Catalytic upgrading of levulinic acid to ethyl levulinate using reusable silica-included Wells-Dawson heteropolyacid as catalyst

Article abstract of DOI:10.1016/j.catcom.2011.12.004

In this paper we report, for the first time, the direct incorporation of a heteropolyacid (HPA) with Wells-Dawson structure during the synthesis of silica by the sol-gel technique, in acidic media, using tetraethyl orthosilicate. The catalyst characterization was carried out by ³¹P MAS-NMR, FT-IR, XRD, N₂ adsorption-desorption measurements, and the acidic properties were determined through potentiometric titration with n-butylamine. The synthesized catalysts were used in the esterification of levulinic acid with ethanol, at 78 °C, to obtain ethyl levulinate. The synthesis of silica-included HPAs was satisfactory, and the samples kept their HPA structure intact after synthesis. The catalytic tests for the esterification reaction between levulinic acid and ethanol to produce ethyl levulinate have shown that the silica-included Wells-Dawson HPA is an active and selective catalyst for this reaction. It must be noted that silica-included HPAs also kept their structure and catalytic activity after three consecutive reaction cycles. These results indicated that these solid acids are promissory catalysts for the esterification reaction of levulinic acid and ethanol to ethyl levulinate.

Full text of DOI:10.1016/j.catcom.2011.12.004

Catalysis Communications 18 (2012) 115–120
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Catalytic upgrading of levulinic acid to ethyl levulinate using reusable silica-included
Wells-Dawson heteropolyacid as catalyst
Gustavo Pasquale a, Patricia Vázquez b, Gustavo Romanelli b, Graciela Baronetti a,
⁎
a
Departamento de Ingeniería Química, Facultad de Ingeniería, Universidad de Buenos Aires, Ciudad Universitaria, C1428BG, Buenos Aires, Argentina
Centro de Investigación y Desarrollo en Ciencias Aplicadas “Dr. J.J. Ronco” (CINDECA), Facultad de Ciencias Exactas, UNLP-CCT La Plata-CONICET, Calle 47 No. 257, B1900AJK,
b
La Plata, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper we report, for the first time, the direct incorporation of a heteropolyacid (HPA) with Wells-Dawson
structure during the synthesis of silica by the sol–gel technique, in acidic media, using tetraethyl orthosilicate. The
catalyst characterization was carried out by 31P MAS-NMR, FT-IR, XRD, N2 adsorption–desorption measurements,
and the acidic properties were determined through potentiometric titration with n-butylamine. The synthesized
catalysts were used in the esterification of levulinic acid with ethanol, at 78 °C, to obtain ethyl levulinate. The
synthesis of silica-included HPAs was satisfactory, and the samples kept their HPA structure intact after synthesis.
The catalytic tests for the esterification reaction between levulinic acid and ethanol to produce ethyl levulinate
have shown that the silica-included Wells-Dawson HPA is an active and selective catalyst for this reaction. It
must be noted that silica-included HPAs also kept their structure and catalytic activity after three consecutive
reaction cycles. These results indicated that these solid acids are promissory catalysts for the esterification reaction
of levulinic acid and ethanol to ethyl levulinate.
Received 22 September 2011
Received in revised form 22 November 2011
Accepted 1 December 2011
Available online 8 December 2011
Keywords:
Ethyl levulinate production
Wells-Dawson heteropolyacid incorporated
in silica
Upgrading of levulinic acid
Heterogeneous catalysis
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
is well known that the use of conventional liquid acids has environmen-
tal problems associated with handling, containment, disposal and
Biomass is defined as any organic matter that is available on a
renewable basis, including dedicated energy crops and trees,
agricultural food and feed crop residues, aquatic plants, wood and
wood residues, animal wastes and other waste materials [1,2]. The
annual production of biomass is about 1.7–2.0×1011 tons [3]; however,
only 6×109tons is currently used for food and non-food applications.
The biomass carbohydrates are the most abundant renewable resources
available, and they are currently viewed as a feedstock for the Green
Chemistry of the future [4–6].
Levulinic acid (LA), derived from the acid catalysis from lignocellu-
losic biomass, one of the top-twelve building blocks [2], is a potentially
versatile building block for the synthesis of several chemicals for
applications such as fuel additives, polymer and resin precursors
[7–11]. Several reviews have been published describing the properties
and potential industrial applications of LA and its derivatives [1,12,13].
There are numerous useful compounds derived from levulinic acid.
For example, ethyl levulinate (EL), obtained by esterifying LA with
ethanol, can be used as an oxygenate additive in fuels. EL and other
levulinic acid esters can be obtained by esterification reaction in
the presence of an acid catalyst, such as sulphuric, polyphosphoric or
p-toluenesulfonic acid in homogeneous medium [14–16]. However, it
regeneration due to their corrosive and toxic nature. The use of solid
acid (heterogeneous) catalysts in organic synthesis and in the industrial
manufacture of chemicals is increasingly important since they provide
green alternatives to homogeneous catalysts [17,18]. On the other
hand, catalysis by heteropolyacids (HPAs) is a well-established area
and it has the potential of a great economic reward and green benefits
[19].
Although there are many structural types of heteropolycompounds,
the majority of the catalytic applications use the most common
commercial Keggin-type structure and then the Wells-Dawson one,
owing to their availability and chemical stability [20,21]. However, the
application of phosphotungstic acid with Wells-Dawson structure as
catalyst is a field of growing importance in sustainable acid catalysis
[22].
Catalytic processes using HPAs as solid acid catalysts have many
advantages over homogeneous liquid acid catalysis. They are no
corrosive, cheap and environmentally friendly, presenting fewer
disposal problems. However, their high solubility in water and polar
solvents, as well as the low specific area of bulk HPAs, limits their
application as heterogeneous catalysts. So attempts to stabilize
HPAs by supporting them on several carriers such as silica, alumina,
titania, among others, have been made [23,24]. The interaction of
the acids with these supports has led to a sharp improvement of the
catalytic properties; however, the supported HPA catalysts obtained
by impregnation techniques are unsuitable for their use as catalysts
⁎
Corresponding author.
E-mail address: baroneti@di.fcen.uba.ar (G. Baronetti).
1566-7367/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.12.004

116
G. Pasquale et al. / Catalysis Communications 18 (2012) 115–120
in polar reaction media due to a continuous leaching of the HPAs. In
order to avoid any HPA leaching, recently, Castanheiro's group has
reported the esterification of free fatty acids with methanol using
Keggin heteropolyacids immobilized on silica [25].
scanning angle between 5 and 60° of 2θ at a scanning rate of 2° per
min. Specific surface areas (SBET) and textural characteristics of the
catalysts were determined by the nitrogen adsorption/desorption
technique using Micromeritics ASAP 2020 equipment.
Moreover, Popa et al. [26] studied the synthesis of heteropolyacid–
mesoporous silica composites, which was carried out in acidic media
by different methods. The effect of incorporating heteropolyacids
(H3PMo12O40 and H4PVMo11O40) species into organized mesoporous
silica was studied by using nonionic and cationic surfactants. Besides,
these authors included a Keggin-type heteropolyacid in silica using a
sol–gel technique by the hydrolysis of ethyl orthosilicate [27].
In this paper we report, for the first time, the direct incorporation of
a heteropolyacid with Wells-Dawson structure during the synthesis of
silica by the sol–gel technique, in acidic media, using tetraethyl
orthosilicate. The catalyst characterization was carried out by 31P
MAS-NMR, FT-IR, XRD, N2 adsorption–desorption measurements, and
the acidic properties were determined through potentiometric titration
with n-butylamine. In addition, the esterification reaction (in heteroge-
neous conditions) of levulinic acid with absolute ethanol at 78 °C, by
incorporating the phosphotungstic HPA with Wells-Dawson structure
(WD) as active phase in a silica framework (Scheme 1), is reported.
The acidic properties of the samples were determined by poten-
tiometric titration using a solution of n-butylamine in acetonitrile
in a Metrohm 794 Basic Titrino apparatus. A 0.05 mL portion of
n-butylamine (0.1 N), in acetonitrile, was added to a known mass of
solid (between 0.1 and 0.05 g) using acetonitrile as solvent, and stirred
for 3 h. Later, the suspension was titrated with the same base at
0.05 mL/min. The electrode potential variation was measured in a
Metrohm 794 Basic Titrino apparatus with a double junction electrode.
Catalytic tests were carried out in a stirred batch reactor at 78 °C.
In a typical experiment, the reactor was loaded with 30 mmol of
absolute ethanol, 250 mg of 40WD-S and 2 mmol of levulinic acid.
The reaction was followed by TLC (thin layer chromatography until
a reaction time of 10 h, using a 4:1 mixture of hexane:ethylacetate
as solvent). TLC aluminum sheets (silica gel 60 F254 Merck) were
used. The catalyst was filtered off and washed twice with absolute
ethanol, 2 mL each/with 2 mL absolute ethanol. The filtrate and the
washing liquids were combined and concentrated in vacuum. The
residue was dissolved in CH2Cl2 (10 mL) and the solution was washed
with NaHCO3 5% (3×2 mL) and water (1×5 mL) to separate the
levulinic acid. The organic phase was dried with anhydrous Na2SO4
and the solvent was evaporated to afford crude ethyl levulinate. The
product was identified via comparison with an authentic sample of
ethyl levulinate and mass spectra analysis. The reaction yield was
expressed as the ratio of % moles of product to moles of initial
substrate. Ion impact mass spectrum of ethyl levulinate: MS (EI),
70 eV, m/z (rel. intensity): 144 (4) [M+], 129 (22), 99 (67), 74 (20),
43 (100), 29 (17) [29].
2. Experimental
2.1. Synthesis of silica-included Wells-Dawson HPA
The Wells-Dawson HPA included in silica was synthesized by the
sol–gel technique. A mixture of n-butanol and Wells-Dawson acid
(WD) and, finally, water was added to tetraethyl orthosilicate. The
mixture was stirred under nitrogen atmosphere for 1 h, at room
temperature. Then, this mixture was stirred at 40–60 °C for 24 h.
The hydrogel obtained was dehydrated in an oven at 80 °C and the
catalyst samples were used without calcination in the catalytic tests,
because the acidity of this type of materials depends on their
hydration state and on the treatment temperature [22,28].
We synthesized two catalysts with two different amounts of WD
acid included in silica. For the catalyst named 20WD-S, 2.7 g of pure
WD acid were included, and for 40WD-S catalyst, 5 g of pure WD
acid. In both cases, the amount of pure silica was 14.7 g. By sol–gel
technique, for each sample, the total amount of WD acid is incorpo-
rated in the silica framework. Before being incorporated in silica,
the bulk pure Wells-Dawson acid was synthesized according to the
method developed in our laboratory [28]. Keggin-structure HPAs
included in silica catalysts were also synthesized by the sol–gel
technique in order to compare them with the Wells-Dawson HPA.
Stability tests of the silica-included HPA catalysts were carried out
by running four consecutive experiments under the same reaction
conditions. After each test, the catalyst was separated from the
reaction mixture by filtration, washed with absolute ethanol
(2×2 mL), dried under vacuum and then reused. The filtrate was
analyzed for W content via ICP, but no W traces were detected.
3. Results and discussion
3.1. Catalysis characterization
3.1.1. SBET measurements
Table 1 lists the surface area, average pore diameter and pore
volume of the different synthesized samples and silica obtained by
sol–gel. Bulk Wells-Dawson acid has a very low value of SBET (2–4 m2/
g), and pure silica 507 m2/g, respectively. For 40WD-S and 20WD-S,
the SBET are 531 and 580 m2/g, with average pore diameter of 25.8
and 26.7 Å, respectively. The values of SBET are higher than the pure
silica because WD incorporated in framework of silica acts as a pore
forming. This can also be observed in the pore volume values
(Table 1). The difference between 20WD-S and 40WD-S may be due
to the catalyst contains the highest amount of WD could be incorporated
into the framework silica and, also, remains in the mouth of the pores,
reducing the three measured parameters (Table 1).
In this case, a commercial phosphomolybdic acid (H3PMo12O40
,
sample 40K-S) was used. Wells-Dawson synthesized samples were
characterized by 31P MAS-NMR, FTIR, SBET
, and potentiometric
titration by n-butylamine techniques.
31P MAS-NMR spectra were recorded in Bruker MSL-300 equipment.
Chemical shifts were expressed in parts per million with respect to 85%
H3PO4 as an external standard. FTIR spectra were obtained in Nicolet
IR.200 equipment. Pellets in KBr and
a measuring range of
400–4000 cm−1 were used to obtain the FT-IR spectra of the solid
samples at room temperature. Power XRD patterns were recorded in
Phillips PW-1732 equipment with built-in recorder, using Cu Kα radia-
tion, nickel filter, 20 mA and 40 kV in the high voltage source, and
Table 1
Surface area of different samples.
Sample
Surface area:
BET, m2/g
Average pore
diameter, Å
Pore volume,
cm3/g
S
Bulk WD
Pure SiO2
20WD-S
40WD-S
2–4
507
580
531
–
–
21.9
26.7
25.8
0.18
0.35
0.20
Scheme 1. Ethyl levulinate synthesis.

G. Pasquale et al. / Catalysis Communications 18 (2012) 115–120
117
3.1.2. 31P MAS-NMR of silica-included WD samples
In order to verify the incorporation of WD in the silica framework,
the synthesized samples were characterized by 31P MAS-NMR. This
technique is the “fingerprint” of the HPA compounds. The pure bulk
WD acid has two equivalent phosphorus atoms and consequently, it
shows only one main peak in the 31P MAS-NMR spectrum in the
range of −12.8, −13 ppm [30]. The spectra of 31P MAS-NMR of WD
included in SiO2 catalysts with two different loadings of WD
(40WD-S and 20WD-S, a and b curves, respectively), are shown in
Fig. 1. The results indicated that silica-included WD samples display
the main peak with a chemical shift at −13.2 ppm, which indicates
that after the synthesis and drying, the acid maintains its Wells-
Dawson heteropolyanion structure. Two new small signals at −12.4
and −11.7 ppm approximately are evident in the samples. These
signals could be related to the presence of different Dawson species,
a
b
PO
4
W=O
1000
W-O-W
800
1200
600
Wavenumber, cm-1
such as H6P2W18O62 strongly interacting with the Si–OH groups of
Fig. 2. FTIR spectra of bulk WD (curve a) and 40WD-S (curve b) samples.
6−
the support, and to species such as P2W21
O
71
, respectively [30].
3.1.4. Potentiometric titration measurements using n-butylamine
3.1.3. FT-IR measurements
The n-butylamine is considered a strong base, so its adsorption on
sites of different acid strength could be expected. The total solid acidity
is titrated without distinguishing the type of acidity. At this point, it is
interesting to comment that the development of a potentiometric
method was thought of taking into account that the acid and basic
functions of a solid are not properly defined thermodynamically.
Moreover, an indicator test such as the Hammet method, though
considered as a reference technique, is difficult to apply to certain solids.
The method used here is based on the observation that the potential
difference is mainly determined by the acidic environment around the
electrode membrane. The measured electrode potential is an indicator
of the acidic properties of dispersed solid particles. The acidic properties
of the solid samples measured by this technique enable the evaluation
of the number of acid sites and their acid strength. In order to interpret
the results, it is suggested that the initial electrode potential (E)
indicates the maximum acid strength of the surface sites, and the values
(meq/g solid), where the plateau is reached, indicate the total number
of acid sites. The acid strength of surface sites can be assigned according
to the following ranges: very strong site, E>100 mV; strong site,
0bEb100 mV; weak site, −100bEb0 mV, and very weak site, Eb
−100 mV.
On the other hand, the FTIR spectra (Fig. 2) also show that after
synthesis the WD keeps its Dawson structure. This figure shows the
spectra for the bulk WD and for the 40WD-S sample, after subtraction
of the spectra that correspond to the support. The characteristic
bands of the HPA with Dawson structure are 1091 (stretching
frequency of the PO4 tetrahedron), 963 (W_O terminal bonds), 911
and 778 cm−1 (“inter” and “intra” W–O–W bridges, respectively)
[30]. It can be observed that the acid included in silica displays the
same characteristic bands. Nevertheless, for these samples
a
broadening of the band at 1091 cm−1 is observed. This fact can be
due to a loss of tetrahedron symmetry [30] because of the interaction
between WO6 octahedral and sylanol groups of silica. A shift of the
778 cm−1 band (“intra” W–O–W bridges) can also be observed,
which could be attributed to the same effect.
On the other hand, XRD measurements of the silica-included
samples only showed a broad diffraction peak corresponding to the
silica support, without the characteristic lines for WD acid as we
reported in previous work for supported catalysts [30]. This indicates
that WD HPAs are included as small particles not detectable by XRD.
Similar results were obtained by Popa et al. [27]
Both 31P MAS-NMR and FTIR measurements show that the Wells-
Dawson acid keeps its Dawson structure after its incorporation in the
silica framework.
The acidity of bulk Wells-Dawson and Keggin HPAs is Brönsted in
nature, and their initial electrode potential indicated that both are
superacids [31,32]. Furthermore, the potentiometric curves obtained
by titration of silica-included WD catalysts with n-butylamine and
their comparison with that of the Keggin structure (40K-S) are
summarized in Fig. 3. Silica-included HPA catalysts also show very
31P MAS-NMR
500
400
300
200
a
b
40DW-S
100
20DW-S
-5
-8
-11
-14
-17
-20
40K-S
0
0,0
0,2
0,4
0,6
0,8
1,0
, ppm
δ
mEq n-butylamine/ g solid
Fig. 1. 31P MAS-NMR spectra of silica-included Wells-Dawson HPA samples, a) 40WD-S
Fig. 3. Curve of potentiometric titration of different silica-included samples: 40WD-S;
and b) 20WD-S.
20WD-S; and 40K-S.

118
G. Pasquale et al. / Catalysis Communications 18 (2012) 115–120
Table 3
strong acid sites after their incorporation in the silica framework; this
could be related to the proton mobility, which in part depends on the
negative charge distribution in the heteropolyanion oxygen's. It is
observed that the increase in the concentration of WD does not cause
an important variation in the acidic properties but there is a variation
of plateau start for both cases. Compared with the Keggin structure,
this property shows a similar behavior when HPA is included in the
silica network. It is possible that negative charges of WD anions are
located on the external oxygen atoms, so the attraction of the ternary
structure to protonated OH groups is probable. For each WD anion
there are 6H+ eventually capable of acting on OH groups of sol–gel
silica.
If an interaction occurs between these OH groups and the protons,
giving rise to OH2+ groups, there are probably enough OH groups to
complete the interaction, but only a small amount of them (those that
are sterically favorable) can be available for the interaction with both
the WD anions and the protons. The acidity of these bulk compounds
is higher than that of mineral acids and it depends on their hydration
state, the strength of acid sites and proton accessibility [22].
Stability studies of the 40WD-S catalyst on the esterification of levulinic acid with
ethanol. Reuse of the catalyst in four consecutive experiments.
Entry
Catalytic cycle
Yield to EL (%)
1
2
3
4
First use
76
68
68
68
1st re-use
2nd re-use
3rd re-use
Experimental conditions: 0.01 mmol catalyst, 2 mmol of levulinic acid, 7.5 mL
(128 mmol) of absolute ethanol; 10 h, 78 °C, molar ratio ethanol/LA: 64/1.
HPA structures, the catalytic activity decreases from bulk HPA to the
included ones. These results are in agreement with the acidity
measurements explained in detail in Section 3.1.2.
Besides, experiences using different amounts of active 40WD-S
catalyst (20, 100, 500 and 1000 mg) were performed (Table 2, entries
7 to 10). The results showed that an increase in the amount of
catalysts from 100 to 1000 mg only produces a slightly increase in
the yield to EL.
3.2. Catalytic tests
3.2.2. Catalyst recycling
As mentioned before, the disadvantages of HPAs as heterogeneous
catalysts lie in their low specific area and high solubility in polar
media, such as ethanol used as a reactant for the esterification
reaction. After the direct incorporation in silica, WD samples have a
higher specific surface area, but it is necessary to avoid any leaching
of the WD acid during the reaction. In fact, the catalytic stability of
the fresh synthesized 40WD-S catalyst was evaluated in the esterifi-
cation reaction using a molar ratio of absolute ethanol/EL of 64:1, by
performing consecutive batch reactions with the same samples, as is
described in detail in the experimental section. After the first run
the yield decreased, and after the second batch the catalyst activity
stabilized during three reaction cycles. The EL yields to the four cycles
were 76%, 68%, 68% and 68%, respectively (Table 3). The filtrate was
analyzed for W content via ICP, but no W traces were detected.
According to these results, an ethanol pretreatment of the fresh
synthesized WD sample before the esterification reaction was done.
The sample was refluxed in absolute ethanol, for 48 h, and then this
washed sample was filtered and dried under vacuum. ICP analysis of
the filtrate indicated that the 10% of WD active phase was lost after
this pretreatment with ethanol. Catalytic tests using this ethanol-
washed sample were done. The results indicated that after this
3.2.1. Performance using different catalysts
This work describes, for the first time, the application of heteropo-
lyacids with Wells-Dawson structure, included in silica for the direct
esterification of levulinic acid, in heterogeneous conditions, with excess
of ethanol. It must be noted that, if the WD active phase is deposited by
impregnation, there will be always leaching during the reaction in polar
solvent like ethanol due to WD is very soluble, and simultaneously there
will be heterogeneous and homogeneous catalysis. Table 2 shows the
obtained results for the yield% of EL using different bulk and silica-
included catalysts. The catalytic tests were carried out in a stirred
batch reactor at 78 °C, as has been described in the experimental sec-
tion. The reaction yield was expressed as the ratio of % moles of product
to moles of initial substrate. In our reaction conditions, EL was obtained
as an only product (selectivity of 100%). Secondary products such as
angelica lactone, 4-etoxy γ-valerolactone and/or aldehydes were no
detected. Blank experiments, without catalyst and silica sol–gel (SiO2)
as catalysts, were performed under similar reaction conditions
(Table 2, entries 1 and 2, respectively). No conversion of LA for both
samples was detected. Besides, the esterification reaction of LA using
two bulk catalysts with different structure, Wells-Dawson (WD) and
Keggin (K) HPA, (Table 1, entries 3 and 5), were performed. In these
conditions, both catalysts are very soluble in the reaction medium
(homogeneous catalysis). The obtained yields were 93% and 92% for
bulk WD and K HPA, respectively.
31P MAS-NMR
On the other hand, the yields (%) of EL using K and silica-
incorporated WD heteropolyacids are also listed in Table 2, entries 4
and 6 to 10, respectively. It can be observed that for both types of
Table 2
Catalytic performance of different heteropolyacids in levulinic acid esterification at
78 °C.
Entry
Catalyst type
Catalyst
EL yield (%)
1
2
3
4
5
6
7
8
9
None
Support
KBulk
–
SiO2
Keggin HPA
–
a
92
38
93
76
–
b
40K-S
WDBulk
40WD-S
40WD-Sa
40WD-Sb
40WD-Sc
40WD-Sd
-5
-8
-11
-14
-17
-20
Wells-Dawson HPA
72
70
79
10
ppm
δ,
Experimental conditions: 2 mmol of LA, 7.5 mL (128 mmol) of absolute ethanol (ratio
ethanol/LA: 64/1); reaction time, 10 h, 78 °C. a,b,c,dUsing 20, 100, 500 and 1000 mg of
catalyst.
Fig. 4. 31P MAS-NMR spectra of silica-included Wells-Dawson HPA. a) Fresh 40WD-S
and b) the same sample after three reaction cycles.

G. Pasquale et al. / Catalysis Communications 18 (2012) 115–120
119
80
70
60
50
40
30
20
10
0
prewashed 40WD-S sample was put in contact with ethanol for
48 h at reflux, under stirring, without LA. After this time, the catalyst
was separated from ethanol by centrifugation, and LA was added to
the reaction mixture. The reaction was carried out for 10 h, and
ethyl levulinate was not detected. This behavior indicates that the
reaction only occurs heterogeneously.
78oC
60oC
40oC
3.2.3. Influence of the different operative conditions on the esterification
reaction
In order to find the best reaction conditions for the esterification
reaction, several experiences using different molar ratios of absolute
ethanol to LA, reaction times, temperatures and different loadings of
WD included in silica were done. For these tests, the ethanol-
prewashed 40WD-S samples were used.
0
2
4
6
8
10
Time (h)
Fig. 5. Yield to ethyl levulinate (EL) vs. time, at different reaction temperatures:
a) 40 °C; b) 60 °C and c) 78 °C.
The molar ratio of ethanol to LA was varied from 64 to 3:1, and the
results showed that a ratio of 3 is enough to obtain similar yields of
ethyl levulinate at 78 °C and 10 h, 76% and 78%, respectively.
Fig. 5 shows the yield of EL versus time, at different reaction
temperatures 40, 60 and 78 °C, using the ethanol-prewashed 40WD-
S sample. In this case, a ratio of ethanol/LA of 3 was used, according
to the results mentioned above. An increase of ethyl levulinate is
observed as temperature and time increase.
On the other hand, Fig. 6 shows the effect of a lower loading of WD
incorporated in silica, 40WD-S and 20WD-S samples, as a function of
time, at 78 °C and an ethanol/LA molar ratio of 3. A diminution of the
EL yield for the sample with the lowest WD loading with respect to
the 40WD-S sample can be observed. It must be noted that the
number of acidic sites for the 20WD-S sample is half that of 40WD-
S, according to the titration measurements (see Fig. 3).
Fig. 6. Yield to ethyl levulinate (EL) vs. time, at two different loadings of included
Wells-Dawson acid.
pretreatment the decrease in activity after the first cycle was not ob-
served, and the catalyst kept the activity constant during four cycles.
The reused sample was characterized by 31P MAS-NMR and
compared with the fresh synthesized sample before the catalytic
test, and the HPAs also kept their HPA structure after four reaction
cycles (see Fig. 4a and b, respectively).
3.2.4. Proposal for a possible mechanism
The reaction mechanism for the esterification reaction of levulinic
acid with ethanol could be described as a general Fischer esterification
with several steps: first, an adsorption of levulinic acid on Brönsted
sites on the catalyst surface, forming a protonated levulinic acid inter-
mediate, increases the electrophilicity of carbonyl carbon. This carbonyl
carbon would be attacked by the nucleophilic oxygen atom of ethanol
leading to the formation of an oxonium ion. Then a proton transfer in
In order to verify heterogeneous conditions of the esterification
reaction, additional catalytic experiments were done. The ethanol-
Scheme 2. Proposed mechanism for levulinic acid esterification with ethanol.

120
G. Pasquale et al. / Catalysis Communications 18 (2012) 115–120
[4] A. Corma, S. Iborra, A. Velty, Chemical Reviews 107 (2007) 2411–2502.
[5] F.W. Lichtenthaler, Accounts of Chemical Research 35 (2002) 728–737.
[6] F.W. Lichtenthaler, Carbohydrate Research 313 (1998) 69–89.
[7] B. Timokhim, V. Baransky, G. Eliseeva, Russian Chemical Reviews 68 (1999)
73–84.
the oxonium ion gives a new oxonium ion and finally the loss of water
from the latter oxonium ion, and subsequent deprotonation leads to the
ester, with the acid site on the catalyst surface being regenerated
[33,34] (Scheme 2).
[8] V. Ghorpade, M. Hanna, US Patent 5859263, 1996.
[9] J. Horvat, B. Klaic, B. Metelko, V. Sunjic, Tetrahedron Letters 26 (1985)
2111–2114.
[10] S. Fritzpatrick, WO Patent 8910362, 1989.
4. Conclusions
[11] S. Frtizpatrick, WO Patent 9640609, 1996.
For the first time, Wells-Dawson HPA structures were incorporated
in the silica framework by the sol–gel technique. The synthesis of
catalysts by the sol–gel technique was satisfactory and the samples
kept their HPA structure intact after synthesis. The catalytic tests for
the esterification reaction between levulinic acid and ethanol to
produce ethyl levulinate using these silica-incorporated Wells-
Dawson catalysts were done. It was observed that the esterification of
levulinic acid with ethanol only occurs in the heterogeneous phase. It
must be noted that silica-included HPAs also kept their structure and
catalytic activity after three consecutive reaction cycles. These results
indicated that these solid acids are promissory catalysts for the
esterification reaction of levulinic acid with ethanol to ethyl levulinate
and they would be active and selective catalysts for this reaction.
[12] R.H. Leonard, Industrial and Engineering Chemistry 48 (1956) 1331–1341.
[13] M. Kitano, F. Tanimoto, M. Okabayashi, Chemical Economic Engineering Review 7
(1975) 25–29.
[14] P. Ayoub, WO Patent 2005070867, 2005.
[15] A.R. Bacler, A. Kontowicz, Journal of the American Chemical Society 75 (1953)
5416–5417.
[16] G. Olah, J. Welch, Synthesis 9 (1974) 652.
[17] A. Corma, H. García, Chemical Reviews 103 (2003) 4307–4365.
[18] I. Ledneczki, M. Darányi, F. Fulop, A. Molnár, Catalysis Today 100 (2005) 437–440.
[19] I. Kozhevnikov, Journal of Molecular Catalysis A: Chemical 262 (2007) 86–92.
[20] I. Kozhevnikov, Catalysis Reviews-Science and Engineering 37 (1995) 311–352.
[21] M. Pope, Heteropoly and Isopoly Oxometalates, Springer, Berlin, 1983.
[22] J. Sambeth, G. Romanelli, J.C. Autino, J. Thomas, G. Baronetti, Applied Catalysis A:
General 378 (2010) 114–118.
[23] P. Vázquez, M. Blanco, C. Cáceres, Catalysis Letters 60 (1999) 205–215.
[24] D. Bennardi, G. Romanelli, A. Sathicq, J.C. Autino, G. Baronetti, H. Thomas, Applied
Catalysis A: General 404 (2011) 68–73.
[25] C. Caetano, I. Fonseca, A. Ramos, J. Vital, J. Castanheiro, Catalysis Communications
9 (2008) 1996–1999.
[26] A. Popa, V. Sasca, E. Kiss, R. Marinkovic-Neducin, I. Holclajtner-Antunovic, Materials
Research Bulletin 46 (2011) 19–25.
Acknowledgments
[27] A. Popa, V. Sasca, E. Kiss, R. Marinkovic-Neducin, M. Bokorov, I. Holclajtner-
Antunovic, Materials Chemistry and Physics 119 (2010) 465–470.
[28] G. Baronetti, L. Briand, U. Sedran, H. Thomas, Journal Applied Catalysis A 172
(1998) 265–272.
[29] A. Leea, N. Chaibakhsha, M.B.A. Rahmana, M. Basria, B.A. Tejo, Industrial Crops and
Products 32 (2010) 246–251.
Authors thank Prof. Dr. Horacio Thomas for his helpful discussions.
The research leading to these results has received funding from the
European Union Seventh Framework Programme (FP7/2007-2013)
under Grant Agreement No. 227248. GR, PV and GB are members of
CONICET.
[30] G. Baronetti, H. Thomas, C. Querini, Applied Catalysis A: General 217 (2001)
131–141.
[31] P. Villabrille, G. Romanelli, L. Gassa, P. Vázquez, C. Cáceres, Applied Catalysis A:
General 324 (2007) 69–76.
References
[32] D. Ruiz, G. Romanelli, G. Pasquale, J. Autino, G. Baronettiy, H. Thomas, Actas del
XXII CICAT–Congreso Iberoamericano de Catálisis, Con-Cón (Chile), 2010, p. 256.
[33] A. Khder, E. El-Sharkawy, S. El-Hakam, A. Ahmed, Catalysis Communications 9
(2008) 769–777.
[34] L. Osiglio, G. Romanelli, M. Blanco, Journal of Molecular Catalysis A: Chemical 316
(2010) 52–58.
[1] B. Girisuta,. Theses: Levulinic acid from lignocellulosic biomass. University of
Groningen, 2007, ISBN 978-90-367-3229-1 (electronic version).
[2] B. Kamm, M. Kamm, P.R. Gruber, S. Kromus, in: B. Kamm, P.R. Gruber, M. Kamm
(Eds.), Biorefinery Systems — An Overview, Biorefineries — Industrial Processes and
Products: Status Quo and Future Directions, Volume 1, Wiley-VCH, Weinheim, 2006.
[3] H. Zoebelin, Dictionary of Renewable Resources, Wiley VCH, Weinheim, 2001.