Conversion of cellobiose into sorbitol in neutral water medium over carbon nanotube-supported ruthenium catalysts

Article abstract of DOI:10.1016/j.jcat.2010.01.024

Carbon nanotube (CNT)-supported ruthenium catalysts were studied for the hydrogenation of cellobiose in neutral water medium. The acidity of catalysts and the size of Ru particles played key roles in the conversion of cellobiose to sorbitol. A higher concentration of nitric acid used for CNT pretreatment provided a better sorbitol yield, suggesting an important role of catalyst acidity. The catalysts with larger mean sizes of Ru particles and abundant acidic sites exhibited better sorbitol yields, while those with smaller Ru particles and less acidic sites favored the formation of 3-β-d-glucopyranosyl-d-glucitol. We elucidated that cellobiose was first converted to 3-β-d-glucopyranosyl-d-glucitol via the hydrogenolysis, and then sorbitol was formed through the cleavage of β-1,4-glycosidic bond in 3-β-d-glucopyranosyl-d-glucitol over the catalysts. The catalyst with smaller Ru particles favored the first step but was disadvantageous to the second step due to the less acidity. Smaller Ru particles also accelerated the degradation of sorbitol.

Full text of DOI:10.1016/j.jcat.2010.01.024

Journal of Catalysis 271 (2010) 22–32
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Conversion of cellobiose into sorbitol in neutral water medium over carbon
nanotube-supported ruthenium catalysts
*
*
Weiping Deng, Mi Liu, Xuesong Tan, Qinghong Zhang , Ye Wang
State Key Laboratory of Physical Chemistry of Solid Surfaces, National Engineering Laboratory for Green Chemical Productions of Alcohols, Ethers and Esters,
Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Carbon nanotube (CNT)-supported ruthenium catalysts were studied for the hydrogenation of cellobiose
in neutral water medium. The acidity of catalysts and the size of Ru particles played key roles in the con-
version of cellobiose to sorbitol. A higher concentration of nitric acid used for CNT pretreatment provided
a better sorbitol yield, suggesting an important role of catalyst acidity. The catalysts with larger mean
sizes of Ru particles and abundant acidic sites exhibited better sorbitol yields, while those with smaller
Received 30 August 2009
Revised 6 January 2010
Accepted 26 January 2010
Available online 24 February 2010
Ru particles and less acidic sites favored the formation of 3-b-
that cellobiose was first converted to 3-b- -glucopyranosyl-
sorbitol was formed through the cleavage of b-1,4-glycosidic bond in 3-b-
over the catalysts. The catalyst with smaller Ru particles favored the first step but was disadvantageous
to the second step due to the less acidity. Smaller Ru particles also accelerated the degradation of sorbitol.
Ó 2010 Elsevier Inc. All rights reserved.
D
-glucopyranosyl-
-glucitol via the hydrogenolysis, and then
-glucopyranosyl- -glucitol
D-glucitol. We elucidated
Keywords:
Biomass conversion
Cellobiose
D
D
D
D
Sorbitol
Ruthenium catalyst
Carbon nanotubes
Hydrogenation
Hydrolysis
1. Introduction
Pt/Al2O3 catalyst, providing a yield of 31% to hexitols (sorbitol
and mannitol, 25% and 6%, respectively) at 463 K [12]. Liu and
The production of fuels and chemicals from renewable biomass
resources has attracted much attention in recent years [1–4]. As
the most abundant source of biomass and because of the non-
edible feature, lignocellulosic biomass may become an important
feedstock to replace or partially replace the fossil feedstock for
the sustainable production of fuels and chemicals [5–8]. However,
the effective utilization of lignocellulosic biomass, which contains
cellulose as a main component, is still a challenge because of the
robust crystalline structure of cellulose [9,10]. So far, processes
for hydrolysis of cellulose to glucose in the presence of strong
mineral acids (e.g., H2SO4) and for high-temperature pyrolysis or
gasification of cellulose to bio-oils or synthesis gas have been
developed, but these processes suffer from problems of high-en-
ergy input and low selectivity [5–8]. It would be highly desirable
to develop a catalytic route for the conversion of cellulose selec-
tively into a platform or building block molecule such as sorbitol,
ethylene glycol or 5-hydroxymethylfurfural (HMF) [11], which
may be facilely transformed into fuels or chemicals.
coworkers [13] developed a two-step transformation of cellulose
to polyols catalyzed by reversibly formed acids and activated car-
bon-supported Ru nanoclusters in hot water, and they obtained a
yield of polyols of ꢀ40% (sorbitol, ꢀ30%) at 518 K. Ni-promoted
tungsten carbide was demonstrated to catalyze the conversion of
cellulose into ethylene glycol with a yield as high as 61% at
518 K [14]. Zhang and coworkers recently developed an effective
route for the rapid conversion of cellulose to sugars and further
to HMF (HMF yield, ꢀ55%) catalyzed by CuCl2/CrCl2 catalysts in
1-ethyl-3-methyl-imidazolium chloride solvent at 353–393 K
[15]. Very recently, we found that a multi-walled carbon nanotube
(CNT)-supported Ru catalyst could catalyze the conversion of cellu-
lose to hexitols with a yield of 40% (sorbitol, 36%) in the presence of
H2 in water medium at 458 K [16]. However, basic understanding
of catalyst requirements for the conversion of cellulose is very lim-
ited. Undoubtedly, more extensive studies are needed to gain in-
sights into the requirements for the rational design of more
efficient catalysts for selective transformations of cellulose.
A few studies have succeeded in converting cellulose into such a
platform molecule under mild conditions [12–16]. The hydrogena-
tion of cellulose in water medium was found to be catalyzed by a
However, because cellulose is a very complex macromolecule
and is insoluble in most solvents, it is not easy to perform funda-
mental research directly with cellulose. In this context, the funda-
mental studies with a model molecule would be helpful in the
present stage. Cellobiose, which is a D-glucose dimer connected
by a b-1,4-glycosidic bond (see Fig. 1 for structural formula),
represents the simplest model of cellulose. The studies on catalytic
* Corresponding authors. Fax: +86 592 2183047.
E-mail addresses: zhangqh@xmu.edu.cn (Q. Zhang), wangye@xmu.edu.cn (Y.
Wang).
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.01.024

W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
23
post-treatments. For this purpose, the dried sample was either di-
rectly reduced by H2 at 623 and 773 K or was first calcined at 623 K
in air and then reduced by H2 at different temperatures (623–
773 K). An ethylene glycol reduction method [24] was also applied
to the preparation of the Ru/CNT catalysts with different sizes of Ru
particles. In this method, RuCl3 was first dissolved in ethylene gly-
col, and then, the CNTs after pretreatment were added into the
RuCl3 solution. After being treated ultrasonically for 0.5 h, the mix-
ture was refluxed at 453 or 483 K for 1 h. The solid product was
then recovered by filtration followed by drying.
Fig. 1. Structure formulas of cellobiose and some typical products.
2.2. Catalyst characterization
conversion of cellobiose may also be useful for transformations of
the decrystallized or the soluble oligosaccharides released in
hydrothermal or acidic treatments of cellulose, which contain b-
1,4-glycosidic bonds. However, there only exist scattered studies
on catalytic conversion of cellobiose. Kou and coworkers [17] dis-
closed that Ru nanoclusters dispersed in water were efficient for
the hydrogenation of cellobiose to sorbitol in an acidic aqueous
medium (pH = 2.0), whereas under neutral or basic conditions
(pH = 7.0 or 10.0), the selectivity of sorbitol was significantly low-
er. Thus, the protons in the liquid phase might participate in the
hydrolysis of cellobiose. Bootsma and Shanks [18] reported that a
kind of solid acid catalysts, i.e., organic–inorganic hybrid mesopor-
ous materials containing acidic functional groups, could catalyze
the hydrolysis of cellobiose into glucose.
Supported Ru catalysts are known as efficient catalysts for the
hydrogenation of glucose to sorbitol [19,20]. The Ru/CNT catalyst
was once reported to be more active for the hydrogenation of glu-
cose than the Ru/Al2O3 and Ru/SiO2 [21]. As mentioned earlier, in
our preceding work, we found that the Ru/CNT catalyst could effi-
ciently catalyze the conversion of cellulose to sorbitol in the pres-
ence of H2 in water medium [16]. However, there is still little
knowledge about the effect of the Ru/CNT catalyst on the conver-
sion of cellulose to sorbitol. Very recently, we chose cellobiose as
a model molecule of cellulose and performed detailed studies on
catalytic conversion of cellobiose. The present article reports the
effects of key factors of Ru/CNT catalysts on the catalytic hydroge-
nation of cellobiose to sorbitol. We will also discuss the possible
reaction mechanism for this catalytic reaction.
Transmission electron microscopy (TEM) measurements were
performed on a FEI Tecnai 30 electron microscope (Phillips Analyt-
ical) operated at an acceleration voltage of 300 kV. The mean sizes
of Ru particles in Ru/CNT samples were estimated from TEM
micrographs by counting ca. 150–200 particles. X-ray photoelec-
tron spectra (XPS) were recorded with a Quantum 2000 Scanning
ESCA Microprob instrument (Physical Electronics) using Al K radi-
a
ation. The binding energy was calibrated using C1s photoelectron
peak at 284.6 eV as a reference. Ru dispersions were measured
by H2AO2 titration using an ASAP2010C Micromeritics apparatus
with the procedures reported in literature [25].
NH3-temperature-programmed desorption (NH3-TPD) was per-
formed on a Micromeritics AutoChem 2920 II instrument. Typi-
cally, the sample loaded in a quartz reactor was first pretreated
with high-purity He at 623 K for 1 h. After the sample was cooled
to 393 K, NH3 adsorption was performed by switching the He flow
to a NH3AHe (10 vol.% NH3) gas mixture and then keeping at 393 K
for 1 h. Then, the gas phase or the weakly adsorbed NH3 was
purged by high-purity He at the same temperature. NH3-TPD was
performed in the He flow by raising the temperature to 973 K at
a rate of 10 K minꢁ1, and the desorbed NH3 molecules were
detected by ThermoStar GSD 301 T2 mass spectrometer with the
signal of m/e = 16.
Titration method was also used to evaluate the acidity of
Ru/CNT catalysts. In a typical experiment, 0.15 g Ru/CNT catalysts
was added into a 25 cm3 0.01 mol dmꢁ3 NaOH aqueous solution
and stirred overnight. The mixture was titrated with a 0.01
mol dmꢁ3 HCl solution to determine the excess NaOH in the solu-
tion to quantify the concentration of the acidic sites on Ru/CNT cat-
alysts. For comparison, the acidity of CNT samples without Ru was
also evaluated by the titration method.
2. Experimental
2.1. Catalyst preparation
2.3. Catalytic reaction
The CNTs with outer diameters of 20–80 nm and inner diame-
ters of 3–5 nm were prepared by a method reported previously
[22]. The prepared CNTs were typically pretreated in concentrated
HNO3 (68 wt.%) at 383 K under refluxing conditions to remove the
remaining Ni catalyst used for CNT preparation, the amorphous
carbon, and to create function groups (e.g., hydroxyl and carboxylic
groups) for anchoring metal precursors [23]. To investigate the role
of CNT functionalization, CNTs were also pretreated by HNO3 with
different concentrations (5–68 wt.%) or by concentrated HCl
(37 wt.%). No Ni was detected after these pretreatments. Stan-
dardly, CNT-supported Ru catalysts were prepared by an impreg-
nation method. The CNTs after pretreatment were added into a
RuCl3 aqueous solution and then were dispersed ultrasonically
for 0.5 h. After being further stirred for 5 h, the suspension was
evaporated at 343 K to remove water. The dried sample was cal-
cined at 623 K in air, followed by H2 reduction at 623 K for 0.5 h
to obtain the Ru/CNT catalyst. The loading of Ru was 1.0 wt.% un-
less otherwise stated.
The conversion of cellobiose was performed with a batch-type
high-pressure autoclave reactor. Typically, the catalyst (0.050 g)
and cellobiose (0.50 mmol) were added into a Teflon-lined stain-
less steel reactor pre-charged with H2O (20 cm3), and then the
reaction was carried out at 458 K under 5 MPa H2 for 3 h. After
the reaction, the solid catalyst was separated by centrifugation,
and the liquid products were analyzed by a HPLC (Shimazu LC-
20A) equipped with a RI detector and a Transgenomic™ CARBON-
Sep CHO-620 column (10
l
m, 6.5 ꢂ 300 mm). The eluent was
water with a flow rate of 0.5 cm3 minꢁ1. The column was thermo-
stated at 338 K by a column heater. Sampling loop has a volume of
20
l
L. The pH value of the reaction solution was ꢀ7 after the con-
version of cellobiose. Chemicals including sorbitol, mannitol,
erythritol [C4H6(OH)4], HMF purchased from Alfa Aesar, and glu-
cose, glycerol, ethylene glycol purchased from Sinopharm Chemi-
cal Reagent Co. Ltd. were used for calibrations without further
treatment. 3-b-D-Glucopyranosyl-D-glucitol synthesized in our lab-
We have attempted to prepare Ru/CNT catalysts with different
sizes of Ru particles by the impregnation followed by different
oratory, which was characterized by mass spectroscopy, was also
used for the calibration.

24
W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
3. Results and discussion
3.1. Role of CNT functionalization in catalytic conversion of cellobiose
over Ru/CNT catalysts
Because the protons in liquid phase were indispensable for sor-
bitol formation in the conversion of cellobiose catalyzed by water-
dispersed Ru nanoclusters [17], the hydrolysis and hydrogenation
were proposed to be two requisite steps for the conversion of cel-
lobiose to sorbitol. To realize the conversion of cellobiose to sorbi-
tol in neutral water medium, we selected Ru supported on acid-
functionalized CNTs as the catalyst for this reaction. We prepared
Ru/CNT catalysts, in which the CNT was pretreated by the concen-
trated HCl solution (37 wt.%) or the HNO3 solutions with concen-
trations in the range of 5–68 wt.% to generate acidic functional
groups [26].
Characterizations with XPS and TEM were performed for this
series of catalysts to gain information about the state of Ru species.
We did not find significant differences in the chemical state and
the mean size of Ru particles in these catalysts. XPS studies re-
vealed that the binding energy of Ru 3d5/2 over each catalyst
was around 280.3 eV, suggesting that the Ru species loaded on
the CNTs pretreated differently were all in metallic (Ru0) state
[27,28]. Fig. 2 shows the TEM micrographs of the Ru/CNT catalysts
with CNTs pretreated by HNO3 with different concentrations. The
size distributions for Ru particles in these catalysts, derived from
the TEM micrographs by counting ꢀ150–200 particles, are also
shown in Fig. 2. With changing the concentration of HNO3 used
for CNT pretreatment from 5 to 68 wt.%, the mean sizes of Ru par-
ticles in these catalysts were almost the same (8.6–8.9 nm).
NH3-TPD results in Fig. 3 show that almost no desorption of NH3
occurs over the CNT pretreated by HCl. On the other hand, desorp-
tion of NH3 was observed from the CNTs pretreated by HNO3, and
the peak intensity increased with increasing the concentration of
HNO3. Desorption of NH3 was also observed from the Ru/CNT cat-
alysts prepared using CNTs pretreated by HNO3 with different con-
centrations (Fig. 3B). This indicates that the acidic sites generated
on CNT surfaces could be sustained on the prepared Ru/CNT cata-
lysts. Similar phenomenon was observed in our recent studies on
the same catalysts for Fischer–Tropsch synthesis [29].
We performed the hydrogenation of cellobiose over the acid-
functionalized Ru/CNT catalysts. As shown in Fig. 4, the Ru/CNT
could catalyze the formation of sorbitol from cellobiose at
458 K, and the catalytic performance depended on the acid used
for CNT pretreatment. Sorbitol yield was only 26% when the
CNT pretreated by HCl (37 wt.%) was used as the support of Ru
catalyst. Sorbitol yield rose from 56% to 87% when the concentra-
tion of HNO3 for CNT pretreatment increased from 5 wt.% to
68 wt.%. Therefore, the acidity generated on CNTs during the pre-
treatment by concentrated HNO3 plays an important role in the
conversion of cellobiose to sorbitol. This result is in essence the
same with that obtained in our previous studies for the conver-
sion of cellulose to sorbitol [16]. The following studies have been
focused on the catalysts using the CNT pretreated by 68 wt.%
HNO3 as the support.
Fig. 2. TEM micrographs and Ru particle size distributions of the Ru/CNT catalysts
with CNTs pretreated by HNO3 with different concentrations. Concentration of
HNO3 used for CNT pretreatment: (A) 5 wt.%, (B) 19 wt.%, (C) 37 wt.%, (D) 52 wt.%,
and (E) 68 wt.%.
may be an important intermediate for sorbitol formation from cel-
lobiose over the Ru/CNT catalyst.
3.2. Ru/CNT catalysts with different sizes of Ru particles and their
catalytic behaviors
To gain further information on the conversion of cellobiose to
sorbitol, we have investigated the temperature dependence of
product distributions for cellobiose conversion over the Ru/CNT
catalyst with the CNT pretreated by 68 wt.% HNO3. Fig. 5 shows
3.2.1. Preparation of Ru/CNT catalysts with different sizes of Ru
particles
that 3-b-D-glucopyranosyl-D-glucitol is formed as the main product
Besides the acidity of the catalyst, Ru nanoparticles are believed
to play important roles in the formation of sorbitol from cellobiose.
Because the size of metal nanoparticles is one of the most impor-
tant factors dominating the performances of nanoparticles-based
catalysis [30], we have prepared a series of Ru/CNT samples with
at lower temperatures, and it is transformed to sorbitol with
increasing the reaction temperature up to 458 K. A further higher
temperature favored the formation of degradation products includ-
ing C6H10(OH)4, C4H6(OH)4, C3H5(OH)3, C2H4(OH)2, and CH4. From
these results, we suggest that 3-b-D-glucopyranosyl-D-glucitol

W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
25
Fig. 5. Dependence of catalytic performances with reaction temperature for the
conversion of cellobiose over the Ru/CNT catalyst with a mean size of Ru particles at
8.7 nm. Reaction conditions: cellobiose, 0.50 mmol; catalyst, 0.050 g; H2O, 20 cm3;
H2, 5 MPa; time, 3 h.
reduction temperatures of 623 K (Fig. 6B, Ru/CNT-C623-H623) and
773 K (Fig. 6C, Ru/CNT-C623-H773), respectively. On the other
hand, Ru/CNT catalysts with mean sizes of Ru particles at 5.1 and
6.8 nm could be obtained by using the method of ethylene glycol
reductions at 483 K (Fig. 6D, Ru/CNT-EG483) and 453 K (Fig. 6E,
Ru/CNT-EG453), respectively. In short, we have succeeded in pre-
paring the Ru/CNT catalysts with mean sizes of Ru particles varying
from 2.4 to 12 nm.
Fig. 3. NH3-TPD profiles of the CNTs pretreated by concentrated HCl or by HNO3
with different concentrations (A) and the Ru/CNT catalysts prepared using CNTs
pretreated by HNO3 with different concentrations (B).
3.2.2. Acidity of the prepared Ru/CNT catalysts with different sizes of
Ru particles
From the results described previously, we know that the forma-
tion of sorbitol is affected by the catalyst acidity, which arises from
the CNT pretreatment by concentrated HNO3. Therefore, we have
evaluated the acidity of the prepared Ru/CNT catalysts with differ-
ent mean sizes of Ru particles by both the NH3-TPD and the titra-
tion methods.
Fig. 7 shows the NH3-TPD profiles of these catalysts. Only a low-
er-temperature NH3 desorption peak (ꢀ480 K) was observed for
the Ru/CNT catalyst with a mean size of Ru particles at 2.4 nm,
and the intensity of this peak was lower than that for other cata-
lysts. Moreover, for the catalysts with mean sizes of Ru particles
P5.1 nm, in addition to the lower-temperature peak, another
NH3 desorption peak at higher temperatures (>700 K) could be ob-
served, indicating the presence of acid sites with stronger acidity,
and the peak associated with the stronger acid sites further shifts
to higher temperatures over the samples with Ru particles at
8.7 nm or 12 nm. From these NH3-TPD results, it becomes clear
that there exist differences in the acidity among the Ru/CNT cata-
lysts with different mean sizes of Ru particles. When compared to
the Ru/CNT catalysts with larger Ru particles, the Ru/CNT catalyst
with a mean size of Ru particles at 2.4 nm (Ru/CNT-H773) pos-
sesses much lower acidity. This observation has further been con-
firmed by the result obtained from the titration method. As shown
in Table 1, the amount of acidic sites for the Ru/CNT-H773 (2.4 nm)
catalyst was significantly lower than those for the other Ru/CNT
catalysts. The amounts of acidic sites over the CNTs alone, which
underwent post-treatments with the same procedures as those
for the preparation of the Ru/CNT catalysts with different Ru sizes,
were also measured by the titration method. When compared to
the CNTs after different post-treatments, unexpectedly, the Ru/
CNT catalysts showed larger amounts of acidic sites.
Fig. 4. Sorbitol yield in the conversion of cellobiose over the Ru/CNT catalysts
prepared using CNTs pretreated by concentrated HCl or by HNO3 with different
concentrations. Reaction conditions: cellobiose, 0.50 mmol; catalyst, 0.050 g; H2O,
20 cm3; H2, 5 MPa; temperature, 458 K; time, 3 h.
different mean sizes of Ru particles to clarify the size effect in the
Ru/CNT-catalyzed conversion of cellobiose. With TEM observa-
tions, we clarified that the direct H2 reduction at 773 K without
calcination resulted in smaller Ru nanoparticles finely dispersed
on the CNT surfaces (Fig. 6A, Ru/CNT-H773). The mean size of Ru
(D) in this sample estimated from TEM images by counting ca.
150–200 particles was 2.4 nm. The size of Ru particles increased
significantly if the calcination at 623 K was adopted before H2
reduction. Moreover, a change of the temperature for H2 reduction
could change the mean size of Ru particles. The catalysts with
mean sizes of Ru particles at 8.7 and 12 nm were obtained by using

26
W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
Fig. 7. NH3-TPD profiles of the Ru/CNT catalysts with different mean sizes of Ru
particles. (A) Ru/CNT-H773 (Ru, 2.4 nm); (B) Ru/CNT-EG483 (Ru, 5.1 nm); (C) Ru/
CNT-EG453 (Ru, 6.8 nm); (D) Ru/CNT-C623-H623 (Ru, 8.7 nm); (E) Ru/CNT-C623-
H773 (Ru, 12 nm).
Table 1
Amount of acidic sites over the CNTs after different post-treatments and the Ru/CNT
samples with different mean sizes of Ru particles.a
Sampleb
Mean size of
Ru (nm)
Amount of acidic sites
(mmol gꢁ1
)
CNT-H773
CNT-EG483
CNT-EG453
CNT-C623-H623
CNT-C623-H773
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
–
–
–
–
0.12
0.20
0.22
0.25
0.17
0.37
0.50
0.55
0.56
0.51
–
2.4
5.1
6.8
8.7
12
a
Measured by the titration method.
The numbers after H, C, and EG denote temperatures for H2 reduction, calci-
b
nation and ethylene glycol reduction, respectively (also see the main text).
Fig. 6. TEM micrographs and Ru particle size distributions of the Ru/CNT catalysts
prepared by the impregnation (A–C) and the ethylene glycol reduction (D and E)
methods. (A) Direct reduction by H2 at 773 K after impregnation (without
calcination); (B) and (C) with calcination at 623 K after impregnation, followed by
H2 reductions at 623 and 773 K, respectively; (D) and (E) reductions by ethylene
glycol at 483 and 453 K, respectively.
Table 2. The percentage of the composition at a binding energy
of 534.2 eV, which could be ascribed to the acidic carboxylic group
on the CNT surface, increased from 8.8% to 16%, 17%, 22%, and 18%
when the mean size of Ru particles rose from 2.4 to 5.1, 6.8, 8.7,
and 12 nm, respectively. These results further suggest that the cat-
alyst with a mean size of Ru particles at 2.4 nm possesses less acid
sites. However, from Table 2, the differences in the fraction of the
acidic carboxylic groups among the catalysts with other mean sizes
of Ru particles are not very significant. Moreover, it is still difficult
to explain the differences in the peak positions of the higher-tem-
perature peak observed in NH3-TPD profiles among the catalysts
with different mean sizes of Ru particles (Fig. 7).
Although the reason for the differences in the acidity among the
Ru/CNT catalysts with different Ru sizes and the CNTs is still un-
clear at this moment, the acidity of our catalysts is believed to stem
from the oxygen-containing functional groups on CNT surfaces
generated during the HNO3 pretreatment or the post-treatments.
Several groups [31–33] reported that the analysis of O1s XPS spec-
tra could give useful information on the functional groups on CNT
surfaces. For example, it was proposed that the O1s peak with
binding energies at 531.1, 532.3, 533.3, and 534.2 eV could be
attributed to the carbonyl groups, the carbonyl oxygen atoms in es-
ters and anhydrides or the oxygen atoms in hydroxyls or ethers,
the ether oxygen atoms in esters and anhydrides, and the oxygen
atoms in carboxylic groups, respectively [32–34]. Thus, we have
performed XPS studies for our catalysts with different mean sizes
of Ru particles, and the results are shown in Fig. 8. The broad fea-
ture of O1s peak in Fig. 8 implies that several types of oxygen-con-
taining functional groups co-exist on our catalysts. The results
derived from the deconvolution of O1s peaks are summarized in
3.2.3. Catalytic conversions of cellobiose to sorbitol over Ru/CNT
catalysts
Fig. 9 shows the catalytic performances of the Ru/CNT catalysts
with different mean sizes of Ru for the conversion of cellobiose at
458 K for 3 h. The catalyst with a smaller mean size of Ru (2.4 nm)
showed a lower sorbitol yield. The sorbitol yield increased with
increasing the mean size of Ru particles up to 8.7 nm, and a further
increase in the mean size of Ru particles to 12 nm only slightly
changed sorbitol yield. 3-b-D-Glucopyranosyl-D-glucitol, mannitol,
and degradation products (including C6H10(OH)4, C4H6(OH)4,

W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
27
Fig. 9. Product yields in the conversion of cellobiose over the Ru/CNT catalysts with
different mean sizes of Ru particles. Reaction conditions: cellobiose, 0.50 mmol;
catalyst, 0.050 g; H2O, 20 cm3; H2, 5 MPa; temperature, 458 K; time, 3 h.
3.3. Possible reaction mechanism for the conversion of cellobiose over
Ru/CNT catalysts
3.3.1. Reaction pathways
To understand the possible reaction pathways for the conver-
sion of cellobiose over the Ru/CNT catalysts, we performed kinetic
studies. Fig. 10 shows the time courses for cellobiose conversions
over the Ru/CNT catalysts with mean sizes of Ru particles at 2.4,
5.1, 8.7, and 12 nm. Over all of these catalysts, it is found that 3-
b-
D
-glucopyranosyl-
D
-glucitol is formed as the main product at
-glucopyranosyl- -gluc-
Fig. 8. O1s XPS spectra of the Ru/CNT catalysts with different mean sizes of Ru
particles. The dotted lines are deconvolution results. (A) Ru/CNT-H773 (Ru, 2.4 nm);
(B) Ru/CNT-EG483 (Ru, 5.1 nm); (C) Ru/CNT-EG453 (Ru, 6.8 nm); (D) Ru/CNT-C623-
H623 (Ru, 8.7 nm); and (E) Ru/CNT-C623-H773 (Ru, 12 nm).
the initial reaction stage. The yield of 3-b-
D
D
itol could reach 93% over the catalyst with a mean size of Ru par-
ticles at 2.4 nm after 20 min of reaction (Fig. 10A). With prolonging
the reaction time, the yield of sorbitol, the target product, in-
creased significantly along with a decrease in that of 3-b-
pyranosyl- -glucitol, indicating that sorbitol was formed from
the consecutive conversion of 3-b- -glucopyranosyl- -glucitol.
D-gluco-
D
C3H5(OH)3, and C2H4(OH)2) were formed with higher yields over
the catalysts with smaller Ru particles (<8.7 nm). These observa-
tions suggest that the mean size of Ru nanoparticles is one of the
important factors, which influence the conversion of cellobiose
into sorbitol. However, the acidity of these catalysts is different
as shown in Table 1, Figs. 7 and 8, while the result in Fig. 4 has indi-
cated the important role of acidity in the formation of sorbitol. The
contribution of acidity in these catalysts will be discussed in Sec-
tion 3.3.3.
We have performed the recycling uses of the Ru/CNT-C628-
H628 (Ru, 8.7 nm) catalyst, which can provide the highest sorbitol
yield. No decreases in sorbitol yield were observed in the repeated
uses, and a sorbitol yield of 88% was obtained after four recycling
tests. Thus, our Ru/CNT catalyst could be used repeatedly.
D
D
For the Ru/CNT catalysts with mean Ru sizes of 8.7 (Fig. 10C) and
12 nm (Fig. 10D), the yield of sorbitol could reach 80–90%, while
the highest sorbitol yields were lower than ꢀ40% and ꢀ60% over
the catalysts with mean Ru sizes of 2.4 nm (Fig. 10A) and 5.1 nm
(Fig. 10B), respectively. The lower yield of sorbitol over the Ru/
CNT catalysts with smaller Ru particles was likely due to the rapid
conversion of sorbitol consecutively to mannitol and other degra-
dation products over these catalysts. Glucose was formed with
very lower yields (<5%) over all of the catalysts in Fig. 10, except
for that over the Ru/CNT-C623-H773 catalyst with a larger mean
size of Ru (12 nm), where a relatively higher yield of glucose
(ꢀ20%) could be achieved at the initial reaction stage (Fig. 10D).
These observations strongly suggest that the main reaction in the
Table 2
Deconvolution results of XPS O1s peaks for the Ru/CNT catalysts with different mean sizes of Ru particles.
Catalyst
Mean sizes of Ru (nm)
Fraction of each O1s component (%)
531.1 eVa
532.3 eVb
533.3 eVc
534.2 eVd
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
2.4
5.1
6.8
8.7
12
40
11
20
19
12
35
43
37
27
46
16
30
26
32
24
8.8
16
17
22
18
a
b
c
The C@O groups at 531.1 eV.
The carbonyl oxygen atoms in esters, amides, anhydrides or oxygen atoms in hydroxyls or ethers at 532.3 eV.
The ether oxygen atoms in esters and anhydrides at 533.3 eV.
d
The oxygen atoms in carboxyl groups at 534.2 eV.

28
W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
Fig. 10. Time courses for the conversions of cellobiose over the Ru/CNT catalysts with different mean sizes of Ru particles. (A) Ru/CNT-H773 (Ru, 2.4 nm); (B) Ru/CNT-EG483
(Ru, 5.1 nm); (C) Ru/CNT-C623-H623 (Ru, 8.7 nm); and (D) Ru/CNT-C623-H773 (Ru, 12 nm). Reaction conditions: cellobiose, 0.50 mmol; catalyst, 0.050 g; H2O, 20 cm3; H2,
5 MPa; temperature, 458 K.
first step for cellobiose conversions over the Ru/CNT catalysts is the
formation of 3-b- -glucopyranosyl- -glucitol via the hydrogenoly-
sis of the CAO bond in one glucose ring but not the formation of
glucose by the cleavage of the b-1,4-glycosidic bond. Over Ru/
CNT catalysts with smaller Ru particles, it is hard to obtain high
sorbitol yield because of its rapid degradation.
On the basis of these results, we propose reaction pathways for
the conversion of cellobiose in Fig. 11. Over the Ru/CNT catalysts,
3-b-D-glucopyranosyl-D-glucitol is first formed by the hydrogenol-
ysis of cellobiose as the main primary product (Step I). Only over
the catalyst with a larger mean size of Ru particles (12 nm), where
the hydrogenation activity of the Ru particles may be not very high,
glucose can be formed as a minor primary product through hydro-
lysis catalyzed by the acid sites (Step I0). The subsequent transfor-
radation products could be completed even in 10 min over the Ru/
CNT catalysts with any mean size of Ru particles (Step IV and Step
V).
The reaction pathways we have proposed in Fig. 11 are quite
different from those suggested in a previous study [17]. During cel-
lobiose conversions over Ru particles dispersed in neutral aqueous
solutions (pH = 7), Kou and coworkers [17] also found the forma-
D
D
tions of 3-b-
tol (minor product). However, they suggested that the formation of
sorbitol had no relations with the 3-b- -glucopyranosyl- -glucitol
D-glucopyranosyl-D-glucitol (major product) and sorbi-
D
D
formed. Instead, they proposed that sorbitol was formed by the
cleavage of b-1,4-glycosidic bond through the direct hydrogenation
of cellobiose based on the dideoxyhexitol detected. In our work,
although dideoxyhexitol has really been detected as a minor prod-
uct (selectivity, ꢀ1%) over some Ru/CNT catalysts, our results in
Figs. 5 and 10 clearly indicate that the consecutive conversion of
3-b-D-glucopyranosyl-D-glucitol is the main path for sorbitol for-
mation. We speculate that the very small amount of dideoxyhexi-
tol may be formed by the consecutive dehydroxylation of sorbitol.
mation of 3-b-D-glucopyranosyl-D-glucitol may provide a mole of
glucose together with the formation of sorbitol through hydrolysis
(Step II). However, we only detected a small amount of glucose in
our systems (Figs. 5 and 10). This allows us to speculate that glu-
cose may be converted to sorbitol very rapidly over the Ru/CNT
catalysts under the current reaction conditions (Step III). This has
been confirmed by the experimental results for the conversion of
glucose over the Ru/CNT catalysts (Table 3). The hydrogenation
of glucose to sorbitol, mannitol, and a small amount of other deg-
3.3.2. Rate of cellobiose conversion over Ru/CNT catalysts
To gain information about the functions of the Ru/CNT catalyst
in each reaction steps in Fig. 11, we have investigated the intrinsic

W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
29
Fig. 11. Proposed reaction pathways for the conversion of cellobiose over the Ru/CNT catalysts.
Table 3
Hydrogenation of glucose over the Ru/CNT catalysts with different mean sizes of Ru
nanoparticles.a
Catalyst
Mean size
of Ru (nm)
Conversion
(%)
Selectivity (%)
Sorbitol
Mannitol
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-C623-H623
Ru/CNT-C623-H773
2.4
5.1
8.7
12
100
100
100
100
86
86
87
88
4.0
3.6
1.0
2.0
a
Reaction conditions: glucose, 1.0 mmol; catalyst, 0.050 g; H2, 5 MPa; H2O,
20 cm3; temperature, 458 K; time, 10 min.
activities of our catalysts in cellobiose conversions. To obtain the
genuine activity of each catalyst, we have chosen reaction condi-
tions (relatively milder temperature) to keep a relatively low cello-
biose conversion. Within initial 30 min at 423 K over the Ru/CNT
catalysts (Fig. 12), the main product for cellobiose conversions
was 3-b-D-glucopyranosyl-D-glucitol. We have also examined in
detail the blank reactions and the reactions over the CNTs after dif-
ferent post-treatments similar to those used for the preparation of
Ru/CNT catalysts with different mean Ru sizes. The results summa-
rized in Table 4 show that the blank reaction does not occur under
the conditions of Table 4. CNTs could provide a small amount of
glucose and HMF probably due to the hydrolysis of cellobiose to
glucose and the subsequent dehydration of glucose to HMF over
the acidic sites on CNTs. However, no 3-b-D-glucopyranosyl-D-gluc-
itol or sorbitol was formed over the CNTs alone. Thus, we can con-
lcude that Ru nanoparticles are responsible for the formations of 3-
b-D-glucopyranosyl-D-glucitol and sorbitol. From the relationship
between the conversion and the reaction time in Fig. 12A, the rate
of cellobiose conversion has been calculated, and the results are
summarized in Table 5. The conversion rate decreased significantly
with increasing the mean size of Ru particles. Because Ru is ex-
Fig. 12. Catalytic behaviors for the conversion of cellobiose at the initial stage over
the Ru/CNT catalysts with different mean sizes of Ru particles. (A) Conversion and
(B) selectivity. Reaction conditions: cellobiose, 0.79 mmol; catalyst, 0.020 g; H2O,
20 cm3; H2, 2 MPa; temperature, 423 K.
pected to catalyze the hydrogenation of cellobiose to 3-b-
D-gluco-
measured by TEM [34]. As shown in Table 5, the catalysts with
smaller mean sizes of Ru particles (2.4 and 5.1 nm) exhibited high-
er TOFs. In other words, these two catalysts are more active toward
pyranosyl- -glucitol, we have evaluated the turnover frequency
D
(TOF) for each catalyst on the basis of the rate of cellobiose conver-
sion per surface Ru atom. The dispersion of Ru, i.e., the fraction of
surface Ru atoms in the whole Ru atoms, was measured by a
H2AO2 titration technique [25]. The value thus measured was in
good agreement with the results estimated from the size of Ru par-
ticles by the following equation assuming spherical Ru particles:
dispersion = 1.32/D (nm), where D is the mean size of Ru particles
the conversion of cellobiose to 3-b-D-glucopyranosyl-D-glucitol
(i.e., hydrogenolysis of the CAO bond of one glucose ring, see
Step I in Fig. 11). Smaller metal particles are generally believed
to possess larger fractions of coordinately unsaturated Ru atoms.
Our present result indicates that the coordinately unsaturated Ru
atoms are more active toward the hydrogenolysis of cellobiose.

30
W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
Table 4
Conversions of cellobiose in the blank reactions and the reactions over CNTs after different post-pretreatments and Ru/CNTs with different mean sizes of Ru particles.a
Sample
Mean size of Ru (nm)
Cellobiose conversion (%)
Yield (%)
Sorbitol
Glucose
HMF
Glucitolb
Blank
CNT-H773
CNT-EG483
CNT-EG453
CNT-C623-H623
CNT-C623-H773
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
–
–
–
–
–
–
2.4
5.1
6.8
8.7
12
<0.1
1
5
5
7
0
0
0
0
0
0
1
2
2
4
2
0
0
0
0
0
0
0
0.8
2.4
3
2.9
2
0
0
0
0
0.1
1.4
1.2
4
2
0
0
0
0
0
4
0
91
65
38
15
5
88
58
36
11
3
0
a
Reaction conditions: cellobiose, 0.79 mmol; catalyst, 0.020 g; H2O, 20 cm3; H2, 2 MPa; temperature, 423 K; time, 30 min.
Glucitol denotes 3-b-D-glucopyranosyl-D-glucitol.
b
Table 5
Rate and turnover frequency for cellobiose conversions at initial reaction stage over the Ru/CNT catalysts with different mean sizes of Ru particles.
Cellobiose conversion ratec ðmmol gꢁ1 hꢁ1
Þ
Catalyst
Mean size of Ru (nm)
Ru dispersiona
Ru dispersionb
TOFd (10ꢁ3 sꢁ1
)
cat
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
2.4
5.1
6.8
8.7
12
0.52
0.24
0.17
0.15
0.11
0.55
0.26
0.19
0.15
0.11
98.4
45.6
30.4
12.0
6.4
532
533
500
222
163
a
b
c
Measured from a H2AO2 titration technique [25].
Calculated by the following equation: dispersion = 1.32/D (nm).
Calculated using the data in Fig. 12A.
Evaluated on the basis of the rate of cellobiose conversion per surface Ru atom by using Ru dispersion measured from H2AO2 titration technique.
d
3.3.3. Rate of 3-b-
D
-glucopyranosyl-
D-glucitol conversion over Ru/CNT
catalysts
Similar studies were performed using 3-b-
D
-glucopyranosyl-
D-
glucitol as a reactant to gain information about Step II in Fig. 11.
Because the hydrolysis over the Ru/CNT catalysts proceeds not as
quickly as the hydrogenolysis step (Step I in Fig. 11), relatively
strict reaction conditions have been employed for the conversion
of 3-b-
ferent Ru/CNT catalysts, the conversions of 3-b-
-glucitol all increased almost proportionally to the reaction time
in 120 min, and sorbitol was the main product (Fig. 13B). The rates
of 3-b- -glucopyranosyl- -glucitol conversions calculated from
D
-glucopyranosyl-
D-glucitol. As shown in Fig. 13A, over dif-
D-glucopyranosyl-
D
D
D
Fig. 13A are summarized in Table 6. The combination of the results
in Table 6 and Fig. 7 or Table 1 suggests that the acidity of the cat-
alysts play an important role in the transformation of 3-b-
D-gluco-
pyranosyl- -glucitol to sorbitol and glucose (Step II in Fig. 11).
D
Almost no glucose was observed during these experiments because
glucose could be subsequently converted to sorbitol over Ru parti-
cles very rapidly (Step III). Thus, the higher acidity of the catalyst
results in a higher rate of 3-b-
sion via the cleavage of the b-1,4-glycosidic bond.
The distribution of the products (Fig. 13B) showed that the
selectivity of sorbitol produced from the conversion of 3-b- -glu-
copyranosyl- -glucitol varied largely over different Ru/CNT cata-
D-glucopyranosyl-D-glucitol conver-
D
D
lysts. The catalysts with higher acidity and larger size of Ru
particle (Ru/CNT-C623-H623 with a mean Ru size of 8.7 nm and
Ru/CNT-C623-H773 with a mean Ru size of 12 nm) afforded higher
sorbitol selectivity. We speculate that the higher acidity which
could provide higher rate of 3-b-D-glucopyranosyl-D-glucitol con-
versions (Table 6) and the lower rate of large Ru particles in sorbi-
tol degradation both contribute to the higher sorbitol selectivity
over these catalysts. To further confirm this speculation, we have
carried out conversions of sorbitol over different Ru/CNT catalysts.
Fig. 13. Catalytic behaviors for the conversion of 3-b-
over the Ru/CNT catalysts with different mean sizes of Ru particles. (A) Conversion
and (B) selectivity. Reaction conditions: 3-b- -glucopyranosyl- -glucitol,
0.174 mmol; catalyst, 0.050 g; H2O, 20 cm3; H2, 5 MPa; temperature, 458 K.
D-glucopyranosyl-D-glucitol
D
D

W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
31
Table 6
copyranosyl-D-glucitol further increased over the Ru/CNT catalysts.
Rate for 3-b-D-glucopyranosyl-D-glucitol conversions over the Ru/CNT catalysts with
different mean sizes of Ru particles.
This is in consistent with the experimental result that the Ru/CNT
catalysts exhibit higher acidity than CNTs (Table 1). However, no
glucose was obtained over the Ru/CNT, further confirming that glu-
cose could be transformed to sorbitol very rapidly over the Ru
nanoparticles. Only the catalysts with larger-sized Ru particles
(7.8 and 12 nm) could provide higher sorbitol yield (Table 8).
The results described earlier suggest that the catalysts with
smaller Ru particles favor the hydrogenolysis of cellobiose to form
Catalyst
Mean size of
Ru (nm)
3-b-D-Glucopyranosyl-D-glucitol
conversion ratea ðmmol gꢁ1 hꢁ1
Þ
cat
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
2.4
5.1
6.8
8.7
12
71.3
90.5
99.2
139
127
3-b-D-glucopyranosyl-D-glucitol (Step I in Fig. 11). Moreover, these
a
Calculated using the data in Fig. 13A.
catalysts are detrimental to keeping sorbitol from degradation
(Steps IV and V in Fig. 11). Therefore, it is understandable that
the catalysts with smaller mean sizes of Ru particles (<8.7 nm)
can achieve higher yields of 3-b-D-glucopyranosyl-D-glucitol in a
Table 7
Conversion of sorbitol over the Ru/CNT catalysts with different mean sizes of Ru
short reaction time but are less efficient for sorbitol formation
(Fig. 9). On the other hand, over the catalysts with larger Ru parti-
cles (8.7 and 12 nm), the higher acidity of these catalysts is bene-
ficial for forming sorbitol through the cleavage of the b-1,4-
particles.a
Catalyst
Mean size of
Ru (nm)
Reaction time (h)
Conversionb (%)
glycosidic bond of 3-b-D-glucopyranosyl-D-glucitol. The lower deg-
Ru/CNT-H773
2.4
5.1
8.7
2.4
5.1
8.7
1
1
1
3
3
3
25
12
4
45
23
4
Ru/CNT-EG483
Ru/CNT-C623-H623
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-C623-H623
radation rate of sorbitol over the larger Ru particles is also an
important factor for obtaining the high yield of sorbitol over these
catalysts. In short, our results suggest that, for an efficient conver-
sion of cellobiose to sorbitol, we should consider not only the size
of Ru particles, which may play key roles in the hydrogenolysis
step and degradation of the target product, but also the acidic
properties of the catalysts, which are crucial for the conversion
a
Reaction conditions: sorbitol, 1.0 mmol; catalyst, 0.050 g; H2, 5 MPa; H2O,
20 cm3; temperature, 458 K.
b
The products include mannitol, degradation compounds, and some unknown
of 3-b-D-glucopyranosyl-D-glucitol. We hope that these insights
products.
may be helpful for the design of efficient catalysts for the transfor-
mation of cellulose. Detailed mechanistic studies on cellulose con-
versions are underway.
The results shown in Table 7 revealed that the conversion of sorbi-
tol proceeded significantly faster over the catalysts with smaller
sizes of Ru particles, forming mannitol and other degradation prod-
ucts. In other words, sorbitol was more stable over the catalyst
with a larger mean size of Ru particles (e.g., 8.7 nm).
4. Conclusions
In a previous communication [13], protons generated by the
autoprotolysis of water were proposed to participate in the hydro-
lysis of b-1,4-glycosidic bond in hot water. Thus, we have also
The Ru/CNT can efficiently catalyze the direct conversion of cel-
lobiose into sorbitol in the presence of hydrogen in neutral water
medium. A sorbitol yield of 87% has been attained at 458 K. The
mean size of Ru nanoparticles and the acidity of the catalysts are
key factors dominating the catalytic performances. The CNTs pre-
treated by concentrated nitric acid possess higher concentrations
of acidic functional groups and are better catalyst support of Ru
catalyst for sorbitol formation. The catalyst with a larger mean size
of Ru nanoparticles (P8.7 nm) and higher acidity exhibits a better
sorbitol yield, while that with a smaller mean Ru size and lower
investigated the conversions of 3-b-D-glucopyranosyl-D-glucitol
without any catalyst in hot water and over the CNTs with different
post-treatments. These results are compared with those over the
Ru/CNT catalysts under the same reaction conditions in Table 8. In-
deed, 3-b-D-glucopyranosyl-D-glucitol could be converted to sorbi-
tol, glucose, and HMF with yields of 22%, 13%, and 1%, respectively,
in the blank reaction. The use of CNTs with different post-treat-
ments as catalysts further raised the conversion of 3-b-
anosyl- -glucitol and the yield of sorbitol and glucose, suggesting
that the acidic sites over the CNTs played roles in the conversion
D
-glucopyr-
acidity can afford a better yield of 3-b-D-glucopyranosyl-D-glucitol
(as high as 93% at the initial reaction stage). It is elucidated that the
reaction involves two key steps. In the first step, cellobiose is trans-
D
of 3-b-
D
-glucopyranosyl-
D
-glucitol. The conversions of 3-b-
D-glu-
formed into 3-b-D-glucopyranosyl-D-glucitol via the hydrogenoly-
Table 8
Catalytic performances of the CNTs after different post-pretreatments and the Ru/CNT catalysts with different mean sizes of Ru particles for the conversions of 3-b-D-
glucopyranosyl-D-glucitol.a
Sample
Mean size of Ru (nm)
Conversion (%)
Yield (%)
Sorbitol
Glucose
HMF
Blank
CNT-H773
CNT-EG483
CNT-EG453
CNT-C623-H623
CNT-C623-H773
Ru/CNT-H773
Ru/CNT-EG483
Ru/CNT-EG453
Ru/CNT-C623-H623
Ru/CNT-C623-H773
–
–
–
–
–
–
2.4
5.1
6.8
8.7
12
48
71
78
78
84
82
83
87
88
98
90
22
26
32
38
32
41
13
34
36
60
58
13
16
20
20
11
17
0
0
0
0
0
1
3
3.5
4
3
6
0
0
0
0
0
a
Reaction conditions: 3-b-D-glucopyranosyl-D
-glucitol, 0.174 mmol; catalyst, 0.050 g; H2O, 20 cm3; H2, 5 MPa; temperature, 458 K; time, 3 h.

32
W. Deng et al. / Journal of Catalysis 271 (2010) 22–32
[7] P.L. Dhepe, A. Fukuoka, ChemSusChem 1 (2008) 969.
[8] M. Stöcker, Angew. Chem. Int. Ed. 47 (2008) 9200.
[9] K. Fleming, D.G. Gray, S. Matthews, Chem. Eur. J. 7 (2001) 1831.
[10] Y. Nishiyama, P. Langan, H. Chanzy, J. Am. Chem. Soc. 124 (2002) 9074;
Y. Nishiyama, J. Sugiyama, H. Chanzy, P. Langan, J. Am. Chem. Soc. 125 (2003)
14300.
sis of CAO bond in one glucose ring, and then the hydrolysis of the
b-1,4 glycosidic bond in 3-b- -glucopyranosyl- -glucitol affords
sorbitol in the second step. The catalyst with a smaller mean size
of Ru particles favors the hydrogenolysis of cellobiose to 3-b- -glu-
copyranosyl- -glucitol. However, it is less active in the subsequent
hydrolysis of 3-b- -glucopyranosyl- -glucitol into sorbitol due to
the lower acidity and being detrimental to keeping sorbitol from
degradation. Both the size of Ru nanoparticles and the acidity of
catalyst should be considered for an efficient conversion of cellobi-
ose into sorbitol.
D
D
D
D
[11] B. Kamm, Angew. Chem. Int. Ed. 46 (2007) 5056.
D
D
[12] A. Fukuoka, P.L. Dhepe, Angew. Chem. Int. Ed. 45 (2006) 5161.
[13] C. Luo, S. Wang, H. Liu, Angew. Chem. Int. Ed. 46 (2007) 7636.
[14] N. Ji, T. Zhang, M. Zheng, A. Wang, H. Wang, X. Wang, J.G. Chen, Angew. Chem.
Int. Ed. 47 (2008) 8510.
[15] Y. Su, H.M. Brown, X. Huang, A.-D. Zhou, J.E. Amonette, Z.C. Zhang, Appl. Catal.
A 361 (2009) 117.
[16] W. Deng, X. Tan, W. Fang, Q. Zhang, Y. Wang, Catal. Lett. 133 (2009) 167.
[17] N. Yan, C. Zhao, C. Luo, P.J. Dyson, H. Liu, Y. Kou, J. Am. Chem. Soc. 128 (2006)
8714.
Acknowledgments
[18] J.A. Bootsma, B.H. Shanks, Appl. Catal. A 327 (2007) 44.
[19] P. Gallezot, N. Nicolaus, G. Flèche, P. Fuertes, A. Perrard, J. Catal. 180 (1998) 51.
[20] B.W. Hoffer, E. Crezee, P.R.M. Mooijman, A.D. Langeveld, F. Kapteijn, J.A.
Moulijn, Catal. Today 79 (2003) 35.
[21] J. Pan, J. Li, C. Wang, Z. Yang, React. Kinet. Catal. Lett. 90 (2007) 233.
[22] P. Chen, H.B. Zhang, G.D. Lin, Q. Hong, K.R. Tsai, Carbon 35 (1997) 1495.
[23] A.J. Plomp, D.S. Su, K.P. de Jong, J.H. Bitter, J. Phys. Chem. C 113 (2009) 9865.
[24] A. Miyazaki, I. Balint, K. Aika, Y. Nakano, J. Catal. 204 (2001) 364.
[25] K.C. Taylor, J. Catal. 38 (1975) 299.
This work was supported by the National Natural Science Foun-
dation of China (Nos. 20625310, 20873110 and 20923004), the Na-
tional Basic Program of China(2010CB732303 and 2005CB221408),
and the Key Scientific Project of Fujian Province (2009HZ0002-1).
We thank Profs. H.B. Zhang and G.D. Lin for providing CNT.
[26] P. Serp, M. Corrias, P. Kalck, Appl. Catal. A 253 (2003) 337.
[27] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, Handbook of X-ray
Photoelectron Spectroscopy, Physical Electronics, Inc., Eden Prairie, 1995.
[28] F. Li, J. Chen, Q. Zhang, Y. Wang, Green Chem. 10 (2008) 553.
[29] J. Kang, S. Zhang, Q. Zhang, Y. Wang, Angew. Chem. Int. Ed. 48 (2009) 2565.
[30] G.A. Somorjai, J.Y. Park, Top. Catal. 49 (2008) 126.
[31] U. Zielke, K.J. Hüttinger, W.P. Hoffman, Carbon 34 (1996) 983.
[32] P.V. Laksgminarayanan, H. Toghiani, C.U. Pittman Jr., Carbon 42 (2004) 2433.
[33] T.I.T. Okpalugo, P. Papakonstantinou, H. Murphy, J. McLaughlin, N.M.D. Brown,
Carbon 43 (2005) 153.
References
[1] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 107 (2007) 2411.
[2] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007) 7164.
[3] C.H. Christensen, J. Rass-Hansem, C.C. Marsden, E. Taarning, K. Egeblad,
ChemSusChem 1 (2008) 283.
[4] D.A. Simonetti, J.A. Dumesic, Catal. Rev. Sci. Eng. 51 (2009) 441.
[5] D. Klemm, B. Heublein, H.-P. Fink, A. Bohn, Angew. Chem. Int. Ed. 44 (2005)
3358.
[6] A.J. Ragauskas, C.K. Williams, B.H. Davison, G. Britovsek, J. Cairney, C.A. Eckert,
W.J. Frederick Jr., J.P. Hallett, D.J. Leak, C.L. Liotta, J.R. Mielenz, R. Murphy, R.
Templer, T. Tschaplinski, Science 311 (2006) 484.
[34] J. Álvarez-Rodríguez, A. Guerrero-Ruiz, I. Rodríguez-Ramos, A. Arcoya-Martín,
Catal. Today 107 (2005) 302.