Production of 5-hydroxymethylfurfural from fructose in ionic liquid efficiently catalyzed by Cr(III)-Al2O3 catalyst

Article abstract of DOI:10.1002/cjoc.201400054

A series of metal-Al₂O₃ catalysts were prepared simply by the conventional impregnation with Al₂O₃ and metal chlorides, which were applied to the dehydration of fructose to 5-hydroxymethylfurfural (HMF). An agreeable HMF yield of 93.1% was achieved from fructose at mild conditions (100°C and 40 min) when employing Cr(III)-Al₂O₃ as catalyst in 1-butyl-3-methylimidazolium chloride ([Bmim]Cl). The Cr(III)-Al₂O₃ catalyst was characterized via XRD, DRS and Raman spectra and the results clarified the interaction between the Cr(III) and the alumina support. Meanwhile, the reaction solvents ([Bmim]Cl) collected after 1^st reaction run and 5 ^th reaction run were analyzed by ICP-OES and LC-ITMS and the results confirmed that no Cr(III) ion was dropped off from the alumina support during the fructose dehydration. Notably, Cr(III)-Al₂O₃ catalyst had an excellent catalytic performance for glucose and sucrose and the HMF yields were reached to 73.7% and 84.1% at 120°C for 60 min, respectively. Furthermore, the system of Cr(III)-Al₂O₃ and [Bmim]Cl exhibited a constant stability and activity at 100°C for 40 min and a favorable HMF yield was maintained after ten recycles. The schematic diagram for fructose dehydration to HMF on Cr(III)-Al₂O₃ catalyst. Copyright

Full text of DOI:10.1002/cjoc.201400054

FULL PAPER
DOI: 10.1002/cjoc.201400054
Production of 5-Hydroxymethylfurfural from Fructose in Ionic
Liquid Efficiently Catalyzed by Cr(III)-Al2O3 Catalyst
Changhua Song, Hui Liu, Yuan Li, Shuai Ge, Hongwei Wang, Wenshuai Zhu, Yonghui Chang,
Changri Han, and Huaming Li*
School of Chemistry and Chemical Engineering, Jiangsu University, Zhenjiang, Jiangsu 212013, China
A series of metal-Al2O3 catalysts were prepared simply by the conventional impregnation with Al2O3 and metal
chlorides, which were applied to the dehydration of fructose to 5-hydroxymethylfurfural (HMF). An agreeable HMF
yield of 93.1% was achieved from fructose at mild conditions (100 ℃ and 40 min) when employing Cr(III)-Al2O3
as catalyst in 1-butyl-3-methylimidazolium chloride ([Bmim]Cl). The Cr(III)-Al2O3 catalyst was characterized via
XRD, DRS and Raman spectra and the results clarified the interaction between the Cr(III) and the alumina support.
Meanwhile, the reaction solvents ([Bmim]Cl) collected after 1st reaction run and 5th reaction run were analyzed by
ICP-OES and LC-ITMS and the results confirmed that no Cr(III) ion was dropped off from the alumina support dur-
ing the fructose dehydration. Notably, Cr(III)-Al2O3 catalyst had an excellent catalytic performance for glucose and
sucrose and the HMF yields were reached to 73.7% and 84.1% at 120 ℃ for 60 min, respectively. Furthermore, the
system of Cr(III)-Al2O3 and [Bmim]Cl exhibited a constant stability and activity at 100 ℃ for 40 min and a fa-
vorable HMF yield was maintained after ten recycles.
Keywords 5-hydroxymethylfurfural, [Bmim]Cl, Cr(III)-Al2O3, fructose
Introduction
of HMF in the dehydration of carbohydrates, leading to
many byproducts like levulinic acid and lactic acid. As
reported by Yang et al., only a HMF yield of 22% was
attained in water without combining tetrahydrofuran
(THF) as extracting organic layer to protect HMF from
further reactions with glucose.[11] The organic solvents
can suppress the unwanted side reactions and highlight
the selectivity of HMF in the dehydration of hexose.
However, higher activation energy may be necessary in
such catalytic systems, thus, high temperature would be
inevitable. According to the recent researches, 170 ℃[12]
and 140 ℃ cooperated with microwaves[13] were re-
quired to activate the catalytic systems in the solvents of
butanol and DMSO, respectively. Meanwhile, the ap-
plication of high boiling organic solvents increases the
difficulty in the subsequent separation of HMF from the
final products.
Consequently, the ILs appear to be a better choice.
Zhao et al. first reported the Lewis acid catalyst (CrCl2)
could cooperate with the ionic liquid 1-ethyl-3-methyl-
imidazolium chloride ([Emim]Cl) to achieve the HMF
yield about 70%.[14] And then many types of ILs have
been attempted. Hu et al. studied the production of
HMF from sucrose in 1-ethyl-3-methylimidazolium
tetrafluoroborate ([Emim]BF4) using SnCl4 as a catalyst
and achieved a HMF yield of 65% at 120 ℃ for 24
The irreversible consumption of the fossil fuels and
the environmental problems have spurred the research
of fuels and chemicals from biomass.[1,2] In this context,
5-hydroxymethylfurfural (HMF), obtained by the cata-
lytic dehydration of carbohydrates based on C6 sugars,
has been early envisaged as a versatile and promising
intermediate.[3] Either 2,5-diformylfuran (2,5-DFF)[4] or
2,5-furandicarboxylic acid (2,5-FDCA)[5] are produced
by the selective oxidation of HMF. 2,5-FDCA is a de-
sirable substitute for terephthalic or isophthalic acids in
the fields of polyester production and a promising in-
termediate in the preparation of polyamides and poly-
urethane.[6] Alternatively, levulinic acid synthesized by
the hydrolysis of HMF can be converted into methyl-
tetrahydrofuran (MTHF) and various levulinate esters,
which can be used as additives in gasoline and biodiesel,
respectively.[7] However, the challenges have been
posed to optimize the conversion process in an envi-
ronmentally benign, economic and efficient way for the
biomass into HMF.
From the previous researches, the preparation of
HMF from carbohydrates has been conducted in aque-
ous media,[8] organic solvents[9] and ionic liquids
(ILs).[10] The utilization of water fulfills the principles of
the green chemistry, but it shows unsatisfying selectivity
*
E-mail: lihm@ujs.edu.cn; Tel.: 0086-511-88791800; Fax: 0086-511-88791708
Received January 25, 2014; accepted April 13, 2014; published online May 22, 2014.
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cjoc.201400054 or from the author.
434
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2014, 32, 434—442

Production of 5-Hydroxymethylfurfural from Fructose in Ionic Liquid
h.[15] In subsequent studies, the conversion of fructose
was also accomplished in [Emim]Cl with addition of
H3PW12O40 encapsulated in MIL-101 and HMF with
high yield of 79% was attained at 80 ℃ for 2.5 h.[9]
Nearly 54% yield of HMF from glucose was achieved
in a 50∶50 (w/w) 1-hexyl-3-methylimidazolium chlo-
ride ([Hmim]Cl)-water mixture in the presence of ZrO2
within 10 min at 200 ℃.[16] More recently, Yi et al.
studied the dehydration of girasol, as a source of inulin,
into HMF in 1-octyl-3-methylimidazolium chloride
([Omim]Cl) catalyzed by HCl at 120 ℃ for 1 h, and
the yield of HMF was about 58.2%.[17] However, high
temperature and long reaction time may speed up the
self- or cross-polymerization between fructose and gen-
erated HMF, demonstrating that HMF was no longer the
favored product. Thus, the fructose dehydration to HMF
based on mild reaction conditions should be explored.
In our previous work, ion-exchanged resins after
acidification were employed as the catalyst, which
achieved the fast conversion of fructose to HMF but
failed in the conversion of glucose to HMF,[18] possibly
due to the fact that the acidic resins had a poor activity
in the isomerization of glucose to fructose before the
dehydration of fructose to HMF. Thus, besides focusing
on an efficient conversion of fructose to HMF, the de-
sign and exploration of catalysts employed in this work
were aiming at synthesis of HMF from more complex
carbohydrates. In this context, the common Lewis acids,
such as chromium halide, caught our eyes, due to their
excellent activity in the dehydration of sugars into HMF
even the substrate was rude biomass. Yi et al. reported
that a yield of HMF about 60% was achieved from raw
tapioca root biomass with high starch in [OMIM]Cl
catalyzed by CrF3 at 120 ℃ for 90 min.[19] But the
chromium ion is toxic and the separation of the Cr(III)
or Cr(II) is still a problem. Fortunately, the heterogene-
ous catalysts have broadened the horizon of the study in
sugar dehydration. Degirmenci et al. found that the
catalyst of CrCl2 supported on the mesoporous silica
zeolite (CrCl2-Im-SBA-15) highlighted the selectivity of
HMF to 50% and realized the conversion to HMF from
glucose in the aqueous medium.[20] Furthermore, Zhang
and Zhao revealed that a HMF yield of 40% was ob-
tained from the dehydration of glucose under micro-
wave irradiation in 2.5 min in the presence of hy-
droxyapatite supported chromium chloride (Cr-HAP).[21]
However, in the synthesis of CrCl2-Im-SBA-15 and
Cr-HAP, the high costs of CrCl2 and hydroxyapatite
(HPA) have posed the challenges on the industrializa-
tion and commercialization of these catalytic systems.
Therefore, suitable support and the precursor of chro-
mium are important for the preparation of supported
chromium catalyst.
satisfying catalytic performance has been obtained. To
the best of our knowledge, metal-Al2O3 catalysts were
the first time used in the dehydration reaction from car-
bohydrates.
Herein, a series of metal-Al2O3 catalysts were pre-
pared in the experiment, which were easily attained via
the conventional impregnation with Al2O3 as the support
in the aqueous metal chlorides solution. The metal chlo-
rides included MgCl2, ZnCl2, MnCl2, CrCl3, WCl6,
MoCl5, CuCl2, FeCl3, SnCl4 and SnCl2. And then these
catalysts were applied to the dehydration of fructose to
HMF. Additionally, the activity and stability of Cr(III)-
Al2O3 in the sugar dehydration reaction and the opti-
mum reaction conditions were investigated in our work.
The aim was to study the effects of the introduction of
metal ions into Al2O3 support on the dehydrating activ-
ity of sugar, thus exploring an optimal heterogeneous
catalyst for the dehydration based on mild reaction con-
ditions.
Experimental
Materials
HMF (99% purity) was purchased from Wutong
Spice Co., Ltd. Chromium chloride hexahydrate (CrCl3•
6H2O, 99% purity), molybdenum chloride (MoCl5,
99.6% purity) and tungsten chloride (WCl6, 99.9% pu-
rity) were purchased from Aladdin. Fructose (B.R.
grade), glucose (A.R. grade), sucrose (A.R. grade),
maltose (B.R. grade), starch (A.R. grade), cellulose
(column chromatography), aluminum oxide (Al2O3, A.R.
grade), iron chloride (FeCl3, A.R. grade), stannic chlo-
ride dihydrate (SnCl2•2H2O, A.R. grade), tin chloride
pentahydrate (SnCl4•5H2O, A.R. grade), zinc chloride
(ZnCl2, A.R. grade), manganese chloride tetrahydrate
(MnCl2•4H2O, A.R. grade), magnesium chloride hexa-
hydrate (MgCl2•6H2O, A.R. grade), copper chloride
dihydrate (CuCl2•2H2O, A.R. grade), barium sulfate
(BaSO4), ethyl acetate (CH3COOC2H5, A.R. grade) and
methanol alcohol (CH3OH, L.C. grade) were purchased
from Sinopharm Chemical Reagent Co., Ltd. 1-Butyl-3-
methylimidazolium chloride ([Bmim]Cl) was purchased
from Shanghai Cheng Jie Chemical Co., Ltd. All of
them were used as received without further purification.
Experimental methods
Preparation of the catalysts In a typical impreg-
nation, Al2O3 was impregnated at 80 ℃ for 4 h in the
aqueous solution of metal chlorides, including CrCl3,
FeCl3, SnCl2, SnCl4, ZnCl2, MnCl2, MgCl2, CuCl2,
WCl6 and MoCl5, and was dried at 80 ℃ overnight
sequentially. Thus the supported metal-Al2O3 powder
was obtained.
In recent years, aluminium oxide (Al2O3), as an
abundant, cheap and commercially available material,
has been widely used as the catalyst carrier in many
organic reactions, such as reduction of NOx,[22] reform-
ing of methane[23] and oxidation of CO2.[24] And much
Characterization The X-ray diffraction (XRD)
was recorded on a Bruker D8 diffractometer with Cu Kα
radiation (λ＝1.54 Å) in the range of 2θ＝10°－80°.
The diffuse reflectance spectra (DRS) were measured by
a UV-vis spectrometer (UV-2450, Shimadzu) in the
Chin. J. Chem. 2014, 32, 434—442
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
435

Song et al.
FULL PAPER
range of 200－800 nm. BaSO4 was used as the reflec-
tance standard material. Raman spectra were carried out
at room temperature using a DXR Raman microscope
with a 532 nm laser source. Temperature-programmed
desorption of ammonia (NH3-TPD) was performed on
an apparatus TP-5079 (Tianjin Xianquan Limited Cor-
poration). The pyridine adsorption infrared spectra
(Py-IR) were recorded on a Perkin-Elmer Spectrum
2000. The pretreatment of samples in the NH3-TPD
analysis and the pyridine-absorbed FT-IR analysis were
referred to literature.[25] The metal content in solvent
was determined by inductively coupled plasma optical
emission spectrometry (ICP-OES) (VISTA-MPX, Var-
ian, USA). The m/z parameters of samples were identi-
fied by liquid chromatography-linear ion trap mass
spectrometer (LC-ITMS) (Thermo LXQ, USA).
Typical reaction procedure [Bmim]Cl (1.0 g) was
preheated at the reaction temperature. Then, fructose
(0.1 g) and the given amount of catalysts (metal-Al2O3)
were charged into the [Bmim]Cl. Meanwhile, a mag-
netic stirrer started. After the desired time, the reaction
mixture was cooled soon by an ice bath and diluted by
pure water at the same time. The diluted solution was
filtered before the HPLC analysis.
Analysis High performance liquid chromatography
(HPLC) was employed to analyze the product. The
HPLC setup was equipped with two Varian ProStar210
pumps, an Agilent TC-C (18) column, and a Varian
ProStar325 UV-Vis Detector. 283 nm was the measured
absorbance wavelength. The mobile phase was a solu-
tion of methanol and H2O in the ratio 50∶50 (V/V) at
0.8 mL/min flow rate. The temperature of column was
kept at 35 ℃ by using a column oven. The yield of
HMF was calculated as follows.
Al2O3 frameworks and/or the changes in the symmetry
of Cr(III) in the Al2O3 material.[9,26] Diffraction peaks
appeared at the angular values 2θ around 22°, 24°, 26°
and 27°, which were ascribed to chromium oxide
formed during the preparation of catalyst.
UV-vis analysis Figure S2 showed the UV-vis dif-
fuse reflectance spectra of the catalysts. No absorption
peak was observed for the alumina support in the 200－
800 nm region. Two absorption peaks at about 450 nm
and 640 nm in the pure CrCl3•6H2O were found, which
were assigned to the absorption of Cr(III).[27] In addition,
it was observed that the maximum absorption band of
Cr(III) was shifted to longer wavelength in the Cr(III)-
Al2O3 catalyst, indicating that some Cr(III) interacted
with the Al2O3 support in the impregnation.
FT-Raman spectroscopy The Raman spectra of
different samples were presented in Figure 1. Any Ra-
man feature in the alumina support was not found in the
200－1200 cm−1 region, therefore, all the observed Ra-
man bands in other samples were assigned to chromium
species vibrations. In the spectra of CrCl3•6H2O, three
bands at around 261, 324 and 460 cm−1 were observed,
which were attributed to the water librations and the
Cr－Cl stretching.[28] After loading CrCl3 on Al2O3 car-
rier to form Cr(III)-Al2O3 catalyst, these bands all
shifted to lower wave number. What’s more, a small and
sharp Raman band at around 362 cm−1 was found in the
spectra of Cr(III)-Al2O3 catalyst, which was attributed
to Cr－O bending modes.[29] It was proved that the in-
teraction between the Cr(III) and the alumina support
was formed. Moreover, a band at 539 cm−1 belonged to
the formation of crystalline Cr2O3 was also observed in
Cr(III)-Al2O3 catalyst,[29,30] which was consistent with
the result of the XRD patterns. But the band was broad-
er than that reported in the literature, maybe due to the
overlap with the band at 453 cm−1.
For fructose and glucose, HMF yield was defined
as follows:
NH3-TPD analysis The temperature programmed
desorption of ammonia was measured to characterize
the acidity of the Cr(III)-Al2O3 catalyst. Figure 2
showed the NH3-TPD profiles of different samples and
the results were summarized in Table 1. Two NH3 de-
sorption peaks at 272 and 549 ℃ were observed in the
nHMF
Yield＝
×100%
nmonosaccharide
For disaccharides and polysaccharides, HMF yield
was defined as follows:
nHMF
Yield＝
×100%
nhexose-unites
where nHMF was the moles of HMF; nmonosaccharide was
the initial moles of fructose or glucose; nhexose-unites was
the initial moles of hexose units in disaccharides or
poly-saccharides.
Results and Discussion
Characterization of Cr(III)-Al2O3 catalyst
XRD patterns The XRD patterns, as presented in
Figure S1, confirmed that the structure of Al2O3 in
Cr(III)-Al2O3 catalyst was maintained. However, some
redistribution of the XRD intensities in the catalyst took
place compared with the pure Al2O3 material. Some
changes at the angular values 2θ around 15°, 38° and
49° were attributed to the interaction of Cr(III) with the
Figure 1 Raman spectra of different samples: (a) Al2O3, (b)
Cr(III)-Al2O3, (c) CrCl3•6H2O.
436
www.cjc.wiley-vch.de
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2014, 32, 434—442

Production of 5-Hydroxymethylfurfural from Fructose in Ionic Liquid
profile of Al2O3, which were attributed to weak acid site
come from Al2O3 and the immobilized chromium ion.
In addition, it was reported that the chromium ion might
coordinate with the alcohol group and the oxygen atom
on the surface of Al2O3 through the impregnation, lead-
ing to initiating some Brønsted acidity for the Cr(III)-
Al2O3 catalyst.[32]
and strong acid site on the surface of Al2O3, respec-
＋
tively.[31] After supporting Cr3 , the higher-temperature
desorption peak areas decreased significantly and the
peak temperature shifted to lower temperature, showing
the weakening of strong acidity. Meanwhile, the desorp-
tion peak at low-temperature shifted to higher tempera-
ture and new peaks were found, indicating the increase
of weak acidity. A new desorption peak at 379 ℃ in
the profiles of Cr(III)-Al2O3 was found, which belonged
to the moderate acid site. The results indicated that the
addition of Cr changed the acidity of catalyst. Moreover,
it could be found in Table 1 that the total acidity from
the Cr(III)-Al2O3 catalyst was on a rise with modifica-
tion of Cr. It showed that the Cr incorporation indeed
created more acidic sites in the aluminum framework.
But it was interesting that the acid strength distribution
of Cr(III)-Al2O3 was uniform among the weak, moder-
ate and strong acidity in comparison with aluminum
support from the data in Table 1.
Pyridine-adsorbed FT-IR analysis The type of
acid on Cr(III)-Al2O3 catalyst was determined by pyri-
dine adsorption infrared spectroscopy and the result was
illustrated in Figure 3. Four major bands at 1447, 1488,
1537 and 1587 cm−1 were found in the spectroscopy of
Cr(III)-Al2O3 catalyst. While referring from the reports
of Wang et al., the bands at about 1450 and 1540 cm−1
were assigned to Lewis acidic site and Brønsted acidic
site, respectively, and the band at 1490 cm−1 was char-
acteristic of pyridine adsorbed on both Lewis and Brøn-
sted acid sites.[25] The results proved that Cr(III)-Al2O3
catalyst contained two kinds of acidity sites: Lewis and
Brønsted acidity sites. Comparing the spectra of Cr(III)-
Al2O3 catalyst with that of Al2O3, the intensity and peak
area of the band at about 1450 cm−1 increased, which
indicated that the Lewis acidity of Cr(III)-Al2O3 might
Figure 3 FT-IR pyridine adsorption spectra of Al2O3 and
Cr(III)-Al2O3 catalyst.
Sugar dehydration reactions and recycle experi-
ments
Effects of metal chlorides Initially, the effects of
different metal ions supported on Al2O3 on the conver-
sion of fructose to HMF were explored and the results
were shown in Figure 4. In the blank experiment (with-
out loading any metal ion on the Al2O3), the HMF yield
was quite low, indicating that the catalytic activity for
the dehydration of fructose from the catalyst carrier
(Al2O3) was negligible. As shown in Figure 4, different
HMF yields, ranging from 4.2% to 84.3%, were ob-
served after supporting different metal ions on Al2O3. It
illustrated that the catalytic activities of metal-Al2O3
catalysts were depended on the properties of metal ions
largely. An excellent yield of 84.3% was achieved when
Cr(III)-Al2O3 was employed. Only moderate yields
were obtained with the Sn(IV)-Al2O3, Cu(II)-Al2O3,
W(VI)-Al2O3 and Mo(V)-Al2O3 catalysts and even low
yields were reached with the Fe(III)-Al2O3 and
Sn(II)-Al2O3 catalysts. However, not all the metal ions
supported on Al2O3 were effective, the modification of
Al2O3 with Zn(II), Mn(II) and Mg(II) did not trigger the
conversion of fructose to HMF, respectively.
Recognizing that the dehydration of fructose to HMF
was an acid-catalyzed process, therefore, it became
more easily with the increasing of the acidity of the cat-
alytic metal centers. According to the researches of
Figure 2 NH3-TPD profiles of Al2O3 and Cr(III)-Al2O3 catalyst.
Table 1 The acid strength distribution of the catalysts calculated by the results of NH3-TPD profiles
Total acidity/
(mmol NH3/g)
Weak acidity/
(mmol NH3/g)
Moderate acidity/
(mmol NH3/g)
Strong acidity/
(mmol NH3/g)
Sample
Temp./℃
Temp./℃
Temp./℃
Al2O3
0.87
1.35
0.09
0.38
272
282
－
－
0.78
0.40
549
501
Cr(III)-Al2O3
0.57
379
Chin. J. Chem. 2014, 32, 434—442
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
437

Song et al.
FULL PAPER
framework were created by immobilization of Cr, which
was in accordance with the result of NH3-TPD profiles.
In addition, the result was due to that part of Cr(III) ions
could be distributed on surface of Al2O3 carrier and the
others might be entered into the channel of Al2O3 or
exchanged into the framework of Al2O3 through the im-
pregnation method,[25,35] which was consistent with the
result of UV-vis diffuse reflectance spectra and Raman
spectra. Therefore, as shown in Figure 5, fructose could
react with not only the Cr(III) ion on the surface of
Al2O3 carrier but also those in the porous channel in the
conversion of fructose to HMF. Thus, the HMF gener-
ated in the porous channel of the Al2O3 carrier, might be
inhibited from further self- and cross-polymerization
with fructose. Furthermore, the pyridine-adsorbed FT-IR
spectrum demonstrated the existence of two types of
acid sites on the Cr(III)-Al2O3 catalyst, favoring the de-
hydration of fructose to HMF.[36]
Figure 4 The fructose dehydration to HMF catalyzed by dif-
ferent metal-Al2O3 catalysts.
Zhang et al., Cr(III) showed the stronger Lewis acidity
in comparison with other metal cations,[33] possibly ac-
celerating the dehydration of fructose. Thus, Cr(III)-
Al2O3 catalyst was selected as the optimum catalyst in
the subsequent experiments.
To investigate the effect of metal ion and the carrier
on the dehydration of fructose, some controlled experi-
ments were carried out and the results were shown in
Table 2. Obviously, only HMF yield of 4.3% was ob-
tained when 0.1 g Al2O3 (entry 3) was used as catalyst
due to the low Lewis acidity of Al2O3. HMF yield
boosted to 74.9% when 0.1 g CrCl3 (entry 2) was ap-
plied, possibly resulting from that CrCl3 owned stronger
acidity and new catalyst ([Bmim]CrCl4) could be
formed by the interaction between CrCl3 and
[Bmim]Cl.[34] However according to the proportion of
CrCl3 and Al2O3 in supported Cr(III)-Al2O3 catalyst,
directly adding 0.1 g mixture of CrCl3 and Al2O3 (entry
4) in the system as catalyst, HMF yield slightly de-
creased in comparison with entry 2. It might be due to
the decrease of CrCl3 dosage and the hindrance effect of
Al2O3 for the mass transfer process.
Figure 5 The schematic diagram for fructose dehydration to
HMF on Cr(III)-Al2O3 catalyst.
However, a question arose whether Cr(III) ion was
able to drop off from the Al2O3 support during the reac-
tion. The results listed in Table 2 did not satisfactorily
confirm that the actual activity center for the fructose
dehydration was the Cr(III) ion distributed on Al2O3
rather than the dissociative Cr(III) ion dropped from the
Cr(III)-Al2O3 catalyst. It was reported that the dissocia-
tive Cr(III) ion could react with [Bmim]Cl to form
[Bmim]CrCl4 like Fe(III) ion[37] at such reaction condi-
tions (100 ℃ and 40 min). Therefore, according to the
literature,[38] [Bmim]CrCl4 was prepared by mixing
[Bmim]Cl and CrCl3, and the reaction solvents after 1st
reaction run and 5th reaction run were collected through
filtration, respectively. Meanwhile, a control experiment
was conducted by mixing [Bmim]Cl and CrCl3 at 100
Table 2 The effects of different catalysts on the dehydration of
fructosea
Entry
Catalyst
HMF yield/%
1
2
3
4
Cr(III)-Al2O3
CrCl3
84.3
74.9
4.3
Al2O3
CrCl3＋Al2O3
67.6
a Conditions: catalyst 0.1 g, fructose 0.1 g, [Bmim]Cl 1.0 g, 90
℃, 40 min.
℃
for 40 min in order to test that whether
[Bmim]CrCl4 could be formed at such reaction condi-
tions. Two characterizations including ICP-OES and
LC-ITMS were further employed to characterize these
samples. Firstly, the reaction solvents were detected by
ICP-OES and no Cr(III) ion was traced in both of them.
Secondly, all samples were characterized by LC-ITMS
method and the results were shown in Figure 6. It was
confirmed that [Bmim]＋ cation existed in all of samples
from the result of Figure 6a and the reaction conditions
Furthermore, it was noteworthy that the Cr(III)-
Al2O3 catalyst (entry 1) prepared by impregnation
method showed the highest dehydration activity and
HMF yield reached to 84.3%. It clearly indicated that
the distributed loading of Cr(III) ion on Al2O3 carrier by
the impregnation method facilitated the dehydration of
fructose and the formation of HMF. It was possibly
arisen from that more acidic sites in the aluminum
−
(100 ℃ and 40 min) were enough to form CrCl4 by
438
www.cjc.wiley-vch.de
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2014, 32, 434—442

Production of 5-Hydroxymethylfurfural from Fructose in Ionic Liquid
comparing Figure 6b-1 with Figure 6b-2. But no mo-
and then remained unchanged with the further increase
of catalyst dosage. The results might be ascribed to that
the side reactions in the catalytic system were also ac-
celerated by the over-using catalysts. Thus, the dosage
of Cr(III)-Al2O3 was chosen as 0.075 g in the subse-
quent experiments.
−
lecular ion peak of CrCl4 was found in both of Figure
−
6b-3 and Figure 6b-4, indicating that no CrCl4 was
formed during the dehydration. The results indicated
that the Cr(III) ion preferred to be loaded on the carrier
Al2O3 rather than be dropped off during the dehydration
in the ionic liquid [Bmim]Cl, although the Cr(III) ion
had strong dispersibility in the aqueous media.
Table 3 The effects of the dosages of Cr(III)-Al2O3 on the de-
hydration of fructosea
Entry
Cr(III)-Al2O3/g
0
HMF yield/%
5.3
1
2
3
4
5
6
7
0.025
76.1
0.050
81.5
0.075
87.7
0.100
84.3
0.150
80.8
0.200
80.8
a Conditions: fructose 0.1 g, [Bmim]Cl 1.0 g, 90 ℃, 40 min.
Effects of reaction temperature and time Figure
7 showed the influence of the reaction temperature on
the dehydration of fructose to HMF. Obviously, the ini-
tial yields of HMF were on the rise with the elevation of
the temperature. More precisely, the high temperature
shortened the time to reach the maximum value of HMF
yields. HMF yield of 93.1% was obtained at 100 ℃ for
40 min, while it only took 20 and 5 min to reach HMF
yield of 91.2% at 110 ℃ and 91.3% at 120 ℃, respec-
tively. The results indicated that high temperature may
speed up the dehydration of fructose.
Figure 6 LC-ITMS spectra of the cations (a) and anions (b) in
all of samples: (a-1 and b-1) [Bmim]CrCl4 prepared by the
method reported in the literature, (a-2 and b-2) the mixture of
[Bmim]Cl and CrCl3, (a-3 and b-3) reaction solvent after 1st reac-
tion run and (a-4 and b-4) reaction solvent after 5th reaction run.
Effects of catalyst dosage The effects of different
catalyst dosages on the HMF yields were examined and
the results were shown in Table 3. As seen in Table 3,
HMF yields first increased and then decreased with the
increasing dosages of catalyst. The HMF yield was only
5.3% in control experiment, however, the yield of HMF
boosted to 76.1% when 0.025 g Cr(III)-Al2O3 was
added. The result confirmed the excellent catalytic per-
formance of Cr(III)-Al2O3 in this dehydration reaction.
As shown in Table 3, the yield of HMF maximized to
87.7% when increasing the catalyst dosage to 0.075 g.
The reason might be that the kinetics of fructose dehy-
dration was faster with the increase of catalytic sites
provided by more catalyst dosage in the reaction. How-
ever, the HMF yields decreased smoothly when the
amount of catalyst was increased from 0.075 to 0.150 g
Figure 7 HMF yield as a function of the reaction time at dif-
ferent temperatures.
Moreover, the change of HMF yields was found at
different temperatures with the prolonging of reaction
time in Figure 7. HMF yields increased quickly in the
first 40 min at lower temperature (80 and 90 ℃) and
the lower the temperature was, the more apparent the
tendency was. In the following 20 min, the rise of HMF
yield became smoothly. It showed that at lower tem-
perature the prolonging of reaction time could highlight
HMF yields effectively. When raising the temperature to
Chin. J. Chem. 2014, 32, 434—442
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
439

Song et al.
FULL PAPER
100 and 110 ℃, HMF yields increased firstly and then
decreased with extending the reaction time. It was re-
ported that some side reactions occurred in the process
of the dehydration of fructose, resulting in forming
some insoluble humins and soluble polymers by the
self- and cross-polymerization of HMF.[13] Furthermore,
with the formation of H2O in the process of the dehy-
dration inevitably, the alternation of the solvent from
[Bmim]Cl to [Bmim]Cl/H2O mixture solvent gave the
chance to the rehydration of the HMF, leading to the
generation of levulinic acid.[39] Therefore, high tem-
perature might accelerate not only the dehydration reac-
tion to HMF, but also some undesirable side reaction,
demonstrating that HMF was no longer the favored
product. When reaction temperature was further in-
creased (120 ℃), the HMF yield reached the maximum
in the first 5 min, and then declined continuously with
the extension of time, which confirmed the conclusion
again. Thus, 100 ℃ and 40 min were considered as the
optimized reaction temperature and time, respectively.
Effects of initial fructose concentration The ef-
fects of the initial concentration of fructose on HMF
yields in [Bmim]Cl were illustrated in Table 4. HMF
yield decreased continuously with the increase of initial
concentration of fructose. High HMF yield was ob-
served at the extremely low concentration of fructose
(4.8 wt%) in ILs. When the concentration of fructose
increased to 9.1 wt%, the HMF yield decreased slightly
to 93.1%. Further increasing the initial concentration of
fructose to 16.7 wt%, an obvious change of HMF yield
was observed. As discussed above, the result could be
ascribed to the polymerization of HMF and fructose to
form humins.[13] However, additional increase of the
fructose did not lead to a significant decline of the HMF
yield in the ensuing process. When the initial concentra-
tion of fructose increased to 23.1 wt%, HMF yield only
decreased to 84.1%. The reason might be due to the ex-
istence of [Bmim]Cl, as a benign catalytic media in the
biomass dehydration, retarding the formation of un-
wanted side products. An HMF yield of 82.5% was still
achieved with the initial fructose concentration of 28.6
wt% (0.4 g) at 100 ℃ for 40 min, indicating that the
Cr(III)-Al2O3/[Bmim]Cl system owned a high catalytic
efficiency for this dehydration reaction.
Conversion of different substrates to HMF with
Cr(III)-Al2O3 catalyst In general, a one-pot synthesis
of HMF from disaccharide and polysaccharide was
more complicated than fructose and the hydrolysis,
isomerization and dehydration reactions would take
place one by one in the process, resulting in the diffi-
culty in the conversion of biomasses to HMF. Therefore,
to evaluate the scope of application of Cr(III)-Al2O3
catalyst, monosaccharide (fructose and glucose), disac-
charide (sucrose and maltose) and polysaccharide
(starch and cellulose) were selected as substrates in the
experiment and the results were exhibited in Table 5.
As shown in Table 5, Cr(III)-Al2O3 catalyst showed a
favorable catalytic activity for glucose and sucrose be-
sides fructose, attributed to the fact that Cr(III)-Al2O3
catalyst contained two types of acid sites (Lewis and
Brønsted acid sites). The hydrolysis of sucrose and the
isomerization of glucose into fructose were catalyzed
more easily by Lewis acid, whereas Brønsted acid fa-
vored the dehydration of fructose into HMF.[36,40] Un-
fortunately, the yields of HMF from starch and cellulose
were quite poor under the same reaction conditions,
possibly due to the complicacy in the hydrolysis of
these macromolecules.
Table 5 The influence of different sugars on the yield of the
reaction
Entry
1a
2b
3b
4b
Sugar
HMF yield/%
91.3
Fructose
Glucose
Sucrose
Maltose
Starch
73.7
84.1
48.3
5b
23.3
6b
Cellulose
12.2
a Conditions: Cr(III)-Al2O3 0.075 g, sugar 0.1 g, [Bmim]Cl 1.0 g,
120 ℃, 5 min. b Conditions: Cr(III)-Al2O3 0.075 g, sugar 0.1 g,
[Bmim]Cl 1.0 g, 120 ℃, 60 min.
Recycling of [Bmim]Cl and Cr(III)-Al2O3 catalyst
Long-term reusability was of great significance for em-
ploying a catalyst in industrial process. In this work, the
reuse of Cr(III)-Al2O3 catalyst and the [Bmim]Cl was
carried out to investigate the stability and activity of
catalyst. 1 mL of pure water should be added into the
reaction mixture after the 1st run in order to decrease the
viscosity of [Bmim]Cl and facilitate the extraction of
HMF. After the extraction of HMF by ethyl acetate, the
mixture was heated at 60 ℃ for 24 h, aiming to remove
the residual water and ethyl acetate after extraction. The
extracted liquid was examined by UV-Vis detector in an
external standard to attain the amount of HMF. Then the
dried [Bmim]Cl solvent with Cr(III)-Al2O3 catalyst was
utilized directly in the next run by adding 0.1 g fructose
under the same reaction conditions. As shown in Figure
8, the decrease of the catalytic activity was negligible
after 10th run, which demonstrated that the Cr(III)-Al2O3
Table 4 The influence of different initial fructose loadings on
the yield of the reactiona
Initial fructose
concentration/wt% loading/g
Initial fructose
Entry
HMF yield/%
1
2
3
4
5
4.8
0.05
0.10
0.20
0.30
0.40
96.0
93.1
85.9
84.1
82.5
9.1
16.7
23.1
28.6
a Conditions: Cr(III)-Al2O3 0.075 g, [Bmim]Cl 1.0 g, 100 ℃, 40
min.
440
www.cjc.wiley-vch.de
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2014, 32, 434—442

Production of 5-Hydroxymethylfurfural from Fructose in Ionic Liquid
catalyst exhibited an excellent stability for the fructose
dehydration into HMF. Surprisingly, the HMF yields in
some recycling tests were even higher than that in the 1st
run, but the change range was very slight, which might
be due to the acceptable experimental error.
References
[1] Deutschmann, R.; Dekker, R. F. H. Biotechnol. Adv. 2012, 30, 1627.
[2] Lange, J. P.; van der Heide, E.; van Buijtenen, J.; Price, R.
ChemSusChem 2012, 5, 150.
[3] Karinen, R.; Vilonen, K.; Niemela, M. ChemSusChem 2011, 4, 1002.
[4] Sádaba, I.; Gorbanev, Y. Y.; Kegnæs, S.; Putluru, S. S. R.; Berg, R.
W.; Riisager, A. ChemCatChem 2013, 5, 284.
[5] Gupta, N. K.; Nishimura, S.; Takagaki, A.; Ebitani, K. Green Chem.
2011, 13, 824.
[6] Gubbels, E.; Jasinska-Walc, L.; Koning, C. E. J. Polym. Sci., Part A:
Polym. Chem. 2013, 51, 890.
[7] Zhang, J.; Weitz, E. ACS Catal. 2012, 2, 1211.
[8] Deng, T. S.; Cui, X. J.; Qi, Y. Q.; Wang, Y. X.; Hou, X. L.; Zhu, Y. L.
Chem. Commun. 2012, 48, 5494.
[9] Zhang, Y. M.; Degirmenci, V.; Li, C.; Hensen, E. J. M.
ChemSusChem 2011, 4, 59.
[10] Zhang, Z. H.; Wang, Q.; Xie, H. B.; Liu, W. J.; Zhao, Z. B.
ChemSusChem 2011, 4, 131.
[11] Yang, Y.; Hu, C. W.; Abu-Omar, M. M. Green Chem. 2012, 14, 509.
[12] Jiang, N.; Huang, R. L.; Qi, W.; Su, R. X.; He, Z. M. Bioenerg. Res.
2011, 5, 380.
[13] De, S.; Dutta, S.; Saha, B. Green Chem. 2011, 13, 2859.
[14] Zhao, H. B.; Holladay, J. E.; Brown, H.; Zhang, Z. C. Science 2007,
316, 1597.
Figure 8 Recycle of IL and Cr(III)-Al2O3 catalyst.
Conclusions
[15] Hu, S. Q.; Zhang, Z. F.; Song, J. L.; Zhou, Y. X.; Han, B. X. Green
Chem. 2009, 11, 1746.
In summary, a series of metal-Al2O3 catalysts were
prepared easily via conventional impregnation method.
By screening different metal-Al2O3 catalysts, it was
demonstrated that Cr(III)-Al2O3 and [Bmim]Cl were an
excellent combination for the conversion of fructose
into HMF and a satisfactory HMF yield of 93.1% was
achieved at mild conditions (100 ℃, 40 min). The
Cr(III)-Al2O3 catalyst was characterized via XRD, DRS
and Raman spectra to clarify the interaction between the
Cr(III) and the alumina support. Meanwhile, the reac-
tion solvents after 1st reaction run and 5th reaction run
were collected through filtration and analyzed by
ICP-OES and LC-ITMS, which confirmed that no Cr(III)
ion was dropped off from the alumina support during
the dehydration of fructose. Furthermore, visible results
were also obtained when glucose, sucrose, maltose,
starch and cellulose were as the feedstock. Notably, a
favorable catalytic activity was observed in Cr(III)-
Al2O3/[Bmim]Cl system for the glucose and sucrose and
the HMF yields reached to 73.7% and 84.1% at 120 ℃
for 60 min, respectively. Meanwhile, the system with
Cr(III)-Al2O3 as catalyst and [Bmim]Cl as solvent can
be efficiently reused and the catalytic activity retained
steadily after 10 times recycling. Hence, the Cr(III)-
Al2O3 and [Bmim]Cl catalytic system has an excellent
potential in the synthesis of HMF from biomass for its
high efficiency, reusability and low price.
[16] Qi, X. H.; Watanabe, M.; Aida, T. M.; Smith, R. L. Bioresour.
Technol. 2012, 109, 224.
[17] Yi, Y. B.; Lee, J. W.; Choi, Y. H.; Park, S. M.; Chung, C. H. Biomass
Bioenergy 2012, 39, 484.
[18] Li, Y.; Liu, H.; Song, C. H.; Gu, X. M.; Li, H. M.; Zhu, W. S.; Yin,
S.; Han, C. R. Bioresour. Technol. 2013, 133, 347.
[19] Yi, Y. B.; Ha, M. G.; Lee, J. W.; Chung, C. H. Chem. Eng. J. 2012,
180, 370.
[20] Degirmenci, V.; Pidko, E. A.; Magusin, P. C. M. M.; Hensen, E. J. M.
ChemCatChem 2011, 3, 969.
[21] Zhang, Z. C.; Zhao, Z. B. Bioresour. Technol. 2011, 102, 3970.
[22] Doronkin, D. E.; Fogel, S.; Tamm, S.; Olsson, L.; Khan, T. S.;
Bligaard, T.; Gabrielsson, P.; Dahl, S. Appl. Catal., B. 2012, 113－
114, 228.
[23] Duarte, R. B.; Nachtegaal, M.; Bueno, J. M. C.; van Bokhoven, J. A.
J. Catal. 2012, 296, 86.
[24] Lang, S.; Türk, M.; Kraushaar-Czarnetzki, B. J. Catal. 2012, 286,
78.
[25] Wang, J. J.; Ren, J. W.; Liu, X. H.; Xi, J. X.; Xia, Q. N.; Zu, Y. H.;
Lu, G. Z.; Wang, Y. Q. Green Chem. 2012, 14, 2506.
[26] Zhang, S. M.; Chen, Y. Y.; Yu, Y.; Wu, H. H.; Wang, S. R.; Zhu, B.
L.; Huang, W. P.; Wu, S. H. J. Nanopart. Res. 2007, 10, 871.
[27] Galindo, R.; Badenes, J. A.; Llusar, M.; Tena, M. A.; Monrós, G.
Mater. Res. Bull. 2007, 42, 437.
[28] Šopotrajanov, B.; Stefov, V.; Žugić, M.; Petruševski, V. M. J. Mol.
Struct. 1999, 482, 109.
[29] Ayari, F.; Mhamdi, M.; Delahay, G.; Ghorbel, A. J. Porous Mater.
2010, 17, 265.
[30] Grzybowska, B.; Słoczyński, J.; Grabowski, R.; Wcisło, K.;
Kozłowska, A.; Stoch, J.; Zieliński, J. J. Catal. 1998, 178, 687.
[31] Gao, W. L.; Yang, T.; Wang, Y. X.; Li, G. B.; Liao, F. H.; Lin, J. H. J.
Phys. Chem. B 2005, 109, 22775.
Acknowledgment
[32] Boorman, P. M.; Kydd, R. A.; Sarbak, Z.; Somogyvari, A. J. Catal.
1985, 96, 115.
[33] Zhang, Y. M.; Pidko, E. A.; Hensen, E. J. Chem. Eur. J. 2011, 17,
5281.
[34] Bali, S.; Tofanelli, M. A.; Ernst, R. D.; Eyring, E. M. Biomass
Bioenergy 2012, 42, 224.
The authors thank the financial support for this
study by the National Natural Science Foundation of
China (Nos. 21106057 and 21376111), Society Devel-
opment Fund of Zhenjiang (SH2012009) and Advanced
Talents of Jiangsu University (10JDG143).
Chin. J. Chem. 2014, 32, 434—442
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
441

Song et al.
FULL PAPER
[35] Moliner, M.; Román-Leshkov, Y.; Davis, M. E. Proc. Natl. Acad. Sci.
U. S. A. 2010, 107, 6164.
[38] Xu, L.; Xia, J. X.; Wang, K.; Wang, L. G.; Li, H. M.; Xu, H.; Huang,
L. Y.; He, M. Q. Dalton Trans. 2013, 42, 6468.
[36] Zhao, S.; Cheng, M. X.; Li, J. Z.; Tian, J.; Wang, X. H. Chem.
Commun. 2011, 47, 2176.
[37] Panja, S. K.; Saha, S. RSC Adv. 2013, 3, 14495.
[39] Asghari, F. S.; Yoshida, H. Ind. Eng. Chem. Res. 2007, 46, 7703.
[40] Wang, J. J.; Xu, W. J.; Ren, J. W.; Liu, X. H.; Lu, G. Z.; Wang, Y. Q.
Green Chem. 2011, 13, 2678.
(Pan, B.)
442
www.cjc.wiley-vch.de
© 2014 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2014, 32, 434—442