Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
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]
The slightly endothermic dehydrocondensa-
tion (∆H = + 8 kJ/mol [187]) of benzene to bi-
t, ◦ C 100 200 300 350
phenyl and terphenyl has been investigated in-
Vapor pressure, kPa 25.43 246.77 558.06 tensively [188, 189]. The formation of biphenyl
Liquid density, g/cm3 0.97 0.889 0.801 0.751 from benzene starts at ca. 480 ◦ C and increases
Heat capacity, J/g 1.786 2.129 2.468 2.640
with increasing temperature and residence time.
Industrially, the reaction is carried out in a
multitube reactor, heated electrically [190] or
Chemical Properties. Biphenyl is a very sta-
by heat exchanger, at 700 – 850 ◦ C using resi-
ble organic compound; in an inert atmosphere
dence times of the order of seconds. An optimal
it remains unchanged even at a temperature
tube length of 45 – 75 m and diameter of 100 –
>300 ◦ C. The compound sublimes, distills with
120 mm have been reported [191–194]. None
steam, and undergoes various substitution reac-
of the published procedures work in equilib-
tions; the reactivity is similar to, but less than that
rium because carbon black formation increases
of benzene. Substitution reactions, e.g., chlori-
with increasing temperature and longer resi-
nation, nitration, and sulfonation, occur at the
dence time.
2- and 4-positions, the latter being the preferred
The preferred residence time is ca. 10 s,
site of attack in Friedel – Crafts reactions.
which results in a benzene conversion of ca.
10 % [192, 195]. By recycling the benzene,
Production. Hydrodealkylation of Benzene. 1 kg of starting material may yield 0.8 – 0.85 kg
Biphenyl is currently obtained mainly as a by- biphenyl, 0.07 – 0.12 kg terphenyl, and some
product in the production of benzene by ther- higher polyphenyls. The major technical prob-
mal or catalytic hydrodealkylation of toluene lem in the production of biphenyl by dehydro-
(→Benzene, Chap. 5.3.1.) [182–185]. The reac- condensation of benzene is carbonization at the
tion is carried out in a hot tube reactor at 700 ◦ C high temperature required [196]. The carbon de-
with a hydrogen pressure of 4 MPa; the ratio of posit on the reactor tube walls and the release of
hydrogen to toluene is 4 : 1. Toluene conver- large carbon flakes at bends or narrow passages
sion ranges from 35 to 85 %, with methane being within the tubes may lead to blockages. Counter-
formed at 2.5 mol % above the stoichiometric. 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
Biphenyl is obtained as the pot residue, fol- means of circulating gases, and high feedstock
lowing removal of gas and distillation of ben- velocity, and the use of turbulent flow [192]. A
zene and toluene; about 1 kg of biphenyl is ob- proposal for further improving the process is to
tained per 100 kg benzene. The biphenyl can be carry out the reaction in fluidized beds. Quartz
enriched to a purity of 93 – 97 % by distillation. sand [202] or sintered corundum [203] is used
Thermal Dehydrocondensation of Benzene. as energy carrier in this process; these materials
Until the early 1970s, biphenyl was produced are recycled after the carbon deposits have been
exclusively by specific thermal dehydroconden- removed by burning them outside of the reactor.
sation of benzene [186]. This process yields a Other Synthetic Routes. In addition to these two
higher quality product than the toluene route, industrial processes for the production of biphe-
e.g., distilled material with a purity >99.5 %; nyl, other synthetic routes have been described:
varying quantities of terphenyl isomers and sev-
eral higher homologues are also obtained. 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].
Hydrocarbons 39
4) Heating phenylmagnesium halides with cop- with a biphenyl content of >99.5 %; it is trans-
per(II) salts [208]. ported in bags or fiber drums; in the molten state
5) Oxidative dehydrocondensation of benzene (ca. 120 ◦ C) in tank cars. It is not a regulated ma-
[209–214]. Benzene is oxidized to biphenyl terial for transportation, on account of its toxi-
by platinum- or palladium salts in glacial city; there are no restrictions for transportation
acetic acid at 120 ◦ C; the salts are reduced by sea, land, or air. The greatest hazards in han-
to the pure metals, and stoichiometric quan- dling biphenyl are the risk of dust explosions
tities of the salts are required. and the ignition of biphenyl vapor – air mixtures
6) Dehydrocondensation of benzene in the pres- over the molten product.
ence of catalytic quantities of noble metal
salts requires reoxidation of the metals dur- Uses. Biphenyl is an important heat transfer
ing the course of the reaction. This may be agent, because of its high thermal stability. It is
done with pure oxygen, but a partial pres- used as a composite with diphenyl ether, espe-
sure of 6 MPa is required [215]. Additives cially as a melting point depressant. The eutectic
such as β-diketones [216], heteropolyacids mixture of 26.5 % biphenyl and 73.5 % diphenyl
[217], or organic copper salts [218] are inca- ether is marketed under the trade names Diphyl
pable of significantly lowering the required (Bayer), Dowtherm A (Dow), Thermic (ICI),
oxygen pressure, so that a potentially explo- Gilotherm (Rhône-Poulenc), Therm S 300 (Nip-
sive regime cannot be avoided [217]. Yields pon Steel), Santotherm VP and Therminol VP-1
of biphenyl are ca. 260 mol %, calculated on (both Monsanto) for use at up to 400 ◦ C. The eu-
the palladium salt charge. tectic begins to boil at 256 ◦ C; the vapor pressure
7) Hydrodimerization of benzene to phenylcy- at the maximum operating temperature (400 ◦ C)
clohexane, followed by dehydrogenation to is ca. 1.1 MPa.
biphenyl [219–223]: 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-
Suitable catalysts for this dimerization are droxy derivatives, o- and p-hydroxybiphenyl and
noble metals on molecular sieve, alkali met- p,p -dihydroxybiphenyl are used industrially. o-
als on alumina or carbon, and transition met- Hydroxybiphenyl is used as a preservative and
als such as nickel, tungsten, or zinc on the fungicide. It is synthesized mainly from cyclo-
zeolites faujasite or mordenite. A selectiv- hexanone (→Phenol Derivatives). p-Hydroxybi-
ity of up to 87 % has been claimed. Sub- phenyl and p,p -dihydroxybiphenyl are gener-
sequent dehydrogenation is carried out at ated by sulfonation of biphenyl, followed by fu-
about 400 ◦ C over Pt – Al2 O3 [223], Cr2 O3 – sion with alkali.
Na2 O – Al2 O3 [222, 224], Cr2 O3 – K2 CO3 –
Fe2 O3 [225], or Cr2 O3 – MgO – Al2 O3 [226] Economic Aspects. According to a 1976 es-
giving yields of up to 99 %. Data concern- timate, 67 % of the annual United States pro-
ing residence times over the catalyst are not duction of 40 000 t biphenyl was produced by
available, either for the hydrodimerization or dealkylation of toluene and 33 % by thermal de-
the dehydrogenation processes. hydrocondensation of benzene [229].
The annual production for 1984 is estimated
Analysis. Gas chromatography is used to an- at 15 000 t, with an increasing proportion from
alyze biphenyl and its derivatives. dealkylation of toluene. The production capaci-
ties of the major producers, Monsanto, (United
Storage and Transportation. Biphenyl is States), Bayer (Federal Republic of Germany),
marketed as a solid in the form of plates, flakes, and Monsanto (United Kingdom) have not been
or pellets, containing <10 ppm benzene and
40 Hydrocarbons
Table 19. Physical properties of terphenyls
disclosed. Based on the legal restrictions im- steam volatile, and show the typical substitu-
posed on the use of PCBs, capacity should far tion reactions of aromatic hydrocarbons, such
exceed current demand. as bromination and nitration. The substitution
pattern for the preferred positions of the three
terphenyls is shown in Table 20.
4.2. Terphenyls
Production. Terphenyls are byproducts in
Physical Properties. The three isomeric ter- the production of biphenyl by dehydroconden-
phenyls, C18 H14 , M r 230.29, are colorless to sation of benzene; they are found in the high-
pale yellow crystalline solids [230–235]. Some boiling fraction of the pyrolysis products. The
physical properties are listed in Table 19. 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.
tained by zone refining [245]. Terphenyl iso- 5. Hydrocarbons from Coal Tar
mers, including hydrogenated terphenyl, can be
analyzed by gas chromatography. The sole pro- Anthracene and Naphthalene and Hydronaph-
ducer of terphenyl in the United States is Mon- thalenes are separate keywords.
santo. The market amounts to several thousand Hydrocarbons from coal tar are generally
tons per year. Terphenyl is available in Europe recovered in large-scale operations by distil-
from Bayer. lation and crystallization, after the polar, co-
boiling components, phenols and nitrogen bases,
Storage and Transportation. Solid ter- have been separated by extraction (→Tar and
phenyl is shipped as flakes in laminated bags or Pitch). Supplies of coal tar as a raw material
in fiber drums, liquid, hydrogenated terphenyl have been sufficient to date to meet the demand
in tank cars, steel drums, or barrels. Terphenyls for these hydrocarbons, so that no other produc-
and hydrogenated terphenyls possess a low toxi- tion processes, such as syntheses, have yet been
city, and are therefore not regarded as dangerous developed to the stage of commerciality. Hydro-
goods for the purpose of transportation. carbons from coal tar are used predominantly in
the synthesis of dyes, plastics, and pharmaceu-
Uses. Terphenyl is used as a heat transfer ticals.
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- 5.1. Acenaphthene
terphenyl, and 2 – 18 % higher polyphenyls is
used. This composition approximates the pot Acenaphthene [83-32-9], C12 H10 , M r 154.21,
residue from the distillation of biphenyl. This was discovered in coal tar in 1867 by Berth-
mixture has bp ca. 360 ◦ C; a disadvantage is elot.
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 Physical Properties. Acenaphthene forms
dye carrier for pressure sensitive recording ma- colorless needles; readily soluble in chloroform,
terials and copy paper, and as a heat transfer toluene, and hot ethanol; soluble in cold ethanol,
oil. Heat transfer oils consisting of partially hy- hot diethyl ether, and benzene; insoluble in wa-
drogenated terphenyl isomers are marketed un- ter.
der the following trade names: Santotherm 66 mp 95.4 ◦ C
or 88, Therminol (Monsanto, United States), bp (101.3 kPa) 277.5 ◦ C
Gilotherm (Rhône-Poulenc, France), and Therm (25 ◦ C) 1.2195 g/cm3
n100 1.6048
S 900 (Nippon Steel, Japan). Their usage range D
T crit 530 ◦ C
extends to 340 ◦ 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
4.3. Polyphenyls.
Polyphenyls with four or more phenyl residues Chemical Properties. Acenaphthene reacts
have generally been synthesized as p-linked iso- with halogens, preferentially at 3-, 5-, and 6-
mers and are used as scintillators [247]. They positions, or at the 1-position when irradi-
are economically unimportant. ated with visible light. Nitration and sulfona-
A symmetrical hydrocarbon within this class tion occur at the 3-, 5-, or 6-positions. Cat-
is 1,3,5-triphenylbenzene, C24 H18 , mp 174 ◦ C, alytic hydrogenation leads to tetra- and decahy-
prepared by condensation of three molecules of droacenaphthene. Decahydroacenaphthene can
acetophenone.
42 Hydrocarbons
Derivatives.
Alkylated indans are used in the production of zene, diethyl ether, carbon disulfide, and hot eth-
synthetic lubricants (→Lubricants and Related anol; soluble with difficulty in cold ethanol.
Products). Indan is used for the synthesis of in-
mp 115 ◦ C
danols (→Phenol Derivatives). bp (101.3 kPa) 298 ◦ C
2-Indanol [4254-29-9], C9 H10 O, M r 134.18; (20 ◦ C) 1.181 g/cm3
mp 69 ◦ C; colorless crystals. It is obtained by hy- Heat of fusion 1.21×105 J/kg
drogenation of 2-indanone prepared by reaction Heat of combustion (25 ◦ C) 3.994×107 J/kg
the synthesis of anti-inflammatory drugs [293]. can be used as starting material for ther-
Reaction of fluorene with sulfur produces elec- mally stable plastics [308]. 9,9-Bis(4-methyl-
trically conductive polymers [294]. The azine aminophenyl)fluorene (from fluorenone by re-
dye Sirius Light Violet FRL is produced from 2- action with phenyl methyl amine) has been sug-
aminofluorene (→Azine Dyes, Chap. 5.2.). The gested as a hardener for epoxy resins [309]. Flu-
fluorene analog of malachite green dye is ob- orenone is also suitable as an oxidizing agent in
tainable from 3,6-bis-(dimethylaminofluorene) the Oppenauer reaction.
[295]. 2,7-Diiodofluorene can be employed in
the production of styryl dyes for optical bright-
eners [296]; heptabromofluorene can be used as 5.4. Fluoranthene
a fire-retardant for plastics [297]. Fluorene it-
Fluoranthene [206-44-0], C16 H10 , M r 206.26,
self can be used in admixture with biphenyl and
was isolated from coal tar by Fittig and Geb-
phenyltoluene as a carrier for dyeing polyester
hardt in 1878.
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 Physical Properties. Colorless crystals with
light blue fluorescence; insoluble in water; read-
9-Fluorenone forms yellow crystals; insol- ily soluble in diethyl ether, boiling ethanol, chlo-
uble in water; soluble in hot sulfuric acid; roform, carbon disulfide, and glacial acetic acid.
very readily soluble in ethanol and diethylether;
volatile in steam. Fluorenone can be produced by mp 110 ◦ C
bp (101.3 kPa) 384 ◦ C
catalytic oxidation of fluorene [261, 299–302], (20 ◦ C) 1.236 g/cm3
or of fluorene fractions in the presence of a qua- Heat of vaporization 3.29×105 J/kg
ternary ammonium salt [303], or by catalytic ox- Heat of combustion (25 ◦ C) 3.913×107 J/kg
idative cracking (oxicracking) of suitable aro-
matic [304]. 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
Di- and trinitrofluorenones are synthesized
mixture of 7,8- and 8,9-phthaloyl fluoranthene.
from fluorenone for use as electron accep-
tors in electrophotography, e.g., in copying Production. Fluoranthene belongs to the
systems [305]. 2,4,7-Trinitrofluorenone is used main constituents of high-temperature coal tar,
as an analytical reagent for polynuclear aro- which contains, on average, 3.3 % of fluoran-
matic hydrocarbons. Fluorenone can be used thene. It is recovered from the fluoranthene frac-
as starting material for the synthesis of im- tion, which boils at 373 – 385 ◦ C, and is obtained
idazolylfluorene salts with antimycotic activ- from the high-boiling anthracene oil II and from
ity [306]. 9,9-Bis(4-hydroxylphenyl)fluorene pitch distillate by redistillation. Subsequent re-
(from fluorenone by reaction with phenol [307]) crystallization of the fluoranthene fraction from
46 Hydrocarbons
mp 101 ◦ C
solvent naphtha gives 95 % pure fluoranthene. bp (101.3 kPa) 338.5 ◦ C
Further refining is carried out by recrystalliza- (20 ◦ C) 1.172 g/cm3
tion from xylene with simultaneous partial sul- Heat capacity (25 ◦ C) 1.361×103 J/kg
fonation of the impurities with ca. 1 % of con- Heat of fusion 1.04×105 J/kg
Heat of combustion (25 ◦ C) 3.982×107 J/kg
centrated sulfuric acid, or by recrystallization
from a pyridine – water mixture.
Chemical Properties. Phenanthrene can
Uses. Fluoranthene is used in the pro- be oxidized to phenanthrenequinone, diphenic
duction of fluorescent dyes (→Fluorescent acid, or to phthalic acid, depending on the ox-
Dyes). Yellow vat dyes are obtained by idant. Hydrogenation produces 9,10-di-, tetra-,
condensation with phthalic anhydride, via octa-, or perhydrophenanthrene. Phenanthrene
phthaloyl fluoranthenes, whereas bromina- is halogenated predominantly in the 9,10-
tion [310] and condensation with 1-amino-4- positions, and nitrated in the 9-position. Sul-
benzoyl-amino anthraquinone produces olive fonation leads to mixtures of 2-, 3-, and 9-
dyes [311]. 2-Benzylfluoranthene (from 2,3- phenanthrene sulfonic acids. For acylation see
dihydrofluoranthene [312] and benzaldehyde) →Acylation and Alkylation, Chap. 2.2.2.
can also be used for the production of dye
intermediates [313]. Fluoranthene and alkyl Production. Phenanthrene, at a concentra-
fluoranthenes can be used as stabilizers in tion of 5 %, is the second most important coal
epoxy resin adhesives [314], as additives in tar constituent in terms of quantity after naph-
electrically-insulating epoxy resin compositions thalene. During primary distillation of coal tar,
[315], and in electrically-insulating oils [316]. it is concentrated in the anthracene oil fraction.
2,3-Dihydrofluoranthene and 1,2,3,10 b-tetra- After crystallization of the anthracene residues
hydrofluoranthene [317] are used for numerous the phenanthrene is recovered as a fraction from
derivatives, some of which (e.g., their deriva- the filtrate of this crystallization, or from the top
tives of phenylethyl amine) are of pharmaceuti- fraction of crude anthracene distillation, by re-
cal interest. 7-Fluoranthenyl aminoalcohols (for distillation. Technically pure grades of phenan-
cytostatics, anthelmintics, or bactericides [318]) threne are obtained by sulfuric acid refining and
and bis-aminoketones, e.g., 3,8-bis(4-piperi- recrystallization from methanol, or by repeated
dinobutyryl)fluoranthene (as antiviral agents rectification of the phenanthrene fraction. The
[319]) can be synthesized from fluoranthene. accompanying substances can be separated ei-
ther by partial sulfonation, or by partial conden-
sation with formaldehyde and hydrogen chlo-
5.5. Phenanthrene ride. The most persistent accompanying sub-
stance, diphenylene sulfide, can be completely
Phenanthrene [85-01-8], C14 H10 , M r 178.24 removed by treating the melt with sodium and
was discovered in coal tar by Fittig and Os- maleic anhydride.
termayer in 1872.
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,
Physical Properties. Colorless flakes with can be produced by the electrochemical con-
weak blue-violet fluorescence; readily sub- version of phenanthrene diazonium salts in a
limable; insoluble in water; slightly soluble in solvent containing a conductive salt, and sub-
ethanol; readily soluble in diethyl ether, ben- sequent doping with various ions (e.g., Na+ ,
zene, carbon disulfide, and glacial acetic acid. Ba2+ , H+ , etc.) [321]. Liquid-crystalline 7-
n-alkyl-9,10-dihydrophenanthrene-2-carboxyl-
ic acid ester, used for optical-electronic ap-
Hydrocarbons 47
sure to a concentration of 400 – 600 ppm leads to of Cincinnati, Ohio, in 1980, revealed a con-
neuropathies [375], which are characterized by centration of 45 ng/L hemimellitene, 127 ng/L
degeneration of both myelin sheaths and axons pseudocumene, and 36 ng/L mesitylene [386].
[376–378]. Trimethylbenzenes are toxic to aquatic organ-
MAK (1987): 50 ppm. isms; LC50 values between 10 and 100 ppm have
n-Heptane has narcotic properties. In hu- been reported [387].
mans 0.1 % in the inhaled air leads to dizzi- The TLV (TWA) level for trimethylbenzene is
ness and 0.5 % to equilibrial disturbances with 25 ppm [388]. The MAK for Trimethylbenzenes
loss of motor coordination [365]. Within 3 min (all isomers (1,2,3–;1,2,4–;1,3,5–) has been set
4.8 vol % in the air inhaled leads to asphyxia at 20 ppm. [389]. The United States toxicity as-
[379]. Like hexane, heptane is a skin and mu- sigment ranges from low to moderately toxic
cosal irritant [365]. The biotransformation of n- for pseudocumene, to highly toxic for the iso-
heptane, as with hexane, is by oxidation. Hep- mer mixture; disturbances in the central nervous
tane apparently also possesses neurotoxic prop- system and abnormal blood pictures have been
erties [380]. Myocardial sensitization to cate- reported [387].
cholamines has also been found [381]. Tetramethylbenzenes. The oral toxicity of
MAK (1987): 500 ppm. tetramethylbenzenes is low; LD50 >5000 mg/kg
n-Octane is similar to n-heptane in regard to (rat, oral). Durene is additionally classified as in-
its narcotic effect; however, it does not seem to travenously highly toxic; LD50 180 mg/kg (i.v.,
cause any other effects on the nervous system mice). Tetramethylbenzenes cause mild skin re-
[372]. Octane is also metabolized by oxidation actions [387]. Two values are cited for the odor
[382]. threshold concentration of durene: 0.083 and
MAK (1987): 500 ppm 0.087 mg/m3 [385].
Only limited toxicological data are available Hexamethylbenzene. Hexamethylbenzene
on the higher molecular mass alkanes. has low oral toxicity; LDLo 5000 mg/kg (rat).
It is, however, suspected of causing neoplastic
effects. Addition of nitromethane to hexam-
6.2. Alkenes ethylbenzene can lead to an explosive reaction
[387].
The alkenes under consideration are toxicolog- Diethylbenzenes. Diethylbenzene is more
ically not very active. Higher molecular mass toxic than monoethylbenzene (→Ethylbenzene)
compounds possess narcotic properties. The [390]. DEB is emitted into the environment from
alkenes are not neurotoxic. The α-olefins gener- engine fuel and heating oil [386]. Diethylben-
ally seem to be more reactive and toxic than the zenes are toxic to aquatic organisms; the LC50
β-isomers. 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-
6.3. Alkylbenzenes eral Republic of Germany DEB is classified as
an aquatic hazard; aquatic hazard class 2. DEB
Trimethylbenzenes. The toxicological data causes mild to moderate irritation of the eye
available on trimethylbenzenes is mainly rather and mucous membranes. The oral toxicity is
old. The acute inhalatory toxic dose for the 1,2,4- low: LDLo 5000 mg/kg (rat) [387]. The vapor
and the 1,3,5-isomers is in the range 7000 – possesses an anesthetic action and may cause
9000 ppm [383, 384]. headache, vertigo, or vomiting; the odor thresh-
As a result of their occurrence in automo- old concentration is <10 ppm [391].
tive fuel and heating oil, low concentrations Triethylbenzene. The LC50 of triethylben-
of trimethylbenzenes can be demonstrated in zene (aquatic organisms) is between 100 and
air and water. For example, measurements in 1000 mg/L. The oral toxicity is low: LDLo
1974 in a tunnel in Rotterdam indicated a 5000 mg/kg (rat). Implant experiments with hex-
level of 0.015 ppm pseudocumene; see [385] for aethylbenzene in mice have raised the suspicion
more details. An analysis of the drinking water that the compound may be carcinogenic [387].
50 Hydrocarbons
Ethyltoluene. Ethyltoluene can, as with aro- which originate from combustion emissions and
matic hydrocarbons, be detected in the envi- industrial processes. For this reason toxic phe-
ronment [385]. The oral toxicity of 2- and 4- nomena possibly caused by exposure to humans
ethyltoluene is low: LDLo 5000 mg/kg (rat) are often unlikely to be attributed to one sin-
[387]. gle aromatic compound. Incorporation of these
Cumene. Cumene appears to be orally as non- compounds into the body occurs predominantly
toxic as propylbenzene, it is even less so by by way of the respiratory tract. Skin contact is
inhalation [383, 392, 393]. Cumene has a pro- important with occupational exposure.
longed depressant effect on the central nervous
system [394].
The MAK for cumene has been set at 6.5.1. Biological Effects
250 mg/m3 , corresponding to a cumene vapor
concentration in air of 50 mL/m3 . The vapor 6.5.1.1. Carcinogenicity and Mutagenicity
pressure of cumene at 20 ◦ C is 0.5 kPa. There
is a risk of cumene being absorbed through the The aromatic compounds under discussion are
skin [395]. Inhalation of cumene vapor leads to considered as noncarcinogens, but pyrene and
the delayed appearance of a long lasting narcotic fluoranthene are able to enhance the carcino-
effect [396]. genic potential of benzo [a] pyrene [50-32-8]
when applied simultaneously to mouse skin (co-
carcinogenic effect) (→Carcinogenic Agents)
6.4. Biphenyls and Polyphenyls [408]. Phenanthrene resulted in just one positive
tumor-initiation test in a series of carcinogenic-
Biphenyls. Biphenyl dust or vapor is irritat- ity studies [408, 409]. Acenaphthene, known as
ing to the eye and mucous membrane at a con- an inhibitor of mitosis in plant cells [410], in-
centration as low as 3 – 4 ppm [390], and to duced the initial phase of bronchocarcinoma in
the skin after extended exposure. A concen- rats after an inhalation period (4 months) of 4
tration of >5 mg/m3 for long periods is con- hours per day with 0.5 – 1.25 mg/m3 [411], but
sidered a health hazard; systemic toxic effects was negative in repeated topical application to
were elicited in humans by a concentration mouse skin [412]. With their low activity in in
with inhalatory exposure maxima of 128 ppm vivo and in vitro assays, there is limited evi-
[400, 401]. The olfactory threshold is ca. 0.06 – dence that phenanthrene and pyrene are muta-
0.3 mg/m3 . Some toxicity data are listed below genic [408, 409, 413–415]. As for fluoranthene,
[397–399]: several more recent positive results indicate a
mutagenic potential [408, 416–419].
LD50 (rat, oral) 3280 mg/kg So far, tests on fluorene have given negative
LD50 (rabbit, oral) 2400 mg/kg
LD50 (rabbit, dermal) 2500 mg/kg
results [408].
TLm (fathead minnow, 96 h) 1.5 mg/L
Triphenyls are only minimally toxic [402]. 6.5.1.2. Mammalian Toxicity and
Terphenyls. Toxicologically, terphenyl iso- Toxicokinetics
mers should be treated like biphenyl. The oral
LD50 is ca. 4.6 – 4.7 g/kg [403]. The toxicity of The acute toxicity of the aromatic compounds
partially (40 %) hydrogenated terphenyl (LD50 under consideration is low: LD50 (oral and der-
17.5 g/kg) is much less than that of the fully aro- mal) is 700 – 2700 mg/kg body weight in rodents
matic terphenyls [404–407]. [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
6.5. Hydrocarbons from Coal Tar [413]. No fatalities occurred in rats exposed
to indene vapor (800 – 900 ppm) for six in-
The aromatic hydrocarbons under consideration halation periods of 7.5 h each; however, sys-
are usually found in complex mixtures of various temic pathological changes in the vascular sys-
other polycyclic aromatic hydrocarbons (PAH), tem and several organs were observed [412].
Hydrocarbons 51
24. H.-W. v. Gratkowski, Erdöl Kohle Erdgas 50. W. Dimmling, R. Seipenbusch, Hydrocarbon
Petrochem. Brennst. Chem. 28 (1975) 81. Process. 54 (1975) no. 9, 169.
25. H. Hiller, Erdöl Kohle Erdgas Petrochem. 51. H. Riegel, H. D. Schindler, M. C. Sze, Erdöl
Brennst. Chem. 28 (1975) 74. Kohle Erdgas Petrochem. Brennst. Chem. 26
26. P. D. Bartlett, K. E. Schueller, J. Am. Chem. (1973) 704.
Soc. 90 (1968) 6071. 52. R. P. Lowry, A. Aguilo, Hydrocarbon Process.
27. Houben-Weyl, 4th ed., vol. IV, parts 3 and 4. 53 (1974) no. 11, 103.
28. J. H. Lucas, F. Baardman, A. P. Kouwenhoven, 53. B. V. Vora, T. Imai, Hydrocarbon Process. 61
Angew. Chem. 87 (1975) 740, 88 (1976) 412. (1982) no. 4, 171.
29. W. C. van Ziill Langhout, Proc. World Pet. 54. L. B. Phillips, Hydrocarbon Process. 42
Congr. 1975, vol. 5, 197. (1963) no. 6, 159.
30. W. L. Lafferty, Jr., W. R. Stokeld, Adv. Chem. 55. A. A. Di Giacomo, J. B. Maerker, J. W. Schall,
Ser. 103 (1971) no. 7, 130. Chem. Eng. Prog. 57 (1961) no. 5, 35.
31. L. F. Albright, Chem. Eng. 73 (1966) no. 14, 56. Universal Oil Products, US 3 649 177, 1972
119; 73 (1966) no. 19, 205; 73 (1966) no. 21, (D. H. Rosback).
209.
32. L. Alcock, R. Martin, K. H. Bourne, W. A. General References
Peet, J. Winsor, Oil Gas J. 72 (1974) no. 27, 57. F. Asinger: Die Petrolchemische Industrie,
102. vol. 1 (Gewinnung der Olefine), vol. 2
33. D. P. Thornton, D. J. Ward, R. A. Erickson, (Verwendung der Olefine) Akademie-Verlag,
Hydrocarbon Process. 51 (1972) no. 8, 81. Berlin 1971.
34. H. Jockel, B. E. Triebskorn, Hydrocarbon 58. J. Falbe (ed.): Methodium Chimicum, vol. 4,
Process. 52 (1973) no. 1, 93. Thieme Verlag, Stuttgart 1980, pp. 35, 111.
35. Hydrocarbon Process. 52 (1973) no. 11, 145. 59. Houben-Weyl, 4th ed., vol. 5/1 b, Thieme
36. P. R. Pujado, B. V. Vora, Energy Prog. 4 Verlag, Stuttgart 1972.
(1984) no. 3, 186. 60. S. Patai, J. Zabicky: The chemistry of alkenes,
37. M. S. Scurrell, Appl. Catal. 32 (1987) 1. vol. 1 (1964), vol. 2 (1970), Interscience
38. P. Leprince, A. Chauvel, L. Castex: Procédés Publishers, New York.
de Pétrochimie, Editions Technip, Paris 1971, 61. E. S. Strauss, G. Hollis, O. Kamatari: Chemical
p. 269. Economics Handbook Marketing Research
39. K. Ito, I. Dogane, K. Tanaka, Hydrocarbon Report “Linear alpha-olefins,” Stanford
Process. 64 (1985) no. 10, 83. Research Institute, Menlo Park, CA, 1984.
40. S. C. Arnold, G. D. Sucui, L. Verde, A. Neri,
62. B. Fell, Tenside Deterg. 12 (1975) 3.
Hydrocarbon Process. 64 (1985) no. 9, 123.
63. H. Isa, CEER Chem. Econ. Eng. Rev., 9
41. Japan. Chemical Reporter, 1981, October, p. 2.
(1977) 26.
42. F. Solymosi, I. Tombácz, G. Kutsan, J. Chem.
64. A. H. Turner, J. Am. Oil Chem. Soc. 60 (1983)
Soc. Chem. Commun. 1985, 1455.
623.
43. R. V. Porcelli, B. Juran, Hydrocarbon Process.
65. A. Behr, W. Keim, Arabian J. Sci. Eng. 10
65 (1986) no. 3, 37.
(1985) 377.
44. W. Hinsen, M. Baerns, Chem. Zt. 107 (1982)
9.
45. Kh. E. Khcheyan, O. M. Revenko, A. N. Specific References
Shatalova: “Oxidative Methylation, a new 66. K. G. Mittal, J. Chem. Technol. Biotechnol. 36
route to petrochemicals through alternative (1986) 291 – 299.
feedstock,” Special Paper No. 16, World 67. D. B. Broughton, R. C. Berg, Chem. Eng. 77
Petroleum Congress 11th, 1983 (London). (1970) 86.
46. G. de Pontanel: Proteins from Hydrocarbons, 68. B. V. Vora, P. R. Pujado, J. B. Spinner, T. Imai,
Academic Press, London 1972. Proc. World Tenside Congr. 1984 (Munich)
47. S. Gutcho: Proteins from Hydrocarbons, vol. II, 16 – 26.
Noyes Data Corp., Park Ridge, N.J., 1973. 69. R. C. Berg, B. V. Vora in McKetta,
48. P. J. Rockwell: Single Cell Proteins from Cunningham (ed.): Encyclopedia of Chemical
Cellulose and Hydrocarbons, Noyes Data Processing and Design, vol. 15, Marcel
Corp., Park Ridge, N.J., 1976. Dekker, New York 1982, 266 – 284.
49. G. Voss, Erdöl Kohle Erdgas Petrochem. 70. H. D. Wulff, Fette Seifen Anstrichm. 69 (1967)
Brennst. Chem. 26 (1973) 249. 32.
Hydrocarbons 53
71. R. Stöbele, Chem. Ing. Tech. 36 (1964) 858. 97. W. Keim, A. Behr, G. Kraus, J. Organomet.
72. K. Ziegler et al., Angew. Chem. 64 (1952) 323, Chem. 251 (1983) 377.
67 (1955) 542. 98. A. Behr, V. Falbe, U. Freudenberg, W. Keim,
73. Gulf Res. Dev. Co., DE 1 443 927, 1961. Isr. J. Chem. 27 (1986) 277 – 279.
74. Ethyl Corp., US 3 906 053, 1973. 99. W. Keim, Ann. N.Y. Acad. Sci. 415 (1983)
75. W. Keim, A. Behr, M. Röper in G. Wilkinson 191 – 200.
(ed.): Comprehensive Organometallic 100. B. Bogdanović, B. Spliethoff, G. Wilke,
Chemistry, vol. 5, Pergamon Press, Oxford Angew. Chem. Int. Ed. Engl. 19 (1980) 622.
1982, 371. 101. G. Speier, J. Organomet. Chem. 97 (1975)
76. W. Keim, Chem. Ing. Tech. 56 (1984) 850. 109.
77. Shell Development Comp., US 3 635 937, 102. M. G. Barlow, M. J. Bryant, R. N. Haszeldine,
1972; 3 637 636, 1972; 3 644 564, 1972; A. G. Mackie, J. Organomet. Chem. 21 (1970)
3 644 563, 1972; 3 647 914, 1972; 3 647 915, 215.
1972; 3 661 803, 1972; 3 686 159, 1972 (W. 103. S. Datta, M. B. Fisher, S. S. Wreford, J.
Keim et al.). Organomet. Chem. 188 (1980) 353.
78. E. R. Freitas, R. C. Gum, Chem. Eng. Progress. 104. R. Schrock, S. McLain, J. Sancho, Pure Appl.
75 (1979) no. 1, 73. Chem. 52 (1980) 729.
79. E. L. T. M. Spitzer, Seifen öle Fette Wachse 105. J. A. Brennan, Ind. Eng. Chem. Prod. Res. Dev.
107 (1981) 141 – 144. 19 (1980) 2 – 6.
80. M. Sherwood, Chem. Ind. (London) (1982) 106. R. L. Shubkin, M. S. Baylerlan, A. R. Maler,
994 – 995. Ind. Eng. Chem. Prod. Res. Dev. 19 (1980)
81. W. Keim, F. H. Kowaldt, Erdöl Kohle Erdgas 15 – 19.
Petrochem. Compendium (1978/79) 107. F. J. Karol, CHEMTECH 1983 (April) 222.
453 – 462. 108. B. W. Werdelmann, Fette Seifen Anstrichm. 76
82. Shell Oil Comp., US 3 647 906, 1972 (F. F. (1974) 1.
Farley). 109. Chem. Eng. News, (1988) May 30, 10.
83. M. Peuckert, W. Keim, Organometallics 2 110. Eur. Chem. News, (1987) Dec. 21/28, 21.
(1983) 594. 111. Eur. Chem. News, (1981) July 6, 23.
84. W. Keim, F. H. Kowaldt, R. Goddard, C. 112. Eur. Chem. News, (1987) Jan. 5/12, 19.
Krüger, Angew. Chem. 90 (1978) 493. 113. Eur. Chem. News, (1987) Jan. 19, 8.
85. W. Keim, A. Behr, B. Limbäcker, C. Krüger, 114. Eur. Chem. News, (1981) Jan. 26, 27.
Angew. Chem. 95 (1983) 505. 115. Eur. Chem. News, (1982) July 5, 26.
86. W. Keim et al., J. Mol. Catal. 6 (1979) 79. 116. G. Wilke, Angew. Chem. Int. Ed. Engl. 2
87. W. Keim et al., Organometallics 5 (1986) (1963) 105.
2356. 117. A. C. L. Su, Adv. Organomet. Chem. 17
88. Chem. Eng. News (1972), April 10, 16 – 17. (1979) 269.
89. C. Commereuc, Y. Chauvin, J. Gaillard, J. 118. S. Warwel, Erdöl Erdgas Kohle 103 (1987)
Léonard, J. Andrews, Hydrocarbon Process. 238.
63 (1984) no. 11, 118 – 120. 119. R. C. Weast, M. J. Astle (eds.): CRC Handbook
90. W. J. Benedek, J.-L. Mauleon, Hydrocarbon of Data on Organic Compounds, CRC Press,
Process. 59 (1980) no. 5, 143. Boca Raton, Florida, 1985.
91. J. Léonard, J. F. Gaillard, Hydrocarbon 120. S. Coffey (ed.): Rodd’s Chemistry of Carbon
Process. 60 (1981) no. 3, 99. Compounds, III A, Elsevier, Amsterdam 1971,
92. J. F. Boucher, G. Follain, D. Fulop, J. Gaillard, pp. 228 – 231.
Oil Gas J. 80 (1982) 84 – 86. 121. A. D. Rudkovskii et al., Chem. Technol. Fuels
93. Y. Chauvin, J. Gaillard, J. Léonard, P. Oils (Engl. Transl.) 19 (1983) 387 – 391.
Bonnifay, J. W. Andrews, Hydrocarbon 122. F. P. McCandless, Ind. Eng. Chem. Prod. Res.
Process. 61 (1982) no. 5, 110 – 112. Dev. 19 (1980) 612 – 616; 21 (1982)
94. D. Commereux, Y. Chauvin, G. Leger, J. 483 – 488.
Gaillard, Rev. Inst. Fr. Pet. 37 (1982) 123. Idemitsu Kosan Co. Ltd., EP 0 104 607, 1983
639 – 649. (H. Sato).
95. J. Gaillard, Pet. Tech. 314 (1985) 20 – 27. 124. Universal Oil Products, US 4 487 984, 1984 (T.
96. B. Bogdanović, Adv. Organomet. Chem. 17 Imai).
125. O. A. Anunziata et al., Appl. Catal. 15 (1985)
(1979) 105.
235 – 245.
54 Hydrocarbons
126. J. Hanotier, M. Hanotier-Bridoux, J. Mol. 150. Mitsui Petrochemical Ind., EP 0 046 678, 1981
Catal. 12 (1981) 133 – 147. (M. Hisaya).
127. L. I. Smith, Org. Synth., coll. vol. 2 (1948) 151. J. P. Fortuin, M. J. Waale, R. P. Oosten, Pet.
248 – 253. Refiner 38 (1959) no. 6, 189.
128. Toyo Soda Manufacturing, DE-OS 3 330 912, 152. A. H. Olzinger, Chem. Eng. NY 82 (1975)
1983 (T. Sakamoto, T. Hironaka, K. Sekizawa, no. 6, 50.
Y. Tsutsumi). 153. Universal Oil Products, US 3 763 259, 1973
129. Toa Nenryo, EP 0 101 651, 1983 (T. Maejima, (G. L. Hervert).
N. Tagaya, S. Sakurada). 154. Mobil Oil Corp., EP-A 0 149 508, 1985 (W. W.
130. J. Haggin, Chem. Eng. News, (1987) June 1, Kaeding).
22 – 28; (1987) June 22, 22 – 25. 155. J. Ewers, H. W. Voges, G. Maleck, Erdöl Kohle
131. International Flavors and Fragrances, Erdgas Petrochem. Compendium 74/75 (1975)
US 3 869 524, 1974 (K. K. Light, J. B. 487.
Anglim). 156. K. Ito, I. Dogane, K. Tanaka, Hydrocarbon
132. Phillips Petroleum, US 4 568 784, 1985 (R. L. Process. 65 (1986) no. 11, 89.
Cobb). 157. H. A. Colvin, J. Muse, CHEMTECH 1986,
133. S. A. Shama, C. C. Wamser, Org. Synth. 61 no. 500.
(1983) 62 – 64. 158. K. Ito, Hydrocarbon Process. 52 (1973) no. 8,
134. H. F. Davis, S. K. Chattopadhyay, P. K. Das, J. 89.
Phys. Chem. 88 (1984) 2798–2803. 159. Phillips Petroleum Co., US 3 291 847, 1966
135. W. W. Kaeding, G. C. Barile, M. M. Wu, Catal. (P. F. Warner); US 3 316 315, 1967 (W. A.
Rev. Sci. Eng. 26 (1984) 597 – 612. Jones).
136. Dow Chemical, US 2 763 702, 1956 (J. L. 160. K. Weissermel, H.-J. Arpe: Industrielle
Amos, K. E. Coulter). Organische Chemie, 3rd ed., VCH
137. W. W. Kaeding, L. B. Young, C.-C. Chu, J. Verlagsgesellschaft, Weinheim 1987,
Catal. 89 (1984) 267 – 273. pp. 72 – 79.
138. W. W. Kaeding, G. C. Barile, Polym. Sci. 161. A. C. Olson, Ind. Eng. Chem. 52 (1960)
Technol. (Plenum) 25 (1984) 223 – 241. no. 10, 833.
139. Eur. Chem. News, Oct. 17 (1988) p. 4. 162. R. Ströbele, Chem. Ing. Tech. 36 (1964) 858.
140. Physical Constants of Hydrocarbons C1 to 163. H.-D. Wulf, T. Böhm-Gößl, L. Rohrschneider,
C10 , “ASTM Spec. Techn. Publ. STP” 106 Fette Seifen Anstrichm. 69 (1967) 32.
(1963). 164. California Research Corp., US 2 477 382, 1949
141. Handbook of Chemistry and Physics, R. C. (A. H. Lewis, A. C. Ettling); Chem. Abstr. 44
Weast (ed.), 63rd ed., CRC Press, Boca Raton (1950) 1137 e.
1982 – 1983, p. D-213. 165. H. Hartig, Chem. Ztg. 99 (1975) 175.
142. D. R. Stull, E. F. Westrum, G. C. Sinke: The 166. K. Ichikawa, H. Shingu, Kogyo Kagaku Zasshi
Chemical Thermodynamics of Organic 54 (1951) 183 – 184; Chem. Abstr. 47 (1953)
Compounds, J. Wiley & Sons, New York 8662 b.
1969, p. 370. 167. S. Murohashi, Nagaoka Kogyo Koto Senmon
143. J. C. Butler, W. P. Webb, Chem. Eng. Data Gakko Kenkyu Kiyo 9 (1973) 103 – 106;
Series 2 (1957) no. 1, 42 – 46. Chem. Abstr. 80 (1974) 132917 r.
144. G. Sorbe: Sicherheitstechnische Kenndaten 168. J. G. Matveev, D. A. Drapkina, R. L. Globus,
chemischer Stoffe, vol. 2, Tr. IREA 21 (1956) 83 – 89, Chem. Abstr. 52
ecomed-Verlagsgesellschaft mbH, Landsberg (1958) 15474 e.
1983, pp. 15 – 16. 169. A. D. Harford, H. W. Vernon, US 2 897 112,
145. P. R. Pujado, J. R. Salazar, C. V. Berger, 1956.
Hydrocarbon Process. 55 (1976) no. 3, 91. 170. British Petroleum Co., Ltd., DE-AS 1 094 035,
146. R. C. Canfield, T. C. Unruh, Chem. Eng. NY 1957 (A. Douglas, H. W. Vernon).
90 (1983) no. 3, 32. 171. N. K. Moshinskaya, V. S. Olifer, L. I. Zhuoiev,
147. Toyo Soda Manufacturing Co., SU 137 993, 1960.
DE-AS 2 418 038, 1974 (T. Yanagihera et al.). 172. E. Siggel, US 2 710 849, 1952.
148. Hitachi Chemical Co., JP-Kokai 77 139 030, 173. BASF, DAS 1 142 338, 1960 (W. Braun, E.
1974 (E. Kudo et al.); Chem. Abstr. 88 (1978) Hartwig, R. Krallmann).
120757 r. 174. FMC Corp., US 3 539 640, 1967 (M. I.
149. BASF, DE-OS 2 141 491, 1971 (F. Neumayr). Stewart, O. K. Carlson).
Hydrocarbons 55
175. Ashland Oil & Refining Co., US 3 529 944, 201. Monsanto Co., US 3 228 994, 1962 (W. D.
1967 (A. M. Leas). Robinson).
176. Japan Petrochemicals Co., Ltd., 202. N. M. Indyukov, R. I. Gasanova, Khim.
DE-OS 2 210 133, 1972 (A. Sato, Y. Aida, I. prom-st. (Moscow) 1975, no. 7, 555.
Shimizi). 203. Farbenfabr. Bayer AG, DE-AS 1 012 595, 1955
177. H. Winkler: Der Steinkohlenteer, Verlag (H. Holzrichter, O. Tegtmeyer).
Glückauf, Essen 1951, p. 143. 204. F. Ullmann, Justus Liebigs Ann. Chem. 332
178. J. Chipman, S. B. Peltier, Ind. Eng. Chem. 21 (1904) 48.
(1929), 1106. 205. E. I. El’bert, SU 215 911, 1968.
179. M. McEwen: Organic Coolant Handbook, 206. R. G. Ismailow, D. A. Aliev, Azerb. Neft. Khoz.
sponsored by U.S. Atomic Energy Comm., 44 (1965) no. 1, 30.
publ. by Monsanto Co., St. Louis, Mo., 1958. 207. L. Gattermann, R. Ehrhardt, Ber. Dtsch. Chem.
180. J. M. Cork, H. Mandel, N. Ewbank (North Ges. 23 (1890) 1226.
American Aviation Inc.), U.S. At. Energy 208. I. H. Gardner, P. Bergstrom, J. Am. Chem. Soc.
Comm. Sci. Rept. 5129 (1960) 1 – 28. 51 (1962) 3376.
Recalculated by A. N. Syverud, The Dow 209. Asahi Kasei Kogyo K. K., DE-AS 2 039 904,
Chemical Co. Thermal Laboratory, Midland, 1970.
Mich. 210. Shell Oil Co., US 3 145 237, 1962 (R. van
181. M. McEwen, E. Wiederhold, Nucl. Eng. Part Helden, B. Balder).
IV Symp. Ser. 5 (1959) no. 22, 9 – 15. 211. R. van Helden, G. Verberg, Rec. Trav. Chim.
182. M. J. Foole, Chem. Eng. Prog. 58 (1962) no. 4, Pays-Bas 84 (1965) 1263.
37 – 40. 212. Gulf Research & Developm. Co.,
183. Hydrocarbon Process. 46 (1967) no. 11, 187. US 3 401 207, 1966 (C. M. Selwitz).
184. S. Feigelman, C. B. O’Connor, Hydrocarbon 213. E. S. Rudakov, A. I. Lutsyk, R. I. Rudakova,
Process. 45 (1966) no. 5, 140 – 144. Kinet. Katal. 18 (1977) 525; Kinet. Catal.
185. Hydrocarbon Process. Pet. Refin. 42 (1963) (Engl. Transl.) 18 (1977) 441.
no. 11, 204. 214. M. K. Starchevskii, M. N. Vargaftik, I. I.
186. Bayer, BIOS Final Rep. Nr. 1787. Moiseev, Izv. Akad. Nauk SSSR 1979,
187. R. Egbert, Trans. AICHE vol. 34 (1938) 242;(Engl. Transl.) 227.
435 – 488. 215. Teijin Ltd., US 3 963 787, 1973 (Y. Ichikawa,
188. R. Dasgupta, B. R. Maiti, Ind. Eng. Chem. 25 T. Yamaji).
(1986) 381 – 386. 216. Ube Industries, JP 55 079 324, 1978.
189. Otkrytiya/Itobret, SU 1 165 677, 1985. 217. L. N. Rachkovskaya et al., Kinet. Katal. 18
190. BIOS Final Rep. Nr. 893. (1977) 792; Kinet. Catal. (Engl. Transl.) 18
191. Soc. Progil, US 3 227 525, 1964 (M. E. (1977) 660.
Degeorges, M. Jaymond). 218. Ube Industries, JP 80 141 417, 1979, Chem.
192. Soc. Progil, US 3 112 349, 1961 (M. E. Abstr. 94 (1981) 156 561 k.
Degeorges, M. Jaymond). 219. British Petroleum, DE-OS 1 909 045, 1969 (D.
193. Soc. Progil, DE-AS 1 249 244, 1961 (M. Fiolka, C. R. Jentsch).
Jaymond). 220. L. H. Slaugh, Tetrahedron 24 (1968) 4523.
221. L. H. Slaugh, J. A. Leonard, J. Catal. 13
194. Dow Chemical Co., US 1 938 609, 1933 (J. H.
(1969) 385.
Reilly).
222. G. N. Koshel et al., Dokl. Akad. Nauk SSSR
195. Monsanto Co., US 3 228 994, 1962 (W. D.
237 (1977) no. 1, 164 (Chem. Tech.); Chem.
Robinson, W. Groves).
Abstr. 88 (1978) 50 394 u.
196. Hawker Siddeley Nuclear Power Co.,
223. Phillips Petroleum, US 4 123 470, 1977 (T. P.
FR 1 272 068, 1961.
Murths).
197. Montecatini, BE 586 826, GB 865 302, IT
224. Farberov M I, SU 416 339, 1972; Chem. Abstr.
Anm. 1144/59, 1960, IT Prior. 1959 (C. Ferry,
88 (1978) 62 124 u.
B. Ponzini).
225. Standard Oil, US 4 218 572, 1978 (S. R.
198. Monsanto Co., US 2 143 509, 1939 (C.
Dolhyj, L. J. Velenyi).
Conover, A. E. Huff).
226. Y. Yamazaki, T. Kawai, A. Yagishita, Sekiyu
199. Montecatini, IT 586 113, 1962.
Gakkaishi 21 (1978) no. 4, 229.
200. Kambaror YU G; SU 6 79 565, 1977, Chem.
227. Koppers Co., US 3 059 987 1962 (H. P.
Abstr. 92 (1980) 58407 d.
Baumann).
56 Hydrocarbons
228. N. Hendrich, Melliand Textilber. 43 (1962) 249. H.-G. Franck, J. W. Stadelhofer: Industrielle
258. Aromatenchemie, Springer Verlag,
229. W. M. Meylan, P. H. Howard: Chemical Berlin – Heidelberg – New
Market Input/Output Analysis of Biphenyl and York – London – Paris – Tokyo 1987.
Diphenyl Oxide to Assess Sources of 250. Beilstein, 5, 586; 5 (1), 274; 5 (2), 494; 5 (3),
Environmental Contamination, Syracuse NY, 1776; 5 (4), 1834.
EPA Contract No. 68-01-3224, 1976. 251. Beilstein, 5, 515; 5 (1), 248; 5 (2), 410; 5 (3),
230. W. E. Bachmann, H. T. Carke, J. Am. Chem. 1355; 5 (4), 1532.
Soc. 49 (1927) 2089 – 98. 252. Beilstein, 5, 625; 5 (1), 300; 5 (2), 531; 5 (3),
231. L. W. Pickett, Proc. R. Soc. London A 142 1936; 5 (4), 2142.
(1933) 333 – 346. 253. Beilstein, 5, 685; 5 (1), 344; 5 (2), 609; 5 (3),
232. H. G. Römer, E. Hertel, Z. Phys. Chem. Abt. B 2276; 5 (4), 2463.
21 (1933) 292 – 296. 254. Beilstein, 5, 667; 5 (1), 327; 5 (2), 579; 5 (3),
233. R. J. Good, E. E. Hardy, A. M. Ellenburg, H. B. 2136; 5 (4), 2293.
Richards, Jr., J. Am. Chem. Soc. 75 (1953) 255. Beilstein, 5, 693; 5 (1), 343; 5 (2), 609; 5 (3),
436. 2279; 5 (4), 2467.
234. J. A. Ellard, W. H. Yanko: Thermodynamic
Properties of Biphenyl and the Isomeric Specific References
Terphenyls, Final Report IDO-11,008, 256. Rütgerswerke, DE 3 203 818, 1982 (K.
Monsanto Research Corp., U.S. At. Energy Stolzenberg, K. H. Koch, R. Marrett).
Comm. Contract AT (10-1)-1088,Oct. 31, 257. Nippon Steel, JP 82 98 219, 1980.
1963. 258. IG Farbenindustrie, DE 428 088, 1921.
235. R. O. Bolt, R. O. Johnson, R. S. Kirk: Organic 259. Selden, GB 318 617, 1929.
Materials for Use as Reactor Coolant 260. Standard Oil, US 2 966 513, 1957 (E. K.
Moderators, California Research Corp., Fields).
June 13, 1956. 261. Rütgerswerke, DE 1 262 268, 1959 (E. Marx,
236. P. L. Geiringer: Handbook of Heat Transfer H.-G. Franck, E. A. Dierk, K. Fiebach).
Media, Reinhold Publ., New York 1962, 262. Ullmann, 4th ed., 18, 685.
p. 173. 263. Farbwerke Hoechst, DE 2 209 692, 1972 (K.
237. W. H. Hedley, M. V. Mines, J. Chem. Eng. Eiglmeier); DE 2 343 964, 1973 (F.
Data 15 (1970) 122. Röhrscheid).
238. H. Mandel, N. Eubank: Critical Constants of 264. Institute of Chemical Sciences, Academy of
Diphenyl and Terphenyl, Report No. NAA-SR Sciences/Kazakh S.S.R., SU 1 084 275, 1982
5129, Atomics International, Dec. 1960, p. 18. (R. U. Umarova, A. K. Ergalieva, D. K.
239. N. Andrews, R. Ubbelohde, Proc. R. Soc. Sembaev, B. V. Suborov)
London 228 A (1955) 435 – 37. 265. BASF, DE 2 721 640, 1977 (A. Hettche, M.
240. E. Pajda, J. Rusin, Chemik Wkladka 1972, 422. Patsch); DE 2 653 346, 1976 (W. Eisfeld, W.
241. R. Balent, Proc. Am. Power Conf. 21 (1959) Disteldorf, A. Hettche).
120. 266. BASF, DE 2 721 639, 1977 (A. Hettche, M.
242. Soc. Progil, DE-AS 1 272 281, 1962 (M.
Patsch).
Jaymond).
267. Ullmann, 4th ed., 18, 693.
243. Farbenfabr. Bayer AG,
268. Farbwerke Hoechst, DE 2 304 873, 1973 (K.
GB 957 065 = FR 1 297 408, 1961.
Eiglmeier, H. Lübbers).
244. Hawker Siddeley Nuclear Power Co.,
269. Toyo Soda Mfg., JP 85 94 928, JP 85 139 630,
FR 1 272 068, 1961.
1983 (M. Kubo, H. Satsuka, Y. Tsutsumi);
245. D. B. Orechkin, SU 570 584, 1977.
EP 217 316, 1986 (M. Kubo, K. Kawabata, Y.
246. Monsanto Co., DE-OS 2 306 454, 1973 (J. K.
Tsutsumi).
Sears).
270. Coal Industry, EP 47 069, 1981 (J. G.
247. M. He, Huaxue Shiji 8 (1986) no. 1, 60 – 61;
Robinson, S. A. Brain).
Chem. Abstr. 105 (1986) 208 532.
271. Nippon Steel, JP 83 176 210, 1982.
272. Institute of Heteroorganic Compounds,
General References
Academy of Sciences/USSR, DE 3 526 155,
248. H.-G. Franck, G. Collin: Steinkohlenteer,
1985 (V. V. Korshak et al.).
Springer Verlag, Berlin – Heidelberg – New
York 1968.
Hydrocarbons 57
273. Sun Research and Development, DE 1593 398, 300. Agency of Industrial Sciences and Technology,
1966 (A. Schneider, E. J. Janoski, R. N. JP 81 82 430, 1979.
McGinnis). 301. Magyar Tudomanyos Akademia, HU 36 764,
274. G. Collin, M. Zander, Erdöl Kohle Erdgas 1983 (L. Simandi et al.).
Petrochem. 33 (1980) 560. 302. Nippon Shokubai Kagaku Kogyo,
275. BASF, DE 3 044 111, 1980 (H. Naarmann, K. JP 85 233 028, 1984 (T. Yoneda, S. Nakahara).
Penzien, D. Nägele, J. Schlag). 303. Dow Chemical, US 4 297 514, 1980 (W. M.
276. Ullmann, 4th ed., 21, 70. King).
277. R. Heering, P. Lobitz, K. Gehrke, Plaste 304. Hydrocarbon Research, DE 3 141 645, 1981.
Kautsch. 31 (1984) 412 – 413; J. Stein, G. (D. Huibers, R. Barclay, R. C. Shah).
Liebisch, S. Helling, Chem. Tech. (Leipzig) 31 305. Fuji-Xerox, JP 82 144 567, 1981; Ricoh,
(1979) 302 – 305. JP 79 109 438, 1978 (E. Akutsu, A. Kojima);
278. Gulf Research and Development, Oce-van der Grinten, DE 2 525 381, 1975 (J.
US 4 280 881, 1977 (J. C. Montagna, R. D. Hagendorn, H. Knibbe); EP 24 762, 1981
Galli). (M. T. J. Peters, J. A. Verbunt).
279. Dow Chemical, US 4 093 672, 1977 (Y. Ch. 306. Bayer, DE 2 430 476, 1974 (W. Draber, K. H.
Sun). Büchel, M. Plempel).
280. Rütgerswerke, DE 2 443 114, 1974 (K. Rühl, 307. Röhm, DE 3 439 484, 1984 (A. Riemann, W.
G. Storch, H. Steinke, B. Charpey). Ude).
281. American Cyanamid, US 4 145 409, 1976 308. Isovolta, EP 41 496, 1981 (W. Rieder, M.
(N. R. Pasarela). Fehrle).
282. Manufacture de Produits Pharmaceutique A. 309. Minnesota Mining and Mfg., EP 203 828, 1986
Christiaens, DE 2 060 721, 1970 (P. Vanhoff, P. (W. J. Schultz, G. B. Portelli, J. P. Tane).
Clarebout). 310. BASF, DE 3 209 426, 1982 (M. Patsch, H.
283. Takeda Chemical Industries, JP 82 123 157, Flohr).
1981. 311. BASF, DE 3 536 259, 1985 (M. Patsch et al.).
284. Rütgerswerke, DE 2 208 253, 1972 (E. 312. Rütgerswerke, DE 2 114 219, 1971 (C. Finger,
Pastorek, H. Miels, W. Orth). H.-G. Franck, M. Zander).
285. Pfizer, DE 2 244 556, 1973 (D. P. O’Shea, R. C. 313. Rütgerswerke, DE 2 166 862, 1971 (C. Finger,
Adams, S. Nakanishi). H.-G. Franck, M. Zander).
286. Meiji Seika Kaisha, DE 2 911 492, 1979 (Y. 314. Kureha Chemical Industry, JP 75 140 535,
Sekizawa et al). 1974 (M. Takahashi, A. Ito, A. Nobuo).
287. Wella, DE 3 215 341, 1982 (E. Konrad, H. 315. Kureha Chemical Industry, DE 2 447 894,
Mager). 1973 (Y. Shisakaiso, A. Ito, M. Takahashi).
288. Rütgerswerke, DE 2 635 401, 1976 (P. 316. Kureha Chemical Industry, DE 2 446 591,
Stäglich, D. Hausigk). 1974 (M. Takahashi et al.).
289. H. J. V. Winkler: Der Steinkohlenteer und 317. Rütgerswerke, DE 1 931 946, 1969 (C. Finger,
seine Bedeutung, Verlag Glückauf, Essen H.-G. Franck, M. Zander).
1951, p. 156. 318. Wellcome Foundation, EP 125 702, 1984
290. Nippon Steel, JP 82 98 220, 1980. (K. W. Bair).
291. Rütgerswerke, DE 3 243 981, 1982 (W. Orth, 319. Richardson Merell, US 3 907 791, 1975 (W. L.
E. Pastorek, W. Fickert). Albrecht, R. W. Fleming).
292. Celamerck, DE 2 816 388, 1979 (J. Gante et 320. Rütgerswerke, DE 1 126 372, 1958 (J.
al.). Altpeter, H. Sauer).
293. Squibb, US 3 856 977, 1974 (E. T. Stiller, S. D. 321. BASF, DE 3 218 229, 1982 (H. Naarmann, J.
Levine, P. A. Principe, P. A. Diassi). Schlag, G. Koehler).
294. Bayer, DE 3 530 819, 1985 (J. Hocker, L. 322. VEB Kombinat Mikroelektronik,
Rottmeier). DE 3 131 625, 1981 (H.-J. Deutscher et al.).
295. Ullmann, 4th ed., 23, 415. 323. Fuji Standard Research, EP 196 169, 1986 (S.
296. Nippon Kayaku, DE 2 325 302, 1973 (D. Otani, Y. Nagai).
Matsunaga, M. Sumitani). 324. Coal Industry, EP 36 710, 1981 (J. G.
297. Chemische Fabrik Kalk, DE 2 725 857, 1977 Robinson, D. I. Barnes).
(H. Jenkner). 325. Hercules, US 3 926 897, 1972 (C. L. Cessna).
298. Sandoz, DE 2 225 565, 1972 (H. P. Baumann). 326. Licentia, DE 2 406 540, 1974 (M. Saure, H.
299. Rütgerswerke, DE 2 704 648, 1977 (C. Finger). Wagner, J. Wartusch).
58 Hydrocarbons
327. Union Carbide, US 3 834 166, 1974 (R. A. 355. British Petroleum, GB 1 399 814, 1971 (A. J.
Cupper, G. S. Somckh). Burn, T. A. Lighting).
328. Bayer, DE 3 232 063, 1982 (H. Herzog, W. 356. Wellcome Foundation, EP 125 701, 1984
Hohmann). (K. W. Bair).
329. Burroughs Wellcome, US 4 511 582, 1983 357. Pentel, JP 8 632 352, 1976 (K. Takemura).
(K. W. Bair). 358. W. R. von Oettingen: “Toxicity and Potential
330. Rütgerswerke, DE 1 240 065, 1965 (E. Egerer, Dangers of Aliphatic and Aromatic
W. Fickert, L. Rappen, G. Grigoleit); Hydrocarbons,” Public Health Bull. (1940)
DE 3 305 528, 1983 (D. Tonnius, W. Orth, H. no. 255.
Miele, E. Pastorek). 359. E. A. Wahrenbrock, E. I. Eger, R. G. Laravuso,
331. US Steel, US 2 956 065, 1958 (C. W. De Walt, G. Maruschak, Anesthesiology 40 (1974)
K. A. Schowalter). no. 1, 19.
332. Mitsubishi Metal, JP 60 238 432, 1984 (H. 360. A. H. Nuckolls, Underwriters Laboratory
Yoshida, M. Morikawa, T. Iwamura). Report (Nov. 13, 1933) no. 2375.
333. Nihon Joryu Kogyo, Nippon Shokubai Kagaku 361. Y. Henderson, H. W. Haggard: Noxious Gases,
Kogyo, JP 85 255 747, 1984 (Y. Saito et al.). 2nd ed., Reinhold Publ. Co., New York 1943.
334. Nippon Shokubai Kagaku Kogyo, 362. J. C. Krantz, Jr., C. J. Carr, J. F. Vitcha, J.
JP 85 233 029, 1984 (I. Maeda, H. Mitsui, T. Pharmacol. Exp. Therap. 94 (1948) 315.
Nakamura). 363. T. Ikoma, Nichidai Igaku Zasshi 31 (1972)
335. Nippon Shokubai Kagaku Kogyo, no. 2, 71.
JP 8 112 335, 1979. 364. D. M. Aviado, S. Zakhari, T. Watanabe:
336. Merck, US 3 506 434, 1970 (E. Jacobi et al.). Non-fluorinated Propellants and Solvents for
337. Ullmann, 4th ed. 12, 614. Aerosols, CRC Press, Cleveland, Ohio 1977,
338. Agfa-Gevaert, EP 2 546, 1979 (F. C. pp. 49 – 81.
Heugebaert, W. Janssens). 365. F. A. Patty, W. P. Yant: “Odor Intensity and
339. Fuji Photo Film, GB 2 009 955, 1979 (M. Symptoms Produced by Commercial Propane,
Kitajima, H. Tachikawa, F. Shinozaki, T. Butane, Pentane, Hexane and Heptane Vapor,”
Ikeda). U.S. Bur. Mines, Rep. Invest. 1929 no. 2979.
340. PPG Industries, US 4 054 683, 1976 (G. W. 366. B. B. Shugaev, Arch. Environ. Health 18
Gruber). (1969) 878.
341. National Starch and Chemical, US 4 069 123, 367. Documentation of the Threshold Limit Values
1972 (M. M. Skoultchi, I. J. Davis). for Substances in Workroom Air, 3rd ed.,
342. Canadian Industries, US 4 012 280, 1976 American Conference of Governmental
(H. H. Holton). Industrial Hygiene, Cincinnati 1971.
343. Ciba-Geigy, CH 580 673, 1972 (W. Müller). 368. M. G. Rumsby, J. B. Finean, J. Neurochem. 13
344. Eli Lilly, DE 2 631 888, 1977 (J. B. Campbell). (1966) 1513.
345. EGIS, DE 3 618 715, 1986 (P. Benko et al.). 369. D. A. Haydon, B. M. Hendry, S. R. Levinson,
346. H. J. V. Winkler: Der Steinkohlenteer und J. Requena, Biochim. Biophys. Acta 470
seine Bedeutung, Verlag Glückauf, Essen (1977) 17.
1951, p. 220. 370. N. Battistini et al., Riv. Patol. Nerv. Ment. 95
347. Rütgerswerke, DE 2 914 670, 1979 (G.-P. (1974) 871.
Blümer, M. Zander). 371. G. W. Paulson, G. W. Waylonis, Arch. Intern.
348. Hoechst, DE 3 532 882, 1985 (S. Schießler, E. Med. 136 (1976) 880.
Spietschka). 372. N. W. Lazarew, Arch. Exp. Pathol. Pharmakol.
349. Rütgerswerke, DE 2 709 015, 1977 (J. Omran, 143 (1929) 223.
M. Zander). 373. E. T. Kimura, D. M. Elery, P. W. Dodge,
350. Hoechst, DE 3 134 131, 1981; DE 3 324 881, Toxicol. Appl. Pharmacol. 19 (1971) 699.
1983 (S. Schießler, E. Spietschka). 374. C. H. Hine, H. H. Zuidema, Ind. Med. 39
351. Ullmann, 4th ed., 17, 471 – 472. (1970) no. 5, 215.
352. Bayer, EP 20 817, 1979 (H. Harnisch). 375. H. H. Schaumburg, P. S. Spencer, Brain 99
353. Ricoh, JP 7 908 527, 1977 (A. Kojima, M. (1976) 182.
Hashimoto, M. Ohta). 376. K. Kurita, Jpn. J. Ind. Health Exp. 9 (1974)
354. Hoechst, DE 2 734 288, 1977 (W. 672.
Wiedemann); DE 2 726 116, 1977 (H. J. 377. A. K. Ashbury, S. L. Nielsen, R. Telfer, J.
Schlosser, E. Lind). Neuropathol. Exp. Neurol. 33 (1974) 191.
Hydrocarbons 59
378. E. G. Gonzales, J. A. Downey, Arch. Phys. Dept. of H. E. W., U.S. Government Printing
Med. Rehabil. 53 (1972) 333. Office, Washington, D.C., 1975.
379. H. E. Swann, Jr., B. K. Kwon, G. H. Hogan, 398. J. Opdyke, Food Cosmet. Toxicol. 12 (1974)
W. M. Snellings, Am. Ind. Hyg. Assoc. J. 35 707.
(1974) no. 9, 311. 399. J. M. Haas, H. W. Earhart, A. S. Todd: Book of
380. W. R. F. Notten, P. W. Henderson, Biochem. Papers, 1974 AATCC National Technical
Pharm. 24 (1975) 1093. Conference, American Association of Textile
381. C. F. Reinhardt et al., Arch. Environ. Health Chemists and Colorists, Research Triangle
22 (1971) 265. Park, N. C., p. 427.
382. A. Y. H. Lu, H. W. Strobel, M. J. Coon, Mol. 400. I. Hakkinen et al., Arch. Environ. Health 26
Pharamcol. 6 (1970) 213. (1973) 70.
383. N. W. Lazarew, Arch. Exp. Pathol. Pharmakol. 401. A. M. Seppalainen, I. Hakkinen, J. Neurol.
143 (1929) 223. Neurosurg. Psychiatry 38 (1975) 248.
384. Criteria for a Recommended Standard – 402. Patty’s Industrial Hygiene and Toxicology,
Occupational Exposure to Xylene, U.S. Dept. vol. 2 B, Toxicology, 3rd ed.
Health, Education and Welfare, Natl. Inst. 403. Z. F. Khromenko, Gig. Tr. Prof. Zabol. 1976,
Occup. Safety and Health, Washington, D.C. 44.
(1975). 404. Y. R. Adamson, J. L. Weeks, Arch. Environ.
385. O. A. Anunziata et al., Appl. Catal. 15 (1985) Health 27 (1973) no. 8, 68.
235 – 245. K. Verschueren: Handbook of 405. Y. R. Adamson, J. M. Furlong, Arch. Environ.
Environmental Data on Organic Chemicals, Health 28 (1974) no. 3, 155.
2nd ed., Van Nostrand Reinhold, New York 406. R. J. Hawkins, J. L. Weeks, At. Energy Can.
1983. Ltd. (Rep.) 4439 (1973); Chem. Abstr. 80
386. W. E. Coleman et al., Arch. Environ. Contam. (1974) 65 963.
Toxicol. 13 (1984) 171 – 178. 407. J. S. Henderson, J. L. Weeks, Ind. Med. Surg.
387. N. I. Sax: Dangerous Properties of Industrial 40 (1973) 10.
Materials, 6th ed., Van Nostrand Reinhold, 408. International Agency for Research on Cancer
New York 1984. (ed.): IARC Monographs on the Evaluation of
388. Threshold Limit Values and Biological the Carcinogenic Risk of Chemicals to
Exposure Indices for 1987 – 1988, American Humans, vol. 32, IARC, Lyon 1983.
Conference of Governmental Industrial 409. Phenanthrene, Dangerous Prop. Ind. Mat.
Hygienists. Cincinnati, Ohio, 1987. Rep. 6 (1986) 68 – 69.
389. DFG, MAK- und BAT-Werte Liste 2001, 410. E. E. Sandmeyer in G. D. Clayton, F. E.
Mitteilung 37, Wiley-VCH, Weinheim, 2001. Clayton (eds.): Patty’s Industrial Hygiene and
390. H. W. Gerarde: Toxicology and Biochemistry Toxicology, 3rd ed., vol. 2 B, J. Wiley & Sons,
of Aromatic Hydrocarbons, Elsevier New York 1981, p. 3353.
Publishing Co., London 1960. 411. Y. S. Rotenberg, F. D. Mashbits, Gig. Tr. Prof.
391. G. Hommel: Handbuch der gefährlichen Zabol. 9 (1965) 53 – 54.
Güter, Springer Verlag, Berlin Heidelberg 412. H. W. Gerarde: Toxicology and Biochemistry of
1987. Aromatic Hydrocarbons, Elsevier Publishing,
392. H. W. Gerarde, Am. Med. Assoc. Arch. Ind. Amsterdam – London – New York 1960.
Health 19 (1959) 403. 413. S. D. Keimig, R. W. Pattillo, Brookhaven Natl.
393. M. A. Wolf et al., Am. Med. Assoc. Arch. Ind. Lab., Rep. BNL-51885, Upton 1984.
Health 14 (1956) 387. 414. Z. Matijasevic, E. Zeiger, Mutat. Res. 142
394. H. W. Werner, R. C. Dunn, W. F. von (1985) 149 – 152.
Oettingen, Ind. Health Toxicol. 26 (1974) 264. 415. M. Sakai, D. Yoshida, S. Mizusaki, Mutat.
395. Maximale Arbeitsplatzkonzentrationen und Res. 156 (1985) 61 – 63.
biologische Arbeitsstofftoleranzwerte, Dt. 416. J. E. Rice, T. J. Hosted, E. J. Lavoie, Canc.
Forschungsgemeinschaft, Verlag Chemie, Lett. (Shannon Irel.) 24 (1984) 327 – 333.
Weinheim 1984. 417. F. Palitti et al., Mutat. Res. 174 (1986)
396. N. I. Sax: Dangerous Properties of Industrial 125 – 130.
Materials, 6th ed., van Nostrand Reinhold. 418. J. R. Babson et al., Toxicol. Appl. Pharmacol.
New York 1984, p. 819. 85 (1986) 355 – 366.
397. H. E. Christensen, T. T. Luginbyhl: Registry of 419. R. P. Bos et al., Mutat. Res. 187 (1987)
Toxic Effects of Chemical Substances, U.S. 119 – 125.
60 Hydrocarbons
Hydrochloric Acid
Severin Austin, Hüls Aktiengesellschaft, Marl, Federal Republic of Germany
Arndt Glowacki, Hüls Aktiengesellschaft, Marl, Federal Republic of Germany
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
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
The most important thermodynamic data are (i.e., the total heat of solution ∆sol H HCl ), the
as follows [8]: partial molar enthalpy of solution
Enthalpy of formation ∂∆sol HHCl
∆H 0f (298.15 K) −92.31 kJ/mol ∆sol HHCl =
∂nHCl p,T ,nH2 O
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) and the partial molar enthalpy of dilution
Specific heat cp (298.15 K) 29.12 J mol−1 K−1
Free energy of formation ∂∆sol HHCl
∆G 0 (293.15 K) −95.25 kJ/mol ∆dil HHCl =
∂nH2 O p,T ,nHCl
p, MPa Temperature, ◦ C
−10 0 10 20 30 40 50 60
Table 3. Density of hydrochloric acid, in grams per cubic centimeter, as a function of temperature
tration, wt % 0 10 20 30 50 75 100
The viscosity η is considerably higher than that such as chlorine or chlorinated organic com-
of pure water. For 10 wt % acid, the viscosity pounds.
varies with temperature as follows: In nature, hydrogen chloride is liberated in
volcanic eruptions or other releases of subter-
t, ◦ C 0 10 20 ranean gas. The amount of hydrogen chloride
η, mPa · s 1.89 1.45 1.16 produced can be of the same order as that of
sulfur dioxide. Hydrochloric acid occurs in the
The variation of viscosity with acid concen- gastric juice of mammals; in the human stomach,
tration is given below (20 ◦ C): 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)
temperature and pressure in the pipeline. Here Sulfuric acid is now the exclusive starting ma-
the pressure must be low enough, or the temper- terial. In the Mannheim process, sodium sulfate
ature high enough, to prevent condensation of is produced; in the Berlin hydrochloric acid pro-
aqueous hydrochloric acid (see Fig. 10 and Ta- cess, sodium hydrogen sulfate.
ble 2). The steel furnaces operated at Hüls can In the Mannheim process, brick-lined exter-
each produce 30 – 40 t/d hydrogen chloride. The nally fired muffle furnaces are used. These con-
hydrogen chloride gas is generated at a pressure tain a stirrer with a scraping action which pre-
up to 0.7 MPa, the water content being < 50 ppm vents agglomeration of the pasty mass and also
(by volume); steel pipes are, therefore, used to removes the sodium sulfate produced. Hydro-
convey it from the plant. gen chloride passes through the side of the muf-
Chemically related processes [19] include the fle and is propelled by a fan to the next stage,
reaction of chlorine with carbon (in the form of the furnace being maintained at a slightly re-
coke) and water vapor duced pressure. The mixture of gases exhausted
from the furnace contains up to 85 % hydrogen
C + 2 H2 O + 2 Cl2 −→ 4 HCl + CO2 (10) chloride, as well as air, sulfuric acid mist, and
fine particles of salt. As a rule, the gas must
or the reaction of chlorine with sulfur dioxide
undergo several purification stages such as fil-
and water
tration through coke and activated carbon, wet
SO2 + 2 H2 O + Cl2 −→ 2 HCl + H2 SO4 (11) 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
3.2.2. Reaction of Metal Chlorides with sodium hydrogen sulfate at about 300 ◦ C in cast-
Sulfuric Acid iron retorts. Alternatively, potassium chloride
can be used, to give hydrogen chloride and potas-
The sulfate process has a high energy consump- sium hydrogen sulfate. This process has retained
tion. It is, therefore, gradually decreasing in some commercial importance because the prin-
importance compared with other processes, es- cipal products are the potassium salts formed
pecially organic chlorination, which inevitably (i.e., KHSO4 and K2 S2 O7 ) [21], whereas hy-
produces large amounts of hydrogen chloride drogen chloride is considered as a byproduct.
as a byproduct. Moreover, the old argument Gas from the retorts is more concentrated be-
that sulfate–hydrochloric acid provides a prod- cause the equipment is more leak proof, and less
uct with higher purity no longer holds; byprod- sulfuric acid mist is present as a result of the
uct acid can now be purified sufficiently to sat- lower working temperature.
isfy even requirements for high quality. Hence,
only the most important sulfate–hydrochloric
acid process is described here. Further details 3.2.3. Recovery as a Chlorination Byproduct
are given in [20].
The reaction of common salt with con- By far the largest amounts of hydrogen chlo-
centrated sulfuric acid according to Equa- ride and hydrochloric acid are produced as by-
tion (2) occurs at comparatively low temperature products of chlorination. Equation (6) is merely
(150 – 300 ◦ C); hydrogen chloride and sodium an example of the varied range of industrial
hydrogen sulfate are formed. The latter reacts chlorination processes. Nevertheless, the ma-
with excess sodium chloride at a minimum tem- jor source of hydrogen chloride is the crack-
perature of 550 – 600 ◦ C, according to Equa- ing of 1,2-dichloroethane to give vinyl chlo-
tion (3), forming neutral sodium sulfate. The use ride (→ Chlorinated Hydrocarbons). The quan-
of sodium hydrogen sulfate as a starting material tity of hydrogen chloride produced and con-
(Eq. 3) is quite feasible but has been discontin- sumed can be extremely large, in many installa-
ued because the high temperature (600 – 800 ◦ C) tions reaching the order of 10 000 m3 /h (16 t/h or
required and the high energy consumption ren- 140 000 t/a) and more. Consequently, hydrogen
der the process uneconomic. chloride is often fed directly to a chemical plant
Hydrochloric Acid 9
which uses it as a raw material; installed stor- pressure (usually 1 – 2 MPa) for subsequent pro-
age capacities for hydrogen chloride are usually cessing, thereby avoiding the necessity for sub-
comparatively small. It is, therefore, essential to sequent mechanical compression of the gas.
operate the plants at a balanced rate of produc- Aqueous absorption of hydrogen chloride
tion and consumption. If production problems gives hydrochloric acid as an intermediate prod-
or market variations affect one of the plants or uct and leads to inevitable problems with con-
products, the other plant can only continue to op- struction materials. These problems have be-
erate if reserve production capacity (normally come less severe since the introduction of flu-
not fully used) is installed. Buffer storage for oropolymers. In many installations, however,
liquid hydrogen chloride is used to overcome aqueous treatment has been replaced by non-
short-term inbalances in production versus con- aqueous methods (e.g., fractional distillation).
sumption. The absorption of hydrogen chloride can be ef-
Hydrogen chloride produced by organic chlo- fected with water or 20 wt % hydrochloric acid.
rination can be treated in various ways: The use of water as the absorption medium is
appropriate if the primary aim is the produc-
1) simple isolation of the gas by condensing the
tion of 30 – 35 wt % hydrochloric acid as the
chlorinated hydrocarbons;
end product. If demand for the latter is insuffi-
2) isolation and purification of hydrogen chlo-
cient, the 20 % azeotrope is used for the absorp-
ride by distillation;
tion of hydrogen chloride, thereby concentrating
3) aqueous absorption of hydrogen chloride.
the acid to 30 – 35 wt %. The concentrated hy-
Condensation of the chlorinated hydrocar- drochloric acid is first purified, if necessary. It is
bons is used when the presence of residual chlo- then distilled to give gaseous hydrogen chloride
rinated hydrocarbons or other residues from the and azeotropic hydrochloric acid. The hydrogen
chlorination process does not cause any prob- chloride is dried and delivered to the consumer;
lems in the hydrogen chloride, which is fed the azeotrope is recycled to the absorption stage.
as a gas. The degree of purification attainable If the presence of chlorinated hydrocarbons in
depends on condensation pressure and tem- hydrochloric acid is undesirable, they can be re-
perature. One example is the chlorination of moved by introducing a stripping step. A sim-
chloromethane to dichloromethane; the hydro- plified scheme is shown in Figure 11. A small
gen chloride formed is used only to convert amount of water is fed to the absorption column
methanol to chloromethane, which in turn is to render the gas stream free of hydrogen chlo-
used to produce dichloromethane: ride. A corresponding amount of acid must be
separated for other use.
CH3 Cl + Cl2 −→ CH2 Cl2 + HCl (12) A number of process and equipment param-
CH3 OH + HCl −→ CH3 Cl + H2 O (13) eters are important for assessing the economics
of the aqueous absorption of hydrogen chloride
The chloromethane and dichloromethane con- from a production gas stream:
tent of hydrogen chloride produced according
to Equation (12) does not have any adverse ef- 1) The concentration of hydrogen chloride in
fect on the reaction of methanol with hydrogen the gas stream determines the pressure and
chloride (Eq. 13). temperature that must be maintained in the
Fractional distillation of the liquefied gas absorption column.
may be employed if a purer product is required. 2) The energy requirement for liberating pure
For example, the gaseous products from the hydrogen chloride gas decreases with in-
cracking of 1,2-dichloroethane are liquefied and creasing concentration of the concentrated
then separated by distillation into vinyl chloride acid formed.
and hydrogen chloride. Liquefaction of these 3) Production of concentrated hydrochloric
gases by reducing the temperature to extremely acid from a gas stream with a low hydro-
low values is uneconomical, so condensation gen chloride concentration requires compar-
and distillation are carried out at elevated pres- atively great effort because either the pres-
sure. An advantage of this procedure is that hy- sure must be increased, the temperature re-
drogen chloride gas is produced at a high enough 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- of equipment is suitable only for operation at at-
ter vapor in the production gas stream can, mospheric or possibly slightly higher pressure.
under certain circumstances, render the pro- The same is true for the chlorinated hydro-
duction of concentrated hydrochloric acid carbon stripping column. If the concentration
impossible. of chlorinated hydrocarbons is low, or if they
have been effectively removed from the acid,
Industrial processes have been reviewed by other materials such as poly(vinyl chloride) can
Schmidt [10] and Vieweg [22]. The latter pro- be used for storage vessels. The desorption col-
poses the use of magnesium chloride solution to umn, however, is made of graphite because of
absorb low levels of hydrogen chloride from gas the high operating temperature. Alternatively,
containing a high level of water vapor. rubber-lined columns with inner brick linings
The choice of construction materials for hy- can be used. Steel columns lined with perflu-
drogen chloride absorption equipment is re- oropolymer, especially polytetrafluoroethylene
stricted because of the possibility of attack both (PTFE), have been used with generally excel-
by concentrated hydrochloric acid and by chlo- lent results, although differences exist among the
rinated hydrocarbons. Graphite heat exchangers products from different manufacturers with re-
are used to cool the acid. For a long time a special gard to microporosity, tear resistance, and flow
ceramic was the only economically viable ma- properties. Compared with graphite columns,
terial for the pumps. The distillation column is coated steel columns have the advantage that
lined with acid-proof bricks or graphite blocks, they can operate at elevated pressure, but failure
and the bricks are cemented with phenolic resin. of a PTFE lining would lead to rapid destruc-
Frequently, a special acid-resistant rubber lining tion of the steel casing. However, these columns
is applied between the steel casing and the brick can be constructed of fairly small individual
wall to prevent attack by acid that might pen- elements that may be replaced quickly, reduc-
etrate pores and cracks. Vessels and equipment ing maintenance downtime. Heat exchangers are
made entirely of phenolic resin are chemically made of graphite in all cases. Another possibil-
very resistant, but brittle and mechanically frag- ity is to use tantalum for lining and evapora-
ile, and require support (e.g., by an outer shell of tors, although special precautions must be taken
glass-fiber-reinforced polyester resin). This type to avoid embrittlement due to absorption of hy-
Hydrochloric Acid 11
drogen. Furthermore, hydrogen fluoride must be reaction (∆S 0r = − 64.7 J K−1 mol−1 ). With
absent. increasing temperature, the equilibrium
Completely dry hydrogen chloride will not shifts from right to left, thus favoring forma-
attack steel pipes or compressors. Hydrogen tion of hydrogen chloride. The temperature
chloride gas was formerly dried by using con- is maintained as high as construction mate-
centrated sulfuric acid, but another approach is rials permit, to prevent formation of chlo-
to make use of the properties of the HCl – H2 O rine; even if the partial pressure of water is
system (see Fig. 10 and Table 2). At low temper- kept high and the oxygen excess low, a tem-
ature, the partial vapor pressure of water over perature > 1000 ◦ C is necessary to shift the
highly concentrated hydrochloric acid is very equilibrium to the left. This keeps the chlo-
low; hydrogen chloride can therefore be dried by rine at an acceptable level and also ensures
cooling to sufficiently low subzero temperature. complete combustion.
At slightly elevated pressure (pabs ≈ 0.3 MPa), it 2) Reaction temperature must not exceed
is sufficient to attain −10 ◦ C to give satisfactory 1200 ◦ C to prevent formation of nitrogen
results. Hydrochloric acid mist must, however, oxides, which would contaminate the hy-
be removed completely, or severe corrosion can drochloric acid product and, which is worse,
result. cause waste gas purification problems. A
possible solution is to use pure oxygen in-
stead of air [16].
3.2.4. Recovery from Waste Incineration
For these reasons, incineration of waste chlo-
All chlorination processes produce undesirable rinated hydrocarbons is carried out between
chlorinated hydrocarbons in addition to the 1000 and 1200 ◦ C. Cooling the flue gases com-
product desired. These unusable and often nox- bined with steam generation can cause prob-
ious substances must be disposed of for environ- lems with construction materials. The absence
mental reasons. One possibility is incineration of corrosive combustion products must, there-
to hydrogen chloride, water, and carbon diox- fore, be assured. The treated wastes should have
ide; the hydrogen chloride should be recovered a well-known composition (down to trace con-
for both economic and environmental reasons. taminants) for proper prediction of the expected
Although the methods used vary in detail combustion products, their concentration, and
[23–25], the general principle remains the same. behavior. Slow cooling can cause chlorine to re-
Combustion takes place in a refractory-lined fur- form, so rapid cooling is preferred.
nace at > 1000 ◦ C. Waste chlorinated hydro- Hydrogen chloride is recovered from the
carbons containing up to 70 wt % chlorine nor- combustion gases by absorption, i.e., the process
mally possess good combustion properties, but is similar to that used in recovery from chlori-
any further increase in chlorine content reduces nation processes, except for the following:
the calorific value to such an extent that fuel oil
1) A dry process is impossible because water
or gas must be added to support combustion.
vapor is always present.
Optimum reaction conditions lie within the
2) The gases should be condensed in such a way
following limits:
as to form the most concentrated hydrochlo-
1) Excess oxygen must always be present to ric acid possible so that gaseous hydrogen
guarantee complete decomposition of chlori- chloride or concentrated hydrochloric acid
nated hydrocarbons. The hydrogen chloride can be obtained. Dilute hydrochloric acid has
gas formed is consequently able to react with little commercial value.
this oxygen to form chlorine and water in ac- 3) Special corrosion problems must be over-
cordance with the Deacon equation: come at places of possible, but not neces-
sarily regular, acid condensation.
−→ 4) If the gaseous combustion products are used
2 HCl + 1/2 O2 ←− Cl2 + H2 O (14) directly, the operating temperature must al-
The reaction has a negative enthalpy of reac- ways be kept above the condensation tem-
tion (∆H 0r = − 57.1 kJ/mol) and entropy of perature of hydrochloric acid.
12 Hydrochloric Acid
One of the oldest plants producing commer- iron; thus, steel equipment and pipelines can be
cial quantities of 33 wt % hydrochloric acid by used. Hydrogen chloride gas is regarded as dry
incinerating chlorinated hydrocarbons is that of if no danger exists of hydrochloric acid conden-
Atochem at St. Auban in the south of France sation at operating temperature and pressure.
[25]. This plant opened in 1975 and consumes Hydrochloric acid has no generally accepted
approximately 16 000 t of chlorinated hydrocar- quality specification. In the Federal Repub-
bons material per year. Many additional chlo- lic of Germany, the only official standard is
rinated hydrocarbon incineration plants have DIN 19 610 (November 1975), which gives the
since been built, and the hydrochloric acid pro- following specifications for hydrochloric acid
duced is finding a wider range of uses (e.g., ad- for boiler feed water treatment:
dition to ethylene) [27].
Hydrogen chloride content ≥ 30 wt %
Sulfate (SO2−
4 ) content ≤ 0.5 wt %
Iron (Fe3+ ) content ≤ 0.002 wt %
3.3. Purification Content of organic chlorine compounds (ex-
Many impurities in hydrogen chloride gas (e.g., pressed as
SO2 , As, and Cl2 ) can be removed by adsorp- Cl− ) ≤ 0.02 g/L
tion on activated carbon [28]. As the importance Quality requirements for medicinal use are
of the sulfate process decreases, the removal of given in DAB 9 [29] for both concentrated
chlorinated hydrocarbons from hydrogen chlo- (36 wt %) and dilute (10 wt %) hydrochloric
ride gas or hydrochloric acid is of greater prac- acid. Prescribed purity tests include free chlorine
tical relevance. Gaseous hydrogen chloride can (4 ppm), arsenic (2 ppm), heavy metals (2 ppm),
be purified by low-temperature scrubbing with sulfate (20 ppm), and residue on evaporation
a high-boiling solvent, which may be another (0.01 %). Other countries have their own reg-
chlorinated hydrocarbon (e.g., hexachlorobuta- ulations [30].
diene or tetrachloroethane), or special oil frac-
tions. The use of activated carbon is often un-
necessary after such treatment. Chlorine may be 5. Storage and Transportation
removed with carbon tetrachloride, in which it
is much more soluble than is hydrogen chloride Wrought- and cast-iron, steel, and steel alloys
gas. have good resistance toward dry hydrogen chlo-
Hydrochloric acid, used as such or for gen- ride gas or liquid. Steel is, therefore, suitable for
eration of hydrogen chloride, contains mainly equipment and containers up to 280 ◦ C, includ-
volatile impurities such as chlorinated hydro- ing pressure vessels used for storage and trans-
carbons. In these cases the impurities may be portation of dry hydrogen chloride.
removed from the acid by stripping (see Fig. 11). Even traces of moisture significantly increase
Stripping with an inert gas stream, which results the corrosion rate, and alloyed austenitic steels
in lower energy consumption, is possible, al- become liable to stress crack corrosion. If the
though in most cases heating of the gas is prefer- working temperature cannot be raised above the
able for environmental reasons. Inorganic im- condensation point, either nonmetallic materi-
purities, in particular iron, are removed by ion als or more exotic metals must be used, such
exchange. as nickel–molybdenum alloys with high molyb-
denum content, zirconium, niobium, tantalum,
or noble metals such as gold or the platinum-
4. Quality Specifications group metals. In practice, it is imperative to spec-
ify precisely the operating limits for metals to
The quality specifications for gaseous hydrogen be used with hydrochloric acid or moist hydro-
chloride depend on intended use. There are no gen chloride. In all cases, limits are given for
general quality standards. Only a trace (a few acid concentration or temperature. Contamina-
parts per million) of water is usually permitted, tion with oxidizing or complexing agents must
particularly if the gas is to be compressed and be avoided. Traces of fluoride ions in hydrochlo-
liquefied. Dry hydrogen chloride does not attack ric acid, for example, can lead to rapid corrosion
Hydrochloric Acid 13
of zirconium or tantalum, although tantalum has kaline products or waste materials, acidification
proved suitable for hydrochloric acid distillation of brine in chlor-alkali electrolysis, and many
columns in the absence of damaging impurities. others.
A wide choice of nonmetallic construction Hydrochloric acid and hydrogen chloride
materials is available. For high temperature, ce- production figures for various countries are
ramics, graphite, or polytetrafluoroethylene may given in Table 4.
be suitable. Phenolic resins are used as cement
for brick lining. At lower acid concentration (be-
low 25 %), enameled or glass apparatus can be 7. Toxicology and Occupational
used. At lower temperature, (e.g., for storage and
transportation), vessels may be made of rubber-
Health
lined steel, polyolefin, or poly(vinyl chloride),
Hydrogen chloride causes severe irritation of
the latter usually with glass-fiber reinforcement.
the eye and respiratory tract and, to a lesser
If the acid contains traces of chlorinated hydro-
extent, the skin. The vapor can produce kera-
carbons such as dichloromethane, vessels fabri-
toconjunctivitis. Inhalation causes irritation and
cated in, or lined with, phenolic resin are pre-
damage to mucous membranes. Hydrogen chlo-
ferred. Plastics such as poly(vinyl chloride) and
ride is usually detectable by odor at 1 – 5 ppm
polyethylene can withstand use for several years.
and becomes objectionable at 5 – 10 ppm. Pro-
For transportation of hydrochloric acid and
longed occupational exposure to excessive hy-
liquefied hydrogen chloride, special labels must
drogen chloride concentration causes increased
be used that comply with the international reg-
incidence of chronic bronchitis or stomach and
ulations for transportation of hazardous goods:
intestinal disorders due to abnormal acid levels.
Hydrochloric acid: Hazard code number 80 Increased incidence of a characteristic dental de-
UN no. 1789 cay has also been observed with prolonged expo-
Warning sign 8 (causes burns)
sure, even at apparently tolerable levels. A con-
CFR 49: 172.101 Cor. M
RID/ADR, class 8, no. 5 b centration > 10 ppm is strongly irritating, even
Hydrogen chloride: Hazard code number 286 when some acclimatization has occurred [33].
UN no. 1050 Suitable protective goggles should be worn
Warning signs 6.1 (poisonous) and
8 (causes burns) 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-
6. Uses and Economic Aspects lic of Germany, code letter B, code color gray),
Hydrochloric acid and hydrogen chloride are as is the hydrogen chloride and sulfur dioxide
two of the most important basic industrial chem- filter (code letter E, code color yellow [34] ).
icals. In the Federal Republic of Germany,
The largest proportion of hydrogen chloride hydrogen chloride and hydrochloric acid
is used immediately by the producer himself, (> 10 wt %) are controlled by the Verordnung
because the hydrogen chloride byproduct of or- über gefährliche Stoffe (GefStoffV, regulations
ganic chlorination processes has to be used as for the control of dangerous substances), which
completely as possible. This is done by means have been adjusted to the corresponding direc-
of oxychlorination or hydrochlorination, or to a tives of the EEC [35]:
small extent, recovery of chlorine (e.g., by the 1) Anhydrous hydrogen chloride: hazard sym-
KEL process, a modified Deacon process [31] bol C (corrosive). Hazard information (“R”
or by electrolysis of hydrochloric acid [32]). notices): 35 (causes serious burns); 37
Aqueous hydrochloric acid has many uses as (causes irritation of respiratory tract). Safety
a strong inorganic acid: manufacture of chlo- Instructions (“S” notices): 7/9 (keep con-
rides, dissolution of minerals, pickling and etch- tainers tightly closed and store in a well-
ing of metals, regeneration of ion-exchange ventilated place); 26 (on contact with the eye,
resins for water treatment, neutralization of al-
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- to guidelines issued by the VDI, 2310 sheet 4,
cal advice). sensitive vegetation such as lettuce, strawber-
2) Hydrochloric acid containing over 25 wt % ries, spruce, birch, and apple trees should not
hydrogen chloride: hazard symbol C (corro- be exposed to a hydrogen chloride concentra-
sive, causes burns).“R” notices: 34 (causes tion higher than 0.8 mg/m3 as a daily average or
burns); 37 (see above).“S” notices: 2 (keep 0.1 mg/m3 as a monthly average.
away from children), 26 (see above). The concentration of hydrogen chloride in in-
3) Hydrochloric acid with 10 – 25 wt % hydro- dustrial waste gases must (according to regula-
gen chloride: hazard symbol Xi , irritant.“R” tions in the Federal Republic of Germany) nor-
notices: 36/38 (causes eye and skin irrita- mally not exceed 30 mg/m3 , although the limit is
tion)“S” notices: 2 (see above), 28 (on con- 50 mg/m3 for waste incineration plants; the cost
tact with the skin, immediately wash with of gas cleaning can, therefore, vary considerably.
copious amounts of water). A simple water scrubbing operation is often suf-
In the event of skin or eye contact, the af- ficient, but more than one stage may be neces-
fected area should be rinsed immediately and sary, sometimes including a final scrubbing with
thoroughly with a large amount of water. Con- an alkaline solution. In many cases it is not possi-
taminated clothing should be treated similarly, ble to produce a commercial hydrochloric acid.
and alkaline cleansing solution may also be used. The dilute acid is then neutralized before be-
For work with hydrogen chloride, especially at ing discarded in the wastewater. Neutralization
elevated pressure, breathing apparatus with an with caustic soda occasionally results in a salt
independent air supply and full protective cloth- solution that is pure enough to be recycled to a
ing must be readily available in case of accident. chlorine plant.
In the Federal Republic of Germany, the
MAK value for hydrogen chloride is 5 ppm by
volume, corresponding to 7 mg/m3 [36]. The 8. References
same values are valid as TLV in the United
States. In the Soviet Union 3 ppm (5 mg/m3 ) has 1. H. Chihara, A. Inaba, J. Chem. Thermodyn. 8
been established. (1976) 915 – 934.
In clean air regulations (TA Luft, 1986), con- 2. S. Miskiewicz, K. Rieser, T. Dorfmüller, Ber.
centration limits for hydrogen chloride in the Bunsenges. Phys. Chem. 80 (1976) 395 – 405.
atmosphere are 0.1 mg/m3 as an annual aver- 3. K. Clusius, G. Wolf, Z. Naturforsch. A (1947)
age and 0.2 mg/m3 for short periods. Phytotox- 495.
icological investigation has shown that plants 4. L. S. Adler, C. L. Yaws, Solid State Technol.
19 (1976) 35 – 38.
have differing tolerances [37–39]. According
Hydrochloric Acid 15
5. K. K. Tremper, J. M. Pransnitz, J. Chem. Eng. 24. Hoechst, US 4 073 871, 1977 (W. Opitz, H.
Data 21 (1976) 295 – 299. Hennen); DT 2 611 671, 1976.
6. W. Gerrard, E. D. Macklen, Chem. Ind. 25. Rhône-Poulenc, Company brochure,
(London) 1959, 1070 – 1071. Combustion of Chlorinated Residues (1978).
7. S. M. Khodeeva, M. B. Rozovskii, Russ. J. 26. Stauffer Chemical O., Hydrocarbon Process.
Phys. Chem. (Engl. Trans.) 49 (1975) (1981) 170.
824 – 827. 27. Uhde GmbH, EP 0 137 160, 1985, 1987 (U.
8. Codata Recommended Key Values, J. Chem. Rieker, G. Link).
Thermodyn. 10 (1978) 904. 28. Ullmann, 3rd ed., 15, 80.
9. D. D. Wagman, W. H. Evans, I. Halow, V. B. 29. DAB, vol. 9.
Parker, S. M. Bailey, R. H. Schumm, NBS 30. Europäisches Arzneibuch, vol. 2, 137 – 138,
Tech. Note (U.S.) 270 (1965) no. 1, 25 – 27. Deutscher Apotheker-Verlag, Stuttgart 1975.
10. A. Schmidt, Chem. Ing. Tech. 25 (1953) 31. Ullmann, 4th ed., 9, 357.
455 – 466. 32. Ullmann, 4th ed., 9, 355.
11. Ullmann, 3rd ed., 15, 70. 33. D. Henschler (ed.): Gesundheitsschädliche
12. J. J. Fritz, C. R. Fuget, Ind. Eng. Chem. Data 1 Arbeitsstoffe, Verlag Chemie GmbH,
(1956) 110 – 112. Weinheim 1972 – 1979.
13. J. F. F. Kao, J. Chem. Eng. Data 15 (1970) 34. Datenbank für wassergefährdende Stoffe
362 –367. (DABAWAS) vom 23. 03. 1987, 2989.
14. B. G. Staples, J. M. Procopio, G. J. Su, Chem.
35. Verordnung über gefährliche Arbeitsstoffe
Eng. (N.Y.) 77 (1970) 113 – 115.
vom 26. 08. 1986 (BGBL 1, S. 1470).
15. W. Kindler, G. Wuster, H. Rau, Ber.
36. Mitteilung 23 der Senatskommission zur
Bunsenges. Phys. Chem. 82 (1978) 543 – 545.
Prüfung gesundheitsschädlicher Arbeitsstoffe,
16. R. W. Bach, H. A. Friedrichs, H. Rau, High
VCH Verlagsgesellschaft mbH, Weinheim
Temp. High Pressures 9 (1977) 305 – 312.
17. O. F. Raper, C. B. Farmer, R. A. Toth, B. D. 1987.
Robbins, Geophys. Res. Lett. 4 (1977) 37. H. J. Grecelius, W. Forwer, R. Zahn, Staub,
531 – 534. Reinhalt. Luft 36 (1976) 201 – 205.
18. Ullmann, 4th ed., 10, 431 – 433. 38. H. van Haut, Schriftenr. LJB Landesanst.
19. Ullmann, 3rd ed., 15, 75. Immissions Bodennutzungsschutz Landes
20. Ullmann, 3rd ed., 15, 71 – 73. Nordrhein Westfalen 33 (1974) 39 – 43.
21. Ullmann, 4th ed., 13, 482. 39. C. Holsenberg, Schriftenr. Ver. Wasser Boden
22. R. Vieweg, Chem. Tech. (Leipzig) 12 (1960) Lufthyg. 41 (1974) 267 – 284.
no. 4, 188 – 195. 40. K. H. Büchel, H.-H. Moretto, P. Woditsch:
23. J. C. Zimmer, R. Guaitella, Hydrocarbon Industrial Inorganic Chemistry, Wiley-VCH,
Process. (1976) 117 – 118. 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
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
1. Introduction
gle units go from 0.1 to 7200 m3 /h, and pressure solids from solids in a suspension according to
drops vary from 0.34 to 6 bar. The efficiency, particle size. Sometimes, either coarse or fine
reported as cut size, lies between 2 and 250 µm. particles may be removed from the product, and
The main advantages of the hydrocyclone are hydrocyclones have performed such operations
its simple, compact, and inexpensive construc- known as degritting or refining, and desliming
tion, its ease of operation (no moving parts), its or washing, respectively. This classifying char-
short residence time, and its manageability of acteristic of hydrocyclones is also employed to
sizes finer than those treated in mechanical clas- improve the performance of other filtration or
sifiers. Some disadvantages are its high power separation equipments. The coarse and fine par-
consumption, its high wear, its inflexibility and ticles separated from hydrocyclones normally
its problem with blockage, especially in small follow different paths in a processing plant, be-
units. Figure 2 presents typical dimensions and ing sometimes fed into other units, to finish as a
standard nomenclature of hydrocyclones. different product.
are of the utmost importance. For a detailed de- This is an approximation, which gives the
scription of fluid dynamics and rheology see → best results for Rep →0; the upper limit of its va-
Fluid Mechanics. 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
3.1.1. Dynamics of Particles Submerged in the terminal settling velocity. Equations 1, 2 and
Fluids 3 combined give another form of Stokes’ law as
If a particle moves relative to the fluid in which follows:
it is suspended, the force opposing the motion 24
CD = (Rep <0.2) (4)
is known as the drag force. Knowledge of the Rep
magnitude of this force is essential if the parti-
cle motion is to be studied. Conventionally, the For Reynolds numbers > 1000 the flow is
drag force F D is expressed as: fully turbulent with inertial forces prevailing,
and C D becomes constant and equal to 0.44 (the
ρu2
F D = CD A (1) Newton region). The region in-between Rep =
2
where u is the particle–fluid relative velocity, ρ 0.2−1000 is known as the transition region, and
is the fluid density, A is the area of the particle C D is either described in a graph or by one or
projected in direction of the motion, and C D is a more empirical equations.
coefficient of proportionality known as the drag For a particle of mass m under the influence of
coefficient. Assuming the drag force is due to the a field of acceleration a, the equation of motion
inertia of the fluid, C D would be constant and di- is:
mensional analysis shows that C D is generally a du ρ
m = ma−ma −FD (5)
function of the particle Reynolds number, i.e.: dt ρs
uxρ
Rep = (2) where ρs is the density of the solids and t is the
µ
where x is the particle size and µ the medium time.
viscosity; the form of the function depends on In applications of hydrocyclones concerned
the regime of the flow. This relationship for rigid with separation of fine particles, which are all
spherical particles is shown in Figure 3. At low most difficult to separate, the Reynolds numbers
Reynolds number under laminar flow conditions are low, often less than 0.2, due to the low val-
when viscous forces prevail, C D can be deter- ues of x and u. Therefore, it is reasonable also
mined theoretically from Navier–Stokes equa- that the time necessary for the particle velocity
tions (see → Fluid Mechanics, Section 3.2) and to reach values very close to the terminal set-
the solution is known as Stokes’ law and repre- tling velocity is very short and being the field of
sented by: acceleration gravity, Equation (5) can be solved
to give:
FD = 3πµux (3)
x2 (ρs −ρ) g
ut = (6)
18µ
velocity at the same point. The inclusion of the century. The first research was done by Ein-
centrifugal terms in Equation (6) leads simply stein[4] in his classical study of the viscosity
to: of dilute suspensions of rigid spheres. His ap-
x2 −ρ) Rω 2
(ρs proach is purely hydrodynamic, and his model
vr = (7) consists of an isolated sphere situated in a simple
18µ
where R is the radius of the particle position and shear flow field in an infinite fluid. A number of
ω is the angular velocity. workers have been extending Einstein’s analysis
As the concentration of the suspension in- [3, 5]. Three approaches have been presented:
creases, particles get closer together and inter- the theoretical basis of viscosity computation,
fere with each other. If the particles are not the effect of particle texture and shape, and the
disturbed uniformly, the overall effect is a net effect of concentration. Several relations have
increase in settling velocity because the return been developed for suspensions in with nearest
flow caused by volume displacement predom- neighbor interactions cannot be neglected. Guth
inates in particles-sparse regions. This is the and Simha [6] considered the first order effect of
well-known effect of cluster formation, which spheres interacting with one another. They found
is significant only in nearly monosized suspen- that:
sions. With most practical widely dispersed sus- µ
= 1+2.5ø+14.1ø2 (9)
pension, clusters do not survive long enough µo
to affect the settling behavior and, as the re- where µ is the viscosity of the suspension, µo
turn flow is more uniformly distributed, the set- is the viscosity of the pure solvent and φ is the
tling rate steadily declines with increasing con- volume fraction of spheres in the suspension. A
centration. This phenomenon is referred to as later correction gave a value of 12.6 for the last
hindered settling and can be theoretically ap- constant [7].
proached in three different manners: as a Stokes’ Several authors have made experimental
law correction by introduction of a multiply- studies of the effects of concentration on the
ing factor; by adopting effective fluid properties viscosity of suspension of spheres. The data
for the suspension different from these of the they have proposed are considerably scattered.
pure fluid; and by determination of bed expan- Thomas [8] has tried to find the sources of scatter
sion with a modified version of the well know in the data and found that scattering is caused
Carman–Kozeny equation. All the approaches in large part by variations in particle size. For
can be shown to yield essentially identical re- small particle sizes (diameters < 1–10 µm) col-
sults. Some important correlations accounting loidal forces become important and the viscos-
for the hindered settling effect have been re- ity begins to increase as particle size decreases.
viewed in [3]. It was shown that their differences The viscosity is also shear rate-dependent in this
are minimal and, according to this, the Richard- case. For the particles larger than 1 to 10 µm,
son and Zakii equation is an obvious choice in the particle Reynolds number becomes signifi-
practice. Such relation can be expressed as: cant. The inertial effects results in an increase in
u
= (1−C)4.65 (8) relative viscosity with increasing particle size.
ut
Thomas has developed a unique curve, by elim-
where u is the settling velocity at concentration inating diameter and shear rate effects. For low
C and ut is the settling velocity of a single par- concentration data, such a graph can be repre-
ticle. sented by:
The relationship above applies only to free,
µ
particulate separation unaffected by coagulation = 1+2.5ø+11.4ø2 (10)
µo
or flocculation, and where all particles are of uni-
form density. where all the components are already defined.
The behavior of some specific systems has
also been reported in the literature. Lin and
3.1.2. Rheology of Suspensions Brodkey [9] studied rheological properties of
The rheological properties of suspension have
been studied since the beginning of the 20th
Hydrocyclones 7
slurry fuel, concluding that they are shear- if the fluid obeys the power-law, n = n’ and:
thinning under most conditions, but turn shear- n
3n+1
thickening at extreme conditions. Griskey et al. K = K (13)
4n
[10] investigated the rheological behavior of
cornstarch dispersions and found that they go Since Equation (11) rigorously portrays the
from dilatant (shear-thickening) behavior at low laminar flow behavior of the fluid (provided n’
shear rates to Newtonian behaviors at higher and K’ are evaluated at the correct shear stress),
shear rates. The effect of particle aggregation one may use it to define a Reynolds number ap-
on suspension has also been studied. Devries plicable to all purely viscous fluids under lam-
[11] presented a number of experimental results inar flow conditions. This dimensionless group
concerning the particle aggregation in polymer can be derived simply by the substitution of
lattices induced or accelerated by simple shear (D∆P/4L) from Equation (11) into the usual
flow. No stable flow is possible under these con- definition of the fanning friction factor, i.e.:
ditions, and the apparent viscosity is an increas- D∆P/4L
ing function of the time of shearing. f = (14)
ρv 2 /2
suspensions. They presented a viscous correla- as well as modifications of original Stokes law
tion in terms of parameters from shear-stress– for flocculated suspensions and particles of odd
shear-rate data and concluded that except for the shapes, have been studied by Thomas [16]. The
carboxymethyl cellulose, the behavior of these turbulent regime is sometimes preferred to trans-
materials could be correlated with the following port slurries within pipe in order to prevent set-
generalization of the Newtonian friction factor: tling, and to provide high throughputs. In this
case, the viscosity of the suspension has been
1
= A1n log Re∗(f )1−0.5n +Cn
∗
(18) normally used to correlate the turbulence fric-
f tion factors.
where A1n and Cn are parameters dependent The problem of distribution and mixing of
on the flow-behavior index n. While the choice solids in turbulent slurry flow has been stud-
of power-law equation in the derivation requires ied. The magnitude of flow velocity or Reynolds
that the Reynolds number used in Equation (17) number required for maintaining turbulent mix-
be the power-law special case of Equation (16), ing of solid–liquid slurries, may be computed
it has been shown [14] that any errors due to from the following equation:
this approximation are less important under tur- ρv 2 ρm vD
= K 0.775 (19)
bulent flow conditions than in the laminar region. (ρs −ρ) gx µ
The utility of Equation (17) and of Reynolds
number outside the laminar region has been where ρ is the liquid density, ρs the solid density,
shown diagrammatically in [14]. The authors ρm the slurry density, v the minimum standard
found excellent agreement between experimen- velocity, g the acceleration due to the gravity, D
tal and extrapolated data. Furthermore, devia- the mean particle diameter, µ the viscosity and
tions from smooth curves were no greater for K’ a constant. When densities of solids and liq-
those fluids, which did not obey the power law, uids are known, along with the volume fraction
than for those which did. The parameters A1n of solids in the slurry, the density of the mixture
and Cn in Equation (17) must be evaluated em- may be calculated by:
pirically, just as in the simpler case of Newtonian ρm = Cρs + (1−C) ρ (20)
behavior.
By far, most of the study of the flow behav- where C is the volume fraction of solids in the
ior of non-Newtonian suspensions has been de- slurry. Equation (18) is recommended for mean
voted to monosized and monoshaped systems, diameters between 50 and 500 µm.
mainly of spherical particle geometry. In real-
ity, however, solid–liquid mixtures usually con-
sist of dissolved solids with a variety of particle 3.2. Flow Patterns in Hydrocyclones
sizes, shapes, and concentrations. The available
theory for the behavior of non-Newtonian sus- The flow pattern in a hydrocyclone is com-
pensions is, therefore, of a limited application, plex. However, it is normally accepted that the
and when dealing with real systems, empirical hydrocyclone fluid flow consists of the three ve-
corrections are necessary to fully explain the sus- locity components: tangential, vertical, and ra-
pension properties. dial. Figure 4 shows a diagrammatically repre-
Studies of some real applications such as sentation of these three velocity distributions
slurry transport are also encountered in the liter- within a hydrocyclone. With respect to the tan-
ature. The slurry to be transported may have set- gential component vt , the linear velocity of the
tling or nonsettling characteristics. Sand in wa- feed stream at the inlet point is given by:
ter, salt in benzene, and clay in drilling mud are Q
typical examples of settling slurries, while pa- vt = (21)
Ai
per pulp is a good example of nonsettling slurry.
The law governing the settling rate of particles at where Q is the volumetric flow rate and Ai is the
low Reynolds numbers is the well know Stokes area of the inlet nozzle.
law. Settling phenomena of slurries in pipelines,
Hydrocyclones 9
where x is the particle diameter or size, and the where x c is known as the cut point, q equals
densities and velocities are as previously de- (Q/A) and Z is the acceleration factor. Including
fined. If vt ∼ (v/Rn ) and vr ∼ (v/Rm ), where further considerations, e.g., for capacity and the
v is the inlet velocity, then: separating area (surface of the cylinder formed
by the air core), Trawinski [17] also proposed:
x2 ρs −ρ K 1/2 1/2
= R1+2n−m (29) 18µ Dc Di
18 µ v xc = k 1 (H)−0.25 (34)
(ρs −ρ) g l
For given inlet velocity and liquid, the radius where k 1 is a constant, H is the liquid head, and
of the orbit is proportional to the particle size. If all the geometric characteristics are as defined
R is greater than the radial position for zero ver- in Figure 2.
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- 3.3.3. Crowding Theory
flow. Thus, a particle will have a 50 % chance
of removal if the radius of its equilibrium orbit Due originally to Fahlstrom [18], this theory
is equal to that of the zero vertical velocity. Us- proposes that the cut size is a function of the
ing this approach, Bradley [7] derived a relation capacity of the underflow orifice and the par-
making assumptions to the position of the locus ticle size distribution. The crowding effect can
of zero vertical velocity profile. 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-
3.3.2. Residence-Time Theory ory has been given mainly by Bloor And Ing-
ham [19], by means of the mathematical mod-
This theory advocates the importance of the time els developed to describe the fluid mechanics of
spent in the hydrocyclone. Assuming the spiral hydrocyclones.
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 3.3.4. Turbulent Two-phase Flow Theory
residence time will be:
This theory states the importance of the effect
2πRav caused by turbulent flow in hydrocyclones. Ac-
t= N (30)
v cording to this, it is proposed that separation
a particle of size x will travel a distance L in this takes place under the influence of the centrifu-
time t where: gal field and the turbulent transport in the two-
phase flow. The influence of turbulence can be
L x2 (ρs −ρ) v2
=v= (31) described by an eddy diffusion coefficient equal
t 18µ R
to, or smaller than the diffusion of the fluid parti-
Combining Equations 30 and 31 the following cles. The assessment of the influence of such dif-
relationship is obtained: fusion on separation is very difficult in practice.
Hydrocyclones 11
Bloor and Ingham [19] have also contributed to found in the specialized literature [21, 22] (see
the theoretical background of this theory. They → Particle Size Analysis and Characterization
derived an analytical form of the radial velocity of a Classification Process).
and found an analytical solution to the equation In solid–liquid separation it is important to
of motion for the tangential speed. They also know the particle size distribution in order to
developed a more elaborated and physically re- identify which part of it will be separated, and
alistic model, for the eddy viscosity based on the transform this quantity in a measure of efficiency
Prandtl mixing theory, which greatly simplified (→ Particle Size Analysis and Characterization
the equation of motion. of a Classification Process, Section 2.1). The
Schubert and Neesse [20] proposed a set of common way of presenting particle size data for
equations and derived two models for turbulent solid–liquid separation purposes is in form of
cross-flow wet classification: the suspension- a plot. The most common ones are equivalent
partition model and the suspension-tapping sphere diameters, equivalent circle diameter, and
model. In the former the flow is divided into statistical diameters. The equivalent sphere di-
overflow and underflow without changes in the ameter is the diameter of a sphere which has
total cross-section, whereas in the latter over- the same property as the particle itself. Such a
flow and underflow are tapped from the main property could be the settling velocity. A diam-
flow through small outlet openings. According eter derived from the settling velocity is known
to them, the suspension-tapping model is analo- as Stokes diameter and is a very useful quantity
gously applicable to the separation in hydrocy- for solid–liquid separation, especially to those
clones. techniques in which the particle motion relative
to the fluid is the governing mechanism.
The amount of particle matter which belongs
3.4. Characteristics of Performance of to specified size classes on the particle axis may
Hydrocyclones be also represented in several ways. Number of
particles and mass of particles are the most com-
The most important features of performance of mons ones, but surface area and volume are used
hydrocyclones are the particle cut size, the pres- as well. For solid–liquid separation techniques,
sure drop and the volume throughput. Previous the most convenient form of expressing amount
work has established the cut size as the best way of particle matter is by mass, because the bal-
to predict efficiency in hydrocyclones. In order ances necessary to define performance are nor-
to understand and utilize the cut size concept, it mally mass balances.
is necessary to deal with at least two topics, i.e., In general, particle size distributions can be
particle size analysis and efficiency of solid–liq- presented as frequencies, f(x), or cumulative fre-
uid separation (see → Particle Size Analysis and quencies, F(x), which are related to each other
Characterization of a Classification Process; → by following equation:
Solid–Liquid Separation, Introduction). dF (x)
f (x) = (35)
dx
3.4.1. Separation Efficiency, Cut Size the graphical representation of a particle size dis-
tribution is usually plotted in a cumulative form.
The solid phase of suspensions to be treated In a typical cumulative plot, points are entered
in solid–liquid separation equipment generally showing the amount of particulate material con-
consists of an immense number of particles of tributed by particles below or above a specified
diverse sizes and shapes. All this population of size. Hence, the curve presents a continuously
particles needs to be identified or characterized. rising or deceasing character. These oversize and
The frequency of occurrence of particles of ev- undersize distributions, as illustrated in Figure
ery size present, arranged and presented in a sta- 5, are simply related by:
tistical manner, is known as the particle size dis-
F (x)oversize = 1−F (x) (36)
tribution. The most common way of presenting
such distributions as well as the types of distri- where F(x) is the cumulative fraction undersize.
butions important to particles technology, can be
12 Hydrocyclones
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 aration efficiency is simply the overall mass re-
x
F (x) = 1− (42) covery as a fraction of the feed flow rate. Ac-
xo cording to Figure 6:
where F(x) is cumulative percentage oversize,
Mc
x o is the maximum size in the sample, s is a pa- Et = (43)
M
rameter concerned with the slope of the log–ln
plot in the fine region and r is a parameter con- where all the components are as defined in Fig-
cerned with the shape of the log–ln plot in the ure 6.
coarse region. If there is no accumulation of solids in the
The ways for measuring particle size are di- separator then:
verse and have been largely improved in re-
M = Mc +Mf (44)
cent years. Sieving and sedimentation are the
traditional ones, but microscopy, laser diffrac- and there is a choice of three possible combi-
tion, and stream scanning techniques are cur- nations of the material streams for the total ef-
rently spreading and finding applications. Sev- ficiency testing. It can be shown [3] that if all
eral methods for analytical determination of par- the operating conditions are equal, the most ac-
ticle size analysis have been reviewed [23]. Al- curate estimation of the local efficiency comes
though the most novel techniques have the ad- from the two leaving streams.
vantage of quick response, they have proved to As has been already mentioned, hydrocy-
be very selective. Therefore, most of them are clones act as flow dividers. Thus, they also split
of quite limited application. A rule of thumb the solids in at least the same ratio, as is the ratio
should be to use the analytical technique whose of underflow to throughput, Rf :
indirect measurement agrees, as much as pos-
U
sible, with the specific application. For exam- Rf = (45)
Q
ple, in solid–liquid separations one is concerned
14 Hydrocyclones
Mc
G(x) = (48)
M x
fine particles is, therefore, equal to the underflow gives a 50 % efficiency in such a curve is called
to throughput ratio Rf as shown in Figure 7. A the reduced cut size and is represented by x50 .
reduced grade efficiency, similar to the reduced The maximum obtainable efficiency related
efficiency represented by Equation 46 can then to particle size would be that minimum particle
be obtained as: size with 100 % probability of being reported to
G(x)−Rf
the underflow. Graphically, by extrapolating the
G (x) = (49) end part of the curve to the horizontal axis, such
1−Rf
size will be obtained. It has been proved that
As can be seen in Figure 7, the correspondent in practice, the maximum of the efficiency is
curve to this definition does cross the origin. around 98 %, and the minimum size correspon-
Hydrocyclones are among those equipments dent to this efficiency is represented by x 98 and
whose separation performance is highly depen- known as the approximate limit of separation.
dent on particle size. Thus, according to this, an Several correlations have been proposed in
overall efficiency of separation is not the best literature to evaluate the cut size. They are ei-
way to evaluate their performance. For this rea- ther empirical or theoretical, and are expressed
son, it is advisable to express their efficiency for in the traditional British System of units, as well
each size of particle in the range measured. As as in SI units, even in both. Perhaps the most
experience has shown, the only single number, important criterion for classifying the proposed
which in some way gives the separation power correlation for cut size, is by dividing them into
of a hydrocyclone, is the cut size x 50 (for a def- two groups considering whether the feed con-
inition of cut size see → Particle Size Analy- centration has been taken into account. The main
sis and Characterization of a Classification Pro- correlations, which do not consider the feed con-
cess, Section 3.1.4). All particles above the cut centration as a variable, will be reviewed as fol-
size will be generally going to the underflow, lows.
whereas those below the cut size will be nor- Bradley [2] proposed:
mally reported to the overflow. 1/2 n
18πµ(1−Rf ) 2.3Do Di2
As can be concluded, the cut size is derived x50 = (51)
16LQ(ρs −ρ) Dc α
from a grade efficiency curve. As the calculation
and plotting of a full grade efficiency curve is a where Q is the flow rate, Do , Dc and Di are the
time-consuming task, alternative ways of deter- dimensions defined in Figure 2, and (and n are
mining the cut size have been proposed [28, 29]. empirical constants depending on the cyclone
The sharpness of separation can be related design and the fluid properties.
to the shape of the curve in a number of man- Rietema [30] derived a dimensionless, char-
ners. Knowing the grade efficiency curve, com- acteristic cyclone number C y50 represented by:
parisons of classification sharpness of different L∆P
Cy50 = x250 (ρs −ρ) (52)
equipments can be made by plotting G(x) against µρQ
the dimensionless size x/x 50 , and comparing the where ∆P is the pressure drop. Cy50 is deter-
resultant normalized curves. To define the curve mined experimentally and is not directly related
steepness, a ratio of two sizes corresponding to to the more important hydrocyclone variables.
two different percentages on the grade efficiency However, Rietema has demonstrated the use-
curve on either side, and 50 % equidistant, can fulness of the parameter and quotes a value of
be used. This parameter is called the sharpness 3.50 for optimal separation, corresponding to
index and is represented by: the conditions: L/Dc = 5, Di /Dc = 0.28,
Do /Dc = 0.34 and l/Dc = 0.4, where all the
x25
H25/75 = (50) dimensions are as defined in Figure 2.
x75
Dahlstrom [31] proposed:
It was previously stated that the grade effi-
ciency curve does not go through the origin. The C(Do Di )0.68 1.73 1/2
x50 = 0.53
(53)
cut size as has been defined is derived from this Q ρs −ρ
curve. If, in order to practically assess the per- where C is a constant sensitive only to major
formance of hydrocyclones, a reduced grade ef- changes in the cyclone dimensions. This equa-
ficiency curve is used, the particle size which tion holds well for cyclones in which the length
16 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.
Hydrocyclones 19
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
for the case of Newtonian suspensions, i.e., low (Fig. 10.), in order to verify the effect that varia-
solids contents, the above mentioned equations tions in the underflow orifice diameter may have
would reduce to the common definitions so they in the cut size. Once a number of tests have been
conclude their model is of wide applicability. performed, the possible scale-up to industrial
All the above described relationships have size may follow.
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 11. Diagram of radial (spider) arrangement of hydrocyclones for industrial scale
clone bodies. For cyclone cavities with a cylin- closed off individually by insertion of a blinding
drical diameter of 10 mm the diameter of the device in order to control the capacity. For feed
deck is 300 mm. The whole cyclone deck can be flows above 5 m3 /h, the decks can be stacked on
molded by injection in one piece. The feed zone top of each other and held together by connecting
is the annular space closest to the periphery. The rods through the inner annular space. There is a
overflow is through peripheral screw-in-vortex limit to the number of decks that can be stacked
finders. The underflow leaves via the central cir- on top of each other. If the stack of decks be-
cular space where the small apex openings enter. comes too high, the pressure drop over the feed
The design feed flow per cyclone cavity is zone will result in a lower inlet pressure to the cy-
about 250 L/h corresponding to a pressure drop clones furthest downstream and, consequently,
of 600 kPa, approximately. A deck with 20 cy- these cyclones will give lower throughputs and
clones in parallel consequently may have a ca- lower efficiency. Some other arrangements have
pacity of 5 m3 /h. This is, however, not the max- been tested with different degrees of success.
imum capacity. The cyclone cavities can be Some of them have been reviewed in [1].
Hydrocyclones 25
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)
This review deals with chemical and physical 1. Introduction and Historical
properties of hydrogen peroxide and gives de- Aspects
tailed descriptions about the various production
processes. Today, almost all hydrogen peroxide The industrial production of hydrogen peroxide
is produced by organic autoxidation processes, passed through three phases, starting with wet
primarily the anthraquinone process. Other as- chemical processes, followed by electrochem-
pects in this review are handling, safety and us- ical processes, and then by organic autoxida-
age of hydrogen peroxide. Hydrogen peroxide tion processes. Almost all hydrogen peroxide
occurs naturally. Because of its chemical reac- is now produced by organic autoxidation (AO)
tivity, hydrogen peroxide also has toxic prop- processes, primarily the anthraquinone process.
erties. The formation and degradation of atmo-
spheric hydrogen peroxide is photochemically Wet Chemical Processes. In 1818, L. J.
mediated. Degradation of aquatic hydrogen per- Thenard [9] obtained hydrogen peroxide for the
oxide is catalyzed by biological material or by first time by treating barium peroxide [1302-29-
transition metals. Toxicity data are collected for 6] with nitric acid. This process was improved by
various organisms. using hydrochloric acid to release hydrogen per-
oxide. The water-soluble barium chloride which
is formed simultaneously, was precipitated with based on this new technology went on stream in
sulfuric acid. 1908 at the Österreichische Chemische Werke
in Weissenstein. The Weissenstein process was
BaO2 +2 HCl→BaCl2 +H2 O2
followed in 1910 by the Münchner process de-
veloped by A. Pietzsch and G. Adolph at the
BaCl2 +H2 SO4 →BaSO4 +2 HCl Elektrochemische Werke, Munich, and then in
1924 by the Riedel–Loewenstein process de-
BaO2 +H2 SO4 →BaSO4 +H2 O2 veloped by L. Loewenstein and first used by
With this reaction, Thenard established the Riedel-de-Haen. In these processes, an ammo-
foundation for the commercial manufacture of nium sulfate solution was electrolyzed instead
aqueous hydrogen peroxide solutions based on of free sulfuric acid, and the resulting am-
wet chemical processes, which began around monium peroxodisulfate (Riedel–Loewenstein)
1880. or potassium peroxodisulfate derived therefrom
Industrial plants using the barium peroxide (Pietzsch–Adolph) was hydrolyzed. After intro-
process were still operating until the middle of duction of these processes, production (calcu-
the 20th century. Around 1900, approximately lated as 100 % hydrogen peroxide) increased
10 000 t/a of barium peroxide was processed by steadily and, in 1950, reached approximately
this technology, yielding ca. 2000 t/a of hydro- (30–35) × 103 t/a [13].
gen peroxide. Sales opportunities for the coprod-
Organic Autoxidation Processes. The deci-
uct barium sulfate (BaSO4 , blanc fixe) had a de-
sive breakthrough in industrial production of hy-
cisive effect on the profitability of the process.
drogen peroxide, which enabled the construc-
The 3 % aqueous hydrogen peroxide solu-
tion of modern large-scale plants, came with the
tions manufactured by the barium peroxide pro-
development of organic autoxidation processes,
cess had only a limited market because of their
especially the anthraquinone process.
high production cost, low hydrogen peroxide
In 1901, Manchot discovered that hydro-
content, and unsatisfactory stability due to im-
quinones (or hydrazobenzenes) react quantita-
purities.
tively with oxygen to form quinones (or azoben-
zenes) and hydrogen peroxide [14]. In 1932,
Electrochemical Processes. The introduc-
Walton and Filson in the United States went
tion of electrochemical processes eliminated the
back to this work and proposed a cyclic process
disadvantages of the barium peroxide process.
for the manufacture of hydrogen peroxide based
In 1853, Meidinger discovered the forma-
on azobenzene–hydrazobenzene [15].
tion of hydrogen peroxide during electrolysis
Pfleiderer of I. G. Farbenindustrie AG, Lud-
of aqueous sulfuric acid [10]. In 1878, Berth-
wigshafen (BASF), subsequently developed a
elot showed that peroxodisulfuric acid [13445-
process for the alkaline autoxidation of hydra-
49-3] is formed during this electrolysis and then
zobenzene in which sodium peroxide is ob-
hydrolyzed by water to give sulfuric acid and
tained instead of free hydrogen peroxide [16].
hydrogen peroxide via peroxomonosulfuric acid
This process was used on an industrial scale by
[12188-01-1] [11]:
Kymmene AB in Kuusankoski (Finland) under
2 H2 SO4 →H2 S2 O8 +H2 BASF licence (see Section 4.4.4.
The azobenzene process had two major tech-
H2 S2 O8 +H2 O→H2 SO5 +H2 SO4 nical drawbacks: hydrogenation of azobenzene
was performed with sodium amalgam, and ox-
H2 SO5 +H2 O→H2 SO4 +H2 O2
idation of hydrazobenzene proceeded satisfac-
torily only in alkaline solution. Pfleiderer and
Riedel overcame these problems by using alky-
2 H2 O→H2 O2 +H2
lated anthraquinones instead of azobenzene. Be-
In 1905, following the work of Elbs and Schoen- tween 1935 and 1945, BASF developed the
herr [12], G. Teichner developed an electrolytic Riedel–Pfleiderer process (generally referred to
process at the Konsortium für Elektrochemis- as the anthraquinone or AO process) in a pi-
che Industrie. The first hydrogen peroxide plant 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].
ice–H2 O2 · 2 H2 O at 45.2 wt % hydrogen per- Table 4. Hydrogen peroxide as a reducing agent [20]
Quinones. Criteria similar to those used for Although “tetra” is more readily hydro-
selecting solvents also apply to the quinone or genated than a 2-alkylanthraquinone, the result-
quinone mixture which is used as the working ing “tetra” hydroquinone is much more difficult
compound: to oxidize than the readily oxidizable anthrahy-
droquinone. The kinetics and basic physico-
1) Good solubility of the quinone form
chemical parameters of tetrahydroquinone ox-
2) Good solubility of the hydroquinone form
idation have been studied [55–57]. Small
3) Good resistance to oxidation
amounts of amines accelerate hydroquinone ox-
4) Good availability
idation [58].
The choice of quinone depends mainly on the The formation of “tetra” in the synthesis loop
properties, especially the solubility, of the prod- depends on process conditions and has led to two
ucts formed in the AO process. Examples of methods for carrying out this process: the “an-
working compounds are 2-ethylanthraquinone thra” system and the “all-tetra” system.
[84-51-5] [45], 2-tert-butylanthraquinone [84-
47-9] [46], 2-neopentylanthraquinone [123257- Anthra System. The slower oxidation rate
96-6] [47], 2-isohexenylanthraquinone [71308- of the “tetra” hydroquinone causes difficulties
16-2] [48], diethylanthraquinones [49], mixtures in the oxidizer. Actions have therefore been pro-
of 2-amylanthraquinones [54], and mixtures of posed which suppress the formation of “tetra” or
different alkylanthraquinones [50–53]. dehydrogenate it to anthraquinone.
The formation of degradation products and When the “tetra” content of the working solu-
their ability to be regenerated to active quinones tion is kept low (i.e., if 2-alkylanthrahydroqui-
also play a role in the decision. In addition to none is formed almost exclusively during hy-
hydroquinone formation, a number of secondary drogenation), the process is referred to as the
reactions occur during the hydrogenation step; “anthra” system.
of these, hydrogenation of the anthraquinone “Tetra” formation is suppressed by using se-
ring system is particularly important. One ring lective catalysts [59], special solvents [60], spe-
of the anthraquinone molecule, preferably the cial working compounds [61], and mild hydro-
one that is not substituted by an alkyl group, is genation conditions [62] (e.g., the use of olefins
hydrogenated. In the case of R=ethyl, this is ac- has been recommend for the steady dehydroge-
complished with selectivities higher than 90 %. nation of “tetra” [63]).
“Tetra” can be dehydrogenated in the pres-
ence of activated aluminum oxide:
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.
tion), (2) during hydrogenation (by cooling the trated to the commercial concentration of 50–70
reactor), or (3) after hydrogenation (by cooling wt %; it is then collected in a storage tank (l).
the hydrogenated working solution). Water vapor produced during distillation is
condensed and fed into the water storage tank
Oxidation. Before the hydrogenated work- (m).
ing solution that contains hydroquinone can be
fed to the oxidation step, it must pass through Auxiliary Processes. A number of addi-
a safety filtration stage (c). This is particularly tional processes are required to maintain the AO
important because the hydrogenation catalysts operation. For example, to maintain hydrogen-
used in the AO process (palladium and Raney ation activity, part of the catalyst is removed,
nickel) also catalyze the decomposition of hy- regenerated in the catalyst regeneration area (q),
drogen peroxide. Even a small amount of these and returned to the hydrogenator. To compensate
catalysts in the oxidation and extraction steps for quinone and solvent losses, working solution
would lead to considerable loss of hydrogen per- is periodically made up with anthraquinone and
oxide and serious disturbances. solvent (o, p).
During the oxidation step (d), the hydro-
genated working solution is gassed with air. Dis-
4.1.2.1. Hydrogenation
solved hydroquinones are oxidized to quinones,
and hydrogen peroxide is formed. The hydrogenation step is the most important
After the working solution has been separated step of modern AO processes. Quinone decom-
(e1 ), air from the oxidation step passes over ac- position products that cannot be regenerated into
tivated carbon adsorbers (f), and adsorbed sol- active quinone are formed during this step (cf.
vent is recovered from them. Several adsorbers Section 4.1.1). New hydrogenation catalysts and
are usually loaded and regenerated alternately, hydrogenation reactors have been developed that
for example, with steam. deviate totally from the BASF principle. Here,
design of the hydrogenator depends largely on
Extraction and Recovery of the Working the type of catalyst used.
Solution. The oxidized working solution is then Four typical reactors for the three usual cat-
treated with water to extract hydrogen peroxide alyst systems (suspended, supported, and fixed-
(g). bed catalysts) are discussed in this section.
The working solution leaving the extraction
unit must be adjusted to a specific water con- BASF Hydrogenation Step. The hydrogen-
tent before being returned to the hydrogenation ation step in the BASF plant (Fig. 6) uses a Raney
step. Free water taken up by the working solu- nickel catalyst at a slight excess pressure of ap-
tion during extraction is separated (e2 ) and the proximately 0.2 MPa and at 30–36 ◦ C. Because
water content is adjusted to the desired level in Raney nickel is sensitive to oxygen, the work-
the drier (h). ing solution from the extraction or drying and
To purify the working solution and regenerate purification steps cannot be fed directly into the
the quinone decomposition products into active hydrogenator. This working solution still con-
quinones, some or all of the working solution is tains residual hydrogen peroxide and must be
passed through a regeneration step (n). passed over a decomposition catalyst (e.g., sup-
ported Ni–Ag), together with a fraction of the
Hydrogen Peroxide Concentration. Crude hydrogenated working solution (which also con-
aqueous hydrogen peroxide from the extraction tains hydroquinone), to remove hydrogen perox-
stage (H2 O2 concentration 15–40 wt %) is fed ide completely:
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-
Hydrogen Peroxide 11
The solutions are passed through the precon- cent years, more selective nickel catalysts, based
tact column (Fig. 6, a) and collected in the hy- on Ni-B or Ni-Cr-B alloys were developed [80].
drogenator feed tank (b). The working solution The pyrophoric properties of Raney nickel
is then pumped into the stirred vessel reactor also require more stringent safety procedures
(c) and is gassed with hydrogen in the pres- when handling the material. Raney nickel is still
ence of Raney nickel. Periodic addition of small used today in some AO plants, but palladium
amounts of hydrogenation catalyst from the cat- catalysts are preferred because of their higher
alyst feed tank (d) allows a constant rate of hy- selectivity and simpler handling.
drogen conversion in the hydrogenator. Hydro-
genated working solution is collected in the ox- Degussa Hydrogenation Step. Degussa
idizer feed tank (h) through the internal filters proposed the use of palladium black as the
in the stirred vessel, thus exploiting the excess hydrogenation catalyst. This exploits the advan-
pressure in the reactor. The solution is then led tages offered by a suspended catalyst and avoids
into the oxidation step via the safety filter (e). the disadvantages of Raney nickel. Equipment
A sidestream of hydrogenated working solution that allows good conversion of hydrogen with
is withdrawn and recycled to the precontact col- very finely distributed palladium black is shown
umn (a). in Figure 7 [81].
When the concentration of Raney nickel in The most important feature of the loop reac-
the hydrogenation reactor reaches a certain limit, tor (c) is the connection in series of pipes with
the content of the reactor is drained into the cat- different diameters. The working solution flows
alyst separator (f). Raney nickel settles to the downward in the large pipes at a rate of 0.7–1.5
bottom, and catalyst-free supernatant is pumped m/s and flows upward in the narrower pipes at
back to the hydrogenator. 1.5–3 m/s. Efficiency can be improved by in-
A significant disadvantage of Raney nickel stallation of internal plate-stacks with a channel
as catalyst is its limited selectivity, i.e., the ratio width of 10-50 mm [82] or external premixing
of hydroquinone formation to “tetra” formation. of gas and liquid [83].
BASF largely eliminated this by pretreating the Degussa proposed a carbon filter [84]. Ce-
catalyst with ammonium formate [76]. ramic [85] and cross-flow [86] filters can also be
Alternatives were subsequently suggested for used. A decline in filter performance can be over-
pretreating Raney nickel (e.g., nitriles [77], come by periodic back flush with hydrogenated
amines [78], and aldehyde solutions [79]). In re- working solution through the filter into the hy-
drogenator.
12 Hydrogen Peroxide
In recent years, especially AO plants built in oxidation step uses a nitrogen–oxygen mixture.
China were equipped with fixed-bed hydrogen- The mixture is cycled in a closed circuit, and
ation. its composition is kept constant by adding pure
oxygen to compensate for the amount consumed
in the formation of hydrogen peroxide.
4.1.2.2. Oxidation Hydrogenated working solution enters a sep-
arator (a) and then flows through four oxidation
In industrial plants, the catalyst-free hydro- columns (b) arranged in series as a cascade. Ox-
genated working solution is usually oxidized idized working solution flows into the extractor
with slightly pressurized air (up to 0.5 MPa). feed tank (d) via a siphon (c). The nitrogen–
The oxidizer off-gas is then directed to a series oxygen mixture is compressed and fed into each
of activated carbon adsorbers to purify it and re- of the four reactors.
cover the solvents escaping with it. Off-gas from the reactors is collected in a pipe
For economic reasons, the following objec- and separated from entrained droplets of work-
tives are sought in designing industrial oxidation ing solution. After compensation for consumed
reactors: oxygen, the gas mixture is returned to the com-
pressor (e).
1) Efficient utilization of atmospheric oxygen to
reduce the volume of the off-gas and the size
Degussa Oxidation Step (Fig. 11) [102].
of activated carbon adsorbers
The hydrogenated working solution flows cocur-
2) Low compressor pressure to decrease energy
rently upward with the oxidizing air through the
costs
first reactor. The working solution and air are
3) Small reactor volume or reduced holdup of
separated in the upper part of the reactor. Off-
working solution to lower investment costs
gas from the reactor is fed to the activated car-
for equipment and working solution
bon adsorption unit, and working solution flows
into the second and third reactors, which are in-
BASF Oxidation Step (Fig. 10). A working stalled in series. The air needed for oxidation is
solution containing benzene as the quinone sol- pressurized with a compressor (c) and fed to the
vent has been used at BASF. For safety reasons third reactor. The working solution and air flow
and to minimize the volume of off-gas, the BASF
Hydrogen Peroxide 15
countercurrently, but both streams are divided 4.1.2.3. Extraction and Drying
into cocurrent flow segments.
In the BASF plant, a sieve-tray extraction col-
Laporte Oxidation Step (Fig. 12). In La- umn extracts hydrogen peroxide from the work-
porte’s Warrington plant, oxidation was carried ing solution. A number of other extractors, such
out in a packed cocurrent flow column [103, as packed columns, pulsed packed columns, and
104]. The entire volume of the oxidation reac- Podbielniak extractors, have been proposed.
tor (a) was used for air gassing. The air and the Besides extraction column design, the ex-
working solution leaving the top of the column traction efficiency is strongly influenced by the
were fed together into a separator (b). The air distribution coefficient, which is depending on
then reached the alternately operated two-stage working solution composition (e.g., type of hy-
activated carbon adsorbers (c), and the work- droquinone solvent).
ing solution passed to the extraction stage. To It has been found, that by injection of a
enhance gas-liquid mass transfer, premixing of small amount of air into the extraction column
working solution and air by a venturi tube is pro- (gas-agitated sieve plate column) extraction effi-
posed [105]. The use of a tubular reactor with ciency is increased by 30–40 % [109]. Processes
static mixing elements is described by Kemira which integrate oxidation and extraction in a sin-
in [106]. 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).
Figure 12. Laporte oxidation step
a) Oxidation reactor; b) Separator; c) Activated carbon ad-
sorber; d) Air compressor
4.1.2.4. Working Solution Purification and
Regeneration
Allied Chemical Oxidation Step. To During the constant circulation of working solu-
shorten the residence time of the working so- tion, degradation products are formed not only
lution in the oxidation step, Allied Chemical from the working compound, but also from the
suggested a counterflow reactor in which res- solvents. These degradation products must be
idence times are less than 2.5 min at a partial removed from the working solution to prevent
oxygen pressure of 70–90 kPa [107]. Space– (1) deterioration of the crude hydrogen peroxide
time yield can be improved by using perforated (color, smell, dissolved organic compounds), (2)
trays with small hole size (cross-sectional area
of holes below 3 mm2 ) [108].
16 Hydrogen Peroxide
an increase in density and viscosity of the work- 5) Treatment of the working solution with aque-
ing solution, and (3) a decrease in interfacial ten- ous sodium hydroxide or potassium hydrox-
sion of the working solution, which promotes ide solution (8–17 mol/L) in the presence of
the formation of an emulsion during extraction. oxidants [120]
In addition, degradation products decrease the 6) Treatment with solvent and liquid carbon
activity and lifetime of the hydrogenation cata- dioxide [121]
lysts. Regeneration and purification can be per- 7) Treatment with a noncyclic hydrocarbon
formed with hydrogenated working solution or [122]
with working solution after extraction.
Numerous methods have been suggested for
purifying the working solution and regenerat- 4.1.2.5. Purification of Crude Hydrogen
ing active quinone from the quinone degradation Peroxide
products. Examples include the following:
The aqueous hydrogen peroxide leaving the ex-
1) Treatment of the hydrogenated working solu-
traction is impure. Separators are usually in-
tion with alkaline substances [71] or aqueous
stalled after extraction (Section 4.1.2.3) [123].
alkali hydroxide [72]
Measures for reducing the amount of dissolved
2) Treatment of the working solution with
organic compounds include treatment of the
sodium aluminum silicates at 50–200 ◦ C
crude product with polyethylene [124], ion ex-
[114]
changers [125, 126], or hydrocarbons [127,
3) Treatment of the working solution after hy-
128]. Water-soluble organic compounds can
drogenation or extraction with active alu-
also be oxidized by heating and then extracted
minum oxide at 20–150 ◦ C [115–118]
with suitable solvents (e.g., the quinone solvent)
4) Treatment with alkali hydroxide, calcium hy-
[129]. The effect of purification on carbon con-
droxide, ammonia, or amines in the presence
tent of crude hydrogen peroxide is illustrated in
of oxygen or hydrogen peroxide [119]
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 dilute solutions has been used industrially to pro-
a distillation unit where it is purified further and duce commercial 50–70 wt % hydrogen perox-
concentrated to the usual commercial concen- ide solution; the technology is described in [3, 5,
tration of 50–70 wt % hydrogen peroxide. This 146–152]. The use of structered packings [150]
product can be used directly in a number of ap- in distillation columns enhances perfomance.
plications. Reverse osmosis membranes can also Concentrations higher than 90 % are preferably
be used to purify crude, distilled or concentrated produced by crystallization [153, 154].
(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- 4.2. 2-Propanol Process (Shell Process)
meate, while the impurities are accumulated in
the retentate. In 1945 Harris discovered that primary and sec-
For use in the electronics industry, hydrogen ondary alcohols react with oxygen to form hy-
peroxide with a very low content (ppt level) of drogen peroxide and an aldehyde or ketone, re-
metal impurities is needed. Many patents deal spectively [155]:
with further purification of crude or prepuri-
RCH2 OH+O2 →RCO+H2 O2
fied hydrogen peroxide with a combination of
cation- and anion- exchange resins [136–140].
Some methods make use of of reverse osmosis R2 CHOH+O2 →R2 C=O+H2 O2
in combination with other methods [134, 141, Oxidation of the alcohol can be carried out in
142], adsorption of metal impurities on activated the liquid or the gas phase. Because aldehydes
carbon [143], stannic oxide [144], or zirconium formed in the reaction with primary alcohols are
phosphate [145]. oxidized easily, only the oxidation of secondary
alcohols, especially 2-propanol [67-63-0], has
industrial importance.
4.1.2.6. Concentration of Hydrogen A process for the simultaneous production
Peroxide of hydrogen peroxide and acetone [67-64-1] by
atmospheric oxidation of 2-propanol was first
Water and hydrogen peroxide do not form an suggested by Rust [156, 157].
azeotrope; their boiling point difference at at- Reaction of 2-propanol with oxygen in the
mospheric pressure is 50.2 ◦ C (Table 1). Dis- liquid phase does not require a special catalyst
tillation of water from aqueous hydrogen per- because it is catalyzed by the hydrogen peroxide
oxide can therefore be used to concentrate the product. However, a small amount (0.5–1 wt %)
solution or to completely separate hydrogen per- of hydrogen peroxide is added to 2-propanol to
oxide from water. A number of technical safety shorten the induction period. The reaction pro-
requirements must be observed. Distillation of ceeds as follows:
18 Hydrogen Peroxide
distillation of hydrogen peroxide are carried out oxide, giving a distillate of 60 wt % which then
in two steps (c, d). Hydrolysis is 80–90 % com- drops to 20 wt %.
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 4.4. Other Processes
expelled from the concentrated sulfuric acid so-
lution by countercurrent steam. Entrained acid is A number of reactions and processes for the pro-
separated in an acid separator (e), and hydrogen duction of hydrogen peroxide are described in
peroxide–water vapors are separated in a frac- the literature. Some of them are no longer eco-
tionating column (f). An aqueous solution with nomical (e.g., the barium peroxide, azobenzene.
a maximum content of 45 wt % hydrogen per- Only a few examples can be described here;
oxide is obtained. To purify the circulating so- detailed information may be found in the litera-
lution, part of the stream is distilled constantly ture [1, 3, 5].
(g).
4.4.1. Production from Peroxy Compounds
4.3.2. Münchner Process Peroxy compounds can be decomposed to form
hydrogen peroxide. Examples of such reac-
In the Münchner process (Fig. 15), acid ammoni- tions, which include the historically important
um sulfate is electrolyzed in electrolyzers with- Thenard reaction of barium peroxide with nitric
out diaphragms. or sulfuric acid, are listed in Table 7.
The circulating solution contains 300 g of
(NH4 )2 SO4 , 40 g of K2 SO4 , 100 g of H2 SO4 , Table 7. Production of hydrogen peroxide from peroxy compounds
and 57.6 g of S2 O2−8 per liter before entering the Peroxide Reaction Refer-
electrolyzer. Its peroxodisulfate concentration is ence
increased to 115.2 g/L by the electrolytic pro- BaO2 reaction of barium peroxide with [90]
H2 SO4 in the presence of a small
cess. After electrolysis, the circulating solution amount of HCl allows quantitative
is cooled and concentrated in a vacuum cooler conversion; ca 6 % aqueous H2 O2
(b) by evaporation of water and then fed into solutions are obtained
BaO2 as above, but in the presence of HCl [91]
a crystallizer (c). Half of the peroxodisulfate is and H3 PO4
precipitated as potassium peroxodisulfate by ad- Na2 O2 sodium peroxide solutions are passed [92]
dition of potassium hydrogen sulfate and sepa- through acidic ion exchangers to
exchange Na+ for H+
rated in a centrifuge (d). The supernatant mother (CH3 )3 CO–OH tert-butyl hydroperoxide is [93]
liquor diluted with washing water is returned hydrolized to tert-butanol and H2 O2
in the presence of H2 SO4 in a
to the electrolyzer. Solid potassium peroxodi- fractionating column
sulfate is hydrolyzed (e) and hydrogen peroxide (CH3 )3 CO–OH H2 O2 and isobutene are formed from[94]
is steam distilled. tert-butyl hydroperoxide in an
aqueous medium in the presence of
The remaining potassium hydrogensulfate– an acidic ion-exchange resin
sulfuric acid residue is separated in a centrifuge peracetic acid is hydrolyzed to [95]
(f). Sulfuric acid is recycled to the hydrolysis CH3 COOH and H2 O2 in the
presence of H2 SO4 ; to shift the
and distillation stage (e) via a purification stage equilibrium to H2 O2 , CH3 COOH is
(g). Potassium hydrogensulfate is returned to reacted with an alcohol and the ester
is distilled
the crystallizer (c) for precipitation of peroxo-
peracetic acid is hydrolyzed with [96]
disulfate. acidic ion exchangers, and the
To purify the circulating solution, a side equilibrium mixture is separated by
distillation
stream is mixed with ammonia in the purifica-
tion step (h) to precipitate iron. The hydrogen
peroxide–water vapors formed during hydroly- 4.4.2. Production from Hydrogen and
sis and distillation are led through an acid sep- Oxygen
arator (i) and separated in the rectification step
(j). Because hydrolysis is carried out batchwise, Hydrogen peroxide can be formed by direct re-
the initial vapor mixture is rich in hydrogen per- action of oxygen and hydrogen. However, the
Hydrogen Peroxide 21
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
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 — Liquid (Aque- For practical use, safety tests of the intended
ous) Phase [224–226]. The explosive risk of working area are very important. When systems
pure, highly concentrated hydrogen peroxide so- such as hydrogen peroxide–formic acid–water
lutions has been examined by various methods, are used, other products may be formed (in this
but the sometimes contradictory findings are dif- case performic acid), that are more dangerous
ficult to interpret consistently. than the original mixture [228].
Results are influenced primarily by test con-
ditions and not by hydrogen peroxide concentra-
tion. For example, 90.7 wt % hydrogen peroxide 7. Uses
in a tube with an internal diameter of 2.67 cm
(1.05 inches) could not be exploded by the deto- Development of the AO process permitted large-
nation shock of a tetryl booster, whereas explo- scale production of hydrogen peroxide, which
sion occurred in a tube with an internal diam- is now used widely in almost all industrial ar-
eter of 4.09 cm (1.61 inches). Similarly, 86 % eas. Its main use is in bleaching (→ Bleaching,
hydrogen peroxide in a tube of 4.09 cm inter- Chap. 2.2.3). Uses in the chemical industry and
nal diameter was exploded by raising the tem- in environmental protection are increasing be-
perature above 50 ◦ C [224, 225]. The decisive cause its great advantage is that the degradation
factors for explosion of pure aqueous hydrogen product is water.
peroxide seem to be the degree of occlusion, the Hydrogen peroxide is used in the textile in-
insulation, and the energy of detonation. dustry for bleaching cotton, linen, bast fibers,
wool, silk, polyester fiber, and polyurethane
Mixtures or Solutions of Aqueous Hydro- fiber. In the pulp and paper industry, it is used to
gen Peroxide with Organic Compounds [219, bleach sulfate and sulfite cellulose, wood pulp,
227–234]. The hazards associated with mix- and wastepaper, and to brighten wood veneers
tures of hydrogen peroxide and organic chem- and wooden structures.
icals are important for its industrial production The chemical industry employs hydrogen
and, especially, for its uses. As an example, peroxide for the production of peroxy com-
Figure 18 shows results obtained with the hy- pounds such as sodium perborate, sodium per-
drogen peroxide–acetone–water system [227]. carbonate, metallic peroxides, or percarboxyl-
Many other organic compounds give similar re- ic acids. Hydrogen peroxide is very important
sults. The size of the explosive region depends in organic chemistry [229] for epoxidation and
on the organic compound and the test conditions. 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
Figure 18. Explosive range (hatched area) of hydrogen
is used in wet chemical processes during the
peroxide–acetone–water mixtures manufacture of silicon semiconductor devices
Dashed line, indicates the stoichiometry to clean silicon wafers and to remove photore-
8 H2 O2 +CH3 COCH3 →3 CO2 +11 H2 O sist layers. It is also used for in situ leaching in
underground uranium mining.
26 Hydrogen Peroxide
Hydrogen peroxide is increasingly used in decomposition to water and oxygen in the un-
environmental protection to detoxify effluents derlying tissue. This may lead to the formation
containing formaldehyde, phenols, or cyanide of small gas bubbles and reversible blanching
(e.g., wastewater from mines and tempering of the exposed tissue. Formation of larger vol-
works, galvanizer concentrate, photochemical umes of gaseous oxygen can lead to detachment
effluents), and to deodorize sulfur-containing ef- of cell layers and the rupture of tissues and or-
fluents. Smoke and exhaust gases containing sul- gans. Locally formed oxygen is removed by the
fur dioxide and even dioxins can be completely blood. However, the increase in oxygen content
detoxified with hydrogen peroxide. in the blood results in a hyperbaric response.
Hydrogen peroxide is a highly efficient dis- There is no evidence that hydrogen peroxide it-
infectant (especially for packaging materials); it self is absorbed in amounts sufficient to produce
is also used as a bleaching agent in hair prepa- systemic effects [236]. In in vitro studies with
rations and as a propellant [226] in space tech- rat blood plasma additions of up to 1 mM of hy-
nology. drogen peroxide were degraded to nondectable
concentrations within less than 10 min [238].
This experimental result supports the assump-
8. Toxicology and Occupational tion that hydrogen peroxide is not systemically
Health available.
The major adverse effects of hydrogen per-
Hydrogen peroxide plays a major role in physio- oxide are related to its irritating properties.
logical oxidation processes in living organisms. The acute oral toxicity of hydrogen perox-
Hydrogen peroxide is produced during normal ide in animals depends on the strength of its so-
aerobic cell metabolism, which involves a num- lution. In the rat LD50 values ranged from >
ber of enzymes, in particular superoxide dismu- 5000 mg/kg for a 10 % solution to 1200 mg/kg
tase. Humans may exhale up to 1 mg/m3 of hy- for 35 % solution. In humans, toxic effects ob-
drogen peroxide resulting from endogenous for- served after accidental ingestion of concentrated
mation [236]. solutions are generally related to the irritant or
Reactive oxygen species, including hydrogen corrosive action and the rapid generation of large
peroxide, may also lead to oxidative damage to volumes of oxygen in the gastrointestinal tract.
cells. Therefore a number of defense systems Single cases of deaths due to gas embolism fol-
can be found in biological systems which lead lowing accidental swallowing of hydrogen per-
to detoxification of reactive oxygen species. The oxide solutions have been reported [239], how-
most important detoxification reactions for hy- ever, in other case reports complete recovery was
drogen peroxide are its conversion to oxygen observed within 2-3 weeks [236].
and water by enzymes such as catalase, perox- Complete recovery has occurred within two
idase, and selenium-dependent glutathione per- to three weeks even in near fatal cases [236].
oxidase. The activity of these enzymes varies Effects of exposure to hydrogen peroxide by
between tissues and between different animal inhalation are limited to irritation of the respi-
species or strains [236]. Additionally, a variety ratory tract and the eyes. Exposure to hydrogen
of nonenzymatic antioxidants, such as Vitamin peroxide vapor (338–427 mg/m3 ) induced signs
E and A, ubiquinols, ascorbic acid, glutathione, of irritation in rodents. In humans exposed for
complete an efficient intra- and extracellular net- 4 h the irritation threshold for the respiratory
work of defenses [237]. tract was 10 mg/m3 and for the skin 20 mg/m3 .
There is limited information concerning the At these concentrations, eye and throat irritation
absorption of exogenously applied hydrogen were reported [236].
peroxide. The permeability of biological mem- Slight nasal irritation was reported in fac-
branes to hydrogen peroxide is comparable to tory workers for short-term exposure to ca. 3.5
that of water, but less than that of oxygen [236]. mg/m3 [236]. Gradual bleaching of hair was re-
Although significant amounts of topically ap- ported for exposure levels of 0.7–1.4 mg/m3 , but
plied hydrogen peroxide can penetrate the epi- it was uncertain whether bleaching was due to
dermis or mucous membranes, the substance un- vapor exposure or to transfer of hydrogen per-
dergoes rapid spontaneous or enzyme-catalyzed oxide from the hands (gloves) to the hair [236].
Hydrogen Peroxide 27
The dermal toxicity of hydrogen peroxide so- ferent production plants for up to 40 years with
lutions of 35–75 % in rabbits was low (LD50 , recent exposures below 1.4 mg/m3 (8-h TWA),
dermal, rabbit: > 2000 mg/kg). LD50 values of but occasional peak exposures up to 5 mg/m3
a 90 % solution ranged from ca. 690 mg/kg in confirmed the absence of effects on lung func-
rabbits to > 2000 mg/kg in rats, pigs, and cats tion or any other clinical effects apart from sin-
[236]. gle observations of skin or throat irritation, skin
The irritating properties of hydrogen perox- or hair whitening after acute incidents of high
ide are dependent on the concentration of the short-term exposures [239]. In a population ex-
solution. Hydrogen peroxide solutions of 10 % posed to levels of 2 - 3 mg/m3 (8-h TWA) and
or less were nonirritating, and solutions of 35 % frequent peak exposure levels up to 11 mg/m3
slightly irritating to rabbit skin. Higher concen- symptoms including eye and respiratory irrita-
trations (≥ 50 %) are corrosive to rabbit skin. tion, headaches and temporary loss of olfaction
Hydrogen peroxide (50 %) was corrosive to rab- as well as skin and hair blanching and three cases
bit skin after 1 h of exposure, whereas 70 % hy- of recurring bronchitis–sinusitis were reported.
drogen peroxide showed corrosive effects after All persons recovered completely after the ex-
only 3 min of exposure [236]. Dermal exposure posure was reduced to below the occupational
of humans to dilute solutions has been reported exposure limit of 1.4 mg/m3 and the repeated
to result in transient symptoms without perma- peak exposures were avoided. The authors as-
nent effects, such as paresthesia, blanching, and sume that the recurrent peak exposure levels of
blistering [236]. up to 11 mg/m3 may have contributed to the ob-
With regard to eye irritation solutions of 5 % served symptoms [241].
or less are not irritant to rabbit eyes; a concen- More detailed studies of the effects of hydro-
tration of 6 % caused slight reversible eye irrita- gen peroxide after prolonged exposure were per-
tion in rabbits. Concentrations of 10 % or higher formed by administration in drinking water. Af-
caused irreversible corneal damage. In humans, ter prolonged exposure (146 d), decreased body-
1–3 % solutions have been used for eye treat- weight gain was observed in rats at concentra-
ment without causing injury. However, solutions tions of 0.25 % and above [242] (other studies
containing more than 200 mg/L have been re- observed first effects at higher levels [236]).
ported to cause hyperemia, and concentrations In a 90-d drinking water study in catalase-
of greater 100 mg/L transient pain and stinging deficient mice receiving 100–3000 ppm of hy-
[236]. Hydrogen peroxide (6 %) was not a skin drogen peroxide, irritation of the mucous mem-
sensitizer in guinea pigs [236]. Human experi- branes of the gastrointestinal tract was observed.
ence from dermatological hospital departments The no observed adverse effect level (NOAEL)
confirms that hydrogen peroxide is no skin sen- was 100 ppm, corresponding to 26 and 37 mg
sitizer [239]. kg−1 d−1 for male and female mice, respec-
After repeated inhalation exposure (6 h/d, 2 tively. After cessation of treatment these effects
to 3 d per week, 3 weeks), mortality in mice was were fully reversible. No systemic effects were
observed at 80 mg/m3 hydrogen peroxide. Af- observed [243]. In studies on the same mouse
ter prolonged exposure, in rats and dogs (5 or 6 strain with a 6–7 month exposure period, signs
h/d, 5 d per week for 16 or 26 weeks, respec- of gastrointestinal irritation were observed at ap-
tively), respiratory irritation and transient skin proximately the same dose level as in the 90-d
thickening were observed at 10 mg/m3 hydro- study. Therefore it can be assumed that the effect
gen peroxide [236]. In a 28-day inhalation study is not cumulative and the NOAEL does not de-
in rats the NOAEL for respiratory tract irritation crease with increasing exposure time. A NOAEL
was 2.9 mg/m3 . Clinical signs of irritation were of 56.2 mg kg−1 d−1 was reported in a 12-week
observed from 14.6 mg/m3 [240]. gavage study in rats [236].
In humans, pulmonary function monitoring Results of mutagenicity studies indicate that
showed no evidence of adverse effects from oc- hydrogen peroxide induces gene mutations in
cupational exposure for several years at levels vitro without metabolic activation in bacteria,
ranging from not detectable to 0.79 mg/m3 (8-h yeast, and mammalian cells. Hydrogen perox-
TWA) [236]. Further surveys of workers occu- ide also induced chromosome aberrations and
pationally exposed to hydrogen peroxide in dif- micronuclei in mammalian cell cultures without
28 Hydrogen Peroxide
a metabolic activation system. However, addi- On the basis of these studies ACGIH has
tion of exogenous metabolic activators or cata- grouped hydrogen peroxide into category A3 for
lase lowered or abolished the genotoxic response carcinogenicity: “The agent is carcinogenic in
[236]. In a mechanistic study, it has be shown experimental animals at relatively high doses,
that hydrogen peroxide induced controlled cell by routes of administration, at sites of histologi-
death (apoptosis) rather than genotoxic lesions cal types, or by mechanisms that are not consid-
in cell cultures of human lymphocytes and did ered relevant to worker exposure. Available epi-
not induce 8-hydoxyguanosin formation [244]. demiological studies do not confirm an increased
In in vivo tests, no chromosomal aberrations risk of cancer in exposed humans. Available ev-
or micronuclei were observed in the bone mar- idence suggests that the agent is not likely to
row of rats and mice after repeated oral or sin- cause cancer in humans except under uncom-
gle intraperitoneal administration [236]. An in mon or unlikely routes of exposure” [246]. Sim-
vivo/ex vivo study for unscheduled DNA syn- ilar conclusions were drawn in the EU risk as-
thesis (UDS) after intravenous administration of sessment on hydrogen peroxide [239].
the maximum tolerated dose to rats did not show Occupational exposure limits have been es-
an increase in UDS compared to the solvent con- tablished for hydrogen peroxide in different
trol group [245]. In a study of local genotoxic- countries, for example by ACGIH [246], the
ity to mouse skin, 70 % hydrogen peroxide did Dutch Social Economic Council [247], the UK
not cause any mutations, hydroxylation of DNA HSE [251], Belgium [248], the French Ministry
bases, or gave an indication of cellular prolifer- of Labour [249], and the German AGS (Auss-
ation, thus suggesting that the substance has no chuss für Gefahrstoffe) [250]: 8-h TWA: 1.4
local genotoxic potential in vivo [239]. In con- mg/m3 or 1.5 mg/m3 . A 15-min STEL of 2.8
clusion, the available studies indicate that hy- mg/m3 was established by HSE (max. 1 h in a
drogen peroxide has no genotoxic or mutagenic 24-h period) [251], and a 15-min STEL of 1.4
potential under in vivo conditions. mg/m3 by AGS [250].
The carcinogenic effects of hydrogen perox-
ide were studied in rats and mice. Hydrogen per-
oxide induced an increase of duodenal tumors 9. Environmental Effects
in mice chronically exposed to 0.1–0.4 % of hy-
Hydrogen peroxide occurs naturally. Its forma-
drogen peroxide in drinking water. The tumor
tion, use, and degradation are strongly related to
incidence correlated with a specific inflamma-
the oxygen-containing atmosphere of our planet
tory response in this tissue and was more pro-
and adaptation of nature to oxygen. The concen-
nounced in those mice strains which had a low
tration of hydrogen peroxide in environmental
catalase activity [236]. The absence of preneo-
compartments results from a dynamic equilib-
plastic changes and the reversibility of the ef-
rium between its natural production and degra-
fects in a 90-d drinking water study in mice of
dation in photochemical, chemical, and biolog-
the same strain [243] (with 0.01–0.3 % hydro-
ical processes.
gen peroxide in drinking water), suggest that the
Because of its chemical reactivity, hydrogen
tumors observed in the two-year study seem to
peroxide also has toxic properties that are some-
be a secondary response to a prolonged inflam-
times used as a natural defense mechanism. The
matory process in this specific strain of mice.
toxicity of hydrogen peroxide to organisms and
Rats exposed to near lethal concentrations in
cells is dependent on the balance of intoxication
their drinking water developed forestomach pa-
and detoxification. However, detoxification re-
pilloma but no tumors in the glandular stomach
actions can in turn lead to toxic effects due to
and duodenum [236]. Chronic hydrogen perox-
the release of oxygen gas [252].
ide exposure at concentrations which induce sig-
nificant inflammation of the exposed tissue has
been shown to give a weak tumor-promoting ef- 9.1. Atmospheric Fate
fect. Repeated topical application of 15 % hy-
drogen peroxide to Sencar mice did not induce The atmospheric fate of hydrogen peroxide is
skin tumors; this demonstrates the absence of governed by generation and degradation reac-
initiating properties [236]. tions. The formation of atmospheric hydrogen
Hydrogen Peroxide 29
Fe2+ +O− • +
2 +2 H →Fe
3+
+H2 O2
•
OH+HO2 • →H2 O+O2
Hydrogen peroxide concentrations in surface
The rate for the direct photolysis of hydrogen water can also originate from precipitation of
peroxide for moderate light intensities is directly rain [259].
30 Hydrogen Peroxide
significant effect on any of the above parameters 11. H. Berthelot, C. R. Hebd. Seances Acad. Sci.
[252]. 86 (1878) 71.
When assessing the risk of hydrogen perox- 12. K. Elbs, O. Schönherr, Z. Elektrochem. 1
ide to environmental organisms, the background (1895) 417, 468.
concentrations in the environment and the ca- 13. W. M. Weigert, H. Delle, G. Käbisch, Chem.
pacity to adapt to increased hydrogen perox- Ztg. 99 (1975) 101.
14. W. Manchot, Liebigs Ann. Chem. 314 (1901)
ide concentrations have to be considered. The
177;
medium used in laboratory tests is quite different 316 (1901) 318, 331.
from the natural conditions. Depending on the 15. J. H. Walton, G. W. Filson, J. Am. Chem. Soc.
environmental conditions, many materials (e.g., 54 (1932) 3228.
dissolved transition metals, solid metals, solid 16. O. v. Schickh, Chem. Ing. Tech. 32 (1960) 462.
metal oxides, hydroxides, activated carbon, en- 17. J. Lacson, S. Schlag and G. Toki: “CEH
zymes) may contribute to rapidly reduced hy- Product Review Hydrogen Peroxide” in
drogen peroxide levels. It is thus assumed that Chemical Economics Handbook, SRI
under laboratory conditions the test organisms International, Menlo Park 2002, 741. 5000A.
are more exposed and thus more sensitive than 18. W. R. Busing, H. A. Levy, J. Chem. Phys. 42
under natural conditions to the prolonged expo- (1965) 3054.
sure of H2 O2 . In field conditions, the more sen- 19. P. A. Giguere, J. Chem. Ed. 60 (1983) 399.
20. C. L. Yaws, H. S. N. Setty, Chem. Eng. (1974)
sitive species may also be protected from a direct
67.
effect of hydrogen peroxide by the presence of 21. G. Scatchard, G. M. Karanagh, L. B. Ticknor,
the total biomass and the suspended solids. J. Am. Chem. Soc. 74 (1952) 3715.
However, it is obvious that accidental re- 22. W. M. Latimar: The Oxidation States of the
leases, in particular of highly concentrated hy- Elements and their Potentials in Aqueous
drogen peroxide, can immediately be lethal to Solutions, 2nd ed., Prentice Hall, New York
aquatic organisms. With regard to long-term ef- 1952.
fects of lower levels, it can be expected that adap- 23. Hydrogen Peroxide for Industrial Use –
tation of ecosystems may well be possible. Determination of Content, International
Organization for Standardization, ISO/DIS
7157.
24. Degussa, DE 1 039 500, 1956 (G. Käbisch, E.
10. References Richter, A. Gross).
25. I.G. Farbenindustrie AG, GB 508 081, 1939;
General Reference US 2 215 883, 1939.
1. Gmelin, System no. 3, Sauerstoff. 26. Degussa, DE-AS 1 945 750, 1969 (G.
2. Kirk-Othmer, 5th ed., 14, Wiley, 2005. Käbisch, R. Trübe).
3. W. Machu: Das Wasserstoffperoxid und die 27. EDOGAWA, DE 1 112 051, 1957 (K.
Perverbindungen, 2nd ed., Springer, Wien Hiratsuka, S. Matsumura, M. Matsuda, K.
1951. Takagi, T. Abe).
4. Wasserstoffperoxid, Berufsgenossenschaft der 28. Laporte Chemicals Ltd., GB 747190, 1953 (C.
chemischen Industrie, Merkblatt M 009, BGI W. Lefeuvre).
792, Jedermann-Verlag, Heidelberg 2000. 29. Akzo Nobel, WO 2001098204, 2000 (M.
5. W. C. Schumb, C. N. Satterfield, R. L. Nystroem, C. Jaernvik).
Wentworth: Hydrogen Peroxide, Reinhold 30. Degussa, DE 1 261 838, 1963 (G. Käbisch).
Publ. Co., New York 1955. 31. RO 108956, 1991 (M. Mihalcescu, C. Andrei,
6. Ullmann, 5th ed., A13, 443–465. C. Dragoi).
7. W. M. Weigert: Wasserstoffperoxid und seine 32. E. I. du Pont de Nemours & Co., DE 888 840,
Derivate, Hüthig, Heidelberg 1978. 1950 (C. R. Harris, J. W. Sprauer).
8. Winnacker-Küchler, 5th ed., 3, 682–760, 33. BECCO, US 2 537 655, 1950 (L. H. Dawsey,
Wiley-VCH, Weinheim, Germany 2005. C. K. Muehlhauser, R. R. Umhoefer).
Specific Reference 34. Laporte Chemicals Ltd., DE 933 088, 1953 (C.
9. L. J. Thenard, Ann. Chim. Phys. 8 (1818) 306. W. Lefeuvre).
10. H. Meidinger, Ann. Chem. Pharm. 88 (1853) 35. E. I. du Pont de Nemours & Co., US 4046868,
57. 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. 61. Degussa, DE 1 051 257, 1957 (G. Käbisch, E.
Suokas). Richter).
38. EKA, EP 284 580, 1988 (U. M. Andersson, E. 62. E. I. du Pont de Nemours & Co., US
A. Bengtsson). 2 673 140, 1951 (J. W. Sprauer).
39. Degussa, DE 3 027 253, 1980 (G. Goor, W. 63. EDOGAWA, FR 1 319 025, 1962.
Kunkel). 64. EDOGAWA, DE-OS 2 141 234, 1971 (H. Shin
40. FMC Corp., US 4 394 369, 1983 (R. Wayne). et al.).
41. EKA, EP 286 610, 1988 (U. M. Andersson, E. 65. J. Petr, L. Kurc, Z. Belohlav, L. Cerveny,
A. Bengtsson). Chem. Eng. Proc. 43 (2004) 887–894.
42. Degussa, DE 2 018 686, 1970 (G. 66. E. I. du Pont de Nemours & Co., US
Giesselmann, G. Schreyer, W. Weigert). 4 668 499, 1986 (J. D. Rushmere).
43. J. Fang, Huagong Keji 6 (1998) 18–23. 67. Solvay Interox, EP 546 616, 1992 (V.
44. I. G. Farbenindustrie AG, US 2 215 883, 1937 Deremince, C. Vogels).
(H.-J. Riedl, G. Pfleiderer). 68. T. Berglin, N. H. Schöön, Ind. Eng. Chem.
45. I.G. Farbenindustrie AG, US 2 158 525, 1936 Process Des. Rev. 22 (1983) 150–153.
(H.-J. Riedl, G. Pfleiderer). 69. EKA, DE 2 419 534, 1973 (B. G. Franzen).
46. E. I. du Pont de Nemours & Co., US 70. Degussa, EP 1274651, 2000 (T. Haas, J.
2 689 169, 1949 (W. S. Hinegardner). Glenneberg, R. Wagner, R. Vanheertum).
47. BASF, EP 348464, 1990 (M. Eggersdorfer, J. 71. Degussa, DE 1 273 499, 1964 (G. Schreyer, O.
Henkelmann, W. Grosch). Weiberg, G. Giesselmann).
48. Kawasaki Kasei, JP 58180452, 1983 (S. 72. Degussa, WO 2005 014471, 2004 (G. Goor, E.
Numata). Staab, J. Glenneberg).
49. Montecatini, IT 601 435, 1960 (C. Ferri, A. 73. F. Sandelin, P. Oinas, T. Salmi, H. Haario,
Niglierina, B. Ponzini). Stud. Surf. Sci. Catal. 88 (1994) 635.
BASF, DE 43 39 649, 1993 (H. Hiller). 74. A. Drelinkiewicz, J. Mol. Catal. A 101 (1995)
50. Degussa, DE 1 063 581, 1956 (L. R. Darbee, 61;
D. F. Kreuz). Bull. Pol. Acad. Sci. Chem. 43 (1995) 51.;
51. Kvaerner, EP 958 238, 1997 (E. Bengtsson, A. A. Drelinkiewicz, R. Laitinen, R. Kangas, J.
Casen, A. Unum). Pursiainen, Appl. Catal. A: General 284
52. Degussa, US 6 153 169, 1998 (J. Glenneberg, (2005), 59–67.
G. Goor, E. Staab, H. Angert). 75. Solvay Interox, WO 98 15 350, 1997 (V.
53. Akzo Nobel, EP 1101733, 1999 (M. Mathieu, P. Pennetreau, N. Vanlautem).
Nystroem, C. Jaernvik, H. Thor, S. Saari). 76. BASF, DE 801 840, 1951 (G. Pfleiderer, H.-J.
54. EDOGAWA, DE-AS 1 085 861, 1957 (K. Riedl, W. Deuschel).
Hiratsuka et al.). 77. FMC Corp., US 2 720 532, 1952 (R. R.
55. E. Santacesaria, R. Ferro, S. Ricci, S. Carra, Umhoefer).
Ind. Eng. Chem. Res. 26 (1987) 155; 78. FMC Corp., US 2 730 533, 1952 (R. R.
E. Santacesaria, M. Di Serio, A. Russo, U. Umhoefer).
Leone, R. Velotti, Chem. Eng. Sci. 54 (1999) 79. Kali-Chemie, DE-AS 1 137 719, 1960 (V.
2799–2806. Hauschild, M. Stahl).
56. P. Oinas, G. Wild, N. Midoux, H. Haario, 80. J. Fang et al., J. Catal. 229 (2005) 97–204.
Chem. Eng. Proc. 34 (1995) 503. 81. Degussa, DE 1 542 089, 1965 (G. Käbisch, H.
57. A. Golja, M. Krajnc, S. Flajs, M. Grcar: Herzog).
Zbornik Referatov s Posvetovanja Slovenski Degussa, US 4 428 923, 1982 (W. Kunkel, J.
Kemijski Dnevi, Maribor, Slovenia, Sept. Kemnade, D. Schneider).
20-21, 2001, Issue Part 2, 723–729. 82. Degussa, WO 2000 015550, 1999 (H.
58. L. Kurc, V. Vopravil, L. Cerveny, Res. Chem. Eickhoff, R. Schütte).
Intermed. 27 (2001) 249–258. 83. Degussa, EP 1 334 062, 2001 (B. Maurer, P.
Degussa, DE 10114 982, 2001 (T. Haas, J. Korl).
Glenneberg, R. Wagner, R. Vanheertum). 84. Degussa, GB 1 096 260, 1965.
59. Laporte Chemicals Ltd., GB 741 444, 1953 85. Degussa, EP 687 649, 1995 (H. Birkenbeil, U.
(W. R. Holmes, C. W. Lefeuvre, R. Lait, W. S. Brand, G. Goor, W. Kunkel).
Wood). 86. Degussa, EP-A 571 762, 1993 (H. Eickhoff, R.
60. Solvay & Cie., DE 1 041 928, 1955 (M. Goedecke, H. Huber, K. Mangartz, R.
Gonze, E. Leblon). Schirrmacher).
Hydrogen Peroxide 33
87. Laporte Chemicals Ltd., GB 718 306, 1953 (C. 111. EKA, EP 078 780, 1982 (B. G. Franzen, W.
A. Morgan). Herrmann).
88. Du Pont, US 5 772 977, 1995 (C.L. Jenkins, F. 112. Kemira, EP-A 726 227, 1996 (I. Turunen).
B. Kirby, T. A. Koch). 113. Research Disclosure 40735, March 1998
89. Mitsubishi, DE 19713376, 1997 (K. (Derwent No. 98-205225).
Ogasawara, T. Kato, N. Okuda, T. Konishi, K. 114. Pittsburgh Plate Glass Co., US 3 055 838,
Kato). 1957 (J. W. Moore).
90. Laporte Chemicals Ltd., GB 746 385, 1953 (C. 115. E. I. du Pont de Nemours & Co., US 2 739 875,
W. Lefeuvre, C. A. Morgan); 1951 (J. W. Sprauer, T. V. Williams).
GB 922 022, 1959 (R. Lait, D.R. 116. Foret, US 4 566 998, 1981 (V. M. Ochoa, R.
Lloyd-Owen). Pueyo).
91. Laporte Chemicals Ltd., GB 718 307, 1952 (J. 117. Mitsubishi Gas, DE-OS 19 715 034, 1997 (K.
A. Williams, C. W. Lefeuvre). Ogasawara, T. Kato, K. Minato, K. Kato, M.
92. Du Pont, US 4 374 820, 1981 (T. E. Guenter). Sumitani).
93. Solvay, WO 99 040024, 1998 (D. Vandenberg, 118. FMC, US 6 180 804, 1998 (E. F. Dougherty).
J. P. Ganhy, N. Vanlautem). 119. Laporte Chemicals Ltd., DE-AS 1 138 743,
94. FMC Corp., US 3 009 782, 1958 (D. H. 1959 (B. K. Howe, S. Andrus, G. N. Davis).
Porter). 120. Laporte Chemicals Ltd., GB 991 413, 1961 (B.
95. EKA, US 5 063 043, 1990 (E. A. Bengtsson). K. Howe).
EKA, US 4 552 748, 1983 (T. Berglin, W. 121. FMC Corp., US 4 824 609, 1988 (D. S. Sethi).
Herrmann). 122. FMC Corp., US 4 668 436, 1985 (D. S. Sethi).
96. A. Drelinkiewicz, A. Pukkinen, R. Kangas, R. 123. Du Pont Australia, WO 95 04 702, 1994 (P. J.
Laitinen, Catal. Lett. 94 (2004) 157–170. Collins, P. B. Humphries, G. J. J. Prendergast).
97. A. Drelinkiewicz, R. Laitinen, R. Kangas, J. 124. Columbia-Southern Chemical Corp., GB
Pursiainen, Appl. Catal. A: General 284 795 433, 1955 (J. R. Cox, M. Korach).
(2005) 59–67. 125. Pittsburgh Plate Glass Co., GB 924 625, 1958
98. FMC Corp., US 4 428 922, 1982 (Q. G. (M. Q. Freeland, P. R. Johnston).
Hopkins). 126. Mitsubishi Gas, US 4 999 179, 1989 (Y.
99. Kemira, US 5 071 634, 1990 (T. Maunula, Sugihara, S. Shimokawa).
E.-L. Mustonen, I. Turunen, P. Virta); 127. E. I. du Pont de Nemours & Co., DE-AS
Kemira, WO 93 02 960, 1992 (E. Maekiniemi, 1 036 225, 1956 (J. W. Sprauer).
T. Maunula, A. Pukkinen, I. Turunen). 128. Laporte Chemicals Ltd., GB 841 323, 1956
100. Q. Wang, L. Wang, Y. Wang, F. Hei, Z. Li, Z. (H. B. Hopkins).
Mi, React. Kinet. Catal. Lett. 81 (2004) 129. FMC Corp., US 3 043 666, 1960 (G. W.
297–304. Siwinski).
101. EKA, EP 778 085, 1996 (M. Nystroem, M. 130. Laporte Chemicals Ltd., DE-AS 1 467 091,
Silverstroem). 1964 (S. D. Williams).
131. E. I. du Pont de Nemours & Co., US 3 016 290,
102. Degussa, DE 2 003 268, 1970 (M. Liebert, H.
1956 (A. A. Elston, W. S. Hinegardner).
Delle, G. Käbisch).
132. Oxysynthèse, FR 1 539 843, 1967 (P. Thirion).
103. Chem. Process Eng. (London) 40 (1959) no. 1,
133. Foret, ES 547 571, 1985 (R. P. Gracia, V. O.
5.
Bendicho, J. M. Lopez Martinez).
104. Br. Chem. Eng. 4 (1959) 88.
134. E. I. du Pont de Nemours & Co., US
105. Österr. Chemische Werke, WO 86 06710,
4 879 043, 1988 (J. H. Boughton, R. A. Butz,
1986 (J. Kemnade, B. Maurer).
H. C. T. Cheng, J. R. Dennis, B. T. Hannon, J.
106. Kemira, DE-OS 4 029 784, 1990 (I. Turunen,
H. Weigel).
E. L. Mustonen).
135. Kemira, EP-A 703 189, 1995 (I. Turunen).
107. Allied Chemical Corp., US 2 902 347, 1954 (J.
136. Mitsubishi Gas, EP 774 442, 1996 (M. Hattori,
N. Cosby, R. L. Harris, E. H. Rissmiller).
Y. Minamikawa, S. Murakami).
108. Degussa, WO 2000 017098, 1999 (H.
137. Ausimont, EP 930 269, 1999 (U. P. Bianchi,
Eickhoff, R. Schütte).
U. Leone, M. Lucci).
109. S. Lue, Z. Mi, Y. Wang, L. Wang, Chem. Eng. 138. Mitsubishi Gas, US 5 534 238, 1995 (S.
Technol. 27 (2004) 903–908. Kajiwara, H. Serizawa).
110. Pittsburgh Plate Glass Co., DE 977 130, 1965 139. Solvay Interox, DE-OS 4 222 109, 1992 (H.
(J. R. Cox, T. E. Brown). Honig).
34 Hydrogen Peroxide
140. Solvay Interox, WO 92 06 918, 1991 (J. R. 166. ARCO, US 5 039 508, 1990 (R. N. Cochran,
Crampton, F. R. F. Hardy, M. H. Millar, G. W. L. M. Candela).
Morris). 167. D. F. Dong, J. R. Jackson, Int. Forum
141. Sumitomo, US 5 906 738, 1994 (Y. Matsuda, Electrolysis Chem. Ind., 9th Electrosynthesis,
A. Morisaki, Y. Sawaguri). Lancaster, NY 1995, p. 215–228.
142. Solvay, EP 1520839, 2003 (R. Owen, J. Bosse, 168. P. Askenasy, R. Rose, Z. Anorg. Allg. Chem.
M. Sell). 189 (1930) 1, 10.
143. E. I. du Pont de Nemours & Co., US 169. J. Lawrence, R. B. Peacock, C. W. Lefeuvre,
4 985 228, 1990 (K. Kirksey). Bios Final Report 1432, (1946), no22.
144. Solvay Interox, WO 90 11 967, 1990 (N. D. 170. D. J. Campbell, E. F. Lefson, US 2 497 812,
Feasey, G. W. Morris). 1950.
145. FMC Corp., US 5 266 298, 1992 (T. 171. Atlantic Richfield Co., US 3 891 748, 1973.
Frianeza-Kullberg, D. Gibilisco, J. L. 172. Atochem, Oxysynthese, EP 237 402, 1987
Manganaro, J. R. Reed). (J.-P. Schirmann, M. Pralus).
146. Laporte Chemicals Ltd., US 3 073 755, 1963 173. Wacker-Chemie, US 312 421, 1964 (W.
(R. H. Banfield, A. M. Hildon). Lohringer, J. Sixt).
147. Columbia-Southern Chemical Corp., US 174. Celanese Corp., US 3 341 297, 1967 (A. F.
2 960 448, 1960 (W. S. Straub, C. J. MacLean, A. L. Stautzenberger).
Sindlinger). 175. Elektrochem. Werke München, US 2 015 040,
148. Degussa, US 3 060 105, 1962 (J. Müller); 1935 (A. Pietzsch).
Degussa, US 3 714 342, 1970 (G. Käbisch). 176. Mitsubishi Denki K. K., DE-OS 4 106 980,
149. FMC Corp., US 3 152 052, 1964 (T. M. 1991 (S. Endoh, K. Namba, S. Yagi, K.
Jenney, D. H. Porter, R. H. Rice). Maeda).
150. Sulzer, EP 419 406, 1990 (A. Meili). 177. Chemical Week 1987, December 9, 20.
Sulzer, EP 835 680, 1996 (A. Meili). 178. E. I. du Pont de Nemours & Co., US
151. Kemira, EP-A 727 388, 1996 (P. Johnsson, T. 4 681 751, 1985 (L. W. Gosser).
Mattila, K. Saari). 179. Chem. Eng. News, 17 Oct 2005, 83 (42), p 13.
152. Solvay Interox, EP 529 722, 1992 (L. 180. Atomic Energy of Canada Ltd., EP 366 419,
Signorini, K. Glinos). 1989 (K. T. Chuang).
Solvay, EP 1 090663, 1999 (S. Bloomfield, G. 181. J. R. Kosak, Chem. Ind. (Dekker) 62 (1995)
E. Williams). 115–124.
153. FMC, US 2 724640, 1947 (G.G. Crewson, J. 182. Mitsubishi Gas Chemical Company Inc., EP
R. Ryan). 492 064 (Y. Hiramatsu, Y. Ishiuchi, H.
154. Degussa, WO 2002 044083, 2001, (M. Creutz, Nagashima).
S. Gross, S. Nordhoff, R. Wagner). 183. Mitsubishi Gas Chemical Company Inc., EP
155. E. I. du Pont de Nemours & Co., US 498 166 (Y. Hiramatsu, Y. Ishiuchi, H.
2 479 111, 1949 (C. R. Harris). Nagashima).
156. Shell Development Co., US 2 871 104, 1951 184. Hydrocarbon Technologies Inc., US 6 168 775
(F. F. Rust). (B. Zhou, L-K. Lee).
157. N. V. De Bataafsche Petroleum Maatschappij, 185. E. I. du Pont de Nemours & Co., US 5 194 242
GB 708 339, 1954. (M. A. Paoli).
158. N. V. De Bataafsche Petroleum Maatschappij,
186. Advanced Peroxide Technology Inc., WO
BE 1 002 295, 1953 (J. R. Skinner, S. E.
98/31629 (H. Huckins).
Steinle).
187. Atofina, WO 01/05498 (M. Devic).
159. Chem. Eng. News 35 (1957) no. 48, 26.
188. Hydrocarbon Technologies Inc., US
160. Chem. Week 125 (1979) no. 18, 13.
2004/0018143 A1 (B. Zhou, M. Rueter).
161. Ullmann, 4th ed., 17, 694.
189. Eka Chemicals AB, WO 99/32398 (M.
162. E. J. Pneva, Chimitscheskoja Promyschlenost
Nyström, J. Wanngard, W. Herrmann).
9 (1976) 19.
190. Eka Chemicals AB, WO 99/32399 (M.
163. Khimprom, SU 1 685 865, 1991 (V. V.
Nyström, J. Wanngard, W. Herrmann)
Chikurov, I. A. Militsin, L. K. Zharikov).
191. Halcon, US 4 462 978, 1982 (W. F. Brill).
164. ARCO, US 4 897 252, 1989 (R. N. Cochran,
192. BOC, US 4 711 772, 1986 (S. E. Jacobson).
L. M. Candela).
193. Enichem, EP-A 788 998, 1997 (R. Bortolo, R.
165. ARCO, US 4 975 266, 1989 (R. S. Albal, R.
D’Aloisio, D. Bianchi).
N. Cochran, L. M. Candela).
Hydrogen Peroxide 35
194. Enichem, EP-A 808 796, 1997 (D. Bianchi, R. 215. Laporte Chemicals Ltd., GB 893 069, 1959 (P.
Bortolo, R. D’Aloisio, M. Ricci, S. Soattini). V. Rutledge, D. S. Bell).
D. Bianchi, R. Bortolo, R. D’Aloisio, M. 216. Columbia-Southern Chemical Corp., US
Ricci, Angew. Chem. 111 (1999) 734; 2 904 517, 1959 (P. H. Baker).
J. Mol. Catal. A: Chem 150 (1999) 87. 217. S. Croft, B. C. Gilbert, J. R. L. Smith, J. K.
195. C. Querci, R. D’Aloisio, R. Bortolo, M. Ricci, Stell, W. R. Sanderson, J. Chem. Soc. Perkin
D. Bianchi, J. Mol. Catal. A: Chem. 176 Trans. 2 1992, 153.
(2001) 95–100. 218. E. Rabald, D. Behrens (ed.),
196. Solvay Interox, US 5 545 391, 1992 (P. A. DECHEMA-Werkstoff-Tabelle
Clapp, D. F. Evans, A. F. E. McKenna). Wasserstoffperoxid, DECHEMA, Frankfurt
197. Canadian Industries, BE 660 719, 1965 (D. 1971; 43.
Hudson, D. R. Lusby, F. Van Zeggeren). Ergänzungslieferung, DECHEMA, Frankfurt,
198. Kymmene AB, DE 1 054 970, 1954 (R. M. R. 2000;
Örnhjelm). R. Bloom, L.E. Weeks, Ch. W. Raleigh,
199. Kymmene AB, DE 1 036 829, 1955 (R. M. R. Corrosion 16 (1960) 164t–170t.
Örnhjelm). 219. J. Mackenzie, Chem. Eng. 1990, 84–90.
200. Kymmene AB, DE 1 063 131, 1956 (R. M. 220. C. N. Satterfield, G. M. Kavanagh, H. Resnick,
Rafaelson). Ind. Eng. Chem. 43 (1951) no. 11, 2507.
201. T. S. Sheriff, J. Chem. Soc. Dalton Trans. 221. J. M. Monger, H. J. Baumgartner, G. C. Hood,
1992, 1051. C. E. Sanborn, Journ. Chem. Eng. Data 9
202. M. Traube, Ber. Dtsch. Chem. Ges. 15 (1882) (1964) no. 1, 119.
2434. 222. E. S. Shanley, P. F. Strong, L. R. Darbee, N. D.
203. F. Fischer, O. Priess, Ber. Dtsch. Chem. Ges. Lee, Ind. Eng. Chem. Fundam. 9 (1970) no. 4,
46 (1913) 698. 574.
204. B. Kastening, W. Faul, Chem. Ing. Tech. 49 223. C. N. Satterfield, P. J. Cocotti, A. H. R.
(1977) no. 11, 911. Feldbrugge, Ind. Eng. Chem. 47 (1955) 1040.
205. J. A. McIntyre, R. F. Philips in T. R. Beck et 224. Bur. Mines Inf. Circ. 8387 (1968).
al. (eds.): Electrochemical Process and Plant 225. E. Haeuseler, Explosionsstoffe 6/7 (1953) 64.
Design, Pennington, NJ, 1983, pp. 79–97. 226. a) N. Nimmerfroh, D. Pauls, S. McMahon,
206. Dow Chemical, EP 86 896, 1982 (J. A. Propulse HTP Manufaxture: Quality,
McIntyre, R. F. Philips). Transportation and Handling, 36 th AIAA /
207. H-D Tech, EP 216 428, 1986 (A. L. Clifford, ASME / SAE / ASEE Joint Propulsion
D. F. Dong). Conference, Huntsville / AL, 2000; b) N.
208. Metallgesellschaft, WO 94 24 336, 1994 (E. Nimmerfroh, E. Walzer, C. Brossmer,
Hillrichs, M. Kienberger, U. Sander). European Space Agency, Special Publication
209. Occidental, US 4 384 931, 1981 (R. J.
SP-484 (2001) 65–73.
Jasinski, C. G. Kuehn).
227. E. S. Shanley, F. P. Greenspan, Ind. Eng.
210. P. C. Foller, R. T. Bombard, J. Appl.
Chem. 39 (1947) no. 12, 1536.
Elektrochem. 25 (1995) 613–627.
228. J. M. Monger, H. Selbo, D. C. Lehwalder, J.
211. C. Oloman: Electrochemical Processing for
Chem. Eng. Data 6 (1961) no. 1, 23.
the Pulp and Paper Industry, The
229. C. W. Jones: Applications of Hydrogen
Electrochemical Consultancy Ltd., Romsey,
Peroxide and Derivatives, RSC Clean
Hants, England, 1996, pp. 143–152.
Technology Monographs, The Royal Society
212. H-D-Tech, US 5 106 464, 1990 (D. J. Rogers,
of Chemistry, Cambridge 1999.
A. L. Clifford).
230. Bur. Mines Rep. Invest. 5877 (1961).
213. S. P. Webb, J. A. McIntyre: “Generation of
231. O. Klais, Thermochimica Acta 225 (1993)
Hydrogen Peroxide in a Shorted Fuel Cell”,
213–222.
Int. Forum Electrolysis Chem. Ind., 10th
232. P. Urban, L. Bretherick: Handbook of Reactive
Electrosynthesis, Lancaster, NY 1996, p.
Chemical Hazards, 6th ed.,
213–225.
Butterworth-Heinemann, Oxford 1999.
(1996), p. 213–225.
214. Dow Chemical, WO 95 30 474, 1995 (J. A. 233. Shell Chemical Company, Concentrated
McIntyre, S. P. Webb); Hydrogen Peroxide, Summary of Research
Dow Chemical, WO 97 13 006, 1996 (J. A. Data on Safety Limitations, New York 1961.
McIntyre, S. P. Webb).
36 Hydrogen Peroxide
234. A. Schreck, A. Knorr, K. D Wehrstadt, P. A. 250. TRGS 900, Technische Regeln für
Wandrey, T. Gmeinwieser, J. Steinbach, Gefahrstoffe, Grenzwerte in der Luft am
Journal of Hazardous Materials A108 (2004) Arbeitsplatz, Bekanntmachung des BMA
1–7. Ausschuß für Gefahrstoffe, BArbBl 7/8
235. PetroChem. News 43 (2005) no. 44, 1. (2004).
236. European Center of Ecotoxicology and 251. Health and Safety Executive, EH40/ 2005
Toxicology of Chemicals (ECETOC), Special Occupational Exposure Limits (2005).
Report No. 10, Hydrogen Peroxide OEL 252. European Center of Ecotoxicology and
Criteria Document, ECETOC, Brussels 1996. Toxicology of chemicals (ECETOC), Joint
237. S. A. Kelly, C. M. Havrilla, T. C. Brady, K. H. Assessments of Commodity Chemicals No. 22,
Abramo, E. D. Levin, Env. Health Perspect. Hydrogen Peroxide, ECETOC, Brussels 1993.
106 (1998) 375–384. 253. R. P. Wayne: The photochemistry of ozone. In:
238. CEFIC Peroxygens hydrogen peroxide risk O. Hutzinger (ed.): The handbook of
assessment task force, Preliminary environmental chemistry 2E, reactions and
Experiments for the Determination of processes, Springer, Berlin, 10 (1988).
Hydrogen Peroxide in Blood, unpublished 254. H. Sakugawa, I. R. Kaplan, W. Tsai, and Y.
report, (1999). Cohen, Envir Sci. Technol. 24 (1990)
239. European Union, Institute for Health and 1452–1462.
Consumer Protection, European Chemicals 255. J. G. Calvert and W. R. Stockwell, Envir. Sci.
Bureau, Existing Substances 2nd priority list Technol. 17 (1983) 28A–443A.
38 (2003) Risk Assessment Report Hydrogen 256. W. C. Schumb, C. N. Satterfield, R. L.
peroxide, European Commission Joint Wentworth, (eds): Hydrogen Peroxide.
Research Centre, EUR 20844 EN. Reinhold Publishing, New York, 1955, 64–68,
240. CEFIC Peroxygen Sector Group, Hydrogen 181, 458–464, 535–547.
Peroxide: 28-day Inhalation Study in Rats 257. L. Kleinman, J. Geophys. Res. 91 (10) (1986)
unpublished report (2002). 889–904.
241. V. Riihimäki, A. Toppila, P. Piirilä, E. 258. E. Hellpointner, S. Gaeb, Nature 337 (1989)
Kuosoma, P. Pfäffli, P. Tuomela, J. Occup. 631–634.
Health 44 (2002) 433–438. 259. W. J. Cooper, R. G. Zika, Science 220 (1983)
242. A. Romanovski, J. R. Murray, M. J. Huston,
711–712.
Pharm. Acta Helv. 35 (1960) 354–357.
260. R. G. Zika, J. W. Moffett, R. G. Pestane, W. J.
243. CEFIC Peroxygen Sector Group: Hydrogen
Cooper, E. S. Saltzman, Geochim. Cosmochim.
Peroxide 13-Week Drinking Water Study with
Acta 49 (1985) 1173–1184.
6-Week Recovery Period in C57BL/6NCrlBR
261. J. M. Moffett, and R. G. Zika, Environ. Sci.
Mice, unpublished report, (1997).
244. M. Fenech, J. Crott, J. Turner, S. Brown, Technol. 21 (1987) 804–810.
262. W. J. Cooper and D. R. S. Lean, Environ. Sci.
Mutagenesis, 14 (6) (1999), 605–612.
245. CEFIC peroxygen sector group, Coming Technol. 23 (11) (1989) 1425–1428.
Hazelton Report No. 514/24, 1997 263. W. Guhl, A. G. Berends, Surfactants 2 (2000)
(unpublished). 98–102.
246. Threshold Limit Values for Chemical 264. A. H. C. Groeneveld, W. A. de Groot:
Substances and Biological Exposure Indices, Activated sludge, respiration test with
ACGIH, Cincinnati 2005. hydrogen peroxide, Unpublished report
247. Sociaal Economische Raad (NL), MAC (1999), Solvay pharmaceuticals, report
Waarden, Waterstoffperoxide 90 % (2005). A.SOL.S.003.
248. Koninklijk besluit van 11 maart 2002 265. Degussa AG, US-IT No. 91-0114-DKO, 1991
betreffende de bescherming van de gezondheid (unpublished).
en de veiligheid van de werknemers tegen de 266. C. A. Clark: Ph. D. Thesis, University of Bath,
risico’s van chemische agentia op het werk 1991.
(B.S. 14.3.2002, ed. 2; erratum B.S. 26.6.2002, 267. M. Fitzpatrick, C. B. Schreck, R. L. Chitwood,
ed. 2; erratum B.S. 4.12.2002) BIJLAGE I The Progressive Fish Culturist, 57 (1995)
Grenswaarden voor beroepsmatige 153–155.
blootstelling. 268. P. Klerks, P. C. Fraleigh, J. AWWA 1991,
249. Réferences françaises, INRS, ND2098, 92–100.
Valeurs limites d’exposition professionnelle 269. M. A. Xenopoulos, D. F. Bird, Photochem.
aux agents chimiques en France (2005). 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)
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- 3.3.1. Hydrogen Sulfide as a Nucleophilic
idizing agents to form sulfur compounds with Reagent
varying degrees of oxidation. Depending on op-
erating conditions and the oxidizing species con- Hydrogen sulfide reacts as a nucleophilic
cerned, sulfur oxides (SO2 , SO3 ), salts (S2 O2−
3 , reagent in the presence of acidic, basic, or metal-
SO2−3 , SO 2−
4 ), polythionic compounds, sulfur, lic oxide catalysts.
and sulfanes are obtained. Compounds such as alcohols or olefins react
with hydrogen sulfide to form thiols (RSH). Thi-
1) Aqueous solutions of hydrogen sulfide can ols (with the exception of tertiary thiols) react
be oxidized in air. like hydrogen sulfide to give the corresponding
2) In air, solutions of ammonium sulfide form organic sulfides (RSR).
orange polysulfides and even oxygen – sulfur
compounds. Substitution Reactions. Because of its
3) In an acid medium, iodine quantitatively oxi- widespread industrial applications, the best
dizes hydrogen sulfide to sulfur. This reaction known substitution reaction is sulfhydrolysis
is used to analyze hydrogen sulfide. (thiolation) of alcohols. This reaction occurs cat-
4) Strong oxidizing agents (e.g., MnO− −
4 , ClO , alytically in the gas phase at ca. 300 ◦ C. Thiols
Cl2 , and Br2 ) react with hydrogen sulfide to were first obtained by reacting hydrogen sul-
give sulfate ions. fide with alcohols on thorin or on pumice im-
5) Hydrogen sulfide can react with sulfur to pregnated with thorin [73]. Subsequently, more
form sulfanes. The sulfanes can be converted selective, more active catalysts were developed
back to hydrogen sulfide by base catalysis based on activated alumina containing promot-
[70]. ers such as soda or potash [74], metallic oxides,
6) Hydrogen sulfide or alkaline hydrosulfides alkaline salts of transition-metal oxo acids [75],
can reduce polysulfides to thiols [71]: and heteropolyacids or their alkaline salts [76].
Primary alcohols are converted very effi-
R−S−S−R + 2 NaSH −→ 2 RSH + Na2 S2 ciently into the corresponding thiols:
This reaction is used for the preparation of
H2 S + ROH −→ RSH + H2 O
thiokol polymers.
7) Hydrogen sulfide can reduce sulfoxides to However, with secondary and especially ter-
sulfides [72]: tiary alcohols, dehydration reactions are more
prevalent.
R−SO−R + H2 S −→ R−S−R + H2 O + S Under catalytic reaction conditions simi-
8) The reaction of hydrogen sulfide and sulfur lar to those used for alcohols, hydrogen sul-
dioxide (Claus reaction) is used for the pro- fide reacts with aliphatic ethers to produce thi-
duction of sulfur (→ Sulfur): ols. Cyclic ethers are converted to the cor-
responding sulfides by catalysis in the gas
2 H2 S + SO2 −→ 3/8 S8 + 2 H2 O phase. For example, reaction of hydrogen sulfide
with tetrahydrofuran yields tetrahydrothiophene
[110-01-0] [77]:
reaction of carbon dioxide with water and this rather than to hydrosulfide or sulfur ions. For ex-
step is much slower than direct reaction. An im- ample, bacterial metabolism is inhibited when
portant principle of selective absorption is there- the hydrogen sulfide content of the surround-
fore the selection of a gas – liquid contact time ings reaches 70 – 200 mg/L [130]. Because hy-
that is long enough to eliminate hydrogen sul- drogen sulfide does not accumulate in living or-
fide, but short enough to minimize the coabsorp- ganisms, it cannot contaminate the food chain
tion of carbon dioxide. [131]. Natural elimination of hydrogen sulfide
If the partial pressure of hydrogen sulfide occurs mainly by diffusion into the atmosphere,
in the sour gas is sufficiently high (≥ 0.3 MPa) but it is also subjected to biological or chemi-
and significant quantities of hydrocarbons (a few cal degradation to form sulfur oxides and sulfur.
percent) can be tolerated, a physical solvent pro- In water, hydrogen sulfide is converted into ele-
cess may also be chosen. mental sulfur or insoluble heavy-metal sulfides.
Accidental leaks of hydrogen sulfide gas can
Distillation. The industrial distillation of an be absorbed directly with aqueous solutions of
acid gas in which the purity of hydrogen sulfide sodium carbonate, sodium hydroxide, sodium
is insufficient was developed by Elf Aquitaine hypochlorite, or iron(III) chloride. When liquid
at its Lacq plant [128]. Hydrogen sulfide is ob- hydrogen sulfide is spilled on the ground, adsor-
tained either as a liquid or as a gas with a purity bents such as activated carbon should be used to
> 99.5 %. The acid gas is first treated with ac- prevent the spread of liquid and reduce the emis-
tivated carbon to eliminate heavy hydrocarbons sion of vapor [132]. Water polluted with hydro-
and dried with a solid adsorbent (alumina, sil- gen sulfide can be stripped and neutralized with
ica beads, or molecular sieves). It is then com- lime, sodium carbonate, or sodium bicarbonate.
pressed to 2.5 – 3.5 MPa and distilled in a tray Sulfur ions are then eliminated by precipitation.
column or packed column. Liquid hydrogen sul- Industrial effluents containing hydrogen sul-
fide is obtained at the bottom of the column and fide must be treated or incinerated if their hy-
easily vaporized for more convenient distribu- drogen sulfide content exceeds 0.001 vol %. The
tion. final procedure depends on the sulfur content
If the acid gas has a very low hydrogen sulfide and local discharge regulations. For example,
content (H2 S/CO2 <0.2), separation of hydro- for residual gases in sulfur plants, treatment can
gen sulfide and carbon dioxide can be improved consist of
by introducing a C3 – C6 alkane above the col-
1) hydrogenation of residual sulfur compounds,
umn feed [129]. The hydrogen sulfide obtained
followed by washing with amines that selec-
is slightly less pure.
tively absorb hydrogen sulfide and recycling
of the effluent from the amine regenerator
to the inlet of the sulfur plant: Scot (Shell),
5. Environmental Protection Elfapur (Elf Aquitaine) processes;
Hydrogen sulfide is not only produced industri- 2) incorporation of an additional treatment
ally, it also originates from natural sources (see stage based on the catalytic Claus reaction
also Chap. 1): at low temperature (135 –150 ◦ C). The sul-
fur obtained remains trapped in the cata-
1) diffusion into underground water from oil lyst and must be desorbed during a subse-
and natural gas wells, quent regeneration cycle: Sulfreen process
2) formation in lakes by aerobic bacterial fer- (Elf Aquitaine, Lurgi).
mentation,
When the sulfur concentration is not too high,
3) production during biological wastewater
washing the residual gas with chemically regen-
treatment, and
erable substances (as used in the LO-CAT pro-
4) formation as a degradation product in the
cess) or with nonregenerable substances (e.g.,
agricultural food industry, slaughterhouses,
sodium hypochlorite or sodium hydroxide) may
and the paper industry.
be more economical. Alternatively, the gases
In the environment, the toxicity of hydrogen sul- may be treated by conventional adsorption (ac-
fide is due to the hydrogen sulfide molecule itself tivated carbon, metal oxides) or by adsorption
Hydrogen Sulfide 11
combined with biological regeneration [133]. Anhydrous hydrogen sulfide does not react
The latter involves immobilizing bacteria on an with steel at ambient temperature. However, in
adsorbent to incite autopurification. 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,
6. Analysis causing blistering and embrittlement. The de-
There is no universal method of analysis for the gree of corrosion depends on the grade of steel
determination of hydrogen sulfide. Each method and is very limited with stainless steel. Fer-
depends not only on the medium in which the ritic steels are much more sensitive to hydro-
quantity of hydrogen sulfide is to be deter- gen embrittlement than austenitic steels. Embrit-
mined (gaseous mixtures, liquid effluents) and tlement can be important with highly resistant
on the concentration of hydrogen sulfide in the steel. High-strength steel exposed to hydrogen
medium, but also on the nature of the desired sulfide is likely to crack under stress. Stress-
results. relieving treatment after welding reduces resid-
The following examples are by no means ex- ual stresses. Thus, the steel used for hydrogen
haustive: sulfide storage vessels and pipes must be chosen
1) Determination in a sour gas by iodometry carefully.
[134], gas-liquid chromatography [135], and Because of the problems associated with
argentimetry [136]. transportation of hydrogen sulfide and its toxic-
2) Determination in a commercial purified gas ity, storage capacities are restricted; the product
by iodometry [137], polarography [138]. must be used close to its production site when-
3) Determination in a fuel gas with lead acetate ever possible. Nevertheless, hydrogen sulfide is
[139]. transported by pipeline over several kilometers
4) Determination in the atmosphere [140]. in Europe and the United States. Liquid hydro-
gen sulfide is also transported in specially de-
signed containers. Particular attention should be
7. Storage, Handling, and given to emptying devices (valves) and their pro-
tection during transportation.
Transportation Storage tanks for liquid hydrogen sulfide
should be equipped with a safety device system
7.1. Storage and Handling
as shown in Figure 4. The blowdown tank is a
The most important problem in storing hydro- reserve storage capacity that enables rapid emp-
gen sulfide is the material used for the storage tying of liquid hydrogen sulfide in case of acci-
vessel. dent. It can be operated by an “emergency stop”
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
12 Hydrogen Sulfide
switch that automatically actuates the safety state, itmust first be degassed to remove all traces
valves. The gas phase in the storage tank is con- of hydrogen sulfide. The sales specification of
nected directly to a flare by means of a relief sulfur in Europe requires a hydrogen sulfide con-
valve that can be bypassed via a safety valve. tent of < 10 ppm. This value is obtained by de-
gassing liquid sulfur into sulfur pits with ammo-
nia before dispatch.
7.2. Transportation
Production of Metal Sulfides. The reaction
Transportation of liquid hydrogen sulfide by of hydrogen sulfide with soda or lime solutions
air (IATA) is strictly prohibited. Transporta- results in formation of the corresponding sul-
tion in bulk containers by sea is also prohibited fides. Sodium sulfide is generally marketed in
(IMDG). Small containers of < 1 t or small bot- a solid hydrated form. Acid sulfides or hydro-
tles are generally used for road and rail trans- sulfides [e.g., NaSH or Ca(SH)2 ] are marketed
portation. The transport of hydrogen sulfide in either as crystals or as concentrated aqueous so-
Europe by rail has been abandoned; it is used lutions.
instead at the site of production. In the United Sodium sulfide is used as a lignin demethyla-
States, liquid hydrogen sulfide is transported in tion agent in the Kraft process for manufacturing
specially designed tankers. Hazard classifica- paper pulp. Mixtures of sodium and calcium sul-
tions for hydrogen sulfide are as follows: fides are powerful reducing agents of disulfide
IMDG Code, class 2.1, label 2.1 + 6 bonds. They attack the cystine molecules of ker-
RID/ADR, class 2, label 3 + 6.1 atin and are used in the leather industry for chem-
United States, CFR 49 : 172.02, flammable ical depilation of animal hides (→ Leather).
UN no. 1053
Flotation. Preliminary treatment with hy-
The sulfur content of crude oil can result in the
drogen sulfide is sometimes required for the pu-
production of hydrogen sulfide, representing a
rification of ore by selective flotation. Certain
risk for field personnel and for the transporter.
mineral species in the ore are sulfurized, which
In sea transport, maintenance of storage tanks
considerably increases the selectivity of flotation
under a blanket of inert gas does not eliminate
collectors and facilitates separation of valuable
the risk of explosion. The risk of sulfur forma-
metallic species. This process is used in the pu-
tion in the tankers should also be kept in mind.
rification of nickel and manganese.
Wearing gas masks in the pump room of tankers
is compulsory when the hydrogen sulfide con-
tent exceeds 10 ppm. The same measures should Catalyst Activation and Poisoning. Certain
be taken for personnel connecting up the loading catalysts reach their maximum activity only
arms. 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.
8. Uses However, hydrogen sulfide also poisons some
Production of Sulfur. Natural gas con- hydrogenation and reforming catalysts with a
taining hydrogen sulfide is currently one of platinum or palladium base. This can be used to
the most important sources of elemental sulfur. advantage to reduce initial catalyst activity and
For commercial use, sour natural gas must first limit coking (i.e., formation of carbonaceous de-
be sweetened to remove the acid compounds, posits). Gasoline reforming catalysts, for exam-
mainly hydrogen sulfide and carbon dioxide. ple, are always selectively poisoned with a small
The resulting acid gas is used for the produc- quantity of hydrogen sulfide prior to use.
tion of sulfur by the Claus process [141, 142]:
Treatment of Metallic Surfaces. Superfi-
2 H2 S + SO2 −→ 3/n Sn + 2 H2 O cial sulfurization of metals with hydrogen sul-
fide allows modification of their physical or
For further details, see → Sulfur. Sulfur is chemical properties. Hydrogen sulfide is used
marketed in liquid or solid form. In the liquid
Hydrogen Sulfide 13
for passivation of the walls of reactors operating n-octanethiol. These are used as intermediates
at high temperature in petrochemical operations in the synthesis of plant protection agents and as
such as steam cracking and hydrodealkylation. polymerization agents.
This treatment prevents desirable secondary re-
actions.
Another use of hydrogen sulfide is the forma- 9. Economic Aspects
tion of a layer of sulfide on the surface of steel
wires or plates that are to be coated with paint Total world sulfur production was 56.6×106 t in
or plastic. This treatment protects the coating or 1986, compared with 51.7×106 t in 1977. Sul-
improves the action of adhesive primers. fur recovered from the treatment of sour gas (i.e.,
from hydrogen sulfide) accounts for 25 % of this
Analytical Chemistry. Hydrogen sulfide (ca. 14×106 t in 1986, 11.8×106 t in 1977).
has long been used to selectively precipitate The primary producing countries are listed
group 1, 2, and 3 metal cations. The metals are below, with their 1986 production figures in 106
qualitatively identified by the color of the sul- metric tonnes per year:
fides formed. Western Canada 5.2
United States 2.2
Production of Thioorganic Compounds. former Soviet Union 1.8
The most common industrial application of hy- Saudi Arabia 1.2
Federal Republic of Germany 1.0
drogen sulfide is the production of organic com- France 0.9
pounds with a thiol function (i.e., RSH, where
R is a hydrocarbon group) or a thiol group and Recovery of sulfur from hydrogen sulfide will
another function (e.g., alcohol, acid). continue to expand worldwide. The former So-
Table 6 shows the most important thioor- viet Union is expected to become the main pro-
ganic compounds (in terms of tonnage) pro- ducer of sulfur from sour gas due to development
duced from hydrogen sulfide in the United States of the Astrakhan gas field (hydrogen sulfide con-
and Europe. Other important thioorganic com- tent: 26 %). Astrakhan I started producing early
pounds derived from hydrogen sulfide are tetra- in 1987 and Astrakhan II is due to be put on
hydrothiophene and mixtures of thiols or sul- stream early in 1989.
fides (often containing tert-butanethiol or di- Canada should continue to be a major pro-
ethylsulfide). These are used as strong-smelling ducer because of the development of the Car-
additives for gas odorization. oline gas field in Alberta. The Middle East is
Long-chain thiols also represent a range of also expected to play a major role, with substan-
useful products, especially n-dodecanethiol and tial tonnage from Iran and Qatar. Production in
Table 6. Thioorganic compounds produced from hydrogen sulfide
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 eliminated by the respiratory tract binds to cir-
Germany is likely to stabilize at (2.3 – 2.5)×106 culating proteins in the blood and is distributed
and 1.0×106 t, respectively. In France, produc- rapidly to all tissues, including the nervous sys-
tion will decline gradually due to depletion of the tem. It is subsequently metabolized and detoxi-
Lacq gas field (hydrogen sulfide content: 15 %). fied as shown in Figure 5 [144].
Sulfur recovered from hydrogen sulfide has been Hydrogen sulfide is detoxified by oxidation
the major source of accumulated stocks, which to sulfate (the primary reaction) or by methy-
are used to swing production to balance the mar- lation (not yet confirmed). The toxicity of hy-
ket when needed. At the end of 1987, the stock drogen sulfide is due to its reaction with protein
of sulfur from sour gas, in tonnes, was constituents of essential enzymes or metallopro-
teins, especially the iron-containing metallopro-
Western Canada 6.7×106
France 2.2×106
teins of the respiratory cell chain.
Saudi Arabia 1.5×106 During chronic exposure, the amount of hy-
drogen sulfide entering the organism is so small
that it is neutralized by the above detoxification
10. Toxicology reactions. However, during massive acute expo-
sure, the detoxification capacity is exceeded; the
Hydrogen sulfide is a toxic gas. Human poison- reaction of hydrogen sulfide with proteins can
ing is related to its acute toxicity. Chronic toxic- thus lead to poisoning.
ity has not been demonstrated in people exposed
to concentrations lower than those that induce Human Risk and Pathology. The acute
acute or subacute poisoning [143, 144]. toxic risk of hydrogen sulfide and the pathology
observed in human poisoning are summarized
Animal Data. Much toxicological data has in Tables 7 and 8 (see next pages) [148, 149].
been published on animals exposed to massive
dosages of hydrogen sulfide. These data confirm
observations made on humans, but information 11. Occupational Health and Safety
on chronic or long-term effects remains limited.
The acute toxicity of hydrogen sulfide is com- The main safety precautions listed here result
parable in animals and humans. from thirty years’ experience at Lacq [150] and
Chronic toxicity studies have been carried out from the literature. Hydrogen sulfide has several
on rabbits, rats, and mice that inhaled daily doses properties that make it a highly dangerous gas:
of hydrogen sulfide at concentrations in air up
to 80 ppm over three to five consecutive months 1) high toxicity;
[143–145]. Despite numerous detailed exami- 2) flammability: it ignites spontaneously at
nations, including neurological tests, no specific 260 ◦ C (compared to 537 ◦ C for methane)
chronic effects were found. Mutagenicity testing and forms an explosive mixture with air when
has shown a lack of genotoxic effects on bacteria the hydrogen sulfide content is 4 – 44 vol %
in vitro [146] and on rats in vivo [147]. (compared to 5 – 15 % for methane) [151];
3) density: being heavier than air, it accumulates
Metabolism. Lack of a chronic toxic effect in low-lying areas of installations;
of hydrogen sulfide in animals and most prob- 4) ability to embrittle steel (Section 7.1); and
ably in humans may be explained by the fact 5) insidiousness: persons intoxicated with hy-
that it does not build up in the body. Hydro- drogen sulfide no longer discern its char-
gen sulfide is metabolized rapidly, mainly into acteristic smell of rotten eggs before the
less toxic, oxidized metabolites that are quickly dangerous threshold concentration (50 –
eliminated via the kidneys. 100 ppm) is reached.
Hydrogen sulfide is readily absorbed by the
pulmonary tract, which at the same time repre- General Precautions. Areas in which hy-
sents an excretory pathway for untransformed drogen sulfide is produced, utilized, and stored
hydrogen sulfide. Hydrogen sulfide that is not must be isolated and delimited.
Hydrogen Sulfide 15
Degree of poisoning
Early signs equilibrium problems warning signs of short duration immediate coma
Immediate effects loss of consciousness for a sudden loss of consciousness lasting a prolonged deep coma with cyanosis,
few seconds; rapid regain few minutes, followed by headache, respiratory distress, occasionally cardiac
of consciousness respiratory distress, gastric pain, distress; regain of consciousness is difficult
diarrhea (following day) and agitated
Long-term effects recovery recovery or bronchial fragility; asthenia bronchopulmonary sequelae
Emergency treatment evacuation from danger evacuation from danger area; evacuation from danger area; administration
area; administration of administration of oxygen; respiratory of oxygen; respiratory assistance if needed;
oxygen assistance if needed; intravenous intravenous vitamin C; cardiotonics and
vitamin C; medical observation for neurosedatives if needed; hospitalization
several hours
Only authorized personnel should have ac- – Nature of special risks attaching to danger-
cess to these areas, i.e., personnel having un- ous substances:
dergone medical tests to detect ocular, respira- – R.13 Extremely flammable liquefied gas
tory, or cardiovascular disease, and signs of al- – R.26 Very toxic by inhalation
coholism. All authorized personnel should un- – Safety advice concercing dangerous chemi-
dergo special training to deal with the inherent cal substances:
risks of working in a hydrogen sulfide environ- – S7/9 Keep container tightly closed in a well-
ment. They should be equipped with gas masks ventilated place
and undergo regular training in their use. – S.25 Avoid contact with eyes
The usual precautions should be taken in all – S.45 In case of accident or clinical symp-
areas where an explosive atmosphere is likely to toms medical advice should be sought im-
be found (sources of flame, electric equipment, mediately.
ect.). “No smoking” signs should be displayed
and strictly observed. Installation and Handling. Apparatus
Only piping, apparatus, and containers made should be properly closed and tightly sealed.
from special steel or alloys should be used; the Waste gas from discharge and pressure valves
grade should be suitable for the conditions of should be sent to a flare to avoid emission of
use. The use of copper and its alloys should be hydrogen sulfide into the atmosphere. This flare
prohibited. can burn off hydrogen sulfide in dangerous sit-
uations, eliminate air from the discharge piping
Storage and Transportation (see also before discharge, and purge hydrogen sulfide
Chap. 7). The storage area should be well ven- after discharge.
tilated and, if possible, equipped with a water Emissions produced under normal conditions
cooling system in case of fire. should be recovered at source, and low-lying ar-
Hydrogen sulfide containers should be stored eas should be ventilated thoroughly.
separately from incompatible products (oxi- Personnel working on apparatus containing
dizing agents, inflammable products, powerful (or having contained) hydrogen sulfide should
acids, etc.) and away from direct sunlight. A spe- be equipped with insulated gas masks and wear
cial parking area should be provided for vehicles protective clothing, goggles, and long gloves.
transporting hydrogen sulfide. No one should be allowed to work alone in a
Each container should carry a label indicat- zone where hydrogen sulfide is handled.
ing that the contents are highly flammable and Frequent checks of the atmosphere should
toxic. The label should also list the special risks be made. Hydrogen sulfide detection devices
and include cautionary advice. should be installed (see below); smell should not
be relied on for detection.
– Modified EEC directive 67/548
Programmed preventive maintenance should
– EEC No. 016-001-00-4
be carried out in all hydrogen sulfide installa-
– Danger symbols: highly flammable and toxic
tions, particularly on sealing devices. A fire per-
Hydrogen Sulfide 17
mit should be obtained to use tools producing A simple but effective method of detecting a
flame, heat, or sparks. leak is to use lead acetate paper. The paper be-
Pyrophoric sulfides should be kept constantly comes black in the presence of hydrogen sulfide
humid. They should be extracted and incinerated but gives no indication of concentration.
as soon as possible. After maintenance work, all
traces of humidity should be eliminated. Exposure Limit Values [152]. The average
permitted concentration of hydrogen sulfide in
Emergency Measures. A plan should be air varies from country to country:
drawn up for evacuation of personnel if the con- TLV (TWA) 5 – 10 ppm (vol)
centration of hydrogen sulfide exceeds the norm. TLV (STEL) 10 – 15 ppm
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 12. References
displayed prominently in the installation.
In the event of leakage without fire, the leak 1. E. Eriksson, Tellus 12 (1960) 63. H.
Roffmann, M. C. Bender: “H2 S: new sources
should be isolated and sprayed with water. Un-
and potential impacts,” 68th Annual Meeting
controllable gas should be sent to a flare pit, and
of the Air Pollution Control Association , June
all sources of heat should be removed or shut 1975.
down. 2. B. Bonsang, La Recherche 13 (1982) no. 137,
If the leaking material has caught fire, the fire 1132 – 1142.
should be put out only when it presents an espe- 3. D. Möller, Atmos. Environ. 18 (1984) no. 1,
cially serious threat to the environment or when 29 – 39.
the leak can be rapidly controlled. 4. T. P. Kudryautseva, Deposited document –
In the event of accidental gassing, lifesaving VINITI 160, 7 – 15; Chem. Abstr. 94 (1981)
intervention should be undertaken only after the no. 22, 177 614 v.
rescuer has donned an insulating gas mask. The 5. A. I. Kravtsov, G. I. Voitov, Izv. Vyssh. Uchebn.
victim should be removed rapidly from the dan- Zaved., Geol. Razved. 19 (1976) no. 4, 18 – 25;
ger zone and should receive necessary first aid Chem. Abstr. 85 (1976) no. 24, 179 877 w.
6. H. G. Spratt, Jr., M. D. Morgan, R. E. Good,
and medical treatment as soon as possible.
Appl. Environ. Microbiol. 53 (1987) no. 7,
1406 – 1411; Chem. Abstr. 107 (1987) no. 11,
Detection of Hydrogen Sulfide in Air. Per- 95 855 x.
manent detection devices are necessary both in 7. J. P. Hollis, Jr., Acta Phytopathol. Acad. Sci.
and around the hydrogen sulfide storage area. Hung. 20 (1985) no. 3 – 4, 321 – 326; Chem.
They should set off both audible and visual Abstr. 105 (1985) no. 21, 190 028 x.
alarms as soon as the concentration of hydrogen 8. I. Benda, Arch. Mikrobiol. 27 (1957)
sulfide in air exceeds 10 ppm. Several models are 337 – 374; Chem. Abstr. 53 (1959) 22 208 g.
available. An electrochemical sensor functions 9. R. Vàmos, M. Ando, Acta Univ. Acta Biol.,
on the gas diffusion principle: hydrogen sulfide Szeged. 5 (1959) 61 – 69; Chem. Abstr. 54
in air reacts chemically with an electrolyte (Sul- (1960) 17 770 f.
fytron 4010, Draeger, FRG). In a semiconduc- 10. V. G. Popov, Gazov. Prom-st Ser.: Geol.
Razved. Gazov. Gazokondens. Mestorozhd.
tor detector, air containing hydrogen sulfide dif-
1981, no. 10, 10 – 18; Chem. Abstr. 96 (1982)
fuses into the specially adapted semiconductor
no. 20, 165 091 a.
and modifies its electrical resistance (General 11. E. E. London, Geol. Nefti Gaza 1975, no. 8,
Monitors, United States). 24 – 30; Chem. Abstr. 84 (1976) no. 6,
Portable detectors can be used for random 33 375 p.
measurement of hydrogen sulfide concentra- 12. C. A. Brook, Geol. Surv. Prof. Pap. U.S. 1141
tion. An electrically or manually operated pump (1981) 193 – 203.
sucks air through a reactive tube divided into 13. R. J. Ferek, R. B. Chatfield, M. O. Andrae,
zones that become colored on contact with the Nature (London) 320 (1986) 514 – 516.
polluting gas (Draeger model 21/31 with specific 14. E. R. Lippincott, R. V. Eck, M. O. Dayhoff, C.
tube for hydrogen sulfide). Sagan, Astrophys. J. 147 (1967) 753 – 764;
Chem. Abstr. 67 (1967) 58 795 n.
18 Hydrogen Sulfide
49. Yu. N. Sokolov, B. V. Yakubenko, T. 71. J. C. Jorczak, E. M. Fettes, Ind. Eng. Chem. 43
Ya. Deputatova, Geol. Str. Neftegazonosn. (1951) 324.
Nizh. Povolzh’ya, Saratov 1981, no. 4, 72. D. Martin, H. G. Hauthal: Dimethylsulfoxide,
106 – 109; Chem. Abstr. 97 (1982) no. 26, Van Nostrand Reinhold Ltd., Wokingham,
219 164 t. England 1971.
50. A. P. Krezub, V. V. Kainash, Burenie (Moskva) 73. P. Sabatier, A. Mailhes, C. R. Acad. Sci. 150
1981, no. 12, 7 – 9; Chem. Abstr. 98 (1983) (1910) 1217.
no. 16, 128 805 g. 74. H. O. Folkins, E. L. Miller, Ind. Eng. Chem.
51. B. A. Andreson, G. P. Bochkarev, E. I. Prod. Res. Dev. 1 (1962) 271.
Ogarkova, R. Kh. Khazipov, SU 1127893 A1, 75. Pure Oil, US 2 820 062, 1958 (H. O. Folkins,
1984; Chem. Abstr. 102 (1985) no. 14, E. L. Miller).
116 396 n. 76. Pennsalt Chemicals Corp., US 3 035 097, 1962
52. J. Raczkowski, A. Dudek, A. Hejnar, Nafta (T. E. Deger, B. Buchholz, R. H. Goshorn).
(Katowice Pol.) 42 (1986) no. 4, 106 – 110; 77. Y. K. Yurev, M. N. Prokina, J. Gen. Chem.
Chem. Abstr. 106 (1987) no. 6, 35 660 z. USSR (Engl. Transl.) 7 (1937) 1868. Pennsalt,
53. Oil Gas J. 82 (1984) 76. GB 913 584, 1960 (T. E. Deger, R. H. Goshorn,
54. Ullman, 4th ed., 21, 169. B. Buchholz).
55. R. Goodwin: “Hydrogen sulfide provisional 78. Pennsalt, US 3 197 483, 1961 (T. E. Deger,
thermophysical properties from 188 to 700 K R. H. Goshorn, B. Buchholz).
at pressures to 75 MPa,” NBS Report no. 79. Société Nationale des Pétroles d’Aquitaine,
NBSIR 83 – 1694 (1983) 175. FR 1 417 154, 1964 (P. Vannel). Shell,
56. “L’Air liquide,” Encyclopédie des gaz, EP 122 654, 1980 (F. H. Elliott).
Elsevier, Amsterdam 1976, p. 935. 80. Société Nationale Elf Aquitaine (Production),
57. J. W. Hennel, K. Kikrynicki, Acta Phys. Pol. Demande brevet français 88 01879, 1988 (E.
18 (1959) 523 – 526. Arretz).
58. Gmelin, 4 a/b, p. 108. 81. Pennsalt, US 2 951 875, 1960 (B. Loev, R. H.
59. P. Pascal, “Nouveau traité de chimie minérale,” Goshorn).
vol. XIII, Masson Ed., Paris 1956. 82. Pennsalt, US 4 102 931, 1978 (B. Buchholz).
60. R. C. Weast (ed.): Handbook of Chemistry and 83. Société Nationale Elf Aquitaine (Production),
Physics, 53rd ed., CRC Press, Cleveland, EP 101 356, 1982 (E. Arretz, A. Mirassou, C.
1972/73, p. 222. Landoussy, P. Augé).
61. C. Huiszoom, A. Dynanus, Physica 31 (1965) 84. Société Nationale Elf Aquitaine (Production),
1049 – 1052. FR 2 330 674, 1975 (E. Arretz, A. Pfister).
62. A. Barua, A. Manna, P. Mukhopadhyay, J. Phillips, US 3 963 785, 1975 (D. H. Kubicek).
Chem. Phys. 49 (1968) 2422 – 2425. Seitetsu, JP 79 132 551, 1978 (M. Kawamura,
63. L’Air liquide: Encyclopédie des gaz, Elsevier, M. Hatta, M. Nakamura, N. Koune, Y.
Amsterdam 1976, p. 936. Iwasaki).
64. M. Burgess, R. Germann. AIChE J 15 (1969) 85. R. Dahlbom, Acta. Chem. Scand. 5 (1951)
272 –275. 690.
86. Toyo Chemical Industry, JP 81 147 763, 1980.
65. J. D. Zel’venski, A. V. Strunina, Gazov.
Phillips, EP 208 323, 1985 (J. S. Roberts).
Promyšlennost 6 (1960) no. 1, 42 – 47.
87. Henkel, BE 731 879, 1968. Henkel, BE
66. E. Isaacs, F. Otto, A. Mather, Can. J. Chem.
731 881, 1968.
Eng. 55 (1977) 751 – 752.
88. Société Nationale des Pétroles d’Aquitaine,
67. A. Kohl, F. Riesenfeld: Gas Purification, 3rd
FR 1 359 678, 1963 (B. Pflugfelder, P. Vannel).
ed., Gulf Pub. Co., Houston 1976, p. 783.
Toyo Kasei, JP 81 110 667, 1980.
68. M. King, H. Al Najjar, Chem. Eng. Sci. 32
89. Pennwalt, US 4 281 202, 1980 (B. Buchholz,
(1977) 1241 – 1246.
C. B. Welsh, H. C. Henry).
69. G. Charlot: Les réactions chimiques en
90. Société Nationale Elf Aquitaine (Production),
solutions aqueuses, Masson, Paris 1983.
EP 274 934, 1988 (E. Arretz, P. Augé, A.
70. Société Nationale des Pétroles d’Aquitaine,
Mirassou, C. Landoussy).
FR 1435 788, 1964 (J. Solinhac). Shell, FR
91. GB 585 655, 1947 (F. N. Woodward). Röhm
2 134 677, 1972 (R. S. Anderson). Exxon, EP
and Haas, DE 1 221 217, 1965 (H. Walter, W.
45 636, 1981 (L. T. Howard). Elf Aquitaine,
Gruber). Société Nationale Elf Aquitaine
FR 2 601 350, 1986 (R. Voirin, A. Pepy).
(Production), FR 2 483 917, 1980 (E. Arretz).
20 Hydrogen Sulfide
Société Nationale Elf Aquitaine (Production), 116. R. C. Forney, J. M. Smith, Ind. Eng. Chem. 43
US 4507505, 1985 (E. Arretz). (1951) 1841.
92. T. L. Cairns et al., J. Am Chem. Soc. 74 (1952) 117. W. A. Bryce, J. Chem. Soc. 4 (1949) 3379.
3982. R. Mayer, Angew. Chem. Int. Ed. Engl. 118. F. Seel, Angew. Chem. 68 (1956) 789.
2 (1963) 277. 119. F. Gomez Vigide et al., Inf. Quim. Anal.
93. Société Nationale des Pétroles d’Aquitaine, (Madrid) 13 (1959) 61.
FR 1 578 891, 1968 (Y. Labat). 120. M. A. Korshunov et al., Zavod. Lab. 34 (1968)
94. B. Magnusson, Acta Chem. Scand. 13 (1959) 285; Chem. Abstr. 69 (1968) 7944.
1031. 121. E. N. Konovalov et al., Gasov. Prom. 6 (1972)
95. Du Pont, US 2 551 813, 1949 (P. S. Pinkney). 47; Chem. Abstr. 77 (1972) 77 435.
96. Stauffer, US 3 689 568, 1969 (R. J. Eletto). 122. R. M. Levit, Zavod. Lab. 27 (1961) 352;
97. Phillips, US 3 223 738, 1965 (D. L. Crain). Chem. Abstr. 55 (1961) 25 387.
98. Phillips, US 3 050 452, 1960 (R. P. Louthan). 123. A. L. Kohl, F. C. Riesenfeld: Gas Purification,
99. Pennwalt, US 4 140 604, 1978 (D. A. Dimmig). 4th ed., Gulf Publishing Co., Houston 1985.
100. Société Nationale Elf Aquitaine (Production), 124. J. I. Lee, F. D. Otto, A. E. Mather, J. Chem.
EP 5 400, 1979 (J. Ollivier, G. Souloumiac, J. Eng. Data 17 (1972) 465.
Suberlucq). 125. R. N. Maddox: Gas and Liquid Sweetening,
101. Société Nationale Elf Aquitaine (Production), 3rd ed., Campbell Petroleum Series, Norman
EP 60754, 1981 (E. Arretz, C. Landoussy, A. (OK) 1982.
Mirassou, J. Ollivier). 126. Proceedings of Gas Conditioning
102. Akzo, US 3 927 085, 1974 (A. G. Zenkel, A. Conferences, University of Oklahoma –
Toth, H. Nagerlein, G. Meyer). Norman (OK) (since 1981).
103. Dow Corning Corp., US 4 082 790, 1975 (J. L. 127. C. Blanc, J. Elgue, Hydrocarbon Process. 60
Speier). Stauffer, US 4 740 623, 1988 (J. B. (1981) no. 8, 111 – 116.
Heather). Société Nationale Elf Aquitaine 128. Elf Aquitaine proprietary documents,
(Production) Demande brevet françois Licensee: Elf Aquitaine.
8 804 693, 1988 (Y. Labat). 129. J. M. Ryan, A. S. Holmes, US 4 293 322, 1981;
104. J. E. Bittel, J. L. Speier, J. Org. Chem. 43 Chem. Abstr. 96 (1982) P 37667.
(1978) 1687. 130. Dangerous Properties of Industrial Materials,
105. P. Noble, D. S. Tarbell, Org. Synth. 4 (1963) N-I-SAX-1979. Van Nostrand, New York
924. 1979.
106. Société Nationale des Pétroles d’Aquitaine, 131. Oil and Hazardous Materials. Technical
US 3 781 366, 1973 (E. Arretz, A. Donche, A. Assistance Data System U.S., Environmental
Pfister). Protection Agency, Oil and special materials
107. Wakunaga, Yakuhin Kabushiki Kaisha, US Control Division, Washington 1980.
4 212 826, 1979 (H. Yamaguchi). 132. Sulfure d’hydrogène – Collection Enviroguide.
108. Sankyo Organic Chemicals Cy, US 3 840 586, Environnement Canada-Ottawa 1984.
1974 (T. Chiba et al.). Degussa, US 4 490 307, 133. Société SAPS – Anticorrosion – Brevet
1984 (H. Kuenk). Français no. 2 591 122.
109. S. Dushman, J. Am. Chem. Soc. 43 (1921) 427.
134. Determination of H2S in Sour Gas by
110. C. M. Thacker, E. Miller, Ind. Eng. Chem. 36
Iodometry, SNEA (P) Internal Method no. 52.
(1944) 182.
135. Determination of H2S in Sour Gas by
111. K. T. Zimmermann, Angew. Chem. 74 (1962)
Gas-liquid Chromatography, SNEA (P)
151.
Internal Method no. 80.
112. K. Pohl, G. Rebentish, Chem. Tech. (Leipzig)
136. Determination of H2S in Sour Gas by
19 (1967) 503.
Argentimetry, SNEA (P) Internal Method
113. L. T. Filimonova et al., Neftepererab.
no. 53.
Neftekhim. (Moscow) 1 (1976) 42; Chem.
137. Determination of H2 S in Natural Gas by a
Abstr. 85 (1976) no. 96, 155.
Cadmium-Sulfate-Iodometric Method, ASTM
114. Girdler, Company publication, “Compact
D 2385.
Package H2S generators” Licensee: Chemical
138. Determination of H2S in a Purified
and Industrial Corp., Girdler, Zurich,
Commercial Natural Gas by a Polarographic
Switzerland.
Method, SNEA (P) Internal Method no. 134.
115. R. F. Bacon, E. S. Boe, Ind. Eng. Chem. 37
(1945) 469.
Hydrogen Sulfide 21
Hydrometallurgy → Metallurgy
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
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.
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].
page 4 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Protium Deuterium Tritium
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).
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.
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.
p-Hydrogen n-Hydrogen
Triple point
Temperature, K 13.803 13.957
page 6 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Density (vapor), kg/m3 0.126 0.130
Heat of vaporization,
J/mol 898.30 899.1
Liquid phase
Density, kg/m3 70.78 70.96
Thermal conductivity,
Thermal conductivity,
page 7 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Properties at STP
(273.15 K, 101.3 kPa)
Thermal conductivity,
* The reference state for enthalpy is zero for the ideal gas at 0 K.
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
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.
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.
page 9 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Dielectric constant 1.286
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.
T, K Solubility, Ostwald –
103 mol % coefficient ×102
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.
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
• 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].
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
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.
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
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).
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).
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].
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
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
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
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
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
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.
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
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.
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
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
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.
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
Feedstock
Coal type North Dakota lignite Bituminous coal Bituminous coal
Volatile matter, wt % 26.9 34.3 33.5
page 22 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Carbon, wt % (waf*) 73.9 83.6 76.7
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).
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
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)
(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]
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
Shift reaction
CO + H2O CO2+ H2 (14) –41.16 220 250 2.0 –37.19
Carbonyl formation
Ni + 4 CO Ni(CO)4 (23) –159.50
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
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
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.
(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.
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:
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
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.
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.
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.
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
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.
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.
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:
Pressurized Fireboxes. The pressure in the fireboxes of conventional tubular reformers is atmospheric or slightly below. The reasons for this are:
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].
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].
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.
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].
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
Production and recovery of hydrogen from the byproducts of the following types of refinery processes are possible:
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
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
Naphthenes, vol % 30 – 50
Reformate, wt % 75 – 88
Paraffins, vol % 30 – 20
Aromatics, vol % 65 – 85
Hydrogen, wt % 2–3
Composition of H2-rich gas typical
H2,vol % 75 – 85 82
page 36 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
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 %
CH4, mol % 5 32 30
C2H4, mol % 32 32 32
C2H6, mol % 28 5 6
C3H6, mol % 1 11 16
C3H8, mol % 1 1
H2S, mol % 1 3 5
H2-fraction of demethanizer
H2, mol % 80 – 95
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)
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
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)
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).
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.
(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)
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.
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):
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).
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
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.
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.
Normal-Pressure Electrolysis Units. The advantages of a process at almost ambient pressure are:
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
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
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.
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].
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
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]
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:
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.
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)
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.
(36)
The reaction is endothermic, the thermodynamic standard values are given below [129]:
Reaction
enthalpy
Reaction
entropy
Gibbs free
energy
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.
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 , %
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
2 HBr + Hg HgBr2+ H2
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
CO + 2 H2 CH3OH
½ As2O5 ½ As2O3+ ½ O2
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
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)
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].
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].
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.
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
CO + 2 H2 CH3OH
CH3OH CH2O + H2
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:
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.
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.
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].
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]
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
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
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
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 %.
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.
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.
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.
mL mmol
page 58 of 133
Hydrogen : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Purple bacteria
Rhodospirillum rubrum 146 6.5
mL mmol time, h
Chp + Fd + H2-ase (C. pasteur) 200 – 350 9.0 – 15.5 3.0 – 6.0
*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
(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.
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].
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].
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].
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]
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.
(54a)
(54b)
(55a)
(55b)
(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]).
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.
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)
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]
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.
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.
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
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.
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].
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).
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]
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]
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.
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.
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.
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.
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
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).
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).
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
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
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:
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.
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.
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.
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.
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 %
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:
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.
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.
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:
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.
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.
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]
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.
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 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:
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.
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.
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)
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.
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:
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.
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
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.
MPa composi-
% H2 others tion, % H2
Ammonia synthesis purge gas 60 N2, CH4, Ar, NH3 13.0 H2, 85 water wash, NH3– conc.
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).
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
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.
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.
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)
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:
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.
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
CO ≤0.1
CO2 ≤0.1
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.
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.
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].
Designation Compressed hydrogen gas storage Absorption by metal Cryo- Liquid hydrogen
hydrides adsorption storage
Operating conditions
Pressure, MPa 10 1.2 15 0.3 – 5 0.3 – 5 4.2 0.15 0.15
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)
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].
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:
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.
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.
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.
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]
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].
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.
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
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
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
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
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.
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.
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:
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.
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:
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).
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.
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].
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
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.
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
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].
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
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:
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.
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.
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).
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.
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.
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)
For technical and economic reasons the aforementioned methods play only a small part in the production of hydrogen for energy systems.
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:
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.
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 ( , , . . .).
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
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
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
Elemental metals
Mg – MgH2 77.4 0.1 286 7.7 0.107 9.9
Alloys
Mg2Ni – Mg2NiH4 64.5 1.0 350 3.16 0.081 4.5
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:
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
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:
operating temperature 80 °C
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
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].
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].
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].
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].
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].
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.
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].
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)
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.,
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.
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
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
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
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)
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
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]:
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.
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
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).
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].
a) Injection head; b) Gas distribution; c, d) Injection manifold; e) Combustion chamber; f) Injection rings; g) Radial water injection
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].
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
* 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]
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:
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
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.
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).
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]
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
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):
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.
The concentration of hydrogen at the catalyst Many metals and metal oxides have general
surface can vary widely, depending on the ratio hydrogenation activity, but marked differences
of the rate of hydrogen consumption to the rate exist among them. Nickel, copper, cobalt, chro-
of hydrogen supply [7]. mium, zinc, iron, and the platinum group (plat-
This rate ratio can have an important influ- inum, palladium, rhodium, and ruthenium) are
ence on selectivity. It is, for instance, an im- among the elements frequently found in com-
portant factor in controlling selective removal mercial hydrogenation catalysts. Combinations
of multiple unsaturation and the extent of cis – of these or other elements are sometimes used
trans isomerization during hydrogenation of nat- to confer additional activity, selectivity, stability,
ural oils to make edible products. or life.
Palladium, platinum, rhodium, and ruthe-
nium make exceptionally active hydrogenation
1.2. Catalysts catalysts and are used frequently when equip-
ment or stability limitations permit operation
Hydrogenation catalysts are of two types, het- only under mild conditions. They can effectively
erogeneous and homogeneous. Heterogeneous reduce most functional groups. Nickel is used to
catalysts are solids that form a distinct phase reduce a variety of functional groups, usually
in the gas or liquid environment. Homogeneous under more vigorous conditions than the plat-
catalysts dissolve in the liquid medium, forming inum group. Copper chromite is used mainly to
only a single phase; a well-known catalyst of this reduce esters to alcohols. Cobalt or iron, in the
type is the Wilkinson catalyst, (Ph3 P)3 RhCl. presence of ammonia, is an effective catalyst for
the reduction of nitriles to primary amines.
Homogeneous Catalysts. Homogeneous
hydrogenation catalysts are used infrequently Catalyst Selectivity. In series reactions of
in industry. They are difficult to separate from the type A → B → C, both mechanistic and ther-
the reaction media and are not employed if modynamic selectivity are involved in the for-
a suitable heterogeneous catalyst is available. mation of B. These selectivities can vary widely
Nonetheless, they are useful in a variety of with the catalyst used. Thermodynamic selec-
hydrogenation processes that are not possible tivity, which is related to the relative free en-
with heterogeneous catalysts [8], [9]. Expanded ergies of adsorption of A and B, refers to the
use of homogeneously catalyzed hydrogenation competition of A and B for catalyst sites. The
awaits only industrial demand for its products. tendency for B to occupy a catalyst site rela-
A noteworthy industrial use of a homoge- tive to A increases as B accumulates and as A
neous catalyst is the asymmetric hydrogenation is depleted. High thermodynamic selectivity de-
of an achiral olefinic precusor of l-Dopa (3,4-di- mands that A be adsorbed much more strongly
hydroxyphenylalanine, used in the treatment of than B and that the rate of displacement of B
Parkinson’s disease; see → Parkinsonism Treat- by A be high relative to the rate of hydrogen-
ment) over a complex rhodium catalyst carrying ation of B to C. Thermodynamic selectivity is
chiral ligands. Optical yields are nearly quanti- affected by depletion of A in the vicinity of the
tative. catalyst because of intra- and interparticle dif-
fusion resistance. Mechanistic selectivity refers
Heterogeneous Catalysts. Heterogeneous to the extent to which C is formed from A with-
catalysts can be divided into two types: those out the formation of desorbed B; it is related to
used in fixed-bed processing, in which the cat- the relative rates of A → B and A → C in the
alyst is stationary and the reactants pass over equation B ← A → C. Either thermodynamic or
it, and those used in fluidized-bed or slurry mechanistic selectivity may be more important,
processing. The former have a relatively large but both must be high if B is to be obtained in
particle size, 0.79 – 6.35 mm, in various forms. good yield.
Slurry or fluidized-bed catalysts are fine pow-
ders that can be suspended readily in a liquid or
gas.
Hydrogenation and Dehydrogenation 3
Hydrogenation is exothermic, and some means Bubble columns resemble trickle-bed reactors
of removing the reaction heat must be provided except for the direction of flow. In bubble col-
to avoid excessive temperature. In liquid-phase umn reactors, flow is upward; liquid fills the
slurry reactions, cooling is usually accomplished void space, and the catalyst is completely wet-
by coils within the reactor. The reaction mass can ted. Ebulliated bed reactors are similar except
also be cooled by circulating the mixture over that the catalyst is fluidized by high gas and liq-
external cooling elements. The latter save reac- uid flow rates. This type of reactor is used in
tor space and permit the reactor to be cleaned H-oil and H-coal processes (see also → Coal
more easily. Liquefaction, Chap. 3.9.5.).
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- 1.4. Hydrogenation of Individual
lyst suspended in the liquid medium. Agitation Compounds
also helps in the removal of heat, and agitators
can be designed to favor either hydrogen trans- Alkynes. Alkynes are adsorbed on a cata-
port or heat removal, as required. Externally lyst surface much more strongly than the cor-
cooled systems can be engineered to return the responding alkenes, thus permitting selective
reaction slurry to the reactor by a venturi tube hydrogenation of traces of alkynes in massive
through which hydrogen is drawn into the pass- amounts of alkenes. Hydrogenation of trace
ing stream. Such a system gives efficient mixing, alkynes, usually over a low percentage of palla-
and much of the hydrogenation is thought to take dium on an alumina support with restricted hy-
place within the nozzle itself. Alternatively, the drogen supply, is used widely to purify various
slurry can be returned to the reactor through a alkene-containing streams.
shower-type arrangement in which substrate and The strong adsorption of alkynes also per-
catalyst particles are sprayed into hydrogen in mits selective reduction to the cis alkenes in high
the head space. 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-
1.3.2. Trickle-Bed Reactors
lar catalyst (Pd – Pb on CaCO3 ), with or without
additional modifiers, is often used for this selec-
Continuous hydrogenation is carried out com-
tive reduction. Ideally, the reaction comes to a
monly in trickle-bed reactors (columns packed
4 Hydrogenation and Dehydrogenation
virtual halt after absorption of 1 mol of hydro- moconjugated dienes and trienes, and an aim of
gen, but terminating the reaction after consump- hydrogenation is preferential removal of multi-
tion of the stoichiometric amount of hydrogen ple unsaturation. This task is facilitated by the
may be necessary. The largest volume hydrogen- tendency of hydrogenation catalysts to isomer-
ations are of alkynols and alkynediols. Butane- ize homoconjugated into conjugated olefins; the
1,4-diol [110-63-4], derived from acetylene, has latter, relative to monoenes, are strongly ad-
long been made by hydrogenation of butyne- sorbed on the catalyst. Catalysts differ widely
diol over a base-metal catalyst such as nickel – in their ability to promote isomerization, and
copper – manganese on silica (→ Butanediols, a close parallel exists between this ability and
Butenediol, and Butynediol, Chap. 1.3.). Nickel, the selective removal of polyunsaturation. As a
the major component, is an effective catalyst for consequence of double-bond migration, isomer-
hydrogenation of both acetylenic and olefinic ization of the naturally occurring all-cis olefins
bonds. At the butenediol stage, some aldehyde is into trans olefins may or may not occur, depend-
formed by rearrangement; this is removed effec- ing on the conformation of the chain at the time
tively by copper. A small amount of manganese of migration. Many catalysts have been stud-
stabilizes the structure. The active binary cata- ied for these reactions, but industrial practice
lyst palladium – ruthenium on carbon behaves in overwhelmingly favors some form of nickel. Hy-
similar fashion. drogenation of edible oils is carried out almost
entirely in batchwise slurry operation at 140 –
Alkenes. Hydrogenation of olefins has ex- 190 ◦ C and 0.2 – 0.3 MPa.
tensive industrial application. Dripolene, or Notwithstanding the huge volume of prod-
pyrolysis gasoline, is stabilized against oxida- uct, processing is largely carried out in a slurry
tion by selective reduction of dienes in a com- batch mode. Continuous fixed-bed operation is
plex mixture of olefins and aromatics. The pre- made difficult by intermittent production, vari-
ferred catalyst consists of pellets of palladium able feedstocks, and the problem of adjusting
on alumina. If the aim of reduction is to produce conditions to compensate for declining catalyst
gasoline, olefins are preserved to boost the oc- activity.
tane rating, but if the product is to be used as a
source of pure aromatics, olefins are removed in Aromatic and Heteroaromatic Rings.
a second stage over a base-metal catalyst, such Compounds with aromatic or heterocyclic rings
as nickel or cobalt – molybdenum. Two steps are are hydrogenated readily and usually without
used for complete olefin hydrogenation, because difficulty if other reducible groups are absent.
the more vigorous conditions required for com- Cyclohexane [110-82-7] is obtained mainly
plete reduction cause severe fouling of the cata- by hydrogenation of sulfur-free benzene over
lyst by polymerization of the diene. supported fixed-bed nickel or platinum cata-
The largest single application of olefin hy- lysts (170 – 230 ◦ C, 2.5 MPa, → Cyclohexane,
drogenation is the partial hydrogenation of Chap. 4.1.). The reaction is reversed at higher
natural oils to produce margarine, shortening, temperature, which sets an upper limit to a per-
toppings, salad oil, and other edible products missible operating temperature of ca. 230 ◦ C.
(→ Margarines and Shortenings). A huge mar- With catalysts of controlled acidity, the prod-
ket also exists for products derived by more com- uct can be obtained virtually free of benzene
plete hydrogenation; the problem of selectivity and methylcyclopentane. Tetralin [119-64-2] is
described below is not present. produced by hydrogenation of naphthalene over
Edible products are derived by partial hydro- nickel sulfide or nickel – molybdenum at 400 ◦ C
genation of various oils (e.g., cottonseed, soy- and 2 – 6 MPa; a higher pressure gives decalin.
bean, sunflower, and rapeseed oil) all of which Hexahydrobenzoic acid [98-89-5], an inter-
are glycerides of long-chain saturated and cis- mediate for caprolactam (nylon 6 monomer), is
unsaturated fatty acids (→ Fats and Fatty Oils, made by hydrogenation of benzoic acid over pal-
Chap. 4.3.). Partial hydrogenation of these prod- ladium on carbon powder by using the product
ucts improves oxidative stability and alters plas- as solvent to help prevent excessive temperature
tic properties suitably. Oxidative instability is (→ Caprolactam, Chap. 4.3.). When only base-
caused primarily by the presence of various ho- metal catalysts were available, this was consid-
Hydrogenation and Dehydrogenation 5
ered a very difficult reduction; base-metal cat- Reactor temperature is held between 155 and
alysts are attacked by the carboxylic acid func- 170 ◦ C, pressure at 0.5 – 1.4 MPa [12].
tion.
Carbonyl Compounds. Aldehydes and ke-
tones behave similarly on hydrogenation, and
the same order of catalyst activity usually ap-
An important monomer 1,4-bis(hydroxyme- plies to both functional groups. The difference
thyl)cyclohexane is made by two-stage hydro- between aromatic and aliphatic carbonyls is
genation of dimethyl terephthalate. The first marked, however, and the preferred catalysts dif-
stage (160 – 180 ◦ C, 30 MPa), carried out in a fer.
fixed-bed reactor over supported palladium, pro- Aliphatic carbonyl compounds are reduced
duces dimethyl hexahydroterephthalate. Part of to the corresponding alcohols with little danger
this product is recycled to moderate the temper- of overhydrogenation. Alcohols are in fact an
ature increase resulting from the highly exother- excellent solvent for hydrogenation, even under
mic hydrogenation (197.6 kJ/mol). The second vigorous conditions.
stage uses a copper chromite catalyst under sim- Hydrogenation of mesityl oxide [141-79-7]
ilar conditions. 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 two-stage process makes temperature control in vapor-phase fixed-bed reactors as well as in
easier and yields better product quality (see also fluidized beds. Catalysts for vapor-phase hydro-
→ 2-Ethylhexanol, Chap. 3.). genation vary with the producer and include sul-
Sorbitol [50-70-4] is produced on a large fided nickel chromates, nickel – aluminum, sil-
scale by hydrogenation of d-glucose over a ver – manganese oxide, and copper – silica. The
nickel or ruthenium catalyst. Sorbitol is used in reaction temperature is high (300 – 475 ◦ C), and
a variety of foods, pharmaceuticals, and cosmet- the pressure is 0.5 – 1.5 MPa [14]. Aniline is
ics, as well as in the production of polyurethanes, also produced in continuous slurry reactors by
resins, and vitamin C. 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
This hydrogenation is not without complica- on carbon, inhibited by addition of dimethyl
tions. The temperature must be high enough to sulfoxide to prevent overhydrogenation [15].
give acceptable rates, but not so high as to pro- Phenylhydroxylamines are key intermedi-
duce various byproducts resulting mainly from ates in several important processes. In aqueous
cyclization. The process must be operated within acid solution, phenylhydroxylamine rearranges
a narrow pH range. Nickel catalysts are damaged to yield 4-aminophenol. If the hydrogenation
if the solution becomes too acidic. If the solu- of nitrobenzene is carried out in sulfuric acid,
tion becomes too alkaline, gluconic acid is pro- the intermediate phenylhydroxylamine can be
duced by the Cannizzaro reaction, and mannitol made to rearrange as it forms. This process has
is formed through epimerization at the α-carbon become the preferred route to 4-aminophenol
[13]. [123-30-8] (see also → Aminophenols). The hy-
Aromatic carbonyl compounds are reduced droxylamine is susceptible to further hydrogen-
easily, and either reduction to the carbinol or ation, a reaction whose rate must be made min-
loss of the hydroxyl group can be achieved at imal relative to the rate of rearrangement.
will.
catalysts at a temperature lower than that nor- toward the reactor exit; redistribution of potas-
mally used in ester reduction (→ Fatty Alcohols, sium occurs within the catalyst particles as well.
Chap. 2.3.3.2., → Fatty Alcohols, Chap. 3.). Styrene formation is endothermic, and as a con-
sequence, the center of a catalyst pellet in use
becomes cooler than the edges; this temperature
2. Dehydrogenation gradient causes inward migration of potassium,
resulting in catalyst deactivation near both the
Dehydrogenation is an important industrial re- center and the edges due to excess and deficient
action, although the numbers of applications are potassium, respectively.
few compared to hydrogenation. The lesser use Superheated steam plays an important role in
of dehydrogenation stems from several aspects styrene production and serves several purposes:
of its chemistry. it is introduced directly into the reactor in large
Dehydrogenation is an endothermic reaction, molar excess to provide part of the necessary
and the high temperature necessary to obtain a heat to drive the endothermic reaction; it serves
useful conversion to the dehydrogenated prod- as an oxidant to maintain the iron oxide cata-
uct restricts the reaction to those substrates and lyst in the appropriate oxidation state and re-
products with suitable thermal stability. In gen- moves carbonaceous deposits from the catalyst
eral, industrial dehydrogenation falls into one through the water – gas shift reaction; it reduces
of the following categories: (1) the acceptable the partial pressure of hydrocarbons, thereby in-
product is a complex mixture, as in the mix- creasing conversion to styrene. Optimum con-
ture of olefins obtained from alkane dehydro- version is determined by the balance between the
genation, used for detergent alkylate; (2) an un- advantage of increased conversion and the cost
saturated system is extended, as in the conver- of increased steam. Over the last two decades,
sion of benzene to biphenyl or ethylbenzene to improved processing has permitted a continual
styrene; (3) aromatization occurs, a reaction that decrease in the molar ratio of steam to hydro-
provides a driving force as well as a convenient carbon. The ratio is between 6 and 12, with
stopping place; (4) a point of attack exists, as in 65 – 70 % conversion and 90 – 95 % selectivity,
the conversion of methanol or ethanol to form- at 600 – 650 ◦ C and 51 – 101 kPa. Some reactors
aldehyde or acetaldehyde, respectively; (5) the operate in an isothermal mode, heat being sup-
choice of product is limited, as in the conversion plied by a heat-transfer medium on the outside
of butane or butene to butadiene [9]. of the reactor tubes, which offers the advan-
tage of reduced steam requirements; however,
this advantage is offset by higher capital costs.
2.1. Styrene Most processes operate in the adiabatic mode.
In practice, conversion is limited by the equi-
Styrene [100-42-5], an important monomer, is librium between styrene, ethylbenzene, and hy-
obtained largely by dehydrogenation of ethyl- drogen. As styrene approaches equilibrium con-
benzene in the presence of superheated steam centration, its rate of formation decreases but
(→ Styrene). 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
The catalyst for this reaction usually contains to combine with liberated hydrogen and remove
some major active ingredient, a stabilizer, and it from the equilibrium). Despite much effort, no
an alkali-metal promoter. Common commercial large-scale plant operates in this mode.
catalysts contain iron oxide promoted by potas- The technology of ethylbenzene dehydrogen-
sium and a structural stabilizer. Potassium is a ation has been extended to produce both α-
very effective promoter and increases the reac- methylstyrene and divinylbenzene by dehydro-
tion rate by an order of magnitude. Potassium genation of cumene and diethylbenzene, respec-
is volatile under reaction conditions and moves tively.
10 Hydrogenation and Dehydrogenation
supported on alumina (150 – 300 m2 /g), as well escapes while other volatile products are con-
as a second metal such as rhenium. Polymetal- densed and returned, with or without removal
lic catalysts permit operation under more severe of the product. If the product is removed, the
conditions; they also possess improved stability operation can be made continuous.
and activity. All the considerations involving the influence
Typical operating conditions are 430 – of mass transport on the rate of reaction (see Sec-
530 ◦ C and 1.0 – 3.5 MPa. A unit consists of tion 1.1) apply in reverse to dehydrogenation.
several, frequently four, fixed-bed reactors in Good stirring helps to remove liberated hydro-
series. Heated naphtha enters the first reactor, gen from the vicinity of the catalyst, and reflux
where the major reaction is endothermic dehy- facilitates hydrogen removal from the liquid. If
drogenation of cycloalkanes accompanied by a the substrate or product is thermally unstable at
decrease in temperature. The effluent is reheated normal reflux temperature, the temperature can
and the process continued for all reactors. The be lowered by reducing the pressure. Liquid-
liquid product contains 60 – 70 wt % aromatics; phase dehydrogenation is used in the production
the gaseous product contains light hydrocar- of a number of fine chemicals of limited thermal
bons admixed with 60 – 90 mol % hydrogen. The stability.
gaseous portion is recycled to the first reactor in Dehydrogenation may be facilitated by the
a ratio of 5 – 10 mol of recycle gas per mole of use of a hydrogen acceptor to remove hydrogen
naphtha to maintain a high partial pressure of hy- from the system as it is formed [23].
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 The hydrogen transfer reaction above may be
as possible, commensurate with good catalyst written generally as follows:
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 These hydrogenation – dehydrogenation reac-
formed by dehydrogenation of alkanes under re- tions are useful for carrying out hydrogenations
forming conditions, because of an unfavorable in places where hydrogen is unavailable or its
equilibrium. Nonetheless, dehydrogenation is of use is not permitted; only simple equipment is
considerable importance because olefins appear required [24]. Hydrogen transfer reactions often
to be intermediates in the overall process [22]. give better yields of hydrogenated products than
are obtained by direct hydrogenation.
2.4. Liquid-Phase Dehydrogenation
The constraints imposed on dehydrogenation by 3. References
equilibrium considerations can be loosened if
hydrogen is removed from the products as they 1. R. L. Augustine: Catalytic Hydrogenation,
Marcel Dekker, New York 1985.
are formed. Dehydrogenation in the liquid phase
2. M. Freifelder: Practical Catalytic
in an open vessel fulfills this goal. Hydrogen es-
Hydrogenation, Wiley-Interscience, New
capes from the system, whereas products and
York 1971.
reactants remain in the vessel or are returned af- 3. P. N. Rylander: Catalytic Hydrogenation over
ter condensation. The lower temperature limit is Platinum Metals, Academic Press, New York
then set not by the equilibrium but by a conve- 1967.
nient reaction rate. 4. P. N. Rylander: Catalytic Hydrogenation in
In practice, the substrate, neat or dissolved in Organic Syntheses, Academic Press, New
an inert solvent of convenient boiling point, is York 1979.
refluxed in the presence of catalyst. Hydrogen
12 Hydrogenation and Dehydrogenation
5. P. N. Rylander: Hydrogenation Methods, 15. P. N. Rylander et al., Ann. N.Y. Acad. Sci. 172
Academic Press, London 1985. (1970) 266.
6. F. Zymalkowski: Katalytische Hydrierungen, 16. C. C. Coe, J. W. Brockington in P. N. Rylander,
Enke Verlag, Stuttgart 1965. H. Greenfield (eds.): Catalysis of Organic
7. G. W. Roberts in P. N. Rylander, H. Greenfield Reactions, Marcel Dekker, New York 1987.
(eds.): Catalysis in Organic Syntheses, 17. G. M. Cor, M. VandeMoesdijk in J. R. Kosak
Academic Press, New York 1976, p. 1. (ed.): Catalysis of Organic Reactions, Marcel
8. P. A. Chaloner: Handbook of Coordination Dekker, New York 1984.
Catalysis in Organic Chemistry, Butterworths, 18. F. S. Dovell, H. Greenfield, J. Am. Chem. Soc.
London 1986. 87 (1965) 2767.
9. P. N. Rylander, Organic Syntheses with Noble 19. C. L. Thomas: Catalytic Processes and Proven
Metal Catalysts, Academic Press, New York Catalysts, Academic Press, New York 1970.
1973. 20. E. H. Lee, Catal. Rev.-Sci. Eng. 8 (1973) 284.
10. C. N. Satterfield, Am. Inst. Chem. Eng. 21 21. K. K. Kearby in P. H. Emmett (ed.): Catalysis,
(1975) no. 2, 209. vol. 3, Reinhold Publishing Corp., New York
11. J. B. Montgomery et al., Ind. Eng. Chem. 50
1955.
(1958) 313.
22. J. H. Sinfelt in J. R. Anderson, M. Boudart
12. J. F. VanPeppen et al. in R. L. Augustine (ed.)
(eds.): Catalysis Science and Technology,
Catalysis of Organic Reactions, Marcel
vol. 1, Springer-Verlag, New York 1981.
Dekker, New York 1985.
13. L. W. Wright, Chemtech. 1974, 42. 23. J. F. Marschik, P. N. Rylander, US 3 517 021,
14. A. M. Stratz in J. R. Kosak (ed.): Catalysis of 1970.
Organic Reactions, Marcel Dekker, New York 24. G. Brieger, T. J. Nestrick, Chem. Rev. 74
1984. (1974) 567.
Hydroquinone
Phillip M. Hudnall, Tennessee Eastman Company, Kingsport, Tennessee 37662, United States
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.
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 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
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
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
man exposure are well documented (see below). 9. Celanese, EP-A 178 929, 1986; US-A 661 552,
Direct contact of hydroquinone dust and quinone 1984 (H. R. Gerberich).
vapor with the eye may cause irritation. Chronic 10. Agency of Industrial Sciences & Technology,
exposure at high levels has caused brownish dis- JP-Kokai 8 678 394, 1986 (A. Yoshikawa).
coloration of the cornea and conjunctiva, as well 11. J. E. Lyons, C-Y. Hsu, Biological and Inorganic
as distortion of the cornea, leading in some cases Copper Chemistry, in K. D. Karlin, J. Zubieta
to decreased visual acuity and even blindness. (eds.): Proc. Conf. Copper Coord. Chem. (July
Exposure to levels below the TLV of 2 mg/m3 23 – 27, 1984), vol. 2, Adenine Press,
Guilderland, New York 1986, pp. 57 – 76.
has not produced these effects [50].
12. J. Van Gent, A. A. Wismeijer, A. W. P. G.
A low incidence of allergic contact dermatitis Peters-Rit, H. Van Bekkum, Tetrahedron Lett.
has been reported from handling large amounts 27 (1986) 1059 – 1062.
of hydroquinone, where exposure to dust oc- 13. Mitsui Petrochemicals Industries, JP-Kokai
curred, and from repeated contact with alkaline 859 889, 1985.
photographic developer solutions [23]. Skin sen- 14. C. H. Kuo, H. S. Soong, Chem. Eng. J.
sitization was noted in 2 members of a group of (Lausanne) 28 (1984) 163 – 171.
48 middle-aged males who used a depigmen- 15. Upjohn, DE 2 214 971, 1972, and
tation ointment containing 5 % hydroquinone; GB 1 344 602, 1974 (P. S. Charleton).
38 subjects treated with a 2 % ointment and 26 16. Ajinomoto, US 3 420 895, 1969 (H.
subjects treated with a 3 % ointment showed no Wakamatsu).
sensitization [51], [52]. 17. Lonza, GB 1 215 568, 1970 (P. Pino).
In reported poisonings, the lethal dosage 18. Koppers, US 3 953 509, 1974 (N. P. Greco).
for adults is 5 – 12 g (70 – 170 mg/kg). Lower 19. Mitsui Toatsu, JP-Kokai 77 102 235, 1976 (F.
doses cause no adverse effects: two men in- Matsuda).
20. Upjohn, US 3 676 503, 1972 (W. v. E. Doering,
gested 500 mg/d for five months and 17 men
W. J. Farrisey, Jr.).
and women ingested 300 mg/d for three to five 21. M. Ya. Fioshin, Elektrokhimiya 13 (1977) 381.
months without abnormal symptoms or any ab- 22. M. Fremery, H. Hoever, G. Schwarzlose,
normalities in blood and urine [43]. Chem. Ing. Technol. 46 (1974) 635.
23. Unpublished results, Health & Environment
Laboratories, Eastman Kodak, Rochester, New
9. References York.
24. K. Verschueren: Handbook of Environmental
1. Beilstein, VI3 , 4374. Data on Organic Chemicals, 2nd ed., Van
2. Sadtler Standard Spectra, Sadtler Research Nostrand Reinhold, New York 1983,
Laboratories, Division of Bio-Rad pp. 746 – 748.
Laboratories, a. Infrared Prism Spectrum 25. Environmental Effects of Photoprocessing
nos. 153, 13206; b. Infrared Grating Spectrum Chemicals, vols. 1 and 2, National
no. 47; c. Carbon-13 NMR Spectrum no. 38; d. Association of Photographic Manufacturers,
Proton NMR Spectrum no. 10350; e. and Hydroscience, Harrison, New York 1974.
Ultraviolet Spectrum no. 60. 26. G. M. De Graeve, D. L. Geiger, J. S. Meyer,
3. C. J. Pouchert (ed.), The Aldrich Library of H. L. Bergman, Arch. Environ. Contam.
FTIR Spectra, Aldrich Chemical Company, Toxicol. 9 (1980) 557.
Milwaukee, Wisconsin 1984, vol. 1, 27. J. E. McKee, H. W. Wolf (eds.): Water Quality
no. 1107D. Criteria, State of California, 1963,
4. A. Seidell: Solubilities of Organic Publication no. 3-A.
Compounds, 3rd ed., vol. 2, Van Nostrand 28. I. Juhnke, D. Luedemann, Z. Wasser Abwasser
Publishing, New York 1941, p. 396. Forsch. 11 (1978) 161.
5. H. Musso, H. Doepp, Chem. Ber. 100 (1967) 29. A. J. Leo, C. Hansch (eds.): Chemical
3267. Parameter Data Base, Medicinal Chemistry
6. P. J. Elving, A. F. Krivis, Anal. Chem. 30 Project, Pomona College, Seaver Chemistry
(1958) 1645. Laboratory, Claremont, Calif. 1985.
7. FMC, EP-A 132 783, 1985 (R. A. Bull). 30. The United States Pharmacopeia XXI, The
8. Mobil Oil, US 4 578 521, 1986 (C. D. Chang, United States Pharmacopeial Convention,
S. D. Hellring). Rockville, Maryland, 1985.
8 Hydroquinone
31. Hydroquinone, Technical Grade, Sales 45. IARC Monographs on the Evaluation of the
Specification no. 3503-9, Tennessee Eastman Carcinogenic Risk of Chemicals to Humans,
Company, Kingsport, Tenn. vol. 15, International Agency for Research on
32. Hydroquinone, Photographic Grade, Sales Cancer (IARC), Lyon, France 1977,
Specification no. 4305-1, Tennessee Eastman pp. 155 – 175.
Company, Kingsport, Tenn. 46. I. Florin, L. Rutberg, M. Curvall, C. R. Enzell,
33. Ishihara Sangyo Kaisha, DE 2 812 571, 1979 Toxicology 15 (1980) 219.
(R. Nishiyama et al.). 47. E. Gocke, M.-T. King, K. Eckhardt, D. Wild,
34. Dow Chemical Co., DD 217 694, 1985 (H. Mutat. Res. 90 (1981) 91.
Johnson, L. H. Troxell). 48. K. Morimoto, S. Wolff, A. Koizumi, Mutat.
35. Dow Chemical Co., EP-A 97 460, 1984 (H. Res. 119 (1983) 355.
Johnson, L. H. Toxell). 49. G. D. Divincenzo, M. L. Hamilton, R. C.
36. Bayer AG, DE 3 219 789, 1983 (H. Foerster et
Reynolds, D. A. Ziegler, Toxicology 33 (1984)
al.).
9.
37. Zoecon Corp., US 4 529 438, 1985 (S. Lee).
38. Farbenfabriken Bayer Aktiengesellschaft and 50. F. L. Oglesby, R. L. Raleigh: Eye Injury
Mobay Chemical Company, US 3 016 364, Associated with Exposure to Hydroquinone
1962 (E. Muller). and Quinone, 2nd ed., Laboratory of
39. S. Romaine, I. J. Cotton, Proceedings Industrial Medicine, Tennessee Eastman Co.,
American Power Conference 48 (1986) 1066. Kingsport, Tenn. 1973, p. 2.
40. D. G. Baird, Fiber Producer (October1984) 51. NIOSH Criteria for a Recommended Standard
66. for Hydroquinone, National Institute for
41. A. S. Wood, Modern Plastics (April1985) 78. Occupational Safety & Health (NIOSH), U.S.
42. B. F. J. Goodwin, R. W. R. Crevel, A. W. Government Printing Office, Wahington, D.C.,
Johnson, Contact Dermatitis 7 (1981) 248. 1978.
43. A. J. Carlson, N. R. Brewer, Proc. Soc. Exp. 52. M. C. Spencer, J. Am. Med. Assoc. 194 (1965)
Biol. Med. 84 (1953) 684. 962.
44. B. L. van Duuren, B. M. Goldschmidt, J. Nat.
Cancer Inst. 56 (1976) 1237.
Most naturally occurring hydroxycarboxylic Hydroxycarboxylic acids readily form the ex-
acids are chiral and thus exhibit optical activ- pected salts at the carboxyl group. The acid-
ity. In pure form, they are usually crystalliz- ity of α- and β-hydroxycarboxylic acids is en-
able. Hydroxycarboxylic acids have exception- hanced relative to the corresponding unsubsti-
ally high boiling points relative to corresponding tuted acids because of the proximity of the po-
unsubstituted acids with the same chain length. lar hydroxyl group. Compounds in this category
This is a consequence of association among the may be esterified at either the carboxyl or the hy-
hydroxyl groups. For this reason, many of these droxyl function by reaction with other alcohols
compounds cannot be distilled without decom- or acids, respectively. The actual course of such
position, even under vacuum. reactions depends once again on the placement
Table 2 summarizes the physical properties of of the hydroxyl group relative to the carboxylic
the commercially important hydroxycarboxylic acid group.
acids and their derivatives listed in Table 1. 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 .
5. Specific Aliphatic
Hydroxycarboxylic Acids of
Commercial Significance
5.1. Glycolic Acid (Hydroxyacetic Acid,
2-Hydroxyethanoic Acid) [10]
Numerous branched and unbranched hy- The chemical structure, molecular mass, and
droxycarboxylic acids may be prepared by CAS registry number of glycolic acid are given
cleavage or oxidation (with alkaline perman- in Table 1. Its physical properties are listed in
ganate, chromic acid, dilute nitric acid, or con- Table 2 and in the following material:
centrated nitric acid) of a variety of starting ma-
mp:
terials [6–9]. Enzymatic production by fermen-
α-Modification 78 – 80 ◦ C
tation is becoming increasingly important. β-Modification, metastable 63 ◦ C
Heat of combustion −697.23 kJ/mol
Heat of solution in water
(infinite dilution) −11.71 kJ/mol
4. Analysis Dissociation constant (at 25 ◦ C) 1.54×10−4
pH (1 M solution) 2.4
The literature describes an array of color re-
actions for qualitative determination of hy- Solid glycolic acid forms colorless, mono-
droxycarboxylic acids. These include the yel- clinic, prismatic crystals. The acid is freely sol-
low color obtained in the presence of iron(II) uble in water, methanol, ethanol, acetone, and
Hydroxycarboxylic Acids, Aliphatic 7
ethyl acetate. Glycolic acid is only slightly steam corrosion, and this characteristic, coupled with
volatile and cannot be destilled under vacuum; its bactericidal properties, makes it suitable for
attempts result in self-esterification with loss of incorporation into acidic cleansing agents. It
water, leading to di- and polyglycolides. How- is especially well-adapted to cleansing opera-
ever, newer results show that glycolic acid can tions involving milk containers, milk-processing
be vaporized without decomposition. In the va- equipment, and drinking fountains, as well as
por phase it can be dehydrogenated catalytically rust and scale removal in heat exchangers and
to glyoxylic acid [11]. pipelines. Glycolic acid inhibits the growth of
iron-oxidizing bacteria. The use of glycolic acid
eliminates the need for simultaneous addition of
5.1.1. Preparation chelating agents and bactericides. The effective-
ness of glycolic acid as a complexing agent also
Glycolic acid is usually produced by hydroly- contributes to its use in copper polishes, as an
sis of molten monochloroacetic acid with 50 % etching agent for lithographic plates, and in the
aqueous sodium hydroxide at 90 – 130 ◦ C. The preparation of electropolishing and galvanizing
resulting glycolic acid solution has a concentra- baths.
tion of ca. 60 % and contains 12 – 14 % sodium Safety Considerations. According to section
chloride. The salt may be removed by evapo- 1.4 of the hazardous materials regulation,
rative concentration, followed by extraction of Gef Stoff V (Federal Republic of Germany), and
the acid with acetone [12]. Attempts have also to the EEC guidelines, glycolic acid solutions are
been made to conduct the hydrolysis with acid “corrosive.”
catalysts at 150 – 200 ◦ C with water or steam un- Analysis. Glycolic acid may be detected
der pressure [13]. In this case, the byproduct is qualitatively by the violet color formed with
hydrogen chloride, rather than sodium chloride, 2,7-dihydroxynaphthalene [17]. The preferred
which can be removed by distillation. The prin- method of quantitative analysis (in the absence
cipal disadvantage of the method is the need for of other acidic or hydrolyzable substances) is
relatively large volumes of water. acidimetric titration. Because of the tendency of
Glycolic acid is produced commercially in lactide formation free and total acid must be de-
the United States (Du Pont) by treating formal- termined separately.
dehyde or trioxymethylene with carbon monox-
ide and water in the presence of acid catalysts at
>30 MPa [14]. Another process, previously uti- 5.1.3. Derivatives
lized by Degussa, involves electrolytic reduction
of oxalic acid [15]. The compound can also be The glycolic acid esters methyl glycolate
prepared in 90 % yield by hydrolysis of the cor- [96-35-5], bp 147 – 149 ◦ C, and ethyl glycolate
responding nitrile, obtained by reaction of form- [623-50-7], bp 158 – 159 ◦ C, both serve as start-
aldehyde with hydrocyanic acid [16]. ing materials for the laboratory preparation of
pure glycolic acid [18] and were formerly em-
ployed as solvents for resins and for nitro- or
5.1.2. Uses acetyl cellulose. Apart from these two materi-
als, only carboxymethyl cellulose [9004-32-4],
Glycolic acid is available commercially as either (→ Cellulose Ethers, Chap. 6. and n-butyl gly-
a 57 % (Hoechst) or a 70 % (Du Pont) aqueous colate have commercial significance.
solution. Total annual consumption worldwide
is ca. 2000 – 3000 t of solution. n-Butyl Glycolate (see Table 1).
Glycolic acid is used in textile dyeing, print- Physical Properties. n-Butyl glycolate is a
ing, and creaseproofing. The fact that it can colorless liquid, which is miscible with most or-
form a chelate with calcium(II) ions makes it ganic solvents. Its physical properties are sum-
well-suited to hide deliming in the leather in- marized in Table 2. The solubility in water is lim-
dustry, as well as to inclusion in alum and ited to 8 wt % (20 ◦ C), although the compound
chrome mordants and in fur-processing opera- itself may contain up to 25 wt % water.
tions. The compound has little tendency to cause
8 Hydroxycarboxylic Acids, Aliphatic
All three acids are soluble in water, etha- [26]. Biodegradable poly(−(R)-(−)-β-hy-
nol, diethyl ether, and numerous organic sol- droxybutyric acids (named PHB) have been
vents, and decompose readily on dry distilla- used in the context of controlled long-term
tion. β-Hydroxybutyric acid easily loses water parenteral application of drugs through im-
to give crotonic acid, and γ-hydroxybutyric acid plantation. These materials offer a number
is transformed even at low temperature to γ- of advantages relative to other implantation
butyrolactone. polymers [e.g., poly(methacrylic acid) deriva-
tives, poly(glycolic acid), poly(lactic acid)] [27].
(R,S-)-α-Hydroxybutyric Acid [(R,S- Poly(β-hydroxybutyric acid) has also been con-
)-2-Hydroxybutanoic Acid] [23], (see Ta- sidered as a potential textile material.
ble 1). The physical properties of (R,S)-α-hy- γ-Hydroxybutyric acid has been used as an
droxybutyric acid are listed in Table 2. anesthetic and pain killer. It has a stabilizing ef-
Preparation. (R,S)-α-Hydroxybutyric acid fect on blood pressure and induces a state of un-
results as main product from treating butyr- consciousness, characteristics that have encour-
aldehyde with alkaline sodium hypochlorite aged its application especially in geriatrics [28].
solution or from heating α-bromobutyric acid
with formamide at 150 ◦ C.
5.4. (R,S)-Malic Acid
(R,S-)-β-Hydroxybutyric Acid [(R,S-)-3- [(R,S)-Hydroxysuccinic Acid,
Hydroxybutanoic Acid] [24], (see Table 1).
(R,S)-Hydroxybutanedioic Acid] [29],
Physical Properties. The mp of racemic β-hy-
droxybutyric acid is 48 – 50 ◦ C; it is commonly [30](see Table 1.)
described in the literature as a syrupy liquid (see
Physical Properties (see Table 2). Malic
Table 2). Pure (R)-(−)-β-hydroxybutyric acid
acid is a colorless, odorless, crystalline sub-
[625-72-9] is a solid with mp 49 – 50 ◦ C.
stance. It is highly soluble in water (100 g
Preparation. (R,S)-β-Hydroxybutyric acid
of water dissolves 126.3 g of malic acid at
results from hydrolysis of β-butyrolactone,
20 ◦ C), methanol, ethanol, acetone, ether, and
which in turn can be produced by the addi-
other polar solvents. Because it is a dibasic
tion of ketene to acetaldehyde in the presence of
acid, it has two dissociation constants: at 20 ◦ C,
boron trifluoride or zinc salts [5]. The compound
K A1 = 3.9×10−4 and K A2 = 1.4×10−5 .
is also prepared by reduction of acetoacetic acid
with sodium amalgam in alkaline solution. Preparation. (R,S)-Malic acid is prepared
commercially in the United States and Canada
γ-Hydroxybutyric Acid (4-Hydroxybutanoic
by hydration of maleic anhydride (→ Maleic and
Acid) [25] (see Table 1).
Fumaric Acids) [31]. The sole manufacturer in
Physical Properties. The most important
the United States is Alberta Gas, with an annual
physical properties of γ-hydroxybutyric acid
capacity of ca. 5000 t. In this process maleic acid
are listed in Table 2. The dissociation constant
is heated at ca. 180 ◦ C (under a pressure of ca.
at 25 ◦ C is 1.93×10−5 . At ambient temperature,
1 MPa), malic acid is yielded as the main prod-
the acid is a liquid and consists in part of lactone.
uct. Byproducts are maleic and fumaric acids.
Preparation. γ-Hydroxybutyric acid is ob-
The latter can be separated by filtration and re-
tained from γ-butyrolactone by alkaline hydro-
turned to the process stream because of its low
lysis and isolated as the alkali-metal salt. γ-
water solubility.
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 The filtrate is then concentrated; this causes sep-
materials or intermediates in organic synthesis. aration of the malic acid, which is purified by
Optically active β-hydroxybutyric acid multiple washings, evaporation, and recrystal-
is an important pharmacological agent lization until the contents of fumaric and maleic
10 Hydroxycarboxylic Acids, Aliphatic
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
protected (e.g., by esterification with sulfuric prisms (from alcohol). Important physical prop-
acid). The alkyl moiety of the sulfuric acid es- erties follow: mp 203 ◦ C; d 25 1.473; dissocia-
ter is then oxidized easily by using potassium tion constants K1 8.71×10−5 , K 2 1.18×10−10
permanganate or chromic acid [14]. Aromatic (at 19 ◦ C); solubility (in 100 g of solution): wa-
hydroxycarboxylic acids can also be produced ter 6.11 g (69 ◦ C), 98 % ethanol 39.6 g (65 ◦ C),
by oxidative alkali melt of appropriately substi- and n-butanol 20.7 g (36.5 ◦ C).
tuted phenols. Alkyl groups ortho to the phenolic
hydroxyl group are oxidized most easily [15]. Chemical Properties. 3-Hydroxybenzoic
The oxidative alkali melt of aromatic hydroxy- acid does not undergo decarboxylation at ele-
aldehydes gives the corresponding hydroxycar- vated temperature and is stable up to 300 ◦ C.
boxylic acids in good yield [16]. In the presence This is in contrast to 2- and 4-hydroxybenzoic
of silver catalysts, oxidation is possible even in acids. Electrophilic substitution (e.g., nitration,
aqueous alkali solution [17]. Hydroxyacetophe- halogenation, sulfonation) occurs ortho or para
nones can be converted into hydroxycarboxylic to the hydroxyl group. Thus, nitration with 62 %
acids by using iodine in pyridine (haloform re- aqueous nitric acid gives 2-nitro-3-hydroxyben-
action), followed by alkaline hydrolysis of the zoic acid as the main product, along with 4-nitro-
intermediate pyridinium iodide [18]. and 6-nitro-3-hydroxybenzoic acids [24].
Direct introduction of a hydroxyl group into
the aromatic nucleus by oxidation is possible in Production. 3-Hydroxybenzoic acid can be
the case of ammonium benzoates by using hy- obtained by alkali melt of the sodium salt of
drogen peroxide [19]. Thus, the first stage of the 3-sulfobenzoic acid at 210 – 220 ◦ C. The crude
Dow process (production of phenols from ben- acid precipitates on acidification. It can be puri-
zoic acids) involves hydroxylation by air of the fied by recrystallization from water with the ad-
metal benzoate ortho to the carboxyl group [20]. dition of activated carbon [11]. 3-Hydroxyben-
Introduction of a carboxyl group is achieved zoic acid can also be obtained from 3-nitroben-
by reaction of phenols with tetrachloromethane zoic acid esters. This occurs via catalytic hydro-
in the presence of alkali and powdered copper genation followed by diazotization of the ami-
or copper salts [21]. Alkyl ethers of hydroxy- nobenzoic acid ester formed. The diazonium salt
carboxylic acids can be obtained directly by car- is then heated [25] with aqueous sulfuric acid.
boxylation of alkoxyphenyl metal halides (e.g., Another production method is oxidation of 3-
magnesium or lithium compounds) with carbon hydroxybenzaldehyde [100-83-4] by bubbling
dioxide. Amides and thioamides of alkoxycar- air through a hot aqueous sodium hydroxide sus-
boxylic acids are formed by conversion of phe- pension [26].
nol ethers with carbamoyl chlorides, isocya-
nates, or isothiocyanates in the presence of alu- Uses. 3-Hydroxybenzoic acid is used as
minum chloride [22]. Rearrangement of arylur- an intermediate in the manufacture of pharma-
ethanes in the presence of Lewis acids (e.g., alu- ceuticals and pesticides. The 3-hydroxybenzo-
minum chloride) may give good yields of the ic acid ester of tropine is used in ophthalmol-
amides of the corresponding hydroxycarboxylic ogy to dilate the pupils. The sodium salt of 3-
acids [23]. hydroxybenzoic acid is a cholepoietic agent. Es-
ters and metal salts have germicidal and preser-
vative properties, and are used in food preser-
4. Individual Aromatic vation. Ether – esters are suitable as plasticizers
Hydroxycarboxylic Acids for vinyl and cellulose resins.
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- of alkali produces salicylic acid [69-72-7] as
marized here: mp 216.3 ◦ C; d 20 1.497; dissoci- the main product and smaller amounts of 4-
ation constants K 1 3.3×10−5 , K 2 4.8×10−10 hydroxybenzoic acid [99-96-7] [21].
(19 ◦ C); solubility (in 100 g of solution): wa-
ter 0.49 g (20 ◦ C), 33.5 g (100 ◦ C); 99 % ethanol Uses. 4-Hydroxybenzoic acid is an interme-
38.75 g (67 ◦ C); and n-butanol 19.5 g (32.5 ◦ C). diate in the manufacture of dyes, pharmaceuti-
4-Hydroxybenzoic acid is sparingly soluble in cals, pesticides [30], and plastics. It is used as
chloroform. an emulsifier and also as a corrosion-protection
Table 2 lists properties of the most commonly agent [31]. Its esters and corresponding sodium
used esters. salts are antimicrobial substances with a high
no-effect level (1200 mg per kilogram of body
Chemical Properties. Above 200 ◦ C, 4- weight) [32]. They are, therefore, used under
hydroxybenzoic acid undergoes decarbox- weakly acidic to weakly basic conditions as food
ylation to produce phenol and carbon diox- preservatives (e.g., in fruit juice), cosmetics, and
ide. Electrophilic substitution (e.g., nitration, pharmaceuticals, as well as in technical products
sulfonation, or halogenation) occurs ortho to (e.g., in lubricants, anti-freezing agents) [33].
the hydroxyl group. With iron(III) chloride, no Benzyl esters are relatively soluble in prepa-
color reaction occurs but a yellow precipitate is rations containing sugar. Table 2 lists observed
formed. bacteriostatic effects [34] compared to phenol.
4-Hydroxybenzoic acid is also used in poly-
Production. 4-Hydroxybenzoic acid is pro-
mers as a component in the manufacture of
duced by carboxylation of potassium phenolate
polyester and as a constituent of liquid crystal
above 220 ◦ C and with a carbon dioxide pres-
polymers [35] (→ Liquid Crystals).
sure of 0.45 MPa. The yield of 4-hydroxyben-
Esters. The food preservative “ PHB-Ester”
zoic acid (partly present as the dipotassium salt)
is a mixture of the ethyl and n-propyl es-
is 80 % of the theoretical amount in which the
ters of 4-hydroxybenzoic acid. According to
conversion of phenolate is 60 %. 4-Hydroxyben-
the preservatives regulations in the different
zoic acid can also be obtained by isomeriza-
countries [EEC guideline 64/52 date: 5. 11. 63,
tion of dipotassium salicylate [27] at temper-
24th amendment 85/585 date: 20. 12. 85; United
atures up to 240 ◦ C and under carbon diox-
States: Food and Drug Administration, HHS,
ide pressure (slightly above atmospheric pres-
21 CFR Ch.I, 4. 1. 86, § 172.145, § 184.1490,
sure) [28], or by oxidation of the alkali-metal §
184.1670; Federal Republic of Germany:
salt of p-cresol [106-44-5] on metal oxides
Verordnung über die Zulassung von Zusatzstof-
at 260 – 270 ◦ C [29]. Reaction of phenol with
fen zu Lebensmitteln, date: 22. 12. 81 (BGBl.
tetrachloromethane [56-23-5] in the presence
Hydroxycarboxylic Acids, Aromatic 5
is also obtained by using the enzyme tannase 58 ◦ C, is used as a spasmolytic and to stimulate
[9025-71-2] or molds (Penicillium glaucum, As- blood circulation. Mandelic acid tropine ester
pergillus niger) to cleave tannin by fermenta- [51-56-9] (homatropine – HBr), C16 H22 BrNO3 ,
tion. Both the metabolism of gallic acid and its M r 356.26, mp 212 – 215 ◦ C, is readily soluble in
impact on plant growth enzymes have been stud- water and ethanol, and is used in ophthalmology
ied [52]. to dilate the pupils. Hexamethylenetetramine
Uses. Gallic acid is used for the manu- mandelate [587-23-5] ( mandelamine, Man-
facture of iron gallnut ink (see above) and dropine, diuramine, hexydaline) C14 H20 N4 O3 ,
as an intermediate in the production of dyes M r 292.34, mp 127 – 130 ◦ C, forms colorless
(e.g., anthragallol [620-64-2], gallocyanine crystals that are readily soluble in diethyl ether.
[1562-85-2], galloflavin [568-80-9], and rufigal- This salt is used to treat urinary tract infections.
lic acid [82-12-2]). Gallate esters, particularly
methyl gallate [99-24-1] ( gallicin), mp 157 ◦ C, Levorotatory (R)(−)-Mandelic Acid
and propyl gallate [121-79-9], mp 150 ◦ C, are [611-71-2]. (R)-Mandelic acid forms color-
used as antioxidants and food preservatives (e.g., less crystals, mp 132 – 135 ◦ C, specific rotation
for fats) [53]. Gallic acid is also employed as [α]20 ◦
D (c = 5) −154 to 157 C. It is produced
a reducing agent for pharmaceutical purposes ( by resolution of racemic (R,S)-mandelic acid
Dermatol, Airol, bismuth salt of gallic acid) [54] with cinchonine [118-10-5] in aqueous solu-
and for the production of photographic develop- tion, whereby the salt of (S)-(+)-mandelic acid
ers and pyrogallol. crystallizes. If resolution is performed with S-
(+)-aminodiol [3306-06-7], (R)-(−)-mandelic
acid crystallizes, while (S)-(+)-mandelic acid
4.4. Mandelic Acid [17199-29-0] remains in solution. (R)-(−)- and
(S)-(+)-mandelic acid can again be subjected to
Racemic (R,S)-Mandelic Acid [611-72-3]. racemic resolution after isomerization.
Physical Properties. (R,S)- Mandelic acid
(α-hydroxyphenylacetic acid, phenylglycolic
acid), C6 H5 −CH(OH)−COOH, C8 H8 O3 , M r 5. Toxicology
152.15, mp 118 – 121 ◦ C, is a colorless crys-
talline powder. It is readily soluble in ethanol Toxicological and physiological properties of
and diethyl ether, less soluble in chloroform, and the different aromatic hydroxycarboxylic acids
insoluble in petroleum ether. Its water solubility are summarized in the following material.
is 15 g per 100 mL of water. In alkaline solution, 3-Hydroxybenzoic Acid. LD50 3700 mg/kg
mandelic acid dissolves to form salts. (rat, i.p.), TDLo 400 mg/kg (rat, s.c., 11 d preg-
Chemical Properties. Mandelic acid turns nant); classification: drug, teratogen [55, ; date:
brown when exposed to light. It can be esterified 1/84].
easily with alcohols in the presence of hydrogen 4-Hydroxybenzoic Acid. LD50 2200 mg/kg
chloride. (mouse, oral), LD50 210 mg/kg (mouse, i.p.),
Production. (R,S)-mandelic acid is pro- LD50 1050 mg/kg (mouse, s.c.); classification:
duced by hydrolysis of mandelic acid nitrile drug, mutagen [55, ; date: 8/83].
[532-28-5] (mp 10 ◦ C) with hydrochloric acid. o-Cresotic Acid. LD50 1000 mg/kg (mouse,
Mandelic acid nitrile is obtained by the conver- oral). LD50 345 mg/kg (mouse, i.v.); classifica-
sion of benzaldehyde [100-52-7] and hydrogen tion: drug [55, ; date: 6/85].
cyanide in the nascent state (NaCN + HCl) un- m-Cresotic Acid. LD50 1800 mg/kg (mouse,
der refrigeration. Another method of produc- oral); classification: drug [55, ; date: 6/85].
ing mandelic acid is the hydrolysis of α,α-di- p-Cresotic Acid. LD50 1000 mg/kg (mouse,
chloroacetophenone with alkali. oral); classification: drug [55, ; date: 6/85].
Uses. Mandelic acid esters have analgesic, 4-Hydroxy-3-methoxybenzoic Acid. Classi-
antirheumatic, and spasmolytic effects. For ex- fication: mutagen [55, ; date: 8/82].
amples, cis-3,5,5-trimethylcyclohexyl mande- 2,4-Dihydroxybenzoic Acid. LDLo 642 mg/kg
late [456-59-7] ( cyclandelate, Spasmocyclon, (rat, s.c., 11 d pregnant); classification: teratogen
Clan-dilon), C17 H24 O3 , M r 276.38, mp 54 – [55, ; date: 3/84].
8 Hydroxycarboxylic Acids, Aromatic
3,5-Dihydroxybenzoic Acid. LD50 2000 mg/kg 8. Rhône-Poulenc, DE-OS 27 54 239, 1976 (J. C.
(mouse, i.v.) [55, ; date: 6/84]. Choulet, J. Nouvel).
2,5-Dihydroxybenzoic Acid. 2,5 Dihydroxyben- 9. F. Wessely, Monatsh. Chem. 81 (1950)
zoic acid is an intermediate in the metabolism of 1071 – 1091.
salicyclic acid [56]: LD50 800 mg/kg (rat, oral) 10. S. E. Hunt, J. I. Jones, A. S. Lindsey, Chem.
LD50 3000 mg/kg (rat, i.p.), LD50 374 mg/kg Ind. (London) 1955, 417 – 418.
(mouse, i.v.), LDLo 380 mg/kg (rat, s.c., 9 d 11. M. F. Clarke, L. N. Owen, J. Chem. Soc. 1950,
2108 – 2115.
pregnant), LDLo 642 mg/kg (rat, s.c., 11 d pre-
12. K. W. Rosenmund, H. Harms, Ber. Dtsch.
gant); classification: teratogen, natural product
Chem. Ges. 53 (1920) 2226 – 2240.
[55, ; date: 5/84]. 13. F. F. Blicke, F. D. Smith, J. L. Powers, J. Am.
3,4,5-Trihydroxybenzoic Acid. LD50 Chem. Soc. 54 (1932) 1460 – 1471.
320 mg/kg (mouse, i.v.), LD50 5000 mg/kg (rab- 14. B. Heymann, W. König, Ber. Dtsch. Chem.
bit, oral), LDLo 800 mg/kg (mouse, i.p.), LDLo Ges. 19 (1886) 704 – 706.
5 mg/kg (mouse, s.c., 1 d pregnant); classifica- 15. O. Jakobson, Ber. Dtsch. Chem. Ges. 11
tion: agricultural chemical, mutagen, teratogen (1878) 376, 570.
[55, ; date: 3/85]. 16. G. Lock, Ber. Dtsch. Chem. Ges. 62 (1929)
1177 – 1188.
17. I. A. Pearl, J. Am. Chem. Soc. 67 (1945)
6. References 1628 – 1629. I. A. Pearl, J. Am. Chem. Soc. 68
(1946) 429 – 432. I. A. Pearl, J. Org. Chem. 12
General References (1947) 79 – 84.
1. Beilstein 10, 43, 134, 149, 192, 214, 375, 464, 18. L. C. King, M. McWhirter, D. M. Barton, J.
10 (1), 20, 63, 68, 83, 95, 173, 232, 10 (2), 25, Am. Chem. Soc. 67 (1945) 2089 – 2095.
79, 88, 112, 248, 331, 10 (3), 87, 422, 1363, 19. H. D. Dakin, M. D. Herter, J. Biol. Chem. 3
2057, 10 (4), 125, 315, 345, 1414, 1971. (1907) 419 – 434.
2. E. Rodd: Chemistry of Carbon-Compounds, 20. G. Navazio, A. Scipioni, Chim. Ind. (Milan)
vol. 3, Elsevier, Amsterdam-London-New 50 (1968) 1086 – 1090.
York 1956, pp. 756 – 815. 21. J. Zelter, M. Landau, DE 258 887, 1912.
3. A. S. Lindey, H. Jeskey, Chem. Rev. 57 (1957) 22. L. Gattermann, Justus Liebigs Ann. Chem.
583 – 620. Houben Weyl 8,372 – 377. 244 (1888) 29 – 76.
23. G. J. Gershzon, J. Gen. Chem. USSR (Engl.
Specific References Transl.) 13 (1943) 82.
4. J. E. Heestermann, Chem. Weekbl. 32 (1935) 24. H. Schäfer, Memorial des Poudres 27 (1937)
463. 153, 156, 157.
5. C. G. Barroso, R. C. Torrijos, J. A. 25. H. E. Ungnade, A. S. Henick, J. Am. Chem.
Perez-Bustamante, Chromatographia 17 Soc. 64 (1942) 1737 – 1738.
(1983) no. 5, 249 – 252. R. Shinohara, 26. I. G. Farben, DE 506 438, 1926 (F. Hans);
Suishitsu Odaku Kenkyu 7 (1984) no. 9, Friedländer 17, 472.
549 – 552; Chem. Abstr. 102 (1985) no. 26, 27. H. Kupferberg, J. Prakt. Chem. 16 (1877) no.
225 717 p. 2, 424 – 447.
6. H. Kolbe, Justus Liebigs Ann. Chem. 113 28. C. A. Buehler, W. E. Cate, Org. Synth. Coll. 2
(1860) 125 –127. H. Kolbe, E. Lautermann, (1943) 341 – 343. Dow Chemical, US
Justus Liebigs Ann. Chem. 115 (1860) 1 937 477, 1932 (L. E. Milk, W. W. Allen).
157 – 206. H. Kolbe, J. Prakt. Chem. 10 S. E. Hunt, J. Chem. Soc. 1958, 3152 – 3160.
(1874) no. 2, 89 – 111. O. Hartmann, J. Prakt. A. J. Rostron, A. M. Spivey, J. Chem. Soc.
Chem. 16 (1877) no. 2, 35 – 59. H. Kolbe, DE 1964, 3092 – 3096.
426, 1877. F. W. v. Heyden, DE 29 939, 1884; 29. P. Friedländer, O. Löw-Beer, DT 170 230,
31 240, 1884(R. Schmitt). F. W. v. Heyden, DE 1905.
33 635, 1885 (R. Schmitt). F. W. v. Heyden, 30. G. W. K. Cavill, J. M. Vincent, J. Chem. Soc.
DE 38 742, 1886 (R. Seifert). S. Marassé, DE Ind. 66 (1947) 175 – 182.
73 279, 1893. 31. BASF Wyandotte, US 4 234 440, 1979 (St. T.
7. D. Cameron, H. Jeskey, O. Baine, J. Org. Hirozawa, E. F. O’Brien, J. C. Wilson).
Chem. 15 (1950) 233 – 236. O. Baine, J. Org. 32. G. M. Cramer et al., Food, Cosmet. Toxicol. 16
Chem. 19 (1954) 510 – 514. (1978) 255.
Hydroxycarboxylic Acids, Aromatic 9
33. A. J. Maga, K. Lorenz, Cereala Sci. Today 18 44. A. Cerami, J. Graziano, R. W. Grady, C. M.
(1973) no. 10, 326 – 328, 350; Chem. Abstr. 84 Peterson, Proc. Symp.: Dev. Iron Chelaters
(1976) no. 23, 163 164 e. H. P. Fiedler: Lexikon Clin. Use 1975, 261 – 273; Chem. Abstr. 87
der Hilfsstoffe für Pharmazie, Kosmetik und (1977) no. 7, 47 789 a.
angrenzende Gebiete, 2nd ed., Editio Cantor 45. K. Brunner, Justus Liebigs Ann. Chem. 351
Aulendorf 1981, pp. 480 ff. (1907) 313 – 331.
34. F. J. Bandelin, Drug Cosmet. Ind. 64 (1949) 46. Beecham Inc., US 4 601 898, US 4 601 899,
430. 1985 (R. E. Stier, W. H. Dunn, J. D. Vidra).
35. F. Hoffmann-La Roche, DE-OS 2 447 098, 47. Kaiser Aluminium & Chemical Corp., US
1974 (A. Boller, H. Scherrer). Hitachi Ltd., EP 3 714 000, 1971 (G. A. Dorsey, Jr.).
110 299, 1983 (S. Hattori, K. Iwasaki, T. 48. Texaco Inc., US 3 537 992, 1967 (J. H.
Kitamura, A. Muko et al.). Chisso Corp., Kolaian).
DE-AS 2 751 403, 1977 (T. Inukai, H. Sato, K. 49. L. Barth, K. Senhofer, Justus Liebigs Ann.
Yokohama). Chem. 159 (1871) 216 – 230.
36. C. S. Ough, Chem. Eng. News 65 (1987) 50. Chem. Fabr. AG vorm. E. Schering, DE
19 – 28. 81 297, 1894. F. Mauthner, J. Prakt. Chem.
37. Yoshitomi Pharmaceutical Industries, JP 156 (1940) 150 – 153.
62 76 651, 1987 (T. Inoue, M. Horie, T. Sato). 51. Fuji Photo Film Co., GB 1 360 581, 1971 (E.
38. Mallinckrodt Inc., US 4 309 564, 1980 (D. F.
Katsusuke, I. Sadao).
Loncrini, J. J. Taylor).
52. E. A. Haddock, Phytochemistry 21 (1982)
39. Mitsubishi Paper Mills Ltd., JP 58-138 689, JP
no. 5, 1049 – 1062. E. Hasalam, Fortschr.
58-151 292, JP 58-8 686, EP-AP 86638, 1983
Chem. Org. Naturst. 41 (1982) 1 – 46.
(H. Tsukahara, T. Torii).
40. Ueno Pharma Co., JP 58-158 289, 1983 (R. S. Kobayashi, M. Ariga, T. Ozawa, H.
Keno, H. Tsuchiya, S. Itou, T. Tsuchida). Imagawa, Agric. Biol. Chem. 48 (1984) no. 2,
Honshupaper Co., JP 58-158 290, 1983 (M. 389 – 395.
Jsoda, S. Nishimatsu, K. Kuniyama). Fuji 53. J. W. Thompson, E. R. Sherwin, J. Am. Oil
Photo Film Co., JP 58-87 089, 1983 (M. Chem. Soc. 43 (1966) no. 12, 683 – 686.
Watanabe). Konzaki Paper Mfg Co. Ltd., JP-P 54. P. Pfeiffer, E. Schmitz, Pharmazie 5 (1950)
210 057-82, 1982 (H. A. T. Suzuki, R. O. 517 – 520.
Hayashi, N. I. O. Arai, Y. I. N. Oeda). 55. Cis. Inc., 7215 York Road, Baltimore, MD
41. Mitsui Toatsu Kagaku KK, JP 49-19 275, 1974 21212, CIS-Data bank (RTECS and
(R. Fujihara, K. Nitta). OHMTADS).
42. Siemens, DE-OS 3 049 417, 1980 (H. 56. D. R. Boreham, B. K. Martin, Br. J.
Januschkowetz, H. Laub). Pharmacol. 37 (1969) no. 1, 294 – 300; Chem.
43. I. A. Pearl, J. Am. Chem. Soc. 68 (1946) Abstr. 71 (1969) 121 980 u.
2180 – 2181. I. A. Pearl, J. Am. Chem. Soc. 68
(1946) 1100 – 1101.
Hypnotics
Hartmund Wollweber, Bayer AG, Wuppertal, Federal Republic of Germany
who display larger differences (>1 ◦ C) are long when alterations in thinking or behavior are nec-
sleepers. Body tissues are regenerated during or- essary. Confrontation with problems also inten-
thodox sleep, and cerebral functions during the sifies REM sleep, e.g., an astronaut experiences
REM phase. greatly increased REM sleep in his first night
The REM phases account for ca. 20 – 25 % in space. Continuous suppression of REM sleep
of sleep and appear to be periods of relatively leads to psychological disturbances and psycho-
high cerebral activity in which dreams occur. somatic symptoms.
They are detected by measuring the spontaneous
movements of the closed eye. In the absence of
these dream phases sleep is not restful. Dreams 1.1. Amino Acids and Hormones
are of great importance for learning and mem- Influencing Sleep
ory; they assimilate the experiences of the pre-
ceding days. Generally, dreams can be remem- Three neurohormones are involved in control-
bered slightly or not at all. With increasing age, ling sleep: serotonin, noradrenaline, and acetyl-
the duration of REM sleep steadily decreases choline. Serotonin regulates orthodox sleep,
to about 20 – 30 % of its original duration. The whereas acetylcholine and noradrenaline influ-
REM phase is especially activated when a per- ence the REM phase. The rates of cell division
son is confronted with an emotionally arousing, are the highest during sleep, and large amounts
intellectual task calling for creative effort, i.e., of growth hormone (GH), which stimulates pro-
Hypnotics 3
tein synthesis, are secreted during the third and all, because of their many adverse effects. Alco-
fourth levels of orthodox sleep. The EEG waves hol should not be taken with sedatives or hyp-
at this stage are also large and slow. notics because the reactions of the patient are
Melatonin is rhythmically secreted by the then adversely affected.
pineal gland. It stimulates specific receptors in Prolonged use of sedatives and hypnotics re-
the body’s biological clock, the suprachiasmatic duces their effect on the cells of the central ner-
nuclei of the hypothalamus and may thus regu- vous system (see also Section 5.1). Most hyp-
late the human circadian rhythm by increasing notics, particularly highly active preparations,
serotonin levels. The human biological clock can can create psychological and physical depen-
be “adjusted” with exogenous melatonin. Conti- dence, resulting in withdrawal symptoms when
nous exposure of humans to light leads to sleep treatment is terminated. In addition, many hyp-
disorders by inhibiting melatonin synthesis [16], notic drugs that suppress and shorten the REM
[29], [30]. phase produce a rebound effect, i.e., a sudden
A sleep inducing nonapeptide, delta-sleep- withdrawal after extended periods of use causing
inducing peptide [62568-57-4] (DSIP), has been a prolongation of the REM phase. As in alcohol
isolated from rabbit serum. This peptide, which withdrawal, this effect can even lead to delirium
also occurs in humans, has been synthesized tremens. In this way, the body tries to make up
and has been prepared by genetic engineering for lost REM sleep.
techniques. DSIP coordinates the rhythmic cir-
cadian sleep – wake cycles by promoting sleep
functions (without having a sedative effect), and 1.3. Requirements of Hypnotics
enhances mental performance when awake. At
the same time, DSIP invokes a higher tolerance There is still no completely satisfactory hypnotic
to stress. The action of this nonapeptide is at- drug. The following requirements can be defined
tributed to modulation of the adrenergic recep- for an ideal hypnotic:
tors [31].
l-Tryptophan, a common amino acid found – reliable effect
in the human body is a precursor of serotonin. In – wide therapeutic range
daily doses of 4 – 7 g it causes sedation during – limited duration of action, no longer than 8 h
the day and a shortening of latency of sleep at – physiological sleep
night. It is registered in Switzerland as a “bio- – no disturbance of deep sleep or REM sleep
logical sleep-inducing drug.” However, studies – low toxicity
have questioned the sleep-modulating effect of – no enzyme induction
this amino acid [32]. – no respiratory depression
– no additive effect with alcohol
– no amnesia, confusion, or ataxia after arousal
1.2. Influence of Hypnotics on Sleep in the night
– no excessive daytime sedation
Most hypnotic drugs shorten both phases of – no loss of potency on repeated use
sleep and suppress the paradoxical (REM) – no risk of dependency, e.g., insomnia after
phase. Thus, the sleep curve is levelled out, and termination of treatment
the EEG resembles that of a narcotized per- – no after effects (e.g., hangover)
son. Hence, drug-induced sleep does not have
the restorative quality of natural sleep and is
no substitute for the treatment of insomnia. A 1.4. Applications of Hypnotics
research report of the Max Planck Institut für
Psychiatrie (München, 1980) states that conven- Sleep-inducing drugs are used by the following
tional sleep-inducing drugs do not produce rest- groups:
ful sleep; since drug-induced sleep is not iden-
tical to natural sleep, taking sleeping pills is not 1) healthy people with demanding jobs or irreg-
advised. Pharmacologists also recommend the ular working hours, who often take sleeping
use of hypnotic drugs only for a few days, if at pills after stimulants such as coffee;
4 Hypnotics
2) sick people with symptoms such as fever, 3) other drugs with sedative effects, (e.g.,
sore throat, or cold; tranquilizers, antidepressants, and antihis-
3) old people with insomnia due to cerebral tamines); and
sclerosis usually have an inverted day – night 4) natural preparations, e.g., valerian, ethanol,
rhythm; they are awake and mobile at night, and DSIP.
and tired and peaceful in the morning.
4) sick people with uncomfortable and sleep- There is no sharp distinction between seda-
less nights due to pain, who have difficulty tives and hypnotic drugs; differences in activity
in breathing, heart trouble, etc.; and are quantitative rather than qualitative. Both in-
5) bedridden people with serious illnesses, who hibit the activation system in the reticular for-
are deprived of physical exercise. mation. Drugs belonging to groups 1 and 2 are
true hypnotics. Drugs belonging to group 3 are
People belonging to groups 1 and 2 can be only occasionally used as hypnotics.
treated with mild relaxants by reducing the sus- Der Bewertende Arzneimittelindex (The Val-
ceptibility of the reticular formation (site of the uating Drug Index) assesses “benzodiazepine
center of wakefulness in the brain) to external sleep-inducing drugs” positively, especially be-
stimuli. These substances include plant compo- cause of their wide therapeutic range [37]. Other
nents with sedative properties, such as valerian recommended drugs are chloral and its deriva-
and hop, and tranquilizers or sedatives which, tives (see Sections page 7page 7), diphenhy-
even in high doses, do not have narcotic prop- dramine, and doxylamine (see Sectionpage 5).
erties. People belonging to groups 4 and 5 may Derivatives of barbituric acid, piperidinediones,
have to be treated for extended periods with hyp- and quinazolinones are not recommended. In
notics and even with antipsychotic and antide- comparison with the benzodiazepines, these
pressant drugs. cause more undesirable effects (respiratory de-
pression, negative influence on REM sleep, en-
zyme induction, and problems of dependence).
1.5. Consumption of Hypnotics Their therapeutic range is narrow. Compounds
containing bromine are also to be avoided be-
The importance of hypnotic drugs is reflected in cause they may cause bromism (see Chap. 2).
the fact that 21 % of the population of France Valerian preparations (e.g., Valepotriate) and
(February 1975) and 32 % of the inhabitants of their components are not favored because of
Los Angeles (1973) suffered from insomnia. In their in vitro cytotoxic and alkylating action. The
France, 2.5×106 tablets, capsules, and dragées therapeutic importance of these drugs cannot be
are taken every evening to promote sleep [33– assessed at present [37].
38]. The dramatic increase in the consumption The lists of trade names cited for individual
of hypnotics with age can only be partly justi- hypnotics are not complete.
fied because older people require less sleep at
night. Instead, they have a more pronounced en-
dogenous sleep profile with increased cerebral 2. Sedatives, Antihistamines, and
readiness to sleep at 9.00, 13.00, and 17.00 h. Tranquilizers
Sedatives and Tranquilizers. There is no
1.6. Classification and Evaluation of significant distinction between sedatives and
Hypnotics hypnotics; the same drugs are often used for both
purposes. Sedatives (→ Sedatives) and tranquil-
Hypnotics may be classified as follows [34]: izers (→ Psychopharmacological Agents) are
used to sedate and relieve anxiety during the day
1) benzodiazepine hypnotics; and, in addition, to induce sleep at night. Thera-
2) nonbenzodiazepine hypnotics, e.g., chloral peutic doses of tranquilizers have a calming, re-
hydrate, barbiturates, piperidinediones, and laxing effect and alleviate psychological symp-
methaqualone; toms such as fear, anxiety, restlessness, and ten-
sion. Ideally, they should not produce sedative
Hypnotics 5
or hypnotic effects. The first therapeutically used Trade Names: Calmonal (Heiden, FRG), Bonine
hypnotics were plant extracts with either seda- (Pfizer, USA).
tive (valerian root and hop) or analgesic action Diphenhydramine hydrochloride [147-24-0],
(opium alkaloids). In spite of the sedative effect 2-(benzhydryloxy)-N,N-dimethylethylamine
of opium alkaloids [39], they should not be used (→Antiallergic Agents)
for obvious reasons (addiction, euphoria, poor Trade Names: Benadryl (Parke-Davis, USA),
dose adjustment, and legality). Of all the plant Halbmond (Much, FRG).
extracts available, only one substance, valmane, Doxylamine succinate [562-10-7], 2-di-
is exactly defined from a chemical standpoint. methylaminoethoxyphenylmethyl-2-picoline
Valepotriate is a mixture of valtratum Trade Names. Decapryn, Meteprine (Merrell,
[18296-44-1] (1), acevaltratum [25161-41-5] USA), Unisom (Pfizer, USA).
(2), and didrovaltratum [18296-45-2] (3):
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.
The sodium salt, triclofos sodium [7246-20-0], These compounds are mild hypnotic agents.
C2 H3 Cl3 NaO4 P, is soluble in water. This prepa- They are generally used to promote the onset
ration has the same physiological properties as of sleep or as daytime sedatives, often in com-
chloral hydrate, but it does not have the unpleas- bination with weak analgesics. They are well-
ant physical characteristics. The compound is absorbed and, depending on the structure, have
metabolized in the body mainly to the hypnot- few side effects. Dependence on these com-
ically active 2,2,2-trichloroethanol. A dose of pounds occurs less frequently than on barbitu-
1.5 g is equivalent to 1 g of chloral hydrate. rates and benzodiazepines.
Trade Names. Triclos (Lakeside, USA), Triclo- Urethane [51-79-6], ethyl carbamate, a mild
syl (Glaxo, UK), Tricloran (C.T.S. Israel). hypnotic agent, is no longer prescribed because
Petrichloral [78-12-6], pentaerythritchloral, its cytostatic properties can cause agranulocyto-
C13 H16 Cl12 O8 , M r 725.76, mp 52 – 54 ◦ C [46]. 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
Chloralodol [3563-58-4], chlorhexadol, (→Psychopharmacological Agents).
2-methyl-(2 ,2 ,2 -trichloro-1 -hydroxy-4- Ureides are still widely prescribed as mild
ethoxy)-pentan-2-ol, C8 H15 Cl3 O3 , M r 265.58, sedatives and hypnotic agents. They are used
mp 102 – 124 ◦ C [47]. 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
Trade Names. Lora (Wallace, USA), Mechloral, sleep.
Mecloral (Dumex, Denmark).
Chloralose [15879-93-3], α-O-(2,2,2-tri- Ethinamate [126-52-3], 1-ethynylcyclo-
chloroethylidene)-1,2-α-d-glucofuranose-(R), hexylcarbamate, C9 H13 NO2 , M r 167.21,
C8 H11 Cl3 O6 , M r 309.54, mp 187 ◦ C, [α]22
D + 19 mp 96 – 98 ◦ C, solubility in water 0.25 %, in
(5 % solution in 98 % ethanol) [48]. ethanol 35 %, and
Trade Names. Valamin (Schering, FRG), broken down to bromide. The same applies to
Valmid, Valmidate (Lilly, USA). acecarbromal and bromisoval.
Trade Names. Benuride (Bengue, UK; Vinas, diacylated cyclic ureas: barbiturates (8) and
Spain), Deturid (Polfa, Poland; Sapos, Switzer- thiobarbiturates (9)
land). N-(β-carbonyl)ethyl derivatives: 2,4-piperi-
dinediones (6)
Capuride [5579-13-5], 2-ethyl-3- N-iminoacyl derivatives: quinazolinones (7)
methylvalerylurea, C9 H18 N2 O2 , M r 186.25, phenyl N-iminoacyl derivatives: benzodiaze-
mp 172 ◦ C [60]. pines (11)
Trade Name. Pacinox (McNeil, USA).
All groups, except 5, 7, and 11, have a doubly
substituted carbon atom adjacent to the carbonyl
Valdettamil, Novonal [512-48-1], 2,2-
group of the amide structure. These substituents
diethyl-4-pentenamide, C9 H17 NO, M r 155.23,
greatly influence the action and the distribution
mp 75 – 76 ◦ C, is soluble in 120 parts of water
of the drug in the body.
and freely soluble in alcohol and ether [61].
Trade Names. Insomnia (ICN, USA), and
as combinations: Arantil (Hoechst, FRG),
Dentigoa (Scheurich, Switzerland).
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].
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
countries. Important properties are listed in Ta- The duration of action is primarily deter-
ble 1. As hypnotics, barbituric acids all pro- mined by the rate of enzymatic degradation in
duce almost identical pharmacodynamic ef- the liver, which depends on the chemical struc-
fects. However, they show wide differences ture of the barbiturate. Continual use causes en-
in pharmacokinetics and metabolism. Differ- zyme induction in the liver and thus a gradual
ences in onset and duration of activity are re- increase in the rate of barbiturate degradation.
lated to lipid and water solubility, and rate Increasing doses are required with prolonged
of metabolism. Lipophilic barbiturates, e.g., use (barbiturate habituation). In some patients
the narcotically active, lipid-soluble thiobarbitu- barbiturates not only have a sedative and hyp-
rates, and the N-monoalkylbarbiturates such as notic action, but also produce excitement and
hexobarbital rapidly disappear in adipose tissue. euphoria. This sense of well-being may easily
As a result of their redistribution in lipophilic lead to barbiturate addiction, a physical depen-
compartments of the body, these compounds dence with withdrawal symptoms. Barbiturate
have a very short duration of action and are addicts take the hypnotic because of its euphoric
rapidly broken down (short-acting narcotics, properties and try to compensate for the central
→Anesthetics, General, Chap. 3.1.). depression by using psychoanaleptic drugs.
From a therapeutic standpoint, barbiturates Cases of accidental and suicidal barbiturate
used as hypnotic agents are classified according poisoning are fairly common. They are charac-
to their duration of action (see also Table 1): terized by unconsciousness (which can precede
a delirious stage), central respiratory depression
1) long-acting compounds (barbital, phenobar- accompanied by oxygen deficiency symptoms,
bital), and in severe poisoning circulatory collapse with
2) medium-acting compounds (amobarbital, cy- death in 12 h to 4 d. The prognosis for barbitu-
clobarbital), rate poisoning is favorable if treatment is imme-
3) short-acting compounds (hexobarbital), and diate. Essential measures include artificial res-
4) compounds whose effects begin after 15 – piration, irrigation of the stomach, circulatory
20 min (secobarbital).
Hypnotics 11
support, forced osmotic diuresis using manni- Trade Names. Dormileno (Faes, Spain), Hyp-
tol, exchange transfusion, and hemodialysis. nox (Püschel, Austria), Veronal (Bayer, FRG;
Barbiturates produce a condition superfi- Merck, FRG), Veroletten (EAS-Labor, Austria).
cially resembling natural sleep. However, the Sodium salt: Barbinetten (Hormosani, Aus-
REM phase, which is very important for rest tria).
and recuperation, is shortened. Continued use
of barbiturates leads to a very low REM value, Phenobarbital [50-06-6], phenemalum,
and sudden withdrawal increases the REM phase phenobarbitone, 5-ethyl-5-phenylbarbituric
(rebound effect). The adverse effects of barbitu- acid, C12 H12 N2 O3 , M r 232.23, mp 174 –
rates are well-documented after more than 50 178 ◦ C. The acid crystallizes from water in
years of experience. This is an important advan- the form of flakes. It is odorless, with a slightly
tage over newly developed hypnotics [68]. bitter taste; an aqueous solution reddens litmus
The pK values of the hypnotically active bar- paper. One part dissolves in 1100 parts of water
bituric acids, range between 7.78 and 8.30. Dis- at 20 ◦ C, 40 parts of boiling water, 10 parts of
ubstituted barbituric acids are present, to ca. ethanol, 18 parts of diethyl ether, and 60 parts
50 %, in the undissociated form under physio- of chloroform. It is synthesized by the reaction
logical conditions (pH 7.4). Only this form of the of urea with ethyl phenyl malonate in sodium
drug can pass through biological membranes. ethoxide solution [71].
The degree of dissociation is strongly influenced The sodium salt of phenobarbital [57-30-7],
by slight alterations in pH. C12 H11 N2 NaO3 , M r 254.23, is a colorless, crys-
Production. The 5,5-disubstituted barbituric talline, hygroscopic, freely water-soluble pow-
acids are still synthesized by the classical con- der. It is slightly soluble in ethanol, and insolu-
densation of substituted malonic esters (12), ble in diethyl ether and chloroform. The aqueous
malonic amides, or cyanoacetic esters (13) with solution is stable for only a short time.
urea (14), thiourea (15), guanidine (16), or di- Phenobarbital is a long-acting hypnotic with
cyanodiamide (17), followed by the hydrolysis anticonvulsant properties. Long-term use in
of the resulting imino- or cyanobarbiturate [69] higher than therapeutic doses can cause phys-
(Fig. 3). ical addiction. An overdose resulting in blood
Under acidic or neutral conditions, substi- levels of 8 – 12 mg/100 mL can cause death.
tuted malonic acids or malonic chlorides can Trade Names. Agrypnal (Eggochemia, Austria),
also be used for the condensation with urea Aphenylbarbit (Streuli, Switzerland), Calminal
or thiourea. It is preferable for the substituents (Wolfs, Belgium), Epsylone (Powell, Canada),
R1 and R2 to be present in the malonic ester Eskabarb Span (Smith, Kline & French, USA),
or cyanoacetic ester molecule; disubstitution of Fenemal (DAK, Denmark), Gardenal (Spe-
barbituric acids at the 5-position is only possi- cia, Canada), Gardenale (Farmitalia Carlo
ble with highly reactive halides, such as allyl Erba, Italy), Hypnolone (Hartz, Canada),
halides. Pepinal, Lepinaletten (Arzneimittelwerk Dres-
den, GDR), Luminal (Bayer, FRG; E. Merck,
Barbital [57-44-3], barbitone, 5,5- FRG; Winthrop, USA), Mediphen (Medic,
diethylbarbituric acid, C8 H12 N2 O3 , M r 184.19, Canada), Phen Bar (Saunders, Canada), Phe-
mp 188 ◦ C, (sodium salt [144-02-5], nobarbyl (Synochem, FRG), Seda-Tablinen
C8 H11 N2 NaO3 , Mr 206.18), has a bitter taste. (Beiersdorf, FRG), Solfoton (Poythress, USA),
One part of the acid dissolves in 170 parts of Teolaxin (Paul Maney, Canada); Phenobarbi-
water at 25 ◦ C or 17 parts of water at 100 ◦ C. tal diethylamine salt: Gratusminal (Farmasimes,
It is freely soluble in acetone, ammonia, and Spain); Phenobarbital calcium salt: Fenileal
lye. Barbital is prepared by the condensation (Turon, Spain), Lumcalcio (Abello, Spain).
of diethyl malonate with urea in the presence
of sodium ethoxide [70]. Both the acid and its Butobarbital [77-28-1], butobarbitone, 5-
sodium salt are used as long-acting hypnotics; butyl-5-ethylbarbituric acid, C10 H16 N2 O3 ,
abuse can cause habituation or addiction. 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
acting hypnotic; abuse can cause habituation or in water; preparation is described in [74], and
addiction. metabolism in [75].
Trade Names. Butenil, Butynoct (Interpharm, Amobarbitol is a widely used medium- to
Austria), Etoval, Longanoct, Sonabarb (Protea, long-acting hypnotic. It resembles secobarbi-
Australia), Soneryl (Specia, Italy; May & Baker, tal and pentobarbital in action with a some-
USA). what longer duration. A blood level of 3 –
6 mg/100 mL of blood is fatal; abuse can cause
Probarbital [76-76-6], 5-ethyl-5-isopro- habituation or addiction.
pylbarbituric acid, C9 H14 N2 O3 , M r 198.22, Trade Names. Amsal (Adams, Australia), Amy-
mp 197 – 198 ◦ C, is slightly soluble in cold cal (AFI, Sweden), Amydorm, Amytal (Lilly,
water, but more soluble in hot water and eth- USA), Isobec (Pharbec, Canada), Isonal (ICN,
anol. Sodium salt [143-82-8], C9 H13 N2 NaO3 , Canada), Neur-Amyl (Fawns & McAllen, Aus-
M r 220.20, is a hygroscopic powder that is solu- tralia), Placidel (Miquel, Spain). Stadadorm
ble in water. The calcium trihydrate [545-74-4], (Stada, FRG).
C18 H26 CaN4 O6 · 3 H2 O, M r 488.55, forms Amobarbital sodium: Altinal, Amsal
slightly bitter crystals; 1 g dissolves in 40 g wa- (Adams, Australia), Amsebarb (Paul Maney,
ter; preparation is described in [73]. Probarbital Canada), Amylbarb Sodium (Protea, Aus-
is a long- to medium-acting hypnotic; abuse can tralia), Amylobeta (Bramble, Australia), Amy-
cause habituation or addiction. tal Sodium (Lilly, USA), Barbamyl (Teva,
Trade Names. Ipral Sodium, Ipral Calcium Israel), Dorminal (Boots, UK), Neur-Amyl
(Squibb, USA). (Fawns & McAllen, Australia), Novamobarb
(Novopharm, Canada).
Amobarbital [57-43-2], 5-ethyl-5-isopen-
tylbarbituric acid, C11 H18 N2 O3 , M r 226.28, mp Cyclobarbital [52-31-3], cyclobarbitone,
156 – 158 ◦ C, forms slightly bitter crystals; 1 g 5-(1-cyclohexen-1-yl)-5-ethylbarbituric
dissolves in 1300 mL water and in 5 mL ethanol. acid,C12 H16 N2 O3 , M r 236,26, mp 171 –
The sodium salt, C11 H17 N2 NaO3 [64-43-7], 174 ◦ C, forms bitter crystals and is moderately
M r 248.25, is hygroscopic and readily soluble soluble in water. The calcium salt [143-76-0],
C24 H30 CaN4 O6 , M r 510.59, is prepared by heat-
Hypnotics 13
Trade Names. Fabadorm (Bayer, FRG), Phan- Heptabarb [509-86-4], heptabarbital, hept-
odorm (Bayer, FRG; Merck, FRG; Winthrop, abarbitone, 5-ethyl-5-(1-cycloheptenyl)barbitu-
USA), Phanotal (Kwizda, Austria). ric acid, C13 H18 N2 O3 , M r 250.29, mp 174 ◦ C;
one part dissolves in 13 parts of methanol,
Aprobarbital [77-02-1], 5-allyl-5-isopro- 20 parts of ethanol, 13 parts of acetone, 40
pylbarbituric acid, C10 H14 N2 O3 , M r 210.23, parts of diethyl ether, and 5000 parts of wa-
mp 140 – 141.5 ◦ C, forms slightly bitter crystals, ter. It is synthesized by the condensation of
and is insoluble in water, but soluble in ethanol. ethyl(cyclohepten-1-yl)-cyanoacetic ester with
The sodium salt [125-88-2], C10 H13 N2 NaO3 , urea [81].
M r 232.21, a hygroscopic, bitter powder, is sol- This compound is a short- to medium-acting
uble in water. Preparation is described in [77]. hypnotic; abuse can cause habituation or addic-
This compound is a medium-acting hypnotic; tion.
abuse can cause habituation or addiction. Trade Names. Medomin, Medapan (Ciba-
Trade Names. Alurate (Hoffmann-La Roche, Geigy, Switzerland).
Switzerland), Numal (United Chemicals, USA).
Secbutabarbital [125-40-6], secbutobarbi-
Allobarbital [52-43-7], allobarbitone, tal, butabarbital, secbutobarbitone, 5-ethyl-5-(2-
5,5-diallylbarbituric acid, C10 H12 N2 O3 , butyl)barbituric acid, C10 H16 N2 O3 , M r 212.24,
M r 208.21, mp 171 – 173 ◦ C, is slightly bitter. mp 165 – 168 ◦ C; the sodium salt [143-81-7],
One part dissolves in 300 parts of cold water C10 H15 N2 NaO3 , M r 234.23, is a bitter powder;
and in 50 parts of boiling water. It is prepared 1 g dissolves in 2 mL of water, and the pH of a
by heating diallyl malonate with urea in sodium 1 % aqueous solution is 9.0 – 10.2. Preparation is
ethoxide. It can also be synthesized by the re- described in [82]. Secbutabarbital is a medium-
action of barbituric acid with allyl bromide in to long-acting hypnotic. It is widely used in the
the presence of sodium acetate or copper sulfate United States as a sedative. Blood levels of 3.5 –
[78], [79]. It is a mediumacting hypnotic; abuse 6 mg/100 mL are fatal; abuse can cause habitu-
can lead to habituation or addiction. ation or addiction.
Trade Names. Diadol (Durst, USA), Dial (Ciba-
Geigy, Switzerland); it is contained in Spasmo-
Cibalgine (Ciba, Switzerland).
5.2. Piperidinediones
with codeinelike, cough-suppressant, analgesic way), Sovinal (ND & K, Denmark), Toraflon
properties; they are useful in patients who do not (Toraude, Canada), Tualone (ICN, Canada), Tu-
tolerate or respond to barbiturates. Prolonged azol (Strasenburgh, USA).
use may cause habituation and psychological de-
pendence. No specific antagonists are available Etaqualone [7432-25-9], 3-(2-ethylphenyl)-
for quinazolinone poisoning. 2-methyl-4(3 H)-quinazolinone, C17 H16 N2 O,
Mr 264.31, mp 81 ◦ C. Hydrochloride
Methaqualone [72-44-6], 2-methyl-3- [97979-65-2], C17 H17 ClN2 O, mp 247 ◦ C.
o-tolyl-4-(3 H)quinazolinone, C16 H14 N2 O, Etaqualone is prepared by the condensation
M r 250.30, mp 120 ◦ C (also 114 – 116 ◦ C), is of N-acetylanthranilic acid with 2-ethylaniline
soluble in ethanol, diethyl ether, and chloro- in the presence of phosphoryl chloride [99].
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. Trade Name. Aolan (Beiersdorf, FRG).
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 tative differences could be of importance. Stud-
are constantly being marketed. Properties of the ies point to an interaction between benzodiaze-
most important benzodiazepines are listed in Ta- pines and their receptors. Thus, they activate the
ble 2. postsynaptic receptors for γ-aminobutyric acid
(GABA, an inhibitory neurotransmitter) and in-
Mode of Action. The benzodiazepines have crease the frequency of the opening of the chlo-
the same hypnotic action as the barbiturates, ride ion canals [106].
with differences in selectivity, metabolism, and In other words, the benzodiazepines act
adverse effects. Large doses do not possess the in the central nervous system by interfering
narcotic properties exhibited by most barbitu- with GABA-mediated neurotransmission [107],
rates. Benzodiazepines do not solely have a [108]. Traces of benzodiazepine derivatives
sleep-inducing action; they exert a combination (e.g., diazepam and lorazepam) have been found
of effects: sedation, relief of anxiety (produc- in wheat and potato and are identical to the syn-
tion of a better state of mind), muscle relax- thetic substances. Benzodiazepines also occur
ation, and anticonvulsant action. These effects in untreated rats, as well as in human serum and
shorten both the intermittent waking times and brain. However, the concentration is very low
deep sleep (reduction of δ sleep). In addition, (several nanograms per gram of potato) and a
a hangover always occurs on the next day. The direct pharmacological effect on the central ner-
effects of barbiturates and benzodiazepines are vous system has not been demonstrated [108],
compared in Table 3. From a pharmacological [109].
standpoint, the hypnotically active benzodiaze-
pines have a depressant action on the electrical Metabolism. All benzodiazepines are me-
discharges in the areas controlled by the limbic tabolized in the body [110], but in some cases
system and block the function of the reticular the retention times are very long. In addition,
formation in the brain stem. They are similar to the pharmacologically active metabolites of the
the barbiturates in this respect; however, quanti- 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
Sweden), Uskan (Desitin, FRG), Vaben (Rafa, Carlo Erba, Italy; Aesca, Austria), Mabertin
Israel), Zapex (Riva, Canada). (Sidus, Argentina), Normison (Wyeth, UK),
Planum (Farmitalia Carlo Erba, Italy), Remestan
Lorazepam [846-49-1], 7-chloro-5-(2- (Wyeth, FRG), Restoril (Sandoz, Switzerland;
chlorophenyl)- 1, 3- dihydro- 3- hydroxy- 2H- Anca, Canada), Somaz (Quantum, USA), Tenso
1,4-benzodiazepin-2-one, (20b), (Castejon, Spain), Veroqual (Lek, Yugoslavia).
C15 H10 Cl2 N2 O2 , M r 321.17, mp 166 – 168 ◦ C,
is prepared from 18 (R = Cl, R1 = H, R2 = Cl), Doxefazepam [40762-15-0], 7-chloro-5-(2-
analogously to oxazepam. Lorazepam is pri- fluorophenyl)- 1, 3- dihydro- 3- hydroxy- 1- (2-
marily an anti-anxiety drug; it also has sleep- hydroxyethyl)-2H-1,4-benzodiazepin-2-one,
inducing properties and is used in anesthesiol- (20e), C17 H14 ClFN2 O3 , M r 348.57, mp 138 –
ogy [123]. 140 ◦ C, is prepared by the alkylation of 7-
Trade Names. Ativan (Wyeth, USA), Bonton chloro-5-(2-fluorophenyl)-1,3-dihydro-2H-1,4-
(Unipharma, Israel), Control (Sigurta, Italy), benzodiazepin-2-one N-oxide with 2-bromoeth-
Donix (Llorens, Spain), Emotion (Alpes, Ar- anol or 2-bromoethyl acetate, followed by
gentina), Emotival (Armstrong, Argentina), Lar- Polonowski rearrangement and saponification
pose (Cipla, India), Laubeel (Desitin, FRG), [128], [129]. Doxefazepam is a hypnotic drug
Loram (Lek, Yugoslavia), Lorans (Schiapar- with two to four times the activity and half the
elli, Italy), Lorax (Wyeth, UK), Lorivan (Disco, toxicity of flurazepam; it causes significantly
Israel), Merlit (Ebewe, Austria), Nervistop L less hangover [130].
(Ingram, Argentina), Orfidal (Orfi, Spain), Pro Trade Name. Doxans (Schiaparelli, Italy).
Dorm (Schürholz, FRG), Securit (Perrel, Italy),
Sedatival (Raffo, Argentina), Tavor (Wyeth, Cinolazepam [75696-02-5], 7-chloro-5-
FRG), Temesta (Ferrosan, Denmark; Wyeth, (2-fluorophenyl)- 2, 3- dihydro- 3- hydroxy-2-
Netherlands), Tolid (Dologiet, FRG), Tra- oxo-1H-1,4-benzodiazepine-1-propionitrile,
pax (Wyeth, Argentina), Wypax (Yamanouchi, (20f),C18 H13 ClFN3 O2 , M r 356.84, mp 190 –
Japan). 193 ◦ C, is prepared by the cyanoethylation
Trimethylacetate: Drupal (Novag, Spain), Pi- of compound 20 (R = Cl, R1 = H, R2 = F) in
ralone (Ferrer, Spain), Placinor (Robert, Spain). the presence of trimethylammonium hydrox-
ide [131]. This drug resembles doxefazepam in
Lormetazepam [848-75-9], 7-chloro- action [132].
5-(2-chlorophenyl)- 1, 3-dihydro-3-hydroxy-
1-methyl-2H-1,4-benzodiazepin-2-one (N- Quazepam [36735-22-5], 7-chloro-5-(2-
methyllorazepam), (20c), C16 H12 Cl2 N2 O2 , fluorophenyl)-1-(trifluoro-2,2,2-ethyl)-1, 3-di-
M r 335.19, mp 192 – 194 ◦ C, is prepared analo- hydro-2H-1,4-benzodiazepine-2-thione, (19),
gously to oxazepam [124]. The hypnotic effect C17 H11 ClF4 N2 S, M r 386.6, mp 138 – 139 ◦ C,
of 1 – 2 mg is similar to that of nitrazepam [125]. is synthesized by the reaction of compound
Trade Names. Lembrol (Vinas, Spain), Loramet 18 (R = Cl, R1 = CH2 CF3 , R2 = F) with phos-
(Wyeth, Belgium, Netherlands), Minias (Far- phorus pentasulfide [133]. Quazepam is a sleep-
mades, Italy), Noctamid, Pronoctan (Schering, inducing drug; it has a long duration of action
FRG), Nocton (Eurolab, Chile). and causes a hangover. It also produces long-
acting metabolites in the body [134].
Temazepam [846-50-4], 7-chloro-1,3-di- Trade Names. Dormalin (Schering Corp., USA),
hydro-3- hydroxy- 1- methyl- 5- phenyl-2H- Prosedar, Selepam, Cetrane (Schering Plough,
1,4-benzodiazepin-2-one (3-hydroxydiazepam, USA).
N-methyloxazepam, oxydiazepam), (20d),
C16 H13 ClN2 O2 , M r 300.74, mp 119 – 121 ◦ C, Haloxazolam [59128-97-1], 10-bromo-
prepared analogously to oxazepam [126]. It is 11b-(2- fluorophenyl)- 2, 3, 7- 11b- tetra-
used as a hypnotic drug and should be taken hydrooxazolo-[3,2-d][1,4]benzodiazepin-
1 – 2 h before going to sleep [127]. 6(5H)-one,C17 H14 BrFN2 O2 , M r 377.22,
Trade Names. Euhypnos (Farmitalia Carlo Erba, mp 185 ◦ C, is synthesized by the reaction of
Italy; Sigma, Australia), Levanxene (Farmitalia bromo- (or tosyloxy-) acetamidobenzophenone
24 Hypnotics
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
Hypnotics 27
15. A. A. Borbély: “Schlafmittel und Schlaf: Peptide (DSIP): An Update,” Peptides (N.Y.) 7
übersicht und therapeutische Richtlinien,” (1986) 1165 – 1187; Neurosci. Biobehav. Rev.
Ther. Umschau 43 (1986) 509 – 516. 8 (1984) 83 – 93.
16. E. Soos: “Schlaf und Schlafmittel,” OAZ 32. A. A. Borbély, G. Youmbi-Balderer: “Ist
Österr. Apoth. Ztg. 28 (1974) 93 – 101. Tryptophan ein biologisches Schlafmittel,”
17. A. N. Nicholson: “Hypnotics Their Place in Schweiz. Rundsch. Med. Prax. 75 (1986)
Therapeutics,” Drugs 31 (1986) 164 – 171. 785 – 790.
18. E. A. Swinyard: Sedatives und Hypnotics in 33. J. P. Olié, S. Lepastier, S. Dana, J. M. Birée,
Remington Pharmaceutical Sciences, 7th ed., Med. Actuelle 5 (1978) 53.
Mack Publ., Easton, Pa., 1985, 34. A. A. Borbély, Ther. Umsch. 43 (1986) 509.
pp. 1059 – 1074. 35. M. B. Scharf, E. O. Bixler, A. Kales, C. R.
19. D. Schneider-Helmert: Schlafmittel: Soldatos, Pharmacology 19 (1979) 173 – 181.
“Symptomatisch oder kurativ?” Schweiz. A. Kales et al., Pharmacology 26 (1983)
ärzteztg. 67 (1986) 1477 – 1483. 138 – 149.
20. W. P. Koella: “The Organization and 36. K. Laemmel, Schweiz. Rundsch. Med. Prax.
Regulation of Sleep,” Experientia 40 (1984) 74 (1985) 1305 – 1310.
309 – 338. 37. E. Greiser: Bewertender Arzneimittelindex:
21. W. Keup: “Mißbrauch und Abhängigkeit von Hypnotika, vol. 2, Med. Pharm. Verlag,
Sedativa und Hypnotika,” Internist 27 (1986) Wiesbaden 1983, pp. 1 – 25.
746 – 756. 38. J. Krieglstein, D. Grusla, Dtsch. Apoth. Ztg.
22. K. Laemmel: “Der Schlaf-Seismograph von 128 (1988) 2041.
Körper und Seele,” Schweiz. Rundsch. Med. 39. E. F. Domino, Annu. Rev. Pharmacol. 2
Prax. 74 (1974) 1305 – 1310. (1962) 215.
23. A. Borbély: Das Geheimnis des Schlafs, 40. F. Hunzicker, F. Künzle, J. Schmitz, Helv.
Deutsche Verlagsanstalt GmbH, Stuttgart Chim. Acta 49 (1966) 1 433. Wander,
1984. GB 1 006 156, 1965; US 3 389 139, 1968.
24. T. Zilker, M. Clarmann: 41. G. Stille, A. Sayers, H. Lauener, E.
“Schlafmittelvergiftungen,” MMW Münch. Eichenberger, Psychopharmakologia 28
Med. Wochenschr. 126 (1984) 297 – 300. (1973) 325; Chem. Abstr. 78 (1973) 131 982 c.
25. V. B. Meyer-Rochow: “Vom Winterschlaf
Y. Take et al., Takeda Kenkyushoho 29 (1970)
lernen,” Selecta 1984, 1116 – 1120.
416; Chem. Abstr. 74 (1971) 21 815 n. H.
26. Drug, Facts and Comparisons, Facts and
Yokotani et al., Takeda Kenkyushoho 29
Comparisons Div., St. Louis 1982, 976 – 1016
(1970) 441; Chem. Abstr. 74 (1971) 21 816 p.
(a division of J. B. Lippincott Co.,
42. W. M. McLamre, S. Y. P. An, A. Barley, J. Org.
Philadelphia-Toronto).
Chem. 20 (1955) 109. Pfizer, US 2 746 900,
27. Martindale, The Extra Pharmacopoeia,
1953.
Hypnotic and Sedatives, 28th ed.,
43. Smith, Kline and French, DE-AS 1 183 333,
Pharmaceutical Press, London 1982,
1963.
pp. 792 – 817.
28. E. A. Swinyard in Remington’s 44. H. C. Caldwell, H. J. Adams, D. E. Rivard, J. V.
Pharmaceutical Sciences, 17th ed., Mack Swintovsky, J. Pharm. Sci. 56 (1967) 920.
Publ. Easton, Pa., 1985, 1059 –1074. 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.
Specific References
48. A. Pictet, F. H. Reichel, Helv. Chim. Acta 6
29. A. A. Borbély, J. Neural. Trans. Suppl. 21
(1923) 621. W. Freudenberg, A. M. Vajda, J.
(1986) 243 – 254.
Am. Chem. Soc. 59 (1937) 1955.
30. S. H. Reppert et al., Science (Washington,
49. C. Bischoff, Ber. Dtsch. Chem. Ges. 7 (1874)
D.C.) 242 (1988) 78 – 81.
31. M. Monnier et al., Experientia 33 (1977) 548. 628.
A. E. Kornmüller, H. B. Lux, K. Wink, M. 50. Rheinpreussen, DE-AS 1 021 843, 1951.
Klee, Naturwissenschaften 48 (1961) 51. Schering, US 2 816 910, 1953, Dtsch. Prior.
503 – 505. D. Schneider-Helmert: “Die 1952.
körpereigene Schlafsubstanz DSIP,” Schweiz. 52. P. Läuger, M. Prost, R. Charlier, Helv. Chim.
Rundsch. Med. 74 (1985) 1311 – 1315. M. V. Acta 42 (1959) 2379. Labaz, US 2 931 828,
Graf, A. J. Kastin: “Delta-Sleep-Inducing 1960.
Hypnotics 29
53. Carlo Erba, GB 736 340, 1955. 78. Ciba, DE 268 158, 1911.
54. G. Newbery, J. Chem. Soc. 127 (1925) 295. 79. Hoffmann-La Roche, DE 526 854, 1930.
Miles, US 2 854 379, 1958; US 2 931 832, 80. M. Seefelder, Justus Liebigs Ann. Chem. 652
1960. (1962) 107. BASF, Ludwigshafen, Festschr.
55. Bayer, DE 225 710, 1909. Friedländer 10, Carl Wurster 60. Geburtstag 1960, 71; Chem.
1 160. Abstr. 56 (1962) 9 927 i; BASF FP 869 Méd.
56. Bayer, DE 286 760, 1913. Friedländer 12, 1961.
704. Bayer, DE 327 129, 1917. Friedländer 81. Geigy, FR 870 714, 1941; DE 756 489, 1941.
13, 809. 82. Eli Lilly, US 1 856 792, 1932.
57. Knoll, DE 185 962, 1906. Friedländer 8, 1217; 83. Bayer, DE 293 163, 1915. Geigy, GB 650 354,
US 914 518, 1909. 1948.
58. Merck Index 10 (1983) 773. E. Schröder, C. 84. J. D. Riedel AG, US 1 739 662, 1929.
Rufer, R. Schmiechen: Pharmazeutische 85. A. C. Cope, E. M. Hancock, J. Am. Chem. Soc.
Chemie, Thieme Verlag, Stuttgart – New York 61 (1939) 776. Sharp & Dohme, US 2 187 703,
1982, pp. 251 – 252. 1940.
59. Bayer, DE 249 241, 1912. Friedländer 10, 86. E. H. Volwiler, J. Am. Chem. Soc. 47 (1925)
1 164. Labs. Aub., CH 374 644, 1964. 2236.
60. Millmaster Onyx, US 3 282 998, 1966. E. H. 87. Compagnie de Bethune, DE 589 947, 1930.
Volwiler, D. L. Tabern, J. Am. Chem. Soc. 58 Friedländer 19, 1 198.
(1936) 1 352. 88. J. D. Riedel AG, US 1 622 129, 1927;
61. M. Bockmühl, O. Schaumann, Dtsch. Med. DE 481 733, 1923.
Wochenschr. 54 (1928) 270. 89. Eli Lilly, US 1 954 429, 1934.
62. Lion Dentifrice (Tokio), DE-OS 1 768 445, 90. O. Schneider, H. Frick, A. H. Lutz, Experientia
1967; US 3 639 457, 1967; DE-OS 2 753 792, 10 (1954) 135. Festschrift F. Reinhardt AG,
1977. Basel 1936, p. 195. Hoffmann-La Roche,
63. N. Okudeira, S. Torii, S. Endo, US 2 151 047, 1938; US 2 680 116, 1954.
Psychopharmacology 70 (1980) 117. K. 91. Ciba-Geigy, US 2 673 205, 1954,
Hillier, Drugs Future 6 (1981) 536 – 537. Schweiz.-Prior. 1951; DE 950 193, 1952. L.
64. W. J. Doran in F. F. Blicke, R. H. Cox (eds.): Hoffmann, E. Tagmann, Helv. Chim. Acta 35
“Barbituric Acid Hypnotics,” Med. Chem. Ser. (1952) 1 541.
Monogr. 4 (1959). 92. Richardson-Merrell Inc. BE 753 846, 1971;
65. R. Y. Levina, F. K. Velichko: “Advances in the DE-AS 2 035 636, 1969.
Chemistry of Barbituric Acids,” Russ. Chem. 93. H. Koch, J. Kotlan, Monatsh. Chem 97 (1966)
Rev. (Engl. Transl.) 29 (1960) 437 – 459. 1 648;Kwizda, DE-AS 1 695 346, 1966.
66. Kirk-Othmer, 2nd ed., 3, 60 – 77. 94. L. Stockinger, H. Koch, Arzneim. Forsch. 19
67. A. Schlieper, Justus Liebigs Ann. Chem. 55 (1969) 167.
(1845) 251; 56 (1845) 1. 95. G. Argyropoulos, G. Lanner, E. Pasquali,
68. R. Gross, Internist 17 (1976) 229. Wien. Med. Wochenschr. 126 (1976) 75; Med.
69. R. J. Levina, F. K. Velichko: “Advances in the Actual. 16 (1980) 388.
Chemistry of Barbituric Acids,” Russ. Chem. 96. K. H. Boltze et al., Arzneim. Forsch. 13 (1963)
Rev. (Engl. Transl.) 29 (1960) 437. 688.
70. E. Fischer, A. Dilthey, Justus Liebigs Ann. 97. B. Dubnick, C. A. Towne in A. Cerletti, F. J.
Chem. 335 (1904) 338. Merck, DE 146 496, Bove (eds.): “The Present Status of
1903.Friedländer 7, 651. Psychotropic Drugs,” Proc. of the VI. Internat.
71. Bayer, DE 247 952, 1911, Friedländer 11, Congr. of the Coll. Internat.
926. Neuro-Psychopharmacologicum, Tarragona
72. Poulenc Frères, US 1 609 520, 1926.
Apr. 1968 pp. 24 – 27; Excerpta Med. 1969,
73. Lambert Thorpe Co., US 1 255 951, 1917; US
77 – 83.
1 576 014,1926.
98. I. K. Kacker, S. H. Zaheer, J. Indian Chem.
74. E. Leyraud, GB 191 008, 1922. Eli Lilly,
Soc. 28 (1951) 344; Chem. Abstr. 46 (1952)
US 1 514 573, 1926; US 1 856 792, 1932.
11 210. General Aniline u. Film Corp.,
75. H. H. Frey, M. P. Magnussen, Arzneim. Forsch.
US 3 213 094, 1963. Lab. Toraude,
16 (1966) 612.
GB 843 073, 1960.
76. Bayer, DE 442 655, 1924; DE 231 150, 1924.
77. Hoffmann-La Roche, US 1 444 802, 1923.
30 Hypnotics
99. Beiersdorf AG, BE 615 282, 1962; Chem. Med. Assoc. 247 (1982) 2147 – 2151. U.
Abstr. 58 (1963) 13 971 ef. S. S. Parmar, K. Klotz: Progress in Clinical Biochemistry and
Kishok, P. K. Seth, R. C. Arona, J. Med. Chem. Medicin, Springer, Berlin – Heidelberg – New
12 (1969) 138. York – Tokyo 1984, pp. 127 – 129. T. A.
100. J. Klosa, J. Prakt. Chem. 14 (1961) 84; 20 Hamor, J. L. Martin in G. P. Ellis, G. B. West
(1963) 283. (eds.): Progress in Medicinal Chemistry 20
101. E. Szirmai, Ther. Umsch. 20 (1963) 542. (1985) p. 183.
102. L. H. Sternbach: “The Benzodiazepine Story,” 107. W. E. Müller, MMW Münch. Med. Wochenschr.
Progr. Drug Res. 22 (1978) 229 – 266. W. 124 (1982) 879 – 884; Dtsch. Apoth. Ztg. 126
Schallek, W. D. Horst, W. Schlosser: (1986) 1036 – 1044.
“Mechanism of Action of Benzodiazepines,” 108. W. E. Müller, Med. Actual. 24 (1988)
Adv. Pharmacol. Chemother. 16 (1979) 649 – 663.
45 – 87. L. O. Randall, W. Schallek, L. H. 109. E. Müller, Dtsch. Apoth. Ztg. 128 (1988)
Sternbach, R. Y. Ning in M. Gordon (ed.): 672 – 674. Roche Magazin 1988, no. 31, 3 – 13
“Chemistry and Pharmacology of the (Hoffmann-La Roche).
1,4-Benzodiazepines,” 110. S. Ebel, H. Schütz: “Biotransformation von
Psychopharmacological Agents, A Series of Benzodiazepin-Derivaten,” Pharm. Unserer
Monographs in Medicinal Chemistry, vol. 3, Zeit 8 (1979) 87. U. Klotz, L. Kangas, J.
Academic Press, New York – San Kanto: “Clinical Pharmacocineties of
Francisco – London 1974, p. 175. V. C. Benzodiazepines,” Prog. Pharmacol. 3 (1980)
Williams, G. C. Varnado, P. U. Nwangwu: 1 – 72. W. E. Müller: “über die Bedeutung der
“Perspectives on the Clinical Pharmacology of Plasmaproteinbindung für Verteilung und
Benzodiazepines,” Med. Actual. 21 (1985) Wirkung von Tranquillantien vom
75 – 96. T. A. Hamor, J. L. Martin in G. P. Ellis, Benzodiazepintyp,” Klin. Wochenschr. 55
G. B. West (eds.): “The Benzodiazepines,” (1977) 105 – 110.
Progress in Medicinal Chemistry 20 (1985) 111. R. G. Borland, A. N. Nicholson, Br. J. Clin.
158 – 223. A. Vela-Bueno, A. Kales: Pharmacol. 2 (1975) 9. S. C. Harvey in
“Benzodiazepine Hypnotics in the Goodman Gilman, The Pharmacological Basis
Multidimensional Treatment of Insomnia,” of Therapeutics, 7th ed., Macmillan Publ.,
Med. Actual. 22 (1986) 271 – 281. C. Pina, A. New York, Toronto, London 1985, p. 352; see
Catedra: “Newer Benzodiazepines, Med. also [102] p. 82.
Actual. 11 (1975) 176; “Chemical Synthesis of 112. Hoffmann-La Roche, DE-AS 1 136 709, 1959;
Benzodiazepines (nur Literatur),” Med. DE-AS 1 145 626, 1959; US 3 109 843, 1961;
Actual. 11 (1975) 211 – 212. H. Schütz: US 3 116 203, 1962; US 3 123 529, 1964.
Benzodiazepines: A Handbook, Springer Delmar Chem., DE-AS 1 811 785, 1968.
Verlag, Berlin 1982. C. Bellantuono, V. Reggi, 113. L. H. Sternbach et al., J. Med. Chem. 6 (1963)
G. Tognoni, S. Garattini: “Benzodiazepines – 261. Hoffmann-La Roche, US 3 116 203,
Clinical Pharmacology and Therapeutic Use,” 1963; US 3 123 529, 1964; US 3 203 900,
Drugs 19 (1980) 195 – 219. S. Garattini: The 1964; GB 1 039 372, 1966.
Benzdiazepines, Raven Press, New York 1973. 114. M. W. Johns, Drugs 9 (1975) 448. J. Rieder,
103. M. Lader (ed.): “New Perspectives in Arzneim. Forsch. 23 (1973) 212. Martindale:
Benzodiazepine Therapy,” Arzneim. Forsch. The Extrapharmacopoea, 28th ed.,
30 (1980) no. 5 a, 851 – 913. U. Klotz: Pharmaceutical Press, London 1982,
“Clinical Pharmacology of Benzodiazepines,” pp. 807 –809. Med. Actual. 1 (1964)
in Progress in Clinical Biochemistry and 132 – 138; 9 (1973) 237.
Medicine, vol. 1, Springer, 115. J. M. Monti, H. Altier, Psychopharmacologia
Berlin – Heidelberg – New York – Tokyo 1984, 32 (1973) 343. W. Schallek, J. Kovacs, A.
pp. 117 – 166. Kuehn, J. Thomas, Arch. Int. Pharmacodyn.
104. H. Schrum: “Benzodiazepine als Schlafmittel,”
Ther. 206 (1973) 161. S. A. Kaplan et al., J.
Schweiz. Rundsch. Med. Prax. 74 (1985)
Pharm. Sci. 63 (1974) 527. R. M. Joye, A. J.
1301 – 1304.
Hance, K. F. Killam, Neuropharmacology 10
105. R. Weber, Dtsch. Apoth. Ztg. 128 (1988)
(1971) 483. J. Freuchen, J. Østergaard, B. O.
2643 – 2652.
106. L. L. Iversen, Nature (London) 275 (1978) Mikkelsen, Curr. Ther. Res. 23 (1978) 90.
447. R. E. Study, J. L. Barker, JAMA J. Am. R. T. Owen, Med. Actual 19 (1983) 163 – 167;
Hypnotics 31
(1977) 1378. Med. Actual. 17 (1981) 147. A. Walser et al., J. Org. Chem. 43 (1978) 936.
283 – 286. P. J. Thorpe, Drugs Future 6 (1981) Hoffmann-La Roche, DE-OS 2 540 522, 1976.
414 – 415. 148. M. Gemper et al., Arzneim. Forsch. 31 (1981)
137. J. B. Hester, D. J. Duchamp, G. C. Chichester, 2177 – 2288. D. A. C. Hegelbach-Feller, J. M.
Tetrahedron Lett. 1971, 1609. J. B. Hester, Tschopp, S. Christeller, J. Fabre, Arzneim.
A. D. Rudzick, B. V. Kamdar, J. Med. Chem. Forsch. 38 (1988) 387. J. Kanto, Med. Actual,
14 (1971) 1078. K. Meduro, H. Tawada, Y. 19 (1983) 376 – 381.
Kuwada, Chem. Pharm. Bull. 21 (1973) 1619, 149. K. H. Weber, W. Sirrenberg, O. Spohn, H.
2382. K. Kamiya, Y. Wada, M. Nishikawa, Daniel, Arzneim. Forsch. 36 (1986) 518.
Chem. Pharma. Bull. 21 (1973) 1520. K. Boehringer, DE-OS 2 410 030,
Megruo, Y. Kuwada, Tetrahedron Lett. 1970, 1974 = US 4 094 984, 1975; DE-OS 2 503 235,
4039. D. L. Loffen, R. J. Fryer, D. A. Katonak, 1975; DE-OS 2 533 924, 1975. Hoffmann-La
F. Wong, J. Org. Chem. 40 (1975) 894. Roche, DE-OS 2 221 623, 1971. K. Gewald,
138. Upjohn, DE-OS 2 012 190, 1969; Chem. Ber. 98 (1965) 3571.
US 3 701 782, 1969; DE-OS 2 302 525, 1973; 150. N. Ferrara et al., Curr. Ther. Res. 37 (1985)
US 4 018 788, 1977; US 3 987 052, 1969. 295. J. Grünberger et al., Curr. Ther. Res. 24
Toyama, DE-OS 1 955 349, 1971. Takeda, (1978) 427. B. Saletu, J. Grünberger, J.
DE-OS 2 114 441, 1971; DE-OS 1 965 894, Volavka, P. Berner, Arzneim. Forsch. 29
1969; US 4 116 956, 1976. (1979) 700. M. S. Langley, S. P. Clissold,
139. S. Tanayama, Y. Shirakawa, Y. Kanai, Z. Drugs 35 (1988) 104 – 122. Pharmcast 87-10,
Suzuoki, Xenobiotika 4 (1974) 33. Y. Kanai, Technomic Information Service, Tokyo 1987,
Xenobiotika 4 (1974) 441. F. S. Eberts, pp. 6-II-151 – 6-II-152.
Pharmacologist 16 (1974) 196; Drug Metab. 151. R. Agar et al., J. Med. Chem. 20 (1977) 1035.
Dispos. 5 (1977) 547. H. Isozaki, M. Tanaka, Russel-UCLAF, DE-OS 2 605 652, 1976; US
K. Inanaga, Curr. Ther. Res. 20 (1976) 493. T. 4 044 142, 1977.
Momose, S. Shii, T. Kuge, Curr. Ther. Res. 19 152. T. G. Johns, D. C. Piper, G. W. L. James, Arch.
(1976) 277. W. Ehrenstein, M. Int. Pharmacodyn. Ther. 240 (1979) 53. J.
Müller-Limmroth, M. Opfermann, F. Weber, Hindmarch, C. A. Clyde, Drugs Exp. Clin.
Arzneim. Forsch. 32 (1982) 280. J. K. Walsh Res. 2 (1980) 61. R. T. Owen, Med. Actual. 20
et al., Curr. Ther. Res. 36 (1984) 866. Med. (1984) 59 – 63. Loprazolam Clinical
Actual. 12 (1976) 355 – 357. Annual Drug Workshop, The Medicine Publishing
Data Report 2 (1979/1980) 85 – 86. Foundation, Oxford, 1983.
140. J. B. Hester, A. D. Rudzick, B. V. Kamdar, J. 153. K. Hirai et al., Heterocycl. Chem. 19 (1982)
Med. Chem. 14 (1971) 1078. 1363; J. Med. Chem. 23 (1980) 764; 24 (1981)
141. A. Kales, M. B. Scharf, J. D. Kales, Science 20. Shionogi, DE-OS 2 725 164, 1977;
(Washington, D.C.) 201 (1978) 1039. US 4 159 374, 1977. Hoffmann-La Roche,
142. T. Roth, M. Kramer, T. Lutz, Drugs Exp. Clin. DE-OS 2 537 069, 1975.
Res. 1 (1977) 279. D. A. C. Hegelbach-Feller, 154. K. Hirose, A. Matsushita, M. Eigyo, Arzneim.
J. M. Tschopp, S. Christeller, J. Fabre, Forsch. 31 (1981) 63. Drugs Future 9 (1984)
Arzneim. Forsch. 38 (1988) 387. 522 – 524; 11 (1986) 626 – 628. Pharmcast
143. M. Leibowitz, A. Sunshine, J. Clin. 87-1, Technomic Information Service, Tokyo
Pharmacol. 18 (1978) 302. K. K. Okawa, 1987, pp. 5-III-139 – 5-III-141.
G. S. Allen, J. Int. Med. Res. 6 (1978) 343. 155. Rhône-Poulenc, DE-OS 2 300 491, 1973; US
B. B. Knap, E. L. Boyd, G. Linsenmeyer, O. J. 3 862 149, 1975. C. Jeanmart, C. Cotrel, C. R.
Linet, Curr. Ther. Res. 23 (1978) 230. Ser. C 287 (1978) 377.
144. A. N. Sing, Curr. Ther. Res. Clin. Exp. 27
156. R. Duriez, Therapie 34 (1979) 317. E.
(1980) 627. A. N. Nicholson, Br. J. Clin.
Wickström, K. E. Giercksky, Eur. J. Clin.
Pharmacol. 9 (1980) 223.
Pharmacol. 17 (1980) 93; Clin. Therap. 3
145. L. Lasagnon, Lancet 1 (1980), 815.
(1978) 59. . J. C. Blanchard, A. Boireau, C.
146. E. A. Swinyard in Remingtons Pharmaceutical
Garrett, L. Julou, Life Sci. 25 (1979) 2417.
Sciences, 17th ed., Mack Publ., Easton, Pa.
P. H. Wu, J. W. Phillis, A. S. Bender, Life Sci.
1985, p. 1064. AMA: Drug Evaluation, 6th
28 (1981) 1023. J. Castaner, Drugs Future 4
ed., Am. Med. Assoc., Chicago 1986, p. 103. J.
(1979) 59 – 60; 5 (1980) 56; 6 (1981) 62 – 63;
Castaner, Drugs Future 1 (1976) 393 – 396; 2
7 (1982) 69 – 70; 8 (1983) 79; 9 (1984) 81; 10
(1977) 558.
Hypnotics 33
(1985) 88 – 89; 11 (1986) 76 – 78; 12 (1987) 158. Synthelabor, EP 50 563, 1981. P. George, H.
95; 13 (1988) 88 – 89. Pharmacast 87-7, Depoortere, J. Pharm-Sci. (in Press).
Technomic Inf. Service, Tokyo 1987, 1989, 159. S. Arbilla et al., Naunyn-Schmiedebergs Arch.
6-I-87 – 6-I-89. Int. Pharmacopsychiatry 17 Pharmacol. 330 (1985) 248. M. Dimsdale et
(1982) (Suppl. 2), 1 – 252. K. A. Goa, R. C. al., Drugs Future 12 (1987) 777 – 780.
Heel, Drugs 32 (1986) 48 – 65. 160. M. Dimsdale et al., Drugs Future 13 (1988)
157. G. Houghton et al., Int. J. Clin. Pharmacol. 106 – 109.
Biopharm. 23 (1985) 97. R. Duriez et al., 161. J. Maillard, M. Vincent, M. Bernard, Bull. Soc.
Therapie 34 (1979) 317. J. Gaillot et al., Chim. Fr. 1961, 529; 1962, 1580.
Pharmacology 27 (1983) (Suppl. 2) 76 – 91. Logeais-Labs., GB 902 388, 1962;Fr. M. 379,
1962; Chem. Abstr. 57 (1962) 15251 g.
Hypnotics
Hartmund Wollweber, Bayer AG, Wuppertal, Federal Republic of Germany
who display larger differences (>1 ◦ C) are long when alterations in thinking or behavior are nec-
sleepers. Body tissues are regenerated during or- essary. Confrontation with problems also inten-
thodox sleep, and cerebral functions during the sifies REM sleep, e.g., an astronaut experiences
REM phase. greatly increased REM sleep in his first night
The REM phases account for ca. 20 – 25 % in space. Continuous suppression of REM sleep
of sleep and appear to be periods of relatively leads to psychological disturbances and psycho-
high cerebral activity in which dreams occur. somatic symptoms.
They are detected by measuring the spontaneous
movements of the closed eye. In the absence of
these dream phases sleep is not restful. Dreams 1.1. Amino Acids and Hormones
are of great importance for learning and mem- Influencing Sleep
ory; they assimilate the experiences of the pre-
ceding days. Generally, dreams can be remem- Three neurohormones are involved in control-
bered slightly or not at all. With increasing age, ling sleep: serotonin, noradrenaline, and acetyl-
the duration of REM sleep steadily decreases choline. Serotonin regulates orthodox sleep,
to about 20 – 30 % of its original duration. The whereas acetylcholine and noradrenaline influ-
REM phase is especially activated when a per- ence the REM phase. The rates of cell division
son is confronted with an emotionally arousing, are the highest during sleep, and large amounts
intellectual task calling for creative effort, i.e., of growth hormone (GH), which stimulates pro-
Hypnotics 3
tein synthesis, are secreted during the third and all, because of their many adverse effects. Alco-
fourth levels of orthodox sleep. The EEG waves hol should not be taken with sedatives or hyp-
at this stage are also large and slow. notics because the reactions of the patient are
Melatonin is rhythmically secreted by the then adversely affected.
pineal gland. It stimulates specific receptors in Prolonged use of sedatives and hypnotics re-
the body’s biological clock, the suprachiasmatic duces their effect on the cells of the central ner-
nuclei of the hypothalamus and may thus regu- vous system (see also Section 5.1). Most hyp-
late the human circadian rhythm by increasing notics, particularly highly active preparations,
serotonin levels. The human biological clock can can create psychological and physical depen-
be “adjusted” with exogenous melatonin. Conti- dence, resulting in withdrawal symptoms when
nous exposure of humans to light leads to sleep treatment is terminated. In addition, many hyp-
disorders by inhibiting melatonin synthesis [16], notic drugs that suppress and shorten the REM
[29], [30]. phase produce a rebound effect, i.e., a sudden
A sleep inducing nonapeptide, delta-sleep- withdrawal after extended periods of use causing
inducing peptide [62568-57-4] (DSIP), has been a prolongation of the REM phase. As in alcohol
isolated from rabbit serum. This peptide, which withdrawal, this effect can even lead to delirium
also occurs in humans, has been synthesized tremens. In this way, the body tries to make up
and has been prepared by genetic engineering for lost REM sleep.
techniques. DSIP coordinates the rhythmic cir-
cadian sleep – wake cycles by promoting sleep
functions (without having a sedative effect), and 1.3. Requirements of Hypnotics
enhances mental performance when awake. At
the same time, DSIP invokes a higher tolerance There is still no completely satisfactory hypnotic
to stress. The action of this nonapeptide is at- drug. The following requirements can be defined
tributed to modulation of the adrenergic recep- for an ideal hypnotic:
tors [31].
l-Tryptophan, a common amino acid found – reliable effect
in the human body is a precursor of serotonin. In – wide therapeutic range
daily doses of 4 – 7 g it causes sedation during – limited duration of action, no longer than 8 h
the day and a shortening of latency of sleep at – physiological sleep
night. It is registered in Switzerland as a “bio- – no disturbance of deep sleep or REM sleep
logical sleep-inducing drug.” However, studies – low toxicity
have questioned the sleep-modulating effect of – no enzyme induction
this amino acid [32]. – no respiratory depression
– no additive effect with alcohol
– no amnesia, confusion, or ataxia after arousal
1.2. Influence of Hypnotics on Sleep in the night
– no excessive daytime sedation
Most hypnotic drugs shorten both phases of – no loss of potency on repeated use
sleep and suppress the paradoxical (REM) – no risk of dependency, e.g., insomnia after
phase. Thus, the sleep curve is levelled out, and termination of treatment
the EEG resembles that of a narcotized per- – no after effects (e.g., hangover)
son. Hence, drug-induced sleep does not have
the restorative quality of natural sleep and is
no substitute for the treatment of insomnia. A 1.4. Applications of Hypnotics
research report of the Max Planck Institut für
Psychiatrie (München, 1980) states that conven- Sleep-inducing drugs are used by the following
tional sleep-inducing drugs do not produce rest- groups:
ful sleep; since drug-induced sleep is not iden-
tical to natural sleep, taking sleeping pills is not 1) healthy people with demanding jobs or irreg-
advised. Pharmacologists also recommend the ular working hours, who often take sleeping
use of hypnotic drugs only for a few days, if at pills after stimulants such as coffee;
4 Hypnotics
2) sick people with symptoms such as fever, 3) other drugs with sedative effects, (e.g.,
sore throat, or cold; tranquilizers, antidepressants, and antihis-
3) old people with insomnia due to cerebral tamines); and
sclerosis usually have an inverted day – night 4) natural preparations, e.g., valerian, ethanol,
rhythm; they are awake and mobile at night, and DSIP.
and tired and peaceful in the morning.
4) sick people with uncomfortable and sleep- There is no sharp distinction between seda-
less nights due to pain, who have difficulty tives and hypnotic drugs; differences in activity
in breathing, heart trouble, etc.; and are quantitative rather than qualitative. Both in-
5) bedridden people with serious illnesses, who hibit the activation system in the reticular for-
are deprived of physical exercise. mation. Drugs belonging to groups 1 and 2 are
true hypnotics. Drugs belonging to group 3 are
People belonging to groups 1 and 2 can be only occasionally used as hypnotics.
treated with mild relaxants by reducing the sus- Der Bewertende Arzneimittelindex (The Val-
ceptibility of the reticular formation (site of the uating Drug Index) assesses “benzodiazepine
center of wakefulness in the brain) to external sleep-inducing drugs” positively, especially be-
stimuli. These substances include plant compo- cause of their wide therapeutic range [37]. Other
nents with sedative properties, such as valerian recommended drugs are chloral and its deriva-
and hop, and tranquilizers or sedatives which, tives (see Sections page 7page 7), diphenhy-
even in high doses, do not have narcotic prop- dramine, and doxylamine (see Sectionpage 5).
erties. People belonging to groups 4 and 5 may Derivatives of barbituric acid, piperidinediones,
have to be treated for extended periods with hyp- and quinazolinones are not recommended. In
notics and even with antipsychotic and antide- comparison with the benzodiazepines, these
pressant drugs. cause more undesirable effects (respiratory de-
pression, negative influence on REM sleep, en-
zyme induction, and problems of dependence).
1.5. Consumption of Hypnotics Their therapeutic range is narrow. Compounds
containing bromine are also to be avoided be-
The importance of hypnotic drugs is reflected in cause they may cause bromism (see Chap. 2).
the fact that 21 % of the population of France Valerian preparations (e.g., Valepotriate) and
(February 1975) and 32 % of the inhabitants of their components are not favored because of
Los Angeles (1973) suffered from insomnia. In their in vitro cytotoxic and alkylating action. The
France, 2.5×106 tablets, capsules, and dragées therapeutic importance of these drugs cannot be
are taken every evening to promote sleep [33– assessed at present [37].
38]. The dramatic increase in the consumption The lists of trade names cited for individual
of hypnotics with age can only be partly justi- hypnotics are not complete.
fied because older people require less sleep at
night. Instead, they have a more pronounced en-
dogenous sleep profile with increased cerebral 2. Sedatives, Antihistamines, and
readiness to sleep at 9.00, 13.00, and 17.00 h. Tranquilizers
Sedatives and Tranquilizers. There is no
1.6. Classification and Evaluation of significant distinction between sedatives and
Hypnotics hypnotics; the same drugs are often used for both
purposes. Sedatives (→ Sedatives) and tranquil-
Hypnotics may be classified as follows [34]: izers (→ Psychopharmacological Agents) are
used to sedate and relieve anxiety during the day
1) benzodiazepine hypnotics; and, in addition, to induce sleep at night. Thera-
2) nonbenzodiazepine hypnotics, e.g., chloral peutic doses of tranquilizers have a calming, re-
hydrate, barbiturates, piperidinediones, and laxing effect and alleviate psychological symp-
methaqualone; toms such as fear, anxiety, restlessness, and ten-
sion. Ideally, they should not produce sedative
Hypnotics 5
or hypnotic effects. The first therapeutically used Trade Names: Calmonal (Heiden, FRG), Bonine
hypnotics were plant extracts with either seda- (Pfizer, USA).
tive (valerian root and hop) or analgesic action Diphenhydramine hydrochloride [147-24-0],
(opium alkaloids). In spite of the sedative effect 2-(benzhydryloxy)-N,N-dimethylethylamine
of opium alkaloids [39], they should not be used (→Antiallergic Agents)
for obvious reasons (addiction, euphoria, poor Trade Names: Benadryl (Parke-Davis, USA),
dose adjustment, and legality). Of all the plant Halbmond (Much, FRG).
extracts available, only one substance, valmane, Doxylamine succinate [562-10-7], 2-di-
is exactly defined from a chemical standpoint. methylaminoethoxyphenylmethyl-2-picoline
Valepotriate is a mixture of valtratum Trade Names. Decapryn, Meteprine (Merrell,
[18296-44-1] (1), acevaltratum [25161-41-5] USA), Unisom (Pfizer, USA).
(2), and didrovaltratum [18296-45-2] (3):
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.
The sodium salt, triclofos sodium [7246-20-0], These compounds are mild hypnotic agents.
C2 H3 Cl3 NaO4 P, is soluble in water. This prepa- They are generally used to promote the onset
ration has the same physiological properties as of sleep or as daytime sedatives, often in com-
chloral hydrate, but it does not have the unpleas- bination with weak analgesics. They are well-
ant physical characteristics. The compound is absorbed and, depending on the structure, have
metabolized in the body mainly to the hypnot- few side effects. Dependence on these com-
ically active 2,2,2-trichloroethanol. A dose of pounds occurs less frequently than on barbitu-
1.5 g is equivalent to 1 g of chloral hydrate. rates and benzodiazepines.
Trade Names. Triclos (Lakeside, USA), Triclo- Urethane [51-79-6], ethyl carbamate, a mild
syl (Glaxo, UK), Tricloran (C.T.S. Israel). hypnotic agent, is no longer prescribed because
Petrichloral [78-12-6], pentaerythritchloral, its cytostatic properties can cause agranulocyto-
C13 H16 Cl12 O8 , M r 725.76, mp 52 – 54 ◦ C [46]. 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
Chloralodol [3563-58-4], chlorhexadol, (→Psychopharmacological Agents).
2-methyl-(2 ,2 ,2 -trichloro-1 -hydroxy-4- Ureides are still widely prescribed as mild
ethoxy)-pentan-2-ol, C8 H15 Cl3 O3 , M r 265.58, sedatives and hypnotic agents. They are used
mp 102 – 124 ◦ C [47]. 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
Trade Names. Lora (Wallace, USA), Mechloral, sleep.
Mecloral (Dumex, Denmark).
Chloralose [15879-93-3], α-O-(2,2,2-tri- Ethinamate [126-52-3], 1-ethynylcyclo-
chloroethylidene)-1,2-α-d-glucofuranose-(R), hexylcarbamate, C9 H13 NO2 , M r 167.21,
C8 H11 Cl3 O6 , M r 309.54, mp 187 ◦ C, [α]22
D + 19 mp 96 – 98 ◦ C, solubility in water 0.25 %, in
(5 % solution in 98 % ethanol) [48]. ethanol 35 %, and
Trade Names. Valamin (Schering, FRG), broken down to bromide. The same applies to
Valmid, Valmidate (Lilly, USA). acecarbromal and bromisoval.
Trade Names. Benuride (Bengue, UK; Vinas, diacylated cyclic ureas: barbiturates (8) and
Spain), Deturid (Polfa, Poland; Sapos, Switzer- thiobarbiturates (9)
land). N-(β-carbonyl)ethyl derivatives: 2,4-piperi-
dinediones (6)
Capuride [5579-13-5], 2-ethyl-3- N-iminoacyl derivatives: quinazolinones (7)
methylvalerylurea, C9 H18 N2 O2 , M r 186.25, phenyl N-iminoacyl derivatives: benzodiaze-
mp 172 ◦ C [60]. pines (11)
Trade Name. Pacinox (McNeil, USA).
All groups, except 5, 7, and 11, have a doubly
substituted carbon atom adjacent to the carbonyl
Valdettamil, Novonal [512-48-1], 2,2-
group of the amide structure. These substituents
diethyl-4-pentenamide, C9 H17 NO, M r 155.23,
greatly influence the action and the distribution
mp 75 – 76 ◦ C, is soluble in 120 parts of water
of the drug in the body.
and freely soluble in alcohol and ether [61].
Trade Names. Insomnia (ICN, USA), and
as combinations: Arantil (Hoechst, FRG),
Dentigoa (Scheurich, Switzerland).
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].
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
countries. Important properties are listed in Ta- The duration of action is primarily deter-
ble 1. As hypnotics, barbituric acids all pro- mined by the rate of enzymatic degradation in
duce almost identical pharmacodynamic ef- the liver, which depends on the chemical struc-
fects. However, they show wide differences ture of the barbiturate. Continual use causes en-
in pharmacokinetics and metabolism. Differ- zyme induction in the liver and thus a gradual
ences in onset and duration of activity are re- increase in the rate of barbiturate degradation.
lated to lipid and water solubility, and rate Increasing doses are required with prolonged
of metabolism. Lipophilic barbiturates, e.g., use (barbiturate habituation). In some patients
the narcotically active, lipid-soluble thiobarbitu- barbiturates not only have a sedative and hyp-
rates, and the N-monoalkylbarbiturates such as notic action, but also produce excitement and
hexobarbital rapidly disappear in adipose tissue. euphoria. This sense of well-being may easily
As a result of their redistribution in lipophilic lead to barbiturate addiction, a physical depen-
compartments of the body, these compounds dence with withdrawal symptoms. Barbiturate
have a very short duration of action and are addicts take the hypnotic because of its euphoric
rapidly broken down (short-acting narcotics, properties and try to compensate for the central
→Anesthetics, General, Chap. 3.1.). depression by using psychoanaleptic drugs.
From a therapeutic standpoint, barbiturates Cases of accidental and suicidal barbiturate
used as hypnotic agents are classified according poisoning are fairly common. They are charac-
to their duration of action (see also Table 1): terized by unconsciousness (which can precede
a delirious stage), central respiratory depression
1) long-acting compounds (barbital, phenobar- accompanied by oxygen deficiency symptoms,
bital), and in severe poisoning circulatory collapse with
2) medium-acting compounds (amobarbital, cy- death in 12 h to 4 d. The prognosis for barbitu-
clobarbital), rate poisoning is favorable if treatment is imme-
3) short-acting compounds (hexobarbital), and diate. Essential measures include artificial res-
4) compounds whose effects begin after 15 – piration, irrigation of the stomach, circulatory
20 min (secobarbital).
Hypnotics 11
support, forced osmotic diuresis using manni- Trade Names. Dormileno (Faes, Spain), Hyp-
tol, exchange transfusion, and hemodialysis. nox (Püschel, Austria), Veronal (Bayer, FRG;
Barbiturates produce a condition superfi- Merck, FRG), Veroletten (EAS-Labor, Austria).
cially resembling natural sleep. However, the Sodium salt: Barbinetten (Hormosani, Aus-
REM phase, which is very important for rest tria).
and recuperation, is shortened. Continued use
of barbiturates leads to a very low REM value, Phenobarbital [50-06-6], phenemalum,
and sudden withdrawal increases the REM phase phenobarbitone, 5-ethyl-5-phenylbarbituric
(rebound effect). The adverse effects of barbitu- acid, C12 H12 N2 O3 , M r 232.23, mp 174 –
rates are well-documented after more than 50 178 ◦ C. The acid crystallizes from water in
years of experience. This is an important advan- the form of flakes. It is odorless, with a slightly
tage over newly developed hypnotics [68]. bitter taste; an aqueous solution reddens litmus
The pK values of the hypnotically active bar- paper. One part dissolves in 1100 parts of water
bituric acids, range between 7.78 and 8.30. Dis- at 20 ◦ C, 40 parts of boiling water, 10 parts of
ubstituted barbituric acids are present, to ca. ethanol, 18 parts of diethyl ether, and 60 parts
50 %, in the undissociated form under physio- of chloroform. It is synthesized by the reaction
logical conditions (pH 7.4). Only this form of the of urea with ethyl phenyl malonate in sodium
drug can pass through biological membranes. ethoxide solution [71].
The degree of dissociation is strongly influenced The sodium salt of phenobarbital [57-30-7],
by slight alterations in pH. C12 H11 N2 NaO3 , M r 254.23, is a colorless, crys-
Production. The 5,5-disubstituted barbituric talline, hygroscopic, freely water-soluble pow-
acids are still synthesized by the classical con- der. It is slightly soluble in ethanol, and insolu-
densation of substituted malonic esters (12), ble in diethyl ether and chloroform. The aqueous
malonic amides, or cyanoacetic esters (13) with solution is stable for only a short time.
urea (14), thiourea (15), guanidine (16), or di- Phenobarbital is a long-acting hypnotic with
cyanodiamide (17), followed by the hydrolysis anticonvulsant properties. Long-term use in
of the resulting imino- or cyanobarbiturate [69] higher than therapeutic doses can cause phys-
(Fig. 3). ical addiction. An overdose resulting in blood
Under acidic or neutral conditions, substi- levels of 8 – 12 mg/100 mL can cause death.
tuted malonic acids or malonic chlorides can Trade Names. Agrypnal (Eggochemia, Austria),
also be used for the condensation with urea Aphenylbarbit (Streuli, Switzerland), Calminal
or thiourea. It is preferable for the substituents (Wolfs, Belgium), Epsylone (Powell, Canada),
R1 and R2 to be present in the malonic ester Eskabarb Span (Smith, Kline & French, USA),
or cyanoacetic ester molecule; disubstitution of Fenemal (DAK, Denmark), Gardenal (Spe-
barbituric acids at the 5-position is only possi- cia, Canada), Gardenale (Farmitalia Carlo
ble with highly reactive halides, such as allyl Erba, Italy), Hypnolone (Hartz, Canada),
halides. Pepinal, Lepinaletten (Arzneimittelwerk Dres-
den, GDR), Luminal (Bayer, FRG; E. Merck,
Barbital [57-44-3], barbitone, 5,5- FRG; Winthrop, USA), Mediphen (Medic,
diethylbarbituric acid, C8 H12 N2 O3 , M r 184.19, Canada), Phen Bar (Saunders, Canada), Phe-
mp 188 ◦ C, (sodium salt [144-02-5], nobarbyl (Synochem, FRG), Seda-Tablinen
C8 H11 N2 NaO3 , Mr 206.18), has a bitter taste. (Beiersdorf, FRG), Solfoton (Poythress, USA),
One part of the acid dissolves in 170 parts of Teolaxin (Paul Maney, Canada); Phenobarbi-
water at 25 ◦ C or 17 parts of water at 100 ◦ C. tal diethylamine salt: Gratusminal (Farmasimes,
It is freely soluble in acetone, ammonia, and Spain); Phenobarbital calcium salt: Fenileal
lye. Barbital is prepared by the condensation (Turon, Spain), Lumcalcio (Abello, Spain).
of diethyl malonate with urea in the presence
of sodium ethoxide [70]. Both the acid and its Butobarbital [77-28-1], butobarbitone, 5-
sodium salt are used as long-acting hypnotics; butyl-5-ethylbarbituric acid, C10 H16 N2 O3 ,
abuse can cause habituation or addiction. 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
acting hypnotic; abuse can cause habituation or in water; preparation is described in [74], and
addiction. metabolism in [75].
Trade Names. Butenil, Butynoct (Interpharm, Amobarbitol is a widely used medium- to
Austria), Etoval, Longanoct, Sonabarb (Protea, long-acting hypnotic. It resembles secobarbi-
Australia), Soneryl (Specia, Italy; May & Baker, tal and pentobarbital in action with a some-
USA). what longer duration. A blood level of 3 –
6 mg/100 mL of blood is fatal; abuse can cause
Probarbital [76-76-6], 5-ethyl-5-isopro- habituation or addiction.
pylbarbituric acid, C9 H14 N2 O3 , M r 198.22, Trade Names. Amsal (Adams, Australia), Amy-
mp 197 – 198 ◦ C, is slightly soluble in cold cal (AFI, Sweden), Amydorm, Amytal (Lilly,
water, but more soluble in hot water and eth- USA), Isobec (Pharbec, Canada), Isonal (ICN,
anol. Sodium salt [143-82-8], C9 H13 N2 NaO3 , Canada), Neur-Amyl (Fawns & McAllen, Aus-
M r 220.20, is a hygroscopic powder that is solu- tralia), Placidel (Miquel, Spain). Stadadorm
ble in water. The calcium trihydrate [545-74-4], (Stada, FRG).
C18 H26 CaN4 O6 · 3 H2 O, M r 488.55, forms Amobarbital sodium: Altinal, Amsal
slightly bitter crystals; 1 g dissolves in 40 g wa- (Adams, Australia), Amsebarb (Paul Maney,
ter; preparation is described in [73]. Probarbital Canada), Amylbarb Sodium (Protea, Aus-
is a long- to medium-acting hypnotic; abuse can tralia), Amylobeta (Bramble, Australia), Amy-
cause habituation or addiction. tal Sodium (Lilly, USA), Barbamyl (Teva,
Trade Names. Ipral Sodium, Ipral Calcium Israel), Dorminal (Boots, UK), Neur-Amyl
(Squibb, USA). (Fawns & McAllen, Australia), Novamobarb
(Novopharm, Canada).
Amobarbital [57-43-2], 5-ethyl-5-isopen-
tylbarbituric acid, C11 H18 N2 O3 , M r 226.28, mp Cyclobarbital [52-31-3], cyclobarbitone,
156 – 158 ◦ C, forms slightly bitter crystals; 1 g 5-(1-cyclohexen-1-yl)-5-ethylbarbituric
dissolves in 1300 mL water and in 5 mL ethanol. acid,C12 H16 N2 O3 , M r 236,26, mp 171 –
The sodium salt, C11 H17 N2 NaO3 [64-43-7], 174 ◦ C, forms bitter crystals and is moderately
M r 248.25, is hygroscopic and readily soluble soluble in water. The calcium salt [143-76-0],
C24 H30 CaN4 O6 , M r 510.59, is prepared by heat-
Hypnotics 13
Trade Names. Fabadorm (Bayer, FRG), Phan- Heptabarb [509-86-4], heptabarbital, hept-
odorm (Bayer, FRG; Merck, FRG; Winthrop, abarbitone, 5-ethyl-5-(1-cycloheptenyl)barbitu-
USA), Phanotal (Kwizda, Austria). ric acid, C13 H18 N2 O3 , M r 250.29, mp 174 ◦ C;
one part dissolves in 13 parts of methanol,
Aprobarbital [77-02-1], 5-allyl-5-isopro- 20 parts of ethanol, 13 parts of acetone, 40
pylbarbituric acid, C10 H14 N2 O3 , M r 210.23, parts of diethyl ether, and 5000 parts of wa-
mp 140 – 141.5 ◦ C, forms slightly bitter crystals, ter. It is synthesized by the condensation of
and is insoluble in water, but soluble in ethanol. ethyl(cyclohepten-1-yl)-cyanoacetic ester with
The sodium salt [125-88-2], C10 H13 N2 NaO3 , urea [81].
M r 232.21, a hygroscopic, bitter powder, is sol- This compound is a short- to medium-acting
uble in water. Preparation is described in [77]. hypnotic; abuse can cause habituation or addic-
This compound is a medium-acting hypnotic; tion.
abuse can cause habituation or addiction. Trade Names. Medomin, Medapan (Ciba-
Trade Names. Alurate (Hoffmann-La Roche, Geigy, Switzerland).
Switzerland), Numal (United Chemicals, USA).
Secbutabarbital [125-40-6], secbutobarbi-
Allobarbital [52-43-7], allobarbitone, tal, butabarbital, secbutobarbitone, 5-ethyl-5-(2-
5,5-diallylbarbituric acid, C10 H12 N2 O3 , butyl)barbituric acid, C10 H16 N2 O3 , M r 212.24,
M r 208.21, mp 171 – 173 ◦ C, is slightly bitter. mp 165 – 168 ◦ C; the sodium salt [143-81-7],
One part dissolves in 300 parts of cold water C10 H15 N2 NaO3 , M r 234.23, is a bitter powder;
and in 50 parts of boiling water. It is prepared 1 g dissolves in 2 mL of water, and the pH of a
by heating diallyl malonate with urea in sodium 1 % aqueous solution is 9.0 – 10.2. Preparation is
ethoxide. It can also be synthesized by the re- described in [82]. Secbutabarbital is a medium-
action of barbituric acid with allyl bromide in to long-acting hypnotic. It is widely used in the
the presence of sodium acetate or copper sulfate United States as a sedative. Blood levels of 3.5 –
[78], [79]. It is a mediumacting hypnotic; abuse 6 mg/100 mL are fatal; abuse can cause habitu-
can lead to habituation or addiction. ation or addiction.
Trade Names. Diadol (Durst, USA), Dial (Ciba-
Geigy, Switzerland); it is contained in Spasmo-
Cibalgine (Ciba, Switzerland).
5.2. Piperidinediones
with codeinelike, cough-suppressant, analgesic way), Sovinal (ND & K, Denmark), Toraflon
properties; they are useful in patients who do not (Toraude, Canada), Tualone (ICN, Canada), Tu-
tolerate or respond to barbiturates. Prolonged azol (Strasenburgh, USA).
use may cause habituation and psychological de-
pendence. No specific antagonists are available Etaqualone [7432-25-9], 3-(2-ethylphenyl)-
for quinazolinone poisoning. 2-methyl-4(3 H)-quinazolinone, C17 H16 N2 O,
Mr 264.31, mp 81 ◦ C. Hydrochloride
Methaqualone [72-44-6], 2-methyl-3- [97979-65-2], C17 H17 ClN2 O, mp 247 ◦ C.
o-tolyl-4-(3 H)quinazolinone, C16 H14 N2 O, Etaqualone is prepared by the condensation
M r 250.30, mp 120 ◦ C (also 114 – 116 ◦ C), is of N-acetylanthranilic acid with 2-ethylaniline
soluble in ethanol, diethyl ether, and chloro- in the presence of phosphoryl chloride [99].
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. Trade Name. Aolan (Beiersdorf, FRG).
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 tative differences could be of importance. Stud-
are constantly being marketed. Properties of the ies point to an interaction between benzodiaze-
most important benzodiazepines are listed in Ta- pines and their receptors. Thus, they activate the
ble 2. postsynaptic receptors for γ-aminobutyric acid
(GABA, an inhibitory neurotransmitter) and in-
Mode of Action. The benzodiazepines have crease the frequency of the opening of the chlo-
the same hypnotic action as the barbiturates, ride ion canals [106].
with differences in selectivity, metabolism, and In other words, the benzodiazepines act
adverse effects. Large doses do not possess the in the central nervous system by interfering
narcotic properties exhibited by most barbitu- with GABA-mediated neurotransmission [107],
rates. Benzodiazepines do not solely have a [108]. Traces of benzodiazepine derivatives
sleep-inducing action; they exert a combination (e.g., diazepam and lorazepam) have been found
of effects: sedation, relief of anxiety (produc- in wheat and potato and are identical to the syn-
tion of a better state of mind), muscle relax- thetic substances. Benzodiazepines also occur
ation, and anticonvulsant action. These effects in untreated rats, as well as in human serum and
shorten both the intermittent waking times and brain. However, the concentration is very low
deep sleep (reduction of δ sleep). In addition, (several nanograms per gram of potato) and a
a hangover always occurs on the next day. The direct pharmacological effect on the central ner-
effects of barbiturates and benzodiazepines are vous system has not been demonstrated [108],
compared in Table 3. From a pharmacological [109].
standpoint, the hypnotically active benzodiaze-
pines have a depressant action on the electrical Metabolism. All benzodiazepines are me-
discharges in the areas controlled by the limbic tabolized in the body [110], but in some cases
system and block the function of the reticular the retention times are very long. In addition,
formation in the brain stem. They are similar to the pharmacologically active metabolites of the
the barbiturates in this respect; however, quanti- 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
Sweden), Uskan (Desitin, FRG), Vaben (Rafa, Carlo Erba, Italy; Aesca, Austria), Mabertin
Israel), Zapex (Riva, Canada). (Sidus, Argentina), Normison (Wyeth, UK),
Planum (Farmitalia Carlo Erba, Italy), Remestan
Lorazepam [846-49-1], 7-chloro-5-(2- (Wyeth, FRG), Restoril (Sandoz, Switzerland;
chlorophenyl)- 1, 3- dihydro- 3- hydroxy- 2H- Anca, Canada), Somaz (Quantum, USA), Tenso
1,4-benzodiazepin-2-one, (20b), (Castejon, Spain), Veroqual (Lek, Yugoslavia).
C15 H10 Cl2 N2 O2 , M r 321.17, mp 166 – 168 ◦ C,
is prepared from 18 (R = Cl, R1 = H, R2 = Cl), Doxefazepam [40762-15-0], 7-chloro-5-(2-
analogously to oxazepam. Lorazepam is pri- fluorophenyl)- 1, 3- dihydro- 3- hydroxy- 1- (2-
marily an anti-anxiety drug; it also has sleep- hydroxyethyl)-2H-1,4-benzodiazepin-2-one,
inducing properties and is used in anesthesiol- (20e), C17 H14 ClFN2 O3 , M r 348.57, mp 138 –
ogy [123]. 140 ◦ C, is prepared by the alkylation of 7-
Trade Names. Ativan (Wyeth, USA), Bonton chloro-5-(2-fluorophenyl)-1,3-dihydro-2H-1,4-
(Unipharma, Israel), Control (Sigurta, Italy), benzodiazepin-2-one N-oxide with 2-bromoeth-
Donix (Llorens, Spain), Emotion (Alpes, Ar- anol or 2-bromoethyl acetate, followed by
gentina), Emotival (Armstrong, Argentina), Lar- Polonowski rearrangement and saponification
pose (Cipla, India), Laubeel (Desitin, FRG), [128], [129]. Doxefazepam is a hypnotic drug
Loram (Lek, Yugoslavia), Lorans (Schiapar- with two to four times the activity and half the
elli, Italy), Lorax (Wyeth, UK), Lorivan (Disco, toxicity of flurazepam; it causes significantly
Israel), Merlit (Ebewe, Austria), Nervistop L less hangover [130].
(Ingram, Argentina), Orfidal (Orfi, Spain), Pro Trade Name. Doxans (Schiaparelli, Italy).
Dorm (Schürholz, FRG), Securit (Perrel, Italy),
Sedatival (Raffo, Argentina), Tavor (Wyeth, Cinolazepam [75696-02-5], 7-chloro-5-
FRG), Temesta (Ferrosan, Denmark; Wyeth, (2-fluorophenyl)- 2, 3- dihydro- 3- hydroxy-2-
Netherlands), Tolid (Dologiet, FRG), Tra- oxo-1H-1,4-benzodiazepine-1-propionitrile,
pax (Wyeth, Argentina), Wypax (Yamanouchi, (20f),C18 H13 ClFN3 O2 , M r 356.84, mp 190 –
Japan). 193 ◦ C, is prepared by the cyanoethylation
Trimethylacetate: Drupal (Novag, Spain), Pi- of compound 20 (R = Cl, R1 = H, R2 = F) in
ralone (Ferrer, Spain), Placinor (Robert, Spain). the presence of trimethylammonium hydrox-
ide [131]. This drug resembles doxefazepam in
Lormetazepam [848-75-9], 7-chloro- action [132].
5-(2-chlorophenyl)- 1, 3-dihydro-3-hydroxy-
1-methyl-2H-1,4-benzodiazepin-2-one (N- Quazepam [36735-22-5], 7-chloro-5-(2-
methyllorazepam), (20c), C16 H12 Cl2 N2 O2 , fluorophenyl)-1-(trifluoro-2,2,2-ethyl)-1, 3-di-
M r 335.19, mp 192 – 194 ◦ C, is prepared analo- hydro-2H-1,4-benzodiazepine-2-thione, (19),
gously to oxazepam [124]. The hypnotic effect C17 H11 ClF4 N2 S, M r 386.6, mp 138 – 139 ◦ C,
of 1 – 2 mg is similar to that of nitrazepam [125]. is synthesized by the reaction of compound
Trade Names. Lembrol (Vinas, Spain), Loramet 18 (R = Cl, R1 = CH2 CF3 , R2 = F) with phos-
(Wyeth, Belgium, Netherlands), Minias (Far- phorus pentasulfide [133]. Quazepam is a sleep-
mades, Italy), Noctamid, Pronoctan (Schering, inducing drug; it has a long duration of action
FRG), Nocton (Eurolab, Chile). and causes a hangover. It also produces long-
acting metabolites in the body [134].
Temazepam [846-50-4], 7-chloro-1,3-di- Trade Names. Dormalin (Schering Corp., USA),
hydro-3- hydroxy- 1- methyl- 5- phenyl-2H- Prosedar, Selepam, Cetrane (Schering Plough,
1,4-benzodiazepin-2-one (3-hydroxydiazepam, USA).
N-methyloxazepam, oxydiazepam), (20d),
C16 H13 ClN2 O2 , M r 300.74, mp 119 – 121 ◦ C, Haloxazolam [59128-97-1], 10-bromo-
prepared analogously to oxazepam [126]. It is 11b-(2- fluorophenyl)- 2, 3, 7- 11b- tetra-
used as a hypnotic drug and should be taken hydrooxazolo-[3,2-d][1,4]benzodiazepin-
1 – 2 h before going to sleep [127]. 6(5H)-one,C17 H14 BrFN2 O2 , M r 377.22,
Trade Names. Euhypnos (Farmitalia Carlo Erba, mp 185 ◦ C, is synthesized by the reaction of
Italy; Sigma, Australia), Levanxene (Farmitalia bromo- (or tosyloxy-) acetamidobenzophenone
24 Hypnotics
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
Hypnotics 27
15. A. A. Borbély: “Schlafmittel und Schlaf: Peptide (DSIP): An Update,” Peptides (N.Y.) 7
übersicht und therapeutische Richtlinien,” (1986) 1165 – 1187; Neurosci. Biobehav. Rev.
Ther. Umschau 43 (1986) 509 – 516. 8 (1984) 83 – 93.
16. E. Soos: “Schlaf und Schlafmittel,” OAZ 32. A. A. Borbély, G. Youmbi-Balderer: “Ist
Österr. Apoth. Ztg. 28 (1974) 93 – 101. Tryptophan ein biologisches Schlafmittel,”
17. A. N. Nicholson: “Hypnotics Their Place in Schweiz. Rundsch. Med. Prax. 75 (1986)
Therapeutics,” Drugs 31 (1986) 164 – 171. 785 – 790.
18. E. A. Swinyard: Sedatives und Hypnotics in 33. J. P. Olié, S. Lepastier, S. Dana, J. M. Birée,
Remington Pharmaceutical Sciences, 7th ed., Med. Actuelle 5 (1978) 53.
Mack Publ., Easton, Pa., 1985, 34. A. A. Borbély, Ther. Umsch. 43 (1986) 509.
pp. 1059 – 1074. 35. M. B. Scharf, E. O. Bixler, A. Kales, C. R.
19. D. Schneider-Helmert: Schlafmittel: Soldatos, Pharmacology 19 (1979) 173 – 181.
“Symptomatisch oder kurativ?” Schweiz. A. Kales et al., Pharmacology 26 (1983)
ärzteztg. 67 (1986) 1477 – 1483. 138 – 149.
20. W. P. Koella: “The Organization and 36. K. Laemmel, Schweiz. Rundsch. Med. Prax.
Regulation of Sleep,” Experientia 40 (1984) 74 (1985) 1305 – 1310.
309 – 338. 37. E. Greiser: Bewertender Arzneimittelindex:
21. W. Keup: “Mißbrauch und Abhängigkeit von Hypnotika, vol. 2, Med. Pharm. Verlag,
Sedativa und Hypnotika,” Internist 27 (1986) Wiesbaden 1983, pp. 1 – 25.
746 – 756. 38. J. Krieglstein, D. Grusla, Dtsch. Apoth. Ztg.
22. K. Laemmel: “Der Schlaf-Seismograph von 128 (1988) 2041.
Körper und Seele,” Schweiz. Rundsch. Med. 39. E. F. Domino, Annu. Rev. Pharmacol. 2
Prax. 74 (1974) 1305 – 1310. (1962) 215.
23. A. Borbély: Das Geheimnis des Schlafs, 40. F. Hunzicker, F. Künzle, J. Schmitz, Helv.
Deutsche Verlagsanstalt GmbH, Stuttgart Chim. Acta 49 (1966) 1 433. Wander,
1984. GB 1 006 156, 1965; US 3 389 139, 1968.
24. T. Zilker, M. Clarmann: 41. G. Stille, A. Sayers, H. Lauener, E.
“Schlafmittelvergiftungen,” MMW Münch. Eichenberger, Psychopharmakologia 28
Med. Wochenschr. 126 (1984) 297 – 300. (1973) 325; Chem. Abstr. 78 (1973) 131 982 c.
25. V. B. Meyer-Rochow: “Vom Winterschlaf
Y. Take et al., Takeda Kenkyushoho 29 (1970)
lernen,” Selecta 1984, 1116 – 1120.
416; Chem. Abstr. 74 (1971) 21 815 n. H.
26. Drug, Facts and Comparisons, Facts and
Yokotani et al., Takeda Kenkyushoho 29
Comparisons Div., St. Louis 1982, 976 – 1016
(1970) 441; Chem. Abstr. 74 (1971) 21 816 p.
(a division of J. B. Lippincott Co.,
42. W. M. McLamre, S. Y. P. An, A. Barley, J. Org.
Philadelphia-Toronto).
Chem. 20 (1955) 109. Pfizer, US 2 746 900,
27. Martindale, The Extra Pharmacopoeia,
1953.
Hypnotic and Sedatives, 28th ed.,
43. Smith, Kline and French, DE-AS 1 183 333,
Pharmaceutical Press, London 1982,
1963.
pp. 792 – 817.
28. E. A. Swinyard in Remington’s 44. H. C. Caldwell, H. J. Adams, D. E. Rivard, J. V.
Pharmaceutical Sciences, 17th ed., Mack Swintovsky, J. Pharm. Sci. 56 (1967) 920.
Publ. Easton, Pa., 1985, 1059 –1074. 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.
Specific References
48. A. Pictet, F. H. Reichel, Helv. Chim. Acta 6
29. A. A. Borbély, J. Neural. Trans. Suppl. 21
(1923) 621. W. Freudenberg, A. M. Vajda, J.
(1986) 243 – 254.
Am. Chem. Soc. 59 (1937) 1955.
30. S. H. Reppert et al., Science (Washington,
49. C. Bischoff, Ber. Dtsch. Chem. Ges. 7 (1874)
D.C.) 242 (1988) 78 – 81.
31. M. Monnier et al., Experientia 33 (1977) 548. 628.
A. E. Kornmüller, H. B. Lux, K. Wink, M. 50. Rheinpreussen, DE-AS 1 021 843, 1951.
Klee, Naturwissenschaften 48 (1961) 51. Schering, US 2 816 910, 1953, Dtsch. Prior.
503 – 505. D. Schneider-Helmert: “Die 1952.
körpereigene Schlafsubstanz DSIP,” Schweiz. 52. P. Läuger, M. Prost, R. Charlier, Helv. Chim.
Rundsch. Med. 74 (1985) 1311 – 1315. M. V. Acta 42 (1959) 2379. Labaz, US 2 931 828,
Graf, A. J. Kastin: “Delta-Sleep-Inducing 1960.
Hypnotics 29
53. Carlo Erba, GB 736 340, 1955. 78. Ciba, DE 268 158, 1911.
54. G. Newbery, J. Chem. Soc. 127 (1925) 295. 79. Hoffmann-La Roche, DE 526 854, 1930.
Miles, US 2 854 379, 1958; US 2 931 832, 80. M. Seefelder, Justus Liebigs Ann. Chem. 652
1960. (1962) 107. BASF, Ludwigshafen, Festschr.
55. Bayer, DE 225 710, 1909. Friedländer 10, Carl Wurster 60. Geburtstag 1960, 71; Chem.
1 160. Abstr. 56 (1962) 9 927 i; BASF FP 869 Méd.
56. Bayer, DE 286 760, 1913. Friedländer 12, 1961.
704. Bayer, DE 327 129, 1917. Friedländer 81. Geigy, FR 870 714, 1941; DE 756 489, 1941.
13, 809. 82. Eli Lilly, US 1 856 792, 1932.
57. Knoll, DE 185 962, 1906. Friedländer 8, 1217; 83. Bayer, DE 293 163, 1915. Geigy, GB 650 354,
US 914 518, 1909. 1948.
58. Merck Index 10 (1983) 773. E. Schröder, C. 84. J. D. Riedel AG, US 1 739 662, 1929.
Rufer, R. Schmiechen: Pharmazeutische 85. A. C. Cope, E. M. Hancock, J. Am. Chem. Soc.
Chemie, Thieme Verlag, Stuttgart – New York 61 (1939) 776. Sharp & Dohme, US 2 187 703,
1982, pp. 251 – 252. 1940.
59. Bayer, DE 249 241, 1912. Friedländer 10, 86. E. H. Volwiler, J. Am. Chem. Soc. 47 (1925)
1 164. Labs. Aub., CH 374 644, 1964. 2236.
60. Millmaster Onyx, US 3 282 998, 1966. E. H. 87. Compagnie de Bethune, DE 589 947, 1930.
Volwiler, D. L. Tabern, J. Am. Chem. Soc. 58 Friedländer 19, 1 198.
(1936) 1 352. 88. J. D. Riedel AG, US 1 622 129, 1927;
61. M. Bockmühl, O. Schaumann, Dtsch. Med. DE 481 733, 1923.
Wochenschr. 54 (1928) 270. 89. Eli Lilly, US 1 954 429, 1934.
62. Lion Dentifrice (Tokio), DE-OS 1 768 445, 90. O. Schneider, H. Frick, A. H. Lutz, Experientia
1967; US 3 639 457, 1967; DE-OS 2 753 792, 10 (1954) 135. Festschrift F. Reinhardt AG,
1977. Basel 1936, p. 195. Hoffmann-La Roche,
63. N. Okudeira, S. Torii, S. Endo, US 2 151 047, 1938; US 2 680 116, 1954.
Psychopharmacology 70 (1980) 117. K. 91. Ciba-Geigy, US 2 673 205, 1954,
Hillier, Drugs Future 6 (1981) 536 – 537. Schweiz.-Prior. 1951; DE 950 193, 1952. L.
64. W. J. Doran in F. F. Blicke, R. H. Cox (eds.): Hoffmann, E. Tagmann, Helv. Chim. Acta 35
“Barbituric Acid Hypnotics,” Med. Chem. Ser. (1952) 1 541.
Monogr. 4 (1959). 92. Richardson-Merrell Inc. BE 753 846, 1971;
65. R. Y. Levina, F. K. Velichko: “Advances in the DE-AS 2 035 636, 1969.
Chemistry of Barbituric Acids,” Russ. Chem. 93. H. Koch, J. Kotlan, Monatsh. Chem 97 (1966)
Rev. (Engl. Transl.) 29 (1960) 437 – 459. 1 648;Kwizda, DE-AS 1 695 346, 1966.
66. Kirk-Othmer, 2nd ed., 3, 60 – 77. 94. L. Stockinger, H. Koch, Arzneim. Forsch. 19
67. A. Schlieper, Justus Liebigs Ann. Chem. 55 (1969) 167.
(1845) 251; 56 (1845) 1. 95. G. Argyropoulos, G. Lanner, E. Pasquali,
68. R. Gross, Internist 17 (1976) 229. Wien. Med. Wochenschr. 126 (1976) 75; Med.
69. R. J. Levina, F. K. Velichko: “Advances in the Actual. 16 (1980) 388.
Chemistry of Barbituric Acids,” Russ. Chem. 96. K. H. Boltze et al., Arzneim. Forsch. 13 (1963)
Rev. (Engl. Transl.) 29 (1960) 437. 688.
70. E. Fischer, A. Dilthey, Justus Liebigs Ann. 97. B. Dubnick, C. A. Towne in A. Cerletti, F. J.
Chem. 335 (1904) 338. Merck, DE 146 496, Bove (eds.): “The Present Status of
1903.Friedländer 7, 651. Psychotropic Drugs,” Proc. of the VI. Internat.
71. Bayer, DE 247 952, 1911, Friedländer 11, Congr. of the Coll. Internat.
926. Neuro-Psychopharmacologicum, Tarragona
72. Poulenc Frères, US 1 609 520, 1926.
Apr. 1968 pp. 24 – 27; Excerpta Med. 1969,
73. Lambert Thorpe Co., US 1 255 951, 1917; US
77 – 83.
1 576 014,1926.
98. I. K. Kacker, S. H. Zaheer, J. Indian Chem.
74. E. Leyraud, GB 191 008, 1922. Eli Lilly,
Soc. 28 (1951) 344; Chem. Abstr. 46 (1952)
US 1 514 573, 1926; US 1 856 792, 1932.
11 210. General Aniline u. Film Corp.,
75. H. H. Frey, M. P. Magnussen, Arzneim. Forsch.
US 3 213 094, 1963. Lab. Toraude,
16 (1966) 612.
GB 843 073, 1960.
76. Bayer, DE 442 655, 1924; DE 231 150, 1924.
77. Hoffmann-La Roche, US 1 444 802, 1923.
30 Hypnotics
99. Beiersdorf AG, BE 615 282, 1962; Chem. Med. Assoc. 247 (1982) 2147 – 2151. U.
Abstr. 58 (1963) 13 971 ef. S. S. Parmar, K. Klotz: Progress in Clinical Biochemistry and
Kishok, P. K. Seth, R. C. Arona, J. Med. Chem. Medicin, Springer, Berlin – Heidelberg – New
12 (1969) 138. York – Tokyo 1984, pp. 127 – 129. T. A.
100. J. Klosa, J. Prakt. Chem. 14 (1961) 84; 20 Hamor, J. L. Martin in G. P. Ellis, G. B. West
(1963) 283. (eds.): Progress in Medicinal Chemistry 20
101. E. Szirmai, Ther. Umsch. 20 (1963) 542. (1985) p. 183.
102. L. H. Sternbach: “The Benzodiazepine Story,” 107. W. E. Müller, MMW Münch. Med. Wochenschr.
Progr. Drug Res. 22 (1978) 229 – 266. W. 124 (1982) 879 – 884; Dtsch. Apoth. Ztg. 126
Schallek, W. D. Horst, W. Schlosser: (1986) 1036 – 1044.
“Mechanism of Action of Benzodiazepines,” 108. W. E. Müller, Med. Actual. 24 (1988)
Adv. Pharmacol. Chemother. 16 (1979) 649 – 663.
45 – 87. L. O. Randall, W. Schallek, L. H. 109. E. Müller, Dtsch. Apoth. Ztg. 128 (1988)
Sternbach, R. Y. Ning in M. Gordon (ed.): 672 – 674. Roche Magazin 1988, no. 31, 3 – 13
“Chemistry and Pharmacology of the (Hoffmann-La Roche).
1,4-Benzodiazepines,” 110. S. Ebel, H. Schütz: “Biotransformation von
Psychopharmacological Agents, A Series of Benzodiazepin-Derivaten,” Pharm. Unserer
Monographs in Medicinal Chemistry, vol. 3, Zeit 8 (1979) 87. U. Klotz, L. Kangas, J.
Academic Press, New York – San Kanto: “Clinical Pharmacocineties of
Francisco – London 1974, p. 175. V. C. Benzodiazepines,” Prog. Pharmacol. 3 (1980)
Williams, G. C. Varnado, P. U. Nwangwu: 1 – 72. W. E. Müller: “über die Bedeutung der
“Perspectives on the Clinical Pharmacology of Plasmaproteinbindung für Verteilung und
Benzodiazepines,” Med. Actual. 21 (1985) Wirkung von Tranquillantien vom
75 – 96. T. A. Hamor, J. L. Martin in G. P. Ellis, Benzodiazepintyp,” Klin. Wochenschr. 55
G. B. West (eds.): “The Benzodiazepines,” (1977) 105 – 110.
Progress in Medicinal Chemistry 20 (1985) 111. R. G. Borland, A. N. Nicholson, Br. J. Clin.
158 – 223. A. Vela-Bueno, A. Kales: Pharmacol. 2 (1975) 9. S. C. Harvey in
“Benzodiazepine Hypnotics in the Goodman Gilman, The Pharmacological Basis
Multidimensional Treatment of Insomnia,” of Therapeutics, 7th ed., Macmillan Publ.,
Med. Actual. 22 (1986) 271 – 281. C. Pina, A. New York, Toronto, London 1985, p. 352; see
Catedra: “Newer Benzodiazepines, Med. also [102] p. 82.
Actual. 11 (1975) 176; “Chemical Synthesis of 112. Hoffmann-La Roche, DE-AS 1 136 709, 1959;
Benzodiazepines (nur Literatur),” Med. DE-AS 1 145 626, 1959; US 3 109 843, 1961;
Actual. 11 (1975) 211 – 212. H. Schütz: US 3 116 203, 1962; US 3 123 529, 1964.
Benzodiazepines: A Handbook, Springer Delmar Chem., DE-AS 1 811 785, 1968.
Verlag, Berlin 1982. C. Bellantuono, V. Reggi, 113. L. H. Sternbach et al., J. Med. Chem. 6 (1963)
G. Tognoni, S. Garattini: “Benzodiazepines – 261. Hoffmann-La Roche, US 3 116 203,
Clinical Pharmacology and Therapeutic Use,” 1963; US 3 123 529, 1964; US 3 203 900,
Drugs 19 (1980) 195 – 219. S. Garattini: The 1964; GB 1 039 372, 1966.
Benzdiazepines, Raven Press, New York 1973. 114. M. W. Johns, Drugs 9 (1975) 448. J. Rieder,
103. M. Lader (ed.): “New Perspectives in Arzneim. Forsch. 23 (1973) 212. Martindale:
Benzodiazepine Therapy,” Arzneim. Forsch. The Extrapharmacopoea, 28th ed.,
30 (1980) no. 5 a, 851 – 913. U. Klotz: Pharmaceutical Press, London 1982,
“Clinical Pharmacology of Benzodiazepines,” pp. 807 –809. Med. Actual. 1 (1964)
in Progress in Clinical Biochemistry and 132 – 138; 9 (1973) 237.
Medicine, vol. 1, Springer, 115. J. M. Monti, H. Altier, Psychopharmacologia
Berlin – Heidelberg – New York – Tokyo 1984, 32 (1973) 343. W. Schallek, J. Kovacs, A.
pp. 117 – 166. Kuehn, J. Thomas, Arch. Int. Pharmacodyn.
104. H. Schrum: “Benzodiazepine als Schlafmittel,”
Ther. 206 (1973) 161. S. A. Kaplan et al., J.
Schweiz. Rundsch. Med. Prax. 74 (1985)
Pharm. Sci. 63 (1974) 527. R. M. Joye, A. J.
1301 – 1304.
Hance, K. F. Killam, Neuropharmacology 10
105. R. Weber, Dtsch. Apoth. Ztg. 128 (1988)
(1971) 483. J. Freuchen, J. Østergaard, B. O.
2643 – 2652.
106. L. L. Iversen, Nature (London) 275 (1978) Mikkelsen, Curr. Ther. Res. 23 (1978) 90.
447. R. E. Study, J. L. Barker, JAMA J. Am. R. T. Owen, Med. Actual 19 (1983) 163 – 167;
Hypnotics 31
(1977) 1378. Med. Actual. 17 (1981) 147. A. Walser et al., J. Org. Chem. 43 (1978) 936.
283 – 286. P. J. Thorpe, Drugs Future 6 (1981) Hoffmann-La Roche, DE-OS 2 540 522, 1976.
414 – 415. 148. M. Gemper et al., Arzneim. Forsch. 31 (1981)
137. J. B. Hester, D. J. Duchamp, G. C. Chichester, 2177 – 2288. D. A. C. Hegelbach-Feller, J. M.
Tetrahedron Lett. 1971, 1609. J. B. Hester, Tschopp, S. Christeller, J. Fabre, Arzneim.
A. D. Rudzick, B. V. Kamdar, J. Med. Chem. Forsch. 38 (1988) 387. J. Kanto, Med. Actual,
14 (1971) 1078. K. Meduro, H. Tawada, Y. 19 (1983) 376 – 381.
Kuwada, Chem. Pharm. Bull. 21 (1973) 1619, 149. K. H. Weber, W. Sirrenberg, O. Spohn, H.
2382. K. Kamiya, Y. Wada, M. Nishikawa, Daniel, Arzneim. Forsch. 36 (1986) 518.
Chem. Pharma. Bull. 21 (1973) 1520. K. Boehringer, DE-OS 2 410 030,
Megruo, Y. Kuwada, Tetrahedron Lett. 1970, 1974 = US 4 094 984, 1975; DE-OS 2 503 235,
4039. D. L. Loffen, R. J. Fryer, D. A. Katonak, 1975; DE-OS 2 533 924, 1975. Hoffmann-La
F. Wong, J. Org. Chem. 40 (1975) 894. Roche, DE-OS 2 221 623, 1971. K. Gewald,
138. Upjohn, DE-OS 2 012 190, 1969; Chem. Ber. 98 (1965) 3571.
US 3 701 782, 1969; DE-OS 2 302 525, 1973; 150. N. Ferrara et al., Curr. Ther. Res. 37 (1985)
US 4 018 788, 1977; US 3 987 052, 1969. 295. J. Grünberger et al., Curr. Ther. Res. 24
Toyama, DE-OS 1 955 349, 1971. Takeda, (1978) 427. B. Saletu, J. Grünberger, J.
DE-OS 2 114 441, 1971; DE-OS 1 965 894, Volavka, P. Berner, Arzneim. Forsch. 29
1969; US 4 116 956, 1976. (1979) 700. M. S. Langley, S. P. Clissold,
139. S. Tanayama, Y. Shirakawa, Y. Kanai, Z. Drugs 35 (1988) 104 – 122. Pharmcast 87-10,
Suzuoki, Xenobiotika 4 (1974) 33. Y. Kanai, Technomic Information Service, Tokyo 1987,
Xenobiotika 4 (1974) 441. F. S. Eberts, pp. 6-II-151 – 6-II-152.
Pharmacologist 16 (1974) 196; Drug Metab. 151. R. Agar et al., J. Med. Chem. 20 (1977) 1035.
Dispos. 5 (1977) 547. H. Isozaki, M. Tanaka, Russel-UCLAF, DE-OS 2 605 652, 1976; US
K. Inanaga, Curr. Ther. Res. 20 (1976) 493. T. 4 044 142, 1977.
Momose, S. Shii, T. Kuge, Curr. Ther. Res. 19 152. T. G. Johns, D. C. Piper, G. W. L. James, Arch.
(1976) 277. W. Ehrenstein, M. Int. Pharmacodyn. Ther. 240 (1979) 53. J.
Müller-Limmroth, M. Opfermann, F. Weber, Hindmarch, C. A. Clyde, Drugs Exp. Clin.
Arzneim. Forsch. 32 (1982) 280. J. K. Walsh Res. 2 (1980) 61. R. T. Owen, Med. Actual. 20
et al., Curr. Ther. Res. 36 (1984) 866. Med. (1984) 59 – 63. Loprazolam Clinical
Actual. 12 (1976) 355 – 357. Annual Drug Workshop, The Medicine Publishing
Data Report 2 (1979/1980) 85 – 86. Foundation, Oxford, 1983.
140. J. B. Hester, A. D. Rudzick, B. V. Kamdar, J. 153. K. Hirai et al., Heterocycl. Chem. 19 (1982)
Med. Chem. 14 (1971) 1078. 1363; J. Med. Chem. 23 (1980) 764; 24 (1981)
141. A. Kales, M. B. Scharf, J. D. Kales, Science 20. Shionogi, DE-OS 2 725 164, 1977;
(Washington, D.C.) 201 (1978) 1039. US 4 159 374, 1977. Hoffmann-La Roche,
142. T. Roth, M. Kramer, T. Lutz, Drugs Exp. Clin. DE-OS 2 537 069, 1975.
Res. 1 (1977) 279. D. A. C. Hegelbach-Feller, 154. K. Hirose, A. Matsushita, M. Eigyo, Arzneim.
J. M. Tschopp, S. Christeller, J. Fabre, Forsch. 31 (1981) 63. Drugs Future 9 (1984)
Arzneim. Forsch. 38 (1988) 387. 522 – 524; 11 (1986) 626 – 628. Pharmcast
143. M. Leibowitz, A. Sunshine, J. Clin. 87-1, Technomic Information Service, Tokyo
Pharmacol. 18 (1978) 302. K. K. Okawa, 1987, pp. 5-III-139 – 5-III-141.
G. S. Allen, J. Int. Med. Res. 6 (1978) 343. 155. Rhône-Poulenc, DE-OS 2 300 491, 1973; US
B. B. Knap, E. L. Boyd, G. Linsenmeyer, O. J. 3 862 149, 1975. C. Jeanmart, C. Cotrel, C. R.
Linet, Curr. Ther. Res. 23 (1978) 230. Ser. C 287 (1978) 377.
144. A. N. Sing, Curr. Ther. Res. Clin. Exp. 27
156. R. Duriez, Therapie 34 (1979) 317. E.
(1980) 627. A. N. Nicholson, Br. J. Clin.
Wickström, K. E. Giercksky, Eur. J. Clin.
Pharmacol. 9 (1980) 223.
Pharmacol. 17 (1980) 93; Clin. Therap. 3
145. L. Lasagnon, Lancet 1 (1980), 815.
(1978) 59. . J. C. Blanchard, A. Boireau, C.
146. E. A. Swinyard in Remingtons Pharmaceutical
Garrett, L. Julou, Life Sci. 25 (1979) 2417.
Sciences, 17th ed., Mack Publ., Easton, Pa.
P. H. Wu, J. W. Phillis, A. S. Bender, Life Sci.
1985, p. 1064. AMA: Drug Evaluation, 6th
28 (1981) 1023. J. Castaner, Drugs Future 4
ed., Am. Med. Assoc., Chicago 1986, p. 103. J.
(1979) 59 – 60; 5 (1980) 56; 6 (1981) 62 – 63;
Castaner, Drugs Future 1 (1976) 393 – 396; 2
7 (1982) 69 – 70; 8 (1983) 79; 9 (1984) 81; 10
(1977) 558.
Hypnotics 33
(1985) 88 – 89; 11 (1986) 76 – 78; 12 (1987) 158. Synthelabor, EP 50 563, 1981. P. George, H.
95; 13 (1988) 88 – 89. Pharmacast 87-7, Depoortere, J. Pharm-Sci. (in Press).
Technomic Inf. Service, Tokyo 1987, 1989, 159. S. Arbilla et al., Naunyn-Schmiedebergs Arch.
6-I-87 – 6-I-89. Int. Pharmacopsychiatry 17 Pharmacol. 330 (1985) 248. M. Dimsdale et
(1982) (Suppl. 2), 1 – 252. K. A. Goa, R. C. al., Drugs Future 12 (1987) 777 – 780.
Heel, Drugs 32 (1986) 48 – 65. 160. M. Dimsdale et al., Drugs Future 13 (1988)
157. G. Houghton et al., Int. J. Clin. Pharmacol. 106 – 109.
Biopharm. 23 (1985) 97. R. Duriez et al., 161. J. Maillard, M. Vincent, M. Bernard, Bull. Soc.
Therapie 34 (1979) 317. J. Gaillot et al., Chim. Fr. 1961, 529; 1962, 1580.
Pharmacology 27 (1983) (Suppl. 2) 76 – 91. Logeais-Labs., GB 902 388, 1962;Fr. M. 379,
1962; Chem. Abstr. 57 (1962) 15251 g.
sucrose is replaced by sugar substitutes such as run refers to volume expansion caused by air in-
fructose or sorbitol to ensure that the ice cream corporation, it being 80 % when 1 L of liquid mix
retains desirable freezing and melting character- yields 1.8 L of frozen product.) Product extruded
istics. from the freezer can be dispensed into cones or
Ice cream is sold in a variety of forms, shapes, cups to be consumed in soft form which cannot
and flavors. Forms of supply include: be stored or transported. Alternatively, it can be
packaged and placed in a colder environment (at
1) single portions (sticks, cups, tubs, cones, below −18 ◦ C) to allow quiescent freezing of
sandwiches, puppet ice), most of the remaining water to yield ice cream
2) family-sized portion (bricks, bombes, multi- in its typically hardened, packaged form. The
packs), and latter process is a combination of freezing quies-
3) catering packs. cently and under agitation. Hardened ice cream
A further differentiation among products is (including varieties that are easily scoopable at
based on whether freezing occurs quiescently, −20 ◦ C) can be stored and transported.
and/or under agitation. Most stick confections
are subjected to quiescent freezing by pouring
3. Legal Aspects
liquid mix into molds immersed in a brine so-
lution at −35 ◦ C. The finished product with in- International compositional requirements for ice
serted stick has the geometry of the mold, which cream and related products are summarized in
allows for a great variety of shapes and sizes. Table 1.
Freezing under agitation refers to products in In the United States ice cream, ice milk, sher-
which 35 – 55 % of the water content is frozen bet, fruit ice, and mellorine are defined under the
as the temperature is lowered to between −4 and standards of the federal Food and Drug Admin-
−6 ◦ C in a swept-surface heat-exchange cylin- istration [2]. Guidelines for other products are
der, which also distributes air as overrun. (Over- specified by individual states. Key provisions for
4 Ice Cream and Frozen Desserts
ice cream are a minimum of 10 % milk fat and air cell [5]. Evidence exists for accumulation of
20 % total milk solids, and a product weight of at protein and fat near the interface between air
least 4.5 lb per gallon (0.54 kg/L). The weight re- and the liquid, unfrozen continuous phase (also
quirement prevents more than a doubling of mix called serum) [5]. Emulsions are thermodynam-
volume through air incorporation during manu- ically unstable [6] and changes occurring are the
facture. Ice milk is low-fat ice cream (with a fat result of the shearing of the mix during freezing
content of 2 – 7 wt %). In mellorine some or part and whipping in the swept-surface cylinder. Par-
of the milk fat is replaced by another plant or an- ticle size distribution for fat tends towards larger
imal fat. Sherbet must contain 1 – 2 % milk fat globules and formation of agglomerates is evi-
with not more than 5 % total milk solids and a dent. Depending on the amount of change, the
minimum weight of 6 lb per gallon (0.71 kg/L). physical characteristics of the frozen emulsion
Most sherbets have a tart taste due to the addi- matrix are profoundly influenced.
tion of an edible acid, usually citric acid. Fruit
ice requirements are similar to those of sherbet,
except that milk solids are not allowed.
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.
crose in most frozen dessert products. Corn of use as emulsifiers for ice cream. The polysor-
syrups of the low conversion 36 DE type com- bates are especially effective in hastening con-
pensate to a degree for the gradual reduction over trolled emulsion destabilization leading to desir-
many years in the milk solids content of econ- able extrusion properties for ice cream [11].
omy and regular ice cream brands. Since their
introduction in the 1970s, fructose syrups ob-
tained by hydrolysis of starch syrup to glucose 7. Manufacture of Mix
and enzyme isomerization have gained a strong
foothold. Conventional corn syrups are not as Preparation of a mix for freezing into ice cream
sweet as sucrose, but fructose syrups are equally is a multistep process involving the assembly of
sweet or sweeter. Although they are the most the mix ingredients, pasteurization of the result-
economical source of sweetness, fructose syrups ing fluid mix, homogenization to stabilize the
increase the molar concentration of solutes, thus emulsion, and finally the cooling and storage of
requiring lower freezing and storage tempera- the finished mix to await freezing. Equipment
tures to maintain proper hardness in packaged can be sized to meet production needs. Mix out-
products. For a detailed description of syrups, put varies from 2000 L/h for a small ice cream
see →Glucose and Glucose-Containing Syrups. 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-
6.4. Stabilizers and Emulsifiers gienic preparation of ice cream requires not only
pasteurization, but also regular cleaning and dis-
In most countries the addition of stabilizers and infection of equipment, proper packing and stor-
emulsifiers to ice cream is legally restricted. age, i.e., Good Manufacturing Practice. Pasteur-
Most ice cream products contain 0.1 – 0.2 wt % ization is not compulsory in all countries, e.g., in
hydrocolloid stabilizer which functions as a wa- the Federal Republic of Germany it is only nec-
ter control agent and increases the viscosity of essary for the types “Eiskrem” (the main type
the unfrozen aqueous phase. By mechanisms of industrially produced edible ice) and “Ein-
which are poorly understood, the hydrocolloids facheiskrem”. The microbiological state of ice
delay growth of large ice crystals and prevent de- cream is subject to very stringent regulations in
terioration of quality during storage and distri- most countries, e.g., total microbial count, co-
bution. Emulsifiers, used at the same or slightly liforms, and pathogenic microorganisms (espe-
higher levels than hydrocolloids, serve as a man- cially Salmonella).
ufacturing aid by helping to control emulsion
changes and lead to good extrusion properties Pasteurization. The primary goal of pas-
for ice cream from the freezing cylinder. For a teurization is the guaranteed destruction of
description of emulsifiers, see →Foods, 3. Food pathogenic microorganisms. In the process, al-
Additives, Chap. 3.6. most all nonpathogens are simultaneously de-
A large number of hydrocolloids function in stroyed, but the mix is not sterile and must be
ice cream systems; guar gum, locust bean gum, stored at 4 ◦ C to prevent multiplication of sur-
cellulose gum, and carrageenan are used most viving microorganisms. Pasteurization regimes
frequently. Most are used in combination as sta- are defined and strictly controlled by federal and
bilizer blends, selection being based on the rhe- state regulatory agencies.
ological properties desired by a particular ice In batch pasteurization, all ingredients in
cream manufacturer. Generally, carrageenan is the desired proportions are fed into a jacketed
a component of stabilizer blends, e.g., with guar vat, heated for 10 – 30 min under agitation to
or locust bean gum at a ratio of 1 : 10 [9] or at 68.3 ◦ C, and held at this minimum pasteuriza-
a concentration of 0.018 wt % [10]. It functions tion temperature for at least 30 min. The vat con-
as a manufacturing aid to prevent separation of tents must then be pumped through the homog-
mix into layers during quiescent storage prior to enizer before a new batch can be started.
freezing into ice cream. Continuous high temperature – short time
Monoglycerides in mixture with diglyc- (HTST) pasteurization is more energy efficient
erides, and polysorbate-80 have a long history than batch pasteurization and better suited to
Ice Cream and Frozen Desserts 7
high-volume plants. Mix ingredients are assem- centration. Dissolved components in the aque-
bled in a vat similar to the batch process. How- ous phase that control this colligative property
ever, the mix is pasteurized by pumping it from are the lactose and soluble salts of milk, as well
the vat as a thin film through a plate heat ex- as the carbohydrates supplied by sweetening and
changer. Thin film heating is almost instanta- flavoring ingredients. Measured and calculated
neous and continuous HTST pasteurization is values for the freezing point depression are in
the preferred approach to pasteurization in the close agreement, at least in regard to the initial
ice cream industry. Temperature – time condi- freezing temperature [12]. As water freezes, the
tions for HTST pasteurization are 78 – 80 ◦ C solutes become more concentrated, thus requir-
for 20 – 40 s (US minimum requirements are ing even lower temperatures for the continuing
79.4 ◦ C for 25 s). change of state of water to ice.
The times and temperatures for batch and Consider a mix containing 15 wt % sucrose,
HTST pasteurization are acceptable minimum 7 wt % lactose and milk salts (i.e., 22 wt % solu-
combinations. Often, a higher temperature or ble solids), and 38 wt % total solids (62 wt % wa-
longer hold is employed to provide added pro- ter). The ratio of soluble solids to unfrozen wa-
tection relative to microbial kill. Furthermore, a ter is 35 : 100 at the start of freezing. When soft
significant number of manufacturers believe that ice cream at −5.5 ◦ C leaves the swept-surface
an extra measure of heat energy, especially at cylinder of the freezer to be packaged, the ra-
elevated temperature, improves ice cream body tio will be 70 : 100 since 50 % of the mix water
and texture. will have crystallized as ice. At a typical stor-
age temperature near −20 ◦ C the ratio will be
Homogenization. Following batch pasteur- 275 : 100. Obviously, as temperature decreases
ization or as an integral part of the HTST and water freezes, ice cream stiffens progres-
process, mix passes through a homogenizer. sively to a hardened mass.
Homogenizers are high-pressure pumps which
force mix through a restricted valve under de-
fined conditions of turbulence and shear. During 8.1. Freezing as Mix Agitates
homogenization, fat globules are broken up to
less than 2 µm in diameter; the resulting small Ice cream freezers are of two types, batch and
globules become coated with milk protein, emul- continuous. In batch freezers, the mix is frozen
sifier, and other surface-active constituents. The and whipped in single batches of 10 – 20 L de-
temperature of the mix should be at least 50 ◦ C pending on machine size. The refrigerant (ei-
to ensure that the fat is completely melted. ther ammonia or Freon) in the jacketed rim of
the cylinder, absorbs heat from the mix causing
Storage of Mix. The prepared mix is held it to freeze as a thin film. The film is contin-
at 4 ◦ C or less in a storage tank until it can be uously scraped off the cylinder wall by revolv-
frozen into ice cream. In the United States, freez- ing scraper blades. Depending on conditions, 4 –
ing is usually carried out the same day the mix is 10 min is required to reach −4.4 ◦ C, which is a
made, with only a 2 – 3 h hold in the storage tank typical discharge temperature for batch-frozen
to allow complete crystallization of fat globules. ice cream. The proportion of mix water frozen
The former practice of holding the cold mix for (ca. 40 %) at this temperature results in a vis-
at least 6 h (usually overnight) before freezing cosity or thickness for the whipping mass that
commenced has been been superseded by the is ideal for incorporation and retention of air as
use of hydrocolloid stabilizers. 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
8. Freezing Mix to Produce Ice through the freezing cylinder and exits
Cream semifrozen in a continuous stream. Because the
cylinder barrel is pressurized, product stiffness
Mix water begins to freeze at −2.5 ◦ C, the exact is not nearly as important for air incorporation as
temperature being a function of the solution con- in batch freezing. Consequently, ice cream flow-
8 Ice Cream and Frozen Desserts
ing from a continuous freezer can be at −6 ◦ C pallets for storage. The main advantages of tun-
or even lower, which means more extensive nu- nel hardening are efficiencies of time, labor, and
cleation during the fast freezing stage of ice space utilization.
cream manufacture. This produces smaller ice
crystals and a smoother texture. Because con-
tinuous freezers use a pumping system, they can 9. References
be adapted for molding ice cream by extrusion,
for incorporating variegating syrups and a great 1. Codex Standard for Edible Ices and Ice Mixes
variety of particulate flavor ingredients, and the 137, 1981; IDF Bulletin 172 (1984) Appendix
automatic filling of packages. Several sizes of II.
continuous freezers allow mix pumping rates 2. Code of Federal Regulations 21, Food and
between 200 and 2000 L/h over an almost infi- Drugs Parts 110 – 199, U.S. Government
nite range of overruns. Because of their many Printing Office, Washington, D.C., 1987,
advantages, continuous freezers are preferred pp. 228 – 236.
over batch machines, except in the smallest ice 3. The Latest Scoop, International Ice Cream
cream factories. Association, Washington, D.C., 1987.
4. K. C. Berger in S. Friberg (ed.): Food
Emulsions, vol. 5, Dekker, New York 1976,
pp. 141 – 213.
8.2. Quiescent Hardening 5. P. G. Keeney, J. A. Maga, J. Dairy Sci. 48
(1965) 1591 – 1596.
The final stage of ice cream manufacture in- 6. S. Friberg in S. Friberg (ed.): Food Emulsions,
volves further freezing to convert packages of vol. 5, Dekker, New York 1976, pp. 1 – 37.
soft ice cream to a hardened form. Hardening 7. P. G. Keeney: Commercial Ice Cream and
takes place in either a hardening room or tun- Other Frozen Desserts, Extension Circular
nel. Ideally in a hardening room, currents of air 553, The Pennsylvania State University,
at −30 ◦ C sweep across all sides of a carton or University Park, Pa., 1972, pp. 1 – 50.
package. For this to occur, the packages must 8. P. G. Keeney, D. V. Josephson, Ice Cream
be kept separated on shelves until core temper- Trade J. 57 (1961) no. 5, 28 – 34.
atures are well below −20 ◦ C, this may require 9. K. A. Hyde, J. Rothwell: Ice Cream, Churchill
several hours depending on package size. Only Livingstone, Edinburgh 1973.
then can the packages be bundled and placed 10. P. G. Keeney, M. Kroger in B. H. Webb, A. H.
on pallets for storage until shipment from the Johnson, J. A. Alford (eds): Fundamentals of
factory. In tunnel hardening, packages move Dairy Chemistry, AVI Publishing 6;
through the tunnel on a conveyor in −50 ◦ C high Washington 1974.
11. P. G. Keeney, Food Technol. (Chicago) 36
velocity air. Hardening is much faster than in
(1982) no. 11, 65 – 70.
a hardening room, and upon exiting the tunnel
12. W. Arbuckle: Ice Cream, 4th ed., AVI
packages are ready to be bundled and placed on
Publishing Co., Westport, CN, 1986, p. 46.
Illites → Clays
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
My Profile Log
HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry
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
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.
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]
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).
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].
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
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.
page 6 of 83
Imaging Technology : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
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].
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.
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.
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.
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.
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.
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.
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].
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.
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.
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
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).
Silicon Mesa Technology. The basic structure of the silicon mesa (patented by Texas Instrument) is shown in Figure 22
[87].
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.
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.
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.
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].
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.
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.
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
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].
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).
A) Printing process: a) Paper; b) Print head; c) Ribbon supply; d) Ribbon take-up; e) Pinch rollers
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).
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.
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].
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.
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.
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.
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).
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.
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
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).
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.
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].
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].
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.
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
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.
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].
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 41. A) Varying lightness: a) Saturated yellow; b) Hue yellow saturation unchanged, 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.
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.
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.
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.
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.
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
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
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.
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
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.
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.
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].
The exposed negative and the receiving material are passed through the developing bath while being pressed together
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
= 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].
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
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].
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.
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.
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.
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:
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.
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
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.
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:
[Top of Page]
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.
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.
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].
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:
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].
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).
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.
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].
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.
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.
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].
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.
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.
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.
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.
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.
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.
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).
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].
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).
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].
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:
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].
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.
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.
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).
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.
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].
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]).
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 %.
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.
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]
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.
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
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.
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.
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.
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.
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
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.
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.
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.
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.
Positive Negative
Du Pont Riston x
Morton Thiocol
(Dynachem) Laminar x
Hercules Aquamer x
Hitachi Chemical Photec x
Hoechst Ozatec x x
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)
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].
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).
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.
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].
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.
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.
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
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.
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.
Negative Positive
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.
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:
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.
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
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].
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).
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
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.
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
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.
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.
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
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
Related Articles
Photography is a separate keyword.
[Top of Page]
page 83 of 83
Imidazole and Derivatives 1
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
3. Chemical Properties
Imidazole is a moderately strong base (pK b = 7.0
2. Physical Properties [1]), and a weak acid (pK a = 14.9 [1]). Imi-
dazoles substituted with electron-withdrawing
Some important physical properties of imida- groups are stronger acids than imidazole itself;
zole are listed in Table 1. Molecular masses, e.g., 4(5)-nitroimidazole has a pK a of 9.3 [1].
boiling points, melting points, and CAS registry Imidazole is stable at 400 ◦ C, possesses con-
numbers of the most important imidazole deriva- siderable aromatic character, and undergoes the
tives are listed in Table 2. Substitution at nitro- usual electrophilic aromatic substitution reac-
gen generally causes a reduction in the melting tions. Nitration and sulfonation require, how-
and boiling points, because the N−H group is ever, far more drastic conditions than the corre-
no longer available for intermolecular hydrogen sponding reactions with benzene. Other substi-
bonding. Distillation of nitroimidazoles is ac- tution reactions of imidazole include halogena-
companied by the risk of spontaneous decom- tion, hydroxymethylation, coupling with aro-
position. matic diazonium salts, and carboxylation. Cat-
alytic hydrogenation has not been described [1],
[2].
∗ At 1.3 kPa.
∗∗ Decomposes.
∗ LD50 , µL/kg.
6 Imidazole and Derivatives
role. The materials so far investigated are mostly be due to the inadequate metabolic activation
eye irritants, but not skin irritants. achieved in the test systems. Materials bearing a
The acute toxicities (LD50 , mg/kg) for some carboxylate or sulfate group may in fact be non-
nitroimidazoles are as follows: mutagenic [43], [45], but the data are insufficient
4(5)-Nitroimidazole [26] 1660 (rat, oral)
for a firm conclusion.
1,2-Dimethyl-5-nitroimidazole 3000 (rat, oral) Urine and other body fluids of patients treated
hydrochloride (dimetridazole with metronidazole in doses of 750 mg/d showed
hydrochloride) [26] 200 – 2000 (cat, mutagenic activity in the Ames test [52], [53].
oral)
1-Methyl-5-nitroimidazole [26] ca. 200 (rat, oral) Patients who received 200 – 1200 mg/d for sev-
1-(2-Hydroxymethyl)-2-methyl-5-nitroimida- 3800 (mouse, oral) eral months were found to have two- to four-fold
zole increases in chromosomal abnormalities in pe-
(metronidazole) [41]
α-Methoxymethyl-2-nitroimidazole-1-etha- 2130 (rat, oral)
ripheral leukocytes [54].
nol Efforts have been undertaken to lower the
(misonidazole) [42] mutagenicity of nitroimidazoles by derivatisa-
2-Methyl-4(5)-nitroimidazole [26] 1540 (rat, oral)
tion and substitution [47], [48], [55]. Note, how-
ever, that mutagenicity in vitro is not quantita-
Four-hour dust inhalation tests on rats with tively extrapolable to in vivo situations. Further-
4(5)-nitroimidazole and dimetridazole gave more, a lower mutagenicity does not necessarily
LC50 values of 6.56 and 5.3 mg/L, respec- indicate a decrease in the carcinogenic potential
tively. Dimetridazole and 1-methyl-5-nitroimi- as well.
dazole have an irritant action on the eyes of rab-
bits, while 4(5)-nitroimidazole is non-irritating Carcinogenicity. The carcinogenic potential
[26]. None of the three compounds exhibits irri- has been investigated for some nitroimidazoles
tancy towards the skin of rabbits [26]. used or foreseen as drugs. However, the study de-
signs were not fully in accordance with present
Mutagenicity. Several nitroimidazoles have standards. The carcinogenic responses obtained
been subjected to bacterial point mutation as- so far were moderate or low. Human experi-
says (such as the Ames test), and, in most cases, ence with nitroimidazole drugs has not revealed
were found to exert mutagenicity [37], [43–51]. cancer cases due to therapeutic usage. Metron-
Among these were the following: idazole is listed as a carcinogen by IARC [56].
1-Methyl-5-nitroimidazole Mice receiving 0.06, 0.15, 0.3, or 0.5 % (ca. 90,
2-Methyl-4(5)-nitroimidazole 225, 450, or 750 mg/kg) of metronidazole in the
1,2-Dimethyl-5-nitroimidazole (dimetridazole) diet throughout their life-span showed a dose-
1-(2-Hydroxymethyl)-2-methyl-5-nitroimida- dependent increase in the incidence of lung tu-
zole (metronidazole) mors and lymphomas. The lowest dose did not
2-Isopropyl-4-nitroimidazole significantly increase the incidence of tumors
1-Methyl-2-isopropyl-5-nitroimidazole [57].
(ipronidazole) No carcinogenic effects were found in rats in
1-(3-Chloro-2-hydroxypropyl)-2-methyl-5- a feeding experiment with 0.135 % (ca. 200 mg
nitroimidazole (ornidazole) kg −1 d−1 ) of metronidazole in the diet for 66
(1-Methyl-5-nitroimidazole-2-yl)methyl carba- weeks and a post-observation period of 10 weeks
mate (ronidazole) [58]. This administration period is, however, too
α-Methoxymethyl-2-nitroimidazole-1-ethanol short to fully rule out carcinogenic effects.
(misonidazole) More recent investigations with metron-
idazole in the diet of mice (75, 150, or
Nitroimidazoles also induce streptomycin- 600 mg kg−1 d−1 for 92 weeks) and rats
resistant mutants (forward mutations) in Kleb- (0.06, 0.3 or 0.6 %, equivalent to 30, 150 or
siella pneumoniae [44], [45]. 300 mg kg−1 d−1 for 150 weeks) showed dose-
Conflicting results were earlier reported with related increases in tumor incidence (liver and
4(5)-nitroimidazole, 2-methyl-4(5)-nitroimida- mammary tumors, Leydig cell and pituitary tu-
zole, and 2-methyl-4(5)-nitroimidazole-1-yl- mors) [59], [60].
acetate [43], [46], [47]. Negative results may
8 Imidazole and Derivatives
After gavage administration of metronida- 5. BASF, DE 2 512 513, 1976 (A. Frank, T.
zole to mice during pregnancy and lactation Dockner).
(2 mg per animal per day for 5 d per week) an 6. BASF, EP 0 111 073, 1987 (A. Frank, T.
increased tumor incidence was observed in the Dockner).
offspring [61]. 7. BASF, EP 0 036 521, 1983 (T. Dockner, U.
Metronidazole (0.15 or 0.3 %) in the diet of Kempe, H. Krug, P. Magnussen, W. Prätorius).
hamsters did not increase the incidence of tu- 8. Houdry Process Corp., US 2 847 417, 1958
mors; however, published details of the study (W. E. Erner).
are insufficient for a reliable assessment of the 9. BASF, EP 0 000 208, 1979 (T. Dockner, A.
Frank).
carcinogenic effect [62].
10. C. Corsar et al., Arzneim. Forsch. 16 (1966)
A group of patients who had been treated with
23.
metronidazole during the period 1960 –1969 did 11. BASF, DE 2 645 172, 1982 (M. Wetzler, T.
not show increased tumor rates in a follow-up Dockner).
study published in 1988 [63]. 12. BASF, DE 1 670 293, 1976 (H. Spänig, A.
Dimetridazole was found to have tumori- Steimmig, J. Sand).
genic activity in rats receiving 0.2 % in the diet 13. BASF, DE-AS 2 233 908, 1977 (T. Dockner,
for 46 weeks. Of 35 Sprague – Dawley rats, 25 H. Krug).
developed a benign, but possibly premalignant 14. M. Häring, Helv. Chim. Acta 42 (1959)
mammary tumor (fibroadenoma) compared to 4 1845 – 1846.
from 35 rats in the control group [58]. The ad- 15. A. Grimison, J. H. Ridd, B. V. Smith, J. Chem.
ministration period was, however, too short to Soc. 1960, 1357 – 1362.
fully determine the tumorigenic potential. 16. W. Reppe, Justus Liebigs Ann. Chem. 601
Ornidazole (25, 100, or 400 mg kg−1 d−1 ) (1956) 81 – 138.
was reported to cause no tumors in a two-year 17. Rhône-Poulenc, DE-OS 1 470 200, 1970 (C.
feeding experiment in rats [64]. Podesva, K. Vagi).
The overall conclusion from all experiments 18. Y. Masashige, M. Masaichiro, Chem. Pharm.
is that the carcinogenic potential of nitroimida- Bull. 24 (1976) 1480 – 1484.
zoles has to be taken into account, but the amount 19. BASF, EP 0 015 516, 1982 (U. Kempe, T.
and quality of the available data do not allow firm Dockner, F. Frank, H. Karn).
20. Delalande S.A., DE-OS 2 934 925, 1978 (Y. A.
conclusions.
Hubert-Brierre).
21. E. R. Shepard, H. A. Shonle, J. Am. Chem.
Neurotoxic effects. There are several case
Soc. 69 (1947) 2269 – 2270.
reports on peripheral neuropathy due to Metron-
22. W. Perkow: Wirksubstanzen der
idazole treatment. The symptoms appear to be
Pflanzenschutz- und
reversible upon ceasing treatment [65]. Schädlingsbekämpfungsmittel, 2nd ed., Verlag
Paul Parey, Berlin – Hamburg 1985.
23. M. D. Nair, K. Nagarajan, Prog. Drug Res. 27
8. References (1983) 163 – 252.
24. Shell Oil Co., US 3 538 039, 1970 (W. L.
General References
Lantz, J. P. Manasia).
1. M. R. Grimmett in K. T. Potts: Comprehensive
25. F. Ferrari, Arch. Int. Pharmacodyn. Ther. 227
Heterocyclic Chemistry, vol. 5, Pergamon
(1985) 303 – 312.
Press, 1984, pp. 345 – 456.
2. K. Hofmann: The Chemistry of Heterocyclic 26. BASF Aktiengesellschaft, unpublished results.
Compounds, Imidazole and Derivatives, Part 27. L. E. Voogd et al., Mutat. Res. 66 (1979)
I, Intersci. Publ. Inc., New York 1953. 207 – 221.
28. A. Momii et al., Iyakuhin Kenkyu 10 (1979)
351 – 357.
Specific References
29. E. Grzybowska et al., Acta Microbiol. Pol. 34
3. BASF, DE-OS 2 360 175, 1979 (A. Frank, H.
(1985) 111 – 120.
Karn, H. Spänig).
4. BASF, EP 0 012 371, 1982 (T. Dockner, U. 30. J. A. Ruddick et al., Teratology 13 (1976)
Kempe, H. Krug, P. Magnussen, W. Prätorius, 263 – 266.
H. Szymanski).
Imidazole and Derivatives 9
31. K. Nishie et al., Toxicol. Appl. Pharmacol. 14 50. P. G. Wislocki et al., Chem. Biol. Interactions
(1969) 301 – 307. 49 (1984) 27 – 38.
32. T. E. Fairbrother et al., Vet. Hum. Toxicol. 29 51. H. S. Rosenkranz et al., Biochem. Biophys.
(1987) 312 – 315. Res. Commun. 66 (1975) 520 – 525.
33. M. H. Idaka, Okayama Igakkai Zasshi 88 52. M. S. Legator et al., Science 188 (1975)
(1976) 673 – 680. 1118 – 1119.
34. Ullmann, 4th ed., 23, 617. 53. E. Mohtashamipur et al., Mutagenesis 1
35. Niosh: Registry of Toxic Effects of Chemicals. (1986) 371 – 374.
36. M. Hidaka, Okayama Igakkai Zasshi 88 54. F. Mitelman et al., Lancet, 1976, 802.
(1976) 653 – 657. 55. G. Cantelli-Forti et al., Teratog. Carcinog.
37. D. G. Lindmark et al., Antimicrob. Agents Mutag. 3 (1983) 51 – 63.
Chemother. 10 (1976) 476 – 482. 56. IARC Monograph 13, 113 – 121.
38. D. I. Edwards, M. Dye, H. Carne, J. Gen. 57. M. Rustia, P. Shubik, J. Nat. Cancer Inst. 48
Microbiol. 76 (1973) 135. (1972) 721 – 729.
39. J. L. Blumer et al., Cancer Res. 40 (1980) 58. S. M. Cohen et al., J. Nat. Cancer Inst. 51
4599 – 4605. (1973) 403 – 417.
40. M. Isildar et al., Toxicol. Appl. Pharmacol. 94 59. J. H. Rust, An assessment of metronidazole
(1988) 305 – 310. tumorigenicity studies in the mouse and the
41. B. Cavalleri et al., J. Med. Chem. 20 (1977) rat, in S. M. Finegold (ed.): Metronidazole,
1522. Proc. Intern. Metronidazole Conf. Montreal,
42. National Technical Information Service USA, Excerpta Medica, Amsterdam (1977)
PB 81-121212. 138 – 144.
43. C. E. Voogd et al., Mutat. Res. 86 (1981) 60. M. Rustia, P. Shubik, J. Nat. Cancer Inst. 63
243 – 277. (1979) 863 – 868.
44. C. E. Voogd et al., Mutat. Res. 26 (1974) 61. M. Chacko et al., Cancer Res. Clin. Oncol.
483 – 490. 112 (1986) 135 – 140.
45. C. E. Voogd et al., Mutat. Res. 66 (1979) 62. F. C. Roe, Proc. Intern. Metronidazole Conf.
207 – 221. Montreal, Excerpta Medica, Amsterdam
46. H. S. Rosenkranz et al., Mutat. Res. 38 (1976) (1977) 132 – 137.
203 – 206. 63. C. M. Beard et al., Mayo Clin. Proc. 63 (1988)
47. J. B. Chin et al., Mutat. Res. 58 (1978) 1 – 10. 147 – 153.
48. J. S. Walsh et al., J. Med. Chem. 30 (1987) 64. R. Richle et al., Arzneim. Forsch. 28 (1978)
150 – 156. 612 – 625.
49. M. Suzangar et al., Biochem. Pharmacol. 36 65. W. G. Bradley et al., Br. Med. J. 2 (1977)
(1987) no. 21, 3743 – 3749. 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
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
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
1) modified into a water-insoluble form,
Year Biocatalyst, process, or event Comments 2) retained by an ultrafiltration membrane in-
side a reactor, or
1979 immobilization of plant cells first immobilization
(Morinda, Catharanthus, Digitalis) of plant cells 3) bound to another macromolecule to restrict
by the calcium alginate method their mobility.
[17]
1979 slice of porcine kidney first immobilization Several schemes have been suggested for
immobilized in the membrane of of animal cells
an ammonia electrode for amino classifying immobilization techniques [24],
acid detection [18] [30], [40], [41]. The classification used here is
1987 Corynebacterium spp. immobilized first reported
in polyacrylamide gel for industrial production
based on that of Kennedy and Cabral [41]
acrylamide production [19] of a commodity and attempts to combine the nature of the inter-
chemical by action responsible for immobilization with that
immobilized cells
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
2. Immobilization Techniques as the purpose of immobilization, activity, sta-
bility, simplicity, and economic feasibility [42],
2.1. Introduction [43].
Since the late 1960s a variety of techniques have
been developed for immobilizing biocatalysts.
2.2.1. Carriers for Enzyme Immobilization
Complete coverage of the vast number of pa-
pers, books, and patents on this topic is almost
The major components of an immobilized en-
impossible. The reader is referred to comprehen-
zyme system are the enzyme, the support, and
sive reviews and monographs [20–38].
the mode of interaction of the enzyme with the
The definition of biocatalysts used in this
support. The support may be a membrane, a
chapter is that drawn up by the Working Party on
water-insoluble solid, or a polymer matrix. The
Applied Biocatalysts within the European Fed-
support (carrier) should have the following prop-
eration of Biotechnology: “Immobilized biocat-
erties [44], [45]:
alysts are enzymes, cells, or organelles (or com-
binations of them) which are in a state that per- 1) Large surface area and high permeability
mits their reuse” [39]. 2) Sufficient functional groups for enzyme at-
Biocatalysts dissolved in aqueous buffer so- tachment under nondenaturing conditions
lutions are usually referred to as soluble or na- 3) Hydrophilic character
tive enzymes, cells, cell parts, or organelles. The 4) Water insolubility
terms immobilized, fixed, or insolubilized en- 5) Chemical and thermal stability
zymes, cells, etc., denote biocatalysts that are 6) Mechanical strength
bound to a support. The material to which en- 7) High rigidity and suitable particle form
zymes or cells are bound is called the carrier, 8) Resistance to microbial attack
support, or matrix. The agent enabling enzyme – 9) Regenerability
support binding is referred to as a cross-linking 10) Toxicological safety
agent, bifunctional agent, or carrier activator. 11) Low or justifiable price
For clarity, techniques for immobilization of en-
zymes and cells are discussed separately (Sec- Carriers can be classified according to their
tions 2.2 and 2.3, respectively). morphology (porous, nonporous, or gel type) or
their chemical composition. The most widely
used supports are listed in Table 2. The charac-
2.2. Methods of Enzyme Immobilization teristics of these materials are presented in [44],
(see also → Enzymes, Chap. 3.3.) [45]. Sources of commercial carriers are listed
in [46].
The definition of immobilized biocatalysts given
in Section 2.1 includes enzymes that are
Immobilized Biocatalysts 5
alumina, activated carbon, clay, diatomaceous beads ( AffiGel 10) to produce a very spe-
earth, glass, and hydroxyapatite. The problems cific, highly active biocatalyst starting from a
of enzyme immobilization by physical adsorp- crude enzyme preparation [57]. In the immo-
tion are discussed in [47], [48], and examples bilization of carboxypeptidase A [11075-17-5]
are given in [49], [50]. In some cases such prob- (E.C. 3.4.17.1), the highest enzyme activity was
lems can be eliminated by cross-linking the ad- shown with a mouse monoclonal antibody –
sorbed enzyme using multifunctional reagents enzyme complex on Eupergit C (Röhm Pharma)
(Section 2.2.2.2). or on a Protein A-Sepharose CL 4B (Pharmacia)
[58].
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].
an inert protein such as albumin (co-cross- generally not chemically bound to the matrix as
linking) and precipitating the resulting aggre- in enzyme copolymerization. Entrapment meth-
gate [24], [25], [49]. This method can also be ods can be subdivided into gel entrapment and
used in combination with a carrier such as a microencapsulation and reverse micelles.
membrane, where the physically adsorbed en-
zymes are cross-linked on the membrane sur- Gel Entrapment. In this procedure, free en-
face [65]. Bi- or multifunctional reagents such zyme is entrapped within the interstitial spaces
as glutaraldehyde [111-30-8], toluene diisocya- of a cross-linked, water-insoluble polymeric gel.
nate [1321-38-6], or bisdiazobenzidine deriva- Gel-forming materials such as polysaccharides,
tives can be used as cross-linking agents. The proteins, or synthetic polymers can be em-
primary disadvantages of this method are the ployed. Among the natural polysaccharides, al-
relatively severe reaction conditions and diffi- ginate, agar, and κ-carrageenan are widely used
culties in controlling the reaction. Hence, activ- in preparing immobilized biocatalysts.
ities of such biocatalysts are generally low [24], Polyacrylamide gel was the first material em-
[25], [40], [41], [66]. 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-
2.2.2.3. Enzyme Copolymerization diphosphate aldolase (E.C. 4.1.2.13)] [72]. The
polymerization of acrylamide can be initiated
The enzyme can be immobilized by copolymer- with potassium persulfate and by X- or γ-
izing it with the polymer matrix. Although the radiation [73]. Another method uses prepoly-
enzyme is then a covalently bound part of the mers that can be cross-linked by UV radi-
matrix, this method can be regarded as a modifi- ation; photo-cross-linkable resin prepolymers
cation of gel entrapment (see Section 2.2.2.4). with varying hydrophilic – hydrophobic ratios,
In the original method, proteins were viny- chain lengths, and anionic or cationic func-
lated with acylating or alkylating monomers and tional groups are available [74]. Such biocata-
copolymerized with other monomers. The ad- lysts can also be used for bioconversion of highly
vantages of this method over matrix entrapment lipophilic or slightly water-soluble substrates in
are the higher activity and higher stability due reaction media consisting of water and a water-
to the lack of enzyme bleeding [67], [68]. miscible organic cosolvent or water-containing
In enzyme graft copolymerization, enzyme organic solvents. Hydrophilic gel matrixes and
copolymer is grafted onto a matrix such as a organic solvents of low polarity are generally
polysaccharide carrier [69], [70]. High mechan- preferred for bioconversion of lipophilic sub-
ical strength can be attained, depending on the strates [75]. For further details of this method,
carrier chosen. High intra particle diffusion re- the literature should be consulted [20], [24],
sistances can be circumvented if the enzyme [41], [43], [45], [55].
copolymer is grafted to the surface of a carrier A special modification of this method in-
with low porosity [55]. For example, horseradish volves enzyme entrapment during the spinning
peroxidase (E.C. 1.11.1.7) was immobilized un- of synthetic fibers such as cellulose triacetate
der strictly controlled conditions to give a bio- [76], [77]. High surface areas can be achieved by
catalyst with kinetic constants close to those of using very fine fibers, which also show chemical
the native enzyme but with higher thermal sta- resistance to weak acid, alkali, and some organic
bility [71]. 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-
2.2.2.4. Entrapment biotics with a β-lactam ring) and food industry
(β-D-galactosidase, E.C. 3.2.1.23, is used for the
Entrapment may be regarded as the physical con- hydrolysis of lactose). The risk of microbial con-
finement of an enzyme in a semipermeable ma- tamination is one disadvantage of this method;
trix. The matrix lattice must be tight enough for a reactor densely packed with fibrous material
the enzyme to be retained but must allow perme- cannot be sterilized easily.
ation of substrate and product(s). The enzyme is
Immobilized Biocatalysts 11
growing cells can also be immobilized; this tech- 1) when the desired enzymes are intracellular
nique is used especially with eukaryotic cells and the extracted, purified enzymes become
where the whole metabolic machinery is often unstable after immobilization;
required for their specific application. 2) when the microorganism does not contain in-
The state of the art in immobilized cell tech- terfering enzymes or when such enzymes can
nology is described in detail in [21], [22], [25], be inactivated without loss of desired cat-
[40], [43], [66], [87–102]. As in enzyme immo- alytic activity; and
bilization, no “ideal” general method exists. This 3) when substrates and products do not have a
chapter merely gives an overview of the advan- high molecular mass and can diffuse through
tages and disadvantages of immobilized whole- the cell membrane.
cell systems and the most generally applied im-
mobilization techniques. The use of whole cells instead of isolated
The primary advantage of immobilized cell enzymes obviates costly, tedious purification
cultures as opposed to suspension cultures is procedures. In multienzyme reactions, the cell
the reduced cost of bioprocessing as a result structure provides the proper geometric en-
of higher cell densities, easy separation, and re- zyme arrangement which is extremely difficult
peated or continuous use of biocatalyst. Immo- to achieve by enzyme co-immobilization. For re-
bilization may also increase the shear resistance actions requiring cofactor regeneration or for the
of shear-sensitive eukaryotic cells. The enzymes production of primary or secondary metabolites,
retain their native geometry and microenviron- the immobilized cells must be viable. If cell vi-
ment, and are protected against changes in the ability is not a prerequisite, diffusion problems
bulk system (temperature, pH, ionic strength, can be solved by chemical or electrical perme-
mechanical stress). Support materials generally abilization of the cell membrane.
stabilize biocatalyst activity and storage stabil- The major advantage of such biocatalytic sys-
ity, but this phenomenon is not fully understood. tems is that the operational stability can often
Correct choice of the hydrophobicity – hy- be markedly enhanced by reactivating the en-
drophilicity balance of the support can alter the zymes in the immobilized cells [104]. Such re-
microenvironment by increasing the concentra- activation can be achieved either by brief incuba-
tion of limiting substrate or by decreasing the tion with growth medium or by continuous con-
concentration of inhibiting product in the vicin- trolled growth of the immobilized cells [105].
ity of the biocatalyst [92], [103]. Poor diffu-
sion and permeability into the carrier matrix and
through the cell membrane limit the applicabil- 2.3.2. Cell Supports
ity of such biocatalyst systems to low molecular
mass substrates or products. This transport bar- A variety of inorganic and organic matrices are
rier can protect the cells against microbial or available for cell immobilization; they interact
viral infection if the pore width of the matrix is with the cell surface by physical or chemical
sufficiently small [92]. bonds [106], [107]. A comprehensive list of
The reactions catalyzed by immobilized commercial support materials is given in [46].
whole-cell biocatalysts can be classified as fol- A number of factors must be considered in the
lows: selection of a suitable carrier for large-scale use
[98], [100], [103]:
1) Reactions involving single enzymes (biocon-
versions) 1) The material should be available in sufficient
2) Reactions involving multienzyme systems quantities and at low price.
with or without cofactors 2) The material should have a large surface area
3) Reactions involving a complete metabolic accessible to cells and reactants.
pathway yielding primary or secondary 3) The material must be mechanically, chemi-
metabolites cally, and thermally stable under process and
storage conditions.
The use of immobilized microbial cells is ad- 4) The matrix should contain a sufficient num-
vantageous in the following areas [96]: ber of functional groups to bind the cells.
Immobilized Biocatalysts 13
5) The material should not reduce cell activity 2.3.3.2. Binding of Cells to a Carrier
or initiate cell lysis.
6) The material should be easy to handle in the Physical Adsorption. Immobilization of
immobilization procedure. microorganisms or cells by adhesion or adsorp-
7) The material should be capable of recycling tion to a preformed matrix is based on their in-
or safe disposal. herent affinity for solid surfaces during growth.
8) In the case of viable growing cells, the matrix Such immobilization can generally be achieved
should have a sufficiently large void volume by keeping the inert carrier and the actively
or be elastic enough to accommodate new growing cells in contact for a defined period.
cells. Additional chemicals are usually unnecessary,
and fixation is carried out under growth con-
ditions; viable cell preparations can, therefore,
2.3.3. Immobilization Techniques be obtained. Examples of cells immobilized
by adhesion or adsorption to various support
Most of the common cell immobilization tech- materials are listed in [100], [104].
niques are extensions or modifications of those Although the method is very simple, mild,
discussed in Section 2.2.2 for enzymes. These and versatile, its use is limited by adhesion ca-
methods can be grouped according to the way pacity and adhesive strength, which usually de-
in which cells bind to the matrix (Fig. 3). An al- pend on environmental conditions (pH, ionic
ternative way of classifying these methods is to strength) [93]. Use of a high flow rate as a
use the route by which the biocatalyst is prepared method of selection in a fluidized bed yields
[103]: good results. Such surface-attached cell systems
1) Carrier-free immobilization ( biofilms) have been used extensively in waste-
2) Immobilization of a given biomass onto a water treatment [95], [109], [110].
preformed carrier surface A promising application of this method is the
3) Immobilization of a given biomass during the use of mammalian cells bound to preformed sur-
course of carrier formation (e.g., by polymer- faces to produce therapeutic biochemicals. The
ization) cells are immobilized on microcarriers (small-
4) Immobilization by controlled growth of an diameter beads, 100 – 200 µm) manufactured
inoculum or by germination of immobilized from different synthetic polymers (e.g., polysty-
spores rene, gelatin, dextran, polyacrylamide, or glass)
that offer a large specific surface area for cell
growth [111–113]. The primary weakness of this
2.3.3.1. Cell Immobilization without a method is the exposure of surface-grown cells
Support to mechanical stress resulting from bead – bead
collisions and impeller impact. The number of
The simplest way to immobilize biocatalysts is cell divisions is also restricted by the limited
to use their intrinsic tendency to aggregate or outer surface area. Both problems have been
flocculate at high cell densities. Yeast, mold, and partially solved by the development of macrop-
plant cells aggregate [90], [95], [99], [104] and orous microcarriers prepared from gelatin, dex-
are relatively stable to shear fields in fluidized- tran, polyacrylamide, or agarose [113], [114].
bed reactors [99], [108]. Flocculation can also Cells grow on the much larger interior surface
be induced by polyelectrolytes such as chitosan of the porous structure and are also protected
[93], [99], [104]. against mechanical stress.
Cell aggregation can be induced by low A monolithic ceramic matrix has been de-
molecular mass bi- or multifunctional reagents veloped for large-scale cell cultures. It consists
such as glutaraldehyde, diazotized diamines, or of a ceramic cylinder with uniform longitudinal
toluene diisocyanate, but the toxicity and the channels of square cross section. The matrix has
high reactivity of these substances limit their ap- a surface area of 32 cm2 per cubic centimeter of
plicability in immobilizing viable cells. Exam- volume. A matrix with a very smooth surface
ples of immobilization by intercellular aggrega-
tion are given in [93], [100], [104].
14 Immobilized Biocatalysts
is used for the growth of anchorage-dependent ity of the coupling agents often results in loss of
cells. Nonadherent cell types (e.g., hybridoma cell viability or enzyme activity [66], [104].
cells) are cultured in a matrix with a very rough The main advantages of this method are re-
surface and are physically entrapped in the pores duced or eliminated cell washout and improved
of the channel walls. Unlike microcarrier cell mechanical strength of the biocatalyst without
cultures, no shear problems occur in the ceramic increasing diffusion barriers. The problems as-
matrix [113], [115]. sociated with various organic and inorganic sup-
For adherent cell growth, the scalability is ports, along with coupling techniques, are dis-
almost linear and depends on the available sur- cussed in [104]; applications are reviewed in
face area. For immobilized hybridoma cultures, [100]. The preparation, properties, and exper-
scaleup is based on the volume of the porous imental applications of eukaryotic cells cova-
matrix [116]. lently bound to solid supports are described in
[117].
Ionic Binding. Ionic binding is a special
case of physical adsorption where charged mi-
crobial cells can electrostatically interact with 2.3.3.3. Immobilization of Cells by
the ions on a carrier surface to form stable com- Entrapment
plexes. Synthetic ion-exchange resins, modified Cells may be immobilized by entrapment with a
cellulose derivatives, or inorganic materials can polymer lattice, a membrane, or microcapsules.
be used as carriers. Cell adsorption is mainly
affected by factors such as pH, ionic strength, Lattice Entrapment. The widely used
surface charge, cell age, or composition of the method of lattice entrapment involves retention
carrier surface. Examples for the application of of cells within the network of a polymer matrix.
this method are given in [93], [98]. The lattice is tight enough to prevent release of
cells, while allowing diffusion of substrates and
Covalent Binding. Cells can also bind to the products. The cell does not bind to the matrix,
functional groups of the carrier surface by cova- and the polymer network is generally formed
lent bonds. This technique is frequently used for in the presence of the cells to be entrapped.
enzyme immobilization but has only limited use Hydrophilic porous matrices are formed under
in whole-cell immobilization because the toxic- mild enough conditions to allow cell entrapment
Immobilized Biocatalysts 15
with minimum loss of viability and activity. The integrate, the beads. Another drawback of the
prepared biocatalyst forms pellets which should alginate method for viable animal cells is the
have good mechanical and chemical stability, as difficulty of producing sufficiently small beads
well as high cell loading capacity. The high vi- to overcome oxygen limitation in their interior
able cell density results in an enhanced demand [113].
for nutrients and oxygen, which must overcome Strengths and weaknesses of the calcium al-
the diffusion barriers of the gel structure. ginate method are discussed in [122]; examples
In this section, only the most extensively used of applications are given in [96], [98], [100].
matrices (e.g., polyacrylamide gel, alginate gel, κ-Carrageenan [11114-20-8] is a readily
κ-carrageenan, and photo-cross-linkable resins) available, nontoxic, high molecular mass, sul-
are discussed. More detailed information on fonated polysaccharide extracted from seaweed.
other systems can be found in [21], [90]. It consists of β-d-galactose 4-sulfate and 3,6-
Polyacrylamide Gel. Polyacrylamide gel anhydro-d-galactose units. Only the sodium salt
was used to immobilize whole cells in the of κ-carrageenan is soluble in cold water. The
first successful application of lattice entrapment gelation characteristics of κ-carrageenan are dis-
used for the biotechnological production of bio- cussed in [123]. Induction of gelation with potas-
chemicals [118]. The same technique was also sium ions can be performed under very mild
employed in the production of l-aspartic acid conditions without the use of chemicals that de-
[119]. The method is based on free-radical poly- crease enzyme activity. Another advantage of
merization in an aqueous acrylamide monomer this technique is that the immobilized cell bio-
solution containing the cells at low tempera- catalysts can be tailor-made into pellets, sheets,
ture. The degree of cross-linking determines the beads, etc. [96].
porosity and mechanical properties of the pel- Two entrapment procedures are available. In
lets. The major disadvantage of the method is the the one-step procedure, a large number of cells
toxicity of the acrylamide monomer, the cross- are entrapped in the gel. The cell suspension
linking agent (e.g., N,N -methylenebisacrylam- in physiological saline is mixed with soluble κ-
ide [110-26-9]), and the polymerization initia- carrageenan gel under sterile conditions, and the
tor (e.g., N,N,N ,N -tetramethylethylenediamine gel beads are incubated in a nutrient medium
[110-18-9]) which can decrease cell viability [98], [122]. Another very elegant method in-
and enzyme activity [96], [120]. Hydroxyethyl volves the use of photo-cross-linkable resins
methacrylate [868-77-9] can be used as a less [74], [124], [125]. Active groups (e.g., vinyl
toxic alternative to acrylamide [43], [121]. groups) are coupled to polyethylene or poly-
With acrylamide monomers, optimal results (propylene glycol) oligomers of controlled chain
are achieved in isotonic buffered solution at length to prepare prepolymers. The prepolymers
low temperature (4 – 12 ◦ C) by keeping expo- are then mixed with the cell culture, and cross–
sure time to a minimum (3 – 4 min). Examples of linking is initiated with UV radiation. The ad-
successful applications of this method are given vantages of this method are:
in [96], [98].
Alginate [9005-38-3] is extracted from sea- 1) entrapment is simple and proceeds under
weed and is a linear copolymer of β-d- very mild conditions;
mannuronic acid and α-l-guluronic acid linked 2) prepolymers do not contain toxic monomers;
by 1,4-glycosidic bonds. It forms a gel in the 3) the network structure of the gels can be
presence of multivalent ions, usually calcium or adapted as required; and
aluminum. The controlled entrapment of cells 4) optimal physicochemical gel properties can
is simple and generally nontoxic. Various cell be achieved by selecting suitable prepoly-
types can be immobilized with negligible loss mers [125].
of viability. However, the matrix can be solubi-
lized in the presence of Ca2+ -chelating agents Membrane Entrapment. In membrane en-
such as phosphate, citrate, or ethylenediamine- trapment, the cells are not held in a porous ma-
tetraacetic acid. The alginate matrix is mechan- trix (as in lattice entrapment), but are retained by
ically weak so that the growing cells (especially a semipermeable membrane barrier. The mem-
plant cells) can be released from, or even dis- brane allows the diffusion of soluble material
16 Immobilized Biocatalysts
to and from the immobilized cells while retain- into organelles such as mitochondria, chloro-
ing and protecting enclosed organisms. Hollow- plasts, peroxisomes (microbodies), vacuoles,
fiber bioreactors, in which the organism is con- and nuclei. Such organelles have their own dis-
fined on one side of the porous fiber and the sol- tinct metabolic functions and, therefore, pro-
uble substrate and products on the other, have vide excellent immobilized biocatalysts that can
found the widest application. Higher cell densi- carry out multistep, multifunctional reactions.
ties can be achieved than with lattice-entrapped Immobilization generally increases organelle
biocatalysts [126], but growth must be con- longevity.
trolled to prevent excessive buildup of biomass Examples of the uses of immobilized or-
and rupture of the membrane [99]. The supply ganelles are adenosine 5 -triphosphate (ATP) re-
of oxygen and nutrients to the cells and the re- generation, metabolite oxidation, and photopro-
moval of carbon dioxide from the cells occur by duction of hydrogen [136]. Most research has
diffusion and may be insufficient. Model equa- been devoted to the isolation and immobilization
tions for mass transfer in hollow-fiber reactors of chloroplasts as potential solar energy photo-
are discussed in [127], [128]. converters [136], [137]. Older work on organelle
The state of the art of membrane reactors is immobilization is reviewed in [137], [138]. En-
described in [84]. Developments in hollow-fiber trapment techniques using photo-cross-linkable
reactors are discussed in [129], and for flat-bed prepolymers are described in [125].
hollow-fiber reactors in [130]. A flat membrane Immobilization of rat liver mitochondria on
reactor with different compartments for cells, a solid support consisting of alkysilanized glass
medium, and product for large-scale cultivation beads was first reported in 1975 [140]. Rat liver
of mammalian cells is described in [131]. mitochondria have also been entrapped in poly-
acrylamide gel [141]; yeast mitochondria have
Microencapsulation. In microencapsula- been entrapped in a polyurethane matrix [142].
tion, cells are placed inside an alginate bead
that is coated with a semipermeable membrane.
The membrane is usually composed of calcium 2.5. Co-immobilization of Biocatalysts
alginate and poly-l-lysine. After being coated
the alginate gel can be solubilized with sodium Co-immobilization refers to the simultaneous
citrate, leaving the cells entrapped in hollow immobilization of specific combinations of
spheres where they can be cultured. Different cells, organelles, or enzymes that show com-
capsule diameters (from 20 nm to 2 mm) and plementary biocatalytic activity. The synergis-
controlled membrane porosity can be achieved tic advantages of co-immobilization are, how-
[127]. In larger microcapsules, however, radial ever, limited by the biological compatibility of
concentration gradients of nutrients or oxygen the components [143].
can result in a necrotic core [127] . Co-immobilization can be achieved by coen-
This well-known enzyme immobilization trapment of preimmobilized enzymes and whole
technique (Section 2.2.2.3) was first adapted for cells in a single matrix or by direct coupling of
the whole-cell immobilization of Micrococcus enzymes to the microorganisms [25], [143]. The
denitrificans [132] and of viable mammalian former method was used to immobilize galac-
cells [133]. Applications of this type of immo- tosidase and Saccharomyces cerevisiae for the
bilized biocatalyst [134] include the production conversion of cellobiose to ethanol [144]. Yeast
of monoclonal antiboides from hybridoma cells cells or dead mycelia have been employed for
[135]. direct coupling of enzymes to cell walls in food-
processing applications [25], [143]. Another in-
teresting application is the bioproduction of hy-
2.4. Methods of Organelle drogen gas with hydrogenase and chloroplasts
Immobilization [145]. Co-immobilization results in a fivefold
increase in hydrogen production compared with
Enzymes in living cells may be associated in the system employing two separately immobi-
supramolecular structures, which are assembled lized catalytic species. Recent results on integral
Immobilized Biocatalysts 17
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.
for enzymatic reactions of positive order such (i.e., external and internal mass transfer; see
as Michaelis – Menten kinetics or competitive Chap. 4).
product inhibition. A catalyst whose activity Mass-transfer resistance must be considered
increases toward its core (concave distribu- in kinetic measurements or when modeling
tion) can give greater utilization efficiency for processes with immobilized biocatalysts. The
substrate-inhibited reactions [155]. familiar Michaelis – Menten equation can be
adapted by using the effectiveness factor η:
vmax CSA
3.2. Kinetics of Immobilized rS =η
KM +CSA
Biocatalysts
or in its linearized form
The rate of reaction catalyzed by a soluble en-
1 KM 1 1
zyme can usually be described by a Michaelis – = +
rS ηvmax CSA ηvmax
Menten type of kinetic equation (→ Enzymes,
Chap. 2.). The simplest mechanism for such a re- where C SA is the substrate concentration at the
action assumes the reversible formation of an en- surface. Since both the substrate concentration
zyme – substrate (ES) complex, which decom- at the surface C SA and the effectiveness fac-
poses irreversibly into product (P) and free en- tor η can vary with the bulk substrate concen-
zyme (E): tration C S , the double reciprocal Lineweaver –
Burk diagram and other linearized forms of the
S + E ES −→ E + P Michaelis – Menten equation are no longer lin-
ear. In Figure 6 it can be seen that the influence of
The well-known Michaelis – Menten equa-
mass transfer decreases at high substrate concen-
tion is
tration and disappears totally when this concen-
vmax CS tration becomes infinite. The effectiveness factor
rS = η changes with the bulk substrate concentration,
KM +CS
reaching its lowest value η (0) when C SA = 0.
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:
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].
Consider the reaction of a substrate which is concentration. Thus, the apparent kinetics of an
transferred from the bulk liquid to an enzyme immobilized enzyme differ from the intrinsic ki-
immobilized on a nonporous surface (point B in netics of the native enzyme. This can be seen
Fig. 7). For a first-order reaction (or Michaelis – most easily from Equation (5) when S B is low. In
Menten kinetics with S < K M ) at steady state this case, the surface rate reduces to vmax S B /K M
(i.e., diffusion equal to surface reaction rate), and the apparent rate becomes
the apparent reaction rate per unit surface area
(r app ) is (vmax /KM ) SA
rapp = (6)
(vmax /KM kSL ) +1
rapp =kSL (SA −SB ) =ks SB (1)
i.e., the apparent rate depends on the transfer
where S A is the substrate concentration at point coefficient k SL .
A, S B the substrate concentration at point B, k SL Generally, true kinetics can be determined
the mass-transfer coefficient for the liquid film only by operating in the kinetic regime with-
at the solid – liquid interface, and k s the surface out diffusion. This can be achieved by reducing
reaction rate constant. Solving for S B and rear- the ratio of vmax /k SL sufficiently as seen from
ranging give Equation (6). Thus, transfer-promoting param-
eters (flow rate in packed and fluidized beds,
ks stirring speed in tanks) should be increased, and
rapp = ·SA (2)
(ks /kSL ) +1 those that promote reaction rate should be de-
creased (enzyme loading, temperature).
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 4.2. Prediction and Correlation of
approaches S A , and the reaction is kinetically External Mass-Transfer Coefficients
controlled with an apparent rate of k s S A . In the
intermediate situation given by Equation (2), the 4.2.1. Catalyst Beds
apparent reaction rate is influenced by both the
rate constant (k s ) and the mass-transfer coeffi- The approximate value of k SL for a bed of
cient (k SL ). packed or fluidized particles can be predicted by
For a zero-order reaction, using data from the literature, which are usually
presented as dimensionless correlations having
kSL (SA −SB ) =kS (3) the form
where k s is a zero-order rate constant. The sub- n1 n2
strate concentration at the reaction surface (S B ) kSL dp dp vs µ
∼ (7)
is thus D µ D
Correlations of the above form can be found being used to calculate a terminal velocity mass-
in [161–163]. For the region of low relative ve- transfer coefficient k SLt [164]. Much of the data
locity, theory and experiment show that when can be approximated by
the Péclet number N Pe (= d/D) is greater than
500 [165], the following holds: kSL = 2kSLt
Variable Exponent
4.3. Internal Mass Transfer [161], [166], diffusion rates in and out must be equal. If reac-
[167] tion occurs, the difference between the diffusion
rates is equal to the reaction rate. Balances of this
The uptake of substrates and products within form can be written for any component; a com-
solid matrices (internal mass transfer) occurs by plete description usually requires balances for
diffusional transport, whose driving force is a every component participating in the reaction.
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 Figure 8. Flux of substrate A (initial bulk concentration
C A0 ) across a solid matrix
causes a flux of substrate A from left to right, a) No reaction; b) Reaction
and the flux rate jA at any point in the matrix is
given by Fick’s law The reaction rate at each position in the im-
mobilized biocatalyst matrix is generally a func-
dCA tion of local substrate and product concentra-
jA = −DA−M
dZ tions and thus varies from position to position.
where DA−M is the diffusivity of A in the matrix These local rates cannot normally be measured;
material. If no reaction occurs in the matrix, a only the net rate can be determined by monitor-
steady-state concentration gradient with a con- ing concentration changes in the bulk liquid.
stant slope is established (solid line, a, in Fig. 8). Consider a situation in which the substrate
Because the gradient is constant, the amount of diffuses into the matrix from one side and does
A entering any region from the left is equal to not penetrate completely because it is consumed
the amount of A leaving at the right and no ac- by reaction with a biocatalyst (Fig. 9). Under
cumulation occurs within the matrix. steady-state conditions, the diffusion rate of sub-
If cells or enzymes are present within the ma- strate into the matrix equals the net reaction rate.
trix which convert substrate A to product, the Thus, the flux J A multiplied by the matrix sur-
substrate profile is curved (curve b in Fig. 8). face area A can be equated to the average volu-
This occurs because the substrate is consumed metric reaction rate r avg multiplied by the vol-
in the reaction and the flux of substrate from the ume of the matrix V M
left exceeds that leaving at the right. A mass bal-
ance can be written over a segment of the matrix. jA,Z=0 A=ravg VM
Under steady-state conditions, the accumula- In both batch reactors and well-mixed contin-
tion is zero, and in the absence of reaction the uous reactors, the diffusion flux at the interface
24 Immobilized Biocatalysts
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.
Effectiveness Factor. Analytical solutions greater than 1. For zero-order kinetics, the lim-
of the general diffusion – reaction equation for ited penetration of substrate causes η to become
simple reaction orders are given in [161], [166]. less than unity.
In Figure 13 the values of η for zero, first, and
second reaction order are plotted against the Experimental Strategies for Overcoming
square root of the Thiele modulus ϕ where Internal Diffusion. In practice, η must be in-
creased to 1 if diffusion-free enzyme or cell ki-
ϕ= L2 kS0n−1 /D 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
where DS−M is the diffusivity of substrate in the 7) control of environmental conditions (temper-
matrix and S 0 is the concentration in the liquid. ature, pH, substrates, inhibitory and toxic
The final model for numerical solution then products);
consists of N equations for each component in 8) substrate availability and waste treatment;
the solid matrix plus equations for each compo- and
nent in the liquid phase. Solution can easily be 9) aseptic operation.
made using a simulation language on a digital
Most of the reactors can be run in a batch, fed-
computer [166].
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
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
(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
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 external mass transfer to small reactive parti-
production units are of the packed- or fixed- cles consisting of immobilized cells, enzymes,
bed type (Fig. 15D) [172–174]; they are usually or biofilms. This is important because grow-
tubular plug-flow reactors. In addition to axial ing organisms would otherwise block fixed- or
gradients, internal and external diffusion pro- packed-bed columns. Fluidized-bed reactors are
cesses may cause gradients at or within the bio- of the tubular type and models are, therefore,
catalyst. In large-scale reactors, the mechanical similar to those for an ideal plug-flow reac-
strength of biocatalyst particles, pressure drop, tor. Figure 18 shows a large-scale anaerobic
and control are important. Deviations from ideal fluidized-bed reactor for wastewater treatment
plug flow may be observed, especially if chan- [175]. A minimum flow rate is required because
neling occurs in the bed. Applications of fixed- fluidization occurs at a minimum superficial liq-
bed reactors are listed in Table 8. 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.
5.5. Fluidized-Bed Reactors The recycle ratio decreases with increasing re-
actor size. Small laboratory-scale fluidized-bed
Fluidized-bed reactors (Fig. 15E) have found in- biocatalyst reactors are, therefore, often run as
creased application because they provide high
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]
requires measuring the steady-state difference flow rate on the overall reaction rate is, there-
between outlet and inlet concentrations and the fore, of interest; a convenient reactor configu-
mean residence time (Eq. 17). The CSTR usu- ration is shown in Figure 19. The system has a
ally yields better results, except when reaction reactor that holds the biocatalyst particles, either
rates are very low. as a packed or fluidized bed. Variable flow rates
In a batch experiment, substrate and product through the reactor are achieved by recycle. The
concentrations vary with time and may be dif- reactor is usually operated at low residence time
ficult to monitor and interpret. Suitable design (high flow, small volume) so that the conversion
of experiments (e.g., variation of initial concen- per pass is small. The axial gradients are then
trations, supply of an excess of one component) sufficiently low for concentrations in the reactor
is, therefore, required to establish reaction rate to be approximated by the average between inlet
kinetics and determine rate constants. and outlet. Thus, steady-state balances take the
At steady state the CSTR can be used to form
measure rates. New steady states can be estab-
lished by changing the flow rate, inlet concen- 0 =FR (Ci0 −Ci1 ) ±ri,avg Vr
tration,etc. where F R is the flow rate of the recycle stream,
Evaluation of experimental data may be per- ri, avg is the reaction rate at the mean concentra-
formed in serveral ways: tion Ci, avg , and V r is the volume of the reactor.
1) Differential equations can be integrated al- The difference Ci0 − Ci1 must be large enough
gebraically and rearranged to give a lin- for accurate measurement but small enough for
ear function for kinetic parameters. In this average concentration to be a valid approxima-
way, model applicability and model pa- tion.
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-
Figure 19. Differential recycle loop reactor for kinetic stud-
tion methods may help greatly in optimal de- ies
sign of kinetic experiments. a) Sampling sensor; b) Biocatalyst bed; c) Sensor and control
(pH, temperature, etc.); d) Flow meter
A short summary of experimental strategies
to determine rate parameters for complex bio- The recycle system shown in Figure 18 can
chemical kinetics is given in [179]. be operated batchwise, with no flow entering or
leaving the recycle loop and a small amount of
conversion occurring per pass. The balances be-
5.7.2. Experimental Differential Recycle come
Reactor
dCi
VT = ± ri Vr
Reactions occurring in and on particulate solids dt
(e.g., in beads of immobilized enzymes or whole where V T is the volume of the total system. Cor-
cells) are generally sensitive to external mass responding balances for a continuously fed re-
transfer (Section 4.1). Study of the influence of cycle system can be set up [169].
Immobilized Biocatalysts 33
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 gen ions (pH), oxygen, hydrogen peroxide, and
detection ammonia]. Because such compounds and ions
4) Optodes: optical detection by using light ab- are often involved in enzymatic reactions as re-
sorption, fluorometry, or luminometry actants or products, these electrodes can be em-
5) Thermistor probes: calorimetric measure- ployed in conjunction with a suitable enzyme to
ment of heat of reaction monitor substrate concentration. Important re-
6) Transistors: similar to electrodes, but semi- views on enzymes and microbial sensors are
conductor devices given in [232–237].
An extensive literature exists on analysis via
immobilized biocatalysts (see Table 11).
Table 11. Reviews on analytical application of immobilized
biocatalysts
ence
electrode detects oxygen consumption which is outside the body (intra- or extracorporeal). Re-
proportional to reaction rate and thus to glucose view articles in this rapidly developing field are
concentration. At low glucose levels, the result- given in [238–240].
ing electrode signal is linearly related to this con- Intracorporeal methods require immobiliza-
centration. Alternatively, a hydrogen peroxide tion in or on microparticles or blood cell frag-
electrode can be used. ments, or with macromolecules [238]. Distinc-
Enzyme electrodes have been constructed for tion is made between enzymes circulated with
at least 40 different substances, but very few the blood and those used for local treatment of
have been commercialized (Table 12). The most specific organs. Soluble polymers and micropar-
important factor is enzyme stability, which is ticles can be used for immobilization in the blood
critical for the lifetime of the enzyme electrode, stream. Slow release by diffusion or biodegrada-
and depends largely on the immobilization pro- tion from particles is employed for local use.
cedure. Stabilities of three to four months have Enzymes introduced into the body may have
often been attained. the advantage of targeting specific sites but in-
Table 12. Examples of enzyme electrodes
crease the danger of toxic or immune response.
The carrier and the immobilized enzyme must
Analysis Enzyme system Electrode exhibit a minimum of allergic and immuno-
Alcohols alcohol oxidase H 2 O2 genic effects. Fortunately, immobilization often
Amino acids l-amino acid oxidase H 2 O2 reduces immunogenicity. A partial list of intra-
Cholesterol cholesterol oxidase O 2 , H 2 O2 corporeally employed enzymes, their therapeu-
Creatinine creatinase NH+ 4 tic use, and the type of carrier employed is given
Glucose glucose oxidase on poly(vinyl O 2 , H 2 O2
chloride) membrane in Table 13. Microencapsulation methods are
Sucrose invertase, mutarotase, and important for a variety of enzyme drug applica-
glucose oxidase on collagen O2 tions and have the following advantages [238],
Urea urease on Teflon NH+4
[240]:
1) ultrathin (20 nm) synthetic membrane cap-
Microorganisms can also be immobilized on sules;
the outside of the electrode. Thus, for exam- 2) small particle size (100 µm);
ple, phenol can be detected by using an oxy- 3) high transfer rates;
gen electrode together with phenol-oxidizing 4) biological particles (normal cells, cell ghosts,
microorganisms. Stabilities of 5 – 30 d have been synthetic liposomes);
achieved for over 20 compounds by using immo- 5) wide range of immobilization applications
bilized organisms [237]. (single enzymes, enzyme mixtures, whole
For certain analyses, the immobilized en- cells, absorbents);
zyme is better employed in a reactor. The sample 6) intra- and extracorporeal use;
is passed through the reactor, and the concentra- 7) possibility of biodegradable materials (pro-
tion of the desired substance in the reactor efflu- teins, lipids);
ent is then measured downstream by an appro- 8) wide range of administration methods (oral,
priate electrode sensor. This arrangement allows intravenous, local injection); and
adjustment of the enzyme quantity. 9) minimal immune reactions.
Extracorporeal application requires access
7.2. Immobilized Enzymes in to the blood stream and anticoagulant treatment
Therapeutic Medicine [239]. The equipment is usually a small immo-
bilized enzyme reactor connected to the patient
The ability of enzymes to carry out specific
(much like a hemodialysis unit) and, therefore, is
biological reactions is being exploited as a
not suitable for continuous treatment. To obtain
therapeutic method in medicine (→ Enzymes,
the necessary activity, a large surface area of par-
Chap. 5.4.). A number of applications have been
ticles (packed and fluidized beds) or membranes
tested on animals, but very few on humans.
(hollow-fiber cartridges) is necessary, which in-
The enzymes require immobilization and con-
creases the danger of blood clotting. Extracor-
tact with body fluids and organs, either inside or
36 Immobilized Biocatalysts
Table 13. Methods of immobilizing intracorporeal enzymes
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
Immobilized Biocatalysts 37
Table 15. Industrial applications of immobilized cells [245]
close to the theoretical maximum for the equilib-
Product Cell system 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
Acetic acid Acetobacter on wood shavings 1880
in a trickle-filter reactor pellets (0.3 – 1 mm), and dried [246]. Activity
l-Aspartic acid Escherichia coli in various gels 1973 – decreases to 10 % in 127 d.
1976 One of the largest enzyme suppliers for this
High-fructose corn various systems 1966 –
syrup 1974 process, the Novo Company in Denmark, gives
l-Malic acid Brevibacterium in carrageenan 1974 – the following information (Table 16) [246]:
gel 1977
Penicillin gel-entrapped Escherichia coli 1969 – Table 16. Operational details of the Novo Sweetzyme reactors
1978 [246]
Raffinose hydrolysis pellets of Mortierella fungus 1968 –
1974 Parameter Comments
Prednisolone Curvularia and Arthrobacter 1978
(steroid coentrapped in polyacrylamide Biocatalyst mutant Streptomyces murinus cells, disrupted
transformation) gel 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 )
7.4.1. Production of High-Fructose Corn two lines of four reactors in series (with recycle)
or parallel
Syrup 5240 kg biocatalyst in all reactors, 30 kg replaced
per day, 16 day replacement
High-fructose (42 %) corn syrup (HFCS) from flow rate of syrup (dry basis) 100 t/d
cornstarch is by far the most important prod- Feed 40 % solids (95 % glucose)
composition
uct of immobilized biocatalyst technology; the Product 42 % fructose, 53 % glucose, 5 %
estimate annual worldwide production for 1982 composition oligosaccharides, 70 – 75 % total solids after
was 6×106 t on a dry basis (5.5×106 t in the evaporation
United States by 11 companies and 0.4×106 t Control cycle flow adjustment necessary due to
decreasing catalyst activity to keep fructose
in Japan by 16 companies). Glucose from corn- product concentration constant. Flow varies
starch is converted to fructose with glucose iso- linearly from 10 to 8 m3 /h over 400 h
Catalyst activity decreases to 10 % in 127 days, catalyst
merase [9055-00-9] (E.C. 5.3.1.18) in the form lifetime replaced alternately in the eight reactors every 16
of a variety of immobilized whole cells. Im- days
mobilization methods include gel entrapment Conditions reactor outlet pH 7 – 7.5, feed pH 7.5 – 8.2,
Mg : Ca ratio 15, 60 ◦ C, oxygen-free
with cross-linking, binding on ion-exchange Yield 0.44 t of fructose per tonne of glucose
resin, flocculation with electrolytes, extrusion Productivity 3.3 t HFCS (dry) per kilogram of catalyst
of cells and cross-linking, and heat treat- 1.4 t fructose (dry) per kilogram of catalyst
Daily 100 t HFCS (dry)
ment (→ Enzymes, Chap. 4.2.1., → Enzymes, production
Chap. 4.3.; → Glucose and Glucose-Containing rates
Syrups, Chap. 6.2.). Developed in Japan in 1966, 43 t fructrose
16.4 t fructose per kilogram of catalyst
the rights to the process were acquired by a U.S. 5.7 t/m3 HFCS (dry)
company which by 1978 had improved the glu-
cose isomerase enzyme production of Strepto-
myces by mutation by a factor of 200 [245]. The syrup is purified by filtration, concen-
The process starts with a relatively inexpen- trated from 30 % to 40 % by evaporation, treated
sive raw material, corn. Cornstarch is converted with activated carbon to remove soluble impuri-
to glucose syrup by soluble amyloglucosidase ties (peptides and amino acids), and by ion ex-
from Aspergillus niger in a batch process. The change to remove calcium ions. Oxygen must
glucose-containing syrup is then purified and finally be removed before the syrup enters the
treated by a continuous process using immo- fixed-bed isomerization reactors. Purification is
bilized glucose isomerase. The resulting fruc- critical for obtaining high product quality and
tose composition is typically 42 %, which is maximum enzyme stability.
38 Immobilized Biocatalysts
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 (2) absence of clogging and thus low-pressure
Pseudomonas dacunhae in a pressurized drop flow properties for packed beds, and (3)
packed-bed column. The pressure keeps the sufficiently high-density particles for fluidized
carbon dioxide in solution, which results in a beds.
smaller pH change and less axial mixing. The l- Plastic foam and other light porous material
aspartic acid substrate is produced by immobi- can be added to conventional activated sludge
lized E. coli cells as described previously, mak- reactors to entrap biofilms for biomass reten-
ing this method the first use of two different im- tion. Another type of reactor, the upflow anaero-
mobilized cells in series (Fig. 23). bic sludge blanket reactor, uses the tendency of
anaerobic cultures to form granules by natural
flocculation and to thus be retained in a column
7.4.3. Environmental Applications reactor with slow upward flow.
Most of the above reactors are classified as
Mixed microbial cultures can be adapted by con- high-rate processes and have been tested on var-
tinuous techniques under anaerobic and aero- ious scales. They have found limited large-scale
bic conditions to attach and adhere to almost application in the biological treatment of indus-
any surface and to degrade a variety of wastes. trial waste. The field is still in the developmental
The organisms develop into a thin film termed a stage. Review articles provide more detailed in-
biofilm, which is typically 0.1 – 0.5 mm thick. formation [251–255].
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- 8. Economic Aspects
actor volume to give a high activity per unit vol-
ume. Retention of the biomass as a biofilm is par- The world biotechnology market has a very high
ticularly important for slow-growing mixed bac- growth rate: 25 % of the world pharmaceuti-
terial cultures, such as those involved in anaer- cal market is already covered by biotechno-
obic treatment of food industry wastes, aerobic logically manufactured products. In the chem-
nitrification for ammonia removal, and removal ical industry, this fraction is approaching 10 %.
of toxic organic wastes, such as phenol. Immobi- According to market growth projections, U.S.
lization allows the process to be operated at high biotechnology industry sales increased from
flow rates, which exceed the rates that would about $ 400×106 in 1987 to $ 2.5×109 by 1990,
otherwise wash suspended organisms out of the and more than $ 25×109 by 2000 [256]. The
reactor. Retention of the active biomass in large world biotechnology market in the year 2000
quantities allows the reactor to operate at rates was predicted to be $ 65×109 [257].
up to 50 times higher than those for nonimmo- An algorithm has been developed for prelim-
bilized systems. inary analysis of the economic viability of an
Biofilm reactors can be designed in a vari- immobilized biocatalyst production process for
ety of ways, depending on whether the process high-value, intermediate-value, and commodity
is anaerobic or aerobic. Column reactors lend products [258]. A detailed list of market values,
themselves to both processes. Packed beds of in dollars per kilograms, and U.S. market sizes,
solid particles (e.g., bricks, stones, or plastic me- in kilograms per year, of potential biotechnolog-
dia) retain organisms by entrapment and film ical products is given in that article. Manufactur-
growth, and are known as aerobic trickling fil- ing costs of immobilized cell systems are dis-
ters or anaerobic filters. Expanded or fluidized cussed along with the effects of feedstock costs
beds use small particles (e.g., sand) which are and biocatalyst characteristics (e.g., selectivity,
lifted by the upflowing liquid. The organisms longevity, inherent productivity, and yield) on
must form a tight biofilm, otherwise they would process economics [258].
be swept out of the reactor. A variety of other ma-
terials (plastic media, honeycomb-type packing, Immobilized Enzymes. The total market for
wound steel wire particles, activated charcoal) industrial enzymes was about $ 400×106 in
have been used for biofilm development. Impor- 1987 [259] and is growing by about 15 % each
tant requirements are (1) maximal surface area, 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 tremely sparse. A theoretical economic analysis
[260]. Soluble use-and-discard enzymes have, of the advantages and disadvantages of a contin-
by far, the most dominant market share. Only uous immobilized enzyme process, compared to
one immobilized enzyme product, immobilized a batchwise soluble enzyme system, is presented
glucose isomerase, is used in amounts > 50 t in [262].
a year; 1500 – 1750 t are used worldwide an-
nually [261]. The three other major immobi- Immobilized microorganisms are pre-
lized enzymes are aminoacylase (about 5 t/a), ferred biocatalysts for processes with multiple
lactase (< 5 t/a), and penicillin G acylase (3 – enzyme reactions especially where cofactor re-
4 t/a) [261]. Other immobilized enzymes are generation is essential. However, ascertaining
also available but are used in even smaller which processes have been industrialized is dif-
amounts: glucoamylase, hydantoinase, inver- ficult; in many cases, whether immobilized cells
tase, nitrilase, RNAse, and penicillin V acylase or enzymes are used is unclear [263].
[260]. Recombinant Escherichia coli can be used
Many immobilized enzyme biocatalysts are for the production of relatively simple pep-
developed and manufactured by the company tide hormones with pharmaceutical applications
that uses them. Most economic data for such such as human insulin, human growth factor,
processes are confidential; published data are ex- or human interferons (→ Genetic Engineering,
Chap. 5.1., → Genetic Engineering, Chap. 5.2.).
42 Immobilized Biocatalysts
An immobilized biocatalyst is reportedly used ity and process economics for large-scale cell
for the conversion [264]. A hepatitis B surface cultures is described in [268].
antigen vaccine prepared in recombinant yeast
was approved for marketing, and many other Immobilized Plant Cells. Plant cell cul-
products from genetically engineered microor- tures can also be used for the production of
ganisms and yeast are in the clinical test stage metabolites such as pharmaceuticals, chemi-
(interferons, lymphokines, hormones, enzymes, cals, flavors, and fragrances. The first prod-
and vaccines) [265]. uct obtained from mass plant cell cultures was
shikonin [517-89-5], a red pigment composed
Immobilized Mammalian Cells. Many of eight naphthoquinone molecules. Shikonin is
complex proteins of pharmacological interest produced by a two-stage fermentation process
have precise folding and glycosylation require- and is a high-value chemical ($ 4000/kg) with
ments that cannot be fulfilled in simple microor- a limited annual market capacity of ca. 15 kg
ganisms. By using hybridoma or recombinant [270].
DNA techniques, mammalian cells can often be Immobilized plant cell systems will be used
persuaded to produce and secrete useful quanti- mainly for products of cells in the station-
ties of these compounds [265]. ary growth phase. The release of intracellularly
Tissue plasminogen activator (TPA) is one stored products by intermittent permeabilization
of the latest products of this technology. It dis- of immobilized cells can be a great economic
solves blood clots that cause heart attacks and advantage, allowing reutilization of the biomass
can be produced in recombinant E. coli or yeast [271]. The continuous immobilized plant cell
or in glycosylated form in mammalian cells. The process in combination with strain selection and
first three months’ sales in the United States in improved product leakage allows production of
1987 were almost $ 60×106 . Worldwide sales plant-derived chemicals in the range of $ 20 –
are projected to reach about $ 500×106 in the 25/kg [272]. However, in the present industrial
next few years [266]. state of technology for plant cell cultures, a
Monoclonal antibodies represent another relatively small number of products have both
major group of genetically engineered mam- high value per weight and sufficient market size
malian cell products. More than 100 mono- [273].
clonal antibodies are offered, with annual sales
of nearly $ 500×106 [266].
Generally, the economics of microbial sys- 9. Safety, Environmental, and Legal
tems are more favorable than those of mam-
malian cell culture systems for recombinant pro-
Aspects
teins. When the use of recombinant microbial
systems is impossible or impractical, cell cul- 9.1. Immobilized Enzymes
ture systems will gain a market position if the
products are high-value biologicals (> $ 100/g) Although enzymes are found throughout nature,
[265]. they cannot always be regarded as harmless. En-
Whether immobilized cell processes will play zymes do not show systemic toxicity in animal
an important role in classical fermentation is still tests and have not been reported to be muta-
an open question. Large-scale cell culture tech- genic, carcinogenic, or teratogenic [274]. How-
nologies are discussed in [267], [268]. Immobi- ever, an enzyme can cause allergic reactions in
lized cell cultures using hollow fibers, ceramic humans; proteolytic enzymes such as papain can
matrices, or microcarriers and microencapsu- be deleterious to skin or mucous membranes.
lated cell systems for industrial processes are be- Futhermore, commercial enzyme preparations
ing developed. β-Interferon was produced from contain foreign materials (e.g., additives, stabi-
human diploid fibroblasts grown on microcarri- lizers). In the safety evaluation of commercial
ers at a scale > 1000 L. Anchorage-independent enzyme preparations three major areas should
hybridoma cells can be immobilized in perfusion be considered:
systems [269]. Calculation of reactor productiv- 1) catalytic activity of the pure enzyme,
Immobilized Biocatalysts 43
2) allergenic reactions produced by the enzyme mitted to the FDA. Lists of enzymes permitted in
itself or by other proteins in the preparation, food manufacture in the United States, Canada,
and and the United Kingdom can be found in [276].
3) presence of toxic metabolites such as myco- The carrier and immobilization techniques
toxins or antibiotics. used for immobilized enzymes must also be
studied in safety evaluations [275]. The U.S.
The preparation and use of industrial en- regulations require that the fixing agents con-
zymes are subject to a variety of national laws sist of GRAS substances or that the immobi-
and regulations. Generally, regulations concern- lized enzyme preparation be washed thoroughly
ing biocatalysts are determined primarly by the to remove fixing-agent residues. The only two
end use of the enzyme, such as food-processing, fixing agents approved for the preparation of
agricultural, or medical applications. immobilized glucose isomerase in high-fructose
Safety aspects of enzyme preparations in corn syrup production are diethylaminoethyl
food processing were considered by the Joint cellulose (DEAE cellulose) and glutaraldehyde
FAO/WHO Expert Committee on Food Addi- [280]. Examples of pending or unsuccessful pe-
tives (JECFA) in several sessions. The commit- titions for GRAS affirmation are given in [280],
tee laid down criteria for evaluating enzymes [281]. Cross-linking agents are highly reactive
according to the source material [275]. En- and often toxic. Many of them lack sufficient
zymes obtained from animal, plant, or micro- toxicity data and are not approved for use in
bial sources commonly used as food, or used in food. On the other hand, the use of nontoxic
the preparation of food are accepted as foods if natural matrix materials such as κ-carrageenan,
satisfactory chemical and microbiological spec- alginate, or monomer-free prepolymers for en-
ifications can be established. No toxicological zyme entrapment seems to be a feasible way of
studies are required for such enzymes. For en- producing safe bioctalyst preparations for food
zymes obtained from nonpathogenic microor- processing.
ganisms commonly found as contaminants in The safety program required to obtain af-
food, JECFA recommended short-term toxic- firmation of GRAS status for immobilized
ity experiments. More extensive toxicological Lipozyme through scientific procedures is pre-
studies, as well as chemical and microbiologi- sented in [281]. Lipozyme is a lipase from Mu-
cal specifications, are required for enzymes ob- cor miehei that is immobilized on a macroporous
tained from less well-known microorganisms anion-exchange resin; it is used for modification
[276]. of edible fats and oils. The information required
In the United States and Canada, enzymes to obtain official approval of immobilized en-
from microbial sources can be marketed only if zymes should be presented by enzyme producers
they are produced according to the guidelines for to the authorities as discussed in [282].
“good manufacturing practices” (GMP) [277] As long as the materials used for immobiliz-
and the U.S. Food Chemicals Codex (FCC) ing enzymes are toxicologically acceptable or
[278]. In many other countries, regulations are do not to leak measurably into the substrate or
based on these guidelines or on the JECFA rec- reactor effluent streams, the use of immobilized
ommendations. enzymes represents a very safe technique [274].
In the United States, enzymes used in food,
animal feed, medical devices, and diagnostic
systems, as well as enzyme-based drugs and bio- 9.2. Immobilized Cells
engineered food products, are regulated by the
Food and Drug Administration (FDA). Gener- A major area of risk assessment in biotechnol-
ally, any product that is considered food can be ogy is associated with the use of living cells,
marketed without undergoing premarket clear- especially if they are genetically engineered.
ance if it is generally recognized as safe (GRAS) Extensive public discussion has focused on the
[279]. Some enzymes are assigned to this cate- possible danger to humans and the environment
gory. Enzymes that are not GRAS are subject from the intended or accidental release of re-
to approval as food additives and must undergo combinant microorganisms (→ Genetic Engi-
comprehensive safety testing before being sub- neering, Chap. 6.).
44 Immobilized Biocatalysts
The safety of mammalian cell products used These guidelines conclude that the hazards
in medicine is discussed in [283], [284]. Bio- associated with recombinant DNA microorgan-
chemical purity and the hypothetical risk of isms can be assessed and managed in a way
transmitting biologically active agents to prod- similar to that used for other organisms. For
uct recipients should be considered. The risk of EFB Class 1 microorganisms (nonpathogens)
contamination by adventitious agents such as free of adventitious agents and having limited
fungi, bacteria, and viruses, and the putative risk capacity for survival without adverse conse-
of tumorigenicity can be eliminated by effec- quences in the environment, the use of good in-
tive process design, characterization studies, and dustrial largescale practice (GILSP) is allowed.
routine biochemical and microbiological testing For Classes 2 and 3 with higher risks, more strin-
[284]. gent rules must be obeyed [287], [289].
Substantial experimental evidence indicates With immobilized living cell biocatalysts,
that even if mammalian tumor cells are intro- safety problems exist only in the case of leakage
duced accidentally into a human, tumor forma- or escape. To prevent such escape, any waste gas
tion does not occur because the foreign cells should be sterilized by filtering and heat treat-
are rejected and destroyed. Only large inoc- ment if necessary. Contaminated effluents must
ula (> 106 cells) can initiate tumor formation be inactivated by chemical or physical means,
in immunosuppressed rodents [283]. Scientific and accidental spills must be collected. Biotech-
methods are available to assure that recombinant nological operations can be performed with the
products derived from continuous mammalian same level of safety as microbiological labora-
cell lines are safe for use as human therapeutics tory work if good microbiological techniques
[284]. (GMT) are applied [289].
Because the highly selected mammalian or The history of the development of govern-
plant cell lines used in fermentation cannot sur- ment regulation of recombinant DNA research
vive in the environment without optimized cul- and manufacturing processes in the United
ture media, accidental release of such cells does States is reviewed in [290], [291]. Efforts to co-
not present a serious problem. However, unin- ordinate scientific review procedures, approval
tended release can become a central problem policies, and rulings have resulted in a “coordi-
with genetically engineered microorganisms. nated framework” [292] that defines the opera-
In industrial processes employing microorgan- tion of a flexible case-by-case risk assessment
isms, nonpathogens are preferred. If the use of [293].
pathogens is unavoidable, they should be phys- The fundamental question in work with re-
ically and biologically contained to prevent or combinant DNA is whether genetically engi-
limit their release. Safety limits depend on the neered organisms pose environmental risks that
risk class of microorganism involved; a risk clas- require entirely new, specific regulations. Over-
sification list has been drawn up by the Safety in all consensus has been reached that the nature
Biotechnology Working Party of the European of the organism to be released is what matters,
Federation of Biotechnology (EFB) [285]. Con- not the method by which it was produced [294],
tainment is achieved by employing good labora- [295].
tory practice, appropriate equipment, and spe- In Europe, an EEC commission is planning
cial design. Immobilization should contribute to announce a series of biotechnology directives
positively to this. that define the levels of physical and biological
For recombinant DNA work, large-scale in- containment and accident control, as well as pro-
dustrial processes are arbitrarily considered to vide the basis for authorizing the planned release
involve culture volumes > 10 L and are sub- of genetically engineered organisms into the en-
ject to special regulations. Guidelines have been vironment [295]. The European Committee on
established by the U.S. National Institutes of Regulatory Aspects of Biotechnology (ECRAB)
Health (NIH) [286], [287] and by the Organi- has suggested a stepwise framework for risk
zation for Economic Cooperation and Develop- analysis in biotechnological processes from ini-
ment (OECD) [284], [288], [289]. tiation to commercialization [296].
Immobilized Biocatalysts 45
42. Y. Y. Linko. P. Linko in J. Tramper, H. C. van Methods in Enzymology, vol. 135, Academic
der Plas, P. Linko (eds.): Biocatalysis in Press, Orlando 1987, pp. 160 – 170.
Organic Synthesis, Elsevier, Amsterdam 60. M. Marek, O. Valentova, J. Kas, Biotechnol.
1985, pp. 159 – 178. Bioeng. 26 (1984) 1223 – 1226.
43. A. Rosevear, J. Chem. Technol. Biotechnol. 61. G. P. Royer in K. Mosbach (ed.): Methods in
Biotechnol. 34 B (1984) 127 – 150. Enzymology, vol. 135, Academic Press,
44. J. Tramper in W. H. Scouten (ed.): Solid Phase Orlando 1987, pp. 141 – 146.
Biochemistry, J. Wiley & Sons, New York 62. R. Axen, J. Porath, S. Ernbaeck, Nature
1983, pp. 393 – 477. (London) 214 (1967) 1302 – 1304.
45. J. F. Kennedy, J. M. S. Cabral in H.-J. Rehm, 63. J. Kohn, M. Wilchek in W. H. Scouten (ed.):
G. Reed (eds.): Biotechnology, A Solid Phase Biochemistry, J. Wiley & Sons,
Comprehensive Treatise in 8 Volumes, vol. 7 a, New York 1983, pp. 599 – 630.
VCH Verlagsgesellschaft, Weinheim 1987, 64. I. Ugi, Angew. Chem. Int. Ed. Engl. 1 (1962)
pp. 347 – 404. 8 – 21.
46. W. H. Scouten in K. Mosbach (ed.): Methods 65. J. E. Prenosil, T. Hediger, Biotechnol. Bioeng.
in Enzymology, vol. 135, Academic Press, 31 (1988) 913 – 921.
Orlando 1987, pp. 30 – 65. 66. I. Chibata, T. Tosa, T. Sato, J. Mol. Catal. 37
47. R. A. Messing (ed.): Immobilized Enzymes for (1986) 1 – 24.
Industrial Reactors, Academic Press, New 67. D. Jaworek in E. K. Pye, L. B. Wingard Jr.
York 1975. (eds.): Enzyme Engineering, vol. 2, Plenum
48. R. A. Messing in M. Moo-Young (ed.): Press, New York 1974, pp. 105 – 114.
Comprehensive Biotechnology, vol. 2, 68. D. Jaworek, H. Botsch, J. Maier in K.
Pergamon Press, Oxford 1985, pp. 191 – 201. Mosbach (ed.): Methods in Enzymology,
49. J. F. Kennedy, C. A. White in A. Wiseman vol. 44, Academic Press, New York 1976,
(ed.): Handbook of Enzyme Biotechnology, pp. 195 – 201.
2nd ed., Ellis Horwood, Chichester UK 1985, 69. L. D’Angiuro et al., Biotechnol. Bioeng. 22
pp. 147 – 207, 1985 a. (1980) 2251 – 2272.
50. J. F. Kennedy, C. A. White in A. Wiseman 70. P. Cremonesi, L. D’Angiuro, Biotechnol.
(ed.): Handbook of Enzyme Biotechnology, Bioeng. 25 (1983) 735 – 744.
2nd ed., Ellis Horwood, Chichester UK 1985, 71. L. D’Angiuro, C. de Lalla, P. Cremonesi,
pp. 380 – 420, 1985 b. Biotechnol. Bioeng. 27 (1985) 1548 – 1553.
51. A. N. Emery, J. S. Hough, J. M. Novais, T. P. 72. P. Bernfeld, J. Wan, Science (Washington,
Lyons, Chem. Eng. (London) 259 (1972) D.C.) 142 (1963) 678 – 679.
71 – 76. 73. K. Kawashima in K. Mosbach (ed.): Methods
52. J. M. Novais: Ph. D. Thesis, University of in Enzymology, vol. 135, Academic Press,
Birmingham 1971. Orlando 1987, pp. 146 – 153.
53. J. F. Kennedy, Chem. Soc. Rev. 8 (1979) 74. S. Fukui, A. Tanaka, Adv. Biochem. Eng.
221 – 257. Biotechnol. 29 (1984) 1 – 33.
54. J. F. Kennedy, J. M. S. Cabral in J. Woodward 75. S. Fukui, A. Tanaka, T. Iida in C. Laane, J.
(ed.): Immobilized Cells and Enzymes – A Tramper, M. D. Lilly (eds.): Biocatalysis in
Practical Approach, IRL Press, Oxford UK Organic Media, Elsevier, Amsterdam 1987,
1985, pp. 19 –37. pp. 21 – 41.
55. F. R. Bernath, K. Venkatasubramanian in A. L. 76. D. Dinelli, Process Biochem. 7 (1972) 9 – 12.
Demain, N. A. Solomon (eds.): Manual of 77. D. Dinelli et al. in E. K. Pye, H. H. Weetall
Industrial Microbiology and Biotechnology, (eds.): Enzyme Engineering, vol. 3, Plenum
American Society for Microbiology, Press, New York 1978, pp. 477 – 481.
Washington D.C. 1986, pp. 230 – 247. 78. T. M. S. Chang in K. Mosbach (ed.): Methods
56. J. Iqbal, M. Saleemuddin, Enzyme Microb. in Enzymology, vol. 44, Academic Press, New
Technol. 4 (1982) 175 – 178. York 1976, pp. 201 – 218.
57. K. Ikura et al., J. Appl. Biochem. 6 (1984) 79. H. E. Klein, D. W. Sundstrom, D. Shim in J.
222 – 231. Woodward (ed.): Immobilized Cells and
58. B. Solomon, R. Koppel, E. Katchalski-Katzir, Enzymes – A Practical Approach, IRL Press,
Biotechnol. Bioeng. 28 (1986) 1213 – 1221. Oxford UK 1985, pp. 49 – 54.
59. B. Solomon, Z. Hollander, R. Koppel, E. 80. Y. T. Tu, T. M. S. Chang, Enzyme Microb.
Katchalski-Katzir in K. Mosbach (ed.): Technol. 4 (1982) 327 – 331.
Immobilized Biocatalysts 47
81. P. L. Luisi et al. in [37] , p. 549. 100. C. Akin in G. E. Russell (ed.): Biotechnology
82. P. L. Luisi, L. J. Magid, CRC Crit. Rev. and Genetic Engineering Reviews, vol. 5,
Biochem. 20 (1986) no. 4, 409 – 474. Intercept, Wimborne UK 1987, pp. 319 – 367.
83. R. Wichmann, C. Wandrey, A. F. Buchmann, 101. P. Brodelius in M. Moo-Young (ed.):
M. R. Kula, Biotechnol. Bioeng. 23 (1981) Bioreactor Immobilized Enzymes and Cells,
2789 – 2802. Fundamentals and Applications, Elsevier
84. V. G. Gekas, Enzyme Microb. Technol. 8 Applied Science, London 1988, pp. 167 – 196.
(1986) 450 – 460. 102. J. Klein in M. Moo-Young (ed.): Bioreactor
85. T. Takeuchi, K. Makino, Biotechnol. Bioeng. Immobilized Enzymes and Cells,
29 (1987) 160 – 164. Fundamentals and Applications, Elsevier
86. M. Fujimura, T. Mori, T. Tosa, Biotechnol. Applied Science, London 1988, pp. 1 – 8.
Bioeng. 29 (1987) 747 – 752. 103. J. Klein, F. Wagner in I. Chibata, L. B. Wingard
87. I. Chibata, L. B. Wingard Jr. (eds.): Jr. (eds.): Immobilized Microbial Cells,
Immobilized Microbial Cells, Applied Academic Press, New York 1983, pp. 11 – 51.
Biochemistry and Bioengineering, vol. 4, 104. J. F. Kennedy, J. M. S. Cabral in I. Chibata,
Academic Press, New York 1983. L. B. Wingard Jr. (eds.): Immobilized
88. B. Mattiasson (ed.): Immobilized Cells and Microbial Cells, Academic Press, New York
Organelles, vol. 1, 2, CRC Press, Boca Raton, 1983, pp. 189 – 280.
Florida 1983. 105. B. Mattiasson in B. Mattiasson (ed.):
89. C. Webb, G. M. Black, B. Atkinson (eds.): Immobilized Cells and Organelles, vol. 2,
Process Engineering Aspects of Immobilized CRC Press, Boca Raton, Florida 1983,
Cell Systems, The Institution of Chemical pp. 23 – 40.
Engineers, Rugby UK 1986. 106. F. B. Kolot, Process Biochem. 16 (1981) no. 5,
90. J. Tampion, M. D. Tampion: Immobilized 2 – 9.
Cells: Principles and Applications, 107. F. B. Kolot, Process Biochem. 16 (1981) no. 6,
Cambridge University Press, Cambridge 1987. 30 – 33.
91. M. Moo-Young (ed.): Bioreactor Immobilized 108. M. E. Hegglin: Ph. D. Thesis no. 8628, Swiss
Enzymes and Cells, Fundamentals and Federal Institute of Technology, Zürich 1988.
Applications, Elsevier Applied Science, 109. H. Tanaka, S. Uzman, I. J. Dunn, Biotechnol.
London 1988. Bioeng. 23 (1981) 1683 – 1702.
92. S. Birnbaum, P.-O. Larsson, K. Mosbach in 110. I. J. Dunn in K. Mosbach (ed.): Immobilized
W. H. Scouten (ed.): Solid Phase Biochemistry, Enzymes and Cells, Part B, Methods in
J. Wiley & Sons, New York 1983, Enzymology, vol. 135, Academic Press,
pp. 679 – 762. Orlando 1987, pp. 300 – 307.
93. J. Klein, K. D. Vorlop in M. Moo-Young (ed.): 111. A. L. van Wezel, Nature (London) 216 (1967)
Comprehensive Biotechnology, vol. 2, 64 – 65.
Pergamon Press, Oxford 1985, pp. 203 – 224. 112. J. M. Clark, M. D. Hirtenstein, J. Interferon
94. J. Tramper, Trends Biotechnol. 3 (1985) Res. 1 (1981) 391 – 400.
45 – 50. 113. K. Nilsson, Trends Biotechnol. 5 (1987)
95. S. F. Karel, S. B. Libicki, C. R. Robertson, 73 – 78.
Chem. Eng. Sci. 40 (1985) 1321 – 1354. 114. K. Nilsson, F. Buzaky, K. Mosbach,
96. I. Chibata, T. Tosa, T. Sato in A. L. Demain, Bio/Technology 4 (1986) 989 – 990.
N. A. Solomon (eds.): Manual of Industrial 115. B. K. Lydersen et al. in J. Feder, W. R. Tolbert
Microbiology and Biotechnology, American (eds.): Large-Scale Mammalian Cell Culture,
Society for Microbiology, Washington D.C. Academic Press, Orlando 1985, pp. 39 – 58.
1986, pp. 217 – 229. 116. G. G. Pugh, G. J. Berg, C. H. J. Sear in G. W.
97. P. K. Robinson, A. L. Mak, M. D. Trevan, Moody, P. B. Baker (eds.): Bioreactors and
Process Biochem. 21 (1986) 122 – 127. Biotransformations, Elsevier Applied
98. P. Brodelius, E. J. Vandamme in H.-J. Rehm, Science, London 1987, pp. 121 – 131.
G. Reed (eds.): Biotechnology, A 117. V. Jirku, J. Turková in K. Mosbach (ed.):
Comprehensive Treatise in 8 Volumes, vol. 7 a, Immobilized Enzymes and Cells, Part B,
VCH Verlagsgesellschaft, Weinheim 1987, Methods in Enzymology, vol. 135, Academic
pp. 405 – 464. Press, Orlando 1987, pp. 341 – 357.
99. C. D. Scott, Enzyme Microb. Technol. 9 (1987) 118. K. Mosbach, R. Mosbach, Acta Chem. Scand.
66 – 73. 20 (1966) 2807 – 2810.
48 Immobilized Biocatalysts
119. I. Chibata, T. Tosa, T. Sato, Appl. Microbiol. Methods in Enzymology, vol. 135, Academic
27 (1974) 878 – 885. Press, Orlando 1987, pp. 454 – 472.
120. G. K. Skryabin, K. A. Koshcheenko in K. 137. A. Tanaka, S. Fukui in B. Mattiasson (ed.):
Mosbach (ed.): Immobilized Enzymes and Immobilized Cells and Organelles, vol. 1,
Cells, Part B, Methods in Enzymology, CRC Press, Boca Raton, Florida 1983,
vol. 135, Academic Press, Orlando 1987, pp. 101 – 119.
pp. 198 – 222. 138. H. Ochiai, A. Tanaka, S. Fukui in I. Chibata,
121. M. Kumakura, M. Yoshida, I. Kaetsu, L. B. Wingard Jr. (eds.): Immobilized
Biotechnol. Bioeng. 21 (1979) 679 – 688. Microbial Cells, Academic Press, New York
122. C. Bucke in K. Mosbach (ed.): Immobilized 1983, pp. 153 – 187.
Enzymes and Cells, Part B, Methods in 139. P. E. Gispy, K. K. Rao, D. O. Hall in K.
Enzymology, vol. 135, Academic Press, Mosbach (ed.): Immobilized Enzymes and
Orlando 1987, pp. 175 – 189. Cells, Part B, Methods in Enzymology,
123. I. Chibata, T. Tosa, T. Sato, I. Takata in K. vol. 135, Academic Press, Orlando 1987,
Mosbach (ed.): Immobilized Enzymes and pp. 440 – 454.
Cells, Part B, Methods in Enzymology, 140. B. Arkles, W. S. Brinigar, J. Biol. Chem. 250
vol. 135, Academic Press, Orlando 1987, (1975) 8856 – 8862.
pp. 189 – 198. 141. S. S. Godbole, R. Kaul, S. F. D’Souza, G. B.
124. T. Omata, A. Tanaka, T. Yamane, S. Fukui, Nadkarni, Biotechnol. Bioeng. 25 (1983)
Eur. J. Appl. Microbiol. Biotechnol. 6 (1979) 217 – 224.
207 – 215. 142. A. Tanaka, N. Hagi, G. Gellf, S. Fukui, Agric.
125. S. Fukui, K. Sonomoto, A. Tanaka in K. Biol. Chem. 44 (1980) 2399 – 2405.
Mosbach (ed.): Immobilized Enzymes and 143. W. Hartmeier, Trends Biotechnol. 3 (1985)
Cells, Part B, Methods in Enzymology, 149 – 153.
vol. 135, Academic Press, Orlando 1987, 144. B. Hahn-Hägerdal, Biotechnol. Bioeng. 26
pp. 230 – 252. (1984) 771 – 774.
126. M. L. Shuler in K. Mosbach (ed.): Immobilized 145. M. F. Cocquempot et al., Biotechnol Lett. 4
Enzymes and Cells, Part B, Methods in (1982) 313 – 318.
Enzymology, vol. 135, Academic Press, 146. B. Hahn-Hägerdal in B. Mattiasson (ed.):
Orlando 1987, pp. 372 – 387. Immobilized Cells and Organelles, vol. 2,
127. C. Heath, G. Belfort, Adv. Biochem. Eng. CRC Press, Boca Raton, Florida 1983,
Biotechnol. 34 (1987) 1 – 31. pp. 79 – 94.
128. K. Ku et al, Biotechnol. Bioeng. 23 (1981) 147. M. Charles, J. Phillips in M. Moo-Young (ed.):
79 – 95. Comprehensive Biotechnology, vol. 2,
129. J. Van Brunt, Bio/Technology 4 (1986) 45 – 50. Pergamon Press, Oxford 1985, pp. 225 – 229.
130. J. Feder in K. Mosbach (ed.): Immobilized 148. K. Buchholz (ed.): “Characterization of
Enzymes and Cells, Part B, Methods in Immobilized Biocatalysts,” DECHEMA
Enzymology, vol. 135, Academic Press, Monogr. 84 (1979) 198 – 200.
Orlando 1987, pp. 393 – 399. 149. J. E. Prenosil, I. J. Dunn, E. Heinzle in H.-J.
131. G. Klement, W. Schreier, H. W. D. Katinger in
Rehm, G. Reed (eds.): Biotechnology, vol. 7 a,
M. Moo-Young (ed.): Bioreactor Immobilized
Chap. 10, VCH Verlagsgesellschaft, Weinheim
Enzymes and Cells, Fundamentals and
1987.
Applications, Elsevier Applied Science,
150. A. Bar-Eli, E. Katschalski, Nature (London)
London 1988, pp. 53 – 58.
188 (1960) 856.
132. R. R. Mohan, N. N. Li, Biotechnol. Bioeng. 17
151. P. Monsan, D. Combes in K. Mosbach (ed.):
(1975) 1137 – 1156.
Methods in Enzymology, vol. 137, Academic
133. F. Lim, A. Sun, Science (Washington, D.C.)
Press, New York 1988, pp. 584.
210 (1980) 908 – 910.
152. K. Martinek, V. P. Torchilin in K. Mosbach
134. A. Jarvis, T. Gridina, Bio Techniques 1 (1983)
(ed.): Methods in Enzymology, vol. 137,
22 –27.
Academic Press, New York 1988, pp. 615.
135. R. G. Rupp in J. Feder, W. K. Tolbert (eds.):
153. L. Goldstein, Y. Levin, E. Katchalski,
Large-Scale Mammalian Cell Culture,
Biochemistry 3 (1964) 1913 – 1919.
Academic Press, Orlando 1985, pp. 19 – 36.
136. J. N. Barbotin et al. in K. Mosbach (ed.): 154. D. I. C. Wang et al., Fermentation and Enzyme
Immobilized Enzymes and Cells, Part B, Technology, J. Wiley & Sons, New York 1979.
Immobilized Biocatalysts 49
155. Hor-Da Juang, Hung-Shan Weng, Biotech. 176. J. Haggin, Chem. Eng. News 66 (1988) no. 23,
Bioeng. 26 (1984) 623 – 626. 7.
156. K. Mosbach (ed.): Methods in Enzymology, 177. E. C. Steiner, G. E. Blau, G. L. Agin: Simusolv,
vol. 44, Acad. Press, New York 1976. Modeling and Simulation Software, Dow
157. S. Furusaki, J. Chem. Eng. Japan 21 (1988) , Chemical Company, Midland, Michigan 1986.
219 –230. 178. I. Chibata, T. Tosa, T. Sato in J. F. Kennedy
158. K. Buchholz: “Reaction engineering (ed.): Biotechnology, vol. 7 a, VCH
parameters for immobilized biocatalysts,” Verlagsgesellschaft, Weinheim 1987, p. 653.
Adv. Biochem. Eng. 24 (1982) 39. 179. R. A. Messing, Biotechnol. Bioeng. 24 (1982)
159. S. F. Karel, S. B. Libicki, C. R. Robertson, 1115.
Chem. Eng. Sci. 40 (1985) 1321 – 1354. 180. I. Nilsson, S. Ohlson, Eur. J. Appl. Microbiol.
160. J. M. Radovich, Enzyme Microb. Technol. 7 Biotechnol. 14 (1982) 86.
(1985) 2 – 10. 181. M. A. Gencer, R. Mutharasan, Biotechnol.
161. C. N. Satterfield, T. K. Sherwood: The Role of Bioeng. 25 (1983) 2243.
Diffusion in Catalysis, Addision Wesley, 182. K. Hashimoto, S. Adachi, H. Noujima, Y.
Reading, Mass. 1963. Ueda, Biotechnol. Bioeng. 25 (1983) 2371.
183. J. P. Guiraud, A. M. Bajon, P. Chautard, P.
162. C. N. Satterfield, Mass Transfer in
Galzy, Enzyme Microb. Technol. 5 (1983) 185.
Heterogeneous Catalysis, MIT Press,
184. P. Monsan, D. Combes, Biotechnol. Bioeng.
Cambridge, Mass. 1970.
26 (1984) 347.
163. P. H. Calderbank in V. W. Uhl, J. B. Gray 185. J. E. Prenosil, E. Stuker, T. Hediger, J. R.
(eds.): Mixing, vol. 2, Academic Press, London Bourne, Ann. N.Y. Acad. Sci. 434 (1984) 136.
1967. 186. F. Godia, C. Casas, C. Sola, Biotechnol.
164. T. K. Sherwood, R. L. Pigford, C. Wilke: Mass Bioeng. 30 (1987) 836.
Transfer, McGraw-Hill, New York 1975. 187. J. M. S. Cabral, J. M. Novais, J. P. Cardoso,
165. K. Buchholz: “Mass transfer influence on J. F. Kennedy, J. Chem. Tech. Biotechnol. 36
effectiveness,” DECHEMA-Monogr. 84 (1986) 247.
(1979) 212. 188. J. B. Griffiths in R. E. Spier, J. B. Griffiths
166. S. F. Karel, S. B. Libicki, C. R. Robertson, (eds.): Animal Cell Biotechnology, vol. 3,
Chem. Eng. Sci. 40 (1985) 1321 – 1354. Academic Press, London 1988, p. 179.
167. H. Tanaka, I. J. Dunn: “Kinetics of Biofilm 189. O.-W. Merten, Trends Biotechnol. 5 (1987)
Nitrification,” Biotechnol. Bioeng. 24 (1982) 230.
669. 190. C. G. J. Baker, A. Margaritis, M. A.
168. D. I. C. Wang et al.: Fermentation and Enzyme Bergougnou in M. Moo-Young (ed.):
Technology, J. Wiley & Sons, New York 1979. Advances in Biotechnology, vol. 1, Pergamon
169. J. E. Prenosil, I. J. Dunn, E. Heinzle in H.-J. Press, Toronto 1981, p. 635.
Rehm, R. Reed (eds.): Biotechnology, vol. 7 a, 191. P. F. Cooper, B. Atkinson: Biological Fluidized
VCH Verlagsgesellschaft, Weinheim 1987, Bed Treatment of Water and Wastewater, Ellis
p. 489. Horwood, Chichester UK 1981.
170. R. E. Spier, J. B. Griffiths: Animal Cell 192. M. Nagashima et al., Biotechnol. Bioeng. 26
Biotechnology, vol. 3, Academic Press, (1984) 992.
London 1988. 193. G. F. Andrews, Biotechnol. Prog. 2 (1986) 16.
171. C. L. Cooney, Science (Washington, D.C.) 219 194. G. H. Schoutens et al., J. Chem. Technol.
(1983) 728. Biotechnol. 36 (1986) 335.
172. I. Chibata in P. Linko, J. Larinkari (eds.): Food 195. S. J. Chen, C. T. Li, W. K. Shieh, J. Chem.
Process Engineering, vol. 2, Applied Science Technol. Biotechnol. 35 B (1985) 101.
Publ., London 1979, p. 1. 196. M. Denac, I. J. Dunn, Biotechnol. Bioeng. 32
173. K. Venkatasubramanian in P. Linko, J. (1988) 159.
197. G. F. Andrews, J. Przezdziecki, Biotechnol.
Larinkari (eds.): Food Process Engineering,
Bioeng. 28 (1986) 802.
vol. 2, Applied Science Publ., London 1979,
198. W. K. Shieh, J. D. Keenan, Adv. Biochem. Eng.
p. 162.
Biotechnol. 33 (1986) 131.
174. H. Samejima et al., Ann. N.Y. Acad. Sci. 434 199. P. W. Runstadler Jr., S. R. Cernek in R. E.
(1984) 394. Spier, J. B. Griffiths (eds.): Animal Cell
175. J. J. Heijnen et al., GWF Gas Wasserfach: Biotechnology, vol. 3, Academic Press,
Wasser/Abwasser 126 (1985) 81. London 1988, p. 305.
50 Immobilized Biocatalysts
245. P. Linko, Y.-Y. Linko, CRC Crit Rev. vol. 4, VCH Verlagsgesellschaft, Weinheim
Biotechnol. 1 (1984) 289 – 337. 1985, pp. 39 – 72.
246. Novo Technical Report, 175d-GB, Novo Allé, 265. A. S. Lubiniecki, Dev. Ind. Microbiol. 28
Bagsvaerd, Denmark 1985. (1987) 133 – 138.
247. I. Chibata, T. Tosa, T. Sato in A. I. Laskin 266. L. H. Smithson, CHEMTECH 18 (1988) no. 3,
(ed.): Enzymes and Immobilized Cells in 168 – 173.
Biotechnology, Benjamin/Cummings, Menlo 267. W. R. Arathoon, J. R. Birch, Science
Park, California 1985, pp. 37 – 70. (Washington, D.C.) 232 (1986) 1390 – 1395.
248. I. Chibata, T. Tosa, T. Sato in J. F. Kennedy 268. R. W. Swartz in M. Phillips, S. P. Shoemaker,
(ed.): Biotechnology, vol. 7 a, VCH R. D. Middlekauff, R. M. Ottenbrite (eds.):
Verlagsgesellschaft, Weinheim 1987, “The Impact of Chemistry on Biotechnology,”
pp. 653 – 684. ACS Symp. Ser. 362 (1988) 102 – 120.
249. I. Chibata, T. Tosa, T. Sato, J. Mol. Cat. 37 269. H. Katinger in O. M. Neijssel, R. R. van der
(1986) 1 – 24. Meer, K. Ch. A. M. Luyben (eds.): Proceedings
250. Y.-Y. Linko, P. Linko in J. Tramper, H. C. van 4th European Congress on Biotechnology,
der Plas, P. Linko (eds.): Biocatalysts in vol. 4, Elsevier, Amsterdam 1987, pp. 21 – 42.
Organic Synthesis, Elsevier, Amsterdam 270. V. Petiard, P. Steck in J. M. Cardoso et al.
1985, pp. 159 – 178. (eds.): Perspect. in Biotechnol. NATO ASI Ser.
251. M. S. Switzenbaum, Enzyme Microb. Technol. A 128 (1987) 139 – 153.
5 (1983) 242 – 250. 271. P. Brodelius in A. I. Laskin (ed.): Enzymes and
252. P. F. Cooper, B. Atkinson (eds.): Fluidized Bed Immobilized Cells in Biotechnology,
Treatment of Water and Wastewater, Ellis Benjamin/ Cummings, Menlo Park, California
Horwood, Chichester 1981. 1985, pp.109 – 147.
253. C. Webb, B. M. Black, B. Atkinson (eds.): 272. O. Sahai, M. Knuth, Biotechnol. Prog. 1
Process Engineering Aspects of Immobilized (1985) 1 – 9.
Cell Systems, Inst. Chem. Engineers, Rugby 273. V. Petiard et al. in O. M. Neijssel, R. R. van der
1986. Meer, K. Ch. A. M. Luyben (eds.):
254. C. Webb in C. F. Forster, D. A. J. Wase (eds.): Proceedings 4th European Congress on
Environmental Biotechnology, Ellis Horwood, Biotechnology, 1987, vol. 4, Elsevier,
Chichester 1987, pp. 347 – 376. Amsterdam 1987, pp. 157 – 170.
255. D. Barnes, P. A. Fitzgerald in C. F. Forster, 274. P. F. Noordervliet, D. A. Toet in H.-J. Rehm,
D. A. J. Wase (eds.): Environmental G. Reed (eds.) Biotechnology, A
Biotechnology, Ellis Horwood, Chichester Comprehensive Treatise in 8 Volumes, vol. 7 a,
1987, pp. 57 – 113. VCH Verlagsgesellschaft, Weinheim 1987,
256. R. E. Shamel, J. J. Chow, Chem. Eng. Prog. 83 pp. 711 – 741.
(1987) no. 10, 41 – 44. 275. “Evaluation of Certain Food Additives,” 21st
257. C. G. Rosén, CHEMTECH 17 (1987) no. 10, Report of the Joint FAO/WHO Expert
612 – 618. Committee on Food Additives (JECFA),
258. W. E. Goldstein, Biotechnol. Bioeng. Symp. 17 W.H.O. Tech. Rep. Ser. 617 (1978).
(1986) 763 – 776. 276. W. H. B. Denner in T. Godfrey, J. R. Reichelt
259. P. L. Layman, Chem. Eng. News 64 (1986) (eds.): Industrial Enzymology, The Nature
no. 37, Sept. 15, 11 – 14. Press, New York 1983, pp. 111 – 137.
260. P. B. Poulsen in [38] , pp. 413 – 419. 277. Code of Federal Regulations, 21 CFR Sec.
261. P. B. Poulsen: Third European Congress on 110.1, Office of Federal Register, National
Biotechnology, München 1984, Proceedings, Archives and Records Service, Washington
vol. 4, VCH Verlagsgesellschaft, Weinheim D.C. 1981.
1985, pp. 339 – 344. 278. Food Chemicals Codex, 3rd ed., National
262. M. M. Domach, T. J. Gricus, Biotechnol. Academy of Science/National Research
Bioeng. 27 (1985) 1458 – 1465. Council, Food and Nutrition Board,
263. P. Linko: Third European Congress on Committee on Codex Specifications, National
Biotechnology, München 1984, Proceedings, Academy Press, Washington D.C. 1981.
vol. 4, VCH Verlagsgesellschaft, Weinheim 279. J. S. Kahan, J. N. Gibbs, Appl. Biochem.
1985, pp. 357 – 363. Biotechnol. 11 (1985) 489 – 506.
264. K. Kieslich: Third European Congress on 280. H. H. Weetall, W. H. Pitcher Jr., Science
Biotechnology, München 1984, Proceedings, (Washington D.C.) 232 (1986) 1396 – 1403.
52 Immobilized Biocatalysts
281. J. R. Fordham, N. H. Block, Dev. Ind. 289. S. P. Vranch in O. M. Neijssel, R. R. van der
Microbiol. 28 (1987) 25 – 31. Meer, K. Ch. A. M. Luyben (eds.):
282. J. R. Reichelt in T. Godfrey, J. R. Reichelt Proceedings 4th European Congress on
(eds.): Industrial Enzymology, The Nature Biotechnology, vol. 4, Elsevier, Amsterdam
Press, New York 1983, pp. 138 – 156. 1987, pp. 355 – 363.
283. J. C. Petricciani, Swiss Biotech. 5(2a) (1987) 290. R. L. Greenstein, Appl. Biochem. Biotechnol.
19 – 20. 11 (1985) 489 – 506.
284. A. S. Lubiniecki, Dev. Ind. Microbiol. 28 291. T. O. McGarity, K. O. Bayer in M. Moo-Young
(1987) 133 – 138. (ed.): Comprehensive Biotechnology, vol. 4,
285. “Safe Biotechnology – General Pergamon Press, Oxford 1985, pp. 587 – 608.
Considerations,” report prepared by the Safety 292. “Coordinated Framework for the Regulation of
in Biotechnology Working Party of the Biotechnology,” U.S. Fed. Reg. 51 (June 26,
European Federation of Biotechnology, Third 1986) 23 302 – 23 393.
European Congress on Biotechnology, vol. 3, 293. D. T. Kingsbury, Trends Biotechnol. 6 (1988)
VCH Verlagsgesellschaft, Weinheim 1984, 39 – 42.
pp. 647 – 662. 294. Introduction of Genetically Engineered
286. “Guidelines for Research Involving Organisms into the Environment: Key Issues,
Recombinant DNA Molecules,” National US National Academy of Sciences,
Institute of Health (NIH) revised version, US Washington D.C. 1987.
Fed. Reg. 49 (Nov. 23, 1984) 46 266. 295. B. P. Ager, Trends Biotechnol. 6 (1988)
287. C. Sekhar, Chem. Eng. (N.Y.) 63 (1985) 42 – 44.
no. 17, 57 – 59. 296. European Committee on Regulatory Aspects
288. Recombinant DNA Safety Considerations, of Biotechnology (ECRAB), Swiss Biotech.
OECD, Paris 1986. 4(5) (1986) 15 – 20.
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
the “fit” between the immunoglobulin binding sites; it represents approximately 80 % of all im-
site and the epitope. 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.
Figure 2. Schematic representation of an immunoglobulin
G (IgG) molecule
Upon binding with invading bacteria, IgG ac-
tivates the complement system (a group of inter-
Antibodies produced by a single clone of acting serum proteins) which attracts phagocytic
B cells are termed monoclonal antibodies and cells. The binding of complement components
recognize only a single epitope (see Chap. 6). to the Fc region of IgG permits the uptake and
Polyclonal antibodies are produced by several killing of the bacteria by phagocytes.
B cell clones which recognize the same antigen Immunoglobulin M (IgM) is a 900 000 dal-
but bind to different epitopes. ton pentamer whose five IgG-like units are held
Immunoglobulins are divided into five together by intramolecular disulfide bonds and
classes termed IgG, IgM, IgA, IgD, and IgE stabilized by a polypeptide “anchor” termed the
based upon physical and structural differences; “J-chain.” IgM comprises 5 – 10 % of normal cir-
they are all glycoproteins. The vast majority of culating immunoglobulins (1 – 2 mg/mL). Due
immunoglobulins circulate in the plasma frac- to its large size, IgM is confined to the in-
tion, but certain cells of the immune system can travascular space. It binds complement but, un-
express immunoglobulin on their surface. like IgG, does not bind directly to phagocytic
Immunoglobulin G (IgG) has a molecular cells. Immunoglobulin M is usually the first an-
mass of 150 000 and has two antigen binding tibody class formed in response to foreign anti-
Immunotherapy and Vaccines 5
gen exposure. Due to its multivalency, IgM is ex- asitic infections, most notably schistosomiasis
tremely efficient at agglutinating bacteria. This (see Chap. 4) where high levels (> 100 µg/mL)
phenomenon is important in the control of bac- of serum IgE have been noted. Immunoglobu-
teremia and the neutralization of LPS released lin E induces an inflammatory response and “re-
by gram-negative bacteria. cruits” effector cells to the area. In contrast, IgE-
Immunoglobulin A (IgA) occurs as 160 000 mediated degranulation of effector cells leads to
dalton monomers or 320 000 dalton dimers. the release of vasoreactive molecules. Individu-
Monomers are found primarily in the intravascu- als suffering from certain allergies can also have
lar space and comprise approximately 10 – 15 % elevated serum IgE levels.
(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 1.2.3. Immune Response
by noncovalent interaction with the “secretory
piece,” a glycoprotein of about 60 000 daltons. The chain of events leading to an immune
Secretory IgA is the predominant immunoglob- response is extremely complicated and not
ulin found in mucous secretions, such as tears, yet completely understood. A simple model
colostrum, pulmonary, intestinal, and urinogen- is shown in Figure 3. The human is capable
ital fluids; IgA can bind complement and react of forming an immune response to thousands
with phagocytic cells. of foreign antigens. The humoral (antibody-
Although parenteral immunization stimu- mediated) response depends on the B cells
lates a vigorous serum IgA response, secretory which are lymphocytes derived from bone mar-
IgA is considered to be of greater importance row stem cells. Upon maturation, B cells form
due to its “first line” defensive role in body se- antibody-secreting plasma cells.
cretions. Secretory IgA plays a critical role in Each clone of B cells has an immunoglob-
providing protection against respiratory, intesti- ulin molecule with a recognition site for a spe-
nal, and urinogenital tract bacterial pathogens. cific antigen on its surface. Binding of the appro-
Immunoglobulin D (IgD) is a monomer of ca. priate antigen to the B cell causes proliferation
180 000 daltons; concentrations vary widely in of the clone whereby the progeny cells secrete
humans ranging from 0 to ca. 0.4 mg/mL. Com- antibody whose specificity is identical to that
pared to other immunoglobulin classes, IgD is of the cell-surface immunoglobulin. A foreign
very susceptible to proteolysis and possesses a antigen can be taken up by macrophages, pro-
short half-life (about 3 days). Since IgD does not cessed, and presented upon the cell surface, or
fix complement or bind to phagocytic cells, it is may remain in a soluble state. All antigens can be
not considered to be a “protective” immunoglob- termed T-dependent or T-independent depend-
ulin as are IgA, IgG, and IgM. However, IgD is ing on whether they require or do not require
abundant on the surface of B lymphocytes and the interaction of T cells for antibody synthe-
may play a central role in the activation pro- sis (T cells are lymphocytes derived from the
cess which leads to the development of antibody- thymus). In the case of a T-independent anti-
secreting plasma cells. gen (usually polymers, such as bacterial capsular
Immunoglobulin E (IgE) is a monomer of polysaccharides), the antigen can cross-link the
200 000 daltons; its average serum concentra- B cell surface antibody molecules of the B cell;
tion (ca. 250 ng/mL) is the lowest of any im- this initiates proliferation and antibody synthe-
munoglobulin. Although IgE does not fix com- sis. A T-dependent antigen requires, in addition
plement or react directly with phagocytic cells, to binding to B-cell surface immunoglobulin, the
it has a very high affinity for most cells and ba- release of T cell factors which act upon the B cell
sophils. Immunoglobulin E appears to mediate to initiate proliferation.
both a protective and a detrimental immune re- A human exposed to an antigen in this way
sponse. At the mucoid surface, pathogens bind- is considered primed. This is a critical factor
ing to IgE stimulate an acute inflammatory re- in immunization. After initial exposure to a T-
sponse by triggering the release of potent medi- dependent antigen by immunization, the induced
ators from mast cells and basophils. A protective serum antibody concentration is low and re-
role for IgE is indicated in several chronic par- 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, population of T cells synthesizes interleukin 2
whereas IgM is ca. 5 days). Upon revaccination (IL-2), also known as T-cell growth factor,
(boosting) or natural exposure to the pathogen which causes a second subpopulation of acti-
or toxin, a vigorous antibody response occurs vated T cells expressing the IL-2 receptor to pro-
in which high antibody levels are synthesized liferate and become cytotoxic T cells. Cytotoxic
over a longer period of time. This is termed an T cells recognize the specific antigen they are
anamnestic response. Therefore, high levels of activated against when expressed on the surface
serum antibodies do not have to be present for of infected cells; they thus kill these infected
the individual to be protected by prior vaccina- cells by lysis. A third subpopulation of T cells
tion. A T-independent antigen does not evoke an evolve to be “primed” memory cells which un-
anamnestic response. dergo rapid proliferation upon reexposure to the
The induction of a cell-mediated immune same antigen.
(CMI) response is rather more complicated and Immunization with a given vaccine may in-
less well defined. The CMI response plays a crit- duce a CMI, a humoral antibody response, or
ical role in immunity to pathogens able to live both depending on the infecting pathogen. Both
and proliferate within host cells. The first step types of immunity are usually desirable and
is “activation” of T cells. Like B cells, T cells may act synergistically. For example, vaccine-
also recognize specific antigens. Recognition of induced immunity to many viral diseases, such
an antigen on the surface of a macrophage by as rabies, measles, mumps, and rubella, is con-
T cells results in the release of interleukin 1 (IL- firmed by measuring serum antibody levels.
1), a lymphokine which activates them. One sub- However, infected cells can only be destroyed
Immunotherapy and Vaccines 7
by cytotoxic T cells. Circulating antibody prob- (e.g., by passaging the virus in cell culture or
ably prevents the spread of the virus, while cy- deletion of bacterial genes), but is still able to
totoxic T cells eliminate already infected cells. multiply to a limited degree, thereby eliciting
Tuberculosis vaccines stimulate a good antibody a protective immune response in the absence
response but are less proctective due to the sup- of disease symptoms.
pression of the critical CMI response.
One major problem associated with highly
purified antigens or subunit vaccines is re-
1.2.4. Active Immunization duced immunogenicity. Unfortunately, as pro-
tective antigens are purified from either bacte-
Active immunization entails the administration ria, viruses, or parasites in order to free them
of one or more antigens to a host in the form from nontoxic substances such as LPS, they lose
of a vaccine in an attempt to elicit a protec- their ability to evoke an immune response. Their
tive immune response. In this way an individ- immunogenicity can, however, be improved by
ual can be rendered immune to a variety of dis- using an adjuvant. Adjuvants function either to
eases. Most vaccines are administered to infants present the antigen to the immune system for a
or young children. The trend is to combine sev- prolonged period of time, or to nonspecifically
eral monovalent vaccines (vaccines containing stimulate the immune system by releasing im-
a single antigen or vaccine strain) to form mul- mune modulators such as lymphokines. To date,
tivalent vaccines capable of simultaneously in- only aluminum gels have been licensed as adju-
ducing immunity to several diseases. Infants are vants for human use. Tetanus or diphtheria tox-
routinely immunized against diphtheria, tetanus, oid is absorbed onto the gel which allows the
pertussis, and, in some countries, polio by ad- toxoid to persist longer at the injection site. Bio-
ministering the vaccines in a single injection. logically active peptides that can nonspecifically
Similarly, young children are immunized simul- stimulate the immune system via lymphokine
taneously with a multivalent measles, mumps, release or downregulate the suppression of the
and rubella vaccine. Administration of multi- immune system, are being evaluated.
valent vaccines reduces the number of visits to
healthcare centers, a critical point in developing
countries. 1.2.5. Passive Immunization
Historically, parenteral vaccines have usu-
ally been used (i.e., injected with a syringe and Passive immunization entails the transfer of pre-
needle). This approach is extremely success- formed immunoglobulins to a host. Passive im-
ful for systemic diseases such as diphtheria or munization is usually employed after known or
measles. However, in localized diseases (e.g., suspected exposure to a given pathogen (for ex-
intestinal infections such as cholera), parenteral ample, after being bitten by a rabid animal). It
vaccines are of limited use since the immune is used in cases were disease progress is rapid
system needs to be stimulated at the site of in- if the pathogen or toxin is not neutralized. The
fection (e.g., local immune system). Therefore, time needed for the host to mount a protective
effort is now being directed at controlling intesti- immune response following active immuniza-
nal infectious diseases with orally administered tion (7 – 14 days) would be too long. In some
vaccines. Oral vaccines can consist of instances, globulin may be passively adminis-
tered as a prophylactic measure, as in the case
1) killed intact bacteria, of travellers entering an area where hepatitis A
2) toxoids (i.e., toxins that are still immuno- is endemic.
genic but are rendered biologically inactive Immunoglobulin preparations are assayed to
by treatment with a chemical, heat, or muta- confirm that they contain a high titer of neu-
tion) tralizing antibodies against the disease in ques-
3) subunit vaccines, in which only the nontoxic tion. Such hyperimmune globulins must contain
portion of a molecule is used, or at least five-fold higher levels than normal glob-
4) live-attenuated vaccines, in which a viral or ulin and are obtained by screening plasma units
bacterial strain is rendered nonpathogenic (human) for a given antibody. Alternatively, vol-
8 Immunotherapy and Vaccines
unteers may be vaccinated to enrich their plasma gen vaccines due to increased safety and lower
for a given antibody. costs.
Immunoglobulin is administered either intra- Numerous licensed, live-attenuated vaccines
muscularly or intravenously. Only a limited vol- (measles, polio, mumps, rubella, varicella, and
ume of globulin can be given intramuscularly. typhoid fever) have been obtained by rather em-
Intravenous administration allows for compara- pirical techniques, such as passage on tissue cul-
tively large quantities (400 mg of immunoglob- ture or chemical mutagenesis. Surprisingly, the
ulin per kilogram body weight) of immunoglob- precise genetic alterations responsible for aviru-
ulin to be administered quickly. lence are unknown. Several live-attenuated bac-
Several problems are associated with the terial vaccine strains have been developed by
preparation of immunoglobulin for passive ther- using genetic engineering. The most advanced,
apy. Firstly, the identification of plasma donors in regards to clinical testing, are for cholera
who are “hyperimmune” to a given antigen and typhoid fever. These strains have been at-
requires expensive and tedious screening pro- tenuated by inactivating a gene or genes es-
grams. Secondly, the administration of im- sential for virulence. A related approach has
munoglobulins entails the possibility of trans- been to clone a given protective antigen and ex-
mitting viral diseases, such as hepatitis B and press it in a suitable, nonvirulent “carrier” strain.
non-A, non-B hepatitis. Additional screening For example, surface glycoprotein from HIV-
procedures are therefore required. 1 and rabies virus have been introduced into
An alternative to antibody obtained from vaccinia virus (smallpox vaccine). In addition,
donors is the use of human monoclonal anti- antigens from several enteric pathogens such
bodies (humabs, see Chap. 6). Humabs can be as Shigella sonnei, S. dysenteriae, and Vibrio
synthesized by hybridoma cell lines that are pro- cholerae have been introduced into the licensed
duced by fusing a human B cell secreting a de- live-oral typhoid vaccine strain, Ty21a. Such re-
sired antibody to a nonsecreting cell line. Hy- combinant strains may prove useful as “biva-
bridomas can be grown in fermentors containing lent” vaccines, conferring protection against two
up to 1000 L of serum-free medium. Antibody enteric pathogens.
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 2. Bacterial Vaccines
have been produced against a variety of infec-
tious agents (e.g., cytomegalovirus, diphtheria The chemotherapy of bacterial dis-
and tetanus toxins) and will be tested in the near ease is treated elsewhere (→ Antibiotics ;
future. → Chemotherapeutics, Chap. 2.).
History of Immunization. Studies by Roux booster doses, together with tetanus toxoid, are
and Yersin in 1888 demonstrated that diphthe- recommended every 7 – 10 years.
ria is a “toxicosis.” Two years later, Behring
and Kitasato established that the disease could Adverse Reactions. The absolute reacto-
be prevented in animals by immunization with genicity of diphtheria toxoid is difficult to deter-
a crude diphtheria toxoid. Attempts to extend mine since it is usually administered with tetanus
these findings to humans used a “toxoid” pre- toxoid or pertussis vaccines. Overall, the vaccine
pared by combining diphtheria toxin and anti- is considered to be safe; it primarily evokes mild,
toxin. The first large-scale immunization pro- transient, local reactions.
gram with such a product was performed on New
York school children in 1922. Soon after, crude Vaccine Efficacy. No large-scale field trials
toxoids prepared by formaldehyde treatment of have been performed to determine vaccine ef-
C. diphtheriae filtrates replaced toxin-antitoxin ficacy. This is because it would be unethical to
toxoids. Routine mass vaccination against diph- withhold the vaccine from control subjects in
theria was initiated in Western Europe and North view of the dramatic decrease in disease fol-
America shortly after World War II. lowing immunization in the 1920s. Furthermore,
overwhelming data indicate that proper vacci-
Production and Properties of Diphthe- nation can provide absolute protection. For ex-
ria Vaccine. Diphtheria toxoid is prepared by ample, mass vaccination with diphtheria toxoid
detoxification of diphtheria toxin with formalde- in Romania in 1958 showed that within seven
hyde. Derivatives of the hypertoxinogenic Park years morbidity and mortality due to diphtheria
Williams 8 strain are used by most manufac- declined by more then 99 %. In Western Europe
turers. Large quantities of toxin are produced and North America, where vaccination is univer-
in fermentors (ca. 200 – 5000 L) and yields ap- sal, diphtheria has been virtually eradicated [11].
proach 500 mg/L. Although manufacturing pro- Rare cases occur almost exclusively in individ-
cesses vary, most employ the following steps. uals lacking a proper history of vaccination.
Formaldehyde [50-00-0] (0.4 – 0.6 %) is added
to cell-free culture supernatants which are then Future Prospects. Given the combination of
stored at 35 – 37 ◦ C for 3 – 5 weeks. Detoxifica- vaccine safety, efficacy, and cost, there has been
tion is accomplished first by reaction of form- little impetus to develop and test a “second gen-
aldehyde with the ε-amino groups of lysine eration” diphtheria toxoid. Any such vaccine
and then by formation of irreversible methylene would have to completely eliminate the possibil-
bridges between aromatic amino acids. Upon ity of toxic reversion and be more economical to
confirmation of detoxification by animal test- produce. In another approach, synthetic peptides
ing, the toxoid is purified by diafiltration, ammo- that express key epitopes of diphtheria toxin are
nium sulfate or ethanol fractionation, or anion- conjugated to carrier proteins; these conjugates
exchange chromatography. can elicit neutralizing antibodies [12]. The draw-
back with this approach is the cost of purifying
Immunization Recommendations. Diph- the carrier protein.
theria toxoid is rarely used as a monovalent The most promising approach to toxoid de-
vaccine, but is usually combined with tetanus velopment is the use of nontoxic mutant proteins
toxoid (DT, see Section 2.2), pertussis vaccine [13], obtained by recombinant DNA technology.
(DPT, see Section 2.3), or inactivated polio vac- Such proteins can be constructed by deletion
cine (DPT-Pol, see Section 3.5) absorbed onto of specific regions responsible for toxicity [14].
an aluminum salt adjuvant. Toxoid content is In addition, these “toxoids” can be synthesized
usually expressed in limits of flocculation, Lf in yields equal to those obtained by production
(1 Lf ≈ 2 µg of toxoid). For primary immuniza- strains of C. diphtheriae. The efficacy of such
tion of infants, ca. 25 Lf of toxoid is adminis- proteins is being evaluated.
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,
10 Immunotherapy and Vaccines
Future Prospects. As with diphtheria tox- unvaccinated controls. Routine large-scale im-
oid, tetanus vaccine is an extremely economic, munization against pertussis began shortly after
effective, safe vaccine. Efforts directed to a “new World War II using standardized whole-cell vac-
generation” of tetanus vaccine are aimed at pro- cines of known potency. Even though the safety
viding a toxoid more amenable to mass vacci- of these vaccines has been under attack, they
nation. The main thrust has been to produce a are still used in most areas of the world except
safer, more immunogenic toxoid capable of in- Japan, where acellular vaccines have been used
ducing long-lasting immunity after one or two since 1980 (see below).
administrations without reactogenic adjuvants.
The substitution of glutaraldehyde for formal- Production and Properties of Pertussis
dehyde yields a safe, immunogenic vaccine not Vaccine. At present, there is no standardized
requiring adjuvants [19]. Production of toxoids method, culture medium, or bacterial strain for
by recombinant DNA techniques similar to those producing pertussis vaccine. A given manufac-
described for diphtheria toxin is also feasible turer’s production procedure is designed to yield
[14]. a vaccine which will meet the minimal require-
ments of the appropriate regulatory agency. B.
pertussis is usually grown in fermentors on a
2.3. Pertussis Vaccine synthetic or semisynthetic medium. The cells
are harvested by centrifugation and resuspended
Etiological Agent and Pathogenesis. Bor- to a given opacity using a reference standard.
detella pertussis, the cause of pertussis or Inactivation is accomplished by heating and/or
“whooping cough,” is a gram-negative rod that adding formaldehyde or thimerosal. In view
was first isolated by Bordet and Gengou in of the diversity of production techniques em-
1906. Humans are the only known host for ployed, toxic components which escape inacti-
B. pertussis. The disease is spread by infectious vation can vary considerably from manufacturer
droplets with the vast majority of cases occurring to manufacturer.
in young children. B. pertussis shows a marked
trophism for the ciliated epithelial cells of the up- Immunization Recommendations. Pertus-
per respiratory tract. Initial symptoms are sim- sis vaccine is used almost exclusively in combi-
ilar to those of the common cold, but worsen nation with diphtheria and tetanus vaccines (see
within 10 – 20 days. The paroxysmal stage, last- Sections 2.1 and 2.2). The vaccine is adsorbed
ing an average of 15 – 20 days, is characterized onto an aluminum salt adjuvant. Immunization
by severe bouts of coughing. The convalescent usually commences at 6 – 12 weeks of age and
stage can last for many months when secondary consists of three doses of vaccine given intra-
infections can present a major problem. B. per- muscularly at 4 – 8 week intervals with a fourth
tussis remains localized within the initial dis- dose given 4 – 12 months later. Vaccination with
ease stages which implies that the severe symp- the acellular vaccine used in Japan consists of
toms are toxin-induced [20]. This is supported 2 – 3 doses of vaccine given at 4 – 8 week inter-
by the fact that B. pertussis can synthesize a large vals starting at two years of age.
number of toxic extracellular factors including
pertussis toxin (also referred to as lymphocy- Adverse Reactions. The majority of chil-
tosis promoting factor), adenylate cyclase, der- dren receiving pertussis vaccine experience an
monecrotic toxin, and tracheal cytotoxin [21]. adverse reaction. Approximately 40 – 50 % of
children have a local reaction and 30 – 40 %
History of Immunization. Attempts at im- a systemic reaction (anorexia, vomiting, fret-
munization against pertussis were initiated in fulness, fever, or persistent crying) [23]. Of
the 1920s using killed whole-cell vaccines [22]. greater concern is the infrequent occurrence
These studies demonstrated that vaccination of convulsions and hypotonia. As a result of
could not only prevent a substantial proportion concern regarding vaccine safety, a large-scale
of disease, but that symptoms in vaccinated indi- study was initiated in England to determine
viduals who became infected were milder than in 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 Case-contact studies have shown the whole-
be 1/110 000 vaccinations and that for en- cell vaccine to be 63 – 95 % effective at prevent-
cephalopathy, 1/310 000 vaccinations. It is ex- ing overt disease. Efficacy of acellular vaccines
tremely difficult to establish a causal relation- is discussed below.
ship in individual cases in such a study; inci-
dence rates were evaluated on a temporal basis. Future Prospects. Efforts to develop a “sec-
Furthermore, neurological symptoms believed ond generation” pertussis vaccine have centered
to be vaccine-related constitute only a small pro- on using purified detoxified antigen preparations
portion of similar cases seen in this age group. containing a minimum amount of lipopolysac-
Review of these data by health authorities has charide. Several such acellular vaccines have
led to the conclusion that the benefit of vacci- been developed and clinically evaluated. They
nation outweighs the attendant risk in view of are composed of detoxified pertussis toxin
the fact that clinical pertussis can often result in (PT) alone or in combination with filamentous
neurological sequela. hemagglutinin (FHA), two key protective anti-
gens [25]. The first generation acellular vaccines
Vaccine Efficacy. The first indication of vac- were produced in Japan using supernatant from
cine efficacy came from a trial conducted in 1929 static cultures as a source of antigen. Cell-free
during an epidemic on the Faroe Islands; vac- supernatants were subjected to ammonium sul-
cination afforded 73 % protection against clini- fate precipitation followed by sucrose density
cal disease [22]. Extensive information has been gradient ultracentrifugation to simultaneously
obtained in Japan, England, and Sweden where enrich PT and FHA and eliminate lipopolysac-
disease incidence correlates inversely with the charide; PT was inactivated by formaldehyde
overall immune status of the general population. treatment. Preliminary testing in Japan showed
Routine immunization against pertussis in the vaccine to be far better tolerated than whole-
Japan was initiated in 1947 – 1949 with an ac- cell vaccine; case-contact studies revealed that
ceptance rate of roughly 90 %. The number of it was ca. 90 % effective.
pertussis cases declined from 152 072 in 1947 Second generation acellular vaccines of a
to less than 400 by 1971. In 1975, controversy greater purity have been produced on a large
concerning vaccine safety briefly halted vaccine scale using fermentor-grown cultures. A mono-
usage. When it was resumed, the acceptance rate valent formalin PT toxoid and a bivalent PT
declined to ca. 25 – 30 %. Pertussis returned to toxoid – FHA vaccine have been evaluated in a
epidemic proportions by 1979 with more than placebo-controlled trial in 5 – 11 month old chil-
13 000 cases reported nationally. At the urging dren in Sweden [26]. Children received 2 doses
of federal health authorities, vaccine acceptance of vaccine 8 – 12 weeks apart. The small num-
rates increased to 65 – 70 % by 1982 with a con- ber of children vaccinated (ca. 3000) does not
comitant decline in cases. allow evaluation of vaccine safety regarding the
Similarly, concern about vaccine safety in rare neurological reactions. However, the vac-
England resulted in a dramatic decline in vaccine cine evoked far fewer local reactions in compar-
acceptance from ca. 80 % in 1973 to ca. 30 % in ison to the whole-cell vaccine. After 15 months
1978. As in Japan, the disease became epidemic of observation, the PT toxoid vaccine was 54 %
with more than 100 000 cases reported between effective whereas the PT toxoid-FHA vaccine
1977 and 1980. In Sweden, routine vaccination was 69 % effective against all forms of clinical
against pertussis was initiated in the early 1950s. pertussis. Both vaccines were equally effective
One decade later, > 90 % of children were con- (ca. 80 %) against severe disease.
sidered to be immune. Due to changes in the
manufacturing technique, the pertussis vaccine
used in the mid-1970s was not potent. Shortly 2.4. Typhoid Fever Vaccine
after, a marked increase of pertussis cases was
seen. When an effective vaccine was reintro- Etiological Agent and Pathogenesis.
duced, disease incidence was dramatically re- Salmonella typhi, the causative agent of typhoid
duced. fever, is a gram-negative bacillus first isolated
Immunotherapy and Vaccines 13
by Gaffky in 1884. Infection is due to inges- [30]. A lyophilized Ty21a preparation afforded
tion of the organisms in contaminated food or > 90 % protection for three years in a field trial
water. The bacteria penetrate the epithelium of in Egypt [31].
the small intestine and are ingested by reticu-
loendothelial cells. Unlike most bacteria, the ty- Production and Properties of Typhoid
phoid bacillus can survive and multiply within Vaccines. Killed whole-cell vaccines are usu-
phagocytic cells and then spreads to the spleen, ally produced from the Ty2 strain of S. typhi.
liver, lymph nodes, and gallbladder. Symptoms Fermentor-grown cultures are inactivated either
appear 1 – 2 weeks after exposure when the bac- by heating in combination with phenol or form-
teria enter the bloodstream. Despite appropriate aldehyde or by acetone drying. Although the lat-
treatment, 1 – 2 % of those infected will become ter method yields a more efficacious vaccine (see
chronic asymptomatic carriers serving as an in- below), only the vaccine produced by the former
fectious reservoir. Typhoid primarily occurs in method is readily available. Each 0.5 mL dose
developing countries or in areas of poor sani- of vaccine contains (1 – 3)×109 killed organ-
tation. In endemic areas, the disease is primar- isms. Vaccine potency is estimated by a mouse-
ily contracted by school-aged children, but trav- protection test.
ellers of all ages entering an endemic area are at The attenuated live-oral Ty21a vaccine strain
risk. is produced from fermentor-grown cultures. The
harvested cells are suspended in a cryoprotec-
History of Immunization. Attempts at im- tive medium consisting of sugar and amino acids
munizing against typhoid fever began in 1896. and lyophilized. The lyophilizate is placed in
The first mass vaccination was conducted by the gelatin capsules which are then coated with
British Army during World War I using a killed an acid-resistant layer. Each capsule contains
whole-cell vaccine. The attack rate among vac- (1 – 5)×109 viable bacteria. Vaccine potency is
cinated soldiers was far less than for nonimmu- based upon the number of viable organisms per
nized soldiers. Consequently, the use of whole- capsule [32].
cell vaccines became common practice in the
armed forces. Immunization Recommendations. People
In the 1960s and 1970s, attempts were made living in or travelling to an endemic area are
to use killed oral immunization. Killed bacte- vaccinated with two doses of whole-cell vac-
ria were administered to volunteers at 1×1011 cine administered subcutaneously or intramus-
per dose; twelve doses resulted in a modest, but cularly 14 days apart. A single booster dose is
significant protection rate of 30 % [27]. Field tri- advised for individuals travelling to an endemic
als in India showed that such killed preparations area if 2 – 3 years have elapsed since primary
were ineffective at preventing disease. immunization.
An alternative approach to killed oral vac- Three doses of attenuated live-oral vaccine
cines against typhoid was to use attenuated live- are ingested, each dose on an alternate day. An
oral vaccine strains. The first candidate was additional complete immunization course is ad-
a streptomycin-dependent mutant developed in vised when entering an endemic area if more
1967 [28]. When orally administered to volun- than 1 – 2 years have elapsed since primary im-
teers in doses of up to 1011 bacteria, this strain munization. The vaccine is for use in children
did not evoke adverse reactions. Good levels six years of age or older and in all ages of non-
of protection (66 – 78 %) were achieved with immunocompromised adults.
freshly harvested cultures. However, lyophilized
preparations afforded little or no protection [28], Adverse Reactions. Reactions following
which diminished interest in this strain. immunization with the parenteral whole-cell
The second live-oral vaccine candidate was a vaccine are frequent and consist of local pain,
gal ε mutant termed Ty21a which lacked glucose swelling, and redness, often accompanied by
1,4-epimerase activity and therefore could not fever, chills, and malaise. Transient debilitating
synthesize complete lipopolysaccharide [29]. reactions occur in ca. 10 % of vaccine recipi-
Freshly harvested cultures of this strain were ents. Due to the high rate of adverse reactions,
safe and effective in volunteer challenge studies parenteral typhoid vaccine is not widely used.
14 Immunotherapy and Vaccines
Reactions following ingestion of the live oral pneumococcal pneumonia. Streptococcus pneu-
Ty21a vaccine are rare (< 1 per 100 000 doses). moniae is also a leading cause of otitis media in
Reactions are usually mild, transient, and re- young children. The pneumococcus often colo-
solve of their own accord. Gastrointestinal dis- nizes the nasopharynxes of healthy individuals
turbances and rash are most frequently reported. which then serve as infectious foci. The organ-
isms can then spread to another person by in-
Vaccine Efficacy. To accurately assess the fected droplets or translocate to the inner ear or
efficacy of parenteral typhoid vaccine, several lungs following a viral infection.
field trials were conducted in the 1960s [33].
In most instances, two doses of either heat– History of Immunization. Early attempts
phenol or acetone-dried vaccine were admin- at immunization against pneumococcal pneu-
istered. Protection rates ranged from 51 – 77 % monia employed heat-killed, whole-cell vac-
and 79 –93 %, respectively, for up to seven years. cines (reviewed in [37] ). These trials and re-
A further trial was conducted in Tonga in 1966 – lated studies led to the conclusion that immunity
1973 where school-aged children received one is mediated by serospecific antibody produced
or two doses of acetone-dried vaccine. Surpris- against the polysaccharide capsule surround-
ingly, one dose gave no protection whereas two ing the pneumococcus. A large-scale trial per-
doses afforded only 40 % protection. formed in the mid-1930s with a bivalent capsular
Efficacy of the live oral Ty21a vaccine ad- polysaccharide (CPS) vaccine showed a modest
ministered in enteric-coated capsules has been decrease in disease. Conclusive proof of vaccine
evaluated in field trials in school-aged children efficacy came from trials conducted with polyva-
in Chile [34]. Three doses of vaccine provided lent vaccines; protection was serospecific. Fur-
about 70 % protection over a three-year period. ther studies showed that vaccines comprising
up to six distinct capsular antigens could be
Future Prospects. Two additional typhoid effectively used [38]. The first pneumococcal
vaccines are undergoing clinical evaluation. The polysaccharide vaccines were hexavalent and
first is a double auxotrophic mutant of S. typhi re- were licensed in the early 1950s. However, ow-
quiring 4-aminobenzoate and adenine [35]. Al- ing to the belief that antibiotics could effectively
though ingestion of up to 1010 organisms only control pneumococcal disease, vaccine accep-
evoked a poor humoral immune response to tance was low, leading to a halt in production.
S. typhi cellular antigens, most vaccines mani-
fested a specific cell-mediated immune response Production and Properties of Streptococ-
to S. typhi lipopolysaccharide. cus pneumoniae Vaccine. This vaccine is com-
The second candidate vaccine is a capsular posed of 23 types of purified CPS that have
polysaccharide termed Vi . Purified Vi antigen been selected on the basis of seroepidemiolog-
is safe upon parenteral immunization. A single ical surveillance studies of bacteremic isolates
dose of vaccine provided ca. 70 % protection for [39].
ca. one year in Nepal [36]. Pneumococcal CPSs are purified by a va-
riety of techniques, depending upon serotype,
which include precipitation with organic sol-
2.5. Streptococcus pneumoniae Vaccine vents, treatment with detergent, digestion with
DNAse, RNAse, and protease, and ultracentrifu-
Etiological Agent and Pathogenesis. Strep- gation. The vaccine contains only traces of pro-
tococcus pneumoniae, often referred to as tein and nucleic acids (≤ 2 wt %) and is analyzed
“pneumococcus,” is a gram-positive diplococ- for its CPS constituents, serological purity, and
cus first isolated by Sternberg and Pasteur pyrogenicity. Vaccine potency is based upon the
in 1881. The pneumococcus usually causes dis- determination of the molecular mass of the CPS,
ease subsequent to a viral infection or in a because a minimum antigen size is required for
compromised individual, i.e., is an opportunistic immunogenicity in humans. One human dose
pathogen. Pneumonia is the most common dis- consists of 25 µg of each antigen in 0.5 mL ad-
ease syndrome caused by S. pneumoniae; bac- ministered intramuscularly or subcutaneously.
teremia and meningitis are often sequelae to
Immunotherapy and Vaccines 15
vaccines were not pursued due to the necessity History of Immunization. Parenteral
of administering four doses for primary immu- whole-cell cholera vaccines were employed by
nization, yearly boosting to maintain efficacy, Fenan as early as 1885. Starting in the 1960s,
and genetic instability. several field trials were performed using inac-
Attempts have been made to produce live- tivated whole-cell vaccines or cholera antigens
oral shigella vaccines by conjugal transfer of (see “Vaccine Efficacy”).
genes coding for S. flexneri surface antigens into Numerous attempts have been made to orally
E. coli. Multiple doses evoked no significant ad- vaccinate against cholera. Multiple oral doses
verse reactions when fed to volunteers, but failed of heat-killed V. cholerae evoked both a local
to protect against subsequent challenge [52]. and humoral antibody response [58]. Ten daily
The identification of critical Shigellae- oral doses of 1.6×1010 killed vibrios provided
protective antigens has allowed their expression protection against a homologous challenge [59].
in the attenuated live-oral typhoid vaccine strain, However, parenteral vaccine gave slightly bet-
S. typhi Ty21a [53]. One strain, S. typhi Ty21a- ter protection. Although promising, the massive
5076-1C, carries a plasmid coding for the form I quantities of organisms used and the need for
(O-polysaccharide) antigen of S. sonnei. This multiple doses made the oral approach expen-
strain is safe and has afforded significant pro- sive and cumbersome.
tection in volunteer studies [54]. Several environmental isolates of V. cholerae
with reduced virulence have been evaluated as
Future Prospects. At present, no vaccines live-oral vaccines. These strains failed to pro-
are available for use against shigellosis. As noted vide significant protection due to their inability
above, S. typhi 5076-1C that expresses S. sonnei to colonize the small intestine and evoke a pro-
form I antigen has shown promise in volunteer tective immune response [60]. A hypotoxino-
studies. However, efficacy varies from lot to lot genic mutant of V. cholerae isolated by chemi-
and probably depends on the degree of flagella- cal mutagenesis afforded significant protection
tion of the bacteria. The construction of a S. typhi against diarrhea. However, due to its propen-
Ty21a strain expressing S. flexneri 2a type and sity to revert, it was not studied further [61].
group antigens has been described [55], but this Chemical mutagenesis has provided a strain of
strain has not yet been clinically evaluated. V. cholerae 3083 (Texas Star-SR), that produces
Genes coding for the production of the O- only the B subunit of cholera toxin responsible
antigen of S. dysenteriae 1 have been cloned and for binding to cells and lacks the enzymatically
inserted in a plasmid [56]. The S. dysenteriae active A subunit. Although vaccination with this
O-antigen can be expressed in E. coli K12 and strain afforded good levels of protection against
S. typhi Ty21a by introduction of the recom- cholera, about 25 % of vaccinees experienced
binant plasmid. Volunteer studies to determine mild to moderate diarrhea [62].
their safety are scheduled.
Production and Properties of Cholera Vac-
cine. The licensed cholera vaccine is composed
2.7. Cholera Vaccine of inactivated whole cells of V. cholerae and
is administered parenterally. Fermentor-grown
Etiological Agent and Pathogenesis. The cultures are inactivated by a combination of
causative agent of cholera is Vibrio cholerae, heat and phenol or formaldehyde. Usually,
a gram-negative motile bacillus which is trans- two cultures of Ogawa-classical and Inaba-El
mitted via the fecal–oral route. Upon passage Tor serotypes-biotypes are produced and then
through the stomach acid barrier, V. cholerae mixed. Cell concentration is adjusted to (5 –
colonizes the ileum where it rapidly multiplies. 8)×109 cells/dose (0.5 mL). Vaccine potency is
The release of a potent enterotoxin (cholera estimated by an intraperitoneal mouse challenge
toxin) causes massive watery diarrhea [57]. In test. Two doses of vaccine given at two-week in-
most cases, oral or intravenous fluid replacement tervals are recommended. A booster dose is rec-
is sufficient treatment. ommended whenever entering an endemic area.
Immunotherapy and Vaccines 17
Future Prospects. No vaccines are licensed Serospecific Vaccine against Klebsiella spp.
against the four pathogens listed above. At- Antibody to Klebsiella capsular polysaccharide
tempts to control bacterial nosocomial infec- (CPS) is highly protective against experimental
tions by immunological means are complicated infections [76]. Several experimental vaccines
because of the diverse patient populations who composed of purified Klebsiella CPS were found
are at high risk and the substantial number to be safe and immunogenic in human volun-
of bacterial pathogens of varying serotypes in- teers [70]. Based upon the seroepidemiology of
volved. Furthermore, vaccination of at-risk pa- Klebsiella blood isolates, a 24-valent CPS vac-
tients may not be feasible because of the time cine has been developed which is safe and im-
needed to mount a protective immune response munogenic in humans. It also elicits production
and because a substantial portion of these pa- of antibody to 11 “cross-reactive” CPS serotypes
tients are unable to mount a significant anti- not included in the vaccine and therefore, of-
body response. A more effective approach is fers potential “coverage” against ca. 75 % of all
to passively immunize patients by using a hy- Klebsiella bacteremic isolates.
perimmune globulin for intravenous use (see
Chap. 5): vaccines developed against nosoco- Serospecific Vaccine against Escherichia
mial pathogens could then be used to vacci- coli. Antibodies to both O (LPS) and K (cap-
nate healthy donors whose plasma would be pro- sular) antigens can provide protection against
cessed into such a globulin [70]. experimental E. coli infections [77], [78]. How-
Two different approaches are being taken ever, ca. 40 % of E. coli bacteremic isolates
to develop appropriate hyperimmune products. do not have a polysaccharide capsule [79]. In
The first is based upon developing vaccines addition, the two most frequently encountered
against the relevant serospecific bacterial anti- K serotypes among blood isolates, K 1 and
gens. The second is to evoke an antibody re- K 5, are nonimmunogenic in humans. Approxi-
sponse to a common epitope expressed by the mately 90 % of E. coli bacteremic isolates can be
lipopolysaccharide of gram-negative bacteria. typed according to their O-antigen and most of
them can be grouped within 11 serotypes, mak-
Serospecific Vaccine against Pseudomonas ing a polyvalent formulation feasible [79]. The
aeruginosa. Human immunity to P. aeruginosa toxicity of native LPS precludes its use as a vac-
depends on the presence of humoral anti- cine. However, nontoxic, serologically reactive
body directed against serospecific lipopolysac- O-polysaccharide can be isolated from E. coli
charide (LPS) determinants and toxin A [72]. LPS and O-polysaccharide – protein conjugates
Several lipopolysaccharide-containing vaccines are being constructed in a manner similar to that
have been developed and clinically evaluated described for P. aeruginosa (see above).
[73]. These vaccines have not gained wide ac-
ceptance due to their toxicity, poor immuno- Serospecific Staphylococcus aureus Vac-
genicity, and poorly characterized antigenic con- cine. Staphylococcus aureus produces numer-
tent. Two serotypes of P. aeruginosa high molec- ous somatic antigens and extracellular toxins
ular mass polysaccharides have been purified which have been implicated as virulence fac-
and found to be safe and immunogenic in hu- tors. These include teichoic acid, lipoteichoic
mans [74]. In an attempt to produce both an anti- acid, exopolysaccharide, capsular polysaccha-
toxin A and an antiLPS antibody response, toxin ride, α-toxin, and β-toxin. Patients with deep-
A – O-polysaccharide conjugate vaccines were seated S. aureus disease usually mount an anti-
synthesized. Such conjugates are nontoxic, non- body response to teichoic acid, α-toxin, and β-
pyrogenic, safe when administered to humans, toxin [80]. However, the relative protective ca-
and evoke an immune response to both vac- pacities of antibodies against these antigens are
cine moieties [75]. A polyvalent vaccine based unknown.
upon O-polysaccharide serotypes is undergoing A serotyping scheme has been described
clinical evaluation; this vaccine would “cover” for S. aureus based upon cell-surface capsular
>90 % of P. aeruginosa bacteremic isolates. polysaccharides [81]. Approximately 80 – 90 %
of S. aureus clinical isolates belong to one of
these serotypes [82]. Serospecific CPS is pro-
Immunotherapy and Vaccines 19
duced during experimental S. aureus infections meningococcal meningitis with antibiotics ap-
and can be isolated from blood [83]. The protec- peared to be highly effective. However, the ap-
tive capacity of antibody to CPS is being evalu- pearance of resistant strains in the early 1960s
ated. revived interest in vaccine development. In the
late 1960s, highly purified group A and group C
Crossreactive Anticore Glycolipid Vac- capsular polysaccharides with a high molecular
cine. Antibody produced against the “core” gly- mass were shown to be safe and immunogenic
colipid of E. coli strain J 5 crossreacts with the in humans [91]. Human immune sera contained
LPS of virtually all gram-negative bacteria [84]. elevated levels of antibody capable of lysing
Polyclonal or monoclonal antiJ 5 antibody (see meningococci in the presence of complement.
also Section 6.1) can afford significant protec-
tion against several gram-negative pathogens Production and Properties of Meningo-
[85]. Protective antibody is directed against the coccal Vaccines. Available vaccines contain
lipid A moiety of LPS and neutralizes the toxic serogroup A, C, W 135, and Y capsular anti-
activities of LPS. When used prophylactically gens. The most widely used formulations are
or therapeutically, human plasma enriched with the tetravalent A, C, W 135, and Y, and the bi-
antiJ 5 antibody provides significant protection valent A and C vaccines. Capsular antigens are
against death due to “gram-negative shock” [86], purified from fermentor-grown cultures by co-
[87]. precipitation with detergents followed by eth-
anol fractionation, extraction with organic sol-
vents, and ultracentrifugation. Preservation of
2.9. Meningococcal Meningitis Vaccine the high molecular mass of the capsular anti-
gens is of critical importance since this governs
Etiological Agent and Pathogenesis. The the immune response. Each vaccine dose con-
causative agent of meningococcal disease is tains 50 µg of capsular antigen per serotype in
Neisseria meningitidis, a gram-negative diplo- lyophilized form. Potency of the vaccine is based
coccus. Humans are the only known reservoir upon various physicochemical properties (size
for N. meningitidis. The meningococcus col- and purity). No animal test is required.
onizes the nasopharynx; infrequently, it en-
ters the bloodstream and infects the meninges. Immunization Recommendations.
Meningococcal disease is most often seen in Meningococcal vaccine can be administered
children less than 18 months of age and in to any immunocompetent individual over 18
settings such as military recruit camps where months of age. A single dose is given either
individuals from different geographical areas subcutaneously or intramuscularly. The group A
come into constant close contact [88]. The dis- antigen is immunogenic in children ≥6 months
ease is endemic worldwide with an annual in- of age, but two doses must be administered
cidence of about 3/100 000 population. Numer- within a 2 – 3 month interval. Due to the low in-
ous epidemics occurred in the 1970s and 1980s cidence of the disease in the general population,
with attack rates exceeding 500/100 000. Eleven vaccine acceptance has been poor, it is usually
meningococcal capsular serotypes are known, used to immunize military recruits or travellers.
with the majority of disease (ca. 95 %) caused
by groups A, B, C, W 135, and Y [88], [89]. Adverse Reactions. The meningococcal
vaccines are very safe–about 250×106 doses
History of Immunization. The first attempts of vaccine have been administered worldwide
at immunizing against meningococcal disease with no vaccine-associated fatalities reported.
used vaccines made of killed bacteria and were Mild, transient local reactions occur in about
unsuccessful [90]. Attention then centered on 25 % of vaccinees.
using purified capsular antigens. In 1945 Kabat
demonstrated that up to 1 mg of purified antigen Vaccine Efficacy. Conclusive proof that
could be administered to humans with no unto- group C meningococcal vaccine prevents dis-
ward reactions. However, these studies were not ease was presented in 1970 [92]. Subsequently,
continued because the prophylactic treatment of the group A vaccine was found to be effective
20 Immunotherapy and Vaccines
bilizer solution, homogenized, and lyophilized. not immunocompromised as the vaccine organ-
The bulk material is standardized by opacity, isms can then rapidly multiply and lead to death
dry mass, and number of viable bacteria per unit [105].
mass. The vaccine is usually filled into glass am-
pules which are sealed under vacuum. Vaccine Efficacy. In numerous field trials
Controls of BCG vaccine are stringent. First, the efficacy of BCG vaccine ranged from 0 –
the culture is examined for purity, viability, and 80 % [100]. Although three trials failed to doc-
identity. Innocuity is tested by injection of vac- ument significant vaccine-induced protection,
cine into guinea pigs. The number of live or- considerable support still exists for large-scale
ganisms are determined by viability counting. vaccination in developing countries [106].
General safety tests to confirm absence of unex-
pected toxicity are also performed. Resistance of Future Prospects. At present, no “second-
the bacteria to elevated temperature is also docu- generation” tuberculosis vaccines are being clin-
mented. Vaccine potency is based upon the total ically evaluated. Although the tuberculin skin
viable bacteria per dose. However, this is not test offers a practical method to measure “im-
totally satisfactory because (1) different BCG munity” to disease, the molecular mechanism re-
strains may vary in their ability to evoke a protec- sponsible for inducing a tuberculin-positive state
tive cell-mediated immune response, and (2) the is not known.
ratio of viable to nonviable bacteria can influ- Several M. tuberculosis cell surface anti-
ence the magnitude of both the humoral and gens have been identified which can induce
cell-mediated immune response to vaccination a cell-mediated response in experimental ani-
[102]. mals [107]. The cloning of such antigens and
their high-level expression in bacteria or yeast
Immunization Recommendations. A sin- would permit economical large-scale purifica-
gle intradermal dose containing (4 – 8)×105 vi- tion. Alternatively, such cloned genes could be
able organisms in 0.1 mL is normally adminis- inserted into a suitable vector such as vaccinia
tered. The preferred use of BCG vaccine de- virus [108]. Since most tuberculosis vaccine are
pends on the disease incidence in the area in used in underdeveloped nations, any new vac-
question. In developed areas where disease is cine must not only be safe and effective, but also
infrequent, vaccination should be limited to in- economical to produce and control.
dividuals at high risk to exposure, i.e., family
members of tuberculosis patients, health care
workers expected to come into contact with tu- 2.11. Escherichia coli Vaccines
berculosis patients or infected clinical speci-
mens, or travellers entering an area of high en- Etiological Agent and Pathogenesis. Es-
demicity for a prolonged period of time. In de- cherichia coli, a gram-negative motile rod, is a
veloping areas with a high rate of disease inci- major cause of diarrheal disease, urinary tract
dence, routine immunization of infants is a cost- infections, neonatal meningitis, and nosocomial
effective method of controlling tuberculosis. In infections (see Section 2.8). Escherichia coli
addition, all tuberculin-negative children upon produces a variety of virulence factors including
entry into school and tuberculin-negative adults a polysaccharide capsule antigen (K-antigen),
should be vaccinated. lipopolysaccharide antigen (O-antigen), fim-
briae, enterotoxins, and cytotoxins. The major-
Adverse Reactions. Correctly administered ity of E. coli strains causing a given disease fall
BCG vaccine is considered to be safe; one vac- within a limited number of K or O serotypes.
cination in 100 000 will result in a noticeable Escherichia coli is the major cause of diar-
reaction [103]. A significant proportion of these rhea worldwide. Three “categories” are recog-
reactions are ulcers caused by inadvertent subcu- nized, i.e., enterotoxigenic, enteroinvasive, and
taneous administration of the vaccine. Of greater enteropathogenic E. coli strains.
concern are the reports of mycobacterial disease
following immunization [104]. Great care must 1) Enterotoxigenic E. coli invariably causes wa-
be taken to insure that a potential vaccinee is tery diarrhea which may be accompanied
22 Immunotherapy and Vaccines
by cramps, fever, or vomiting. The dis- infecting organisms are thought to be acquired
ease is acquired by ingesting contaminated from the flora colonizing the vagina and peri-
food and affects infants, young children, and urethral areas. P and type I fimbriae are believed
travellers to underdeveloped countries. The to play an important role by mediating tissue
strains produce either a heat-stable entero- colonization [109].
toxin and/or a heat-labile enterotoxin; they
also possess fimbriae which mediate attach- History of Immunization. Initial attempts
ment to the ileal epithelium. Three distinct to vaccinate against E. coli diarrhea used crude
fimbriae (CFA/I, CFA/II, and E 8775) have cellular extracts derived from O types 111, 55,
been identified on enterotoxigenic E. coli and 86 [110]. In one study, hospitalized infants
strains isolated from human disease. up to one year of age received multiple doses
2) Enteroinvasive E. coli causes a dysentery- of oral vaccine. Overall efficacy was 41 %. Ef-
like disease characterized by fever, diarrhea, forts to develop vaccines against E. coli diar-
cramps, and stools containing blood and mu- rhea center around the use of purified antigens
cus. The disease is acquired by ingestion of or live-oral attenuated strains to elicit an anti-
contaminated food. The bacteria readily pen- toxic or antiadhesion response. Calves or neona-
etrate the ileal and colonic epithelia where tal piglets reared by mothers immunized with
they rapidly multiply resulting in tissue de- purified K 88, K 99, or 987 P fimbriae, proc-
struction. They can then infect adjacent cells holeragnoid (heat-aggregated cholera toxin with
or penetrate the underlying lamina propria. reduced toxicity) or heat-labile enterotoxin are
3) Enteropathogenic E. coli affects primarily protected against E. coli bacillosis [111–113].
infants and young children after ingestion of Graded doses (45 – 1800 µg) of purified type I
contaminated food causing watery diarrhea fimbriae have been parenterally administered to
often accompanied by fever and/or vomit- human volunteers [114]. The vaccine was well
ing. These E. coli strains rarely produce a tolerated and elicited both a humoral IgG and
heat-labile or heat-stable enterotoxin and are an intestinal secretory IgA antibody response.
unable to invade eukaryotic cells. They col- A moderate level of serospecific protein was ob-
onize the ileal epithelium with destruction tained in volunteer challenge studies.
of the surrounding brush border in the ab- Oral administration of 1 mg of purified CFA/I
sence of invasion. Enteropathogenic E. coli evoked a significant secretory IgA response in
strains can synthesize a cytotoxin similar to about 50 % of vaccinated humans [115]. Re-
that produced by Shigella dysenteriae 1. This sponders were protected against challenge with
so-called Shiga-like toxin is thought to elicit an St+ , LT+ , CFA/I+ E. coli strain.
the clinical symptoms.
Future Prospects. No vaccines are available
Approximately 50 % of neonatal meningi- to prevent diarrhea urinary tract infections, or
tis is due to E. coli transmitted from mother meningitis caused by E. coli. At present, both
to infant. The bacteria apparently penetrate the attenuated live-oral and inactivated vaccines are
intestinal epithelium of the neonate, enter the being evaluated as to their ability to prevent
bloodstream, and are then translocated to the E. coli diarrhea. The production of a killed, inac-
meninges. The majority of strains (>80 %) caus- tivated, oral, whole-cell vaccine from a CFA/I+
ing meningitis express the K 1 capsular antigen. strain has been described [115]. The majority of
However, the O-antigen also appears to play a vaccinees responded with a good secretory IgA
role in virulence. Escherichia coli meningitis antibody response to CFA/I. The diarrhea attack
has a high attendant fatality rate. rate for volunteers challenged with a toxinogenic
Escherichia coli is the causative agent of ca. CFA/I+ E. coli strain was 89 % for those in the
90 % of nonobstructive urinary tract infections placebo group and 20 % for those in the vaccine
which are much more common among women group.
than men. The clinical syndromes can include A spontaneous mutant of E. coli has been iso-
bacteriuria, cystitis, and pyelonephritis. Renal lated which has lost the genes for production
scarring associated with pyelonephritis occurs of the heat-labile and heat-stable enterotoxins
most often in children ≤5 years of age. The but still expresses CFA/II. This strain has been
Immunotherapy and Vaccines 23
proposed as a likely carrier for cloned genes ex- can cause disease [123]. The first step in the
pressing relevant protective antigens such as col- infectious process is pili-mediated attachment
onization factors or “enterotoxoids” [116]. Sev- of the gonococcus to mucus-secreting epithelial
eral purified antigens show promise as oral vac- cells. Shortly thereafter, the bacteria are internal-
cine candidates including conjugates of the two ized and enter the submucosa [121]. Treatment
enterotoxins [117], the B subunit of the heat- of gonorrhea has been complicated by the oc-
labile toxin [118], and toxoid termed “procol- currence of multiple antibiotic-resistant clinical
igenoid” produced by heating the heat-labile en- isolates.
terotoxin [119].
Since the majority of E. coli strains associ- History of Immunization. The first attempt
ated with neonatal meningitis possess the K 1 to prevent gonorrhea by immunization used
capsule antigen, this would appear to be an ex- a killed whole-cell vaccine administered par-
cellent vaccine candidate. However, the K 1 anti- enterally [124], but did not provide protection
gen (which is identical to the group B meningo- in Eskimos. Subsequent human vaccine trials
coccal capsule) is not immunogenic in humans have used purified intact pili: antipili antibody
(see Section 2.9). blocks attachment of the gonococcus to epithe-
The most promising vaccines to prevent seri- lial cells, and, an antipili secretory IgA anti-
ous E. coli urinary tract infections (pyelonephri- body response is mounted during natural in-
tis) contain type I fimbrial antigens. Two fimbrial fection [125]. Vaccination of human volunteers
amino acid sequences have been shown to pre- with purified N. gonorrhoeae pili elicited a hu-
vent colonization by a homologous E. coli strain moral and a local vaginal antibody response.
in a murine pyelonephritis model [120]. Anti- These antibodies blocked the attachment of
body to one sequence also afforded protection N. gonorrhoeae to epithelial cells in vitro [125].
against an E. coli strain expressing a distinct fim- Two 2 mg doses of a monovalent pili vaccine
brial serotype. resulted in significant protection against chal-
lenge with the homologous strain [126]. How-
ever, no protection was observed when the chal-
2.12. Neisseria gonorrhoeae Vaccine lenge strain possessed a pilus serotype modestly
cross-reactive with the vaccine [121]. A subse-
Etiological Agent and Pathogenesis. quent field trial with this monovalent vaccine
Neisseria gonorrhoeae was identified as the failed to show any protection [125].
causative agent of gonorrhea by Bumm in 1885.
It is a gram-negative diplococcus with hu- Future Prospects. No vaccine has been li-
mans serving as its only known natural host. censed for the prevention of gonorrhea. The
Gonorrhea is a sexually transmitted disease of above human volunteer studies showed that pro-
epidemic proportions worldwide. More than tection against gonorrhea mediated by antipili
500 000 cases are reported annually in the antibody is serospecific [125]. Two different ap-
United States. The infection is usually restricted proaches are being taken to circumvent the prob-
to the entry site, i.e., the genitalia, pharynx, lem of antigenic variation among pili:
and/or rectum. More than 90 % of infected males
display symptoms characterized by pain upon 1) Use of a polyvalent vaccine containing those
urination and an urethral exudate [121]. How- types of pili which predominate in a given
ever, up to 50 % of infected women remain area [126].
asymptomatic and can serve as carriers. In ap- 2) Use of conserved antigenic domains from
proximately 10 % of women with genital gon- the antigenic pili molecule. A cyanogen bro-
orrhea, the bacteria invade the bloodstream and mide cleavage fragment can induce produc-
result in pelvic inflammatory disease. Dissemi- tion of an antibody that can recognize var-
nated gonococcal infections are far less frequent ious pili types [127]. Peptides correspond-
in male patients. ing to this conserved region block attach-
When cultured on solid medium, four colo- ment of the gonococcus to epithelial cells
nial variants (T 1 – T 4) can be discerned [122] ; in vitro [128]. If such peptides can induce
however, only the piliated T 1 and T 2 variants a broadly protective immune response in
24 Immunotherapy and Vaccines
polysaccharide conjugate are being evaluated in strains are included in current measles (and com-
clinical trials [135], [136]. A large-scale Finnish bined measles – mumps – rubella) vaccines. At-
trial is being performed to evaluate the diphthe- tenuation of the original Edmonston strain in
ria toxoid conjugate vaccine. Vaccine efficacy human diploid cells yielded the Edmonston-
was 83 % in children vaccinated at 3, 4, 6, and Zagreb strain, also in current use [139]. With the
14 months of age [137]. In contrast, this vaccine introduction of general immunization against
afforded insignificant protection when used to measles, the disease has almost disappeared in
immunize Alaskan Eskimo children at 2, 3, and many countries (see below).
4 months of age.
Production and Properties of Measles Vac-
cines. The attenuated measles vaccine strain is
3. Viral Vaccines usually grown in primary culture using chick
embryo cells or human diploid cells. The WHO
3.1. Measles Vaccine test requirements for the seed virus and cell sub-
strates are described in[140, pp. 52 – 73],[141,
Etiological Agent and Pathogenesis. p. 179 ]. Tests for monitoring measles vaccine
Measles (rubeola) is caused by a paramyxovirus grown in chick embryo cells include tests for
first isolated by Enders and Peebles in 1954. nonadsorbing viruses and avian leukosis viruses.
This highly contagious disease is transmitted For vaccines grown in human diploid cells, chro-
in an aerosol of virus-infected droplets from mosomal monitoring requirements have been
the respiratory tract. During the prevaccination formulated. Potency is checked by titration of
era, children in industrialized countries gener- the virus in tissue culture, (minimum require-
ally had a self-limiting disease characterized ment, 1000 TCID50 per human dose). To test
by conjunctivitis, bronchitis, fever, and a rash; for stability, a sample of the final freeze-dried
complications included pneumonia (mostly bac- vaccine is incubated at 37 ◦ C for seven days: the
terial) and encephalitis. A late complication of sample must then contain at least 1000 TCID50
measles in early childhood is the development of in each human dose. The vaccine is lyophilized;
a progressive, fatal neurological disease, termed reconstituted vaccine should be used immedi-
subacute sclerosing panencephalitis. Death due ately or stored at 0 – 10 ◦ C for not more than
to complications following measles infection 8 h.
remains one of the leading causes of child mor-
tality in developing countries. Immunization Recommendations. The
vaccine is given at the age of 15 – 24 months
History of Immunization. Two approaches [142] or even younger in developing coun-
to vaccine development were adopted, one was tries [139]. In the United States and many
based on killed (inactivated) strains and the other European countries, a combined measles –
on live, attenuated strains [138]. The formalde- mumps – rubella (MMR) is usually given (see
hyde-inactivated, alum-precipitated vaccine was Section 3.4). As for all live vaccines, immu-
licensed in the United States in 1963 and with- nization of immuncompromised individuals is
drawn in 1967 because its protective effect was not generally recommended. Immunization of
insufficient. In addition, the vaccinees exposed pregnant women should be avoided although
to the wild virus developed atypical measles with no increased risks have been documented for
fever and rash but also pneumonitis. measles vaccine.
The attenuated live vaccine with the Edmon-
ston B strain was also licensed in 1963 in the Adverse Reactions. Current attenuated
United States. A large-scale field trial in the measles strains give a few mild reactions, mostly
United Kingdom showed high seroconversion consisting of fever and rash 7 – 10 days after
rates, a long antibody response, and a 84 – 94 % immunization. The neurologic reaction of the
decrease in attack rate among the 36 000 im- Guillian-Barré syndrome has been reported but
munized children. Further attenuation of the is extremely rare.
vaccine strain in chick embryo cells yielded
the Schwartz and the Moraten strains. Both
26 Immunotherapy and Vaccines
Vaccine Efficacy. The rapid decrease of infected individuals. Most cases occur in chil-
measles morbidity in the United States after in- dren 5 – 10 years of age. The most frequent com-
troduction of general immunization is a good plication is meningoencephalitis, giving clinical
example of vaccine efficacy [142], [143]. Over symptoms in >10 % of patients. Orchitis, a less
a period of 20 years the rate of notified cases fell common but feared complication in postpuber-
by more than 99 %. Subacute sclerosing panen- tal males, results in impairment of fertility in
cephalitis also decreased. The overall protective 10 – 20 % of cases but absolute sterility is rare.
efficacy of the current strains is 90 % but is lower
in children of 12 months of age due to the pres- History of Immunization. Isolation of the
ence of maternal antibody. Life-long immunity causative agent for mumps by Habel in 1945
after one dose of vaccine has not been proven. from chick embryo was followed by attenuation
Increased disease incidence in the United States studies for vaccine development. Development
has raised the question of the possible need for of an inactivated (killed) vaccine, was also ini-
a two-dose schedule [142]. tiated; a killed vaccine was licensed in the US
in 1950 – 1978 but had low long-term protective
Future Prospects. Given the efficacy and efficacy [146]. An inactivated mumps vaccine
acceptability of the currently used attenuated was also produced in Finland for immunizing
measles strains, there has been little incentive military recruits; it decreased disease incidence
for vaccine development. However, current vac- by 94 % [147]. Immunization with an attenu-
cines are not very stable at high temperature, ated mumps vaccine (Jeryl Lynn strain cultured
which is a concern in many developing coun- in chick embryonic cells), licensed in 1967 in
tries. Furthermore, a large proportion of mor- the United States resulted in a 97 % decline of
bidity and mortality in children in developing disease incidence in the general population by
countries occurs before the recommended age 1981 [146], [148], [149]. Chick embryonic cells
for immunization. Research is centered on solv- are used to propagate another highly attenuated
ing these two problems. More immunogenic strain, Urabe Am 9 developed in Japan [150].
strains and/or higher doses of attenuated vac- The Rubini strain, developed in Switzerland, is
cines are being investigated [139]. Another ap- used in a vaccine in which the virus is grown in
proach would be to renew work on the devel- human diploid cells [151]. Vaccines based on the
opment of a killed (inactivated) vaccine. Inacti- Leningrad-3 strain have been used in the former
vated vaccines have so far failed, this is probably Soviet Union and other countries since 1974.
the result of a lack of neutralizing antibodies to
the viral fusion surface protein that is destroyed Production and Characterization of
in the inactivation process [144]. Subunit vac- Mumps Vaccines. The WHO requirements for
cines based on the hemagglutinin and the fusion the above-mentioned live mumps vaccine have
proteins of measles virus have successfully been been formulated [152, pp. 139 – 164]. Controls,
tested in animals and may provide a vaccine es- thermostability tests, and requirements are as
pecially suited for developing countries [145]. 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
3.2. Mumps Vaccine requirement of potency has been established
by the WHO but a commonly used minimum
Etiological Agent and Pathogenesis. dosage is 5000 TCID50 per human dose. The
Mumps is caused by a member of the paramyx- vaccine is lyophilized, reconstituted vaccine
ovirus group and is also known as (epidemic) should be used without delay.
parotitis due to its predominant symptom–
infection of the salivary (parotid) gland. Trans- Immunization Recommendations.
mission occurs via infected aerosol droplets. Mumps vaccine is recommended to be given
The disease is generally characterized by mod- to all susceptible individuals over the age of
erate fever and swelling of the salivary glands. 12 months [153]. It is usually given in com-
Subclinical infection occurs in ca. one third of bination with measles and rubella vaccines in
industrialized countries (see Section 3.4).
Immunotherapy and Vaccines 27
Adverse Reactions. The RA 27/3 strain has months. Two doses are given in some Euro-
few and mild side effects. The joint manifesta- pean countries [162]. In Sweden, a first dose is
tions, mainly arthralgia, are more common in administered at 18 months and a second at 12
adults. Intrauterine infections during pregnancy years of age, replacing the monovalent programs
have been documented for all three strains (Cen- for measles and rubella immunization respec-
dehill, HPV-77, and RA 27/3) [163]. No abnor- tively. In the United States, the combination of
malities related to congenital rubella infections MMR with oral polio vaccine and diphteria –
have been documented for any of the strains. tetanus – pertussis (DTP) at 15 months has been
recommended [164], [165].
Vaccine Efficacy. The efficacy of the current
strain is ca. 90 %. The incidence of congeni- Adverse Reactions. The side effects are the
tal rubella virus infection decreased in countries same as for the monovalent vaccines, the
with a routine immunization program against the measles component being the most reactogenic.
disease. The duration of immunity has not yet The rate of adverse reactions is <0.5 – 4 % [166].
been determined. The risks to the fetus upon a
maternal reinfection are also unknown and may Vaccine Efficacy and Future Prospects.
be very small. The efficacy of the combined preparation is the
same as for the monovalent vaccines with an
Future Prospects. The current vaccine is overall efficacy of 90 – 95 % [143]. The life-long
considered safe and efficient and no efforts are protective efficacy of a single injection has not
devoted to further development. been proven. Further combinations with an at-
tenuated varicella component were investigated
[167].
3.4. Combined Measles – Mumps –
Rubella Vaccine
3.5. Polio Vaccine
History of Immunization. The first com- Etiological Agent and Pathogenesis. Polio
bined measles – mumps – rubella (MMR) vac- (infantile paralysis) is caused by a picornavirus
cine was licensed in the US in 1971. It con- of the genus enteroviruses. A poliovirus strain
tains the Moraten measles strain (Section 3.1), was first isolated in cell culture by Enders,
the Jeryl Lynn mumps strain (Section 3.2), both Weller, and Robbins in 1949. In 1951, polio
grown in chick embryo cultures, and the RA virus isolates were officially grouped into three
27/3 rubella strain (Section 3.3), grown in hu- serotypes, type 1 (Brunhilde), type 2 (Lansing),
man diploid cells. Three other combined vac- and type 3 (Leon). Transmission is mainly via
cines were licensed later. The triple vaccine has the oral – fecal route and the virus reaches the
replaced the monovalent vaccines used in gen- central nervous system by way of the blood
eral immunization programs for children in the stream. The vast majority of infections are sub-
United States and in many European countries. clinical. The nonparalytic disease has a mild or
minor form with fever and general malaise and
Production and Properties of Combined a more severe or major form with additional
Vaccines. The minimum potencies for the at- symptoms of meningitis/meningoencephalitis.
tenuated strains of measles, mumps, and rubella Paralytic polio, with its most severe bulbar form,
virus, are the same as in the monovalent vac- is estimated to represent 5 – 10 % of clinical
cines, i.e., 1000 TCID50 for measles, >5000 cases. The mortality rate is 5 – 10 % of clini-
TCID50 for mumps, and >1000 TCID50 for cal cases, i.e., 1 – 2 % of all infections. Severe
rubella. The vaccine is lyophilized and should paralysis is seen in 10 – 20 % of clinical cases,
be stored at 2 – 8 ◦ C. and mild or moderate paralysis in 30 %.
With improved sanitation and standards of
Immunization Recommendations. In living, a shift towards infection at a higher
many countries, one dose of vaccine is ad- age was observed in industrialized countries in
ministered to children at the age of 15 – 24 the prevaccination era. During the first half of
Immunotherapy and Vaccines 29
this century, peak incidence was noted in the Saukett strain. Primary and secondary monkey
age group 5 – 14 years. A large proportion of (Cynomolgus) kidney cultures or human diploid
cases occurred in young adults. In the develop- cells are most commonly used for culture. The
ing countries, polio has maintained its character WHO requirements for both OPV and IPV pro-
of infantile paralysis with the majority of chil- duced in primary monkey kidney cells include
dren being infected during the first few years of tests
life. Cases of paralytic polio occur only in the
1) for the absence of cytopathogenic viruses
youngest age groups.
with the exception of some foamy
History of Immunization. The first large- viruses[141, pp. 40 – 84] , [159, pp. 157 –
scale field trial of an inactivated (killed) polio 173], [176],[182, pp. 108 – 110]
vaccine was launched in the US in the early 2) the absence of simian B virus in rabbits, and
1950s, only a few years after successful prop- 3) the absence of SV 40 in sensitive cells.
agation of the virus in tissue culture. Inactivated (usually primary green monkey Cercopite-
polio vaccine (IPV) was licensed for general use cus kidney cultures). Minimal potency require-
in the United States in 1955 [168], [169]. In ments for OPV were formulated by the WHO in
1955, cases of atypical paralytic polio were re- 1987[152, pp. 165 – 166]. A single human dose
ported, most of them were associated with two of trivalent oral vaccine should contain approx-
lots of vaccine from Cutter [170]. Although imately 105.5 – 106.5 infectious units of type 1,
live poliovirus was recovered from vaccine sup- 104.5 – 105.5 of type 2, and 105.0 – 106.0 of type
plied by other manufacturers, this failure of in- 3. For IPV, recommendations of 40 : 8 : 32 D-
activation is known as the Cutter incident. The antigen units per human dose for types 1, 2, and
total toll was 269 cases, of which 192 were par- 3 respectively were issued in 1981 [141]. The
alytic (with ten deaths). Clinical trials with a potency tests for OPV (and IPV prior to inactiva-
live attenuated oral polio vaccine (OPV) were tion) are performed by titration in tissue culture.
started in 1958 and the vaccine was licensed in In vivo potency tests are also required for IPV
the United States in 1962 [171], [172]. Immu- but neither the animal species nor the number of
nization with OPV alone has been used from the injections is specified. Inactivation of polio vac-
late 1950s – early 1960s in Europe. One of the cine by formaldehyde is controlled by titration
largest immunization campaigns was launched of polio-sensitive cells (usually Cynomolgus or
in the former Soviet Union in 1960 when OPV Cercopitecus) in tissue culture. Oral polio vac-
was given to ca. 77×106 people. A few Euro- cines are supplemented with a stabilizer, usually
pean countries have used only IPV [173–175]. magnesium chloride or sorbitol. Both OPV and
In Sweden, clinical trials with IPV were started IPV contain antibiotics, usually neomycin; OPV
in 1955 and general immunization was intro- is best stored at −20 ◦ C but can be stored at 2 –
duced in 1957 [173]. All three strategies were 4 ◦ C for a variable length of time; IPV is stored
effective in reducing the incidence of paralytic at 2 – 8 ◦ C.
polio. With IPV alone, an 80 % reduction rate
was achieved in the US with less than 50 % of Immunization Recommendations. Rec-
the population immunized [168], [169], [171]. ommendations vary but at least three doses of
The decrease of polio continued at the same rate OPV are given. In the US, five doses are rec-
after introduction of OPV. Introduction of the ommended, in Sweden four, and in Finland six.
massive IPV campaign in Sweden resulted in a The WHO recommends four doses for infants
93 % decrease of paralytic disease after six years in developing countries.
of immunization and elimination of the disease
by 1964 [173]. Adverse Reactions. Serious adverse reac-
tions of paralytic polio have only been reported
Production and Properties of Polio Vac- for OPV with a predominance for type 3 [169],
cines. Oral polio vaccines are manufactured [177] (incidence = 1/1×106 vaccine recipients,
with the attenuated Sabin strains. Inactivated po- including both nonimmune and immune indi-
lio vaccines are mostly produced with the type 1 viduals). No cases of IPV-induced paralytic po-
Mahoney strain, type 2 MEF-1 strain, and type 3 lio have been reported after 30 years of use (with
30 Immunotherapy and Vaccines
the exception of the Cutter incident). IPV is con- with infectivity; they are found in the 42-nm
sidered to be the least reactogenic of all the vac- Dane particle, i.e., the infectious virion.
cines used for childhood immunization. The HBV is a pathogen only for humans, but
certain monkeys, in particular chimpanzees, are
Vaccine Efficacy. Both OPV and IPV have also susceptible to infection. Transmission in hu-
proven highly efficient in eliminating polio from mans occurs mainly by inoculation with infected
industrialized countries with general immuniza- blood. The virus is excreted in body fluids, caus-
tion programs. The protective efficacy is close to ing infection through saliva and sexual contact.
100 % after at least three doses of vaccine with A chronic carrier state is established in 5 – 10 %
adequate immunogenicity. Serologic data re- of adult cases; chronic active hepatitis develop in
ported for OPV from developing countries have 25 – 30 % of the carriers, often leading to cirrho-
been less than encouraging. Both vaccines have sis of the liver: 90 % of infants infected at birth
eliminated the circulation of wild type strains in become carriers. Hepatitis B occurs throughout
industrialized countries with high immunization the world; the incidence varies from <0.5 % in
rates. Outbreaks among groups refusing immu- Western Europe and the US to 5 – 15 % in South-
nization have occurred [168]. In 1984, nine cases east Asia and Southern Africa [183], [184].
of paralytic polio caused by type 3 occurred Hepatitis B virus plays an important role
in Finland [177]. The low immunogenicity of in hepatocellular carcinoma although the exact
the IPV used in Finland has been known since mechanism has not yet been determined [183].
the late 1960s [178]. The new IPV developed
in the Netherlands with its high antigen content History of Immunization. Heat-inactivated
is given in European countries in at least three HBsAg-positive serum was used by Krugman
doses [179]. Two doses or even one dose of the in 1971 to immunize children. A 70 % protec-
new IPV preparations have been claimed to be tion rate was found upon challenge with active
highly protective. A two dose-schedule has been virus. The high protective efficacy of hepatitis B
used in Africa and protective efficiency was ca. vaccine derived from purified, inactivated HB-
89 % [180]. Inactivated polio vaccine is more sAg from positive plasma was demonstrated in
heat-stable and can be combined with diphte- 1980 [185]. The vaccine, developed in the US,
ria – tetanus – pertussis vaccines. was licensed for general use in 1981. Studies
in 1980 – 1983 also showed that the HBV vac-
Future Prospects. Further development of cine in combination with human antiHBV im-
OPV is aimed at reducing production costs by mune globulin (HBIG) prevented development
use of microcarrier cultures (cell cultures on of the carrier state in infants born to carrier moth-
beads etc.) in fermentors [179]. Use of contin- ers. Hepatitis B vaccine was the first vaccine
uous cell lines is being investigated. An IPV to be routinely produced by DNA recombinant
produced in a continuous cell line is currently technology in yeast (→ Genetic Engineering,
licensed in France [181]. The WHO require- Chap. 5.2.). Serologic studies indicate that the
ments for production of vaccines in continuous protective efficacy of the recombinant vaccines
cell lines are formulated in [182, pp. 93 – 107]. is the same as that of the plasma-derived vaccine.
Genetically engineered vaccines were licensed
in the US in 1986 and subsequently in several
3.6. Hepatitis B Vaccine European countries.
by urea (8 mol/L), and formaldehyde or by heat Vaccine Efficacy. The overall vaccine effi-
with or without formaldehyde. After controls for cacy is 90 % with no response upon immuniza-
purity and antigen content, an alum adjuvant is tion in 5 – 10 % of the vaccines. In individuals
added; the vaccine must be stored at 5 ± 3 ◦ C. with seroconversion to HBsAg, protective effi-
No minimal requirement of antigen content has cacy is almost 100 %. Studies on the duration
been formulated but a commonly used vaccine of protection beyond 5 – 10 years are not yet
contains 20 µg HBsAg per human dose. available. The protective efficacy of HBIG and
The WHO requirements for HBV made by vaccine administered to newborns born to car-
recombinant DNA technology state that such rier mothers has been estimated to be more than
vaccines may contain the S gene products or 90 % [183].
the S/pre-S combination [152, pp. 106 – 138]; li-
censed recombinant vaccines contain only the S Future Prospects. Current HBV vaccines
gene product. A full description of the host cell are safe and have a high efficacy but also high
(Saccharomyces cerevisiae) and the expression production costs. Research is aimed at the de-
vector is required. velopment of cheaper, more immunogenic vac-
The HBsAg is commonly purified by precipi- cines. The inclusion of the pre-S (P 31) re-
tation, ultrafiltration, and chromatography. Tests gion of the viral genome in addition to the
for HBsAg and residual cell or plasmid DNA present S (P 25) region is being investigated
are required before addition of adjuvant. No in yeast-derived recombinant vaccines [190],
minimal antigen requirements have been formu- [191]. Other antigens such as HBcAg are also
lated but a potency assay in mice has been out- being studied. Polypeptide vaccines, based on
lined. Commonly used recombinant yeast vac- the major determinants of HBsAg, are undergo-
cines contain 10 or 20 µg HBsAg per human ing clinical trials. Synthetic peptides may pro-
dose. vide the ultimate solution to the problem of im-
munogenic, readily available HBV vaccines.
Immunization Recommendations. Most
Western European countries and the US have
issued recommendations for vaccination of risk 3.7. Rabies Vaccine
groups [187–189]. In general, pre-exposure pro-
phylaxis is recommended for medical and other Etiological Agent and Pathogenesis. Ra-
staff in frequent contact with high-risk groups bies is a lethal disease caused by a neu-
or with blood from such groups. Immunization tropic rhabdovirus affecting humans and warm-
of patient groups frequently receiving blood or blooded animals. The most common route of
blood products, as well as patients in certain transmission to humans is by infected saliva
institutions is also recommended. through the bite from a rabid animal. The virus
For pre-exposure prophylaxis, three 20 µg ascends along peripheral nerves to the central
doses of plasma-derived vaccine is given at day nervous system. The incubation period varies
0, 30 and 6 months intramuscularly in the del- from ten days to several months. Symptoms in-
toid region. Children <11 years of age should clude a nonspecific prodromal stage followed by
receive 5 µg at each injection. Postexposure pro- an acute neurological phase ending in coma and
phylaxis for infants to HBsAg, HBeAg-positive death. The disease is endemic in animals in most
mothers is 0.5 mL intramuscular injection of hu- parts of the world with some exceptions such as
man antiHBV immune globulin (HBIG, see also the United Kingdom and parts of Scandinavia
Section 5.3). Vaccine (10 µg) should be given at [192]. In the urban form, stray dogs and cats act
birth and then at one and six months of age. For as vectors, whereas the sylvatic form involves
postexposure prophylaxis of adults, 1 mL HBIG wild animals. Urban rabies is well-controlled
should be given immediately together with three in most countries, but progressive spread of the
doses of vaccine administrated as for preexpo- sylvatic form by foxes is a major problem in
sure prophylaxis. Europe. In the US, other animal species are in-
volved including bats. In Europe and Africa,
Adverse Reactions. Side effects are mild bats are the vector for an antigenic variant of the
and mainly local at the site of injection. virus [193].
32 Immunotherapy and Vaccines
Adverse Reactions. The incidence of sys- during the first trimester of pregnancy. Neonatal
temic (febrile) and local reactions is <10 % and varicella with a 30 % mortality rate occurs with
<20 % for the whole virus vaccines and split maternal varicella within one week prior to term
vaccines, respectively. An increased rate of the without prophylaxis. Varicella in the compro-
neurological reaction of Guillain – Barre’s syn- mised host is a severe disease with a 10 – 20 %
drome was reported in the United States af- mortality rate. Herpes zoster, a localized, often
ter large-scale immunization against swine in- painful vesicular rash is most common in the
fluenza in 1976. elderly; it occurs in 10 % of the population.
Fever and rash occurred in 40 % of vaccine re- tissue was used for mass immunization in Sene-
cipients with chemotherapy suspended for two gal and the chick embryo vaccine was tested in
weeks. Brazil. The vaccines were administered by scar-
ification using normal human serum as stabi-
Vaccine Efficacy. Protective efficacy in lizer. The chick embryo cultured vaccine (first
healthy children is 95 – 100 % with persistence the 17 E vaccine and later the 17 D vaccine)
of immunity over a 5 – 10 year period. In adults, has been most widely used. Numerous prob-
protective efficacy is 60 – 80 % and possibly of lems and accidents were associated with both
shorter duration. In children with malignancies, vaccines. The human serum used as stabilizer
protective efficacy is 60 – 92 %. Spread from caused hepatitis affecting many vaccinees in the
vaccine-induced vesicular rash has been docu- armed forces during World War II. Systemic re-
mented in household contacts. actions were common. Severe postvaccination
encephalitis with a high mortality rate occurred
Future Prospects. The currently investi- mainly with the neural tissue cultured vaccine.
gated, attenuated Oka strain vaccine is effective This reaction (mostly in children < one year
but long-term immunity and zoster incidence re- of age) was also reported with some lots of the
main to be established. The problems of latency 17 D vaccine with increased neurotropism; fur-
and of vaccine-induced rash could be overcome thermore, loss of protective efficacy was noted
by the development of a subunit vaccine pro- in tropical climates due to low thermostability.
duced either by a recombinant DNA technique The 17 D vaccine is the result of developments
or by using synthetic peptides. aimed at careful definition of the properties of
the seed virus and at improved thermostability
[215].
3.10. Yellow Fever Vaccine
Production and Properties of Yellow Fever
Etiological Agent and Pathogenesis. Yel- Vaccines. Only certain institutes are approved
low fever is a hemorrhagic fever caused by a by the WHO for production of yellow fever
member of the toga virus group [212], [213]. vaccine[159, pp. 34 – 53]. The seed lot virus,
The virus was transmitted to rhesus monkeys usually a substrain of 17 D-204, has to be
by Mathis and coworkers in 1927. The virus shown to be free from neurotropism by test-
strain was then propagated by serial passages ing in monkeys[159, pp. 34 – 53],[182, pp. 113 –
of intracerebral inoculations in white mice. Yel- 141],[216] . Most producers use seed lots that are
low fever is transmitted by mosquitoes of the free of leukosis virus for production but their
genus Haemagogus in South America and of use is not mandatory. Virus-infected embryos
the genus Aedes in Africa. The animal reser- are harvested, homogenized, and the supernatant
voir is mainly monkeys. Subclinical infections is used as vaccine. Several tests for adventitious
are common. The incubation period in clini- agents are performed. Virus titrations are per-
cal cases is 3 – 6 days; symptoms range from formed in a mouse assay with a minimal potency
transient fever and headache to high fever with requirement of 1000 LD50 per human dose. The
meningoencephalitis, followed by jaundice and vaccines are lyophilized in the presence of stabi-
hemorrhagic manifestations. In the malignant lizer. Most current vaccines retain the minimal
form all these symptoms are present and death requirement for two weeks at 22 ◦ C. A vaccine
occurs within one week. Mortality rates are stable for two weeks at 37 ◦ C is requested by
40 – 50 % in the severe forms of the disease. WHO for use in tropical areas.
younger children but also elderly people with protective efficacy is estimated to be 90 – 95 %
mortality rates of 50 %. Neurological sequelae from serologic studies.
have been reported in 30 – 40 % of survivors of
the severe clinical forms.
3.13. Smallpox Vaccine
History of Immunization. A formaldehyde-
killed vaccine, consisting of a 5 % suspension Smallpox (variola) was caused by a member of
of infected mouse brain tissue, was introduced the pox virus group, which also includes the
for human use in Japan in 1954 [219]. The vac- vaccinia virus used for immunization. The dis-
cine (Nakayama strain) has been purified by pro- ease was one of the most devastating infections
tamine sulfate precipitation since the late 1950s in human history. Eradication of smallpox is
and by absorption with charcoal or kaolin since the success story of immunization. In 1967 the
the early 1960s. A protective efficacy of 80 – WHO launched a massive eradication program–
90 % for the vaccine was shown. The highly pu- smallpox was still reported from 42 countries. In
rified, mouse brain cultured vaccine produced May 1980, WHO officially declared the world
in Japan is also used for immunizing travellers free from smallpox. No proven cases of variola
to endemic areas. In China, a vaccine produced have occurred in the past decade.
in primary hamster kidney cells has been exten- Two types of vaccine were manufactured, calf
sively used since the late 1950s. lymph vaccine and egg vaccine [220]. Both liq-
uid and lyophilized forms were used. Require-
Production and Properties of Japanese ments for manufacturing, control, and potency
Encephalitis Vaccines. Several Japanese were formulated by the WHO. Immunization by
manufacturers produce the purified vaccine from multiple puncture with a bifurcated needle was
culture of mouse neural tissue by similar meth- most commonly used. Severe adverse reactions
ods. No WHO requirements have been formu- included postvaccination encephalitis and dis-
lated. A vaccine used for immunizing travellers seminated vaccinia.
to endemic areas is prepared by infecting mouse General immunization against smallpox was
brain with the Nakayama strain. The brain ho- withdrawn in most European countries in the
mogenate is purified by protamine sulfate treat- mid 1970s. Requirement for smallpox vaccina-
ment and then inactivated with formaldehyde. tion was abandoned for international travel in
Further purification involves ultracentrifugation 1982.
on a sucrose density gradient. The vaccine is Recommendations for civilian immunization
lyophilized; the reconstituted vaccine must be in the United States include only laboratory
used immediately. workers handling variola virus or other closely
related orthopox viruses. Military personnel in
Immunization Recommendations. Pri- the United States and the former Soviet Union
mary immunization consists of subcutaneous are routinely vaccinated against smallpox [221].
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 3.14. Rift Valley Fever Vaccine
every 1 – 3 years. Extensive immunization in en-
demic areas has been considered by the WHO. Rift Valley fever is an arthropod-borne disease
Children less than three years of age should re- known only in Africa [212], [213]. It is caused
ceive 0.5 mL doses. Immunization is generally by a member of the Bunya virus group. The ma-
recommended for health care workers and other jor vectors are mosquitoes (Culex theileri and
people with extended stay in endemic areas. Aedes cabbalus). The main natural hosts are
sheep, cattle, and goats. The incubation period
Adverse Reactions. Only a few mild, local
is 2 – 6 days. Rift Valley fever is a febrile dis-
and systemic reactions have been reported.
ease with headache and abdominal pain lasting
less than one week. Hemorrhagic fever with liver
Protective Efficacy. Immunization has dras-
tically reduced disease incidence in Japan. The
38 Immunotherapy and Vaccines
necrosis and encephalitis are the severe mani- Another vaccine against hookworms in dogs,
festations causing mortality and sequelae. Out- although immunologically efficient, has failed
breaks occurred in Africa during the 1970s, with commercially.
the largest outbreak in Egypt in 1977 – 1978. Important in the development of vaccines is
An inactivated (killed) vaccine (NDBR-103) the identification, isolation, and testing of puta-
was produced by the US army in 1967 [222]. The tive protective antigens. Successful vaccination
Entebbe strain of the virus was grown in primary against helminth infections requires the prim-
monkey kidney cells, inactivated with formalde- ing of those responses which may subsequently
hyde, and lyophilized. The adverse reactions are be triggered during natural infection. Since anti-
few and mild. One case of Guillain – Barré oc- gen presentation plays a central role in the ac-
curred in a Swedish military vaccinee [223]. Se- quired immune response, the development of
roconversion rates after subcutaneous injection accessory cells and the activation of T and B
of three 1 mL doses given at 1 – 2 week inter- lymphocytes has to be taken into account [224].
vals were over 95 %. A newer vaccine (GSD- Developments have focussed on vaccines pro-
200) is based on a cloned version of the orig- duced by recombinant DNA techniques. In vitro
inal seed virus (Entebbe strain) and grown in cultivation methods have been used for produc-
diploid rhesus monkey cells. The WHO has for- ing helminth antigens for immunoprophylaxis
mulated requirements for inactivated Rift Val- and combined with recombinant DNA technol-
ley fever vaccines produced in primary monkey ogy [225].
kidney cells and in human or non-human pri- The helminthoses described in this chapter
mate diploid cells[141, pp. 104 – 143]. No mini- were chosen on the basis of the parasites’ im-
mal potency requirements have been formulated. portance to human health or because of their in-
teresting biology.
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 large variety of nonlymphoid effector cells oc-
of polypeptides for vaccination trials. cur within the intestine, including natural killer
One of the most promising candidates is cells, macrophages, neutrophils, eosinophils and
a schistosomula surface polypeptide with a basophils; their numbers increase during para-
molecular mass of 28 000. The gene coding site infections.
for this polypeptide has been cloned [242] and No vaccine against gastrointestinal nema-
expressed in E. coli. Immunization with this todes is presently available for human applica-
recombinant antigen induced a high level of tion. However, promising, successful trials in
serum cytotoxicity towards schistosomulas in veterinary parasitology have initiated interesting
the rat, hamster, and monkey. Significant pro- work in human gastrointestinal nematode infec-
tection against a natural challenge infection with tions and will probably result in vaccine supply
live cercariae was obtained in rats and hamsters in the near future.
[242]. These results can be viewed with opti-
mism for the development of a schistosomiasis Hookworm Disease. Various species of the
vaccine. family Ancylostomatidae are responsible for
hookworm disease in humans, the most impor-
tant being Ancylostoma duodenale and Necator
4.1.2. Vaccines against Nematodes americanus. Over a fifth of the world’s popula-
tion is afflicted by this disease, mainly in trop-
4.1.2.1. Gastrointestinal Nematodes ical and subtropical regions. Adult hookworms
have a length of about 0.7 – 1.8 cm and a hook-
The most prevalent and pathogenic gastroin- like anterior end with a distinct mouth area. Fe-
testinal nematode parasites of humans belong to males release eggs in the small intestine which
the genera Ascaris, Strongyloides, Trichinella, are excreted in the feces of the host. In humid
Trichuris, Ancylostoma, and Necator. Although surroundings first-stage larvae hatch from the
many species parasitize deeper tissues of the eggs and develop into infective larvae, which
body, the majority are intestinal. The intestine penetrate through the skin into new hosts. After
has been maintained as a site for adult stage de- migration through lymph or blood vessels, the
velopment, whereas larval stages often invade larvae finally develop into mature adult worms
other host tissues. The intestine consists of a se- in the small intestine, completing the parasite’s
ries of distinct parasite habitats (gut sections, life cycle. Hookworms suck blood from mi-
lumen, mucosa, etc.) each having its own char- crolesions in the host’s small intestine, thereby
acteristics. Large worms such as Ascaris must causing chronic gastrointestinal blood loss, ane-
live within the lumen, smaller species such as mia, and hypoalbuminemia. With a large worm
hookworms are associated with the mucosa. For burden death may occur. Patients usually suffer
a long time research on immunity against in- from extreme weakness, pallor, secondary res-
testinal worms was given low priority due to piratory tract infections, skin irritations, heart
the lack of knowledge about local immune re- palpitations, and gastrointestinal distress.
sponses in the gut. Studies have shown that in- Immunity and Vaccine Design. Nematode
testinal worms are indeed subject to protective parasites present special problems for the host’s
immune responses, although these responses protective immune response, because they pos-
differ somewhat from classical immunity in the sess a tough, protective, external cuticle. The
body because the worms live in the gut lumen. cuticle is both antigenic and immunogenic, but
Special features are (1) macromolecular antigen it is doubtful whether responses directed against
uptake across the intact mucosal epithelium or its surface play a major role in immunity against
by specialized epithelial cells overlying Peyer’s intestinal species. Protective responses are more
patches and (2) complexation of the antigen by likely initiated by antigens released through the
dimeric IgA secreted from the mucosa or in- orifices of the parasite [243]. One such antigen
testinal IgG Fab fragments. Cells from the un- with a potential vaccine function is a secreted
derlying lamina propria participate in cytotoxic- proteolytic enzyme [244]. Hookworms attached
ity and hypersensitivity reactions and thus affect to the mucosa secrete the enzyme from glands
mucosal structure and function. Furthermore, a in their mouths. The enzyme degrades the host
Immunotherapy and Vaccines 41
proteins and inhibits blood coagulation which persons. Its early symptoms are fever, lymphan-
permits the hookworms to feed for an indefi- gitis, and lymphadenitis. A following chronic
nite period of time. Immunizing the host against stage is frequently characterized by more se-
this enzyme would result in inhibition of any rious clinical manifestation including elephan-
enzyme secreted by worms after a subsequent tiasis, hydrocele, and pulmonary eosinophilia.
challenge infection. The worms would then be Adult worms are found in the lymphatic system,
unable to feed. The gene coding for the enzyme microfilariae may be found in blood.
in question has been cloned. Future investiga- Filiarial infections caused by Onchocerca
tions will have to demonstrate the applicability volvulus occur in Africa and South and Central
of this kind of vaccine. America, affects millions of patients, many of
A vaccine has been developed for the control whom become blind. Adult worms are generally
of hookworm infections in dogs [245]. It was located under the skin, forming typical nodules;
based upon irradiation-attenuated infective lar- less often they penetrate deeply into the tissues.
vae and was administered parenterally. Although Pathology in onchocerciasis is due entirely to
the vaccine was highly effective in preventing the microfilarial stage. Microfilariae are present
hookworm disease, it was withdrawn because it in the skin and may penetrate into the eye, thus
did not completely prevent infection, and effec- leading to severe eye lesions and blindness.
tive anthelminthic chemotherapy was available. Loa loa is prevalent in the forest areas of
This example suggests that protective immunity West Africa. Adult worms penetrate into the tis-
may also occur in human hookworm infections. sues, provoking transient edema. Severe clin-
ical symptoms are rare, loiasis being gen-
erally regarded as a benign infection [246].
4.1.2.2. Tissue-Invading Nematodes Dipetalonema perstans, D. streptocerca, and
(Filariidae) Mansonella ozzardi infect humans (Africa,
South America) but usually asymptomatically
Many nematode species which live as adults and rarely cause significant diseases [247].
in the intestine, (e.g., Ascaris, hookworms, and Many of the changes associated with filarial
Trichinella) undergo development in parenteral infection are immunopathological in origin and
tissues. Other species are wholly confined to hypersensitivity reactions are important in their
these tissues and have no contact with the in- development.
testine (tissue-invading nematodes). This closed Immunity and Vaccine Design. Although fi-
habitation site requires special conditions in or- larial parasites provoke a strong immune re-
der to obtain biological contact with the outside sponse in the human host, the chronicity of these
world, especially for reproduction. In the major infections implies the absence of a protective re-
group of tissue-invading nematodes, the Filari- sponse or the evasion of such responses by the
idae, the worms overcome this problem by using worms. There is also no direct evidence that fi-
bloodfeeding arthropods as intermediate hosts. larial infections in nature confer resistance to
The female worms release microfilariae larvae reinfection with the same parasite species.
which circulate in the blood or accumulate in No vaccines against filarial parasites are
the skin of the host. The arthropod feeds on the available. The development of new strategies for
blood and takes up the microfilariae which de- immunological control depends on a thorough
velop into infective larvae. At a following blood understanding of immunological host – parasite
meal, infective larvae are reinoculated into new relationship. Many studies on protective im-
human hosts. munity in animals have concentrated upon re-
sponses directed against larval stages. A vaccine
Filariasis. The human disease filariasis com- against microfilariae would inhibit transmission
prises an extremely heterogeneous group of of the disease, a vaccine against infective larvae
diseases. The main filiarial parasite species would provide protective immunity against pri-
in humans are Wuchereria bancrofti and Bru- mary infection of hosts. Inoculation with infec-
gia malayi, which are both transmitted by tive larvae attenuated by irradiation [248], [249],
mosquitoes. The disease is widespread in tropi- only confers partial protection to a subsequent
cal and subtropical regions affecting ca. 90×106 challenge infection. Filarial vaccines based on
42 Immunotherapy and Vaccines
irradiated larvae cannot be used in humans with- tive hosts. The diseases caused by the metaces-
out first determining whether these attenuated todes are referred to as cystic echinococco-
larvae induce pathological changes [250]. sis for E. granulosus and alveolar echinococ-
The target antigens of antimicrofilarial im- cosis for E. multilocularis. Echinococcus gran-
mune responses are probably located on the sur- ulosus is prevalent throughout the world and
face of the microfilariae; they are currently being is a public health and economic problem in
characterized [251]. Antibodies to surface anti- many areas. Echinococcus multilocularis only
gens mediate adherence of host cells to the mi- exists in the northern hemisphere and is rela-
crofilariae; this can result in worm killing [243]. tively frequently seen in the former Soviet Union
Using Dipetalonema vitae as a model, IgE was (Siberia), central Europe, northern China, Japan,
found to be the primary immunoglobulin in- and Alaska. The fully developed metacestode
volved in cell adherence to the worm cuticle. of E. granulosus is a typically unilocular, fluid-
The first cell type to adhere is the eosinophil. filled cyst, which is located in the liver, lungs,
Subsequent, adherence of macrophages is fol- and other organs. Echinococcus multilocularis
lowed by release of lysosomal enzymes which metacestode conforms a vesiculated parasitic
degrade the cuticle. In humans, a comparatively mass in the liver of the host, it proliferates by
long time is needed to establish such immunity. continuous exogenous budding with possible
This may be due to pronounced immunosup- metastasis formation in other organs.
pression or because the microfilariae cover their The causative agent of cysticercosis is the
surface with host components and thus render tapeworm Taenia solium. Humans are the oblig-
their antigenic surface epitopes less accessible atory definitive hosts, pigs act as intermediate
to the host’s immune system. The search for fi- hosts. Metacestode stage infection can also oc-
larial antigens that can safely and successfully cur in humans and may result in infection of the
be used in humans is still continuing; it is still central nervous system by parasite larvae (neu-
not known how restricted series of antigens can rocysticercosis). Morbidity includes intracranial
be used to protect natural hosts against first or hypertension, basal arachnoiditis, focal neuro-
persistent infections with a complex, adaptable, logical deficits, and dementia. Hyperendemic ar-
genetically diverse parasite. Recombinant DNA eas are found in Latin America, Africa, and Asia;
technology and hybridoma technology may pro- areas of lower endemicity occur in southern and
vide potential candidates for successful filarial eastern Europe.
vaccines.
Immunity and Vaccine Design. Host pro-
4.1.3. Vaccines against Cestodes tective immunity is a striking feature of re-
peated infection with cestodes in mammalian
Etiologic Agents, Pathogenesis, and Epi- intermediate hosts [225]. It plays a major role
demiology. This section deals with the hy- in regulating natural transmission of these para-
datidosis and cysticercosis disease complexes, sites, and substantial research efforts have been
which are caused by the larval stages of tape- undertaken towards development of vaccines
worms belonging to the family Taeniidae. The against cestodes of veterinary importance [252].
most important causative agents of hydatido- For human cestode infections, it is debatable
sis (echinococcosis) are Echinococcus granu- whether there is sufficient importance to war-
losus and E. multilocularis, whose life cycles rant the research required to develop a vaccine.
involve a definitive and an intermediate mam- In areas of high egg contamination (e.g., the
malian host. The definitive hosts are carni- Turkana district (Kenya) for E. granulosus, St
vores (mainly dogs for E. granulosus and foxes Lawrence Island for E. multilocularis, or Mex-
for E. multilocularis) in whose intestines the ico for T. solium), vaccination should be given to
adult stage worms occur. Intermediate hosts are very young persons, as patients usually become
herbivorous and omnivorous species in which infected at very young age. An ideal human vac-
the larvae (metacestodes) develop. Humans and cine requires complete, long-lasting protection.
other intermediate hosts become infected by Experiments in veterinary parasitology, how-
ingesting eggs passed in the feces of defini- ever, demonstrated exactly opposite results. Vac-
Immunotherapy and Vaccines 43
cination of animals against Taenia species re- 4.2.1. Strategy for Malaria Vaccine
sulted only in marked reduction of cyst numbers Development
which persisted only for a maximum of one year.
In addition, strong adjuvants had to be employed More than 100 species of malarial plasmodia
which are not tolerated by humans. Other cri- are known, but only four infect humans: Plas-
teria may be important; a review of protective modium falciparum, which is responsible for the
immune mechanisms is given in [253]. majority of human deaths, P. vivax, P. malariae,
Little attention has been paid to the vacci- and P. ovale.
nation of definitive hosts (dogs and foxes for The life cycle of the plasmodia is complex
Echinococcus, and humans for T. solium). Ex- (Fig. 4). The female anopheline mosquitoes in-
periments [254], [255] demonstrated an im- oculate sporozoites into the blood of the verte-
mune response after infection as well as highly brate host. Within minutes the sporozoites in-
significant suppression of egg production by vade the liver parenchymal cells (hepatocytes)
E. granulosus after immunization of dogs with where they divide asexually and develop into
secretory antigens derived from adult tape- merozoites which rupture the hepatocytes and
worms. This approach seems to be the most reenter the blood. In the subsequent erythrocytic
likely control measure for reducing infection cycle, the merozoites invade the red blood cells
risk in humans. 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
4.2. Malaria Vaccine clinical manifestations of malaria. Some mero-
zoites differentiate into sexual stages called ga-
Malaria remains a major health problem in many metocytes which are ingested by the mosquito.
tropical and subtropical countries and affects Fertilization of the gametes occurs solely in
hundreds of million of people each year. A ma- the midgut of mosquito. The resulting zygotes
jor effort was made to control malaria from develop into ookinetes and then into oocysts.
1950 – 1970 with insecticides (→ Insect Con- Sporozoites are released from mature oocysts
trol) and antimalarial drugs (→ Chemotherapeu- and migrate to the mosquito salivary glands. The
tics, Chap. 3.3.). The initial remarkable results cycle is then repeated.
have been difficult to maintain. The advent of This complex life cycle involves continuous
drug-resistant parasite strains and of insecticide- morphologic, enzymatic, and antigenic changes
resistant mosquito vectors are major obstacles which are linked to the parasite’s environmen-
in the effort to control malaria. Since the early tal adaptation to the host. The invasive stages
1970s new approaches have been explored such of the parasite (sporozoites, merozoites, ga-
as vector control through biological agents and metes) have unique, stage-specific surface de-
control of malaria infection through vaccines. terminants. Furthermore, immunologic crossre-
The development of malaria vaccine has re- activity exists between plasmodia species and
ceived considerable impetus: first, because im- between the developmental stages of a given
munization of animals with whole parasites can species. However, immunization experiments in
induce a degree of protection equal or superior animal models have demonstrated that the anti-
to that induced following natural infection [256– genic determinants involved in protective re-
258]; second, because in vitro culture systems sponses are species- and stage-specific. Three
have been developed for the blood and hepatic types of malaria vaccine can therefore be de-
stages of the malarial parasite Plasmodium fal- vised, based on:
ciparum [259], [260]; and third, because mono-
clonal antibody and recombinant DNA tech- 1) sporozoites,
niques have been used to identify and produce 2) asexual stages (merozoites, schizonts), and
the parasite polypeptides possibly involved in 3) sexual stages (gametes).
the development of protective immunity. 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- 1) Identification and selection of malaria anti-
velopment of a multivalent vaccine containing gens that are the target of protective immune
components of several malaria stages. The se- responses.
lection of defined parasite components versus 2) Functional, biochemical, and immunological
whole parasites (sporozoites, merozoites, ga- characterization of these antigens, including
metes) for vaccine development is indicated by identification of B and T cell epitopes.
the following reasons: 3) Cloning of the genes coding for protective
antigens, determination of DNA and amino
1) Large-scale production and purification of acid sequences–what is the level of antigenic
whole parasites is not feasible. diversity?
2) Parasites cannot be obtained free of host 4) Production of candidate protective antigens
components (e.g., mosquito salivary glands, or epitopes by genetic engineering or chem-
erythrocyte membranes) which may induce ical synthesis.
adverse autoimmune reactions. 5) Evaluation of candidate protective antigens
3) Most of the parasite components are irrele- in terms of production of antigens, safety
vant to the induction of protective responses tests, adjuvants, carriers, etc.
and their inclusion may impair truly protec- 6) Immunization trials in monkeys and human
tive responses or induce immunopathologic volunteers.
lesions in the host.
The strategy for the development of malaria
4.2.2. Sporozoite Vaccines
vaccine is as follows:
Following invasion of the hepatocytes, the
sporozoites develop into thousands of mero-
Immunotherapy and Vaccines 45
zoites, each of which may invade an erythrocyte. asparagine – proline (NANP) and four copies
Obviously a vaccine which neutralizes sporo- of the tetrapeptide asparagine – valine – aspartic
zoites before their entry into liver cells or within acid – proline (NVDP). The regions flanking
liver cells would optimally prevent malaria in- the repeats are more highly conserved between
fection. Vaccination with attenuated, irradiated species than the repeats. Within a species, lim-
sporozoites in rodents, monkeys, and humans ited variations also occur outside the repeats
induced complete protection against malaria [274].
[262], [263]. The induced immunity is species- The repeats cover the surface of mature
and stage-specific but not strain-specific. It is sporozoites; for P. falciparum ca. 108 molecules
at least partially mediated by antibodies as is of NANP are expressed on the membrane of
shown by the protection against sporozoite chal- mature sporozoites. The NANP repeats are the
lenge afforded by the passive transfer of anti- target of protective monoclonal antibodies pas-
sporozoite monoclonal antibody [268] and by sively transferred in vivo. The antibody response
neutralization of sporozoite infectivity follow- against sporozoites in humans is also mainly di-
ing incubation with the serum of protected ani- rected against the NANP repeats [275], [276].
mals. Deposition of specific antibodies on the In view of these findings, and because the
sporozoite surface results in the formation of NANP repeats are present on all the isolates
a tail-like precipitate; this is called the circum- of P. falciparum tested, two malaria vaccines
sporozoite (CS) reaction. Indirect evidence sug- based on NANP repeats have been prepared and
gests that antisporozoite antibodies may play a tested on human volunteers. The first, produced
role in human malaria [264], [265]. by DNA recombinant technology, consisted of
The control of sporozoite-induced infection 32 repeats of NANP and NVDP fused to a 32
also involves T cell-dependent mechanisms. amino acid tail. The second was composed of
In animal models, effector T cells can me- three NANP repeats (NANP 3) conjugated to
diate antisporozoite immunity via antibody- tetanus toxoid. Both formulations use aluminum
independent mechanisms; for example, immu- hydroxide as adjuvant [277], [278]. The vac-
nization with irradiated sporozoites can protect cines were safe and did not induce adverse reac-
B cell-deficient mice against subsequent chal- tions. Volunteers with high antisporozoite anti-
lenge infection with viable sporozoites [266]. body titers were challenged with sporozoites and
Activation of malaria-specific T cells by malaria some were protected or presented a delay in ap-
antigens induces the secretion of lymphokines pearance of parasitemia. Protection was shown
which may act directly on the malaria parasite in individuals with the highest antibody titers.
or indirectly by activating host effector systems An antisporozoite vaccine must induce high
[267]. antibody titers for a prolonged period and ideally
The CS protein has been identified in several a boosting effect should occur following expo-
plasmodia species. Passive transfer of mono- sure to sporozoites. Higher antibody responses
clonal antibodies directed against the CS protein can be obtained by changing the formulation
protects mice from sporozoite challenge infec- and concentration of the immunogens and by
tion [268]. Indirect evidence suggests that CS using other adjuvants or live-attenuated vectors
protein is involved in the binding and penetra- (vaccinia virus, salmonella) that carry and ex-
tion of sporozoites into liver cells. Similarly, press the gene coding for the CS protein. Opti-
fragments of monoclonal antibodies against CS mal antibody formation is dependent on the col-
prevent the attachment of sporozoites to hepato- laboration of T helper cells and B cells. In the
cytes in vitro [269]. two sporozoite vaccines tested, T cell help was
The gene coding for the CS protein has provided by foreign protein (tetanus toxoid) but
been cloned in several plasmodial species [270], was unable to boost the antiNANP antibody re-
[271]. The protein contains a central block sponse following exposure to sporozoites. More
of tandemly repeated amino acids which vary efficient sporozoite vaccines should contain T
in number and sequence among the different cell epitopes derived from sporozoites and ide-
malaria species [272], [273]. For example, in ally from the CS protein; in this context, proper
P. falciparum the central area consists of 37 help is provided by sporozoite-specific, primed
copies of the tetrapeptide asparagine – alanine – T cells and can lead to optimal antiNANP anti-
46 Immunotherapy and Vaccines
4.2.3.1. Merozoite Surface Antigens chor; hydrolysis of its anchor activates the pro-
teolytic activity of the 76 kDa polypeptide and
A protein with a molecular mass of 190 – may play a role in the invasion of erythrocytes
200 kDa has been identified at the surface of by merozoites [305]. The 41 kDa polypeptide
P. falciparum schizonts and merozoites [289], displays aldolase activity [306]. Interestingly,
[290]. During maturation of schizonts this both the 76 kDa and the 41 kDa components
polypeptide is processed into several compo- can induce at least partial protection against
nents, one of them (M r 83 000) being the main P. falciparum infection in monkeys [307], [308].
surface component of the merozoites [290]. An The gene coding for the 41 kDa protein has been
important feature of the 190 – 200 kDa protein cloned and presents two interesting characteris-
is its genetic polymorphism [291–294]. The tics in terms of vaccine development; absence
gene coding for the protein can be divided into of variable amino acid repeats and almost com-
blocks ranging in homology among different plete conservation of the amino acid sequences
P. falciparum isolates from 10 – 87 % at the ami- among isolates from P. falciparum [306].
no acid level. A relatively short region of vari- Another rhoptry antigen of P. falciparum
able tripeptide repeats is found close to the N with a molecular mass of 225 kDa has been iden-
terminus. The blocks encoding for the N and C tified in the peduncle of the rhoptries. It is syn-
terminal sequences are highly conserved. thesized as a 240 kDa polypeptide which is pro-
Immunization with the 190 – 200 kDa protein cessed into a 225 kDa protein during schizogony
derived from P. falciparum in monkeys [295– and is quantitatively recovered in the culture su-
297] and with an analogous protein from P. yoelii pernatant following merozoite invasion [309].
[298] can induce at least partial protection. Im- A third set of rhoptry-associated proteins
munization with synthetic polypeptides corre- is the 105 – 130 – 140 kDa complex composed
sponding to defined parts of the molecule (for of three coprecipitating but unrelated proteins
example, the N terminus and amino acids 277 – [310]. The 225 kDa proteins and the 105 – 130 –
287) also confer partial protection in monkeys 140 kDa complex have not been evaluated in im-
[299], [300]. munization trials.
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- 4.2.3.3. Antigens Associated with the
clonal antibodies. It contains variant and con- Membrane of Infected Erythrocytes
stant epitopes for various P. falciparum isolates.
There is considerable antigenic diversity The ring-infected erythrocyte surface antigen
among P. vivax isolates as regards the compo- (RESA) is a P. falciparum antigen with a molec-
nents exposed at the surface of merozoites. ular mass of 155 kDa. It is synthesized in tropho-
zoites, accumulates in the merozoite, and fol-
lowing invasion becomes associated with the
4.2.3.2. Rhoptry Antigens membrane of erythrocytes containing ring forms
of the parasite [311], [312] but is not directly ac-
Rhoptries are apical organelles of the mero- cessible on the external erythrocyte surface. An-
zoites which release their contents onto the ery- tiRESA antibodies inhibit the multiplication of
throcyte membrane during invasion. A mono- asexual blood stages in vitro and may interfere
clonal antibody directed against a rhoptry pro- with the invasion process [313]. The gene cod-
tein of a rodent malaria, P. yoelii, reduced the ing for RESA has been cloned and sequenced
virulence of the infection and a monoclonal an- [314]. It contains two blocks of repetitive ami-
tibody directed against 82 and 41 kDa compo- no acid sequences which are the immunodomi-
nents of P. falciparum inhibited the growth of nant regions of the molecules in terms of anti-
P. falciparum in vitro [301–303]. The 82 kDa body response. Antibodies directed against the
component is processed into 82 and 65 kDa com- RESA repeats crossreact with at least six other
ponents [304]. asexual blood stage components. Aotus mon-
The 82 kDa polypeptide is membrane-bound keys have been immunized with fusion proteins
through a glycosyl – phosphatide – inositol an- corresponding to various areas of the RESA and
48 Immunotherapy and Vaccines
with synthetic polypeptides corresponding to Several immunization trials have been con-
the repetitive sequences [315]. Partial protection ducted in monkeys using synthetic peptides de-
was observed in some groups of animals; work rived from asexual blood stages of P. falciparum
is in progress to optimize the efficacy of immu- coupled to carrier proteins [298], [299], [315].
nization based on RESA-derived molecules. A partial protective response was observed with
Erythrocytes containing mature asexual peptides corresponding to various areas of the
blood stages of P. falciparum attach to endothe- 190 – 200 kDa protein, to RESA, and to frag-
lial cells lining the venules of deep tissues. ments of parasite components identified by their
This cytoadherence of mature parasites pre- molecular mass of 55 and 35 kDa [299]. Syn-
vents their passage through the spleen and thus thetic hybrid polymer – proteins containing sev-
their exposure to localized destructive mech- eral peptides corresponding to epitopes of 195 –
anisms. Electron-dense protuberances (knobs) 200 kDa, RESA, 55 kDa, 35 kDa, and CS pro-
on the plasma membranes of infected erythro- tein have been used for immunization of human
cytes are implicated in cytoadherence [316]. volunteers [324]. The vaccine was well toler-
The genes coding for two knob components ated and no adverse effects were observed. All
(knob-associated histidine-rich protein M r 85 – the immunized and control volunteers had patent
105 kDa) and mature parasite-infected erythro- parasitemia but the majority of the immunized
cyte surface antigen (M r 240 – 300 kDa) have volunteers were able to control their parasitemia
been cloned [317], [318]. The two proteins dif- in the absence of drug therapy. The immune re-
fer antigenically among isolates and contain re- sponse was low in terms of specific antibody
peated amino acid sequences. Cytoadherence production, and cell mediated responses as mea-
can be inhibited by antisera in a strain-specific sured by proliferation assays was undetectable.
manner [319]. However, an antigenically invari-
ant epitope has also been identified on the sur-
face of infected erythrocyte isolates and may be 4.2.4. Sexual Stages–Transmission Blocking
an important antigen for vaccine development Immunity
[320].
The transmission of malaria from the vertebrate
host to the mosquito vector is effected by sexual
4.2.3.4. Other Proteins and Synthetic parasite stages–the gametocytes–which develop
Peptides from merozoites. Within the vertebrate host the
gametocytes are surrounded by the erythrocyte
A number of other antigens are also candidates membrane; following ingestion by the mosquito
for vaccine development. P. falciparum requires vector, the gametes become extracellular. The
exogenous iron in the form of ferrotransferrin. A female gametes are fertilized by the male ga-
malaria transferrin receptor at the surface of in- metes in the midgut of the mosquito to pro-
fected erythrocytes transports bound ferrotrans- duce zygotes which develop into ookinetes. The
ferrin to the parasite and may be used as a target ookinetes penetrate the midgut wall where they
for the vaccine [321]. Glycophorins exposed at remain to form oocysts in which the sporozoites
the erythrocyte surface may act as ligands for develop.
P. falciparum merozoites and P. falciparum pro- In the vertebrate host, the sexual stages do not
teins have been identified which either bind to produce illness, and their intracellular localiza-
glycophorins or to human erythrocytes [322], tion prevents direct attack by host effector mech-
[323]. A prominent antigen of P. falciparum with anisms. Within the midgut of the mosquito the
an apparent molecular mass of 126 – 140 kDa extracellular gametes are exposed to antigamete
is associated with merozoite release. The gene and/or antizygote antibodies from the vertebrate
coding for this protein has been cloned and con- host that are ingested with the mosquito’s blood
tains at least two stretches of amino acid repeats, meal. The antibodies partially or completely pre-
one being composed of polyserine repeats [324]. vent the development of sexual stages and subse-
Monkeys immunized with this protein are pro- quent production of sporozoites. Transmission
tected from a lethal challenge infection [295]. of the parasite is therefore blocked. This phe-
Immunotherapy and Vaccines 49
nomenon is termed transmission blocking im- tetanus, diphtheria, and rabies. Human serum
munity. was first used in 1907 by Cenci for the modifi-
The development of vaccines based on sexual cation of measles and later for mumps and per-
blood stages could have an important impact on tussis [334]. Placental extracts prepared by am-
the epidemiology of malaria in endemic areas monium sulfate precipitation [335], [336] were
by reducing the level of malaria transmission. also employed and may be considered as the first
Ideally, transmission blocking vaccines have to immunoglobulins prepared for human therapy
be used in combination with vaccines based on [336]. Placental material is still used as a source
sporozoite and/or asexual blood stages (see Sec- of immunoglobulin.
tions 4.2.2 and 4.2.3). The serious hypersensitivity reactions as-
Transmission blocking immunity has been sociated with animal serum proteins and the
induced by immunization with extracellular ga- risk of viral hepatitis with convalescent human
metes in several animals [325–328]. The in- serum limited the use of serum therapy to life-
duced antigamete response is long lasting and threatening infections. In the 1920s attempts
in some cases is boosted by malaria infection, were made to separate the immune substances
probably due to the gametocyte antigens in the from animal serum by alcohol or acetone treat-
circulation of the vertebrate host [328], [329]. ment. One such preparation, Huntoon’s antibody
There is also evidence that in P. vivax malaria in solution, was administered intravenously to over
humans the antigamete response is boosted dur- 400 patients without the occurrence of ana-
ing natural infection [330]. Addition of sera of phylaxis or serum sickness; however, pyrexia,
previously infected individuals to gametes can cyanosis, and dyspnea did occur and were im-
prevent fertilization and development of oocysts plicated in the deaths of three patients [337].
in mosquitos. The introduction of antibiotics in the 1930s de-
Specific targets for antigamete immunity creased the demand for serum therapy [338].
have been identified using monoclonal antibod- During this period, however, the experimental
ies in species including the human parasites basis for combination therapy with antimicro-
P. falciparum and P. vivax [331], [332]. The an- bials and hyperimmune animal serum was es-
tibodies act against the gametes by prevent- tablished [339], [340].
ing fertilization and against the zygotes and A wide variety of biological products is
ookinetes by preventing further development. In now available for immunotherapy, the most im-
P. falciparum the target antigens for inhibition of portant being purified gamma globulin for in-
fertilization are polypeptides with a molecular tramuscular injection (standard immune serum
mass of 230 kDa and 45 – 48 kDa [331], [332]. globulin, ISG) and globulin for intravenous in-
Some of the epitopes on the 45 – 48 kDa anti- jection (standard intravenous immunoglobulin,
gen have been defined and are the targets of in- IVIG). Hyperimmune globulins with a high an-
hibitory monoclonal antibodies [333]. New anti- tibody titer against specific pathogens are also
gens are expressed at the surface of the zygote used. Additional preparations include antitox-
and ookinete and one of them (M r = 25 kDa) is ins, plasma, and other blood products.
a probable target of inhibitory monoclonal anti-
bodies [331].
5.1. Gamma Globulin Preparations
Since the late nineteenth century considerable Historical Aspects. In 1936 Arne Tiselius
progress has been made in our concepts of pas- separated serum proteins into four major frac-
sive immunotherapy and in the development of tions by electrophoresis; subsequently he and
preparations safe for human use; however, the Kabat found that immunoglobulin occurred
proper role for such therapy in clinical medicine predominantly in the gamma electrophoretic
still needs to be defined. fraction [341]. Cohn and colleagues devised
By 1900 immune serum from various ani- a procedure for recovering immunoglobulins
mal species had been used to treat pneumonia, (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.
Standard immune serum globulin is usually was attributed to contamination of the prepara-
prepared as a 16.5 % injectable solution (165 mg tion by staphylococcal enterotoxin [352]. The
per mL) and represents a ca. 20 – 25-fold con- intravenous administration of ISG was repeated
centration of plasma IgG [336], [343]. This pro- by Janeway in 1970 [336], and induced se-
vides an effective dose of antibody in a relatively vere cardiovascular (tachycardia, arrhythmias,
small volume [343]. Because of the viscosity of hypotension, severe chest pain), tachypneic,
the preparation, it can only be given intramuscu- and pyrogenic (fever, chills, malaise) reactions
larly or subcutaneously [343], usually through a [343]. These adverse effects were attributed to
large gauge needle (16 – 18 ga). It has been esti- the presence of immunoglobulin aggregates that
mated that 1 g of ISG contains 4×1018 antibody activated complement [353], [354].
molecules with >107 specificities [334]. The To avoid the above-mentioned severe reac-
product, although highly stable at 4 ◦ C, can still tions, methods were introduced to rid the prepa-
undergo proteolysis, presumably due to plasmin rations of the high molecular mass immunoglob-
contamination [336]. The ISG may also contain ulin aggregates. Although they could be re-
blood group substances, IgA, and IgG dimers. moved by centrifugation, reaggregation usually
The presence of IgA can result in anaphylactic occurred and this method was not practical on
reactions in patients who lack IgA [336]. a large scale [355]. Consequently, attention was
turned to chemical modification.
Various methods of chemical modification
5.1.2. Immunoglobulin for Intravenous Use have been tried. Degradation with the enzymes
Intravenously administered gamma globulin is pepsin and plasmin resulted in the formation of
needed to allow greater patient comfort, increase antibody fragments with immunologic activity
acceptability, and provide larger quantities of an- [336], [356]. In addition, since the proteolytic
tibody. This is particularly important for patients enzyme was not removed, it may have contin-
who have a small muscle mass (children) or in- ued to be active in the absence of lyophiliza-
sufficient skin surface (burns), who are at risk tion; porcine pepsin also induced antibody for-
from uncontrollable bleeds (Wiskott-Aldrich or mation [354]. Reaggregation was also prevented
other bleeding diatheses), or who need either by acidification with hydrochloric acid [357]
large repeated doses (immunodeficient) or rapid and reduction – oxidation with dithiothreitol,
onset of peak levels (intoxicated patients). Fur- iodoacetamide, sulfite, or tetrathionite. Alkyla-
thermore, intravenous administration avoids the tion and acylation of immunoglobulin has also
local degradation of ISG antibody that may oc- been accomplished with β-propiolactone.
cur at the injection site [351]. The IVIG preparations that have been re-
In 1948 Cohn administered 25 – 50 mL of duced and alkylated or otherwise modified may
ISG intravenously and saw no adverse effects have impaired complement binding and altered
[336], [352]. Janeway and Cohn repeated the subclass distribution [358], impaired opsonic ac-
experiment with a new preparation and induced tivity in vitro [359], shortened serum half-life,
severe reactions after only 2 mL. This reaction and decreased protective efficacy [360]. Con-
52 Immunotherapy and Vaccines
sequently, a new generation of native or in- 6) are as unmodified as possible so that opsonic,
tact IVIG preparations has been produced with- complement, and other biologic activities are
out modification by methods which include ad- maintained;
justment to pH 4, use of poly(ethylene glycol), 7) contain all IgG subclasses;
ethanol precipitation, ultra- or diafiltration, and 8) have high levels of antibody to at least two
ion-exchange chromatography [353]. Four un- bacterial species or toxins and two viruses (to
modified products are licensed in the United be ascertained by neutralization tests), and
States: one prepared by adjustment to pH 4 in 9) contain at least 0.1 International Units (IU)
the presence of traces of pepsin, one prepared of antibody to hepatitis B and have a 1 : 1000
by ion-exchange chromatography and ultrafil- titer to hepatitis A [364].
tration, and a third by diafiltration, ultrafiltra-
tion, and adjustment of pH to 4 – 4.5. A fourth In addition, manufacturers should specify the
product (Venoglobulin-1 from Alpha Therapeu- diluent and any chemical modifications used. Fi-
tic Corporation) licensed in the United States but nally a product can only be classified as a hyper-
produced in Japan is prepared by poly(ethylene immune immunoglobulin if the antibody level is
glycol) fractionation and ion-exchange adsorp- five times that of standard ISG preparations.
tion. There is no consensus on which laboratory
Commercially prepared IVIG is usually sta- tests should be used to predict the safety of IgG
bilized with a mono- or disaccharide, such as (contact activation, anticomplementary activ-
10 % maltose [338] or glucose. This consider- ity) [336]. Safety requirements should, however,
ably decreases the incidence of side effects that consider the IgG half-life and virus transmission
accompany the infusion of IVIG [336], [357], (particularly of human immunodeficiency virus
[361], perhaps by minimizing precipitation or HIV, but also of hepatitis B and non-A, non-B
aggregation. Maltose can cause a mild diuretic hepatitis viruses) [353].
effect [362].
that ISG modified, but did not prevent this dis- titers of antibody are needed [365]. Neverthe-
ease [365]. The ISG dose for treating hepatitis less, the use of ISG is recommended for women
A is 0.02 mL/kg body weight [336]. A hyper- exposed to the disease during early pregnancy
immune antibody preparation is also available [364].
commercially. In the case of hepatitis B, ISG is
not recommended. Instead, hyperimmune glob- Clinical Studies. On the basis of data avail-
ulin is given (see Section 5.3). Although ISG can able in 1980, a committee of the WHO observed
decrease the incidence of post-transfusion hep- that it was “inappropriate” to use standard im-
atitis [366], its use for non-A, non-B hepatitis is munoglobulin for the prevention of infection in
considered optional [364]. premature infants, during the physiologic hy-
pogammaglobulinemia of infancy, or for mal-
Measles. The ability of convalescent serum nutrition. Its use is contraindicated in patients
to modify the course of measles was demon- with selective IgA deficiency [364]. Previous at-
strated in 1907 [335]. The ISG has also been tempts in the 1960s showed that ISG did not
shown to prevent or attenuate the disease [343]. prevent infection in a variety of clinical situa-
The introduction of active immunization with a tions [370], [371] including multiple myeloma
measles vaccine in 1963 (see Section 3.1) sig- [372]. Although ISG was unable to prevent up-
nificantly decreased the incidence of this disease per or lower respiratory tract infections in chil-
in the United States. Standard ISG is now rec- dren or institutionalized, elderly adults, a sig-
ommended for infants under one year of age or nificant decrease was seen in the incidence of
for immunodeficient patients within six days of mumps in children and fevers of unknown origin
acute exposure to a case of measles [364]. in adults [373]. Its use in the prevention of in-
fection in burned patients has yielded conflicting
Polio. If given early, ISG can modify the par- data [335]. In a study on Peruvian children with
alytic complications of polio [335], [346]. An burns over at least 10 % of their body surface
unexposed individual should receive 0.15 mL/kg area, there was a 41 % incidence of septicemia
body weight. and 15 % mortality in control patients compared
to a 21 % incidence of septicemia and 6 % mor-
Prevention of Infection in Patients with tality among those who received either plasma
Hypogammaglobulinemia. An intramuscular or ISG [374]. However, in another study in the
ISG dose of 100 mg/kg given every 3 – 4 weeks United States administration of ISG did not ben-
is recommended for patients with hypogam- eficially affect either the rate of septicemia or
maglobulinemia. This maintains a serum level mortality in burned patients [375]. More re-
of circulating IgG above 200 mg/mL [336] and cently, patients admitted to a burn unit in India
confers protection against a wide variety of in- were randomized into groups that received ac-
fections [342]. This minimum recommended tive immunization against Pseudomonas aerug-
dose is based on a study by the British Medical inosa, passive immunization with a cold etha-
Research Council conducted between 1956 – 66 nol precipitate of plasma from normal immu-
on 176 patients [367]. Although serum IgG lev- nized volunteers, both immunologic treatments,
els at this dose rarely rise to the normal range or neither. In children, but not adults, there was
[368], total replacement of IgG does not appear a significant reduction in mortality following a
necessary for preventing infection [369]. Al- daily dose of just over 20 mg protein for three
though a monthly dose of 200 mg/kg was found days. There was a decrease in the incidence
superior to 100 mg/kg, most patients did not tol- of not only P. aeruginosa, but also other gram-
erate more than the 100 mg/kg [361]. Two other negative bacilli among those treated immuno-
studies also showed that higher doses of ISG de- logically [376].
creased the incidence of acute infections [334]. The ability of ISG to prevent infections has
Individualization of doses and their frequency been most extensively studied in high-risk, pre-
has also been suggested [368]. mature infants. This population has a hypogam-
maglobulinemia that exposes them to a high inci-
Rubella. Standard ISG is unreliable in the dence of bacterial infection [377]. The more pre-
modification of rubella [335], [337], [365]. High mature the infant the lower the IgG, since most of
54 Immunotherapy and Vaccines
the transplacental transfer of IgG occurs during rabies [364] (see Section 3.7). Combined active
the last six weeks of gestation [370]. Children and passive immunization against rabies have
do not attain adult levels of immunoglobulin un- been shown to be superior to active immuniza-
til two years of age. In two studies using 0.5 – tion alone [336].
3 mL/kg, no prophylactic effect was observed
[371], [378], although there was a suggestion Rho(D) immune globulin. This hyperim-
that the high dose regimen may have had a ben- mune globulin is recommended for rhesus-
eficial effect. negative mothers who deliver rhesus-positive
infants [364].
used to immunize Apache infants. There was ability to screen large numbers of samples for
a significant reduction in the incidence of sys- antibody levels or to immunize volunteers with
temic disease caused by H. influenzae and pneu- an increasing number of vaccines, an increas-
mococci during the first six followup months as ing number of publications have examined the
well as a significant decrease in the incidence of efficacy of passive immunotherapy, particularly
bacteremia [380]. with hyperimmune IVIG preparations, in the
treatment and prophylaxis of infectious diseases.
Intravenous immunoglobulin G is effective in
5.4. Therapy with Immune Serum the prevention of infection in patients with hy-
Globulin pogammaglobulinemia [358]. It has also been
shown to be effective for treating chronic in-
Data suggest ISG is active in bacterial infections fection in such patients who developed sinopul-
in animal models [379], [381] and may be syner- monary infection despite ISG maintenance ther-
gistic with antibiotics [339], [340], [382]. How- apy [351].
ever in 1968 Schless and Harrell [379] and
others [336] observed that there was little evi-
dence to support its therapeutic efficacy in es- 5.5.1. Viral Infection
tablished infection in humans; they suggested
the need for a controlled clinical trial of ISG Since cytomegalovirus (CMV) is a frequent
in the treatment of systemic infection in patients cause of infection in patients undergoing organ
that were not deficient in antibody [379]. Indeed, transplantation and effective antiviral therapy is
since the most functionally-active antibody to lacking, there has been considerable interest in
gram-negative bacteria is IgM and not the IgG the use of IVIG that is hyperimmune in CMV
found in ISG, little benefit was to be expected antibody for both the prophylaxis and treatment
[342]. The ISG is active against a wide vari- of CMV infections.
ety of human pathogens in animal models of
infection [379]. When used with antimicrobial Prophylaxis. Hyperimmune CMV-IVIG (to-
agents, ISG has possible benefits in experimen- tal dose 550 mg/kg) was shown to decrease
tal infection with a wide variety of organisms the attack rate of symptomatic CMV infec-
[382], [383] and in clinical infections in humans tion among CMV-negative patients who re-
[339], [340], [384], [385]. Large daily doses of ceived kidneys from CMV antibody-positive
ISG administered intravenously for 10 days to donors from 60 – 21 % [388]. This treatment
patients with leukemia and fever was well toler- also decreased the incidence of fungal and
ated [386]. However, no benefit was found for parasitic infection. Hyperimmune CMV-IVIG
patients who received the ISG in addition to anti- (3×200 mg/kg doses) also prevented mortality
biotics (compared to antibiotics alone). The ad- and interstitial pneumonia from CMV for 120
ministration of ISG to shorten the course of in- days in leukemic patients who underwent bone
fection in children under two years of age was marrow transplantation [389]; however in an-
ineffective [387]. 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-
5.5. Prophylaxis and Therapy with
row transplantation patients [391]; however, the
Intravenous Immunoglobulin (IVIG) incidence of symptoms and interstitial pneumo-
nia decreased. In contrast, CMV-IVIG given to
In studies on the prophylactic or therapeutic ef- CMV-negative patients undergoing bone mar-
ficacy of ISG, the possibility that higher doses row transplantation had no effect on either the
of standard immunoglobulin might improve ef- prevention of new disease or the amelioration
ficacy was a recurring theme. With the availabil- of established disease [392]. Therapy of bacte-
ity of IVIG, larger volumes of immunoglobulin rial infections following infection with human
could be administered directly into the blood- immunodeficiency virus are discussed in Sec-
stream (see Section 5.1.2). In addition, with the tion 5.5.2.
56 Immunotherapy and Vaccines
Therapy. Immunoglobulin has been used it is not surprising that similar studies have now
sporadically in the treatment of viral disease. been reported with IVIG. In one nursery with
High-titered CMV-IVIG from screened donors a high rate of infection, a single dose of IVIG
did not show efficacy in the treatment of bone (120 mg/kg within 2 h of birth) decreased ac-
marrow transplant patients with documented quisition of infection and mortality in preterm,
CMV infection [393]. IVIG has been shown to lowbirthweight neonates [404]. A second dose
alter the course of echovirus encephalitis infec- at eight days, conferred no further advantage.
tion in three patients with hypogammaglobuline- Prophylaxis with IVIG (0.5 mg/kg/week for four
mia. Although no benefit occurred from giving weeks) was significantly better in preventing in-
the IVIG intravenously in two of these patients, fection and death in neonates, but only in the
intraventricular administration resulted in clin- subpopulation that weighed less than 1500 g and
ical cures [394], [395]. The use of IVIG was had a gestational age of less than 34 weeks [370].
unable to alter the lethal course of polymyositis In another study, antibiotics were given either
secondary to echovirus in another patient with alone or with IVIG to women 27 – 36 weeks
hypogammaglobulinemia [396]. IVIG has also pregnant who had chorioamnionitis. Only high
been used experimentally in the successful treat- doses (24 g/d for five days) of IVIG given after
ment of herpes infection in mice [397]. the 32nd week of pregnancy prevented infection
in the delivered babies [405]. The investigators
concluded that little transplacental transfer of
5.5.2. Bacterial Infection IgG occurred before the 32nd week of gestation.
Data on the use of standard IVIG for the pre-
Numerous studies have demonstrated the ef- vention of bacterial infections in patients with-
ficacy of hyperimmune IVIG in the prophy- out hypogammaglobulinemia is limited. High
laxis and, if used early after infection, the treat- levels of specific antibody are needed; these may
ment of bacterial infection. In 1943 Alexan- not be found in standard, nonimmune IVIG, de-
der showed that the combination of a sulfa drug spite the possibility of delivering larger amounts
and animal hyperimmune sera was more effec- of IVIG than was the case with ISG. Hyper-
tive in reducing mortality than either agent alone immune products have been shown to prevent
[339], [340]. A number of studies have demon- specific infections in experimental models (see
strated the efficacy of IVIG in both the pre- Sections 5.4, 5.5.2, and 5.6). The administration
vention and treatment of experimental infection of nonimmune IVIG (1000 mg/kg) before bone
with H. influenzae in neonatal rat models [398]; marrow transplantation and weekly for 17 weeks
with E. coli [399] and Klebsiella [400] in mouse thereafter had no effect on the acquisition of ei-
models; and with P. aeruginosa in neutropenic ther bacterial or fungal infection [406]. Clinical
and burned rodent models [401], [402]. Intra- trials are being performed to test the efficacy of
venous immunoglobulin hyperimmune to group IVIG in the prophylaxis of bacterial infection
B streptococcal surface antigens can prevent and in adults with chronic lymphocytic leukemia,
treat serious bacteremia in experimental infec- a condition which may be complicated by hy-
tion in monkeys [403]. pogammaglobulinemia [407].
Prophylaxis. Earlier data with ISG in both Therapy. Patients with human immunodefi-
experimental bacterial infection in animals and ciency virus (HIV) infection (i.e., acquired im-
in clinical infection in humans indicated that in mune deficiency syndrome, AIDS) have a dys-
some situations (e.g., patients with hypogam- function in their humoral immune system which
maglobulinemia), exogenous standard gamma includes both a decreased antibody response to
globulin could prevent the acquisition of serious bacterial antigens and an altered distribution of
bacterial infection. Similar studies with IVIG IgG subclasses [408]. Unlike adults, who tend
have established the efficacy of these prepara- to acquire opportunistic infections (infections
tions [358]. in which immunodeficient individuals are in-
Since earlier investigators believed that larger fected by organisms that are normally with-
doses of ISG might be effective in the preven- stood by immunocompetent individuals), young
tion of bacterial infection in high-risk neonates, children and particularly infants infected with
Immunotherapy and Vaccines 57
HIV often resemble patients with primary hu- recovered from bleeding secondary to a co-
moral immunodeficiency and tend to suffer from incidental platelet deficiency (thrombocytope-
bacterial infections. Consequently, prophylaxis nia) following the infusion [412]. Following
with monthly doses of IVIG has been used in this fortuitous observation, experiments showed
the management of these patients. In one pi- that a high dose of IVIG could indeed re-
lot study, a decreased incidence in episodes of verse thrombocytopenia in the absence of hy-
fever and bacteremia was noted. This was ac- pogammaglobulinemia; furthermore, a regimen
companied by clinical improvement, prolonga- of 400 mg/kg/d for five days was effective treat-
tion of life, and improvements in other immuno- ment for idiopathic thrombocytopenic purpura
logic parameters [408]. In a 37 month old child (ITP) in children [413]. The use of IVIG and oral
with AIDS, monthly doses of IVIG produced steroids was also compared. Among those who
increases in IgG 2 antibody and antibody to 12 responded rapidly to treatment (62 %), IVIG was
pneumococcal serotypes, and prevented subse- as efficacious as steroids; however the slower
quent episodes of bacteremia [409]. On the basis responders responded better to the IVIG treat-
of these preliminary data, IVIG prophylaxis has ment. Administration of IVIG was accompanied
been advocated in the treatment of childhood by a doubling of serum IgG; IgM levels rose
AIDS [408]. Periodic administration of IVIG under both treatment regimens. The IVIG may
decreased lactic dehydrogenase activity (a pro- possibly reduce clearance of antibody-coated
posed indicator of pulmonary interstitial inflam- platelets via an Fc-mediated mechanism (see
mation) in adults and children with HIV infec- Section 5.5.3.2) [414]. The dosage regimen used
tion [410]. in the above study (400 mg/kg/d for five days)
Attempts to demonstrate a significant effect has been used in many other studies on the non-
in the therapy of bacterial infections with IVIG infectious uses of IVIG. However, in another
have been much less successful than with its use study 800 – 1000 mg/kg was given as a single in-
in prophylaxis. This may be attributable to the fusion to children with ITP with similar results
shorter half-life of IVIG in the blood during in- and no reported untoward effects [415]. Experi-
fection [368], [381]. The addition of IVIG to ence with IVIG for ITP has not been uniformly
standard regimens of antibiotics decreased the successful however [416].
mortality from documented bacteremia, particu- IVIG has been used in the treatment of
larly among preterm infants. The number of sub- immunologically-mediated blood disorders in
jects studied was too small for statistical analy- both pediatric and adult populations. These
sis, however [411]. include autoimmune hemolytic anemia [417],
In adults, little data is available to support [418], autoimmune neutropenia [419], post-
the use of IVIG in the treatment of bacterial in- transfusion purpura [420], ITP in adults [421],
fections. This may be due to the need for high [422], chronic ITP in adults and children
levels of antibody specific for the invading or- [423], [424], thrombocytopenia secondary to al-
ganism (rather than simply high levels of non- loimmunization in leukemic patients receiving
specific antibody) as well as the need for prompt platelet transfusions who became refractory to
initiation of therapy. In most experimental stud- subsequent platelet transfusions [425], thrombo-
ies with hyperimmune IVIG, little benefit can cytopenia secondary to transplacental passage of
be shown if the exogenous immunoglobulin is antiplatelet antibodies [426], ITP of pregnancy
given more than 8 h after infection (see [381], [427], antibody-mediated red cell aplasia [428],
[383]). In clinical medicine, however, identifi- during pregnancy in women with severe rhesus
cation of the time of onset of infection is often immunization [429], and in conjunction with cy-
difficult. clophosphamide used to treat antibody to factor
VIII in hemophilia [430].
In addition to these hematologic disorders,
5.5.3. Noninfectious Diseases two large trials conducted in patients with
5.5.3.1. Therapeutic Effect of IVIG Kawasaki’s disease have compared the use
of aspirin alone to that of aspirin and IVIG
In 1981 it was reported that a patient who re- (400 mg/kg/d for 4 and 3 days) [431], [432].
ceived IVIG for hypogammaglobulinemia also
58 Immunotherapy and Vaccines
The IVIG reduced the fever and the incidence the protection of platelets or megakaryocytes
of coronary artery disease [431], [432]. It sup- against antiplatelet antibodies [441], antiidio-
pressed the T and B cell activation characteristic typic suppression of antibody synthesis [436],
of patients with this disease and decreased the [442], blocking of the Fc receptor by antibod-
levels of spontaneous immunoglobulin synthe- ies, and the production of antilymphocyte an-
sis in vitro [433]. tibody [443]. However, patients with ITP have
Placentally derived immunoglobulin has responded to IVIG without demonstrable alter-
been administered to patients with severe ation in Fc-receptor mediated clearance [441]. In
rheumatoid arthritis because the symptoms of addition, preincubation of erythrocytes in IVIG
some patients improved during pregnancy [434]. failed to inhibit the phagocytosis of sensitized
None of five patients given IVIG derived from erythrocytes by cultured macrophages [444].
control plasma improved whereas 3/5 improved Thus, IVIG may lead to improvement in such
under the placentally derived globulin. An an- patients by multiple mechanisms.
tihuman leukocyte antigen – DR surface antigen IVIG is a potent modulator of antibody pro-
(anti HLA-DR) antibody in the placental prepa- duction both in vitro and in vivo. It inhibits
ration was presumed to be a possible mechanism lectin-driven B cell differentiation in vitro [445],
for this improvement. [446] and immunoglobulin production by pe-
High dose immunoglobulin treatments have ripheral blood mononuclear cells stimulated
also been used in patients with Felty’s syndrome with pokeweed mitogen [446]. For these ef-
[435], myasthenia gravis [436], epilepsy [437], fects the Fc portion of immunoglobulin must
and multiple sclerosis [438]. In the latter study be present: the Fc portion alone was 100 fold
one-third of patients worsened with IVIG ther- more effective than the intact IgG. The observa-
apy. tion that IgM antibody rises after IVIG infusion
has raised the possibility that it may stimulate
some immunoglobulin-producing cell popula-
5.5.3.2. Mechanism of Action tions [423]. Interestingly, the monthly admin-
istration of ISG to premature infants during the
Many immunologic mechanisms have been in- first year of life resulted in a significantly lower
voked to explain the beneficial effects of IVIG level of immunoglobulin compared to the con-
in noninfectious illnesses. IVIG is a potent im- trol group [371], [447].
mune modulator; the role of the Fc portion of In adults with ITP, the infusion of IVIG
the immunoglobulin molecule in this respect is (400 mg/kg) led to a decrease in T 4 (helper)
still a subject of active investigation. lymphocytes and elevation in T 8 (suppressor)
Antibody-coated material is removed by lymphocytes (and decrease in T 4/T 8 ratio)
phagocytes in a process known as Fc-receptor- [421]. Similarly, the administration of IVIG to
mediated clearance. The Fc portion of the an- patients with hypogammaglobulinemia led to in-
tibody coating the bacterium cell, protein etc. creased suppressor T cell activity, decreased to-
binds to the Fc receptor of the phagocyte. The tal T cells, a significant decrease in the T 4/T 8
material is then taken up by the phagocyte and ratio, and a decrease in lectin-induced im-
digested. Initially, the Fc receptors in the retic- munoglobulin syntheses in vitro [448].
uloendothelial system were thought to become The ability of IVIG to decrease antibody pro-
saturated by the high dose of immunoglobulins duction may be desirable in patients producing
in IVIG resulting in a decreased clearance of autoantibody, but dangerous in those with infec-
IgG-coated particles due to competitive inhibi- tion [445]. In patients with acute otitis, repeated
tion [417], [439]. This is consistent with studies monthly [379] infusions of IVIG resulted in
in which radiolabelled, autologous erythrocytes higher IgG levels if the initial antipneumococcal
were cleared more slowly after IVIG infusion antibodies were low. However, in the presence of
than before [414], [423]; this treatment altered high initial levels of antibody, specific antipneu-
Fc receptor affinity, not receptor number [440]. mococcal antibody levels decreased [449]. The
Other suggested mechanisms include decreased suggestion has also been made that high levels
synthesis of autoantibodies [441], the clearance of nonspecific antibody could lead to a reduc-
of persistent, occult viral infections [396], [441], tion in the survival of specific antibody [450],
Immunotherapy and Vaccines 59
[451]. Evidence suggests that IVIG may even shock. The rapid infusion of large volumes
exacerbate infections. A patient with neutrope- of lyophilized human plasma from blood
nia died due to acceleration of the yeast infec- with >40 µg/mL of antibody to a mixture of
tion shortly after infusion of IVIG; the IVIG may lipopolysaccaride serotypes resulted in nearly
have blocked the normal Fc-mediated clearance 7-fold decrease in mortality in South African
mechanism [452]. women treated in an obstetrical–gynecological
Immunoglobulin G can bind both native C 3 ward for septic shock [459]. Equine plasma simi-
and C 3 b during complement activation by sol- larly screened for antilipopolysaccaride antibod-
uble immune complexes and by bacteria [453], ies has also been used routinely in veterinary
[454]. Finally since immunoglobulin has unique practice in South Africa [460].
antigenic determinants, antibodies to the im-
munoglobulin may develop [455], [456]. These Opsonins. The defective neonatal antibody
antibodies may play a role in the anti-idiotype (opsonic) response that mediates the uptake and
network. killing of bacteria by phagocytes can be par-
In summary, exogenous immunoglobulin has tially corrected with gamma globulin [461]. The
a diverse, potent effect on a wide variety of ability passively administered opsonins to pre-
immune regulatory mechanisms. These interac- vent infection was first suggested in one study
tions may often work to the patient’s benefit, but in which fresh whole blood was administered to
may also result in previously unsuspected ad- infants to prevent group B streptococcal sepsis.
verse effects. All nine infants transfused with blood having
antibody to group B streptococci lived versus
3/6 transfused with blood having undetectable
5.6. Prophylaxis and Therapy with antibody titers. Protection was correlated with
Plasma and Other Blood Products opsonic antibody titers in the infants’ blood and
with having >40 % of their blood volume re-
Although the regular, large-scale, clinical use of placed [462].
plasma and other blood products is considered
impractical (both for production and safety rea- Postimmune Serum. Donors were immu-
sons), passive immunotherapy with these agents nized with a J 5 mutant of E. coli to elicit
may provide some efficacy in the prophylaxis antibodies to widely shared core epitopes in
and treatment of infections in humans. the lipopolysaccharides of gram-negative bacte-
ria. The passive administration of postimmune
Plasma has been used to treat infections, par- serum from these individuals to patients in bac-
ticularly those caused by Pseudomonas [374], terial endotoxic shock decreased the incidence
[457]. Plasma was superior to ISG in the pre- of death as well as the need for pressor ther-
vention of infection among patients who suf- apy in patients with profound shock. However,
fered >30 % burns [457]. Similar observations since this protection did not correlate with the
were made during experimental infection [458]. presence of antibody to the core epitopes, it is
One possible explanation for the higher efficacy not clear whether the protection was conferred
of plasma is that it contains IgG, IgA, and IgM by the antibody [463]. In a follow-up study,
(ISG contains only IgG). This may improve the the prophylactic administration of postimmune
distribution of antibody at different anatomic plasma appeared to significantly prevent shock
sites. Since the isotypes have functional differ- and death from gram-negative infections in pa-
ences, the antibacterial activity is also increased tients admitted to a surgical intensive care unit.
(natural antibody against gram-negative bacilli Such treatment had no effect on infection rate,
is thought to be predominantly of the IgM iso- however [464]. In both studies the protective
type). However, plasma infusions carry the risk moiety in the plasma and serum was presumed
of transmitting hepatitis (there is no manufactur- to be antibody to the endotoxin. In any event,
ing process for plasma that inactivates the virus for this to be a practical therapy, the protec-
as is the case with ISG). tive portion of the postimmune product must be
Further efforts in the use of passive im- mass produced in the form of a safe, standard-
munotherapy have also focused on septic ized preparation (e.g., made into an IVIG from
60 Immunotherapy and Vaccines
a large pool of donors immunized with the J 5 Viral Transmission. Immune serum globu-
vaccine or into a monoclonal antibody prepara- lin G has had an outstanding safety record with
tion). Attempts to show a protective effect from regard to lack of transmission of virus to recip-
an IVIG prepared from immunized donors have ients [466]. For example, ISG is free of serum
been unsuccessful to date [465]. Nevertheless, hepatitis, even if the virus occurs in the origi-
monoclonal antibody preparations directed to- nal plasma. Ten children who received plasma
ward a J 5 epitope were evaluated in human vol- for prophylaxis against measles in England be-
unteers (see also Section 6.1). came jaundiced and three died. In contrast, when
ISG prepared from the same plasma was given
to 56 children, only one child became jaun-
5.7. Adverse Effects of Gamma Globulin diced [343]. In the United States, 4 out of 15
Preparations volunteers who were inoculated subcutaneously
with hepatitis-positive plasma developed hepati-
Immune serum globulin G is one of the safest tis. Five further volunteers received large doses
biological products available with an overall in- of ISG and none developed hepatitis [343]. The
cidence of reactions of 3 – 12 % [334]. Rare sys- ability of the Cohn – Oncley fractionation (see
temic (anaphylactic) reactions may occur during Fig. 5) to inactivate virus may in part be due
or within minutes of administration of ISG (ca. to the high concentration of antivirus antibody
1/500 – 1/1000 injections). Late-occurring sys- [353]. Non-A, non-B hepatitis has been reported
temic reactions (within hours or days) include in pastes produced in this procedure: in one, the
arthralgias, pyrexia, and diarrhea and are not un- lyophilization step was omitted, and in two oth-
common in immunodeficient patients [364]. Lo- ers an ion-exchange step was added to lower IgA
cal reactions occur with intramuscular adminis- but it lowered IgG 4 levels as well [353].
tration of ISG and are related primarily to the In contrast, non-A, non-B (NANB) hepati-
large volume administered. Local reactions to tis has been reported following IVIG infusion
ISG can be prevented by using small volumes or [467–469]. In one such episode 16 out of 77
by pretreatment with analgesics. Systemic re- patients who received IVIG for immunodefi-
actions may be modified by the use of aspirin, ciency acquired hepatitis with death occurring
hydrocortisone, or antihistamines. in 5 cases. Interestingly, no hepatitis was ob-
Systemic reactions with IVIG are more com- served following administration of ISG prepared
mon than with ISG and often depend on the from the same serum pool [467]. Since neither
rate of infusion and the type of preparation. of the IVIG products was subjected to a recog-
Patients receiving IVIG for hypogammaglobu- nized virucidal finishing step, such as treatment
linemia have a reported incidence of 2.5 % re- with acid, β-propiolactone, or ultraviolet radia-
actions which are usually mild [358]; the inci- tion, the virus may still have been viable. Thus
dence of these reactions decreases after the first ethanol fractionation by itself might not fully
two months of therapy. The cause of reactions to inactivate putative virus [470], though it is viru-
IVIG are not clear but may depend on aggregate cidal for enveloped viruses [471]. Supplemen-
formation, IgA contamination, and activated en- tary virucidal procedure could make IVIG safe
zymes that initiate the release of inflammatory from NANB hepatitis transmission [471]. Un-
mediators [364]. like plasma, IVIG has not been known to trans-
The experimental work discussed above and mit hepatitis B [472].
clinical experience give reason for caution in the There have been no confirmed cases of trans-
use of immunoglobulin preparations. Clinical mission of HIV virus in IVIG but antibody to
experience primarily involves concern about vi- HIV has been isolated from two patients with
ral transmission, anaphylactic reactions among primary hypogammaglobulinemia [471]. When
patients with IgA deficiency, and isolated case HIV was added to plasma and then processed
reports of unusual occurrences. The indications into IVIG, >104 PFU/mL (PFU denotes plaque-
for the use of gamma globulin should be well forming units) of HIV were inactivated dur-
founded [363]. ing alcohol fractionation and poly-(ethylene gly-
col) fractionation to IVIG [473]. Antibody to
HIV was detected by an enzyme-linked im-
Immunotherapy and Vaccines 61
munoabsorbent assay (ELISA) in all 10 lots now known to be a potent modulator of the im-
of a reduced, alkylated IVIG and in 4/8 lots mune system, IVIG has been used as a form of
of a pH 4/pepsin IVIG; 8/10 and 3/8 lots were therapy for many noninfectious diseases. How-
also positive by Western blot analysis. The HIV ever, the impact of such treatment on the abil-
antibody-positive lots, which were negative for ity to respond to infectious diseases often com-
HIV by culture and reverse transcriptase activity, plicates these conditions and still has to be as-
were infused into patients. Comparison of pre- sessed. Studies to date do not clearly show that
and post-infusion sera demonstrated that anti- the indications for the use of IVIG in infectious
body to HIV was detectable for up to one month diseases should be expanded beyond those al-
before converting to a negative antibody status ready shown for ISG. In some conditions involv-
[472]. ing functional hypogammaglobulinemia (e.g.,
One woman who had received hyperim- HIV infection in young children, chronic lym-
mune Rho (D) immunoglobulin subsequently phocytic leukemia), prophylaxis with standard
was found to be infected with HIV but belonged IVIG may ultimately demonstrate efficacy. In
to a group with an increased risk of acquiring certain subpopulations that are at risk of infec-
infection with HIV [474]. tion with specific organisms over a finite period
of time, immunoglobulin preparations enriched
Anaphylaxis. Anaphylaxis has been docu- in antibody to those organisms (i.e., hyperim-
mented following the administration of IVIG mune preparations) may prove useful. The most
to some patients with hypogammaglobulinemia. promising use appears to be in the prophylaxis
Indeed, patients with antibody deficiency have of specific infections; this usually involves hy-
an increased incidence of adverse reactions to perimmune serum and, as shown by studies in
IVIG [342]. Low levels of IgA are associated neonates, the substantial replacement of blood
with an increased risk of such anaphylactic re- volume.
actions [475]; IVIG products differ markedly in The use of IVIG to treat established infections
their IgA content [476]. In two patients with with pharmacological (not replacement) doses
common variable immunodeficiency, IgE anti- of immunoglobulin, (with the possible excep-
body to IgA in the infused IVIG was felt to ac- tion of antitoxin treatment of septic shock), still
count for these reactions [475]. In addition, IgE has no proven role. This may be related to the
has been detected in commercial preparations of need for prompt intervention and a large amount
IVIG [477]. of a specific antibody, requiring rapid identifi-
cation of the microorganisms involved. Prepa-
Other clinical problems have been de- rations that are known to be immunotherapeu-
scribed which involve the use of IVIG. IVIG tically effective are less efficacious when given
prophylaxis precipitated cryoglobulinemic 8 h or more after the onset of infection. Substan-
nephropathy in a patient with hypogammaglob- tial evidence suggests that passive immunother-
ulinemia secondary to B cell neoplasm [478]. apy may provide an enhanced therapeutic bene-
The passive transfer of antibodies against fit; however, this still remains to be proved in a
specific blood types in IVIG may also be associ- prospective, controlled therapeutic trial.
ated with hemolytic anemia or cause problems Finally, more recent studies that show some
in the crossmatching of blood prior to surgery benefit from passive immunotherapy have not
[479]. Immunologically normal patients may re- been performed with IVIG preparations, e.g.,
ceive antibody to immunoglobulin that is of a [376], [463], [464], [459]. Thus despite the sub-
different genotype. stantial progress made in the development of
IVIG preparations for human use, their firm rec-
ommendation for widespread use in infectious
5.8. Future Prospects disease requires further well-designed studies.
Humabs have been reported that recognize (The development of drugs for rare diseases is
Hemophilus influenza type B capsular polysac- supported by the US government; these pharma-
charide [498]. The successful introduction of ceuticals are called orphan drugs.)
an H. influenza type B capsular polysaccharide The acquired immune deficiency syndrome
vaccine [499] (see Section 2.13) has made the ( AIDS) is caused by the human immunodefi-
clinical use of these humabs less attractive than ciency viruses (HIV) and is currently the major
originally anticipated. However, a safe humab infectious viral disease problem. Many groups
for this life-threatening disease may still be de- are generating humabs that recognize one or
veloped for clinical use in conjunction with the more HIV strains. Many of these humabs rec-
vaccines. ognize gene product gp 160 that is found in the
Anti-mycobacterium leprae humabs have viral envelope [511], [512]. At present, there is
been reported [500]. In view of the central role of no evidence that a humoral immune response can
cell-mediated immunity in resistance and recov- prevent or change the pattern of HIV infection
ery from this disease, these reagents are unlikely in people or animals. The viruses undergo rapid
to find clinical application. Nevertheless, these mutation (five times the rate of influenza). Fur-
antibodies and those made by patients suffering thermore, they can spread by cell to cell contact
from other infectious diseases can be used to and reside permanently out of reach of humoral
probe the human humoral immune response, to immunity inside the mononuclear phagocytes
clone antigens recognized by the humabs, and scattered throughout the body. Trials are under-
to investigate antibody-mediated autoimmunity way to test the efficacy of a high-titer anti-HIV
initiated by microorganisms. antiserum in the prevention of neonatal AIDS.
The dismal failure of all attempted vaccines and
passive serum to delay the spread of the virus
6.3. Viral and Chlamydial Targets in subhuman primate models is very discourag-
ing. Antiviral humabs attached to toxins or engi-
Antiviral humabs are especially attractive ther- neered to bring cytotoxic cells into contact with
apeutic targets. Immunoprophylaxis of cy- infected cells might improve the potency of the
tomegalovirus (CMV) infections in immuno- antibody approach.
suppressed patients (e.g., transplantation) is of Several humabs have been generated against
major clinical interest. Several groups have gen- chlamydia [513] and other viruses:
erated CMV-neutralizing humabs [504], [505]
and clinical trials were scheduled to begin in Epstein Barr Virus [514]
1989. Herpes simplex [515–517]
Humabs that neutralize hepatitis B virus Human T cell leukemia virus I [518]
[506–508] will replace the antisera currently Measles (SSPE) virus [519], [520]
used after acute exposure to this agent. These Rabies [521]
humabs may also be useful for “active – passive” Rubella [522]
immunization of at-risk infants. A very large X 31 influenza virus [523]
population of infants, particularly in less devel-
oped countries, become chronically infected in However, clinical interest in the therapeutic
the perinatal period. Evidence suggests [509] purposes and use of these antibodies is limited.
that administration of immune sera will prevent
lifelong infection and could have an impact on
the prevalence of the most common human can- 6.4. Parasite Targets
cer in the world, hepatitis-B-positive hepatoma
[509]. The widespread use of anti-hepatitis A The major therapeutic target of interest among
antisera suggests that the humabs generated to parasitic diseases is malaria (see Section 4.2).
this target may also find clinical application. Much less is known about the humoral immune
A minor yet important target for humabs is response of humans suffering from other par-
the treatment of varicella zoster infections [510]. asitic diseases (Section 4.1). Passively admin-
These humabs may have to be developed as an istered humabs may have therapeutic potential
orphan drug given the small numbers of patients. in acute Falciparum malaria. Immunoglobulin
Immunotherapy and Vaccines 65
M and IgG humabs have been generated [524], 14. L. Greenfield, H. L. Dovey, F. C. Lawyer, D. H.
[525]. Monoclonal antibodies are also useful re- Gelfand: “High level expression of diphtheria
search tools in the development of malaria vac- toxin peptides in Escherichia coli,”
cines (Chap. 4). Bio/Technology 4 (1986) 1006 – 1011.
15. B. Bizzinin in R. Germanier (ed.): “Tetanus,”
Bacterial Vaccines, Academic Press, Orlando,
7. References FL 1984, pp. 37 – 68.
16. M. C. Hardegree et al.: “Titration of tetanus
1. I. Roitt: Essential Immunology, 5th ed., toxoids in international units; relationship to
Blackwell Scientific Publications, Oxford, antitoxin responses of rhesus monkeys,” 6th
United Kingdom 1984. Proc. Int. Conf. Tetanus 1981, 1982,
2. B. D. Davis, R. Dubecco, H. N. Eisen, H. S. pp. 409 – 423.
Ginsberg: Microbiology: Including 17. M. Huet: “La standardisation des vaccins
Immunology and Molecular Genetics, 3rd ed., tetaniques,” 6th Proc. Int. Conf. Tetanus 1981,
Harper and Row, Hagerstown, MD, 1980. 1982 pp. 425 – 433.
3. R. H. Schwartz, A. Yano, W. E. Paul: 18. K. W. Newell, A. Dueñas-Lehmann, D. R.
“Interaction between antigen-presenting cells Leblanc, N. Garces-Osorio: “The use of toxoid
and primed T-lymphocytes,” Immunol. Rev. 40 for the prevention of tetanus neonatorum; final
(1978) 153. report of a double-blind controlled field trial,”
4. W. Stroler, L. A. Hanson, K. W. Sell (eds.): Bull. W.H.O. 35 (1966) 863 – 871.
Recent Advances in Mucosal Immunity, Raven 19. N. Guerin, C. Fillastre: “Vaccin D.T.C. Polan,”
Press, New York 1982. 6th Proc. Int. Conf. Tetanus 1981, 1982,
5. T. Yamamura, T. Tada: Progress in pp. 477 – 479.
Immunology V, Academic Press, Tokyo 1984. 20. M. Pittman: “Pertussis toxin; the cause of the
6. J. B. Robbins, J. C. Hill, J. C. Sadoff (eds.): harmful effects and prolonged immunity of
Bacterial Vaccines, Seminars in Infectious whooping cough,“ Rev. Infect. Dis. 1 (1979)
Disease, vol. IV, Thieme-Stratton, New York 401 – 412.
1982. 21. C. R. Manclark, J. L. Cowell in R. Germanier
7. U. E. Nydegger (ed.): Immunotherapy, A (ed.): “Pertussis,” Bacterial Vaccines,
Guide to Immunoglobulin Prophylaxis and Academic Press, Orlando, FL 1984,
Therapy, Academic Press, London 1981. pp. 69 – 106.
8. R. E. Black, M. M. Levine, M. L. Clements, G. 22. T. Madsen: “Vaccination against whooping
Losonsky et al.: “Prevention of shigellosis by a cough,” JAMA J. Am. Med. Assoc. 101 (1933)
Salmonella typhi – Shigella sonnei bivalent 187 – 188.
vaccine,” J. Infect. Dis. 155 (1987) 23. C. L. Cody et al.: “Nature and rates of adverse
1260 – 1265. reactions associated with DTP and DT
9. M. M. Levine, J. B. Kaper, D. Herrington, J. immunizations in infants and children,”
Ketley et al.: “Safety, immunogenicity, and Pediatrics 68 (1981) 650 – 660.
efficacy of recombinant live oral cholera 24. R. Aldersdale et al.: Whooping Cough. Reports
vaccines, CVD 103 and CVD 103-HgR,” from the Committee on Safety of Medicines
Lancet 2 (1988) 467 – 470. and the Joint Committee on Vaccination and
10. R. J. Collier: “Diphtheria toxin: mode of action Immunization, vol. 4, H. M. Stationary Office,
and structure,” Bacteriol. Rev. 39 (1975) London, England 1981, pp. 79 – 169.
54 – 85. 25. M. Oda, J. L. Cowell, D. G. Burstyn, R.
11. A. M. Pappenheimer, Jr. in R. Germanier (ed.):
Manclark: “Protective activities of the
“Diphtheria,” Bacterial Vaccines, Academic
filamentous hemagglutinin and the
Press, Orlando, FL 1984, pp. 1 – 36.
lymphocytosis-promoting factor of Bordetella
12. F. Audibert et al.: “Active antitoxic
pertussis in mice,” J. Infect. Dis. 150 (1984)
immunization by a diphtheria toxin synthetic
823 – 833.
oligopeptide,” Nature (London) 289 (1981)
26. Ad Hoc Group for the study of Pertussis
593 – 594.
13. T. Uchida, D. M. Gill, A. M. Pappenheimer, Vaccines: “Placebo controlled trial of two
Jr.: “Mutation in the structural gene for acellular pertussis vaccines in
diphtheria toxin carried by temperate phage Sweden–protective efficacy and adresse
β,” Nature (London) 283 (1971) 8 – 11. events.” Lancet 1 (1988) 955 – 960.
66 Immunotherapy and Vaccines
27. C. S. Chuttani et al.: “Controlled field trial of a 39. J. B. Robbins, R. Austrian, C.-J. Lee, S. C.
high-dose oral killed typhoid vaccine in India,” Rastogi et al.: “Considerations for formulating
Bull. W.H.O. 55 (1977) 643 – 644. the second-generation pneumococcal capsular
28. M. Reitman: “Infectivity and antigenicity of polysaccharide vaccine with emphasis on the
streptomycin-dependent Salmonella typhosa,” cross-reactive types,” J. Infect. Dis. 148
J. Infect. Dis. 117 (1967) 101 – 107. (1983) 1136 – 1159.
29. R. Germanier, E. Fürer: “Isolation and 40. Health and Public Policy Committee,
characterization of gal ε mutant Ty 21 a of American College of Physicians:
Salmonella typhi; a candidate strain for a live, “Pneumococcal vaccine,” Ann. Intern. Med.
oral typhoid vaccine,” J. Infect. Dis. 131 104 (1986) 118 – 120.
(1975) 553 – 558. 41. P. Smit, D. Oberholzer, S. Hayden-Smith, H. J.
30. R. H. Gilman, R. B. Hornick, W. E. Woodward, Koornhof et al.: “Protective efficacy of
H. L. Dupont et al.: “Evaluation of a pneumonococcal polysaccharide vaccines,”
UDP-glucose-4-epimeraseless mutant of JAMA J. Am. Med. Assoc. 238 (1977)
Salmonella typhi as a live oral vaccine,” J. 2613 – 2616.
Infect. Dis. 136 (1977) 717 – 723. 42. G. Bolan, C. V. Broome, R. R. Facklam, B. D.
31. M. H. Wahdan, C. Sérié, Y. Cerisier, S. Sallam Plikaytis et al.: “Pneumococcal vaccine
et al.: “A controlled field trial of live efficacy in selected populations in the United
Salmonella typhi Ty 21 a oral vaccine against States,” Ann. Intern. Med. 104 (1986) 1 – 6.
typhoid; three year results,” J. Infect. Dis. 145 43. M. S. Simberkoff, A. P. Cross, M. Al-Ibrahim,
(1982) 292 – 295. A. L. Baltch et al.: “Efficacy of pneumococcal
32. World Health Organization Committee on vaccine in high-risk patients,” N. Engl. J. Med.
Biological Standardization: Requirement for 315 (1986) 1318 – 1327.
typhoid vaccine (live-attenuated, Ty21 a, oral). 44. J. A. Leech, A. Gervais, F. L. Ruben: “Efficacy
World Health Organization Technical Report, of pneumococcal vaccine in severe chronic
series 700; Geneva, Switzerland: 34th report, obstructive pulmonary disease,” Can. Med.
1984, pp. 48 – 68. Assoc. J. 136 (1987) 361 – 365.
33. R. Germanier in R. Germanier (ed.): “Typhoid 45. C. Chu, R. Schneerson, J. B. Robbins, S. C.
fever,” Bacterial Vaccines, Academic Press, Rastogi: “Further studies on the
Orlando, FL 1984, pp. 137 – 165. immunogenicity of Haemophilus influenzae
34. M. M. Levine, C. Ferreccio, R. E. Black, R. type b and pneumococcal type 6 A
Germanier, Chilean Typhoid Committee: polysaccharide-protein conjugates,” Infect.
“Large-scale field trial of Ty21 a live oral Immun. 40 (1983) 245 – 256.
typhoid vaccine in enteric-coated capsule 46. R. Schneerson, J. R. Robbins, C. Chu, A.
formulation,” Lancet ii (1987) 1049 – 1052. Sutton et al.: “Serum antibody response of
35. M. M. Levine, D. Herrington, J. R. Murphy, juvenile and infant rhesus monkeys injected
J. G. Morris et al.: Safety, infectivity, with Haemophilus influenzae type b and
immunogenicity, and in vivo stability of two pneumococcus type 6 A capsular
auxotrophic mutant strains of Samonella typhi, polysaccharide-protein conjugates,” Infect.
541 Ty and 543 Ty, as live oral vaccine in Immun. 45 (1984) 582 – 591.
humans, J. Clin. Invest. 79 (1987) 888 – 902. 47. R. Schneerson, J. B. Robbins, J. C. Parke, Jr.,
36. I. L. Acharya, C. V. Lowe, R. Thapa, V. L.
C. Bell et al.: “Quantitative and qualitative
Gurubacharya et al.: “Prevention of typhoid
analyses of serum antibodies elicited in adults
fever in Nepal with the Vi capsular
by Haemophilus influenzae type b and
polysaccharide of Salmonella typhi,” N. Engl.
pneumococcus type 6 A capsular
J. Med. 317 (1987) 1101 –1104.
polysaccharide-tetanus toxoid conjugates,”
37. R. Austrian: “Of gold and pneumococci; a
Infect. Immun. 52 (1986) 519 – 528.
history of pneumococcal vaccines in South
48. A. D. O’Brien, M. R. Thompson, P. Gemski,
Africa,” Trans. Am. Clin. Climatol. Assoc. 89
B. P. Doctor et al.: “Biological properties of
(1977) 141 – 161.
Shigella flexneri 2 A toxin and its serological
38. C. M. MacLeod, M. R. Krauss: “Stepwise
relationship to Shigella dysenteriae 1 toxin,”
intratype transformation of pneumococcus
Infect. Immun. 15 (1977) 796 – 798.
from R to S by way of a variant intermediate in
49. A. R. Higgins, T. M. Floyd, M. A. Kader:
capsular polysaccharide production,” J. Exp.
“Studies in shigellosis. III. A controlled
Med. 86 (1947) 439 – 453.
Immunotherapy and Vaccines 67
months to five years of age,” N. Engl. J. Med. Mycobacterium bovis BCG,” Infect. Immun.
297 (1977) 686 – 691. 55 (1987) 2378 – 2382.
96. H. J. Jennings, C. Lugowski: 108. B. Moss, C. Flexner: “Vaccinia virus
“Immunochemistry of groups A, B and C expression vectors,” Ann. Rev. Immunol. 5
meningococcal polysaccharide-tetanus toxoid (1987) 305 – 324.
conjugates,” J. Immunol. 127 (1981) 109. H. Leffler, C. Svanborg-Eden: “Glycolipid
1011 – 1018. receptors for uropathogenic E. coli attaching to
97. C. E. Frasch: “Prospects for the prevention of human urinary tract epithelial cells and
meningococcal disease: special reference to agglutinating human erythrocytes,” FEMS
group B,” Vaccine 5 (1987) 3 – 4. Microbiol. Lett. 8 (1980) 127 – 134.
98. W. D. Zollinger, R. E. Mandrell, J. M. Griffis in 110. K. Rauss, I. Kétyi, E. Matusovits, L. Szendrei
J. B. Robbins, J. C. Hill, J. C. Sadoff (eds.): et al.: “Specific oral prevention of infantile
“Enhancement of immunological activity by enteritis. III. Experiments with corpuscular
monovalent complexing of meningococcal vaccine,” Acta. Microbiol. Acad. Sci. Hung.
group B polysaccharide and outer membrane 19 (1972) 19 – 28.
protein,” Seminars in Infectious Disease, 111. E. Fürer et al.: “Protection against
vol. 4, Thieme-Stratton, New York 1982, colibacillosis in neonatal piglets by
pp. 254 – 262. immunization of dams with procholeragenoid,”
99. C. E. Frasch, G. Coetzee, D. M. Zahradrik, Infect. Immun. 35 (1982) 887 – 894.
L. Y. Wang in G. K. Schoolnik (ed.): “New 112. L. Doberescu, C. Huygelen: “Protection of
developments in meningococcal vaccines,” piglets against neonatal E. coli enteritis by
Pathogenic Neisseria, American Society for immunization of the sow with a vaccine
Microbiology, Washington, D.C. 1985, containing heat-labile enterotoxin (LT),
pp. 633 – 640. Zentralbl. Veterinärmed. Reihe A 23 (1976)
100. F. M. Collins in R. Germanier (ed.): 79 – 88.
“Tuberculosis,” Bacterial Vaccines, Academic 113. R. L. Morgan, R. E. Isaacson, H. W. Micon,
Press, Orlando, FL 1984, pp. 373 – 418. C. C. Brinton et al.: “Immunization of suckling
101. R. G. Ferguson, A. B. Simes: “BCG pigs against enterotoxigenic Escherichia
vaccination of Indian infants in coli-induced diarrheal disease by vaccinating
Saskatchewan,” Tubercle 30 (1949) 5 – 11. dams with purified 987 or K 99 pili; protection
102. S. V. Boyden: “The effect of previous correlates with pilus homology of vaccine and
injections of tuberculoprotein on the challenge,” Infect. Immun. 22 (1978)
development of tuberculin sensitivity 771 – 777.
following B.C.G. vaccination in guinea pigs,” 114. M. M. Levine, R. E. Black, C. C. Brinton, Jr.,
Br. J. Exp. Pathol. 38 (1957) 611 – 617. M. L. Clements et al.: “Reactogenicity,
103. A. Lotte, O. Wasz-Höckert, F. Lert, N. immunogenicity and efficacy studies of
Dumitrescu et al.: “Balance of risks and Escherichia coli type 1 somatic pili parenteral
balance of cost regarding tuberculosis and vaccine in man, Scand. J. Infect. Dis. Suppl. 33
antituberculosis vaccination,” Dev. Biol. Stand. (1982) 83 – 95.
43 (1979) 111 – 119. 115. D. G. Evans, D. J. Evans, Jr., A. R. Opekun,
104. R. Torriani, A. Zimmermann, A. Morell: “Die D. Y. Graham in A. Tagliabue, R. Rappuoli,
BCG-Sepsis als letale Komplikation der S. E. Piazzi (eds.): “Oral whole cell vaccine
BCG-Impfung,” Schweiz. Med. Wochenschr. protective against enterotoxigenic Escherichia
109 (1979) 708 –713. coli diarrhea,” Bacterial Vaccines and Local
105. W. Hennessen, H. Freudenstein, H. Immunity, Edita Da Sclavo Sp.A., Siena, Italy
Engelhardt: “Observations on a BCG vaccine 1986, pp. 155 – 156.
causing adverse reactions in newborns,” J. 116. M. M. Levine in R. Germanier (ed.):
Biol. Stand. 5 (1977) 139 – 146. Escherichia coli infections,” Bacterial
106. D. B. Travers: “BCG vaccination,” Lancet i Vaccines, Academic Press, Orlando, FL 1984
(1981) 1001 – 1002. pp. 187 – 235.
107. M. C. Lu, M. H. Lien, R. E. Becker, H. C. 117. F. A. Klipstein, R. F. Engert, R. A. Houghton:
Heine et al.: “Genes for immunodominant “Protection in rabbits immunized with a
protein antigens are highly homologous in vaccine of Escherichia coli heat-stable toxin
Mycobacterium tuberculosis, Mycobacterium cross-linked to the heat-labile toxin B
africanum, and the vaccine strain subunit,” Infect. Immun. 40 (1983) 888 – 893.
70 Immunotherapy and Vaccines
118. J. Holmgren: “Actions of cholera toxin and the Bacterial Infections, Springer-Verlag, Berlin
prevention and treatment of cholera,” Nature 1985, pp. 51 – 66.
(London) 292 (1981) 413 – 417. 130. T. M. Buchanan, M. S. Siegel, K. C. S. Chen,
119. R. A. Finkelstein, C. V. Sciortino, L. C. Rieke, W. A. Pearce in J. B. Robbins, J. Hill, J. C.
M. F. Burks et al.: “Preparation of Sadoff (ed.): “Development of a vaccine to
procholigenoids from Escherichia coli prevent gonorrhea,” Seminars in Infectious
heat-labile enterotoxins,” Infect. Immun. 45 Disease, vol. 4, Bacterial Vaccines,
(1984) 518 – 521. Thieme-Stratton, Inc., New York 1982,
120. M. A. Schmidt, P. O’Hanley, G. K. Schoolinik pp. 160 – 164.
in A. Tagliabue, R. Rappuoli, S. E. Piazzi 131. J. B. Robbins, R. Schneerson, M. Pittman in R.
(eds.): Synthetic peptides protect against Germanier (ed.): Haemophilus influenzae type
infection in a model of experimental b infections,” Bacterial Vaccines, Academic
pyelonephritis in mice,” Bacterial Vaccines Press, Orlando, FL 1984, pp. 289 – 316.
and Local Immunity, Edita Da Sclavo Sp.A., 132. A. Sutton, R. Schneerson, S. Kendall-Morris,
Siena, Italy 1986, pp. 389 – 398. J. B. Robbins: “Differential complement
121. E. C. Gotschlich in R. Germanier (ed.): resistance mediates virulence of Haemophilus
‘’Gonorrhea,” Bacterial Vaccines, Academic influenzae type b,” Infect. Immun. 35 (1982)
Press, Orlando, FL 1984, pp. 353 – 371. 95 – 104.
122. D. S. Kellogg, Jr., I. R. Cohen, L. C. Norins, 133. H. Peltola, H. Käyhty, A. Sivonen, P. H.
A. L. Schroeter et al.: “Neisseria gonorrhoeae. Mäkelä: “Haemophilus influenzae type b
II. Colonial variation and pathogenicity during capsular polysaccharide vaccine in children; a
35 months in vitro,” J. Bacteriol. 96 (1968) double-blind field study of 100 000 vaccines 3
596 – 605. months to 5 years of age in Finland,”
123. J. Swanson, S. J. Kraus, E. C. Gotschlich: Pediatrics 60 (1977) 730 – 737.
“Studies on gonococcus infection. I. Pili and 134. E. A. Mortimer, “Efficacy of Haemophilus b
zones of adhesion; their relation to gonococcal Polysaccharide Vaccine: an Enigma,” J. Am.
growth patterns,” J. Exp. Med. 134 (1971) Med. Assoc. 260 (1988) 1454 – 1455.
886 – 906. 135. H. Käyhty, J. Eskola, H. Peltola, M. G. Stout et
124. L. Greenberg, B. B. Diena, F. A. Ashton, R. al.: “Immunogenicity in infants of a vaccine
Wallace et al.: “Gonococcal vaccine studies in composed of Haemophilus influenzae type b
Inuvik,” Can. J. Public Health 65 (1974) capsular polysaccharide mixed with DPT or
29 – 33. conjugated to diphtheria toxoid,” J. Infect. Dis.
125. E. C. Tramont, J. W. Boslego: “Pilus vaccines,” 155 (1987) 100 – 106.
Vaccine 3 (1985) 3 – 10. 136. A. A. Lenoir, P. D. Granoff, D. M. Granoff:
126. C. C. Brinton, S. W. Wood, A. Brown, A. M. “Immunogenicity of Haemophilus influenzae
Labik et al. in : J. B. Robbins, J. Hill, J. C. type b polysaccharide-Neisseria meningitidis
Sadoff (eds.): “The development of a Neisseria outer membrane protein conjugate vaccine in
pilus vaccine for gonorrhea and meningococcal 2- to 6-month-old infants,” Pediatrics 80
meningitis,” Seminars in Infectious Disease, (1987) 283 – 287.
vol. 4, Bacterial Vaccines, Thieme-Stratton, 137. J. Eskola, H. Peltola, A. K. Takala, H. Käyhty
Inc., New York 1982, pp. 140 – 159. et al.: “Efficacy of Haemophilus influenzae
127. G. K. Schoolinik, J. Y. Tai, E. C. Gotschlich: type b polysaccharide-diphtheria toxoid
“A pilus peptide vaccine for prevention of conjugate vaccine in infancy,” N. Engl. J. Med.
gonorrhea,” Prog. Allergy 33 (1983) 317 (1987) 717 –722.
314 – 331. 138. C. D. Mitchell, H. H. Balfour, Jr.: “Measles
128. J. B. Rothbard, R. Fernandez, L. Wang, control: So near and yet so far,” Prog. Med.
N. N. H. Teng et al.: “Antibodies to peptides Virol. 31 (1985) 1 – 42.
corresponding to conserved sequence of 139. L. E. Markowitz, R. H. Bernier:
gonococcal pilins block bacterial adhesion,” “Immunization of young infants with
Proc. Nat. Acad. Sci. U.S.A (2) (1985) Edmonston-Zagreb measles vaccine,” Pediatr.
915 – 919. Infect. Dis. J. 6 (1987) 809 – 812.
129. M. S. Blake in G. G. Jackson, H. Thomas (ed.): 140. WHO Expert Committee on Biological
“Functions of the outer membrane proteins of Standardization: WHO Tech. Rep. Ser. 329
Neisseria gonorrhoeae,” The Pathogenesis of (1966) 52 – 73.
Immunotherapy and Vaccines 71
141. WHO Expert Committee on Biological 154. K. Penttinen, E.-P. Helle, E. Norrby:
Standardization: WHO Tech. Rep. Ser. 673 “Differences in antibody response induced by
(1982). formaldehyde inactivated and live mumps
142. Centers for Disease Control: “Measles – vaccines,” Dev. Biol. Stand. 43 (1979)
United States, 1986,” Morb. Mort. Wkly. Rep. 265 – 268.
36 (1987) 301 – 305. 155. A. Löve, R. Rydbeck, G. Utter, C. örvell et al.:
143. K. J. Bart, W. A. Orenstein, A. R. Hinman: ‘’Monoclonal antibodies against fusion protein
“The virtual elimination of rubella and mumps are protective in necrotizing mumps
from the United States and the use of meningoencephalitis,” J. Virol. 58 (1986)
combined measles, mumps and rubella 220 – 222.
vaccines (MMR) to eliminate measles,” Dev. 156. F. T. Perkins: “Licensed vaccines,” Rev. Infect.
Biol. Stand. 65 (1986) 45 – 52. Dis. 7 (1985) Suppl. 1, 73 – 76.
144. E. Norrby, G. Enders-Ruckle, V. ter Meulen: 157. S. A. Plotkin, F. Buser: “History of RA 27/3
“Differences in the appearance of antibodies to rubella vaccine,” Rev. Infect. Dis. 7 (1985)
structural components of measles virus after Suppl. 1, 77 – 78.
immunization with inactivated and live virus,” 158. WHO Expert Committee on Biological
J. Infect. Dis. 132 (1975) 262 – 269. Standardization: WHO Tech. Rep. Ser. 610
145. T. M. Varsanyi, B. Morein, A. Löve, E. (1977) 54 – 84.
Norrby: “Protection against lethal measles 159. WHO Expert Committee on Biological
virus infection in mice by immune-stimulating Standardization: WHO Tech. Rep. Ser. 658
complexes containing the hemagglutinin or (1981).
fusion protein,” J. Virol. 61(1987) 160. J. A. Dudgeon: “Selective immunization;
3896 – 3901. Protection of the individual,” Rev. Infect. Dis.
146. M. R. Hilleman: “Mumps vaccination,” in: 7 (1985) Suppl. 1, 185 – 190.
Modern Trends in Medical Virology, vol. 2, 161. Recommendation of the Immunization
Butterworths, London (1970) pp. 241 – 261. Practices Advisory Committee (ACIP):
147. K. Penttinen, K. Cantell, P. Somer, A. “Rubella prevention,” Morb. Mort. Wkly. Rep.
Poikolainen: “Mumps vaccination in the 33 (1984) 301 – 318.
Finnish defense forces,” Am. J. Epidemiol. 88 162. B. Christenson, M. Böttiger, L. Heller: “Mass
(1968) 234 – 244. vaccination program aimed at eradicating
148. M. R. Hilleman, R. E. Weibel, E. B. Buynak, J. measles, mumps, and rubella in Sweden; first
Stokes, Jr. et al.: “Live, attenuated experience,” Br. Med. J. 287 (1983) 389 – 391.
mumps-virus vaccine; 4. Protective efficacy as 163. Centers for Disease Control: “Rubella
measured in a field evaluation,” N. Engl. J. vaccination during pregnancy – United States,
Med. 276 (1967) 252 – 258. 1971 – 1986,” Morb. Mort. Wkly. Rep. 36
149. Centers for Disease Control: “Efficacy of (1987) 457 – 461.
mumps vaccine – Ohio,” Morb. Mort. Wkly. 164. Recommendation of the Immunization
Rep. 32 (1983) 391–398. Practices Advisory Committee (ACIP): “New
150. F. E. Andre, J. Peetermans: “Effect of recommended schedule for active
simultaneous administration of live measles immunization of normal infants and children,”
vaccine on the “take rate” of live mumps Morb. Mort. Wkly. Rep 35 (1986) 577 – 579.
vaccine,” Dev. Biol. Stand. 65 (1986) 165. A. Deforest, S. S. Long, H. W. Lischner,
101 – 107. J. A. C. Girone et al.: “Simultaneous
151. R. Glück, J. M. Hoskins, A. Wegmann, M. Just administration of measles-mumps-rubella
et al.: “Rubini, a new live attenuated mumps vaccine with booster doses of
vaccine virus strain for human diploid cells,” diphtheria-tetanus-pertussis and poliovirus
Dev. Biol. Stand. 65 (1986) 29 – 35. vaccines,” Pediatrics 81 (1988) 237 – 246.
152. WHO Expert Committee on biological 166. H. Peltola, O. P. Heinonen: “Frequency of true
standardization: WHO Tech. Rep. Ser. 760 adverse reactions to measles-mumps-rubella
(1987). vaccine,” Lancet i (1986) 939 – 942.
153. Recommendation of the Immunization 167. A. M. Arbeter, L. Baker, S. E. Starr, S. A.
Practices Advisory Committee (ACIP): Plotkin: “The combination measles, mumps,
“Mumps vaccine,” Morb. Mort. Wkly. Rep. 31 rubella and varicella vaccine in healthy
(1982) 617 – 625. children,” Dev. Biol. Stand. 65 (1986) 89 – 93.
72 Immunotherapy and Vaccines
197. Centers for Disease Control: “Human rabies vaccine. Efficacy trial in healthy children,” N.
despite treatment with rabies immune globulin Engl. J. Med. 310 (1984) 1409 – 1415.
and human diploid cell rabies vaccine – 211. A. A. Gershon, S. P. Steinberg, L. Gelb, and
Thailand,” Morb. Mort. Wkly. Rep. 36 (1987) The National Institute of Allergy and
759 – 765. Infectious Diseases Varicella Vaccine
198. Centers for Disease Control: “Systemic Collaborative Study Group: “Live attenuated
allergic reactions following immunization with varicella vaccine use in immunocompromised
human diploid cell rabies vaccine,” Morb. children and adults,” Pediatrics 78 (1986)
Mort. Wkly. Rep. 33 (1984) 185 – 187. Suppl., 757 – 762.
199. M. C. Swanson, E. Rosanoff, M. Gurwith, M. 212. Arthropod-borne and rodent-borne viral
Deitch et al.: “IgE and IgG antibodies to diseases: WHO Tech. Rep. Ser. 719 (1985).
β-propiolactone and human serum albumin 213. Viral haemorrhagic fevers: WHO Tech. Rep.
associated with urticarial reactions to rabies Ser. 721 (1985).
vaccine,” J. Infect. Dis. 155 (1987) 909 – 913. 214. K. C. Smithburn, C. Durieux, R. Koerber,
200. M. Granström, M. Eriksson, G. Edevåg: “A H. A. Penna et al.: Yellow Fever Vaccination,
sandwich ELISA for bovine serum in viral World Health Organization, Geneva 1956.
vaccines,” J. Biol. Stand. 15 (1987) 193 – 197. 215. J. C. Roche, A. Jouan, B. Brisou, R. Rodhain
201. B. Morein, K. Simons: “Subunit vaccines et al.: “Comparative clinical study of a new
against enveloped viruses; virosomes, micelles 17 D thermostable yellow fever vaccine,”
and other protein complexes,” Vaccine 3 Vaccine 4 (1986) Suppl., 163 – 165.
(1985) 83 – 93. 216. WHO Expert Committee on Biological
202. C. H. Stuart-Harris, G. C. Schild: “Influenza,” Standardization: WHO Tech. Rep. Ser. 594
The Viruses and Disease, E. Arnoldt Ltd, (1976) 23 – 49.
London 1976. 217. A. A. Smorodintsev, A. V. Dubov, V. I. Ilyenko,
203. P. F. Wright, D. T. Karzon: “Live Attenuated V. G. Platonov: “A new approach to
Influenza Vaccines,” Prog. Med. Virol. 34 development of live vaccine against tick-borne
(1987) 70 – 88. encephalitis,” J. Hyg. 67 (1969) 13 – 20.
204. WHO Expert Committee on Biological 218. C. Kuntz, F. X. Heinz, H. Hofmann:
Standardization: WHO Techn. Rep. Ser. 638 “Immunogenicity and reactogenicity of a
(1979). highly purified vaccine against tick-borne
205. Recommendations of the Immunization encephalitis,” J. Med. Virol. 6 (1980)
Practices Advisory Committee (ACIP): 103 – 109.
“Prevention and Control of Influenza,” Morb. 219. A. Oya in H. Fukumi (ed.): “Japanese
Mort. Wkly. Rep. 36 (1987) 373 – 387. encephalitis vaccine,” The vaccination. Theory
206. P. A. Gross, A. L. Gould, A. E. Brown: “Effect and practice, International Medical
of cancer chemotherapy on the immune Foundation of Japan (1975) pp. 69 – 82.
response to influenza virus vaccine: Review of 220. WHO Expert Committee on Biological
published studies,” Rev. Infect. Dis. 7 (1985) Standardization: WHO Tech. Rep. Ser. 323
613 – 618. (1966) 56 – 71.
221. Recommendation of the Immunization
207. M. A. Strassburg, S. Greenland, F. J. Sorvillo,
Practices Advisory Committee (ACIP):
L. E. Lieb et al.: “Influenza in the elderly:
Smallpox Vaccine Morb. Mort. Wkly. Rep. 34
report of an outbreak and a review of vaccine
(1985) 341 – 342.
effectiveness reports,” Vaccine 4 (1986)
222. R. Randall, L. N. Binn, V. R. Harrison:
38 – 44.
“Immunization against Rift Valley fever
208. J. M. Ostrove, G. Inchauspe: “The biology of
virus,” J. Immunol. 93 (1964) 293 – 299.
varicella-zoster virus” in: S. E. Straus
223. B. Niklasson: “Rift Valley fever virus vaccine
“Varicella-zoster virus infections: biology,
trial; Study of side effects in humans,” Scand.
natural history, treatment, and prevention,”
J. Infect. Dis. 14 (1982) 105 – 109.
Ann. Intern. Med. 108 (1988) 221 –237.
224. P. M. Kaye, Parasitol. Today 3 (1987)
209. M. Takahashi: “Clinical overview of varicella
293 – 299.
vaccine; Development and early studies,” 225. M. D. Rickard, M. J. Howell in E. E. R. Taylor,
Pediatrics 78 (1986) Suppl., 736 – 741. J. R. Baker (ed.): In Vitro Methods for Parasite
210. R. E. Weibel, B. J. Neff, B. J. Kuter, H. A. Cultivation, Academic Press, London 1987,
Guess et al.: “Live attenuated varicella virus pp. 407 – 451.
74 Immunotherapy and Vaccines
312. G. V. Brown, I. G. Culvenor, P. E. Crewther, 335. E. R. Stiehm, Pediatrics 63 (1979) 301 – 19.
A. E. Bianco et al., J. Exp. Med. 162 (1985) 336. J. A. Finlayson, in C. S. F. Easmon, J.
774 – 779. Jeljaszewics (eds.): Medical Microbiology,
313. B. Wahlin, M. Wahlgren, H. Perlamnn, H. vol. 1, Academic Press, London 1982,
Berzins et al., Proc. Natl. Acad. Sci. USA 81 pp. 129 – 82.
(1984) 7912 – 7916. 337. R. Heffron: Pneumonia, With Special
314. J. M. Favarolo, R. L. Coppel, L. M. Corcoran, Reference to Pneumococcus Lobar
S. J. Foote et al., Nucl. Acid. Res. 14 (1986) Pneumonia, The Commonwealth Fund, New
8265 –8277. York 1939.
315. W. E. Collins, R. F. Anders, M. Pappaioanou, 338. R. H. Rousell, M. S. Collins, M. B. Dobkin,
G. H. Campbell et al., Nature (London) 323 R. E. Louie et al., Am. J. Med. Suppl. 76
(1986) 259 – 262. (1984) 40 – 145.
316. S. A. Luse, L. H. Miller, Am. J. Trop. Med. 339. H. E. Alexander, Am. J. Dis. Child. 66 (1943)
Hyg. 20 (1971) 655 – 670. 172 – 187.
317. T. Triglia, H. D. Stahl, P. E. Crewther, D. 340. H. E. Alexander, Am. J. Dis. Child. 66 (1943)
Scanlon et al., EMBO J. 6 (1987) 1413 – 1419. 160 – 171.
318. M. Koenen, A. Scherf, O. Mercereau, G. 341. E. Merler, F. S. Rosen, N. Engl. J. Med. 275
Langsley et al., Nature (London) 311 (1984) (1966) 480 – 86; 536 – 542.
382 – 385. 342. C. A. Janeway, F. S. Rosen, N. Engl. J. med.
319. I. J. Udeinya, L. H. Miller, I. A. McGregor,
275 (1966) 826 – 831.
J. B. Jensen, Nature (London) 231 (1983)
343. P. A. M. Gross, D. Gitlin, C. A. Janeway, N.
429 – 431.
Engl. J. Med. 260 (1959) 170 – 178.
320. K. Marsh, R. J. Howard, Science 231 (1986)
344. J. Stokes, E. P. Maris, S. S. Gellis, J. Clin.
150 – 153.
321. S. Pollack, J. Fleming, Brit. J. Haematol. 58 Invest. 23 (1944) 531 – 540.
(1984) 289 – 294. 345. J. Stokes, J. R. Neefe, JAMA J. Am. Med.
322. M. E. Perkins, J. Exp. Med. 160 (1984) Assoc. 127 (1945) 531 – 540.
788 – 797. 346. W. M. Hammon, L. L. Coriell, J. Stokes et al.,
323. D. Camus, T. J. Hadley, Science 230 (1985) JAMA J. Am. Med. Assoc. 150 (1950) 739 – 49;
553 – 557. 750 – 756.
324. P. Delplace, B. Fortier, G. Tronchin, J. F. 347. C. A. Janeway, L. Apt, D. Gitlin, Trans. Assoc.
Dubremetz et al., Mol. Biochem. Parasitol. 23 Am. Physician 66 (1953) 200 – 202.
(1987) 193 – 202. 348. P. A. M. Gross, D. Gitlin, C. A. Janeway, N.
325. M. E. Patarroyo, R. Amador, P. Lavijo, A. Engl. J. Med. 260 (1959) 121 – 125.
Moreno et al., Nature (London) 332 (1988) 349. R. G. Hamilton, Clin. Chem. (Winston Salem
158 – 161. N.C.) 33 (1987) 1707 – 1725.
326. R. W. Gwadz, Science 193 (1976) 350. G. R. Siber, P. H. Schur, A. C. Weitzman, G.
1150 – 1153. Schiffman, N. Engl. J. Med. 303 (1980)
327. K. N. Mendis, G. A. T. Targett, Nature 178 – 182.
(London) 277 (1979) 289 – 391. 351. C. M. Roifman, H. M. Lederman, S. Lavi, L. D.
328. R. W. Gwadz, L. C. Koontz, Infect. Immun. 44 Stein et al., Am. J. Med. 79 (1985) 171 – 174.
(1984) 137 – 145 352. S. Barandun, H. Isliker, Vox Sang. 51 (1986)
329. N. Kumar, R. Carter, Mol. Biochem. Parasitol. 157 – 160.
13 (1984) 333 – 341. 353. A. Hassig, Vox Sang. 51 (1986) 10 – 17.
330. K. Mendis, P. Udayama, R. Carter, P. H. David, 354. D. D. Schroeder, M. L. Dumas, Am. J. Med.
J. Cell Biochem. Suppl. 10 A (1986) 149 – 156. Suppl. 76 (1984) 33 – 39.
331. A. N. Vermuelen, T. Ponnudurai, P. J. A. 355. C. A. Janeway, E. Merler, F. S. Rosen, S.
Beckers, J. P. Verhave et al., J. Exp. Med. 162 Salmon et al., N. Engl. J. Med. 278 (1968)
(1985) 1460 – 1476. 919 – 923.
332. J. Rener, P. M. Graves, R. Carter, J. L. Williams 356. J. Passwell, F. S. Rosen, E. Merler in B.
et al., J. Exp. Med. 158 (1983) 976 – 980. Alving, J. Finlayson (eds.): Immunoglobulins;
333. P. M. Graves, R. Carter, T. R. Burkot, J. Rener characteristics and uses of intravenous
et al., Infect. Immun. 48 (1985) 611 – 619. preparations, U.S.Dept. H.H.S., Washington
334. E. R. Stiehm, E. Ashida, K. S. Kim, D. J.
1979, pp. 139 – 142.
Winston et al., Ann. Intern. Med. 107 (1987)
367 – 82.
Immunotherapy and Vaccines 77
450. K. K. Christensen, P. Christensen, Pediatric 474. Morb. Mort. Wkly. Rep. 36 (1987) 728 – 729.
Infectious Diseases 5 (1986) S 189 – 192. 475. A. W. Burks, H. A. Sampson, R. H. Buckley,
451. J. G. Kelton, C. J. Carter, C. Rodger, G. N. Engl. J. Med. 314 (1986) 560 – 564.
Bebenek et al., Blood 63 (1984) 1434 – 1438. 476. R. Apfelzweig, D. Piskiewicz, J. A. Hooper, J.
452. A. S. Cross, B. M. Alving, J. C. Sadoff, P. Clin. Immunol. 7 (1987) 46 – 50.
Baldwin et al., Lancet 1 (1984) 912. 477. R. Paganelli, I. Quinti, G. P. D’Offizi, C.
453. M. Berger, P. Rosencranz, C. Y. Brown, Clin. Papetti et al., Vox Sang. 51 (1986) 87 – 91.
Immunol. Immunopathol. 34 (1985) 227 – 236. 478. J. C. Barton, G. A. Herrera, J. H. Galla, L. F.
454. J. Kulics, E. Rajnavolgyi, G. Fust, J. Gergely, Bertoli et al., Am. J. Med. 82 (1987) 624 – 629.
Mol. Immunol. 20 (1983) 805 – 810. 479. A. G. Brox, D. Cournoyer, M. Sternbach, G.
455. E. F. Ellis, C. S. Henney, J. Allergy 43 (1969) Spurll, Am. J. Med. 82 (1987) 633 – 635.
45 – 54. 480. R. Schroff, K. Foon, S. Beatty, R. Oldham et
456. R. C. Williams, O. J. Mellbye, G. Kronvall, al.: “Human anti-murine immunoglobulin
Infect. Immun. 6 (1972) 316 – 323. responses in patients receiving monoclonal
457. D. S. Feingold, F. Oski, Arch. Intern. Med. antibody therapy,” Can. Res. 45 (1985)
116 (1965) 226 – 228. 879 – 885.
458. R. C. Millican, J. D. Rust, J. Infect. Dis. 107 481. M. Nose, H. Wigzell: “Biological significance
(1960) 389 – 394. of carbohydrate chains on monoclonal
459. E. Lachman, S. B. Pitsoe, S. L. Gaffin, Lancet antibodies,” Proc. Natl. Acad. Sci. U.S.A. 80
1 (1984) 981 – 983. (1983) 6632 – 6636.
460. M. A. Thomson, J. S. Afr. Vet. Assoc. 54 482. E. Englemann, S. Foung, J. Larrick, A
(1983) 279 – 281. Raubitschek (eds.): Human Hybridomas and
461. M. L. Forman, R. Stiehm, N. Engl. J. Med. Monoclonal Antibodies, Plenum Press, New
281 (1969) 926 – 931. York 1985.
462. A. S. Shigeoka, R. T. Hall, H. R. Hill, Lancet 1 483. A. Strelkelkaus (ed.): Human Hybridomas;
(1978) 636 – 638. Diagnostic and Therapeutic Applications,
463. E. J. Ziegler, J. A. McCutchan, J. Fierer, M. P. Marcel Dekker, New York 1986.
Glauser et al., N. Engl. J. Med. 307 (1982) 484. D. Kozbor, J. Roder, T. Chang, Z. Stplewski et
1225 – 1230. al.: “Human antitetanus toxoid monoclonal
464. J.-D. Baumgartner, M. P. Glauser, J. A. antibody secreted by EBV-transformed human
McCutchan, E. J. Ziegler et al., Lancet 2 B cells fused with murine myeloma,”
(1985) 59 –63. Hybridoma 1 (1982) 323 – 328.
465. T. Calandra, J. Schellekens, J. Verhoef, M. P. 485. J. Larrick, K. Truitt, A. Raubitschek, G. Senyk
Glauser, Program and Abstracts for the 4th et al., “Characterization of human hybridomas
International Symposium on Infections in the secreting antibody in tetanus toxoid,” Proc.
Immunocompromised Host, Ronneby Brunn, Natl. Acad. Sci. U.S.A. 80 (1983) 6376 – 6380.
Sweden, 1986. Abstract No. 128. 486. N. Chiorazzi, R. Wasserman, H. Kunkel: “Use
466. A. H. Levy, J. Chronic Dis. 15 (1962) of Epstein-Barr virus-transformed B cell lines
589 – 598. for the generation of
467. J. Bjorkander, C. Cunningham-Rundles, P.
immunoglobulin-producing human B cell
Lundin, R. Olsson et al., Am. J. Med. 84
hybridomas,” J. Exp. Med. 156 (1982)
(1988) 107 – 111.
930 – 935.
468. R. S. Lane, Lancet 2 (1983) 974 – 975.
487. F. Gigliotti, R. Insel: “Protective human
469. H. D. Ochs, S. H. Fischer, F. S. Virant, M. L.
hybridoma antibody to tetanus toxin,” J. Clin.
Lee et al., Lancet 1 (1985) 404 – 405.
Invest. 70 (1982) 1306 – 1309.
470. C. L. S. Leen, P. L. Yap, G. Neill, D. B. L.
488. L. Olsson, T. Mazauric, J. Vincent-Falquet, J.
McClelland et al., Vox Sang. 50 (1986)
Armand: “A human monoclonal antibody
26 – 32.
specific for tetanus toxoid,” Dev. Biol. Stand.
471. B. Cuthbertson, R. J. Perry, P. R. Foster, K. G.
57(1984) 87 – 91.
Reid, J. Infect. 15 (1987) 125 – 133.
489. S. Tsuchiya, S. Yokoyama, O. Yoshie, Y. Ono:
472. C. C. Wood, A. E. Williams, J. G. McNamara,
“Production fo diphtheria antitoxin antibody in
J. A. Annunziata et al., Ann. Intern. Med. 105
Epstein-Barr virus induced lymphoblastoid
(1986) 536 – 538.
473. Y. Hamamoto, S. Harada, N. Yamamoto, Y. cell lines.” J. Immunol. 124 (1980)
Uemura et al., Vox Sang. 53 (1987) 65 – 69. 1970 – 1976.
80 Immunotherapy and Vaccines
490. E. Ziegler, J. McCutchan, J. Fierer et al.: 499. H. Peltola, H. Kayhty, M. Virtanen, P. Makela:
“Treatment of Gram-negative bacteremia and “Prevention of Hemophilus influenza type b
shock with human anti-serum to a mutant bacteremic infections with the capsular
Escherichia coli,” N. Engl. J. Med. 307 (1982) polysaccharide vaccine,” N. Engl. J. Med. 310
1225 – 1230. (1984) 1561 –1565.
491. W. Bogard, E. Hornberger, P. Kung in E. 500. T. Atlaw, D. Kozbor, J. Roder: “Human
Englemann, S. Foung, J. Larrick, A. monoclonal antibodies against Mycobacterium
Raubitscheck (eds.): “Production and Leprae,” Infect. Immun. 49 (1985) 104 – 110.
characterization of human monoclonal 501. B. Brodeur, L. Lagace, Y. Larose, M. Martin et
antibodies against Gram-negative bacteria,” al. in B. Schook (ed.): “Mouse-human
Human Hybridomas and Monoclonal myeloma partner for the production of
Antibodies, Plenum Press, New York 1985, heterohybridomas,” Monoclonal Antibodies ,
pp. 95 – 112. Marcel Dekker, 1986, p. 51.
492. N. Teng, H. Kaplan, J. Hebert et al.: 502. M. Steinitz, S. Tamir, A. Goldfarb: “Human
“Protection against Gram-negative bacteremia anti-pneumococci antibody produced by an
and endotoxemia with human monoclonal IgM Epstein-Barr virus (EBV)-immortalized cell
antibodies,” Proc. Natl. Acad. Sci. U.S.A. 82 line,” J. Immunol. 132 (1984) 877 – 882.
(1985) 179 – 184. 503. J. Schwaber, M. Posner, S. Schlossman, H.
493. J. Larrick, M. Jahnsen, G. Senyk, S. Weiss et Lazarus: “Human-human hybrids secreting
al. in H. Friedman (ed.): Production of human pneumococcal antibodies,” Hum. Immunol. 9
monoclonal antibodies recognizing (1984) 137 – 142.
cross-reactive determinants on 504. D. Emanuel, J. Gold, J. Colacino, C. Lopez et
lipopolysaccharides,” The Immunology and al.: “A human monoclonal antibody to
Immunopharmacology of Endotoxins, 1986, cytomegalovirus (CMV),” J. Immunol. 133
pp. 75 – 81. (1984) 2202 – 2205.
494. M. Pollack, A. Raubitschek, J. Larrick: 505. C. Amadei, S. Michelson, J. Frot, M. Fruchart
“Cross-reactive human monoclonal antibodies et al.: “Human anticytomegalovirus (CMV)
that recognize conserved epitopes in the immunoglobulins secreted by
core-lipid A region of endotoxin,” J. Clin. EBV-transformed B-lymphocytes cell lines,”
Invest. 79 (1987) 1421 – 1430. Dev. Biol. Stand. 57 (1984) 283 – 286.
495. J. Baumgartner, M. Glauser, J. McCutchan, E. 506. K. Burnett, J. Leung, J. Marinis in E.
Zeigler et al.: “Prevention of gramnegative Engleman, S. Foung, J. Larrick, A.
shock and death in surgical patients by Raubitschek (eds.): “Human monoclonal
antibody to endotoxin core glycolipid.” Lancet antibodies to defined antigens,” Human
2 (1985) 59 – 63. Hybridomas and Monoclonal Antibodies,
496. J. Larrick, S. Hart, D. Lippman, M. Plenum Press, New York 1985, pp. 113 – 133.
Glembourtt et al. in A. Strelkelkaus (ed.): 507. Y. Ichimori, K. Sasano, H. Itoh, S.
“Generation and characterization of human Hitosumachi et al.: “Establishment of
monoclonal anti-Pseudomonas aeruginosa hybridomas secreting human monoclonal
antibodies,” Human Hybridomas; Diagnostic antibodies against tetanus toxin and hepatitis B
and Therapeutic Applications, Marcel Dekker, virus surface antigen,” Biochem. Biophys. Res.
New York 1986, pp. 65 – 80. Commun. 129 (1985) 26 – 38.
497. J. Larrick, B. Dyer, G. Senyk et al. in E. 508. E. Stricker, R. Tiebout, P. Lelie, W.
Engleman, S. Foung, J. Larrick, A. Zeijlemaker: “A human monoclonal IgG
Raubitschek (eds.): “In vitro expression of antihepatitis B surface antibody; production,
human B cells for the production of human properties and applications,” Scand. J.
monoclonal antibodies,” Human Hybridomas Immunol. 22 (1985) 337 – 345.
and Monoclonal Antibodies, Plenum Press, 509. R. Beasley et al.: “Hepatitis B immune
New York 1985, pp. 149 – 165. globulin (HBIG) efficacy in the interruption of
498. K. Hunter, Jr., G. Fischer, V. Hemming, S. perinatal transmission of hepatitis B virus
Wilson et al.: “Antibacterial activity of a carrier state,” Lancet 2 (1981) 388 – 393.
human monoclonal antibody to Haemophilus 510. S. Foung et al.: “Human monoclonal
influenzae type B capsular polysaccharide,” antibodies neutralizing Varicella Zoster virus,”
Lancet 2 (1982) 789 – 799. J. Infect. Dis. 52 (1985) 280 – 285.
Immunotherapy and Vaccines 81
My Profile Log
HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry
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
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.
⊕ = 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.
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.
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.
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.
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.
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.
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]
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
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
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)
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
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.
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.
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 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)
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)
5.8 – 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)
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)
9.4 – 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.
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
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
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.
page 8 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
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
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).
page 10 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Methyl Green (0.1 % in ethanol) 2
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.
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
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
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].
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
[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].
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).
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)
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.
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
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.
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:
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
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].
Alizarin Complexone [3952-78-1] Ba, Ca, Cd, Co, Hg, In, fluoride
Pb, Sr, Zn
Alizarin S [130-22-3] Al, lanthanides, Mo, Sc, Al, lanthanides, Sc, Th, Y,
Th, Y, Zr Zr
Arsenazo III [62337-00-2] lanthanides, Th, Y lanthanides, Sc, Th, U, Y,
Zr
Aurintricarboxylic acid
(Aluminon, C.I. 43 810) [4431-00-9] Al, Ca, Fe, Mg Al, Be, Fe, Ga,
lanthanides, Pd, Y
2,2′-Bipyridine [366-18-7] Ag, Cd, Cu, Fe, Mo
Bromopyrogallol Red [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
Calcon (Eriochrome Blue
Black R) [2538-85-4] Ca, Cd, Mg, Mn, Zn
Calconcarboxylic acid [3737-95-9] Ca
Chrome Azurol S (C.I. [1667-99-8] Al, Ca, Cu, Fe, Al, Be, Bi, Cr, Cu, Fe, Ga,
43 825) lanthanides, Mg, Ni, Th, In, lanthanides, Mg, Mn,
U, V, Y, Zr Ni, Pb, Pd, Sc, Sn, Th, Y,
Zr
Chromotropic acid, disodium
salt [5808-22-0] Ti
Cuprizone [370-81-0] Cu
5-(4- [536-17-4] Ag, Au, Cu, Hg, Pd
Dimethylaminobenzylidene)
rhodanine
page 19 of 31
Indicator Reagents : Ullmann's Encyclopedia of Industrial Chemistry : Wiley InterScience
Dimethylglyoxime [95-45-4] Co, Fe, Ni, Pd
1,5-Diphenylcarbazide [140-22-7] Hg, Ni, Pb, V Cr, Cu, Hg
Dithizone [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
Eriochrome Black T [1787-61-7] Ba, Ca, Cd, Hg, In,
lanthanides, Mg, Mn, Pb,
Zn, Zr
Eriochrome Blue SE [1058-92-0] Ca, Cd, Mg, Mn, Ni, Pb,
Zn
Eriochrome Blue Black B [3564-14-5] Ca, Cd, Mg, Mn, U, Zn, Zr
Eriochrome Cyanine R
(C.I. 43 810) [3564-18-9] Al, Ca, Cu, Fe, Mg, Th, Zr Al, Be, Cr, Ga, Mg, Th, Tl,
Zr
Fluorescein Complexone
(Calcein) [1461-15-0] Ba, Ca, Cr, Cu, Mg, Sr Ca
Glyoxalbis(2-hydroxyanil) [1149-16-2] Ca, Sc, U
Hematoxylin [517-28-2] Al, Bi, Cu, Th, Zr Sn
8-Hydroxyquinoline [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
2-Mercaptobenzothiazole [149-30-4] Au, Cd, Cu, Pb, Pd
Methylthymol Blue [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
Murexide [3051-09-0] Ag, Ca, Co, Cu, Mn, Ni, Ca, Cu, Sc
Sc, Th, Zn
1-Nitroso-2-naphthol [131-91-9] Co, Fe, Mo, Pd
2-Nitroso-1-naphthol [132-53-6] Co, Pd
Nitroso-R-salt [525-05-3] Cu, Ni Co, Fe, Pd
1,10-Phenanthroline [66-71-7] Ag, Cu, Fe, Pd, V, Zn
Phenylfluorone [975-17-7] Ge, Mo, Nb, Sb, Sn, Ta, U
Phthalein Purple Ba, Ca, Cd, Mg, Sr Ba, Ca, Mg, Mo
1-(2-Pyridylazo)-2-naphthol [85-85-8] Al, Bi, Cd, Co, Cu, Fe, Ag, Bi, Cd, Co, Fe, Ga,
(PAN) Ga, Hg, In, Mn, Ni, Pb, lanthanides, Pd, Sc, U, V,
Th, Tl, U, V, Zn Y, Zn, Zr
4-(2-Pyridylazo)resorcinol [1141-59-9] Al, Bi, Cd, Cu, Fe, Ga, Ag, Bi, Cd, Co, Cr, Cu, Fe,
(PAR) Hg, In, lanthanides, Mn, Ga, Hg, La, Mn, Nb, Ni,
Ni, Pb, Sr, Th, Tl, Zn Pb, Pd, Sc, U, V, Y, Zn, Zr
Pyrocatechol Violet [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
Pyrogallol Red [32638-88-3] Bi, Co, Ni, Pb Ag, Mo
Sulfonazo III [1738-02-9] Ba, Ba
5-Sulfosalicylic acid [5965-83-3] Fe, Tl Fe, Ti
4-(2-Thiazolylazo)resorcinol [2246-46-0] Co, Cu, lanthanides, Ni, Tl as for PAR
(TAR)
Thorin [3688-92-4] Bi, Sc, Th, U, Y lanthanides
Thymolphthalexon [1913-93-5] Ag, Ba, Ca, Cr, Mn, Sr Ba, Ca, lanthanides, Mg,
Mo, Sr
Tiron [149-45-1] Fe Fe, lanthanides, Mo, Ti, U,
W
Toluene-3,4-dithiol [496-74-2] Ge, Mo, Sn, W
Xylenol Orange [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
Zincon [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.
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.
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.
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
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):
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].
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.
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
Alizarin Complexone
Fluorescein Complexone
[Top of Page]
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. 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.
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:
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:
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
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
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
[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]
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]
[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]
[Top of Page]
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.
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]
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]
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]
page 10 of 10
Indium and Indium Compounds 1
dium content of 2 g/L by double neutralization [16]. The flow sheet for the process is illus-
with calcium carbonate. The enriched solution trated in Figure 3. A mixture of fumes, coke,
is then subjected to solvent extraction followed and antimony-containing slag is melted in a
by back extraction with hydrochloric acid. The direct-fired rotary furnace. Arsenic is removed
aqueous solution is concentrated to 20 – 30 g/L by adding scrap iron to form iron speiss. The
indium. Indium is recovered by immersing alu- residual tin – indium – antimony – lead alloy is
minum plates in the solution. Indium sponge ad- cast into anodes and electrolyzed in a fluorosil-
hering to the aluminum plates is removed for icate solution, giving a solder cathode (Sn – Pb
melting and casting into anodes. Indium metal alloy) and indium – antimony anode slimes. The
of 99.99 % purity is then obtained by electrore- slimes are roasted with sulfuric acid. Indium is
fining. dissolved from the sulfated product by leaching
with water. Copper is removed from the indium
solution by cementation on indium sheets. In-
4.2. Production from Lead Circuits dium is then recovered from the solution as a
sponge by addition of aluminum or zinc. The
At Cominco’s integrated lead – zinc smelter at indium sponge is melted and electrorefined to
Trail in Canada, tin and indium accumulate in give indium of 99.99 % purity.
the fumes of the continuous drossing furnace
Indium and Indium Compounds 5
∗ Sublimes
Fusible indium alloys are usually based on They are also used as temperature indicators
bismuth in combination with lead, cadmium, tin, in situations where other methods of tempera-
and indium. ture measurement are impracticable. Small rods
Fusible indium alloys are used to bend thin- of alloys of known melting point are inserted
walled tubes without wrinkling the wall or into equipment such as aircraft header tanks, test
changing the original cross section. The alloy bearings, and experimental rigs. The melting of
is cast in the tube, rendering it effectively solid. the alloy thus indicates the temperature attained
After bending, the alloy can be melted out using in particular parts of the test system.
hot water. Addition of indium to lead – tin bearings for
Table 4. Estimated western-world consumption of indium in 1985
heavy-duty and high-speed applications pro-
[9] vides particularly high resistance to fatigue and
Use 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-
Low-melting alloys 12
Bearings 9
gines.
Dental alloys 7 The addition of indium to gold dental al-
Nuclear reactor control rods 3.3 loys improves their mechanical characteristics
Low-pressure sodium lamps 4–5
and increases resistance to discoloring. Small
Electrical contacts 8
Alkaline dry batteries 6.8 ∗ amounts of indium are used to improve the
Phosphors 6 ∗∗ machinability of gold alloys for jewelry.
Semiconductors Indium is used in nuclear reactor control-rod
Lasers, photodetectors, integrated
circuits (Ga – As In – P, In – P) 5
alloys (80 % Ag, 15 % In, and 5 % Cd). The con-
CCDs in infra-red video cameras sumption of indium in this field is estimated at
(image orthicons, In – Sb) 3.5 3.3 t/a for the replacement of control rods in ex-
Liquid crystal displays 7 isting nuclear power plants [9].
∗ 1987 data [40]. In low pressure sodium lamps ( SOX lamps),
∗∗ 1986 data for Japan [41]. indium is applied as a thin layer of indium – tin
oxide on the inside of the outer glass envelope.
These alloys are also used in fire-control sys- The indium – tin oxide film increases the operat-
tems. Restraining links that hold alarm, water ing temperature and the efficiency of the lamp.
valve, and door operating mechanisms in place Low pressure sodium lamps, characterized by
are soldered with a low-melting alloy. A rise in their orange color, are mainly used in the United
temperature sufficient to melt the alloy results in Kingdom, Belgium, and the Netherlands. The
the system coming into operation. production of SOX lamps accounts for 4 – 5 t/a
Fusible alloys are used for making disposable indium [9].
patterns and dies for the founding of ferrous and Indium is now used as a corrosion inhibitor in
non-ferrous metals. 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
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
5. Uses
My Profile Log
HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry
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
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.
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.
(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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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].
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.
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.
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.
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.
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.
The total values of media delivered worldwide during 1997 are given below. Figures are in $ × 109.
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
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
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.
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] ).
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.
Red Book defines the physical properties of the CD Audio (Aug 1995, Red Book);
(CD Audio) compact disc and the digital audio encoding: subcode/control and display system
z audio specification for 16-bit PCM. (Nov 1991, extension to Red Book for
z disc specification, including physical CD Graphics); CD Text (Sep 1996,
parameters. extension to Red Book)
z optical stylus and parameters.
z deviations and block error rate. the Red Book is also available as the
z modulation system and error international standard ISO/IEC 60908
correction.
z control and display system and CD
graphics
allows for up to 74 min of digital sound.
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
directory structure
z image data coding including a
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
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
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.)
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.
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.
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 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.
Parameter CD DVD
* 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).
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.
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
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.
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.
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.
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.
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).
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
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.
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.
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.
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).
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.
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.
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.
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.
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.
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:
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.
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
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.
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.
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.
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)
C) Nickel negative or father stamper after removal of photoresist, adhesive layer and silver
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.
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].
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]
Optical
Light transmission, %
630 nm ≥90 ≥90
300 – 380 nm ≥80 a
Thickness tolerance, µm ± 20 c ± 50 d
Distortion, degrees ± 0.4 ± 0.6
Thermal
Vicat softening temperature, °C 96 – 107 145
a For photopolymerization.
b Parallel light beam.
c Disc diameter 305 mm.
d Disc diameter 120 mm.
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.
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).
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%
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)
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.
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.
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].
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?
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.
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.
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].
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
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].
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.
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.
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.
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.
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]:
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.
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
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].
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.
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].
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.
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]
page 62 of 62
Infrared and Raman Spectroscopy 1
However, the appearance of the first research- [5]; Koningstein [6]; and Long [7]. The tech-
grade Fourier transform infrared (FT-IR) spec- nique and applications of FT-IR spectroscopy
trometers in the early 1970s initiated a renais- are discussed in detail by Griffiths and de
sance of infrared spectrometry. After analytical Haseth [8], and by Mackenzie [9]. An in-
instruments (since the late 1970s) and routine troduction to infrared and Raman spectroscopy,
instruments (since the mid 1980s), dedicated in- dealing with theoretical as well as experimen-
struments are now available at reasonable prices. tal and interpretational aspects, is provided by
With its fundamental multiplex or Fellgett’s ad- Colthup, Daly, and Wiberley [10] and by
vantage and throughput or Jacquinot’s advan- Schrader [11]. Books about Raman spec-
tage, FT-IR offers a versatility of approach to troscopy in particular are those of Kohlrausch
measurement problems often superior to other [12]; Baranska, Labudzinska, and Terpinski
techniques. Furthermore, FT-IR is capable of ex- [13]; and Hendra, Jones, and Warnes [14].
tracting from samples information that is diffi- Practical FT-IR and Raman spectroscopy is co-
cult to obtain or even inaccessible for nuclear vered by Ferraro and Krishnan [15]; Pel-
magnetic resonance and mass spectrometry. Ap- letier [16]; and Grasselli and Bulkin [17].
plications of modern FT-IR spectrometry in- For the interpretation of IR and Raman spectra
clude simple, routine identity and purity exami- of organic compounds, see Socrates [18]; Bel-
nations (quality control) as well as quantitative lamy [19], [20]; Dollish, Fateley, and Bent-
analysis, process measurements, the identifica- ley [21], and Lin-Vien, Colthup, Fateley,
tion of unknown compounds, and the investiga- and Grasselli [22]. Vibrational spectroscopy
tion of biological materials. of inorganic and coordination compounds is
Raman and infrared spectra give images of described by Nakamoto [23]. Biological ap-
molecular vibrations which complement each plications of IR and Raman spectroscopy are
other; i.e., the combined evaluation of both covered by Gremlich and Yan [24]. Various
spectra yields more information about molec- collections of vibrational spectra are available.
ular structure than when they are evaluated Both Raman and IR spectra of about 1000 or-
separately. The 1990s witnessed a tremendous ganic compounds are offered as an atlas by
growth in the capabilities and utilization of Schrader [25]. The most comprehensive refer-
Raman scattering measurements. The parallel ence-spectra libraries, digital and hardcopy, can
growth of Fourier transform Raman (FT-Raman) be obtained from Sadtler [26]. A book on mod-
with charge-coupled device (CCD) based dis- ern techniques in applied optical spectroscopy
persive instrumentation has provided the spec- was edited by Mirabella [27].
troscopist with a wide range of choice in instru-
mentation. There are clear strengths and weak-
nesses that accompany both types of instrumen- 2. Techniques
tation. While FT-Raman systems virtually elimi-
nate fluorescence by using near-IR laser sources, 2.1. Fourier Transform Infrared
this advantage is paid for with limited sensitiv-
ity. On the other hand, fluorescence interferences
Technique
always play a role in dispersive measurements
The basis of Fourier transform infrared (FT-
with higher sensitivity.
IR) spectroscopy is the two-beam interferom-
As an alternative to infrared spectroscopy,
eter, designed by Michelson in 1891 [28], and
Raman spectroscopy can be easier to use in some
shown schematically in Figure 1A. Broadband
cases; for example, whereas water and glass are
infrared radiation is emitted by a thermal source
strong infrared absorbers they are weak Raman
(globar, metal strips, Nernst glower) and falls
scatterers, so that it is easy to produce a good-
onto a beam splitter which, in the ideal case,
quality Raman spectrum of an aqueous sample
transmits half the radiation and reflects the other
in a glass container.
half. The reflected half, after traversing a dis-
General References. Details of vibrational tance L, falls onto a fixed mirror M1. The ra-
theory are given in the books by Herzberg [2], diation is reflected by M1 and, after travers-
[3]; Wilson, Decius, and Cross [4]; Steele 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 spectrometer, all frequencies emitted by the IR
of a perfectly aligned Michelson interferome- source reach the detector simultaneously, which
ter has the shape of a (sin x)/x or sinc x func- results in considerable time saving and a large
tion (Fig. 2), exhibiting strong side lobes or feet, signal-to-noise ratio advantage over dispersive
which may even have negative intensities. To instruments; this is the most important advan-
attenuate these spurious feet, so-called apodiza- tage, known as the multiplex or Fellgett’s ad-
tion functions are used as weighting functions vantage [8]. A further advantage stems from
for the interferogram. Numerous such functions the fact that the circular apertures employed in
exist (e.g., triangular, Happ – Genzel, or Black- FT-IR spectrometers have larger surface areas
man – Harris) which produce instrumental line than the linear slits of grating spectrometers and
shapes with lower intensity side lobes, but also hence permit radiation throughputs at least six
with broader main lobes than the sinc function. times greater; this is known as the throughput
Hence, side-lobe suppression is only possible at or Jacquinot’s advantage [8]. In FT-IR spec-
the cost of resolution, because the full width at troscopy, the measuring time is the time required
half-height of the ILS defines the best resolu- to move the mirror M2 the distance necessary
tion achievable with a given apodization func- to give the desired resolution; since the mirror
tion [31]. In contrast to dispersive spectroscopy, can be moved very rapidly, complete spectra can
where the triangular apodization function is de- be obtained within fractions of a second. By
termined by the slit of the grating spectrom- contrast, the measuring time for a conventional
eter, in FT-IR spectroscopy, a choice between spectrum is on the order of minutes. In an FT
highest resolution (no apodization) and best side spectrum, the accuracy of each wavenumber is
lobe suppression (Blackman – Harris apodiza- coupled to the accuracy with which the position
tion function) is possible. 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.
Figure 2. The sinc x function: instrumental line shape func- 2.2. Dispersive and Fourier Transform
tion of a perfectly aligned Michelson interferometer with no Raman Techniques
apodization
ν̃ = wavenumber, cm−1 ; ∆ = maximum retardation, cm
In modern dispersive Raman spectrometers, ar-
In comparison with conventional spec- rays of detector elements such as photodiode
troscopy, the FT-IR method possesses significant arrays or charge-coupled devices (CCDs) are
advantages. In grating spectrometers, the spec- arranged in a polychromator. Here, each ele-
trum is measured directly by registering the radi- ment records a different spectral band at the
ation intensity as a function of the continuously same time, thus making use of the multichannel
changing wavelength given by a monochromator advantage. In dispersive instruments, the very
[10]. Depending on the spectral resolution cho- strong radiation at and near the exciting line, the
sen, only a very small part (in practice less than Rayleigh radiation, may produce stray radiation
0.1 %) of the radiation present in the monochro- in the entire spectrum with an intensity much
mator reaches the IR detector. In the FT-IR higher than that of the Raman lines. In order
to remove this unwanted radiation, line filters
Infrared and Raman Spectroscopy 5
in the form of so-called notch filters or holo- with six to ten orders of magnitude higher inten-
graphic filters are used which specifically reflect sity than the Raman signal. The shot noise of
the Rayleigh line. In conventional, dispersive the Rayleigh line and Rayleigh wings produces
Raman spectroscopy visible lasers are usually a background noise which would completely
used to excite the Raman effect (see Section 3.1), mask all Raman lines. Therefore, an effective fil-
while in FT-Raman spectroscopy near-IR lasers ter that removes the Rayleigh line and Rayleigh
are employed. Excitation with visible lasers of- wings in the range 750 – 100 cm−1 is one of the
ten produces strongly interfering fluorescence most important components of a powerful FT-
originating from the sample or impurities; how- Raman instrument [32]. In addition, optimiza-
ever, this is eliminated by using a neodymium tion of the sampling technique is indispensable
yttrium aluminum garnet (Nd : YAG) laser pro- in order to convert the flux of exciting radiation
viding a discrete line excitation at 1.064 µm, as into a maximum flux of Raman radiation to be
at this wavelength in the near-IR region no elec- received by the optical system of the spectrom-
tronic transitions are induced. eter [34]. The 180◦ backscattering arrangement
An FT-Raman spectrometer is shown in sche- (Fig. 3B), rather than the 90◦ arrangement used
matic form in Figure 3A. The beam of the laser in visible, dispersive Raman technique, is un-
source passes to the sample compartment where doubtedly the most efficient way of collecting
it is focused, typically to a spot of 100 µm di- the Raman signal and provides the best quality
ameter, onto the sample, either in 90◦ or 180◦ data in the near-IR FT-Raman experiment.
geometry (Fig. 3B). The scattered Raman light
is collected by an aspherical lens and passed
through a filter module to remove the Rayleigh 3. Basic Principles of Vibrational
line and Rayleigh wings [32]. It is then directed
through the infrared input port into the interfer-
Spectroscopy
ometer which is optimized for the near-IR region
as is the liquid-nitrogen-cooled germanium or 3.1. The Vibrational Spectrum
room-temperature indium gallium arsenide (In-
GaAs) detector [33]. The procedure to obtain the To explain the origins of a vibrational spectrum
FT-Raman spectrum from the interferogram, the the vibration of a diatomic molecule is consid-
signal registered by the detector, is essentially ered first. This can be illustrated by a molecular
the same as that described above for an FT-IR model in which the atomic nuclei are represented
spectrum. by two point masses m1 and m2 , and the inter-
Before FT-Raman spectrometers could be atomic bond by a massless spring. According
successfully used several difficulties had to be to Hooke’s law, the vibrational frequency ν (in
overcome. As the intensity of Raman scattering s−1 ) determined by classical methods is given
is proportional to the fourth power of the fre- by
quency of excitation a loss in sensitivity by a
1 f
factor between 20 and 90 must be compensated ν= (1)
for when the exciting radiation is shifted from 2π µ
the visible to the near-IR region [32]. Moreover, where f is the force constant of the spring in
the noise equivalent power (NEP) of near-IR de- N/m, and µ is the reduced mass in kg:
tectors is greater by about one to two orders of
magnitude than that of a photomultiplier tube, m1 m2
µ= (2)
the usual detector for conventional Raman spec- m1 +m2
troscopy. However, both disadvantages are com- Thus, the vibrational frequency is higher
pensated for by the throughput (Jacquinot) and when the force constant is higher, i.e., when
multiplex (Fellgett) advantages of Fourier trans- the bonding of the two atoms is stronger. Con-
form spectroscopy (see Section 2.1). Another versely, the heavier the vibrating masses are, the
major problem is the fact that Raman scatter- smaller is the frequency ν. The frequency of fun-
ing is always accompanied by very strong ra- damental vibrations is on the order of magnitude
diation, the Rayleigh line and Rayleigh wings, of 1013 s−1 . Quantum-mechanical methods give
occurring at and near the frequency of excitation
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.
Infrared and Raman Spectroscopy 7
Therefore, the transition energy decreases with which the wavenumber is measured (the refrac-
increasing vibrational quantum number until tive index of air is 1.0003), and λ is the wave-
the molecule finally dissociates. The quantum length in centimeters. In the infrared region of
mechanical energy of a diatomic inharmonic os- the electromagnetic spectrum the practical unit
cillator is given in good approximation by: of wavelength is 10−4 cm or 10−6 m (i.e., µm),
2
and wavenumber and wavelength are related as
1 1 follows:
Evib =hν0 v+ −x0 v+ (6)
2 2
104
ν̃/cm−1 = (8)
where x 0 is the inharmonicity constant. λ/µm
A complex molecule is considered as a sys- In infrared spectroscopy, both wavenumber
tem of coupled inharmonic oscillators. If there and wavelength are used, whereas in Raman
are N atomic nuclei in the molecule, there will be spectroscopy only wavenumber is employed.
a total of 3 N degrees of freedom of motion for all The infrared and microwave regions of the
the nuclear masses in the molecule. Subtracting electromagnetic spectrum are shown in Table 1.
the pure translations and rotations of the entire In the near-infrared region, overtones and com-
molecule leaves 3 N − 6 internal degrees of free- bination bands are observed. Since these transi-
dom for a nonlinear molecule and 3 N − 5 inter- tions are only possible because of inharmonicity,
nal degrees of freedom for a linear molecule. The their probability is diminished. A good rule of
internal degrees of freedom correspond to the thumb is that the intensity of the first overtone
number of independent normal modes of vibra- is about one-tenth that of the corresponding fun-
tion; their forms and frequencies must be calcu- damental vibration. The fundamentals occur in
lated mathematically [10]. In each normal mode the mid-infrared range, which is the most im-
of vibration all the atoms of the molecule vibrate portant infrared region. In the far-infrared re-
with the same frequency and pass through their gion, the fundamentals of heavy, single-bonded
equilibrium positions simultaneously. The true atoms and the absorptions of inorganic coordi-
internal motions of the molecule, which consti- nation compounds [23] are found.
tute its vibrational spectrum, are composed of Infrared spectroscopy is based on the interac-
the normal vibrations as a coupled system of tion of an oscillating electromagnetic field with
these independent inharmonic oscillators. Thus, a molecule, and it is only possible if the dipole
a molecule is unambiguously characterized by moment of the molecule changes as a result of a
its vibrational spectrum. Real-world molecules molecular vibration. While the absorption fre-
are best described by the model of inharmonic quency depends on the molecular vibrational
oscillators because here transitions are allowed frequency, the absorption intensity depends on
which are forbidden to the harmonic oscilla- how effectively the infrared photon energy is
tor: transitions from the ground state (v = 0) to transferred to the molecule. It can be shown [3]
states with v = 2, 3, . . . (overtones), transitions that the intensity of infrared absorption is pro-
from an excited level of a vibration (difference portional to the square of the change in the dipole
bands), and transitions between states belonging moment µ, with respect to the change in the nor-
to different types of normal mode (combination mal coordinate q describing the corresponding
bands). molecular vibration:
In vibrational spectroscopy, instead of the fre- 2
quency ν (in s−1 ) the so-called wavenumber ν̃ IIR ∼
∂µ
(9)
expressed in cm−1 (reciprocal centimeters) is ∂q
mostly used; this signifies the number of waves A second way to induce molecular vibrations
in a 1 cm wavetrain: is to irradiate a sample with an intense source
ν 1
of monochromatic radiation, usually in the visi-
ν̃= = (7) ble or near-infrared region, this leads to the Ra-
(c/n) λ
man effect, which can be regarded as an inelastic
where c is the velocity of light in a vacuum collision between the incident photon and the
(2.997925 × 1010 cm/s), c/n is the velocity of molecule. As a result of the collision, the vi-
light in a medium with refractive index n in brational or rotational energy of the molecule
8 Infrared and Raman Spectroscopy
Table 1. Infrared and microwave regions of the electromagnetic spectrum
Infrared
near 7.8×10−1 to 2.5 12 800 to 4000 3.8×1014 to 1.2×1014
mid 2.5 to 5.0×101 1.2×1014 to 6.0×1012
4000 to 200
far 5.0×101 to 1.0×103 6.0×1012 to 3.0×1011
200 to 10
Microwave 1.0×103 to 1.0×106 1.0×1012 to 3.0×108
10 to 0.01
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 With modern sampling techniques, good
is characteristic of a specific sample at a spe- quantitative infrared analysis with virtually ev-
cific wavelength. As the transmittance T does ery type of sample is practicable; however, liq-
not vary linearly with the concentration of an uids are ideal for this purpose, being measured
absorbing species the equation above is usually in a liquid cell of fixed thickness, either as 100 %
transformed by taking the logarithms of both sample or diluted with solvent. In this connec-
sides of the equation and replacing I/I 0 with I 0 /I tion it should be taken into account that there are
to eliminate the negative sign: no ideal solvents for infrared spectroscopy [35].
In addition, because absorption bands and path
I0
log =abc (14) length can be influenced by the temperature of
I the transmission cell, it is advisable to control
The term log (I 0 /I ) is the absorbance A, the temperature.
which changes linearly with changes in concen- Polymers are usually analyzed as pressed
tration of an absorbing species: films, and solids are often difficult to exam-
ine quantitatively by infrared methods; in these
A=abc (15) cases, absorption band ratios [10] give the best
results. In fact, the main application of multi-
This fundamental equation for spectromet- component quantitative infrared analysis is gas
ric quantitative analysis is known as Beer – analysis. If the difficulty in handling the gases
Lambert – Bouguer law, sometimes shortened to is overcome, multicomponent analysis can be
Beer’s law or Beer – Lambert law. readily performed.
In practice, several steps are involved in the A simple chemical system can consist of a
development of a quantitative method [35]. The pure single component, or of a single component
first and probably most crucial for the ultimate or more than one component in a mixture with no
success of the analytical method is to understand spectral interference; it is assumed that the radia-
the system, i.e., to obtain reference spectra of tion absorption by one component is not affected
the analyte and all other components. In the sec- by the presence of other components. In simple
ond step, the best method of sampling should be systems, absorbance peak height measurements,
determined. Then, with standards having been directly or by using a selected baseline [35], are
prepared, the system must be calibrated. The last often employed for calibration and analysis. Be-
step before analyzing samples is to prepare val- cause of intrinsic instrumental errors the practi-
idation samples and to evaluate the method. cal limit for usable absorbance values is about
10 Infrared and Raman Spectroscopy
three. Peak height measurements are also sen- chiefly in the band intensities that the two types
sitive to changes in instrumental resolution and of spectra differ, sometimes markedly so. On
can vary considerably from instrument to instru- the other hand, when a center of symmetry is
ment. To circumvent these problems, an alterna- present, i.e., for molecules with a high degree of
tive method is the use of integrated absorbance symmetry, bands that are infrared active are Ra-
or peak area [10]. man inactive and vice versa. By using laser ex-
Complex chemical systems are composed of citation (i.e., linearly polarized light) (see Sec-
one or more components in a mixture with a sig- tion 2.2), Raman spectroscopy also provides a
nificant degree of spectral interference, or of sev- means of recognizing totally symmetrical vibra-
eral components with a large amount of mutual tions [3]. The intensity of a Raman line depends
physical and/or chemical interaction. In these on the direction of the exciting electric field in re-
cases, quantitative analysis is best performed by lation to the orientation of the analyzer (Fig. 5).
statistical methods such as principal component The latter can be either perpendicular or parallel
regression (PCR) or partial least squares (PLS) to the direction of the electric field. The depolar-
[36]; these are offered in the software packages ization ratio is the ratio between the intensities
of instrument manufacturers and software sup- of scattered light measured at each of these two
pliers. Artificial neural networks (ANNs) should positions:
be primarily used when a data set is nonlinear
I⊥
[37]. = (16)
I
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.
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.
Figure 13. Characteristic group frequencies of aromatic Figure 12. FT-infrared (A) and FT-Raman (B) spectra of
rings 2,3-dimethyl-2-butene
The C−C stretching vibration of CH2 chains stretching vibration generally occurs above
is seen only in the Raman spectrum, at 1131 3000 cm−1 . In symmetrical trans or symmet-
and 1061 cm−1 [40], while the CH2 rocking vi- rical tetrasubstituted double-bond compounds,
bration at 720 cm−1 , which is characteristic for the C=C stretching frequency near 1640 cm−1 ,
longer (CH2 )n chains with n ≥ 4, is observed usually a medium intensity band, is infrared in-
only in the infrared spectrum (e.g., polyethyl- active because, in this case, no change of dipole
ene) [25]. moment occurs as a result of the vibration. How-
Aliphatic ring compounds are best character- ever, the C=C stretching vibration gives rise to a
ized by their Raman spectra, since in the infrared strong Raman signal in the region between 1680
only very weak characteristic ring frequencies and 1630 cm−1 in all types of alkenes (Fig. 12).
due to C−C ring stretching vibrations are ob- In vinyl and vinylidine groups, the =CH2 in-
served near 1000 cm−1 (Fig. 10). Raman spec- plane deformation gives rise to a medium inten-
tra, in contrast, show prominent lines due to ring sity band in both the infrared and Raman spec-
stretching vibrations: cyclopropane 1188 cm−1 trum near 1415 cm−1 . In the case of the vinyl
[25], cyclobutane 1001 cm−1 [40], cyclopen- group, an additional infrared and Raman defor-
tane 886 cm−1 [25], and cyclohexane 802 cm−1 mation band is observed near 1300 cm−1 [10].
(Fig. 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-
4.3. Alkene Groups trum. In the corresponding trans compounds, the
in-plane bending vibrations occur at 1305 cm−1
Characteristic group frequencies of alkenes are in the Raman and at 1295 cm−1 in the infrared
shown in Figure 11. In alkenes, the =CH spectrum. In these trans compounds, the =CH
Infrared and Raman Spectroscopy 15
out-of-plane bending vibration gives rise to a a very strong signal at 1000 cm−1 occurs for
strong infrared band, of high diagnostic value, mono and meta substitution, a strong line is
occurring between 980 and 965 cm−1 . 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].
4.6.1. Ethers
4.6.2. Alcohols and Phenols key bands are the NH stretching bands, although
they are often not very intense. Primary amines
Characteristic group frequencies of alcohols and exhibit two bands due to antisymmetric (3550 –
phenols are shown in Figure 22. Because of the 3330 cm−1 ) and symmetric (3450 – 3250 cm−1 )
strongly polarized OH group, alcohols generally stretching, while for secondary amines only one
show a weak Raman effect (and therefore are ex- NH stretching band is observed. The intensities
cellent solvents for Raman spectroscopy). In the of NH and NH2 stretching bands are generally
liquid or solid state, alcohols and phenols exhibit weaker in aliphatic than in aromatic amines. The
strong infrared bands due to O−H stretching vi- C−N stretching vibration is not necessarily di-
brations. Due to hydrogen bonding of the OH agnostic, because in aliphatic amines it gives rise
groups, these bands are very broad in the pure to absorption bands of only weak to medium in-
liquid or solid, as well as in concentrated solu- tensity between 1240 and 1000 cm−1 . In aro-
tions or mixtures. The position of the similarly matic amines, strong C−N stretching bands are
intense C−O stretching band permits distinction observed in the range 1380 – 1250 cm−1 .
between primary, secondary, and tertiary alco-
hols, as well as phenols.
4.8.1. Ketones
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
4.8.2. Aldehydes
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
4.8.9. Amides
β-Lactones 1840
α, β-unsaturated
1760 – 1750 and 1740 – 1715
β, γ-unsaturated
Figure 42. FT-infrared (A) and FT-Raman (B) spectra of Figure 44. FT-infrared (A) and FT-Raman (B) spectra of
l-cystine dimethyl sulfone
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
Figure 45. Optical diagram of a White gas cell (A mode)
Spectroscopy 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
5.1. Infrared Spectroscopy light beam enters and exits on one and the same side of the
field mirror, the objective mirrors are coupled and are rotated
5.1.1. Transmission Spectroscopy together linearly over the entire operational range
(Reproduced by permission of Harrick Scientific Corpora-
tion, Ossining, NY 10562)
Transmission spectroscopy is the simplest sam-
pling technique in infrared spectroscopy, and Apart from general laboratory gas analysis,
is generally used for routine spectral measure- gas-phase transmission spectroscopy is used for
ments on all kinds of samples. It requires only the characterization of air pollutants as well as
simple accessories, such as infrared-transparent the monitoring of stack gas and of air quality in
windows [e.g., potassium bromide, sodium chlo- work places. As air quality has become a major
ride, or thallium bromoiodide (KRS-5)] for concern for environmentalists, industrial manu-
gases and liquids, or a sample holder for solids. 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 29
wavelengths where the sample absorbs a frac- high sensitivity also for microsamples. As the
tion of the incident radiation, a modulated tem- signal is inversely proportional to the cell vol-
perature fluctuation at the same frequency, but ume, cells optimized for the measurement of
not necessarily with the same phase, as that of condensed samples have a small volume, typ-
the incident radiation is generated in the sam- ically well under 1 cm3 , to enhance the signal
ple, and the surrounding inert gas produces pe- amplitude [99]. Water vapor or carbon dioxide
riodic pressure waves in the sealed cell. These generate good photoacoustic signals, so even
fluctuations in pressure can be detected by a mi- traces of these impurities cause serious interfer-
crophone because the modulation frequency of ence in the photoacoustic spectroscopy signal;
the incident beam usually lies in the acoustic therefore, the purity of the gas in the cell is es-
range, e.g., between 100 and 1600 Hz. The low sential for good-quality photoacoustic spectra.
power available from dispersive instruments re- Important applications of photoacoustic
sults in spectra with a rather poor signal-to-noise spectroscopy are near-surface analysis [100] and
ratio; hence, photoacoustic mid-infrared spectra depth profiling of polymers and polymer lay-
of condensed substances are preferably obtained ers [101–103]. The theory of depth profiling
by using an FT-IR spectrometer with a pho- is described in detail in [92]. It is based on
toacoustic cell accessory. In a rapid-scanning the fact that thermal diffusion length (the dis-
Michelson interferometer with a mirror veloc- tance a thermal wave travels in the sample be-
ity of v (in cm/s), radiation of wavenumber ν̃ fore its intensity is reduced by 1/e in magni-
(in cm−1 ) is modulated with a frequency of tude) and hence intensity of the photoacoustic
2 v ν̃ Hz [8]. Therefore, FT-IR spectrometers are signal are increased by lowering the modula-
ideal for photoacoustic measurements in the tion frequency. In a rapid-scanning interferome-
mid-infrared, where mirror velocities on the or- ter this is achieved by decreasing the mirror ve-
der of 0.05 – 0.2 cm s−1 provide modulation fre- locity. To date, rapid-scanning interferometers
quencies in the acoustic range. As carbon black that provide modulation frequencies which con-
absorbs all incident radiation, the sample single- tinuously change with wavenumber have gener-
beam spectrum is usually ratioed against a car- ally been used. However, step-scan interferom-
bon black single-beam spectrum to obtain an eters accommodating any desired modulation
absorbance-like spectrum. 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.
graphic elution time. This is especially so when search with a vapor-phase spectra database is
molecules that have various conformational iso- commonly used. Helpful for the identification in
mers are to be distinguished. the vapor phase is the virtual absence of interac-
Supercritical fluid chromatography tion between molecules, and therefore the non-
(SFC)/FT-IR spectroscopy, generally with car- destructive IR detection enables in many cases
bon dioxide as the mobile phase, bridges the an easy and fast interpretation of single and over-
gap between GC/FT-IR and LC/FT-IR, and is lapping absorption bands.
particularly useful for separating nonvolatile or Providing structural information for sample
thermally labile materials not amenable to gas components down to nanogram levels, TG/FT-
chromatographic separation [109]. Flow cells, IR applications include decomposition studies
mobile phase elimination and matrix-isolation of polymers and laminates [123], the analysis
techniques are used as SFC/FT-IR interfaces. of coal, oil shales, and petroleum source rocks
Applications of chromatography/FT-IR spec- [124], [125], and the determination of activa-
troscopy involve toxins and carcinogens, waste- tion energies [126] and thermal decomposition
water constituents, sediment extracts, airborne of energetic materials [127].
species, pesticides and their degradation prod- The IR detection of evolved gases opens
ucts, fuels and fuel feedstocks, flavors and new possibilities in the quantification of ther-
fragrances, natural products, pharmaceuticals, mogravimetric analysis. Besides measuring the
biomedicine, and polymers. 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
5.1.8. Thermogravimetry/Fourier further improvement in quantitative TG/FT-IR
Transform Infrared Spectroscopy is achieved by PulseTA [129]. Here, a certain
amount of gas or liquid is injected into the sys-
Thermogravimetry (TG), also called thermo- tem and used as internal calibration standard
gravimetric analysis (TGA), measures changes for the quantification of the unknown amount
in sample mass as a function of increasing of evolved gas.
temperature and has become an important tool
for the analysis and characterization of mate-
rials [118], [119]. To correlate weight changes 5.1.9. Vibrational Circular Dichroism
with chemical changes, it is necessary to iden-
tify gases evolved from the thermogravimet- Vibrational circular dichroism (VCD) extends
ric analyzer furnace. This is made possible by the functionality of electronic circular dichroism
coupling thermogravimetry with FT-IR spec- (CD) into the infrared spectral region, with VCD
troscopy, which provides the necessary speed bands of enantiomeric pairs of chiral molecules
and sensitivity. having opposite signs. Early VCD spectrome-
A system has been described in which the ters were dispersive instruments equipped with
infrared beam of the FT-IR spectrometer is led a CaF2 photoelastic modulator for creating left
directly into the TG equipment [120], but in and right circularly polarized radiation and a liq-
commercially available systems [121], [122] the uid-nitrogen-cooled InSb detector. FT-IR-VCD
evolved gases are conducted from the TG sys- measurements are carried out by using the so-
tem to the spectrometer by carrier gas flow via a called double modulation technique [130] in
transfer line. The interface is heated to prevent which the VCD interferogram is carried by the
recondensation and incorporates the transfer line polarization modulation frequency in the region
and a flow-through gas cell with nitrogen or he- of 50 kHz, and the IR transmission interfero-
lium as carrier gas. gram occurs in the usual range of Fourier fre-
After infrared spectra at low resolution (usu- quencies, typically below 5 kHz.
ally 4 cm−1 ) have been recorded as a func- Combining the structural specificity of FT-
tion of time and temperature, the further proce- IR spectroscopy with the stereosensitivity of cir-
dure is similar to GC/FT-IR (see Section 5.1.7) cular dichroism, VCD allows the determination
and, in fact, GC/FT-IR software, such as Gram – of optical purity without enantiomeric separa-
Schmidt thermogram reconstruction and library tion or derivatization, as well as of absolute
36 Infrared and Raman Spectroscopy
configuration without crystallization. Simulta- thickness exceeds 2 mm the intensity does not
neous monitoring of the optical purity of multi- increase. Optical fibers are used for remote sam-
ple chiral species, such as reactant and product pling, and for FT-Raman spectroscopy this is es-
molecules in a reaction process is also possible, pecially advantageous as the transmission by the
as is the determination of conformations of bio- fibers, excited by a near-infrared Nd : YAG laser
logical molecules such as proteins, peptides, and [33], [135], can be at its highest just in the range
DNA [131], [132]. of the Raman spectrum. Guidelines for optimiz-
ing FT-Raman recording conditions are given
in [34]. Conventional dispersive Raman spec-
5.2. Raman Spectroscopy troscopy has not been widely used for quantita-
tive measurements, mainly because of difficul-
In mid-infrared spectroscopy, Fourier transform ties with band reproducibility. However, by us-
instruments are used almost exclusively. How- ing an FT-Raman spectrometer, spectra with ex-
ever, in Raman spectroscopy both conventional cellent reproducibility, both in band position and
dispersive and Fourier transform techniques intensity, are obtained, so that this technique has
have their applications, the choice being gov- great potential for quantitative analysis [136–
erned by several factors [133], [134]. Conse- 138].
quently, a modern Raman laboratory is equipped Along with general reviews [139–142] and
with both Fourier transform and CCD-based dis- surveys of major areas of Raman spectroscopy
persive instruments. For a routine fingerprint [16], [17], various reviews covering special ap-
analysis, the FT system is generally used, be- plications of Raman spectroscopy, such as po-
cause it requires less operator skill and is quicker larization measurements [143], polymers [144],
to set up; the FT system is also be tried first proteins and nucleic acids [145], pharmaceu-
if samples are highly fluorescent or light sen- ticals and biomaterials [146–150], paint sys-
sitive. However, if the utmost sensitivity is re- tems [151], [152], inorganic and organometallic
quired, or if Raman lines with a shift smaller materials [153], and thin-layer chromatography
than 100 cm−1 are to be recorded, conventional [154], have been published. For further appli-
spectrometers are usually preferred. cations and special Raman techniques, such as
Raman spectroscopy offers a high degree of resonance Raman (RR), surface-enhanced Ra-
versatility, and sampling is easy because no spe- man spectroscopy (SERS), coherent anti-Stokes
cial preparation is required. As it is a scatter- Raman scattering (CARS), and stimulated Ra-
ing technique, samples are simply placed in the man scattering (SRS), see [14], [155].
laser beam and the backscattered radiation is an-
alyzed.
Glass gives only a weak Raman signal, so that 5.3. Fourier Transform Infrared and
FT-Raman spectra from liquids are recorded by Raman Microspectroscopy
using a capillary cell [34], a sapphire sphere [34],
or glass cuvettes with the back surface silvered In conventional infrared microsampling, beam
for the favored 180◦ backscattering arrangement size is reduced four to six times by using beam
(see Section 2.2). Moreover, the study of aque- condensers. These accessories, which permit
ous samples in glass or quartz vessels, which submicrogram samples of between 2 and 0.5 mm
is difficult in the mid-infrared region owing to in diameter to be measured, have, however,
strong infrared absorption bands of water and drawbacks such as heating effects, since the en-
quartz, is possible in Raman spectroscopy be- tire power of the infrared beam is condensed on
cause there are no significant bands causing in- the sample, and the fact that the sample can-
terference in the solution spectrum. Powders are not be visibly located, so that a skilled operator
either measured by using microspherical cells is needed for this time-consuming work. These
[33] or are filled into cells consisting of a brass disadvantages are largely overcome by coupling
cylinder with a 3 mm hole drilled through, into an optical microscope to an FT-IR spectrome-
which a reflective metal rod is placed to com- ter. As infrared lenses are poor in performance,
press the powder [14]. The Raman intensity is infrared microscopes use reflecting optics such
related to the sample thickness, but once the 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
cialists. Regardless of the specific method used sensitive to alterations in protein secondary and
to collect data during a vibrational spectroscopic tertiary structures. Spectral shifts could there-
imaging experiment, the final representation is fore be used to image specifically the distribu-
typically expressed as a three-dimensional im- tions of specific structural moieties within a bi-
age cube, or hypercube, consisting of two spa- ological sample as opposed to measuring sim-
tial dimensions and one spectral dimension. The ply an overall distribution of a general class of
image cube concept, shown in Figure 56, is in- molecules. Multivariate approaches may also be
terpreted in one of two ways, either (1) a series used to generate a composite metric indicative of
of images collected as a function of wavenum- either disease progression or biological function.
ber (Figure 56A), or (2) a vibrational spectrum Although these metrics can not be readily in-
corresponding to each spatial location, or pixel, terpreted in terms of biochemical changes, they
within the image plane (Figure 56B). The abil- may often be statistically correlated with dis-
ity to merge and recall, within a single analyti- ease. In addition, by simultaneously recording
cal technique, information corresponding to the data on a two-dimensional focal-plane array de-
spatial and spectral axes of the image cube pro- tector from a myriad of spatial locations within a
vides extraordinary flexibility in probing and ex- biological matrix, pooled spectra may be treated
tracting structural and compositional informa- statistically. In this manner, analysis of variance
tion from samples. Thus, a data set may be either can be employed to more robustly test the signif-
summarized as a single image or expressed as a icance of observed spectral changes appearing,
series of images derived from one or more spec- for example, because of a perturbed or diseased
troscopic parameters, such as spectral peak in- state.
tensity, band frequency, or linewidth. The chem- Raman and infrared imaging techniques as
ical and morphological interpretations of the re- well as their biological applications are de-
sulting images are then based on the spatial vari- scribed in detail in [170].
ations of the intrinsic molecular property being
highlighted and their relevance to the biologi-
cal questions being pursued. In contrast to im- 5.6. Infrared and Raman Spectroscopy
ages generated by conventional point mapping, of Polymers (see also → Plastics Analysis,
which generally suffer from low spatial resolu- Chap. 4.)
tion due to relatively few sampling points, direct
imaging approaches using the advantages inher- As polymer analysis is a discipline of its own
ent in two-dimensional array detectors derive an beyond the scope of this article, the reader is re-
enormous benefit from the highly parallel na- ferred to the comprehensive works by Hummel
ture of the data collection. That is, a single data [171], Koenig [172], and Zerbi et al. [173].
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 6. Near-Infrared Spectroscopy
spectral variations that are often overlooked or
misinterpreted in traditional spectroscopic stud- 6.1. Comparison of Mid-Infrared and
ies to be tested and revealed. Near-Infrared Spectroscopy
Vibrational spectroscopic imaging has been
widely applied to diverse materials, includ- Near-infrared spectroscopy as a valuable tool
ing polymers, semiconductors, pharmaceuticals, for qualitative and quantitative analysis [174],
cosmetics, consumer products, and biological [175] has experienced a revival through the de-
materials. The strengths and adaptability of vi- velopment of chemometric methods and the in-
brational spectroscopic imaging rest not only troduction of fiber optics during the 1980s, and
on the ability to determine chemical composi- high-quality near-infrared spectra are provided
tions and component distribution within a sam- by improved conventional and by Fourier trans-
ple, but also on being able to extract localized form instruments.
molecular information relevant to the sample ar- Various aspects differentiate a mid-infrared
chitecture. For example, vibrational spectra are spectrum from the corresponding near-infrared
40 Infrared and Raman Spectroscopy
spectrum (Table 6). In the mid-infrared region Table 6. Comparison of mid-infrared and near-infrared
fundamental molecular vibrations always oc- spectroscopy
cur between 4000 and 200 cm−1 , while in the
Mid-Infrared Near-Infrared
near-infrared region overtones and combina-
tion bands of these fundamentals are observed 4000 – 200 cm−1 12 800 – 4000 cm−1
between 12 800 and 4000 cm−1 . Near-infrared Fundamentals
High intensity
overtones and combinations
low intensity
bands are due primarily to hydrogenic stretches High sensitivity (trace analysis low sensitivity (not suited for
of C−H, N−H, and O−H bonds whose funda- possible) trace analysis)
Sharp, separated bands strongly overlapping bands
mental vibrations give rise to strong bands bet- Easy quantitation with isolated quantitation complex,
ween 4000 and 2000 cm−1 , so that their over- bands chemometrics necessary
tones and combinations occur above 4000 cm−1 FT-instruments grating, filter, FT and
acousto-optical tunable scanning
as the most intense absorption bands in the near- instruments
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 6.2. Applications of Near-Infrared
mid-infrared spectroscopy, which shows all (ac- Spectroscopy
tive) fundamentals and the overtones and com-
binations of low-energy vibrations. The near- Applications of near-infrared spectroscopy in-
infrared overtone and combination bands de- clude chemistry [179], [180], the oil industry
pend more on their environment than does the [181], clinical analysis [182], biological and
fundamental of the same vibration; slight per- medical analysis [183], and the pharmaceutical
turbation in the bonding scheme causes only industry [184], [185].
small changes in the fundamental but signifi- Since the intensities of characteristic near-
cant frequency shifts and amplitude changes in infrared bands are independent of or only
the near-infrared. In going from the fundamental slightly dependent on the state of the system,
to the first overtone the intensity of an absorp- near-infrared spectroscopy is widely applicable
tion band decreases by a factor of about 10 – for the quantitative study of liquid and com-
100, so that the sensitivity of near-infrared spec- pressed gaseous systems, including fluid sys-
troscopy is reduced in comparison with the mid- tems, up to high pressures and temperatures.
infrared. This is a disadvantage when gases are The application of near-infrared data to rou-
to be measured, but for liquids, it is a consider- tine quantitative analysis was initiated by Nor-
able advantage as regards sample handling be- ris [186], who used diffuse reflectance measure-
cause cells with convenient path lengths between ments to quantitatively determine major com-
1 mm and 10 cm can be used. Glass or quartz is ponents, such as moisture, protein and fat, in
used as window material. Near-infrared bands agricultural commodities. In comparison with
of liquids and solids have relatively large band- mid-infrared, near-infrared diffuse reflectance
widths between 30 and 60 cm−1 , so that they [187] is able to measure powdered samples
strongly overlap and direct assignment of bands with minimal sample preparation, and lends it-
is generally not possible for larger, complex self extremely well to quantitative analysis be-
molecules. This is why, in the near-infrared, easy cause the smaller absorptivities and larger scat-
quantitative analysis using isolated bands, as in tering coefficients at shorter wavelengths result
the mid-infrared, is not possible. Chemometrics in a larger scattering/absorption ratio. Although
[37], [176–178], however, by the application of near-infrared reflectance analysis was initially
mathematical and statistical procedures, gener- developed for agricultural and food applications
ates correlations between experimental data and [188], in combination with chemometrics it is
the chemical composition or physical properties now applied to many other areas, e.g., poly-
of the sample, and can be used in a general man- mers [189], pharmaceuticals [190], organic sub-
ner to solve quantitative and qualitative analyti- stances [191], geological samples [192], and
cal problems. thin-layer chromatography [193], [194].
Infrared and Raman Spectroscopy 41
The excellent transmittance of quartz in the dustries, typical applications of near-infrared re-
near-infrared range has led to a further en- mote spectroscopy are the determination of the
hancement of the potential of near-infrared spec- octane number and oxygenates in gasoline, aro-
troscopy by the introduction of fiber optics. matics in fuels, the composition of solvent mix-
Fiber-optic waveguides are used to transfer light tures [197], [198], the hydroxyl number of poly-
from the spectrometer to the sample and, af- ols, low contents of water in solvents, and for
ter transmission or reflection, back to the spec- polymer analysis.
trometer. Most fiber optic cables consist of three
concentric components: the inner part, through
which the light propagates, is called the core, 7. References
the middle layer is the cladding, and the outer
protective layer is the jacket. Crucial to the light 1. W. W. Coblentz: Investigations of Infrared
throughput of the fiber optic cable are the op- Spectra, Carnegie Institution, Washington
tical properties of the core and cladding. The 1905, re-published, The Coblentz Society,
difference in refractive index of the two ma- Norwalk 1962.
terials defines the highest angle at which the 2. G. Herzberg: Molecular Spectra and
core/cladding interface exhibits total internal re- Molecular Structure I, Van Nostrand,
flection, and hence allows the light to propagate Princeton 1953.
through the cable. This angle is related to the 3. G. Herzberg: Molecular Spectra and
numerical aperture (N.A.) of the fiber by the fol- Molecular Structure II, Van Nostrand,
lowing equation: Princeton 1960.
4. E. B. Wilson, Jr., J. C. Decius, P. C. Cross:
1/2
sinα= N.A. = n2core −n2cladding (20)
Molecular Vibrations, McGraw-Hill, New
York 1955, re-published Dover, Mineola 1980.
where α is the half-angle of acceptance and n is 5. D. Steele: Theory of Vibrational Spectroscopy,
the refractive index of the material. The greater Saunders, Philadelphia 1971.
the difference in refractive index the greater the 6. J. A. Koningstein: Introduction to the Theory
of the Raman Effect, Reidel, Dordrecht 1972.
half-angle of acceptance. Any light entering at
7. D. A. Long: Raman Spectroscopy,
an angle greater than the half-angle of accep-
McGraw-Hill, New York 1977.
tance is lost to the cladding through absorption.
8. P. R. Griffiths, J. A. de Haseth: Fourier
The core and cladding are usually of quartz (of Transform Infrared Spectroscopy, J. Wiley &
different refractive index), while the outer pro- Sons, New York 1986.
tective layer is a polymer; other fiber optic ma- 9. M. W. Mackenzie: Advances in Applied
terials are zirconium fluoride (ZrF) and chalco- Fourier Transform Infrared Spectroscopy, J.
genide (AsGeSe) [195]. The obvious advantage Wiley & Sons, New York 1988.
of the optical fiber technique is that the loca- 10. N. B. Colthup, L. H. Daly, S. E. Wiberley:
tion of measurement can be separated from the Introduction to Infrared and Raman
spectrometer by between two and several hun- Spectroscopy, 3rd ed., Academic Press, San
dred meters, depending on the length of the Diego 1990.
waveguide. Since almost all practically relevant 11. B. Schrader (ed.): Infrared and Raman
substances show characteristic near-infrared ab- Spectroscopy, VCH Verlagsgesellschaft,
sorption bands, quantitative analysis by near- Weinheim 1995.
infrared spectroscopy is generally applicable to 12. K. W. F. Kohlrausch: Ramanspektren,
on-line concentration measurements in connec- Akademische Verlagsgesellschaft, Leipzig
tion with chemical reactions, chemical equilib- 1943, re-published Heyden & Son, London
ria, and phase equilibria. Various probes can be 1973.
integrated into different reactors or bypasses and 13. H. Baranska, A. Labudzinska, T. Terpinski:
Laser Raman Spectroscopy, J. Wiley & Sons,
offer numerous remote-sensing and on-line pro-
New York 1983.
cess control applications [196]. In the analysis of
14. P. Hendra, C. Jones, G. Warnes: Fourier
hazardous or toxic materials this has proved of
Transform Raman Spectroscopy, Ellis
outstanding importance. In the refining, petro- Horwood, Chichester 1991.
chemical, polymer, plastics, and chemicals in-
42 Infrared and Raman Spectroscopy
15. J. R. Ferraro, K. Krishnan (eds.): Practical Spectroscopy, Academic Press, San Diego
Fourier Transform Infrared Spectroscopy, 1990, p. 167.
Academic Press, San Diego 1990. 34. B. Schrader, A. Hoffmann, S. Keller,
16. M. J. Pelletier (ed.): Analytical Applications of Spectrochim. Acta 47 A (1991) 1135 – 1148.
Raman Spectroscopy, Blackwell Science, 35. J. Coates in W. O. George, H. A. Willis (eds.):
Oxford 1999. Computer Methods in UV, Visible and IR
17. J. G. Grasselli, B. J. Bulkin (eds.): Analytical Spectroscopy, The Royal Society of
Raman Spectroscopy, J. Wiley & Sons, New Chemistry, Cambridge 1990, p. 95.
York 1991. 36. P. S. Wilson in W. O. George, H. A. Willis
18. G. Socrates: Infrared Characteristic Group (eds.): Computer Methods in UV, Visible and
Frequencies, J. Wiley & Sons, New York IR Spectroscopy, The Royal Society of
1980. Chemistry, Cambridge 1990, p. 143.
19. L. J. Bellamy: The Infrared Spectra of 37. F. Despagne, D. L. Massart, Analyst 123
Complex Molecules, 3rd ed., vol. 1, Chapman (1998) 157R – 178R.
& Hall, London 1975. 38. G. M. Barrow: Introduction to Molecular
20. L. J. Bellamy: The Infrared Spectra of Spectroscopy, McGraw-Hill, New York 1962.
Complex Molecules, 2nd ed., vol. 2, Chapman 39. S. R. La Paglia: Introductory Quantum
& Hall, London 1980. Chemistry, Harper & Row, New York 1971.
21. F. R. Dollish, W. G. Fateley, F. F. Bentley: 40. Ullmann, 4th ed., 5, 303.
Characteristic Raman Frequencies of Organic 41. Available from Sadtler Division, Bio-Rad
Compounds, Wiley Interscience, New York Laboratories, Inc., 3316 Spring Garden Street,
1974. Philadelphia, PA 19104.
22. D. Lin-Vien, N. B. Colthup, W. G. Fateley, 42. Available from LabControl GmbH, D-50858
J. G. Grasselli: The Handbook of Infrared and Köln 40, Max-Planck-Str. 17 a, Germany.
Raman Characteristic Frequencies of Organic 43. Available from Galactic Industries
Molecules, Academic Press, Boston 1991. Corporation, 395 Main Street, Salem NH
23. K. Nakamoto: Infrared and Raman 03079.
Spectroscopy of Inorganic and Coordination 44. J. P. Coates, R. W. Hannah in T. Theophanides
Compounds, 5th ed., J. Wiley & Sons, (ed.): Fourier Transform Infrared
Chichester 1997. Spectroscopy, D. Reidel Publ., Dordrecht
24. H.-U. Gremlich, B. Yan (eds.): Infrared and 1984, p. 167.
Raman Spectroscopy of Biological Materials, 45. S. R. Lowry, D. A. Hupplar, Anal. Chem. 53
Marcel Dekker, New York 2000. (1981) 889 – 893.
25. B. Schrader: Raman/Infrared Atlas of Organic 46. J. T. Clerc in W. O. George, H. A. Willis (eds.):
Compounds, 2nd ed., VCH Computer Methods in UV, Visible and IR
Verlagsgesellschaft, Weinheim 1989. Spectroscopy, The Royal Society of
26. Sadtler IR Digital Spectra Libraries, Sadtler Chemistry, Cambridge 1990, p. 13.
Division, Bio-Rad Laboratories, Inc., 3316 47. D. A. Cirovic, Trends Anal. Chem. 16 (1997)
Spring Garden Street, Philadelphia, PA 19104. 148 – 155.
27. F. M. Mirabella (ed.): Modern Techniques in 48. H. Yang, P. R. Griffiths, Anal. Chem. 71
Applied Molecular Spectroscopy, J. Wiley & (1999) 751 – 761.
Sons, New York 1998. 49. J. U. White, J. Opt. Soc. Am. 32 (1942)
28. A. A. Michelson, Philos. Mag. 31 (1891) 285 – 288.
256 – 259. 50. R. T. Graf, J. L. Koenig, H. Ishida, Appl.
29. W. Herres, J. Gronholz, Comput. Appl. Lab. 4 Spectrosc. 39 (1985) 405 – 412.
(1984) 216 – 220. 51. R. G. Greenler, J. Chem. Phys. 44 (1966)
30. J. W. Cooley, J. W. Tukey, Math. Comput. 19 310 – 315.
(1965) 297 – 301. 52. R. G. Greenler, J. Chem. Phys. 50 (1969)
31. J. Gronholz, W. Herres, Instrum. Computers 3 1963 – 1968.
(1985) 10 – 19. 53. D. L. Allara, J. D. Swalen, J. Phys. Chem. 86
32. B. Schrader, A. Hoffmann, A. Simon, J. (1982) 2700 – 2704.
Sawatzki, Vib. Spectrosc. 1 (1991) 239 – 250. 54. N. J. Harrick, Appl. Opt. 10 (1971)
33. B. Schrader in J. R. Ferraro, K. Krishnan 2344 – 2349.
(eds.): Practical Fourier Transform Infrared 55. I. Newton: Opticks, Dover, New York 1952.
Infrared and Raman Spectroscopy 43
103. C. Q. Yang, R. R. Bresee, W. G. Fateley, Appl. 128. E. Post, S. Rahner, H. Möhler, A. Rager,
Spectrosc. 41 (1987) 889 – 896. Thermochim. Acta 263 (1995) 1 – 6.
104. R. Rubinovitz, J. Seebode, A. Simon, Bruker 129. M. Maciejewski, F. Eigenmann, A. Baiker, E.
Report 1 (1988) 35 – 38. Dreser, A. Rager, Thermochim. Acta (2000) in
105. M. J. Smith, C. J. Manning, R. A. Palmer, J. L. press.
Chao, Appl. Spectrosc. 42 (1988) 546 – 555. 130. L. A. Nafie in M. W. Mackenzie (ed.):
106. R. A. Crocombe, S. V. Compton, FTS/IR Notes Advances in Applied FTIR Spectroscopy, John
no. 82, Bio-Rad, Digilab Division, Cambridge, Wiley and Sons, New York 1988, p. 67.
MA 02139, 1991. 131. L. A. Nafie, T. B. Freedman in H.-U. Gremlich,
107. K. H. Michaelian, Appl. Spectrosc. 45 (1991) B. Yang (eds.): Infrared and Raman
302 –304. Spectroscopy of Biological Materials, Marcel
108. L. Bertrand, Appl. Spectrosc. 42 (1988) Dekker, New York 2000, p. 15.
134 – 138. 132. T. A. Keiderling in H.-U. Gremlich, B. Yang
109. R. White: “Chromatography/Fourier (eds.): Infrared and Raman Spectroscopy of
Transform Infrared Spectroscopy and Its Biological Materials, Marcel Dekker, New
Applications,” Practical Spectroscopy Series, York 2000, p. 55.
vol. 10, Marcel Dekker, New York 1990. 133. B. Schrader, S. Keller, Proc. SPIE Int. Soc.
110. W. Herres: HRGC-FTIR: Capillary Gas Opt. Eng. 1575 (Int. Conf. Fourier Transform
Chromatography-Fourier Transform Infrared Spectrosc., 8th, 1991) (1992) 30 – 39.
Spectroscopy, Hüthig Verlag, Heidelberg 134. N. J. Everall, J. Lumsdon, Spectrosc. Eur. 4
1987. (1992) 10 – 21.
111. S. Bourne et al., FTS/IR Notes no. 75, 135. C. G. Zimba, J. F. Rabolt, Appl. Spectrosc. 45
Bio-Rad, Digilab Division, Cambridge, MA (1991) 162 – 165.
02139, 1990. 136. T. Jawhari, P. J. Hedra, H. A. Willis, M.
112. R. L. White, G. N. Giss, G. M. Brissey, C. L. Judkins, Spectrochim. Acta 46 A (1990)
Wilkins, Anal. Chem. 53 (1981) 1778 – 1782. 161 – 170.
113. P. Coffey, D. R. Mattson, J. C. Wright, Am. 137. F. T. Walder, M. J. Smith, Spectrochim. Acta
Lab. 10 (1978) 126 – 132. 47 A (1991) 1201 – 1216.
114. K. S. Chiu, K. Biemann, K. Krishnan, S. L. 138. H. Sadeghi-Jorabchi, R. H. Wilson, P. S.
Hill, Anal. Chem. 56 (1984) 1610 – 1615. Belton, J. D. Edwards-Webb, D. T. Coxon,
115. Available from Lab Connections, Inc., 201 Spectrochim. Acta 47 A (1991) 1449 – 1458.
Forest Street, Marlborough, MA 01752. 139. L. A Lyon et al., Anal. Chem. 70 (1998)
116. Available from Bourne Scientific, Inc., 65 341R – 361R.
Bridge Street, Lexington, MA 02421-7927. 140. B. Chase in A. M. C. Davies, C. S. Creaser
117. S. Bourne, Am. Lab. 16 (1998) 17F – 17J. (eds.): Anal. Appl. Spectrosc. 2, The Royal
118. C. M. Earnest, Anal. Chem. 56 (1984) Society of Chemistry, Cambridge 1991, p. 13.
1471 A –1486 A. 141. K. P. J. Williams, S. M. Mason, TRAC 9
119. D. Dollimore, Anal. Chem. 64 (1992) (1990) 119 –127.
147 R – 153 R. 142. P. J. Hendra, J. Mol. Struct. 266 (1992)
120. J. A. J. Jansen, J. H. van der Maas, A. 97 –114.
Posthuma de Boer, Appl. Spectrosc. 46 (1992) 143. A. Hoffmann et al., J. Raman Spectrosc. 22
88 – 92. (1991) 497 – 503.
121. K. Ichimura, H. Ohta, T. Tajima, T. Okino,
144. P. J. Hendra, C. H. Jones, Makromol. Chem.
Mikrochim. Acta 1 (1988) 157 – 161.
Macromol. Symp. 52 (1991) 41 – 56.
122. R. C. Wieboldt, S. R. Lowry, R. J. Rosenthal,
145. G. J. Thomas, Ann. Rev. Biophys. Biomol.
Mikrochim. Acta 1 (1988) 179 – 182.
Struct. 28 (1999) 1 – 27.
123. J. Mullens et al., Bull. Soc. Chim. Belg. 101
146. A. M. Tudor et al., J. Pharm. Biomed. Anal. 8
(1992) 267 – 277.
(1990) 717 – 720.
124. R. M. Carangelo, P. R. Solomon, D. J. Gerson,
147. P. R. Carey, J. Biol. Chem. 274 (1999)
Fuel 66 (1987) 960 – 967.
26 625 – 26 628.
125. J. K. Whelan, Energy Fuels 2 (1988) 65 – 73.
148. J. Góral, V. Zichy, Spectrochim. Acta 46 A
126. L. A. Sanchez, M. Y. Keating, Appl. Spectrosc.
(1990) 253 – 275.
42 (1988) 1253 – 1258.
127. M. D. Timken, J. K. Chen, T. B. Brill, Appl. 149. I. W. Levin, E. N. Lewis, Anal. Chem. 62
Spectrosc. 44 (1990) 701 – 706. (1990) 1101 A – 1111 A.
Infrared and Raman Spectroscopy 45
150. S. Nie et al., Spectroscopy 5 (1990) 24 – 32. 172. J. L. Koenig: Spectroscopy of Polmers,
151. G. Ellis, M. Claybourn, S. E. Richards, American Chemical Society, Washington,
Spectrochim. Acta 46 A (1990) 227 – 241. D.C. 1992.
152. D. Bourgeois, S. P. Church, Spectrochim. Acta 173. G. Zerbi et al.: Modern Polymer Spectroscopy,
46 A (1990) 295 – 301. Wiley-VCH, Weinheim 1998.
153. T. N. Day, P. J. Hendra, A. J. Rest, A. J. 174. J. Workman, Jr., J. Near Infrared Spectrosc. 1
Rowlands, Spectrochim. Acta 47 A (1991) (1993) 221.
1251 – 1262. 175. I. Murray, I. A. Cowe (eds.): Making Light
154. N. J. Everall, J. M. Chalmers, I. D. Newton, Work: Advances in Near Infrared
Appl. Spectrosc. 46 (1992) 597 – 601. Spectroscopy, VCH Verlagsgesellschaft,
155. D. L. Gerrard, J. Birnie, Anal. Chem. 64 Weinheim 1992.
(1992) 502 R – 513 R. 176. P. Geladi, E. Dabbak, J. Near Infrared
156. B. Yan, Acc. Chem. Res. 31 (1998) 621 – 630. Spectrosc. 3 (1995) 119.
157. R. G. Messerschmidt, M. A. Harthcock (eds.):
177. D. L. Massart, B. G. M. Vandeginste, L. M. C.
Infrared Microspectroscopy: Theory and
Buydens, S. de Jong, P. J. Lewi, J.
Applications, Marcel Dekker, New York 1988.
Smeyers-Verbecke: Handbook of
158. K. Krishnan, S. L. Hill in J. R. Ferraro, K.
Chemometrics and Qualimetrics, Part A,
Krishnan (eds.): Practical Fourier Transform
Elsevier, Amsterdam 1997.
Spectroscopy, Academic Press, San Diego
178. D. L. Massart, B. G. M. Vandeginste, L. M. C.
1990, p. 103.
159. S. Compton, J. Powell, FTS 7/IR Notes no. 2, Buydens, S. de Jong, P. J. Lewi, J.
Bio-Rad, Digilab Division, Cambridge, MA Smeyers-Verbecke: Handbook of
02139, 1991. Chemometrics and Qualimetrics, Part B,
160. N. J. Harrick, M. Milosevic, S. L. Berets, Appl. Elsevier, Amsterdam 1998.
Spectrosc. 45 (1991) 944 – 948. 179. S. M. Donahue, C. W. Brown, B. Caputo, M. D.
161. B. Schrader, Fresenius J. Anal. Chem. 337 Modell, Anal. Chem. 60 (1988) 1873 – 1878.
(1990) 824 – 829. 180. J. Lin, C. W. Brown, Vibr. Spectrosc. 7 (1994)
162. J. D. Louden in D. J. Gardiner, P. R. Graves 117 – 123.
(eds.): Practical Raman Spectroscopy, 181. S. R. Westbrook, SAE Tech. Paper Ser. 930734
Springer Verlag, Berlin 1989, p. 119. (1993) 1 – 10.
163. J. Sawatzki, Fresenius J. Anal. Chem. 339 182. H. M. Heise in H.-U. Gremlich, B. Yan (eds.):
(1991) 267 –270. Infrared and Raman Spectroscopy of
164. A. Hoffmann, B. Schrader, R. Podschadlowski, Biological Materials, Marcel Dekker, New
A. Simon, Proc. SPIE-Int. Soc. Opt. Eng. 1145 York 2000, p. 259.
(Int. Conf. Fourier Transform Spectrosc., 7th, 183. R. J. Dempsey, D. G. Davis, R. G. Buice, Jr.,
1989) (1989) 372 – 373. R. A. Lodder, Appl. Spectrosc. 50 (1996)
165. F. Bergin, Proc. Annu. Conf. Microbeam Anal. 18A – 34A.
Soc. 25 (1990) 235 – 239. 184. J. D. Kirsch, J. K. Drennen, Appl. Spectrosc.
166. B. W. Cook in A. M. C. Davies, C. S. Creaser Rev. 30 (1995) 139 – 174.
(eds.): Anal. Appl. Spectrosc. 2, The Royal 185. W. Plugge, C. van der Vlies, J. Pharm.
Society of Chemistry, Cambridge 1991, p. 61. Biomed. Anal. 14 (1996) 891 – 898.
167. A. J. Sommer, J. E. Katon, Appl. Spectrosc. 45 186. I. Ben-Gera, K. H. Norris, Israel J. Agric. Res.
(1991) 527 – 534. 18 (1968) 125 – 132.
168. D. Gernet, W. Kiefer, H. Schmidt, Appl. 187. D. L. Wetzel, Anal. Chem. 55 (1983)
Spectrosc. 46 (1992) 571 – 576. 1165 A – 1176 A.
169. K. Gerwert in H.-U. Gremlich, B. Yang (eds.): 188. P. Williams, K. Norris (eds.): Near-Infrared
Infrared and Raman Spectroscopy of Technology in the Agricultural and Food
Biological Materials, Marcel Dekker, New Industries, AACC, St. Paul 1987.
York 2000, p. 193. 189. C. E. Miller, Appl. Spectrosc. Rev. 26 (1991)
170. M. D. Schaeberle, I. W. Levin, E. N. Lewis in 277 –339.
H.-U. Gremlich, B. Yang (eds.): Infrared and 190. B. F. MacDonald, K. A. Prebble, J. Pharm.
Raman Spectroscopy of Biological Materials, Biomed. Anal. 11 (1993) 1077 – 1085.
Marcel Dekker, New York 2000, p. 231. 191. L. G. Weyer, Appl. Spectrosc. Rev. 21 (1985)
171. D. O. Hummel: Atlas of Polymer and Plastics
1 – 43.
Analysis, 3rd ed., VCH Verlagsgesellschaft,
Weinheim 1991.
46 Infrared and Raman Spectroscopy
192. D. E. Honigs, T. B. Hirschfeld, G. M. Hieftje, 195. M. J. Smith, T. E. May, Int. Lab. 22 (1992)
Appl. Spectrosc. 39 (1985) 1062 – 1065. 18 – 24.
193. E. W. Ciurczak, L. J. Cline-Love, D. M. 196. L. P. McDermott, Ad. Instrum. Control. 45
Mustillo, Spectrosc. Int. 3 (1991) no. 5, (1990) 669 – 677.
39 – 42. 197. Guided Wave, Inc., Application Note
194. E. W. Ciurczak, W. R. Murphy, D. M. Mustillo, no. A3-987, El Dorado Hills, CA 95630.
Spectrosc. Int. 3 (1991) no. 7, 39 – 44. 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
based on silicon – oxygen polymers, including Production. Polysilanes have been synthe-
the majority of rocks, soil, and sand. sized by four methods:
The alternative of utilizing heteroatom back-
bones or noncarbon-based polymers is attractive 1) Reductive coupling
because of the great variety of available reac- 2) Dehydrogenative coupling
tants. Thus, a wide range of compounds can be 3) “Masked” dienes
exploited to obtain materials with specific prop- 4) Anionic ring opening
erties. Furthermore, these materials are of inter- Attempts to prepare polysilanes in the late
est because the resulting bonds may have bond 1940s and 1950s gave insoluble, intractable ma-
energies greater than that of the C−C single terials; the major route employed was the reac-
bond in alkanes (ca. 343 kJ/mol). Approximate tion of dichlorosilanes with sodium (Fig. 1). For
bond energies for other elements, in kilojoules example, polydimethylsilane [28883-63-8] was
per mole, are B−N 389, Si−N 439, Si−O 795, investigated at Union Carbide but found to be
P−N 615, and Be−O 448. Bonding in these ma- insoluble [28]. The key developments leading
terials is a combination of ionic (nondirectional) to soluble polysilanes were the introduction of
and covalent (directional) components; the ionic large substituent groups to decrease order, the
character increases as the electronegativity dif- use of two different substituents, and the co-
ference between adjacent atoms increases. Fur- polymerization of different monomers (Fig. 1)
thermore, the tendency to form multiple bonds [29], [30]. This resulted in high molecular mass
is high for many of the elements involved. The (> 400 000), tractable, soluble polymers, which
thermal stability of many inorganic and organo- could be purified easily, cast into films, and
metallic materials has been described [15]. molded.
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].
Carbon has an unique ability to form strong sin- Improved methods for preparing high molec-
gle, double, and triple bonds to itself. Because ular mass polysilanes are being sought, because
of the relatively high bond energies (C–C sin- the reaction of dichlorosilanes with sodium fre-
gle bonds are of the order of 340 kJ/mol) com- quently gives low yields of high molecular mass
pounds containing carbon in their backbone do polymers and because binodal molecular mass
not readily cleave. By comparison, fellow mem- distributions are obtained [28], [29],[31–35].
bers of group 14 have only recently been synthe- Carrying out the coupling reaction at low tem-
sized because of the low bond energies (gener- perature by using ultrasonication gives monon-
ally on the order of 85 – 125 kJ/mol). Polysilanes odal polymers with low polydispersity [35].
may have been prepared by Kipping in 1924 [2]. Alternating copolymers are formed from the
Only recently have soluble, well-characterized coupling of dichlorodisilanes with dilithium
polysilanes been synthesized [3–5]. (Topics re- salts of 1,2-diethynyldisilanes [36] or dilithio
lated to polysilanes have been recently reviewed
[6–14].)
Inorganic Polymers 3
polythiophenes [36], [37]. These products ex- or deep UV exposure [99–101]. The unusual ab-
hibit σ – π conjugation. Some of these materi- sorption spectra of polysilanes have indicated
als also undergo solid-state transitions to liquid- potential uses in a number of conducting areas.
crystalline mesophases [36]. The copolymer of polysilastyrene, when doped
Water-soluble products are formed by cou- with arsenic pentafluoride, becomes a semicon-
pling silanes with oligomeric poly(ethylene gly- ductor. Conducting films have been reported
col) side chains [38]. [102].
Dehydrogenative coupling of dialkyl and Polysilanes display UV absorption due to sil-
monoalkyl silanes in the presence of transi- icon – silicon bonds of the backbone. For exam-
tion metal catalysts such as titanocene and zir- ple, poly(cyclohexylmethylsilane) exhibits an
conocene dialkyls has given polysilylenes [36– absorption band at 326 nm [30]. This absorption
48]. Functionalized polysilanes have also been is not found in saturated carbon chains. Polysi-
produced. When the side chains contain amine lanes with phenyl side chains show strong ab-
functional groups, the products can be made sorption near 330 nm, resulting from interaction
water-soluble by formation of an amine salt net- between the phenyl groups and the silicon back-
work [49–51]. Dendrimeric [52–54] products bone, which acts as a σ → σ∗ or σ → π∗ chro-
can also be formed by dehydrogenative cou- mophore.
pling. These products exhibit σ conjugation in Polysilanes have a high sensitivity to degra-
many directions and show good conductivity. dation by UV light. After light is absorbed, chain
They have been used as negative-type photore- cleavage occurs with measured quantum yields
sists because exposure to the air or light gives of 0.20 – 0.97. This has led IBM to examine
insoluble products. polysilanes as resists for bilayer UV lithogra-
Polygermanes and polystannanes have also phy [103–105]. They serve as excellent reactive-
been prepared by similar routes [55–58]. Many ion etching barriers for bilevel resist applications
of these materials show better σ conductivity because a protective layer of silicon dioxide is
than the corresponding polysilanes and have formed during exposure of the polymers to oxy-
been used as light-emitting diodes [59]. gen plasmas [104], [105].
The wavelength at which maximum absorp-
Properties and Uses. Much of the interest tion occurs (λmax ) depends on the polysi-
in the polysilanes, polygermanes, and polystan- lane chain length. During irradiation nonlinear
nanes involves their σ delocalization and their bleaching (decrease in UV absorption) occurs,
σ – π delocalization when coupled with arenes because the polymer undergoes chain cleavage
or acetylenes. This is not unexpected since sil- to give shorter fragments of lower absorptiv-
icon exists as a covalent network similar to ity. As the irradiation continues, the polysilanes
diamond. In exhibiting electrical conductivity, become increasingly transparent. This unique
germanium and tin show more typical “metal- bleaching latency is useful in contrast-enhanced
lic” bonding. Some polystannanes have been re- lithography [104], [105].
ferred to as “molecular metals” [60]. Conduc- The λmax for UV absorption also depends
tivity is increased by doping [61], illumination on the conformation of the polysilane backbone
[62], and application of an electric field. [107], [108]. Polysilanes having a planar zigzag
Because of a number of interesting electronic backbone conformation absorb at much longer
and physical properties exhibited by polysilanes, wavelengths than those in which the backbone
a number of potential uses have been suggested is disordered or helical. The solid materials are
and/or shown to exist, including precursors of strongly thermochromic [107]. For example, the
β-SiC fibers [63–71], impregnation of ceramics λmax of poly(di-n-hexylsilane) in solution or im-
[71], [72], polymerization initiators [73], [74], mediately after baking at 100 ◦ C is 317 nm, but
photoconductors for electrophotography [75– after 3 h at 21 ◦ C films exhibit λmax = 371 nm.
78], nonlinear optical materials [79–87], mid- Polysilanes are excellent photocatalysts for
UV photolithography bilayer materials [88–91], making high molecular mass vinyl polymers.
contrast enhancement layers in photolithogra- On irradiation, chain cleavage occurs to give
phy [92], deep UV-sensitive photoresists [90], free radicals [106], which initiate vinyl poly-
[93–100], and self-developing by excimer laser merization and curing. The 3M Company has
4 Inorganic Polymers
commercialized this process. Polysilane initia- nology. The discovery that poly(methylphe-
tors are relatively insensitive to termination of nylsilane) has nonlinear optical properties sug-
polymerization by oxygen. 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
Pioneering work by Yajima, [109] showed well as in electrophotography [115] and reprog-
that poly(dimethylsilane) can be converted into raphy [116]. Although these applications are
β-SiC fibers with very high tensile strength largely in the research stage, the unique all-
(350 kg/mm2 ) by a series of pyrolysis steps. silicon backbone provides an alternative to car-
Thermolysis of poly(dimethylsilane) at 400 – bon and siloxane polymers.
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 3. Poly(Carborane – Siloxanes),
350 ◦ C (cross-linking occurs) to provide rigid-
ity, then pyrolyzed at 1200 ◦ C under nitrogen Polycarbosilanes and Polysilazanes
to give crystalline β-SiC (see also → Fibers, 5.
Synthetic Inorganic, Chap. 3.2.2.). 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).
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
The condensation polymerization of appro-
priate N-silylphosphoramines has resulted in the
successful preparation of the first polyphospha- Polyphosphazenes exhibit some of the low-
zene to have only alkyl substituents on phos- est known rotational barriers for skeletal back-
phorus [151]: 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,
Inorganic Polymers 7
polyphosphazenes are usually soft just above be used as a visually interpretable oxygen sen-
this lower transition temperature, which allows sor. One potential use is to follow differences in
compression molding of films to be carried out. pressure across a surface area such as a wing of
Table 1 lists representative data. a plane.
The thermal stability of some polyphos-
phazenes is outstanding. Several classes ex-
hibit onset of weight loss above 300 ◦ C, as 5. Poly(Boron Nitride)
shown by thermogravimetric analysis (TGA),
but these studies overestimate the stability. Boron nitride [10043-11-5] occurs in three crys-
Poly(diphenoxyphosphazene) [28212-48-8], for talline forms – α, β, and γ (→ Boron Carbide,
example, undergoes rapid depolymerization Boron Nitride, and Metal Borides, Chap. 2.).
above 150 ◦ C. Fluoroalkoxyphosphazene vul- The α-form is a layered, hexagonal polymer
canizates have excellent resistance to solvents much like graphite. A crystalline, tetrahedral
(e.g., lubricants and fuels) and display low network polymer similar to diamond also exists
volume swells after long immersion in lu- [169–172]. Although boron nitride is isoelec-
bricants [159]. Polyphosphazenes also display tronic with the diamond and graphite forms of
good flame-retardant properties, with oxygen in- carbon, it exhibits different properties. For ex-
dex values in the range 20 – 70, and they give off ample, layered boron nitride is white and an elec-
only moderate amounts of smoke [160], [161]. trical insulator. Thus its electrons appear to be
highly localized in contrast to those of graphite.
Uses. The favorable properties of polyphos- Like graphite, layered boron nitride is soft and
phazenes, as well as the high elastomer elon- easily machined. Articles are fabricated by sin-
gations, which are greater than those of poly- tering or machining. These articles can be used
siloxanes from −60 to 200 ◦ C, have resulted in air up to 800 ◦ C or in an inert atmosphere up
in the use of various polyphosphazenes as gas- to 1600 ◦ C. Boron nitride has a Young’s modu-
kets, specialty damping materials, and petro- lus : density ratio of 20 (versus 4 for iron) and
leum piping for arctic applications. They have a tensile strength : density ratio of 8 (versus 1.7
also been used as specialty membrane mate- for glass and 0.3 for iron). The tetrahedral form
rials [150]. Polyphosphazene – salt complexes, of boron nitride is almost as hard as diamond
which have the highest ionic conductivities at and is therefore employed as a substitute for di-
ambient temperature of all polymer – salt elec- amond in both cutting and grinding and in jew-
trolytes, are reported to be semiconducting ma- elry. Industrial applications include its use as an
terials [150]. Other uses include catalytic tem- abrasive, insulator, and refractory material.
plates [150], biomedical materials [160], and The hexagonal modification of boron nitride
their most prominent use as specialty elastomers has been prepared from combinations of cheap
in O-rings, gaskets, and fuel hoses [143], [161– boron and nitrogen compounds, such as B(OH)3 ,
165]. (NH2 )2 CO, and N2 , or KBH4 and NH4 Cl [162–
The first commercial polyphosphazene elas- 167, 169–175]. More recently, chemical vapor
tomers were synthesized and subsequently de- deposition (CVD) methods have used mixtures
veloped by the Firestone Tire and Rubber Co. of BCl3 and NH3 , BF3 and NH3 , or B2 H6 and
[166] and represent the first new class of semi- NH3 to produce hexagonal boron nitride [175,
inorganic elastomers to be developed commer- 178–180].
cially since silicone rubber. Ethyl Corporation Polymeric precursors to boron nitride are
holds the rights to many commercial uses and beginning to attract attention. The reaction
produces both commercial polymers and fin- of difunctionalized borazines with bis(trime-
ished products from this interesting class of ma- thylsilyl)-amines gives soluble oils that form
terials [167]. films which can be pyrolyzed at 1200 ◦ C to
Thin films of amino-substituted polythionyl yield a boron carbide – boron nitride ceramic,
phosphazenes exhibit an unusual permeabil- but removal of all of the carbon has not yet
ity to molecular oxygen [168]. When a phos- been achieved [173]. The use of trifunctional bo-
phorescent dye that is quenched by oxygen is razines (shown below) leads to hexagonal boron
dispersed in the polymer film, the product can
8 Inorganic Polymers
nitride in good yields, with good crystallinity via crystal axis have been measured [184], [185].
intermediates such as 5: 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-
The formation of boron nitride nanocompos- bility and because S4 N4 is explosive.
ites has been reviewed [176]. Further, polymerization is slow (6 – 8 weeks).
Efforts have aimed at developing alternate syn-
thetic routes.
6. Poly(Sulfur Nitride) Poly(sulfur nitride) has been prepared from
solution by addition of trimethylsilyl azide to
The topic of poly(sulfur nitride) has been re- (NSCl)3 and other similar routes [187–189]. In
viewed [177]. these processes, a powdered product is obtained
Poly(sulfur nitride) [56422-03-8] was first rather than the larger crystalline grains. Micro-
prepared in 1910 by Burt, who passed vapors crystalline layers of poly(sulfur nitride) have
of cyclic tetrasulfur tetranitride [28950-34-7], been prepared on platinum electrodes by elec-
S4 N4 , over heated silver gauze or silica wool trolysis of [S5 N5 ]+ [190], [191]. Additional syn-
[181]. However, it was not studied extensively thetic routes have also been found [177]. Even
until the 1970s, when purer products were ob- so, the classical route appears to be the method
tained [182]. Poly(sulfur nitride) films have a of choice for the synthesis of pure poly(sulfur
lustrous golden color and are very striking in nitride).
appearance. Polymeric sulfur nitride is a crys- The fine structure of poly(sulfur nitride) con-
talline, fibrous material that is malleable under tinues to be elucidated [177]. Currently, it is
mild pressure. Electron micrographs reveal that believed the polymer chains are planar and lie
the crystals are composed of fiber layers which along the (crystallographic) b axis that develops
are stacked parallel to one another along the along the a axis of S2 N2 . There are two chains
crystal axis; X-ray studies show that the poly- in each unit cell with two SN units per chain.
mer consists of almost planar chains of alter- Disordered chains are present between the “or-
nating sulfur and nitrogen atoms. All the S−N dered” chains. Secondary interaction between
bond lengths are similar and correspond to a sul- the chains is significant.
fur – nitrogen bond order intermediate between Polymer crystals are stable in air and water
a single and a double bond. The polymer decom- for several days. Stress and strain properties are
poses on heating at ca. 140 ◦ C (see below). like those of an anisotropic metal. The Young’s
The polymer exhibits highly anisotropic elec- modulus parallel to the axis of the chain is on
trical conductivity. Room-temperature conduc- the order of 1010 N m−2 .
tivities as high as 3700 Ω−1 cm−1 along the
Inorganic Polymers 9
Poly(sulfur nitride) becomes superconduc- and crystalline poly(sulfur nitride) can convert
tive at 0.26 K [192]. This property has been ethyne to ethene at rate 106 times that achieved
extensively studied [177]. The onset of super- by chemical systems [212]. Paste electrodes
conductivity T c increases with increasing pres- of poly(sulfur nitride) exhibit metallic electro-
sure up to about 9000 bar, whereupon it de- chemical behavior in nonaqueous liquids [213].
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 7. Polysulfur
to a –SNNS– polymer [193]. Poly(sulfur ni-
tride) exhibits the Meissner – Ochsenfeld ef- Because of its removal from industrial waste to
fect (→ Superconductors, Chap. 2.2.) below meet environmental standards, sulfur has been
about 20 K. Below 4.2 K negative transverse stockpiled and is available in large quantities
magnetoresistance occurs in the presence of at a modest price. Although the stable form at
weak magnetic fields (< 3 T), though a normal room temperature is cyclooctasulfur, S8 , lin-
positive magnetoresistance is found above 77 K ear polysulfur is formed on heating. Unfortu-
or in stronger fields [177]. nately, the thermodynamically stable form of
Thin films have been produced with proper- sulfur is the S8 monomer and the polymer under-
ties that at times are similar to and under certain goes depolymerization after some time. Various
conditions appears to differ from those of the methods have been used to retard depolymer-
crystals themselves. The reason(s) for these dif- ization to the S8 monomer. Removal of minute
ferences are not currently fully understood. The amounts of the monomer through judicious ex-
ability to make large crystals has been reported traction with carbon disulfide retards depoly-
[194], [195]. merization to some extent. Probably the most
Recent interest has focused on electrical ap- effective method involves the addition of olefins,
plications because of its unusual magnetic and such as limonene, cyclopentadiene, and myrcene
electrical properties. Suggested uses include [214]. Some of these result in polysulfur that is
light-emitting diodes, solar cells, and transis- stable for more than five years. Such stabilized
tors [196], [197]. A poly(sulfur nitride) – Ga – polysulfur has been incorporated into concrete
As solar cell with an efficiency of 6.2 % has and asphalt mixes to strengthen them. Concrete
been described [198]. Blue light-emitting diodes blocks, posts, highway pavement, and parking
have been made on ZnS with either gold or tire restrainers that contain stabilized polysulfur
poly(sulfur nitride) barrier electrodes [199]. The are produced.
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
8. Organometallic Polymers
found for a lithium cell with a poly(sulfur ni-
A variety of organometallic polymers have been
tride) electrode [192]. Patents for the use of
synthesized for a number of applications. Ex-
poly(sulfur nitride) have been issued for a va-
amples are shown in Figure 2. The metals may
riety of uses including the cathode in batter-
be an integral part of the polymer chain (6) or
ies [200], [201], coatings on an image record-
side chain (7); they can be connected through
ing sheet [202], etching mask [203–205], ex-
carbon (8) or other atoms (9); they may be
plosive initiator [206], electroconductive resins
bonded by coordination (9), covalent (6–8), or
[207], multilayered wiring patterns [208], sub-
π – metallocene-like bonding (10); or they can
limable polishing grains for semiconductor de-
be part of a linear (6, 8–10) or cross-linked (7)
vices [209], and in light-emitting diodes [210].
product. A classical coordination polymer is de-
Extensive work has been done on poly(sulfur
scribed by 11, and stacked metal-chain com-
nitride) electrodes [177]. A ruthenium(II)-
plexes are depicted by 12, where the circles re-
modified poly(sulfur nitride) electrode gener-
present organic macrocycles such as phthalocya-
ated hydrogen continuously in aqueous sulfu-
nine.
ric acid solution at − 0.1 V under visible light
[211]. Molybdenum(VI) electrodes using film
10 Inorganic Polymers
Applications of ceramers include adhesives lens materials have also been developed [293].
for glass surfaces [297], protective coatings Another area of practical application is the re-
for medieval stained glass [300], and scratch- inforcement of plastics and elastomers [293].
resistant coatings for plastic eyeglass lenses. Recent production of sol – gels with nanoscale
Sol – gel preparations of tetraethoxysilane – pores allows their potential use as catalysts,
water – alcohol – HCl can be spun into fibers porous supports, and as selective absorbents
once the appropriate viscosity has been reached. [293].
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- 10. Inorganic Fibers and Whiskers
C3 H7 O)4 – H2 O – ethanol – HCl gives thin film
coatings of controlled thickness which exhibit Inorganic fibers have been important materials
interference colors that vary with film thickness for over 50 years. Asbestos, glass, and carbon
and they are n-type semiconductors [301]. fibers are particularly well known. Many of the
Hybrid materials have been made by incorpo- materials described herein are discussed in more
rating end-capped poly(tetramethylene oxide) detail under → Fibers, 5. Synthetic Inorganic.
(PTMO) blocks of molecular mass 650 – 2000 Many inorganic fibers exhibit very high ten-
into tetraethoxysilane sol – gel glasses [302]. 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
These materials exhibit high extensibility. and can cause lung cancer and other lung dam-
They contain interdispersed organic polymer age. Asbestos-containing shingles, paper, weld-
and inorganic polymer regions. Although the ing rods, floor tiles, and brake linings are, how-
PTMO blocks are well dispersed, some degree ever, still manufactured. These products avoid
of local phase separation occurs. Titanium iso- fibers of length 5 – 20 µm, which are the most
propoxide has been used to incorporate TiO2 dangerous. Glass fibers are used widely as in-
into these block systems [302]. sulation and in fiber-reinforced plastics. Car-
Thus, although sol – gel chemistry has been bon fibers are of increasing importance in high-
extensively investigated since the early 1960s performance composites. Several metallocene-
[304], [305], only now is it rapidly becoming type condensation polymers (e.g., 21) exhibit
important in material synthesis. fiber formation directly upon polymerization or
upon application of stress [306].
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- Inorganic fibers composed of alumina, sil-
tective coatings are being considered in a num- ica, titania, zirconia, aluminum silicate, boron,
ber of areas. Hot-melt adhesives for glass con- boron carbide, boron nitride, silicon boride, sili-
tainers have been developed [293]. Hard contact con carbide, silicon nitride and others are known.
14 Inorganic Polymers
Polymeric alumina, titania, and zirconia fibers less dense and are, therefore, useful in the con-
are produced industrially [307–310]. Zirconia struction of lightweight composites [318] found
textiles are produced by Union Carbide under in tennis racket frames and aircraft. However,
the tradename Zircar. their cost is high because of production com-
Alumina fibers with a high Al2 O3 content plexities. Processes involve chemical vapor de-
have high temperature stabilities and high mod- position of boron onto carbon fiber and other
uli. They have been made by dispersing alu- cores, followed by a variety of treatments [328].
minum salts on rayon fiber, heating to burn Boron carbide fibers are made from mixtures of
off the organic material, and sintering. Alterna- methane and boron trichloride by using CVD
tively, slurries of hydrated alumina can be ex- to create a boron carbide layer on tungsten core
truded, dried, and heated to give polycrystalline filaments [302]. Boron nitride fibers are used
filaments [311], [312]. Fibers may also be ex- with boron nitride matrices to give composites
truded and drawn from a viscous organic solu- that are employed as electrical insulators which
tion, followed by heating to high temperature are also good thermal conductors. Boron nitride
[313]. Polyaluminoxanes are spun into fibers fiber mats are useful as electrtic cell separators
and sintered to give Al2 O3 fibers, or mixed with in lithium sulfide batteries [302]. The fibers are
silicon compounds, spun, and calcined at 1000 – made from boric oxide fibers that are nitrided
1200 ◦ C to give aluminosilicate fibers. by ammonia at ca. 800 ◦ C followed by stretch-
High-purity silicon dioxide fibers can be used ing at 2000 ◦ C. They have a tensile strength of
up to 1100 ◦ C and are of interest for advanced 2100 MPa and a modulus of 345 GPa.
composites [314]. They are made by melting a Silicon carbide can withstand 1800 ◦ C un-
glass composed of 75 % SiO2 and 25 % Na2 O, der oxidizing conditions. Two methods of pro-
and then spinning and stretching the fiber simul- duction are employed. In the first, CVD is
taneously at 1100 ◦ C. The fibers are then acid- used to deposit silicon carbide on tungsten fil-
leached to remove Na2 O and dried above 300 ◦ C aments. In the second method, polysilanes are
[315]. In another process, silicon alkoxide poly- melt-spun to give a precursor fiber, which on
mers are spun by a sol – gel process to make heat treatment yields fibers 10 – 30 µm in di-
high-purity SiO2 fibers. The fibers are heated ameter (see Chap. 2), [109], [319], [320]. Sili-
to 1000 ◦ C to give a 99.999 % pure quartz fiber con carbide fibers are more resistant to oxida-
which is stable at 1000 ◦ C [316]. tion at high temperature than carbon fibers and
A variety of metal oxide fibers are now have greater compressive strength and electri-
available. Du Pont has developed an alumina – cal resistance. Silicon carbide – carbon fibers,
zirconia fiber with a tensile strength > 2100 MPa can be spun from mixtures of pitch and a
and a modulus of 380 GPa for use in high- poly(dimethylsilane), followed by vacuum heat-
performance composites [317]. Even after expo- ing at 1400 ◦ C [321]. In a similar fashion, melt-
sure to 1400 ◦ C for 100 h, its tensile strength was spinning of a polycarbosilazane followed by
1400 MPa and its modulus remained unchanged. high-temperature treatment gives silicon car-
Stabilized zirconia and alumina – silica – boria bide – nitride fibers with a tensile strength of
fibers have also been produced [315]. The 1.2 MPa and a modulus of 187 MPa, which resist
patent literature reports fibers of such materi- oxidation up to 1200 ◦ C [332].
als as alumina – chromia, thoria (thorium ox- Silicon nitride fibers are used in advanced
ide), titania, titania – silica, thoria – silicate, zir- composites for aircraft, radomes, and electrical
conia – alumina, zirconium silicate, and zirco- components. They are prepared either by reduc-
nia – yttria, as well as whiskers of these and tion of silica in the presence of nitrogen in an
many other metal oxides [318]. electric furnace [315] or by nitriding polycar-
Boron – tungsten fibers have densities of bosilane precursor fibers with ammonia [323].
2.4 – 2.6 g/cm3 and elastic moduli as high Sol – gel polytitanocarbosilanes have been
as 390 GPa. However, they lose both tensile prepared and used to melt-spin a precursor fiber
strength and modulus at high temperature. When which, on pyrolysis in nitrogen at 800 – 1500 ◦ C,
coated with silicon carbide and boron carbide, gives Si – C – Ti – O fibers. At 1300 ◦ C, a tensile
they retain about half their initial strength in strength of 3.0 GPa and a modulus of 220 GPa
air up to 800 – 1000 ◦ C. Boron fibers are much were achieved [324]. Polycarbosilanes can be
Inorganic Polymers 15
treated with titanium tetrabutoxide to give a system. The biomedical applications of poly-
polymer from which fibers may be spun. Af- siloxanes are reviewed in [232]. Monomeric
ter heat treatment, Si – C – Ti – C fibers with a and oligomeric siloxanes are generally not sold
Young’s modulus of 75 – 150 GPa and tensile as medical-grade materials and may contain
strength of 0.1 – 11 GPa were obtained [335]. toxic trace contaminants that are not important
More recently, a number of inorganic materi- in industrial applications but may be critical
als have also been produced as whiskers (see also in biomedical applications. Proper procedures
→ Whiskers). In general, the tensile strengths must be adopted to ensure quality control and
of whiskers are greater than those of fibers. absence of contaminants that may be toxic and
For instance the approximate tensile strength may adversely affect long-term, repeated use of
of graphite in the bulk form is 1000 MPa, for the end product.
fibers it is 2800 MPa, and for whiskers it is Polysiloxanes are extensively employed in
15 000 MPa. Carbon whiskers are produced by biomedical devices as specialty surface treat-
depositing a layer of pyrocarbon from the vapor ments [332–334]. For example, octadecylpoly-
phase on a catalytically grown carbon filament siloxane ( Siliclad, Glassclad 18) is applied to
[325]. hydrophilic surfaces to reduce blood – surface
Inorganic fibers and whiskers exhibit some interactions and protein adsorption [334], [335].
of the highest tensile strengths recorded. They Other polysilane-derived surface treatments
are useful when light weight and high strength are sold under tradenames such as Glassclad
are needed. Uses include turbine blades, heat- HP, D.C. 5700, Glassclad 6C, Dri-Film, and
resistant re-entry vessels, golf club shafts, au- Glassclad IM [232, 340]. Silicone membranes
tomotive radio aerials, fishing poles, dental fill- play critical roles in oxygenation and dialysis.
ings, and parts of aircraft, including stealth air- Polysiloxanes are about ten times more perme-
craft [326]. Many of these inorganic fibers and able to oxygen than low-density polyethylene
whiskers are used as reinforcing agents in com- and a hundred times more permeable than ny-
posites [326]. lon or butyl rubber [336]. Therefore, poly(di-
methylsiloxane) [1646-73-7] soft lenses and
poly(dimethylsiloxane-co-methacrylate) hard
11. Biomedical and Antimicrobial lenses constitute a major portion of the con-
tact lens market. Silicone-based interpenetrating
Polymers polymer networks give clean, resilient gels with
cohesive properties, which are used in breast
Silicones. Silicones (polysiloxanes) are
implants [337–339, 232]. Bilayer membranes
polymers having a silicon – oxygen backbone.
employing a top layer of cured silicone polymer
Among the numerous materials having biomed-
are used as artificial skin [348], [349].
ical applications, silicone polymers play an
Low-consistency, room-temperature vulcan-
important and versatile role. Medical devices
izing silicones cure to give low modulus
made from polysiloxanes include artificial wrist,
mechanical plugs when injected into the fallop-
toe, and finger joints; hip implants; oviductal
ian tubes [350], [351]. This blocks the ovum
plugs [350], [351]; brain membranes; mam-
from reaching the uterus and serves to prevent
mary implants; tracheotomy vents; and artificial
conception. A retrieval ring at the uteral end
hearts. For a detailed description of polysilox-
makes the procedure reversible. Silicone poly-
anes, see → Silicones [328–331]. This chapter
mers are used to make dental impression ma-
deals with their biomedical applications [332].
terials [342–345]. Vinyl addition cured systems
Polysiloxanes are widely applied in medicine
are used in biomedical encapsulation and mold
for numerous reasons including their generally
making [346] and in maxillofacial prostheses.
extremely low toxicity, high gas permeabil-
High-consistency elastomers are used in com-
ity, good mechanical properties, and the wide
pression molding of shunts, flexing joints, and
range of fabrication processes that can be used
catheter tubing [347]. Silicone – urethanes (e.g.,
(e.g., coating, encapsulation, casting, molding,
cardiothane 51) are used in blood pumps, in-
sealing, and extrusion). Furthermore, they are
traaortic balloons, and artificial hearts because
generally inert to the body and to the immune
of their good fatigue strength, flexibility, tough-
16 Inorganic Polymers
ness, and retarded interaction with plasma pro- A variety of boron-containing polymers have
teins [348]. been made by heating trialkyl phosphates with
Silicone – polycarbonate block copolymers boron trichloride at 300 ◦ C [378], [379]. At
can be injection-molded, extruded, and cast into 900 ◦ C this reaction produces boron phosphate
films. Their synthesis is described in [349–351]. glass. The boron polymers form networks prior
They have high tear strength as well as high to glass formation.
oxygen and water permeability. Applications in-
clude blood oxygenation, dialysis, and micro-
electrode materials [352–354].
gives subboric acid, B2 (OH)4 , which when phite bridges [394] between rings have also been
heated at 220 ◦ C loses water to yield a white prepared.
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.
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:
Metal-Containing Polymeric Systems, Plenum 49. P. Bianconi, T. Weidman, J. Amer. Chem. Soc.
Press, New York 1985; C. U. Pittman, Jr., 110 (1988) 2342.
“Vinyl Polymerization of Organic Monomers 50. W. Szymanski, G. Visscher, P. Bianconi,
Containing Transition Metals” in E. J. Becker, Ultrasonics 28 (1990) 310.
M. Tsutsui (eds.): Organometallic Reactions, 51. W. Szymanski, G. Visscher, P. Bianconi,
vol. 6, Wiley-Interscience, New York 1977, Macromolecules 26 (1993) 869.
p. 1; C. U. Pittman, Jr. “Polymer Supported 52. A. Sekignchi, M. Naugo, C. Kabuto, H.
Catalysts” in G. Wilkinson and F. G. A. Stone Sakurai, J. Amer. Chem. Soc. 117 (1995) 4195.
(eds.): Comprehensive Organometallic 53. H. Suzuki, Y. Kimata, S. Satoh, A. Kuriyama,
Chemistry, vol. 8, Chap. 55, Pergamon Press, Chem. Lett. (1995) 293.
Oxford 1982. 54. J. Lambert, J. Pfing, C. Stern, Angew. Chem.
28. J. P. Wesson, T. C. Williams, J. Polym. Sci. Int. Ed. Engl. 34 (1995) 98.
Polym. Chem. Ed. 19 (1981) 65. 55. T. Imori, T. Tilley, J. Chem. Soc. Commun.
29. X. H. Zhang, R. West, J. Polym. Sci. Polym. (1993) 1607.
Chem. Ed. 22 (1984) 159, 225. 56. N. Devylder, M. Hill, K. Molloy, C. Price,
30. R. West, L. D. David, P. I. Djurovich, H. Yu, R. Chem. Commun. (1996) 711.
Sinclair, Am. Ceram. Soc. Bull. 62 (1983) 57. V. Lu, T. Tilley, Macromolecules 29 (1996)
no. 8, 899. 5763.
31. R. West, J. Organomet. Chem. 300 (1986) 58. J. Reichi, C. Popoff, L. Gallagher, E. Remsen,
327. D. Berry, J. Amer. Chem. Soc. 118 (1996)
32. R. West et al., J. Am. Chem. Soc. 103 (1981) 9430.
7352. 59. M. Fujino, T. Hisaki, N. Matsumoto,
33. L. A. Harrah, J. M. Zeigler, J. Polym. Sci. Macromolecules 28 (1995) 5017.
Polym. Lett. Ed. 23 (1985) 209. 60. S. Adams, J. Drager, Angew. Chem. Int. Ed.
34. J. M. Zeigler, Polym. Prepr. Am. Chem. Soc. Engl. 26 (1987) 1255.
Div. Polym. Chem. 28 (1987) no. 1, 424. 61. J. Mark, H. Allcock, R. West: Inorganic
35. K. Matyjaszewski, Y. L. Chen, H. K. Kim: Polymers, Prentice Hall, New York 1992.
“Inorganic and Organometallic Polymers,” 62. H. Suzuki, H. Meyer, J. Simmerer, J. Yang, D.
ACS Symp. Ser. 360 (1988). Naarer, Adv. Mater. 5 (1993) 743.
36. R. West, S. Hayase, T. Iwahara, J. Inorg. 63. R. West, L. David, P. Djurovick, H. Yu, Am.
Organometal. Polym. 1 (1991) 545. Ceram. Soc. Bull. 62 (1983) 899.
37. J. Wilderman, J. Herrema, G. Hadziioannou, 64. M. Kumada, K. Tamao, Adv. Organomet.
E. Schomaker, J. Inorg. Organometal. Polym. Chem. 6 (1968) 19.
1 (1991) 567. 65. R. Barney, G. Chandra: Encyclopedia of
38. C. Van Wakreem, T. Cleiji, J. Zwikker, L. Polymer Science and Engineering, 2nd ed.,
Jenneskens, Macromolecules 28 (1995) 8696. vol. 13, John Wiley & Sons, New York 1988,
39. J. Herrema et al., Macromolecules 28 (1995) p. 312.
8102. 66. C. Schilling, Br. Polymer J. 18 (1986) 355.
40. C. Aitken, J. Harrod, E. Samuel, Organomet. 67. J. Pilot et al., in A. Bassindale, P. Gaspar
Chem. 279 (1985) C11. (eds.): Frontiers in Organosilicon Chemistry,
41. C. Aitken, J. Harrod, E. Samuel, J. Amer. Royal Society of Chemistry, Cambridge 1991.
Chem. Soc. 108 (1986) 4059. 68. W. Schmidt et al., Chem. Mater. 3 (1991) 257.
42. J. Harrod, ACS Symposium Ser. 360 (1988) 89.
69. Y. Shieh, S. Sawan, Polymer Preprints 33
43. H. Woo, T. Tilley, J. Amer. Chem. Soc. 111
(1992) 1042.
(1989) 3757.
70. D. Seyferth et al., Polymer Reprints 34 (1992)
44. T. Tilley, Accounts Chemical Research 26
no. 1, 223.
(1993) 22.
71. D. Seyferth, P. Czubarow, Chem. Mater. 6
45. J. Banovetz, K. Stein, R. Waymouth,
(1994) no. 1, 10.
Organometallics 10 (1991) 3430.
72. S. Yajima, Y. Hasegawa, L. Hayasi, M. Iimura,
46. J. Corey, X. Zhu, T. Bedard, L. Lange,
J. Mater. Sci. 13 (1978) 2569.
Organometallics 10 (1991) 924.
73. R. West, A. Wolff, D. Peterson, J. Radiation
47. H. Li, F. Gauvin, J. Harrod, Organometallics
Curing 35 (1986) 40.
12 (1993) 575.
48. C. Forsyth, S. Nolan, T. Marks, 74. A. Wolf, R. West, Applied Organomet. Chem.
Organometallics 10 (1991) 2543. 1 (1987) 7.
Inorganic Polymers 25
123. M. B. Roller, J. K. Gillham, Polym. Eng. Sci. 8 146. H. R. Allcock, CHEMTECH 5 (1975) 552.
(1974) 567. 147. H. R. Allcock, Angew. Chem. Int. Ed. Engl. 16
124. E. N. Peters et al., J. Polym. Sci. Polym. Phys. (1977) 147.
Ed. 15 (1977) 723. 148. S. H. Rose, J. Polym. Sci. Part. B 6 (1968) 837.
125. L. Interrante et al., J. Amer. Chem. Soc. 116 149. C. Carraher, J. Macromol. Sci. Chem. A17
(1994) 12025. (1982) no. 8, 1293.
126. H. Wu. L. Interrante, Macromolecules 25 150. H. R. Allcock, Polymer Prepr. Am. Chem. Soc.
(1992) 1840. Div. Polym. Chem. 28 (1987) no. 1, 437.
127. I. Rushkin, L. Interrante, Macromolecules 28 151. R. H. Neilson et al., in M. Zeldin, K. J. Wynne
(1995) 5160. and H. R. Allcock (eds.): “Inorganic and
128. M. Lienhard et al., J. Amer. Chem. Soc. 116 Organometallic Polymers,” Chap. 22, ACS
(1994) 12020. Symp. Ser. 360 (1988).
129. M. Lienhard, C. Weigand, T. Apple, B. 152. H. Allcock, J. Nelson, S. Reeves, C.
Farmer, L. Interrante, Polymer Preprints 39 Honeyman, I. Manners, Macromolecules 29
(1998) no. 2, 808. (1996) 7740.
130. S. Yamjima, Y. Hasegawa et al., J. Mater. Sci. 153. J. Nelson, H. Allcock, Macromolecules 30
12 (1978) 2569. (1997) 1854.
131. K. Wynne, R. Rice, Ann. Rev. Mater. Sci. 14 154. H. Allcock, S. Reeves, J. Nelson, C. Crane,
(1984) 297. Macromolecules 30 (1997) 22B.
132. R. Laine, Y. Blum et al., ACS Symp. Ser. 360 155. J. Nelson, H. Allcock, Polymer Reprints 39
(1988) 142. (1998) no. 2, 631.
133. J. Mark, H. Allcock, R. West, Inorganic 156. I. Dez, R. de Jaeger, J. Inorg. Organometal.
Polymers, Prentice Hall, Englewood Cliffs, NJ Polym. 6 (1996) no. 2, 111.
1992, Chap. 6. 157. D. Gates, I. Manners, J. Chem. Soc., Dalton
134. D. Seyferth, G. Wiseman, C. Prud’homme, J. Trans. (1997) 2525.
Am. Ceram. Soc. 66 (1983) C13. 158. H. Allcock, S. Kuharcik, J. Inorg.
135. Rhône-Poulence, EP 197 863 B1, 1988 (J. Organometal. Polym. 5 (1995) no. 4, 307.
Lebrun, H. Porte). 159. J. C. Vicic, K. A. Reynard, J. Appl. Polym. Sci.
136. Bayer, US 3 853 567, 1974 (W. Verbeck). 21 (1977) 3185.
137. Dow Corning, US 4 340 619, 1982 (J. Gaul, 160. R. Singler, G. L. Hagnauer in C. Carraher, J.
Jr.). Sheats, C. U. Pittman, Jr. (eds.):
138. H. Allcock, Adv. Mater. 6 (1994) 106. Organometallic Polymers, Chap. 26,
139. H. Allcock, S. Kuharcik, J. Inorg.
Academic Press, New York 1978.
Organometal. Polym. 6 (1996) no. 1, 1.
161. R. E. Singler, N. S. Schneider, G. L. Hagnauer,
140. P. Bortolus, M. Gleria, J. Inorg. Organometal.
Polym. Eng. Sci. 15 (1975) 321.
Polym. 4 (1994) no. 2, 95 and 4 (1994) no. 3,
162. H. R. Penton, Pol. Prepr. Am. Chem. Soc. Div.
205.
Polym. Chem. 28 (1987) no. 1, 437.
141. J. Inorg. Organometal. polym. 6 (1996) no. 3,
163. C. Kim, H. R. Allcock, Polym. Prepr. Am.
and 6 (1996) no. 4.
Chem. Soc. Div. Polym. Chem. 28 (1987)
142. G. Bosscher, A. Jekel, J. van der Grampel, J.
no. 1, 446.
Inorg. Organometal. Polym. 7 (1997) no. 1,
164. R. Singler, R. Willingham, R. Lenz, A.
19.
Furukana, H. Finkelmann, Polym. Prepr. Am.
143. H. R. Allcock: Phosphorus-Nitrogen
Chem. Soc. Div. Polym. Chem. 28 (1987)
Compounds, Academic Press, New York
no. 1, 448.
1972.
165. G. S. Kyker, T. A. Antkowiak, Rubber Chem.
144. L. Hagnauer, C. Carraher, J. Sheats, C. U.
Technol. 47 (1974) 32.
Pittman, Jr. (eds.): Advances in
166. D. F. Lohr, J. A. Beckman, Rubber and
Organometallic and Inorganic Polymer
Plastics News 16 (1982).
Science, Chap. 17, Marcel Dekker, New York
167. H. R. Penton in M. Zeldin, K. J. Wynne and
1982.
145. R. Neilson, P. Wisian-Neilson in C. E. H. R. Allcock (eds.): “Inorganic and
Carraher, Jr., J. E. Sheats, C. U. Pittman, Jr. Organometallic Polymers,” Chap. 21, ACS
(eds.): Advances in Organometallic and Symp. Ser. 360 (1988).
Inorganic Polymer Sciences, Chap. 19, Marcel 168. Z. Pang et al., Adv. Mater. 8 (1996) 768.
Dekker, New York 1982, p. 425. 169. W. Balmain, Phil. Mag. 21 (1842) 10.
Inorganic Polymers 27
170. R. Wentorf, J. Chem. Phys. 26 (1957) 956. 195. H. Kahlert, B. Kundu, Mater. Res. Bull. 11
171. T. O’Connor, J. Amer. Chem. Soc. 84 (1962) (1976) 967.
1753. 196. R. Scranton, J. Appl. Phys. 48 (1977) 3838.
172. F. Bundy, R. Wentorf, J. Chem. Phys. 38 197. R. Scranton, J. Best, J. McCaldin, J. Vac. Sci.
(1963) 1144. Technol. 14 (1977) 930.
173. Gmelin, Boron Compounds, 2nd Suppl. 1 198. M. Cohen, J. Harris, Appl. Phys. Lett. 33
(1983) 304. (1978) 812.
174. N. J. Archer, Spec. Publ. Chem. Soc. No. 30 199. A. Thomas, J. Woods, Z. Hauptman, J. Phys.
(1977) 167. D: Appl. Phys. 16 (1983) 1123.
175. C. S. Kalyoncu, Ceram. Eng. Sci. Proc. (1985) 200. Asahi Chemical Industry Co. Ltd.,
1356. JP 55 137 671, 1980; 58 93 164, 1983;
176. K. Gonslaves, X. Chen, in J. Salamone (ed.): 58 93 177, 1983; 59 49 155, 1984.
Polymeric Materials Encyclopedia, vol. 5, 201. Electricité de France Service National,
CRC Press, Boca Raton, FL 1996, p. 3259. EP 97 078 B1, 1986 (J. Badoz et al.).
177. A. J. Banister, I. B. Gorrell, Adv. Mater. 10 202. Sony Corp., JP 55 108 944 A, 1980 (T.
(1998) no. 17, 1415. Yoshioka, M. Ogawa, K. Hayashi).
178. T. Takahashi, H. Itoh, A. Takeuchi, J. Cryst. 203. Sony Corp., EP 517 165 A1, 1992; EP 517 165
Growth 47 (1979) 245. B1, 1997; JP 07 230 984, 1955 (S. Kadomura).
179. M. Sano, M. Aoki, Thin Solid Films 83 (1981) 204. Sony Corp., JP 07 263 420, 1995 (T.
247. Nagayama).
180. K. L. Chopra et al., Thin Solid Films 126 205. Sony Corp., JP 07 263 426, 1995 (N. Akiba).
(1985) 307. 206. The United States of America as represented
181. F. P. Burt, J. Chem. Soc. 1910, 1171. by the Secretary of the Army, US 4 206 705,
182. C. Mikulski et al., J. Am. Chem. Soc. 97 1980 (Z. Iqbal, H. Fair, D. Downs).
(1975) 6358. 207. Nippon Denso Co. Ltd., JP 59 129 261, 1984
183. M. Akhtar et al., in C. Carraher, J. Sheats, C. U. (H. Murametsu).
Pittman, Jr. (eds.): Organometallic Polymers, 208. Sony Corp., JP 07 130 735, 1995 (T.
Chap. 30, Academic Press, New York 1978. Nagayama).
184. R. H. Baughman, P. A. Apgar, R. R. Chance, 209. Sony Corp., JP 09 148 285, 1997 (J. Sato).
A. G. MacDiarmid, A. F. Garito, J. Chem. 210. Toshiba Corp., JP 05 136 458, 1993 (T.
Phys. 66 (1977) 401. Kamakura).
185. A. G. MacDiarmid in R. B. King (ed.): 211. H. B. Mark, A. Voulgaropoulos, C. Meyer, J.
Inorganic Compounds with Unusual Chem. Soc., Chem. Commun. (1981) 1021.
Properties, Chap. 6, American Chemical 212. J. Rubinson, T. Behymer, H. Mark, J. Amer.
Society, Washington D.C. 1976. Chem. Soc. 104 (1982) 1224.
186. R. L. Green et al., Lect. Notes Phys. 65 (1977) 213. R. Nowak, C. Joyal, D. Weber, J. Electroanal.
603. Chem. 143 (1983) 413.
187. J. Passmore, M. Rao, J. Chem. Soc., Chem.
214. B. R. Currell, A. J. Williams, A. J. Mooney and
Commun, 1980 1268.
B. J. Nash, Adv. Chem. Ser. 1 (1974) 1.
188. F. Kennett, G. MacLean, J. Passmore, M. Rao,
215. J. Sheats, C. Carraher, C. Pittman, M. Zeldin,
J. Chem. Soc., Dalton Trans 1982 851.
B. Currell: Inorganic and Metal-Containing
189. A. Banister, Z. Hauptam, J. Passmore, C.
Polymeric Materials, Plenum, New York
Wong, P. White, J. Chem. Soc., Dalton Trans
1990.
1986 2371.
216. C. Pittman, C. Carraher, M. Zeldin, B.
190. A. Banister, Z. Haupton, A. Kendrick, J.
Culbertson, J. Sheats: Metal-Containing
Chem. Soc., Chem. Commun. 1983 1016.
Polymeric Materials, Plenum, New York
191. A. Banister, Z. Hauptman, A. Kendrick, R.
1996.
Small, J. Chem. Soc., Dalton Trans 1987 915.
217. M. Zeldin, K. Wynne, H. Allcock: Inorganic
H. Fritz, R. Bruchhaus, Z. Naturforsch. B. 38
and Organometallic Polymers, ACS,
(1983) 1375.
Washington, DC 1998.
192. R. L. Greene, G. B. Street, L. J. Suter, Phys.
218. Macromolecule-Metal Complexes III,
Rev. Lett. 34 (1975) 577.
Seclected Papers from seminar of July 1989; J.
193. W. H. Jones, R. Bardo, J. Phy. Chem. 97
Macromol. Sci. Chem. A27 (1990) nos. 9 – 11,
(1993) 4974.
194. I. Kakada, J. Cryst. Growth 55 (1981) 447. 1109.
28 Inorganic Polymers
219. I. Manners, Ann. Rep. Prog. Chem., Sect. A. Metal-Containing Polymeric Materials,
Inorg. Chem. 88 (1991) 77; 89 (1992) 93; 90 Plenum, NY 1996.
(1993) 103; 91 (1994) 131; 92 (1995) 127; 93 239. C. Carraher et al., Polym. Mater. 55 (1986)
(1996) 129. 469.
220. J. Mark, H. Allcock, R. West, Inorganic 240. C. Carraher, V. Foster, R. Linville, D. Stevison,
Polymers, Prentice Hall, New York 1992. Polym. Mater. 56 (1987) 401.
221. J. Sheats, C. Carraher, C. Pittman, M. Zeldin, 241. C. Carraher, R. Schwarz, J. Schroeder, M.
in C. Craver, C. Carraher (eds.): Applied Schwarz, H. M. Molloy, Org. Coat. Plast.
Polymer Science, Elsevier, New York 1999. Chem. 43 (1981) 798.
222. S. Takahashi, K. Sonogashira, in J. Salamone 242. C. Carraher, R. Schwarz, J. Schroeder, M.
(ed.): Polymeric Materials Encyclopedia, vol. Schwarz in C. Carraher, J. Preston (eds.):
6, CRC Press, Boca Raton, FL 1996, p. 4804. Interfacial Synthesis, vol. 3, Chap. 6, Marcel
223. A. D. Hunter, A. Guo, in J. Salamone (ed.): Dekker, New York 1982.
Polymeric Materials Encyclopedia, vol. 6, 243. C. Carraher et al., Org. Coat. Plast. Chem. 44
CRC Press, Boca Raton, FL 1996, p. 4813. (1982) 753.
224. T. Endo, I. Tomita, in J. Salamone (ed.): 244. J. Wrobleski, D. B. Brown, Inorg. Chem. 18
Polymeric Materials Encyclopedia, vol. 6, (1979) 498, 2738.
CRC Press, Boca Raton, FL 1996, p. 4822. 245. H. Matsuda, H. Nakaniski, M. Kato, J. Polym.
225. R. Wei, L. Ya, W. Jinguo, X. Qifeng, in, J. Sci. Polym. Lett. Ed. 22 (1984) 107.
Salamone (ed.): Polymeric Materials 246. J. E. Sheats, C. Carraher, C. Pittman (eds.):
Encyclopedia, vol. 6, CRC Press, Boca Raton, Metal-Containing Polymeric Systems, Plenum
FL 1996, p. 4826. Press, New York 1985.
226. W. L. Driessen (ed.): Macromolecule-Metal 247. X. Zhang, R. West, J. Polym. Sci. Polym.
Complexes, Huthig & Wepf, Oxford, CT 1998. Chem. Ed. 22 (1984) 159.
227. F. Ciardelli, E. Tsuchida, D. Wohrle (eds.): 248. R. Liepins, M. Timmons, N. Morosoff, J.
Macromolecule-Metal Complexes, Surles in J. Sheats, C. Carraher, C. Pittman
Springer-Verlag, Berlin 1996. (eds.): Metal-Containing Polymeric Systems,
228. A. Pomogailo, V. Savost’yanov: Synthesis and Plenum Publishing, New York 1985,
Polymerization of Metal-Containing pp. 225 – 236.
Monomers, CRC Press, Boca Raton, New 249. H. Yasuda et al., in J. Sheats, C. Carraher,
York 1994. C. U. Pittman (eds.): Metal-Containing
229. B. P. Block, Inorg. Macromol. Rev. 1 (1970) Polymeric Systems, Plenum Publishing, New
no. 2, 115. York 1985, pp. 275 – 290.
230. C. Carraher, R. Schwarz, J. Schroeder, M. 250. C. Carraher et al., Organic Coatings Plastics
Schwarz, J. Macromol. Sci. Chem. A15 Chemistry 44 (1981) 753.
(1981) no. 5, 773. 251. E. Savinova et al., J. Mol. Catal. 32 (1985)
231. E. Neuse, H. Rosenberg: Metallocene 149.
Polymers, Marcel Dekker, New York 1970. 252. M. Kaneko, A. Yamada, Adv. Polym. Sci. 55
232. Dow Corning, GB 1 582 081, 1980. (1983).
233. C. Carraher, F. Li, D. Siegman-Louda, C.
253. M. S. Wrighton (ed.): Interfacial
Butler, J. Ross, PMSE 77 (1997) 499.
Photoprocesses: Energy Conversion and
234. C. Carraher, V. Saurino, C. Butler, D. Sterling,
Synthesis, American Chemical Society,
PMSE 72 (1995) 192.
Washington 1980.
235. C. Carraher, D. Sterling, C. Butler, T.
254. U. T. Müller-Westerhoff, A. I. Nazzal, US
Ridgway, Biotechnical Polymers, Chap. 13,
4 379 740, 1983.
Technomic, Lancaster, PA 1995.
255. F. C. Anson et al., J. Am. Chem. Soc. 106
236. C. Carraher, M. Nagata, H. Stewart, S. Miao,
(1984) 59.
S. Carraher, A. Gaonkar, C. Highland, F. Li,
256. E. R. Savinova et al., J. Mol. Catal. 32 (1985)
PMSE 79 (1998) 52.
149, 159.
237. C. Carraher, A. Gaonkar, H. Stewart, S. Miao,
257. in [23] pp. 137 – 147.
D. Mitchell, C. Barosy, M. Colbert, R.
258. B. J. Spalding, Chemical Week 135 (1986)
Duffield, Polymer Reprints 38 (1997) no. 2,
no. Oct. 8, 29.
572.
238. C. Carraher, H. Stewart, W. Soldani, J. Dela 259. C. Carraher, S. Tsuji, J. DiNunzio, W. Feld,
Torre, B. Pandya, L. Reckleben, Polym. Mater. 55 (1986) 875.
Inorganic Polymers 29
260. C. Carraher, J. Schroeder, Polym. Prepr. 16 283. C. Carraher, H. Zhuang, F. Medina, D. Baird,
(1975) 659, J. Polym. Sci. Polym. Lett. Ed. 13 R. Pennisi, B. Landreth, F. Nounou, PMSE 62
(1975) 215. (1990) 633.
261. L. Donaruma, Polym. Prepr. Am. Chem. Soc. 284. M. Kaneko, A. Yamada, Adv. Polym. Sci.
Div. Polym. Chem. 22 (1981) 1. (1983) 55.
262. B. Jiang et al., Coord. Chem. Rev. 171 (1998) 285. US 4 379 740, 1983 (U. Muller-Westerhoff, A.
365. Nazzal).
263. W. E. Jones, L. Hermans, B. Jiang, in V. 286. F. Anson, J. Amer. Chem. Soc. 106 (1984) 59.
Ramamurthy, K. Schanze (eds.): Molecular 287. B. J. Spaulding, Chemical Week, Oct 8
and Supramolecular Photochemistry, vol. 2, (1986)29 and C. M. Elliott, Abstracts, 193 ACS
Marcel Dekker, New York 1999. Meeting, Denver CO, April, 1987, INOR 232.
264. C. Pittman, J. Paint Technol. 39 (1967) 585. 288. M. MacLachlan, P. Aroca, N. Coombs, I.
265. N. E. Wolff, R. J. Pressley, Appl. Phys. Lett. 2 Manners, G. Ozin, Adv. Mater. 10 (1998) 144.
(1963) 152. 289. Proceedings of the IIIrd Internat. Conference
266. in [17] p. 425. on Ultrastructure Processing of Glasses,
267. J. E. Sheats et al., in B. M. Culbertson, C. U. Ceramics and Composites, Feb. 24 – 27, 1987.
Pittman, Jr. (eds.): New and Unusual 290. C. J. Brinker, D. Clark, D. R. Ulrich (eds.):
Monomers and Polymers, Plenum Press, New Better Ceramics Through Chemistry II,
York 1983, pp. 83 – 98; Polym. Mater. Sci. Eng. Materials Res. Soc., Pittsburgh, Pa., 1986.
49 (1983) no. 2, 363. 291. L. L. Hench, D. R. Ulrich (eds.): Science of
268. N. Bilow, H. Rosenberg, J. Polym. Sci. Part Ceramic Chemical Processing,
A-1 7 (1969) 2689. Wiley-Interscience, New York 1986.
269. N. Bilow, H. Rosenberg, US 3 640 961, 1972. 292. J. O’Reilly, B. Coltrain, in J. Salamone (ed.):
270. C. U. Pittman, Jr., R. F. Felis, J. Organomet. Polymeric Materials Encyclopedia, vol. 6,
Chem. 72 (1974) 389. CRC Press, Boca Raton, FL 1996, p. 4772.
271. C. U. Pittman, R. F. Felis, J. Organomet. 293. G. Wilkes, J. Wen, in J. Salamone (ed.):
Chem. 72 (1974) 399. Polymeric Materials Encyclopedia, vol. 6,
272. R. Tannenbaum, E. P. Goldberg, C. L. CRC Press, Boca Raton, FL 1996, p. 4782.
Flenniken in J. Sheats, C. Carraher, C. Pittman 294. Y. Chujo, in J. Salamone (ed.): Polymeric
(eds.): Metal-Containing Polymeric Systems, Materials Encyclopedia, vol. 6, CRC Press,
Plenum Publishing, New York 1985, Boca Raton, FL 1996, p. 4793.
pp. 303 – 340. 295. C. J. Brinker et al., “Inorganic and
273. C. Carraher, X. Xu, PMSE 75 (1996). Organometallic Polymers,” ACS Symp. Ser.
274. C. Carraher, X. Xu: Modification of Polymers, 360 (1988); D. R. Ulrich, CHEMTECH 18
Plenum, New York 1997. (1988) 242.
275. C. Carraher, V. Forster, R. Linville, D. 296. H. K. Schmidt, “Inorganic and Organometallic
Steverson, R. Ventatachalam, Adhesives, Polymers,” ACS Symp. Ser. 360 (1988).
Sealants, and Coatings for Space and Harsh 297. H. Schmidt, H. Scholze, G. Tunker, J. Non
Environments, Chap. 19, Plenum, NT 1988. Cryst. Solids 80 (1986) 557.
276. C. Carraher, A. Taylor-Murphy, PMSE 76 298. H. Schmidt, H. Scholze, Springer Proceedings
(1997) 409. in Physics, vol. 6, Aerogels, Springer
277. S. Yajima, H. Hayashi, M. Omori, Chem. Publishing Co., 1986, p. 49.
Letters. (1975) 931 and 1209. 299. G. Philipp, H. Schmidt, J. Non Cryst. Solids
278. R. West, Am. Ceram. Soc. Bull. 62 (1983) 825. 63 (1984) 261.
279. D. Seyferth, G. H. Wiseman, J. Amer. Ceram. 300. G. Tunker, H. Patzelt, H. Schmidt, H. Scholze,
Soc. 67 (1984); Massachusetts Institute of Glastechn. Ber. 59 (1986) 272.
Technology, US 4 482 669, 1984 (D. Seyferth, 301. S. Sakka, Polym. Prepr. Am. Chem. Soc. Div.
G. H. Wiseman). Polym. Chem. 28 (1987) no. 1, 430.
280. R. Laine, Y. Blum, A. Chow, R. Hamlin, K. 302. P. Brooke, H. Schurmans, J. Verhoest:
Schwartz, D. Rowecliff, Polymer Reprints 28 Inorganic Fibers and Composite Materials: A
(1987) no. 1, 393. Survey of Recent Developments, Pergamon
281. M. Arai, S. Sakurada, T. Isoda, T. Tomizawa, Press, Oxford, UK, 1984.
Polymer Reprints 28 (1987) no. 1, 407. 303. H. Huang, R. H. Glaser, G. L. Wilkes,
282. C. Narula, R. Paine, R. Schaeffer, Polymer “Inorganic and Organometallic Polymers,”
Reprints 28 (1987) no. 1, 454. ACS Symp. Ser. 360 (1988).
30 Inorganic Polymers
304. H. Dislich, Angew. Chem. Int. Ed. Engl. 106 Encyclopedia, vol. 2, CRC Press, Boca Raton,
(1970) 363. FL 1996, p. 973.
305. B. E. Yoldas, J. Mater. Sci. 12 (1977) 203; 14 326. C. Carraher: Polymer Chemistry: An
(1979) 1843. Introduction, 4th ed., Dekker, New York 1996,
306. C. Carraher, CHEMTECH 1972, 741. p. 122.
307. V. C. Farmer, A. R. Fraser, J. M. Tait, J. Chem. 327. K. Okamura, Composites 18 (1987) 107.
Soc. Chem. Commun. 1977, 462. 328. W. Noll: Chemistry and Technology of
308. K. Wada et al., Clay Miner. 8 (1970) 487. Silicones, Academic Press, New York 1968.
309. G. Alberti, U. Costantino, J. Chromatogr. 102 329. W. Lynch: Handbook of Silicon Rubber
(1974) 5. Fabrication, Van Nostrand, New York 1978.
310. N. H. Ray: Inorganic Polymers, Chap. 7, 330. M. Ranney: Silicones, Noyes Data Corp., Park
Academic Press, New York 1978. Ridge, N.J., 1977.
311. J. D. Birchall, J. A. Bradbury, J. Dinwoodie in 331. B. Arkles, W. Peterson, R. Anderson:
W. Watt, B. V. Perov (eds.): Handbook of Organosilicon Compounds: Register and
Composites, North Holland, Amsterdam Review, Petrarch Systems, Bristol, Pa., 1982.
1985, p. 115. 332. B. Arkles, CHEMTECH 13 (1983) 542.
312. J. D. Birchall: “Inorganic Fibers” in M. B. 333. A. Isquith, L. Abbot, P. Walters, Appl.
Bever (ed.): Encyclopedia of Material Microbiol. 24 (1972) 859.
Sciences and Engineering, Pergamon Press, 334. B. Arkles, W. Brinigar, A. Miller in D.
Oxford, UK 1986, p. 2333. Leyden, W. Collin (eds.): Silylated Surfaces,
313. Sumitomo Chemical Co., US 4 101 615, 1973 Gordon Breach, New York 1980.
(S. Horikiri, K. Tsuji, Y. Abe); 4 152 149, 1977 335. E. Sako, O. Kawamura, US 3 746 196, 1973.
(S. Horikiri, K. Tsuji, Y. Abe). 336. W. Robb, Ann. N.Y. Acad. Sci. 146 (1968) 119.
314. J. Tretzel et al., “Production, Processing and 337. M. Nelson, US 3 020 260, 1962.
Application of Enka Silica Fibers,” 57th 338. E. Jeram, US 4 072 635, 1978.
Annual TRI Conference, Charlotte, N.C., April 339. T. Cronin, US 3 293 663, 1965.
21 – 22, 1987. 340. I. V. Yannas, J. F. Burke, J. Biomed. Mater.
315. W. C. Miller in M. Grayson (ed.): Res. 14 (1980) 65, 107.
Encyclopedia of Textiles, Fibers and 341. N. Dagalakis et al., J. Biomed. Mater. Res. 14
Nonwoven Fabrics, J. Wiley and Sons, New (1980) 511.
York 1984, p. 443. 342. R. A. Erb, US 4 245 623, 1981.
316. Chem. Week 13 (1987) no. April 22. 343. R. A. Erb, T. P. Reed, J. Reprod. Med. 23
317. J. C. Romine, Ceram. Eng. Sci. Proc. 8 (1987) (1979) 65.
755. 344. A. Zaffaroni, US 3 905 360, 1975.
318. J. O. Carlsson: “Boron Fibers” in M. B. Bever 345. P. Hittmaier, W. Hechtel, H. Eckhart, US
(ed.): Encyclopedia of Material Science and 4 035 453, 1977.
Engineering, Pergamon Press, Oxford, UK, 346. C. N. Raptis, R. Yu, J. G. Knapp, J. Prosthet.
1986, p. 402. Dent. 4414 (1980) 447.
319. J. O. Carlsson: “Silicon Carbide Fibers” in
347. A. B. Swanson, US 3 875 594, 1975.
M. B. Beuer (ed.): Encyclopedia of Material
348. T. Honda et al., J. Thorac. Cardiovasc. Surg.
Science and Engineering, Pergamon Press,
69 (1975) 92.
Oxford, UK, 1986, p. 4406.
349. D. Laurin, US 3 994 988, 1976.
320. S. Yajima, Y. Hasegawa, J. Hayashi, M. Imura,
350. D. G. LeGrand, J. Magila, Eng. Sci. 10 (1970)
J. Mater. Sci. 13 (1978) no. 12, 2569.
349.
321. Y. Hasegawa, K. Okamura, Yogyo Kyokaishi
351. J. Riffle, R. Freelin, A. Banthia, J. McGrath, J.
95 (1987) 99, 103.
Macromol. Sci. Chem. A15 (1981) 967.
322. W. Verbeck, DE-OS 2 218 960, 1973.
352. E. Pierce, N. Dibelus, Trans. Am. Soc. Artif.
323. K. Okamura, M. Sato, Y. Hasegawa, Ceram.
Intern. Organs 14 (1968) 220.
Int. 13 (1987) 55.
353. J. Montalvo, US 3 869 354, 1975.
324. T. Yamamura et al., “Proc. World Congress on
354. J. Brown, G. Slusarczuk, O. LeBlanc, US
High Tech Ceramics, 6th Int. Meeting on
3 743 588, 1971.
Modern Ceramics Technologies,” Mater. Sci.
355. H. Allcock in C. Carraher, J. E. Sheats, C. U.
Monogr. 38 (1987).
325. N. Tsubokawa, T. Yoshihara, H. Ueno, in J. Pittman, Jr. (eds.): Organometallic Polymers,
Salamone (ed.): Polymeric Materials Academic Press, New York 1978, p. 283.
Inorganic Polymers 31
356. H. Allcock, R. Allen, J. O’Brien, J. Chem. Soc. 378. W. Gerrard, J. Chem. Soc. 1960, 3170.
Chem. Commun. 1976, 717. 379. R. Bedell, M. J. Frazer, W. Gerrard, J. Chem.
357. C. Carraher, W. J. Scott, J. Schroeder, D. J. Soc. 1960, 4037.
Giron, J. Macromol. Sci. Chem. A15 (1981) 380. R. W. Upson, US 2 511 310, 1950.
no. 4, 625. 381. K. Hizawa, E. Nojimoto, JP 4791, 1953.
358. C. Carraher et al., in C. Gebelein, C. Carraher 382. J. M. Thompson: “Other Boron-Containing
(eds.): Polymeric Materials in Medication, Polymers” in M. F. Lappert, G. J. Leigh (eds.):
Chap. 15, Plenum Publishing, New York 1985. Developments in Inorganic Polymer
359. C. Carraher in C. Gebelein, C. Carraher (eds.): Chemistry, Elsevier, Amsterdam 1962, p. 57.
Bioactive Polymeric Systems, Chap. 22, 383. R. M. Washburn, E. Levens, C. F. Albright,
Plenum Publishing, New York 1985. F. A. Billig: “Metal-Organic Compounds,”
360. C. Carraher, I. Lopez, D. Giron, Polym. Mater. Adv. Chem. Ser. 23 (1959) 139, 153.
53 (1985) 644. 384. W. Kuchen, R. Brinkmann, Angew. Chem. 72
361. C. Carraher, I. Lopez, D. Giron in C. Gebelein (1960) 564.
(ed.): Advances in Biomedical Polymers, 385. A. B. Burg, Angew. Chem. 72 (1960) 183.
Plenum Publishing, New York 1987, p. 311. 386. R. I. Wagner, F. F. Caserio, J. Inorg. Nucl.
362. C. Carraher, N. Bigley, M. Trombley, D. Chem. 11 (1959) 259.
Giron, Polym. Mater. 57 (1987) 177. 387. T. Wartik, E. F. Apple, J. Am. Chem. Soc. 77
363. A. Winston, in C. Gebelein, C. Carraher (eds.): (1955) 6400.
Bioactive Polymeric Systems, Chap. 21, 388. H. Rosenberg, F. L. Hedberg, Tech. Rept.
Plenum Publishing, New York 1985. AFML-TR-69-68, AD 863, 817 (1969).
364. A. Winston, Polym. News 10 (1984) 6. 389. H. Steinberg, R. J. Brotherton: Organoboron
365. C. Carraher et al., in C. Carraher, C. Gebelein Chemistry, vol. 2, J. Wiley and Sons, New
(eds.): Biological Activities of Polymers, York 1966, pp. 329 – 340, 363 – 364, 418.
Chap. 2, American Chemical Society, 390. B. M. Mikhailov, A. F. Galkin, Bull. Acad. Sci.
Washington, D.C., 1982. USSR, Div. Chem. Sci. (Engl. Transl.) 12
366. D. Anderson, J. Mendoza, B. Garg, R. V. (1963) 575.
Subramanian in C. Carraher, C. Gebelein 391. R. H. Toeniskoetter, R. Didchenko, CA
(eds.): Biological Activities of Polymers, 685 580, 1964 to Union Carbide.
Chap. 3, American Chemical Society, 392. R. H. Toeniskoetter, K. A. Killip, J. Am. Chem.
Washington, D.C., 1982. Soc. 86 (1964) 690.
367. R. V. Subramanian, B. Garg, J. Corredor in C. 393. V. Gutmann, A. Metter, R. Schlegel, Monatsh.
Carraher, J. Sheats, C. U. Pittman, Jr. (eds.): Chem. 94 (1963) 1071.
Organometallic Polymers, Chap. 19, 394. G. Chainani et al., J. Appl. Chem. 15 (1965)
Academic Press, New York 1978. 372.
368. C. Carraher et al., Polym. Mater. 57 (1987) 395. K. O. Knollmuller, R. N. Scott, H. Kwasnik,
186. J. F. Sieckhaus, J. Polym. Sci. Polym. Chem.
369. C. Carraher, D. Giron, J. Schroeder, C. Ed. A1, 9 (1971) 1071.
McNeely, US 4 312 981, 1982. 396. S. Papetti et al., J. Polym. Sci. Polym. Chem.
370. J. A. Montemarano, E. J. Dyckman, J. Paint Ed. 4 (1966) 1623.
Technol. 47 (1975) 59. 397. R. P. Alexander, H. Schroeder, Angew. Chem.
371. A. T. Phillip, Prog. Org. Coat. 2 (1973/74) 76 (1964) 278.
159. 398. H. A. Schroeder et al., Inorg. Chem. 8 (1969)
372. J. Leebrick, US 3 167 473, 1965. no. 11, 2444.
373. D. Atherton, J. Verborgt, M. A. Winkeler, J. 399. H. A. Schroeder, Inorg. Macromol. Rev. 1
Coat. Technol. 51 (1979) 88. (1970) 45.
374. S. Matsuda, H. Kudara, US 4 174 339, 1979. 400. I. B. Atkinson, D. B. Clapp, W. F. Flavell,
375. Shipcare International, 10 (1978) no. 11, 23. Polym. Prepr. Am. Chem. Soc. Div. Polym.
376. R. W. Drisko, T. B. O’Neill, L. K. Schwab, Chem. 13 (1972) 770.
Technical Note no. N-1480, Naval Facilities 401. N. Semenuk, S. Papetti, H. Schroeder, Inorg.
Engineering Command, Point Hueneme, CA, Chem. 8 (1969) no. 11, 2441.
May 1977. 402. J. Rinse, Ind. Eng. Chem. 56 (1964) no. 5, 42.
377. Y. Chujo, in J. Salamone (ed.): Polymeric 403. DuPont, US 2 744 071, 1956 (C. N. Theobald).
Materials Encyclopedia, vol. 6, CRC Press, 404. K. C. Pande, R. C. Mehrotra, Z. Anorg. Allg.
Boca Raton, FL 1996, p. 4795. Chem. 286 (1956) 291.
32 Inorganic Polymers
405. Peter Lunt and Co., GB 573 083, 1945 (E. J. 432. E. Whalley, Can. J. Chem. 38 (1960) 2105.
Lush). 433. Y. Okamoto, Z. Iqbal, R. H. Baughman, J.
406. Hardman and Holden Ltd., GB 772 144, 1957 Macromol. Sci. Chem. A25 (1988) no. 5 – 7,
(J. Rinse). 799.
407. A. L. McCloskey et al., WADC Tech. Rept. 434. W. S. Chan, A. K. Jonsher, Phys. Status Solidi
59-761 (1959) pp. 117 – 200. 32 (1969) 749.
408. Koninklijke Industrieele Maatschappij 435. J. Tsukamoto, A. Takahashi, Jpn. J. Appl.
voorheen Noury, GB 783 679, 1957 (N. V. Van Phys. 25 (1986) L338.
der Lande). 436. M. Berthelot, Ann. Chim. Phys. 11 (1936) 15.
409. K. A. Andrianov: Metalorganic Polymers, 437. Y. Asano, Jpn. J. Appl. Phys. 22 (1983) 1618.
Chap. 5, Interscience Publishers, New York 438. L. A. Wall, D. W. Brown, J. Polym. Sci. Part C
1965. 4 (1964) 1151.
410. K. A. Andrianov, V. M. Novikov, Vysokomol. 439. Z. Iqbal et al., J. Chem. Phys. 85 (1986) 4019.
Soedin. 1 (1959) 1390. 440. G. Grimm, H. Metzger, Ber. Dtsch. Chem.
411. V. Kugler, J. Polym. Sci. 29 (1958) 637. Ges. 69 (1936) 1356.
412. T. R. Patterson et al., J. Am. Chem. Soc. 81 441. A. J. Brown, E. Whalley, Inorg. Chem. 7
(1959) 4213. (1968) 1254.
413. J. I. Jones: “Polymetallosiloxanes” in M. L. 442. Y. Okamoto, P. S. Wojciechowski, J. Chem.
Lappert, G. J. Leigh (eds.): Developments in Soc. Chem. Commun. 1982, 386.
Inorganic Polymer Chemistry, Part I, Part II, 443. H. Kobayashi et al., Chem. Lett. 1983, 1407.
Chap. 7, 8, Elsevier, New York 1962. 444. H. Kobayashi, A. Kobayashi, Y. Sasaki, Mol.
414. D. C. Bradley: “Polymeric Metal Alkoxides, Cryst. Liq. Cryst. 118 (1985) 427.
Organometalloxanes and 445. A. Kobayashi et al., Bull. Chem. Soc. Jpn. 59
Organometalloxanosiloxanes” in F. G. A. (1986) 3821.
Stone, W. A. G. Graham (eds.): Inorganic 446. H. Emeleus, J. Anderson: Modern Aspects of
Polymers, Chap. 7, Academic Press, New York Inorganic Chemistry, 3rd ed.,
1962. Routledge & Kegan, London 1960.
415. I. K. Stavitskii, Vysokomol. Soedin. 1 (1959) 447. F. Cataldo, J. Inorg. Organomet. Polym. 7
1502. (1997) no. 1, 35.
416. S. N. Borisov, N. G. Sviridova, Vysokomol. 448. Y. L. Gefter: Organophosphorus Monomers
Soedin. 3 (1961) 50. and Polymers, Pergamon Press, London 1962.
417. R. D. Crain, P. E. Koenig, WADC Tech. Rept. 449. H. Noth, W. Schragle, Z. Naturforsch. B 166
69-427 (1960) p. 15. (1961) 473.
418. F. A. Henglein, R. Lang, L. Schmack, 450. C. U. Pittman, Jr., J. Polym. Sci. Polym. Chem.
Makromol. Chem. 22 (1957) 103. Ed. 5 (1967) 371.
419. Du Pont, US 2 512 058, (H. C. Gulledge). 451. C. Shimasaki, H. Kitano, in J. Salamone (ed.):
420. T. Boyd, GB 728 751, 1955.
Polymeric Materials Encyclopedia, vol. 7,
421. T. Boyd, US 2 716 656, 1955.
422. A. Hancock, R. Sidlow, J. Oil Colour Chem. CRC Press, Boca Raton, FL 1996.
452. Phosphorus, Sulfur, Silicon, Relat. Elem.,
Assoc. 35 (1952) 28.
423. W. H. Post, W. T. Schwartz, Jr., AD 255 545, ongoing journal.
U.S. Dept. Commerce Office Tech. Serv., 453. D. Corbridge: “Studies in Inorganic Chemistry
Washington D.C., 1961. 10: Phosphorus, An Outline of Its Chemistry,”
424. K. A. Andrianov, S. V. Pichkhadze, Biochemistry and Technology, 4th ed.,
Vysokomol. Soedin. 3 (1961) 577. Elsevier, Amsterdam 1990.
425. C. Ungurenasu, I. Haidec, Rev. Roumaine 454. G. Borissov, in J. Salamone (ed.): Polymeric
Chim, 13 (1968) 957. Materials Encyclopedia, vol. 7, CRC Press,
426. C. C. Currie, US 2 672 455, 1954. Boca Raton, FL 1996, p. 5092.
427. Monsanto Inc., GB 723 989, 1955. 455. D. Liaw, D. Wang, in J. Salamone (ed.):
428. R. M. Kary, US 2 646 440, 1953. Polymeric Materials Encyclopedia, vol. 9,
429. W. P. Gill et al., Phys. Rev. Lett. 38 (1977) CRC Press, Boca Raton, FL 1996, p. 6569.
1305. 456. E. W. Neuse, G. J. Chris, J. Macromol. Sci.
430. F. B. Bundy, K. T. Dunn, J. Chem. Phys. 71 A-1 (1967) 371.
(1979) 1550. 457. H. R. Allcock: “Rings, Clusters and Polymers
431. P. W. Bridgman, Proc. Am. Acad. Arts Sci. 74 of the Main Group Elements,” ACS Symp. Ser.
(1941) 399. 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 [1] eye; about 70 000 of its victims are considered
blind. Other important human diseases trans-
Insects attack humans and domestic animals; act mitted by insects and related arthropod vec-
as vectors for human, animal, and plant diseases; tors include plague (causative organism Yersinia
destroy structures (man-made buildings, etc.); pestis) transmitted by the asiatic rat flea Xenop-
and compete for supplies of food and fiber. The sylla cheopis and other fleas, endemic typhus
ca. 106 insect species that have been described (Rickettsia prowazeki) spread by the human
represent about 70 % of all animal species. At body louse Pediculus humanus humanus, and
least 600 insect species are important pests in scrub typhus (R. orientalis) spread by Trombic-
the United States; the world total exceeds 10 000 ula mites [3].
species [1]. Arthropods are particularly adept at transmit-
Worldwide crop losses caused by insect pests ting viral disease (arborviruses); insects, mites,
have been estimated as wheat 5.0 %, rice 26.7 %, and ticks are vectors for at least 450 viruses.
maize 12.4 %, potatoes 6.5 %, sugar beet and The most feared are yellow fever, dengue,
cane 16.5 %, vegetables 8.7 %, fruit 5.8 %, oil and dengue hemorrhagic fevers (transmitted
crops 11.5 %, and fiber and rubber crops 14.2 % largely by the domestic mosquito Aedes ae-
[2]; see also → Crop Protection. Total produc- gypti), as well as the human encephalitides
tion loss is estimated at 13.8 % or $ 29.7×109 . such as St. Louis, western and eastern equine,
In the United States, total losses from insect at- and Venezuelan encephalitis for which the
tacks calculated in 1988 dollars are estimated at mosquitoes Culex tarsalis and C. pipiens are im-
$ 14×109 [1]. portant vectors [4].
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
1.1. Development of Insecticides [5–13]
cause human malaria; worldwide, they are re-
Humans have been striving to ameliorate the
sponsible for as many as 200 million clinical
enormous depredations caused by insects since
cases and a million deaths annually. At least
the dawn of history. Chinese writings of the
250 million people suffer from lymphatic filari-
Shang Dynasty (1523 – 1027 b.c.) describe orga-
asis caused by the nematode parasites Wuchere-
nized efforts to use fire to destroy plagues of the
ria bancrofti and Brugia malayi. This disease
migratory locust Locusta migratoria manilensis.
is transmitted largely by the common house
However, the causal agents of malaria, Plasmod-
mosquito Culex fatigans that breeds in pol-
ium spp., and the role of Anopheles mosquitoes
luted water. Trachoma, a viral disease spread
in their transmission have been recognized only
by the housefly Musca domestica is the most
since 1871.
common cause of human blindness, infecting
Plant-derived insecticides (e.g., rotenone, ve-
an estimated 40 million victims. American try-
ratridines, pyrethrins, and nicotine) have been
panosomiasis ( Chagas’ disease) is caused by
used for insect control since antiquity. Major
Trypanosoma cruzi and transmitted to as many
developments in the use of chemicals for plant
as 40 million humans by about 95 species of
protection began with Paris green or copper ace-
blood-sucking Triatominae bugs, with a variety
toarsenite in 1865, used as a stomach poison
of animal reservoirs. The dozen or so species
to control the Colorado potato beetle Leptino-
of tsetse flies (Glossina spp.) are vectors of
tarsa decemlineata. Lead arsenate, introduced in
trypanosomiasis in Central Africa. Blackflies
1892, and calcium arsenate, introduced in 1907,
of the Simulium damnosum complex are vec-
became the most important insecticides during
tors of the filarian Onchocerca volvulus that in-
the early twentieth century; production reached
fects about a million people in the Volta River
a maximum of 41.1×106 kg of lead arsenate in
Basin of Africa. The parasite may attack the
Insect Control 3
1944 and 38.1×106 kg of calcium arsenate in synthetic insecticides. Large-scale use of other
1942 [11]. Much of this was applied to cotton organochlorine compounds followed (e.g., lin-
to control the boll weevil Anthonomus grandis dane; toxaphene; and the cyclodienes chlordane,
and the pink bollworm Pectinophora gossyp- heptachlor, aldrin, dieldrin, and endrin).
iella. Arsenicals had the pronounced disadvan-
tages of high toxicity to humans and domes-
tic animals, considerable phytotoxicity, and ex- 1.2. Insect Resistance to Insecticides
treme environmental persistence [6], [11]. Cry- [16–18]
olite (sodium fluoroaluminate) was introduced
as a stomach poison insecticide in 1928 be- The ability of insects to rapidly develop resis-
cause of widespread concern about toxic ar- tance to the lethal action of insecticides has be-
senical residues on fruit and vegetables. These come a major factor in the control of insect pests.
stomach poison insecticides were effective only Insecticide resistance depends on random muta-
against chewing insects and had little contact ac- tion that establishes an R-allele (a gene coding
tion. for resistance) in the natural population of the
Although potassium dinitro-o-cresylate, the insect species. The mutant individual possesses
first synthetic organic insecticide, was intro- biochemical, physiological, or even behavioral
duced in Germany in 1892 as a dormant spray for characteristics that confer some degree of immu-
tree fruit insects, synthetic organic insecticides nity to the insecticide. Widespread application of
had little success until the patenting of 1,1,1- the insecticide propagates the R-allele through
trichloro-2,2-bis(4-chlorophenyl)ethane in 1939 preferential survival, and it is dispersed through-
[14]. This insecticide was enormously success- out the population. As the R-allele becomes suf-
ful in vector control of malaria, typhus, and ficiently common, insecticide effectiveness is re-
other human diseases during World War II, and duced. If the R-allele is partially dominant, com-
widespread publicity attending its civilian de- pletely resistant RR homozygotes are selected
but in 1945 led to the synthesis and evalua- rapidly. Thus insecticide resistance has been de-
tion of hundreds of thousands of synthetic or- scribed as accelerated microevolution. These R-
ganic insecticides. With its efficient contact ac- alleles are thought to be present in typical insect
tion, DDT became the mainstay of vector con- populations at frequencies of 10−4 – 10−2 , with
trol programs and rapidly replaced the arsenicals RR homozygotes present at 10−8 – 10−4 . With
in crop protection. It became the standard rem- the resistance of Anopheles gambiae mosquito to
edy for control of the codling moth Laspeyre- dieldrin used for malaria eradication, R-alleles
sia pomonella, which attacked deciduous fruit, were present in African populations at frequen-
and for the pink bollworm P. gossypiella, which cies of 0.04 – 6.0 %, and a single spray treatment
attacked cotton. For the first time, the use of produced a population of 86 % RS and RR geno-
DDT offered prospects for chemical control types, with virtual immunity [16].
of the two great defoliators of North Ameri- The first description of insecticide resistance
can forests, the gypsy moth Porthetria dispar in 1914 was that of the San Jose scale Quadras-
and the spruce budworm Choristoneura fumifer- pidiotus perniciosus to lime sulfur spray. In
ana. The name DDT became a household word 1916, resistance to hydrogen cyanide fumigant
for controlling insect pests in flower and veg- was reported in the California red scale Aoni-
etable gardens and for mothproofing woolens. diella aurantii and the black scale Saissetia
The success of DDT residual spray operations, oleae which attacked citrus. The number of sci-
which eliminated malaria in the Latina province entifically validated examples of insecticide re-
of Italy and on Sardinia within four years, in- sistance increased gradually, and by 1946 re-
augurated the malaria eradication campaign of sistance was known in eleven species including
the World Health Organization in 1955. The the codling moth Laspeyresia pomonella and the
plan was to eliminate malaria by spraying, at peach twig borer Anarsia lineatella to lead arse-
a level of 1 g/m2 , all human dwellings where nate, the citrus thrips Scirtothrips citri to potas-
malaria transmission occurred [15]. The enor- sium antimonyl tartrate (tartar emetic), the wal-
mous success of DDT in insect control stim- nut husk fly Rhagoletis completa to cryolite, and
ulated the research and development of other the cattle ticks Boophilus microplus and B. de-
4 Insect Control
Table 1. Rates of development of insect resistance to insecticides
Insecticide 1938 1948 1954 1970 1975 1980 1984 1989 Average
doubling time,
years
coloratus to sodium arsenite dip. However, little persistence of R genes in the species genome
scientific attention was given to insecticide resis- and of their interaction through a variety of re-
tance until the introduction of DDT in 1945, with sistance mechanisms that affect both the detoxi-
resistant strains of the housefly Musca domes- fication of and the target-site insensitivity to var-
tica appearing almost simultaneously in Swe- ious insecticides.
den and Denmark in 1946, of the mosquitoes
Culex tarsalis in Italy and Aedes sollicitans in
Florida in 1947, of the bedbug Cimex lectular- 1.2.1. Cross-Resistance and Multiple
ius in Hawaii in 1947, and of the human body Resistance
louse Pediculus humanus humanus in Korea and
Japan in 1951 [16]. The exponential growth in numbers of insect
Burgeoning use of insecticides after World species resistant to insecticides does not ade-
War II resulted in an exponential increase in quately describe the impact of resistance on in-
the number of scientifically documented exam- sect control problems. Two general types of re-
ples of resistance, which reached 224 species sistance are common: cross-resistance and mul-
in 1970, 364 in 1975, 447 in 1984, and 505 in tiple resistance.
1988 (Table 1). Most of the early examples of Cross-resistance enables resistant species to
resistance were found in insect vectors of hu- survive exposure to chemically related insecti-
man disease because of the widespread use of cides, such as DDT and methoxychlor, lindane
DDT, lindane, and dieldrin in vector control pro- and heptachlor, parathion and diazinon, carbaryl
grams. By 1970, resistance was documented in and carbofuran, or permethrin and fenvalerate.
119 pests of crops, forests, and stored products, Cross-resistance results from a common detoxi-
compared to 166 pests of human and animal fication mechanism or a change in susceptibility
health; in 1980 the respective figures were 283 to a common biochemical lesion.
agricultural pests compared to 198 pests of hu- Multiple resistance enables resistant species
mans and animals [18]. to survive exposure to several classes of in-
As each new class of insecticides was intro- secticide with differing detoxification systems
duced and used widely, the rate of development and biochemical modes of action. Multiple re-
of resistant species followed an almost iden- sistance results from interactions among sev-
tical pattern of exponential growth until high- eral types of evolutionarily developed survival
level resistance or legal restrictions slowed fur- mechanisms so that entire categories of insec-
ther usage (Table 1). Doubling time for the de- ticides become useless. Examples include the
velopment of resistant species has decreased altered acetylcholinesterase target site that pro-
steadily with the introduction of each new class duces multiple resistance to organophosphorus
of insecticide (i.e., DDT and methoxychlor, lin- and carbamate insecticides, which is found in
dane and cyclodienes, organophosphates, carba- the housefly, the green rice leafhopper Nephotet-
mates, and pyrethroids). This accelerated devel- tix cincticeps, the cattle tick Boophilus mi-
opment of multiple resistance is a result of the croplus, and the red spider mite Tetranychus ur-
ticae. Another type of multiple resistance is the
Insect Control 5
“knockdown” mechanism of nerve axon insen- portant Anopheles mosquito vectors of malaria
sitivity that promotes resistance to both DDT are resistant to one or more of the few residual
and the pyrethroids, and has been demonstrated insecticides that can be used safely in human
in the housefly; a variety of mosquitoes; bed- habitations. This resistance has so increased the
bugs of Cimex spp.; lepidopteran pests includ- cost and decreased the effectiveness of malaria
ing Spodoptera, Plutella, and Heliothis; and the control that the World Health Organization aban-
cattle tick B. microplus. Multiple resistance is doned its widely heralded program for global
now present in at least 44 families of insects in eradication of malaria in 1976 [20]. Important
10 orders [19]. An increasing number of ma- insect pests of man, such as the housefly and
jor insect pests are resistant to all the principal the German cockroach, can scarcely be con-
classes of insecticides (Table 1). These include trolled by insecticides. Practical chemical con-
the housefly; the malaria vectors Anopheles albi- trol of agricultural pests such as the Colorado
manus, A. sacharovi, and A. stephensi; the Ger- potato beetle, the diamondback, and the He-
man cockroach Blattella germanica; the Col- liothis cotton bollworms is in jeopardy. Resis-
orado potato beetle Leptinotarsa decemlineata; tance has severely affected the economy of in-
the green peach aphid Myzus persicae; the pear sect control in agriculture and public health by
psylla Psylla pyricola; the diamondback moth necessitating replacement of the relatively in-
Plutella xylostella; the cotton bollworms Helio- expensive organochlorine insecticides with the
this spp. and Spodoptera spp.; the red flour bee- increasingly expensive organophosphates, car-
tle Tribolium castaneum; and the granary weevil bamates, and pyrethroids. Adverse effects on
Sitophilus granarius. human health followed the withdrawal of rel-
In summary, cross-resistance limits the atively safe organochlorines, such as DDT and
choice of available insecticides, whereas mul- lindane, and their replacement by highly toxic
tiple resistance reflects the past history of insec- organophosphates, such as parathion and methyl
ticide use and precludes a return to those used parathion. Insecticide resistance is a major factor
previously. Both types of insecticide resistance in efforts to replace programs for the chemical
severely deplete insecticide resources. control of insects by more ecologically oriented
programs for insect pest management.
enormous outbreak of redspider mites (Tetrany- on the environment. This concern results from
chidae) in deciduous and citrus fruit orchards their intrinsic toxicity to animals and humans,
and on cotton and other row crops. These mites as well as their purposeful application to fields,
were rarely abundant enough before 1946 to re- orchards, forests, ponds and streams, and human
quire pesticide application. The sudden appear- habitations. Insecticide application is generally
ance in 1947 of the new words miticide and aca- inefficent: much less than 0.1 % of the amount
ricide for special pesticides to control these mite applied is required to kill pests; the remainder
pests provided dramatic proof of the impact of (> 99.9 %) is distributed throughout the ecosys-
broad-spectrum insecticides on the balance bet- tem.
ween phytophagous mites and their predators. With rapidly increasing use, environmen-
The imbalances created were all the more dra- tal contamination from insecticides has be-
matic because of the very short generation time come massive. In 1976, ca. 73.5×106 kg of ac-
of the mites and the rapidity with which they tive ingredients of insecticides was applied to
developed pesticide resistance. 30.335×106 ha of United States cropland; av-
Today, the use of broad-spectrum pesticides erage contamination was about 2.4 kg/ha [28].
has almost inevitably been followed by pest re- For cotton, the most heavily treated crop,
sistance, pest resurgence, and the outbreak of 29.5×106 kg of active ingredient was applied to
secondary pests. These occur most dramatically 2.835×106 ha, an average of 10.3 kg/ha. Early
on crops that are treated heavily with pesticides, insecticides such as the arsenicals and nicotine
such as cotton, rice, and tree fruit. Invariably, were highly toxic, and lead arsenate persisted
ecological imbalance has been related to the and accumulated in orchard soil to levels that
suppression of key natural enemies. Cotton cul- killed young trees.
ture worldwide has been almost destroyed by The organochlorine insecticides had the
the resurgence of Heliothis spp. and Spodoptera greatest impact on environmental quality be-
spp. bollworms that became devastating sec- cause of their relatively long persistence, their
ondary pests when their predators Chrysopa, very high lipid – water partition coefficients, and
Orius, Geocoris, and Nabis spp. were destroyed. their relative resistance to biological degrada-
Attempts to control cotton bollworms with more tion. Estimated average soil half-lifes in years
frequent applications of insecticides led to the are for DDT 2.8, dieldrin 2.5, endrin 2.2, lindane
destruction of populations of the Encarsia and 1.2, chlordane 1.0, and heptachlor 0.8; in some
Eretmocerus parasites of the whitefly Bemisia soil types, persistence is much longer. These in-
tabaci. This secondary pest, which transmits a secticides have very low water solubility (in the
variety of plant diseases, has become a critical ppb-ppm range) and have octanol – water parti-
factor in profitable cotton production in several tion values of 103 – 106 .
parts of the world. They accumulate in living organisms
The green rice leafhopper Nephotettix cincti- (bioconcentration), and trace levels in water or
ceps and related leafhoppers are vectors of rice food increase by a factor of 103 – 106 in fish or
diseases and have emerged as important sec- animals at the top level of the food chain. During
ondary pests of rice as a result of the very heavy their maximum use in 1950 – 1970, organochlo-
use of benzene hexachloride (BHC), parathion, rine insecticides became ubiquitous micropol-
and other broad-spectrum insecticides to con- lutants of the environment (see Chap. 17). Their
trol the rice stem borers Chilo suppressalis perturbing effects on wildlife and humans have
and Tryporyza incertulas. Insecticide applica- been major factors in their stringent regulation
tion destroyed predatory spiders that suppressed or prohibition in the United State, Canada, and
leafhopper populations. both Eastern and Western Europe: e.g., U.S.
government withdrawal of registration for DDT
(1973), aldrin and dieldrin (1974), heptachlor
1.4. Insecticides and Environmental and chlordane (1976), chlordecone (1980),
Quality [22–27] endrin (1981), mirex (1982), and toxaphene
(1983). The average levels of organochlorine
Insecticides have been highly controversial be- insecticide residues in water, soil, and human
cause of their widespread and persistent effects tissue have begun to decline, and threatened
Insect Control 7
populations of birds of prey have been restored. Pyrethroids have the advantage of extraor-
However, worldwide use of these compounds dinary effectiveness against insect pests, so
between 1946 and 1976 was so extensive (DDT that the required dosages are only about one-
> 2.3×106 t; toxaphene 4.5×105 t; cyclodienes tenth to one-twentieth those of organochlo-
including aldrin, dieldrin, heptachlor, chlordane, rines, organophosphates, and carbamates. How-
and endrin > 2.7×105 t) that traces will remain ever, the newer pyrethroids are more persis-
in the environment for many years. tent and lipophilic; they are extremely toxic to
Environmental pollution resulting from beneficial insects and fish. Furthermore, unlike
overuse of organochlorine insecticides focused organophosphates and carbamates, which have
attention on the importance of biodegradability. atropine as an antidote, no effective antidote ex-
Organophosphate insecticides react in the pres- ists for pyrethroid poisoning.
ence of light, air, and moisture, and are readily
degraded in living organisms by oxidation and
hydrolysis. Therefore, the use of these insecti- 1.5. Integrated Pest Management [30–36]
cides has not led to serious problems of bioac-
cumulation and food chain transfer. Their soil The discovery of DDT and subsequent de-
persistence is low, with half-lifes ranging from velopment of hundreds of organochlorine,
1 – 2 weeks for malathion, to 3 – 6 months for organophosphate, carbamate, and pyrethroid in-
parathion, diazinon, azinphosmethyl, and chlor- secticides produced a euphoria about chemical
pyrifos. pest control that led to the pest eradication phi-
However, organophosphates are highly toxic losophy prevalent during the 1950s and 1960s.
to almost all animals including humans because Widespread application of cheap, effective in-
they irreversibly inhibit the enzyme acetyl- secticides was supposed to provide a rapid solu-
cholinesterase. Acetylcholinesterase is required tion to every insect pest control problem. The
to break down the neurotransmitter acetyl- World Health Organization began the global
choline after it has been released from the eradication of malaria through annual resid-
nerve ending. Unlike most organochlorine insec- ual (long-lasting) house spraying directed at
ticides, many organophosphate insecticides are Anopheles mosquito vectors. The United States
almost as toxic upon dermal exposure as upon in- undertook the insecticide eradication of pests
gestion. Furthermore, chronic exposure can pro- such as the gypsy moth Porthetria dispar, the
gressively deplete vital acetylcholinesterase lev- Japanese beetle Popillia japonica, the yellow
els to a point where slight additional exposure fever mosquito Aedes aegypti, and the red im-
can result in severe illness or death from cholin- ported fire ant Solenopsis invicta. As each erad-
ergic crises. Human intoxication resulting from ication program faltered because of insecticide
the use of the organophosphates is intensified resistance, destruction of wildlife, and wide-
by insecticide resistance, which necessitates in- spread environmental pollution, the initial reac-
creased dosage and more frequent application. tion was to develop another insecticide of differ-
The incidence of human intoxication after use ent chemistry.
of organophosphorus insecticides has become The publication of Rachel Carson’s Silent
acute; the World Health Organization has esti- Spring in 1962 [37] inaugurated public chal-
mated that 500 000 hospitalizations and as many lenge to the notion that all insect control prob-
as 20 000 deaths occur annually worldwide [29]. lems could be solved solely with insecticides.
Introduction of carbamate insecticides pro- Surveys demonstrated the continuing, unabated
vided some amelioration of both environmental damage resulting from insect pest attack despite
persistence and acute toxic hazards. Carbamates a greater than 10-fold increase in insecticide use
are generally biodegradable and do not bioaccu- and a 20-fold proliferation in chemicals avail-
mulate. Soil persistence is low, with half-lifes able for insect control (Table 2). Unbridled use
of 1 – 2 weeks for carbaryl and 1 – 4 months for of broad-spectrum insecticides over the past 40
carbofuran and aldicarb. Many carbamates are years had obviously not solved the manifold
highly toxic to humans and animals, but acetyl- problems of insect pest control. The benefits to
cholinesterase inhibition is readily reversible so cost ratio of pest control activities to the farmer
that severe intoxication is unlikely. 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 ∗
Alfalfa 10 5 9 15
Apple 20 10 14 13
Cabbage 10 20 8 17
Corn 8 12 4 12
Cotton 10 15 15 19
Potato 10 22 16 4
Calcium arsenate [7778-44-1], Ca3 (AsO4 )2 , Mode of Action. Inorganic fluorides com-
was introduced in 1920 to control the plex magnesium ions as magnesium fluorophos-
boll weevil on cotton. It contains 37.6 % phate. They inhibit enzymes such as enolase
arsenic and has an LD50 (rat, oral) of (E.C. 4.2.1.1) that require Mg2+ as a prosthetic
298 mg/kg. Basic calcium arsenate [74966-85- group. Thus, fluoride poisoning prevents phos-
1], [Ca3 (AsO4 )2 ] 3 · Ca(OH)2 , contains 25 % phate transfer in oxidative metabolism.
arsenic with ca. 0.4 – 0.5 % of it water-soluble.
It is a safer form that does not hydrolyze readily.
3. Natural Product Insecticides
Arsenic trioxide [1327-53-3], As2 O3 , with
75.7 % arsenic has an LD50 (rat, oral) In this and the following chapters two values are
of 14.1 mg/kg. Sodium arsenite [7784-46-5], often cited for the LD50 . The first value refers to
NaAsO2 , with 55.7 % arsenic has an LD50 (rat, male animals and the second to female animals.
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 3.1. Nicotinoids [8], [9]
cattle ticks.
Nicotine [54-11-5], l-1-methyl-2-(3 -
Mode of Action. The As3+ ion of arsenicals pyridyl)-pyrrolidine, bp 247 C, d 20
◦
4 1.009, one
specifically inhibits the sulfhydryl-containing of the earliest insecticides, was recommended in
enzymes involved in cell respiration, e.g., pyru- 1793 as a tea for the destruction of aphids. Nico-
vate oxidase (E.C. 1.2.3.3.). 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.
2.2. Fluorides [6] It occurs as the principal alkaloid along with
small amounts of 12 other alkaloids, includ-
Cryolite [15096-52-3], sodium fluoroalumi- ing the water-miscible insecticidal compounds
nate, Na3 AlF6 , 54.2 % fluorine, water-soluble nornicotine and anabasine. Nicotine is readily
fluorine 0.035 – 0.061 %, was developed as a volatile (vapor pressure 5.5 Pa at 25 ◦ C) and
safer stomach poison insecticide than the arseni- dibasic (K b1 = 1×10−6 , K b2 = 1×10−11 ), easily
cals, especially because of concern about per- forming salts with acids and many metals. Its
sistent lead arsenate residues on edible produce LD50 (rat, oral) is 30 – 55 mg/kg.
such as apples. The LD50 (rat, oral) of cryolite
has been given as 200 – 13 500 mg/kg. Nicotine sulfate [65-30-5], (C10 H14 N2 )2 ·
H2 SO4 , has been used as an aphicide on fruit,
Sodium fluorosilicate [16893-85-9], vegetables, and ornamentals because it is less
Na2 SiF6 , 60.6 % fluorine, water-soluble fluo- volatile than nicotine and safer to handle.
rine 0.65 %, LD50 (rat, oral) 125 mg/kg, is used Nornicotine [494-97-3], 2-(3 -pyridyl)pyrro-
for mothproofing. lidine, bp 270 ◦ C, d 20
4 1.07, occurs naturally as
both the d and the l isomer, the former in D.
Barium fluorosilicate [17125-80-3], hopwoodii and the latter in Nicotiana.
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.
10 Insect Control
0.04 – 0.25 % active ingredient, together with for effective synthetic derivatives of these costly
five to ten times this amount of piperonyl butox- natural products. Since the early 1940s, struc-
ide or other synergists to retard detoxification tural optimization of pyrethroid molecules has
in insects. Deodorized oil or water-based sprays led to a series of broad-spectrum insecticides
are also used. The principal areas of use for with greatly enhanced activity and, in many
pyrethrum products are cases, extended residual action. Pyrethroids
have become one of the most important classes
1) household insecticides for the control of of insecticides.
mosquitoes, flies, cockroaches, bedbugs,
ants, silverfish, and fabric pests;
2) livestock or cattle sprays to protect animals 3.5.1.1. Chrysanthemate Esters
from annoyance by biting and nonbiting flies;
3) protective sprays for grain in storage; and Allethrin [584-79-2], 2-methyl-4-oxo-3-
4) limited use on fruit and vegetables to con- (2-propenyl)-2-cyclopenten-1-yl 2,2-dimethyl-
trol insect pests prior to harvest when toxic 3-(2-methyl-1-propenyl)cyclopropanecarbox-
residues are very objectionable. ylate, the first successful synthetic pyrethroid,
was developed in 1949 and is the allyl homo-
Toxic Constituents. Complete identifica- logue of cinerin I. The product is a brownish liq-
tion of the active principles of pyrethrum re- uid, d 20
4 1.005 – 1.015, containing 75 – 95 % of a
quired approximately 50 years, dating from the mixture of eight optical and geometric isomers
work of Staudinger and Ruszicka in 1924. [70 % (±)-trans and 30 % (±)-cis acids] esteri-
There are six esters of the two cyclopropanecar- fied with (±)-methylcyclopentenolone alcohol.
boxylic acids, chrysanthemic acid and pyrethric Allethrin has an LD50 (rat, oral) of 920 mg/kg;
acid (Fig. 1), and three cyclic keto alcohols, it is as effective as natural pyrethrum against
pyrethrolone, cinerolone, and jasmolone. The flies and mosquitoes but has a narrow spec-
esters are optically active with absolute config- trum of activity. Bioallethrin is the (+)-trans-
urations of 1R, 3R, and 4S; the double bond in chrysanthemate ester of the same alcohol with
the acid side chain is cis and in the alcohols, enhanced “ knockdown ” activity.
trans.
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.
sential. The C-3 substituent in cyclopropane C=C double bonds of pyrethrins, the cinerin side
can be isobutenyl, 2,2-dihalovinyl, or tri- chains, and methyl groups of the isobutenyl moi-
fluoromethylchlorovinyl. The configuration ety. Synergists of methylenedioxyphenyl(1,3-
at C-1 and C-3 of the cyclopropane must be benzodioxole) react with the iron atom of cy-
either 1R, 3S or 1R, 3R. tochrome P450 and prevent detoxification. Such
2) Alcohol Moiety. The alcohol moiety should synergists are usually included with natural
contain a cyclopentenolone ring or stereo- pyrethrins at ratios of (10 – 20) : 1 in fly sprays
chemically similar furan or benzyl groups and aerosols. They may increase the activity of
and an unsaturated side chain or aromatic pyrethrins as much as 30-fold. Extensive struc-
group. The α-cyano group apparently acts as tural optimization of synthetic pyrethroids has
a dimensional spacer to promote maximum eliminated most of the detoxification sites. Syn-
complementarity with the axonal receptor. ergists are generally ineffective with the persis-
tent pyrethroids.
in the United States. The compound has an ap- DDT in alkaline solution or in biological sys-
propriate combination of properties: (1) broad tems. However, the methoxy groups are readily
spectrum of insecticidal activity, (2) long resid- dealkylated in vivo by microsomal oxidases to
ual persistence, (3) low acute toxicity to mam- produce phenols that are eliminated easily. Thus,
mals, and (4) low cost. The enormous publicity methoxychlor does not bioaccumulate like DDT
it received after World War II heightened its use and is favored for general environmental use.
as a panacea for all insect pest control problems, Methoxychlor is one of the safest of all insecti-
and DDT quickly became the standard remedy cides with an LD50 (rat, oral) of > 6 000 mg/kg.
for control of the codling moth on deciduous Unfortunately, insects resistant to DDT are gen-
fruit, the pink bollworm on cotton, the Colorado erally cross-resistant to methoxychlor.
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 The compound known as DDD [72-54-
WHO program for global eradication of malaria, 8] (TDE), 1,1 -(2,2-dichloroethylidene)bis(4-
and it still has major place in malaria control. chlorobenzene), occurs as an impurity in tech-
The compound known as DDT [50-29- nical DDT and has been marketed as an in-
3], 1,1 -(2,2,2-trichloroethylidene)bis(4-chloro- secticide. Technical DDD contains mainly the
benzene), is a white crystalline solid, mp 109 ◦ C, 4,4 -isomer, mp 110 ◦ C, and 7 – 8 % of the 2,4 -
vapor pressure at 20 ◦ C 0.025 mPa. Technical isomer, mp 76 ◦ C. The properties of DDD are
DDT is a white amorphous powder containing almost indistinguishable from DDT and the
65 – 80 % of the active 4,4 -DDT; 14 – 21 % of product is even safer with LD50 values (rat)
the nearly inactive 2,4 -DDT [789-02-6], mp of 3 400 mg/kg (oral) and > 10 000 mg/kg (der-
73 – 74 ◦ C; up to 4 % of the insecticidal DDD mal). Like DDT, DDD is also bioaccumulated
[72-54-8], 1,1 -(2,2-dichloroethylidene)bis(4- and its registration was also canceled in the
chlorobenzene), mp 110 ◦ C; and traces of 2,2 - United States in 1974.
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. Perthane [72-56-0], 1,1 -(2,2-dichloroeth-
This reaction also occurs in biological systems ylidene)bis(4-ethylbenzene), mp 60 – 61 ◦ C, is
and accounts for the storage and accumulation the 4,4 -diethyl analogue of DDT. It is very
of DDE in lipid tissue of humans and higher biodegradable and has low mammalian toxicity,
animals. The rat LD50 values of DDT are 113, LD50 (rat, oral) 8 170 mg/kg. Perthane has been
118 mg/kg (oral) and 2510 mg/kg (dermal). used indoors, where very low acute and chronic
toxicities are desirable.
(4-fluorobenzene), mp 44 ◦ C, has chemical and 1974. The DDT residues in humans declined
insecticidal properties similar to DDT. It was slowly to about one-half the above values by
used as a sanitary insecticide by the German 1978. In vivo DDT is slowly dechlorinated re-
army in World War II but is too phytotoxic ductively to DDD and dechlorinated further to
for agricultural use. The LD50 (rat, oral) is DDA or 4,4 -dichlorodiphenylacetic acid, the
900 mg/kg. 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.
5.1.1. Mode of Action Whereas DDT is highly toxic to fish (LC50 for
trout and bluegill 0.002 – 0.008 mg/kg), it is only
Both DDT and its analogues specifically af- moderately toxic to birds (oral LD50 for mallard
fect the peripheral sensory organs of insects 1 300 mg/kg and for pheasant > 2 240 mg/kg).
to produce violent trains of afferent impulses However, because of its high lipid – water par-
that cause hyperactivity and convulsions. Sub- tition value, DDT bioaccumulates as much as
sequent paralysis and death are the result of 3×106 -fold from the waters of Lake Michigan
metabolic exhaustion or the production of an en- 6 ng/L in the bodies of lake trout Salvelinus na-
dogenous neurotoxin. The insecticidal action of maycush (> 20 mg/kg). In raptorial birds such as
DDT is facilitated by its highly lipophilic char- eagle, osprey, and brown pelican, bioaccumula-
acter, which promotes absorption through the tion of DDE from fish has decreased eggshell
lipoprotein matrix of the chitinous insect cuticle. thickness by 10 – 15 %, resulting in widespread
The DDT molecule is absorbed into the sodium reproductive failure. Because of the many envi-
gate mechanism of the insect nerve axon, pro- ronmental problems resulting from massive use
longing the negative afterpotential and promot- of DDT, its registration for agricultural use was
ing multiple, repetitive nerve impulse firing. The canceled in the United States and Canada; Japan
critical site of action may be the inhibition of and many European countries have enacted sim-
Ca2+ ATPase. ilar prohibitions.
are 100, 162 mg/kg (oral) and 195, 250 mg/kg of hexachlorocyclopentadiene and 2,5-di-
(dermal). Heptachlor is oxidized readily to hep- hydrofuran. It has a mp of 122 – 123 ◦ C and a
tachlor epoxide [1024-57-3], mp 159 – 160 ◦ C, vapor pressure of 0.39 mPa at 20 ◦ C; it is soluble
LD50 (rat, oral) 47 mg/kg, which is an important in water to about 0.1 mg/L at 20 ◦ C. Isobenzan
and highly persistent environmental pollutant. 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.
Bromodan [1715-40-8], 5-bromomethyl- present in human milk and fatty tissue. Dieldrin
1,2,3,4,7,7-hexachlorobicyclo[2.2.1]hept-2- residues in the body fat of U.S. inhabitants av-
ene, mp 75 – 79 ◦ C, is the Diels – Alder adduct eraged 0.22 – 0.35 ppm in 1970 and declined to
of hexachlorocyclopentadiene and allyl bro- 0.12 – 0.14 ppm by 1978, after cancellation of
mide. It has very low mammalian toxicity, LD50 the use of aldrin and dieldrin in 1974. Oxychlor-
(rat, oral) 1.2×104 mg/kg, and has been used dane [26880-48-8], mirex, and chlordecone are
in Europe to control ectoparasites of domestic also highly persistent and bioaccumulative.
animals. 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
Mode of Action [52]. The cyclodienes ap- 381, and endrin 1.6 and 5.6 mg/kg. Lifetime ani-
pear to act by a common mechanism affect- mal feeding studies have shown that heptachlor,
ing the nerve axon to produce hypersensitivity, aldrin, and dieldrin are carcinogens. The data
convulsions, prostration, and death. Spray crew- with endrin are equivocal.
men with lengthy exposure to dieldrin have suf-
fered from prolonged and repeated central ner-
vous system epileptiform convulsions; similar 5.4. Chlorinated Camphenes
nervous disturbances have occurred in workers
Toxaphene, C10 H10 Cl8 , is a mixture of more
exposed to chlordecone. The lesion appears to
than 180 chlorinated cyclic compounds pro-
be blockage of GABA synaptic transmission by
duced by photochlorination of the bicyclic ter-
competitive inhibition of the neurotransmitter
pene camphene. The product is a yellowish,
binding site. This mode of action is common
semicrystalline gum, mp 65 – 90 ◦ C, d 254 1.64,
to the cyclodienes, toxaphene, and lindane. The
vapor pressure 0.13 mPa at 25 ◦ C. Toxaphene
highest toxic action is shown by compounds con-
is soluble in water to about 0.4 mg/L and very
taining a polychlor center and an adjacent, prop-
soluble in petroleum solvents. It is unstable in
erly oriented electronegative site (e.g., a double
alkali, upon prolonged exposure to sunlight, and
bond adjacent to a Cl, O, S, or N atom).
above 155 ◦ C, liberating hydrogen chloride and
Environmental Toxicology. Cyclodienes losing some of its insecticidal potency. The rat
add oxygen readily; e.g., aldrin and heptachlor LD50 values are 80, 90 mg/kg (oral) and 1 075,
form the more stable epoxides dieldrin and hep- 780 mg/kg (dermal). The most active ingredi-
tachlor epoxide. The epoxides are formed in the ents identified from technical toxaphene are
tissues of animals that ingest or absorb the com- 2,2,5- endo-6-exo-8,9,10-heptachlorobornane
pounds, and they are concentrated and stored [51775-36-1], with a mouse intraperitoneal
preferentially in fatty tissue. Epoxidation also LD50 of 6.6 mg/kg, and 2,2,5-endo-6-exo-
occurs in plants and soil. Because dieldrin and 8,9,9,10-octachlorobornane [58002-18-9], with
heptachlor epoxide are more stable (e.g., the a mouse intraperitoneal LD50 of 3.1 mg/kg.
half-life on alfalfa leaves is 7 – 8 d, compared Each constitutes 2 – 6 % of the technical mix-
to < 1 d (for aldrin and heptachlor), in situ ox- ture. Toxaphene is a broad-spectrum persistent
idation markedly affects the residual behavior insecticide with major use on cotton and for the
of these cyclodienes. Average soil half-lifes for control of animal ectoparasites.
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
24 Insect Control
> 2 000 mg/kg (dermal). Trichlorfon is a stom- 42 mg/kg (dermal). Dicrotophos is a contact and
ach poison insecticide used on foliage for chew- systemic insecticide which has been injected
ing insects and in dry sugar bait for fly control. into trees to control bark and foliage pests.
Above pH 6, trichlorfon is converted rapidly to
dichlorvos, which accounts for its toxicity.
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.
consequently accumulates and disrupts the nor- phosphorus per mole of enzyme. An equimolar
mal functioning of the nervous system, pro- quantity of alcoholic, phenolic, or acidic product
ducing typical cholinergic symptoms (i.e., hy- HX is liberated as the leaving group.
peractivity, tremors, convulsions, paralysis, and
death). In higher animals and in humans, these Delay Factor. The reactivity of organophos-
cholinergic effects are translated into muscarinic phorus insecticides is determined by the elec-
effects (e.g., nausea, salivation, lachrymation, trophilicity of the phosphorus atom, the strength
and myosis of the pupil of the eye); nicotinic ef- of the “ high-energy ” P−X bond, and steric ef-
fects (e.g., muscular fasiculations); and central fects of the substituents. The electrophilicity of
effects (e.g., giddiness, tremulousness, convul- the central phosphorus is determined by the rel-
sions, and coma). ative electronegativities of the atoms bonded to
The reaction between the esterase and the it (e.g., P 2.1, O 3.5, S 2.5, N 3.0, and C 2.5).
phosphorus inhibitor involves attack on the elec- Therefore, the phosphorus in a phosphate es-
trophilic phosphorus atom by a nucleophilic ser- ter (P=O) is much more electrophilic than the
ine hydroxyl group with the neighboring imi- oxygen, and these compounds are much more
dazole ring of a histidine residue as the active reactive than the phosphorothioate esters (P=S).
site of acetylcholinesterase. The normal three- The latter are generally so stable as to be unre-
step reaction (A) that occurs at the synapse bet- active with acetylcholinesterase. Their insecti-
ween acetylcholinesterase (EH) and the sub- cide activity and general toxicity result from in
strate acetylcholine is 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]:
high degree of correlation exists between the 6.9.2. Delayed Neuropathy [57], [65], [66]
quantitative measure of electron-withdrawing
power of the substituent (Hammett’s σ) and Certain organophosphorus insecticides [espe-
anticholinesterase activity, and toxicity [58], cially phosphonates and phosphonothionates
[59]. This explains the high insecticidal activ- with P(O)C bonds] phosphorylate an enzyme
ity of molecules with 4-NO2 (methyl parathion), neurotoxic esterase to produce a delayed neu-
4-CN, 4-CH3 SO (fensulfothion), and 2,4,5-Cl ropathy that is essentially irreversible and may
(ronnel) phenyl leaving groups. occur from one to four weeks after acute or
chronic exposure. Neurotoxic esterase is evi-
Activation. In vivo, microsomal oxidation dently required for growth and maintenance of
is also important for activation or lethal syn- the long axons of the central nervous system. Af-
thesis in which electron-donating or inert ter its inhibition, degenerative changes develop
groups (e.g., CH3 S) are converted to strongly in the long axons and are manifest by tingling
electron-withdrawing groups (e.g., CH3 SO and and burning sensations in the limb extremities
CH3 SO2 ). Thus, insecticides such as fenthion that progress to weakness in the lower limbs and
are activated in vivo by microsomal oxidation to ataxia. The final stage is irreversible paraly-
to produce strongly electrophilic phosphorus sis which, in severe poisoning, may affect upper
atoms. This type of reaction involves a built-in limbs as well. Recovery is very slow and seldom
delay that is particularly effective in certain sys- complete in adults, with flaccid paralysis chang-
temic insecticides, such as phorate and disulfo- ing to spasticity. The phosphorylated esterase is
ton, where both P=S and C2 H5 S oxidation occur modified by “ aging ” so that the remaining es-
in plant tissues: 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.
Acetylcholinesterase inhibition in the blood and public health pest control. Safety is also en-
and central nervous system of humans is cumu- hanced by differences in the rates of microsomal
lative and can increase to fatal levels upon re- oxidation of P=S to P=O (see Section 6.9.1) and
peated chronic exposure of agricultural work- in hydrolytic detoxification between insects and
ers. This is due to the irreversible inhibition mammals.
of acetylcholinesterase with O,O-diethyl phos- The user of OP insecticides can enhance
phates. Therefore this vital enzyme may have selectivity and safety by choosing only com-
to be resynthesized. Hazards resulting from use pounds with LD50 values > 100 mg/kg, by us-
of the more toxic organophosphorus insecticides ing O,O-dimethyl instead of O,O-diethyl insec-
(Class I poisons with LD50 < 50 mg/kg) can be ticides, and by using P=S in preference to P=O
reduced by wearing protective clothing includ- insecticides. Great care should be exercised in
ing mask, goggles, and rubber gloves. Work- handling, applying, and storing all organophos-
ers with frequent exposure should have regu- phorus insecticides.
lar determinations of blood acetylcholinesterase
levels and should be removed from exposure,
when these blood acetylcholinesterase levels 6.10. Environmental Toxicology
show progressive decrease. In cases of severe
exposure and incipient cholinergic symptoms, Organophosphorus insecticides are reactive and
atropine should be administered promptly and are degraded readily by oxidation and hydro-
the patient should be hospitalized. lysis in vivo. Therefore, the use of organophos-
Selectivity is implicit in the definition of an phorus esters does not result in serious problems
insecticide as “ an agent for destroying insects ”. of bioaccumulation and food chain transfer. Soil
Nevertheless, many highly toxic organophos- persistence is low, with typical half-life values
phorus compounds are still used in massive ranging from one or two weeks for malathion to
quantities throughout the world. Their selectiv- three to six months for parathion, diazinon, and
ity is apparent only through careful application azinphos-methyl. However, organophosphorus
and confinement to the area of treatment. Selec- insecticides are general biocides and are toxic
tivity is important in providing safety not only to nearly all animals. Oral LD50 values, in mil-
for the user of pesticides and domestic animals, ligrams per kilogram, are phorate, mallard 0.62,
but also for wildlife and beneficial insects: honey pheasant 7.1; parathion, mallard 2.0, pheasant
bees and other pollinators, parasites, and preda- 12; monocrotophos, mallard 4.8, pheasant 2.8;
tors. diazinon, mallard 3.5, pheasant 4.3; fenthion,
The development of malathion in 1950 was mallard 5.9, pheasant 18; chlorpyrifos, mallard
an important milestone in the use of selective in- 75, pheasant 12; and azinphos-methyl, mallard
secticides. Malathion is about one-half to one- 136, pheasant 75. The organophosphorus com-
twentieth as toxic as parathion to insects, but pounds are generally toxic to fish, and LD50 val-
it is only about one one-hundredth as toxic to ues, in parts per million, are phorate, bluegill
humans and higher animals. The use of thou- 0.0055; parathion, bluegill 0.10, trout 0.047; di-
sands of metric tons of malathion over several azinon, bluegill 0.052, trout 0.38; malathion,
decades has demonstrated substantial safety to bluegill 0.11, trout 0.13; fenthion, bluegill 1.4,
users, coupled with effectiveness in insect pest trout 0.93; and azinphos-methyl, bluegill 0.005,
control. trout 0.004. Organophosphorus insecticides are
A high degree of selectivity has been demon- highly toxic to bees, beneficial parasites, and
strated by incorporating chlorine or methyl predators.
groups in the 3-position of the aryl ring of
parathion. Such groups interact stereochemi-
cally with acetylcholinesterase, increasing the 7. Carbamate Insecticides [67–71]
affinity for the insect enzyme and decreas-
ing it for the mammalian enzyme. Such com- The carbamates are synthetic analogues of the
pounds, including fenitrothion, fenthion, ronnel, toxic alkaloid physostigmine [57-47-6] isolated
bromophos, and iodofenphos, have been used from the calabar bean Physostigma venenosum
widely and safely in household, animal health,
38 Insect Control
Oxamyl [23135-22-0], S-methyl 1-(dimeth- chain oxidation, and ring hydroxylation. Meth-
ylcarbamoyl)-N-[(methylcarbamoyl)oxy]thio- ylenedioxyphenyl synergists (Section 3.5.4)
formimidate, mp 108 – 110 ◦ C, vapor pressure prevent this largely by inhibiting the microso-
20 mPa at 25 ◦ C, rat LD50 5.4 mg/kg (oral) and mal enzymes.
2 960 mg/kg (dermal), is a systemic insecticide
and nematocide.
7.5. Environmental Toxicology
three-step hydrolysis of acetylcholine (cf. Sec- LD50 (rat, oral) 238 mg/kg, is used primarily
tion 6.9.1). However, the third reaction step is against eggs and early growth stages of cotton
much slower for the carbamylated enzyme than bollworm and rice stem borer larvae (instars). It
for the acetylated one. The importance of struc- has both vapor and systemic action.
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 Amitraz [33089-61-1], N -(2,4-dimethyl-
are virtually inactive [68], [69]. phenyl)-N-{[(2,4-dimethylphenyl)imino]meth-
Detoxification of carbamate insecticides oc- yl}-N-methylmethanimidamide, mp 86 – 87 ◦ C,
curs in vivo through microsomal hydroxylation, water solubility 1 mg/L, LD50 (rat, oral)
N-demethylation of carbamyl nitrogen, side
Insect Control 43
400 mg/kg, is used principally to control red dodecadienoate, bp 100 ◦ C (6.7 Pa), vapor pres-
spider mites and pear psylla. 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.
9.1. Juvenoids
over 18 months of ca. 2×109 male flies, ster- 13.2. Chemosterilants [99–101]
ilized with γ-radiation from cobalt-60, over an
area of 181 000 km2 . A similar program along The availability of a variety of radiomimetic
the Texas – Mexico border has substantially con- (i.e., mimic the effect of γ-radiation) and an-
trolled the screwworm over the past 25 years, timetabolite cancer suppressive drugs that can
through the annual release of as many as 10×109 sterilize large segments of wild insect popula-
sterile males. Dramatic population reduction of tions has led to many suggestions for their use
the Mediterranean fruit fly Ceratitis capitata in genetic control programs.
was also achieved in California in 1981 when Aziridines have a highly reactive ethylen-
ca. 40×106 flies were released weekly. Ster- imine group. They alkylate deoxyribonucleic
ile male release has been used with limited acid and cause sterility in both sexes of
success to suppress isolated populations of the house flies, mosquitoes, fruit flies, and
codling moth Laspeyresia pomonella in the Pa- other insects. Compounds used experi-
cific Northwest and is proposed as an integral mentally include TEPA [57-39-6], trieth-
part of a cotton boll weevil eradication scheme ylenephosphoramide, mp 41 ◦ C, LD50 (rat,
in the southeastern United States. oral) 37 mg/kg, and its analogue thio-
Success in sterile male release programs de- TEPA [52-24-4] (→ Cancer Chemotherapy,
pends on several factors: Chap. 3.5.), triethylenethiophosphoramide,
mp 51.5 ◦ C. Other radiomimetic compounds
1) method for rearing large numbers of insects, that have been investigated as chemosteri-
2) adequate dispersion of released sterile males, lants include apholate [52-46-0], 2,2,4,4,6,6-
3) a sterilization procedure that does not de- hexakis(1-aziridinyl)-2,2,4,4,6,6-hexahydro-
crease male vigor and competitiveness in 1,3,5,2,4,6-triazaphosphorine, mp 147.5 ◦ C,
mating, LD50 (rat, oral) 98, 113 mg/kg; and HEMPA
4) a pest species that mates only once per life- [680-31-9], hexamethylphosphoric triamide,
time or in which the sperm from sterile males bp 105 – 107 ◦ C (1.5 kPa), d 20 4 1.03, LD50
compete with those of fertile males, and (rat, oral) 2650, 3360 mg/kg. Busulfan
5) a low pest population density or a means of [55-98-1], 1,4-butanediol dimethanesulfonate,
reducing the population to a low level (e.g., CH3 SO2 O(CH2 )4 OSO2 CH3 , LD50 (rat, oral)
by insecticide application). 1.8 mg/kg, has been used in the mass produc-
Many species of insects do not conform to tion of sterile insects. Unfortunately, all of these
these criteria, and the rearing and release of ster- radiomimetic compounds are mutagens and car-
ile males can be a costly procedure. 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
repellent and toxic action. Ideally, such moth- 15. Attractants [104–117]
proofing materials should
With a growing need for more specific, less en-
1) be fixed into the cloth during dyeing by vironmentally polluting pest control, the use of
chemical reaction with the wool protein, specific insect attractants has become an impor-
2) be completely effective against fabric pests, tant part of control technology. Chemical at-
3) give protection for up to 15 years after re- tractants can be used in conjunction with sim-
peated washing and dry cleaning, and ple, inexpensive sticky traps for monitoring in-
4) be nonvolatile, stable to light, odorless, col- sect pest populations in relation to the economic
orless, and nontoxic to humans. threshold, for removal trapping of pests, for lur-
Sodium fluorosilicate [16893-85-9], alone or ing them to toxic bait, or for mating confusion.
with aluminum, magnesium, or ammonium Two general types of attractants have been stud-
salts, is an effective mothproofing agent when ied widely: (1) the sex pheromones used in in-
employed in 0.5 – 0.7 % aqueous solution with traspecies chemical communication leading to
0.3 % potassium alum and 0.03 % oxalic acid. mating, and (2) plant kairomones used in inter-
Such formulations may be applied by spraying species chemical communication for host plant
or dipping and cannot be removed by dry clean- selection.
ing.
Formerly, DDT at 0.25 – 0.75 % and dield-
rin at 0.05 % of the cloth weight were applied 15.1. Sex Pheromone Attractants
by spraying in a volatile petroleum solvent or
by dipping the cloth in an aqueous emulsion. Sex pheromone attractants have now been iden-
This protected the cloth against fabric pests tified in several hundred species of insect
for months to years. However, because of en- pests[113]. For the most part, they are volatile es-
vironmental pollution problems with these per- ters, alcohols, and aldehydes that facilitate sex-
sistent bioaccummulative organochlorines, they ual communication and mating over distances up
are no longer used. Several persistent synthetic to several hundred meters and can be perceived
pyrethroids including permethrin, fenvalerate, in nanogram to microgram quantities. In Lepi-
and fenpropanate are effective in situ mothproof- doptera, sex pheromones are typically produced
ing agents. by the female in eversible glands at the tip of
The most widely used commercial moth- the abdomen. A number of pheromones used in
proofing agents are the (polychloro-2-chlorome- commercial insect control are identified in Ta-
thylsulfamido) diphenyl ethers ( Eulan type) and ble 4. The naturally occurring sex pheromones
the chlorinated diphenyl ether urea sulfonic acid are often complex mixtures, and maximum re-
(Mitin type). They are applied to woolens during sponse to synthetic mixtures depends on specific
fabrication at about 1 – 3 % of the weight of the blends. For example, the male European corn
cloth, and fixed by chemical reaction with the borer Ostrinia nubilalis responds to a mixture of
wool protein. 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
Insect Control 53
(e.g., cotton boll weevil Anthonomus grandis), 15.2. Aggregation Pheromones [108],
a four-component sex pheromone is produced [114]
by the male and liberated in the feces (frass).
Each component is essential for effective aggre- Bark beetles (Coleoptera, Scolytidae) attack dy-
gation, and the mixture is used commercially as ing trees and bring about aggregation and mass
grandlure [11104-05-5]. 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].
These beetle pheromones are used in trapping Scolytus multistriatus, the lesser European
to monitor pest populations. 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].
Similarly, the Mediterranean fruit fly Cerati- the active ingredients, although ground botan-
tis capitata is attracted strongly to the kairomone ical preparations such as pyrethrum, derris, or
trimedlure [12002-53-8], tert-butyl 2-methyl- ryania may be used undiluted. Dust carriers in-
4-chlorocyclohexanoate. Mixtures of apple clude organic flours, sulfur, silicon oxides, lime,
volatiles including butyl 2-methylbutanoate, gypsum, talc, pyrophyllite, bentonites, attapul-
propyl hexanoate, butyl hexanoate, hexyl gite, and volcanic ash. They are selected on the
propanoate, and benzyl butanoate are used to basis of compatibility with the active ingredi-
trap the apple maggot Rhagoletis pomonella; ent (including pH, moisture content, and sta-
dipropyl disulfide [629-19-6], C3 H7 S−SC3 H7 , bility), particle size, abrasiveness, absorbability,
is an attractant for the onion maggot fly Hylemya density, wettability, and cost. The mixture of ac-
antiqua; and allyl isothiocyanate [57-06-7], 3- tive and inactive ingredients is produced by op-
isothiocyanato-1-propene, for the cabbage mag- erations including solvent impregnation, fusing,
got fly H. brassicae. Yellow jackets, Vespula spp. grinding, and milling. Particle diameter is usu-
(Hymenoptera: Vespidae), are trapped with hep- ally 0.5 – 4.0 µm.
tyl butyrate [5870-93-9], C3 H7 COOC7 H15 .
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.
Dusts are the simplest formulations and gen- Slow-release formulations are especially
erally contain low concentrations (1 – 20 %) of useful for compounded pheromone and
56 Insect Control
Table 5. Average organochlorine insecticide residues in U.S. inhabitants (1964) ∗
kairomone attractants, in which reduced volatil- attenuation during food processing and prepa-
ity of the active ingredient is important and ration, measurable amounts of insecticides (i.e.,
decreased toxicity is essential. Formulations are pesticide residues) are detectable in food. A pes-
made with polymeric carriers such as starch, ticide residue is defined by the World Health Or-
cross-linked polyamides, carboxylated poly- ganization as “ any substance . . . resulting from
acrylates, or polyisobutylenes. Incorporation the use of a pesticide [that] includes any spec-
of the active ingredient in poly(vinyl chloride) ified derivatives such as degradation products,
(e.g., dichlorvos), or in hollow microfibers is metabolites, and reaction products which are of
used to release the sex pheromone attractants toxicological significance ” [120].
of the gypsy moth Porthetria dispar or the pink Residues of persistent pesticides in the hu-
bollworm Pectinophora gossypiella for male man diet are absorbed readily, stored in hu-
confusion. man adipose tissue, and secreted in body flu-
ids such as milk [121–123]. The effects of 20
Bait. The use of bait-containing attractants years of contamination of the U.S. diet by per-
and feeding stimulants together with a minimum sistent organochlorine insecticide residues are
amount of insecticide is a growing technology summarized in Table 5. Withdrawal of these in-
for integrated pest management. Examples in- secticides in the 1970s was followed by a de-
clude dried ground apple (pomace) and carbaryl cline in the average residue in human adipose
for the black cutworm Agrotis ipsilon, soybean tissue, as shown by comparison of surveys made
oil and amdro for the imported fire ant Solenop- in 1970 and 1978: DDT(T) 7.5 vs. 3.6 mg/kg,
sis invicta, cue-lure and naled for the melon fly dieldrin 0.44 vs. 0.19 mg/kg, and β-BHC 0.44
Dacus cucurbitae, and methyl eugenol and naled vs. 0.19 mg/kg[124].
for the oriental fruit fly D. dorsalis. The use of
bait greatly reduces the amount of toxic chem- Residue Persistence Curve. The magni-
ical applied and improves the selectivity of the tude and duration of pesticide residues on raw
insect pest management procedure. agricultural commodities depend on the dosage
applied, the chemical nature of the insecticide,
and the physical nature of the treated plant sur-
17. Insecticide Residues in Foods face. Standard technology involves determina-
tion of the content of insecticide residue, in
[118], [119]
milligrams per kilogram, in produce at intervals
after application. Because of the curvilinear na-
17.1. Residue Persistence ture of the results (Fig. 2A), estimation of the rate
of residue loss is difficult. For rapidly degrad-
Approximately 70 % of all insecticides are used able insecticides (most organophosphates and
in agriculture where they are typically applied carbamates), residue persistence follows first-
directly to raw agricultural commodities for order kinetics. When plotted on a semilogarith-
plant and animal protection. Despite weathering mic scale (Fig. 2B), straight lines are obtained
of crop residues during maturation and further 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
structure is altered to form smaller, more water- hydrolyzed principally at the reactive P−O –
soluble moieties. 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.
the Federal Insecticide, Fungicide, and Roden- 12. E. Y. Spencer: Guide to the Chemicals Used in
ticide Act (1972) [132], [133]. This law requires Crop Protection, 7th ed., Research Branch
“ that a pesticide may be registered if it is effec- Agriculture, Canada, 1980.
tive, properly labeled, and when properly used 13. R. Wegler (ed.): Chemie der Pflanzenschutz-
will not cause unreasonable adverse effects on und Schädlingsbekämpfungsmittel, vol. 1,
the environment taking into account the eco- 1970; vol 3, 1976; Springer-Verlag,
nomic, social and environmental costs and bene- Berlin-Heidelberg.
14. R. L. Metcalf, J. Agr. Food Chem. 21 (1973)
fits of the use of any pesticide.” The EPA residue
511.
tolerance values for most insecticides are pre- 15. A. W. A. Brown, J. Hayworth, A. R. Zahar, J.
sented in Table 6. Med. Entomol. 13 (1976) 1.
16. A. W. A. Brown, R. Pal: Insect Resistance in
Carcinogenicity. The Food Additive Arthropods, WHO, Geneva 1971.
Amendment (1958) to the U.S. Food, Drug, 17. G. P. Georghiou, T. Saito: Pest Resistance to
and Cosmetic Act (1938) states “ that no addi- Pesticides, Plenum Publishing, New York
tive shall be deemed to be safe if it is deemed 1984.
to cause cancer when ingested by man or an- 18. G. P. Georghiou, A. Lagunes-Tejeda: “ The
imal.” This concept has been used by EPA to Occurrence of Resistance to Pesticides in
Arthropods,” FAO Plant Protection Series,
cancel registration and tolerance for organochlo-
Rome 1991.
rine insecticides such as DDT, aldrin, dieldrin, 19. National Academy of Sciences: Pesticide
chlordane, heptachlor, and toxaphene. Resistance: Strategies and Tactics for
Management, Washington, D.C., 1986.
20. R. L. Metcalf, Annu. Rev. Entomol. 25 (1980)
18. References 219.
21. R. L. Metcalf in M. Kogan (ed.): Ecological
1. R. L. Metcalf, R. A. Metcalf: Destructive and Theory and Integrated Pest Management
Useful Insects, 5th ed., McGraw-Hill, New Practice, Wiley-Interscience, New York 1986.
York 1993. 22. A. W. A. Brown: Ecology of Pesticides,
2. H. H. Cramer: “ Plant Protection and World Wiley-Interscience, New York 1978.
23. C. A. Edwards (ed.): Environmental Pollution
Crop Production,” Pflanzenschutz Nachr. 20
by Pesticides, Plenum Publishing, New York
(1967).
1973.
3. J. W. Wright in R. L. Metcalf, J. J. McKelvey,
24. C. A. Edwards: Persistent Pesticides in the
Jr. (eds.): The Future for Insecticides,
Environment, 2nd ed., CRC Press, Cleveland
Wiley-Interscience, New York 1976, p. 17.
1973.
4. World Health Organization: 25. M. A. Q. Khan (ed.): Pesticides in Aquatic
“ Arthropod-Borne and Rodent-Borne Viral Environments, Plenum Publishing, New York
Diseases,” W.H.O. Tech. Rep. Ser. 719 (1985). 1977.
5. K. H. Büchel (ed.): Chemistry of Pesticides, 26. F. L. McEwen, G. R. Stephenson: The Use and
Wiley-Interscience, New York 1983. Significance of Pesticides in the Environment,
6. D. E. H Frear: Chemistry of Pesticides, 3rd Wiley-Interscience, New York 1979.
ed., Van Nostrand Princeton, N.J., 1955. 27. D. Pimentel: Ecological Effects of Pesticides
7. N. N. Melnikov: Chemistry of Pesticides (Eng. on Non-Target Species, President’s Office
Transl.), Springer, New York 1971. Science, Technology, Washington, D.C., 1971.
8. R. L. Metcalf: Organic Insecticides, 28. T. R. Eichers, P. A. Andrilenas, T. W.
Interscience, New York 1955. Anderson: “ Farmers’ Use of Pesticides in
9. R. D. O’Brien: Insecticides: Action and 1976,” USDA Agr. Econ. Rept. No. 418,
Metabolism, Academic Press, New York 1967. Washington, D.C., 1978.
10. W. Perkow: Wirksubstanzen der 29. J. F. Copplestone in D. L. Watson, A. W. A.
Pflanzenschutz- und Brown (eds.): Pesticide Management and
Schädlingsbekämpfungsmittel, Parey, Berlin Insecticide Resistance, Academic Press, New
1976. York 1977, p. 147.
11. H. H. Shepard: The Chemistry and Action of 30. M. L. Flint, R. van den Bosch: Introduction to
Insecticides, McGraw-Hill, New York 1951. Integrated Pest Management, Plenum
Publishing, New York 1981.
62 Insect Control
31. C. B. Huffaker (ed.): New Technology of Pest 51. P. Müller (ed.): DDT Insektizid
Control, Wiley-Interscience, New York 1980. Dichlordiphenyltrichloroäthan und seine
32. M. Kogan (ed.): Ecological Theory and Bedeutung, Birkhäuser, Basel 1958.
Integrated Pest Management Practice, 52. J. E. Cassida, L. J. Lawrence, Environ. Health
Wiley-Interscience, New York 1986. Perspect. 61 (1985) 123.
33. R. L. Metcalf, W. H. Luckmann (eds.): 53. G. Schrader: Die Entwicklung neuer
Introduction to Insect Pest Management, 3rd insektizider Phosphorsäure-Ester, 3rd ed.,
ed., Wiley-Interscience, New York 1994. Verlag Chemie, Weinheim 1963.
34. National Academy of Sciences: Scientific 54. C. Fest, K.-J. Schmidt: The Chemistry of
Aspects of Pest Control, Publ. 1402, Organophosphorus Pesticides, 2nd ed.,
Washington, D.C., 1966. Springer-Verlag, New York 1980.
35. National Academy of Sciences: Insect Pest 55. M. Eto: Organophosphorus Pesticides:
Management and Control, Publ. 1695, Organic and Biological Chemistry, CRC
Washington, D.C., 1969. Press, Cleveland 1974.
36. National Academy of Sciences: Pest Control
56. R. D. O’Brien: Toxic Phosphorus Esters,
Strategies for the Future, Washington, D.C.,
Academic Press, New York 1960.
1972.
57. WHO Environmental Health Criteria 63:
37. R. Carson: Silent Spring, Houghton-Mifflin,
Organophosphorus Insecticides, WHO,
Boston 1962.
38. R. L. Metcalf, Annu. Rev. Entomol. 25 (1980) Geneva 1986.
219. 58. W. Lange, G. von Kreuger, Chem. Ber. 65
39. R. L. Metcalf in R. L. Metcalf, W. H. (1932) 1998.
Luckmann (eds.): Introduction to Insect Pest 59. R. L. Metcalf, Bull. Entomol. Soc. Am. 5
Management, 3rd ed., Wiley-Interscience, (1959) 3.
New York 1994. 60. T. R. Fukuto, Bull. W.H.O. 44 (1971) 31.
40. D. L. Watson, A. W. A. Brown: Pesticide 61. T. R. Fukuto in C. F. Wilkinson (ed.): Pesticide
Management and Insecticide Resistance, Biochemistry and Physiology, Plenum
Academic Press, New York 1977. Publishing, New York 1978.
41. M. Jacobsen, D. G. Crosby (eds.): Naturally 62. R. D. O’Brien in C. F. Wilkinson (ed.):
Occurring Insecticides, Marcel Dekker, New Pesticide Biochemistry and Physiology,
York 1973. Plenum Publishing, New York 1976, p. 271.
42. J. E. Cassida (ed.): Pyrethrum, the Natural 63. T. Namba, Bull. W.H.O. 44 (1971) 289.
Insecticide, Academic Press, New York 1971. 64. S. D. Murphy in C. D. Klassen, M. O. Amdur,
43. C. B. Gnadinger: Pyrethrum Flowers, 2nd ed., J. Doull (eds.): Toxicology, 3rd ed., Macmillan
McLaughlin, Gormley, King, Co., Publ. Co., New York 1986, p. 519.
Minneapolis 1945. 65. R. L. Metcalf, Neurotoxicology 3 (1982) 269.
44. M. Elliott in R. L. Metcalf, J. J. McKelvey, Jr. 66. M. K. Johnson: “ The Delayed Neuropathy
(eds.): The Future for Insecticides, Caused by Some Organophosphorus Esters:
Wiley-Interscience, New York 1976, p. 163. Mechanism and Challenge,” CRC Crit. Rev.
45. M. Elliott, Bull. W.H.O. 44 (1971) 315. Toxicol. 3 (1975) 289.
46. R. A. Fuchs, R. Schroder in K. H. Buchel (ed.): 67. R. J. Kuhr, H. W. Dorrough: Carbamate
Chemistry of Pesticides, Wiley-Interscience, Insecticides, Chemistry, Biochemistry, and
New York 1983. Toxicology, CRC Press, Cleveland 1976.
47. T. Narahashi in R. L. Metcalf, J. J. McKelvey, 68. R. L. Metcalf, T. R. Fukuto, J. Agric. Food
Jr. (eds.): The Future for Insecticides, Chem. 13 (1965) 220.
Wiley-Interscience, New York 1976, p. 371. 69. R. L. Metcalf, Bull. W.H.O 44 (1971) 43.
48. C. F. Wilkinson in R. L. Metcalf, J. J.
70. WHO Environmental Health Criteria 64:
McKelvey, Jr. (eds.): The Future for
Carbamate Pesticides: a General Introduction,
Insecticides, Wiley-Interscience, New York
Geneva 1986.
1976, p. 195.
71. R. L. Metcalf, T. R. Fukuto, J. Agric. Food
49. G. T. Brooks: Chlorinated Insecticides, vols.
Chem. 15 (1967) 1022.
1, 2, CRC Press, Cleveland 1974.
50. R. L. Metcalf in R. L. Metcalf, J. J. McKelvey, 72. H. Gysin, Chimia 8 (1954) 205.
Jr. (eds.): The Future for Insecticides: Needs 73. M. J. Kolbezen, R. L. Metcalf, T. R. Fukuto, J.
and Prospects, Wiley-Interscience, New York Agric. Food Chem. 2 (1954) 864.
1976, p. 223.
Insect Control 63
112. J. M. Brand, J. C. Young, R. M. Silverstein, 121. W. F. Durham, Ann. N.Y. Acad. Sci. 160
“ Insect pheromones: a critical review of recent (1969) 183.
advances in their chemistry, biology, and 122. R. E. Duggan, Ann. N.Y. Acad. Sci. 160 (1969)
application,” Prog. Chem. Org. Natural Prod. 173.
37 (1979) 1. 123. A. Curley, R. Kimbrough, Arch. Environ.
113. Y. Tamaki in G. A. Kerkut, L. I. Gilbert (eds.): Health 18 (1969) 156.
Comprehensive Insect Physiology, 124. T. Nakatsugawa, M. A. Morelli in C. F.
Biochemistry, and Pharmacology, “ Sex Wilkinson (ed.): Insecticide Biochemistry and
pheromones,” vol. 9, Pergamon Press, Oxford Physiology, Plenum Publishing, New York
1985, p. 145. 1976, p. 61.
114. J. H. Borden in G. A. Kerkut, L. I. Gilbert 125. C. M. Menzie: Metabolism of Pesticides, U.S.
(eds.): Comprehensive Insect Physiology, Dept. Interior, Spec. Sci. Rept. Wildlife no. 96,
Biochemistry, and Pharmacology, 1966 ; no. 184, 1974, no. 212, 1978.
“ Aggregation pheromones,” vol. 9, Pergamon 126. H. Aizawa: Metabolic Maps of Pesticides,
Press, Oxford 1985, 257. Academic Press, New York 1982.
115. P. A. Hedin, F. G. Maxwell, J. H. Jenkins in 127. G. Vettorazzi: “ Toxicological Decisions and
F. A. Maxwell, F. A. Harris (eds.): Proc. Recommendations Resulting from the Safety
Summer Inst. of Biol. Control of Plants, Assessment of Pesticide Residues in Food,”
Insects, and Diseases, “ Insect plant CRC Crit. Rev. Toxicol. 4 (1975) 125.
attractants, feeding stimulants, repellents, 128. WHO/FAO Tech. Rept. Ser. no. 370, 1963;
deterrents, and other related factors affecting 417, 1969; 458, 1970; 474, 1971; 502, 1972;
insect behavior,” University Press, Jackson, 525, 1973; 545, 1974; 574, 1975.
Missisippi, 1974, p. 494. 129. U.S. Environmental Protection Agency:
116. R. L. Metcalf, E.. R.. Metcalf: Plant Compendium of Registered Pesticides,
Kairomones in Insect Ecology and Control, Washington, D.C.
Chapman & Hall, New York1992. 130. L. P. Van Dyk, I. H. Wiese, J. E. C. Muller,
117. S. Finch in T. H. Coaker (ed.): Applied Biology Residue Rev. 82 (1984) 37.
“ Chemical attraction of plant feeding insects 131. G. Vettorazzi: International Regulatory
to plants,” V., Academic Press, New York Aspects of Pesticide Chemicals, vols. 1, 2,
1980, p. 67. CRC Press, Cleveland, Ohio, 1979.
118. F. A. Gunther, J. Gunther (eds.): Residue Revs., 132. R. A. Merrill in C. D. Klassen, M. O. Amdur, J.
vols. 1 – 86, Springer Verlag, Hamburg Doull (eds.): Toxicology, 3rd ed., McMillan
1962 – 1987. Publ. Co., New York 1986, p. 917.
119. H. Maier-Bode: 133. National Academy of Sciences: Regulating
Pflanzenschutzmittel-Rückstände, E. Ulmer, Pesticides, Washington, D.C., 1980.
Stuttgart 1965. 134. P. Dahl, L. Rattan: J. Assoc. Off. Anal. Chem.
120. N. A. Smart, Residue Rev. 98 (1985) 99. 54 (1971) 1095.
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)
then insulators act as dielectrics and have the modification, often increase their value. Poly-
additional function of storing electrical energy. mers (thermoplasts, thermosets, and elastomers)
Not all materials of high resistivity (see Sec- are mainly synthetically manufactured products.
tion 2.2) are suitable for use as electrical insulat- Electric insulating materials are generally in-
ing materials. Additional chemical, mechanical, dustrial products of complex chemical struc-
thermal, and, depending upon the application in- ture which have been adapted to suit their in-
volved, further properties are required. Stabil- tended application by treatment, modification,
ity toward various electrical and nonelectrical and blending.
influences [e.g., toward electrical discharges in Compared with metallic conductors and mag-
cavities and at interfaces (see Section 2.8), to- netic substances, insulating materials represent
ward heat produced by dielectric dissipation (see the most comprehensive material group within
Section 2.5) and toward ignition and burning] the field of electrotechnics and possess the great-
is of major importance. The working life of an est number of technically interesting properties.
insulated electric circuit depends mainly upon Nonelectrical properties have been considered
the thermal aging behavior. Although a mate- here only to an extent sufficient to permit a brief
rial may be technically suitable as an insulator, characterization.
unfavorable processing properties or the manu- Groups of organic insulating materials are
facturing price can restrict its applications. listed in Chapter 3 according to their chemi-
The selection of an insulating material lies in cal composition. In most cases, organic products
the hands of the designer, provided the proper- based on rubber or other polymers can be modi-
ties necessary for the function in question are fied by processing. Hence an absolute value for
guaranteed. To ensure this, materials are tested a property of a given insulator group does not
according to national standards or IEC publica- exist. For this reason, data are not presented in
tions. The results of these tests permit compari- tabular form but as performance relief diagrams.
son of the insulating materials under the speci- The values shown in each block diagram merely
fied conditions. Data for a specific material can- indicate the range of the properties of main in-
not generally be applied to an insulating compo- terest. Organic products which find application
nent because the component and its insulation in the electrotechnics industry are generally em-
properties are influenced by design and manu- ployed with additives, which can often consid-
facturing processes. The final decision regarding erably influence electrical and dielectric proper-
functionality must be made by testing the com- ties. The most common additives include pro-
ponent itself. cessing aids, stabilizers, pigments, fillers, and
Insulating materials can be subdivided into flame retardants.
inorganic and organic materials. Inorganic insu-
lators include natural products such as mica or
asbestos. Synthetic insulators made of ceram- 2. Properties and Testing Methods
ics, glass, or more recently glass ceramics are of
greater significance, however. Organic insulat- The electrical behavior of electrical insulating
ing materials are primarily synthetic polymers. materials results from the physical processes of
All of these insulating materials have a secure dissociation, polarization, and ionization.
position within the electrical industry, because Certain assumptions about the formation of
of their special material properties or the ef- free electrons and bound charge carriers un-
ficient processes used to manufacture compo- der the influence of electrical fields are asso-
nents from them. Developments have afforded ciated with these terms. They are responsible
new materials with improved properties. Devel- for electrical resistance, dielectric polarization,
opments within the electronics and electrical in- and electrical breakdown. In the broadest sense,
dustries make the availability of new improved these measurable parameters can be regarded as
insulating materials essential physical properties. Electrical insulating mate-
In their original form, natural materials are rials are materials with a resistivity >106 Ω · m
seldom able to meet the diverse requirements (>108 Ω · cm). The energy values which the
demanded of an electric insulating material; electrons in a solid are able to assume can be
however, purification, preparation, or chemical
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- solution B is identical, but a specified amount
dard ANSI/ASTM 3638-1977 [25] forms the ba- of a given wetting agent is also added. The test-
sis for tests according to the testing regulations ing solutions therefore differ in their electrolytic
ANSI/UL 746 A-1984 [26]; these differ from the conductivity because the wetting agent intro-
currently valid international standard (IEC Pub- duces almost the same contribution to the to-
lication 112-1979) in the type of evaluation pro- tal conductivity as the ammonium chloride salt.
cedure employed and in the lack of opportunity The testing solutions also differ in their surface
for using a more aggressive testing solution (B). tension.
During testing, surface contamination is sim- According to the definition given in the stan-
ulated by allowing an appropriate test fluid to dards, a track has developed when the tracking
drip onto the surface. The IEC Publication 112 current on the tracking path is greater than or
specifies five separate determinations of a.c. equal to 0.5 A for at least 2 s, or when the spec-
voltage (50 – 60 Hz) at which 50 drops of test imen begins to burn as a result of the heat gen-
solution between the platinum electrodes do not erated. A circuit used for evaluating tracking is
lead to failure. In contrast, ASTM 3638-1977 shown in Figure 2.
stipulates one measurement at a number of test The test is carried out at voltage steps be-
voltages and comparison with a standard curve. tween 100 and 600 V. The maximum test voltage
The point of intersection of this curve with the which can be divided by 25 without a remainder
50-drop line is determined on the plot of droplet represents the comparative tracking index CTI,
number as a function of the testing voltage; the provided that the test specimen (which must be
test voltage is read off the abscissa (voltage axis) at least 3 mm thick) can also withstand a fur-
and the CTI is thus determined (Fig. 1). 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.
constant end value. The drop in current can be uid insulators. Recommendations for measuring
caused by dielectric polarization and by mi- the insulation resistance at elevated temperature
gration of highly mobile ions toward the elec- are given in IEC Publication 345 [29].
trodes. The insulation resistance is generally de-
termined 1 min after the beginning of the elec-
trification.
layers is also influenced by the nature and struc- pacity of a capacitor filled with a dielectric and
ture of the surface of the material. For example, that of a capacitor of identical dimensions but
in inorganic insulating materials the formation incorporating a vacuum between the electrodes
of adsorbed water layers is related to the phase is known as the relative permittivity εr of the di-
characteristics of the surface: the polished sur- electric. Under d.c. conditions, the relative per-
faces of ceramic insulator components often ex- mittivity is characteristic for most materials, but
hibit a greater surface conductance than the fired, is temperature dependent. Under a.c. conditions,
glass-enriched surface. A significant increase in it moreover depends on frequency and in certain
surface conductance can be observed with in- cases on field strength.
creasing air humidity. The relative permittivity is a dimensionless
Surface resistance is obtained by dividing the comparative value. The procedure for determin-
d.c. voltage between two electrodes in contact ing capacitance is given in the international stan-
with the surface of a sample by the current flow- dard IEC Publication 250 [30].
ing between these electrodes 1 min after electri- Insulating materials can be divided into three
fication (see Section 2.2). The surface resistivity groups according to relative permittivity:
σ is the surface resistance expressed relative to a
square area. The length of the sides of this square 1) With most materials, the positive and nega-
is, in theory, immaterial [27]. tive charge centers coincide. Under the in-
Generally, current mainly flows through the fluence of an electric field, electron polar-
surface layer of the specimen and through the ization, and in some cases ion polarization,
moisture and surface contamination associated occur. Such insulating materials possess low
with it. However, the measured surface resis- relative permittivities of 2 – 10.
tance also includes a nondefinable contribution 2) In the second group, a dipole moment is al-
from the resistance of the bulk of the sample. ready present because of the molecular or
Thus, surface resistance cannot be measured crystal structure. In addition to the electron or
precisely but only approximately. The measured ion polarization, the external field introduces
value is generally an indication of the state of orientation polarization, assuming of course
the surface at the time of measurement. The rel- that the dipoles are sufficiently mobile. How-
ative permittivity (see Section 2.4) of the sample ever, this is rarely the case with organic in-
under investigation affects the form and extent sulating materials, or even with liquid insu-
of contamination [27]. The current which flows lation oils of relatively high viscosity. With
as a result of contamination is, however, also inorganic insulating materials, the dipoles in
influenced by the surface structure and surface the crystal structure may reorientate, which
properties of the specimen. Hence the surface leads to a considerable increase in capacity.
resistivity is not a material property in the nor- Such insulating materials are characterized
mal sense, but it can be regarded as such if the by a significantly higher relative permittivity
effects of contamination are taken into consid- (e.g., titanium dioxide with a relative permit-
eration, e.g., by appropriate preconditioning. tivity of ca. 100).
The test procedure has been specified by the 3) In ferroelectric materials that undergo spon-
international standard IEC 93 [27]. taneous polarization (e.g., barium titanate),
complete regions are reorientated by an ex-
ternal electrical field. This leads to relative
2.4. Relative Permittivity permittivity values of ≥ 1000.
Polarization leads to the internal charging of the The transportation of charge carriers in an
material in a direction opposing that of the ex- electric field and the continuous reversal of po-
isting field. This causes a reduction of the field larization in the a.c. field both incur energy
strength within the capacitor and a correspond- losses. They cause a certain degree of heating of
ing increase in the capacity when the volume the capacitor dielectric. The power requirement
between the electrodes is completely and exclu- for this is taken from the a.c. electrical field.
sively filled with an insulating material as op-
posed to being empty. The ratio between the ca-
Insulation, Electric 7
2.5. Dielectric Dissipation Factor With glasses and ceramic materials incorpo-
rating a glassy phase, the network-modifying al-
The dielectric dissipation factor (dielectric loss) kali ions have a higher mobility than the dou-
is the tangent of the loss angle δ by which the bly bonded alkaline-earth ions. In low-loss in-
phase displacement between current and volt- sulating materials, the fluxing agents (which are
age in a capacitor deviates from π/2 when the normally based on alkali compounds) are sub-
dielectric of the capacitor consists solely of the stituted by equivalent materials based on alka-
insulating material under investigation (Fig. 5). line earths. Examples include “low-loss steatite”
At the same time, it is also the ratio of the loss or and “alkaline-earth porcelain”. Improvements
effective power N e to the reactive power N r of in dielectric properties are, however, associ-
a capacitor completely filled with the insulating ated with manufacturing problems; alkali com-
material and subjected to an a.c. voltage. pounds used as fluxing agents cannot usually be
substituted.
mechanical state of the specimen, tween the two electrodes across which the volt-
homogeneity, age is applied. The term electrical strength is
surrounding medium, and only valid when a uniform field exists between
electrode size and positioning. the electrodes.
Determination of the breakdown voltage is
Hence all of the specifications listed in the
specified in the international standards IEC Pub-
standards (e.g., international standard IEC Pub-
lication 243 [31] for solids and IEC Publication
lication 243) must be met exactly if compara-
156 [32] for liquid insulating materials.
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- 2.8. Further Electrical Tests
age does not in general possess the character of
a material-specific property. It is more a crite- For special applications or investigations, sup-
rion for the degree of defectiveness of the insu- plementary test procedures have proved valu-
lator because the weakest regions (e.g., inhomo- able; they have gained international recognition
geneities, voids, gas inclusions) are monitored and have been standardized.
during measurement.
Breakdown behavior of insulating materials High-Voltage Tracking Test. This test pro-
varies according to material type and testing con- cedure (Fig. 7) is utilized for materials that are
ditions; differences depend on various break- used under severe ambient conditions in electri-
down mechanisms (Fig. 6). In general, a purely cal engineering (e.g., high humidity, high volt-
electrical breakdown is observed during impulse ages). The test enables assessment of the stabil-
tests (e.g., exposure to peak loadings), and the ity towards track formation and erosion. Track
values of breakdown voltage are correspond- formation is defined as a process in which track-
ingly high. During long-term field strength test- ing and erosion are the result of an electrical dis-
ing, thermal breakdown occurs as a result of lo- charge at or adjacent to the contaminated surface
cal instability caused by the heat generated by in- of an insulating material.
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.
a contamination liquid. The contamination liq- Low Voltage – High Current Arc Test. This
uid consists of distilled water containing ammo- test is used to investigate the behavior of heat-
nium chloride and a specified wetting agent. The resistant materials. These are generally inor-
specimen has the dimensions of 5×12 cm. High ganic substances or organic materials which are
voltages of 2.5, 3.5, or 4.5 kV are applied to the highly filled with inorganic substances. The test
two electrodes fastened on the smaller edges of determines (1) the distance over which a car-
the specimen. The contamination is allowed to bon arc can be extended across the surface,
flow at a specified rate over the lower surface of (2) whether the arc path is conducting, and
the specimen. The time (up to 6 h) is observed (3) whether the material underneath the surface
for failure caused by tracking or burning. changes significantly after cooling. The test pro-
The tested materials are classified into four cedure is specified in DIN VDE 0303 part 5 [35].
categories according to their resistance to track-
ing which depends on the test voltage and time. Assessment of Electrostatic Behavior.
A correlation exists between this test and the Electrostatic behavior can be assessed in a test
low-voltage tracking test (see Section 2.1). The method in which either disk-shaped samples
procedure is specified in the international stan- (rigid materials) or endless strips or tapes (flexi-
dard IEC Publication 587 [33]. ble materials), are charged using contacting sub-
strates. The field strength generated by separa-
Arc Resistance. The test procedure accord- tion of the specimen and the substrate is mea-
ing to ASTM-D 495 uses a high-voltage sup- sured as well as the change of the charge as a
ply of 12.5 kV in order to generate arcing across function of time. Since the tests are designed
the surface of the specimen [34]. Contrary to to allow the selection of nonchargeable or only
the high-voltage tracking procedure according to slightly chargeable materials, they are only of
IEC Publication 587 (see above), the samples are advantage for materials with a resistance at the
tested in the absence of arbitrary contaminants. lower end of the insulation range (109 – 1011 Ω).
The electrodes are round tungsten rods. The an- The test is related to the determination of sur-
gled tips of these rods are placed in contact with face resistance (see Section 2.3). Procedures are
the upper or under surface of the specimen under specified in DIN VDE 0303 parts 8 and 14 [36],
investigation at a defined angle and at a distance [37].
of 6.35 mm apart (Fig. 8). The insulating mate-
rial is classified according to the duration of the Determination of Corrosion-Promoting
arc in seconds. At the end of each test minute, Effects. The corrosion-promoting effects of in-
the power effect on the surface of the specimen sulating materials can be determined by three
is doubled. 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
Figure 8. Arc resistance test (ASTM) presence of a corrosion-promoting effect.
Arcing occurs between the tungsten electrodes across the
sample surface. 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- in copolymers; deformation; aging; irradiation;
rode. This can be simulated by testing under cryogenic processing; crystallinity; conducting
the simultaneous affects of a d.c. supply, heat mechanisms; activation energy; stereo com-
(e.g., 40 ◦ C), and environmental humidity (e.g., plexes; thermal histories; cross-linking; mois-
92 % R.H.). The test reveals which combina- ture content; and relationships between mechan-
tions of insulating material and metal are par- ical relaxation, elongation, and calorimetry.
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].
The strength of ceramic insulators can be in- also suitable for application at elevated temper-
creased considerably by application of suitable atures.
glazes. The thermal expansion coefficient of the
glaze must be slightly lower than that of the ce- Thermal Conductivity. The thermal con-
ramic substrate. Following cooling after firing, ductivity of the insulating materials is gener-
the glaze is subjected to compressive stresses. ally low. Hence, they can be simultaneously em-
When an external tensile force is applied to a ployed as electrical and thermal insulators. How-
glazed article, these compressive stresses in the ever, oxidic and nonoxidic materials are avail-
surface must be counteracted before the tensile able which have a similar thermal conductivity
stresses can become effective. In glasses, sim- to metallic substrates. Such insulators are nor-
ilar effects are achieved by thermal hardening, mally employed when heat transfer or removal
i.e., by quenching from temperatures above the is necessary.
glass transition temperature. Chemical harden-
ing (via ion exchange at elevated temperature) Specific Heat. The specific heat of the insu-
is used less widely. lating material may be of significance in heat
With organic polymers, similar increases in storage applications. Magnesium oxide with a
strength can be gained by employing composite specific heat of greater than 1000 J kg−1 K−1 is
materials: glass or carbon fibers are incorporated particularly suitable.
into the polymer.
Combustion Behavior. The possibility of
Material Space Filling. Apart from com- combustion can be a considerable obstacle re-
pletely solid materials, porous substances are garding the use of organic insulators. Materials
also used in electrical engineering. Porosity low- that do not burn at all when placed in contact
ers the breakdown strength, therefore porous with an electrical source of heat are given prior-
materials are not considered for high-voltage ap- ity. Classification criteria include the extinguish-
plications. In contrast to this, an open porous ing of the flame following removal of the heat
structure is favorable for the thermal shock re- source, or the speed at which the flames spread.
sistance because porous materials have a lower
Young’s modulus. Porous materials are there-
fore often employed in thermal electrical engi- 3. Insulating Materials
neering. An increase in the surface conductiv-
ity caused by the larger surface does not pose 3.1. Inorganic Insulating Materials
a threat because the water film is evaporated by
the current flow. Inorganic electrical insulating materials include
Under high-vacuum conditions, porous ma- several natural minerals (generally mica or as-
terials are also utilized as electrical insulators to bestos) but also industrially manufactured prod-
prevent the formation of continuous vapor lay- ucts. The wide range of industrially manufac-
ers. tured materials includes pure glass, materials
containing glassy and crystalline structural com-
Thermal Expansion. The thermal expan- ponents, purely crystalline materials, and even
sion of inorganic insulators is generally rela- single crystals. This broad spectrum of mate-
tively low which is favorable for thermal shock rials ensures a wide variety of properties. De-
resistance. A low thermal expansion also en- pending upon the intended application, one or
sures good shape retention following tempera- more of the following properties may be particu-
ture change. Inorganic insulators can withstand larly pronounced: low or high relative permittiv-
rapid temperature increases better than rapid ity; extremely low dielectric loss; high electrical
cooling. strength at low and high frequencies; excellent
Organic substances are prone to material insulation characteristics even at temperatures
changes as a result of prolonged exposure to high well above 1000 ◦ C; low or high thermal con-
temperature, which affects both the mechanical ductivity; high mechanical strength.
and electrical properties. Special organic com- The ferrites are also inorganic insulating ma-
pounds have recently been developed which are terials. Ferrites are sintered products consisting
12 Insulation, Electric
of iron(III) oxide and oxides of mono- or diva- Composite Insulating Materials. Large
lent metals. Their application within electrical amounts of mica powder and scrap are pro-
engineering is primarily governed by their mag- cessed to nonwoven substrates or sheets. In
netic properties. Ferrites are generally regarded one process, scrap mica is heated to 800 ◦ C; a
as magnetic substances (→ Magnetic Materi- portion of the water of crystallization escapes
als), although their insulation characteristics are and causes partial cleavage of the lamellae. The
of interest for certain applications. mica is then immersed in sodium carbonate so-
lution whilst still hot. Carbon dioxide gas is
evolved during subsequent treatment with di-
3.1.1. Insulating Materials Based on Natural lute acid, and causes complete separation of the
Minerals mica lamellae. The very fine mica platelets can
3.1.1.1. Mica and Related Products be processed into thin, endless, nonwoven sub-
(→ Mica) strates in a procedure analogous to that used in
the manufacture of paper.
Micas are sheet silicates. Their properties are In another process, the mica is cleaved using
determined by mixtures of homo- or het- a high-pressure water jet and again processed to
eropolar bonds within the lamellar planes, a nonwoven fabric. The product is slightly less
which in turn are held together by van der densely packed than that resulting from the pro-
Waals’ forces. This type of bonding is re- cedure mentioned previously.
sponsible for several of mica’s properties, Such nonwoven fabrics, and small rejected,
e.g., the excellent cleaving properties paral- cleaved mica fragments are combined with natu-
lel to the plane of the lamellae. Muscovite ral or synthetic binders to give sheets, films, and
[1318-94-1] , KAl2 [AlSi3 O10 (OH)2 ] , and phlo- tapes (micanite, VDE 0332, IEC 371). Suitable
gopite [12251-58-0] , KMg3 [AlSi3 O10 (OH)2 ] , binders include asphalt, shellac, alkyd resins, or
havefound widespread application in the elec- more recently, silicone and epoxy resins.
tronics industry. Materials selected for insula- Micanite is a binder-containing sheet prod-
tion purposes must not contain contaminants uct that is manufactured using heated hydraulic
which may adversely effect insulation proper- presses. Commutator micanite should not con-
ties. The mica products employed in electronics tain more than 4 % binder. Sheets are ground
mainly originate from India. to meet the required thickness within the toler-
ance limits, and then punched into commutator
Production and Processing. The mica is se- segments.
lected by hand from the mined rock and sorted Heating micanite is designed for use as in-
according to size. Small fragments are ground sulation in electric heaters; here, the binder im-
and processed into powder. Larger specimens mediately ignites when the device is first uti-
are cleaned and cleaved into sheets of thick- lized. The remaining mica layer is responsible
ness as low as 0.01 mm. Commercially available for insulation. With an appropriate binder, only
products are classified according to purity and short-term inconvenience is caused by the com-
surface area. bustion products. The binder should not produce
Cleaned, cleaved mica is generally processed an electrically conducting residue after com-
by punching. The requirements of the punched bustion. Shellac is a very satisfactory binder in
component regarding tolerances and precise fin- amounts up to 3 %. Heating micanite is also man-
ishing are extremely high. During manufacture ufactured from fine mica and contains 10 to 15 %
of capacitors, the metallizing preparation (e.g., silicone resin as an additive. The organic compo-
a silver suspension) is applied directly onto the nents ignite to form gaseous combustion prod-
mica substrate and fired onto the surface at ucts; the residual silica functions as a cement and
600 ◦ C following drying. confers the required strength to the insulator.
Mica products can be processed to vacuum- Rigid micanite, which cannot be shaped, is
sealed components. Lead borate glasses are par- obtained on addition of increased amounts of
ticularly suitable solders because of their low binder. If subsequent shaping is desired, ther-
melting temperatures. Such composites can be moplastic binder materials must be employed.
used up to 300 ◦ C.
Insulation, Electric 13
Mica foil consists of a thin substrate covered insulation capabilities even at elevated temper-
with one or two layers of cleaved mica or fine atures.
nonwoven mica material that are bonded to- Asbestos materials cause severe lung dam-
gether with a binder. Depending upon the type of age (asbestosis), which has led to a ban on their
substrate employed, the products are known as use. They have been replaced by fiber materials
paper micanite, fabric micanite, nonwoven mi- produced from ceramics, particularly alumina
canite, or film micanite (VDE 0332, IEC 371). ceramic fibers. Their properties are equal to or
“Glimmerband” (mica tape) denotes mica foil even better than those of asbestos. It has not yet,
which has been slit into tapes. however, been clarified whether these alterna-
Glass – mica composite materials can also tive fibers can also cause cancer. No regulations
be manufactured with low-melting lead borate have been laid down so far. The IEC standards
glasses. With appropriate selection of the glass and draft standards for the standardization of as-
component, the semifinished article can be ma- bestos products have not yet been withdrawn.
chined.
Asbestos [1332-21-4] is a fibrous form of mag- The glass insulators are classified into six
nesiumsilicate. It has an extremely high thermal groups (see next page).
stability as well as good electricaland thermal
14 Insulation, Electric
G-100 alkali – lime – silica glasses
G-200 chemically resistant borosilicate glasses
components to be estimated from the character-
G-300 electrically resistant borosilicate glasses istic data obtained with standardized samples.
G-400 alumina – lime – silica glasses With glass, the thermal history of the material
G-500 lead oxide – alkali – silica glasses
G-600 barium oxide – alkali – silica glasses is important. Prestressed glass is influenced by
the conditions experienced during thermal hard-
ening. Here, it is even more difficult to deduce
According to IEC 672, ceramic materials are the properties of an insulation component from
those substances which are shaped before fir- those of a standard specimen.
ing; their main components are usually poly-
crystalline silicates, aluminosilicates, titanates,
or oxides. Glasses are dense inorganic materi- 3.1.2.1. Porcelain
als manufactured from a melt or by sintering, Porcelain was introduced as an electrical insulat-
which solidify without crystallization. Glass ce- ing material by W. von Siemens in the middle of
ramics are manufactured from glass, they con- the nineteenth century. Various types of porce-
tain a considerable crystalline content as a re- lain have since become the material of choice
sult of a special treatment. Glazes are coatings for high-voltage insulators in overhead lines and
with a smooth surface which are generated in transformer stations. Normal porcelain is less
a melt process and which essentially consist of suitable for high-frequency and electrical heat-
glass. They can contain coloring components or ing applications because the desired properties
opacifiers. cannot be obtained due to the presence of alkali
The above classification of ceramic insulators from the fluxing agent. For special applications,
is not systematic. Apparent inconsistencies are alkaline-earth porcelains are available in which
due to historical reasons. Because the preced- the alkali elements are eliminated. Only small
ing standard, DIN 40 685, had proven valuable components can be manufactured from these
for several decades and was internationally ac- materials, however, because their firing interval
cepted, its classification system was adopted in (i.e., the temperature range between vitrification
the new IEC 672 standard. Some groups are par- and distortion) is very small.
tially defined on the basis of material aspects.
However, structure is also decisive, in particu- Material Types. The C-100 group according
lar the presence of an open pore system. For to IEC 672 Part 3, comprises five material types
example group C-500 contains only porous ma- which differ with regard to their phase compo-
terials, whereas dense substances of the same sition and required properties, particularly the
chemical composition are classified in another mechanical data:
group. Porous materials are not limited to group C-110 quartz porcelain
C-500 however. Certain groups (e.g., C-100 and C-111 compressed porcelain (with a
C-200) contain both porous and dense varieties permitted residual porosity)
C-112 cristobalite porcelain
of a given composition. C-120 alumina porcelain, strong
Testing.IEC 672, Part 2 describes the testing C-130 alumina porcelain, ultrastrong
procedures used for determining the mechanical,
thermal, and electrical properties of the individ-
ual materials. Precise specification of test pro- The compressed porcelains of type C-111
cedures, including sample preparation, is nec- are only suitable for special low-voltage appli-
essary because data are strongly influenced by cations due to their permitted open porosity.
the manufacturing process, shape, and dimen- Cristobalite porcelains exhibit better mechani-
sions of the sample, and also by the measuring cal properties than the quartz porcelains. Alu-
procedure itself. Accordingly, data are strictly mina porcelains are now the preferred porcelains
valid only for the stipulated sample and cannot for high-voltage insulators. Their classification
be applied to samples that are of another shape or in the two types C-120 and C-130 takes into
are manufactured by another process. However, consideration recent developments with high-
modern fracture mechanics and Weibull statis- strength materials being classified as type C-
tics allow the properties of actual construction 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.
sulators, suspension insulators, and terminal in- a MgO-enriched body and contains forsterite
sulators for overhead line construction; insula- [15118-03-3] (Mg2 SiO4 ) as the main crys-
tor housings able to withstand extreme mechan- talline phase. The porous types (C-230 and C-
ical loading; or as external shielding for voltage 240)allow the manufacture of high precision
transformers of heights up to 5 m and diameters components from fired semifinished products
of 1 m. Lead-through insulators can also be pro- using conventional machining techniques.
duced with multiple bore holes running parallel
to the longitudinal axis. Properties. Several of the most important
The alkaline earth porcelains are mainly em- properties of the C-220 and C-221 types are
ployed as resistor bodies. listed in Table 2.
Type C-220 steatite is manufactured using
feldspar as a fluxing agent. This yields an ad-
3.1.2.2. Steatite equate firing range for the production of larger
items. With low-loss steatite (type C-221), di-
The term steatite [14807-96-6] has two mean- electric properties are considerably improved
ings, firstof all as a synonym for the fine-grained when barium carbonate is used as a fluxing
variety of talc, and secondly fora special ceramic agent. This material is thus also suitable as a
material manufactured from this talc. dielectric for capacitors. For some applications
To avoid confusion, the term steatite is used its relatively high mechanical strength is also of
here to denote the ceramic product. The term significance. With the exception of high-voltage
talc (and not steatite) is used for the ceramic raw insulators, low-loss steatite is the most popular
material regardless of its crystal size. Normally, variety of steatite.
however, the term talc solely denotes the coarse- Compared with the porcelains, the steatites
grained variety of the mineral. have outstanding electrical properties. Only the
best alumina porcelains have mechanical prop-
Material Types. Group C-200 according to erties comparable with those of steatite. Recent
IEC 672 contains six types of material based on investigations have shown that the resistance of
talc: steatite to crack propagation is slightly greater
Type C-210 steatite for low-voltage applications (not dense) than that of alumina porcelain [43].
Type C-220 steatite
Type C-221 low-loss steatite
Type C-230 steatite, porous Raw Materials. The preferred raw mate-
Type C-240 forsterite, porous rial for the manufacture of steatite is talc
Type C-250 forsterite, dense
[14807-96-6] , Mg3 [Si4 O10 (OH)2 ] in its fine
crystalline form (→ Talc). The random orienta-
The most important materials in group C- tion of the crystals in this talc allows the man-
200 are steatite (type C-220) and low-loss ufacture of ceramic products with low particle
steatite (type C-221). Type C-250 is formed from
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
orientation. One of the most important talc de- mosphere. The firing interval is lower than that
posits is found near Göpfersgrün – Thiersheim of porcelain, particularly with low-loss steatite.
in the Federal Republic of Germany. It repre- The lower firing interval for steatites is mainly
sents the main basis for the worldwide manu- responsible for the fact that steatite insula-
facture of large, high-voltage steatite insulators. tors can only be manufactured with maximum
Talc for use in ceramics originates in Australia, heights of ca. 1.2 m.
the United States, Spain, and China. During firing, dry-pressed components
Only talcs with low levels of contamination shrink by only 10 to 12 %, and hence are suitable
can be used as a raw material for the manufacture as insulator components with high dimensional
of steatite. The maximum upper limits which tolerances.
can be tolerated are 1.5 wt% Al2 O3 and 2.0 wt% Metal layers of silver, platinum, iron, or
Fe2 O3 [44]. Recently, calcia has been regarded molybdenum can be used as electrodes on
as less critical and may be present in amounts steatite or as a soldering base for joining steatite
up to 1 wt% [45]. The steatite body contains 80 to metals of other ceramic components.
to 90 % talc along with 5 to 10 wt% of a fluxing During firing, the talc is transformed to pro-
agent (in steatite, feldspar; in low-loss steatite, toenstatite, MgSiO3 . The additives (clay and
barium carbonate) and plastic clay to improve feldspar or barium carbonate) form a melt phase
the processing properties. which vitrifies on cooling; the glass content of
the fired body may be as high as 25 to 50 vol%
Manufacture. Components from steatite are [46]. An excess of silica leads to crystallization
produced according to the general principles em- of cristobalite from the melt; this occurs in cer-
ployed in the ceramics industry (→ Ceramics, tain steatite bodies, but particularly in feldspar
General Survey). All molding and shaping tech- steatites. Improved thermal shock resistance is
niques can be employed such as casting, shaping observed with steatites in which the silica has
in the plastic state, or pressing, and when neces- been completely taken up by the melt phase, for
sary subsequent machining in the leather-hard or example in most low-loss steatite products con-
dried state. The most important technique, par- taining silica-free fluxing agents.
ticularly for small components, is pressing. The
dry pressing process only gained widespread use Uses. Steatite finds widespread use as an in-
after its development with talc bodies. Talc bod- sulating material in electrical engineering. In
ies have proved to be particularly suitable for low-voltage applications, it is employed for the
this technique because the low hardness of talc manufacture of insulation components in which
(Mohs hardness = 1) guarantees a long service a high degree of dimensional precision is re-
life. quired (e.g., for mechanically mounted compo-
The steatite components are usually fired at nents). Typical insulators made from steatite in-
1300 – 1400 ◦ C in an oxidizing or reducing at- clude sockets, inserts for plugs, levers and rollers
18 Insulation, Electric
for rotary and rocker switches, holders, and car- 3.1.2.3. Materials for Electrical Heat
tridge fuses. Steatite has many advantages over Engineering
porcelain when greater mechanical loading has
to be withstood. Two of the most important demands for the use
In high-voltage applications, steatite can now of ceramic materials in electrical heat engineer-
be replaced by alumina porcelain. Abrasion of ing are adequate stability at high temperature
machining tools is, however, considerably lower and favorable thermal shock resistance.
with steatite bodies. On the other hand, alumina The thermal shock resistance is determined
porcelain bodies are easier to fire and are suitable by the phase composition: compositions which
for the manufacture of components much larger approach that of a eutectic melt at relatively low
than those made of steatite. Steatite shows clear temperatures and are therefore unsuitable. Ther-
advantages in high-frequency applications, for mal shock resistance increases with increasing
example as bases for antenna masts and for ter- mechanical strength and thermal conductivity,
minal insulators in transmission masts. but also with decreasing modulus of elasticity
Low-loss steatites are suitable for the man- and decreasing thermal expansion. It also de-
ufacture of capacitors, e.g., the complex thick- pends upon external conditions, such as shape
walled capacitors used in high-voltage applica- and dimensions of the article, as well as on the
tions. rate of heat exchange with the surroundings. In
this respect, the thermal shock resistance is not
Talc Bodies Enriched in Magnesia. The a material constant.
crystalline phase forsterite, Mg2 SiO4 , is de- Young’s modulus is related to the structure of
liberately formed in fired bodies of the material the material; it is lower with porous structures
types C-240 and C-250 by enriching the talc than with dense structures.
with magnesia (e.g., by addition of magnesite, The most important factor controlling ther-
MgCO3 ). Several properties of type C-250 ma- mal shock resistance is thermal expansion. In
terials are summarized in Table 2. materials with a low thermal expansion coeffi-
The outstanding properties of forsterite ce- cient, the product is only exposed to low stresses,
ramics include the particularly low dielectric even if the thermal shock is high.
losses and the large thermal expansion coeffi- Hence the most suitable materials for electri-
cient. The latter permits low-stress soldering to cal heat engineering are those which are porous
iron or nickel substrates which possess coeffi- and possess a low thermal expansion coefficient.
cients of thermal expansion of similar magni-
tude. This advantage of forsterite is, however, Material Types. Group C-400 defined in
obtained at the expense of low thermal shock IEC 672 contains two types of material:
resistance.
C-410 cordierite dense
C-420 celsian dense (BaAl2 Si2 O8 )
Zircon-Containing Steatite. Zircon
[1490-68-2] , ZrSiO4 , can be sinteredto produce
a dense material only with difficulty. When used Type C-410 materials are used more widely.
as an additivefor steatite, and, in some cases, for Group C-500 contains the following porous
porcelain, the thermal expansion of the result- materials:
ing material is reduced, and hence the thermal
shock resistance is improved. The amount of
C-510 materials based on grog
zircon additive can reach 60 wt%. The dielectric (fireclay, chamotte)
properties of the zircon steatites are adversely C-511 materials based on grog
affected because losses increase significantly containing magnesium oxide
C-512 materials based on grog
above 1 MHz. The relative permittivity of zir- containing magnesium oxide
con steatite is 8, this is sometimes undesirably C-520 materials of high cordierite
content
high for certain insulators. C-530 materials of high alumina
content
Insulation, Electric 19
Types C-511 and C-512 contain cordierite of the scarcity of the raw materials, and because
[12182-53-5] , Mg2 Al4 Si5 O18 , and mullite of the extremely narrow firing range. Therefore
[55964-99-3] , Al6 Si2 O13 . They differ in the their excellent thermal shock resistance is only
number andsize of the pores, and hence also in used in exceptional cases.
the strength and thermal shock resistance. The
composition of type C-520 materials approaches
that of cordierite. The high alumina content of
3.1.2.4. Alumina Ceramics (→ Ceramics,
the type C-530 materials enables high applica-
Electronic)
tion temperatures up to 1300 ◦ C.
The inorganic insulating materials discussed in
Properties. Table 3 contains typical proper-
Section 3.1.1 are all produced from natural mate-
ties of porous and dense materials used in elec-
rials. Aluminas and the other materials discussed
trical heat engineering.
in this section are based on synthetic materials
produced by the chemical industry.
Materials Containing Cordierite. Even as
early as the 1930s, addition of talc to clay-based
material had been shown to reduce their thermal Material Types. In IEC 672, alumina mate-
expansion coefficient, and hence increase their rials are classified in group C-700. They differ
thermal shock resistance. Talc and clay react dur- in their minimum alumina contents. In all cases,
ing firing to give cordierite. Optimum results are the alumina content is so high that the insulat-
obtained with approximately 60 % clay and grog ing material contains α-aluminum oxide as the
(chamotte) and 30 % talc (type C-520 material predominant or even exclusive crystal phase.
according to IEC 672). Types C-511 and C-512
contain ca. 10 % talc.
A distinct disadvantage of cordierite materi- Properties. The important physical proper-
als is their narrow firing range. Consequently it ties of alumina ceramics are summarized in Ta-
is normal to refrain from vitrifying the articles, ble 4. Properties generally improve with increas-
and concentrate on porous products that have ing alumina content (higher number).
a lower Young’s modulus and therefore exhibit Alumina ceramics combine a range of re-
improved thermal shock resistance. If dense ma- markable properties. With the high percent-
terials of type C-410 are required for special age types (C-799), a strength of ≥ 500 N/mm2
applications, then small amounts of a fluxing can be achieved, particularly if crystal growth
agent must be added; feldspar has proved par- can be suppressed during sintering. Mechani-
ticularly effective in this respect. Such materi- cal strength is maintained up to temperatures of
als are not only inferior as regards their thermal 1000 ◦ C. Thermal conductivity reaches a level
shock resistance, but also in their electrical in- equivalent to that of metallic lead, especially in
sulation characteristics. This is because of the the high percent types. Alumina substrates are
contribution of the alkali ions to their electri- amongst the best electrical insulating materials
cal conductivity at high temperatures. Further- under all application conditions.
more, the porous materials allow higher applica-
tion temperatures (ca. 1200 ◦ C) than their dense Production. Raw materials include calcined
counterparts. However, dense cordierite materi- aluminas which are obtained from natural baux-
als are required for some applications. ites by the Bayer process (→ Aluminum Ox-
ide). The precipitation and calcination condi-
Lithium Aluminum Silicates. Lithium alu- tions govern the most important properties in
minum silicates possess an even lower thermal ceramic applications, such as crystal size and
expansion coefficient than cordierite. sintering capability. Since the aluminas are not
The materials have not been standardized. plastic, molding and shaping necessitates the ad-
Raw materials are petalite, LiAlSi4 O10 , and to a dition of organic plasticizers which must be re-
lesser extent spodumen, LiAlSi2 O6 , or eucryp- moved prior to firing.
tite, LiAlSiO4 . Widespread application of these
materials has been somewhat hindered because
20 Insulation, Electric
Table 3. Important properties of materials classified in groups C-400 and C-500 of IEC 672
Table 4. Important properties of alumina materials classified in group C-700 according to IEC 672
raw material deposits, means that beryllia com- The materials differ with respect to their rela-
ponents are extremely expensive. Their use is tive permittivity, the associated temperature co-
restricted to a small number of applications that efficient, and the dielectric dissipation factor.
involve particularly high demands, e.g., in the Types C-310, C-320, C-330, C-331, and C-340
aerospace industry. have been classified according to international
regulations as Type I capacitor materials; ac-
Magnesia Ceramics. Magnesia is standard- cording to DIN 41 920, they are subdivided on
ized in IEC 672 as type C-820, a porous ce- the basis of the temperature coefficients of their
ramic insulating material. Its main applications relative permittivities. Material types C-350 and
are found within the field of electrical heat en- C-351 have a high relative permittivity and are
gineering, primarily because of its high specific classified as Type II substances.
heat for use in heat storage elements, but also be-
cause of its favorable thermal conductivity and Titanium Dioxide and Mixed Dielec-
excellent insulation properties for use as an in- trica Based on Rutile. Titanium dioxide
sulator in tubular heating elements. In the latter [13463-67-7] crystallizes in the rutilestructure,
case, the ceramic is inserted into the cavity be- which is always formed at temperatures above
tween the heating wire and the outer tubes and 1000 ◦ C evenwhen the metastable anatase is
subsequently crushed, compacted, and solidified used as a starting material.
by rolling, hammering, and pressing. The relative permittivity of rutile is strongly
anisotropic. In polycrystalline materials with
Zirconia Ceramics. Zirconia ceramics are random crystal orientation, the relative permit-
standardized as type C-830 in IEC 672. They tivity has an average value of 117. The temper-
can only be used as insulators at room temper- ature coefficient of the relative permittivity is
ature because the oxygen ions of zirconia be- highly negative (− 0.001 K−1 ).
come highly mobile at higher temperatures and Properties (see Table 5) can be modified by
the material then acts as a semiconductor. This additives, such as zirconium oxide, lanthanum
principle forms the basis of the widespread ap- oxide, or magnesium titanate. These additives
plication of zirconia as a sensor for determining possess considerably lower relative permittiv-
the partial pressure of oxygen. ities and positive temperature coefficients. By
following a logarithmic mixing relationship for
relative permittivity, or a linear mixing rela-
3.1.2.6. Insulating Materials Containing tionship for the temperature coefficient, specific
Titanium Oxide and Titanates materials with tailored properties can be ob-
tained. The temperature coefficient and the rela-
Whereas the materials discussed so far are gen- tive permittivity decrease with increasing addi-
erally employed purely as insulators, the use of tive concentrations. Dissipation factors are low
components containing titanium dioxide and ti- (≤10−3 ); with mixed dielectrics containing zir-
tanates is governed by their high relative permit- conium oxide, they are independent of frequency
tivity. up to 108 Hz.
Dielectrics of this type are used to compen-
Material Types. Group C-300 of IEC 672 is sate for the increase in capacitance with temper-
subdivided into seven types: ature produced by other components. The pri-
mary criterion for the selection of the dielectrics
C-310 titanium dioxide is the magnitude of the negative temperature co-
C-320 magnesium titanate
C-330 titanium dioxide incorporating efficient of the relative permittivity rather than
other oxides as additives the permittivity itself.
C-331 as for type C-330
C-340 strontium, calcium, and bismuth
titanates Materials Containing Titanates. The
C-350 barium titanate properties of titanate-based insulating materi-
C-351 barium titanate
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
Table 6. Important physical properties of titanate-based dielectrica classified according to IEC 672
feature of these materials is the excellent tem- symmetry is its extremely large relative permit-
perature stability of the relative permittivity, the tivity, particularly in the direction perpendicular
values of the latter lie between 12 and 40. to the main crystal axis (Fig. 11). Maximum rel-
Type I capacitors with extremely high nega- ative permittivity values are observed at each of
tive temperature coefficients for the relative per- the crystallographic transition temperatures, but
mittivity (up to 0.006 K−1 ) can be obtained with are accompanied by a maximum in the dielectric
materials based on strontium and/or calcium ti- loss (Fig. 12). This results in a strong, nonlinear
tanate (type C-340 in IEC 672). temperature dependence of the relative permit-
Barium titanate [12047-27-7] , BaTiO3 , is tivity. The dipoles in the deformed crystal inter-
the most important alkaline-earth titanate act mutually and are aligned in parallel in small
(→ Ceramics, Electronic). It also finds applica- areas known as domains. The direction of dipole
tion as a piezoelectric material. orientation in neighboring domains is reversed
Because of their piezoelectric properties, so that the whole system assumes an energy min-
considerable interest has also been centered on imum and does not have a net external charge.
titanates that contain lead in the form of mixed On applying an external electrical field, the do-
crystals with zirconates, stannates, and niobates. mains become oriented in the direction of the
Like all of the other titanates listed above, field to an extent proportional to the electrical
barium titanate crystallizes to give a deformed field strength. Because of the anisotropy of the
perovskite structure. Slight distortions lead to dielectric properties, this reorientation of the do-
a tetrahedal structure at room temperature, a mains is coupled with a change in the relative
rhombic structure below 0 ◦ C, and trigonal sym- permittivity, i.e., the relative permittivity is also
metry below −90◦ C. The most significant prop- dependent upon field strength. The combination
erty of the distorted structure without a center of of phenomena discussed here is known as fer-
Insulation, Electric 23
roelectricity (see also → Ceramics, Electronic, Certain oxide additives (e.g., a small ex-
Chap. 3.). cess of titanium dioxide) act as grain boundary
Above a certain temperature known as the phases which prevent crystal growth. In small
Curie temperature (in the case of barium titanate crystals, ferroelectric domains are still formed,
120 ◦ C), the structure becomes strictly cubic, but their changes of position and size in an ex-
and the ferroelectric properties are lost. Above ternal field are considerably hindered. Materials
the Curie temperature T c , the changes in the rel- of this kind show reduced hysteresis loss, and
ative permittivity with temperature are described the curve of εT = f (T ) flattens out in the vicinity
by the Curie – Weiss law: of the Curie temperature. Because of their low
C losses, such dielectrics are employed as contact-
εT = protective capacitors. Extremely small crystals
T − Tc
(ca. 50 nm) cannot be reorientated in an external
where C is the Curie constant and εT is the rel-
electric field and hence do not exhibit ferroelec-
ative permittivity at temperature T . Compounds
tric properties.
exhibiting a perovskite structure tend to form
Other additives (e.g., bismuth oxide, bismuth
mixed crystals. Hence Ba2+ can be replaced
titanate, zinc oxide, or zinc titanate) extend the
by Sr2+ , Pb2+ , and to a certain extent also
firing interval, but also change the shape of the
by Ca2+ ; Ti4+ can be substituted by Zr4+ or
curve εT = f (T ). Bismuth titanate additives pro-
Sn4+ . Similarly, ion combinations can be incor-
duce dielectric materials of low temperature de-
porated providing the ionic radii are equivalent
pendency in their electrical properties.
and a balance of charge is attained (complex per-
If a dielectric is made from two perovskite
ovskites). These substitutions produce changes
materials that do not form or only partially form
in the relative permittivity and its temperature
mixed crystals, then the change in the relative
dependency, as well as in the dielectric loss and
permittivity between the relevant Curie tempera-
the Curie temperature. In this way, a wide range
tures is low. Examples of such dielectrics include
of insulating materials with differing properties
those containing calcium and barium titanate.
can be produced.
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-
Figure 11. Relative permittivity of barium titanate along the electrics is as capacitors in high-frequency ap-
a and c crystal axes as a function of temperature
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-
Figure 12. Temperature dependence of the relative permit- pacitors which possess particularly high capac-
tivity (εr ) and the dielectric loss (tan δ) of polycrystalline
barium titanate
24 Insulation, Electric
ities due to the low thickness of the dielectric sistance. This characteristic is of considerable
layer (Type III dielectrics). importance in glow plugs in diesel engines.
More recently, increased interest has been shown Glass materials have been used as insulators in
in nonoxide inorganic materials because of sev- electrical engineering for many years. The sys-
eral unusual properties. They have a very high tematic classification and establishment of min-
mechanical strength, even at elevated temper- imum requirements occurred during their incor-
ature. The most important materials are com- poration in the IEC standard 672.
pounds of elements of groups 3, 4, and 5 of the
periodic table. Silicon nitride (Si3 N4 ), silicon Material Types. Glass materials used as
carbide (SiC), and aluminum nitride (AlN) de- electrical insulators can be classified into six
serve special mention. groups:
The essentially homopolar bonding charac-
G-100 alkali – lime – silica glasses
ter between the atoms of these compounds de- G-200 chemically resistant borosilicate
termines their physical properties to a large glasses
extent, but also causes problems during the G-300 electrically resistant borosilicate
glasses
manufacture of components when using tradi- G-400 alumina – lime – silica glasses
tional procedures employed in the ceramic in- G-500 lead oxide – alkali – silica glasses
G-600 barium oxide – alkali – silica
dustry. Production is thus more complicated and glasses
more expensive. Nonoxide ceramic materials
are only suitable as insulating materials when
their extraordinary properties provide a particu- Group G-100 consists of:
larly elegant solution for a given problem. Sil-
icon nitride [12033-89-5] andaluminum nitride Type G-110 alkali – lime – silica glasses, an-
[24304-00-5] are used as electrical insulators. nealed
Silicon carbide is not suitable for electrical Type G-120 alkali – lime – silica glasses, tough-
insulation purposes because its electrical con- ened
ductivity increases strongly with temperature; it The mechanical strength is increased tremen-
is even employed as a high-temperature heating dously by prestressing (Type G-120).
element. Group G-300 consist of:
The thermal conductivity of aluminum ni-
tride is considerably greater than that of alu- Type G-310 borosilicate glasses, low loss
minum oxide. The low density of 3.2 g/cm3 Type G-320 borosilicate glasses, stable to high
makes it a suitable alternative to beryllium oxide voltage
when the effective removal of heat generated by
dielectric losses is important. At increasing tem- Properties. Several of the requirements
peratures, the decrease in thermal conductivity specified in the IEC 672 are summarized in Ta-
of aluminum nitride is considerably lower than ble 7. A range of properties of glass are governed
that of the oxide materials. A further advantage by its chemical composition. Characteristic data
of aluminum nitride, especially as a substrate for can be determined by multiplying the content of
high-power silicon components, is the fact that a glass component (expressed in percent) by a
the thermal expansion coefficient is closer to that property-specific factor, and summing the cal-
of silicon than any other potential substrate ma- culated values. Properties that can determined
terials. in this way include the thermal expansion and
Silicon nitride exhibits similar insulation the surface energy.
properties to aluminum oxide, but its thermal
expansion coefficient is only half as large, and Production. Glass is produced by melting
correspondingly shows greater thermal shock re- 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- and correspondingly an improved thermal shock
ponent. Residual stress can be removed by an- resistance (→ Glass Ceramics).
nealing. However, compressive stresses can also Glass materials of type G-200 (chemically re-
be deliberately generated in order to increase the sistant borosilicate glass) are used in discharge
mechanical strength of the component. lamps. The low coefficient of thermal expansion
Insulator components are mainly shaped by resulting from the use of boron oxide ensures
pressing the viscous melt in steel molds, al- that the glass has a high thermal shock resis-
though recently, extrusion has been introduced tance. Boron also improves chemical resistance.
as an additional shaping method. Glass can also Type G-310 glass materials exhibit favorable
be processed by blowing. properties for lead through insulators in high-
frequency applications. They are also suitable
Uses. The applications for glass in electrical for glass components designed to have capacitor
insulator are determined by properties that can functions. Type G-320 materials are employed
be particularly well optimized in certain types in valves for high-voltage applications, e.g., for
of glass; these properties are indicated by bold X-ray tubes.
type in Table 7. Type G-400 glass materials were devel-
Glass materials from group G-100 are fre- oped to have optimum properties for high-
quently employed for overhead line insulators. performance halogen lamps; their properties
Cap insulators are most common and are com- have now been standardized.
bined to form long chains. Long rod insulators Type G-500 and G-600 glasses are employed
are also used but are more difficult to manufac- when metallic components, particularly those
ture. based on nickel – iron alloys, have to be sealed
Glass insulators have been increasingly em- into electric lamps and valves. They have a low
ployed with overhead lines. Their major advan- glass transition temperature and a large thermal
tage is that faults can easily be identified because expansion coefficient to ensure that only weak
they are transparent; slightly defective compo- mechanical stresses are generated in the com-
nents can therefore be exchanged before a major posite system. Because of the incorporation of
failure occurs. heavy metal oxides, they are also used in tele-
The high requirements placed on the mechan- visions to absorb the generated X-rays. Glass
ical properties of suspension insulators for over- materials containing lead can change on expo-
head lines are met by toughened glass. Glass sure to X-rays and electron beams, and thus be-
ceramics (which have not yet been included in come discolored or even produce trackìng cur-
the IEC standard) have also been recently em- rents. In applications where these effects must be
ployed because they have a high strength com- excluded, lead-free glasses of type G-600 must
bined with a low thermal expansion coefficient be employed.
26 Insulation, Electric
3.2. Organic Insulating Materials The significance and determination of the pa-
rameters presented in the graphics are described
The following discussion of the most common in Chapters 1, 2, and 4. Values can easily be read
organic materials employed as electrical insula- from the plots. The dielectric dissipation factor
tors includes products that meet all or many of is obtained by dividing the loss index (column
the imposed requirements. This does not mean 4) by the corresponding value for relative per-
that all of the products are employed in large mittivity (column 5).
amounts in the electronics industry. The com-
pilation is by no means complete, and is in-
tended merely to provide an overview. Descrip- 3.2.1. Thermoplasts
tion of the products has been kept brief. Details
are available from manufacturers’ information Thermoplasts, also referred to as plastomers, are
and specialist publications. Examples of trade solid synthetic resins which can be repeatedly
names are given for each group of polymers but softened without chemical modification by heat-
these lists are by no means exhaustive. Most of ing in a relatively narrow temperature range.
the polymers are described in detail elsewhere They can be shaped while soft and generally
in the encyclopedia. Emphasis has been placed consist of linear molecules. Physical properties
on graphic representations–the relief diagrams. of the thermoplasts are compiled in Table 8.
As opposed to tables, these diagrams allow the
rapid visual assessment of the suitability of a ma- Acrylonitrile – Butadiene – Styrene. Phys-
terial for a given electrotechnical project. Differ- ical properties are given in Table 8 and electrical
ent materials can also be compared. The block properties in Figure 13 (see also → Polystyrene
lengths demonstrate the quality variation of a and Styrene Copolymers).
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- Figure 13. Relief diagram of electrical properties of acry-
tion 2.6) lonitrile – butadiene – styrene
5) and the relative permittivity εr measured ac-
cording to IEC 250 at 1 kHz, 23 ◦ C, and 50 % Special properties: high impact toughness,
R.H. (see Section 2.4) thermally stable, unlimited shaping character-
istics, limited stability to UV radiation.
The electric strength is not shown because Uses : casings for office machinery, electrical
although the test results obtained in accordance industry, electronics, telecommunications, pro-
with the standard can be used for detecting tective insulating components.
changes or deviations from normal characteris- Trade Names: Cycolac (Borg Warner Chemi-
tics resulting from processing variables, aging cals, USA); Kralastic (USS Chemicals, USA);
conditions, or other manufacturing or environ- Novodur (Bayer, FRG); Terluran (BASF, FRG);
mental situations, they can seldom be used di- Diapet (Mitsubishi Rayon Co., Japan); Ravikral
rectly to determine the behavior of insulating (EniChem, Italy).
materials in a real application. Measured val-
ues of the electric strength are affected by many Acrylonitrile – Styrene – Acrylic Ester.
factors (Section 2.7), but these factors are not Physical properties are given in Table 8 and
specified when the results are published. 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
Figure 16. Relief diagram of electrical properties of cellu- Figure 17. Relief diagram of electrical properties of cellu-
lose acetate butyrate lose propionate
Special properties: antistatic properties de- Trade Names: Ampol (American Polymers,
spite high electrical resistance, crystal clear, USA); Tenite (Eastman Chemical International,
tough, hard, scratch-resistant, insensitive to USA).
stress cracking, readily dyeable with brilliant
colors, not permanently weather-resistant, less Perfluoro(Ethylene – Propylene). Phys-
suitable for precision components. ical properties are given in Table 8 and
Uses: tool handles, electric insulation films, electrical properties in Figure 18 (see also
lights, casings. → Fluoropolymers, Organic, Chap. 2.3.).
Trade Names: Dexel (Courtaulds Plastic Group, Special properties: nonflammable, weather-
UK); Ampol (American Polymers, USA); Ten- resistant, resistant to solvents, no water uptake,
ite (Eastman Chemical International, USA). physiologically inert, stable to pressure.
Uses: flexible printed circuits, coaxial ca-
Cellulose Acetate Butyrate. Physical prop- bles for computers, cable insulation, shrinkable
erties are given in Table 8 and electrical proper- tubes.
ties in Figure 16 (see also → Cellulose Esters). Trade Names: Teflon (Du Pont, USA); Neoflon
Special properties: antistatic despite high (Daikin Kogyo Co., Japan).
electrical resistance, crystal clear, tough, hard,
scratch resistant, insensitive to stress-cracking, Polyamide. Physical properties are given in
readily and brilliantly dyeable, not suitable for Table 8 and electrical properties in Figure 19
precision components. (see also → Polyamides).
Uses: tool handles, electric insulation films, Special properties: normal minimum water
lights, casings. content of 2.5 %, prone to stress cracking when
Trade Names: Ampol (American Polymers, freshly injected, impact resistant, viscoelastic,
USA); Tenite (Eastman Chemical International, puncture resistant, dynamic-loading capability,
USA). good slip properties.
Uses: high-voltage insulation, multiple con-
necting plugs, relays, extinguishing chambers,
Cellulose Propionate. Physical properties
wear-resistant insulation and sleeve materials
are given in Table 8 and electrical properties in
for cable technology.
Figure 17 (see also → Cellulose Esters).
Trade Names: Ultramid (BASF, FRG);
Special properties: antistatic despite high
Durethan (Bayer, FRG); Technyl (Rhône-
electrical resistance, crystal clear, tough,
Poulenc Specialities Chimiques, France); Grilon
scratch-resistant, insensitive to stress cracking,
(Ems-Chemie, Switzerland); Zytel (Du Pont,
readily dyeable with brilliant colors, not suitable
USA); Maranyl (ICI, UK).
for precision components.
Uses: tool handles, electric insulation films, Polyamide 12-Resin. Physical properties
lights, casings. are given in Table 8 and electrical properties
in Figure 20.
Insulation, Electric 29
Special properties: resistant to organic sol- ditions, printed circuits, suitable for solderable
vents, fuels, oils, weak acids and alkalis. connecting cable.
Uses: casings, components for electrical de-
vices. Polybutene. Physical properties are given in
Trade Names: Vestamid (Hüls, FRG); Ril- Table 8 and electrical properties in Figure 22
san (Atochem, France); Grilamid (Ems-Chemie, (see also → Polyolefins).
Switzerland).
Special properties: high long-term stability, suitability under abrasive conditions, electrical
low tracking propensity at high temperature, properties practically independent of humidity.
high stress cracking stability, resistant to low Uses: moldings, casings, appliances, relay
temperature, easily weldable. elements, condenser housings, plug connec-
Uses: insulation jackets for wires, cable tions, high-voltage insulation, cable distribu-
sleeves, pipes. tion panels, mounting boards, screens, compact
Trade Names: Duraflex (Shell International discs, information storage media.
Chemical Co., UK); Krene (Union Carbide, Trade Names: Makrolon (Bayer, FRG); Orgal-
USA). lan (Atochem, France); Merlon (Mobay Chem-
ical Corp., USA); Lexan (General Electric,
Poly(Butylene Terephthalate) (PBT). USA); Jupilon (Mitsubishi Gas Chem, Japan);
Physical properties are given in Table 8 and Sinvet (EniChem, Italy).
electrical properties in Figure 23 (see also
→ Polyesters). Polycarbonate Alloy (PC – ABS). Physical
Special properties: impact toughness, rigid, properties are given in Table 8 and electrical
maintains shape, scratch-resistant, good slip properties in Figure 25.
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).
Figure 25. Relief diagram of electrical properties of poly-
Polycarbonate (PC). Physical properties carbonate alloy (PC – ABS)
are given in Table 8 and electrical properties
in Figure 24 (see also → Polycarbonates). Special properties: resistant to combustion,
Special properties: low combustion, rigid rigid elastic, tough above 90 ◦ C, long term sta-
elastic, high impact toughness above 90 ◦ C, bility at high temperatures, limited suitability
long term stability at high temperature, limited under abrasive conditions, electrical character-
istics practically independent of humidity.
Insulation, Electric 31
Uses: moldings, casings, appliances, relay Alathone (Du Pont, USA); Novatec (Mit-
elements, condensor housings, plug connec- subishi Chemical Industries, Japan); Dylan
tions, high-voltage insulation, cable distribution (Arco Chemical Co., USA).
panels, mounting boards, screens.
Trade Names: Cycolac (Borg Warner Chemi- Cross-linked Polyethylene. Physical prop-
cals, USA); Bayblend (Bayer, FRG); Cycoloy erties are given in Table 8 and electrical proper-
(Borg Warner Chemicals, USA). ties in Figure 28.
Polyethylene. Physical properties are given Special properties: stable at high tempera-
in Table 8 and electrical properties in Figure 27 ture, high impact strength at low temperature,
(see also → Polyolefins). resistant to stress cracking, greater resistance to
Special properties: resistant to water, saline elevated temperatures than polyethylene.
solutions, alkalis, acids, (exception: strong oxi- Uses: insulation and sleeve materials for ca-
dizing agents), can be subjected to high mechan- bles, shrinkable sleeves.
ical stress, hard, resistant to low temperatures. Trade Names: Lubonyl (N. Lundbergs Fab-
Uses: insulation and sleeve materials for ca- rics AB, Sweden); Alveolit (Sekisui Chemical
bles, pipes, with 30 % carbon black deflects lines Co., Japan); Alveolux (Alveo AG, Switzerland);
of force in conductors, transformers, etc. Blanex (Reichhold Chemicals Inc., USA); Ela-
Trade Names: Lupolen (BASF, FRG); Car- por (Elastogran, FRG).
lona (Shell International Chemical Co., UK);
32 Insulation, Electric
Figure 32. Relief diagram of electrical properties of poly- Figure 33. Relief diagram of electrical properties of
imide poly(methyl methacrylate)
Trade Names: Kapton, Vespel (Du Pont, USA); Poly(4-methyl-1-pentene) (PMP). Physi-
Kinel (Rhône-Poulenc Specialities Chimiques, cal properties are given in Table 8 and electrical
France). properties in Figure 34 (see also → Polyolefins).
Special properties: light weight, crystal
Poly(Methyl Methacrylate). Physical clear, rigid, not suitable for exterior applica-
properties are given in Table 8 and elec- tions, good thermal stability.
trical properties in Figure 33 (see also Uses: insulation for coaxial and electrical ca-
→ Polymethacrylates). bles.
Special properties: hard, tough, highly trans- Trade Names: TPX (Mitsubishi Petrochemical
parent, weather-resistant, chemically inert to Co., Japan).
medium concentrations of acids, resistant to fu-
els, physiologically inert. Polyoxymethylene (POM). Physical prop-
Uses: lamps, traffic lights, nautical aids, opti- erties are given in Table 8 and elec-
cal fibers, high-pressure mercury lamps, optical trical properties in Figure 35 (see also
data storage media. → Polyoxymethylenes).
Trade Names: Lucite (Du Pont, USA); Plexiglas Special properties: high strength and stiff-
(Röhm, FRG); Altulite (Altulor Groupe CdF ness, high dynamic loading capacity over a wide
Chimie, France); Sumipex (Sumitomo Chemi- temperature range, high impact toughness even
cal Co., Japan). 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- Trade Names: Luranyl (BASF, FRG); Noryl
dependent of frequency and temperature. (General Electric, USA); Vestoran (Hüls, FRG).
Uses: precision components for measuring
instruments in electronics and precision engi- Poly(Phenylene Sulfide). Physical proper-
neering, household appliances. ties are given in Table 8 and electrical properties
Trade Names: Delrin (Du Pont, USA); Ultra- in Figure 38.
form (BASF, FRG); Tenac (Asahi Chemical In-
dustries, Japan).
Figure 39. Relief diagram of electrical properties of Figure 40. Relief diagram of electrical properties of poly-
polystyrene sulfone
Special properties: crystal clear; high gloss; Special properties: transparent yellow; high
brittle; resistant to humidity, salt solutions, alkali mechanical loading capacity; dielectric proper-
and nonoxidizing acids; mechanical properties ties only slightly temperature- and frequency-
only slightly time- and temperature-dependent; dependent, resistant to acids, bases, oxidation
dielectric properties are independent of temper- and hydrolysis; highly incombustible.
ature and frequency; susceptible to damage by Uses: insulation films, circuit boards.
UV light. Trade Name: Ultrason (BASF, FRG).
Uses: encapsulation of electronic compo-
nents, insulating films, dielectric for capacitors. Polytetrafluoroethylene. Physical proper-
Trade Names: Styroflex (Norddeutsche Seek- ties are given in Table 8 and electrical proper-
abelwerke AG, FRG); Vestyron (Hüls, FRG); ties in Figure 41 (see also → Fluoropolymers,
Sternite (Chemical Products, UK); Dylene Organic, Chap. 2.2.).
(Arco Chemical Co., USA); Esbrite (Sumit- Special properties: difficult to wet; antiadhe-
omo, Japan); Toporex (Mitsui Toatsu Chemi- sive; favorable coefficient of friction; noncom-
cals, Japan). bustible; weather-resistant; physiologically in-
ert; extremely resistant to corrosion, solvents,
Polysulfone. Physical properties are given in acids, and bases; dielectric properties indepen-
Table 8 and electrical properties in Figure 40 (see dent of temperature and frequency.
also → Polymers, High-Temperature).
36 Insulation, Electric
Uses: films, insulation, insulating and sleeve properties in Figure 44 (see also → Polystyrene
material in cables. and Styrene Copolymers).
Trade Names: Hostaflon TF (Hoechst, FRG);
Teflon (Du Pont, USA); Polyflon (Daikin Ko-
gyo Co., Japan); Fluon (ICI, UK).
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-
Figure 43. Relief diagram of electrical properties of position temperature. Thermosets can be mod-
styrene – acrylonitrile copolymer
ified with fillers. Physical properties are com-
Special properties: resistant to UV light, piled in Table 9.
weather-resistant, good long-term stability at el-
evated temperature, good formability. Epoxy Resins. Physical properties are given
Uses: insulators, insulation material, ventila- in Table 9 and electrical properties in Figure 45
tors, casings for electrical equipment, electrical, (see also → Epoxy Resins).
electronic, telecommunications, protectively in- Special properties: strong adhesion to met-
sulated components. als, good flow properties, no material losses,
Trade Names: Acrylafil (Fiberfil Inc., USA); thermally stable, light colors.
Thermocomp (LNP Corp., USA); Luran (BASF, Uses: jacketing, contact substrates in electri-
FRG); Lustran (Monsanto Co., USA); Litac cal applications, insulation of converters, sub-
(Mitsui Toatsu Chemicals Inc., Japan). strate material for motor armatures, collector
rings in electrical motors and generators.
Trade Names: Beckopox (Hoechst, FRG);
Styrene – Butadiene Copolymer. Physical
Araldit (Ciba-Geigy, Switzerland); Epodite
properties are given in Table 8 and electrical
(Showa High Polymer Co., Japan); Epocast (M
Insulation, Electric 37
Table 9. Physical properties of thermosets
dustries Sweden, Sweden); Dynapor (Hüls- castable; curing takes place even at low temper-
Troisdorf, FRG). ature.
Uses: cable end caps, converters, transform-
Synthetic Resin – Chip Board. Physical ers, ignition coils.
properties are given in Table 9 and electrical Trade Names: Elastan (Elastogran, FRG);
properties in Figure 48. Desmodur (Bayer, FRG); Suprasec
(Polyurethanes Group ICI, UK).
Figure 52. Relief diagram of electrical properties of unsat- Special properties: highly incombustible,
urated polyester leatherlike toughness, weather-resistant,
corrosion-resistant.
Uses: insulating and sleeve material for ca-
Special properties: meltable, can be cast bles.
when liquid, no material losses, dimensionally Trade Names: Baypren (Bayer, FRG); Buta-
stable, not prone to stress cracking, stable to clor (Rhône-Poulenc Specialities Chimiques,
light, easy to process. France); Neoprene (Du Pont, USA).
Uses: mounting electrical components, also
for high-voltage applications. Ethylene – Propylene Terpolymer. Physi-
Trade Names: Palatal (BASF, FRG); Vestopal cal properties are given in Table 10 and electrical
(Hüls, FRG); Alpolit (Hoechst, FRG); Crystic properties in Figure 54.
(Scott Bader Co., UK); Glastic (Glastic Corp., Special properties: rubber-elastic, abrasion
USA); U-Pica (Japan-Upica, Japan). resistant, stable towards aging and weathering
(oxygen, ozone, water, UV radiation, high tem-
perature), inert towards many chemicals.
3.2.3. Elastomers Uses: insulation and sleeve material for ca-
bles, plugs, cable filling, with carbon black
Elastomers are natural or synthetic polymers causes deflection of the lines of force in con-
(→ Rubber, 2. Natural; → Rubber 3., Synthetic) ductors, transformers, etc.
which exhibit rubberlike behavior between at
least 20 ◦ C and their respective decomposition
temperature. They solidify at or below 0 ◦ C. The
40 Insulation, Electric
Table 10. Physical properties of elastomers
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 Uses: insulating and sleeve material for ca-
(Exxon Chemical Co., USA). bles.
Trade Names: Dutral TP (Montepolmeri, Italy);
Ethylene – Propylene Rubber. Physical Intolan (Dow Corning Corp., USA); Vistalon
properties are given in Table 10 and electrical (Exxon Chemical Co., USA).
properties in Figure 55.
Special properties: resistant to chemicals, Ethylene – Vinyl Acetate Rubber. Physical
ozone, aging, water, acids, and alkalis; weather properties are given in Table 10 and electrical
resistant; unstable to mineral oils and organic properties in Figure 56. Special properties: ther-
solvents.
Insulation, Electric 41
mally stable, excellent adhesion, rubber-elastic, operations result in products that frequently do
impact resistant. not require further physical modification. The
Uses: insulating and sleeve material for ca- forming procedures most applicable to single
bles. items or small numbers of insulators include re-
Trade Name: Levapren (Bayer, FRG). shaping and bonding, for example by heat seal-
ing. Thermoplastic materials can also be ma-
Styrene – Butadiene Rubber. Physical chined, including boring and cutting. A molded
properties are given in Table 10 and electrical polymeric insulator, fashioned to the proper
properties in Figure 57. shape in a single operation without recourse to
Special properties: resistant to organic sol- subsequent machining, may often replace con-
vents, weldable, limited stability fuels, oils, ventional insulators consisting of multiple parts,
acids, and alkalis. each requiring its own series of machining op-
Uses: insulating and sleeve material for ca- erations. However, the equipment required to
bles, capacitor components, accumulator boxes produce complicated polymeric objects are usu-
components for telecommunications, connec- ally quite expensive, so economic factors tend
tors, insulators. to limit this approach to mass-produced items.
Trade Name: Buna-Hüls (Hüls, FRG). For a more detailed description of the techniques
described below see → Plastics, Processing.
Silicone Rubber. Physical properties are
given in Table 10 and electrical properties in Fig- Casting without the Use of Pressure. Cer-
ure 58. 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).
Once polymeric materials have been synthesized Casting under Thermoplastic Conditions.
and, if necessary, treated with fillers, pigments, The thermoplastic casting method plays an im-
plasticizers, processing aids, and agents to mini- portant role in polymer processing. It requires a
mize aging and flammability, they are processed temperature in the range of 150 – 250 ◦ C and a
into powders or granulates. Subsequent forming pressure of several hundred bar.
42 Insulation, Electric
Calendering denotes the process of rolling Transfer Molding. In transfer molding the
a polymeric mass in the plastic state until polymer is warmed to a plastic state in a sep-
it reaches a desired thickness, usually 0.06 – arate compartment prior to being forced into a
0.6 mm. forming die.
subsequent electrical conduction warming of the insulators and coolants. In cables, capacitors,
conductor. Rapid heating to the curing tempera- and connections, insulation liquids are used as
ture induces such an impregnating agent finally a functional component of the dielectric. In re-
to harden. lays, they function as insulators, coolants, and
extinguishers for arcing. Most liquid insulators
Varnishing. Metallic electrical conductors consist of mineral oils or synthetic liquids.
can also be electrically insulated with wire var-
nish. This method is often employed in trans- Mineral oils are complex hydrocarbon mix-
formers, motors, generators, relay coils, and tures that are obtained from petroleum oil by
similar devices. The applied polymer (usually a fractional distillation and refining. They com-
polyesterimide, polyurethane, polyamideimide, bine important properties including high break-
or polyhydantoin) is dissolved in an appropri- down strength, good impregnation characteris-
ate solvent, e.g., aromatic and aliphatic hydro- tics, and the ability to conduct large amounts
carbons, chlorinated hydrocarbons, aqueous al- of heat via convection. Mineral oils can also
cohols. The conductor is passed several times bind any gas produced during arcing with pro-
through a varnish bath, and becomes uniformly nounced evaporation and decomposition. De-
coated. Solvent is evaporated in a radiation oven, pending upon the predominant hydrocarbon, the
permitting deposition and annealing of a multi- oils can be classified as naphthene- or paraffin-
layered coating graded according to hardness. based oils; however, the transition between these
Thermoplastic polycarbonate films can be ob- groups is continuous–the mixed petroleum oils.
tained from solution in a fashion similar to var- The naphthene-based oils have a higher density,
nish films, whereby the film is extracted from a and a lower solidification point than the paraffin
drum after evaporation of solvent. oils; their viscosity is also more temperature de-
A more environmentally acceptable coating pendent. Since they are also more stable at low
technique utilizes either a melt varnish of the temperatures, they are preferred for devices used
polyesterimide type or a water-soluble wire var- in the open air. They are also preferable for cable
nish. applications because paraffin oils are more prone
to decomposition at high field strengths with for-
Embedding. Solvent-free, soaking, trick- mation of hydrogen and wax products. Decom-
ling, and casting resins are used for embed- position can lead to partial discharges and gas in-
ding components in electrotechnics. They pro- clusions, which may interact resulting in break-
tect costly components against environmen- down of the cable insulation (see section 2.7).
tal effects, provide mechanical stabilization, Naphthene oils are also suitable for cable im-
and prevent the formation of dangerous con- pregnation systems comprising oils and refined
tacts. Reaction resins, monomers, or prepoly- resins (e.g., collophonium) because the refined
mers are generally used that harden in the resins are incompatible with paraffin oils [47].
presence of suitable reactants. Most important Insulation oils for transformers, converters,
are methyl methacrylate, unsaturated polyesters, and relays are subject to standard regulations,
epoxy resins, polyurethanes, modified pheno- e.g., IEC Publication 296, which only specifies
lic resins, and furan resins. The cross-linkable minimum values for the breakdown voltage [48].
hydrocarbons and methyl methacrylate casting Since the electric strength of oils depends
resins have favorable dielectric values at high strongly on the measuring technique used (Sec-
frequencies (Sections 2.6) and are therefore im- tion 2.7), breakdown behavior is character-
portant for microwave technology (television, ized by the breakdown voltage determined un-
radar, dielectric warming). der standardized conditions (specified electrode
configuration, electrode separation, and test pro-
cedure, IEC Publication 156 [32]).
3.3. Liquid Insulating Materials When insulation liquids simultaneously act
as a coolant, they must have a low viscosity. They
In transformers, starter motors, and other elec- should be able to withstand cooling to − 30 ◦ C
trical equipment that generate a considerable without thickening. Furthermore, the oxidation
amount of heat, liquids are employed both as and aging behavior of new oil and used oil should
44 Insulation, Electric
be monitored during operation when the temper- both chemically and thermally stable. They pos-
ature is elevated and atmospheric oxygen has sess dipole character; their dielectric dissipation
access to the equipment. factor is therefore frequency- and temperature-
The viscosity of cable insulation oils varies dependent and they have a high relative permit-
within wide limits, ranging from low viscosity tivity (εr ≈ 4 – 6 at 50 Hz).
oils for oil cables to highly viscous impregnation Liquid fluorine compounds show excellent
systems for earthing and pressure cables. Fur- properties. They are noncombustible, prevent
thermore, extremely viscous cable immersion arcing, and are more thermally stable than As-
systems can be prepared by addition of thick- carels. They have a lower boiling point and are
eners (natural resins, polymers). highly volatile. Because of their low viscos-
The most important property of cable immer- ity, heat transfer properties are favorable. They
sion systems is the dielectric loss index (see have a high breakdown voltage, high volume
Section 2.6) which should be as low as possi- resistivity (Section 2.2), and an essentially fre-
ble because of the danger of thermal breakdown quencyindependent dielectric dissipation factor.
of the cable insulation at higher field strengths. Liquid silicones are polymeric organosilicon
The electrical strength of the impregnation sys- compounds with a highly variable degree of
tem is less important because the high electri- polymerization. Commercial silicone oils have
cal strength of the cable insulation is a result of differing viscosities, which are only slightly
the combination of the multilayer insulation film temperature-dependent. They combine excel-
(paper or polymeric) and impregnating agent; it lent resistance to low temperatures (setting point
depends upon the breakdown behavior of both −5 ◦ C to −80 ◦ C) with outstanding high temper-
components. ature stability (240 ◦ C) and are not prone to oxi-
The viscosity and electrical properties of dation. They remain noncombustible up to oper-
mineral oils used for high-tension capacitors ating temperatures of 300 ◦ C. Silicone oils have
(phase displacers) are similar to those of the ca- a high breakdown strength, low dielectric dis-
ble insulation oils. sipation factor (≤1×10−4 ), low relative permit-
tivity (2.5 – 2.9), and an average volume resistiv-
Synthetic Liquids. The drawbacks of min- ity (1012 Ωm). The low temperature dependence
eral insulation oils (oxidizability, flammability, of the dielectric dissipation factor (3×10−4 ) is
and aging with formation of detrimental decom- remarkable.
position products and deterioration of electrical Other Liquid Insulators. Apart from the As-
properties) led to the development of synthetic carels and liquid silicones, further synthetic in-
insulation liquids with improved properties. A sulation fluids find application mostly in trans-
general advantage of synthetic insulation liquids formers, capacitors, and cables. The most im-
is the fact that they can be reproducibly manufac- portant commercially available examples in-
tured with identical structures and compositions clude compounds based on alkyl benzenes,
[49]. They include chlorinated hydrocarbons, polybutene, diarylalkanes, esters, alkyl diphenyl
the most common being aromatic compounds ethers, and high molecular mass paraffin hy-
such as chlorinated biphenols, chlorinated ben- drocarbons. These liquids match the excellent
zenes, and their mixtures. These noncombustible electrical properties of the Ascarels but not their
materials are known as Ascarels. Damage to flame retardation. They differ according to appli-
health and the environment can result from care- cation regarding relative permittivity, dielectric
less handling and use of such toxic agents. Con- dissipation factor, ignition temperature, viscos-
sequently, sale of Ascarel-containing products ity, and the behavior toward low temperatures.
is prohibited in many countries. Closed systems
(transformers, capacitors, etc.) have been ex-
empted from these regulations. 4. Selection of an Insulating Material
During decomposition in electric arcs, As-
carels do not emit explosive or combustible gas Within its technical committee 15 for “Insu-
mixtures provided that they contain equivalent lating Materials”, the International Electrotech-
molar amounts of chlorine and hydrogen; hydro- nical Commission (IEC) has installed two
gen chloride is then produced. The Ascarels are subcommittees: subcommittee 15 A for “Short
Insulation, Electric 45
Term Tests” and subcommittee 15 B for “En- should be no greater than 1000 V/cm (see Sec-
durance Tests”. Electrical testing methods for tion 2.6), then a loss index of εr ≈ 3600 can
insulation materials are developed in the work- be tolerated. An increase in the field strength to
ing groups of these subcommittees. The rec- 10 000 V/cm causes a decrease in the threshold
ommendations of national electrotechnical com- value for the loss index by a factor of 100. In
missions serve as the basis for their develop- the same way the loss index threshold value de-
ment. creases linearly with an increase in frequency.
The intention and objective of the IEC is that The magnitude of the relative permittivity
the international standards drawn up by numer- is particularly important in determining the in-
ous experts from the member states should be tended application of a material. To avoid di-
adopted as the national standards. All published electric warming, its value should be low. A
results should then be comparable. lower relative permittivity is also associated with
The values of electrical properties given in a higher breakdown voltage in air. (As a general
this publication have been extracted from nu- rule, the maximum attainable breakdown volt-
merous sources; some have been supplemented age is achieved when the relative permittivity of
and amended using the results of our own in- the material is identical with that of the surround-
vestigations. In the available literature, detailed ing medium). For applications in capacitors, the
information pertaining to testing conditions has relative permittivity should be as large as possi-
often been omitted, consequently the data plots ble to minimize the geometric dimensions.
provide nothing more than rough guides. This The electrical strength is primarily intended
can only be improved by testing identical sam- to allow comparison between materials. How-
ples of the huge range of insulating materials ever, this is only the case when all parameters,
used in the electrotechnics industry in one and including sample thickness, are absolutely iden-
the same laboratory, under identical testing con- tical. The significance of the electrical strength
ditions that strictly comply with the international is thus disputable (Section 2.7). It cannot be used
regulations. Although this undertaking would be directly for design purposes. Test models are im-
highly desirable, it is almost impossible. perative, studies on the effect of loading time are
To solve a specific insulation problem, mate- often required.
rials can be preselected from the relief diagrams. A “single value”, i.e., result at a discrete tem-
The manufacturers must then be consulted, or perature and frequency does not characterize the
reference to technical data sheets can be made. behavior of a material under application con-
Measurements or the results of an independent ditions. The influences of parameters such as
testing institute ensure that the material is able temperature, frequency, loading time, age, or
to meet the requirements. humidity cannot be discerned from this single
What makes an outstanding material for elec- piece of information. A knowledge of the in-
trical applications? The answer to this question terdependence of the various parameters pro-
depends upon the intended use, e.g., whether vides far more information than a single data
it involves low-voltage or high-tension sup- value. In many cases, the development of tech-
plies, d.c., or low-, mid-, or high-frequency nical polymers and their successful application
sources. For applications with low-voltage sup- in the electrical industry has only been possi-
plies, the tracking parameter should be at least ble because the complete characteristics of the
CTI 400, i.e., group II according to IEC Publi- material had been previously compiled and eval-
cation 664 [24], group I includes materials with uated. The application range must be considered
CTI 600. Similarly, the volume resistivity should as a function of temperature, frequency, voltage,
be greater than 1012 Ω · m, the surface resistivity and environmental conditions. Other influences
at 23 ◦ C and 50 % R.H. should be at least 1014 Ω. including manufacturing processes, mechanical
At mains frequency, no particular demands loading, and the effects of fillers and additives
are placed on the dielectric dissipation factor or are better reflected by the functions of the electri-
the relative permittivity except for the applica- cal parameters rather than by individual results
tion as a dielectric in capacitors. If, for exam- at a single temperature or frequency. Automatic
ple, the power loss per cubic centimeter should measurement processes have made the practical
not exceed 0.1 W at 50 Hz and the field strength
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 Figure 60. Relative permittivity (A) and dielectric dissipa-
factor (Section 2.5) and the relative permittiv- tion factor (B) of a thermoplastic polymer blend as functions
ity (Section 2.4) as a function of temperature at of frequency at 23 ◦ C
frequencies of 50 Hz, 1 kHz, and 1 MHz. At tem- Figure 61 shows the plot of volume resistivity
peratures below the long-term thermal stability (Section 2.2) as a function of temperature. The
threshold (90 ◦ C) the curves remain relatively decrease in the volume resistivity by a factor
constant and are typical of a medium-quality of five powers of ten is considerable and illus-
material. In the vicinity of − 50 ◦ C, a relaxation trates that the number of mobile charge carriers
phenomenon occurs, which is indicated by the strongly increases with increasing temperature.
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.
12. Engineering Property Data on Selected 30. IEC Publication 250 – 1969, Bureau Central de
Ceramics: vol. I, Nitrides (1976); vol. II, la Commission Electrotechnique International,
Carbides (1979); vol. III, Single oxides Genève.
(1981); Metals and Ceramics Information 31. IEC Publication 243 – 1967, Bureau Central de
Center, MCIC Report HB-07, Battelle la Commission Electrotechnique International,
Columbus Laboratories, Columbus, Ohio. Genève.
13. W. Espe: Werkstoffkunde der 32. IEC Publication 156 – 1963, Bureau Central de
Hochvakuumtechnik, vol. II: la Commission Electrotechnique International,
Silikatwerkstoffe, VEB Dtsch. Vlg. d. Genève.
Wissensch., Berlin 1960. 33. IEC Publication 587 – 1984, Bureau Central de
14. P. Guillery: Werkstoffkunde für la Commission Electrotechnique International,
Elektroingenieure, Vieweg, Braunschweig Genève.
1971. 34. ANSI/ASTM-D 495, ASTM Philadelphia, Pa.,
15. T. Haase: Keramik, VEB Dtsch. Vlg. f. USA.
Grundstoffind., Leipzig 1970. 35. DIN VDE 0303, 1989.
16. A. Hecht: Elektrokeramik. Springer, Berlin – 36. VDE 0303, part 8, Beuth Verlag, Berlin 1975.
Heidelberg – New York 1976. 37. DIN VDE 0303, 1989.
17. R. Racho et al.: Werkstoffe der Elektrotechnik, 38. IEC Publication 426 – 1973. Bureau Central de
VEB Vlg. Technik, Berlin 1968. la Commission Electrotechnique International,
18. H.-W. Rotter: Glimmer und Genève.
Glimmererzeugnisse, Eigenschaften, 39. IEC Publication 167 – 1964, Bureau Central de
Entwicklungen, Anwendungen, Siemens AG, la Commission Electrotechnique International,
Berlin – München 1985. Genève.
19. E. Ryschkèwitch: Oxide Ceramics, Academic 40. IEC Publication 343 – 1970, Bureau Central de
Press, New York – London 1960. la Commission Electrotechnique International,
20. H. Stäger: Werkstoffkunde der Genève.
elektrotechnischen Isolierstoffe, Bornträger, 41. IEC Publication 672-2 – 1980, Part 1
Berlin 1955. Definition and Classification, Part 2 Methods
21. H. Salmang, H. Scholze: Die physikalischen of Test, Part 3 Individual Materials, Bureau
und chemischen Grundlagen der Keramik, Central de la Commission Electrotechnique
Springer, Berlin – Heidelberg – New York International, Genève.
1968. 42. K. H. Schüller: “Hochfeste Porzellane auf
Quarz- und Cristobalitbasis,” Ber. Dtsch.
Specific References Keram. Ges. 44 (1967) 212 – 223, 284 – 293,
22. A. Kamm, Kunststoffe 75 (1985) no. 12, 387 – 391.
900 – 903. 43. V. Frith, R. O. Heckroodt, K. H. Schüller:
23. IEC Publication 112 – 1979, Bureau Central de “Characterization of Ceramic High-Voltage
la Commission Electrotechnique International, Insulator Materials,” Ber. Dtsch. Keram. Ges.
Genève. 64 (1987) 216 – 219.
24. IEC Publication 664/664 A – 1981, Bureau 44. H. Kromer, D. Rose, K. H. Schüller:
Central de la Commission Electrotechnique “Investigations on Commercial Talcs and
International, Genève. Steatites, IV formation of Mineral Phase
25. ANSI/ASTM 3638 – 1977, ASTM During Firing,” Sprechsaal 121 (1988) no. 8,
Philadelphia, Pa., USA. 1 – 8.
26. ANSI/UL 764 A – 1984, UL Publications 45. H. Schmidt, K. H. Schüller: “Untersuchungen
Stock, Northbrook, Ill., USA. an Sondersteatiten mit Kalk,” Silikattechnik
27. IEC Publication 93 – 1980, Bureau Central de 39 (1988) 137 – 138.
la Commission Electrotechnique International, 46. W. Büssem, C. Schusterius, K. Stuckard: “über
Genève. die Konstitution des Steatis,” Wiss. Veroeff.
28. IEC Publication 247 – 1978, Bureau Central de Siemens-Werken 17 (1938) 59 – 89.
la Commission Electrotechnique International, 47. A. C. M. Wilson: Insulated Liquids, The
Genève. Institute of Electric Engineers, London 1980.
29. IEC Publication 345 – 1971, Bureau Central de 48. IEC Publication 296 – 1969, Bureau Central de
la Commission Electrotechnique International, la Commission Electrotechnique International,
Genève. Genève.
Insulation, Electric 49
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
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
tually unlimited quantities of IFN has permitted In addition to viruses, a number of natu-
detailed biological and structural characteriza- ral and synthetic agents are capable of induc-
tion, as well as extensive clinical trials world- ing IFN production in a wide variety of cell
wide [9], [10]. types [5]. A general summary of cell types and
Genetically engineered IFNα-2 b ( Intron A) IFN inducers for the three IFN classes is pre-
has now been approved for marketing in over sented in Table 1. Interferon-inducing agents
30 countries, including the United States, for include bacteria, protozoa, mycoplasma, rick-
both cancer and viral indications such as hairy ettsia, and bacterial products (e.g., bacterial cell
cell leukemia, acquired immune deficiency syn- wall extracts, lipopolysaccharides, and endotox-
drome (AIDS)-related Kaposi’s sarcoma, renal ins). The effect of synthetic and natural nucleic
cell carcinoma, malignant melanoma, multiple acids and polyanions, including double-stranded
myeloma, laryngeal papillomatosis, and condy- RNA, synthetic polynucleotides, and synthetic
loma acuminata (genital warts). Extensive clin- polycarboxylates, has been of special interest.
ical trials have also been performed on IFNγ Low molecular mass compounds such as anthra-
[11–14] and IFNβ [15], [16]. This article re- quinones, propanediamines, acridines, pyrimi-
views the properties, production, and specifica- dines and fluorenones can also induce IFN titers
tions of recombinant human IFNα, for which the in the serum of certain animals. Specific struc-
most data are available. However, for compara- tural requirements for these low molecular mass
tive purposes selected data are also provided for compounds to induce IFN have not been identi-
human IFNβ and IFNγ, as well as natural leuko- fied yet. Metabolic inhibitors (e.g., actinomycin
cyte IFN. For further information, see [2–4], [9], D), mitogens, antigens, and tumor cells have also
[17–44]. been identified as IFN inducers.
Table 1. Classes, sources, and inducers of IFN
effects as measured by its ability to enhance the Both of these molecules lack a cysteine residue
activity of immune effector cells (e.g., natural at position 1 and, hence, are predicted to con-
killer cells) [46]. The range of biological ac- tain only one disulfide bond (see Chap. 5). The
tivity demonstrated for IFN has increased con- hybrid molecule consists of residues 5 – 62 of
siderably [47]. One of the important features of IFNα-2 and residues 64 – 166 of IFNα-1. De-
IFNs is their species specificity; all classes of hu- spite the high degree of purity of these prepara-
man IFN are most effective on human cells and tions and their amino acid sequence homology
are generally much less potent on animal cells. (at least 70 %), the specific antiviral activities
However, because many of the early studies on span approximately two orders of magnitude.
nonantiviral activity were performed with im- Interferon α-2 has the highest antiviral activity
pure preparations of naturally occurring IFNs, (1.7×108 IU/mg).
these biological effects were not known to be
Table 2. Specific antiviral activity of IFNα subtypes ∗
inherent properties of IFN. Subsequent use of
highly purified recombinant IFN has demon- IFN subtype Activity, IU/mg
strated that most observed biological effects can
α-1 7.1×106
be attributed to IFN [2], [3]. α-2 1.7×108
These properties of IFN provide strong sup- α-4 1.0×108
port for the clinical application of this class of α-7 2.0×107
δ-4 α-1 2.8×106
biological response modifiers as both antiviral δ-4 α-2/α-1 1.0×108
and antitumor agents. Aspects of the three cate-
gories of biological activity of IFN are reviewed ∗ Activity was determined by the CPE-inhibition assay
employing encephalomyocarditis virus and human foreskin cells;
briefly below. data represent the mean of at least nine assays [50].
Synergy may be the result of a complex interac- I MHC antigens play a role in the lysis of virus-
tion of factors, including the nature of the tumor, infected cells and tumor cells by cytotoxic T
the mechanism of action of the cytotoxic agent, lymphocytes. Class II MHC antigens are essen-
and the administration schedule of IFN and the tial for the ability of macrophages to function as
cytotoxic agent. Thus, synergistic interactions antigen-presenting cells. However, both classes
between IFN and cytotoxic drugs have been ob- of MHC antigens are fundamentally important
served in the human tumor clonogenic assay uti- for achieving maximal immune responsiveness.
lizing recombinant IFNα and vinblastine, cis- All types of IFN appear to inhibit antibody pro-
platin, or doxorubicin [54], [55]. However, syn- duction, although under certain conditions of
ergy is highly schedule-dependent. When cells dosing and time of addition, an enhancement of
were first treated with IFN antibody production has been observed.
α followed by doxorubicin or vice versa, no The role of immune stimulation in clinical
synergy was observed. In contrast, dramatic syn- responses to IFN remains unknown. For exam-
ergy was observed when both agents were ad- ple, measurements of natural killer cell activity
ministered simultaneously. Other drugs tested after administration of recombinant IFNα have
in this assay (e.g., bleomycin, methotrexate, or demonstrated both enhancement and depression
vinca alkaloids) were not synergistic with IFNα. of activity, as well as no response [58], [59].
The mechanism of the observed synergy as well However, stimulation by IFN of cytotoxic effec-
as of drug failure remains unknown. tor cell function and antibody production may
Another important aspect of the anticellular still represent major mechanisms of its thera-
activity of IFNα is its ability to modulate the peutic activity.
state of cellular differentiation. Both inhibition
and acceleration of differentiation have been ob-
served [56]. Interferon also interacts synergisti- 4. Mechanism of Action
cally with other agents that inhibit differentia-
tion, e.g., phorbol ester tumor promoters [57]. The first step in expression of the biological ac-
The induction of differentiation in tumor cells tivity of IFN is its binding to specific cell sur-
mediated by IFN may account for at least a part face receptors [60]. The number of receptors
of its antitumor activity because the degree of for IFNα on human cells is reported to be ca.
cellular differentiation appears to be inversely 650 – 13 000 per cell. The ligand and receptor
related to the rate of cell division. 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
3.3. Immunomodulatory Activity mass of 140 000 – 150 000 deduced from chem-
ical cross-linking of radioactively labeled IFN
A variety of immune functions can be affected to the cell surface. Thus, the apparent molecu-
by IFN [41]. Immune response may be mediated lar mass of the receptor is 120 000 – 130 000 if
either by cells or by antibodies. Interferon can one molecule of IFN is bound per receptor. The
activate or inhibit both cellular and humoral (an- failure of certain cells to respond to IFN may be
tibody) immune function: higher concentrations due to absence of the receptor or a low affinity
generally inhibit, whereas lower concentrations of IFN for the receptor.
activate. Natural killer cell activity as well as Ultrastructural studies indicate that bound
macrophage tumoricidal and tumoristatic activ- IFNα enters cells by a clustering of the recep-
ities can be significantly enhanced by both IFNα tor – ligand complex into “coated pits.” These
and IFNγ. Other types of immune effector cells pits, which are surrounded by the protein
activated by IFN include cytotoxic T lympho- clathrin, function in the transport of a variety
cytes, cells involved in antibody-dependent cy- of molecules into the cell. The pits pinch off to
totoxicity, and mast cells. Interferons of all three form vesicles (so-called receptosomes), which
types induce expression of class I antigens of the can transport their contents to a variety of tar-
major histocompatibility complex (MHC), and gets within the cell. The role of internalization of
IFNγ induces expression of class II MHC anti- IFNα into the cell in the mechanism of action is
gens. This property is significant because class
Interferons 5
yet to be elucidated. Interferon alone, the IFN – alyzes the cleavage of both messenger and ri-
receptor complex, or IFN degradation products bosomal RNA. Thus, the localized activation of
may trigger a biological response; alternatively, endonuclease by double-stranded RNA present
binding of IFN to the cell surface may itself be during viral replication may account for the se-
sufficient. lective cleavage of viral versus host RNA.
Binding of IFNα to its receptor induces the Interferon induces another system that also
synthesis of about a dozen different proteins requires double-stranded RNA for activation,
ranging from 15 000 to 120 000 in molecular namely, a protein kinase that catalyzes the phos-
mass. Many of these have not been identified, phorylation of protein initiation factor eIF-2, as
and their role in the mechanism of IFN activ- well as the ribosomal protein P 1 [65]. In cell-
ity is unknown. Interferon also affects the ac- free systems, the kinase is activated by much
tivity of specific proteins or enzymes. For ex- lower concentrations of double-stranded RNA
ample, levels of ornithine decarboxylase and than is 2 ,5 -oligoadenylate synthetase; higher
S-adenosylmethionine decarboxylase, enzymes concentrations of double-stranded RNA inhibit
involved in polyamine biosynthesis, are reduced the kinase but activate the synthetase. Thus, the
by IFN treatment [61]. Because these enzymes level of double-stranded RNA may determine
are critical for cellular proliferation, their inhi- which biochemical pathway is important in the
bition may account for the antigrowth activity expression of antiviral activity.
of IFN. Interferon treatment can increase cyclic
guanosine 5 -monophosphate (cGMP) and cy-
clic adenosine 5 -monophosphate (cAMP); both 5. Physicochemical Properties
of these compounds have also been implicated
in growth regulation [61]. However, the exact Only very low (picogram) quantities of IFN
biochemical mechanism for the antiproliferative can normally be produced from natural sources
effects of IFN remains unknown. because the biological activity is expressed at
As noted above, all three IFN classes can very low protein levels. Thus, characterization
induce the expression of class I MHC anti- of its properties was for many years slow and
gens [62], which mediate the lysis of virus- incomplete. Studies on IFNα prepared from a
infected cells and tumor cells by cytotoxic T variety of natural sources, including virally in-
lymphocytes. In addition, IFN can enhance the duced natural leukocytes, lymphoblastoid cells,
expression of the immunoglobulin G receptor and chronic myelogenous leukemic cells, sug-
on lymphocytes and macrophages; this, in turn, gested that IFNα consists of a family of multi-
may enhance antibody-dependent cytotoxicity ple species with apparent molecular masses de-
or macrophage-mediated phagocytosis [63]. termined by sodium dodecyl sulfate – polyacryl-
An extensively characterized enzyme that is amide gel electrophoresis (SDS – PAGE), rang-
induced by IFN treatment and appears to play ing from 16 000 to 27 000. Amino acid analy-
a critical role in interferon’s antiviral and an- sis of the purified proteins indicated that their
tiproliferative activities is 2 ,5 -oligoadenylate compositions were similar but that IFNα con-
synthetase (E. C. 2.7.7.-) [64]. This enzyme sisted of a family of proteins with different ami-
catalyzes the synthesis of a family of novel no acid sequences. However, further characteri-
2 ,5 -oligoadenylate nucleotides that regulate zation of the physicochemical properties of IFNs
the breakdown of RNA in cells. Reported levels was severely hindered by the fact that IFNs
of induction of the synthetase vary between 10- purified from natural sources were frequently
and 10 000-fold. The enzyme requires double- contaminated with other proteins and IFN sub-
stranded RNA, which may be produced by the types. In addition, the extremely low levels of
virus as a replicative intermediate, for expres- substance produced resulted in large losses of
sion of enzyme activity. Although the exact protein because of adsorption and surface de-
mechanism for this activation is not known, the naturation. These problems were solved by the
double-stranded RNA may bind to the enzyme large-scale production of genetically engineered
and convert it to its biologically active form. IFNα-2 b ( Intron A) in E. coli [3, pp. 1 – 12],
The 2 ,5 -oligoadenylates activate a latent en- [10].
doribonuclease, known as RNase L, which cat-
6 Interferons
three-dimensional structure at this degree of res- and various leukemic cell lines. The principal
olution. 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
Figure 4. Human IFNγ crystals in a suspended droplet 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-
5.3. Solution Properties tocols.
Partially purified human leukocyte IFN was
Based on their predicted amino acid composi- first developed by Cantell and his associates in
tions, the polypeptide molecular masses for hu- Finland and has been available in limited quan-
man IFNα, β, γ are 19 269, 19 990, and 17 145, tities for clinical trials since 1973 [71]. In this
respectively. However, under certain conditions procedure, fresh blood from normal donors is
the apparent molecular mass of IFNα in solu- centrifuged, and the leukocyte fraction (i.e., the
tion appears to be considerably higher than the buffy coat) is removed. Pooled leukocytes are
polypeptide molecular mass [10]. Sedimenta- treated with 0.83 % ammonium chloride to lyse
tion velocity experiments have established that residual erythrocyte contaminants, suspended in
IFNα-2 undergoes self-association to yield a a culture medium (e.g., Eagle’s minimum es-
molecular mass at least three times that of the sential medium), and then primed with high-
monomer. Lowering the pH from 7 to 2 fa- titer crude IFN. Interferons production (primar-
vors monomer formation. Under the low pro- ily by monocytes) is then induced by addition
tein concentration in a biological assay, it is an- of Sendai virus. After overnight incubation, the
ticipated that IFNα is completely monomeric; cells and debris are removed by centrifugation;
thus, the monomer is probably the active species. the supernatant containing crude IFN is the start-
Even under highly concentrated conditions (e.g., ing material for purification.
10 mg/mL or higher), IFNα-2 b appears to be Crude IFN is first precipitated with potassium
very soluble. At its isoelectric point (ca. pH 6.0), thiocyanate at pH 3.4 and extracted with 95 %
IFNα is relatively insoluble and can be crystal- ethanol. Extraction with acidified ethanol inac-
lized readily. In contrast, IFNγ appears to form tivates residual Sendai virus and partially puri-
a stable dimer that does not dissociate on dilu- fies the IFN because not all proteins redissolve.
tion and has a highly basic isoelectric point of The pH is then increased stepwise to precipitate
ca. 9.5. contaminating proteins, which are removed by
centrifugation. The IFN is precipitated at pH 8,
concentrated, and dialyzed against phosphate-
6. Purification buffered saline.
The specific antiviral activity of the IFN
6.1. Natural Leukocyte IFN preparation is approximately 106 IU/mg of pro-
tein, which represents approximately 1 % purity.
Starting materials used for the purification of hu- The net yield per unit of blood (i.e., 0.5 L) is
man IFN include leukocytes isolated from whole ca. (5 – 7)×106 antiviral units. The final product
blood, neonatal fibroblasts, and lymphoblastoid contains a number of different species of IFNα,
Interferons 9
including α-1, α-2, and α-4. Other lymphokines ical activity or structure of the recombinant
(i.e., hormones released by leukocytes such as protein.
interleukin-1 and 2) probably also contaminate 3) The purified IFN must be substantially free
the preparation. of nucleic acids and proteins derived from the
The partially purified leukocyte IFN prepa- host microorganism if it is to be employed in
ration can be purified further to a specific activ- human clinical trials.
ity of ca. 108 IU/mg of protein by chromatog- 4) The product must be a homogeneous IFN
raphy on an immobilized monoclonal antibody species that is free of aggregates and pro-
referred to as NK-2. The IFN is eluted from the teolytic fragments. These IFN-related com-
column by exposure to low pH. However, the ponents are of special concern because they
relative composition of IFNα subtypes in this could render the product antigenic when in-
preparation differs from that in crude IFN prior jected into humans.
to chromatography because the NK-2 antibody 5) To be commercially viable, purification must
does not bind all IFNα species equally. be achieved with a good yield (e.g., 20 %) of
biological activity (e.g., antiviral units).
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 diluted (usually by a factor of two) in the cells
pH gradient are finally removed from the prepa- of a microtiter plate, and target cells (e.g., hu-
ration by crystallization of the protein, gel fil- man foreskin diploid fibroblasts) are then added
tration chromatography, or ammonium sulfate to each well. Cells to which IFN has not been
precipitation. added are also included as controls. After an
incubation period (e.g., 4 h), a challenge virus
(e.g., encephalomyocarditis virus) is added, fol-
7. Quality Specifications lowed by a further period of incubation. In the
presence of IFN, cells are protected from viral
7.1. Purity infection and can be detected by staining with a
dye such as Crystal Violet. However, in the ab-
Purified IFN must be shown to be biologically sence of IFN, the cells lyse and are not stained
active, and its activity must remain constant from by the dye because they do not adhere to the mi-
batch to batch. Because IFN was originally de- crotiter plate. The end point is arbitrarily defined
fined in terms of its antiviral activity, quantita- (e.g., 50 % protection against viral infectivity).
tion of antiviral activity per milligram of puri- An appropriate international reference standard
fied protein is routinely employed for charac- provided by the World Health Organization is in-
terization of purified product. The most com- cluded in each microtiter plate and allows con-
monly employed assay, which is readily adapted version to the international antiviral units (IU)
to screening large numbers of samples, is the cy- established by international accord. Further de-
topathic effect- (CPE-) inhibition assay. This as- tails of the CPE-inhibition assay are described
say is based on the ability of IFN to interfere with in [77].
the cytopathic effects (e.g., cell lysis) caused by
viruses. In a simple CPE assay, IFN is serially
12 Interferons
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
Intermetallics
Gerhard Sauthoff, Max-Planck-Institut für Eisenforschung, Düsseldorf, Germany
tural materials are being developed around the metallics is accompanied by an unusual brittle-
world, in particular in the United States, Japan, ness, and the reasons for this were sought.
and Germany. Besides intermetallics with out- In the 1900s intermetallics were used primar-
standing high-temperature performance, inter- ily as functional materials. The first industrial
metallics with other outstanding physical prop- applications relied on the special magnetic be-
erties led to the development of new functional havior of certain compounds, and materials de-
materials much earlier. velopments led, e.g., to Sendust which shows
The results of the numerous and manifold outstanding magnetic properties and wear resis-
research and development activities on inter- tance and is widely used for magnetic heads in
metallics have been published in the respec- tape recorders. In the second half of the twen-
tive conference proceedings volumes and review tieth century another important application re-
monographs [2 – 8] in addition to scientific jour- sulted from the development of new supercon-
nals and research reports. ducting materials on the basis of the A15 com-
The present overview, which is an updated pounds, which are used for superconducting
summary of a previous report [2], briefly de- magnets. A third group of functional materi-
scribes the various intermetallics which have al- als — the shape-memory alloys — make use of a
ready been selected for materials developments martensitic phase transformation and have found
or which have been and still are regarded as manifold applications since ca. 1970.
promising for materials developments. An important group of functional materials
are the III/V compounds such as InSb, InAs, and
GaAs, which have found applications in elec-
1.2. Historical Remarks tronics and thermoelectric power generation.
The constituent elements of these compounds re-
Intermetallics have been exploited since the be- present the transition from metals to metalloids
ginning of metallurgy, as is exemplified in Table and nonmetals, and the resulting compounds
1. The early intermetallics resulted from alloy are semiconductors. Hence these intermetallic
systems with low melting temperatures, and the semiconductors lie outside the field of true in-
applications relied on the outstanding hardness termetallics with metallic behavior.
and wear resistance of intermetallics together Intermetallics were not used as structural ma-
with their metallic properties. Because of the terials in the past because of their brittleness.
metallic behavior of intermetallics, they could The only noteworthy exception is the continu-
be polished, and thus the decorative aspect was ing use of amalgams as dental restoratives. How-
also important for many applications. Examples ever, various intermetallics have been success-
are the bronze coatings in ancient Egypt and the fully used as strengthening second phases in
mirrors used by the ancient Chinese, Etruscans, conventional alloys for structural applications at
and Romans. Some intermetallics show beauti- high temperatures. Thus it became clear that in-
ful colors, e.g. the violet Cu2 Sb (“Regulus of termetallics are promising as structural materials
Venus”). for high temperatures because of their high hard-
Intermetallics became a subject of scien- ness and stability, and numerous investigations
tific research with the development of physi- began around 1950 with the aim of clarifying the
cal metallurgy. Intensive and broad-scale work potential of intermetallics for structural applica-
on intermetallics was initiated by Tammann in tions. Various promising candidates were iden-
Göttingen, Germany, and K urnakov in St. Pe- tified, but the respective materials developments
tersburg, Russia, around 1900. Work on inter- were hampered by unsolved brittleness prob-
metallics during the first decades of the 1900s al- lems, and these various activities slowly died
ready included studies of phase stabilities, phase out in the 1960s.
equilibria, and phase reactions in order to es- One offspring of these activities was the
tablish phase diagrams, and investigations of development of the electric heating elements
chemical, electrochemical properties, and phys- based on MoSi2 , which rely on the advanta-
ical properties including magnetism, supercon- geous chemical behavior of this compound, that
ductivity, and mechanical properties. It was rec- is, its high oxidation resistance at high tempera-
ognized that the outstanding hardness of inter- tures. In the course of superalloy research, sur-
Intermetallics 3
Table 1. Some applications of intermetallics in past and present times [2]
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- important group of these intermetallics are the
terized by completely filled electron orbitals, A15 compounds. The best known size-factor
in particular, by a full octet shell in the nor- compounds are the Laves phases, which form
mal case. Thus they may be regarded as valence the most numerous group of intermetallics and
compounds which satisfy the familiar chemi- which crystallize in the closely related hexago-
cal valency rules. The Zintl phases have crys- nal C14, cubic C15, or hexagonal C36 structures,
tal structures which are characteristic for typ- whereby MgZn2 , MgCu2 , and MgNi2 , respec-
ical salts, for example, NaTl with cubic B32 tively, are typical examples. The valence elec-
structure (Strukturbericht designations are used tron concentration, however, is also an impor-
here and in the following for the crystal struc- tant factor here, since it decides between the
tures) and Mg2 Si with cubic C1 structure, and three crystal structures. These crystal structures
therefore ionic bonding is expected. However, are characterized by high symmetry, high atomic
all types of bonding — ionic, metallic, and co- coordination, and high density in close analogy
valent — and mixtures thereof have been found, to metals, and indeed the Laves phases show pri-
depending on the particular electron distribu- marily metallic bonding according to the scarce
tion. Hence the Zintl phases may also be re- information available.
garded as electron compounds whose electron Compounds which form by a phase transi-
band-structure energy is a large fraction of the tion with lowering of crystal symmetry from the
total energy and the crystal structure types of solid solution of the constituent elements at a
which are related to particular valence electron transition temperature below the melting tem-
concentrations (average number of valence elec- perature are known as Kurnakov phases. In the
trons per atom). simplest case the phase transition results from an
The best-known electron compounds are the ordering reaction in the solid-solution lattice in
Hume – Rothery phases, which have the cubic which a superlattice structure is established. Ex-
B2 structure (e.g., β-brass at high temperatures amples for compound formation by atomic or-
and the transition metal aluminides FeAl, NiAl, dering are the compounds Fe3 Al with D03 struc-
and CoAl), the complex cubic A13 structure ture, which is formed by two-step ordering from
(β-Mn type; e.g., Zn3 Co, Cu5 Si), or the close- the bcc solid solution, as well as Ni3 Fe with L12
packed hexagonal A3 structure (e.g.; Cu3 Ga, structure, CuAu with L10 structure, and Cu3 Au
Ag3 Al) for the valence electron concentration with L12 structure, the latter three of which form
VEC = 3/2; the complex cubic D82 structure from fcc solid solutions. This atomic ordering
(γ-brass type) for VEC = 21/13 (e.g.; Cu5 Zn8 , with superlattice formation clearly results from
Fe5 Zn2 ); and the A3 structure again for VEC = stronger bonding between unlike atoms com-
7/4 (e.g., CuZn3 , Ag5 Al3 ). The bonding in such pared with bonding to like atoms. The difference
intermetallics is not purely metallic, in spite of in bonding energy between unlike and like atoms
the metal-like band structure. For example, the is usually known as interaction energy and cor-
bonding in the above-mentioned NiAl is essen- responds to an ordering energy. Hence, higher
tially covalent with some metallic character and crystal stability and higher ordering temperature
no ionic component. Recent ab initio calcula- are expected for a higher interaction energy. The
tions showed that, strictly speaking, the B2 alu- degree of order, which depends on temperature
minides are not Hume – Rothery electron com- and composition, as well as the order/disorder
pounds. This means that the Hume – Rothery temperature can be modeled as a function of
rules, which relate crystal structure to valence the interaction energy for a given compound to
electron concentration, represent a very simpli- various degrees of approximation, for example,
fied picture of the bonding situation. Neverthe- to a first approximation by the simple Bragg –
less, such rules are useful since they describe Williams model or to a better approximation by
reality surprisingly well. the cluster variation or the Monte Carlo method.
Other intermetallics are best characterized Apart from the degree of approximation, the pre-
by the size ratios and packing schemes of the dictive power of such models with respect to
constituent atoms and are therefore called size- phase-stability and phase-diagram calculations
factor compounds, topologically close packed relies on the knowledge of the interaction en-
intermetallics, or Frank – Kaspar phases. One ergy and on whether such interaction energies
Intermetallics 5
are sufficient to characterize the atomic bond- It might be supposed that the particular crys-
ing state. Analyses of various transition metal tal structure of an intermetallic reflects the char-
systems have shown that the modeling with the acter and strength of bonding in a systematic
cluster variation or Monte Carlo method gives way and that crystal structure would be a good
good agreements with the experimental obser- criterion for compound classification. However,
vations in most cases, whereas in some com- this does not mean that intermetallics with the
pounds with bcc structures the simple Bragg – same crystal structure are similar with respect
Williams model gave better results than the more to bonding and properties. For example, the B2
sophisticated models. structure is common to the intermetallic NiAl
It follows that single simple atomic param- and to the purely ionic salt CsCl, and in the case
eters like the interaction energy, atomic size, of the transition-metal disilicides with hexago-
or valence electron concentration have only a nal C40 structure, NbSi2 is metallic but CrSi2 is
limited validity and are generally not sufficient a semiconductor. Recent ab initio calculations
for predicting the stability of a compound as a showed that atomic coordination and connectiv-
function of the constituent elements. A better ity is more relevant to crystal energy and sta-
and more general correlation between the atomic bility than symmetry. In view of the complex-
properties and the compound crystal structure ity of the classification of intermetallics, inter-
should be obtained if the particularities of the metallics are often grouped according to more
electron configurations of the constituent atoms practical criteria which refer to similarities in
are taken into account; these are reflected in the behavior.
positions in the periodic table of elements. A
better correlation has indeed been found, but the
resulting structure maps still have their limita- 2.2. Basic Properties
tions because the changes in electron configura-
tion due to compound formation are not consid- A basic property is the melting temperature since
ered. materials parameters which characterize the de-
It must be concluded that the bonding charac- formation behavior are well correlated with the
ter and crystal structure of an intermetallic com- melting temperature. Examples are the elastic
pound can be predicted in a significant way only moduli, which not only control the elastic de-
on the basis of quantum-mechanical ab initio formation, but are also important parameters for
calculations on the respective compound. Much describing plastic deformation, and the diffusion
progress has been made in this respect, and for coefficients, which determine not only the ki-
various important compounds with not too com- netics of phase reactions, but also the kinetics
plicated structures the crystal energy has been of high-temperature deformation, that is, creep.
calculated as a function of lattice structure in or- Furthermore, the melting temperature is intu-
der to determine the compound stability. How- itively regarded as a measure of compound sta-
ever, these calculations are very time-consuming bility, since it limits the application temperature
even in simple cases and more progress is nec- range.
essary in the future in order to give guidance to However, a compound melts when the Gibbs
practical materials developments on the basis of energy of the solid phase, which is a thermo-
multinary compounds with less simple crystal dynamic state function and which controls the
structures. phase stability, is higher than that of the liq-
The intermetallics do not form a homoge- uid phase. The Gibbs energy of a phase is given
neous group of materials. Instead the term in- by the enthalpy and the product of entropy and
termetallics comprises a huge variety of com- temperature, that is, its temperature dependence
pounds which differ drastically with respect is determined primarily by the entropy. Hence,
to bonding, crystal structure, and properties. the melting temperature depends on the differ-
Therefore, properties of intermetallics cannot be ences in enthalpy and entropy between the solid
discussed generally for all intermetallics, but can and liquid states and is a rather complex func-
be discussed only by referring to specific groups tion of the bonding strength. Only the enthalpy
of intermetallics. is directly related to the internal crystal energy,
which is determined by the bonding strength.
6 Intermetallics
Nevertheless, the melting temperature correlates potential wells of the atoms, which cannot be
surprisingly well with the enthalpy of formation characterized by a single parameter. A similar
for sufficiently similar compounds. correlation has been found for the activation en-
It may be concluded that the enthalpy of com- ergy of diffusion.
pound formation may be a better parameter for It must be concluded that there are physically
characterizing bonding strength and compound justified correlations between parameters which
stability and for correlating this with fundamen- characterize bonding strength and compound
tal properties such as elastic moduli. Formation stability and the macroscopic compound behav-
enthalpies have been determined experimen- ior. However, such correlations represent com-
tally, and they can be calculated by quantum- plex functional relationships, and hence they are
mechanical calculations or estimated by more useful only for order-of-magnitude predictions.
or less empirical approaches. More quantitative predictions have to consider
However, the conventional formation en- the character and strength of bonding in a more
thalpy values refer to compound formation from detailed way, that is, they have to rely on quan-
the constituent elements at room temperature, tum mechanical calculations which are cumber-
that is, from the solid elemental crystals, and thus some and time-consuming.
these formation enthalpies characterize only Recently, fundamental properties, especially
bonding strength changes during compound for- elastic moduli, of various simple compounds
mation. For the complete characterization of have been studied by ab initio calculations, and
bonding strength the total compound enthalpy much progress has been made in this field. Thus,
must be considered, which refers to compound a basic understanding of the bonding character
formation from the elements in the gaseous state. of, for example, transition metal aluminides was
The total enthalpy of a compound may be es- reached, and it was shown that the degree of
timated by adding the formation enthalpy and directional d-electron bonding determines the
the enthalpies of fusion and vaporization of the ideal cleavage strength. However, it has also
constituent elements. As an example, data for been shown for B2 transition-metal aluminides
the Ni – Al system show a good agreement of that the ductility or brittleness is too complex a
estimates with experimental data. Since the for- function of strength and character of bonding for
mation enthalpies are small compared with the establishing simple relations between electron
total enthalpies, the enthalpies of the constituent distribution parameters and mechanical proper-
elements are the major contributions to the total ties.
enthalpies.
The correlation between the total enthalpy of
a compound and Young’s modulus, which is one 2.3. Criteria for Compound Selection
of the three elastic moduli, has been checked for
various cubic compounds: for the fcc elements Candidate compounds for materials develop-
Al and Ni (A1), the fcc-ordered Ni3 Al (L12 ), the ments are selected in view of specific applica-
bcc-ordered FeAl, NiAl, and CoAl (B2), and the tions, that is, they must show certain properties
cubic Laves phases CaAl2 , YAl2 , LaAl2 , NbCr2 , which make them promising for a particular ap-
ZrCo2 and HfCo2 (C15). The correlation is quite plication. For permanent magnets the “energy
good for the Laves phases, in that there seems product” (BH)max (B = magnetic induction, H
to be a common scatter band, near which the = magnetic field strength) is a figure of merit
data of some other compounds lie. However, the and should reach high values whereas for other
data points may be subject to appreciable exper- functional materials other physical parameters
imental uncertainties, as indicated by data for are decisive.
Ni3 Al. Such uncertainties for Young’s modulus Intermetallic compounds for structural appli-
may be due to differences in microstructure, that cations must show a sufficient strength at service
is, texture and artefacts. Apart from this, the data temperature which also means a sufficient creep
for the B2 compounds show that Young’s mod- resistance. The creep resistance is correlated
ulus is not a simple function of the enthalpy of with the diffusion coefficient and with the shear
the compound. This must be expected because modulus, and both parameters in turn with the
Young’s modulus is a complex function of the melting temperature. Hence, the candidate com-
Intermetallics 7
pound should have a sufficiently high melting T m for Si and Al2 O3 , that is, the deformation be-
temperature. The limiting temperature for struc- havior of such an intermetallic may be regarded
tural applications of conventional metallic alloys as intermediate between those of metals and ce-
is of the order of 75 % of the melting temperature ramics.
in many cases. The most advanced high-tempe- Plastic deformation is more difficult in inter-
rature alloys are the Ni-based superalloys, which metallics than in metals and conventional metal
are used in gas turbines with service tempera- alloys because of the stronger atomic bond-
tures of up to about 1100 ◦ C. Hence compounds ing and the resulting ordered atom distribution,
with melting temperatures above 1600 ◦ C must which gives rise to more complex crystal struc-
be considered if application temperatures above tures. The brittleness of intermetallics increases
those of superalloys are aimed at. High-melting with decreasing lattice symmetry and increasing
compounds may form low-melting eutectics that unit cell size. Therefore intermetallics with high
can affect the stability of the compound dur- crystal symmetry — possibly cubic compounds
ing processing or service, and therefore such — and small unit cells are preferred for develop-
compounds with low-melting eutectics should ing new structural materials. Indeed the candi-
be avoided. date compounds FeAl, NiAl, Fe3 Al, Ni3 Al, TiAl
For many functional and structural applica- for development of structural materials have
tions density is a very important and sometimes crystal structures which are cubic (B2, D03 or
decisive parameter. Structural intermetallics for L21 , L12 , L 12 ) or nearly cubic, that is, tetrago-
moving components must show a sufficiently nally distorted (L10 , D022 ).
high specific strength, which is the ratio of However, even for such selected compounds,
strength to density and has the dimension of brittleness remains a problem, and the causes
length. The specific fracture strength, that is, for this are manifold. A material fractures in a
the rupture length, of advanced superalloys such brittle manner if there is no plastic deformation
as the mechanically alloyed oxide dispersion and stress relaxation at the crack tip, that is, the
strengthened (ODS) superalloy MA 6000 is of yield stress is higher than the stress for cleav-
the order of 15 km, which would have to be sur- age or fracture. The reason for this may be an
passed by new structural materials. Compounds insufficient number and/or mobility of disloca-
which contain light elements — e.g. Ti, Al, Si, tions and/or an insufficient number of slip sys-
Mg — may have such a low density that they tems. Since these parameters depend not only on
compare favorably with conventional alloys in crystal symmetry but also on the specific char-
spite of lower strength or restricted temperature acteristics of the particular compounds, the aim
range. of compound selection and composition varia-
The main problem of strong intermetallics is tion by alloying must be dislocations with low
their brittleness, which makes processing and energy and high mobility and at least five in-
application difficult or impossible. However, the dependent deformation modes (i.e. dislocation
brittleness of intermetallics should be less severe slip systems and twinning systems), which are
than that of ceramics because the atomic bond- necessary for general homogeneous plastic de-
ing of intermetallics is at least partially metal- formation according to the Von Mises criterion.
lic, whereas it is primarily covalent or ionic in Crystal anisotropy is an important factor in brit-
ceramics. This is illustrated by the temperature tleness, and criteria are used which rely on elas-
dependence of yield stress and deformability of tic moduli relations. Brittle fracture may be a
the intermetallic NiAl in comparison with that result of weak grain boundaries and other mi-
of nickel on the one hand and of silicon and crostructure heterogeneities leading to localized
alumina on the other hand. NiAl which is a can- deformation and stress concentrations. The dis-
didate compound for high-temperature applica- proportion of yield stress and fracture stress may
tions softens with rising temperature at about also be due to a very low surface energy, which
half the melting temperature T m , which is the makes cleavage and cracking easy. This may
range of Si with covalent bonding and which be aggravated by the segregation of impurities,
is higher than that of the ceramic Al2 O3 . How- which usually lower surface energies. Impuri-
ever, the brittle-to-ductile transition temperature ties, in particular oxygen and hydrogen, may
is about 0.4 T m for NiAl whereas it is above 0.8 enter the material from the environment, so en-
8 Intermetallics
vironmental embrittlement is an important issue first discovered by Mishima in 1931, and this
for many intermetallics. A detailed knowledge initiated the development of the AlNiCo al-
of the thermodynamics and kinetics of the par- loys for applications as permanent magnets (→
ticular alloy system is necessary for a reduction Magnetic Materials, → Cobalt and Cobalt Com-
of the various embrittling effects. pounds, Chap. 6.3). The basic alloy contains
Intermetallics for any application must be about 50 atom % Fe, 25 atom % Ni, and 25 atom
corrosion resistant in the operating environment. % Al, to which various other elements (Co,
For high-temperature applications this means Cu as macroalloying elements and Nb, Ti as
oxidation resistance in many cases. Oxidation microalloying elements) are added for prop-
resistance is provided by the presence of ele- erty optimization. The alloys are single-phase at
ments, in particular Cr, Al, Si, which can form high temperatures and decompose spinodally on
protective oxide layers. However, chromium ox- cooling at a controlled rate to form a two-phase
ide is volatile above 1000 ◦ C and silicon oxide microstructure with interlocking phase distribu-
may form low-melting silicates. Therefore, alu- tion. The additional alloying elements Co, Mn,
minides are strongly favored for high-tempera- Ti, and Cu, which dissolve preferentially in the
ture applications. To act as protective layers, the B2 compound, and Cr, Mo, V, Nb, and Si, which
scales must show a sufficient density and adher- dissolve preferentially in the disordered Fe-rich
ence, which may be improved by alloying with compound, control the decomposition process
further elements (macroalloying) and in partic- of the AlNiCo alloys. The constituent phases
ular by adding minor amounts of “active” ele- of these alloys are the weakly ferromagnetic
ments such as Ti, Zr, Hf (microalloying). Fur- NiAl base compound with B2 structure and the
thermore, scales must show sufficient long-term strongly ferromagnetic, disordered bcc Fe-rich
stability, which also means mechanical stabil- compound. The interlocking phase distribution
ity, that is, they must tolerate strains, for exam- and the strong shape anisotropy of the Fe-rich
ple, during creep, without being damaged. In the compound result in high coercivity. The mag-
case of insufficient oxidation resistance protec- netic behavior is a sensitive function of pretreat-
tive coatings may be applied. However, the ther- ment.
mal stability of such coatings decreases with in- AlNiCo alloys, which are produced by cast-
creasing temperature because of increasing re- ing or sintering, are brittle and hard; they can
action and diffusion rates and therefore inter- be machined only by grinding, spark erosion,
metallics for applications above 1100 ◦ C should and electrochemical milling, and they resist at-
be inherently oxidation-resistant. mospheric corrosion well up to 500 ◦ C. The
Finally, it must already be possible to prepare magnetic properties are summarized and com-
an intermetallic with sufficient quality in the lab- pared to those of other alloys in [10 – 12]. Al-
oratory, and from a practical point of view this NiCo alloys are superior to other permanent
is the basic requirement for any materials devel- magnet materials in resisting the effects of tem-
opment. The processing of high-strength inter- perature on magnetic properties. Presently, NiAl
metallics is difficult because of their brittleness, is also considered an attractive candidate com-
and the development of the necessary processing pound for other functional applications, for ex-
techniques is a very demanding task. It is, how- ample, contacts in thin-film electronic devices
ever, also a very important task, because a poor or high-voltage electrodes in electron-optical in-
quality increases the apparent brittleness and re- struments.
duces the strength, and this may then preclude
any applications of the tested material.
3.2. FeCo Alloys
and in high-accuracy positioning devices. Simi- tigue resistance is improved by ordering with
lar Fe – Co – V alloys with more V and less Fe only a small effect of the environment. The
are known as Vicalloy alloys. low ductility can be increased by adding Ni as
The B2 compound FeCo is stable only at tem- fourth element, which changes the microstruc-
peratures below 730 ◦ C and has a broad range ture, and still more by thermomechanical treat-
of homogeneity. At 730 ◦ C there is an order – ment, which results in a Hall – Petch-type de-
disorder transition with 0.2 % change in volume pendence of ductility on grain size or subgrain
from B2 to A2 (bcc), and a further structure size. Even cold rolling can reduce the brittle-
change from A2 to A1 (fcc) occurs at about ness. The brittle-to-ductile transition tempera-
985 ◦ C. The ordering reaction proceeds het- ture in the ordered state can be reduced signifi-
erogeneously or homogeneously, depending on cantly by deviations from stoichiometry, that is,
temperature. The ordering kinetics are affected by decreasing the Co content, which decreases
by alloying additions, and in particular alloy- the critical temperature of ordering. Positive ef-
ing with 2 atom % V reduces the ordering rate fects on cold workability have been produced
significantly and allows the freezing of the high- — apart from alloying with V or Cr which both
temperature disordered structure by quenching slow down the ordering kinetics — by alloying
to low temperatures. The addition of more than with V, Nb, Ta, Cr, Mo, W, and C, whereas Be,
2 % V leads to alloy decomposition, as is indi- Ti, Zr, Mn, Cu, Ag, Au, B, Al, and Si are inef-
cated by the Fe – Co – V phase diagram. A still fective.
larger effect on the ordering kinetics is produced Alloying of Fe and Co with precious met-
by alloying with Cr. Below the critical temper- als leads to the compounds FePt, CoPt, FePd
ature of ordering the degree of long-range order with L10 structure, which is an ordered fcc
at equilibrium increases continuously from 0 at structure with tetragonal distortion. These com-
730 ◦ C to 0.9 at 500 ◦ C and approaches 1 (per- pounds have been of importance for applications
fect order) asymptotically at low temperatures. as permanent-magnetic materials, and in par-
The mechanical behavior of FeCo has been ticular CoPt, which is used in the partially or-
studied intensively including creep, fatigue, and dered state, excels because of isotropy, ductility,
fracture. At room temperature FeCo is ductile, easy machinability, and resistance to corrosion
with 15 % uniform elongation and subsequent and high temperatures (see also [12]). However,
necking, in the disordered state, that is, after CoPt alloy magnets are seldom used now since
quenching from above the ordering temperature, these alloys have been replaced by rare earth
whereas in the ordered state, that is, after an- magnets with superior magnetic properties.
nealing below the ordering temperature, it shows
only little ductility with 5 % uniform elongation
and no necking. Hence cold rolling is difficult in 3.3. Sendust Alloy
the ordered state. Furthermore, FeCo is subject
to hydrogen embrittlement at room temperature. The iron aluminide Fe3 Al with DO3 structure
The flow stress increases with decreasing de- exhibits an outstanding high magnetic perme-
gree of order, that is, with rising temperature, ability which makes it useful as a magnetic
to reach a maximum at about 700 ◦ C just below material. The magnetic Fe3 Al-based alloys are
the ordering temperature when FeCo is only par- known as Alfenol alloys [12]. Fe3 Al is formed
tially ordered. Above the flow stress peak tem- on cooling by ordering reactions in the solid state
perature, normal softening is observed due to that transform the bcc disordered solid solution,
thermal activation. In other words, FeCo, like which is stable above about 800 ◦ C, first into the
other intermetallics, exhibits an anomalous pos- FeAl compound with B2 structure, which is sta-
itive temperature dependence of flow stress. ble between about 800 ◦ C and 550 ◦ C, and then
Ordered FeCo – 2V exhibits more homoge- into Fe3 Al with the D03 structure. The magnetic
neous deformation and a higher strain-hardening properties can be improved by partial substitu-
rate than disordered FeCo – 2V at all temper- tion of Al by Si, which does not change the crys-
atures. The yield stress and fracture stress in- tal structure. This gave rise to the development of
crease with decreasing grain size according the Sendust alloy which is based on Fe3 (Al,Si)
to the usual Hall – Petch relationship. The fa-
10 Intermetallics
difficult, and their corrosion resistance is low, structures. The A15 structure may be regarded
which necessitates protective coatings in corro- as the basis of the family of compounds with tcp
sive environments. SmCo5 suffers an environ- structures which are known as Frank – Kaspar
mental degradation of coercivity due to hydro- phases.
gen absorption. The hydrogenation of the Nd – The compounds with A15 structure form a
Fe – B alloys leads to structure transformation large and comparatively homogeneous group of
and decomposition reactions and is used for pro- intermetallic compounds and are of enormous
ducing fine-grained Nd – Fe – B powders. Be- practical importance since many of them are su-
cause of their most advantageous combination perconductors (→ Superconductors, Chap. 3.1).
of properties, the rare earth magnets have found Some of them, namely, V3 Ga, V3 Si, Nb3 Al,
a huge market with a broad field of applications. Nb3 Ga, Nb3 Ge, and Nb3 Sn, exhibit critical tem-
Research and development is rapidly progress- peratures between 15 and 23 K which are only
ing [10, 13]. 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-
4. Superconducting Materials cal current density, and the best combination of
critical temperature, critical current density, and
4.1. Laves Phases critical magnetic field strength is found in the
A15 superconductors, which have found wide
A transition to superconductivity at low tem- application [12, 14].
peratures has been observed for a large num- The processing of the A15 compounds is dif-
ber of Laves phases with C14 or C15 structure. ficult because of their brittleness. In the follow-
(Hf,Zr)V2 with C15 structure is of particular in- ing, various A15 compounds, which have al-
terest because of its advantageous superconduct- ready been applied or have been considered for
ing properties, since it combines a fairly high application, are briefly overviewed.
critical temperature with a high critical current
density and critical magnetic field strength. Pro-
4.2.1. V3 Si
cessing is difficult because of its poor deforma-
bility. Further alloying, in particular with Nb, V3 Si was the first A15 compound discovered to
improves the mechanical behavior to such an be superconducting, with a critical temperature
extent that multifilament superconductors can of 17 K. Stoichiometric V3 Si melts congruently
be produced which are considered for appli- at 1925 ◦ C and is stable at high and low tempera-
cation in fusion reactor magnets since the su- tures. A deviation from stoichiometry is possible
perconducting properties are comparatively in- on the Si-poor side, but reduces the critical tem-
sensitive to neutron irradiation and mechanical perature due to the resulting constitutional disor-
strain. Room-temperature deformation of HfV2 - der. The critical temperature is also lowered by
base alloys occurs primarily by twinning. The Hf dissolved hydrogen. Likewise the critical cur-
– V – Nb phase diagram is available. rent density and the critical field strength are re-
duced by elastic tensile or compressive strain.
Slightly above the critical temperature, V3 Si
4.2. A15 Compounds transforms martensitically into a weakly tetrag-
onal, still superconducting compound with twin-
The A15 crystal structure is a complex cubic related thin lamellae. Prior plastic deformation
structure derived from the bcc lattice (A2) with with large strains at high temperatures stabilizes
eight atoms per unit cell. It is a typical example the tetragonal martensitic phase even at temper-
of the topologically close packed (tcp) structures atures above the martensite formation tempera-
which result from the stacking of polyhedra of ture. The martensitic transformation gives rise
various shapes to accommodate atoms of dif- to superelasticity and a shape-memory effect.
ferent sizes. These tcp structures allow a closer V3 Si has a high Young’s modulus of 213 GPa
packing of atoms of different sizes than the ge- and an apparent brittle-to-ductile transition tem-
ometrically close packed fcc (A1) and hcp (A3) perature (BDTT) of the order of 1200 ◦ C. The
12 Intermetallics
plastic deformation behavior above the BDTT stability, which is advantageous for fabrication,
has been studied in detail. The creep behavior is and the high critical temperature of about 16 K,
characterized by a comparatively high activation V3 Ga is of interest for applications.
energy in the range of 2 – 11 eV per atom that The superconducting properties and the de-
depends on stress and composition. Creep can formation behavior of V3 Ga are similar to those
be enhanced by superimposed electric currents. of V3 Si. The superconduction transition can be
The transition from superconduction to nor- affected by prior plastic deformation. V3 Ga is
mal conduction in A15 superconductors oc- ductile in compression above 1000 ◦ C and the
curs by the penetration of small fluxoids, that creep behavior has been studied in detail. As in
is, quantized magnetic flux vortices, which is the case of V3 Si, V3 Ga can be deformed plas-
characteristic for the type II high-field super- tically by hot extrusion and at room tempera-
conductors. These fluxoids can be pinned by ture by superimposing a high hydrostatic pres-
microstructural inhomogeneities, in particular sure which makes the production of multifila-
grain boundaries and precipitated particles, and ment superconducting wires feasible for appli-
this is beneficial with respect to critical current cations in solenoid magnets. Hybrid high-field
density and field strength. Consequently, mi- magnets with V3 Ga tape coils have been pro-
crostructural changes by prior plastic straining duced successfully. However, Nb3 Sn has been
with or without recrystallization affect the super- found more favorable with respect to super-
conducting properties. The thin lamellae, which conducting properties, mechanical behavior, and
are produced by the above-mentioned marten- processing, and this has precluded the commer-
sitic transformation, are less effective with re- cialization of V3 Ga up to now.
spect to flux pinning than grain boundaries, so
that their effect on the superconducting proper-
ties becomes relevant only in single crystals. 4.2.3. Nb3 Sn
V3 Si is of interest for applications because
of its favorable superconducting properties and Nb3 Sn with A15 structure forms peritectically
its high stability. Apart from hot extrusion, V3 Si at 2130 ◦ C from the liquid solution and solid
can be deformed plastically at room temperature Nb and is stable only above 775 ◦ C where it de-
in spite of its brittleness by superimposing a high composes eutectically. The decomposition reac-
hydrostatic pressure, which makes the produc- tion can be suppressed by not too slow cooling,
tion of multifilament superconducting wires fea- so that Nb3 Sn with A15 structure is maintained
sible for applications in solenoid magnets. Lam- at lower temperatures in a metastable condi-
inar V3 Si can be prepared by solid-state thin- tion. Nb3 Sn transforms into a twinned tetragonal
film reactions between V and SiO2 on Si sub- martensite at 43 K, and the transition to super-
strates. However, Nb3 Sn has been found more conductivity occurs at about 18 K. This critical
favorable with respect to superconducting prop- temperature is lowered by deviations from stoi-
erties, mechanical behavior and processing, and chiometry, which affect the atomic order, and by
this has precluded the application of V3 Si. elastic strain. Small increases in the critical tem-
perature are possible by addition of ternary al-
loying components. The transition temperature
4.2.2. V3 Ga does not depend sensitively on prior plastic de-
formation.
V3 Ga with A15 structure is stable only up to Like the other A15 compounds, Nb3 Sn can
1300 ◦ C where it transforms congruently into a be deformed readily at high temperatures, e.g.,
V – Ga disordered bcc solid solution. There are by hot extrusion, whereas deformation at low
indications for a martensitic transformation at temperatures is possible only by superimposing
very low temperatures, in close analogy to V3 Si, a high hydrostatic pressure. The high tempera-
although the conditions for martensite formation ture deformation and creep behavior has been
are not yet clear for V3 Ga. It has a broad range of studied in detail, and a beneficial effect of grain
homogeneity centered around the stoichiometric size reduction on the brittle-to-ductile transition
composition, that is, it is very stable with respect temperature (BDTT) has been found.
to deviations from stoichiometry. Because of this
Intermetallics 13
Nb3 Sn superconductors with multifilament the other A15 compounds, but in a less sensitive
geometry, which are needed for electromagnet way. The A15 compound with the highest transi-
applications, are easily fabricated by the so- tion temperature is Nb3 Ge, which can be alloyed
called bronze process which has lead to the with Nb3 Al to form the ternary A15 compound
widespread commercial application of Nb3 Sn. Nb3 (Al0.7 Ge0.3 ) with the highest critical field
In the bronze process Nb rods are inserted in strength.
a bronze rod or tube which is then processed Nb3 Al has long been considered for electro-
to obtain a multifilament wire. After wire pro- magnet applications because of its advantageous
cessing, Nb3 Sn shells are formed around the Nb superconducting properties. However, the pro-
cores by reaction heat treatments. The final mul- cessing of Nb3 Al is very difficult because of
tifilament composite superconductor consists of its brittleness of Nb3 Al. Powder metallurgy has
Nb3 Sn fibers with Nb cores which are embed- been used successfully for preparing Nb3 Al by
ded in a bronze matrix; the latter is surrounded reaction synthesis and producing, e.g., a Nb3 Al
by a coating of Nb, which serves as a diffusion – Ag composite superconductor. Nb3 Al tapes
barrier, and an outer Cu tube, which serves as a have been prepared by rapid solidification. A
stabilizer and carries the current in the case of novel process with thin Nb – Al sandwiches as
transition to normal conduction. Various specific starting material has allowed the fabrication of
processing routes, including powder metallurgy, Nb3 Al superconducting wires with good worka-
have been developed for optimized supercon- bility. Furthermore various other processes have
ductor production, and hybrid superconductor been used successfully for producing Nb3 Al
coils are commercially available. wires.
The superconducting properties of the com-
posite superconductor as well as the mechanical
properties depend on the particular microstruc- 4.2.5. Nb3 Si
ture which is a function of the various prior pro-
cessing steps. Much work has been directed at Nb3 Si is a high-temperature line compound
the details of the various processes, e.g. the ef- which forms with a tetragonal crystal structure
fects of alloying additions, the diffusion of Sn by a peritectic reaction at 1980 ◦ C and is sta-
during Nb3 Sn formation, the grain size depen- ble only above 1770 ◦ C. Rapid solidification of
dence of the flux pinning force and the tensile the liquid produces amorphous Nb3 Si. The A15
strength and fracture at room temperature, or structure is unstable and is obtained only by spe-
the effects of the hardnesses of the constituent cial rapid solidification techniques and by crys-
phases on the composite workability. Recent tallization of amorphous Nb3 Si. Nb3 Si with A15
new developments rely on the beneficial effects structure is of interest because a high transition
of additions of Ti, Hf, Ta, and/or Ge on the su- temperature for superconduction is expected,
perconducting properties. and indeed a transition temperature of 19 K was
reported for near-stoichiometric Nb3 Si with A15
structure, which was prepared by crystallization
4.2.4. Nb3 Al of amorphous Nb3 Si under high pressure.
most eminent example is Pd2 Si which is used Schottky field effect transistors (MESFETs), the
for shallow contacts with very small controlled high electron mobility transistors (HEMTs), the
thickness and extension. The kinetics of thin- heterojunction bipolar transistors (HBTs), the
film formation and growth are controlled by dif- permeable-base transistors (PBTs), the static in-
fusion and can be optimized by alloying with ductance transistors (SITs), and the metal insula-
further elements. The film-formation process af- tor semiconductor field effect transistors (MIS-
fects the distribution of dopants in the Si sub- FETs).
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 6. Electric Materials
form on the silicide films by air exposure at room
temperature, which can be a problem for Ohmic The transition metal disilicides MSi2 (M = Ti,
contacts or a necessity for a tunneling device. Zr, Hf, V, Nb, Ta, Cr, Mo, W (→ Silicon,
The transition metal silicides CrSi, MnSi, Chap. 5) crystallize with the tetragonal C11b
FeSi, and CoSi crystallize with the cubic B20 structure, the hexagonal C40 structure, or the
structure, whereas NiSi and the noble metal sili- orthorhombic C49 and C54 structures and show
cides PtSi, IrSi, and PdSi crystallize with the extended mutual solid solubilities. These crys-
orthorhombic B31 structure. Such MSi com- tal structures are closely related since they only
pounds, in particular PtSi and also NiSi, are also differ in the stacking of the same structural ele-
important thin-film materials for applications in ments.
electronic devices. MoSi2 with C11b structure has found wide-
The disilicides CrSi2 , VSi2 , NbSi2 , and TaSi2 spread application as heating elements in high-
crystallize with the hexagonal C40 structure. temperature furnaces at up to 1700 ◦ C because
VSi2 is of interest for applications in electronic of its advantageous electrical properties and its
devices. TaSi2 has been successfully used in pro- excellent oxidation resistance. Well-known ex-
ducing microprocessors and other electronic de- amples are Kanthal Super and Mosilit heaters.
vices. CrSi2 is of interest for applications as thin- The high oxidation resistance of these materi-
film interconnectors. CoSi2 with the compara- als, which is not shown by the other Mo-rich
tively simple cubic C1 structure is applied in silicides, is due to the formation of a protective
large-scale integrated electronic devices. NiSi2 vitreous highly adherent SiO2 film with prior
is another compound of primary interest for ap- evaporation of the volatile Mo oxide. However,
plications in electronic thin-film devices. MoSi2 MoSi2 is subject to oxidative disintegration by
is advantageous for applications in electronic de- pest oxidation at intermediate temperatures bet-
vices either as substrate for Si films or as thin- ween about 300 ◦ C and 600 ◦ C, which occurs
film interconnectors in VLSI circuits. In view of at grain boundaries and microstructural defects
these applications, the conditions for thin-film and thus can be minimized by avoiding these or
formation as well as for the formation of thin by alloying with small amounts of Fe and Re.
oxide layers have been studied in detail.
7. Thermoelectric and
5.2. Other Compounds
Thermomagnetic Materials
Various aluminides, such as Al3 Ti, Al3 Hf,
Al3 Ta, and Al7 Cr, are used as shunting layers 7.1. Silicides
in integrated circuits to improve the resistance
to failure of Al-based interconnects by electro- The dimagnesium monosilicide Mg2 Si is a semi-
migration stressing [17]. conductor with a high thermoelectric power and
GaAs, InP, InSb, and InAs are compounds a low thermal conductivity which is advanta-
with high electron mobilities and high satura- geous for thermoelectric power generation. The
tion velocities relative to Si and hence are very thermal conductivity can be minimized by al-
attractive materials for application in high-speed loying with Ge to form the ternary Mg2 (Si,Ge)
devices [18]. Such devices are the GaAs metal compound. Doping with Ag and Sb produces p-
Intermetallics 15
and n-type semiconducting Mg2 (Si,Ge), respec- deviation from stoichiometry and doping with
tively, and the combination of both results in a halogens as donors and alkali metals as accep-
thermocouple with a high efficiency for thermo- tors, respectively.
electric energy conversion. The interactions bet- Research and development are leading to
ween doping, point defects, and microstructure novel complex alloys with even more attractive
has been studied in detail. properties [21, 22].
FeSi2 is another disilicide with a high ther-
moelectric power and has been proposed for ap-
plication in a thermoelectric power generator. It 7.3. Thermomagnetic Compounds
has an orthorhombic crystal structure and is sta- [23 – 27]
ble only at lower temperatures, whereas at higher
temperatures it decomposes to form a two-phase Magnetization of a material by application of a
FeSi – Fe2 Si5 alloy. Furthermore, it is subject to strong magnetic field produces magnetic order
pest oxidation. Doping with Al produces p-type with electron spin alignment and is accompanied
conducting FeSi2 , doping with Co produces n- by heat release. Without the magnetic field the
type FeSi2 , and the combination of both materi- material returns to magnetic disorder with heat
als gives a thermocouple which can be used as a consumption from the surroundings. This mag-
thermoelectric power generator. Such a power- netocaloric effect is used for magnetic refriger-
generation system for the efficient conversion ation, which is more efficient than conventional
of low-grade heat into electricity has been pro- gas compression refrigeration. For this purpose
posed [12, 19]. magnetic refrigerant materials with high mag-
The monosilicides MnSi and CoSi exhibit netic moments are needed, since the amount of
large thermoelectric powers and thus are promis- produced or consumed heat is proportional to
ing for thermoelectric power generation. In the magnetic moment apart from the magnetic
particular, a CoSi – CrSi2 thermogenerator has field.
been proposed. Intermetallic rare earth metal compounds
The complex manganese silicide Mn11 Si19 with transition metals show very large magnetic
is promising for thermoelectric applications be- moments and have successfully been used for
cause of its anomalously high anisotropy of the magnetic refrigeration. Present developments
thermopower over an exceptionally wide tem- concentrate on optimizing such rare earth metal
perature range [20]. phases with respect to magnetic moment and
temperature of operation. An outstanding ex-
ample is the phase Gd5 Si2 Ge2 with a magne-
7.2. Other Thermoelectric Compounds tocaloric effect which is twice as large as that
of Gd [25]. The operating temperature of this
Thermoelectric materials on the basis of the
material can be tuned from about −240 to about
semiconducting compound Bi2 Te3 are widely
+20 ◦ C by adjustment of the Si/Ge ratio.
used as thermoelectric coolers (TECs) for opera-
The materials developments aim at magnetic
tion near room temperature in small refrigerators
refrigerators which can be applied, e.g., as small
[20] (→ Refrigeration Technology, Chap. 1.4).
cryocoolers for cooling superconductive devices
Such materials with the rhombohedral C33
to near liquid helium temperatures or as large
structure are solid solutions of Bi2 Te3 with the
commercial liquefiers for liquefying cryofuels.
isomorphous compounds Sb2 Te3 and Bi2 Se3 .
They easily cleave along the basal planes and
show anisotropic mechanical and transport prop-
erties. 8. Optical Materials
Thermoelectric materials for application at
medium temperatures are based on the com- Various intermetallics show beautiful colors
pound PbTe with simple cubic B1 structure, which may be attractive for applications in jew-
which forms solid solutions with the isomor- elry. An early example is the Regulus of Venus
phous compounds PbSe, PbS, SnTe, and GeTe of the alchemists with its bright violet color,
[20]. n- and p-type materials are prepared by which results from the reduction of Sb2 S3 , the
16 Intermetallics
of the B element, primarily by Fe, Co, Ni, Cu, development of La0.8 Nd0.2 Ni2.5 Co2.4 Si0.1 or
Mn, Cr, substitution of the A element, primar- MmNi3.5 Co0.7 Al0.8 (Mm = mixed metal). The
ily by Ti, or any combination of these alloying permanent-magnet compound SmCo5 can also
possibilities. absorb large quantities of hydrogen which, how-
An example for the stoichiometry effect is ever, reduces its coercive force. CaNi5 , which is
given by ZrMn2 which exhibits an increasing isostructural with LaNi5 , shows an analogous
hydrogen vapor pressure with increasing Mn behavior and is considered as a low-cost candi-
content and which is stable with C14 struc- date for hydrogen storage.
ture up to ZrMn3.8 . Another example is C14 Research and development aim at producing
TiMn2 , which reaches its maximum H solubil- nanostructured materials, which can be obtained
ity at the off-stoichiometric composition corre- by powder-metallurgical processing through hy-
sponding to TiMn1.5 , whereas it does not ab- drogenation [32]. Such materials promise more
sorb H at the stoichiometric composition. The advantageous storage properties because of their
effects of substitution on electronic structure, much-increased internal interface area and de-
alloy stability, hydride stability and H storage fect density, as demonstrated for Mg2 X phases
capacity have been studied for various sys- (X = Cu, Ni, Si, Ge, Sn) [37, 38]. The prevail-
tems, e.g., (Zrx Ti1−x )Mn2 , Zr(Mn1−x Fex )2+y , ing problem is the hysteresis accompanying the
Zr(V1−x Cox )2 , in which both V and Co are hydriding and dehydriding processes, which re-
substituted for by other transition metals, and flects irreversibility and contributes to the degra-
Zr(Cr1−x Vx )2 . Zr(Cr1−x Vx )2 exhibits polytyp- dation of the absorption properties during long-
ism, that is, crystal structure variants of either the term service [31].
hexagonal C14 structure or the cubic C15 struc-
ture are formed, which only differ in the stacking
of crystal structure units. The thermodynamics 10. Electrode Materials
of hydrogenation have been studied, and hydride
formation enthalpies determined. The optimiza- 10.1. Mg2 Ni Alloys
tion of properties has led to multinary Laves
phases such as Zr – Mn – Co – V alloys, Zr – Mg2 Ni-base hydrogen-storage alloys are of ma-
Mn – Ni – V alloys, Zr – Mn – Ni – V – Cr al- jor interest for application as active materials of
loys, or Zr – Ti – Mn – Fe and Zr – Ti – V – Fe metal hydride electrodes. Such materials can be
alloys. prepared by powder metallurgy, and their prop-
In summary, the Laves phase materials are erties have been optimized by alloying with fur-
advantageous because of their good hydrogen ther elements [39]. Both partial substitutions of
capacities in the temperature range 0 – 100 ◦ C Ti or Ce for Mg and of Mn or Co for Ni improve
and low costs. discharge capacity and cycle life.
effect due to the formation of thermoelastic Ni content from about − 240 ◦ C for 60 atom %
martensite. The ternary compound has given Ni to about 1000 ◦ C for 70 atom % Ni. Thus,
rise to the development of the Cu – Al – Ni the shape-memory effect can be produced at
shape memory alloys with about 11 – 15 wt % high temperatures, and this allows the devel-
Al and 3 – 5 wt % Ni. The deformation modes of opment of high-temperature shape-memory al-
the martensites include various twinning modes loys. The martensitic transformation of Ni-rich
and stress-induced transformations. The shape NiAl can be induced by external applied stresses;
memory effect is a two-way effect, that is, it is it is thermoelastic, superelasticity has been ob-
completely reversible. The martensite-start tem- served, and the martensite deforms by twinning
perature has an upper limit of about 180 ◦ C and and detwinning. The transformation character-
can be lowered to a large degree by increasing istics do not depend on the processing route.
the Al content and to a smaller degree by in- The problem of the low room-temperature duc-
creasing the Ni content. tility of NiAl has been overcome by alloying
The Cu – Al – Ni alloys are advantageous be- with third elements, in particular Fe, to produce
cause of their higher stability at higher temper- a ductile second phase with fcc structure.
atures compared with the Cu – Zn – Al alloys.
However, second-phase precipitation, which
embrittles the Cu – Al – Ni alloys and precludes 12.3. NiTi Alloys
cold working, cannot be suppressed; hence, such
alloys can only be hot finished. Thermomechan- The intermetallic compound NiTi with B2 struc-
ical treatments and microalloying additions, in ture has been used since about 30 years as
particular Mn, Ti, Zr, are used for improving shape-memory alloy for couplings, fasteners,
the mechanical behavior. An important effect is connectors, and actuators in the automotive
grain refinement, which increases both fracture and aerospace industries, electronics, mechan-
strength and fracture strain according to a Hall – ical engineering, and medical applications. NiTi
Petch relationship. Indeed precipitated particles is known as Nitinol, which refers to Ni and Ti and
have been found to reduce the as-cast grain size the US Naval Ordnance Laboratory where this
and hinder later grain growth during annealing. material was developed. NiTi melts congruently
The alloying additions affect the transformation at 1310 ◦ C and has an extended homogeneity
and decomposition behavior, and this must be range.
taken into account in establishing the optimum NiTi transforms martensitically from the par-
conditions for the induction of the two-way ef- ent B2 phase to a monoclinic martensite phase
fect. Recently, the ductilizing effect of small with an intermediate orthorhombic R phase,
boron additions was discovered and studied in from which a reversible shape-memory effect
detail with respect to transformation behavior with the best shape-memory behavior of all
and microstructure. shape-memory alloys and pseudoelasticity re-
Other similar Cu-base alloys also show the sults. The characteristics of such shape-memory
shape-memory effect. Well-studied examples alloys have been studied in detail with respect to
are the Cu – Al – Mn shape memory alloys, crystallography, microstructure, thermodynam-
which order by a two-step reaction on rapid cool- ics, kinetics, and macroscopic mechanical be-
ing, that is, the initial disordered A2 structure is havior, and applications have been described.
transformed first into the intermediate B2 struc- The shape-memory behavior of NiTi is im-
ture and then into the final L21 structure. proved by prior thermomechanical treatments
which produce advantageous microstructures
and textures. The transformation temperature is
12.2. NiAl Alloys about 110 ◦ C for stoichiometric NiTi and is de-
creased by excess Ni. Likewise the transforma-
Al-deficient NiAl with B2 structure is trans- tion temperature is lowered by alloying with
formed martensitically by rapid cooling. This Fe and Cr, whereas Cu decreases the trans-
transformation is thermoelastic and produces the formation temperature hysteresis. Recently a
shape-memory effect. The martensite formation widened transformation temperature hysteresis
temperature increases linearly with increasing
Intermetallics 21
was found for Ni – Ti – Nb alloys which is at- pedic implants, nails, screws, plates for fixing
tractive with respect to coupling and sealing ap- bones, artificial hip joints, heart valve prosthe-
plications. The transformation temperature is in- ses, and stents. Problems are posed by the nec-
creased to 500 ◦ C by the partial substitution of essary biocompatibility, which is controlled pri-
Ni by Pd, which, however, affects ductility. marily by corrosion resistance and physical alloy
The elastic and plastic deformation behav- properties. The corrosion resistance increases
ior, including mechanical twinning, of NiTi al- with the transition from steels to Co – Cr alloys
loys has been studied intensively. Noteworthy and further to Ti alloys which, however, is par-
is the high ductility of NiTi, which allows de- alleled by respective cost increases.
formations of up to 70 %. Furthermore the yield In view of the advantageous corrosion behav-
strength of NiTi depends on temperature in an ior of the Ti alloys, intermetallic Ti – Al alloys
anomalous way, that is, it increases with increas- (see Section 16.1.2) have recently attracted ma-
ing temperature to reach a maximum, and only jor interest for endoprosthetic applications, such
above this peak temperature does normal soft- as hip and knee prostheses and dental implants
ening occur. [72].
Processing of NiTi alloys is difficult and must Besides the presently used biomaterials, the
be done with care to obtain the advantageous intermetallic phase NiTi known as Nitinol is ex-
properties. Melting must be carried out in vac- tensively used for a wide variety of applications
uum or under inert atmosphere because of the because of its outstanding attractive properties
reactivity of Ti, and hence vacuum-induction, [73 – 75]. In addition to being biocompatible,
plasma-arc, and electron-beam melting are all this material is superelastic and shows the shape-
used commercially. Hot working is possible memory effect (see Section 12.3). The extremely
without problems. Processing at low tempera- good biocompatibility is due to the formation
tures is difficult because of extremely high work of a passive TiO2 layer. Since NiTi is not fer-
hardening, and intermediate annealing treat- romagnetic, NiTi devices are compatible with
ments are required for cold rolling, machining, magnetic resonance imaging (MRI).
etc. The ductility problem of the (Ni,Pd)Ti al- Superelastic NiTi wires are used for connect-
loys with high transition temperature is relieved ing bone fragments and fixing small bone frag-
by alloying with B. Joining by welding, braz- ments. Other applications are filters for large
ing, or soldering is generally difficult. The ex- embolized blood clots. The superelasticity gives
cellent shape-memory behavior of the NiTi al- rise to an enormous flexibility which is made
loys is accompanied by an excellent corrosion use of for producing medical instruments, e.g.,
resistance, comparable to stainless steels, which superelastic NiTi catheters. The most celebrated
makes them the only shape-memory alloys suit- application is the self-expanding stent, which re-
able for implantation in the human body. Re- lies on knitted or welded NiTi wire. The long-
cently NiTi has been used as strengthening phase term stability of such devices is still the subject
in Al matrix composites, which experience an of discussion, and in particular there is the ques-
additional strengthening effect due to the shape- tion whether toxicity may arise after many years
memory effect of NiTi. of service.
These amalgams are formed by alloying Ag – Sn ally known as long-period superlattice (LPS). At
– Cu alloys on the basis of the intermetallic com- the equiatomic composition the tetragonal L10
pounds Ag3 Sn and Cu3 Sn with mercury. The phase CuAu I forms, and again an orthorhombic
amalgamation reaction results in the formation LPS is observed at intermediate temperatures,
of a complex multiphase alloy structure which which is known as CuAu II. All phases are sep-
contains the phases Ag2 Hg3 , Sn8 Hg, Ag3 Sn, arated from each other by two-phase equilibria.
and Cu6 Sn5 , depending on composition. The elastic behavior is known, as is the dif-
Processing involves mixing and compaction fusion behavior. The deformation behavior of
of powders and is comparatively simple and is Cu3 Au was studied in detail to clarify the ef-
analogous to mechanical alloying. The amal- fects of atomic order, and it was found in par-
gams must show a sufficient compressive ticular that order decreases the yield stress and
strength and creep resistance, since the oral tem- increases the work hardening rate while still re-
perature is only about 10 % below the melting taining high ductility. Inversely, heavy cold de-
temperature. The dental amalgams are very brit- formation destroys the order. The ordering reac-
tle which leads to failure with tensile or bend- tion produces antiphase domains which give rise
ing stresses in unsupported margins. The frac- to an additional strengthening effect in this and
ture processes are correlated with corrosion pro- similar phases; strength increases with decreas-
cesses which primarily attack the phases Sn8 Hg ing domain size. The domain size can be varied
and Cu6 Sn5 and lead to the release of mercury by prior processing, and it increases during age-
and SnO, respectively. Attempts to replace mer- ing in analogy to Ostwald ripening. The Portevin
cury in these amalgams have not yet been suc- – Le Chatelier effect with serrated yielding was
cessful. observed for both the ordered and disordered
The alloy systems of interest have been stud- state. Recovery and recrystallization have been
ied in detail with respect to phase reactions, mi- analyzed in detail.
crostructure evolution, mechanical behavior in-
cluding creep and fatigue, corrosion resistance
and clinical performance, and the processing 15. Wear-Resistant Materials
technologies are well established [76, 77].
Intermetallic compounds are attractive in tribo-
logical applications because of various advanta-
14.2. Cu – Au Alloys geous properties, such as hardness, thermal sta-
bility, and oxidation resistance, and because of
Au – Cu – Ag alloys on the basis of the in- their compatibility with metallic matrices [78].
termetallic compounds CuAu and Cu3 Au have Some of the oldest bearing alloys contain in-
found applications in dentistry because of their termetallic compounds as is exemplified by the
extremely high corrosion resistance, their ad- lead-based bearing alloys (babbits) with SbSn
vantageous mechanical properties such as high and possibly Cu6 Sn5 or the bronze bearing al-
strength and ductility, and their decorative gold loys with Cu32 Sn8 or Cu3 Al.
color. These alloys age-harden as a result of The ternary Laves phase MoCoSi, that is,
complex ordering and decomposition reactions Mo(Co,Si)2 , with hexagonal C14 structure has
by which the phases Cu3 Au I, CuAu I, CuAu II, given rise to the development of wear-resistant
and an Ag-rich α2 phase are formed, depending Co – Mo – Cr – Si alloys which contain large
on composition. volume fractions of the Laves phase in coarse
Cu3 Al I is the well-known Cu3 Al compound distribution together with a Co-rich phase and
with L12 structure, which is the prototype phase are known as Tribaloys. Apart from their high
for this crystal structure. Cu3 Au II forms at in- strength and hardness, they show a sufficient
termediate temperatures with off-stoichiometric fracture toughness of the order of 20 MN/m3/2 .
compositions and is described as either tetrag- These advantageous mechanical properties re-
onal or orthorhombic. In any case, the Cu3 Au sult in an excellent wear resistance which, in
II structure results from the L12 structure by combination with an excellent corrosion resis-
the introduction of equally spaced antiphase tance in various environments, makes the Trib-
boundaries (APBs). Such a structure is usu- aloys very appealing for applications in a broad
Intermetallics 23
range of temperatures and environments. These ture and a density of 4.2 g/cm3 , which make it
alloys have been studied in detail including most attractive for materials developments (→
corrosive wear testing, according to which the Titanium, Titanium Alloys, and Titanium Com-
Tribaloys compare favorably with conventional pounds, Chap. 11.2).
wear-resistant materials [78]. The low ductility The aim of the various materials develop-
of Tribaloys, which is due to the large amounts ments on the basis of Ti3 Al has been to im-
of Laves phase, can be increased by proper ad- prove both strength and ductility by alloying
justment of alloy composition and heat treatment with further elements and by controlling the mi-
[79, 80]. crostructure. The most effective element for im-
Apart from the Tribaloys, modified struc- proving ductility is Nb, and indeed Ti3 Al-base
tural intermetallic high-temperature alloys (see alloys with engineering significance, that is, α2
Chap. 16.1) have been proposed for applica- alloys and super-α2 alloys, contain 10 – 30 atom
tion as wear-resistant alloys. An example is the % Nb. Ti3 Al-base alloys readily dissolve oxy-
novel Nicralc alloys which result from alloying gen and hydrogen, which both produce embrit-
Ni3Al with C to increase the hardness and which tlement.
are intended as substitutes for high-tempera- The developments of the Ti3 Al-based alloys
ture abrasion- and erosion-resistant cobalt al- are on the brink of commercialization. They have
loys [81]. Other examples are superelastic NiTi been produced in ingots of up to 4500 kg and
(see Section 12.3) with hardening TiC disper- are available as cast shapes or wrought forms
soids [82] and MoSi2 – SiC [83] and Fe3 Al – including billets, bars, plates, and sheet. Like-
Fe3 AlC0.5 composites [84]. wise the powder metallurgy methods which have
been developed for conventional titanium alloys
are readily adapted and in addition other powder
16. Structural High-Temperature metallurgy methods, in particular reactive pro-
Materials cessing, have been applied to Ti3 Al-base alloys.
The alloys can be machined, sheet can be de-
The aluminides of titanium, iron, and nickel, formed superplastically and diffusion bonded to
as well as some silicides, have been selected form complex sheet parts, and the various manu-
as bases for developments of new structural facturing processes can be accomplished on con-
high-temperature alloys because of their high ventional equipment. The alloys can be joined
strengths at low and high temperatures, high successfully by both diffusion bonding and lin-
thermal stability, high oxidation resistance, ear friction welding.
and/or low density. The problem of insufficient
ductility was overcome by further alloying and
optimization of microstructure, which usually 16.1.2. TiAl Alloys
resulted in multiphase alloys [85]. Research and
development is in progress, and some develop- The titanium aluminide TiAl, often designated
ments are on the brink of commercialization. the γ-phase, crystallizes with the tetragonal L10
The basic properties of the intermetallic com- structure. The density of 3.76 g/cm3 is even
pounds in question and the respective phase di- lower than that of Ti3 Al and has made the ti-
agrams, as well as the physical, mechanical and tanium aluminides most attractive for materials
chemical properties of the respective developed developments (→ Titanium, Titanium Alloys,
or proposed intermetallic alloys, with emphasis and Titanium Compounds, Chap. 11.2). Single-
on powder metallurgy processing, are summa- crystalline TiAl shows an anomalous positive
rized in [4]. temperature dependence of the yield stress, that
is, the yield stress increases with increasing tem-
perature until a maximum is reached at about
16.1. Titanium Aluminide Alloys 600 ◦ C, and only above the peak temperature is
16.1.1. Ti3 Al Alloys a normal negative temperature dependence ob-
served.
The titanium aluminide Ti3 Al, often designated
as the α2 phase, shows the hexagonal D019 struc-
24 Intermetallics
The major problem for the application of TiAl and significantly increased Nb contents of about
is low ductility at room temperature. If the Al 5 – 10 atom % for relieving the ductility prob-
content is reduced below 50 atom % to form lems [90].
Ti3 Al as second phase besides TiAl, the resulting
intermetallic two-phase alloys show ductilities
of a few percent which are acceptable for appli- 16.2. Iron Aluminide Alloys
cations. 48 atom % Al is the optimum compo-
sition for maximum room-temperature ductility 16.2.1. Fe3 Al Alloys
of TiAl alloys. The aim of the various materials
developments on the basis of TiAl is to improve Fe3 Al is formed on cooling by ordering reac-
both strength and ductility first by controlling the tions in the solid state that transform the bcc
microstructure and second by alloying with fur- disordered solid solution, which is stable above
ther elements. Oxygen embrittles TiAl alloys, about 800 ◦ C, first into the FeAl phase with B2
as do other interstitial impurities (N, C, B). In structure, which is stable between about 800 ◦ C
addition, environmental embrittlement has been and 550 ◦ C, and only then into Fe3 Al with the
found for TiAl, as for other intermetallics, which D03 structure. The critical temperature of order-
is clearly caused by the presence of atmospheric ing is much lower than the melting temperature,
moisture. and this indicates weaker bonding between un-
The development of TiAl alloys is less ad- like atoms than in, for example, the nickel alu-
vanced than that of the Ti3 Al-based materials minides. The critical temperature of D03 order-
because of the more severe ductility problems. ing can be shifted to higher temperatures by as
However, the TiAl alloys are regarded as more much as 250 ◦ C by alloying with third elements,
promising, primarily because of the lower den- in particular Cr, Mo, Mn, Ti, and Si. Alloying
sity and the higher application temperatures, and with Ti is of particular interest in view of the
development activities are now concentrated on high solubility of Ti in Fe3 Al with beneficial
TiAl in various countries. Processing is more effects on strength and ductility [91]. Thermal
difficult compared with Ti3 Al alloys because of vacancies are formed readily because of a low
the higher brittle-to-ductile transition tempera- formation enthalpy, which is even lower than
ture, and hot working temperatures are outside the migration enthalpy, and consequently there
the range of conventional titanium processing is a high concentration of vacancies at thermal
equipment, which has limited the practical in- equilibrium. High concentrations of vacancies
got size to about 200 kg. Besides casting and hot formed at high temperatures are trapped even
working, powder metallurgy has been used for by slow cooling and contribute to strengthening.
component fabrication, though there are prob- Strength and creep resistance at high tempera-
lems of quality, availability, and cost of powder. tures can be increased significantly by proper
Cast TiAl alloys can be forged isothermally, alloying with precipitation of second phases, as
fine-grain powder metallurgy material can be has been exemplified by Fe3 Al with precipitated
formed superplastically, as well as special cast perovskite-type carbide and/or Laves phase par-
and thermomechanically treated material, sheet ticles [92, 93].
of 1 – 2 mm thickness can be produced by twin- Fe3 Al is considered for structural applica-
roll casting, special processes have been de- tions since it not only shows a higher strength
veloped for rolling, and machining is possi- than comparable iron alloys, but also a high
ble, as is joining by diffusion bonding and corrosion resistance in oxidizing and sulfidiz-
welding. With respect to turbine-engine appli- ing environments. Problems for application are
cations, component-fabrication techniques have the low ductility at ambient temperatures and
been demonstrated for airfoils and compressor the strength drop above the ordering tempera-
cases. Besides aerospace applications, interest is ture. The ductility of Fe3 Al is very sensitive to
focused on automotive engine applications such the environment, and environmental embrittle-
as valves [86, 87] and turbocharger rotors [88, ment can occur in the presence of moisture in the
89], which are nearer to realization. ambient atmosphere. Alloys have been produced
More recent materials developments concen- by both powder metallurgy and ingot metallurgy,
trated on Ti – Al alloys with reduced Al contents
Intermetallics 25
including arc melting, vacuum-arc remelting, in- 16.3. Nickel Aluminide Alloys
duction melting in air and vacuum, and elec-
troslag remelting. These developments, aimed 16.3.1. Ni3 Al alloys
at applications such as automotive exhaust sys-
tems and resistance heating elements, are close Ni3 Al is the most studied and best known in-
to commercialization. Other potential applica- termetallic compound because it has long been
tions are hot-gas filters in coal gasification plants used as strengthening phase in the superalloys
and chemical processing equipment. and because its ductility problems were over-
come by ductilization by microalloying with
boron. Hence, it can be produced and tested
16.2.2. FeAl Alloys without major experimental difficulties, and thus
it was preferred in the past for studying the be-
FeAl is closely related to NiAl (Section 16.3.2) havior of intermetallics. Ni3 Al, known as the
since both phases show the B2 structure and γ -phase, crystallizes with the cubic L12 struc-
a complete mutual miscibility. However, FeAl ture.
melts incongruently in contrast to NiAl, and its The outstanding feature of the plastic flow
melting temperature is lower, which indicates of Ni3 Al is the anomalous temperature depen-
a lower stability and atomic bonding strength. dence of the flow stress, which increases bet-
The tensile ductility at low temperatures of poly- ween room temperature and about 700 ◦ C with
crystalline FeAl is practically zero at the stoi- increasing temperature to reach a maximum, and
chiometric composition and increases with de- only at higher temperatures does normal soften-
creasing Al content, that is, increasing deviation ing occur. Polycrystalline Ni3 Al is brittle and
from stoichiometry. The reported ductilities of fails by intergranular fracture, whereas Ni3 Al
off-stoichiometric FeAl are obtained only with single crystals are highly ductile. The polycrys-
well-annealed specimens, that is, if the low-tem- tal brittleness of Ni3 Al is removed by microal-
perature deformation is preceded by annealing loying Al-deficient Ni3 Al with B.
at very low cooling rates. This behavior is sup- Exposure to air or oxygen affects the mechan-
posed to be due to retained excess vacancies, ical behavior of Ni3 Al, and various effects of en-
which increase hardness and yield strength and vironmental embrittlement have been observed.
decrease ductility and which are not equilibrated At about 800 ◦ C the ductility of ductilized Ni3 Al
after high-temperature annealing with normal shows a minimum which is more pronounced in
cooling rates. Anneals of several hundred hours air than in vacuum and which is significantly
at about 400 ◦ C are recommended for produc- alleviated by the addition of Cr. Environmental
ing ductility in off-stoichiometric FeAl. FeAl is embrittlement has also been observed at room
subject to environmental embrittlement, which temperature, especially in moist air, depending
is due to the dissolution of hydrogen in FeAl at on Ni3 Al composition and dopant content, and it
the crack tip and is reduced by increasing the was concluded that this embrittlement is caused
deformation rate sufficiently. by hydrogen formation. Accordingly, stress cor-
In view of its reported advantageous prop- rosion cracking has been observed in aqueous
erties and comparatively low density, FeAl is environments.
regarded as a potential structural material for Successful materials developments on the ba-
high-temperature applications, and processing sis of Ni3 Al have resulted in a series of closely
developments make use of both ingot and pow- related Ni3 Al alloys for high-temperature ap-
der metallurgy. FeAl alloy compositions have plications. These alloys, known as nickel alu-
been identified, e.g., for good weldability and minides or advanced aluminides, generally con-
for high corrosion and wear resistance in aggres- tain B at levels below 500 ppm and Al below
sive gasification environments. As in the case of the stoichiometric composition to give ductility
Fe3 Al, the high-temperature strength and creep at room temperature, as well as up to 5 atom
resistance of FeAl can be improved significantly % Hf, Zr, Ta, and Mo for improving strength
by proper alloying with precipitation of harden- at high temperatures and up to 10 atom % Cr for
ing second phases [94, 95].
26 Intermetallics
conversion systems, and petrochemical plants ductile Al matrix, which has given rise to a suc-
[97]. cessful alloy development.
For the strengthening of NiAl second phases The M5 Si3 compounds (M = transition
with not too low Al contents are preferred metal), which adopt a variety of complex crystal
with respect to density and oxidation resistance. structures, are usually very stable congruently
There are hard and brittle ternary Laves phases melting line compounds with melting tempera-
with Al, such as NbNiAl and TaNiAl, which may tures above 2000 ◦ C. The most attractive M5 Si3
be used for strengthening NiAl. A respective de- compound for structural high-temperature ap-
velopment has lead to a NiAl – Ta – Cr alloy plication is Ti5 Si3 with the hexagonal D88 struc-
which is not deformable at room temperature ture because of its high stability with a melting
but excels in high-temperature strength, oxida- temperature of 2130 ◦ C, its high strength and
tion resistance, and shock resistance [98]. The hardness, and its low density. A promising ma-
alloy can be produced by both ingot metallurgy terials development is based on the combination
and powder metallurgy, so that components can of Ti3 Al and Ti5 Si3 which has been further
be produced by investment casting, hot extru- alloyed with Nb. Another candidate phase for
sion, hot isostatic pressing (HIP), powder injec- high-temperature applications is Nb5 Si3 , which
tion molding, and isothermal forging. 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
16.4. Silicide Alloys are alloyed with further elements to achieve an
excellent balance of room-temperature tough-
Silicides (→ Silicon, Chap. 5) were considered ness, high-temperature creep performance, and
for high-temperature applications as early as 50 oxidation resistance [101, 102].
years ago because of their high strength and oxi- Presently, MoSi2 is considered for structural
dation resistance at high temperatures. Presently high-temperature applications, and various ma-
the need for new structural materials for the terials developments are in progress. The brittle-
highest temperatures has revived interest in sili- to-ductile transition occurs at about 1000 ◦ C or
cides for structural applications, and respective higher temperatures depending on microstruc-
developments are under way. ture and impurity content. The ductility and
Cr3 Si with the topologically close packed toughness of MoSi2 can be improved by spe-
cubic A15 structure is a candidate phase for cial processing routes, microalloying with C
high-temperature applications. The brittle Cr3 Si for avoiding SiO2 layers on grain boundaries
is alloyed with Cr to produce a ductile-phase- or insertion of a ductile second phase, usu-
toughened intermetallic composite [99]. Ni3 Si ally an Nb-rich phase. Toughness can also be
with the ordered fcc L12 structure has a very increased by embedding hard second phases,
high potential for structural applications because in particular ceramics, in a MoSi2 matrix to
of its advantageous mechanical properties and hinder crack growth. MoSi2 -based compos-
its outstanding corrosion resistance. Alloying of ites reinforced by TiC, TiCN, and TiB2 in-
Ni3 Si with Nb and Ti results in multiphase Ni3 Si deed showed improved hardness and bending
alloys with dispersed Ni3 Nb and Ni3 Ti precip- strength compared to monolithic MoSi2 [103].
itates which are most promising for high-tem- Most work has concentrated on the MoSi2 –
perature applications because of their favorable SiC system. Other systems of particular interest
mechanical and chemical properties [100]. are MoSi2 – TiC and MoSi2 – Al2 O3 , as well
Mg2 Si crystallizes with the fcc C1 structure, as the eutectics MoSi2 – Mo5 Si3 and MoSi2
and its density is only 1.88 g/cm3 . Mg2 Si has a – Er2 Mo3 Si4 and the ductile-phase-toughened
comparatively high strength and a low thermal MoSi2 – Nb/Ta and MoSi2 – Mo systems. More
expansion coefficient. However, its brittleness, recent developments concentrate on intermetal-
with a brittle-to-ductile transition temperature lic Mo – Si – B alloys consisting primarily of
of 450 ◦ C, precludes the application of single- the phases Mo3 Si and Mo5 SiB2 besides ductile-
phase Mg2 Si. Alloying with Al produces Al- phase-toughening Mo [104 – 106]. Such com-
Mg2 Si alloys with brittle Mg2 Si particles in a posites show attractive oxidation resistances,
fracture toughness, and creep strength, the bal-
28 Intermetallics
26. M. Ilyn et al.: “Magnetocaloric Properties of 39. J. Chen, P. Yao, D. H. Bradhurst, S. X. Dou, H.
the LaFe11.7Si1.3 and LaFe11.2Co0.7Si1.1 K. Liu: “Mg2 Ni-based Hydrogen Storage
Systems,” J. Magn. Magn. Mater. 290 (2005) Alloys for Metal Hydride Electrodes,” J.
712–714. Alloys Compd. 293 – 295 (1999) 675 – 679.
27. D. Vasylyev, J. Prokleska, J. Sebek, V. 40. B. Knosp, L. Vallet, P. Blanchard:
Sechovsky: “Thermal Properties of “Performance of an AB2 Alloy in Sealed
Er(Co1-xSix)(2) Compounds,” J. Alloys Ni-MH Batteries for Electric Vehicles —
Compounds 394 (2005) 96–100. Quantification of Corrosion Rate and
28. I. M. Wolff, M. B. Cortie: “The Development Consequences on the Battery Performance,” J.
of Spangold,” Gold Bull. 27 (1994) 44–54. Alloys Compd. 293–295 (1999) 770 – 774.
29. W. A. McGahan: “Magneto-Optical 41. A. Otto, V. Güther: “Development of Fast
Applications”, in J. H. Westbrook, R. L. Kinetics Metal Hydride Alloys and Battery
Fleischer (eds.): Intermetallic Compounds: Electrodes for High Power Applications,” J.
Principles and Practice, Vol. 2, Wiley, Alloys Compd. 293 – 295 (1999) 734 – 736.
Chichester 1995, pp. 435 – 452. 42. T. Sakai, I. Uehara, H. Ishikawa:
30. V. A. Yartys, I. R. Harris, V. V. Panasyuk: “Research-and-Development on Metal
“New Metal-Hydrides — A Survey,” Mater. Hydride Materials and Ni-MH Batteries in
Sci. (Engl. Transl.) 37 (2001) 219–240. Japan,” J. Alloys Compd. 293 – 295 (1999)
31. P. Dantzer: “Properties of Intermetallic 762 – 769.
Compounds Suitable for Hydrogen Storage 43. F. Zhan, L. J. Jiang, B. R. Wu, Z. H. Xia, X. Y.
Applications,” Mater. Sci. Eng. A Struct. Wei, G. R. Qin: “Characteristics of Ni/MH
Mater. 329 (2002) 313–320. Power Batteries and its Application to Electric
32. M. Okada, T. Kuriiwa, A. Kamegawa, H. Vehicles,” J. Alloys Compd. 293 – 295 (1999)
Takamura: “Role of Intermetallics in 804 – 808.
Hydrogen Storage Materials,” Mater. Sci. Eng. 44. J. M. Joubert, M. Latroche, A.
A Struct. Mater. 329 (2002) 305–312. Percheron-Guegan: “Metallic Hydrides II:
33. R. A. Varin, T. Czujko: “Overview of Materials for Electrochemical Storage,” MRS
Processing of Nanocrystalline Hydrogen Bull. 27 (2002) 694–698.
Storage Intermetallics by Mechanical 45. G. Marsh: “Fuel Cell Materials,” Mater. Today
Alloying/Milling,” Materials and 4 (2001) 20–24.
Manufacturing Process 17 (2002) 129–156. 46. P. K. Datta, J. S. Burnell-Gray, K. Natesan:
34. A. Zuttel: “Materials for Hydrogen Storage,” “Coating Technology,” in J. H Westbrook, R.
Mater. Today 6 (2003) 24–33. L. Fleischer (eds. ): Intermetallic Compounds
35. M. Conte, P. P. Prosini, S. Passerini: — Principles and Practice, Vol. 3 Progress,
“Overview of Energy/Hydrogen Storage: John Wiley & Sons, Chichester 2002, pp.
State-of-the-Art of the Technologies and 561–588.
47. J. R. Nicholls, R. Wing: “Advances in Coating
Prospects for Nanomaterials,” Mater. Sci. Eng.
Systems for Utility Gas Turbines,” in J.
B 108 (2004) 2–8.
Lecomte-Beckers, M. Carton, F. Schubert, P. J.
36. D. M. Kim, K. J. Jang, J. Y. Lee: “A Review
Ennis (eds.): Materials for Advanced Power
on the Development of AB(2)-Type Zr-Based
Engineering 2002 (Proc. 7th Liège
Laves Phase Hydrogen Storage Alloys for
Conference), Forschungszentrum Jülich,
Ni-MH Rechargeable Batteries in the Korea
Jülich, 2002, pp. I.57–I.71.
Advanced Institute of Science and
48. A. Matthews, J. H. Givens, W. L. Ensinger, S.
Technology,” J. Alloys Compd. 293 – 295
L. Rohde (eds.): Metallurgical Coatings and
(1999) 583–592.
Thin Films 1999 (Proc. 26th Int. Conf.
37. H. Y. Shao, Y. T. Wang, H. R. Xu, X. G. Li:
Metallurgical Coatings and Thin Films),
“Preparation and Hydrogen Storage Properties
North-Holland Elsevier Science, Amsterdam
of Nanostructured Mg2Cu Alloy,” J. Solid
2000.
State Chem. 178 (2005) 2211–2217.
49. J. R. Nicholls: “Designing Oxidation-Resistant
38. R. Janot, F. Cuevas, M. Latroche, A.
Coatings,” JOM J. Min. Met. Mat. 52 (2000)
Percheron-Guegan: “Influence of Crystallinity
28–35.
on the Structural and Hydrogenation 50. A. G. Evans, D. R. Mumm, J. W. Hutchinson,
Properties of Mg2X Phases (X=Ni, Si, Ge, G. H. Meier, F. S. Pettit: “Mechanisms
Sn),” Intermetallics 14 (2006) 163–169. Controlling the Durability of Thermal Barrier
30 Intermetallics
Applications,” J. Mater. Sci. 39 (2004) Reinforced by TiC, TiCN, and TiB2,” Mater.
3975–3984. Sci. Eng. A 396 (2005) 277–284.
100. K. Ohira, Y. Kaneno, T. Takasugi: 104. S. Nakatani, K. Yoshimi, S. Hanada, S. H. Ko,
“Microstructure, Mechanical Property and Y. H. Park: “Design of Mo-Si-B-O Pseudo
Oxidation Property in Ni3Si-Ni3Ti-Ni3Nb in-situ Composites for High Temperature
Multi-phase Intermetallic Alloys,” Mater. Sci. Applications,” in PRICM 4: Fourth Pacific
Eng. A 399 (2005) 332–343. Rim International Conference on Advanced
101. B. P. Bewlay, M. R. Jackson, J. C. Zhao, P. R. Materials and Processing, Vols I and II, 2001,
Subramanian: “A Review of pp. 971–974.
Very-High-Temperature Nb-Silicide-Based 105. J. H. Schneibel et al.: “Optimization of
Composites,” Metall. Mater. Trans. A 34 Mo-Si-B intermetallics,” in E. P. George, H.
(2003) 2043–2052. Inui, M. J. Mills, G. Eggeler (eds. ): Defect
102. C. L. Ma, J. G. Li, Y. Tan, R. Tanaka, S. Properties and Related Phenomena in
Hanada: “Microstructure and Mechanical Intermetallic, MRS, Warrendale 2003, pp.
Properties of Nb/Nb5Si3 In Situ Composites 53–58.
in Nb-Mo-Si and Nb-W-Si Systems,” Mater. 106. V. Behrani, A. J. Thom, M. J. Kramer, M.
Sci. Eng. A 386 (2004) 375–383. Akinc: “Microstructure and Oxidation
103. J. Meng, J. Lu, J. Wang, S. Yang: “Preparation Behavior of Nb-Mo-Si-B Alloys,”
and Properties of MoSi2 Composites 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
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.
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]
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
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].
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]
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
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.
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].
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].
[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 ).
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.
7.2.1. 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].
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.
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
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].
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.
Moreover (2 6) oligosaccharides can be clearly distinguished from (2 1) compounds and GFn molecules from Fn
molecules.
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).
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]
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
page 11 of 11
Iodine and Iodine Compounds 1
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
thiocyanate solution with ferric alum as the in- In Oklahoma, the United States, iodine oc-
dicator. Quantities of up to 0.05 µg of iodine can curs in brines with concentrations of 150 –
be determined by colorimetric methods. 1200 ppm in the Morrowan formation of the
Pennsylvanian age [12]. In California, brines
containing 30 – 70 ppm of iodine are found in
4. Occurrence the Monterey Formation of the Middle Miocene
age and in the Repetto Formation of the
According to the United States Geological Sur- Pliocene age. In Michigan, brines containing
vey, iodine is the forty-seventh most abundant about 30 ppm of iodine occur in the Sylvania
element in the earth’s crust, counting the rare Formation of the Devonian age. In playa lakes
earths as a single element. Small quantities of in the western United States, iodine occurs at
iodine are widely distributed in rocks, soils, un- a concentration of 14 ppm [13]. In Japan, the
derground brines, and seawater. major iodine production occurs in the south-
ern Kanto gas field that underlies the Chiba,
Minerals. Iodine is found in nature in Tokyo, and Kanagawa Prefectures. The primary
the following mineral forms: bruggenite, iodine reservoirs occur in the Kiwada, Otadai,
Ca (IO3 )2 · H2 O; dietzeite, Ca2 (IO3 )2 · CrO4 ; and Umegase Formations of the Pliocene to
iodoargyrite, AgI; iodoembolite, Ag (Cl, Br, I); middle Pleistocene. In the Miyazaki Prefecture,
lautarite [7789-80-2], Ca (IO3 )2 ; and marshite iodine and natural gas are produced from the
[24401-69-2], CuI. The most important iodine- late Pliocene Sadowara and Tonogori members
containing mineral is lautarite (anhydrous cal- of the Miyazaki group. Iodine concentrations
cium iodate) [9]. Iodate occurs in the minerals in the Sadowara gas field range between 59
bruggenite, dietzeite, and lautarite, in caliche and 101 ppm. Iodine is contained in the Tomi-
deposits (between 2 and 50 feet below ground gusuku Formation of the Shimajiri group of late
level) in the Atacama Desert of Chile’s Antofa- Miocene to early Pliocene. Iodine concentra-
gasta and Tarapaca Provinces. The iodate is be- tions in the southern Okinawa gas field range
lieved to have formed from oceanic emissions between 40 and 110 ppm. In the Niigata Pre-
during the late Tertiary and Quaternary epochs. fecture, iodine concentrations range between 7
The atmospheric photochemical oxidation of io- and 68 ppm in brines in the Niigata and Hachi-
dide to iodate is thought to have occurred in the mori gas fields. The Niigata basin is of the
troposphere and at ground level. Pliocene to Pleistocene age [14]. In Indonesia,
iodine occurs with trace amounts of bromine in
Seawater contains approximately 0.05 ppm brines associated with oil. The most important
of iodine [11]. The amount of iodine is related iodine-producing area is the Gujangon anticline
to the salinity. For example, the Red Sea con- of sandstone and diatomaceous marls of the Up-
tains 0.06 ppm and the Baltic Sea, 0.01 ppm. Io- per Pliocene, Kailibeng Formation. The brines
dine is concentrated by marine organisms, (e.g., were reported to contain about 50 ppm of io-
sponges and seaweed) and by fish, especially in dine [15]. Information on iodine production in
the thyroid gland and in liver oil, which contain other parts of the world is incomplete, but the
510 – 69 200 ppm [11]. The brown algae, Lami- former Soviet Union has reported iodine con-
naria and Fucus, which grow in temperate zones centrations of 67 ppm in brines associated with
of the Northern Hemisphere, have the highest oil fields at Lake Bejuk-Schor, near Baku in the
iodine content (0.0279 – 0.4535 wt % after dry- Azerbaidzhan area on the Caspian Sea. Brines
ing). The name kelp was originally applied to the in the Slavyansko – Troitsko area were reported
ash obtained by burning seaweed, but has been to have iodine concentrations of 80 – 120 ppm in
extended to include the seaweed itself, partic- the Neftechinski oil field near the Black Sea. At
ularly the varieties yielding potash, iodine, and the Nebit-Dag iodine – bromine plant in Turk-
other commercially important products. menistan, iodine is produced from brines that
contain 250 g/L of a combination of boron, cal-
Brines. Iodine occurs frequently as potas- cium, magnesium, sodium compounds, and io-
sium or sodium iodide in subsurface brines as- dine [16].
sociated with oil and gas deposits.
4 Iodine and Iodine Compounds
chloride. After elution, the resulting iodate – costs. Capital costs include high-quality stain-
iodide solution is acidified with sulfuric acid to less steels because iodine is highly corrosive. In
form iodine. The iodine is then recovered from addition, the brine must be leased from private
the resulting slurry by filtration or centrifuga- landowners over large acreages for many years.
tion, and purified by sublimation or by melting
under sulfuric acid.
In the sloping fluidized-bed process, sodium 7. Uses
iodide is also adsorbed on an ion-exchange resin,
but in this case the resin is mobile and filtration Iodine was first exploited commercially in 1916
of the brine is not necessary. The resin is trans- in the form of potassium iodide as a remedy for
ferred to a recovery plant, and the sodium iodide goiter. In 1916, tincture of iodine and iodoform,
is removed as hydrogen iodide by elution with CHI3 , were used as disinfectants for cuts and
sulfuric acid. Iodine is precipitated by addition abrasions, and for sanitation. In 1939, Louis Da-
of chlorine. guerre published details of his method for mak-
The Wengfu Mine in China has reported io- ing photographs on plated silver by using iodine
dine concentrations of 0.004 %. Calcining at vapor to form a thin coating of light-sensitive sil-
1050 ◦ C yielded an iodine-rich gas that could ver iodide crystals. Clinical uses of iodine have
be further processed. Iodine has been recovered evolved from simple tinctures to radiopaque me-
during process design testing [18]. dia, iodophors, and batteries in subcutaneously
implanted pacemakers [20]. Iodine is also used
in a wide variety of specialty chemicals.
6. Quality Specifications, Establishing an accurate end-use pattern for
Transportation, and Storage iodine is difficult because iodine-containing in-
termediates are marketed before reaching their
Commercial crude iodine normally has a mini- ultimate end uses.
mum purity of 99.5 %. Impurities are chiefly wa- Uses include dietary supplements, catalysts,
ter, sulfuric acid, iron, and insoluble materials. inks and colorants, pharmaceuticals, radiopaque
The U.S. Pharmacopoeia (U.S.P.) XVII specifies diagnostic media, photographic equipment, san-
an iodine content not less than 99.8 %. The Com- itary and industrial disinfectants, and stabiliz-
mittee on Analytical Reagents of the American ers, as well as the production of motor fuels,
Chemical Society (ACS) allows a maximum of smog inhibitors, lubricants, high-purity metals,
0.005 % total bromine and chlorine, and 0.010 % and iodized salt. Iodine also has applications in
nonvolatile matter. cloud seeding; worldwide about 11 000 kg of sil-
Crude iodine is packed in fiber drums lined ver iodide is used in amounts of 10 – 50 g per
with double-ply polyethylene that contain 25 – cloud seeding.
50 kg. There are no specific transportation, ship-
ping, or safety requirements. As low tempera- Catalysts. The major end use of iodine is in
tures reduce vapor pressure and therefore de- catalysts. After the catalytic Monsanto process
crease the rate of loss, a cool, well-ventilated for producing acetic acid was introduced in the
storage area is recommended. 1960s, 90 % of new acetic acid capacity used the
Under normal conditions, the rate of loss process (→Acetic Acid, Chap. 4.2.). The pro-
from intact commercial containers is negligible. cess involves methanol carbonylation with an
Investment and maintenance costs are high for a iodide-promoted rhodium complex as catalyst
new iodine plant using underground brines be- [20]. In the final distillation, the acetic acid of-
cause brine sources usually do not contain high ten contains organic iodides; these are removed
concentrations of iodine and because 1 kg of by crystallization or distillation [21]. Iodide cat-
chlorine is required for the production of 1.4 – alysts, such as titanium tetraiodide [7720-83-4]
1.8 kg of iodine. Electrical costs for maintain- and aluminum iodide [7784-23-8], are also sig-
ing the pumps to bring the brine to the surface, nificant in the dehydrogenation of butane and
for air blowing the iodine, and for reinjecting butene to butadiene [22], and in the preparation
the brine are major components of maintenance of stereoregular polymers [23].
6 Iodine and Iodine Compounds
Stabilizers. Consumption of iodine as a sta- 10 % available iodine and about 5 % inactive io-
bilizer in the manufacture of nylon for tire cord dide [28]. Globaline tablets are used by the army
and carpets, and for converting rosins, tall oil, to disinfect water supplies without boiling [29].
and other wood products to more stable forms is Iodine is used as a disinfectant in swimming
the second major end use [24]. pools and is an efficient microbiocide. It is more
Butadiene reacts with iodine and copper effective than chlorine against the standard fecal
cyanide to yield a copper(I) iodide complex of indicators. High iodine concentrations are eas-
dehydroadiponitrile. Reaction with hydrocyanic ily maintained and are more stable than chlorine
acid gives a high yield of adiponitrile, which [30]. A study in Florida showed that concentra-
can be catalytically hydrogenated to give hexa- tions up to 5 ppm of iodine in potable water are
methylenediamine, a precursor of nylon 66 [26]. not deleterious to health and that 1 ppm is suffi-
In the reaction with HCN the starting materials, cient to safely disinfect a water supply with no
copper(I) cyanide and iodine, are regenerated. loss of effectiveness at high pH values [31].
Animal Feeds. The third area of major io- Pharmaceuticals. The fifth major area of io-
dine consumption is in the form of additives for dine consumption is in pharmaceuticals, primar-
animal feeds. In 1986, the recommended amount ily as radiopaque media, which are injected in-
of ethylene diamine dihydroiodide [5700-49-2] travenously or intracerebrally to produce X-ray
(EDDI) consumed in animal feeds was lowered photographs of the internal structure of the body
from 50 to 10 mg per head per day, in accor- (→Radiopaque Media). Clinical use includes
dance with the U.S. Food and Drug Adminis- cardioangiography, cerebral angiography, arte-
tration Compliance Policy Guideline (7125.18). riography, urography, and phlebography. Media
Cattle and sheep are fed iodine compounds for listed by the U.S. International Trade Commis-
the prevention of soft tissue lumpy jaw. The sion include meglumine diatrizoate [131-49-7],
use of iodized salt has significantly reduced sodium diatrizoate [737-31-5], iopanoic acid
the incident of goiter in livestock in certain [96-83-3], iothalamic acid [2276-90-6], and
iodine-deficient geographical regions. In addi- sodium tyropanoate [7246-21-1]. These sub-
tion, iodized protein in fowl and cattle feed in- stances contain between 47 and 67 wt % io-
creases yields of eggs and milk. About 25 % dine. A nonionic contrast medium, N,N -bis-
of the reported domestic consumption of io- (2,3-dihydroxypropyl-5-[N-(2,3-dihydroxypro-
dine was in animal feeds, primarily as the com- pyl)acetamidol]-2,4,6-triiodoisophthalamide,
pound EDDI, but also as potassium iodide, cal- which contains 46 wt % iodine, has also been
cium iodate [7789-80-2], and calcium periodate marketed [32] . Potassium iodide is widely used
[7790-28-5]. Iodine is one of 14 mineral com- as an expectorant in cough medicines. Hydri-
modities listed by the National Feed Ingredients odic acid and potassium iodide are used in the
Association as being used for domestic meat synthesis of amphetamine, methamphetamine,
production. and ethylamphetamine.
Disinfectant. Iodine is an effective germi- Miscellaneous Uses. One of the oldest in-
cide for a wide range of microorganisms. The use dustrial uses of iodine is in photography. Up
of iodine in sanitary preparations, such as disin- to 7 % of the silver salt in negative emul-
fectants, is the fourth major area of consump- sions is iodide. Ammonium iodide [12027-06-4]
tion. Iodine is often used in conjunction with is used as a photographic developer. Iodine
iodine-complexing nonionic surfactants or poly- is also a constituent of some dyes including
mers (iodophors) in disinfectants that are used in 4 ,5 -diiodofluorescein [38577-97-8], rose ben-
dairies, laboratories, and food processing plants, gal [11121-48-5], and erythrosine [16423-68-0].
and for the sanitation of dishes in restaurants Erythrosine, certified food color Red No. 3,
[26]. Polyvinyl pyrrolidinone – iodine complex contains 58 % iodine and is used in carbonated
[25655-41-8], PVP-iodine, is extensively used soft drinks, powdered drinks, gelatin desserts,
because of its germicidal, bactericidal, fungi- icings, and pet food. 4 ,5 -Diiodofluorescein
cidal, and general disinfecting properties [27] [38577-97-8], C.I. Solvent Red 73, is used
(→Disinfectants). The solution contains about in dyeing and printing cotton, in printing
Iodine and Iodine Compounds 7
half-silk, and in dyeing jute and straw prod- tion of a 150-t/a iodine plant to extract iodine
ucts. Sodium iodate can also be used in the from waste tailings from sodium nitrate produc-
dyeing process as an oxidizing agent. Fluo- tion. These tailings still contain ca. 50 % of the
roiodocarbons, such as CF3 (CF2 )7 I [507-63-1] original iodine content. This new plant will guar-
and CF3 (CF2 )7 CH2 CH2 I [2043-53-0], are in- antee Chile a secure source of iodine from the
termediates in the manufacture of oil and wa- accumulated waste tailings.
ter repellents, surfactants, and fire-extinguishing In the United States, iodine is produced by
foams [33], [34]. IoChem, North American Brine Resources, and
The addition of iodine to aromatic hydrocar- Woodward Iodine.
bons, such as n-butylbenzene [104-51-8], results The United States has been dependent upon
in the formation of charge-transfer complexes imported iodine for roughly half of domestic
that are effective lubricants for some metals, e.g., consumption since ca. 1960. United States im-
titanium or steels [35]. Iodine is also used to ports of iodine and potassium iodide according
produce high-purity forms of titanium, silicon, to the exporting country are listed in Table 3.
hafnium, and zirconium in the van Arkel pro-
Table 3. United States imports of iodine and potassium iodide
cess, which involves the formation of volatile
iodides [36]. Product and country 1985 1986
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.
iodine in a caustic alkali solution and recover- vigorously with dry carbon, phosphorus, and or-
ing the iodate by filtration. Iodine can be oxi- ganic matter.
dized with concentrated nitric acid to iodic acid,
followed by neutralization with an oxide or hy- Polyiodides and polyhalides form when io-
droxide; oxidation of an alkaline iodide solution dine or iodine halides combine with halide
with chlorine; or electrolytic oxidation of a neu- salts. Some examples are RbI3 [12298-69-0],
tral iodide solution containing chromate with a [N (CH3 )4 I9 [3345-37-7], [N (C2 H5 )4 IBr2 ]
platinum anode and an iron cathode in a non- [20445-98-1], and KICl4 [14323-44-5]. Polyio-
diaphragm cell. Iodates are precipitated when dide and polyhalide salts usually form highly
solutions of metal salts react with alkali-metal colored orthorhombic crystals; however, KIF6
iodates. [20916-97-6] is white.
Most iodates are white. The solid metallic
iodates are stable and safe to handle in the pure Iodine Halides. Seven iodine halides
state, but should be kept out of contact with or- are known: IF, IF3 , IF5 [7783-66-6],
ganic substances and other combustible materi- IF7 [16921-96-3], I2 Cl6 [865-44-1], ICl
als because such mixtures can be explosive. The [7790-99-0], and IBr [7789-33-5]. All are read-
salts of the alkali metals and magnesium are sol- ily formed by direct reaction of the two halogen
uble in water. elements.
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 10. Organic Iodine Compounds
in the presence of iodates by adding potassium
iodide to a solution buffered with borax and As with inorganic iodides, organic iodides have
boric acid, and titrating the liberated iodine with higher specific gravities and indices of refraction
sodium arsenite solution. than the corresponding chloro and bromo deriva-
tives, and higher melting and boiling points.
Hydrogen iodide [10034-85-2], HI, Compounds containing an iodine atom attached
M r 127.93, mp −50.9 ◦ C, bp −35.4 ◦ C, is a col- to a primary carbon atom are the most stable.
orless gas which fumes in moist air. It is the Secondary compounds are less stable, and ter-
least stable of the common hydrogen halides, tiary are the least stable of all. Aliphatic iodo
and the best reducing agent. It cannot be pre- compounds can be prepared by reacting alcohols
pared using strong acids because it is readily with a mixture of iodine and red phosphorus.
oxidized; therefore, it must be prepared from a Other methods for preparing aliphatic iodo com-
weak acid and iodine, or by the hydrolysis of pounds include: the addition of iodine halides or
certain iodides. iodine to olefins; the substitution reactions bet-
Hydriodic acid is the colorless solution ween an alkyl bromide or chloride and an alkali
formed when hydrogen iodide gas dissolves in metal iodide; and the reaction of alcohols with
water, and at commercial strength, typically con- triphenyl phosphite and methyl iodide.
tains 47 % hydrogen iodide. Hydriodic acid is Organic iodo compounds are highly reactive.
used in the preparation of iodides and is an im- The alkyl iodides are generally 50 – 100 times
portant reagent in organic chemistry. Dilute hy- more reactive than the corresponding chlorides.
driodic acid reacts with metals or their hydrox- Reactivity of unsaturated organic iodides is af-
ides, carbonates, and other salts to yield metal fected little if iodine is attached to one of the
iodides. carbon atoms involved in the double bond, but
reactivity is promoted if iodine is attached to the
Iodic acid [7782-68-5], HIO3 , M r 175.93, carbon atom in the α position to the double bond.
forms colorless rhombohedral crystals, which Iodine in organic compounds can be determined
decompose at 110 ◦ C. It is prepared by oxidiz- after being converted into an inorganic form by
ing iodine with nitric acid. It is a strong oxidiz- heating the compound with one of the following:
ing agent, oxidizing iodide to iodine, sulfite to (1) fuming nitric acid in a sealed strong glass
sulfate, and hydrogen sulfide to sulfur. It reacts tube, the Carius method; (2) sodium peroxide
10 Iodine and Iodine Compounds
20. F. E. Paulik, Catal. Rev. 6 (1972) 49. R. P. 36. R. F. Rolsten: Iodine Metals and Metal Iodide,
Lowry, A. Aguilo, Hydrocarbon Process. 53 J. Wiley & Sons, New York 1961.
(1974) no. 11, 103. 37. J. W. Thomas et al., Anal. Chem. 22 (1950)
21. Kirk-Othmer, 3rd ed., 1, 135. 726 – 747.
22. Shell Oil Co., US 3 106 590, 1963 (C. W. 38. N. Bagchi et al., Science (Washington D.C.)
Bittner); US 3 119 881, 1964 (R. L. Hodgson). 4723 (1985) 325 – 327.
23. Phillips Petroleum Co. GB 920 244, 1963; 39. N. I. Sax: 47th Dangerous Properties of
GB 931 440, 1963. Industrial Materials, 4th ed., Van Nostrand
24. G and A Laboratories, Inc., US 2 299 577, Reinhold Co., New York 1975, p. 32.
1942 (T. Hasselstrom, E. A. Brenhan). 40. CFR, 29th, 1910.1000, Table Z – 1, July 1,
25. Esso Research & Engineering, J. Chem. Eng. 1986.
78 (1971) 107. 41. Encyclopaedia of International Occupational
26. R. R. Whitehead, Water Pollut. Control (Don
Health and Safety,vol. 1 , International Labor
Mills Can.) 119 (1981) no. 12, 10 – 12.
Office, Geneva 1971.
27. West Chem. Products, Inc., US 4 271 149,
42. Federal Emergency Management Agency:
1981 (M. W. Winicor, M. Oberland).
28. GAF Corp., US 4 128 633, 1978 (D. H. Larenz, “Federal Policy on Distribution of Potassium
E. P. Williams). Iodide Around Nuclear Power Sites for Use as
29. J. C. Morris et al., Ind. Eng. Chem. 45 (1953) a Thyroidal Blocking Agent,” Fed. Regist. 50
1013 – 1015. (1985, Jun. 20) no. 119, 25624 – 25625.
30. M. S. Favero, C. H. Drake, Public Health Rep. 43. Federal Emergency Management Agency:
79 (1964) no. 3 251 – 257. “Federal Policy on Distribution of Potassium
31. W. C. Thomas et al. Trans. Clin. Climatol. Iodide Around Nuclear Power Sites for Use as
Assoc. 90 (1978) 153 – 162. a Thyroidal Blocking Agent,” Fed. Regist. 50
32. E. Lindgren, Acta Radiol. Suppl. 362 (1980) (1985, Jul. 24) no. 142, 30258 –30259.
134. 44. Büchel, Moretto, Woditsch Industrial
33. Kirk-Othmer, 2nd ed., 22, 145 – 146. Inorganic Chemistry Wiley-VCH, Weinheim
34. Kirk-Othmer, 2nd ed., Supplement Volume, 45. http://minerals.usgs.gov/minerals/pubs/
964 – 973. commodity/iodine
35. General Electric Co., US 3 228 880, 1966
(R. S. Owens, R. W. Roberts).
My Profile Log
HOME
ABOUT US
CONTACT US
HELP
Home / Chemistry / Industrial Chemistry
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]
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.
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.
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]
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.
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.
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.
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).
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]
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.
The application of ion beams for insulator modification can be grouped in two categories:
The most important materials treated in practice are different types of polymers, engineering ceramics, and glasses.
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].
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