Fast transformation of glucose and di-/polysaccharides into 5-hydroxymethylfurfural by microwave heating in an ionic liquid/catalyst system

Article abstract of DOI:10.1002/cssc.201000124

An efficient method for converting glucose into 5-hydroxymethylfurfural (5-HMF), in the presence of CrCl₃ catalyst, is developed by using the ionic liquid 1-butyl-3-methyl imidazolium chloride as solvent. A 5-HMF yield of 71% is achieved in 30 s for 96% glucose conversion with microwave heating at 140°C. The activation energy of glucose conversion is determined to be 114.6 kJ mol^-1, with a pre-exponential factor of 3.5A? - 10 ¹⁴ min^-1. Fructose, sucrose, cellobiose, and cellulose are studied and 5-HMF yields of 54% are obtained for cellulose conversion at 150°C during 10 min of reaction time. Recycling of the ionic liquid and CrCl₃ is demonstrated with six cycles of use.

Full text of DOI:10.1002/cssc.201000124

DOI: 10.1002/cssc.201000124
Fast Transformation of Glucose and Di-/Polysaccharides
into 5-Hydroxymethylfurfural by Microwave Heating in an
Ionic Liquid/Catalyst System
Xinhua Qi,*[a, b] Masaru Watanabe,[b] Taku M. Aida,[b] and Richard L. Smith, Jr[b]
An efficient method for converting glucose into 5-hydroxyme-
thylfurfural (5-HMF), in the presence of CrCl3 catalyst, is devel-
oped by using the ionic liquid 1-butyl-3-methyl imidazolium
chloride as solvent. A 5-HMF yield of 71% is achieved in 30 s
for 96% glucose conversion with microwave heating at 1408C.
The activation energy of glucose conversion is determined to
be 114.6 kJmolÀ1, with
a pre-exponential factor of 3.5ꢀ
1014 minÀ1. Fructose, sucrose, cellobiose, and cellulose are stud-
ied and 5-HMF yields of 54% are obtained for cellulose conver-
sion at 1508C during 10 min of reaction time. Recycling of the
ionic liquid and CrCl3 is demonstrated with six cycles of use.
Introduction
With growing concerns about global warming, the search for
sustainable alternative energy sources is critically important.
Biomass resources can possibly supplement our fuel supply
and meet our demands for chemicals, provided that efficient
transformation routes can be developed. 5-Hydroxymethylfur-
fural (5-HMF), for example, is an important intermediate for the
production of fine chemicals, pharmaceuticals, and biofuels,[1]
and its preparation through the dehydration of biomass-de-
rived sugars has received much attention.[2–9] Many researchers
have selected fructose as substrate for the acid-catalyzed pro-
duction of 5-HMF, and high yields of 5-HMF, up to 96%, have
been achieved.[3,4,7,10–13] On the other hand, glucose (an isomer
of fructose) is a better candidate as a resource for 5-HMF be-
cause it is the monomeric unit in cellulose and because it is
the most abundant monosaccharide in nature. Thus, finding ef-
ficient routes for converting glucose into 5-HMF constitutes an
important research topic.
Binder et al. developed a process that converted glucose to 5-
HMF in N,N-dimethylacetamide in the presence of NaBr and
CrCl2, and achieved a 5-HMF yield of 81% at 1008C with a re-
action time of 5 h.[17] Hu et al. were able to convert glucose
into 5-HMF in 1-ethyl-3-methylimidazolium tetrafluoroborate
([EMIM][BF4]) by using SnCl4 as catalyst, and obtained a 5-HMF
yield of ca. 60% at 1008C in 3 h.[16] Although high 5-HMF
yields (68–81%) from glucose were obtained at moderate reac-
tion temperatures of ca. 1008C, all of the proposed methods
have required relatively long reaction times, from 3 to 6 h, and
up to this point kinetic parameters are lacking. Li et al. present-
ed a process for the production of 5-HMF with high yields
from glucose and cellulose in an ionic liquid under microwave
irradiation, but the experimental conditions such as reaction
temperature and the detailed analyses were not well de-
fined.[18]
Herein, we investigate the catalytic conversion of glucose
into 5-HMF in the ionic liquid 1-butyl-3-methyl imidazolium
chloride ([BMIM][Cl]) by using CrCl3 as catalyst and applying
microwave heating. Compared to the strongly reductive CrII,
the trivalent form CrIII possesses a higher environmental stabili-
ty. In addition, although CrIII and CrVI are both chromium spe-
cies that are stable in the environment, their biological activity
and chemical reactivity are very different.[19] CrVI is a toxic com-
A reaction scheme describing the transformation of glucose,
fructose, and other di-/polysaccharides is shown in Scheme 1.
Glucose can be used as-is, or it can be isomerized to fructose.
However, it has been shown that it is difficult to convert glu-
cose into 5-HMF (yields <30%) in water,[13] organic,[14] and or-
ganic/water mixed solvents.[6] The apparent reason is that glu-
cose tends to form a stable six-membered pyranoside structure
that has a low enolization rate.[15]
Zhao et al. proposed an effective conversion technique for
transforming glucose into 5-HMF by using an ionic liquid sol-
vent (1-ethyl-3-methylimidazolium chloride) and CrCl2 as cata-
lyst.[5] A 5-HMF yield of 68% was obtained at a temperature of
1008C in a reaction time of 3 h. Several other papers have re-
ported the conversion of glucose into 5-HMF by using ionic
liquids as solvents and chromium(II) chloride as catalyst.[2,16,17]
Yong et al. studied the production of 5-HMF from glucose in 1-
butyl-3-methyl imidazolium chloride ([BMIM][Cl]) using NHC/
CrCl2 (NHC: N-heterocyclic carbene) as catalyst, and a 5-HMF
yield of 81% was achieved at 1008C in a 6 h reaction time.[2]
[a] Dr. X. Qi
Key Laboratory of Pollution Processes and Environmental Criteria
Ministry of Education, College of Environmental Science and Engineering
Nankai University
Tianjin 300071 (PR China)
[b] Dr. X. Qi, Dr. M. Watanabe, Dr. T. M. Aida, Prof. R. L. Smith, Jr
Research Center of Supercritical Fluid Technology
Tohoku University, 6-6-11 Aoba, Aramaki, Aoba-ku
Sendai 980-8579 (Japan)
Fax: (+81) 022-795-5864
E-mail: qixinhua@scf.che.tohoku.ac.jp
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201000124.
ChemSusChem 2010, 3, 1071 – 1077
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1071

X. Qi et al.
Scheme 1. Reaction scheme for the transformation of glucose, fructose, and other di-/polysaccharides.
pound with teratogenic and carcinogenic effects on humans,
while CrIII is essential to mammals for removing glucose from
the bloodstream.[20] In this work, the [BMIM][Cl]/CrCl3 system
was found to be suitable for the conversion of biomass sub-
strates such as glucose, fructose, sucrose, cellobiose, and cellu-
lose into 5-HMF.
Results and Discussion
Effect of the reaction temperature and kinetics analysis
In our previous work, the ionic liquid [BMIM][Cl] was found to
be a suitable solvent for the acid-catalyzed dehydration of fruc-
tose into 5-HMF.[10] In this work, we examined the conversion
of glucose to 5-HMF in [BMIM][Cl] in the presence of a
chromium(III) chloride catalyst.
Figure 1a and b show the glucose conversion and 5-HMF
yield versus reaction time, respectively, for the range of tem-
peratures studied in this work. It is apparent that the reaction
temperature has a large effect on the glucose conversion and
5-HMF yield. When the reaction temperature was 908C and the
reaction time was 60 min, the conversion of glucose approach-
ed 53% and the yield of 5-HMF was about 40%. In a reaction
time of 10 min, the 5-HMF yield increased to about 69% when
the temperature was increased to 1208C. At a reaction temper-
ature of 1408C, a 5-HMF yield of 71% was obtained in 0.5 min.
Yong et al., in their study on the conversion of glucose into 5-
HMF in [BMIM][Cl] with NHC/CrCl2, reported that the 5-HMF
yield from glucose increased with temperature below 1008C,
Figure 1. Effect of reaction temperature on a) glucose conversion and b) 5-
HMF yield in a system using [BMIM][Cl] as solvent and CrCl3 as catalyst. Con-
ditions: 0.05 g glucose and 0.015 g CrCl3·6H2O were dissolved in 1 g [BMIM]
[Cl] at 708C for 10 min (water bath heating).
but then decreased with increasing temperature from 1008C
to 1408C.[2] However, such a decrease of the 5-HMF yield with
increasing temperature (for conditions below 1408C) was not
observed in our experiments.
The reaction kinetics for the conversion of glucose into 5-
HMF were fitted with first- and second-order reaction rate
equations. First-order kinetics were found to be more applica-
ble, and the rate constants at different temperatures (T) were
used to make an Arrhenius plot (Figure 2). The activation
energy and pre-exponential factor were determined to be
114.6 kJmolÀ1 and 3.5ꢀ1014 minÀ1, respectively. The value of
the activation energy of this work is comparable with the
value given by Antal and Mok, who reported an activation
1072
www.chemsuschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2010, 3, 1071 – 1077

Fast Transformation of Glucose and Di-/Polysaccharides into 5-HMF
For comparison, the reaction was carried out with oil-bath
heating and microwave heating at identical conditions for two
temperatures (Table 1). Microwave heating gave yields that
were roughly 48% higher than oil-bath heating. That higher
yields can be obtained with microwave heating than with oil-
bath heating for this reaction and catalyst system was also re-
ported by Li et al., who obtained a 91% 5-HMF yield with mi-
crowave irradiation at 400 W in a reaction time of 1 min and
under catalytic conditions similar to this work.[18] An unknown
reaction temperature was used by those researchers, and anal-
yses were made by UV/Vis at 282 nm. The results of Li et al.
could not be reproduced in this work. The Supporting Informa-
tion provides other experimental runs and shows clearly sepa-
rated ionic liquid, reactant, and product peaks (by HPLC, with
refractive index detection). Data in the Supporting Information
of this work shows that increasing the temperature or reaction
time gave lower yields than our results in Table 1 at 1408C. Re-
sults obtained on the conversion of glucose into 5-HMF with
Figure 2. Arrhenius plot for glucose conversion into 5-HMF in ionic liquid
[BMIM][Cl] solvent. First-order reaction kinetics constants at different temper-
atures were obtained for glucose conversion (X) at a certain reaction time (t)
via ln(1ÀX) vs. t plots.
energy of 100 kJmolÀ1 for the
dehydration of glucose at sulfu-
ric acid concentrations of
5 mmolLÀ1 in water.[21] Because
no previous kinetic analyses on
the dehydration of glucose into
5-HMF were found in the litera-
ture, we chose pre-exponential
factors for the conversion of
fructose into 5-HMF for compari-
son. The pre-exponential factors
that have been reported show a
wide variation, from 4.8ꢀ106 to
8.7ꢀ1011 minÀ1, depending on
the solvents, catalysts, and even
heating methods used.[3] The
Table 1. Comparison of 5-HMF yields for ionic liquid-catalyst systems.
Solvent
Catalyst, amount[a]
T [8C]
t
5-HMF
Yield [%]
Heating
Ref.
method[b]
[EMIM][Cl]
[BMIM][Cl]
([EMIM][BF4]
[BMIM][Cl]
CrCl2, 6 mol%
100
100
100
Unknown[c]
100
3 h
6 h
3 h
1 min
1 h
68
81
60
91
17
67
67
45
71
48
Oil bath
Oil bath
Oil bath
MW
Oil bath
MW
MW
Oil bath
MW
[5]
[2]
[16]
[18]
[18]
This work
NHC-CrCl2, 9 mol%
SnCl4, 10 mol%
CrCl3, 9 mol%
[BMIM][Cl]
CrCl3, 10 mol%
100
1 h
120
120
140
140
5 min
5 min
0.5 min
0.5 min
Oil bath
[a] Based on glucose. [b] MW: Microwave heating. [c] The sample was treated under MW irradiation at 400 W
for 1 min, and the reaction temperature was unknown.
pre-exponential factor determined in this work is 3.5ꢀ
1014 minÀ1, which is 3–8 orders of magnitude larger than the
values of previous works. Under microwave irradiation, the ef-
fective collision among reactant molecules in the homogene-
ous phase is probably enhanced, and as a result the pre-expo-
nential factor in the Arrhenius law increases.[22]
ionic liquids are summarized in Table 1. The results achieved in
this work show that the conversion is more efficient than in
previous works, with respect to the reaction times. The reac-
tion time required for a 5-HMF yield of 71% was 0.5 min at
1408C, while 3-6 h were required to achieve 5-HMF yield of 60-
81% at 1008C.
The differences between the reaction in the IL system and in
other solvents, such as water or organic solvents, should not
solely attributed to the change of the pre-exponential factor in
the Arrhenius law due to microwave heating effects, as the
ionic liquid itself has a large effect on glucose or fructose de-
hydration. Binder et al. proposed that for ionic liquids that
have a chloride anion, a hexacoordinated chromium complex
catalyzes the isomerization of glucose into fructose through a
enediolate intermediate, and fructose then transforms rapidly
into 5-HMF through a nucleophilic pathway, where the fructo-
furanosyl oxocarbenium ion is readily attacked by the chloride
ion of the ionic liquid to form a 2-deoxy-2-halo intermediate.[17]
This intermediate would be prone to lose HCl to form the enol
and finally yield 5-HMF. Thus, there are strong chemical effects
that affect the kinetics.
Effect of catalyst amount
Table 2 shows how varying the amount of catalyst affects the
glucose conversion and 5-HMF yield. The molar ratios of cata-
lyst to glucose (C/G) were 0, 0.1, 0.2, and 0.4. No glucose con-
version occurred in the absence of catalyst, for a reaction time
of 30 min and a temperature of 1208C. At C/G=0.1, a glucose
conversion of 34% with a 5-HMF yield of 27% was obtained in
a reaction time of 1 min reaction time. In the same reaction
time, the glucose conversion increased to 59% and 61% when
using C/G ratios of 0.2 and 0.4, respectively. The data show
that the use of catalyst leads to a remarkable increase of the
glucose conversion and 5-HMF yield at the beginning of the
reaction (1 min), but that then the 5-HMF yield reaches a pla-
ChemSusChem 2010, 3, 1071 – 1077
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1073

X. Qi et al.
solutions in the conversion of fructose, where losses due to
the formation of humins can amount to up to 35% with a 1m
fructose solution while this value reduces to ca. 20% with
0.25m fructose solutions.[15] Polymerization losses cannot be
avoided with the present techniques, and so achieving high
yields of 5-HMF from highly concentrated glucose stocks re-
mains a challenge.
Table 2. Effect of the C/G ratio on glucose conversion and 5-HMF yield.[a]
Catalyst
[g]
C/G ratio
(mol/mol)
t
Glucose
conv. [%]
5-HMF
yield [%]
5-HMF
selec. [%]
[min]
0
0
0.1
30
1
3
10
1
3
10
1
3
0
34
64
88
59
84
94
85
94
99
0
27
48
66
46
64
69
61
68
70
0
81
75
75
78
75
73
71
72
70
0.0075
We examined the stability of 5-HMF in the system at 1208C
by adding 0.05 g of 5-HMF and 0.015 g CrCl3·6H2O into 1 g of
[BMIM][Cl] in the absence of glucose. After a reaction time of
10 min, 96% of the 5-HMF could be recovered, indicating that
it is stable under these experimental conditions in the absence
of glucose. This is a further indication that the lower yields of
5-HMF can be attributed to the formation of soluble polymers
that result from self-polymerization of glucose and cross-poly-
merization between glucose and 5-HMF.
0.0150
0.0300
0.2
0.4
10
[a] Different amounts of CrCl3·6H2O and 0.05 g glucose were dissolved in
1 g [BMIM][Cl] at 708C for 10 min with a water bath. Reaction mixtures
were placed in a microwave oven and heated to 1208C for different reac-
tion times. Glucose conversions and 5-HMF yields are based on HPLC
analyses with RI detector.
Recycling of the ionic liquid and catalyst
teau value between 66% to 70% at 10 min reaction time, be-
cause most of the glucose has been converted.
The ability to recycle both the solvent and the catalyst is an
important consideration for practical biomass transformations.
The activity of the ionic liquid/catalyst system was studied by
recycling both ionic liquid and catalyst six times with different
glucose loadings. Experiments were conducted at 1208C for a
reaction time of 10 min. Ethyl acetate was used to extract the
5-HMF product from the reaction mixture after each reaction
cycle, since 5-HMF can be extracted into ethyl acetate while
[BMIM][Cl] and glucose have practically no solubility in the
ethyl acetate phase.[2,4,10] After extraction of 5-HMF from the
ionic liquid phase, the reaction mixture was heated at 608C for
24 h in a vacuum oven to remove water and residual ethyl ace-
tate, and an equal amount of glucose was added and used for
the next reaction cycle. As shown in Figure 4, when successive
reactions were performed with recycled solvent and catalyst
comparable 5-HMF yields of 62–70% were obtained, showing
that the catalyst retained most of its activity. When the glucose
loading was increased to 0.2 g, the yield of 5-HMF decreased
somewhat compared to the yield obtained for a lower glucose
loading, giving 5-HMF yields ranging from 52% to 59%. The
oligomeric humin byproducts that remained in the solvent did
not seem to affect the subsequent dissolution and conversion
of glucose, but there did seem to be a systematic decrease in
the obtained yields as the number of recycling iterations in-
creased. The humins were not quantified, but their presence
could be ascertained from the color of reaction mixtures
(changing into a deep-brown color).[11,15] In some cases the 5-
HMF yield increased compared to previous cases, which can
be attributed to the retention of some 5-HMF and unreacted
glucose from the previous cycle. It is clear that separation of
the product and solvent requires energy. In the case of ethyl
acetate as extraction solvent, the energy costs of the separa-
tion may be low due to the use of a low-boiling-point solvent.
Effect of initial glucose concentration
In practical applications it is desirable to use reactants at high
concentrations, and so the initial concentrations of glucose
were varied as shown in Figure 3. When the glucose/[BMIM][Cl]
weight ratio was varied from 0.01 to 0.3, the 5-HMF yield grad-
ually decreased from 74% to 55%. The byproducts were fruc-
tose (<5% yield), 1,6-anhydroglucose (<1% yield), formic acid
(<2% yield), levulinic acid (<1%), and mainly soluble poly-
mers. The soluble polymers were not characterized. The losses
in 5-HMF yield with increasing glucose concentration are most
likely due to cross-polymerization occurring between glucose
and 5-HMF.[15] This is similar to effects observed with aqueous
Figure 3. Effect of glucose loading on 5-HMF yield from glucose. Different
amounts of glucose and corresponding amounts of CrCl3·6H2O (catalyst/glu-
cose molar ratio=0.2) were dissolved in 1 g [BMIM][Cl] at 708C for 10 min
(water bath heating). The mixtures were reacted in a microwave oven at
1208C for 10 min. Glucose conversions and 5-HMF yields are based on HPLC
analysis.
Conversion in various ionic liquids
As shown by the results described herein, CrCl3 is an effective
catalyst for the conversion of glucose into 5-HMF in [BMIM]
1074
www.chemsuschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2010, 3, 1071 – 1077

Fast Transformation of Glucose and Di-/Polysaccharides into 5-HMF
À
acidic ionic liquids with HSO4 as anion, the 5-HMF yields were
below 10%, which is most likely due to the formation of solu-
ble polymers and insoluble humins.
Zhao et al. postulated a mechanism for the chromium-cata-
lyzed transformation of glucose into 5-HMF in [EMIM][Cl] in the
presence of CrCl2.[5] They proposed that a hexacoordinate chro-
mium(II) complex catalyzes the isomerization of glucose into
fructose, probably via an enediolate intermediate, and fructose
is then rapidly converted into 5-HMF. Hu et al. proposed a
mechanism for the conversion of glucose to 5-HMF in [EMIM]-
[BF4] catalyzed by SnCl4, where the formation of a five-mem-
bered-ring chelate complex of the Sn atom with the two
neighboring hydroxyl groups in glucose is thought to play a
key role in the catalytic reaction, and the Sn in SnCl4 interacts
with O atoms and promotes the formation of straight-chain
glucose and the enediol intermediate.[16] The latter intermedi-
ate is further transformed into fructose, which can be dehy-
drated to form 5-HMF rapidly.
Figure 4. Successive use of [BMIM][Cl] and catalyst with different glucose
loadings. Conditions: 0.05 or 0.2 g glucose with corresponding amount of
CrCl3·6H2O (catalyst/glucose molar ratio=0.2) was dissolved in 1 g [BMIM]
[Cl] at 708C for 10 min (water bath heating). The mixture was reacted in a
microwave oven at 1208C for 10 min. 5-HMF was separated from the mix-
ture by extraction with 6 mL of ethyl acetate, 5 times, after 0.5 g of water
was added. After extraction, the ionic liquid solvent was heated to 608C for
24 h in a vacuum oven to remove water and residual ethyl acetate, and the
solvent was used directly in subsequent runs by adding glucose.
In our previous work, base catalyst was shown to favor the
isomerization of glucose to fructose while acid catalyst was
shown to favor the conversion of fructose into 5-HMF.[13] How-
ever, the transformation of glucose becomes nonselective
under acidic reaction environments because many other reac-
tion pathways, such as retro-aldol condensation and polymeri-
zations, become promoted. It has been reported that the acidi-
ty or basicity of alkylimidazolium chloro-metal ionic liquids can
be controlled by changing the amount of metal chloride
added to the ionic liquids. When the molar fraction of metal
chloride (x) was below 0.5, the ionic liquids became basic,
whereas for the cases of x=0.5 and x>0.5, the ionic liquids
exhibited neutral and acidic reaction environments, respective-
ly.[23] In this work, the catalytic amount of CrCl3 dissolved into
[BMIM][Cl] (x !0.5), and the formed [BMIM][CrCl4] ionic liquid
exhibited basicity that probably accelerated the isomerization
of glucose to fructose and the dehydration of fructose into 5-
HMF via the mechanism depicted above. On the other hand,
[HMIM][HSO4] and [BMIM][HSO4] are Brønsted-acidic ionic liq-
uids so that even after a small amount of CrCl3 is added, reac-
tions such as polymerization other than isomerization and de-
hydration are promoted. This probably leads to low 5-HMF
yields and large amounts of soluble polymers and insoluble
humins.
[Cl]. Because some ionic liquids readily dissolve carbohydrates
such as fructose, glucose and cellulose, the CrCl3-catalyzed
conversion of glucose into 5-HMF was studied in a series of
ionic liquids, namely [BMIM][Cl], 1-ethyl-3-methyl imidazolium
chloride ([EMIM][Cl]), 1-hexyl-3-methyl imidazolium chloride
([HexylMIM][Cl]), 1-methyl imidazolium hydrogen sulfate
([HMIM][HSO4]) and 1-butyl-3-methyl imidazolium hydrogen
sulfate ([BMIM][HSO4]). The results, shown in Figure 5, demon-
strate that ionic liquids with ClÀ as anion in the presence of
CrCl3 are the most effective for converting glucose into 5-HMF.
Even though glucose conversions were high for Brønsted-
Catalyzed conversion of various substrates in [BMIM][Cl]/
CrCl3
The [BMIM][Cl]/CrCl3 system was tested for catalyzing other
substrates, including fructose, sucrose, cellobiose, and cellu-
lose. This catalytic system was also effective for fructose con-
version and a 5-HMF yield of 78% was obtained at 1008C in
1 min. Sucrose is a disaccharide consisting of glucose and fruc-
tose. Normally, the 5-HMF yield from sucrose is mainly ob-
tained from the fructose moiety and the glucose moiety is for
the most part wasted.[6,8] Here, both the glucose and fructose
moiety could be converted into 5-HMF and a 5-HMF yield of
76% could be achieved. As a disaccharide consisting of two
molecules of glucose, when cellobiose was used as a substrate,
Figure 5. Catalyzed conversion of glucose into 5-hydroxymethylfurfural (5-
HMF) in various ionic liquids by CrCl3. Conditions: 0.05 g glucose with
0.015 g CrCl3·6H2O was dissolved in 1 g ionic liquid at 708C for 10 min
(water bath heating). The mixture was reacted with microwave heating at
1208C for 10 min. Glucose conversions and 5-HMF yields are based on HPLC
analysis.
ChemSusChem 2010, 3, 1071 – 1077
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1075

X. Qi et al.
a clear homogeneous rose-colored solution. The reaction mixture
was placed into a microwave oven (Shikoku Keisoku mReactor,
SMW-087, 2.45 GHz, maximum power 700 W) with magnetic stir-
ring and could be heated to the desired temperature within about
15 s. The reaction mixture was held at the reaction temperature for
a given period of time. At a given elapsed time, power was cut
and the reaction mixture was cooled quickly by the addition of
10 g of pure water that was followed by solution analyses. For re-
cycling of the ionic liquid and catalyst, 5-HMF was extracted from
the mixture by contact with 6 mL of ethyl acetate 5 times after
0.5 g of water was added. The addition of water to the ionic liquid
greatly promoted the transport of 5-HMF from the IL phase to the
ethyl acetate phase, and a separation efficiency of about 97% for
5-HMF removal from IL to ethyl acetate could be achieved. After
extraction, the ionic liquid was heated to 608C for 24 h in a
vacuum oven to remove water and residual ethyl acetate. It was
then used directly in subsequent runs by adding glucose.
Table 3. Formation of 5-HMF from various substrates in the [BMIM][Cl]
ionic liquid/CrCl3 catalytic system.[a]
Substrate
T [8C]
t [min]
5-HMF yield [%]
Fructose
Sucrose
Cellobiose
100
100
120
140
140
120
150
150
1
5
10
2
5
10
5
78
76
49
52
55
9
Cellulose
49
54
10
[a] 0.05 g sugar and 0.015 g CrCl3·6H2O were dissolved in 1 g [BMIM][Cl]
at 708C for 10 min (water bath heating). Reaction mixtures were trans-
ferred to a microwave oven and heated for different reaction times. 5-
HMF yields are based on HPLC analyses.
Analysis: HPLC equipped with a refractive index detector (HPLC-RI,
SH 1011 column) was employed for the analysis of the liquid sam-
ples. The column oven temperature was 608C, and the mobile
phase was a 0.5 mm sulfuric acid aqueous solution at a flow rate of
55% of cellobiose was converted to 5-HMF at 1408C for 5 min.
When cellulose was used as the feedstock, 54% could be con-
verted into 5-HMF at 1508C in 10 min (Table 3).
1 mLminÀ1
.
Additional experiments: We investigated if water might be re-
moved from the liquid phase due to microwave heating, so that
the dehydration reaction might be promoted. Microwave heating
of solution in open vessels did not affect the general reaction
trends reported in Figure 1.
Conclusions
The ionic liquid [BMIM][Cl] and CrCl3 is an excellent and envi-
ronmentally safe reaction solvent system for converting glu-
cose into 5-HMF. Both microwave heating and convective heat-
ing can be used. The conditions in this work were carefully
controlled to ensure that the kinetics based on microwave
heating are both reliable and reproducible. A high 5-HMF yield
of 69% was obtained from glucose at 1208C in 10 min, and
the 5-HMF yield increased to 71% in a reaction time of 30 s
when the reaction temperature was increased to 1408C. Good
results were achieved when fructose, sucrose, cellobiose, and
cellulose were used as feedstocks. The 5-HMF yields decreased
with increasing initial glucose concentration due to soluble
polymer formation. The ionic liquid [BMIM][Cl] and CrCl3 could
be recycled without loss of activity after the product 5-HMF
was separated. The combination of neutral ionic liquids with
ClÀ as anion, such as [BMIM][Cl], [EMIM][Cl], and [HexylMIM]
[Cl], with CrCl3 was effective for converting glucose into 5-
Acknowledgements
The authors gratefully acknowledge financial support by the Na-
tional Natural Science Foundation of China (NSFC; grant no.
20806041), SRF for ROCS, SEM, and the Japan Society for the Pro-
motion of Science (JSPS).
Keywords: biofuels · biomass · carbohydrates · cellulose ·
sustainable chemistry
[1] J. Lewkowski, Arkivoc 2001, 2, 17–54; Y. Y. Gorbanev, S. K. Klitgaard,
J. M. Woodley, C. H. Christensen, A. Riisager, ChemSusChem 2009, 2,
672–675.
[2] G. Yong, Y. G. Zhang, J. Y. Ying, Angew. Chem. 2008, 120, 9485–9488;
Angew. Chem. Int. Ed. 2008, 47, 9345–9348.
À
HMF, but the acidic ionic liquids with HSO4 as anion had a
[3] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Green Chem. 2008, 10,
799–805; X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Ind. Eng. Chem.
Res. 2008, 47, 9234–9239; M. Bicker, D. Kaiser, L. Ott, H. Vogel, J. Super-
crit. Fluids 2005, 36, 118–126; M. Bicker, J. Hirth, H. Vogel, Green Chem.
2003, 5, 280–284.
[4] S. Q. Hu, Z. F. Zhang, Y. X. Zhou, B. X. Han, H. L. Fan, W. J. Li, J. L. Song, Y.
Xie, Green Chem. 2008, 10, 1280–1283.
[5] H. B. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007, 316,
1597–1600.
[6] J. N. Chheda, Y. Roman-Leshkov, J. A. Dumesic, Green Chem. 2007, 9,
342–350.
[7] Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312, 1933–
1937.
[8] C. Moreau, A. Finiels, L. Vanoye, J. Mol. Catal. A: Chem. 2006, 253, 165–
169.
negative influence on 5-HMF formation.
Experimental Section
Materials and methods: Glucose (99%) and fructose (99%) were
purchased from Wako Pure Chemical Company; sucrose, cello-
biose, cellulose, 5-hydroxymethylfurfural, CrCl3·6H2O, [BMIM][Cl],
[HMIM][Cl], [EMIM][Cl], [HexylMIM][Cl], [HMIM][HSO4], and [BMIM]-
[HSO4] were purchased from Sigma–Aldrich Company. The water
content of the ionic liquids was lower than 1%, although it should
be noted that the reaction under investigation generates water
through dehydration. All chemicals were used without further pu-
rification.
In a typical run, 0.100 g of glucose (0.556 mmol) and 0.015 g
CrCl3·6H2O (0.056 mmol) were dissolved in 1 g of [BMIM][Cl] at
708C using a water bath. The conversion with this procedure was
not more than 3%. Initially, the glucose, CrCl3, and ionic liquid mix-
ture was green, and dissolution caused the mixture to change into
[9] F. Salak Asghari, H. Yoshida, Ind. Eng. Chem. Res. 2006, 45, 2163–2173;
C. Carlini, P. Patrono, A. M. R. Galletti, G. Sbrana, Appl. Catal. A: Gen.
2004, 275, 111–118; A. Takagaki, M. Ohara, S. Nishimura, K. Ebitani,
Chem. Commun. 2009, 6276–6278; F. Ilgen, D. Ott, D. Kralisch, C. Reil,
A. Palmberger, B. Konig, Green Chem. 2009, 11, 1948–1954; X. L. Tong,
Y. D. Li, ChemSusChem 2010, 3, 350–355.
1076
www.chemsuschem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2010, 3, 1071 – 1077

Fast Transformation of Glucose and Di-/Polysaccharides into 5-HMF
[10] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Green Chem. 2009, 11,
1327–1331.
[18] C. Z. Li, Z. H. Zhang, Z. B. K. Zhao, Tetrahedron Lett. 2009, 50, 5403–
5405.
[11] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, ChemSusChem 2009, 2,
944–946.
[12] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Catal. Commun. 2009, 10,
1771–1775; Q. X. Bao, K. Qiao, D. Tomida, C. Yokoyama, Catal. Commun.
2008, 9, 1383–1388; T. S. Hansen, J. M. Woodley, A. Riisager, Carbohydr.
Res. 2009, 344, 2568–2572.
[19] A. Beni, R. Karosi, J. Posta, Microchem. J. 2007, 85, 103–108.
[20] T. M. Florence, Talanta 1982, 29, 345–364.
[21] M. J. Antal, W. S. L. Mok, G. N. Richards, Carbohydr. Res. 1990, 199, 91–
109.
[22] M. Hosseini, N. Stiasni, V. Barbieri, C. O. Kappe, J. Org. Chem. 2007, 72,
1417–1424.
[13] X. H. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Catal. Commun. 2008, 9,
2244–2249.
[14] Y. Romꢂn-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic, Nature 2007, 447,
982–985.
[23] J. S. Wilkes, J. A. Levisky, R. A. Wilson, C. L. Hussey, Inorg. Chem. 1982,
21, 1263–1264; M. Hasan, I. V. Kozhevnikov, M. R. H. Siddiqui, A. Steiner,
N. Winterton, Inorg. Chem. 1999, 38, 5637–5641.
[15] B. F. M. Kuster, Starch/Staerke 1990, 42, 314–321.
[16] S. Q. Hu, Z. F. Zhang, J. L. Song, Y. X. Zhou, B. X. Han, Green Chem. 2009,
11, 1746–1749.
Received: April 28, 2010
Revised: May 26, 2010
[17] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985.
Published online on July 26, 2010
ChemSusChem 2010, 3, 1071 – 1077
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1077