Relationship between the structural properties of supported bimetallic Pt-Rh catalysts and their performances for methylcyclopentane ring opening

Article abstract of DOI:10.1016/j.jcat.2010.09.017

Bimetallic Pt-Rh catalysts supported on alumina and silica were synthesized either by the refilling method or by coimpregnation. They were characterized by transmission electron microscopy (TEM), H₂ chemisorption, temperature-programmed reduction (TPR), Fourier transform infrared (FTIR) of adsorbed CO and the test reaction of cyclohexane dehydrogenation. Varying the preparation procedure and the nature of the support leads to the formation of bimetallic Pt-Rh particles of different size in the range 1.2-2.7 nm. A Pt surface enrichment and the presence of a Pt-Rh alloy were detected by FTIR of adsorbed CO on the largest particles. The catalytic performances of the Pt-Rh systems were evaluated in methylcyclopentane (MCP) ring opening (RO) under pressure (28.5 bar). All the bimetallic catalysts present an activity much higher than that of the Pt monometallic sample, close to that of the Rh one. At 300 °C, the turnover frequencies are the followings: 115 h^-1 for Pt, 4094 h^-1 for Rh, between 2264 and 5563 h^-1 for the Pt-Rh catalysts whatever the preparation procedure and the nature of the support. On Pt-Rh samples presenting particles with large enough size (>1.7 nm), the RO selectivity at iso-conversion (>50%) is higher than those of monometallic Pt/Al₂O₃ and Rh/Al₂O₃ systems, reaching the performances of a monometallic iridium-based sample considered as the most selective catalyst in these conditions. Finally, the refilling method allows obtaining supported Pt-Rh catalysts with the structural properties required to obtain both high activity and RO selectivity during methylcyclopentane hydrogenolysis.

Full text of DOI:10.1016/j.jcat.2010.09.017

Journal of Catalysis 276 (2010) 237–248
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Relationship between the structural properties of supported bimetallic Pt–Rh
catalysts and their performances for methylcyclopentane ring opening
⇑
P. Samoila, M. Boutzeloit, C. Especel , F. Epron, P. Marécot
Laboratoire de Catalyse en Chimie Organique, UMR 6503, CNRS – Université de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Bimetallic Pt–Rh catalysts supported on alumina and silica were synthesized either by the refilling
method or by coimpregnation. They were characterized by transmission electron microscopy (TEM),
H2 chemisorption, temperature-programmed reduction (TPR), Fourier transform infrared (FTIR) of
adsorbed CO and the test reaction of cyclohexane dehydrogenation. Varying the preparation procedure
and the nature of the support leads to the formation of bimetallic Pt–Rh particles of different size in
the range 1.2–2.7 nm. A Pt surface enrichment and the presence of a Pt–Rh alloy were detected by FTIR
of adsorbed CO on the largest particles.
The catalytic performances of the Pt–Rh systems were evaluated in methylcyclopentane (MCP) ring
opening (RO) under pressure (28.5 bar). All the bimetallic catalysts present an activity much higher than
that of the Pt monometallic sample, close to that of the Rh one. At 300 °C, the turnover frequencies are the
followings: 115 hꢀ1 for Pt, 4094 hꢀ1 for Rh, between 2264 and 5563 hꢀ1 for the Pt–Rh catalysts whatever
the preparation procedure and the nature of the support. On Pt–Rh samples presenting particles with
large enough size (>1.7 nm), the RO selectivity at iso-conversion (>50%) is higher than those of monome-
tallic Pt/Al2O3 and Rh/Al2O3 systems, reaching the performances of a monometallic iridium-based sample
considered as the most selective catalyst in these conditions. Finally, the refilling method allows obtain-
ing supported Pt–Rh catalysts with the structural properties required to obtain both high activity and RO
selectivity during methylcyclopentane hydrogenolysis.
Received 16 June 2010
Revised 8 September 2010
Accepted 8 September 2010
Available online 27 October 2010
Keywords:
Pt–Rh catalysts
Ring opening
Methylcyclopentane
FTIR of adsorbed CO
Alloy
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
metallic noble metal-based catalysts, among which Ir-based sys-
tems lead generally to the best performances in terms of MCP
The selective ring opening (RO) of naphthenic molecules is a
reaction of interest for the valorization of various products result-
ing from catalytic reforming and cracking processes [1–3]. Indeed,
the hydrogenation of aromatic compounds followed by a selective
opening of the obtained naphthenes to paraffins allows improving
notably the fuel quality, by the selective formation of substituted
paraffins in the case of gasoline or linear ones in the case of diesel.
The selective ring opening reaction requires the hydrogenolysis of
only one endocyclic C–C bond while avoiding the secondary reac-
tions of deep hydrogenolysis.
In this research field, the ring opening of methylcyclopentane
(MCP) is often used as a probe reaction in the investigation of
the structure sensitivity of several hydrocarbon conversions cata-
lysed by noble metal-based catalysts [4–7]. Nevertheless, in a large
majority of studies, this reaction is not performed under refining
conditions, i.e. under both high pressure and temperature. Most
of the studied catalytic systems are composed of supported mono-
conversion and selectivity to RO products [8]. Nevertheless, irid-
ium is not available in great quantity at the natural state, making
impossible the manufacture of such catalysts to industrial scale
for their implementation in ring opening units. In this context,
the development of bimetallic systems showing catalytic perfor-
mances comparable to those of Ir catalysts presents an interesting
challenge for the researchers. Several bimetallic couples are
reported in the literature concerning the MCP ring opening per-
formed at low or atmospheric pressure, as Pt/Al2O3 or Pt/SiO2 sys-
tems modified by Sn, Ru, Au or Ge [9–13], Rh/Al2O3 or Rh/TiO2
modified by Pd, Ge, Sn or Ag [14–17] or Ir–Au/Al2O3 catalysts
[18]. In these studies, the bimetallic catalysts were prepared by
coimpregnation, sequential impregnation, underpotential deposi-
tion or organometallic grafting. In most cases, the addition of the
modifier onto the monometallic catalyst surface induces signifi-
cant changes in MCP conversion and/or products distribution, in
the degree of deactivation due to coke deposition or in resistance
to poisoning by sulphur molecules introduced in the charge. The
authors interpreted these phenomena in term of a modification
of the ring opening mechanism induced by the presence of the
additive in the based-metal neighbourhood.
⇑
Corresponding author. Fax: +33 5 49 45 37 41.
E-mail address: catherine.especel@univ-poitiers.fr (C. Especel).
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.09.017

238
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
Recently, we have studied the catalytic performances of several
content deposited on the catalyst by the refilling method was
determined by ICP-OES spectroscopy. For the coimpregnation
method, the H2PtCl6 and RhCl3 salts were simultaneously added
to the alumina support maintained at pH = 1 by HCl addition or di-
rectly to the silica support. The solution was gently stirred for 1 h
at room temperature. Then, the slurry was dried on a sand bath un-
til the solid was visibly dry. Drying was completed in an oven at
120 °C overnight. The catalyst was finally conditioned by heating
in air at 300 °C for 4 h, flushing with nitrogen and then reducing
in hydrogen (60 mL minꢀ1) at 500 °C for 4 h. Heating ramps were
bimetallic platinum-based catalysts for MCP ring opening under
high pressure conditions [19]. Two types of additives were used,
either inactive metals for MCP hydrogenolysis, such as copper or
germanium, or active ones, such as ruthenium and rhodium.
Among the studied catalysts, only Pt–Rh/Al2O3 samples showed
both activity and RO selectivity higher than those of monometallic
Pt/Al2O3 and Rh/Al2O3 systems, reaching the performances of an
iridium-based sample tested in the same conditions. In this previ-
ous work, the Pt–Rh/Al2O3 catalysts were prepared by a surface
redox reaction (namely the refilling method) that consists of
depositing the rhodium additive by reduction of its precursor salt
by hydrogen preadsorbed on the platinum surface. However, the
characterization of the synthesized Pt–Rh/Al2O3 catalysts was not
exposed in the previous paper [19] and the arrangement of plati-
num and rhodium atoms at the support surface was not deter-
mined. Consequently, at the end of this previous study, we were
not able to conclude whether the particular behaviour of this bime-
tallic system was directly linked to its particular preparation
method.
The present study will try to highlight the various parameters
that can modulate the catalytic performances for MCP RO of the
supported bimetallic Pt–Rh systems by varying their preparation
methods and the support. The classical preparation method by
coimpregnation was used in parallel with the refilling method, as
well as two oxides (alumina and silica) for the support. The bime-
tallic catalysts were characterized by H2 chemisorption and trans-
mission electron microscopy (TEM), by temperature-programmed
reduction, by Fourier transform infrared spectroscopy of adsorbed
CO and by the test reaction of cyclohexane dehydrogenation. Our
aim was to characterize in depth the bimetallic Pt–Rh samples in
order to explain the relationship between their structure and their
catalytic performances for the RO reaction of methylcyclopentane
under pressure.
programmed at 10 °C minꢀ1
.
The chlorine content on each catalyst was determined at the
‘‘Service central d’analyse” of the CNRS.
2.2. Transmission electron microscopy (TEM) measurements
TEM measurements were performed on a JEOL 2100 electron
microscope operating at 200 kV with a LaB6 source and equipped
with a Gatan Ultra scan camera. All the samples were embedded
in a polymeric resin (spur) and cut into a section as small as
40 nm with an ultramicrotome equipped with a diamond knife.
Cuts were then deposited on an Al grid previously covered with a
thin layer of carbon. Average particle sizes were determined by
measuring at least 100 particles for each sample analysed, from
at least five different micrographs. Microanalysis of Pt and Rh
was carried out by energy dispersive X-ray spectroscopy (EDX) in
the nanoprobe mode.
2.3. Hydrogen chemisorption
The metallic accessibility or dispersion (D%) was determined
from H2 chemisorption measurement, carried out in a pulse chro-
matographic system, using a stoichiometry H/Metal equal to one. It
is known [20] that this may lead to erroneous relative metallic dis-
persions but we have checked that the values of the particle size
estimated by the hydrogen chemisorption measurements are in
accordance with those determined from the TEM pictures.
2. Experimental
2.1. Catalyst preparation
2.4. Temperature-programmed reduction (TPR) experiments
The supports were
c
-alumina (from AXENS) and silica (from
The TPR experiments were performed with a 1.0vol% H2/Ar gas
mixture. The sample was heated in the temperature range of 25–
500 °C with a ramp of 5 °C minꢀ1 and then maintained at 500 °C
for 1 h. The hydrogen uptake was monitored by a thermal conduc-
tivity detector. Prior to the TPR, the catalysts were first pretreated
in situ under oxygen for 30 min at 300 °C and cooled down to room
temperature.
DEGUSSA) with specific surface areas (SBET) of 215 and 175
m2 gꢀ1, respectively. The supports were crushed and sieved, in or-
der to retain particles with sizes between 0.25 and 0.40 mm, and
calcined in flowing air at 450 °C for 4 h.
The monometallic 0.6 wt%Pt/Al2O3, 0.6 wt%Rh/Al2O3 and 0.6
wt%Ir/Al2O3 catalysts were prepared by impregnation of the pre-
cursor salts (H2PtCl6, RhCl3, H2IrCl6) in an HCl solution (pH = 1).
After evaporation of the solvent, the catalyst was activated by
calcination at 450 °C (Pt/Al2O3) or 300 °C (Rh/Al2O3 and Ir/Al2O3)
under air (4 h) and reduced under hydrogen at 500 °C (4 h). Rh/
Al2O3 and Ir/Al2O3 catalysts were characterized in [19]. Briefly,
they contain 1 wt% and 1.2 wt% of chlorine and their dispersions
are 56% and 85%, respectively.
Bimetallic Pt–Rh/Al2O3 catalysts were prepared either by sur-
face redox reaction or by coimpregnation using each time RhCl3
as Rh precursor salt. In the case of SiO2 support, only the coimpreg-
nation method was retained. The deposition of the Rh modifier of
the monometallic Pt/Al2O3 catalyst (called ‘‘parent” catalyst) by
surface redox reaction was performed according to the ‘‘refilling
method” described elsewhere [19]. In this study, the solution con-
taining the RhCl3 modifier salt in hydrochloric acid medium
(pH = 1) was let in contact with the parent catalyst under nitrogen
flow for 5 min. Finally, the solution was filtered out after reaction
and the catalyst was dried overnight before to be reduced under
hydrogen flow at 500 °C for 1 h (2 °C minꢀ1 heating rate). The Rh
2.5. Fourier transform infrared (FTIR) spectroscopy study of adsorbed
CO
Transmission FTIR spectra were collected in the single-beam
mode, with a resolution of 2 cmꢀ1, using a Nicolet Nexus FTIR spec-
trometer. Catalyst samples were prepared as self-supported wafers
with a diameter of 12 mm and a ‘‘thickness” of approximately
20 mg cmꢀ2
.
Prior to CO adsorption experiments, all the catalysts were re-
duced in situ for 1 h under H2 at 500 °C, then outgassed at this tem-
perature before to be cooled to room temperature. Reference
spectra of the clean surfaces were collected at room temperature.
CO was introduced to the cell, which was then purged to remove
any weakly bonded CO. Spectra were collected until steady state
was reached. Difference spectra between the samples and the cor-
responding reference are shown in this paper. Spectral deconvolu-
tion was performed using the Galactic PeakSolve peak-fitting
program. All deconvoluted spectra shown in this paper are

P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
239
converged solutions, with correlation factors above 0.999 and stan-
dard errors below 0.004.
‘‘RM”) or by classical coimpregnation (labelled ‘‘CI”). The refilling
method was used to synthesize a series of bimetallic 0.6 wt%Pt–
x wt%Rh/Al2O3 catalysts by modifying a monometallic 0.6 wt%Pt/
Al2O3 catalyst (‘‘parent” catalyst) with several successive refillings
of 0.2 wt%Rh. Table 1 shows that the Rh content really deposited on
the catalyst increases logically with the theoretical wt%Rh intro-
duced in solution, but the Rh deposit is not total. However, for a
same Rh content introduced in solution, the present RM method al-
lows depositing more important Rh quantities on platinum than
the redox method used in our previous study [19], i.e. a refilling
step followed by a catalytic reduction step under hydrogen (depos-
it of 0.46 wt% vs. 0.17 wt%Rh, respectively, for 0.80 wt%Rh intro-
duced in solution). Coimpregnated 0.6 wt%Pt–0.3 wt%Rh catalysts
were also prepared on alumina and silica, the chosen Rh content
being in the middle of the range of the RM samples series.
2.6. Cyclohexane dehydrogenation
The test reaction of cyclohexane dehydrogenation to benzene
was performed in order to study the effect of the Rh deposit on
the Pt metallic phase. This reaction was carried out in a continuous
flow reactor at 270 °C under atmospheric pressure on 10 mg of cat-
alyst. The detailed protocol was described elsewhere [19].
2.7. Ring opening of methylcyclopentane (MCP)
MCP hydrogenolysis was performed under a total pressure of
H2
28.5 bar with a molar ratio MCP = 7.5 in a temperature range from
In Table 1, the characteristics of the Pt–Rh/Al2O3-RM catalysts
are compared to those of both monometallic Pt/Al2O3 catalysts
named ‘‘parent” and ‘‘blank”, respectively. To prepare the blank
Pt/Al2O3 sample, the parent catalyst was treated under hydrogen
in a HCl solution of pH = 1. This treatment corresponds to the prep-
aration protocol of the RM bimetallic catalysts, but without addi-
tion of the Rh precursor salt. The blank treatment influences
somewhat the characteristics of the parent Pt/Al2O3 catalyst. Be-
sides a slight increase of the chlorine content, an increase of the
metal particles size is noticed. This sintering phenomenon induced
by the actual preparation procedure by surface redox reaction was
already observed on various monometallic and bimetallic catalysts
under similar conditions [21,22]. Thus, this phenomenon can ex-
plain the low metallic accessibility value of the bimetallic
0.6 wt%Pt–0.36 wt%Rh/Al2O3-RM catalyst compared to that of the
0.6 wt%Pt–0.3 wt%Rh/Al2O3-CI sample (0.36 vs. 0.77). On the other
hand, the CI catalyst supported on silica presents a metallic acces-
sibility comparable with that of Pt–Rh/Al2O3-RM system (0.41 vs.
0.36) and thus much lower than that of the CI catalyst supported
on alumina. This result is not surprising since it is known that on
silica the metal–support interactions are rather weak compared
to alumina. Moreover, silica support retains less chlorine, since
the chlorine content measured on the Pt–Rh/SiO2-CI sample is
three to four times lower than that of the Pt–Rh/Al2O3 catalysts.
Fig. 1 shows representative TEM images, as well as the corre-
sponding particle size distributions for the three bimetallic Pt–Rh
catalysts with Rh content close to 0.3 wt%. The average particle
sizes of all the samples examined by TEM are reported in Table
1. A good homogeneity is highlighted on TEM pictures with the
presence of small metal particles well distributed on the alumina
or silica support. As observed previously by H2 chemisorption,
Fig. 1b and c shows that the average particle size of Pt–Rh catalysts
prepared by CI method is notably different according to the sup-
200 to 425 °C according to the experimental protocol described
elsewhere [19]. The absence of diffusion limitations in these exper-
imental conditions was checked by varying independently the
particle size (in the 0.1–0.6 mm range), the amount of catalyst
(from 125 mg to 500 mg) and the reactant flow rates (from
0.0334 to 0.1336 mL minꢀ1). Conversion, defined as the percentage
of MCP converted, was determined as a function of temperature.
The only reaction products were 2-methylpentane, 3-methylpen-
tane, n-hexane and C1–C5 products. Neither cyclohexane nor ben-
zene was detected as reaction products. The selectivity of the
catalysts for ring opening (hydrogenolysis of only one endocyclic
C–C bond) was estimated from the yield in C1–C5 products result-
ing from deep hydrogenolysis: the lowest the C1–C5 yield, the
highest the RO products yield and then the highest the MCP ring
opening selectivity. The yield of C1–C5, defined as the percentage
of MCP converted into C1–C5 products, was reported as a function
of the conversion. It was verified that the yield in C1–C5 obtained
for a given conversion is independent of the reaction conditions by
varying the WHSV instead of the temperature. For all the catalysts,
the ‘‘Conversion = f(Temperature)” and ‘‘C1–C5 yield = f(Conver-
sion)” curves will be always studied simultaneously in order to
compare their catalytic performances since the effects of tempera-
ture and conversion are linked in our experimental conditions.
3. Results and discussion
3.1. Main characteristics of the supported Pt–Rh catalysts
3.1.1. Preparation of the Pt–Rh catalysts by MR and CI methods
Table 1 displays the main characteristics of the various bimetal-
lic Pt–Rh catalysts supported on either alumina or silica and pre-
pared either by surface redox reaction (refilling method labelled
Table 1
Characteristics of the supported 0.6 wt%Pt–x wt%Rh catalysts prepared by refilling method (RM) or by coimpregnation (CI),
compared with those of monometallic parent and blank 0.6 wt%Pt/Al2O3 samples.
H/Ma
b
c
-
-
Catalyst
wt%Rh theoretical
wt%Rh deposited
wt%Cl
d (nm)
d (nm)
Pt/Al2O3-parent
Pt/Al2O3-blank
Pt–Rh/Al2O3-RM
–
–
–
–
1.36
1.54
1.25
1.40
1.40
1.30
1.50
1.70
0.45
0.75
0.55
n.d.
n.d.
0.36
n.d.
n.d.
0.77
0.41
1.3
1.7
n.d.
n.d.
2.6
n.d.
n.d.
1.2
2.3
n.d
n.d
2.3
n.d.
2.4
n.d.
2.7
1.2
2.4
0.20
0.40
0.60
0.80
2.00
0.30
0.30
0.11
0.23
0.36
0.46
0.85
0.30
0.30
Pt–Rh/Al2O3-CI
Pt–Rh/SiO2-CI
n.d.: not determined.
a
Determined from H2 chemisorption.
Deduced from H/M value.
Deduced from TEM analysis.
b
c

240
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
30
25
a
d=2.4 nm
20
15
10
5
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4
d (nm)
5 nm
50
45
40
35
30
25
20
15
10
5
b
d=1.2 nm
0
0.8
1.2
1.6
2.0
2.4
2.8
d (nm)
5 nm
c
30
25
20
15
10
d=2.4 nm
5
0
1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4
d (nm)
5 nm
Fig. 1. TEM images and particle size distributions of: (a) Pt–0.36Rh/Al2O3-RM; (b) Pt–0.30Rh/Al2O3-CI; (c) Pt–0.30Rh/SiO2-CI.
port nature: 2.4 nm for Pt–Rh/SiO2-CI sample against 1.2 nm for
Pt–Rh/Al2O3-CI system. Finally, whatever the studied sample, the
average particle sizes deduced from TEM analysis and chemisorp-
tion measurements are in perfect agreement (Table 1). Concerning
the RM samples, a slight enlargement of the metal particles is ob-
served with the quantity of deposited Rh from 2.3 nm (0.11 wt%Rh)
to 2.7 nm (0.85 wt%Rh). EDX analysis of metal crystallites reveals
the presence of platinum and rhodium in the same particle on all
the Pt–Rh systems. Moreover, for the RM catalysts, the Rh quantity
within a metal particle increases with the deposited Rh content,
which allows explaining the enlargement phenomenon observed
previously by TEM.
3.1.2. Sintering treatments of the Pt–Rh/Al2O3-CI catalyst
The previous results indicated that for similar metallic compo-
sitions, the supported Pt–Rh catalysts displayed average particle
sizes strongly influenced by the preparation procedure and the
support used. Thus, on the alumina support, both RM and CI prep-
aration methods does not lead to bimetallic Pt–Rh/Al2O3 catalysts
of comparable particle sizes. In order to compare latter equitably
the catalytic performances for MCP hydrogenolysis of both sam-
ples, the CI catalyst was submitted to various treatments in order
to enlarge its metal particles. The first one consisted to subject
the Pt–Rh/Al2O3-CI sample to a blank treatment, i.e. a treatment
under hydrogen flow in HCl solution of pH = 1. The second one

P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
241
consisted in activating Pt–Rh/Al2O3-CI system under nitrogen flow
and steam at 400 °C during 2 h.
The main characteristics of the bimetallic catalysts treated in
such ways (named Pt–Rh/Al2O3-CI-HCl and Pt–Rh/Al2O3-CI-T,
respectively) are given in Table 2. Both treatments undergone by
Pt–Rh/Al2O3-CI catalyst involve a slight decrease in the chlorine
content and above all a decrease in metal accessibility values, espe-
cially for the blank treatment (Pt–Rh/Al2O3-CI-HCl catalyst). Never-
theless, these values remain higher than those observed for Pt–Rh/
Al2O3-RM and Pt–Rh/SiO2-CI catalysts. A good agreement is ob-
tained between the average particle sizes measured by TEM and
the metal accessibility values determined by H2 chemisorption.
Moreover, on these catalysts, the presence of platinum and rho-
dium within the same metal particle is still highlighted by EDX
analysis, i.e. none of the treatments induced segregation of the
two metals.
Fig. 2. Relative activity for cyclohexane dehydrogenation of supported bimetallic
Pt–Rh catalysts: (ꢀ) Pt–Rh/Al2O3-RM (reported to Pt/Al2O3-blank sample); (j) Pt–
Rh/Al2O3-CI, (+) Pt–Rh/Al2O3-CI-HCl and (N) Pt–Rh/Al2O3-CI-T (reported to Pt/
Al2O3-parent sample); (h) Pt–Rh/SiO2-CI (reported to
a Pt/SiO2 sample); (s)
3.2. Characterization of the metallic function of the supported Pt–Rh
catalysts
0.6 wt%Rh/Al2O3 catalyst (reported to Pt/Al2O3-parent sample).
its counterparts of same average particle size but prepared by
RM method or on silica support. These results would mean that
the more the bimetallic Pt–Rh particles are large, the less their sur-
face is rich in rhodium.
To conclude this part, cyclohexane dehydrogenation seems to
indicate that small Pt–Rh particles have more rhodium atoms at
their surface than the largest ones, for the same platinum and rho-
dium content. Then, it can be inferred on a platinum enrichment on
large Pt–Rh particles.
3.2.1. Activity for the cyclohexane dehydrogenation
Cyclohexane dehydrogenation towards benzene performed un-
der atmospheric pressure is a reaction catalysed by the metal func-
tion and considered generally as structure insensitive [23].
Therefore, the activity is directly proportional to the number of
surface active atoms [24,25]. The relative activities obtained on
all the supported bimetallic Pt–Rh catalysts reported to that of
the corresponding monometallic sample (i.e. either the Pt/Al2O3-
blank system for the RM series or the Pt parent system supported
on Al2O3 or SiO2 for the CI catalysts) are plotted in function of the
deposited Rh content in Fig. 2. For comparison, the relative activity
of a 0.6 wt%Rh/Al2O3 catalyst is presented on the same figure.
First, it should be noticed that, whatever the preparation meth-
od (MR, CI, CI + sintering treatment) and the support used, the
addition of rhodium inhibits the platinum activity. Concerning
the bimetallic Pt–Rh/Al2O3-RM catalysts, Fig. 2 shows that the
inhibiting effect increases with the quantity of deposited rhodium.
This inhibiting effect indicates the existence of an interaction be-
tween rhodium and platinum, rhodium being less active than plat-
inum for this reaction. However, the bimetallic interaction or more
exactly the nature of the active surface of the Pt–Rh systems seems
to depend on the preparation method and the support. Thus, for a
same content of deposited Rh (approximately 0.3 wt%), the relative
activities for cyclohexane dehydrogenation of the supported Pt–Rh
samples are quite comparable except in the case of the Pt–Rh/
Al2O3-CI system that is less active than all the others. As this last
system possesses the highest metal accessibility among the various
bimetallic catalysts (Table 1), its lowest relative activity could be
explained by a larger amount of rhodium on the bimetallic Pt–Rh
particle surface. After undergoing sintering treatments (in HCl
medium or by thermal activation), the relative activity of the Pt–
0.30Rh/Al2O3-CI catalyst reaches values comparable to those of
3.2.2. Characterization by temperature-programmed reduction
The reduction in the catalysts pre-oxidized at 300 °C was fol-
lowed by the temperature-programmed reduction technique.
Fig. 3A shows the TPR profiles of two bimetallic Pt–Rh/Al2O3-RM
catalysts (with 0.11 and 0.36 wt%Rh, respectively) compared to
the profiles of monometallic 0.6 wt%Pt/Al2O3 (parent and blank)
and 0.3 wt%Rh/Al2O3 samples. The Pt/Al2O3-parent and Pt/Al2O3-
blank catalysts present a reduction peak with maxima located at
236 °C and 214 °C, respectively (Fig. 3A, profiles a and b). This shift
to a lower temperature can be explained by the increase of the me-
tal particle size of the parent catalyst after the blank treatment,
inducing a weaker metal–support interaction on the blank catalyst
[22]. The reduction curves of the bimetallic RM catalysts are quite
comparable with two peaks centred, respectively, at 95 °C and
258 °C (Fig. 3A, profiles c and d). Moreover, the surface of the peak
at lower temperature increases with the rhodium content. Finally,
the TPR profile of the monometallic Rh catalyst presents one reduc-
tion peak between 50 and 250 °C, with a maximum at 130 °C
(Fig. 3A, profile e).
The hydrogen consumptions deduced from TPR curves are re-
ported in Table 3. For the monometallic Pt and Rh catalysts, the
values of H2/M ratios indicate that, after calcination at 300 °C, plat-
inum and rhodium were almost completely oxidized in the PtO2
(theoretical H/M = 2) and Rh2O3 (theoretical H/M = 1.5) forms,
respectively. The addition of rhodium to platinum by the RM tech-
nique contributes to increase the hydrogen consumption com-
pared to the monometallic Pt catalysts, leading to H2/M ratios
intermediate between those of the Rh/Al2O3 and Pt/Al2O3-blank
systems. The quantities of consumed hydrogen by the bimetallic
RM catalysts correspond approximately to those necessary to re-
duce the monometallic 0.6 wt%Pt/Al2O3-blank and 0.3 wt%Rh/
Al2O3 systems, by taking into account the Pt and Rh contents in
these bimetallic samples. Nevertheless, according to Table 3, it is
clear that the first peak of hydrogen consumption observed on
the Pt–Rh/Al2O3-RM catalysts does not correspond only to the
Table 2
Characteristics of the 0.6 wt%Pt–0.30 wt%Rh/Al2O3 catalysts prepared by coimpreg-
nation (CI) and submitted to various sintering treatments.
-
-
Catalyst
Sintering
treatment
wt%Cl H/
Ma
d
d
(nm)b
(nm)c
Pt–Rh/Al2O3-CI
Pt–Rh/Al2O3-CI-
HCl
–
1.70
1.51
0.77
0.50
1.2
1.8
1.2
1.7
HCl (pH = 1)/H2
Pt–Rh/Al2O3-CI-T
H2O/N2/400 °C
1.43
0.60
1.6
1.5
a
b
c
Determined from H2 chemisorption.
Deduced from H/M value.
Deduced from TEM analysis.

242
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
(A)
(B)
25 a.u.
(a)
(b)
.
(a)
(b)
25a.u.
(c)
(c)
(d)
(e)
(d)
(e)
500°C (1 h)
500°C (1 h)
0
100 200 300 400 500
0
100 200 300 400 500
Temperature (°C)
Temperature (°C)
Fig. 3. (A) TPR profiles of Pt–Rh/Al2O3-RM catalysts compared with the Pt/Al2O3 and Rh/Al2O3 ones: (a) Pt/Al2O3-parent, (b) Pt/Al2O3-blank, (c) Pt–0.11 wt%Rh/Al2O3, (d) Pt–
0.36 wt%Rh/Al2O3 and (e) 0.3 wt%Rh/Al2O3. (B) TPR profiles of Pt–Rh catalysts (containing around 0.30 wt%Rh) prepared by different methods: (a) Pt–0.36 wt%Rh/Al2O3-RM,
(b) Pt–0.30 wt%Rh/Al2O3-CI, (c) Pt–0.30 wt%Rh/SiO2-CI, (d) Pt–0.30 wt%Rh/Al2O3-CI-HCl and (e) Pt–0.30 wt%Rh/Al2O3-CI-T.
Table 3
Fig. 3B profile a). Nevertheless, as observed for RM sample, the
presence of rhodium introduced by CI method contributes to in-
Hydrogen consumptions during TPR analysis of the supported 0.6 wt%Pt–x wt%Rh
catalysts prepared by refilling method (RM) or by coimpregnation (CI), compared
crease the quantity of consumed H2 compared to the catalyst
with those of monometallic 0.6 wt%Pt/Al2O3 (parent and blank) and 0.3 wt%Rh/Al2O3
0.6Pt/Al2O3 (Table 3). On bimetallic Pt–Rh/Al2O3-CI catalyst pre-
senting rather small metallic particles, the presence of only one
reduction peak corresponding to such a H2 consumption can be
attributed to the reduction of bimetallic entities, inside of which
rhodium catalyses the reduction of platinum.
samples.
Catalyst
Tmax
(°C)
H2 consumed Total number
H2/M
ꢀ1
of metal atoms
(lmol g
)
cat
ꢀ1
(
lmol g
)
cat
Pt/Al2O3-parent
Pt/Al2O3-blank
Pt–0.11Rh/Al2O3-RM
236
214
95
258
95
258
130
58, 116
60
55
45
24
74
25
83
78
31
31
42
1.9
1.8
1.6
The CI catalyst supported on silica (Fig. 3B, profile c) presents a
particular curve with two reduction maxima at low temperature
(58 and 116 °C). For both Pt–Rh/Al2O3-CI-HCl and Pt–Rh/Al2O3-
CI-T catalysts (Fig. 3B, profiles d and e), TPR profiles are compara-
ble to that of the Pt–Rh/Al2O3-RM catalyst previously observed,
with two reduction peaks: a first major peak located at low tem-
perature (102 °C and 124 °C respectively) and a second minor
one at 287 °C. Finally, bimetallic Pt–Rh catalysts of comparable
particle size (Pt–Rh/Al2O3-RM, Pt–Rh/SiO2-CI, Pt–Rh/Al2O3-CI-HCl
and Pt–Rh/Al2O3-CI-T) tend to present comparable behaviours dur-
ing TPR analysis, and these results being in agreement with the
tendencies observed during cyclohexane dehydrogenation.
In conclusion, the TPR characterization highlights an interaction
between platinum and rhodium for all the bimetallic Pt–Rh cata-
lysts, whatever the preparation method (RM or CI) or the support
used (alumina or silica). However, this technique shows that the
nature of the Pt–Rh phases differs according to the average parti-
cles size of the bimetallic sample, the size being modulated by both
the preparation procedure and the nature of the support.
Pt–0.36Rh/Al2O3-RM
66
1.5
Pt–0.30Rh/Al2O3-CI
Pt–0.30Rh/SiO2-CI
Pt–0.30Rh/Al2O3-CI-HCl 102, 287 73
Pt–0.30Rh/Al2O3-CI-T
Rh/Al2O3
60
60
60
60
29
1.4
1.3
1.2
1.1
1.5
124, 287 69
130 43
reduction of Rh species, but to the reduction of bimetallic Pt–Rh
entities. The second peak of consumption at higher temperature
could then be assigned to the presence of small platinum particles
in strong interaction with the support and/or of bimetallic Pt–Rh
particles highly enriched in Pt.
Fig. 3B compares the TPR curves of supported bimetallic Pt–Rh
catalysts of similar Rh content (near 0.30 wt%Rh) but prepared by
different methods (RM or CI) on alumina or silica support. The
TPR profiles vary notably according to the preparation method
and the treatment undergone by the bimetallic samples. Indeed,
on alumina support, the CI catalyst (not modified by treatment
after its preparation) presents a TPR profile quite similar to that
of monometallic rhodium sample, with only one reduction peak
with a maximum at 130 °C (Fig. 3B, profile b). Therefore, this TPR
profile is different from that of the RM system of same metallic
3.2.3. Characterization by FTIR of adsorbed CO
FTIR spectra of CO adsorbed at room temperature on supported
bimetallic Pt–Rh catalysts are given in Fig. 4 compared to those of
monometallic Pt and Rh catalysts.
All spectra, except spectrum d of Fig. 4B corresponding to the
only catalyst supported on silica, show the existence of a band
compositions as described previously (Fig. 3A profile
d and

P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
243
(A)
(B)
0.01 a.u.
0.01 a.u.
.
.
(a)
(b)
(a)
(b)
(c)
(c)
(d)
(d)
(e)
2250 2200 2150 2100 2050 2000 1950
2250
2200
2150
2100
2050
2000
1950
Wavenumber (cm-1)
Wavenumber (cm-1)
Fig. 4. (A) FTIR spectra of adsorbed CO of Pt–Rh/Al2O3 catalysts compared with the Pt/Al2O3 and Rh/Al2O3 ones: (a) Pt/Al2O3-parent, (b) Pt–0.30 wt%Rh/Al2O3-CI, (c) Pt–
0.36 wt%Rh/Al2O3-RM and (d) 0.3 wt%Rh/Al2O3. (B) FTIR spectra of adsorbed CO of supported Pt–Rh catalysts prepared by CI method compared with the Pt/Al2O3 and Rh/Al2O3
ones: (a) Pt/Al2O3-parent, (b) Pt–0.30 wt%Rh/Al2O3-CI-T, (c) Pt–0.30 wt%Rh/Al2O3-CI-HCl, (d) Pt–0.30 wt%Rh/SiO2-CI and (e) 0.3 wt%Rh/Al2O3.
more or less marked at 2244 cmꢀ1, attributed to CO adsorbed on
the Lewis acid sites of the alumina support [26].
lyst (Fig. 4A, spectrum c), a main dissymmetric band is located at
2077 cmꢀ1, which can be decomposed into three main elementary
peaks (Fig. 5c). The decomposition of this spectrum highlights the
presence of a peak at 2061 cmꢀ1 not observed on the CI system (Ta-
ble 4). Fig. 4B shows that the spectra of the Pt–0.30Rh/Al2O3-CI-T,
Pt–0.30Rh/Al2O3-CI-HCl and Pt–0.30Rh/SiO2-CI catalysts present a
profile comparable to that of the Pt–0.36Rh/Al2O3-RM catalyst,
with a main band in the 2066–2075 cmꢀ1 range and two shoulders
around 2033 cmꢀ1 and 2100 cmꢀ1. Moreover, one band of very low
intensity located at 2132 cmꢀ1 is observed on Pt–0.30Rh/Al2O3-CI-
T and Pt–0.30Rh/Al2O3-CI-HCl catalysts, indicating the presence of
Rh2+ species. After decomposition (example given in Fig. 5d for Pt–
0.30Rh/Al2O3-CI-HCl sample), a band at 2059 cmꢀ1 is once again
observed as on the Pt–0.36Rh/Al2O3-RM catalyst. Lévy et al. [39]
studied the surface of monometallic catalysts containing platinum
and rhodium, as well as bimetallic platinum–rhodium catalysts
supported on alumina, by successive adsorptions of NO and CO fol-
lowed by IR. On the bimetallic systems, these authors also ob-
served the appearance of a band around 2060 cmꢀ1, which they
attributed to a Pt–Rh alloy. Finally, in the present work, this band
is only observed on the supported bimetallic Pt–Rh catalysts pre-
senting the largest particles size, i.e. all the samples except CI cat-
alyst deposited on alumina without subsequent sintering
treatment.
On the monometallic Pt/Al2O3 catalyst (Fig. 4A, spectrum a),
FTIR analysis reveals only one intense band located at 2067 cmꢀ1
characteristic of CO adsorbed in a linear mode onto platinum
(Pt0–CO species) [27,28]. On the monometallic Rh/Al2O3 catalyst
(Fig. 4A, spectrum d), four distinct bands are observed: two intense
ones at 2103 and 2032 cmꢀ1 and two lower ones at higher wave-
number (2145 and 2175 cmꢀ1). Similar results concerning the
adsorption of CO on Rh are reported in the literature, making it
possible to assign these bands in the following way [29–36]: (i)
bands at 2032 and 2103 cmꢀ1 correspond to the vibrations of the
gem-dicarbonyl Rh1+–(CO)2 species (antisymmetric and symmetric
mode, respectively), (ii) bands at 2145 and 2175 cmꢀ1 are attrib-
uted to CO adsorbed onto Rh2+ and Rh3+ species, respectively.
The presence of Rh+ species is generally explained by Rh oxidation
in presence of CO involving hydroxyl groups of the support, accord-
ing to the following reaction [37]: 2Rh + 4CO + 2AlOH ? H2 + 2Al–
O–Rh(CO)2. This phenomenon results from the higher Rh–CO bind-
ing energy than that of Rh–Rh bond (185 kJ molꢀ1 vs. 44.5 kJ
molꢀ1). The presence of Rh2+ and Rh3+ species on catalyst prere-
duced under H2 during 1 h at 500 °C can be surprising. However,
after decomposition of the spectra into elementary bands, their rel-
ative surfaces are very weak (Fig. 5a and Table 4: 2.2% for Rh3+ and
1.9% for Rh2+), the gem-dicarbonyl species being predominant on
this sample (relative surface close to 96%). The presence of Rh3+oxi-
dized species was already observed by Kondarides et al. [38] on Rh
catalysts supported on CeO2 and was attributed to the proximity of
chlorine atoms that stabilize rhodium in the Rh3+ oxidation state.
In addition, no band at 2060 cmꢀ1 corresponding to the CO linearly
adsorbed (Rh0–CO) was observed.
In conclusion, the FTIR analysis highlights the existence of me-
tal phases of different nature between the various Pt–Rh systems,
as previously observed for the probe reaction of cyclohexane dehy-
drogenation and by temperature-programmed reduction. The me-
tal phase nature of the Pt–Rh catalysts seems to be directly
correlated to their particles size, with the formation of particular
Pt–Rh species when the particles are large enough.
The bimetallic Pt–Rh systems present different spectra accord-
ing to the preparation method. The IR spectrum of adsorbed CO
on the Pt–Rh/Al2O3-CI catalyst (Fig. 4A, spectrum b) is composed
of three distinct main bands at 2096, 2081 and 2033 cmꢀ1, and
these vibration frequencies are corresponding after decomposition
to those observed on both monometallic platinum and rhodium
catalysts spectra (Fig. 5b, Table 4). However, on this sample, the
relative surface corresponding to symmetric gem-dicarbonyl spe-
cies becomes very low whereas that of the antisymmetric mode
is higher than that obtained for the Rh/Al2O3 system (51.5% vs.
33.3%, respectively). On the bimetallic Pt–0.36Rh/Al2O3-RM cata-
3.3. Performances of the supported Pt–Rh catalysts during MCP ring
opening (RO)
3.3.1. Bimetallic Pt–Rh/Al2O3 catalysts prepared by refilling method
(RM)
The MCP conversions as function of temperature and yields in
C1–C5 products as function of MCP conversion obtained on the
bimetallic Pt–Rh/Al2O3-RM, and monometallic 0.6Pt/Al2O3 (parent
and blank) and 0.6Rh/Al2O3 catalysts are presented in Fig. 6. The
catalytic performances of a 0.6 wt%Ir/Al2O3 catalyst are also pre-

244
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
0.05
a
b
0.015
0.04
0.03
0.02
0.01
0.010
0.005
0
0
2250
2200
2150
2100
2050
2000
1950
2250
2200
2150
2100
2050
2000
1950
Wavenumber (cm-1)
Wavenumber (cm-1)
c
d
0.04
0.03
0.02
0.01
0
0.06
0.04
0.02
0
2250
2200
2150
2100
2050
2000
1950
2250
2200
2150
2100
2050
2000
1950
Wavenumber (cm-1)
Wavenumber (cm-1)
Fig. 5. FTIR spectra of adsorbed CO decomposed in elementary bands for monometallic Rh and bimetallic Pt–Rh catalysts: (a) 0.3 wt%Rh/Al2O3, (b) Pt–0.30 wt%Rh/Al2O3-CI,
(c) Pt–0.36 wt%Rh/Al2O3-RM and (d) Pt–0.30 wt%Rh/Al2O3-CI-HCl.
Table 4
Peak assignments (cmꢀ1) and relative surface areas under each peak (%) observed during the adsorption of CO on the supported 0.6 wt%Pt–x wt%Rh and 0.3 wt%Rh/Al2O3 catalysts.
Catalyst
Linear on Rh3+
Linear on Rh2+
Gem sym on Rh1+
Linear on Pt0
–
Pt–Rh alloy
–
Gem antisym on Rh1+
Rh/Al2O3
2175
2.2%
–
2145
1.9%
–
2103
62.6%
2103
4.3%
2103
0.2%
2032
33.3%
2032
51.5%
2032
35.7%
2039
44.1%
Pt–0.30Rh/Al2O3-CI
2082
44.2%
2076
34.9%
2076
31.6%
–
Pt–0.36Rh/Al2O3-RM
Pt–0.30Rh/Al2O3-CI-HCl
–
–
–
2061
29.2%
2059
21.4%
2132
0.1%
2103
2.8%
sented, this system being used in the present work as reference
since Ir-based catalysts are known to be the most active and selec-
tive for naphthenic ring opening (RO) [8].
frequencies (TOF) were calculated considering that the dispersion
of the catalysts studied in [8] was similar to those of the catalysts
of the present study, i.e. 75% of dispersion for the platinum catalyst
and 85% for the iridium one. The calculations give a TOF of roughly
1611 hꢀ1 for the 0.3 wt%Pt/Al2O3 of [8] and 1673 hꢀ1 for our
0.6 wt%Pt/Al2O3 at 375 °C, and 4661 hꢀ1 for the 0.9 wt%Ir/Al2O3 of
[8] and 3958 hꢀ1 for our 0.6 wt%Ir/Al2O3 at 275 °C, which can be
considered as similar performances taking into account the differ-
ences in the experimental conditions.
As already observed in our previous study [19], the addition of
rhodium to platinum by surface redox reaction promotes the activ-
ity of the latter, leading to MCP conversions for the bimetallic Pt–
Rh/Al2O3-RM catalysts intermediate between those of monometal-
lic Pt and Rh systems (Fig. 6a). The activity increases with the Rh
content to reach values very close to those of monometallic cata-
lysts containing rhodium or iridium. Moreover, from the addition
of the lowest Rh content to platinum (0.11 wt%), a decrease of
the C1–C5 products formation is observed (Fig. 6b). This beneficial
effect on the RO selectivity increases with the Rh loading, reaching
a plateau from a content of 0.36 wt%Rh. Finally, the bimetallic Pt–
Rh/Al2O3-RM catalysts with the highest Rh contents present RO
selectivities comparable to those of monometallic Ir-based cata-
lyst. The bimetallic Pt–Rh catalysts synthesized by RM method
First, Fig. 6a shows that the blank treatment of the Pt/Al2O3 cat-
alyst involves a drastic decrease in the platinum activity for MCP
conversion. Thus, at 400 °C, the parent catalyst presents a conver-
sion value equal to 50%, against around 10% for the blank catalyst.
This behaviour is related to the enlargement of the Pt particles dur-
ing the blank treatment as described previously. Nevertheless, the
blank treatment does not modify the quantities of C1–C5 products
and consequently the RO selectivity (Fig. 6b). Compared to plati-
num, monometallic Rh and Ir systems are much more active for
MCP hydrogenolysis, but, as expected, only the Ir-based catalyst
presents very high RO selectivity even at high MCP conversion (less
than 2% of C1–C5 products are produced up to 90% of conversion).
In their paper, McVicker et al. [8] obtained 10.4% of MCP conver-
sion with a RO selectivity of 97% with 0.3 wt%Pt/Al2O3 at 350 °C
and 52% of MCP conversion with a RO selectivity of 99% with
0.9 wt%Ir/Al2O3 at 275 °C, in similar experimental conditions, ex-
cept the H2/MCP molar ratio, which was lower (5 instead of 7.5)
and the WHSV, which was equal to 15 hꢀ1 for the platinum catalyst
and 30 hꢀ1 for the iridium one, instead of 12 hꢀ1 in the present
study. In order to compare these results with ours, the turnover

P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
245
100
80
60
40
20
0
equal probability of breaking cyclic bonds, and consequently to a
statistical distribution with 2/5 of n-C6, 2/5 of 2MP and 1/5 of
3MP (i.e. n-C6/3MP and 2MP/3MP ratios both equal to 2); (ii) a
completely selective mechanism producing only 2MP and 3MP,
the breaking of substituted endocyclic C–C bonds being impossible
(i.e. n-C6/3MP = 0 and 2MP/3MP = 2); (iii) a partly selective mech-
anism yielding to a product distribution intermediate between
those obtained with the statistical mechanism and the selective
one. According to the work of Weisang and Gault [41] and McVick-
er et al. [8], the MCP hydrogenolysis on the Ir/Al2O3 catalyst yields
selectively to the formation of 2MP and 3MP (Table 5). On the Rh/
Al2O3 catalyst, a selective mechanism is also observed with a very
low percentage of n-hexane, in accordance with the results of
Teschner et al. [42] and Coq et al. [43]. On the Pt/Al2O3 catalysts
(parent and blank), the product distribution is close to 2/5 of n-
C6, 2/5 of 2MP and 1/5 of 3MP. Consequently, in agreement with
the results of McVicker et al. [8], the RO product distribution can
be considered as resulting from a statistical mechanism, and no
modification was induced by the blank treatment.
Ir
Rh
Pt-xRh
(x≥0.23)
a
Pt-0.11Rh
Pt parent
Pt blank
200 225 250 275 300 325 350 375 400 425
Temperature (°C)
6
b
Pt parent
Pt blank
Rh
5
4
3
2
1
0
The addition of Rh loadings to platinum by RM method de-
creases the n-C6 formation, favouring the production of 2MP and
3MP. Consequently, for the bimetallic Pt–Rh/Al2O3-RM catalysts
series, the n-C6/3MP ratios are close to 0 corresponding to a selec-
tive mechanism as for monometallic Ir- and Rh-based systems.
Pt-0.11Rh
Pt-0.85Rh
Ir
Pt-xRh
(x≥0.23)
3.3.2. Comparison between the bimetallic Pt–Rh catalysts prepared by
different methods
Thereafter, bimetallic Pt–0.30Rh/SiO2-CI, Pt–0.30Rh/Al2O3-CI,
Pt–0.30Rh/Al2O3-CI-HCl and Pt–0.30Rh/Al2O3-CI-T catalysts were
tested in MCP ring opening in order to evaluate the influence of
the preparation method, more exactly the influence of the bimetal-
lic particles size on the RO performances. Compared to the Pt–Rh/
Al2O3-RM sample with a similar Rh content (0.36 wt%), the Pt–
0.30Rh/SiO2-CI catalyst presents the same activity curve for MCP
conversion, whereas lower conversions were observed for the Pt–
0.30Rh/Al2O3-CI system (Fig. 7a). Nevertheless, the MCP conver-
sions obtained on this last catalyst are still higher than those of
the monometallic Pt/Al2O3 catalyst. The two treatments (HCl and
T) performed on the Pt–0.30Rh/Al2O3-CI sample lead to an increase
of its activity, more particularly after the thermal treatment.
Concerning the RO selectivity, the Pt–0.30Rh/SiO2-CI catalyst
presents a similar behaviour than that of the Pt–0.36Rh/Al2O3-
RM system, with C1–C5 product percentages lower than 2% for
more than 90% of conversion (Fig. 7b). On the other hand, the Pt–
0.30Rh/Al2O3-CI catalyst displays RO selectivities comparable or
even lower than those of the monometallic Pt or Rh systems. The
sintering of the metal particles of this bimetallic catalyst, induced
by ‘‘HCl” or ‘‘T” treatment, allows improving the RO selectivity by
decreasing significantly the formation of C1–C5 products at high
MCP conversions.
0
20
40
60
80
100
Conversion (%)
Fig. 6. (a) MCP conversion as function of temperature; (b) yield in C1–C5 products
as function of the MCP conversion for the bimetallic Pt–Rh/Al2O3-RM series: (j)
0.6Pt/Al2O3–parent, (-) 0.6Pt/Al2O3-blank, (+) Pt–0.11Rh/Al2O3-RM, (4) Pt–0.23Rh/
Al2O3-RM, (h) Pt–0.36Rh/Al2O3-RM, (}) Pt–0.46Rh/Al2O3-RM, (ꢁ) Pt–0.85Rh/Al2O3-
RM. In addition: (N) 0.6Rh/Al2O3, (ꢀ) 0.6Ir/Al2O3, (s) experiment at WHSV = 6 hꢀ1
performed with 0.6Pt/Al2O3-parent catalyst in order to complete the C1–C5
products curve shape of Pt-based systems.
are even more selective than the Pt–Rh systems prepared in our
previous study by an alternative surface redox reaction [19], since
the present RM method allows depositing higher Rh contents on Pt
parent sample.
The RO product distribution and the molar ratio of n-hexane to
3-methylpentane (n-C6/3MP) and 2-methylpentane to 3-methyl-
pentane (2MP/3MP) are gathered in Table 5 for the different cata-
lysts, the values being given for MCP conversions as close as
possible. However, previous results concerning platinum-based
catalysts have demonstrated that for MCP conversions higher than
10%, the product distribution does not vary significantly as a func-
tion of the conversion [19]. Gault and et al. [40] showed that the
ring opening product distribution results from three possible
mechanisms: (i) a non-selective mechanism corresponding to an
Finally, all these observations highlight that the catalytic perfor-
mances of the bimetallic Pt–Rh catalysts for MCP hydrogenolysis
Table 5
Yield in n-C6, 2MP, 3MP, n-C6/3MP and 2MP/3MP ratios on monometallic 0.6 wt%M/Al2O3 and supported bimetallic Pt–Rh catalysts prepared by refilling method (RM).
Catalyst
T (°C)
Conversion (%)
2MP (%)
3MP (%)
n-C6 (%)
n-C6/3MP
2MP/3MP
Ir/Al2O3
Rh/Al2O3
Pt/Al2O3-parent
Pt/Al2O3-blank
Pt–0.11Rh/Al2O3-RM
Pt–0.23Rh/Al2O3-RM
Pt–0.36Rh/Al2O3-RM
Pt–0.46Rh/Al2O3-RM
Pt–0.85Rh/Al2O3-RM
275
275
400
400
350
275
275
275
275
73
64
50
10
54
34
45
56
69
68
62
45
44
72
74
75
75
75
29
30
23
24
22
22
21
21
21
3
8
32
32
6
4
4
4
4
0.1
0.3
1.4
1.7
0.3
0.2
0.2
0.2
0.2
2.4
2.1
1.9
1.9
3.3
3.4
3.6
3.6
3.5

246
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
100
80
60
40
20
0
Ir
Rh
a
Pt-0.3Rh/Al2O3-CI-T
Pt-0.36Rh/Al2O3-RM
Pt-0.30Rh/SiO2-CI
Pt-0.3Rh/Al2O3-CI-HCl
Pt-0.3Rh/Al2O3-CI
Pt parent
200
225
250
275
300
325
350
375
400
425
Temperature (°C)
6
5
4
3
2
1
0
Pt parent
Rh
Pt-0.3Rh/Al2O3-CI
b
Pt-0.3Rh/Al2O3-CI-T
Ir
Pt-0.30Rh/SiO2-CI
Pt-0.36Rh/Al2O3-RM
Pt-0.3Rh/Al2O3-CI-HCl
0
20
40
60
80
100
Conversion (%)
9
8
7
6
5
4
c
0.0014
0.0016
0.0018
0.0020
0.0022
1/T
Fig. 7. (a) MCP conversion as function of temperature; (b) yield in C1–C5 products as function of the MCP conversion; (c) Arrhenius plot with TOF in hꢀ1 and T in K for the
supported bimetallic 0.6Pt–Rh catalysts prepared by various methods: (j) 0.6Pt/Al2O3-parent, (-) Pt–0.30Rh/SiO2-CI, (4) Pt–0.30Rh/Al2O3-CI, (h) Pt–0.36Rh/Al2O3-RM, (ꢁ) Pt–
0.30Rh/Al2O3-CI-T, (X) Pt–0.30Rh/Al2O3-CI-HCl. In addition: (N) 0.6Rh/Al2O3, (ꢀ) 0.6Ir/Al2O3, (s) 0.6Pt/Al2O3-parent at WHSV = 6 hꢀ1
.
are closely linked to their preparation procedure that modulates
deeply their average particles size and then the bimetallic particles
structure. In Fig. 7b, the most selective bimetallic Pt–Rh catalysts
for the MCP RO are the Pt–0.36Rh/Al2O3-RM, Pt–0.30Rh/SiO2-CI
and Pt–0.30Rh/Al2O3-CI-HCl catalysts, with RO selectivity compa-
rable to that of Ir/Al2O3 sample. These three bimetallic catalysts
present the largest particles size among all the prepared Pt–Rh sys-
tems (1.7–2.3 nm, Tables 1 and 2). Moreover, the previous charac-
terizations have shown rather close behaviours for these three
samples: (i) similar activities during cyclohexane dehydrogenation
(Fig. 2); (ii) similar H2 consumption profiles during TPR experi-
ments (Fig. 3B) and (iii) similar FTIR spectra of adsorbed CO (Figs.
4 and 5).
3.3.3. Study of the stability of the Pt–Rh catalysts prepared by refilling
method
According to the previous results, the refilling method is a prep-
aration procedure that directly allows synthesizing bimetallic Pt–
Rh/Al2O3 catalysts as selective for the ring opening of methylcycl-
opentane as Ir-based catalysts for Rh content P0.23 wt%. Finally,

P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
247
the stability of such bimetallic systems towards oxidation and un-
der long operating conditions was evaluated and compared with
that of a monometallic Ir catalyst.
100
80
60
40
20
0
First, in order to estimate the effect of an oxidation treatment
onto the optimized bimetallic catalysts, the Pt–0.36Rh/Al2O3-RM
and Pt–0.46Rh/Al2O3-RM catalysts were further submitted to calci-
nation under air at 450 °C during 2 h followed by reduction at
500 °C under H2 during 2 h. Fig. 8 gives the catalytic performances
for MCP conversion obtained for the catalysts after such a treat-
ment (notation: Pt–Rh/Al2O3-RMoxyd.) compared with the fresh
(not oxidized) ones and the monometallic Ir, Rh and Pt catalysts
supported on alumina. Fig. 8a clearly shows that the oxidation
treatment does not influence the activity of the Pt–Rh/Al2O3-RM
catalysts for MCP conversion, since similar conversions are ob-
served for fresh and oxidized catalysts. Moreover, the oxidation
treatment does not modify the RO selectivity since the oxidized
bimetallic catalysts lead to quantities of C1–C5 products still com-
parable with that of the monometallic Ir sample (Fig. 8b).
0
2
4
6
8
10
12
14
16
18
20
Reaction time (h)
Fig. 9. Stability under long operating conditions of bimetallic Pt–0.36Rh/Al2O3-RM
catalyst [mol% of (N) converted MCP and (4) C1–C5 products] compared to 0.6Ir/
Al2O3 catalyst [mol% of (ꢀ) converted MCP and (}) C1–C5 products].
Secondly, the ring opening of MCP was performed on the Pt–
0.36Rh/Al2O3-RM catalyst at 300 °C during 20 h of time-on-stream
(TOS) in order to estimate its stability under long operating condi-
tions. The results were compared with those obtained on the 0.6Ir/
Al2O3 catalyst at 290 °C during the same period. Fig. 9 shows good
stabilities as function of reaction time for both systems, i.e. no sig-
nificant loss of MCP conversion and no significant evolution of the
RO selectivity. Indeed, after 20 h of TOS, the bimetallic Pt–0.36Rh/
Al2O3-RM catalyst always presents a conversion higher than 80%
with less than 2% of C1–C5 products. Consequently, this bimetallic
system always presents high performances for selective ring open-
ing of MCP even after long reaction time.
4. Conclusion
Our previous study [19] revealed that Pt–Rh/Al2O3 catalysts
could be an interesting catalytic system for the ring opening of
methylcyclopentane under pressure, since such samples prepared
by a surface redox method showed RO selectivity higher than those
of monometallic Pt/Al2O3 and Rh/Al2O3 systems, reaching the per-
formances of an iridium-based sample tested in the same condi-
tions. The present work was the continuation of this previous
study, with the aim of characterizing supported bimetallic Pt–Rh
samples prepared by various routes in order to understand the
relationship between their structure properties and their catalytic
performances for the MCP RO.
Two preparation methods were used to prepare Pt–Rh/Al2O3
and Pt–Rh/SiO2 catalysts: (i) the refilling method RM (i.e. a surface
redox reaction) performed in successive steps in order to obtain
sufficient amounts of Rh deposited onto the monometallic Pt cata-
lyst and (ii) the classical technique of coimpregnation CI. Varying
the preparation procedure and the nature of the support leads to
the formation of supported bimetallic Pt–Rh particles of different
average size: >2.0 nm for all the catalysts except those synthesized
on alumina support by CI method (1.2 nm). In order to obtain
comparable sizes by the two preparation techniques, sintering
treatments were performed on the latter inducing a slight enlarge-
ment of the particles.
The performances of the bimetallic Pt–Rh systems for RO of
MCP were compared with those of monometallic catalysts contain-
ing iridium, rhodium and platinum in the same operating condi-
tions under pressure. The bimetallic catalysts present an activity
closed to that of Rh-based monometallic sample, whatever the
preparation method (RM or CI), the support used (Al2O3 or SiO2)
and thus the average size of the metallic particles. The presence
of a bimetallic Pt–Rh interaction was highlighted on all the pre-
pared systems by various characterization techniques: by TEM
coupled with EDX, by temperature-programmed reduction and
by the probe reaction of cyclohexane dehydrogenation. However,
the nature of the Pt–Rh phases differs according to the average par-
ticle sizes of the bimetallic sample. Indeed, cyclohexane dehydro-
genation revealed the presence of a Pt surface enrichment on the
largest bimetallic particles. It is known that under hydrogen atmo-
sphere, the surface of bimetallic Pt–Rh particles is preferably en-
riched with platinum atoms, rhodium atoms being located in the
100
Ir
Rh
a
Pt-0.46Rh/Al2O3-RM
80
Pt-0.46Rh/Al2O3-RMoxyd
Pt-0.36Rh/Al2O3-RM
Pt-0.36Rh/Al2O3-RMoxyd
.
.
60
40
20
0
Pt
200 225 250 275 300 325 350 375 400 425
Temperature (°C)
6
b
Pt
Rh
5
4
3
2
1
0
Ir
Pt-0.46Rh/Al2O3-RM
Pt-0.46Rh/Al2O3-RMoxyd
Pt-0.36Rh/Al2O3-RM
Pt-0.36Rh/Al2O3-RMoxyd
.
.
0
20
40
60
80
100
Conversion (%)
Fig. 8. (a) MCP conversion as function of temperature; (b) yield in C1–C5 products
as function of the MCP conversion after oxidation treatment (specified by oxyd.
index) of bimetallic Pt–Rh/Al2O3-RM catalysts: (4) Pt–0.36Rh/Al2O3-RMoxyd., (+) Pt–
0.46Rh/Al2O3-RMoxyd., (h) Pt–0.36Rh/Al2O3-RM, (}) Pt–0.46Rh/Al2O3-RM. In addi-
tion: (ꢀ) 0.6Ir/Al2O3, (N) 0.6Rh/Al2O3, (j) 0.6Pt/Al2O3.

248
P. Samoila et al. / Journal of Catalysis 276 (2010) 237–248
}
[13] N. Gyorffy, I. Bakos, S. Szabó, L. Tóth, U. Wild, R. Schlögl, Z. Paál, J. Catal. 263
(2009) 372.
core, whereas the reverse is observed under oxidizing atmosphere
[39,44–48]. But in the case of small particles (close to 1.0 nm), all
the atoms are mainly located on the surface inhibiting this surface
enrichment phenomenon. Moreover, FTIR of adsorbed CO showed
that only the bimetallic particles of large enough size (>1.7 nm)
present particular adsorption sites associated in the literature to
the presence of a Pt–Rh alloy. During the MCP ring opening reac-
tion, only these particles lead to RO selectivities comparable with
that of iridium. Consequently, synthesizing Pt–Rh catalysts with
bimetallic interactions does not lead systematically to catalytic
systems as selective as Ir-based catalysts for MCP RO. The particle
size seems to be a determining parameter that induces a different
surface composition of the bimetallic particles. Finally, the pres-
ence of bimetallic particles of small sizes (ꢂ1.0 nm) is not required,
largest particles being more selective for MCP RO, leading to per-
formances similar to those of iridium-based reference catalysts.
In this context, the refilling method allows obtaining Pt–Rh/Al2O3
catalysts with the required structural properties, and moreover,
these systems show a good stability towards oxidation and under
long operating conditions, as well on the level of the MCP conver-
sion as of the RO selectivity.
[14] G. Del Angel, B. Coq, F. Figueras, J. Catal. 95 (1985) 167.
[15] M. Chamam, A. Wootsch, L. Pirault-Roy, I. Boghian, Z. Paál, Catal. Commun. 8
(2007) 686.
[16] M. Chamam, K. Lázár, L. Pirault-Roy, I. Boghian, Z. Paál, A. Wootsch, Appl. Catal.
A 332 (1) (2007) 27.
[17] R.J. Fenoglio, G.M. Nuñez, D.E. Resasco, Appl. Catal. 63 (1990) 319.
[18] R.J. Chimentão, G.P. Valença, F. Medina, J. Pérez-Ramírez, Appl. Surf. Sci. 253
(2007) 5888.
[19] P. Samoila, M. Boutzeloit, C. Especel, F. Epron, P. Marécot, Appl. Catal. A: Gen.
369 (2009) 104.
[20] G.C. Bond, Metal-Catalysed Reactions of Hydrocarbons, Springer, 2005.
[21] B. Nohair, C. Especel, G. Lafaye, P. Marécot, L.C. Hoang, J. Barbier, J. Mol. Catal.
A: Chem. 229 (2002) 117.
[22] C. Carnevillier, F. Epron, P. Marécot, Appl. Catal. A: Gen. 275 (2004) 25.
[23] M. Boudart, A. Aldag, J.C. Benson, N.A. Doughart, C.G. Harkins, J. Catal. 6 (1966)
92.
[24] D.W. Blakely, G.A. Somorjai, J. Catal. 42 (1976) 181.
[25] I. Rodriguez-Ramos, A. Guerrero-Ruiz, J. Catal. 135 (1992) 458.
[26] A. Zecchina, E. Escalona Platero, C. Otero Areán, J. Catal. 107 (1987) 244.
[27] K. Tanaka, J.M. White, J. Catal. 79 (1983) 81.
[28] A. Bourane, O. Dulaurent, D. Bianchi, J. Catal. 196 (2000) 115.
[29] R. Dictor, S. Roberts, J. Phys. Chem. 93 (1989) 2526.
[30] C.A. Rice, S.D. Worley, C.W. Curtis, J.A. Guin, A.R. Tarrer, J. Chem. Phys. 74
(1981) 6487.
[31] D. Panayotov, P. Basu, J.T. Yates Jr, J. Phys. Chem. 92 (1988) 6066.
[32] S. Trautmann, M. Baerns, J. Catal. 150 (1994) 335.
[33] F. Solymosi, M. Pasztor, J. Phys. Chem. 90 (1986) 5312.
[34] P.B. Rasband, W.C. Hecker, J. Catal. 139 (1993) 551.
[35] O. Dulaurent, K. Chandes, C. Bouly, D. Bianchi, J. Catal. 192 (2000) 262.
[36] K. Tomishige, K. Asakura, Y. Iwasawa, J. Catal. 149 (1994) 70.
[37] H.F.J. Van’t Blik, J.B.A.D. Van Zon, T. Hulzinga, J.C. Vis, J.C. Koningsberger, R.
Prins, J. Phys. Chem. 87 (1983) 2264.
References
[1] F.G. Gault, Adv. Catal. 30 (1981) 1.
[2] M. Daage, G.B. McVicker, M.S. Touvelle, C.W. Hudson, D.P. Klein, W.C. Baird,
B.R. Cook, J.G. Chen, S. Hantzer, D.E.W. Vaughan, E.S. Ellis, Stud. Surf. Sci. Catal.
159 (2001) 135.
[3] H. Du, C. Fairbridge, H. Yang, Z. Ring, Appl. Catal. A: Gen. 294 (2005) 1.
[4] M. Boudart, Adv. Catal. 20 (1969) 153.
[38] D.I. Kondarides, Z. Zhang, X.E. Verykios, J. Catal. 176 (1998) 536.
[39] P.J. Lévy, V. Pitchon, V. Perrichon, M. Primet, M. Chevrier, C. Gauthier, J. Catal.
178 (1998) 363.
[40] G. Maire, G. Plouidy, J.C. Prudhomme, F.G. Gault, J. Catal. 4 (1965) 556.
[41] F. Weisang, F.G. Gault, J. Chem. Soc. Chem. Commun. 11 (1979) 519.
[42] D. Teschner, Z. Paál, D. Duprez, Catal. Today 65 (2001) 185.
[43] B. Coq, R. Dutartre, F. Figueras, T. Tazi, J. Catal. 122 (1990) 438.
[44] D.D. Beck, C.L. Di Maggio, G.B. Fisher, Surf. Sci. 297 (1993) 303.
[45] E. Rogemond, N. Essayem, R. Fréty, V. Perrichon, M. Primet, M. Chevrier, C.
Gauthier, F. Mathis, J. Catal. 186 (1999) 414.
[5] G. Leclercq, L. Leclercq, R. Maurel, J. Catal. 50 (1977) 87.
[6] J.M. Dartigues, A. Chambellan, F.G. Gault, J. Am. Chem. Soc. 98 (1976) 856.
[7] Z. Paál, P. Tétényi, Nature 267 (1977) 234.
[8] G.B. McVicker, M. Daage, M.S. Touvelle, C.W. Hudson, D.P. Klein, W.C. Baird Jr.,
B.R. Cook, J.G. Chen, S. Hantzer, D.E.W. Vaughan, E.S. Ellis, O.C. Feeley, J. Catal.
210 (2002) 137.
[9] B. Coq, F. Figueras, J. Mol. Catal. 25 (1984) 87.
[10] M.P. González-Marcos, B. Iñarra, J.M. Guil, M.A. Gutiérrez-Ortiz, Appl. Catal. A:
Gen. 273 (2004) 259.
[11] G. Diaz, F. Garin, G. Maire, S. Alerasool, R.D. Gonzalez, Appl. Catal. A: Gen. 124
(1995) 33.
[12] G. Espinosa, G. Del Angel, J. Barbier, P. Bosch, V. Lara, D. Acosta, J. Mol. Catal. A:
Chem. 164 (2000) 253.
[46] P.S. Dimick, J.L. Kross, E.G. Roberts, R.G. Herman, H.G. Stenger, C.E. Lyman,
Appl. Catal. B: Environ. 89 (2009) 1.
}
[47] N. Gyorffy, A. Wootsch, S. Szabó, I. Bakos, L. Tóth, Z. Paál, Top. Catal. 46 (2007)
57.
}
[48] Z. Paál, N. Gyorffy, A. Wootsch, L. Tóth, I. Bakos, S. Szabó, U. Wild, R. Schlögl, J.
Catal. 250 (2007) 254.