Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostructured silica using DMSO as co-solvent

Article abstract of DOI:10.1016/j.cattod.2016.02.016

The use of sulfonic acid-functionalized heterogeneous catalysts in conjunction with the use of dimethyl sulfoxide (DMSO) as co-solvent in the catalytic transformation of fructose in ethanol to produce 5-ethoxymethyl furfural (EMF) is shown as an interesting alternative route for the production of this advanced biofuel. Arenesulfonic acid-modified SBA-15 mesostructured silica (Ar-SO₃H-SBA-15) has been the most active catalyst, ascribing its higher catalytic performance to the combination of excellent textural properties, acid sites surface concentration and acid strength. Noticeably, DMSO promotes the formation of EMF and HMF, reducing the extent of side reactions. Reaction conditions (temperature, catalyst loading and DMSO concentration) where optimized for Ar-SO₃H-SBA-15 via response surface methodology leading to a maximum EMF yield of 63.4% at 116 °C, 13.5 mol% catalyst loading based on starting fructose and 8.3 vol.% of DMSO in ethanol after 4 h of reaction. Catalyst was reused up to 4 consecutive times, without regeneration treatment, showing a slight gradual decay in activity attributed to the formation of organic deposits on the catalyst's surface.

Full text of DOI:10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Efficient production of 5-ethoxymethylfurfural from fructose by
sulfonic mesostructured silica using DMSO as co-solvent
G. Morales∗, M. Paniagua, J.A. Melero, J. Iglesias
Chemical and Environmental Engineering Group, ESCET, Universidad Rey Juan Carlos, C/Tulipán s/n, Móstoles, E28933 Madrid, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The use of sulfonic acid-functionalized heterogeneous catalysts in conjunction with the use of dimethyl
sulfoxide (DMSO) as co-solvent in the catalytic transformation of fructose in ethanol to produce 5-
ethoxymethyl furfural (EMF) is shown as an interesting alternative route for the production of this
advanced biofuel. Arenesulfonic acid-modified SBA-15 mesostructured silica (Ar-SO3H-SBA-15) has been
the most active catalyst, ascribing its higher catalytic performance to the combination of excellent textural
properties, acid sites surface concentration and acid strength. Noticeably, DMSO promotes the formation
of EMF and HMF, reducing the extent of side reactions. Reaction conditions (temperature, catalyst load-
ing and DMSO concentration) where optimized for Ar-SO3H-SBA-15 via response surface methodology
leading to a maximum EMF yield of 63.4% at 116 ◦C, 13.5 mol% catalyst loading based on starting fructose
and 8.3 vol.% of DMSO in ethanol after 4 h of reaction. Catalyst was reused up to 4 consecutive times,
without regeneration treatment, showing a slight gradual decay in activity attributed to the formation
of organic deposits on the catalyst’s surface.
Received 13 December 2015
Received in revised form 25 February 2016
Accepted 26 February 2016
Available online xxx
Keywords:
Fructose
5-Ethoxymethyl furfural
Dimethyl sulfoxide
Response surface methodology
Sulfonic acid
Mesoporous silica
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
genated compounds, which can be used as blend components for
the reformulation of conventional fuels (gasoline and diesel), in
Diminishing fossil fuel reserves along with the need for greening
of land transport, aiming to reduce air pollution, are key concerns
for the present century. Such environmental objectives are driv-
ing society towards the search for new renewable energy sources
for transport that may advantageously substitute fossil sources.
Lignocellulosic biomass is abundant, and it has the potential to sig-
nificantly displace petroleum in the production of not only fuels but
also valuable chemicals, especially through the transformation of
sugars coming from the hydrolysis of cellulose and hemicellulose
[1]. Limitations of conventional biofuels (first generation biodiesel
and bioethanol) and new trends in legislation have stimulated the
search for new technologies providing energy-dense (i.e., lower
oxygen content) biomass-derived fuels. Such advanced biofuels
could be easily implemented in the existing hydrocarbon-based
transportation infrastructure (e.g. engines, fuelling stations, distri-
bution networks and petrochemical processes) and, importantly,
not depending on edible biomass for their production.
some cases even improving certain fuel properties [2]. In this con-
text, 5-ethoxymethylfurfural (EMF), the main representative of the
5-alkoxymethylfurfural ethers family, is considered as an excel-
lent component for diesel-range fuels. It has relatively high energy
density, similar to regular gasoline and nearly as good as diesel
fuel, and significantly higher than the bioethanol, currently the
most extended biofuel [3]. EMF has been evaluated admixed with
commercial diesel in engine tests, leading to promising results in
terms of engine performance, accompanied by a significant reduc-
tion of soot (fine particulates), and SOx emissions [4]. Consequently,
there is currently a remarkable interest in the synthesis of EMF
from renewable resources such as cellulosic materials. Most of
the studies in literature have been focused on the synthesis of
EMF from 5-hydroxymethylfurfural (HMF). HMF is the dehydra-
tion product of biomass hexoses (mainly glucose and fructose) and
can be easily etherified with ethanol by means of acid catalysis,
advantageously using solid acid catalysts [5]. While relative high
EMF yields can be obtained using this approach, the direct use of
HMF as a precursor for the preparation of EMF is not still industri-
ally interesting. The reason is that HMF is a very reactive molecule,
difficult to isolate in good yields. Another proposed strategy is via
5-chloromethylfurfural by substitution with an alcohol. This pro-
posal seeks to replace the OH group within the HMF molecule by
Among the wide range of possibilities, an interesting approach is
the transformation of lignocellulosic platform molecules into oxy-
∗
Corresponding author.
E-mail address: gabriel.morales@urjc.es (G. Morales).
http://dx.doi.org/10.1016/j.cattod.2016.02.016
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
2
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
a chlorine atom, leading to 5-(chloromethyl)furfural (CMF) [6,7].
Thereafter, the nucleophilic substitution of the Cl with ethanol
easily leads to the formation of EMF, giving HCl as by-product.
Although high EMF yields are achievable, the presence of HCl may
generate serious problems in the industrial processing. Likewise,
SBA-15-supported sulphated zirconia, a bifunctional catalyst show-
ing both Lewis and Brønsted acid sites, has also been successfully
applied to HMF etherification with ethanol, demonstrating the exis-
tence of a quantitative relationship between the concentrations of
each type of acid sites within the catalyst [8].
the production of EMF from fructose in ethanol using a commer-
cial homogeneous heteropolyacid HPW catalyst [24]. On the other
hand, several authors have brought the attention on the main draw-
back of using DMSO in HMF production, which is the difficulty of
the separation of both chemicals by conventional processes such
as distillation, due to the high boiling point of HMF and its sensi-
tivity to high temperatures. However, a recent work on the room
temperature separation of HMF from DMSO by selective adsorption
on porous activated carbons, have provided a cost-effective recov-
ery process [25], avoiding the disadvantages of high temperature
separation.
A more appealing methodology is the one-pot combination of
the dehydration of cheap and renewable source such as fructose
into HMF, together with the etherification into EMF using ethanol as
solvent and a heterogeneous acid catalyst. Both transformations are
driven by acid catalysis, being feasible to optimize the selectivity
towards EMF through the proper selection of the acid catalyst and
the reaction conditions. For instance, a conventional mineral acid
such as sulfuric acid has been used as catalyst in this one-pot sys-
tem [9]. In this work, a mechanistic study of the reaction pathway
indicated that the dehydration of EMF into EL is the slowest trans-
formation, not being the sole pathway responsible for EL formation.
On the other hand, the implementation of solid acid catalysts would
have several advantages over mineral acids, especially in terms of
selectivity and management of the transformation. Therefore, the
use of selective solid catalysts is of great interest in this transfor-
mation. In a pioneering work, Brown and co-workers evaluated
the preparation of ethers of HMF, together with HMF itself and
alkyl levulinates, from fructose using ion-exchange resins in non-
aqueous solvents [10]. However, both the selectivity to EMF and
the reaction rates were low. In a similar way, more recently, a cata-
lyst based on silica-sulfuric acid provided 70% EMF yield at 110 ◦C,
but still requiring excessively long reaction times [11]. In another
example, Liu et al. proposed the use of a heteropolyacid-based
organic-inorganic hybrid catalyst, [MIMBS]3PW12O40, leading to
high EMF yields at moderate temperatures [12]. Kraus et al. used
recyclable sulfonic acid-functionalized ionic liquids (ILs), providing
a biphasic system that was shown as a key factor to significantly
enhance both yield and selectivity, avoiding interferences from
humins-type by-products [13]. As an indication of the current rele-
vance of the investigation on EMF production from fructose, during
the last year several other heterogeneous catalytic systems have
been reported: magnetic sulfonic nanoparticles (Fe3O4@C-SO3H)
[14]; acid-base bifunctional hybrid nanospheres prepared from the
self-assembly of basic amino acids and phosphotungstic acid (HPA)
[15]; H-USY, dealuminated H-beta zeolite, Amberlyst-15, SO3H-
SBA-15 [16].
In this context, sulfonic acid-functionalized mesostructured
materials have demonstrated an excellent behaviour in the
catalytic transformation of biomass-derived compounds into
added-value products, such as biodiesel from non-conventional
feedstocks [26], glycerol derivatives from crude glycerin [27,28],
HMF from glucose [22], levulinates from levulinic acid [29], etc.
These materials, featured by high surface area, large uniform pores,
high thermal stability, and the capability to control the surface
hydrophilic/hydrophobic balance as well as the strength and con-
centration of acid sites, appear as promising catalysts for this sort of
acid-catalyzed reactions. More broadly, solid acids with SO3H acid
sites and tunable surface properties appear to have a large potential
in the valorization of biomass [30].
In this contribution, we have studied the catalytic performance
of several sulfonic-containing heterogeneous acid catalysts in the
conversion of fructose to EMF, investigating the effect of using
DMSO as co-solvent, followed by a multivariate analysis to assess
the optimal reaction conditions – catalyst loading, temperature and
DMSO content – to maximize the production of EMF over these
catalysts.
2. Experimental
2.1. Materials
Fructose (99% purity), 5-(hydroxymethyl) furfural (HMF, 99%
purity), 5-ethoxymethylfurfural (EMF, 97% purity) and ethyl levuli-
nate (EL, 99% purity) were purchased from Sigma-Aldrich. Ethanol
(99.9% purity) and dimethyl sulfoxide (DMSO, 99.8% purity) were
obtained from Scharlab. Decane (99% purity) was acquired from
Across Organics. All the chemicals were used as received without
previous purification.
2.2. Catalysts
On the other hand, during the production of EMF from fruc-
tose in ethanol using solid acid catalysts it is particularly important
to prevent formed HMF from rehydrating to yield levulinic acid,
reaction favoured in the presence of strong acid catalysts even in
the presence of very small amounts of water. Such consumption of
HMF limits the formation of EMF, while increases the production
of ethyl levulinate (EL) from the esterification of levulinic acid in
ethanol medium (Scheme 1). Thus, though EL might also be con-
sidered a target fuel additive [17], if the desired product is EMF,
the rehydration must be prevented. In this sense, many authors
have previously reported on the use of an aprotic organic solvent
such as dimethylsulfoxide (DMSO) in the dehydration of hexoses to
5-hydroxymethylfurfural [18–22]. DMSO can stabilize the formed
HMF, significantly reducing undesired side-reactions leading to the
formation of humins, as well as levulinic acid. Furthermore, DMSO
can play an active role in the reaction, since at high temperature
it has the effect of modifying the tautomeric forms of fructose,
increasing the presence of furanose versus pyranose forms, making
easier the dehydration into HMF, precursor to EMF [23]. A recent
work by Wang et al. demonstrated the benefits of using DMSO in
Several sulfonic acid-containing heterogeneous catalysts have
been evaluated in the dehydration of fructose. Propylsulfonic
acid and arenesulfonic acid functionalized mesostructured sil-
icas (Pr-SO3H-SBA-15 and Ar-SO3H-SBA-15, respectively) were
synthesized following previously reported procedures [31,32]. As
reference catalysts, commercial acid catalysts were also evaluated
in this work. Acidic macroporous resin, Amberlyst-15, and a homo-
geneous catalyst, p-toluenesulfonic acid (PTSA), were supplied by
Sigma-Aldrich.
2.3. Catalysts characterization
The textural properties of the sulfonic acid-modified mesostruc-
tured silicas were obtained by means of nitrogen adsorption-
desorption isotherms recorded at 77 K using a Micromeritics
TRISTAR 3000 system. Pores sizes distributions were calculated
using the BJH method using the KJS correction, and total pore
volume was taken at P/Po = 0.975. Structural characterization was
performed by X-ray powder diffraction (XRD) patterns, which were
acquired on a PHILIPS X‘PERT diffractometer using the Cu K␣ line.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
3
OH
OH
O
Et
O
O
O
O
HO
HO
OH
Et
O
EtOH
H3C
+
O
Ethyl Levulinate (EL)
5 Ethoxymethylfurfural (EMF)
Fructose
Scheme 1. Main products of the acid-catalyzed transformation of fructose in ethanol.
Data were recorded from 0.6 to 5◦ (2␪) with a resolution of 0.02◦.
Acid capacity was measured by the determination of cationic-
exchange capacity using sodium as the cationic-exchange agent.
Thermogravimetric analyses (TGA) were performed in a SDT 2960
Simultaneous DSC-TGA, from TA Instruments with an air flow rate
of 100 mL/min and a heating ramp of 5 ◦C/min. Table 1 summarizes
the most relevant physicochemical properties recorded for the acid
catalysts tested in this work.
Product quantification was based on a previous calibration of the
analysis unit with standard stock solutions of pure commercially
available chemicals.
Reacted mol of fructose
Initial mol of fructose
Xfructose
=
× 100
Formed mol of i
Initial mol of fructose
Yi =
× 100
2.4. Catalytic tests
3. Results and discussion
Reaction runs were performed in ACE pressure glass reactors
immersed in an oil bath under strict temperature control. After
a specific reaction time, tube reactor was removed from the oil
bath and rapidly cooled down to room temperature. For catalysts
screening, fructose dehydrations were performed in ethanol up to
8 h at 110 ◦C, adding the appropriate amount of each catalyst to
achieve a constant acid sites loading of 0.05 mmol H+. For the opti-
mization of reaction conditions, the reaction variables investigated
through a multivariate analysis were: catalyst loading (5–15% of
acid sites based on fructose, mol/mol), temperature (90–130 ◦C),
and DMSO loading (5–10 vol.%). Typically, the composition of the
reaction mixture was 0.18 g of fructose (1 mmol), 0.05 g of decane
as internal standard and 5 mL of solvent (mixture of ethanol and
DMSO), and the respective mass of catalyst. In the optimization
study reaction time was fixed at 4 h. Thereafter, under the opti-
mized reaction conditions, a kinetic study was carried out in the
range 0–24 h.
3.1. Catalysts screening
Fructose was converted into EMF over different sulfonic acid-
based catalysts in ethanol in order to identify the catalyst offering
the optimal performance for this transformation. Catalysts assayed
in this work have included propyl- and arene-sulfonic acid-
modified SBA-15 mesoporous silicas, which have been shown
previously as very active catalysts in other acid-catalyzed reactions
[33]. These synthesized catalysts were benchmarked with a sulfonic
macroporous resin, commercially available and conventionally
used in many acid-catalyzed processes, such as Amberlyst-15;
as well as a homogeneous sulfonic acid, p-toluenesulfonic acid
(PTSA). The nature of the active sulfonic acid sites is similar for
PTSA, Amberlyst-15 and Ar-SO3H-SBA-15, all of them presenting
an aromatic ring directly attached to the SO3H group. The electron-
withdrawing effect introduced by said aromatic ring acts as an acid
strength enhancer, resulting in stronger sulfonic sites [32]. On the
other hand, the alkyl nature of the tethering moiety in the propyl-
sulfonic acid-modified SBA-15 silica (Pr-SO3H-SBA-15) results in a
lower acid strength. A blank reaction experiment, performed in the
absence of catalyst, was also performed, showing negligible pro-
duction of EMF. Fig. 1 depicts the results of the catalyst screening
carried out under arbitrarily selected reaction conditions, based on
preliminary experiments and literature. For a better comparison,
catalysts mass loadings were adjusted as a function of their acid
capacities (Table 1), so that a constant acid sites molar loading was
achieved (0.05 mmol H+).
2.5. Product analysis
Reaction samples were analysed combining gas and liquid chro-
matography, using a Varian 3900 gas chromatograph and a Varian
ProStar HPLC chromatograph, respectively. Fructose was quantified
by HPLC using a Hi-Plex H+ column and a refractive index detector
(Varian 356-LC). 0.005 M aqueous sulfuric acid was used as eluent,
using a flow rate of 0.6 mL/min keeping the column temperature
at 60 ◦C. Reaction products detected by GC included HMF, EMF and
EL. In the experiments without DMSO, GC analysis was performed
using a ZB-WAX Plus column and a FID detector. In the experiments
with DMSO, since EL and DMSO eluted together at the same reten-
tion time, a second GC column was required. In this case, a VARIAN
CP-8944 column and a FID detector was used to quantify EL. Cat-
alytic results are shown either in terms of absolute conversion of
fructose or in terms of yields towards the different products (Yi).
As shown in Fig. 1, after 8 h at 110 ◦C all the catalysts provided
100% fructose conversion, with the exception of Amberlyst-15.
At this temperature, the reaction can be considered fast, espe-
cially with the homogeneous PTSA catalyst, rapidly leading to the
transformation of fructose into HMF as an intermediate that subse-
quently evolves to EMF and EL. The appearance of ethyl levulinate
becomes more significant at higher reaction times. It must be
Table 1
Physicochemical properties of SO3H-based catalysts.
Catalyst
BET area (m2/g)
Pore volume (cm3/g)
Pore Size(Å)
Acid capacity(meq H+/g)
SO3H surface concentration(␮eqH+/m2)
T limit(◦C)
Amberlyst−15a
PTSA
Pr-SO3H-SBA-15
Ar-SO3H-SBA-15
45
–
721
712
0.40
–
1.18
0.97
300
–
81
92
4.70
5.80
1.03
0.92
104
–
1.43
1.29
120
–
>200
>200
a
Properties provided by the suppliers or experimentally determined.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
4
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
PTSA
Amberlyst-15
50
50
40
30
20
10
0
100
100
40
80
60
40
20
0
80
XFructose
YHMF
YEMF
YEL
XFructose
60
40
20
0
30
20
10
0
YHMF
YEMF
YEL
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
t (h)
t (h)
Pr-SO3H-SBA-15
Ar-SO3H-SBA-15
50
40
30
20
10
0
50
40
30
20
10
0
100
80
60
40
20
0
100
80
60
40
20
0
XFructose
YHMF
YEMF
YEL
XFructose
YHMF
YEMF
YEL
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
t (h)
t (h)
Fig. 1. Fructose conversion (left axes) and yields to HMF, EMF and EL (right axes) in the transformation of fructose to EMF in ethanol using different sulfonic acid catalyst.
Reaction conditions: 1 mmol fructose, 0.05 mmol H+ catalyst, 5 mL ethanol, 110 ◦C.
noted that, in order to close reaction mass balance, there must
be other unidentified products, either degradation by-products
(humin-type) or intermediates such as ethyl-fructoside. This set
of data confirms previously reported main reaction mechanism
including the steps of fructose dehydration into HMF, etherification
of HMF to EMF and rehydration of EMF into EL, the final prod-
uct. As expected, due to its homogeneous nature, PTSA is the most
active catalyst, providing the highest production of EMF (maximum
yield to EMF of 49% at 8 h). However, this high EMF production is
accompanied by a sustained production of EL, reaching over 20%
yield at 8 h. On the other hand, the performance of the three solid
acid catalysts can be correlated to their textural properties as well
as to the availability and strength of their acid sites (Table 1). In
this case, the commercial acid resin Amberlyst-15 has the low-
est specific surface, thus leading to a poor accessibility to the acid
sites. Furthermore, surface acid sites concentration is extraordi-
narily elevated for this catalyst (104 ␮eqH+/m2, Table 1). This is
most likely strongly affecting the local hydrophilicity of the solid
surface [34], which is a key parameter controlling the diffusion of
bulky substrates such as monosaccharides and their derivatives.
Mesostructured catalysts, on the contrary, display high BET surface
areas with a much more moderate surface acid concentration. In
this way, Ar-SO3H-SBA-15 led to a higher EMF yield at 8 h than
Amberlyst-15, 29% vs. 22%, respectively. Additionally, the steps of
fructose dehydration and HMF etherification occur faster over this
catalyst. On the contrary, Pr-SO3H-SBA-15 catalyst displays a sim-
ilar behaviour though reaching lower values of EMF and EL final
yields. Since the support matrix (silica mesostructured), the acid
capacity and the surface concentration are very similar, such a
lower catalytic performance as compared to Ar-SO3H-SBA-15 can
be attributed to the lower inherent acid strength of alkyl- vs. arene-
sulfonic acid sites. Therefore, among the heterogeneous catalysts,
the mesostructured Ar-SO3H-SBA-15 shows the best catalytic per-
formance due to the combination of a suitable acid strength and
excellent textural properties, and it was the selected catalysts for
the following optimization of the reaction.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
5
Fig. 2. Effect of reaction temperature and time in the transformation of fructose in ethanol to give HMF, EMF and EL over Ar-SO3H-SBA-15 catalyst. Reaction conditions:
1 mmol fructose, 0.05 mmol H+, 5 mL ethanol.
3.2. Optimization of reaction conditions with Ar-SO3H-SBA-15
From the above catalysts screening, Ar-SO3H-SBA-15- was
selected as the best solid catalyst to further analyse the reaction
variables. The objective was to maximize EMF yield through the
proper selection of the reaction conditions. Significant variables
were first identified. Fig. 2 includes the results of fructose con-
version and yield to the main products (HMF, EMF and EL) at two
different temperatures, 70◦ and 110 ◦C, after 8 h and 24 h. As shown,
temperature dramatically affects the catalytic performance, espe-
cially in terms of yields to EMF and EL. Products distribution is
clearly different at 70 ◦C, favouring the formation of HMF over the
ether and/or the ester. However, as occurred over Amberlyst-15
(Fig. 1), the conversion of fructose is still high at low tempera-
ture. This indicates different activation energies for the dehydration
of fructose into HMF and the subsequent transformations with
ethanol. In order to maximize EMF temperatures higher than 70 ◦C
must be used. Regarding the effect of reaction time, longer reac-
tion times clearly favour the formation of ethyl levulinate, which
eventually becomes the major product at 110 ◦C. As a reaction inter-
mediate, yield to EMF tends to decrease slowly at longer times.
Therefore, selection of reaction time is also important to ensure
Fig. 3. Effect of the addition of DMSO to the reaction medium. Catalyst, Ar-SO3H-
SBA-15. Reaction conditions: 110 ◦C, 4 h, 1 mmol fructose, 0.05 mmol H+, 5 mL total
reaction volume.
maximum EMF production, while keeping the formation of EL as
low as possible. From Figs. 1 and 2, for the catalyst Ar-SBA-15-SO3H
at 110 ◦C, an adequate reaction time around 4 h can be inferred.
On the other hand, a preliminary set of experiments aiming to
analyse the benefits of using dimethylsulfoxide as co-solvent was
performed (Fig. 3). EMF yield is clearly enhanced even by the pres-
ence of 5 vol.% of DMSO, and increasing the concentration up to
10 vol% keeps enhancing the production of EMF. As shown, the yield
towards HMF is also increased, though in a lesser degree than EMF.
In conclusion, reaction temperature and DMSO loading appear
to be most influential variables in this system. On the other hand,
catalyst loading is usually a key parameter in such catalytic pro-
cesses. Furthermore, because of the presence of side reactions,
crossed interactions between variables can be expected. Therefore,
in order to simultaneously analyse the effect of these three main
reaction variables, i.e. temperature, catalyst loading and DMSO
concentration, an experimental design methodology was proposed
[35]. Particularly, a 33 factorial experimental design, with three dif-
Fig. 4. Standardized Pareto chart for yield to EMF.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
6
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
Table 2
Experiments matrix and experimental results for the transformation of fructose in ethanol-DMSO over arenesulfonic acid functionalized mesostructured silica, Ar-SO3H-
SBA-15. Reaction time 4 h.
Run
T (◦C) (X1)
C (%) (X2)
D (%) (X3)
IT
IC
ID
YHMF (%)
YEMF (%)
1
2
3
4
5
6
7
8
9
90
110
130
90
110
130
90
5
5
5
10
10
10
15
15
15
5
5
5
5
5
5
5
5
5
−1
0
1
−1
−1
−1
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
25.5
13.1
0.7
22.4
4.0
0.3
21.8
1.5
14.2
37.5
36.7
24.3
54.0
40.0
34.6
61.4
30.9
−1
0
1
−1
110
130
0
1
0.4
10
11
12
13
14
15
16
17
18
19
20
21
110
110
110
90
110
130
90
110
130
90
110
130
10
10
10
5
5
5
10
10
10
15
15
15
7.5
7.5
7.5
7.5
7.5
7.5
7.5
7.5
7.5
7.5
7.5
7.5
0
0
0
0
0
0
−1
−1
−1
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
6.5
8.7
8.2
24.8
49.6
1.6
42.6
10.7
0.7
56.2
61.4
57.1
8.5
49.6
54.7
28.3
62.9
57.9
31.4
63.8
49.5
−1
0
1
−1
0
1
−1
29.1
3.4
0.5
0
1
1
22
23
24
25
26
27
28
29
30
90
110
130
90
110
130
90
5
5
5
10
10
10
15
15
15
10
10
10
10
10
10
10
10
10
−1
0
1
−1
−1
−1
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
37.8
24.2
4.3
32.3
11.2
1.3
36.4
7.8
1.1
11.0
42.9
44.2
18.3
50.3
56.8
34.4
53.2
56.9
−1
0
1
−1
110
130
0
1
Note: T, temperature; C, catalyst loading (mol% of acid sites based on fructose); D, DMSO loading (vol.% based on ethanol); I, coded value; YHMF , yield to 5-hydroxymethyl
furfural; YEMF , yield to 5-ethoxymethyl furfural. Columns 2–4 represent the factor levels on a natural scale whereas columns 5–7 represent the 0 and 1 encoded factor levels
on a dimensionless scale.
ferent levels for each of the three factors, was carried out. Selection
of the levels was based on the above previous results (Figs. 1–3).
The lower and upper levels of the experimental factors were:
90–130 ◦C for the temperature, 5–15 mol% of acid sites based on
fructose for the catalyst loading (equivalent to 0.05–0.15 mmol H+
in the reaction medium), and 5–10 vol.% based on ethanol for DMSO
loading. Central point experiment (110 ◦C, 0.10 mmol H+, 7.5 vol.%
DMSO) was repeated three times in order to determine the vari-
ability of the results and assess the experimental error. The selected
responses were the yields towards the most relevant products, 5-
hydroxymethylfurfural (YHMF) and 5-ethoxymethylfurfural (YEMF).
Fructose conversion was practically total for all the experiments,
whereas yield to ethyl levulinate was not determined for every run
due to the complexity of analysis in the presence of DMSO (see Sec-
tion 2). Nevertheless, from the analysed samples, yield to EL showed
low variability within the experimental range. Optimization of the
reaction variables targeted the maximization of the yield to EMF.
The standard experiments matrix for the design is shown in Table 2
. Experiments were run randomly to minimize errors due to pos-
sible systematic trends in the variables. Table 2 also includes the
results of the yields to HMF and EMF.
The analysis of experimental data was performed according to
response surface methodology using a second-order polynomial
equation:
Y = ˇ0 + ꢀ3 ˇi · Xi + ꢀ3 ˇii · Xi2 + ꢀ3i=1ˇij · Xi · Xj
i=1
i=1
where Y is each response (yield to HMF, yield to EMF, mol%) and
␤0, ␤i, ␤ii and ␤ij are the regression coefficients of intercept, linear,
quadratic and binary interactions, respectively. Xi and Xj are the
independent factors, temperature, catalyst loading and DMSO con-
centration. To confirm the parameter estimation and for the sake
of the model fitting, an estimation of the statistical error was per-
formed. Eqs. (1)–(4) were obtained by multiple regression analysis
using the software Statgraphics (Table 3). The statistical model was
obtained from encoded levels giving the real influence of each vari-
able on the process, and the technological model was obtained from
the real values. Consequently, the influence of the variables on the
Table 3
Predictive equations obtained by response surface methodology.
Equation
R2
Statistical models
2
2
2
YHMF = 12.2417 − 14.5444·IT − 4.42222·IC + 3.70556·ID + 3.29167·IT − 0.316667·IT·IC − 2.625·IT·ID + 3.29167·IC − 0.45·IC·ID − 2.95833·ID
(1)
(2)
0.7537
0.9374
YEMF = 58.9896 + 12.3667·IT + 6.48889·IC + 1.91111·ID − 17.4479·IT − 5.41667·IT·IC + 4.975·IT·ID − 3.64792·IC2 + 0.658333·IC·ID − 5.94792·ID
2
2
Technological models
YHMF = 126.772 − 2.11848·T − 2.88208·C + 14.7447·D + 0.00825661·T2 − 0.00317583·T·C − 0.052505·T·D + 0.130919·C2 − 0.03598·C·D − 0.475203·D2
(3)
(4)
0.7537
0.9374
YEMF = −597.705 + 10,0184·T + 9.78487·C + 3.57358·D − 0.0436582·T2 − 0.0541383·T·C + 0.0994733·T·D − 0.146244·C2 + 0.0527467·C·D − 0.951857·D2
Note: IT, coded value for temperature; IC, coded value for catalyst loading; ID, coded value for DMSO loading; YHMF , yield to 5-hydroxymethyfurfural; YEMF, yield to 5-
ethoxymethyl furfural. The models include only the significant terms.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
7
Fig. 5. Response surfaces for yield to EMF and HMF predicted by the models for the three different levels of DMSO loading. Catalyst, Ar-SO3H-SBA-15; time, 4 h.
responses is discussed using the statistical models shown in Eqs.
(1) and (2) (Table 3).
In the case of the yield to HMF (Eq. (1)), statistical analysis within
the studied experimental range identifies the temperature as the
most influential factor with a negative effect. In addition, the effect
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
8
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 6. Response surfaces for yield to EMF and HMF predicted by the models for the three different levels of catalyst loading. Catalyst, Ar-SO3H-SBA-15; time, 4 h.
of catalyst loading has also a negative contribution. Considering
both effects together, the yield to HMF can be favoured under low
temperature and catalyst loading, i.e. under low reaction intensi-
ties. On the other hand, as expected from literature, DMSO displays
a positive effect. An analysis of the significance of each parameter by
means of Pareto chart (not shown) identified the linear contribution
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
9
Fig. 7. Response surfaces for yield to EMF and HMF predicted by the models for the three different temperatures. Catalyst, Ar-SO3H-SBA-15; time, 4 h.
of T as the main factor. The interactions between the three factors
fall in the limit of significance. On the other hand, for the response
of yield to EMF (Eq. (2)), the statistical analysis within the studied
experimental range identifies again the temperature as the most
important factor, but in this case with a positive effect. However,
the second factor in importance is the quadratic effect of temper-
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
10
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
70
(A)
(B)
40
60
50
40
30
20
10
0
30
20
10
0
0
10
20
30
40
0
10
20
30
40
50
60
70
Experimental
Experimental
Fig. 8. Experimental versus predicted values for (A) HMF yield; (B) EMF yield. Catalyst, Ar-SBA-15; Fructose, 1 mmol; reaction time, 4 h.
80
70
60
50
40
30
20
10
0
ature with a negative effect. This indicates that an increase in the
temperature does not produce a constant rise in YEMF, because the
100
XFructose
curvature becomes significant. Catalyst and DMSO effects appear
also as positive for YEMF. Fig. 4 shows the standardized Pareto chart
for this response, evidencing the significance of T and C linear con-
tributions, as well as the quadratic effect of temperature with the
three factors and the quadratic effect of DMSO. Therefore, the three
factors effectively influence on the achieved yield to EMF, showing
curvature of the model in the case of temperature and DMSO. This is
an indication of the necessity of evaluating the factors all together
in order to account for the interactions between them.
In order to facilitate the interpretation of the equations and
visualize the variation of the reaction parameters, Figs. 5–7 plot
the response surfaces predicted by the above technological mod-
els (Eqs. (3) and (4), Table 3) for the yields towards HMF and EMF.
In Fig. 5, the surfaces of temperature-catalyst loading are depicted
at the three different levels of DMSO. Fig. 6 shows the effect of
temperature-DMSO for each of the fixed catalyst loadings, and
Fig. 7 includes the surfaces corresponding to the variation of DMSO-
catalyst loading at each temperature. As can be observed from the
response surfaces, the temperature-catalyst interaction has a sig-
nificant negative effect meaning that the enhancement of EMF yield
with the temperature is less significant at the highest values of
catalyst loading, and the corresponding improvement with the cat-
alyst loading is also more substantial at low values of temperature.
On the other hand, increasing the concentration of DMSO in the
reaction medium clearly improves the yield to EMF, though a max-
imum can be deduced between 7.5 vol.% and 10 vol.%. Again, this
confirms that in this system DMSO promotes the dehydration of
fructose to HMF, subsequently etherified to EMF in the presence
of ethanol. Analysing the surface responses corresponding to HMF,
reverse trends can be observed leading to minimum values of YHMF
when high values of YEMF are obtained.
YHMF
YEMF
YEL
80
60
40
20
0
0
5
10
15
20
25
t (h)
Fig. 9. Effect of reaction time in the transformation of fructose in DMSO over
Ar-SO3H-SBA-15 under the optimized reaction conditions: T = 116 ◦C; catalyst load-
ing = 13.5 mol%; DMSO = 8.3 vol.%.
than experimental ones (left graph). As an additional estimation
of error, the arithmetical averages and the standard deviations
of both responses were calculated from the central point repli-
cas (Table 2, runs 10–12, and 17): HMF yield (8.5 1.7) and EMF
yield (59.4 3.3). Consequently, the obtained standard deviations
are low enough to consider the experimental error, especially for
the target variable yield to EMF, as not very significant.
Applying the mathematical model represented by Eq. (2), the
maximal predicted value for the yield to EMF at 4 h within the
experimental region is 63.4 mol%. This optimal value corresponds
to the following reaction conditions: 116 ◦C, 13.5 mol% of catalyst
loading – based on starting fructose – and 8.3 vol.% of DMSO loading
(IT = +0.29, IC = +0.70, ID = +0.32 in coded values). A catalytic exper-
iment carried out under such reaction conditions (Fig. 9) provided
An evaluation of the regression error was performed on the
models, aiming to validate their use for making predictions. An
indication of the goodness of the fit for YEMF is that the regres-
sion coefficient is over 0.93 (Table 3). The fit for YHMF, however,
has a lower regression coefficient of 0.75. Fig. 8 depicts the corre-
lation between experimental results (Table 2) and the respective
predicted values obtained using the mathematical models. For the
yield to EMF, there is an excellent agreement between experi-
mental and predicted values in the whole range of values (graph
on the right), while for HMF predicted values tend to be higher
the following results at 4 h: 100% XF; 3.3% YHMF; 65% YEMF; 8.9% YEL
,
in fair agreement with the model prediction. The kinetic curves of
conversion and yields at optimized reaction conditions shown in
Fig. 9 evidence that selection of reaction time is critical in order
to achieve maximum yield to EMF. After reaching the maximum,
around 4 h in our system, yield to EMF starts to decrease slowly
insofar as the rehydration of EMF into EL takes place. Conversely,
the yield to EL increases progressively with reaction time, being
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
11
OH OH
O
Fructopyranose
HO
OH
Ethyl fructosides
HO
OH
EtOH
- H2O
O
OH
O
HO
OH
O
HO
- 3 H2O
OH
HO
HO
OH
HO
O
Fructofuranose
O
- 3 H2O
HO
OH
O
O
HO
OH
5-Hydroxymethylfurfural (HMF)
+ 2 H2O
- formic acid
- 3 H2O
EtOH
- H2O
O
O
OH
O
O
O
5 Ethoxymethylfurfural (EMF)
Levulinic Acid (LA)
+ 2 H2O
- formic acid
EtOH
- H2O
O
O
O
Ethyl Levulinate (EL)
Fig. 10. Proposed reaction mechanism for the transformation of fructose in ethanol, modified from the one reported by Flannelly et al. [9].
the slowest transformation. Apparently, DMSO does not seem to
prevent the rehydration of EMF into EL, as it does with the rehy-
dration of HMF into levulinic acid [22]. This is in agreement with the
mechanism proposed by Flannelly et al. [9], where EL was reported
to be produced not solely from EMF, being the most feasible alter-
native pathway the formation of levulinic acid (via HMF) and its
rapid esterification to EL with ethanol, as it has been reported for
this type of catalyst at similar temperature [29]. Hence, the over-
all reaction mechanism would be as shown in Fig. 10, wherein a
contribution to the formation of EL via levulinic acid is also consid-
ered. In such reaction network, the enhancing role of DMSO on the
yield to EMF has two contributions, on the one hand, it stabilizes
the furanose form of the fructose [23] and, on the other hand, it
limits the rehydration of HMF into levulinic acid [22].
tion cycles. Fig. 11 displays the obtained results in terms of yields
to the main products, i.e. HMF, EMF and EL (fructose conversion
is total in every run). Catalyst recovery between consecutive runs
was carried out by simple filtration without any further treatment.
Using such a simple procedure the catalyst can be reused several
times with only a small loss of activity in terms of EMF or EL pro-
duction, accompanied by a small increase in the production of the
precursor HMF. Noteworthy, activity loss is larger after the first
reaction cycle, maintaining a more constant performance in the
following cycles. Spent catalysts were analysed by means of ele-
mental analysis, in order to evaluate the possible incorporation of
organic deposits coming from undesired side reactions (humins) as
well as the possible leaching of sulphur species. Elemental analy-
sis provided almost constant sulphur content (from 2.8 to 2.7 wt%),
even after 4 reaction cycles, together with an increase in carbon
content (from 12.6 to 15.9 wt%). This indicates, on the one hand,
the high stability of sulfonic species, which are not washed out
Additionally, aiming to evaluate the reusability of the Ar-
SO3H-SBA-15 catalyst in this reaction system, recycling tests were
conducted under the optimized reaction conditions for four reac-
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016

G Model
CATTOD-10046; No. of Pages12
ARTICLE IN PRESS
G. Morales et al. / Catalysis Today xxx (2016) xxx–xxx
References
12
100
90
80
70
60
50
40
30
20
10
0
EL
EMF
HMF
[1] S.G. Wettstein, D. Martin Alonso, E. Gürbüz, J.A. Dumesic, Curr. Opin. Chem.
Eng. 1 (2012) 218–224.
[2] M.J. Climent, A. Corma, S. Iborra, Green Chem. 16 (2014) 516–547.
[3] M. Mascal, E.B. Nikitin, Angew. Chem. Int. Ed. 47 (2008) 7924–7926.
[4] G.J.M. Gruter, F. Dautzenberg, U.S. Patent Appl. 2011/0082304 A1 (2011).
[5] P. Lanzafame, D. Temi, S. Perathoner, G. Centi, A. Macario, A. Aloise, G.
Giordano, Catal. Today 175 (1) (2011) 435–441.
[6] M. Mascal, E.B. Nikitin, ChemSusChem 2 (9) (2009) 859–861.
[7] M. Brasholz, K. von Kanel, C.H. Hornung, S. Saubern, J. Tsanaktsidis, Green
Chem. 13 (5) (2011) 1114–1117.
[8] K. Barbera, P. Lanzafame, A. Pistone, S. Millesi, G. Malandrino, A. Gulino, S.
Perathoner, G. Centi, J. Catal. 323 (2015) 19–32.
[9] T. Flannelly, S. Dooley, J.J. Leahy, Energy Fuels 29 (11) (2015) 7554–7565.
[10] D.W. Brown, A.J. Floyd, R.G. Kinsman, Y. Roshan-Ali, J. Chem. Biotechnol. 32
(7–12) (1982) 920–924.
[11] M. Balakrishnan, E.R. Sacia, A.T. Bell, Green Chem. 14 (6) (2012) 1626–1634.
[12] B. Liu, Z. Zhang, K. Deng, Ind. Eng. Chem. Res. 51 (47) (2012) 15331–15336.
[13] G.A. Kraus, T. Guney, Green Chem. 14 (6) (2012) 1593–1596.
[14] Z. Yuan, Z. Zhang, J. Zheng, J. Lin, Fuel 150 (2015) 236–242.
[15] H. Li, K.S. Govind, R. Kotni, S. Shunmugavel, A. Riisager, S. Yang, Energy
Convers. Manage. 88 (2014) 1245–1251.
[16] H. Li, S. Saravanamurugan, S. Yang, A. Riisager, Green Chem. (2015), http://dx.
doi.org/10.1039/C5GC01043H (Advance Article).
[17] B.C. Windom, T.M. Lovestead, M. Mascal, E.B. Nikitin, T.J. Bruno, Energy Fuels
25 (2011) 1878–1890.
[18] Y. Roman-Leshkov, J.N. Chheda, J.A. Dumesic, Science 312 (2006) 1933–1937.
[19] J.N. Chheda, Y. Roman-Leshkov, J.A. Dumesic, Green Chem. 9 (2007) 342–350.
[20] R.M. Musau, R.M. Munavu, Biomass 13 (1987) 67–74.
[21] Y.Y. Lee, K.C.W. Wu, Phys. Chem. Chem. Phys. 14 (2012) 13914–13917.
[22] G. Morales, J.A. Melero, M. Paniagua, J. Iglesias, B. Hernández, M. Sanz, Chin. J.
Catal. 35 (2014) 696–707.
Fig. 11. Reusability of catalyst Ar-SO3H-SBA-15 on the transformation of fructose
into EMF in DMSO under the optimized conditions: T = 116 ◦C; catalyst load-
ing = 13.5 mol%; DMSO = 8.3 vol.%; t = 4 h.
[23] A.S. Amarasekara, L.D. Williams, C.C. Ebede, Carbohydr. Res. 343 (18) (2008)
3021–3024.
[24] H. Wang, T. Deng, Y. Wang, Y. Qi, X. Hou, Y. Zhu, Bioresour. Technol. 136
(2013) 394–400.
[25] N. Rajabbeigi, R. Ranjan, M. Tsapatsis, Microporous Mesoporous Mater. 158
(2012) 253–256.
(leached) during the reaction, and, on the other hand, the most
probable cause of deactivation, which seems to be the deposition
of organic carbonaceous by-products on the catalyst’s surface. Such
deposits would come primarily from the formation of humins.
[26] G. Morales, L.F. Bautista, J.A. Melero, J. Iglesias, R. Sánchez-Vázquez, Bioresour.
Technol. 102 (2011) 9571–9578.
4. Conclusions
[27] J.A. Melero, R. van Grieken, G. Morales, M. Paniagua, Energy Fuel 21 (2007)
1782–1791.
[28] J.A. Melero, G. Vicente, G. Morales, M. Paniagua, J.M. Moreno, R. Roldan, A.
Ezquerro, C. Perez, Appl. Catal. A: Gen. 346 (2008) 44–51.
[29] J.A. Melero, G. Morales, J. Iglesias, M. Paniagua, B. Hernández, S. Penedo, Appl.
Catal. A: Gen. 466 (2013) 116–122.
[30] P.A. Russo, M.M. Antunes, P. Neves, P.V. Wiper, E. Fazio, F. Neri, F. Barreca, L.
Mafra, M. Pillinger, N. Pinna, A.A. Valente, J. Mater. Chem. A 2 (2014)
11813–11814.
[31] D. Margolese, J.A. Melero, S.C. Christiansen, B.F. Chmelka, G.D. Stucky, Chem.
Mater. 12 (2000) 2448–2459.
[32] J.A. Melero, G.D. Stucky, R. van Grieken, G. Morales, J. Mater. Chem. 12 (2002)
1664–1670.
[33] J.A. Melero, R. van Grieken, G. Morales, Chem. Rev. 106 (2006) 3790–3812.
[34] G. Morales, G. Athens, B.F. Chmelka, R. van Grieken, J.A. Melero, J. Catal. 254
(2008) 205–217.
The use of Brønsted sulfonic acid heterogeneous catalysts allows
the catalytic transformation of fructose in ethanol to give EMF and
EL in high yields. Arene-sulfonic acid-modified mesoporous SBA-
15 silica catalyst provides the highest yield to EMF due to the
combination of excellent textural properties, acid sites surface con-
centration and acid strength. The use of DMSO as co-solvent in
small quantities (up to 10 vol.%) clearly enhances the production
of HMF and EMF. A factorial experimental design allowed for the
optimization of the reaction conditions over Ar-SO3H-SBA-15 cat-
alyst, with the model predicting a maximum EMF yield of 63.4% at
116 ◦C, 13.5 mol% of catalyst loading and 8.3 vol.% of DMSO in 4 h
of reaction time. A study of catalyst’s reuse, without regeneration
treatment, showed a slight gradual decay in activity attributed to
the formation of organic deposits on the catalyst’s surface.
[35] G.E.P. Box, W.G. Hunter, J.S. Hunter, Statistics for Experiments, an Introduction
to Design, Data Analysis and Model Building, Wiley, New York, 1978.
Acknowledgements
Financial support from the Spanish Ministry of Economy and
Competitiveness through the projects CTQ2011-28216-C02-01 and
CTQ2014-52907-R, as well as from the Regional Government
of Madrid through the project S2013-MAE-2882 is gratefully
acknowledged. Likewise, we want to show our gratitude to grade
students Marta Campos and Hellen Esparragoza for their collabo-
ration in experimental tasks.
Please cite this article in press as: G. Morales, et al., Efficient production of 5-ethoxymethylfurfural from fructose by sulfonic mesostruc-
tured silica using DMSO as co-solvent, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.016