Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst under mild conditions

Article abstract of DOI:10.1016/j.cattod.2013.09.061

A catalyst consisting of dispersed Ru on an ionic liquid (BmimPF ₆)-heteropolyacid (H₃PW₁₂O₄₀· nH₂O) hybrid as a support, i.e. Ru/[Bmim]₃PW ₁₂O₄₀, has been successfully synthesized. The catalyst, which combines the Ru sites for hydrogenation and both Lewis and Br?nsted acidic sites for hydrolysis, exhibits a superior catalytic performance for selective conversion of the microcrystalline cellulose to hexitols over the catalyst of mixing [Bmim]₃PW₁₂O₄₀ and Ru/C. On the Ru/[Bmim]₃PW₁₂O₄₀ catalyst, a sorbitol selectivity of 70.3% with a microcrystalline cellulose conversion of 63.7% was achieved in 24 h at 433 K and 5 MPa H₂. The superior catalytic performance of Ru/[Bmim]₃PW₁₂O₄₀ has been characterized using the hydrogenolysis of cellobiose as a probe reaction and was attributed to the Br?nsted acid sites generated from hydrogen spillover from the Ru sites to the O sites of the support. In situ generation of the Br?nsted acidic sites through hydrogen spillover has been confirmed by FT-IR characterization of pyridine adsorption. Furthermore, pH changes after treating the catalyst in H₂ demonstrated that dissolution of the protons generated on the oxygen sites as a result of hydrogen spillover acidifies the liquid product. These Br?nsted acids work synergistically with the supported Ru and contribute to the enhanced hydrogenolysis activity.

Full text of DOI:10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
Catalysis Today xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Selective conversion of microcrystalline cellulose into hexitols over a
Ru/[Bmim]3PW12O40 catalyst under mild conditions
Xianna Xiea, Jinyu Hana, Hua Wanga,∗, Xinli Zhua, Xiao Liua, Yufei Niua, Zhiqiang Songa,
Qingfeng Gea,b,∗∗
a Key Laboratory for Green Chemical Technology of State Education Ministry, School of Chemical Engineering & Technology, Tianjin University, Tianjin
300072, PR China
b Department of Chemistry and Biochemistry, Southern Illinois University, Carbondale, IL 62901, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A catalyst consisting of dispersed Ru on an ionic liquid (BmimPF6)–heteropolyacid (H3PW12O40·nH2O)
hybrid as a support, i.e. Ru/[Bmim]3PW12O40, has been successfully synthesized. The catalyst, which com-
bines the Ru sites for hydrogenation and both Lewis and Brønsted acidic sites for hydrolysis, exhibits a
superior catalytic performance for selective conversion of the microcrystalline cellulose to hexitols over
the catalyst of mixing [Bmim]3PW12O40 and Ru/C. On the Ru/[Bmim]3PW12O40 catalyst, a sorbitol selec-
tivity of 70.3% with a microcrystalline cellulose conversion of 63.7% was achieved in 24 h at 433 K and
5 MPa H2. The superior catalytic performance of Ru/[Bmim]3PW12O40 has been characterized using the
hydrogenolysis of cellobiose as a probe reaction and was attributed to the Brønsted acid sites gener-
ated from hydrogen spillover from the Ru sites to the O sites of the support. In situ generation of the
Brønsted acidic sites through hydrogen spillover has been confirmed by FT-IR characterization of pyri-
dine adsorption. Furthermore, pH changes after treating the catalyst in H2 demonstrated that dissolution
of the protons generated on the oxygen sites as a result of hydrogen spillover acidifies the liquid prod-
uct. These Brønsted acids work synergistically with the supported Ru and contribute to the enhanced
hydrogenolysis activity.
Received 1 August 2013
Received in revised form
23 September 2013
Accepted 26 September 2013
Available online xxx
Keywords:
Ru/[Bmim]3PW12O40
Cellobiose and cellulose
Hydrogenolysis
Hydrogen spillover
Brønsted acid
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
conversion of cellulose [13]. On the other hand, Geboers et al.
reported a 49% hexitols yield (sugar alcohol 32%, sorbitan 17%)
Selective conversion of renewable lignocellulosic biomasses,
such as cellulose, to low molecular alcohols and/or hydrocarbons is
crucial to a wide spread use of the most abundant, inexpensive and
renewable sources [1]. Producing commodity chemicals or fuels
from biomass-based resources will reduce the reliance on fossil
resources and help to improve a nation’s energy security [2–8].
However, the strong ß-1,4-glycosidic bond in cellulose as well as
intra- and inter-molecular hydrogen bonds result in a robust crys-
talline structure, which made the hydrogenolysis of cellulose under
mild conditions a great challenge [9].
Mixed catalysts that combine active sites for hydrolysis and
hydrogenation are attractive for cellulose conversion [10–12].
Palkovits and coworkers demonstrated a yield of ∼81% toward C4
to C6 sugar alcohols (C6 53.0%, C5 8.9%, C4 18.7%) at 433 K and
5 MPa H2 in 7 h over a combined H4SiW12O40–Ru/C catalyst in the
with a conversion of 82% over the H4SiW12O40–Ru/C catalyst at
463 K and 5 MPa H2 in 24 h [14]. The same group performed an
one-step conversion of ball-milled cellulose over a catalyst mixing
the cesium salts of HPAs calcined at 600 ◦C and Ru/C, yielding a 90%
hexitols (alditol 70%, sorbitan 20%) with a 100% conversion after
48 h [15]. In those catalysts, Ru, the active sites for hydrogenation,
is supported on C and not in direct contact with the acidic sites of the
ployacids, which are active for hydrolysis. Liu et al. reported briefly
a study on the conversion of ball-milled cellulose over the catalysts
prepared by loading Ru nanoparticles directly on the Cs salts of
H3PW12O40 [16]. The authors showed that a sorbitol yield of 43% at
433 K after 24 h reaction can be achieved over the Ru/Cs3PW12O40
catalyst and suggested that the Brønsted acid sites generated in situ
through hydrogen spillover [17–19] made up for the missing intrin-
sic Brønsted acidity and contributed to the observed enhancement
of catalytic performance.
Ionic liquids (ILs) are a class of non-volatile green solvents
and have been studied extensively for dissolution of cellulose
[20]. The IL–HPA hybrids are more stable, and less sensitive to
humidity and temperature than the parent HPAs [21] and have
been reported to exhibit excellent performance in electrochemistry
∗
Corresponding author.
Corresponding author at: Department of Chemistry and Biochemistry, Southern
∗∗
Illinois University, Carbondale, IL 62901, USA. Tel.: +1 618 4536406.
E-mail addresses: tjuwanghua@tju.edu.cn (H. Wang), qge@chem.siu.edu (Q. Ge).
0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.09.061
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
2
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
and catalysis [22–27] due to their superior proton conductivity. In
order to enhance the acidity of the IL–HPA hybrids, several authors
attempted to incorporate acidic functional groups into the organic
cations. For example, Dyson and Kou’s group used the catalytic sys-
tem composed of metal NPs and a functionalized Brønsted acidic IL
immobilized in a nonfunctionalized IL, allowing hydrogenation and
dehydration reactions to occur in tandem [28]. Huang et al. showed
that sulfonated IL–HPA salts ([MIMPSH]nH3−nPW12O40 (n = 1, 2, 3))
were able to dehydrate fructose into 5-hydroxymethyl furfural (5-
HMF) [29]. The same catalyst showed a glucose yield of 36% for
cellulose hydrolysis at 413 K in 5 h with water and methyl isobutyl
ketone (MIBK) as solvents [30]. However, [MIMPSH]nH3−nPW12O40
are soluble in water, which makes the recovery of the catalyst a
challenge. Therefore, an insoluble IL–HPA hybrid would an ideal
support for a Ru hydrogenation catalyst for the selective conversion
of cellulose.
In the present study, we successfully dispersed Ru on an insol-
uble hybrid material of IL(BmimPF6)–HPA(H3PW12O40·nH2O). The
catalyst has been demonstrated to have a high selectivity toward
sorbitol for the hydrogenation/hydrolysis of microcrystalline cel-
lulose in a one-pot process. Our characterization showed that the
Ru/IL–HPA catalyst generates Brønsted acidic sites in situ and acid-
ifies the liquid reactive mixture. On this catalyst, the active sites for
hydrogenation and hydrolysis coexist and work synergistically.
deionized water were added to the reactor. The reactor was then
flushed with hydrogen to 5 MPa and heated to the desired reac-
tion temperature (measured in the Teflon insert). After reaction,
the reactor was quickly cooled in an ice bath. The solid residual con-
taining the catalyst and unreacted cellulose were separated from
the liquid product using centrifugation. The liquid products were
quantified using HPLC equipped with a refractive index detector
and ICSep Coregel-87H column. An aqueous solution of H2SO4 at
0.005 M with a flow rate of 0.6 mL/min was used as the mobile
phase. A series of calibration standards with different concentra-
tions have been prepared and measured with the same column on
the HPLC. By comparing the retention time and peak area of the
sample with that of the calibration standards under the same con-
dition, a particular component of the liquid products corresponding
to each peak can be determined. The amount of the product can be
quantified with the peak area by referring to the standard curve. The
HPLC column used in the present study can identify and quantify
soluble saccharides (sucrose, fructose, xylose, etc.), C1–C6 alcohol
(methanol, propanediol, sorbitol, etc.), carboxylic acids (lactic acid,
levulinic acid, formic acid, etc.), carbonyls (formaldehyde, acetalde-
hyde) and some other small organic molecules. The solid residual
containing the catalyst and the unreacted cellulose were dried
overnight before being weighed.
The conversion of cellulose was calculated as the ratio of the
weight difference between the solids before and after reaction to
the initial weight of cellulose. The conversion of cellobiose was
determined from the HPLC standard curve. The yield and selec-
tivity of a specific product were calculated as: yield (%) = [moles
C in product/total moles C loaded in reactor] × 100%; selectivity
(%) = yield/conversion × 100% [31].
2. Experimental details
2.1. Materials
Microcrystalline cellulose from Sigma-Aldrich, cellobiose (98%)
from Acros, 12-phosphotungstic acid (AR) from Kermel (Tianjin,
China), activated carbon-supported Ru (5 wt% Ru) from Dalian
Tongyonger Chemical Co. Ltd., and 1-butyl-3-methylimidazolium
hexafluorophosphonate ([Bmim]PF6) from Henan Lihua Pharma-
ceutical Co., Ltd were used as received without further purification.
Standard chemicals including glucose, sorbitol, anhydrous sorbitol,
xylitol, erythrol and others were purchased from Sigma–Aldrich.
3. Results and discussion
3.1. Characterization of catalyst
The C, H and N contents were determined with a Vario MICRO
cube elemental analyzer and the W and P contents were measured
using an inductively coupled plasma emission (ICP; ICP-9000, USA
Thermo Jarrell-Ash Corp) spectrometry. The sample was treated at
200 ◦C in a N2 stream for 3 h prior to elemental analysis. The results
showed that the C, N, H, P and W contents in the catalyst are 8.90%,
2.53%, 1.52%, 0.96% and 65.82%, respectively. The theoretical val-
2.2. Catalyst preparation
2.2.1. Synthesis of [Bmim]3PW12O40
[Bmim]3PW12O40 was prepared by following the procedure
reported by Ammam and Fransaer [23]. In a typical preparation,
adding [Bmim]PF6 (6 mmol) directly to a H3PW12O40 solution pre-
pared by dissolving 2 mmol H3PW12O40 in deionized water (10 mL)
at room temperature and stirring the mixture for 24 h results
in precipitates. The precipitates were filtered, then washed with
deionized water and dried in vacuum, resulting in the solid product.
ues calculated based on the molecular formula, [Bmim]3PW12O40
,
are 8.74% for C, 2.55% for N, 1.37% for H, 0.94% for P and 66.98%
for W. Apart from H, the elemental analysis and ICP results showed
that the elemental ratio is within 2% of the [Bmim]3PW12O40 for-
mula value. Therefore, the as-synthesized compound is denoted
as [Bmim]3PW12O40. TG analysis showed that the synthesized
[Bmim]3PW12O40 is stable up to 673 K. Obvious decomposition
occurs at 723 K, as shown in Fig. S1 of the Electronic Supplementary
Information.
The product has been characterized as [Bmim]3PW12O40
.
2.2.2. Synthesis of Ru/[Bmim]3PW12O40
Ru/[Bmim]3PW12O40 was prepared with
a conventional
impregnation method: a mixture of solid [Bmim]3PW12O40 (2 g)
and RuCl3 (0.2632 g) in 10 mL ethanol solution was stirred for 24 h
at room temperature, followed by drying at 353 K overnight. The
resulting solid was reduced in H2 at 473 K for 3 h, corresponding to
5 wt% Ru on the support, before being used as a catalyst in subse-
The SEM and TEM analysis have been used to monitor the
morphology and dispersion of the support and catalysts and the
images (Figs. S2 and S3) are provided in the Electronic Supple-
mentary Information. SEM images show that the crystallines of
[Bmim]3PW12O40 are in the form of block-shaped particles. The
particle sizes range from several hundred nanometers to a few
microns. In contrast, TEM images show that Ru is dispersed uni-
formly on the support and its particles in the Ru/[Bmim]3PW12O40
sample are uniform spherical nanoparticles with a size of 4–5 nm.
quent reactions. The catalyst was denoted as Ru/[Bmim]3PW12O40
.
2.3. Reaction procedures
The one-pot conversion of cellobiose and microcrystalline cel-
lulose into hexitols was conducted in a lab-scale batch reactor
with a Teflon insert (20 mL) heated in an oil bath and stirred by
an electromagnetic stick. In a typical catalytic reaction run, 0.25 g
microcrystalline cellulose or cellobiose, 0.05 g catalyst and 5 mL
The
XRD
patterns
of
(a)
H3PW12O40–nH2O,
(b)
[Bmim]3PW12O40, and (c) Ru/[Bmim]3PW12O40 are shown in
Fig. 1. Long-range order was clearly shown in these XRD patterns.
In addition, the diffraction peaks of (b) and (c) is different from that
of (a) due to the substitution of the protons in H3PW12O40–nH2O by
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
3
Fig. 1. XRD patterns of (a) H3PW12O40
,
(b) [Bmim]3PW12O40
,
and (c)
Ru/[Bmim]3PW12O40
.
[Bmim]+ in [Bmim]3PW12O40. This is because that the crystalline
structure was formed from packing the anions of the heteropoly-
acids, with the [Bmim]+ filling the interstitials. Therefore, the
diffraction peaks in 5–10◦ should be results from the low-index
planes of the crystalline packing as the lattice parameters are
◦
˚
typically larger than 10 A. The peaks in the 2 theta range of 10–40
should be diffraction from higher index planes. As W is a good
X-ray scatter, long rang order originated from W in the crystal
are expected to contribute to the observed diffraction peaks in
this range. These XRD patterns show that the arrangement of the
heteropolyacid cages remains to be the dominant feature of the
hybrid with the ionic liquid as well as the Ru/[Bmim]3PW12O40
sample. The main diffraction peak of metallic Ru corresponding to
the (1 1 0) planes should appear at 2 theta of ∼44◦, corresponding
to the weak feature in Fig. 1(c). As the Ru particles are small,
∼4–5 nm according to TEM micrograph, the diffraction peaks are
expected to be weak and broad. Furthermore, there also peaks
originating from [Bmim]3PW12O40 and H3PW12O40 around this
angle, making the unambiguous assignment of the diffraction peak
difficult.
Fourier transform infrared (FTIR) analysis was performed under
vacuum on the Nicolet6700 spectrometer (KBr method). The FTIR
spectra of the samples are collected in Fig. 2. As shown in Fig. 2, bulk
H3PW12O40 shows characteristic bands associated with the Keggin
unit, i.e. the P-O band at 1079 cm−1, the W–O band at 981 cm−1, the
W–O–W bands at 893 cm−1 and 798 cm−1, and a weak W–O–P band
at 520 cm−1 [32]. The characteristic bands of the Keggin unit can
also be observed in the Bmim]3PW12O40, RuCl3/[Bmim]3PW12O40
Fig. 2. FT-IR spectra of (a) H3PW12O40, (b) [Bmim]3PF6, (c) [Bmim]3PW12O40, and
(d) Ru/[Bmim]3PW12O40
.
results showed that there was a small conversion of 6.6% after
10 h, attributed to the hydrolysis of cellulose at elevated tem-
perature. In the presence of the [Bmim]3PW12O40 + Ru/C catalyst,
the conversion of cellulose is increased to 35.5%, with oligosac-
charide being the main product. There is no glucose formation
with the [Bmim]3PW12O40 + Ru/C catalyst. In contrast, when the
Ru/[Bmim]3PW12O40 catalyst was used, the conversion of cellulose
was further improved to 42.6%. A more pronounced difference from
that on the [Bmim]3PW12O40 + Ru/C catalyst is product distribu-
tion. The Ru/[Bmim]3PW12O40 catalyst exhibited a high selectivity
toward sorbitol (58.0%). The selectivity of oligosaccharide is only
5.8%.
3.2.2. Optimization of reaction conditions
After demonstrating that the Ru/[Bmim]3PW12O40 catalyst is
highly selective toward sorbitol in the conversion of microcrys-
talline cellulose, we further optimized the reaction temperature
and time to maximize the sorbitol yield, by changing the reaction
temperature and controlling reaction time.
Fig. 3 shows cellulose conversion, sorbitol, glucose, and xyli-
tol selectivity on the Ru/[Bmim]3PW12O40 catalyst at temperatures
ranging from 393 K to 453 K and a reaction time of 10 h. Clearly,
the conversion of cellulose increases as the reaction temperature
and Ru/[Bmim]3PW12O40 samples, indicating the cage structure
3−
of PW12O40
unit remains intact. We noted that new absorp-
tion peaks at 3147 cm−1, 2958 cm−1, 1568 cm−1 and 1165 cm−1
appeared for the [Bmim]3PW12O40, RuCl3/[Bmim]3PW12O40 and
Ru/[Bmim]3PW12O40 samples, which can be assigned to the imi-
dazole ring. The absorption peaks at 3147 cm−1, and 2958 cm−1
were attributed to the C H stretching mode of the imidazole ring
and substituent group whereas the absorption peaks at 1568 cm−1
Table 1
Conversion of microcrystalline cellulose and selectivity of different products.
and 1165 cm−1 corresponded to the
stretching mode [33].
C
N
and imidazole ring
Catalyst
Conv. [%]
Selectivity [%]
Oli
Glu
Sor
–
4.5 9.6
58 6.8
Xyl
–
Gly
Meth
Nonea
[Bmim]3PW12O40 + Ru/C
Ru/[Bmim]3PW12O40
6.6
35.5
42.6
90.0
52.6
5.8
0.6
–
7.9
–
1.0
0.3
–
13.2
2.9
3.2. Catalytic activity
3.2.1. Catalytic conversion of microcrystalline cellulose
Reaction conditions: cellulose, 0.25 g; 5 wt% Ru/[Bmim]3PW12O40
,
0.05 g;
The Ru/[Bmim]3PW12O40 catalyst was tested in the conver-
sion of microcrystalline cellulose and compared with the mixed
[Bmim]3PW12O40 + Ru/C catalyst, as shown in Table 1. As a refer-
ence, we also run the reaction without adding any catalyst. Our
[Bmim]3PW12O40
, 0.05 g; 5 wt% Ru/C, 0.05 g; water 5 mL; 10 h; 433 K; H2,
5 MPa.
a
Ambient pressure; Oli = oligosaccharide; Glu = glucose; Sor = sorbitol;
Xyl = xylitol; Gly = glycol; Meth = methanal.
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
4
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
Fig. 4. Variation of cellulose conversion and selectivity for different prod-
ucts with reaction time at 433 K. Reaction conditions: cellulose, 0.25 g; 5 wt%
Ru/[Bmim]3PW12O40, 0.05 g; water, 5 mL; H2, 5 MPa.
Fig. 3. Conversion of cellulose and selectivity for different products at different tem-
peratures. Reaction conditions: cellulose, 0.25 g; 5 wt% Ru/[Bmim]3PW12O40, 0.05 g;
water, 5 mL; H2, 5 MPa; reaction time, 10 h.
is increased. At 453 K, the conversion reaches 88.6%. On the other
hand, the selectivity of sorbitol increases as the temperature is
increased, reaching a maximum at ∼433 K, and then decreases.
Higher reaction temperatures degraded the product into anhydrous
sorbitol (18.6%), erythrol (3.4%) and glycol (2.0%). Therefore, tem-
perature has to be controlled at ∼433 K to maximize the selectivity
toward sorbitol although higher temperatures would result in a
higher conversion of cellulose.
Reaction time is another parameter that can be controlled to
optimize conversion and selectivity for a batch reaction. Fig. 4
shows the conversion of microcrystal cellulose and the selectivity
toward the major products at 433 K with different reaction times.
As shown in Fig. 4, the conversion of cellulose increases almost
linearly with the reaction time. On the other hand, the selectivity
toward sorbitol increases from 20.8% in 2 h to 70.3% in 24 h. Further
increasing the reaction time to 36 h resulted in a sorbitol selec-
tivity of 38.1%, significantly decreased from the maximum value.
Consequently, degradation products including anhydrous sorbitol
(16.1%), 5-HMF (12%), glucose (1.7%), erythrol (2.3%), glycol (1.3%)
1,2-propylene glycol (0.9%), levulinic acid (0.5%) and formaldehyde
(8.0%) as well as a small amount of ethanol start to form.
Ru/[Bmim]3PW12O40
, respectively. As shown in Fig. 5, both
catalysts convert cellobiose effectively, with conversion reach-
ing almost 100% over [Bmim]3PW12O40 + Ru/C and 76.8% over
Ru/[Bmim]3PW12O40 in 30 min. A rough estimate, by assuming a
first-order reaction, shows that [Bmim]3PW12O40 + Ru/C converts
cellobiose almost 3 times faster than Ru/[Bmim]3PW12O40. How-
ever, the selectivity of the reaction with the two catalysts is in stark
contrast: the main product over [Bmim]3PW12O40 + Ru/C is gluci-
tol (3-␤-d-glucopyranosyl-d-glucitol, sugar alcohol of cellobiose),
reaching a maximum value of 88.8% at 30 min and maintaining
essentially constant over the reaction time whereas the selectiv-
ity toward sorbitol is low, reaching only 19.0% with a 3 h reaction
time. On the other hand, both glucitol and glucose were formed
with Ru/[Bmim]3PW12O40 as the catalyst. The glucitol and glu-
cose intermediates went through a maximum of ∼28.0% at ∼30 min
and ∼60 min, respectively, and were consumed almost completely
with a 3 h reaction time. Accompanying the disappearance of the
intermediates, the selectivity toward sorbitol increases and reaches
almost 100% in 3 h. As shown in Fig. 5(b), the sorbitol yield reaches
98.0% with a 3 h reaction time. Clearly, the Ru/[Bmim]3PW12O40
catalyst has a superior selectivity toward sorbitol.
On the basis of the experimental observation, we propose a reac-
tion mechanism on the Ru/[Bmim]3PW12O40 catalyst, as shown in
Fig. 6.
3.3. Mechanism study
We chose cellobiose as a probe molecule to study the mecha-
nism of the reactivity and selectivity of the Ru/[Bmim]3PW12O40
catalyst. Fig. 5(a) and (b) compare the results of hydrogenol-
ysis of cellobiose catalyzed by [Bmim]3PW12O40 + Ru/C and
According to this mechanism, cellobiose is first hydrogenated
to glucitol by the dissociatively adsorbed hydrogen atoms gener-
ated on the metallic Ru sites. Glycosidic bonds of glucitol are then
cleaved over the acid sites to produce glucose and sorbitol. The
Fig. 5. Catalytic conversion of cellobiose under catalysts (a) [Bmim]3PW12O40 + Ru/C, and (b) Ru/[Bmim]3PW12O40. Reaction conditions: cellobiose, 0.25 g; 5 wt%
Ru/[Bmim]3PW12O40, 0.05 g; [Bmim]3PW12O40, 0.05 g; 5 wt% Ru/C, 0.05 g; water, 5 mL; 433 K; H2, 5 MPa.
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
5
Fig. 6. Proposed mechanism of catalytic conversion of cellobiose into sorbitol.
glucose produced will be rapidly hydrogenated to sorbitol by the
available hydrogen atoms on the metallic Ru sites. The proximity of
the hydrogenation sites (metallic Ru sites) and the acid sites (Wn+
)
Fig. 7. FT-IR spectra of the liquid products of cellobiose transformations in 3 h at
on the Ru/[Bmim]3PW12O40 catalyst is key to the observed superior
selectivity. As shown in Fig. 5, the [Bmim]3PW12O40 + Ru/C cata-
lyst shows an even better hydrogenation activity, manifested by
the high initial conversion rate of cellobiose. However, cellobiose
is only hydrogenated to glucitol. The hydrolysis of glucitol was
hindered by the lack of the immediately available acid sites. Conse-
quently, glucitol is the main product of cellobiose hydrogenolysis
catalyzed by the [Bmim]3PW12O40 + Ru/C catalysts.
433 K.
catalyst. The Ru/[Bmim]3PW12O40 catalyst allowing a close con-
tact between the Ru site and the O site, synergistically facilitates
the hydrogenolysis reactions.
We used pyridine adsorption and FT-IR to characterize Brønsted
acidity of the Ru/[Bmim]3PW12O40 and [Bmim]3PW12O40 + Ru/C
catalysts and presented the results in Fig. 8. As shown in Fig. 8(a),
the 1540 cm−1 peak characterizing pyridine adsorption on the
Brønsted sites is clear seen after treating the Ru/[Bmim]3PW12O40
catalyst in a H2 + N2 mixture. This peak was clearly a result of hydro-
gen treatment as there is no peak at this position in Fig. 8(b) when
the catalyst was treated with only N2. Liu et al. have also demon-
strated that hydrogen spillover was the source of Brønsted sites
on the Ru/Cs3PW12O40 catalyst [16]. These authors used pyridine-
adsorbed FT-IR and confirmed the generation of the Brønsted sites
due to hydrogen treatment. In our study, the 1540 cm−1 peak was
not observed in H2 + N2 treated [Bmim]3PW12O40 + Ru/C, indicat-
ing that there is no Brønsted site generated on this catalyst. The
inability of generating Brønsted sites on [Bmim]3PW12O40 + Ru/C is
because C can not act as an acceptor of the proton from H spillover.
Furthermore, the spatial separation between the Ru sites and the O
sites on the mixed catalyst prevented H to migrated to the O sites.
To determine the nature and role of the Ru species in
Ru/[Bmim]3PW12O40 catalyst, we performed a XPS analysis. Due
Hydrogen spillover is the main source of the Brønsted acid sites
on Ru/[Bmim]3PW12O40 catalyst [19]. The hydrogen adatom pro-
duced at the metallic Ru sites will likely spillover onto the O or W
sites of the polyacids, generating protons and hydridic hydrogen,
respectively, on the support. The protons can then migrate through
[Bmim]+ to the solution phase, making the reaction mixture acidic.
We tested the hypothesis by measuring the pHs of reactant and
product from the cellobiose hydrogenolysis at 433 K and 5 MPa H2
for 3 h and summarized the results in Table 1. As shown in Table 2,
the pHs are essentials unchanged before and after the reaction
over the [Bmim]3PW12O40 + Ru/C catalyst. In contrast, the pH was
reduced from 4.4 to 2.5 after reaction over the Ru/[Bmim]3PW12O40
catalyst. We exclude the contribution to the pH reduction from the
acidic products, as HPLC analysis did not show any carboxylic acids
in the product. In fact, the products of the cellobiose degradation
are mainly sorbitol, with a small amount of glucose and glucitol. We
also measured the pH changes of the solution containing only the
catalysts (Ru/[Bmim]3PW12O40 or [Bmim]3PW12O40 + Ru/C) before
and after the treatment in 5 MPa H2 at 433 K for 3 h. The results, also
shown in Table 2, indicate that similar pH changes to the reaction
mixtures starting with cellobiose were achieved when the catalyst
were treated in H2 at the reaction conditions. Furthermore, the IR
spectra of the liquid products of the cellobiose reaction, as shown
in Fig. 7, further confirmed that there are no carboxyl characteristic
peaks (∼1700 cm−1) in the infrared spectra. The absorption peaks at
1415 cm−1 was attributed to hydroxyl group of hextols and those at
1650 cm−1 and 3400 cm−1 were assigned to adsorbed water. These
results demonstrate that hydrogen spillover generates Brønsted
acid sites formed during reaction over the Ru/[Bmim]3PW12O40
Table 2
The pH values of the reactive mixtures before and after the reaction at 433 K and
5 MPa H2 for 3 h with different catalysts.
Catalyst
Initial
After reaction
Catalyst only
With cellobiose
4.4
4.4
2.5
2.4
Fig. 8. Pyridine-adsorbed FT-IR spectra of (a) Ru/[Bmim]3PW12O40 after treat-
Ru/[Bmim]3PW12O40
ing in
a H2 + N2 (10:90) mixture at ambient pressure and 473 K for 1 h, (b)
Ru/[Bmim]3PW12O40 after treating in N2 at ambient pressure and 473 K for 1 h,
(c) Ru/C + [Bmim]3PW12O40 after treating in a H2 + N2 (10:90) mixture at ambient
pressure and 473 K for 1 h.
Catalyst only
With cellobiose
4.6
4.6
4.4
4.4
[Bmim]3PW12O40 + Ru/C
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
6
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
Fig. 11. Reusability of the catalyst for the conversion of cellobiose to sorbitol. Reac-
tion conditions: cellobiose, 0.25 g; 5 wt% Ru/[Bmim]3PW12O40, 0.05 g; water, 5 mL;
433 K; H2, 5 MPa.
Fig. 9. Ru 3p3/2 XPS spectra of (a) RuCl3 and (b) Ru/[Bmim]3PW12O40
.
to the fact that the binding energy of Ru 3d5/2 orbital overlaps with
that of C 1s, the Ru 3p3/2 orbital was used in the analysis. As shown
in Fig. 9, Ru 3p3/2 in Ru/[Bmim]3PW12O40 shows a broad peak cen-
tered at ∼461.05 eV, corresponding to the metallic Ru (∼76%). There
is another weak peak at 463.63 eV, which was assigned to Ru(III)
(∼24%) by comparing with the spectrum from a reference RuCl3
sample. The contribution from formation of RuOx during handling
and preparation of the sample can be excluded as the Ru 3p3/2
lines lie outside the range [34]. The results indicate that the Ru
species on Ru/[Bmim]3PW12O40 was not reduced completely to
the metallic state, and therefore, Ru(III) coexists with the metal-
lic state. Consequently, the Ru(III) sites with Lewis acidity are also
likely to contribute to the hydrolysis activity toward cellulose on
the Ru/[Bmim]3PW12O40 catalyst. In summary, in addition to the
Brønsted acidic sites generated from hydrogen spillover, the Lewis
acidic active centers (Ru(III) and W5+) on the Ru/[Bmim]3PW12O40
catalyst may also contribute to the observed hydrolysis activity.
Interestingly, we observed that the reaction mixture was dark
blue immediately after the reaction over the Ru/[Bmim]3PW12O40
catalyst, as shown in Fig. 10(a). After leaving the mixture in air
overnight, white precipitates formed and the liquid mixture turned
colorless, as shown in Fig. 10(b). The formation of solid precipitates
and change of color correspond to the oxidation of the reduced W
sites in air at room temperature formed during the hydrogenolysis
helping the catalyst to regain it activity [35]. We incorporated this
step in the dashed box of the mechanism shown in Fig. 6.
The formation of Brønsted acid active centers and the exis-
tence of Lewis acidic sites are in synergy with the metallic Ru
sites, making the Ru/[Bmim]3PW12O40 catalyst active and selective
for hydrogenolysis of cellobiose to sorbitol. The same mechanism
should be operative for the hydrogenolysis of cellulose on the
Ru/[Bmim]3PW12O40 catalyst, resulted in high selectivity to sor-
bitol.
3.4. Reusability
The reusability of catalysts has been examined using the
hydrogenolysis of cellobiose. After run of the reaction, the remain-
ingsolidwerecollectedandcentrifugedandwashedwithdeionized
water. The resulting solid was used as catalyst in the next run.
Fig. 11 summarizes the results in the bar chart the conversion
and yields of the reaction. Obviously, the conversion of cellobiose
remained almost constant in the second and third run. However,
the yield of sorbitol decreased significantly. Meanwhile, the glu-
citol yield was increased, reaching 25% in the third run. In order
to understand the change of the selectivity in different runs, we
analyzed the liquid products using the inductively coupled plasma
(ICP). The results showed that 20 wt% of the W and 0.06 wt% of
the Ru content (based on the fresh catalyst) were leached to liquid
products. We speculate that the loss of more active W reduces the
number of receiving site of spillover hydrogen, and thereby, the
hydrolysis activity. Consequently, it becomes difficult to break up
glycosidie linkages after cellobiose hydrogenation, resulting in the
increased yield of glucitol.
4. Conclusions
A catalyst consisting of supported Ru and an ionic liquid
(BmimPF6)–Heteropoly acid (H3PW12O40·nH2O) hybrid as a sup-
port has been synthesized. The catalyst, with the Ru sites being
in close contact with the oxygen sites of the support, facilitates
hydrogen spillover and generates Brønsted acid sites in situ and
acidifies the reactive mixture. On this catalyst, the Ru sites for
hydrogenation combine with both Lewis and Brønsted acidic sites
for hydrolysis, produced a synergy toward a superior catalytic per-
formance for the selective conversion of microcrystalline cellulose
to hexitols, resulted in a sorbitol selectivity of 70.3% with a micro-
crystalline cellulose conversion of 63.7% at 433 K with 5 MPa H2 in
24 h. The proximity of the hydrogenation sites and the acid sites
on the Ru/[Bmim]3PW12O40 catalyst is key to the observed supe-
rior selectivity. The synergy between the IL–HPA hybrid and the
Fig. 10. Photographs of the liquid products (a) immediately after reaction and (b)
after leaving overnight in air.
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061

GModel
CATTOD-8632; No. of Pages7
ARTICLE IN PRESS
X. Xie et al. / Catalysis Today xxx (2013) xxx–xxx
7
supported Ru was revealed through a mechanistic analysis: hydro-
gen spilled-over from the Ru site to the surface sites of the IL–HPA
hybrid and its subsequent acidification of the reactive mixture via
proton transfer through [Bmim]+ played key roles in the enhanced
catalytic performance.
[10] N. Yan, C. Zhao, C. Luo, P.J. Dyson, H.C. Liu, Y. Kou, J. Am. Chem. Soc. 128 (2006)
8714–8715.
[11] A. Fukuoka, P.L. Dhepe, Angew. Chem. Int. Ed. 45 (2006) 5161–5163.
[12] C. Luo, S.A. Wang, H.C. Liu, Angew. Chem. Int. Ed. 46 (2007) 7636–7639.
[13] R. Palkovits, K. Tajvidi, A.M. Ruppert, J. Procelewska, Chem. Commun. 47 (2011)
576–578.
[14] J. Geboers, S. Van de Vyver, K. Carpentier, K. de Blochouse, P. Jacobs, B. Sels,
Chem. Commun. 46 (2010) 3577–3579.
Acknowledgments
[15] J. Geboers, S. Van de Vyver, K. Carpentier, P. Jacobs, B. Sels, Green Chem. 13
(2011) 2167–2174.
[16] M. Liu, W.P. Deng, Q.H. Zhang, Y.L. Wang, Y. Wang, Chem. Commun. 47 (2011)
9717–9719.
[17] R.B. Levy, M. Boudart, J. Catal. 32 (1974) 304–314.
[18] W.C. Conner, J.L. Falconer, Chem. Rev. 95 (1995) 759–788.
[19] R. Prins, Chem. Rev. 112 (2012) 2714–2738.
The work was supported by Nature Sciences Foundation of China
(Grant No. 21076152 and No. 21276191), Special Research Fund
for the Doctoral Program of Higher Education of China (Grant No.
2100032110018), and the Program of Introducing Talents of Disci-
pline to Universities, China (Grant No. B06006).
[20] R.P. Swatloski, S.K. Spear, J.D. Holbrey, R.D. Rogers, J. Am. Chem. Soc. 124 (2002)
4974–4975.
[21] Z. Li, Q. Zhang, H. Liu, P. He, X. Xu, J. Li, J. Power Sources 158 (2006) 103–109.
[22] A. Corma, S. Iborra, F. Xamena, R. Monton, J.J. Calvino, C. Prestipino, J. Phys.
Chem. C 114 (2010) 8828–8836.
Appendix A. Supplementary data
[23] M. Ammam, J. Fransaer, J. Electrochem. Soc. 158 (2011) A14–A21.
[24] Y. Leng, J. Wang, D.R. Zhu, X.Q. Ren, H.Q. Ge, L. Shen, Angew. Chem. Int. Ed. 48
(2009) 168–171.
[25] Y. Leng, J. Wang, D.R. Zhu, Y.J. Wu, P.P. Zhao, J. Mol. Catal. A: Chem. 313 (2009)
1–6.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.cattod.2013.09.061.
[26] C.Z. Li, Z.K.B. Zhao, Adv. Synth. Catal. 349 (2007) 1847–1850.
[27] H. Wang, G. Gurau, R.D. Rogers, Chem. Soc. Rev. 41 (2012) 1519–1537.
[28] N. Yan, Y.A. Yuan, R. Dykeman, Y.A. Kou, P.J. Dyson, Angew. Chem. Int. Ed. 49
(2010) 5549–5553.
[29] Y.S. Qu, C.P. Huang, J. Zhang, B.H. Chen, Bioresour. Technol. 106 (2012) 170–172.
[30] Z. Sun, M.X. Cheng, H.C. Li, T. Shi, M.J. Yuan, X.H. Wang, Z.J. Jiang, RSC Adv. 2
(2012) 9058–9065.
[31] M. Liu, H. Wang, J. Han, Y. Niu, Carbohydr. Polym. 89 (2012) 607–612.
[32] C. Pazé, S. Bordiga, A. Zecchina, Langmuir 16 (2000) 8139–8144.
[33] S. Yang, J.-z. Gui, D. Liu, X.-l. Peng, Y.-h. Gao, Y.-j. Zhang, J. Petrochem. Univ. 24
(2011) 31–39.
[34] A.V. Naumkin, A. Kraut-Vass, S.W. Gaarenstroom, C.J. Powell, NIST X-ray Pho-
toelectron Spectroscopy Database, in: NIST Standard Reference Database, vol.
20, National Institute of Standards and Technology, Gaithersburg, 2012.
[35] Z.J. Tai, J.Y. Zhang, A.Q. Wang, M.Y. Zheng, T. Zhang, Chem. Commun. 48 (2012)
7052–7054.
References
[1] D.L. Klass, Biomass for renewable energy and fuels, in: J.C. Cutler (Ed.), Ency-
clopedia of Energy, Elsevier, New York, 2004, pp. 193–212.
[2] A.M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem. Int. Ed. 51 (2012)
2564–2601.
[3] D.M. Alonso, J.Q. Bond, J.A. Dumesic, Green Chem. 12 (2010) 1493–1513.
[4] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007)
7164–7183.
[5] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
[6] W.P. Deng, Q.H. Zhang, Y. Wang, Dalton Trans. 41 (2012) 9817–9831.
[7] C.H. Zhou, X. Xia, C.X. Lin, D.S. Tong, J. Beltramini, Chem. Soc. Rev. 40 (2011)
5588–5617.
[8] S. Van De Vyver, J. Geboers, P.A. Jacobs, B.F. Sels, ChemCatChem 3 (2011) 82–94.
[9] H. Kobayashi, H. Ohta, A. Fukuoka, Catal. Sci. Technol. 2 (2012) 869–883.
Please cite this article in press as: X. Xie, et al., Selective conversion of microcrystalline cellulose into hexitols over a Ru/[Bmim]3PW12O40 catalyst
under mild conditions, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.061