Direct conversion and NMR observation of cellulose to glucose and 5-hydroxymethylfurfural (HMF) catalyzed by the acidic ionic liquids

Article abstract of DOI:10.1016/j.molcata.2010.10.006

The hydrolysis of cellulose was catalyzed over a variety of acidic ionic liquids (ILs). It is found that the hydrolysis activity is directly associated with the acidity of catalysts, as evidenced by IR spectroscopy. ¹³C NMR characterization results confirm the majority product of cellulose hydrolysis is glucose, and the resulting carbohydrates undergo further degradation, possibly also catalyzed by the acidic ILs, to 5- hydroxymethylfurfural (HMF). Moreover, in situ ¹³C NMR measurements clearly exhibit that the evolution of products is dependent on the reaction process. We attempt to study the kinetics of cellulose hydrolysis over the most active catalyst of [C₄SO₃Hmim]HSO₄ at different temperatures (80-120 °C) to obtain the important kinetic parameters such as apparent activation energies of consecutive reaction steps.

Full text of DOI:10.1016/j.molcata.2010.10.006

Journal of Molecular Catalysis A: Chemical 334 (2011) 8–12
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Direct conversion and NMR observation of cellulose to glucose and
5-hydroxymethylfurfural (HMF) catalyzed by the acidic ionic liquids
Feng Jianga, Qingjun Zhua, Ding Maa,b,∗, Xiumei Liua, Xiuwen Hana
a State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Sciences, Dalian 116023, PR China
b Beijing National Laboratory for Molecular Science, State Key Laboratory for Structural Chemistry of Stable and Unstable Species, College of Chemistry and Molecular Engineering,
Peking University, Beijing 100871, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 May 2010
Received in revised form 30 August 2010
Accepted 4 October 2010
Available online 8 December 2010
The hydrolysis of cellulose was catalyzed over a variety of acidic ionic liquids (ILs). It is found that the
hydrolysis activity is directly associated with the acidity of catalysts, as evidenced by IR spectroscopy.
13C NMR characterization results confirm the majority product of cellulose hydrolysis is glucose, and
the resulting carbohydrates undergo further degradation, possibly also catalyzed by the acidic ILs, to 5-
hydroxymethylfurfural (HMF). Moreover, in situ 13C NMR measurements clearly exhibit that the evolution
of products is dependent on the reaction process. We attempt to study the kinetics of cellulose hydrolysis
over the most active catalyst of [C4SO3Hmim]HSO4 at different temperatures (80–120 ◦C) to obtain the
important kinetic parameters such as apparent activation energies of consecutive reaction steps.
© 2010 Elsevier B.V. All rights reserved.
Keywords:
Hydrolysis of cellulose
Ionic liquids
In situ 13C NMR
Acidic catalysis
Apparent activation energy
1. Introduction
future to circumvent the cost and scalability hurdles as encoun-
tered in the traditional gasification and pyrolysis processes for the
Cellulose constitutes the major components of the so-called
biomass that has attracted worldwide attention as a potentially
widespread approach for sustainable energy production [1], but
there are very few practical processes for converting cellulosic
biomass directly into more value-added products. The reason for
this difficulty is related with their strong hydrogen-bond networks
in the cellulose crystalline forms, leading to the low efficiency
and environmental unfriendliness at the hydrolysis process in
which the bulk cellulose are transformed to their monomers or
fermentable sugars. Therefore, a novel, green and cost-effective
process under mild reaction conditions is urgently required for
cellulose utilization.
The application of ionic liquids (ILs) as solvents for the disso-
lution of cellulose was firstly reported by Rogers and co-workers
[2]. The unique reaction performances were attributed to the inter-
action of the chloride in the studied IL of [C4mim]Cl with the
hydrogen-bonding networks of cellulose, resulting in the decrys-
tallization. The characteristic properties of the ionic liquids such
as low vapor pressure and good thermal stability offer promising
biomass transformation. Moreover, the components of cation and
anion in the ILs can be fine-tuned to adjust the physicochemical
properties, such as hydrophobicity/hydrophilicity, of the resulting
ILs, thus affecting their performance in the interaction with the cel-
lulose [3]. Furthermore, hydrolysis of cellulose to the sugars such
as monosaccharides and disaccharides was observed when liquid
acids or solid acids were added to the reaction system with ILs
as solvents [4–6]. The separation of the products (from ILs) can
be achieved by extraction, e.g., supercritical CO2, or the so-called
nanofiltration by the membrane [7]. It is interesting to note that the
product of resulting monosaccharides could be subsequently con-
verted to other valuable products, e.g., 5-hydroxymethylfurfural
(HMF), which could be used for the future liquid fuel [8]. The
acidities of the employed acids undoubtedly played crucial roles
for those transformations, but the detailed studies of the reaction
mechanisms have been scarce in the literature.
The acidic ionic liquids can be prepared by modifying the ILs
with acidic groups, and such materials were used as catalysts for
the acid-catalyzed reactions [9–11]. In this report, we investigate
the hydrolysis of cellulose over several acidic ILs in the solvent of
1-butyl-3-methylimidazolium chloride ([Bmim]Cl). Infrared spec-
troscopy is used to study the acidity of the catalysts with pyridine as
the probe molecule to correlate the acidity and hydrolysis activity.
Moreover, in situ 13C NMR spectroscopy is applied to identify the
reaction products and to monitor the hydrolysis process to under-
∗
Corresponding author at: Beijing National Laboratory for Molecular Science,
State Key Laboratory for Structural Chemistry of Stable and Unstable Species, College
of Chemistry and Molecular Engineering, Peking University, Beijing 100871, China.
Tel.: +86 10 62758603; fax: +86 10 62758603.
E-mail address: dma@pku.edu.cn (D. Ma).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.10.006

F. Jiang et al. / Journal of Molecular Catalysis A: Chemical 334 (2011) 8–12
9
Fig. 1. Ionic liquids used in the experiments. (A) [Bmim]Cl, (B) [C1COOHmim]Cl, (C)
[Bmim]HSO4, (D) [C4SO3Hmim]Cl, and (E) [C4SO3Hmim]HSO4.
stand the underlying reaction mechanism. Based on the knowledge
of the characterization results, we propose a simplified reaction
model, and perform kinetic studies to obtain important reaction
parameters. To our knowledge, it is the first time to report in situ
study of 13C NMR to monitor the hydrolysis process of cellulose
with the acidic ionic liquids as catalysts.
Fig. 2. Yield of TRS/glucose/HMF at 100 ◦C (1 h) with different acid ionic liquids as
catalysts in the solvent of [Bmim]Cl. (a) Blank, (b) [C1COOHmim]Cl, (c) [Bmim]HSO4,
(d) [C4SO3Hmim]Cl, and (e) [C4SO3Hmim]HSO4.
2.3. Characterization
2.3.1. IR
2. Experimental
The interaction of pyridine molecule with the acidic sites of
studied ionic liquids was studied by infrared spectroscopy at room
temperature. The probe molecule of pyridine was mixed with the
ionic liquids with the ratio of 1:5, and the mixture was spreaded
into the liquid film between KBr windows. All the IR spectra were
recorded on a Bruker Vector 22 spectrometer with an MCT detector
at a resolution of 1 cm−1 and 32 scans.
2.1. Materials
The structures of the ionic liquids used for this study
are displayed in Fig. 1. 1-Butyl-3-methylimidazolium chlo-
ride ([Bmim]Cl), 1-carboxyethyl-3-methylimidazolium chloride
([C1COOHmim]Cl), 1-butyl-3-methylimidazolium hydrogensulfate
([Bmim]HSO4), 1-butyl sulfonic acid-3-methylimidazolium chlo-
ride ([C4SO3Hmim]Cl), 1-butyl sulfonic acid-3-methylimidazolium
hydrogensulfate ([C4SO3Hmim] HSO4) were provided from Meis-
ibei company, China, and purities of all ILs were ca. 98%. We
speculate that [C1COOHmim]Cl and [Bmim]HSO4 belong to the
weakly acidic ILs since it− is well known that the function
group of –COOH and HSO4 are relatively weak acids, whereas
[C4SO3Hmim]Cl and [C4SO3Hmim]HSO4 are suggested to be rel-
atively strong ILs because –SO3H group are widely used to modify
the acidity of the polymers, leading to the strong acidic catalysts.
To remove the water content in the ILs, the ILs have been kept in
the vacuum drying oven at 50 ◦C, and reactant of microcrystalline
cellulose (AR, ACROS) is dried at 100 ◦C oven overnight before use.
2.3.2. NMR
Liquid-state 13C nuclear magnetic resonance (NMR) spectra
were obtained on a Varian DRX-400 spectrometer. The resonance
frequency employed in 13C NMR was 100.6 MHz. The sample was
added with 10% DMSO-d6 in the sample tube, and the chemical
shifts were referenced to tetramethylsilane (TMS). In the standard
sample measurements, 4% of glucose or cellulose was dissolved in
the [Bmim]Cl. The samples were heated at 90 ◦C to reduce the vis-
cosities of the mixture. In contrast, the in situ NMR experiment was
carried out in a home-made glass NMR sample tube, with the tem-
perature accurately controlled. The sample tube was loaded into
NMR machine and heated quickly to 80 ◦C and reacted at that tem-
perature for 24 h. In situ 13C NMR spectra were recorded during the
reaction with the accumulation of 800 scans.
2.2. Hydrolysis reaction
The cellulose hydrolysis was carried out in a round bottom
flask that was heated in the oil-bath in the range from 80 ◦C to
120 ◦C. Typically, 0.4 g microcrystalline cellulose was added into
8.0 g [Bmim]Cl solvent that was preheated under vigorous stirring,
followed by the addition of desired 1 ml acidic ionic liquid cata-
lyst, 0.08 g H2O and 2 ml co-solvent of dimethylformamide (DMF)
at the reaction temperature. Blank experiment was also performed
without the addition of the acidic IL catalyst for comparison. The
reaction was then vigorously stirred until a transparent solution
was obtained. At different time intervals, 0.11 g samples (0.1 ml)
were extracted, and quenched immediately with 0.9 ml NaOH solu-
tion (0.03 mol/l). The solution was centrifuged at 15,000 rpm for
10 min and further employed for the product analysis. The anal-
ysis of total reducing sugar (TRS) was based on the adsorption
spectroscopy, and the details are elaborated in the literature [12].
In a typical measurement, 0.5 ml resultant solution was mixed
with 0.5 ml predetermined 3,5-dinitrosalicylic acid (DNS) reagent,
heated for 10 min at 100 ◦C and cooled to room temperature. 4 ml
deionized water was then added to the sample, and the mixture
was subjected for the quantitative analysis on a JASCO V-550 spec-
trophotometer at 520 nm with a slit width of 0.1 mm.
3. Results and discussion
The cellulose hydrolysis reaction results (100 ◦C, 1 h) over the
various acidic IL catalysts are exhibited in Fig. 2. For the blank
experiment without the presence of acidic IL, i.e., the reaction in
the solvent of [Bmim]Cl, the yield of TRS is negligible. Indeed, lit-
tle activity is observed even when the reaction is prolonged to
100 h, confirming that [Bmim]Cl is not active for the production
of reducing sugars. In contrast, the presence of weakly acidic ILs of
[C1COOHmim]Cl and [Bmim]HSO4 catalyze the hydrolysis of cel-
lulose, and TRS of 3% and 5% are achieved after 1 h reaction. We
note further reaction time at 24 h enhance the release of reduc-
ing sugars to 17% and 28% over [C1COOHmim]Cl and [Bmim]HSO4,
respectively. It is remarkable that the presence of the sulfonic
acid group in the ILs, i.e., [C4SO3Hmim]Cl and [C4SO3Hmim]HSO4,
greatly increase the reaction rates of the cellulose hydrolysis, and
the hydrolysis are nearly complete in 1 h. TRS of 95% and 85% are
observed on [C4SO3Hmim]Cl and [C4SO3Hmim]HSO4, respectively.
The major products of the cellulose transformation, as revealed by
chromato analysis, are glucose and HMF, further careful inspec-
tion discloses a variety of organic substances (containing carbon,

10
F. Jiang et al. / Journal of Molecular Catalysis A: Chemical 334 (2011) 8–12
Fig. 3. FT-IR spectra of ionic liquids with pyridine as probe molecule. From bottom
to top: [Bmim]Cl, [Bmim]HSO4, [C4SO3Hmim]HSO4, and [C4SO3Hmim]Cl.
Fig. 4. 13C NMR of (a) cellulose and (b) glucose in the solvent [Bmim]Cl.
hydrogen, and/or oxygen) are present as by-products. However,
the amounts of the by-products are minor and the nature of those
molecules is very difficult to identify. We suggest that the hydrol-
ysis discrepancy between the same volume of [C4SO3Hmim]Cl
and [C4SO3Hmim]HSO4 does not indicate the acid strength of
[C4SO3Hmim]Cl is stronger than [C4SO3Hmim]HSO4. On the con-
trary, it might implicate that [C4SO3Hmim]HSO4 has less amount
of the active acid sites.
90 and 100 ppm indicates that cellulose is not transformed into the
glucose in the presence of only solvent, supporting the conclusion
that the hydrolysis of cellulose to reducing sugars is nearly negli-
gible without the acidic catalyst. Therefore, cellulose and glucose
undoubtedly exhibit characteristic bands in the 13C NMR spec-
troscopy and they do not interfere. We take this as a strong proof
to use 13C NMR spectroscopy to monitor the cellulose hydrolysis
process. For the sake of simplicity, we take characteristic bands at
101.8 ppm for the presence of cellulose, 96.6 and 91.9 ppm for the
presence of glucose, respectively.
It is reported by Yang and Kou that the application of pyridine on
the ionic liquids provides an easy approach for studying the acidi-
ties of ionic liquids, and the appearance of a band near 1540 cm−1
is the characteristic band for the formation of pyridinium ions,
resulted from the presence of Brønsted acidity [13]. Fig. 3 shows
the IR spectra of the employed ionic liquids with pyridine as
the probe molecule. Since [C1COOHmim] is a solid under room
temperature, it is not included in the figure for the comparison. No
signal at 1540 cm−1 can be detected over the mixture of [Bmim]Cl
and pyridine, implicating that the Brønsted acidity is nearly negli-
gible on the solvent of [Bmim]Cl. Nevertheless, a small but distinct
band is clearly observed after [Bmim]HSO4 is added with pyridine,
confirming the presence of Brønsted acidity on [Bmim]HSO4. More-
over, such a band is more pronounced over [C4SO3Hmim]HSO4
and [C4SO3Hmim]Cl. Despite the difficulty in the quantitative
analysis under the present experimental conditions, we suggest
that the increase of the intensity of the band at 1540 cm−1, i.e.,
[C4SO3Hmim]Cl > [C4SO3Hmim]HSO4 > [Bmim]HSO4 > [Bmim]Cl,
might be related with their Brønsted acid amounts, and such
increase is consistent with their increasing reaction yields of TRS
in Fig. 2.
The in situ 13C NMR investigation of the cellulose hydrolysis
upon the acidic catalyst of [C4SO3Hmim]Cl at 80 ◦C is presented
in Fig. 5. The 13C NMR spectra are plotted against the reaction
time to show the evolution of the reaction products as the reac-
tion proceeds. For the first 0.5 h, only the resonance attributed
to cellulose is observed, implicating that the cellulose hydroly-
sis is insignificant or the product formation is not high enough
for the 13C NMR detection. However, after the reaction proceeds
for 1 h, two slight but distinct resonances are clearly observed
13C NMR spectroscopy is employed as a characterization tool
to identify the products of the cellulose products and to study
the reaction process, this is based on the observation that the sol-
vent of [Bmim]Cl does not display the 13C resonance signal in the
studied region between 55 and 120 ppm [14]. Since glucose is the
building unit for the cellulose, [15] their 13C NMR spectra in the
solvent of [Bmim]Cl are displayed in Fig. 4. The glucose possess
two diastereomers of ␣-glucopyranose and ␤-glucopyranose in the
solution, this is reflected by the characteristic bands at 91.9 ppm
and 96.6 ppm, which are assigned to the C1 atom in ␣ and ␤ isomers,
respectively [16]. The peaks between 60 and 80 ppm are attributed
to various carbon atoms (C2–C6) in the glucose structure. In con-
trast to the spectrum of glucose, the chemical shift of C1 and C4
atoms in the cellulose shift to 101.8 and 78.8 ppm, respectively,
because of the formation of ␤-1,4 glycosidic bonds among glucose
to form the cellulose, which are similar to that reported in the lit-
erature [14]. Moreover, the absence of resonance signal between
Fig. 5. In situ 13C NMR study of the cellulose hydrolysis over [C4SO3Hmim]Cl. The
reaction was conducted at 80 ◦C under atmosphere pressure in a home-made sample
tube in NMR machine.

F. Jiang et al. / Journal of Molecular Catalysis A: Chemical 334 (2011) 8–12
11
at 96.6 and 91.9 ppm, which are related to the presence of glu-
cose. In other words, the cellulose undergoes hydrolysis to form
glucose. Although it is reported that various products are plausi-
ble such as cellobiose, cellotriose, in the cellulose hydrolysis [17],
we tentatively draw the conclusion that the majority of the cellu-
lose hydrolysis products in our system is glucose because no other
product can be identified by 13C NMR. This suggestion is further
underpinned by the observation of the decrease of cellulose signal
(101.8 ppm) and the simultaneous increase of the glucose signal
(96.6 and 91.9 ppm) at the reaction time of 1.5 h. Moreover, a slight
resonance at 109.3 ppm at 1.5 h is distinguishable. Longer reac-
tion time (2–2.5 h) leads to further increase of the characteristic
peaks assigned to glucose and the decrease of the resonance related
with cellulose. When the reaction is prolonged to 3 h, the cellulose
nearly disappears, indicating that it is completely transformed. In
contrast, four characteristic bands at 177.7, 162.4, 151.3 and 109.3
are clearly observed, and those characteristic bands are assigned to
the formation of 5-hydroxymethylfurfural (HMF) [18]. Clearly, the
degradation of glucose to HMF takes place on the acidic environ-
ment employed in this research. As the reaction further proceeds
from 3.5 h to 5 h, we observe pronounced increase of the charac-
teristic resonance signal associated with HMF, accompanied by the
decrease of the peak intensities related with glucose. The 13C NMR
measurements suggest that the cellulose hydrolysis can be sim-
plified as two consecutive steps: (1) cellulose is hydrolyzed in the
presence of acidic catalyst to the glucose, a kind of reaction inter-
mediate; (2) the produced glucose is further converted into HMF,
also catalyzed by the catalyst of [C4SO3Hmim]Cl. The detailed iso-
merization mechanisms from glucose to HMF in the presence of IL
have been discussed by Zhao et al. [19].
Fig. 6. Cellulose hydrolysis over [C4SO3Hmim]Cl at different temperatures: (ꢀ)
80 ◦C, (᭹) 100 ◦C, and (ꢁ) 120 ◦C.
Table 1
Reaction rate constants at different temperatures.
Temperature (◦C)
80
100
120
K1 (min−1
K2 (min−1
)
)
7.9 × 10−3
1.6 × 10−3
6.6 × 10−2
8.5 × 10−3
4.1 × 10−1
4.1 × 10−2
A series of cellulose hydrolysis upon the acidic catalyst of
[C4SO3Hmim]Cl have been carried out at 80 ◦C, 100 ◦C and 120 ◦C,
respectively. The relations between the TRS (glucose) yield and the
reaction time are displayed in Fig. 6. They seem to follow the same
pattern, i.e., the increase of the glucose concentration at the begin-
ning is owing to the hydrolysis of cellulose to glucose, and further
decrease of the TRS yield is due to the degradation of glucose to the
HMF. The results are in accordance with the conclusions from the
13C NMR observations in Fig. 5. Clearly, short reaction time favors
the production of glucose and longer reaction time facilitates for-
mation of HMF in a given temperature. Moreover, the hydrolysis
is remarkably dependent on the reaction temperature. Hydrolysis
of cellulose seems to be faster at higher temperature as well as the
degradation of the glucose to HMF. Thus, the cellulose hydrolysis in
the acidic ILs is both dependent on the reaction time and the reac-
tion temperature. Undoubtedly, the kinetic parameters are crucial
to optimize the reaction process to obtain desired reaction prod-
ucts. The simulated reaction rate constants are listed in Table 1.
The hydrolysis of cellulose upon the acidic catalysts is an
extremely complex process in which a variety of reactions can take
place together and many products are formed. Based on our obser-
vation that the products detected are only glucose and HMF, we
tentatively simplify the cellulose reaction process in our system as
K
K
2
1
Cellulose−→ glucose−→ HMF
in which K1 and K2 are reaction rate constants for the hydrolysis and
further degradation steps, respectively. As proposed by a number
of reports, [5,20] such a simplified model can provide a reasonable
estimation for the study of kinetic parameters. Li et al. concluded
that the kinetics above follow a consecutive first order sequence [5].
Assume the initial concentration of cellulose is ␣ at the beginning
of the reaction, the concentrations of cellulose, glucose and HMF in
a given reaction time t can be written as
dx
= −K1x
(1)
(2)
(3)
dt
dy
dt
= K1x − K2y
dz
dt
= K2y
where x, y and z are the concentration of cellulose, glucose and HMF,
respectively. After further integration, we can obtain
x = ae−Kt
(4)
More importantly, the glucose concentration y can be expressed as
K1a
t
t
2
(e−K − e−K
)
(5)
1
y =
K2 − K1
Therefore, for a given reaction at a certain temperature, the reaction
rate constants K1 and K2 can be simulated by plotting the glucose
concentration (determined by experiments) and the reaction time
[5].
Fig. 7. Arrhenius plot of different reaction steps in the cellulose hydrolysis over
[C4SO3Hmim]Cl. (ꢂ) cellulose hydrolysis to glucose, (ꢀ) glucose degradation to HMF.

12
F. Jiang et al. / Journal of Molecular Catalysis A: Chemical 334 (2011) 8–12
The reaction rates of both reaction steps increase at elevated tem-
perature, implying relatively high temperature might be necessary
for the practical application. In addition, Arrhenius plots for the
cellulose hydrolysis and glucose degradation are shown in Fig. 7.
We estimate that the apparent activation energies for the cellulose
hydrolysis to glucose and glucose degradation to HMF are 114 and
95 kJ mol−1, respectively.
References
[1] M. Stöcker, Angew. Chem. Int. Ed. 47 (2008) 9200–9211.
[2] R.P. Swatloski, S.K. Spear, J.D. Holbrey, R.D. Rogers, J. Am. Chem. Soc. 124 (2002)
4974–4975.
[3] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A. Broker, R.D.
Rogers, Green Chem. 3 (2001) 156–164.
[4] C. Li, Z.K. Zhao, Adv. Synth. Catal. 349 (2007) 1847–1850.
[5] C. Li, Q. Wang, Z.K. Zhao, Green Chem. 10 (2008) 177–182.
[6] R. Rinaldi, R. Palkovits, F. Schüth, Angew. Chem. Int. Ed. 47 (2008) 8047–
8050.
[7] J. Kröckel, U. Kragl, Chem. Eng. Technol. 26 (2003) 1166–1168.
[8] Y. Román-Leshkov, C.J. Barrett, Z.Y. Liu, J.A. Dumesic, Nature 447 (2007)
982–985.
[9] A.C. Cole, J.L. Jensen, I. Ntai, K.L.T. Tran, K.J. Weaver, D.C. Forbes, J.H. Davis, J.
Am. Chem. Soc. 124 (2002) 5962–5963.
[10] D.C. Forbes, K.J. Weaver, J. Mol. Catal. A 214 (2004) 129–132.
[11] A.S. Amarasekara, O.S. Owereh, Ind. Eng. Chem. Res. 48 (2009) 10152–
10155.
4. Conclusion
The acidic ionic liquids readily catalyze the cellulose hydrolysis
in the solvent of [Bmim]Cl in the range of 80–120 ◦C, resulting in
the dominant productions of glucose and 5-hydroxymethylfurfural
(HMF). The Brønsted acidity, particularly the acid strength, in the
catalysts are crucial to break the ␤-1,4 glycosidic bonds in the cel-
lulose to form glucose, and further convert glucose to HMF. The
hydrolysis process and the product formation are both dependent
on the reaction time and the reaction temperature. In situ 13C NMR
spectroscopy is proven to be an efficient characterization method
to study the cellulose hydrolysis in the ionic liquids.
[12] G.L. Miller, Anal. Chem. 31 (1959) 426–428.
[13] Y.-l. Yang, Y. Kou, Chem. Commun. (2004) 226–227.
[14] J.S. Moulthrop, R.P. Swatloski, G. Moyna, R.D. Rogers, Green Chem. 7 (2005)
1557–1559.
[15] B.E. Binder, R.T. Raines, J. Am. Chem. Soc. 131 (2009) 1979–1985.
[16] R.L. Dudley, C.A. Fyfe, P.J. Stephenson, Y. Deslandes, G.K. Hamer, R.H. Marches-
sault, J. Am. Chem. Soc. 105 (1983) 2469–2472.
[17] M. Sasaki, B. Kabyemela, R. Malaluan, S. Hirose, N. Takeda, T. Adschiri, K. Arai,
J. Supercrit. Fluids 13 (1998) 261–268.
Acknowledgements
[18] http://riodb01.ibase.aist.go.jp/sdbs/cgi-bin/direct frame top.cgi
April 2010).
(accessed
This work received financial support from the Ministry of Sci-
ence and Technology of China (2010CB732202), and the Natural
Science Foundation of China. D.M. thanks the CAS for support
through the Bairen project.
[19] H. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Science 316 (2007) 1597–
1600.
[20] L. Vanoye, M. Fanselow, J.D. Holbrey, M.P. Atkins, K.R. Seddon, Green Chem. 11
(2009) 390–396.