Zeolite-catalysed conversion of C3 sugars to alkyl lactates

Article abstract of DOI:10.1039/b921284a

The direct conversion of C₃ sugars (or trioses) to alkyl lactates was achieved using zeolite catalysts. This reaction represents a key step towards the efficient conversion of bio-glycerol or formaldehyde to added-value chemicals such as lactate derivatives. The highest yields and selectivities towards the desired lactate product were obtained with Ultrastable zeolite Y materials having a low Si/Al ratio and a high content of extra-framework aluminium. Correlating the types and amounts of acid sites present in the different zeolites reveals that two acid functions are required to achieve excellent catalysis. Bronsted acid sites catalyse the conversion of trioses to the reaction intermediate pyruvic aldehyde, while Lewis acid sites further assist in the intramolecular rearrangement of the aldehyde into the desired lactate ester product. The presence of strong zeolitic Bronsted acid sites should be avoided as much as possible, since they convert the intermediate pyruvic aldehyde into alkyl acetals instead of lactate esters. A tentative mechanism for the acid catalysis is proposed based on reference reactions and isotopically labelled experiments. Reusability of the USY catalyst is demonstrated for the title reaction.

Full text of DOI:10.1039/b921284a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
Zeolite-catalysed conversion of C sugars to alkyl lactates
3
a
a
b
a
a
Paolo P. Pescarmona, Kris P. F. Janssen, Chlo e¨ Delaet, Christophe Stroobants, Kristof Houthoofd,
a
b
c
a
a
An Philippaerts, Chantal De Jonghe, Johan S. Paul, Pierre A. Jacobs* and Bert F. Sels*
Received 13th October 2009, Accepted 22nd March 2010
First published as an Advance Article on the web 5th May 2010
DOI: 10.1039/b921284a
The direct conversion of C sugars (or trioses) to alkyl lactates was achieved using zeolite catalysts.
3
This reaction represents a key step towards the efficient conversion of bio-glycerol or
formaldehyde to added-value chemicals such as lactate derivatives. The highest yields and
selectivities towards the desired lactate product were obtained with Ultrastable zeolite Y materials
having a low Si/Al ratio and a high content of extra-framework aluminium. Correlating the types
and amounts of acid sites present in the different zeolites reveals that two acid functions are
required to achieve excellent catalysis. Brønsted acid sites catalyse the conversion of trioses to the
reaction intermediate pyruvic aldehyde, while Lewis acid sites further assist in the intramolecular
rearrangement of the aldehyde into the desired lactate ester product. The presence of strong
zeolitic Brønsted acid sites should be avoided as much as possible, since they convert the
intermediate pyruvic aldehyde into alkyl acetals instead of lactate esters. A tentative mechanism
for the acid catalysis is proposed based on reference reactions and isotopically labelled
experiments. Reusability of the USY catalyst is demonstrated for the title reaction.
1
. Introduction
environment. This could be achieved through the development
of sustainable processes, e.g. using microorganisms, viz. algae, as
The synthesis of lactic acid or alkyl lactates is a process of
great appeal in the areas of renewable biomass conversion and
green chemistry: these versatile products find applications in the
13,14
a source of triglycerides.
Considering the expected increase
in the worldwide production of glycerol, its efficient conversion
to valuable chemicals, such as propylene glycol, acrolein,
propylene oxide, 1,3-propanediol, but also to lactates becomes
an environmentally and economically attractive reaction, and
represents an important step towards a complete, sustainable
1
chemical-, food-, pharmaceutical- and cosmetic industries. In
particular, they are increasingly employed as environmentally
friendly solvents and as starting materials for the synthesis of
polylactic acid (PLA), a biodegradable polymer with a wide
1,15-29
process for the use of biomass feedstock.
2,3
range of applications. The most common commercial pro-
cesses to produce lactic acid are based on carbohydrate fermen-
tation. Drawbacks of such processes include low productivities
and the need for expensive and energy-inefficient separation
and purification steps. In this context, the development of an
alternative, sustainable method for producing lactic acid and
lactates from renewable sources would be of great relevance.
Here, we report the catalytic formation of alkyl lactates from
triose monosaccharides in alcohol solvents. The two trioses, i.e.
dihydroxyacetone and glyceraldehyde, can be obtained either
In addition to the glycerol-route, dihydroxyacetone synthesis
is also feasible from natural gas sources via the thiazolium-ylide-
30
catalysed condensation of formaldehyde.
The first catalysts reported for the conversion of trioses
to lactic acid or to alkyl lactates contained transition metals
either as homogenous complexes or supported on an ion
31–34
exchange resin.
The highest yields were obtained with Cr-
based catalysts, making the process unattractive for health- and
environmental reasons. More recently, tin halides were found to
catalyse the reaction of trioses with alcohols to produce alkyl
4–9
biochemically or catalytically from glycerol, which is the
main by-product of the transesterification of triglycerides used
35
lactates with high yield and selectivity. However, the intrinsic
obstacle represented by the separation of homogeneous catalysts
from the products and the corrosive nature of tin halides pose
serious limitations to this system. Therefore, the development
of an efficient heterogeneous catalyst for a sustainable process
to convert trioses to lactic acid or its derivatives free of any
heavy metal is of utmost interest. The present work describes
the conversion of dihydroxyacetone and glyceraldehyde to alkyl
lactate in alcohol solvents, using a range of zeolites as catalysts.
Zeolites are renowned stable heterogeneous catalysts, employed
as environmentally benign alternatives to hazardous catalysts
10–12
to produce bio-diesel.
Bio-diesel represents a promising
alternative to fossil fuels provided that its production does
not affect food supply and does not cause damage to the
a
Centre for Surface Chemistry and Catalysis, K.U. Leuven, Kasteelpark
Arenberg 23, 3001, Heverlee, Belgium.
E-mail: bert.sels@biw.kuleuven.be, pierre.jacobs@biw.kuleuven.be;
Fax: (+)32 16321998
b
GROEP T Leuven Engineering College (associated with K.U. Leuven),
A. Vesaliusstraat 13, 3000, Leuven, Belgium
36,37
c
in many chemical processes,
biomass conversion.
including recent applications in
Flanders Materials Centre (FLAMAC), Technologiepark 903, 9052,
22,38,39
Zwijnaarde, Belgium
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1083–1089 | 1083

View Article Online
2
. Experimental
In a typical catalytic test, 5 ml of an alcohol solu-
tion containing dihydroxyacetone (0.4 M, prepared from the
2
.1 Materials
1
,3-dihydroxyacetone dimer, which readily converts into the
35
monomer ) and 1,4-dioxane (GC internal standard) was added
to 0.2 g of catalyst. The reaction was performed in tightly
closed glass vials (363 K) or in stainless-steel autoclaves (383
and 393 K), with stirring. The same reaction conditions and
concentrations were used when glyceraldehyde, pyruvic aldehyde
Zeolite NH
4
-Y 450 [FAU] was obtained by calcination of the
◦
NH
4
-Y zeolite ZF110 provided by Zeocat at 450 C. Na-Y
[
FAU] (CBV 100), Ultrastable zeolites Y [FAU] (CBV 500, CBV
6
00, CBV 712, CBV 720, CBV 760 and CBV 780) and ZSM-5
-ZSM-5 CBV 8020) were obtained from PQ Zeolites.
[
MFI] (NH
4
(
as a 40 wt% solution in H O) or phenyl glyoxal were used as
2
Zeolite-Beta [BEA] (H-Beta PB-1) was obtained from Uetikon.
◦
substrate. The product yields (Y, %) were determined by Gas
Chromatography (GC) on a Finnigan Trace GC Ultra (RTX-5
column, 5 m, 0.1 mm), allowing very fast analysis of each sample
(2 min) as a consequence of the rapid heating and cooling system
of the short column (Ultra-Fast Module). The assignment of
the products was confirmed by Gas Chromatography-Mass
Spectrometry analysis (GC-MS) on an Agilent 6890 N Gas
Chromatograph (WCOT fused silica column, 30 m, 0.25 mm)
coupled to an Agilent 5973 MSD Mass Spectrometer. The
conversion of dihydroxyacetone (XDHA, %) was calculated from
the sum of the yields of the two observed products, alkyl
lactate and dialkyl acetal of pyruvic aldehyde, because the direct
quantification of dihydroxyacetone by GC analysis is somewhat
Prior to use, all the zeolites were calcined in air at 450 C (from
◦
◦
-1
◦
◦
◦
-1
2
6
1
5 to 80 C at 1 C min ; 6 h at 80 C; 80 to 450 C at 1 C min ;
◦
h at 450 C). After calcination, the solids were kept at
◦
00 C.
2
.2 Characterisation
The fraction of Al located in the framework of FAU zeolite was
determined by means of solid-state Si MAS NMR analysis
2
9
40
of the calcined zeolites using the formula of Thomas. The
2
9
Si MAS NMR spectra were recorded on a Bruker MX300
spectrometer (7.0 T). 3900 scans were accumulated with a recycle
delay of 60 s. The samples were packed in 4 mm rotors. The
spinning frequency of the rotor was 5000 Hz. Tetramethylsi-
lane (TMS) was used as shift reference. Deconvolution was
performed with the Bruker WinFit software. Gaussian curves
were used for the simulation of the peaks, with the exception
of the Si(0Al) line of USY CBV600, 712 and 720, for which
a Lorentzian curve was used. The fraction of framework Al
determined for USY CBV 760 and 780 contain a higher level
of uncertainty owing to the presence of a broad peak due to
amorphous silica, which hindered the accurate simulation of the
other peaks.
13
1
hindered by its low response factor. C{ H} NMR spectra were
◦
measured at 25 C on a Bruker Avance 300 spectrometer (7.05
T, 300 MHz).
Two alternative approaches were followed for the recycling
of the catalysts. The first one is a washing procedure: initially
the reaction solution was removed from each sample; then,
5
ml of ethanol were added to each vial and the samples were
stirred for 5 min. The samples were centrifuged for 5 min at
500 rpm to deposit the solid catalyst. The supernatant ethanol
solution was removed. The washing procedure was repeated
times. After the last washing step, the supernatant solution
3
3
The fraction of Al located in the framework of H-ZSM-5 and
H-Beta zeolite was determined from Al MAS NMR spectra
was analysed by GC to confirm the absence of residual reaction
components. Finally, the samples were dried for 16 h in an oven
2
7
2
7
assuming negligible amounts of invisible Al. The Al MAS
NMR were recorded on a Bruker DSX400 spectrometer (9.4 T).
The samples were packed in 2.5 mm zirconia rotors. 36000 scans
were accumulated with a recycle delay of 100 ms. The spinning
frequency of the rotor was 20 kHz. A 0.1 M aqueous solution
◦
at 100 C. A calcination procedure was also evaluated: a washing
step with ethanol was carried out as described above, followed
◦
◦
◦
-1
by calcination in air at 450 C (from 25 to 80 C at 1 C min ;
6
◦
◦
◦
-1
◦
h at 80 C; 80 to 450 C at 1 C min ; 6 h at 450 C).
of Al(NO
3
)
3
·9H
2
O was used as reference.
3
. Results and discussion
2
.3 High-throughput experimentation catalytic tests
3
.1 Conversion of trioses over zeolites
Catalytic tests were performed by means of High-Throughput
41,42
43
Experimentation (HTE)
on a FLAMAC platform. The
The conversion of trioses in various alcohol solvents was studied
with different zeolites as catalysts. The employed zeolites were
selected in order to investigate the effect of the type and
amount of Brønsted and Lewis acid sites and of different porous
HTE platform included a Powdernium solid-handling robot,
used to weigh the solids, and a Tecan Cavro liquid-handling
workstation with integrated heating and magnetic stirring units,
used to dispense the liquids and to perform the catalytic tests.
A tailored approach was used for the kinetic studies: a separate
sample was used to measure the activity of each zeolite at each
reaction time. This approach guarantees an increased reliability
of the obtained results as opposed to conventional kinetic
experiments. The effect of concentration variations generated by
consecutive removal of sample aliquots for GC analysis during
structures: zeolite Na-Y and NH -Y 450, a series of commercial
4
+
Ultrastable zeolite Y in the H -form (USY, CBV series) with
different level of steaming and acid-leaching treatment, an
H-ZSM-5 and an H-Beta zeolite.
In a first screening of the zeolites in EtOH, the same two
main products were always observed (Table 1): ethyl lactate and
the diethyl acetal of pyruvic aldehyde. The presence of these
products suggests that the reaction starts with the formation
of a pyruvic aldehyde intermediate, from which it can proceed
towards the lactate by rearrangement with incorporation of an
43
the course of the reaction is efficiently avoided. A number of
HT catalytic tests were performed in duplicate or in triplicate:
in those cases, the average product yields are reported.
1
084 | Green Chem., 2010, 12, 1083–1089
This journal is © The Royal Society of Chemistry 2010

View Article Online
Table 1 Conversion of dihydroxyacetone (DHA) and product yields
while the dialkyl acetal becomes the major product with zeolites
a
over different zeolites in ethanol
with high Si/Al. USY CBV 600, with a 59% yield and 77%
selectivity towards ethyl lactate at 363 K, exhibits results similar
b
Bulk AlFramework
Si/Al (%)
Y
lactate
Y
acetal
35
to the best tin halide catalysts, with the advantage of being a
commercially available, structured heterogeneous catalyst.
These results evidence that the different product distributions
obtained when traversing the USY zeolite series is determined
by the nature of their aluminium sites. The USY zeolites
Entry Zeolite
(%)
(%)
S
lactate (%)
1
2
3
4
5
6
7
8
9
1
1
Na-Y
NH -Y 450
2.6
2.7
100
89
56
27
35
68
99
99
95
76
—
0
1
0
4
18
54
59
56
35
3
9
4
31
1
74
41
18
23
41
34
20
6
20
57
77
71
46
8
32
37
54
n.a.
USY CBV 500 2.6
USY CBV 600 2.6
USY CBV 712 5.8
USY CBV 720 15
USY CBV 760 27
USY CBV 780 40
employed here are prepared from zeolite NH -Y by different
4
degrees of dealumination generated by steaming and acid
leaching treatments, which influence the type and amount of
44
H-ZSM-5
H-Beta
40
9.9
—
their aluminium sites. Despite many characterisation studies,
the nature of the different aluminium sites in zeolites USY is
0
1
27
0
c
(Blank)
still a matter of debate and each of them can play a different
a
45–49
Conditions: 5 ml ethanol solution containing dihydroxyacetone (0.4 M)
catalytic role in different reactions.
While framework Al
and 0.2 g of zeolite, 363 K (6 h), batch reactor. Ylactate = mol% yield of
species are unequivocally associated with Brønsted acid sites, Al
ethyl lactate. Yacetal = mol% yield of diethyl acetal of pyruvic aldehyde.
3+
in extra-framework positions may occur in different forms (Al ,
S
lactate = ethyl lactate selectivity in mol%, calculated on the basis of lactate
2
+
+
b
Al(OH) , Al(OH) , Al(OH) , AlO(OH) and Al O ) displaying
and acetal yields (Slactate = Ylactate/(Ylactate + Yacetal). % of framework Al
2
3
2
3
29
40
27
as measured by Si NMR in the case of FAU or by Al NMR in the
Lewis acid properties. In the series from CBV 600 to CBV 780,
by increasing the Si/Al ratio, the number of Brønsted acid sites
decreases, while their strength increases (up to Si/Al ~ 5.8) and
46 c
case of H-ZSM-5 and H-Beta. At 383 K (4 h).
50,51
the amount of extra-framework aluminium is reduced.
In order to correlate the catalytic results in the conversion
of trioses with the type of Al species in the zeolite catalysts,
the fraction of the extra-framework aluminium was determined
2
9
27
using Si MAS NMR and Al MAS NMR. A clear trend can
be observed in the plot of ethyl lactate yield against the fraction
of framework Al in the tested zeolites (Fig. 1): the yield of lactate
shows a proportionality to the amount of extra-framework Al
species. Therefore, we conclude that the formation of the desired
ethyl lactate product is catalysed by the presence of Lewis acid
sites in the form of extra-framework aluminium species. On the
other hand, it is known that the reaction of aldehydes with
5
2,53
Scheme 1 Conversion of trioses over acidic zeolites. LA = Lewis Acid;
BA = Brønsted Acid.
alcohols generating acetals is catalysed by Brønsted acids.
Therefore, we infer that the formation of the diethyl acetal of
pyruvic aldehyde is due to Brønsted acid sites located in the
framework, the acetal yield being proportional to the number
and strength of these acid sites.
ethanol molecule or towards the acetal by addition of ethanol
(
Scheme 1).
While all tested zeolites in their H -form are active in
+
the conversion of dihydroxyacetone, the product yields and
distributions vary considerably with the catalyst (Table 1). The
fraction of dihydroxyacetone converted is inversely proportional
to the Si/Al ratio of the zeolite, indicating that aluminium sites
are the catalytic active centres.
Zeolite NH -Y 450 exhibits very high conversion of dihy-
4
droxyacetone, though with a very low selectivity towards the
desired ethyl lactate (Table 1, entry 2). USY CBV 500 combines
a similarly high conversion with high yield of both the ethyl
lactate and the acetal (Table 1, entry 3). High conversions are
obtained with zeolites USY CBV 600, CBV 712 and CBV 720
(
Table 1, entries 4–6), while the conversion with USY CBV 760
and CBV 780 are markedly lower (Table 1, entries 7 and 8).
When no zeolite was present, or Na-Y was used, negligible
dihydroxyacetone conversion and alkyl lactate formation was
observed (Table 1, entries 1 and 11).
The product distribution varies along the series of USY
zeolites: high selectivities towards ethyl lactate are observed
with zeolites with low values of Si/Al ratio and with a large
population of extra-framework Al species (CBV 600, CBV 712),
Fig. 1 Ethyl lactate yield over different zeolites as a function of the
fraction of aluminium located in the framework of the zeolites employed
in this work. The fraction of framework aluminium was determined
29
using Si NMR for the FAU zeolites (NH
based on Al NMR data for H-Beta and H-ZSM-5.
4
-Y 450 and USY) and was
27
46
Thus, returning to Scheme 1, the formation of pyruvic
aldehyde, i.e. the first step in the reaction pathway, is catalysed by
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1083–1089 | 1085

View Article Online
Brønsted acids. The formed pyruvic aldehyde then further reacts
either to the desired ethyl lactate on the Lewis acid sites (extra-
framework Al) or into the ethyl acetal on the strong Brønsted
acid sites. In other words, the reaction selectivity is determined
by the ratio of Lewis vs. strong Brønsted acids.
The above differentiation in catalytic active sites provides a
consistent explanation of the results obtained with the various
zeolites. This is clearly exemplified by the comparison of the
results obtained with NH
4
-Y 450, USY CBV 500 and USY CBV
6
00, which present the same type of framework [FAU] and have
analogous bulk Si/Al ratio, but have different distribution of
type of aluminium sites (Table 1). USY CBV 500 and USY CBV
6
00 are both generated by steaming of NH -Y zeolite, the degree
4
of the treatment and, therefore, the ratio of extra-framework
Al species vs. framework Al being higher for USY CBV 600
(
Table 1). The increasing selectivity towards ethyl lactate passing
from zeolite NH -Y 450 to USY CBV 500 and USY CBV
00 thus evidences how the removal of aluminium from the
4
6
framework of zeolite Y upon steaming generates catalytic species
that favour the formation of ethyl lactate while decreasing the
number of Brønsted acid sites that catalyse the formation of the
acetal product (Table 1).
The catalytic behaviour of the H-ZSM-5 and H-Beta follows
the description of the role of the different aluminium sites
discussed above. Poor conversion and low selectivity towards
ethyl lactate were obtained with H-ZSM-5 (Table 1, entry 9),
in agreement with its high Si/Al ratio (low conversion) and
high fraction of framework Al (low selectivity). H-Beta displays
intermediate conversion and selectivity (Table 1, entry 10) in line
with the Si/Al ratio and the presence of extra-framework species
Fig. 2 Conversion of dihydroxyacetone catalysed by different USY
zeolites at 363 K in ethanol: yield of ethyl lactate (a) and yield of diethyl
acetal of pyruvic aldehyde (b) as a function of reaction time.
46
in this type of zeolite.
The four catalysts that showed the highest yields of ethyl
lactate, USY CBV 500, 600, 712 and 720, were studied in detail
by following the kinetics of the formation of the two products
during the reaction (Fig. 2). Ethyl lactate and the diethyl acetal
of pyruvic aldehyde show rather different kinetic profiles: while
the yield of lactate is still increasing after 6 h of reaction,
the yield of acetal tends to reach a plateau after 3 h. This
implies that the ratio between lactate and acetal increases as
the reaction proceeds: with all the four catalysts the selectivity
towards the lactate displays an increase with increasing degree
of conversion of dihydroxyacetone (Fig. 3). This indicates that,
if there are sites suitable for catalysing its formation, ethyl
lactate is the thermodynamically favoured product of the studied
reaction. The argument is supported by the fact that the acetal
formation under acid catalysis is reversible. Therefore, the
formed diethyl acetal of pyruvic aldehyde can be reconverted
to pyruvic aldehyde and, from there, yield the desired lactate
product. In this sense, the acetal can be considered as a reservoir
of the reaction intermediate pyruvic aldehyde (Scheme 1). This
observation led us to propose that higher lactate yields would
be obtained at higher temperature. Indeed, when the reaction is
carried at 383 K and 393 K higher lactate yields are observed
Fig. 3 Selectivity towards ethyl lactate (Slactate) as a function of the
conversion of dihydroxyacetone (XDHA = Ylactate + Yacetal) over zeolites
USY CBV 500, 600, 712 and 720 at 363 K in ethanol.
52
dihydroxyacetone, while the selectivity towards the alkyl lactate
is lower with linear alcohols with longer chains.
Further tests showed that USY CBV 600 also catalyses the
selective conversion of glyceraldehyde to the corresponding alkyl
lactate (Table 3, entry 18). This observation is very interesting in
view of our challenge to convert glycerol directly to alkyl lactates.
As both trioses, viz. glyceraldehyde and dihydroxyacetone, can
be transformed into the desired lactate product, the difficult
regioselective oxidation of either the primary or the secondary
alcohol function of glycerol is not further required.
(
9
Table 2, entries 12–17). At the highest temperature, more than
0% of the starting dihydroxyacetone is converted into the
desired ethyl lactate (entry 17). This value is the highest yield
achieved for the ethyl lactate yield in this work, and rivals the best
35
soluble-tin chemistry. The use of different alcohols as solvent
does not influence significantly the degree of conversion of
1
086 | Green Chem., 2010, 12, 1083–1089
This journal is © The Royal Society of Chemistry 2010

View Article Online
Table 2 Conversion of dihydroxyacetone and product selectivity over
a
zeolite USY CBV 600 in various alcohols
Entry
ROH
Y
lactate (%)
Y
acetal (%)
Slactate (%)
b
1
1
1
1
1
1
2
3
4
5
6
7
MeOH
EtOH
n-PrOH
i-PrOH
n-BuOH
EtOH
82
81
69
83
71
91
1 (7)
18
29
16
28
9
91
81
69
83
71
91
c
a
Conditions: 5 ml of an alcohol solution containing dihydroxyacetone
b
(
0.4 M) and 0.2 g of zeolite USY CBV 600, 383 K (4 h). Yield of
c
tetramethyl diacetal of pyruvic aldehyde. At 393 K (3 h).
Fig. 4 Recycling of zeolite USY CBV 600 as catalyst for the conversion
of dihydroxyacetone, using either a washing or a calcination procedure.
Table 3 Conversion of glyceraldehyde, pyruvic aldehyde and phenyl
3.3 Reaction mechanism
a
glyoxal and product selectivity over zeolite USY CBV 600 in ethanol
The proposed mechanism for the reaction involves the formation
of pyruvic aldehyde as a reaction intermediate by dehydration
and rearrangement of dihydroxyacetone (Scheme 1). Pyruvic
aldehyde further rearranges with incorporation of a solvent
alcohol molecule to yield the corresponding alkyl lactate. This
last step bears a clear resemblance to the conversion of pyruvic
aldehyde to D-lactate catalysed by the glyoxalase system. Traces
of pyruvic aldehyde are indeed detected in the reaction mixture
and the mechanism is further supported by the presence of the
diethyl acetal product of pyruvic aldehyde, which is formed by
addition of the alcohol to the aldehyde functional group. In
Reaction
Entry T (K) time (h) Substrate
c
c
Yester (%) Sester (%)
1
1
8
9
0
383
4
76
33
95
83
89
97
54
Glyceraldehyde
363 1.5
b
Pyruvic aldehyde
addition, the reaction of pyruvic aldehyde (40 wt% in H O)
2
in ethanol using USY CBV 600 in similar reaction conditions
selectively produces ethyl lactate (Table 3, entry 19). Similarly,
phenyl glyoxal is converted selectively into the corresponding
ethyl mandelate (Table 3, entry 20).
2
363 2.5
Isotope labelling experiments using CD OD showed no
3
Phenyl glyoxal
incorporation of deuterium into the hydrocarbon backbone of
methyl lactate when obtained from pyruvic aldehyde (step 2 top
reaction, Scheme 1). This is clear from the spectral data (MS
a
Conditions: 5 ml of an ethanol solution containing the chosen substrate
b
(
0.4 M) and 0.2 g of zeolite USY CBV 600. As a 40 wt% solution
c
in water. Formation of ethyl lactate from glyceraldehyde and pyruvic
aldehyde; formation of ethyl mandelate from phenyl glyoxal.
1
3
and C NMR data, summarized in ref. 55) of the formed methyl
1
3
lactate product: there is no triplet signal observed in C NMR
for the –CH group, while the typical fragment bearing the –
3
3.2 Recycling of the catalysts
CH group is not deuterated. As no deuterium incorporation is
3
observed in the lactate product, the mechanism can be regarded
In view of a possible industrial application, it is important
to evaluate the reusability of the catalysts: for this purpose,
commercial zeolite USY CBV 600, the best catalyst identified
in this work, was tested in several catalytic cycles at 363 K in
ethanol (6 h). Before each successive catalytic test, the zeolites
were recovered either by washing with ethanol and drying at
as a formal intramolecular Cannizzaro reaction of the a-keto
aldehyde, viz. pyruvic aldehyde. The mechanism is very similar
to that of the glyoxalase system and their reported Lewis-acid
31,32,54,56
catalysed biomimetic reactions.
(Scheme 2).
3
73 K or by washing with ethanol followed by calcination in air
at 723 K. The calcination step guarantees a complete removal
of organics present on the catalyst but might as well induce
undesired variations in the zeolite structure.
USY CBV 600 can be recycled without significant loss of
activity both by washing and by calcination (Fig. 4), proving
the stability of this material under the employed reaction and
recycling conditions. However, it should be noted that the
selectivity towards ethyl lactate slightly decreases from the first
to the second run and then remains constant in the successive
cycles. Since both recycling procedures prove equally efficient,
for energetic reasons the washing procedure might be the most
favourable for future applications.
Scheme 2 Conversion of pyruvic aldehyde to methyl lactate in
CD OD. Spectral data of the product in ref. 55.
3
Usually, the reaction in Scheme 2 has been reported in
57
presence of a base catalyst, viz. NaOH, but also a Lewis
acid is capable of catalysing the reaction by activating the
carbonyl function of the aldehyde, allowing the fast attack of
a methoxide moiety to the C=O carbonyl function, followed by
an intramolecular rearrangement with a shift of the hydride. It is
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1083–1089 | 1087

View Article Online
known that when such an internal Cannizzaro reaction is carried
out in deuterated solvents, no deuteration of the a position to the
alcohol function occurs. This is indeed observed for the reaction
of Scheme 2 in presence of the USY zeolite catalyst.
CBV 600 displayed the best catalytic results, achieving similar
lactate yields as the best homogeneous catalysts known for
this reaction, with the advantage of being a commercially
available heterogeneous catalyst that can be reused in various
successive cycles without loss of activity. Besides the reusability,
the presented heterogeneous catalyst is free of any heavy metals
like Sn or Cr, which is key factor on the way to a sustainable
process.
When the triose dihydroxyacetone or glyceraldehyde instead
of pyruvic aldehyde was used as substrate in CD OD, it was
3
converted to methyl lactate with one deuterium atom incorpo-
rated into the methyl group of the lactate product (the MS and
1
3
C NMR data are collected in ref. 58). Since in the second step
A subtle combination of acid sites, viz. mild Brønsted acidity
and Lewis acidity in the form of extra-framework aluminium
species, catalyses the formation of alkyl lactates, while strong
Brønsted acidity, as typically present in zeolites with high Si/Al
of the mechanism deuterium is not substituting hydrogen in the
hydrocarbon backbone (see Scheme 2), deuterium incorporation
likely occurs after the dehydration step, catalysed by the
Brønsted acid sites, via tautomerisation of the enol intermediate
5
9
60
53
ratio, in sulfonated silica or carbon, in acid resins and
52
into CH
2
D-labelled methyl(d
3
) lactate (Scheme 3).
in soluble strong acids, viz. H
2
SO , is detrimental because it
4
catalyses the formation of the diethyl acetal of pyruvic aldehyde
side product, thus reducing the selectivity towards the desired
lactate product.
The method reported here offers a valid alternative route to
the slow and complex fermentation process presently used to
produce lactic acid. The next step will involve the extension of
the catalytic process to the direct conversion of glycerol to lactic
acid and derivatives, with the aim of finding a more suitable
process compared with the recently reported hydrothermal
reaction requiring stoichiometric amounts of NaOH and high
61
temperatures or with the liquid-phase reaction via oxidative
dehydrogenation catalysed by noble metals supported on car-
bon, which has moderate selectivity and also requires relatively
62,63
high temperatures and high concentrations of a strong base.
3
Scheme 3 Conversion of trioses to methyl lactate in CD OD with
incorporation of deuterium during the dehydration/tautomerization
step. A selection of the spectral data of the lactate product is summarized
in ref. 58.
Acknowledgements
The authors appreciate sponsorship in the framework of a
FLAMAC-IWT project, and are grateful to the following
institutions and research programs for their support: FWO, IAP,
GOA, VIRKAT, IDECAT, CECAT and Methusalem (CASAS,
long-term financing from the Flemish government).
On the basis of the above-mentioned mechanism and of
the catalytic results, a complete description of the reactions
occurring is achieved. The dehydration of trioses leading to the
pyruvic aldehyde intermediate is catalysed by Brønsted acid sites,
which are present in all the tested materials with the exception of
the inactive Na-Y. Mild Brønsted acidity is sufficient to catalyse
this step, as indicated by the formation of pyruvic aldehyde as the
sole product when acetic acid is used as a reference weak acid
catalyst. From pyruvic aldehyde, the reaction proceeds to the
dialkyl acetal in the presence of strong zeolitic Brønsted acidity,
while the formation of alkyl lactate from pyruvic aldehyde is
catalysed by extra-framework aluminium sites that act as Lewis
acid sites, in agreement with previous work from other groups on
References
1 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411.
2
R. E. Drumright, P. R. Gruber and D. E. Henton, Adv. Mater., 2000,
1
2, 1841.
3
E. T. H. Vink, K. R. R a´ bago, D. A. Glassner, B. Springs, R. P.
O’Connor, J. Kolstad and P. R. Gruber, Macromol. Biosci., 2004, 4,
551.
4
5
A. Gupta and K. Singh, J. Mol. Microbiol. Biotechnol., 2001, 3, 445.
H. Kimura, K. Tsuto, T. Wakisawa, Y. Kazumi and Y. Inaya, Appl.
Catal., A, 1993, 96, 217.
31–34,56
homogeneous systems.
This explains why USY CBV 600,
presenting a high amount of extra-framework with the lowest
fraction of framework aluminium species (Table 1), displays
good activity combined with the highest lactate selectivity. It
is thus concluded that a material with a suitable combination
of Lewis acid and Brønsted acid sites, like in USY CBV 600,
is required to efficiently and selectively catalyse the dual-site
mechanism for the conversion of trioses to alkyl lactates.
6 R. Garcia, M. Besson and P. Gallezot, Appl. Catal., A, 1995, 127,
165.
7
8
9
R. Ciriminna, G. Palmisano, G. Della Pinna, M. Rossi and M.
Pagliaro, Tetrahedron Lett., 2006, 47, 6993.
S. Demirel, K. Lehnert, M. Lucas and P. Claus, Appl. Catal., B, 2007,
70, 637.
A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int. Ed., 2007,
4
5, 7896.
1
1
0 H. Fukuda, A. Kondo and H. Noda, J. Biosci. Bioeng., 2001, 92, 405.
1 M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi and C. Della Pinna,
Angew. Chem., Int. Ed., 2007, 46, 4434.
4
. Conclusions
1
2 P. T. Vasudevan and M. Briggs, J. Ind. Microbiol. Biotechnol., 2008,
3
5, 421.
Zeolites with different number and type of acid sites were
studied as catalysts for the liquid-phase conversion of dihy-
droxyacetone or glyceraldehyde to alkyl lactate. Zeolite USY
1
1
3 B. E. Rittmann, Biotechnol. Bioeng., 2008, 100, 203.
4 B. Smith, H. C. Greenwell and A. Whiting, Energy Environ. Sci.,
2009, 2, 262.
1
088 | Green Chem., 2010, 12, 1083–1089
This journal is © The Royal Society of Chemistry 2010

View Article Online
1
5 B. F. Sels, E. D’Hondt and P. A. Jacobs, in: Catalysis for Renewables,
ed. G. Centi and R. A. Van Santen, Wiley-VCH, Weinheim, 2007,
ch. 11, p. 223.
6 A. Behr, J. Elting, K. Irawadi, J. Leschinski and F. Lindner, Green
Chem., 2008, 10, 13.
7 M. Pagliaro and M. Rossi, in: Future of Glycerol: New Usages for a
Versatile Raw Material, 2008, RSC Publishing.
40 J. M. Thomas, J. Mol. Catal., 1984, 27, 59.
41 P. P. Pescarmona, J. C. van der Waal, I. E. Maxwell and T.
Maschmeyer, Catal. Lett., 1999, 63, 1.
42 W. F. Maier, K. St o¨ we and S. Sieg, Angew. Chem., Int. Ed., 2007, 46,
6016.
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
43 K. P. F. Janssen, J. S. Paul, B. F. Sels and P. A. Jacobs, Appl. Surf.
Sci., 2007, 254, 699.
8 D. A. Simonetti, J. Rass-Hansen, E. L. Kunkes, R. R. Soares and
J. A. Dumesic, Green Chem., 2007, 9, 1073.
44 M. J. Remy, D. Stanica, G. Poncelet, E. J. P. Feijen, P. J. Grobet,
J. A. Martens and P. A. Jacobs, J. Phys. Chem., 1996, 100,
12440.
9 S. Liu, M. Rebros, G. Stephens and A. C. Marr, Chem. Commun.,
2
009, 2308.
45 G. Catana, D. Baetens, T. Mommaerts, R. A. Schoonheydt and B. M.
Weckhuysen, J. Phys. Chem. B, 2001, 105, 4904.
46 A. Omegna, J. A. van Bokhoven and R. Prins, J. Phys. Chem. B,
2003, 107, 8854.
47 C. J. A. Mota, D. L. Bhering and N. Rosenbach Jr., Angew. Chem.,
Int. Ed., 2004, 43, 3050.
0 J. Chaminand, L. Djakovitch, P. Gallezot, Ph. Marion, C. Pinel and
C. Rosier, Green Chem., 2004, 6, 359.
1 Z. Yu, L. Xu, Y. Wei, Y. Wang, Y. He, Q. Xia, X. Zhang and Z. Liu,
Chem. Commun., 2009, 3934–3936.
2 E. D’Hondt, S. Van de Vyver, B. F. Sels and P. A. Jacobs, Chem.
Commun., 2008, 6011.
48 J. Jiao, S. Altwasser, W. Wang, J. Weitkamp and M. Hunger, J. Phys.
Chem. B, 2004, 108, 14305.
3 W. Girke, H. Klenk, D. Arntz, T. Haas and A. Neher, US Pat.
5
387720, 1995.
49 J. Huang, Y. Jiang, V. R. R. Marthala, B. Thomas, E. Romanova and
M. Hunger, J. Phys. Chem. C, 2008, 112, 3811.
50 D. Barthomeuf, Mater. Chem. Phys., 1987, 17, 49.
51 W. E. Farneth and R. J. Gorte, Chem. Rev., 1995, 95, 615.
52 P. Y. Bruice, in: Organic Chemistry (5th edn), Pearson Prentice Hall,
2007.
53 S. K. Gupja, Br. Pat. GB 1473782, 1977.
54 F. Himo and P. E. M. Siegbahn, J. Am. Chem. Soc., 2001, 123,
10280.
4 Y. Gu, A. Azzouzi, Y. Pouilloux, F. Jerome and J. Barrault, Green
Chem., 2008, 10, 164.
5 J. Barrault and F. Jerome, Eur. J. Lipid Sci. Technol., 2008, 110,
8
25–830.
6 S.-H. Chai, H.-P. Wang, Y. Liang and B.-Q. Xu, Green Chem., 2008,
1
0, 1087.
7 M. G. Musolino, L. A. Scarpino, F. Mauriello and R. Pietropaolo,
Green Chem., 2009, 11, 1511.
8 A.-Y. Yin, X.-Y. Guo, W.-L. Dai and K.-N. Fan, Green Chem., 2009,
55 Own selected data for methyl lactate prepared from pyruvic aldehyde
1
3
1
1
1, 1514.
in the presence of USY in CD
3
OD as solvent: C{ H} NMR
9 H. Garc ´ı a-Mar ´ı n, J. C. van der Toorn, J. A. Mayoral, J. I. Garc ´ı a and
I. W. C. E. Arends, Green Chem., 2009, 11, 1605.
(300 MHz, CD
46 ([CH
3
OD): d 68.28 ppm (s, –CH(OD)–); GC-MS: m/z =
+
3
CH(OD)] , 100%).
0 R. Dostalek, B. Knuth and R. Schneider, Ger. Pat. DE 19543415 A1,
56 A. E. Russell, S. P. Miller and J. P. Morken, J. Org. Chem., 2000, 65,
1
996.
8381.
1 J. Eriksen, O. Mønsted and L. Mønsted, Transition Met. Chem.,
998, 23, 783.
2 E. Bang, J. Eriksen, L. Mønsted and O. Mønsted, Acta Chem. Scand.,
57 J. March, Advanced Organic Chemistry: Reactions, Mechanisms and
Structures (4th edn), John Wiley & Sons, 1992.
58 Own selected data for methyl lactate prepared from dihydroxyacetone
1
13
1
1
994, 48, 12.
in the presence of USY in CD
(300 MHz, CD O+D): d 20.71 ppm (t, –CH
([CDH CH(OD)] , 100%).
3
OD as solvent: C{ H} NMR
3
3
3
3
3 R. L. Kelly, Eur. Pat. Appl. EP 460831 A2, 1991.
3
2
D); GC-MS: m/z = 47
4 R. L. Kelly, Eur. Pat. Appl. EP 541330 A2, 1993.
2
5 Y. Hayashi and Y. Sasaki, Chem. Commun., 2005, 1716.
6 ‘Introduction to Zeolite Science and Practice’, in Studies in Surface
Science and Catalysis, Vol. 137 (2nd edn), ed. H. van Bekkum,
E. M. Flanigen, P. A. Jacobs and J. C. Jansen, Elsevier, Amsterdam,
59 W. M. Van Rhijn, D. E. De Vos, B. F. Sels, W. D. Bossaert and P. A.
Jacobs, Chem. Commun., 1998, 317.
60 L. Peng, A. Philippaerts, X. Ke, J. Van Noyen, F. De Clippel, G.
Van Tendeloo, P. A. Jacobs and B. F. Sels, Catal. Today, 2010, 150,
140.
2
001.
3
3
3
7 P. Gallezot, Green Chem., 2007, 9, 295.
61 H. Kishida, F. Jin, Z. Zhou, T. Moriya and H. Enomoto, Chem.
8 A. Fukuoka and P. L. Dhepe, Angew. Chem., Int. Ed., 2006, 45, 5161.
9 A. Philippaerts, S. Paulussen, S. Turner, O. I. Lebedev, G. Van
Tendeloo, H. Poelman, M. Bulut, F. De Clippel, P. Smeets, B. Sels
and P. Jacobs, J. Catal., 2010, 270, 172.
Lett., 2005, 34, 1560.
62 E. P. Maris and R. J. Davis, J. Catal., 2007, 249, 328.
63 E. P. Maris, W. C. Ketchie, M. Murayama and R. J. Davis, J. Catal.,
2007, 251, 281.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1083–1089 | 1089