1

Published in Topics in Catalysis, vol. 27 (2004) 105-117

The original publication is available at www.springerlink.com

http://www.springerlink.com/content/p106389747319371/

FULL TITLE:

Solid catalysts for the synthesis of fatty esters of glycerol, polyglycerols and

sorbitol from renewable resources

SHORT TITLE:

Fatty acid esters of polyols

AUTHORS:

Carlos Márquez-Alvarez, Enrique Sastre and Joaquín Pérez-Pariente*

Instituto de Catálisis y Petroleoquímica, CSIC. C/ Marie Curie, S/N. Cantoblanco,

28049-Madrid, Spain

*To whom correspondence should be addressed. E-mail: jperez@icp.csic.es

2

ABSTRACT:

Partial esters of polyols such as glycerol, polyglycerols and sorbitol are non-

ionic surfactants used as emulsifiers in the food, detergents and cosmetics industries.

The commercial large-scale production of these chemicals relies on traditional

homogeneous catalysis processes making use of strong mineral acids or alkalis. This

technology possesses severe drawbacks, related to the generation of large amounts of

by-products, high energy demand, and heterogeneity of the ester mixtures obtained. The

development of new processes based on more selective solid catalyst has therefore a

great economical interest. This contribution reviews the scientific and technical

literature in this field. It is proposed that solid catalysts based on MCM-41 and other

mesostructures could give rise to the development of new, more efficient processes for

the large-scale synthesis of fatty acid esters of polyols.

KEYWORDS:

Fatty acid esters; polyols; monoglycerides; sorbitol; polyglycerol; non-ionic surfactants;

emulsifiers; heterogeneous catalysts; zeolites; mesoporous catalysts; esterification;

transesterification; glycerolysis.

3

1. Introduction

There is an increasing trend in the chemical industry to introduce new processes

that should meet requirements such as generation of nearly zero waste chemicals,

reduced energy input and use, whenever it is possible, of raw materials derived from

non fossil primary sources. Besides, a higher efficiency is generally required, in terms

of optimum transformation of the feedstocks into the desired product. Many of these

requirements challenge conventional processes based upon old technology, and act

therefore as a strong driving force for innovation, which includes the use of

heterogeneous catalysts.

Despite the enormous impact that solid catalysts have on the production of

chemicals, ranging from large scale refining and petrochemical processes to fine and

specialty compounds, their potential is far from being fully developed in some key areas

of the chemical industry. A notorious example of the prevalence of aged technology

based on homogeneous catalysts is provided by the commercial production of fatty acid

esters of polyols. Among them, those of glycerol, polyglycerols and sorbitol will be

most relevant to this review.

Fatty acid esters of these polyols, particularly those of glycerol (about 80% of

the total), are widely used as emulsifiers in a large variety of applications, particularly in

food processing. The hydrophobic acyl group of the ester is formed by a hydrocarbon

chain containing from 10 to 20 carbon atoms, whereas the polar “head” of the surfactant

is formed by hydroxyl groups, the number of which varies from 2 for monoglycerides,

to around 5 for some polyglycerol esters.

The emulsifying properties of the esters are basically dependent upon their

hydrophilic-lipophilic balance, best known as HLB [1]. According to Griffin’s scale, the

4

higher the HLB value, the greater the hydrophilicity of the surfactant. A large range of

HLB values can be covered by proper combination of the number and nature of the

hydrophobic hydrocarbon “tail” and the number of free hydroxyl groups contained in

the polar “head” of the ester. Therefore, it is possible to select specific surfactants for

targeted applications as a function of their HLB values. For practical purposes, and due

to the strong hydrophobic character of the fatty acyl group, esters containing an

OH/fatty chain ratio equal or higher than two are in general best preferred as

emulsifiers. Thus, in the case of glycerol, only the monoesters are good emulsifying

agents. The same applies to sorbitol, despite this polyol contains six hydroxyl groups.

This is so because, due to its low thermal stability, sorbitol experiences anhydrization at

the temperature at which commercial esterification is currently carried out (about 150ºC

or higher). Under these conditions, sorbitol transforms into a mixture of inner ethers,

sorbitans and sorbides, which possess 4 and 2 free hydroxyl groups, respectively.

Therefore, only the so-called “sorbitan monoesters” exhibit suitable emulsifying

properties.

Details on polyglycerol chemistry will be given below, but it suffices to say here

that the number of hydroxyl groups is given by n + 2, being n the number of glycerol

units making the polyglycerol molecule. For lower polyglycerols, monoesters are

preferred, whereas diesters and eventually even triesters of polyglycerols of high

molecular weight could still have good emulsifying properties.

Figure 1 displays the evolution of the number of publications on

monoglycerides, polyglycerol esters and sorbitol monoesters, covering application,

synthesis and processing, both in journals and patents. Noticeable differences in

scientific and technological activities among the three groups of esters are observed.

Significant research effort in monoglyceride chemistry started in the thirties of the last

5

century, increasing steadily since then. A relatively large number of journal publications

have been continuously released, although the relative number compared to patents

started to decrease in recent years. Research on fatty esters of polyglycerol showed a

significant level in the sixties and seventies, but it started to grow in the early eighties at

a much faster rate than that of monoglycerides. It is remarkable that this intense

research activity has given rise to a large number of patents, while very few

contributions in this field have been published in scientific journals. The

scientific/technological activities concerning sorbitol esters have grown at much lower

rate that those of the two other polyol esters, and very few journal reports have been

published so far on sorbitol monoesters. This fact might reflect the relative commercial

impact of these esters.

The statistics of Figure 1 clearly show strong and increasing innovation activities

on polyglycerol esters, which nevertheless remain virtually unexplored outside the

industrial R + D departments. An analysis of the nationalities of the patent applicants

(Fig 1b, inset) reveals the leadership of Japanese companies, particularly in the

expanding polyglycerol esters market.

The commercial synthesis of the fatty acid esters of polyols is carried out

basically by two different procedures: direct esterification of the polyol, usually

catalysed by a homogeneous acid, such as sulfuric and sulfonic acids, or

transesterification of triglycerides and polyols, catalysed by alkaline hydroxides like

NaOH, KOH, Ca(OH)2, or the sodium salts of lower aliphatic alcohols, like methanol. It

can be estimated that approximately 80% of the manufacturing processes belong to the

first pathway.

The fatty acids and the glycerol are obtained basically by splitting of fats and

vegetable oils with water at high temperature [2]. Sorbitol is also derived from non-

6

fossil sources, as it is obtained by catalytic hydrogenation of glucose, being starch the

basic commercial source of this sugar. Therefore, a variety of abundant raw materials

resulting from agriculture activities are readily available for the synthesis of the polyol

esters described in this work.

Here we provide an overview of the synthesis procedures of these polyol esters

and discuss the potential that exists for developing new, more efficient synthesis

processes based on solid catalysts. Enzyme catalysed synthesis of polyol esters is not

covered by this review. An intense research effort is being carried out on the use of

lipases to synthesise esters of glycerol, sorbitol and sugars [3-6] and even polyglycerol

[7,8]. These studies show that enzymes have an enormous potential as catalysts in those

processes in which high regioselectivity is required. However, for large-scale synthesis

of the more common surfactants, these processes are not yet competitive due to the high

cost of enzymes and their low productivity.

2. Monoglycerides.

The product of the direct esterification or transesterification of glycerol is a

mixture of mono- di- and some triglyceride, and dissolved glycerol, which typically

contains 40-60% monoglyceride and 35-45% diglyceride. Most of the fatty

monoglycerides are obtained by transesterification of triglycerides, either from

vegetable or animal sources, and glycerol. The basic characteristics of the process have

been comprehensively reviewed [9]. Scheme 1 describes the overall simplified chemical

equation for glycerolysis of fats and oils.

The actual transesterification process is more complex than that described by

Scheme 1, in such a way that di- and even triester are simultaneously obtained. Typical

composition of the esters blends has been given above. Besides, Scheme 1 does not

7

reflect the formation of the 2-acyl or β-monoglyceride, which proportion remains below

10% at low temperature, but increases with the reaction temperature.

The basic catalysts used in the process, generally KOH and Ca(OH)2, are

neutralised with phosphoric acid after the reaction, and the formed salts removed.

Therefore, undesired waste chemicals are obtained. The use of solid catalysts suitable

for regeneration would obviously avoid these drawbacks. However, this is not the only

driving force to modify the classical process. The glycerolysis reaction requires a high

temperature, in order to increase the solubility of glycerol in the fat phase. An additional

energy-consuming step is needed to separate the monoglyceride from the crude reaction

product by vacuum molecular distillation, in order to obtain the monoester with a purity

of 90% at least. Therefore, the key aspect of the process is the low selectivity to the

monoglyceride. It must be stressed that direct esterification of the glycerol with a fatty

acid, by using homogeneous acid catalysts, does not improve the monoglyceride

selectivity.

These considerations have led to explore alternative synthesis routes that involve

in most cases a change of the nature of the catalyst and the chemical feedstocks. The use

of inorganic solid catalysts will now be discussed.

2.1. Base catalysis.

Solid basic materials such as MgO, Al-Mg hydrotalcites as well as Cs-

exchanged sepiolite and mesoporous MCM-41 have been used as catalysts for the

glycerol transesterification with triglycerides [10]. MgO and Al-poor hydrotalcites

(Al/(Al+Mg) = 0.20) behave as active catalysts for triolein glycerolysis, but a

glycerol/fat molar ratio of as high as 12 is required to obtain a monoglyceride yield of

70% at 97% conversion (MgO catalysts at 513 K). The large excess of glycerol requires

8

an increase of the reactor capacity of over 60%, compared to stoichiometric reaction

mixtures, and increases the thermal demand by a factor of nearly 2. This example

illustrates the negative influence of the low solubility of the hydrophilic glycerol in the

fat phase. The use of organic solvents would have benefits on the overall reaction but,

for safety reasons, it is avoided when producing monoglycerides to be used as food

additives.

Other esters have been used for glycerol transesterification. Metal oxides like

MgO, CeO2, La2O3 and ZnO have been used as solid base catalysts for the

transesterification of glycerol with stoichiometric amounts of methyl stearate [11] in the

absence of solvent. The catalysts are active, but the selectivity to mono-, di- and

triesters is similar to that obtained by using homogeneous basic catalysts (40%

monoester at 80% conversion).

2.2. Acid catalysis.

A variety of solid acids have been used to catalyse the esterification of glycerol

with fatty acids. Zeolite materials meet, in principle, the basic requirements for

successful replacement of homogeneous acids, as they possess sufficient acid strength.

Indeed, they have been used for the esterification of monoalcohols with low molecular

weight carboxylic acids [12]. When applied to glycerol esterification with fatty acids,

the relatively small pore aperture of zeolite materials would have the potential to reduce

the formation of the bulky di- and triesters, increasing therefore the selectivity to

monoglycerides. Table 1 summarises the most relevant results reported by using zeolite

catalysts for the esterification of glycerol with lauric and oleic acids.

Large-pore tridirectional 12-membered-ring structures, like faujasite and Beta,

are more active and selective than 10-membered-ring structures, like ZSM-5, or

9

unidirectional 12-MR materials like mordenite [13-18]. Influence of the aluminum

content of the framework has been reported for faujasite [14] and Beta [18]. For both

structures, the conversion and monoglyceride selectivity increase with the Si/Al ratio,

and a maximum is observed at Si/Al ratios in the range 35-40. Absence of crystal size

effect for dealuminated faujasite (Si/Al = 14) in the range ∼ 0.3 – 3.5 μm has been taken

as a proof that the esterification with lauric acid takes place inside the intracrystalline

voids of the zeolite [16]. In this way, transition state selectivity inside the supercages

suppresses the consecutive esterification of the monoglycerides to bulkier glycerides. A

similar effect would explain the high selectivity of zeolite Beta. However, as di- and

even traces of triglycerides are readily observed even at conversion higher than 10%, a

significant contribution of the external surface to the overall reaction can not be totally

excluded. Indeed, both the activity and the selectivity with oleic acid are much lower

(Table 1), which suggests a severe pore size limitation to the reaction rate.

In order to avoid the pore size constraints of the zeolite materials on the reaction,

cross-linked porous polymers, like ion-exchange resins containing sulfonic acid groups,

have been used as catalysts for the esterification of glycerol with fatty acids [19]. It has

been observed that gel resins, like Amberlyst, are more active than macroporous resins,

due to their swelling capacity. Both the activity and selectivity show a strong

dependency on the glycerol/fatty acid ratio (R), as well as on the chain length of the

fatty acid. The reaction is faster with lauric acid than with oleic acid. For lauric acid,

using Amberlyst 31 as catalyst, the activity decreases with R, but the monolaurate

selectivity increases up to R = 3. For R = 6 and 140ºC, the selectivity to monolaurate is

83% at 65% conversion. The selectivity for oleic acid at the same conversion level is

lower. However, the selectivity is higher than that obtained with zeolite materials. No

information on catalyst regeneration is provided in these works.

10

2.3. Acidic mesoporous materials.

The benefits of the swelling of the ion-exchange resins on the overall catalyst

performance in glycerol esterification evidence the positive influence of increasing the

accessibility of the reagents to the active centres. Therefore, the family of the ordered

mesoporous materials appears as a very interesting alternative for the development of

solid catalysts, as it offers the possibility to overcome the pore size limitation

characteristic of zeolite materials. The pore size of these materials is typically

comprised in the range 2-10 nm, and several structures possessing different pore

dimensions and channel and voids architecture are available. Unfortunately, the strength

of the acid sites present in conventional MCM-41 materials is low [20] and

consequently their activity and monoglycerides selectivity are low [14].

As the sulfonic acid groups present in the ion-exchange resins are very active in

catalysing the esterification reaction, they have been attached to pure silica mesoporous

frameworks in order to obtain mesoporous catalysts. The behaviour of mesoporous

materials containing R-SO3H groups in the esterification of glycerol with fatty acids has

been reviewed recently [21], and some of the most relevant aspects are briefly

commented here. MCM-41 materials containing propylsulfonic acid groups have been

reported to catalyse selectively the esterification of glycerol with lauric acid for

equimolar amounts of both reagents, and in the absence of solvent. A monolaurin yield

of 53% was obtained, whereas for oleic acid the monoglyceride yield dropped to 31%

[22]. Following this pioneering work, further improvement of the SO3H-containing

MCM-41 catalysts prepared by a one-pot synthesis procedure was obtained through the

optimisation of the pore arrangement by using mixtures of surfactants having different

chain length [23], as well as mixtures of n-dodecylamine and

11

hexadecyltrimethylammonium cations in the synthesis gel [24]. While these

modifications increase the activity and monoglyceride yield of the reaction, a major

improvement in catalyst efficiency results by tuning the hydrophilic/lipophilic balance

of the catalyst surface [25]. The anchoring of hydrocarbon moieties like methyl and

propyl groups decreases substantially the strong affinity of the catalyst to hydrophilic

molecules, like water and glycerol. As a consequence, the monolaurin yield raises to

66%, while that of monoolein is close to 50% [26]. The catalysts can be reused,

although they show a small loss of activity [23]. Sulfur leaching is small. After 8 h at

100ºC, the S content in the reaction mixture is lower than 80 ppm, although it increases

with the temperature and reaction time.

Organosulfonic groups other than propylsulfonic have been also attached to the

MCM-41 surface by one-pot synthesis procedures. For example, the intrinsic activity of

the phenylsulfonic acid-containing material is higher than that with propylsulfonic acid,

although the selectivity is lower [27].

MCM-41 catalysts have been also prepared from gels containing 3-

mercaptopropyl(methyl)dimethoxysilane. The presence of a methyl group attached to

the silicon atom bonded to the propylsulfonic moiety increases substantially both the

activity and selectivity. At 50% conversion, the selectivities to monolaurin and

monoolein were 77% and 65%, respectively [28].

More recent results evidence the influence of the n-alkyl chain length of the

organosulfonic acid moiety in glycerol esterification [29]. Catalysts obtained by

sulfonation of MCM-41 materials containing vinyl groups exhibit better activity and

selectivity than propylsulfonic acid-containing materials.

The influence of the pore size of functionalised MCM-41 catalysts containing

sulfonic acid groups has not yet been properly addressed. Catalysts having lower

12

activity and selectivity are achieved by replacing methyl groups by propyl groups in

catalysts containing both n-alkyl and propylsulfonic moieties [26]. When performing

studies on the influence of pore size in functionalised materials, it has to be remarked

that the effective pore size varies with the S-loading and with the size of the functional

group [30].

Comparison of the results obtained from lauric and oleic acids suggests that the

optimum pore size for the reaction varies according to the chain length of the fatty acid

[26]. Indeed, it has been shown recently that the selectivity to glycerol monostearate of

MCM-41 functionalised materials having a S-loading of 0.3 meq·g-1 decreases from

47% to 40% at 50% acid conversion, as the pore size increases from 2.6 to 3.7 nm [31].

A further decrease of the pore size to 2.0 nm does not change the selectivity, but

decreases the turnover values by 80%. These results suggest that there is a noticeable

contribution of the external surface to the overall reaction for small pore materials.

Mesopore structures other than MCM-41 have been explored as catalysts for

glycerol esterification. SBA-15, which possesses unidirectional channels arranged in

hexagonal symmetry and interconnected by micropores, has been functionalised with

propylsulfonic acid groups (0.78 meq·g-1). Its specific activity (TON) and monoolein

selectivity are lower than that of MCM-41 catalysts, which has been attributed to both

the presence of non accessible micropores, and to the large mesopore size, reaching 3.2

nm [32]. SBA-12, a structure with interconnected spherical cavities in a cubic-close

packing [33], and SBA-2, an intergrowth of hexagonal and cubic mesostructures, both

of which contain a complex set of narrow channels that interconnect the large cavities,

show low activity and selectivity [32,34]. Therefore, none of these complex structures

afford better catalysts than conventional MCM-41.

13

2.4. Basic MCM-41 catalysts.

Despite the improvement in monoglyceride selectivity achieved by using

mesoporous acid catalysts, a monoglyceride yield of 90%, that would be needed in

order to avoid the costly molecular distillation of the ester mixtures, has not yet been

reached.

Nucleophilic ring opening of glycidol with fatty acids (Scheme 2) has been

explored as a selective and convenient route toward nearly pure monoglycerides [35].

To accomplish this purpose, MCM-41 all-silica materials have been functionalised by

grafting R-propyl moieties, where R stands for NH2 and a variety of cyclic secondary

amines. In this synthesis route, capping of the residual silanol groups with trimethylsilyl

groups, for example, is required to prevent glycidol polymerisation. However, reused

catalysts also show improved activity, due to the grafting of the residual silanol groups

with glycerol molecules during the first run. Catalysts containing tertiary and hindered

amines are more efficient than those having primary amines, owing to steric effects

[36]. Indeed, the activity correlates approximately with the base strength of the organic

base. Following this procedure, monolaurate yields of 90% have been reported. This

synthesis route would be attractive for monoglyceride production, were it not for two

reasons: 1) the glycidol is much more expensive than glycerol, the price of which goes

down as its production increases due to the increasing use of biodiesel, and 2) toluene is

used as solvent in the addition reaction of glycidol to fatty acid. Both factors make this

route unattractive for commercial monoglyceride production.

3. Polyglycerol esters.

As it has been described above, the emulsifying properties of the polyglycerol

esters (PGE) depend basically upon the length of the polyglycerol chain, the degree of

14

esterification and the molecular weight of the fatty acid. Indeed, the preparation of PGE

suitable as additives in foodstuffs is a process much more complex than that of

monoglycerides, and in general involves two different steps: 1) polymerisation of

glycerol in the presence of a small amount of alkali (base catalysis), and 2) esterification

of the resulting polyglycerol. Scheme 3 offers an overview of the process.

3.1. Polyglycerol synthesis.

Glycerol polymerisation is carried out at a temperature above 200ºC under

alkaline conditions. The condensation reaction involves the terminal hydroxyl groups of

glycerol molecules, as they are more reactive than the 2-hydroxy group. However, some

ramified and even cyclic by-products are always formed. Carbon dioxide or nitrogen is

bubbled through the reaction mixture to prevent dehydration of glycerol to undesired

acrolein. Several bases have been tested as catalysts, including hydroxides, carbonates

and oxides of several metals [37]. It has been found that the carbonates are more active

than hydroxides, despite the weaker base character of the former. This has been

attributed to the better solubility of carbonates in the glycerol and in the polymeric

product at elevated temperatures. Oxides like MgO, CaO and ZnO are less active

because of lack of solubility.

The polymerisation reaction proceeds by chance under homogeneous conditions

and therefore a mixture of products with different molecular size is obtained [38-44].

Low amount of catalyst (< 1 mol%) and relatively short reaction time are required to

exert some control on the polymerisation degree, in order to favour a mixture of

products having, as average, low mean polymerisation degree. Product distribution of

polyol is important, as the polyglycerol moiety of the ester should meet some

specifications to be used as food additive. According to EC regulations, the content of

15

di-, tri- and tetraglycerol should predominate, whereas the content of polyglycerols

equal to or higher than heptaglycerol must be low [45]. FDA regulations allow the use

of polyglycerols up to and including decaglycerol [46].

The alkaline polymerisation of glycidol affords a more selective process,

provided that a strict control of the reaction parameters is exerted [47]. As an example,

linear octaglycerol has been prepared from glycidol by using KOH as catalyst, at a

reaction temperature of ca. 120ºC [48]. The secondary hydroxyl content was 72%

(theoretical maximum = 80%). In contrast, the octaglycerol obtained by conventional

glycerol condensation at 230-250ºC, using NaOH as catalyst, contained 41% secondary

hydroxyls, due to the presence of residual glycerol and oligomers of lower molecular

weight. Besides, the condensation reaction of glycerol and glycidol catalysed by acids

has been claimed to render polyglycerols of high polymerisation degree with low colour

values [49-51].

Porous solid catalysts could exert some shape-selective effect on the course of

the polymerisation reaction. Nevertheless, reports on such catalysts are extremely

scarce. Zeolites NaX and NaA have been used as catalysts for the selective production

of diglycerol [52]. Cs-exchanged zeolite X shows high selectivity to di- (62%) and

triglycerol (33%), at 70% conversion, while only a 4% of tetraglycerol is obtained

[53,54]. On the other hand, medium pore Cs-containing zeolites like ZSM-5 are less

active and selective.

The influence of pore size on the polymerisation reaction is evidenced by the use

of basic MCM-41 catalysts [55]. The Cs-impregnated material affords the best results,

whereas Mg and La-containing catalysts favoured glycerol dehydration and the

formation of acrolein. As an example, for a conversion of 80%, the selectivity to di- +

triglycerol (di-/triglycerol > 1) is close to 90%. However, a severe leaching of Cs is

16

observed, and a decrease of the Cs loading of the catalysts between 95% and 70% has

been reported. Indeed, no X-ray diffraction peak at low angle is observed for Cs loading

equal or higher than 18.6·10-4 eq. g –1 . Despite the severe Cs leaching, the catalysts still

show a remarkable high selectivity to di- + triglycerol.

These results are promising, but nevertheless the selectivity of these mesoporous

catalysts approaches that of Cs-exchanged faujasite X. They also evidence that there is

much need for investigation into reusable, shape-selective catalysts for controlled

glycerol polymerisation.

3.2. Esterification of polyglycerols.

The esterification of polyglycerols leads to mixtures with a broad range of

different compounds, owing to the large number of hydroxyl groups and the intrinsic

complexity of the commercially available polyglycerols mixtures. The hydrophilic-

lipophilic properties of the esters, expressed by means of the HLB concept [1], are a

function of the molecular weight of the esterified fatty acid, the degree of esterification

of hydroxyl groups and the number of glycerol units forming the polyglycerol

backbone. As an example, Figure 2 shows the range of HLB values of triglycerol esters

as a function of the carboxylic acid chain length and the degree of esterification. It can

be concluded that a decrease of the molecular weight of the fatty acid results in an

increase in the hydrophilic nature of the ester (the HLB value increases). On the other

hand, an increase in the number of fatty acid moieties that react with the polyol

molecule will result in a more lipophilic product. For a given fatty acid and

esterification degree, the hydrophilic character increases with the molecular weight of

the polyglycerol (higher hydroxyl/fatty acid ratio). This is depicted in Figure 3 for

stearic acid esters of several polyglycerols. As shown in Figure 3, an HLB range from

17

about 3 to 8 can be covered by different polyesters of tri- to decaglycerol, but

monoesters of high molecular weight polyols are required to get highly hydrophilic

compounds (HLB ≥ 12).

Figure 3 underlines the strong impact that the chemical composition of the

polyglycerol has on the emulsifying properties of the esters resulting from the

esterification process. A narrow distribution of the molecular weight of polyglycerols

would be needed in order to obtain products having well-defined HLB values.

The processes for manufacturing PGE are summarised in Scheme 4. The esters

are most simply made by direct esterification of the polyol with a free organic acid. The

reaction can proceed at T > 200ºC without using a catalyst [37,56-58], or at T < 200ºC

by using either alkali [37,56,59-65] or acid catalysts [40,56,66-68]. The reaction can

also be carried out by transesterification of a polyol with a triglyceride or a fatty acid

alkyl ester (methyl ester usually), generally in the presence of a suitable alkaline catalyst

[40,42-44,69]. Furthermore, it has been reported that partial esters of polyglycerol can

be obtained by acid catalysed polycondensation of monoglyceride or monoglyceride and

glycerol [70,71]. Finally, the addition polymerisation of glycidol to a fatty acid or to a

fatty acid monoglyceride catalysed by acids also renders polyglycerol esters [72-77].

These reactions must be carried out under an inert atmosphere to obtain good colour and

organoleptic properties. Some examples describe the use of moderate vacuum as well.

In the commercial manufacture of partial esters, the reaction will give a mixture

in which the fatty acid radicals are distributed among all available hydroxyl groups,

whose proportions are dictated basically by the thermodynamic constraints of the

reaction, namely temperature and molar ratio of the reagents. As an example, the

product obtained by reacting equimolar quantities of triglycerol and stearic acid

contains 41 wt% and 31 wt% of mono- and distearate, respectively [41]. As a

18

consequence, the HLB value of the product is lower than that of the pure monoester. To

approach the theoretical HLB values of the monoesters, a large excess of polyglycerol is

required.

Due to the inherent complexity of the esters mixtures that are obtained by these

methods, in most of the publications these mixtures are characterised in terms of their

saponification and OH values. These numbers are related to the mean polyglycerol

chain length and the average number of hydroxyl groups esterified, but do not provide

information on the distribution of products. To illustrate this, in Figure 4 the

saponification numbers have been plotted versus their corresponding hydroxyl numbers

for a series of oleic acid esters of polyglycerols. Figure 4 shows that many different

mixtures of esters can exhibit a given couple of saponification and hydroxyl values, and

that, especially in the region of lower values, these mixtures could contain a very large

number of esters with different esterification degree and polyglycerol chain length.

In Table 2 we have collected some representative examples on the manufacture

of PGE, selected among those reported both in journal and patent literature that provide

a more detailed composition of the ester mixture.

Several procedures have been reported to increase significantly the monoester

yield. The transesterification between polyglycerols and methyl esters of fatty acid has

been successfully achieved by using an emulsion medium, in the presence of alkaline

catalysts (NaOH) [69]. Anionic emulsifiers, Na-stearate particularly, are very effective

in promoting the reaction between tetraglycerol and methyl stearate. For a molar ratio of

polyol/acid = 2, a selectivity to the monoester of 98% at 100% conversion has been

reported by using 10 wt% (of the oil phase) of Na-stearate as emulsifier agent.

The reaction between glycidol and a fatty acid or fatty monoglyceride has also

been claimed to produce a high yield of polyglycerol monoester [73]. The reaction is

19

carried out at temperatures around 150ºC, in the presence of phosphoric acid as catalyst

(< 0.5 wt%), and the glycidol is slowly added to the mixture of fatty acid and catalyst.

Products containing more than 70% of the monoester are obtained using different

glycidol/fatty acid ratios. However, there is no indication on the polyglycerol chain

length distribution. Earlier reports on this same reaction use Lewis acids, like BF3 and

SnCl4, as catalysts [77].

To our knowledge, the use of porous, zeolite-like materials as catalysts for the

esterification of polyglycerols has not yet been reported. This is not surprising, owing to

the low performance of these materials in the selective esterification of the smaller

glycerol molecule. However, new opportunities for shape-selective catalysis would be

offered by the new generation of mesoporous materials, as their functionalisation with

appropriate acid or base organic moieties would probably afford active and selective

catalysts, as it has been outlined above for glycerol esterification.

4. Sorbitol monoesters

The sorbitol molecule (a hexitol) contains six hydroxyl groups able to be

esterified with carboxylic acids. However, this molecule dehydrates partially at

moderate temperature, forming as main products the hexitans (mainly 1,4-sorbitan and

minor amounts of 2,5-sorbitan), which contain four hydroxyl groups, and hexides

(mainly isosorbide), which contain two hydroxyl groups only (Scheme 5). Therefore,

the result of the esterification process of sorbitol with fatty acyl groups is a complex

mixture of different compounds, whose nature and relative proportions are governed by

the specific reaction conditions. As a consequence, the HLB value of the esterification

product is also strongly affected by the conditions prevailing in the esterification

process.

20

Sorbitan fatty acid esters can be produced by direct, base- or acid-catalysed

reaction of sorbitol with fatty acids at elevated temperature, or by transesterification of

sorbitol with triglycerides or fatty acid methyl esters catalysed by basic compounds [78-

87]. Homogeneous catalysts, either acids (p-toluenesulfonic acid, sulfuric acid,

phosphoric acid) or alkalis (NaOH, KOH, alkaline carbonates) are generally used, in the

absence of solvents. The use of a nitrogen atmosphere or vacuum is required to remove

the water and avoid excessive oxidation of the reagents.

The anhydrization degree of sorbitol increases in the presence of an acid, and

therefore acid catalysts tend to produce mixtures with a high proportion of esters of

sorbides, while alkaline catalysts lead mainly to esters of sorbitans [78]. Unfortunately,

most of the scientific publications and patents only report the saponification and

hydroxyl values of the ester mixtures obtained. A comparison of these values with the

theoretical values calculated for pure esters (as an example, in Figure 5, those of oleic

acid esters as well as the non-esterified polyols are reported) provide the average

anhydrization degree of the sorbitol and the average number of hydroxyl groups

esterified in these mixtures, but the actual distribution of products can not be obtained.

In Table 3 we have collected some representative examples of sorbitol esterification and

transesterification processes for which the composition of the esters mixtures obtained

is reported.

A disadvantage of the base catalysed direct reaction is that the product is usually

highly coloured [78]. Therefore, treatment with a bleaching agent such as hydrogen

peroxide and adsorbents is required to obtain acceptable colour values. Two-step

processes have been established where the acid catalysed anhydrization of sorbitol takes

place before the esterification or transesterification reaction [80,83,88,89]. It has also

been reported that lower colour values are obtained when reducing acids, like

21

phosphorous and hypophosphorous acids, are used as catalysts [83,86], or by addition of

reducing agents like sodium borohydride [83].

The reports on the use of heterogeneous solid catalysts for sorbitol esterification

are scarce. Van Rhijn et al. [85] described the formation of dilauryl isosorbide by

reacting lauric acid and D-sorbitol (acid to polyol molar ratio equal to 6) in the presence

of a siliceous MCM-41 catalyst functionalised with propylsulfonic acid groups. At a

lauric acid conversion of 33%, the selectivity to the isosorbide diester was 95%. At

lower conversion (17%) isosorbide monolaurate is also detected.

Interestingly, the large pore zeolite HBeta (Si/Al = 12.5) is reported to be

inactive under the same reaction conditions [85], due to the preferential adsorption of

the polyol on the zeolite. The better performance of the SO3H-MCM-41 material is

attributed to its higher hydrophobicity. Therefore, it would be expected that this

catalyst, functionalised with hydrophobic alkyl groups, exhibits an improved activity in

the sorbitol esterification. However, the predominance of isosorbide esters reveals that

the acid catalyst favours the deep sorbitol dehydration due to its strong acidity.

Furthermore, the hydrophilic character of the isosorbide monoesters is low (very low

HLB value), owing to their low hydroxyl content, which makes the process unattractive

for the synthesis of emulsifiers (nevertheless isosorbide esters find application in other

fields; for example, the diester with 2-ethylhexanoic acid has been claimed to be

promising as a green PVC plasticizer [90]). A hydrophobic catalyst containing acid

centres of medium strength would be more appropriate for the direct esterification of

sorbitol. Moreover, solid base catalysts would also be useful, as they will probably

decrease the anhydrization of the hexitol.

A more systematic exploration of the behaviour of different zeolites in the

esterification of sorbitol with oleic acid has been reported recently [91]. Zeolite HBeta

22

(Si/Al = 13) turns to be active at 200 ºC (sorbitol/oleic acid = 1, 14.6 wt% of catalyst),

but the ester selectivity is similar to that of p-toluenesulfonic acid. Decreasing the

crystal size by one order of magnitude (to 12 nm) do not change the activity and

selectivity, but both decrease when the Si/Al ratio is increased to 50. Zeolite Beta free

of defects (that is, free of silanol groups) synthesised in fluoride medium is more active,

which has been attributed to its higher hydrophobicity. Zeolite ITQ-2 (delaminated

MCM-22) is also active in the esterification reaction.

In all these examples the OH content of the esters was determined according to

standard procedures, and a value of 100 mg KOH·g-1 was obtained. Such a low OH

content of the product is consistent with the formation of isosorbide esters and/or

sorbitan esters having a high degree of esterification (Figure 5). Therefore, it can be

concluded that a low yield of monoesters is obtained with these zeolitic catalysts, for

sorbitol conversion levels in the range of 48 to 99%.

Cs-containing heteropolyacids (Hx Cs3-x PW12 O40 , x = 0.5 and 1) have also been

used as esterification catalysts [91]. At 150 ºC and a sorbitol/oleic acid molar ratio of 1,

with 13.6 wt% of catalyst, quantitative conversion of both reagents is obtained.

However, the OH value of the product is about 110, which suggests the predominance

of isosorbide monoesters in the mixture.

To summarise these results, it is concluded that no sorbitol or sorbitan

monoesters are obtained by using these acid catalysts in the direct esterification of

sorbitol with oleic acid. In order to improve the selectivity to the monoacyl derivatives,

part of the hydroxyl groups of the sorbitol were protected by forming the corresponding

ketals by reaction with acetone in the presence of HCl as catalyst [91]. This

modification increases the solubility of sorbitol in oleic acid, which allows the

esterification reaction to take place at lower temperature (135 ºC). Under these

23

conditions, the oleic acid conversion level obtained with the zeolite mordenite (Si/Al =

10) was higher than 90% at 48 h, whereas the OH value of the product was as high as

360. OH values of the esters mixtures obtained with zeolites Beta and ITQ-2 and

heteropolyacids also increased by means of the ketalization of sorbitol. Therefore, this

procedure probably reduces the transformation of sorbitol into isosorbide, and also

decreases the average esterification degree of the product.

5. Concluding remarks

The current technology for the production of fatty acid esters of glycerol,

polyglycerol and sorbitol suffers from severe drawbacks due to the use of homogeneous

catalysts. The synthesis of these emulsifiers generates large amounts of waste residues

because the mineral acids or alkalis that act as catalysts must be neutralised and

bleaching agents and adsorbents are usually needed to obtain products with low colour

values. Due to the low selectivity of the homogeneous catalysts, a large excess of polyol

is needed in order to increase the yield of the valuable esters, those with a low

esterification degree, therefore requiring separation of the reactant in excess and

increasing the energy demand. Furthermore, costly molecular distillation of the

glyceride mixtures obtained must be carried out to obtain the required concentration of

monoglycerides. In the case of sorbitol and polyglycerol, complex mixtures of esters

with a wide hydrophilic-lipophilic balance distribution are being currently produced.

However, ester mixtures with a more defined HLB would exhibit improved emulsifying

properties.

The potential of solid catalysts to overcome most of these drawbacks has hardly

been explored. A relatively important research effort has been devoted to the study of

solid catalysts for the synthesis of monoglycerides. Significant improvements have been

24

reported by using zeolitic catalysts and, especially, mesostructured solids, like siliceous

MCM-41 structures, with acid functionalities. As the highest yields of monoglycerides

that have been obtained with these catalysts are still below those required for

commercial use of these esters mixtures, there are still several issues that need further

research. The possibility to tune the pore size and generate different pore architectures

in these mesostructures, can allow the tailoring of the catalyst according to the fatty acid

used in order to optimise the monoester yield. Different functionalities can be

incorporated on the surface of these mesostructures. Beneficial effects of the control of

the catalyst hydrophilic properties by attachment of alkyl chains on its surface have

been reported that deserves further study. Likewise, these mesostructures are interesting

catalysts for the synthesis of esters of sorbitol and polyglycerol. Acid-functionalised

mesostructures are good candidates to be tested as catalysts for the selective synthesis of

polyglycerol monoesters at low fatty acid to polyglycerol ratio. On the other hand, base-

functionalised mesostructures have been reported to catalyse the etherification reaction

of glycerol to produce polyglycerol with a narrow chain length distribution. These

materials could be excellent catalysts for the esterification or transesterification of

sorbitol to obtain esters of high HLB, as they would produce esters with lower sorbitol

anhydrization degree than those obtained with acid catalysts.

ACKNOWLEDGEMENTS

The financial support of the Spanish CICyT (projects MAT2000-1167-C02-02 and

PETRI95-0537-OP), CAM (project 07M/48/2002) and Betaquímica, S.A. is gratefully

acknowledged.

30

TABLES

Table 1. Esterification of glycerol with fatty acids catalysed by zeolites.

Catalyst Si/Al

ratio

Fatty

acid

Glycerol to

fatty acid

molar ratio

Reaction

temperature

(ºC)

Fatty acid

conversion a

(%)

Monoglyceride

selectivity a

(%)

Ref.

FAU 10.0 Oleic 6 90 5 67 13

FAU 37.5 Oleic 1 100 10.5 53 14

FAU 7.3 Oleic 1 100 6.9 53 14

FAU 7.7 Oleic 1 180 80 b 76 b 15

FAU 14.8 Lauric 1 112 45 69 c 16

BEA 14.0 Lauric 1 112 38 64 c 16

MFI 16.5 Lauric 1 112 33 57 c 16

FAU 35.0 Lauric 1 100 13 68 17

MOR 62.0 Lauric 1 100 16 55 17

BEA 42.0 Lauric 1 100 32 d 68 d 18

a Reaction time = 24 h. b Reaction time = 5 h. c At 50% conversion. d Reaction time = 10 h.

31

Table 2. Synthesis of fatty acid esters of polyglycerol.

Polyglycerol esterification and transesterification

Fatty compd. Fatty c./

PGL

ratio

Catalyst [wt%] Reaction

temp.

(ºC)

Fatty acid

conversion

(%)

Glycerol

units

Ester

composition

Ref.

Capric acid 1 dodecylbenzene

sulfonic acid+

H3PO2 [0.2]

145-165 100 1-4

(88%)

50% mono-,

40% diester

66

Lauric acid 1.1 dodecylbenzene

sulfonic acid+

H3PO2 [0.2]

145-165 100 1-4

(88%)

35% mono-,

55% diester

66

Erucic acid 6 LiCl 230 6 75% mean

esterifn.

65

Methyl

stearate

0.5 NaOH [0.15] 130 100 4 (mean) 98% monoester 69

Methyl

stearate

1 NaOH [0.15] 130 89 4 (mean) 56% monoester 69

Addition polymerisation of glycidol to fatty compounds

Fatty compd. Fatty c./

glycidol

ratio

Catalyst [wt%] Reaction

temp.

(ºC)

Fatty acid

conversion

(%)

Glycerol

units

Ester

composition

Ref.

Lauric

acid

1/6 H3PO4

[0.02]

140 6 (mean) 91% monoester 73

Lauric

acid

1/10 H3PO4

[0.02]

140 10 (mean) 77% monoester 73

32

Table 3. Synthesis of fatty acid esters of sorbitol and sorbitol anhydrides.

Fatty compd. Fatty c./

sorbitol

ratio

Catalyst Reaction

temp.

(ºC)

Fatty acid

conversion

(%)

Esters mixture

composition

Ref.

Coconut oil

fatty acids

1 None 235 ~100 70% sorbitan monoester 78

Coconut oil

fatty acids

1 H3PO4

(pH=2.0)

225 ~100 Mostly sorbide monoester 78

Oleic acid 1.45 KOH 200-220 Mostly sorbitan diester a 83

Oleic acid 1.47 H3PO3 + NaOH

(0.6:1 molar

ratio)

2.6 wt%

245 ~100 Mostly sorbitan diester 86

Lauric acid 6 Sulfonic acid

functionalised

MCM-41

110 33 95% isosorbide diester 85

a Sorbitol dehydration is carried out in a previous step using hypophosphorous acid as catalyst.

33

SCHEMES

Scheme 1

OOCOOCOOC

RRR

OHOHOH

OOCOHOH

R

+ 32

34

Scheme 2

OOCOHOH

RO

OH

R COOH+

35

Scheme 3

OHOHOH

OH

O

OHOH

OHOH

O

n

OH2OH

O

OHOH

OH

O

OOC R

n

R-COOH

OH2

(n+2)- (n+1)

-

36

Scheme 4

Direct esterification

Transesterification

Polycondensation

Addition polymerisation of glycidol

OH OHO

OHO

OH OH

n

RCOOH

OH OHO

OHO

OH OOC-R

n

OH2+ +

OH OHO

OHO

OH OH

n

RCOOMe

OH OHO

OHO

OH OOC-R

n

MeOH

OOCOOCOOC

RRR

OH OHO

OHO

OH OH

n

OH OHO

OHO

OH OOC-R

n

OHOOCOOC

RR

+ +

+ +

RCOOR'OH OH

OOH

OOH OOC-R

n

OH2

O OH O OOC-RR'OHO OH

R' CH2 CH CH2

OH OH

+- (n+1)

= H ,

OHOHOH

OH OHO

OHO

OH OOC-R

n

OHOHOOC R

OH2+ +(n+1) (n+1)

37

Scheme 5

OH2

OHOH

OH OH

OH OH

O

O OH

OH

O

OH OH

OHOH

O

OH OH

OHOH

OH2

&

-2

-

1,4-sorbitan 2,5-sorbitan

isosorbide

sorbitol

38

FIGURE CAPTIONS

Figure 1. Scientific and technological publications on the synthesis, processing and use

of fatty acid monoglycerides (a), polyglycerol esters (b) and sorbitol monoesters (c).

Insets: distribution of patent applicants nationalities. Data collected using the Chemical

Abstracts Service research tool SciFinder®. *Data on 2003 correspond to the period

from January to April.

Figure 2. HLB values of triglycerol fatty acid esters.

Figure 3. HLB values of polyglycerol stearic acid esters.

Figure 4. Saponification and OH numbers of oleic acid esters of polyglycerols. Figures

indicate the esterification degree (number of fatty acyl groups per molecule) of each

ester.

Figure 5. Saponification and OH numbers of oleic acid esters of sorbitol and its

anhydrides. Figures indicate the esterification degree (number of fatty acyl groups per

molecule) of each ester.

39

0

20

40

60

80

100

120

140

160

180

200

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

Publ

icat

ions

TotalPatents

* 20

03

UE20%

USA17%

Others9%

Japan54%

a

0

20

40

60

80

100

120

140

160

180

200

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

Publ

icat

ions

TotalPatents

* 20

03

UE14%

Japan76%

Others1%

USA9%

0

5

10

15

20

25

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

Year

Publ

icat

ions

TotalPatents

* 20

03

UE15%

Japan61%

Others5%USA

19%

b

c

Figure 1

40

Figure 2

2 4 6 8 10 12 14 16 18 200

2

4

6

8

10

12

14

16

H

LB

Fatty acid chain length (No. of C atoms)

Esterificationdegree

1

2

345

41

Figure 3

0 2 4 6 8 100

2

4

6

8

10

12

14

16

H

LB

Esterification degree

Polymerizationdegree

10

7

3

4

5

42

Figure 4

0 100 200 300 400 500 60050

100

150

200

1

23

1

2

34

1

2

34

5

1

2

3

45

1

2

3

45

67

89

decaglycerol

glycerol

diglycerol

triglycerol

esterification degree pentaglycerol

Sapo

nific

atio

n N

o. (m

g KO

H·g

-1)

OH No. (mg KOH·g-1)

43

Figure 5

0 500 1000 1500 2000

0

50

100

150

200

0

1

2

0

1

23

456

0

1

23

4

esterification degree

sorbitolsorbitans

Sap

onifi

catio

n N

o. (m

g K

OH

·g-1)

OH No. (mg KOH·g-1)

sorbides

25

REFERENCES

[1] W.C. Griffin, J. Soc. Cosmetic Chem. 1 (1949) 311

[2] G. Dieckelmann, H.J. Heinz, The Basics of Industrial Oleochemistry (Petr Pomp,

Essen, 1988)

[3] U.T. Bornscheuer, Enzyme Microb. Technol., 17 (1995) 578

[4] F.J. Plou, M. Barandiaran, M.V. Calvo, A. Ballesteros, E. Pastor, Enzyme Microb.

Technol., 18 (1996) 66

[5] C. Otero, J.A. Arcos, M.A. Berrendero, C. Torres, J. Mol. Catal. B, 11 (2001) 883

[6] W. Tsuzuki, Y. Kitamura, T. Suzuki, S. Kobayashi, Biotechnol. Bioeng., 64 (1999)

267

[7] D. Charlemagne, M.D. Legoy, J. Am. Oil Chem. Soc., 72 (1995) 61

[8] A. Brock, B. Grüning, G. Hills, European Patent Appl. EP 1250917 (2003)

[9] N.O.V. Sonntag, J. Am. Oil Chem. Soc., 59 (1982) 795A

[10] A. Corma, S. Iborra, S. Miquel, J. Primo, J. Catal. 173 (1998) 315

[11] S. Bancquart, C. Vanhove, Y. Pouilloux, J. Barrault, Appl. Catal. A, 218 (2001) 1

[12] A. Corma, H. García, S. Iborra, J. Primo, J. Catal. 120 (1989) 78

[13] S. Abro, Y. Pouilloux, J. Barrault, Stud. Surf. Sci. Catal. 108 (1997) 539

[14] I. Díaz, A. Martínez-Guereñu, F. Mohíno, J. Pérez-Pariente, E. Sastre, Actas do

XVII Simpósio Ibero-americano de Catálise (FEUP, Porto, 2000) p. 559

[15] N. Sánchez, M. Martínez, J. Aracil, Ind. Eng. Chem. Res., 36 (1997) 1524

[16] E. Heykants, W.H. Verrelst, R.F. Parton, P.A. Jacobs, Stud. Surf. Sci. Catal., 105

(1997) 1277

[17] M. da Silva-Machado, J. Pérez-Pariente, E. Sastre, D. Cardoso, A. Martínez-

Guereñu, Appl. Catal. A, 203 (2000) 321

26

[18] M. da Silva-Machado, D. Cardoso, J. Pérez-Pariente, Stud. Surf. Sci. Catal. 130

(2000) 3417

[19] Y. Pouilloux, S. Abro, C. Vanhove, J. Barrault, J. Mol. Catal. A, 149 (1999) 243

[20] A. Corma, V. Fornés, M.T. Navarro, J. Pérez-Pariente, J. Catal. 148 (1994) 569

[21] J. Pérez-Pariente, I. Díaz, F. Mohíno, E. Sastre, Appl. Catal. A, in press

[22] W.D. Bossaert, D.E. de Vos, W.M. Van Rhijn, J. Bullen, P.J. Grobet, P.A. Jacobs,

J. Catal. 182 (1999) 156

[23] I. Díaz, F. Mohíno, J. Pérez-Pariente, E. Sastre, Appl. Catal. A, 205 (2001) 19

[24] I. Díaz, C. Márquez-Alvarez, F. Mohíno, J. Pérez-Pariente, E. Sastre, Microporous

Mesoporous Mater. 44-45 (2001) 295

[25] I. Díaz, C. Márquez-Alvarez, F. Mohíno, J. Pérez-Pariente, E. Sastre, J. Catal. 193

(2000) 283

[26] I. Díaz, C. Márquez-Alvarez, F. Mohíno, J. Pérez-Pariente, E. Sastre, J. Catal. 193

(2000) 295

[27] F. Mohíno, I. Díaz, J. Pérez-Pariente, E. Sastre, Stud. Surf. Sci. Catal. 142 (2002)

1275

[28] I. Díaz, F. Mohíno, J. Pérez-Pariente, E. Sastre, Appl. Catal. A, 242 (2003) 161

[29] I. Díaz, F. Mohíno, M.T. Blasco, E. Sastre, J. Pérez-Pariente, J. Catal., submitted

[30] K. Wilson, A.F. Lee, D.J. Macquarrie, J.H. Clark, Appl. Catal. A, 228 (2002) 127

[31] M. Boveri, J. Agúndez, I. Díaz, J. Pérez-Pariente, E. Sastre, Actas de la Reunión de

la SECAT (Málaga-Torremolinos, 2003)

[32] I. Díaz, F. Mohíno, E. Sastre, J. Pérez-Pariente, Stud. Surf. Sci. Catal. 135 (2001)

1383

[33] Y. Sakamoto, I. Díaz, O. Terasaki, D. Zhao, J. Pérez-Pariente, J.M. Kim, G.D.

Stucky, J. Phys. Chem. B, 106 (2002) 3118

27

[34] I. Díaz, F. Mohíno, J. Pérez-Pariente, E. Sastre, P.A. Wright, W. Zhou, Stud. Surf.

Sci. Catal. 135 (2001) 1248

[35] A. Cauvel, G. Renard, D. Brunel, J. Org. Chem., 62 (1997) 749

[36] X. Lin, G.K. Chuah, S. Jaenicke, J. Mol. Catal. A, 150 (1999) 287

[37] N. Garti, A. Aserin, B. Zaidman, J. Am. Oil Chem. Soc. 58 (1981) 878

[38] E. Symmes, US Patent Appl. US 1696337 (1928)

[39] H.A. Bruson, US Patent Appl. US 2284127 (1942)

[40] K. Babayan, US Patent Appl. US3637774 (1972)

[41] R.T. McIntyre, J. Am. Oil Chem. Soc., 56 (1979) 835A

[42] J. Van Heteren, P. Cornelis, F. Reckweg, M.F. Stewart, European Patent Appl. EP

0070080 (1983)

[43] J. Klein, P. Daute, U. Hees, B. Beuer, PCT Int. Appl. WO 9302124 (1993)

[44] G. Hillion, R. Stern, European Patent Appl. EP 0518765 (1992)

[45] European Parliament and Council Directive No. 95/2/EC

[46] Code of Federal Regulation, Vol. 3, Title 21, Food and Drug Administration, US

Government (2002)

[47] H. Helfert, A. Hettche,S. Weiss, W.D. Balzer, W. Dietsche, G. Hatzmann, German

Patent Appl. DE 3118417 (1982)

[48] K.S. Dobson, K.D. Williams, C.J. Boriack, J. Am. Oil Chem. Soc., 70 (1993) 1089

[49] S. Akimoto, Japanese Patent Appl. JP58198429 (1983)

[50] S. Uda, E. Takemoto, Japanese Patent Appl. JP61140534 (1986)

[51] S. Uda, E. Takemoto, Japanese Patent Appl. JP61043627 (1986)

[52] U. Hees, R. Bunte, J.W. Hachgenei, P. Kuhm, E.G. Harris, PCT Int. Appl.

WO9325511 (1993)

28

[53] K. Cottin, J.M. Clacens, Y. Pouilloux, J. Barrault, Ol., Corps Gras, Lipides, 5

(1998) 407

[54] J.M. Clacens, Y. Pouilloux, J. Barrault, C. Linares, M. Goldwasser, Stud. Surf. Sci.

Catal., 118 (1998) 895

[55] J.M. Clacens, Y. Pouilloux, J. Barrault, Appl. Catal. A, 227 (2002) 181

[56] T. Nakamura, M. Yamashita, US Patent Appl. US 2002/0035238 (2002)

[57] N. Garti, G.F. Remon, B. Zaidman, Tenside Deterg. 23 (1986) 6

[58] S.B. Harvey, US Patent Appl. US 5585506 (1996)

[59] D. Lemke, PCT Int. Appl. WO 0236534 (2002)

[60] UK Patent Appl. GB1025265 (1966)

[61] M. Ishitobi, US Patent Appl. US2002/123439 (2002)

[62] H. Tomokazu, Y. Akifumi, Japanese Patent Appl. JP 8217725 (1996)

[63] P. Seiden, R.A. Woo, Patent Appl. CA 1185610 (1985)

[64] B.K. Tjurin, German Patent Appl. DE 2511807 (1976)

[65] K. Matsuda, M. Kitao, US Patent Appl. US 5773073 (1998)

[66] G. Jakobson, W. Siemanowski, K.H. Uhlig, US Patent Appl. US 5424469 (1995)

[67] P. Denecke, G. Börner, V. Von Allmen, UK Patent Appl. GB 2073232 (1981)

[68] G. Jakobson, W. Siemanowski, K.H. Uhlig, US Patent Appl. US 5466719 (1995)

[69] V.R. Kaufman, N. Garti, J. Am. Oil Chem. Soc. 59 (1982) 471

[70] I. Beseda, P.E. de Detrich, US Patent Appl. US 4950441 (1990)

[71] V. Eychenne, L. Debrauwer, Z. Mouloungui, J. Surfact. Deterg. 3 (2000) 173

[72] E. Ulsperger, L. Richter, F. Mainas, Patent Appl. DD 41661 (1965)

[73] T. Endo, T. Daito, European Patent Appl. EP0758641 (1997)

[74] T. Endo, K. Maruo, Japanese Patent Appl. JP 2000143794 (2000)

[75] T. Endo, T. Ueno, Japanese Patent Appl. JP 2000143584 (2000)

29

[76] K. Maruo, H. Omori, Japanese Patent Appl. JP 11222521 (1999)

[77] G. Vanlerberghe, H. Sebag, US Patent Appl. US 4515775 (1985)

[78] K.R. Brown, US Patent Appl. US 2322820 (1943)

[79] L.D. D'Hinterland, G. Mouzin, H. Cousse, French Patent Appl. FR 2176487 (1973)

[80] G.J. Stockburger, US Patent Appl. US 4297290 (1981)

[81] T. Masuda, M. Honjiyou, K. Watanabe, Japanese Patent Appl. JP 57133197 (1982)

[82] S. Ropuszynski, E. Sczesna, Tenside Deterg. 22 (1985) 190

[83] N. Milstein, PCT Int. Appl. WO 9200947 (1992)

[84] G.O. Gruetzmacher, J.W. Raggon, B. Wlodecki, US Patent Appl. US 5612080

(1997)

[85] W.M. Van Rhijn, D.E. De Vos, B.F. Sels, W.D. Bossaert, P.A. Jacobs, Chem.

Commun. (1998) 317

[86] J.M.H. Ellis, J.J. Lewis, R.J. Beattie, PCT Int. Appl. WO 9804540 (1998)

[87] PCT Int. Appl. WO 0128961 (2001)

[88] L. Szotyori, A. Ujhidy, I. Szabo, G. Bozoki, J. Bathory, Hungarian Patent Appl.

HU 5159 (1972)

[89] R.V. Vladea, M.A. Vineze, T.L. Simandan, L.M. Rusnac, A.T. Fenesiu, M.

Mosincat, Romanian Patent Appl. RO 87381 (1985)

[90] H. Luitjes, J.C. Jansen, PCT Int. Appl. WO 9945060 (1999)

[91] A. Velty, Ph. D. Thesis, Universidad Politécnica de Valencia (2003)