Influence of the functional groups of multiwalled carbon nanotubes on performance of Ru catalysts in sorbitol hydrogenolysis to glycols

Article abstract of DOI:10.1016/j.molcata.2016.11.003

Different functional groups (i.e. [sbnd]NH₂, [sbnd]COOH, [sbnd]OH and nitrogen-doping) modified CNTs (denoted as AMCN, CMCN, HMCN and NMCN, respectively) supported ruthenium catalysts (Ru/AMCN, Ru/CMCN, Ru/HMCN and Ru/NMCN) were prepared by incipient wetness impregnation method. They were fully characterized by XRD, TG, Raman, XPS, TPD and TEM to elucidate the relationship between the physical property and their catalytic performance. TEM results shown that Ru particles were well dispersed on the surface for all the samples with the size of 1.48–1.99 nm. The effects of functional groups of carbon nanotubes (CNTs), nitrogen doping and base additive types on activity and selectivity of ethylene glycol (EG) and propylene glycol (1,2-PD) were investigated. In addition, the activity and final products distribution were much influenced by the properties of functional groups on CNTs and the type of metal cation of the base promoters, which probably participated in the reaction for accelerating a retro-aldol reaction for C[sbnd]C cleavage. Among the catalysts, Ru supported on AMCN exhibited the best catalytic activities and glycols selectivities than on MCN, CMCN, HMCN and NMCN.

Full text of DOI:10.1016/j.molcata.2016.11.003

Accepted Manuscript
Title: Influence of the functional groups of multiwalled carbon
nanotubes on performance of Ru catalysts in sorbitol
hydrogenolysis to glycols
Author: Xingcui Guo Huihuan Dong Bin Li Linlin Dong
Xindong Mu Xiufang Chen
PII:
S1381-1169(16)30459-9
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2016.11.003
MOLCAA 10100
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
27-7-2016
8-10-2016
2-11-2016
Please cite this article as: Xingcui Guo, Huihuan Dong, Bin Li, Linlin Dong, Xindong
Mu, Xiufang Chen, Influence of the functional groups of multiwalled carbon nanotubes
on performance of Ru catalysts in sorbitol hydrogenolysis to glycols, Journal of
Molecular Catalysis A: Chemical http://dx.doi.org/10.1016/j.molcata.2016.11.003
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Influence of the functional groups of multiwalled carbon nanotubes on
performance of Ru catalysts in sorbitol hydrogenolysis to glycols
a
a,b,1
a
a
a,
Xingcui Guo , Huihuan Dong , Bin Li , Linlin Dong , Xindong Mu *, Xiufang
Chena,*
a Key Laboratory of Bio-based Materials, Qingdao Institute of Bioenergy and
Bioprocess Technology, Chinese Academy of Sciences, Qingdao 266101, China
b School of Chemical and Environmental Engineering, Shanghai Institute of
Technology, Shanghai, 201418, China
*
Corresponding author. Fax: +86 53280662724; Tel: +86 53280662723
E-mail addresses: muxd@qibebt.ac.cn (X. Mu), chenxf@qibebt.ac.cn (X. Chen)

Graphical Abstract
Functional groups of CNTs acted as anchoring site for Ru and obstructed the sintering
of ruthenium clusters which therefore maintained high dispersion on these catalysts.
The nature of these groups played crucial role on the activity of sorbitol hydrogenolysis.
The presence of a solid base promoter or nitrogen doping facilitates dehydration of
polyols and subsequent deoxygenation while restraining the methanation reaction.
Highlights



Modification of carbon nanotubes supported Ru catalysts were prepared, nature
of acidic or basic sites of functional groups played crucial role.
With nitrogen doping to carbon nanotubes, the activity of Ru was improved in
sorbitol hydrogenolysis.
All the alkali hydroxide promoters significantly enhanced the sorbitol
conversion, while NaOH promoter gave the highest conversion and glycol
yields.
ABSTRACT Different functional groups (i.e. -NH , -COOH, -OH and nitrogen-doping)
2

modified CNTs (denoted as AMCN, CMCN, HMCN and NMCN, respectively)
supported ruthenium catalysts (Ru/AMCN, Ru/CMCN, Ru/HMCN and Ru/NMCN)
were prepared by incipient wetness impregnation method. They were fully
characterized by XRD, TG, Raman, XPS, TPD and TEM to elucidate the relationship
between the physical property and their catalytic performance. TEM results shown that
Ru particles were well dispersed on the surface for all the samples with the size of 1.48-
1
.99 nm. The effects of functional groups of carbon nanotubes (CNTs), nitrogen doping
and base additive types on activity and selectivity of ethylene glycol (EG) and
propylene glycol (1,2-PD) were investigated. In addition, the activity and final products
distribution were much influenced by the properties of functional groups on CNTs and
the type of metal cation of the base promoters, which probably participated in the
reaction for accelerating a retro-aldol reaction for C-C cleavage. Among the catalysts,
Ru supported on AMCN exhibited the best catalytic activities and glycols selectivities
than on MCN, CMCN, HMCN and NMCN.
Keywords: Hydrogenolysis; Sorbitol; Ruthenium catalyst; Glycols
1
. Introduction
Biomass as sustainable and renewable raw materials was considered to be one of
the most potential substitutes for the production of liquid renewable fuels and chemicals
to displace petroleum in the next few years [1-3]. Lignocellulosic biomass-derived
sugar alcohols were considered as versatile feedstocks for the production of useful
chemicals [4]. The hydrogenolysis of sorbitol to EG and 1,2-PD has aroused
considerable interest due to the wide applications of EG and 1,2-PD in the manufacture
of polymers, resins, functional fluids, perfumes, cosmetics, etc [5,6]. Furthermore,

producing EG and 1,2-PD from renewable raw materials can reduce human dependence
on fossil energy resources and contribute to sustainable development for the
environment and the economy.
Over the past decades, a lot of studies have been performed on the effects of
different metal catalysts (Ni, Pd, Pt, Rh and Ru) [7-10], supports and base promoters
on the structure and their catalytic activity in sorbitol hydrogenolysis to glycols. Among
the proposed catalyst systems, Ru was found to be the most active metal for polyols
hydrogenolysis [11]. Meanwhile, Carbon materials (carbon nanotubes (CNTs), carbon
nanofibers (CNFs), graphenes or graphite) were often used as supports for Ru based
catalysts. In this regard Sun et al. [12] studied the influence of activated carbon as
supports of ruthenium for xylitol hydrogenolysis and found that Ru/AC showed high
activities and glycol selectivities. Zhao et al. [13] prepared carbon nanofibers (CNFs)
and used the material to improve the selectivity of EG and 1,2-PD, however, the nano-
scale CNFs will induce severe difficulty in catalyst separation and product purification
for practical application. Gallegos-Suarez et al. [14] suggested that the electron-donor
properties of HSAG and CNTs could stable the electron-rich Ru nanoparticles and that,
in comparison with Ru/AC, favor the formation of 1,2-propanediol from glycerol
hydrogenolysis on metal sites.
Among them, Carbon nano-tubes exhibit electron donor character that can modify
the properties of the supported metal nanoparticles and they can be also functionalized
in order to change their surface chemical properties [15,16]. Studying and clarifying
the catalytic effect of a specific type of functional groups on carbon materials is very
important and can provide valuable guidance to the design and application of carbon-
based catalysts. Coloma et al. [17,18] studied the influence of oxygen containing
surface groups on the hydrogenation of crotonaldehyde or benzene with Pt/AC or Pt

and Ru-supported on graphite as catalysts, and it was found that the catalytic activities
have been greatly promoted by these groups. Recently, Gallegos-Suarez [19]
investigated the effect of the concentration of surface -COOH groups on the
hydrogenolysis of glycerol over Ru/C catalyst, and found that the presence of -COOH
groups increased the activity of the catalyst. But there are several different types groups
on the surface of activated carbon, including acidic carboxyl, anhydride, lactone, and
the neutral or weakly basic carbonyl, quinone, ether and pyrone. The concentration and
distribution of these different groups have some major impact on the adsorptive and
catalytic performance of AC. However, due to the complexity of activated carbon
surface chemistry, it is difficult to distinguish the role of some single functional group
on catalytic application.
The surface of carbon nanotubes can be selective chemical modification through
introducing some specific group. The role of different functional groups formed on
carbon nanotubes can be compared in detail. Various surface functional groups should
have the possibility of affecting the catalytic performances of the carbon nanotubes-
based catalysts for many reactions [20,21]. Nitrogen doping has also been studied as a
feasible strategy for tuning the physical and chemical properties of CNTs in a well-
defined way [22]. But, the influence of different functional groups or nitrogen-doping
CNTs on the properties of Ru/CNTs catalyst and the catalytic performance of Ru-based
catalyst for sorbitol hydrogenolysis to glycols have not yet been evaluated
systematically.
In the present paper, different carbon nanotubes (MCN, AMCN, CMCN, HMCN
and NMCN) supported ruthenium catalysts (Ru/MCN, Ru/AMCN, Ru/CMCN,
Ru/HMCN and Ru/NMCN) were prepared by incipient wetness impregnation method.
Then, the effect of functional groups or nirogen-doping on the structure, metal

dispersion, and catalytic performance in sorbitol hydrogenolysis to glycols was
investigated carefully.
2
. Experimental
2
.1. Materials
D-sorbitol, EG, 1,2-PD, LiOH, NaOH and KOH were purchased from Aladdin
Industrial Inc. Multi-walled carbon nanotubes (length: 0.5-2 μm, ID: 5-10 nm, OD: 10-
0 nm), including MCN, HMCN (the content of hydroxyl group is 3.7 wt%), AMCN
the content of amino group is 2.5 wt%), CMCN (the content of carboxyl group is 2.6
2
(
wt%) and NMCN (the content of nitrogen is 2.8 wt%) were obtained from Chengdu
Organic Chemicals Co., LTD and used as received, Ruthenium chloride hydrate
(RuCl
3
·xH O) (Sinopharm Chemical Reagent Co., Ltd, Shanghai, China) was used as
2
the precursor of ruthenium.
2
.2. Catalysts preparation
To investigate the role of CNTs functionalization, various functionalized CNTs
(
AMCN, CMCN, HMCN and NMCN) supported Ru catalysts were prepared by an
impregnation method. Standardly, the modified CNTs were added into the aqueous
solution of RuCl ·xH O and dispersed with stirring for 5 h. After being further
impregnation for 12 h at ambient temperature and then the suspension was evaporated
3
2
o
o
at 70 C to remove water and dried at 110 C overnight. Reduction of the catalysts were
o
performed in pure H
2
at a flow rate of 20 mL/min at 250 C for 2 h. The obtained
catalysts are generally labeled as Ru/MCN, Ru/AMCN, Ru/CMCN, Ru/HMCN and
Ru/NMCN, respectively. The loading of Ru was 4.0 wt.%.
2
.3. Catalyst characterization
Powder X-ray diffraction (XRD) patterns of catalyst samples were recorded on a
D8 Advance X-ray diffractometer (Bruker, Germany) operated with Cu K irradiation

o
o
and 2θ ranged from 10 to 80 . Fourier transform infrared spectroscopy (FT-IR) was
-1
taken on a Nicolet 380 FT-IR spectrometer with a resolution of 2 cm . The High
Resolution Transmission Electron Microscopy (HRTEM) images were performed on a
JEOL JEM-2100F operated at 200 KV. For Raman spectroscopy tests, the 532-nm line
from a Kimmon IK3201 R-F He-Cd laser was used for excitation with an intensity of
~20 Mw measured at the source. X-ray photoelectron spectroscopy (XPS) analyzes
were carried out using a PHI-560 ESCA (Perkin Elmer) spectrometer equipped with an
Mg Kα target. A Thermo IRIS Intrepid II XSP atomic emission spectrometer was
applied to determine the chemical composition of catalysts and metal leaching after
reaction. N
2
adsorption–desorption experiments were performed with a Micromeritics
-TPD and CO -TPD experiments
ASAP 2020 surface area and porosity analyzer. NH
3
2
were performed with a Micro-meritics Autochem II chemisorption analyzer.
Thermogravimetric analyses were performed on a DuPont TGA-2950TA
thermogravimetric analyzer. Samples were analyzed in platinum pans at a heating rate
o
o
of 5 C/min from room temperature to 1000 C in an atmosphere of air. ICP analyses
were carried out on a Thermo IRIS Intrepid II XSP atomic emission spectrometer to
determine the chemical composition of catalysts and to examine metal leaching during
reactions.
2
.4. Catalytic hydrogenolysis reaction
The hydrogenolysis of sorbitol was performed in a 50 mL stainless-steel autoclave
with magnetic stirring. After sorbitol aqueous solution, Ru catalyst and an appropriate
amount of base were charged into the reactor, the autoclave was purged with hydrogen
four times and then pressurized to the desired pressure at room temperature. Then the
reaction was performed at certain temperature under the stirring speed of 800 r/min.
After the reaction, the reactor was cooled down and the used catalyst was separated

form the reaction mixture by centrifugation. The samples were filtered through 0.22
μm-pore-size filters (Membrana) prior to analysis. The obtained products such as 1,2-
PG, EG and GLY were determined using a gas chromatography (GC, 7890A, Agilent,
USA) equipped with a CP-Wax 58 (FFAP) capillary column (0.25 mm × 25 m) and a
flame ionization detector. Other products like glucose, sugar alcohols were quantified
by Anion-Exchange Chromatography (IC, Dionex ICS-3000) equipped with pulsed
amperometric detector and an Aminex HPX-87H column (Bio-Rad, 7.8 × 300 mm),
-1
o
using 500 mM NaOH as eluent with a flow rate of 0.4 mL min at 30 C. The obtained
products in resultant solutions were also identified by GC-MS (6890N, Agilent, USA).
The conversion of sorbitol and yields of products were calculated on the carbon basis
and defined as follows:
moles of sorbitol (in) − moles of sorbitol (out)
Conversion (%) =
× 100
moles of sorbitol (in)
(
moles of the product) × (number of carbon atoms in the product)
Yield (%) =
× 100
(
moles of sorbitol charged) × 3
3
3
3
. Results and discussion
.1. Characterization of the catalysts
.1.1. Structural properties of the catalysts
The Raman patterns of all the supports with different functional groups are
examined to explore the influence of functionalization on the extent of disorder or
defects on these CNTs. As illustrated in Fig. 1, all the five MWCNTs exhibit two
−
1
−1
characteristic peaks at around 1340 and 1575 cm . The peak at near 1575 cm (G-
band) represents the tangential mode vibrations of carbon atoms in graphene sheets,
-1
while the D peak (around 1342 cm ) is corresponding to the vibrations of carbon atoms
in the disordered graphite structure. The ratio (I /I ) of the integral value of D band
intensity (I ) to the integral value of the G band intensity (I ) of raman spectrum can
D
G
D
G

be used to measure the disorder degree or the concentration of functionalized groups
and defects in CNTs samples [23]. There is an obvious increase of I /I ratio for
samples HMCN (1.21), CMCN (1.28) or AMCN (1.32) compared to the pristine MCN
0.91) (Fig. 1). This indicates a great increase in defects along the nanotubes’body after
D
G
(
functionalization. Because Ru nanoparticles prefer anchoring onto the defect structures
rather than the perfect sites of CNTs [24]. So, with the increasing of defects sites,
smaller Ru particles in the range of 1~3 nm can more effectively be attached or loaded
on the carbon support surface and agglomeration of catalyst particles is inhibited.
However, a different behaviour of the band intensity at “D band” and “G band” can be
observed for NMCN. The defects are induced by uniformly doped nitrogen in CNTs,
which provide more nucleation sites, and thus promoting a higher dispersion of
discharge products. On the other hand, it can be seen that the nitrogen-doping has some
influence on the shift of “D band” and “G band” for all the given samples. The D band
is upshifted with nitrogen doping which attributes to the strong positive inductive effect
of nitrogen atoms, which indicates that new types of disorder is appeared for NMCN.
The shifts of the G-band have been ascribed to C–C expansion and the changes of
electronic structure [25]. The lower intensity peak of the NMCN sample as compared
to the other samples demonstrates a low content of organic molecules covalently
attached [26].
XRD patterns of different Ru/CNTs catalysts with the same Ru loadings are
checked and displayed in Fig. 2. The diffraction patterns of all Ru/CNTs samples show
a strong reflection and a weak reflection peaks at 2θ =25.9 º and 43.2 º, ascribed to the
(002) and (100) planes of graphite, respectively [27]. The ICP-OES results show that
the ruthenium contents for Ru/MCN, Ru/AMCN, Ru/CMCN, Ru/HMCN and
Ru/NMCN are 4.1, 4.2, 4.0, 4.2 and 3.8, respectively. Besides that, no diffraction

signals related to the Ru species are observed from XRD, confirming that ruthenium
has a good deposition on the CNTs and the size of Ru crystallites was lower the
limitation of XRD decection.
The nitrogen physisorption isotherms of CNTs and their corresponding Ru
catalysts are shown in Fig. 3A and B, respectively. All the CNTs show type IV
adsorption isotherms with a type III hysteresis loop at higher relative pressures,
according to IUPAC-classification [28], which is typical of tubular capillaries in a
mesoporous material.
The catalysts exhibit similar isotherms to those of their supports, indicating that
the pore structure are maintained after Ru loading. The amount of N adsorbed on these
2
CNTs at relative pressures less than 0.1 is very small, suggesting that micropores are
insignificant in the samples [29]. As shown in Fig. 3B, The catalysts exhibit similar
isotherms to their corresponding supports, indicating that the pore structure of the CNTs
is not apparently changed after Ru loading. Indeed, nitrogen adsorption isotherms
results demonstrate that mesopores are accounted for the majority of the total pores in
the catalysts.
The morphologies and structures of the different Ru/CNTs catalysts were
determined by TEM and displayed in Fig. 4. It can be seen that all the CNTs have a
diameter of about 10–20 nm. The ruthenium particles are distributed uniformly over the
surface of different CNTs and the particle size of ruthenium species are in the range of
0
.5-3 nm as shown in Fig. 4, consistent well with the XRD results.
.1.2. Surface acidity/basicity of catalysts
3
TPD analysis is carried out to elucidate surface acid-base properties of various
Ru/CNTs catalysts. All catalysts NH
3
-TPD profiles are displayed in Fig. 5A. The main
o
ammonia TPD peak appeared at about 220~299 C for the Ru/MCN and Ru/AMCN

catalysts, and the peak intensity is lower than that of the other three catalysts, indicating
that most of the ammonia molecules are weakly chemisorbed. Moreover, for Ru/HMCN,
in addition to the lower-temperature peak, another ammonia TPD peak at higher
o
temperatures (>330 C) can be observed, indicating ammonia chemisorbs strongly, and
the peak associated with the stronger acid sites further shifts to higher temperatures
o
(
~336 C) over Ru/CMCN sample. From these NH
3
-TPD results, it is clear that there
exists differences in the acidity among these catalysts with different functional groups.
The Ru/AMCN catalyst possesses much lower acidity than those for the other Ru/CNTs
catalysts [30]. The acid properties of the parent carbon nanotubes (MCN, AMCN,
CMCN, HMCN and NMCN) were also investigated by NH
3
-TPD, and the results are
shown in Fig. S1. It can be seen the NH -TPD profiles of the various CNTs (Fig. S1)
3
are remarkably different from each other. The desorption peaks for AMCN and NMCN
during 200-400 °C are obviously lower than the other three samples. The lower total
acid sites of AMCN and NMCN is mainly due to incorporation of basic amino groups
or nitrogen, respectively. So, the increase of low-temperature peaks for the series of
ruthenium catalysts can be attributed to residual chloride from precursors RuCl
desorption of NH from Ru sites absorption.
Similarly, basicity measurements through CO
3
or the
3
2
-TPD shows that carbon dioxide
also reacts differently with various catalyst samples (Fig. 5B). Most of these TPD
o
patterns show two desorption peaks at about 230~320 C that differ from a sample to
another by their positions, heights and widths, indicating changes not only in the total
basicity, but also in the distribution of the base strength. It seems that Ru/AMCN
o
catalyst possess much base sites at 250~320 C introduced by amino groups. More
recently, Chia et al. showed that the acidity and basicity on catalysts was responsible
for the enhanced C–O hydrogenolysis [31]. So, the number of acid and base sites arises

from the functional group of CNTs will be correlated with the catalyst activity and
product distribution below [32].
3
.1.3. Surface chemical states
Due to the inherently surface sensitivity of XPS technical, XPS analysis can also
be used to discern the chemical modification on the surface of various functionalized
CNTs. As shown in Fig. 6a, C1s peaks are deconvoluted into several bands based on
2
these literatures: the main peak at 284.5 eV is assigned to sp -hybridized carbon atoms,
peaks with higher binding energies at 286.2 eV, 287.5 eV, and 288.9 eV are considered
as C-O (phenols and ethers), C=O (ketones and quinones), -COO (carboxyls and esters)
species, respectively [33], which are in agreement with FT-IR spectrum of different
functionalized CNTs (Fig. S2, Supporting information). The π–π* transition loss peak
at 291.0 eV corresponds to the characteristic shake-up line of carbon in aromatic
compounds.
The chemical nature of the oxygenated groups on the surface of CNTs was also
investigated by XPS measurements. The XPS spectra of different CNTs showing the
deconvolution of the O1s peak are given in Fig. 6b. The fraction of each surface oxygen
groups was obtained by deconvolution of the O 1s peak in three components, namely,
carbonyl oxygen (peak at 531.4 eV), oxygen atoms in C-O bonds, like hydroxyl and
ether groups (peak at 532.6 eV) and oxygen atoms in acidic carboxyl groups (peak at
5
34.6 eV) [34]. The O/C atomic ratio for the sample MCN (0.0187) is lower than that
of the CMCN (0.127) support, but very similar to that found for NMCN (0.0178). In
comparison with the other supports the sample CMCN exhibits the highest population
of oxygen at the surface, most of them being carboxylic acid species (COOH). All these
observations are in agreement with the TPD-MS measurements above discussed.
In addition, there are N 1s signals for Ru/AMCN and Ru/NMCN catalysts. The

peak can be deconvoluted in two components: pyridinic nitrogen (399.5–398.5 eV) and
pyrrolic nitrogen (400.8–399.8 eV) [35] as shown in Fig. 6c. The two components of
the N 1s spectra, prove the presence of the nitrogen atoms within the graphite plane
bonded to the carbon atoms, forming part of the carbon nanotubes structure.
3
.2. Catalytic performance of the catalysts
3
.2.1. Effects of functional groups of carbon nanotubes on sorbitol hydrogenolysis
o
Table 1 shows the activities and selectivities of sorbitol hydrogenolysis at 210 C
and 6.0 MPa H over different supported ruthenium catalysts with Ru loadings of ca. 4
2
wt%. The supports include MCN, AMCN, CMCN and HMCN with different surface
functional groups. On the four Ru catalysts, ethylene glycol, 1,2-propylene glycol and
glycerol are formed as the major products in the presence of NaOH. 1,3-propylene
glycol is also detected with a significant amount. Other products mainly include ethanol,
1
,2-butanediol, CH
4
, CO and some unknown compounds. As shown in Table 1,
2
Ru/AMCN is more active and selective for the formation of 1,2-PD and EG compared
o
with the other catalysts. After 0.5 h at 210 C and 6 MPa H
2
, Ru/AMCN provided a
2
7.3% sorbitol conversion and a 58.9% combined selectivities to 1,2-PD and EG, the
highest selectivity to the two target glycols among the four catalysts examined. The
glycols selectivities over Ru/CMCN catalyst also stayed at a high level (54.2%) at
similar sorbitol conversions (∼30%). While the combined selectivities to 1,2-PD and
EG over Ru/MCN and Ru/HMCN catalysts decreased to 51.0% and 43.9 %,
respectively. The catalytic activities over Ru-supported catalysts can be ranked as
Ru/AMCN > Ru/CMCN > Ru/MCN > Ru/HMCN. Due to Ru particles are evenly
dispersed on the surface of carbon nanotubes for all the catalysts as shown in Fig. 4 and
the characteristic diffraction peaks of Ru species are not found by XRD (Fig. 2). So the
size of Ru particle is not a determining factor for the activity difference in this study.

As shown in Table 1, both acidic and alkali functional groups (carboxyl or amino groups)
of the surface of CNTs as binding sites for ruthenium particles have significant effect
on enhancing the selectivities to 1,2-PD and EG. In order to convert ruthenium to the
active species, Ru2+ cations must be reduced up to Ru . Amino groups possessing
0
0
electron-donor properties [36] could stabilize high-dispersed Ru , and prevent its re-
oxidation, mostly likely, due to possible partial electron density transfer from nitrogen
atoms of amine groups on modified CNTs to ruthenium. In contrast, carboxylic groups
0
2+
having electron-acceptor character favor re-oxidation of small Ru particles as Ru . So
Ru catalysts supported on AMCN show higher activity than Ru/CMCN. The
Ru/HMCN sample shows much lower activity than that of Ru/AMCN and Ru/MCN.
This dues to the presence of intermolecular interactions between the hydroxyl groups
of sorbitol and CNTs, such as hydrogen bond and orientation force, etc., which hinders
the sorbitol molecular adsorption on Ru metal particles [37].
3
.2.2. Effects of nitrogen doping on the activity of Ru/MCN in sorbitol hydrogenolysis
The catalytic performance of Ru/NMCN catalyst is tested in a stainless-steel
o
autoclave at 210 C and 6.0 MPa H
2
, the result is summarized in Table 1 (entry 6). For
Ru/NMCN catalyst, the two glycols with a combined selectivity of 57.2% at 26.8%
sorbitol conversion can be obtained, which are higher than Ru/MCN catalyst under the
same conditions. So after nitrogen is doped into CNTs support, the catalytic activity and
glycols selectivity are improved greatly. As shown in Fig. 4e, Ru particles are
distributed uniformly for Ru/NMCN with the size about 1.99 nm, almost as large as
that in Ru/MCN (1.48 nm). Based on these, one can conclude that the intrinsic
properties of the supports are important factors affecting the activity and selectivity of
Ru catalysts for sorbitol hydrogenolysis in water.
XRD (Fig. 2) and XPS results (Fig. 6) show that the structure of Ru/NMCN

changes slightly after nitrogen doping. Nevertheless, no obvious difference is observed
for surface texture of CNTs due to low nitrogen content. The much higher I /I ratio
1.21) of NMCN (Fig. 1) indicates that it contains more defects than MCN, which are
D
G
(
induced by uniformly doped nitrogen in CNTs and can provide more nucleation sites.
On the other hand, it has been reported the basicity and the electronic density of NMCN
could be increased after the doping of nitrogen, which could promote electron transfer
from CNTs to Ru [17] and improve the reaction activity of Ru/NMCN catalyst in
sorbitol hydrogenolysis.
3
.2.3. Effect of base type on sorbitol hydrogenolysis
As reported before, that the activity of sorbitol hydrogenolysis depended strongly
on the basicity of reaction solution [38]. The relations between the type alkali and
sorbitol hydrogenolysis activity are investigated by using various alkali hydroxide at
-
the same OH groups content and the results are shown in Fig. 7. It can be seen clearly
that alkaline promoters have a strong influence on product distribution and all the bases
can enhance sorbitol hydrogenolysis activity of Ru/AMCN catalyst. The conversion of
sorbitol on Ru/AMCN catalyst with these promoters decreases in the order of LiOH >
NaOH > KOH, which is reversed to the strength of the basicity. The two investigated
bases LiOH and NaOH outperform the base KOH in terms of sorbitol conversion and
product selectivity. Notably, sorbitol conversion reaches 97.3% and the combined
o
yields to 1,2-PD and EG is 61.4% under 210 C for 2 h when with NaOH as promoter.
However, only about 84.9% sorbitol conversion is achieved and the 1,2-PD and EG
yields decrease to 29.8% and 21.7% when KOH is added under the same reaction
conditions.
The reaction network for the hydrogenolysis of sorbitol involves complex
pathways due to the multiple C–C and C–O bond cleavage reactions that can occur on

metal catalysts or the base promoters at elevated temperatures. The conversion of
sorbitol to glycols initiates with the base promoted dehydrogenation and C=O bond
formation. Then, the unsaturated intermediate undergoes C–C cleavage via retro-
aldolization to form some small molecules, which undergo further hydrogenation or
hydrodeoxygenated to1,2-PD, EG and glycerol as the major products in the presence of
a base [38]. This indicates that the role of NaOH promoter is to provide a favorable
environment for the activation of the initial polyols in retro-aldol condensation but also
in the resulting intermediates.
+
+
It is found that in comparison with K , Na display a better promotional effects on
sorbitol conversion as well as glycols selectivity for Ru/AMCN catalyst. This is may
+
be attribute to some of the active sites of Ru are covered by K ions due to larger ionic
+
diameter of K [39], or it can be due to deposing carbon on the surface of Ru-K, which
leads to the formation of inactive carbon covering the active sites.
3
.2.4. Catalytic recyclability of the Ru/AMCN in the hydrogenolysis of sorbitol
The stability of Ru/AMCN is studied by thermogravimetric technique before
reaction, the TG profiles of the samples in flowing air are shown in Fig. S3. There is no
o
weight loss of Ru/AMCN catalyst at temperatures below 450 C, suggesting that
o
Ru/AMCN catalyst is stable for sorbitol hydrogenolysis at 210 C as used in the present
o
study. The 4.3% residual mass for the Ru/AMCN catalysts (above 650 C) corresponds
to the amount of ruthenium residue.
The thermal behavior of Ru/AMCN is measured by thermogravimetric technique
before reaction, the TG profiles of the sample AMCN in flowing air are also compared
o
and shown in Fig. S3. AMCN starts to oxidize at around 200 C, which is much lower
than the initial oxidation temperature of Ru/AMCN in the presence of Ru. This
indicated that the initial weight loss temperature of AMCN was improved after

ruthenium nanoparticles loaded on AMCN and Ru/AMCN exhibited high thermal
stability, which is attributed to the synergetic effect that occurred at the interface of Ru
o
and CNTs. There is no weight loss of Ru/AMCN catalyst at temperatures below 450 C,
o
suggesting that Ru/AMCN catalyst is stable for sorbitol hydrogenolysis at 210 C as
used in the present study. The 4.3% residual mass for the Ru/AMCN catalysts (above
o
6
50 C) corresponds to the amount of ruthenium residue.
As stability was critical for the practical application for sorbitol hydrogenolysis,
the reusability of Ru/AMCN catalyst was conducted. After completion of the reaction,
the catalyst was collected from the final solution and followed by washing with water
o
3
times, then dried at a temperature below 110 C before the next run. As shown in Fig.
S4, sorbitol conversion and glycols yields decreased to 82.9% and 40.3% after five
repeated runs. No leaching of Ru was detected in the reaction solutions by ICP-OES
analysis, which confirmed that Ru/AMCN catalyst was well tolerated during sorbitol
hydrogenolysis, in well agreement with the TG result. The used catalyst was further
characterized by XRD and TEM. It can be seen that (Fig. S5), no characteristic
diffraction peaks associated with Na
2
CO
3
structure were found clearly for the spent
CO produced in the reaction was removed
Ru/AMCN catalyst by XRD, indicating Na
2
3
completely by washing with water. The TEM (Fig. S6) results indicated that the size of
Ru species had some growth compared with the fresh catalyst and their average size
was 2.83 nm after reaction. The pore volume and surface area from N physisorption
2
data (Table S1) showed a drop of 15.6% and 16.3%, respectively for the Ru/AMCN
catalyst as a result of 3.0 wt% coke build-up on the catalyst surface after reaction, as
determined by TGA(-MS). These results demonstrated that the partial agglomeration
of Ru species caused by high pressure or the liquid phase nature of the reaction and the
deposition of organic by products on the surface of Ru surface might be the main reason

of catalyst deactivation.
. Conclusion
In summary, the effects of different functionalized of CNTs on the catalytic
4
performance in sorbitol hydrogenolysis is investigated. Results indicate that these
functional groups act as anchoring site for Ru and obstruct the sintering of ruthenium
clusters. So Ru nanoparticles for all the samples maintain high dispersion on these
CNTs. It is found that acidity/basicity of functional groups significantly improve the
activity of Ru catalysts during hydrogenolysis of sorbitol. NaOH, as solid base promoter,
o
are found to favor hydrodeoxygenation reactions at a temperature range of 210 C in
the presence of hydrogen. The presence of a solid base promoter facilitates dehydration
of polyols and subsequent deoxygenation while restraining the methanation reaction.
The reaction liquid is almost a clear solution after filtration of the catalyst and that no
leaching of Ru is detected by inductive coupled plasma (ICP) analysis. This shows that
the Ru/AMCN catalyst is highly active and can be recycled to use with regeneration in
the hydrogenolysis of sorbitol.
Acknowledgements
The authors would like to acknowledge the financial support from the National Natural
Science Foundation of China (21201174, 21303238, 21406251, 21403265, 21306216,
and 31470609), and the Taishan Scholars Climbing Program of Shandong (No. tspd
2
0150210).

References
[
[
[
1] S. Zhu, X. Gao, Y. Zhu, Y. Zhu, H. Zheng, Y. Li, Promoting effect of boron oxide
on Cu/SiO , catalyst for glycerol hydrogenolysis to 1,2-propanediol, J. Catal. 303
2013) 70-79.
2
(
2] A.M. Ruppert, W. Kamil, P. Regina, Cheminform abstract: hydrogenolysis goes
bio: from carbohydrates and sugar alcohols to platform chemicals, Angew. Chem.
Int. Ed. 43 (2012) 2564-2601.
3] N.A. Khan, D.K. Mishra, I. Ahmed, W.Y. Ji, J.S. Hwang, S.H. Jhung, Liquid-phase
dehydration of sorbitol to isosorbide using sulfated zirconia as a solid acid catalyst,
Appl. Catal. A: Gen. 93 (2013) 34-38.
[
[
[
4] M. Besson, P. Gallezot, C. Pinel, Conversion of biomass into chemicals over metal
catalysts, Chem. Rev. 114 (2013) 1827-1870.
5] I.M. Leo, M.L. Granados, J.L.G. Fierro, R. Mariscal, Sorbitol hydrogenolysis to
glycols by supported ruthenium catalysts, Chin. J. Catal. 35 (2014) 614-621.
6] H. Liu, Z. Huang, C. Xia, Y. Jia, J. Chen, H. Liu, Selective hydrogenolysis of
xylitol to ethylene glycol and propylene glycol over silica dispersed copper
catalysts prepared by a precipitation-gel method, ChemCatChem 6 (2014) 2918-
2
928.
[
[
7] X. Chen, X. Wang, S. Yao, X. Mu, Hydrogenolysis of biomass-derived sorbitol to
glycols and glycerol over Ni-MgO catalysts, Catal. Commun. 39 (2013) 86-89.
8] L. Ye, X. Duan, H. Lin, Y. Yuan, Improved performance of magnetically
recoverable Ce-promoted Ni/Al
2
3
O catalysts for aqueous-phase hydrogenolysis of
sorbitol to glycols, Catal. Today 183 (2012) 65-71.

[
[
[
9] M. Banu, P. Venuvanalingam, R. Shanmugam, B. Viswanathan, S. Sivasanker,
Sorbitol hydrogenolysis over Ni, Pt and Ru supported on NaY, Top Catal. 55
(2012) 897-907.
10] K.L. Deutsch, D.G. Lahr, B.H. Shanks, Probing the ruthenium-catalyzed higher
polyol hydrogenolysis reaction through the use of stereoisomers. Green Chem.
1
4 (2012) 1635-1642.
11] F. Mauriello, H. Ariga, M.G. Musolino, R. Pietropaolo, S. Takakusagi, K.
Asakura, Exploring the catalytic properties of supported palladium catalysts in
the transfer hydrogenolysis of glycerol, Appl. Catal. B: Environ. 166-167 (2015)
1
21-131.
[
[
[
12] J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene
glycol and propylene glycol on supported Ru catalysts, Green Chem. 13 (2011)
135-142.
13] L. Zhao, J.H. Zhou, Z.J. Sui, X.G. Zhou, Hydrogenolysis of sorbitol to glycols
over carbon nanofiber supported ruthenium catalyst, Chem. Eng. Sci. 65 (2010)
30-35.
14] E. Gallegos-Suarez, A. Guerrero-Ruiz, I. Rodriguez-Ramos, A. Arcoya,
Comparative study of the hydrogenolysis of glycerol over ru-based catalysts
supported on activated carbon, graphite, carbon nanotubes and KL-zeolite,
Chem. Eng. J. 262 (2015) 326-333.
[
[
15] J. Álvarez-Rodríguez, I. Rodríguez-Ramos, A. Guerrero-Ruiz, E. Gallegos-
Suarez, A. Arcoya, Influence of the nature of support on ru-supported catalysts
for selective hydrogenation of citral. Chem. Eng. J. 204-206 (2012) 169-178.
16] A.B. Dongil, B. Bachiller-Baeza, A. Guerrero-Ruiz, I. Rodríguez-Ramos, A.
Martínez-Alonso, J.M.D. Tascón, Surface chemical modifications induced on

high surface area graphite and carbon nanofibers using different oxidation and
functionalization treatments, J. Colloid Interface Sci. 355 (2011) 179-189.
17] F. Coloma, J. Narciso-Romero, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso,
Gas phase hydrogenation of crotonaldehyde over platinum supported on
oxidized carbon black, Carbon 36 (1998) 1011-1019.
[
[
18] F. Coloma, A. Sepúlveda-Escribano, J.L.G. Fierro, F. Rodríguez-Reinoso, Gas
phase hydrogenation of crotonaldehyde over Pt/activated carbon catalysts.
influence of the oxygen surface groups on the support, Appl. Catal. A: Gen. 150
(1997) 165-183.
[
[
19] E. Gallegos-Suarez, M. Pérez-Cadenas, A. Guerrero-Ruiz, I. Rodriguez-Ramos,
A. Arcoya, Effect of the functional groups of carbon on the surface and catalytic
properties of Ru/C catalysts for hydrogenolysis of glycerol, Appl. Surf. Sci. 287
(2013) 108-116.
20] L. Wang, J. Zhao, S. Bai, H. Zhao, Z. Zhu, Significant catalytic effects induced
by the electronic interactions between carboxyl and hydroxyl group modified
carbon nanotube supports and vanadium species for no reduction with NH
low temperature. Chem. Eng. J. 254 (2014) 399-409.
3
at
[
[
[
21] X. Wang, N. Li, J.A. Webb, L.D. Pfefferle, G.L. Haller, Effect of surface oxygen
containing groups on the catalytic activity of multi-walled carbon nanotube
supported pt catalyst. Appl. Catal. B: Environ. 101 (2010) 21-30.
22] J. Chen, H.Z. Zhong, S. Wang, Q. Ma, V. Rudolph, Q.L. Gao, Effects of nitrogen
doping on the structure of carbon nanotubes (CNTs) and activity of Ru/CNTs in
ammonia decomposition. Chem. Eng. J. 156 (2010) 404-410.
23] B.K. Dong, D.H. Lim, H.J. Chun, H.H. Kwon, H.I. Lee, Nitrogen-containing
graphitized carbon support for methanol oxidation Pt catalyst, Carbon 48 (2010)

673-679.
[
[
24] J.A. Rodríguez-Manzo, O. Cretu, F. Banhart, Trapping of metal atoms in
vacancies of carbon nanotubes and graphene, Acs Nano 4 (2010) 3422-3428.
25] R. Mi, H. Liu, H. Wang, K.W. Wong, J. Mei, Y. Chen, W.M. Lau, H. Yan, Effects
of nitrogen-doped carbon nanotubes on the discharge performance of li-air
batteries, Carbon 67 (2014) 744-752.
[
[
[
[
[
[
[
26] G. Gabriel, G. Sauthier, J. Fraxedas, M. Moreno-Mañas, M.T. Martínez, C.
Miravitlles, J. Casabó, Preparation and characterisation of single-walled carbon
nanotubes functionalised with amines, Carbon 44 (2006) 1891-1897.
27] B. Li, C. Wang, G. Yi, H. Lin, Y. Yuan, Enhanced performance of Ru
nanoparticles confined in carbon nanotubes for CO preferential oxidation in a
2
H -rich stream, Catal. Today 164 (2010) 74-79.
28] K.S.W. Sing, R.T. Williams, Physisorption hysteresis loops and the
characterization of nanoporous materials, Adsorpt. Sci. Technol. 22 (2004) 773-
7
82.
29] Z.T. Liu, C.X. Wang, Z.W. Liu, L. Jian, Selective hydrogenation of
cinnamaldehyde over pt-supported multi-walled carbon nanotubes: insights into
the tube-size effects. Appl. Catal. A Gen. 344 (2008) 114-123.
30] W. Deng, M. Liu, X. Tan, Q. Zhang, Y. Wang, Conversion of cellobiose into
sorbitol in neutral water medium over carbon nanotube-supported ruthenium
catalysts, J. Catal. 271 (2010) 22-32.
31] M. Chia, M.A. Haider, P.G. Rd, G.A. Kraus, M. Neurock, J.A. Dumesic,
Mechanistic insights into ring-opening and decarboxylation of 2-pyrones in
liquid water and tetrahydrofuran, J. Am. Chem. Soc. 135 (2013) 5699-5708.
32] F. Auneau, M. Berchu, G. Aubert, C. Pinel, M. Besson, D. Todaro, M. Bernardi,

T. Ponsetti, R.D. Felice, Exploring the reaction conditions for Ru/C catalyzed
selective hydrogenolysis of xylitol alkaline aqueous solutions to glycols in a
trickle-bed reactor, Catal. Today 234 (2014) 100-106.
[
[
[
[
[
33] L. Wang, L. Ge, T.E. Rufford, J. Chen, W. Zhou, Z. Zhu, V. Rudolph, A
comparison study of catalytic oxidation and acid oxidation to prepare carbon
nanotubes for filling with Ru nanoparticles, Carbon 49 (2011) 2022-2032.
34] W. Zheng, J. Zhang, B. Zhu, R. Blume, Y. Zhang, K. Schlichte, R. Schlögl, F.
Schüth, D. Su, Structure-function correlations for Ru/CNT in the catalytic
decomposition of ammonia, ChemSusChem 3 (2010) 226-230.
35] J. Chen, H.Z. Zhong, S. Wang, Q. Ma, V. Rudolph, Q.L. Gao, Effects of nitrogen
doping on the structure of carbon nanotubes (CNTs) and activity of Ru/CNTs in
ammonia decomposition, Chem. Eng. J. 156 (2010) 404-410.
36] Y. Chen, Z. Guo, T. Chen, Y. Yang, Surface-functionalized TUD-1 mesoporous
molecular sieve supported palladium for solvent-free aerobic oxidation of benzyl
alcohol. J. Catal. 275 (2010) 11-24.
37] A.B. Castle, E. Graciaespino, C. Nietodelgado, H. Terrones, M. Terrones, S.
Hussain, Hydroxyl-functionalized and N-doped multiwalled carbon nanotubes
decorated with silver nanoparticles preserve cellular function, Acs Nano 5 (2011)
2
458-2466.
38] X. Jin, J. Shen, W. Yan, M. Zhao, P.S. Thapa, B. Subramaniam, R.V. Chaudhari,
Sorbitol hydrogenolysis over hybrid Cu/CaO-Al catalysts: tunable activity
[
[
2
O
3
and selectivity with solid base incorporation, Acs. Catal. 5 (2015) 6545-6558.
39] H. Xiong, M.A. Motchelaho, M. Moyo, L.L. Jewell, N.J. Coville, Effect of Group
I alkali metal promoters on Fe/CNT catalysts in Fischer-Tropsch synthesis, Fuel
1
50 (2015) 687-696.

Figure Captions:
Figures
Fig. 1. Typical raman spectra for the raw and functionalized CNTs.

Fig. 2. XRD patterns of various supported Ru catalysts with different functionalized
carbon nanotubes as supports.

Fig. 3. The adsorption–desorption isotherms of N
2
on different MWCNTs (A): MCN
(); AMCN (); CMCN (○); HMCN (☆); NMCN () and supported ruthenium
catalysts (B): Ru/MCN (); Ru/AMCN (); Ru/CMCN (○); Ru/HMCN ( ☆ );
Ru/NMCN ().

☆
Fig. 4. HRTEM images and ruthenium particle size distribution of the catalysts: (a)
Ru/MCN; (b) Ru/AMCN; (c) Ru/CMCN; (d) Ru/HMCN and (e) Ru/NMCN.

Fig. 5. NH
3
-TPD (A) and CO -TPD (B) profiles of Ru/MCN catalysts with different
2
functional groups: (a) Ru/MCN; (b) Ru/AMCN; (c) Ru/CMCN; (d) Ru/HMCN; (e)
Ru/NMCN.

(a) C1s