An investigation of the NO/H2/O2 (Lean-deNOx) reaction on a highly active and selective Pt/La0.5Ce0.5MnO3 catalyst

Article abstract of DOI:10.1006/jcat.2000.3101

The NO/H₂/O₂ reaction has been studied under lean-burn conditions in the 100-400°C range over 0.1 wt% Pt supported on La_0.5Ce_0.5MnO₃ (mixed oxide containing LaMnO₃, CeO₂, and MnO₂ phases). For a critical comparison, 0.1 wt% Pt was supported on γ-Al₂O₃ and tested under the same reaction conditions. The maximum in the NO conversion has been observed at 140°C (74% conversion) for the Pt/La_0.5Ce_0.5MnO₃ and at 125°C (66% conversion) for the Pt/γ3-Al₂O₃ catalyst using a GHSV of 80,000 h^-1. Addition of 5% H₂O in the feed stream influenced the performance of the catalyst in a positive way. In particular, it widened the operating temperature window of the catalyst above 200°C with appreciable NO conversion and had no negative effect on the stability of the catalyst for a 20-h run on reaction stream. Remarkable N₂ selectivity values in the 80-90% range have been observed on the Pt/La_0.5Ce_0.5MnO₃ catalyst in the 100-200°C range either in the absence or in the presence of water in the feed stream. This result is reported for the first time for the NO/H₂/O₂ lean-deNO_x reaction at least on Pt-based catalysts. A maximum specific integral reaction rate of 397 μmol of N₂/s.g of Pt metal was measured at 140°C during reaction with 0.25% NO/1% H₂/5% O₂/5% H₂O/He gas mixture on the 0.1 wt% Pt/La_0.5Ce_0.5MnO₃ catalyst. This value was found to be higher by 40% than that observed on the 0.1 wt% Pt/γ-Al₂O₃ catalyst at 125°C, and it is the highest value ever reported in the 100-200°C range. A TOF value of 0.49 s^-1 was calculated at 140°C for the Pt/La_0.5Ce_0.5MnO₃ catalyst. Temperature-programmed desorption (TPD) of NO and transient titration experiments of the catalyst surface following reaction have revealed important information concerning several mechanistic steps of the present catalytic system. A hydrogen-assisted NO dissociation step and a nitrogen-assisted mechanism for N₂ and N₂O formation are proposed to explain all the transient experiments performed in a satisfactory manner.

Full text of DOI:10.1006/jcat.2000.3101

Journal of Catalysis 197, 350–364 (2001)
doi:10.1006/jcat.2000.3101, available online at http://www.idealibrary.com on
An Investigation of the NO/H /O (Lean-deNO ) Reaction on a Highly
2
2
x
Active and Selective Pt/La Ce MnO Catalyst
0.5
0.5
3
�
�
,1
C. N. Costa, V. N. Stathopoulos,† V. C. Belessi,† and A. M. Efstathiou
�
Department of Chemistry, University of Cyprus, P.O. Box 20537, CY 1678 Nicosia, Cyprus; and †Department of Chemistry,
University of Ioannina, Ioannina 45 110, Greece
Received August 7, 2000; revised October 27, 2000; accepted October 30, 2000; published online January 12, 2001
oxidizing conditions has been undertaken and reported in
The NO/H2/O2 reaction has been studied under lean-burn con- the literature as recently reviewed (1–3). Today’s great con-
�
ditions in the 100–400 C range over 0.1 wt% Pt supported on cern about the increasing emissions of carbon dioxide to the
La0.5Ce0.5MnO3 (mixed oxide containing LaMnO3, CeO2, and MnO2
phases). For a critical comparison, 0.1 wt% Pt was supported on � -
atmosphere and the problems faced resulting from the use
of NH3 as a reducing agent (4) lead to a demand for find-
Al2O3 and tested under the same reaction conditions. The maximum
ing appropriate non-carbon-containing reducing molecules
�
in the NO conversion has been observed at 140 C (74% conversion)
for the catalytic removal of NO from combustion exhaust
�
for the Pt /La0.5Ce0.5MnO3 and at 125 C (66% conversion) for the
streams. Hydrogen is one of the gases present in the exhaust
stream of automobiles and has been reported to be very ef-
fective as a reducing agent for the NO/H2 reaction (5–16).
�
1
Pt /� -Al2O3 catalyst using a GHSV of 80,000 h . Addition of 5%
H2O in the feed stream influenced the performance of the catalyst in
a positive way. In particular, it widened the operating temperature
�
x
Hydrogen could also be used to reduce NO emissions of
window of the catalyst above 200 C with appreciable NO conver-
sion and had no negative effect on the stability of the catalyst for stationary sources. However, only a few attempts have been
a 20-h run on reaction stream. Remarkable N2 selectivity values in reported for the reduction of NO by H2 under oxygen-rich
the 80–90% range have been observed on the Pt /La0.5Ce0.5MnO3 conditions (17–20) because H can be easily consumed by
2
�
catalyst in the 100–200 C range either in the absence or in the
reacting with O2 to form water under the applied reaction
presence of water in the feed stream. This result is reported for
the first time for the NO/H2/O2 lean-deNOx reaction at least on
Pt-based catalysts. A maximum specific integral reaction rate of
97 �mol of N2/s.g of Pt metal was measured at 140 C during re-
action with 0.25% NO/1% H2/5% O2/5% H2O/He gas mixture on
the 0.1 wt% Pt /La0.5Ce0.5MnO3 catalyst. This value was found to
be higher by 40% than that observed on the 0.1 wt% Pt /� -Al2O3
conditions (21, 22).
Supported platinum catalysts have been found to be the
most active for the NO/H2/O2 lean-NOx reaction at low
temperatures (T < 200 C) (17–20). Yokota et al. (20) have
reported catalytic results on the NO reduction with H2 in
the presence of O2 over the Pt–Mo–Na/SiO2 catalyst, while
�
3
�
�
Frank et al. (17) reported kinetic results of the NO/H2/O2
catalyst at 125 C, and it is the highest value ever reported in the
�
� 1
�
1
00–200 C range. A TOF value of 0.49 s was calculated at 140 C reaction on the Pt–Mo–Co/�-Al2O3 catalyst. The latter two
for the Pt /La0.5Ce0.5MnO3 catalyst. Temperature-programmed des- catalyst formulations were found to produce substantially
orption (TPD) ofNO and transient titration experimentsofthe cata- lower amounts of N O (N selectivity at about 75% ) as
2
2
lyst surface following reaction have revealed important information
concerningseveral mechanistic stepsof the present catalytic system.
A hydrogen-assisted NO dissociation step and a nitrogen-assisted
mechanism for N2 and N2Oformation areproposed toexplain allthe
transient experiments performed in a satisfactory manner.
�
c 2001 Academic Press
Key Words: lean de-NOx; NO reduction; NO TPD; perovskites.
compared to conventional supported Pt catalysts (e.g., Pt /
Al2O3, Pt /SiO2) which are known to present lower N2 se-
lectivity values (SN = 40–50% ). On the other hand, it was
2
reported (20) that the Pt–Mo–Na/SiO2 catalyst shows a rel-
atively low stability under lean-burn conditions in the pres-
ence of water.
In our laboratory we have initiated a research program
concerning the use of certain perovskite-type materials as
potential catalysts or supports of active metals for the selec-
tive reduction of NO under oxygen-rich conditions. It was
1
. INTRODUCTION
Much research concerning the selective catalytic reduc- found that La0.5Ce Sr
tion (SCR) of NO by NH3 and hydrocarbons under net low catalytic activity and high selectivity toward N forma-
3
MnO solids present a relatively
x
0.5� x
2
�
tion for the NO/H2/O2 reaction in the 250–500 C range.
1
However, the combination of such a selective system with
a very small amount of Pt (0.1 wt% ) results in a catalyst
To whom correspondence should be addressed. Fax: (+3572)-339060.
E-mail: efstath@ucy.ac.cy.
350
0021-9517/01 $35.00
Copyright �c 2001 by Academic Press
All rights of reproduction in any form reserved.

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
351
composition of high activity and N2 selectivity in the 100– 2.2. Catalysts Characterization
�
2
00 C range. Thiscatalytic system exhibitsalso a high stabil-
Surface area and metal dispersion measurements. The
specific surface area of the La0.5CexSr0.5� xMnO3 and � -
Al2O3 solids was checked by N2 adsorption at 77 K (BET
method) using a multipoint Fisons Sorpty 1900 System.
Metal dispersions of the catalysts were determined by H2
chemisorption at room temperature followed by TPD in
He flow.
ity in the presence of 5% H2O in the feed stream. These fea-
tures suggest that the present catalyst may be an important
candidate for low-temperature lean-deNOx applications.
The present work concerns catalytic and transient re-
activity studies on a 0.1 wt% Pt /La0.5Ce0.5MnO3 catalyst
for both the NO/H2/O2 and NO/H2 reactions. The same
studies were also conducted on a 0.1 wt% Pt /Al2O3 cata-
lyst for a critical comparison. Transient surface reactivity
XRD analysis. The crystal structure of the prepared
studies have been carried out on both metal-supported La0.5CexSr0.5� xMnO3 solids was checked by XRD measure-
catalysts and the corresponding supports in order to ob- ments using a SIEMENS Diffract 500 system employing
tain fundamental information about the nature and reac- CuK� radiation (� = 1.5418
tivity of intermediate species involved in the mechanism
A˚ ).
2
.3. Catalytic and Transient Studies
of the NO/H2/O2 reaction. The transient studies involved
(
(
a) temperature-programmed desorption (TPD) of NO,
b) temperature-programmed surface reaction (TPSR) in
The flow system used for conducting catalytic measure-
ments for the NO/H2/O2 (lean-deNOx) reaction at 1 atm
total pressure consisted of a flow measuring and control sys-
tem (MKS Instruments, Model 247C), mixing chambers, a
quartz fixed-bed microreactor (2 ml nominal volume), and a
GC–MS analysis system. The flow system, the microreactor,
and the analysis system used have been recently described
in detail(23). The feed stream consistingof0.25% NO, 1.0%
H2, 5% O2, and He as balance gas was used in all experi-
ments. The amount of supported Pt catalyst sample used
for the catalytic experiments was 0.15 g and the total flow
He flow, and (c) TPSR in H2/He flow following the
NO/H2/O2 reaction. The latter two kinds of transient ex-
periments allowed for the estimation of the surface cov-
erage of molecularly adsorbed NO and that of total N-
containing intermediate species formed under NO/H2/O2
reaction conditions. In addition, the surface reactivity of
these species toward He and H2/He treatment has been
studied. The transient experiments performed support the
view for a hydrogen-assisted dissociation step of NO and
a nitrogen-assisted reduction mechanism of NO to N2 and
N2O gas products.
3
rate was 100 cm (STP)/min, resulting in a GHSV of about
� 1
8
0,000 h . For the experiments to study the effects of H2O
in the feed stream on the NO conversion, a feed gas mix-
ture of about 5 mol% H2O/He was prepared by flowing He
through a deionized water saturator held at 35 C (� 1 C).
The exit stream was then mixed with NO/He, H2/He, and
O2/He gas mixtures at the appropriate mass flow rates in
2
. EXPERIMENTAL
�
�
2
.1. Catalysts Preparation
The La0.5CexSr0.5� xMnO3 solids were prepared using order to produce the desired feed gas composition. Reac-
the ceramic method as follows: calculated amounts of tion rates (kinetic or integral) were calculated using the
�
1
� 1
La(NO3)3 � 6H2O (Merck), Sr(NO3)2 (Ferak), Mn(NO3)3 � relationship Rate (mol � g � s ) = N y /W � N (mol/s)
T
i
T
9
H2O (Merck), and CeO2 (Aldrich) were mixed thoroughly is the total molar exit flow rate, y is the molar fraction of
i
�
� 6
in an agate mortar and heated slowlyto 400 C for nitrate de- component i (e.g., N ) expressed in ppm � 10 , and W is
2
composition. At the end of this step (no visible NO2 fumes), the weight of the catalyst (g). Reaction turnover frequen-
�
� 1
the system was further heated for 4 h at 900 C at 1 atm. The cies (TOF, s ) were calculated on the basis of the exposed
mixture was then removed from the furnace, cooled, and Pt atoms (4.6 �mol of Pt /g ).
s
cat
�
after grinding was heated again for another 4 h at 1050 C.
2
For the present catalytic reaction system, H reacts with
The sample was then slowly cooled to room temperature NO and O on the basis of the following competitive reac-
2
and stored for further use. It is mentioned that the notation tion scheme:
La0.5CexSr0.5� xMNO3 indicates just the nominal composi-
tion of the solids. This has nothing to do with particular
2
NO + 4H2 + O2 → N2 + 4H2O
[1]
[2]
crystal phases existing in each sample.
The metal-supported catalysts (0.1 wt% Pt /La0.5Ce0.5
MnO3 and 0.1 wt% Pt /Al2O3) were prepared by the incip-
ient wetness impregnation method using the H2Pt(IV)Cl6
2NO + 3H + O → N O + 3H O
2
2
2
2
1
H2 + O2 → H2O.
[3]
2
(
Aldrich) precursor for both catalysts. The � -Al2O3 support
was supplied by Aldrich (standard grade, 150 mesh). After
�
impregnation and drying (overnight at � 120 C), the cata- The only N-containing product species formed in the
�
lyst sample was calcined in air at 400 C for 2 h prior to use. present catalytic reaction system were N2 and N2O. The

3
52
COSTA ET AL.
TABLE 1
Description of Sequential Step Changes of Gas Flow during
of which measurements by online mass spectrometry were
recorded.
Temperature-Programmed Desorption (TPD) and Surface Reac-
tion (TPSR) Experiments Performed on La0.5Ce0.5MnO3, � -Al2O3,
3
. RESULTS
0
.1% Pt /La0.5Ce0.5MnO3, and 0.1% Pt /Al2O3 Catalysts
3
.1. Catalysts Characterization
BET surface area measurements. The specific surface
Sequence of step changes of gas
flow over the catalyst sample
Experiment
A
2
areas (BET, m /g) of the La0.5CexSr0.5� xMnO3 solids were
evaluated on the basis of results of the measured nitro-
gen isotherms at 77 K. The BET areas were found to be in
the 3–4 m /g range. In the case of � -Al2O3, the BET area
was found to be 155 m /g. Before any measurements were
0.5% NO/He (room temperature, 20 min)
→
He (5 min, room temperature) → TPD in He
�
B
C
0.25% NO/1.0% H2/5% O2 (140 C, 30 min)
2
→
cool quickly to room temperature in reaction mixture
2
�
→
He (5 min, 30 C) → TPSR in He
�
taken, the sample was outgassed at 250 C under vacuum
�
0.25% NO/1.0% H2/5% O2 (140 C, 30 min)
P � 1.3 � 10� mbar) overnight.
3
(
→
cool quickly to room temperature in reaction mixture
He (5 min, 30 C) → TPSR in 10% H2/He
�
→
X-ray diffraction studies. The crystal phase composi-
tion of the La0.5CexSr0.5� xMnO3 solids as determined by
XRD with reference to ASTM standards is shown in
Table 2. The main crystal phases detected are the LaMnO3
and SrMnO3� x of perovskite structure and the oxidic phases
of CeO2 and MnO2.
selectivity, �, for reduction of NO by H2 to N2, e.g., the
ratio of the consumption rate of H2 for reaction [1] to the
total consumption rate of H2 (reactions [1]–[3]), is calcu-
lated on the basis of the following:
Metal dispersion. The dispersion of platinum metal
for each of the 0.1 wt% Pt /Al2O3 and 0.1 wt% Pt /
La0.5Ce0.5MnO3 catalysts was found to be about 90%
(4.6 �mol of Pts/gcat) on the basis of H2 chemisorption (10%
0
.5yf XNO
NO
�
ꢀ% ) =
SN � 100.
[4]
f
2
y
XH2
H2
�
3
H2/He, T = 20 C, t = 30 min, q = 30 cm /min) followed by
TPD in He flow. This value gives an average particle size of
about 12 A˚ . It is noted that no H2 spillover effect was found
In Eq. [4], yf and yf are the feed molar fractions of NO
NO
H2
and H2, respectively, XNO and XH are the NO and H2 con-
2
versions, respectively, and SN is the selectivity of the reac-
2
under the hydrogen chemisorption conditions applied.
tion to N2 gas product. The value of 0.5 is the stoichiometric
ratio of NO to H2 appeared in reaction [1].
Temperature-programmed desorption (TPD) and sur-
face reaction (TPSR) experiments were conducted in a spe-
cially designed flow system that has been recently described
3
.2. Catalysts Performance and Stability with Time on
Stream for the NO/H2/O2 Reaction
3
.2.1. La0.5CexSr0.5� xMnO3 solids. The four solids of
the La0.5CexSr0.5� xMnO3 series prepared (Table 2) were
tested for the NO/H2/O2 reaction in the 200–500 C range.
(
23). For transient experiments, the amount of catalyst sam-
�
ple used was 0.15–0.2 g. The total flow rate was kept con-
3
The La0.5Ce0.5MnO3 solid exhibits the best catalytic per-
formance in terms of NO conversion and N2 selectivity as
shown in Table 3. In this table, Tmax is the temperature at
which a maximum in NO conversion wasobserved. N2 selec-
stant at 30 cm (STP)/min. Chemical analysis of the gas ef-
fluent stream of reactor during transients was done with an
online quadrupole mass spectrometer (Omnistar, Balzers)
equipped with a fast response inlet capillary/leak valve (SVI
�
tivity values are in the 83–92% range (at T = 350–400 C),
0
50, Balzers) and data acquisition systems. The gaseous
while the NO conversion varies between 20 and 24% under
the indicated reaction conditions. The La0.5Ce0.5MnO3 solid
responses obtained by mass spectrometry were calibrated
against standard mixtures. The mass numbers (m/z) 15, 28,
3
0, 32, 44, and 46 were used for NH3, N2, NO, O2, N2O, and
NO2, respectively.
Before any measurements were taken, the support ma-
terials were pretreated in 5% O2/He for 2 h at 500 C, while
the supported Pt catalysts were pretreated at 400 C for 2 h
TABLE 2
�
�
Prepared La0.5CexSr0.5� xMnO3 Solids and Crystal Phases
Detected by XRD
followed by treatment with a 10% H2/He gas mixture at
Solid composition
Crystal phases
�
3
5
00 C for 2 h. The feed was then switched to pure He at
00 or 300 C, respectively, for 15 min and the reactor was
�
La0.5Ce0.5MnO3
La0.5Sr0.5MnO3
La0.5Sr0.3Ce0.2MnO3
La0.5Sr0.2Ce0.3MnO3
LaMnO /MnO /CeO /La(OH)�
3
2
2
3
�
LaMnO3/SrMnO3� x/MnO2/La(OH)3
cooled to the appropriate temperature of the experiment
to be followed. Table 1 describes the necessary sequence
of steps performed for each kind of transient experiments
presented in this work. The underlined step is that during
�
LaMnO3/MnO2/CeO2/SrMnO3� x/La(OH)
3
�
�
LaMnO3/MnO2/CeO2/La(OH) /SrMnO
3
3� x
�
Traces

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
353
TABLE 3
Catalytic Activity of the La0.5CexSr0.5� xMnO3 Solids for the
a
NO/H2/O2 Reaction
�
Support
Tmax ( C)
XNO (% )
SN2 (% )
La0.5Sr0.5MnO3
400
400
350
400
20.8
23.7
19.9
24.3
82.9
91.6
89.2
92.2
La0.5Ce0.2Sr0.3MnO3
La0.5Sr0.2Ce0.3MnO3
La0.5Ce0.5MnO3
a
W = 0.2 g. Feed conditions: 0.25% NO, 1% H2, 5% O2, He as balance
gas; GHSV = 30,000 h�
1
.
was chosen as a support for the preparation of the 0.1 wt%
Pt /La0.5Ce0.5MnO3 catalyst on the basis of the results of
Table 3.
3
.2.2. Pt/La0.5Ce0.5MnO3 and Pt/Al2O3 catalysts. The
catalytic behavior of the 0.1 wt% Pt /La0.5Ce0.5MnO3 solid
as a function of reaction temperature for the reduction of
NO by H2 in the presence of 5% O2 is presented in Figs. 1
and 2. Catalytic results are given in terms of NO conversion
�
(
XNO, % ) and N2 selectivity (SN , % ) in the 100–400 C
2
�
range (Fig. 1a). The temperature of 400 C was the high-
est investigated in order to avoid sintering of Pt metal.
FIG. 2. Temperature profiles of (a) N2 and (b) N2O integral pro-
duction rates for the NO/H2/O2 “lean-NOx” reaction on the 0.1 wt%
Pt /La0.5Ce0.5MnO3 (᭹) and 0.1 wt% Pt /Al2O3 (᭡) catalysts. Reac-
tion conditions: H2 = 1.0% , NO = 0.25% , O2 = 5% , W = 0.15 g, GHSV =
�
1
8
0,000 h .
The XNO behavior of the catalyst can be distinguished in
�
three different temperature ranges. In the 100–160 C low-
temperature range the catalyst presents high catalytic ac-
tivity (XNO > 20% ) with a maximum conversion of 75%
�
�
at 140 C. Between 200 and 300 C the activity of the cata-
�
lyst is very small, while in the 300–400 C high-temperature
range the NO conversion is between 5 and 15% . The latter
behavior is mostly due to the activity of support itself as
illustrated by the catalytic results shown in Table 3. The N2
selectivity behavior of the Pt /La0.5Ce0.5MnO3 catalyst is as
follows. The catalyst is highly selective toward N2 formation
with selectivity values in the 75–87% range. The N2 selectiv-
�
ity remains practically unchanged in the 120–160 C range
(
SN = 75% ), while it increases to 80–87% at temperatures
2
�
�
higher than 160 C and lower than 120 C.
Figure 1b presents the temperature profile of H2 conver-
sion (XH , % ) for the corresponding experiments of Fig. 1a,
2
and the corresponding profile of the H2 selectivity of reac-
tion parameter (�H , % ) given by Eq. [4]. The H2 conver-
2
�
sion increases with reaction temperature, where at 250 C
FIG. 1. Temperature profiles of (a) the NO conversion, XNO (᭹),
and N2 selectivity, SN2 (᭡), and (b) the H2 conversion, XH2 (᭡), and H2
reaction selectivity, �H2 (᭹), of the NO/H2/O2 “lean-NOx” reaction on
the value obtained is found to be about 95% and increases
only slightly in the 250–400 C range. This temperature pro-
�
file of H2 conversion is totally different than that of NO
conversion as seen in Fig. 1a. As the temperature of the
the 0.1 wt% Pt /La0.5Ce0.5MnO3 catalyst. Reaction conditions: H2 = 1.0% ,
NO = 0.25% , O2 = 5% , W = 0.15 g, GHSV = 80,000 h�
1
.

3
54
COSTA ET AL.
reaction increases, the percentage of H2 used for the desired
reaction [1] decreases due to the H2 combustion reaction
[3]. It will be shown later that the rate of H2 combustion
is significantly higher than the rate of NO/H2/O2 reaction,
and this explains the temperature profile of H2 reaction se-
lectivity (Fig. 1b).
Figure 2 presents comparative results concerning the in-
tegral rates per gram of catalyst of N2 (Fig. 2a) and N2O
(
Fig. 2b) formation among the 0.1 wt% Pt /La0.5Ce0.5MnO3
�
and 0.1 wt% Pt /Al2O3 catalysts in the 100–400 C range.
As shown in Fig. 2a, the two catalysts present profiles
similar in quality in the 100–200 C range, where the
�
Pt /La0.5Ce0.5MnO3 exhibits significantly higher rates of N2
formation than the Pt /Al2O3 catalyst. There is practically
no activity of Pt /Al2O3 in the 250–400 C range, a behav-
FIG. 3. Effect of 5% H2O in the feed stream on NO conversion
᭹) and N2 selectivity (᭡) as a function of reaction temperature on the
(
�
0
.1 wt% Pt /La0.5Ce0.5MnO3 catalyst. Reaction conditions: NO = 0.25% ,
� 1
ior different than that observed on the Pt /La0.5Ce0.5MnO3 H = 1.0% , O = 5% , H O = 5% , W = 0.15g, GHSV = 80,000 h
.
2
2
2
catalyst. For the latter case, an increasing reaction rate is
�
�
observed in the 250–400 C range, where at 400 C a rate
�
mum NO conversion, while values in the 90–95% range are
similar to that obtained at 100 C was estimated. As clearly
�
obtained at T > 250 C.
shown in Fig. 2b, the Pt /Al2O3 catalyst presents significantly
�
Figure 4 presents comparative results of the integral
rate per gram of catalyst for N2 formation (Fig. 4a) and
N2O formation (Fig. 4b) among the Pt /La0.5Ce0.5MnO3 and
higher rates of N2O production in the 100–140 C range
than the Pt /La0.5Ce0.5MnO3 catalyst does. In particular, at
�
Tmax = 140 C the rate of N2O formation on the Pt /Al2O3 is
�
Pt /Al2O3 catalysts in the 100–400 C range in the case
3
times higher than that obtained on the Pt /La0.5Ce0.5MnO3
where H2O is present in the feed stream. As shown in
catalyst. Thus, it is apparent from the results of Fig. 2 that
the use of La0.5Ce0.5MnO3 solid instead of � -Al2O3 as sup-
port of Pt metal has significantly improved the N2 selectivity
�
of Pt metal in the 100–200 C low-temperature range.
It was interesting to study the rate of NO decomposition
�
in the 200–400 C range where the H2 conversion is very
high (Fig. 1b). After using a 0.25% NO/He mixture over the
Pt /La0.5Ce0.5MnO3 catalyst, it was found that a maximum
�
�
rate of 0.022 �mol of N2/g.s is obtained at 350 C. At 200 C
a rate of 0.004 �mol of N2/g.s was estimated. These results
clearly indicate that the decomposition reaction ofNO itself
contributes only to a small extent in the case where low
concentrations of hydrogen are present in the feed stream.
It is noted that the N2 selectivity of the NO/He reaction was
found to be in the 85–98% range at temperatures between
�
2
00 and 400 C.
The catalyticbehavior ofthe Pt /La0.5Ce0.5MnO3 solid was
also studied in the presence of 5 mol% water in the feed
stream containing 0.25 mol% NO, 1.0 mol% H2, 5 mol% O2,
and He as balance gas. The temperature profile obtained is
presented in Fig. 3. By comparing the results presented in
Figs. 1 and 3, it can be seen that the presence of water in
the reaction stream only slightly affected the activity of the
�
catalyst at temperatures lower than 250 C. On the other
hand, an enhanced catalytic activity at temperatures above
�
�
2
50 C is obtained. In particular, at 400 C the NO conver-
sion increases remarkably from 15 to 45% in the presence
of water. The N2 selectivity is also significantly affected by
the presence of water in the feed stream. A N2 selectivity
value of 87% is now obtained at the temperature of maxi- NO = 0.25% , O2 = 5% , H2O = 5% , W = 0.15 g, GHSV = 80,000 h
FIG. 4. Effect of 5% H2O in the feed stream on (a) integral reac-
tion rate of N2 and (b) N2O formation on the 0.1 wt% Pt /La0.5Ce0.5MnO3
(
᭹) and 0.1 wt% Pt /Al2O3 (᭡) catalysts. Reaction conditions: H2 = 1.0% ,
� 1
.

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
355
�
cantly in the 100–150 C range, while a smaller increase is
�
observed in the 200–400 C range over both catalysts. A
5% conversion of NO is achieved at 400 C in the case
�
9
of Pt /La0.5Ce0.5MnO3 catalyst. It is clearly seen in Fig. 6a
that the Pt /La0.5Ce0.5MnO3 exhibits significantly higher NO
conversion values than the Pt /Al2O3 catalyst in the 150–
�
4
00 C range. A similar behavior is obtained for the N2
selectivity of the NO/H2 reaction but at lower tempera-
�
tures (100–200 C, Fig. 6b). N2 selectivity values in the 55–
�
9
5% range are obtained for temperatures in the 100–400 C
�
� 3 � 1
range. At 100 C, a TOF value of 10.6 � 10
s
is obtained
FIG. 5. Stability test on the 0.1 wt% Pt /La0.5Ce0.5MnO3 (᭿) and
on both catalysts. It is noted that the temperature profile
0
0
.1 wt% Pt /Al2O3 (᭡) catalysts. Reaction conditions: H2 = 1.0% , NO =
�
of XNO for the NO/H2 reaction in the 100–400 C range is
.25% , O2 = 5% , H2O = 5.0% , W = 0.15 g, GHSV = 80,000 h�
1
.
completely different than that obtained for the NO/H2/O2
reaction (Fig. 1).
Fig. 4a, the Pt /La0.5Ce0.5MnO3 catalyst exhibits higher rates
of N2 formation than the Pt /Al2O3 catalyst over the whole 3.4. Kinetic Rates and Activation Energies of the
�
temperature range of 100–400 C. It is also important to
2 2 2 2 2
NO/H /O , NO/O , and H /O Reactions
�
note that at temperatures higher than 250 C the activ-
Kinetic aspects of the reduction of NO by H2 in the pres-
ence of O2 have been studied by measurements of the in-
trinsic rates of the NO/H2/O2, NO/O2, and H2/O2 reactions
ity of Pt /La0.5Ce0.5MnO3 increases with increasing reac-
tion temperature, while there is no activity in the case of
Pt /Al2O3 catalyst. In the case of the integral rate of N2O
formation, the results presented in Fig. 4b are similar to
those of Fig. 2b but the differences between the two cata-
lysts are more pronounced. Figure 4b clearly indicates that
the Pt /Al2O3 presents much higher rates of N2O produc-
�
at 140 C. In addition, the apparent activation energy of the
NO/H2/O2 reaction for both catalyst formulations has been
estimated. The rate results obtained are given in Table 4 in
�
tion than the Pt /La0.5Ce0.5MnO3 catalyst in the 100–200 C
range. It should be pointed out that the positive effect of
the La0.5Ce0.5MnO3 support on the activity and N2 selectiv-
ity of supported Pt catalyst is significantly improved by the
presence of water.
The stability of the catalyst performance with time on
stream was examined on both the Pt /La0.5Ce0.5MnO3 and
Pt /Al2O3 catalysts under a feed containing 0.25% NO/1.0%
H2/5% O2/5% H2O/He. Experiments were conducted for
2
0 h of continuous reaction. Figure 5 presents results of
these experiments in terms of N2 product yield (YN , % ). It
2
is observed that for both catalysts the N2 yield changes only
by 5 percentage units during the first 10 h of reaction. In the
case of Pt /Al2O3, the N2 yield remains the same during the
next 10 h on stream but in the case of Pt /La0.5Ce0.5MnO3 it
drops by 6 percentage units. For the latter catalyst, there is
about a 12% drop of its initial N2 yield after 20 h on stream.
It is noted that after 20 h on stream the Pt /La0.5Ce0.5MnO3
still exhibits a value of N2 yield higher by almost 20 per-
centage units than that of the Pt /Al2O3 catalyst (Fig. 5).
3
.3. Catalysts Performance for the NO/H2 Reaction
The catalytic activity of Pt /La0.5Ce0.5MnO3 and Pt /Al2O3
solids was also tested in the absence of oxygen using a feed
composition of 0.25% NO/1.0% H2/He. Figure 6 presents
FIG. 6. Temperature profiles of (a) the NO conversion and (b) the
N2 selectivity for the NO/H2 reaction on the 0.1 wt% Pt /La0.5Ce0.5MnO3
results of the NO conversion (Fig. 6a) and N2 selectivity
�
(
Fig. 6b) obtained on both catalysts in the 100–400 C range.
(
᭹) and 0.1 wt% Pt /Al2O3 (᭡) catalysts. Reaction conditions: H2 = 1.0% ,
� 1
As shown in Fig. 6a, the NO conversion increases signifi- NO = 0.25% , W = 0.15 g, GHSV = 80,000 h
.

3
56
COSTA ET AL.
3
TABLE 4
Q = 200 cm /min). In the case of the Pt /La0.5Ce0.5MnO3
catalyst, after 2 h on stream a steady state conversion of
hydrogen (XH2 = 16.5% ) was achieved. It is noted that no
hydrogen combustion activity was observed for the sup-
ports alone under the same reaction conditions.
Intrinsic Reaction Rates of NO/H2/O2, NO/O2, and H2/O2
�
Reactions on Supported Pt Catalysts at T = 140 C
Rate
(�mol/g.s)
TOF(s 1)
�
Catalyst
Reaction
Figure 7 shows Arrhenius plots for the estimation of the
apparent activation energy of the NO/H2/O2 reaction. Mea-
0
.1 wt% Pt /Al2O3
0.25% NO/1.0%
H2/5.0% O2
�
1.22a
surements were taken in the 90–130 C range with a feed
0.265
0
.1 wt% Pt /
consisting of 0.25% NO/1% H2/5% O2/He. Apparent ac-
tivation energy values of 21.5 and 24.8 kcal/mol were cal-
culated for the Pt /La0.5Ce0.5MnO3 and Pt /Al2O3 catalysts,
respectively. These results are in harmony with the higher
reaction rates observed on the former than the latter cata-
lyst formulation (see Fig. 2).
La0.5Ce0.5MnO3
.1 wt% Pt /Al2O3
.1 wt% Pt /
La0.5Ce0.5MnO3
.1 wt% Pt /Al2O3
.1 wt% Pt /
2.26a
0.490
0.0037
0
0
0.25% NO/5.0% O2
1.0% H2/5.0% O2
0.017a
0.043a
0.0093
13.6
b
0
0
62.7
La0.5Ce0.5MnO3
51.2b
11.1
a
3.5. Surface Reactivity Studies by Transient Methods
Micromoles of NO per gram per second.
Micromoles of H2 per gram per second.
b
3
.5.1. NO TPDs. The chemical interaction of NO with
the surface of La0.5Ce0.5MnO3 and Al2O3 supports as well as
of the 0.1 wt% Pt /La0.5Ce0.5MnO and 0.1 wt% Pt /Al O -
�
1
3
2
3
terms of micromoles per gram per second and TOF (s )
along with the feed composition used for each reaction. In
the case of NO/O2 reaction, only NO2 was observed with
a rate value 2 orders of magnitude smaller than that ob-
served for the desired NO/H2/O2 reaction. On the other
hand, in the case of H2 combustion (1% H2/5% O2/He) the
observed rate is now 50 times higher than that of the desired
NO/H2/O2 reaction in the case of Pt /Al2O3 and 22.5 times
higher in the case of Pt /La0.5Ce0.5MnO3 catalyst. The sig-
nificance of these results will be discussed later. It is worth
mentioning here that a spillover mechanism of H2 was ob-
served during H2/O2/He reaction on the basis of monitoring
H2 consumption. In particular, in the case of Pt /Al2O3 cata-
lyst the H2 conversion as a function of time on stream was
supported catalysts was studied by adsorption experiments
of NO at room temperature followed by temperature-
programmed desorption (TPD) in He flow. Before NO
chemisorption took place the La0.5Ce0.5MnO3 and Al2O3
3
samples were pretreated with air (30 cm (STP)/min) at
�
6
50 C for 2 h, while the supported Pt catalyst samples were
�
pretreated in air at 400 C for 2 h followed by H2 treatment
at 300 C for 2 h.
�
Figure 8 shows TPD response curves of NO and N2O
for all the studied materials. In the case of La0.5Ce0.5MnO3
solid (Fig. 8a), one distinct NO desorption peak is obtained
�
at TM = 365 C with a shoulder at the rising part of it. A
�
broad N2O peak is obtained at TM = 620 C, while a very
�
small peak at TM = 280 C is also observed. However, in
as follows: XH (% ) = 100, 100, 80, 55, 32, 20.5, and 20.2 after
2
the case of 0.1 wt% Pt /La0.5Ce0.5MnO3 catalyst (Fig. 8b)
two distinct NO desorption peaks with peak maxima at low
temperatures are observed. The first peak (the numbering
of desorption peaks follows the direction of increasing tem-
0
.1, 30, 60, 90, 120, 180, and 300 min, respectively. Thus, a
true steady state rate for H2 combustion was achieved after
h on stream (5 mg of Pt /Al2O3 diluted in 45 mg of Al2O3,
3
�
perature) iscentered at 65 C, while the second sharper peak
is centered at 120 C. A small N2O peak appears in the 130–
85 C range as opposed to the case of La0.5Ce0.5MnO3 sup-
�
�
1
port where a much larger N2O peak appears at much higher
temperatures. The amount of desorbed NO and N2O over
the two solids is substantially different. The quantities ob-
tained (micromoles per gram) are given in Table 5. In the
case of supported Pt catalyst an amount of desorbed NO
2
.1 times lower than that obtained on La0.5Ce0.5MnO3 sup-
port is noted.
In the case of Al2O3 (Fig. 8c), only NO is observed to
desorb. Two main desorption peaks are obtained centered
�
�
at 375 C and 525 C, while a very small peak is observed
�
at 75 C. In the case of 0.1 wt% Pt /Al2O3 catalyst (Fig. 8d),
FIG. 7. Arrhenius plots of the rate (TOF, s 1) of N2 formation of
the NO/H2/O2 lean de-NOx reaction on the 0.1 wt% Pt /La0.5Ce0.5MnO3
᭡) and 0.1 wt% Pt /Al2O3 (᭹) catalysts. Feed composition: H2 = 1.0% ,
NO = 0.25% , O2 = 5% , He as balance gas.
�
besides NO, nitrous oxide (N2O) is observed in the TPD
�
spectrum. A large NO peak is centered at about 240 C and
(
�
�
two smaller NO peaks are centered at 420 C and 470 C.

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
357
FIG. 8. Temperature-programmed desorption (TPD) profiles of NO, N2, and N2O in He flow according to Experiment A (Table 1) on
a) La0.5Ce0.5MnO3 (W = 0.2 g), (b) 0.1 wt% Pt /La0.5Ce0.5MnO3 (W = 0.15 g), (c) Al2O3 (W = 0.15 g), and (d) 0.1 wt% Pt /Al2O3 (W = 0.15 g);
(
�
�
3
= 30 C/min, QHe = 30 cm /min.
�
A small N2O peak is observed at 535 C and a very small case of Pt /La0.5Ce0.5MnO3, while only 10% of NO desorbs
�
�
one appears in the 120–230 C range. The NO desorption at T < 200 C in the case of Pt /Al2O3 catalyst. Further dis-
spectrum observed on Pt /Al2O3 shifted to lower tempera- cussion on the TPDs of Fig. 8 will be provided in a following
tures compared to that obtained with the Al2O3 support. section.
By comparing the TPD results obtained on both supported
3
.5.2. TPSR in He flow following NO/H2/O2 reaction.
Pt catalysts, two distinct differences must be noted. First,
the Pt /Al2O3 desorbs about 5 times more NO than the
Pt /La0.5Ce0.5MnO3 catalyst (see Table 5, Expt A). Second,
The measurement of NO chemisorption under NO/H2/O2
reaction conditions on the 0.1 wt% Pt /La0.5Ce0.5MnO3 and
.1 wt% Pt /Al2O3 catalysts was attempted as follows. After
reaction with the 0.25% NO/1.0% H2/5% O2/He gas mix-
ture at 140 C for 30 min (steady state), the reactor was
0
�
practically all the adsorbed NO desorbs at T < 200 C in the
�
quickly cooled to room temperature under the reaction
mixture. The feed was then changed to He for 5 min to
purge the gas phase from reactor and the lines. The tem-
TABLE 5
Amount of Species Desorbed (�mol/gcat) during Various Kinds
of Temperature-Programmed Desorption (TPD) and Surface Re-
action (TPSR) Experiments in Different Gas Atmospheres as a
Function of Catalyst Composition
perature of the catalyst was then increased at the rate of
�
3
0 C/min to carry out a TPSR experiment. Figures 9a–9d
present transient response curves of NO, N2O, N2, and NO2
obtained over La0.5Ce0.5MnO3, Pt /La0.5Ce0.5MnO3, Al2O3,
and Pt /Al O catalysts. In the case of La0.5Ce0.5MnO solid,
Total
N-containing
species
Expt
2
3
3
Catalyst sample (Table1) NO N2O N2 NO2 NH3
two NO desorption peaks of about the same peak maximum
�
intensity are obtained. The first peak is centered at 100 C,
0
.1% Pt /La0.5
Ce0.5MnO3
�
while the second one is centered at 265 C. In the case of
A
B
C
2.95 0.03
3.23 0.02 0.47
3.53 0.28
3.01(� = 0.65)
4.21(� = 0.92) Pt /La0.5Ce0.5MnO , transient response curves of NO, N O,
3
2
0.02 4.11(� = 0.89)
and N2 are observed. More precisely, a broad NO peak is
�
�
0
.1% Pt /Al2O3
A
B
C
14.25 1.6
13.12 0.81 12.43
18.83
17.45
39.6
37.7
observed in the 25–150 C range and a smaller one at 175 C.
A very small and broad N2O peak is observed in the 100–
1
�
�
90 C range, and a N2 peak at TM = 200 C.
La0.5Ce0.5MnO3
Al2O3
A
B
6.26 0.8
5.65
7.86
5.65
In the case of Al2O3 (Fig. 9c), a very large NO desorp-
�
�
tion peak centered at 250 C is observed along with a smaller
NO2 peak centered at 625 C. A totally different TPSR spec-
A
B
7.31
25.01
7.31
36.03
11.02
trum is observed in the case of Pt /Al2O3 (Fig. 9d). Two very

3
58
COSTA ET AL.
FIG. 9. Transient response curves of NO, N2, N2O, and NO2 obtained during TPSR in He flow following reaction in NO/H2/O2 on the
3
�
(
a) La0.5Ce0.5MnO3, (b) 0.1 wt% Pt /La0.5Ce0.5MnO3, (c) Al2O3, and (d) 0.1 wt% Pt /Al2O3 catalysts; QHe = 30 cm /min, � = 30 C/min.
�
�
large and intense peaks centered at 305 C and 470 C cor-
responding to N2 and NO, respectively, are observed. A
�
small N2O peak is also seen at 310 C. We note the fact that
the amount of NO desorbed from Al2O3 is about twice the
amount of N2 desorbed from Pt /Al2O3, while the amount of
NO2 desorbed from Al2O3 is similar to that of NO desorbed
from Pt /Al2O3. In addition, the amount of N2 produced
by the latter catalyst is found to be about 26 times larger
than the amount observed on Pt /La0.5Ce0.5MnO3 catalyst.
Table 5 (Expt B) reports the amounts of NO, N2O, NO2,
and N2 measured by the He TPSR experiments presented
in Figs. 9a–9d.
3
.5.3. TPSR in H2/He flow following NO/H2/O2 reaction.
Figure 10 describes TPSR response curves of NO, N2O, N2,
and NH3 obtained in H2/He flow (Expt C, Table 1) on the
Pt /La0.5Ce0.5MnO3 and Pt /Al2O3 catalysts following reac-
�
tion in NO/H2/O2 at 140 C and quick cooling of the sample
to room temperature in the reaction mixture. In the case of
Pt /La0.5Ce0.5MnO3 (Fig. 10a), response curves correspond-
ing to NO, N2O, and NH3 are obtained. A desorption peak
�
�
of NO centered at 40 C with a tail out to 225 C is observed.
�
A very small peak of NH3 centered at 50 C and a very broad
N2O peak in the 25–175 C range are also observed. How-
�
FIG. 10. Transient response curves of NO, NH3, N2, and N2O gaseous
species obtained during TPSR in 10% H2/He flow on the (a) 0.1 wt%
Pt /La0.5Ce0.5MnO3 and (b) 0.1 wt% Pt /Al2O3 catalysts according to the
ever, in the case of Pt /Al2O3 (Fig. 10b) only a N2 response
curve is obtained with four peaks in the 125–425 C range.
The amounts of gaseous species formed in the H2 TPSR
experiments (Fig. 10) are reported in Table 5 (Expt C).
�
3
sequence of steps described in Experiment C (Table 3); Q = 30 cm /min,
�
� = 30 C/min.

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
359
�
4
. DISCUSSION
the 100–200 C low-temperature range. According to the
results of Table 6, the present Pt /La0.5Ce0.5MnO3 catalyst
appears to be superior in all respects of its performance to
all the other catalysts listed. Given the fact that the reac-
4
.1. Catalysts Performance
The remarkable catalytic performance of the present tion orders with respect to all three reactants may not be
0
.1 wt% Pt /La0.5Ce0.5MnO3 catalyst toward NO/H2/O2 expected to have values greater than unity, from the results
�
(
lean-deNOx) reaction in the 100–200 C low-temperature of Table 6 it is clear that the Pt /La0.5Ce0.5MnO3 exhibits
range and in the presence of 5% H2O in the feed stream the highest activity and selectivity ever reported for the
is best illustrated in Table 6 by comparison with other Pt- NO/H2/O2 reaction.
based catalysts reported in the open literature. In Table 6,
On the basis of the fact that the present supported Pt
Tmax is the reaction temperature at which maximum NO catalysts exhibit similar dispersions (� 90% ), a purely sup-
conversion is observed, while 1T1 and 1T2 are the tem- port effect must be invoked to explain the differences
perature ranges in which XNO appears to be greater than in the catalytic performances of these catalyst formula-
1
/2 or 1/10 of the maximum conversion of NO observed, tions. It is suggested that one of the main reasons that the
respectively. The latter parameters could be used to de- Pt /La0.5Ce0.5MnO3 catalyst exhibits remarkable N2 selec-
fine the quality of the operating temperature window. For tivity values compared to Pt /Al2O3 may be the ability of
example, the larger the value of 1T1 the better the desired La0.5Ce0.5MnO3 support to act as an oxygen reservoir. The
operation of the catalyst under practical conditions. Table 6 latter material consists of CeO2, MnO2, and LaMnO3 (of
reports also the integral rate of N2 production per gram of perovskite structure) phases within the pores in which Pt
Pt metal (RN ) evaluated according to the XNO and SN val- was highly dispersed. These phases contain inherently oxy-
2
2
ues reported. In addition, results of the turnover frequency gen vacancies according to O2 chemisorption experiments
�
(
based on Pt dispersion values reported) of the N2 produc- performed. More precisely, following O2 adsorption at 30 C
�
1
�
tion, TOFN (s ) at 130 or 140 C (XNO is less than 20% at for 0.5 h a TPD run was performed. A large and broad
the feed composition reported) are given in Table 6. First, O2 desorption peak starting at 600 C was obtained that
2
�
the comparison among the two catalysts presently investi- amounted to 32.5 �atom of O/gcat. This value is about 7
gated is rather straightforward. There is about a 40% in- times larger than the amount of exposed Pt atoms in the
crease in the maximum integral rate when Pt is supported present supported catalysts. A recent detailed kinetic study
on La0.5Ce0.5MnO3 than � -Al2O3 solid, while in terms of of the NO/H2/O2 reaction on Pt–Mo–Co/�-Al2O3 (17) re-
TOF the increase is even larger (e.g., 85% ). In addition, the vealed that for O2 feed concentrationsbelow2% the ratesof
N2 selectivity value of 88% obtained on Pt /La0.5Ce0.5MnO3, NO conversion and N2 selectivity both increase significantly
to our knowledge, is the highest value ever reported at the with increasing O2 feed concentration, while co-adsorbed
temperature at which maximum activity is observed for the hydrogen was found to displace molecularly adsorbed NO.
NO/H2/O2 reaction with 5% H2O in the feed stream in The authors have suggested that the oxygen promoting
TABLE 6
�
Catalytic Activity of Various Supported Pt Catalysts for the NO/H2/O2 Reaction in the Low-Temperature Range of 90–200 C
Reaction conditions
2
RN2
1T1
1T2
XNO
(% ) (% )
SN2 TOFN � 10
2
�
�
�
a
�
b
� 1
Catalyst
% Pt /SiO2
% Pt /Al2O3
% Pt /TiO2
% Pt /Al2O3
% Pt–Mo–Co/Al2O3
.1% Pt /Al2O3
.1% Pt /
NO (% ) H2 (% ) O2 (% ) GHSV (h 1) Tmax ( C) (�mol/s.gm) ( C) ( C)
(s
)
Ref.
1
1
1
1
1
0
0
0.05
0.05
0.1
0.1
0.3
0.2
0.2
0.3
0.3
0.8
1.0
1.0
6
6
5.0c
5.0c
8.0
240,000
240,000
40,000
40,000
6500
90
140
100
100
150
125
140
7.6
10.1
2.2
25
40
50
65
30
45
60
60
90
100
120
70
75
50
50
62.2
55
30
60
21
11
50
60
79
1.63f
(18)
(18)
(19)
(19)
(17)
g
0.87
0.17
0.22
0.31
f
f
1.4
g
12.1
285.8
396.9
d
f
0.25
0.25
5.0
5.0
80,000
80,000
100
300
66
88
26.5
This work
This work
d
e
f
49.0
La0.5Ce0.5MnO3
a
1
T1: temperature range where XNO > XNOmax /2.
T2: temperature range where XNO > XNOmax /10.
b
c
d
e
f
1
1
0% H2O is also present in the feed stream.
1
0% H2O is also present in the feed stream.
�
Except at T = 250 C.
�
g
�
T = 140 C. T = 130 C.

3
60
COSTA ET AL.
ring on Pt /Al2O3, if the surface coverage of NO does not
much influence this rate. In situ FTIR studies provide the
means to probe in a direct manner for the presence of such
an adsorbed NO structure. This kind of investigation will
be conducted and results will soon be reported.
It has been shown (24) that after passing NO/O2 gas mix-
ture over the MnOy–ZrO2 system the MnOy phase acts
as a reservoir of lattice oxygen (due to its oxygen vacant
sites) that can oxidize NO to NO2. The latter species is
then transferred and stored on the ZrO2 surface. It has also
been suggested that on the Pt /Al2O3 catalyst reduction of
NO by C3H6 to produce N2 gas in the presence of 5% O2
proceeds via an NO2-adsorbed intermediate species that is
more active than the adsorbed NO species (30). The re-
SCHEME 1. Interaction of gaseous NO, O2, and H2 with the 0.1 wt% sults in Table 4, aiming to shed some light on the role of
Pt /La0.5Ce0.5MnO3 catalyst.
NO2 on the activity behavior of the present supported Pt
catalysts, need to be interpreted with care. If one assumes
effect is attributed to the lowering of abundant adsorbed that the NO2 is an important intermediate species in the
hydrogen atoms by reaction with oxygen to form H2O, per- sequence of steps for the formation of N2 gas during the
mitting therefore the pairing of two adjacent adsorbed N NO/H2/O2/He reaction, the measured steady state produc-
atomsto form N2 gas. Thisexplanation might also be offered tion rates of NO2 during the NO/O2/He reaction cannot di-
to explain the N2 selectivity behavior of the present sup- rectly be associated with the former rate. The low value of
ported Pt catalysts. In fact, higher N2 selectivity values are the steady state rate of NO2 formation reported in Table 4
�
obtained in the 100–200 C range in the presence than the does not imply the absence of a reservoir of adsorbed NO2
absence of O2 in the feed stream (see Figs. 2a and 6a). It on the metal and/or support that might be more active
is suggested that the rate of removal of adsorbed hydro- than the molecularly adsorbed NO under lean-NOx con-
gen atoms on the small Pt particles might be enhanced by ditions. The He TPSR experiments reported in Fig. 9 and
the participation of oxygen atoms present at the interface Table 5 suggest the formation of a large amount of NO2 on
between the Pt particles and support. Such oxygen atoms the alumina surface but the Pt /Al2O3 appears less active
could be more active than oxygen atoms on the Pt surface and selective than the Pt /La0.5Ce0.5MnO3 catalyst.
toward reaction with adsorbed hydrogen. Scheme 1 shows
Two maxima in the temperature profile of NO conversion
�
the formation of adsorbed oxygen species on the Pt metal (Tmax = 100 and 300 C) for the NO/H2/O2 reaction have
and La0.5Ce0.5MnO3 support. For the latter, anionic vacant been reported on Pd/TiO2 catalyst (19). The authors have
sites present in LaMnO3, CeO2, and MnO2 but not in � - suggested that the cause of the appearance of the two max-
Al2O3 support phases are offered for O2 chemisorption (16, ima is the change of the reaction paths between the direct
2
4, 25).
reduction ofNO byH2 and the reduction ofin situ generated
The adjustment of reaction N2 selectivity by the catalyst NO2 by H2. According to the catalytic results of the present
is an aspect the understanding of which requires precise Pt /La0.5Ce0.5MnO3 and La0.5Ce0.5MnO3 solids, the appear-
information about the catalytic mechanism and intrinsic ki- ance of two maxima in the activity vs temperature profile in
�
netics of the reaction. This matter was out of the scope of the 100–500 C range is due to the bifunctional operation of
�
the present work. However, some remarks based on the the catalyst. The Pt metal operates in the 100–200 C range,
present results could be made. On supported Rh catalysts while the La0.5Ce0.5MnO3 support in the 250–500 C range
the formation of Rh–NO species is well documented (26– where the operation of the latter is influenced by the Pt
�
�
2
8). Dissociation of the latter species is expected to occur metal.
easier than of other chemisorbed NO species (e.g., Rh–NO
A maximum in the activity vs temperature profile ob-
+
�
(
neutral) and Rh–NO ). The presence of F-type defects served at 140 C (Figs. 1a and 2a) is rather clear that is due
in the La0.5Ce0.5MnO3 support as previously discussed and to the significant rate ofH2 combustion to H2O on the Pt sur-
�
the fact that NO can be adsorbed onto these sites (29) make face in the 100–250 C range according to the kinetic rates
possible the likelihood of formation of negatively charged reported in Table 4. Other researchers (17, 19) have also
adsorbed NO species at the metal–support interface as il- suggested this explanation. In particular, Ueda et al. (19)
lustrated in Scheme 1. In this chemisorbed state the N atom have measured on the 1 wt% Pt /Al2O3 catalyst the rate of
�
is bonded to the Pt metal and the O atom is bonded to the H2 combustion at 300 C. The latter rate was 30 times larger
oxygen vacancy. This possibility is not offered in the case than that obtained for the NO/H2 reaction.
of Pt /Al2O3. Therefore, the rate of dissociation of NO on
Hodjati et al. (31) and Eguchi et al. (32) have reported
Pt /La0.5Ce0.5MnO3 is expected to be larger than that occur- results on the beneficial effect of H2O on the adsorptive

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
361
capacity of some perovskites ABO3 (A = Ca, Sr, Ba and a result quite different than that obtained on the present
B = Sn, Zr, Ti) and Mn–Zr mixed oxides, respectively, to- Pt /La0.5Ce0.5MnO3 catalyst. In addition, the TM value of
ward NO and NO2 under lean-burn conditions. The au- the second NO peak was found to be significantly higher
thors have suggested that adsorbed water stabilizes ad- than that observed on the present Pt /La0.5Ce0.5MnO3 sys-
�
sorbed NO3 species through hydration, the latter species tem, and about 40% of adsorbed NO participated in the
formed via NO and NO2 adsorption. A similar explanation formation of N2 and N2O (34). This result is opposite to
might be offered for the significant increase of NO con- that observed here and it shows again the great effect of
�
version in the 300–400 C range (Fig. 3) that takes place support on the chemistry of NO adsorption and dissocia-
mainly on the La0.5Ce0.5MnO3 support. On the other hand, tion over supported Pt catalysts.
the less pronounced beneficial effect of water observed in
Pt/Al2O3 catalyst. It is found that the Pt /Al2O3 cata-
�
the 100–200 C range (Figs. 3 and 4) could be explained on
lyst adsorbs a much higher amount of NO than that cor-
the basis of the role of adsorbed hydrogen (produced by
responding to the Pt surface and alumina alone (Table 5).
H2O dissociation on the Pt metal) to regulate the surface
This difference cannot be justified even in the case of for-
atomic composition of Hs and Os. The rate of water disso-
mation of dinitrosyl (Pt–(NO)2) species only. A reasonable
ciation under the reaction conditions may be small enough
mechanism by which the above-mentioned results can be
to significantly affect the NO conversion and N2 reaction
selectivity.
explained is that of NO spillover from the Pt metal to the
Al2O3. By comparing the NO response curves in Figs. 8c
and 8d, it can be stated that the large NO desorption peak
centered at 240 C represents a new adsorbed state of NO
4
.2. Transient Studies
�
4
.2.1. NO TPDs. Bycomparingthe TPD spectra shown formed by the spillover mechanism.
in Figs. 8a–8d and the quantities of each species desorbed
The formation of N O on the Pt /Al O catalyst in the
2
2
3
�
(
Expt A, Table 5) from the supported Pt catalysts and their 425–625 C range is derived via the dissociation of NO onto
supports alone, the following are noted.
2 3
the Pt metalafter desorption ofNO from the Al O support.
The appearance of the same peak maximum temperature
Pt/La0.5Ce0.5MnO3 catalyst. In the case of Pt /La0.5
Ce0.5MnO3 and its support alone, it is found that the lat-
ter adsorbs higher amounts of NO than the supported
Pt catalyst. Furthermore, the desorption spectra (shape
and position) are quite different among the two sam-
ples. It is suggested that the small Pt clusters deposited
on the La0.5Ce0.5MnO3 support have considerably altered
the adsorptive characteristics of certain type(s) of its
NO chemisorption sites (probably via short-range elec-
tronic interactions), while at the same time they may have
blocked adsorption sites to a given extent. It is known that
chemisorption ofNO on metaloxidesproceedswith the par-
ticipation of metal cations, oxygen anions, and oxygen va-
cant sites forming various kinds of nitrate and nitrite struc-
tures (33). From Figs. 8a and 8b it can be stated that break-
ing of the N–O bond proceeds with much more difficulty on
the La0.5Ce0.5MnO3 support alone than on the supported Pt
metal in order to form N2O gas. The following widely ac-
cepted elementary steps describe the formation of N2O on
the Pt surface.
�
(
TM = 525 C) in the NO response of Al2O3 (Fig. 8c) and
that of N2O of Pt /Al2O3 (Fig. 8d) strongly supports this
view. Even though a spillover mechanism of NO from sup-
ported Pt catalysts has not been reported previously to our
knowledge, this mechanism was reported to operate in the
case of the SCR reaction of NO on the V2O5/TiO2 catalyst
(
35). An alternative explanation for the NO TPD resultsob-
served for the Pt /Al2O3 catalyst could be based on the fact
that additional sites for NO chemisorption, besides those
on Pt and alumina surfaces alone, may have been formed
at the periphery of metal–support interaction. However,
considering the fact that after subtracting the sum of the
amount of NO chemisorption corresponding to the alumina
and Pt surface (4.6 �mol/g) alone from the total amount of
NO chemisorption, the remaining amount of NO exceeds
that which would correspond to adsorption on the Pt sur-
face. This result rather supports the view that the enhanced
chemisorption of NO on the Pt /Al2O3 catalyst is at least
partly due to a spillover mechanism. The N2O produced in
�
the Pt /Al2O3 catalyst in the 425–625 C range (Fig. 8d) is
NOꢀg) + s ↔ NO-s
NO-s + s → N-s + O-s
NO-s + N-s → N2Oꢀg) + 2s.
[5] due to reaction of readsorbed NO (after being desorbed
from the alumina) with N atoms on Pt according to the
[6]
[7]
elementary reaction steps [5]–[7].
4
.2.2. TPSR experiments in He and H2 flow following
NO/H2/O2 reaction. The experiments presented in Figs. 9
and 10 reveal important kinetic and mechanistic informa-
tion for the present catalytic reaction when compared to
the ordinary NO TPDs presented in Fig. 8.
Two clear adsorbed states of NO are formed on the Pt clus-
ters as indicated in Fig. 8b. A NO TPD spectrum very sim-
ilar in shape to that of Fig. 8b has been reported by Burch
et al. (34) for the 5 wt% Pt /SiO2 catalyst after adsorption
of NO at room temperature. However, on the latter TPD
Pt/La0.5Ce0.5MnO3 catalyst. First, the total amount of
study two N2 and N2O desorption peaks were observed, equivalent N atoms stored on the catalyst surface under

3
62
COSTA ET AL.
SCHEME 2. (a) H-assisted NO desorption and dissociation mechanism and (b) N-assisted reduction mechanism of NO to N2 and N2O on the 0.1
wt% Pt /La0.5Ce0.5MnO3 catalyst.
reaction conditions and measured under He (Fig. 9b) and reported (37) that the activation energy of the N–O bond
H2 flow (Fig. 10a) in the TPSR experiments are in excellent dissociation step [6] is higher than that of the H-assisted
agreement (see Table 5, last column). This result confirms N–O bond dissociation step [8] given below.
the reproducibility and accuracy of the experiments per-
NO-s + H-s → N-s + OH-s.
[8]
formed. Second, the differences in the spectra shown in
Figs. 9b and 10a concern the composition and amounts of
gases evolved and therefore reflect the different kinetics
of reactions involved under He or H2 flow. Third, the spec-
trum of desorbed NO in the H2 TPSR experiment (Fig. 10a)
shifts significantly to lower temperatures compared to the
case of He TPD (Fig. 8b) and TPSR (Fig. 9b) experiment.
Based on the above-mentioned remarks and the fact that
two distinct NO adsorbed states (types I and II) are very
likely to have been formed on this catalyst (see Figs. 8b and
In Scheme 2b a nitrogen-assisted NO reduction to N2
and N2O is illustrated according to the literature (8, 37).
It has been shown that formation of N2 and N2O via these
mechanismsproceedswith a lower activation energybarrier
than the recombination reaction step [9]ofadsorbed atomic
N to form N2 gas.
N-s + N-s → N2ꢀg) + 2s.
[9]
9
b), the following discussion explains the results of Figs. 9b In Scheme 2b both kinds of adsorbed NO species shown
and 10a in a satisfactory manner.
in Scheme 1 could participate in this mechanism. In the
Scheme 2a illustrates that under NO/H2/O2 reaction con- ordinary NO TPD (Fig. 8b) it is clearly shown that very lit-
ditions adsorbed atomic N, O, and H species can be pop- tle NO dissociation occurs (12% of adsorbed NO, Table 5)
ulated on the catalyst surface in addition to molecularly that favors the formation of N2O. On the other hand,
adsorbed NO. The surface coverage of the adsorbed NO in the case of the He TPSR experiment some coverage
species and its binding energy will depend on the oxida- of adsorbed atomic N is suggested to be present at the
tion state of Pt. An assisted hydrogen effect on the lower- start of the experiment. This species had been formed dur-
ing of bond strength of Pt–NO might be expected as it has ing NO/H2/O2 reaction. According to Scheme 2b, the rate
been reported for supported Rh catalysts (36). In the H2 of nitrogen-assisted NO dissociation becomes significant
TPSR experiment, the coverage of adsorbed atomic hydro- even at low temperatures and competes favorably with
gen species is expected to be much higher than that found the rate of NO desorption. Given the fact that desorp-
under NO/H2/O2 reaction conditions, while under NO/He tion of NO can now occur at lower temperatures com-
adsorption conditionsthisspeciescannot be present. There- pared to the case of NO TPD (Fig. 8b), the possibility
fore, the shift of desorbed NO response to lower tempera- of N2O formation now decreases (k1 > k4 in Scheme 2),
tures in the He TPSR (Fig. 9b) and H2 TPSR (Fig. 10a) ex- while N2 can be formed at higher temperatures (k3 < k4 in
periments with respect to the ordinary NO TPD response Scheme 2).
(
Fig. 8b) can be justified. In Scheme 2a a hydrogen-assisted
In the case of H2 TPSR experiment (Fig. 10a) it is seen
NO dissociation mechanism, first suggested by Hecker and that the formation of N2O dominates that of N2, while N2O
Bell (8) on the Rh/SiO2 catalyst, is also proposed. It was is formed in the same temperature range as desorption of

NO/H2/O2 (LEAN-deNOx) REACTION ON SUPPORTED Pt CATALYSTS
363
NO occurs. This behavior can be understood on the basis of
5. CONCLUSIONS
Scheme 2 where the presence of H2 gas (10% H2/He) pro-
motes significantly the insertion mechanism of N atom to
the Pt–NO bond to form N2O gas. Thus, when adsorbed
N species are present at the start of the H2 TPSR ex-
periment, the rate of N2O formation competes favorably
with that of NO desorption, while it is higher than that
of N2 formation (k3 < k4 according to the mechanisms of
Scheme 2).
According to the He TPSR results reported in Table 5,
the NO TPDs of Fig. 8b, and Scheme 2, an upper value
for the coverage of adsorbed NO present in the NO/H2/O2
reaction can be suggested for the Pt /La0.5Ce0.5MnO3 cata-
lyst. This value is �NO < 0.75. Much smaller amounts of NHx
The following conclusions can be derived from the results
of the present work:
(
1) A 0.1 wt% Pt supported on La0.5Ce0.5MnO3 (a mixed
oxide containing LaMnO3, CeO2 and MnO2 phases) cata-
lyst has proven to give the highest integral rate of NO
conversion to N2 (micromole per second per gram of Pt)
�
during NO/H2/O2 lean-NOx reaction in the 100–200 C low-
temperature range ever reported in the open literature.
A similar result has been obtained for the site reactivity,
� 1
TOFN (s ) and N2 selectivity parameters.
2
(
2) The 0.1 wt% Pt /La0.5Ce0.5MnO3 catalyst has shown
also very high N2 yields (>80% ) in the presence of 5% H2O
in the feed stream and for 20 h on stream. This performance
is by far better than that exhibited by the 0.1 wt% Pt /Al2O3
catalyst.
(
x = 0–2) seem to be formed, while the coverage of O and
H atoms on Pt seems to be less than about 0.1. These results
strongly suggest that dissociation of NO must be considered
as a slow step.
(
3) A hydrogen-assisted NO dissociation step and a
nitrogen-assisted mechanism for N2 and N2O formation
Pt/Al2O3 catalyst. In the H2 TPSR experiment are proposed to explain the various NO TPD and TPSR
(
Fig. 10b), only N2 gas is formed, a result largely differ- experiments in He and H2/He flow following reaction in a
ent than that obtained on the Pt /La0.5Ce0.5MnO3 catalyst satisfactory manner.
Fig. 10a), while the latter catalyst is found to exhibit sig- (4) The remarkable catalytic behavior of 0.1 wt%
(
nificantly higher N2 selectivity values during the NO/H2/O2 Pt /La0.5Ce0.5MnO3 toward NO/H2/O2 reaction is suggested
�
reaction in the 100–200 C range. These two experimental to be due to various direct and indirect influences of sup-
results do not contradict each other. It should be noted port composition. In particular, oxygen vacant sites of sup-
that the oxidation state of surface Pt atoms is an important port located next to the small Pt clusters could provide
surface parameter that will determine the bond strength the means for the formation of adsorbed NO with the N
among the various adsorbed species and the Pt atoms. This atom located on the Pt metal and the O atom on the oxy-
parameter is expected to be affected by the support compo- gen vacancy. Such an adsorbed state is expected to give
sition in the case of highly dispersed supported metal cata- rise to a higher rate of N2 formation compared to the
lysts, as the ones studied in the present work, in addition to state of molecularly adsorbed NO on the Pt metal alone.
the composition and concentration of adsorbed reaction in- In addition, adsorbed oxygen formed on the oxygen va-
termediate species. In fact, according to the He TPD and He cancies of support located at the metal–support interface
�
TPSR experiments in the 25–225 C range on the Pt /Al2O3 could be considered to be more active in removing ad-
catalyst, the Pt surface adsorbs very little NO, if any, as op- sorbed hydrogen from the Pt surface. This result could reg-
posed to the case of Pt /La0.5Ce0.5MnO3 catalyst. Therefore, ulate in the appropriate manner the surface coverage of
under the present reaction conditions the proposed mecha- adsorbed hydrogen toward activity and N2 selectivity maxi-
nism for N2 and N2O formation is associated with different mization.
kinetics on the present two supported Pt catalysts, each in-
(5) Deposition of 0.1 wt% Pt onto the Al2O3 surface gives
fluenced by the oxidation state of Pt. It seems reasonable to rise to a significant enhancement of NO adsorption on the
suggest according to the H2 TPSR results that in the case of Al2O3 support at room temperature. This is very likely to
Pt /Al2O3 it appears that Pt can be more readily reduced by be the result of a spillover phenomenon of NO from Pt to
H2 treatment than in the case of Pt /La0.5Ce0.5MnO3 cata- the support. The creation of new adsorption sites for NO
lyst. This then allows the NO reduction by H2 to completely on the metal–support interface can only partly account for
proceed toward N2 and H2O formation. The various N2 this behavior.
peaks observed in the H2 TPSR experiment on Pt /Al2O3
in the 150–425 C range are due to reduction of NO that is Pt /La0.5Ce0.5MnO3 catalyst are observed to desorb in the
spilled-over from the alumina support to the Pt surface. In 25–225 C range in He flow. Only the more strongly bound
(6) Two distinct adsorbed states of NO on the
�
�
the case of He TPSR experiment, part of this NO desorbs adsorbed state of NO was found to react with H2 in the
�
at very high temperatures due to decomposition reactions 25–225 C range.
of the various nitrate and nitrito structures formed on the
(7) The surface coverage of adsorbed NHx (x = 0–2)
�
alumina support upon NO/H2/O2 treatment of the catalyst species formed during NO/H2/O2 reaction at 140 C was
�
at 140 C.
found to be very small compared to the surface coverage

3
64
COSTA ET AL.
of molecularly adsorbed NO on the Pt /La0.5Ce0.5MnO3 15. Lindsteld, A., Str o¨ mberg, D., and Milh, M. A., Appl. Catal. 116, 109
1994).
(
catalyst.
1
6. Ferri, D., Forni, L., Dekkers, M. A. P., and Nieuwenhuys, B. E., Appl.
Catal. B: Environ. 16, 339 (1998).
ACKNOWLEDGMENTS
17. Frank, B., Emig, G., and Renken, A., Appl. Catal. B: Environ. 19, 45
(
1998).
Financial support by the Research Committee of the University
of Cyprus is gratefully acknowledged. The experimental assistance of
Chrystalla Charalambous (final year project student of this University)
is also acknowledged.
1
1
8. Burch, R., and Coleman, M. D., Appl. Catal. B:Environ. 23, 115(1999).
9. Ueda, A., Nakao, T., Azuma, M., and Kobayashi, T., Catal. Today 45,
1
35 (1998).
0. Yokota, K., Fukui, M., and Tanaka, T., Appl. Surf. Sci. 121/122, 273
1997).
21. Burch, R., Millington, P. J., and Walker, A. P., Appl. Catal. B: Environ.
, 160 (1994).
2
(
REFERENCES
4
1
2
. Fritz, A., and Pitchon, V., Appl. Catal. B: Environ. 13, 1 (1997).
. P aˆ rvulescu, V. I., Grange, P., and Delmon, B., Catal. Today 46, 233
22. Fukui, M., Yokota, K., and Shokubai, K., Catalysts Catal. 36, 160
(1994).
(
1998).
. Busca, G., Lietti, L., Ramis, G., and Berti, F., Appl. Catal. B: Environ.
8, 1 (1998).
23. Costa, C. N., Anastasiadou, T., and Efstathiou, A. M., J. Catal. 194,
250 (2000).
24. Eguchi, K., Kondo, T., Hayashi, T., and Arai, H., Appl. Catal. B: Env-
iron. 16, 69 (1998).
3
4
1
. Janssen, F. J., in “Handbook of Heterogeneous Catalysis” (G. Ertl,
H. Kn o¨ zinger, and J. Weitkamp, Eds.), p. 1633. VCH, Weinheim,
25. Tejuca, L. G., Fierro, J. L. G., and Tascon, J. M. D., Adv. Catal. 36, 237
(1989).
1
997.
5
6
. Rausenberger, B., Swiech, W., Schmid, A. K., Rastomjee, C. S., Emgel,
W., and Bradshaw, A. M., J. Chem. Soc. Faraday Trans. 94(7), 963
26. Ari, H., and Tominga, H., J. Catal. 43, 131 (1976).
27. Solymosi, F., Bansagi, T., and Novak, E., J. Catal. 112, 183 (1988).
28. Srinivas, G., Chuang, S. S. C., and Debnath, S., J. Catal. 148, 748 (1994).
29. Busca, G., and Lorenzelli, V., J. Catal. 72, 303 (1981).
30. Burch, R., and Sullivan, J. A., J. Catal. 182, 489 (1999).
31. Hodjati, S., Vaezzadeh, K., Petit, C., Pitchon, V., and Kiennemann, A.,
Appl. Catal. B: Environ. 26, 5 (2000).
(
1998).
. Dumpelmann, R., Cant, N. W., and Trimm, D. L., in “3rd ICC and
Automotive Polution Control” (A. Frennet and J.-M. Bastin, Eds.).
Vol. 2, p. 13. ULB Brussels, 1994.
. Tomishige, K., Asakura, K., and Iwasawa, Y., J. Catal. 157, 472 (1995).
. Hecker, W. C., and Bell, A. T., J. Catal. 92, 247 (1985).
. Hornung, A., Muhler, M., and Ertl, G., Catal. Lett. 53, 77 (1998).
7
8
9
32. Eguchi, K., Watabe, M., Ogata, S., and Arai, H., J. Catal. 158, 420
(1996).
1
1
0. Kobylinski, T. P., and Taylor, B. W., J. Catal. 33, 376 (1974).
1. Huang, S. J., Walters, A. B., and Vannice, M. A., J. Catal. 173, 229
33. Klingenberg, B., and Vannice, M. A., Appl. Catal. B: Environ. 21, 19
(1999), and references therein.
(
1998).
34. Burch, R., Shestov, A. A., and Sullivan, J. A., J. Catal. 186, 353 (1999).
35. Odriozola, J. A., Heinemann, H., Somorjai, G. A., Garcia de la Banda,
J. F., and Pereira, P., J. Catal. 119, 71 (1989).
1
1
2. Burch, R., and Scire, S., Catal. Lett. 27, 177 (1994).
3. Salama, T. M., Ohnishi, R., Shido, T., and Ichikawa, M., J. Catal. 162,
1
69 (1996).
4. Tanaka, T., Yokota, K., Doi, H., and Sugiura, M., Chem. Lett. 273
1997).
36. Solymosi, F., and Kn o¨ zinger, H., J. Chem. Soc. Faraday Trans. 86(2),
389 (1990).
37. Shustorovich, E., and Bell, A. T., Surf. Sci. 289, 127 (1993).
1
(