Current Opinion in Colloid & Interface Science 6 Ž2001.

148᎐159

Natural surfactants

Krister HolmbergU
Department of Applied Surface Chemistry, Chalmers Uni¨ ersity of Technology, SE-412 96 Goteborg,
¨ Sweden

Abstract

The ever-increasing environmental concern about surfactants triggers an interest in natural surfactant. This review, which
has an emphasis on work published since 1998, covers three categories of natural surfactants: amphiphiles produced by yeast
or bacteria, amphiphiles containing a natural polar headgroup and amphiphiles containing a natural hydrophobic tail.
Microorganisms produce both high molecular weight and low molecular weight surfactants. Only the low molecular weight
compounds are included in the review. Sugars and amino acids are the two most important examples of surfactant polar
headgroups of natural origin. The research is particularly intense in the area of sugar surfactants and the review covers three
types: alkylglucosides, alkylglucamides and sugar esters. Surfactants based on two types of natural hydrophobic tails are
included: fatty acid monoethanolamides and sterol ethoxylates. Routes of preparation as well as physico-chemical properties
are discussed for the surfactants prepared by organic synthesis. 䊚 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Surfactant; Natural; Biotechnology; Sugar; Polyol; Amino acid; Fatty acid; Sterol

1. Introduction
The term ‘natural surfactant’ is not unambiguous.
Taken strictly a natural surfactant is a surfactant
taken directly from a natural source. The source may
be of either plant or animal origin and the product
should be obtained by some kind of separation proce-
dure such as extraction, precipitation or distillation.
No organic synthesis should be involved, not even as
an after-treatment. There are in fact not many surfac-
tants in use today that fulfil these requirements.
Lecithin, obtained either from soybean or from egg
yolk, is probably the best example of a truly natural
surfactant.
The main reason why natural surfactants in the real
sense of the word are so scarce is not a lack of
availability. Amphiphiles are abundant in both the
U
Tel.: q46-31-7722969; fax: q46-31-160062.
E-mail address: kh@surfchem.chalmers.se ŽK. Holmberg..
plant and the animal kingdom where they are often
referred to as polar lipids. In biological systems the
surface active agents are used in very much the same
way as surfactants are employed in technical systems:
to overcome solubility problems, as emulsifiers, as
dispersants, to modify surfaces, etc.
The factor that works against production of truly
natural surfactants is the cost of work-up. The
products are usually present in small quantities and
the separation process tends to be tedious. In most
instances the cost of separationrisolation will by far
exceed the manufacturing cost of equivalent synthetic
surfactants.
The unfavorable cost situation for truly natural
surfactants may change if fermentation processes can
be developed that yield biosurfactants in high yields.
Both yeast and bacteria can be efficient producers of
surface active agents and there is considerable cur-
rent interest in biotechnological processes for produc-
tion of amphiphilic compounds. These substances can
be either low molecular weight, such as acylated

1359-0294r01r$ - see front matter 䊚 2001 Elsevier Science Ltd. All rights reserved.
PII: S 1 3 5 9 - 0 2 9 4 Ž 0 1 . 0 0 0 7 4 - 7
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159 149

oligosaccharides, or high molecular weight, most of-

ten lipopolysaccharides. So far most interest has fo-
cused on the high molecular weight products, with
䊛
Emulsan, a highly charged heterolipopolysaccharide
with a molecular weight of approximately 10 6 , as the
prime example w1x.
The term ‘natural surfactant’ is often used in a
broader sense than that discussed above, however.
Surfactants synthesized from natural raw materials
are usually referred to as natural surfactants. Proper
examples of surfactants belonging to this category are
fatty acid esters of sugars and fatty acid esters or
amides of amino acids. It is, in fact, common practice
to be even more generous in the use of the term
‘natural surfactant’. A surfactant with one of the main
building blocks, the polar headgroup or the hy-
drophobic tail, obtained from a natural source is
often referred to as a natural surfactant. For example,
alkyl glucosides which are prepared from a ‘natural’
sugar unit and a ‘non-natural’ fatty alcohol are often Fig. 1. A trehalose ester.
regarded as natural surfactants. ŽSo-called ‘natural
fatty alcohols’ should not be seen as natural starting
hydroxy fatty acids connected to disaccharides by es-
materials, although this is often done. Fatty alcohols
ter bonds. They are extracellular compounds pro-
derived from coconut oil and other triglycerides are
duced by actinomycetes such as Mycobacterium,
made by two consecutive chemical reactions:
Corynebacterium and Bre¨ ibacterium w3x. The acylpoly-
methanolysis of the triglyceride followed by reduction
ols are abundant in bacterial cell walls. A typical
of the methyl ester..
example of an acylpolyol is the trehalose ester shown
In this review the term ‘natural surfactant’ is used
in Fig. 1.
in its broadest sense. The paper covers true natural
surfactants prepared by fermentation, synthetic sur-
2.2. Glycolipids
factants with a natural polar head group and synthetic
surfactants with a natural hydrophobic tail. Only low
Glycolipids are usually hydroxy fatty acids attached
molecular weight surfactants with a well-defined
to a sugar via a glycosidic bond. Sophorolipids and
structure are included. The ‘natural surfactants’ used
rhamnolipids are well known examples, produced by
to treat the respiratory distress syndrome, a chapter
Candida and by Pseudomonas, respectively w4,5x. Fig. 2
of its own, are in reality mixtures of amphiphiles with
shows a cyclic and an acyclic sophorolipid. Rham-
structures not fully elucidated w2x and are, hence, not
nolipids are attractive in that they can be efficiently
covered. The review is organized correspondingly.
produced during growth on either hydrocarbon or
carbohydrates as the sole carbon source. There has
been considerable interest in the molecular genetics
2. Surfactants prepared by fermentation of rhamnolipids and this topic has recently been re-
viewed in this journal w6x.
2.1. Acylpolyols A sucrose lipid produced by Serratia marcescens has
recently been isolated and characterized w7 x. The 䢇

Acylpolyols prepared by fermentation are usually product was found to be an excellent emulsifier for a

Fig. 2. A cyclic and an acyclic sophorolipid.

150 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
Fig. 3. Surfactin from Bacillus subtilis.
wide range of hydrocarbons including crude oil. It has
been suggested as a surfactant for cleaning of oil
tanks.
2.3. Acylpeptides
Acylpeptides or lipopeptides, are usually cyclic
compounds based on a hydroxy acid and a short
peptide chain. By far the most studied compound is
䊛
Surfactin, produced from Bacillus subtilis w8,9x.
Acylpeptides, such as Surfactin and Lichenysin pro-
duced by Bacillus licheniformis, are efficient am-
phiphiles when it comes to surface tension reduction,
etc. Surfactin is one of the few biosurfactants that has
found commercial use. It is used in a variety of
pharmacological applications w10x. The formula of
Surfactin is shown in Fig. 3. Surfactin and rham-
nolipids are the only two biosurfactants for which
there is any regulatory information and the molecular
genetics of Surfactin has recently been reviewed w6x.
Fig. 4. Examples of polyol surfactants. From top to bottom: an
alkyl glucoside, an alkylglucamide, and the ethyl acetal of a fatty
acid glucose ester.
3. Surfactants based on a natural polar headgroup
3.1. Sugar as polar headgroup
Various types of sugars or polyols derived from
sugar have been used as surfactant polar headgroup
for many years. Sorbitane alkanoates and ethoxylated
sorbitane alkanoates, well known under the trade
names of 䊛 Span and 䊛 Tween, respectively, have been
around for a long time. In recent years there has been
a focus on three classes of surfactants with sugar or a
polyol derived from sugar as polar headgroup: alkyl
polyglucosides ŽAPGs., alkyl glucamides and sugar
esters. Representative structures of the three surfac-
tant types are shown in Fig. 4.
There is currently a very strong interest in ex-
ploring alkyl polyglucosides ŽAPGs. as surfactant for
several types of applications. APGs are synthesized by
direct reaction of glucose with fatty alcohol, using a
large excess of alcohol in order to minimize sugar
oligomerization. Alternatively, they are made by
transacetalization of a short chain alkyl glucoside,
such as ethyl or butyl glucoside, with a long chain
alcohol. An acid catalyst is used in both processes.
Either glucose or a degraded starch fraction is used as
starting material. Fig. 5 illustrates the synthesis. Alkyl
Fig. 5. Pathway for APG synthesis Žfrom von Rybinski and Hill w18x
with permission..
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
glucosides can also be made by enzymatic synthesis,
using ␤-glucosidase as catalyst which yields only the
␤-anomer Žand gives low yield.. The corresponding
␣-anomer can more readily be obtained by ␤-gluco-
sidase catalyzed hydrolysis of the racemate. There are
considerable differences between the ␣␤ mixture ob-
tained by organic synthesis and the pure enantiomers
obtained by the bio-organic route. The ␤-anomer of
n-octyl glucoside has found use as a surfactant in
biochemical work.
Alkyl glucosides are stable at high pH and sensitive
to low pH where they hydrolyze to sugar and fatty
alcohol. A sugar unit is more water-soluble and less
soluble in hydrocarbons than the corresponding poly-
oxyethylene unit; hence, APGs and other polyol-based
surfactants are more lipophobic than their poly-
oxyethylene-based surfactant counterparts w11x. This
makes the physico-chemical behavior of APG surfac-
tants in oil᎐water systems distinctly different from
that of conventional non-ionics. Furthermore, APGs
do not show the pronounced inverse solubility vs.
temperature relationship that normal non-ionics do.
This makes an important difference in solution behav-
ior between APGs and polyoxyethylene-based surfac-
tants. The main attractiveness of APGs lies in their
favorable environmental profile: the rate of biodegra-
dation is usually high and the aquatic toxicity is low
w12,13x. In addition, APGs exhibit favorable dermato-
logical properties, being very mild to skin and eye
w14x. The mildness makes this surfactant class attrac-
tive for personal care products but APGs have also
found a range of technical applications.
APGs have been the topic of several excellent
overviews w15᎐18x. In addition, the physico-chemical
properties of APGs, as well as other polyhydroxyl-
based surfactants, have recently been reviewed in this
journal w19x. Only a few brief notes on recent develop-
ments in the field will therefore be given here.
The solution behavior of APGs, both anomeric
mixtures, prepared by organic synthesis, and pure
enantiomers have been studied in detail w20 ,21᎐23x.
䢇
The micellar phase region is usually large and at
higher concentrations the normal pattern of liquid
crystalline phases appears. It interesting to note, as
was described in a previous review w19x, that some
APG surfactants exhibit a miscibility gap in the micel-
lar region. As with polyoxyethylene-based non-ionics
two micellar phases appear in this region, one dilute
and one concentrated. For a pure decyl glucoside
surfactant it was found that the dilute phase consisted
of micelles of aggregation number in the range
200᎐400 while the concentrated phase contained
larger aggregates, probably branched micelles that
formed a network through entanglement w20x. A char-
acteristic feature of the liquid crystalline region which
appears at higher surfactant concentration is that the
151
borders between the different crystalline phases in a
temperature vs. surfactant concentration diagram are
almost vertical, indicating a temperature-independent
behavior. This is very different from the behavior of
polyoxyethylene-based non-ionics.
Mixed micelle formation of two APGs with differ-
ent length of the alkyl chain with both anionic and
cationic surfactants have been studied w24 x. The ex-
䢇
perimental data were fitted to Rubingh’s model for
non-ideal mixed micelle formation w25x. A value of the
interaction parameter, ␤ is obtained and the value of
␤, positive or negative, indicates whether there is a
repulsive or an attractive interaction involved in the
formation of mixed micelles between the two surfac-
tant types. Combination of an APG with the cationic
surfactant dodecyltrimethylammonium bromide gave
a value of the ␤-parameter of y4.1 indicating rather
strong attractive interaction. The combination of the
same APG with the anionic surfactant sodium dode-
cylsulfate gave a ␤-value of y2.3, indicative of a
smaller attraction. These results are interesting and
not easily explained. They tend to support the view,
proposed by several workers, that APGs carry a nega-
tive net charge, the origin of which is unclear w24 x. It
䢇
is interesting that whereas a combination of an APG
and an alcohol ethoxylate gave a value of the ␤-
parameter of zero, the same APG when combined
with another alkyl glycoside, an alkyl maltoside of the
same alkyl chain length ŽC10. gave a ␤-value of y2.0.
This attractive interaction was tentatively explained as
due to favorable packing of the headgroups of the two
surfactants w24 x. 䢇
The majority of work on glycoside surfactants has
been made on alkyl glucosides, i.e. surfactants based
on glucose or slightly oligomerized glucose as polar
headgroup. Recently, alkyl maltosides have been stud-
ied with a specific focus on its behavior in mixed
micelles together with anionic and non-ionic surfac-
tants w26,27x. A combination of a decyl maltoside with
sodium dodecylbenzene sulfonate gave a value of the
␤-parameter of y2.1 which can be compared with the
value of y3.3 for a combination of an alcohol ethoxy-
late, octaŽethylene glycol.monodecyl ether ŽC10E8.
with the same anionic surfactant. A combination of
the alkyl maltoside with the alcohol ethoxylate gave a
very small ␤-value Žy0.3.. A synergistic effect with
regard to surface tension reduction was also found for
combinations of either of the non-ionic surfactants
and sodium dodecylbenzene sulfonate. Thus, the re-
sults obtained with the maltoside surfactant agree
very well with those previously obtained with gluco-
side surfactants of the same alkyl chain length. A full
explanation of the results still remains, however.
Alkylglucamides Žsee Fig. 4., or more strictly named
as N-alkanoyl-N-metylglucamines, are commercially
important products. The product sold in large quanti-
152 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
ties for the detergent sector is N-dodecanoyl-N-meth-
ylglucamines, i.e. the C12-derivative. The product is
prepared from glucose, metyl amine, hydrogen and
methyl laurate by a two-step reaction.
Alkylglucamides have been the subject of an exten-
sive review recently w28x. Few major investigations on
this surfactant class seem to have been made recently.
In a paper by Zhu et al. a glucamide was compared
with polyol surfactants of similar structure, a xylamide
and a glyceramide, all having a C12 alkyl chain w29 x.
The glucamide has four hydroxyl groups in the polar
head, the xylamide three and the glyceramide one.
This was reflected both in the area per molecule
values and the CMC values which were 30, 28 and 26
˚2 and 0.347, 0.331 and 0.234 mM, respectively.
A contrast, at lower temperature, when activation en-
There is considerable current interest in sugar es-
ters Žsee Fig. 4. and a recent review covers the topic
until 1998 w30x. Esters of glucose can be made either
by enzymatic synthesis, using a lipase catalyst, or by
an organic chemical route. Using the right enzyme
the bio-organic route can give esterification almost
exclusively at the 6-position of the sugar moiety. The
organic synthesis requires extensive use of protecting
groups to obtain high selectivity. Selective enzymatic
synthesis of other sugar esters than glucose esters
have been difficult to achieve without the use of
protective groups. Starting from sugar acetals and
fatty acids, monoesters of sugars were obtained in
good yield, after a deprotection step, from several
starting materials, both mono- and disaccharides
w31,32x.
The advantage of the organic synthesis over the
enzymatic route is that the sugar unit can be varied at
will. The group of Drummond has recently made
important contributions in this area with a specific
focus on degradation mechanisms w33 ,34 ,35x but
䢇 䢇䢇
also covering physico-chemical behavior w36 x. A se-
䢇
ries of sugar ester surfactants were synthesized having
different headgroup size Žglucose, sucrose and raffi-
nose, consisting of one, two and three anhydroglucose
rings, respectively. and with C12 and C16 acyl chains.
Sulfonated sugar esters were also prepared. It was
found that all unsubstituted sugar esters degraded
rapidly. Variations in the sugar headgroup size or the
acyl group chain length did not significantly affect the
biodegradability. The anionic derivatives degraded
more slowly, however w33 x.
䢇
A detailed study on the mechanism of biodegrada-
tion of the unmodified and the sulfonated sugar esters
showed that sucrose laurate biodegradation occurs via
initial ester hydrolysis. The sucrose ␣-sulfonyl laurate,
on the other hand, degrades by initial alkyl chain
oxidation. This indicates that the ester hydrolysis
pathway is blocked by the sulfonyl group adjacent to
the ester bond so that biodegradation is forced to
proceed via the slower alkyl chain oxidation pathway
w34 x. An ethyl substituent in ␣-position instead of
䢇䢇
the sulfonate group also slowed down the biodegrada-
tion rate, probably for the same reason. The two
degradation routes are shown in Fig. 6.
Base catalyzed hydrolysis was also investigated.
Similar to the results from the biodegradation, the
unsubstituted sugar esters degraded faster than the
derivatives with either a sulfonate or an ethyl group in
the ␣-position. The difference in hydrolysis rate
䢇
between the three surfactants sucrose laurate, sucrose
␣-sulfonyl laurate and sucrose ␣-ethyl laurate, varied
with temperature. It was concluded that at higher
temperature, when steric hindrance is the dominant
factor, hydrolysis of ␣-sulfonyl laurate is slowest. By
ergy is more important, hydrolysis is faster in the
presence of an electron-withdrawing ␣-sulfonyl group
than in the presence of an electron-attracting ␣-ethyl
group w35x.
Dimeric and trimeric sugar ester surfactants have
been synthesized by a combination of enzymatic and
organic chemical methods w37 x. An impressive se-
䢇䢇
ries of products were prepared with different spacer
units between the sugar rings. For instance, a dimeric
Žgemini. glucose ester with C12 acyl chains attached
at positions 6 of the two anhydroglucose units and
with the two moieties connected via a six carbon
spacer was prepared in high yield. The product is
shown in Fig. 7.
3.2. Amino acid as polar headgroup
Amino acids and short peptides constitute an alter-
native to sugars as a natural polar headgroup for
surfactants. Several amino acids have been investi-
gated for this purpose with the majority of papers
dealing with basic amino acids such as arginine and
lysine from which cationic surfactants can easily be
prepared. Most synthetic procedures are based on
organic synthesis but enzymatic processes have also
been explored w38,39x. Arginine-based surfactants have
recently been reviewed w40x.
Arginine surfactants, such as N-dodecanoylarginine
metyl ester salt are conveniently prepared by reaction
between the oil soluble dodecanoic acid and the water
soluble arginine methyl ester hydrochloride using di-
cyclohexylcarbodiimide as condensing agent. The re-
action is normally carried out in an aprotic, polar
solvent such as dimethylformamide ŽDMF.. The reac-
tion is performed at room temperature and works
well in a laboratory scale. However, DMF is an un-
suitable solvent for large scale operations, partly due
to cost and toxicity and partly due to difficulties in
removing it from the product by simple vacuum evap-
oration. An alternative process, using a high-
internal-phase-ratio emulsion, has been investigated
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
Fig. 6. Degradation pathways of a sugar fatty acid ester Žfrom Baker et al. w34x with permission..
for the synthesis w41 x. Such emulsions can be formu-
䢇䢇
lated with a large amount, sometimes above 99%, of
internal phase and with very low amount of surfac-
tant, often less than 0.5%. The structure of water-in-
oil emulsions of this type is that of closed-packed
Fig. 7. A sugar fatty acid gemini surfactant Žfrom Gao et al. w37x
with permission..
153
water droplets of a broad size distribution separated
by a thin film of the continuous oil phase. The vis-
cosity is usually high and such emulsions are
sometimes referred to as gel emulsions. It was de-
monstrated that the concentrated emulsions were
good reaction media with yields equivalent to those
obtained in DMF. The yield depended on the water-
to-oil ratio. Too high water content at constant sur-
factant concentration gave large water droplets with
corresponding small interfacial area which was found
to be unfavorable.
A papain-catalyzed synthesis of arginine-based sur-
factants have been reported w42,43x. Papain deposited
on a solid support was used as catalyst for the amide
and ester bond formation between arginine methyl
ester and various long-chain alkyl amines and fatty
154 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
Fig. 8. Route of preparation of a tryptophan-based surfactant Žfrom Pegiaduo et al. w46x with permission..
alcohols in organic solvent. High yield of arginine
amide was obtained from reaction in acetonitrile. For
the synthesis of the arginine fatty acid ester the best
yield was obtained in a solvent-free system kept at a
temperature above the melting temperature of the
fatty alcohol.
Using a combination of organic and bioorganic
synthesis arginine-based gemini surfactants have been
synthesized by the group of Infante w44x and the
solution properties have been investigated w45x. The
gemini surfactants consisted of two N-acylarginine
moieties connected via the arginine carboxyl groups
by a diamino spacer molecule, such as 1,3-di-
aminopropane or 1,3-diamino-2-hydroxypropane. A
N-acyl-L-arginine alkyl ester was used as starting
material. This compound was first reacted with the
diamino spacer to form the amino-amide, i.e. the
spacer reacted with one of its amino group to form a
N-acyl-L-arginine amide. In a second step, catalyzed
by papain, the free amino group reacted with another
molecule of N-acyl-L-arginine alkyl ester to form the
diamide product. Overall yields of the order of 65%
were reported.
A series of tryptophan-based surfactants have been
synthesized w46 x. The surfactants were synthesized by
䢇
reaction of tryptophan with the epichlorohydrine
adduct of a fatty alcohol, as shown in Fig. 8. The
series of surfactants based on fatty alcohols with from
9 to 16 carbon atoms was evaluated with respect to
CMC, surface tension at CMC and area per molecule
at the air᎐water interface. The results are summa-
rized in Table 1. As can be seen, the CMC values are
low, considerably lower than for simple alkanoate
surfactants of the same hydrocarbon chain length. For
instance, sodium laurate has a CMC of 20 mM to be
compared with the value of 0.038 mM of the C12
surfactant of Table 1. The surface tension values are
generally low, as expected for these hydrophobic sur-
factants. The areas per molecule are within the ex-
pected range, the increase in area for the shorter
chain length surfactants most probably reflecting a
less ordered packing at the surface.
The amino acid-based surfactants are possible can-
didates for pharmaceutical applications and the
hemolytic action of some lysine-based anionic surfac-
tants has been investigated. The surfactants were salts
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159 155

Table 1
Physico-chemical data for arginine-based surfactants having the structure given in Fig. 8 Ždata taken from Pegiaduo et al. w46x.

Compound CMC Surface tension at CMC Area per molecule

ŽmM. ŽmNrm. ˚2 .
ŽA

Rs C9H19 0.5 35 78
Rs C10H21 0.36 33 67
Rs C11H23 0.065 32 53
Rs C12H25 0.038 32 51
Rs C13H27 0.022 32 64
Rs C14H29 0.020 31 64
Rs C15H31 0.017 30 66
Rs C16H33 0.012 30 68

of N␣ ,N␧-dialkanoyl lysine. In one series of experi- size of the polar group Ždecrease in the critical pack-
ments the alkyl chain lengths were varied w47x. In ing parameter value. which is unfavorable for close
another work the surfactant was N␣ ,N␧-octanoyl ly- alignment of surfactant molecules at the air᎐water
sine and the counterion was varied w48x. As expected, interface. The same effect was seen in adsorption at a
the hemolytic activity depended on the alkyl chain hydrophobic, solid surface, as studied by ellipsometry.
length and on the choice of counterion. With lysine as The presence of unsaturations in the hydrophobic
counterion to the lysine-based anionic surfactant a tail gave an increase in CMC. This may partly be due
protective effect against hypotonic hemolysis was ob- to the surfactants becoming more hydrophilic and
tained. partly to the increased bulkiness of the chains render-
ing packing into closed aggregates and formation of
hydrogen bonds between amide groups more difficult.
4. Surfactants based on a natural hydrophobic tail The presence of one double bond, cis Žoleyl. or trans
Želaidyl., did not much affect the molecular cross-sec-
4.1. Fatty acid as hydrophobic tail tional area at the surface. Two double bonds gave
much larger areas, as can be seen from Fig. 9.
Self-diffusion NMR was used to study the effect of
Fatty amide ethoxylates are easily prepared by temperature and surfactant concentration on the size
ethoxylation of the fatty amide monoethanolamide. of the surfactant aggregates. It was found that at both
The ethanolamide, in turn, is prepared by aminolysis
of the fatty acid metyl ester by ethanolamine. The
fatty amide ethoxylates are of interest as an alterna-
tive to fatty alcohol ethoxylates for several reasons: Ži.
they biodegrade readily to fatty acid and amino-
terminated polyŽethylene glycol.; Žii. the amide bond
in the structure may improve surfactant packing due
to hydrogen bond formation; and Žiii. double bonds in
the fatty acid chain can be preserved in the product
Žalthough during the ethoxylation step 1,3-cis,cis-dou-
ble bonds undergo migration to conjugated trans,cis
structures ..
A series of fatty amid ethoxylates, all based on C18
fatty acids but with varying number of double bonds,
has been synthesized and evaluated with regard to the
effect of the amide bond and the double bonds on
physico-chemical properties w49 ,50 x. Surfactants
䢇 䢇

with different polyoxyethylene chain lengths were pre-

pared and the corresponding saturated fatty alcohol
ethoxylates, stearyl ethoxylates, were used as refer-
ences. It was found that the presence of the amide
group led to a decrease in CMC, possibly due to
hydrogen bond formation. The surface tension at the
CMC was higher when an amide bond was present,
however, which was assumed to reflect the increase in
Fig. 9. Area per molecule for a series of fatty amide ethoxylates
with 10, 15 or 20 oxyethylene units. The number of double bonds in
the acyl chain vary from zero Žstearyl. to one, cis Žoleyl. and one,
trans Želaidyl. to two Žlinoleyl.. Stearyl alcohol erhoxylates ŽC18En.
are included as references.
156 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
high and low surfactant concentration the hydrody-
namic radius of a stearyl amide ethoxylate was some-
what larger than the radius of the stearyl alcohol
ethoxylate with the same number of oxyethylene units.
For the higher surfactant concentration, where the
micelles are non-spherical, this may be due to the
larger polar headgroup size of the amide surfactants
Žsee discussion above. giving slightly less elongated
micelles. The results are more difficult to rationalize
for the low surfactant concentration where both am-
phiphiles give spherical micelles.
4.2. Sterol as hydrophobic tail
Sterol-based surfactants have been reviewed re-
cently w51x and a new review is in progress w52x.
Sterol-based surfactants are of interest because of the
large hydrophobic group of fully natural origin which,
due to its rather planar four ring structure, may
induce good packing at interfaces. Two recent papers
deal with phytosterol ethoxylates w53 ,54x. Phytos-
䢇 oxyethylene units. The surfactant having 10 oxyethy-
terols are sterols of plant origin and their structure is
similar to that of cholesterol, the prime example of
sterols from animal sources. The sterols contain a
secondary hydroxyl group that can be ethoxylated.
The alcohol is sterically hindered, however, and the
reaction with ethylene oxide is not straightforward.
The best procedure seems to be to start the ethoxyla-
tion with a Lewis acid catalyst, cease the reaction
after 3᎐5 mol of ethylene oxide have been added, and
continue with KOH as initiator w55x. The structure of
a representative sterol ethoxylate is shown in Fig. 10.
Surface tension plots of a series of phytosterol
ethoxylates with 10, 20 and 30 oxyethylene units
showed that Ži. the CMC values were very low; and
Žii. the CMC increased with increasing polyoxyethy-
lene chain length. The latter, which has previously
been observed by others as accounted for in Folmer
et al. w53 x, is unexpected and is difficult to explain.
䢇
The values are collected in Table 2. Another interest-
ing observation is that the time to reach surface
tension equilibrium is very long, more than 2 h. The
decay of surface tension with time is shown in Fig. 11.
Most likely, the long equilibrium time is due to ex-
Fig. 10. ␤-Sitosterol ethoxylates.
Table 2
Physico-chemical data for sterol-based surfactants having the
structure given in Fig. 10 Ždata taken from Folmer et al. w53x.
Number of CMC Surface
oxyethylene Ž ␮ M. tenstion at
CMC ŽmNrm.
10 10 31
20 7 34
30 3 42
change reactions of sterol ethoxylates with varying
polyoxyethylene chain length at the air᎐water inter-
face and to slow conformational changes of individual
molecules at the surface. The multiring structure of
sterols is rigid and the time required to adapt a
favorable conformation at an interface may be long.
The phase behavior of the sterol ethoxylates has
been studied in some detail w53 ,54x. Fig. 12 shows
䢇
phase diagrams for phytosterols with 5, 10, 20 and 30
lene units gives a lamellar phase over a very wide
concentration range at elevated temperature. The
phase diagrams of the more hydrophilic surfactants
with large headgroups are dominated by hexagonal
and cubic phases. The high thermal stability of the
liquid crystalline phases is noteworthy. It is particu-
larly striking to see a cubic phase for a non-ionic
surfactant that extends above 100⬚C. Cubic phases are
normally seen as being trapped in a three-dimen-
sional lattice with no possibility of movement. Usually
when the temperature is increased a cubic structure
therefore melts at a relatively low temperature. The
high thermal stability of the cubic phase formed by
the ethoxylated sterols is probably related to an effi-
cient packing of the sterol moieties. Self-diffusion
NMR measurements of samples from within the cubic
Fig. 11. Surface tension for phytosterol with 20 oxyethylene units
as a function of time at fixed concentration Ž3.44= 10y6 M9 Žfrom
Folmer et al. w53x with permission..
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
Fig. 12. Phase diagrams for phytosterol with Ža. five oxyethylene
units; Žb. 10 oxyethylene units; Žc. 20 oxyethylene units; and Žd. 30
oxyethylene units. L1, L2, L ␣ , L ␤ , I1, Nc , and H1 indicate the
following phases: micellar, reversed micellar, lamellar, gel, closed-
packed micellar cubic, nematic, and hexagonal, respectively Žfrom
Folmer et al. w53x with permission..
phase has shown the presence of elongated aggre-
gates with an axial ratio of 1.6 w54x.
5. Conclusions
The review has covered work performed in recent
years in three areas, i.e. surfactants produced from
bacteria and yeast, surfactants based on either sugar
or amino acid as polar headgroup, and surfactants
based on either fatty acid or sterol as hydrophobic
tail. All three areas are important and subject to
considerable research activity. The natural surfactants
obtained by these approaches are generally well char-
acterized and the performance are up to the stan-
dards of conventional surfactants. The biotechnologi-
cally produced surfactants suffer from high cost due
to relatively low yield in the fermentation and cum-
bersome work-up procedures. The surfactants based
on either a natural polar headgroup or a natural
hydrophobic tail show great promise for the future.
157
Acknowledgements
Dan Lundberg at Chalmers University of Tech-
nology is acknowledged for help with the literature
search. This review on natural surfactants and some
of the work referred to are part of the activities
within SNAP, the Centre for Surfactants from Natu-
ral Products which is sponsored by VINNOVA and by
several companies.
References and recommended reading
䢇 of special interest
䢇䢇 of outstanding interest
w1x Kosaric N. Biosurfactants. Surfactant Science Series 48. New
York: Marcel Dekker, 1993.
w2x Halliday HL. Controversies: synthetic or natural surfactant.
The case for natural surfactant. J Perinatal Med
1996;24:417᎐426.
w3x Haferburg D, Hommel R, Claus R, Kleber H-P. Extracellular
microbial lipids as biosurfactants. Adv Biochem EngrBio-
technol 1986;33:53᎐93.
w4x Weber L, Stach J, Haufe G, Hommel R, Kleber H-P. Eluci-
dation of the structure of an unusual cyclic glycolipid from
Torulopsis apicola. Carbohydrate Res 1990;206:13᎐19.
w5x Guerra-Santos LH, Kappeli O, Fiechter A. Dependence of
Pseudomonas aeruginosa continuous culture biosurfactant
production on nutritional and environmental factors. Appl
Microbiol Biotechnol 1986;24:443᎐448.
w6x Sullivan ER. Molecular genetics of biosurfactant production.
Curr Opin Biotechnol 1998;9:263᎐269.
w7x Pruthi V, Cameotra SS. Novel sucrose lipid produced by
䢇 Serratia marcescens and its application in enhanced oil recov-
ery. J Surf Det 2000;3:533᎐537.
Sucrose lipids with excellent ability to emulsify hydrocarbons, in-
cluding crude oil, were produced. The surfactant was a mixture of
two hydroxyalkanoic acid esters of sucrose.
w8x Zuber P, Nakano MM, Marahiel MA. Peptide antibiotics. In:
Zuber P, Nakano MM, Marahiel MA, editors. Bacillus subtilis
and other gram-positive bacteria. Washington DC: American
Society for Microbiology, 1993:897᎐916.
w9x Sen R, Swaminathan T. Application of response-surface
methodology to evaluate the optimal environmental condi-
tions for the enhanced production of surfactin. Appl Mi-
crobiol Biotechnol 1997;47:358᎐363.
w10x Vollenbroich D, Ozel M, Vater J, Kamp RM, Pauli G.
Mechanism of inactivation of enveloped viruses by the bio-
surfactant surfactin from Bacillus subtilis. Biologicals
1997;25:289᎐297.
w11x Shinoda K, Carlsson A, Lindman B. On the importance of
hydroxyl groups in the polar headgroup of nonionic surfac-
tants and membrane lipids. Adv Colloid Interface Sci
1996;64:253᎐271.
w12x Garcia MT, Ribosa I, Campos E, Leal JS. Ecological proper-
ties of alkylglucosides. Chemosphere 1997;35:545᎐556.
w13x Uppgard L, Sjostrom
¨ ¨ M. Multivariate quantitative structure-
activity relationships for the aquatic toxicity of alkyl polyglu-
cosides. Tenside Surf Det 2000;37:131᎐138.
w14x Busch P, Hensen H, Tesmann H. Alkylpolyglycoside-eine
neue Tensidgeneration fur ¨ die Kosmetik. Tenside Surf Det
1993;30:116᎐121.
w15x Hill K, von Rybinski W, Stoll G. Alkyl polyglycosides: tech-
nology, properties and applications. Weinheim, Germany:
VCH, 1997.
158 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
w16x von Rybinski W. Alkyl glycosides and polyglycosides. Curr
Opin Colloid Interface Sci 1996;1:597.
w17x von Rybinski W, Hill K. Alkyl polyglycosides ᎏ properties
and applications of a new class of surfactants. Angew Chem
Int Ed 1998;14:4699᎐4709.
w18x von Rybinski W, Hill K. Alkyl polyglycosides. In: Holmberg
K, editor. Novel surfactants. New York: Marcel Dekker,
1998:31᎐85.
w19x Soderman
¨ O, Johansson I. Polyhydroxyl-based surfactants
and their physicochemical properties and applications. Curr
Opin Colloid Interface Sci 2000;4:391᎐401.
w20x Nilsson F, Soderman
¨ O, Reimer J. Phase separation and
䢇 aggregate᎐aggregate interactions in the C9G1rC10G1 ␤-al-
kylglucosiderwater system. A phase diagram and NMR self-
diffusion study. Langmuir 1998;14:6396᎐6404.
The paper discusses the mechanism behind the liquid᎐liquid phase
separation that occurs in the micellar region; this region is split
into one concentrated and one dilute phase, both containing mi-
celles.
w21x Nilsson F, Soderman
¨ O, Johansson I. Four different C8G1
alkylglucosides. Anomeric effects and the influence of straight
vs. branched hydrocarbon chains. J Colloid Interface Sci
1998;203:131᎐139.
w22x Dorfler HD, Gopfert A. Lyotropic liquid crystals in binary
systems n-alkylglucosidesrwater. J Dispersion Sci Technol
1999;20:35᎐58.
w23x Hantzschel D, Schulte J. Thermotropic and lyotropic proper-
ties of n-alkyl-␤-D-glucopyranoside surfactants. PCCP
1999;1:895᎐904.
w24x Sierra ML, Svensson M. Mixed micelles containing alkylgly-
䢇 cosides: effect of the chain length and the polar head group.
Langmuir 1999;15:2301᎐2306.
Formation of micelles in mixtures of alkylglycosides and a conven-
tional surfactant was studied. A strong negative value of the inter-
action parameter, ␤ , indicating attractive interaction, was obtained
for the mixture of alkylglycoside and the cationic surfactant. A
somewhat smaller negative ␤-value was obtained for the mixture of
a glycoside surfactant with the anionic surfactant. No interaction
was seen for the alkylglycoside᎐alcohol ethoxylate mixture.
w25x Holland PM, Rubingh DN. Mixed surfactant systems. ACS
Symp Ser 501. Washington: American Chemical Society, 1992.
w26x Liljeqvist P, Kronberg B. Comparing decyl ␤-maltoside and
octaŽethylene glycol.mono n-decyl ether in mixed micelles
with dodecylbenzene sulphonate-1. Formation of mixed mi-
celles. J Colloid Interface Sci 2000;222:159᎐164.
w27x Liljeqvist P, Kronberg B. Comparing decyl ␤-maltoside and
octaŽethylene glycol.mono n-decyl ether in mixed micelles
with dodecylbenzene sulphonate-2. Interaction of mixed mi-
celles with polyvinylpyrrolidone. J Colloid Interface Sci
2000;222:165᎐169.
w28x Laughlin RG, Fu Y-C, Wireko FC, Scheibel JJ, Munyon RL.
The physical science of N-dodecanoyl-N-metylglucamine and
its aqueous mixtures. In: Holmberg K, editor. Novel surfac-
tants. New York: Marcel Dekker, 1998:31᎐85.
w29x Zhu Y-P, Rosen MJ, Vinson PK, Morral SW. Surface proper-
䢇 ties of N-alkanoyl-N-metylglucamines and related materials.
J Surf Det 1999;2:357᎐362.
The effect of the number of hydroxyl groups in polyol surfactants
on physico-chemical parameters such as CMC, area per molecule
and Krafft point was investigated. The values of all three parame-
ters were reduced with decreasing number of hydroxyl groups.
w30x Allen DK, Tao BY. Carbohydrate-alkyl ester derivatives as
biosurfactants. J Surf Det 1999;2:383᎐390.
w31x Fregapane G, Sarney DB, Vulfson EN. Facile chemoenzy-
matic synthesis of monosaccharide fatty acid esters. Biocatal-
ysis 1994;11:9᎐18.
w32x Sarney DB, Kapeller H, Fregapane G, Vulfson EN.
Chemoenzymatic synthesis of disaccharide fatty acid esters. J
Am Oil Chem Soc 1994;71:711᎐714.
w33x Baker IJA, Matthews B, Suares H et al. Sugar fatty acid ester
䢇 surfactants: structure and ultimate aerobic biodegradation. J
Surf Det 2000;3:1᎐12.
A series of sugar esters with different sized sugar headgroups were
investigated with respect to ultimate aerobic biodegradation. Sugar
esters with either a sulfonate or an ethyl group in ␣-position of the
fatty acyl chain were also included. It was found that variations of
the sugar headgroup did not much affect the rate of biodegrada-
tion. The ␣ substituted surfactants degraded at a lower rate.
w34x Baker IJA, Willing I, Furlong DN, Grieser F, Drummond CJ.
䢇䢇 Sugar fatty acid ester surfactants: biodegradation pathways. J
Surf Det 2000;3:13᎐28.
The paper discusses biodegradation pathways for sugar ester sur-
factants. It was shown that normal nonionic sugar esters degraded
via initial ester bond hydrolysis. Sugar ester surfactants with either
a sulfonate or an ethyl group in ␣-position of the fatty acyl chain
degraded by another, slower mechanism, ␻-oxidation of the acyl
chain.
w35x Baker IJA, Furlong DN, Grieser F, Drummond CJ. Sugar
fatty acid ester surfactants: base-catalyzed hydrolysis. J Surf
Det 2000;3:29᎐32.
w36x Boyd BJ, Drummond CJ, Krodkiewska I, Grieser F. How
䢇 chain length, headgroup polymerization, and anomeric con-
figuration govern the thermotropic and lyotropic liquid crys-
talline phase behavior and the air᎐water interfacial adsorp-
tion of glucose-based surfactants. Langmuir 2000;16:
7359᎐7367.
The effect of alkyl chain length, headgroup polymerization and
anomeric configuration of a series of sugar esters on physico-chem-
ical properties were investigated.
w37x Gao C, Millqvist-Fureby A, Whitcombe MJ, Vulfson EN.
䢇䢇 Regioselective synthesis of dimeric Žgemini. and trimeric
sugar-based surfactants. J Surf Det 1999;2:293᎐302.
An impressive series of gemini sugar surfactants Žand some trimeric
surfactants . were synthesized by a combination of enzymatic and
organo-chemical routes.
w38x Vulfson E. Enzymatic synthesis of surfactants. In: Holmberg
K, editor. Novel surfactants. New York: Marcel Dekker,
1998:279᎐300.
w39x Valivety R, Jauregi P. Chemo-enzymatic synthesis of amino
acid-based surfactants. J Am Oil Chem Soc 1997;74:879᎐886.
w40x Infante MR, Peres ´ L, Pinazo A. Novel cationic surfactants
from arginine. In: Holmberg K, editor. Novel surfactants.
New York: Marcel Dekker, 1998:87᎐114.
w41x Pinazo A, Infante MR, Izquierdo P, Solans C. Synthesis of
䢇䢇 arginine-based surfactants in highly concentrated water-in-oil
emulsions. J Chem Soc Perkin Trans 2000;2:1535᎐1539.
A highly concentrated water-in-oil emulsion was used as a reaction
medium for acylation of arginine methyl ester with dodecanoic
acid. It was demonstrated that such emulsions should be seen as an
alternative to apropic, polar solvents such as dimethylformamide
and also to microemulsions.
w42x Clapes P, Moran C, Infante MR. Enzymatic synthesis of
arginine-based cationic surfactants. Biotechnol Bioeng
1999;63:333᎐343.
w43x Piera E, Dominguez C, Clapes P, Erra P, Infante MR. Quali-
tative and quantitative analysis of new alkyl amide arginine
surfactants by high performance liquid chromatography and
capillary electrophoresis. J Chromatogr A 1999;852:499᎐506.
w44x Piera E, Infante MR, Clapes P. Chemo-enzymatic synthesis
of arginine-based gemini surfactants. Biotechnol Bioeng
2000;70:323᎐331.
 K. Holmberg r Current Opinion in Colloid & Interface Science 6 (2001) 148᎐159
w45x Pinazo A, Wen XY, Perez L, Infante MR, Franses EI. Aggre-
gation behavior in water of monomeric and gemini cationic
surfactants derived from arginine. Langmuir 2000;15:
3134᎐3142.
w46x Pegiaduo S, Perez L, Infante MR. Synthesis, characterization
䢇 and surface properties of 1-N-L-tryptofan-glycerol-ether sur-
factants. J Surf Det 1999;3:517᎐525.
A new type of tryptophan-based surfactants were synthesized and
characterized with respect to CMC, surface tension and area per
molecule at the air᎐water interface. The tryptophane surfactants
showed very low CMC values.
w47x Vives MA, Macian M, Seguer J, Infante MR, Vinardell MP.
Hemolytic action of anionic surfactants of the diacyl lysine
type. Comp Biochem Physiol C Pharmacol Toxicol En-
docrinol 1997;118:71᎐74.
w48x Vives MA, Infante MR, Garcia E, Selve C, Maugras M,
Vinardell MP. Erythrocyte hemolysis and shape changes in-
duced by new lysine-derivate surfactants. Chem Biol Interact
1999;118:1᎐18.
w49x Folmer BM, Holmberg K, Gottberg Klingskog E, Bergstrom ¨ time to reach equilibrium surface tension was very long, more than
䢇 K. Fatty amide ethoxylates; synthesis and self-assembly. J
Surf Det 2001 Žin press..
A series of fatty acid monoethanolamide ethoxylates with varying
number of double bonds in the acyl chain and with varying number
of oxyethylene groups in the polar headgroup was investigated. The
influence of the amide bond and of the unsaturations on adsorption
at the air᎐water interface and on self-assembly was investigated.
w50x Folmer BM, Nyden ´ M, Holmberg K. Micellization and ad-
159
B 䢇 sorption of a series of fatty amide ethoxylates. J Colloid
Interface Sci 2001; in press.
A series of fatty acid monoethanolamide ethoxylates with varying
number of double bonds in the acyl chain and with varying number
of oxyethylene groups in the polar headgroup was investigated. The
influence of the amide bond and of the unsaturations on adsorption
at a hydrophobic solid surface and on micellization was investigated
by ellipsometry and by NMR.
w51x Svensson M. Surfactants based on sterols and other alicyclic
compounds. In: Holmberg K, editor. Novel surfactants. New
York: Marcel Dekker, 1998:179᎐200.
w52x Folmer BM. Sterol surfactants; from synthesis to applica-
tions. In: Karsa DR, editor. Annual surfactant reviews, Vol. 3.
Sheffield, UK: Sheffield Academic Press, 2001 Žin press..
w53x Folmer BM, Svensson M, Holmberg K, Brown W. The physic-
䢇 ochemical behavior of phytosterol ethoxylates. J Colloid In-
terface Sci 1999;213:112᎐120.
A series of sterol ethoxylates was investigated. It was found that the
two hours. This probably reflects both exchange reactions at the
surface and a slow reorientation of molecules at the air᎐water
interface.
w54x Folmer BM, Nyden ´ M. Structure and dynamics in the micel-
lar and cubic phase of an ethoxylated phytosterol surfactant.
Langmuir Žsubmitted..
w55x Folmer BM. Ph.D. Thesis. Royal Institute of Technology,
Stockholm, Sweden, 2000.