Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phosphotungstic acid

Article abstract of DOI:10.1016/j.cattod.2016.02.018

The dehydration of fructose to 5-hydroxymethylfurfural (HMF) is a way to produce a very versatile intermediate of the green chemistry. Heteropolyacid compounds are acid catalysts with singular properties (strong acidity and well-defined structure). This paper aims to analyze the activity of phosphotungstic acid (HPW) and Cs-exchanged phosphotungstic acid (HCsPW) in HMF synthesis, as well as HPW supported on MCM-41. The reactions were conducted in a batch reactor, using dimethyl sulfoxide (DMSO) as solvent, for 2 h. The initial tests indicated that HMF is formed in the absence of catalyst, but in presence of the catalysts, the maximum yield to HMF is higher and achieved in less time. The HPW showed promising results, reaching 100% of fructose conversion and 92% of HMF yield, after 30 min of reaction at 120 °C, with a catalyst:fructose mass ratio of 1:10. However, this heteropolyacid is soluble in DMSO. Besides, the Cs-exchanged HPW resulted in a reduction in HMF yield, because of the reduction in solid acidity. HPW supported in MCM-41 (HPW/MCM) showed to be a good solid acid catalyst for the production of HMF from fructose; moreover, it can be recovered and reused. The variables evaluated for this catalyst were the reaction temperature (100, 120 and 140 °C) and the catalyst:fructose mass ratio (1:10; 1:30 and 1:50). The best results obtained for HPW/MCM were 100% of fructose conversion and 80% of HMF yield in 60 min, at 120 °C.

Full text of DOI:10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by
phosphotungstic acid
Filipe N.D.C. Gomesa, Fabiana M.T. Mendesb, Mariana M.V.M. Souzaa,∗
a Escola de Química—Universidade Federal do Rio de Janeiro (UFRJ), Centro de Tecnologia, Bloco E, sala 206, CEP 21941-909 Rio de Janeiro/RJ, Brazil
b Instituto Nacional de Tecnologia, Divisão de Catálise e Processos Químicos, Av. Venezuela, 82, sala 110, CEP 20081-312 Rio de Janeiro/RJ, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
The dehydration of fructose to 5-hydroxymethylfurfural (HMF) is a way to produce a very versatile inter-
mediate of the green chemistry. Heteropolyacid compounds are acid catalysts with singular properties
(strong acidity and well-defined structure). This paper aims to analyze the activity of phosphotungstic
acid (HPW) and Cs-exchanged phosphotungstic acid (HCsPW) in HMF synthesis, as well as HPW sup-
ported on MCM-41. The reactions were conducted in a batch reactor, using dimethyl sulfoxide (DMSO)
as solvent, for 2 h. The initial tests indicated that HMF is formed in the absence of catalyst, but in presence
of the catalysts, the maximum yield to HMF is higher and achieved in less time. The HPW showed promis-
ing results, reaching 100% of fructose conversion and 92% of HMF yield, after 30 min of reaction at 120 ◦C,
with a catalyst:fructose mass ratio of 1:10. However, this heteropolyacid is soluble in DMSO. Besides,
the Cs-exchanged HPW resulted in a reduction in HMF yield, because of the reduction in solid acidity.
HPW supported in MCM-41 (HPW/MCM) showed to be a good solid acid catalyst for the production of
HMF from fructose; moreover, it can be recovered and reused. The variables evaluated for this catalyst
were the reaction temperature (100, 120 and 140 ◦C) and the catalyst:fructose mass ratio (1:10; 1:30 and
1:50). The best results obtained for HPW/MCM were 100% of fructose conversion and 80% of HMF yield
in 60 min, at 120 ◦C.
Received 1 December 2015
Received in revised form 3 February 2016
Accepted 10 February 2016
Available online xxx
Keywords:
5-Hydroxymethylfurfural
Dehydration
Fructose
Phosphotungstic acid
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
HMF. However, this synthesis is complex due to the high instability
of HMF; depending on the reaction conditions, there occur paral-
The energy generation from biomass has been widely used in
substitution of fossil fuels, such as using ethanol instead of gaso-
line. However, the use of biomass as an alternative to replace oil in
the production of chemical supplies is walking in the slow steps. In
this scenario, it is highlighted the production of furan derivatives.
Such compounds can be obtained from biomass with properties
equivalent to petrochemical derived ones, and thus are consid-
ered as fundamental building blocks for biorefineries. In particular,
5-hydroxymethylfurfural (HMF) is considered a key intermedi-
ate for the development of biomass-based products, because a
series of compounds derives from it, like: organic acids (formic,
levulinic and adipic), polymer precursors (furan 2, 5 dicarboxylic
acid, 2-hydroxymethyl-furan) and biofuels (2-methylfuran, 2,5-
dimethylfuran) [1,2].
lel reactions that lead HMF and fructose to soluble polymers and
humins. In addition, HMF can be converted to levulinic and formic
acids by rehydration resulting in yield loss [3,4].
The literature about dehydration of fructose to HMF is fairly
extensive, with studies of the reaction mechanism, search for opti-
mal synthesis conditions and also the market analysis. However,
no catalytic system with potential for industrial application has
been developed yet. Regarding this issue, there is a range of cata-
lysts, reaction conditions and solvents that have been evaluated and
the results provide a great understanding of how these variables
influence the HMF synthesis.
Researches about reaction temperature and solvent used for
fructose dehydration showed that temperatures from 120 to 180 ◦C
and the water reducing content of the reaction system are ideal con-
ditions to produce HMF. The fructose conversion is very low at low
temperatures; however, at temperatures above 180 ◦C the degra-
dation of HMF is favored. On the other hand, the water content
directly influences the rehydration of the product to levulinic and
formic acids [5,6].
Dehydration of carbohydrates, particularly fructose, in acid
medium, as shown in Fig. 1, is the most promising way to produce
∗
Corresponding author.
E-mail address: mmattos@eq.ufrj.br (M.M.V.M. Souza).
http://dx.doi.org/10.1016/j.cattod.2016.02.018
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
2
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 1. Reaction scheme of fructose dehydration to HMF and parallel reactions.
A lot of system reactions have been studied and the biphasic
and organic media have greater advantages for HMF production
over aqueous medium. The continuous removal of the HMF into
the organic phase provides the displacement of the reaction equi-
librium for production of this compound, as well as prevents its
rehydration and polymerization. Among the solvents used in the
literature, compounds with four carbons (tetrahydrofuran and 1-
butanol, for example) produce higher HMF yields [7–9]. However,
DMSO (dimethyl sulfoxide) showed similar results to those sol-
vents with four carbons [10,11]. Thus, its use as solvent is more
attractive because the fructose is soluble in DMSO, unlike other
organic solvents, reducing parallel reactions and the possibility of
HMF oxidation in the same medium.
(A)
100
95
HCsPW
HPW
90
MCM-41
85
The acid strength of the catalysts (homogeneous or heteroge-
neous) is another variable that governs the HMF selectivity. The
use of strong acids favors deprotonation of the hydroxyl groups
present in the hexoses and therefore the parallel and rehydration
reactions [5]. Thus, it was found that the use of moderate acids
provides increased formation of HMF. Another factor that hinders
the production of this furan derivative is the reaction time. The use
of heterogeneous catalysts requires a longer reaction time for con-
version of fructose and this can cause an increase in polymerization
reactions of both substrates and products, which necessitates the
continuous extraction of HMF from the reaction medium. In addi-
tion, the HMF yield also decreased with the increase of substrate or
catalyst concentration [5–7].
80
100
200
300
400
500
600
Temperature (°C)
(B)
HCsPW
HPW
The heteropolyacids (HPAs) of Keggin type are superacids, with a
well-defined structure, consisting of a central cation (heteroatoms:
P, As, Si, Ge, B etc.) and surrounded by polyanions. Due to their
structural characteristics and heteroatoms, the acidity and other
properties can vary considerably. In addition, the exchange of pro-
tons by other cations, such as Cs+, can cause deep modifications in
their properties [12–14]. Therefore, this class of compounds, which
already has a wide application in catalysis, has a great potential for
use in the HMF synthesis and its derivatives. Although the HPAs
have high acidity, they present weak thermal stability, high solu-
bility in polar solvents and low specific surface area (<10 m2 g−1),
which can limit their use in several reactions. Because of that, HPAs
supported over classical porous materials (e.g., mesoporous silica)
are usually more effective for catalytic reactions [15–17].
MCM-41
100
200
300
400
500
600
Temperature (°C)
Fig. 2. TG/DTA profiles of the catalysts.
There are few reports of the use of HPAs in HMF synthesis.
The studies with pure phosphotungstic acid (HPW) and cesium-
modified HPW – H3PW12O40 and H0.5Cs2.5PW12O40, respectively –
have reported almost total fructose conversions, as well as yields in
the range of 90–95% with DMSO, butanol or methyl isobutyl ketone
(MIBK) as the solvents. The studies report the acid strength of the
catalyst as the principal property that rules the yield and selectivity
to HMF, which is linked to the proton exchange. The exchange by
cesium, for example, enables to modify their acid-base properties
and solubility in polar solvents, consequently changing the HMF
yield [18,19]. HPW supported on MCM-41 has been tested for sev-
eral reactions [16,17], but not for dehydration of fructose. As far as
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
3
7
2.0
(A)
(B)
Adsorption
Desorption
Adsorption
Desorption
6
5
4
3
2
1
0
1.5
1.0
0.5
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure (P/Po)
Relative Pressure (P/Po)
30
25
20
15
10
5
7
6
5
4
3
2
1
0
(D)
(C)
Adsorption
Desorption
Adsorption
Desorption
0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure (P/Po)
Relative Pressure (P/Po)
Fig. 3. N2 adsorption–desorption isotherms of the catalysts: (A) HPW, (B) HCsPW, (C) MCM-41 and (D) HPW/MCM.
we know, the use of HPW supported on MCM-41 for dehydration
of fructose is reported here for the first time.
sponded to 50 wt.% of HPW in the catalyst. The mixture was stirred
at room temperature for 12 h, and the methanol was evaporated
under reduced pressure. The resulting solid was dried at 120 ◦C
overnight and then calcined using the previous conditions.
All researches on hexose dehydration to HMF using HPAs found
in the literature were carried out in organic medium or ionic liquids
without evaluating the role of solvent, the need for proton exchange
and the influence of the support [10,18–20]. Thus, this study aims
to assess the catalytic activity of pure and cesium-exchanged phos-
photungstic acids in HMF synthesis from fructose, with DMSO as
solvent. The activity of the HPW supported on MCM-41 will also be
investigated in fructose dehydration.
Cesium-exchanged
12-tungstophosphoric
acid
(H0.5Cs2.5PW12O40) was synthesized by dropwise addition of
an aqueous solution of cesium carbonate to an aqueous solu-
tion of 12-tungstophosphoric acid, in the desired stoichiometry,
according to the literature [22]. In a typical preparation protocol of
H0.5Cs2.5PW12O40 (HCsPW), 1 mmol of H3PW12O40 was dissolved
in 30 mL of distilled water. This solution was added dropwise
into 1.25 mL, 1 mol L−1 of Cs2CO3 aqueous solution with vigorous
stirring at room temperature. The mixture was continuously
stirred for about 30 min, the water was evaporated at 80 ◦C under
reduced pressure and then the precipitate was aged for 12 h at
room temperature. The white solid was dried at 120 ◦C overnight
and then calcined as described above.
2. Experimental
2.1. Catalyst preparation
Fructose, Keggin-type 12-tungstophosphoric acid (HPW),
cesium carbonate, methanol and dimethyl sulfoxide (DMSO) were
obtained from Sigma-Aldrich Corporation. MCM-41 (mesoporous
aluminosilicate) was purchased from Tianjin Chemist Scientific
Ltd. (Chemist). All reagents were used as supplied without further
purification.
12-tungstophosphoric acid and MCM-41 were firstly calcined
in air for 3 h at 350 ◦C. MCM-41 supported 12-tungstophosphoric
acid (HPW/MCM) was prepared by incipient wetness method [21].
Typically, 1.5 g of HPW was dissolved in 90 mL of methanol. Then
1.5 g of MCM-41 was added into the HPW solution, which corre-
2.2. Catalyst characterization
N2 adsorption-desorption experiment was carried out at
−196 ◦C using an ASAP 2020 equipment (Micromeritics). The sam-
ples were outgassed for 24 h at 300 ◦C. The specific surface area was
obtained by using the Brauner-Emmett-Teller (BET) method, and
pore volume and mean pore diameter by Barrett-Joyner-Halenda
(BJH) method.
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
4
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
Table 1
Physico-chemical properties of the catalysts determined by N2 adsorption/desorption and NH3-TPD measurements.
−1
)
Catalyst
SBET (m2 g−1
)
Pore volume (cm3 g−1
)
Mean pore diameter (Å)
Total acidity (␮molNH3 gcat
HPW
HCsPW
MCM-41
HPW/MCM
<5
0.00
0.17
≥0.70
0.08
0.00
79.4
30.0
87.8
466
52
403
816
135
≥800
114
Fig. 4. SEM micrographs of the catalysts: (A) HPW, (B) HCsPW, (C) MCM-41 and (D) HPW/MCM.
Thermogravimetric (TG) and differential thermal (DTA) anal-
yses were performed under a nitrogen flow at constant rate of
10 ◦C min−1 using a TA Instrument SDT-Q600.
Acid properties of the support and catalysts were determined
by temperature-programmed desorption (TPD) of NH3. Ammonia
adsorption was carried out at room temperature with 4% NH3/He
(30 mL min−1) for 30 min, followed by purging with He for 1 h.
The desorption of chemisorbed ammonia was realized by heat-
ing up to 1000 ◦C at 10 ◦C min−1 under He flow (30 mL min−1). The
desorbed NH3 was detected using a mass spectrometer Pfeiffer
Vacuum PRISMA PLUS.
Fourier transform infrared spectroscopy (FTIR) measurements
were recorded in KBr discs on a Shimadzu PRESTIGE-21 spectrom-
eter with the wavenumber range of 400–1500 cm−1
.
X-ray powder diffraction (XRD) patterns were determined with
a Rigaku Miniflex II diffractometer equipped with a graphite
monochromator and CuK␣ radiation (30 kV and 15 mA). The mea-
surements were carried out with a speed of 0.05◦ s−1 and over the
2␪ range of 2–90◦. XRD in low angle was performed in Rigaku Dmax
2200 diffractometer, with a speed of 0.02◦ s−1 from 1 to 5◦.
Morphology of the samples and the energy dispersive X-ray
analysis (EDS) of W, P and Cs elements were investigated by
means of Scanning Electron Microscope Hitachi TM-1000. Chemical
composition of the HPW/MCM and MCM was determined by X-ray
fluorescence (XRF) using a Rigaku PRIMINI spectrometer.
2.3. Catalytic activity tests
All fructose dehydration reactions were carried out in 100 mL
glass reactor equipped with a reflux condenser. This system was
heated with a thermostatically controlled oil bath. In a typical
procedure, fructose (500 mg—2.77 mmol) was dissolved in DMSO
(20 mL) and then the catalyst (50 mg) was added into the fructose
solution, which corresponded to 10 wt.% of active phase in the fruc-
tose solution. The resulting mixture was charged into a reactor and
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
5
it was placed in the oil bath at 120 ◦C. The samples were collected
from 0 to 120 min. Once defined the catalyst, other reaction con-
ditions were tested, like temperature (100, 120 and 140 ◦C) and
catalyst:substrate mass ratio (1:10, 1:30 and 1:50).
HPW
HCsPW
2.4. Analysis
After each fructose dehydration reaction, the catalyst was
filtered and all products were analyzed by high-performance
liquid chromatography (HPLC) in a Shimadzu Prominence sys-
tem, equipped with diode-array (DAD) and refractive index
(RID) detectors and an Aminex HPX-87H column (Biorad), using
MCM-41
0.01 mol L−1 H2SO4 as the mobile phase at
a flow rate of
1.0 mL min−1. The column temperature was maintained at 40 ◦C.
The fructose conversion (Cf) and the HMF yield (YHMF) are
described in Eqs. (1) and (2), respectively.
HPW/MCM
1200
1400
1000
800
600
400
moles of fructose reacted
moles of fructose initial
Wavenumber (cm-1)
Ct % =
( )
× 100
(1)
(2)
moles of HMF prodced
× 100
moles of fructose initial
Fig. 5. FTIR spectra of the catalysts.
Y
% =
HMF ( )
3. Results and discussion
3.1. Catalyst characterization
Chemical composition of the pure phosphotungstic acid, mea-
sured by energy dispersive X-ray (EDS), is consistent with the
literature. The molar ratio between phosphorus and tungsten was
11 (the theoretical value is 12). Considering that this analysis is
superficial, a small variation in the real value is acceptable. Chemi-
cal composition of the MCM-41 (89 wt.% SiO2 and 5 wt.% Al2O3) and
the mass percentage between HPW/MCM (45% HPW and 55% MCM-
41), measured by X-ray fluorescence (XRF), are in agreement with
the real value and indicate that impregnation of the active phase on
the support by incipient wetness method is a simple method with
good results.
The stability of HPW, HCsPW and MCM-41, before calcination,
was examined by TG/DTA measurements, under N2 atmosphere,
as displayed in Fig. 2. The profile obtained for MCM-41 exhibits
only a weight loss up to 150 ◦C, corresponding to the loss of water
molecules adsorbed on the surface of material [23]. Typical TG/DTA
patterns for the Keggin-type heteropolyacids were presented by
pure phosphotungstic acid. For this catalyst, the major weight loss
occurred before 100 ◦C, which is related to the elimination of water
molecules. These molecules are responsible for the crystallization
of heteropolyanions into a hydrate. The second weight loss for
HPW occurred from 150 to 250 ◦C and corresponds to the loss of
constitutional water. After, there was a plateau between 250 and
400 ◦C, which can be described as the formation of anhydrous acid.
Finally, the third weight loss is reported as decomposition of het-
eropolyanion, with crystallization of oxides [24]. These three events
are endothermic, as seen in DTA profile. As the protons of the
Keggin structure are replaced by cesium atoms, the weight loss is
decreased, as shown by HCsPW profile, indicating that the proton-
exchange improves the thermal stability of phosphotungstic acid
[14,24].
The textural properties of the catalysts were obtained by N2
adsorption-desorption isotherms (Fig. 3). The BET specific surface
area, pore volume and mean pore diameter of all catalysts are
depicted in Table 1. MCM-41 presented a type IV isotherm, typi-
cal of mesoporous materials. The profile of adsorption-desorption
isotherm is completely changed as cesium is inserted in the HPW
structure. The pure acid presented a type II isotherm, characteristic
of non-porous or macroporous solids [25]; however, the HCsPW
presented a type IV isotherm. Moreover, the BET specific sur-
Fig. 6. XRD patterns of the catalysts.
face area of the HCsPW (135 m2 g−1) was much higher than HPW
(<5 m2 g−1).
The decrease in the BET specific surface area and pore volume
of the supported catalyst (HPW/MCM) is significant, about 90% of
reduction when compared to MCM-41, but it is still much higher
than the unsupported compound (Table 1). This reduction is con-
sistent with the literature and suggests that the HPW is deposited
inside the mesochannels and well dispersed on the surface of MCM-
41 [16].
SEM images of solids are illustrated in Fig. 4. The micrographs
enabled to observe a decrease of the particle size with the inser-
tion of Cs in the structure of phosphotungstic acid and indicate that
HPW/MCM is well dispersed in homogeneous particles in compar-
ison with pure MCM-41.
The HPAs of Keggin structure have four different types of oxygen,
which are responsible for characteristic IR bands of heteropoly-
acids of this class: 1060–1080, 960–980, 880–860, 740–790 cm−1
,
attributable to ꢀas (P-Oa), ꢀas (M = Ot), ꢀas (M-Ob-M) and ꢀas (M-
Oc-M) vibrations, respectively, where M is the metal ion (W, Mo,
etc.). The four distinct oxygen sites correspond to the following
description: oxygen of the central tetrahedral PO4 (Oa), terminal
oxygen linked to metal atom alone (Ot), bridge oxygen between
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
100
6
(A)
HPW/MCM
80
60
40
20
0
HCsPW
HPW
Blank
MCM-41
HPW
HCsPW
HPW/MCM
MCM-41
0
20
40
60
80
100
120
100
200
300
400
500
600
700
800
900
1000
Temperature (°C)
Temperature (°C)
Fig. 7. NH3-TPD profiles of the catalysts.
100
(B)
two different M3O13 groups (Ob) and bridge oxygen in the same
M3O13 group (Oc) [19,22,26].
80
60
40
20
0
Fig. 5 shows the FTIR spectra of the catalysts. The FIIR bands
characteristics of the Keggin structure can be clearly seen for HPW
sample. A minor band at 993 cm−1 is observed for the HCsPW. This
split in the ꢀas (M = Ot) band can be attributed to interactions of Cs
atoms in the HPW structure. Li et al. [22] reported that the band at
984 cm−1 is associated with hydrogen bonding between W = Oa and
residual protons (H+), while the band at 993 cm−1 results from the
interaction between W = Oa and counter cations (Cs2+). FTIR spec-
trum of 50 wt.% HPW/MCM indicates that the Keggin structure is
preserved on MCM support. The characteristic peaks of the HPW
and MCM-41 appear without any shift in the position for the sup-
ported catalyst (HPW/MCM-41), although there is an overlapping
of the peaks characteristics of unsupported HPW and MCM-41.
X-ray diffraction patterns of the catalysts are displayed in Fig. 6.
The structure of the calcined HPW is very close to the hydrated
acid and is in agreement with the literature (JCPDS—#7521-25).
For this catalyst, the main XRD lines are at 2␪ = 10.5◦, 18.3◦, 23.7◦,
26.1◦, 30.2◦, 35.6◦ and 38.8◦, which are commonly assigned to cubic
structure. The pattern for the acid modified with cesium (HCsPW)
presented a significant reduction in its intensity, together with a
broader and small shift toward higher angles. The literature reports
these facts and associated with a decrease in the value of the cubic
lattice parameters due to cesium insertion into the secondary struc-
ture [25,27].
Blank
MCM-41
HPW
HCsPW
HPW/MCM
0
20
40
60
Temperature (°C)
80
100
120
Fig. 8. Conversion (A) and HMF yield (B) for fructose dehydration reaction with
different catalysts. Reaction conditions: 500 mg of fructose, 20 mL of DMSO, 50 mg
of catalyst, 120 ◦C.
fication is associated with different sites of ammonia adsorption
in bulk HPAs. The profile of HPW exhibits two small peaks at
180 and 400 ◦C ascribed to the desorption of hydrogen-bonded
NH3 molecules (weak-strength) and the decomposition of ionically
+
bonded NH4 ions with contact with oxygen atom of one Keggin
The XRD pattern of MCM-41 shows a sharp peak at 2ꢁ = 2.2◦
resulted from a strong reflection (see insert of Fig. 6), and two small
peaks at 2ꢁ = 4.0◦ and 4.4◦, which are indexed to (1 0 0), (1 1 0)
and (2 0 0) planes, respectively [15,17]. The XRD pattern of 50 wt.%
HPW/MCM shows no reflections corresponding to HPW. Therefore,
this is a clear evidence of the uniform distribution of HPW on MCM-
41, as described in the literature [16,17]. This result shows that
catalyst synthesis by incipient wetness method with methanol as
solvent is more effective in the distribution of the heteropolyacid on
the support as compared to the use of water as solvent. The aque-
ous impregnation method for higher loadings of HPW (>20 wt.%)
on MCM-41 can cause support hydrolysis or a nonhomogeneous
distribution of anionic species [15].
The acidity of the catalysts was investigated by NH3-TPD, includ-
ing the amount of acid sites and acid strength. Fig. 7 shows
the NH3-TPD profiles of HPW, HCsPW, MCM-41 and HPW/MCM.
According to NH3 desorption temperature, the acid sites can
be classified as weak- (150–300 ◦C), medium- (300–500 ◦C), and
strong- (500–650 ◦C) strength. For the heteropolyacids, this classi-
unit (medium-strength), respectively. Furthermore, the HPW has
an intense peak at 630 ◦C (strong-strength), which is ascribed to the
+
decomposition of ionically bonded NH4 ions in contact with oxy-
gen atom between two neighboring Keggin units [28,29]. Table 1
presents the total acid density of HPW and the other catalysts, cal-
culated from the amount of NH3 desorbed from room temperature
to 800 ◦C, which is provided in a weight-specific value (␮mol g−1).
The Cs substitution in the HPW modified the NH3-TPD profile:
there was a significant reduction of desorption peaks, in particular
of the strong-strength acid site, due to partial replacement of the
protons of phosphotungstic acid by Cs+ ions. The total acid den-
sity was reduced from 466 to 52 ␮mol g−1. The NH3-TPD profile of
the MCM-41 material displays a small and wide peak from 180 to
500 ◦C, which corresponds to weak and medium-strength acid sites
(403 ␮mol g−1). This acid strength is associated with the presence
of only hydrogen bonding due to connection of silanol groups in the
hexagonal structure. The same observation was made by pyridine
gaseous adsorption [15]. After impregnation of HPW in MCM-41 it
is observed an increase in the total amount of acid sites of this new
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
7
100
(A)
100
80
60
40
20
0
(A)
80
60
40
100 oC
20
1:10
1:30
1:50
120 oC
140 oC
0
0
20
40
60
80
100
120
0
20
40
60
80
100
120
Time (min)
Time (min)
100
80
60
40
20
0
(B)
100
80
60
40
20
0
(B)
100 0C
120 0C
140 0C
1:10
1:30
1:50
0
20
40
60
80
100
120
Time (min)
0
20
40
60
80
100
120
Time (min)
Fig. 9. Effect of temperature on conversion (A) and HMF yield (B) for the fruc-
tose dehydration using HPW/MCM catalyst. Reaction conditions: 500 mg of fructose,
20 mL of DMSO, 50 mg of catalyst.
Fig. 10. Effect of catalyst: fructose mass ratio on conversion (A) and HMF yield (B)
for the fructose dehydration using HPW/MCM catalyst. Reaction conditions: 500 mg
of fructose, 20 mL of DMSO, 120 ◦C.
catalyst (816 ␮mol g−1), which is ascribed to the higher amount
of available protons that were not bonded directly to the MCM-
41 framework, for large amounts of HPW supported in MCM-41
[15,25].
ity for the conversion of fructose to HMF: the fructose is totally
converted in 40 min of reaction and the maximum HMF yield is 80%
in 80 min. These results can be due to solvating ability of DMSO
towards cations, that is, the solvent may stabilize the positively
charged intermediates of fructose conversion and thus inhibit their
side reaction to byproducts, such as cited by Dias et al. [25] in the
dehydration of xylose to furfural. This observation can be extended
to explain the slow HMF formation. Some works reported good HMF
yield from fructose in absence of catalyst in different systems, but
with much longer reaction times (at least several hours) than in
this work [10,11].
The use of 10% of HPW resulted in an increased fructose conver-
sion and HMF yield with much shorter time when compared with
the blank test: 100% of fructose conversion in only 10 min and max-
imum yield of 92% in 30 min. Qu et al. [20] reported 98% of fructose
conversion and 97% of HMF yield in 120 min with the same cata-
lyst, temperature and solvent. However, these authors used a very
diluted fructose solution (0.5%), higher ratio of catalyst:fructose
(1:2) and they did not report the results of blank test.
3.2. Catalytic tests of fructose dehydration
The dehydration of fructose to HMF by heteropolyacid catalysts
was evaluated in dimethyl sulfoxide (DMSO) as aprotic solvent.
According to the literature [10,11], DMSO is a good solvent that
could suppress unwanted side reactions in hexose dehydration.
In addition to the formation of HMF, by elimination of three
water molecules from fructose, the main byproducts reported
in hexose dehydration are humins, soluble polymers, furfural,
levulinic and formic acids. These compounds are products of cross-
polymerization, condensation and rehydration reactions.
In DMSO, the furanose conformation of fructose is prevalent,
which favors the formation of HMF because this form is more reac-
tive than other structures of fructose. In addition, DMSO also acts as
catalyst in fructose dehydration. A NMR study on fructose dehydra-
tion in DMSO by Amareseka et al. [30] showed that the mechanism
is very similar to cyclic pathway in aqueous medium.
The exchange of protons by Cs in HPW resulted in a great
decrease of catalytic activity, which may be related to its lower
acidity (Table 1). It should be noticed that HPW forms a homoge-
neous system, while HCsPW is heterogeneous. Qu et al. [20] also
observed a decrease in catalytic activity of Cs3PW12O40 in relation
The results of the blank test (without any catalyst) and catalytic
evaluation of fructose dehydration in DMSO at 120 ◦C are displayed
in Fig. 8. It is noteworthy that DMSO alone has a high catalytic activ-
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
8
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
(141 kJ mol−1) using zeolites or Bicker et al. [34] (99 kJ mol−1), using
sulfuric acid as catalyst. On the other hand, it is higher than that
calculated by Xiao et al. [35] (31.88 kJ mol−1) using HPW in ionic
liquid.
to HPW for fructose dehydration in DMSO. For the same catalysts
and solvent used in this work, Shimizu et al. [3] did not report a
decrease in catalytic activity of HCsPW in relation to HPW, but the
reaction was carried out under evacuation. Evaluating the catalytic
activity of FePW12O40, these authors concluded that the removal
of water by a mild evacuation suppresses two side reactions: the
hydrolysis of HMF to levulinic acid and the reaction of partially
dehydrated intermediates to condensation products.
It is surprising that the conversion and yield of MCM-41 are
lower than those obtained in the blank test, despite the presence of
acid sites. This can be related to the strong adsorption of fructose
on the high surface area of MCM-41 [31], making its conversion to
HMF slower. When HPW is supported on MCM-41, the HMF yield
is lower than that of pure HPW, but considering the advantages of
a heterogeneous catalyst, this system seems to be a good option for
fructose dehydration.
The effect of catalyst:fructose mass ratio was evaluated in the
range of 1:10–50, for HPW/MCM, and the results are shown in
Fig. 10. The amount of catalyst practically has no influence on the
fructose conversion. The highest HMF yield (80%) was obtained
using 1:10 ratio in 60 min, achieved in a shorter time than in blank
test (80 min). However, a further increase in the mass ratio to 1:30
or 50 decreased the HMF yield to 65 and 60%, respectively. This
result can be associated with the absence of enough amount of cat-
alytic sites necessary for converting the intermediate compounds
to HMF, since there was no formation of byproducts, such as lev-
ulinic and formic acids. Therefore, the optimal catalyst:fructose
mass ratio was 1:10 under the reaction conditions used.
Initial reaction rates were calculated based on HMF concen-
tration vs. time curves. The obtained values were 0.34, 0.08 and
0.22 mol h−1 g−1 for HPW, HCsPW and HPW/MCM, respectively.
These values clearly show that HCsPW is less active for fruc-
tose dehydration to HMF, and HPW/MCM is a good choice due
to its relatively large reaction rate and heterogeneity on reaction
media. Shimizu et al. [3] reported a reaction rate of approximately
0.9 mol h−1 g−1 for HPW using DMSO as solvent, at 120 ◦C, but
under evacuation conditions. Other catalysts, such as Amberlyst-
15, WO3/ZrO2 and H-BEA showed reaction rates much lower, in
the order of 0.05–0.2 mol h−1 g−1 [3]. Moreau et al. [32] reported
reaction rates of approximately 0.03 mol h−1 g−1 for fructose dehy-
dration using H-MOR as catalyst, in MIBK/water solvent, at 165 ◦C.
The effect of the temperature on dehydration of fructose cat-
alyzed by HPW/MCM was also evaluated, as shown in Fig. 9.
According to the experimental data, it is possible to observe that
the reaction temperature influences both fructose conversion and
HMF yield. At 140 ◦C, 98% of fructose conversion was obtained in
10 min of reaction, while at lower temperatures, 120 and 100 ◦C, in
the same time, the fructose conversion reached about 86 and 27%,
respectively. This result confirms that an increase of the temper-
ature favors the fructose conversion, as reported in the literature
[7,18–21].
Regarding the HMF yield, the temperature increase presents an
ambiguous effect. When the reaction temperature is 100 ◦C, the
HFM yield is 65% in 80 min, and when the temperature increases
to 120 or 140 ◦C, the same yield is reached in 10 min. According to
the literature, the formation rate of HMF is increased by a higher
enolization rate, as well as by a higher proportion of the acyclic and
furanose forms of fructose at higher temperature [33]. This effect
is further increased in the presence of DMSO, since the furanose
proportion is quite high (70% at 20 ◦C) and increases with temper-
ature [31]. Thus, it can be said that the temperature increase plays
a positive role on the HMF yield; however, this observation is only
valid for times lower than 10 min. After 20 min, there is almost no
difference in HMF yield with increasing temperature from 120 to
140 ◦C, which suggests a change of reaction pathway. The elevated
temperature together with long reaction times cause a decrease in
HMF yield because of side reactions that lead fructose and HMF
to byproducts. The activation energy for HMF degradation, as rehy-
dration reactions, is favored by increasing the temperature [7,8,18].
The maximum yield of 80% was obtained at 120 ◦C after a reaction
time of 60 min.
4. Conclusions
Aiming to transform the biomass rich in carbohydrates into
chemicals that can be used as direct substitutes of non-renewable
source compounds, this study evaluated the synthesis of HMF (5-
hydroxymethylfurfural) from fructose in the presence of DMSO as
solvent. The use of HPW (phosphotungstic acid) in the synthesis of
HMF has shown to be promising (92% of HMF yield at 120 ◦C), but
this acid is soluble in aprotic solvent. Thus, it was also evaluated the
Cs-exchanged phosphotungstic acid (HCsPW) as catalyst, since this
heteropolyacid is turned into insoluble compound by substituting
a fraction of their protons, and the HPW supported on mesoporous
MCM-41. Both catalysts showed similar results at 120 ◦C in 1 h of
reaction (80% of HMF yield). Thus, it was demonstrated that the
HPW/MCM-41 shows high catalytic activity for the conversion of
fructose to HMF, with advantage that its synthesis is much easier
than the HCsPW.
Acknowledgements
The authors thank the Foundation for Research of the State
of Rio de Janeiro (FAPERJ) for financial support granted to carry
out this work, the National Institute of Technology (INT) for N2
adsorption-desorption analyses and Carlos André de Castro Perez
from NUCAT/PEQ/COPPE/UFRJ for XRD in low angle.
References
[1] A. Gandini, Biocatalysis in Polymer Chemistry, first ed., Wiley-VCH, New York,
2011.
[2] A.C. Boisen, W. Fu, Y.Y. Gorbanev, T.S. Hansen, J.S. Jensen, S.K. Klitgaad, S.
Pedersen, A. Riiasger, T. Stahlberg, J.M. Woodley, Chem. Eng. Res. Des. 87
(2009) 1318–1327.
[3] K. Shimizu, R. Uozumi, A. Satsuma, Catal. Commun. 10 (2009) 1849–1853.
[4] C. Moreau, R. Durand, S. Razigade, J. Duhamet, P. Faugeras, P. Rivalier, P. Ros,
G. Avignon, Appl. Catal. 145 (1996) 211–224.
[5] Y. Takeuchi, F. Jin, K. Tohji, H. Enomoto, J. Mater. Sci. 43 (2007) 2472–2475.
[6] C. Carlini, M. Giuttari, A.M.R. Galletti, G. Sbrana, Appl. Catal. A 183 (1999)
295–302.
[7] F.N.D.C. Gomes, L.R. Pereira, N.F.P. Ribeiro, M.M.V.M. Souza, Braz. J. Chem. Eng.
32 (2015) 119–126.
[8] Y. Román-Leshkov, J.A. Dumesic, Top. Catal. 52 (2009) 297–303.
[9] A. Gandini, Polym. Chem. 1 (2010) 245–251.
[10] K.-I. Shimizu, R. Uozumi, A. Satsuma, Catal. Comm. 10 (2009) 1849–1853.
[11] R.-J. van Putten, J.C. van der Wall, E. de Jong, C.B. Rasrendra, H.J. Heeres, J.G. de
Vries, Chem. Rev. 113 (2013) 1499–1597.
[12] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171–198.
The rate constants at different temperatures for fructose conver-
sion to HMF using HPW/MCM were obtained by plotting ln(1 − X)
vs. time, where X is the fractional conversion. The first-order
kinetic model was assumed based on previous reports [34,35].
The activation energy calculated from the slope of Arrhenius plot
was 58.7 kJ mol−1, with a correlation coefficient of 0.997. This
value is much lower than those reported by Moreau et al. [32]
[13] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199–217.
[14] H. Eom, D. Lee, S. Kim, S. Chung, Y.G. Hur, K.H. Lee, Fuel 126 (2014) 263–270.
[15] P.R.S. Braga, A.A. Costa, E.F. de Freitas, R.O. Rocha, J.L. de Macedo, A.S. Araujo,
J.A. Dias, S.C.L. Dias, J. Mol. Catal. A: Chem. 358 (2012) 99–105.
[16] Y. Chen, X.-L. Zhang, X. Chen, B.-B. Dong, X.-C. Zheng, Solid State Sci. 24 (2013)
21–25.
[17] J.C. Juan, J. Zhang, M.A. Yarmo, J. Mol. Catal. A: Chem. 267 (2007) 265–271.
[18] Q. Zhao, L. Wang, S. Zhao, X. Wang, S. Wang, Fuel 90 (2011) 2289–2293.
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018

G Model
CATTOD-10048; No. of Pages9
ARTICLE IN PRESS
F.N.D.C. Gomes et al. / Catalysis Today xxx (2016) xxx–xxx
9
[19] C. Fan, H. Guan, H. Zhang, J. Wang, S. Wang, X. Wang, Biomass Bioenergy 35
(2012) 2659–2665.
[27] S. Zhang, Y. Zu, Y. Fu, M. Luo, D. Zhang, T. Efferth, Bioresour. Technol. 101
(2010) 931–936.
[20] Y. Qu, C. Huang, J. Zhang, B. Chen, Bioresour. Technol. 106 (2012) 170–172.
[21] I.V. Kozhevnikov, K.R. Kloetstra, A. Sinnema, H.W. Zandbergen, H. Van
Bekkum, J. Mol. Catal. A: Chem. 114 (1996) 287–298.
[22] X.K. Li, J. Zhao, W.J. Ji, Z.B. Zhang, Y. Chena, C.T. Au, S. Han, H. Hibst, J. Catal.
237 (2006) 58–66.
[28] L. Shen, H. Yin, A. Wang, Y. Feng, Y. Shen, Z. Wu, T. Jiang, Chem. Eng. J. 180
(2012) 277–283.
[29] J.K. Kim, J.H. Choi, J.H. Song, J. Yi, I.K. Song, Catal. Commun. 27 (2012) 5–8.
[30] A.S. Amarasekara, L.D. Williams, C.C. Ebede, Carbohydr. Res. 343 (2008)
3021–3024.
[23] F. Chen, M. Hong, W. You, C. Li, Y. Yu, Appl. Surf. Sci. 357 (2015) 856–865.
[24] M. Fournier, C. Feumi-Jantou, C. Rabia, G. Hervé, S. Launay, J. Mater. Chem. 2
(1992) 917–978.
[25] A.S. Dias, S. Lima, M. Pillinger, A.A. Valente, Carbohydr. Res. 341 (2006)
2946–2953.
[31] J.L. Figueiredo, F.R. Ribeiro, Catálise Heterogênea, first ed., Fundac¸ ão Calouste
Gulbenkian, Lisboa, 1987.
[32] C. Moreau, R. Durand, S. Razigade, J. Duhamet, P. Faugeras, P. Rivalier, P. Ros,
G. Avignon, Appl. Catal. A: Gen. 145 (1996) 211–224.
[33] B.M.F. Kuster, Starch 42 (1990) 314–321.
[26] N. Essayem, A. Holmqvist, P.Y. Gayraud, J.C. Vedrine, Y.B. Taarit, J. Catal. 197
(2001) 273–280.
[34] M. Bicker, J. Hirth, H. Vogel, Green Chem. 5 (2003) 280–284.
[35] Y. Xiao, Y.-F. Song, Appl. Catal. A: Gen. 484 (2014) 74–78.
Please cite this article in press as: F.N.D.C. Gomes, et al., Synthesis of 5-hydroxymethylfurfural from fructose catalyzed by phospho-
tungstic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.02.018