, mp 61 ◦ C, LD50 (rat,
oral) 635, 460 mg/kg, is an animal systemic in-
secticide. It is used primarily for cattle grub
control either as a 0.75 % spray or when fed at
20 – 25 mg/kg body weight.
Phoxim [14816-18-3], O,O-diethylO-(α-
cyanobenzilideneamino) phosphorothioate, bp
102 ◦ C (1,3 Pa), d 20
4 1.176, LD50 (rat, oral)
2 500 mg/kg, is especially useful for control of
stored product pests.
Methamidophos [10265-92-6], O,S-di-
methyl phosphoramidothioate, mp 39 – 41 ◦ C,
d 44.5
4 1.31, vapor pressure 39 mPa at 30 ◦ C,
LD50 (rat, oral) 19, 21 mg/kg, is a contact and
systemic insecticide for insect pests of cotton,
potato, and vegetable crops.
Acephate [30560-19-1], O,S-dimethyl N-
acetyl phosphoramidothioate is the N-acetyl
derivative of methamidophos, mp 72 – 80 ◦ C,
water soluble to 650 g/L, LD50 (rat, oral) 866,
945 mg/kg. It has much lower acute toxicity
than methamidophos, and is effective against a
wide variety of agricultural and household insect
pests.
Ethoprophos [13194-48-4], Prophos, Etho-
prop, O-ethyl S,S-dipropyl phosphorodithioate,
bp 86 – 91 ◦ C (26.6 Pa), d 20 1.094, vapor
4
pressure 45 mPa at 26 ◦ C, soluble in water
to 750 mg/L, rat LD50 61 mg/kg (oral) and
26 mg/kg (dermal), is used to control insect pests
and nematodes in vegetables and ornamentals.
6.9. Mode of Action and Toxicity [60–66]
6.9.1. Inhibition of Acetylcholinesterase
The biological activity of the organophosphorus
insecticides results from the capacity of the cen-
tral phosphorus atom to phosphorylate the ac-
tive site of acetylcholinesterase (E.C. 3.1.1.7),
which is an essential enzyme of invertebrates
and vertebrates. The phosphorylated enzyme is
irreversibly inhibited and, therefore, cannot per-
form its normal function of rapidly hydrolyz-
ing and removing the neurotransmitter acetyl-
choline from the nerve synapse. Acetylcholine

consequently accumulates and disrupts the nor-
mal functioning of the nervous system, pro-
ducing typical cholinergic symptoms (i.e., hy-
peractivity, tremors, convulsions, paralysis, and
death). In higher animals and in humans, these
cholinergic effects are translated into muscarinic
effects (e.g., nausea, salivation, lachrymation,
and myosis of the pupil of the eye); nicotinic ef-
fects (e.g., muscular fasiculations); and central
effects (e.g., giddiness, tremulousness, convul-
sions, and coma).
The reaction between the esterase and the
phosphorus inhibitor involves attack on the elec-
trophilic phosphorus atom by a nucleophilic ser-
ine hydroxyl group with the neighboring imi-
dazole ring of a histidine residue as the active
site of acetylcholinesterase. The normal three-
step reaction (A) that occurs at the synapse bet-
ween acetylcholinesterase (EH) and the sub-
strate acetylcholine is
In the normal process (A), step 3 occurs very
rapidly and step 1 is the rate-determining step.
However, in the inhibition process (B), step 3 is
rate-determining because it occurs very slowly
(from a matter of hours with O,O-dimethyl phos-
phates to days with O,O-diethyl phosphates).
The organophosphorus insecticides may, there-
fore, be regarded as synthetic neurotransmit-
ters with very slow turnover numbers. In vitro,
the organophosphorus insecticide (e.g., tetra-
ethyl pyrophosphate, mevinphos, or dichlorvos)
engages in a first-order reaction with acetyl-
cholinesterase. The inhibited phosphorylated
enzyme is relatively stable and contains 1 mol of
Insect Control 35
phosphorus per mole of enzyme. An equimolar
quantity of alcoholic, phenolic, or acidic product
HX is liberated as the leaving group.
Delay Factor. The reactivity of organophos-
phorus insecticides is determined by the elec-
trophilicity of the phosphorus atom, the strength
of the “ high-energy ” P−X bond, and steric ef-
fects of the substituents. The electrophilicity of
the central phosphorus is determined by the rel-
ative electronegativities of the atoms bonded to
it (e.g., P 2.1, O 3.5, S 2.5, N 3.0, and C 2.5).
Therefore, the phosphorus in a phosphate es-
ter (P=O) is much more electrophilic than the
oxygen, and these compounds are much more
reactive than the phosphorothioate esters (P=S).
The latter are generally so stable as to be unre-
active with acetylcholinesterase. Their insecti-
cide activity and general toxicity result from in
vivo oxidation by a microsomal oxidase in the
insect gut and fat body tissue and in the mam-
malian liver. A typical example is the oxidation
of parathion to paraoxon [311-45-5]:
The time required for this in vivo oxidation
or activation (usually several hours) is called the
delay factor and is important in human toxi-
cology because at least a portion of the insec-
ticide can be detoxified and eliminated during
this period. The delay factor also provides op-
portunity for therapy with atropine which pro-
tects the postsynaptic nerve membrane from the
accumulating acetylcholine. Because of the de-
lay factor, phosphorothionates (P=S) are almost
always less toxic and less hazardous to the user
than their equivalent phosphates (P=O).
The substituents on phosphorus strongly af-
fect its electrophilic character by inductive
and mesomeric effects. If the substituents are
electron-attracting, the phosphorus atom be-
comes more reactive and the insecticide more
effective. For the simple O,O-dimethyl and
O,O-diethyl phenyl phosphates and phosphoro-
thioates of the parathion – paraoxon type, a very
36 Insect Control
high degree of correlation exists between the
quantitative measure of electron-withdrawing
power of the substituent (Hammett’s σ) and
anticholinesterase activity, and toxicity [58],
[59]. This explains the high insecticidal activ-
ity of molecules with 4-NO2 (methyl parathion),
4-CN, 4-CH3 SO (fensulfothion), and 2,4,5-Cl
(ronnel) phenyl leaving groups.
Activation. In vivo, microsomal oxidation
is also important for activation or lethal syn-
thesis in which electron-donating or inert
groups (e.g., CH3 S) are converted to strongly
electron-withdrawing groups (e.g., CH3 SO and
CH3 SO2 ). Thus, insecticides such as fenthion
are activated in vivo by microsomal oxidation
to produce strongly electrophilic phosphorus
atoms. This type of reaction involves a built-in
delay that is particularly effective in certain sys-
temic insecticides, such as phorate and disulfo-
ton, where both P=S and C2 H5 S oxidation occur
in plant tissues:
The originial insecticide is relatively
lipophilic and stable. It is absorbed readily and
translocated through plant tissue without de-
structive hydrolysis until microsomal oxidation
occurs. The activated (oxidized) insecticide is
highly toxic to sucking insects and relatively
unstable so that it hydrolyzes to nontoxic prod-
ucts. Microsomal oxidation can also take place
in the mammalian liver, which makes many of
these systemic insecticides extremely toxic to
humans.
6.9.2. Delayed Neuropathy [57], [65], [66]
Certain organophosphorus insecticides [espe-
cially phosphonates and phosphonothionates
with P(O)C bonds] phosphorylate an enzyme
neurotoxic esterase to produce a delayed neu-
ropathy that is essentially irreversible and may
occur from one to four weeks after acute or
chronic exposure. Neurotoxic esterase is evi-
dently required for growth and maintenance of
the long axons of the central nervous system. Af-
ter its inhibition, degenerative changes develop
in the long axons and are manifest by tingling
and burning sensations in the limb extremities
that progress to weakness in the lower limbs and
to ataxia. The final stage is irreversible paraly-
sis which, in severe poisoning, may affect upper
limbs as well. Recovery is very slow and seldom
complete in adults, with flaccid paralysis chang-
ing to spasticity. The phosphorylated esterase is
modified by “ aging ” so that the remaining es-
ter bond of the organophosphorus inhibitor at-
tached to the active site of the neurotoxic es-
terase is cleaved to form a negatively charged
residue bound to the enzyme.
Delayed neuropathy was originally observed
in humans after excessive exposure to tri-
o-cresyl phosphate and has been produced
by mipafox, EPN, leptophos, and methami-
dophos. Many other related compounds such as
cyanofenphos, trichlormat, and isofenphos pro-
duce delayed neuropathy in hens.
6.9.3. Selective Toxicity
The organophosphorus insecticides are gener-
ally highly toxic to nontarget organisms, and
their use can be hazardous to humans. Human
poisoning can result from acute inhalation, der-
mal exposure, or oral ingestion. Many of the in-
secticides are nearly as toxic when applied to the
skin as when ingested (cf. oral and dermal LD50
values). Fatalities have resulted from ingestion
of as little as 2 mg of parathion in small children
to 120 mg in adults. Some of the most acutely
toxic organophosphorus insecticides (e.g., phor-
ate, disulfoton, and terbufos) are marketed only
as granular or pelletized products that are ap-
plied directly to the soil and thus reduce the toxic
hazard.

Acetylcholinesterase inhibition in the blood
and central nervous system of humans is cumu-
lative and can increase to fatal levels upon re-
peated chronic exposure of agricultural work-
ers. This is due to the irreversible inhibition
of acetylcholinesterase with O,O-diethyl phos-
phates. Therefore this vital enzyme may have
to be resynthesized. Hazards resulting from use
of the more toxic organophosphorus insecticides
(Class I poisons with LD50 < 50 mg/kg) can be
reduced by wearing protective clothing includ-
ing mask, goggles, and rubber gloves. Work-
ers with frequent exposure should have regu-
lar determinations of blood acetylcholinesterase
levels and should be removed from exposure,
when these blood acetylcholinesterase levels
show progressive decrease. In cases of severe
exposure and incipient cholinergic symptoms,
atropine should be administered promptly and
the patient should be hospitalized.
Selectivity is implicit in the definition of an
insecticide as “ an agent for destroying insects ”.
Nevertheless, many highly toxic organophos-
phorus compounds are still used in massive
quantities throughout the world. Their selectiv-
ity is apparent only through careful application
and confinement to the area of treatment. Selec-
tivity is important in providing safety not only
for the user of pesticides and domestic animals,
but also for wildlife and beneficial insects: honey
bees and other pollinators, parasites, and preda-
tors.
The development of malathion in 1950 was
an important milestone in the use of selective in-
secticides. Malathion is about one-half to one-
twentieth as toxic as parathion to insects, but
it is only about one one-hundredth as toxic to
humans and higher animals. The use of thou-
sands of metric tons of malathion over several
decades has demonstrated substantial safety to
users, coupled with effectiveness in insect pest
control.
A high degree of selectivity has been demon-
strated by incorporating chlorine or methyl
groups in the 3-position of the aryl ring of
parathion. Such groups interact stereochemi-
cally with acetylcholinesterase, increasing the
affinity for the insect enzyme and decreas-
ing it for the mammalian enzyme. Such com-
pounds, including fenitrothion, fenthion, ronnel,
bromophos, and iodofenphos, have been used
widely and safely in household, animal health,
Insect Control 37
and public health pest control. Safety is also en-
hanced by differences in the rates of microsomal
oxidation of P=S to P=O (see Section 6.9.1) and
in hydrolytic detoxification between insects and
mammals.
The user of OP insecticides can enhance
selectivity and safety by choosing only com-
pounds with LD50 values > 100 mg/kg, by us-
ing O,O-dimethyl instead of O,O-diethyl insec-
ticides, and by using P=S in preference to P=O
insecticides. Great care should be exercised in
handling, applying, and storing all organophos-
phorus insecticides.
6.10. Environmental Toxicology
Organophosphorus insecticides are reactive and
are degraded readily by oxidation and hydro-
lysis in vivo. Therefore, the use of organophos-
phorus esters does not result in serious problems
of bioaccumulation and food chain transfer. Soil
persistence is low, with typical half-life values
ranging from one or two weeks for malathion to
three to six months for parathion, diazinon, and
azinphos-methyl. However, organophosphorus
insecticides are general biocides and are toxic
to nearly all animals. Oral LD50 values, in mil-
ligrams per kilogram, are phorate, mallard 0.62,
pheasant 7.1; parathion, mallard 2.0, pheasant
12; monocrotophos, mallard 4.8, pheasant 2.8;
diazinon, mallard 3.5, pheasant 4.3; fenthion,
mallard 5.9, pheasant 18; chlorpyrifos, mallard
75, pheasant 12; and azinphos-methyl, mallard
136, pheasant 75. The organophosphorus com-
pounds are generally toxic to fish, and LD50 val-
ues, in parts per million, are phorate, bluegill
0.0055; parathion, bluegill 0.10, trout 0.047; di-
azinon, bluegill 0.052, trout 0.38; malathion,
bluegill 0.11, trout 0.13; fenthion, bluegill 1.4,
trout 0.93; and azinphos-methyl, bluegill 0.005,
trout 0.004. Organophosphorus insecticides are
highly toxic to bees, beneficial parasites, and
predators.
7. Carbamate Insecticides [67–71]
The carbamates are synthetic analogues of the
toxic alkaloid physostigmine [57-47-6] isolated
from the calabar bean Physostigma venenosum
38 Insect Control

in 1864. Its structure was elucidated by Sted- 7.1. Dimethylcarbamates of

man and Barker (1925), and its cholinergic na- Heterocyclic Enols
ture was discovered by Loewi and Nawratil
in 1926. Stedman found that 3-dimethylami- Isolan [119-38-0], 1-isopropyl-3-methyl-5-
nophenyl methylcarbamate methiodide was a pyrazoyl dimethylcarbamate, is a colorless liq-
highly effective cholinergic drug in aqueous so- uid, bp 105 – 107 ◦ C (40 Pa), LD50 (rat, oral),
lution. The more stable 3-dimethylaminophenyl 54 mg/kg, used as a systemic aphicide.
dimethylcarbamate methiodide was developed
as the drug prostigmine. Because of their posi-
tively charged quaternary structure, prostigmine
and its medicinally effective relatives cannot
penetrate the lipid sheath of insect nerves and
are noninsecticidal. Gysin (1954) developed the
N,N-dimethylcarbamate of heterocyclic enols
as insecticides [72]. Kolbezen, Metcalf, and Dimetilan [644-64-4], 2-dimethyl-
Fukuto (1954) produced the first N-methylcar- carbamoyl-3-methyl-5-pyrazolyl dimethylcar-
bamates of phenols that have a broader spectrum bamate, mp 68 – 71 ◦ C, is a water-soluble car-
of activity and are now the most widely used bamate used in sugar bait for housefly control.
carbamate insecticides [73]. Subsequently, N- The rat LD50 values are 25, 64 mg/kg (oral) and
methylcarbamates of a variety of oximes were 600 mg/kg (dermal).
found to be extraordinarily toxic and have been
used extensively as systemic insecticides.
Carbamates were the third major group of
synthetic organic insecticides, developed af-
ter organochlorine and organophosphorus com-
pounds. Because of their high degree of
biodegradability, they are free from the per-
sistence and bioaccumulation associated with
organochlorines. Carbamates, although often
highly toxic to humans and higher animals, Pirimicarb [23103-98-2], 2-(dimethylami-
are reversible inhibitors of acetylcholinesterase no)-5,6-dimethyl-4-pyrimidinyl dimethylcarba-
and much safer to use than the irreversible mate, mp 90.5 ◦ C, vapor pressure 4 mPa at
organophosphorus inhibitors. Additionally, car- 30 ◦ C, is water-soluble to 0.25 %. The rat LD50
bamates provided a welcome alternative to values are 147 mg/kg (oral) and > 500 mg/kg
the serious problem of insect pest resistance (dermal). Pirimicarb is a systemic aphicide used
that had developed with organochlorines and largely on grain crops.
organophosphates. Therefore, carbamates have
become widely used for soil insects and in fo-
liar application to field and vegetable crops. Ap-
proximately 11×106 kg (active ingredient) were
applied to U.S. farm crops in 1976. Carbaryl pro-
duction was estimated at 25×106 kg in 1971, and
global consumption of carbamate insecticides
has been estimated to be as high as 35×106 kg.
Propoxur was developed as a residual insecti-
cide for controlling adult mosquitoes in human 7.2. Methylcarbamates of Phenols
habitations and has been used as a replacement
for DDT. Aminocarb [2032-59-9], (4-dimethylami-
no-3-methyl)phenyl methylcarbamate, mp
93 ◦ C, rat LD50 40, 38 mg/kg (oral) and 280,
320 mg/kg (dermal), is particularly effective
against lepidopterous larvae, snails, and slugs.

Bendiocarb [22781-23-3], 2,2-dimethyl-
1,3-benzodioxol-4-yl methylcarbamate, mp
129 –130 ◦ C, vapor pressure 0.7 mPa at 25 ◦ C, is
water-soluble to 0.04 g/L. The rat LD50 values
are 179 mg/kg (oral) and 1000 mg/kg (dermal).
Bendiocarb is used to control cockroaches, other
household pests, and soil insects.
BPMC [3766-81-2], 2-(1-methylpropyl)-
phenyl methylcarbamate, mp 32 ◦ C, bp 112 –
113 ◦ C (40 Pa), d 20
4 1.050, 48 mPa at 20 C,
◦
LD50 (rat) 410 mg/kg, is used to control leafhop-
pers and plant hoppers on rice.
Butacarb [2655-19-8], 3,5-di-tert-
butylphenyl methylcarbamate, mp 97 – 99 ◦ C,
LD50 (rat, oral) > 4 000 mg/kg, has been used in
veterinary entomology to control sheep blowfly
larvae in wounds.
Carbaryl [63-25-2], 1-naphthyl methylcar-
bamate, mp 142 ◦ C, vapor pressure < 5 mPa
Insect Control 39
at 20 ◦ C, is soluble in water to 0.12 g/L. The
rat LD50 values are 540 mg/kg (oral) and
> 4 000 mg/kg (dermal). Carbaryl is a broad-
spectrum insecticide registered on more than
100 crop plants in the United States. It is detox-
ified and eliminated rapidly in animals, and is
neither concentrated in fat nor secreted in the
butterfat of milk. Therefore, carbaryl is favored
for application to food crops.
Carbofuran [1563-66-2], 2,3-dihydro-2,2-
dimethylbenzofuranyl methylcarbamate, mp
150 – 152 ◦ C, vapor pressure 2.7 mPa at 33 ◦ C, is
soluble in water to 0.7 g/L. The rat LD50 values
are 4.8 mg/kg (oral) and > 10 000 mg/kg (der-
mal). Carbofuran is a broad-spectrum insecti-
cide used largely to control soil insects.
Dioxacarb [6988-21-2], 2-(1,3-dioxolan-2-
yl)phenyl methylcarbamate, mp 114 – 115 ◦ C,
vapor pressure 0.05 mPa at 20 ◦ C, is soluble
in water to ca. 6 g/L. The rat LD50 values
are 125 mg/kg (oral) and 3 000 mg/kg (dermal).
Dioxacarb is used to control cockroaches and a
variety of household or stored product insects.
Ethiofencarb [29973-13-5], 2-(ethylthio-
methyl)phenyl methylcarbamate, is a yellowish
oil, d 20 ◦
4 1.147, vapor pressure 13 mPa at 30 C,
water solubility 1.8 g/L. The rat LD50 values
are 411, 499 mg/kg (oral) and > 1 150 mg/kg
40 Insect Control

(dermal). Ethiofencarb is a systemic insecticide Metalkamate [8065-36-9], bufencarb, a

effective against aphids. mixture of three parts 3-(1-methylbutyl)phenyl
carbamate and one part 3-(1-ethylpropyl)phenyl
methylcarbamate, mp 26 – 36 ◦ C, LD50 (rat,
oral) 87, 170 mg/kg, has been used extensively
as a soil insecticide.

Etrofol [3942-54-9], CPMC, 2-chlorophe-

nyl methylcarbamate, mp 87 ◦ C, LD50 (rat, oral)
648 mg/kg, LD50 (mouse) 150 mg/kg, is used to
control leafhoppers and plant hoppers in rice.
Methiocarb [2032-65-7], (3,5-dimethyl-
4-methylthio)phenyl methylcarbamate, mp
121 ◦ C, 15 mPa at 25 ◦ C, rat LD50 70, 60 mg/kg
(oral) and > 2 000 mg/kg (dermal), is used to
control insect pests of fruit and vegetables.

Formetanate [22259-30-9], N,N-dimethyl-

N -{3-[(methylamino)carbonyl]oxy}phenyl
methanimidamide, 3-dimethylaminomethy-
leneiminophenyl methylcarbamate, vapor pres-
sure 0.027 mPa at 25 ◦ C, water solubility
6.3 g/L, rat LD50 15 mg/kg (oral) and > 5 60
house flies
0 mg/kg (dermal), is marketed as the water- Mexacarbate [315-18-4], (4-dimethylami-
miscible hydrochloride and used as an acaricide no-3,5-dimethyl)phenyl methylcarbamate, mp
for deciduous fruit. 85 ◦ C, is soluble in water to 0.1 g/L. Rat LD50
values are 37, 25 mg/kg (oral) and 1 500 mg/kg
(dermal). Mexacarbate has been used principally
to control lepidopterous forest pests.

Isoprocarb [2631-40-5], 2-(1-methylethyl)-

phenyl methylcarbamate, mp 93 ◦ C, LD50 (rat,
oral) 403, 485 mg/kg, is used to control pests of
MPMC [2425-10-7], 3,4-dimethylphenyl
rice and cacao.
methylcarbamate, mp 79 – 80 ◦ C, vapor pres-
sure 70 mPa at 25 ◦ C, water solubility 0.58 g/L,
LD50 (rat, oral) 380 mg/kg, has been used to
control leafhoppers and plant hoppers that at-
tack rice.

MTMC [1129-41-5], 3-methylphenyl
methylcarbamate, mp 76 – 77 ◦ C, 145 mPa at
25 ◦ C, LD50 (rat, oral) 268 mg/kg, has also been
used to control leafhoppers and plant hoppers
that attack rice.
Promecarb [2631-37-0], 3-methyl-5-(1-
methylethyl)phenyl methylcarbamate, mp 87 –
88 ◦ C, water solubility 92 mg/L, LD50 (rat, oral)
74 mg/kg, has been used to control orchard pests.
Propoxur [114-26-1], 2-(1-methylethoxy)-
phenyl methylcarbamate, mp 91 ◦ C, vapor pres-
sure 0.84 mPa at 20 ◦ C, water solubility 2 g/L,
has rat LD50 values of 83 mg/kg (oral) and
> 1 000 mg/kg (dermal). Propoxur is used to
control household insects such as cockroaches,
bedbugs, and wasps, and as a replacement for
DDT in malaria control.
TBPMC [780-11-0], 3-tert-butylphenyl
methylcarbamate, mp 140 – 142 ◦ C, LD50
(mouse, oral) 470 mg/kg, has been used to con-
trol rice and fruit pests.
Insect Control 41
Trimethacarb [2655-15-4], 3,4,5-trimeth-
ylphenyl methylcarbamate, mp 117 – 119 ◦ C,
LD50 (rat, oral) 178 mg/kg, is used as a soil in-
secticide for corn.
7.3. Methylcarbamates of Oximes [74]
Aldicarb [116-06-3], 2-methyl-2-(methyl-
thio)-propanal O-[(methylamino)carbon-
yl]oxime, mp 99 – 100 ◦ C, vapor pressure
13 mPa at 25 ◦ C, water solubility 6 g/L, rat LD50
values 0.8, 0.65 mg/kg (oral) and 3, 2.5 mg/kg
(dermal) is the most toxic commercial insec-
ticide. Aldicarb is a broad-spectrum systemic
insecticide used for seed and soil treatment, and
as a nematocide.
Methomyl [16752-77-5], S-methyl-N-
[(methylcarbamoyl)oxy]thioacetimidate, mp
78 – 79 ◦ C, vapor pressure 9.1 mPa at 25 ◦ C,
water solubility 58 g/L, rat LD50 17, 26 mg/kg
(oral) and > 15 000 mg/kg (dermal), is a broad-
spectrum insecticide.
42 Insect Control
Oxamyl [23135-22-0], S-methyl 1-(dimeth-
ylcarbamoyl)-N-[(methylcarbamoyl)oxy]thio-
formimidate, mp 108 – 110 ◦ C, vapor pressure
20 mPa at 25 ◦ C, rat LD50 5.4 mg/kg (oral) and
2 960 mg/kg (dermal), is a systemic insecticide
and nematocide.
Thiofanox [39196-18-4], 3,3-dimethyl-1-
(methylthio)-2-butanone-O-[(methylamino)-
carbonyl]oxime, vapor pressure 22.6 mPa at
25 ◦ C, water solubility 5.2 g/L, LD50 (rat, oral)
8.5 mg/kg, is a systemic soil insecticide effective
against red spider mites.
7.4. Mode of Action [62], [69]
The insecticidal carbamates are cholinergic.
Poisoned insects and animals exhibit violent
convulsions and other neuromuscular distur-
bances. These insecticides carbamylate acetyl-
cholinesterase and may also have a direct ac-
tion on acetylcholine receptors because of their
pronounced structural resemblance to acetyl-
choline. The mechanism of interaction with
acetylcholinesterase is analogous to the normal
three-step hydrolysis of acetylcholine (cf. Sec-
tion 6.9.1). However, the third reaction step is
much slower for the carbamylated enzyme than
for the acetylated one. The importance of struc-
tural complementarity of the insecticidal carba-
mates to the active site of acetylcholinesterase
is demonstrated by the pronounced difference
in activities of d- and l-2-(sec-butylphenyl)
methylcarbamate (the l isomer is five times
more toxic) and of the 2-, 3-, and 4-substituted
phenyl methylcarbamates where the 4-isomers
are virtually inactive [68], [69].
Detoxification of carbamate insecticides oc-
curs in vivo through microsomal hydroxylation,
N-demethylation of carbamyl nitrogen, side
chain oxidation, and ring hydroxylation. Meth-
ylenedioxyphenyl synergists (Section 3.5.4)
prevent this largely by inhibiting the microso-
mal enzymes.
7.5. Environmental Toxicology
The N-methylcarbamates are generally
biodegradable and have low to moderate soil
persistence (one to four months). Aldicarb ox-
idizes readily to toxic CH3 SO and CH3 SO2
derivatives which persist in groundwater. Cer-
tain carbamates are highly toxic to birds with the
following oral LD50 values, in milligrams per
kilogram: carbofuran, mallard 0.40, pheasant
4.2; mexacarbate, mallard 3.0, pheasant 4.5; and
methomyl, mallard 16, pheasant 15; compared
to carbaryl > 2 000. Fish toxicity is generally
low, but carbamates are highly toxic to bees.
Carbamates with low mammalian LD50 values
(e.g., isolan, aldicarb, methomyl, oxamyl, car-
bofuran, and mexacarbate) have caused severe
poisoning in humans and domestic animals.
8. Formamidines [75–77]
The formamidine insecticides act as competitive
agonists of the neurotransmitter octopamine in
insects.
Chlordimeform [6164-98-3], N  -(4-
chloro-2-methylphenyl)-N,N-dimethylmethan-
imidamide, mp 35 ◦ C, bp 156 – 157 ◦ C (53.3 Pa),
d 25
4 1.105, vapor pressure 0.045 Pa at 20 C,
◦
LD50 (rat, oral) 238 mg/kg, is used primarily
against eggs and early growth stages of cotton
bollworm and rice stem borer larvae (instars). It
has both vapor and systemic action.
Amitraz [33089-61-1], N -(2,4-dimethyl-
phenyl)-N-{[(2,4-dimethylphenyl)imino]meth-
yl}-N-methylmethanimidamide, mp 86 – 87 ◦ C,
water solubility 1 mg/L, LD50 (rat, oral)

400 mg/kg, is used principally to control red
spider mites and pear psylla.
9. Insect Growth Regulators [78–85]
Most conventional insecticides are neurotox-
ins that produce biochemical lesions at spe-
cific target sites (e.g., acetylcholinesterase or the
nerve axon). Therefore, many organochlorines,
organophosphates, carbamates, and pyrethroids
lack desirable insecticide selectivity. Insect
growth regulators interfere with unique bio-
chemical processes that are peculiar to in-
sects: molting or ecdysis, and formation of
the chitinous exoskeleton; they are truly selec-
tive insecticides (see also → Biological Control,
Chap. 5.1.1.).
9.1. Juvenoids
Juvenoids are structurally optimized derivatives
of the insect juvenile hormones (JH) or neotenin:
JH I [13804-51-8], methyl 7-ethyl-9-(3-ethyl-
3-methyloxiranyl)-3-methyl-2,6-nonadienoate,
JH O [55333-14-7], (7-ethyl, 11-ethyl), JH II
[34218-616], (7-methyl, 11-ethyl), and JH III
[22963-93-5], (7-methyl, 11-methyl) have also
been isolated from various insect species.
The JH hormones are insecticidal when ap-
plied exogenously to the egg and last instar
larval stage, and interfere with the develop-
ment of physiologically competent adults. How-
ever, they are of short persistence, and a variety
of structurally optimized analogues have much
longer residual action and improved specificity.
Methoprene [40596-69-8], 1-methyl-
ethyl (E,E)-11-methoxy-3,7,11-trimethyl-2,4-
Insect Control 43
dodecadienoate, bp 100 ◦ C (6.7 Pa), vapor pres-
sure 3.1 Pa at 25 ◦ C, soluble in water to 1.4 mg/L,
has an LD50 (rat, oral) of > 34 000 mg/kg and
is effective against mosquito and fly larvae.
Hydroprene [41096-46-2], ethyl (E,E)-
3,7,11-trimethyl-2,4-dodecadienoate, wa-
ter solubility 0.54 mg/L, LD50 (rat, oral)
> 34 000 mg/kg, is especially effective against
aphids and in bait for cockroach control.
Epophenonane [57342-02-6] is 2-
ethyl-3-[3-ethyl-5-(4-ethylphenoxy)pentyl]-
2-methyloxirane. The LD50 (rat oral) is
4 000 mg/kg.
Fenoxycarb [79127-80-3], ethyl [2-(4-
phenoxyphenoxy)-ethyl] carbamate, mp 53 ◦ C,
vapor pressure 7.8 µPa at 20 ◦ C, water solubil-
ity 6 mg/L, has a broad spectrum of activity. Its
LD50 (rat, oral) is > 16 800 mg/kg.
9.2. Chitin Synthesis Inhibitors [85], [86]
Some insect growth regulators inhibit the forma-
tion of the chitinous insect exoskeleton and thus
produce a critical biochemical lesion at the time
of molting, pupation, or ecdysis.
44 Insect Control
Diflubenzuron [35367-38-5], N-{[(4-
chlorophenyl)amino]carbonyl}-2,6-difluoro-
benzamide, mp 239 ◦ C, is soluble in water
to 0.3 mg/L and has an LD50 (rat, oral) of
> 4 640 mg/kg. Diflubenzuron is the best known
compound with the same general structure that
inhibits the synthesis of chitin and is thus very
specific for the control of arthropods. Difluben-
zuron is an ovicide and a slow-acting broad-
spectrum insecticide.
10. Other Types of Insecticides
Hydramethylnon [67485-29-4], tetra-
hydro-5,5-dimethyl-2(1H)-pyridinone(3-[4-
(trifluoromethyl)-phenyl]-1-{2-[4-(trifluoro-
methyl)phenyl]ethenyl}-2-propenylidene)hy-
drazone, mp 189 – 191 ◦ C, is a slow-acting tox-
icant used successfully in bait for the fire ant,
other ants, and cockroaches.
Azadirachtin [11141-17-6], is the most ac-
tive of the growth regulators (limonoids), oc-
curring at 0.2 – 0.4 % in the seeds of the neem
tree, Azadirachta indica (Meliaceae). Neem ex-
tracts are wide-spectrum insecticides and they
act by blocking the action of the molting hor-
mone ecdysone [87].
11. Fumigants [1]
Fumigants are distributed to the target site as
gases and, therefore, must exist in air at a con-
centration that is lethal to the insect pest. This
physical requirement limits the insecticides that
can be employed usefully as fumigants. Com-
pounds that boil around room temperature are
the most generally useful and include hydrogen
cyanide, methyl bromide, methyl formate, and
ethylene oxide. For soil fumigation, the slower
release of vapor from ethylene dibromide, di-
bromochloropropane, and dichlorodiethyl ether,
which boil as high as 180 ◦ C, has proved effec-
tive. Organic compounds with higher vapor pres-
sures, such as 1,4-dichlorobenzene and naph-
thalene, sublime readily and have fumigant ac-
tion in tight enclosures such as museum cases
and cedar chests. Certain contact insecticides,
such as azobenzene, lindane, and dichlorvos,
may kill insects by vapor action especially when
volatilized by atomization, heat, and combustion
in pyrotechnic mixtures.
The major uses of fumigants are for insect-
infested mills, warehouses, grain elevators, gro-
ceries, museums, and habitations to control
structural pests; for stored products in bulk or
package; for soil fumigation of grubs, wire-
worms, ants, nematodes, etc.; and for living
plants in greenhouses and nursery stock.
The proper choice of fumigant for insect pest
control is determined not only by the effective-
ness of the fumigant gas, but also by safety to
humans, animals, and plants; flammability, pen-
etrability, and reactivity with furnishings; effect
on seed germination; and cost. Many of the com-
pounds listed below are described in detail else-
where in this encyclopedia.
Acrylonitrile [107-13-1], CH2 =CHCN, bp
78 ◦ C, liquid d 20
4 0.797, gas d 1.8, vapor pres-
sure 11 kPa at 20 ◦ C, TLV 20 ppm, is used to
fumigate mills and commodities.
Carbon disulfide [75-15-0], CS2 , bp
46.3 ◦ C, liquid d 20
4 1.263, gas d 2.6, vapor
pressure 41.9 kPa at 20 ◦ C, TLV-TWA 10 ppm,
MAK 10 ppm, has been used as a household
fumigant.

Carbon tetrachloride [56-23-5], CCl4 , bp
76 ◦ C, liquid d 20
4 1.595, gas d 5.3, vapor pres-
sure 12 kPa at 20 ◦ C, TLV-TWA 5 ppm, MAK
10 ppm, is a grain fumigant and carrier for other
fumigants.
Chloropicrin [76-06-2], Cl3 CNO2 , bp
112 ◦ C, liquid d 20
4 1.651, gas d 5.7, vapor pres-
sure 2.7 kPa at 20 ◦ C, TLV-TWA 0.1 ppm, MAK
0.1 ppm, has been used as a grain and soil fumi-
gant, but is an obnoxious lachrymator.
1,2-Dibromo-3-chloropropane [96-12-8],
CH2 BrCHBrCH2 Cl, bp 199 ◦ C, liquid d 20 4 very hazardous general fumigant.
2.08, vapor pressure 0.08 kPa at 20 ◦ C, was used
widely as a soil fumigant, but U.S. registration
was withdrawn because of its carcinogenicity.
2,2
-Dichlorodiethyl ether [111-44-4],
ClCH2 CH2 OCH2 CH2 Cl, bp 178 ◦ C, liquid d 20
4 Methyl bromide [74-83-9], CH3 Br, bromo-
1.222, gas d 4.9, vapor pressure 0.10 kPa, at
20 ◦ C, TLV 15 ppm, MAK 10 ppm, is a soil
fumigant.
1,1-Dichloro-1-nitroethane [594-72-9],
CH3 CCl2 NO2 , bp 111 ◦ C, liquid d 20
4 1.224,
vapor pressure 2.47 kPa at 20 ◦ C, TLV 10 ppm,
MAK 10 ppm, has been used to fumigate grain
and stored products.
1,2-Dichloropropane [78-87-5], propylene
dichloride, CH2 ClCHClCH3 , bp 95.4 ◦ C, liquid
d 20 ◦
4 1.159, vapor pressure 28 kPa at 20 C, TLV-
TWA 50 ppm, MAK 75 ppm, is a soil fumigant.
trans-1,3-Dichloropropene [542-75-6],
ClCH=CHCH2 Cl, bp 111 ◦ C, liquid d 20
4 1.224,
vapor pressure 2.47 kPa at 20 ◦ C, TLV-TWA
1 ppm, is a soil fumigant.
Ethylene chlorobromide [107-04-0],
ClCH2 CH2 Br, 1-bromo-2-chloroethane, bp
107 ◦ C, liquid d 20
4 1.689, vapor pressure 5.3 kPa
at 20 ◦ C, has been used as a soil fumigant.
Ethylene dibromide [106-93-4], BrCH2 CH2 Br,
1,2-dibromoethane, bp 131.6 ◦ C, liquid d 20 4 Sulfuryl fluoride [2699-79-8], SO2 F2 , bp
2.172, gas d 6.5, vapor pressure 1.0 kPa at 20 ◦ C,
has been used widely as a soil and commodity
fumigant, but its U.S. registration was canceled
because of carcinogenicity.
Insect Control 45
Ethyl formate [109-94-4], HCOOC2 H5 , bp
54 ◦ C, liquid d 20
4 0.917, TLV-TWA 100 ppm,
MAK 100 ppm, is used to fumigate packaged
foods.
Ethylene oxide [75-21-8], (CH2 )2 O, bp
10.7 ◦ C, liquid d 74 0.887, gas d 1.5, vapor pres-
sure 146 kPa at 20 ◦ C, TLV-TWA 1 ppm, is used
to fumigate packaged foods.
Hydrogen cyanide [74-90-8], HCN, bp
26 ◦ C, liquid d 20
4 0.688, gas d 0.9, vapor pres-
sure 84 kPa at 20 ◦ C, TLV-TWA 10 ppm, is a
β-Methallyl chloride [563-47-3], CH2 =
C (CH3 )CH2 Cl, bp 72 ◦ C, liquid d 20
4 0.925, is
no longer used.
methane, bp 4.3 ◦ C, liquid d 04 1.732, gas d 3.3,
vapor pressure 189.2 kPa at 20 ◦ C, TLV-TWA
5 ppm, MAK 5 ppm, is a widely used general
fumigant.
Methyl formate [107-31-3], HCOOCH3 , bp
32 ◦ C, liquid d 20
4 0.974, vapor pressure 83.2 kPa
at 20 ◦ C, TLV-TWA 100 ppm, is used to fumi-
gate dried fruit.
Naphthalene [91-20-3], C10 H8 , bp 218 ◦ C,
gas d 4.4, vapor pressure 0.01 kPa at 20 ◦ C, TLV-
TWA 10 ppm, is used to fumigate fabric and mu-
seum specimens.
1,4-Dichlorobenzene [106-46-7], ClC6 H4 Cl,
bp 173.4 ◦ C, mp 53 ◦ C, gas d 5.1, vapor pressure
0.1 kPa at 20 ◦ C, TLV-TWA 75 ppm, is used to
fumigate fabric and museum specimens.
Phosphine [7803-51-2], PH3 , bp − 87.4 ◦ C,
liquid d −90 0.746, gas d 1.2, TLV-TWA
0.3 ppm, is used as a grain fumigant, generated
from aluminum phosphide and ammonium car-
bamate in the presence of moisture.
− 55.2 ◦ C, liquid d 25
4 1.342, gas d 3.5, vapor
pressure 1601.3 kPa at 20 ◦ C, TLV-TWA 5 ppm,
is used in tent fumigation for drywood termites
and powder post beetles.
46 Insect Control
Trichloroacetonitrile [545-06-2], Cl3 CCN,
bp 85 ◦ C, liquid d 25
4 1.44, has been used as a
grain fumigant.
Trichloroethylene [79-01-6], ClCH=CCl2 ,
bp 86.7 ◦ C, liquid d 15
4 1.470, gas d 4.5, vapor
pressure 9.7 kPa at 20 ◦ C, TLV-TWA 50 ppm, is
a grain fumigant.
12. Microbial Insecticides [88–92]
Insects are attacked by numerous pathogens; ca.
450 viruses, 80 bacteria, 460 fungi, 250 pro-
tozoa, and 20 rickettsial insect diseases have
been identified. A number of these can be
adapted for mechanical dissemination as micro-
bial insecticides in the form of spores, micro-
bial toxins, or viral suspensions (→ Biological
Control, Chap. 2.3., → Biological Con-
trol, Chap. 5.1.2.,→ Biological Con-
trol, Chap. 5.1.3.,→ Biological Con-
trol, Chap. 5.2.1.,→ Biological Control,
Chap. 5.2.2.). These microbial insecticides are
being used increasingly widely for integrated
pest management programs because of their
selectivity and freedom from environmental
hazards and toxic residues. Possible disadvan-
tages are their instability in light and air, their
dependence upon suitable climatic conditions,
the precise timing of application for maximum
effectiveness, and their slow lethal action.
Bacillus thuringiensis insecticide is the crys-
talline endotoxin that is made from the bac-
terium as a fermentation product (→ Biological
Control, Chap. 2.3.). Commercial wettable pow-
ders are standardized against a reference strain
and are especially effective against lepidopter-
ous larvae such as the bagworm, cabbage looper,
and gypsy moth. Bacillus thuringiensis sub-
species isralensis is a related toxin that is es-
pecially effective against mosquito and blackfly
larvae; B. sphaericus is a spore-forming bac-
terium that produces a specific toxin against
mosquito larvae.
Bacillus popillae is a spore-forming bac-
terium that causes “ milky disease ” in the lar-
vae of the Japanese beetle Popillia japonica and
related species. The spores have been used for
many years as a standardized talc dust, con-
taining 108 spores per gram, for treatment of
turf to kill Japanese beetle larvae.
Avermectins and Ivermectin. The aver-
mectins are pentacyclic lactones isolated as
fermentation products of Streptomyces aver-
mitilis. Ivermectin is 22,23-dihydroavermectin.
These complex products (→Biological Control,
Chap. 5.1.3., →Biological Control) are broad-
spectrum acaricides, insecticides, and parasiti-
cides, and are effective in quantities of grams
per hectare for control of ants, mites, and other
pests.
Other microbial insecticides include the
nuclear polyhedrosis viruses (NPV) of the cotton
bollworms Heliothis, the armyworms Prodenia
and Spodoptera, the cabbage looper Trichoplu-
sia, and the sawfly Neodiprion. The spores of the
microsporidan Nosema locustae are marketed
for grasshopper control. The fungal spores of
Beauvaria bassiana are standardized and are
used in China to control the European corn
borer and in the Soviet Union to control the
Colorado potato beetle.
13. Genetic Control [93–98]
Genetic control refers to a variety of methods
for suppressing insect pests through manipula-
tion of the inheritance mechanisms of the popu-
lation.
The mass release of sterilized males, ster-
ile male release, has proved most success-
ful, but other techniques have been evaluated
including the use of chemosterilants and the
mass introduction of deleterious mutations (e.g.,
conditional lethals, cytoplasmic incompatibility,
chromosomal translocations, and meiotic drive;
(→ Biological Control, Chap. 5.2.2.).
13.1. Sterile Male Release
The release of relatively large numbers of males,
sterilized either by exposure to γ-radiation or by
chemosterilants, into a population of virgin fe-
males has reduced the number of fertile female
offspring, as shown in Table 3.
The most effective use of sterile male release
was in the eradication of the screwworm fly
Cochliomyia hominivorax from the southeast-
ern United States in 1959 by periodic release

Table 3. Rate of decrease in wild pest population after sustained release of sterile males a
Generation Wild female population
Unsuppressed Suppressed
Parental 1×106 1×106
F1 5×106 0.5×106
F2 25×106 131 580
F3 1.25×108 9 535
F4 1.25×108 c 50
a
After Knipling [93].
b
Each female produces five reproductive daughters.
c
Maximum sustainable population.
over 18 months of ca. 2×109 male flies, ster-
ilized with γ-radiation from cobalt-60, over an
area of 181 000 km2 . A similar program along
the Texas – Mexico border has substantially con-
trolled the screwworm over the past 25 years,
through the annual release of as many as 10×109
sterile males. Dramatic population reduction of
the Mediterranean fruit fly Ceratitis capitata
was also achieved in California in 1981 when
ca. 40×106 flies were released weekly. Ster-
ile male release has been used with limited
success to suppress isolated populations of the
codling moth Laspeyresia pomonella in the Pa-
cific Northwest and is proposed as an integral
part of a cotton boll weevil eradication scheme
in the southeastern United States.
Success in sterile male release programs de-
pends on several factors:
1) method for rearing large numbers of insects,
2) adequate dispersion of released sterile males,
3) a sterilization procedure that does not de-
crease male vigor and competitiveness in
mating,
4) a pest species that mates only once per life-
time or in which the sperm from sterile males
compete with those of fertile males, and
5) a low pest population density or a means of
reducing the population to a low level (e.g.,
by insecticide application).
Many species of insects do not conform to
these criteria, and the rearing and release of ster-
ile males can be a costly procedure.
Insect Control 47
Sterile male popu- Ratio of Number of
lation released sterile to fertile reproducing
males females b
9×106 9:1 100 000
9×106 18 : 1 26 316
9×106 68 : 1 1 907
9×106 942 : 1 10
9×106 180 000 : 1 0
13.2. Chemosterilants [99–101]
The availability of a variety of radiomimetic
(i.e., mimic the effect of γ-radiation) and an-
timetabolite cancer suppressive drugs that can
sterilize large segments of wild insect popula-
tions has led to many suggestions for their use
in genetic control programs.
Aziridines have a highly reactive ethylen-
imine group. They alkylate deoxyribonucleic
acid and cause sterility in both sexes of
house flies, mosquitoes, fruit flies, and
other insects. Compounds used experi-
mentally include TEPA [57-39-6], trieth-
ylenephosphoramide, mp 41 ◦ C, LD50 (rat,
oral) 37 mg/kg, and its analogue thio-
TEPA [52-24-4] (→ Cancer Chemotherapy,
Chap. 3.5.), triethylenethiophosphoramide,
mp 51.5 ◦ C. Other radiomimetic compounds
that have been investigated as chemosteri-
lants include apholate [52-46-0], 2,2,4,4,6,6-
hexakis(1-aziridinyl)-2,2,4,4,6,6-hexahydro-
1,3,5,2,4,6-triazaphosphorine, mp 147.5 ◦ C,
LD50 (rat, oral) 98, 113 mg/kg; and HEMPA
[680-31-9], hexamethylphosphoric triamide,
bp 105 – 107 ◦ C (1.5 kPa), d 20 4 1.03, LD50
(rat, oral) 2650, 3360 mg/kg. Busulfan
[55-98-1], 1,4-butanediol dimethanesulfonate,
CH3 SO2 O(CH2 )4 OSO2 CH3 , LD50 (rat, oral)
1.8 mg/kg, has been used in the mass produc-
tion of sterile insects. Unfortunately, all of these
radiomimetic compounds are mutagens and car-
cinogens, and produce degenerative changes in
human germ plasm at dosages of one-hundredth
or less of the LD50 . Their casual and indiscrim-
inate use in insect control is precluded.
48 Insect Control
Antimetabolites suggested for use as
insect sterilants include methotrexate
[59-05-2], amethopterin, N-{4-[(2,4-diami-
no-6-pteridinyl)methyl]methylamino}benzoyl-
l-glutamic acid, a folic acid antimetabo-
lite; and fluorouracil [51-21-8], 5-fluoro-2,4-
(1H,3H)-pyrimidinedione, a uracil antimetabo-
lite (→ Cancer Chemotherapy, Chap. 2.2.).
13.3. Host Plant Resistance
Breeding plants to incorporate genetic factors
that deter insect pest attack is a most impor-
tant technology for integrated pest management
(IPM). Genetic host plant resistance is pro-
duced by antixenosis, antibiosis, or by toler-
ance. Antixenosis (nonpreference) results from
the removal of kairomone attractants (see Sec-
tion 15.3) or alteration of morphological charac-
ters such as pubescence. Antibiosis results from
the production of deleterious allomones that de-
ter insect growth and development. Tolerance is
produced by enhanced plant vigor that results in
rapid repair of insect injury.
Genetic engineering provides opportunities
to produce host plant resistance through trans-
fer of genes for the production of antibiotic fac-
tors such as Bacillus thuringiensis (Bt) toxin
into the genomes of crop plants such as tobacco,
tomato, potato, corn, and soybean. The insecti-
cidal properties of the toxin should protect the
crop plant from damage by insect herbivores.
However, widespread utilization of this technol-
ogy may also provide increased opportunities for
the development of Bt resistance.
14. Repellents [102], [103]
Repellents are substances that protect humans
and domestic animals, plants, or products from
insect attack by making food or living conditions
unattractive or offensive to the insect. Repellents
usually have low toxicity and are rarely, if ever,
effective against a wide variety of insect pests.
They can sometimes be employed to advantage
in insect control where use of toxic insecticides
is impossible. Examples of the use of insect re-
pellents are
1) repellents against crawling insects (e.g., cre-
sote lines and dinitro-o-cresol barriers for
migrating chinch bugs, pentachlorophenol
and trichlorobenzene soil barriers for subter-
ranean termites, and sticky bands around tree
trunks to protect against gypsy moth larvae);
and
2) repellents against insect feeding (e.g., Bor-
deaux mixture to prevent the feeding of
leafhoppers, flea beetles, and psyllids on
plants; fly and mosquito repellents to
lessen the attack of blood-sucking insects,
mites, and ticks; treatment of logs with
pentachlorophenol to prevent attack of bark
beetles; and mothproofing treatment to pro-
tect woollen goods from attack by clothes
moths and carpet beetles).
14.1. Repellents for Plant-Feeding
Insects
Bordeaux mixture originated in France as a
spray to control downy mildew disease, caused
by the fungus Peronospora viticola of grapes
in about 1882 (→ Fungicides, Agricultural). Al-
though primarily a fungicide, Bordeaux mix-
ture is repellent to leafhoppers, psyllids, and flea
beetles when sprayed on leaves of grapes and
potatoes. The mixture is formed by mixing 3.6 –
4.5 kg of hydrated lime and 1.9 – 2.7 kg of cop-
per sulfate in 380 L of cold water to produce a
colloidal precipitate of tetracupric sulfate. The
colloid remains in suspension for several hours
and, when sprayed on foliage, covers the leaves
with a thin tenacious film that gradually pro-
duces soluble copper to give some residual toxic
effect.
Two fungicides that have repellent ef-
fects on the Japanese beetle Popillia japonica,
which feeds on foliage, are thiram [137-26-8],
bis(dimethylthiocarbamoyl) disulfide, mp 155 –
156 ◦ C, d 1.29, LD50 (rat, oral) 640 mg/kg, and
nabam [142-59-6], ethylenebis(dithiocarbamic
acid) disodium salt, LD50 (rat, oral) 395 mg/kg.
 Insect Control 49

N-Butylacetanilide [91-49-6], bp 277 –

281 ◦ C, mp 22 ◦ C, d 0.99, LD50 (rat, oral)
14.2. Repellents for Blood-Feeding 2 830 mg/kg, is used as a clothing treatment
Insects to repel fleas and ticks.
The requirements for effective repellents against
blood-feeding pests in order of importance are:
1) effective protection of the treated area for
several hours on all types of subjects and un-
der all climatic conditions; 2,3,4,5-bis(Butyl-2-ene)tetrahydrofurfural
2) complete freedom from toxicity and irrita- [126-15-8], bp 307 ◦ C, d 1.121, LD50 (rat, oral)
tion when regularly applied to human or an- 2 500 mg/kg, repels biting flies on cattle.
imal skin;
3) cosmetic acceptability including freedom
from unpleasant odor, taste, or touch, and
harmlessness to clothing;
4) protection against a wide variety of biting
mosquitoes, flies, fleas, mites and ticks; and Butopyronoxyl [532-34-3], butyl-3,4-di-
5) low cost and ready availability. hydro-2,2-dimethyl-4-oxo-2H-pyran-6-carbox-
Evaluation of tens of thousands of chemicals ylate, bp 256 – 270 ◦ C, d 25
4 1.052 – 1.060, LD50
as pest repellents have shown that few meet all (rat, oral) 7 800 mg/kg, is a mosquito and fly
requirements satisfactorily. The routine use of repellent.
repellents for arthropod vectors of human dis-
ease has been an important part of military hy-
giene to protect troops from such diseases as
malaria and scrub typhus. The use of fly repel-
lents on cattle and milk cows has been shown
to decrease weight loss and to improve milk
production in animals subject to severe annoy- Dibutyl adipate [105-99-7], bp 183 ◦ C
ance and blood loss from constant infestation. (1.9 kPa), d 0.965, LD50 (rat, oral)
The following repellents have found widespread 12 900 mg/kg, is a tick repellent.
practical application:

Benzyl benzoate [120-51-4], bp 323 ◦ C, mp

21 C, d 25
◦
4 1.12, LD50 (rat, oral) 1900 mg/kg, is
used as a clothing treatment for chigger mites
Trombicula spp. Dibutyl phthalate [84-74-2], bp 340 ◦ C,
d 20
4 1.045, LD50 (rat, oral) 21 000 mg/kg, is used
on clothing to repel chiggers.

Butoxypoly(propylene glycol) [9003-13-8],

400 and 800 molecular mass fractions, d 0.973 –
0.990, LD50 (rat, oral) 11, 200 mg/kg, is a fly
repellent for cattle.
50 Insect Control

Di-n-butyl succinate [141-03-7], bp

274 ◦ C, d 20
4 0.9768, LD50 (rat, oral)
8 000 mg/kg, is a fly repellent for cattle.
Application. Repellents are often applied to
clothing, gloves, and head nets, where their ac-
tion may remain effective for a week or more.
Application can be made by rubbing, spraying,
Deet [134-62-3], N,N-diethyl-3-methyl- or dipping the clothing in an emulsion of the re-
benzamide, bp 160 ◦ C (2.5 kPa), d 20 pellent. A repellent mixture of three parts each
4 0.996,
LD50 (rat, oral) 2 000 mg/kg, is a general- of benzyl benzoate, N-butylacetanilide, and 2-
purpose repellent that is used widely by outdoor ethyl-2-butyl-1,3-propanediol with one part of
recreationists. nonionic emulsifier is the standard clothing im-
pregnant (M – 1960) for the U.S. Armed Ser-
vices.

14.3. Repellents for Wood-Feeding

Insects
Dimethyl phthalate [131-11-3], bp 284 ◦ C, Wooden building materials have been pro-
d 25
25 1.189, LD50 (rat, oral) 8 200 mg/kg, tected against termites by treatment with
is a general-purpose repellent for flies and pentachlorophenol [87-86-5].
mosquitoes. Trichlorobenzene, bp 205 – 250 ◦ C, d 25 25
1.460, has been used as a soil treatment to re-
pel or poison termites around the foundation of
buildings. The technical product is a mixture
of the 1,2,3- [87-61-6] and 1,2,4- [120-82-1]
isomers.
2-Butyl-2-ethyl-1,3-propanediol [115-84-4],
mp 40 ◦ C, d 20 4 0.931, LD50 (rat, oral)
5 040 mg/kg, is an effective, persistent mosquito 14.4. Repellents for Fabric-Eating Pests
repellent.
Two major groups of insect pests can utilize
the keratinaceous proteins of wool and animal
hides as complete dietary sources and, thus,
are major pests of woolen clothes, rugs, uphol-
stered furniture, and museum specimens. The
clothes moths (family Tineidae) include four
2-Ethyl-1,3-hexanediol [94-96-2], “ 612,” species with worldwide distribution: the case-
bp 244 ◦ C, d 20 0.942, LD50 (rat, oral) making clothes moth Tinea pellionella, the web-
4
2 400 mg/kg, is a widely used mosquito repel- bing clothes moth Tineola bisselliella, the car-
lent. pet moth Trichophaga tapetzella, and the brown
house moth Hoffmannophila pseudospretella.
Carpet beetles (family Dermestidae) include the
black carpet beetle Attagenus piceus, along with
the carpet beetle Anthrenus scrophulariae, the
varied carpet beetle A. verbasci, and the furniture
carpet beetle A. flavipes. These fabric pests cause
Di-n-propylpyridine 2,5-dicarboxylate direct damage amounting to more than $ 109 an-
[136-45-8], bp 186 ◦ C (0.13 kPa), d 1.08, LD50 nually in the United States alone. A number of
(rat, oral) 6 230 mg/kg, is a fly repellent used on mothproofing agents are available that protect
cattle. woolen fabrics from attack by a combination of

repellent and toxic action. Ideally, such moth-
proofing materials should
1) be fixed into the cloth during dyeing by
chemical reaction with the wool protein,
2) be completely effective against fabric pests,
3) give protection for up to 15 years after re-
peated washing and dry cleaning, and
4) be nonvolatile, stable to light, odorless, col-
orless, and nontoxic to humans.
Sodium fluorosilicate [16893-85-9], alone or
with aluminum, magnesium, or ammonium
salts, is an effective mothproofing agent when
employed in 0.5 – 0.7 % aqueous solution with
0.3 % potassium alum and 0.03 % oxalic acid.
Such formulations may be applied by spraying
or dipping and cannot be removed by dry clean-
ing.
Formerly, DDT at 0.25 – 0.75 % and dield-
rin at 0.05 % of the cloth weight were applied
by spraying in a volatile petroleum solvent or
by dipping the cloth in an aqueous emulsion.
This protected the cloth against fabric pests
for months to years. However, because of en-
vironmental pollution problems with these per-
sistent bioaccummulative organochlorines, they
are no longer used. Several persistent synthetic
pyrethroids including permethrin, fenvalerate,
and fenpropanate are effective in situ mothproof-
ing agents.
The most widely used commercial moth-
proofing agents are the (polychloro-2-chlorome-
thylsulfamido) diphenyl ethers ( Eulan type) and
the chlorinated diphenyl ether urea sulfonic acid
(Mitin type). They are applied to woolens during
fabrication at about 1 – 3 % of the weight of the
cloth, and fixed by chemical reaction with the
wool protein.
Insect Control 51
15. Attractants [104–117]
With a growing need for more specific, less en-
vironmentally polluting pest control, the use of
specific insect attractants has become an impor-
tant part of control technology. Chemical at-
tractants can be used in conjunction with sim-
ple, inexpensive sticky traps for monitoring in-
sect pest populations in relation to the economic
threshold, for removal trapping of pests, for lur-
ing them to toxic bait, or for mating confusion.
Two general types of attractants have been stud-
ied widely: (1) the sex pheromones used in in-
traspecies chemical communication leading to
mating, and (2) plant kairomones used in inter-
species chemical communication for host plant
selection.
15.1. Sex Pheromone Attractants
Sex pheromone attractants have now been iden-
tified in several hundred species of insect
pests[113]. For the most part, they are volatile es-
ters, alcohols, and aldehydes that facilitate sex-
ual communication and mating over distances up
to several hundred meters and can be perceived
in nanogram to microgram quantities. In Lepi-
doptera, sex pheromones are typically produced
by the female in eversible glands at the tip of
the abdomen. A number of pheromones used in
commercial insect control are identified in Ta-
ble 4. The naturally occurring sex pheromones
are often complex mixtures, and maximum re-
sponse to synthetic mixtures depends on specific
blends. For example, the male European corn
borer Ostrinia nubilalis responds to a mixture of
the Z- and E-isomers of 11-tetradecenyl acetate.
Populations of the corn borer in the midwestern
United States respond optimally to a 97 : 3 iso-
mer blend, but those in the eastern United States
respond to a 3 : 97 isomer blend [113].
The male tobacco budworm Heliothis
virescens produces a pheromone blend of
(Z)-11-hexadecenal with traces of (Z)-
9-hexadecenal, (Z)-7-hexadecenal, (Z)-9-
tetradecenal [53939-27-8], (Z)-11-hexadecenol
[56683-54-6], hexadecanal [629-80-1], and
tetradecanal [124-25-4].
In the Coleoptera, sex pheromones are gener-
ally produced by the female, but in some species
52 Insect Control
Table 4. Sex pheromones (Lepidoptera) used in insect control

(e.g., cotton boll weevil Anthonomus grandis),
a four-component sex pheromone is produced
by the male and liberated in the feces (frass).
Each component is essential for effective aggre-
gation, and the mixture is used commercially as
grandlure [11104-05-5].
The sex pheromone of the female Japanese
beetle Popillia japonica is (R,Z)-5-(1-decenyl)-
dihydro-2-(3H)-furanone [64726-91-6]; that of
the western corn rootworm Diabrotica vir-
gifera virgifera and its close relative the north-
ern corn rootworm D. barberi is (2R,8R)- 8-
methyl-2-decyl propanoate [104487-05-0]. The
female sex pheromone of the southern corn root-
worm D. undecimpunctata is (R)- 10-methyl-2-
tridecanone [82621-53-2].
These beetle pheromones are used in trapping
to monitor pest populations.
Insect Control 53
15.2. Aggregation Pheromones [108],
[114]
Bark beetles (Coleoptera, Scolytidae) attack dy-
ing trees and bring about aggregation and mass
attack by using a blend of pheromones produced
by males and females, together with kairomones
produced by weakened trees. Three types of ag-
gregation pheromone blends have been used in
mass trapping of bark beetle pests.
In Dendroctonus brevicomis, the west-
ern pine beetle, oxygenated terpenoid
pheromones exo-brevicomin [20290-99-7],
(1R)-exo-7-ethyl-5-methyl-6,8-dioxabicy-
clo[3.2.1]octane (female pheromone), and
frontalin [28401-39-0], 1,5-di-methyl-(1S)-6,8-
dioxabicyclo[3.2.1]octane (male pheromone),
are produced in the hind gut of the beetles af-
ter contact with the tree kairomone terpenoid
myrcene [123-35-3].
In Ips typographus, the Norway spruce
beetle, boring males convert the tree ter-
penoid myrcene to the pheromone ipsdienol
[35628-00-3], 2-methyl-6-methylene-(S)-2,7-
octadien-4-ol. These beetles also produce the
pheromones 2-methyl-3-buten-2-ol [115-18-4]
and cis-verbenol [473-67-6], cis-4,6,6-tri-
methylbicyclo[3.1.1]hept-3-en-2-ol.
Scolytus multistriatus, the lesser European
bark beetle, is a vector of the pathogen Cerato-
cystis ulmi, which causes Dutch elm disease. The
beetle frass pheromones 4-methyl-3-heptanol
[14979-39-6] (female) and α-multistriatin
[59014-03-8], (1S)-endo, endo-5-ethyl-2,4-di-
54 Insect Control

methyl-6,8-dioxabicyclo[3.2.1]octane, (male)
are produced after exposure to the
tree terpenoid α-cubebene [17699-14-8],
[3aS-(3aα,3bβ,4β,7α,S)]-3a,3b,4,5,6,7-hexa-
hydro-3,7-dimethyl-4-methyl-ethyl-1H-cyclo-
penta[1,3]cyclopropa[1,2]benzene. The three-
component aggregation pheromone is called
multilure [60018-97-5].

Adult corn rootworms are attracted strongly

to a variety of kairomones, including in-
dole [120-72-9], estragole [140-67-0], and 4-
methoxycinnamaldehyde [71277-11-7] for the
western corn rootworm; eugenol and trans-
cinnamyl alcohol [104-54-1] for the northern
15.3. Kairomone Attractants
corn rootworm; and cinnamaldehyde [104-55-2]
for the southern corn rootworm.
Kairomone attractants are volatile chemicals re-
leased by plants that serve as major cues for
host plant selection by phytophagous insects.
Increasing numbers of kairomones have been
identified and found to be useful insect at-
tractants[109](see also → Biological Control).
Among the most successful are mixtures of
eugenol [97-53-0] with geraniol [106-24-1], or
phenylethyl propanoate [122-70-3], which are
used widely for monitoring new infestations of
the Japanese beetle Popillia japonica and for re-
moval trapping in orchards.

Attractants have proved especially effective

Other kairomone lures for Coleoptera include
caproic acid [142-62-1] for the green June beetle
Cotinus nitida; propyl 1,4-benzodioxane-2-car-
boxylate [24902-02-1] for the European chafer
Amphimallon majalis; and ethyl 3-isobutyl-2,2-
dimethylcyclopropanoate [33419-38-4] for the
coconut rhinoceros beetle Oryctes rhinoceros.
for the control of fruit flies (Diptera: Tephriti-
dae). Methyl eugenol [93-15-2], eugenol methyl
ether, is very strongly attractive for the oriental
fruit fly Dacus dorsalis and many closely related
species. Raspberry ketone [5471-51-2], 1-(4-hy-
droxyphenyl)-2-butanone, is a strong attractant
for the melon fly D. cucurbitae and a large group
of its close relatives. The 1-(4-acetyl)derivative,
cue-lure [3572-06-3], is more volatile and is
a synthetic kairomone especially effective for
monitoring and for bait traps. Methyl eugenol
and cue-lure have been used for male annihi-
lation programs for several Dacus spp. by ap-
plying them to fiberboard blocks containing the
volatile attractants at 15 g/ha and contact insec-
ticides such as malathion or naled at 1 g/ha.

Similarly, the Mediterranean fruit fly Cerati-
tis capitata is attracted strongly to the kairomone
trimedlure [12002-53-8], tert-butyl 2-methyl-
4-chlorocyclohexanoate. Mixtures of apple
volatiles including butyl 2-methylbutanoate,
propyl hexanoate, butyl hexanoate, hexyl
propanoate, and benzyl butanoate are used to
trap the apple maggot Rhagoletis pomonella;
dipropyl disulfide [629-19-6], C3 H7 S−SC3 H7 ,
is an attractant for the onion maggot fly Hylemya
antiqua; and allyl isothiocyanate [57-06-7], 3-
isothiocyanato-1-propene, for the cabbage mag-
got fly H. brassicae. Yellow jackets, Vespula spp.
(Hymenoptera: Vespidae), are trapped with hep-
tyl butyrate [5870-93-9], C3 H7 COOC7 H15 .
16. Insecticide Formulation
Effective use of any chemical for insect control
depends on its formulation into a product that
guarantees safety to the applicator, to plants, and
to domestic animals. Insecticides are commonly
formulated as dust, water-dispersible powder,
emulsive concentrates, solution, and bait. Prepa-
ration of such formulations involves the use of
accessory agents such as dust carriers, solvents,
emulsifiers, wetting and dispersing agents, stick-
ers, deodorants or masking agents, and attrac-
tants or feeding stimulants.
Dusts are the simplest formulations and gen-
erally contain low concentrations (1 – 20 %) of
Insect Control 55
the active ingredients, although ground botan-
ical preparations such as pyrethrum, derris, or
ryania may be used undiluted. Dust carriers in-
clude organic flours, sulfur, silicon oxides, lime,
gypsum, talc, pyrophyllite, bentonites, attapul-
gite, and volcanic ash. They are selected on the
basis of compatibility with the active ingredi-
ent (including pH, moisture content, and sta-
bility), particle size, abrasiveness, absorbability,
density, wettability, and cost. The mixture of ac-
tive and inactive ingredients is produced by op-
erations including solvent impregnation, fusing,
grinding, and milling. Particle diameter is usu-
ally 0.5 – 4.0 µm.
Granulars are pelleted mixtures containing
2 – 20 % active ingredients in a carrier of absorp-
tive clay, bentonite, or diatomaceous earth, with
particle diameter ranging from 250 – 420 µm
(40 – 60 mesh) to 420 – 840 µm (10 – 40 mesh).
Ground organic matter such as corn grits or corn
cobs is also used. Granulars are prepared by im-
pregnation of the carrier with a solution or slurry
of the active ingredient, and are used chiefly as
soil insecticides or mosquito larvicides.
Water-dispersible powders contain the ac-
tive ingredient in high concentration (15 –
95 %), with a dust carrier such as attapulgite
which wets and suspends well in water. From 1
to 2 % of a surfactant is usually added to improve
suspension and wetting of the carrier. Wettable
powders are used widely for spraying plants and
animals because of their relative safety.
Emulsives are emulsions of the active in-
gredient in water-immiscible organic solvents,
commonly at 15 – 50 %, with a small percent-
age of surfactant to promote emulsification,
wetting, and spreading. Factors influencing the
choice of solvent are solubility, safety to plants
and animals, volatility, flammability, compati-
bility, odor, and cost. The most commonly used
solvents are kerosene, xylene, and related pe-
troleum fractions; methyl isobutyl ketone; and
amyl acetate. Water emulsion sprays are used
widely for plant protection and household insect
control.
Slow-release formulations are especially
useful for compounded pheromone and
56 Insect Control
Table 5. Average organochlorine insecticide residues in U.S. inhabitants (1964) ∗
Insecticide Food, daily
intake, mg
DDT 0.037
DDE 0.024
Aldrin, dieldrin 0.008
Heptachlor, heptachlor epoxide 0.003
BHC 0.004
∗ Data from [121–123].
kairomone attractants, in which reduced volatil-
ity of the active ingredient is important and
decreased toxicity is essential. Formulations are
made with polymeric carriers such as starch,
cross-linked polyamides, carboxylated poly-
acrylates, or polyisobutylenes. Incorporation
of the active ingredient in poly(vinyl chloride)
(e.g., dichlorvos), or in hollow microfibers is
used to release the sex pheromone attractants
of the gypsy moth Porthetria dispar or the pink
bollworm Pectinophora gossypiella for male
confusion.
Bait. The use of bait-containing attractants
and feeding stimulants together with a minimum
amount of insecticide is a growing technology
for integrated pest management. Examples in-
clude dried ground apple (pomace) and carbaryl
for the black cutworm Agrotis ipsilon, soybean
oil and amdro for the imported fire ant Solenop-
sis invicta, cue-lure and naled for the melon fly
Dacus cucurbitae, and methyl eugenol and naled
for the oriental fruit fly D. dorsalis. The use of
bait greatly reduces the amount of toxic chem-
ical applied and improves the selectivity of the
insect pest management procedure.
17. Insecticide Residues in Foods
[118], [119]
17.1. Residue Persistence
Approximately 70 % of all insecticides are used
in agriculture where they are typically applied
directly to raw agricultural commodities for
plant and animal protection. Despite weathering
of crop residues during maturation and further
Human fat, Human blood, Human milk,
mg/kg mg/L mg/L
2.3 0.0068 0.022
8.0 0.0114 0.041
0.29 0.0014 0.0073
0.24 0.0008 0.0027
0.60 0.0031 0.0024
attenuation during food processing and prepa-
ration, measurable amounts of insecticides (i.e.,
pesticide residues) are detectable in food. A pes-
ticide residue is defined by the World Health Or-
ganization as “ any substance . . . resulting from
the use of a pesticide [that] includes any spec-
ified derivatives such as degradation products,
metabolites, and reaction products which are of
toxicological significance ” [120].
Residues of persistent pesticides in the hu-
man diet are absorbed readily, stored in hu-
man adipose tissue, and secreted in body flu-
ids such as milk [121–123]. The effects of 20
years of contamination of the U.S. diet by per-
sistent organochlorine insecticide residues are
summarized in Table 5. Withdrawal of these in-
secticides in the 1970s was followed by a de-
cline in the average residue in human adipose
tissue, as shown by comparison of surveys made
in 1970 and 1978: DDT(T) 7.5 vs. 3.6 mg/kg,
dieldrin 0.44 vs. 0.19 mg/kg, and β-BHC 0.44
vs. 0.19 mg/kg[124].
Residue Persistence Curve. The magni-
tude and duration of pesticide residues on raw
agricultural commodities depend on the dosage
applied, the chemical nature of the insecticide,
and the physical nature of the treated plant sur-
face. Standard technology involves determina-
tion of the content of insecticide residue, in
milligrams per kilogram, in produce at intervals
after application. Because of the curvilinear na-
ture of the results (Fig. 2A), estimation of the rate
of residue loss is difficult. For rapidly degrad-
able insecticides (most organophosphates and
carbamates), residue persistence follows first-
order kinetics. When plotted on a semilogarith-
mic scale (Fig. 2B), straight lines are obtained
whose statistical reliability can be determined
 Insect Control 57
Table 6. FAO/WHO acceptable daily intakes (ADI) and maximum residue limits (MRL) together with EPA residue tolerances for various
insecticides

Insecticide Trade name ADI, mg kg−1 d−1 MRL, mg/kg Tolerance, mg/kg

Organochlorines
Aldrin/dieldrin 0.0001 0 – 0.1 a
Chlordane 0.001 0.3 a
DDT 0.005 7a
Endosulfan Thiodan 0.0075 0.1 – 2
Endrin 0.0002
Heptachlor 0.0005 0 – 0.1 a
Lindane 0.0125 1–7
Methoxychlor Marlate 0.1 1.25 – 100
Chlorobenzilate 0.02 0.2 – 5
Dicofol Kelthane 0.025 0.1 – 30
Toxaphene 0.3 – 7 a
Chlorbenside 0.01
Tetradifon Tedion 1 – 100
Ovex Chlorfenson, Ovotran 0.01 3–5
Organophosphates
Acephate Orthene 0.0005 b 0.1 – 10 0.5 – 15
Azinphos methyl Guthion 0.0025 0.2 – 4 0.2 – 10
Bromophos Nexion 0.04 0.5 – 5
Carbophenothion Trithion 0.0005 0.2 – 2 0.1 – 10
Chlorfenvinphos Supona, Birlane 0.002 0.05 – 1
Chlorpyrifos Dursban, Lorsban 0.01 0.05 – 2 0.05 – 1.5
Coumaphos Co-ral, Asuntol 0.0005 c 0.1 – 10
Cruformate Ruelene 0.1 1
Demeton Systox 0.005 c 0.1 – 1 0.3 – 12
Dialifor Torak 0.003 c 1 – 40 3
Diazinon Basudin, Spectracide 0.003 0.5 – 2 0.1 – 60
Dichlorvos Vapona, DDVP 0.004 0.1 – 2 0.02 – 1
Dicrotophos Bidrin 0.05
Dimethoate Cygon, Rogor 0.002 b 0.05 – 2 0.002 – 5
Dioxathion Delnav 0.0015 0.1 – 5 0.14 – 4.9
Disulfoton Thiodemeton 0.002 0.1 – 5 0.1 – 12
Ethion 0.0005 b 0.1 – 2 0.1 – 14
Ethoprop Mocap 0.02 0.02
Fenaminphos 0.0003 b 0.05 – 0.5
Fenchlorphos Ronnel, Korlan, Nankor 0.01 0.01 0 – 10
Fenitrothion Sumithion, Folithion 0.003 b 0.05 – 2
Fensulfothion Dasanit 0.0003 0.02 – 0.1 0.02 – 5
Fenthion Baytex 0.001 0.05 – 2 0.01 – 18
Fonofos Dyfonate 0.1 – 0.5
Formothion 0.02 0.2 0.2
Isofenphos Amaze 0.0005 b 0.02 – 1
Malathion Cythion 0.02 0.5 – 8 8 – 135
Methamidophos Monitor, Tamaron 0.0006 1–2 0.1 – 10
Methidathion Supracide 0.005 0.2 – 5 0.2 – 6
Mevinphos Phosdrin 0.0015 0.05 – 1 0.24 – 4
Monocrotophos Azodrin 0.0006 0.05 – 1 0.05 – 5
Naled Dibrom 0.5 – 3
Omethoate Folimat 0.0003 0.05 – 2 0.5 – 3
Parathion, ethyl Alkron, Niron 0.005 0.5 – 1 0.1 – 5
Parathion, methyl Metacide 0.02 0.05 – 0.2 0.1 – 5
Phorate Thimet 0.0002 0.1 – 1 0.05 – 1.5
Phosalone Zolone 0.006 5 – 10 0.05 – 85
Phenthoate Cidial 0.003 1
58 Insect Control
Table 6. (Continued)

Insecticide Trade name ADI, mg kg−1 d−1 MRL, mg/kg Tolerance, mg/kg

Phosphamidon Dimecron 0.0005 0.05 – 0.5 0.01 – 1

Phosmet Imidan 0.02 0.05 – 15 0.1 – 4
Pirimiphos methyl Actellic 0.01 0.05 – 5
Thiometon Morphothion, Ekatin M 0.003
Trichlorfon Dylox, Dipterex 0.01 0.05 – 2 0.1 – 240
Temephos Abate 0.1
Stirofos Rabon, Gardona 0.1 – 110
Phoxim Volaton 0.001 0.05 – 2
Leptophos Phosvel 0.001 b
Triazophos Hostathion 0.0002 0.05 – 2
Carbamates
Aldicarb Temik 0.005 0.1 – 0.5 0.002 – 1
Bendiocarb Ficam 0.002 b
Carbaryl Sevin 0.01 0.1 – 2.5 0.5 – 100
Carbofuran Furadan 0.01 0.02 – 20
Ethiofencarb Crotenon 0.1
Methomyl Lannate 0.1 – 10
Oxamyl Vydate 0.03
Propoxur Baygon 0.02 0.05 – 5
Pirimicarb Pirimin 0.02
Pyrethroids
Allethrin Pynamin 2–4
Deltamethrin Decis 0.01 0.02 – 0.2
Permethrin Pounce, Ambush 0.05 0.1 – 1
Pyrethrins 0.04 1–2 1–3
Cypermethrin Ripcord 0.05 0.01 – 0.2
Fenvalerate Pydrin, Sumicidin 0.02 0.1 – 10
Miscellaneous
Chlordimeform Galecron, Fundal 0.01 0.5 – 25
Binapacryl Morocide 0.0025
Methoprene Altosid 0.01 – 1
Piperonyl butoxide 0.1 – 20
a
Cancelled. b Temporary. c Withdrawn. a Cancelled. b Temporary. c Withdrawn.

by linear regression analysis. With highly per- 17.2. Environmental Chemistry of

sistent insecticides such as organochlorines and
certain pyrethroids, residue persistence may fol-
low second-order kinetics due to initial photo-
chemical or hydrolytic effects and the formation
of persistent metabolites, as well as to gradual
compartmentalization of the residue in plant
waxes or sequestration in oil glands.
Insecticide Residues [125][126]
Exposure to air, light, and moisture, and bio-
chemical interactions with enzymes in plant
and animal products, cause insecticide residues
to undergo inactivating or activating reactions,
generally within periods of days or weeks.

Safe Interval. The linear, semilogarithmic

residue persistence curve (Fig. 2) provides ready
determination of the residue half-life (R1/2 ),
which is independent of application rate. From
such residue persistence curves, the required in-
terval after application for residue attenuation
to toxicologically safe levels (maximum residue
limits) can be determined easily and expressed
as a safe interval before harvest on the pesticide
label directions.
17.2.1. Inactivation
Inactivation processes result in the formation of
chemical residues that no longer exert specific
target site interactions (e.g., inhibition of acetyl-
cholinesterase by an organophosphate or carba-
mate). Inactivation processes typically attack the
chemical bonds of degradophores (i.e., chemi-
cal groups sensitive to hydrolysis, oxidation, re-
duction, or conjugation) so that the insecticide

structure is altered to form smaller, more water-
soluble moieties.
Figure 2. Residue persistence curves for carbaryl on corn,
Zea mays
A) Degradation of residue vs. time; B) Semilogarithmic
transformation [134]
Hydrolysis. Highly lipophilic, organochlo-
rine insecticides with strong C−Cl bonds are
not subject to rapid environmental hydrolysis,
which accounts for their lack of biodegradabil-
ity. On the other hand, the three C−O−P es-
ter links in organophosphate insecticides are hy-
drolyzed readily, which accounts for their gen-
eral biodegradability. Parathion, for example, is
Insect Control 59
hydrolyzed principally at the reactive P−O –
aryl bond to form diethylphosphate and 4-
nitrophenolate. More profound hydrolysis may
split P−O – alkyl bonds of organophosphates
to form partial phosphate esters. The reduced
toxic hazard associated with malathion is the re-
sult of hydrolysis at both carbethoxy and phos-
phorothiolate ester bonds.
Hydrolysis is also a major detoxification
pathway for carbamates; for example, carbaryl
is hydrolyzed to 1-naphthol and methylamine.
Oxidation occurs readily via light, atmo-
spheric oxidants, and microsomal processes cat-
alyzed by cytochrome P 450 enzymes in plant
and animal tissues where · OH is the initiating
radical. Common oxidation reactions include the
following [125]:
1) Desulfuration of phosphorothionates (P=S)
to phosphates (P=O)
2) Oxidation of thioethers (C−S−C) to sulfox-
ide (C−SO−C) and sulfone (C−SO2 C) (see
Section 6.9.1)
3) Ring hydroxylation of aromatic rings (e.g.,
carbaryl to naphthalene diols, permethrin and
fenvalerate to phenols)
4) Side chain oxidation of
CH3 → CH2 OH → COOH (e.g., in
pyrethrins, carbamates)
5) O-Dealkylation of aryl methoxy (CH3 O) to
phenol (HO) (e.g., in methoxychlor)
6) N-Dealkylation of NHCH3 to NH2 (e.g., in
carbamates)
7) Epoxidation of C=C characteristic of cy-
clodienes (e.g., aldrin to dieldrin, and hep-
tachlor to heptachlor epoxide)
Reduction commonly occurs only under
anaerobic conditions (e.g., reduction of the NO2
group of parathion to NH2 in the animal rumen).
Dehalogenation occurs by glutathione con-
jugation of organochlorines (e.g., DDT is con-
verted to DDE; lindane, to trichlorothiophenol).
Conjugation by enzymatic action enhances
water solubility and excretion in animals.
Phenols are conjugated with glucosides, glu-
curonides, sulfates, and phosphates; carboxylic
acids with glycine.
60 Insect Control
17.2.2. Activation
Activation is a special residue reaction in which
the first transformation produces an insecticide
derivative that is of equal or enhanced toxicity:
1) Epoxidation of cyclodienes (e.g., aldrin to
dieldrin and heptachlor to heptachlor ep-
oxide) occurs rapidly on plant surfaces, in
the soil, or in animal tissues. The epoxides
are more persistent and usually more toxic
than their precursors.
2) Phosphorothioates are oxidized to phos-
phates which are over a thousand fold
more reactive with the target site, acetyl-
cholinesterase (see Section 6.9.1).
3) Sulfides are oxidized readily to sulfoxides,
and the enhanced electron withdrawal acti-
vates the phosphorus atom of many alkylthio-
phosphates such as fenthion, demeton, disul-
foton, and phorate (see Section 6.9.1).
4) N-Dealkylation of carboxamide organophos-
phates enhances their toxicity (e.g., dicro-
tophos to monocrotophos).
17.3. Toxicology of Insecticide Residues
[128]
No Observed Adverse Effect Level
(NOAEL). Determination of the NOAEL for
a pesticide fed over the lifetime of at least two
and preferably three species of laboratory an-
imal (mouse, rat, dog) provides the basis for
risk assessment in predictive toxicology. The
NOAEL, in milligrams of pesticide ingested
per kilogram of body weight per day, is deter-
mined by lifetime (two-year) feeding studies
with a range of dosages. Evaluated factors in-
clude general health, food intake, weight gain,
blood chemistry, enzyme activity, organ weight,
and gross histopathology of organs and tissues.
The NOAEL is the highest dosage administered
that does not produce adverse (toxicological)
effects.
Acceptable Daily Intake (ADI) [120],
[127–129]. To determine the safe dosage for hu-
mans exposed to pesticide residues, the NOAEL
obtained from animal feeding studies is divided
by an additional safety factor to correct for both
interspecies and intraspecies variations. The re-
sult is the acceptable daily intake (ADI). Three
levels of safety factors are recognized:
1) a safety factor of 10 when valid chronic ex-
posure data for humans are available,
2) a safety factor of 100 when no valid human
exposure data exist, but adequate chronic ex-
posure data are available from studies with
laboratory animals, and
3) a safety factor of 1000 when no adequate
chronic exposure data exist.
The ADI is an estimate of the daily intake
of a pesticide, in milligrams per kilogram per
day, which over an entire human lifetime ap-
pears to be without appreciable risk. Leader-
ship in the establishment of ADI values is taken
by the Joint FAO – WHO Meeting on Pesticide
Residues, which periodically reviews and re-
vises ADI values as pertinent scientific data be-
come available. ADI values for many insecti-
cides are presented in Table 6.
Maximum Residue Limits, Residue Toler-
ances [120], [129–134]. The basic guidelines
for estimating the risk from human exposure to
pesticide residues in food are called the maxi-
mum residue limits (MRL) or tolerances. They
are established by various national and interna-
tional agencies. The MRL value represents the
smallest amount of pesticide that is practical and
toxicologically acceptable [128]. The MRL val-
ues (in parts per million) for agricultural com-
modities are determined as the residues resulting
after a pesticide has been used according to good
agricultural practice, given the minimum quan-
tities necessary to achieve adequate pest con-
trol. These definitions leave substantial latitude,
and differing MRL values have been established
for similar uses in various countries. However,
standardization is highly desirable because of
international trade in agricultural commodities,
and the Joint FAO – WHO Codex Committee
on Pesticide Residues has established interna-
tional MRL values for more than 100 pesticides
in > 1 600 specific foods. These values are re-
viewed and revised at intervals as new informa-
tion becomes available [131].
In the United States, registration and regula-
tion of pesticide tolerance are the responsibility
of the Environmental Protection Agency, under

the Federal Insecticide, Fungicide, and Roden-
ticide Act (1972) [132], [133]. This law requires
“ that a pesticide may be registered if it is effec-
tive, properly labeled, and when properly used
will not cause unreasonable adverse effects on
the environment taking into account the eco-
nomic, social and environmental costs and bene-
fits of the use of any pesticide.” The EPA residue
tolerance values for most insecticides are pre-
sented in Table 6.
Carcinogenicity. The Food Additive
Amendment (1958) to the U.S. Food, Drug,
and Cosmetic Act (1938) states “ that no addi-
tive shall be deemed to be safe if it is deemed
to cause cancer when ingested by man or an-
imal.” This concept has been used by EPA to
cancel registration and tolerance for organochlo-
rine insecticides such as DDT, aldrin, dieldrin,
chlordane, heptachlor, and toxaphene.
18. References
1. R. L. Metcalf, R. A. Metcalf: Destructive and
Useful Insects, 5th ed., McGraw-Hill, New
York 1993.
2. H. H. Cramer: “ Plant Protection and World
Crop Production,” Pflanzenschutz Nachr. 20
(1967).
3. J. W. Wright in R. L. Metcalf, J. J. McKelvey,
Jr. (eds.): The Future for Insecticides,
Wiley-Interscience, New York 1976, p. 17.
4. World Health Organization:
“ Arthropod-Borne and Rodent-Borne Viral
Diseases,” W.H.O. Tech. Rep. Ser. 719 (1985).
5. K. H. Büchel (ed.): Chemistry of Pesticides,
Wiley-Interscience, New York 1983.
6. D. E. H Frear: Chemistry of Pesticides, 3rd
ed., Van Nostrand Princeton, N.J., 1955.
7. N. N. Melnikov: Chemistry of Pesticides (Eng.
Transl.), Springer, New York 1971.
8. R. L. Metcalf: Organic Insecticides,
Interscience, New York 1955.
9. R. D. O’Brien: Insecticides: Action and
Metabolism, Academic Press, New York 1967.
10. W. Perkow: Wirksubstanzen der
Pflanzenschutz- und
Schädlingsbekämpfungsmittel, Parey, Berlin
1976.
11. H. H. Shepard: The Chemistry and Action of
Insecticides, McGraw-Hill, New York 1951.
Insect Control 61
12. E. Y. Spencer: Guide to the Chemicals Used in
Crop Protection, 7th ed., Research Branch
Agriculture, Canada, 1980.
13. R. Wegler (ed.): Chemie der Pflanzenschutz-
und Schädlingsbekämpfungsmittel, vol. 1,
1970; vol 3, 1976; Springer-Verlag,
Berlin-Heidelberg.
14. R. L. Metcalf, J. Agr. Food Chem. 21 (1973)
511.
15. A. W. A. Brown, J. Hayworth, A. R. Zahar, J.
Med. Entomol. 13 (1976) 1.
16. A. W. A. Brown, R. Pal: Insect Resistance in
Arthropods, WHO, Geneva 1971.
17. G. P. Georghiou, T. Saito: Pest Resistance to
Pesticides, Plenum Publishing, New York
1984.
18. G. P. Georghiou, A. Lagunes-Tejeda: “ The
Occurrence of Resistance to Pesticides in
Arthropods,” FAO Plant Protection Series,
Rome 1991.
19. National Academy of Sciences: Pesticide
Resistance: Strategies and Tactics for
Management, Washington, D.C., 1986.
20. R. L. Metcalf, Annu. Rev. Entomol. 25 (1980)
219.
21. R. L. Metcalf in M. Kogan (ed.): Ecological
Theory and Integrated Pest Management
Practice, Wiley-Interscience, New York 1986.
22. A. W. A. Brown: Ecology of Pesticides,
Wiley-Interscience, New York 1978.
23. C. A. Edwards (ed.): Environmental Pollution
by Pesticides, Plenum Publishing, New York
1973.
24. C. A. Edwards: Persistent Pesticides in the
Environment, 2nd ed., CRC Press, Cleveland
1973.
25. M. A. Q. Khan (ed.): Pesticides in Aquatic
Environments, Plenum Publishing, New York
1977.
26. F. L. McEwen, G. R. Stephenson: The Use and
Significance of Pesticides in the Environment,
Wiley-Interscience, New York 1979.
27. D. Pimentel: Ecological Effects of Pesticides
on Non-Target Species, President’s Office
Science, Technology, Washington, D.C., 1971.
28. T. R. Eichers, P. A. Andrilenas, T. W.
Anderson: “ Farmers’ Use of Pesticides in
1976,” USDA Agr. Econ. Rept. No. 418,
Washington, D.C., 1978.
29. J. F. Copplestone in D. L. Watson, A. W. A.
Brown (eds.): Pesticide Management and
Insecticide Resistance, Academic Press, New
York 1977, p. 147.
30. M. L. Flint, R. van den Bosch: Introduction to
Integrated Pest Management, Plenum
Publishing, New York 1981.
62 Insect Control
31. C. B. Huffaker (ed.): New Technology of Pest
Control, Wiley-Interscience, New York 1980.
32. M. Kogan (ed.): Ecological Theory and
Integrated Pest Management Practice,
Wiley-Interscience, New York 1986.
33. R. L. Metcalf, W. H. Luckmann (eds.):
Introduction to Insect Pest Management, 3rd
ed., Wiley-Interscience, New York 1994.
34. National Academy of Sciences: Scientific
Aspects of Pest Control, Publ. 1402,
Washington, D.C., 1966.
35. National Academy of Sciences: Insect Pest
Management and Control, Publ. 1695,
Washington, D.C., 1969.
36. National Academy of Sciences: Pest Control
Strategies for the Future, Washington, D.C.,
1972.
37. R. Carson: Silent Spring, Houghton-Mifflin,
Boston 1962.
38. R. L. Metcalf, Annu. Rev. Entomol. 25 (1980)
219.
39. R. L. Metcalf in R. L. Metcalf, W. H.
Luckmann (eds.): Introduction to Insect Pest
Management, 3rd ed., Wiley-Interscience,
New York 1994.
40. D. L. Watson, A. W. A. Brown: Pesticide
Management and Insecticide Resistance,
Academic Press, New York 1977.
41. M. Jacobsen, D. G. Crosby (eds.): Naturally
Occurring Insecticides, Marcel Dekker, New
York 1973.
42. J. E. Cassida (ed.): Pyrethrum, the Natural
Insecticide, Academic Press, New York 1971.
43. C. B. Gnadinger: Pyrethrum Flowers, 2nd ed.,
McLaughlin, Gormley, King, Co.,
Minneapolis 1945.
44. M. Elliott in R. L. Metcalf, J. J. McKelvey, Jr.
(eds.): The Future for Insecticides,
Wiley-Interscience, New York 1976, p. 163.
45. M. Elliott, Bull. W.H.O. 44 (1971) 315.
46. R. A. Fuchs, R. Schroder in K. H. Buchel (ed.):
Chemistry of Pesticides, Wiley-Interscience,
New York 1983.
47. T. Narahashi in R. L. Metcalf, J. J. McKelvey,
Jr. (eds.): The Future for Insecticides,
Wiley-Interscience, New York 1976, p. 371.
48. C. F. Wilkinson in R. L. Metcalf, J. J.
McKelvey, Jr. (eds.): The Future for
Insecticides, Wiley-Interscience, New York
1976, p. 195.
49. G. T. Brooks: Chlorinated Insecticides, vols.
1, 2, CRC Press, Cleveland 1974.
50. R. L. Metcalf in R. L. Metcalf, J. J. McKelvey,
Jr. (eds.): The Future for Insecticides: Needs
and Prospects, Wiley-Interscience, New York
1976, p. 223.
51. P. Müller (ed.): DDT Insektizid
Dichlordiphenyltrichloroäthan und seine
Bedeutung, Birkhäuser, Basel 1958.
52. J. E. Cassida, L. J. Lawrence, Environ. Health
Perspect. 61 (1985) 123.
53. G. Schrader: Die Entwicklung neuer
insektizider Phosphorsäure-Ester, 3rd ed.,
Verlag Chemie, Weinheim 1963.
54. C. Fest, K.-J. Schmidt: The Chemistry of
Organophosphorus Pesticides, 2nd ed.,
Springer-Verlag, New York 1980.
55. M. Eto: Organophosphorus Pesticides:
Organic and Biological Chemistry, CRC
Press, Cleveland 1974.
56. R. D. O’Brien: Toxic Phosphorus Esters,
Academic Press, New York 1960.
57. WHO Environmental Health Criteria 63:
Organophosphorus Insecticides, WHO,
Geneva 1986.
58. W. Lange, G. von Kreuger, Chem. Ber. 65
(1932) 1998.
59. R. L. Metcalf, Bull. Entomol. Soc. Am. 5
(1959) 3.
60. T. R. Fukuto, Bull. W.H.O. 44 (1971) 31.
61. T. R. Fukuto in C. F. Wilkinson (ed.): Pesticide
Biochemistry and Physiology, Plenum
Publishing, New York 1978.
62. R. D. O’Brien in C. F. Wilkinson (ed.):
Pesticide Biochemistry and Physiology,
Plenum Publishing, New York 1976, p. 271.
63. T. Namba, Bull. W.H.O. 44 (1971) 289.
64. S. D. Murphy in C. D. Klassen, M. O. Amdur,
J. Doull (eds.): Toxicology, 3rd ed., Macmillan
Publ. Co., New York 1986, p. 519.
65. R. L. Metcalf, Neurotoxicology 3 (1982) 269.
66. M. K. Johnson: “ The Delayed Neuropathy
Caused by Some Organophosphorus Esters:
Mechanism and Challenge,” CRC Crit. Rev.
Toxicol. 3 (1975) 289.
67. R. J. Kuhr, H. W. Dorrough: Carbamate
Insecticides, Chemistry, Biochemistry, and
Toxicology, CRC Press, Cleveland 1976.
68. R. L. Metcalf, T. R. Fukuto, J. Agric. Food
Chem. 13 (1965) 220.
69. R. L. Metcalf, Bull. W.H.O 44 (1971) 43.
70. WHO Environmental Health Criteria 64:
Carbamate Pesticides: a General Introduction,
Geneva 1986.
71. R. L. Metcalf, T. R. Fukuto, J. Agric. Food
Chem. 15 (1967) 1022.
72. H. Gysin, Chimia 8 (1954) 205.
73. M. J. Kolbezen, R. L. Metcalf, T. R. Fukuto, J.
Agric. Food Chem. 2 (1954) 864.

74. T. A. Magee in J. R. Coats (ed.): Insecticide
Mode of Action, Academic Press, New York
1982, p. 71.
75. C. O. Knowles in J. R. Coats (ed.): Insecticide
Mode of Action, Academic Press, New York
1982, p. 243.
76. R. M. Hollingworth, A. E. Lund in J. R. Coats
(ed.): Insecticide Mode of Action, Academic
Press, New York 1982, p. 190.
77. F. Matsumura, R. W. Beeman in J. R. Coats
(ed.): Insecticide Mode of Action, Academic
Press, New York 1982, p. 231.
78. V. B. Wigglesworth, J. Insect Physiol. 15
(1969) 73.
79. L. I. Gilbert (ed.): The Juvenile Hormones,
Plenum Publishing, New York 1976.
80. W. S. Bowers, Bull. W.H.O. 44 (1971) 381.
81. G. B. Stahl, Annu. Rev. Entomol. 20 (1975)
417.
82. J. P. Edwards, J. J. Menn in R. Wegler (ed.):
Chemie der Pflanzenschutz- und
Schädlingsbekämpfungsmittel, vol. 6, Springer
Verlag, Berlin 1981, p. 185.
83. N. Wakaboyashi, R. M. Waters in E. D.
Morgan, B. Mandava (eds.): Handbook of
Natural Pesticides, vol. 1, CRC Press, Boca
Raton, Florida, 1987.
84. E. P. Marks, T. Leighton, F. Leighton in J. L.
Coats (ed.): Insecticide Mode of Action,
Academic Press, New York 1982, p.283.
85. L. C. Post, B. J. deJong, W. R. Vincent, Pestic.
Biochem. Physiol. 4 (1974) 473. C. A.
Hendrick in J. L. Coats (ed.): Insecticide Mode
of Action, Academic Press, New York 1982,
p. 315.
86. A. Verloop, C. D. Ferrell in J. R. Plimmer (ed.):
Pesticide Chemistry for the 20th Century, Am.
Chem. Soc., Washington, D.C., 1977.
87. H. Schmutterer: The Neem Tree, VCH,
Weinheim1995.
88. E. A. Steinhaus: Insect Pathology: An
Advanced Treatise, vol. 1, 2, Academic Press,
New York 1973.
89. E. Kurstak (ed.): Microbial and Viral
Pesticides, Marcel Dekker, New York 1982.
90. C. M. Ignoffo in M. Jacobsen (ed.):
Insecticides of the Future, Marcel Dekker,
New York 1975.
91. R. A. Dyback in P. Doyle, T. Fujita (eds.):
Pesticide Chemistry: Human Welfare and the
Environment, Synthesis and Structure Activity
Relationships, vol. 1, Pergamon Press, Oxford
1982, p. 83.
92. C. C. Beegle, L. C. Lewis, T. Yamamoto in
C. M. Ignoffo (ed.): Handbook of Naturally
Insect Control 63
Occurring Pesticides: Microbial Insecticides,
CRC Press, Boca Raton, Florida, 1986.
93. E. F. Knipling: “ The Basic Principles of Insect
Population Suppression and Control,” USDA
Agricultural Handbook, no. 512, Washington,
D.C., 1979.
94. M. A. Hoy, J. J. McKelvey, Jr.: Genetics in
Relation to Insect Management, Rockefeller
Foundation, New York 1979.
95. A. H. Baumhover, JAMA J. Am. Med. Assoc.
196 (1966) 240.
96. R. Pal, M. J. Whitten (eds.): The Use of
Genetics in Insect Control, Elsevier,
Amsterdam 1974.
97. J. W. Wright, R. Pal (eds.): Genetics of Insect
Vectors of Disease, Elsevier, Amsterdam
1967.
98. R. Pal, L. E. LaChance, Annu. Rev. Entomol.
19 (1974) 269.
99. A. B. Borkovic: Insect Chemosterilants, vol. 7,
Adv. Pest Control Res., Wiley-Interscience,
New York 1966.
100. G. C. LaBrecque, C. N. Smith: Principles of
Chemosterilization, Elsevier, Amsterdam
1968.
101. D. G. Campion, Bull. Entomol. Res. 61 (1972)
577.
102. S. A. Hall, N. Green, M. Beroza, J. Agr. Food.
Chem. 5 (1957) 66.
103. C. N. Smith, “ Repellents for Anopheline
Mosquitoes,” Misc. Publ. Entomol. Soc. Am. 7
(1970) 99.
104. W. J. Bell, R. T. Carde (eds.): Chemical
Ecology of Insects, Sinauer Assoc.,
Sunderland, Mass., 1984.
105. M. Beroza: “Attractants and Repellents for
Insect Control,” in Pest Control Strategies for
the Future, Nat. Acad. Sci., Washington, D. C.,
1972.
106. W. H. Bossert, E. O. Wilson, J. Theor. Biol. 5
(1963) 443.
107. V. G. Dethier: The Physiology of Insect Senses,
Wiley-Interscience, New York 1963.
108. E. R. Mitchell (ed.): Management of Insect
Pests with Semiochemicals, Plenum
Publishing, New York 1981.
109. D. A. Nordlund, R. L. Jones, W. J. Lewis
(eds.): Semiochemicals: their Role in Pest
Control, Wiley-Interscience, New York 1981.
110. H. H. Shorey, J. J. McKelvey, Jr. (eds.):
Chemical Control of Insect Behavior,
Wiley-Interscience, New York 1977.
111. M. C. Birch: Pheromones, Elsevier,
Amsterdam 1974.
64 Insect Control
112. J. M. Brand, J. C. Young, R. M. Silverstein,
“ Insect pheromones: a critical review of recent
advances in their chemistry, biology, and
application,” Prog. Chem. Org. Natural Prod.
37 (1979) 1.
113. Y. Tamaki in G. A. Kerkut, L. I. Gilbert (eds.):
Comprehensive Insect Physiology,
Biochemistry, and Pharmacology, “ Sex
pheromones,” vol. 9, Pergamon Press, Oxford
1985, p. 145.
114. J. H. Borden in G. A. Kerkut, L. I. Gilbert
(eds.): Comprehensive Insect Physiology,
Biochemistry, and Pharmacology,
“ Aggregation pheromones,” vol. 9, Pergamon
Press, Oxford 1985, 257.
115. P. A. Hedin, F. G. Maxwell, J. H. Jenkins in
F. A. Maxwell, F. A. Harris (eds.): Proc.
Summer Inst. of Biol. Control of Plants,
Insects, and Diseases, “ Insect plant
attractants, feeding stimulants, repellents,
deterrents, and other related factors affecting
insect behavior,” University Press, Jackson,
Missisippi, 1974, p. 494.
116. R. L. Metcalf, E.. R.. Metcalf: Plant
Kairomones in Insect Ecology and Control,
Chapman & Hall, New York1992.
117. S. Finch in T. H. Coaker (ed.): Applied Biology
“ Chemical attraction of plant feeding insects
to plants,” V., Academic Press, New York
1980, p. 67.
118. F. A. Gunther, J. Gunther (eds.): Residue Revs.,
vols. 1 – 86, Springer Verlag, Hamburg
1962 – 1987.
119. H. Maier-Bode:
Pflanzenschutzmittel-Rückstände, E. Ulmer,
Stuttgart 1965.
120. N. A. Smart, Residue Rev. 98 (1985) 99.
Insect Repellents → Crop Protection
Insect Repellents → Insect Control
Instant Photography → Photography
121. W. F. Durham, Ann. N.Y. Acad. Sci. 160
(1969) 183.
122. R. E. Duggan, Ann. N.Y. Acad. Sci. 160 (1969)
173.
123. A. Curley, R. Kimbrough, Arch. Environ.
Health 18 (1969) 156.
124. T. Nakatsugawa, M. A. Morelli in C. F.
Wilkinson (ed.): Insecticide Biochemistry and
Physiology, Plenum Publishing, New York
1976, p. 61.
125. C. M. Menzie: Metabolism of Pesticides, U.S.
Dept. Interior, Spec. Sci. Rept. Wildlife no. 96,
1966 ; no. 184, 1974, no. 212, 1978.
126. H. Aizawa: Metabolic Maps of Pesticides,
Academic Press, New York 1982.
127. G. Vettorazzi: “ Toxicological Decisions and
Recommendations Resulting from the Safety
Assessment of Pesticide Residues in Food,”
CRC Crit. Rev. Toxicol. 4 (1975) 125.
128. WHO/FAO Tech. Rept. Ser. no. 370, 1963;
417, 1969; 458, 1970; 474, 1971; 502, 1972;
525, 1973; 545, 1974; 574, 1975.
129. U.S. Environmental Protection Agency:
Compendium of Registered Pesticides,
Washington, D.C.
130. L. P. Van Dyk, I. H. Wiese, J. E. C. Muller,
Residue Rev. 82 (1984) 37.
131. G. Vettorazzi: International Regulatory
Aspects of Pesticide Chemicals, vols. 1, 2,
CRC Press, Cleveland, Ohio, 1979.
132. R. A. Merrill in C. D. Klassen, M. O. Amdur, J.
Doull (eds.): Toxicology, 3rd ed., McMillan
Publ. Co., New York 1986, p. 917.
133. National Academy of Sciences: Regulating
Pesticides, Washington, D.C., 1980.
134. P. Dahl, L. Rattan: J. Assoc. Off. Anal. Chem.
54 (1971) 1095.
 Insulation, Electric 1

Insulation, Electric
Adolf Kamm, Prüfinstitut für elektrisches Verhalten, Leverkusen, Federal Republic of Germany (Chaps. 1 and
2 in part, 3.2, 3.3, and 4)
Karl Heinz Schüller, Fachhochschule Nürnberg, Fachbereich Werkstofftechnik, Nürnberg, Federal
Republic of Germany (Chaps. 1 and 2 in part, Section 3.1)

1. Introduction . . . . . . . . . . . . . . 2 3.1.2.1. Porcelain . . . . . . . . . . . . . . . . . 15

2. Properties and Testing Methods . 3 3.1.2.2. Steatite . . . . . . . . . . . . . . . . . . 17
2.1. Resistance to Tracking . . . . . . . 4 3.1.2.3. Materials for Electrical Heat Engi-
2.2. Volume Resistivity . . . . . . . . . . 5 neering . . . . . . . . . . . . . . . . . . 19
2.3. Surface Resistivity . . . . . . . . . . 6 3.1.2.4. Alumina Ceramics . . . . . . . . . . . 20
2.4. Relative Permittivity . . . . . . . . 7
3.1.2.5. Special Oxide Ceramics . . . . . . . 21
2.5. Dielectric Dissipation Factor . . . 7
2.6. Loss Index . . . . . . . . . . . . . . . 8 3.1.2.6. Insulating Materials Containing Ti-
2.7. Breakdown Voltage . . . . . . . . . 8 tanium Oxide and Titanates . . . . . 22
2.8. Further Electrical Tests . . . . . . 9 3.1.2.7. Nonoxide Ceramics . . . . . . . . . . 25
2.9. Mechanical and Thermal Proper- 3.1.2.8. Glass . . . . . . . . . . . . . . . . . . . 25
ties . . . . . . . . . . . . . . . . . . . . 11 3.2. Organic Insulating Materials . . . 26
3. Insulating Materials . . . . . . . . . 12 3.2.1. Thermoplasts . . . . . . . . . . . . . . 27
3.1. Inorganic Insulating Materials . . 12 3.2.2. Thermosets . . . . . . . . . . . . . . . 37
3.1.1. Insulating Materials Based on Natu- 3.2.3. Elastomers . . . . . . . . . . . . . . . . 40
ral Minerals . . . . . . . . . . . . . . . 13
3.2.4. Polymer Processing . . . . . . . . . . 42
3.1.1.1. Mica and Related Products . . . . . 13
3.1.1.2. Asbestos and Related Products . . . 14 3.3. Liquid Insulating Materials . . . . 44
3.1.2. Synthetic Inorganic Insulating Ma- 4. Selection of an Insulating Material 45
terials . . . . . . . . . . . . . . . . . . . 14 5. References . . . . . . . . . . . . . . . 48

1. Introduction Within the field of electrotechnics, noncon-

All nonmetallic organic and inorganic materi-
als are semiconductors. They provide a vastly
greater resistance towards a flowing electric cur-
rent than metallic conductors. Their resistivity
may be 27 powers of ten greater than that of a
conductor.
The differing electrical conduction behavior
of conductors and semiconductors is a result of
their atomic structure. In metallic conductors,
the electric current results from the movement of
free electrons. In contrast, the extremely small,
but nevertheless existant conductivity of the so-
called nonconductors arises from electrons and
ions which migrate under the influence of elec-
tric fields. Since dissociation induced by cosmic
radiation always generates a significant level of
freely moving electrons and ions, an absolute
nonconductor is not possible.
ductors (insulators) are generally employed to
insulate the current-carrying components of
conducting circuits from their surroundings. In-
sulators should prevent contact between conduc-
tors and energy losses caused by stray tracking
currents. The function of insulators therefore is
to prevent the undesirable flow of current be-
tween two conductors subjected to a potential
difference. Insulators must possess a maximum
electrical resistance under all the conditions to
which the system can be exposed, e.g., at ele-
vated temperature (electrothermal applications),
at high voltages (energy technology), or high fre-
quency (telecommunications). Insulation can be
improved by appropriate design of the insula-
tor. The choice of a suitable insulating material,
however, remains critical for successful insula-
tion. If insulating materials are used to fill the
free space between the electrodes of capacitors,

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 321
2 Insulation, Electric
then insulators act as dielectrics and have the
additional function of storing electrical energy.
Not all materials of high resistivity (see Sec-
tion 2.2) are suitable for use as electrical insulat-
ing materials. Additional chemical, mechanical,
thermal, and, depending upon the application in-
volved, further properties are required. Stabil-
ity toward various electrical and nonelectrical
influences [e.g., toward electrical discharges in
cavities and at interfaces (see Section 2.8), to-
ward heat produced by dielectric dissipation (see
Section 2.5) and toward ignition and burning]
is of major importance. The working life of an
insulated electric circuit depends mainly upon
the thermal aging behavior. Although a mate-
rial may be technically suitable as an insulator,
unfavorable processing properties or the manu-
facturing price can restrict its applications.
The selection of an insulating material lies in
the hands of the designer, provided the proper-
ties necessary for the function in question are
guaranteed. To ensure this, materials are tested
according to national standards or IEC publica-
tions. The results of these tests permit compari-
son of the insulating materials under the speci-
fied conditions. Data for a specific material can-
not generally be applied to an insulating compo-
nent because the component and its insulation
properties are influenced by design and manu-
facturing processes. The final decision regarding
functionality must be made by testing the com-
ponent itself.
Insulating materials can be subdivided into
inorganic and organic materials. Inorganic insu-
lators include natural products such as mica or
asbestos. Synthetic insulators made of ceram-
ics, glass, or more recently glass ceramics are of
greater significance, however. Organic insulat-
ing materials are primarily synthetic polymers.
All of these insulating materials have a secure
position within the electrical industry, because
of their special material properties or the ef-
ficient processes used to manufacture compo-
nents from them. Developments have afforded
new materials with improved properties. Devel-
opments within the electronics and electrical in-
dustries make the availability of new improved
insulating materials essential
In their original form, natural materials are
seldom able to meet the diverse requirements
demanded of an electric insulating material;
however, purification, preparation, or chemical
modification, often increase their value. Poly-
mers (thermoplasts, thermosets, and elastomers)
are mainly synthetically manufactured products.
Electric insulating materials are generally in-
dustrial products of complex chemical struc-
ture which have been adapted to suit their in-
tended application by treatment, modification,
and blending.
Compared with metallic conductors and mag-
netic substances, insulating materials represent
the most comprehensive material group within
the field of electrotechnics and possess the great-
est number of technically interesting properties.
Nonelectrical properties have been considered
here only to an extent sufficient to permit a brief
characterization.
Groups of organic insulating materials are
listed in Chapter 3 according to their chemi-
cal composition. In most cases, organic products
based on rubber or other polymers can be modi-
fied by processing. Hence an absolute value for
a property of a given insulator group does not
exist. For this reason, data are not presented in
tabular form but as performance relief diagrams.
The values shown in each block diagram merely
indicate the range of the properties of main in-
terest. Organic products which find application
in the electrotechnics industry are generally em-
ployed with additives, which can often consid-
erably influence electrical and dielectric proper-
ties. The most common additives include pro-
cessing aids, stabilizers, pigments, fillers, and
flame retardants.
2. Properties and Testing Methods
The electrical behavior of electrical insulating
materials results from the physical processes of
dissociation, polarization, and ionization.
Certain assumptions about the formation of
free electrons and bound charge carriers un-
der the influence of electrical fields are asso-
ciated with these terms. They are responsible
for electrical resistance, dielectric polarization,
and electrical breakdown. In the broadest sense,
these measurable parameters can be regarded as
physical properties. Electrical insulating mate-
rials are materials with a resistivity >106 Ω · m
(>108 Ω · cm). The energy values which the
electrons in a solid are able to assume can be
 Insulation, Electric 3

classified in an energy step scheme known as path [23]. The international report of the Interna-
the “band model”. Various energy levels exist: tional Electrotechnical Commission (IEC Pub-
lication 664/664 A) classifies insulation materi-
1) energy levels which are occupied by elec-
als for low-voltage technology into four groups,
trons in the ground state of the atom;
according to their comparative tracking indices
2) energy levels which are accessible to elec-
(CTI) [24]. The electric strength and the insula-
trons after introduction of energy (excited
tion resistance are disregarded, but they should
state); and
be in a range appropriate for an insulating ma-
3) energy levels which cannot be occupied be-
terial.
cause of the Pauli exclusion principle.
Insulating material I: CTI = 600
The description of conducting processes in Insulating material II: CTI = 400 – 575
a solid material involves only the outermost Insulating material III a: CTI = 175 – 375
Insulating material III b: CTI = 100 – 150
electron-filled energy levels (the basic or va-
lence band) and the next highest permissible en-
ergy range. In an insulating material, the valence
band is completely filled with electrons, whereas
the conduction band is empty. Following the in-
troduction of energy, for example in the form
of thermal energy, electrons are excited from
the valence band and enter the conduction band.
Here, they possess increased mobility and con-
sequently lead to electrical conductance. In this
way, an insulating material can attain a semicon-
ductor character at elevated temperature.
In an insulating material, a prohibited energy
zone occurs between the valence and conduction
bands which is so large that an activation energy
of >2 eV is required for an electron to make the
transition. For this reason, practically no elec-
trons occupy the conduction levels at room tem-
perature.
Test procedures exist for reproducing or sim-
ulating technical conditions. The principles of
these tests are often not evident, but can also be
traced back to physical processes. The most im-
portant process in low-voltage technology is that
used to determine resistance to tracking. Addi-
tionally, methods are used to determine promo-
tion of corrosion, to observe electrostatic charg-
ing behavior, to investigate behavior towards
partial discharges, and to assess the damage
caused by arcing. Studies of thermally stimu-
lated discharge currents provide further physical
information [22].
Figure 1. Diagram for determining the comparative tracking
index (CTI) according to ASTM D 3638/UL 764 A
2.1. Resistance to Tracking Curves are shown for two materials of differing behavior but
with a CTI of 325.
Tracking is the progressive decomposition of the
surface of a solid insulating material caused by The international standard IEC Publication
local discharge and subsequent development of 112-1979 stipulates conditions for measuring
a conducting or partially conducting tracking
4 Insulation, Electric
resistance to tracking. The United States stan-
dard ANSI/ASTM 3638-1977 [25] forms the ba-
sis for tests according to the testing regulations
ANSI/UL 746 A-1984 [26]; these differ from the
currently valid international standard (IEC Pub-
lication 112-1979) in the type of evaluation pro-
cedure employed and in the lack of opportunity
for using a more aggressive testing solution (B).
During testing, surface contamination is sim-
ulated by allowing an appropriate test fluid to
drip onto the surface. The IEC Publication 112
specifies five separate determinations of a.c.
voltage (50 – 60 Hz) at which 50 drops of test
solution between the platinum electrodes do not
lead to failure. In contrast, ASTM 3638-1977
stipulates one measurement at a number of test
voltages and comparison with a standard curve.
The point of intersection of this curve with the
50-drop line is determined on the plot of droplet
number as a function of the testing voltage; the
test voltage is read off the abscissa (voltage axis)
and the CTI is thus determined (Fig. 1).
Figure 2. Schematic representation of measurement of low-
voltage tracking
a) Current limiting resistor; b) Excess current switch is acti-
vated by relay (c) when I ≥ 0.5 A for 2 s; c) Relay; d) Con-
tamination fluid; e) Platinum electrode; f) Specimen; g) Volt-
age supply
Test solution A comprises 1 g of ammonium
chloride dissolved in 1 L of deionized water. The
concentration of ammonium chloride in testing
solution B is identical, but a specified amount
of a given wetting agent is also added. The test-
ing solutions therefore differ in their electrolytic
conductivity because the wetting agent intro-
duces almost the same contribution to the to-
tal conductivity as the ammonium chloride salt.
The testing solutions also differ in their surface
tension.
According to the definition given in the stan-
dards, a track has developed when the tracking
current on the tracking path is greater than or
equal to 0.5 A for at least 2 s, or when the spec-
imen begins to burn as a result of the heat gen-
erated. A circuit used for evaluating tracking is
shown in Figure 2.
The test is carried out at voltage steps be-
tween 100 and 600 V. The maximum test voltage
which can be divided by 25 without a remainder
represents the comparative tracking index CTI,
provided that the test specimen (which must be
at least 3 mm thick) can also withstand a fur-
ther five tests with 100 drops at a voltage 25 V
lower than the maximum value. Should a sample
not fulfill this requirement, the voltage must be
reduced in steps of 25 V and the value must be
determined at which 100 drops can be withstood
without failure. The magnitude of this voltage is
given in parentheses following the CTI.
To avoid the incorrect characterization of
strongly eroding materials, the standard recom-
mends that the maximum depth of erosion from
the five evaluated results of the tracking path for-
mation is determined and indicated along with
the CTI. This also applies when no tracking
path is generated, e.g., due to sample ignition
or when, although tracking paths are formed at
a test voltage of 600 V, a current of 0.5 A is not
exceeded.
2.2. Volume Resistivity
Volume resistance is obtained by dividing the
d.c. voltage between two electrodes placed on
two opposite faces of a specimen by the steady-
state current between these electrodes. The vol-
ume resistivity  is defined as the volume resis-
tance relative to a cube with unit volume [27].
When a d.c. voltage is placed across two
electrodes which are in intimate contact with a
specimen (Fig. 3), the current passing through
the sample decreases asymptotically towards a

constant end value. The drop in current can be
caused by dielectric polarization and by mi-
gration of highly mobile ions toward the elec-
trodes. The insulation resistance is generally de-
termined 1 min after the beginning of the elec-
trification.
Figure 3. Circuit diagram for measuring volume resistivity
a) Voltage supply; b) Switch; c) Voltmeter; d) Shielded am-
meter; e) Specimen; f) Electrode
The current flowing along the surface, and the
current arising from the polarization phenom-
ena on the electrodes during careful and exact
measurements must be excluded from the mea-
surement. This is achieved by using appropriate
circuits and measurement procedures.
Electrodes used for these tests are metal-like
conductors or metallic conducting materials of
suitable shape, size, and design; they are placed
in intimate contact with the specimen and form
the transition regions between the electron cur-
rent and the ion-mediated current flow in the in-
sulator. With liquid insulators, the electrodes are
provided by the completely wetted container in
which the measurements are being carried out
(fluid electrode). With solid insulators, the elec-
trodes must be adhered to the material (Fig. 4).
Even under pressure, placing metallic compo-
nents of dimensions stipulated in the standards
on or beneath the sample material is not ade-
quate in order to determine publishable results.
None of the insulating materials has a perfectly
flat surface. Therefore, in extreme cases, disk-
shaped electrodes only make defective contact
at three elevations on the sample surface.
The international standard testing process is
specificied in IEC Publication 93 [27] for solid
materials and IEC Publication 247 [28] for liq-
Insulation, Electric 5
uid insulators. Recommendations for measuring
the insulation resistance at elevated temperature
are given in IEC Publication 345 [29].
Figure 4. Top view (A) and side view (B) of volume re-
sistivity measurement in a solid specimen using adhering
electrodes
a) Sample; b) To voltage supply; c) Unprotected electrode;
d) Guard electrode; e) Guard gap; f) Connection for amme-
ter; g) Protected electrode
2.3. Surface Resistivity
The electrical insulation behavior of an insula-
tor is influenced considerably by the nature of
its surface, as well as by adsorbed or contam-
ination layers. Conduction across the surface
produced by these effects is known collectively
as surface conductance. To decrease this prop-
erty, the length of the conduction pathway can
be increased, for example by mounting ribs or
shielding on the insulator. The removal of con-
tamination layers is simplified by using a smooth
surface, which is achieved in ceramic insulating
components by glazing.
The contribution of adsorbed and contamina-
tion layers to surface conduction illustrates that
this parameter is not a material property. How-
ever, the deposition and enrichment of adsorbed
6 Insulation, Electric
layers is also influenced by the nature and struc-
ture of the surface of the material. For example,
in inorganic insulating materials the formation
of adsorbed water layers is related to the phase
characteristics of the surface: the polished sur-
faces of ceramic insulator components often ex-
hibit a greater surface conductance than the fired,
glass-enriched surface. A significant increase in
surface conductance can be observed with in-
creasing air humidity.
Surface resistance is obtained by dividing the
d.c. voltage between two electrodes in contact
with the surface of a sample by the current flow-
ing between these electrodes 1 min after electri-
fication (see Section 2.2). The surface resistivity
σ is the surface resistance expressed relative to a
square area. The length of the sides of this square
is, in theory, immaterial [27].
Generally, current mainly flows through the
surface layer of the specimen and through the
moisture and surface contamination associated
with it. However, the measured surface resis-
tance also includes a nondefinable contribution
from the resistance of the bulk of the sample.
Thus, surface resistance cannot be measured
precisely but only approximately. The measured
value is generally an indication of the state of
the surface at the time of measurement. The rel-
ative permittivity (see Section 2.4) of the sample
under investigation affects the form and extent
of contamination [27]. The current which flows
as a result of contamination is, however, also
influenced by the surface structure and surface
properties of the specimen. Hence the surface
resistivity is not a material property in the nor-
mal sense, but it can be regarded as such if the
effects of contamination are taken into consid-
eration, e.g., by appropriate preconditioning.
The test procedure has been specified by the
international standard IEC 93 [27].
2.4. Relative Permittivity
Polarization leads to the internal charging of the
material in a direction opposing that of the ex-
isting field. This causes a reduction of the field
strength within the capacitor and a correspond-
ing increase in the capacity when the volume
between the electrodes is completely and exclu-
sively filled with an insulating material as op-
posed to being empty. The ratio between the ca-
pacity of a capacitor filled with a dielectric and
that of a capacitor of identical dimensions but
incorporating a vacuum between the electrodes
is known as the relative permittivity εr of the di-
electric. Under d.c. conditions, the relative per-
mittivity is characteristic for most materials, but
is temperature dependent. Under a.c. conditions,
it moreover depends on frequency and in certain
cases on field strength.
The relative permittivity is a dimensionless
comparative value. The procedure for determin-
ing capacitance is given in the international stan-
dard IEC Publication 250 [30].
Insulating materials can be divided into three
groups according to relative permittivity:
1) With most materials, the positive and nega-
tive charge centers coincide. Under the in-
fluence of an electric field, electron polar-
ization, and in some cases ion polarization,
occur. Such insulating materials possess low
relative permittivities of 2 – 10.
2) In the second group, a dipole moment is al-
ready present because of the molecular or
crystal structure. In addition to the electron or
ion polarization, the external field introduces
orientation polarization, assuming of course
that the dipoles are sufficiently mobile. How-
ever, this is rarely the case with organic in-
sulating materials, or even with liquid insu-
lation oils of relatively high viscosity. With
inorganic insulating materials, the dipoles in
the crystal structure may reorientate, which
leads to a considerable increase in capacity.
Such insulating materials are characterized
by a significantly higher relative permittivity
(e.g., titanium dioxide with a relative permit-
tivity of ca. 100).
3) In ferroelectric materials that undergo spon-
taneous polarization (e.g., barium titanate),
complete regions are reorientated by an ex-
ternal electrical field. This leads to relative
permittivity values of ≥ 1000.
The transportation of charge carriers in an
electric field and the continuous reversal of po-
larization in the a.c. field both incur energy
losses. They cause a certain degree of heating of
the capacitor dielectric. The power requirement
for this is taken from the a.c. electrical field.

2.5. Dielectric Dissipation Factor
The dielectric dissipation factor (dielectric loss)
is the tangent of the loss angle δ by which the
phase displacement between current and volt-
age in a capacitor deviates from π/2 when the
dielectric of the capacitor consists solely of the
insulating material under investigation (Fig. 5).
At the same time, it is also the ratio of the loss or
effective power N e to the reactive power N r of
a capacitor completely filled with the insulating
material and subjected to an a.c. voltage.
Figure 5. Sinusoidal alternating current: voltage (U), cur-
rent strength (I ) and power (P) as a function of time (t)
δ = loss angle; ϕ = phase angle; ω = frequency
The following equation describes the rela-
tionship between the dielectric dissipation factor
(tan δ) and the power factor (cos ϕ) of a capaci-
tor, where ϕ = 90◦ − δ and the apparent power,

electric loss performance Pv , relative to a unit
Na = Ne2 +Nr2
Ne Ne tanδ
cosϕ= = =
Na Ne2 +Nr2 1 + tan2 δ
The dielectric dissipation factor alone only
permits comparison between the specific dissi-
pation of different materials when the relative
permittivity is the same (see Section 2.6).
The procedures for measuring the dielectric
dissipation factor are described in the interna-
tional standard IEC Publication 250 [30].
The highest-quality dielectric materials pos-
sess dissipation factors (tan δ) of < 0.0002
within a temperature range of 20 to 100 ◦ C and
a frequency range of 50 Hz to 1 MHz. With
some insulating materials the dielectric dissipa-
tion factor is greater at low frequencies than at
high frequencies. The dissipation factor gener-
ally increases with increasing temperature. In
high-power systems, the heat generated by the
energy losses has detrimental effects.
Insulation, Electric 7
With glasses and ceramic materials incorpo-
rating a glassy phase, the network-modifying al-
kali ions have a higher mobility than the dou-
bly bonded alkaline-earth ions. In low-loss in-
sulating materials, the fluxing agents (which are
normally based on alkali compounds) are sub-
stituted by equivalent materials based on alka-
line earths. Examples include “low-loss steatite”
and “alkaline-earth porcelain”. Improvements
in dielectric properties are, however, associ-
ated with manufacturing problems; alkali com-
pounds used as fluxing agents cannot usually be
substituted.
2.6. Loss Index
The loss index ε r of an insulating material is
defined as the product of the relative permittivity
and the dissipation factor
ε


r =εr ·tanδ
When comparing different materials, the dielec-
tric loss is to be preferred to the dissipation fac-
tor because the dielectric losses of identical vol-
umes under identical field strengths and at iden-
tical frequencies behave in the same way as the
respective loss indices (see Section 3.2). The di-
volume of, for example, 1 cm3 can be calculated
using the following equation:
   
Pv percm−3 =ω·ε0 ·ε

2 −14
W/cm3
r ·E ·10
where ω = angular frequency = 2 π f (frequency
in Hz), ε0 = electric constant (0.08854 pF/cm),
ε r = tan δ · εr , E = field strength (V/cm).
2.7. Breakdown Voltage
The breakdown voltage is the voltage level at
which the potential between the electrodes can-
not be maintained and the specimen is destroyed.
It is not a material-specific property and depends
on numerous parameters:
sample thickness,
rate of voltage increase
type of test voltage (d.c. or a.c.),
duration of voltage exposure,
environmental temperature,
humidity,
8 Insulation, Electric
mechanical state of the specimen,
homogeneity,
surrounding medium, and
electrode size and positioning.
Hence all of the specifications listed in the
standards (e.g., international standard IEC Pub-
lication 243) must be met exactly if compara-
ble results are to be obtained. Even if all sec-
ondary effects can be discounted from the re-
sults (e.g., parasite discharges at the edges of
the electrodes), the short-term breakdown volt-
age does not in general possess the character of
a material-specific property. It is more a crite-
rion for the degree of defectiveness of the insu-
lator because the weakest regions (e.g., inhomo-
geneities, voids, gas inclusions) are monitored
during measurement.
Breakdown behavior of insulating materials
varies according to material type and testing con-
ditions; differences depend on various break-
down mechanisms (Fig. 6). In general, a purely
electrical breakdown is observed during impulse
tests (e.g., exposure to peak loadings), and the
values of breakdown voltage are correspond-
ingly high. During long-term field strength test-
ing, thermal breakdown occurs as a result of lo-
cal instability caused by the heat generated by in-
creased conductivity losses and dielectric losses
(see Section 2.6). Hence the values of break-
down voltage obtained are lower, possibly also
because of the low thermal conductivity. The dis-
charge breakdown is of particular importance,
especially during the testing of compacted ma-
terials and liquids. It is caused by discharge and
destructive processes located around inhomo-
geneities or gas inclusions.
Figure 6. Electrical strength as a function of loading time
The electrical strength is defined as the ra-
tio of the breakdown voltage to the distance be-
tween the two electrodes across which the volt-
age is applied. The term electrical strength is
only valid when a uniform field exists between
the electrodes.
Determination of the breakdown voltage is
specified in the international standards IEC Pub-
lication 243 [31] for solids and IEC Publication
156 [32] for liquid insulating materials.
2.8. Further Electrical Tests
For special applications or investigations, sup-
plementary test procedures have proved valu-
able; they have gained international recognition
and have been standardized.
High-Voltage Tracking Test. This test pro-
cedure (Fig. 7) is utilized for materials that are
used under severe ambient conditions in electri-
cal engineering (e.g., high humidity, high volt-
ages). The test enables assessment of the stabil-
ity towards track formation and erosion. Track
formation is defined as a process in which track-
ing and erosion are the result of an electrical dis-
charge at or adjacent to the contaminated surface
of an insulating material.
Figure 7. Schematic representation of measurement of high-
voltage tracking
a) Current limiting resistor; b) Excess current switch acti-
vated by relay (c) when I ≥ 60 mA for 2 s; c) Relay; d) Spec-
imen (inclined plane); e) Contamination fluid; f) Voltage
supply
The test must be carried out with plane speci-
mens that are inclined at 45◦ and moistened with

a contamination liquid. The contamination liq-
uid consists of distilled water containing ammo-
nium chloride and a specified wetting agent. The
specimen has the dimensions of 5×12 cm. High
voltages of 2.5, 3.5, or 4.5 kV are applied to the
two electrodes fastened on the smaller edges of
the specimen. The contamination is allowed to
flow at a specified rate over the lower surface of
the specimen. The time (up to 6 h) is observed
for failure caused by tracking or burning.
The tested materials are classified into four
categories according to their resistance to track-
ing which depends on the test voltage and time.
A correlation exists between this test and the
low-voltage tracking test (see Section 2.1). The
procedure is specified in the international stan-
dard IEC Publication 587 [33].
Arc Resistance. The test procedure accord-
ing to ASTM-D 495 uses a high-voltage sup-
ply of 12.5 kV in order to generate arcing across
the surface of the specimen [34]. Contrary to
the high-voltage tracking procedure according to
IEC Publication 587 (see above), the samples are
tested in the absence of arbitrary contaminants.
The electrodes are round tungsten rods. The an-
gled tips of these rods are placed in contact with
the upper or under surface of the specimen under
investigation at a defined angle and at a distance
of 6.35 mm apart (Fig. 8). The insulating mate-
rial is classified according to the duration of the
arc in seconds. At the end of each test minute,
the power effect on the surface of the specimen
is doubled.
Figure 8. Arc resistance test (ASTM)
Arcing occurs between the tungsten electrodes across the
sample surface.
Insulation, Electric 9
Low Voltage – High Current Arc Test. This
test is used to investigate the behavior of heat-
resistant materials. These are generally inor-
ganic substances or organic materials which are
highly filled with inorganic substances. The test
determines (1) the distance over which a car-
bon arc can be extended across the surface,
(2) whether the arc path is conducting, and
(3) whether the material underneath the surface
changes significantly after cooling. The test pro-
cedure is specified in DIN VDE 0303 part 5 [35].
Assessment of Electrostatic Behavior.
Electrostatic behavior can be assessed in a test
method in which either disk-shaped samples
(rigid materials) or endless strips or tapes (flexi-
ble materials), are charged using contacting sub-
strates. The field strength generated by separa-
tion of the specimen and the substrate is mea-
sured as well as the change of the charge as a
function of time. Since the tests are designed
to allow the selection of nonchargeable or only
slightly chargeable materials, they are only of
advantage for materials with a resistance at the
lower end of the insulation range (109 – 1011 Ω).
The test is related to the determination of sur-
face resistance (see Section 2.3). Procedures are
specified in DIN VDE 0303 parts 8 and 14 [36],
[37].
Determination of Corrosion-Promoting
Effects. The corrosion-promoting effects of in-
sulating materials can be determined by three
standardized procedures specified in IEC Publi-
cation 426 [38]. A visual procedure, a tensile
strength method, and an insulation resistance
method are available to select between materials
which have corrosion-promoting effects on non-
precious metals. With the first two tests, effects
are evaluated according to the damage following
contact with copper or brass. For special applica-
tions, other metals can be employed. During the
evaluation of the resistance change, a decrease
in resistance between taper pin electrodes or bar
electrodes (measurement procedure in IEC Pub-
lication 167 [39]) after storage under humid con-
ditions is regarded as being a criterion for the
presence of a corrosion-promoting effect.
Within the fields of telecommunications, pro-
cess control, and data processing, insulating ma-
terials are often employed to insulate conductors
placed under a d.c. voltage. Under unfavorable
10 Insulation, Electric
conditions, the metallic components may cor-
rode. This can be simulated by testing under
the simultaneous affects of a d.c. supply, heat
(e.g., 40 ◦ C), and environmental humidity (e.g.,
92 % R.H.). The test reveals which combina-
tions of insulating material and metal are par-
ticularly susceptible to corrosion. Judgement is
made by observing color changes, destruction of
test metal strips, or decreases in insulation resis-
tance. Testing procedures are specified in IEC
Publication 426 [38].
Relative Resistance to Breakdown by Sur-
face Discharges. This method is used to esti-
mate the working life of materials by exposing
them to voltages above the discharge voltage. As
with the determination of the breakdown voltage
(Section 2.7), the working life decreases rapidly
with increasing field strength. The partial dis-
charges at the surface damage the material in a
way similar to discharges at defects within the
material.
The resistance breakdown by discharges is
tested. Electric strengths refer to one year and
are only comparable if sample thicknesses are
identical and all test parameters are kept con-
stant. A minimum of nine samples are tested at
a minimum of three voltage settings. Large de-
viations must be expected because defects are
irregularly distributed and the test area is rela-
tively small. Hence the median of each set of
results is used for evaluation; extrapolation to
one year is performed either graphically or us-
ing regression calculations.
The test procedure is described in IEC Pub-
lication 343 [40].
Detection of Transition Regions. The de-
tection of regions of relaxation transition us-
ing thermally stimulated discharge currents is an
extremely effective method for studying inelas-
tic effects in polymers. The high resolution al-
lows the complex behavior of the materials to be
broken down into elementary processes (Fig. 9)
[22].
The interpretation of the informative dis-
charge curves requires further intensive inves-
tigations with physical or other methods. Ther-
mal stimulation of discharge currents can be
used to solve a variety of problems [22]; ex-
amples include effects of additives on elec-
trical and mechanical properties; compatibility
in copolymers; deformation; aging; irradiation;
cryogenic processing; crystallinity; conducting
mechanisms; activation energy; stereo com-
plexes; thermal histories; cross-linking; mois-
ture content; and relationships between mechan-
ical relaxation, elongation, and calorimetry.
Figure 9. Relaxation regions of an acrylonitrile –
butadiene – styrene (ABS) product
a) Dielectric damping (1 kHz); b) Thermally stimulated dis-
charging (0.01 Hz); c) Mechanical damping (1 Hz)
2.9. Mechanical and Thermal Properties
Strength. The applicability of an insulator
is primarily determined by its electrical proper-
ties. In many applications, however, mechanical
properties are of equal importance.
The assessment of the mechanical strength of
insulating materials is carried out using special
specimens. However, the strength characteristics
determined in this way cannot simply be applied
to an insulation component because the strength
decreases with increasing sample volume. This
particularly applies to brittle, nonmetallic, inor-
ganic materials. During the design of compo-
nents and equipment, this must be taken into
account and safety corrections must be made.
These are known as a result of experience but
they can also be approximately calculated by us-
ing techniques such as fracture mechanics and
Weibull statistics.
Due to experimental difficulties however, ac-
ceptably reproducible results are only obtained
in bending strength tests. Therefore standards
have only been drawn up for this test. However,
tensile tests are also routinely carried out when
testing insulators.

The strength of ceramic insulators can be in-
creased considerably by application of suitable
glazes. The thermal expansion coefficient of the
glaze must be slightly lower than that of the ce-
ramic substrate. Following cooling after firing,
the glaze is subjected to compressive stresses.
When an external tensile force is applied to a
glazed article, these compressive stresses in the
surface must be counteracted before the tensile
stresses can become effective. In glasses, sim-
ilar effects are achieved by thermal hardening,
i.e., by quenching from temperatures above the
glass transition temperature. Chemical harden-
ing (via ion exchange at elevated temperature)
is used less widely.
With organic polymers, similar increases in
strength can be gained by employing composite
materials: glass or carbon fibers are incorporated
into the polymer.
Material Space Filling. Apart from com-
pletely solid materials, porous substances are
also used in electrical engineering. Porosity low-
ers the breakdown strength, therefore porous
materials are not considered for high-voltage ap-
plications. In contrast to this, an open porous
structure is favorable for the thermal shock re-
sistance because porous materials have a lower
Young’s modulus. Porous materials are there-
fore often employed in thermal electrical engi-
neering. An increase in the surface conductiv-
ity caused by the larger surface does not pose
a threat because the water film is evaporated by
the current flow.
Under high-vacuum conditions, porous ma-
terials are also utilized as electrical insulators to
prevent the formation of continuous vapor lay-
ers.
Thermal Expansion. The thermal expan-
sion of inorganic insulators is generally rela-
tively low which is favorable for thermal shock
resistance. A low thermal expansion also en-
sures good shape retention following tempera-
ture change. Inorganic insulators can withstand
rapid temperature increases better than rapid
cooling.
Organic substances are prone to material
changes as a result of prolonged exposure to high
temperature, which affects both the mechanical
and electrical properties. Special organic com-
pounds have recently been developed which are
Insulation, Electric 11
also suitable for application at elevated temper-
atures.
Thermal Conductivity. The thermal con-
ductivity of the insulating materials is gener-
ally low. Hence, they can be simultaneously em-
ployed as electrical and thermal insulators. How-
ever, oxidic and nonoxidic materials are avail-
able which have a similar thermal conductivity
to metallic substrates. Such insulators are nor-
mally employed when heat transfer or removal
is necessary.
Specific Heat. The specific heat of the insu-
lating material may be of significance in heat
storage applications. Magnesium oxide with a
specific heat of greater than 1000 J kg−1 K−1 is
particularly suitable.
Combustion Behavior. The possibility of
combustion can be a considerable obstacle re-
garding the use of organic insulators. Materials
that do not burn at all when placed in contact
with an electrical source of heat are given prior-
ity. Classification criteria include the extinguish-
ing of the flame following removal of the heat
source, or the speed at which the flames spread.
3. Insulating Materials
3.1. Inorganic Insulating Materials
Inorganic electrical insulating materials include
several natural minerals (generally mica or as-
bestos) but also industrially manufactured prod-
ucts. The wide range of industrially manufac-
tured materials includes pure glass, materials
containing glassy and crystalline structural com-
ponents, purely crystalline materials, and even
single crystals. This broad spectrum of mate-
rials ensures a wide variety of properties. De-
pending upon the intended application, one or
more of the following properties may be particu-
larly pronounced: low or high relative permittiv-
ity; extremely low dielectric loss; high electrical
strength at low and high frequencies; excellent
insulation characteristics even at temperatures
well above 1000 ◦ C; low or high thermal con-
ductivity; high mechanical strength.
The ferrites are also inorganic insulating ma-
terials. Ferrites are sintered products consisting
12 Insulation, Electric
of iron(III) oxide and oxides of mono- or diva-
lent metals. Their application within electrical
engineering is primarily governed by their mag-
netic properties. Ferrites are generally regarded
as magnetic substances (→ Magnetic Materi-
als), although their insulation characteristics are
of interest for certain applications.
3.1.1. Insulating Materials Based on Natural
Minerals
3.1.1.1. Mica and Related Products
(→ Mica)
Micas are sheet silicates. Their properties are
determined by mixtures of homo- or het-
eropolar bonds within the lamellar planes,
which in turn are held together by van der
Waals’ forces. This type of bonding is re-
sponsible for several of mica’s properties,
e.g., the excellent cleaving properties paral-
lel to the plane of the lamellae. Muscovite
[1318-94-1] , KAl2 [AlSi3 O10 (OH)2 ] , and phlo-
gopite [12251-58-0] , KMg3 [AlSi3 O10 (OH)2 ] ,
havefound widespread application in the elec-
tronics industry. Materials selected for insula-
tion purposes must not contain contaminants
which may adversely effect insulation proper-
ties. The mica products employed in electronics
mainly originate from India.
Production and Processing. The mica is se-
lected by hand from the mined rock and sorted
according to size. Small fragments are ground
and processed into powder. Larger specimens
are cleaned and cleaved into sheets of thick-
ness as low as 0.01 mm. Commercially available
products are classified according to purity and
surface area.
Cleaned, cleaved mica is generally processed
by punching. The requirements of the punched
component regarding tolerances and precise fin-
ishing are extremely high. During manufacture
of capacitors, the metallizing preparation (e.g.,
a silver suspension) is applied directly onto the
mica substrate and fired onto the surface at
600 ◦ C following drying.
Mica products can be processed to vacuum-
sealed components. Lead borate glasses are par-
ticularly suitable solders because of their low
melting temperatures. Such composites can be
used up to 300 ◦ C.
Composite Insulating Materials. Large
amounts of mica powder and scrap are pro-
cessed to nonwoven substrates or sheets. In
one process, scrap mica is heated to 800 ◦ C; a
portion of the water of crystallization escapes
and causes partial cleavage of the lamellae. The
mica is then immersed in sodium carbonate so-
lution whilst still hot. Carbon dioxide gas is
evolved during subsequent treatment with di-
lute acid, and causes complete separation of the
mica lamellae. The very fine mica platelets can
be processed into thin, endless, nonwoven sub-
strates in a procedure analogous to that used in
the manufacture of paper.
In another process, the mica is cleaved using
a high-pressure water jet and again processed to
a nonwoven fabric. The product is slightly less
densely packed than that resulting from the pro-
cedure mentioned previously.
Such nonwoven fabrics, and small rejected,
cleaved mica fragments are combined with natu-
ral or synthetic binders to give sheets, films, and
tapes (micanite, VDE 0332, IEC 371). Suitable
binders include asphalt, shellac, alkyd resins, or
more recently, silicone and epoxy resins.
Micanite is a binder-containing sheet prod-
uct that is manufactured using heated hydraulic
presses. Commutator micanite should not con-
tain more than 4 % binder. Sheets are ground
to meet the required thickness within the toler-
ance limits, and then punched into commutator
segments.
Heating micanite is designed for use as in-
sulation in electric heaters; here, the binder im-
mediately ignites when the device is first uti-
lized. The remaining mica layer is responsible
for insulation. With an appropriate binder, only
short-term inconvenience is caused by the com-
bustion products. The binder should not produce
an electrically conducting residue after com-
bustion. Shellac is a very satisfactory binder in
amounts up to 3 %. Heating micanite is also man-
ufactured from fine mica and contains 10 to 15 %
silicone resin as an additive. The organic compo-
nents ignite to form gaseous combustion prod-
ucts; the residual silica functions as a cement and
confers the required strength to the insulator.
Rigid micanite, which cannot be shaped, is
obtained on addition of increased amounts of
binder. If subsequent shaping is desired, ther-
moplastic binder materials must be employed.

Mica foil consists of a thin substrate covered
with one or two layers of cleaved mica or fine
nonwoven mica material that are bonded to-
gether with a binder. Depending upon the type of
substrate employed, the products are known as
paper micanite, fabric micanite, nonwoven mi-
canite, or film micanite (VDE 0332, IEC 371).
“Glimmerband” (mica tape) denotes mica foil
which has been slit into tapes.
Glass – mica composite materials can also
be manufactured with low-melting lead borate
glasses. With appropriate selection of the glass
component, the semifinished article can be ma-
chined.
Properties. The properties of mica products
are strongly influenced by the nature of the
binder. Fine mica articles are more flexible than
their cleaved mica counterparts. The applica-
tion range is essentially governed by the type
of binder used, but generally lies below 120 ◦ C.
With silicone resin binders, prolonged exposure
to temperatures as high as 180 ◦ C is possible.
Heating micanite bound with silicone resins is
able to withstand temperatures of 500 ◦ C.
Uses. Mica wafers in their naturally occur-
ring form are generally employed to fix elec-
trode systems in valves and guarantee a high
degree of stability, independent of the operat-
ing temperature. Further areas of application in-
clude discharge windows for high frequency en-
ergy, e.g., in transmitters operating in the mi-
crowave range, or as admission windows in
counter tubes. Their reasonably high relative
permittivity means that they can be used as in-
sulators in capacitors in constant-frequency de-
vices. The good thermal stability is utilized for
insulating heating systems.
Micanites and mica foils are suitable for a
variety of insulation applications, particularly
in the construction of electrical equipment and
plants.
3.1.1.2. Asbestos and Related Products
(→ Asbestos)
Asbestos [1332-21-4] is a fibrous form of mag-
nesiumsilicate. It has an extremely high thermal
stability as well as good electricaland thermal
Insulation, Electric 13
insulation capabilities even at elevated temper-
atures.
Asbestos materials cause severe lung dam-
age (asbestosis), which has led to a ban on their
use. They have been replaced by fiber materials
produced from ceramics, particularly alumina
ceramic fibers. Their properties are equal to or
even better than those of asbestos. It has not yet,
however, been clarified whether these alterna-
tive fibers can also cause cancer. No regulations
have been laid down so far. The IEC standards
and draft standards for the standardization of as-
bestos products have not yet been withdrawn.
3.1.2. Synthetic Inorganic Insulating
Materials
Synthetic inorganic materials which are com-
monly used as electrical insulators are described
in IEC Publication 672 [41], which has also been
recently adopted as the German Standard DIN
VDE 0335. Part 1 specifies terms and classifica-
tion, Part 2 describes basic testing procedures,
and Part 3 contains characteristic data for the
individual material types.
Classification.The IEC 672 standard con-
cerns insulating materials made from ceramics,
glass, and glass ceramics, although the latter
have not yet been standardized as a material type.
The ceramic insulators are classified into eight
groups:
C-100 materials based on
alkali – aluminum silicate
(porcelains)
C-200 materials based on magnesium
silicate (steatites and forsterites)
C-300 materials based on titanium
dioxide, titanates, stannates, or
niobates (ceramics of high
relative permittivity)
C-400 materials based on alkaline
earth – aluminum silicates
C-500 porous materials based on
aluminum silicate and
magnesium silicate
C-600 materials based on aluminum
silicate (mullite ceramics)
C-700 materials of high alumina
content
C-800 special oxide ceramics
The glass insulators are classified into six
groups (see next page).
14 Insulation, Electric
G-100 alkali – lime – silica glasses
G-200 chemically resistant borosilicate glasses
G-300 electrically resistant borosilicate glasses
G-400 alumina – lime – silica glasses
G-500 lead oxide – alkali – silica glasses
G-600 barium oxide – alkali – silica glasses
According to IEC 672, ceramic materials are
those substances which are shaped before fir-
ing; their main components are usually poly-
crystalline silicates, aluminosilicates, titanates,
or oxides. Glasses are dense inorganic materi-
als manufactured from a melt or by sintering,
which solidify without crystallization. Glass ce-
ramics are manufactured from glass, they con-
tain a considerable crystalline content as a re-
sult of a special treatment. Glazes are coatings
with a smooth surface which are generated in
a melt process and which essentially consist of
glass. They can contain coloring components or
opacifiers.
The above classification of ceramic insulators
is not systematic. Apparent inconsistencies are
due to historical reasons. Because the preced-
ing standard, DIN 40 685, had proven valuable
for several decades and was internationally ac-
cepted, its classification system was adopted in
the new IEC 672 standard. Some groups are par-
tially defined on the basis of material aspects.
However, structure is also decisive, in particu-
lar the presence of an open pore system. For
example group C-500 contains only porous ma-
terials, whereas dense substances of the same
chemical composition are classified in another
group. Porous materials are not limited to group
C-500 however. Certain groups (e.g., C-100 and
C-200) contain both porous and dense varieties
of a given composition.
Testing.IEC 672, Part 2 describes the testing
procedures used for determining the mechanical,
thermal, and electrical properties of the individ-
ual materials. Precise specification of test pro-
cedures, including sample preparation, is nec-
essary because data are strongly influenced by
the manufacturing process, shape, and dimen-
sions of the sample, and also by the measuring
procedure itself. Accordingly, data are strictly
valid only for the stipulated sample and cannot
be applied to samples that are of another shape or
are manufactured by another process. However,
modern fracture mechanics and Weibull statis-
tics allow the properties of actual construction
components to be estimated from the character-
istic data obtained with standardized samples.
With glass, the thermal history of the material
is important. Prestressed glass is influenced by
the conditions experienced during thermal hard-
ening. Here, it is even more difficult to deduce
the properties of an insulation component from
those of a standard specimen.
3.1.2.1. Porcelain
Porcelain was introduced as an electrical insulat-
ing material by W. von Siemens in the middle of
the nineteenth century. Various types of porce-
lain have since become the material of choice
for high-voltage insulators in overhead lines and
transformer stations. Normal porcelain is less
suitable for high-frequency and electrical heat-
ing applications because the desired properties
cannot be obtained due to the presence of alkali
from the fluxing agent. For special applications,
alkaline-earth porcelains are available in which
the alkali elements are eliminated. Only small
components can be manufactured from these
materials, however, because their firing interval
(i.e., the temperature range between vitrification
and distortion) is very small.
Material Types. The C-100 group according
to IEC 672 Part 3, comprises five material types
which differ with regard to their phase compo-
sition and required properties, particularly the
mechanical data:
C-110 quartz porcelain
C-111 compressed porcelain (with a
permitted residual porosity)
C-112 cristobalite porcelain
C-120 alumina porcelain, strong
C-130 alumina porcelain, ultrastrong
The compressed porcelains of type C-111
are only suitable for special low-voltage appli-
cations due to their permitted open porosity.
Cristobalite porcelains exhibit better mechani-
cal properties than the quartz porcelains. Alu-
mina porcelains are now the preferred porcelains
for high-voltage insulators. Their classification
in the two types C-120 and C-130 takes into
consideration recent developments with high-
strength materials being classified as type C-
130. Their manufacture is, however, difficult
which explains why C-130 type materials are
 Insulation, Electric 15

only employed when extreme mechanical de- sion of a pug with a vacuum extruder and sub-
mands have to be met by the insulator. For nor- sequent cutting either after partial drying in a
mal requirements type C-120 materials, which “leather-hard” state (Fig. 10), or after complete
present fewer manufacturing problems, suffice. drying. The pug itself can be manufactured by
The use of the various porcelain types is isostatic pressing to give a dry pug, suitable for
based on differing philosophies regarding se- immediate machining. Care must be taken to en-
lection criteria. In western industrial nations, sure that any dust generated during machining
quartz porcelain is only used when the insulator is removed completely.
is not exposed to mechanical loading. In general,
alumina porcelains are applied even when they
more than satisfy the requirements of the prob-
lem in hand. However, the universal utilization
of alumina porcelains eliminates a range of prob-
lems that could arise if insulators were made in
a single factory from different materials. More-
over, this improves the safety of overhead lines
and transformer stations in which failure of an
insulator could have a serious effect on energy
supply. In contrast, in other countries (e.g., the
Comecon countries), the minimization of the im-
port of expensive synthetic alumina is the pre-
dominant factor. Here, quartz porcelain is used
whenever acceptable.

Properties. Important properties of the

porcelain types C-110, C-120, and C-130 are
summarized in Table 1.

Raw Materials and Body Composition.

The traditional hard porcelain is made from
china clay, quartz, and feldspar. The plasticity
of the china clays is more important than its
ability to produce whiteness after firing. China
clay can thus be partially substituted by plastic
clay in order to improve the workability of the
porcelain body.
Quartz porcelains have a high quartz content
of up to 40 % which accounts for their high
strength. During firing, certain types of quartz
are readily converted to cristobalite, which gives
rise to the formation of cristobalite porcelain
types [42]. The substitution of quartz by alu-
mina produces the alumina porcelains; type C-
120 contains up to 30 wt% and type C-130 up to
40 wt% alumina.
Manufacture (→ Ceramics, General Sur-
vey). Molding and shaping are carried out us-
ing conventional processes employed in the ce-
ramic industry. Low-voltage insulation compo-
nents are manufactured by pressing techniques.
High-voltage insulators are produced by extru-
Figure 10. Machining of an insulator with photoelectric con-
trol
The shaped components are coated with a raw
glaze, dried, and fired in a tunnel, chamber, or
oven kiln.
Semiconducting glazes coated onto insula-
tors may influence the shape of the electric
field and contribute to the prevention of short
circuiting due to condensation. However, they
can cause problems during operation, and have
therefore not yet made a general breakthrough,
although they would permit the more compact
construction of transmission lines.
Uses. Low-voltage insulation components
made from compressed porcelain are employed
in a variety of applications including coupler
terminal strips, cartridge fuses, terminal blocks,
and inserts for plugs and sockets. They may be
used in combination with organic polymers.
A wide range of porcelain insulators are em-
ployed in high-voltage applications: support in-
16 Insulation, Electric
Table 1. Important physical properties of porcelain types C-110, C-120, and C-130 classified according to IEC 672

Property C-110 C-120 C-130

Open porosity, vol% 0 0 0

Bulk density, g/cm3 >2.2 >2.3 >2.5
Bending strength, N/mm2
Unglazed >50 >90 >140
Glazed >60 >110 >160
Linear expansion coefficient (20 – 600◦ C), 10−6 K−1 4–7 4–7 5–7
Electrical strength, kV/mm >20 >20 >20
Withstand voltage (1 min), kV >30 >30 >30
Relative permittivity at 20 ◦ C and 50 Hz 6–7 6–7 6 – 7.5
Dielectric dissipation factor (tan δ) at 20 ◦ C and 50 Hz <0.025 <0.025 <0.030
Volume resistivity, Ω · cm
at 20 ◦ C >1013 >1013 >1013
at 600 ◦ C >104 >104 >104

sulators, suspension insulators, and terminal in-
sulators for overhead line construction; insula-
tor housings able to withstand extreme mechan-
ical loading; or as external shielding for voltage
transformers of heights up to 5 m and diameters
of 1 m. Lead-through insulators can also be pro-
duced with multiple bore holes running parallel
to the longitudinal axis.
The alkaline earth porcelains are mainly em-
ployed as resistor bodies.
3.1.2.2. Steatite
The term steatite [14807-96-6] has two mean-
ings, firstof all as a synonym for the fine-grained
variety of talc, and secondly fora special ceramic
material manufactured from this talc.
To avoid confusion, the term steatite is used
here to denote the ceramic product. The term
talc (and not steatite) is used for the ceramic raw
material regardless of its crystal size. Normally,
however, the term talc solely denotes the coarse-
grained variety of the mineral.
Material Types. Group C-200 according to
IEC 672 contains six types of material based on
talc:
Type C-210 steatite for low-voltage applications (not dense)
Type C-220 steatite
Type C-221 low-loss steatite
Type C-230 steatite, porous
Type C-240 forsterite, porous
Type C-250 forsterite, dense
The most important materials in group C-
200 are steatite (type C-220) and low-loss
steatite (type C-221). Type C-250 is formed from
a MgO-enriched body and contains forsterite
[15118-03-3] (Mg2 SiO4 ) as the main crys-
talline phase. The porous types (C-230 and C-
240)allow the manufacture of high precision
components from fired semifinished products
using conventional machining techniques.
Properties. Several of the most important
properties of the C-220 and C-221 types are
listed in Table 2.
Type C-220 steatite is manufactured using
feldspar as a fluxing agent. This yields an ad-
equate firing range for the production of larger
items. With low-loss steatite (type C-221), di-
electric properties are considerably improved
when barium carbonate is used as a fluxing
agent. This material is thus also suitable as a
dielectric for capacitors. For some applications
its relatively high mechanical strength is also of
significance. With the exception of high-voltage
insulators, low-loss steatite is the most popular
variety of steatite.
Compared with the porcelains, the steatites
have outstanding electrical properties. Only the
best alumina porcelains have mechanical prop-
erties comparable with those of steatite. Recent
investigations have shown that the resistance of
steatite to crack propagation is slightly greater
than that of alumina porcelain [43].
Raw Materials. The preferred raw mate-
rial for the manufacture of steatite is talc
[14807-96-6] , Mg3 [Si4 O10 (OH)2 ] in its fine
crystalline form (→ Talc). The random orienta-
tion of the crystals in this talc allows the man-
ufacture of ceramic products with low particle
 Insulation, Electric 17
Table 2. Important physical properties of steatite types C-220 and C-221, as well as forsterite type C-250 classified according to IEC 670

Property C-220 C-221 C-250

Open porosity, vol% 0 0 0

Bulk density, g/cm3 >2.6 >2.7 >2.8
Bending strength, unglazed, N/mm2 >120 >140 >140
Linear expansion coefficient (20 – 600 ◦ C), 10−6 K−1 7–9 7–9 9 – 11
Electrical strength, kV/mm >15 >20 >20
Withstand voltage (1 min), kV >20 >30 >30
Dielectric dissipation factor (tan δ) at 20 ◦ C
50 Hz <0.005 <0.0015 <0.0015
1 MHz <0.003 <0.0012 <0.0005
Volume resistivity, Ω · cm
at 20 ◦ C >1013 >1013 >1013
at 600 ◦ C >105 >107 >107
Relative permittivity at 20 ◦ C and 50 Hz 6 6 7

orientation. One of the most important talc de-
posits is found near Göpfersgrün – Thiersheim
in the Federal Republic of Germany. It repre-
sents the main basis for the worldwide manu-
facture of large, high-voltage steatite insulators.
Talc for use in ceramics originates in Australia,
the United States, Spain, and China.
Only talcs with low levels of contamination
can be used as a raw material for the manufacture
of steatite. The maximum upper limits which
can be tolerated are 1.5 wt% Al2 O3 and 2.0 wt%
Fe2 O3 [44]. Recently, calcia has been regarded
as less critical and may be present in amounts
up to 1 wt% [45]. The steatite body contains 80
to 90 % talc along with 5 to 10 wt% of a fluxing
agent (in steatite, feldspar; in low-loss steatite,
barium carbonate) and plastic clay to improve
the processing properties.
Manufacture. Components from steatite are
produced according to the general principles em-
ployed in the ceramics industry (→ Ceramics,
General Survey). All molding and shaping tech-
niques can be employed such as casting, shaping
in the plastic state, or pressing, and when neces-
sary subsequent machining in the leather-hard or
dried state. The most important technique, par-
ticularly for small components, is pressing. The
dry pressing process only gained widespread use
after its development with talc bodies. Talc bod-
ies have proved to be particularly suitable for
this technique because the low hardness of talc
(Mohs hardness = 1) guarantees a long service
life.
The steatite components are usually fired at
1300 – 1400 ◦ C in an oxidizing or reducing at-
mosphere. The firing interval is lower than that
of porcelain, particularly with low-loss steatite.
The lower firing interval for steatites is mainly
responsible for the fact that steatite insula-
tors can only be manufactured with maximum
heights of ca. 1.2 m.
During firing, dry-pressed components
shrink by only 10 to 12 %, and hence are suitable
as insulator components with high dimensional
tolerances.
Metal layers of silver, platinum, iron, or
molybdenum can be used as electrodes on
steatite or as a soldering base for joining steatite
to metals of other ceramic components.
During firing, the talc is transformed to pro-
toenstatite, MgSiO3 . The additives (clay and
feldspar or barium carbonate) form a melt phase
which vitrifies on cooling; the glass content of
the fired body may be as high as 25 to 50 vol%
[46]. An excess of silica leads to crystallization
of cristobalite from the melt; this occurs in cer-
tain steatite bodies, but particularly in feldspar
steatites. Improved thermal shock resistance is
observed with steatites in which the silica has
been completely taken up by the melt phase, for
example in most low-loss steatite products con-
taining silica-free fluxing agents.
Uses. Steatite finds widespread use as an in-
sulating material in electrical engineering. In
low-voltage applications, it is employed for the
manufacture of insulation components in which
a high degree of dimensional precision is re-
quired (e.g., for mechanically mounted compo-
nents). Typical insulators made from steatite in-
clude sockets, inserts for plugs, levers and rollers
18 Insulation, Electric
for rotary and rocker switches, holders, and car-
tridge fuses. Steatite has many advantages over
porcelain when greater mechanical loading has
to be withstood.
In high-voltage applications, steatite can now
be replaced by alumina porcelain. Abrasion of
machining tools is, however, considerably lower
with steatite bodies. On the other hand, alumina
porcelain bodies are easier to fire and are suitable
for the manufacture of components much larger
than those made of steatite. Steatite shows clear
advantages in high-frequency applications, for
example as bases for antenna masts and for ter-
minal insulators in transmission masts.
Low-loss steatites are suitable for the man-
ufacture of capacitors, e.g., the complex thick-
walled capacitors used in high-voltage applica-
tions.
Talc Bodies Enriched in Magnesia. The
crystalline phase forsterite, Mg2 SiO4 , is de-
liberately formed in fired bodies of the material
types C-240 and C-250 by enriching the talc
with magnesia (e.g., by addition of magnesite,
MgCO3 ). Several properties of type C-250 ma-
terials are summarized in Table 2.
The outstanding properties of forsterite ce-
ramics include the particularly low dielectric
losses and the large thermal expansion coeffi-
cient. The latter permits low-stress soldering to
iron or nickel substrates which possess coeffi-
cients of thermal expansion of similar magni-
tude. This advantage of forsterite is, however,
obtained at the expense of low thermal shock
resistance.
Zircon-Containing Steatite. Zircon
[1490-68-2] , ZrSiO4 , can be sinteredto produce
a dense material only with difficulty. When used
as an additivefor steatite, and, in some cases, for
porcelain, the thermal expansion of the result-
ing material is reduced, and hence the thermal
shock resistance is improved. The amount of
zircon additive can reach 60 wt%. The dielectric
properties of the zircon steatites are adversely
affected because losses increase significantly
above 1 MHz. The relative permittivity of zir-
con steatite is 8, this is sometimes undesirably
high for certain insulators.
3.1.2.3. Materials for Electrical Heat
Engineering
Two of the most important demands for the use
of ceramic materials in electrical heat engineer-
ing are adequate stability at high temperature
and favorable thermal shock resistance.
The thermal shock resistance is determined
by the phase composition: compositions which
approach that of a eutectic melt at relatively low
temperatures and are therefore unsuitable. Ther-
mal shock resistance increases with increasing
mechanical strength and thermal conductivity,
but also with decreasing modulus of elasticity
and decreasing thermal expansion. It also de-
pends upon external conditions, such as shape
and dimensions of the article, as well as on the
rate of heat exchange with the surroundings. In
this respect, the thermal shock resistance is not
a material constant.
Young’s modulus is related to the structure of
the material; it is lower with porous structures
than with dense structures.
The most important factor controlling ther-
mal shock resistance is thermal expansion. In
materials with a low thermal expansion coeffi-
cient, the product is only exposed to low stresses,
even if the thermal shock is high.
Hence the most suitable materials for electri-
cal heat engineering are those which are porous
and possess a low thermal expansion coefficient.
Material Types. Group C-400 defined in
IEC 672 contains two types of material:
C-410 cordierite dense
C-420 celsian dense (BaAl2 Si2 O8 )
Type C-410 materials are used more widely.
Group C-500 contains the following porous
materials:
C-510 materials based on grog
(fireclay, chamotte)
C-511 materials based on grog
containing magnesium oxide
C-512 materials based on grog
containing magnesium oxide
C-520 materials of high cordierite
content
C-530 materials of high alumina
content

Types C-511 and C-512 contain cordierite
[12182-53-5] , Mg2 Al4 Si5 O18 , and mullite
[55964-99-3] , Al6 Si2 O13 . They differ in the
number andsize of the pores, and hence also in
the strength and thermal shock resistance. The
composition of type C-520 materials approaches
that of cordierite. The high alumina content of
the type C-530 materials enables high applica-
tion temperatures up to 1300 ◦ C.
Properties. Table 3 contains typical proper-
ties of porous and dense materials used in elec-
trical heat engineering.
Materials Containing Cordierite. Even as
early as the 1930s, addition of talc to clay-based
material had been shown to reduce their thermal
expansion coefficient, and hence increase their
thermal shock resistance. Talc and clay react dur-
ing firing to give cordierite. Optimum results are
obtained with approximately 60 % clay and grog
(chamotte) and 30 % talc (type C-520 material
according to IEC 672). Types C-511 and C-512
contain ca. 10 % talc.
A distinct disadvantage of cordierite materi-
als is their narrow firing range. Consequently it
is normal to refrain from vitrifying the articles,
and concentrate on porous products that have
a lower Young’s modulus and therefore exhibit
improved thermal shock resistance. If dense ma-
terials of type C-410 are required for special
applications, then small amounts of a fluxing
agent must be added; feldspar has proved par-
ticularly effective in this respect. Such materi-
als are not only inferior as regards their thermal
shock resistance, but also in their electrical in-
sulation characteristics. This is because of the
contribution of the alkali ions to their electri-
cal conductivity at high temperatures. Further-
more, the porous materials allow higher applica-
tion temperatures (ca. 1200 ◦ C) than their dense
counterparts. However, dense cordierite materi-
als are required for some applications.
Lithium Aluminum Silicates. Lithium alu-
minum silicates possess an even lower thermal
expansion coefficient than cordierite.
The materials have not been standardized.
Raw materials are petalite, LiAlSi4 O10 , and to a
lesser extent spodumen, LiAlSi2 O6 , or eucryp-
tite, LiAlSiO4 . Widespread application of these
materials has been somewhat hindered because
Insulation, Electric 19
of the scarcity of the raw materials, and because
of the extremely narrow firing range. Therefore
their excellent thermal shock resistance is only
used in exceptional cases.
3.1.2.4. Alumina Ceramics (→ Ceramics,
Electronic)
The inorganic insulating materials discussed in
Section 3.1.1 are all produced from natural mate-
rials. Aluminas and the other materials discussed
in this section are based on synthetic materials
produced by the chemical industry.
Material Types. In IEC 672, alumina mate-
rials are classified in group C-700. They differ
in their minimum alumina contents. In all cases,
the alumina content is so high that the insulat-
ing material contains α-aluminum oxide as the
predominant or even exclusive crystal phase.
Properties. The important physical proper-
ties of alumina ceramics are summarized in Ta-
ble 4. Properties generally improve with increas-
ing alumina content (higher number).
Alumina ceramics combine a range of re-
markable properties. With the high percent-
age types (C-799), a strength of ≥ 500 N/mm2
can be achieved, particularly if crystal growth
can be suppressed during sintering. Mechani-
cal strength is maintained up to temperatures of
1000 ◦ C. Thermal conductivity reaches a level
equivalent to that of metallic lead, especially in
the high percent types. Alumina substrates are
amongst the best electrical insulating materials
under all application conditions.
Production. Raw materials include calcined
aluminas which are obtained from natural baux-
ites by the Bayer process (→ Aluminum Ox-
ide). The precipitation and calcination condi-
tions govern the most important properties in
ceramic applications, such as crystal size and
sintering capability. Since the aluminas are not
plastic, molding and shaping necessitates the ad-
dition of organic plasticizers which must be re-
moved prior to firing.
20 Insulation, Electric
Table 3. Important properties of materials classified in groups C-400 and C-500 of IEC 672

Property Fine fire Materials containing cordierite Materials containing

clay alumina

C-510 C-410 C-511 C-512 C-520 C-530

Open porosity, vol% 30 0.5 20 40 20 30

Bulk density, g/cm3 >1.9 >2.1 >1.9 >1.8 >1.9 >2.1
Bending strength, N/mm2 >25 >60 >25 >15 >30 >30
Linear expansion coefficient (20 – 600 ◦ C), 10−6 K−1 3–6 2–4 4–6 3–6 2–4 4–6
Thermal shock resistance, K >150 >250 >200 >250 >300 >350
Electrical strength, kV/mm >10
Dielectric dissipation factor (tan δ) at 50 Hz <0.025
Volume resistivity, Ω · cm
at 200 ◦ C >109 >108 >109 >109 >109 >1010
at 600 ◦ C >105 >105 >105 >105 >105 >106

Table 4. Important properties of alumina materials classified in group C-700 according to IEC 672

Property C-780 C-786 C-795 C-799

Open porosity, vol% 0 0 0 0

Bulk density, g/cm3 >3.2 >3.4 >3.5 >3.7
Bending strength, N/mm2 >200 >250 >280 >300
Coefficient of linear expansion (20 – 600 ◦ C), 10−6 K−1 6–8 6–8 6–8 7–8
Thermal conductivity, W m−1 k−1 10 – 16 14 – 24 16 – 28 19 – 30
Electrical strength, kV/mm >10 >15 >15 >17
Withstand voltage (1 min), kV >15 >18 >18 >20
Dielectric dissipation factor at 20 ◦ C
50 Hz <0.001 <0.0005 <0.0005 <0.0002
1 MHz <0.0015 <0.001 <0.001 <0.001
Volume resistivity, Ω · cm
at 20 ◦ C >1014 >1014 >1014 >1014
at 600 ◦ C >107 >107 >108 >108

For solder connections, alumina compo- 3.1.2.5. Special Oxide Ceramics

nents are metallized, preferably by the moly –
manganese paste process (→ Ceramics, Ce-
ramic – Metal Systems, Chap. 4.1.). Alternatives
are active soldering with titanium hydride or
use of solders containing titanium under con-
ditions of high vacuum or in an oxygen-free at-
mosphere.
Uses. Alumina ceramics find many applica-
tions within the field of electronics and electrical
engineering due to their outstanding insulation
properties even at very high frequencies. Fur-
thermore, their large thermal conductivity means
that any heat generated as a result of the dielec-
tric losses in electrical circuits and components
can easily be removed. Examples of uses include
housings for high-frequency electron tubes, rec-
tifier housings, high-frequency coupling win-
dows, solderable bypasses, resistor substrates,
substrates for potentiometers, and substrates for
microelectronics.
Group C-800 of IEC 672 comprises several spe-
cial oxide ceramic insulating materials: beryllia
(type C-810), magnesia (type C-820), and zirco-
nia (type C-830).
Beryllia Ceramics. Beryllia ceramics (type
810 according to IEC 672) possess similar prop-
erties to alumina ceramics; the main differ-
ences are the lower density (<3.0 g/cm3 ) and
the large thermal conductivity which is approx-
imately one order of magnitude higher than that
of alumina. In substrates, potentiometer mount-
ing plates, and housings for semiconductor com-
ponents, the material effectively removes heat
produced by dielectric losses. However, beryllia
dust is highly toxic and attacks the lungs, neces-
sitating comprehensive safety precautions dur-
ing manufacturing (beryllia in finished products
is of no danger provided that no dust is gener-
ated). This, along with the limited availability of

raw material deposits, means that beryllia com-
ponents are extremely expensive. Their use is
restricted to a small number of applications that
involve particularly high demands, e.g., in the
aerospace industry.
Magnesia Ceramics. Magnesia is standard-
ized in IEC 672 as type C-820, a porous ce-
ramic insulating material. Its main applications
are found within the field of electrical heat en-
gineering, primarily because of its high specific
heat for use in heat storage elements, but also be-
cause of its favorable thermal conductivity and
excellent insulation properties for use as an in-
sulator in tubular heating elements. In the latter
case, the ceramic is inserted into the cavity be-
tween the heating wire and the outer tubes and
subsequently crushed, compacted, and solidified
by rolling, hammering, and pressing.
Zirconia Ceramics. Zirconia ceramics are
standardized as type C-830 in IEC 672. They
can only be used as insulators at room temper-
ature because the oxygen ions of zirconia be-
come highly mobile at higher temperatures and
the material then acts as a semiconductor. This
principle forms the basis of the widespread ap-
plication of zirconia as a sensor for determining
the partial pressure of oxygen.
3.1.2.6. Insulating Materials Containing
Titanium Oxide and Titanates
Whereas the materials discussed so far are gen-
erally employed purely as insulators, the use of
components containing titanium dioxide and ti-
tanates is governed by their high relative permit-
tivity.
Material Types. Group C-300 of IEC 672 is
subdivided into seven types:
C-310 titanium dioxide
C-320 magnesium titanate
C-330 titanium dioxide incorporating
other oxides as additives
C-331 as for type C-330
C-340 strontium, calcium, and bismuth
titanates
C-350 barium titanate
C-351 barium titanate
Insulation, Electric 21
The materials differ with respect to their rela-
tive permittivity, the associated temperature co-
efficient, and the dielectric dissipation factor.
Types C-310, C-320, C-330, C-331, and C-340
have been classified according to international
regulations as Type I capacitor materials; ac-
cording to DIN 41 920, they are subdivided on
the basis of the temperature coefficients of their
relative permittivities. Material types C-350 and
C-351 have a high relative permittivity and are
classified as Type II substances.
Titanium Dioxide and Mixed Dielec-
trica Based on Rutile. Titanium dioxide
[13463-67-7] crystallizes in the rutilestructure,
which is always formed at temperatures above
1000 ◦ C evenwhen the metastable anatase is
used as a starting material.
The relative permittivity of rutile is strongly
anisotropic. In polycrystalline materials with
random crystal orientation, the relative permit-
tivity has an average value of 117. The temper-
ature coefficient of the relative permittivity is
highly negative (− 0.001 K−1 ).
Properties (see Table 5) can be modified by
additives, such as zirconium oxide, lanthanum
oxide, or magnesium titanate. These additives
possess considerably lower relative permittiv-
ities and positive temperature coefficients. By
following a logarithmic mixing relationship for
relative permittivity, or a linear mixing rela-
tionship for the temperature coefficient, specific
materials with tailored properties can be ob-
tained. The temperature coefficient and the rela-
tive permittivity decrease with increasing addi-
tive concentrations. Dissipation factors are low
(≤10−3 ); with mixed dielectrics containing zir-
conium oxide, they are independent of frequency
up to 108 Hz.
Dielectrics of this type are used to compen-
sate for the increase in capacitance with temper-
ature produced by other components. The pri-
mary criterion for the selection of the dielectrics
is the magnitude of the negative temperature co-
efficient of the relative permittivity rather than
the permittivity itself.
Materials Containing Titanates. The
properties of titanate-based insulating materi-
als are listed in Table 6.
Type C-320 IEC 672 comprises materials
based on magnesium titanate. The outstanding
22 Insulation, Electric
Table 5. Important physical properties of dielectrics based on titanium dioxide classified according to IEC 672

Type C-310 C-330 C-331

Open porosity, vol% 0 0 0

Bulk density, g/cm3 >3.5 >4 >4.5
Bending strength, N/mm >70 >80 >80
Electrical strength,
kV/mm >8 >10 >10
Withstand voltage
(1 min), kV >15 >15 >15
Relative permittivity 40 – 100 25 – 50 30 – 70
Dielectric dissipation
factor (tan δ)
1 kHz <0.0065 <0.020 <0.007
1 MHz <0.002 <0.0008 <0.001
Resistivity at 20 ◦ C, Ω · cm >1012 >1011 >1011
Temperature coefficient −280 to +70 to −120 to
of the relative per- −900 −120 −700
mittivity, 10−6 K−1

Table 6. Important physical properties of titanate-based dielectrica classified according to IEC 672

Type C-320 C-340 C-350 C-351

Open porosity, vol% 0 0 0 0

Bulk density in raw state, g/cm3 >3.1 >3.0 >4.0 >4.0
Bending strength, N/mm2 >70 >70 >50 >50
Electrical strength, kV/mm >8 >6 >2 >2
Withstand voltage (1 min), kV >15 >8 >2 >2
Relative permittivity 12 – 40 100 – 700 350 – 3000 >3000
Dielectric dissipation factor (tan δ) at 1 MHz <0.0015 <0.005 <0.035 <0.035
Volume resistivity at 20 ◦ C, Ω · cm >1011 >1011 >1010 >1010
Temperature coefficient of the relative permittivity, +130 to −150 −1200 to −6000 nonlinear nonlinear
10−6 K−1

feature of these materials is the excellent tem-
perature stability of the relative permittivity, the
values of the latter lie between 12 and 40.
Type I capacitors with extremely high nega-
tive temperature coefficients for the relative per-
mittivity (up to 0.006 K−1 ) can be obtained with
materials based on strontium and/or calcium ti-
tanate (type C-340 in IEC 672).
Barium titanate [12047-27-7] , BaTiO3 , is
the most important alkaline-earth titanate
(→ Ceramics, Electronic). It also finds applica-
tion as a piezoelectric material.
Because of their piezoelectric properties,
considerable interest has also been centered on
titanates that contain lead in the form of mixed
crystals with zirconates, stannates, and niobates.
Like all of the other titanates listed above,
barium titanate crystallizes to give a deformed
perovskite structure. Slight distortions lead to
a tetrahedal structure at room temperature, a
rhombic structure below 0 ◦ C, and trigonal sym-
metry below −90◦ C. The most significant prop-
erty of the distorted structure without a center of
symmetry is its extremely large relative permit-
tivity, particularly in the direction perpendicular
to the main crystal axis (Fig. 11). Maximum rel-
ative permittivity values are observed at each of
the crystallographic transition temperatures, but
are accompanied by a maximum in the dielectric
loss (Fig. 12). This results in a strong, nonlinear
temperature dependence of the relative permit-
tivity. The dipoles in the deformed crystal inter-
act mutually and are aligned in parallel in small
areas known as domains. The direction of dipole
orientation in neighboring domains is reversed
so that the whole system assumes an energy min-
imum and does not have a net external charge.
On applying an external electrical field, the do-
mains become oriented in the direction of the
field to an extent proportional to the electrical
field strength. Because of the anisotropy of the
dielectric properties, this reorientation of the do-
mains is coupled with a change in the relative
permittivity, i.e., the relative permittivity is also
dependent upon field strength. The combination
of phenomena discussed here is known as fer-

roelectricity (see also → Ceramics, Electronic,
Chap. 3.).
Above a certain temperature known as the
Curie temperature (in the case of barium titanate
120 ◦ C), the structure becomes strictly cubic,
and the ferroelectric properties are lost. Above
the Curie temperature T c , the changes in the rel-
ative permittivity with temperature are described
by the Curie – Weiss law:
C losses, such dielectrics are employed as contact-
εT =
T − Tc
where C is the Curie constant and εT is the rel-
ative permittivity at temperature T . Compounds
exhibiting a perovskite structure tend to form
mixed crystals. Hence Ba2+ can be replaced
by Sr2+ , Pb2+ , and to a certain extent also
by Ca2+ ; Ti4+ can be substituted by Zr4+ or
Sn4+ . Similarly, ion combinations can be incor-
porated providing the ionic radii are equivalent
and a balance of charge is attained (complex per-
ovskites). These substitutions produce changes
in the relative permittivity and its temperature
dependency, as well as in the dielectric loss and
the Curie temperature. In this way, a wide range
of insulating materials with differing properties
can be produced.
Figure 11. Relative permittivity of barium titanate along the
a and c crystal axes as a function of temperature
Figure 12. Temperature dependence of the relative permit-
tivity (εr ) and the dielectric loss (tan δ) of polycrystalline
barium titanate
Insulation, Electric 23
Certain oxide additives (e.g., a small ex-
cess of titanium dioxide) act as grain boundary
phases which prevent crystal growth. In small
crystals, ferroelectric domains are still formed,
but their changes of position and size in an ex-
ternal field are considerably hindered. Materials
of this kind show reduced hysteresis loss, and
the curve of εT = f (T ) flattens out in the vicinity
of the Curie temperature. Because of their low
protective capacitors. Extremely small crystals
(ca. 50 nm) cannot be reorientated in an external
electric field and hence do not exhibit ferroelec-
tric properties.
Other additives (e.g., bismuth oxide, bismuth
titanate, zinc oxide, or zinc titanate) extend the
firing interval, but also change the shape of the
curve εT = f (T ). Bismuth titanate additives pro-
duce dielectric materials of low temperature de-
pendency in their electrical properties.
If a dielectric is made from two perovskite
materials that do not form or only partially form
mixed crystals, then the change in the relative
permittivity between the relevant Curie tempera-
tures is low. Examples of such dielectrics include
those containing calcium and barium titanate.
In titanate-based products, the raw materi-
als are thoroughly mixed and calcined onto a
frit, which is subsequently ground, shaped, and
fired. In order to prevent chemical reduction of
the titanium, which would cause the article to
have semiconductor properties, capacitor mate-
rials must be fired in an oxidizing atmosphere.
The electrodes and contact layers are formed by
depositing metal coatings, generally silver.
Types C-350 and C-351 in IEC 672 have a
relative permittivity below and above 3000, re-
spectively.
The main use of titanate-containing di-
electrics is as capacitors in high-frequency ap-
plications. Examples include the bridging of d.c.
resistors and the manufacture of capacitive cou-
plers. Furthermore, perovskite-based materials
containing titanate have gained considerable sig-
nificance as piezoelectric components.
When titanates are fired under reducing con-
ditions, semiconductor materials are obtained
which reoxidize at the interfacial regions dur-
ing the firing of metallizing layers. This process
is used to produce intergranular barrier layer ca-
pacitors which possess particularly high capac-
24 Insulation, Electric
ities due to the low thickness of the dielectric
layer (Type III dielectrics).
3.1.2.7. Nonoxide Ceramics
More recently, increased interest has been shown
in nonoxide inorganic materials because of sev-
eral unusual properties. They have a very high
mechanical strength, even at elevated temper-
ature. The most important materials are com-
pounds of elements of groups 3, 4, and 5 of the
periodic table. Silicon nitride (Si3 N4 ), silicon
carbide (SiC), and aluminum nitride (AlN) de-
serve special mention.
The essentially homopolar bonding charac-
ter between the atoms of these compounds de-
termines their physical properties to a large
extent, but also causes problems during the
manufacture of components when using tradi-
tional procedures employed in the ceramic in-
dustry. Production is thus more complicated and
more expensive. Nonoxide ceramic materials
are only suitable as insulating materials when
their extraordinary properties provide a particu-
larly elegant solution for a given problem. Sil-
icon nitride [12033-89-5] andaluminum nitride
[24304-00-5] are used as electrical insulators.
Silicon carbide is not suitable for electrical
insulation purposes because its electrical con-
ductivity increases strongly with temperature; it
is even employed as a high-temperature heating
element.
The thermal conductivity of aluminum ni-
tride is considerably greater than that of alu-
minum oxide. The low density of 3.2 g/cm3
makes it a suitable alternative to beryllium oxide
when the effective removal of heat generated by
dielectric losses is important. At increasing tem-
peratures, the decrease in thermal conductivity
of aluminum nitride is considerably lower than
that of the oxide materials. A further advantage
of aluminum nitride, especially as a substrate for
high-power silicon components, is the fact that
the thermal expansion coefficient is closer to that
of silicon than any other potential substrate ma-
terials.
Silicon nitride exhibits similar insulation
properties to aluminum oxide, but its thermal
expansion coefficient is only half as large, and
correspondingly shows greater thermal shock re-
sistance. This characteristic is of considerable
importance in glow plugs in diesel engines.
3.1.2.8. Glass (→ Glass)
Glass materials have been used as insulators in
electrical engineering for many years. The sys-
tematic classification and establishment of min-
imum requirements occurred during their incor-
poration in the IEC standard 672.
Material Types. Glass materials used as
electrical insulators can be classified into six
groups:
G-100 alkali – lime – silica glasses
G-200 chemically resistant borosilicate
glasses
G-300 electrically resistant borosilicate
glasses
G-400 alumina – lime – silica glasses
G-500 lead oxide – alkali – silica glasses
G-600 barium oxide – alkali – silica
glasses
Group G-100 consists of:
Type G-110 alkali – lime – silica glasses, an-
nealed
Type G-120 alkali – lime – silica glasses, tough-
ened
The mechanical strength is increased tremen-
dously by prestressing (Type G-120).
Group G-300 consist of:
Type G-310 borosilicate glasses, low loss
Type G-320 borosilicate glasses, stable to high
voltage
Properties. Several of the requirements
specified in the IEC 672 are summarized in Ta-
ble 7. A range of properties of glass are governed
by its chemical composition. Characteristic data
can be determined by multiplying the content of
a glass component (expressed in percent) by a
property-specific factor, and summing the cal-
culated values. Properties that can determined
in this way include the thermal expansion and
the surface energy.
Production. Glass is produced by melting
appropriate mixtures of raw materials and mold-
ing them in the molten state (→ Glass). On cool-
 Insulation, Electric 25
Table 7. Important physical properties of glass-insulating materials classified according to IEC 672. Figures in bold face indicate properties
that are of importance for intended applications.
Property G-110 G-120 G-200 G-310 G-320 G-400 G-500 G-600

Bending strength, N/mm2 >30 >150 >30 >30 >30 >40 >30 >30
Linear expansion
coefficient, 10−6 K−1 8.5 – 10 8.5 – 10 3–5 4.6 – 5.1 4.6 – 5.5 4 – 4.6 8 – 10 9 – 10
Glass transition
temperature, ◦ C 500 – 560 500 – 560 520 – 560 480 – 510 620 – 730 430 – 470 430 – 500
Electrical strength, kV/mm >25 >25 >30 >30 >30 >30
Relative permittivity
at 20 ◦ C and 1 MHz 6.5 – 7.6 7.3 – 7.6 4.0 – 5.5 4.9 – 5.5 5–6 5.5 – 7.5 6–8 6.5 – 7.5
Dielectric dissipation factor
at 20 ◦ C
50 Hz <0.030 <0.060 <0.020 <0.0035 <0.030 <0.0025 <0.003 <0.004
1 MHz <0.010 <0.060 <0.010 <0.002 <0.008 <0.003 <0.002 <0.0025
Volume resistivity, Ω · cm
at 20 ◦ C >1012 >1012 >1014 >1014 >1014 >1014 >1014 >1014
at 200 ◦ C >107 >107 >109 >1010 >108 >1012 >1010 >1010

ing, the melt vitrifies to form the desired com-
ponent. Residual stress can be removed by an-
nealing. However, compressive stresses can also
be deliberately generated in order to increase the
mechanical strength of the component.
Insulator components are mainly shaped by
pressing the viscous melt in steel molds, al-
though recently, extrusion has been introduced
as an additional shaping method. Glass can also
be processed by blowing.
Uses. The applications for glass in electrical
insulator are determined by properties that can
be particularly well optimized in certain types
of glass; these properties are indicated by bold
type in Table 7.
Glass materials from group G-100 are fre-
quently employed for overhead line insulators.
Cap insulators are most common and are com-
bined to form long chains. Long rod insulators
are also used but are more difficult to manufac-
ture.
Glass insulators have been increasingly em-
ployed with overhead lines. Their major advan-
tage is that faults can easily be identified because
they are transparent; slightly defective compo-
nents can therefore be exchanged before a major
failure occurs.
The high requirements placed on the mechan-
ical properties of suspension insulators for over-
head lines are met by toughened glass. Glass
ceramics (which have not yet been included in
the IEC standard) have also been recently em-
ployed because they have a high strength com-
bined with a low thermal expansion coefficient
and correspondingly an improved thermal shock
resistance (→ Glass Ceramics).
Glass materials of type G-200 (chemically re-
sistant borosilicate glass) are used in discharge
lamps. The low coefficient of thermal expansion
resulting from the use of boron oxide ensures
that the glass has a high thermal shock resis-
tance. Boron also improves chemical resistance.
Type G-310 glass materials exhibit favorable
properties for lead through insulators in high-
frequency applications. They are also suitable
for glass components designed to have capacitor
functions. Type G-320 materials are employed
in valves for high-voltage applications, e.g., for
X-ray tubes.
Type G-400 glass materials were devel-
oped to have optimum properties for high-
performance halogen lamps; their properties
have now been standardized.
Type G-500 and G-600 glasses are employed
when metallic components, particularly those
based on nickel – iron alloys, have to be sealed
into electric lamps and valves. They have a low
glass transition temperature and a large thermal
expansion coefficient to ensure that only weak
mechanical stresses are generated in the com-
posite system. Because of the incorporation of
heavy metal oxides, they are also used in tele-
visions to absorb the generated X-rays. Glass
materials containing lead can change on expo-
sure to X-rays and electron beams, and thus be-
come discolored or even produce trackìng cur-
rents. In applications where these effects must be
excluded, lead-free glasses of type G-600 must
be employed.
26 Insulation, Electric
3.2. Organic Insulating Materials
The following discussion of the most common
organic materials employed as electrical insula-
tors includes products that meet all or many of
the imposed requirements. This does not mean
that all of the products are employed in large
amounts in the electronics industry. The com-
pilation is by no means complete, and is in-
tended merely to provide an overview. Descrip-
tion of the products has been kept brief. Details
are available from manufacturers’ information
and specialist publications. Examples of trade
names are given for each group of polymers but
these lists are by no means exhaustive. Most of
the polymers are described in detail elsewhere
in the encyclopedia. Emphasis has been placed
on graphic representations–the relief diagrams.
As opposed to tables, these diagrams allow the
rapid visual assessment of the suitability of a ma-
terial for a given electrotechnical project. Differ-
ent materials can also be compared. The block
lengths demonstrate the quality variation of a
particular material class. Each diagram shows:
1) the comparative tracking index (CTI) mea-
sured according to IEC 112 (see Section 2.1)
2) the volume resistivity () measured accord-
ing to IEC 93 at 23 ◦ C and 50 % R.H. (see
Section 2.2)
3) the surface resistivity (σ) measured accord-
ing to IEC 93 at 23 ◦ C and 50 % R.H. (see
Section 2.3)
4) the loss index ε r measured according to IEC
250 at 1 kHz, 23 ◦ C, and 50 % R.H. (see Sec-
tion 2.6)
5) and the relative permittivity εr measured ac-
cording to IEC 250 at 1 kHz, 23 ◦ C, and 50 %
R.H. (see Section 2.4)
The electric strength is not shown because
although the test results obtained in accordance
with the standard can be used for detecting
changes or deviations from normal characteris-
tics resulting from processing variables, aging
conditions, or other manufacturing or environ-
mental situations, they can seldom be used di-
rectly to determine the behavior of insulating
materials in a real application. Measured val-
ues of the electric strength are affected by many
factors (Section 2.7), but these factors are not
specified when the results are published.
The significance and determination of the pa-
rameters presented in the graphics are described
in Chapters 1, 2, and 4. Values can easily be read
from the plots. The dielectric dissipation factor
is obtained by dividing the loss index (column
4) by the corresponding value for relative per-
mittivity (column 5).
3.2.1. Thermoplasts
Thermoplasts, also referred to as plastomers, are
solid synthetic resins which can be repeatedly
softened without chemical modification by heat-
ing in a relatively narrow temperature range.
They can be shaped while soft and generally
consist of linear molecules. Physical properties
of the thermoplasts are compiled in Table 8.
Acrylonitrile – Butadiene – Styrene. Phys-
ical properties are given in Table 8 and electrical
properties in Figure 13 (see also → Polystyrene
and Styrene Copolymers).
Figure 13. Relief diagram of electrical properties of acry-
lonitrile – butadiene – styrene
Special properties: high impact toughness,
thermally stable, unlimited shaping character-
istics, limited stability to UV radiation.
Uses : casings for office machinery, electrical
industry, electronics, telecommunications, pro-
tective insulating components.
Trade Names: Cycolac (Borg Warner Chemi-
cals, USA); Kralastic (USS Chemicals, USA);
Novodur (Bayer, FRG); Terluran (BASF, FRG);
Diapet (Mitsubishi Rayon Co., Japan); Ravikral
(EniChem, Italy).
Acrylonitrile – Styrene – Acrylic Ester.
Physical properties are given in Table 8 and
electrical properties in Figure 14.
 Insulation, Electric 27
Table 8. Physical properties of thermoplasts

Polymer CAS registry Density, Tensile strength, Ultimate strain, Long term
number g/cm3 N/mm2 % thermal
stability, ◦ C

Acrylonitrile – butadiene – styrene copolymer [9003-56-9] 1.1 45 20 80

(ABS)
Acrylonitrile – styrene – acrylic ester (ASA) [9003-54-7] 1.1 50 15 85
Cellulose acetate (CA) [9004-35-7] 1.2 35 30 40
Cellulose acetate butyrate (CAB) [9004-36-8] 1.3 35 30 40
Cellulose propionate (CP) [9004-64-2] 1.2 20 – 50 4 40
Perfluoro(ethylene – propylene) (FEP) [25067-11-2] 2.2 20 300 70
Polyamide (PA) 1.1 60 200 60 – 90
Polyamide 12-resin (PA) [24937-16-4] 1 55 200 150
Polyamideimide (PAI) 1.4 220 6 220
Polybutene (PB) [9003-29-6] 0.92 10 15 100
Poly(butylene terephthalate) (PBTP) [24968-12-5] 1.3 60 5 – 200 120
Polycarbonate (PC) [25037-45-0] 1.2 60 60 – 100 90
Polycarbonate alloy (PC – ABS) 1.2 60 60 – 100 90
Polycarbonate alloy (PC – PBTP) 1.3 60 60 – 100 90
Polyethylene (PE) [9002-88-4] 0.92 – 0.96 9 – 23 650 60 – 80
Cross-linked polyethylene (PE-X) 1.3 9 (low density) 350 130
23 (high density)
Polyetherimide (PEI) [61128-24-3] 1.3 105 60 170
Polyethersulfone (PES) [25667-42-9] 1.4 85 20 200
Poly(ethylene terephthalate) (PETP) [25038-59-9] 1.4 240 90 130
Polyimide (PI) [25036-53-7] 1.4 90 10 250
Poly(methyl methacrylate) (PMMA) [9011-14-7] 1.2 80 5 70
Poly(4-methyl-1-pentene) (PMP) [9016-80-2] 0.8 25 10 – 50 150
Polyoxymethylene (POM) [9002-81-7] 1.4 – 1.7 2 – 20 30 100
Polypropylene (PP) [9003-07-0] 0.9 30 600 100
Poly(phenylene oxide) (PPO) [9041-80-9] 1.1 60 30 175
Poly(phenylene sulfide) (PPS) [9016-75-5] 1.7 200 2 240
Polystyrene (PS) [9003-53-6] 1.1 60 3 60
Polysulfone (PSU) [25135-51-7] 1.2 85 2 – 70 160
Polytetrafluoroethylene (PTFE) [9002-84-0] 2.2 25 300 260
Poly(vinyl chloride) (PVC) [9002-86-2] 1.4 55 30 60
Styrene – acrylonitrile copolymer (SAN) [9003-54-7] 1.1 70 4 85
Styrene – butadiene copolymer (SB) [9003-55-8] 1.1 40 35 60

Trade Names: Richform (Richmond Div. of Dix-

ico, USA); Luran S (BASF, FRG); Geloy (Gen-
eral Electric, USA).

Cellulose Acetate. Physical properties are

given in Table 8 and electrical properties in Fig-
ure 15 (see also → Cellulose Esters).

Figure 14. Relief diagram of electrical properties of acry-

lonitrile – styrene – acrylic ester

Special properties: stable to UV radiation,

weather-resistant, thermally stable over long pe-
riods, unlimited shaping characteristics.
Uses: insulators, insulating material, ventila-
tors, casings for electronic equipment, electrical
industry, electronics, telecommunications, pro-
Figure 15. Relief diagram of electrical properties of cellu-
tective insulating components. lose acetate
28 Insulation, Electric
Figure 16. Relief diagram of electrical properties of cellu-
lose acetate butyrate
Special properties: antistatic properties de-
spite high electrical resistance, crystal clear,
tough, hard, scratch-resistant, insensitive to
stress cracking, readily dyeable with brilliant
colors, not permanently weather-resistant, less
suitable for precision components.
Uses: tool handles, electric insulation films,
lights, casings.
Trade Names: Dexel (Courtaulds Plastic Group,
UK); Ampol (American Polymers, USA); Ten-
ite (Eastman Chemical International, USA).
Cellulose Acetate Butyrate. Physical prop-
erties are given in Table 8 and electrical proper-
ties in Figure 16 (see also → Cellulose Esters).
Special properties: antistatic despite high
electrical resistance, crystal clear, tough, hard,
scratch resistant, insensitive to stress-cracking,
readily and brilliantly dyeable, not suitable for
precision components.
Uses: tool handles, electric insulation films,
lights, casings.
Trade Names: Ampol (American Polymers,
USA); Tenite (Eastman Chemical International,
USA).
Cellulose Propionate. Physical properties
are given in Table 8 and electrical properties in
Figure 17 (see also → Cellulose Esters).
Special properties: antistatic despite high
electrical resistance, crystal clear, tough,
scratch-resistant, insensitive to stress cracking,
readily dyeable with brilliant colors, not suitable
for precision components.
Uses: tool handles, electric insulation films,
lights, casings.
Figure 17. Relief diagram of electrical properties of cellu-
lose propionate
Trade Names: Ampol (American Polymers,
USA); Tenite (Eastman Chemical International,
USA).
Perfluoro(Ethylene – Propylene). Phys-
ical properties are given in Table 8 and
electrical properties in Figure 18 (see also
→ Fluoropolymers, Organic, Chap. 2.3.).
Special properties: nonflammable, weather-
resistant, resistant to solvents, no water uptake,
physiologically inert, stable to pressure.
Uses: flexible printed circuits, coaxial ca-
bles for computers, cable insulation, shrinkable
tubes.
Trade Names: Teflon (Du Pont, USA); Neoflon
(Daikin Kogyo Co., Japan).
Polyamide. Physical properties are given in
Table 8 and electrical properties in Figure 19
(see also → Polyamides).
Special properties: normal minimum water
content of 2.5 %, prone to stress cracking when
freshly injected, impact resistant, viscoelastic,
puncture resistant, dynamic-loading capability,
good slip properties.
Uses: high-voltage insulation, multiple con-
necting plugs, relays, extinguishing chambers,
wear-resistant insulation and sleeve materials
for cable technology.
Trade Names: Ultramid (BASF, FRG);
Durethan (Bayer, FRG); Technyl (Rhône-
Poulenc Specialities Chimiques, France); Grilon
(Ems-Chemie, Switzerland); Zytel (Du Pont,
USA); Maranyl (ICI, UK).
Polyamide 12-Resin. Physical properties
are given in Table 8 and electrical properties
in Figure 20.
 Insulation, Electric 29

Figure 19. Relief diagram of electrical properties of

Figure 18. Relief diagram of electrical properties of perflu- polyamide
oro(ethylene – propylene)

Figure 21. Relief diagram of electrical properties of

Figure 20. Relief diagram of electrical properties of polyamideimide
polyamide 12-resin

Special properties: resistant to organic sol-
vents, fuels, oils, weak acids and alkalis.
Uses: casings, components for electrical de-
vices.
Trade Names: Vestamid (Hüls, FRG); Ril-
san (Atochem, France); Grilamid (Ems-Chemie,
Switzerland).
ditions, printed circuits, suitable for solderable
connecting cable.
Polybutene. Physical properties are given in
Table 8 and electrical properties in Figure 22
(see also → Polyolefins).

Polyamideimide. Physical properties are

given in Table 8 and electrical properties in Fig-
ure 21.
Special properties: resistant to elevated and
cryogenic temperatures; nonmeltable; resistant
to hydrocarbons, weak acids, and weak alkalis;
resistant to highly energetic radiation sources;
very low coefficient of friction.
Uses: special material for relays and multi-
ple connection plugs used under extreme con- Figure 22. Relief diagram of electrical properties of poly-
butene
30 Insulation, Electric
Figure 23. Relief diagram of electrical properties of
poly(butylene terephthalate)
Special properties: high long-term stability,
low tracking propensity at high temperature,
high stress cracking stability, resistant to low
temperature, easily weldable.
Uses: insulation jackets for wires, cable
sleeves, pipes.
Trade Names: Duraflex (Shell International
Chemical Co., UK); Krene (Union Carbide,
USA).
Poly(Butylene Terephthalate) (PBT).
Physical properties are given in Table 8 and
electrical properties in Figure 23 (see also
→ Polyesters).
Special properties: impact toughness, rigid,
maintains shape, scratch-resistant, good slip
and abrasion behavior, electrical properties
are essentially frequency- and temperature-
independent.
Uses: electrical insulation in household
equipment, relay capstans, connecting cable,
components for switches, sparking plug cases.
Trade Names: Hostadur (Hoechst, FRG); Pocan
(Bayer, FRG); Vestodur (Hüls, FRG); Ultradur
(BASF, FRG); Celanex (Celanese Engineering
Resins, USA); Novadur (Mitsubishi Chemical
Industries, Japan).
Polycarbonate (PC). Physical properties
are given in Table 8 and electrical properties
in Figure 24 (see also → Polycarbonates).
Special properties: low combustion, rigid
elastic, high impact toughness above 90 ◦ C,
long term stability at high temperature, limited
Figure 24. Relief diagram of electrical properties of poly-
carbonate
suitability under abrasive conditions, electrical
properties practically independent of humidity.
Uses: moldings, casings, appliances, relay
elements, condenser housings, plug connec-
tions, high-voltage insulation, cable distribu-
tion panels, mounting boards, screens, compact
discs, information storage media.
Trade Names: Makrolon (Bayer, FRG); Orgal-
lan (Atochem, France); Merlon (Mobay Chem-
ical Corp., USA); Lexan (General Electric,
USA); Jupilon (Mitsubishi Gas Chem, Japan);
Sinvet (EniChem, Italy).
Polycarbonate Alloy (PC – ABS). Physical
properties are given in Table 8 and electrical
properties in Figure 25.
Figure 25. Relief diagram of electrical properties of poly-
carbonate alloy (PC – ABS)
Special properties: resistant to combustion,
rigid elastic, tough above 90 ◦ C, long term sta-
bility at high temperatures, limited suitability
under abrasive conditions, electrical character-
istics practically independent of humidity.
 Insulation, Electric 31

Figure 26. Relief diagram of electrical properties of poly-

carbonate alloy (PC – PBTP) Figure 27. Relief diagram of electrical properties of
polyethylene

Uses: moldings, casings, appliances, relay
elements, condensor housings, plug connec-
tions, high-voltage insulation, cable distribution
panels, mounting boards, screens.
Trade Names: Cycolac (Borg Warner Chemi-
cals, USA); Bayblend (Bayer, FRG); Cycoloy
(Borg Warner Chemicals, USA).
Alathone (Du Pont, USA); Novatec (Mit-
subishi Chemical Industries, Japan); Dylan
(Arco Chemical Co., USA).
Cross-linked Polyethylene. Physical prop-
erties are given in Table 8 and electrical proper-
ties in Figure 28.

Polycarbonate Alloy (PC – PBTP). Physi-

cal properties are given in Table 8 and electrical
properties in Figure 26.
Special properties: resistant to combustion,
rigid elastic, tough above 90 ◦ C, long term sta-
bility at high temperatures, limited suitability
under abrasive conditions, electrical character-
istics practically independent of humidity.
Uses: moldings, casings, consumer equip-
ment, relay elements, condensor housings, plug
connections, high voltage insulation, cable dis-
tribution panels, mounting boards, screens.
Trade Names: Makroblend (Bayer, FRG); Figure 28. Relief diagram of electrical properties of cross-
Xenoy (General Electric, USA). linked polyethylene

Polyethylene. Physical properties are given
in Table 8 and electrical properties in Figure 27
(see also → Polyolefins).
Special properties: resistant to water, saline
solutions, alkalis, acids, (exception: strong oxi-
dizing agents), can be subjected to high mechan-
ical stress, hard, resistant to low temperatures.
Uses: insulation and sleeve materials for ca-
bles, pipes, with 30 % carbon black deflects lines
of force in conductors, transformers, etc.
Trade Names: Lupolen (BASF, FRG); Car-
lona (Shell International Chemical Co., UK);
Special properties: stable at high tempera-
ture, high impact strength at low temperature,
resistant to stress cracking, greater resistance to
elevated temperatures than polyethylene.
Uses: insulation and sleeve materials for ca-
bles, shrinkable sleeves.
Trade Names: Lubonyl (N. Lundbergs Fab-
rics AB, Sweden); Alveolit (Sekisui Chemical
Co., Japan); Alveolux (Alveo AG, Switzerland);
Blanex (Reichhold Chemicals Inc., USA); Ela-
por (Elastogran, FRG).
32 Insulation, Electric
Polyetherimide. Physical properties are
given in Table 8 and electrical properties
in Figure 29 (see also → Polymers, High-
Temperature).
Figure 29. Relief diagram of electrical properties of
polyetherimide
Special properties: transparent yellow; high
heat resistance; very high resistance to combus-
tion; resistant to alcohols; fuels; lubricants; de-
tergents, acids and weak alkalis; resistant to UV
light, high energy sources of radiation, hydroly-
sis, hot water, and steam.
Uses: casings, equipment, protective relays
for high-voltage supplies, multiple connecting
plugs, soldered connecting cables, components
for integrated circuits.
Trade Names: Ultem (General Electric Co.,
USA).
Polyethersulfone. Physical properties are
given in Table 8 and electrical properties
in Figure 30 (see also → Polymers, High-
Temperature).
Figure 30. Relief diagram of electrical properties of
polyethersulfone
Special properties: transparent yellow, high
long-term stability, low tracking propensity
until the glass transition temperature is ap-
proached, resistant to acids and alkalis, not sus-
ceptible to hydrolysis, resistant to oxidation,
highly incombustible, dielectric values essen-
tially temperature- and frequency-independent.
Uses: involving high demands in electrical
engineering, insulation films.
Trade Names: Udel (Amaco Chemicals, USA);
Ultrason (BASF, FRG); Victrex (ICI, UK).
Poly(Ethylene Terephthalate). Physical
properties are given in Table 8 and electrical
properties in Figure 31 (see also → Polyesters).
Figure 31. Relief diagram of electrical properties of poly-
(ethylene terephthalate)
Special properties: tough, stiff, maintains
shape at high temperature, good slip and abra-
sion behavior, electrical properties are only
slightly temperature- and frequency-dependent.
Uses: insulation and sleeve materials for ca-
bles, sonic substrates, insulation materials, insu-
lation films.
Trade Names: Hostaphan (Hoechst, FRG);
Crastin (Ciba-Geigy, Switzerland); Melinar
(ICI, UK); Rynite (Du Pont, USA).
Polyimide. Physical properties are given in
Table 8 and electrical properties in Figure 32
(see also → Polyimides).
Special properties: high thermal stability, re-
sistant to most dilute or weak acids, attacked by
strong alkalis and ammonia, permanent humid-
ity may cause cracking, pronounced resistance
to high-energy radiation.
Uses: electrical and electronic components,
films, shrinkage films.

Figure 32. Relief diagram of electrical properties of poly-
imide
Figure 34. Relief diagram of electrical properties of poly(4-
methyl-1-pentene)
Trade Names: Kapton, Vespel (Du Pont, USA);
Kinel (Rhône-Poulenc Specialities Chimiques,
France).
Poly(Methyl Methacrylate). Physical
properties are given in Table 8 and elec-
trical properties in Figure 33 (see also
→ Polymethacrylates).
Special properties: hard, tough, highly trans-
parent, weather-resistant, chemically inert to
medium concentrations of acids, resistant to fu-
els, physiologically inert.
Uses: lamps, traffic lights, nautical aids, opti-
cal fibers, high-pressure mercury lamps, optical
data storage media.
Trade Names: Lucite (Du Pont, USA); Plexiglas
(Röhm, FRG); Altulite (Altulor Groupe CdF
Chimie, France); Sumipex (Sumitomo Chemi-
cal Co., Japan).
Insulation, Electric 33
Figure 33. Relief diagram of electrical properties of
poly(methyl methacrylate)
Figure 35. Relief diagram of electrical properties of poly-
oxymethylene
Poly(4-methyl-1-pentene) (PMP). Physi-
cal properties are given in Table 8 and electrical
properties in Figure 34 (see also → Polyolefins).
Special properties: light weight, crystal
clear, rigid, not suitable for exterior applica-
tions, good thermal stability.
Uses: insulation for coaxial and electrical ca-
bles.
Trade Names: TPX (Mitsubishi Petrochemical
Co., Japan).
Polyoxymethylene (POM). Physical prop-
erties are given in Table 8 and elec-
trical properties in Figure 35 (see also
→ Polyoxymethylenes).
Special properties: high strength and stiff-
ness, high dynamic loading capacity over a wide
temperature range, high impact toughness even
at −40 ◦ C, high surface hardness, low coefficient
of friction, good wear properties, low gas perme-
ability, not resistant to UV light, physiologically
34 Insulation, Electric

Figure 36. Relief diagram of electrical properties of Figure 37. Relief diagram of electrical properties of
polypropylene poly(phenylene oxide)

inert, dielectric characteristics are practically in-
dependent of frequency and temperature.
Uses: precision components for measuring
instruments in electronics and precision engi-
neering, household appliances.
Trade Names: Delrin (Du Pont, USA); Ultra-
form (BASF, FRG); Tenac (Asahi Chemical In-
dustries, Japan).
Trade Names: Luranyl (BASF, FRG); Noryl
(General Electric, USA); Vestoran (Hüls, FRG).
Poly(Phenylene Sulfide). Physical proper-
ties are given in Table 8 and electrical properties
in Figure 38.

Polypropylene. Physical properties are

given in Table 8 and electrical properties in
Figure 36 (see also → Polyolefins).
Special properties: stiff, impact toughness,
no stress cracking, resistant to hot water and
chemicals.
Uses: accumulator boxes, casings for elec-
trotechnical equipment, electrotechnical com-
ponents.
Trade Names: Vestolen P (Hüls, FRG); Hostalen
PP (Hoechst, FRG); Propathene (ICI, UK); Pro- Figure 38. Relief diagram of electrical properties of
fax (Himont, USA); Noblen (Mitsubishi Petro- poly(phenylene sulfide)
chemical Co., Japan).
Special properties: transparent, requires spe-
Poly(Phenylene Oxide). Physical proper- cial reinforcement, low glass transition temper-
ties are given in Table 8 and electrical properties ature, high thermal loading capacity, dielectric
in Figure 37 [see also → Poly(Phenylene Ox- properties essentially independent of tempera-
ides)]. ture and frequency.
Special properties: high long-term stability; Uses: encapsulation of electronic compo-
low tendency to tracking until the glass transition nents, insulating films.
temperature is approached; resistant to acids, al- Trade Names: Ryton (Phillips Petroleum Chem-
kalis, hydrolysis, and oxidation; highly incom- icals, Belgium).
bustible; transparent.
Uses: cable insulation for data processing Polystyrene. Physical properties are given in
equipment and equipment for telecommunica- Table 8 and electrical properties in Figure 39 (see
tions. also → Polystyrene and Styrene Copolymers).

Figure 39. Relief diagram of electrical properties of
polystyrene
Figure 41. Relief diagram of electrical properties of polyte-
trafluoroethylene
Special properties: crystal clear; high gloss;
brittle; resistant to humidity, salt solutions, alkali
and nonoxidizing acids; mechanical properties
only slightly time- and temperature-dependent;
dielectric properties are independent of temper-
ature and frequency; susceptible to damage by
UV light.
Uses: encapsulation of electronic compo-
nents, insulating films, dielectric for capacitors.
Trade Names: Styroflex (Norddeutsche Seek-
abelwerke AG, FRG); Vestyron (Hüls, FRG);
Sternite (Chemical Products, UK); Dylene
(Arco Chemical Co., USA); Esbrite (Sumit-
omo, Japan); Toporex (Mitsui Toatsu Chemi-
cals, Japan).
Polysulfone. Physical properties are given in
Table 8 and electrical properties in Figure 40 (see
also → Polymers, High-Temperature).
Insulation, Electric 35
Figure 40. Relief diagram of electrical properties of poly-
sulfone
Figure 42. Relief diagram of electrical properties of
poly(vinyl chloride)
Special properties: transparent yellow; high
mechanical loading capacity; dielectric proper-
ties only slightly temperature- and frequency-
dependent, resistant to acids, bases, oxidation
and hydrolysis; highly incombustible.
Uses: insulation films, circuit boards.
Trade Name: Ultrason (BASF, FRG).
Polytetrafluoroethylene. Physical proper-
ties are given in Table 8 and electrical proper-
ties in Figure 41 (see also → Fluoropolymers,
Organic, Chap. 2.2.).
Special properties: difficult to wet; antiadhe-
sive; favorable coefficient of friction; noncom-
bustible; weather-resistant; physiologically in-
ert; extremely resistant to corrosion, solvents,
acids, and bases; dielectric properties indepen-
dent of temperature and frequency.
36 Insulation, Electric
Uses: films, insulation, insulating and sleeve
material in cables.
Trade Names: Hostaflon TF (Hoechst, FRG);
Teflon (Du Pont, USA); Polyflon (Daikin Ko-
gyo Co., Japan); Fluon (ICI, UK).
Poly(Vinyl Chloride). Physical properties
are given in Table 8 and electrical properties in
Figure 42 [see also → Poly(Vinyl Chloride)].
Special properties: highly incombustible,
rigid, weldable, thermoformable, resistant to
salt solutions, bases, most acids, and fuels.
Uses: insulation, casings, appliance compo-
nents, insulating and sleeve material in cables.
Trade Names: Vestolit (Hüls, FRG); Trosiplast
(Hüls-Troisdorf, FRG); Vinoflex (BASF, FRG);
Duval (Alpha Chemical & Plastics Corp., USA);
Corvic (ICI, UK); Nipolit (Chisso Corp., Japan).
Styrene – Acrylonitrile Copolymer. Physi-
cal properties are given in Table 8 and electrical
properties in Figure 43 (see also → Polystyrene
and Styrene Copolymers).
Figure 43. Relief diagram of electrical properties of
styrene – acrylonitrile copolymer
Special properties: resistant to UV light,
weather-resistant, good long-term stability at el-
evated temperature, good formability.
Uses: insulators, insulation material, ventila-
tors, casings for electrical equipment, electrical,
electronic, telecommunications, protectively in-
sulated components.
Trade Names: Acrylafil (Fiberfil Inc., USA);
Thermocomp (LNP Corp., USA); Luran (BASF,
FRG); Lustran (Monsanto Co., USA); Litac
(Mitsui Toatsu Chemicals Inc., Japan).
Styrene – Butadiene Copolymer. Physical
properties are given in Table 8 and electrical
properties in Figure 44 (see also → Polystyrene
and Styrene Copolymers).
Figure 44. Relief diagram of electrical properties of
styrene – butadiene copolymer
Special properties: impact toughness, long-
term thermal stability, good formability, limited
resistance to UV light.
Uses: casings for office equipment, elec-
trical, electronic, telecommunications, protec-
tively insulated components.
Trade Names: Superflex (Carl Gordon Ind.,
USA).
3.2.2. Thermosets
Thermosets are curable synthetic resins and
polymers that undergo irreversible cross-linking
and become insoluble on heating within a certain
temperature range or on addition of catalysts.
The term also includes cured products which
do not change significantly below their decom-
position temperature. Thermosets can be mod-
ified with fillers. Physical properties are com-
piled in Table 9.
Epoxy Resins. Physical properties are given
in Table 9 and electrical properties in Figure 45
(see also → Epoxy Resins).
Special properties: strong adhesion to met-
als, good flow properties, no material losses,
thermally stable, light colors.
Uses: jacketing, contact substrates in electri-
cal applications, insulation of converters, sub-
strate material for motor armatures, collector
rings in electrical motors and generators.
Trade Names: Beckopox (Hoechst, FRG);
Araldit (Ciba-Geigy, Switzerland); Epodite
(Showa High Polymer Co., Japan); Epocast (M
 Insulation, Electric 37
Table 9. Physical properties of thermosets

Polymer Density, g/cm3 Tensile strength, Ultimate strain, % Long-term thermal

N/mm2 stability, ◦ C

Epoxy Resins (EP) 1.2 60 4 150

Melamine – formaldehyde resins (MF) 1.5 30 1 65 – 100
Phenol – formaldehyde resins (PF) 1.5 25 1 150
Synthetic resin (chip board) 1.2 130 5 100
Polyurethane (PUR) 1.3 45 50 80
Silicone resin (SI) 1.6 90 2 180
Urea – formaldehyde resins (UF) 1.5 30 1 120
Unsaturated polyester (UP) 1.2 60 2 70

Figure 45. Relief diagram of electrical properties of epoxy

resins
Figure 46. Relief diagram of electrical properties of
melamine – formaldehyde resins

& T Chemicals Inc., USA); Epikote (Shell In-

ternational Chemical Co., UK).

Melamine – Formaldehyde Resins. Physi-

cal properties are given in Table 9 and electrical
properties in Figure 46 (see also → Amino
Resins).
Special properties: flame resistant; high me-
chanical loading capacity; resistant to fuels, oils,
solvents, alcohols; low stability to acids and
bases; prone to stress cracking; can be dyed with
light colors.
Uses: electrical installations, electrical com-
ponents, fuse elements. Figure 47. Relief diagram of electrical properties of phe-
Trade Names: Ultrapas (Hüls-Troisdorf, FRG); nol – formaldehyde resins
Tufnol (Tufnol Ltd., UK); Textolite (Gen-
Special properties: insensitive to heat and
eral Electric, USA); Synthane (Synthane-Taylor
humidity, thermally stable, dimensionally sta-
Corp., USA); Nikamelamine (Nippon Carbide
ble, only dark colors.
Ind., Japan).
Uses: insulating material, casings for electri-
cal and electronic applications.
Phenol – Formaldehyde Resins. Physical
Trade Names: Acell (BP Chemicals, UK);
properties are given in Table 9 and electrical
Alveophen (Synthetic Resins Ltd., UK); Bake-
properties in Figure 47 (see also → Amino
lite (Bakelite, FRG); Compolet (Nobel In-
Resins).
38 Insulation, Electric
dustries Sweden, Sweden); Dynapor (Hüls-
Troisdorf, FRG).
Synthetic Resin – Chip Board. Physical
properties are given in Table 9 and electrical
properties in Figure 48.
Figure 48. Relief diagram of electrical properties of syn-
thetic resin – chip board
Special properties: properties are mostly
those of the impregnating agent.
Uses: high-voltage insulation, slot liners.
Trade Names: Delignit (Blomberger Holzindus-
trie, FRG); Pag-Holz (Pag-Preßwerk, FRG).
Polyurethane. Physical properties are given
in Table 9 and electrical properties in Figure 49
(see also → Polyurethanes).
Figure 49. Relief diagram of electrical properties of
polyurethane
Special properties: constant mechanical be-
havior over a wide temperature range; resistant
to surface and sea water; resistant to biologi-
cal contamination; limited resistance to hydrol-
ysis by steam, hot water, acids and bases; phys-
iologically inert when completely cross-linked;
castable; curing takes place even at low temper-
ature.
Uses: cable end caps, converters, transform-
ers, ignition coils.
Trade Names: Elastan (Elastogran, FRG);
Desmodur (Bayer, FRG); Suprasec
(Polyurethanes Group ICI, UK).
Silicone Resin. Physical properties are given
in Table 9 and electrical properties in Figure 50
(see also → Silicones).
Figure 50. Relief diagram of electrical properties of silicone
resin
Special properties: chemically inert, physi-
ologically inert, water-repellent, physical and
electrical properties remain almost unchanged
from very low to high temperatures.
Uses: thermally stable impregnating resins.
Trade Names: Baysilon (Bayer, FRG); Wacker
Silicone (Wacker Chemie, FRG); GE-Silicone
(General Electric, USA); DC-Silicone (Dow
Corning Corp., USA).
Urea – Formaldehyde Resins. Physical
properties are given in Table 9 and electrical
properties in Figure 51 (see also → Amino
Resins).
Special properties: stable to light, resistant
to solvents and oils, but not to acids and bases,
light colors possible.
Uses: electrical installations.
Trade Names: Pollopas (Hüls-Troisdorf, FRG);
Beetle, Scarab (BIP Chemicals, UK); Urochem
(Chemiplastica S.p.A., Italy).
Unsaturated Polyester. Physical properties
are given in Table 9 and electrical properties in
Figure 52 (see also → Polyester Resins, Unsat-
urated).

Figure 51. Relief diagram of electrical properties of urea –
formaldehyde resins
Figure 52. Relief diagram of electrical properties of unsat-
urated polyester
Special properties: meltable, can be cast
when liquid, no material losses, dimensionally
stable, not prone to stress cracking, stable to
light, easy to process.
Uses: mounting electrical components, also
for high-voltage applications.
Trade Names: Palatal (BASF, FRG); Vestopal
(Hüls, FRG); Alpolit (Hoechst, FRG); Crystic
(Scott Bader Co., UK); Glastic (Glastic Corp.,
USA); U-Pica (Japan-Upica, Japan).
3.2.3. Elastomers
Elastomers are natural or synthetic polymers
(→ Rubber, 2. Natural; → Rubber 3., Synthetic)
which exhibit rubberlike behavior between at
least 20 ◦ C and their respective decomposition
temperature. They solidify at or below 0 ◦ C. The
Insulation, Electric 39
rubber – elastic behavior is a result of the tan-
gled arrangement of the molecular chains or of
their degree of cross-linking. Physical properties
of elastomers are compiled in Table 10.
Chloroprene Rubber. Physical properties
are given in Table 10 and electrical properties in
Figure 53.
Figure 53. Relief diagram of electrical properties of chloro-
prene rubber
Special properties: highly incombustible,
leatherlike toughness, weather-resistant,
corrosion-resistant.
Uses: insulating and sleeve material for ca-
bles.
Trade Names: Baypren (Bayer, FRG); Buta-
clor (Rhône-Poulenc Specialities Chimiques,
France); Neoprene (Du Pont, USA).
Ethylene – Propylene Terpolymer. Physi-
cal properties are given in Table 10 and electrical
properties in Figure 54.
Special properties: rubber-elastic, abrasion
resistant, stable towards aging and weathering
(oxygen, ozone, water, UV radiation, high tem-
perature), inert towards many chemicals.
Uses: insulation and sleeve material for ca-
bles, plugs, cable filling, with carbon black
causes deflection of the lines of force in con-
ductors, transformers, etc.
40 Insulation, Electric
Table 10. Physical properties of elastomers

Polymer Density, g/cm3 Tensile strength, Ultimate strain, % Long-term thermal

N/mm2 stability, ◦ C

Chloroprene rubber (CR) 1.3 – 1.7 55 250 60 – 90

Ethylene – propylene – terpolymer 0.9 18 380 140
(EPM – EPDM)
Ethylene – propylene rubber (EPR) 1.3 4 200 90
Ethylene – vinyl acetate rubber (EVA) 1.4 7 200 120
Styrene – butadiene rubber (SBR) 0.94 5 – 25 500 70
Silicone rubber (SI-R) 1.3 4–9 300 – 700 200

Figure 54. Relief diagram of electrical properties of ethy- Figure 55. Relief diagram of electrical properties of ethy-
lene – propylene terpolymer lene – propylene rubber

Figure 56. Relief diagram of electrical properties of ethy- Figure 57. Relief diagram of electrical properties of
lene – vinyl acetate rubber styrene – butadiene rubber

Trade Names: Buna AP (Hüls, FRG); Vistalon
(Exxon Chemical Co., USA).
Ethylene – Propylene Rubber. Physical
properties are given in Table 10 and electrical
properties in Figure 55.
Special properties: resistant to chemicals,
ozone, aging, water, acids, and alkalis; weather
resistant; unstable to mineral oils and organic
solvents.
Uses: insulating and sleeve material for ca-
bles.
Trade Names: Dutral TP (Montepolmeri, Italy);
Intolan (Dow Corning Corp., USA); Vistalon
(Exxon Chemical Co., USA).
Ethylene – Vinyl Acetate Rubber. Physical
properties are given in Table 10 and electrical
properties in Figure 56. Special properties: ther-

mally stable, excellent adhesion, rubber-elastic,
impact resistant.
Uses: insulating and sleeve material for ca-
bles.
Trade Name: Levapren (Bayer, FRG).
Styrene – Butadiene Rubber. Physical
properties are given in Table 10 and electrical
properties in Figure 57.
Special properties: resistant to organic sol-
vents, weldable, limited stability fuels, oils,
acids, and alkalis.
Uses: insulating and sleeve material for ca-
bles, capacitor components, accumulator boxes
components for telecommunications, connec-
tors, insulators.
Trade Name: Buna-Hüls (Hüls, FRG).
Silicone Rubber. Physical properties are
given in Table 10 and electrical properties in Fig-
ure 58.
Figure 58. Relief diagram of electrical properties of silicone
rubber
Special properties: difficult to wet, physio-
logically safe, electrical characteristics show lit-
tle change over large regions.
Uses: insulating and sleeve material for ca-
bles, electric insulation, casting resins.
Trade Names: Silastic (Dow Corning Corp.,
USA); Silopren (Bayer, FRG); Wacker-
Silicongummi (Wacker-Chemie, FRG).
3.2.4. Polymer Processing
Once polymeric materials have been synthesized
and, if necessary, treated with fillers, pigments,
plasticizers, processing aids, and agents to mini-
mize aging and flammability, they are processed
into powders or granulates. Subsequent forming
Insulation, Electric 41
operations result in products that frequently do
not require further physical modification. The
forming procedures most applicable to single
items or small numbers of insulators include re-
shaping and bonding, for example by heat seal-
ing. Thermoplastic materials can also be ma-
chined, including boring and cutting. A molded
polymeric insulator, fashioned to the proper
shape in a single operation without recourse to
subsequent machining, may often replace con-
ventional insulators consisting of multiple parts,
each requiring its own series of machining op-
erations. However, the equipment required to
produce complicated polymeric objects are usu-
ally quite expensive, so economic factors tend
to limit this approach to mass-produced items.
For a more detailed description of the techniques
described below see → Plastics, Processing.
Casting without the Use of Pressure. Cer-
tain polymeric materials can be formed without
the use of pressure by methods such as the rota-
tional casting of hollow objects in heated forms
and centrifugal casting. Such materials include
thermoplastic polymers such as poly(methyl
methacrylate), polyamide 6, poly(vinyl chlo-
ride) mixed with liquid plasticizers, certain elas-
tomers, and thermosetting polymers (casting
resins, e.g., epoxy resins and polyurethanes).
Powder Techniques. The powder method
entails fusing a low-melting polymer powder to a
heated metal surface, for example by whirl sin-
tering. This approach lends itself to materials
such as polyolefins, cellulose acetobutyrate, and
casting resins.
Low-Pressure Casting. Low-pressure cast-
ing is appropriate for polymers reinforced with
glass or carbon fibers. The precursors are
resin mats and prepregs made from unsaturated
polyesters, epoxides, and other casting resins.
These are converted to their final form either by
the hand lay-up or the plane-type press process.
A mixture of resin and fibers can also be sprayed
at low pressure.
Casting under Thermoplastic Conditions.
The thermoplastic casting method plays an im-
portant role in polymer processing. It requires a
temperature in the range of 150 – 250 ◦ C and a
pressure of several hundred bar.
42 Insulation, Electric
Calendering denotes the process of rolling
a polymeric mass in the plastic state until
it reaches a desired thickness, usually 0.06 –
0.6 mm.
Extruding. The most versatile of the pro-
cesses used for forming thermoplastic materials
is extrusion with a continuous screw extruder.
Proper choice of dies permits the manufacture
of films, pipes, tubes, sheets, slabs, and even
items of more complex cross-sectional shape.
Screw extruders are often coupled with calibra-
tion devices, smoothing rolls, and cooling facil-
ities. With this approach it is possible to com-
pletely surround a metallic wire, conductor, or
cable with a sprayed plastic insulator.
Pressing. Thermosetting laminates (e.g.,
laminated paper, glass – fabric laminates, cotton
laminates, melamine – formaldehyde laminates,
or phenolformaldehyde – fiberglass fabrics) are
prepared with the aid of a multiplaten press in
which the carrier sheet is continuously soaked
with a resin solution. Suitably sized pieces of the
material are then stacked in layers and warmed
under pressure in a heated press, causing the
resin to flow and subsequently solidify. Provided
the material is cooled while still under pressure,
such a press can also be used to produce ther-
moplastic blocks of any desired thickness with
a high-quality surface finish.
Injection Molding. Injection molding ma-
chines operate on the batch principle. A heated
injection cylinder operates in tandem with a
completely closed forming die, the two units be-
ing separated spatially. Plastic material is intro-
duced in portions into the die through a noz-
zle. Once the substance has hardened, the die
is opened and the molded object ejected. The
injection molding process is also applicable to
thermosetting polymers.
Compression molding requires continuous
heating of the forming die. Appropriate amounts
of a preheated thermosetting material are first in-
troduced into an open die. Once the die is closed,
heat and pressure cause the polymer to flow and
expand, entirely filling the available space. Af-
ter curing is complete, the die is opened and the
finished product is ejected while still hot.
Transfer Molding. In transfer molding the
polymer is warmed to a plastic state in a sep-
arate compartment prior to being forced into a
forming die.
Reshaping of a semifinished thermoplastic
polymer into a thermoelastic state requires rel-
atively little applied force. Cut sheets or blocks
of material are brought to working temperature
with the aid of infrared radiation and then drawn
into monolithic forming dies where they acquire
the desired shape. Finished parts are ejected after
cooling.
Heat sealing complements the technique of
molding. Small numbers of complex objects can
be economically assembled from components
by heat-sealing either with the aid of warm air
or by techniques based on high-frequency radi-
ation, heat impulses, or ultrasonics.
Soaking is especially suitable for coating
electronic products with casting resins. How-
ever, it may also be utilized with insulating li-
quids and certain thermoplastic materials, where
the latter are either preheated or applied as low-
viscosity solutions. The purpose of soaking is to
reduce the risk of short circuiting, for example
in the windings of a current-carrying metallic
conductor, by ensuring that all gaps are filled.
Improved heat transfer is achieved simultane-
ously (the specific thermal conductivity of air
is 0.026 W m−1 K−1 , while that of synthetic
resin filler is 0.6 W m−1 K−1 ). The most com-
mon soaking resins are solvent-free unsaturated
polyester resins, which may be cured by activa-
tion with a setting agent. Resins generally have
a higher filling power than impregnating var-
nishes.
Impregnation provides almost total protec-
tion against undesired environmental influences.
It also confers increased mechanical stability
and protection against the adverse effects of mo-
tion. Impregnation itself is normally carried out
under vacuum. Most impregnating agents are
liquid in the working state, but they can be trans-
formed into a thermosetting state by curing. The
method is also applicable to solvent-free trick-
ling resins prepared from unsaturated polyesters.
These materials break into droplets in the course
of winding and revert to a gel-like state with

subsequent electrical conduction warming of the
conductor. Rapid heating to the curing tempera-
ture induces such an impregnating agent finally
to harden.
Varnishing. Metallic electrical conductors
can also be electrically insulated with wire var-
nish. This method is often employed in trans-
formers, motors, generators, relay coils, and
similar devices. The applied polymer (usually a
polyesterimide, polyurethane, polyamideimide,
or polyhydantoin) is dissolved in an appropri-
ate solvent, e.g., aromatic and aliphatic hydro-
carbons, chlorinated hydrocarbons, aqueous al-
cohols. The conductor is passed several times
through a varnish bath, and becomes uniformly
coated. Solvent is evaporated in a radiation oven,
permitting deposition and annealing of a multi-
layered coating graded according to hardness.
Thermoplastic polycarbonate films can be ob-
tained from solution in a fashion similar to var-
nish films, whereby the film is extracted from a
drum after evaporation of solvent.
A more environmentally acceptable coating
technique utilizes either a melt varnish of the
polyesterimide type or a water-soluble wire var-
nish.
Embedding. Solvent-free, soaking, trick-
ling, and casting resins are used for embed-
ding components in electrotechnics. They pro-
tect costly components against environmen-
tal effects, provide mechanical stabilization,
and prevent the formation of dangerous con-
tacts. Reaction resins, monomers, or prepoly-
mers are generally used that harden in the
presence of suitable reactants. Most important
are methyl methacrylate, unsaturated polyesters,
epoxy resins, polyurethanes, modified pheno-
lic resins, and furan resins. The cross-linkable
hydrocarbons and methyl methacrylate casting
resins have favorable dielectric values at high
frequencies (Sections 2.6) and are therefore im-
portant for microwave technology (television,
radar, dielectric warming).
3.3. Liquid Insulating Materials
In transformers, starter motors, and other elec-
trical equipment that generate a considerable
amount of heat, liquids are employed both as
Insulation, Electric 43
insulators and coolants. In cables, capacitors,
and connections, insulation liquids are used as
a functional component of the dielectric. In re-
lays, they function as insulators, coolants, and
extinguishers for arcing. Most liquid insulators
consist of mineral oils or synthetic liquids.
Mineral oils are complex hydrocarbon mix-
tures that are obtained from petroleum oil by
fractional distillation and refining. They com-
bine important properties including high break-
down strength, good impregnation characteris-
tics, and the ability to conduct large amounts
of heat via convection. Mineral oils can also
bind any gas produced during arcing with pro-
nounced evaporation and decomposition. De-
pending upon the predominant hydrocarbon, the
oils can be classified as naphthene- or paraffin-
based oils; however, the transition between these
groups is continuous–the mixed petroleum oils.
The naphthene-based oils have a higher density,
and a lower solidification point than the paraffin
oils; their viscosity is also more temperature de-
pendent. Since they are also more stable at low
temperatures, they are preferred for devices used
in the open air. They are also preferable for cable
applications because paraffin oils are more prone
to decomposition at high field strengths with for-
mation of hydrogen and wax products. Decom-
position can lead to partial discharges and gas in-
clusions, which may interact resulting in break-
down of the cable insulation (see section 2.7).
Naphthene oils are also suitable for cable im-
pregnation systems comprising oils and refined
resins (e.g., collophonium) because the refined
resins are incompatible with paraffin oils [47].
Insulation oils for transformers, converters,
and relays are subject to standard regulations,
e.g., IEC Publication 296, which only specifies
minimum values for the breakdown voltage [48].
Since the electric strength of oils depends
strongly on the measuring technique used (Sec-
tion 2.7), breakdown behavior is character-
ized by the breakdown voltage determined un-
der standardized conditions (specified electrode
configuration, electrode separation, and test pro-
cedure, IEC Publication 156 [32]).
When insulation liquids simultaneously act
as a coolant, they must have a low viscosity. They
should be able to withstand cooling to − 30 ◦ C
without thickening. Furthermore, the oxidation
and aging behavior of new oil and used oil should
44 Insulation, Electric
be monitored during operation when the temper-
ature is elevated and atmospheric oxygen has
access to the equipment.
The viscosity of cable insulation oils varies
within wide limits, ranging from low viscosity
oils for oil cables to highly viscous impregnation
systems for earthing and pressure cables. Fur-
thermore, extremely viscous cable immersion
systems can be prepared by addition of thick-
eners (natural resins, polymers).
The most important property of cable immer-
sion systems is the dielectric loss index (see
Section 2.6) which should be as low as possi-
ble because of the danger of thermal breakdown
of the cable insulation at higher field strengths.
The electrical strength of the impregnation sys-
tem is less important because the high electri-
cal strength of the cable insulation is a result of
the combination of the multilayer insulation film
(paper or polymeric) and impregnating agent; it
depends upon the breakdown behavior of both
components.
The viscosity and electrical properties of
mineral oils used for high-tension capacitors
(phase displacers) are similar to those of the ca-
ble insulation oils.
Synthetic Liquids. The drawbacks of min-
eral insulation oils (oxidizability, flammability,
and aging with formation of detrimental decom-
position products and deterioration of electrical
properties) led to the development of synthetic
insulation liquids with improved properties. A
general advantage of synthetic insulation liquids
is the fact that they can be reproducibly manufac-
tured with identical structures and compositions
[49]. They include chlorinated hydrocarbons,
the most common being aromatic compounds
such as chlorinated biphenols, chlorinated ben-
zenes, and their mixtures. These noncombustible
materials are known as Ascarels. Damage to
health and the environment can result from care-
less handling and use of such toxic agents. Con-
sequently, sale of Ascarel-containing products
is prohibited in many countries. Closed systems
(transformers, capacitors, etc.) have been ex-
empted from these regulations.
During decomposition in electric arcs, As-
carels do not emit explosive or combustible gas
mixtures provided that they contain equivalent
molar amounts of chlorine and hydrogen; hydro-
gen chloride is then produced. The Ascarels are
both chemically and thermally stable. They pos-
sess dipole character; their dielectric dissipation
factor is therefore frequency- and temperature-
dependent and they have a high relative permit-
tivity (εr ≈ 4 – 6 at 50 Hz).
Liquid fluorine compounds show excellent
properties. They are noncombustible, prevent
arcing, and are more thermally stable than As-
carels. They have a lower boiling point and are
highly volatile. Because of their low viscos-
ity, heat transfer properties are favorable. They
have a high breakdown voltage, high volume
resistivity (Section 2.2), and an essentially fre-
quencyindependent dielectric dissipation factor.
Liquid silicones are polymeric organosilicon
compounds with a highly variable degree of
polymerization. Commercial silicone oils have
differing viscosities, which are only slightly
temperature-dependent. They combine excel-
lent resistance to low temperatures (setting point
−5 ◦ C to −80 ◦ C) with outstanding high temper-
ature stability (240 ◦ C) and are not prone to oxi-
dation. They remain noncombustible up to oper-
ating temperatures of 300 ◦ C. Silicone oils have
a high breakdown strength, low dielectric dis-
sipation factor (≤1×10−4 ), low relative permit-
tivity (2.5 – 2.9), and an average volume resistiv-
ity (1012 Ωm). The low temperature dependence
of the dielectric dissipation factor (3×10−4 ) is
remarkable.
Other Liquid Insulators. Apart from the As-
carels and liquid silicones, further synthetic in-
sulation fluids find application mostly in trans-
formers, capacitors, and cables. The most im-
portant commercially available examples in-
clude compounds based on alkyl benzenes,
polybutene, diarylalkanes, esters, alkyl diphenyl
ethers, and high molecular mass paraffin hy-
drocarbons. These liquids match the excellent
electrical properties of the Ascarels but not their
flame retardation. They differ according to appli-
cation regarding relative permittivity, dielectric
dissipation factor, ignition temperature, viscos-
ity, and the behavior toward low temperatures.
4. Selection of an Insulating Material
Within its technical committee 15 for “Insu-
lating Materials”, the International Electrotech-
nical Commission (IEC) has installed two
subcommittees: subcommittee 15 A for “Short

Term Tests” and subcommittee 15 B for “En-
durance Tests”. Electrical testing methods for
insulation materials are developed in the work-
ing groups of these subcommittees. The rec-
ommendations of national electrotechnical com-
missions serve as the basis for their develop-
ment.
The intention and objective of the IEC is that
the international standards drawn up by numer-
ous experts from the member states should be
adopted as the national standards. All published
results should then be comparable.
The values of electrical properties given in
this publication have been extracted from nu-
merous sources; some have been supplemented
and amended using the results of our own in-
vestigations. In the available literature, detailed
information pertaining to testing conditions has
often been omitted, consequently the data plots
provide nothing more than rough guides. This
can only be improved by testing identical sam-
ples of the huge range of insulating materials
used in the electrotechnics industry in one and
the same laboratory, under identical testing con-
ditions that strictly comply with the international
regulations. Although this undertaking would be
highly desirable, it is almost impossible.
To solve a specific insulation problem, mate-
rials can be preselected from the relief diagrams.
The manufacturers must then be consulted, or
reference to technical data sheets can be made.
Measurements or the results of an independent
testing institute ensure that the material is able
to meet the requirements.
What makes an outstanding material for elec-
trical applications? The answer to this question
depends upon the intended use, e.g., whether
it involves low-voltage or high-tension sup-
plies, d.c., or low-, mid-, or high-frequency
sources. For applications with low-voltage sup-
plies, the tracking parameter should be at least
CTI 400, i.e., group II according to IEC Publi-
cation 664 [24], group I includes materials with
CTI 600. Similarly, the volume resistivity should
be greater than 1012 Ω · m, the surface resistivity
at 23 ◦ C and 50 % R.H. should be at least 1014 Ω.
At mains frequency, no particular demands
are placed on the dielectric dissipation factor or
the relative permittivity except for the applica-
tion as a dielectric in capacitors. If, for exam-
ple, the power loss per cubic centimeter should
not exceed 0.1 W at 50 Hz and the field strength
Insulation, Electric 45
should be no greater than 1000 V/cm (see Sec-
tion 2.6), then a loss index of εr ≈ 3600 can
be tolerated. An increase in the field strength to
10 000 V/cm causes a decrease in the threshold
value for the loss index by a factor of 100. In
the same way the loss index threshold value de-
creases linearly with an increase in frequency.
The magnitude of the relative permittivity
is particularly important in determining the in-
tended application of a material. To avoid di-
electric warming, its value should be low. A
lower relative permittivity is also associated with
a higher breakdown voltage in air. (As a general
rule, the maximum attainable breakdown volt-
age is achieved when the relative permittivity of
the material is identical with that of the surround-
ing medium). For applications in capacitors, the
relative permittivity should be as large as possi-
ble to minimize the geometric dimensions.
The electrical strength is primarily intended
to allow comparison between materials. How-
ever, this is only the case when all parameters,
including sample thickness, are absolutely iden-
tical. The significance of the electrical strength
is thus disputable (Section 2.7). It cannot be used
directly for design purposes. Test models are im-
perative, studies on the effect of loading time are
often required.
A “single value”, i.e., result at a discrete tem-
perature and frequency does not characterize the
behavior of a material under application con-
ditions. The influences of parameters such as
temperature, frequency, loading time, age, or
humidity cannot be discerned from this single
piece of information. A knowledge of the in-
terdependence of the various parameters pro-
vides far more information than a single data
value. In many cases, the development of tech-
nical polymers and their successful application
in the electrical industry has only been possi-
ble because the complete characteristics of the
material had been previously compiled and eval-
uated. The application range must be considered
as a function of temperature, frequency, voltage,
and environmental conditions. Other influences
including manufacturing processes, mechanical
loading, and the effects of fillers and additives
are better reflected by the functions of the electri-
cal parameters rather than by individual results
at a single temperature or frequency. Automatic
measurement processes have made the practical
46 Insulation, Electric

determination of such relationships relatively in- assumes a value equal to the square of the re-
expensive. fractive index at the optical relative permittivity.
The following diagrams represent exam- The dielectric dissipation factor reaches a broad
ples for characterizing a technical insulation maximum at 10 MHz.
material–a thermoplast with typical properties:
high flame resistance, rigid-elastic, high impact
toughness above 90 ◦ C, mechanical durability
at 90 ◦ C. On the basis of the short-term electri-
cal properties, the product quality lies within the
following limits:
Tracking CTI 250 – CTI 500
Volume resistivity (0.5 – 5)×1016 Ω · cm
Surface resistivity (0.5 – 5)×1015 Ω
Loss index (0.3 – 3)×10−2
Relative permittivity 2.9 – 3.2
Electrical strength 25 – 35 kV/mm

Figure 59 illustrates the dielectric dissipation
factor (Section 2.5) and the relative permittiv-
ity (Section 2.4) as a function of temperature at
frequencies of 50 Hz, 1 kHz, and 1 MHz. At tem-
peratures below the long-term thermal stability
threshold (90 ◦ C) the curves remain relatively
constant and are typical of a medium-quality
material. In the vicinity of − 50 ◦ C, a relaxation
phenomenon occurs, which is indicated by the
decrease in the dielectric dissipation factor at
low frequencies. Above 90 ◦ C, the dielectric dis-
sipation factor increases sharply; the effects of
conductivity losses are apparent.
Figure 60. Relative permittivity (A) and dielectric dissipa-
tion factor (B) of a thermoplastic polymer blend as functions
of frequency at 23 ◦ C
Figure 61 shows the plot of volume resistivity
(Section 2.2) as a function of temperature. The
decrease in the volume resistivity by a factor
of five powers of ten is considerable and illus-
trates that the number of mobile charge carriers
strongly increases with increasing temperature.

Figure 61. Volume resistivity of a thermoplastic polymer

blend as a function of temperature

The surface of the sample is hygroscopic.

Figure 59. Relative permittivity (A) and dielectric dissipa-
tion factor (B) of a thermoplast as functions of temperature
a) 50 Hz; b) 1 kHz; c) 1 MHz
Figure 60 illustrates both dielectric param-
eters as functions of frequency. The decrease
in the relative permittivity with increasing fre-
quency is low. The relative permittivity must
decrease with increasing frequency because it
This is shown in Figure 62 by the decrease in
surface resistivity with increasing air humidity
(Section 2.3).
Because of the large deviation in measured
values, the variation in the measurement of sur-
face resistivity as a function of temperature (see
Section 2.3) has been included in Figure 63. A
pronounced decrease in resistance surprisingly

Figure 62. Surface resistivity of a thermoplastic polymer
blend as a function of atmospheric humidity
appears at approximately 90 ◦ C but the reason
for this is not known.
The electric strength is shown as a function
of temperature in Figure 64. A pronounced de-
crease in the breakdown voltage occurs between
90 and 100 ◦ C. This represents the limitations in
thermal stability as determined using nonelectri-
cal procedures (see also Section 2.7).
Figure 64. Electrical strength of a thermoplastic polymer
blend as a function of temperature
On the basis of the above properties, the insu-
lating material would be suitable for applications
in the low- and middle-voltage range in a normal
climate (e.g., 20 ◦ C, 55 % R.H.), and at a work-
ing temperature up to 70 ◦ C. Small distances be-
tween conductors subjected to potential differ-
ences are possible. The material is also qualified
for integrated circuits and for high frequencies. It
is highly suited for applications in audio, video,
and telecommunication equipment.
Insulation, Electric 47
Figure 63. Surface resistivity of a thermoplastic polymer
blend as a function of temperature
5. References
General References
1. J. B. Birksetal.(eds.): Progress in Dielectrics,
vols. 1 – 7, Heywood, London 1959 – 1967.
2. S. Flügge (ed.): Handbuch der Physik,
Springer Verlag, Berlin Göttingen Heidelberg,
vol. 17, Dielektrika, 1956; vols. 19, 20,
Elektrische Leitungsphänomene, 1958, 1959.
3. S. Flügge,F. Trendelenburg:Ergebnisse der
exakten Naturwissenschaft, vol. 25,1951;
vol. 27, 1953;vol. 31, 1959, Springer Verlag,
Berlin .
4. H. Fröhlich: Theory of Dielectrics, Oxford
University Press, Oxford 1987.
5. P. M. Pflier, H. Jahn, G. Jentsch: Elektrische
Meßgeräte und Meßverfahren, Springer
Verlag, Berlin 1978.
6. K. Küpfmüller: Einführung in die theoretische
Elektrotechnik, Springer Verlag Berlin,
Göttingen, Heidelberg 1984.
7. A. von Hippel: Dielectric Materials and
Applications, M.I.T. Press, Cambridge, USA,
1954.
8. A. von Hippel: Molecular Science, Molecular
Engineering, Chapman and Hall, London
1959.
9. C. P. Smith: Dielectric Behaviour and
Structure, McGraw-Hill, New York 1955.
10. K. Potthoff, W. Widmann: Meßtechnik der
hohen Wechselspannungen, Vieweg & Sohn,
Braunschweig – Wiesbaden 1965.
11. H. Bilke: Handbuch der Keramik, Z. Vlg.
Schmid, Freiburg i. Brg.
48 Insulation, Electric
12. Engineering Property Data on Selected
Ceramics: vol. I, Nitrides (1976); vol. II,
Carbides (1979); vol. III, Single oxides
(1981); Metals and Ceramics Information
Center, MCIC Report HB-07, Battelle
Columbus Laboratories, Columbus, Ohio.
13. W. Espe: Werkstoffkunde der
Hochvakuumtechnik, vol. II:
Silikatwerkstoffe, VEB Dtsch. Vlg. d.
Wissensch., Berlin 1960.
14. P. Guillery: Werkstoffkunde für
Elektroingenieure, Vieweg, Braunschweig
1971.
15. T. Haase: Keramik, VEB Dtsch. Vlg. f.
Grundstoffind., Leipzig 1970.
16. A. Hecht: Elektrokeramik. Springer, Berlin –
Heidelberg – New York 1976.
17. R. Racho et al.: Werkstoffe der Elektrotechnik,
VEB Vlg. Technik, Berlin 1968.
18. H.-W. Rotter: Glimmer und
Glimmererzeugnisse, Eigenschaften,
Entwicklungen, Anwendungen, Siemens AG,
Berlin – München 1985.
19. E. Ryschkèwitch: Oxide Ceramics, Academic
Press, New York – London 1960.
20. H. Stäger: Werkstoffkunde der
elektrotechnischen Isolierstoffe, Bornträger,
Berlin 1955.
21. H. Salmang, H. Scholze: Die physikalischen
und chemischen Grundlagen der Keramik,
Springer, Berlin – Heidelberg – New York
1968.
Specific References
22. A. Kamm, Kunststoffe 75 (1985) no. 12,
900 – 903.
23. IEC Publication 112 – 1979, Bureau Central de
la Commission Electrotechnique International,
Genève.
24. IEC Publication 664/664 A – 1981, Bureau
Central de la Commission Electrotechnique
International, Genève.
25. ANSI/ASTM 3638 – 1977, ASTM
Philadelphia, Pa., USA.
26. ANSI/UL 764 A – 1984, UL Publications
Stock, Northbrook, Ill., USA.
27. IEC Publication 93 – 1980, Bureau Central de
la Commission Electrotechnique International,
Genève.
28. IEC Publication 247 – 1978, Bureau Central de
la Commission Electrotechnique International,
Genève.
29. IEC Publication 345 – 1971, Bureau Central de
la Commission Electrotechnique International,
Genève.
30. IEC Publication 250 – 1969, Bureau Central de
la Commission Electrotechnique International,
Genève.
31. IEC Publication 243 – 1967, Bureau Central de
la Commission Electrotechnique International,
Genève.
32. IEC Publication 156 – 1963, Bureau Central de
la Commission Electrotechnique International,
Genève.
33. IEC Publication 587 – 1984, Bureau Central de
la Commission Electrotechnique International,
Genève.
34. ANSI/ASTM-D 495, ASTM Philadelphia, Pa.,
USA.
35. DIN VDE 0303, 1989.
36. VDE 0303, part 8, Beuth Verlag, Berlin 1975.
37. DIN VDE 0303, 1989.
38. IEC Publication 426 – 1973. Bureau Central de
la Commission Electrotechnique International,
Genève.
39. IEC Publication 167 – 1964, Bureau Central de
la Commission Electrotechnique International,
Genève.
40. IEC Publication 343 – 1970, Bureau Central de
la Commission Electrotechnique International,
Genève.
41. IEC Publication 672-2 – 1980, Part 1
Definition and Classification, Part 2 Methods
of Test, Part 3 Individual Materials, Bureau
Central de la Commission Electrotechnique
International, Genève.
42. K. H. Schüller: “Hochfeste Porzellane auf
Quarz- und Cristobalitbasis,” Ber. Dtsch.
Keram. Ges. 44 (1967) 212 – 223, 284 – 293,
387 – 391.
43. V. Frith, R. O. Heckroodt, K. H. Schüller:
“Characterization of Ceramic High-Voltage
Insulator Materials,” Ber. Dtsch. Keram. Ges.
64 (1987) 216 – 219.
44. H. Kromer, D. Rose, K. H. Schüller:
“Investigations on Commercial Talcs and
Steatites, IV formation of Mineral Phase
During Firing,” Sprechsaal 121 (1988) no. 8,
1 – 8.
45. H. Schmidt, K. H. Schüller: “Untersuchungen
an Sondersteatiten mit Kalk,” Silikattechnik
39 (1988) 137 – 138.
46. W. Büssem, C. Schusterius, K. Stuckard: “über
die Konstitution des Steatis,” Wiss. Veroeff.
Siemens-Werken 17 (1938) 59 – 89.
47. A. C. M. Wilson: Insulated Liquids, The
Institute of Electric Engineers, London 1980.
48. IEC Publication 296 – 1969, Bureau Central de
la Commission Electrotechnique International,
Genève.
 Insulation, Electric 49

49. VDEW: ölbuch, Versorgungs- und 50. H. Saechtling: Kunststoff-Taschenbuch, 23rd

Wirtschaftsgesellschaft der Elektrizitätswerke ed., C. Hanser Verlag, München 1986.
mbH, Frankfurt 1983. 51. R. Vieweg et al., Kunststoff Handbuch,
vols. I – XI, Hanser, München 1963.

Insulin → Peptide and Protein Hormones

Integrated Circuits → Imaging Technology
Integrated Circuits → Semiconductors
 Interferons 1

Interferons
Tattanahalli L. Nagabhushan, Schering – Plough Research, Bloomfield, New Jersey 07003, United States
Paul P. Trotta, Schering – Plough Research, Bloomfield, New Jersey 07003, United States

1. Introduction . . . . . . . . . . . . . . . . . 1 5.3. Solution Properties . . ... . . . . . . . 8

2. Classification and Origin . . . . . . . . 2 6. Purification . . . . . . . ... . . . . . . . 8
3. Biological Activity . . . . . . . . . . . . . 3 6.1. Natural Leukocyte IFN .. . . . . . . . 8
3.1. Antiviral Activity . . . . . . . . . . . . . 3 6.2. Recombinant IFNα . . ... . . . . . . . 9
3.2. Anticellular Activity . . . . . . . . . . . 4 6.3. Removal of Pyrogens . ... . . . . . . . 10
3.3. Immunomodulatory Activity . . . . . . 4 7. Quality Specifications . ... . . . . . . . 12
4. Mechanism of Action . . . . . . . . . . . 5 7.1. Purity . . . . . . . . . . . ... . . . . . . . 12
5. Physicochemical Properties . . . . . . . 6 7.2. Pyrogenicity Testing . ... . . . . . . . 14
5.1. Primary Structure . . . . . . . . . . . . . 6 8. Clinical Studies . . . . . ... . . . . . . . 14
5.2. Secondary and Tertiary Structure . . 7 9. References . . . . . . . . ... . . . . . . . 15

Abbreviations used in this article: antiviral agents because they act directly on the
AIDS acquired immune deficiency syndrome target cell to confer a state of resistance to vi-
cAMP cyclic adenosine 5
-monophosphate ral infectivity at one or more phases of the viral
cGMP cyclic guanosine 5
-monophosphate replication cycle, but have no direct toxicity to-
CD circular dichroism ward the virus itself. The biological effects of
CPE cytopathic effect IFN are not limited to antiviral activity, how-
DEAE diethylaminoethyl ever: highly purified preparations of genetically
DNA deoxyribonucleic acid engineered IFN have potent antiproliferative and
HPLC high-performance liquid chromatogra- immunomodulatory effects both in vitro and in
phy animal models [3], [4]. Many types of animal
IFN interferon cells can produce IFN in response to a variety
IU international antiviral units of external stimuli or “inducers,” including cer-
MHC major histocompatibility complex tain types of double-stranded ribonucleic acid
RIEF recycling isoelectric focusing (RNA), antigens, and mitogens [5].
RNA ribonucleic acid Purification of IFN from natural sources for
SDS – PAGE sodium dodecyl sulfate – poly- structural and biological characterization, as
acrylamide gel electrophoresis well as for initiation of clinical trials, proved
Tris tris(hydroxymethyl)aminomethane difficult because IFNs exhibit a high specific
biological activity and are normally produced
by cells in low absolute quantities. Thus, only
relatively impure preparations containing up to
1. Introduction 1 % IFN could be partially purified from nor-
mal human leukocytes for use in clinical trials
In 1957, A. Isaacs and J. Lindenmann ob- in the mid 1970s. Not until the late 1970s could
served that mammalian cells incubated with a sufficient quantities of naturally occurring IFNs
heat-inactivated influenza virus produced a sub- be purified from lymphoblastoid cells for initi-
stance that could confer on fresh cells a re- ation of structural studies [6], [7]. However, the
sistance to infection by live virus [1]. This achievement of Weissmann and coworkers in
substance, called interferon, was subsequently cloning the gene for a single subtype of human
shown to consist of a system of distinct proteins, leukocyte IFN, IFNα-2, in Escherichia coli al-
many of which are structurally related, that are lowed the preparation of gram quantities through
products of multigene families as well as sep- genetic engineering [8]. The availability of vir-
arate genes [2]. Interferons (IFNs) are unique

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 365
2 Interferons

tually unlimited quantities of IFN has permitted
detailed biological and structural characteriza-
tion, as well as extensive clinical trials world-
wide [9], [10].
Genetically engineered IFNα-2 b ( Intron A)
has now been approved for marketing in over
30 countries, including the United States, for
both cancer and viral indications such as hairy
cell leukemia, acquired immune deficiency syn-
drome (AIDS)-related Kaposi’s sarcoma, renal
cell carcinoma, malignant melanoma, multiple
myeloma, laryngeal papillomatosis, and condy-
loma acuminata (genital warts). Extensive clin-
ical trials have also been performed on IFNγ
[11–14] and IFNβ [15], [16]. This article re-
views the properties, production, and specifica-
tions of recombinant human IFNα, for which the
most data are available. However, for compara-
tive purposes selected data are also provided for
human IFNβ and IFNγ, as well as natural leuko-
cyte IFN. For further information, see [2–4], [9],
[17–44].
In addition to viruses, a number of natu-
ral and synthetic agents are capable of induc-
ing IFN production in a wide variety of cell
types [5]. A general summary of cell types and
IFN inducers for the three IFN classes is pre-
sented in Table 1. Interferon-inducing agents
include bacteria, protozoa, mycoplasma, rick-
ettsia, and bacterial products (e.g., bacterial cell
wall extracts, lipopolysaccharides, and endotox-
ins). The effect of synthetic and natural nucleic
acids and polyanions, including double-stranded
RNA, synthetic polynucleotides, and synthetic
polycarboxylates, has been of special interest.
Low molecular mass compounds such as anthra-
quinones, propanediamines, acridines, pyrimi-
dines and fluorenones can also induce IFN titers
in the serum of certain animals. Specific struc-
tural requirements for these low molecular mass
compounds to induce IFN have not been identi-
fied yet. Metabolic inhibitors (e.g., actinomycin
D), mitogens, antigens, and tumor cells have also
been identified as IFN inducers.
Table 1. Classes, sources, and inducers of IFN

2. Classification and Origin Class Cell source Inducers

Alpha macrophages viruses

The first classification of IFN subtypes was (leukocyte, B cells foreign cells and
based on the observation that exposure to pH 2 type 1) their products
inactivated the antiviral activity of some IFN null cells ∗ low molecular mass
inducers (e.g. tilorone)
preparations. The IFN stable at pH 2 was des- Beta epithelial cells viruses
ignated class I, whereas IFN labile at pH 2 was (fibroblast, fibroblasts polynucleotides
designated class II. Thus, heterogeneity exists type I) null cells ∗ foreign cells and
within the IFN system. Subsequently the type their products
lymphoblasts metabolic inhibitors
I IFNs produced by leukocytes and fibroblasts B-cell mitogens
were shown to differ in their antigenicity and Gamma T cells T-cell mitogens
in selected chemical properties. As a result of (immune, null cells ∗ foreign antigens
type II) metabolic inhibitors
this observation, type I IFNs were renamed ac-
cording to the cells of origin, i.e., leukocyte and ∗ Null cells are non-B, non-T lymphocytes.
fibroblast IFN, respectively. These definitions,
however, were not adequate because IFN identi-
fied as fibroblast with specific antisera was also
shown to be produced by leukocytes; leukocyte 3. Biological Activity
IFN could likewise be produced by fibroblasts.
Therefore, international agreement was reached As previously noted, IFN was first discovered as
that IFNs should be categorized as alpha and an antiviral agent that could protect cells from
beta, depending principally on their antigenic viral infectivity. However, a variety of other bi-
properties. A third antigenic category, referred ological effects can be elicited by both natural
to as gamma (or immune) IFN, was also recog- and recombinant IFNs. Utilizing relatively crude
nized; this IFN class is produced principally by preparations, Paucker et al. [45] described the
T lymphocytes stimulated by specific antigens first nonantiviral activity of IFN, i.e., the ability
or mitogens and is categorized as type II (i.e., to inhibit cell growth. Subsequently, IFN was
acid-labile). also shown to exert potent immunomodulatory

effects as measured by its ability to enhance the
activity of immune effector cells (e.g., natural
killer cells) [46]. The range of biological ac-
tivity demonstrated for IFN has increased con-
siderably [47]. One of the important features of
IFNs is their species specificity; all classes of hu-
man IFN are most effective on human cells and
are generally much less potent on animal cells.
However, because many of the early studies on
nonantiviral activity were performed with im-
pure preparations of naturally occurring IFNs,
these biological effects were not known to be
inherent properties of IFN. Subsequent use of
highly purified recombinant IFN has demon-
strated that most observed biological effects can
be attributed to IFN [2], [3].
These properties of IFN provide strong sup-
port for the clinical application of this class of
biological response modifiers as both antiviral
and antitumor agents. Aspects of the three cate-
gories of biological activity of IFN are reviewed
briefly below.
3.1. Antiviral Activity
Early studies indicated that the replication of a
broad spectrum of both RNA- and DNA-con-
taining viruses, either oncogenic or nononco-
genic, could be inhibited by IFN [37]. Exam-
ples of such DNA viruses are herpes virus types
1 and 2 and cytomegalovirus. Examples of RNA
viruses are rhinovirus and respiratory syncytial
virus. In addition, IFNα can act synergistically
with other antiviral agents [48]. However, the
degree of antiviral activity depends not only on
the nature of the virus, but also on the character-
istics of the target cell, the type of IFN, and the
ratio of infecting virus to cell number. In vivo
studies in nonhuman primate models have also
demonstrated the antiviral activity of IFNα and
IFNβ against herpesvirus [49].
The specific antiviral activities of highly pu-
rified recombinant IFNα subtypes utilizing hu-
man foreskin fibroblast cells and encephalomy-
ocarditis virus are summarized in Table 2 [50].
Four of these (α-1, -2, -4, and -7) are natu-
rally occurring gene products, whereas the re-
maining two subtypes (δ-4 α-1 and the δ-4
α-2/α-1 hybrid) are products of recombinant
DNA technology in which the first four ami-
no acids of the amino terminus are deleted.
Interferons 3
Both of these molecules lack a cysteine residue
at position 1 and, hence, are predicted to con-
tain only one disulfide bond (see Chap. 5). The
hybrid molecule consists of residues 5 – 62 of
IFNα-2 and residues 64 – 166 of IFNα-1. De-
spite the high degree of purity of these prepara-
tions and their amino acid sequence homology
(at least 70 %), the specific antiviral activities
span approximately two orders of magnitude.
Interferon α-2 has the highest antiviral activity
(1.7×108 IU/mg).
Table 2. Specific antiviral activity of IFNα subtypes ∗
IFN subtype Activity, IU/mg
α-1 7.1×106
α-2 1.7×108
α-4 1.0×108
α-7 2.0×107
δ-4 α-1 2.8×106
δ-4 α-2/α-1 1.0×108
∗ Activity was determined by the CPE-inhibition assay
employing encephalomyocarditis virus and human foreskin cells;
data represent the mean of at least nine assays [50].
Interferon confers resistance to viral infectiv-
ity by affecting various stages of the viral repli-
cation cycle. It may inhibit early events (e.g.,
attachment, receptor-mediated endocytosis, un-
coating, and transcription), the translation of
messenger RNA, and virus maturation, includ-
ing budding of the virion at the cell membrane.
Other evidence also indicates that IFN lowers
the infectivity of the progeny virions.
3.2. Anticellular Activity
Interferon can inhibit the growth of a variety
of normal and malignant cells in vitro. For ex-
ample, it can inhibit the growth of fresh tumor
cells in the clonogenic assay [51]. This assay is
an in vitro cloning technique in which cancer
cells form colonies in a semisolid medium. The
antigrowth activity of IFN is generally cytostatic
(i.e., prevents cell division) rather than cytotoxic
(i.e., kills cells directly). Sensitivity to growth in-
hibition by IFN is variable, even within cells of
the same histologic type; this may be due to dif-
ferences in IFN receptor levels or in the affinity
of IFN for the receptor.
Combinations of IFN and chemotherapeutic
agents may have a synergistic (enhancing) or ad-
ditive interaction as antitumor agents [52], [53].
4 Interferons
Synergy may be the result of a complex interac-
tion of factors, including the nature of the tumor,
the mechanism of action of the cytotoxic agent,
and the administration schedule of IFN and the
cytotoxic agent. Thus, synergistic interactions
between IFN and cytotoxic drugs have been ob-
served in the human tumor clonogenic assay uti-
lizing recombinant IFNα and vinblastine, cis-
platin, or doxorubicin [54], [55]. However, syn-
ergy is highly schedule-dependent. When cells
were first treated with IFN
α followed by doxorubicin or vice versa, no
synergy was observed. In contrast, dramatic syn-
ergy was observed when both agents were ad-
ministered simultaneously. Other drugs tested
in this assay (e.g., bleomycin, methotrexate, or
vinca alkaloids) were not synergistic with IFNα.
The mechanism of the observed synergy as well
as of drug failure remains unknown.
Another important aspect of the anticellular
activity of IFNα is its ability to modulate the
state of cellular differentiation. Both inhibition
and acceleration of differentiation have been ob-
served [56]. Interferon also interacts synergisti-
cally with other agents that inhibit differentia-
tion, e.g., phorbol ester tumor promoters [57].
The induction of differentiation in tumor cells
mediated by IFN may account for at least a part
of its antitumor activity because the degree of
cellular differentiation appears to be inversely
related to the rate of cell division.
3.3. Immunomodulatory Activity
A variety of immune functions can be affected
by IFN [41]. Immune response may be mediated
either by cells or by antibodies. Interferon can
activate or inhibit both cellular and humoral (an-
tibody) immune function: higher concentrations
generally inhibit, whereas lower concentrations
activate. Natural killer cell activity as well as
macrophage tumoricidal and tumoristatic activ-
ities can be significantly enhanced by both IFNα
and IFNγ. Other types of immune effector cells
activated by IFN include cytotoxic T lympho-
cytes, cells involved in antibody-dependent cy-
totoxicity, and mast cells. Interferons of all three
types induce expression of class I antigens of the
major histocompatibility complex (MHC), and
IFNγ induces expression of class II MHC anti-
gens. This property is significant because class
I MHC antigens play a role in the lysis of virus-
infected cells and tumor cells by cytotoxic T
lymphocytes. Class II MHC antigens are essen-
tial for the ability of macrophages to function as
antigen-presenting cells. However, both classes
of MHC antigens are fundamentally important
for achieving maximal immune responsiveness.
All types of IFN appear to inhibit antibody pro-
duction, although under certain conditions of
dosing and time of addition, an enhancement of
antibody production has been observed.
The role of immune stimulation in clinical
responses to IFN remains unknown. For exam-
ple, measurements of natural killer cell activity
after administration of recombinant IFNα have
demonstrated both enhancement and depression
of activity, as well as no response [58], [59].
However, stimulation by IFN of cytotoxic effec-
tor cell function and antibody production may
still represent major mechanisms of its thera-
peutic activity.
4. Mechanism of Action
The first step in expression of the biological ac-
tivity of IFN is its binding to specific cell sur-
face receptors [60]. The number of receptors
for IFNα on human cells is reported to be ca.
650 – 13 000 per cell. The ligand and receptor
form a high-affinity complex with a dissocia-
tion constant of 10−10 – 10−11 mol/L. The IFN-
receptor complex has an apparent molecular
mass of 140 000 – 150 000 deduced from chem-
ical cross-linking of radioactively labeled IFN
to the cell surface. Thus, the apparent molecu-
lar mass of the receptor is 120 000 – 130 000 if
one molecule of IFN is bound per receptor. The
failure of certain cells to respond to IFN may be
due to absence of the receptor or a low affinity
of IFN for the receptor.
Ultrastructural studies indicate that bound
IFNα enters cells by a clustering of the recep-
tor – ligand complex into “coated pits.” These
pits, which are surrounded by the protein
clathrin, function in the transport of a variety
of molecules into the cell. The pits pinch off to
form vesicles (so-called receptosomes), which
can transport their contents to a variety of tar-
gets within the cell. The role of internalization of
IFNα into the cell in the mechanism of action is

yet to be elucidated. Interferon alone, the IFN –
receptor complex, or IFN degradation products
may trigger a biological response; alternatively,
binding of IFN to the cell surface may itself be
sufficient.
Binding of IFNα to its receptor induces the
synthesis of about a dozen different proteins
ranging from 15 000 to 120 000 in molecular
mass. Many of these have not been identified,
and their role in the mechanism of IFN activ-
ity is unknown. Interferon also affects the ac-
tivity of specific proteins or enzymes. For ex-
ample, levels of ornithine decarboxylase and
S-adenosylmethionine decarboxylase, enzymes
involved in polyamine biosynthesis, are reduced
by IFN treatment [61]. Because these enzymes
are critical for cellular proliferation, their inhi-
bition may account for the antigrowth activity
of IFN. Interferon treatment can increase cyclic
guanosine 5
-monophosphate (cGMP) and cy-
clic adenosine 5
-monophosphate (cAMP); both
of these compounds have also been implicated
in growth regulation [61]. However, the exact
biochemical mechanism for the antiproliferative
effects of IFN remains unknown.
As noted above, all three IFN classes can
induce the expression of class I MHC anti-
gens [62], which mediate the lysis of virus-
infected cells and tumor cells by cytotoxic T
lymphocytes. In addition, IFN can enhance the
expression of the immunoglobulin G receptor
on lymphocytes and macrophages; this, in turn,
may enhance antibody-dependent cytotoxicity
or macrophage-mediated phagocytosis [63].
An extensively characterized enzyme that is
induced by IFN treatment and appears to play
a critical role in interferon’s antiviral and an-
tiproliferative activities is 2
,5
-oligoadenylate
synthetase (E. C. 2.7.7.-) [64]. This enzyme
catalyzes the synthesis of a family of novel
2
,5
-oligoadenylate nucleotides that regulate
the breakdown of RNA in cells. Reported levels
of induction of the synthetase vary between 10-
and 10 000-fold. The enzyme requires double-
stranded RNA, which may be produced by the
virus as a replicative intermediate, for expres-
sion of enzyme activity. Although the exact
mechanism for this activation is not known, the
double-stranded RNA may bind to the enzyme
and convert it to its biologically active form.
The 2
,5
-oligoadenylates activate a latent en-
doribonuclease, known as RNase L, which cat-
Interferons 5
alyzes the cleavage of both messenger and ri-
bosomal RNA. Thus, the localized activation of
endonuclease by double-stranded RNA present
during viral replication may account for the se-
lective cleavage of viral versus host RNA.
Interferon induces another system that also
requires double-stranded RNA for activation,
namely, a protein kinase that catalyzes the phos-
phorylation of protein initiation factor eIF-2, as
well as the ribosomal protein P 1 [65]. In cell-
free systems, the kinase is activated by much
lower concentrations of double-stranded RNA
than is 2
,5
-oligoadenylate synthetase; higher
concentrations of double-stranded RNA inhibit
the kinase but activate the synthetase. Thus, the
level of double-stranded RNA may determine
which biochemical pathway is important in the
expression of antiviral activity.
5. Physicochemical Properties
Only very low (picogram) quantities of IFN
can normally be produced from natural sources
because the biological activity is expressed at
very low protein levels. Thus, characterization
of its properties was for many years slow and
incomplete. Studies on IFNα prepared from a
variety of natural sources, including virally in-
duced natural leukocytes, lymphoblastoid cells,
and chronic myelogenous leukemic cells, sug-
gested that IFNα consists of a family of multi-
ple species with apparent molecular masses de-
termined by sodium dodecyl sulfate – polyacryl-
amide gel electrophoresis (SDS – PAGE), rang-
ing from 16 000 to 27 000. Amino acid analy-
sis of the purified proteins indicated that their
compositions were similar but that IFNα con-
sisted of a family of proteins with different ami-
no acid sequences. However, further characteri-
zation of the physicochemical properties of IFNs
was severely hindered by the fact that IFNs
purified from natural sources were frequently
contaminated with other proteins and IFN sub-
types. In addition, the extremely low levels of
substance produced resulted in large losses of
protein because of adsorption and surface de-
naturation. These problems were solved by the
large-scale production of genetically engineered
IFNα-2 b ( Intron A) in E. coli [3, pp. 1 – 12],
[10].
6 Interferons
5.1. Primary Structure
The amino acid sequence of IFNα-2 deduced
from the DNA sequence of the cloned gene is de-
picted in Figure 1. This sequence has been con-
firmed by N-terminal sequencing through auto-
mated Edman degradation of the native protein
and of fragments produced by either proteolytic
or cyanogen bromide cleavage. The DNA se-
quences of 16 other IFN genes indicate an ami-
no acid sequence homology of 75 % or more
and an identity of amino acids at approximately
60 % of all positions for the majority of IFNα
subtypes. However, some IFNα subtypes differ
by only one amino acid and yet are produced
by distinct, nonallelic genes, e.g., IFNα-A and
IFNα-2. The IFNα subtypes generally consist
of 165 or 166 amino acid residues preceded by
a 23-amino acid signal sequence. These signal
sequences are not found in the mature protein
because they are removed by proteolysis during
secretion of IFN by the cell.
All IFNα subtypes contain four cysteine
residues. As shown in Figure 1, for IFNα-2 b
these residues occur at positions 1, 29, 98, and
138. Chromatographic analysis of proteolytic
fragments has shown that these amino acids form
two disulfide bonds between residues 1 and 98,
and residues 29 and 138. Pairing of the cysteine
residues is probably the same in all IFNα sub-
types. Studies with reducing agents indicate that
the 29 – 138 disulfide bond is required for the full
expression of biological activity, whereas the 1 –
98 disulfide bond may be disrupted without a
significant loss in activity.
Based on examination of the biological ac-
tivity of various recombinant hybrid IFNα
molecules, IFNα appears to contain at least two
distinct binding sites located at the amino and
the carboxyl terminals, respectively. Each bind-
ing site is thought to interact with the IFN re-
ceptor, and the nature of the biological response
is determined by the quality of fit between lig-
and and receptor [66]. However, other studies
suggest that the final 13 carboxyl-terminal ami-
no acids of IFNα are not required for biological
activity.
Two forms of IFNβ have been cloned, and the
amino acid sequence predicted from the comple-
mentary DNA sequence [67], [68]. The principal
form of IFNβ produced by fibroblasts, IFNβ-1,
contains 166 amino acids and exhibits approxi-
mately 30 % homology in amino acid sequence
with IFNα-1. However, the region of highest ho-
mology between IFNβ-1 and IFNα-1 lies bet-
ween residues 115 and 151, which suggests an
important functional role for this region.
A single form of IFNγ has been identified
[69]. The gene structure predicts that this protein
contains 146 amino acids, significantly fewer
than IFNα or IFNβ. Virtually no homology ex-
ists at the amino acid level with either IFNα or
IFNβ. Another form of IFNγ that lacks three
amino acids (cysteine – tyrosine – cysteine) at
the amino terminus has also been cloned. The
specific antiviral activity of this species appears
to be higher than that of the 146-amino acid
form.
5.2. Secondary and Tertiary Structure
The secondary structure of purified recombinant
IFNα-2 b has been examined by circular dichro-
ism (CD) in both the near and the far UV range
(Fig. 2).
Intense, negative bands at 218 and 208 nm
are characteristic of an alpha helix. Calcula-
tions based on the amplitude of these bands in-
dicate an alpha helicity of ca. 50 %. Although
the CD spectra do not indicate the presence of
beta sheets, a small percentage (e.g., 0 – 10 %)
may be present but undetectable in the spectrum.
Near-ultraviolet CD spectra indicate two intense
negative bands at 291 and 286 nm, which can
be assigned to transitions of tryptophan. The in-
tensity and location of these bands suggest that
at least one of the two tryptophan residues is
present in a hydrophobic environment. Weak,
negative CD bands at 255, 262, and 268 nm can
be assigned to phenylalanine. Titration of IFNα-
2 b solution to pH 2 results in virtually complete
loss of tryptophan CD bands as well as substan-
tial depression in the CD bands representative
of the alpha helix. However, the apparent loss in
native conformation that occurs on acidification
is completely reversible on retitration to pH 7.4.
In parallel with these findings, the antiviral ac-
tivity of IFNα-2 is also retained completely after
exposure to pH 2 and neutralization.
Precise information on the tertiary structure
of IFNα is not yet available. However, crystal-
lization of recombinant IFNα-2 b (Fig. 3) has
provided a basis for X-ray diffraction analysis

Figure 1. Amino acid sequence of human IFNα-2 b (Intron A)
Figure 2. Circular dichroism spectra of human IFNα-2 b (Intron A)
A) Far ultraviolet; B) Near ultraviolet
of three-dimensional structure. Needle-shaped
crystals, with dimensions of ca. 0.1×0.01 mm2 ,
are prepared by incubation of a purified IFNα-
2 solution at pH 6.0, which is near the isoelec-
tric point of the protein [10]. Crystal dimensions
vary depending on the conditions of crystalliza-
tion.
Figure 3. Human IFNα-2 b (Intron A) crystals
Crystallization of a form of human IFNγ
(IFNγ-D
), in which the five amino acids at the
carboxyl terminus are deleted, has also been re-
ported [70].
Interferons 7
Crystals suitable for high-resolution X-
ray diffraction have been prepared by vapor-
diffusion equilibration against a solution of
30 % saturated ammonium sulfate in 50 mmol/L
sodium acetate, pH 5.9 [70]. Rhombohedral
crystals (0.45×0.45×0.40 mm) are formed af-
ter 5-d incubation at room temperature (Fig. 4).
The X-ray precession photographs indicate that
the crystals are trigonal. The crystals are mod-
erately stable to X radiation and diffracted to a
resolution of 0.33 nm on a rotating anode gener-
ator. The three dominant faces of the rhombohe-
dral crystal correspond to the (110), (101), and
(011) planes of the rhombohedral unit cell. The
measured density of the crystals is 1.224 g/cm3
in a Ficoll – 45 % ammonium sulfate solution.
The calculated mass of the protein per unit
volume is 63 200, which indicates that four
molecules of IFNγ-D
are present in the asym-
metric unit. Diffraction data from native crys-
tals and from several heavy-atom derivatives
have been collected by oscillation photogra-
phy with synchrotron radiation [70]. The inten-
sity data sets from native crystals (resolution to
0.285 nm) are being analyzed to determine the
8 Interferons
three-dimensional structure at this degree of res-
olution.
Figure 4. Human IFNγ crystals in a suspended droplet
5.3. Solution Properties
Based on their predicted amino acid composi-
tions, the polypeptide molecular masses for hu-
man IFNα, β, γ are 19 269, 19 990, and 17 145,
respectively. However, under certain conditions
the apparent molecular mass of IFNα in solu-
tion appears to be considerably higher than the
polypeptide molecular mass [10]. Sedimenta-
tion velocity experiments have established that
IFNα-2 undergoes self-association to yield a
molecular mass at least three times that of the
monomer. Lowering the pH from 7 to 2 fa-
vors monomer formation. Under the low pro-
tein concentration in a biological assay, it is an-
ticipated that IFNα is completely monomeric;
thus, the monomer is probably the active species.
Even under highly concentrated conditions (e.g.,
10 mg/mL or higher), IFNα-2 b appears to be
very soluble. At its isoelectric point (ca. pH 6.0),
IFNα is relatively insoluble and can be crystal-
lized readily. In contrast, IFNγ appears to form
a stable dimer that does not dissociate on dilu-
tion and has a highly basic isoelectric point of
ca. 9.5.
6. Purification
6.1. Natural Leukocyte IFN
Starting materials used for the purification of hu-
man IFN include leukocytes isolated from whole
blood, neonatal fibroblasts, and lymphoblastoid
and various leukemic cell lines. The principal
source for purification of IFN prior to the use of
genetically engineered microorganisms was hu-
man cell cultures treated with an appropriate in-
ducer (e.g., a virus or a double-stranded polynu-
cleotide). The principal problem with the isola-
tion of IFN from such sources is the low level
of expression of IFN due to its potent biologi-
cal activity. Large starting volumes and multi-
ple purification steps are thus required, and only
small quantities of material can be purified. For
example, in 1978, two laboratories in Helsinki
together produced only 1.3 g of IFNα of <1 %
purity from 90 000 bags (45 000 L) of donated
blood. Lack of precision in the antiviral assay
commonly employed for monitoring the degree
of purification (see Section 7.1) has also been a
problem in the development of purification pro-
tocols.
Partially purified human leukocyte IFN was
first developed by Cantell and his associates in
Finland and has been available in limited quan-
tities for clinical trials since 1973 [71]. In this
procedure, fresh blood from normal donors is
centrifuged, and the leukocyte fraction (i.e., the
buffy coat) is removed. Pooled leukocytes are
treated with 0.83 % ammonium chloride to lyse
residual erythrocyte contaminants, suspended in
a culture medium (e.g., Eagle’s minimum es-
sential medium), and then primed with high-
titer crude IFN. Interferons production (primar-
ily by monocytes) is then induced by addition
of Sendai virus. After overnight incubation, the
cells and debris are removed by centrifugation;
the supernatant containing crude IFN is the start-
ing material for purification.
Crude IFN is first precipitated with potassium
thiocyanate at pH 3.4 and extracted with 95 %
ethanol. Extraction with acidified ethanol inac-
tivates residual Sendai virus and partially puri-
fies the IFN because not all proteins redissolve.
The pH is then increased stepwise to precipitate
contaminating proteins, which are removed by
centrifugation. The IFN is precipitated at pH 8,
concentrated, and dialyzed against phosphate-
buffered saline.
The specific antiviral activity of the IFN
preparation is approximately 106 IU/mg of pro-
tein, which represents approximately 1 % purity.
The net yield per unit of blood (i.e., 0.5 L) is
ca. (5 – 7)×106 antiviral units. The final product
contains a number of different species of IFNα,

including α-1, α-2, and α-4. Other lymphokines
(i.e., hormones released by leukocytes such as
interleukin-1 and 2) probably also contaminate
the preparation.
The partially purified leukocyte IFN prepa-
ration can be purified further to a specific activ-
ity of ca. 108 IU/mg of protein by chromatog-
raphy on an immobilized monoclonal antibody
referred to as NK-2. The IFN is eluted from the
column by exposure to low pH. However, the
relative composition of IFNα subtypes in this
preparation differs from that in crude IFN prior
to chromatography because the NK-2 antibody
does not bind all IFNα species equally.
6.2. Recombinant IFNα (→Genetic
Engineering, Chap. 5.1.)
Purification of recombinant IFN from a genet-
ically engineered microorganism poses unique
problems distinct from those encountered in pu-
rification from natural sources [72]. The follow-
ing requirements must be satisfied:
1) A procedure for efficient extraction of IFN
from the cell that does not denature the pro-
tein must be developed. High concentrations
of a conformation-disrupting agent such as
urea or guanidine should be avoided for ex-
traction because renaturation of the extracted
protein may be difficult or impossible. This
consideration is especially important for hu-
man IFNα because two disulfide bonds must
be reformed in the native structure and the use
of a chaotrope may lead to incorrect disul-
fide pairing. Furthermore, even if the final
product is biologically active and appears to
be properly refolded, small quantities of in-
correctly folded material may be highly anti-
genic and remain undetected by standard an-
alytical methods.
2) The extraction and purification procedure
must yield a protein that is uncontaminated
with bacterial pyrogens (gram-negative bac-
terial lipopolysaccharides that increase body
temperature when injected parenterally). Py-
rogens are not readily removed from purified
IFN preparations because effective methods
for their removal from solution (e.g., adsorp-
tion on activated charcoal or destruction with
ionizing radiation) may damage the biolog-
Interferons 9
ical activity or structure of the recombinant
protein.
3) The purified IFN must be substantially free
of nucleic acids and proteins derived from the
host microorganism if it is to be employed in
human clinical trials.
4) The product must be a homogeneous IFN
species that is free of aggregates and pro-
teolytic fragments. These IFN-related com-
ponents are of special concern because they
could render the product antigenic when in-
jected into humans.
5) To be commercially viable, purification must
be achieved with a good yield (e.g., 20 %) of
biological activity (e.g., antiviral units).
Purification of recombinant IFNα from
E. coli for the IFNδ-4 α-2/α-1 hybrid that was
genetically engineered to contain amino acid
residues 4 – 62 of IFNα-2 and 54 – 166 of IFNα-
1 is summarized in Table 3. The vector for
this species was a PBR-322-based plasmid un-
der control of the lac promoter [73] (see also
→Genetic Engineering, Chap. 4.1.1.). The pu-
rification of IFN is monitored by following
changes in specific antiviral activity (IU) ex-
pressed per milligram of protein.
The E. coli extract is prepared by acidifica-
tion of the bacteria to pH 2, followed by adjust-
ment to neutral pH (see Section 6.1) [74]. This
procedure exploits the acid stability of type I
(α and β) IFNs (see chap. 2) and avoids high
concentrations of denaturants. The first purifi-
cation step removes nucleic acids and contami-
nating E. coli proteins by acidification to pH 4.5
followed by centrifugation and is accompanied
by a 20-fold increase in specific antiviral ac-
tivity. The stability of IFNα at low pH and in
organic solvents permits subsequent precipita-
tion by addition of trichloroacetic acid and ex-
traction from the precipitate with ethanol. This
step increases the specific activity by an ad-
ditional 20-fold and also concentrates the IFN
solution with virtually no loss in activity. The
118 % yield indicates complete recovery of an-
tiviral activity within the experimental accuracy
of the assay. The next step consists of affin-
ity chromatography on a blue dextran affinity
medium [Matrex Gel Blue A (Cibracron Blue
F 3 GA) Sepharose]. This resin is used because
both human and murine IFNs bind to blue dex-
tran. Elution from this column is achieved with
10 Interferons
Table 3. Purification of recombinant human IFNδ-4 α-2/α-1
Purification step Volume, L
Escherichia coli extract ∗∗ 765.0
Acidification (pH 4.5) 765.0
Trichloroacetic acid – ethanol extract 21.0
Matrex Gel Blue A Sepharose 15.3
Concentration by ultrafiltration followed by 1.0
Selective precipitation 0.069
DEAE Sepharose chromatography 0.031
∗ Based on antiviral units.
∗∗ Extract was prepared by acidification of the E. coli, followed by adjustment to neutral pH for release of the IFN.
sodium chloride solution (2 mol/L). The final
purification step can be performed in two ways:
(1) repeated cycles of selective precipitation by
dialyzing to low ionic strength and redissolving
in 2 mol/L sodium chloride or (2) selective pre-
cipitation followed by ion-exchange chromatog-
raphy on diethylaminoethyl (DEAE) Sepharose.
The specific activity of the final product prepared
by either procedure is 7×107 IU/mg, which re-
presents a 140-fold purification.
An alternative approach for purifying re-
combinant IFNα employs immunosorbent chro-
matography with an immobilized monoclonal
antibody [72]. Purification of IFNα-5 by this
technique is summarized in Table 4.
The IFNα-5 is extracted by acidification
of E. coli followed by neutralization (see Sec-
tion 6.1). Extracted IFN is adsorbed on sil-
ica gel and eluted by lowering the pH to 2.0,
which results in an almost tenfold increase in
specific activity. Silica gel chromatography also
removes the majority of contaminating nucleic
acids in the unbound fraction. Matrex Gel Blue
A chromatography results in a further 3.3-fold
purification. Next, IFN is adsorbed onto an
anion-exchange resin (i.e., DEAE Sepharose)
and eluted with a pH gradient from 8.0 to 6.0,
resulting in an additional tenfold increase in spe-
cific activity. The DEAE Sepharose eluate is
purified further on an immobilized monoclonal
antibody, YOK Sepharose. Elution is achieved
with 0.1 mol/L sodium citrate, pH 2.0, to give a
final specific activity of 1.9×108 IU/mg.
Antiviral activity Purifiction ∗,
fold
107 IU/mg Total, 109 IU Yield, %
0.05 629 100 (1.0)
1.12 466 67 2.4
2.1 740 118 42
7.7 177 28 154
8.2 192 31 164
7.0 30 5 140
7.0 18 3 140
6.3. Removal of Pyrogens
The final IFN preparation should be low in py-
rogenic substances if it is to be employed in
clinical trials (see Section 6.2). Although the
purification protocols described in Sections 6.1
and 6.2 generally result in a product that is very
low in pyrogens (i.e., 5 – 10 ng of endotoxin per
milligram of protein, determined as described
in Section 7.2), preparations with high pyrogen
levels are occasionally encountered. An effec-
tive procedure for lowering the endotoxin level
of these preparations is recycling isolelectric fo-
cusing (RIEF). Isoelectric focusing is based on
the principle that charged substances migrate in
an electric field and a stable pH gradient to a pH
at which their net change is zero. In the RIEF
apparatus designed by Bier et al. [75], the pro-
tein solution is continuously recycled between a
multichannel heat exchange reservoir and a mul-
tichannel focusing cell. Thus, heat is dissipated
in the heat exchange reservoir and not in the fo-
cusing cell. As shown in Figure 5, when RIEF
is applied to a solution of IFNα-2 with a high
pyrogen content, the majority of endotoxins and
the IFNα-2 migrate to different pH values [76].
Thus, whereas endotoxins reach maximum lev-
els at pH 5.07 (channel 5), the antiviral activity
of IFNα-2 peaks at pH 6.4 (channel 10).
In the removal of contaminating pyrogens
from IFNα-2, a 40-fold purification is obtained
from a single RIEF experiment, which results in
a pyrogen level of 12.5 ng/mg of protein, com-
pared to 500 ng/mg of protein prior to RIEF. No
apparent loss in specific antiviral activity is ob-
served, and the total recovery of IFN activity
 Interferons 11
Table 4. Purification of recombinant human IFNα-5

Purification step Volume, L Antiviral activity Purification,

fold
107 IU/mg Total, 109 IU Yield, %

Escherichia coli extract 200.0 0.019 42 (100) (1.0)

Silica (Si 200) 1.8 0.18 32 76 9.5
Matrex Gel Blue A 0.8 0.6 25 60 31.6
DEAE Sepharose CL-6 B 0.2 5.8 10 24 305
YOK Sepharose 0.04 19.0 8 19 1000

Figure 5. Recycling isoelectric focusing (RIEF) of human recombinant IFNα-2 b (Intron A) containing Escherichia coli
pyrogens (endotoxins)
The apparatus was prefocused with 0.1 % ampholines (pH range 3 – 10) in urea (4 mol/L) for 60 min followed by addition of
the sample. Protein concentration was determined from the optical density reading at 280 nm; pyrogen levels were assayed by
the limulus amebocyte lysate assay, and antiviral activity was determined by CPE-inhibition assay.

is >90 %. The ampholines used to establish the
pH gradient are finally removed from the prepa-
ration by crystallization of the protein, gel fil-
tration chromatography, or ammonium sulfate
precipitation.
7. Quality Specifications
7.1. Purity
Purified IFN must be shown to be biologically
active, and its activity must remain constant from
batch to batch. Because IFN was originally de-
fined in terms of its antiviral activity, quantita-
tion of antiviral activity per milligram of puri-
fied protein is routinely employed for charac-
terization of purified product. The most com-
monly employed assay, which is readily adapted
to screening large numbers of samples, is the cy-
topathic effect- (CPE-) inhibition assay. This as-
say is based on the ability of IFN to interfere with
the cytopathic effects (e.g., cell lysis) caused by
viruses. In a simple CPE assay, IFN is serially
diluted (usually by a factor of two) in the cells
of a microtiter plate, and target cells (e.g., hu-
man foreskin diploid fibroblasts) are then added
to each well. Cells to which IFN has not been
added are also included as controls. After an
incubation period (e.g., 4 h), a challenge virus
(e.g., encephalomyocarditis virus) is added, fol-
lowed by a further period of incubation. In the
presence of IFN, cells are protected from viral
infection and can be detected by staining with a
dye such as Crystal Violet. However, in the ab-
sence of IFN, the cells lyse and are not stained
by the dye because they do not adhere to the mi-
crotiter plate. The end point is arbitrarily defined
(e.g., 50 % protection against viral infectivity).
An appropriate international reference standard
provided by the World Health Organization is in-
cluded in each microtiter plate and allows con-
version to the international antiviral units (IU)
established by international accord. Further de-
tails of the CPE-inhibition assay are described
in [77].
12 Interferons
Figure 6. Polyacrylamide gel electrophoresis of purified re-
combinant IFNα subtypes
Before loading, 0.01 mg of each subtype was boiled for 2 min
in 1 % 2-mercaptoethanol, 1 % sodium dodecyl sulfate, and
10 % glycerol dissolved in 0.06 mol/L Tris hydrochloride,
pH 6.8, containing 0.001 % bromophenol blue. Discontinu-
ous SDS – PAGE was run according to [79].
a) Standard proteins (lysozyme, β-lactoglobulin, carbonic
anhydrase, ovalbumin, and bovine serum albumin); b) IFNα-
1; c) IFNδ-4 α-1; d) IFNα-2; e) IFNδ-4 α-2/α-1; f) IFNα-4;
g) IFNα-7
Sodium dodecyl sulfate – polyacrylamide gel
electrophoresis is a powerful technique for
demonstrating the purity of the final IFN prepa-
ration [78]. The quantitation of impurities and
sensitivity of the method depend on the nature of
the stain employed for protein detection. Quan-
tities of protein as low as 0.05 – 0.1 µg can be de-
tected by staining with Coomassie Blue R 250,
whereas silver staining may result in detection of
levels of impurities up to 100-fold lower. How-
ever, silver-stained gels must be interpreted with
caution because staining intensity varies widely
among proteins (up to a factor of ten). A typi-
cal gel electrophoresis of purified recombinant
IFNα subtypes stained with Coomassie Blue
R 250 is presented in Figure 6; SDS-PAGE re-
veals a purity of ≥95 % for each subtype, which
was confirmed by silver staining. The two com-
ponents observed for the δ-4 α-2/α-1 hybrid
(Fig. 6, lane e) were both demonstrated to be
IFN-related. Approximate quantitation of im-
purities can be achieved by electrophoresis at
IFN loading levels ranging from 25 to 0.05 µg.
The loading level at which an impurity disap-
pears relative to the main component can be used
to calculate the approximate percentage of im-
purity. The SDS-PAGE technique can also be
employed for estimation of apparent molecular
mass by comparison to the migration of stan-
dards with known molecular mass. Thus, for ex-
ample, the average apparent molecular mass of
IFNα-2 was observed to be 18 500 ± 600, the
standard deviation representing gel-to-gel vari-
ation. The calculated molecular mass of IFNα-2
is 19 270, in good agreement with the experi-
mental value.
Reversed-phase high-performance liquid
chromatography (HPLC) is also employed to
characterize the purity of the final product [80].
Reversed-phase HPLC of a clinical-grade sam-
ple of IFNα-2 b prior to formulation is presented
in Figure 7. Results indicate a high degree of
homogeneity, in agreement with those obtained
by SDS-PAGE.
Figure 7. Reversed-phase HPLC of purified recombinant
human IFNα-2 b (Intron A)
Chromatography was performed on a Water Associates
µ-Bondapak C-18 column (30 cm×3.9 mm). The eluting
gradient extends from 31.4 to 62.7 % acetonitrile con-
taining 0.01 mol/L trifluoroacetic acid at a flow rate of
1 mL/min for 30 min. Detection was accomplished by mon-
itoring fluorescence after post-column derivatization with
o-phthalaldehyde. The first peak is caused by the solvent.
Amino acid analysis can also be used to char-
acterize the purified preparation. Table 5 con-
tains the average amino acid analysis of 15
batches of purified IFNα-2 b. Interferon was
first hydrolyzed with 6.7 mol/L hydrochloric
acid for 18 h at 160 ◦ C, followed by chro-
matographic resolution of the individual ami-
no acids on a C-18 HPLC reversed-phase col-
umn utilizing paired ion chromatography [81].
The amino acids were detected by fluores-
cence following postcolumn derivatization with
o-phthalaldehyde. Excellent agreement was ob-
served between experimental and theoretical
values.
Amino acid sequencing provides a more sen-
sitive method than amino acid analysis for de-
tecting contaminating proteins, as well as for
demonstrating that the recombinant protein has
the correct amino acid sequence predicted from
 Interferons 13

the nucleotide sequence of the gene. For exam- 7.2. Pyrogenicity Testing
ple, N-terminal sequencing has been achieved
for recombinant IFNα-2 by using automated Ed- Amebocytes circulating in the horseshoe crab
man degradation in a gas-phase sequenator [82]. (Limulus) hemolymph contain a coagulation
system that is highly sensitive to gram-negative
Table 5. Amino acid analysis of recombinant human IFNα-2 b
bacterial endotoxins [83]. The Limulus amebo-
Amino acid Theoretical con- Measured con- cyte lysates form a firm gel in the presence of low
tent, mol % tent, mol % levels of endotoxin. The “ limulus test” for detec-
Aspartic acid, tion of endotoxins, based on this gelation reac-
asparagine 7.74 9.06 ± 0.46 tion, is extensively employed in the pharmaceu-
Serine 9.03 10.33 ± 0.48 tical industry. As shown in Figure 8, extremely
Glycine 3.23 3.80 ± 0.91
Glutamic acid,
low endotoxin levels are detectable in purified
glutamine 16.77 19.03 ± 0.79 preparations of recombinant IFNα-2, in which
Threonine 6.45 5.54 ± 0.58 the values generally are significantly lower than
Alanine 5.16 5.61 ± 0.58 1.0 ng/mg of protein.
Valine 4.52 2.62 ± 0.23
Methionine 3.23 2.82 ± 0.59
Tyrosine 3.23 2.81 ± 0.45
Isoleucine 5.16 3.52 ± 0.39
Leucine 13.55 13.96 ± 0.46
Phenylalanine 6.45 6.51 ± 0.20
Histidine 1.94 1.95 ± 0.15
Lysine 7.10 6.53 ± 0.58
Arginine 6.45 5.99 ± 0.70

The phenylthiohydantoin – amino acids were

resolved by reversed-phase HPLC on a
cyanopropyl silane column, with acetonitrile –
methanol (4 : 1) as eluent. Sequencing was per-
formed after reduction with dithiothreitol and
alkylation with iodoacetamide to permit identi-
fication of cysteine residues. The first 57 amino
acids agreed completely with the amino acid se-
quence predicted from the complementary DNA
(Fig. 1). In addition, by utilizing peptide frag-
ments produced by cyanogen bromide treat-
ment (which cleaves the protein at methionyl Figure 8. Pyrogenicity of recombinant human IFNα-2 b (In-
residues), 90 % of the amino acid sequence tron A)
IFN dose, IU/kg:
0.75; • 1.5;  2.5
was determined. Tryptic digestion of the C-
terminal cyanogen bromide fragment followed An alternative, traditional method for in-
by reversed-phase HPLC and amino acid analy- vestigating pyrogen content is to measure the
sis of the separated components was performed increase in body temperature following intra-
to identify glutamic acid as the C-terminal ami- venous injection into rabbits. However, as shown
no acid. The amino acid sequence was also sur- in Figure 8, no apparent correlation existed bet-
veyed by gas chromatography – mass spectrom- ween temperature increase and either IFN dose
etry of fragments of IFNα-2 produced by sub- or endotoxin level measured in the limulus test.
tilisin cleavage followed by trifluoroacetylation The in vivo test is further complicated by the fact
and permethylation. Selected sequences with that IFN itself has inherent pyrogenic properties.
IFNα-2 were identified and demonstrated to cor- Thus, the limulus test appears to be a convenient
respond to the gene structure. and easily performed in vitro test for estimation
of endotoxin levels.
14 Interferons
8. Clinical Studies
Although IFN was first discovered in 1957, not
until 1981 were large enough quantities of pu-
rified recombinant IFNα available to permit the
initiation of phase I/II human clinical trials. To
date over 2000 patients with either advanced ma-
lignancies or viral infections have been treated
with IFNα-2 b ( Intron A). Despite the fact that
cancer patients treated with Intron A generally
displayed poor prognosis as well as resistance
to standard chemotherapy, a number of patients
demonstrated complete recovery of long-term
duration.
Phase I trials with Intron A were generally
performed on patients with advanced malignan-
cies and on a small number of normal volun-
teers [9], [84]. These studies indicate that Intron
A is safe in humans; its toxic effects are dose-
related and completely reversible. No evidence
for cumulative toxicity or drug intolerance was
obtained. Three routes of administration (sub-
cutaneous, intramuscular, and intravenous) were
employed in a variety of schedules. For example,
Intron A was administered in an escalating daily
dose for 28 d to determine the maximum toler-
able dosage. Dosages were also elevated on a
weekly schedule. The total daily doses examined
were generally in the range of (1 – 100)×106 IU.
The principal toxic effect observed in 90 % of
all patients consisted of flu-like symptoms (e.g.,
fatigue, fever, headache). However, these dose-
related effects reversed rapidly when the dosage
was lowered or treatment discontinued. Other
hematologic toxicities, including mild depres-
sion of white blood cell and platelet counts, and
elevation of hepatic enzymes were also observed
in some patients but were all reversible upon dis-
continuance of IFN administration. Single In-
tron A doses of up to 200×106 IU per square
meter of body surface area could be tolerated
when administered intravenously. However, for
5-d intravenous scheduling, 50×106 IU/m2 was
the maximum dose tolerated.
The pharmacokinetics of Intron A were also
evaluated in phase I trials. Intravenous infusion
for 30 min produced peak serum levels within
15 – 30 min. Because of a relatively short half-
life, serum levels returned to base line within 4 h.
However, sustained levels could be achieved by
continuous intravenous infusion. Subcutaneous
and intramuscular dosing demonstrated distinct
pharmacokinetics with slowly rising blood lev-
els that peaked at 4 h and maintained sustained
levels for 8 – 12 h. Based on in vitro studies, the
peak serum levels obtained by any of these three
routes of administration should be sufficient to
activate immune effector cells or to exert a direct
antiproliferative effect on tumor cells.
Phase II clinical trials of Intron A have been
conducted on a variety of malignancies. Results
are as follows:
Active: hairy cell leukemia
non-Hodgkin’s lymphoma
mycosis fungoides
Kaposi’s sarcoma
multiple myeloma
malignant melanoma
renal cell cancer
bladder cancer (superficial tumors)
ovarian cancer
Inactive: breast cancer
colon cancer
non-small cell lung cancer
prostate cancer
acute myelogenous leukemia
To be determined: chronic myelogenous leukemia
chronic lymphocytic leukemia
preleukemia
Hodgkin’s disease
osteogenic sarcoma
astrocytoma
Efficacy has been observed most often in
hematological malignancies, particularly those
of pre-B-cell and B-cell origin. For example, in
hairy-cell leukemia, Intron A administered sub-
cutaneously at relatively low doses for continu-
ous periods produced hematological responses
in ca. 90 % of patients and objective responses
in ca. 75 %. Complete responses of relatively
long duration have also been observed in multi-
ple myeloma and low-grade lymphomas.
Most adult solid tumors (e.g., breast, lung,
and colon) have not responded well to Intron
A therapy. However, clear activity has been ob-
served in malignant melanoma, in which the
therapeutic effects were apparently schedule-
dependent because continuous dosing was more
efficacious than intermittent dosing. Tumor bur-
den is probably also important in determining
the response to Intron A because malignant
melanoma patients with large tumor burdens ap-
peared to have a lower incidence of response.
Objective responses have also been observed
in superficial bladder cancer, renal cell cancer,

ovarian cancer, and AIDS-related Kaposi’s sar-
coma.
Intron A has also demonstrated activity in
viral indications. For example, intralesional
administration into genital warts (three times
weekly for three weeks) can result in complete
regression of the injected warts between four and
twelve weeks. Therapeutic activity has also been
observed in laryngeal papillomatosis, in chronic
hepatitis B, and as a prophylactive treatment for
the common cold.
In all these clinical trials, the incidence of
human antibody formation to Intron A was very
low [85]. An antibody response was elicited in
<3 % of more than 500 patients treated sys-
temically and in <1 % of more than 1000 pa-
tients treated intralesionally or intranasally. Sig-
nificantly higher rates of neutralizing antibody
production have been reported for recombinant
IFNβ (ca. 65 % of treated patients) and re-
combinant IFNα-2 a (Roferon) (ca. 25 – 30 % of
treated patients).
The presence of neutralizing antibodies to
Roferon has been reported in 7 out of 12 re-
sponding patients and in 9 out of 29 nonre-
sponding patients with metastatic renal cell car-
cinoma [86]. A comparison of the medium du-
ration of remission between antibody-positive
and antibody-negative patients in this study sug-
gested that antibody formation contributed to
early relapse. Similarly, 17 out of 23 patients in a
phase I clinical trial with a form of recombinant
IFNβ containing a genetically engineered serine
residue at position 17 showed evidence for an-
tibody formation [87]. These data strongly sup-
port that Intron A is significantly less antigenic
in human subjects than other recombinant IFNs.
9. References
1. A. Isaacs, J. Lindenmann, Proc. R. Soc.
(London) Sec. B 147 (1957) 258 – 267.
2. P. P. Trotta, S. K. Narula in S. Baron, F.
Dianzani, G. J. Stanton, W. R. Fleischmann, Jr.
(eds.): Interferon Genes, Univ. of Texas Press,
Austin 1987, pp. 137 – 147.
3. P. P. Trotta in P. L. Kisner, J. F. Smyth (eds.):
Interferon Alpha-2: Preclinical and Clinical
Evaluation, Martinus Nijhoff Medical
Division, The Hague 1985, pp. 13 – 27.
4. P. P. Trotta, R. J. Spiegel in F. M. Muggia (ed.):
Concepts, Clinical Developments, and
Interferons 15
Therapeutic Advances in Cancer
Chemotherapy, Martinus Nijhoff Publishers,
Boston 1987, pp. 4 – 159.
5. D. A. Stringfellow (ed.): Interferon and
Interferon Inducers-Clinical Applications,
Marcel Dekker, New York 1980.
6. K. C. Zoon, D. Miller, D. Zur Nedden, M. W.
Hunkapiller, Proc. Natl. Acad. Sci. U.S.A. 76
(1979) 5601 – 5605.
7. K. C. Zoon in E. DeMaeyer, G. Galasso, H.
Schellekens (eds.): The Biology of the
Interferon System 1981,
Elsevier/North-Holland Biomedical Press,
Amsterdam 1981, pp. 47 – 55.
8. M. Streuli, S. Nagata, C. Weissmann, Science
209 (1980) 1343 – 1349.
9. R. J. Spiegel in [3]pp. 43 – 50.
10. T. L. Nagabhushan et al. in K. C. Zoon, P. D.
Noguchi, T. Y. Liu (eds.): Interferon:
Research, Clinical Application and Regulatory
Consideration, Elsevier, New York 1984,
pp. 79 – 88.
11. C. L. Bennett, N. J. Vogelzang, M. J. Ratain,
S. D. Reich, Cancer Treat. Rep. 70 (1986)
1081 – 1084.
12. K. Foon et al., Cancer Immunol. Immunother.
20 (1985) 193 – 197.
13. E. S. Kleinerman et al., Cancer Res. 46 (1986)
5401 –5405.
14. R. Kurzrock et al., Cancer Res. 45 (1985)
2866 – 2872.
15. J. Rinehart, L. Malspeis, D. Young, J.
Neidhart, Cancer Res. 46 (1986) 5364 – 5367.
16. E. C. Bordon et al., Cancer Res. 45 (1985)
5914 – 5920.
17. L. H. Kronenberg in P. N. Cheremisinoff, R. P.
Quellette (eds.): Biotechnology, Technomic
Publishing Co., Lancaster 1985, pp. 451 – 462.
18. S. Pestka, J. H. Langer, K. C. Zoon, C. E.
Samuel in C. C. Richardson, P. D. Boyer, I. B.
David, A. Meister (eds.): Annual Reviews of
Biochemistry, vol. 56, Annual Reviews, Palo
Alto 1987, pp. 727 – 777.
19. S. Pestka, Arch. Biochem. Biophys. 221
(1983) 1 –37.
20. S. Baron, F. Dianzani, G. J. Stanton (eds.): The
Interferon System: A Review to 1982, Univ. of
Texas Medical Branch Publishers, Galveston
1982, Part 1. Reference 20, Part 2.
21. S. Baron, F. Dianzani, G. J. Stanton (eds.): The
Interferon System: A Review to 1982, Univ. of
Texas Medical Branch Publishers, Galveston
1982, Part 2.
16 Interferons
22. S. Baron, F. Dianzani, G. J. Stanton, W. R.
Fleischmann, Jr. (eds.): The Interferon
System-A Current Review to 1987, Univ. of
Texas Press, Austin 1987.
23. P. E. Came, W. A. Carter: Interferons and
Their Applications, Springer Verlag, New
York 1984.
24. E. DeMaeyer, H. Schellekins (eds.): The
Biology of the Interferon System 1983,
Elsevier, New York 1983.
25. R. M. Friedman, T. M. Merigan, T. Sreevalson:
Interferons as Cell Growth Inhibitors and
Antitumor Factors, Alan R. Liss, New York
1986.
26. I. Gresser (ed.): Interferon 1-1979, Academic
Press, London-New York 1979.
27. I. Gresser (ed.): Interferon 2-1980, Academic
Press, London-New York 1980.
28. I. Gresser (ed.): Interferon 3-1981, Academic
Press, London-New York 1981.
29. I. Gresser (ed.): Interferon 4-1982, Academic
Press, London-New York 1982.
30. I. Gresser (ed.): Interferon 5-1983, Academic
Press, London-New York 1983.
31. I. Gresser (ed.): Interferon 6-1985, Academic
Press, London-New York 1985.
32. I. Gresser (ed.): Interferon 7-1986, Academic
Press, London-New York 1986.
33. H. Kirchner, H. Schellekens (eds.): The
Biology of the Interferon System 1984,
Elsevier, New York 1985.
34. T. C. Merigan, R. M. Friedman (eds.):
Interferons, Academic Press, New York 1982.
35. S. Pestka (ed.): Methods in Enzymology,
vol. 78, Academic Press, New York 1981, Part
A.
36. S. Pestka (ed.), Methods in Enzymology,
vol. 79, Academic Press, New York 1981, Part
B.
37. W. E. Stewart II: The Interferon System, 2nd
ed., Springer-Verlag, New York 1981.
38. W. E. Stewart II, H. Schellekins (eds.): The
Biology of the Interferon System 1985,
Elsevier, Amsterdam-New York 1986.
39. K. C. Zoon, P. D. Noguchi, T.-Y. Liu (eds.):
Interferon: Research, Clinical Application and
Regulatory Consideration, Elsevier Science
Publishers, New York 1984.
40. A. Billiau (ed.): Interferon 1. General and
Applied Aspects, Elsevier, Amsterdam 1984.
41. J. Vilcek, E. DeMaeyer (eds.): Interferon 2.
Interferons and the Immune System, Elsevier,
Amsterdam 1984.
42. R. M. Friedman (ed.): Interferon 3.
Mechanisms of Production and Action,
Elsevier, Amsterdam 1984.
43. N. B. Finter, R. K. Oldham (eds.): Interferon 4.
In Vivo and Clinical Studies, Elsevier,
Amsterdam 1985.
44. J. Taylor-Papadimitriou (ed.): Interferons.
Their Impact in Biology and Medicine,
Oxford University Press, Oxford 1985.
45. K. Paucker, K. Cantell, W. Henle, Virology 17
(1962) 324 – 334.
46. P. Lindahl, P. Leary, I. Gresser, Proc. Natl.
Acad. Sci. U.S.A. 69 (1972) 721 – 726.
47. J. Taylor-Papadimitriou in A. Billiau (ed.):
Interferon 1. General and Applied Aspect,
Elsevier, Amsterdam 1984, pp. 139 – 166.
48. D. A. Eppstein, Y. V. Marsh, Biochem.
Biophys. Res. Commun. 120 (1984) 66 – 73.
49. E. Lvovsky et al., J. Natl. Cancer Inst. 66
(1981) 1013 – 1019.
50. N. B. Lydon et al., Biochemistry 24 (1985)
4131 – 4141.
51. S. E. Salmon et al., J. Clin. Oncol. 1 (1983)
217 – 225.
52. I. Gresser, C. Maury, M. Tovey, Eur. J. Cancer
14 (1978) 97 – 99.
53. M. Tozawa et al., Cancer Treat. Rep. 66
(1982) 1575 – 1577.
54. M. S. Aapro, D. S. Alberts, S. E. Salmon,
Cancer Chemother. Pharmacol. 10 (1983)
161 – 166.
55. C. E. Welander, T. M. Morgan, H. D.
Homesley, P. P. Trotta et al., Int. J. Cancer 35
(1985) 721 – 729.
56. J. L. Taylor, J. L. Sabran, S. E. Grossberg, Eur.
J. Immunol. 10 (1980) 3.
57. P. B. Fisher, R. H. Mufson, I. B. Weinstein,
Biochem. Biophys, Res. Commun. 100 (1981)
823 – 830.
58. E. Lotzova, C. A. Saavary, J. U. Gutterman,
E. M. Hersh, Cancer Res. 42 (1982)
2480 – 2488.
59. A. E. Maluish et al., J. Immunol. 131 (1983)
503 – 507.
60. K. C. Zoon, H. Arnheiter, Pharmacol. Ther.
24 (1984) 259 – 278.
61. T. Sreevalson in R. M. Friedman (ed.):
Interferon-3. Mechanisms of Production and
Action, Elsevier, Amsterdam 1984,
pp. 343 – 387.
62. M. E. Hokland, Acta Pathol. Microbiol.
Immunol. Scand. ,[Suppl.] 93 (1985) 4 – 35.
63. K. Stoh, M. Inoue, S. Kataoka, K. Kumagai, J.
Immunol. 124 (1980) 2589 – 2595.
64. B. R. G. Williams, R. H. Silverman (eds.): The
2-5 A System: Molecular and Clinical Aspects
of the Interferon-Regulated Pathway, Alan R.
Liss, New York 1985.

65. B. Lebleu, G. C. Sen, S. Shaula, B. Carrer,
Proc. Natl. Acad. Sci. U.S.A. 73 (1976)
3107 – 3111.
66. M. Streuli et al., Proc. Natl. Acad. Sci. U.S.A.
78 (1981) 2848 – 2852.
67. T. Taniguchi, S. Ohno, Y. Fuji-Kuriyama, M.
Muramatsu, Gene 10 (1980) 11 – 15.
68. A. Zilberstein, R. Ruggieri, J. H. Horn, M.
Revel, EMBO J. 5 (1986) 2529 – 2537.
69. P. W. Gray, D. V. Goeddel, Nature (London)
298 (1982) 859 – 863.
70. S. Vijay-Kumar et al., “Crystallization and
Preliminary X-ray Investigation of a
Recombinant Form of Human Gamma
Interferon,” J. Biol. Chem. 262 (1987)
4804 – 4805.
71. K. E. Mogensen, K. Cantell, Pharmacol. Ther.
1 (1977) 369 – 383.
72. P. P. Trotta, H. V. Le, B. Sharma, T. L.
Nagabhushan in G. Pierce (ed.): Developments
in Industrial Microbiology, vol. 27, Elsevier,
Amsterdam 1987, pp. 53 – 64.
73. P. Leibowitz, M. Ryan, Eur. Patent Publication
No. 146 (1985) 903.
74. Schering Corp., US 4 315 852, 1982 (P. J.
Leibowitz, M. J. Weinstein).
75. M. Bier et al.: “New Developments in
Isoelectric Focusing,” in E. Gross, J.
Meienhofer (eds.): Peptides: Structures and
Biological Function, Pierce Chem. Co.,
Rockford 1979, pp. 79 – 89.
Interferons 17
76. T. L. Nagabhushan, B. Sharma, P. P. Trotta,
Electrophoresis 7 (1986) 552 – 557.
77. S. Rubinstein, P. C. Fameletti, S. Pestka, J.
Virol. 37 (1981) 755 – 758.
78. K. Weber, M. Osborn, J. Biol. Chem. 244
(1969). 4406 – 4412
79. U. K. Laemmli, Nature (London) 227 (1970)
680 – 685.
80. M. Rubinstein, S. Chen-Kiang, S. Stein, S.
Odenfriend, Anal. Biochem. 111 (1981)
184 – 188.
81. M. K. Radjai, R. T. Hatch, J. Chromatogr. 196
(1980) 319 – 322.
82. P. Edman, Acta Chem. Scand. 4 (1956)
761 – 768. P. Edman, G. Begg, Eur. J.
Biochem. 1 (1967) 80 – 91.
83. J. Levin, F. B. Bang, Bull. Johns Hopkins
Hosp. 115 (1975) 265 – 274.
84. R. J. Spiegel in K. Hellman (ed.): Cancer
Treatment Reviews, vol. 12, Academic Press,
New York 1985, supplement B, pp. 5 – 16.
85. R. J. Spiegel et al., Am. J. Med. 80 (1986)
223 – 228.
86. R. R. Quaseda et al., J. Clin. Oncol. 3 (1985)
1522 –1528.
87. M. Hawkins et al., (eds.): The Biology of the
Interferon System 1984, Elsevier, New York
1985, pp. 503 – 508.
 Intermetallics 1

Intermetallics
Gerhard Sauthoff, Max-Planck-Institut für Eisenforschung, Düsseldorf, Germany

1. Introduction . . . . . . . . . . . . . . 1 9.2. Laves Phases . . . . . . . . . . . . . . 16

1.1. Definition of Intermetallics . . . . . 1 9.3. Rare Earth Metal Compounds and
1.2. Historical Remarks . . . . . . . . . . 2 Other Phases . . . . . . . . . . . . . . 17
2. General Considerations . . . . . . . 3 10. Electrode Materials . . . . . . . . . . 17
2.1. Crystal Structure and Compound 10.1. Mg2 Ni Alloys . . . . . . . . . . . . . . 17
Stability . . . . . . . . . . . . . . . . . 3 10.2. Laves Phase Alloys . . . . . . . . . . 17
2.2. Basic Properties . . . . . . . . . . . . 5 10.3. RNi5 Alloys . . . . . . . . . . . . . . . 18
2.3. Criteria for Compound Selection . 6 11. Coating Materials . . . . . . . . . . . 18
3. Magnetic Materials . . . . . . . . . . 8 11.1. Aluminides . . . . . . . . . . . . . . . 18
3.1. AlNiCo Alloys . . . . . . . . . . . . . 8 11.2. Silicides . . . . . . . . . . . . . . . . . . 19
3.2. FeCo Alloys . . . . . . . . . . . . . . . 8 12. Shape-Memory Materials . . . . . 19
3.3. Sendust Alloy . . . . . . . . . . . . . . 9 12.1. Cu-Base Alloys . . . . . . . . . . . . . 19
3.4. Laves Phases . . . . . . . . . . . . . . 10 12.1.1. Cu – Zn – Al Shape Memory Alloys 19
3.5. Rare-Earth Compounds . . . . . . . 10 12.1.2. Cu – Al – Ni Shape-Memory Alloys 19
4. Superconducting Materials . . . . . 11 12.2. NiAl Alloys . . . . . . . . . . . . . . . 20
4.1. Laves Phases . . . . . . . . . . . . . . 11 12.3. NiTi Alloys . . . . . . . . . . . . . . . 20
4.2. A15 Compounds . . . . . . . . . . . . 11 13. Medical Biomaterials . . . . . . . . 21
4.2.1. V3 Si . . . . . . . . . . . . . . . . . . . . 11 14. Dental Materials . . . . . . . . . . . . 21
4.2.2. V3 Ga . . . . . . . . . . . . . . . . . . . 12 14.1. Amalgams . . . . . . . . . . . . . . . . 21
4.2.3. Nb3 Sn . . . . . . . . . . . . . . . . . . . 12 14.2. Cu – Au Alloys . . . . . . . . . . . . . 22
4.2.4. Nb3 Al . . . . . . . . . . . . . . . . . . . 13 15. Wear-Resistant Materials . . . . . . 22
4.2.5. Nb3 Si . . . . . . . . . . . . . . . . . . . 13 16. Structural High-Temperature
5. Electronic Materials . . . . . . . . . 13 Materials . . . . . . . . . . . . . . . . . 23
5.1. Silicides . . . . . . . . . . . . . . . . . . 13 16.1. Titanium Aluminide Alloys . . . . . 23
5.2. Other Compounds . . . . . . . . . . 14 16.1.1. Ti3 Al Alloys . . . . . . . . . . . . . . . 23
6. Electric Materials . . . . . . . . . . . 14 16.1.2. TiAl Alloys . . . . . . . . . . . . . . . . 23
7. Thermoelectric and 16.2. Iron Aluminide Alloys . . . . . . . . 24
Thermomagnetic Materials . . . . 14 16.2.1. Fe3 Al Alloys . . . . . . . . . . . . . . . 24
7.1. Silicides . . . . . . . . . . . . . . . . . . 14 16.2.2. FeAl Alloys . . . . . . . . . . . . . . . 25
7.2. Other Thermoelectric Compounds 15 16.3. Nickel Aluminide Alloys . . . . . . . 25
7.3. Thermomagnetic Compounds . . 15 16.3.1. Ni3 Al alloys . . . . . . . . . . . . . . . 25
8. Optical Materials . . . . . . . . . . . 15 16.3.2. NiAl Alloys . . . . . . . . . . . . . . . 26
9. Hydrogen-Storage Materials . . . 16 16.4. Silicide Alloys . . . . . . . . . . . . . 27
9.1. B2 Compounds . . . . . . . . . . . . . 16 17. References . . . . . . . . . . . . . . . . 28

1. Introduction cording to the simple definition in [1] inter-

1.1. Definition of Intermetallics
Intermetallics is the short and summarizing
designation for intermetallic compounds and
phases, which result from the combination of
various metals and which form a tremendously
numerous and manifold class of materials. Ac-
metallics are compounds of metals whose crys-
tal structures differ from those of the constituent
metals, and intermetallic phases and ordered al-
loys are thus included. During the last 20 years
intermetallic compounds (IMCs) have aroused
enormous and still-increasing interest in materi-
als science and technology with respect to appli-
cations at high temperatures. Various new struc-

c 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.e14 e01.pub2
2 Intermetallics
tural materials are being developed around the
world, in particular in the United States, Japan,
and Germany. Besides intermetallics with out-
standing high-temperature performance, inter-
metallics with other outstanding physical prop-
erties led to the development of new functional
materials much earlier.
The results of the numerous and manifold
research and development activities on inter-
metallics have been published in the respec-
tive conference proceedings volumes and review
monographs [2 – 8] in addition to scientific jour-
nals and research reports.
The present overview, which is an updated
summary of a previous report [2], briefly de-
scribes the various intermetallics which have al-
ready been selected for materials developments
or which have been and still are regarded as
promising for materials developments.
1.2. Historical Remarks
Intermetallics have been exploited since the be-
ginning of metallurgy, as is exemplified in Table
1. The early intermetallics resulted from alloy
systems with low melting temperatures, and the
applications relied on the outstanding hardness
and wear resistance of intermetallics together
with their metallic properties. Because of the
metallic behavior of intermetallics, they could
be polished, and thus the decorative aspect was
also important for many applications. Examples
are the bronze coatings in ancient Egypt and the
mirrors used by the ancient Chinese, Etruscans,
and Romans. Some intermetallics show beauti-
ful colors, e.g. the violet Cu2 Sb (“Regulus of
Venus”).
Intermetallics became a subject of scien-
tific research with the development of physi-
cal metallurgy. Intensive and broad-scale work
on intermetallics was initiated by Tammann in
Göttingen, Germany, and K urnakov in St. Pe-
tersburg, Russia, around 1900. Work on inter-
metallics during the first decades of the 1900s al-
ready included studies of phase stabilities, phase
equilibria, and phase reactions in order to es-
tablish phase diagrams, and investigations of
chemical, electrochemical properties, and phys-
ical properties including magnetism, supercon-
ductivity, and mechanical properties. It was rec-
ognized that the outstanding hardness of inter-
metallics is accompanied by an unusual brittle-
ness, and the reasons for this were sought.
In the 1900s intermetallics were used primar-
ily as functional materials. The first industrial
applications relied on the special magnetic be-
havior of certain compounds, and materials de-
velopments led, e.g., to Sendust which shows
outstanding magnetic properties and wear resis-
tance and is widely used for magnetic heads in
tape recorders. In the second half of the twen-
tieth century another important application re-
sulted from the development of new supercon-
ducting materials on the basis of the A15 com-
pounds, which are used for superconducting
magnets. A third group of functional materi-
als — the shape-memory alloys — make use of a
martensitic phase transformation and have found
manifold applications since ca. 1970.
An important group of functional materials
are the III/V compounds such as InSb, InAs, and
GaAs, which have found applications in elec-
tronics and thermoelectric power generation.
The constituent elements of these compounds re-
present the transition from metals to metalloids
and nonmetals, and the resulting compounds
are semiconductors. Hence these intermetallic
semiconductors lie outside the field of true in-
termetallics with metallic behavior.
Intermetallics were not used as structural ma-
terials in the past because of their brittleness.
The only noteworthy exception is the continu-
ing use of amalgams as dental restoratives. How-
ever, various intermetallics have been success-
fully used as strengthening second phases in
conventional alloys for structural applications at
high temperatures. Thus it became clear that in-
termetallics are promising as structural materials
for high temperatures because of their high hard-
ness and stability, and numerous investigations
began around 1950 with the aim of clarifying the
potential of intermetallics for structural applica-
tions. Various promising candidates were iden-
tified, but the respective materials developments
were hampered by unsolved brittleness prob-
lems, and these various activities slowly died
out in the 1960s.
One offspring of these activities was the
development of the electric heating elements
based on MoSi2 , which rely on the advanta-
geous chemical behavior of this compound, that
is, its high oxidation resistance at high tempera-
tures. In the course of superalloy research, sur-
 Intermetallics 3
Table 1. Some applications of intermetallics in past and present times [2]

Since ca. Material or process Phase Application

2500 b.c. cementation Cu3 As coating of bronze tools, etc. (Egypt, Anatolia, Britain)
100 b.c. yellow brass CuZn coins, ornamental parts (Rome)
0 high-tin bronze Cu31 Sn8 mirror (China)
600 amalgam Ag2 Hg3 + Sn6 Hg dental restorative (China)
1500 amalgam Cu4 Hg3 dental restorative (Germany)
1505 amalgam Sn8 Hg mirror surface (Venice)
1540 type metal SbSn printing
1910 Acutal (Cu,Mn)3 Al fruit knife (Germany)
1921 Permalloy Ni3 Fe high-permeability magnetic alloy
1926 Permendur FeCo( – 2V) high permeability magnetic alloy
1931 AlNiCo NiAl – Fe – Co permanent magnet material
1935 Sendust Fe3 (Si,Al) magnetic head material
1938 Cu – Zn – Al, Cu – Al – CuZn – Al, (Cu,Ni)3 Al shape memory alloys
Ni
1950 pack aluminide coating NiAl, CoAl surface coating for protection from environment
1956 Kanthal Super, Mosilit MoSi2 electric heating elements
1961 A15 compound Nb3 Sn superconductors
1962 Nitinol NiTi shape-memory alloy
1967 Co – Sm magnets Co5 Sm permanent magnets

face coatings that protect the coated materials with the aim of lower density and/or higher ap-
against high-temperature corrosion and which plication temperatures.
are based primarily on such intermetallics as
NiAl and CoAl were developed.
The above-mentioned molybdenum disili- 2. General Considerations
cide is a borderline case of an intermetallic, since
silicon is not a metal, but a semiconductor. It is 2.1. Crystal Structure and Compound
known that the combination of silicon with met- Stability
als gives rise to compounds with metallic prop-
Intermetallics form because the strength of
erties like MoSi2 , as well as compounds with
bonding between unlike atoms is larger than that
semiconductor properties, that is, silicides repre-
between the respective like atoms. Accordingly,
sent the transition from intermetallics to com-
intermetallics have particular crystal structures
pounds of metals and nonmetals. Nevertheless,
with ordered atom distributions in which atoms
silicides are traditionally included in the field of
are preferentially surrounded by unlike atoms.
intermetallics because of their many similarities
The known intermetallic compounds are listed
with metals.
with their crystal structures in [9].
The first intensive and successful develop-
The crystal structure of an intermetallic is de-
ments of structural materials were based on the
termined by the strength and character of bond-
titanium aluminides Ti3 Al and TiAl and began
ing in the crystal, which depends on the particu-
in the early 1970s, although the potential of
lar electron configuration. The relation between
these compounds for high-temperature applica-
structure type and atomic properties of the con-
tions had already been recognized in the 1950s.
stituent atoms, however, is not a simple one, and
Other successful structural materials were
therefore various criteria have been used for cor-
based on the nickel aluminide Ni3 Al, which is
relating structure type and compound type. The
the strengthening second phase in the super-
grouping of intermetallics according to simple
alloys and was also a candidate compound in
criteria is not unambiguous, and it cannot be ex-
the 1950s. This development was given impe-
pected that all intermetallics will show metal-
tus by the discovery of the “ductilizing” effect
lic bonding like the constituent metals. In the
of small additions of boron, which led to major
following a brief survey of three important and
research programs on intermetallics and excited
rather different groups of intermetallics is given
the general and still-increasing interest in inter-
to illustrate this.
metallics. These successful developments ini-
The Zintl phases are formed by metals on
tiated work on other less common compounds
the left-hand side and on the right-hand side of
4 Intermetallics
the periodic table of elements and are charac-
terized by completely filled electron orbitals,
in particular, by a full octet shell in the nor-
mal case. Thus they may be regarded as valence
compounds which satisfy the familiar chemi-
cal valency rules. The Zintl phases have crys-
tal structures which are characteristic for typ-
ical salts, for example, NaTl with cubic B32
structure (Strukturbericht designations are used
here and in the following for the crystal struc-
tures) and Mg2 Si with cubic C1 structure, and
therefore ionic bonding is expected. However,
all types of bonding — ionic, metallic, and co-
valent — and mixtures thereof have been found,
depending on the particular electron distribu-
tion. Hence the Zintl phases may also be re-
garded as electron compounds whose electron
band-structure energy is a large fraction of the
total energy and the crystal structure types of
which are related to particular valence electron
concentrations (average number of valence elec-
trons per atom).
The best-known electron compounds are the
Hume – Rothery phases, which have the cubic
B2 structure (e.g., β-brass at high temperatures
and the transition metal aluminides FeAl, NiAl,
and CoAl), the complex cubic A13 structure
(β-Mn type; e.g., Zn3 Co, Cu5 Si), or the close-
packed hexagonal A3 structure (e.g.; Cu3 Ga,
Ag3 Al) for the valence electron concentration
VEC = 3/2; the complex cubic D82 structure
(γ-brass type) for VEC = 21/13 (e.g.; Cu5 Zn8 ,
Fe5 Zn2 ); and the A3 structure again for VEC =
7/4 (e.g., CuZn3 , Ag5 Al3 ). The bonding in such
intermetallics is not purely metallic, in spite of
the metal-like band structure. For example, the
bonding in the above-mentioned NiAl is essen-
tially covalent with some metallic character and
no ionic component. Recent ab initio calcula-
tions showed that, strictly speaking, the B2 alu-
minides are not Hume – Rothery electron com-
pounds. This means that the Hume – Rothery
rules, which relate crystal structure to valence
electron concentration, represent a very simpli-
fied picture of the bonding situation. Neverthe-
less, such rules are useful since they describe
reality surprisingly well.
Other intermetallics are best characterized
by the size ratios and packing schemes of the
constituent atoms and are therefore called size-
factor compounds, topologically close packed
intermetallics, or Frank – Kaspar phases. One
important group of these intermetallics are the
A15 compounds. The best known size-factor
compounds are the Laves phases, which form
the most numerous group of intermetallics and
which crystallize in the closely related hexago-
nal C14, cubic C15, or hexagonal C36 structures,
whereby MgZn2 , MgCu2 , and MgNi2 , respec-
tively, are typical examples. The valence elec-
tron concentration, however, is also an impor-
tant factor here, since it decides between the
three crystal structures. These crystal structures
are characterized by high symmetry, high atomic
coordination, and high density in close analogy
to metals, and indeed the Laves phases show pri-
marily metallic bonding according to the scarce
information available.
Compounds which form by a phase transi-
tion with lowering of crystal symmetry from the
solid solution of the constituent elements at a
transition temperature below the melting tem-
perature are known as Kurnakov phases. In the
simplest case the phase transition results from an
ordering reaction in the solid-solution lattice in
which a superlattice structure is established. Ex-
amples for compound formation by atomic or-
dering are the compounds Fe3 Al with D03 struc-
ture, which is formed by two-step ordering from
the bcc solid solution, as well as Ni3 Fe with L12
structure, CuAu with L10 structure, and Cu3 Au
with L12 structure, the latter three of which form
from fcc solid solutions. This atomic ordering
with superlattice formation clearly results from
stronger bonding between unlike atoms com-
pared with bonding to like atoms. The difference
in bonding energy between unlike and like atoms
is usually known as interaction energy and cor-
responds to an ordering energy. Hence, higher
crystal stability and higher ordering temperature
are expected for a higher interaction energy. The
degree of order, which depends on temperature
and composition, as well as the order/disorder
temperature can be modeled as a function of
the interaction energy for a given compound to
various degrees of approximation, for example,
to a first approximation by the simple Bragg –
Williams model or to a better approximation by
the cluster variation or the Monte Carlo method.
Apart from the degree of approximation, the pre-
dictive power of such models with respect to
phase-stability and phase-diagram calculations
relies on the knowledge of the interaction en-
ergy and on whether such interaction energies

are sufficient to characterize the atomic bond-
ing state. Analyses of various transition metal
systems have shown that the modeling with the
cluster variation or Monte Carlo method gives
good agreements with the experimental obser-
vations in most cases, whereas in some com-
pounds with bcc structures the simple Bragg –
Williams model gave better results than the more
sophisticated models.
It follows that single simple atomic param-
eters like the interaction energy, atomic size,
or valence electron concentration have only a
limited validity and are generally not sufficient
for predicting the stability of a compound as a
function of the constituent elements. A better
and more general correlation between the atomic
properties and the compound crystal structure
should be obtained if the particularities of the
electron configurations of the constituent atoms
are taken into account; these are reflected in the
positions in the periodic table of elements. A
better correlation has indeed been found, but the
resulting structure maps still have their limita-
tions because the changes in electron configura-
tion due to compound formation are not consid-
ered.
It must be concluded that the bonding charac-
ter and crystal structure of an intermetallic com-
pound can be predicted in a significant way only
on the basis of quantum-mechanical ab initio
calculations on the respective compound. Much
progress has been made in this respect, and for
various important compounds with not too com-
plicated structures the crystal energy has been
calculated as a function of lattice structure in or-
der to determine the compound stability. How-
ever, these calculations are very time-consuming
even in simple cases and more progress is nec-
essary in the future in order to give guidance to
practical materials developments on the basis of
multinary compounds with less simple crystal
structures.
The intermetallics do not form a homoge-
neous group of materials. Instead the term in-
termetallics comprises a huge variety of com-
pounds which differ drastically with respect
to bonding, crystal structure, and properties.
Therefore, properties of intermetallics cannot be
discussed generally for all intermetallics, but can
be discussed only by referring to specific groups
of intermetallics.
Intermetallics 5
It might be supposed that the particular crys-
tal structure of an intermetallic reflects the char-
acter and strength of bonding in a systematic
way and that crystal structure would be a good
criterion for compound classification. However,
this does not mean that intermetallics with the
same crystal structure are similar with respect
to bonding and properties. For example, the B2
structure is common to the intermetallic NiAl
and to the purely ionic salt CsCl, and in the case
of the transition-metal disilicides with hexago-
nal C40 structure, NbSi2 is metallic but CrSi2 is
a semiconductor. Recent ab initio calculations
showed that atomic coordination and connectiv-
ity is more relevant to crystal energy and sta-
bility than symmetry. In view of the complex-
ity of the classification of intermetallics, inter-
metallics are often grouped according to more
practical criteria which refer to similarities in
behavior.
2.2. Basic Properties
A basic property is the melting temperature since
materials parameters which characterize the de-
formation behavior are well correlated with the
melting temperature. Examples are the elastic
moduli, which not only control the elastic de-
formation, but are also important parameters for
describing plastic deformation, and the diffusion
coefficients, which determine not only the ki-
netics of phase reactions, but also the kinetics
of high-temperature deformation, that is, creep.
Furthermore, the melting temperature is intu-
itively regarded as a measure of compound sta-
bility, since it limits the application temperature
range.
However, a compound melts when the Gibbs
energy of the solid phase, which is a thermo-
dynamic state function and which controls the
phase stability, is higher than that of the liq-
uid phase. The Gibbs energy of a phase is given
by the enthalpy and the product of entropy and
temperature, that is, its temperature dependence
is determined primarily by the entropy. Hence,
the melting temperature depends on the differ-
ences in enthalpy and entropy between the solid
and liquid states and is a rather complex func-
tion of the bonding strength. Only the enthalpy
is directly related to the internal crystal energy,
which is determined by the bonding strength.
6 Intermetallics
Nevertheless, the melting temperature correlates
surprisingly well with the enthalpy of formation
for sufficiently similar compounds.
It may be concluded that the enthalpy of com-
pound formation may be a better parameter for
characterizing bonding strength and compound
stability and for correlating this with fundamen-
tal properties such as elastic moduli. Formation
enthalpies have been determined experimen-
tally, and they can be calculated by quantum-
mechanical calculations or estimated by more
or less empirical approaches.
However, the conventional formation en-
thalpy values refer to compound formation from
the constituent elements at room temperature,
that is, from the solid elemental crystals, and thus
these formation enthalpies characterize only
bonding strength changes during compound for-
mation. For the complete characterization of
bonding strength the total compound enthalpy
must be considered, which refers to compound
formation from the elements in the gaseous state.
The total enthalpy of a compound may be es-
timated by adding the formation enthalpy and
the enthalpies of fusion and vaporization of the
constituent elements. As an example, data for
the Ni – Al system show a good agreement of
estimates with experimental data. Since the for-
mation enthalpies are small compared with the
total enthalpies, the enthalpies of the constituent
elements are the major contributions to the total
enthalpies.
The correlation between the total enthalpy of
a compound and Young’s modulus, which is one
of the three elastic moduli, has been checked for
various cubic compounds: for the fcc elements
Al and Ni (A1), the fcc-ordered Ni3 Al (L12 ), the
bcc-ordered FeAl, NiAl, and CoAl (B2), and the
cubic Laves phases CaAl2 , YAl2 , LaAl2 , NbCr2 ,
ZrCo2 and HfCo2 (C15). The correlation is quite
good for the Laves phases, in that there seems
to be a common scatter band, near which the
data of some other compounds lie. However, the
data points may be subject to appreciable exper-
imental uncertainties, as indicated by data for
Ni3 Al. Such uncertainties for Young’s modulus
may be due to differences in microstructure, that
is, texture and artefacts. Apart from this, the data
for the B2 compounds show that Young’s mod-
ulus is not a simple function of the enthalpy of
the compound. This must be expected because
Young’s modulus is a complex function of the
potential wells of the atoms, which cannot be
characterized by a single parameter. A similar
correlation has been found for the activation en-
ergy of diffusion.
It must be concluded that there are physically
justified correlations between parameters which
characterize bonding strength and compound
stability and the macroscopic compound behav-
ior. However, such correlations represent com-
plex functional relationships, and hence they are
useful only for order-of-magnitude predictions.
More quantitative predictions have to consider
the character and strength of bonding in a more
detailed way, that is, they have to rely on quan-
tum mechanical calculations which are cumber-
some and time-consuming.
Recently, fundamental properties, especially
elastic moduli, of various simple compounds
have been studied by ab initio calculations, and
much progress has been made in this field. Thus,
a basic understanding of the bonding character
of, for example, transition metal aluminides was
reached, and it was shown that the degree of
directional d-electron bonding determines the
ideal cleavage strength. However, it has also
been shown for B2 transition-metal aluminides
that the ductility or brittleness is too complex a
function of strength and character of bonding for
establishing simple relations between electron
distribution parameters and mechanical proper-
ties.
2.3. Criteria for Compound Selection
Candidate compounds for materials develop-
ments are selected in view of specific applica-
tions, that is, they must show certain properties
which make them promising for a particular ap-
plication. For permanent magnets the “energy
product” (BH)max (B = magnetic induction, H
= magnetic field strength) is a figure of merit
and should reach high values whereas for other
functional materials other physical parameters
are decisive.
Intermetallic compounds for structural appli-
cations must show a sufficient strength at service
temperature which also means a sufficient creep
resistance. The creep resistance is correlated
with the diffusion coefficient and with the shear
modulus, and both parameters in turn with the
melting temperature. Hence, the candidate com-

pound should have a sufficiently high melting
temperature. The limiting temperature for struc-
tural applications of conventional metallic alloys
is of the order of 75 % of the melting temperature
in many cases. The most advanced high-tempe-
rature alloys are the Ni-based superalloys, which
are used in gas turbines with service tempera-
tures of up to about 1100 ◦ C. Hence compounds
with melting temperatures above 1600 ◦ C must
be considered if application temperatures above
those of superalloys are aimed at. High-melting
compounds may form low-melting eutectics that
can affect the stability of the compound dur-
ing processing or service, and therefore such
compounds with low-melting eutectics should
be avoided.
For many functional and structural applica-
tions density is a very important and sometimes
decisive parameter. Structural intermetallics for
moving components must show a sufficiently
high specific strength, which is the ratio of
strength to density and has the dimension of
length. The specific fracture strength, that is,
the rupture length, of advanced superalloys such
as the mechanically alloyed oxide dispersion
strengthened (ODS) superalloy MA 6000 is of
the order of 15 km, which would have to be sur-
passed by new structural materials. Compounds
which contain light elements — e.g. Ti, Al, Si,
Mg — may have such a low density that they
compare favorably with conventional alloys in
spite of lower strength or restricted temperature
range.
The main problem of strong intermetallics is
their brittleness, which makes processing and
application difficult or impossible. However, the
brittleness of intermetallics should be less severe
than that of ceramics because the atomic bond-
ing of intermetallics is at least partially metal-
lic, whereas it is primarily covalent or ionic in
ceramics. This is illustrated by the temperature
dependence of yield stress and deformability of
the intermetallic NiAl in comparison with that
of nickel on the one hand and of silicon and
alumina on the other hand. NiAl which is a can-
didate compound for high-temperature applica-
tions softens with rising temperature at about
half the melting temperature T m , which is the
range of Si with covalent bonding and which
is higher than that of the ceramic Al2 O3 . How-
ever, the brittle-to-ductile transition temperature
is about 0.4 T m for NiAl whereas it is above 0.8
Intermetallics 7
T m for Si and Al2 O3 , that is, the deformation be-
havior of such an intermetallic may be regarded
as intermediate between those of metals and ce-
ramics.
Plastic deformation is more difficult in inter-
metallics than in metals and conventional metal
alloys because of the stronger atomic bond-
ing and the resulting ordered atom distribution,
which gives rise to more complex crystal struc-
tures. The brittleness of intermetallics increases
with decreasing lattice symmetry and increasing
unit cell size. Therefore intermetallics with high
crystal symmetry — possibly cubic compounds
— and small unit cells are preferred for develop-
ing new structural materials. Indeed the candi-
date compounds FeAl, NiAl, Fe3 Al, Ni3 Al, TiAl
for development of structural materials have
crystal structures which are cubic (B2, D03 or
L21 , L12 , L 12 ) or nearly cubic, that is, tetrago-
nally distorted (L10 , D022 ).
However, even for such selected compounds,
brittleness remains a problem, and the causes
for this are manifold. A material fractures in a
brittle manner if there is no plastic deformation
and stress relaxation at the crack tip, that is, the
yield stress is higher than the stress for cleav-
age or fracture. The reason for this may be an
insufficient number and/or mobility of disloca-
tions and/or an insufficient number of slip sys-
tems. Since these parameters depend not only on
crystal symmetry but also on the specific char-
acteristics of the particular compounds, the aim
of compound selection and composition varia-
tion by alloying must be dislocations with low
energy and high mobility and at least five in-
dependent deformation modes (i.e. dislocation
slip systems and twinning systems), which are
necessary for general homogeneous plastic de-
formation according to the Von Mises criterion.
Crystal anisotropy is an important factor in brit-
tleness, and criteria are used which rely on elas-
tic moduli relations. Brittle fracture may be a
result of weak grain boundaries and other mi-
crostructure heterogeneities leading to localized
deformation and stress concentrations. The dis-
proportion of yield stress and fracture stress may
also be due to a very low surface energy, which
makes cleavage and cracking easy. This may
be aggravated by the segregation of impurities,
which usually lower surface energies. Impuri-
ties, in particular oxygen and hydrogen, may
enter the material from the environment, so en-
8 Intermetallics
vironmental embrittlement is an important issue
for many intermetallics. A detailed knowledge
of the thermodynamics and kinetics of the par-
ticular alloy system is necessary for a reduction
of the various embrittling effects.
Intermetallics for any application must be
corrosion resistant in the operating environment.
For high-temperature applications this means
oxidation resistance in many cases. Oxidation
resistance is provided by the presence of ele-
ments, in particular Cr, Al, Si, which can form
protective oxide layers. However, chromium ox-
ide is volatile above 1000 ◦ C and silicon oxide
may form low-melting silicates. Therefore, alu-
minides are strongly favored for high-tempera-
ture applications. To act as protective layers, the
scales must show a sufficient density and adher-
ence, which may be improved by alloying with
further elements (macroalloying) and in partic-
ular by adding minor amounts of “active” ele-
ments such as Ti, Zr, Hf (microalloying). Fur-
thermore, scales must show sufficient long-term
stability, which also means mechanical stabil-
ity, that is, they must tolerate strains, for exam-
ple, during creep, without being damaged. In the
case of insufficient oxidation resistance protec-
tive coatings may be applied. However, the ther-
mal stability of such coatings decreases with in-
creasing temperature because of increasing re-
action and diffusion rates and therefore inter-
metallics for applications above 1100 ◦ C should
be inherently oxidation-resistant.
Finally, it must already be possible to prepare
an intermetallic with sufficient quality in the lab-
oratory, and from a practical point of view this
is the basic requirement for any materials devel-
opment. The processing of high-strength inter-
metallics is difficult because of their brittleness,
and the development of the necessary processing
techniques is a very demanding task. It is, how-
ever, also a very important task, because a poor
quality increases the apparent brittleness and re-
duces the strength, and this may then preclude
any applications of the tested material.
3. Magnetic Materials
3.1. AlNiCo Alloys
The advantageous magnetic properties of Fe –
Ni – Al alloys with suitable composition were
first discovered by Mishima in 1931, and this
initiated the development of the AlNiCo al-
loys for applications as permanent magnets (→
Magnetic Materials, → Cobalt and Cobalt Com-
pounds, Chap. 6.3). The basic alloy contains
about 50 atom % Fe, 25 atom % Ni, and 25 atom
% Al, to which various other elements (Co,
Cu as macroalloying elements and Nb, Ti as
microalloying elements) are added for prop-
erty optimization. The alloys are single-phase at
high temperatures and decompose spinodally on
cooling at a controlled rate to form a two-phase
microstructure with interlocking phase distribu-
tion. The additional alloying elements Co, Mn,
Ti, and Cu, which dissolve preferentially in the
B2 compound, and Cr, Mo, V, Nb, and Si, which
dissolve preferentially in the disordered Fe-rich
compound, control the decomposition process
of the AlNiCo alloys. The constituent phases
of these alloys are the weakly ferromagnetic
NiAl base compound with B2 structure and the
strongly ferromagnetic, disordered bcc Fe-rich
compound. The interlocking phase distribution
and the strong shape anisotropy of the Fe-rich
compound result in high coercivity. The mag-
netic behavior is a sensitive function of pretreat-
ment.
AlNiCo alloys, which are produced by cast-
ing or sintering, are brittle and hard; they can
be machined only by grinding, spark erosion,
and electrochemical milling, and they resist at-
mospheric corrosion well up to 500 ◦ C. The
magnetic properties are summarized and com-
pared to those of other alloys in [10 – 12]. Al-
NiCo alloys are superior to other permanent
magnet materials in resisting the effects of tem-
perature on magnetic properties. Presently, NiAl
is also considered an attractive candidate com-
pound for other functional applications, for ex-
ample, contacts in thin-film electronic devices
or high-voltage electrodes in electron-optical in-
struments.
3.2. FeCo Alloys
Permendur, an FeCo alloy containing 2 % V, is
a soft magnetic material with high saturation in-
duction (e.g., for electromagnet pole tips [12]
(→ Magnetic Materials). These alloys also ex-
hibit a high positive magnetostrictive coefficient,
which has made them useful in sonar transducers

and in high-accuracy positioning devices. Simi-
lar Fe – Co – V alloys with more V and less Fe
are known as Vicalloy alloys.
The B2 compound FeCo is stable only at tem-
peratures below 730 ◦ C and has a broad range
of homogeneity. At 730 ◦ C there is an order –
disorder transition with 0.2 % change in volume
from B2 to A2 (bcc), and a further structure
change from A2 to A1 (fcc) occurs at about
985 ◦ C. The ordering reaction proceeds het-
erogeneously or homogeneously, depending on
temperature. The ordering kinetics are affected
by alloying additions, and in particular alloy-
ing with 2 atom % V reduces the ordering rate
significantly and allows the freezing of the high-
temperature disordered structure by quenching
to low temperatures. The addition of more than
2 % V leads to alloy decomposition, as is indi-
cated by the Fe – Co – V phase diagram. A still
larger effect on the ordering kinetics is produced
by alloying with Cr. Below the critical temper-
ature of ordering the degree of long-range order
at equilibrium increases continuously from 0 at
730 ◦ C to 0.9 at 500 ◦ C and approaches 1 (per-
fect order) asymptotically at low temperatures.
The mechanical behavior of FeCo has been
studied intensively including creep, fatigue, and
fracture. At room temperature FeCo is ductile,
with 15 % uniform elongation and subsequent
necking, in the disordered state, that is, after
quenching from above the ordering temperature,
whereas in the ordered state, that is, after an-
nealing below the ordering temperature, it shows
only little ductility with 5 % uniform elongation
and no necking. Hence cold rolling is difficult in
the ordered state. Furthermore, FeCo is subject
to hydrogen embrittlement at room temperature.
The flow stress increases with decreasing de-
gree of order, that is, with rising temperature,
to reach a maximum at about 700 ◦ C just below
the ordering temperature when FeCo is only par-
tially ordered. Above the flow stress peak tem-
perature, normal softening is observed due to
thermal activation. In other words, FeCo, like
other intermetallics, exhibits an anomalous pos-
itive temperature dependence of flow stress.
Ordered FeCo – 2V exhibits more homoge-
neous deformation and a higher strain-hardening
rate than disordered FeCo – 2V at all temper-
atures. The yield stress and fracture stress in-
crease with decreasing grain size according
to the usual Hall – Petch relationship. The fa-
Intermetallics 9
tigue resistance is improved by ordering with
only a small effect of the environment. The
low ductility can be increased by adding Ni as
fourth element, which changes the microstruc-
ture, and still more by thermomechanical treat-
ment, which results in a Hall – Petch-type de-
pendence of ductility on grain size or subgrain
size. Even cold rolling can reduce the brittle-
ness. The brittle-to-ductile transition tempera-
ture in the ordered state can be reduced signifi-
cantly by deviations from stoichiometry, that is,
by decreasing the Co content, which decreases
the critical temperature of ordering. Positive ef-
fects on cold workability have been produced
— apart from alloying with V or Cr which both
slow down the ordering kinetics — by alloying
with V, Nb, Ta, Cr, Mo, W, and C, whereas Be,
Ti, Zr, Mn, Cu, Ag, Au, B, Al, and Si are inef-
fective.
Alloying of Fe and Co with precious met-
als leads to the compounds FePt, CoPt, FePd
with L10 structure, which is an ordered fcc
structure with tetragonal distortion. These com-
pounds have been of importance for applications
as permanent-magnetic materials, and in par-
ticular CoPt, which is used in the partially or-
dered state, excels because of isotropy, ductility,
easy machinability, and resistance to corrosion
and high temperatures (see also [12]). However,
CoPt alloy magnets are seldom used now since
these alloys have been replaced by rare earth
magnets with superior magnetic properties.
3.3. Sendust Alloy
The iron aluminide Fe3 Al with DO3 structure
exhibits an outstanding high magnetic perme-
ability which makes it useful as a magnetic
material. The magnetic Fe3 Al-based alloys are
known as Alfenol alloys [12]. Fe3 Al is formed
on cooling by ordering reactions in the solid state
that transform the bcc disordered solid solution,
which is stable above about 800 ◦ C, first into the
FeAl compound with B2 structure, which is sta-
ble between about 800 ◦ C and 550 ◦ C, and then
into Fe3 Al with the D03 structure. The magnetic
properties can be improved by partial substitu-
tion of Al by Si, which does not change the crys-
tal structure. This gave rise to the development of
the Sendust alloy which is based on Fe3 (Al,Si)
10 Intermetallics
and which has found widespread application as
magnetic head material in recorders.
3.4. Laves Phases
The Laves phases, sometimes designated as Fri-
auf – Laves phases, with AB2 composition in
the binary case form a very large group of in-
termetallics which crystallize with the hexago-
nal C14 structure, the cubic C15 structure, or
the dihexagonal C36 structure. These structures
are topologically close packed (tcp) structures,
that is, the Laves phases belong to the family
of the Frank – Kaspar phases. The stability of
the three crystal structures C14, C15, and C36 is
controlled by the atomic-size ratio of the atoms
A and B and by the valence electron concentra-
tion of the Laves phase.
Various Laves phases have been regarded as
promising for both functional and structural ap-
plications. Some transition-metal Laves phases,
such as TFe2 (T = Ti, Zr, Hf, Nb, Mo) and
NbCo2 , as well as the closely related rare-earth
Laves phases, show interesting magnetic prop-
erties as a result of the particular electronic
structures. The rare-earth compounds TbFe2 and
SmFe2 , both with C15 structure, exhibit huge
magnetostrictions at room temperature. In addi-
tion there is a magnetoelastic interaction, that
is, unusual changes in elastic moduli are ob-
served in a magnetic field. The ternary C15
compound (Tb1−x Dyx )Fe2 , which is known
as Terfenol-D, has much potential for applica-
tions as a high-power transducer, since it com-
bines a large magnetostriction with a vanish-
ing magnetic anisotropy (for x ≈ ± 0.7). Like
other Laves phases, this material is brittle, with
an apparent brittle-to-ductile transition temper-
ature of about 875 ◦ C.
Various rare earth (R) Laves phases, espe-
cially RFe2 , can absorb large amounts of hy-
drogen (see Section 10.2) which may affect
their structures, that is, they can undergo struc-
ture transformations on hydrogenation, includ-
ing amorphization, which may be accompanied
by decomposition reactions. Such hydrogen-
ation and dehydrogenation reactions have been
used for producing fine-grained powders for the
powder metallurgical production of so-called
giant-magnetostrictive RFe2 alloys.
3.5. Rare-Earth Compounds
There is a multitude of intermetallic rare earth
compounds. The complex crystal structures of
these compounds are derived from few basic
structures, such as the cubic D21 structure or the
hexagonal D2d structure, by stacking the struc-
tural units in different ways from which the vari-
ous stoichiometries result. The crystallographic
data of most binary, ternary, and quaternary rare
earth compounds have been compiled. The rare
earth intermetallics are mostly line compounds
and have only very narrow composition ranges.
The preparation of the compounds is difficult
because of the high affinity of the rare earths for
oxygen, which leads to reactions with common
crucible materials. Various processes have been
developed for the preparation of intermetallic
rare earth compounds.
The rare earth intermetallics have won a
great practical importance because of their out-
standing magnetic properties, which depend on
composition and crystal structure. SmCo5 and
Sm2 Co17 led to the development of the rare
earth permanent magnets (REPM) which excel
by their high energy product (BH)max and their
high coercivity (→ Magnetic Materials). Opti-
mized property spectra have been achieved by
substituting other rare earth metals for Sm and
Fe and other transition metals and Cu for Co and
by appropriate processing. The resulting multi-
nary materials have complex multiphase struc-
tures, and the effects of the various structural
features on the magnetic properties have not yet
been understood completely.
Alloying with the interstitials B, C, and N led
to ternary rare earth compounds with very attrac-
tive magnetic properties, and outstanding exam-
ples are Sm2 Fe17 Cx , Sm2 Fe17 Nx , Nd2 Fe14 B,
and Nd2 Fe14 C. Nd2 Fe14 B has become the ba-
sis for the development of the Nd – Fe – B per-
manent magnets, which have an even higher
energy product (BH)max at room temperature
than the SmCo5 -type and Sm2 Co17 -type mate-
rials at lower cost; however their behavior at el-
evated temperatures is less satisfactory. The var-
ious REPM materials — SmCo5 type, Sm2 Co17
type, Nd2 Fe14 B type — are produced primarily
by powder metallurgy, though ingot metallurgy
and rapid solidification have also been used. The
materials are brittle, which makes machining

difficult, and their corrosion resistance is low,
which necessitates protective coatings in corro-
sive environments. SmCo5 suffers an environ-
mental degradation of coercivity due to hydro-
gen absorption. The hydrogenation of the Nd –
Fe – B alloys leads to structure transformation
and decomposition reactions and is used for pro-
ducing fine-grained Nd – Fe – B powders. Be-
cause of their most advantageous combination
of properties, the rare earth magnets have found
a huge market with a broad field of applications.
Research and development is rapidly progress-
ing [10, 13].
4. Superconducting Materials
4.1. Laves Phases
A transition to superconductivity at low tem-
peratures has been observed for a large num-
ber of Laves phases with C14 or C15 structure.
(Hf,Zr)V2 with C15 structure is of particular in-
terest because of its advantageous superconduct-
ing properties, since it combines a fairly high
critical temperature with a high critical current
density and critical magnetic field strength. Pro-
cessing is difficult because of its poor deforma-
bility. Further alloying, in particular with Nb,
improves the mechanical behavior to such an
extent that multifilament superconductors can
be produced which are considered for appli-
cation in fusion reactor magnets since the su-
perconducting properties are comparatively in-
sensitive to neutron irradiation and mechanical
strain. Room-temperature deformation of HfV2 -
base alloys occurs primarily by twinning. The Hf
– V – Nb phase diagram is available.
4.2. A15 Compounds
The A15 crystal structure is a complex cubic
structure derived from the bcc lattice (A2) with
eight atoms per unit cell. It is a typical example
of the topologically close packed (tcp) structures
which result from the stacking of polyhedra of
various shapes to accommodate atoms of dif-
ferent sizes. These tcp structures allow a closer
packing of atoms of different sizes than the ge-
ometrically close packed fcc (A1) and hcp (A3)
Intermetallics 11
structures. The A15 structure may be regarded
as the basis of the family of compounds with tcp
structures which are known as Frank – Kaspar
phases.
The compounds with A15 structure form a
large and comparatively homogeneous group of
intermetallic compounds and are of enormous
practical importance since many of them are su-
perconductors (→ Superconductors, Chap. 3.1).
Some of them, namely, V3 Ga, V3 Si, Nb3 Al,
Nb3 Ga, Nb3 Ge, and Nb3 Sn, exhibit critical tem-
peratures between 15 and 23 K which are only
surpassed by those of the ceramic superconduc-
tors on the basis of perovskite-type oxides. How-
ever, the latter are inferior with respect to criti-
cal current density, and the best combination of
critical temperature, critical current density, and
critical magnetic field strength is found in the
A15 superconductors, which have found wide
application [12, 14].
The processing of the A15 compounds is dif-
ficult because of their brittleness. In the follow-
ing, various A15 compounds, which have al-
ready been applied or have been considered for
application, are briefly overviewed.
4.2.1. V3 Si
V3 Si was the first A15 compound discovered to
be superconducting, with a critical temperature
of 17 K. Stoichiometric V3 Si melts congruently
at 1925 ◦ C and is stable at high and low tempera-
tures. A deviation from stoichiometry is possible
on the Si-poor side, but reduces the critical tem-
perature due to the resulting constitutional disor-
der. The critical temperature is also lowered by
dissolved hydrogen. Likewise the critical cur-
rent density and the critical field strength are re-
duced by elastic tensile or compressive strain.
Slightly above the critical temperature, V3 Si
transforms martensitically into a weakly tetrag-
onal, still superconducting compound with twin-
related thin lamellae. Prior plastic deformation
with large strains at high temperatures stabilizes
the tetragonal martensitic phase even at temper-
atures above the martensite formation tempera-
ture. The martensitic transformation gives rise
to superelasticity and a shape-memory effect.
V3 Si has a high Young’s modulus of 213 GPa
and an apparent brittle-to-ductile transition tem-
perature (BDTT) of the order of 1200 ◦ C. The
12 Intermetallics
plastic deformation behavior above the BDTT
has been studied in detail. The creep behavior is
characterized by a comparatively high activation
energy in the range of 2 – 11 eV per atom that
depends on stress and composition. Creep can
be enhanced by superimposed electric currents.
The transition from superconduction to nor-
mal conduction in A15 superconductors oc-
curs by the penetration of small fluxoids, that
is, quantized magnetic flux vortices, which is
characteristic for the type II high-field super-
conductors. These fluxoids can be pinned by
microstructural inhomogeneities, in particular
grain boundaries and precipitated particles, and
this is beneficial with respect to critical current
density and field strength. Consequently, mi-
crostructural changes by prior plastic straining
with or without recrystallization affect the super-
conducting properties. The thin lamellae, which
are produced by the above-mentioned marten-
sitic transformation, are less effective with re-
spect to flux pinning than grain boundaries, so
that their effect on the superconducting proper-
ties becomes relevant only in single crystals.
V3 Si is of interest for applications because
of its favorable superconducting properties and
its high stability. Apart from hot extrusion, V3 Si
can be deformed plastically at room temperature
in spite of its brittleness by superimposing a high
hydrostatic pressure, which makes the produc-
tion of multifilament superconducting wires fea-
sible for applications in solenoid magnets. Lam-
inar V3 Si can be prepared by solid-state thin-
film reactions between V and SiO2 on Si sub-
strates. However, Nb3 Sn has been found more
favorable with respect to superconducting prop-
erties, mechanical behavior and processing, and
this has precluded the application of V3 Si.
4.2.2. V3 Ga
V3 Ga with A15 structure is stable only up to
1300 ◦ C where it transforms congruently into a
V – Ga disordered bcc solid solution. There are
indications for a martensitic transformation at
very low temperatures, in close analogy to V3 Si,
although the conditions for martensite formation
are not yet clear for V3 Ga. It has a broad range of
homogeneity centered around the stoichiometric
composition, that is, it is very stable with respect
to deviations from stoichiometry. Because of this
stability, which is advantageous for fabrication,
and the high critical temperature of about 16 K,
V3 Ga is of interest for applications.
The superconducting properties and the de-
formation behavior of V3 Ga are similar to those
of V3 Si. The superconduction transition can be
affected by prior plastic deformation. V3 Ga is
ductile in compression above 1000 ◦ C and the
creep behavior has been studied in detail. As in
the case of V3 Si, V3 Ga can be deformed plas-
tically by hot extrusion and at room tempera-
ture by superimposing a high hydrostatic pres-
sure which makes the production of multifila-
ment superconducting wires feasible for appli-
cations in solenoid magnets. Hybrid high-field
magnets with V3 Ga tape coils have been pro-
duced successfully. However, Nb3 Sn has been
found more favorable with respect to super-
conducting properties, mechanical behavior, and
processing, and this has precluded the commer-
cialization of V3 Ga up to now.
4.2.3. Nb3 Sn
Nb3 Sn with A15 structure forms peritectically
at 2130 ◦ C from the liquid solution and solid
Nb and is stable only above 775 ◦ C where it de-
composes eutectically. The decomposition reac-
tion can be suppressed by not too slow cooling,
so that Nb3 Sn with A15 structure is maintained
at lower temperatures in a metastable condi-
tion. Nb3 Sn transforms into a twinned tetragonal
martensite at 43 K, and the transition to super-
conductivity occurs at about 18 K. This critical
temperature is lowered by deviations from stoi-
chiometry, which affect the atomic order, and by
elastic strain. Small increases in the critical tem-
perature are possible by addition of ternary al-
loying components. The transition temperature
does not depend sensitively on prior plastic de-
formation.
Like the other A15 compounds, Nb3 Sn can
be deformed readily at high temperatures, e.g.,
by hot extrusion, whereas deformation at low
temperatures is possible only by superimposing
a high hydrostatic pressure. The high tempera-
ture deformation and creep behavior has been
studied in detail, and a beneficial effect of grain
size reduction on the brittle-to-ductile transition
temperature (BDTT) has been found.

Nb3 Sn superconductors with multifilament
geometry, which are needed for electromagnet
applications, are easily fabricated by the so-
called bronze process which has lead to the
widespread commercial application of Nb3 Sn.
In the bronze process Nb rods are inserted in
a bronze rod or tube which is then processed
to obtain a multifilament wire. After wire pro-
cessing, Nb3 Sn shells are formed around the Nb
cores by reaction heat treatments. The final mul-
tifilament composite superconductor consists of
Nb3 Sn fibers with Nb cores which are embed-
ded in a bronze matrix; the latter is surrounded
by a coating of Nb, which serves as a diffusion
barrier, and an outer Cu tube, which serves as a
stabilizer and carries the current in the case of
transition to normal conduction. Various specific
processing routes, including powder metallurgy,
have been developed for optimized supercon-
ductor production, and hybrid superconductor
coils are commercially available.
The superconducting properties of the com-
posite superconductor as well as the mechanical
properties depend on the particular microstruc-
ture which is a function of the various prior pro-
cessing steps. Much work has been directed at
the details of the various processes, e.g. the ef-
fects of alloying additions, the diffusion of Sn
during Nb3 Sn formation, the grain size depen-
dence of the flux pinning force and the tensile
strength and fracture at room temperature, or
the effects of the hardnesses of the constituent
phases on the composite workability. Recent
new developments rely on the beneficial effects
of additions of Ti, Hf, Ta, and/or Ge on the su-
perconducting properties.
4.2.4. Nb3 Al
Nb3 Al with A15 structure forms peritectically
at 2060 ◦ C and is stable below this temperature
in the whole temperature range only with Al-
deficient off-stoichiometric compositions. The
stoichiometric composition corresponds to the
maximum solubility of Al in Nb3 Al and is
reached only at 1940 ◦ C in eutectic equilibrium
with the liquid and the σ-phase Nb2 Al. The tran-
sition to superconduction occurs at 19 K and
there is no martensitic transformation. The tran-
sition temperature of Nb3 Al is affected by devi-
ations from stoichiometry and perfect order like
Intermetallics 13
the other A15 compounds, but in a less sensitive
way. The A15 compound with the highest transi-
tion temperature is Nb3 Ge, which can be alloyed
with Nb3 Al to form the ternary A15 compound
Nb3 (Al0.7 Ge0.3 ) with the highest critical field
strength.
Nb3 Al has long been considered for electro-
magnet applications because of its advantageous
superconducting properties. However, the pro-
cessing of Nb3 Al is very difficult because of
its brittleness of Nb3 Al. Powder metallurgy has
been used successfully for preparing Nb3 Al by
reaction synthesis and producing, e.g., a Nb3 Al
– Ag composite superconductor. Nb3 Al tapes
have been prepared by rapid solidification. A
novel process with thin Nb – Al sandwiches as
starting material has allowed the fabrication of
Nb3 Al superconducting wires with good worka-
bility. Furthermore various other processes have
been used successfully for producing Nb3 Al
wires.
4.2.5. Nb3 Si
Nb3 Si is a high-temperature line compound
which forms with a tetragonal crystal structure
by a peritectic reaction at 1980 ◦ C and is sta-
ble only above 1770 ◦ C. Rapid solidification of
the liquid produces amorphous Nb3 Si. The A15
structure is unstable and is obtained only by spe-
cial rapid solidification techniques and by crys-
tallization of amorphous Nb3 Si. Nb3 Si with A15
structure is of interest because a high transition
temperature for superconduction is expected,
and indeed a transition temperature of 19 K was
reported for near-stoichiometric Nb3 Si with A15
structure, which was prepared by crystallization
of amorphous Nb3 Si under high pressure.
5. Electronic Materials
5.1. Silicides
Metal(M)-like M2 Si silicides, as well as other
transition metal silicides (→ Silicon, Chap. 5),
are of major importance for applications as thin-
film materials in electronic devices, for exam-
ple, as low-resistivity interconnects and gates,
Ohmic contacts, and Schottky barriers in very
large scale integrated (VLSI) circuits [16]. The
14 Intermetallics
most eminent example is Pd2 Si which is used
for shallow contacts with very small controlled
thickness and extension. The kinetics of thin-
film formation and growth are controlled by dif-
fusion and can be optimized by alloying with
further elements. The film-formation process af-
fects the distribution of dopants in the Si sub-
strate. Other M2 Si compounds have also been
studied with respect to thin-film formation, e.g.,
Ni2 Si, Co2 Si,and Mg2 Si. Thin oxide layers may
form on the silicide films by air exposure at room
temperature, which can be a problem for Ohmic
contacts or a necessity for a tunneling device.
The transition metal silicides CrSi, MnSi,
FeSi, and CoSi crystallize with the cubic B20
structure, whereas NiSi and the noble metal sili-
cides PtSi, IrSi, and PdSi crystallize with the
orthorhombic B31 structure. Such MSi com-
pounds, in particular PtSi and also NiSi, are also
important thin-film materials for applications in
electronic devices.
The disilicides CrSi2 , VSi2 , NbSi2 , and TaSi2
crystallize with the hexagonal C40 structure.
VSi2 is of interest for applications in electronic
devices. TaSi2 has been successfully used in pro-
ducing microprocessors and other electronic de-
vices. CrSi2 is of interest for applications as thin-
film interconnectors. CoSi2 with the compara-
tively simple cubic C1 structure is applied in
large-scale integrated electronic devices. NiSi2
is another compound of primary interest for ap-
plications in electronic thin-film devices. MoSi2
is advantageous for applications in electronic de-
vices either as substrate for Si films or as thin-
film interconnectors in VLSI circuits. In view of
these applications, the conditions for thin-film
formation as well as for the formation of thin
oxide layers have been studied in detail.
5.2. Other Compounds
Various aluminides, such as Al3 Ti, Al3 Hf,
Al3 Ta, and Al7 Cr, are used as shunting layers
in integrated circuits to improve the resistance
to failure of Al-based interconnects by electro-
migration stressing [17].
GaAs, InP, InSb, and InAs are compounds
with high electron mobilities and high satura-
tion velocities relative to Si and hence are very
attractive materials for application in high-speed
devices [18]. Such devices are the GaAs metal
Schottky field effect transistors (MESFETs), the
high electron mobility transistors (HEMTs), the
heterojunction bipolar transistors (HBTs), the
permeable-base transistors (PBTs), the static in-
ductance transistors (SITs), and the metal insula-
tor semiconductor field effect transistors (MIS-
FETs).
6. Electric Materials
The transition metal disilicides MSi2 (M = Ti,
Zr, Hf, V, Nb, Ta, Cr, Mo, W (→ Silicon,
Chap. 5) crystallize with the tetragonal C11b
structure, the hexagonal C40 structure, or the
orthorhombic C49 and C54 structures and show
extended mutual solid solubilities. These crys-
tal structures are closely related since they only
differ in the stacking of the same structural ele-
ments.
MoSi2 with C11b structure has found wide-
spread application as heating elements in high-
temperature furnaces at up to 1700 ◦ C because
of its advantageous electrical properties and its
excellent oxidation resistance. Well-known ex-
amples are Kanthal Super and Mosilit heaters.
The high oxidation resistance of these materi-
als, which is not shown by the other Mo-rich
silicides, is due to the formation of a protective
vitreous highly adherent SiO2 film with prior
evaporation of the volatile Mo oxide. However,
MoSi2 is subject to oxidative disintegration by
pest oxidation at intermediate temperatures bet-
ween about 300 ◦ C and 600 ◦ C, which occurs
at grain boundaries and microstructural defects
and thus can be minimized by avoiding these or
by alloying with small amounts of Fe and Re.
7. Thermoelectric and
Thermomagnetic Materials
7.1. Silicides
The dimagnesium monosilicide Mg2 Si is a semi-
conductor with a high thermoelectric power and
a low thermal conductivity which is advanta-
geous for thermoelectric power generation. The
thermal conductivity can be minimized by al-
loying with Ge to form the ternary Mg2 (Si,Ge)
compound. Doping with Ag and Sb produces p-

and n-type semiconducting Mg2 (Si,Ge), respec-
tively, and the combination of both results in a
thermocouple with a high efficiency for thermo-
electric energy conversion. The interactions bet-
ween doping, point defects, and microstructure
has been studied in detail.
FeSi2 is another disilicide with a high ther-
moelectric power and has been proposed for ap-
plication in a thermoelectric power generator. It
has an orthorhombic crystal structure and is sta-
ble only at lower temperatures, whereas at higher
temperatures it decomposes to form a two-phase
FeSi – Fe2 Si5 alloy. Furthermore, it is subject to
pest oxidation. Doping with Al produces p-type
conducting FeSi2 , doping with Co produces n-
type FeSi2 , and the combination of both materi-
als gives a thermocouple which can be used as a
thermoelectric power generator. Such a power-
generation system for the efficient conversion
of low-grade heat into electricity has been pro-
posed [12, 19].
The monosilicides MnSi and CoSi exhibit
large thermoelectric powers and thus are promis-
ing for thermoelectric power generation. In
particular, a CoSi – CrSi2 thermogenerator has
been proposed.
The complex manganese silicide Mn11 Si19
is promising for thermoelectric applications be-
cause of its anomalously high anisotropy of the
thermopower over an exceptionally wide tem-
perature range [20].
7.2. Other Thermoelectric Compounds
Thermoelectric materials on the basis of the
semiconducting compound Bi2 Te3 are widely
used as thermoelectric coolers (TECs) for opera-
tion near room temperature in small refrigerators
[20] (→ Refrigeration Technology, Chap. 1.4).
Such materials with the rhombohedral C33
structure are solid solutions of Bi2 Te3 with the
isomorphous compounds Sb2 Te3 and Bi2 Se3 .
They easily cleave along the basal planes and
show anisotropic mechanical and transport prop-
erties.
Thermoelectric materials for application at
medium temperatures are based on the com-
pound PbTe with simple cubic B1 structure,
which forms solid solutions with the isomor-
phous compounds PbSe, PbS, SnTe, and GeTe
[20]. n- and p-type materials are prepared by
Intermetallics 15
deviation from stoichiometry and doping with
halogens as donors and alkali metals as accep-
tors, respectively.
Research and development are leading to
novel complex alloys with even more attractive
properties [21, 22].
7.3. Thermomagnetic Compounds
[23 – 27]
Magnetization of a material by application of a
strong magnetic field produces magnetic order
with electron spin alignment and is accompanied
by heat release. Without the magnetic field the
material returns to magnetic disorder with heat
consumption from the surroundings. This mag-
netocaloric effect is used for magnetic refriger-
ation, which is more efficient than conventional
gas compression refrigeration. For this purpose
magnetic refrigerant materials with high mag-
netic moments are needed, since the amount of
produced or consumed heat is proportional to
the magnetic moment apart from the magnetic
field.
Intermetallic rare earth metal compounds
with transition metals show very large magnetic
moments and have successfully been used for
magnetic refrigeration. Present developments
concentrate on optimizing such rare earth metal
phases with respect to magnetic moment and
temperature of operation. An outstanding ex-
ample is the phase Gd5 Si2 Ge2 with a magne-
tocaloric effect which is twice as large as that
of Gd [25]. The operating temperature of this
material can be tuned from about −240 to about
+20 ◦ C by adjustment of the Si/Ge ratio.
The materials developments aim at magnetic
refrigerators which can be applied, e.g., as small
cryocoolers for cooling superconductive devices
to near liquid helium temperatures or as large
commercial liquefiers for liquefying cryofuels.
8. Optical Materials
Various intermetallics show beautiful colors
which may be attractive for applications in jew-
elry. An early example is the Regulus of Venus
of the alchemists with its bright violet color,
which results from the reduction of Sb2 S3 , the
16 Intermetallics
ore of antimony, through the addition of cop-
per. A more recent example is the nickel disili-
cide NiSi2 , which can be alloyed with Ni and Al
to produce compounds with distinct pale blue,
yellow, and white colors and variations thereof.
Au – Cu – Ag alloys on the basis of the inter-
metallic compounds CuAu and Cu3 Au, which
have found applications in dentistry, excel by
virtue of their decorative gold color. Likewise,
PdIn with its gold color is promising for den-
tal prosthetics and jewelry. A further example is
the bright-purple AuAl2 known as purple plague
which gave rise to the Spangold alloy [28].
Apart from color, various intermetallics show
optical properties such as reflectance, trans-
mittance, absorbance and refraction, which are
attractive for applications in optical devices
[12]. GaAs is used for optical waveguides
in integrated IR light circuits. Other applica-
tions of GaAs are optical high-power switches,
high-efficiency solar cells, light-emitting diodes
(LEDs), and semiconductor diode lasers. Vari-
ous tellurides such as HgTe, CdTe, PbTe, SnTe,
and corresponding solid solutions in which Zn
and Mn beneficially substitute for Cd have been
used for developing photodetectors for light
from the far infrared to the visible region.
Thin films of various intermetallics such as
AuIn2 , AuSn, InSb, BiTe, SnAs, CaSb, GeTe,
and others are being developed for use as record-
ing media in optical devices [12]. The local re-
flectance is changed by short laser beam pulses
with subsequent slow or rapid cooling to pro-
duce a crystalline or amorphous structure, re-
spectively. Likewise thin films of the shape-
memory alloy NiTi (see Section 12.3) can be
used as bistable optical memory elements.
Various ferromagnetic intermetallics are of
interest in magneto-optical applications as ma-
terials for erasable optical data storage [29]. This
relies on the polar Kerr effect, by which linearly
polarized light experiences a rotation of the po-
larization plane when reflected at a magnetized
material. The Kerr effect has been studied for a
large number of Mn-based, Co-based, and Fe-
based intermetallics. The highest potential for
magneto-optical storage is attributed to MnBi
and PtMnSb, besides amorphous TbFeCo-based
alloys.
9. Hydrogen-Storage Materials
[30 – 35]
9.1. B2 Compounds
FeTi with B2 structure is one of the most promis-
ing intermetallic compounds for application as
a hydrogen-storage material and related appli-
cations (see also → Hydrogen, Chap. 9.3.1.2).
Dissolved hydrogen orders to form a substruc-
ture, whereby the B2 structure becomes rhom-
bic, that is, a stable hydride is formed reversibly.
FeTi with excess Ti is more easily hydrogenized
then stoichiometric FeTi, since Ti activates the
absorption and desorption of hydrogen catalyti-
cally. Second phases, including oxides, shorten
the incubation time of hydrogenization, but oxy-
gen may poison the FeTi surface at room tem-
perature. Other B2 compounds of interest for
applications as hydrogen-storage materials are
CoTi and the ternary intermetallics (Fe,Co)Ti
and (Fe,Ni)Ti, which show more advantageous
storage properties than the binary compounds.
The FeTi-base materials have good hydrogen ca-
pacities and low costs, but the problems with re-
spect to activation, impurities, and instabilities
have prevented large-scale commercial applica-
tion.
9.2. Laves Phases
Various Laves phases can absorb large amounts
of hydrogen and are considered for applications
as hydrogen storage materials. Such AB2 com-
pounds contain a strong hydride former as the
A element and crystallize with the C14 or C15
structure. The Laves phases with the highest
sorption capacities are C15 ZrV2 with a hydro-
gen-to-metal ratio H/M = 1.8, C14 or C15 ZrCr2
with H/M = 1.3, C14 ZrMn2 and C14 or C15
TiCr2 with H/M = 1.2. LaNi2 , which was known
as a C15 compound, with H/M = 1.5 is a tetrago-
nal pseudo-Laves phase which is stabilized only
by the presence of impurities according to recent
findings.
The hydrides formed in these binary Laves
phases are too stable for easy hydrogen desorp-
tion, which is a prerequisite for hydrogen stor-
age. The hydride stability can be reduced and ad-
justed to practical hydrogen storage conditions
by deviations from stoichiometry, substitution

of the B element, primarily by Fe, Co, Ni, Cu,
Mn, Cr, substitution of the A element, primar-
ily by Ti, or any combination of these alloying
possibilities.
An example for the stoichiometry effect is
given by ZrMn2 which exhibits an increasing
hydrogen vapor pressure with increasing Mn
content and which is stable with C14 struc-
ture up to ZrMn3.8 . Another example is C14
TiMn2 , which reaches its maximum H solubil-
ity at the off-stoichiometric composition corre-
sponding to TiMn1.5 , whereas it does not ab-
sorb H at the stoichiometric composition. The
effects of substitution on electronic structure,
alloy stability, hydride stability and H storage
capacity have been studied for various sys-
tems, e.g., (Zrx Ti1−x )Mn2 , Zr(Mn1−x Fex )2+y ,
Zr(V1−x Cox )2 , in which both V and Co are
substituted for by other transition metals, and
Zr(Cr1−x Vx )2 . Zr(Cr1−x Vx )2 exhibits polytyp-
ism, that is, crystal structure variants of either the
hexagonal C14 structure or the cubic C15 struc-
ture are formed, which only differ in the stacking
of crystal structure units. The thermodynamics
of hydrogenation have been studied, and hydride
formation enthalpies determined. The optimiza-
tion of properties has led to multinary Laves
phases such as Zr – Mn – Co – V alloys, Zr –
Mn – Ni – V alloys, Zr – Mn – Ni – V – Cr al-
loys, or Zr – Ti – Mn – Fe and Zr – Ti – V – Fe
alloys.
In summary, the Laves phase materials are
advantageous because of their good hydrogen
capacities in the temperature range 0 – 100 ◦ C
and low costs.
9.3. Rare Earth Metal Compounds and
Other Phases
Various rare earth compounds can absorb large
amounts of hydrogen and are of major inter-
est for applications as hydrogen-storage mate-
rials, as is exemplified by RCo3 and RFe3 . The
most outstanding examples are the RNi5 com-
pounds, in particular LaNi5 , which may be re-
garded as a prototype hydrogen storage material
(see also Chapter 10). The stability of LaNi5
is affected by absorption – desorption cycling
due to surface degradation processes. The prop-
erty spectrum has been optimized by alloying
with further elements, as is exemplified by the
Intermetallics 17
development of La0.8 Nd0.2 Ni2.5 Co2.4 Si0.1 or
MmNi3.5 Co0.7 Al0.8 (Mm = mixed metal). The
permanent-magnet compound SmCo5 can also
absorb large quantities of hydrogen which, how-
ever, reduces its coercive force. CaNi5 , which is
isostructural with LaNi5 , shows an analogous
behavior and is considered as a low-cost candi-
date for hydrogen storage.
Research and development aim at producing
nanostructured materials, which can be obtained
by powder-metallurgical processing through hy-
drogenation [32]. Such materials promise more
advantageous storage properties because of their
much-increased internal interface area and de-
fect density, as demonstrated for Mg2 X phases
(X = Cu, Ni, Si, Ge, Sn) [37, 38]. The prevail-
ing problem is the hysteresis accompanying the
hydriding and dehydriding processes, which re-
flects irreversibility and contributes to the degra-
dation of the absorption properties during long-
term service [31].
10. Electrode Materials
10.1. Mg2 Ni Alloys
Mg2 Ni-base hydrogen-storage alloys are of ma-
jor interest for application as active materials of
metal hydride electrodes. Such materials can be
prepared by powder metallurgy, and their prop-
erties have been optimized by alloying with fur-
ther elements [39]. Both partial substitutions of
Ti or Ce for Mg and of Mn or Co for Ni improve
discharge capacity and cycle life.
10.2. Laves Phase Alloys
Various rare earth (R) Laves phases undergo
amorphization on hydrogenation. RNi2 with
C15 structure is easily transformed into amor-
phous RNi2 which is very stable and is being
considered for applications as long-life elec-
trodes in electrochemical nickel – hydrogen bat-
teries.
There are many other binary and ternary
Laves phases which can absorb large amounts
of hydrogen (see Section 9.2). The optimiza-
tion of properties has led to multinary Laves
phases such as Zr – Mn – Co – V alloys, Zr –
18 Intermetallics
Mn – Ni – V alloys, Zr – Mn – Ni – V – Cr al-
loys, or Zr – Ti – Mn – Fe and Zr – Ti – V – Fe
alloys, which have been studied in detail in view
of applications as electrodes in electrochemical
nickel – hydrogen batteries.
These developments have led to Zr(Ni,Mn)2 -
base materials with high rate capability and high
hydrogen capacity which are regarded as highly
promising for application as negative electrode
in nickel – metal hydride (NiMH) batteries for
electric vehicles [36, 40].
10.3. RNi5 Alloys
The rare earth (R) compounds RNi5 , in partic-
ular LaNi5 , of Section 9.3 are highly promis-
ing for applications as rechargeable electrodes
in nickel – metal hydride (NiMH) batteries. The
rare earths can be substituted by the cheaper
Mm, and further alloying for optimiztion has
led, for example, to Mm(Ni,Co,Mn,Al)5 , which
is intended for application in high-power batter-
ies [41]. Such batteries are presently commer-
cialized and are used in newly developed pure
electric vehicles (PEV) [33, 42 – 44].
Electrode materials are of particular interest
for fuel cells, which are the subject of intense
worldwide developments [45].
11. Coating Materials
Coatings are used in many cases to protect
high-temperature materials in corrosive environ-
ments. Various coating systems have been de-
veloped which differ in processing and proper-
ties, as exemplified for high-temperature coat-
ings for gas turbines [46, 47]. Such coatings usu-
ally contain intermetallic compounds as a major
constituent. Present-day research and develop-
ment is directed not only to high-temperature
corrosion resistance, but also to high-tempera-
ture wear resistance and applications as thermal
barriers [48 – 50].
11.1. Aluminides
An exceptionally high oxidation resistance is
shown by the B2 compound NiAl. Accordingly
NiAl has long been used as a coating material.
Indeed, one of the oldest processes for produc-
ing coatings is pack aluminizing, in which Al
diffuses from the Al-rich pack into the alloy to
be coated and forms an aluminide surface layer,
which consists of NiAl in the case of Ni-base
alloys. Such diffusion coatings are still exten-
sively used for protecting turbine blades under
various turbine conditions. The coating perfor-
mance is improved by adding alloying additions
of Cr, Si, Ta, rare earths, or precious metals by
modifying the pack aluminizing process.
Besides such diffusion coatings, overlay
coatings are used which are produced by
depositing a specifically designed corrosion-
resistant alloy on the component surface (→
High-Temperature Materials). These coating al-
loys, which are described by the general for-
mula MCrAlY (M = Fe, Ni, and/or Co), again
rely on the presence of MAl phases with B2
structure. The advantage of these coatings is the
negligibly small chemical interaction between
coating and substrate during deposition, choice
of corrosion-resistant alloy composition, and the
ability to deposit thicker coatings for extended
service lives. During service interdiffusion oc-
curs between coating and substrate and may re-
sult in Al depletion of the coating. Al-deficient
NiAl may transform martensitically because of
cooling cycles, and this promotes spalling, that
is, it contributes to the degradation of the coat-
ing.
Newly developed structural high-tempera-
ture materials may lack sufficient oxidation re-
sistance, which initiates the development of pro-
tective coating systems. For example, the tita-
nium aluminide alloys based on TiAl have insuf-
ficient oxidation resistance (see Section 16.1).
Such alloys were coated successfully through
application of a protective layer of the ternary
intermetallic compound Al22 Ti8 Cr3 with L12
structure by using the pack cementation tech-
nique [51]. Other successful coatings on TiAl
alloys relied on applying a Ti – Al – Cr alloy for
formation of the t phase together with the Laves
phase [52] or an Ni – Al alloy with NiAl forma-
tion [53] as well as aluminizing with formation
of the Al-rich titanium aluminide TiAl2 [54].

11.2. Silicides
The alloys of the refractory metals molybde-
num, niobium, and tantalum, which have ma-
jor potential for structural applications at very
high temperatures, must be protected from ox-
idation in oxidizing environments, and this is
best achieved by applying silicide coatings [55].
Niobium alloys have successfully been coated
by using a newly developed pack cementation
process with codeposition of Si, Ti, Cr, and Fe,
which relies on the formation of the quaternary
silicide Nb3 Fe3 CrSi6 with partial substitution of
Nb by Ti and of Fe by Co or Ni [56].
Similar silicide diffusion coatings were de-
veloped for Cr – Nb alloys, which rely on the
formation of NbSi2 or CrSi2 , depending on dop-
ing with Ge, which suppresses formation of
chromium disilicide [57]. Likewise a boron-
modified titanium disilicide diffusion coating
process with Ge doping by pack cementation
was developed for protecting Ti alloys and
Ti3 Al-based alloys (see Section 16.1) from oxi-
dation [58]. TiAl-based alloys were successfully
coated by codeposition of Al and Si by pack ce-
mentation [59] or simple liquid-phase siliconiz-
ing [60].
12. Shape-Memory Materials
[61 – 65]
12.1. Cu-Base Alloys
12.1.1. Cu – Zn – Al Shape Memory Alloys
The Cu – Zn – Al shape memory alloys are de-
rived from the intermetallic compound CuZn
with B2 structure by alloying with Al,
and the composition of the resulting ternary
compounds is approximately described by
Cu1 + 1.3x Zn1 − 2.3x Alx . The complete ternary
Cu – Zn – Al phase diagram is available. The
crystal structure of the Cu – Zn – Al alloy is the
disordered A2 structure at high temperatures,
which transforms into the B2 structure on cool-
ing and into the L21 structure on further cool-
ing. These structures are stable only at elevated
temperatures and are only metastable at room
temperature, that is, these alloys decompose if
annealed at too low temperatures.
Intermetallics 19
The L21 structure can be transformed marten-
sitically to form various types of martensite. The
crystallography and thermodynamic stability of
the martensite variants have been analyzed in
detail. The martensite stability is correlated with
elastic anomalies, as in the case of CuZn. It was
shown that different types of martensite can co-
exist for certain compositions. In any case, the
martensite formation temperature is a very sen-
sitive function of composition and can be varied
by prior thermal cycling. In particular, alloying
with Mn is used to depress the martensite for-
mation temperature.
A special feature of the Cu – Zn – Al alloys
is the so-called two-way memory effect which
produces a reversible shape change on cooling
and heating. This effect results from martensitic
transformation and retransformation after spe-
cific thermomechanical treatments, which are
designated as “training”.
Processing of Cu – Zn – Al alloys poses few
problems. The alloys are usually produced by in-
duction melting. Various dopants are in use for
grain refinement to improve formability. Apart
from this, powder metallurgy is used for obtain-
ing fine-grained materials. The alloys are readily
hot worked in air, and cold workability decreases
with increasing Al content.
12.1.2. Cu – Al – Ni Shape-Memory Alloys
The Cu – Al – Ni shape memory alloys are
based on the intermetallic compound Cu3 Al
with disordered A2 structure, which is usually
known as β-phase and which is stable only at
high temperatures between 567 and 1049 ◦ C
with compositions between 71 and 82 atom %
Cu. At 567 ◦ C this phase decomposes by a eutec-
toid reaction at such a sluggish rate that it can be
retained with the then metastable A2 structure by
cooling below this temperature. At about 500 ◦ C
the β-phase undergoes an ordering reaction to
form the metastable β 1 -phase with D03 struc-
ture. Both phases can be transformed marten-
sitically by quenching to form various types of
martensite, depending on Al content and quench
rate.
Cu3 Al can be alloyed with Ni without chang-
ing the transformation characteristics of the
compound. Both the binary Cu3 Al and the
ternary (Cu,Ni)3 Al show the shape-memory
20 Intermetallics
effect due to the formation of thermoelastic
martensite. The ternary compound has given
rise to the development of the Cu – Al – Ni
shape memory alloys with about 11 – 15 wt %
Al and 3 – 5 wt % Ni. The deformation modes of
the martensites include various twinning modes
and stress-induced transformations. The shape
memory effect is a two-way effect, that is, it is
completely reversible. The martensite-start tem-
perature has an upper limit of about 180 ◦ C and
can be lowered to a large degree by increasing
the Al content and to a smaller degree by in-
creasing the Ni content.
The Cu – Al – Ni alloys are advantageous be-
cause of their higher stability at higher temper-
atures compared with the Cu – Zn – Al alloys.
However, second-phase precipitation, which
embrittles the Cu – Al – Ni alloys and precludes
cold working, cannot be suppressed; hence, such
alloys can only be hot finished. Thermomechan-
ical treatments and microalloying additions, in
particular Mn, Ti, Zr, are used for improving
the mechanical behavior. An important effect is
grain refinement, which increases both fracture
strength and fracture strain according to a Hall –
Petch relationship. Indeed precipitated particles
have been found to reduce the as-cast grain size
and hinder later grain growth during annealing.
The alloying additions affect the transformation
and decomposition behavior, and this must be
taken into account in establishing the optimum
conditions for the induction of the two-way ef-
fect. Recently, the ductilizing effect of small
boron additions was discovered and studied in
detail with respect to transformation behavior
and microstructure.
Other similar Cu-base alloys also show the
shape-memory effect. Well-studied examples
are the Cu – Al – Mn shape memory alloys,
which order by a two-step reaction on rapid cool-
ing, that is, the initial disordered A2 structure is
transformed first into the intermediate B2 struc-
ture and then into the final L21 structure.
12.2. NiAl Alloys
Al-deficient NiAl with B2 structure is trans-
formed martensitically by rapid cooling. This
transformation is thermoelastic and produces the
shape-memory effect. The martensite formation
temperature increases linearly with increasing
Ni content from about − 240 ◦ C for 60 atom %
Ni to about 1000 ◦ C for 70 atom % Ni. Thus,
the shape-memory effect can be produced at
high temperatures, and this allows the devel-
opment of high-temperature shape-memory al-
loys. The martensitic transformation of Ni-rich
NiAl can be induced by external applied stresses;
it is thermoelastic, superelasticity has been ob-
served, and the martensite deforms by twinning
and detwinning. The transformation character-
istics do not depend on the processing route.
The problem of the low room-temperature duc-
tility of NiAl has been overcome by alloying
with third elements, in particular Fe, to produce
a ductile second phase with fcc structure.
12.3. NiTi Alloys
The intermetallic compound NiTi with B2 struc-
ture has been used since about 30 years as
shape-memory alloy for couplings, fasteners,
connectors, and actuators in the automotive
and aerospace industries, electronics, mechan-
ical engineering, and medical applications. NiTi
is known as Nitinol, which refers to Ni and Ti and
the US Naval Ordnance Laboratory where this
material was developed. NiTi melts congruently
at 1310 ◦ C and has an extended homogeneity
range.
NiTi transforms martensitically from the par-
ent B2 phase to a monoclinic martensite phase
with an intermediate orthorhombic R phase,
from which a reversible shape-memory effect
with the best shape-memory behavior of all
shape-memory alloys and pseudoelasticity re-
sults. The characteristics of such shape-memory
alloys have been studied in detail with respect to
crystallography, microstructure, thermodynam-
ics, kinetics, and macroscopic mechanical be-
havior, and applications have been described.
The shape-memory behavior of NiTi is im-
proved by prior thermomechanical treatments
which produce advantageous microstructures
and textures. The transformation temperature is
about 110 ◦ C for stoichiometric NiTi and is de-
creased by excess Ni. Likewise the transforma-
tion temperature is lowered by alloying with
Fe and Cr, whereas Cu decreases the trans-
formation temperature hysteresis. Recently a
widened transformation temperature hysteresis

was found for Ni – Ti – Nb alloys which is at-
tractive with respect to coupling and sealing ap-
plications. The transformation temperature is in-
creased to 500 ◦ C by the partial substitution of
Ni by Pd, which, however, affects ductility.
The elastic and plastic deformation behav-
ior, including mechanical twinning, of NiTi al-
loys has been studied intensively. Noteworthy
is the high ductility of NiTi, which allows de-
formations of up to 70 %. Furthermore the yield
strength of NiTi depends on temperature in an
anomalous way, that is, it increases with increas-
ing temperature to reach a maximum, and only
above this peak temperature does normal soft-
ening occur.
Processing of NiTi alloys is difficult and must
be done with care to obtain the advantageous
properties. Melting must be carried out in vac-
uum or under inert atmosphere because of the
reactivity of Ti, and hence vacuum-induction,
plasma-arc, and electron-beam melting are all
used commercially. Hot working is possible
without problems. Processing at low tempera-
tures is difficult because of extremely high work
hardening, and intermediate annealing treat-
ments are required for cold rolling, machining,
etc. The ductility problem of the (Ni,Pd)Ti al-
loys with high transition temperature is relieved
by alloying with B. Joining by welding, braz-
ing, or soldering is generally difficult. The ex-
cellent shape-memory behavior of the NiTi al-
loys is accompanied by an excellent corrosion
resistance, comparable to stainless steels, which
makes them the only shape-memory alloys suit-
able for implantation in the human body. Re-
cently NiTi has been used as strengthening phase
in Al matrix composites, which experience an
additional strengthening effect due to the shape-
memory effect of NiTi.
13. Medical Biomaterials [66 – 69]
Classical metallic biomaterials for medical im-
plants and other devices are stainless steels, in
particular the austenitic AISI 316L steel, Co
– Cr( – Mo) alloys including Vitallium, mani-
fold Ti alloys, i.e., commercially pure (cp) Ti,
Ti6Al4V and other Ti alloys with Zr and/or Nb
and/or Ta, cp Nb, cp Ta [70], and there are devel-
opments for using Mg alloys [71]. Such alloys
are used usually for applications such as ortho-
Intermetallics 21
pedic implants, nails, screws, plates for fixing
bones, artificial hip joints, heart valve prosthe-
ses, and stents. Problems are posed by the nec-
essary biocompatibility, which is controlled pri-
marily by corrosion resistance and physical alloy
properties. The corrosion resistance increases
with the transition from steels to Co – Cr alloys
and further to Ti alloys which, however, is par-
alleled by respective cost increases.
In view of the advantageous corrosion behav-
ior of the Ti alloys, intermetallic Ti – Al alloys
(see Section 16.1.2) have recently attracted ma-
jor interest for endoprosthetic applications, such
as hip and knee prostheses and dental implants
[72].
Besides the presently used biomaterials, the
intermetallic phase NiTi known as Nitinol is ex-
tensively used for a wide variety of applications
because of its outstanding attractive properties
[73 – 75]. In addition to being biocompatible,
this material is superelastic and shows the shape-
memory effect (see Section 12.3). The extremely
good biocompatibility is due to the formation
of a passive TiO2 layer. Since NiTi is not fer-
romagnetic, NiTi devices are compatible with
magnetic resonance imaging (MRI).
Superelastic NiTi wires are used for connect-
ing bone fragments and fixing small bone frag-
ments. Other applications are filters for large
embolized blood clots. The superelasticity gives
rise to an enormous flexibility which is made
use of for producing medical instruments, e.g.,
superelastic NiTi catheters. The most celebrated
application is the self-expanding stent, which re-
lies on knitted or welded NiTi wire. The long-
term stability of such devices is still the subject
of discussion, and in particular there is the ques-
tion whether toxicity may arise after many years
of service.
14. Dental Materials
14.1. Amalgams
Dental amalgams have been in use for more than
a thousand years. Amalgams are still a most im-
portant class of material for dental restoration in
spite of corrosion problems with mercury release
because they offer an advantageous combination
of properties with respect to processing and use.
22 Intermetallics
These amalgams are formed by alloying Ag – Sn
– Cu alloys on the basis of the intermetallic com-
pounds Ag3 Sn and Cu3 Sn with mercury. The
amalgamation reaction results in the formation
of a complex multiphase alloy structure which
contains the phases Ag2 Hg3 , Sn8 Hg, Ag3 Sn,
and Cu6 Sn5 , depending on composition.
Processing involves mixing and compaction
of powders and is comparatively simple and is
analogous to mechanical alloying. The amal-
gams must show a sufficient compressive
strength and creep resistance, since the oral tem-
perature is only about 10 % below the melting
temperature. The dental amalgams are very brit-
tle which leads to failure with tensile or bend-
ing stresses in unsupported margins. The frac-
ture processes are correlated with corrosion pro-
cesses which primarily attack the phases Sn8 Hg
and Cu6 Sn5 and lead to the release of mercury
and SnO, respectively. Attempts to replace mer-
cury in these amalgams have not yet been suc-
cessful.
The alloy systems of interest have been stud-
ied in detail with respect to phase reactions, mi-
crostructure evolution, mechanical behavior in-
cluding creep and fatigue, corrosion resistance
and clinical performance, and the processing
technologies are well established [76, 77].
14.2. Cu – Au Alloys
Au – Cu – Ag alloys on the basis of the in-
termetallic compounds CuAu and Cu3 Au have
found applications in dentistry because of their
extremely high corrosion resistance, their ad-
vantageous mechanical properties such as high
strength and ductility, and their decorative gold
color. These alloys age-harden as a result of
complex ordering and decomposition reactions
by which the phases Cu3 Au I, CuAu I, CuAu II,
and an Ag-rich α2 phase are formed, depending
on composition.
Cu3 Al I is the well-known Cu3 Al compound
with L12 structure, which is the prototype phase
for this crystal structure. Cu3 Au II forms at in-
termediate temperatures with off-stoichiometric
compositions and is described as either tetrag-
onal or orthorhombic. In any case, the Cu3 Au
II structure results from the L12 structure by
the introduction of equally spaced antiphase
boundaries (APBs). Such a structure is usu-
ally known as long-period superlattice (LPS). At
the equiatomic composition the tetragonal L10
phase CuAu I forms, and again an orthorhombic
LPS is observed at intermediate temperatures,
which is known as CuAu II. All phases are sep-
arated from each other by two-phase equilibria.
The elastic behavior is known, as is the dif-
fusion behavior. The deformation behavior of
Cu3 Au was studied in detail to clarify the ef-
fects of atomic order, and it was found in par-
ticular that order decreases the yield stress and
increases the work hardening rate while still re-
taining high ductility. Inversely, heavy cold de-
formation destroys the order. The ordering reac-
tion produces antiphase domains which give rise
to an additional strengthening effect in this and
similar phases; strength increases with decreas-
ing domain size. The domain size can be varied
by prior processing, and it increases during age-
ing in analogy to Ostwald ripening. The Portevin
– Le Chatelier effect with serrated yielding was
observed for both the ordered and disordered
state. Recovery and recrystallization have been
analyzed in detail.
15. Wear-Resistant Materials
Intermetallic compounds are attractive in tribo-
logical applications because of various advanta-
geous properties, such as hardness, thermal sta-
bility, and oxidation resistance, and because of
their compatibility with metallic matrices [78].
Some of the oldest bearing alloys contain in-
termetallic compounds as is exemplified by the
lead-based bearing alloys (babbits) with SbSn
and possibly Cu6 Sn5 or the bronze bearing al-
loys with Cu32 Sn8 or Cu3 Al.
The ternary Laves phase MoCoSi, that is,
Mo(Co,Si)2 , with hexagonal C14 structure has
given rise to the development of wear-resistant
Co – Mo – Cr – Si alloys which contain large
volume fractions of the Laves phase in coarse
distribution together with a Co-rich phase and
are known as Tribaloys. Apart from their high
strength and hardness, they show a sufficient
fracture toughness of the order of 20 MN/m3/2 .
These advantageous mechanical properties re-
sult in an excellent wear resistance which, in
combination with an excellent corrosion resis-
tance in various environments, makes the Trib-
aloys very appealing for applications in a broad

range of temperatures and environments. These
alloys have been studied in detail including
corrosive wear testing, according to which the
Tribaloys compare favorably with conventional
wear-resistant materials [78]. The low ductility
of Tribaloys, which is due to the large amounts
of Laves phase, can be increased by proper ad-
justment of alloy composition and heat treatment
[79, 80].
Apart from the Tribaloys, modified struc-
tural intermetallic high-temperature alloys (see
Chap. 16.1) have been proposed for applica-
tion as wear-resistant alloys. An example is the
novel Nicralc alloys which result from alloying
Ni3Al with C to increase the hardness and which
are intended as substitutes for high-tempera-
ture abrasion- and erosion-resistant cobalt al-
loys [81]. Other examples are superelastic NiTi
(see Section 12.3) with hardening TiC disper-
soids [82] and MoSi2 – SiC [83] and Fe3 Al –
Fe3 AlC0.5 composites [84].
16. Structural High-Temperature
Materials
The aluminides of titanium, iron, and nickel,
as well as some silicides, have been selected
as bases for developments of new structural
high-temperature alloys because of their high
strengths at low and high temperatures, high
thermal stability, high oxidation resistance,
and/or low density. The problem of insufficient
ductility was overcome by further alloying and
optimization of microstructure, which usually
resulted in multiphase alloys [85]. Research and
development is in progress, and some develop-
ments are on the brink of commercialization.
The basic properties of the intermetallic com-
pounds in question and the respective phase di-
agrams, as well as the physical, mechanical and
chemical properties of the respective developed
or proposed intermetallic alloys, with emphasis
on powder metallurgy processing, are summa-
rized in [4].
16.1. Titanium Aluminide Alloys
16.1.1. Ti3 Al Alloys
The titanium aluminide Ti3 Al, often designated
as the α2 phase, shows the hexagonal D019 struc-
Intermetallics 23
ture and a density of 4.2 g/cm3 , which make it
most attractive for materials developments (→
Titanium, Titanium Alloys, and Titanium Com-
pounds, Chap. 11.2).
The aim of the various materials develop-
ments on the basis of Ti3 Al has been to im-
prove both strength and ductility by alloying
with further elements and by controlling the mi-
crostructure. The most effective element for im-
proving ductility is Nb, and indeed Ti3 Al-base
alloys with engineering significance, that is, α2
alloys and super-α2 alloys, contain 10 – 30 atom
% Nb. Ti3 Al-base alloys readily dissolve oxy-
gen and hydrogen, which both produce embrit-
tlement.
The developments of the Ti3 Al-based alloys
are on the brink of commercialization. They have
been produced in ingots of up to 4500 kg and
are available as cast shapes or wrought forms
including billets, bars, plates, and sheet. Like-
wise the powder metallurgy methods which have
been developed for conventional titanium alloys
are readily adapted and in addition other powder
metallurgy methods, in particular reactive pro-
cessing, have been applied to Ti3 Al-base alloys.
The alloys can be machined, sheet can be de-
formed superplastically and diffusion bonded to
form complex sheet parts, and the various manu-
facturing processes can be accomplished on con-
ventional equipment. The alloys can be joined
successfully by both diffusion bonding and lin-
ear friction welding.
16.1.2. TiAl Alloys
The titanium aluminide TiAl, often designated
the γ-phase, crystallizes with the tetragonal L10
structure. The density of 3.76 g/cm3 is even
lower than that of Ti3 Al and has made the ti-
tanium aluminides most attractive for materials
developments (→ Titanium, Titanium Alloys,
and Titanium Compounds, Chap. 11.2). Single-
crystalline TiAl shows an anomalous positive
temperature dependence of the yield stress, that
is, the yield stress increases with increasing tem-
perature until a maximum is reached at about
600 ◦ C, and only above the peak temperature is
a normal negative temperature dependence ob-
served.
24 Intermetallics
The major problem for the application of TiAl
is low ductility at room temperature. If the Al
content is reduced below 50 atom % to form
Ti3 Al as second phase besides TiAl, the resulting
intermetallic two-phase alloys show ductilities
of a few percent which are acceptable for appli-
cations. 48 atom % Al is the optimum compo-
sition for maximum room-temperature ductility
of TiAl alloys. The aim of the various materials
developments on the basis of TiAl is to improve
both strength and ductility first by controlling the
microstructure and second by alloying with fur-
ther elements. Oxygen embrittles TiAl alloys,
as do other interstitial impurities (N, C, B). In
addition, environmental embrittlement has been
found for TiAl, as for other intermetallics, which
is clearly caused by the presence of atmospheric
moisture.
The development of TiAl alloys is less ad-
vanced than that of the Ti3 Al-based materials
because of the more severe ductility problems.
However, the TiAl alloys are regarded as more
promising, primarily because of the lower den-
sity and the higher application temperatures, and
development activities are now concentrated on
TiAl in various countries. Processing is more
difficult compared with Ti3 Al alloys because of
the higher brittle-to-ductile transition tempera-
ture, and hot working temperatures are outside
the range of conventional titanium processing
equipment, which has limited the practical in-
got size to about 200 kg. Besides casting and hot
working, powder metallurgy has been used for
component fabrication, though there are prob-
lems of quality, availability, and cost of powder.
Cast TiAl alloys can be forged isothermally,
fine-grain powder metallurgy material can be
formed superplastically, as well as special cast
and thermomechanically treated material, sheet
of 1 – 2 mm thickness can be produced by twin-
roll casting, special processes have been de-
veloped for rolling, and machining is possi-
ble, as is joining by diffusion bonding and
welding. With respect to turbine-engine appli-
cations, component-fabrication techniques have
been demonstrated for airfoils and compressor
cases. Besides aerospace applications, interest is
focused on automotive engine applications such
as valves [86, 87] and turbocharger rotors [88,
89], which are nearer to realization.
More recent materials developments concen-
trated on Ti – Al alloys with reduced Al contents
and significantly increased Nb contents of about
5 – 10 atom % for relieving the ductility prob-
lems [90].
16.2. Iron Aluminide Alloys
16.2.1. Fe3 Al Alloys
Fe3 Al is formed on cooling by ordering reac-
tions in the solid state that transform the bcc
disordered solid solution, which is stable above
about 800 ◦ C, first into the FeAl phase with B2
structure, which is stable between about 800 ◦ C
and 550 ◦ C, and only then into Fe3 Al with the
D03 structure. The critical temperature of order-
ing is much lower than the melting temperature,
and this indicates weaker bonding between un-
like atoms than in, for example, the nickel alu-
minides. The critical temperature of D03 order-
ing can be shifted to higher temperatures by as
much as 250 ◦ C by alloying with third elements,
in particular Cr, Mo, Mn, Ti, and Si. Alloying
with Ti is of particular interest in view of the
high solubility of Ti in Fe3 Al with beneficial
effects on strength and ductility [91]. Thermal
vacancies are formed readily because of a low
formation enthalpy, which is even lower than
the migration enthalpy, and consequently there
is a high concentration of vacancies at thermal
equilibrium. High concentrations of vacancies
formed at high temperatures are trapped even
by slow cooling and contribute to strengthening.
Strength and creep resistance at high tempera-
tures can be increased significantly by proper
alloying with precipitation of second phases, as
has been exemplified by Fe3 Al with precipitated
perovskite-type carbide and/or Laves phase par-
ticles [92, 93].
Fe3 Al is considered for structural applica-
tions since it not only shows a higher strength
than comparable iron alloys, but also a high
corrosion resistance in oxidizing and sulfidiz-
ing environments. Problems for application are
the low ductility at ambient temperatures and
the strength drop above the ordering tempera-
ture. The ductility of Fe3 Al is very sensitive to
the environment, and environmental embrittle-
ment can occur in the presence of moisture in the
ambient atmosphere. Alloys have been produced
by both powder metallurgy and ingot metallurgy,

including arc melting, vacuum-arc remelting, in-
duction melting in air and vacuum, and elec-
troslag remelting. These developments, aimed
at applications such as automotive exhaust sys-
tems and resistance heating elements, are close
to commercialization. Other potential applica-
tions are hot-gas filters in coal gasification plants
and chemical processing equipment.
16.2.2. FeAl Alloys
FeAl is closely related to NiAl (Section 16.3.2)
since both phases show the B2 structure and
a complete mutual miscibility. However, FeAl
melts incongruently in contrast to NiAl, and its
melting temperature is lower, which indicates
a lower stability and atomic bonding strength.
The tensile ductility at low temperatures of poly-
crystalline FeAl is practically zero at the stoi-
chiometric composition and increases with de-
creasing Al content, that is, increasing deviation
from stoichiometry. The reported ductilities of
off-stoichiometric FeAl are obtained only with
well-annealed specimens, that is, if the low-tem-
perature deformation is preceded by annealing
at very low cooling rates. This behavior is sup-
posed to be due to retained excess vacancies,
which increase hardness and yield strength and
decrease ductility and which are not equilibrated
after high-temperature annealing with normal
cooling rates. Anneals of several hundred hours
at about 400 ◦ C are recommended for produc-
ing ductility in off-stoichiometric FeAl. FeAl is
subject to environmental embrittlement, which
is due to the dissolution of hydrogen in FeAl at
the crack tip and is reduced by increasing the
deformation rate sufficiently.
In view of its reported advantageous prop-
erties and comparatively low density, FeAl is
regarded as a potential structural material for
high-temperature applications, and processing
developments make use of both ingot and pow-
der metallurgy. FeAl alloy compositions have
been identified, e.g., for good weldability and
for high corrosion and wear resistance in aggres-
sive gasification environments. As in the case of
Fe3 Al, the high-temperature strength and creep
resistance of FeAl can be improved significantly
by proper alloying with precipitation of harden-
ing second phases [94, 95].
Intermetallics 25
16.3. Nickel Aluminide Alloys
16.3.1. Ni3 Al alloys
Ni3 Al is the most studied and best known in-
termetallic compound because it has long been
used as strengthening phase in the superalloys
and because its ductility problems were over-
come by ductilization by microalloying with
boron. Hence, it can be produced and tested
without major experimental difficulties, and thus
it was preferred in the past for studying the be-
havior of intermetallics. Ni3 Al, known as the
γ  -phase, crystallizes with the cubic L12 struc-
ture.
The outstanding feature of the plastic flow
of Ni3 Al is the anomalous temperature depen-
dence of the flow stress, which increases bet-
ween room temperature and about 700 ◦ C with
increasing temperature to reach a maximum, and
only at higher temperatures does normal soften-
ing occur. Polycrystalline Ni3 Al is brittle and
fails by intergranular fracture, whereas Ni3 Al
single crystals are highly ductile. The polycrys-
tal brittleness of Ni3 Al is removed by microal-
loying Al-deficient Ni3 Al with B.
Exposure to air or oxygen affects the mechan-
ical behavior of Ni3 Al, and various effects of en-
vironmental embrittlement have been observed.
At about 800 ◦ C the ductility of ductilized Ni3 Al
shows a minimum which is more pronounced in
air than in vacuum and which is significantly
alleviated by the addition of Cr. Environmental
embrittlement has also been observed at room
temperature, especially in moist air, depending
on Ni3 Al composition and dopant content, and it
was concluded that this embrittlement is caused
by hydrogen formation. Accordingly, stress cor-
rosion cracking has been observed in aqueous
environments.
Successful materials developments on the ba-
sis of Ni3 Al have resulted in a series of closely
related Ni3 Al alloys for high-temperature ap-
plications. These alloys, known as nickel alu-
minides or advanced aluminides, generally con-
tain B at levels below 500 ppm and Al below
the stoichiometric composition to give ductility
at room temperature, as well as up to 5 atom
% Hf, Zr, Ta, and Mo for improving strength
at high temperatures and up to 10 atom % Cr for
26 Intermetallics
enhancing ductility at intermediate temperatures
between 400 ◦ C and 900 ◦ C.
The Ni3 Al alloys compare favorably with
many superalloys with respect to mechanical
properties. Thus, Ni3 Al alloys enlarge the su-
peralloy spectrum. However, they cannot com-
pete with the advanced aerospace superalloys
used in aircraft engines. Nevertheless, Ni3 Al al-
loys are promising for less demanding applica-
tions where good combinations of strength and
resistance against fatigue, wear, including ero-
sion and cavitation, and oxidation are needed.
Ni3 Al alloys can be melted by air-induction
melting with acceptable quality and by vacuum-
induction melting, vacuum-arc remelting, and
electroslag remelting for better qualities. Ingots
up to 1500 kg and components have been cast
with low or no porosity, depending on bound-
ary conditions, hot working is possible by con-
ventional hot forging, isothermal forging, and
hot extrusion, the alloys can be cold worked,
and property data are available. Most of the al-
loys can be welded with good quality, provided
welding is performed with care and cutting is
accomplished by high-speed abrasive wheels. In
addition, powder-metallurgy methods have been
applied successfully. The transfer process from
the laboratory to large-scale production is hand-
icapped by the limited deformability at elevated
temperatures, which makes hot working diffi-
cult. The feasibility of commercial large-scale
manufacturing has successfully been demon-
strated by producing rolls of a Ni3 Al alloy for a
steel hardening furnace by conventional foundry
practice and automated welding [96].
16.3.2. NiAl Alloys
NiAl is the best known example of the inter-
metallic compounds with the cubic B2 structure
which form one of the largest group of inter-
metallics. NiAl has an extended range of homo-
geneity and melts congruently at about 1640 ◦ C
for the stoichiometric composition with 50 atom
% Al. This melting point is higher than those
of the constituent elements, and this indicates
strong bonding between Ni and Al and corre-
spondingly a high phase stability with a strong
tendency for atomic ordering.
NiAl shows high strength at low tempera-
tures, a comparatively steep strength decrease
at intermediate temperatures around half the
melting temperature of about 680 ◦ C for NiAl,
and low strength with further softening at high
temperatures. The softening at high temper-
atures is related to thermally activated creep
processes. Deviations from stoichiometry pro-
duce constitutional defects which enhance diffu-
sion and lead to softening at high temperatures,
whereas at low temperatures these defects act as
strengthening deformation obstacles. Polycrys-
talline NiAl is usually brittle with practically no
ductility.
Unlike most other aluminides, NiAl with B2
structure exhibits excellent oxidation resistance
since a protective Al2 O3 scale is readily formed
during oxidation. NiAl seems to be the only truly
oxidation resistant intermetallic besides some
silicides. The physical reason for this high ox-
idation resistance is that the Al content is suf-
ficiently high and Al diffusion is sufficiently
fast in NiAl at all temperatures for stable Al2 O3
scales to form at the surface and to avoid internal
oxidation in the bulk. Environmental embrittle-
ment, which affects the mechanical behavior of
various other intermetallics (e.g., the other B2
aluminide FeAl) has not been found for NiAl.
NiAl is advantageous for high-temperature
applications because of its low density, good
thermal conductivity, high melting point, and ex-
cellent oxidation resistance, whereas its limited
deformability at room temperature and its low
strength and creep resistance at temperatures
above 1000 ◦ C are disadvantages compared to
the superalloys. The low strength at high tem-
peratures is related to the excellent oxidation
resistance, since both effects result from easy
high-temperature diffusion. NiAl is regarded as
highly promising for developing alloys which
are competitive with superalloys, because the
strength deficit is compensated by advantages in
density and thermal conductivity, and respective
materials developments are under way.
As conventional disordered alloys, NiAl al-
loys can be strengthened appreciably by second
phases. Apart from the effects on strength, sec-
ond phases may also be beneficial for ductil-
ity and toughness. An example for the latter is
the successful development of an NiAl – Fe – Cr
alloy with sufficient ductility at room tempera-
ture and strength at low and high temperatures
for applications in industrial furnaces, coal-

conversion systems, and petrochemical plants
[97].
For the strengthening of NiAl second phases
with not too low Al contents are preferred
with respect to density and oxidation resistance.
There are hard and brittle ternary Laves phases
with Al, such as NbNiAl and TaNiAl, which may
be used for strengthening NiAl. A respective de-
velopment has lead to a NiAl – Ta – Cr alloy
which is not deformable at room temperature
but excels in high-temperature strength, oxida-
tion resistance, and shock resistance [98]. The
alloy can be produced by both ingot metallurgy
and powder metallurgy, so that components can
be produced by investment casting, hot extru-
sion, hot isostatic pressing (HIP), powder injec-
tion molding, and isothermal forging.
16.4. Silicide Alloys
Silicides (→ Silicon, Chap. 5) were considered
for high-temperature applications as early as 50
years ago because of their high strength and oxi-
dation resistance at high temperatures. Presently
the need for new structural materials for the
highest temperatures has revived interest in sili-
cides for structural applications, and respective
developments are under way.
Cr3 Si with the topologically close packed
cubic A15 structure is a candidate phase for
high-temperature applications. The brittle Cr3 Si
is alloyed with Cr to produce a ductile-phase-
toughened intermetallic composite [99]. Ni3 Si
with the ordered fcc L12 structure has a very
high potential for structural applications because
of its advantageous mechanical properties and
its outstanding corrosion resistance. Alloying of
Ni3 Si with Nb and Ti results in multiphase Ni3 Si
alloys with dispersed Ni3 Nb and Ni3 Ti precip-
itates which are most promising for high-tem-
perature applications because of their favorable
mechanical and chemical properties [100].
Mg2 Si crystallizes with the fcc C1 structure,
and its density is only 1.88 g/cm3 . Mg2 Si has a
comparatively high strength and a low thermal
expansion coefficient. However, its brittleness,
with a brittle-to-ductile transition temperature
of 450 ◦ C, precludes the application of single-
phase Mg2 Si. Alloying with Al produces Al-
Mg2 Si alloys with brittle Mg2 Si particles in a
Intermetallics 27
ductile Al matrix, which has given rise to a suc-
cessful alloy development.
The M5 Si3 compounds (M = transition
metal), which adopt a variety of complex crystal
structures, are usually very stable congruently
melting line compounds with melting tempera-
tures above 2000 ◦ C. The most attractive M5 Si3
compound for structural high-temperature ap-
plication is Ti5 Si3 with the hexagonal D88 struc-
ture because of its high stability with a melting
temperature of 2130 ◦ C, its high strength and
hardness, and its low density. A promising ma-
terials development is based on the combination
of Ti3 Al and Ti5 Si3 which has been further
alloyed with Nb. Another candidate phase for
high-temperature applications is Nb5 Si3 , which
is even more stable with a melting temperature of
2484 ◦ C and which is alloyed with Nb to relieve
its brittleness. Such Nb – Nb5 Si3 composites
are alloyed with further elements to achieve an
excellent balance of room-temperature tough-
ness, high-temperature creep performance, and
oxidation resistance [101, 102].
Presently, MoSi2 is considered for structural
high-temperature applications, and various ma-
terials developments are in progress. The brittle-
to-ductile transition occurs at about 1000 ◦ C or
higher temperatures depending on microstruc-
ture and impurity content. The ductility and
toughness of MoSi2 can be improved by spe-
cial processing routes, microalloying with C
for avoiding SiO2 layers on grain boundaries
or insertion of a ductile second phase, usu-
ally an Nb-rich phase. Toughness can also be
increased by embedding hard second phases,
in particular ceramics, in a MoSi2 matrix to
hinder crack growth. MoSi2 -based compos-
ites reinforced by TiC, TiCN, and TiB2 in-
deed showed improved hardness and bending
strength compared to monolithic MoSi2 [103].
Most work has concentrated on the MoSi2 –
SiC system. Other systems of particular interest
are MoSi2 – TiC and MoSi2 – Al2 O3 , as well
as the eutectics MoSi2 – Mo5 Si3 and MoSi2
– Er2 Mo3 Si4 and the ductile-phase-toughened
MoSi2 – Nb/Ta and MoSi2 – Mo systems. More
recent developments concentrate on intermetal-
lic Mo – Si – B alloys consisting primarily of
the phases Mo3 Si and Mo5 SiB2 besides ductile-
phase-toughening Mo [104 – 106]. Such com-
posites show attractive oxidation resistances,
fracture toughness, and creep strength, the bal-
28 Intermetallics
ance of which depends sensitively on compo-
sition adjustment and processing. Furthermore,
MoSi2 has been considered as reinforcing phase
in other composite systems, e.g. in an SiC matrix
or in a beryllide matrix. TiSi2 with orthorhom-
bic C54 structure has also been considered for
structural applications at high temperatures be-
cause of its high oxidation resistance and low
density.
17. References
1. G. E. R. Schulze: Metallphysik,
Akademie-Verlag, Berlin 1967.
2. G. Sauthoff: Intermetallics, VCH
Verlagsgesellschaft, Weinheim 1995.
3. N. S. Stoloff, V. K. Sikka (eds.): Physical
Metallurgy and Processing of Intermetallic
Compounds, Chapman & Hall, London 1996.
4. G. Sauthoff: “Intermetallic Materials”, in P.
Beiss, R. Ruthardt, H. Warlimont (eds.):
Landolt-Börnstein (New Series) Group VIII
Subvolume VIII/2A - Powder Metallurgy Data,
Part 2, Refractory Hard and Intermetallic
Materials, Springer, Berlin 2001, pp. 221–267.
5. J. H. Westbrook, R. L. Fleischer (eds.):
Intermetallic Compounds, 4 Volume Set, John
Wiley & Sons, Chichester 2000.
6. J. H. Westbrook, R. L. Fleischer (eds.):
Intermetallic Compounds, Principles and
Practice, Vol. 3, Progress, John Wiley & Sons,
Chichester 2002.
7. K. J. Hemker et al. (eds.): Structural
Intermetallics 2001 (Proc. ISSI-3), TMS,
Warrendale 2001.
8. J. H. Schneibel, K. J. Hemker, R. D. Noebe, S.
Hanada, G. Sauthoff (eds.): High-Temperature
Ordered Intermetallic Alloys IX, MRS,
Warrendale 2001.
9. P. Villars: Pearsons’s Handbook Desk Edition
— Crystallographic Data for Intermetallic
Phases, ASM International, Materials
Park/OH 1997.
10. K. H. J. Buschow: Permanent Magnet
Materials and their Applications, TTP
Transtech Publications, Ütikon-Zürich 1998.
11. H. Harada et al.: “Magnetic Materials,” in P.
Beiss, R. Ruthardt, H. Warlimont (eds.):
Landolt-Börnstein (New Series) Group VIII:
Advanced Materials and Technologies,
Subvolume VIII/2A: Powder Metallurgy Data,
Part 1: Metals and Magnets, Springer-Verlag,
Berlin 2003, pp. 9.1–9.83.
12. J. H. Westbrook, R. L. Fleischer (eds.):
Intermetallic Compounds, Vol. 4, Magnetic,
Electrical and Optical Properties and
Applications of Intermetallic Compounds,
John Wiley & Sons, Chichester 2000.
13. R. Harris: “The Attractions of Rare-Earth
Magnets,” Materials World 7 (1999) 478 –
481.
14. K. Tachikawa, A. Nagata, K. Togano:
“Advances in Superconducting Materials,” J.
Jpn. Inst. Metal. 66 (2002) 199–200.
15. K. S. Kumar: “Silicides: Science, Technology
and Applications”, in J. H. Westbrook, R. L.
Fleischer (eds.): Intermetallic Compounds:
Principles and Practice, Vol. 2, Wiley,
Chichester 1995, pp. 211 – 236
16. S. P. Murarka: “Silicide Thin Films and Their
Applications in Microelectronics,”
Intermetallics 3 (1995) 173–186.
17. J. H. Westbrook: “Miscellaneous
Applications”, in J. H. Westbrook, R. L.
Fleischer (eds.): Intermetallic Compounds:
Principles and Practice, Vol. 2, Wiley,
Chichester 1995, pp. 645 – 656.
18. K. Masumoto, A. Katsui, T. Matsuoka:
“Semiconductor Applications”, in J. H.
Westbrook, R. L. Fleischer (eds.): Intermetallic
Compounds: Principles and Practice, Vol. 2,
Wiley, Chichester 1995, pp. 323 – 350.
19. K. Ono, R. O. Suzuki: “Thermoelectric-Power
Generation — Converting Low-Grade Heat
into Electricity,” JOM J. Min. Met. Mat. Soc.
50 (1998) 49 – 51.
20. M. V. Vedernikov: “Thermoelectric and
Electrical Applications”, in J. H. Westbrook,
R. L. Fleischer (eds.): Intermetallic
Compounds: Principles and Practice, Vol. 2,
Wiley, Chichester 1995, pp. 453 – 472.
21. G. Nicolas, J. Sharp, J. Goldsmid:
Thermoelectrics, Springer, Berlin 2001.
22. G. Chen, M. S. Dresselhaus, G. Dresselhaus, J.
P. Fleurial, T. Caillat: “Recent Developments
in Thermoelectric Materials,” Int. Mater. Rev.
48 (2003) 45–66.
23. J. A. Barclay: “Active and Passive Magnetic
Regenerators in Gas Magnetic Refrigerators,”
J. Alloys Compounds 207 (1994) 355–361.
24. E. Bruck: “Developments in Magnetocaloric
Refrigeration,” J. Phy. D Appl. Phys. 38
(2005) R381–R391.
25. K. A. Gschneidner, V. K. Pecharsky, A. O.
Tsokol: “Recent Developments in
Magnetocaloric Materials,” Rep. Progr. Phys.
68 (2005) 1479–1539

26. M. Ilyn et al.: “Magnetocaloric Properties of
the LaFe11.7Si1.3 and LaFe11.2Co0.7Si1.1
Systems,” J. Magn. Magn. Mater. 290 (2005)
712–714.
27. D. Vasylyev, J. Prokleska, J. Sebek, V.
Sechovsky: “Thermal Properties of
Er(Co1-xSix)(2) Compounds,” J. Alloys
Compounds 394 (2005) 96–100.
28. I. M. Wolff, M. B. Cortie: “The Development
of Spangold,” Gold Bull. 27 (1994) 44–54.
29. W. A. McGahan: “Magneto-Optical
Applications”, in J. H. Westbrook, R. L.
Fleischer (eds.): Intermetallic Compounds:
Principles and Practice, Vol. 2, Wiley,
Chichester 1995, pp. 435 – 452.
30. V. A. Yartys, I. R. Harris, V. V. Panasyuk:
“New Metal-Hydrides — A Survey,” Mater.
Sci. (Engl. Transl.) 37 (2001) 219–240.
31. P. Dantzer: “Properties of Intermetallic
Compounds Suitable for Hydrogen Storage
Applications,” Mater. Sci. Eng. A Struct.
Mater. 329 (2002) 313–320.
32. M. Okada, T. Kuriiwa, A. Kamegawa, H.
Takamura: “Role of Intermetallics in
Hydrogen Storage Materials,” Mater. Sci. Eng.
A Struct. Mater. 329 (2002) 305–312.
33. R. A. Varin, T. Czujko: “Overview of
Processing of Nanocrystalline Hydrogen
Storage Intermetallics by Mechanical
Alloying/Milling,” Materials and
Manufacturing Process 17 (2002) 129–156.
34. A. Zuttel: “Materials for Hydrogen Storage,”
Mater. Today 6 (2003) 24–33.
35. M. Conte, P. P. Prosini, S. Passerini:
“Overview of Energy/Hydrogen Storage:
State-of-the-Art of the Technologies and
Prospects for Nanomaterials,” Mater. Sci. Eng.
B 108 (2004) 2–8.
36. D. M. Kim, K. J. Jang, J. Y. Lee: “A Review
on the Development of AB(2)-Type Zr-Based
Laves Phase Hydrogen Storage Alloys for
Ni-MH Rechargeable Batteries in the Korea
Advanced Institute of Science and
Technology,” J. Alloys Compd. 293 – 295
(1999) 583–592.
37. H. Y. Shao, Y. T. Wang, H. R. Xu, X. G. Li:
“Preparation and Hydrogen Storage Properties
of Nanostructured Mg2Cu Alloy,” J. Solid
State Chem. 178 (2005) 2211–2217.
38. R. Janot, F. Cuevas, M. Latroche, A.
Percheron-Guegan: “Influence of Crystallinity
on the Structural and Hydrogenation
Properties of Mg2X Phases (X=Ni, Si, Ge,
Sn),” Intermetallics 14 (2006) 163–169.
Intermetallics 29
39. J. Chen, P. Yao, D. H. Bradhurst, S. X. Dou, H.
K. Liu: “Mg2 Ni-based Hydrogen Storage
Alloys for Metal Hydride Electrodes,” J.
Alloys Compd. 293 – 295 (1999) 675 – 679.
40. B. Knosp, L. Vallet, P. Blanchard:
“Performance of an AB2 Alloy in Sealed
Ni-MH Batteries for Electric Vehicles —
Quantification of Corrosion Rate and
Consequences on the Battery Performance,” J.
Alloys Compd. 293–295 (1999) 770 – 774.
41. A. Otto, V. Güther: “Development of Fast
Kinetics Metal Hydride Alloys and Battery
Electrodes for High Power Applications,” J.
Alloys Compd. 293 – 295 (1999) 734 – 736.
42. T. Sakai, I. Uehara, H. Ishikawa:
“Research-and-Development on Metal
Hydride Materials and Ni-MH Batteries in
Japan,” J. Alloys Compd. 293 – 295 (1999)
762 – 769.
43. F. Zhan, L. J. Jiang, B. R. Wu, Z. H. Xia, X. Y.
Wei, G. R. Qin: “Characteristics of Ni/MH
Power Batteries and its Application to Electric
Vehicles,” J. Alloys Compd. 293 – 295 (1999)
804 – 808.
44. J. M. Joubert, M. Latroche, A.
Percheron-Guegan: “Metallic Hydrides II:
Materials for Electrochemical Storage,” MRS
Bull. 27 (2002) 694–698.
45. G. Marsh: “Fuel Cell Materials,” Mater. Today
4 (2001) 20–24.
46. P. K. Datta, J. S. Burnell-Gray, K. Natesan:
“Coating Technology,” in J. H Westbrook, R.
L. Fleischer (eds. ): Intermetallic Compounds
— Principles and Practice, Vol. 3 Progress,
John Wiley & Sons, Chichester 2002, pp.
561–588.
47. J. R. Nicholls, R. Wing: “Advances in Coating
Systems for Utility Gas Turbines,” in J.
Lecomte-Beckers, M. Carton, F. Schubert, P. J.
Ennis (eds.): Materials for Advanced Power
Engineering 2002 (Proc. 7th Liège
Conference), Forschungszentrum Jülich,
Jülich, 2002, pp. I.57–I.71.
48. A. Matthews, J. H. Givens, W. L. Ensinger, S.
L. Rohde (eds.): Metallurgical Coatings and
Thin Films 1999 (Proc. 26th Int. Conf.
Metallurgical Coatings and Thin Films),
North-Holland Elsevier Science, Amsterdam
2000.
49. J. R. Nicholls: “Designing Oxidation-Resistant
Coatings,” JOM J. Min. Met. Mat. 52 (2000)
28–35.
50. A. G. Evans, D. R. Mumm, J. W. Hutchinson,
G. H. Meier, F. S. Pettit: “Mechanisms
Controlling the Durability of Thermal Barrier
30 Intermetallics
Coatings,” Prog. Mater. Sci. 46 (2001)
505–553.
51. H. Mabuchi, H. Tsuda, T. Kawakami, S.
Nakamatsu, T. Matsui, K. Morii:
“Oxidation-Resistant Coating for
Gamma-Titanium Aluminides by Pack
Cementation,” Scr. Mater. 41 (1999) 511 –
516.
52. G. S. Fox-Rabinovich, D. S. Wilkinson, S. C.
Veldhuis, G. K. Dosbaeva, G. C. Weatherly:
“Oxidation Resistant Ti-Al-Cr Alloy for
Protective Coating Applications,”
Intermetallics 14 (2006) 189–197.
53. T. Izumi, T. Nishimoto, T. Narita: “Superior
Long-Term Oxidation Resistance of Ni-Al
Coated TiAl Alloys,” Intermetallics 13 (2005)
727–732.
54. Z. Liu, G. Wang: “Improvement of Oxidation
Resistance of [gamma]-TiAl At 800 and 900
[deg]C in Air by TiAl2 Coatings,” Mate. Sci.
Eng. A 397 (2005) 50–57.
55. C. M. Packer: “Overview of Silicide Coatings
for Refractory Metals”, in T. Grobstein, J.
Doychak (eds.): Oxidation of
High-Temperature Intermetallics, TMS,
Warrendale 1989, pp. 235 – 244.
56. M. Vilasi, M. Francois, R. Podor, J. Steinmetz:
“New Silicides for New Niobium Protective
Coatings,” J. Alloys Compd. 264 (1998) 244 –
251.
57. Y. R. He, R. A. Rapp, P. F. Tortorelli:
“Oxidation-Resistant Ge-Doped Silicide
Coating on Cr-Cr2 Nb Alloys by Pack
Cementation,” Mater. Sci. Eng. A 222 (1997)
109 – 117.
58. B. Cockeram, R. A. Rapp: “Development and
Growth of Boron Modified and Germanium
Doped Titanium Silicide Diffusion Coatings
by the Halide Activated, Pack Cementation
Method,” Oxid. Met. 45 (1996) 375–425.
59. Z. D. Xiang, S. R. Rose, P. K. Datta:
“Codeposition of Al and Si to Form
Oxidation-Resistant Coatings on
[gamma]-TiAl by the Pack Cementation
Process,” Mater. Chem. Phys. 80 (2003)
482–489.
60. H. P. Xiong, W. Mao, Y. H. Xie, Y. Y. Cheng,
X. H. Li: “Formation of Silicide Coatings on
the Surface of a TiAl-Based Alloy and
Improvement in Oxidation Resistance,” Mater.
Sci. Eng. A 391 (2005) 10–18.
61. J. Vanhumbeeck: “High-Temperature
Shape-Memory Alloys,” J. Eng. Mater.
Technol. Trans. ASME 121 (1999) 98 – 101.
62. J. Vanhumbeeck: “Nonmedical Applications
of Shape-Memory Alloys,” Mater. Sci. Eng. A
Struct. Mater. 275 (1999) 134–148.
63. R. D. James, K. F. Hane: “Martensitic
Transformations and Shape-Memory
Materials,” Acta Mater. 48 (2000) 197–222.
64. J. Vanhumbeeck: “Shape-Memory Alloys —
A Material and a Technology,” Adv. Eng.
Mater. 3 (2001) 837–850.
65. D. S. Grummon: “Thin-Film Shape-Memory
Materials for High-Temperature Applications,”
JOM J. Min. Met. Mat. 55 (2003) 24–32.
66. J. A. Helsen, H. J. Breme (eds.): Metals as
Biomaterials, John Wiley & Sons Ltd,
Chichester, New York 1998.
67. A. Fischer: “Biomaterials — Fundamentals
and Clinical-Applications,” Materialwiss.
Werkstofftech. 32 (2001) 109.
68. J. B. Park, J. D. Bronzino (eds.): Biomaterials:
Principles and Applications, CRC Press, Boca
Raton, FL, 2002.
69. B. D. Ratner, A. Hoffman, F. Schoen, J.
Lemons (eds.): Biomaterials Science: An
Introduction to Materials in Medicine,
Academic Press, New York 2004.
70. K. S. Katti: “Biomaterials in Total Joint
Replacement,” Colloids Surf. B: Biointerf. 39
(2004) 133–142.
71. M. P. Staiger, A. M. Pietak, J. Huadmai, G.
Dias: “Magnesium and its Alloys as
Orthopedic Biomaterials: A Review,”
Biomaterials 27 (2006) 1728–1734.
72. M. L. Escudero, M. A. Munoz-Morris, M. C.
Garcia-Alonso, E. Fernandez-Escalante: “In
Vitro Evaluation of a [gamma]-TiAl
Intermetallic for Potential Endoprothesic
Applications,” Intermetallics 12 (2004)
253–260.
73. T. Duerig, A. Pelton, D. Stockel: “An
Overview of Nitinol Medical Applications,”
Mater. Sci. Eng. A Struct. Mater. 275 (1999)
149–160.
74. F. Elfeninat, G. Laroche, M. Fiset, D.
Mantovani: “Shape-Memory Materials for
Biomedical Applications,” Adv. Eng. Mater. 4
(2002) 91–104.
75. A. R. Pelton, S. M. Russell, J. DiCello: “The
Physical Metallurgy of Nitinol for Medical
Applications,” JOM J. Min. Met. Mat. 55
(2003) 33–37.
76. R. M. Waterstrat, T. Okabe: “Dental
Amalgams”, in J. H. Westbrook, R. L.
Fleischer (eds.): Intermetallic Compounds:
Principles and Practice, Vol. 2, Wiley,
Chichester 1995, pp. 575 – 590.

77. C. Guminski: “Amalgams,” in J. H.
Westbrook, R. L. Fleischer (eds.):
Intermetallic Compounds — Principles and
Practice, Vol. 3 Progress, John Wiley & Sons,
Chichester 2002, pp. 21–35.
78. W. A. Glaeser: “Intermetallics in Tribology”,
in J. H. Westbrook, R. L. Fleischer (eds.):
Intermetallic Compounds: Principles and
Practice, Vol. 2, Wiley, Chichester 1995, pp.
591 – 602.
79. R. Liu, W. Xu, M. X. Yao, P. C. Patnaik, X. J.
Wu: “A Newly Developed Tribaloy Alloy with
Increased Ductility,” Scripta Mater. 53 (2005)
1351–1355.
80. R. Liu, M. X. Yao, P. C. Patnaik, X. J. Wu:
“Effects of Heat Treatment on Mechanical and
Tribological Properties of Cobalt-Base
Tribaloy Alloys,” J. Mater. Eng. Perform. 14
(2005) 634–640.
81. H. Goldenstein, Y. Nunes Silva, H. N.
Yoshimura: “Designing a New Family of High
Temperature Wear Resistant Alloys Based on
Ni3Al IC: Experimental Results and
Thermodynamic Modelling,” Intermetallics 12
(2004) 963–968.
82. D. Y. Li: “Development of Novel Tribo
Composites with TiNi Shape Memory Alloy
Matrix,” Wear 255 (2003) 617–628.
83. P. La, Q. Xue, W. Liu: “Study of Wear
Resistant MoSi2-SiC Composites Fabricated
by Self-Propagating High Temperature
Synthesis Casting,” Intermetallics 11 (2003)
541–550.
84. J. Yang, P. La, W. Liu, J. Ma, Q. Xue:
“Tribological Properties of Fe3Al-Fe3AlC0.5
Composites under Dry Sliding,” Intermetallics
13 (2005) 1184–1189.
85. G. Sauthoff: “Multiphase Intermetallic Alloys
for Structural Applications,” Intermetallics 8
(2000) 1101 – 1109.
86. M. Blum, G. Jarczyk, H. Scholz, S. Pleier, P.
Busse, H. J. Laudenberg, K. Segtrop, R.
Simon: “Prototype Plant for the Economical
Mass Production of TiAl-Valves,” Mater. Sci.
Eng. A Struct. Mater. 329 (2002) 616–620.
87. K. Gebauer: “Performance, Tolerance and
Cost of TiAl Passenger Car Valves,”
Intermetallics 14 (2006) 355–360.
88. M. Nazmy, V. Lupinc: “Gamma TiAl
Intermetallics for Turbomachinery
Applications,” in J. Lecomte-Beckers, M.
Carton, F. Schubert, P. J. Ennis (eds.):
Materials for Advanced Power Engineering
2002 (Proc. 7th Liège Conference),
Intermetallics 31
Forschungszentrum Jülich, Jülich, 2002, pp.
I.43–I.56.
89. T. Tetsui: “Development of a TiAl
Turbocharger for Passenger Vehicles,” Mater.
Sci. Eng. A Struct. Mater. 329 (2002) 582–588.
90. F. Appel: “Recent Developments in the Design
and Processing of Gamma-Based Titanium
Aluminide Alloys,” in 4th International
Conference on Processing and Manufacturing
of Advanced Materials, JUL 07-11, 2003
(THERMEC’2003), PTS 1-5, 2003, pp. 91–98.
91. M. Palm, G. Sauthoff: “Deformation
Behaviour and Oxidation Resistance of
Single-Phase and Two-Phase L21 Fe-Al-Ti
Alloys,” Intermetallics 12 (2004) 1345–1359.
92. D. G. Morris, M. A. Munoz-Morris, C.
Baudin: “The High-Temperature Strength of
Some Fe3Al Alloys,” Acta Mater. 52 (2004)
2827–2836.
93. M. Palm, A. Schneider, F. Stein, G. Sauthoff:
“Strengthening of Iron Aluminide Alloys for
High-Temperature Applications,” in M. J.
Mills, H. Inui, H. Clemens, C. L. Fu (eds.):
Integrative and Interdisciplinary Aspects of
Intermetallics, MRS, Warrendale, 2005, pp.
3–14.
94. D. G. Morris, M. A. Munoz-Morris, J. Chao:
“Development of High Strength, High
Ductility and High Creep Resistant Iron
Aluminide,” Intermetallics 12 (2004)
821–826.
95. D. D. Risanti, G. Sauthoff:
“Iron-Aluminium-Base Alloys with
Strengthening Laves Phase for Structural
Applications at High Temperatures,” Mater.
Sci. Forum 475–479 (2005) 865–868.
96. V. K. Sikka et al.: “Large-Scale Manufacturing
of Nickel Aluminide Transfer Rolls for Steel
Austenitizing Furnaces,” Intermetallics 12
(2004) 837–844.
97. D. Letzig, J. Klöwer, G. Sauthoff: “Screening
of NiAl-Base Ni-Fe-Al Alloys for Structural
High Temperature Applications and
Development of a New Ni-30Fe-10Al-Cr
Alloy,” Z. Metallkd. 90 (1999) 712 – 721.
98. M. Palm, G. Sauthoff: “Characterization and
Processing of an Advanced Intermetallic
NiAl-Base Intermetallic Alloy for
High-Temperature Applications,” in K. J.
Hemker et al. (eds. ): Structural Intermetallics
2001 (Proc. ISSI-3), TMS, Warrendale, 2001,
pp. 149–156.
99. H. Bei, G. M. Pharr, E. P. George: “A Review
of Directionally Solidified Intermetallic
Composites for High-Temperature Structural
32 Intermetallics
Applications,” J. Mater. Sci. 39 (2004)
3975–3984.
100. K. Ohira, Y. Kaneno, T. Takasugi:
“Microstructure, Mechanical Property and
Oxidation Property in Ni3Si-Ni3Ti-Ni3Nb
Multi-phase Intermetallic Alloys,” Mater. Sci.
Eng. A 399 (2005) 332–343.
101. B. P. Bewlay, M. R. Jackson, J. C. Zhao, P. R.
Subramanian: “A Review of
Very-High-Temperature Nb-Silicide-Based
Composites,” Metall. Mater. Trans. A 34
(2003) 2043–2052.
102. C. L. Ma, J. G. Li, Y. Tan, R. Tanaka, S.
Hanada: “Microstructure and Mechanical
Properties of Nb/Nb5Si3 In Situ Composites
in Nb-Mo-Si and Nb-W-Si Systems,” Mater.
Sci. Eng. A 386 (2004) 375–383.
103. J. Meng, J. Lu, J. Wang, S. Yang: “Preparation
and Properties of MoSi2 Composites
Reinforced by TiC, TiCN, and TiB2,” Mater.
Sci. Eng. A 396 (2005) 277–284.
104. S. Nakatani, K. Yoshimi, S. Hanada, S. H. Ko,
Y. H. Park: “Design of Mo-Si-B-O Pseudo
in-situ Composites for High Temperature
Applications,” in PRICM 4: Fourth Pacific
Rim International Conference on Advanced
Materials and Processing, Vols I and II, 2001,
pp. 971–974.
105. J. H. Schneibel et al.: “Optimization of
Mo-Si-B intermetallics,” in E. P. George, H.
Inui, M. J. Mills, G. Eggeler (eds. ): Defect
Properties and Related Phenomena in
Intermetallic, MRS, Warrendale 2003, pp.
53–58.
106. V. Behrani, A. J. Thom, M. J. Kramer, M.
Akinc: “Microstructure and Oxidation
Behavior of Nb-Mo-Si-B Alloys,”
Intermetallics 14 (2006) 24–32.
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Inulin Save title to My
Standard Article Profile Topics

Anne Franck1 and Bart Levecke1 Email this page

1ORAFTI, Tienen, Belgium
Print this page SEARCH THIS TITLE
Copyright © 2003 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights
reserved.
DOI: 10.1002/14356007.g14_g02 Advanced Product Search
Article Online Posting Date: July 31, 2003 Search All Content
Acronym Finder
Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Introduction
2. Chemical Structure
3. Physical and Chemical Properties
4. Physiological Properties
5. Resources and Raw Materials
5.1. Dahlia
5.2. Jerusalem Artichoke
5.3. Chicory
6. Production
6.1. First Step
6.2. Second Step
6.3. Environmental Aspects
7. Uses
7.1. Food Applications
7.2. Inulin Derivatives for Non-Food Applications
7.2.1. Oxidation
7.2.2. Carboxymethylation
7.2.3. Hydrophobization
8. Analysis
8.1. HPLC
8.2. Gas Chromatography
8.3. HPAEC Analysis (Dionex)
8.4. Permethylation
8.5. Quantitative Determination
9. Legal Aspects
10. Economic Aspects and Trade Names

[Top of Page]

1. Introduction
Inulin is a storage carbohydrate found in many plants. It has formed part of the daily diet of humans for some centuries as
it naturally occurs in many vegetables, fruit, and cereals. Inulin is industrially obtained from chicory roots and used as a
food ingredient offering both interesting nutritional properties and important technological benefits. It significantly improves
the organoleptic characteristics of food and drinks, increases the stability of foams and emulsions, and shows a fat-like
behavior when used as a gel in water.

Inulin was first isolated from Inula helenium by ROSE, a German scientist, in 1804 [1] but it was THOMSON who called
this substance inulin [2]. One pioneer in fructan research was the German plant physiologist JULIUS SACHS [3] who
detected inulin in the tubers of Dahlia, Jerusalem artichoke (Helianthus tuberosus), and Inula helenium after ethanol

page 1 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
precipitation. Already in the beginning of the 20th century, feeding diabetic patients with inulin was reported to be beneficial
[4]. Especially since the 1990s, a spectacular increase in the number of publications dealing with the technological and
nutritional benefits of inulin has been observed [5].

[Top of Page]

2. Chemical Structure
Inulin [9005-80-5] is a polydisperse carbohydrate material consisting mainly, if not exclusively, of fructose [57-48-7] units,
linked by (2 1) bonds [6]. A starting glucose [50-99-7] moiety can be present in the chain but is not necessary. Levan,
formed by certain bacteria, consists mainly or exclusively of (2 6) linked fructosyl-fructose units. As for inulin, glucose
can be present, but again is not necessary. Fructan is a more general name, used for any compound in which fructosyl-
fructose bonds constitute the majority of linkages. The term fructan therefore covers both inulin and levan.

Referring to the definition of inulin, both GFn and Fn compounds (F being a fructosyl unit and G a glucosyl unit) are
included under that same nomenclature. In chicory inulin, n, the number of fructose units, can vary from 2 to 60 [7], i.e.,
inulin is a mixture of oligomers and polymers. Its molecular structure is shown in Figure 1.

Figure 1. Chemical structure of inulin

The degree of polymerization (DP) and the possible presence of branches are important properties because they influence
the functionality of inulin. Therefore a strict distinction must be made between inulin from plant origin and inulin from
bacterial origin. The DP of plant inulin is rather low (DP < 200) and varies according to the plant species, weather
conditions and the physiological age of the plant. Standard chicory inulin has an average DP of about 10–12. Long-chain
chicory inulin, from which the lower DP fraction has been physically removed and which has an average DP of about 25, is
also available. In contrast, bacterial inulin has a very high DP, ranging from 10000 to over 100000 (106 to 107 dalton). It is
branched (≥15 %) which results in a completely different behavior in aqueous systems.

[Top of Page]

3. Physical and Chemical Properties

Chicory inulin is available as a white, odorless powder with a high purity and a defined chemical composition. The taste is
neutral, without any off-flavor or aftertaste. Standard chicory inulin is slightly sweet (10 % sweetness in comparison with
sucrose), whereas long-chain inulin is not sweet at all [5] (see also Physiological Properties). The physicochemical
properties of chicory inulin are summarized in Table 1.

Table 1. Physicochemical properties of chicory inulin

Standard inulin Long-chain inulin

Chemical structure GFn (2 ≤ n ≤ 60) GFn (10 ≤ n ≤ 60)

Average degree of polymerization 12 25
Dry matter (%) > 95 > 95
Inulin content (% of d.m.) 92 99.5
Sugar content (% of d.m.) 8 0.5
pH (10 wt %) 5-7 5-7
Sulfated ash (% of d.m.) < 0.2 < 0.2
Heavy metals (ppm of d.m.) < 0.2 < 0.2
Appearance white powder white powder
Taste neutral neutral
Sweetness (vs sucrose = 100 %) 10 % none
Solubility in water at 25 °C (g/L) 120 10
Viscosity in water (5 %) at 10 °C
(mPa·s) 1.6 2.4
Heat stability good good
Acid stability fair good
Functionality in foods fat replacement

page 2 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
body and mouthfeel
texture improvement
foam stabilization
emulsion stabilization
synergy with gelling agents

*G = glucosyl unit. F = fructosyl unit. d.m. = dry matter.

Physical Properties. Chicory inulin is moderately soluble in water (about 10 % at room temperature) which allows its
incorporation in watery systems without precipitation. The viscosity of chicory inulin solutions is rather low, e.g.,
at 10 °C for a 5 % dry matter (d.m.) solution and for a 30 % d.m. solution [5]. Dissolved inulin exerts a small
effect on the freezing and boiling point of water (e.g., 15 % chicory inulin decreases the freezing point by 0.5 °C).

At high concentration (>25 % in water for standard chicory inulin and >15 % for long-chain inulin), inulin has gelling
properties and forms a particle gel network after shearing.

The gel strength obtained depends on the inulin concentration and total dry substance content, on the shearing
parameters (such as temperature, time, speed, or pressure) and also on the type of shearing device used, but is not
influenced by pH (between pH 4 and 9). Electron cryomicroscopy has shown that such an inulin gel is composed of a
three-dimensional network of insoluble submicron inulin particles in water. These particles of about 100 nm in size
aggregate to form larger clusters having a diameter of 1 to 5 µm. Large amounts of water are immobilized in that network,
as determined by NMR experiments. X-ray diffraction confirmed the crystalline nature of the gel particles whereas the
starting inulin powder is essentially amorphous.

An inulin gel exhibits a viscoelastic rheological behavior and shows shear-thinning and thixotropic properties. The gel is
characterized by a relatively low yield stress (e.g., 1540 Pa for a gel of 30 % standard inulin in water at 25 °C). Inulin also
displays synergistic effects with most gelling agents (e.g., gelatin, alginate, - and -carrageenan, gellan gum,
maltodextrin) [10].

Furthermore, inulin improves the stability of foams and emulsions, e.g., in aerated dairy desserts, ice creams, table
spreads, and sauces. It can therefore replace other stabilizers in different food products.

Chemical Properties. Under strongly acidic conditions, the (2 1) bonds between the fructose units can be (partially)
hydrolyzed and fructose is formed. Inulin is stable in applications with a pH > 4. Even at lower pH values, the hydrolysis of
inulin is limited to less than 10 % if the products either have a high dry substance content (>70 %), are stored at a low
temperature (<10 °C), or have a short shelf-life.

[Top of Page]

4. Physiological Properties
Inulin behaves like a bulk ingredient, contributing to body and mouthfeel. It is often used in combination with intense
sweeteners to provide a rounder mouthfeel and a better-sustained flavor with reduced aftertaste.

Inulin provides several interesting nutritional benefits to animals and humans (see Table 2) [11]. Thanks to its (2 1)
bonds, which the digestive enzymes of humans cannot hydrolyze, inulin passes through the mouth, the stomach, and the
small intestine without undergoing any significant change and without being absorbed [12]. Inulin thus enters the colon
almost quantitatively and there it is completely metabolized by the intestinal bacteria [13], mainly to short-chain fatty acids
(SCFA), bacterial biomass and gases. Only the SCFA contribute to the host's energy metabolism, which explains the
reduced caloric value of inulin (between 1 and 1.5 kcal/g). Inulin has no influence on blood glucose and insulin levels when
ingested orally and it has been known as a food for diabetics since the beginning of the 20th century [4], [14].

Table 2. Physiological properties of inulin

Strong evidence:
Nondigestibility and low caloric value (1–1.5 kcal/g)
Suitable for diabetics
Soluble dietary fiber
Stool bulking effect: increase in stool weight and stool frequency, relief of constipation
Modulation of the gut flora composition, stimulating beneficial bacteria (Bifidobacteria) and
repressing harmful ones (Clostridia): prebiotic / bifidogenic effect
Improvement of calcium (and magnesium) bioavailability
Promising evidence:
Reduction of serum and liver triglycerides and serum insulin levels
Reduction of colon cancer risk (in animal models)

page 3 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Modulation of immune response and resistance
Protection against intestinal disorders and infections
Improved well-being

Inulin is a soluble dietary fiber [15] which induces typical effects on the gut function, such as a reduction of the intestinal
pH, a relief of constipation, and an increase in stool weight and frequency (or bulking effect). Its fecal bulking effect is
similar to that of other soluble fibers such as pectin and guar gum [16]: each gram ingested increases the fecal wet weight
by about 2 g. Inulin also has modulating effects on the lipid metabolism, e.g., by reducing serum and liver triglycerides [19],
[20].

In the colon, inulin selectively promotes the growth and metabolic activity of beneficial bacteria, mainly Bifidobacteria, while
repressing harmful ones, e.g., Clostridia (prebiotic or bifidogenic effect, Health Value Added Foods – Prebiotics ) [17],
[21], [22]. A daily intake of 5 to 10 g of inulin is sufficient to significantly enhance Bifidobacteria in humans [23].

Recent research indicates that inulin also has a significant chemopreventive potential [24]. Indeed, it can prevent the
formation of precancerous lesions and tumors in the colon of rats. Long-chain inulin is more effective than shorter-chain
fructans. Synergistic effects were observed when inulin and Bifidobacteria were administered together as a synbiotic [25].
It was furthermore demonstrated that inulin inhibits the development of cancer cells transplanted in the thigh or in the
peritoneum of mice [26].

Inulin also increases the intestinal absorption of calcium, iron, and magnesium as well as the bone mineral density in rats
[27], [28]. Studies in human volunteers also show an increased calcium absorption upon administration of inulin [29], [30],
[31].

[Top of Page]

5. Resources and Raw Materials

After starch, fructans are the most abundant nonstructural polysaccharides found in nature. They are present in a wide
variety of plants and in some bacteria. Most literature on the natural occurrence of fructans does not differentiate, however,
between levan and inulin.

Fructan-producing plants are commonly present among the grasses (1200 species), whereas 15 % of the flowering plants
produce fructans in significant amounts. Fructan-producing plants are widely spread within the Liliaceae (3500 species)
and most frequently among the Compositae (25 000 species) [32]. A plant with a special status is the Agave Azul Tequila
Weber (Liliaceae): Tequila is an alcoholic drink made by fermentation from highly branched fructan present in this plant.

Inulin-containing plants that are commonly used in the human diet belong mainly to either the Liliaceae (leek, onion, garlic,
and asparagus) or the Compositae (Jerusalem artichoke, chicory, and yacon) as shown in Table 3 [33].

Table 3. Inulin content (% of fresh weight) of plants that are commonly used in human nutrition

Source Edible parts Dry solids content Inulin content

Artichoke leave-heart 14–16 3–10

Banana fruit 24–26 0.3–0.7
Barley cereal NA 0.5–1.5*
Burdock root 21–25 3.5–4.0
Camas bulb 31–50 12–22
Chicory root 20–25 15–20
Dandelion leaves 50–55* 12–15
Garlic bulb 40–45* 9–16
Jerusalem artichoke tuber 19–25 14–19
Leek bulb 15–20* 3–10
Murnong root 25–28 8–13
Onion bulb 6–12 2–6
Rye cereal 88–90 0.5–1*
Salsify root 20–22 4–11
Yacon root 13–31 3–19

NA: figures not available.

*Estimated.

Bacterial inulin is produced by Streptococcus mutans, the spores of Aspergillus sydowi and Lactobacillus reuteri. A high

page 4 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
degree of branching (≥15 %) is observed in these fructans.

Given their high inulin content (>10 %), dahlia, Jerusalem artichoke, and chicory (Cichorium intybus) have been
considered as candidates for industrial production of inulin [34].

5.1. Dahlia
Many dahlia cultivars are available, but they have all been selected for their flowers, rather than for inulin production. The
tuberous roots can be propagated only if attached to a piece of stem tissue. When propagated from seed, sowing has to
be delayed until late spring, given dahlia's extreme sensitivity to frost. Mechanical harvesting of the tubers is feasible only
on sandy grounds. Although the DP of dahlia inulin is higher than that of chicory inulin, inulin yield from dahlia is only half
that of chicory (see Table 4). For all these reasons, dahlia does not appear as an interesting crop for inulin production.

Table 4. Yields and composition of Dahlia, Jerusalem artichoke and Chicory

Dahlia Jerusalem artichoke Chicory

Roots or tubers (t/ha)

average 25 42 43.5

variation 35–60 25–75

d.m. (%)
average 18 22 22.5

variation 15–22 19–25 20–25

Inulin (%)
average 11 16 16.5

variation 10–12 14–18 15–18

Inulin (t/ha) 2.5–3 4.5–8.5 5–11
Mean DP 13–20 6–10 10–14

d.m. = dry matter. DP = degree of polymerization.

5.2. Jerusalem Artichoke

Jerusalem artichoke shows a rather high inulin content (14–19 %). It is not recommended to cultivate Jerusalem artichoke
in clay-soil as the tubers are small and irregular, hence bringing a lot of soil attached to them. Although Jerusalem
artichoke is frost tolerant, its inulin metabolism is very sensitive to cold. As Jerusalem artichoke inulin has only 20 % of
chains with a degree of polymerization longer than 10, it is unsuited for several food and non-food applications (see Table
4).

5.3. Chicory
Chicory is a biennial plant. During the first season, the plants remain in the vegetative phase and make only leaves, tap
roots, and fibrous roots. The roots look like small oblong sugar beets. The inulin content is high (16–18 %) and fairly
constant from year to year for a given region. Yields in tons of roots per hectare (around 45 t/ha) show a wider variation
(Table 4). In order to keep the chicory culture healthy a strict crop rotation is imposed (once every five years). Industrially,
inulin is produced from chicory. The chicory used for inulin production (Cichorium intybus) is of the same type as the one
used for the production of a coffee substitute (after roasting).

[Top of Page]

6. Production
The production of chicory inulin consists of two steps. The first step includes the extraction and primary purification,
resulting in a raw impure syrup. The second step is the refining phase, which results in a purified commercial end product
[5], [35].

6.1. First Step

The first process phase is virtually similar to sugar beet processing. Roots are harvested and stored in piles on the field.
To minimize losses, no more than 7 days should elapse between harvest and delivery to the factory. The roots are then
transported to the factory by truck, weighed, and carefully stored on the court. From there, they are transported with a
stream of water inside the factory, where they are washed and sliced. Raw inulin is extracted from the resulting “chips”
with hot water in a countercurrent diffuser. The leached “chips” are dried and sold as animal feed. A first purification step is
applied to the extraction juice by liming and carbonation at high pH. The resulting CaCO3 sludge precipitates easily.
Peptides, some anions, degraded proteins, and colloids are trapped in its flocs. This generates a foam-type product, which
is used by the farmers to improve the soil structure, as it is rich in calcium and organic matter.

6.2. Second Step

The raw juice is further refined using cationic and anionic ion exchange resins for demineralization and active carbon for
decolorization. This technology is comparable to the one used in starch processing. After demineralization and

page 5 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
decolorization, the juice is passed over a 0.2 µm filter for sterilization, before being evaporated and spray-dried. The
resulting standard inulin (RAFTILINE ST, ORAFTI) has an average chain length which reflects the original DP distribution
present in the chicory root (between 10 and 12). A special long-chain grade inulin (RAFTILINE HP, ORAFTI), with an
average DP >23 is also available. It is made by physical elimination of the small DP fraction [36].

6.3. Environmental Aspects

The chemicals used for regeneration of the ion exchangers are not NaOH and HCl as usual, but NH3 and H2SO4. The
advantage is that the effluents can be converted into reusable byproducts. At high concentrations, easily crystallizing salts
such as (NH4)2SO4 and K2SO4 precipitate, are separated from the mother liquor and sold commercially as fertilizers. The
mother liquor is further evaporated into a stable product that is sold as animal feed because of its high organic matter
content. This way the circle between the processing industry and the agriculture is fully closed.

[Top of Page]

7. Uses
7.1. Food Applications
In food and drink products, inulin can be used for either its nutritional advantages or its technological properties, but it is
often applied to offer a double benefit: an improved organoleptic quality and a better-balanced nutritional composition.
Table 5 gives an overview of the applications of inulin in foods and drinks.

The use of inulin as a dietary fiber ingredient is easy and often leads to an improved taste and texture [37]. Its solubility
allows fiber incorporation in watery systems such as drinks, dairy products, and table spreads. Inulin is more and more
used in “functional foods”, especially in a whole range of dairy products, as a prebiotic ingredient ( Health Value Added
Foods – Prebiotics ).

Table 5. Applications of inulin in foods and drinks

Application Functionality Dosage level (wt %)

Dairy products fat replacement 2–10

(yogurts, cheeses, desserts, drinks) body and mouthfeel
foam stabilization
fiber and prebiotic
Frozen desserts fat replacement 2–10
texture improvement
melting behavior
low caloric value
Table spreads and butter-like products fat replacement 2–10
texture and spreadability
emulsion stabilization
replacement of gelatine
fiber and prebiotic
Baked goods and breads fiber and prebiotic 2–15
moisture retention
Breakfast cereals and extruded snacks fiber and prebiotic 2–20
crispness and expansion
low caloric value
Fillings fat replacement 2–30
texture improvement
Salad-dressings and sauces fat replacement 2–10
mouthfeel and body
emulsion stabilization
Meat products fat replacement 2–10
texture and stability
fiber
Dietetic products and meal replacers fat replacement 2–15
synergy with intense sweeteners
body and mouthfeel
fiber and low calorie
Chocolate sugar replacement and 5–30

page 6 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
fiber and heat resistance
Tablets sugar replacement 5–75
fiber and prebiotic

Thanks to its specific gelling characteristics, inulin allows the development of low-fat foods without compromising on taste
and texture. When the fructan is thoroughly mixed with water or another aqueous liquid, using a shearing device such as a
rotor-stator mixer or a homogenizer, a white creamy structure results which can easily be incorporated in foods to replace
fat up to 100 % [5], a process patented by RAFFINERIE TIRLEMONTOISE/ORAFTI [9]. In table spreads, both fat and
water-continuous, inulin allows the replacement of significant amounts of fat and the stabilization of the emulsion, while
providing a short spreadable texture. In low-fat dairy products, such as fresh cheese, cream cheese or processed cheese,
the addition of a few percents of inulin gives a creamier mouthfeel and imparts a better-balanced round flavor. Inulin also is
destined to be used as a fat replacer in frozen desserts, providing an easy processing, a fatty mouthfeel, excellent melting
properties, as well as freeze-thaw stability, without any unwanted off-flavor. Fat-reduced meat products, such as sausages
and pâtés with a creamier and juicier mouthfeel and an improved stability due to better water immobilization, can be
obtained using inulin as ingredient. The synergistic effect of inulin with other gelling agents constitutes an additional
advantage in all these applications.

Special “instant” qualities, which do not require shearing to give stable homogeneous gels also have been developed
(using specific spray-drying technology) and patented.

The incorporation of inulin (1 to 3 %) in diet fruit yogurts, possibly through the fruit preparation, improves the mouthfeel,
and offers a synergistic taste effect in combination with aspartame and/or acesulfame K [8]. Inulin furthermore increases
the stability of foams and mousses: its incorporation at 1 to 5 % into dairy-based aerated desserts improves the
processibility and upgrades the quality.

Inulin also has found an interesting application as a low calorie bulk ingredient in chocolate without added sugar, often in
combination with a polyol. It is also used as a dietary fiber or sugar replacer in tablets.

7.2. Inulin Derivatives for Non-Food Applications

In the context of product development based on renewable resources, different types of chemical modifications have been
performed using inulin as raw material [38]. From the industrial point of view, three reaction types are most interesting:
oxidation, carboxymethylation, and hydrophobization.

7.2.1. Oxidation
Although inulin can be oxidized in several ways, a one-step reaction using sodium hypochlorite as the oxidant together
with sodium bromide as a catalyst is generally used (see Fig. 2) [38-41].

Figure 2. Oxidation of inulin

Under well-controlled conditions (slow addition of the oxidant in 4 to 6 h, constant pH of 9.5, addition of a concentrated
NaOH solution) oxidation occurs almost exclusively at the C3-C4 fructose bond, resulting in a ring-opened inulin known as
dicarboxyinulin (DCI). Due to this fructose ring-opening and the direct formation of the calcium-complexing oxydiacetate
(ODA) structure, DCI shows excellent calcium sequestering properties (SC) (2.0–2.5 mmol Ca/g active substance) [39].
Moreover, due to the linear relationship between the degree of oxidation, the formation of the ODA structure and hence the
SC value, it is easy to compromise between the oxidation degree and the biodegradability.

7.2.2. Carboxymethylation
Heating an aqueous solution of inulin in the presence of an excess of sodium hydroxide with monochloroacetic acid results
in the production of O-(carboxymethyl) inulin (CMI) (see Fig. 3) [42, 43]. The reaction efficiency of this etherification very
much depends on parameters such as reaction temperature and water content: lower water content and temperature
result in a higher efficiency.

Figure 3. Carboxymethylation of inulin

CMI is a good inhibitor of the crystallization of calcium carbonate. Even at very low concentrations (5–200 ppm), CMI can
extend the induction period for crystallization by absorption of the carboxylate groups onto the crystal surfaces.

7.2.3. Hydrophobization
Hydrophobization of inulin, e.g., grafting of alkyl chains on the inulin backbone, can be done in two different ways: by
esterification or by carbamoylation.

The esterification of inulin is carried out by reaction with acid chlorides or anhydrides; the introduced alkyl chain can range
from 1 to 22 carbon atoms (see Fig. 4) [44]. The reaction is usually performed either in solvents such as pyridine, N,N-
dimethylformamide or dimethyl sulfoxide, or solvent-free, using a kneader or (preferably) an extruder. As catalyst 4-
(dimethylamino)pyridine, sodium acetate, or potassium carbonate can be used.

page 7 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 4. Esterification of inulin

The carbamoylation of inulin (modification of inulin with an alkyl-isocyanate) is done in a polar organic solvent such as
N,N-dimethylformamide or N-methylpyrolidinone, as shown in Figure 5 [45].

Figure 5. Carbamoylation of inulin

Both the esterified and carbamoylated inulins show interesting interfacial properties. Depending on the length of the alkyl
chain and the degree of substitution these inulin derivatives are able to decrease the surface tension of aqueous solutions
and the interfacial tension at an oil–water interface [46]. Moreover, the critical micelle concentration (cmc) of
hydrophobized inulins is approximately 0.01 % (w/v). The above described tensioactive properties make these molecules
suitable for use in detergent applications, cosmetics, or crop protection formulations.

[Top of Page]

8. Analysis
Although several analytical techniques are available for the determination of inulin, no single method gives a complete and
quantitative analysis of all the components present. A combination of different methods is therefore often necessary.

8.1. HPLC
Two columns in series in the K+ form are used (Aminex HPX-87 K+) for optimal separation. The separation into fructose,
glucose, DFA (difructose-dianhydride), GF, F2, and F3 is optimal, but further separation into DP3, DP4, and DP ≥ 5 is not
very precise. DP3 and DP4 fractions are not pure, but include GF2 plus F4 and GF3 plus F5, respectively. The fraction of
DP ≥ 5 is the integrated sum of DP5 and higher DP molecules. As this fraction might include small oligosaccharides as
well as high DP inulins, a fixed response factor can not be determined, which makes this analysis unsuited for quantitative
inulin determination. The method is well adapted to evaluate the relative amounts of the different compounds present;
especially the amounts of the non-inulin compounds glucose, fructose, and sucrose.

8.2. Gas Chromatography

A high temperature capillary gas chromatographic method has been developed for the quantitative determination of inulin-
type oligosaccharides with DP <10 [47]. Sample preparation involves oximetrimethylsilylation (oxymation) and silylation of
the extracted sugars. The method is accurate and specific.

Moreover (2 6) oligosaccharides can be clearly distinguished from (2 1) compounds and GFn molecules from Fn
molecules.

8.3. HPAEC Analysis (Dionex)

A differentiation between GFn and Fn compounds can be made using High Pressure Anion Exchange Chromatography
(HPAEC) which furthermore provides a fingerprint of the molecular mass distribution of inulin. This analytical technique
uses a Dionex series 4000 ion chromatograph (Carbo-Pac PA-1 column) coupled with a Pulsed Amperometric Detector
(PAD). During the analysis, the carbohydrates are eluted with a NaOH/NaAc gradient [48]. The major drawback of
HPAEC-PAD is that it is very difficult to quantify the high DP compounds, due to the lack of appropriate standards and the
reduced sensitivity of the PAD detector for high DP polymers.

When coupling HPAEC with a Pulsed Electrochemical Detector (PED), oligomers up to DP17 can be quantified and the
weight fraction of each compound present can be calculated [49].

The average degree of polymerization is determined by the fructose to glucose ratio after complete hydrolysis with
inulinase (Fructozyme, NOVO) (at pH 4.5 and 60 °C for 30 min): .

8.4. Permethylation
The type of linkage and the occurrence of branching are checked by permethylation followed by reductive cleavage and in
situ acetylation [50]. The reaction products are analyzed using capillary gas chromatography (CGC).

8.5. Quantitative Determination

The official AOAC method no. 997.08 was developed to quantitatively measure inulin, e.g., in food [51]. An overview of the
method is given in Figure 6.

page 8 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 6. Flow diagram of the enzymatic fructan determination method (AOAC no. 997.08)

If inulin is the only compound present in the sample, the method consists only of steps 1 and 3. The inulin is extracted from
a substrate at 85 °C for 10 min; part of the extract is put apart for determination of free fructose, glucose, and sucrose by
any reliable chromatographic method available (HPLC, HGC, or HPAEC-PAD), the other part is submitted to an enzymatic
hydrolysis, identical to the one described above for DP determination. After this hydrolysis step, resulting fructose, and
glucose are determined again by chromatography. By subtracting the initial glucose, fructose, and sucrose contents from
the final ones, the following formula can be applied:

where k (<1) depends on the DP of the inulin analyzed and corrects for the water gain after hydrolysis. Ginu and Finu are
glucose and fructose, respectively, strictly originating from inulin.

If a complex sample needs to be analyzed, as is often the case when dealing with food products, an amyloglucosidase
treatment must be included before the last step and an extra sugar analysis performed to avoid overestimation of the
glucose originating from starch and maltodextrins present.

Although AOAC method 997.08 is very reliable, it is very labor intensive and requires the use of chromatographic
apparatus. AOAC method no. 999.03, based on the use of enzymes for hydrolysis and sugars determinations, just
requires a spectrophotometer and some other standard lab equipment [52]. This method is reliable for the determination of
inulin, but it cannot be used for the determination of oligosaccharides obtained by hydrolysis of inulin, as Fn-type
compounds are significantly underestimated.

[Top of Page]

9. Legal Aspects
Inulin is legally classified as food or food ingredient, and not as additive, in all countries where it is used. As a
consequence, inulin is not listed as “accepted food additive” in the standard positive lists, e.g., from the European Union or
from Codex Alimentarius. In the USA, inulin was declared “Generally Recognized As Safe” in 1992.

Inulin is classified as ‘dietary fiber’ in all European countries and in most other countries. It also complies with the Codex
Alimentarius definition of Dietary Fiber (Codex Guidelines on Nutrition Labelling CAC/GL 2-1985, Rev. 1 - 1993).

[Top of Page]

10. Economic Aspects and Trade Names

Currently, inulin is produced industrially by three companies, all of them using chicory as raw material: ORAFTI, Belgium, a
subsidiary of RAFFINERIE TIRLEMONTOISE (trade names RAFTILINE for food applications, RAFTIFEED for animal feed
and INUTEC for non-food uses), WARCOING, Belgium (trade name FIBRULINE) and SENSUS, The Netherlands, a
subsidiary of COSUN (trade name FRUTAFIT).

Chicory processing has evolved from a laboratory and pilot plant project to a full industrial scale. While the production of
inulin as such is not regulated by quota, the production of chicory fructose (obtained by hydrolysis of inulin, see
Fructose) is. As these quota are officially attributed by the European Union, they give an idea of the production scale [5].

The fact that the land surface assigned to chicory for inulin/fructose production has evolved from a few hundred hectares
in 1990 to more than 20 000 ha in 2000 is another indication for the growth of the inulin-related business.

[Top of Page]

References
1. V. Rose, Über eine eigenthümliche vegetablische Substanz, Gehlens Neues Algem. Jahrb. Chem. 3 (1804) 217.
2. T. Thomson: A System of Chemistry, 5th London edition, vol. 4, Abraham Small, Philadelphia 1818, p. 65.
3. J. Sachs, Über die Sphärokristalle des Inulins und dessen mikroskopische Nachweisung in den Zellen, Bot. Ztg. 22
(1864) 77.
4. H. Root, M. Baker, Inulin and artichokes in the treatment of diabetes, Arch. Intern. Med. 36 (1925) 126–145. Links
5. A. Franck, L. De Leenheer: “Inulin”, in A. Steinbuchel (ed.): Biopolymers, vol. 6, VCH Weinheim, Germany 2002, pp.
439–479.
6. A. L. Watherhouse, N. J. Chatterton: “Glossary of fructan terms”, in M. Suzuki, N. J. Chatteron (eds.): Science and

page 9 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Technology of fructans, CRC Press, Florida 1993, pp. 2–7.
7. L. De Leenheer, H. Hoebregs, Progress in the elucidation of the composition of Chicory inulin, Starch 46 (1994) 193–
196.
8. M. Wiedmann, M. Jager, Synergistic sweeteners, Food Ingredients and Analysis International, Nov.–Dec. (1997) 51–
56.
9. Tiense Suikerraffinaderij, Compositions having a creamy structure and containing fructan, preparation method
therefor and uses thereof, Patent Application WO 93 06744, 1993 (A. Frippiat, G. Smits).
10. Tiense Suikerraffinaderij, Inulin based hydrocolloid compositions, Patent Application EP 867 470, 1998 (A. Frippiat).
11. J. Van Loo, J. Cummings, N. Delzenne, H. Englyst, A. Franck, M. Hopkins, N. Kok, G. Macfarlane, D. Newton, M.
Quigley, M. Roberfroid, T. van Vliet, E. van den Heuvel, Functional food properties of non-digestible oligosaccharides:
a consensus report from the ENDO project (DGXII AIRII-CT94-1095), Br. J. Nutr. 81 (1999) 121–132. Links
12. L. Ellegard, H. Andersson, I. Bosaeus, Inulin and oligofructose do not influence the absorption of cholesterol, and the
excretion of cholesterol, Fe, Ca, Mg and bile acids but increase energy excretion in man. A blinded, controlled cross-
over study in ileostomy subjects, Eur. J. Clin. Nutr. 51 (1997) 1–5. Links
13. M. Roberfroid, G. R. Gibson, N. Delzenne, Biochemistry of oligofructose, a non-digestible fructo-oligosaccharide: an
approach to estimate its caloric value, Nutrition Reviews 51 (1993) 137–146. Links
14. A. Beringer, R. Wenger, Inulin in der Ernährung des Diabetikers, Deutsch. Zeitschr. F. Verdauungs-u.
Stoffwechselkrankh. 15 (1955) 268–272.
15. L. Prosky, Inulin and Oligofructose are Part of the Dietary Fiber Complex, J. AOAC Int. 82 (1999) 223–226. Links
16. M. B. Roberfroid: “Health benefits of non-digestible oligosaccharides,” in Kritchevsky and Bonefield (Eds.): Dietary
Fiber in Health and Disease, Plenum Press, New York 1997, pp. 211–219.
17. B. Kleessen, B. Sykura, H. J. Zunft, Effect of inulin and lactose on fecal microflora, microbial activity and bowel habit
in elderly constipated persons, Am. J. Clin. Nutr. 65 (1997) 1397–1402. Links
18. E. Den Hond, B. Geypens, Y. Ghoos, Effect of high performance chicory inulin on constipation, Nutrition Research 20
(2000) 731–736. Links
19. N. M. Delzenne, C. Daubioul, A. Neyrinck, M. Lasa, H. S. Taper, Inulin and Oligofructose modulate lipid metabolism in
animals: review of biochemical events and future prospects, Br. J. Nutr. 87 (2002) Suppl.2 S225–S259.
20. K. G. Jackson, G. R. J. Taylor, A. M. Clohessy, C. M. Williams, The effect of the daily intake of inulin on fasting lipid,
insulin and glucose concentrations in middle-aged men and women, Br. J. Nutr. 82 (1999) 23–30. Links
21. G. R. Gibson, X. Wang, Bifidogenic properties of different types of fructo-oligosaccharides, Food Microbiology 11
(1994) 491–498.
22. G. R. Gibson, E. R. Beatty, X. Wang, J. H. Cummings, Selective stimulation of bifidobacteria in the human colon by
oligofructose and inulin, Gastroenterology 108 (1995) 975–982.
23. M. Roberfroid, J. Van Loo, G. Gibson, The Bifidogenic nature of chicory inulin and its hydrolysis products, J. Nutr. 128
(1998) 11–19.
24. D. S. Reddy, R. Hamid, C. V. Rao, Effect of dietary oligofructose and inulin on colonic preneoplastic aberrant crypt
foci inhibition, Carcinogenesis 18 (1997) 1371–1374.
25. I. R. Rowland, C. J. Rummey, J. T. Coutts, L. Lievense, Effects of Bifidobacterium longum and inulin on gut bacterial
metabolism and carcinogen-induced aberrant crypt foci in rats, Carcinogenesis 19 (1998) 281–285.
26. H. Taper, N. Delzenne, M. B. Roberfroid, Growth inhibition of transplantable mouse tumours by non-digestible
carbohydrates, Int. J. Cancer 71 (1997) 1109–1112. Links
27. N. Delzenne, J. Aertssens, N. Verplaetse, M. Roccaro, M. Roberfroid, Effect of fermentable fructo-oligosaccharides
on energy and nutrients absorption in the rat, Life Science 57 (1995) 1579–1587.
28. M. B. Roberfroid, J. Cumps, J. P. Devogelaer, Dietary Chicory Inulin increases Whole-Body Mineral Density in
Growing Male Rats, J. Nutr. 132 (2002) 3599–3602. Links
29. C. Coudray, J. Bellanger, C. Castiglia-Delavaud, C. Rémésy, M. Vermorel, Y. Rayssignuier, Effects of soluble or
partly soluble dietary fibres supplementation on absorption and balance of calcium, magnesium, iron and zinc in
healthy young men, Eur. J. Clin. Nutr. 51 (1997) 375–380. Links
30. I. J. Griffin, P. M. Davila, S. A. Abrams, Non-digestible oligosaccharides and calcium absorption in girls with adequate
calcium intakes, Br. J. Nutr. 87 Suppl. 2 (2002) S187–S191.
31. A. Franck, Prebiotics stimulate calcium absorption: a review, Milchwissenschaft 53 (1998) 427–429. Links
32. G. A. Hendry, R. K. Wallace: “The origin, distribution and evolutionary significancy of fructans”, in M. Suzuki, N. J.
Chatteron (eds.): Science and Technology of fructans, CRC Press, Florida 1993, pp. 119–139.
33. J. Van Loo, P. Coussement, L. De Leenheer, H. Hoebregs, G. Smits, On the presence of inulin and oligofructose as
natural ingredients in the Western diet, Crit. Rev. in Food Sci. Nutr. 35 (1995) 525–552.
34. W. J. M. Meyer, E. W. J. M. Matthysen, G. E. L. Borm: “Crop characteristics and inulin production of Jerusalem
Artichoke and Chicory”, in A. Fuchs (ed.): Inulin and inulin-containing crops, Studies in Plant Science, 3rd ed.,
Elsevier Science Publishers, 1993, pp. 29–44.
35. L. De Leenheer: “Production and use of inulin: Industrial reality with a promising future”, in H. Van Bekkum, H. Röper,
F. Voragen (eds.): Carbohydrates as organic raw materials III, VCH Publ. Inc., New York 1996, pp. 67–92.
36. Tiense Suikerraffinaderij, Fractionated polydisperse compositions, Patent Application WO 96 01849, 1994 (G. Smits,
L. Daenekindt, K. Booten).
37. A. Franck, Technological functionality of inulin and oligofructose, Br. J. Nutr. 87 Suppl. 2 (2002) S287–S291. Links
38. C. Stevens, A. Meriggi, K. Booten, Chemical modification of inulin, a valuable renewable resource and its industrial
applications, Biomacromolecules 2 (2001) 1–16. Links
39. A. Besemer, H. van Bekkum, The relation between calcium sequestering capacity and oxidation degree of dicarboxy-

page 10 of 11
Inulin : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

starch and dicarboxy-inulin, Starch/Stärke 46 (1994) 419–422.

40. TNO, Method for preparation of calcium-binding polycarboxy compounds based on polysaccharides and
replacements for phosphates in detergents, based on these polycarboxy compounds, PCT WO 91/17189, 1991 (A.
Besemer, H. Van Bekkum).
41. D. Verraest, J. Peters, H. van Bekkum, Oxidation and carboxymethylation of sucrose and inulin, Zuckerindustrie 120
(1995) 799–803. Links
42. D. Verraest, J. Peters, J. Batelaan, H. van Bekkum, Carboxymethylation of inulin, Carbohydrate Research 271 (1995)
101–112. Links
43. Akzo Nobel, Carboxymethyl inulin, PCT WO 95/15984, 1995 (D. Verraest, J. Batelaan, J. Peeters, H. van Bekkum).
44. Südzucker, Aliphatic Carboxylate esters of inulin, US Patent US 5 877 144, 1999 (S. Ehrhardt, A. Haji Begli, M. Kunz,
L. Sheiwe).
45. Tiense Suikerraffinaderij, Surface-active alkyl-urethanes of fructans, PCT WO 99/64549, 1999 (C. Stevens, K.
Booten, L. Laquiere, L. Daenekindt).
46. C. Stevens, A. Meriggi, M. Peristeropoulou, P. Christov, K. Booten, B. Levecke, A. Vandamme, N. Pittevils, T. Tadros,
Polymeric surfactants based on inulin, a polysaccharide extracted from chicory. 1. Synthesis and interfacial
properties, Biomacromolecules 2 (2001) 1256–1259. Links
47. D. Joye, H. Hoebregs, Determination of oligo-fructose, a soluble dietary fiber, by high-temperature capillary gas
chromatography, J. AOAC Int. 83 (2000) 1020–1025. Links
48. N. J. Chatteron, P. A. Harrison, W. R. Thornley, J. H. Bennet: “Separation and quantification of fructan (inulin)
oligomers by anion exchange chromatography”, in A. Fuchs (ed.): Inulin and Inulin-containing crops, 3rd ed., Elsevier
Science Publishers, 1993, pp. 93–99.
49. J. W. Timmermans, M. B. van Leeuwen, H. Tournois, D. de Wit, J. F. G. Vliegenthart: Proceedings of the fourth
seminar on inulin, Wageningen, 1993, pp. 12–15.
50. I. Ciucanu, F. Kerek, A simple and rapid method for the permethylation of carbohydrates, Carbohydr. Res. 131 (1984)
209–217. Links
51. H. Hoebregs, Fructans in foods and food products, ion-exchange chromatographic method: Collaborative study, J.
AOAC Int. 80 (1997) 1029–1037. Links
52. B. McCleary, A. Murphy, D. Mugford, Measurement of total fructan in foods by enzymatic/ spectrophotometric
method: Collaborative study, J. AOAC Int. 83 (2000) 356–364. Links
53. P. A. A. Coussement, Inulin and Oligofructose: Safe intakes and Legal Status, J. Nutr. 129 (1999) 1412S–1417S.
Links
[Top of Page]

About Wiley InterScience | About Wiley | Privacy | Terms & Conditions

Copyright © 1999-2007John Wiley & Sons, Inc. All Rights Reserved.

page 11 of 11
 Iodine and Iodine Compounds 1

Iodine and Iodine Compounds

Phyllis A. Lyday, US Bureau of Mines, Washington, DC 20241, United States

1. Introduction . . . . . . . . . . . . . . . . . 1 7. Uses . . . . . . . . . . . . . . . . . . . . . . . 5
2. Physical Properties . . . . . . . . . . . . . 1 8. Economic Aspects . . . . . . . . . . . . . 7
3. Chemical Properties . . . . . . . . . . . . 2 9. Inorganic Iodine Compounds . . . . . . 8
4. Occurrence . . . . . . . . . . . . . . . . . . 3
10. Organic Iodine Compounds . . . . . . . 10
5. Production . . . . . . . . . . . . . . . . . . 4
6. Quality Specifications, Transportation, 11. Toxicology and Occupational Health . 10
and Storage . . . . . . . . . . . . . . . . . . 5 12. References . . . . . . . . . . . . . . . . . . 12

1. Introduction por exhibits fluorescence at low pressure, emit-

ting a yellowish-green light when illuminated
Iodine, I [7553-56-2], atomic mass 129.9044, by sunlight or the light from an electric arc. It
atomic number 53, a nonmetallic element of the also rotates the plane of polarized light when
halogen family, appears in group 17 of the pe- placed in a magnetic field between crossed Nicol
riodic table between bromine and astatine. The prisms, and is thermoluminescent above 500 ◦ C.
electronic configuration of the iodine atom is Some physical properties of iodine are as fol-
[Kr] 4d 10 5s2 5p5 . The relative atomic mass of lows:
the only stable isotope of iodine is 127. There are
22 artificial isotopes with masses between 117 Atomic radius 133.1 pm
Covalent radius 136 pm
and 139, and fourteen of these yield significant Ionic radius 220 pm
radiation, 122 I through 123 I and 126 I through 136 I Electronegativity (Pauling scale) 2.66
(→Radionuclides). The half-lifes of the isotopes Electron affinity 301.9 kJ/mol
vary; the shortest, 136 I, has a half-life of 86 s and Ionization potential 10.451 eV
Solid iodine
the longest, 129 I, has a half-life of 1.6×107 a. mp 113.6 ◦ C
The isotope 129 I is a product of the fission of Density at 101.3 kPa
235 d 20 4.93 g/cm3
U with slow neutrons. Because of their use 4
d 60
4 4.866 g/cm3
in radioactive tracer work and radiotherapy, the Refractive index, n20 3.34
most important radioisotopes are 131 I, which has Vapor pressure
D

a half-life of 8 d, and 125 I, which has a half-life at 25 ◦ C 0.031 kPa

of 60 d. at 113.6 ◦ C 9.17 kPa
Electrical resistivity
Iodine was first discovered by Courtois in at 25 ◦ C 5.85×106 Ωcm
1811. While cleaning kelp pans with hot sul- at 110 ◦ C 8.33×105 Ωcm
furic acid, he noticed the evolution of a violet Thermal conductivity 0.421 W m−1 K−1
vapor that condensed as crystals on the cooler Dielectric constant at 25 ◦ C 10.3
Entropy at 298.2 K 116.81 JK−1 mol−1
parts of the vessel. Courtois provided samples Heat of fusion at mp 62.17 J/g
of this unknown material to other scientists, and Heat of sublimation at mp 238.4 J/g
in 1813 Davy and Gay-Lussac identified it as a Specific heat, J g−1 K−1 0.21163 + 19.615
at 25 – 113.6 ◦ C ×10−5 t
new element. Gay-Lussac named the substance
Liquid iodine
“iode” after the Greek word for violet-colored. bp 184.35 ◦ C
In 1814 Gay-Lussac published the results of his Critical temperature 546 ◦ C
investigations in “Mémoire sur l’Iode”. Critical pressure 11.6 MPa
Critical compressibility factor, 0.268
Zc
Density
2. Physical Properties d 120 3.960 g/cm3
d 180 3.736 g/cm3
Iodine is the first member of the halogen family
that is solid at ordinary temperature. Iodine va-

c 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10.1002/14356007.a14 381
2 Iodine and Iodine Compounds
Electrical resistivity at 140 ◦ C 1.1×105 Ωcm
lenium, but only reacts indirectly with carbon,
Kinematic viscosity
at 116 ◦ C 0.5727 mm2 /s (=cSt) nitrogen, oxygen, and some of the noble met-
at 184 ◦ C 0.3785 mm2 /s (=cSt) als. Unlike chlorine, iodine does not form com-
Dynamic viscosity pounds with carbon monoxide, nitric oxide, or
at 116 ◦ C 2.268 mPa · s
at 184 ◦ C 1.414 mPa · s
sulfur dioxide. Ferric and cupric salts, and com-
Dielectric constant at 118 ◦ C 11.08 pounds of vanadium, chromium, and manganese
Heat of vaporization at bp 164.46 J/g in their highest oxidation states are reduced in
Specific heat at 113.6 – 184 ◦ C 0.3163 J g−1 K−1 acid solution by iodide ion, with the liberation
Gaseous iodine
Vapor density
of free iodine.
at 101.3 MPa, 185 ◦ C 6.75 g/L
Entropy at 298.2 K 260.63 JK−1 mol−1 Reaction with Organic Compounds. Io-
Specific heat 25 – 1200 ◦ C 0.1464 J g−1 K−1
dine reacts with hydrocarbons but the equilibria
are often unfavorable because hydrogen iodide
is liberated, which may reduce the alkyl iodide
The solubilities of iodine in water and organic to the hydrocarbon with formation of iodine.
solvents and the color of the solutions are given Iodination can sometimes be made to go to
in Table 1. completion by preventing the formation of free
Table 1. Solubility of iodine in various solvents at 25 ◦ C hydrogen iodide. The following three methods
are commonly used:
Solvent Solubility, g/kg Color

Benzene 164.0 red 1) Addition of an oxidizing agent, such as nitric

Butan-2-ol 97 brown acid, iodic acid, or a peroxide
Carbon disulfide 197.0 red 2) Neutralization of hydrogen iodide with a
Carbon tetrachloride 19.2 violet
Chloroform 49.7 violet
base
Ethanol 271.7 brown 3) Combination with mercuric salts to form
Ether 337.3 brown mercuric iodide or an iodomercurate(II)
Ethyl acetate 157.0 brown complex [7], [8]
Hexane 13.2 violet
Toluene 182.5 red
Water 0.34 brown
Iodine also adds to carbon – carbon double
bonds in compounds such as ethylene, but the
reactions often do not go to completion due to
their reversibility.
Iodine is used as a catalyst in many organic
3. Chemical Properties reactions (e.g., acylation, halogenation, dehy-
dration, isomerization, and pyrolytic decompo-
The principal oxidation states of iodine are −1,
sition).
+1, +3, +5, and +7, but an oxide of formula IO2
with a formal oxidation state of +4 also exists.
Except for the +4 state, thermodynamically sta- Detection of Iodine. Iodine as the free ele-
ble compounds are formed in all these oxidation ment can be detected by the characteristic blue
states. Iodine is the least electronegative of the color given with a starch solution, or by the vi-
halogens and is usually less violent in its reac- olet color of its vapor and of solutions in car-
tions than fluorine, chlorine, or bromine. It ex- bon tetrachloride or carbon disulfide. Iodine is
hibits mild oxidizing properties in acid solution. also detected by the precipitation of yellow sil-
Iodine is a stronger oxidizing agent than astatine, ver iodide, black palladium iodide, or yellow or
the fifth halogen. A basic chemical characteris- red mercuric iodide. Iodine can be quantitatively
tic of the halogens is the formation of a salt by determined in acidic aqueous solutions by titrat-
direct reaction with a metal. Iodine reacts slowly ing with a standard solution of sodium thiosul-
with those metals that form nonvolatile iodides, fate using starch as the indicator. Iodides can
for example, magnesium, especially if they are also be assayed by the Volhard method in which
not finely divided. an excess of a standard silver nitrate solution
Iodine forms compounds with all the ele- is added and then backtitrated with a standard
ments except the noble gases, sulfur, and se-

thiocyanate solution with ferric alum as the in-
dicator. Quantities of up to 0.05 µg of iodine can
be determined by colorimetric methods.
4. Occurrence
According to the United States Geological Sur-
vey, iodine is the forty-seventh most abundant
element in the earth’s crust, counting the rare
earths as a single element. Small quantities of
iodine are widely distributed in rocks, soils, un-
derground brines, and seawater.
Minerals. Iodine is found in nature in
the following mineral forms: bruggenite,
Ca (IO3 )2 · H2 O; dietzeite, Ca2 (IO3 )2 · CrO4 ;
iodoargyrite, AgI; iodoembolite, Ag (Cl, Br, I);
lautarite [7789-80-2], Ca (IO3 )2 ; and marshite
[24401-69-2], CuI. The most important iodine-
containing mineral is lautarite (anhydrous cal-
cium iodate) [9]. Iodate occurs in the minerals
bruggenite, dietzeite, and lautarite, in caliche
deposits (between 2 and 50 feet below ground
level) in the Atacama Desert of Chile’s Antofa-
gasta and Tarapaca Provinces. The iodate is be-
lieved to have formed from oceanic emissions
during the late Tertiary and Quaternary epochs.
The atmospheric photochemical oxidation of io-
dide to iodate is thought to have occurred in the
troposphere and at ground level.
Seawater contains approximately 0.05 ppm
of iodine [11]. The amount of iodine is related
to the salinity. For example, the Red Sea con-
tains 0.06 ppm and the Baltic Sea, 0.01 ppm. Io-
dine is concentrated by marine organisms, (e.g.,
sponges and seaweed) and by fish, especially in
the thyroid gland and in liver oil, which contain
510 – 69 200 ppm [11]. The brown algae, Lami-
naria and Fucus, which grow in temperate zones
of the Northern Hemisphere, have the highest
iodine content (0.0279 – 0.4535 wt % after dry-
ing). The name kelp was originally applied to the
ash obtained by burning seaweed, but has been
extended to include the seaweed itself, partic-
ularly the varieties yielding potash, iodine, and
other commercially important products.
Brines. Iodine occurs frequently as potas-
sium or sodium iodide in subsurface brines as-
sociated with oil and gas deposits.
Iodine and Iodine Compounds 3
In Oklahoma, the United States, iodine oc-
curs in brines with concentrations of 150 –
1200 ppm in the Morrowan formation of the
Pennsylvanian age [12]. In California, brines
containing 30 – 70 ppm of iodine are found in
the Monterey Formation of the Middle Miocene
age and in the Repetto Formation of the
Pliocene age. In Michigan, brines containing
about 30 ppm of iodine occur in the Sylvania
Formation of the Devonian age. In playa lakes
in the western United States, iodine occurs at
a concentration of 14 ppm [13]. In Japan, the
major iodine production occurs in the south-
ern Kanto gas field that underlies the Chiba,
Tokyo, and Kanagawa Prefectures. The primary
iodine reservoirs occur in the Kiwada, Otadai,
and Umegase Formations of the Pliocene to
middle Pleistocene. In the Miyazaki Prefecture,
iodine and natural gas are produced from the
late Pliocene Sadowara and Tonogori members
of the Miyazaki group. Iodine concentrations
in the Sadowara gas field range between 59
and 101 ppm. Iodine is contained in the Tomi-
gusuku Formation of the Shimajiri group of late
Miocene to early Pliocene. Iodine concentra-
tions in the southern Okinawa gas field range
between 40 and 110 ppm. In the Niigata Pre-
fecture, iodine concentrations range between 7
and 68 ppm in brines in the Niigata and Hachi-
mori gas fields. The Niigata basin is of the
Pliocene to Pleistocene age [14]. In Indonesia,
iodine occurs with trace amounts of bromine in
brines associated with oil. The most important
iodine-producing area is the Gujangon anticline
of sandstone and diatomaceous marls of the Up-
per Pliocene, Kailibeng Formation. The brines
were reported to contain about 50 ppm of io-
dine [15]. Information on iodine production in
other parts of the world is incomplete, but the
former Soviet Union has reported iodine con-
centrations of 67 ppm in brines associated with
oil fields at Lake Bejuk-Schor, near Baku in the
Azerbaidzhan area on the Caspian Sea. Brines
in the Slavyansko – Troitsko area were reported
to have iodine concentrations of 80 – 120 ppm in
the Neftechinski oil field near the Black Sea. At
the Nebit-Dag iodine – bromine plant in Turk-
menistan, iodine is produced from brines that
contain 250 g/L of a combination of boron, cal-
cium, magnesium, sodium compounds, and io-
dine [16].
4 Iodine and Iodine Compounds
5. Production
There have been three phases in the history of
iodine production:
1) Production from seaweed began in 1817 and
continued until 1959.
2) Production as a byproduct of sodium nitrate
production began in 1852.
3) Production from brines began in 1854.
Six processes have been used to produce io-
dine from brines; the three that are currently used
are described here in detail. Other, obsolete pro-
cesses include the nitrate – sulfite and silver io-
dide methods [16].
Production from Chile Salpeter. Chile is
the only location in the world where an iodine-
containing deposit is commercially mined and
processed. Iodine is present in the deposits as
calcium iodate (lautarite and dietzeite). Process-
ing of the deposits began prior to the 1800s. The
first iodine plant opened in 1852, but closed
during the first year of production. Iodine pro-
duction did not begin again until about 1868 and
has continued ever since. Iodine is obtained as
a byproduct of sodium nitrate production.
Crushed ore is leached to give a solution con-
taining sodium nitrate and calcium iodate. After
removal of sodium nitrate by precipitation, the
iodate-rich mother liquor is treated with aqueous
sodium bisulfite to reduce the iodate to iodide as
follows [17]:
2 NaIO3 + 6 NaHSO3 → 2 NaI + 3 Na2 SO4 + 3 H2 SO4
Addition of an equivalent quantity of the
mother liquor then liberates free iodine.
5 NaI + NaIO3 + 3 H2 SO4 → 3 Na2 SO4 + 3 H2 O + 3 I2
The iodine precipitate is collected by filtra-
tion and is then fed into a melting tank where it
is indirectly heated with steam to 120 ◦ C. The
molten iodine flows to a settler tank for flaking.
The flaker consists of two water-cooled cylindri-
cal drums made of Hastelloy C, an alloy which is
resistant to corrosion by iodine. The molten io-
dine flows into the trough formed by the drums
and is removed by scraping. The iodine flakes
are carried by conveyor belt to be packed in fiber
drums lined with double-ply polyethylene.
Production from Brine. In the United
States, iodine is produced in Oklahoma by two
brine processes. The brines contain ca. 300 ppm
of iodine. In the blowing-out process, brine is
pumped from the production well into a surge
tank. The iodide-rich brine is fed into a column,
where it is acidified with hydrochloric acid and
treated with chlorine to oxidize the iodide to
iodine. The iodine is stripped from the brine by
steam and sent to a second column, where it is
reduced and absorbed in an aqueous solution
of hydriodic acid and sulfuric acid. The iodine
is reduced to iodide and absorbed into solution.
The iodide solution is continuously fed to a stor-
age tank where chlorine gas precipitates crystals
of elemental iodine. The iodine is filtered from
the solution, and the acid stream is returned to
the absorption column. The iodine is purified
by melting under concentrated sulfuric acid to
remove water and organic contaminants. It is
then dried, and packaged in polyethylene-lined
cardboard cartons for shipping [12].
The second process is an adsorption process
that is used with brines from oil field disposal
plants. The brines are skimmed in holding tanks
to remove suspended oil. Oil and suspended mat-
ter are removed by filtration through sand. Hy-
drochloric acid is added to the brine and chlorine
gas to oxidize the iodide to iodine. The brine
is then passed through the absorber, where io-
dine is absorbed, and the waste brine is neutral-
ized and sent to a disposal well. The adsorbent
is treated with sodium hydroxide solution that
elutes the iodine as sodium iodide. The iodide-
rich eluate is sent to a refinery that uses a process
similar to the blowing-out process.
In Japan, iodine is produced from under-
ground brines associated with gas fields by
a blowing-out process and by two adsorption
processes (the ion-exchange and the sloping
fluidized-bed processes) [16]. The blowing-out
process is similar to that used in the United
States.
In the ion-exchange process, the resin bed is
fixed, and filtration of the brine is necessary to
remove any solids before adsorption. The acid-
ified brine is treated with chlorine to liberate
iodine, which is adsorbed on an ion-exchange
resin (Amberlite) as polyiodides. The iodine is
released from the resin in two stages by elution
with sodium hydroxide and then with sodium

chloride. After elution, the resulting iodate –
iodide solution is acidified with sulfuric acid to
form iodine. The iodine is then recovered from
the resulting slurry by filtration or centrifuga-
tion, and purified by sublimation or by melting
under sulfuric acid.
In the sloping fluidized-bed process, sodium
iodide is also adsorbed on an ion-exchange resin,
but in this case the resin is mobile and filtration
of the brine is not necessary. The resin is trans-
ferred to a recovery plant, and the sodium iodide
is removed as hydrogen iodide by elution with
sulfuric acid. Iodine is precipitated by addition
of chlorine.
The Wengfu Mine in China has reported io-
dine concentrations of 0.004 %. Calcining at
1050 ◦ C yielded an iodine-rich gas that could
be further processed. Iodine has been recovered
during process design testing [18].
6. Quality Specifications,
Transportation, and Storage
Commercial crude iodine normally has a mini-
mum purity of 99.5 %. Impurities are chiefly wa-
ter, sulfuric acid, iron, and insoluble materials.
The U.S. Pharmacopoeia (U.S.P.) XVII specifies
an iodine content not less than 99.8 %. The Com-
mittee on Analytical Reagents of the American
Chemical Society (ACS) allows a maximum of
0.005 % total bromine and chlorine, and 0.010 %
nonvolatile matter.
Crude iodine is packed in fiber drums lined
with double-ply polyethylene that contain 25 –
50 kg. There are no specific transportation, ship-
ping, or safety requirements. As low tempera-
tures reduce vapor pressure and therefore de-
crease the rate of loss, a cool, well-ventilated
storage area is recommended.
Under normal conditions, the rate of loss
from intact commercial containers is negligible.
Investment and maintenance costs are high for a
new iodine plant using underground brines be-
cause brine sources usually do not contain high
concentrations of iodine and because 1 kg of
chlorine is required for the production of 1.4 –
1.8 kg of iodine. Electrical costs for maintain-
ing the pumps to bring the brine to the surface,
for air blowing the iodine, and for reinjecting
the brine are major components of maintenance
Iodine and Iodine Compounds 5
costs. Capital costs include high-quality stain-
less steels because iodine is highly corrosive. In
addition, the brine must be leased from private
landowners over large acreages for many years.
7. Uses
Iodine was first exploited commercially in 1916
in the form of potassium iodide as a remedy for
goiter. In 1916, tincture of iodine and iodoform,
CHI3 , were used as disinfectants for cuts and
abrasions, and for sanitation. In 1939, Louis Da-
guerre published details of his method for mak-
ing photographs on plated silver by using iodine
vapor to form a thin coating of light-sensitive sil-
ver iodide crystals. Clinical uses of iodine have
evolved from simple tinctures to radiopaque me-
dia, iodophors, and batteries in subcutaneously
implanted pacemakers [20]. Iodine is also used
in a wide variety of specialty chemicals.
Establishing an accurate end-use pattern for
iodine is difficult because iodine-containing in-
termediates are marketed before reaching their
ultimate end uses.
Uses include dietary supplements, catalysts,
inks and colorants, pharmaceuticals, radiopaque
diagnostic media, photographic equipment, san-
itary and industrial disinfectants, and stabiliz-
ers, as well as the production of motor fuels,
smog inhibitors, lubricants, high-purity metals,
and iodized salt. Iodine also has applications in
cloud seeding; worldwide about 11 000 kg of sil-
ver iodide is used in amounts of 10 – 50 g per
cloud seeding.
Catalysts. The major end use of iodine is in
catalysts. After the catalytic Monsanto process
for producing acetic acid was introduced in the
1960s, 90 % of new acetic acid capacity used the
process (→Acetic Acid, Chap. 4.2.). The pro-
cess involves methanol carbonylation with an
iodide-promoted rhodium complex as catalyst
[20]. In the final distillation, the acetic acid of-
ten contains organic iodides; these are removed
by crystallization or distillation [21]. Iodide cat-
alysts, such as titanium tetraiodide [7720-83-4]
and aluminum iodide [7784-23-8], are also sig-
nificant in the dehydrogenation of butane and
butene to butadiene [22], and in the preparation
of stereoregular polymers [23].
6 Iodine and Iodine Compounds
Stabilizers. Consumption of iodine as a sta-
bilizer in the manufacture of nylon for tire cord
and carpets, and for converting rosins, tall oil,
and other wood products to more stable forms is
the second major end use [24].
Butadiene reacts with iodine and copper
cyanide to yield a copper(I) iodide complex of
dehydroadiponitrile. Reaction with hydrocyanic
acid gives a high yield of adiponitrile, which
can be catalytically hydrogenated to give hexa-
methylenediamine, a precursor of nylon 66 [26].
In the reaction with HCN the starting materials,
copper(I) cyanide and iodine, are regenerated.
Animal Feeds. The third area of major io-
dine consumption is in the form of additives for
animal feeds. In 1986, the recommended amount
of ethylene diamine dihydroiodide [5700-49-2]
(EDDI) consumed in animal feeds was lowered
from 50 to 10 mg per head per day, in accor-
dance with the U.S. Food and Drug Adminis-
tration Compliance Policy Guideline (7125.18).
Cattle and sheep are fed iodine compounds for
the prevention of soft tissue lumpy jaw. The
use of iodized salt has significantly reduced
the incident of goiter in livestock in certain
iodine-deficient geographical regions. In addi-
tion, iodized protein in fowl and cattle feed in-
creases yields of eggs and milk. About 25 %
of the reported domestic consumption of io-
dine was in animal feeds, primarily as the com-
pound EDDI, but also as potassium iodide, cal-
cium iodate [7789-80-2], and calcium periodate
[7790-28-5]. Iodine is one of 14 mineral com-
modities listed by the National Feed Ingredients
Association as being used for domestic meat
production.
Disinfectant. Iodine is an effective germi-
cide for a wide range of microorganisms. The use
of iodine in sanitary preparations, such as disin-
fectants, is the fourth major area of consump-
tion. Iodine is often used in conjunction with
iodine-complexing nonionic surfactants or poly-
mers (iodophors) in disinfectants that are used in
dairies, laboratories, and food processing plants,
and for the sanitation of dishes in restaurants
[26]. Polyvinyl pyrrolidinone – iodine complex
[25655-41-8], PVP-iodine, is extensively used
because of its germicidal, bactericidal, fungi-
cidal, and general disinfecting properties [27]
(→Disinfectants). The solution contains about
10 % available iodine and about 5 % inactive io-
dide [28]. Globaline tablets are used by the army
to disinfect water supplies without boiling [29].
Iodine is used as a disinfectant in swimming
pools and is an efficient microbiocide. It is more
effective than chlorine against the standard fecal
indicators. High iodine concentrations are eas-
ily maintained and are more stable than chlorine
[30]. A study in Florida showed that concentra-
tions up to 5 ppm of iodine in potable water are
not deleterious to health and that 1 ppm is suffi-
cient to safely disinfect a water supply with no
loss of effectiveness at high pH values [31].
Pharmaceuticals. The fifth major area of io-
dine consumption is in pharmaceuticals, primar-
ily as radiopaque media, which are injected in-
travenously or intracerebrally to produce X-ray
photographs of the internal structure of the body
(→Radiopaque Media). Clinical use includes
cardioangiography, cerebral angiography, arte-
riography, urography, and phlebography. Media
listed by the U.S. International Trade Commis-
sion include meglumine diatrizoate [131-49-7],
sodium diatrizoate [737-31-5], iopanoic acid
[96-83-3], iothalamic acid [2276-90-6], and
sodium tyropanoate [7246-21-1]. These sub-
stances contain between 47 and 67 wt % io-
dine. A nonionic contrast medium, N,N -bis-
(2,3-dihydroxypropyl-5-[N-(2,3-dihydroxypro-
pyl)acetamidol]-2,4,6-triiodoisophthalamide,
which contains 46 wt % iodine, has also been
marketed [32] . Potassium iodide is widely used
as an expectorant in cough medicines. Hydri-
odic acid and potassium iodide are used in the
synthesis of amphetamine, methamphetamine,
and ethylamphetamine.
Miscellaneous Uses. One of the oldest in-
dustrial uses of iodine is in photography. Up
to 7 % of the silver salt in negative emul-
sions is iodide. Ammonium iodide [12027-06-4]
is used as a photographic developer. Iodine
is also a constituent of some dyes including
4 ,5 -diiodofluorescein [38577-97-8], rose ben-
gal [11121-48-5], and erythrosine [16423-68-0].
Erythrosine, certified food color Red No. 3,
contains 58 % iodine and is used in carbonated
soft drinks, powdered drinks, gelatin desserts,
icings, and pet food. 4 ,5 -Diiodofluorescein
[38577-97-8], C.I. Solvent Red 73, is used
in dyeing and printing cotton, in printing
 Iodine and Iodine Compounds 7

half-silk, and in dyeing jute and straw prod-
ucts. Sodium iodate can also be used in the
dyeing process as an oxidizing agent. Fluo-
roiodocarbons, such as CF3 (CF2 )7 I [507-63-1]
and CF3 (CF2 )7 CH2 CH2 I [2043-53-0], are in-
termediates in the manufacture of oil and wa-
ter repellents, surfactants, and fire-extinguishing
foams [33], [34].
The addition of iodine to aromatic hydrocar-
bons, such as n-butylbenzene [104-51-8], results
in the formation of charge-transfer complexes
that are effective lubricants for some metals, e.g.,
titanium or steels [35]. Iodine is also used to
produce high-purity forms of titanium, silicon,
hafnium, and zirconium in the van Arkel pro-
cess, which involves the formation of volatile
iodides [36].
tion of a 150-t/a iodine plant to extract iodine
from waste tailings from sodium nitrate produc-
tion. These tailings still contain ca. 50 % of the
original iodine content. This new plant will guar-
antee Chile a secure source of iodine from the
accumulated waste tailings.
In the United States, iodine is produced by
IoChem, North American Brine Resources, and
Woodward Iodine.
The United States has been dependent upon
imported iodine for roughly half of domestic
consumption since ca. 1960. United States im-
ports of iodine and potassium iodide according
to the exporting country are listed in Table 3.
Table 3. United States imports of iodine and potassium iodide
Product and country 1985 1986

Quan- Value, Quan- Value,

tity, t 103 $ tity, t 103 $
8. Economic Aspects Crude iodine
Belgium 10 174
From 1982 to 1986, world iodine production Canada 1 3
remained fairly static and amounted to ca. Chile 749 8 105 627 7 622
12 500 t/a (Table 2). In 1986, Japan accounted Japan 1 496 18 479 746 9 576
Total ∗ 2 254 26 761 1 373 17 199
for 57 % of world production, followed by Chile Resublimed iodine
(23 %), the Soviet Union (15 %), and China Federal Republic
(3.5 %); production figures for the United States of Germany 4 3
were withheld. Japan 185 2 158 1 204 15 530
Mexico 1 13
Table 2. Production of crude iodine a Sweden 1 17 1 24
Total ∗ 186 2 193 1 205 15 530
Country Production, t/a Potassium iodide
Belgium 1 17 3
1994a 1996b 2000b 2001b Brazil 1 10
Canada 3 37
Chile 5 600 5 000 9 800 10 500
Federal Republic
China 500 500 500
of Germany 3 30 3
Indonesia 80 70 70
India 5 64 23 279
Japan 7 200 6 200 6 760 6 100
Italy 1 3 28 5
Former Soviet Union 760 570 570
Japan 19 226 8 111
Total 16 700 13 800 19 100 19 600
Switzerland 1
a United Kingdom 1 54 1 42
[44]
b
[45] Total ∗ 30 396 37 492
Grand total 2 470 29 350 2 615 33 253

∗ Data may not add to totals shown because of independent

In Japan, iodine is produced entirely as a by-
product of natural gas production. When gas and
oil prices are low, the import of world supplies
of petrochemicals becomes economically viable
for Japan and therefore leads to a decrease in
production of domestic natural gas and the as-
sociated iodine.
Chile has large reserves of iodine associated
with sodium nitrate deposits, but until 1986 had
not provided capital for expansion or process im-
provements. In 1987, Chile completed construc-
rounding.
U.S. companies began importing large
amounts of potassium iodide in 1984, due to in-
creased demand (see Table 3).
Changes in the official price of iodine have
in the past been initiated by Japan followed
by Chile and the United States. The price of
crude iodine in the United States increased from
$ 10.40/kg in 1980 to $ 17.40/kg in 1988.
8 Iodine and Iodine Compounds

Table 4 summarizes the consumption of Iodides range from completely ionic to co-
crude iodine in the United States according to valent structures; for example, potassium iodide
product; iodine consumption according to use is is ionic and titanium tetraiodide is covalent. In
given in Table 5. general, the ionic iodides are the most soluble of
the halides of a given element.
Table 4. Consumption of crude iodine in the United States.
Commercially, iodides are the most impor-
Product Consumption, t/a tant class of iodine compounds. In general, they
are less fusible and volatile than chlorides or
1985 1986
bromides. At higher temperature, even the more
Resublimed iodine 87 70 stable iodides show dissociation. Volatile cova-
Potassium iodide 488 474
Sodium iodide 66 62
lent iodides, such as those of aluminum and tita-
Other inorganic compounds 566 565 nium, form inflammable mixtures with air. Dry
Ethylenediamine dihydroiodide 595 461 nitrogen triiodide, NI3 , formed by reacting io-
Other organic compounds 595 698 dine with excess ammonia, will explode at ordi-
Total 2346 2329 ∗
nary temperature under the slightest shock.
∗ Data do not add to total shown because of independent Potassium iodide [7681-11-0], KI,
M r 166.02, mp 686 ◦ C, forms colorless cubic
rounding.

crystals, which are soluble in water, ethanol,

methanol, and acetone. It is used in animal
Table 5. Estimated consumption of iodine in the United States
according to use feeds, catalysts, photographic chemicals, and
for sanitation. Potassium iodide is produced by
Use Consumption, t/a
reaction of potassium hydroxide with iodine.
1981 1982 1983 1984 1985 The product is purified by crystallization from
Animal feed additives 816 726 635 544 499 water.
Catalysts 727 635 816 862 816 Sodium iodide [7681-82-5], NaI, M r 149.92,
Inks and colorants 227 136 181 181 181 mp 662 ◦ C, forms colorless cubic crystals which
Pharmaceuticals 544 317 317 317 363
Photography 181 181 272 272 272 are soluble in water, ethanol, acetone and glyc-
Sanitary uses 635 454 544 544 544 erol. It is used in photography, for the produc-
Stabilizers 726 544 726 635 680 tion of organic chemicals, and as an expectorant
Other 181 181 181 181 181
Total 3991 3129 3673 3537 3537
in cough medicines. Sodium iodide is produced
by the addition of sodium hydroxide or sodium
carbonate to an acidic iodide solution. The an-
Consumption of iodine in the United States hydrous product is obtained from the solution
has decreased for of a variety of reasons. Some by evaporation.
companies have turned to importing interme- Ammonium iodide [12027-06-4], M r 144.96,
diate products or sell products produced in forms white tetragonal crystals which are solu-
other countries. For example, the majority of ra- ble in water, alcohol, glycerol, and methanol.
diopaque media are now imported. Furthermore, Upon heating, ammonium iodide sublimes with
use of the major animal feed additive, EDDI, decomposition. It is used as an expectorant and
has now been limited by the United States De- in photographic chemicals. Ammonium iodide
partment of Agriculture. In 1987, IoChem began is prepared from ammonia, iodine, and hydrogen
production at a new plant in Oklahoma, in a joint peroxide; from ammonia and hydrogen iodide;
venture with Schering (Federal Republic of Ger- or from ammonium carbonate and hydrogen io-
many). A major portion of all iodine produced dide. To prevent the crystals or solution from
at the plant is exported to Schering. turning yellow due to formation of free iodine,
up to 1 % of ammonium hypophosphite is added
as a stabilizer.
9. Inorganic Iodine Compounds
Iodates have the general formula MIO3 ,
Iodine forms binary compounds with hydrogen, where M is a univalent metal or electropositive
metals, other halogen elements, and oxygen. radical. Iodates can be prepared by dissolving

iodine in a caustic alkali solution and recover-
ing the iodate by filtration. Iodine can be oxi-
dized with concentrated nitric acid to iodic acid,
followed by neutralization with an oxide or hy-
droxide; oxidation of an alkaline iodide solution
with chlorine; or electrolytic oxidation of a neu-
tral iodide solution containing chromate with a
platinum anode and an iron cathode in a non-
diaphragm cell. Iodates are precipitated when
solutions of metal salts react with alkali-metal
iodates.
Most iodates are white. The solid metallic
iodates are stable and safe to handle in the pure
state, but should be kept out of contact with or-
ganic substances and other combustible materi-
als because such mixtures can be explosive. The
salts of the alkali metals and magnesium are sol-
uble in water.
Iodates and periodates can be determined by
addition of excess alkali iodide in an acid solu-
tion followed by titration of the liberated iodine
with thiosulfate. Periodates can be determined
in the presence of iodates by adding potassium
iodide to a solution buffered with borax and
boric acid, and titrating the liberated iodine with
sodium arsenite solution.
Hydrogen iodide [10034-85-2], HI,
M r 127.93, mp −50.9 ◦ C, bp −35.4 ◦ C, is a col-
orless gas which fumes in moist air. It is the
least stable of the common hydrogen halides,
and the best reducing agent. It cannot be pre-
pared using strong acids because it is readily
oxidized; therefore, it must be prepared from a
weak acid and iodine, or by the hydrolysis of
certain iodides.
Hydriodic acid is the colorless solution
formed when hydrogen iodide gas dissolves in
water, and at commercial strength, typically con-
tains 47 % hydrogen iodide. Hydriodic acid is
used in the preparation of iodides and is an im-
portant reagent in organic chemistry. Dilute hy-
driodic acid reacts with metals or their hydrox-
ides, carbonates, and other salts to yield metal
iodides.
Iodic acid [7782-68-5], HIO3 , M r 175.93,
forms colorless rhombohedral crystals, which
decompose at 110 ◦ C. It is prepared by oxidiz-
ing iodine with nitric acid. It is a strong oxidiz-
ing agent, oxidizing iodide to iodine, sulfite to
sulfate, and hydrogen sulfide to sulfur. It reacts
Iodine and Iodine Compounds 9
vigorously with dry carbon, phosphorus, and or-
ganic matter.
Polyiodides and polyhalides form when io-
dine or iodine halides combine with halide
salts. Some examples are RbI3 [12298-69-0],
[N (CH3 )4 I9 [3345-37-7], [N (C2 H5 )4 IBr2 ]
[20445-98-1], and KICl4 [14323-44-5]. Polyio-
dide and polyhalide salts usually form highly
colored orthorhombic crystals; however, KIF6
[20916-97-6] is white.
Iodine Halides. Seven iodine halides
are known: IF, IF3 , IF5 [7783-66-6],
IF7 [16921-96-3], I2 Cl6 [865-44-1], ICl
[7790-99-0], and IBr [7789-33-5]. All are read-
ily formed by direct reaction of the two halogen
elements.
10. Organic Iodine Compounds
As with inorganic iodides, organic iodides have
higher specific gravities and indices of refraction
than the corresponding chloro and bromo deriva-
tives, and higher melting and boiling points.
Compounds containing an iodine atom attached
to a primary carbon atom are the most stable.
Secondary compounds are less stable, and ter-
tiary are the least stable of all. Aliphatic iodo
compounds can be prepared by reacting alcohols
with a mixture of iodine and red phosphorus.
Other methods for preparing aliphatic iodo com-
pounds include: the addition of iodine halides or
iodine to olefins; the substitution reactions bet-
ween an alkyl bromide or chloride and an alkali
metal iodide; and the reaction of alcohols with
triphenyl phosphite and methyl iodide.
Organic iodo compounds are highly reactive.
The alkyl iodides are generally 50 – 100 times
more reactive than the corresponding chlorides.
Reactivity of unsaturated organic iodides is af-
fected little if iodine is attached to one of the
carbon atoms involved in the double bond, but
reactivity is promoted if iodine is attached to the
carbon atom in the α position to the double bond.
Iodine in organic compounds can be determined
after being converted into an inorganic form by
heating the compound with one of the following:
(1) fuming nitric acid in a sealed strong glass
tube, the Carius method; (2) sodium peroxide
10 Iodine and Iodine Compounds
in a combustion container; or (3) a mixture of
chromic and sulfuric acids, followed by addi-
tion of phosphoric acid and distillation into a
solution of sodium arsenite [37].
Iodoform [75-47-8], triiodomethane, CHI3 ,
mp 119 ◦ C, crystallizes from acetone as yellow
hexagonal plates, which have a characteristic
penetrating odor and are insoluble in water. The
compound can be prepared by the reaction of
iodine chloride with acetone in the presence of
potassium hydroxide, or by reaction of chloro-
form with methyl iodide. Iodoform is used as a
disinfectant.
Methylene iodide [75-11-6], CH2 I2 ,
M r 267.85, mp 5 ◦ C, bp 180 ◦ C (101.3 kPa), is
a colorless liquid when pure. It has a relative
density of 3.325 at 20 ◦ C and can be mixed with
benzene (d 20
20 0.879) to produce liquids with dif-
ferent densities. These mixtures are used in the
determination of the refractive indices of miner-
als and for the separation of minerals according
to density.
Ethyl iodide [75-03-6], C2 H5 I, M r 155.98,
bp 72.3 ◦ C, is a colorless liquid, which is com-
pletely miscible with ethanol and ether, but only
partially miscible with benzene. Ethyl iodide is
prepared by reaction of ethanol with iodine and
red phosphorus.
Methyl iodide [74-88-4], CH3 I, M r 141.95
mp −66.1 ◦ C, bp 42.5 ◦ C, is a colorless, trans-
parent liquid, which turns brown upon exposure
to light. Methyl iodide is prepared from methyl
alcohol, iodine, and red phosphorus. Methyl io-
dide is used as a methylating agent.
Iodobenzene, [591-50-4], M r 204.02,
mp −30 ◦ C, bp 188.7 ◦ C, is a colorless liquid
that rapidly turns yellow and has a characteristic
odor. Iodobenzene is prepared by diazotization
of aniline followed by treatment of the solution
with an aqueous solution of potassium iodide,
or by the action of nitric acid on a benzene and
iodine mixture.
11. Toxicology and Occupational
Health
Iodine is absorbed by the body and concen-
trated in the form of diiodotyrosine and triiodo-
thyronine, which form thyroxine that is stored
as thyroglobulin. Thyroxine is secreted by the
thyroid, enters the circulation, and is carried to
the peripheral tissues where it controls tissue
metabolism primarily through regulation of en-
zyme activities (→ Hormones, Chap. 2.1.). Io-
dine is essential to higher animals and humans.
A normal person requires about 75 mg of iodine
per year, which is usually consumed as iodized
salt that contains one part sodium or potassium
iodide to 100 000 parts of sodium chloride. Io-
dine deficiency is a major cause of goiter. Stud-
ies on chickens link autoimmune thyroiditis di-
rectly with consumption of large amounts of io-
dine [38]. These results are also applicable to
humans.
Iodine is safer to handle at ambient tem-
perature than chlorine or bromine because it
is a solid with a lower vapor pressure. When
handling properly packaged containers, usual
work clothes are sufficient. In the handling of
solid, unpacked iodine, rubber gloves, a rub-
ber apron, and safety goggles are recommended;
more drastic situations may necessitate the use
of a gas mask. The maximum safe concentration
for short-term exposure of up to 1 h is 1.0 ppm.
Exposure of the lungs and eyes can be irritat-
ing at concentrations of 0.1 ppm and should be
avoided. Exposure in concentrations higher than
0.1 ppm for extended periods causes severe irri-
tation to the eyes and the respiratory tract, and
may lead to pulmonary edema [39]. The mean
lethal oral dose for an adult is 2 – 4 g. Prolonged
contact with the skin may be harmful.
Treatment includes irrigating the eyes with
water or washing the contaminated body area
with a 5 % solution of sodium thiosulfate, fol-
lowed by saline catharsis. Iodine and concen-
trated solutions, including U.S.P. tincture, are
classed as poisons because of their irritating ef-
fects to the gastrointestinal tract. The gastric
treatment antidote is a 5 % solution of sodium
thiosulfate followed by saline catharsis. Medi-
cal attention should be provided.

In the United States, the Occupational Safety
and Health Administration (OSHA) has set the
maximum concentration of iodine vapor permit-
ted in the working atmosphere as 0.1 ppm [40].
The American Conference of Government
and Industrial Hygienists (ACGIH) established
0.1 ppm as the TLV (TWA) for iodine. The MAK
value is also 0.1 ppm. To detect iodine at this
concentration, air monitoring is necessary [41].
Empty containers may be destroyed in an in-
cinerator or decontaminated by washing with a
dilute sodium thiosulfate solution. Bulk waste
should be treated by controlled chemical opera-
tions to reclaim the iodine or convert it to harm-
less waste.
Iodine can be mixed safely with ammonia so-
lutions in the presence of a reactive organic sub-
stance; however, contact of gaseous ammonia
or its solutions with free iodine and its halogen
derivatives in any form can form explosive ni-
trogen triiodide [13444-85-4], NI3 . If this dark
brown to black solid forms, it should be dis-
carded while wet or destroyed by adding a re-
ducing agent, such as ethanol, or more of the
compound to be iodinated.
In general, iodides are acutely toxic only
when ingested in large amounts. Chronic inges-
tion or absorption through the skin may produce
iodism, which is manifested by a rash, nasal drip,
or headache. In severe cases, weakness, anemia,
and general depression may occur. Care should
be taken in handling organic iodine compounds.
In 1979 and 1986, nuclear accidents caused
the release of radioactive iodine, 131 I, into the
atmosphere. Ingestion of potassium iodide, KI,
can reduce the radiation dose to the thyroid gland
by up to 90 % if administered before or shortly
after exposure. A dosage of 130 mg was recom-
mended for adults. If administered 4 h after ex-
posure, KI can block 50 % of the uptake of 131 I.
The United States Federal Emergency Manage-
ment Agency provided guidelines to state and
local agencies in the United States regarding the
distribution of KI for use as a thyroidal block-
ing agent by the general public in the vicinity of
nuclear power plants [42]. Guidance on the use
of radioprotective substances was delegated to
the Food and Drug Administration (FDA). The
FDA recommended that KI be used by people
who are likely to receive more than 10 – 20 rad
to the thyroid [43].
Iodine and Iodine Compounds 11
12. References
General References
1. G. L. Clark (ed.): The Encyclopedia of
Chemistry, 2nd ed., Van Nostrand, New York
1966, pp. 556 – 559.
2. D. M. Considein et al. (ed.): Van Nostrand’s
Scientific Encyclopedia, 6th ed. Van Nostrand,
New York 1983, pp. 1550 – 1652.
3. McGraw-Hill Encyclopedia of Science &
Technology, 5th ed., vol. 7, McGraw-Hill,
New York 1982, pp. 321 – 325.
4. Kirk-Othmer, 3rd ed., 13, 649 – 676.
5. M. Windholz (ed.): The Merck Index, 9th ed.,
Merck & Co., Rahway, N.J., 1976.
6. C. A. Hampel: Encyclopedia of Chemical
Elements, Reinhold Book Corp., New York
1968.
Specific References
7. L. Jurd, Aust. J. Sci. Res., Ser. A: 2 (1949)
111 – 1169.
8. R. G. Shepherd, C. E. Fellows, J. Am. Chem.
Soc. 70 (1948) 157 – 160.
9. G. E. Ericksen, Geol. Surv. Prof. Pap. (U.S.)
(1981) 1188.
10. F. G. Sawyer, M. F. Ohman, F. E. Lusk, Ind.
Eng. Chem. Fundam. 41 (1949) no. 8,
1547 – 1552.
11. A. Gene Collins: Geochemistry of Oilfield
Waters, Elsevier Scientific Publishing Co.,
New York 1975, p. 228.
12. H. M. Cotten in K. S. Johnson, J. A. Russel
(eds.): “Iodine in Northwestern Oklahoma,”
Thirteenth Annual Forum on the Geology of
Industrial Minerals, Norman, OK 1978
pp. 89 – 94.
13. P. A. Lyday: Mineral Facts and Problems,
BuMines Bulletin 675, Washington 1985,
pp. 377 – 384.
14. O. Fukuta in W. J. Schlitt ed.: Japanese
Iodine–Geology and Geochemistry, Salts &
Brines 85, New York 1985, pp. 151 – 168.
15. R. W. von-Bemmelen: Geology of Indonesia,
vol. 2, The Hague, Netherlands 1949,
pp. 103 – 111.
16. P. A. Lyday, Indus. Min. Mag. 222, London
1986, pp. 65, 66, 69, 70, 75, 76.
17. R. D. Crozier, Min. Mag. 146 (1982) no. 4,
282 – 289.
18. World Mining Equipment, 10 (1986) no. 4,
40 – 44.
19. P. A. Lyday, J. Chem. Ed. 64 (1987) 152.
12 Iodine and Iodine Compounds
20. F. E. Paulik, Catal. Rev. 6 (1972) 49. R. P.
Lowry, A. Aguilo, Hydrocarbon Process. 53
(1974) no. 11, 103.
21. Kirk-Othmer, 3rd ed., 1, 135.
22. Shell Oil Co., US 3 106 590, 1963 (C. W.
Bittner); US 3 119 881, 1964 (R. L. Hodgson).
23. Phillips Petroleum Co. GB 920 244, 1963;
GB 931 440, 1963.
24. G and A Laboratories, Inc., US 2 299 577,
1942 (T. Hasselstrom, E. A. Brenhan).
25. Esso Research & Engineering, J. Chem. Eng.
78 (1971) 107.
26. R. R. Whitehead, Water Pollut. Control (Don
Mills Can.) 119 (1981) no. 12, 10 – 12.
27. West Chem. Products, Inc., US 4 271 149,
1981 (M. W. Winicor, M. Oberland).
28. GAF Corp., US 4 128 633, 1978 (D. H. Larenz,
E. P. Williams).
29. J. C. Morris et al., Ind. Eng. Chem. 45 (1953)
1013 – 1015.
30. M. S. Favero, C. H. Drake, Public Health Rep.
79 (1964) no. 3 251 – 257.
31. W. C. Thomas et al. Trans. Clin. Climatol.
Assoc. 90 (1978) 153 – 162.
32. E. Lindgren, Acta Radiol. Suppl. 362 (1980)
134.
33. Kirk-Othmer, 2nd ed., 22, 145 – 146.
34. Kirk-Othmer, 2nd ed., Supplement Volume,
964 – 973.
35. General Electric Co., US 3 228 880, 1966
(R. S. Owens, R. W. Roberts).
Ion Exchange Membranes → Membranes and Membrane Separations
36. R. F. Rolsten: Iodine Metals and Metal Iodide,
J. Wiley & Sons, New York 1961.
37. J. W. Thomas et al., Anal. Chem. 22 (1950)
726 – 747.
38. N. Bagchi et al., Science (Washington D.C.)
4723 (1985) 325 – 327.
39. N. I. Sax: 47th Dangerous Properties of
Industrial Materials, 4th ed., Van Nostrand
Reinhold Co., New York 1975, p. 32.
40. CFR, 29th, 1910.1000, Table Z – 1, July 1,
1986.
41. Encyclopaedia of International Occupational
Health and Safety,vol. 1 , International Labor
Office, Geneva 1971.
42. Federal Emergency Management Agency:
“Federal Policy on Distribution of Potassium
Iodide Around Nuclear Power Sites for Use as
a Thyroidal Blocking Agent,” Fed. Regist. 50
(1985, Jun. 20) no. 119, 25624 – 25625.
43. Federal Emergency Management Agency:
“Federal Policy on Distribution of Potassium
Iodide Around Nuclear Power Sites for Use as
a Thyroidal Blocking Agent,” Fed. Regist. 50
(1985, Jul. 24) no. 142, 30258 –30259.
44. Büchel, Moretto, Woditsch Industrial
Inorganic Chemistry Wiley-VCH, Weinheim
45. http://minerals.usgs.gov/minerals/pubs/
commodity/iodine
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

My Profile Log

HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry

Ullmann's Encyclopedia of Industrial Chemistry Recommend to Your BROWSE THIS TITLE

Librarian
Article Titles A–Z
Ion Beam Techniques Save title to My
Standard Article Profile Topics

Gerhard K. Wolf1 and Wolfgang Ensinger2 Email this page

1Universität Heidelberg, Physikalisch Chemisches Institut,
Print this page SEARCH THIS TITLE
Heidelberg, Germany
2Universität Marburg, Analytische und Kernchemie, Marburg,
Germany Advanced Product Search
Copyright © 2003 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights Search All Content
reserved. Acronym Finder
DOI: 10.1002/14356007.g14_g01
Article Online Posting Date: March 15, 2003

Abstract | Full Text: HTML

Abstract
The article contains sections titled:

1. Introduction
2. Interaction of Ions with Matter
3. Equipment for Ion Beam Generation and Application
3.1. Gaseous Ion Sources
3.2. Solid Sources
3.3. IBAD Facilities
4. Applications
4.1. Semiconductor Processing
4.2. Metal Modification
4.3. Ion Beam Modification of Insulators
4.3.1. Ion Implantation and Ion Bombardment
4.3.2. Metallization and Ion Beam Assisted Coatings
5. Outlook

[Top of Page]

1. Introduction
Energetic ions and ion beams are very flexible tools being used in material research, biology and medicine,
microelectronics and micromechanics, optics, analysis, and basic research. For most of these fields they provide a certain
breadth of applications. In some areas like microelectronics they also play the role of a key technology. The connection
and network formation of the ion technology with different areas of science, applications, and materials is shown
schematically in Figure 1. Special advantages of ion beams are the flexibility of size ranging from focused beams (≤ 1µm
to broad beams ≈ 1 m wide. The penetration depth of the ions in the material ranges roughly from 0.3 nm to 10 µm and
depends on their energy. Finally, the beam can consist of ions of nearly any element from hydrogen to bismuth. Because
of the possible energy variation from several electronvolts up to 109 eV, the differences in mass, and the specific chemical
properties of the ions, a huge number of variations exist to generate physical, chemical, or biological effects or analytically
useful signals. Ion beams represent an excellent and flexible tool for surface modification of small samples or for basic and
applied science applications. In industry they are used for semiconductor processing and for niche applications in different
materials finishing procedures. Treatment of large areas is under development but not yet ready for industrial use. Finally,
ion beams may be used for analysis of thin films and surfaces ( Surface and Thin-Film Analysis – Static Secondary Ion
Mass Spectrometry (SSIMS)).

Figure 1. Connection and network formation of ion beam technology with different areas of science, materials, and
application

page 1 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
In the following at first the basics of the interaction of ions with solid materials and the principles of the different ion beam
techniques will be described (Chap. Interaction of Ions with Matter). Then a short survey on the necessary equipment and
tools will be presented (Chap. Equipment for Ion Beam Generation and Application). In Section Semiconductor
Processing . the applications of ion beams for the processing of semiconductors are briefly reviewed. More details are
given in Semiconductors – Wafer Fabrication.

Sections Metal Modification and Ion Beam Modification of Insulators deal with the application of ion beam techniques in
the modification of metals and insulators, respectively, with special emphasis on features such as wear, friction, corrosion,
adhesion, conductivity, and the corresponding microstructural changes. In Chapter Outlook some ideas about future trends
will be presented.

For a general introduction into the ion beam technique see [1], [2], [3], [4], [5], [6] and the

z Proceedings of the biennial International Conference on Ion Beam Modification of Materials IBMM, published in
Nucl. Instr. Meth. B (Semiconductors, Insulators, Metals).
z Proceedings of the biennial International Conference on Ion Implantation Technology IIT, mainly published in Nucl.
Instr. Meth. B (Semiconductors).
z Proceedings of the biennial International Conference on Surface Modification of Metals by Ion Beam SMMIB,
published in Mat. Sci. Eng. or Surf. Coat Technol. (Metals, Insulators).
z Proceedings of the biennial International Conference on Radiation Effects in Insulators REI, mainly published in
Nucl. Instr. Meth. B (Insulators).

[Top of Page]

2. Interaction of Ions with Matter

An energetic ion beam impinging on a solid surface may lose its energy by a number of processes [1], [7]:

1. Electronic excitation of the target atoms (so called electronic stopping).

2. Elastic collisions with the nuclei or atoms of the targets (so called nuclear stopping).
3. Charge exchange and inelastic collisions.

In the energy regime relevant for the applications of ion beam techniques the processes (1) and (2) are of major
importance. In the electronic stopping regime, the ions (low mass and higher energy) travel straight through the
bombarded material, in the nuclear regime (lower energy, higher mass) the collisions lead to strong directional spread.

To give an example: An ( -particle (He ion) of 1 MeV energy loses ≈ 99.9 % of its energy by electronic stopping, an Xe
ion of 100 keV energy ≈ 90 % by nuclear stopping.

The straggling of the ions during the slowing down process is responsible for a theoretically Gaussian distribution of the
stopped ions in the bombarded material. The half width of the distribution as well as the range depend on the mass and
energy of the ions and on the target material. The range is usually expressed as projected range defined as the position of
the maximum of the Gaussian curve with respect to the target surface. Referring to the examples given, the projected
range of 1 MeV He+ in Al is 2.6 ± 0.12 µm and of 100 keV Xe+ only 40 ± 11 nm. The detailed values for most ions and
materials may be found in tables or can be calculated by the TRIM code (a computer code based on the LSS theory
performed by Lindhard, Scharff, and Schiøtt [10]) [8].

The two dominant stopping modes may cause a number of modifications in the bombarded materials.

– Chemical changes such as cleavage or formation of bonds in compounds.

– Heating of the target.
– Doping or alloying with the implanted ions.
– Generation of various types of defects.
– Enhanced diffusion (surface and bulk).
– Densification of the bombarded region.
– Phase changes and formation of special textures and surface morphologies.

Of special importance is the sputtering effect, namely the emission of surface atoms from the target as a consequence of
the nuclear stopping process near the surface [2]. Therefore, the sputtering rate is high for low to medium energy of the
beam, high ion mass, and low target binding energy. In some cases one impinging ion may sputter 10 – 100 target atoms.
Consequently the sputtering process can be used for surface ablation and surface cleaning.

The intensity of the interactions naturally increases with increasing ion dose or ion fluence, expressed as ions/cm2. Above
1012 ions/cm2 chemical changes and defect formation near the surface of the target can be effected. At fluences around
1015 ions/cm2, modifications of the crystal structure, surface morphology, and electrical properties takes place, and above
1015 ions/cm2 the mechanical surface properties are affected, substantial surface ablation and degradation occur, and the
doping and compound formation level approaches the percent regime [3], [9].

As compared with the impact of electrons and photons, ions show the following differences and advantages.

page 2 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
1. Ions interact much more strongly with the target than electrons and photons.
2. The interaction depth of ions is restricted to a few micrometers, however the limits of the interaction zone are sharp
and very well defined. There is nearly no depth straggling or interaction zone broadening as in the case of
electrons.
3. Ions can be focused to spots < 1 µm allowing direct nanostructuring of surfaces.
4. Ion beams are universal doping tools because any stable ion can form a beam and the energy may be varied from
several electronvolts to several megaelectronvolts.

Figure 2. maps schematically the most important basic applications of ion beams. The analytical applications mentioned in
the figure are treated in a different keyword ( Surface and Thin-Film Analysis – Static Secondary Ion Mass
Spectrometry (SSIMS)).

Figure 2. Schematic of the most important properties of the “tool” ion beam

Material modification with ion beams may either be performed by direct bombardment of the subsurface region or by
creating a new surface under bombardment by sputtering or by ion assisted deposition. The different techniques are
illustrated in Figure 3. [10]:

1. Ion implantation for modifying or doping a region of typically ≤ 1 µm below the surface.
2. Ion beam mixing for improving the interface between an already existing coating and the sample by intermixing and
densification (usually restricted to layers ≤ 1 µm thick). For these two techniques an ion implanter or ion source is
required that delivers ions of 100 keV – 1 MeV.
3. Ion beam assisted (enhanced) deposition (IBAD). An element or compound is deposited from the gas phase and
simultaneously bombarded with an ion beam. This improves the adhesion to the sublayer and densifies or modifies
the coating. By choosing the right deposition parameters the structure and texture of the growing film can be
affected in a controlled way as compared with pure evaporation. If an element is deposited and the ion beam
consists of reactive species or if a reactive gas is fed to the surface during bombardment the coating may be
transformed into a compound (Cr + N+ beam CrXN, Ti + N2 + Ar+ beam TiN; see also Nitrides – Boron
Nitride).

Figure 3. Different ion beam techniques: ion implantation, ion beam mixing and ion beam assisted deposition compared
with pure evaporation

For IBAD ion sources are required, that generate low- to medium-energy ions, and an apparatus for evaporating elements
and compounds. Some possible combinations are presented in Figure 4.

Figure 4. Possible combinations of depositing a coating under ion bombardment (IBAD)

In the case of a focused ion beam or by using masks, structuring or local treatment of a surface is possible as shown in
Figure 5. Ions of higher energy may also penetrate deeper into a material, as illustrated in the lower part of Figure 5. This
effect can be used for destruction of degenerated tissue for medical purposes or formation of deep implanted layers for
microelectronics applications.

Figure 5. Use of focussed and high energy ion beams for A) microstructuring ( -focussed treatment and alloying <
1 µm); B) formation of deep implants; C) destruction of local regions of tissue

page 3 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Techniques, which are comparable to ion implantation and IBAD, are plasma immersion ion implantation (PIII) and ion
plating [11]. In the case of plasma immersion implantation the sample is not bombarded by an ion beam but by ions
generated in a plasma. This is performed by giving the sample a pulsed negative bias of several 10 kV. Plasma immersion
implantation is less flexible than ion implantation, however, three-dimensional or shaped samples are easier to treat. The
advantage of IBAD over ion plating and comparable physical vapor deposition (PVD) techniques (see Thin Films –
Physical Vapor Deposition (PVD) Processes) is its lower working pressure and the precise control of the energy, direction
of incidence of the ions, and within certain limits also the treatment temperature. However, a disadvantage is the difficulty
to treat bigger and odd-shaped items.

[Top of Page]

3. Equipment for Ion Beam Generation and Application

Equipment for ion beam generation and application usually consists of a vacuum chamber, where the beam is formed, and
a target station. The probably most important part of an apparatus for ion beam modification is the ion source [4, 5, 12]. An
ion source is an ion reservoir from which ions are extracted and form a directed ion beam.

3.1. Gaseous Ion Sources

Here, the reservoir usually is a plasma which is confined in a certain volume. The plasma is sustained by ionized gaseous
species, preferably the ones which form the ion beam. However, there may also be an auxiliary gas forming the main part
of the plasma while the species of interest is only diluted in it. The most important part of a gaseous ion source is the
plasma generator which is basically a vacuum vessel filled with a gas to a certain pressure at a controlled flow rate by
means of a valve from an external gas reservoir. Figure 6. shows a schematic presentation of a typical ion source.

In order to ionize the gas atoms or molecules, energy has to be supplied to the gas. A number of processes can be used
for ionization. The most common one is electron impact. Electrons are accelerated in an electrical field to kinetic energies
high enough to overcome the first ionization potential of the gas particles when they encounter the particles. The electrons
can be supplied thermionically by a hot filament. A wire, often made of tungsten or tantalum, is electrically heated to a
temperature high enough to emit electrons. In an electrical circuit, this filament acts as cathode, while another part of the
ion source, often the vessel itself, acts as anode. Thus, an electric field is established in which the electrons are
accelerated. This is the principle of hot cathode ion sources.

In cold cathode ion sources, ions impinging on the cathode create secondary electrons which in turn are accelerated in the
electrical field and ionize the gas. Here, an externally induced ionizing process has to take place in order to ignite the
discharge. Often, cosmic rays are sufficient for this purpose.

Mostly, the plasma is confined in a magnetic field. The field elongates the path of the ionizing electrons by forcing them
into a spiral movement, thus increasing the number of collisions with gas particles and hence the number of ions.

Apart from electron impact, electromagnetic waves, e.g., radiofrequency (rf) waves can be used for ionization. The rf
power is coupled inductively or capacitively into the gas volume. The radiofrequency energy excites the particles which
become ionized in collisions. In microwave ion sources, another wavelength regime often used, electrons absorb the
microwave energy and become accelerated to energies high enough for ionization. Further methods of ionization are
surface ionization, field ionization, photoionization, spark ionization, etc.

Once ions have been created, they have to be formed to a directed beam. For this purpose, an extraction system is used.
An external electrode near the plasma boundary is negatively biased with respect to the plasma and accelerates ions out
of the plasma. The extraction electrode is supposed to extract the ions from the plasma, but not to collect them. For
collection of ions it contains an aperture. The ions move towards the electrode and through the aperture and then drift on.
Thus, an ion beam is formed.

For generating ions of gaseous species in the described way, a number of different ion sources are available. Apart from
the kind of ionization, there are other types of classifications which can be applied to ion sources. One is to distinguish
sources for narrow but dense ion beams from sources for broad beams, another one makes a distinction between ion
sources designed for accelerating the ions to high energies and those which emit low energy ions.

The most prominent source for broad beams of low energy ions is the Kaufman type ion source (see Fig. 6). It is a hot-
filament ion source with a large-volume plasma chamber where the ions are extracted by a multi-aperture system. Rather
than using an extraction electrode with a single drill-hole, as described below, a large-area grid is used. In this way, an ion
beam with a diameter up to several tens of a centimeter can be generated. Another type of ion source capable of creating
large area ion beams is the above mentioned radiofrequency-driven source. In contrast to the hot cathode ion sources,
which contain one or several filaments made of an oxidizable metal, an rf driven source may be operated with reactive
gases such as oxygen or hydrocarbons.

page 4 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Figure 6. Schematic cross-section of an ion source (hot cathode type for gaseous ions)

Most types of large-area ion sources require a multi-aperture (grid extraction) system. In the end-Hall ion sources, the
expensive extraction grids are replaced by a potential difference generated by a magnetic field and a strong electron
current.

As mentioned before, the beam shape is determined by the extraction system. Most broad-beam ion sources and focused
beam sources, both for low and high energy ions, are equipped with a circular extraction system. These sources emit a
circular symmetric ion beam with either a bell-shaped intensity distribution in cross-section of the beam, or with a beam
with a more or less flat plateau in the center. When flat elongated workpieces or coil material are to be processed,
elongated extraction systems for line-shaped ion beams up to several 10 cm in length are available. In this case, the
workpiece is moved through the ion beam. These systems, also called sheet or ribbon beam ion sources, may be scaled-
up to a length of 1 m or more.

Broad beam ion sources are used when the beam is directly applied to the workpiece to be modified. However, when the
beam has to be guided, swept, or manipulated, its diameter is restricted. In this case, different ion sources with narrow
extraction systems have to be used. These sources include the Penning ion gun (PIG) which is of the cold cathode type
and therefore capable of generating reactive ions, the Duoplasmatron ion source of the hot cathode type with a particularly
intense ion beam with a small diameter ( Surface and Thin-Film Analysis – Ion Sources), the hot-cathode CHORDIS
with a comparatively large volume and powerful beam, and the Freeman ion source. These ion sources have typically a
single aperture extractor which may be round or elongated. The Freeman source is usually equipped with a slot-shaped
extractor.

3.2. Solid Sources

With gaseous feeds, ions of solid materials can only be generated by a molecular species, e.g., B2H6, when boron ions
are required, or CO for generating carbon ions.

However, in some of the above mentioned ion sources also solid feed materials can be used, namely in sputtering sources
and in oven sources. A typical example for the latter is the Freeman ion source in the oven version. In the plasma vessel
an electrically heated crucible is placed, out of which solid material with a sufficiently high vapor pressure can be
vaporized. Metals such as silver or copper are thermally evaporated into a plasma of an auxiliary gas, mostly argon. Part
of the evaporated metal atoms are ionized and can be extracted together with ions of the gas. Another possibility is to use
an additional cathode made of the solid material to be ionized. Ions of the auxiliary gas are accelerated towards the
cathode, impinge onto it and sputter atoms into the plasma. There, again, the metal atoms are ionized and can be
extracted. In this way, ions of metals with a low vapor pressure, such as refractory metals like Cr or Ta, can be generated.
Sputtering devices are available in a number of source types, including Duoplasmatron and Penning ion source.

To form an ion beam from a solid without auxiliary gas plasma, metal vapor vacuum arc (MEVVA) ion sources may be
applied. Figure 7. shows a schematic cross-section. In this source, an electrical arc is drawn to a cathode. The cathode
material is vaporized and forms a high-intensity plasma. These ion sources may also be used for generating broad beams
when a broad beam extractor is used. MEVVA ion sources are usually operated in a pulsed mode, in contrast to other ion
sources which mostly work in the d.c. mode.

Figure 7. Schematic cross-section of a metal vapor vacuum arc ions source for metal ions

A problem with ion sources operating with auxiliary plasmas is that, as mentioned before, a large number of ions of the
auxiliary gas are extracted together with the desired ions of the solid material. Also a certain amount of ions from the ion
source construction material may be present as impurity. For this reason, the ion beam has to be purified [6]. This is
mostly achieved by means of a sector magnet working as a mass analyzer. When ions with a defined kinetic energy travel
through a magnetic field, they are separated according to their mass differences. Light ions are strongly deflected in the
magnetic field, while heavy ions follow a smaller deflection radius. By choosing a suitable magnetic field certain ionic
species can be selected and pass the magnet, while others impinge on metal plates inside the tube. Another possibility for
mass separation is to use a velocity filter based on the Wien principle.

Consequently, ion implantation facilities can be divided in those with mass separation and those without mass separation.
Figure 8. shows a schematic representation of a set-up without mass separation. Ions are extracted from a large-area ion
source and are accelerated towards the workpiece. For a uniform treatment, the workpiece usually has to be manipulated
in the ion beam, e.g., by x-y rotation or motion. In case of an implantation facility with mass separation, an ion beam-line

page 5 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
has to be established (see Fig. 9). The ion beam is extracted from the ion source and passes the magnet. After separation,
the desired ionic species is accelerated to high kinetic energies by applying a high voltage between the ion
source/separation magnet and ground. To this end the ion source is kept at a positive high voltage with respect to the
workpiece which is at ground level. Another possibility is to first accelerate the ions and then carry out the separation. In
both cases, a beam of highly energetic ions — with kinetic energies from typically several 10 keV up to several
megaelectronvolts, depending on the application — is generated. Usually, the ion beam-line contains beam-focusing and
deflecting elements such as quadrupoles, Einzel lenses, and electrostatic deflection plates. The ion beam can be focused
to a thin intense beam or defocused to a beam with a larger diameter. For a uniform treatment, the ion beam may be
electromagnetically or electrostatically scanned in x- and y-direction and/or the workpiece is moved.

Figure 8. Schematic presentation of broad-beam ion implantation facility (without mass separation)

Figure 9. Ion implantation facility with a beam line with mass-separation and ion beam guiding elements

The implantation dose is usually determined by means of measuring the ion current integrated over the process time. Ion
beam charge collection is done by a Faraday cup or directly on the conducting workpiece with suppression of secondary
electrons.

The workpiece holder is usually a metal plate which can be cooled by water or heated resistively or by IR radiation. Often,
it is mounted on a moving facility which allows x/y-motion or rotation by means of electromotors. For laterally defined ion
irradiation, usually a beam collimator plate with a round or rectangular aperture is mounted in front of it.

All the vessels and tubes are made of vacuum-compatible material, usually with stainless steel seal. For generating the
vacuum required for ion beam transport, different types of vacuum pumps such as turbomolecular, oil diffusion or
cryopumps, combined with forepumps such as rotary pumps are used. The residual gas pressure is measured by means
of Pirani, Penning, membrane, or other vacuum gauges.

3.3. IBAD Facilities

A facility for ion beam assisted deposition is in principle a set-up for broad beam ion implantation, combined with a vapor
deposition device (see Fig. 10). Mostly, electron beam evaporation is used because of the versatility and high power input.
Electron beam heating allows for vaporizing materials with high melting point and low vapor pressure, both elements such
as titanium or carbon, and compounds such as oxides or nitrides. In analogy to the above described sputter ion sources,
vaporization can also be achieved by using a secondary ion source which sputters material onto the workpiece (see Fig.
11).

Figure 10. Cross-section of an apparatus for ion beam assisted deposition with electron beam evaporator and ion source

Figure 11. Apparatus for ion beam assisted deposition with one ion source for sputter deposition and another one for ion
irradiation of the deposit

Apart from electron and ion irradiation, the materials can also be evaporated electrically by resistive or inductive heating,
or by irradiation with intense light (laser).

In IBAD facilities, in addition to gas pressure and ion flux measurement, the deposition rate has to be determined. This is

page 6 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
usually achieved by depositing the film onto microbalances of the quartz crystal monitor type, mounted close to the
workpiece, or by measuring optically the atomic vapor density over the crucible. Another possibility is to monitor a property
of the deposited film related to its thickness such as its resistivity. Ion and atom measurements define the ion to atom
arrival ratio (I/A), an important quantity for IBAD deposition.

In addition to simple IBAD facilities with one ion and one atom source, more sophisticated ones exist with several sources
for vapor and ions, these are used particularly when multi-component multi-layer films have to be deposited.

The process chambers for workpiece handling range from volumes of several 10 liters up to cubic meters. The
corresponding workpiece holders have sizes up to 1 m2 capable of handling workpieces up to several 10 kg load.

Facilities for ion beam treatment require specific safety considerations with respect to high voltage and electron induced X-
rays. Usually, interlocks and shieldings are used for preventing these hazards. Additionally, toxic feed materials for ion
source or evaporator may cause problems. Appropriate gas handling, exhaust and cleaning procedures and devices have
to be applied.

For details on ion sources and ion implantation facilities see [4-12, 6].

[Top of Page]

4. Applications
4.1. Semiconductor Processing ( Semiconductors – Wafer Fabrication)
Semiconductor processing is the by far most widespread and important industrial application of ion beam technology. For
fabrication of electronic elements the semiconducting material has to be doped, i.e., a small amount of an impurity has to
be introduced. Two methods for doping are used, one being diffusion where the dopant is introduced into the
semiconductor by thermal motion, the other one being ion implantation. The latter has become more and more important
and now plays a greater role. The unique features of ion implantation such as a very good reproducibility and versatility
and the possibility to circumvent to a certain extent the restrictions of thermodynamics in order to create uniformly doped
regions renders it an excellent tool for well defined doping. Nowadays, worldwide several thousand ion implanters are
being used for semiconductor processing. Without ion implantation, the present state of semiconductor technology and the
extremely widespread application of semiconductors would not have been reached.

By combination of different technological steps, with ion implantation being the doping step, microelectronic integrated
circuit chip technology (IC) is realized. IC manufacturing includes wafer fabrication, wafer probing and sorting, chip
assembly, chip testing and burn-in. Both passive electronic elements, namely resistors and capacitors and active circuit
devices such as transistors are fabricated in large numbers on a wafer by doping of selected areas by means of ion
implantation. The selected areas are produced in a multi-step process by lithography thus creating a desired pattern. The
parts which are left free by the mask are doped. The ion beam is directed towards the surface and forms the electronic
element in the desired shape and region of the wafer.

The development of IC technology included bipolar transistors and metal-oxide-silicon (MOS) transistors, later, for more
power efficiency, complementary MOS technology (CMOS) was established. With this technology, static or dynamic
random access memories (SRAMs or DRAMs) are fabricated.

Ion implantation is used for standard CMOS applications, including threshold voltage adjust implants, n- and p-well
implants, field isolation implants, source/drain implants, lightly doped drain implants, punchthrough stop implants,
polysilicon doping and others. Advanced applications include ultrashallow source/drain junction formation, channel and
drain optimization, trench capacitors and other three-dimensional structures. Figure 12 illustrates the implanted regions
and their functions.

Figure 12. Standard ion implantation applications for CMOS technology

Implanted Regions nd their Functions:

A = NMOS source/drain; basic transistor structure.
B = NMOS channel threshold voltage adjust; sets n-channel Vt (or Vp).
C = NMOS LDD; hot carrier suppression.
D = p-well (“tub”) structure; contains NMOS transistors.
E = p-type “channel stop” for p-well; intra-well (E) and inter-well (E') field isolation.
F = PMOS source/drain; basic transistor structure.
G = PMOS buried-channel threshold voltage adjust; sets p-channel Vt (or Vp).
H = PMOS “punchthrough” suppression.
I = n-well (“tub”) structure; contains PMOS transistors.
J = n-type “channel stop” for n-well; intra-well (J) and inter-well (J') field isolation.
K = NMOS “punchthrough” suppression.
L = PMOS LDD; hot carrier suppresion.
M = polysilicon gate doping (typically n+); improves conductance.

Another ion beam process for semiconductor fabrication is the silicon-on-insulator (SOI) technology. By means of ion
implantation at large dose, it is possible to create a silicon layer on top of an insulating silicon oxide layer. Another
possibility is a combination of ion implantation and wafer bonding. The surface of a wafer is oxidized. Then hydrogen is
implanted through the oxide and a defect layer is formed. Another wafer is bonded to the implanted one. The latter is cut
along the defect plane (smart cut process) by means of a heat treatment, forming a new surface of silicon on silicon oxide
which is annealed and polished. Thus, a monocrystalline silicon layer on an insulator is fabricated. The SOI material is

page 7 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
dedicated to low-voltage, low-power, high speed systems.

Detailed information on such applications for semiconductors may be found in many monographs [13], [14], articles [16]
and the proceedings of the international conferences on, e.g., Ion Beam Modification of Materials (IBMM) and Ion
Implantation Technology (IIT).

4.2. Metal Modification

Ion beam modification of metals began in the mid-1970s, mainly for improving wear and corrosion performance.
Advantages were seen in the low temperature and the nonequilibrium nature of the process. The conventional
metallurgical solubility limits could be exceeded with the possibility to create materials with improved properties.

In contrast to semiconductor processing, ion beam modification of metals has so far only found limited industrial
application. The main obstacle for this is the high cost of ion beam treatment. In contrast to most applications in
semiconductor processing, metal treatment requires high implantation doses. Additionally, ion beam treatment is in
competition with other techniques such as galvanic deposition, and chemical or physical vapor deposition, which yield
larger depths of modification or coating. In the following, a number of examples for modification of metals with ion beams
will be given.

Nitriding. The certainly most widely investigated application of ion beam treatment of metals is nitriding by nitrogen ion
implantation for increased wear life [17-22] ( Metals, Surface Treatment – Nitriding and Nitrocarburizing Methods). The
implantation doses typically range from 1 to 8×1017 N ions/cm2, the ion energies from 50 to 100 keV. In general, nitrogen
ion implantation leads to hardening. This may be due to different effects, the most important ones being pinning of
dislocations, and precipitation of hard phases. In general, nitrides are hard and increase the hardness of the metal in the
near-surface zone which, in turn, reduces wear. With ion implantation, nitriding may be performed at room temperature
which allows to treat temperature-sensitive metals. Typically, the temperatures can be held below 200 °C. At such
temperatures, no phase changes or segregations occur. No distortion of the workpiece is expected, neither does it change
its dimensions. Additionally, the surface finish is almost unchanged. Since no coating is applied, there is no adhesive
failure to be expected, even at high loads.

Nitriding has been carried out with a large number of alloys, including steels such as austenitic stainless steels and high
speed steels, aluminum alloys, titanium alloys such as Ti6Al4V, alloys of other transition metals, including chromium,
molybdenum, and tantalum, and hard metals such as tungsten carbide. The implanted nitrogen doses ranged from 1017 to
1018 N ions/cm2.

Wear tests under different conditions showed that despite the shallowness of the nitrided zone (well below 1 µm) good
results were obtained (see Table 1). This lead to several industrial applications, e.g., in the USA in biomedical applications,
where endoprostheses were commercially treated [17], [18]; or in Denmark, where tools for, e.g., plastics processing were
modified [19], [20], [24]. Reports on application-oriented studies from other countries such as China, Japan, and Spain
may be found in [20], [22], [25-27]. Ion beam processing of tools has been focused on precision tools to avoid alterations in
the dimensions or shape by coating or by heating; tools with a refined finish, i.e., highly polished tools; tools heat treated at
low temperatures; and very expensive tools because the risk of negative effects in processing is small.

Table 1. Applications of ion implantation in order to increase wear life of tools [17-22]

Application Life-time increase

Tools for piercing, stamping or punching metal 3–5 ×

e.g., manufacturing of beverage cans (punches) 5–10 ×
Transformation of plastics (moulds, nozzles, valves) 3–4 ×
e.g., extrusion dies for polymer fiber 3–4 ×
Blades for cutting food (vegetables, bread, meat) 4–5 ×
Blades for cutting paper, cardboards, rubber, and plastic 3–7 ×
e.g., blades for cutting synthetic fiber 6×
Drills for metal and polymers 2–3 ×
e.g., carbide twist drills for printed circuit boards 2–3 ×

So far, the modified zones have been quite shallow. Their thickness has mainly been governed by ballistic effects which
determine the travel distance of the ion in the metal. When the sample temperature is increased, either by an external
heating source or, more easily, by energy input via an intense ion beam, diffusion adds as a transport mechanism [28]. In
this case, even low ion energies may be used. Energies around 1 keV at ion current densities of several miliamperes per
square centimeter were applied. At temperatures of 400 °C, the range of the incorporated nitrogen in austenitic stainless
steel has been found to increase by an order of magnitude or more, beyond 10 µm, thus yielding a greatly increased wear
life. In the process, a metastable expanded austenite phase may form, another mechanism being the precipitation of CrN,
depending on the temperature. The hard nitride phase increases wear resistance, however, it deteriorates the corrosion
performance. By an appropriate selection of the process parameters, particularly the temperature, an optimum in wear and
corrosion performance can be obtained.

page 8 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Apart from nitrogen, carbon, oxygen, boron, and a number of metals ( Steel – Vacuum Vapor Deposition) have been
implanted for wear reduction.

Formation of Metastable Alloys. As mentioned initially, a particularly strong feature of ion implantation is that virtually
any element can be incorporated into any solid. This means that apart from nitrogen any other metalloid and any metal can
be implanted into any metallic target. In this way, a metal can be alloyed with any other metal even if they are not soluble
in each other. Thus, metastable alloys can be formed. An example is the alloying of iron with mercury, lead, or gold.

Amorphization. The formation of metastable phases is not only possible from a chemical, but also from a structural point
of view. Metals are well known for their crystallinity. Ion bombardment can be used to induce an instability in the crystalline
structure of metals, thus leading to amorphization [29]. Accumulation of radiation damage may force a metal into a
structure without long-range order. Glassy metals exhibit interesting properties, e.g., with respect to corrosion or
magnetism ( Metallic Glasses). In conventional techniques the glassy state is obtained by transferring the metal into
the liquid state and quenching it by rapidly cooling (splat cooling) the melt on a spinning wheel. This yields ribbons or films
of glassy metals. The structural disorder in the melt is frozen in to a maximum amount by using quenching rates as high as
possible. When the temperature is far below the recrystallization temperature, the amorphous state will be stable due to a
kinetic barrier. The interaction of energetic ions with matter does also yield high quenching rates and is therefore suitable
for forming glassy metals. In general, ion irradiation needs large enough ions for creating the required radiation damage
and low enough temperatures for preventing recrystallization. For fabricating thin films of amorphous metals, different
methods can be used. When intermetallics, e.g., Al2Au, AlMn, AlFe, Al3Ti, AuIn2, NiAl, NiZr, FeNd, FeCu, Ni2Al3, CuTi, or
Zr3Al, are irradiated at low temperatures with heavy rare gas ions, they accumulate radiation damage and eventually
become glassy. Another possibility is to implant metalloids such as P or B which stabilize the glassy state. Rather than
bombarding alloys with ions, also multilayers can be irradiated.

Films and Coatings by Ion Beam Mixing. In the case of ion implantation, the base material is always a component of the
modified zone. If an alloy or compound of different materials is needed as coating, ion beam mixing can be used. The
deposited film is modified by ion irradiation, either structurally or in addition chemically. Thus, a metal can be coated with
another metal or a nonmetal which is converted by ion implantation into a compound, e.g., by depositing a chromium film
on steel, and implanting nitrogen ions into the film, thus creating a chromium nitride coating. Apart from a chemical
change, the microstructural changes induced by ion irradiation play an important role. Prior to ion irradiation the features of
the coating are defined by the deposition process. Among these are, e.g., the grain size and the crystal orientation, or, as
a result of the microstructure, the porosity and the state of stress. These features can be affected by ion irradiation. By a
radiation induced lattice rearrangement, the crystal orientation may be altered. Ion irradiation usually induces a preferential
crystallographic orientation. By grain boundary motion, often a change in grain size is observed. The stress state can be
influenced. Microporosity, often a consequence of open grain boundaries, can be reduced. In this way, properties such as
hardness or corrosion resistance can be influenced. Apart from the film itself its interface to the substrate will be modified
when ions travel through it. Figure 13. shows a schematic representation. Chemical bonds will be broken and formed.
Impurities can be “diluted”, a mixed zone can be created both by ballistic mixing and by radiation enhanced diffusion.
Depending on the combination film-substrate-ion a phase can be generated (interphase). Interface stress can be reduced.
These effects may influence important properties of the film-substrate system, such as its porosity or its ohmic resistance.
The probably most important property is the adhesion between film and substrate. The above mentioned effects usually
lead to substantial enhancement of adherence between the film and the metal substrate.

Figure 13. Ion beam mixing of a coating on a substrate, forming a transition zone (gradient) or an interface. Both may
enhance adhesion.

As ion beam mixing is essentially also limited by the range of the ions, it suffers from similar restrictions with respect to
thickness of the modified zone as ion implantation. For metal modification often thicker modified zones are needed, e.g.,
for corrosion protection. The problem with shallowness of the modified zone can be overcome when thin films are
deposited from the vapor phase under simultaneous rather than postdeposition bombardment with energetic ions (IBAD).
Simultaneous ion irradiation or implantation influences the film in a way similar to ion beam mixing. In situ ion
bombardment is particularly powerful with respect to film densification, preferential crystallographic orientation, grain size
and morphology, and stress [30-32]. Thin films usually suffer from stress, either due to a mismatch in the thermal
expansion coefficient or intrinsic stress, which arises from defects, impurities etc. Often, the stress state of physical vapor
deposited films is tensile. By means of simultaneous ion irradiation, it may be reduced or changed to compressive stress.
Figure 14. shows a typical dependence of the stress state on the I/A-ratio. The stress state changes from tensile to
compressive, with a state of zero stress in the transition. This allows to deposit films with low stress levels, thus enhancing
adhesion.

Figure 14. General dependence of stress of thin IBAD films on the ion to atom arrival ratio (I/A)

IBAD coatings range in their thickness from 0.1 µm up to 10 µm with a typical thickness of 1 to 2 µm. The coatings can be
divided in elemental or compound layers and in single layers (monolithic) and multilayers. Films of a single element consist
mostly of metals such as Al or transition metals like Ti, Cr or Ta, nonmetals such as Si, B, or, mostly, carbon. The latter is

page 9 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
usually an amorphous material with varying ratio in sp2/sp3 hybridization, sometimes including hydrogen. Among the
compound films, the nitrides play a major role. TiN, ZrN, CrN, TaN and other transition metal nitrides, AlN from the main
group metals, and nonmetal nitrides, particularly Si3N4, CNx, and BN, especially the hard metastable cubic modifications,
have been deposited on many kinds of metal substrates. All these can comparatively easily be formed by evaporating the
element and irradiating the growing film with nitrogen ions. Another group of compounds are the oxides, such as Al2O3,
SiO2, TiO2, or Ta2O5. When the element is highly reactive, the film can be formed by nonmetal ion implantation and
reactions of the growing film with gas from the atmosphere. In this case, both reactive ions (nitrogen or oxygen) or rare gas
ions, mostly argon, can be used. The most important and typical examples of nitride and oxide films on metals formed by
reactive ion beam assisted deposition are TiN and Al2O3. When the element is not reactive enough, particularly in case of
the main group elements, the nitrogen or oxygen has to be incorporated by ion implantation. Here, the I/A ratio directly
determines the composition. Examples are AlN, SiO2, Si3N4 and BN. While the metallically bonded transition metal nitrides
can be grown crystalline over a wide temperature range, the covalent oxides and nitrides of the main group elements tend
to grow amorphous.

Microlaminates. When different materials are deposited in alternating order or the ion beam is modulated, multilayers can
be formed. These exhibit often considerably better features then single films of either of the materials, such as enhanced
elasticity and hardness. Particularly, their mechanical strength can be greatly improved. Such so-called microlaminates
have a great stability both against crack initiation and propagation. A typical examples is the system Al-Al2O3. The
microlaminate shows a much lower crack initiation and propagation probability than either of the monolithic layers.

Wear protection. Most of the nitrides and oxides mentioned above are harder than metals. Therefore, they may serve as
protection against wear. The most prominent of the nitrides is certainly TiN. It may be deposited by evaporation of titanium
either under nitrogen ion bombardment or under rare gas, usually Ar, irradiation in an atmosphere of backfilled nitrogen.
Depending on the process parameters titanium nitride may have a broad range in composition and also in hardness. Both
comparatively soft and ductile films may be obtained with hardness well below that of bulk TiN (20 GN/m2), but also a
quality with considerably larger hardness can be obtained, where the films usually have intrinsic compressive stress.
Friction and wear tests of TiN films on different substrates, including stainless steel, aluminum alloys, but also tungsten
carbide, gave reductions both in friction coefficient and wear. Figure 15. shows the dependence of hardness and wear life
of TiN formed by reactive IBAD on the Ar+/Ti arrival ratio.

Figure 15. Dependence of hardness (left scale) and wear life (right scale, given as number n of passes of the wear test)
of TiN formed by reactive IBAD on the Ar+/Ti arrival ratio [32]

Of particular interest is the group of (potentially) ultrahard materials, namely diamond-like carbon (DCL) and cubic boron
nitride (c-BN). Both can be grown by ion beam assisted deposition. Thin DLC films are often comparatively elastic and
exhibit particularly low friction coefficients. Cubic boron nitride films show similar features and may improve the tribological
performance of metals. However, the process window is comparatively narrow, and the films suffer from large compressive
stresses which may cause adhesion problems.

Lubricant Coatings. Apart from hard coatings, solid lubricants can be used. Molybdenum sulfide (MoS2) grown either as
monolithic film or as multilayer in combination with hard coatings such as B, B4C, or BN, show very low friction coefficients.

Corrosion Resistant Alloys. Metals may easily be alloyed by ion implanting an alloying element which leads to a more
corrosion resistant alloy [33], [34]. An example is the implantation of chromium into iron which, if the implantation
concentration exceeds 12 at%, becomes corrosion resistant in neutral aqueous media similar to stainless steels. In the
literature, a large number of examples of ion implantation into various metals and alloys, including alloys based on iron,
aluminum, titanium, nickel, and others are described. The implanted species were all kinds of metals and nonmetals
including noble metals such as platinum or palladium, and metalloids such as phosphorus and boron. The layers formed
protect against uniform corrosion, localized corrosion such as pitting and crevice corrosion, and hydrogen embrittlement,
which may occur when metals are exposed to water, acids, or salt brines. However, the effect is often not long-lasting
because the modified zone below the surface is too shallow. When corrosion progresses, the thin modified zone may be
consumed in a short time, ranging from a few days to weeks. For particular cases a long-term protection can be achieved.
The cathodic protection offered by implanting catalytically active noble metals such as platinum is very effective for some
metals with active/passive transition behavior. These metals such as titanium are protected by a native oxide which leads
to passive corrosion with low dissolution rates. In aggressive environment, the oxide may be dissolved with the result that
active corrosion with high rates starts. The noble metal shifts the potential into a region where the oxide film is recovered
and the metal remains passive. When the implanted protective material itself is not consumed by corrosion, it may remain
on the metal surface and exert a long-term protection effect. In this way, titanium can be protected against corrosion in
mineral acids or salt brines, and tantalum can be protected against corrosion and hydrogen embrittlement in hot
concentrated mineral acids.

When the implanted zones are too thin and/or the effect is not long-lasting, ion beam assisted deposition of thicker films
may be applied. Usually, a coating more corrosion resistant than the base material is used. It may consist of nonmetals
such as boron, silicon, or carbon, resistant metals such as noble metals, but also aluminum, titanium or tantalum, or
nitrides, carbides and oxides. It turns out that the most important characteristics of corrosion-resistant thin films deposited
under ion bombardment are low microporosity and high adhesion [34]. These features can be optimized by proper
selection of the main deposition process parameters which are I/A ratio, ion energy, and ion incidence angle. In the

page 10 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
beginning of film growth, ion beam mixing effects lead to a strengthening of the interface and to enhanced adhesion.
During film growth, ion bombardment increases the density of the film material and closes micropores. Often, the closing of
pores is enhanced when an off-normal ion incidence angle is chosen.

Again, multilayer films may be more effective than monolithic films. Figure 16. shows the iron dissolution currents of an
iron sample coated with a Ti/TiN multilayer sequence. The multilayers are less porous than the corresponding single films
of either Ti or TiN. It can be assumed that monolithic films contain pores from the substrate through the film to the surface.
TiN grows in columns parallel to the surface normal. Between the columns, more or less open channels may exist. In a
multilayer film with a sequence of a cubic nitride and a hexagonal metal crystalline structure, such continuous open
structures are less likely to form.

Figure 16. Critical current densities deduced from current-potential plots for Ti/TiN multilayers deposited by IBAD on steel
vs numbers of potential cycles

( ) CK45 steel, (*) Ti, (+) TiN, (♦) Ti/TiN ratio: 100/25, ( ) ratio: 100/50, ( ) ratio: 100/100, ( ) ratio: 50/100, ( )
ratio: 25/100 [36]

Not only aqueous corrosion, but also high temperature corrosion in gases can be reduced. Coatings such as Si3N4 are
effective in reducing oxidation of metals such as Ni and Cr and metallic alloys.

Films with Special Electric and Magnetic Properties. Apart from mechanical and chemical features, electric and
magnetic properties can be influenced. For metal films, a strong dependence of the electric resistivity on ion energy, I/A-
ratio, and type of metal has been found [35]. Magnetic properties such as coercivity can be influenced and films with giant
magnetoresistance can be formed.

4.3. Ion Beam Modification of Insulators

The effects of ion beams in insulators are somewhat different from those in metals because:
– The electronic stopping of ions causes much more changes in insulators than in metals.
– Insulators are mostly compounds, which means that the incoming ions interact with chemical bonds during slowing
down.
– The bombardment of insulators with ions may cause strong charging of the samples, an effect which has to be taken
into account during the experiment.

The application of ion beams for insulator modification can be grouped in two categories:

1. Ion implantation and ion bombardment.

2. Metallization and ion beam assisted coatings.

The most important materials treated in practice are different types of polymers, engineering ceramics, and glasses.

4.3.1. Ion Implantation and Ion Bombardment

Polymers. Most polymers are not very stable to radiation, which means that already moderate fluences of heavy energetic
ions may partly destroy the polymer structure. Figure 17. indicates the chemical changes caused by the ions with
increasing dose and the corresponding variations of properties. In the low dose range mainly properties such as solubility
or wettability are affected, in the higher dose range — near the region where graphite is formed — also hardness,
conductivity, and optical properties are modified.

Figure 17. Dose dependence of the chemical and physical properties of polymers and the corresponding structural
changes

Generation of Local Conductivity. Many researchers tried to induce conductivity in polymers, because this is an interesting
possibility to obtain locally conductive regions in an insulating material, for example for O-electronics applications. Figure
18. illustrates the variations of surface conductivity of the photoresistant polymer HPR-204 as function of the impinging
dose of 2 MeV Ar+ ions. The conductivity increases by four to five orders of magnitude and reaches the values of ion
bombarded amorphous carbon or disordered metals [37].

Figure 18. Resistivity changes of 2 MeV Ar+ ion bombarded polymers (HPR-204) as function of ion dose in comparison
with amorphous carbon

page 11 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience

Another possibility for obtaining local conductivity is the irradiation of intrinsically conductive polymers such as polypyrrole
or polythiophene. In this case the irradiation destroys the conductive properties, and, if done through a mask, leaves only
the shaded areas conductive. Figure 19. shows an example of such a structure with subsequent copper deposition on the
conductive areas [38], [39].

Figure 19. Galvanic deposition of copper on the shaded areas of a conducting polymer bombarded with argon ions
through a mask (size of the structures 13.5 µm)

Production of Microfilters. Microfilters can be produced by high energy ion beams (typically 100 MeV). The stopping track
of the ions in polymers can be etched by alkaline solutions into very small channels suitable for special filtering or
separation problems [40].

Glass. In the case of glass ion beam techniques have mainly been used to change its optical properties. These alterations
range from the implantation of color centers (rare earth ions, for example) with the aim of flat screen production to the
generation of wave guides for light. In the latter field a number of groups are working. The basic principle is the local
implantation of metals which are not soluble in the glass and thus form small precipitates, causing local changes in the
reflectivity of the materials [41], [42].

In addition ion beams are also used for polishing or machining high precision optical lenses in the nanometer-regime by
sputtering.

Ceramics. Here the improvement of the mechanical surface properties like hardness, friction, and cracking behavior were
the main subjects of application oriented studies. Especially nitrogen implantation in hard metals and in silicon nitride
showed a considerable potential for crack reduction and the formation of intermediate solid lubrication layers [43], [44].

4.3.2. Metallization and Ion Beam Assisted Coatings

Ion beam mixing and ion beam assisted deposition are excellent methods for metallizing polymers, ceramics, and glass. In
addition also compound layers or multilayers can be deposited on these substrates.

The advantage of ion beam techniques in comparison with pure evaporation is the greatly improved adhesion, and the low
processing temperature, in comparison with PVD methods. The latter is especially important for the polymer case.

Metallizing of Polymers. Metallized polymers find broad applications in microelectronics as base materials for circuit
boards, for electronic shielding in several industrial branches and for decorative and package purpose. Well established
techniques are galvanic deposition and vacuum evaporation with or without prior plasma cleaning and roughening.
Because of the low processing temperature and because they are not entailed with surface roughening, ion beam
techniques can be used in cases were very smooth surfaces are needed, or for polymers which are difficult to metallize,
such as high temperature thermoplastics [45], [46].

The factors affecting adhesion of metal layers on substrates are compiled in Figure 20. Most of these factors can be
modified by ion bombardment, especially the film stress and the chemistry and morphology of the interface. This is
illustrated in Figure 21. were the adhesion of thin copper layers on polyethersulfone (PES), measured by the peel test, is
plotted as a function of the mixing ion species and the ion energy [47]. Before peeling the original Cu-layers have to be
thickened to 20 µm by galvanic means.

Figure 20. Schematic representation of the factors affecting the adhesion of metal films on polymer or ceramic substrates

Figure 21. Adhesion of copper deposited on polyethersulfone (PES) after bombardment with different ion species
measured by the peel test (ion dose 1014/cm2; ion energy 60 keV or 120 keV; PVD: deposition without bombardment)

page 12 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
A layer deposited by vacuum evaporation shows initially no adhesion, upon bombardment the adhesion first increases with
increasing electronic energy loss of the ion beam and goes through a maximum. The decrease after the maximum is
caused by the increasing importance of nuclear energy loss in the higher mass region, leading to a destruction of the
polymer skeleton. Similar effects have also been observed using ions with higher energy in the range of a few 100 keV–
10 GeV [48]. The IBAD technique can also be used for improving the metal adhesion on polymers. The obtainable effects
are very much dependent on the chemical composition of the polymer. In the case of polyethersulfone only very small ion
doses (expressed in terms of the ion to atom arrival ratio I/A) of argon ions in the energy range of 2.5 – 5 keV or higher
doses of low energy ions of 0.5 – 1 keV lead to considerable improvements. In Figure 22. the adhesion force, measured by
the pull off test, is plotted as function of I/A for several energies. Interestingly, especially polyphenylenesulfide (PPS) is
rather difficult to metallize with the established techniques. In some cases also a Ti intermediate layer improves the
situation.

Figure 22. Adhesion force of a copper layer deposited on polyphenylensulfide (PPS) by ion beam assisted deposition as
function of the ion/atom arrival ratio, measured by the pull off test (parameter is the argon ion energy)

(♦) 0.5 keV; (|) 1 keV ( ); 2.5 keV; ( ) 5 keV; ( ) 7.5 keV; ( ) 10 keV

Metallization by IBAD has been used also for medical purposes. Polymer tubes which are applied for “in body” therapy of
kidney diseases have been coated with thin silver layers of excellent adhesion. Silver has bactericidal properties and thus
helps to avoid unwanted inflammations.

Metallization of Ceramics. Only little work has been done on metallization of ceramics by ion beam mixing or IBAD. One
example for applications is again the formation of conductive layers on an insulating substrate, another one are adhesive
coatings with low friction coefficient. Si3N4 and Y stabilized ZrO2 (PSZ) were coated with a number of metals and
subsequently irradiated with 140 keV Ar+ ions.

Under the running conditions of a Diesel engine the friction coefficient was reduced by nearly one order of magnitude for
the case of a Ni–Ti double layer [49].

[Top of Page]

5. Outlook

At present the cost of non mass analyzed ion implantation (nitrogen for example) is ≈ 0.20 € for an implanted dose of 1015
ions/cm2 and ≈ 0.70 – 1 € for 1017 ions/cm2. Metal ion implantations with the MEVVA source was estimated in the USA to
range around 0.10 €/cm2.

These figures show that modifications requiring low ion doses, like in the case of polymers, are relatively cheap to perform.
Also the PIII technique is somewhat less expensive than ion beam implantation. Ion beam assisted deposition is already
used routinely in the optical industry and for small series or development work in many other industrial fields. The costs of
the modification depend on one hand on the investment, which is usually 10–30 % higher than that for a conventional PVD
plant. Secondly the beam power of the ion source and the thickness of the coating determine the cost per square
centimeter. A 3–5 µm thick TiN layer may be realized at present for 1–2 €/cm2. With an optimized dedicated machine
0.25 €/cm2 should be no problem.

At present plants with continuous coil coating or large area capability using powerful slit extraction sources are under
development [50]. With such plants cost reductions may be realized, especially for cases where the ion beam is only used
for precleaning by sputtering or for the production of thin interface or barrier coatings. The thickening of the coating is then
performed by pure evaporation.

The future trends for applications of ion beam techniques are on the one hand dependent on the further development of
high power broad beam sources and continuous working plants, on the other hand on the future requirements for high-tech
coatings. Fields of high potential for special surfaces and coatings are:
– Implants and materials for medical and biological use like internal prostheses, screws, fixations, infusion tubes,
catheters etc.
– UV- and X-ray optics, specially surface finishing of lenses by sputtering with nanometer precision.
– Texturing and metallizing of polymers and ceramics for circuit boards and sensors.
– Coatings and nitriding of highly temperature sensitive materials like Al, Mg, and their alloys.
– Surface modification and machining of micromechanical devices and creation of nanostructures by using mask
techniques or highly focused ion beams.

[Top of Page]

References
General References
1. G. K. Wolf: “Chemical Effects of Ion Bombardment” in: Topics in Current Chemistry, vol. 85. (1979), p. 1.
2. G. Carter, J. S. Colligon: “Ion Bombardment of Solids, Chapter Sputtering”, Heinemann Educational Books, London

page 13 of 15
Ion Beam Techniques : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
1968, p. 310.
3. H. Oechsner, J. Walldorf, G. K. Wolf: Teilchenstrahlgestützte Verfahren in G. Kienel, K. Röll (eds.):
Vakuumbeschichtung 2, VDI-Verlag Düsseldorf, 1995, p 193.
4. J. J. Cuomo, S. M. Rossnagel, H. R. Kaufman (eds.): Handbook of Ion Beam Processing Technology, Noyes, Park
Ridge 1989.
5. J. F. Ziegler (ed.): Handbook of Ion Implantation Technology, North-Holland, Amsterdam 1992.
6. B. H. Wolf (ed.): Handbook of Ion Sources, CRC Press, Boca Raton 1995.
Specific References
7. J. Lindhard, J. M. Scharff, H. E. Schiott, Mat.-Fys. Medd.-K. Dan. Vidensk. Selsk. 33, (1963) 14.
8. J. F. Ziegler, J. P. Biersack, U. Littmark: The Stopping and Range of Ions in Solids, Pergamon Press, New York 1985.
9. G. K. Wolf: Ionenstrahltechnologie – Werkzeug für das 21. Jahrhundert, Vakuum in Forschung und Praxis 3 (1996)
143.
10. G. K. Wolf: “Erzeugung verschleißmindernder Randschichten durch Ionenstrahlverfahren”, Materialwis. Werkstofftech.
24 (1993) 109.
11. J. R. Conrad, “Plasma Source Ion Implantation: A New Approach to Ion Beam Modification of Materials”, Mater. Sci.
Eng. A 116 (1989) 197.
12. W. Ensinger, “Ion Sources for Ion Beam Assisted Deposition of Thin Films and Coatings”, Rev. Sci. Instrum. 63
(1992) 5217.
13. W. S. Ruska: Microelectronic Processing: an Introduction onto the Manufacture of Integrated Circuits, McGraw Hill,
New York, 1987.
14. D. J. Elliott: Integrated Circuit Fabrication Technology, McGraw Hill, New York, 1989.
15. T. Seidel, “MOS Process Review with Implications for Ion Implantation”, Nucl. Instr. Methods B21 (1987) 96.
16. E. Chason et al., “Ion Beams in Silicon Processing and Characterisation”, J. Appl. Phys. 81 (1997) 6513.
17. P. Sioshansi, “Surface Modification of Industrial Components by Ion Implantation”, Mater. Sci. Eng. 90 (1987) 373–
383.
18. J. K. Hirvonen: “Ion Beam Processing for Industrial Applications”, Mater. Sci. Eng. A116 (1989) 167–175.
19. B. Torp et al.: “A New Industrial Ion Implanter for Surface Modification of Metals”, Nucl. Instr. Meth. Phys. Res.
B127/128 (1997) 940–944.
20. G. I. Zhang: “Present Status of Ion Implantation in Non-Semiconductor Materials in China”, Mater. Sci. Eng. A115
(1989) 377.
21. M. Geng et al.: “Modification Technology of N+ Implantation of Carbide Twist Drills Used in Integrated Circuit Boards
and its Industrial Application”, Surf. Coat. Technol. 51 (1992) 551–555.
22. H. Li, C. Ji: “Surface Modification of Non-Semiconductors by Ion Beams in China”, Surf. Coat. Technol. 65 (1994)
180–193.
23. C. A. Straede, N. J. Mikkelsen: “Implementation of Ion Implantation in European Industry”, Surf. Coat. Technol. 84
(1996) 567–573.
24. C. A. Straede, N. J. Mikkelsen: “Industrial Implementation of Ion Implantation on Tools and Wear Parts Based on a
Strategic Approach”, Surf. Coat. Technol. 103–104 (1998) 191–194.
25. M. Iwaki: “Ion Implantation in Japan in Non-Semiconductor Fields”, Mater. Sci. Eng. A115 (1989) 369–376.
26. R. J. Rodriguez, A. L. Sanz, A. M. Medrano: “The Search for New Applications of Ion Implantation Treatments”, Surf.
Coat. Technol. 84 (1996) 594–599.
27. J. I. Onate, F. Alonso, I. Braceras, A. L. Sanz, R. J. Rodriguez: “Current Status of Commercial Ion Implantation in
Spain”, Surf. Coat. Technol. 103–104 (1998) 185–190.
28. R. Wei: “Low Energy, High Current Density Ion Implantation of Materials at Elevated Temperatures for Tribological
Applications”, Surf. Coat. Technol. 83 (1996) 218–227.
29. P. M. Ossi: “Ion Induced Crystal-to-Glass Transition in Alloys”, Mater. Sci. Eng. A115 (1989) 107-121.
30. F. A. Smidt, G. K. Hubler: “Recent Advances in Ion Beam Modification of Metals”, Nucl. Instr. Meth. Phys. Res.
B80/81 (1993) 207–216.
31. J. K. Hirvonen: “Current Topics of Ion Beam R&D”, Surf. Coat. Technol. 65 (1994) 84–89.
32. W. Ensinger: “Ion Bombardment Effects during Deposition of Metal and Nitride Films”, Surf. Coat. Technol. 99 (1998)
1.
33. G. K. Wolf: “A Historical Perspective of Ion Bombardment Research for Corrosion Studies”, Surf. Coat. Technol. 83
(1996) 1–9.
34. W. Ensinger, “The Influence of Ion Irradiation during Film Growth on the Chemical Stability of Film/Substrate
Systems”, Surf. Coat. Technol. 80 (1996) 35.
35. R. Roy: “Low-Energy Ion Assisted Deposition of Metal Films”, Surf. Coat. Technol. 51 (1992) 203–211.
36. R. Hübler, A. Schröer, W. Ensinger, G. K. Wolf, W. H. Schreiner, I. J. R. Baumvol: “Plasma and Ion – Beam –
Assisted Deposition of Multilayers”, Surface and Coatings Technol. 60 (1993) 561.
37. T. Venkatesan: “High Energy Ion Beam Modification on Polymer Films”, Nucl. Instr. Methods B. 7/8 (1985) 461.
38. S. Schiestel, P. Banniza, G. K. Wolf, K. Edinger: “Modification of Intrinsically Conducting Polymers by Ion
Implantation”, Nucl. Instr. Methods B. 116 (1996) 164.
39. T. Asmus, G. K. Wolf: “Modification and Structuring of Conducting Polymer Films on Insulating Substrates by Ion
Beam Treatment”, Nucl. Instr. Methods B 166/167 (2000) 732.
40. R. Sudowe, P. Vater, R. Brandt, J. Vetter, W. Ensinger: “Filters with < 100 nm Radius Pores for Gas Separation

page 14 of 15