Efficient production of 5-hydroxymethylfurfural and alkyl levulinate from biomass carbohydrate using ionic liquid-based polyoxometalate salts

Article abstract of DOI:10.1039/c3ra45632c

Direct conversion of fructose into 5-hydroxymethylfurfural (HMF) and alkyl levulinate is achieved by making use of ionic liquid-based polyoxometalate salts (IL-POMs) as solid acid catalysts. Among these solid acids, phosphotungstic acid-derived IL-POM shows the highest catalytic performance in both the HMF and ethyl levulinate (EL) formation. A study for optimizing the reaction conditions such as the reaction time and the temperature has been performed. High HMF and EL yields of up to 99% and 82%, respectively, are obtained from fructose under the investigated conditions. Moreover, the generality of the catalyst is further demonstrated by processing representative di- and polysaccharides such as sucrose and inulin with good yields to HMF (76% from inulin and 48% from sucrose) and EL (67% from inulin and 45% from sucrose), again under mild conditions, thereby eliminating the separate hydrolysis step before the dehydration reaction. The catalyst recycling experiment indicates that the adsorption and accumulation of oligomeric products on the catalyst surface results in a partial deactivation of catalyst. The mechanism research reveals that a major pathway for EL formation involves a fructose-to-HMF transformation followed by HMF etherification and rehydration of HMF-ether to give EL. The research highlights an efficient, environment-friendly and recyclable solid acid for biomass valorization.

Full text of DOI:10.1039/c3ra45632c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Efficient production of 5-hydroxymethylfurfural
and alkyl levulinate from biomass carbohydrate
using ionic liquid-based polyoxometalate salts†
Cite this: RSC Adv., 2014, 4, 4194
a
a
b
Jinzhu Chen,* Guoying Zhao and Limin Chen
Direct conversion of fructose into 5-hydroxymethylfurfural (HMF) and alkyl levulinate is achieved by making
use of ionic liquid-based polyoxometalate salts (IL-POMs) as solid acid catalysts. Among these solid acids,
phosphotungstic acid-derived IL-POM shows the highest catalytic performance in both the HMF and ethyl
levulinate (EL) formation. A study for optimizing the reaction conditions such as the reaction time and the
temperature has been performed. High HMF and EL yields of up to 99% and 82%, respectively, are obtained
from fructose under the investigated conditions. Moreover, the generality of the catalyst is further
demonstrated by processing representative di- and polysaccharides such as sucrose and inulin with
good yields to HMF (76% from inulin and 48% from sucrose) and EL (67% from inulin and 45% from
sucrose), again under mild conditions, thereby eliminating the separate hydrolysis step before the
dehydration reaction. The catalyst recycling experiment indicates that the adsorption and accumulation
of oligomeric products on the catalyst surface results in a partial deactivation of catalyst. The mechanism
research reveals that a major pathway for EL formation involves a fructose-to-HMF transformation
followed by HMF etherification and rehydration of HMF-ether to give EL. The research highlights an
efficient, environment-friendly and recyclable solid acid for biomass valorization.
Received 7th October 2013
Accepted 5th December 2013
DOI: 10.1039/c3ra45632c
www.rsc.org/advances
biomass-based carbohydrate, especially monosaccharide fruc-
Introduction
7,8
9
10
tose, using a variety of mineral acids, organic acids, acidic
11
12
13
14
Catalytic routes to fuels and chemicals from sustainable feed- resins, oxides, phosphates, and zeolites as catalysts.
stock have attracted much attention over the last few years, Dehydration was carried out in many solvents including water,
15
owning to diminishing fossil reserves as well as degradation of organic solvents, and biphasic mixed systems. Recently, the
1
–3
the environment.
provides an abundant alternative to fossil resources for the liquids (ILs) as catalysts or in ILs.
sustainable production of fuels and chemicals; particularly,
In addition to HMF, alkyl levulinate is well known to have
biomass carbohydrate is one of the largest components of widespread applications as odorous substances, plasticizing
Renewable and carbon-neutral biomass synthesis of HMF was developed rapidly with acidic ionic
16
6,17,18
woody biomass. Current research in this area has led to agents, and fuel additives.
Generally, the synthetic routes
signicant progress in exploring efficient routes for the for alkyl levulinate from biomass feedstock involve typically
19
conversion of carbohydrate into biomass-based platform either the esterication of levulinic acid with alcohol or a
chemical compound. Among these platform molecules, direct acid-catalyzed alcoholysis of biomass-based feedstocks
4,5
6
20–22
23–27
5
-hydroxymethylfurfural (HMF) and alkyl levulinates are of such as furfuryl alcohol
particular interest due to their versatile functions and benign former route requires acid catalyst such as mineral liquid acid
properties.
or solid acid for esterication of levulinic acid with corre-
and monosaccharide.
The
As a sustainable precursor for petrochemical industry and sponding alcohol. The one-step, direct alcoholysis of biomass
biofuel chemistry, HMF is obtained mainly by dehydration of feedstock to alkyl levulinate is preferable, as it avoids the
isolation and purication steps of the intermediate levulinic
acid. Moreover, the latter route shows a minimized wastewater
a
as well as higher grade products over the route of esterication
CAS Key Laboratory of Renewable Energy, Guangzhou Institute of Energy Conversion,
Chinese Academy of Sciences, Guangzhou, 510640, PR China. E-mail: chenjz@ms.giec. of levulinic acid.
ac.cn; Fax: +86-20-3722-3380; Tel: +86-20-3722-3380
In recent years, solid acid catalysts for direct production of
b
Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution
alkyl levulinate through biomass carbohydrate have undergone
an impressive development. Zhao and co-workers made
impressive progress on the direct alkyl levulinate production
from furfuryl alcohol using methylimidazolebutylsulfate
Control, College of Environment and Energy, South China University of Technology,
Guangzhou, 510006, PR China
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/c3ra45632c
1
4194 | RSC Adv., 2014, 4, 4194–4202
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
21
phosphotungstate ([MIMBS]
3
PW12O
40) as solid acid catalyst.
in one pot without isolation of intermediate of mono-
The plausible reaction pathway of furfuryl alcohol alcoholysis saccharides, and with excellent efficiency.
20
was subsequently revealed by Dumesic et al. Riisager and
co-workers did nice work on a direct conversion of mono- and
disaccharides to ethyl levulinate (EL) with sulfonic acid-func-
Experimental
Materials
23
24
tionalized ILs and sulfonic acid-functionalized SBA-15 as
catalysts. Lin and Liu reported a direct conversion of glucose to Unless otherwise stated, all chemicals in this work were
EL over sulfated zirconia catalysts with a moderate yield of 30% commercial available and used without further purication.
ꢀ
25,26
at 200 C.
Tominaga group reported mixed-acid systems of Keggin-type phosphotungstic acid (H PW O ), silicotungstic
3 12 40
Lewis and Brønsted acids for direct synthesis of methyl levuli- acid (H SiW O ), phosphomolybdic acid (H PMo O ), fruc-
4
12 40
3
12 40
27
nate from cellulose. Recently, we achieved a direct conversion tose, glucose, sucrose, methanol, ethanol, diethyl ether, and
of fructose into HMF and alkyl levulinate by using poly(p-styr- petroleum ether were purchased from Sinopharm Chemical
enesulfonic acid)-graed multi-walled carbon nanotubes as Reagent Co. Ltd. (Shanghai, PR China). Inulin, cellobiose, cellu-
8
solid acid catalyst. Therefore, a study on developing various lose, 5-hydroxymethylfurfural, ethyl levulinate, methyl levulinate,
kinds of green and efficient acid catalysts is very important.
n-propyl levulinate, n-butyl levulinate, n-propanol, n-butanol,
In recent years, solid acids, especially IL-based poly- dimethyl sulfoxide, toluene, tetramethylethylenediamine
oxometalate salts (IL-POMs), have attracted much attention due (TMEDA), 1,3-propanesulfonate, 1-methylimidazole, 2-methyl-
0
to the exible adaptability of both the inorganic and the organic imidazole, 4,4 -bipyridine, triphenylphosphine, methyl hex-
21,28–30
groups.
Tethering organic groups to POMs can realize adecanoate (internal standard), and sodium ethoxide were
heterogenization of heteropolyacid (HPA)-promoted reactions, purchased from Aladdin Industrial Inc. (Shanghai, PR China).
which has become an attractive topic for catalyst recycling. In Hydrogen gas (>99.999%) and nitrogen gas (>99.999%) were
particular, IL-cations with varieties of structures and properties obtained from Huate Co. Ltd (foshan, PR China). De-ionized pure
can be applied to organic blocks to modify POM catalysts with water from Millipore-Milli Q Plus System was used as solvent.
their acidity, polarity, solubility, redox properties and surface
structures. Herein we report on the behaviour of a series of Catalyst preparation
IL-POMs as solid acid catalysts for the fructose dehydration to
The ILs 3–7 were prepared according to literature procedure
HMF and one-pot conversion of fructose into alkyl levulinate
with slight modications (Scheme 2) (see the ESI for details†).
(
Scheme 1). The HPA and IL used in IL-POMs, the catalyst
The IL-POMs were prepared by the treatment of ILs 3–7 with
HPA in water. In typical synthetic procedure of
3$2H] [PW12 , IL 3 (0.36 g, 1 mmol) dissolved in water
5 mL) was slowly dropped into the aqueous solution (5 mL) of
loading and the reaction temperature were found to have
signicant effects on these catalytic systems. Under the optimal
conditions, high yields of HMF and EL of up to 99 and 82% were
obtained, respectively, from direct fructose conversion. More-
over, in comparison with commonly used monocationic
IL-POMs, dicationic IL-POMs in this research allow greater
variety and control of the properties of an IL-POM, including its
higher melting point and lower solubility in the reaction
system, which achieves its heterogenization and recycling. In
addition, by using preferred reaction conditions for represen-
tative di- and polysaccharides such as sucrose (from sugarcane)
and inulin (from chicory), that are abundantly available,
renewable and inexpensive, we can directly process these
saccharides with good yields to HMF (76% from inulin and 48%
from sucrose) and EL (67% from inulin and 45% from sucrose)
by using IL-POMs as catalysts. Notably, these processes thereby
eliminate the need to obtain simple carbohydrate molecules by
acid hydrolysis as a separate processing step. The catalyst
system described in this article thus achieves an integrated
chemical process where multi-step reactions are consolidated
a
[
(
3
40 2
O ]
H
3
PW12O40 (1.92 g, 0.67 mmol) under vigorous stirring. When it

nished, the mixture continued to stir at room temperature for
4 hours. Water was removed under vacuum, and the nal
product, [3$2H] [PW O ] was obtained as a white solid and
2
3
12 40 2
characterized by FT-IR spectroscopy. [3$2H] [PW O ] : FT-IR
3
12 40 2
(
KBr): n ¼ 1482 (C–H), 1228, 1155 (S]O), 1080, 1041 (P–O
a
), 981
Scheme 1 Conversion of fructose into HMF and alkyl levulinate.
Scheme 2 Synthesis of IL-POMs.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 4194–4202 | 4195

View Article Online
RSC Advances
Paper
ꢁ1
(
W–O
POMs were prepared following the same procedures as for
3$2H] [PW O ] .
d
), 978 (W ¼ O
d
), 899, 811 cm (P–Ob/c). The other IL- Table
1
Dehydration of fructose to HMF catalyzed by various
a
catalysts
[
3
12 40 2
Run
Catalyst
Conversion [%]
Yield [%]
General procedure for the reaction
1
2
3
4
5
6
7
8
9
[3$2H]
[4$2H]
3
[PW12
[PW12
O
O
40
40
]
]
2
2
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
92
92
91
90
90
90
89
88
86
83
48
42
38
39
40
3
Fructose conversion into HMF. In a typical run, fructose
[5$H]
3
[PW12
O
40
]
]
]
3 40 2
(50 mg), [3$2H] [PW12O ] (25 mg) and DMSO (2 mL) were
[6$H] [PW12O40
3
ꢀ
added to an Ace pressure tube. The mixture was heated at 100 C
with vigorous stirring for 1.0 h. Aer the reaction, the mixture
was ltered and then decanted into a volumetric ask using
[7$H]
3
[PW12
O
[SiW12
[SiW12
40
[3$2H]
[4$2H]
2
O
O
40
]
]
2
40
[5$H]
[6$H]
4
[SiW12
[SiW12
O
O
40
]
]
]
water as diluents, and analyzed by high-performance liquid
4
40
1
chromatography (HPLC). HMF: H NMR (400 MHz, D
2
O) d 9.33 10
4
[7$H] [SiW12O40
(
s, 1H), 7.40 (d, J ¼ 3.6 Hz, 1H), 6.55 (d, J ¼ 3.1 Hz, 1H), 4.57 11
[3$2H]
3
[PMo12
O
40
]
]
2
1
3
1
12
13
[4$2H] [PMo12O
3
40 2
(s, 2H). C { H} NMR (101 MHz, D
2
O) d 180.5, 161.3, 151.8,
ꢁ
+
1
[5$H]
[6$H]
[7$H]
3
3
3
[PMo12
[PMo12
[PMo12
O
O
O
40
]
]
]
1
26.8, 110.9, 56.0. IR (KBr, cm ): n ¼ 3377 (O–H), 1670 (C]O).
1
1
4
5
40
40
Mass spectrum (EI) m/z 126 (M ).
Fructose conversion into EL. In a typical run, fructose
a
Reaction conditions: fructose (50 mg, 0.28 mmol), catalyst (2.5 mol%
(
50 mg), [3$2H] [PW O ] (50 mg) and ethanol (4 mL) were
ꢀ
3
12 40 2
relative to fructose), DMSO (2 mL), 100 C, 1.0 h.
added to an Ace Pressure Tube. The mixture was heated at
ꢀ
1
20 C with vigorous stirring for 12 h. Aer the reaction, the
mixture was ltered and analyzed by GC, GC-MS and HPLC. EL:
1
H NMR (400 MHz, CDCl
3
) d 4.12 (q, J ¼ 7.1 Hz, 2H), 2.74 (t, J ¼
(
Scheme 1). Generally, tungsten IL-POMs (IL ¼ 3–7), such as
6
.5 Hz, 2H), 2.56 (t, J ¼ 6.6 Hz, 2H), 2.18 (s, 3H), 1.24 (t, J ¼ 7.1
IL-H [PW12 40] and IL-H [SiW12 40], showed excellent to good
3
O
4
O
1
3
1
Hz, 3H). C{ H} NMR (101 MHz, CDCl ) d 206.7, 172.7, 60.6,
3
spectrum (EI) m/z 144 (M ).
3
activity toward the fructose dehydration with HMF yield ranging
from 92 to 83% (Table 1, Runs 1–10). In the case of molyb-
denum IL-POMs, the HMF was obtained with very low yields
ꢁ
1
7.9, 29.8, 27.9, 14.1. FT-IR (KBr, cm ): n ¼ 1726 (C]O). Mass
+
(
Table 1, Runs 11–15). The main by-products in the fructose
Analytic methods
dehydration reaction were polymeric humans, which were not
1
H NMR spectra were recorded on a Bruker AVANCE Digital 400 observed in HPLC analyses. Currently, humins could not be
ꢀ
31
MHz spectrometer at 25 C with D O or CDCl as solvent. IR quantied according to their composition. The inuence of
2
3
spectra were measured on a Bruker Tensor 27 FT-IR spectrom- various HPAs in the IL-POM on fructose dehydration thus
eter as KBr pellets. The HPLC analysis was performed on Shi- revealed that with a same IL organic cation, the catalytic activ-
madzu LC-20AT equipped with a UV-Vis/refractive index ities were in the order of IL-H [PW O ] > IL-H [SiW O ] [
3
12 40
4
12 40
detector and a Shodex Sugar SP-0810 column (ø 8 ꢂ 300 mm). IL-H [PMo O ], which is in accordance with that of the
3
12 40
3
2
H
2
SO
4
(0.005 M) solution was used as the mobile phase at a ow Brønsted acidity of HPAs. In addition to relatively lower acid
ꢁ1
rate of 0.5 mL min , and the column temperature was main- strength, lower thermal stability, and higher oxidation potential
tained at 50 C. The amounts of fructose and HMF were of [PMo O ] cation compared with tungsten heteropolyacid
calculated based on external standard curves constructed with cations ([PW O ] and [SiW O ] ) are presumably related
authentic standards. The GC analysis was performed by Agilent to low catalytic performance of IL-H [PMo O ]. Nevertheless,
ꢀ
3+
1
2 40
3
+
4+
1
2
40
12 40
32
3
12 40
6
890 or Shimadzu 2010 equipped with a ame ionization for the IL-POMs with the same heteropolyanion, various
detector (FID) and a KB-5 capillary column (internal diameter protonated IL cation groups have very limited inuence on the
.32 mm, length 30 m) using nitrogen as the carry gas. The GC- catalytic activity of IL-POMs.
MS analysis was performed by Trace GC-MS 2000 as well as by
The inuence of catalyst loading amounts on the fructose-to-
0
comparing the retention times to respective standards in GC HMF transformation was investigated by varying the amount of
traces. The operating conditions for GC and GC-MS were as [3$2H] [PW O ] ranging from blank to 5 mol% based on
3
12 40 2
ꢀ
follows: injector port temperature, 260 C; column temperature, fructose used (Fig. 1). As expected, the more catalyst there is, the
initial temperature 50 C (2 min), gradient rate 30 C min higher both the conversion of fructose and the yield of HMF are.
7 min), nal temperature 260 C (3 min), ow rate 75 mL The conversion of fructose is enhanced from 24% to 100% in
min . The amounts of products were calculated based on 30 min. The yields of HMF could reach to maxima of 93% with
internal standard curves constructed with authentic standards. 5 mol% of catalyst. Under the investigated reaction conditions,
ꢀ
ꢀ
ꢀ
(
ꢁ
1
2
.5 mol% of catalyst was required to reach a complete conver-
sion of fructose within 30 min with an HMF yield of 90%
Fig. 1). The increase in the HMF yield with an increasing
catalyst dosage should be attributable to an increase in the
Results and discussion
Conversion of fructose into HMF
(
Table 1 shows the dehydration of fructose to HMF in DMSO at availability and number of catalytically active sites of
ꢀ
1
00 C for 1 h with different IL-POMs as solid acid catalysts [3$2H]
3
[PW12
O
40
]
2
during dehydration of fructose.
4196 | RSC Adv., 2014, 4, 4194–4202
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
phase reactor system of water and methylisobutylketone (MIBK)
36
was explored with solid Ag PW O as catalyst. The above
3
12 40
researches thus developed POM-based catalyst system for fruc-
tose-to-HMF transformation with IL or bi-phase solvent as an
alternative to DMSO.
Conversion of fructose into alkyl levulinate
In addition to fructose-to-HMF transformation, in order to
explore the application scope of IL-POM catalyst, a one-pot
conversion of fructose into alkyl levulinates was investigated
(
Scheme 1). Table 2 shows the one-pot conversion of fructose
ꢀ
into ethyl levulinate (EL) catalyzed by various IL-POMs at 120 C
in ethanol (Scheme 1). Initial experiments were performed with
Fig. 1 Influence of the catalyst amount on fructose dehydration.
IL-H
3
[PW12
O
40] (IL ¼ 3–7) as solid acid catalyst. In all these
(
Reaction conditions: fructose, 50 mg, 0.28 mmol; catalyst, specified
ꢀ
reactions EL was found to be the predominant product with a
yield ranging from 80 to 74% (Table 2, Runs 1–5). The main by-
products in the reaction were HMF, 5-ethoxymethylfurfural
amount of [3$2H]
3
[PW12O
40
]
2
; DMSO, 2 mL; 100 C; 30 min.)
(
EMF), and 2-(diethoxymethyl)-5-(ethoxymethyl)furan (DEEF),
Fig. 2 shows the inuence of reaction temperature and time
40 2
on the dehydration of fructose catalyzed by [3$2H] [PW12O ]
which were observed in GC and GC-MS analyses (Fig. S4†). To
evaluate the inuence of the HPA on the formation of EL,
IL-H
were introduced into the reaction as catalyst as well. The
IL-H [SiW O ] gave a relatively lower EL yield than the corre-
3
using DMSO as solvent. Both the conversion of fructose and the
yield of HMF increased with reaction time at all temperatures
investigated. The yield of HMF increased slowly to 55% aer
4
[SiW12
O40] (IL ¼ 3–7) and IL-H
3
[PMo12
O
40] (IL ¼ 3–7)
ꢀ
6
1
0 min at 60 C, while it increased quickly to 97% aer 15 min at
4
12 40
ꢀ
sponding IL-H [PW O ] analogue (Table 2, Runs 1–10). In the
3 12 40
case of IL-H [PMo O ], the observed EL yields were remark-
00 C, conrming that increasing the reaction temperature
promotes the conversion of fructose to HMF. Moreover, an HMF
3
12 40
ably inferior under the investigated conditions (Table 2, Runs
11–15). The inuence of various HPAs in the IL-POMs on fruc-
tose alcoholysis to EL revealed a similar trend to fructose-to-
HMF transformation.
yield of 99% with a full fructose conversion was obtained in
ꢀ
DMSO using [3$2H]
reaction time of 30 min (Fig. 2).
It is well known that the presence of DMSO can stabilize the
3
[PW12
O
40
]
2
as catalyst at 100 C aer a
33,11b
3 12 40
When H PW O was used as a catalyst, an EL yield of 84%
HMF and increase its selectivity and yield;
however, this
with a full fructose conversion was obtained (Table 2, Run 16).
However, in this system, H PW O behaved homogeneously
approach necessitates energy intensive isolation procedures
34
owning the high boiling point of DMSO. Recently, phospho-
tungstic acid/metal–organic framework hybrid (PTA/MOFs) was
developed as a novel solid acid catalyst for fructose dehydration
3
12 40
35
Table 2 Conversion of fructose into alkyl levulinate catalyzed by
to HMF in 1-ethyl-3-methylimidazolium chloride (EMIMCl). In
addition, the catalytic conversion of fructose into HMF in a two-
a
various catalysts
ꢀ
Run Catalyst
R
T [ C] Conversion [%] Yield [%]
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
[3$2H]
[4$2H]
3
[PW12
[PW12
O
O
40
]
]
2
2
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
120
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
80
76
79
76
74
72
67
67
66
60
33
30
47
28
32
84
75
83
77
3
40
[5$H]
[6$H]
[7$H]
3
[PW12
3
[PW12
3
[PW12
O
O
O
40
]
]
]
40
40
[3$2H]
[4$2H]
2
[SiW12
O
O
40
]
]
2
[SiW12
40
[5$H]
[6$H]
[7$H]
4
[SiW12
4
[SiW12
4
[SiW12
O
O
O
40
]
]
]
40
40
0
1
2
3
4
5
6
7
8
9
[3$2H]
[4$2H]
3
[PMo12
O
O
40
40
]
]
2
2
3
[PMo12
[5$H]
[6$H]
[7$H]
3
[PMo12
3
[PMo12
3
[PMo12
O
O
O
40
]
]
]
40
2
40
2
3 40
H PW12O
[3$2H]
[3$2H]
[3$2H]
3
[PW12
3
[PW12
3
[PW12
O
40
O
40
O
40
]
]
]
2
2
2
Me 100
n
Pr 120
Fig. 2 Influence of the temperature and time on fructose dehydration
to HMF. Reaction conditions: fructose (20 mg, 0.11 mmol) and
n
Bu 120
a
[
3$2H]
1 mL).
3
[PW12O
]
40 2
(10 mg, 2.5 mol% relative to fructose) in DMSO
Reaction conditions: fructose (50 mg, 0.28 mmol), catalyst (5.0 mol%
(
relative to fructose), alcohol (4 mL), 12 h.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 4194–4202 | 4197

View Article Online
RSC Advances
Paper
Table 3 Recycling of [3$2H]
into EL
3
[PW12
O
40
]
2
in the fructose conversion washed with ethanol and then recycled for further reaction.
The yield of EL decreased from 80% to 62%, indicating that
a
the catalyst partially lost activity during the reaction process.
Run
Number of previous cycles
Conversion [%]
Yield [%]
A similar phenomena of catalyst deactivation was also
2
1
8,37
1
2
3
4
5
0
1
2
3
4
>99
>99
>99
>99
>99
80
78
73
68
62
previously reported study by Zhao and us. It is important
to note that the reusability of [3$2H] [PW12 is still higher
than that achieved with regards to [MIMBS] 40. A fast
deactivation process of furfuryl alcohol conversion into n-
butyl levulinate was observed with [MIMBS]
catalyst. The catalyst recycling showed
decreased yields of n-butyl levulinate from 83 to 65% for the
3
40 2
O ]
3
PW12O
3
PW12O40 as
a
3 40 2
Reaction conditions: fructose (50 mg, 0.28 mmol), [3$2H] [PW12O ]
sharply
ꢀ
50 mg, 5.0 mol% relative to fructose), ethanol (4 mL), 120 C, 12 h.
(
2
1
second run.
As illustrated in Fig. 3, the prole of FT-IR spectra for the
recovered [3$2H] [PW12 aer the h run was well
consistent with that of the fresh one. Both the fresh and the
in the ethanol, and the separation and recovery of the
40 from the reaction medium was problematic. On the
3
40 2
O ]
3
H PW12O
other hand, all of the dicationic IL-POMs (IL ¼ 3–5) developed
in this research are insoluble in ethanol leading to heteroge-
neous catalysis, which can be readily recovered and reused.
Therefore, in contrast to the HPA, the propylsulfonic acid group
in the organic cation of IL-POMs provides the active acid site
which is responsible for catalytic performance. Whereas, owing
to high valence and large volume, the heteropolyanions of
IL-POMs can trap the ILs and show a high melting point as well
recovered [3$2H]
3
[PW12
O
40
]
2
show characteristic bands corre-
2+
ꢁ1
sponding to [3$2H] (1481 cm
assigned to CH
3
, 1227–
ꢁ1
3ꢁ
1
153 cm assigned to SO ) and [PW O ] species (observed
3
12 40
ꢁ1
from 1080 to 811 cm ) (Fig. S1†). Notably, the spectrum of the
recovered catalyst shows peaks at 1717 and 1623 cm , assigned
to the stretching frequencies of C]O groups (Fig. 3b). This
indicated the presence of organic carbonyl species on the
ꢁ1
3 40 2
surface of [3$2H] [PW12O ] aer reuse. Therefore, our results
28
as low solubility.
suggest that the adsorption and accumulation of oligomeric
products on the catalyst surface, presumably blocking acidic
catalytic sites, was the main reason for the partial deactivation
of [3$2H] [PW O ] .
In addition to EL, in order to explore the scope of this
method (Scheme 1), the preparations of methyl levulinate,
n-propyl levulinate, and n-butyl levulinate were achieved as well
with corresponding alkyl alcohol as solvent (Table 2, Runs 17–
3
12 40 2
Reaction pathways for the [3$2H] [PW O ] -catalyzed
3
12 40 2
1
9). Previously, [MIMBS] PW O was used as solid acid cata-
3 12 40
fructose-to-EL transformation in ethanol were further investi-
gated using liquid chromatography-mass spectrometry (LC-MS)
and gas chromatography-mass spectrometry (GC-MS). From
LC-MS and GC-MS analysis, three species of HMF (m/z 126.1),
EMF (m/z 154.0), and DEEF (m/z 228.1) were observed to be
formed and later consumed in the interval of time, suggesting
an intermediate behavior; whereas, EL (m/z 144.0) was observed
to increase with time (Fig. S4†). The GC analysis and a plausible
pathway for EL formation are shown in Fig. 4 and Scheme 3,
respectively.
lyst for alcoholysis of furfuryl alcohol to give alkyl levulinate.
Methyl levulinate and n-propyl levulinate were obtained with
21
yields of 5% and 80%, respectively. In our case, yields of 75%
for methyl levulinate, and 83% for n-propyl levulinate were
achieved by direct alcoholysis of fructose, further indicating a
clean and facile catalytic process.
In order to demonstrate the reusability of catalyst, a ve-
cycle experiment of the [3$2H] [PW O ] was investigated
3
12 40 2
for the fructose-to-EL transformation as described in Table 3.
The [3$2H] [PW12 catalyst was recovered by lter,
3
40 2
O ]
Fig. 4 GC analysis of EL formation (Reaction conditions: fructose, 50
Fig. 3 FT-IR of (a) the fresh [3$2H]
3$2H] [PW12
3
[PW12
O
40
]
2
, and (b) the recovered mg, 0.28 mmol; [3$2H]
3
[PW12O
]
40 2
, 50 mg, 5.0 mol%; ethanol, 4 mL;
ꢀ
[
3
O
]
40 2
after a five-cycle in the fructose conversion into EL. 120 C).
4198 | RSC Adv., 2014, 4, 4194–4202
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Scheme 3 Plausible pathway for the formation of EL from fructose.
In principle, our reported results seemed to agree with the
mechanism of fructose alcoholysis proposed by Riisager and
23
co-workers with sulfonic acid-functionalized ILs as catalysts.
According to the literature, a plausible major pathway for EL
formation involves a fructose-to-HMF transformation followed
by HMF etherication and rehydration of HMF-ether to give EL
(Scheme 3). Moreover, in our case, EMF was seen to be in a
reversible equilibrium with DEEF via acetalisation, the rehy-
dration of DFFE was presumably another minor pathway to
form EL (Fig. 4 and Scheme 3). When xylose and furfural were
subjected to the reaction conditions, the expected product EL
was unobserved. However, EL was quantitatively obtained with
furfuryl alcohol or HMF as starting materials, respectively.
These results thus further indicate that the etherication step of
HMF to EMF plays a key role for EL formation.
Fig. 5 shows the inuences of temperature and time on the
conversion of the fructose into EL in the presence of ethanol
with [3$2H] [PW O ] as solid catalyst. The conversion of
3
12 40 2
fructose increased with time at all temperatures investigated
Fig. 5a). Moreover, with a xed reaction time point, the fructose
(
conversion increased as the temperature increased. The
conversion of fructose increased slowly to 97% aer 24 h at
ꢀ
ꢀ
8
0 C, while it increased quickly to 99% aer 1 h at 140 C.
As shown in Scheme 3, both HMF and EMF are “interme-
diates” to the nal product EL, the inuences on both the
formation and consumption of HMF (Fig. 5b) and EMF (Fig. 5c)
were observed as well during the conversion of the fructose into
EL. Generally, the HMF yields increased as the temperature
ꢀ
increased with a maximum of 74% at 120 C (Fig. 5b). A further
ꢀ
increased temperature to 140 C, however, led to a slightly
reduced HMF yield to 58%, indicating acid catalyzed HMF
degradation under higher reaction temperature. The reaction
times for maximum yields of “intermediate” HMF are different,
based on the different reaction temperatures; that is, the higher
the temperature is, the shorter the reaction time needs to be. At
Fig. 5 Influence of the reaction temperature and time on fructose
conversion into EL. (a) The conversion of fructose, (b) the yield of HMF,
(c) the yield of EMF, (d) the yield of EL. (Reaction conditions: fructose,
ꢀ
the higher temperature of 120 C, the yields of HMF reach
5
0 mg, 0.28 mmol; [3$2H] [PW12 , 50 mg, 5.0 mol%; ethanol, 4
3
40 2
O ]
maximum values of 74% in 5 min. The HMF yields decrease to mL; specified temperature).
ꢀ
3
3% at 100 C in the same time and reach a maximum of 64% in
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 4194–4202 | 4199

View Article Online
RSC Advances
0 min. For a reaction temperature of 80 C, the reaction time
Paper
ꢀ
3
Fig. 6 shows a typical reaction prole obtained using
was much longer to 60 min as the yields of HMF reach [3$2H] [PW O ] for the catalytic transformation of the fruc-
3
12 40 2
ꢀ
maximum values.
tose to EL at 100 C, which further reveals the inuences of
The effect of the reaction temperature on EMF is similar to reaction time on EL formation. Fructose is immediately con-
that of HMF (Fig. 5c). With the HMF as the “intermediate” for verted to HMF. The yield of HMF has a maximum value of 64%
the subsequently generated EMF, the reaction time was, aer 30 min. Aer 3 hours, the second transformation of the
however, much longer as the yields of EMF reach maximum HMF to EMF reaches a maximum yield of 17%. Aer this time,
values than that of HMF. For example, the yields of HMF reach the yield of EMF decreases with the concomitant production of
ꢀ
maximum values of 74% in 5 min at 120 C. Aer 60 min, the EL. Therefore, the yield trends of both HMF and EMF increase
second transformation of the HMF to EMF reaches a maximum rst, and then decrease with reaction time. However, both the
yield of EMF at 17%.
fructose conversions and EL yields increase as the reaction
In the case of nal production EL, as discussed in Scheme 3, time. These results further suggest that the conversion of fruc-
the one-pot conversion of fructose with ethanol requires the tose to EL is going stepwise with observed HMF and EMF as the
formation of HMF rstly and EMF secondly, which then was “intermediates”.
converted into the nal product EL. When the reaction was
ꢀ
carried out at 80 C, the rate for the EL formation was rather
Conversion of other feedstocks to HMF and EL
slow (Fig. 5d). Even with a prolonged reaction time, we only
obtained EL in a low yield of 26% within 24 h. Moreover, EMF
remained unchanged at 80 C within 24 h under the investi-
gated conditions (Fig. 5c), indicating that the EMF-to-EL
transformation was difficult at 80 C. When the reaction
temperature was elevated to 100 C, both the EMF conversion
Fig. 5c) and the EL yield (Fig. 5d) enhanced signicantly. In
comparison with the reaction at 80 C, the yield of EL within
3 40 2
Besides fructose, the [3$2H] [PW12O ] catalyst was further
ꢀ
extended to fructose based di- and polysaccharides such as
sucrose and inulin, and the results of HMF and EL formations
are presented in Table 4. Both sucrose and inulin are the most
abundant and renewable saccharides. When inulin was used,
the HMF yield reached 76%, which is higher than the HMF yield
of 48% obtained from sucrose (Table 4, Runs 1–2). According to
the literature, one molecule of inulin contains one molecule of
glucose and 1–59 molecules of fructose; whereas, one molecule
of sucrose is composed of one molecule of glucose and one
ꢀ
ꢀ
(
ꢀ
2
4 h increased sharply, from 26% to 65%. When the reaction
ꢀ
temperature was further increased to 120 C, its inuence was
much more obvious. As shown in Fig. 5c and d, intermediate
EMF was almost completely converted into the product EL in a
yield of 80% within just 12 h at 120 C. A further increased
temperature to 140 C led to a reduced reaction time to 10 h to
38,39
molecule of fructose.
obtained from inulin. These results further indicate that
[3$2H] [PW12 effectively promotes the hydrolysis of
Thus much higher HMF yield was
ꢀ
ꢀ
3
O ]
40 2
sucrose and inulin under mild conditions. When glucose,
cellobiose, and a-cellulose were subjected to our reaction
conditions, the HMF yield was remarkably inferior (Table 4,
Runs 3–5). This result clearly indicates that only fructose in
sucrose and inulin was converted to HMF, and
reach a full conversion of EMF with the EL yield of 80%. Fig. 5d
ꢀ
further reveals higher EL yields at 120 C aer 13.5 h than in the
ꢀ
case of 140 C, presumably owing to an acid catalyzed EL
degradation under higher reaction temperature as a prolonged
reaction time. The above results demonstrated that the optimal
reaction temperature for the one-pot fructose conversion into
3 40 2
[3$2H] [PW12O ] , functions as Brønsted acid, can hardly
ꢀ
promotes the isomerization of glucose to fructose.
EL was 120 C with a maximum EL yield of 82% and an optimal
In the case of EL production, the catalyst [3$2H] [PW O ]
3
12 40 2
reaction time of 18 h.
yielded EL of 67% from inulin (Table 4, Run 6), 45% from
a
b
Table 4 Conversion of other feedstocks to HMF and EL catalyzed by
3$2H] [PW12
[
3
40 2
O ]
ꢀ
Run
Substrate
T [ C]
t [h]
Yield [%]
a
1
Inulin
Sucrose
Glucose
Cellobiose
a-Cellulose
Inulin
Sucrose
glucose
Glucose
Cellobiose
a-Cellulose
120
120
120
150
150
120
120
120
150
150
150
2
2
2
2
2
12
12
24
24
24
24
76
48
18
30
10
67
45
7
2
3
4
5
b
6
7
8
9
20
18
14
1
1
0
1
a
Fig. 6 Product distribution for transformation of fructose into EL.
Reaction conditions: HMF production (Run 1–5), substrate (50 mg),
b
(
Reaction conditions: fructose, 50 mg, 0.28 mmol; [3$2H]
3
ꢀ
[PW12
O
40 2
] ,
3 40 2
[3$2H] [PW12O ] , (25 mg), DMSO (1 mL). EL production (Run 6–
11), substrate (50 mg), [3$2H] [PW12O ] (50 mg), ethanol (4 mL).
3 40 2
5
0 mg, 5.0 mol% relative to fructose; ethanol, 4 mL; 100 C).
4200 | RSC Adv., 2014, 4, 4194–4202
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
sucrose (Table 4, Run 7), and only 7% from glucose (Table 4,
Run 8), respectively. Even under the conditions of increased
reaction temperature and prolonged reaction time, the EL yield,
obtained from glucose, cellobiose, and a-cellulose, were still
remarkable inferior (Table 4, Runs 9–11). Both sucrose and
inulin can be readily obtained from biomass such as sugarcane
7 A. Ranoux, K. Djanashvili, I. W. C. E. Arends and
U. Hanefeld, ACS Catal., 2013, 3, 760–763.
8 R. L. Liu, J. Z. Chen, X. Huang, L. M. Chen, L. L. Ma and
X. J. Li, Green Chem., 2013, 15, 2895–2903.
9 Y. Rom ´a n-Leshkov, J. N. Chheda and J. A. Dumesic, Science,
2006, 312, 1933–1937.
sugarbeet, chicory, and cassava. For these reason, this process is 10 F. S. Asghari and H. Yoshida, Ind. Eng. Chem. Res., 2006, 45,
an economically viable and environmentally attractive route to 2163–2173.
produce HMF and EL as sustainable chemicals using renewable 11 (a) Z.-Z. Yang, J. Deng, T. Pan, Q.-X. Guo and Y. Fu, Green
biomass as the raw material in one-pot reaction under reason-
ably mild reaction conditions.
Chem., 2012, 14, 2986–2989; (b) K.-I. Shimizu, R. Uozumi
and A. Satsuma, Catal. Commun., 2009, 10, 1849–1853; (c)
J. N. Chheda and J. A. Dumesic, Catal. Today, 2007, 123,
5
9–70; (d) X. H. Qi, M. Watanabe, T. M. Aida and
Conclusions
R. L. Smith, Jr, Green Chem., 2008, 10, 799–805.
In conclusion, IL-POMs are prominent solid acid catalysts for 12 M. Watanabe, Y. Aizawa, T. Iida, R. Nishimura and
one-pot conversion of fructose into HMF and alkyl levulinate. H. Inomata, Appl. Catal., A, 2005, 295, 150–156.
The catalytic activities of IL-POMs for fructose dehydration to 13 Y. Rom ´a n-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic,
HMF as well as fructose alcoholysis to EL followed the order of
Nature, 2007, 447, 982–985.
their acid strength. A major pathway for EL formation involves a 14 V. V. Ordomsky, J. van der Schaaf, J. C. Schouten and
fructose-to-HMF transformation followed by HMF ether- T. A. Nijhuis, J. Catal., 2012, 287, 68–75.
ication and rehydration of HMF-ether to give EL. Aer the 15 F. W. Lichtenthaler, Acc. Chem. Res., 2002, 35, 728–737.
reaction, the catalyst can easily be separated from the reaction 16 (a) D. A. Kotadia and S. S. Soni, Catal. Sci. Technol., 2013, 3,
mixture for reuse. Moreover, the developed method herein
provides a catalytic process for converting abundant, cheap and
renewable di- and polysaccharide such as sucrose and inulin to
biomass-based ne chemicals HMF and EL using recoverable
and cheap IL-POMs as catalysts. In addition, these processes
eliminate the need to obtain simple monosaccharide molecules
469–474; (b) H. Zhao, J. E. Holladay, H. Brown and
Z. C. Zhang, Science, 2007, 316, 1597–1600; (c) T. St ˚a hlberg,
W. J. Fu, J. M. Woodley and A. Riisager, ChemSusChem,
2011, 4, 451–458; (d) T. Okano, K. Qiao, Q. X. Bao,
D. Tomida, H. Hagiwara and C. Yokoyama, Appl. Catal., A,
2013, 451, 1–5.
by acid hydrolysis as a separate processing step. Therefore, solid 17 A. Lee, N. Chaibakhsh, M. B. A. Rahman, M. Basri and
acids IL-POMs are a step toward the more economical produc- B. A. Tejo, Ind. Crops Prod., 2010, 32, 246–251.
tion of HMF and EL through the processing of inexpensive, 18 A. Demirbas, Appl. Energy, 2011, 88, 17–28.
abundantly available and renewable biomass feedstocks.
19 F. Su, L. Ma, D. Y. Song, X. H. Zhang and Y. H. Guo, Green
Chem., 2013, 15, 885–890.
2
0 G. M. G. Maldonado, R. S. Assary, J. A. Dumesic and
L. A. Curtiss, Energy Environ. Sci., 2012, 5, 8990–8997.
Acknowledgements
This work was supported by National Natural Science Founda- 21 Z. H. Zhang, K. Dong and Z. B. Zhao, ChemSusChem, 2011, 4,
tion of China (21172219 and 21207039), National Basic 112–118.
Research Program of China (973 Program, 2012CB215304), 22 J.-P. Lange, W. D. van de Graaf and R. J. Haan,
Scientic Research Foundation for the Returned Overseas ChemSusChem, 2009, 2, 437–441.
Chinese Scholars (State Education Ministry), and Guangdong 23 S. Saravanamurugan, O. N. Van Buu and A. Riisager,
Natural
Science
Foundation
(S2013010012986
and
ChemSusChem, 2011, 4, 723–726.
S2011010002274).
24 S. Saravanamurugan and A. Riisager, Catal. Commun., 2012,
17, 71–75.
2
5 L. C. Peng, L. Lin, J. H. Zhang, J. B. Shi and S. J. Liu, Appl.
Catal., A, 2011, 397, 259–265.
Notes and references
1
2
P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538–1558.
A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411–
502.
G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106,
044–4098.
A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and
C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
R.-J. van Putten, J. C. van der Waal, E. de Jong,
C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev.,
26 L. C. Peng, L. Lin, H. Li and Q. L. Yang, Appl. Energy, 2011, 88,
4590–4596.
27 K.-I. Tominaga, A. Mori, Y. Fukushima, S. Shimada and
K. Sato, Green Chem., 2011, 13, 810–812.
28 Y. Leng, J. Wang, D. R. Zhu, X. Q. Ren, H. Q. Ge and L. Shen,
Angew. Chem., Int. Ed., 2008, 48, 168–171.
29 R. Kreiter, M. D. A. Rietkerk, H. L. Castricum, H. M. van
Veen, J. E. ten Elshof and J. F. Vente, ChemSusChem, 2009,
2, 158–160.
2
3
4
5
4
2
013, 113, 1499–1597.
J. Zhang, S. B. Wu, B. Li and H. D. Zhang, ChemCatChem,
012, 4, 1230–1237.
30 (a) Z. Sun, M. X. Cheng, H. C. Li, T. Shi, M. J. Yuan,
X. H. Wang and Z. J. Jiang, RSC Adv., 2012, 2, 9058–9065;
(b) K. X. Li, L. Chen, H. L. Wang, W. B. Lin and Z. C. Yan,
6
2
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 4194–4202 | 4201

View Article Online
RSC Advances
Paper
Appl. Catal., A, 2011, 392, 233–237; (c) M. J. dos Reis,
V. Pr ´e vot, F. Leroux, F. Silv ´e rio and J. B. Valim, J. Porous
Mater., 2010, 17, 443–451.
F. Qin, H. Xu and W. Shen, ChemSusChem, 2013, 6, 1063–
1069.
35 Y. M. Zhang, V. Degirmenci, C. Li and E. J. M. Hensen,
ChemSusChem, 2011, 4, 59–64.
3
1 (a) X. H. Qi, H. X. Guo, L. Y. Li and R. L. Smith Jr,
ChemSusChem, 2012, 5, 2215–2220; (b) K.-I. Shimizu, 36 C. Y. Fan, H. Y. Guan, H. Zhang, J. H. Wang, S. T. Wang and
R. Uozumi and A. Satsuma, Catal. Commun., 2009, 10,
849–1853.
2 (a) I. V. Kozhevnikov, Appl. Catal., A, 2003, 256, 3–18; (b)
M. N. Timofeeva, Appl. Catal., A, 2003, 256, 19–35.
3 J. N. Chheda, Y. Rom ´a n-Leshkov and J. A. Dumesic, Green
Chem., 2007, 9, 342–350.
X. H. Wang, Biomass Bioenergy, 2011, 35, 2659–2665.
37 J. Z. Chen, S. P. Wang, J. Huang, L. M. Chen, L. L. Ma and
X. Huang, ChemSusChem, 2013, 6, 1545–1555.
38 S. Sirisansaneeyakul, N. Worawuthiyanan, W. Vanichsriratana,
P. Srinophakun and Y. Chisti, World J. Microbiol. Biotechnol.,
2007, 23, 543–552.
1
3
3
3
4 (a) F. W. Lichtenthaler and S. R ¨o ninger, J. Chem. Soc., Perkin 39 Y. Yang, M. M. Abu-Omar and C. W. Hua, Appl. Energy, 2012,
Trans. 2, 1990, 1489–1497; (b) Z. Huang, W. Pan, H. Zhou, 99, 80–84.
4202 | RSC Adv., 2014, 4, 4194–4202
This journal is © The Royal Society of Chemistry 2014