Selective C?C Hydrogenolysis of Alkylbenzenes to Methylbenzenes with Suppression of Ring Hydrogenation

Article abstract of DOI:10.1002/cctc.201801118

Selective reduction of alkylbenzenes, which are supplied by refining heavy hydrocarbons, to methylbenzenes by selective C?C hydrogenolysis was accomplished by Ru/CeO₂ catalyst. Hydrogenolysis of ethylbenzene was carried out with gas-phase fixed-bed flow reactor and various noble metal catalysts. The catalyst with high Ru loading (4 wt%) on CeO₂ support pre-calcined at high temperature (1073 K; Ru(4)/CeO₂¹⁰⁷³ catalyst) showed both good activity and selectivity to toluene. The yield of toluene reached 65 %-C in the conditions of 533 K and H₂ partial pressure of 0.075 MPa. Hydrogenolysis of other alkylbenzenes including p-ethyltoluene, propylbenzene and cumene was also carried out. Hydrogenolysis of C_aryl-C_alkyl bond and hydrogenation of benzene ring was slow in all substrate cases. Hydrogenolysis of p-ethyltoluene can give p-xylene as the main product (highest yield: 53 %-C). These high yields of methylbenzenes make a marked contrast to the traditional bifunctional hydrocracking systems which preferably dissociate C_aryl-C_alkyl bond in alkylbenzenes. Characterization of Ru catalysts suggests that the block-shaped Ru particles with small size (<2 nm) in Ru(4)/CeO₂¹⁰⁷³ give good catalytic performance. Too strong interaction between Ru species and CeO₂ support induced by decreasing the loading amount rather gives ring-hydrogenation activity and decreases methylbenzenes selectivity.

Full text of DOI:10.1002/cctc.201801118

Accepted Article
Title: Selective C-C Hydrogenolysis of Alkylbenzenes to
Methylbenzenes with Suppression of Ring Hydrogenation
Authors: Shin Yanatake, Yosuke Nakaji, Mii Betchaku, Yoshinao
Nakagawa, Masazumi Tamura, and Keiichi Tomishige
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201801118
Link to VoR: http://dx.doi.org/10.1002/cctc.201801118
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201801118
Selective C-C Hydrogenolysis of Alkylbenzenes to Methylbenzenes
with Suppression of Ring Hydrogenation
a
a
a
a,b,*
Shin Yanatake, Yosuke Nakaji, Mii Betchaku, Yoshinao Nakagawa,
Masazumi
a,b
a,b,*
Tamura and Keiichi Tomishige
a
Shin Yanatake, Yosuke Nakaji, Mii Betchaku, Dr. Yoshinao Nakagawa, Dr. Masazumi Tamura,
Prof. Dr. Keiichi Tomishige
Department of Applied Chemistry, School of Engineering, Tohoku University, 6-6-07 Aoba,
Aramaki, Aoba-ku, Sendai, 980-8579, Japan
b
Dr. Yoshinao Nakagawa, Dr. Masazumi Tamura, Prof. Dr. Keiichi Tomishige
Research Center for Rare Metal and Green Innovation, Tohoku University, 468-1, Aoba,
Aramaki, Aoba-ku, Sendai 980-0845, Japan
*
Corresponding authors.
E-mail: yoshinao@erec.che.tohoku.ac.jp, Tel/Fax: +81-22-795-7215 (Y. N.);
tomi@erec.che.tohoku.ac.jp, Tel&Fax: +81-22-795-7214 (K. T.)
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Abstract
Selective reduction of alkylbenzenes, which are supplied by refining heavy hydrocarbons, to
methylbenzenes by selective C-C hydrogenolysis was accomplished by Ru/CeO catalyst.
2
Hydrogenolysis of ethylbenzene was carried out with gas-phase fixed-bed flow reactor and
various noble metal catalysts. The catalyst with high Ru loading (4 wt%) on CeO support
2
1073
pre-calcined at high temperature (1073 K; Ru(4)/CeO2
catalyst) showed both good activity
and selectivity to toluene. The yield of toluene reached 65 %-C in the conditions of 533 K and
H partial pressure of 0.075 MPa. Hydrogenolysis of other alkylbenzenes including
2
p-ethyltoluene, propylbenzene and cumene was also carried out. Hydrogenolysis of C -C
aryl alkyl
bond and hydrogenation of benzene ring was slow in all substrate cases. Hydrogenolysis of
p-ethyltoluene can give p-xylene as the main product (highest yield: 53 %-C). These high
yields of methylbenzenes make a marked contrast to the traditional bifunctional
hydrocracking systems which preferably dissociate C -C
bond in alkylbenzenes.
aryl alkyl
Characterization of Ru catalysts suggests that the block-shaped Ru particles with small size
1073
(<2 nm) in Ru(4)/CeO2
give good catalytic performance. Too strong interaction between
Ru species and CeO support induced by decreasing the loading amount rather gives
2
ring-hydrogenation activity and decreases methylbenzenes selectivity.
Keywords
Heterogeneous catalysis, Cleavage reactions, BTX, Ruthenium, ceria support
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
1
. Introduction
Conversions of heavy hydrocarbon resources into value-added chemicals are becoming
more and more important, since heavier crude oil with high mining cost has been more
[1]
supplied. Monoaromatic hydrocarbons with a long side chain are included in products of
catalytic cracking of polycyclic aromatic hydrocarbons which are one class of main
[2-4]
components of heavy oil.
Except ethylbenzene, monoaromatic hydrocarbons with long
alkyl chains are not so useful chemicals and are mainly used as fuel. Such monoaromatic
hydrocarbons are also present in heavy naphtha (5-10%), and the amount is increased by
[
5,6]
reforming (about twice).
Although the percentages are not so high, the absolute amount is
quite large because heavy naphtha is utilized in very large scale. On the other hand, aromatics
with methyl groups, especially p-xylene, are highly useful chemicals. Therefore, conversion
of the alkyl groups of monoaromatic hydrocarbons to methyl groups is a valuable process;
however, such process has been hardly reported. Conventional methods for C-C dissociations
[
7,8]
include thermal cracking, catalytic cracking and hydrocracking (hydrogenolysis).
The
main products of both thermal cracking and catalytic cracking of aromatic hydrocarbons with
[
9]
long side chains are not methylbenzenes: formation of styrenes and dissociation of
[
10,11]
C -C
aryl alkyl
bond
proceeds instead, respectively. There are several hydrogenolysis systems
of aromatic hydrocarbons. Reactions of tetralin over metal-acid bifunctional catalysts such as
[12,13]
[14]
[15]
noble metal/zeolite,
noble metal/SiO -Al O ,
noble metal/SiO -ZrO ,
and
2
2
3
2
2
[16]
Ni/zeolite are typical examples. Ni-Mo catalysts with acidic property, which are analogues
of hydrodesulfurization catalysts, are also tested as bifunctional catalysts for
1
7,18
hydrogenolysis.
These bifunctional catalysts cleave the C-C bond mainly via
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Friedel-Crafts-type reaction over the acid sites, and C -C
bond is preferably dissociated
aryl alkyl
(butylbenzene from tetralin). In addition, bifunctional catalysts have hydrogenation activity
and the benzene ring tends to be hydrogenated, especially at lower reaction temperature.
Recently, the use of H-beta as acid component such as Ni/H-beta has been reported to produce
[
19,20]
BTX fractions in significant yield at high reaction temperature (~673 K),
and the
[
21,22]
reported yields of toluene + xylenes from tetralin are 20~30%.
On the other hand,
[23]
monofunctional metal catalysts were also tested for C-C bond cleavage by hydrogenolysis.
[
24]
Ir, Rh and Ru are typical active metals, and Ru is known to have higher activity. Recently,
we have developed Ru/CeO catalyst with very small Ru particles (< 1.5 nm) prepared by
2
[
25,26]
treatment with inert gas at 573 K after impregnation.
This C-C dissociation catalyst
shows good regioselectivity toward inner C-C bonds instead of terminal C-C bonds of
[
25-27]
alkanes.
We also developed Ru-VO /SiO catalyst with small Ru ensembles which give
x 2
[
28]
similar good regioselectivity, although the modification of V decreases the activity of Ru.
Because of the high hydrogenation activity, the monofunctional catalysts have been mainly
applied to saturated substrates such as naphthenes, higher n-alkanes and hydrogenated
terpenoids, and the C-C dissociations of aromatic substrates are hardly tested. In this paper,
we tested various noble metal catalysts including Ru/CeO for hydrogenolysis of C-C bods in
2
alkyl chain of monoaromatics. We found that Ru/CeO catalyst with appropriate Ru loading
2
can effectively convert ethylbenzene to toluene (eq. (1)) with suppression of ring
hydrogenation (eq. (2)) and C -C
dissociation (eq. (3)) under appropriate reaction
aryl alkyl
conditions. Production of p-xylene from 4-ethyltoluene (eq. (4)) is also possible with this
catalyst.
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
°
298
-1
°
-1
C H C H (g) + H → C H CH (g) + CH H
= -55 kJ mol , G
= -53 kJ mol
298
6
5
2
5
2
6
5
3
4
(1)
°
-1
°
-1
C H C H (g) + 3H → C H C H (g) H
= -203 kJ mol , G
= -92 kJ mol
298
(2)
(3)
6
5
2
5
2
6
11
2
5
298
°
-1
°
-1
C H C H (g) + H → C H (g) + C H H
= -31 kJ mol , G
= -17 kJ mol
298
6
5
2
5
2
6
6
2
6
298
p-CH C H C H + H → p-CH C H CH + CH
(4)
3
6
4
2
5
2
3
6
4
3
4
2
2
. Experimental section
.1. Catalyst preparation
Ru/support catalysts (Ru: typically 4 wt%) were prepared by impregnation method with
various supports and Ru(NO)(NO ) (OH) in diluted nitric acid (Sigma Aldrich). Supports
3
3-x
x
used in experiment were as follows: SiO (Fuji Silysia Chemical Ltd., G-6, calcined at 973 K
2
for 1 h), CeO (Daiichi Kigenso Kagaku Kogyo Co., Ltd., HS, calcined for 3 h (typically 1073
2
K)), Al O (Nippon Aerosil Co., Ltd., AEROXIDE Alu C, calcined at 973 K for 1 h), MgO
2
3
(
Ube Industries Ltd., 500A, calcined at 973 K for 1 h), TiO (Nippon Aerosil Co., Ltd.,
2
AEROXIDE P25, calcined at 973 K for 1 h), and ZrO (Daiichi Kigenso Kagaku Kogyo Co.,
2
Ltd., RC-100P, calcined at 973 K for 1 h). M/CeO (M = Rh, Ir, Pt, Pd; M 4 wt%) were also
2
prepared by impregnation method with various precursors. Metal precursors used in
experiment were as follows: Rh(NO ) (Wako Pure Chemicals Industries, Ltd.), Ir(NO )
3 3
3
3
(
Furuya Metal Co., Ltd.), Pt(NO ) (NH ) (Tanaka Kikinzoku Kogyo), Pd(NO ) (N.E.
2 2 3 2 3 2
CHEMCAT Corp.). After evaporating the solvent, all the prepared catalysts were dried at 383
K for 12 h, and pretreated for 1 h in N flow at 573 K. The catalysts are denoted as
2
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
n
M(m)/support , where m and n represents the loading amount in wt% for M and the
calcination temperature in K for support, respectively.
2
4
.2. Activity tests
Hydrogenolysis of ethylbenzene (Tokyo Chemical Industry Co., Ltd., > 99%),
-ethyltoluene (Tokyo Chemical Industry Co., Ltd., > 95%), propylbenzene (Tokyo Chemical
Industry Co., Ltd., > 99%), cumene (Tokyo Chemical Industry Co., Ltd., > 99%), p-xylene
Wako Pure Chemicals Industries, Ltd., > 98%) and toluene (Wako Pure Chemicals Industries,
(
Ltd., > 99.5%) was performed in a fixed-bed quartz tube reactor under atmospheric pressure.
Before the activity test, the catalyst (100 mg; 60-80 mesh) was reduced at 533 K for 0.5 h in
H flow (50 ml/min). After the pretreatment, the feed gas was changed to hydrocarbon/N /H
2
2
2
−1
(
typically in a 1:26:78 molar ratio, total gas feed rate 128 mmolꢀh (W/F=0.8
−1
−1
−total
g−catꢀh ꢀmol
)). The feed was maintained for 2 h to reach a steady state. The product gas
was collected in a hot syringe heated at 353 K and analyzed with GC (Shimadzu GC-2014)
equipped with flame ionization detector (FID). An Rtx®-1 PONA column (Restek Corp.) was
used for separation. 0.1 ml of gas sample was injected into GC. The conversion and
selectivity were calculated on carbon-basis with eqs. (5) and (6) as shown below. The carbon
balance (eq. (7)) was always in the range of the experimental error (±10%), except in the
cases of low conversion of substrate where the carbon balance was sometimes fluctuated
(120% in the worst case) probably because of the condensation of substrate. In the cases of
very high conversion level (>90%), the conversion was calculated by the decrease of substrate
amount instead of eq. (5).
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
∑
Amount of products [mol-C]
Conversion (%)= ꢀ
ꢁ ×100 (5)
Amount of substrate input [mol-C]
Amount of each product [mol-C]
Selectivity (%)= ꢂ ∑ Amount of products [mol-C] ꢃ ×100 (6)
Amount of unreacted substrate [mol-C]+∑ Amount of products [mol-C]
Carbon balance (%)= ꢂ
ꢃ ×100
(7)
Amount of substrate input [mol-C]
2
.3. Characterization
BET surface area was measured with Micromeritics Gemini VII 2390 instrument. H2
temperature programmed reduction (H -TPR) was carried out in a fixed-bed reactor equipped
2
with a thermal conductivity detector (TCD) and frozen acetone trap using 5% H diluted with
2
Ar (30 ml/min). The amount of sample was 0.05 g or 0.1 g, and the temperature was increased
from 223 K (cooled with frozen acetone) to 973 K at a heating rate of 10 K/min.
Thermogravimetry-differential thermal analysis (TG-DTA) was carried out with a Rigaku
Thermo Plus EVO-II under air atmosphere. X-ray diffraction (XRD) patterns were recorded
by a diffractometer (Rigaku MiniFlex600). Cu K ( 0.154 nm, 45 kV, 40 mA) radiation

was used as an X-ray source. The amount of H chemisorption was measured in a
2
high-vacuum system using a volumetric method in a vacuum line. Before adsorption
measurements, the catalysts were reduced with H at 533 K for 2 h. The adsorption was
2
conducted at room temperature and the final H pressure was about 4 kPa. The measurement
2
of H adsorption was repeated after evacuation at room temperature in order to determine the
2
physisorption amount. The chemisorption amount was determined by the difference of these
two adsorption amounts. Transmission electron microscope (TEM) images were obtained on a
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
JEOL JEM-2100F (for Ru/CeO samples) or Hitachi HD-2700 (for Ru/(other support)
2
samples) instrument. The catalysts after the reduction with H at 533 K or after reaction were
2
used as samples for the TEM observation. The samples were dispersed in ethanol under
supersonic waves and were dropped on grids. X-ray photoelectron spectra (XPS) were
measured with Shimadzu AXIS-ULTRA DLD spectrometer under high vacuum at room
temperature. The catalyst sample was mixed with SiO (1:1 in weight), and pressed into a disk.
2
The disk was reduced with H at 533 K for 0.5 h, and transferred to the measurement chamber
2
in inert atmosphere. The binding energy was corrected by Si 2s peak at 154.4 eV because of
the overlap between C 1s and Ru 3d signals.
3
/2
3
3
. Results and discussion
.1. Hydrogenolysis of ethylbenzene over noble metal catalysts
Ethylbenzene was selected as a model substrate in this study (eq. (1)). The target product is
toluene, and the maximum carbon-based selectivity of toluene is 88% (=7/8). Methane is
always co-produced with toluene (methane:toluene carbon ratio = 1:7), while some side
reactions also produce methane. Ethylcyclohexane (ECH) and methylcyclohexane (MCH) are
main by-products via hydrogenation of benzene ring. Other by-products are benzene,
cyclohexane, and smaller hydrocarbons that are produced by overhydrogenolysis. Figure 1
shows the results of hydrogenolysis of ethylbenzene over 4 wt% noble metal catalysts
1073
supported on CeO pre-calcined at 1073 K (M(4)/CeO2 ). Ru catalyst showed the highest
2
1073
activity among the tested noble metal catalysts: The conversion level over Ru(4)/CeO2
was
1073
comparable or higher than those over other M(4)/CeO2
8
catalysts at one fourth of contact
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
1
073
time. The selectivity to toluene over Ru(4)/CeO2
was higher than those over other noble
metal catalysts, and the high selectivity was almost maintained at higher conversion level. The
carbon ratio of produced methane:toluene (1:5.9 at 40% conversion) was only slightly smaller
the stoichiometrical one (1:7), indicating that amount of methane produced by side-reactions
was small. Rh and Ir catalysts showed some activity in toluene formation; however,
ethylcyclohexane was also formed in similar selectivity to toluene. In the cases of Pd and Pt
catalysts, the main product was ethylcyclohexane. Therefore, Ru is the most suitable active
metal species for hydrogenolysis of ethylbenzene to toluene without hydrogenation of the
benzene ring.
Next, we investigated the effect of support for Ru catalysts in hydrogenolysis of
ethylbenzene. Figure 2 shows the reaction results over various Ru(4)/support catalysts. All the
tested Ru catalysts showed good selectivity to toluene. The activity much depended on the
used support. CeO supports gave much higher activity than other supports. This is in contrast
2
873
to the hydrogenolysis of alkanes where Ru(4)/CeO2 catalyst shows significantly lower
9
73
[25-27]
activity than Ru(4)/SiO2 catalyst.
As described later, the higher activity of Ru/CeO2
than other Ru catalysts shown in Figure 2 can be mostly (although not totally) explained by
the difference of dispersion. The dependence of the performance of Ru(4)/CeO catalysts on
2
1
073
the calcination temperature of support was also investigated, and Ru(4)/CeO2
showed the
are shown in Figure
S1, supporting information). According to our previous reports on the CeO catalysts, CeO
673
1273
highest activity (The performances of Ru(4)/CeO2 and Ru(4)/CeO2
2
2
calcined at 873 K or higher temperature has sufficiently high crystallinity, and higher
calcination temperature than 873 K simply decreases both the surface area and catalytic
9
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
[29]
activity in acid/base reactions.
In order to further investigate the effect of calcination
873
temperature of CeO support, the Ru loading amount was also changed for both CeO2
2
2
-1
1073
2
-1
(
pre-calcined at 873 K; BET surface area 83 m g ) and CeO
(41 m g ) supports. The
2
results of the reaction tests are shown in Figure 3. The activity (conversion) fairly
corresponded to the Ru loading amount. In view of selectivity, considering that selectivity to
toluene was gradually decreased at higher conversion level, catalysts with low Ru loading
amount showed much lower selectivity to toluene, and ethylcyclohexane became the main
product instead. The selectivity change suddenly occurred at some level of Ru loading
1
073
amount: For CeO2
support, the selectivity change occurred between Ru loading amount of
873
0
.5 and 1 wt%, while for CeO2 support it occurred between 1 and 4 wt%. Considering that
8
73
1073
the surface area of CeO2 is about twice of that of CeO2 , the selectivity of Ru/CeO is
2
suggested to be determined by the Ru concentration on the CeO surface ([Ru loading
2
amount]/[surface area of CeO support]). High Ru surface concentration on CeO above ~1
2
2
2
2
-1
atom/nm (corresponding to 0.75 wt% Ru on support with 40 m g surface area) gives good
1073
selectivity to toluene. After all, we selected Ru(4)/CeO2
reaction.
as the optimized catalyst for this
3
.2. Optimization of reaction conditions
Figure 4 shows the effect of reaction temperature on hydrogenolysis of ethylbenzene over
1073
Ru(4)/CeO2 . Lower reaction temperature (513 K) gave higher selectivity to
ethylcyclohexane which is the hydrogenation product of the benzene ring. In view of both
activity and selectivity, 533 K is better reaction temperature than 513 K. Here, it should be
1
0
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
discussed whether the equilibrium of ring hydrogenation/dehydrogenation limited the
formation of ethylcyclohexane or not. In Figure 1, Pt/CeO catalyst showed almost 100%
2
selectivity to ethylcyclohexane at 40% conversion at 533 K, which means that the equilibrium
limitation of the ethylcyclohexane formation is not lower than this yield. On the other hand,
1073
Ru(4)/CeO2
catalyst showed <10% selectivity to ethylcyclohexane at similar conversion
1073
level. Therefore, the formation amount of ethylcyclohexane over Ru(4)/CeO2
was much
lower than the equilibrium level, at least in the case of 533 K. The high selectivity to toluene
and low selectivity to ethylcyclohexane were kinetically controlled.
At 553 K, the activity was increased from 533 K; however, the difference was not so large.
The selectivity to toluene was similar to that at 533 K. The smaller increase of activity
between 533 and 553 K than between 513 and 533 K may be due to the decrease of substrate
coverage on the active site at higher temperature. The selectivity to ethylcyclohexane was
decreased and those to smaller hydrocarbons including methane were increased instead.
Considering the decrease of selectivity to ethylcyclohexane from 513 K to 533 K, the very
low selectivity to ethylcyclohexane at 553 K was probably due to the suppression of
hydrogenation of benzene ring rather than rapid overhydrogenolysis of ethylcyclohexane.
However, overhydrogenolysis of toluene to smaller hydrocarbons more proceeded at high
conversion levels at 553 K, and the toluene yield at 553 K was rather lower than that 533 K.
We selected 533 K as the standard reaction temperature.
[
26]
873
In our previous report, hydrogenolysis of alkanes over Ru(5)/CeO2 was carried out at
-1
much lower reaction temperature (433 K) under similar flow conditions (W/F ~0.1 gcat
h
-1
mol , alkane:N :H ~ 1:60:60, ambient pressure). The conversion of n-hexane at W/F = 0.08
2
2
11
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
-1
-1
26
g h mol was about 20%, and about 40 mol% of the total products was methane. These
cat
data indicate that the reactivity to ethylbenzene is much lower than n-alkanes. Strong
adsorption of benzene ring on the catalyst surface may lower the reactivity to hydrogenolysis
over Ru.
Figure 5 shows the effect of H :N ratio (H partial pressure) on hydrogenolysis of
2
2
2
ethylbenzene. Higher H concentration (higher H /N ratio) gave higher activity. Although the
2
2
2
selectivity to ethylcyclohexane was slightly increased with increase of H concentration, we
2
selected H /N ratio = 3 (P(H ) = 0.075 MPa) as the standard reaction conditions because of
2
2
2
the high reaction rate. Too high H pressure may lead to high selectivity to ring-hydrogenated
2
products. In fact, hydrogenation of aromatic ring over Ru/CeO catalysts under high-pressure
2
[
30,31]
H (a few MPa) in batch reactor has been reported.
2
At 533 K under H /N = 3, 65% yield
2
2
-1
-1
of toluene (theoretical maximum = 88%) was obtained at W/F = 0.4 g h mol . This value
cat
is much higher than the reported yield values (<~30%) of toluene + xylene by hydrocracking
[
21,22]
of aromatics.
Longer contact time (W/F) increased the selectivity to smaller
hydrocarbons including methane and decreased the selectivity to toluene.
The effect of substrate (ethylbenzene) partial pressure was also investigated, and the
detailed data are shown in Table S1. The selectivity pattern was not changed by the substrate
1073
1073
partial pressure for both Ru(4)/CeO2
and Ru(0.5)/CeO2 . The substrate partial pressure
has weak positive effect on the activity; the reaction order with respect to ethylbenzene partial
pressure was about 0.3 for both catalysts, indicating that the active sites on the both catalysts
fairly adsorbed the ethylbenzene molecule. The low reaction order was consistent with the
data that almost total conversion was obtained in Figures 4 and 5.
1
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
3
.3. Hydrogenolysis of other substrates
The hydrogenolysis system of Ru(4)/CeO2
1
073
catalyst was applied to various aromatic
hydrocarbons (Table 1). Because hydrogenolysis of p-dialkylbenzenes can give p-xylene
which is a very important raw material of PET, we carried out hydrogenolysis of
p-ethyltoluene (Table 1, entries 5-8). The reactivity of p-ethyltoluene was slightly lower than
that of ethylbenzene. The highest p-xylene selectivity was 53% (theoretical maximum = 8/9 =
-1
-1
8
9%), which was obtained around W/F = 0.4~0.8 g h mol .
cat
In the case of propylbenzene (Table 1, entries 9-11), both toluene and ethylbenzene were
formed, and toluene was more formed (toluene/ethylbenzene molar ratio ~ 3) (Scheme 1A).
Comparable amount of ethane to toluene was actually detected (ethane/toluene molar ratio ~
0
.7), which indicated that most toluene was produced by direct dissociation of PhC-C bond
rather than stepwise hydrogenolysis via ethylbenzene. The ratios were gradually changed by
the overhydrogenolysis of ethylbenzene. The reactivity of propylbenzene was much lower
than those of ethylbenzene and p-ethyltoluene.
Hydrogenolysis of cumene (isopropylbenzene) gave ethylbenzene as the main product at
short contact time (Table 1, entry 12). As the contact time was increased, the selectivity to
ethylbenzene was decreased and that to toluene was increased, suggesting that toluene was
mainly produced by successive hydrogenolysis via ethylbenzene (Table 1, entries 12-14)
(Scheme 1B). The reactivity of cumene was also lower than those of ethylbenzene and
4
-ethyltoluene.
The reactions of target products, toluene and p-xylene, were also investigated (Table 1,
1
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
entries 15 and 16). The reactivities were surely lower than alkylbenzenes with ≥C2 alkyl
chains. In particular, hydrogenolysis of toluene to benzene was very slow, and rather
hydrogenation to methylcyclohexane more proceeded. The results agreed with that benzene
was hardly formed from ethylbenzene, propylbenzene and cumene. On the other hand, the
main product from p-xylene was toluene, and the selectivity to ring-hydrogenated products
was low. The adsorption of the substrate with the benzene ring might be blocked to some
extent by the two methyl groups.
The reactivity and selectivity trends are related to the reaction mechanism. According to the
[
32]
[26]
literature
and our previous studies,
the C-C hydrogenolysis of alkanes over metal
surface starts with dissociation of C-H bond and formation of C-metal bond. In alkylbenzenes,
the weakest C-H bond is that at the benzyl position, and thus (Ph-)C-metal bond is preferably
formed. Next, the dissociation of C-C bond produces two alkyl or alkylidene surface species
in hydrogenolysis of alkanes. In the case of ethylbenzene (Scheme 2A), PhCH(CH )- surface
3
species is formed first. There can be two C-C dissociation reactions: formation of Ph- + C H -
2
5
surface species and that of PhCH- + CH - surface species. The former pair of surface species
3
is less stable than the latter pair because C-H bond in benzene ring is much stronger than that
at the benzyl position. Therefore toluene and methane are the main products from
ethylbenzene. In the case of propylbenzene, the main reaction is the formation of toluene and
ethane in similar mechanism. However, the steric hindrance around the benzyl position can
weaken the formation of the surface intermediate species, leading to the lower reactivity than
ethylbenzene. In the case of p-ethyltoluene (Scheme 2B), p-CH C H CH(CH )- surface
3
6
4
3
species is the precursor of p-xylene. However, p-C H C H CH - surface species can be also
2
5
6
4
2
1
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
formed, which is not reactive in further hydrogenolysis. The competitive formation of these
surface species can decrease the reactivity of p-ethyltoluene in comparison with ethylbenzene.
3
.4. Catalyst stability
1
073
We conducted the stability test of Ru(4)/CeO2
for hydrogenolysis of ethylbenzene
(Figure 6). The conversion and selectivity pattern were not changed. Thermogravimetric (TG)
analysis of the used catalyst showed that the amount of deposited organic substance (coke)
was negligible (Figure S2). The TEM image of the used catalyst was also essentially
unchanged from the fresh reduced catalyst (Figure S3). These results indicate that the
1073
Ru(4)/CeO2
catalyst is stable during hours of use. The good stability of Ru/CeO catalyst
2
for hydrocarbon hydrogenolysis has been also reported in alkane hydrogenolysis in both batch
[
25-27]
and fixed-bed reactions.
3
.5. Catalyst characterization
The Ru catalysts on various supports were characterized by N adsorption (BET surface
2
area), XRD, TEM and H chemisorption. The results are summarized in Table 2, and the raw
2
data are shown in Figures S4 (XRD) and S5 (TEM). Although measurement of Ru particles
was not possible for all catalysts with XRD and TEM because of the detection limit or the
overlap of signals with support, the size of Ru particles determined by XRD and TEM fairly
agreed with the dispersion determined by H chemisorption (H/Ru). From the comparison
2
between the dispersion and activity, higher activity of Ru(4)/CeO catalysts than other Ru(4)
2
catalysts can be mostly explained by the higher dispersion. However, the activity of
1
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Ru/support catalysts cannot be explained only by dispersion. Catalysts with acidic support
such as Al O have relatively lower activity.
2
3
As described in section 3.1 and Figure 3, the selectivity of Ru/CeO catalyst much
2
depends on the concentration of Ru on the surface of CeO support, and higher concentration
2
gives higher selectivity to toluene from ethylbenzene. From Table 2, there is little correlation
between the selectivity and dispersion of the six Ru/CeO catalysts. Therefore we further
2
characterize the Ru/CeO catalysts with different Ru loadings.
2
1
2
073
The H -TPR profiles of Ru/CeO
catalysts are shown in Figure 7. All the catalysts
2
showed reduction signals below 473 K, and thus the reduction completed below the reaction
-1
temperature (≥513 K). The H consumption amount was 0.40, 0.56 and 1.4 mmol g for
2
1073
1073
1073
Ru(0.5)/CeO2 , Ru(1)/CeO2
and Ru(4)/CeO2 , respectively, and was higher than the
amount required for the reduction of oxidized Ru such as RuO to Ru metal. The surface of
2
1073
CeO was also reduced. The reduction temperature was similar for all the Ru/CeO2
2
catalysts.
1
073
The TEM images of reduced Ru/CeO2
catalysts are shown in Figure 8. In Figure 8(a)
1
073
1073
and (b) which represents Ru(4)/CeO2
clearly observed. On the other hand, in Ru(0.5)/CeO2
activity in benzene ring hydrogenation and low selectivity to toluene, Ru particles were not
and Ru(1)/CeO2 , respectively, Ru particles are
1073
catalyst, which has relatively high
1
073
observed. The Ru species in Ru(0.5)/CeO2
is too small in particle size or amorphous. We
1073
also carried out XAFS measurement of reduced Ru/CeO2
catalysts to determine the
coordination environment of the Ru particles. However, in spite of repeated trials for the
measurement of reduced catalysts without exposure to air, the obtained Ru K-edge spectrum
1
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
1
073
of Ru(0.5)/CeO2
were always the pattern of oxidized Ru species (Figure S6). Considering
the TPR results, it is unlikely that the Ru species are not reduced by the treatment with H at
2
1073
5
33 K. The Ru species in Ru(0.5)/CeO2
catalyst is probably very easily oxidized by trace
O (cf. the detection limit of O meter in the glove bag was 0.1%; carrying the sample reduced
2
2
in our lab to the synchrotron apparatus took a few days). The absence of Ru particles in TEM
1
073
image of Ru(0.5)/CeO2
catalyst (Figure 8(c)) agrees with the idea that the Ru species is
1073
very easily oxidized with air. On the other hand, the XAFS spectra of Ru(4)/CeO2
and
1073
Ru(1)/CeO2
were the patterns of metallic Ru, and the Ru-Ru coordination numbers (CNs)
in EXAFS curve fitting analysis were 7.8 and 4.2, respectively (Table S2). The much lower
CNs than that of bulk Ru (CN = 12) agreed with the formation of small Ru particles (< 2 nm)
as observed in TEM.
We also carried out XPS measurement of the reduced catalysts (Figure S7). The samples
after reduction were quickly transferred to vacuum chamber in inert atmosphere. All the
1073
873
catalysts including Ru(0.5)/CeO2
and Ru(0.5)/CeO2 showed stronger signal of metallic
0
IV
Ru than that of Ru O in Ru 3d region, confirming that Ru was surely reduced to the
2
5/2
0
metallic state by heating under H . The peak position of Ru signal seems to be shifted to
2
lower binding energy (~0.3 eV) by decreasing the Ru loading amount, suggesting that the
electronic state of Ru particles was negatively charged in low Ru loading due to reduced CeO2
surface. However, the determination of the electronic state of metallic Ru species by these
spectra is not decisive because there was still some signal of oxidized Ru species; surface
oxidation of Ru particles probably proceeded during the sample transfer process, and the
electronic state of Ru metal particles can be changed by this oxidation. We do not include the
1
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
detailed curve-deconvolution results of these XPS data, because we do not think that the ratio
n+
0
of Ru /Ru in the data represents the exact state of reduced catalysts.
1
073
1073
In the three Ru/CeO2
catalysts, only Ru(0.5)/CeO2
showed low selectivity to
873
873
873
toluene. Similarly, for Ru/CeO2 catalysts, Ru(1)/CeO2 and Ru(0.5)/CeO2 which had
low selectivity to toluene showed XAFS spectra of oxidized Ru species, while Ru metal was
8
73
detected in XAFS spectra of Ru(4)/CeO2 which had good selectivity to toluene. The Ru
species that are easily oxidized by O seem to have activity in hydrogenation of benzene ring.
2
1
073
Here, the TEM images are closely investigated again. The Ru particles in Ru(4)/CeO2
1
073
are block-shaped (Figure 8(a)). On the other hand, the Ru particles in Ru(1)/CeO2
are low
considering the much
indicates the small volume of each Ru particle,
1073
hill-shaped. The relatively large number of particles in Ru(1)/CeO2
1073
lower Ru loading amount than Ru(4)/CeO2
agreeing with the low shape. Lower height of supported particles means that they are more
sensitive to the electronic effect by the support. Lowering the Ru loading amount decreases
the height of supported particles rather than decreases the particle size with similar shape.
Further lowering the Ru loading amount can give thin flat particles. The electronic state of
such particles is highly affected by the CeO support, and the particles are very easily
2
oxidized. The flat particles have activity in benzene ring hydrogenation. The schematic
representation of catalyst structure and catalytic performance described above are summarized
in Figure 9. The hydrogenation activity of flat Ru species on low-loading-amount Ru/CeO2
catalysts can be due to electronic effect by the CeO support or geometric effect of flat Ru
2
surface that can adsorb flat aromatic ring. Considering that very large Ru particles which also
have flat Ru surface have low hydrogenation selectivity, the electronic effect by the CeO2
1
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
surface is more plausible. Although the hydrogenation is not the target reaction in this work,
the investigation of the electronic effect can be an interesting topic in future works.
4
. Conclusions
1073
Ru(4)/CeO2
catalyst (4 wt% Ru loading on CeO support pre-calcined at 1073 K) shows
2
good activity and selectivity in hydrogenolysis of ethylbenzene to toluene with suppressing
hydrogenation of the benzene ring. Catalysts supporting another noble metal showed
significant selectivity to ring-hydrogenation products. The highest yield of toluene was
6
5 %-C (theoretical maximum yield: 88 %-C) at 533 K and H partial pressure of 0.075 MPa
2
1
073
over Ru(4)/CeO2
catalyst. This yield value is much higher than those of methylbenzenes
yield from large alkylbenzenes in literature hydrogenolysis systems such as Ni/H-beta. The
selections of reaction temperature and H pressure are very important: too low temperature
2
and too high H pressure increases selectivity to ring-hydrogenated products; too high
2
temperature promotes overhydrogenolysis to smaller hydrocarbons; and too low H pressure
2
decreases the activity. The catalyst has excellent stability during hours of reaction, and no
coke deposition was observed. Other aromatic hydrocarbons with ≥C2 chains were also
converted to methylated aromatic hydrocarbons. The reactivity of aromatic hydrocarbons with
C3 chain is lower than that of ethylbenzene probably because of the steric hindrance. The
1
073
good activity of Ru(4)/CeO2
catalyst is due to the high dispersion of Ru particles. However,
the interaction of Ru particles and CeO is not wholly advantageous to the catalysis: too low
2
loading amount of Ru on CeO support gives hydrogenation activity of benzene ring to the
2
catalyst by formation of Ru particles highly interacted with CeO support.
2
1
9
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Acknowledgements
We thank the Technical Division in School of Engineering, Tohoku University for TEM and
XPS measurements.
References
[
[
[
1] M. S. Rana, V. Sámano, J. Ancheyta, J. A. I. Diaz, Fuel 2007, 86, 1216-1231.
2] Y. Miki, Y. Sugimoto, Fuel Proc. Technol. 1995, 43, 137-146.
3] I. Shimada, K. Takizawa, H. Fukunaga, N. Takahashi, T. Takatsuka, Fuel 2015, 161,
2
07-214.
[
[
4] R. Pujro, M. Falco, U. Sedra, Energy Fuels 2015, 29, 1543-1549.
5] J. Ancheyta-Juárez,; E. Villafuerte-Marcías, L. Díaz-García, E. González-Arredondo,
Energy Fuels 2001, 15, 887-893.
6] A. A. Al-Shammari, S. A. Ali, N. Al-Yassir, A. M. Aitani, K. E. Ogunronbi, K. A.
Al-Majnouni, S. S. Al-Khattaf, Fuel Proc. Technol. 2014, 122, 12-22.
[
[
[
[
[
[
[
[
7] A. Corma, B. W. Wojciechowski, Catal. Rev. 1985, 27, 29-150.
8] J. Weitkamp, ChemCatChem 2012, 4, 292-306.
9] Q. Chen, G. F. Froment, J. Anal. Appl. Pyrolysis 1991, 21, 51-77.
10] H. B. Mostad, T. Riis, O. H. Ellestad, Appl. Catal. 1990, 63, 345-364.
11] A. T. Townsend, J. Abbot, Appl. Catal. A 1992, 90, 97-115.
12] M. J. Girgis, Y. P. Tsao, Ind. Eng. Chem. Res. 1996, 35, 386-396.
13] S. J. Ardakani, K. J. Smith, Appl. Catal. A 2011, 403, 36-47.
2
0
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
[
14] a) S. Nassreddine, L. Massin, M. Aouine, C. Geantet, L. Piccolo, J. Catal. 2011, 278,
2
1
53-265; b) L. Piccolo, S. Nassreddine, M. Aouine, C. Ulhaq, C. Geantet, J. Catal. 2012, 292,
73-180.
[
15] E. Rodríguez-Castellón, J. Mérida-Robles, L. Díaz, P. Maireles-Torres, D. J. Jones, J.
Rozière, A. Jiménez-López, Appl. Catal. A 2004, 260, 9-18.
16] Y. Wang, B. Shen, L. Wang, B. Feng, J. Li, Q. Guo, Fuel Proc. Technol. 2013, 106,
[
1
41-148.
[
17] R. Hernández-Huesca, J. Mérida-Robles, P. Maireles-Torres, E. Rodríguez-Castellón, A.
Jiménez-López, J. Catal. 2001, 203, 122-132.
18] S. G. A. Ferraz, B. M. Santos, F. M. Z. Zotin, L. R. R. Araujo, J. L. Zotin, Ind. Eng.
Chem. Res. 2015, 54, 2646-2656.
19] G. C. Laredo, P. Pérez-Romo, J. Escobar, J. L. Garcia-Gutierrez,; P. M. Vega-Merino, Ind.
Eng. Chem. Res. 2017, 56, 10939-10948.
[
[
[
[
20] Y.-S. Kim, K.-S. Cho, Y.-K. Lee, J. Catal. 2017, 351, 67-78.
21] a) Y. Choi, J. Lee, J. Shin, S. Lee, D. Kim, J. K. Lee, Appl. Catal. A 2015, 492, 140-150;
b) J. Lee, Y. Choi, J. Shin, J. K. Lee, Catal. Today 2016, 265, 144-153; c) J. Shin, Y. Oh, Y.
Choi, J. Lee, J. K. Lee, Appl. Catal. A 2017, 547, 12-21.
[
22] D. P. Upare, S. Park, M. S. Kim, Y.-P. Jeon, J. Kim, D. Lee, J. Lee, H. Chang, S. Choi, W.
Choi, Y.-K. Park, C. W. Lee, J. Ind. Eng. Chem. 2017, 46, 356-363.
[
[
23] G. C. Bond, Metal-Catalyzed Reactions of Hydrocarbons, Springer, New York, 2005.
24] G. B. McVicker, M. Daage, M. S. Touvelle, C. W. Hudson, D. P. Klein, W. C. Baird, B.
R. Cook, J. G. Chen, S. Hantzer, D. E. W. Vaughan, E. S. Ellis, O. C. Feeley, J. Catal. 2002,
10, 137-148.
2
2
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
[
25] S. Oya, D. Kanno, H. Watanabe, M. Tamura, Y. Nakagawa, K. Tomishige,
ChemSusChem 2015, 8, 2472-2475.
26] Y. Nakagawa, S. Oya, D. Kanno, Y. Nakaji, M. Tamura, K. Tomishige, ChemSusChem
017, 10, 189-198.
27] Y. Nakaji, S. Oya, H. Watanabe, M. M. Watanabe, Y. Nakagawa, M. Tamura, K.
Tomishige, ChemCatChem 2017, 9, 2701-2708.
28] Y. Nakaji, Y. Nakagawa, M. Tamura, K. Tomishige, Fuel Proc. Technol. 2018, 176,
49-257.
29] (a) Y. Yoshida, Y. Arai, S. Kado, K. Kunimori, K. Tomishige, Catal. Today 2006, 115,
5-101; (b) M. Honda, A. Suzuki, B. Noorjahan, K. Fujimoto, K. Suzuki, K. Tomishige,
[
2
[
[
2
[
9
Chem. Commun. 2009, 4596-4598; (c) M. Honda, S. Kuno, S. Sonehara, K. Fujimoto, K.
Suzuki, Y. Nakagawa, K. Tomishige, ChemCatChem 2011, 3, 365-370; (d) M. Honda, M.
Tamura, Y. Nakagawa, S. Sonehara, K. Suzuki, K. Fujimoto, K. Tomishige, ChemSusChem
2
013, 6, 1341-1344.
[
[
30] R. S. Suppino, R. Landers, A. J. G. Cobo, Appl. Catal. A 2013, 452, 9-16.
31] T. Mizugaki, K. Togo, Z. Maeno, T. Mitsudome, K. Jitsukawa, K. Kaneda, Sci. Rep. 2017,
7
, 14007.
32] (a) D. W. Flaherty, E. Iglesia, J. Am. Chem. Soc. 2013, 135, 18586-18599; (b) D. W.
Flaherty, D. D. Hibbitts, E. Iglesia, J. Am. Chem. Soc. 2014, 136, 9664-9676.
[
2
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
1073 [a]
Table 1. Hydrogenolysis of various aromatic hydrocarbons over Ru(4)/CeO2
Entry Substrate W/F [gcat Conv. Selectivity [%-C]
-1
h mol ]
[%]
40
73
95
99
22
53
76
99
13
27
41
27
42
81
20
24
p-Xy EB
T
B
2
Hy
7
Methane Others
1
2
3
4
5
6
7
8
9
1
Ethylbenzene
Ethylbenzene
Ethylbenzene
Ethylbenzene
0.1
-
-
77
77
69
53
7
13
10
15
20
14
13
14
18
12
10
14
16
20
27
20
15
1
0.2
-
-
2
8
3
0.4
-
-
3
8
5
0.8
-
-
4
12
2
11
<1
2
p-Ethyltoluene 0.1
p-Ethyltoluene 0.2
p-Ethyltoluene 0.4
p-Ethyltoluene 0.8
Propylbenzene 0.2
Propylbenzene 0.4
Propylbenzene 0.8
73
72
69
54
-
3
<1
<1
1
3
9
2
1
12
19
45
44
42
23
35
48
-
2
2
<1
17
15
12
49
33
9
1
2
6
[
b]
1
15
15
12
8
10(9 )
[b]
0
-
1
16(8 )
[b]
11
-
2
18(9 )
1
1
1
1
1
2
3
4
5
6
Cumene
Cumene
Cumene
Toluene
p-Xylene
0.2
0.4
0.8
0.4
0.4
-
1
4
-
1
4
7
-
3
4
9
-
-
15
2
37
15
27
4
-
-
64
[
a] Reaction conditions: catalyst 0.1 g, substrate/N /H = 1/26/78, ambient pressure, 533 K.
2 2
p-Xy = p-xylene, EB = ethylbenzene, T = toluene, B = benzene, Hy = products with
cyclohexane ring, Others = C2-C5 hydrocarbons. [b] Selectivity to ethane.
2
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Table 2. Summary of characterization data
[e]
Catalyst
BET surface dXRD
dTEM
Chemisorption Catalytic reaction
2
-1
[f]
area [m g ] [nm]
[nm]
H/Ru
0.89
1.0
Conv. [%] Select. [%]
8
8
73
73
[a]
[c]
Ru(4)/CeO2
84
-
-
-
-
-
-
-
95
27
19
79
22
14
20
18
14
10
4
69
66
18
62
22
13
62
68
67
65
59
[a]
[a]
[a]
[a]
[a]
[c]
Ru(1)/CeO2
85
-
873
[c]
Ru(0.5)/CeO2
83
-
0.56
0.72
0.71
0.56
0.19
0.03
0.29
0.27
0.35
1
073
Ru(4)/CeO2
Ru(1)/CeO2
42
1.8
1.1
1
073
40
1073
[d]
Ru(0.5)/CeO2
39
-
9
73
Ru(4)/ZrO2
44
6.0
15
7.7
>10
4.3
4.5
7.9
9
73
Ru(4)/MgO
146
44
9
73
[b]
Ru(4)/TiO2
-
9
73
Ru(4)/SiO2
435
116
3.6
6.9
973
Ru(4)/Al O
2
3
[
a] XRD peaks for Ru metal were too small to observe. [b] XRD peaks for Ru metal were
overlapped with support peaks. [c] Difficult to determine because of the low contrast between
support and Ru particles. [d] No Ru particles were observed. [e] Data from Figures 2 and 3.
-1
W/F = 0.4 g h mol . [f] C-based selectivity to toluene.
cat
2
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Scheme 1. Reaction route of hydrogenolysis of propylbenzene (A) and cumene (B).
2
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
Scheme 2. Reaction mechanisms of ethylbenzene (A) and p-ethyltoluene (B) reductions.
2
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
CH4
Toluene MCH
Others
100
80
60
40
20
0
ECH
Conv.
Ru
Ru*
Rh
Ir
Pd
Pt
M in M(4)/CeO 1
073
2
Figure 1. Hydrogenolysis of ethylbenzene over various noble metal/CeO catalysts. Reaction
2
-1
-1
conditions: catalyst 0.1 g, W/F = 0.4 gcat h mol (*: 0.1 g h mol ), ethylbenzene/N /H =
cat
2
2
1
/26/78, 533 K. Others = C2-C6 hydrocarbons.
2
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
MCH
CH4
100
Others
80
60
40
20
0
ECH
Toluene
Conv.
Support in Ru(4)/Support catalyst
Figure 2. Hydrogenolysis of ethylbenzene over Ru/Support catalysts. Reaction conditions:
-1
-1
catalyst 0.1 g, W/F = 0.4 g h mol (*: 0.1 g h mol ), ethylbenzene/N /H = 1/26/78, 533
cat
cat
2
2
K. Others = C2-C6 hydrocarbons.
2
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801118
CH4
Others
100
80
60
40
20
0
MCH
ECH
Conv.
Toluene
Catalysts
Figure 3. Hydrogenolysis of ethylbenzene over various Ru/CeO catalysts with different Ru
2
loading amount and pre-calcination temperature for CeO support. Reaction conditions:
2
-1
catalyst 0.1 g, W/F = 0.4 g h mol , ethylbenzene/N /H = 1/26/78, 533 K. Others = C2-C6
cat
2
2
hydrocarbons.
2
9
This article is protected by copyright. All rights reserved.