Methane combustion activity of Pd-PdOx-Pt/Al2O 3 catalyst: The role of platinum promoter

Article abstract of DOI:10.1016/j.molcata.2011.12.023

The effect of Pt promoter on Pd-PdO_x/Al₂O₃ catalyst was investigated to improve understanding of the role of promoter formation of the active phase of Pd-PdO_x-Pt/Al₂O ₃ methane combustion catalysts. The influence of Pt addition on the composition of the active catalyst surface was characterized by NH₃ temperature programmed desorption procedure. Addition of promoter improved the low temperature CH₄ combustion activity of the aged catalyst, though it caused a decrease in activity of the fresh catalyst. The changes in activity were related to changes in the ratio of the metal to metal oxide phases in the surface composition. Pt tends to create metal sites, whereas Pd prefers the formation of metal oxide sites. Catalyst preparation parameters affect the Pd/PdO_x ratio of the non-promoted monometallic Pd catalyst, and the ratio needs to be known so that the proper amount of Pt promoter can be added.

Full text of DOI:10.1016/j.molcata.2011.12.023

Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Methane combustion activity of Pd–PdOx–Pt/Al2O3 catalyst: The role
of platinum promoter
Niko M. Kinnunen, Janne T. Hirvi, Mika Suvanto∗, Tapani A. Pakkanen∗
University of Eastern Finland, Department of Chemistry, P.O. Box 111, FI-80101 Joensuu, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
The effect of Pt promoter on Pd–PdOx/Al2O3 catalyst was investigated to improve understanding of the
role of promoter formation of the active phase of Pd–PdOx–Pt/Al2O3 methane combustion catalysts.
The influence of Pt addition on the composition of the active catalyst surface was characterized by NH3
temperature programmed desorption procedure. Addition of promoter improved the low temperature
CH4 combustion activity of the aged catalyst, though it caused a decrease in activity of the fresh catalyst.
The changes in activity were related to changes in the ratio of the metal to metal oxide phases in the
surface composition. Pt tends to create metal sites, whereas Pd prefers the formation of metal oxide sites.
Catalyst preparation parameters affect the Pd/PdOx ratio of the non-promoted monometallic Pd catalyst,
and the ratio needs to be known so that the proper amount of Pt promoter can be added.
© 2011 Elsevier B.V. All rights reserved.
Received 22 June 2011
Received in revised form
17 December 2011
Accepted 20 December 2011
Available online 31 December 2011
Keywords:
Methane
Combustion
Alumina
PdO
Pt
Pd
1. Introduction
support materials in methane combustion catalysts are alumina
and zirconium oxides and their mixtures. Noble metals are applied
Natural gas and different blends of natural gas are of great inter-
est nowadays in the effort to move toward less polluting transport
fuels. Liquefied natural gas, which is natural gas in liquid form, con-
tains methane as the main component. Methane is the most difficult
of all hydrocarbons to oxidize completely [1] and it is a 20 times
more potent greenhouse gas than CO2 [2]. Moreover, high energy
density of methane increases the combustion temperature, which
causes more NOx emissions. Other pollutants, such as unburned
hydrocarbons (UHCs) and CO, are released simultaneously if the
conditions at a running engine are unfavorable for complete oxida-
tion [3,4]. A blend of compressed natural gas and hydrogen, called
Hythane®, has been developed to reduce emissions particularly in
heavy duty vehicles [4]. After-treatment of the exhaust gases is typ-
ically done in a three-way catalytic converter system consisting of
oxidation and reduction catalysts where CO is oxidized to CO2 and
hydrocarbons (HCs) to H2O and CO2, while NOx is reduced to N2
and O2 [5,6].
as active metals [7–9], and platinum group metals have proved
to be a good option for methane combustion [3,10,11]. Promot-
ers improve catalytic activity [12–14], increase thermal stability of
the support material [15,16], and retard sintering of active metal
particles [17].
Platinum promoter has been shown to increase low temperature
methane combustion activity and hinder the growth of Pd and PdO
particle size [18] and crystallite size [19]. Small active metal cluster
size has been linked to longer durability of the promoted catalyst in
hydrocarbon combustion, but no correlation between particle size
and methane combustion activity has been found for either the
fresh or aged monometallic Pd–PdOx catalysts [20,21]. Although
too high Pt content suppresses the activity of the catalyst [22], no
simple correlation has been indicated between the amount of Pt
and methane combustion activity. The role of the Pt promoter is
still unclear [19,22].
The methane combustion activity and thermal stability of the
active metal particles can also be tailored without a promoter
through proper choice of the active metal precursor and the
impregnation solvent [20,21]. Support is thereby provided for a
relationship between methane combustion activity of monometal-
lic Pd–PdOx catalysts and the ratio of metal (M) to metal oxide
(MOx). The target of the present study is to improve understanding
of the effect of Pt promoter on active phase formation and par-
ticularly on the M/MOx ratio of the alumina supported Pd–PdOx
catalyst. The reason for the improved thermal durability of the aged
Catalysts for oxidation of CO and HCs consist of support material,
active metal(s) and promoter(s). The support material has a high
specific surface area to facilitate good dispersion of the active metal.
It also increases the thermal stability of the metal [1]. Conventional
∗
Corresponding authors. Tel.: +358 13 251 3312; fax: +358 13 251 3390.
E-mail addresses: Mika.Suvanto@uef.fi (M. Suvanto), Tapani.Pakkanen@uef.fi
(T.A. Pakkanen).
1381-1169/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.12.023

N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
21
Table 1
Compositions and preparation parameters of the catalysts.
Catalyst
Amountofactivemetalprecursor(g)
Amount of active
metals (mmol)
Total active
metal loading
(wt.%)
Impregnation
time (h)
Volume of impregnation
solution (cm3)
Pd(acac)2
Pt(acac)2
Pd
Pt
Pd(24 h)
Pd(48 h)
Pd(72 h)
0.54
0.54
0.54
0.54
0.54
0.54
0.54
0.17
–
–
–
–
–
1.77
1.77
1.77
1.77
1.77
1.77
1.77
0.56
–
–
–
–
–
2.3
2.3
2.3
2.3
2.3
2.6
3.0
3.0
2.3
4.1
24
48
72
24
16
16
16
25
Pd(25 cm3)
Pd(35 cm3)
Pd0.92–Pt0.08
Pd0.85–Pt0.15
Pd0.37–Pt0.63
Pt(2.3 wt.%)
–
–
24
35
a
a
b
0.06
0.13
0.38
0.38
0.69
0.15
0.33
0.97
0.97
1.75
24 + 24
24 + 24
24 + 24
24
16 + 16
16 + 16
16 + 16
16
Pt(4.1 wt.%)c
–
–
24
16
a
Based on monometallic Pd(24 h) catalyst.
Based on monometallic Pt(2.3 wt.%) catalyst.
Pt mole fraction is the same as the Pd mole fraction of Pd(24 h) catalyst.
b
c
promoted Pd–PdOx/Al2O3 catalyst is also of interest. The influence
of parameters used in the preparation of monometallic Pd–PdOx
catalysts is discussed.
bath (T = 323 K), since heating improves solubility of the precursor
and decreases the surface tension of the impregnation solution [24].
After 15 min mixing, freshly prepared and calcined Pd(24 h) catalyst
was added slowly to the precursor solution. The mixture was stirred
for 24 h. The impregnation solution was removed under vacuum by
heating the flask in a water bath (T = 313 K). After evaporation, the
catalyst was calcined and aged by the procedures described above
(Section 2.1.1). The same stepwise procedure was used to prepare
the Pd0.37–Pt0.63 catalyst with high Pt proportion, but now the Pd
precursor was added to the Pt(2.3 wt.%) catalyst to avoid complete
masking of the active Pd sites.
2. Experimental
Catalysts were prepared with alumina as a support material. The
size distribution of alumina particles was <25 ␮m 37.8%, <45 ␮m
64.4%, and <90 ␮m 92.0%. The amount of support material was
8.0 g for each batch, and total active metal loading varied between
2.3 and 4.1 wt.%. The active metal precursors were Pd(C5H7O2)2
and Pt(C5H7O2)2, hereafter referred to as Pd(acac)2 and Pt(acac)2.
Pure propionic acid (Fluka, p.a.) was diluted with water to prepare
impregnation solution of 50 vol.% propionic acid–water solution
(P50).
2.2. Characterization
Crystallinity and crystallite size of the active metal and metal
oxide species were determined by powder X-ray diffraction (XRD).
The XRD measurements were carried out with a Bruker-AXD D8
Advance device using Cu K␣ as radiation source. The diffraction
pattern was collected between 9◦ and 93◦ at 2ꢀ scale with a step
size of 0.11◦ min−1. The crystallite sizes were evaluated from the
diffraction data with TOPAS 2.0 software [25].
2.1. Catalyst preparation
Monometallic Pd catalysts with different preparation parame-
ters and monometallic Pt catalysts with different loadings were
prepared by wet impregnation method. Bimetallic catalysts were
prepared by a stepwise impregnation method. In the stepwise
impregnation, the catalyst undergoes the wet impregnation cycle
twice to ensure controlled distribution and better contact between
active metals [23]. The composition of the catalysts and details of
the preparation are presented in Table 1.
Dispersion was measured by CO chemisorption. The sam-
ple (∼200 mg) was first pretreated under H2 at a flow rate of
30 cm3 min−1 and heated to 773 K at a heating rate of 20 K min−1
.
Cooling to room temperature was carried out under helium at a
flow rate of 30 cm3 min−1. The dispersion was then measured by CO
pulse chemisorption at room temperature. The flow rate of helium
during the chemisorption measurement was 10 cm3 min−1. Evalu-
ation of the dispersion was carried out by GRAMS/32 software [26]
by using a stoichiometric factor of 2 for Pd [27–30] and 1 for Pt.
Specific surface area of the catalyst was measured with
a Micromeritics ASAP 2010 device by Brunauer–Emmett–Teller
method (BET). A sample (∼200 mg) was pretreated in vacuum at
623 K for 2 h and the measurement was carried out at the temper-
ature of liquid nitrogen. More specific surface characterization and
identification of the concentrated active species were carried out
with a Hitachi S4800 FE-SEM. Detailed surface analysis was done
with a Thermo Electron EDS device. Working distance was 15 mm,
accelerating voltage 20 kV, and current 20 ␮A.
Activity of the catalyst in methane combustion was measured
under lean burn conditions. The test gas, produced by AGA, con-
sisted of CH4 (1%), O2 (16%), and N2 (83%). A catalyst sample of
600 mg was used. The activity performance was recorded three
times for each catalyst, each time with a new sample. A detailed
description of the methane combustion procedure and device
can be found elsewhere [20,31]. Methane conversion rate (MCR)
2.1.1. Wet impregnation
The Pd(acac)2 or Pt(acac)2 was weighed into a 50-cm3 flask and
stirred with 10 cm3 of impregnation solution. The alumina support
was weighed in a weighing boat and slowly added to the active
metal solution. The remaining impregnation solution was then
added. The total amounts of impregnation solutions and impreg-
nation times used in catalyst preparation are presented in Table 1.
The impregnation solution was removed during the evaporation
step by heating the flask at 313 K in a water bath. When the batch
was completely dry, it was removed to a drying tray. The tray was
placed in a furnace, where the batch was dried 10 h at 418 K. After
drying, the batch was calcined under air for 2 h at 823 K. Half of the
batch was aged under air for 4 h at 1173 K.
2.1.2. Stepwise impregnation
A stepwise method was used to prepare Pt promoted catalysts,
Pd0.92–Pt0.08 and Pd0.85–Pt0.15, in which the monometallic Pd(24 h)
catalyst was impregnated with promoter. The platinum precursor
was dissolved in the P50 solution, which was heated in a water

22
N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
Fig. 1. XRD graphs of fresh (a) Pd(24 h), (b) Pd0.92–Pt0.08, and (c) Pd0.85–Pt0.15 catalysts and aged (d) Pd(24 h), (e) Pd0.92–Pt0.08, and (f) Pd0.85–Pt0.15 catalysts.
[molCH4 g−1 −1] based on mass of the active metal (M) was calcu-
s
Pd
3. Results and discussion
lated by Eq. (1):
3.1. Characterization by XRD, CO chemisorption and BET methods
((XCH4 /100) × 8.17 × 10−6 molCH4 s−1
)
MCR =
(1)
mM
The effect of Pt promoter addition on the X-ray diffraction
pattern of the Pd–PdOx/Al2O3 catalyst and the characteristic
reflections of the metallic Pd and Pt, PdO, and alumina sup-
port are presented in Fig. 1. No reflections for platinum oxides
were observed. Diffraction patterns of the five monometallic
Pd–PdOx/Al2O3 catalysts were similar to each other and also to
those of our previous study [21]. Addition of Pt promoter causes a
new reflection at 2ꢀ scale of 82◦ on the aged catalyst (marked with
an arrow in Fig. 1). Clear shoulders at 40◦ and 46◦ can be seen. The
intensities of the corresponding peaks increase with the amount of
Pt promoter and hence can be attributed to metallic platinum. As is
clear from Fig. 2, neither crystalline PtO nor PtO2 phase was present
for the monometallic Pt catalysts, indicating that the Pt exists only
in metallic form. Fig. 2 also suggests that a high proportion of Pt
influences the ratio of the Pd and PdO phases, for the PdO peaks do
not appear for the Pd0.37–Pt0.63 catalyst.
The PdO crystallite size was evaluated from the characteris-
tic peak at 2ꢀ scale of 34◦ and Pt crystallite size from the peak
at 82◦. The results in Table 2 indicate that neither the impregna-
tion parameters nor the promoter affect the PdO crystallite size
of the fresh or aged catalysts. Earlier, the introduction of Pt pro-
moter was found to increase slightly the crystallite size of PdO [22].
In contrast, the Pt crystallite size of the aged bimetallic catalysts
was clearly smaller than that of the monometallic Pt catalysts. We
note, however, that the particles may consist of multiple crystalline
domains, and the particle and crystallite sizes are not necessarily
comparable [33].
As can be seen in Table 2, the dispersions of active metal are
between 23.5% and 46.6% for the fresh catalysts and between 2.3%
and 11.7% for the aged catalysts. The preparation parameters do
not affect the dispersion of the monometallic Pd–PdOx/Al2O3 cat-
alysts. Moreover, the bimetallic fresh and aged catalysts exhibit
similar active metal dispersion to the fresh and aged monometal-
lic Pd–PdOx catalysts. However, dispersion of the monometallic Pt
catalysts drops dramatically with aging. In contrast to our result,
Castellazzi et al. [22] observed the addition of Pt promoter to
decrease the dispersion of the active phase. BET surface areas of
fresh catalysts (Table 2) are about 210 m2 g−1 and those of the aged
catalysts about 150 m2 g−1. The decrease in BET surface area dur-
ing aging is due to the high temperature and gas atmosphere. The
surface areas of the support material [34] and the catalysts [35]
where XCH is methane conversion at the measurement tempera-
4
ture [%], 8.17 × 10−6 molCH4 s−1 is the total methane flow rate per
second, and mM is the mass of active metal [g]. Turnover rate (TOR)
[s−1] was calculated by Eq. (2):
((XCH4 /100) × 8.17 × 10−6 mol 4 s−1
)
CH
TOR =
(2)
((DM/100) × nM
)
where DM is active metal dispersion [%] and nM is the amount of
active metal [mol].
Thermal stability of the bulk active metal oxides was studied
by temperature programmed oxidation (TPO). Catalyst samples
were not pretreated. The measurement was carried out under gas
mixture of He (20 cm3 min−1) and O2 (3 cm3 min−1). The TPO curve
was recorded while the catalyst was heated at a rate of 20 K min−1
from 303 K to 1323 K. Before recording of the data, the signal
was stabilized for 18 min. During the measurement, a cold trap
(T = 213 K) was used to remove water vapor. The composition of the
desorbing gas, below 500 K, was identified with a Hewlett-Packard
gas chromatograph.
Characterization of acid sites of the catalyst was carried out by
temperature programmed ammonia desorption (NH3 TPD), which
is a quantitativemethod[32]. However, the presentNH3 TPDresults
are reported and compared at a semi-quantitative level, as in our
previous work [21], because the exact composition of the desorb-
ing gas was unknown. An Autochem 2100 device was utilized for
the TPD measurements. A sample of 150 mg was heated to 823 K
(20 K min−1) under helium at a flow rate of 30 cm3 min−1. After
cooling, NH3 gas was introduced to the catalyst, using 10 vol.%
blend of NH3 and He gas (20 cm3 min−1) at 313 K for 1 h. Physically
adsorbed NH3 was removed by flushing the sample with helium
(20 cm3 min−1) for 1 h. Finally, the sample was heated at 10 K min−1
from 313 K to 823 K to record the NH3 desorption curve. Integration
of the peaks was carried out using GRAMS/32 program [26]. The
ammonia desorption results were confirmed with XPS measure-
ments (ESCA3000, VG Microtech Inc., UK) of fresh and aged Pd(24 h)
and fresh Pt(4.1 wt.%) catalysts. The XPS spectra were recorded
before and after annealing treatment at 823 K for 1 h under vacuum,
which simulates the pretreatment of the NH3 TPD measurement.
The binding energy region around 335 eV has been considered.

N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
23
Fig. 2. XRD graphs of fresh (a) Pd0.37–Pt0.63, (b) Pt(2.3 wt.%), and (c) Pt(4.1 wt.%) catalysts and aged (d) Pd0.37–Pt0.63, (e) Pt(2.3 wt.%), and (f) Pt(4.1 wt.%) catalysts. Note the
difference in scale bars.
Fig. 3. SEM images (×30,000) of the aged (A) Pd(24 h), (B) Pt(2.3 wt.%), (C) Pt(4.1 wt.%), and (D) Pd0.37–Pt0.63 catalysts.
Table 2
PdO and Pt crystallite sizes, active metal dispersions, and BET surface areas of the studied catalysts.
Catalyst
PdO crystallite size (nm)
Pt crystallite size (nm)
Activemetaldispersion(%)
BET surface area (m2 g−1
)
Fresh
Aged
Fresh
Aged
Fresh
Aged
Fresh
Aged
Pd(24 h)
Pd(48 h)
Pd(72 h)
2.7
3.1
3.2
3.0
3.1
3.4
3.3
n.d.a
–
12.9
14.7
15.9
14.5
14.7
12.1
12.6
n.d.a
–
–
–
–
–
–
–
–
–
25.6
26.6
23.5
25.7
26.7
27.4
24.2
29.0
24.0
46.6
11.7
9.2
7.8
206
216
214
211
211
202
210
213
211
212
148
150
143
141
166
151
144
142
145
146
Pd(25 cm3)
Pd(35 cm3)
Pd0.92–Pt0.08
Pd0.85–Pt0.15
Pd0.37–Pt0.63
Pt(2.3 wt.%)
Pt(4.1 wt.%)
8.5
–
–
10.1
12.1
9.5
11.6
4.2
n.d.a
n.d.a
35.9
33.9
30.8
32.4
36.5
43.2
68.3
74.8
–
–
2.3
a
No peak was detected.

24
N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
Fig. 4. Methane conversion curves of the fresh (solid lines) and aged (dashed lines)
Pd(24 h) and Pd0.85–Pt0.15 catalysts.
decrease considerably more in aging under oxidizing conditions
than under reducing conditions.
3.2. SEM/EDS characterization
The support material of the catalysts consisted of nano- and
microsized alumina particles. The SEM studies indicated that long
impregnation time (72 h) may decrease the particle size of the
microsized support, whereas the amount of impregnation solu-
tion has no effect. Fig. 3 presents SEM images of the microsized
support particles of the aged monometallic Pd(24 h), Pt(2.3 wt.%),
Pt(4.1 wt.%) and bimetallic Pd0.37–Pt0.63 catalysts. Maintenance of
dispersion is an important property. Inspection of Fig. 3A and B
shows how aging increases the particle size of the different active
metals at equal metal loadings (2.3 wt.%). The growth of Pt metal
particles is still clearer in a comparison of catalysts with equal
Fig. 6. Effect of (A) impregnation time, (B) volume of impregnation solvent, and
(C) Pt promoter on TPO profiles of the fresh (solid lines) and aged (dotted lines)
catalysts.
metal mole fractions (Fig. 3A and C). Large Pt-rich particles were
confirmed by EDS. In general, alumina-supported Pt catalysts were
fairly unstable in aging. However, Fig. 3D shows that the growth
of particles is hindered in the bimetallic catalyst even when the
Fig. 5. Methane conversion curves of the fresh (solid lines) and aged (dashed lines)
Pt(2.3 wt.%), Pt(4.1 wt.%), and Pd0.37–Pt0.63 catalysts.

N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
25
proportion of Pt is large. These observations are in good agreement
with our dispersion results (Table 2).
and the Pd0.37–Pt0.63 catalyst with high Pt proportion do not show
the thermal decomposition peak of the oxide.
3.3. Methane combustion activity
3.5. Determination of active sites
Methane conversion curves for the fresh and aged Pd(24 h)
and Pd0.85–Pt0.15 catalysts are compared in Fig. 4. Activity is bet-
ter for the aged Pt promoted Pd0.85–Pt0.15 catalyst than for the
aged monometallic Pd(24 h) catalyst, showing that Pt promoter
improves the durability of the catalyst. The conversion profiles of
the fresh catalysts show the reverse order. Activities of the fresh and
aged Pd0.37–Pt0.63 catalysts, with high Pt proportion, are notably
lower than the activities of the Pd0.85–Pt0.15 (cf. Figs. 4 and 5)
and other bimetallic catalysts. In fact, like both the fresh and aged
Pt(2.3 wt.%) and Pt(4.1 wt.%) catalysts, the aged Pd0.37–Pt0.63 cata-
lyst is inactive below 623 K (Fig. 5). This is an interesting finding,
since the active metal dispersion of the aged Pd0.37–Pt0.63 is clearly
higher than that of the monometallic Pt catalysts (see Table 2). Evi-
dently dispersion is not necessarily the dominant factor for low
temperature activity in methane combustion.
Table 3 summarizes methane light-off temperatures, methane
conversions, methane conversion rates (MCR), and turnover rates
(TOR) of the catalysts. The best TOR values are achieved for the
fresh monometallic Pd–PdOx catalysts. The values are slightly bet-
ter than those reported in the literature [36,37] and in agreement
with our previous results [21]. The impregnation parameters do
not have a significant effect on the TORs. The turnover rates of
the fresh and aged monometallic Pt catalysts are low, as expected,
since Pt is a less active metal in methane combustion [42]. The
fresh Pd0.92–Pt0.08 and Pd0.85–Pt0.15 catalysts exhibit TORs between
those of the monometallic Pd–PdOx and Pt catalysts, whereas the
aged Pd0.85–Pt0.15 catalyst exhibits highest activity of all the aged
catalysts.
The NH3 TPD profiles of noble metal catalysts show three peaks
corresponding to desorption from alumina, metal oxide (MOx), and
metal (M) sites (Fig. 7). Table 4 summarizes the peak integrals and
calculated M/MOx ratios obtained by NH3 TPD and XPS methods.
Although the preparation parameters affect the M/MOx ratio in
some degree, no clear trends are evident, at least not for the fresh
catalysts. Small differences in homogeneity and calcination of the
catalysts may screen the effects for the fresh catalysts. A slight trend
in the M/MOx ratios can be seen after aging.
The effect of Pt promoter on the NH3 TPD profile of the fresh
Pd–PdOx catalyst (Pd(24 h)) is presented in Fig. 7. Addition of Pt
increases the integral of the last broad desorption peak related
to the metal (M) sites. This is in agreement with the finding that
the fresh and aged monometallic Pt catalysts show only a peak for
metallic Pt (Fig. 8). Furthermore, XPS measurement of the fresh
Pt(4.1 wt.%) catalyst after the thermal treatment at 823 K gives no
indication for the existence of PtOx. In general, the M/MOx ratio
is highest for the catalyst with highest Pt proportion. The effect
of aging on the NH3 TPD profiles of the catalysts is shown in
Figs. 7 and 8. Previously we have concluded that low temperature
activity in methane combustion decreases with aging because the
relative amount of PdOx sites increases and thus the M/MOx ratio
decreases [21]. The observation is also supported by the present
XPS data shown in Fig. 9. The same conclusion can be drawn for
the bimetallic catalysts, though addition of Pt promoter may in fact
lead to slightly higher M/MOx ratio by sustaining the metallic form.
As an extreme, the high proportion of Pt in the Pd0.37–Pt0.63 cata-
lyst seems to prevent the re-oxidation of palladium, as was earlier
suggested for Pt-promoted Pd catalysts in general [22]. Comparison
of the NH3 desorption curves in Fig. 8 shows the MOx related peak
to exist for the fresh Pd0.37–Pt0.63 catalyst, but it disappears during
aging.
3.4. Thermal stability of active species
TPO profiles of the supported monometallic and Pt-promoted
Pd–PdOx catalysts include two desorption peaks (Fig. 6). The first
peak occurs between 350 K and 550 K and is due to the desorption of
O2 [38,39]. Our gas chromatograph analysis showed the desorbing
gas to contain CO2 as well. The presence of CO2 is observed because
the sample is stored under air and CO2 binds to the acid sites of
the catalyst. The second desorption peak is due to the thermal
decomposition of PdO to metallic Pd, which is reported generally to
occur between 973 K and 1173 K [38,40,41]. According to Gélin and
Primet [42], PdO is stable up to 1073 K in air under normal pres-
sure, whereas the most stable platinum oxide (PtO2) under similar
conditions decomposes at 673 K.
The maximum of the PdO decomposition peak for the
monometallic Pd–PdOx catalysts lies between 1200 K and 1300 K.
The change in the peak position may be due to high O2 concentra-
tion of the measurement gas mixture or to strong metal–support
interaction [43]. An increase in impregnation time improves the
thermal stability of the both fresh and aged catalysts (Fig. 6A). The
improvement in thermal stability of PdO can be explained by the
progress of reaction between carboxylic acid, aluminum oxide and
palladium precursor [44]. Increase in the volume of impregnation
solution also increases the thermal stability of the fresh catalysts,
but not that of the aged catalysts (Fig. 6B). After addition of the Pt
promoter, the thermal decomposition peak of oxide of the fresh
catalysts becomes broader, and the maximum appears between
1150 K and 1250 K (Fig. 6C). With aging, the maximum of the peak
moves noticeably toward lower temperature and sharpens, as was
also observed earlier [22]. An increase in the amount of Pt causes a
decrease in the area of the decomposition peak, which corresponds
to the smaller amount of the oxide. The monometallic Pt catalysts
3.6. Active phase of methane combustion catalyst
Coexistence of metallic Pd and PdOx has been observed on
methane combustion catalysts [45,46]. Several groups also sug-
gest that the simultaneous presence of these phases is essential
for formation of the active catalyst in CH4 oxidation [37,47,48]. We
observed earlier for the monometallic Pd catalyst that low tempera-
ture methane combustion activity correlates with the Pd/PdOx ratio
[21]. We concluded, moreover, that an important reason for the loss
of activity under oxidizing aging conditions is decrease in the num-
ber of metal sites. In the present study, the negative effect of aging
was successfully reduced through addition of the Pt promoter. As
also indicated by our XRD, TPO, and NH3 TPD results, platinum is
hard to oxidize and it remains in metallic state even under oxidizing
conditions [42]. We conclude that the addition of Pt increases the
number of metal sites and slightly improves the low temperature
activity of the catalyst because of the larger M/MOx ratio. Overall,
platinum provides metal sites, whereas palladium is responsible
for both metal and metal oxide sites, and the number of PdOx sites
increases noticeably during aging.
Fig. 10 shows correlation between TOR and the M/MOx ratio.
As is evident, the active form of the Al2O3-supported Pd–PdOx cat-
alyst contains both metal and metal oxide phases, and good low
temperature methane combustion activity is achieved with proper
M/MOx ratio. The results show as well that combustion activity
decreases with too high M/MOx ratio. Such activity behavior can be
explained by the lower activity of metallic Pd and Pt than of MOx
in low temperature methane combustion [43,49]. The M/MOx ratio

26
N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
Fig. 7. NH3 TPD curves of the fresh and aged monometallic Pd(24 h) and bimetallic Pd0.92–Pt0.08 and Pd0.85–Pt0.15 catalysts.
Fig. 8. NH3 TPD curves of fresh and aged monometallic Pd(24 h), Pt(2.3 wt.%), and Pt(4.1 wt.%) catalysts and the bimetallic Pd0.37–Pt0.63 catalyst.
Fig. 9. XPS spectra and oxidation states of fresh and aged monometallic Pd(24 h) catalysts.

N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
27
Table 3
Methane light-off temperatures together with methane conversions, methane conversion rates (MCRs), and turnover rates (TORs) at 573 K for the catalysts.
Catalyst
Light-offtemperature(K)
Light-off difference (K)
Methaneconversion(%)
MCR (␮mol gM−1 s−1
)
TOR (×10−2 s−1
)
Fresh
Aged
Fresh
Aged
Fresh
Aged
Fresh
Aged
Pd(24 h)
Pd(48 h)
Pd(72 h)
548
552
555
547
549
556
559
617
n.d.a
740
610
622
633
614
620
613
598
773
n.d.a
n.d.a
62
70
78
67
71
57
39
156
–
96.0
79.0
72.3
94.1
89.1
71.6
66.6
12.0
1.2
6.6
5.9
6.0
5.5
5.8
8.9
11.7
1.8
0.7
0.7
566
465
426
555
525
368
301
53
39
35
35
32
34
46
53
8
23.5
18.7
19.3
23.0
20.9
14.7
14.2
2.6
3.5
4.0
4.8
4.0
3.6
4.1
6.4
1.0
1.0
1.9
Pd(25 cm3)
Pd(35 cm3)
Pd0.92–Pt0.08
Pd0.85–Pt0.15
Pd0.63–Pt0.37
Pt(2.3 wt.%)
Pt(4. 1 wt.%)
7
7
4
2
0.3
0.3
–
2.1
a
No methane light-off was observed at measurement temperatures between 298 K and 773 K.
Table 4
Integrals and calculated M/MOx ratios of the fresh and aged catalysts obtained by NH3 desorption and XPS methods.
Catalyst
Fresh catalysts
M
Aged catalysts
M
MOx
M/MOx ratio
MOx
M/MOx ratio
Pd(24 h)
Pd(48 h)
2.08 (0.20)a
1.53
0.89 (0.05)a
0.74
2.34 (4.00)a
2.07
1.59 (0.11)a
1.32
1.40 (0.05)a
1.58
1.13 (2.20)a
0.83
Pd(72 h)
1.58
0.47
3.37
1.42
1.76
0.81
Pd(25 cm3)
Pd(35 cm3)
Pd0.92–Pt0.08
Pd0.85–Pt0.15
Pd0.37–Pt0.63
Pt(2.3 wt.%)
Pt(4.1 wt.%)
1.99
1.63
3.12
3.58
1.64
0.83
1.74
0.92
0.58
1.58
0.94
2.16
2.81
1.97
3.83
1.22
1.23
1.28
2.09
1.29
0.51
0.89
1.30
1.67
1.12
0.92
0.74
1.15
1.40
1.49
c
0.33
4.95
n.d.b
–
–
–
n.d.b
–
n.d.b
c
c
n.d.b
–
n.d.b
c
c
a
The at.% values, presented in parentheses, were obtained by deconvolution from XPS spectra presented in Fig. 9.
No corresponding peak was detected.
Not calculated, since data was incomplete.
b
c
•
The active phase of the catalyst for low temperature CH4 com-
bustion contains both M and MOx forms.
•
•
•
The preparation parameters of the monometallic base catalyst
influence the M/MOx ratio, but no clear trends were observed.
The platinum promoter prefers metallic form, while palladium
exists in both metallic and oxide forms.
The M/MOx ratio decreases in aging due to the formation of sta-
ble PdO, but the addition of Pt improves the ratio by increasing
the number of M sites and thus enhances the low temperature
activity.
•
Re-oxidation of palladium is prevented at high platinum pro-
portions, and this is reflected in a loss of low temperature CH4
combustion activity.
Acknowledgment
Ecocat Oy is thanked for supplying the alumina support material.
Fig. 10. Turnover rates (TORs) as a function of M/MOx ratio. The results shown as
gray squares are included for comparison [21].
References
[1] A.K. Neyestanaki, F. Klingstedt, T. Salmi, D.Y. Murzin, Fuel 83 (2004) 395–408.
[2] R.E. Hayes, Chem. Eng. Sci. 59 (2004) 4073–4080.
[3] G. Centi, J. Mol. Catal. A 173 (2001) 287–312.
of the non-promoted monometallic catalyst needs to be known in
order that the proper proportion of promoter can be added.
[4] L.A. Graham, G. Rideout, D. Rosenblatt, J. Hendren, Atmos. Environ. 42 (2008)
4665–4681.
[5] J. Kasˇpar, P. Fornasiero, N. Hickey, Catal. Today 77 (2003) 419–449.
[6] R. Di Monte, J. Kasˇpar, Top. Catal. 1–4 (2004) 47–57.
[7] V.R. Choudhary, V.P. Patil, P. Jana, B.S. Uphade, Appl. Catal. A 350 (2008)
186–190.
4. Conclusions
[8] R.J.H. Grisel, P.J. Kooyman, B.E. Nieuwenhuys, J. Catal. 191 (2000) 430–437.
[9] J.H. Lee, D.L. Trimm, Fuel Process. Technol. 42 (1995) 339–359.
[10] P. Gélin, L. Urfels, M. Primet, E. Tena, Catal. Today 83 (2003) 45–57.
[11] J.K. Lampert, M.S. Kazi, R.J. Farrauto, Appl. Catal. B 14 (1997) 211–233.
[12] Y. Yazawa, H. Yoshida, S.-i. Komai, T. Hattori, Appl. Catal. A 233 (2002) 113–124.
[13] B. Yue, R. Zhou, Y. Wang, X. Zheng, Appl. Catal. A 295 (2005) 31–39.
[14] T. Kolli, K. Rahkamaa-Tolonen, U. Lassi, A. Savimäki, R.L. Keiski, Catal. Today
100 (2005) 303–307.
The effects of preparation parameters and Pt promoter on
monometallic Pd–PdOx catalysts were studied. Aging durability of
the catalyst was successfully improved by adding the promoter. In
general, the activity was related to the proportions of metal (M) and
metal oxide (MOx) surface sites, allowing the following conclusions
to be drawn:

28
N.M. Kinnunen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 20–28
[15] B. Ersoy, V. Gunay, Ceram. Int. 30 (2004) 163–170.
[16] O.V. Mokhnachuk, S.O. Soloviev, A.Y. Kapran, Catal. Today 119 (2007) 145–151.
[17] Y. Ozawa, Y. Tochihara, A. Watanabe, M. Nagai, S. Omi, Appl. Catal. A 258 (2004)
261–267.
[18] K. Narui, H. Yata, K. Furuta, A. Nishida, Y. Kohtoku, T. Matsuzaki, Appl. Catal. A
179 (1999) 165–173.
[19] H. Yamamoto, H. Uchida, Catal. Today 45 (1998) 147–151.
[20] N.M. Kinnunen, M. Suvanto, M.A. Moreno, A. Savimäki, K. Kallinen, T.-J.J. Kin-
nunen, T.A. Pakkanen, Appl. Catal. A 307 (2009) 78–87.
[21] N.M. Kinnunen, J.T. Hirvi, T. Venäläinen, M. Suvanto, T.A. Pakkanen, Appl. Catal.
A 397 (2011) 54–61.
[22] P. Castellazzi, G. Groppi, P. Forzatti, Appl. Catal. B 95 (2010) 303–311.
[23] M. Rassoul, F. Gaillard, E. Garbowski, M. Primet, J. Catal. 203 (2001) 232–241.
[24] E. Álvarez, G. Vázquez, M. Sánchez-Vilas, B. Sanjurjo, J.M. Navaza, J. Chem. Eng.
Data 42 (1997) 957–960.
[33] R.E. Dinnebier, S.J.L. Billinge (Eds.), Powder Diffraction: Theory and Practice,
RSC Publishing, Cambridge, UK, 2008.
[34] A. Piras, A. Trovarelli, G. Dolcetti, Appl. Catal. B 28 (2000) L77–L81.
[35] U. Lassi, R. Polvinen, S. Suhonen, K. Kallinen, A. Savimäki, M. Härkönen, M.
Valden, R.L. Keiski, Appl. Catal. A 263 (2004) 241–248.
[36] G. Zhu, K.-i. Fujimoto, D.Y. Zemlyanov, A.K. Datye, F.H. Ribeiro, J. Catal. 225
(2004) 170–178.
[37] K.-i. Fujimoto, F.H. Ribeiro, M. Avalos-Borja, E. Iglesia, J. Catal. 179 (1998)
431–442.
[38] R. Zhou, B. Zhao, B. Yue, Appl. Surf. Sci. 254 (2008) 4701–4707.
[39] B. Zhao, C. Yang, Q. Wang, G. Li, R. Zhou, J. Alloys Compd. 494 (2010) 340–346.
[40] K. Persson, P.O. Thevenin, K. Jansson, J. Agrell, S.G. Järås, L.J. Pettersson, Appl.
Catal. A 249 (2003) 165–174.
[41] G. Lapisardi, L. Urfels, P. Gélin, M. Primet, A. Kaddouri, E. Garbowski, S. Toppi,
E. Tena, Catal. Today 117 (2006) 564–568.
[25] Bruker AXS, TOPAS V2.0: General Profile Analysis Software for Powder Diffrac-
tion Data, in: User Manual, Bruker AXS, Karlsruhe, Germany, 2000.
[26] GRAMS/32, v4.0, Galactic Industries Corporation.
[42] P. Gélin, M. Primet, Appl. Catal. B 39 (2002) 1–37.
[43] G. Groppi, C. Cristiani, L. Lietti, P. Forzatti, Stud. Surf. Sci. Catal. 130 (2000)
3801–3806.
[27] S.D. Jackson, L.A. Shaw, Appl. Catal. A 134 (1996) 91–99.
[28] C.A. Müller, M. Maciejewski, R.A. Koeppel, A. Baiker, Catal. Today 47 (1999)
245–252.
[29] G. Fagherazzi, P. Canton, P. Riello, N. Pernicone, F. Pinna, M. Battagliarin, Lang-
muir 16 (2000) 4539–4546.
[44] D. Wickham, R. Cook, United States Patent Application Publication, Pub. No.:
US 2003/0,176,278 A1.
[45] A.K. Datye, J. Bravo, T.R. Nelson, P. Atanasova, M. Lyubovsky, L. Pfefferle, Appl.
Catal. A 198 (2000) 179–196.
[46] S. Colussi, A. Trovarelli, G. Groppi, J. Llorca, Catal. Commun. 8 (2007) 1263–1266.
[30] P. Canton, G. Fagherazzi, M. Battagliarin, F. Menegazzo, F. Pinna, N. Pernicone,
Langmuir 18 (2002) 6530–6535.
[47] D. Gao, C. Zhang, S. Wang, Z. Yuan, S. Wang, Catal. Commun.
2583–2587.
9 (2008)
[31] M.J. Koponen, T. Venäläinen, M. Suvanto, K. Kallinen, T.-J.J. Kinnunen, M. Härkö-
nen, T.A. Pakkanen, J. Mol. Catal. A 258 (2006) 246–250.
[32] A. Corma, Chem. Rev. 95 (1995) 559–614.
[48] Y. Yazawa, H. Yoshida, N. Takagi, S.-i. Komai, A. Satsuma, T. Hattori, Appl. Catal.
B 19 (1998) 261–266.
[49] R. Burch, F.J. Urbano, Appl. Catal. A 124 (1995) 121–138.