Oxide-supported Rh catalysts for H2 generation from low-temperature ethanol steam reforming: Effects of support, Rh precursor and Rh loading on catalytic performance

Article abstract of DOI:10.1039/c5ra14608a

A variety of oxide-supported Rh catalysts prepared from frequently used oxides (γ-Al₂O₃, MgO, SiO₂, CeO₂, ZrO₂, La₂O₃, Y₂O₃, TiO₂ and ZnO) and Rh

Full text of DOI:10.1039/c5ra14608a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Oxide-supported Rh catalysts for H2 generation
from low-temperature ethanol steam reforming:
effects of support, Rh precursor and Rh loading on
catalytic performance†
Cite this: RSC Adv., 2015, 5, 99461
*
Lin Huang, Catherine Choong, Luwei Chen, Zhan Wang, Ziyi Zhong, Kee Ann Chng‡
*
and Jianyi Lin§
A variety of oxide-supported Rh catalysts prepared from frequently used oxides (g-Al2O3, MgO, SiO2, CeO2,
ZrO2, La2O3, Y2O3, TiO2 and ZnO) and Rh compounds (Rh(NO3)3$xH2O, RhCl3$xH2O, Rh(acac)3 and
Rh4(CO)12) for H2 generation from ethanol steam reforming (ESR) at low temperatures (250–400 ꢀC)
have been studied. Catalytic screening shows that Rh/CeO2 is advantageous over other frequently used
oxide-supported Rh catalysts. Rh/CeO2 behaves as the most effective catalyst for the water gas shift
(WGS) pathway during ESR. A combination of CeO2 and Rh4(CO)12 with a 1% Rh loading results in an
optimal-performance catalyst that brings about a CO-free H2 yield of 4 mol molꢁ1 C2H5OH at 350 ꢀC
with good catalytic stability. A comparison of the thermodynamic and catalytic data of ESR indicates that
low-temperature ESR is strongly kinetically controlled over Rh/CeO2 in favour of H2 production via
acetaldehyde steam reforming, acetone steam reforming, methane decomposition, steam reforming of
adsorbed CHx (x ¼ 1–3) and WGS pathways. Combined studies by catalytic stability, thermogravimetric
analysis, X-ray photoelectron spectroscopy, transmission electron microscopy and X-ray diffraction
suggest that the catalytic stability of Rh/CeO2 is markedly affected by coking which is the only cause of
catalyst deactivation during low-temperature ESR and that both CeO2-supported Rh0 and Rh+ are the
active sites for ESR to produce H2. The effects of support, Rh precursor, Rh loading and calcination of
the precatalyst on the catalytic performance are discussed. In addition, catalytic roles played by typical
oxides in reaction pathways during ESR are elucidated.
Received 23rd July 2015
Accepted 10th November 2015
DOI: 10.1039/c5ra14608a
www.rsc.org/advances
temperature ESR that meets the needs of applications, low-
temperature ESR is facing challenges of efficiency of H2
Introduction
production and formation of undesirable products. Although
thermodynamic considerations are substantial, kinetic control
of ESR in favour of a high H2 yield and low yields of undesirable
products is more important. Under identical thermodynamic
conditions, a dozen of studies mostly on high temperature ESR
clearly indicated that distinct yields of H2 and distributions of
products (H2, CH4, CO, CO2, C2H4, C2H6, CH3CHO, (CH3)2CO
and coke, etc.) could be obtained with diverse types of cata-
The ethanol steam reforming (ESR) process yields pure H2 and
CO2 only under ideal conditions, and inevitably produces CO
and CH4 as the main undesirable products under usual condi-
tions. Thermodynamically, ESR is strongly endothermic.1 High
temperatures are benecial to direct C2H5OH conversion to H2,
methane steam reforming (MSR) to H2 and decoking. At the
same time, high temperatures are detrimental to water gas shi
(WGS), reverse WGS takes off to produce CO and lower H2 yield.
On the contrary, low temperatures are favourable to WGS,
methanation and coking. Because of the high demand for low-
lysts.2–7 Furthermore, Aupretre et al. reported that in the pres-
ˆ
ence of 1% Rh/g-Al2O3 or 9.7% Ni/g-Al2O3 at 600 ꢀC, the
selectivity to H2 was obviously higher than the thermodynamic
equilibrium value while the selectivities to CO and to CH4 were
greatly lower than the thermodynamic equilibrium values.8 Roh
et al. found that in the presence of 1% Rh/ZrO2–CeO2 at 450 ꢀC,
the H2 yield signicantly surpassed the thermodynamic equi-
librium value.4 Rossi et al., who compared their thermodynamic
analyses with experimental results reported by others over
the supported metal catalysts Rh/g-Al2O3,9 Ni/MgO10 and
Ni/g-Al2O3 11–13 at 500–800 ꢀC, noticed that the selectivity values
Institute of Chemical and Engineering Sciences, Agency for Science, Technology and
Research,
1 Pesek Road, Jurong Island, Singapore 627833, Singapore. E-mail:
huang_lin@ices.a-star.edu.sg; lijy@ntu.edu.sg; Fax: +65-6316-6182
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c5ra14608a
‡ Present address: Ngee Ann Polytechnic, 535 Clementi Rd, Singapore 599489,
Singapore.
§ Present address: Energy Research Institute @ NTU, Nanyang Technological
University, 1 Cleantech Loop, Singapore 637141, Singapore.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99461

View Article Online
RSC Advances
Paper
to H2 over these catalysts were signicantly higher than those
calculated by their proposed methodology for high-temperature
ESR.14 Such results illuminate that ESR is kinetically controlled
under normal reaction conditions. Hence, it can be expected to
control the kinetics of ESR through appropriate catalysts at low
temperatures and thus increase the selectivity to H2 against
undesirable products.
Ethanol dehydrogenation to acetaldehyde (EDA):
Ethanol decomposition to acetone (EDAC):
C2H5OH / CH3CHO + H2
(1)
2C2H5OH / (CH3)2CO + CO + 3H2 (2)
Ethanol dehydration to ethylene (EDE):
C2H5OH / C2H4 + H2O
(3)
The catalytic ESR reaction for the production of H2 has been
extensively studied with supported transition metal and Cu
catalysts.2,15–17 Catalytic activity, selectivity (or product distri-
bution) and stability during ESR are virtually dependent on the
nature of the metal and support and the metal–support inter-
action. Towards H2 production from steam reforming reactions
with supported metal catalysts, it was admitted that the
supports more or less play a real catalytic role18–22 whereas the
metals may function as promoters of supports.6 On the other
hand, Homs's group demonstrated that the mixture of Co0 and
CoO derived from bulk Co3O4 is active and selective for H2
production from ESR,23,24 which hinted that the Co metal alone
may be the true catalyst. Very recently, the evidence that the
metals alone can independently fulll H2 production from ESR
was reported by Divins et al., working with Rh–Pd bimetallic
nanoparticles.25 They meanwhile demonstrated with CeO2-
supported Rh–Pd bimetallic nanoparticles that the support can
play not only an important catalytic role through new active
sites but also a pivotal promotional role via the metal–support
interaction in ESR.25 The new understanding of the activation of
C2H5OH and H2O on the surfaces of metal particles and
supports has been summarized by several recent reviews.26–30
The support, nature of the metal, metal particle size and metal
oxidation state determine the reaction pathways. A better
understanding of the reaction mechanism and the inuence of
these factors on the reaction pathways should lead toward the
development of high performance supported metal catalysts for
ESR.
Over supported metal catalysts, ESR would follow a bifunc-
tional mechanism where C2H5OH is activated on the metal
particles while H2O is activated mainly on the supports.29 On the
metal particles, C2H5OH is adsorbed as surface ethoxy or oxa-
metallacycle, which favours the C–C bond cleavage. On the
supports, H2O is adsorbed as surface OH, which facilitates
steam reforming of different surface species. The supports may
serve to promote migration of surface OH towards the metal
particles for steam reforming and stabilization of the metal
particles. In addition, the metal–support interaction, an indis-
pensable factor determining the catalytic properties, should be
considered.29,31 It is envisioned that ESR primarily occurs in
a bifunctional manner on both the dispersed metal particles
and the support and that most of the activity is at the metal–
support interface. The catalytic properties may be dependent on
the nature of the interaction between the metal particles and
the support.
Ethanol dehydrogenation to adsorbed oxametallacycle:
C2H5OH / CH2CH2O(ads) + H2
(4)
The secondary reactions implicate further reactions of primary
products, as represented below:
Adsorbed oxametallacycle decarbonylation:
CH2CH2O(ads) / CH4 + CO
(5)
(6)
(7)
Acetaldehyde steam reforming (ASR): CH3CHO + H2O
/ 2CO + 3H2
Acetaldehyde decarbonylation (AD):
CH3CHO / CH4 + CO
Acetaldehyde oxidation to adsorbed acetate:
CH3CHO + OH(ads) / CH3COO(ads) + H2 (8)
Acetone steam reforming (ACSR):
(CH3)2CO + 2H2O / 5H2 + 3CO (9)
Adsorbed ethoxy oxidation to adsorbed acetate:
C2H5O(ads) + H2O / CH3COO(ads) + 2H2 (10)
Adsorbed acetate decomposition to methane and carbon dioxide:
CH3COO(ads) + OH(ads) / CH4 + CO2 + O(ads) (11)
WGS: CO + H2O / CO2 + H2
(12)
(13)
Reverse WGS: CO2 + H2 / CO + H2O
Carbon monoxide methanation: CO + 3H2 / CH4 + H2O (14)
Carbon dioxide methanation: CO2 + 4H2 / CH4 + 2H2O (15)
MSR: CH4 + H2O / CO + 3H2
Methane decomposition (MD): CH4 / 2H2 + C
Ethylene steam reforming (ETSR):
(16)
(17)
C2H4 + 2H2O / 2CO + 4H2 (18)
Ethylene decomposition (ED): C2H4 / 2H2 + 2C
Boudouard reaction (BR): 2CO / CO2 + C
(19)
(20)
In usual cases, reforming catalysts merely allow yielding
nite H2 with the concurrent production of CO, CH4 and CO2,
etc. If a high yield of H2 is required with the minimal formation
of CO and CH4, development of catalysts containing effective
components for the secondary reactions must be considered.
A complex ESR reaction system may consist of primary
reactions and secondary reactions. Herein, the primary reac-
tions refer to C2H5OH reactions that give primary products, as
depicted below:
99462 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Quite a number of studies have been reviewed concerning
integrated functional ESR catalysts allowing for both primary
and secondary reactions in view of activity, selectivity and
stability.2,15–17 Bicomponent metals such as Cu + Ni, Rh + Pt, Rh
+ Pd, Rh + Ni, Rh + Fe and Rh + Co are apt to not only ensure
dissociation of the C–C bond but also improve efficiency of WGS
and coke tolerance of catalysts.25,32–40 Alkaline metals such as K,
Li, Na and Ca are able to neutralize acid sites of supports and
thus reduce coke formation.32,38,41–44 Basic and redox oxides like
MgO, ZnO, V2O5, CeO2 and CeO2–ZrO2 as supports are suitable
to promote EDA and/or WGS.4,18,20,21,45,46 Use of CeO2 permits
diminution of coke formation.47–50 CeO2 and CeO2–ZrO2 have
the ability to promote CO oxidation.51–53 CeO2–ZrO2 is also
helpful to MSR to H2.54
Previous studies have shown that Rh catalysts are more
active than other metal catalysts for H2 generation from
ESR.3,8,9,22,43,55,56 Although numerous catalytic investigations of
oxide-supported Rh catalysts in ESR have been under-
taken,2–4,9,41,55–69 few papers have been published on the
comparative study of different types of oxide-supported Rh
catalysts.8,58,63,70 Moreover, almost all these catalytic data pub-
lished are restricted to ESR at high temperatures (above 400 ꢀC).
The present paper reports comparative work on various types of
oxide-supported Rh catalysts for H2 generation from ESR at
ꢀ
250–400 C. In our work, we systematically examined the cata-
lytic properties of typical oxides such as g-Al2O3, MgO, SiO2 and
CeO2, as well as Rh supported on various types of oxides
including g-Al2O3, MgO, SiO2, CeO2, ZrO2, La2O3, Y2O3, TiO2
and ZnO in low-temperature ESR, in order to illustrate the roles
of Rh and the various types of oxides in low-temperature ESR.
Through a screening of these oxide-supported Rh catalysts, an
optimal combination between Rh precursors and supports was
reached. By comparing the catalytic performances of Rh/CeO2 at
different Rh loadings, the effect of dispersion of Rh metal on H2
generation from ESR was demonstrated. To learn about the
reductive behaviour of the Rh precatalysts and the Rh–support
interaction, we performed a temperature-programmed reduc-
tion (TPR) investigation on all the Rh precatalysts studied. To
shed light on the causes of deactivation of Rh catalysts during
ESR, we conducted a series of analyses including thermogravi-
metric analysis (TGA), X-ray photoelectron spectroscopy (XPS),
transmission electron microscopy (TEM) and X-ray diffraction
(XRD) on Rh/CeO2. In addition, the inuence of strong Rh–
support interaction on H2 generation from ESR and the kinetic
control of H2 generation from ESR with Rh/CeO2 were
discussed.
Fig. 1 Catalytic properties of (a) g-Al2O3 and (b) Rh(N)/g-Al2O3 as
a function of reaction temperature.
CH3CHO (76.6% S) and C2H4 (23.3% S) were also detected. The
results suggest that EDA occurs as the major primary reaction
with the concomitant EDE as the minor primary reaction at
lower temperatures over g-Al2O3, consistent with the dominant
formation of CH3CHO in a C2H5OH oxidation reaction over
Al2O3 at 227 ꢀC reported earlier.71 From 250 to 300 ꢀC, the
selectivity to CH3CHO dramatically dropped to 14.4% with the
concurrent quick rise to 85.6% in the selectivity to C2H4 as the
ꢀ
C2H5OH conversion increased. From 300 C onwards, a rapid
increase was noted in the C2H5OH conversion, which reached
78.4% at 400 ꢀC. Meanwhile, only trace amounts of H2 were
ꢀ
detected at 400 C. This demonstrates that C2H5OH is mostly
converted to C2H4 via EDE as a primary reaction once the
C2H5OH decomposition proceeds at higher temperatures over
g-Al2O3, in agreement with reported product distributions of
ESR and C2H5OH oxidation over Al2O3 at 427–450 ꢀC.21,71,72
Above 400 ꢀC, C2H5OH was quickly fully converted and H2
gradually rose with increasing temperature. An H2 yield of
1.35 mol molꢁ1 C2H5OH was obtained at 600 ꢀC. The selectivity
to C2H4 rose to 94.5% at 500 ꢀC but fell to 87.6% at 600 ꢀC. The
selectivity to CH3CHO fell progressively to 0.6% at 600 ꢀC. From
the variations of selectivities to CH3CHO, C2H4 and H2 at
400–600 ꢀC, H2 appears to arise as a secondary product from ED
instead of AD and ETSR. Rh(N)/g-Al2O3 already had a fairly good
activity at 250 ꢀC, which enabled conversion of C2H5OH at
64.7%, concomitant with an H2 yield at 1.76 mol molꢁ1
Results and discussion
Behaviour of catalysts in ESR
Typical oxides and their respective oxide-supported Rh
catalysts. Fig. 1–4 illustrate the overall catalytic results of most
frequently used oxides (g-Al2O3, MgO, SiO2 and CeO2) and their
respective oxide-supported Rh catalysts prepared from
Rh(NO3)3$xH2O in ESR. In the case of the acidic oxide g-Al2O3
(Fig. 1), g-Al2O3 alone brought about only a C2H5OH conversion
of 0.7% and a negligible H2 yield at 250 ꢀC. At this temperature,
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99463

View Article Online
RSC Advances
Paper
Fig. 2 Catalytic properties of (a) MgO and (b) Rh(N)/MgO as a function
of reaction temperature.
Fig. 3 Catalytic properties of (a) SiO2 and (b) Rh(N)/SiO2 as a function
of reaction temperature.
C2H5OH. At the same time, CH3CHO (19.1% S), CH4 (34.8% S)
and CO (45.5% S) were produced. This shows that the Rh
obviously catalyzes not only EDA but also ASR and AD. At 300 ꢀC,
the conversions of C2H5OH and CH3CHO were complete, the
selectivities to CH4 and CO started to fall in favour of the rise in
the selectivities to H2 and CO2, evidently due to the occurrence
of MD and WGS. At 400 ꢀC, the H2 yield attained to 4.5 mol
molꢁ1 C2H5OH with the nearly complete removal of CO.
Throughout the ESR process over Rh(N)/g-Al2O3, no C2H4 was
detected, probably because of predominant EDA over EDE by
competition in the presence of the Rh.
In the case of the basic oxide MgO (Fig. 2), MgO alone did not
produce a C2H5OH conversion of 4.9% with trace amounts of H2
until 400 ꢀC. Besides, a selectivity to CH3CHO of 100% at 400 ꢀC
and below indicates that only a rather slow EDA occurs as
a primary reaction over MgO at low temperatures. Above 400 ꢀC,
the C2H5OH conversion increased rapidly, the selectivity to
CH3CHO decreased progressively with the concurrent slight
increase in the selectivity to C2H4. At 600 ꢀC, 85.0% of C2H5OH
was converted mainly to CH3CHO (78.0% S), C2H4 (12.7% S),
CO2 (5.9% S), CO (1.5% S), CH4 (1.9% S) and H2 (32.1% S). This
suggests that EDA still dominates the reaction pathways over
MgO at high temperatures, in accordance with reported product
distributions of ESR and C2H5OH oxidation over MgO at 227–
450 ꢀC.21,71 The H2 yield was low, being just 1.64 mol molꢁ1
C2H5OH. Only a small part of H2 may arise from ASR followed by
WGS, since MgO is known as a catalyst for WGS.73,74 Over Rh(N)/
MgO, 36.6% of C2H5OH was already converted, concomitant
with an H2 yield of 1.02 mol molꢁ1 C2H5OH at 250 ꢀC, at which
CH3CHO (27.6% S), CH4 (31.3% S), CO (33.9% S) and CO2 (7.2%
S) were observed as well. We infer from the general evolutions of
CH3CHO, CO, CO2 and H2 that the Rh promotes WGS besides
EDA, ASR and AD. At 350 ꢀC, C2H5OH was completely converted
and CH3CHO nearly disappeared. From 300 to 350 ꢀC, the
selectivities to CH4, CO2 and H2 rose at the expense of the
selectivities to CH3CHO and CO, which may be indicative of the
occurrence of ASR, AD and WGS. At 400 ꢀC, the H2 yield attained
to 4.28 mol molꢁ1 C2H5OH. The selectivities to CO and CH4 fell
with the concomitant increase in the selectivities to CO2 and H2.
Based on the variations of selectivities to the products from 350
to 400 ꢀC, we infer that MD, WGS and BR take place to give H2,
CO2 and coke at 400 ꢀC. Throughout the ESR process over
Rh(N)/MgO, limited CH3CHO was observed, likely because ASR
and AD proceed quickly in the presence of the Rh.
In the case of the weakly acidic oxide SiO2 (Fig. 3), SiO2 alone
did not bring about a C2H5OH conversion of 3.3% until 350 ꢀC,
at which only 0.03 mol H2 per mol C2H5OH was produced
together with CH3CHO (100% S). At 250–350 ꢀC, only CH3CHO
and H2 were detected as products, suggesting that EDA takes
place as a primary reaction over weakly acidic oxides like SiO2 in
99464 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
In the case of the redox oxide CeO2 (Fig. 4), CeO2 alone gave
rise to a C2H5OH conversion of 0.1% with high selectivities to
ꢀ
H2 and CO2 at 300 C. This may imply that C2H5OH decompo-
sition occurs in low speed followed by steam reforming and
WGS in high speed at lower temperatures. At 350 ꢀC, (CH3)2CO
was produced as a primary product in a selectivity as high as
68.8% concomitant with H2 (62.7% S) and CO2 (31.1% S). This
signies that (CH3)2CO is an important reaction intermediate
and that both EDAC and WGS are accelerated at 350 ꢀC. The
C2H5OH conversion did not come to 8.4% until 400 ꢀC, at which
CH3CHO (8.0% S) and just 0.23 mol H2 per mol C2H5OH were
produced together with (CH3)2CO (38.0% S), C2H4 (38.1% S),
C2H6 (4.4% S) and CO2 (11.5% S), indicating that EDA, ASR,
EDAC, ACSR, EDE and WGS take place simultaneously. At 500
ꢀC, the C2H5OH conversion was full and the H2 yield was as high
as 2.16 mol molꢁ1 C2H5OH. In addition, (CH3)2CO (21.5% S),
C2H4 (49.3% S), C2H6 (4.6% S), CH4 (9.0% S) and CO2 (15.6% S)
were observed with the disappearance of CH3CHO, indicating
that AD and/or acetone decomposition (ACD) take place besides
ASR, ACSR, EDE and WGS. At 600 ꢀC, (CH3)2CO disappeared
with the concurrent appearance of CO and slight rise in the
selectivities to H2 and CO2. Meanwhile, the selectivity to CH4
continued to rise and the selectivity to C2H4 remained almost
unchanged. This means that while (CH3)2CO transformation
ꢀ
proceeds, WGS may slow down at 600 C. The overall observa-
tions during ESR over CeO2 postulate that EDE takes place and
becomes more important than EDA from 400 ꢀC onwards.
Possibly because of the weak acidity of CeO2,75 its activity for
EDE is much lower than that of g-Al2O3. However, the activities
of CeO2 for EDA and EDAC are fairly high despite the weak
basicity of CeO2.75 This may be explained by strong catalysis by
CeO2 for CH3CHO transformation (ASR and AD), (CH3)2CO
transformation (ACSR and ACD) and WGS, which stimulates the
chemical equilibrium shis of EDA and EDAC. The observed
higher selectivity to (CH3)2CO at 350–500 ꢀC is in accordance
with reported product distributions of ESR over CeO2 at 400–
Fig. 4 Catalytic properties of (a) CeO2 and (b) Rh(N)/CeO2 as a func-
tion of reaction temperature.
low speed at lower temperatures. Above 350 ꢀC, the selectivity to
C2H4 ascended with the concurrent fall in the selectivity to
CH3CHO, as the C2H5OH conversion ascended. At 600 ꢀC,
85.0% of C2H5OH was converted mainly to CH3CHO (50.8% S),
C2H4 (46.6% S) and H2 (10.3% S). The results indicate that both
EDA and EDE occur as primary reactions over weakly acidic
oxides like SiO2 at higher temperatures. Increasing the
temperature increases the rate of EDE relative to EDA in
competition over SiO2. Possibly due to both the weak acidity and
basicity of SiO2,75 its activity for EDE is incomparable to that of
g-Al2O3 and its activity for EDA is lower than that of MgO based
on this work. At 600 ꢀC, the H2 yield came to only 0.52 mol
molꢁ1 C2H5OH. It is thus inferred that H2 results from EDA
rather than other reaction pathways, in accordance with what
was noted by Carrero et al. in ESR over silicate SBA-15.76 Rh(N)/
SiO2 was active enough to totally convert C2H5OH with the
production of 3.83 mol H2 per mol C2H5OH at 400 ꢀC, sug-
gesting strong catalysis or promotion by the Rh. From 250 to
450 C.21,72 The obvious observation of CO2 should be an indi-
ꢀ
cation of the efficient occurrence of WGS over CeO2, since CeO2
is known as an effective catalyst for WGS.18–20 Over Rh(N)/CeO2,
40.6% of C2H5OH was converted and 1.33 mol H2 per mol
C2H5OH was produced together with CH3CHO (16.3% S),
(CH3)2CO (0.6% S), CH4 (32.3% S), CO2 (5.1% S) and CO (45.7%
ꢀ
S) at 250 C. The formation of (CH3)2CO suggests a promoting
effect of the Rh on EDAC occurring over CeO2, although Rh
alone is unable to catalyze the EDAC pathway in C2H5OH
reactions.77–79 Above 250 ꢀC, the selectivity to CO declinedꢀin
favour of the rise in the selectivities to CO2 and H2. At 350 C,
C2H5OH was fully converted and the H2 yield continued to rise.
CH3CHO, (CH3)2CO and CO vanished with the concomitant
enhancements of CH4, CO2 and H2. The results imply that the
Rh effectively promotes WGS in addition to EDA, EDCA, ASR,
AD, ACSR and ACD. At 400 ꢀC, the H2 yield went up to 4.70 mol
molꢁ1 C2H5OH. The selectivity to CH4 declined markedly
together with the appearance of small amounts of CO and
a marked increase in the selectivity to CO2. The transformation
of CH4 may go via MD and steam reforming of adsorbed CHx
(x ¼ 1–3) followed by WGS to give H2, CO and CO2. During the
ꢀ
400 C, a low selectivity to CH3CHO was accompanied by high
selectivities to CH4, CO and H2, hinting that the primary
product from EDA is rapidly transformed to the secondary
products via ASR and AD in the presence of the Rh. From 350 to
400 ꢀC, the selectivity to CO fell slightly in favour of the rise in
the selectivities to both CO2 and H2, indicating the occurrence
of WGS. During the ESR process over Rh(N)/SiO2, only trace
amounts of C2H4 were discerned, probably because EDA
dominates over EDE by competition in the presence of the Rh.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99465

View Article Online
RSC Advances
Paper
ESR process over Rh(N)/CeO2, C2H4 was not detected, probably activity for H2 production can be generally ranked as Rh(N)/
owing to predominant EDA and EDAC over EDE by competition CeO2 > Rh(N)/g-Al2O3 > Rh(N)/MgO > Rh(N)/SiO2. Rh(N)/g-Al2O3
in the presence of the Rh. Trace (CH3)2CO (S # 1.4%) was is more active than the others for EDA (or EDA plus EDAC) at all
ꢀ
detected only at 250 and 300 C, probably because of predom- the temperatures, and for CH3CHO transformation at 300–
ꢀ
inant (CH3)2CO transformation over EDAC by competition in 400 C. The order of the catalyst activity for EDA (or EDA plus
the presence of the Rh.
Overall, all these four typical oxides themselves are able to MgO > Rh(N)/SiO2, while the order for CH3CHO trans-
catalyze the ESR process to generate H2 to a different extent, formation at 300–400 ꢀC can be ranked as Rh(N)/g-Al2O3
EDAC) can be ranked as Rh(N)/g-Al2O3 > Rh(N)/CeO2 > Rh(N)/
>
though the H2 yield is very low compared to supported metal Rh(N)/CeO2 > Rh(N)/MgO > Rh(N)/SiO2. However, the residual
catalysts. The acidic oxide g-Al2O3 favours EDE followed by ED, amounts of CO and CH4 in the reforming products vary with the
the basic oxide MgO benets EDA, the weakly acidic oxide SiO2 catalyst selectivity to the secondary reactions. Over all these
catalyzes EDA to a small extent, and the redox oxide CeO2 eases catalysts, CH3CHO is transformed to CH4, CO and H2 via AD
WGS following ASR, AD, ACSR and ACD. Towards C2H5OH and ASR. Over Rh(N)/CeO2, (CH3)2CO may be converted to CH4,
decomposition over these oxides, the reaction pathways are H2, CO and CO2 via ACSR and ACD.80–82 CO is further converted
likely related to the acid–base properties of oxide surfaces. The to H2 via WGS to a different extend, whereas CH4 is transformed
rates of EDE and EDA (or EDA plus EDAC) are likely dependent to H2 via MD or steam reforming of adsorbed CHx (x ¼ 1–3) to
on the acidity and basicity of oxides, respectively. The rates of a limited extend as the high selectivities of 30.1–37.3% at 400 ꢀC
EDA and EDAC rely also on the redox property of the oxides. The are obtained. Among all these catalysts, Rh(N)/CeO2 is most
relative rate between EDE and EDA over an oxide may basically active and Rh(N)/SiO2 least active for CH4 transformation to H2.
rely on both the acidity and basicity of the oxide. The order of The S(CO2)/S(CO) value can represent the ability of Rh(N)/oxide
the catalyst activity for EDE can be ranked as g-Al2O3 > CeO2 > to catalyze WGS in ESR, and the S(CO2)/S(H2) value can reect
SiO2 > MgO, while the order for EDA (or EDA plus EDAC) can be the contribution of WGS to the total production of H2 in ESR,
ranked as CeO2 > MgO > SiO2 > g-Al2O3. The presence of Rh on assuming that BR is unimportant. In the whole temperature
these oxides leads to strong catalysis or substantial promotion range, both the S(CO2)/S(CO) and the S(CO2)/S(H2) values
for all the reaction pathways as a whole, since H2 production increase substantially as the reaction temperature increases for
from ESR over oxide-supported metal catalysts is realized all the four catalysts, signifying that both the activity for WGS
through the metal-oxide concerted action.25,29 In the presence of and the contribution of WGS to the total production of H2
Rh on these oxides, EDA (or EDA plus EDAC) dominates over increase generally with increasing temperature. It is evident
EDE as a primary reaction pathway.
that the orders of the average catalyst activity for WGS at 250–
ꢀ
ꢀ
Table 1 summarizes the catalytic properties of the four 400 C and the catalyst activity for WGS at 300 and 350 C are
Rh(N)/oxide catalysts in ESR at 250–400 ꢀC. Rh(N)/CeO2 ranked as Rh(N)/CeO2 > Rh(N)/MgO > Rh(N)/g-Al2O3 > Rh(N)/
performs better, producing a higher H2 yield than the other SiO2. This order matches with those of the average contribution
three Rh(N)/oxide catalysts at 250–400 ꢀC, except a lower H2 of WGS to the total production of H2 at 250–400 ꢀC and the
yield than Rh(N)/g-Al2O3 at 250 ꢀC. The order of the catalyst contribution of WGS to the total production of H2 at 300 and
Table 1 Properties of typical oxide-supported Rh catalysts in ESRa
Selectivity%
C2H5OH
H2 yield
Catalyst
T (ꢀC) conv%
(mol molꢁ1 C2H5OH) H2
CO2 C2H4 C2H6 CH4 CO
CH3CHO S(CO2)/S(CO) S(CO2)//S(H2)
Rh(N)/g-
Al2O3
250
300
350
400
250
300
350
400
250
300
350
400
250
300
350
400
64.7
100
100
100
36.5
53.3
100
100
32.4
48.6
86.2
100
40.9
94.1
100
100
1.76
3.37
4.05
4.50
1.02
1.56
3.56
4.28
0.90
1.40
2.86
3.83
1.35
3.60
4.18
4.70
45.3
38.3
67.5 16.4
75.0 64.5
31.3
48.8 10.6
0.6
2.8
0
0
0
0
0
1
0
0.7
0
0
0
0
0
0
0
0
0
0.1
0
0.2
0
34.8 45.5 19.1
0.01
0.05
0.33
24.6
0.21
0.30
0.92
—
0.02
0.02
0.06
0.37
0.12
0.45
—
0.01
0.05
0.24
0.86
0.16
0.22
0.47
0.88
0.02
0.02
0.06
0.27
0.10
0.30
0.88
0.87
56.1 57.7
33.1 50.5
0.5
0
32.9
2.6
0
Rh(N)/MgO
Rh(N)/SiO2
Rh(N)/CeO2
7.2
46.3 33.9 27.6
35.4 35.7 17.3
59.3 27.5 0.7
71.3 62.8
40.8 30
37.3
0.90
0
0
0
0
46.1
48.0
55.3
63.9 17.2
55.2 5.6
63.8 19.1
69.7 61.2
78.4 68.4
1.0
1.8 1.5
4.7 1.0
39.3 49.5 10.2
36.6 48.2 11.9
37.5 53.7
35.8 46.9
3.1
0
0
0
0
0
0
b
32.9 45.5 16.0
36.3 42.7
1.7
0
0
38.8
30.1
0
1.5
0
42.6
a
0.15 g of precatalyst, 1% Rh loading in precatalysts, Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, gas phase hourly space velocity (GHSV) ¼ 26 200
b
h
ꢁ1, data taken 20 min later aer initiation of reaction at each reaction temperature. Excluding trace (CH3)2CO at 250 and 300 ꢀC.
99466 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
ꢀ
350 C. On the other hand, the S(CO2)/S(CO) maximum corre- active than Rh(N)/CeO2 for CH4 transformation to H2, display-
sponds to the minimum of remaining CO in ESR, and the ing lower selectivities to CH4 at 350 and 400 ꢀC. The trans-
S(CO2)/S(H2) maximum means the maximal contribution of formation of CH4 over Rh(N)/ZrO2 is assumed to proceed via
WGS to the total production of H2 in ESR. From Table 1, the MD. The weaker ability of Rh(N)/ZrO2 (vs. Rh(N)/CeO2) to cata-
S(CO2)/S(CO) maxima for Rh(N)/SiO2, Rh(N)/g-Al2O3, Rh(N)/ lyze WGS can gain support from the S(CO2)/S(CO) and S(CO2)/
MgO and Rh(N)/CeO2 correspond to the selectivities to CO of S(H2) values at 350 ꢀC and below. This outcome rendered Rh(N)/
28.6 at 400 ꢀC, 1.5 at 400 ꢀC, 0 at 400 ꢀC and 0 at 350 ꢀC, ZrO2 to disable complete CO removal in ESR until 400 ꢀC.
respectively. The corresponding S(CO2)/S(H2) maxima (which is Rh(N)/La2O3 and Rh(N)/Y2O3 exhibited low activities for H2
inferior to 1) reach 0.27 at 400 ꢀC, 0.86 at 400 ꢀC, 0.88 at 400 ꢀC production below 350 ꢀC, because of their weak abilities in EDA,
and 0.88 at 350 ꢀC, respectively. In the end, using Rh/CeO2 and CH3CHO transformation and WGS (as indicated by the S(CO2)/
Rh(N)/MgO can completely remove CO via WGS in ESR at 350 S(CO) and S(CO2)/S(H2) values at 250 and 300 ꢀC). However,
ꢀ
and 400 C, respectively, whereas using Rh(N)/SiO2 and Rh(N)/ their activities increased from 350 ꢀC. Similar H2 yields to those
g-Al2O3 cannot. Using Rh(N)/g-Al2O3, Rh(N)/MgO and Rh(N)/ over Rh(N)/CeO2 were found at 350 and 400 ꢀC, which are
CeO2, WGS can play a major part in the production of H2 in ESR consistent with their abilities to transform C2H5OH and
at 400 ꢀC or below, whereas using Rh(N)/SiO2, WGS can play CH3CHO and catalyze WGS that increased with increasing
a minor part only. In the literature, the comparative study of Rh/ temperature (as indicated by the S(CO2)/S(CO) values). At 350
Al2O3, Rh/MgO, Rh/SiO2 and Rh/CeO2 in WGS is not available and 400 ꢀC, Rh(N)/La2O3 had similar activities to those of Rh(N)/
yet. But it has been reported that CeO2 is the most active CeO2 for WGS and CH4 transformation. CO was thoroughly
support for supported Pt catalysts in WGS at 250–400 ꢀC among removed at 350 ꢀC. The transformation of CH4 may go via MD.
Al2O3, MgO, SiO2 and CeO2.83
At 350 and 400 ꢀC, Rh(N)/Y2O3 had a lower activity for WGS than
Transition metal oxide-supported Rh catalysts. We in turn Rh(N)/CeO2 and a higher activity for CH4 transformation than
extended our catalytic screening to frequently used transition Rh(N)/CeO2. The transformation of CH4 may proceed via MD
metal oxide-supported Rh catalysts (Rh(N)/ZrO2, Rh(N)/La2O3, and steam reforming of adsorbed CHx (x ¼ 1–3). Rh(N)/TiO2 and
Rh(N)/Y2O3, Rh(N)/TiO2 and Rh(N)/ZnO) in ESR, for the sake of Rh(N)/ZnO apparently showed rather low activities for H2
demonstrating the roles of transition metal oxides in supported production in the whole temperature range, due to their rather
Rh catalysts. Table 2 shows the catalytic results of these oxide- weak abilities in EDA, CH3CHO transformation and WGS (as
ꢀ
supported Rh catalysts in ESR at 250–400 C. Rh(N)/ZrO2 per- indicated by the S(CO2)/S(CO) and S(CO2)/S(H2) values). They
formed similarly to Rh(N)/CeO2 in H2 production. Nevertheless, gave rise to only 3.84 and 2.41 mol H2 per mol C2H5OH at
Rh(N)/ZrO2 was less active than Rh(N)/CeO2 for CH3CHO 400 ꢀC, respectively. At the same time that higher selectivities to
transformation and WGS, which led to higher selectivities to CO CO were observed, lower selectivities to CH4 were achieved
at 350 ꢀC and below. On the other hand, Rh(N)/ZrO2 was more owing to a higher activity for MSR. During the ESR process over
Table 2 Properties of transition metal oxide-supported Rh catalysts in ESRa
Selectivity%
C2H5OH H2 yield
Catalyst
T (ꢀC) conv%
(mol molꢁ1 C2H5OH) H2
CO2 C2H4 C2H6 CH4 CO
CH3CHO S(CO2)/S(CO) S(CO2)//S(H2)
Rh(N)/ZrO2
250
300
350
400
53.1
89.4
1.48
2.96
4.34
4.79
0.52
1.55
4.30
4.77
0.88
2.04
4.10
4.76
0.42
0.74
1.99
3.84
0.15
0.29
1.40
2.41
46.5
55.0
72.4 57.9
79.8 71.5
29.2
43.5
71.5 62.6
79.5 68.9
35.9
49.0
72.1 33.4
79.5 67.7
24.8
29.5
47.3 10.6
68.4 33.8
0.8
3.5 0.1
0
0
0.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
27.7 41.3 30.2
0.02
0.07
7.3
0.02
0.06
0.80
0.90
0.04
0.14
0.88
0.87
0.01
0.04
0.46
0.85
0.03
0.05
0.22
0.50
0
35.9 53.5
6.9
0
0
100
100
29.6
59.2
100
100
41.1
69.1
95.6
100
28.2
41.7
70.1
93.6
12.4
23.2
57.3
80.5
0
0
0
0
0
0
0
34.2
28.5
7.9
0
—
Rh(N)/La2O3 250
1.3
6.2
13.8 14.2 70.7
23.5 26.4 43.9
0.09
0.23
—
96.0
0.01
0.05
1.1
19.8
0.10
0.14
1.1
1.5
0
0.07
0.87
1.5
300
350
400
37.4
30.2
0
0.7
0
0
Rh(N)/Y2O3
Rh(N)/TiO2
Rh(N)/ZnO
250
300
350
400
250
300
350
400
250
300
350
400
0.3
1.9 0.6
13.6 20.2 66.2
21.9 36.1 39.5
0
0
29.6 30.6
6.4
0
28.9
6.1
3.4
7.9 85.1
0.7 0.2
1.6 0.6
8.2 11.1 78.5
0
0
0
0.4
0
11.5
9.4 68.5
28.3 22.9 15.0
20.0
20.8
40.6
0
0
8.7
5.9
5.1
6.4 87.7
5.6 88.7
0
0.21
0.31
10.3 10.0 71.0
14.7 10.0 60.0
49.9 15.3
0
a
0.15 g of precatalyst, 1% Rh loading in precatalysts, Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, GHSV ¼ 26 200 hꢁ1, data taken 20 min later aer
initiation of reaction at each reaction temperature.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99467

View Article Online
RSC Advances
Paper
Rh(N)/ZnO, (CH3)2CO was not observed, possibly due to the use conversions, concomitant with lower H2 yields below 400 ꢀC.
of a GHSV as high as 26 200 hꢁ1. It is documented that The results are evidently attributed to a lower activity of Rh(Cl)/
(CH3)2CO can be produced during ESR over ZnO and that the CeO2 for CH3CHO transformation and WGS, as seen in the
selectivity to (CH3)2CO drops remarkably with increasing selectivities to CH3CHO and CO, the S(CO2)/S(CO) and S(CO2)/
GHSV.21
S(H2) values. The activity lowering is caused probably by the
Based on the catalytic results in Tables 1 and 2, the following modication of the Rh/CeO2 surface properties in the presence
can be summarized on the frequently used oxide-supported Rh of Cl.84–86 It is known that in a Rh(Cl)/CeO2 sample, all of the Cl
catalysts for low-temperature ESR.
can be retained on the CeO2 lattice via substitutioꢀn with the
(1) Rh(N)/CeO2, Rh(N)/ZrO2 and Rh(N)/g-Al2O3 are generally lattice O aer thermal treatment in H2 even at 900 C.84,87 The
advantageous in H2 production, followed by Rh(N)/La2O3 and presence of Cl in the CeO2 lattice can alter the redox behaviour
Rh(N)/Y2O3.
of CeO2 and inuence the surface chemistry of Rh/CeO2.85,86 As
(2) Typical oxide-supported Rh catalysts in Table 1 are more a result, the catalytic activity of Rh/CeO2 for ESR can be inu-
active than the transition metal oxide-supported ones in Table 2 enced. Meanwhile, it was noticed that none of the four catalysts
in CH3CHO transformation including AD and ASR.
(3) Rh(N)/CeO2 is advantageous in WGS, followed by Rh(N)/ and below in terms of the selectivity to CH4.
was active enough for ASR to convert more CH4 to H2 at 400 ꢀC
La2O3 and Rh(N)/ZrO2.
Nonetheless, Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2
(4) All these catalysts are disadvantageous in ASR and CH4 apparently perform better than Rh(Cl)/CeO2 for H2 generation
transformation. Rh(N)/ZnO, Rh(N)/TiO2, Rh(N)/Y2O3 and Rh(N)/ from ESR.
ZrO2 are slightly more active for CH4 transformation, followed
by Rh(N)/La2O3 and Rh(N)/CeO2.
CeO2-supported Rh catalysts with different Rh loadings. To
reveal the effect of dispersion of Rh metal on CeO2 on H2
Various Rh precursor-derived CeO2-supported Rh catalysts. generation from ESR, we examined the catalytic properties of
As observed above, CeO2 behaves as an advantageous support CeO2-supported Rh catalysts with different Rh loadings in ESR.
over the other oxides to produce an effective supported Rh Table 4 gives the catalytic results of CeO2-supported Rh catalysts
catalyst for H2 generation from ESR. We next studied the effect prepared from Rh(NO3)3$xH2O ꢀwith Rh loadings varying from
ꢀ
of Rh precursor on the catalytic behaviour of a CeO2-supported 0.1 to 2.8% in ESR at 250–400 C. Below 350 C, the C2H5OH
Rh catalyst in ESR. Table 3 compares the catalytic results of conversion, CH3CHO transformation, (CH3)2CO transformation
CeO2-supported Rh catalysts derived from Rh(NO3)3$xH2O, and H2 yield increased with increasing Rh loading. This rela-
RhCl3$xH2O, Rh(acac)3 and Rh4(CO)12 in ESR at 250–400 ꢀC. tionship is in line with the fact that the supported Rh can
Similar to Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2 had good promote all the reaction pathways on CeO2. However, it looks as
catalytic performances, producing higher H2 yields in the whole if the specic catalytic activity of Rh(N)/CeO2 decreases mark-
temperature range. Both of them were quite active for WGS in edly with increasing Rh loading. This suggests that the disper-
terms of S(CO2)/S(CO) and S(CO2)/S(H2). Above all, CO could be sion of Rh metal on CeO2 declines with increasing Rh loading.
fully removed at 350 and 400 ꢀC over Rh(C)/CeO2. In contrast to Although a lower Rh loading led to more efficieꢀnt H2 produc-
these three catalysts, Rh(Cl)/CeO2 led to lower C2H5OH tion, it disabled complete removal of CO at 400 C and below.
Table 3 Properties of various Rh precursor-derived CeO2-supported Rh catalysts in ESRa
Selectivity%
C2H5OH H2 yield
Catalyst
T (ꢀC) conv%
(mol molꢁ1 C2H5OH) H2
CO2 C2H4 C2H6 CH4 CO
CH3CHO S(CO2)/S(CO) S(CO2)//S(H2)
Rh(N)/CeO2
250
300
350
400
40.9
94.1
100
100
23.6
48.9
85.9
100
1.35
3.60
4.18
4.70
0.61
1.63
3.55
4.66
1.42
2.98
4.27
4.12
1.82
3.74
4.18
4.73
55.2
5.6
63.8 19.1
69.7 61.2
78.4 68.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
0.3
0
0
0
0.6
0
32.9 45.5 16.0
0.12
0.45
—
42.6
0
0.03
0.21
3.1
0.11
0.52
25.1
117
0.04
0.21
—
0.10
0.30
0.88
0.87
0
0.03
0.18
0.69
0.09
0.32
0.85
0.80
0.04
0.17
0.88
0.89
36.3 42.7
1.7
0
0
38.8
30.1
0
1.5
Rh(Cl)/CeO2 250
42.7
55.5
68.9 12.7
77.6 53.5
0
1.4
24.0 34.2 41.8
29.4 49.3 19.9
300
350
400
29.1 52.5
29.1 17.2
5.7
0.2
Rh(A)/CeO2
Rh(C)/CeO2
250
300
350
400
250
300
350
400
43.4
81.8
54.8
5.0
36.4 46.1 17.5
60.5 19.6
71.2 60.3
68.9 59.5
37.2 37.4
5.5
0
0
100
100
59.4
99.9
100
100
37.3
40.0
2.4
0.5
51.0
2.1
35.5 48.6 13.8
62.4 10.8
69.7 61.4
78.8 70.4
36.3 52.3
0
0
0
38.5
29.6
0
0
0
—
a
0.15 g of precatalyst, 1% Rh loading in precatalysts, Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, GHSV ¼ 26 200 hꢁ1, data taken 20 min later aer
initiation of reaction at each reaction temperature, excluding trace (CH3)2CO at 250 and 300 ꢀC.
99468 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Table 4 Properties of CeO2-supported Rh catalysts with different Rh loadings in ESRa
Selectivity%
Rh loading
(%)
C2H5OH H2 yield
T (ꢀC) conv% (mol molꢁ1 C2H5OH) H2
CO2 C2H4 C2H6 CH4 CO CH3CHO (CH3)2CO S(CO2)/S(CO) S(CO2)//S(H2)
0.1
0.3
1.0
2.8
250
300
350
400
250
300
350
400
250
300
350
400
250
300
350
400
30.2
46.4
83.2
0.74
40.5 4.0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
25.5 26.3 42.4
19.3 20.3 41.6
18.8 21.0 18.3
1.8
10.5
24.2
0
0.3
1.8
0
0.15
0.41
0.84
22.0
0.09
0.16
2.68
26.7
0.12
0.41
—
0.10
0.19
0.31
0.86
0.07
0.13
0.65
0.85
0.10
0.28
0.88
0.87
0.05
0.15
0.86
0.87
1.26
2.88
4.76
0.81
1.96
4.03
4.52
1.33
3.56
4.17
4.70
1.58
3.63
4.14
4.71
44.6 8.3
57.7 17.7
79.4 68.2
44.6 3.0
50.7 6.5
67.3 44.0
75.4 64.1
54.6 5.4
63.2 17.6
69.7 61.2
78.4 68.4
56.7 2.9
63.9 9.4
69.1 59.3
78.5 68.1
100
28.7 3.1
0
31.1
64.5
99.9
100
40.6
94.0
100
100
46.4
94.7
100
100
27.3 33.2 36.2
34.0 40.9 16.8
38.2 16.4 0.5
33.5 2.4
0
0
32.3 45.7 16.3
35.9 43.2 1.7
0.7
1.4
0
0
0
0.3
0
0
38.8
30.1 1.5
0
0
0
42.6
0.06
0.17
—
31.7 51.4 14.0
34.0 54.8 1.5
40.7
30.3 1.6
0
0
0
42.6
a
0.15 g of Rh3+(N)/CeO2, Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, GHSV ¼ 26 200 hꢁ1, data taken 20 min later aer initiation of reaction at each
reaction temperature.
From the S(CO2)/S(CO) values, a lower Rh loading leads to CeO2 decreased with increasing Rh loading, which further
a higher catalytic activity of Rh(N)/CeO2 for WGS at 250 ꢀC. But suggests that (CH3)2CO transformation dominates over EDAC
its catalytic activity for WGS rises slowly with increasing reac- by competition in the presence of Rh. The 1.0% Rh loading on
tion temperature, differing from that with a higher Rh loading CeO2 appears more favourable for ESR in terms of H2 produc-
which rises more rapidly with increasing reaction temperature. tion, CH4 transformation and CO removal. Rh loadings above
The variation of S(CO2)/S(H2) with Rh loading indicates that the 1% seem helpless for improving the catalytic performance of
higher the Rh loading on CeO2, the more the contribution of Rh(N)/CeO2 in ESR.
WGS to the total production of H2 increases with increasing
CeO2-supported Rh catalysts from calcined precatalysts. To
ꢀ
reaction temperature at 250–350 C. Meanwhile, a very low Rh demonstrate the effect of strong metal–support interaction on
loading (0.1%) led to a slight decrease in the selectivity to CH4 H2 generation from ESR, we looked into the catalytic properties
and thus a slight increase in the selectivity to H2 at 400 ꢀC, of CeO2-supported Rh catalysts from calcined precatalysts in
which suggests that the high Rh dispersion on CeO2 favours ESR. Table 5 presents the catalytic results of CeO2-supported Rh
accelerating MD and steam reforming of adsorbed CHx (x ¼ catalysts from Rh3+(N)/CeO2 calcined at different temperatures
1–3). It was noticed that the selectivity to (CH3)2CO over Rh(N)/ in ESR at 250–400 ꢀC. At 250 and 300 ꢀC, Rh3+(N)/CeO2 calcined at
Table 5 Properties of CeO2-supported Rh catalysts from calcined precatalysts in ESRa
Selectivity%
Calcination Reaction T C2H5OH H2 yield
T (ꢀC)
(ꢀC)
conv%
(mol molꢁ1 C2H5OH) H2
CO2 C2H4 C2H6 CH4 CO
CH3CHO S(CO2)/S(CO) S(CO2)//S(H2)
21
250
300
350
400
250
300
350
400
250
300
350
400
40.9
94.1
100
1.35
3.60
4.18
4.70
0.91
2.53
4.17
4.66
0.67
1.81
4.49
4.82
55.2
5.6
63.8 19.1
69.7 61.2
78.4 68.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
0
0
0
0
0
32.9 45.5 16.0
0.12
0.45
—
42.6
0.03
0.10
2.8
30.6
0.12
0.49
4.8
0.10
0.30
0.88
0.87
0.03
0.09
0.71
0.91
0.09
0.30
0.78
0.91
36.3 42.7
1.7
0
0
38.8
30.1
0
1.5
100
400
700
27.0
68.7
99.1
100
21.9
47.2
98.4
100
56.0
61.3
70.1 49.8
77.6 70.7
1.6
5.6
29.2 53.2 16.0
31.5 59.0
31.6 18.0
3.9
0.6
0
27.0
2.3
50.6
4.4
19.4 36.7 39.5
22.6 38.9 19.0
63.9 19.2
76.1 59.2
80.2 73.3
0.3
0
0
26.6 12.3
24.5 2.2
1.9
0
33.9
a
15 g of calcined Rh3+(N)/CeO2, 1% Rh loading in Rh3+(N)/CeO2, Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, GHSV ¼ 26 200 hꢁ1, data taken 20 min
later aer initiation of reaction at each reaction temperature, excluding trace (CH3)2CO at 250–350 ꢀC.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99469

View Article Online
RSC Advances
Paper
400 and 700 ꢀC led to signicant decreases in the C2H5OH CeO2-supported Rh catalysts at 350 ꢀC. Rh(N)/CeO2 deactivated
conversion, CH3CHO transformation and H2 production. The gradually with reaction time (Fig. 5). The C2H5OH conversion
C2H5OH conversion, CH3CHO transformation and H2 yield and H2 yield did not deplete within the initial 10 h but dropped
decreased obviously with increasing calcination temperature. to 73.2% and 3.20 mol molꢁ1 C2H5OH at 25 h. During 25 h of
Trace (CH3)2CO was detected only at 250 and 300 ꢀC. In the whole reaction, the selectivities to CO and CH3CHO increased to 18.5
temperature range, calcined Rh3+(N)/CeO2 gave rise to a nitely and 17.6% from 3.4 and 0%, respectively, while the selectivity to
decreased selectivity to CH4 with increasing calcination CH4 decreased to 18.5% from 42.8%. The drop in the H2 yield
temperature. Meanwhile, Rh3+(N)/CeO2 calcined at 400 and 700 coincided well with the decrease in the selectivity to CO2 and the
ꢀC led to an increased selectivity to CO so that CO could not be increase in the selectivity to CO, accounting for the deteriora-
thoroughly eliminated in ESR, consistent with the S(CO2)/S(CO) tion of catalytic activity for WGS. Meanwhile, the increase in the
values. Accordingly, calcination of Rh3+(N)/CeO2 is helpless for selectivity to CH3CHO was accompanied by the decrease in the
improving the catalytic performance of Rh(N)/CeO2 in H2 selectivity to CH4, indicating the catalyst deactivation for AD.
generation from ESR. Comparing Tables 4 and 5, it is noted that Rh(A)/CeO2 deactivated more rapidly than Rh(N)/CeO2 with
the catalytic behaviour of Rh(N)/CeO2 with increasing calcina- reaction time (Fig. 6). The C2H5OH conversion and H2 yield
tion temperature is similar to that of Rh(N)/CeO2 with decreasing started to decline aer 6 h, and dropped to 41.5% and 1.56 mol
Rh loading, which hints that the dispersion of Rh metal on CeO2 molꢁ1 C2H5OH at 25 h. At 25 h, the selectivities to CO, CH4 and
may increase with increasing calcination temperature.
CH3CHO were observed at 37.5, 12.2 and 29.8%, respectively. In
Catalytic stability of CeO2-supported Rh catalysts. Since Rh/ contract, Rh(C)/CeO2 displayed stable catalytic performance
CeO2 turned out to be more effective than the other oxide- during at least 24 h of reaction (Fig. 7). At 24 h, a C2H5OH
supported Rh catalysts in H2 generation from low-temperature conversion of 100% and an H2 yield of 3.96 mol molꢁ1 C2H5OH
ESR, we nally paid our attention to the catalytic stability of were obtained with the concomitant production of CH4 (39.7%
Rh/CeO2 in low-temperature ESR. Fig. 5–7 illustrate the on S) and no CO. During the 24 h reaction, all the catalytic prop-
stream reaction proles over three typical Rh precursor-derived erties including C2H5OH conversion and selectivities to all the
products remained substantially unchanged, which signi-
cantly showed
a good catalytic stability of Rh(C)/CeO2.
Fig. 5 On stream ESR profiles over Rh(N)/CeO2 at 350 ꢀC.
Fig. 7 On stream ESR profiles over Rh(C)/CeO2 at 350 ꢀC.
Fig. 6 On stream ESR profiles over Rh(A)/CeO2 at 350 ꢀC.
99470 | RSC Adv., 2015, 5, 99461–99482
Fig. 8 On stream ESR profiles over Rh(C)/CeO2 at 300 ꢀC.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Nevertheless, no satisfactory catalytic stability could be ach- of ESR applied comparable conditions to those used in our ESR
ieved at 300 ꢀC, as shown in Fig. 8. The catalyst deactivated experiments. Hence, it can provide some useful information for
rapidly. The C2H5OH conversion and H2 yield dropped from 100 our reference. As shown in Table 6, the thermodynamic calcu-
to 51.3% and from 3.95 to 1.85 mol molꢁ1 C2H5OH within 19 h. lations in equilibrium conditions predict that the C2H5OH
Kinetic control. It is of signicance to stress kinetic control conversion is complete and that the H2 yield increases with
under thermodynamically favourable conditions for the increasing reaction temperature at 300–500 ꢀC. At the same
purpose of achieving a high H2 yield against undesirable time, the selectivity to CO increases with increasing reaction
products in low-temperature ESR. Several studies by different temperature while the selectivity to CH4 decreases with
authors4,8,14 have reported that in the presence of their sup- increasing reaction temperature. The prominent increase in
ported metal catalysts at 450–800 ꢀC, the selectivity values to H2 S(CO2)/S(CO) with decreasing reaction temperature implies that
are signicantly higher than the thermodynamic equilibrium low temperatures favour WGS in thermodynamics. When using
values.4,8–13 In low-temperature ESR, the different yields of H2 our Rh(N)/CeO2 or Rh(C)/CeO2 as a catalyst, the H2 yields exceed
and selectivities to products can be obtained over the various the thermodynamic equilibrium values at 300–500 ꢀC. The value
oxides and various oxide-supported Rh catalysts, as indicated in differences increase as the reaction temperature decreases and
our catalytic results. Although these catalysts accelerate all the become salient at 300 ꢀC. This accounts for that H2 generation
related reaction pathways during ESR as a whole, the extents to from ESR is strongly controlled by kinetics in the presence of
which the reaction pathways are accelerated are different. The either Rh(N)/CeO2 or Rh(C)/CeO2 at low temperatures. But as
CeO2-supported highly dispersed Rh catalysts favour enhancing reaction temperature increases, the H2 yields gradually
the H2 yield and reducing the selectivity to CH4, however they approach the thermodynamic equilibrium values, which indi-
unfavour reducing the selectivity to CO. Rh(C)/CeO2 with a 1% cates that H2 generation from ESR tends to lose kinetic control
Rh loading acts as an appropriate catalyst to lead to complete at high temperatures. On the other hand, the selectivity to CO is
removal of CO although it is less active for CH4 transformation. much higher than the thermodynamic equilibrium value and
In the complex ESR process consisting of multiple reaction thereby the S(CO2)/S(CO) value is much lower than the ther-
pathways, the roles of catalysts should be to not just speed up modynamic equilibrium value at 300 ꢀC. The comparative
the formation of H2 but also kinetically limit the formation of selectivity to CO and S(CO2)/S(CO) values are contrary at 400 ꢀC.
undesired products. Desired reaction pathways should be Meanwhile the selectivity to CH4 is remarkably inferior to the
ꢀ
accelerated whereas undesired reaction pathways should be thermodynamic equilibrium value from 300 to 500 C. Out of
depressed. We expect a successful catalyst design to kinetically the thermodynamic predictions, our catalytic results suggest
control the relative rates of reaction pathways during low- that Rh(N)/CeO2 and Rh(C)/CeO2 promote H2 generation from
temperature ESR and thus reach a desired product distribution. ESR via ASR, ACSR and WGS at 300 ꢀC, and via ASR, ACSR, MD,
Now let us make a comparison between the catalytic steam reforming of adsorbed CHx (x ¼ 1–3) and WGS at 400 ꢀC,
performance of our typical Rh/CeO2 catalysts and the thermo- which are also reected in the different S(CO2)/S(H2) values.
dynamic analysis for low-temperature ESR. Like in almost all
the reported catalytic studies of ESR, we adopted dilution of the
reaction feed with an inert gas as carrier gas in our ESR reaction
at atmospheric pressure. Although addition of inert gas has
a positive effect on the equilibrium H2 yield at low tempera-
tures,14,88–90 only a more complete study on the thermodynamic
analysis of ESR with diluted systems has been reported by
Surface properties of catalysts
TPR pertinent to Rh precatalysts. Fig. 9 shows the TPR
patterns of various oxide-supported Rh(NO3)3 precatalysts
following 2 h of drying in vacuum (10ꢁ2 torr) at 21 C. Under
ꢀ
equivalent conditions, there was no H2 consumption below
450 ꢀC in the TPR processes of all the supports used alone. Aer
2 h of drying in vacuum at 21 ꢀC, all the Rh3+(N)/oxide samples
Hernandez and Kafarov to date.90 This thermodynamic analysis
´
Table 6 Thermodynamica and catalyticb data of ESR
Selectivity%
C2H5OH
conv%
H2 yield
T (ꢀC)
(mol molꢁ1 C2H5OH)
H2
CO
CO2
CH4
S(CO2)/S(CO)
S(CO2)//S(H2)
Thermodynamic equilibrium
Rh(N)/CeO2
300
400
500
300
400
500
300
400
500
100
100
100
94.1
100
100
99.9
100
100
2.4
4.2
5.5
3.60
4.70
5.86
3.74
4.73
5.82
34
70
92
63.8
78.4
97.7
62.4
78.8
97.0
1
4
9
42.7
1.5
9.8
52.3
0
33
51
72
19.1
68.4
86.3
10.8
70.4
88.0
66
45
19
36.3
30.1
3.8
36.3
29.6
5.0
33
13
8.0
0.45
42.6
8.7
0.21
—
0.97
0.73
0.78
0.30
0.87
0.88
0.17
0.89
0.91
Rh(C)/CeO2
6.9
12.4
a
From ref. 90, Ar : C2H5OH : H2O ¼ 49 : 1 : 9 molar ratio, Y(H2) ¼ X(C2H5OH) ꢂ S(H2) ꢂ 6. b 0.15 g of precatalyst, 1% Rh loading in precatalysts,
Ar : C2H5OH : H2O ¼ 49 : 1 : 10 molar ratio, GHSV ¼ 26 200 hꢁ1, data taken 20 min later aer initiation of reaction at each reaction temperature.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99471

View Article Online
RSC Advances
Paper
TPR, the strong Rh3+–support interaction observed in Rh3+(N)/
Y2O3, Rh3+(N)/La2O3 and Rh3+(N)/MgO may be caused by heat-
ing during the TPR experiments, which may produce surface Rh
oxides before reduction at higher temperatures. It is noteworthy
that except the fractions of surface Rh oxides in Rh3+(N)/Y2O3,
Rh3+(N)/La2O3 and Rh3+(N)/MgO, all of the Rh(NO3)3 normally
ꢀ
supported on the oxides seems to be fully reduced at 400 C.
From the above TPR results, Rh3+(N)/CeO2 has weaker Rh3+–
support interaction than the other Rh3+(N)/oxide precatalysts
studied. To reveal the effect of Rh precursor on the Rh3+–
support interaction, we next conducted further TPR experi-
ments with Rh3+(Cl)/CeO2, Rh3+(A)/CeO2 pretreated in vacuum
at 21 ꢀC, and Rh4(C)/CeO2 pretreated in air at 21 and 110 ꢀC (see
ESI†). As shown in Fig. 10, the presence of Rh inuenced little
the reduction peaks of Ce4+ so that the H spillover effect from
Rh to CeO2 could be neglected. For Rh3+(Cl)/CeO2, an intense
peak at 158 ꢀC and two weak ones around 335 and 461 ꢀC were
observed, which may be ascribed to the reduction of Rh3+ in
weak interaction with CeO2, bulk-like RhCl3 and a surface Rh
oxide formed via the strong Rh3+–CeO2 interaction, respectively.
Fig. 9 TPR profiles of (a) Rh3+(N)/CeO2; (b) Rh3+(N)/SiO2; (c) Rh3+(N)/
Al2O3; (d) Rh3+(N)/ZrO2; (e) Rh3+(N)/TiO2; (f) Rh3+(N)/ZnO; (g) Rh3+(N)/
Y2O3; (h) Rh3+(N)/La2O3 and (i) Rh3+(N)/MgO.
remained unchanged in colour. Moreover the amounts of H2
consumed in the TPR processes of all the Rh3+(N)/oxide samples
highly exceeded the stoichiometric value for the complete
reduction of Rh3+. It is thus assumed that Rh(NO3)3 on these
For Rh3+(A)/CeO2, two broad, intense peaks at 251 and 300 C
ꢀ
appeared together with a shoulder one near 431 ꢀC. It is known
that oxide-supported transition metal acetylacetonates are
stable at room temperature94 and that decomposition of
Rh(acac)3 is complete at 267 ꢀC under Ar95 and acac groups are
fully away from Rh(acac)3/TiO2 at 250 ꢀC under vacuum.96 Thus,
ꢀ
supports would not decompose aer drying at 21 C and that
the surplus H2 consumption would be due to the reduction of
NO3 (plus Ce4+ for Rh3+(N)/CeO2). As expected, the number
ꢁ
and temperature of TPR peaks of such Rh3+(N)/oxide depended
on the nature of the interaction of Rh(NO3)3 with the support
surface. The TPR pattern of Rh3+(N)/CeO2 exhibited an intense
peak at 160 ꢀC, a weak one around 252 ꢀC and a very weak one
around 380 ꢀC. The rst peak may be assigned to the reduction
ꢀ
the two peaks at 251 and 300 C may correspond to the reduc-
tion of Rh3+ in weak interaction with CeO2 and Rh3+ in bulk-like
Rh(acac)3, respectively. The reduction of acac groups in Rh3+(A)/
ꢀ
ꢀ
CeO2, if possible, may occur near 251 C. The peak at 431 C
may be due to the reduction of a surface Rh oxide formed via the
strong Rh3+–CeO2 interaction. As for Rh4(C)/CeO2, the TPR
pattern aer 5 h of exposure to air at 21 ꢀC showed only a couple
of tiny signals, which implies that the Rh0 in Rh4(C)/CeO2 is
oxidized in air at 21 ꢀC negligibly. Aer treatment in air at
110 ꢀC, the TPR pattern exhibited a strong peak at 148 ꢀC,
a medium one at 285 ꢀC and a shoulder one near 440 ꢀC. These
three peaks can be attributed to the reduction of Rh3+ in weak
ꢁ
of NO3 (ref. 40 and 91) and Rh3+ which is in weak interaction
with CeO2, the second one may be attributed to the reduction of
Rh3+ in bulk-like Rh(NO3)3, while the last one may be a result of
the shoulder peak shi to lower temperatures for the reduction
of surface Ce4+ and carbonates because of H spillover from Rh
to the support.92 Since the CeO2 used here has a very low surface
area and moreover the Rh3+–CeO2 interaction is very weak, the
observed H spillover effect is negligible (see ESI†). On Rh3+(N)/
SiO2, Rh3+(N)/g-Al2O3, Rh3+(N)/ZrO2, Rh3+(N)/TiO2 and Rh3+(N)/
ZnO, similarly, the TPR patterns gave a peak or two overlapping
ꢁ
peaks at lower temperatures due to the reduction of NO3 and
Rh3+ in weak interaction with the supports, and a peak at higher
temperatures due to the reduction of Rh3+ in bulk-like
Rh(NO3)3. On Rh3+(N)/Y2O3, Rh3+(N)/La2O3 and Rh3+(N)/MgO,
in contrast, one or two extra peaks appeared at 427–560 ꢀC
apart from the above mentioned normal ones due to supported
Rh(NO3)3. These extra peaks may be related to the strong Rh3+
–
support interaction, as described in the literature.93 For
Rh3+(N)/oxide, the relative strength of the interaction between
Rh3+ and support can be roughly evaluated by the relative Rh3+
reduction peak shi to higher temperatures. In virtue of the
TPR results in Fig. 9, the order of the strength of Rh3+–support
interaction can be ranked as Rh3+(N)/MgO > Rh3+(N)/La2O3
>
Rh3+(N)/Y2O3 > Rh3+(N)/SiO2 > Rh3+(N)/ZnO > Rh3+(N)/g-Al2O3 >
Rh3+(N)/ZrO2 > Rh3+(N)/TiO2 > Rh3+(N)/CeO2. As these Rh3+(N)/
oxide samples are not subjected to any treatments prior to
Fig. 10 TPR profiles of (a) Rh3+(N)/CeO2, (b) Rh3+(Cl)/CeO2, (c)
Rh3+(A)/CeO2, (d) Rh4(C)/CeO2 pretreated in air at 21 ꢀC and (e) Rh4(C)/
CeO2 pretreated in air at 110 ꢀC.
99472 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
interaction with CeO2, a bulk-like Rh oxide and a surface Rh even smaller Rh2O3 particles on the one hand, and to heavy
oxide formed via the strong Rh3+–CeO2 interaction, respectively. sintering of CeO2 leading to a reduced surface area of CeO2 and
It is deemed that the reduction of CO groups in Rh4(C)/CeO2 thus decreased amounts of a surface Rh oxide on the other
pretreated in air does not take place during TPR, in that oxide- hand.98 It is known that the surface area of CeO2 is almost
supported Rh carbonyls seem to fully decarbonylate below 200 unaffected by calcination below 400 ꢀC, but decreases obviously
ꢀC.97 From Fig. 10, the TRP proles of Rh3+(Cl)/CeO2, Rh3+(A)/ with increasing calcination temperature above 400 C.98 These
ꢀ
CeO2 and oxidized Rh4(C)/CeO2 displayed a weak peak at 432– TPR results suggest that higher calcination temperatures
465 C possibly due to the reduction of a surface Rh oxide in strengthen the Rh3+–CeO2 interaction in Rh3+(N)/CeO2, which
ꢀ
small amounts formed via the strong Rh3+–CeO2 interaction. may lead to the formation of smaller supported Rh particles,
This was not the case for Rh3+(N)/CeO2 and slightly oxidized coinciding with the catalytic behaviour of Rh(N)/CeO2 in ESR
Rh4(C)/CeO2. Except these small amounts of the surface Rh with increasing calcination temperature (Table 5). However, it is
oxide, all of the Rh3+ normally supported on CeO2 seems to be difficult to estimate the variation of the amounts of surface
ꢀ
fully reduced at 400 C.
Rh2O3 and an irreducible surface Rh oxide on CeO2 caused by
We nally checked the inuence of calcination on the Rh3+– the stronger Rh3+–CeO2 interaction from the TPR results. It is
CeO2 interaction by TPR, since calcination of oxide-supported unclear why the calcination at 700 ꢀC results in an increased
metal precatalysts plays an important role in determining intensity in the reduction peak of surface Rh2O3 on CeO2, as
ꢀ
catalytic performance of oxide-supported metal catalysts. TPR compared with the case of the calcination at 400 C.
experiments of Rh3+(N)/CeO2 aer calcination in air were con-
Overall, the TPR processes pertinent to the various oxide-
ducted. Fig. 11 shows the TPR patterns of Rh3+(N)/CeO2 supported Rh precatalysts studied produce supported Rh
calcined at different temperatures. The calcination pretreat- particles from supported Rh3+ in weak interaction with the
ments did not seem to alter the reductive behaviour of Ce4+ and supports, supported bulk-like Rh compounds and surface Rh
reinforce the H spillover effect from Rh to CeO2. The calcination oxides, depending on the strength of the interaction of Rh
at 110 ꢀC enabled decomposition of the supported Rh(NO3)3 to precursors with oxides. Rh3+(N)/CeO2 has weaker Rh3+–support
give the supported Rh2O3 with the evolution of NO2.39 The interaction than the other oxide-supported Rh precatalysts, so
resulting Rh2O3/CeO2 had a sharp peak at 129 ꢀC due to the that its TPR process gives rise to supported Rh particles only
reduction of Rh3+ in weak interaction with CeO2, and a very from supported Rh3+ in weak interaction with CeO2 and sup-
weak one at 218 ꢀC due to the reduction of trace amounts of ported bulk-like Rh(NO3)3. Rh3+(N)/MgO, Rh3+(N)/La2O3 and
bulk-like Rh2O3. However, the calcination at 400 ꢀC led to Rh3+(N)/Y2O3 possess strong Rh3+–support interaction, so that
ꢀ
a downward shi of the rst peak from 129 to 108 C and the their TPR processes yield supported Rh particles from surface
ꢀ
appearance of a new, broad one centred at 285 C. The down- Rh oxides besides supported bulk-like Rh compounds. Use of
ward shi appears to account for the formation of smaller Rh RhCl3$xH2O and Rh(acac)3 as precursors slightly promotes the
particles via smaller Rh2O3 particles formed via strengthened strong Rh3+–CeO2 interaction which results in the formation of
3+
ꢀ
Rh –CeO2 interaction at 400 C. The appearance of the broad small amounts of supported surface Rh oxides. Use of
peak at 285 C may implicate the emergence of stronger Rh – Rh4(CO)12 as a precursor avoids the strong Rh3+–CeO2 interac-
3+
ꢀ
CeO2 interaction for Rh3+ at the interface. When the calcination tion and thus the formation of a surface Rh oxide. With high-
temperature was raised to 700 C, the rst peak continued to temperature calcination pretreatments, the Rh3+–support
ꢀ
shi downward to 102 C and the peak at 285 C dropped in interaction in Rh3+/oxide gets efficiently strengthened and thus
ꢀ
ꢀ
intensity. The results may be attributed to further strengthened the TPR processes can produce smaller supported Rh particles
3+
ꢀ
Rh –CeO2 interaction upon calcination at 700 C resulting in from surface Rh oxides as well as supported bulk-like Rh2O3.
The nature of the interaction of metal precursor with support
also determines the catalytic properties of supported metal
catalysts in ESR. It is necessary to mention that we chose the
reduction temperature at 400 ꢀC for all precatalysts and
calcined precatalysts prior to the ESR reaction according to the
highest desired reaction temperature. This reduction tempera-
ture may be inferior to those of certain surface Rh oxides due to
the strong Rh3+–support interaction mentioned above.
However, this would not affect the actual investigation of cata-
lytic properties with the Rh active sites in low-temperature ESR.
It is reasonable to reckon that the states of the Rh active sites in
ESR rely on the reaction temperature. Even if certain surface Rh
oxides undergo higher-temperature reduction, the resulting
ꢀ
supported Rh particles, once oxidized during ESR at 400 C or
below, cannot restore their active states at the identical
temperature. Therefore, we gave up using higher reduction
temperatures for precatalysts and calcined precatalysts. The
Fig. 11 TPR profiles of Rh3+(N)/CeO2 calcined at (a) 110 ꢀC, (b) 400 ꢀC
and (c) 700 ꢀC.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99473

View Article Online
RSC Advances
Paper
catalytic contributions of these surface Rh oxides were ignored C2H5OH conversions and H2 yields in low-temperature ESR over
in this work.
Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2 (Table 3).
H2 chemisorption of fresh CeO2-supported Rh catalysts. In
TGA of spent Rh catalysts. TGA in air can be used to evaluate
Table 7 are presented the results of H2 chemisorbed on different the amounts of coke formed on solid catalysts during ESR
Rh precursor-derived fresh CeO2-supported Rh catalysts. The through combustion of carbonaceous species. In Table 8 are
apparent H/Rh value can be used to represent the dispersion of compared the amounts of coke deposited on all the spent Rh
zero-valent Rh on the support or the reduction degree of surface catalysts studied aer 5 h of ESR at 350 ꢀC calculated from TGA
Rh for a supported Rh catalyst. On Rh(N)/CeO2 obtained from data. Larger amounts of coke (10.7–29.4 mg gꢁ1 cat.) were
uncalcined Rh3+(N)/CeO2, the apparent H/Rh value decreased formed on Rh(N)/SiO2, Rh(N)/g-Al2O3, Rh(N)/MgO, Rh(N)/
from 0.61 to 0.19 with increasing Rh loading from 0.1 to 2.8%. La2O3, Rh(N)/TiO2 and Rh(N)/ZnO, while smaller amounts of
This relationship implies that increasing the Rh loading on coke (6.7 mg gꢁ1 cat. and below) were observed on Rh(N)/ZrO2,
CeO2 is susceptible to render the Rh metal less dispersed or the Rh(N)/Y2O3 and Rh/CeO2. Above all on the Rh/CeO2 catalysts,
Rh particle size enlarged, which conrms that the specic remarkably tiny amounts of coke (0.8 mg gꢁ1 cat. and below)
catalytic activity of Rh/CeO2 for the C2H5OH conversion were noted. These comparative results are consistent with the
increases with increasing dispersion of Rh metal. However, on reported stronger ability of CeO2 to suppress coke deposition in
Rh(N)/CeO2 obtained from Rh3+(N)/CeO2 calcined at 400 and ESR.47–50 The comparative amounts of coke deposited on the
ꢀ
700 C, the apparent H/Rh values were 0.24 and 0.19, respec- spent oxide-supported Rh catalysts show that Rh/CeO2 is of
tively. This may imply that the calcination of Rh3+(N)/CeO2 may a signicant advantage over the other Rh/oxide catalysts in the
lead to incomplete reduction of Rh3+ on CeO2 due to the strong resistance against coking during low-temperature ESR, which
Rh3+–CeO2 interaction, assuming that the Rh particle size can reasonably predict a better catalytic stability of Rh/CeO2
remains unchanged. Some similar situations with irreducible than that of the other Rh/oxide catalysts. The strong resistance
supported metals at given reduction temperatures are well of Rh/CeO2 against coking consists in a high oxygen storage
documented for oxide-supported metal catalysts because of capacity of CeO2.102,103
strong metal–support interaction aer calcination.99–101 The
Of all the spent Rh/CeO2 catalysts aer 5 h of ESR at 350 ꢀC,
incomplete reduction of Rh3+ actually lowers the active Rh the amount of coke formed on Rh(C)/CeO2 was not detectable,
loading on CeO2. Either the calcination or the incomplete while those on the other Rh/CeO2 catalysts attained to 1.8–2.3
reduction of Rh3+ can lead to the formation of smaller sup- mol molꢁ1 surf. Rh. For spent Rh(C)/CeO2 aer 5 h of ESR at
ported Rh particles, which is in line with the decreased C2H5OH 300 ꢀC, the amount of coke deposited went up to 3.0 mol molꢁ1
conversion, CH3CHO transformation and H2 yield in low- surf. Rh. The results apparently show that coking on Rh/CeO2 is
temperature ESR over the catalysts obtained from Rh3+(N)/ sensitive to the ESR temperature and low temperatures are
CeO2 calcined at 400 and 700 ꢀC (Table 5). On Rh/CeO2 derived favourable to coking, in conformity with thermodynamic anal-
from any of the four Rh precursors with a 1% Rh loading, the ysis of ESR.88,89 On the other hand, the results indicate that use
apparent H/Rh value was 0.28–0.31, which hints that the of Rh/CeO2 can suppress ꢀcoke deposition during ESR at
dispersion of Rh metal or the Rh particle size on CeO2 is little temperatures as low as 350 C depending on the Rh precursor
affected by the Rh precursor. This further accounts for that the and the precatalyst pretreatment. At 300 ꢀC, Rh/CeO2 has a high
poorer catalytic performance of Rh(Cl)/CeO2 for ESR arises from probability of blockage of Rh metal sites by coke deposition
the detrimental effect of Cl, regardless of the dispersion of Rh
metal on CeO2. This is also in agreement with the comparable
Table 8 Coke deposition on spent Rh catalysts after 5 h of ESR at
350 ꢀC evaluated by TGA in aira
Table 7 H2 chemisorption measurements of fresh CeO2-supported
Amount of coke formed
Rh catalystsa
Spent catalyst (mg gꢁ1 cat.) (mol molꢁ1 Rh) (mol molꢁ1 surf. Rh)
Chemisorbed
Apparent H/Rh
(atomic ratio)
H2 (mmol gꢁ1
)
Rh(N)/SiO2
Rh(N)/g-Al2O3 24.5
29.4
25.2
21.0
11.7
4.5
12.5
5.7
9.6
9.2
0.5
0.5
—
—
—
—
—
—
—
—
1.8
1.9
2.3
—
Catalyst
Rh loading%
Rh(N)/MgO
Rh(N)/ZrO2
Rh(N)/La2O3
Rh(N)/Y2O3
Rh(N)/TiO2
Rh(N)/ZnO
Rh(N)/CeO2
Rh(Cl)/CeO2
Rh(A)/CeO2
Rh(C)/CeO2
13.6
5.2
14.6
6.7
11.2
10.7
0.6
CeO2
0
0
3.0
6.9
14.1
11.5
9.4
26.3
15.4
14.7
15.1
—
Rh(N)/CeO2
Rh(N)/CeO2
Rh(N)/CeO2
Rh(N)/CeO2
Rh(N)/CeO2
Rh(N)/CeO2
Rh(Cl)/CeO2
Rh(A)/CeO2
Rh(C)/CeO2
0.1
0.3
1.0
1.0
1.0
2.8
1.1
1.0
1.0
0.61
0.47
0.29
0.24
0.19
0.19
0.28
0.30
0.31
b
c
0.6
0.8
0.7
—
0.9
n.d.b
1.1
c
Rh(C)/CeO2
3.0
a
b
Rh/CeO2 obtained from uncalcined precatalysts. Obtained from
c
Rh3+(N)/CeO2 calcined at 400 ꢀC. Obtained from Rh3+(N)/CeO2
a
1% Rhloadinginprecatalysts.b Not detected. c Aer5hofESRat300ꢀC.
calcined at 700 ꢀC.
99474 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
during ESR. At 350 ꢀC, the probability of blockage is negligible
for Rh(C)/CeO2 and still present for Rh(N)/CeO2, Rh(Cl)/CeO2
and Rh(A)/CeO2. Coking on Rh/CeO2 is in line with the catalytic
stability of Rh/CeO2 at 350 ꢀC, in correlating the amount of coke
formed with the extent of catalyst deactivation in ESR on stream
over Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2. The strong
resistance of Rh(C)/CeO2 against coking may be related to the
unique structural and surface properties of the CeO2-supported
Rh metal particles derived from Rh4(CO)12, which make
a distinct difference from those of the CeO2-supported Rh metal
particles derived from the other Rh precursors. When using
Rh(C)/CeO2 as a catalyst, ESR has turned out to be strongly
kinetically controlled in favour of suppression of coking on the
catalyst in this work. The relevant aspects need in-depth
investigation in the future.
Rh 3d XPS analysis of CeO2-supported Rh catalysts. Rh 3d
XPS was used to determine the oxidation valence of surface Rh
and the ratio among various surface Rh species with different
valences in Rh/CeO2. Fig. 12 and 13 show the Rh 3d XPS spectra
of different Rh precursor-derived CeO2-supported Rh catalysts
ꢀ
before and aer 5 h of ESR at 350 C. Before XPS experiments,
the fresh catalyst samples Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/
CeO2 were exposed to air at 21 ꢀC for 5 h. The spectra of all the
three fresh catalyst samples presented a pair of symmetric
peaks at 306.7–307.4 and 311.4–312.1 eV, which are assigned to
Rh0 via curve-tting (see Fig. 12 and Table 9).104,105 The results
clearly indicate that the surface Rh of all the three fresh cata-
lysts is zero-valent and that the CeO2-supported Rh metal
particles are stable enough not to be oxidized in air at 21 ꢀC. The
spectrum of spent Rh(N)/CeO2 displayed a pair of symmetric
peaks at 308.5 and 313.2 eV while that of spent Rh(A)/CeO2
exhibited a pair of symmetric ones at 306.9 and 311.6 eV. These
two pairs of peaks correspond to the binding energies of Rh+
and of Rh0 via curve-tting (see Fig. 13 and Table 9), respec-
tively, which postulates that the supported Rh metal particles
are oxidized to Rh+ for Rh(N)/CeO2 and remain as Rh0 for Rh(A)/
CeO2 during ESR at 350 ꢀC. For spent Rh(C)/CeO2, two pairs of
peaks can be tted out at 307.3 and 312.0, 308.7 and 313.5 by
deconvolution (see Fig. 13 and Table 9), which are attributed to
a mixture of Rh0 and Rh+. This postulates part oxidation of the
supporꢀted Rh metal particles to Rh+ for Rh(C)/CeO2 during ESR
at 350 C.
However, the Rh 3d XPS spectra of two fresh catalyst samples
obtained from Rh3+(N)/CeO2 calcined at 400 and 700 ꢀC
(Fig. 14), present two pairs of peaks that can be tted out by
deconvolution (see ESI† and Table 9), which are attributed to
a mixture of Rh0 and Rh+. This suggests that the surface Rh3+ in
H2 at 400 ꢀC is incompletely reduced to Rh0 with the formation
of a large fraction of Rh+ because of the strong Rh3+–CeO2
interaction. At the same time, the peak intensity of the catalyst
sample obtained from Rh3+(N)/CeO2 calcined at 700 ꢀC was
much lower than that at 400 ꢀC under equivalent measurement
Fig. 12 Rh 3d XPS spectra of (a) Rh(N)/CeO2, (b) Rh(A)/CeO2 and (c)
Rh(C)/CeO2. The black and red curves stand for the raw experimental
data and the sum of the Gaussian components, respectively. The peak
intensity is adjusted to facilitate the comparison among the different
samples.
Fig. 12–14, no peaks due to Rh2+ and Rh3+ other than those of
Rh0 and Rh+ can be tted out by deconvolution.
The XPS results in Fig. 14 are in accordance with those of the
catalytic evaluation, TPR and H2 chemisorption of the same
catalysts obtained by calcination of Rh3+(N)/CeO2. At this stage,
however, it is difficult to verify the assumption that the
ꢀ
conditions, assuming that the calcination at 700 C results in
a reduced fraction of surface Rh0–1. This may be caused by
a heavy sintering (ꢀdecrease in the surface area) of CeO2 upon
calcination at 700 C.97 In all the Rh 3d XPS spectra shown in
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99475

View Article Online
RSC Advances
Paper
Table 9 Valences of surface Rh in CeO2-supported Rh catalysts
estimated by XPS
Before ESR
Rh0%
Aer ESR
Catalyst
Rh+%
Rh0%
Rh+%
Rh(N)/CeO2
Rh(N)/CeO2
Rh(N)/CeO2
Rh(A)/CeO2
Rh(C)/CeO2
100
50
60
100
100
0
50
40
0
0
100
—
—
0
a
b
—
—
100
73
0
27
a
b
Obtained from Rh3+(N)/CeO2 calcined at 400 ꢀC. Obtained from
Rh3+(N)/CeO2 calcined at 700 ꢀC.
Fig. 14 Rh 3d XPS spectra of Rh(N)/CeO2 obtained from Rh3+(N)/
CeO2 calcined at (a) 400 ꢀC (b) 700 ꢀC.
support interaction (SMSI).106 CeO2-supported noble metal
catalysts are long known to exhibit SMSI effects.106,107 Possibly
because of the Rh3+–CeO2 SMSI at high temperatures, the Rh3+
is partly buried or wrapped by CeO2 particles aer the calcina-
tion like in other CeO2-supported noble metal systems,106,107
which limits the access of H2 to the supported Rh3+ in TPR and
the access of C2H5OH to the active Rh sites in ESR.
From Table 9, the supported Rh0 is fully oxidized to Rh+ for
Rh(N)/CeO2, fully retained for Rh(A)/CeO2 and partly oxidized to
+
ꢀ
Rh for Rh(C)/CeO2 aer 5 h of ESR at 350 C. Since none of
these three catalysts deactivates but all of them are highly active
instead for H2 generation from ESR when the reaction proceeds
to 5 h (Fig. 5–7), it is difficult to rule out the possibility that both
the supported Rh0 and Rh+ participate in catalysis or promotion
for H2 generation from ESR. It is unlikely that oxidation of the
supported Rh0 to Rh+ causes the catalyst deactivation on Rh/
CeO2 during ESR.
Fig. 13 Rh 3d XPS spectra after 5 h of ESR at 350 ꢀC of (a) Rh(N)/CeO2,
(b) Rh(A)/CeO2 and (c) Rh(C)/CeO2. The black and red curves stand for
the raw experimental data and the sum of the Gaussian components,
respectively. The peak intensity is adjusted to facilitate the comparison
among the different samples.
To our knowledge, the roles of Rh sites with different
valences in supported Rh catalysts in ESR remain unclear. By
contrast, the roles of Co sites with different valences in sup-
ported Co catalysts in ESR have been increasingly concerned
about in recent years.108–112 O' Shea et al. investigated the
evolution of Co3O4 under operando conditions with P-XRD, and
observed that Co3O4 was not active but the evolved material
with CoO and Co0 was active and selective in ESR.24 To identify
dispersion of Rh metal on CeO2 increases with increasing
calcination temperature by TEM. Due to the strong inuence of
the dark background of small CeO2 particles in TEM images, it
is difficult to properly measure Rh particle sizes and thus
acquire accurate Rh particle size distributions.106 On the other
hand, CeO2 has been recognized as a material for strong metal–
99476 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
the roles of Co0 and Co2+ in ESR, Tuti and Pepe investigated
bulk Co3O4 (both oxidized and reduced), Co/MgO and CoO/MgO
with TPR.108 They suggested that Co0 is mainly responsible for
ESR while Co2+ is active for EDA. Hyman and Vohs studied the
reaction of C2H5OH on metallic and oxidized Co single crystal
surfaces using TPD and XPS, to determine the dependence of
the reaction pathways on the Co oxidation state.109 They found
that the primary reaction for ethoxide species on Co0 was
decarbonylation producing CO, H2 and C, CoOx that predomi-
nantly contained Co2+ was selective for dehydrogenation of
ethoxide groups producing CH3CHO, and more highly oxidized
Co that contained both Co2+ and Co3+ was active for complete
oxidation of C2H5OH producing CO, H2 and H2O. Passos et al.
examined the effect of the balance between Co0 and Co2+ on the
catalytic activity of Co/Al2O3 for ESR using time-resolved in situ
X-ray absorption spectroscopy.110 They correlated the catalytic
performance with the ratio of Co2+/Co0, and pointed out that
control of Co2+/Co0 ratio helps maintain the balance between
C2H5OH conversion and C oxidation, resulting in stable cata-
lysts. Wang's group studied the roles of Co0 and Co2+ in ESR
over oxide-supported Co catalysts.111,112 For Co/MgO, they
observed with the aid of TPR and in situ XPS that Co0 is much
more active than Co2+ for C2H5OH conversion, C–C cleavage
and WGS, whereas C2H5OH decomposition to CH4 is dominant
on Co2+.111 For Co/ZrO2, they concluded with TPR, in situ XPS
and theoretical density functional theory calculations that Co0
is mainly responsible for ACSR, while Co2+ plays a key role in
converting C2H5OH into (CH3)2CO via dehydrogenation and
condensation/ketonization.112 All these studies clearly indicate
that both Co0 and Co2+ in supported Co catalysts act as active
sites for different reaction pathways during ESR. It appears that
an appropriate Co2+/Co0 ratio may lead to good catalytic
performance and stability in ESR. Because of the similarity
between Co and noble metal catalysts in the activity for C–C
bond cleavage and in the selectivity to H2, CO2 and CH4 in
ESR,47,113–118 different catalytic roles of Rh sites with different
valences in ESR deserve to be considered.
Fig. 15 TEM images of Rh(N)/CeO2 (a) before and (b) after 25 h of ESR
at 350 ꢀC.
TEM analysis of CeO2-supported Rh catalysts. TEM was used
to observe the microscopic morphology of CeO2-supported Rh
metal particle catalysts. Fig. 15 and 16 show the TEM pictures of
Rh(N)/CeO2 and Rh(A)/CeO2 before and aer 25 h of ESR at
ꢀ
350 C. Roughly, the particle size of CeO2 was estimated to be
6–40 nm. The small dark spots were identied as the images of
Rh metal particles on the CeO2 support by energy dispersive
X-ray analysis (EDX). It was observed that the Rh metal particles
were well dispersed on the CeO2 surface with a size of 2–5 nm in
the cases of both Rh(N)/CeO2 and Rh(A)/CeO2. Due to the strong
inuence of the CeO2 dark background, it is difficult to properly
measure the particle sizes of both Rh metal and CeO2 and thus
acquire their particle size distributions. From the comparative
particle sizes of both Rh metal and CeO2 before and aer ESR,
there were no visible particle growths occurring aer ESR for
both Rh metal and CeO2 in the cases of both Rh(N)/CeO2 and
Rh(A)/CeO2. The observations suggest that sintering of Rh metal
and CeO2 particles in Rh/CeO2 does not take place under the
ESR conditions. Thus, the possibility that sintering of Rh metal
and CeO2 particlesꢀin Rh/CeO2 leads to the catalyst deactivation
during ESR at 350 C can be ruled out.
Our XPS results suggest that the different Rh precursors on
CeO2 lead to different extents of oxidation of Rh metal particles
during ESR. Although the Rh/CeO2 catalysts derived from
Rh(NO3)3$xH2O, Rh(acac)3 and Rh4(CO)12 by H2 treatment at
400 ꢀC have much the same Rh dispersion on CeO2 and the
similar initial activity for ESR, the natures of the interaction
between Rh metal particles and CeO2 may be discrepant in the
three cases, so that Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2
behave differently against Rh oxidation as observed in this
work. More recently, Simson et al. reported through in situ XPS
ꢀ
measurements that during ESR at 650 C, a 4% Rh/Al2O3/ZrO2
monolith catalyst underwent a gradual Rh oxidation.119 Coronel
et al. showed by a comparative XPS study that aer ESR at 500
ꢀC, Rh0 was oxidized to Rhn+ in 0.6% Rh/La2O3(40%)–SiO2 and
0.6% Rh/CeO2–La2O3(10%)–SiO2 catalysts, and Rh0 remained
unchanged in a 0.6% Rh/La2O3(15%)–SiO2 catalyst.68 These
results demonstrate that supported Rh has a propensity to be
oxidized under ESR conditions, depending on the nature of the
Rh–support interaction or the dispersion of Rh metal particles.
XRD analysis of CeO2-supported Rh catalysts. XRD was used
to check the phase structure of CeO2 in Rh/CeO2. In ESI† are
shown the XRD patterns of Rh(N)/CeO2 and Rh(A)/CeO2 before
and aer 25 h of ESR at 350 ꢀC. Since the 1% Rh loading in Rh/
CeO2 was rather low, the diffraction peaks of Rh metal particles
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99477

View Article Online
RSC Advances
Paper
active among all the oxide-supported Rh catalysts studied for
WGS during low-temperature ESR.
(3) Likely due to the detrimental inuence of Cl retained in
the CeO2 lattice on catalysis, Rh(Cl)/CeO2 has poorer catalytic
performance than Rh(N)/CeO2, Rh(A)/CeO2 and Rh(C)/CeO2 in
H2 generation from low-temperature ESR.
(4) For Rh/CeO2, the efficiency of H2 production increases
with decreasing Rh loading, the 0.1% Rh loading enables slight
decrease in the selectivity to CH4, but Rh loadings below 1%
disable complete removal of CO from low-temperature ESR. The
1% Rh loading appears optimal for H2 generation from low-
temperature ESR.
(5) Calcination of Rh3+/CeO2 leads to decreases in the H2
yield and selectivity to CH4 and an increase in the selectivity to
CO in low-temperature ESR, which may be due to reduced
amounts of surface Rh0 caused by the strong Rh3+–CeO2 inter-
action. The strong Rh3+–CeO2 interaction at high temperatures
such as 700 ꢀC can render the Rh3+ to be partly buried or
wrapped by CeO2 particles aer the calcination, which limits
the access of H2 to the supported Rh3+ in TPR and the access of
C2H5OH to the active Rh sites in ESR.
(6) Rh(C)/CeO2 has stable catalytic performance during at
least 24 h of ESR at 350 ꢀC with a CO-free H2 yield of 4 mol
molꢁ1 C2H5OH. In contrast, Rh(N)/CeO2 and Rh(A)/CeO2 deac-
tivate gradually as the reaction proceeds under equivalent
conditions. The discrepancy in the catalytic stability can be
related to the different amounts of coke formed on these cata-
lysts which cause the catalyst deactivation to a different extent.
Other factors causing the catalyst deactivation such as oxidation
of the Rh metal particles and sintering of the Rh metal particles
and CeO2 in Rh/CeO2 during ESR, can be ruled out.
Fig. 16 TEM images of Rh(A)/CeO2 (a) before and (b) after 25 h of ESR
at 350 ꢀC.
did not appear. The diffraction peaks of CeO2 with uorite
structure remained unchanged in intensity and width aer ESR.
The results demonstrate that sintering of CeO2 particles in Rh/
CeO2 does not occur under the ESR conditions, in agreement
with the TEM observations. Thus, it is conrmed that the
possibility that sintering of CeO2 particles in Rh/CeO2 leads to
the catalyst deactivation during ESR at 350 ꢀC can be also ruled
out.
(7) Both CeO2-supported Rh0 and Rh+ may participate in
catalysis for ESR to produce H2, based on the fact that the
catalytic reaction is observed to proceed efficiently in the pres-
ence of either CeO2-supported Rh0 or Rh+.
(8) H2 generation from low-temperature ESR is strongly
kinetically controlled in the presence of Rh/CeO2. Rh/CeO2
promotes H2 production via ASR, ACSR and WGS at 300 ꢀC, and
via ASR, ACSR, MD, steam reforming of adsorbed CHx (x ¼ 1–3)
ꢀ
and WGS at 400 C.
Conclusions
Experimental
The following can be concluded from the whole work.
(1) Various types of oxides themselves are catalytically active Rh4(CO)12 and Rh(acac)3 were supplied by Strem. Rh(NO3)3$xH2O,
for H2 generation from the ESR process though the H2 yield is RhCl3$xH2O, silica gel (SiO2, Merck), g-Al2O3, Mg(OH)2 (95%),
low. The acidic oxide g-Al2O3 favours EDE followed by ED to CeO2 (99.9%), ZrO2 (99.99%), La2O3 (99.999%), Y2O3 (99.99%),
generate H2, the basic oxide MgO catalyzes EDA to release H2, TiO2 (99.99%), ZnO (99.99%) and other reagents were purchased
the weakly acidic oxide SiO2 catalyzes EDA to give H2 to a small from Aldrich. MgO was derived from Mg(OH)2 by 4 h of calcina-
ꢀ
extent, and the redox oxide CeO2 eases WGS as well as EDA, tion in owing air at 400 C, which was conrmed by the struc-
EDAC, ASR and ACSR to produce H2. The presence of Rh on tural transition in XRD (see ESI†). H2 and Ar gases used in
oxides greatly enhances the catalytic activity for all the reaction experiments had a purity of 99.999%.
pathways involved in ESR.
The above Rh compounds and metal oxides were used as
(2) Rh(N)/CeO2, Rh(N)/ZrO2 and Rh(N)/g-Al2O3 are generally precursors and supports to prepare oxide-supported Rh pre-
advantageous in H2 production from low-temperature ESR. catalysts and catalysts. The physical properties of the oxides
Rh(N)/CeO2, Rh/g-Al2O3, Rh/MgO and Rh/SiO2 are advanta- were determined by N2 adsorption–desorption, as shown in
geous over Rh/ZrO2, Rh/La2O3, Rh/Y2O3, Rh/TiO2 and Rh/ZnO in Table 10. In preparing precatalysts, the oxides were previously
AD and ASR during low-temperature ESR. Rh/CeO2 is most dried at 110 ꢀC in air overnight. Precatalysts from Rh4(CO)12
99478 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
were obtained through impregnation in dry hexanes under Ar as gas chromatography aer 10 min of reaction at each reaction
described previously.40 Precatalysts from Rh(acac)3 were ob- temperature. For an assessment of catalytic stability, the ESR
tained through impregnation in dry toluene under Ar following reaction was operated at a couple of xed temperatures. The GC
a similar procedure.40 Aer evacuation of organic solvents, the was equipped with two channels connected to two thermal
precatalysts were stored in Schlenk tubes under Ar. Precatalysts conductivity detectors, respectively. One was made up of
˚
from inorganic salts were obtained through incipient wetting. a molecular sieve 5 A column using Ar as carrier gas for analysis
Aer removal of water under vacuum, the precatalysts were of H2 and CO, the other of a Hayesep Q column using He as
dried under vacuum (10ꢁ2 torr) overnight at 21 ꢀC and stored in carrier gas for analysis of CO2, CH4, C2H4, C2H6, C2H5OH,
a vacuum desiccator. The Rh loading in the precatalysts was set CH3CHO, (CH3)2CO and H2O.
at 1% unless mentioned where necessary. The Rh4(CO)12
-
Conversion of C2H5OH, selectivity to a C-containing product,
derived CeO2-supported Rh precatalysts and catalysts are selectivity to H2 and yield of H2 were estimated according to the
denoted as Rh4(C)/CeO2 and Rh(C)/CeO2, the Rh(acac)3-derived following formulae:
CeO2-supported Rh precatalysts and catalysts as Rh3+(A)/CeO2
and Rh(A)/CeO2, the Rh(NO3)3$xH2O-prepared CeO2-supported
XðC2H5OHÞ% ¼
!
Rh precatalysts and catalysts as Rh3+(N)/CeO2 and Rh(N)/CeO2,
the RhCl3$xH2O-prepared CeO2-supported Rh precatalysts and
catalysts as Rh3+(Cl)/CeO2 and Rh(Cl)/CeO2. For the purpose of
studying the strong metal–support interaction, Rh3+(N)/CeO2
was subjected to 4 h of calcination in air at 400 and 700 ꢀC,
respectively.
2ðmol C2H5OHÞ
X
ꢀ
1 ꢁ
ꢂ 100
ꢁ
2ðmol C2H5OHÞ þ
x mol CxHyOz
ꢀ
ꢁ
ꢀ
ꢁ
x mol CxHyOz
ꢀ
X
S CxHyOz % ¼
ꢂ 100
ꢁ
x mol CxHyOz
ESR catalysts were tested using a customized ve channel
xed-bed quartz reactor from Bel Japan Inc that was connected
to a Varian GP 380 gas chromatography (GC). Four of the ve
reaction tubes were used for catalyst tests, with one serving as
a reference for blank tests. Precatalysts or calcined precatalysts
(0.15 g each) were reduced in H2 using a heating program with
10 ꢀC minꢁ1 from 21 to 400 ꢀC and holding at 400 ꢀC for 30 min
in the reaction tubes and ushed with Ar for 30 min while
cooled to 250 ꢀC, followed by introduction of a reactant mixture
with Ar. A liquid mixture of C2H5OH and H2O (1 : 3 volumetric
ratio or 1 : 10 molar ratio) at a ow rate of 0.005 ml minꢁ1 was
2ðmol H2Þ
X
ꢀ
SðH2Þ% ¼
ꢂ 100
ꢁ
2ðmol H2Þ þ
y mol CxHyOz
Y(H2) ¼ X(C2H5OH) ꢂ S(H2) ꢂ 6
where CxHyOz stands for C-containing product or
a
H-containing organic product. X(C2H5OH) was gained based on
unreacted C2H5OH and all detected C-containing products,
S(CxHyOz) on all detected C-containing products, and S(H2) on
H2 and all detected H-containing organic products. Y(H2) was
expressed in terms of the moles of H2 produced per mole of
C2H5OH. Blank tests veried the absence of gas phase reactions.
Under the reaction conditions used in this work, only H2, CO,
CO2, CH4, C2H4, C2H6, CH3CHO and (CH3)2CO were detected as
the products.
ꢀ
fed into the vaporizer at 170 C with Ar at a ow rate of 39 ml
minꢁ1 before admission to the reaction tubes (Ar : C2H5-
OH : H2O ¼ 49 : 1 : 10 molar ratio, 1.7 mol% C2H5OH). The
reactant composition with a C2H5OH : H2O molar ratio of 1 : 10
was similar to a bioethanol stream from biomass fermentation.
The Ar : (C2H5OH + H2O) molar ratio and GHSV amounted to
4.6 and 26 200 hꢁ1 (namely 17 700 ml gꢁ1 hꢁ1), respectively. For
a general test of catalytic performance as a function of reaction
temperature, the ESR reaction was consecutively run from 250
to 500 ꢀC or 600 ꢀC and each reaction temperature was held for
30 min. The reaction products were analyzed twice on line by
Surface areas of supports were measured by N2 adsorption–
desorption on a Quantachrome Autosorb-6B analyzer.
Reductive behaviour of precatalysts or calcined precatalysts
with a 1% Rh loading was studied by TPR on a Thermo Scientic
TPROD 1100 instrument. Prior to TPR experiments, all pre-
catalyst and calcined precatalyst samples except the Rh4(C)/
CeO2 sample were dried under vacuum at 21 ꢀC for 2 h, the
Rh4(C)/CeO2 sample was oxidized in air followed by 2 h of
drying under vacuum (see ESI†). A TPR prole was obtained by
passing 5% H2/Ar at a ow rate of 50 ml minꢁ1 through
a sample of 0.10 g from 30 to 800 or 850 ꢀC using a heating rate
Table 10 Physical properties of supports determined from N2
adsorption–desorption
BET surface
area (m2 gꢁ1
Pore volume
Pore
diameter (A)
˚
Support
)
(ml gꢁ1
)
SiO2
312.0
146.8
10.5
7.7
5.0
4.2
4.3
14.7
5.2
1.048
0.207
0.091
0.035
0.046
0.015
0.021
0.062
0.026
120.5
34.2
35.6
40.5
35.6
174.2
25.6
28.6
231.8
of 10 C minꢁ1. H2 consumption during TPR was recorded in
ꢀ
g-Al2O3
MgO
CeO2
ZrO2
La2O3
TiO2
thermal conductivity signals. Quantication of H2 consumed
was done by calibration with the aid of TPR proles of known
amounts of bulk CuO. H2 consumption was obtained from the
integrated peak area relative to the average calibration value.
Apparent H/Rh values were determined by H2 pulse chemi-
sorption, as stated previously.40
ZnO
Y2O3
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99479

View Article Online
RSC Advances
Paper
Amounts of carbon formed on catalysts during ESR were
evaluated by TGA in air, as stated previously.40
18 K. Otsuka, M. Hatano and A. Morikawa, J. Catal., 1983, 79,
493–496.
Rh oxidation states in CeO2-supported Rh catalysts were
determined by XPS, as stated previously.40
19 T. Bunluesin, R. J. Gorte and G. W. Graham, Appl. Catal., B,
1998, 15, 107–114.
Microscopic images of CeO2-supported Rh catalysts were
observed by TEM, as stated previously.40
20 S. Hilaire, X. Wang, T. Luo, R. J. Gorte and J. Wagner, Appl.
Catal., A, 2001, 215, 271–278.
Phase structures of solid materials were analyzed by XRD on
a Siemens D5005 spectrometer.
21 J. Llorca, P. Ramirez de la Piscina, J. Sales and N. Homs,
Chem. Commun., 2001, 641–642.
Rh contents in precatalysts and catalysts were determined by
the inductively coupled plasma (ICP) technique on a Varian
Vista-MPX CCD simultaneous ICP-OES spectrograph.
22 J. P. Breen, R. Burch and H. M. Coleman, Appl. Catal., B,
2002, 39, 65–74.
´
23 J. Llorca, P. Ramırez de la Piscina, J.-A. Dalmon and
N. Homs, Chem. Mater., 2004, 16, 3573–3578.
24 V. A. D. O'Shea, N. Homs, E. B. Pereira, R. Nafria and
P. Ramirez de la Piscina, Catal. Today, 2007, 126, 148–152.
Acknowledgements
´
We acknowledge the Science and Engineering Research Council
(SERC) of the Agency for Science, Technology and Research
(A*STAR) for nancial support.
25 N. J. Divins, I. Angurell, C. Escudero, V. Perez-Dieste and
J. Llorca, Science, 2014, 346, 620–623.
26 O. O. James, S. Maity, M. A. Mesubi, K. O. Ogunniran,
T. O. Siyanbola, S. Sahu and R. Chaubey, Green Chem.,
2011, 13, 2272–2284.
27 L. V. Mattos, G. Jacobs, B. H. Davis and F. B. B. Noronha,
Chem. Rev., 2012, 112, 4094–4123.
28 J. Sun and Y. Wang, ACS Catal., 2014, 4, 1078–1090.
29 D. Zanchet, J. B. O. Santos, S. Damyanova, J. M. R. Gallo and
J. M. C. Bueno, ACS Catal., 2015, 5, 3841–3863.
30 J. Lin, L. Chen, C. K. S. Choong, Z. Zhong and L. Huang, Sci.
China: Chem., 2015, 58, 60–78.
References
1 I. Fishtik, A. Alexander, R. Datta and D. Geana, Int.
J. Hydrogen Energy, 2000, 25, 31–45.
2 P. D. Vaidya and A. E. Rodrigues, Chem. Eng. J., 2006, 117,
39–49.
˜
´
´
´
¨ ¨
3 A. Erdohelyi, J. Rasko, T. Kecskes, M. Toth, M. Domok and
´
K. Baan, Catal. Today, 2006, 116, 367–376.
4 H.-S. Roh, Y. Wang, D. L. King, A. Platon and Y.-H. Chia,
Catal. Lett., 2006, 108, 15–19.
31 C. T. Campbell, Nat. Chem., 2012, 4, 597–598.
˜
32 F. Marino, M. Boveri, G. Baronetti and M. Laborde, Int.
5 B. Zhang, X. Tang, Y. Li, Y. Xu and W. Shen, Int. J. Hydrogen
Energy, 2007, 32, 2367–2373.
6 A. C. Basagiannis, P. Panagiotopoulou and X. E. Verykios,
Top. Catal., 2008, 51, 2–12.
7 E. B. Pereira, N. Homs, S. Marti, J. L. G. Fierro and
P. Ramirez de la Piscina, J. Catal., 2008, 257, 206–214.
J. Hydrogen Energy, 2001, 26, 665–668.
33 P. Y. Sheng, W. W. Chiu, A. Yee, S. J. Morrison and H. Idriss,
Catal. Today, 2007, 129, 313–321.
34 J. Kugai, V. Subramani, C. Song, M. H. Engelhard and
Y.-H. Chin, J. Catal., 2006, 238, 430–440.
35 M. Scott, M. Goeffroy, W. Chiu, M. A. Blackford and
H. Idriss, Top. Catal., 2008, 51, 13–21.
ˆ
8 F. Aupretre, C. Descorme and D. Duprez, Catal. Commun.,
2002, 3, 263–267.
9 D. K. Liguras, D. I. Kondarides and X. E. Verykios, Appl.
Catal., B, 2003, 43, 345–354.
10 F. Frusteri, S. Freni, V. Chiodo, L. Spadaro, O. Di Blasi,
G. Bonura and S. Cavallaro, Appl. Catal., A, 2004, 270, 1–7.
11 A. N. Fatsikostas and X. E. Verykios, J. Catal., 2004, 225,
439–452.
36 H. Idriss, M. Scott, J. Llorca, S. C. Chan, W. Chiu,
P.-Y. Sheng, A. Yee, M. A. Blackford, S. J. Pas, A. J. Hill,
F. M. Alamgir, R. Rettew, C. Petersburg, S. D. Senanayake
and M. A. Barteau, ChemSusChem, 2008, 1, 905–910.
37 A. Simson, E. Waterman, R. Farrauto and M. Castaldi, Appl.
Catal., B, 2009, 89, 58–64.
38 L. Chen, C. Choong, Z. Zhong, L. Huang, T. P. Ang, L. Hong
and J. Lin, J. Catal., 2010, 276, 197–200.
˜
12 J. Comas, F. Marino, M. Laborde and N. Amadeo, Chem.
Eng. J., 2004, 98, 61–68.
13 A. Aboudheir, A. Akande, R. Idem and A. Dalai, Int.
J. Hydrogen Energy, 2006, 31, 752–761.
14 C. C. R. S. Rossi, C. G. Alonso, O. A. C. Antunes,
R. Guirardello and L. Cardozo-Filho, Int. J. Hydrogen
Energy, 2009, 34, 323–332.
39 A. Simson, R. Farrauto and M. Castaldi, Appl. Catal., B,
2011, 106, 295–303.
40 L. Huang, C. Choong, L. Chen, Z. Wang, Z. Zhong,
C. Campos-Cuerva and J. Lin, ChemCatChem, 2013, 5,
220–234.
ˆ
41 F. Aupretre, C. Descorme, D. Duprez, D. Casanave and
15 A. Haryanto, S. Fernando, N. Murali and S. Adhikari, Energy
Fuels, 2005, 19, 2098–2106.
D. Uzio, J. Catal., 2005, 233, 464–477.
ˆ
ˆ
42 R. M. Navarro, M. C. Alvarez-Galvan and M. C. Sanchez-
ˆ
Sanchez, Appl. Catal., B, 2005, 55, 229–241.
16 A. Bshish, Z. Yaakob, B. Narayanan, R. Ramakrishnan and
A. Ebshish, Chem. Pap., 2011, 65, 251–266.
ˆ
ˆ
43 M. C. Sanchez-Sanchez, R. M. Navarro and J. L. G. Fierro,
Int. J. Hydrogen Energy, 2007, 32, 1462–1471.
´
˜
17 J. L. Contreras, J. Salmones, J. A. Colın-Luna, L. Nuno,
´
B. Quintana, I. Cordova, B. Zeifert, C. Tapia and
44 A. Denis, W. Grzegorczyk, W. Gac and A. Machocki, Catal.
Today, 2008, 137, 453–459.
G. A. Fuentes, Int. J. Hydrogen Energy, 2014, 39, 18835–
18853.
99480 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
45 F. Frusteri, S. Freni, L. Spadaro, V. Chiodo, G. Bonura,
S. Donato and S. Cavallaro, Catal. Commun., 2004, 5, 611–
615.
46 H.-S. Roh, Y. Wang and D. L. King, Top. Catal., 2008, 49, 32–
37.
72 M. Kourtelesis, P. Panagiotopoulou, S. Ladas and
X. E. Verykios, Top. Catal., 2015, 58, 1202–1217.
73 J. J. F. Scholten, P. Mars, P. G. Menon and R. van Hardereld,
Proceedings of the 3rd International Congress on Catalysis,
North-Holland, Amsterdam, 1965, July 1964, vol. 2, p. 881.
74 T. Shido, K. Asakura and Y. Iwasawa, J. Catal., 1990, 122,
55–67.
47 J. Llorca, N. Homs, J. Sales and P. Ramirez de la Piscina,
J. Catal., 2002, 209, 306–317.
48 Z. He, M. Yang, X. Wang, Z. Zhao and A. Duan, Catal. Today,
2012, 194, 2–8.
75 K. Tanabe, M. Misono, Y. Ono and H. Hattori, Stud. Surf.
Sci. Catal., 1989, 51, 27–127.
´
49 W. Cai, F. Wang, E. Zhan, A. C. van Veen, C. Mirodatos and
W. Shen, J. Catal., 2008, 257, 96–107.
76 A. Carrero, J. A. Calles and A. J. Vizcaıno, Appl. Catal., A,
2007, 327, 82–94.
50 H. Wang, Y. Liu, L. Wang and Y. N. Qin, Chem. Eng. J., 2008,
145, 25–31.
77 M. Mavrikakis and M. A. Barteau, J. Mol. Catal. A: Chem.,
1998, 131, 135–147.
51 H. Idriss, Platinum Met. Rev., 2004, 48, 105–115.
52 T. X. T. Sayle, S. C. Parker and C. R. A. Catlow, Surf. Sci.,
1994, 316, 329–336.
53 N. Kamiuchi, M. Haneda and M. Ozawa, Catal. Today, 2013,
201, 79–84.
54 C. Diagne, H. Idriss and A. Kiennemann, Catal. Commun.,
2002, 3, 565–571.
55 A. M. Karim, Y. Su, J. Sun, C. Yang, J. J. Strohm, D. L. King
and Y. Wang, Appl. Catal., B, 2010, 96, 441–448.
56 W.-C. Chiu, R.-F. Horng and H.-M. Chou, Int. J. Hydrogen
Energy, 2013, 38, 2760–2769.
57 N. Palmeri, S. Cavallaro, V. Chioda, S. Freni, F. Frusteri and
J. C. J. Bart, Int. J. Hydrogen Energy, 2007, 32, 3335–3342.
58 T. Montini, L. de Rogatis, V. Gombac, P. Fornasiero and
M. Graziani, Appl. Catal., B, 2007, 71, 125–134.
59 A. Birot, F. Epron, C. Descorme and D. Duprez, Appl. Catal.,
B, 2008, 79, 17–25.
78 H.-L. Chen, S.-H. Liu and J.-J. Ho, J. Phys. Chem. B, 2006,
110, 14816–14823.
79 J. Zhang, Z. Zhong, X.-M. Cao, P. Hu, M. B. Sullivan and
L. Chen, ACS Catal., 2014, 4, 448–456.
80 C. Houtman and M. A. Barteau, J. Phys. Chem., 1991, 95,
3155–3164.
81 H. Song, X. Bao, C. M. Hadad and U. S. Ozkan, Catal. Lett.,
2011, 141, 43–54.
82 J. Sun, H. Zhang, N. Yu, S. Davidson and Y. Wang,
ChemCatChem, 2015, 7, 2932–2936.
83 P. Panagiotopoulou and D. Kondarides, Catal. Today, 2006,
112, 49–52.
´
84 S. Bernal, J. J. Calvino, G. A. Cifredo, J. M. Gatica, J. A. Perez
Omil, A. Laachir and V. Perrichon, Stud. Surf. Sci. Catal.,
1995, 96, 419–429.
85 D. I. Kondaride and X. E. Verykios, J. Catal., 1998, 174, 52–
64.
60 Z. Zhong, H. Ang, C. Choong, L. Chen, L. Huang and J. Lin,
Phys. Chem. Chem. Phys., 2009, 11, 872–880.
61 X. Wu and S. Kawi, Energy Environ. Sci., 2010, 3, 334–342.
62 A. M. da Silva, K. R. de Souza, G. Jacobs, U. M. Graham,
B. H. Davis, L. V. Mattos and F. B. Noronha, Appl. Catal.,
B, 2011, 102, 94–109.
63 J. S. Moura, M. O. G. Souza, J. D. A. Bellido, E. M. Assaf,
M. Opportus, P. Reyes and M. do Carmo Rangel, Int.
J. Hydrogen Energy, 2012, 37, 3213–3224.
86 C. Force, J. P. Belzunegui, J. Sanz, A. Martinez-Arias and
J. Soria, J. Catal., 2001, 197, 192–199.
87 F. Fajardie, J.-F. Tempere, J.-M. Manoli, G. D. Mariadassou
and G. Blanchard, J. Chem. Soc., Faraday Trans., 1998, 94,
3727–3735.
88 H. Song, L. Zhang, R. B. Watson, D. Braden and U. S. Ozkan,
Catal. Today, 2007, 129, 346–354.
89 W. Wang and Y. Q. Wang, Int. J. Energy Res., 2008, 32, 1432–
1443.
´
64 Y. Ando, K. Matsuoka, H. Takagi and K. Kuramoto, Bull.
Chem. Soc. Jpn., 2012, 85, 517–521.
90 L. Hernandez and V. Kafarov, J. Power Sources, 2009, 192,
195–199.
´
65 O. A. Gonzalez Vargas, J. A. de los Reyes Heredia, J. A. Wang,
91 S. Yuvaraj, F.-Y. Lin, T.-H. Chang and C.-T. Yeh, J. Phys.
Chem. B, 2003, 107, 1044–1047.
L. F. Chen, A. Montesinos Castellanos and M. E. Llanos, Int.
J. Hydrogen Energy, 2013, 38, 13914–13925.
66 M. Bilal and S. D. Jackson, Catal. Sci. Technol., 2013, 3, 754–
766.
ˇ
92 A. Trovarelli, G. Dolcetti, C. de Leitenburg, J. Kaspar,
P. Finetti and A. Santoni, J. Chem. Soc., Faraday Trans.,
1992, 88, 1311–1319.
67 D. V. Andreev, V. V. Radaev, L. L. Makarshin,
A. G. Gribovskii, V. I. Zaikovskii and V. N. Parmon, Kinet.
Catal., 2014, 55, 798–808.
93 E. Ruchenstein and H. Y. Wang, J. Catal., 2000, 190, 32–38.
94 J. von Hoene, R. G. Charles and W. M. Hickam, J. Phys.
Chem., 1958, 62, 1098–1101.
´
68 L. Coronel, J. F. Munera, A. M. Tarditi, M. S. Moreno and
95 S. Poston and A. Reisman, J. Electron. Mater., 1988, 17, 57–
61.
96 J. A. Nario, M. Macias, F. J. Marchena and C. Real, Stud.
Surf. Sci. Catal., 1992, 72, 423–433.
L. M. Cornaglia, Appl. Catal., B, 2014, 160–161, 254–266.
69 T. Hou, B. Yu, S. Zhang, T. Xu, D. Wang and W. Cai, Catal.
Commun., 2015, 58, 137–140.
´
70 C. Diagne, H. Idriss and A. Kiennemann, Catal. Commun.,
2002, 3, 565–571.
71 R. W. McCabe and P. J. Mltchell, Ind. Eng. Chem. Prod. Res.
Dev., 1984, 23, 196–202.
97 A. Theolier, A. K. Smith, M. Leconte, J. M. Basset,
G. M. Zanderighi, R. Psaro and R. Ugo, J. Organomet.
Chem., 1980, 191, 415–424.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 99461–99482 | 99481

View Article Online
RSC Advances
Paper
98 X. Yu, F. Li, X. Ye and X. Xin, J. Am. Ceram. Soc., 2000, 83, 109 M. P. Hyman and J. M. Vohs, Surf. Sci., 2011, 605, 383–389.
964–966.
110 A. R. Passos, L. Martins, S. H. Pulcinelli, C. V. Santilli and
99 A. Roman and B. Delmon, J. Catal., 1973, 30, 333–342.
V. Briois, Catal. Today, 2014, 229, 88–94.
100 P. Turlier, H. Praliaud, P. Moral, G. A. Martin and 111 A. M. Karim, Y. Su, M. H. Engelhard, D. L. King and
J. A. Dalmon, Appl. Catal., 1985, 19, 287–300. Y. Wang, ACS Catal., 2011, 1, 279–286.
101 B. Mile, D. Stirling, M. A. Zammitt, A. Lovell and M. Webb, 112 J. Suna, A. M. Karim, D. Meib, M. Engelhard, X. Bao and
J. Catal., 1988, 114, 217–229. Y. Wang, Appl. Catal., B, 2015, 162, 141–148.
102 J. Z. Shyu, K. Otto, W. L. H. Watkins, G. W. Graham and 113 F. Haga, T. Nakajima, H. Miya and S. Mishima, Catal. Lett.,
R. K. Belitz, J. Catal., 1988, 114, 23–33. 1997, 48, 223–227.
103 A. Trovarelli, C. de Leitenburg, M. Boaro and G. Dolcetti, 114 J. R. Mielenz, Curr. Opin. Microbiol., 2001, 4, 324–329.
Catal. Today, 1999, 50, 353–367.
115 J. Llorca, P. Ramirez de la Piscina, J. A. Dalmon, J. Sales and
104 D. Briggs and M. P. Seah, Practical Surface Analysis, vol. 1:
N. Homs, Appl. Catal., B, 2003, 43, 355–369.
Auger and X-Ray Photoelectron Spectroscopy, Wiley, New 116 M. S. Batista, R. K. S. Santos, E. M. Assaf, J. M. Assaf and
York, 1990. E. A. Ticiasnelli, J. Power Sources, 2003, 124, 99–103.
105 D. I. Kondarides and X. E. Verykios, J. Catal., 1998, 174, 52– 117 H. Song, L. Z. Zhang and U. S. Ozkan, Green Chem., 2007, 9,
64. 686–694.
106 R. Herrmann, M. Rennhak and A. Reller, Beilstein 118 S. S. Y. Lin, D. H. Kim and S. Y. Ha, Catal. Lett., 2008, 122,
J. Nanotechnol., 2014, 5, 2413–2423. 295–301.
107 C. M. Y. Yeung, K. M. K. Yu, Q. J. Fu, D. Thompsett, 119 A. Simson, S. G. Podkolzin and M. J. Castaldi, Presented in
M. I. Petch and S. C. Tsan, J. Am. Chem. Soc., 2005, 127,
18010–18011.
part at 23rd North American Catalysis Society Meeting, P-W-
BRC-39, Louisville, June, 2013.
108 S. Tuti and F. Pepe, Catal. Lett., 2008, 122, 196–203.
99482 | RSC Adv., 2015, 5, 99461–99482
This journal is © The Royal Society of Chemistry 2015