Low-temperature oxidation of carbon monoxide on Pd(Pt)/CeO2 catalysts prepared from complex salts

Article abstract of DOI:10.1134/S0023158411020182

Catalysts containing cerium oxide as a support and platinum and palladium as active components for the low-temperature oxidation of carbon monoxide were studied. The catalysts were synthesized in accordance with original procedures with the use of palladium and platinum complex salts. Regardless of preparation procedure, the samples prepared with the use of only platinum precursors did not exhibit activity at a low temperature because only metal and oxide (PtO, PtO₂) nanoparticles were formed on the surface of CeO₂. Unlike platinum, palladium can be dispersed on the surface of CeO₂ to a maximum extent up to an almost an ionic (atomic) state, and it forms mixed surface phases with cerium oxide. In a mixed palladium-platinum catalyst, the ability of platinum to undergo dispersion under the action of palladium also increased; as a result, a combined surface phase with the formula Pd _x Pt _y CeO_{2 - δ}, which exhibits catalytic activity at low temperatures, was formed.

Full text of DOI:10.1134/S0023158411020182

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2011, Vol. 52, No. 2, pp. 282–295. © Pleiades Publishing, Ltd., 2011.
Original Russian Text © E.M. Slavinskaya, R.V. Gulyaev, O.A. Stonkus, A.V. Zadesenets, P.E. Plyusnin, Yu.V. Shubin, S.V. Korenev, A.S. Ivanova, V.I. Zaikovskii, I.G. Danilova,
A.I. Boronin, 2011, published in Kinetika i Kataliz, 2011, Vol. 52, No. 2, pp. 291–304.
LowꢀTemperature Oxidation of Carbon Monoxide
on Pd(Pt)/CeO Catalysts Prepared from Complex Salts
2
a
a
a,c
b,c
b,c
E. M. Slavinskaya , R. V. Gulyaev , O. A. Stonkus , A. V. Zadesenets , P. E. Plyusnin ,
b,c
b,c
a
a
a
a,c
Yu. V. Shubin , S. V. Korenev , A. S. Ivanova , V. I. Zaikovskii , I. G. Danilova , and A. I. Boronin
a
Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia
b
Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia
c
Novosibirsk State University, Novosibirsk, 630090 Russia
eꢀmail: lesla@catalysis.ru
Received July 2, 2010
Abstract—Catalysts containing cerium oxide as a support and platinum and palladium as active components
for the lowꢀtemperature oxidation of carbon monoxide were studied. The catalysts were synthesized in accorꢀ
dance with original procedures with the use of palladium and platinum complex salts. Regardless of preparaꢀ
tion procedure, the samples prepared with the use of only platinum precursors did not exhibit activity at a low
temperature because only metal and oxide (PtO, PtO ) nanoparticles were formed on the surface of CeO .
2
2
Unlike platinum, palladium can be dispersed on the surface of CeO to a maximum extent up to an almost an
2
ionic (atomic) state, and it forms mixed surface phases with cerium oxide. In a mixed palladium–platinum
catalyst, the ability of platinum to undergo dispersion under the action of palladium also increased; as a result,
a combined surface phase with the formula Pd Pt CeO , which exhibits catalytic activity at low temperaꢀ
x
y
2 –
δ
tures, was formed.
DOI: 10.1134/S0023158411020182
INTRODUCTION
ters at which lowꢀtemperature catalytic oxidation
takes place. The Pd/CeO catalysts are superior to
2
The removal of harmful gases, primarily, carbon
monoxide and hydrocarbons, from the atmosphere is
a problem of considerable current importance. The
catalytic neutralization of harmful substances by comꢀ
plete oxidation can be an efficient way for solving this
environmental problem. Catalytic oxidation at low
temperatures is most difficult to perform. The lowꢀ
temperature oxidation of CO and hydrocarbons is
required in the neutralization of exhaust emissions in
the cold start process, the purification of indoor atmoꢀ
Pt/CeO in terms of activity in CO oxidation. This is
2
likely due to different degrees of chemical interactions
between the above metals or metal oxides and the surꢀ
face of cerium oxide. The efficiency of this interaction
mainly depends on the degree of dispersion of an
active component and the temperature of catalyst calꢀ
cination. As a rule, an increase in the treatment temꢀ
perature of catalysts based on noble metals is accomꢀ
panied by a decrease in the catalyst activity. It is
believed that a possible reason for this phenomenon
consists in changes in the size of metal particles and
spheres, the operation of lasers based on CO , and
2
other processes occurring at low temperatures [1].
Heterogeneous metal–oxide systems, in which the
metal and the oxide exhibit a synergistic effect, belong
the particleꢀsize distribution [8]. Thus, in Pd/Al O
2
3
samples calcined at 350°С, the average Pd particle size
to the most promising lowꢀtemperature catalysts for was 2.4
±
0.4 nm and the particleꢀsize distribution had
CO oxidation. The constituents of these systems are a unimodal character. As the treatment temperature
platinum group metals or IB Group metals (Cu, Ag, was increased, the distribution became bimodal and
and Au) in a highly dispersed state supported on tranꢀ the average particle size increased. Bimetallic catalysts
sition element oxides [2].
whose active component consists of two metals are
considered more stable. In particular, Yang et al. [9]
It is well known that highly dispersed palladium or
platinum supported on cerium oxide and/or ceriumꢀ
containing compositions exhibit considerable cataꢀ
lytic activity and stability [3, 4]. The results of recent
studies on these catalysts [5–7] made it possible to
studied Pd1 – Au , Pd Cu , Au xCux, and
x
x
1 –
x
x
1 –
Au1 ⎯ xAg compositions prepared by the laser evaporaꢀ
x
tion of precursors under controlled conditions followed
by the condensation of particles and demonstrated that
conclude that the synergistic effect of a metal–oxide nanosized bimetallic compositions are promising for
system is due to the chemical interaction of metal and use in CO oxidation. The temperature ( 50) at which a
oxide particles and the formation of new phases on the 50% conversion of CO was reached increased in the
surface. These new surface phases contain active cenꢀ order Cu Pd 120 С) < Cu Pd 135 С) <
Т
(
∼
°
(
∼
°
50
50
12
88
282

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
283
Cu Au (
∼
145°С) < Pd Au ( 190°С). The activity Examples of the preparation of nanoparticles in binary
∼
9
0
10
88
12
of bimetallic compositions, including Pd–Pt catalysts and monometallic catalytic systems based on noble
with different ratios between the noble metals, can and other transition metals are well known [27–31].
increase upon supporting the metals onto an appropriꢀ The described procedure makes it possible to form a
ate carrier.
metal phase from a supported precursor at a low temꢀ
The activity of supported metals depends on a perature. The intensity of the surface diffusion of
number of factors including the average particle size, nanoparticles formed under these conditions noticeꢀ
the nature of metal–support interactions, the nature ably decreases to facilitate the retention of the metal
of used precursors, and the preparation procedure. phase in a highly dispersed state.
Various oxides or mixed oxide compositions are used
as supports. Among them, oxides that possess redox
In this work, we purposefully studied the effect of
the activation procedure of precursor compounds on
the structural and electronic states of Pd and Pt and on
properties, for example, CeO , are the most interestꢀ
2
ing species. As found previously [10–17], Pd and Pt
the activity of Pd(Pt)/CeO catalysts in the reaction of
2
can interact with CeO to form a Ce1 – M O
(M =
2
x
x
2 –
δ
CO oxidation and compared the monoꢀ and bimetalꢀ
lic catalysts. For this purpose, we used a set of physicꢀ
ochemical and kinetic techniques (see below).
Pd and Pt) solid solution, which changes the catalytic
activity, as compared with that of monometallic cataꢀ
lysts.
The most widespread initial noble metal comꢀ
pounds whose solutions are used for impregnating a
support are H PtCl [18, 19], Pd(NO ) [20], and
EXPERIMENTAL
2
6
3 2
Catalyst Preparation Methods
Pd(CH COO) [21]. However, the use of a chlorineꢀ
3
2
containing precursor of platinum seems unreasonable
because chloride ions poison the surface of cerium
oxide. The formation of cerium oxo chloride
Cerium dioxide prepared by precipitation from a
solution of cerium nitrate with an aqueous ammonia
solution at certain pH and temperature followed by filꢀ
tration and washing the precipitate with distilled water
until the absence of nitrates was used as a support. The
resulting sample was dried initially in air and then in a
3+
(Ce OCl) prevents the stabilization of supported
metals in an active state; because of this, the activity of
such catalysts is low [22]. Methods for the formation
of nanosized noble metal particles by the decomposiꢀ
tion of Pdꢀcontaining precursors of another type,
namely, Pd(Mes)2, where Mes is mesitylene (1,3,5ꢀtriꢀ
methylbenzene), in a trioctylphosphine solution or
Pd(CH COO) in an octylamine solution were develꢀ
drying oven at 110
calcined in a flow of dried air at 450
specific surface area of the support was 113 m /g.
°
С
for 12–14 h; thereafter, it was
for 4 h. The
°С
2
Two procedures were used for supporting an active
component.
3
2
oped [8]. The process is performed at 300°С followed
by the steps of centrifugation and washing with methaꢀ
Procedure I. The support was incipient wetness
nol. The preparation of supported catalysts with the use impregnated with a solution of a complex salt with a
of coordination compounds as precursors makes it posꢀ given concentration followed by drying in air for 24 h.
sible to eliminate many disadvantages characteristic of To prepare monometallic catalysts, aqueous solutions
traditional procedures. This area has been intensively of a [Pd(NH ) ](NO ) or [Pt(NH ) ](NO ) complex
3 4 3 2 3 4 3 2
developed in our laboratory in recent years [23
In this work, we used isostructural complex salts of
Pt(II) and Pd(II)—the tetraammine complex nitrates
⎯
25].
salt were used. The Pd–Pt binary active component
with the molar ratio Pd : Pt = 2 : 1 in the resulting solid
solution was supported by impregnating the support
with an aqueous solution containing both of the comꢀ
[
M(NH ) ](NO3)2, where M = Pd or Pt—as precurꢀ
3 4
plex salts, [Pd(NH ) ](NO ) and [Pt(NH ) ](NO ) ,
sor compounds for supporting metals onto CeO2.
These compounds are readily soluble in water to facilꢀ
itate the impregnation of a support. Note that a solid
3 4
3 2
3 4
3 2
in a ratio of 1 : 2.
The starting complex salts were prepared by the
solution rather than a mixture of individual complexes reactions
is formed on the support surface upon the simultaꢀ
neous crystallization of Pd and Pt compounds. Thus,
platinum and palladium become mixed at an atomic
level even at the step of supporting a precursor comꢀ
pound. Upon the thermal decomposition of these
solid solutions, bimetallic alloy particles in a high
HNO (conc)
3
Pd
[Pd(H O) (NO ) ]
2 2 3 2
NH (conc)
3
[
Pd(NH ) ](NO ) ,
3 4 3 2
2AgNO3
[
Pt(NH ) ]Cl
[Pt(NH ) ](NO ) .
3 4 3 2
3
4
2
degree of dispersion remain on the surface.
Along with thermal decomposition, we additionꢀ
Procedure II. The support was impregnated with a
ally tested a procedure of the reduction of supported complex salt solution as described in procedure I;
complex solid solutions by hydrazine hydrate (N H4 however, after the step of drying, the sample was
H O) at room temperature. Because hydrazine has a reduced with hydrazine hydrate (N H4 H O) at room
·
2
·
2
2
2
high reduction potential, this reagent is widely used in temperature; thereafter, the samples were washed with
various methods for the reduction of metals [26]. water on boiling and dried in air at 130–150
°C.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

284
SLAVINSKAYA et al.
Table 1. Consumption of hydrogen in the course of the temperatureꢀprogrammed reduction of Pd, Pt, and PdPt catalysts supꢀ
2
ported on CeO with hydrogen
2
Supported metal content
Consumption of H , μmol/g
H /(Pd+Pt)
2
2
Catalyst
wt %
μmol/g
total
at the 53
°
С peak
total
at the 53 С peak
°
Pd/CeO ꢀI
2
0.98
0.97
92.1
91.2
609.3
540.7
899.1
443.0
781.7
401.8
–
6.6
5.9
8.6
2.9
3.3
2.0
–
Pd/CeO ꢀII
2
99.2
–
1.1
–
Pt/CeO ꢀI
2
2.03
104.1
Pt/CeO ꢀII
2
2.95
151.2
–
–
PdPt /CeO ꢀI
2
1.1 + 2.54
0.98 + 2.13
103.4 + 130.2
92.1 + 109.2
–
–
2
PdPt /CeO ꢀII
2
–
–
2
The resulting samples were thermally treated in a trations of metal components in the samples were
flow of oxygen on heating at a rate of 10 K/min to determined from the 111 peak area of the fcc phase
4
50
°
C
and kept at this temperature for 0.5 h.
(
2 = 39.80°–40.15°) with normalization to the 002
θ
peak area of the support (CeO ). In the course of the
The catalysts prepared in accordance with the
2
quantitative phase analysis, we considered the ratio
between the corundum numbers of platinum and palꢀ
ladium (Icorr (Pt)/Icorr (Pd)) to be 1.49.
The catalytic properties of the synthesized samples
were tested using a temperatureꢀprogrammed reaction
above procedures are referred to as M/CeO ꢀX, where
2
M = Pd, Pt, or Pd–Pt and X denotes the procedure of
supporting an active component (I or II). Table 1 sumꢀ
marizes the concentrations of Pd (Pt) in the samples.
(
TPR) in an automated system with a flow reactor and
Investigation Techniques
massꢀspectrometric analysis of a gas mixture. A 0.38 g
Atomic absorption spectrometry (AAS) was used to sample with a particle size of 0.25–0.5 mm and a volꢀ
3
determine main component concentrations in the ume of 0.25 cm was placed in a stainless steel reactor.
samples [32].
A reaction mixture containing 0.2 vol % CO, 1.0 vol %
O , and 0.5 vol % Ne (the balance was helium) was
The samples were studied by highꢀresolution transꢀ
mission electron microscopy (HRTEM) on a JEMꢀ
2
3
supplied at a rate of 1000 cm /min to the initial cataꢀ
lyst cooled to –15
2
°
С
. The space velocity was
40000 h . In the reaction mixture, the catalyst was
heated from –15 to 450 at a rate of 10 K/min and
2010 microscope (JEOL, Japan).
–1
The electronic diffuse reflectance spectra of oxide
powders in the UV and visible regions were measured
°
С
then cooled in a flow of helium. In the course of the
–
1
in the spectral range of 11000–54000 cm in air with
no pretreatment using a UVꢀ2501 PC spectrometer
from Shimadzu with an IRSꢀ250A diffuse reflectance
attachment [33].
The Xꢀray photoelectron spectra (XPS) of the samꢀ
ples, which were ground in an agate mortar and
mechanically pressed into an iridium surface of a
holder, were obtained on a KRATOS ES 300 XPS
spectrometer from Kratos Analytical (United Kingꢀ
reaction, the concentrations of CO, О2, and CO were
2
determined. The concentrations were measured at a
frequency of 0.34 Hz.
The temperatureꢀprogrammed reduction with
hydrogen (TPRꢀH ) was performed in an automated
2
system with a flow reactor and massꢀspectrometric
analysis of a gas mixture. The initial sample was
treated in a flow of a gas mixture containing 20 vol %
O2 in He to remove water and carbonates from the surꢀ
dom). The
К line of aluminum (or magnesium) with
α
face. A reaction mixture containing 0.1 vol % H and
2
a photon energy of 1486.6 eV (1253.6 eV) was used as
primary radiation. The instrument was calibrated
based on the position of the Au 4
U''' line in the Ce 3 spectrum was used as a calibration
standard, the binding energy of which was taken equal
to 916.7 eV.
0
.5 vol % Ne (the balance was helium) was supplied at
3
a rate of 1000 cm /min to the initial catalyst (0.2 g)
cooled to –15°C. In the reaction mixture, the catalyst
f line (84.0 eV). The
d
was heated from –15 to 450°С at a rate of 10 K/min;
thereafter, it was cooled in a flow of helium.
The specific surface areas of the samples were meaꢀ
sured using the thermal desorption of argon [34]; the
RESULTS
HRTEM Data
error of this method was
Xꢀray diffraction (XRD) analysis was performed on
a DRONꢀSEIFERTꢀRM4 diffractometer with Cu
±
10%.
K
We failed to detect palladium particles in the elecꢀ
α
radiation. The crystallite size was determined by meaꢀ tron micrographs of sample Pd/CeO ꢀI (Fig. 1a).
2
suring the physical broadening of individual diffracꢀ Nevertheless, the EDX spectra measured in the same
tion profiles using the Scherrer formula. The concenꢀ support regions suggest the occurrence of a small
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
Intensity, arb. units
285
(a)
(
b)
Ce
L
1350
1200
1050
900
750
600
450
300
150
Pd
L
02
3
4
5
6
7
keV
5
nm
Intensity, arb. units
(
c)
(d)
Cu
K
7
6
5
4
4
3
2
1
470
640 Pt M
810
980
150
320
Ce L
Pt
M
490
660
30
Fe
K
8
04
.80
3.52 5.24 6.96 8.68
keV
50 nm
Intensity, arb. units
00
10
20
30
40
50
60 Pt
(e)
(f)
9
8
7
6
5
4
3
2
1
Ce
L
M
Pd
L
70
80
9
0
10 nm
0 1
.40
2.72 4.04 5.36 6.60
keV
Fig. 1. (a) HRTEM image of a CeO particle in sample Pd/CeO ꢀI; (b) EDX spectrum, which suggests the presence of palladium
2
2
in CeO particles shown in Fig. 1a; (c) micrograph of PtOx particle aggregates in catalyst Pt/CeO ꢀI; (d) EDX spectrum of the
2
2
region marked in Fig. 1c; (e) HRTEM image of the support in catalyst PdPt /CeO ꢀI; and (f) EDX spectrum of support particles
2
2
shown in Fig. 1e.
amount of Pd (Fig. 1b). The atomic concentrations of support regions, and they were absent from other
palladium calculated from the EDX spectra measured regions; this suggests a nonuniform distribution of an
in various support regions were always close to the active component over the support surface. We failed
value specified in the catalyst preparation; this fact to detect palladium and platinum particles in the elecꢀ
suggests a uniform distribution of Pd on cerium dioxꢀ tron micrograph of PdPt /CeO2ꢀI (Fig. 1e). In this
2
ide. The micrograph of catalyst Pt/CeO ꢀI exhibited case, the calculation based on Ce
L
, Pd
platinum oxide (PtOx) particle aggregates of size 100– lines of the EDX spectrum (Fig. 1f) demonstrated that
00 nm (Fig. 1c); in this case, the size of particles that the atomic ratio between Pd and Pt was close to the
L, and Pt L
2
2
formed these aggregates was 10–30 nm. Note that the specified value. Consequently, both palladium and
aggregates of PtOx particles were detected in some platinum were uniformly distributed on the support in
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

286
SLAVINSKAYA et al.
Intensity, arb. units
(a)
(b)
7
6
5
4
4
3
2
1
38
56
74
92
10
28
46
64
Pt
M
Cu
L
8
2
Ce L
0 0
.50 0.99 1.48 1.97 2.46 2.95 3.44 3.93 4.42 4.9
keV
20 nm
Intensity, arb. units
(d)
(c)
1
100
990
880
770
660
550
440
330
220
110
Pt
M
Pd
L
Ce L
0 1
.1
1.5 1.9 2.3 2.7 3.1 3.5 3.9 4.9 4.7
keV
100 nm
Fig. 2. (a) HRTEM image of a Pt metal particle in sample Pt/CeO ꢀII; (b) EDX spectrum of the particle shown in Fig. 2a;
2
(
c) HRTEM image of a bimetallic particle in catalyst PdPt /CeO ꢀII; and (d) EDX spectra of two different particles of sample
2 2
PdPt /CeO ꢀII, which suggest the inhomogeneous composition of alloy particles (the spectra of these particles differed in color
2
2
intensity).
sample PdPt /CeO ꢀI. Figures 2a and 2b show the
Electronic Diffuse Reflectance Spectroscopic Data
2
2
HRTEM images of catalyst Pt/CeO ꢀII and the EDX
2
The electronic diffuse reflectance spectrum of a
spectrum of the marked region, respectively. It can be
seen that platinum in this catalyst occurred as coarse
metal particles of sizes from 50 to 200 nm. Coarse
metal particles of size 100–300 nm were observed in
PdPt /CeO ꢀII. One of these particles is shown in an
CeO sample (Fig. 3a, curve 1) exhibited two intense
2
–1
absorption bands at 36000 and 29000 cm , which
were attributed to Ce
4+
2–
O
charge transfer (CT)
bands and a interband transition in CeO , respectively
2
–1
2
2
[
35]. An additional absorption band at 22000 cm
HRTEM image (Fig. 2c). Figure 2d shows the EDX
spectra of two different particles of size about 200 nm
appeared in the spectrum of catalyst Pd/CeO ꢀI
2
(
Fig. 3a, curve
2); this band can be attributed to the
normalized to the Pt
M
line. It can be seen that the
2+
d⎯d
transition of Pd ions in an oxygen environment
Pd/Pt ratios in these particles were different; this fact
suggests an inhomogeneous composition of alloy parꢀ
ticles.
in D4h symmetry, which is characteristic of small PdO
clusters [36]. Considerable changes were observed in
the spectrum of catalyst Pd/CeO ꢀII, which was preꢀ
2
pared with the use of hydrazine (Fig. 3a, curve 3).
Thus, an analysis of the results of electronꢀmicroꢀ
scopic studies demonstrated that ultrafine palladium
particles and coarse aggregates of platinum oxide parꢀ
ticles were formed in monometallic catalysts upon
supporting the noble metals from complex salts. The
combined supporting of palladium and platinum in a
certain ratio increased the degree of dispersion of platꢀ
inum in bimetallic catalysts. Upon reduction with
hydrazine, coarse Pd and Pt particles were present on
the support surface in monometallic catalysts, whereas
Stronger background absorption in the longꢀwaveꢀ
length region was characteristic of reduced cerium
oxides, and the appearance of a new broad absorption
–1
band in the range of 14500–16000 cm can be
4+
Се3+ interꢀ
ascribed to a CT band due to the Ce
valent transition [37]. Thus, we can conclude that the
treatment of a sample with hydrazine facilitated the
reduction of cerium oxide.
In addition to the absorption bands of cerium
oxide, the electronic diffuse reflectance spectrum of
Pd–Pt alloy particles with various Pd/Pt ratios Pt/CeO ꢀI (Fig. 3b, curve
2) exhibited intense absorpꢀ
2
–1
occurred in bimetallic catalysts.
tion bands at 16500 and 21000 cm , and the specꢀ
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
287
trum of Pt/CeO ꢀII exhibited an absorption band at
2
(a)
–1
2
[
1000 cm (Fig. 3b, curve
38], the most intense absorption band corresponding
to the d–d transitions of Pt(IV) in the bulk oxide PtO2
3). According to Lietz et al.
·
8
4
0
–1
H O was observed at 23500 cm , and it was shifted to
2
shorter wavelength in the spectrum of the supported
oxide PtO2. The appearance of intense absorption
bands in the lowꢀfrequency region, which were
ascribed to spin forbidden d–d transitions, is characꢀ
teristic of Pt(II) compounds, for example, squareꢀplaꢀ
nar halide complexes [39]. Thus, based on the results
of analysis performed by electronic diffuse reflectance
spectroscopy, we can hypothesize that Pt(II) oxide was
predominant in samples Pt/CeO ꢀII and Pt/CeO ꢀI.
3
2
1
2
2
1
5000 20000 25000 30000 35000
The spectrum of PdPt /CeO ꢀI (Fig. 3c, curve 1)
2
2
(b)
was close to the spectrum of Pt/CeO ꢀI in terms of
2
absorption band positions. It contained broad absorpꢀ
tion bands with maximums at 16500, 21000, and
8
6
4
2
0
9
–1
2
5000 cm in addition to the absorption bands of
cerium dioxide. The electronic diffuse reflectance
spectroscopic data suggest that Pt(II) oxide was preꢀ
dominant in the bimetallic sample. Based on the
results of analysis performed using this technique, it is
difficult to identify the oxidized states of palladium
because the spectra of platinum oxide compounds
superimposed on the spectra of palladium oxide comꢀ
pounds. In the spectrum of PdPt /CeO ꢀII (Fig. 3c,
2
2
2
3
1
curve
2), an increase in the background absorption in
the longꢀwavelength region and the appearance of an
15000 20000 25000 30000 35000 40000
(c)
–1
absorption band in the region of 14500–16000 cm ,
3+
which is indicative of the appearance of Ce ions,
were observed. Thus, the treatment of both
PdPt /CeO ꢀII and Pd/CeO ꢀII with hydrazine facilꢀ
2
1
2
2
2
itated the reduction of the support.
6
3
0
XPS Data
Figure 4 shows the Pd 3d and Pt 4f XPS spectra of
all of the catalysts. It can be seen that, in Pd/CeO ꢀI
2
Fig. 4a), palladium is mainly in the Pd2 state with
+
(
BE(Pd 3d5/2) = 338.2 eV. This peak cannot be attribꢀ
uted to PdO nanoparticles because BE(Pd 3 5/2) in the
structure of PdO is 337.0 0.2 eV [40–42]. According
d
1
5000 20000 25000 30000 35000 40000
−
Wave number, cm
±
1
to published data [10, 12–14, 43, 44], this peak charꢀ
acterizes the ionic state of palladium as a constituent
of a phase obtained by the interaction of palladium
Fig. 3. Electronic diffuse reflectance spectra of (a) (
CeO , ( ) Pd/CeO ꢀI, and ( ) Pd/CeO ꢀII; (b) ( ) CeO2,
) Pt/CeO ꢀII; and (c) (1)
1)
2
3
1
2
2
2
(
2
) Pt/CeO ꢀI, and (
3
with the surface of CeO ; the composition of this
2
2
2
PdPt /CeO ꢀI and (2) PdPt /CeO ꢀII.
phase can be formally described as Pd Ce xO2 –
.
δ
≈
2
2
2 2
x
1 –
In the Pd 3
d
spectrum, a peak at BE(Pd 3 5
d
/2)
3
36.7 eV can also be recognized. This peak was
The treatment of a sample with hydrazine was
accompanied by a dramatic growth of palladium metal
observed previously [43, 44], and it belongs to
extremely small palladium clusters consisting of a few
atoms, mainly as О–Pd –О dimers [45]. This high
value of BE was due to the participation of oxygen ions
from the support surface, which affected the pallaꢀ
dium clusters to convert them into an ionized state trum of an active component on a support is deterꢀ
particles in catalyst Pd/CeO ꢀII, which were not
2
2
detected by XPS (Fig. 4b). It is well known [46] that
the signal intensity of the core line in the XPS specꢀ
+
(
the Pd state in the case of dimers).
mined from the formula I ~ (Cac)/d Ssp, where I is the
ac
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

288
SLAVINSKAYA et al.
(a)
(b)
(c)
(d)
338.2
338.0
Pd 3d
Pd 3d
336.7
336.5
335
340
73.2
345
335
340
72.7
345
335
340
73.0
345
335 340 345
72.6
Pt 4
f
Pt 4
f
70.9
71.0
71.0
70 72 74 76 78 80
70 72 74 76 78 80
70 72 74 76 78 80
Binding energy, eV
70 72 74 76 78 80
Fig. 4. The (top) Pd 3
d
and (bottom) Pt 4f XPS spectra of catalysts (a) Pd/CeO ꢀI and Pt/CeO ꢀI; (b) Pd/CeO ꢀII and Pt/CeO ꢀ
2 2 2 2
II; and (c) PdPt /CeO ꢀI and PdPt /CeO ꢀII.
2
2
2
2
signal intensity of the core line that characterizes the BE(Pt 4f7/2) = 73.2 eV, which is characteristic of the
active component, Cac is the concentration of the platinum oxide PtO [40].
active component, dac is the particle diameter of the
active component, and Ssp is the specific surface area
of the support.
Catalyst Pt/CeO ꢀII, which was reduced with
2
hydrazine, exhibited an unusual behavior. Both divaꢀ
lent platinum with BE(Pt 4f7/2) = 72.7 eV and platiꢀ
num metal with BE(Pt 4f7/2) = 70.9 eV occurred on its
surface. In addition, an insignificant amount of tetꢀ
ravalent platinum with BE(Pt 4f7/2) = 74.2 eV was
present; this is consistent with electronic diffuse
reflectance spectroscopic data. The Pt/Ce ratios in
The observed decrease in the Pd 3d signal intensity
in the spectrum of sample Pd/CeO ꢀII can be
2
explained only by a decrease in the degree of disperꢀ
sion of palladium because all of the other parameters
remained unchanged. The decrease in degree of disꢀ
persion suggests the agglomeration of palladium with
the formation of coarse metal particles, as demonꢀ
strated below by HRTEM.
samples Pt/CeO ꢀI and Pt/CeO ꢀII were calculated
2
2
from the results of XPS analysis (Table 2). This ratio in
sample Pt/CeO ꢀI was 1.6% at a Pt content of 2.03 wt
2
%
, whereas it was 2.27% in sample Pt/CeO ꢀII at a Pt
2
An analysis of the Pt 4f spectra of Pt/CeO ꢀI indiꢀ
2
cated (Fig. 4a) that platinum occurred in a state with
content of 2.95 wt % or 2.32% as converted to a platiꢀ
num content of 2.03 wt %. These data suggest the same
degree of dispersion of platinum in catalysts Pt/CeO2ꢀ
II and Pt/CeO ꢀI.
2
Table 2. Ratios between metal atoms and cerium atoms in Pd,
In PdPt /CeO ꢀI (Fig. 4c), BE(Pd 3d5/2) =
Pt, and PdPt catalysts supported on CeO (according to XPS
2
2
2
2
data)
338.0 eV, and this suggests the strong interaction of
supported palladium with the support and the formaꢀ
tion of an interaction phase Pd Ce xO2 – . This conꢀ
Catalyst
Pd/Ce, at %
Pt/Ce, at %
x
1 –
δ
clusion is consistent with HRTEM data because no
extended palladium structures were detected on the
surface of the above catalyst. Nevertheless, an insignifꢀ
icant portion of palladium occurred as twoꢀdimenꢀ
Pd/CeO ꢀI
2
1.23
0
–
Pd/CeO ꢀII
2
–
Pt/CeO ꢀI
2
–
1.60
2.27
1.60
0.91
sional cluster structures (BE(Pd 3d3/2) = 336.5 eV),
which cannot be detected by HRTEM. The state of
platinum as a constituent of catalyst PdPt /CeO ꢀI
Pt/CeO ꢀII
2
–
2
2
2+
PdPt /CeO ꢀI
2
1.45
0
corresponds to bivalent Pt , and only an insignificant
2
portion of platinum was present as metal particles
(BE(Pt 4f7/2) = 71.0 eV). It is believed that the major
PdPt /CeO ꢀII
2
2
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
289
portion of platinum as a constituent of this catalyst sition of the most intense reflections of oxide phases
occurred as a solid solution rather than PtO particles; and the reflections of the support phase and by the
this solid solution was present in the subsurface layers weak intensity of these reflections because of low conꢀ
of CeO along with palladium cations. The Pt/Ce ratio centrations of active components. Another possible
2
in catalyst PdPt /CeO ꢀI was somewhat lower (1.6%) explanation can consist in the binding of oxidized
2
2
than that in Pt/CeO ꢀI (upon converting to platinum platinum and palladium in Pd(Pt) Ce xO2 – phases.
2
x
1 –
δ
content, the Pt/Ce ratio becomes 2.0%). Hence, it folꢀ
lows that the degree of dispersion of platinum someꢀ
what decreased at high metal contents.
TPRꢀH Data
2
The reduction of bimetallic catalyst PdPt /CeO ꢀ
The temperatureꢀprogrammed reduction of the
2
2
II with hydrazine (Fig. 4d) resulted in the formation of initial samples with hydrogen was performed at a low
coarse Pd metal particles similarly to that upon the concentration of H2 (0.1 vol %) and a high flow rate of
reduction of catalyst Pd/CeO ꢀII. In both cases, the a reaction mixture because the sensitivity of this
2
particles were large so that the signal from the Pd 3
line was not observed. Platinum is less prone to reducꢀ of hydrogen in the mixture [48].
tion than palladium, and it remained in an oxidized
According to the experimental results (Fig. 6a), one
state as PtO with BE(Pt 4 7/2) = 72.6 eV after the action peak with a maximum at a temperature of about 145
°С
d
method is inversely proportional to the concentration
f
of the reducing agent (N H ). In this case, PtO partiꢀ was observed on the reduction of Pd/CeO ꢀI. The
2
4
2
cles are sufficiently large, as evidenced by a low signal amount of consumed hydrogen was 609
intensity of the Pt 4
µ
mol/(g cataꢀ
f
line, a decrease in the Pt/Ce ratio, lyst) (Table 1). In this case, the ratio of the amount of
and HRTEM data, which are given in Fig. 2c.
consumed hydrogen to the amount of palladium
H /Pd) was 6.6 in the reduction of Pd/CeO ꢀI. The
(
2
2
TPRꢀH curve of Pt/CeO ꢀI exhibited two peaks
2
2
XRD Data
(
Fig. 6a) with maximums at 150 and 250
°С. The amount
According to XRD analysis data, the average crysꢀ of consumed hydrogen was 899
µ
mol/(g catalyst),
tallite size of the support phase was ~10 nm. The crysꢀ and the H /Pt ratio was 8.6. As can be seen, the
2
tal lattice parameters of CeO in the catalyst samples amount of consumed hydrogen was much greater than
2
were the same as the parameters of a polycrystalline the stoichiometric amount required for the reduction
reference sample (JCPDS, file no. 34ꢀ394) within the of PdO, PtO2, and PtO (H /PdO (or PtO) = 1;
2
limits of experimental error (
±
0.002 Å); this suggests a H /PtO = 2). The superstoichiometric consumption
2 2
3+
low concentration of reduced cerium (Ce ).
of hydrogen suggests the participation of oxygen from
According to XRD data, Pd/CeO ꢀI did not conꢀ cerium dioxide in the reduction [49]. Thus, the peak in
2
tain a metal phase, and about 80% platinum in the TPR spectrum of catalyst Pd/CeO ꢀI corresponds
2
Pt/CeO ꢀI occurred as metal particles with an average to the reduction of cerium oxide, and the presence of
2
crystallite size of 30 nm (Fig. 5a). Only an insignificant two peaks in the spectrum of Pt/CeO ꢀI can be
2
amount of a metal phase (5% on a total platinum and explained by the reduction of PtOx (Tmax = 150°C) and
palladium basis) was detected in PdPt /CeO ꢀI. Thus, cerium oxide (Tmax = 250
we can conclude that the total palladium and the
°
C
).
Another character of reduction was observed in the
major portion of platinum occurred in an oxidized study of PdPt /CeO ꢀI (Fig. 6a). The TPRꢀH specꢀ
2
2
2
2
2
state. The crystallite size in metal particles was 25 nm. trum of this sample exhibited only a single peak with a
According to Darby and Myles [47], who studied the maximum at 125
°С. Additionally, a lowꢀtemperature
dependence of the crystal lattice parameters of solid shoulder was observed in the main peak and the conꢀ
solutions on the composition of these solid solutions sumption of hydrogen came into play at a much lower
in the palladium–platinum system, the composition temperature of about
of the detected metal phase is Pd0.10Pt0.90
0
°
C
. Thus, the simultaneous
.
presence of palladium and platinum shifted the reducꢀ
In the samples of series II, the major portion of an tion temperature of cerium oxide to the lowꢀtemperaꢀ
introduced metal occurred in an unoxidized state ture region. The amount of consumed hydrogen was
(
Fig. 5b). In Pd/CeO ꢀII, 60% palladium occurred as 782 mol/(g catalyst) at the ratio H /PdPt = 3.3.
µ
2
2 2
the constituent of a metal phase. The crystallite size
The character of reduction of the samples reduced
was 16 nm. Platinum in sample Pt/CeO ꢀII almost with hydrazine was dramatically different (Fig. 6b).
2
completely occurred in metal particles with a crystalꢀ The TPRꢀH spectrum of Pd/CeO ꢀII exhibited a
2
2
lite size of 22 nm. The lattice parameter of a metal narrow peak with a maximum at 53
°С
and a broad
phase in PdPt /CeO ꢀII is 3.908 Å, which corresponds peak with a maximum at 170
°С. The occurrence of a
2
2
to solid solution Pd0.35Pt0.65 [47]. This phase contained narrow peak suggests a phase transition of oxide strucꢀ
about 50% platinum and palladium. The crystallite tures to a reduced state under the action of hydrogen.
size was 18 nm.
The amount of hydrogen consumed at this peak was
Attempts to detect and identify the reflections of 99.2 mol/(g catalyst), which corresponds to H /Pd =
µ
2
platinum and palladium oxide phases did not give posꢀ 1.1 (Table 1). This suggests that the palladium oxide
itive results. This fact can be explained by the superpoꢀ PdO was reduced in this temperature region. The total
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

290
SLAVINSKAYA et al.
a)
(
1
5000
0000
1
PdPt /CeO ꢀI
2
2
5000
Pd/CeO ꢀI
2
Pt/CeO ꢀI
2
0
30
40
50
2 , deg
θ
(b)
1
5000
0000
1
PdPt /CeO ꢀII
2
2
5000
Pd/CeO ꢀII
2
Pt/CeO ꢀII
2
0
30
40
50
2 , deg
θ
Fig. 5. XRD patterns of catalysts (a) Pd/CeO ꢀI, Pt/CeO ꢀI, and PdPt /CeO ꢀI and (b) Pd/CeO ꢀII, Pt/CeO ꢀII, and
2
2
2
2
2
2
PdPt /CeO ꢀII.
2
2
amount of absorbed hydrogen was as high as is higher than that required for the reduction of PtO2
40.7 mol/(g catalyst), which corresponds to and PtO. The occurrence of peaks at a low temperaꢀ
5
µ
H /Pd = 5.9. Hence, we can conclude that the oxygen ture suggests the reduction of oxidized platinum comꢀ
2
of cerium oxide also participated in the interaction pounds; this is consistent with XPS data, and an H /Pt
2
with hydrogen. According to XRD data, only about ratio higher than the stoichiometrically required one
6
0% Pd in the sample occurred as metal particles. suggests the participation of cerium oxide in the proꢀ
Hence, it follows that about 40% palladium was in the cess.
oxide state (PdO) and interacted with the support.
Because of this, the reduction occurred over a broad
temperature range, and a maximum was observed at a
sufficiently high temperature of 170 . In this case,
not only oxidized palladium but also the surface of
CeO participated in the reduction.
The reduction curve of PdPt /CeO ꢀII exhibited
2
2
two peaks at 14 and 135
.0. This suggests that Pt oxide particles, which were
reduced at 14 similarly to the reduction that
°С. In this case, H /PdPt =
2 2
2
°
С
°С
occurred in catalyst Pt/CeO ꢀII, were present in the
catalyst along with metal alloy particles, which were
2
2
The TPRꢀH curve of Pt/CeO ꢀII exhibited two detected by physicochemical techniques. In addition,
2
2
peaks with maximums at 32 and 90
hydrogen consumption was 443.0 mol/(g catalyst), acted with the support and activated oxygen from
°С. The total an insignificant portion of palladium strongly interꢀ
µ
which corresponds to the ratio H /Pt = 2.9. This ratio CeO , as evidenced by the occurrence of highꢀtemperꢀ
2
2
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
291
(a)
(b)
145°
С
53°
С
0.01
0.02
Pd/CeO ꢀI
2
250°
С
170°С
Pd/CeO ꢀII
2
Pt/CeO ꢀI
2
32°С
90°
С
125°С
14°
С
Pt/CeO ꢀII
2
135°С
PdPt /CeO ꢀI
2
PdPt /CeO ꢀII
2
2
2
0
0
50 100 150 200 250 300 350 400 450
c)
0
50 100 150 200 250 300 350 400 450
(d)
(
1
.0
.9
.8
.7
.6
.5
.4
.3
.2
.1
0
1.0
0
0
0
0
0
0
0
0
0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
3
2
1
3
1
2
0
50 100 150 200 250 300 350 400 450
Temperature
0
50 100 150 200 250 300 350 400450
,
°
C
Temperature, C
°
Fig. 6. The temperature dependence of hydrogen consumption by catalysts (a) Pd/CeO ꢀI, Pt/CeO ꢀI, and PdPt /CeO ꢀI and
2
2
2
2
(
b) Pd/CeO ꢀII, Pt/CeO ꢀII, and PdPt /CeO ꢀII in the course of TPRꢀH and the temperature dependence of CO conversion
2 2 2 2 2
on catalysts (c) (
PdPt /CeO ꢀII.
1) Pd/CeO ꢀI, (2) Pt/CeO ꢀI, and (3) PdPt /CeO ꢀI and (d) (1) Pd/CeO ꢀII, (2) Pt/CeO ꢀII, and (3)
2 2 2 2 2 2
2
2
ature peaks in the TPR spectrum and an increased their efficiency in the reaction of CO oxidation, but
value of H /(Pd + Pt).
this increase was insignificant. Thus, for Pt/CeO ꢀI
2
2
and Pt/CeO ꢀII, Т50 = 148 and 134°С, respectively.
2
Catalytic Properties of the Catalysts in CO Oxidation
Table 3. Catalytic activity of Pd, Pt, and PdPt catalysts supꢀ
2
ported on CeO in the reaction of CO oxidation
In Figs. 6c and 6d and Table 3, it can be seen that
the Pdꢀcontaining catalysts prepared by procedures I
and II exhibited different activities. The activity of
catalysts was evaluated from three temperatures at
which 10, 50, and 90% CO conversions were reached.
2
Catalyst
Т10
,
°
С
Т50
,
°
С
Т90, °С
Pd/CeO ꢀI
2
9
42
148
148
134
40
125
185
161
161
110
161
Pd/CeO ꢀII
2
110
119
68
5
They are referred to as Т10
,
Т50, and Т90, respectively;
the value of Т50 was most frequently considered. For
catalysts Pd/CeO ꢀI and Pd/CeO ꢀII, Т50 = 42 and
Pt/CeO ꢀI
2
2
2
Pt/CeO ꢀII
2
1
48°С, respectively. Thus, the treatment with hydraꢀ
zine dramatically decreased the efficiency of catalyst
PdPt /CeO ꢀI
2
2
Pd/CeO . On the contrary, the reduction of Ptꢀconꢀ
2
PdPt /CeO ꢀII
2
68
134
2
taining catalysts with hydrazine somewhat increased
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

292
SLAVINSKAYA et al.
The catalytic properties of bimetallic catalysts heat of CO adsorption on metal particles [51], which
depend on preparation procedures. Thus, the low is as high as 25–35 kcal/mol [52]; because of this, the
value of Т50 = 40
was characteristic of catalyst oxidation of CO cannot be performed at low temperaꢀ
tures.
°С
PdPt /CeO ꢀI. Intermediate reduction with hydrazine
In catalyst Pt/CeO ꢀI, PtOx platinum oxide partiꢀ
2
2
dramatically decreased its efficiency, so that Т50
2
increased to 134°С
. Thus, reduction with hydrazine cle aggregates of size 100–200 nm were present; in this
case, the size of particles in the aggregates was 10–
0 nm. These data are consistent with the results
impaired the performance characteristics of the Pdꢀ
3
and PdPt ꢀcontaining catalysts, whereas the efficiency
2
obtained by electronic diffuse reflectance spectrosꢀ
copy, which indicated the presence of Pt(IV) and
Pt(II) in an oxygen environment in the sample, and by
XPS, which exhibited the presence of a main line
characteristic of PtO and an additional line due to
PtO2 in the spectrum of this sample. In this case, from
XRD data, it follows that about 80% platinum
occurred as metal particles with an average crystallite
size of 30 nm. This contradiction can be explained by
the fact that coarse particles cannot be detected by
electronic diffuse reflectance spectroscopy and XPS,
whereas, on the contrary, XRD analysis cannot detect
platinum in a highly dispersed state. The TPRꢀH2
method demonstrated the presence of highly dispersed
platinum species in the catalyst, as evidenced by the
high temperature of reduction and the superstoichioꢀ
of the Ptꢀcontaining sample increased in this case,
although it did not reach the efficiency of the lowꢀ
temperature catalyst.
DISCUSSION
According to electronic diffuse reflectance specꢀ
troscopic and XPS data, palladium in catalyst
Pd/CeO ꢀI occurred as small clusters (O–Pd –
2
2
2+
O dimers) and isolated Pd ions as the constituents of
the interaction phase Pd Ce xO . The presence of
x
1 –
2 –
δ
only a single peak at 145
°С
and the ratio H /Pd = 6.6
2
in the TPRꢀH curve supports the occurrence of
2
chemical interaction between palladium and CeO .
2
2+
Thus, the presence of palladium as Pd ions as the metric ratio of consumed hydrogen to Pt (8.6). Conseꢀ
constituents of the interaction phase Pd Ce xO2 –
quently, we can conclude that a portion of platinum
x
1 –
δ
and small clusters is responsible for the high activity of interacted with cerium oxide to form dispersed catꢀ
this catalyst in the reaction of CO oxidation at low ionic species. The presence of these platinum species
resulted in the appearance of a small peak in the curve
of CO conversion in the region of low temperatures
temperatures.
The state of palladium in catalyst Pd/CeO ꢀII
treated with hydrazine was essentially different.
2
(Fig. 6c, curve
2). In this case, in general, the temperꢀ
ature dependence of the conversion of CO was similar
HRTEM, electronic diffuse reflectance spectroꢀ to that observed on catalyst Pt/Al O [50]. Its shape
2
3
scopic, and XPS data suggest a decrease in the degree depends on the heat of CO adsorption on coarse metal
of dispersion of Pd because of the formation of coarse particles.
particles. This is consistent with the results of XRD
analysis, which suggest the presence of a metal phase lyst Pt/CeO ꢀII occurred as coarse metal particles of
The HRTEM data indicate that platinum in cataꢀ
2
with an average crystallite size of 16 nm. However, sizes from 50 to 200 nm. According to the results of
only 60% palladium occurred as a metal. It is likely XRD analysis, platinum almost entirely occurred as
that a portion of palladium, which cannot be detected metal particles with a crystallite size of 22 nm. In this
case, both platinum metal particles and the dispersed
particles of PtO and PtO2 were detected by XPS. As
noted above, XRD analysis provides information on
the occurrence of a platinum metal phase, whereas
oxide and dispersed states cannot be detected by this
by XRD analysis because of the superposition of the
most intense reflections of oxide phases and the supꢀ
port phase, occurred as PdO because the TPRꢀH2
curves exhibited a peak with a maximum at 53
°С and
the ratio H /Pd = 1.1. The appearance of a broad and
2
technique. The TPRꢀH method additionally supꢀ
2
lowꢀintensity peak with a maximum at 170°С in the
ported XPS data: the occurrence of reduction peaks at
a low temperature suggests the reduction of platinum
oxide compounds, whereas the somewhat increased
TPRꢀH curves was due to the presence of an insignifꢀ
2
icant fraction of palladium that strongly interacted
with the support. The experimental data allowed us to
conclude that the treatment of the catalyst with hydraꢀ
ratio H /Pt = 2.9 suggests the participation of oxygen
2
from CeO in the process. In this case, the activity of
2
0
zine resulted in the formation of coarse Pd and PdO
catalyst Pt/CeO ꢀII, as well as Pt/CeO ꢀI, depended
2
2
particles and an insignificant amount of Pd2 as the
+
on coarse particles.
constituent of the interaction phase Pd Ce xO
.
x
1 –
2 –
δ
The HRTEM technique did not allow us to detect
palladium and platinum particles in sample
PdPt /CeO ꢀI. In this case, an analysis of EDX specꢀ
In this case, the activity of coarse palladium particles,
which did not interact with the support, was close to
2
2
the activity of catalyst Pd/Al O [50]. Thus, the reacꢀ tra indicated that Pd and Pt were present in the sample
2
3
tion occurs by the Langmuir–Hinshelwood mechaꢀ and their atomic ratio was close to a specified value.
nism, and the initiation of reaction depends on the Consequently, both palladium and platinum in sample
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
293
PdPt /CeO ꢀI are uniformly dispersed on the support. action of palladium and platinum with the surface of
2
2
According to electronic diffuse reflectance spectroꢀ CeO with the formation of a new phase state of cataꢀ
2
scopic data, supported platinum was a mixture of lyst components—Pd CeO
or Pd Pt CeO
surꢀ
x
2 –
δ
x
y
2 –
δ
Pt(IV) and Pt(II) in an oxygen environment. Only an
insignificant amount of a metal phase was detected in
this catalyst by XRD analysis. In general, Pd and Pt
mainly occurred in an oxidized state. This conclusion
was supported by XPS data, which suggest the presꢀ
face solid solutions. The resulting interaction phases
possess two new properties: (1) the oxygen constituent
of these phases exhibits much higher activity in the
catalytic reaction of CO oxidation than the oxygen of
the CeO support or oxide particles of palladium and
2
2+
ence of Pd as the constituent of the interaction phase
Pd Ce xO2 – and in O–Pd –O dimers. The state of
0
0
platinum; (2) Pd and Pt small metal clusters, where
n
n
x
1 –
δ
2
2
+
n 2, are stabilized as the constituents of these surface
≥
platinum corresponds to Pt , and only an insignifiꢀ
phases. The former property is responsible for the
occurrence of oxidation by the stepwise Mars–van
Krevelen mechanism because of the activity of lattice
oxygen [53]. The latter property facilitates the activaꢀ
tion of adsorbed carbon monoxide with the formation
of a weakly bound state, and the presence of small
metal clusters decreases the enthalpy of CO adsorpꢀ
tion to initiate the reaction at a low temperature.
cant portion of platinum occurred as metal clusters.
The TPRꢀH spectra also supported this conclusion:
2
the occurrence of a single peak with a maximum at
125°С in them and the superstoichiometric ratio
H /(Pd + Pt) = 3.3 suggests the presence of the ionic
2
forms of Pd and Pt, which interacted with the support,
and the participation of cerium dioxide in reduction.
Thus, all of the methods suggest the high degree of disꢀ
persion of active components on the support surface
The formation of both an ionic state of platinum as
and their ionic state. This provides the high activity of a constituent of the surface interaction phase and platꢀ
catalyst PdPt /CeO ꢀI, as well as Pd/CeO ꢀI, in the inum metal clusters occurs only under the action of a
2
2
2
course of CO oxidation in the lowꢀtemperature palladium precursor supported in the synthesis of
region.
samples in accordance of procedure I. If platinum preꢀ
cursors are used without the addition of palladium
precursors, platinum as a catalyst constituent cannot
form surface phases of this type. Note that the formaꢀ
According to HRTEM and XRD data, coarse Pd–
Pt alloy metal particles of size 100–300 nm with variꢀ
ous Pd/Pt ratios occurred in catalyst PdPt /CeO ꢀII.
XPS also suggests the formation of the coarse particles
of palladium metal and PtO. The presence of coarse
PtO particles was supported by the TPRꢀH method
because the reduction spectrum exhibited a peak at
14°С. The appearance of a broad peak at a higher temꢀ
perature suggests that only an insignificant portion of
palladium interacted with the support to activate oxyꢀ
gen in CeO . In general, the activity of this catalyst
2
2
tion of the platinum and CeO interaction phases as
2
Pt CeO was detected previously by Bera and coauꢀ
x
2 – x
thors [10–12] with the use of the solution–combusꢀ
tion method. However, the resulting catalysts were not
so active at a low temperature as PdPt /CeO ꢀI. Thus,
2
2
2
the above data indicate that palladium interacts with
the surface of cerium oxide much stronger than platiꢀ
num. In the absence of palladium, platinum does not
form surface interaction phases; only individual
phases of metal and oxide particles, which cannot oxiꢀ
dize CO at a low temperature, are formed. The Pd–Pt
alloy particles also cannot be active sites for the formaꢀ
2
remained low because it depended on the presence of
coarse particles, which do not interact with the supꢀ
port.
The entire set of experimental data suggests that, tion of adsorbed states of CO and oxygen, which parꢀ
first, the active catalysts based on CeO for lowꢀtemꢀ ticipate in lowꢀtemperature oxidation.
2
perature oxidation should contain palladium as an
Thus, in this work, we used complex salts to obtain
active component and the lowꢀtemperature process on Pd–Pt–Ce oxide catalysts for the lowꢀtemperature
platinum is impossible. Second, the reduction of cataꢀ oxidation of CO. In the catalysts prepared with the use
lysts with hydrazine leveled the behaviors of palladium of only platinum precursors, metal and oxide (PtO,
and platinum and coarse metal, bimetallic, or oxide PtO ) nanoparticles, which do not possess lowꢀtemꢀ
2
particles, which do not exhibit lowꢀtemperature activꢀ perature activity, are formed on the surface of CeO . In
2
ity, were formed in this case on the support surface. the mixed palladium–platinum catalyst, under the
Third, the simple incipient wetness impregnation of action of palladium, platinum becomes strongly disꢀ
the support with the solutions of complex precursor persed to result in the formation of a Pd Pt CeO
x
y
2 –
δ
compounds made it possible to obtain very active catꢀ combined surface phase, which is responsible for the
alysts containing Pd or Pd–Pt as active components. A lowꢀtemperature oxidation of CO.
characteristic property of systems Pd/CeO ꢀI and
2
PdPt /CeO ꢀI, which possess good catalytic characꢀ
2
2
REFERENCES
teristics, is an extremely high degree of metal disperꢀ
2+
sion with the formation of the ionic states Pd and
1. Recent Progress in Catalysis by Ceria and Related Comꢀ
pounds, Bernal, S., Kaspar, J., and Trovarelly, A., Eds.,
Catal. Today, 1999, vol. 50, p. 173.
2+
Pt . The coarse metal and oxide particles of pallaꢀ
dium and platinum were not formed in these catalysts.
Lowꢀtemperature activity depends on only the interꢀ
2. Haruta, M., Chem. Rec., 2003, vol. 3, p. 75.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

294
SLAVINSKAYA et al.
3
. Engler, B, Koberstein, E, Lindner, D, and Lox, E, in 25. Shubin, Yu.V. and Korenev, S.V., Katal. Prom–sti.
,
Catalysis and Automotive Pollution Control II, Crucq, A.,
Ed., Amsterdam: Elsevier, 1991, p. 641.
2008, special issue, p. 69.
6. Chen, J.P. and Lim, L.L., Chemosphere, 2002, vol. 49,
p. 363.
7. Wu, S.ꢀH. and Chen, D.ꢀH., J. Colloid Interface Sci.
2003, vol. 259, p. 282.
2
2
2
4. Trovarelli, A., Catal. Rev. Sci. Eng., 1996, vol. 38,
p. 439.
,
5. Colussi, S., Gayen, A., Camellone, M.F., Boaro, M.,
Llorca, J., Fabris, S., and Trovarelli, A., Angew. Chem.
009, vol. 121, p. 8633.
,
8. Wojcieszak, R., Zielinski, M., Monteverdi, S., and Betꢀ
tahar, M.M., J. Colloid Interface Sci., 2006, vol. 299,
p. 238.
2
6
. Pierre, D., Deng, W., and FlytzaniꢀStephanopoulos, M.,
Top. Catal., 2007, vol. 46, p. 363.
2
9. Wojcieszak, R., Zielinski, M., Monteverdi, S., and Betꢀ
tahar, M.M., J. Colloid Interface Sci., 2008, vol. 317,
p. 166.
7. Luo, J.ꢀY., Meng, M., Xian, H., Tu, Y.ꢀB., Li, X.ꢀG.,
and Ding, T., Catal. Lett., 2009, vol. 133, p. 328.
8
. Burton, P.D., Lavenson, D., Johnson, M., Gorm, D., 30. Kim, H.K., Lee, Y.B., Choi, E.Y., Park, H.C., and
Karim, A.M., Conant, T., Datye, A.K., Hernandezꢀ
Sanchez, B.A., and Boyle, T.J., Top. Catal., 2008,
vol. 49, p. 227.
Park, S.S., Mater. Chem. Phys., 2004, vol. 86, p. 420.
3
3
3
1. Wu, M.ꢀLi., Chen, D.ꢀH., and Huang, T.ꢀC., J. Colloid
Interface Sci., 2001, vol. 243, p. 102.
9
. Yang, Y., Saoud, K.M., Abdelsayed, V., Glaspell, G.,
Deevi, S., and ElꢀShall, M.S., Catal. Commun., 2006,
vol. 7, p. 281.
2. Price, W.J., Analytical AtomicꢀAbsorption Spectroscopy
,
London: Heyden, 1972.
3. Boehm, H.ꢀP and Knozinger, H, in Catalysis Science
and Technology, Anderson, J.R.. and Boudart, M.,
Eds., Berlin: Springer, 1983, p. 39.
1
0. Bera, P., Patil, K.C., Jayram, V., Subbanna, G.N., and
Hegde, M.S., J. Catal., 2000, vol. 196, p. 293.
1
1. Bera, P., Gayen, A., Hegde, M.S., Lalla, N.P., Spaꢀ
3
4. Lowell, S., Shields, J.E., Thomas, M.A., and Thomꢀ
mes, M., Characterization of Porous Solids and Powders:
Surface Area, Pore Size and Density, Dordrecht: Kluꢀ
wer, 2006.
daro, L., Frusteri, F., and Arena, F., J. Phys. Chem. B
003, vol. 107, p. 6122.
,
2
1
2. Bera, P., Priolkar, K.R., Gayen, A., Sarode, P.R.,
Hegde, M.S., Emura, S., Kumashiro, R., Jayaram, V.,
and Subbanna, G.N., Chem. Mater., 2003, vol. 15,
p. 2049.
3
3
3
3
3
4
5. Bensalem, A., Muller, J.C., and BozonꢀVerduraz, F.,
J. Chem. Soc., Faraday Trans., 1992, vol. 85, p. 153.
6. Rakai, A., Tessier, D., and BozonꢀVerduraz, F., New J.
Chem., 1992, vol. 16, p. 869.
1
3. Gayen, A., Priolkar, K.R., Sarode, P.R., Jayaram, V.,
Hegde, M.S., Subbanna, G.N., and Emura, S., Chem.
Mater., 2004, vol. 16, p. 2317.
7. Binet, C., Bardi, A., and Lavalley, J.C., J. Phys. Chem.
994, vol. 98, p. 6392.
,
1
1
1
1
4. Roy, S. and Hegde, M.S., Catal. Commun., 2008, vol. 9,
p. 811.
8. Lietz, G., Lieske, H., Spindler, H., Hanke, W., and
Völter, J., J. Catal., 1983, vol. 81, p. 17.
5. Sharma, S. and Hegde, M.S., J. Chem. Phys., 2009,
vol. 130, p. 114706.
9. Mason, W.R. and Gray, H.B., J. Am. Chem. Soc., 1968,
vol. 90, p. 5721.
6. Deganello, G., Giannici, F., Martorana, A., Pantaleo, G.,
and Prestianni, A., J. Phys. Chem. B, 2006, vol. 110,
p. 8731.
0. Handbook of Xꢀray Photoelectron Spectroscopy, Moulꢀ
der, J.F., Stickle, W.F., and Sobol, P.E., Eds., Eden
Prairie. Minn.: PerkinꢀElmer Corporation, Physical
Electronics Division, 1992.
1
7. Longo, A., Balerna, A., Acapito, F., D’Anca, F., Gianꢀ
nici, F., Liotta, L.F., Pantaleod, G., and Martorana, A.,
J. Synchrotron Radiat., 2005, vol. 12, p. 499.
41. Brun, M., Berthet, A., and Bertolini, J.C., J. Electron
Spectrosc. Relat. Phenom., 1999, vol. 104, p. 55.
1
1
2
2
2
8. Damyanova, S. and Bueno, J.M.C., Appl. Catal., A
,
2
003, vol. 253, p. 135.
4
2. Suhonen, S., Valden, M., Pessa, M., Savimaki, A.,
Harkonen, M., Hietikko, M., Pursiainen, J., and
Laitinen, R., Appl. Catal., 2001, vol. 207, p. 113.
9. Graf, P.O., de Vlieger, D.J.M., Mojet, B.L., and Lefꢀ
ferts, L., J. Catal., 2009, vol. 262, p. 181.
4
3. Slavinskaya, E.M., Boronin, A.I., Danilova, I.G.,
Amosov, Yu.I., Ivanova, A.S., Kuznetsov, P.A.,
Polukhina, I.A., Gulyaev, R.V., Stadnichenko, A.I.,
Koshcheev, S.V., Zaikovskii, V.I., and Noskov, A.S.,
Kinet. Katal., 2009, vol. 50, no. 6, p. 853 [Kinet. Catal.
(Engl. Transl.), vol. 50, no. 6, p. 819].
0. Roth, D., Gelin, P., Primet, M., and Tena, E., Appl.
Catal., A, 2000, vol. 203, p. 37.
1. Yoshida, H., Nakajima, T., Yazawa, Y., and Hattori, T.,
Appl. Catal., B, 2007, vol. 71, p. 70.
2. Le Normand, F., Hilaire, L., Kili, K., Krill, G., and
Marie, G., J. Phys. Chem., 1988, vol. 92, p. 2561.
4
4. Boronin, A.I., Slavinskaya, E.M., Danilova, I.G.,
Gulyaev, R.V., Amosov, Yu.I., Kuznetsov, P.A.,
Polukhina, I.A., Koscheev, S.V., Zaikovskii, V.I., and
Noskov, A.S., Catal. Today, 2009, vol. 144, p. 201.
23. RF Patent RU 2294240, Byull. Izobret., 2007, no. 6.
24. Snytnikov, P.V., Yusenko, K.V., Korenev, S.V., Shuꢀ
bin, Yu.V., and Sobyanin, V.A., Kinet. Katal., 2007,
vol. 48, p. 292 [Kinet. Catal. (Engl. Transl.), vol. 48, 45. Boronin, A.I., Ivanova, A.S., Slavinskaya, E.M.,
p. 276]. Gulyaev, R.V., Danilova, I.G., Zaikovskii, V.I., and
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

LOWꢀTEMPERATURE OXIDATION OF CARBON MONOXIDE ON Pd(Pt)/CeO2
295
Noskov, A.S., Proc. 6th World Congr. on Oxidation 50. Ivanova, A.S., Slavinskaya, E.M., Gulyaev, R.V., Zaiꢀ
Catalysis, Lille, France, 2009.
6. Fung, S.C., J. Catal., 1979, vol. 58, p. 454.
7. Darby, J.B. and Myles, Jr.K.M., Metall. Trans. A, 1972,
kovskii, V.I., Stonkus, O.A., Danilova, I.G., Plyaꢀ
sova, L.M., Polukhina, I.A., and Boronin, A.I., Appl.
Catal., B, 2010, vol. 97, p. 57.
4
4
vol. 3, p. 653.
51. Bourane, A. and Bianchi, D., J. Catal., 2002, vol. 209,
p. 126.
4
8. Monti, D.A.M. and Baiker, A., J. Catal., 1983, vol. 83,
p. 323.
52. Bourane, A. and Bianchi, D., J. Catal., 2003, vol. 218,
p. 447.
4
9. Jen, H.ꢀW., Graham, G.W., Chun, W., McCabe, R.W.,
Cuif, J.ꢀP., Deutsch, S.E., and Touret, O., Catal. 53. Mars, P. and van Krevelen, D.W., Chem. Eng. Sci.
,
Today, 1999, vol. 50, p. 309.
1954, vol. 3, Suppl. 1, p. 41.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011