Mesoporous materials as catalysts for the production of chemicals: Synthesis of alkyl glucosides on MCM-41

Article abstract of DOI:10.1006/jcat.1998.2382

The synthesis of alkylglucosides from glucose and n-butanol has been carried out successfully on AL-MCM-41 mesoporous materials. The influence of the chemical composition (Si/Al) and pore dimensionson activity and selectivity has been studied. It has been found that a higher concentration of acid sitesdoes not guarantee a better catalytic performance, and the adsorption-desorption properties of the material play a determinant role in this reaction where the two reactants and the product have very different polarities. On the other hand, in the range of pore sizes studied here, the larger the diameter of the pore at the same level of Al contents, the more activeis the final catalyst. The catalyst loses activity during the process due to the presence of strongly adsorbed molecules. Soxhlet extraction by methanol followed by water does not recover all the initial activity but produces a loss of crystal unity. However, the catalyst could be fully regenerated by calcination in air at 773 K.

Full text of DOI:10.1006/jcat.1998.2382

Journal of Catalysis 183, 76–82 (1999)
Article ID jcat.1998.2382, available online at http://www.idealibrary.com on
Mesoporous Materials as Catalysts for the Production of Chemicals:
Synthesis of Alkyl Glucosides on MCM-41
M. J. Climent, A. Corma,1 S. Iborra, S. Miquel, J. Primo, and F. Rey
Instituto de Tecnologia Qu´ımica (U.P.V.-C.S.I.C.), Universidad Polite´cnica de Valencia, Avenida de Los Naranjos s/n 46022 Valencia, Spain
E-mail: acorma@itq.upv.es
Received July 21, 1998; revised December 9, 1998; accepted December 9, 1998
chemicals and fine chemical products, which often involve
bulky reactants and products. Furthermore, in many cases
the synthesis of the chemicals is carried out in liquid phase
where diffusional problems can be enhanced. While meso-
porous materials have been extensively studied as catalysts
for oil refining and petrochemistry, their use in the syn-
thesis of chemicals and fine chemicals is still in its infancy
(5–13).
Surfactants are present in many products which are com-
monly and largely used in our society. Long-chain alkyl glu-
cosides are non-ionic compounds with excellent surfactant
properties, biodegradability, and lowdegree ofskin and oral
toxicity (14, 15). They have several applications as cosmetic
surfactants, food emulsifiers, and pharmaceutical dispersing
agents (16, 17). Recently we have presented that is possible
to prepare butyl glucosides in very good yields by means
of the Fischer glycosidation reaction, using large pore acid
zeolites as catalysts (18, 19). These lower alkyl glucosides,
being more soluble than the glucose in a fatty alcohol, are
important intermediates for the manufacturing of higher
glucosides (20).
The synthesis of alkylglucosides from glucose and n-butanol has
been carried out successfullyon Al–MCM-41mesoporousmaterials.
The influence of the chemical composition (Si/Al) and pore dimen-
sions on activity and selectivity has been studied. It has been found
that a higher concentration of acid sites does not guarantee a better
catalytic performance, and the adsorption–desorption properties of
the material play a determinant role in this reaction where the two
reactants and the product have very different polarities. On the
other hand, in the range of pore sizes studied here, the larger the di-
ameter of the pore at the same level of Al contents, the more active is
the final catalyst. The catalyst loses activity during the process due
to the presence of strongly adsorbed molecules. Soxhlet extraction
by methanol followed by water does not recover all the initial activ-
ity but produces a loss of crystallinity. However, the catalyst could be
c
fully regenerated by calcination in air at 773 K.
1999 Academic Press
INTRODUCTION
The synthesis of new mesoporous aluminosilicates mole-
cular sieves of the M41-S family (1) with regular pore di-
mensions which can be varied between 2 and 10 nm has
opened new possibilities in the field of solid acid catalysis
(2). The system with a hexagonal array of pores known as
MCM-41 is the most important member of the family and
can be prepared by a liquid crystal templating mechanism
where surfactant molecules act as templates (3). These ma-
terials, with mild acidities, were conceived initially to carry
out catalytic cracking processes of bulky molecules. Unfor-
tunately the hydrothermal stability of the material should
be improved in order to prepare successful fcc catalysts (4).
The advantages of mesoporous materials for acid catalized
In this work we have investigated the possibilities of
mesoporous MCM-41 aluminosilicates in the preparation
ofalkylglucoside surfactants(21). Thus, we have carried out
the acetalization of -D-glucose with butanolin the presence
of MCM-41, and we present the influence of the physico-
chemical characteristics of the catalyst on the activity and
selectivity to butyl glucosides.
EXPERIMENTAL
reactions, are based on the presence of large regular pores Catalysts
which allow the diffusion of reactants and conversely the
The synthesis of Al containing MCM-41 catalysts with
fast diffusion of the products out, minimizing unwanted
consecutive reactions and catalyst decay. If these charac-
teristics are important in the field of oil refining and petro-
chemistry, they are still more important in the synthesis of
different pore diameters and varied aluminium contents
was performed by controlling the chain length of the sur-
factant that templates the MCM-41 formation as well as the
synthesis conditions.
Al–MCM-41 materials were prepared from the follow-
ing general synthesis gel formulation: SiO2, 0.28; TMAOH,
1 To whom correspondence should be addressed.
76
0021-9517/99 $30.00
c
Copyright
1999 by Academic Press
All rights of reproduction in any form reserved.

SYNTHESIS OF ALKYL GLUCOSIDES ON MCM-41
77
TABLE 1
Synthesis Conditions, Chemical Composition, Pore Diameter, and Surface Area of the Catalysts
Crystallization
Time (days)
Porea
diameter (nm)
Catalysts
n
x (102)
T (K)
Si/Al
S (m2/g)b
MCM41-1
MCM41-2
MCM41-3
MCM41-4
MCM41-5
16
16
16
16
16
3.33
2.0
1.0
1.0
1.0
423
423
423
408
408
5
5
5
1
3
14
26
50
51
50
4.9
5.4
5.3
4.5
2.5
720
624
753
835
880
a Measured by Ar.
b Measured by N2.
0.12; CnTABr, 26.2; H2O, xAl2O3, where x was varied in After saturation, the samples were outgased at 423, 523, and
order to obtain Si/Al ratios between 15 and 50, and CnTABr 673 K under vacuum, cooled to room temperature, and then
is an alkyltrimethylammonium bromide in which the linear the spectra were recorded.
alkyl chain contains 10 or 16 carbons atoms. The synthe-
Anhydrous -D-glucose and n-butanol, obtained from
sis gel was prepared by adding a CnTABr aqueous solution Aldrich and Quimon, respectively, with a nominal purity
to an aqueous slurry containing 10 wt% of silica (Aerosil <99% , were used without further purification.
200 from Degussa) and the appropriate weight of TMAOH
Reaction Procedures
(25 wt% solution in water, Aldrich). This slurry was further
homogenized by stirring the mixture for 15 min. Then, the
Acetalization reaction of D-glucose with n-butanol. The
appropriate amount of aluminium source (Alumina Vista catalyst was activated in situ, in a 100-ml batch glass reac-
from Catapal) was added followed by the silica necessary tor, by heating 1.5 g of the solid at 353 K under vacuum
to obtain the gel composition indicated in Table 1. The gel (1 Torr) for 3 h. After this time, the system was cooled to
was homogenized for 1 h at room temperature by stirring room temperature and then n-butanol (50 ml, 0.54 mol)
at 250 rpm and autoclaved at 408 K during different crys- was introduced followed by the addition of glucose (2.50 g,
tallization times (see Table 1) in Teflon-lined stainless steel 0.014 mol). The reaction mixture was heated at 393 K for
autoclaves. After crystallization, the Al–MCM-41 material 4 h in a system equipped with a silicone oil bath, a mag-
was recovered by filtration, washed with distilled water, and netic stirrer, and a condenser. The catalyst was uniformly
dried at 333 K overnight. The final active acid Al–MCM-41 suspended as a slurry by stirring at 600 rpm.
catalyst was obtained by removing the occluded surfactant,
The solubility of the glucose in the n-butanol increased
which fills the pores, by heating the sample under nitrogen once the products were formed, and when 15% of conver-
at 810 K (temperature ramp = 3 C min 1) for 1 h. Then, the sion was achieved the glucose was already completely dis-
remaining organic was burned off in air at 810 K for 6 h. The solved.
synthesis conditions for each particular MCM-41 catalyst,
chemical composition, pore diameter, and surface area are vals. The filtered zeolite remaining in the microsyringe was
reported in Table 1. extracted with methanol and water as the first and second
Samples were taken and analyzed at regular time inter-
A beta zeolite with crystal size of 0.35 m and Si/Al solvents, respectively. The reaction products were analyzed
ratio of 16 was synthesized following the method presented by HPLC, adding a known amount of methyl -D-gluco-
in Ref. (22). Scanning electron microscopy (ISI-SS60) was pyranoside as internal standard.
used to determine crystal sizes and morphology.
At the end of the reaction, the mixture was cooled to
The solids were characterized by X-ray diffraction on room temperature, dissolved in methanol, and filtered. The
a Philips PW diffractometer using CuK radiation. Surface filtered catalyst was submitted to continuous solid–liquid
area measurements were obtained on an ASAP-2000 appa- extractions in micro-soxhlet equipment using methanol and
ratus following the BET procedure. Pore diameter distribu- water as the first and second solvents, respectively. The or-
tion was obtained using argon as adsorbate and following ganic solutions were freed from the solvents and the re-
the Horvath–Kawazoe method (23).
maining butanol by evaporation in vacuum. Then they were
The infrared spectra were recorded at room tempera- combined, weighed, and analyzed by HPLC. Mass balances
ture in a Nicolet 710 FTIR using self-supported wafers of were performed in each experiment and the recovered
10mg cm 2. The calcined sampleswere outgased overnight products accounted for more than 95% of the reactants.
3
at 673 K and 10 Pa dynamic vacuum for 16 h, and then
HPLC analysis were performed at 305 K on a system
pyridine was admitted into the cell at room temperature. composed of a Waters pump (model 510) and a Waters 410

78
CLIMENT ET AL.
differential refractometer using a HYPERSIL-APS-25 mm
(250 0.46 mm) column. In order to deal with the less polar
products the column was eluted with an acetonitrile–water
mobile phase composed of 7% of water, and 20% of water
for the more polar products, the flow of the mobile phase
being 1 ml/min in both cases. Preparative scale HPLC was
performed using a HYPERSIL-APS-25 mm (250 1 mm)
and in this case the flow of the mobile phase was 4 ml/min.
The isolated products of the reaction were analyzed by
SCHEME 1
1
means of H and 13C-NMR spectroscopy using a Varian
400 WB instrument, and they were identified by comparing
with NMR spectra given in the literature (24).
of the five-membered ring compound, followed by the ring
expansion to the pyranoside product (25–29). The presence
of5-hydroxymethylfurfuralcomingfrom the possible dehy-
dration reaction of glucose was not detected in the reaction
mixture.
Transacetalization of diphenylacetaldehyde (5 ) with TOF.
Activation of the catalyst was performed in situ by heating
the solid under vacuum (1 Torr) for 1 h. After this time, the
system was left at room temperature and then a solution
of the carbonyl compound (1.6 mmol) and trimethyl ortho-
formiate (TOF) (8 mmol) in tetrachloromethane (50 ml)
as solvent was poured into the activated catalyst. The re-
sulting suspension was magnetically stirred at reflux tem-
perature for 5 h. Samples were taken at intervals and the
reaction products were analyzed by gas chromatography
(GC). At the end of the reaction the catalyst was filtered
and washed with dichloromethane. The organic solution
was concentrated in vacuum, and the residue was weighed
The catalytic active sites involved in the reaction should
be acid sites associated to bridging hydroxyl groups, which
in the case of the MCM-41 sample should be related to
the presence of tetrahedrally coordinated Al in the amor-
phous walls forming the channels of the MCM-41 structure.
The presence of the Brønsted-acid sites associated to AlIV
has been determined by pyridine adsorption (Fig. 2). The
1
presence of a pyridinium band at 1545 cm indicates the
presence in the MCM41-1 sample of Brønsted-acid sites
able to protonate pyridine which remains adsorbed after
evacuation at 423 K. The presence of the 1455 cm 1 band
corresponds to pyridine coordinated to Lewis sites, proba-
bly related with the presence of AlVI generated during the
calcination of the synthesized sample.
For comparative purposes, we have compared the results
obtained on MCM41-1 with those obtained on a beta zeo-
lite with a Si/Al ratio of 16 and crystal size of 0.35 m. The
results of the initial rates for the disappearance of glucose
calculated from the slope of the curves at zero time and
1
and analyzed by H-RMN and gas chromatograpy-mass
spectrometry (GC-MS) using a Hewlett-Packard 5988 A
spectrometer provided with a 25-m capillary column of
cross-linked 5% phenylmethylsilicone. After reaction, the
catalyst wassubmitted to continuoussolid–liquid extraction
with dichloromethane using micro-soxhlet equipment. Af-
ter removal of the solvent the residue was also weighed and
1
analyzed by GC-MS and H-RMN. In all cases the recov-
ered material accounted for more than 90% of the start-
1
ing material. The H-RMN analysis was carried out with
a 400 MHz Varian VXR-400 spectrometer in deuterated
trichloromethane and TMS as internal standard.
RESULTS AND DISCUSSION
The reaction between the -D-glucose (1) and n-butanol
(2) was carried out in the presence of a MCM-41 sample
with a Si/Al ratio of 14 and a pore diameter of 4.9 nm, at
393 K. After 4 h of reaction the conversion of the glucose
was practically quantitative and the only observed products
were the anomeric mixtures of two butyl glucoside isomers,
i.e., ( , )-butyl glucofuranosides (3) and ( , )-butyl glu-
copyranosides (4) (Scheme 1). In Fig. 1 the evolution of
the reaction products versus time is represented. It can be
seen there that ( , )-butyl glucofuranoside (3) appears as
a primary and unstable product while the behavior of 4 cor-
responds to a secondary and stable product. These results
FIG. 1. Kinetic curves for the reaction of -D-glucose and n-butanol
on MCM-41-1 catalyst, at 393 K. (᭡) -D-glucose; (᭹) ( , )-butyl gluco-
support a reaction mechanism involving the fast formation furanosides; (᭺) ( , )-butyl glucopyranosides.

SYNTHESIS OF ALKYL GLUCOSIDES ON MCM-41
79
FIG. 3. IR spectra of pyridine adsorbed on a beta zeolite (Si/Al = 16)
FIG. 2. IR spectra of pyridine adsorbed on MCM-41-1 at room tem-
perature and desorbed at 423, 523, and 623 K.
at room temperature and desorbed at 423, 523, and 623 K.
the yields of butyl glucosides are summarized in Table 2. It
can be seen that beta zeolite is more active than MCM-41
and this result can be related with the presence of a much
larger amount of bridging hydroxyl groups in the zeolites
(Fig. 3). This observation is in agreement with the results
obtained during the acetalization of aldehydes with differ-
ent kinetic diameters (12), where it has been showed that
when the reactants can diffuse inside the pores of zeolites,
the microporous MCM-41 materials are more active than
the mesoporous MCM-41 materials.
with Si/Al ratios of 14, 26, and 50 which have a similar pore
diameter ( 5.0 nm) (Table 1). In Table 3 the initial rates
for the disappearance of glucose as well as the yields after a
4 h reaction time have been summarized. As we mentioned
above, one would expect the number of acid sites to con-
tinuously decrease in the samples when going from 14 to
50 Si/Al ratios (Table 4), and for this reason the catalytic
activity should decrease in the same way. However it can
be seen in Table 3 that the activity of the catalysts increases
when increasing the Si/Al ratio. It is evident that in order
to explain the catalytic behavior, other factors besides the
number of acid sites have to be taken into account. Thus,
if one considers that the acetalization of glucose involves
a highly hydrophilic reactant (D-glucose) and a much more
hydrophobic n-butanol, we also need to consider the role
played by the hydrophilic–hydrophobic properties of the
catalyst on the relative adsorption of the two reactants. In
this sense, it has been claimed (30, 31) that silica MCM-41
materials are hydrophobic in the sense that they adsorb
a much larger amount of cyclohexane than water. How-
ever, when Al is introduced, the hydrophilicity of the ma-
terial increases. This is even more so when some of the
Al is removed from tetrahedral positions leaving a “nest”
Influence of the Si/Al Ratio of MCM-41
In a recent work, we have presented that large pore ze-
olites were active catalysts for carrying out the acetaliza-
tion of glucose with n-butanol, (19) and it was observed
that the hydrophobic–hydrophilic properties of the solid
play an important role in the kinetics of the process deter-
mining the conversion and product selectivity. It is known
that the framework Si/Al ratio of molecular sieves is not
only responsible for the acid but also for the hydrophobic–
hydrophilic properties of the solids. Then, in order to study
the influence of the Si/Al ratio on the catalytic activity of
MCM-41, the reaction between D-glucose and n-butanol
was carried out in the presence of three catalyst samples
TABLE 3
TABLE 2
Initial Rates for the Disappearance of D-glucose and Yields of
Butyl Glucosides Obtained in the Glucosidation of D-glucose with
n-butanol in the Presence of MCM-41 with Different Si/Al Ratios,
at 393 K, after 4 h Reaction Time
Initial Rates for the Disappearance of D-glucose and Yields of
Butyl Glucosides Obtained in the Glucosidation of D-glucose with
n-butanol in the Presence of Beta Zeolite and MCM-41-1 at 393 K,
after 4 h Reaction Time
Yield (% )
ro 104
(mol min 1 g
Conversion
(% )
Yield (% )
ro 104
Conversion
(% )
Catalyst
)
3
4
1
1
Catalyst
(mol min 1 g
)
3
4
MCM41-1
MCM41-2
MCM41-3
2.7
3.13
3.68
55
61
39
42
33
59
97
94
98
Beta
MCM41-1
5.2
2.7
48
55
52
42
100
97

80
CLIMENT ET AL.
TABLE 4
Brønsted Acidity (B.A.) and Lewis Acidity (L.A.) of the MCM-
41 Samples Measured by IR Spectroscopy Combined with Pyridine
Adsorption and Desorption at Different Temperatures
TABLE 5
Initial Rates for the Disappearance of Glucose and Yields of
Butyl Glucosides Obtained in the Glucosidation of D-glucose with
n-butanol at 393 K after 4 h Reaction Time, in the Presence of
MCM-41 with Si/Al Ratio of 50 and Different Pore Diameters
423 K
523 K
B.A.
623 K
Yield (% )
Catalyst
B.A.a
L.A.a
L.A
B.A.
L.A.
ro 104
(mol min 1 g
Conversion
(% )
1
Catalyst
)
3
4
MCM41-1
MCM41-2
MCM41-3
MCM41-4
MCM41-5
11.8
8.3
7.2
7.1
6.0
37.8
11.0
8.9
10.6
10.6
3.6
2.8
2.4
3.5
2.4
24.6
12.3
5.0
8.5
8.5
0
0
0
0
0
17.4
7.4
3.8
5.1
5.5
MCM41-3
MCM41-4
MCM41-5
3.68
2.30
1.38
39
58
55
59
36
25
98
94
80
a Acidity (molPy/g catalyst) calculated using the extinction coefficients
given in Ref. (32).
in activity when decreasing the pore diameter occurs for
the acetalization of glucose in liquid phase. Indeed, when
the pore diameter decreases by a factor of two the initial
rate becomes 2.5 times smaller.
of silanol groups. If this is so, we could expect that in the
higher Al content MCM-41 the more polar glucose could
be preferentially and strongly adsorbed over the aliphatic
alcohol, influencing negatively the diffusion–adsorption–
desorption rates of reactants and products. This behavior
can be changed by changing the framework Si/Al ratio of
the catalyst. Then, when increasing the Si/Al ratio of the
MCM-41 the hydrophobicity of the catalyst will increase,
and the adsorption of the aliphatic alcohol as well as the
desorption and diffusion of products will occur in a larger
extension improving the catalytic behavior of the meso-
porous material.
Indeed, it can be seen in Table 3 that the MCM-41 sample
with the lowest content of Al, and therefore with the lowest
content of Brønsted-acid sites, give nevertheless the high-
est conversion for the formation of glucosides. We correlate
this increase in activity with the increase in the hydropho-
bicity of the sample, in agreement what we have observed
when using a beta zeolite as catalyst (19).
In order to explain these results we have first of all looked
into the concentration ofacid sitespresent on the three sam-
ples. The pyridine adsorption results (Table 4) show that
the acidity is very similar in all samples and in any case the
small differences observed can not explain the large vari-
ation in activity. A further confirmation of this has been
obtained by carrying out the transacetalization reaction
of diphenylacetaldehyde (5) with trimethyl orthoformiate
(TOF) (Scheme 2) on the three MCM-41with different pore
diameters. The aldehyde selected has no problems diffusing
through the pores and at the same times does not present
large differences in polarity when compared with the other
reactant and with the reaction product 6. The catalytic ac-
tivity results presented in Table 6 shows that, as expected,
the three MCM-41 samples with different pore diameters
show very similar catalytic activity.
Thus, we have to conclude that the differences observed
during the acetalization of D-glucose on MCM-41 samples
with different pore diameter are not due to differences in
the number of active sites, but have to be related with dif-
fusional effects. Since owing to the size of the reactants and
products obtained during the acetalization of glucose we
should not expect differences in the diffusion coefficients
on the three catalysts studied here, the following hypothe-
sis has been made in order to explain the results obtained.
In the case of the MCM-41, even for a sample with a Si/Al
ratio of 50, some glucose is still strongly adsorbed in the
pores of the catalyst. These strongly adsorbed molecules
Influence of the Pore Diameter of the Catalyst
A characteristic of the mesoporous molecular sieves
which is of paramount importance in organic synthesis is
the possibility of modifying the dimensions of the pores in
a wide range. Thus, depending on the molecular size of reac-
tants and products, it is possible to select the most adequate
pore diameter for a specific reaction.
In our case, and with the aim to study the influence of
the mesopore diameter on the glucosidation reaction, the
acetalization of D-glucose with n-butanol was carried out
under the same experimental conditions, in the presence of
three MCM-41 samples with the same Si/Al ratio (50) but
with pore diameters of 2.5, 4.5, and 5.3 nm. If one would be
working in gas phase and taking into account the relative
size of the catalyst pores and the molecular diameter of
reactants and products studied here, little, if any, influence
of the pore diameter on the rate of the reaction could be
expected. However in Table 5 one can see that a decrease
SCHEME 2

SYNTHESIS OF ALKYL GLUCOSIDES ON MCM-41
81
TABLE 6
Initial Rates and Conversion after 4 h Reaction Time for the
Transacetalization Reaction of Diphenylacetaldehyde (5) with
TOF at 351 K in the Presence of MCM-41 with Si/Al Ratio of
50 and Different Pore Diameters
ro 103
(mol min )
Conversion
(% )
1
Catalyst
MCM41-3
MCM41-4
MCM41-5
10.0
9.3
9.5
87
85
88
decreased the mean effective path way in the pores, intro-
ducing, therefore, diffusional limitations. These diffusional
limitations will be smaller when increasing the pore diam-
eter of the catalysts.
FIG. 4. Catalyst decay and regeneration by methanol and water soxh-
let extraction. (᭡) Fresh catalyst. (᭹) After a second use previous regen-
eration by soxhlet extraction. (᭺) After a third use previous a second
regeneration by soxhlet extraction.
This hypothesis is not only reasonable, but it is also sup-
ported when considering the selectivity results obtained at
the same level of conversion. In Table 7 one can see that
the ratio of furanoside to pyranoside (3/4) increases when
decreasing the pore diameter of the catalysts. It appears
then that the formation of butyl glucopyranoside is more
affected than the butyl glucofuranoside by the pore diame-
ter, something which can be related to the lower diffusivity
of the bulkier six-membered ring pyranoside with respect
to the smaller five-membered ring furanoside (19).
loss of crystallinity and surface area of the catalyst, as a con-
sequence of the successive soxhlet extractions with water.
Indeed a strong decrease in the crystallinity of the sample
hasbeen observed after the soxhlet extraction treatments. It
appears then, that a better catalyst regeneration procedure
for this process could be burning of the products remain-
ing adsorbed in the pores after completion of the reaction.
When this was done by calcining the catalyst during 3 h at
773 K in air, the initial activity was restored.
Catalyst Deactivation and Regeneration
In order to study the catalyst deactivation occurring dur-
ing the synthesis of alkyl glucosides we have carried out the
acetalization of D-glucose with n-butanol in the presence
of a MCM41-2 sample with a pore diameter of 5.4 nm and
a Si/Al ratio of 26. After the reaction was completed, the
catalyst sample was subjected to a soxhlet extraction with
methanol and then with water. The resultant sample was
again used in a second and in a third experiment. In Fig. 4
the conversion versus time obtained in each one of the three
catalytic cycles is presented, and a decrease in activity when
using the recovered catalyst is observed. Nevertheless, al-
most complete conversion is achieved in all cases after 4 h
reaction time. This decrease in activity can be related with a
CONCLUSIONS
It has been shown that MCM-41 acid samples are able
to carry out successfully the acetalization of glucose by
aliphatic alcohols to produce biodegradable surfactants.
For this reaction, the hydrophobic properties of the cata-
lysts are as important as the concentration of active acid
sites. In thissense sampleswith higher Si/Alratios, at least in
the range from 15 to 50, give higher conversions. Due to the
strong adsorption of glucose, the mean free path through
the channel could be diminished, and MCM-41 samples
with larger pore diameters give higher conversions. Finally,
the catalyst could be partially regenerated by carrying out
a soxhlet extraction by methanol followed by water of the
solid used in the reaction. However, this process produces
the partial destruction of the MCM-41 and it is not the most
adequate catalyst regeneration procedure. The catalyst can
be conveniently regenerated by calcination in air at 773 K.
TABLE 7
Influence of the Pore Diameter on the Products Distribution
at 70% Glucose Conversion
Yield (% )
Pore
Catalysts
diameter (nm)
3
4
3/4
ACKNOWLEDGMENTS
MCM41-3
MCM41-4
MCM41-5
5.3
4.5
2.5
50
55
60
20
15
10
2.5
3.6
6.0
Financial support by the Spanish MAT97-1016-C02-01 is gratefully ac-
knowledged.

82
CLIMENT ET AL.
REFERENCES
14. de Nijs, P. M., Maat, L., and Kieboom, A. P. G., Recl. Trav. Chim.
Pays-Bas 109, 420 (1990) and references cited therein.
1. Kresge, C. T., Leonowicz, M. E., Roth, W. J., Vartulli, J. C., and Beck, 15. Straathof, A. J. J., Van Bekkum, H., and Kieboom, A. P. G., Starch-
J. S., Nature 359, 710 (1992).
Stra¨rke 40, 229 (1988).
2. Corma, A., Chem. Rev. 97, 2373 (1997).
16. Hurforg, J. R., in “Developments in Food Carbohydrate” (C. K. Lee,
Ed.), p. 327. Appl. Sci., London, 1980.
17. Hughes, F. A., and Lew, B. W., J. Am. Chem. Soc. 47, 162 (1970).
18. Corma, A., Iborra, S., Miquel, S., and Primo, J., J. Catal. 161, 713 (1996).
19. Camblor, M. A., Corma, A., Iborra, S., Miquel, S., Primo, J., and
Valencia, S., J. Catal. 172, 76 (1997).
3. Beck, J. S., Vartulli, J. C., Roth, W. J., Leonowicz, M. E., Kresge, C. T.,
Schmitt, K. D., Chu, C. T.-W., Olson, D. H., Sheppard, E. W., McCullen,
S. B., Higgins, J. B., and Schlenker, J. L., J. Am. Chem. Soc. 114, 10834
(1992).
4. Corma, A., Grande, M. S., Gonzalez-Alfaro, V., and Orchilles, A. V.,
J. Catal. 159, 375 (1996).
20. Boettner, F. E., U.S. Pat. 3,219,656 (1965).
5. Armengol, E., Cano, M. L., Corma, A., Garcia, H., and Navarro, M. T.,
J. Chem. Soc. Chem. Commun. 519 (1995).
6. Armengol, E., Corma, A., Garcia, H., and Primo, J., Appl. Catal. A
126, 391 (1995).
7. Armengol, E., Corma, A., Garcia, H., and Primo, J., Appl. Catal. A
149, 411 (1997).
21. De Goede, A. T. J. W., Van Der Leij, I. G., Van Der Heijden, A. M.,
Van Rantwijk, F., and Van Bekkum, H., Pat. WO 96/36640 (1996).
22. Camblor, M. A., and Pe´rez-Pariente, J., Zeolites 11, 202 (1991).
23. Horvath, G., and Kawazoe, K., J. Chem. Eng. Jpn. 16, 470 (1983).
24. Straathof, A. J. J., Romein, J., Van Rantwijk, F., Kieboom, A. P. G.,
and Van Bekkum, H., Starch /Sta¨rke 39, 362 (1987).
8. Le, Q. N., U.S. Pat. 5,191,134 (1993).
25. Mowery, P. F., J. Org. Chem. 26, 3484 (1961).
9. Gunnewegh, E. A., Gopie, S. S., and Van Bekkum, H., J. Mol. Catal.
A 106, 151 (1996).
10. Van Bekkum, H., Hoefnagel, A. J., Van Koten, M. A., Gunnewegh,
E. A., Vogt, A. H., and Kouwenhoven, H. W., Stud. Surf. Sci. Catal.
83, 379 (1994).
11. Kloestra, K. R., and Van Bekkum, H., J. Chem. Res.(S) 1, 26 (1995).
12. Climent, M. J., Corma, A., Iborra, S., Navarro, M. C., and Primo, J.,
J. Catal. 161, 786 (1996).
13. Climent, M. J., Corma, A., Guil-Lopez, R., Iborra, S., and Primo, J.,
J. Catal. 175, 70 (1998).
26. Capon, B., Chem. Rev. 68, 407 (1969).
27. Bennet, A. J., Sinnot, M. L., and Sulochana Wijesundera, W. S.,
J. Chem. Soc., Perkin Trans. 2, 1233 (1985).
28. Yamaguchi, T., and Kojima, M., Carbohydr. Res. 59, 129 (1977).
29. Post, C. B., and Karplus, M., J. Am. Chem. Soc. 108, 1317 (1986).
30. Chen, C. Y., Li, H. X., and Davis, M. E., Microporous Mater. 2, 17
(1993).
31. Glaser, R., Roesky, R., Boger, T., Eigenberger, G., Ernst, S., and
Weitkcamp, J., Stud. Surf. Sci. Catal. 105, 695 (1997).
32. Emeis, C. A., J. Catal. 141, 347 (1993).