Kinetics of the CO + NO reaction over bimetallic platinum-rhodium on alumina: Effect of ceria incorporation into noble metals

Article abstract of DOI:10.1006/jcat.2002.3519

The influence of ceria on the kinetic behavior of noble metals in NO reduction by CO was studied at 120° and 300°C on a freshly prepared Pt-Rh/Al₂O₃-CeO₂ catalyst, and on an aged catalyst after reaction for 16 hr at 500°C, at 2.3-9 × 10^-3 atm for CO and 1.5-8 × 10^-3 atm for NO, using a differential fixed-bed flow reactor. A rate equation derived from a bifunctional mechanism involving reaction paths either on metal or on ceria could correctly fit rate measurements performed at 120°C. At low temperatures, the refilling step was fast enough to avoid extensive ceria reduction. Consequently, Ce⁴⁺ probably predominated at the surface. The low steady state surface concentration of anionic vacancies prevented strong interaction between ceria and noble metals, preserving adsorption properties similar to those characteristic of Pt-Rh/Al₂O₃. On the other hand, drastic changes in kinetics operated at 300°C, when probably only noble metals catalyze the CO + NO reaction. At high temperatures, ceria was extensively reduced into Ce³⁺, with the subsequent formation of anionic vacancies in the vicinity of metal particles, and that the strong interaction generated with noble metals altered their electronic properties, subsequently modifying their adsorption properties mainly by weakening the CO- and NO-metal bond.

Full text of DOI:10.1006/jcat.2002.3519

Journal of Catalysis 207, 202–212 (2002)
doi:10.1006/jcat.2002.3519, available online at http://www.idealibrary.com on
Kinetics of the CO + NO Reaction over Bimetallic Platinum–Rhodium
on Alumina: Effect of Ceria Incorporation into Noble Metals
P. Granger, ,1 L. Delannoy, J. J. Lecomte, C. Dathy, H. Praliaud,† L. Leclercq, and G. Leclercq
∗
∗
∗
∗
∗
∗
∗
†
Universit e´ des Sciences et Technologies de Lille, Laboratoire de Catalyse, UPRESA 8010, B aˆ t. C3, 59650-Villeneuve d’Ascq C e´ dex, France; and
Laboratoire d’Application de la Chimie a` l’Environnement, UMR 9977, Universit e´ Claude Bernard Lyon I, 69622-Villeurbanne Cedex, France
Received June 15, 2001; revised January 4, 2002; accepted January 4, 2002
ceria increases thermal stability and noble metal dispersion
The influence of ceria on the kinetic behavior of noble metals in (6, 7) and promotes the water–gas shift reaction (8). The
◦
the reduction of NO by CO has been investigated at 120 and 300 C extent of interaction between ceria and noble metals can
on a freshly prepared Pt–Rh/Al2O3–CeO2 catalyst, and on an aged
catalyst after reaction for 16 h at 500 C, within pressure ranges of
induce considerable changes in the reactivity of interme-
diates and their adsorptive properties over noble metals
◦
−
3
−3
2
.3–9 × 10 atm for CO and 1.5–8 × 10 atm for NO, using a diff-
(
9, 10). Presently, the clarification of the mechanisms at
erential fixed-bed flow reactor. These two temperatures have been
selected from temperature-programmed experiments because they
correspond to two very different regimes of activity of noble metals
with ceria. It has been found that a rate equation derived from a
bifunctional mechanism involving reaction paths either on metal or
atomic scale as well as the establishment of reliable kinetic
models (11) are of prime importance for attempts to opti-
mize the interaction between ceria and the noble metals.
Earlier investigations into the kinetics of the CO + NO
reaction over Pt- and Rh-based catalysts, performed in our
◦
on ceria can correctly fit rate measurements performed at 120 C.
In contrast a conventional mechanism earlier proposed for model- laboratory, led to the selection of the following mechanism
◦
ing the CO + NO reaction on Pt–Rh/Al2O3 at 300 C, where only involving the dissociation of NO via step [3], which has been
noble metals are involved, enables modeling of rate measurements assumed to be rate limiting (12, 13).
◦
at 300 C on Pt–Rh/Al2O3–CeO2, which suggests that the interac-
∗
tion between ceria and noble metals is suppressed at that tempera-
ture. The temperature dependency of the rate constants and equi-
librium constants for NO and CO adsorption has been quantified
in order to explain such changes in the kinetic behavior of Pt–Rh/
CO+ ⇔ CO
,
[1]
[2]
[3]
[4]
[5]
[6]
[7]
ads
∗
NO + ⇔ NOads,
∗
NOads + → N + O
ads
ads
,
Al2O3–CeO2.
ꢀc 2002 Elsevier Science (USA)
∗
NOads + Nads → N2O + 2 ,
NOads + Nads → N2 + Oads+ ,
Nads + Nads → N2 + 2 ,
COads + Oads → CO2 + 2 ,
∗
1
. INTRODUCTION
∗
∗
The reduction of CO, NO, and hydrocarbon emissions
fromautomotiveexhaustgaseshasbeenextensivelystudied
in the past two decades (1, 2). Presently the three-way cata- where ∗ stands for a vacant adsorption site.
lyst technology is adopted all over the world and reduces
Contrary to what has been observed on Pt/Al2O3, step
efficiently NOx emissions in a wide range of air-to-fuel ra- [6] can be neglected on Rh-based catalysts (13) but not Pt.
tios. The typical active components of three-way catalysts It was shown that Eq. [8], based on noncompetitive ad-
(
TWCs) are noble metals with various additives, such as ce- sorptions of NO and CO, respectively, on Rh and Pt sites
ria and zirconia (3, 4), which improve their oxygen storage and a surface predominantly populated by adsorbed NO
capacity. However after aging of the catalyst, a loss of oxy- and CO molecules, correctly fits the influence of the par-
gen storage capacity is observed (5). Moreover, the more tial pressures on the rate of NO transformation on Pt–
severe restrictions on NOx emissions set by the legislation Rh/Al2O3 (13). This equation also accounts for the fact
need more-efficient TWCs, which implies a better under- that the nearest-neighbor adsorption site, involved in NO
standing of the nature of the interaction between ceria and dissociation (step [3]), is mainly composed of Pt.
noble metals. It has also been shown that the presence of
2
k3λNO PNO
rNO =
,
[8]
(
1 + λNO PNO) (1 + λCO PCO)
1
To whom correspondence should be addressed. Fax: +33 3 20 33 61 21.
E-mail: Pascal.Granger@univ-lille1.fr.
where k3 is the rate constant of the dissociation of NO
202
0
ꢀ
021-9517/02 $35.00
c 2002 Elsevier Science (USA)
All rights reserved.

THE CO + NO REACTION OVER BIMETALLIC PLATINUM–RHODIUM
203
according to step [3], and λi and Pi are, respectively, the thermal conductivity detector. Products and reactants were
equilibrium adsorption constants and the partial pressures separated on a CTR1 column from Alltech. Prior to reac-
of the reactants (i = CO or NO).
tion, the catalyst was initially in situ reduced in H2, out-
This third paper of the series has been devoted to the gassed in flowing helium at 400 C, and finally cooled at
◦
study of the effect of ceria addition on the catalytic behavior room temperature (RT). Steady state rate measurements at
◦
of noble metals in Pt–Rh/Al2O3–CeO2. Particular attention 300 C were performed on a reduced Pt–Rh/Al2O3–CeO2
◦
has been paid to the synergy effect on the catalytic perfor- further heated in the reacting mixture at 500 C for 16 h. The
mances earlier observed in the CO + NO reaction (14, 15), experimental conditions used for the reaction were a total
−
1
which is generally explained by the involvement of interac- flow rate of 10 L h with 0.05–0.2 g of catalyst mixed with
tions between noble metals and oxygen vacancies (16–18). 0.2–0.8 g of α-Al2O3, corresponding to space velocities in
−
1
Oh (9) suggests that the high rates for the dissociation of the range 25,000–120,000 h
.
NO (step [3]), and the desorption of N2 (step [6]), could The formation of N2 and N2O was only observed from
explain changes in the kinetics of the CO + NO reaction the reduction of NO, according to the following reactions:
when ceria interacts with Rh. Alternately, such a beneficial
effect can be explained from a different mechanism involv-
ing anionic vacancies as active sites for the dissociation of
NO (3, 4, 19).
2
NO + 2CO → N2 + 2CO2,
NO + CO → N2O + CO2.
[9]
2
[10]
This study deals with the kinetics of the CO + NO re-
action on Pt–Rh/Al2O3–CeO2. We first checked to see if
Eq. [8] is still valid for modeling the partial pressure depen-
dency of the reaction rate, or if it implies a rate equation
derived from a bifunctional mechanism involving the re-
dox properties of ceria. The comparison of the values of
k3, λNO, and λCO, with those previously estimated on Pt–
Rh/Al2O3, with additional spectroscopic observations al-
lows discussion of the effective role of ceria in modifying
the properties of noble metals, and their subsequent cat-
alytic performances.
The overall conversions of NO (τNO) and of CO (τCO), re-
lated to the reduction of NO, are given by Eqs. [11] and
[
12], where τN and τN O are, respectively, the conversion of
2
2
0
NO into N2 and N2O. n is the inlet amount of the reactants
i
while nN and nN O are, respectively, the amount of N2 and
2
2
0
N2O produced at any time. P is the inlet partial pressure
i
of the reactant i.
ꢀ
ꢁ
2
nN + nN O
2
2
τNO =
= τN + τN O,
[11]
[12]
0
2
2
n
NO
ꢂ
ꢃn0
2
nN + nN O
τN O
2
. EXPERIMENTAL
2
2
2
NO
τCO =
=
ꢃ P
τN +
0
2
0
n
2
n
CO
CO
2
.1. Catalysts
The catalyst preparation was described in Ref. (15).
ꢂ
0
τN O
2
NO
0
=
τN +
.
2
2
P
CO
Al2O3–12 wt% CeO2 was synthesized by impregnation of
γ -Al2O3 with a solution of cerium nitrate. After water
evaporation, the solid thus obtained was dried overnight
2.3. Spectroscopic Measurements
◦
◦
2.3.1. X-ray photoelectron spectroscopy (XPS). XPS
measurements were performed with a Leybold Heraeus
LHS10 spectrometer equipped with a magnesium anode
at 120 C, then calcined in air at 550 C. X-ray diffraction
XRD) analysis did not reveal bulk detectable cerium ox-
(
ide species. Pt–Rh/Al2O3–CeO2 was prepared according to
(
hν = 1253.6 eV). Before analysis under ultrahigh vacuum
the conventional coimpregnation procedure. The impreg-
◦
(UHV) conditions, the catalyst samples were preheated
in situ in 10 vol% H diluted in helium at 425 C, then trans-
nated sample was successively calcined in air at 350 C and
◦
◦
reduced in pure H2 at 450 C for 2 h. Such a procedure
2
ferred to the analysis chamber. The binding energies (B.E.)
were referenced to the Al 2p level (74.6 eV). The accuracy
on the B.E. values was ± 0.3 eV.
was also applied to the preparation of the reference Pt–
Rh/Al2O3 catalyst (13).
2
.2. Kinetic Experiments
2.3.2. Infrared spectroscopy. Fourier transform infra-
The CO + NO reaction was studied in a fixed-bed flow red (FTIR) spectra were recorded on a Brucker Vector 22
reactor earlier described (13). Differential conditions were spectrometer, under controlled atmosphere, at room tem-
achieved by recycling the outlet gas mixture with a high perature. Prior to analysis, the samples were reduced in situ
◦
recycling ratio, of about 180, in order to minimize exter- in flowing hydrogen at 500 C, then outgassed under vacuum
◦
nal heat and mass transfer restrictions. The catalyst was at 350 C for 1 h. Finally the cell was cooled at RT before
−
2
in a powder form with an average particle size of 80 µm. admitting 1.3 × 10 atm of CO. The catalyst samples were
The feedstream composition was analyzed by means of a maintained in these operating conditions for 1 h; finally CO
Hewlett Packard 5890 series II chromatograph fitted with a was evacuated.

2
04
GRANGER ET AL.
3
. RESULTS
-
-
6
7
-6
-7
-8
-9
1
.6
1.7
1.8
1.9
3
.1. Kinetic Measurements
-
1
E
app = 118.8 kJ mol
-
1
E
app = 84.0 kJ mol
3
.1.1. Preliminary temperature-programmed CO + NO
reaction (TPR). Temperature-programmed experiments
of the CO + NO reaction on a freshly prepared Pt–
Rh/Al2O3–CeO2 were performed from room temperature
◦
to 500 C in stoichiometric conditions with an inlet par-
−3
tial pressure for NO and CO of 5 × 10 atm. Immedi-
ately after admission of the reactant mixture at RT, a sig-
nificant NO conversion was observed without subsequent
CO conversion, which has been assigned to surface re-
-
1
E
app = 69.6 kJ mol
-8
-9
oxidation of Ce3 formed by Ce reduction during the
reductive pretreatment in H2 (20). As shown in Fig. 1,
the CO + NO reaction starts on Pt–Rh/Al2O3–CeO2 be-
+
4+
◦
◦
2
.4
2.5
2.6
2.7
low 50 C (first run) instead of ∼150 C for the less-active
3
-1
1
0 /T, K
Pt–Rh/Al2O3. It is intriguing that the low-temperature con-
versionrangeontheceria-promotedcatalyststabilizesfrom
FIG. 2. Arrh e´ nius plots for the CO + NO reaction on Pt–Rh/Al2O3
◦
2
00 C, while it should rise monotonically with the tem-
◦
(
3
ꢃ) and on Pt–Rh/Al2O3–CeO2 in the range 110–120 C (ꢁ) and 300–
perature. This observation could be related to the con-
◦
−3
30 C (ꢂ), with PNO = PCO = 5 × 10 atm, mCAT = 0.2 g, and total flow
−1
comitant suppression of the synergy effect of ceria on the rate = 10 L h
.
activity of noble metals or/and to the occurrence of deac-
◦
tivation phenomena. Above 280 C, the conversion curves
◦
Rh/Al2O3–CeO2 at 500 C for 72 h; finally the catalyst was
cooled at room temperature in the reactant mixture. A sec-
ond run was performed on the aged Pt–Rh/Al2O3–CeO2.
The low-temperature conversion range is still observable,
which suggests reversible changes in the kinetics of the
CO + NO reaction, with the synergy effect being limited
of both catalysts are superimposed, which suggests that the
mechanism describe on Pt–Rh/Al2O3 probably dominates
on Pt–Rh/Al2O3–CeO2. The reaction was continued on Pt–
1
◦
to the low-temperature range (below 280 C). However a
strongattenuationisalsoobservedforthislow-temperature
conversion range, which probably indicates deactivation
phenomena.
0
0
0
0
0
0
0
0
0
.9
.8
.7
.6
.5
.4
.3
.2
.1
0
3
.1.2. Steady state kinetic measurements. Steady state
◦
kinetic measurements were performed below 140 C, cor-
responding to the low-temperature conversion range, on
a freshly prepared Pt–Rh/Al2O3–CeO2 catalyst in situ re-
duced in hydrogen at 500 C, and above 280 C when the
synergy effect of ceria attenuates.
◦
◦
Effect of temperature. Rate measurements were per-
formed in stoichiometric conditions with inlet NO and CO
−
3
partial pressures of 5 × 10 atm in the temperature range
◦
◦
1
10–120 C and 300–330 C. The apparent activation ener-
gies for the CO oxidation by NO corresponding to these two
ranges of temperature have been estimated from the slope
of the Arrh e´ nius plots shown in Fig. 2. The values obtained
on Pt–Rh/Al2O3–CeO2 are, respectively, 69.6 and 84.0 kJ
5
0
100
150
200
250
300
350
−1
mol . The corresponding value previously estimated on
Temperature,˚C
−1
Pt–Rh/Al2O3, ∼119 kJ mol , emphasizes earlier observa-
tions related to a decrease in the apparent activation energy
after ceria addition to noble metals (9, 16, 21, 22).
FIG. 1. Temperature-programmed conversion curves for the CO +
NO reaction on Pt–Rh/Al2O3 (ꢀ) and on Pt–Rh/Al2O3–CeO2 (ꢁ, first
run); (ꢂ) second run after 72-h aging in the course of the reaction
Influence of the reactant partial pressures on the rate of CO
transformation. The influence of CO and NO partial pres-
◦
−3
at 500 C, with PNO = PCO = 5 × 10 atm, mCAT = 0.2 g, and total flow
rate = 10 L h−
1
.
sures has been investigated at 120 and 300 C, respectively,
◦

THE CO + NO REACTION OVER BIMETALLIC PLATINUM–RHODIUM
205
TABLE 1
CeO also differs according to the temperature range con-
2
ditions (see Table 1). The selectivity for the production of
Effect of Temperature and Partial Pressures of the Reactants on
◦
N2O (SN O), calculated at 300 C, is independent of the NO
the Rate of the CO + NO Reaction on Pt–Rh/Al2O3–CeO2
2
and CO partial pressures, similarly to what was observed
rCOa
NO
N2O
on Pt–Rh/Al2O3 at the same temperature.Conversely SN O
2
(10 3 mol h
−
−1
conv. selectivity
(%)
PNO
PCO
−3
◦
continuously decreases with a rise in PNO at 120 C.
−
3
−1
Temp. (10 atm) (10 atm)
g
cat.)
(%)
00b
3.2. Surface Characterization
3
1.5
5.0
5.0
5.0
5.0
4.9
5.0
5.0
6.0
7.0
8.0
9.0
0.58
6.0
6.0
6.0
5.95
7.6
5.9
6.1
6.5
6.1
6.6
5.3
63
66
66
66
66
67
58
58
58
57
58
2
3
4
4
5
5
5
5
5
5
.5
.5
.5
.9
.0
.0
.0
.0
.0
.0
1.05
1.35
1.81
1.88
1.96
1.96
1.90
1.78
1.75
1.68
3.2.1. X-ray photoelectron spectroscopy observation.
Figure 3 shows X-ray photoelectron (XP) spectra of the Ce
3d core level for Pt–Rh/Al2O3–CeO2 directly introduced
into the analysis chamber and after in situ reduction in H2
◦
at 425 C. Ce 3d5/2 and Ce 3d3/2 photopeaks can be decom-
posed into five contributions: u0, u1, u2, u3, and u4, and
v0, v1, v2, v3, and v4, respectively (23). However the lowest
binding energy contributions, u0 and v0, are usually difficult
to resolve (23–25). As observed in Fig. 3, only four con-
tributions are distinguishable for each group, which have
been assigned in accordance with the literature data. The
spectrum recorded before in situ thermal treatment (see
Fig. 3A) is dominated by the contribution u1(v1), u3(v3),
20c
1
2.1
4.4
4.5
4.2
4.6
4.5
4.2
4.4
3.9
4.5
5.5
7.3
8.4
0.44
0.42
0.55
0.53
0.61
0.67
0.74
0.70
0.95
0.89
0.82
0.92
11.9
10.9
13.1
10.0
9.3
86
78
72
71
66
54
54
70
73
79
79
78
3
3
4
5
6
8
3
4
4
4
4
.2
.4
.3
.2
.6
.0
.5
.3
.3
.1
.2
7.8
6.9
4
+
and u4(v4), which mainly characterizes Ce species in Pt–
Rh/Al O –CeO . As illustrated in Fig. 3B, significant modi-
10.4
14.6
14.4
13.2
13.7
2
3
2
ficationsoccurafterreductionsincethemostintensesignals,
u1(v1) and u2(v2), indicate the predominant formation of
3
+
Ce . ThequantitativeanalysisofXPspectraissummarized
in Table 3 and shows surface Rh-enriched catalysts com-
pared to bulk compositions. It is worthwhile to note that a
reducing treatment induced surface Ce and Pt enrichment,
mainly on Pt–Rh/Al2O3–CeO2, in accordance with the cur-
rent trends reported in the literature (26).
a
Rate of CO oxidation by reaction with NO calculated from the CO
conversion given by Eq. [12].
b
−1
Space velocity, 120,000 h . Rate measurements on an aged catalyst
◦
after reaction at 500 C for 16 h.
c
−1
.
Space velocity, 25,000 h
between 2.1–8.0×10− atm for PNO and 3.9–8.4×10 atm
3
−3
−3
for PCO, and in the range 1.5–5×10 atm for PNO and 4.9–
−
3
9
×10 atm for PCO, respectively. The apparent orders with
respect to NO and CO partial pressures, calculated from the
rate values in Table 1, are listed in Table 2. The comparison
◦
with the CO order value obtained on Pt–Rh/Al2O3 at 300 C
shows significant attenuation of the CO-inhibiting effect on
the reaction rate after ceria addition. The positive value
◦
obtained at 120 C invalidates Eq. [8], which only admits
values between −1 and 0. The selectivity on Pt–Rh/Al2O3–
TABLE 2
Influence of Ceria Addition on the Apparent Reaction Orders
for the CO + NO Reaction over Bimetallic Pt, Rh Catalysts
T
PCO
−3
PNO
−3
◦
ma
na
Catalysts
( C) (10 atm) (10 atm)
b
Pt–Rh/Al2O3
300
300
120
3.0–8.0
4.9–9.0
3.9–8.4
1.5–5.7
1.5–5.0
2.1–8.0
−0.38 0.40
−0.27 0.98
0.16 0.42
Pt–Rh/Al2O3–CeO2
Pt–Rh/Al2O3–CeO2
FIG. 3. X-ray photoelectron spectra of Ce 3d core level in Pt–Rh/
Al2O3–CeO2 without prior thermal treatment (A) and after reduction in
10 vol% He at 425 C (B).
a
b
m
n
NO
rCO = kP × P
.
CO
◦
See Ref. (13).

2
06
GRANGER ET AL.
TABLE 3
Surface Composition of Bimetallic Pt–Rh Supported on γ -Al2O3 and after Modification by Ce Addition
Bulk atm. composition
Surface atm. composition
B.E. (eV)
Catalyst
Pretreatment
Ce/Ce + Al
3.9 × 10−
Rh/Rh + Pt
Pt/Al
Rh/Al
Ce/Ce + Al
Rh/Rh + Pt
Pt 3d3/2
Rh 3d5/2
−
−3
3
−3
−3
Pt–Rh/Al2O3
Initial
0.27
2.0 × 10
2.3 × 10
0.8 × 10
0.9 × 10
2.2 × 10
0.29
0.29
332.4
332.1
308
307.6
Reduceda
2
−3
−3
−3
−3
−2
Pt–Rh/Al2O3–
Initial
0.27
2.0 × 10
5.7 × 10
0.52
0.44
332.9
331.4
307.7
307.4
Reduceda
4.1 × 10
3.2 × 10
6.7 × 10
−2
1
2 wt% CeO2
a
◦
After reduction in 10 vol% H2 in helium at 425 C.
−
1
3
.2.2. Infrared observations. The FTIR spectra in 1840 cm is also detected. According to previous assign-
Figs. 4 and 5 were recorded at room temperature after ments, they mainly characterize, respectively, linear and
−
2
0
exposure to 1.3 × 10 atm of CO for 1 h on prere- bridge-bonded CO mainly adsorbed on Pt (27–30). The
−
1
duced samples and evacuation under vacuum. The desorp- asymmetry of the peak at 2060 cm could be due to a mi-
tion of CO irreversibly adsorbed was further studied be- nor contribution of the asymmetric stretching mode of gem
◦
−1
tween 25 and 300 C. Figure 4a shows IR spectra of CO dicarbonyl species at ∼2030 cm . An increase in temper-
adsorbed on Pt–Rh/Al2O3 mainly characterized by an ab- ature under vacuum activates the desorption of adsorbed
−1
sorption band at 2060 cm ; a broad and weaker band at CO molecules visualized by a subsequent attenuation of
theseabsorptionbandsandashifttowardlowerfrequencies
at the same time. Finally, CO on Pt–Rh/Al2O3 completely
◦
desorbs when the temperature rises to 300 C. Figure 5
FIG. 4. Infrared spectra of CO adsorbed on prereduced Pt–Rh/Al2O3
FIG. 5. Infrared spectra of CO adsorbed on prereduced Pt–Rh/
Al2O3–CeO2 after desorption under vacuum at various temperatures:
◦
◦
after desorption under vacuum at various temperatures: 25 C (a); 100 C
◦
◦
◦ ◦ ◦
25 C (a); 100 C (b); and 200 C (c).
(
b); 200 C (c); and 300 C (d).

THE CO + NO REACTION OVER BIMETALLIC PLATINUM–RHODIUM
207
illustrates changes in the FTIR spectra of CO adsorbed on tion site on the metals; and COads, NOads, and Nads are on
Pt–Rh/Al2O3–CeO2 after desorption at various tempera- a metal site. The direct reaction between gaseous CO and
tures using a similar experimental procedure. As previously oxygen from ceria seems to be in disagreement with ear-
observed on Pt–Rh/Al2O3, linear CO species over noble lier observations. Several workers (3, 4, 9) claimed that the
metals still predominate on the ceria-modified catalyst, with activation of CO by adsorption on the metal is essential to
the presence of bridge-bonded CO species occurring to promote the CO + NO reaction. Previous investigations of
a lesser extent. The highest resolution of the IR band at the CO + NO reaction performed in our laboratory over
−1
+
2
030 cm , assigned to Rh (CO)2 species, is in accor- ceria-promoted catalysts seem also to be in agreement with
dance with the highest surface Rh composition of Pt– this statement (20). The consensus on the refilling of the
Rh/Al2O3–CeO2. No significant changes in the overall oxygen vacancies in ceria either with NO adsorbed or in
intensity of absorption bands is distinguishable in com- the gas phase (steps [14a] and [14b]) seems to be less gen-
parison with Pt–Rh/Al2O3. A further increase in tem- eral. As a matter of fact, Fornasiero et al. (3) preferred step
perature leads to similar features previously observed [14b], but with no really decisive arguments in our opinion.
on Pt–Rh/Al2O3 associated with the desorption of CO. Since the fundamentals of catalysis is based on the acti-
It is interesting that complete desorption of CO on vation of reagents by adsorption, we will assume that NO
Pt–Rh/Al2O3–CeO2 operates at a lower temperature, adsorbed on metals is considerably more reactive than NO
◦
∼
200 C, than that on Pt–Rh/Al2O3, which indicates weaker in the gas phase and consequently will neglect step [14b].
CO adsorption on the prereduced Pt–Rh/Al2O3–CeO2 Recent transient kinetic studies of the CO + NO reaction
catalyst.
on Pt–Rh/γ -Al2O3–CeO2 (32, 33) have shown that NO can
adsorb on ceria but is unable to dissociate, which empha-
sizes the fact that the adsorption of NO over noble metals
is a requisite step for activating its dissociation on anionic
vacancies from ceria.
4
. DISCUSSION
Clearly the kinetic behavior of Pt–Rh/Al2O3–CeO2 for
the CO + NO reaction is strongly dependent on the
temperature-range conditions. As observed, addition of ce-
ria to noble metals lowers the CO-inhibiting effect on the
reactionrateat300 Canddecreasestheapparentactivation
energy. Such tendencies are more pronounced at 120 C,
with a slight positive value for the CO order, which is in-
consistent with Eq. [8], previously established from a mech-
anism involving only steps on the metals (12, 13). Conse-
quently, this shift to positive values should be related to
changes in the reaction mechanism.
Another tentative explanation for the effect of ceria re-
lies on a possible modification of the adsorptive proper-
ties of noble metals in interaction with ceria in accordance
with Oh (9), who evidenced strong modifications in NO
adsorption and dissociation on Rh in the presence of ce-
ria. Finally he explained the promoting effect of ceria with-
out involving the intermediate formation of oxygen vacan-
cies as active sites for NO dissociation. In a recent study of
the CO + NO and CO + N2O reactions on ceria-modified
Pd- and Rh-based catalysts, Holles et al. (34) modeled their
results by assuming a mechanism involving only noble met-
als, but they concluded that the presence of electrophilic
sites at the interface of ceria/noble metals is crucial in the
dissociation of NO. In fact, these earlier investigations as-
sumed the existence of various possible mechanisms which
are sometimes difficult to unambiguously discriminate. This
is well illustrated by the work of Nibbelke et al. (11) on the
CO + O2 reaction on Pt–Rh/γ -Al2O3–CeO2, where they
showed that a monofunctional mechanism, where the reac-
tion is catalyzed only by noble metals, and a bifunctional
one, involving reaction paths between CO adsorbed on the
metal and oxygen from ceria at the metal/ceria interface,
◦
◦
The promotor effect of ceria has already been explained
(
3, 4, 9, 19, 31) by the reactivity of oxygen from ceria, at
the vicinity of CO molecules adsorbed on the metal parti-
cles at the metal/ceria interface, with subsequent formation
of anionic vacancies which are further replenished via the
dissociation of NO adsorbed on the metals. The following
bifunctional mechanism scheme is usually used for explain-
ing the rate enhancement in the CO + NO reaction. It also
involves steps ([13a]–[14b]) on metals previously depicted
in the introduction and on ceria.
∗
COads + O → CO2g + V + ,
[13a]
COg + O → CO2g + V,
NOads + V → Nads + O,
[13b] are equally in agreement with their kinetic results from a
statistical point of view.
[14a]
∗
In the light of these considerations, it seems obvious that
[14b]
NOg + V + → Nads + O,
ceria can exhibit different effects as a consequence either of
∗
NOads + Nads → N2Og + 2 ,
[4]
modifications in the electronic properties of noble metals,
∗
NOads + Nads → N2g + Oads+ ,
[5] and consequently their adsorption properties, or of the par-
∗
ticipation of active oxygen species from ceria. Accordingly
the monofunctional mechanism depicted in the introduc-
Nads + Nads → N2g + 2 ,
[6]
where V and O, respectively, stand for an anionic vacancy tion and the bifunctional one can be equally considered
and a reactive oxygen species from ceria; ∗ is a free adsorp- for modeling rate measurements. Consequently, the rate

2
08
GRANGER ET AL.
equation for CO oxidation by NO could be the sum of the sites or Pt sites covered with adsorbed CO molecules, while
rates of both mechanisms. θNO stands for the coverage of NO adsorbed on Rh. θCO
The rates of steps [3] (r3) and [9] (r9) can easily be es- and θNO can be easily derived:
tablished as functions of the CO and NO partial pressures.
λCO PCO
However the rates of NO conversion into N2 and N2O (rNO)
or of CO oxidation into CO2 (rCO) in the two mechanisms
are not directly related to r3 and r14. According to the over-
all transformation of NO and CO (see Eqs. [9] and [10]),
the expressions of rNO and rCO are
θCO =
θNO =
,
,
[21]
[22]
1 + λCO PCO
λNO PNO
1
+ λNO PNO
and
ꢀ
ꢁ
rNO = 2 rN O + rN = 2(r4 + r5 + r6),
[15]
2
2
1
θ∗ = 1 − θCO =
.
[23]
and
1 + λCO PCO
Equations [21]–[23] can further be introduced into Eq. [20],
which yields Eq. [24]:
rCO = rN O + 2rN = r4 + 2r5 + 2r6,
[16]
2
2
where rN O and rN are, respectively, the rates of NO trans-
2
2
2
k3λNO PNO
formation into N2O and N2 (rN O = r4 and rN = r5 + r6).
r1NO = (
.
[24]
2
2
1 + λCO PCO)(1 + λNO PNO)
Hence, the ratesrCO andrNO can be related to the selectivity
for the production of N2O according to Eq. [17]:
Regarding the rate of CO oxidation via the monofunctional
mechanism, Eqs. [15] and [16] can be simplified, consider-
rN O
r4
2
SN O =
=
.
[17]
2
ing, as mentioned before, that r is negligible, thus Eq. [25]
6
rN O + rN2
r4 + r5 + r6
2
can be obtained,
In the case of the monofunctional mechanism previously es-
tablished on Pt–Rh/Al2O3, involving only metal sites (steps
r1CO
r1NO
r4 + 2r5
k4θNOθN + 2k5θNOθN
2(k4θNOθN + k5θNOθN)
=
=
,
[25]
[26]
[1]–[7]), the following equation can be obtained by applying
2(r4 + r5)
the steady state approximation to adsorbed N atoms:
leading to Eq. [26],
r1CO
dθN
ꢄ
ꢅ
=
0 = r3 − r4 − r5 − 2r6.
[18]
k + 2k
4
5
dt
=
r1NO,
2
(k4 + k5)
Hence,
and finally to Eq. [27],
r1CO = (
r1NO = 2r3 − 2r6.
[19]
kλNO PNO
Now, considering that at high temperature the synergy
effect of ceria is suppressed, it is reasonable to assume
,
[27]
[28]
1 + λCO PCO)(1 + λNO PNO)
◦
that at 300 C the monofuctional mechanism predominates,
with
which is in agreement with the selectivity (SN O) inde-
2
ꢂ
ꢃ
pendence of PNO and PCO observed at that temperature
on Pt–Rh/Al2O3–CeO2 and Pt–Rh/Al2O3. Such selectiv-
ity behavior implies necessarily that the rate of step [6]
is negligible in comparison of that of steps [4] and [5], as
already explained (13). Consequently for the mechanism
on metals, r6 can be neglected in Eq. [19], leading to the
equation
k4 + 2k5
k + k
k = k3
.
4
5
Regarding the bifunctional mechanism involving active
oxygen species from ceria and which is supposed to occur at
low temperature, the rate of CO2 formation is that of step
[13],
∗
r2CO = k θ θ ,
[29]
∼
13 CO O
r1NO = 2r3 = 2k3θNOθ ,
[20]
∗
where θO is the fraction of surface oxygen from ceria. As
demonstrated in the Appendix, Eq. [30] can be established:
where θi and θ are, respectively, coverage for the reactant
i and the fraction of vacant metal sites.
If we still assume for Pt–Rh/Al2O3–CeO2 preferential
adsorptions of NO and CO, respectively, on Rh and Pt,
and that NO dissociates on a nearest-neighbor Pt site, as
r2CO
k13k14λNOλCO PNO PCO
=
.
k14λNO PNO(1 + λCO PCO) + k13λCO PCO(1 + λNO PNO)
we have already shown to probably occur on Pt–Rh/Al2O3
∗
(
13), then θ and θCO represent the fraction of vacant Pt
[30]

THE CO + NO REACTION OVER BIMETALLIC PLATINUM–RHODIUM
TABLE 5
209
Consequently the rate of the overall CO2 production on
Pt–Rh/Al2O3–CeO2 is given by Eq. [31]:
Temperature Dependence on the Kinetic and Thermodynamic
◦
Constants for the CO + NO Reaction at 120 C on Pt–Rh/Al2O3–
rCO
CeO2
kλNO PNO
= (
1 + λCO PCO)(1 + λNO PNO)
Activation Adsorption
Preexponential factor
energy,
enthalpy,
k13k14λNOλCO PNO PCO
A
mol h
fi
E
ꢀHads,i
(kJ mol
+
.
−1 −1
−1
−1
−1
(
g
)
(atm
)
(kJ mol
)
)
k14λNO PNO(1 + λCO PCO) + k13λCO PCO(1 + λNO PNO)
1
1
1
2
[
31] Steps on the metals 1.0 × 10
110.2
104.7
48.5
Step [13a]
6.1 × 10
3.3 × 10
3
Step [14a]
NO adsorption
CO adsorption
Such an equation cannot be linearized and, further, be
solved using a usual graphic method. Only an optimization
method, earlier described (12), using nonlinear regression
analysis can be applied for adjusting the rate constant k, k13,
and k14 and the equilibrium constants of the reactants λi .
The adjustment routine has been achieved using the solver
set up on Excel 5 from Microsoft. It consists of minimizing
the square difference between experimental rates, shown
in Table 1, and those calculated from Eq. [31].
−
4
4.9 × 10
−44.3
−24.9
0.38
the equilibrium constants of the reactants λi and the corre-
sponding values for the activation energies (E, E13, and E14)
and for the adsorption enthalpies of the reactants ꢀHads,i
in Eqs. [32]–[35]:
ꢂ
ꢃ
E
k = A exp −
,
[32]
[33]
[34]
[35]
4
.1. Effect of Ce Additive on the Kinetic Behavior
of Noble Metals at 120 C
RT
◦
ꢂ
ꢃ
E13
k13 = A13 exp −
,
,
The results obtained from the optimization routine, us-
RT
◦
ing the set of rates at 120 C, are reported in Table 4. The
value for k, the rate constant related to the dissociation over
noble metals, is much smaller than for k13 and k14. Conse-
quently, at this temperature the dissociation of NO on the
metals is very slow, which leads to the conclusion that a
reaction path with the conventional mechanism involving
only noble metals is unable to model the CO + NO reac-
ꢂ
ꢃ
E14
k14 = A14 exp −
RT
ꢂ
ꢃ
ꢀ
Hads,i
λi = fi exp −
.
RT
The adjusted values are listed in Table 5. As observed in
Fig. 6, they lead to reasonable agreement between experi-
mental and predicted rates using the optimized parameters.
◦
−5
−1 −1
−1 −1
tion at 120 C (r1CO = 2.2×10 mol h
g
in comparison
−4
−3
with r2CO = 7.1 × 10 mol h
g
for PNO = 4 × 10
−3
atm and PCO = 4 × 10 atm), the bifunctional reaction
paths [13] and [14] being probably the most representative.
Accordingly, at low temperature, the rate equation of CO
transformation rCO ∼r2CO.
6
4
2
0
Using rate measurements obtained at various temper-
atures, as shown in Fig. 1, we have attempted to model
the temperature dependency of parameters included in
Eq. [31]. It consists of determining the preexponential fac-
tors A, A13, A14, and fi of the rate constant k13, k14, and of
TABLE 4
Optimized Kinetic and Thermodynamic Constants for the
◦
CO + NO Reaction at 120 C on Pt–Rh/Al2O3–CeO2
−
1
−1
Rate constants of reaction paths involving ceria (mol h
g )
−
3
k13
k14
89.5 × 10
1.2 × 10−
3
−
1
−1
g )
0
2
4
6
Rate constant of paths involving the metals (mol h
-
3
-1 -1
−
4
k
2.7 × 10
Experimental rate rCO, 10 mol h
g
Equilibrium adsorption constants (atm−1)
FIG. 6. Correlation between experimental and calculated rates on Pt–
λNO
λCO
353
751
◦
Rh/Al2O3–CeO2 in the temperature ranges 110–120 C (ꢀ) and 300–
◦
3
30 C (ꢂ).

2
10
GRANGER ET AL.
Two major comments arise from the examination of these
optimized parameters. First, the energy of activation corr-
esponding to the step of oxygen consumption (step [13a])
is substantially higher than that of the refilling step of an-
ionic vacancies (step [14a]). Consequently at high temper-
ature, the extent of ceria reduction will be high and θO
very low, which will lower the rate of step [13a]. This is in
good agreement with the interpretation of the loss of syn-
ergy effect at high temperature by ceria reduction. Second,
the values for the adsorption enthalpies of CO and NO
are comparable with those calculated on the reference
Pt–Rh/Al2O3 catalyst according to the margin of error
TABLE 6
Kinetic and Thermodynamic Constants for the CO + NO
◦
Reaction at 300 C on Pt–Rh/Al2O3–CeO2
Adsorption
constant (atm
a
−1
)
Rate contant, k
−
1
−1
Catalyst
(mol h
g
)
λ
NO
λ
CO
3
Aged Pt–Rh on
Al2O3–CeO2
Pt–Rh/Al2O3
38.5 × 10−
12.1
505
29.9
4.7 × 10−3
122
a
Noncompetitive adsorption of NO and CO, respectively, on Rh and
Pt site.
(
see Table 7). This observation suggests that no significant
interaction between ceria and noble metals occurs at low
temperature, which could alter their electronic properties
stantiates a lessening of the CO-inhibiting effect after ce-
ria addition. With regard to the rate constant k (per gram
of catalyst), no significant conclusion can be drawn rel-
ative to the ceria-induced effect on the rate of NO dis-
sociation since the determination of the metal dispersion
(
35) and subsequently their adsorption properties.
4
.2. Kinetic Behavior of the Aged Pt–Rh/Al2O3–CeO2
at 300 C
◦
Let us remember that steady state kinetic experiments is difficult to assess from hydrogen or CO chemisorption
were performed after aging of the reactant mixture at on Pt–Rh/Al2O3–CeO2 (36, 37). Additionally XPS and in-
◦
5
00 C for 16 h, which could lead to some modifications frared spectroscopy of CO adsorption observations under-
in its catalytic performances mainly due to deactivation, as line a surface rhodium composition higher than that of
indicated by temperature-programmed (TP) experiments, Pt–Rh/Al2O3. Hence, it is expected that the intrinsic rate
which showed that the beneficial effect of ceria on NO con- constant (per surface Rh atom) varies differently than the
−
3
−3
version is significantly attenuated for a longer period of specific rate constant (4.73 × 10 against 38.5 × 10 mol
−
1
−1
time.
h
g ), with Rh being much more active than Pt for the
As previously mentioned, at 300 C, ceria is probably ex- dissociation of NO (38, 39). Both the highest surface Rh
◦
tensively reduced. At that temperature, the ratio k13/k14 composition and the surface modifications, related to sin-
4
would be ∼1.4 × 10 , as calculated from the values shown tering reactions (40) which might occur during reaction at
◦
in Table 5 and the reaction paths including ceria should be 500 C, could also cause changes in the rate of dissociation
negligible. Consequently, the transformation of NO would of NO, which is known to be structure sensitive (41).
◦
predominantly involve noble metals at 300 C and would
The lower activation energy of step [14a] relative to
obey the classical Langmuir–Hinshelwood mechanism ear- step [13a], calculated using rate measurements at low-
◦
lier described. It is noteworthy that the selectivity values for temperature, and the low value of k14/k13 at 300 C indicate
4
+
3+
the transformation of NO into N2O (Table 1) are insensitive extensive reduction of Ce into Ce during pretreatment
to changes in the partial pressure of the reactants, mimick- at 500 and 300 C. From this explanation it seems obvious
ing the selectivity behavior of Pt–Rh/Al2O3 and Rh/Al2O3 that kinetic data obtained at 300 C reflect the influence of
◦
◦
(
13). Such an observation supports the previous conclusion reduced ceria, with the subsequent formation of anionic
and can probably be related to a preferential adsorption of vacancies interacting with noble metals, on the kinetic be-
NO on Rh in Pt–Rh/Al2O3–CeO2, since the subsequent dis- havior of Pt–Rh/Al2O3–CeO2.
sociation of adsorbed NO molecules yielding chemisorbed
N atoms on Rh is consistent with the fact that only steps λi ), the activation energy E corresponding to the rate of
4]–[6] are involved in the transformation of NO. The in- NO dissociation on the metal, and ꢀHads,i are reported
The values of preexponential factors A and fi (for k and
[
sensitivity of SN O to PNO also suggests that the associa- in Table 7. Figure 6 shows that a good quality of fit be-
2
tive desorption of chemisorbed N atoms via step [6] is slow tween experimental and predicted rates, using the adjusted
in comparison with the fast steps [4] and [5], as already parameters shown in Table 7, is obtained. Comparison of
◦
discussed (13). Consequently, rCO ∼r1CO at 300 C on Pt– these parameters with those previously calculated on Pt–
Rh/Al2O3–CeO2.
Rh/Al2O3 (42) shows a significant decrease in the activa-
The optimized values for k, λCO, and λNO are reported tion energy of NO dissociation and in the adsorption en-
in Table 6. Comparison of the parameters k and λi with thalpies of the reactants. The weakening of the CO–Pt and
◦
those estimated on the reference Pt–Rh/Al2O3 at 300 C NO–Rh bonds could be due to anionic vacancies interacting
shows significant differences mainly associated with a sharp with noble metals in Pt–Rh/Al2O3–CeO2 and leading to a
decrease in λNO and to a lesser extent in λCO, which sub- strong metal-support-type interaction, which usually alters

THE CO + NO REACTION OVER BIMETALLIC PLATINUM–RHODIUM
211
TABLE 7
Optimised Kinetic and Thermodynamic Parameters for the CO + NO Reaction on Aged Pt–Rh/Al2O3–CeO2
Preexponential factor
Adsorption enthalpy
A
fCO
(atm
fNO
−1
ꢀHads,CO
(kJ mol
ꢀHads,NO
(kJ mol
Activation energy,
E
−
1
−1
−1
−1
−1
a
−1
(kJ mol )
Catalyst
(mol g
h
)
)
(atm
)
)
)
9
9
−2
−2
Aged Pt–Rh on Al2O3–CeO2
Pt–Rh/Al2O3
1.6 × 10
1.3 × 10
0.52
0.86
2.3 × 10
5.8 × 10
−19.4
−26.1
−30.1
−43.8
117.2
123.8
a
∗
∗
∗
∗
Activation energy associated with the rate-limiting step (NO + → N + O ).
the chemisorptive capacity of noble metals, particularly to- Consequently Ce4 probably predominates at the surface.
ward hydrogen adsorption (43). Of course no conclusive The low steady state surface concentration of anionic va-
statements relative to either the nature and/or the extent cancies probably prevents strong interaction between ceria
of interaction in Pt–Rh/Al2O3–CeO2 can be drawn from and noble metals, preserving adsorption properties similar
+
◦
examination of our kinetic data at 300 C. Infrared spectro- to those characteristic of Pt–Rh/Al2O3. On the other hand,
◦
scopic measurements give additional arguments for discus- drastic changes in kinetics operates at 300 C, when prob-
sion relative to the difference observed in the strength of ably only noble metals catalyze the CO + NO reaction. It
the CO–metal bond. Clearly CO is more weakly adsorbed was also found that at high temperatures ceria is probably
3
+
over noble metals on prereduced Pt–Rh/Al2O3–CeO2 than extensively reduced into Ce , with the subsequent forma-
on Pt–Rh/Al2O3. These spectroscopic features corroborate tion of anionic vacancies in the vicinity of metal particles,
our kinetic data, which showed a significant decrease in and that the strong interaction generated with noble metals
the adsorption enthalpy of NO and CO on Pt–Rh/Al2O3– alters their electronic properties, subsequently modifying
CeO2. Such observations are not presently explained and their adsorption properties mainly by weakening the CO–
necessitate further investigation. These modifications could and NO–metal bond.
be due either to deactivation in the course of the reaction,
via an accumulation of isocyanate, which was found to alter
the adsorption properties of Pt (44), or to a weakening of
APPENDIX
the metal–CO bond, as previously proposed by Chafik et al.
2
The rate of CO production according to a bifunctional
(
45) for the CO + NO reaction on Rh/TiO2. These authors mechanism (as shown earlier) is
explained their results as the unavailibilty of unoccupied
states of Rh metal to form a σ-bond and a subsequent lesser
back-donation into the 2π of CO even when Rh exhibits
r2CO = k θ θ .
[29]
13
CO O
∗
electron-enriched d orbitals.
θO can be obtained by applying the steady state approxima-
tion to oxygen species from ceria,
5
. CONCLUSION
dθO
=
0 = k14θNOθV − k13θCOθO,
[36]
dt
The influence of Ce additive on the catalytic perfor-
mances of three-way bimetallic Pt–Rh/Al2O3 in the CO +
NO reaction has been investigated. Preliminary TP exper-
iments evidenced a beneficial effect of ceria on the con-
version of NO mainly at low temperatures. Above 280 C,
Pt–Rh/Al2O3–CeO2 behaves apparently similarly to Pt–
with
1
= θV + θO,
[37]
◦
where θV is the fraction of oxygen vacancies on ceria. Thus,
◦
Rh/Al2O3. An extensive kinetic study at 120 and 300 C,
k14θNO
θO =
.
[38]
corresponding to temperatures where, respectively, the
ceria-induced effect on activity is predominant and sup-
pressed, has been performed in an attempt to explain these
results. It was found that a bifunctional mechanism in-
volving reaction of oxygen from ceria with adsorbed CO
molecules on the metals, then dissociation of adsorbed NO
molecules on anionic vacancies, correctly fits rate mea-
surements at 120 C. At low temperatures, the refilling Accounting for preferential adsorptions of CO and NO,
step [14a] is fast enough to avoid extensive ceria reduction. respectively, on Pt and Rh, θNO and θCO can be replaced by
k14θNO + k13θCO
Consequently,
1
3 14 CO NO
k k θ θ
r2CO =
.
[39]
k14θNO + k13θCO
◦

2
12
GRANGER ET AL.
17. Chojnacki, T., Krause, K., and Schmidt, L. D., J. Catal. 128, 161 (1991).
Eqs. [21] and [22], leading ultimately to Eq. [30]:
1
1
2
8. Golunski, S. E., Hatcher, H. A., Rajaram, R. R., and Truex, T. J., Appl.
Catal. B 5/4, 367 (1995).
9. Harrison, B., Diwell, A. F., and Hallett, C., Platinum Met. Rev. 32, 73
(1988).
0. Granger, P., Lamonier, J.-F., Sergent, N., Aboukais, A., Leclercq, L.,
and Leclercq, G., Top. Catal. 16/17, 89 (2001).
r2CO
k13k14λNOλCO PNO PCO
=
.
k14λNO PNO(1 + λCO PCO) + k13λCO PCO(1 + λNO PNO)
[
30]
21. Zafiris, G. S., and Gorte, R. J., J. Catal. 143, 86 (1993).
22. Serre, C., Garin, F., Belot, G., and Maire, G., J. Catal. 141, 9 (1993).
23. Romeo, M., Bak, K., El Fallah, J., Lenormand, F., and Hilaire, L., Surf.
Int. Anal. 2, 508 (1993).
ACKNOWLEDGMENTS
We gratefully acknowledge the Institut Francais du Petrole and the
CNRS, which initiated this work through a research group, “Groupement
de Recherche Pot Catalytique,” and Mr. J. Billy, who conducted infrared
measurements.
2
2
4. Lenormand, F., Hilaire, L., Kili, K., Krill, G., and Maire, G., J. Phys.
Chem. 92, 2561 (1998).
5. Ramarason, E., Tempere, J.-F., Guilleux, M. F., Vergand, F.,
Roulet, H., and Dufour, G., J. Chem. Soc., Faraday Trans. 88, 1211
(
1992).
2
2
6. Nieuwenhuys, B. E., Adv. Catal. 44, 259 (1999).
7. Unpublished results obtained at the Institut de Recherches sur la
Catalyse.
REFERENCES
1
2
3
. Taylor, K. C., Catal. Rev.–Sci. Eng. 35, 457 (1993).
. Schelef, M., and Graham, G. W., Catal. Rev.–Sci. Eng. 36, 433 (1994).
. Fornasiero, P., Ranga Rao, G., Kaspar, J., L’Eratio, F., and
Graziani, M., J. Catal. 175, 269 (1998).
28. Yang, A. C., and Garland, C. W., J. Phys. Chem. 61, 1504 (1957).
29. Yates, D. J. C., Murel, L. L., and Prestridge, E. B., J. Catal. 84, 41 (1979).
30. Solymosi, P., J. Phys. Chem. 89, 4789 (1985).
31. Oh, S. H., and Eickel, C. C., J. Catal. 112, 543 (1988).
32. Harmsen, J. M. A., Hoebink, J. H. B. J., and Shouten, J. C., Catal. Lett.
71, 81 (2001).
4
. Ranga Rao, G., Fornasiero, P., Di Monte, R., Kaspar, J., Vlaic,
G., Balducci, G., Meriani, S., Gubitosa, G., Cremona, A., and
Graziani, M., J. Catal. 162, 1 (1996); Ranga Rao, G., Kaspar, J., Di
Monte, R., Meriani, S., Gubitosa, G., and Graziani, M., Catal. Lett. 24,
33. Harmsen, J. M. A., Hoebink, J. H. B. J., and Shouten, J. C., Stud. Surf.
Sci. Catal. 133, 349 (2001).
1
07 (1996).
5
6
7
8
9
. Schmieg, S. J., and Belton, D. N., Appl. Catal. 58, 131 (1979).
. Sanchez, M. G., and Gazquez, J.-L., J. Catal. 104, 120 (1987).
. Summer, J. C., and Ausen, S. A., J. Catal. 58, 131 (1979).
. Kim, G., Ind. Eng. Chem. Prod. Res. Dev. 21, 267 (1982).
. Oh, S. H., J. Catal. 124, 477 (1990).
34. Holles, J. H., Switzer, M. A., and Davis, R. J., J. Catal. 190, 247
(2000).
35. Martinez-Arias, A., Soria, J., and Conesa, J. C., J. Catal. 168, 364
(1997).
36. Trovarelli, A., Dolcetti, G., de Leitenburg, C., Kaspar, J., Finetti, P.,
and Santoni, A., J. Chem. Soc., Faraday Trans. 88, 1311 (1992).
37. Cordatos, H., and Gorte, R. J., J. Catal. 159, 112 (1996).
38. Gorte, R. J., and Schmidt, L. D., Surf. Sci. 109, 367 (1990).
39. Wickam, D. T., Banse, B. A., and Koel, B. E., Surf. Sci. 243, 83 (1991).
40. Krause, K. R., and Schmidt, L. D., J. Catal. 140, 424 (1993).
41. Peden, C. H. F., Belton, D. N., and Schmieg, S. J., J. Catal. 155, 204
(1995).
42. Granger, P., Lecomte, J.-J., Leclercq, L., and Leclercq, G., Appl. Catal.
208 (2001).
43. Tauster, S. J., Fung, S. C., and Garten, R. L., J. Am. Chem. Soc. 100,
170 (1978).
44. Lorimer, D., and Bell, A. T., J. Catal. 59, 223 (1979).
45. Chafik, T., Kondarides, D. I., and Verykios, J. Catal. 190, 446
(2000).
1
1
1
1
1
1
0. Loof, P., Kasemo, B., Anderson, S., and Frestad, A., J. Catal. 130, 181
(1991).
1. Nibbelke, R. H., Campman, M. A. J., Hoebink, J. H. B. J., and
Marin, G., J. Catal. 171, 358 (1997).
2. Granger, P., Dathy, C., Lecomte, J.-J., Leclercq, L., Prigent, M.,
Mabilon, G., and Leclercq, G., J. Catal. 173, 304 (1998).
3. Granger, P., Dathy, C., Lecomte, J.-J., Leclercq, L., and Leclercq, G.,
J. Catal. 175, 194 (1998).
4. Lecomte, J. J., Granger, P., Leclercq, L., Lamonier, J.-F., and
Aboukais, A., Colloid Surf. A 158, 241 (1999).
5. Leclercq, G., Dathy, C., Mabilon, G., and Leclercq, L., in “Cataly-
sis and Automotive Pollution Control II” (A. Crucq, Ed.), p. 181.
Elsevier, Amsterdam, 1991.
1
6. Shwartz, J. M., and Schmidt, L. D., J. Catal. 148, 22 (1994).