Performance of Pd-Mo/γ-Al2O3 catalysts for the selective reduction of NO by methane

Article abstract of DOI:10.1016/S1381-1169(03)00214-0

Pd-Mo/γ-Al₂O₃ catalysts were prepared by wet impregnation using organometallic precursors. The performance of these catalysts in SCR of NO by methane (NO-CH₄) was compared with corresponding Pd- and Mo-supported catalysts prepared by the same method. An encapsulation effect of partially reduced Mo species over Pd particle was observed. The activity for NO-CH₄ reaction was higher for the binary samples at 200°-500°C. A decrease in the activity (at 500°C) was observed over the Pd-Mo/Al₂O₃ samples after 2 hr under reaction conditions. Higher initial activity on bimetallic catalysts was due to the greater number of active sites and a synergetic effect between Pd and Mo. The main products for the NO-CH₄ reaction were N₂, N₂O, and CO₂. When Pd/Al₂O₃ catalyst was pre-reduced at 300°C, the steady-state conversion of NO was 86%.

Full text of DOI:10.1016/S1381-1169(03)00214-0

Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
Performance of Pd-Mo/␥-Al O catalysts for the
2
3
selective reduction of NO by methane
b,1
a,∗
G.M. Tonetto , D.E. Damiani
a
Chemical Engineering Department, PLAPIQUI (UNS-CONICET), Camino La Carrindanga Km 7,
CC 717 (8000) Bahia Blanca, Argentina
Faculty of Engineering, University of Western Ontario, 1151 Richmond Street, London, Canada N6A 5B9.
b
Received 22 November 2002; received in revised form 16 March 2003; accepted 19 March 2003
Abstract
Pd-Mo/␥-Al2O3 catalysts were prepared by wet impregnation using organometallic precursors. The performance of these
catalysts in the selective catalytic reduction of NO by methane (denoted NO–CH4) was compared with corresponding
palladium- and molybdenum-supported catalysts prepared by the same method.
Characterization by TPR, hydrogen chemisorption and FTIR of CO and NO showed evidence of palladium–molybdenum
interaction. The experimental results indicated an encapsulation effect of partially reduced Mo species over Pd particle.
The catalytic tests revealed that the binary catalysts have different behaviors for the studied reaction. The activity for
◦
NO–CH4 reaction is higher for the binary samples in the temperature range 200–500 C. On the other hand, a decrease in
◦
the activity (at 500 C) was observed over the Pd-Mo/Al2O3 samples after 2 h under reaction conditions, being similar to the
activity displayed by Pd/Al2O3. This is thought to be associated with a poisoning of MoOx sites by the oxygen formed during
the dissociation of NO. The oxygen taken up by the MoOx species does not seem to be removed by methane. Higher initial
activity on bimetallic catalysts is attributed to the greater number of active sites and a synergetic effect between Pd and Mo.
©
2003 Elsevier Science B.V. All rights reserved.
Keywords: Nitrogen oxide; Selective catalytic reduction; Pd catalyst; Molybdenum; Methane
1
. Introduction
to N2 in exhaust gas streams. Methane is one of the hy-
drocarbons of interest. It is the primary component of
natural gas (over 85%), it is relatively inexpensive and
it presents attractive properties as a reducing agent.
Cobalt- and copper-exchanged ZSM-5 catalysts
have been reported to be effective for the CH4-SCR of
The negative impact of man-made NOx emissions
on the environment is well known: nitrogen oxides are
extremely harmful pollutants. Much effort has there-
fore been put into the control of this contaminant.
The selective catalytic reduction by hydrocarbons
◦
NO at reasonable temperatures (400 C) [1]. The poor
(
HC-SCR) is an attractive technique for NOx reduction
thermal and hydrothermal stability of zeolite-based
catalysts, however, makes them impossible to be used
under practical conditions [2]. Alumina-based cata-
lysts, which present high hydrothermal stability, are
potential candidates for the abatement of NO. It is
well known that supported noble metal catalysts (e.g.
Pt, Pd) are active for the reaction of NO with H2, CO
∗
Corresponding author. Tel.: +54-291-4861700;
fax: +54-291-4861600.
E-mail addresses: tonetto@surf.ssw.uwo.ca (G.M. Tonetto),
ddamiani@plapiqui.edu.ar (D.E. Damiani).
1
Tel.: +1-519-6612111x88357; fax: +1-519-6613498.
1
381-1169/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S1381-1169(03)00214-0

2
90
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
or NH3 [3–5]. It has been also reported that they show
a high level of activity for NO–CH4 reaction, but
that they give relatively high selectivity to N2O, NH3
and CO [6]. Finally, the addition of molybdenum to
Pd-supported catalyst has been reported to improve
the catalytic properties for NO reduction [7–10].
Based on these results, and taking into considera-
tion the fact that molybdenum oxide volatilization is
possible at high temperatures, and that molybdenum
stability can be improved by using low Mo loading
Pd-Mo with different catalytic properties to the metals
individually.
As a continuation of those studies, this paper will
report the preparation, characterization and evaluation
in NO–CH4 reaction of Pd-Mo catalysts supported on
alumina.
2. Experimental
[8], we prepared Pd-Mo catalysts by wet impregna-
2.1. Preparation of catalysts
tion. We used acetyl acetonate precursors and low
metallic loading (Mo weight loading was chosen to
be as close at 0.9 wt.% always).
Pd/Al2O3 sample was prepared by wet impregna-
tion of the support with a solution of Pd(C5H7O2)2
(Alfa) in toluene (Merck). The support material was
We have observed the effect of Mo over Pd/Al2O3
catalyst on NO decomposition [11]. The catalytic tests
revealed that the binary catalyst has a different behav-
ior for the NO decomposition, the main differences
being a longer steady-state activity and an increase in
oxygen evolution at high temperatures. Characteriza-
tion results indicated that Pd–Mo interaction appears
to be responsible for the observed modification in NO
activity and selectivity.
In another study [12], adsorption and disso-
ciation of NO on Pd(1 1 1), MoOx-Pd(1 1 1) and
MoOx-␥Al2O3 model surfaces have been investigated
using a molecular orbital approach of the extended
Hückel type (EHMO), including repulsion terms. Our
results of EHMO showed that the NO adsorption
energy on MoOx-Pd(1 1 1) model is lower than the
one for Pd(1 1 1) or MoOx-␥Al2O3. In addition, the
interaction of N2 and O2 with the surface is also
notoriously modified in the bimetallic model. These
results would indicate the existence of an interface
2
␥-Al2O3 (Condea, Puralox, 148 m /g). Prior to im-
◦
pregnation the support was dried under N2 at 150 C
for 2 h. Mo/Al2O3 catalyst was prepared in a similar
manner using a solution of MoO2(C5H7O2)2 (Alfa) in
toluene. After impregnation the catalysts were dried
◦
in Ar at 150 C for 2 h and then calcined in air (chro-
◦
matographic grade) at 500 C during 2 h.
The bimetallic samples were prepared utilizing the
same precursors. The metals were co-impregnated or
impregnated in sequence. The catalyst Pd-Mo/Al2O3
was prepared by wet impregnation of the Mo/Al2O3
with a solution of Pd(C5H7O2)2. On the other hand,
Pd-Mo(c)/Al2O3 catalyst was made by impregna-
tion of alumina with a solution of both precursors
in toluene. After impregnation, the sample was dried
and calcined similarly to the monometallic catalysts.
Table 1 summarizes catalyst preparation data. The Pd
and Mo contents were determined by atomic absorp-
tion spectroscopy (AAS).
Table 1
Characterization of the catalysts
Catalyst
Metal loading (wt.%)
Pd
Preparation
Pdexp (%)a
Mo
Pd/Al2O3
Mo/Al2O3
Pd-Mo(c)/Al2O3
Pd-Mo/Al2O3
0.82
–
0.45
1.00
–
w.i.b
w.i.
55 ± 5
–
0.71
0.74
0.87
w.i.–co-impregnation
w.i.–sequencec
36 ± 5
45 ± 5
Metallic content and hydrogen uptakes as exposed metal percentage.
a
Fraction of exposed palladium.
w.i.: wet impregnation.
Pd impregnated on Mo/Al2O3.
b
c

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
291
2
2
.2. Characterization of catalysts
was tested in flow system using a glass reactor (6 mm
o.d.) operating at atmospheric pressure. The reactor
was mounted horizontally in an electrically heated
oven. The reaction products were analyzed by on
line gas chromatograph operating in TCD mode.
.2.1. H2 chemisorption
The chemisorption uptakes were measured in a con-
ventional glass apparatus [13]. Before the reduction,
the catalysts were oxidized in air at 500 C during 1 h.
The samples were then purged in He and reduced at
5
the samples were evacuated for 20 h at reduction tem-
perature and cooled to adsorption temperature (25 C)
under vacuum. Irreversible uptakes were determined
from dual isotherms measured for hydrogen using the
method of Benson et al. [14]. The fraction of exposed
palladium was calculated assuming that one hydrogen
atom is adsorbed per surface palladium atom.
◦
Two Porapak Q columns operated isothermally at
◦
4
0 and −20 C with He as carrier gas were used for
◦
00 C in flowing H2 during 1 h. Following reduction,
separation. The reaction mixture composition was
35.4 ppm of NO and 183.9 ppm of CH4 in He bal-
ance (stoichiometric mixture, for practical purpose
7
◦
−
11
−3
1
ppm = 4.1 × 10
g mol cm ). Prior to each ac-
tivity measurement, the catalysts were pre-treated (in
◦
situ) under flow of 5% H2 in Ar, at 300 C for 30 min.
Catalyst weight was adjusted in order to keep con-
stant the weight of palladium in the different experi-
ments. The reaction mixture flow rate was maintained
2
.2.2. TPR experiments
TPR experiments were performed in a conventional
3
4
at 35 cm /min with a space velocity of 3.99×10 feed
mol/h mol Pd. Conversion data was measured in a tem-
perature range 150–500 C.
apparatus, as previously described [15]. Before re-
duction, the catalysts were oxidized in flowing chro-
matographic air at 500 C for 1 h and purged and
◦
◦
cooled in Ar. Then, a mixture of 5% hydrogen in ar-
gon was passed through the sample and the tempera-
3
. Results and discussion
◦
ture was raised from −50 to 500 C at a heating rate
3.1. Preparation of catalysts and H2 chemisorption
◦
of 10 C/min.
The choice of the catalysts preparation method re-
2
.2.3. IR studies
The samples were pressed into a 10 mg/cm self-
supporting wafers. Prior to each experiment, the cata-
2
ported here derives from preliminary work carried out
by our laboratory. Different parameters such as the
metal precursors, the metal loading and preparation
technique were analyzed [4,17,18].
Some of the results concerning the characterization
by TPR and hydrogen chemisorption were reported
in our previous work [12] and are here reproduced to
facilitate comparisons.
−
5
lysts were evacuated (10 Torr, 1 Torr = 1.333224 Pa)
◦
at 300 C for 1 h. Then they were exposed to 200 Torr
◦
of O2 at 200 C during 1 h. The O2 was pumped off to
a final pressure of 10 Torr during 1 h at 300 C. Fol-
lowing this, the material was exposed to 300 Torr of
hydrogen at 200 C for 1 h, and finally evacuated for
additional 1 h at 300 C. After that, the catalysts were
exposed to 25 Torr of CO. After adsorption, the sam-
ples were out-gassed and the spectra were recorded at
room temperature. For NO studies, the pre-treatment
of the sample was similar. Following NO adsorption,
the catalysts were out-gassed at 100, 200 and 300 C.
The treatments were carried out using a quartz IR
cell [16]. The infrared spectra were recorded on a FTIR
−
5
◦
◦
◦
Table 1 lists all the prepared catalysts and their
metal content determined by atomic absorption spec-
troscopy. According to the results of Yao [19], the
concentration of saturation of MoO3 on ␥-Al2O3 is
2
4
␮mol Mo/m (BET). Considering this value, 13.2%
◦
of the surface of the Mo/Al2O3 catalyst surface is cov-
ered by molybdenum.
When Pd loadings are compared, palladium con-
centrations are higher for Pd/Al2O3 and Pd-Mo/Al2O3
than in Pd-Mo(c)/Al2O3 samples. Molybdenum load-
ings, however, remain practically constant. This re-
sult is consistent with the type of interactions be-
tween the metal precursors and the support. Several
reports indicate that the fixation of the metal acetyl
−
1
8
300 Shimadzu spectrometer with 4 cm resolution
and 100 scans.
2
.2.4. Catalytic activity measurements
The catalytic activity of the sample was tested for
the reduction of NO with CH4. The catalyst (∼50 mg)

2
92
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
acetonates on alumina occurs by a ligand-exchange
mechanism. A covalented bond is originated and Ha-
surface preventing its contact with H2 [18]. Similar
results were reported for Ru-Mo samples [24] and
Rh-Mo catalysts [25].
cac (C H8O2) is generated, which remains adsorbed
5
on the surface of alumina. This mechanism was found
valid for Pd(acac)2 [17,20,21] and MoO2(acac)2 [22]
and can be represented by the following scheme of
general reaction:
The TPR results for the four samples are showed in
Fig. 1. Mo/Al2O3 presents a very low hydrogen con-
sumption just before 500 C, due to partial reduction
of MoO3. The hydrogen consumption peak is small
as a result of the low Mo loading. In this condition,
the interaction between Mo and alumina is strong and
the reduction of MoO3 requires higher temperatures
[26–28].
◦
S–OH + M(acac)x → S–OM(acac)x−1 + Hacac(ad)
When the catalyst precursor is calcined, the remain-
ing acac ligand combusts and PdO and MoO3 particles
are formed.
On the other hand, the TPR profiles for Pd/Al2O3
◦
Boitiaux et al. [20] reported that acetylacetone
sample presents a low temperature peak, around 25 C.
(
Hacac) generated during fixation of Pd(acac)2 and
This signal is attributed to the reduction of palladium
oxide. In bimetallic samples, we find the signal as-
signed to the reduction of palladium oxide.
adsorbed on the support, affects the impregnation
mechanism of palladium. They found that: (a) a sup-
port capable of retaining 0.3 wt.% of palladium, only
retains 0.07 wt.% when acetylacetone is pre-adsorbed;
and (b) when a support, that has retained 0.3 wt.% of
Pd, is submerged in a solution containing an excess
of acetylacetone, 80% of palladium in the support is
lost into the solution. In view of this, the minimum
palladium loading in the sample Pd-Mo(c)/Al2O3
may be understood as a competition for sites be-
tween the complex of palladium and acetylacetone
generated during the co-impregnating of the precur-
sors Pd(acac)2 and MoO2(acac)2. The molybdenum
complex is not affected.
◦
The negative peak around 60 C is attributed to pal-
ladium hydride decomposition. Its small size shows a
connection with the average particle size presented by
the samples, since the hydride formation is a bulk phe-
nomenon. The hydrogen consumption at high temper-
ature corresponds to the partial reduction of MoO3.
The reduction profiles for the binary samples pre-
sent an additional hydrogen consumption peak in the
◦
80–200 C temperature range. This broad peak would
be related to highly dispersed palladium oxide, which
◦
is reduced at temperatures higher than 100 C [29] or
it could be to assigned to a partial reduction of MoOx
species caused by close contact with the noble metal,
according to similar results reported for Pd-V/Al2O3
catalyst [30]. Another possible explanation would be
the bronze formation, due to the hydrogen spillover
from the noble metal towards MoO3 [31]. This H2
consumption is not present in Pd/Al2O3 sample, which
demonstrates the interaction between Pd and Mo.
Fig. 2 shows the FTIR spectrum of Pd/Al2O3 after
Another bimetallic sample was prepared. Mo-Pd/
Al2O3 catalyst was synthesized by wet impregnation
of Pd/Al2O3 sample with a solution of MoO2(acac)2.
The metal loadings determined by AAS were
0
.95 wt.% of Pd and 0.76 wt.% of Mo. The simi-
larities in metallic contents between Mo-Pd/Al2O3
and Pd-Mo/Al2O3 indicates that Pd and Mo do not
compete for the same sites on alumina.
−
1
1
The hydrogen chemisorption measurement results
exposure to CO. The absorption band at 2094 cm
is assigned to linear CO and the signal at 1975 cm
−
(expressed as Pd exposed fraction) are also summa-
rized in Table 1. Assuming hemispherical particle
is attributed to multiple connected CO [32,33]. The
shoulder present at low frequency (1940 cm ) has
been attributed to multiple coordinated species [34].
19
2
−1
and superficial density of 1.27 × 10 Pds atoms/m ,
metal particle size in Pd/Al2O3 catalyst is 2 nm
(
dp = 1.12/Pdexp) [23]. In relation to Pd monometal-
According to Tessier et al. [35], the absorption band
−
1
2+
lic catalyst, bimetallic samples show a decrease in
the H2 chemisorption capacity of Pd when Mo is
also present. During the preparation procedure, high
oxidation temperature favors the formation of MoO3,
which after exposure to H2 at high temperature re-
duces to a suboxide that tends to migrate onto the Pd
at 2150 cm can be assigned to Pd –C≡O. They
found ionic palladium in catalysts prepared using
◦
acetyl acetonate precursors, after reduction at 300 C.
They consider that the high exposed metallic fraction
(65%) favors a transfer of electrons from the pal-
ladium to the support. The increase in intensity of

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
293
◦
Fig. 1. TPR profiles. Oxidation temperature: 500 C.
bands at 1440 and around 1650 cm 1 is assigned to
−
In the case under discussion, the observed τ for
Pd/Al2O3 and Pd-Mo/Al2O3 samples was 0.5 and
0.7, respectively. These results suggest that molybde-
num addition generates a decrease of multiple con-
nected species on palladium. This observation could
be explained assuming a Pd particle encapsulation by
molybdenum species, and this result is in agreement
with the results of H2 chemisorption. Thus, when the
fraction of exposed Pd decreases, the adsorption of
linear CO is also favored.
carboxilate formation over alumina [36].
Fig. 3 shows the FTIR spectrum of CO adsorbed
on Pd-Mo/Al2O3 sample. the same absorption bands
as in Pd/Al2O3 catalyst were found. Upon Mo addi-
tion, changes were observed in the intensity of linear
and bridged bands. For samples containing the same
species (showing bands at identical positions), it seems
reasonable to admit that the absorption coefficient val-
ues do not vary significantly from one sample to the
other. Therefore, the relationship of relative abundance
between lineal and multiple connected CO could be
estimated integrating the intensities of signals at 2094,
According to Peri [37], characteristic absorption
bands for CO on molybdenum appear around 2025
−
1
0
−1
δ+
and 2045 cm for Mo , and at 2100 cm for Mo .
−
1
1
975 and 1940 cm expressed as follows:
We did not find any signal for CO adsorbed on Mo for
Mo/Al2O3 sample. As shown in Fig. 3, no band cor-
responding to molybdenum for Pd-Mo/Al2O3 catalyst
was found either. This would be due to the absence
I2094
τ = I
+ I1940
1
975

2
94
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
Fig. 2. FTIR spectra of CO absorbed on Pd/Al2O3 sample. References: (a) catalyst; (b) 25 Torr of CO.
Fig. 3. FTIR spectra of CO absorbed on Pd-Mo/Al2O3 sample. References: (a) catalyst; (b) 25 Torr of CO.

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
295
of partially reduced species. No CO adsorption was
observed on calcined Mo/alumina catalysts [37].
Fig. 4 shows the FTIR spectrum of Pd/Al2O3 af-
ter exposure to NO at room temperature, followed
by evacuation at different temperatures. The bands
with the support and making difficult its reducibility
[26–28].
The infrared spectra of NO adsorption on Pd-Mo/
Al2O3 catalyst are presented in Fig. 5. The IR spectra
basically show the same bands present on Pd/Al2O3.
−
1
−1
at 1465, 1314 and 1228 cm
have been reported
The band at 1742 cm corresponding to NO linear
◦
for NO adsorption on alumina due to nitrate and
nitrite complexes [38]. The broad band observed
presents its maximum intensity at 100 C, and dimin-
ishes with temperature. But the Mo addition led to the
suppression of the bands attributed to the adsorption
of NO threefold bridged on Pd. This result may be of
consequence to the smaller space for the adsorption
due to the decoration of palladium particles by molyb-
denum species.
Hoost et al. [42] found a decreasing NO adsorp-
tion on La-promoted Pd/Al2O3 catalyst, which was
attributed to either the coverage of Pd particles with
LaOx patches or to an increase in the NO dissociation.
Noronha et al. [9] did not observe any bands attributed
to the adsorption of NO on Pd in Pd/MoO3/Al2O3 cat-
alyst with 11.6 wt.% of Mo and 1 wt.% of Pd.
Summing up, on the Pd-Mo/Al2O3 catalyst, the de-
crease of H2 adsorption, the suppression of the ad-
sorption of threefold bridged NO–Pd and the increase
−
1
around 1760 cm
sorbed on a-top site over Pd in the Pd/Al2O3 catalysts
39,40]. The intensity of this band increased with
temperature. An intensity increase of the signal is
was assigned to linear NO ad-
[
−
1
observed at 1573 cm , which can be attributed to
NO threefold bridged [38]. This band diminishes with
the temperature and presents its maximum value at
◦
2
00 C.
The FTIR spectra of Mo/Al2O3 after exposure to
NO (not presented here) do not show any characteris-
tic signal for NO adsorbed on Mo species, at 1723 or
−
1
1
819 cm [37,40,41]. It is possible to presume that re-
+
6
duction temperature was too smooth, since Mo does
not adsorb NO [41]. The Mo content in Mo/Al2O3 cat-
alyst is low (0.71%), originating a stronger interaction
◦
◦
◦
Fig. 4. FTIR spectra of NO absorbed on Pd/Al2O3 sample. References: (a) catalyst; (b) 25 Torr of NO at 25 C; (c) 100 C; (d) 200 C;
◦
(e) 300 C.

2
96
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
◦
◦
◦
Fig. 5. FTIR spectra of NO absorbed on Pd-Mo/Al2O3 sample. References: (a) catalyst; (b) 25 Torr of NO at 25 C; (c) 100 C; (d) 200 C;
◦
(e) 300 C.
of the linear/bridged CO ratio are consistent with the
decoration of Pd surface by Mo species.
not present activity for methane, and the same was
observed for pure ␥-alumina. The total conversion of
◦
NO was achieved at 450 C over bimetallic catalysts
◦
and at 500 C for Pd/Al2O3 sample.
3
.2. Catalytic activity measurements
The catalyst’s weight was adjusted in order to keep
the weight of palladium in the different experiments
constant. When Mo was added on Pd/Al2O3 catalyst,
CH4 conversion activity was only slightly better.
The activities of the bimetallic catalysts do not de-
pend on the manner of impregnation.
Fig. 6 shows the NO conversion as a function of
reaction temperature. Pure ␥-alumina was initially
tested as catalyst for NO reduction in the pres-
ence of methane. The experimental data show that
un-promoted alumina exhibits very low activity for
the reduction of NO.
Figs. 8 and 9 show the distribution of products
and reactants as a function of reaction temperature
when the reaction is carried out on Pd/Al2O3 and
Pd-Mo/Al O samples.
The deposition of an active metal on alumina en-
hances the activity. Over Pd/Al2O3 catalyst, the con-
◦
2
3
version of NO increases from 300 C and it reaches
◦
The main products for the NO–CH reaction were
4
8
7% at 500 C. The Mo/Al2O3 catalyst showed low
◦
N , N O and CO . For the purpose of discussing the
2
2
2
activity, it was 20% at 500 C. The bimetallic samples
have better catalytic performance for the reduction of
NO than that of monometallic palladium sample, and
complete conversion of NO was observed over Pd-Mo
product distribution of the CO–CH4 reaction system,
it is sufficient to consider the two following stoichiom-
etry equations:
◦
catalysts at 450 C.
4
NO + CH4 → 2N2 + CO2 + 2H2O
(1)
(2)
Fig. 7 shows the methane conversion as a function
of reaction temperature. The Mo/Al2O3 catalyst did
8NO + CH4 → 4N2O + CO2 + 2H2O

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
297
4
Fig. 6. NO conversion as a function of reaction temperature. SV = 4 × 10 feed mol/h mol Pd.
4
Fig. 7. Methane conversion as a function of reaction temperature. SV = 4 × 10 feed mol/h mol Pd.

2
98
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
4
Fig. 8. Products and reactants distribution as a function of reaction temperature for Pd/Al2O3 catalyst. SV = 4 × 10 feed mol/h mol Pd.
4
Fig. 9. Products and reactants distribution as a function of reaction temperature for Pd-Mo/Al2O3 catalyst. SV = 4×10 feed mol/h mol Pd.

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
299
Over Mo/Al2O3 sample, according Figs. 5 and 6,
the reaction would be:
perature. The total selectivity to N2 was observed at
450 C.
◦
The formation of CO only was detected at
2
2
NO → N2 + O2
(3)
(4)
◦
4
00 C for Pd-Mo/Al2O3 (∼25 ppm) whereas for the
1
2 3
Pd/Al O sample CO was formed between 350 and
NO → N2O + O2
2
◦
4
50 C (∼8 ppm).
In all the cases, the nitrogen balance closed at 7%
Comparing the catalysts, in the bimetallic sample,
on the basis of assuming that species present were N2
and N2O (but it does not allow to discard the possible
formation of NO2 or NH3).
three conclusions emerge:
1. NO production begins at lower temperatures, but
◦
When the reaction was carried out on Pd/Al2O3
sample, the maximum formation of N2O occurred in
from 450 C its formation is not significant.
2. The formation of N always is greater than N O
2
2
◦
the temperature range of 400–450 C. Nitrogen pro-
production.
◦
duction started when the reaction temperature reached
3. Total conversion of NO is reached at 450 C with
total selectivity to N2.
◦
◦
3
50 C and it continued increasing until 400–500 C.
The formation of N2O was a little greater than N2
◦
production, closely at 400 C.
In order to study the effect of the temperature
The catalytic activity of the Pd-Mo/Al2O3 for the
reduction of NO with methane as a function of temper-
ature is presented in Fig. 9. The concentration of N2O
4
pre-treatment on the NO–CH reaction, new samples
were reduced at 300 C (during 30 min), at 500 C
◦
◦
(for 2 h) under flow of 5% H in Ar, or oxidized at
2
◦
◦
went through a maximum at 400 C. Unlike N2O, the
500 C (for 2 h). The catalysts were then purged under
amount of N2 increased constantly along with tem-
He flow. The studied samples were further exposed
◦
4
Fig. 10. NO conversion over Pd/Al2O3 catalyst (after different pre-treatments) at 500 C. SV = 4 × 10 feed mol/h mol Pd.

3
00
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
◦
4
Fig. 11. NO conversion over Pd-Mo/Al2O3 catalyst (after different pre-treatments) at 500 C. SV = 4 × 10 feed mol/h mol Pd.
◦
to the reaction mixture at 500 C for 2 h. Samples of
reaction products were taken regularly.
96% for another 120 min. In bimetallic sample, the
NO conversion was lower, close at 85%.
Figs. 10 and 11 present NO conversion as a function
of the reaction time. Again, since the same palladium
mass was used in the reaction, relative performances
will be discussed.
When samples were oxidized, NO conversion over
Pd/Al2O3 catalyst was ∼77% and for Pd-Mo/Al2O3 it
was initially 83% and then increased slightly (∼90%).
The most significant differences are the following:
Both catalysts present total methane conversion af-
ter the pre-treatments, except when the reaction was
carried out over Pd-Mo/Al2O3 catalyst pre-reduced
◦
•
•
When the samples are reduced at 500 C, the molyb-
denum addition diminishes the activity.
◦
When the samples are reduced at 300 C, the
◦
at 500 C (methane conversion was 95%). Methane
bimetallic catalyst initially presents greater conver-
sion than the monometallic sample, but after 2 h of
reaction, both catalysts exhibited equal activity.
conversion is stoichiometric only on the surface
◦
of Pd/Al2O3 catalyst pre-reduced at 500 C, ac-
cording to the reactions mentioned above ((1) and
(
2)).
◦
When the samples were pre-reduced at 300 C,
Pd-Mo/Al2O3 catalyst exhibited at first total NO
conversion, which then diminished gradually. After
4. Discussion and conclusions
1
20 min, the activity was similar to the one displayed
Characterization by TPR, hydrogen chemisorption
and FTIR of CO and NO adsorbed indicate that palla-
dium physicochemical properties are altered by an ef-
fective interaction with molybdenum. The results sug-
gest a decoration effect of Mo species over Pd parti-
by the monometallic sample. For Pd/Al2O3 catalyst,
the conversion of NO fluctuated near 86%. After re-
duction at 500 C, Pd/Al2O3 catalyst presented total
◦
conversion during 90 min, and then it decreased to

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
301
cle. This interaction appears to be responsible for the
observed modification in NO activity.
After a reducing treatment at 500 C, methane and
Reaction (6) is not considered because no appreciable
formation of CO was observed.
◦
These results are in agreement with those re-
ported by Mouaddib et al. [45]. They observed that
Pd/alumina catalysts could be active in the reformed
reaction of methane with steam at as low temperature
NO conversions over Pd/Al2O3 sample were total. The
only detected products were N2 and CO2. These ob-
servations suggest that the NO–CH4 reaction on Pd
catalysts takes place via NO adsorption and dissocia-
tion on metallic Pd sites forming N2, whereas the gen-
erated oxygen is retained in the catalyst surface and is
removed by reaction with CH4, regenerating the ac-
tive sites. These results are consistent with previous
works of Burch and Scire [43].
◦
as 400 C, in deficient oxygen atmospheres. Pisanu
◦
[44] studied the reduction of NO by C2H at 400 C
6
over Pd/␣-Al2O3 catalyst. She found that the reactions
of reformed with steam are promoted on the surface
of pre-oxidized catalyst, as a result of a dispersion or
reconstruction of oxide of palladium associated to the
reaction of oxidation of C2H6.
After an oxidizing pre-treatment, NO conversion
immediately reaches a 77% steady-state value, while
methane conversion is complete. On the oxidized
surface, the reaction begins with the oxidation of
methane. The removal of oxygen through reaction by
methane is fast at this temperature, thus leading to
an increased availability of sites for the adsorption
and dissociation of NO [44]. The value of conver-
sion for NO in stationary state is smaller than the
◦
When Pd/Al2O3 catalyst is pre-reduced at 300 C,
the steady-state conversion of NO is 86%. This value
was smaller than that one presented by the same cat-
◦
alyst pre-reduced at 500 C. This could be indicating
that the oxidation degree of Pd, and therefore the num-
ber of active sites for NO adsorption, depends on the
reducing temperature of the pre-treatment. Although
TPR study indicated that palladium is in metallic state
◦
◦
one obtained after reduction at 300 and 500 C. This
at temperatures greater than 100 C, FTIR spectrum
2
+
suggests that a different number of palladium atoms
participate in the reaction. The methane conversion is
more than necessary to consume all the oxygen gener-
ated by the decomposition of NO. It could be thought
that the catalyst accumulates oxygen during the ox-
idizing treatment. It is known that O2 is an oxidant
more powerful than NO and it could be cause of the
non-stoichiometric methane consumption. Therefore,
as the reaction advances and the available oxygen
reacts, the activity would have to achieve a level of
stoichiometric stationary state. Methane and NO con-
versions, however, are not modified in 120 min. The
water steam produced during the methane oxidation
of adsorbed CO indicated the existence of Pd –C≡O
◦
for same sample reduced at 200 C. It also has to be
considered that different activities (Fig. 10) could be
the consequence of differences in the conformation of
the surface of the catalyst due to the reduction treat-
ments. The adsorption and dissociation of NO seem to
be sensitive to the presence of edges and steps on the
surface [46–48]. When the catalyst was pre-reduced
◦
under smooth conditions (300 C for 30 min), the re-
action of reformed by steam would also be present,
although in a lesser extent.
Results of characterization performed at different
temperatures suggest the blockage of the Pd surface by
the molybdenum oxide species. When the bimetallic
◦
constitutes an oxygen source at 500 C. Consequently,
◦
it can be considered that the reactions of reformed
with steam (reactions (5) and (6)) are simultaneous
with the reduction of NO with methane (reactions (1)
and (2)).
catalyst is reduced at 500 C, this effect is severe, and
it originates a diminution in the catalytic activity of
the sample.
The fractions of palladium exposed for Pd/Al2O3
and Pd-Mo/Al2O3 samples are 0.55 and 0.45, respec-
tively. These values of dispersion were obtained with
CH4 + 2H2O → CO2 + 4H2
CH4 + H2O → CO + 3H2
(5)
(6)
◦
the catalysts previously reduced at 500 C. Since the
same palladium mass was used in reaction, Pd-Mo/
Al2O3 sample exposes a metallic surface 18.2% lower
than Pd/Al2O3, and its conversion is smaller in 17%.
Molybdenum, on the other hand, is the cause of the
greater initial activity for NO observed on the catalyst
According to reaction (1), when the NO conversion
is 77%, it is generated an amount of water sufficient
to eliminate the excess of methane by reaction (5).

3
02
G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
◦
pre-reduced at 300 C. During the reduction treatment
of Pd-Mo samples, MoOx species are formed. These
sites are proposed in [10] as preferred for the adsorp-
tion and dissociation of NO, this being the first step of
the mechanism postulated for the NO–CH4 reaction.
But, during the course of the reaction, the bimetal-
lic sample is deactivated. The MoOx sites would be
poisoned with the O formed during the dissociation of
NO. The oxygen taken up by the MoOx species does
not seem to be removed by methane.
[7] T. Hoost, G. Graham, M. Shelef, O. Alexeev, B. Gates, Catal.
Lett. 38 (1996) 57.
[
[
8] I. Halasz, A. Brenner, M. Shelef, Appl. Catal. B 2 (1993) 131.
9] F. Noronha, M. Baldanza, R. Monteiro, D. Aranda, A. Ordine,
M. Schmal, App. Catal. A 210 (2001) 275.
[
[
[
[
10] M. Baldanza, L. de Mello, A. Vannice, F. Noronha, M.
Schmal, J. Catal. 192 (2000) 64.
11] A. Eberhardt, E. Benvenutti, C. Moro, G.M. Tonetto, D.E.
Damiani, J. Mol. Catal. A 201 (2003) 247–261.
12] G.M. Tonetto, M. Ferreira, D.E. Damiani, J. Mol. Catal. A
193 (2003) 121.
13] A. Pisanu, C. Gigola, Appl. Catal. B 11 (1996) L37.
[14] J. Benson, H. Hwang, M. Boudart, J. Catal. 30 (1973)
146.
15] G.M. Tonetto, M. Ferreira, D.E. Damiani, J. Mol. Catal. A
171 (2001) 123.
16] J. Foschiera, T. Pizzolato, E. Benvenutti, J. Braz. Chem. Soc.
◦
When Pd-Mo sample was pre-reduced at 300 C,
the activity for NO conversion after 2 h in reaction was
similar to the displayed by Pd/Al2O3 catalyst. Accord-
[
◦
ing to CO and NO FTIR results (performed at 200 C),
[
Pd particles are partially covered by Mo species. Nev-
ertheless, Pd seems to be as active in this case (Pd-Mo
sample) as it is on monometallic Pd where the disper-
sion is higher. The modifications in the catalytic per-
formance should not be attributed to the simple fact
that more active sites are available on the binary sam-
ple than in sample Pd/alumina. A synergetic effect be-
tween palladium and Mo should be invoked.
1
2 (2001) 159.
[17] C. Neyertz, M. Volpe, Colloids Surf., A 136 (1998) 63.
[18] L. Konopny, A. Juan, D.E. Damiani, Appl. Catal. B 15 (1998)
115.
[
[
19] H. Yao, J. Catal. 70 (1981) 440.
20] J.P. Boitiaux, J. Cosyns, S. Vasudevan, in: G. Poncelet, P.
Grange, P.A. Jacobs (Eds.), Preparation of Catalysts, vol. III,
Elsevier, Amsterdam, 1983, p. 123.
[21] E. Lesage-Rosenberg, G. Vlaic, H. Dexpert, P. Legarde, E.
Freund, Appl. Catal. 22 (1986) 211.
[
[
[
[
22] O. Klimov, M. Fedotov, N. Startsev, J. Catal. 139 (1993) 142.
23] P. Aben, J. Catal. 10 (1968) 224.
24] A. Juan, D.E. Damiani, J. Catal. 137 (1992) 77.
25] N. Takahashi, T. Mori, A. Miyamoto, T. Hattori, Y. Murakami,
Appl. Catal. 38 (1988) 301.
Acknowledgements
This work was supported with funding from the
Universidad Nacional del Sur and the Consejo Na-
cional de Investigaciones Cient ´ı ficas y Técnicas
[26] S. Rajagopal, H. Marini, J. Mazari, R. Miranda, J. Catal. 147
(1994) 417.
[
[
27] L. Wang, W. Hall, J. Catal. 66 (1980) 251.
28] H. Knözinger, H. Jezioroswski, J. Phys. Chem. 83 (1979)
1166.
(CONICET). GT thanks Dr. H. de Lasa, from Chemi-
cal Reactor Engineering Centre (Faculty of Engineer-
ing Science, University of Western Ontario) for his
advice and support during the writing of this paper.
Authors thank Dr. E. Benvenutti and Dr. C. Moro
from Universidad Federal de Rio Grande do Sul for
their assistance in FTIR studies and translator-to-be
Melisa Tonetto for proofreading.
[29] H. Lieske, J. Volter, J. Phys. Chem. 89 (10) (1985) 1842.
[
30] C. Neyertz, M.A. Volpe, C. Gigola, Catal. Today 57 (2000)
55.
31] P. Sermon, G. Bond, J. Chem. Soc., Faraday Trans. 1 (72)
1976) 730.
2
[
(
[32] F. Bonzon-Verduraz, D. Tessier, A. Rakai, J. Catal. 127 (1991)
457.
[
[
[
33] A. Ortega, F. Hoffmann, A. Bradsshaw, Surf. Sci. 119 (1982)
9.
34] V. Pitchon, M. Primet, H. Praliaud, Appl. Catal. 62 (1990)
17.
35] D. Tessier, A. Rakai, F. Bozon-Verduraz, J. Chem. Soc.,
Faraday Trans. 88 (5) (1992) 741.
7
References
3
[
1] Y. Li, J.N. Armor, Appl. Catal. B 1 (1992) L31.
2] C. Torre-Abreu, M.F. Ribeiro, C. Enriques, F.R. Ribeiro,
Catal. Lett. 43 (1997) 25.
[
[36] M. Pereira da Silva, R. Mello Viera, M. Schmal, Appl. Catal.
190 (2000) 177.
[
[
[
3] K. Tanaka, A. Sasahara, J. Mol. Catal. A 155 (2000) 13.
4] A. Pisanu, C. Gigola, Appl. Catal. B 20 (1999) 179.
5] I. Balint, A. Miyazaki, K.-I. Aika, Appl. Catal. B 37 (2002)
[37] J. Peri, J. Phys. Chem. 86 (9) (1982) 1615.
[38] T. Hoost, K. Otto, A. Laframboise, J. Catal. 155 (1995) 303.
[39] K. Almusaiteer, S. Chuang, J. Catal. 180 (1998) 161.
[40] F. Noronha, M. Baldanza, M. Schmal, J. Catal. 188 (1999)
270.
217.
[6] R. Burch, A. Ramli, Appl. Catal. B 15 (1998) 49.

G.M. Tonetto, D.E. Damiani / Journal of Molecular Catalysis A: Chemical 202 (2003) 289–303
303
[
[
41] N.-Y. Topsoe, H. Topsoe, J. Catal. 75 (1982) 354.
42] T. Hoost, G. Graham, M. Shelef, O. Alexeev, B.C. Gates,
Catal. Lett. 33 (1996) 57.
[46] M. Demicheli, L. Hoang, J. Ménézo, J. Babier, M.
Pinabiau-Carlier, Appl. Catal. A 97 (1993) L11.
[47] S. Schneider, S. Ringler, P. Girard, G. Maire, F. Garin, D.
Bazin, in: A. Corma, F. Melo, S. Mendioroz, J. Fierro (Eds.),
Studies in Surface Science and Catalysis, vol. 130, Elsevier,
Amsterdam, 2000, p. 641.
[
[
43] R. Burch, S. Scire, Appl. Catal. B 3 (1994) 295.
44] A. Pisanu, Ph.D. Thesis, Universidad Nacional del Sur,
Argentina, 1997.
[
45] N. Mouaddib, C. Feumi-Jantou, E. Garbowski, M. Primet,
[48] T. Zambelli, J. Winterllin, J. Trost, G. Ertl, Science 273
(1996) 1688.
Appl. Catal. 87 (1992) 129.