Microwave assisted conversion of carbohydrates and biopolymers to 5-hydroxymethylfurfural with aluminium chloride catalyst in water

Article abstract of DOI:10.1039/c1gc15550d

The common Lewis acid AlCl₃ has efficiently produced 5-hydroxymethylfurfural (HMF) from carbohydrate and biopolymer substrates in water, DMSO, and water-methylisobutylketone biphasic solvents under microwave irradiation. The yield of HMF in different solvents follows an increasing order from water to water-MIBK biphasic solvent to DMSO. The yield of HMF increased with an increase in catalyst loading whereas it remains unchanged upon increase of the carbohydrate concentration. In most reactions, the maximum yield of HMF is recorded within 2 min of reaction time. The mechanism of the AlCl₃ catalyzed glucose dehydration reaction is proposed to proceed through the isomerization of glucopyranose to fructofuranose, followed by a proton assisted transformation of fructofuranose to HMF.

Full text of DOI:10.1039/c1gc15550d

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 2859
www.rsc.org/greenchem
PAPER
Microwave assisted conversion of carbohydrates and biopolymers to
-hydroxymethylfurfural with aluminium chloride catalyst in water†
5
Sudipta De, Saikat Dutta and Basudeb Saha*
Received 13th May 2011, Accepted 20th July 2011
DOI: 10.1039/c1gc15550d
The common Lewis acid AlCl has efficiently produced 5-hydroxymethylfurfural (HMF) from
3
carbohydrate and biopolymer substrates in water, DMSO, and water–methylisobutylketone
biphasic solvents under microwave irradiation. The yield of HMF in different solvents follows an
increasing order from water to water–MIBK biphasic solvent to DMSO. The yield of HMF
increased with an increase in catalyst loading whereas it remains unchanged upon increase of the
carbohydrate concentration. In most reactions, the maximum yield of HMF is recorded within
2
min of reaction time. The mechanism of the AlCl catalyzed glucose dehydration reaction is
3
proposed to proceed through the isomerization of glucopyranose to fructofuranose, followed by a
proton assisted transformation of fructofuranose to HMF.
◦
8
1
. Introduction
([EMIM]BF
4
) at 100 C. Tungsten salts (WCl
4
, WCl ) acted
6
as Lewis acid catalysts for glucose conversion in 1-butyl-3-
imidazolium chloride [BMIM]Cl solvent and produced 72%
HMF at mild reaction conditions (50 C). In more recent
Irreversible consumption of carbon sources by humankind
resulted in diminishing reserves of fossil fuels and global
warming by CO
of economy to replace fossil-based resources with renewable
◦
9
1
2
emission. This issue has prompted a shift
studies, GeCl
4
in [EMIM]BF
4
◦
catalytic system produced 92.1%
10
HMF from fructose at 100 C. The catalytic conversion of
fructose to HMF with boric acid (B(OH)
2
and sustainable ones. In this context 5-hydroxymethylfurfural
3
) has been reported
(
HMF) has received significant attention as a platform chemical
to produce 46% HMF in biphasic water–methylisobutylketone
for synthesizing a broad range of chemicals and liquid trans-
portation fuels. Carbohydrates constitute 75% of the world’s
renewable biomass and cellulose. The dehydration of fructose
and glucose serve as model reactions for the synthesis of
HMF from biomass-derived carbohydrates since the conversion
to HMF always involves these sugars in the final reaction
sequence. Recently, fructose and glucose have been converted
to HMF in high yields with Cr(II) or Cr(III) halide catalysts
◦
11
(
MIBK) solvent at 150 C. The lanthanide salt-based catalytic
3
systems have been reported to produce 24% HMF yield at
40 C in 6 h from glucose. In limited cases, the synthesis of
HMF using mixed Brønsted and Lewis acid catalysts and the
microwave-assisted reactions from cellulose and sugar deriva-
tives were reported to be successful. Among the reports on the
microwave (MW) assisted green synthesis of HMF, concentrated
HCl catalyzed dehydration of fructose produced 63% HMF in
water. However, the selectivity of HMF decreased when the
reaction was carried out for more than 1 min. Microwave
assisted glucose dehydration with butyl-3-imidazolium chloride
4
◦
12
1
13
14
5
6
in imidazolium ionic liquids. In recent years, research efforts
for developing sustainable methods of HMF production from
various carbohydrates with Lewis acid catalysts have been
limited to the use of ionic liquid solvents. Among these efforts,
HMF synthesis from carbohydrates, including glucose, was
accomplished by various catalytic systems. A catalytic system
consisting of Cr(II) or Cr(III)-chlorides and bulky N-heterocyclic
carbene ligands produced high yields of HMF from fructose
and glucose. A SnCl
produced 60% HMF in 1-ethyl-3-imidazolium tetrafluroborate
14a
[
BMIMCl]/CrCl
3
(10 mol%) catalyst produced 71% HMF in
0
.5 min. The yield of HMF decreased to 67% upon continuing
14d
the reaction for 1 h. However, potential drawbacks of these
methods using ionic liquids is that ionic liquids are expensive
and hence probably not suitable candidates for developing an
economically favourable and scalable HMF production method.
Therefore, it is beneficial to replace expensive ionic liquid sol-
vents with aqueous and/or mixed aqueous biphasic and highly
polar aprotic organic solvents for the sustainable synthesis of
HMF from carbohydrates. Regardless of recent developments
of metal salt catalyzed synthetic methods for HMF production,
it remains highly challenging to develop a sustainable route
for HMF production from carbohydrates and biopolymers
in environmentally benign solvents. In a continuation of our
7
4
catalyzed dehydration reaction of glucose
Laboratory of Catalysis, Department of Chemistry, North Campus,
University of Delhi, Delhi, India. E-mail: bsaha@chemistry.du.ac.in;
Fax: 91 2766 7794; Tel: 011-2766 6646
†
Electronic supplementary information (ESI) available: Results of
HMF yield as a function of fructose concentration, catalyst dosage, reac-
tion time, various aluminium catalysts and UV-Visible and NMR spectra
of isolated HMF products are deposited. See DOI: 10.1039/c1gc15550d
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2859–2868 | 2859

View Article Online
3
Scheme 1 The microwave assisted transformation of fructose or starch to HMF with AlCl catalyst.
ongoing research program in developing Lewis acid and metal
salt catalyzed dehydration of biomass derived sugars and
cellulosic biomass, we have studied the microwave (MW) assisted
technique of HMF synthesis from various sugar derivatives with
Upon completion of the reaction for the set reaction time, the
reactor was opened. The temperature of the reaction mass was
cooled down to room temperature before analyzing the product
using NMR spectroscopic and UV-Visible spectrophotometric
techniques.
In the case of the DMSO solvent mediated reaction, HMF
was isolated from the reaction mixture by distilling out DMSO
at reduced pressure and then extracting the residue as the organic
layer with diethyl ether after adding water into the DMSO
separated residue. The organic layer was collected in a pre-
weighed empty vial and dried under vacuum. After drying off
the solvent, the weight of the vial was recorded. The yield of
isolated HMF was determined by subtracting the weight of the
HMF-containing vial from that of empty vial.
a strong Lewis acid catalyst. AlCl
acid catalyst for HMF synthesis in ionic liquid ([BMIMCl]).
However, the catalytic effectiveness of AlCl
3
is known as an effective Lewis
9
3
for HMF synthesis
has not been reported in aqueous or mixed aqueous solvents
under microwave irradiation. The present paper describes the
transformation of carbohydrates (fructose, glucose, sucrose)
and biopolymers (inulin, starch) into HMF in environmentally
benign solvents with anhydrous and hydrated AlCl
Lewis acid catalysts (Scheme 1).
3
, and AlBr
3
2
. Experimental
Dehydration of carbohydrates using oil-bath heating. The
synthesis of HMF from carbohydrates was also carried out in
a round bottom flask by oil-bath heating. For oil bath heating,
a glass round bottom flask was charged with substrate, solvent
and catalyst, and then refluxed in an oil bath. After completion
of reaction, the reaction mixture was cooled down to room
temperature and analyzed by UV-Visible spectrophotometric
2
.1 Materials and experimental methods
Substrates (fructose, glucose, sucrose, starch and inulin) and
solvents (DMSO, MIBK) were purchased from Sigma Aldrich
and used without further purification. The starch sample, was
dried under vacuum at 100 C for 24 h before use. Anhydrous
and hydrated AlCl
◦
1
3
, AlBr
3
, and Yb(OTf) were purchased from
3
and H NMR spectroscopic techniques.
Sigma Aldrich and used without further purification. SnCl
4
was purchased from Spectrochem, India and used without
further purification. Unless otherwise mentioned, distilled water
Catalyst life-time study
The life-time of the catalyst was studied by recycling the reaction
mixture containing the spent catalyst. Prior to recycling for the
next run, the HMF component was extracted from the reaction
mixture with diethyl ether and fresh substrate was added into
the reaction mixture. Fresh AlCl catalyst was not added to
compensate any loss of the catalyst in the prior runs.
was used as the aqueous phase and anhydrous AlCl was
3
used as catalyst for all reactions. The catalytic conversion of
carbohydrates to HMF was performed in a CEM Matthews
WC Discover Microwave reactor, (model: Discover System,
no. 908010 DV9068) at the standard operating frequency of
3
1
a microwave synthesis reactor (2.45 GHz, power 300 Watt). H
NMR spectra were recorded on a JEOL JNM ECX-400 P 400
MHz instrument and NMR data were processed with JEOL
DELTA program version 4.3.6. Progress of the reaction was
monitored by H NMR spectroscopy in limited cases. HMF
yields were measured by both UV-visible spectrophotometric
Determination of HMF yield
1
The yield of HMF was determined by both H NMR and UV-
Visible spectrophotometric techniques. For H NMR spectro-
scopic 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.
1
1
(
UV-SPECORD 250 analytikjena) and NMR spectroscopic
techniques.
HMF production
1
H NMR method. For quantifying the yield of HMF using
1
H NMR spectroscopic technique, a known concentration of
mesitylene (internal standard) was added into the HMF product
Dehydration of carbohydrate using microwave irradiation
The dehydration reactions of carbohydrates, ca. D-fructose,
glucose, sucrose and biopolymer substrates, were carried out
by charging substrates, solvent and catalyst in a microwave
tube under the reaction conditions mentioned in Tables 1 and
solution in DMSO-d . The percentage of HMF yield was
6
calculated by using the integrated values of the aldehyde proton
(d = 9.58 ppm) of HMF and three aromatic ring protons
of mesitylene (d = 6.79 ppm) (Fig S1†). First, a standard
HMF solution of 99% purity was analyzed for correlating
the percentage of actual and calculated amount of HMF.
2
. The microwave tube was then inserted into the microwave
reactor pre-set at the desired temperature and reaction time.
2
860 | Green Chem., 2011, 13, 2859–2868
This journal is © The Royal Society of Chemistry 2011

View Article Online
Table 1 The AlCl
3
catalyzed dehydration of fructose to HMF under microwave irradiation
HMF Yield (%)
◦
1
Entry
1
Fructose (wt%)
AlCl
3
(mol%)
Solvent
T ( C)
t (min)
H NMR
UV-Vis
Isolated
5
5
5
5
5
5
5
5
5
5
5
5
5
50
50
50
40
25
50
50
50
50
50
50
—
—
50
50
50
50
50
40
40
25
25
25
50
—
Water/MIBK
Water/MIBK
Water
Water
Water
Water
Water
Water
Water
DMSO
DMSO
DMSO
DMSO
Water
Water
Water
Water/THF
Water
Water
Water
Water
130
130
120
120
120
120
120
120
120
100
140
100
140
120
120
120
100
120
120
120
120
120
120
120
120
5
5
5
5
5
15
10
2
20
5
5
5
5
5
5
5
5
0.5
2
15
2
10
15
5
60.1
61.6
51.4
50.6
—
—
—
—
53.3
—
70.1
—
21.4
54.8
51.0
51.4
57.9
—
—
45.7
—
61.2
63.3
53.9
51.5
40.9
55.1
54.6
52.3
55.7
42.9
71.3
3.5
22.0
55.3
52.6
52.1
59.0
47.8
55.1
47.0
38.0
42.3
43.4
53.5
0.8
—
a
2
61.0
50.2
—
—
—
—
—
52.7
—
69.4
—
—
—
50.1
50.8
—
—
—
45.5
—
—
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
5
15
30
5
5
5
5
5
5
5
a
Water
Water
Water
Water
—
42.8
51.2
—
—
51.9
—
10
5
5
a
NaCl (50 mg). For biphasic solvents aqueous : organic = 1 : 2 volume ratio (mL) was used.
Table 2 Conversion of fructose to HMF catalyzed by AlCl
3
under oil-bath heating
HMF Yield (%)
1
Entry
1
2
3
4
Fructose (wt%)
AlCl
3
(mol%)
Solvent
H NMR
UV-Vis
Isolated
5
5
5
5
50
50
50
50
Water/MIBK (2/8)
Water/MIBK (2/8)
Water
32.2
33.5
13.0
—
33.4
36.1
14.1
30.7
—
33.4
—
a
DMSO
—
a
◦
NaCl (100 mg), T = 120 C, t = 60 min.
Once a good correction was established, the extracted HMF
product samples were run and the percentage of HMF yield was
calculated.
conditions, such as the nature of solvents, catalysts, catalyst
concentrations, nature of substrates, substrate concentrations,
reaction time etc. The purpose for designing these experiments
are: (i) identification of an effective substrate for the synthesis of
HMF in an aqueous medium and (ii) optimization of reaction
conditions for maximizing HMF yield. The effect of several
reaction parameters on HMF yield and catalysts effectiveness
are described below.
UV-Visible spectrophotometric method. The UV-Visible
spectrum of pure HMF solution (Fig. S2†) has a distinct
peak at 284 nm with a corresponding extinction coefficient
4
-1
-1
(
e) value of 1.66 ¥ 10 M cm . The percentage of HMF in
each of the reaction products, as tabulated in Table 1, was
calculated from the measured absorbance values at 284 nm and
the extinction coefficient value. Repeated measurement of the
same solution showed the percentage of error associated with
this measurement was ±3%. The yield of HMF obtained from
3.1 Effect of solvents on HMF yield
The dehydration reactions of carbohydrates (fructose, glucose,
1
two different methods ( H NMR and UV-Visible) for the same
sucrose) with the AlCl catalyst were carried out in water, DMSO
3
reaction product agreed very well and the result was within ±5%
error (Table 1).
and water–MIBK biphasic solvents under microwave irradiation
to investigate the effect of different solvents on HMF yield.
Details of reaction conditions and the corresponding HMF
yields are summarized in Tables 1 and 2. Under identical reaction
conditions, water, water–MIBK and DMSO solvent mediated
fructose dehydration reactions produced 53.9%, 61.2% and
71.3% HMF (entries 1, 3, 11 in Table 1), respectively. The higher
3
. Results and discussion
Several experiments were designed for studying the catalytic de-
hydration of carbohydrates to HMF under various experimental
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2859–2868 | 2861

View Article Online
Table 3 Results of glucose dehydration catalyzed by AlCl
3
under MW irradiation
Solvent
Water
Water/MIBK
Water/MIBK
DMSO
DMSO
Water
Water
Water
Water
Water
Water
Water
◦
Entry
AlCl
3
(mol%)
T ( C)
t (min)
HMF yield (%)
1
2
50
50
50
50
50
50
50
50
50
33
33
20
120
130
130
140
140
120
120
120
120
120
120
120
5
5
5
5
5
37.3
43.7
43.0
6.4
52.4
35.1
39.0
39.8
40.3
30.2
33.6
25.4
a
3
4
5
6
7
8
9
1
1
1
2
10
15
20
5
15
5
0
1
2
a
NaCl (50 mg), Glucose = 5 wt%.
HMF yield in DMSO solvent in comparison to that in aqueous
medium for all carbohydrate substrates can be attributed to the
3
Table 4 Results of 30 wt% fructose dehydration with AlCl catalyst at
variable catalyst loading
(
i) higher microwave absorbing ability of DMSO, as measured
a
Fructose (wt%)
AlCl
3
(mol%)
HMF yield (%)
by the higher tangent value of DMSO (tan d 0.825) than water
(
occur in aqueous medium and (iii) catalytic ability of DMSO
in converting fructofuranose to HMF via a 5-membered cyclic
15
tan d 0.123), (ii) avoidance of the side reactions in DMSO that
1
2
3
4
5
30
30
30
30
30
50
40
33
25
20
52.1
48.6
46.0
38.8
35.8
16
◦
mechanism at a higher temperature (150 C) without any added
17
catalyst. These positive characteristics of DMSO accounted
for its use as a solvent so that a comparison of HMF yields in
different solvents can be drawn. The lower HMF yield in water
is due to the rehydration of HMF and the formation of levulinic
and formic acids as side-products.
◦
Solvent = water (2 mL), T = 120 C, t = 5 min, MW.
3
.2 Effect of the starting fructose concentration on HMF yield
On the other hand, the higher HMF yield in water–MIBK
biphasic solvent than in water is due to the driving force of
the biphasic mixture in which HMF accumulates in the organic
phase after its formation in the aqueous phase. This method of
HMF synthesis in a biphasic solvent also produced pure HMF,
The literature report suggests that HMF yield and selectivity
depend on the starting fructose concentrations. Humin is one of
the side reaction products of the fructose dehydration reaction.
The percentage of humin formation is reported to increase with
an increase in starting fructose concentrations, and thereby
lower the desired HMF yield. The formation of humin can be
as high as 35% for 18 wt% fructose solution and 20% for 4.5
wt% fructose solution in water. Therefore, a significant loss of
HMF yield has been reported in aqueous medium. To study this
effect, the starting fructose concentrations were varied from 5
1
as evidenced from the clean H NMR spectrum of the HMF
product collected from the MIBK layer (Fig. S3†).
18
It is reported that the salting-out effect has a positive
influence in increasing HMF yield in a biphasic solvent mediated
dehydration reaction. This effect was observed in boric acid
catalyzed fructose dehydration in aqueous–organic biphasic
medium upon addition of NaCl and similar alkali metal
19
wt% to 30 wt%. Under identical reaction conditions of 50 mol%
◦
of AlCl
3
and at 120 C in water, the yield of HMF changed
11
salts. To check the salting-out effect in the present study, the
dehydration reactions of 5 wt% fructose and 5 wt% glucose
from 53.9% to 52.1% (Fig. 1(a)). Thus, HMF yield remain
unchanged upon increase in fructose concentrations from 5 wt%
to 30 wt%. Additional experiments for the dehydration of highly
were carried out with an AlCl catalyst in biphasic aqueous–
3
MIBK solvent using NaCl salt. The results as tabulated in
Table 1 (entries 1 and 2), Table 2 (entries 1 and 2) and Table
◦
concentrated fructose (30 wt%) at 120 C under microwave
assisted heating (Table 4) showed an appearance of insoluble
brown material in the solution. Attempts were made to analyze
3
(entries 2 and 3) suggested that the effect of NaCl on HMF
yields was insignificant. A comparison of the results between
entry 2 of Table 1 and entry 2 of Table 2 also revealed
that HMF formation from the oil bath heating experiment
for the optimized reaction time was significantly lower (36%)
than that of microwave assisted heating (63%) experiment. To
further check the salting-out effect, additional experiments were
1
the brown material using HMR spectroscopy; however, due
to poor solubility of this brown material in most deuterated
1
solvents, signals were not intense (Fig. S4†). Inspection of the H
NMR spectrum in DMSO-d confirms the presence of signals
6
corresponding to two furan ring protons and a –CHO group
proton (d 9.50 ppm). This brown insoluble material is believed to
performed for a 30% fructose dehydration reaction with an AlCl
catalyst in the presence of Na SO , NaBr, and NaNO salts. The
3
20
be humin as mentioned in the literature. Thus, the probability
2
4
3
of loss of HMF to the form of humin may be high when the
results confirmed no significant effect of these salts on HMF
yield in biphasic aqueous–MIBK solvent.
reaction starts with concentrated fructose solution (ca. 30 wt%)
20
in aqueous medium.
2
862 | Green Chem., 2011, 13, 2859–2868
This journal is © The Royal Society of Chemistry 2011

View Article Online
Fig. 1 (a) The effect of HMF yields on (a) fructose concentrations at fixed AlCl
dehydration in water at 120 C MW heating for 5 min.
3
loading (50 mol%), and (b) AlCl
3
catalyst dosage for 5 wt% fructose
◦
3
.3 Effect of the catalyst dosage on HMF yield
The dehydration reaction of fructose to HMF was studied
at variable catalyst concentrations for optimizing the reaction
conditions and maximizing the HMF yield. The concentrations
of AlCl were varied from 20 mol% to 50 mol% while other
3
reaction parameters were constant at [fructose] = 5 wt%, T =
◦
1
20 C, reaction time = 5 min and solvent = water. As shown in
Fig. 1(b), the yield of HMF increased from 37.2% to 53.9% with
an increase in AlCl concentration from 20 mol% to 50 mol%.
3
The high catalyst loaded experiments improved HMF yield,
meaning that the kinetic of fructose dehydration is fast in the
case of higher catalyst loading. It has been shown in section 3.2
that HMF yields remain unchanged upon increase in starting
fructose concentration from 5 wt% to 30 wt%. This prompted
us to further study the catalyst variation experiments using
concentrated fructose solution (ca. 30 wt%). The results, as
tabulated in Table 4, show the yield of HMF increased from
Fig. 2 The effect of reaction time on HMF yield for the conversion
of carbohydrates (fructose, glucose, sucrose) to HMF in water under
◦
microwave heating at 120 . The concentrations of carbohydrate and
3
5.8% to 52.1% upon increase in catalyst loading from 20 wt%
to 50 wt% (Table 4). Thus, the trend of HMF formation against
AlCl loading for 30 wt% fructose dehydration was similar to
that obtained for 5 wt% fructose dehydration.
3
AlCl are 5 wt% and 50 mol%, respectively.
3
and sucrose dehydration reactions showed 5% and 7% increase
in HMF yield during 2–20 min of reaction time, respectively.
It is important to note that the color of the reaction mixture
gradually turned to deep brown from pale yellow when the
dehydration of fructose was continued for a longer time (20
min). Perhaps, the color change of the reaction mixture is
associated with the decomposition of the product HMF into
side products. To further investigate the HMF decomposition
3
.4 Effect of reaction time on HMF yield
The reaction time of the dehydration reaction was varied to
study the rate of HMF formation as a function of time. As
plotted in Fig. 2, the yield of HMF improved from 47.8%
to 52.3% upon increasing the duration of the reaction from
1
and end product, the reaction mixture was analyzed by H NMR
1
spectroscopy. A H NMR spectral analysis of the vacuum dried
3
0 s to 2 min for AlCl
3
catalyzed dehydration of fructose at
reaction mixture in chloroform-d
1
confirmed the formation 5-
◦
22
120 C. Further increase in reaction time to 20 min under similar
chloromethylfurfural (CMF) and levulinic acid as side prod-
ucts (Fig. 3). The formation of levulinic acid can be explained by
conditions resulted in 55.7% HMF yield. Thus, the kinetics of
HMF formation are rapid in first 0.5 min followed by a slow
reaction. This behavior of increasing HMF yield with an increase
in reaction time agreed well with the trend observed for fructose
dehydration in water and water–DMSO solvent using mineral
acid and an ion-exchange resin catalyst under MW heating.
A similar trend of slow HMF yield after 0.5 min of reaction time
was observed for AlCl catalyzed dehydration of glucose and
20,23
the rehydration of HMF.
The self-polymerization of HMF
molecules or cross-polymerization of HMF with fructose is also
possible, particularly due to the fact that the HMF yield is
20
not significantly increased upon prolonged reaction. Although
HMF rehydration can be suppressed in non-aqueous solvents,
considerable rehydration of HMF is possible in aqueous or
mixed aqueous solvents. This hypothesis supports the literature
data where a lower HMF yield has been reported in aqueous
14a,21
3
sucrose as well. Under comparable reaction conditions, glucose
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2859–2868 | 2863

View Article Online
1
Fig. 3 A representative H NMR (CDCl
AlCl
3
) spectrum of the reaction product obtained from the dehydration reaction of 5 wt% fructose with 50 mol%
◦
3
in water under microwave heating at T = 120 C for 15 min reaction time.
24
medium than in biphasic solvent. The integration of proton
1
signals corresponding to the –CHO group in H NMR spectra
(
Fig. 3) of the reaction mixture suggests the formation of
CMF occurs in a 1 : 2 molar ratio of CMF : HMF in water.
The formation of CMF can be attributed to the ability of
AlCl to chlorinate the intermediate species formed during the
3
dehydration of fructose.
As discussed above, detection of levulinic acid as a by-product
1
in the H NMR spectrum of the reaction mixture in the water
mediated reaction suggested the formation of formic acid as well.
However, the signal for the –COOH group proton of formic acid
(
d 8.05 ppm) was too small in comparison to the levulinic acid
signal. This could be due to the fact that the –COOH proton
signal of formic acid did not resolve well in a reaction mixture or
disappeared due to its consumption in undefined side reactions.
In order to investigate the role of formic acid in fructose
dehydration, a control experiment was performed for the
Fig. 4 The effect of aluminum catalysts and reaction time on HMF
yields for the dehydration of 5 wt% fructose with 50 mol% catalysts in
◦
water under microwave heating at 120 C.
dehydration of 5 wt% fructose with 0.56 mmol formic acid at
◦
1
20 C for 5 min under microwave-assisted heating. This control
produced about 2–7% less HMF than its anhydrous analogue
experiment with formic acid did not produce HMF, which
precludes the possibility of the participation of a formic acid
by-product as a catalyst in the fructose dehydration reaction.
(
Fig. 4), which could be due to the contribution of extra water
from the hydrated salt.
3
.5 Effect of other aluminium catalysts
3.6 Conversion of glucose and sucrose
The dehydration reaction of fructose to HMF was also inves-
tigated with two other commercially available aluminium salts,
Although fructose has been the preferred feedstock for HMF
production, its occurrence in nature is limited. This drives the
attention to utilize more abundant carbohydrates, glucose and
sucrose, as raw materials for HMF synthesis. A recent study
demonstrates that the conversion of glucose to HMF is likely
to proceed via consecutive steps, namely, mutarotation and
isomerization of glucose into fructose followed by dehydration
AlBr
Fig. 4) show that AlBr
production, which produced 35.1% HMF in 5 min as compared
to 53.9% with the anhydrous AlCl catalyst. Under comparable
reaction conditions, the hydrated aluminium chloride catalyst
3
and hydrated aluminium chloride AlCl
3
·6H
2
O. The results
(
3
is a less active catalyst for HMF
3
2
864 | Green Chem., 2011, 13, 2859–2868
This journal is © The Royal Society of Chemistry 2011

View Article Online
3
Scheme 2 The proposed mechanism for glucose dehydration to HMF with the AlCl catalyst in water.
Table 5 Results of sucrose dehydration catalyzed by AlCl
3
under MW
To understand the mechanism of glucose dehydration reac-
irradiation
1
tion, a H NMR study of a reaction mixture of glucose and AlCl
in DMSO-d was carried out at room temperature. It shows a
3
◦
6
Entry AlCl
3
(mg)
Solvent T ( C) t (min)
HMF yield (%)
line broadening of –OH proton signals of glucose in the range
1
2
3
4
5
6
7
8
9
1
50
50
50
50
50
50
50
33
33
20
Water
Water
DMSO
DMSO
Water
Water
Water
Water
Water
Water
120
120
140
140
120
120
120
120
120
120
5
10
5
5
2
15
20
5
15
5
30.0
30.4
4.6
42.5
26.7
31.7
33.2
26.0
27.8
20.6
of d 4–5 ppm. This line broadening of –OH protons indicates
a H-bonding interaction between a Cl atom of AlCl
3
and the
–
OH protons of glucose forming an AlCl –glucose adduct. The
3
line broadening of the –OH protons corresponding –H ◊ ◊ ◊ Cl–
◦
interaction completely disappeared at 100 C (Fig. S5(d)†) and
1
a new signal appeared at d 9.49 ppm in the H NMR spectra
◦
after heating the reaction mixture at 100 C for 40 min. The
0
new signal at d 9.49 ppm corresponds to the –CHO proton of
HMF. This experiment suggests that the dehydration of glucose
in DMSO occurs via the formation of an intermediate species
Sucrose = 5 wt%
between the –OH proton and the Cl atom of AlCl
The dehydration reaction is believed to be initiated by the
hydrolysis of AlCl . The hydrolysis of AlCl in water under mi-
crowave irradiation forms a cationic species [Al(OH)(H ] .
3
.
6
of fructose to HMF. In the present study, the conversion of
glucose and sucrose to HMF was carried out with anhydrous
3
3
2
+
2
O)
O) ] acts as a
5
◦
◦
+
AlCl
3
catalyst in water at 120 C and in DMSO at 140 C under
This active form of the catalyst, [Al(OH)(H
2
5
MW heating. The yield of HMF was monitored as a function
of reaction time in the range of 2–20 min. The results, as shown
in Tables 3 and 5, reveal the maximum HMF yield from glucose
and sucrose are 52.4% (entry 5 in Table 3) and 42.5% (entry 4
in Table 5), respectively, in DMSO at 140 C for 5 min reaction
time. Under similar conditions, the HMF yields in water are
potential electrophile that reacts with a-glucopyranose to form
the intermediate A (Scheme 2). Intermediate A is possibly
converted to ketohexose B via a hydride transfer, which forms
fructofuranose through cyclization. This proposed mechanistic
route for the formation of ketohexose B is similar to the
◦
25
mechanism described by Davis et al. using a SnCl catalyst.
4
3
7.3% and 30.0% from glucose and sucrose, respectively. The
The mechanism of HMF formation from fructose has been a
topic of debate over the years. In our case, the transformation of
fructofuranose to intermediate compound E occurs via a cyclic
glucose dehydration reaction in water–MIBK biphasic solvent
produced 43% HMF as compared to 37.3% in water.
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2859–2868 | 2865

View Article Online
mechanism assisted by a proton that is generated during AlCl
3
fructose was added into the reaction mixture. Fig. 5 shows that
the loss of activity of the catalyst in terms of HMF yield is only
10% after five cycles (53.9% in first cycle versus 44% in fifth
cycle).
-
hydrolysis. It is of note that the possibility of Cl acting as a
base was taken into account for abstracting the proton from
the oxonium ion forming structure D (path A), as proposed by
26
-
Raines et al. In path B, Cl functions as both the nucleophile
and base to form intermediate D. The cyclic intermediate E
17
has been identified recently by NMR study, which supports
our proposed mechanism for the formation of E (Scheme 2). A
similar mechanism involving intermediate E via a cyclic pathway
was also proposed by Riisager et al. in a fructose dehydration
27
reaction with boric acid catalyst. Intermediate compound E
then releases a water molecule to form the final product HMF.
This proposed mechanism was similar to a mechanism
previously described for SnCl catalyzed dehydration of glucose
4
in mixed [EMIM]BF
and DMSO-d
4
(EMIM = 1-ethyl-3-methylimidazolium)
solvents (9 : 1 w/w). Nevertheless, to understand
8
6
the role of other cations in the present reaction, a mechanistic
investigation for glucose (5 wt%) dehydration with SnCl catalyst
4
1
(
20 mol%) was carried out in DMSO-d
6
(0.5 mL). H NMR
◦
spectra revealed the formation of HMF upon heating at 80 C
Fig. 5 Catalyst recyclability study for dehydration of fructose with
for 45 min, when the signals of –OH proton corresponding
AlCl
3
catalyst in water. Reaction conditions: fructose = 5 wt%, AlCl
3
=
◦
–H . . . Cl– interactions (interaction between –OH proton and
50 mol%, T = 120 C, t = 5 min.
Cl atom of SnCl
Fig. S6†). Further heating at 120 C for 30 min ensured signifi-
cant conversion of glucose to HMF, as revealed from H NMR
4
) disappeared. (see supporting information,
◦
1
3.8 Conversion of starch and inulin
(
Fig. S6(d)†). Attempts were made to investigate the mechanism
of glucose dehydration with a lanthanide cation, Yb(OTf) (Tf =
. H NMR
The synthesis of HMF from biopolymers, such as starch and
inulin, was carried out using an anhydrous AlCl catalyst
3
1
3
trifluromethanesulfonate) as a catalyst in DMSO-d
investigation confirmed the low conversion of glucose into HMF
in the presence of Yb(OTf)
6
under microwave irradiation. Table 6 shows details of reaction
conditions for HMF synthesis from biopolymers. However,
prerequisite for successful dehydration of starch to HMF is
the facile depolymerization of biopolymeric units. The depoly-
merization of the biopolymeric unit is reported to follow first-
◦
3
at 120 C for 1 h (Fig. S7†).
This result is consistent with a previously reported Yb(OTf)
catalyzed glucose dehydration reaction in ionic liquid solvent
BMIM]Cl (BMIM = 1-butyl-3-methylimidazolium) where a
3
[
24
order kinetics with respect to catalyst concentration. In the
present work, we used AlCl to hydrolyze the glycosidic linkages
between monomeric unit. AlCl is believed to be effective for
◦
12
maximum 24% HMF yield was realized at 140 C for 6 h.
3
3
3
.7 Catalyst life-time in water medium
hydrolysis of glycosidic linkages because of its strong Lewis
acidic character. As shown in Table 6, a maximum 29.6%
The life-time of the AlCl catalyst was studied for fructose
3
dehydration reaction in aqueous medium by recycling the
reaction mixture containing the spent catalyst. The life time
HMF was obtained from dehydration of starch with AlCl
3
◦
catalyst in DMSO at 140 C under MW heating. The same
reaction in aqueous solvent produced about 21% HMF, which
is clearly less than that obtained in DMSO. The effect of solvent
on the HMF yield has been discussed in section 3.1. The
dehydration of starch to HMF has been reported in the literature
study of the catalyst was carried out with 5 wt% fructose and
◦
5
0 mol% AlCl
3
in water at 120 C for 5 min under microwave
heating. Prior to recycle for the next run, HMF component was
extracted from the reaction mixture with diethyl ether and fresh
3
Table 6 Results of starch and inulin dehydration catalyzed by AlCl under MW irradiation
◦
Entry
Substrate (5 wt%)
AlCl
3
(mol%)
Solvent
T ( C)
t (min)
HMF yield (%)
1
2
3
4
5
6
7
8
9
Starch
Starch
Starch
Starch
Starch
Starch
Starch
Inulin
Inulin
Inulin
Inulin
50
50
50
50
50
50
50
50
50
50
50
Water
Water
Water/MIBK
DMSO
Water
Water
Water
Water
Water
120
120
130
140
120
120
120
120
120
130
140
2
5
5
5
10
5
10
5
10
5
19.6
21.1
24.1
30.6
21.3
14.0
15.2
29.4
31.0
32.5
39.2
1
1
0
1
Water/MIBK
DMSO
5
Aqueous: organic = 1 : 2 volume (mL)
2
866 | Green Chem., 2011, 13, 2859–2868
This journal is © The Royal Society of Chemistry 2011

View Article Online
with SnCl
chloride ([EMIM]Cl) and 1-octyl-3-methylimidazolium chloride
[OMIM]Cl]) solvent, respectively. These reported reactions
produced 47% and 73% HMF with SnCl and CrCl catalysts,
respectively Although the yield of HMF in water is less than
that reported in ionic liquid with different Lewis acid catalysts,
but our reaction in water is more sustainable in terms of avoiding
the use of expensive ionic liquid solvents.
4
and CrCl
2
catalysts in 1-ethyl-3-methylimidazolium
4. Conclusions
In summary, we have demonstrated that AlCl
3
is an excel-
(
lent catalyst for the rapid conversion of carbohydrates and
biopolymers into HMF in water, DMSO and water–MIBK
biphasic solvents under microwave irradiation. The yield of
HMF remains unchanged upon varying the starting fructose
concentrations from 5 wt% to 30 wt%. The optimized reaction
conditions were developed to maximize HMF yield by varying
catalyst loading, reaction time, aluminium salts etc. Satisfactory
results were obtained when glucose, sucrose, starch, and inulin
4
2
28
The direct synthesis of HMF from inulin occurs via hydrolysis
of inulin to fructose followed by dehydration of fructose. The
synthesis of HMF from inulin was carried out with an anhydrous
were used as the substrates. The mechanism of the AlCl
3
AlCl
3
catalyst in three different solvents (water, water–MIBK
for 5 min
catalyzed glucose dehydration reaction is proposed to proceed
through the isomerization of glucopyranose to fructofuranose
followed by a proton assisted transformation of fructofuranose
to HMF. Finally, the Lewis acidic catalyst AlCl deserves further
investigation in less expensive ionic liquids to facilitate sustain-
able conversions of cellulosic materials to value added products.
and DMSO) using 1 : 1 molar ratio of inulin: AlCl
3
microwave heating. The inulin dehydration reaction produced
maximum 39% and 31% HMF in DMSO and water, respectively
3
(
Table 6). In an earlier report, Amberlyst 70 catalyst promoted
hydrolysis and dehydration of inulin produced 60% HMF at
◦
29
1
10 C in [BMIM]Cl/glycerol carbonate (10 : 90). Although
yield of HMF in aqueous solvent is less, but it is more sustainable
and environmentally friendly process..
Acknowledgements
The authors gratefully acknowledge financial support by the
University Grant Commission (UGC), India and the University
of Delhi. SD thanks UGC, India for a DS Kothari Postdoctoral
Research Fellowship.
3
.8 Substrate scope of AlCl
3
catalyst
From the experimental results demonstrated above, we sought to
explore the substrate scope of AlCl catalyst and an estimation
3
of comparative yields of HMF under microwave heating. This
analysis for all substrates showed a trend of decrease in HMF
yield from 53.9% to 21.1% in aqueous medium at 120 C for
References
◦
1
(a) A. J. Ragauskas, C. K. Williams, B. H. Davidson, G. Britovsek, J.
Cairney, W. J. Frederick, J. P. Hallett, D. J. Leak, C. L. Liotta, J. R.
Mielenz, R. Murphy, R. Templet and T. Tschaplinski, Science, 2006,
5
min reaction duration. A similar decreasing trend of HMF
yield is also noted in DMSO for the dehydration reaction of
fructose, glucose, sucrose, inulin and starch (Fig. 6). The yields
of HMF from starch and inulin were comparatively lower than
those obtained from fructose and glucose. However, the direct
synthesis of HMF from starch and inulin certainly is a promising
3
11, 484; (b) G. Stephanopoulos, Science, 2007, 315, 801.
Y. Rom a´ n-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic, Nature,
007, 447, 982.
2
2
3 (a) G. W. Huber, J. N. Chheda, C. J. Barrett and J. A. Dumesic,
Science, 2005, 308, 1446; (b) J. N. Chheda, G. W. Huber and J. A.
Dumesic, Angew. Chem., Int. Ed., 2007, 46, 7164; (c) A. Boisen, T.
Christensen, W. Fu, Y. Gorbanev, T. Hansen, J. Jensen, S. Klitgaard,S.
Pedersen, A. Riisager, T. St a˚ hlberg and J. Woodley, Chem. Eng. Res.
Des., 2009, 87, 1318.
alternative route to HMF production. Nevertheless, the AlCl
3
catalyst is shown to be effective to convert less-expensive and
easily available biopolymers (starch and inulin) to HMF in
aqueous medium under microwave irradiation. This indeed
expands the substrate scope of the catalyst and sustainability.
4
5
A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411.
T. St a˚ hlberg, W. Fu, J. M. Woodley and A. Riisager, ChemSusChem,
2
011, 4, 451.
H. Zhao, J. E. Holladay, H. Brown and Z. C. Zhang, Science, 2007,
16, 1597.
G. Yong, Y. Zhang and J. Y. Ying, Angew. Chem., Int. Ed., 2008, 47,
345.
S. Hu, Z. Zhang, J. Song, Y. Zhou and B. Han, Green Chem., 2009,
1, 1746.
J. Y. Gerentt and Y. Zhang, ChemSusChem, 2009, 2, 731.
6
7
8
9
3
9
1
1
0 Z. Zhang, Q. Wang, H. Xie, W. Liu and Z. K. Zhao, ChemSusChem,
2
011, 4, 131.
1
1
1 T. S. Hansen, J. Mielby and A. Riisager, Green Chem., 2011, 13, 109.
2 T. Stahlberg, M. G. Sorensen and A. Riisager, Green Chem., 2010,
1
2, 321.
3 S. Zhao, M. Cheng, J. Li, J. Tian and X. Wang, Chem. Commun.,
011, 47, 2176.
1
1
2
4 (a) T. S. Hansen, J. M. Woodley and A. Riisager, Carbohydr. Res.,
2
1
009, 344, 2568; (b) Z. Zhang and Z. K. Zhao, Biores. Tech., 2010,
01, 1111; (c) C. Z. Li, Z. H. Zhang and Z. B. K. Zhao, Tetrahedron
Lett., 2009, 50, 5403; (d) X. Qi, M. Watanabe, T. M. Aida and R. L.
Smith Jr., ChemSusChem, 2010, 3, 1071.
1
1
1
5 D. Dallinger and C. O. Kappe, Chem. Rev., 2007, 107, 2563.
6 X. Tong and Y. Li, ChemSusChem, 2010, 3, 350.
7 A. S. Amarasekara, L. D. Williams and C. C. Ebede, Carbohydr. Res.,
Fig. 6 A comparison of HMF yields for dehydration of various
2
008, 343, 3021.
substrates in water and DMSO. Reaction conditions: substrate = 5 wt%,
18 Y. Rom a´ n-Leshkov and J. A. Dumesic, Top. Catal., 2009, 52, 297.
◦
AlCl
3
= 50 mol%, T = 120 C, t = 5 min.
19 B. F. M. Kuster, Starch/Staerke, 1990, 42, 314.
This journal is © The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2859–2868 | 2867

View Article Online
2
2
2
0 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Green Chem.,
009, 11, 1327.
1 X. Zi, M. Watanabe, T. M. Aida and R. L. Smith Jr., Green Chem.,
008, 10, 799.
2 M. Brasholz, K. von K a¨ nel, C. H. Hornung, S. Saubern and J.
Tsanaktsidis, Green Chem., 2011, 13, 1114.
25 Y. Rom a´ n-Leshkov, M. Moliner, J. A. Labinger and M. E. Davis,
Angew. Chem., Int. Ed., 2010, 49, 8954.
26 J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131,
1979.
27 T. St a˚ hlberg, S. Rodriguez-Rodriguez, P. Fristrup and A. Riisager,
Chem.–Eur. J., 2011, 17, 1456.
2
2
2
2
3 F. S. Asghari and H. Yoshida, Ind. Eng. Chem. Res., 2007, 46, 7703.
4 (a) R. Rinaldi, R. Palkovits and F. Sc u¨ th, Angew. Chem., Int. Ed.,
28 J.-A. Chun, J.-W. Lee, Y.-B. Yi, S.-S. Hong and C.-H. Chung,
Starch/Staerke, 2010, 62, 326.
29 M. Benoit, Y. Brissonnet, E. Guelou, K. D. Oliveira Vigier, J. Barrault
and F. Jerome, ChemSusChem, 2010, 3, 1304.
2
008, 47, 8047; (b) R. Rinaldi, N. Meine, J. vom Stein, R. Palkovits
and F. Sc u¨ th, ChemSusChem, 2010, 3, 266.
2
868 | Green Chem., 2011, 13, 2859–2868
This journal is © The Royal Society of Chemistry 2011