Insights into the Major Reaction Pathways of Vapor-Phase Hydrodeoxygenation of m-Cresol on a Pt/HBeta Catalyst

Article abstract of DOI:10.1002/cctc.201501232

Conversion of m-cresol was studied on a Pt/HBeta catalyst at 225-350°C and ambient hydrogen pressure. At 250°C, the reaction proceeds through two major reaction pathways: (1) direct deoxygenation to toluene (DDO path); (2) hydrogenation of m-cresol to methylcyclohexanone and methylcyclohexanol on Pt, followed by fast dehydration on Br?nsted acid sites (BAS) to methylcyclohexene, which is either hydrogenated to methylcyclohexane on Pt or ring-contracted to dimethylcyclopentanes and ethylcyclopentane on BAS (HYD path). The initial hydrogenation is the rate-determining step of the HYD path as its rate is significantly lower than those of subsequent steps. The apparent activation energy of the DDO path is 49.7 kJ mol^-1 but the activation energy is negative for the HYD path. Therefore, higher temperatures lead to the DDO path becoming the dominant path to toluene, whereas the HYD path, followed by fast equilibration to toluene, is less dominant, owing to the inhibition of the initial hydrogenation of m-cresol.

Full text of DOI:10.1002/cctc.201501232

DOI: 10.1002/cctc.201501232
Full Papers
Insights into the Major Reaction Pathways of Vapor-Phase
Hydrodeoxygenation of m-Cresol on a Pt/HBeta Catalyst
[a, b]
[b]
[a]
[a]
[a]
[a, c]
Qianqian Sun,
Xinli Zhu*
Guanyi Chen, Hua Wang, Xiao Liu, Jinyu Han, Qingfeng Ge,*
and
[
a, b]
Conversion of m-cresol was studied on a Pt/HBeta catalyst at
25–3508C and ambient hydrogen pressure. At 2508C, the re-
action proceeds through two major reaction pathways:
1) direct deoxygenation to toluene (DDO path); (2) hydrogena-
hydrogenation is the rate-determining step of the HYD path as
its rate is significantly lower than those of subsequent steps.
The apparent activation energy of the DDO path is
2
À1
(
49.7 kJmol but the activation energy is negative for the HYD
tion of m-cresol to methylcyclohexanone and methylcyclohexa-
nol on Pt, followed by fast dehydration on Brønsted acid sites
path. Therefore, higher temperatures lead to the DDO path be-
coming the dominant path to toluene, whereas the HYD path,
followed by fast equilibration to toluene, is less dominant,
owing to the inhibition of the initial hydrogenation of m-
cresol.
(
BAS) to methylcyclohexene, which is either hydrogenated to
methylcyclohexane on Pt or ring-contracted to dimethylcyclo-
pentanes and ethylcyclopentane on BAS (HYD path). The initial
Introduction
Growing attention has been paid to the conversion of lignin,
a major fraction of biomass, to valuable chemicals and
(and therefore high hydrogen consumption and low liquid fuel
yields) owing to the severe operating conditions, fast catalyst
deactivation owing to sulfur loss, as well as producing a sulfur-
containing stream, which will burden the subsequent
processes.
[
1–4]
fuels.
Depolymerization (such as fast pyrolysis) of lignin to
phenolic compounds (phenol, guaiacol, syringol, and their de-
[
5,6]
rivatives ), followed by catalytic hydrodeoxygenation repre-
sents an important approach to reduce the oxygen content in
In the last decade, many studies have been performed by
using nonsulfur catalysts under various conditions to develop
an efficient deoxygenation process and to understand the re-
[4,5]
fuel products from phenolics.
In earlier work, studies of hydrodeoxygenation of phenolics
were focused on using CoMoS and NiMoS catalysts at operat-
ing conditions similar to the industrial hydrodesulfurization
[
4]
action mechanisms. Among these catalysts, transition metals
supported on acidic supports, particularly zeolites, have been
[
7–12]
[13–22]
processes (200–4008C and high hydrogen pressures).
Two
reported as efficient deoxygenation catalysts.
In liquid-
oxygenation removal pathways have been proposed based on
phase reactions at low temperature and high pressure, it is
generally accepted that hydrodeoxygenation of phenolics fol-
lows the HYD reaction path: the metal sites catalyze the hydro-
genation of phenols to cyclohexanols, and then the zeolite
acid sites catalyze the subsequent dehydration of cyclohexa-
nols to cyclohexenes, which are finally hydrogenated to cyclo-
[7,10]
the product distributions.
One produces aromatics by
direct deoxygenation (the DDO path). The other produces cy-
clohexanes by saturating the phenyl ring, followed by deoxy-
[7,10]
genation (the HYD path).
This approach has the advantage
of using commercially available technologies, but suffers from
drawbacks including over hydrogenation and high cracking
[
13,14]
hexanes on the metal sites.
By contrast, aromatics were
found to be the predominant products for hydrodeoxygena-
tion of phenolics on similar catalysts at relatively higher tem-
[
a] Q. Sun, Dr. H. Wang, Dr. X. Liu, Prof. J. Han, Prof. Q. Ge, Dr. X. Zhu
Collaborative Innovation Center of Chemical Science and Engineering
School of Chemical Engineering and Technology
Tianjin University
Tianjin 300072 (P.R. China)
E-mail: xinlizhu@tju.edu.cn
[5,15,23–25]
peratures and lower pressures.
However, it is not yet
clear whether the difference in product distributions originates
from different reaction pathways, or from the same pathway
but are shifted owing to the thermodynamic equilibrium be-
tween cyclohexanes and aromatics, the initial products of the
reaction, as a result of the change in reaction conditions.
Owing to the delocalization effect of the orbital containing the
lone pair electrons of oxygen mixing with the p bond orbitals
of the phenyl ring, the CaromaticÀOH is strengthened, resulting in
[
b] Q. Sun, Prof. G. Chen, Dr. X. Zhu
School of Environmental Science and Engineering
State Key Laboratory of Engines
Tianjin University
Tianjin 300072 (P.R. China)
[
c] Prof. Q. Ge
Department of Chemistry and Biochemistry
Southern Illinois University
Carbondale, Illinois 62901 (United States)
E-mail: qge@chem.siu.edu
À1
a bond dissociation energy 84 kJmol higher than that of the
[26]
CaliphaticÀOH bond in aliphatic alcohol. Therefore, the direct
deoxygenation path may be less favorable compared with the
HYD path, in which the relatively weak CaliphaticÀOH bond can
ChemCatChem 2016, 8, 551 – 561
551
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
easily break and eliminate water by a dehydration reaction on
the acid sites of the zeolite. Additionally, the aromatics may be
formed by subsequent fast dehydrogenation of cyclohexenes,
which is thermodynamically favored at high temperatures.
Indeed, a recent study reported that at 2 MPa and 3008C, the
may be a result of a large fraction of Pt clusters or even single
Pt atoms being located inside the micropores of zeolite, which
cannot be easily observed in TEM, but are accessible to CO in
the chemisorption study.
The acidity properties of Pt/HBeta are important as they
hydrodeoxygenation of 4-propylphenol on Pt-Re/ZrO catalyst
may influence the reaction pathways. The NH temperature-
2
3
follows the HYD path: hydrogenation to 4-propylcyclohexanol,
followed by dehydration to 4-propylcyclohexene, and then de-
programmed desorption (TPD) profile (Figure 2A) of the bare
HBeta displays two desorption peaks at 205 and 3458C, which
[
27]
hydrogenation to propylbenzene as the final major product.
are typically assigned to NH desorption from weak acid sites
3
The purpose of this work is to gain insights into the reaction
pathways of the hydrodeoxygenation of phenolics over a Pt/
HBeta catalyst. The reaction was carried out in the vapor phase
and in the temperature range 225–3508C to bridge the gap of
the different results in the literature between low-temperature
and high-pressure reactions and high-temperature and low-
pressure reactions. m-Cresol (3-methylphenol) was selected as
a simple model compound because (1) it is in liquid form at
room temperature, which could be easily fed into the continu-
ous-flow reaction system without using any solvent; (2) it is
and strong acid sites, respectively. Introducing 1% Pt to the
HBeta resulted in the NH desorption peaks shifting to lower
3
temperatures with slightly reduced intensities. The quantified
À1
total acid densities are 0.723 and 0.695 mmolg for HBeta
and Pt/HBeta, respectively. Isopropylamine (IPA)-TPD was used
to explore the Brønsted acidic properties of Pt/HBeta. Only the
Brønsted acid sites (BAS) can catalyze the decomposition of
IPA to propylene and ammonia, although Lewis acid sites may
[
32,33]
exist, but do not play a role.
Figure 2B shows the m/z=41
profiles during IPA-TPD. The weak and broad peak at tempera-
[
28–30]
a major product from hydrodeoxygenation of guaiacol,
tures lower than 3008C is generated from a fragment of IPA,
which is the most abundant component in bio-oil; and (3) the
methyl group donates electrons to the phenyl p bond, and
[
24,31]
thus further strengthens the CaromaticÀOH bond,
making
cresol deoxygenation the most difficult step in the deoxygena-
tion of complex phenolics.
Results
Catalyst characterization
The XRD pattern of the reduced Pt/HBeta sample (data not
shown) showed a BEA framework structure without any diffrac-
tion peaks belonging to Pt, indicating that Pt was highly dis-
persed on the HBeta support. Figure 1 shows a representative
TEM image of the reduced sample. Pt particles of <4 nm uni-
formly distribute on the zeolite crystal, with a number-weight-
ed average particle size of 1.7 nm. Some Pt particles with sizes
smaller than 1 nm might reside in the micropores of HBeta. CO
chemisorption studies estimate the Pt dispersion to be 0.77,
giving an estimated Pt particle size of 1.3 nm. The difference in
the Pt particle sizes estimated from TEM and CO chemisorption
Figure 2. (A) NH
catalysts.
3
-TPD and (B) IPA-TPD profiles of HBeta and Pt/HBeta
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. TEM image of reduced Pt/HBeta catalyst.
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
552

Full Papers
whereas the strong peak at temperatures higher than 3008C is
due to propylene formed from the decomposition of IPA on
BAS. The propylene peak centered at 3508C is sharp for HBeta,
whereas it is shifted to the higher temperature of 3838C on Pt/
HBeta with a broadened peak, indicating that the acid strength
was weakened as a result of Pt loading. The measured BAS
À1
densities are 0.695 and 0.674 mmolg
for HBeta and Pt/
HBeta, respectively. The measured BAS density from IPA-TPD is
slightly lower than the total acid density from NH -TPD, which
3
implies that the major acid sites are Brønsted type. Compari-
son between Pt/HBeta and bare HBeta suggests that at least
parts of the Pt clusters are in close contact with the acid sites,
and, therefore, reduce the number of accessible acid sites and
weaken the acid strength owing to the interaction between
the acid sites and Pt clusters. It should be noted that the
molar ratio of BAS to surface Pt site is 17.
Figure 4. Major reaction path of m-cresol conversion on Pt/HBeta. Note that
acid-catalyzed cresol isomerization and transalkylation are not included in
this scheme.
libration with Tol. To test this hypothesis, 3-methylcyclohexanol
(MCHol), an important intermediate of the HYD path, was used
as a reactant. As displayed in Figure 5, MCHol is quickly dehy-
drated to methylcyclohexene (MCHene) with an intrinsic reac-
tion rate two orders of magnitude faster than that of hydro-
deoxygenation of m-cresol. As W/F is increased, the MCHene
formed initially converts to the RC products on BAS, or to
MCHane and Tol on the Pt sites through hydrogenation/dehy-
drogenation reactions. The overall intrinsic reaction rate of
MCHene conversion is one order of magnitude faster than that
of hydrodeoxygenation of m-cresol. Note that Tol is the least
favorable product from MCHene conversion at low W/F. Fur-
ther increases in W/F lead to the RC products and MCHane
converts further to the thermodynamically more favorable Tol
through the ring enlargement and dehydrogenation reactions.
When W/F is higher than 0.1 h, Tol becomes the dominant
product. Importantly, at W/F=0.5 h, MCHol produces the same
Tol yield, of approximately 83%, as that of the m-cresol feed
m-Cresol conversion at 3508C
Figure 3 illustrates the evolution of major products as a func-
tion of space time (W/F)—defined as the ratio of catalyst mass
À1
to organic feed flow rate (gcat greactant h)—at 3508C. Clearly, tol-
uene (Tol) is the major and primary product from hydrodeoxy-
genation of m-cresol. At low W/F, besides hydrodeoxygenation,
the isomerization of m-cresol to p- and o-cresol isomers (Iso-
Cr) and transalkylation of cresol isomers to phenol (Ph) and xy-
lenol isomers (Xol) on BAS also appear to be important reac-
tions. However, they eventually convert to Tol, benzene (Ben),
and xylenes (Xyl) when W/F is increased to 0.4 h. The yields of
methylcyclohexane (MCHane) and ring-contraction (RC) prod-
ucts are close to zero, and the possible intermediates of the
HYD path are not observed. This product distribution suggests
that DDO is the major reaction path (Figure 4), in agreement
(
Figure 3). At first glance, one may argue that hydrodeoxyge-
[5,24]
nation of m-cresol may also follow the HYD path with the in-
termediate products [3-methylcyclohexanone (MCHone),
MCHol, MCHene] being quickly converted and, thus, not
observable.
with the previously reported results at 4008C.
As also shown in Figure 4, the formation of toluene may
follow the HYD path, constrained by the thermodynamic equi-
Figure 3. Effect of W/F on m-cresol conversion and major products distribu-
Figure 5. Effect of W/F on methylcyclohexanol (MCHol) conversion and
major products distribution on Pt/HBeta at 3508C. Reaction conditions:
tion on Pt/HBeta at 3508C. Reaction conditions: T=3508C, P=1 atm, H
cresol=50, Time on stream (TOS)=0.5 h for each W/F.
2
/m-
T=3508C, P=1 atm, H
2
/MCHol=50, TOS=0.5 h for each W/F.
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
553
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
m-Cresol conversion at 2508C
To determine which reaction path the hydrodeoxygenation of
m-cresol follows, we lowered the reaction temperature to
2
508C. At this temperature, the conversion of m-cresol increas-
es almost linearly with W/F, and complete conversion requires
a W/F of 1 h (Figure 6A). At low W/F (Figure 6A, inset), the
yields of MCHone, MCHane, and Tol are almost the same,
whereas the yield of the RC products is rather low. Increasing
W/F leads to fast reduction of MCHone to almost zero and
a significant increase in the RC products, making the latter one
of the major products on the scale of MCHane and Tol (Fig-
ure 6B). Again, isomerization and transalkylation products
(
Ph+Iso-Cr+Xol) are observed at intermediate W/F values,
with yields being lower than those at 3508C. These intermedi-
ates are finally deoxygenated to Ben, Tol, and Xyl. This product
evolution might suggest that the reaction follows the HYD
path (Figure 4): m-cresol is first hydrogenated to MCHone and
MCHol on Pt, MCHol is then quickly dehydrated on the adja-
cent BAS to MCHene, which either hydrogenates to MCHane
or dehydrogenates to Tol on Pt with a faster rate than the ring
contraction to form the RC products on BAS.
Further analysis of the distribution of RC products (Fig-
ure 6C) shows some interesting features. At low W/F, ethylcy-
clopentane (ECP) appears to be the major product. At high W/
F, the yields of dimethylcyclopentane (DMCP) isomers increase
quickly and replace ECP as the major products. The yields of
the DMCP isomers follow the order of 1,3-DMCP>1,2-DMCP>
1
,1-DMCP.
Although Tol also appears to be one of the primary products
at low W/F, the contribution to its formation from the HYD
path, followed by fast equilibration from MCHene and MCHane
to Tol, cannot be excluded. If Tol comes mainly from the HYD
path, one would expect that the conversion of intermediate
products of the HYD path could result in the same Tol yield at
similar reaction conditions as that of m-cresol. Therefore, we
tested all the possible intermediates to explore if Tol was
mainly from the HYD path.
When MCHone is fed into the reaction (Figure 7) at low W/F,
MCHene is the first to appear in a significant amount, followed
by MCHane, whereas Tol, RC, and m-cresol (m-Cr) are almost
zero. This result indicates that the primary and major reaction
is hydrogenation of MCHone to MCHol as well as the subse-
quent fast dehydration to MCHene, followed by hydrogenation
to MCHane. The low yields of Tol and RC products at low W/F
indicate that dehydrogenation on Pt and ring contraction on
BAS are less favorable than hydrogenation of MCHene to
MCHane on Pt under the kinetic control regime of the current
reaction conditions. The yields of RC and Tol increase slowly
with increasing W/F, and they are less than 5% and 4%, re-
spectively, at W/F of 1 h. It is important to note that the Tol
yield of 4% is about eight times lower than that of 31% for m-
cresol at W/F of 1 h (Figure 3). This test indicates that MCHone
cannot produce Tol with a similarly high yield as m-cresol
under the same reaction conditions. It should be noted that
besides the hydrodeoxygenation-related reactions, other side
Figure 6. Effect of W/F on m-cresol conversion (A), major products distribu-
tion (B), and RC products distribution (C) on Pt/HBeta at 2508C. Inset: major
product yields at low W/F. Reaction conditions: T=2508C, P=1 atm, H /m-
2
cresol=50, TOS=0.5 h for each W/F.
[34,35]
reactions (such as aldol condensation
) catalyzed by the
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
554
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
this conclusion, we increased W/F to 2 h, but the yield of Tol
remains at 9%.
To map out the major reaction pathways, MCHene, MCHane
and Tol were fed into the reaction in turn. For the MCHene
feed (Figure 9A), MCHane is the primary and major product,
while RC products and Tol are minor products. The yield of
MCHane is at least six times more than that of Tol. This further
confirms that the hydrogenation of MCHene to MCHane on Pt
is significantly faster than that of dehydrogenation to Tol on Pt
or ring contraction to RC products on BAS, as was observed
when MCHone and MCHol were used as the feeds. It should
be noted that the isomerization of MCHene (double bond
shift) is also notable but has been treated as unconverted
MCHene (Figure 9A). For the MCHane feed (Figure 9B), the RC
products and Tol are observed, with the ring contraction being
more favorable over the dehydrogenation to Tol. When Tol is
fed into the reaction (Figure 9C), hydrogenation of Tol to
MCHane is more favorable than the ring contraction to the RC
products at low W/F. When W/F is high, MCHane formed in the
reaction converts further to the RC products. These results
demonstrate that the interconversion of deoxygenated prod-
ucts of MCHane, MCHene, and Tol to the RC products on Pt/
HBeta (Figure 4) complicates the reaction pathway.
Figure 7. Effect of W/F on methylcyclohexanone (MCHone) conversion and
major products distribution on Pt/HBeta at 2508C. Reaction conditions:
T=2508C, P=1 atm, H /MCHone=50, TOS=0.5 h for each W/F.
2
bare acid sites can lead to a significant amount of other bicy-
clic products (Figure 7).
It is interesting to note that MCHene was observed when
the MCHone and MCHol were the feeds but not when the
feeds were MCHane or Tol. This difference can be attributed to
the different active sites for the formation of MCHene. For
MCHone and MCHol, MCHene is formed through dehydration
on BAS with high rates. High quantities of MCHene could
(i) desorb to the gas phase; (ii) stay on the BAS for RC reaction
with slower rates than dehydration; (iii) diffuse to the adjacent
Pt sites for hydrogenation or dehydrogenation reactions. As for
MCHane and Tol, the formation of MCHene occurs on the Pt
sites. The hydrogenation/dehydrogenation are fast, without
Figure 8 shows the product evolution as a function of W/F
for the feed of MCHol. At very low W/F, MCHene is the only
primary product from dehydration of MCHol on BAS. Increas-
ing W/F results in the hydrogenation of MCHene to MCHane
as the major path over that of ring contraction on BAS or de-
hydrogenation to Tol on Pt. At W/F=0.02 h, MCHene is com-
pletely converted to MCHane (66%), RC products (21%), and
Tol (9%). Note that the Tol yield is three times lower than that
of 31% for m-cresol at W/F=1 h, indicating that MCHol cannot
produce a similar yield of Tol to the m-cresol feed. To confirm
[
36]
MCHene being released to the gas phase. A small quantity
of surface MCHene may diffuse to the adjacent BAS for the
ring-contraction reaction.
As none of the major intermediates (MCHone, MCHol,
MCHene) of the HYD path produce Tol with as high a yield as
that of m-cresol at low W/F under kinetic control of the prod-
uct distribution or at high W/F under thermodynamic control
of the product distribution, we can conclude that two distinct
reaction pathways exist for hydrodeoxygenation of m-cresol
under the current reaction conditions. As shown in Figure 4,
one reaction path is the direct deoxygenation of m-cresol to
Tol (DDO path), the other is hydrogenation of m-cresol to
MCHone and MCHol, dehydration to MCHene, followed by fast
hydrogenation to MCHane on Pt or slow ring contraction to
RC products on BAS (HYD path). The contribution of the HYD
path to the formation of Tol is negligible at low W/F (or low
conversion) and not important at high W/F (or high conver-
sion).
Effect of reaction temperature on m-cresol conversion
Figure 8. Effect of W/F on methylcyclohexanol (MCHol) conversion and
major products distribution on Pt/HBeta at 2508C. Reaction conditions:
It is apparent that the product distributions are significantly
T=2508C, P=1 atm, H
2
/MCHol=50, TOS=0.5 h for each W/F.
different at 2508C and 3508C. To explore how the reaction
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
555
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 9. Effect of W/F on methylcyclohexene (A), methylcyclohexane (B), and toluene (C) conversion on Pt/HBeta at 2508C. Reaction conditions: T=2508C,
P=1 atm, H /reactant=50, TOS=0.5 h for each W/F.
2
temperature affects the product distribution, we further inves-
tigated the conversion of m-cresol over the temperature range
of 225–3508C using a W/F of 0.5 h. Note that this range covers
the broad reaction temperature range of studies of hydrodeox-
ygenation of phenolics in the literature. As shown in Figure 10,
creases to zero. The yields of the isomerization and transalkyla-
tion products are lower than 5% over this temperature range.
They are finally converted to Tol, Ben, and Xyl at 3508C.
m-cresol conversion increases almost linearly from 32 to 99% Discussion
with increasing temperature. MCHane and Tol are the domi-
Role of Brønsted acid sites on the reaction pathway of
m-cresol at 2508C
nant products at 2258C. Increasing the temperature causes the
Tol yield to increase monotonically, whereas the yield of
MCHane gradually decreases to zero at 3508C. The yield of the
RC products increases to a maximum at 3008C, and then de-
In previous work, it has been demonstrated that the major re-
action pathways of m-cresol conversion on a Pt/SiO catalyst at
2
2
508C and at atmospheric H pressure are the direct hydro-
2
deoxygenation to Tol and hydrogenation to MCHone and
MCHol, with the hydrogenation rate faster than that of hydro-
[
37]
deoxygenation. The HYD path stops at the hydrogenation
step because of the lack of acidity on silica to catalyze the de-
hydration of MCHol and low CÀO hydrogenolysis activity of
MCHol on Ni, Pd, and Pt catalysts. The MCHone and MCHol
originally formed have to be dehydrogenated back to m-cresol
[
37]
and then deoxygenated to Tol at high m-cresol conversions.
The reversible interconversion of m-cresol, MCHone, and
MCHol through hydrogenation and dehydrogenation on bare
metal catalysts has also been observed under similar reaction
[
25,38–40]
conditions.
However, when Pt was supported on acidic HBeta, the HYD
reaction path becomes accessible. In contrast to the Pt/SiO2
catalyst, dehydrogenation of MCHone to m-cresol is negligible
on Pt/HBeta (Figure 7). Similarly, dehydrogenation of MCHol to
MCHone and m-cresol is absent on Pt/HBeta (Figure 8). These
results indicate that the presence of acid sites on HBeta ena-
bled the dehydration reaction with a significantly higher rate
than that of the dehydrogenation reaction, making the hydro-
genation products (MCHone and MCHol) almost irreversibly
dehydrate to MCHene. Table 1 compares the intrinsic reaction
rates and turnover frequencies (TOF) of the conversions of the
Figure 10. Effect of reaction temperature on m-cresol conversion on Pt/
HBeta at a W/F value of 0.5 h. Reaction conditions: P=1 atm, H
cresol=50, TOS=0.5 h, W/F=0.5 h.
2
/m-
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
556
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
lyzed hydrogenation of m-cresol, the rate-determin-
ing step for the HYD path is the initial hydrogenation
of m-cresol to MCHol on Pt. Therefore, the selectivity
of following the DDO and HYD path is dependent on
the balance of the initial reactions between the
direct deoxygenation to Tol and hydrogenation to
MCHone and MCHol.
Table 1. Comparison of the reaction rates and turnover frequencies (TOF) for different
feeds on Pt/HBeta at 2508C.
Feed
m-cresol MCHone MCHol MCHene MCHane Tol
[
a]
catÀ1 À1
Rate [mmolg s ]
4.87
0.10
0.0014
1.93
0.04
0.0005
1239
1.84
96
2.05
0.016
48
1.22
25
0.64
[b]
À1
TOFPt [s
]
À1
[c]
TOFBAS [s
]
[
a] The reaction rate and TOF were measured at conversions less than 15%. [b] TOF of
Pt-catalyzed reactions. [c] TOF of Brønsted acid site catalyzed reactions.
Effect of the reaction temperature on the reaction
pathway of m-cresol
When the reaction temperature is increased to
3508C, thermodynamic equilibrium drives the interconversion
between the primary deoxygenation products (Tol, MCHene,
MCHane, and RC) with significantly enhanced reaction rates,
which may mask the true reaction pathway. Detailed analysis
of the product distributions from MCHol (Figure 5) and m-
cresol (Figure 3) may provide insights into the true reaction
pathway. As compared in Table 2, at W/F=0.02 h when
MCHene formed from MCHol is almost completely converted
to the final products (Figure 5), the yields of MCHane and RC
are significantly higher than those observed for the m-cresol
feed. In addition, at W/F=0.1 h when the yield of MCHane is
reduced to a low value for the MCHol feed (Figure 5), the yield
of the RC products is still ~10 times higher than that for the
feed of m-cresol at a similar W/F of 0.125 h. These differences
in product distribution between the MCHol and m-cresol feeds
at similar W/F values indicate that the reaction pathways of m-
cresol and MCHol are different. Furthermore, when both feeds
convert completely with the same yield of Tol at W/F=0.5 h,
the yields of Ben and Xyl from m-cresol are almost twice that
from MCHol. This result provides further evidence that the m-
cresol and MCHol deoxygenation reactions proceed through
different pathways. For m-cresol, the transalkylation reaction
different feeds at 2508C, measured at a conversion of less than
1
5%. The reactions initiated on Pt (direct deoxygenation and
hydrogenation/dehydrogenation) account for the TOF , calcu-
Pt
lated in terms of the number of surface Pt atoms measured by
CO chemisorption, and the reactions that took place on BAS
account for the TOFBAS, calculated on the basis of the IPA-TPD
measured acid density. The acid-catalyzed reactions include:
isomerization and transalkylation for m-cresol, condensation
for MCHone, dehydration for MCHol, and ring contraction for
MCHene. From Table 1, it is evident that the reaction rate of
dehydration of MCHol is about 2–3 orders of magnitude
higher than the other reactions. This high reaction rate can be
attributed to both the high BAS density (i.e., high BAS/Pt ratio)
and the high TOF of dehydration. Indeed, the TOF of MCHol
dehydration is ~10 times higher than the hydrodeoxygenation
and hydrogenation of m-cresol and MCHone. As a result, the
MCHol formed is quickly consumed by dehydration, making it
barely observable when m-cresol and MCHone were fed into
the reaction (Figure 6 and 7). Therefore, one could expect that
once MCHol is formed, the reaction of m-cresol would irreversi-
bly proceed through the HYD path. In addition, the intrinsic re-
action rate and TOF of MCHene conversion are ~20 times
higher than those of m-cresol and MCHone (Table 1), which
leads to a fast conversion of MCHene formed from dehydration
to MCHane preferably at a low conversion. As a consequence
of the significantly higher rates of MCHol dehydration and
MCHene conversion, the formation and consumption of MCHol
and MCHene reach a steady state and their yields are negligi-
bly low in m-cresol conversion. The fast production of MCHane
makes it appear to be one of the primary products, like Tol
and MCHone, at low m-cresol conversions (Figure 6).
[
23,24]
happens readily as a primary reaction on BAS,
and the
subsequent direct deoxygenation leads to the formation of
Ben and Xyl with high yields. In contrast, the transalkylation re-
action is not expected to occur for MCHols, but takes place for
Tol. Tol is not a kinetically favored product from MCHene (see
Figures 5, 8, and 9A). Its formation requires a high W/F value
as a result of the thermodynamic equilibration from the
MCHane and RC products (see Figure 5). This leads to lower
yields of the tertiary products of Ben and Xyl from the secon-
dary product of Tol. Thus, at 3508C, the HYD path followed by
fast thermodynamic equilibration to Tol is not important,
Owing to the fact that the dehydration of MCHol and hydro-
genation of MCHene are significantly faster than the Pt-cata-
Table 2. Comparison of the product distribution of m-cresol and methylcyclohexanol conversion on Pt/HBeta at 3508C and similar W/F values.
Feed
W/F [h]
Conversion [%]
Yield of major products [%]
RC Iso-Cr+Ph+Xol
[a]
Tol
2.56
23.6
82.2
28.2
74.0
84.6
MCHane
MCHene
Ben+Xyl
Others
m-cresol
0.02
5.2
37.8
100
100
100
0.07
0.45
0.19
22.5
4.06
0.29
0
0
0
1.3
0
0.14
2.2
0.02
0.49
9.0
0.97
2.4
0.2
1.1
7.8
3.7
5.8
8.9
0
0
.125
.5
1.46
1.07
10.7
0.07
MCHol
0.02
43.4
13.7
1.13
0
0
0
0
0
.1
.5
100
0
5.1
[
a] Sum of other products.
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
557
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
whereas the DDO path to Tol is the predominant reaction path
for m-cresol conversion.
hanced, the hydrogenation of m-cresol to MCHol, that is, the
rate-determining step of the HYD path, is significantly retard-
ed, resulting in the HYD path not playing an important role for
m-cresol conversion at 3508C.
As discussed previously, the rate-determining step for the
HYD path is the initial hydrogenation of m-cresol. Even though
the dehydration rate of MCHol is increased as the temperature
is raised, the hydrogenation step would become less favorable
It is interesting and important to analyze the activation ener-
gies of the DDO and HYD paths for the hydrodeoxygenation
of m-cresol on Pt/HBeta. As discussed above, Tol is the least fa-
vorable product, with negligible yield from MCHene following
the HYD path in the kinetic control regime (Figure 9), Tol can
therefore be approximately treated as a product exclusively
from the DDO path at low m-cresol conversions. Similarly, the
tiny amounts of Ben and Xyl can also be treated as the prod-
ucts from the direct deoxygenation of Ph and Xol. However,
MCHone, MCHol, MCHene, MCHane, and the RC products are
produced through the HYD path at low m-cresol conversions.
Figure 12 shows the Arrhenius plots between 225–3508C. The
reaction rate was measured under a m-cresol conversion of
less than 7% by adjusting W/F. It should be noted that the iso-
merization and transalkylation products were included with
the unconverted m-cresol, as these reactions are catalyzed by
[25,26,40–42]
as it is a strongly exothermic reaction,
resulting in a re-
duced contribution of the HYD pathway to the overall conver-
sion. To verify this, MCHone, tautomer of the partial hydroge-
nation product of m-cresol, was tested. Compared with
MCHone conversion at 2508C (Figure 7), a significantly differ-
ent product distribution is observed at 3508C (Figure 11). Al-
though the m-cresol and Tol yields are tiny at 2508C, they
become the major products when the reaction temperature is
increased to 3508C. If the yields of isomerization and transalky-
lation products (Iso-Cr+Ph+Xol) of m-cresol and their deoxy-
genation products (Ben+Xyl) are included, the yields of m-
cresol and Tol and their secondary products are the dominant
products in the conversion of MCHone at 3508C. In contrast,
MCHane and MCHene are the major products of MCHone con-
version at 2508C, but they become minor products at 3508C.
These results indicate that the dehydrogenation of MCHone to
m-cresol followed by direct deoxygenation to Tol is more fa-
vorable than its hydrogenation to MCHol to proceed through
the HYD path to MCHene, RC products, and MCHane. As
MCHone, which has four hydrogen atoms added to the phenyl
ring, can be easily dehydrogenated to m-cresol, one would
expect that the surface intermediates with fewer carbon atoms
being hydrogenated in the phenyl ring will be much easier to
dehydrogenate and reproduce m-cresol than hydrogenation to
MCHone and then to MCHol to proceed through the HYD
path. Therefore, although the MCHol dehydration rate is en-
[
24]
BAS without the participation of Pt. The plots yield an appar-
À1
ent activation energy (E ) for the DDO path of 49.7 kJmol ,
a
significantly lower than that obtained on Pt(111) surfaces
[
43,44]
through theoretical calculations.
This relatively low activa-
tion energy could be a result of the participation of BAS in the
Pt-catalyzed direct deoxygenation reaction, as suggested in
[
5]
previous work. We also note that the data for the HYD path
did not fit a single line. We used two segments and obtained
À1
an E value for when the temperature <2908C of À4.2 kJmol
a
À1
and À69.7 kJmol for temperatures >2908C. The negative E
a
value clearly indicates that strongly exothermic elementary
steps contribute to and dominate the hydrogenation of m-
[
25,26,40–42]
[25,26]
cresol
The significant difference in the Ea value before and after
908C can be related to the contribution of the thermodynam-
as well as the subsequent reaction steps.
2
ic control to the overall reaction, which makes hydrogenation
Figure 11. Effect of W/F on methylcyclohexanone (MCHone) conversion on
Figure 12. Arrhenius plots of m-cresol hydrodeoxygenation on Pt/HBeta at
Pt/HBeta at 3508C. Reaction conditions: T=3508C, P=1 atm, H
MCHone=50, TOS=0.5 h for each W/F.
2
/
2
225–3508C. Reaction conditions: P=1 atm, H /m-cresol=50, TOS=0.5 h for
each data point, conversion is kept under 7% by adjusting the W/F value.
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
558 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
unfavorable at higher temperatures. Indeed, Shin and Keane
showed that phenol is predominant over cyclohexanone and
cyclohexanol in the thermodynamic equilibration on a Ni cata-
deoxygenation. Because dehydration of MCHol on BAS is
always faster than hydrogenation of m-cresol on Pt, once the
initial reaction is shifted to hydrogenation, the reaction would
eventually turn to the HYD path. The results of the present
study indicate that temperature and pressure have a strong in-
fluence on the initial reaction (hydrogenation or direct deoxy-
genation) on Pt, and therefore determine the overall reaction
path.
[40]
lyst when temperature is increased to 3008C. Similarly, Push-
karev et al. also observed the Ln(turnover rate) [Ln(TOR)]
values deviated from a linear regression line on the Arrhenius
plot at higher temperatures during benzene hydrogenation on
[
45]
a Pt catalyst. They attributed this behavior to the lowered
surface coverage of hydrogen and reaction intermediates at
higher temperatures as well as thermodynamic limits that
make the forward exothermic hydrogenation unfavorable, par-
Ring contraction
ticularly when the temperature is higher than 2508C. The over-
The isomerization (including ring contraction) of alkanes on
the Pt/zeolite bifunctional catalyst is generally believed to
follow the carbenium ion mechanism: dehydrogenation of al-
kanes on Pt to olefins, the olefin is then protonated by adja-
cent BAS forming the carbenium ion, which is rearranged to
another olefin through a cyclopropyl carbenium ion transition
state followed by the olefin hydrogenation to alkane prod-
À1
all deoxygenation E is only 16.8 kJmol , possible owing to
a
a combined contribution from the two reaction pathways.
Increasing reaction temperature would favor the endother-
mic reaction of the DDO path with a positive apparent activa-
tion energy. In contrast, the hydrogenation of m-cresol be-
comes unfavorable as it is an exothermic reaction with a nega-
tive activation energy for the overall HYD path, particularly at
temperatures higher than 2908C. In addition, increasing the
temperature lowers the surface hydrogen coverage, and thus
also hinders the initial hydrogenation of m-cresol. Therefore,
the HYD path becomes less favorable whereas the DDO path
becomes the major path as the temperature is increased.
There is a minimum at 2758C in the conversion–temperature
[
46]
uct. McVicker et al. showed that MCHane ring contraction
can be used as a probe reaction to characterize the acidity of
Pt-loaded zeolites by analyzing the reaction rates and product
[
47]
distributions. In the current work, the yield of ECP at low
conversion of m-cresol is only slightly higher than the individu-
al yield of DMCP isomers. ECP converts to DMCP at higher con-
versions at 2508C, confirming the strong acidity of Pt/HBeta.
The formation rate of the RC products from MCHene is
slightly faster than that from MCHane but significantly faster
than from Tol (Figure 9). This trend is clearly related to the fea-
sibility of formation of MCHene, which is readily protonated on
BAS to form the carbenium ion for ring contraction. The fact
that RC is favored over Tol when the HYD intermediates were
fed into the reaction (Figures 7–9), but Tol is favored over the
RC products when m-cresol is fed into the reaction (Figure 6),
further supports that the formation of Tol is mainly through
the DDO pathway. The higher RC yield is apparently related to
the high yield of MCHene from dehydration of MCHol, which is
readily ring contracted on BAS along the HYD path; whereas
along the DDO path, the conversion rate of Tol to RC is slow
owing to the undetectably low concentration of MCHene in
the Tol hydrogenation (Figure 9C).
[
37]
plot for the hydrodeoxygenation of m-cresol on Pt/SiO2.
However, this was not observed on Pt/HBeta (Figure 10). It has
been interpreted that the reduction in the hydrogenation rate
of m-cresol cannot be compensated by the increase in the
rates of deoxygenation of m-cresol on Pt/SiO as the tempera-
2
ture is increased. The monotonic increase in m-cresol conver-
sion on Pt/HBeta, therefore, could be a result of the increase in
the direct deoxygenation rate being faster than the decrease
in hydrogenation of m-cresol to MCHone and MCHol. The dif-
ference in the two catalysts indicates that the BAS participate
in the Pt-catalyzed DDO of m-cresol, resulting in an enhanced
reaction rate and compensating for the decrease due to the in-
hibition of the HYD path.
Comparison with literature results at high pressures
It is interesting to note that the 25% yield of the RC prod-
ucts at 2508C is significantly higher than that observed in the
liquid-phase reactions at high pressures (typically lower than
At 200–2508C and 5 MPa hydrogen pressure using a metal/
zeolite catalyst, the HYD path is the only reaction pathway in
hydrodeoxygenation of phenolics, producing cyclohexanes as
[
48]
10%). The RC yield can be further improved, approaching
the thermodynamic limit of 50%, by increasing W/F (data not
shown). If a selective ring-opening catalyst (such as Ir) is pres-
ent, the RC products can be converted to branched alkanes
[
13,14]
the major products.
This work showed that both the HYD
and DDO paths are the major reaction pathways at the same
reaction temperature, but at atmospheric hydrogen pressure.
The different reaction pathways, and therefore the different
product distributions, are clearly only related to the difference
in the hydrogen pressures. Compared with direct deoxygena-
tion, the hydrogenation reaction consumes 3 times more hy-
drogen. Increasing the hydrogen pressure would make the re-
action more favorable toward hydrogenation than direct deox-
ygenation, as high pressure favors the molar reduction reac-
tion. Furthermore, increasing the hydrogen pressure would in-
crease the surface coverage of hydrogen on the Pt particles.
Thus, the hydrogenation reaction is more favorable than direct
[
49]
with desirable octane numbers. Note that the ring opening
[
50]
of cyclopentanes is much easier than that of cyclohexanes.
Therefore, the catalytic processes in the present study provide
a straightforward approach to produce branched alkanes with
desirable fuel properties under mild conditions. Moreover, the
product distribution can be simply adjusted by controlling the
reaction temperature. Relatively high temperatures (>3508C)
are suitable for aromatics when the aromatics are the desirable
products and the hydrogen supply is limited. On the other
hand, low reaction temperatures (<3008C) are desirable when
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
559
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
the hydrogen supply is sufficient and the aromatics are unde-
sirable. This is especially suited to current commercial applica-
tions as very low aromatic concentrations are allowed in fuel,
whereas high octane numbers are required for fuel.
electron microscopy (TEM) observation on a JEM 2010F system.
The Pt dispersion was estimated by CO chemisorption in a micro-
reactor system, equipped with gas delivery system and tempera-
ture control unit, and monitored by a Cirrus 200 mass spectrome-
ter (MKS). The catalyst sample (50 mg, 40–60 mesh) was loaded
into a quartz tube reactor (o.d.=6 mm), reduced by flowing H at
2
3
008C for 1 h, which was followed by He purging for 30 min at the
Conclusions
same temperature to remove chemisorbed H2 from the catalyst
surface. Then, CO/He pulses (5%, 100 mL) were sent onto the cata-
lyst by a six-port valve every 3 min after the temperature reached
room temperature. The pulse was stopped when no more CO ad-
sorption was observed. A CO/Pt stoichiometry of 1:1 was used to
estimate the dispersion. The acid density was quantified by tem-
perature-programmed desorption of NH (NH -TPD) and isopropyl-
amine (IPA-TPD) in the same system as the CO chemisorption. The
sample pretreatment used the same procedure as the chemisorp-
tion. The adsorption, desorption, and quantification followed the
Hydrodeoxygenation of m-cresol at 2508C under atmospheric
hydrogen pressure produces toluene, methylcyclohexanone,
and methylcyclohexane as the major products at low W/F
values, but produces toluene, methylcyclohexane, and ring-
contraction products (dimethylcyclopetanes and ethylcyclo-
pentane) as major products at high W/F values. None of the
possible intermediates (methylcyclohexanone, methylcyclohex-
anol, methylcyclohexene) of the HYD path produce a similar
toluene yield as that of m-cresol either at low W/F values,
when kinetics control the product distribution, or at high W/F
values, when thermodynamic equilibria control the product
distribution. These results indicate that two distinct major reac-
tion pathways during m-cresol hydrodeoxygenation may oper-
ate: one is the direct deoxygenation (DDO) to toluene and the
other is hydrogenation to methylcyclohexanol on Pt followed
by dehydration to methylcyclohexene on the Brønsted acid
sites (HYD) and further to the final products. Because the sub-
sequent steps of dehydration of methylcyclohexanol and con-
version of methylcyclohexene are significantly faster than
those of Pt-catalyzed direct deoxygenation and hydrogenation
of m-cresol, the rate-determining step of the HYD path is the
initial hydrogenation of m-cresol. Consequently, which of the
two pathways dominates the reaction depends on the balance
between the initial direct deoxygenation and hydrogenation
rate of m-cresol on Pt. Increasing the temperature improves
the direct deoxygenation to toluene but inhibits the hydroge-
nation of m-cresol to methylcyclohexanol, resulting in DDO as
the dominant path. The apparent activation energy of the
3
3
[32]
procedure reported in previous work.
Catalytic performance
Hydrodeoxygenation of m-cresol was investigated in a fixed-bed
quartz tube (o.d.=6 mm) reactor at atmospheric hydrogen pres-
sure. The catalyst sample of 40–60 mesh was reduced in flowing
H at 3008C for 1 h prior to the reaction, and then the temperature
2
was adjusted to the reaction temperature. Gases were controlled
by mass flow controllers, and m-cresol was fed into the reaction by
using a syringe pump (Kd Scientific), and was vaporized before en-
tering the reactor. All lines were heated to avoid any condensation
of reactants and products. The products were quantified online in
a gas chromatograph (GC 7890B, Agilent) equipped with an Inno-
wax capillary column (60 m) and a flame ionization detector (FID).
Product identification was performed with a GC-MS (Shimadzu
QP2010SE), accompanied by standard sample injection. Space time
À1
(
W/F, gcat greactant h) is defined as the ratio of catalyst mass (g) to
À1
organic feed flow rate (gh ). To follow the major products evolu-
tion, the W/F value was adjusted from 0 to 2 h by varying both the
catalyst amount and organic feed flow rate. The H /organic feed
2
molar ratio was kept at 50 in all runs. The conversion and yield are
reported in molcarbon %. To verify the major reaction path, the possi-
ble intermediates (3-methylcyclohexanone, 3-methylcyclohexanol,
À1
direct deoxygenation path is 49.7 kJmol at 225–3508C. The
present study also showed that adjusting the reaction condi-
tions provides a simple approach to control the formation of
ring-contraction products and isoalkanes with desirable fuel
properties under mild conditions.
4
-methylcyclohexene, methylcyclohexane, and toluene) were also
studied. The chemicals were purchased form Sigma–Aldrich and
used without further purification, and the purity of the gases used
was higher than 99.999%. The carbon mass balance was higher
than 95% in all runs. Both qualitative and quantitative reproduci-
bility have been tested for several selected operating conditions
and the results showed that the relative uncertainties of conver-
sion and selectivity are within Æ2%.
Experimental Section
Catalyst preparation
The catalyst support of HBeta was obtained by calcination of the
Acknowledgements
ammonium form of the Beta zeolite (Zeolyst, CP814C, Si/Al=19,
2
À1
SBET =710 m g ) at 5508C for 6 h. Pt/HBeta (1 wt%) was prepared
by incipient wetness impregnation of HBeta with an aqueous solu-
tion of H PtCl ·6H O (Tianjin Kemiou Chemical) overnight. The re-
The authors thank the support from the National Natural Science
Foundation of China (21206116 and 21373148) and the Ministry
of Education of China Program of New Century Excellent Talents
in University (NCET-12-0407).
2
6
2
sulting sample was then dried at 1208C for 12 h, and calcined at
008C for 4 h.
4
Catalyst characterization
Keywords: hydrodeoxygenation · hydrogenation · m-cresol ·
Pt/HBeta · ring contraction
The bulk structure of the catalyst sample was characterized by
powder X-ray diffraction (XRD) on a Rigaku D/max 2500 diffractom-
eter. The Pt particle size distribution was evaluated by transmission
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
560
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
1] A. J. Ragauskas, G. T. Beckham, M. J. Biddy, R. Chandra, F. Chen, M. F.
Davis, B. H. Davison, R. A. Dixon, P. Gilna, M. Keller, P. Langan, A. K.
Naskar, J. N. Saddler, T. J. Tschaplinski, G. A. Tuskan, C. E. Wyman, Science
[27] H. Ohta, B. Feng, H. Kobayashi, K. Hara, A. Fukuoka, Catal. Today 2014,
234, 139–144.
[28] S. Boonyasuwat, T. Omotoso, D. E. Resasco, S. P. Crossley, Catal. Lett.
2013, 143, 783–791.
2014, 344, 1246843.
[
[
[
[
[
2] J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius, B. M. Weckhuysen, Chem.
Rev. 2010, 110, 3552–3599.
3] C. Liu, H. Wang, A. M. Karim, J. Sun, Y. Wang, Chem. Soc. Rev. 2014, 43,
[29] S. T. Oyama, T. Onkawa, A. Takagaki, R. Kikuchi, S. Hosokai, Y. Suzuki,
K. K. Bando, Top. Catal. 2015, 58, 201–210.
[30] R. N. Olcese, M. Bettahar, D. Petitjean, B. Malaman, F. Giovanella, A.
Dufour, Appl. Catal. B 2012, 115–116, 63–73.
[31] B. S. Gevert, J. E. Otterstedt, F. E. Massoth, Appl. Catal. 1987, 31, 119–
131.
[32] X. L. Zhu, L. L. Lobban, R. G. Mallinson, D. E. Resasco, J. Catal. 2010, 271,
88–98.
7
594–7623.
4] M. Saidi, F. Samimi, D. Karimipourfard, T. Nimmanwudipong, B. C. Gates,
M. R. Rahimpour, Energy Environ. Sci. 2014, 7, 103–129.
5] X. L. Zhu, L. L. Lobban, R. G. Mallinson, D. E. Resasco, J. Catal. 2011, 281,
2
1–29.
6] G. T. Neumann, B. R. Pimentel, D. J. Rensel, J. C. Hicks, Catal. Sci. Technol.
014, 4, 3953–3963.
[33] T. J. G. Kofke, R. J. Gorte, G. T. Kokotailo, W. E. Farneth, J. Catal. 1989,
115, 265–272.
2
[
[
[
7] E. O. Odebunmi, D. F. Ollis, J. Catal. 1983, 80, 56–64.
8] H. Weigold, Fuel 1982, 61, 1021–1026.
9] E. Laurent, B. Delmon, Ind. Eng. Chem. Res. 1993, 32, 2516–2524.
[34] J. Huang, W. Long, P. K. Agrawal, C. W. Jones, J. Phys. Chem. C 2009, 113,
16702–16710.
[35] L. Brabec, J. Novakova, L. Kubelkova, Stud. Surf. Sci. Catal. 1994, 84,
1889–1896.
[36] S. C. Reyes, J. H. Sinfelt, G. J. DeMartin, J. Phys. Chem. B 2005, 109,
2421–2431.
[37] C. Chen, G. Chen, F. Yang, H. Wang, J. Han, Q. Ge, X. L. Zhu, Chem. Eng.
Sci. 2015, 135, 145–154.
[
10] F. E. Massoth, P. Politzer, M. C. Concha, J. S. Murry, J. Jakowski, J. Simons,
J. Phys. Chem. B 2006, 110, 14283–14291.
11] Y. Romero, F. Richard, S. Brunet, Appl. Catal. B 2010, 98, 213–223.
12] A. L. Jongerius, R. Jastrzebski, P. C. A. Bruijnincx, B. M. Weckhuysen, J.
Catal. 2012, 285, 315–323.
[
[
[
13] D. Y. Hong, S. J. Miller, P. K. Agrawal, C. W. Jones, Chem. Commun. 2010,
[38] L. Nie, D. E. Resasco, J. Catal. 2014, 317, 22–29.
[39] M. Dobrovolszky, Z. TØtØnyi, Z. Paµl, J. Catal. 1982, 74, 31–43.
[40] E. Shin, M. A. Keane, J. Catal. 1998, 173, 450–459.
[41] Y. Yoon, R. Rousseau, R. S. Weber, D. Mei, J. A. Lercher, J. Am. Chem. Soc.
2014, 136, 10287–10298.
4
6, 1038–1040.
[
[
14] C. Zhao, S. Kasokov, J. He, J. A. Lercher, J. Catal. 2012, 296, 12–23.
15] T. Nimmanwudipong, R. C. Runnebaum, D. E. Blockab, B. C. Gates,
Energy Fuels 2011, 25, 3417–3427.
[
[
[
[
16] M. S. Zanuttini, B. O. D. Costa, C. A. Querini, M. A. Peralta, Appl. Catal. A
[42] G. Li, J. Han, H. Wang, X. L. Zhu, Q. Ge, ACS Catal. 2015, 5, 2009–2016.
[43] J. Lu, S. Behtash, O. Mamun, A. Heyden, ACS Catal. 2015, 5, 2423–2435.
[44] K. Lee, G. H. Gu, C. A. Mullen, A. A. Boateng, D. G. Vlachos, ChemSu-
sChem 2015, 8, 315–322.
[45] V. V. Pushkarev, K. An, S. Alayoglu, S. K. Beaumont, G. A. Sormorjai, J.
Catal. 2012, 292, 64–72.
[46] W. Knaeble, R. T. Carr, E. Iglesia, J. Catal. 2014, 319, 283–296.
[47] G. B. McVicker, O. C. Feeley, J. Z. Ziemiak, D. E. W. Vaughan, K. C. Stroh-
maier, W. R. Kliewer, D. P. Leta, J. Phys. Chem. B 2005, 109, 2222–2226.
[48] C. Zhao, J. He, A. A. Lemonidou, X. Li, J. A. Lercher, J. Catal. 2011, 280,
8–16.
2
014, 482, 352–361.
17] W. Zhang, J. Chen, R. Liu, S. Wang, L. Chen, K. Li, ACS Sus. Chem. Eng.
014, 2, 683–691.
18] J. E. Peters, J. R. Carpenter, D. C. Dayton, Energy Fuels 2015, 29, 909–
16.
2
9
19] T. M. Huynh, U. Armbruster, M. Pohl, M. Schneider, J. Radnik, D. Hoang,
B. M. Q. Phan, D. A. Nguyen, A. Martin, ChemCatChem 2014, 6, 1940–
1
951.
[
[
20] L. Wang, J. Zhang, X. Yi, A. Zheng, F. Deng, C. Chen, Y. Ji, F. Liu, X.
Meng, F. Xiao, ACS Catal. 2015, 5, 2727–2734.
21] A. Ausavasukhi, Y. Huang, A. T. To, T. Sooknoi, D. E. Resasco, J. Catal.
[49] M. Santikunaporn, W. E. Alvarez, D. E. Resasco, Appl. Catal. A 2007, 325,
175–187.
2012, 290, 90–100.
[
[
22] A. F. Foster, P. T. M. Do, R. F. Lobo, Top. Catal. 2012, 55, 118–128.
23] X. L. Zhu, R. G. Mallinson, D. E. Resasco, Appl. Catal. A 2010, 379, 172–
[50] G. B. McVicker, M. Daage, M. S. Touvelle, C. W. Hudson, D. P. Klein, W. C.
Baird, Jr., B. R. Cook, J. G. Chen, S. Hantzer, D. E. Vaughan, E. S. Ellis, O. C.
Feeley, J. Catal. 2002, 210, 137–148.
1
81.
24] X. L. Zhu, L. L. Lobban, R. G. Mallinson, D. E. Resasco, Energy Fuels 2014,
8, 4104–4111.
[
2
[
[
25] E. Shin, M. A. Keane, Ind. Eng. Chem. Res. 2000, 39, 883–892.
26] E. Furimsky, Appl. Catal. A 2000, 199, 147–190.
Received: November 9, 2015
Published online on January 12, 2016
ChemCatChem 2016, 8, 551 – 561
www.chemcatchem.org
561
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim