Microwave assisted rapid conversion of carbohydrates into 5-hydroxymethylfurfural catalyzed by mesoporous TiO2 nanoparticles

Article abstract of DOI:10.1016/j.apcata.2011.09.037

Energy efficient and sustainable process for the production of 5-hydroxymethylfurfural (HMF) from carbohydrates is highly demanding. Here, direct conversion of carbohydrates into HMF has been investigated over self-assembled mesoporous TiO₂ nanoparticles (NPs) catalyst. Monosachharides d-fructose and d-glucose, disaccharides sucrose, maltose, cellobiose were successfully converted into HMF with variable yields in both aqueous and organic mediums under microwave-assisted heating conditions. The effects of solvent polarity, microwave absorbing ability, catalyst loading, reaction time, and substrate variations on the HMF yields have been studied. Pyridine-IR and NH₃-TPD analyses confirmed the presence of Lewis acidic sites, whereas N₂ sorption analysis revealed high BET surface area for the self-assembled mesoporous TiO₂ nanomaterials synthesized by using sodium salicylate as template. High surface area, Lewis acidity and uniform nanosphere-like particle morphology are responsible for high catalytic activity in the dehydration of carbohydrate substrates over this mesoporous TiO₂ nanomaterial, whereas commercially available TiO₂ is almost inactive for this dehydration reaction under similar conditions. The higher microwave energy absorbing ability (tan δ) of the DMSO and NMP resulted in higher HMF yield in organic solvents than water. Catalyst life-time analysis suggested that mesoporous TiO₂ NPs catalysts can be recycled for four catalytic cycles without appreciable loss of its catalytic activity.

Full text of DOI:10.1016/j.apcata.2011.09.037

Applied Catalysis A: General 409–410 (2011) 133–139
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Microwave assisted rapid conversion of carbohydrates into
5
-hydroxymethylfurfural catalyzed by mesoporous TiO nanoparticles
2
a
a
b
c
b,∗
a,∗∗
Saikat Dutta , Sudipta De , Astam K. Patra , Manickam Sasidharan , Asim Bhaumik , Basudeb Saha
a
Laboratory of Catalysis, Department of Chemistry, University of Delhi, North Campus, Delhi-110007, India
Department of Materials Science, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700032, India
Department of Chemistry, Faculty of Science and Engineering, Saga University, 1 Honjo-machi, Saga 840-8502, Japan
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 July 2011
Received in revised form 29 August 2011
Accepted 29 September 2011
Available online 6 October 2011
Energy efficient and sustainable process for the production of 5-hydroxymethylfurfural (HMF) from car-
bohydrates is highly demanding. Here, direct conversion of carbohydrates into HMF has been investigated
over self-assembled mesoporous TiO2 nanoparticles (NPs) catalyst. Monosachharides d-fructose and d-
glucose, disaccharides sucrose, maltose, cellobiose were successfully converted into HMF with variable
yields in both aqueous and organic mediums under microwave-assisted heating conditions. The effects
of solvent polarity, microwave absorbing ability, catalyst loading, reaction time, and substrate varia-
tions on the HMF yields have been studied. Pyridine-IR and NH3-TPD analyses confirmed the presence
of Lewis acidic sites, whereas N2 sorption analysis revealed high BET surface area for the self-assembled
mesoporous TiO2 nanomaterials synthesized by using sodium salicylate as template. High surface area,
Lewis acidity and uniform nanosphere-like particle morphology are responsible for high catalytic activity
in the dehydration of carbohydrate substrates over this mesoporous TiO2 nanomaterial, whereas com-
mercially available TiO2 is almost inactive for this dehydration reaction under similar conditions. The
higher microwave energy absorbing ability (tan ı) of the DMSO and NMP resulted in higher HMF yield
in organic solvents than water. Catalyst life-time analysis suggested that mesoporous TiO2 NPs catalysts
can be recycled for four catalytic cycles without appreciable loss of its catalytic activity.
Keywords:
Mesoporous titania
Carbohydrates
Microwave catalysis
5
-Hydroxymethylfurfural
Sustainable chemistry
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
multiple conditions are now established to enable dehydration of
d-fructose to HMF in acidic water [9], sub- and supercritical water,
water–acetone mixed solvents [10,11], organic solvents [12,13],
multicomponent mixed solvents [14], and ionic liquids in combina-
tion with metal catalysts [15]. Biphasic continuous systems using
porous solid acid catalysts and specially nanocatalysts have great
potential in this context [16–19].
Among the approaches for the efficient synthesis of HMF from
sugar derivatives, heterogeneous acid catalyst plays an important
role. In the past, ion-exchange resin-activated carbon in water
and Dowex-type ion-exchange resin in dimethyl sulfoxide (DMSO)
were used as catalysts [20]. High conversion and selectivity of HMF
has been reported in biphasic solvents using ion exchange resin
under microwave-assisted heating conditions [21]. Microwave
assisted dehydration of fructose with Amberlyst-15 catalyst in ionic
liquid [BMIM][Cl], or its combination with hydrotalcite exhibited
high HMF yields [22–24].
Today the gap between the supply and demand of energy has
reached into an alarming level, which motivated the researchers
for new synthetic route and process development for utilizing
renewable feedstock for chemicals and liquid fuels in order to
reduce our dependence on petroleum sources and rising limit
of CO2 level in the atmosphere [1,2]. One compound that has
been under particular scrutiny as a future platform chemical is
5
-hydroxymethylfurfural (HMF), which is promising because of
its identical carbon skeleton to those in most abundant cellulose
and biomass-derived carbohydrates [3,4]. HMF is primarily con-
sidered a starting material for liquid transportation fuels [5,6] and
polyester building block chemicals [7,8]. In this context a rapid
progress has been witnessed in developing catalytic processes to
convert biomass and carbohydrates into HMF [9–15]. More notably,
Syntheses of HMF from hexoses in the presence of solid acid
catalysts are known today [25]. ZrO2 and TiO2 catalyzed conver-
sion of d-fructose and d-glucose into HMF in aqueous medium
∗
Corresponding author. Tel.: +91 3324734971; fax: +91 3324732805.
Corresponding author.
E-mail addresses: msab@iacs.res.in (A. Bhaumik), bsaha@chemistry.du.ac.in
∗
∗
◦
under microwave irradiation at 200 C produced moderate HMF
(
B. Saha).
yields (38.1% and 18.6% respectively) [26,27]. Sulphated zirconia
0
926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.09.037

1
34
S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
Micromeritics ChemiSorb 2720 attached with
conductivity detector.
a
thermal
O
HO
OH
OH
O
TiO
2
NPs
In a typical microwave-assisted reaction, desired amount of car-
bohydrate substrates and catalyst were charged in a microwave
tube containing of 2 mL solvent. The tube was placed in the
microwave reactor. The microwave power was set to 300 W.
Desired temperatures and time was set. After the reaction, the tube
was allowed to cool to room temperature. The reaction mixture
was then analyzed (after separating the reactant) by using NMR
and spectrophotometric techniques.
O
HO
OH
Fructose Catalyst
HO
H
Solvent
1
00-140 o
C
HMF
Scheme 1. Microwave-assisted dehydration of d-fructose into HMF catalyzed by
self-assembled mesoporous TiO2 NPs.
−
The recycling efficiency of the catalyst was determined by using
dehydration of fructose as a representative reaction. In this study,
(
SO42 /ZrO ) was found to be effective catalysts for the HMF pro-
2
◦
duction from fructose under microwave heating at 180–200 C [13].
Mixed metal oxide TiO –ZrO2 under hot compressed water (HCW)
acts as a good catalyst for glucose dehydration in sub-critical water
28].
Despite recent developments in heterogeneous catalysts for
5
wt% fructose solution in water was mixed with 100 mg meso-
2
porous TiO2 sample in a microwave tube. The tube was placed
in the microwave reactor and the mixture was heated for 5 min
[
◦
at 120 C using 300 W microwave power. After the reaction, the
tube was cooled down to room temperature and liquid was col-
lected by decanting the tube. The solid residue left in the tube
was collected and washed several times with distilled water and
dehydration of carbohydrates to HMF, it remains a challenge to
develop a more energy efficient heterogeneous catalyst and pro-
cess for efficient conversion of carbohydrates to HMF. Porous TiO2
with high surface area and unique particle morphology, which
can be synthesized through the templating pathways [29,30] can
be an efficient solid acid catalyst in this context as surface Lewis
acidity together with high surface area could play crucial role in
this dehydration reaction. Herein, we first report the use of self-
assembled mesoporous titania nanoparticles synthesized by using
sodium salicylate as template in efficient catalytic dehydration of
carbohydrates into HMF under microwave-assisted heating in var-
ious solvents such as water, water–MIBK (methylisobutyl ketone),
NMP (N-methylpyrollidone), and DMSO (dimethyl sulfoxide) in
morderate to good yield in very short reaction time (Scheme 1).
◦
dried in oven for 3 h at 100 C. The dried catalyst was re-used
for three more cycles by following the above method and HMF
yields were determined from each run. The yield of HMF was
1
determined by both H NMR and UV–visible spectrophotometric
1
techniques. For H NMR spectroscopic analysis, HMF was extracted
from the reaction mixture with diethyl ether. Pale yellow oily
HMF was obtained after removing the solvent in vacuum at room
temperature.
2
.1. 1H NMR method
For quantifying the yield of HMF using 1H NMR spectroscopic
technique, known concentration of mesitylene (internal standard)
was added into the HMF product solution in DMSO-d6. The per-
centage of HMF yield was calculated by using the integrated values
of the aldehyde proton (ı = 9.58 ppm) of HMF and three aromatic
ring protons of mesitylene (ı = 6.79 ppm). First, standard HMF solu-
tion of 99% purity was analyzed for correlating the percentage of
actual and calculated amount of HMF. Once good correction was
established, extracted HMF product samples were run and the per-
centage of HMF yield was calculated.
2
. Experimental
Fructose, glucose, sucrose, cellobiose, and maltose monohydrate
were purchased from Sigma–Aldrich and were used without fur-
ther purification. MIBK, DMSO and NMP were purchased from
Spectrochem, India. Self-assembled mesoporous TiO2 was pre-
pared by following the literature procedure using sodium salicylate
as template [30]. As-synthesized mesoporous TiO2 samples were
◦
calcined at 500 C for 6 h to remove the template molecules and
this sample was used for all the characterizations except the TPD
analysis, where the sample was activated at 800 C for 3 h before
2.2. UV–visible spectrophotometric method
◦
ammonia adsorption. Unless otherwise stated, freshly distilled
water was used as aqueous phase for reactions in aqueous and
aqueous–organic phase.
The UV–visible spectrum of pure HMF solution has a distinct
peak at 284 nm with corresponding extinction coefficient (ε) value
4
−1
−1
of 1.66 × 10 M cm . The yield of HMF was determined by mea-
suring the absorbance of HMF product solution at 284 nm and the
extinction coefficient value. Repeated measurement of the same
solution showed the percentage of error associated with this mea-
surement was ± 0.3%. The yield of HMF obtained from two different
The catalytic reactions for carbohydrate substrates dehydra-
tion to HMF were performed in a CEM Matthews WC Discover
microwave reactor (model no. 908010 DV9068 equipped with
programmable pressure and temperature controller). 1H NMR
spectral analysis was performed on a JEOL JNM ECX-400 P 400 MHz
instrument and data were processed using a JEOL DELTA program
version 4.3.6. HMF yields were determined by using a UV–visible
spectrophotometer (UV-SPECORD 250 analytikjena spectrometer)
monitoring the absorbance of the product solution at 284 nm (cor-
1
methods ( H NMR and UV–visible) for the same reaction product
agreed very well and the result was within ± 5% error.
3. Results and discussion
4
−1
−1
responding to ꢀmax of HMF with ε = 1.66 × 10 M cm ). A JEOL
JEM 6700F field emission scanning electron microscope was used
for the determination of the particle morphology of the TiO2 NPs.
Fourier transform infrared (FT IR) spectra of the pyridine adsorbed
samples were recorded by using a Perkin Elmer Spectrum 100
spectrophotometer. For the temperature programmed desorption
Self-assembly of tiny mesoporous TiO nanospheres used as cat-
2
alyst was prepared utilizing sodium salicylate (SS) as a template by
following the literature procedure [30]. This material is composed
of very tiny uniform spherical nanoparticles of 12–20 nm in size.
2
−1
N2 sorption analysis revealed high surface area (326 m g ) and
3
−1
large pore volume (0.3 cm g , not shown). FT IR spectra of the
(
TPD) of ammonia studies, sample was activated and then ammo-
pyridine adsorbed mesoporous TiO NPs show two characteristic
2
−
1
−1
nia was injected at room temperature in the absence of carrier gas
flow. The temperature was then raised in a stepwise manner at
a linear heating rate of 10 C/min. The desorbed ammonia in the
bands at 1588 and 1439 cm (Fig. 1). The band at 1439 cm could
be attributed to the adsorbed pyridine at the Lewis acid site (PyL)
[31,32]. This band showed no sign to disappear with increase in
the desorption temperature. This result suggested the presence of
◦
◦
temperature range 100–800 C was analyzed by using
a

S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
135
Table 1
Catalytic activity of various metal oxides in the dehydration reaction under MW
conditions.
Entry
Metal oxides
Loading (mg)
HMF yield (%)
1
2
3
4
CeO2 NPs (6–10 nm)
Al2O3 (commercial)
TiO2 (commercial)
SiO2 (commercial)
50
50
50
50
1.0
1.8
3.4
1.2
◦
Fructose = 100 mg; metal oxides = 50 mg; T = 120 C; solvent = water; t = 5 min.
well-defined nanosphere like particle morphology [see the SEM
images of the reused catalysts below] catalyzed the fructose
dehydration and produced 34.3% HMF in water under microwave-
◦
assisted heating at 120 C in 5 min (Table 2, entry 2). To test
this catalytic effectiveness further, control experiments were per-
formed for fructose dehydration without mesoporous TiO catalyst,
2
which produced only trace amount of HMF under identical reaction
conditions (Table 2, entry 1). Thus, high yield of HMF in the pres-
Fig. 1. FT IR spectra of mesoporous TiO2 NPs (a) and pyridine desorbed samples at
23 K (b), 423 K (c), 323 K (d) and 298 K (e).
5
ence our prepared mesoporous TiO nanoparticulate material again
2
validated its catalytic effectiveness as discussed above.
considerably strong Lewis acid sites in our mesoporous TiO2 NPs.
Further, in Fig. 2 we have plotted the temperature programmed
More experiments were designed for optimizing the reaction
conditions and HMF yields at variable substrate to catalyst ratios
and in different solvents. As shown in Table 2, these experi-
ments produced variable yields of HMF in the range of 25–54%.
These results clearly indicates that mesoporous TiO2 is an effec-
tive catalyst for HMF synthesis from fructose. In an earlier report,
anatase-TiO2 catalyzed fructose dehydration reaction claimed to
desorption of NH . As seen from Fig. 2 that the desorption tem-
3
◦
perature range is very high (400–800 C with peaks at ca. 650 and
◦
7
44 C). Total acidity of this mesoporous TiO NPs corresponds
2
−1
to ca. 0.4 mmol g . Hence this strong surface acidity of the self-
assembled TiO NPs could be effectively utilized in the dehydration
2
of carbohydrates.
◦
produce 33.5% HMF at 200 C in hot compressed water [26,27]. It is
Several solid metal oxides such as nanoparticulate of ceria
to point out that such a high temperature conversion reported by
Qi et al. [27] is energy consuming and not sustainable. It is perti-
nent to mention that, from the present study maximum 36% HMF
yield was achieved in aqueous phase using mesoporous TiO2 cata-
(
CeO ), commercial alumina (Al O ), commercial SiO , commer-
2 2 3 2
cial anatase TiO , and mesoporous TiO2 nanoparticles synthesized
2
here were screened for the catalytic dehydration of fructose in
aqueous medium under identical microwave-assisted heating con-
ditions. Among these metal oxides, except our self-assembled
mesoporous TiO2 NPs, others can rarely be considered as cat-
alysts for HMF production in aqueous or organic medium. As
shown in Table 1, in the presence of other metal oxides as cata-
lyst only 1.0–3.4% HMF have been produced (reaction conditions:
◦
lyst at a considerably lower temperature (120 C). This justifies that
the microwave-assisted process described herein is more sustain-
able. To our knowledge, this is the first report of using mesoporous
TiO2 material as an effective catalyst for the rapid conversion of
carbohydrates into HMF under mild microwave assisted heating
conditions.
The promising catalytic activity of the mesoporous TiO2 NPs for
fructose dehydration has prompted us to test the effectiveness of
this catalyst for HMF synthesis from glucose and other disaccha-
rides such as sucrose, cellobiose and maltose in aqueous and in
organic solvents under microwave heating. The reaction conditions
of glucose, sucrose, cellobiose and maltose dehydration reactions
and corresponding HMF yields in different solvents are summa-
rized in Table 3. Under comparable reaction conditions and 2:1
◦
fructose = 100 mg, catalyst = 50 mg, T = 120 C, reaction time = 5 min
and solvent = water). Commercial anatase TiO was also ineffective
2
to catalyze the dehydration of fructose, suggesting the nanostruc-
ture of our mesoporous TiO2 NPs plays crucial role in this catalytic
reaction. In contrast to the catalytic activity of the metal oxides
listed in Table 1, mesoporous TiO2 nanoparticulate material with
weight ratio of glucose to TiO , 22.1% and 35.5% HMF yields were
2
achieved from glucose dehydration in water and DMSO, respec-
tively. The similar reactions performed in NMP and water–MIBK
biphasic solvents produced 29.6% and 25.9% HMF, respectively. The
data presented in Table 3 shows a variation in HMF yields (%) as a
function of the polarity of the solvents for the glucose conversion
to HMF. Under identical reaction conditions, sucrose dehydra-
tion reaction produced less HMF than that of glucose dehydration
reaction. Using 1:1 substrate to catalyst weight ratio, the sucrose
dehydration reaction produced 15.1% and 20.5% HMF in water and
DMSO, respectively. The NMP and water–MIBK solvents mediated
sucrose dehydration reactions produced 14.9% and 13.9% HMF. The
cellobiose dehydration reaction produced 14.5% and 18.7% HMF in
water and DMSO, respectively. An earlier report claimed 41% HMF
yield from cellobiose using a biphasic [BMIM]Cl–MIBK solvents and
CrCl catalyst [33]. This difference in HMF yields can be attributed to
3
the strong Lewis acidic nature of the chromium chloride catalyst in
comparison to the surface acidity of the mesoporous TiO2 catalyst.
Moreover, HMF yields in ionic liquids are superior to that in water or
Fig. 2. Temperature programmed desorption of ammonia over mesoporous TiO2
nanoparticles.

1
36
S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
Table 2
Microwave-assisted fructose dehydration to HMF with mesoporous TiO2 catalyst.
a
Entry
Fructose (mg)
TiO2 NPs (mg)
Solvent (mL)
t (min)
HMF yield (%)
UV–vis
NMR
Isolated
1
2
3
4
5
6
7
8
9
100
100
100
100
100
50
50
50
50
100
100
100
100
50
50
50
50
150
50
–
100
Water (2)
Water (2)
Water (2)
Water (2)
Water (2)
Water (1)
Water (1)
Water (1)
DMSO (1)
DMSO (2)
DMSO (2)
DMSO (2)
DMSO (2)
DMSO (1)
DMSO (1)
DMSO (1)
DMSO (1)
Water–MIBK (1/2)
NMP (1)
5
5
5
5
5
2
10
15
5
5
5
5
5
0.8
34.3
33.5
30.4
25.6
27.2
35.0
36.3
54.1
53.4
50.8
48.3
40.5
32.3
32.7
29.5
31.9
32.8
30.0
50
25
10
25
25
25
50
50
25
10
5
25
25
25
–
75
25
25
10
10
25
25
33.5
34.9
52.6
51.4
49.1
47.3
53.2
51.7
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
b
2
41.8
10
15
5
5
5
53.7
54.0
20.3
40.2
36.5
29.3
27.4
27.9
49.1
48.3
52.8
38.0
53.2
38.4
50
Acetonitrile (1)
Water (2)
Water (2)
DMSO (2)
DMSO (2)
5
100
100
100
100
10
20
5
c
d
5
◦
◦
◦
◦
T = 120 C (water), 130 C (water/MIBK), 120 C (acetonitrile), 140 C (NMP and DMSO).
a
Microwave power 300 W.
Fructose conversion was 61.0%.
Microwave power 100 W.
Microwave power 200 W.
b
c
d
DMSO. When maltose monohydrate was used as a substrate, HMF
yields were slightly lower than that obtained from other two disac-
charides, cellobiose and sucrose. In case of DMSO solvent mediated
reaction, the slight difference in HMF yields from maltose can be
explained due to the presence of water of crystallization associ-
ated with the maltose molecular structure, which could inhibit its
dehydration into HMF.
in catalyst loading to 25 mg resulted slight improvement in HMF
yield to 28.4%. To test the effect of more catalyst loading, 50 mg
and 100 mg of mesoporous TiO2 were used, which produced 33.5%
and 34.3% HMF, respectively. Further increase in catalyst loading
beyond 1:1 weight ratio of substrate to catalyst showed no signif-
icant increase in HMF yield. The DMSO solvent mediated fructose
dehydration reaction showed similar trend of HMF yield on cata-
lyst loading. Maximum 54.1% HMF was obtained in DMSO solvent
mediated reaction in comparison to 34.3% in water.
Since, maximum HMF yields from fructose were obtained
with a catalyst/substrate ratio of 1:1, results with a low load-
ing of catalyst (catalyst to substrate ratio 1/10) would be more
favourable to demonstrate the catalytic activity against run time.
The additional experiments using fructose as substrate with 1/10
catalyst–substrate ratio in water showed a small increase in HMF
yield from 25.6% to 27.9% (Table 2, entry 5, 21, and 22) upon increas-
ing the run time of reaction from 5 min to 20 min.
Microwave-assisted HMF synthesis from fructose using meso-
porous TiO2 catalyst has been studied by varying catalyst
concentrations for optimizing the reaction conditions and maxi-
mizing the HMF yield. A plot showing the catalyst loadings versus
HMF yield (%) is shown in Fig. 3. When fructose solution (100 mg
in 2 mL water) was treated with 5 mg of TiO2 for 5 min under
◦
microwave-assisted heating at 120 C, an HMF yield of 19.0% was
obtained. To improve the HMF yield, catalyst loading was increased
to 10 mg for same fructose concentration (100 mg in 2 mL water),
which resulted an increase in HMF yield to 25.6%. Further increase
Table 3
Results of glucose, sucrose, maltose, and cellobiose dehydration to HMF catalyzed by mesoporous TiO2 NPs.
◦
Entry
Substrate (mg)
TiO2 NPs (mg)
Solvent (mL)
T ( C)
t (min)
HMF yield (%)
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
Glucose (100)
Glucose (100)
Glucose (50)
Glucose (50)
Glucose (150)
Glucose (50)
Sucrose (50)
Sucrose (100)
Sucrose (100)
Sucrose (50)
Sucrose (150)
Sucrose (50)
Cellobiose (100)
Cellobiose (100)
Maltose (100)
Maltose (100)
50
50
50
50
75
50
50
50
50
50
75
50
100
100
100
100
Water (2)
DMSO (2)
DMSO (1)
NMP (1)
Water/MIBK (1/2)
Water (1)
Water (1)
DMSO (2)
DMSO (1)
NMP (1)
Water–MIBK (1/2)
Water (1)
Water (2)
DMSO (2)
Water (2)
DMSO (2)
120
140
140
140
130
120
120
140
140
140
130
120
120
140
120
140
2
5
5
5
5
5
22.1
35.5
37.2
29.6
25.9
24.8
15.1
20.5
21.0
12.0
14.9
13.9
14.5
18.7
10.7
14.1
10
5
5
5
5
5
5
5
5
5
1
1
1
1
1
1
1

S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
137
Fig. 3. The effect of HMF yield on catalyst loading for the microwave assisted
fructose dehydration in water and DMSO. Reaction conditions: fructose = 100 mg,
TiO2 = 5–100 mg, T = 120 C, t = 5 min.
Fig. 5. Results of TiO2 NPs catalyzed conversion of fructose, glucose, and sucrose to
◦
HMF in different solvents. Reaction conditions: substrate = 100 mg, TiO2 = 100 mg,
◦
T = 120 C, t = 5 min.
The reaction time for the dehydration of carbohydrates was var-
ied to optimize the HMF yield. The yield of HMF was monitored as
a function of reaction time from 2 min to 15 min for a set of exper-
iments with variable catalyst to substrate ratios in aqueous and
DMSO medium under microwave-assisted heating conditions. As
plotted in Fig. 4, maximum yield of HMF recorded within 5 min for
most of the reaction. When these reactions were continued for a
longer period of time, ca. 15 min, no significant increase in HMF
yield was observed. The similar trend in HMF yield was recorded at
different catalyst to substrate ratios as well. Curves a and b in Fig. 4
represents the dehydration of fructose at 2:1 substrate to catalyst
ratio in water and DMSO, respectively. Curve b shows the forma-
tion of 52.9% HMF in 5 min in DMSO. Extending the reaction time to
continuing the reaction for a longer time. The results suggest that
the kinetics of HMF formation is rapid in initial 2 min reaction time
followed by a slower rate of reaction.
Microwave-assisted dehydration reaction of sugar substrates
(fructose, glucose, sucrose, cellobiose, and maltose) were carried
out in aqueous, biphasic aqueous-MIBK, NMP, and DMSO medium.
Details of the reaction conditions and corresponding HMF yields
are summarized in Tables 2 and 3. Under comparable reaction
conditions, mesoporous TiO2 catalyzed fructose dehydration in
water, NMP, water–MIBK, and DMSO produced 34.3%, 36.5%, 40.2%,
54.1% HMF, respectively. Similar trend of increased HMF yields
were observed for glucose and sucrose as substrates. The higher
yield of HMF obtained in DMSO is attributed to the large tangent
(tan ı) value of DMSO (0.825) than that of water (0.123) and NMP
(0.275). The higher tan ı value of DMSO corresponds to its higher
microwave energy absorbing ability than that of water [34]. In
DMSO, more electromagnetic energy is converted into heat energy
at a given frequency and temperature which resulted to improved
15 min resulted only 1% increase in HMF yield. In aqueous media,
32.1% HMF was obtained in 5 min reaction time which increased to
36.3% in 15 min. An increase in HMF yield was also noted for the
conversion of glucose and sucrose in water and DMSO under essen-
tially identical reaction conditions (Fig. 4, curves c–e and f–h) upon
Fig. 4. Results of carbohydrate dehydration to HMF with TiO2 NPs catalysts at var-
◦
ious substrates to catalyst ratios in water at 120 C. (unless otherwise mentioned.)
◦
Reaction conditions: (a) fructose: TiO2 = 2:1; (b) fructose: TiO2 = 1:2, DMSO, 140 C;
Fig. 6. HMF yield versus reusability of the TiO2 catalyst in water and DMSO
(
c) glucose: TiO2 = 1:1; (d) glucose: TiO2 = 1:2; (e) glucose: TiO2 = 1:4; (f) sucrose:
for microwave-assisted fructose dehydration reaction. Reaction conditions: fruc-
tose = 100 mg, TiO2 = 100 mg, T = 120 C, t = 5 min, water or DMSO = 2 mL.
◦
TiO2 = 1:1; (g) sucrose: TiO2 = 1:2; (h) sucrose: TiO2 = 1:4.

1
38
S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
Fig. 8. Comparative HMF yields obtained from a range of carbohydrate sub-
strates from TiO2 catalyzed dehydration in water and DMSO. Reaction conditions:
◦
sugar = 100 mg, TiO2 = 100 mg, T = 120 C, t = 5 min.
The reusability of the TiO NP catalyst was examined for fructose
2
dehydration reaction in aqueous and organic media by recycling the
solid catalyst after oven drying for each run. A plot of the catalyst
cycle number versus HMF yields for fructose dehydration is shown
in Fig. 6. The results show that the loss of activity of the catalyst
in terms of HMF yield is only 2% in fourth catalytic cycles (34.3%
HMF in first cycle and 32.0% HMF yield in fourth cycle). The catalyst
life-time study in DMSO solvent also showed about 2% decrease in
HMF yield in fourth catalytic cycles (Fig. 6). FE-SEM images of the
reused TiO NPs after first and fourth recycling are shown in Fig. 7(a)
2
and (b), which reveals that small uniform nanosphere-like particle
morphology has been retained after fourth reaction cycle.
We have studied a range of carbohydrate substrates to examine
the catalytic performance of mesoporous TiO2 NPs in producing
the HMF. A decreasing trend of HMF yields from 34.3% to 10.4%
is observed from fructose to maltose monohydrate in aqueous
◦
medium at 120 C in 5 min reaction time (Fig. 8). It is interesting
to note that maltose is a disaccharide formed from two units of
glucose joined with an ␣(1 → 4) bond, which is structurally similar
to cellobiose. In this study, we have obtained higher yield of HMF
from cellobiose in comparison to that from maltose. This could be
probably due to the water inhibition effect of the hydrated maltose
monohydrate. Similar trend of HMF yields were observed in DMSO.
In DMSO, fructose and glucose dehydration reactions produced
maximum 54.1% and 37.2% HMF, respectively under identical reac-
tion conditions. This decrease in HMF yield from fructose to maltose
agreed well with the reported data of CrCl2 catalyzed dehydration
of several carbohydrate substrates [35].
Fig. 7. The FE-SEM images of mesoporous TiO2 nanospheres after first (a) and fourth
b) catalytic cycles for the microwave assisted dehydration of fructose.
(
HMF yield in comparison to that of NMP and water. The results of
HMF yields (%) as a function of different solvents is shown in Fig. 5.
The higher yield of HMF in water–MIBK biphasic solvent than that
in water is due to the thermodynamic phenomenon of the biphasic
solvent in which HMF can easily accumulates in the organic phase
after its formation in aqueous phase, and thereby drives the dehy-
dration reaction equilibrium towards the HMF side. Maximum HMF
yield (54.1%) from fructose was obtained over mesoporous TiO cat-
4. Conclusion
2
alyst in DMSO likely due to its high microwave absorption ability
(
tan ı). To study the direct influence of the microwave energy on
From our experimental observations we can conclude that
the self-assembled mesoporous nanoparticulate TiO2 material can
be utilized as catalyst for the microwave-assisted conversion of
carbohydrates into HMF in aqueous and organic media. These
self-assembled spherical TiO2 NPs catalyzed fructose dehydra-
tion reaction produced maximum 34.3% and 54.1% HMF in water
and DMSO, respectively. Moderate yield of HMF (37.2%) was also
obtained from glucose as substrate in DMSO. Optimized reaction
conditions were established by varying the catalyst loading, reac-
tion time, substrates to catalyst weight ratios and solvents. The
the catalytic performances of TiO , fructose dehydration process in
DMSO was carried out by varying microwave power at 100, 200, and
3
at 140 C. No significant difference in HMF yield (Table 2, entry 11,
2
results demonstrates that minimum 100 W microwave energy to
be sufficient for the significant reaction progression using TiO cat-
2
00 W independently with a 4:1 fructose to catalyst ratio for 5 min
◦
3, and 24) was evident upon varying microwave power. These
2
alyst as observed in the case of HCl-catalyzed microwave-assisted
fructose dehydration in aqueous media [9].

S. Dutta et al. / Applied Catalysis A: General 409–410 (2011) 133–139
139
presence of strong Lewis acidic sites in TiO NPs surface was deter-
[8] O. Casanova, S. Iborra, A. Corma, ChemSusChem 2 (2009) 1138–1144.
9] T.S. Hansen, J.M. Woodley, A. Riisager, Carbohydr. Res. 344 (2009) 2568–2572.
10] M. Bicker, J. Hirth, H. Vogel, Green Chem. 5 (2003) 280–284.
11] A. Chuntanapum, T.L. Yong, S. Miyake, Y. Matsumura, Ind. Eng. Chem. Res. 47
(2008) 2956–2962.
2
[
mined by pyridine-IR and NH -TPD studies. Catalyst life-time study
3
[
[
showed the catalyst retained its activity after four repeating cycles
in terms of HMF yields. The ability of mesoporous TiO NPs catalyst
2
[
12] C.V. McNeff, D.T. Nowlan, L.C. McNeff, B.W. Yan, R.L. Fedie, Appl. Catal. A: Gen.
to convert a range of carbohydrate substrates into HMF in aqueous
and organic media under microwave heating promises the oppor-
tunity for utilizing this catalyst for the conversion of abundant and
inexpensive cellulosic biomass into a renewable energy source.
384 (2010) 65–69.
[
13] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr., Catal. Commun. 10 (2009)
1771–1775.
[14] Y. Román-Leshkov, J.A. Dumesic, Top. Catal. 52 (2009) 297–303.
[
15] M.E. Zakrzewska, E. Bogel-Lukasik, R. Bogel-Lukasik, Chem. Rev. 111 (2011)
97–417, and references therein.
3
Acknowledgements
[
[
[
16] M. Mascal, E.B. Nikitin, Angew. Chem. Int. Ed. 47 (2008) 7924–7926.
17] W. Schmidt, ChemCatChem 1 (2009) 53–67.
18] M.C. Avila, N.A. Comelli, E. Rodriguez-Castellon, A. Jimenez-Lopez, R.C. Flores,
E.N. Ponzi, M.I. Ponzi, J. Mol. Catal. A: Chem. 322 (2010) 106–112.
19] M. Nandi, J. Mondal, K. Sarkar, Y. Yamauchi, A. Bhaumik, Chem. Commun. 47
(2011) 6677–6679.
[20] G.A. Halliday, R.J. Young, V.V. Grushin, Org. Lett. 5 (2003) 2003–2005.
21] J.N. Chheda, J.A. Dumesic, Catal. Today 123 (2007) 59–70.
22] X.H. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr., Green Chem. 10 (2008) 799–805.
23] M. Ohara, A. Takagaki, S. Nishimura, K. Ebitani, Appl. Catal. A: Gen. 383 (2010)
We thank University Grant Commission, New Delhi, India for
financial support. SD acknowledges University Grant Commission,
New Delhi, India for supporting this research with a DS Kothari
Postdoctoral Research Grant. AB wishes to thank DST New Delhi for
financial supports. SD thanks University Grant Commission, New
Delhi, India for a Junior Research Fellowship.
[
[
[
[
1
49–155.
[24] X.H. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr., ChemSusChem 2 (2009) 944–946.
Appendix A. Supplementary data
[
[
[
[
25] A.A. Rosatella, S.P. Simeonov, R.F.M. Frade, C.A.M. Afonso, Green Chem. 13
(
2011) 754–793, and references therein.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2011.09.037.
26] M. Watanabe, Y. Aizawa, T. Iida, R. Nishimura, H. Inomata, Appl. Catal. A: Gen.
295 (2005) 150–156.
27] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr., Catal. Commun.
244–2249.
9 (2008)
2
References
28] A. Chareonlimkun, V. Champreda, A. Shotipruk, N. Laosiripojana, Bioresour.
Technol. 101 (2010) 4179–4186.
[
1] A.J. Ragauskas, C.K. Williams, B.H. Davison, G. Britovsek, J. Cairney, C.A. Eck-
ert, W.J. Frederick, J.P. Hallett, D.J. Leak, C.L. Liotta, J.R. Mielenz, R. Murphy, R.
Templer, T. Tschaplinski, Science 311 (2006) 484–489.
[29] N. Pal, M. Paul, A. Bera, D. Basak, A. Bhaumik, Anal. Chim. Acta 674 (2010)
96–101.
[30] A.K. Patra, S.K. Das, A. Bhaumik, J. Mater. Chem. 21 (2011) 3925–3930.
[31] M.M. Mohamed, W.A. Bayoumy, M. Khairy, M.A. Mousa, Micropor. Mesopor.
Mater. 103 (2007) 174–183.
[32] S.K. Das, M.K. Bhunia, A.K. Sinha, A. Bhaumik, ACS Catal. 1 (2011) 493–501.
[33] S. Lima, P. Neves, M.M. Antunes, M. Pillinger, N. Ignatyev, A.A. Valente, Appl.
Catal. A: Gen. 363 (2009) 93–99.
[
[
[
[
[
2] Y.C. Lin, G.W. Huber, Energy Environ. Sci. 2 (2009) 68–80.
3] G.W. Huber, A. Corma, Angew. Chem. Int. Ed. 46 (2007) 7184–7201.
4] X.L. Tong, Y. Ma, Y.D. Li, Appl. Catal. A: Gen. 385 (2010) 1–13.
5] R.M. West, Z.Y. Liu, M. Peter, J.A. Dumesic, ChemSusChem 1 (2008) 417–424.
6] G. Tian, R. Daniel, H. Li, H. Xu, S. Shuai, P. Richards, Energy Fuels 24 (2010)
3
898–3905.
[34] D. Dallinger, C.O. Kappe, Chem. Rev. 107 (2007) 2563–2591.
[35] H.B. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Science 316 (2007) 1597–1600.
[
7] Y. Román-Leshkov, J.N. Chheda, J.A. Dumesic, Science 312 (2006) 1933–1937.