Use of the differential charging effect in XPS to determine the nature of surface compounds resulting from the interaction of a Pt/(BaCO3 + CeO2) model catalyst with so x

Article abstract of DOI:10.1134/S0023158411040161

Changes in the chemical composition of the surface of a Pt/(BaCO ₃ + CeO₂) model NO _x storage-reduction catalyst upon its interaction with SO _x (SO₂ (260 Pa) + O ₂ (2600 Pa) + H₂O (525 Pa)) followed by regeneration in a mixture of CO (2100 Pa) with H₂O (525 Pa) were studied by X-ray photoelectron spectroscopy (XPS). Model catalyst samples were prepared as a thin film (about several hundreds of angstrom units in thickness) on the surface of tantalum foil coated with a layer of aluminum oxide (～100 A). It was found that the Pt/BaCO₃ and Pt/CeO₂ catalyst constituents acquired different surface charges (differential charging) in the course of photoelectron emission; because of this, it was possible to determine the nature of surface compounds formed as a result of the interaction of the catalyst with a reaction atmosphere. It was found that barium carbonate was converted into barium sulfate as a result of reaction with SO _x on the surface of BaCO₃ at 150°C. As the treatment temperature in SO _x was increased to 300°C, the formation of sulfate on the surface of CeO ₂ was observed. The sulfatization of CeO₂ was accompanied by the reduction of Ce(IV) to Ce(III). The regeneration reaction of the catalyst treated in SO _x at 300°C resulted in the consecutive decomposition of cerium(III) sulfate at ≤500°C and then barium sulfate at 600-700°C. Upon the decomposition of BaSO₄, a portion of sulfur was converted into a sulfide state, probably, because of the formation of BaS.

Full text of DOI:10.1134/S0023158411040161

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2011, Vol. 52, No. 4, pp. 595–604. © Pleiades Publishing, Ltd., 2011.
Original Russian Text © M.Yu. Smirnov, A.V. Kalinkin, A.A. Dubkov, E.I. Vovk, A.M. Sorokin, A.I. Nizovskii, B. Carberry, V.I. Bukhtiyarov, 2011, published in Kinetika i Kataliz,
011, Vol. 52, No. 4, pp. 605–614.
2
Use of the Differential Charging Effect in XPS to Determine
the Nature of Surface Compounds Resulting from the Interaction
of a Pt/(BaCO + CeO ) Model Catalyst with SO
3
2
x
a
a
b
a
a
M. Yu. Smirnov , A. V. Kalinkin , A. A. Dubkov , E. I. Vovk , A. M. Sorokin ,
a
b
a, c
A. I. Nizovskii , B. Carberry , and V. I. Bukhtiyarov
a
Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia
b
Ford Forschungszentrum Aachen GmbH, Aachen, Dꢀ52072 Germany
c
Novosibirsk State University, Novosibirsk, 630090 Russia
eꢀmail: smirnov@catalysis.ru
Received November 9, 2010
Abstract—Changes in the chemical composition of the surface of a Pt/(BaCO + CeO ) model NO storꢀ
3
2
x
age–reduction catalyst upon its interaction with SO (SO (260 Pa) + O (2600 Pa) + H O (525 Pa)) followed
x
2
2
2
by regeneration in a mixture of CO (2100 Pa) with H O (525 Pa) were studied by Xꢀray photoelectron specꢀ
2
troscopy (XPS). Model catalyst samples were prepared as a thin film (about several hundreds of angstrom
units in thickness) on the surface of tantalum foil coated with a layer of aluminum oxide (~100 Å). It was
found that the Pt/BaCO and Pt/CeO catalyst constituents acquired different surface charges (differential
3
2
charging) in the course of photoelectron emission; because of this, it was possible to determine the nature of
surface compounds formed as a result of the interaction of the catalyst with a reaction atmosphere. It was
found that barium carbonate was converted into barium sulfate as a result of reaction with SO on the surface
x
of BaCO at 150
°
C. As the treatment temperature in SO was increased to 300
°
C, the formation of sulfate on
3
x
the surface of CeO was observed. The sulfatization of CeO was accompanied by the reduction of Ce(IV) to
2
2
Ce(III). The regeneration reaction of the catalyst treated in SO at 300
°
C resulted in the consecutive decomꢀ
x
position of cerium(III) sulfate at
≤
500
°
C and then barium sulfate at 600–700 C. Upon the decomposition of
°
BaSO , a portion of sulfur was converted into a sulfide state, probably, because of the formation of BaS.
4
DOI: 10.1134/S0023158411040161
INTRODUCTION
and then to barium nitrate under oxidizing conditions
and for the subsequent nitrate reduction to nitrogen
under reducing conditions.
The current development of automobile industry is
oriented toward decreasing fuel consumption and
reducing carbon dioxide emissions into the atmoꢀ
sphere. In particular, this can be achieved by the use of
engines with a lean air–fuel mixture. However, under
these conditions, nitrogen oxides are formed in conꢀ
siderable amounts, and they cannot be efficiently neuꢀ
tralized on commonly used threeꢀway catalysts with a
large excess of oxygen in the reaction mixture. Toyota
developed soꢀcalled NOx storage–reduction (NSR)
catalysts [1] as an approach to solve the problem of the
neutralization of nitrogen oxides. The principle of
operation of the NSR catalyst is the following: the
basic component of the catalyst binds nitrogen oxides
as nitrates in contact with NOx in an oxidizing atmoꢀ
sphere; then, nitrates are reduced to nitrogen during a
short pulse injection of an excess of fuel, which
increases the amount of reducing gases (CO, H2, and
CHx) in the exhaust [1, 2]. Traditionally, Baꢀcontainꢀ
A disadvantage of NSR catalysts of this type is their
high sensitivity to poisoning by sulfur compounds,
which appear in exhaust gases as a result of the oxidaꢀ
tion of Sꢀcontaining impurities in fuel. It was found
that the interaction of a catalyst with sulfur oxides in
an oxidizing atmosphere leads to the formation of staꢀ
ble barium sulfates, as a result of which the absorption
capacity for nitrogen oxides decreases [3–11]. Cataꢀ
lyst regeneration, which consists in the removal of the
sulfate ion, is performed in a reducing atmosphere at
high temperatures, which are reached by the shortꢀ
term injection of an additional amount of fuel. At the
same time, highꢀtemperature regeneration can cause
undesirable changes in the state of the catalyst, such as
a decrease in the specific surface area, the agglomeraꢀ
tion of platinum, and the formation of barium alumiꢀ
ing substances—BaO, Ba(OH)2, and BaCO3—supꢀ nates, which are inert with respect to the absorption of
ported on
γ
ꢀAl O are used as the basic components of NO
x
[8]. The formation of barium sulfides [8, 10, 12,
2 3
the NSR catalyst. Platinum is introduced into the catꢀ 13] and/or platinum sulfides [4, 10, 14, 15] is also posꢀ
alyst for the more efficient oxidation of NO to NO2 sible; it also leads to a decrease in the catalyst activity.
595

596
SMIRNOV et al.
state but is the constituent of different components,
Intensity
291.5
appears in the spectrum as different signals on the
scale of binding energies. Previously, this method was
used in the analysis of the nature of nitrites and nitrates
formed on the surface of a Pt/BaCO /Al O model
3
2
3
NSR catalyst upon its interaction with NOx (a mixture
of NO, O2, and H O) [24].
2
295.9
EXPERIMENTAL
The experiments were performed using a VG
ESCAꢀ3 photoelectron spectrometer at a residual
–
7
pressure of <5
×
10 Pa in the analyzer chamber. Nonꢀ
radiation ( = 1253.6 eV) was
monochromatic Mg
K
hν
α
used to obtain the XPS spectra. Before the experiꢀ
305 ments, the binding energy scale of the spectrometer
was calibrated using the following signal positions of
gold and copper metals: Au 4f7/2 (84.0 eV) and Cu 2p3/2
(932.6 eV). The treatment of the photoemission specꢀ
tra (the deconvolution into individual components,
the measurement of XPS peak areas, and the determiꢀ
nation of binding energies) of the test samples was perꢀ
280
285
290
295
300
Binding energy, eV
Fig. 1. XPS spectrum of the Pt/(BaCO + CeO ) model
catalyst in the initial state measured in the C1
no correction for charging and no background subtraction.
3
2
s region with
A method for increasing the resistance of NSR catꢀ formed after Shirley background subtraction.
alysts to poisoning by sulfurꢀcontaining impurities is
The samples of a model NSR catalyst were preꢀ
pared as a thin film on the surface of tantalum foil. To
exclude the interaction of catalyst components with
the introduction of CeO , TiO , Fe O3, and MgO addiꢀ
2
2
2
tives into the catalysts [16–20] or the replacement of
an Al O support by an oxide less prone to the formaꢀ
2
3
tion of sulfates [21, 22]. It was shown that, in the operꢀ tantalum, the foil was coated with an aluminum oxide
ation of a catalyst containing cerium dioxide along film ~100 Å in thickness by evaporation aluminum
–
3
with a BaO (BaCO3) basic component, a portion of metal in an atmosphere of oxygen at 1.3
× 10 Pa
SOx was bound to the catalyst as cerium sulfate. Thus, before supporting the catalyst components. Barium
the fraction of the Baꢀcontaining component, which
was eliminated from the process of NOx removal as a
result of conversion into barium sulfate, decreased
carbonate and cerium dioxide were supported onto the
surface of the alumina film from a suspension of
BaCO3 and CeO2 (in a molar ratio of 1 : 1) in ethanol
followed by drying in air and calcination in a vacuum
[
16]. It is assumed that cerium dioxide also plays a posꢀ
itive role in the regeneration of an NSR catalyst
because of its high catalytic activity in the reaction of
water gas conversion with the formation of H2 [23].
at 600 С. The BaCO3 and CeO2 film thickness was sufꢀ
°
ficient for the complete shielding of signals from the
Al O /Ta substrate in the measurements of XPS specꢀ
2
3
In this work, we report the results of a study of the
interaction of SOx with the samples of a model NSR
catalyst containing Pt, BaCO3, and CeO2 performed
by Xꢀray photoelectron spectroscopy (XPS). The aim
of this study was to determine the nature of the resultꢀ
ing surface compounds and the conditions of sulfate
formation and decomposition on the surfaces of
BaCO3 and CeO2. We used XPS because this technique
allowed us not only to reliably determine the oxidation
tra. Then, platinum was supported by thermal evapoꢀ
ration in a vacuum. Previously, an analogous method
was used for the preparation of model catalysts for
XPS analysis [24].
The resulting model catalyst samples were treated
with SOx in a mixture of SO2 (260 Pa), O2 (2600 Pa),
and H O (525 Pa) at 150 and 300
°
С
with the subseꢀ
quent regeneration in a mixture of CO (2100 Pa) and
H O (525 Pa) at 500, 600, and 700 in the preparaꢀ
2
state of sulfur from the binding energy (Eb) measured
°
С
2
for the S 2p3/2 signal but also to reveal the component
to which sulfur is bound. To solve the latter problem,
we used a method based on the differential charging
effect, the essence of which consists in different chargꢀ
ing of test catalyst sample components (in this case,
BaCO3 and CeO2) in the process of electron photoeꢀ
mission during the measurement of XPS spectra. As a
tion chamber of the spectrometer. The samples were
heated by passing an alternating current through the
tantalum foil. The temperature was measured using a
Chromel–Alumel thermocouple welded to the reverse
side of the foil. After each treatment in a reaction mixꢀ
ture, the sample was transferred to the analyzer chamꢀ
result of differential charging, the same element (for ber away from contact with the atmosphere for meaꢀ
example, sulfur), which occurs in the same oxidation suring the spectra.
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

USE OF THE DIFFERENTIAL CHARGING EFFECT IN XPS TO DETERMINE
597
Table 1. Binding energies in the XPS spectrum of the initial Pt/(BaCO + CeO ) sample obtained directly from the spectra
3
2
and corrected for differential charging
Measured values of Eb
Corrected values of Eb, eV
,
Signal
Assignment
eV
Δ
= 6.7 eV
Δ
= 4.9 eV
C
B
C1
s
291.5
284.8 (+)
–
–
C/Pt/BaCO3
2
–
2
95.9
535.2
38.0
783.7
289.2 (+)
528.5 (?)
C
O3 /BaCO3
O1
s
530.3
2–
O /CeO
2
2
–
5
531.3 (+)
777.0 (?)
778.9 (+)
89.0 (+)
914.9 (?)
107.5 (?)
–
778.8
–
C
O3 /BaCO3
Ba3d5/2
Ba4d5/2
2+
Ba /CeO
2
785.6
2+
Ba /BaCO
3
95.7
–
2+
Ba /BaCO
3
Ce3
d
d
,
,
u
'''
921.6
114.2
76.8
916.7
109.3
4+
Ce /CeO
2
2
Ce4
A
4+
Ce /CeO
Pt4f7/2
70.1(?)
72.0(+)
71.9
–
Pt/CeO2
78.7
Pt/BaCO3
RESULTS AND DISCUSSION
Analysis of the XPS Spectra of the Pt/(BaCO + CeO )
the conditions of sample preparation, one of these sigꢀ
nals with the greatest binding energy belongs to carbon
as the constituent of carbonate ions in BaCO3, whereas
it is obvious that the other signal belongs to amorphous
3
2
Catalyst in the Initial State
As mentioned above, during the recording of the carbon.
XPS spectra of multicomponent systems, each of the
By analogy with our previous work [24], we anaꢀ
test sample components acquires a stationary positive
potential in the process of electron photoemission; in
this case, as a rule, its values are different for different
components. As a result, the photoemission signals of
an element, which is the constituent of a certain comꢀ
ponent, is shifted from the true position by a value corꢀ
responding to the potential acquired by the given comꢀ
ponent. The internal standard method is the simplest
method for the quantitative evaluation of potential
and the determination of the true positions of binding
energies of photoemission signals under conditions of
the differential charging of the test sample [24, 25–27].
lyzed the binding energies (Eb) of signals in the XPS
spectrum of an initial catalyst sample, which were
obtained taking into account their shifts as a result of
differential charging. Under the assumption that the
C 1
to amorphous carbon with the tabulated value of Eb
of 284.8 eV, the shift is = 6.7 eV. Table 1 summarizes
s signal with a binding energy of 291.5 eV belongs
=
Δ
B
the values of Eb obtained directly from the spectra
column 2) and after correction for the differential
charging (column 3). The plausible values of bindꢀ
(
Δ
B
ing energy are marked by a plus sign. The values that
cannot be ascribed to a particular chemical compound
are indicated by a question mark. Let us consider consecꢀ
utively all of the Eb values given in Table 1 (column 3).
The C 1s signal from amorphous carbon, which is
accumulated on the sample surface upon exposure to
residual gases in the spectrometer chamber, is comꢀ
monly used as an internal standard. The binding
As mentioned above, two signals are observed in the
energy of this C 1
shows the spectrum in the C 1
the Pt/(BaCO + CeO ) model catalyst in the initial its binding energy of 284.8 eV specifies a charging
s
signal is taken to be 284.8 eV. Figure 1 spectrum of the C 1
s region (Fig. 1); the most intense
s
region of a sample of of these signals is attributed to amorphous carbon, and
3
2
state. The spectrum is the superposition of two signals value of 6.7 eV. The other signal after correction for
with binding energies of 291.5 and 295.9 eV. Based on charging acquires the value of Eb = 289.2 eV, which is
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

598
SMIRNOV et al.
and Ba 4d5/ (89.0 eV) signals, which can be related to
Intensity
2
71.4
one of the components in the model catalyst— BaCO3
.
74.7
At the same time, the correction for gives too
Δ
B
low values of binding energies for the signals that charꢀ
acterize cerium. Thus, according to published data for
CeO2, the signal
[32–34] and the signal
31]. For the second signal in the O1
u
''' in the Ce 3
d
region has Eb 916.7 eV
≈
4
3
A
in the Ce 4
d
region, ~109.3 eV
s region, the use of
[
Δ
as a correction also gives a binding energy 2 eV
B
lower that that expected for CeO2 [31]. A lowꢀintensity
signal detected in the Ba 3 region should also be
noted. For this signal, the binding energy corrected for
is Eb(Ba 3d5/2 = 777.0 eV; this anomalously low
value was observed in none of barium compounds
35, 36]. All the above suggests that, indeed, the
d
Δ
B
)
[
potential of CeO2 differed from the potential of BaCO3
in the course of XPS measurements; that is, differenꢀ
71.9
2
tial charging occurred, and the C 1s signal of amorꢀ
phous carbon cannot be used as an internal standard
for CeO2. A possible reason consists in an insignificant
accumulation of amorphous carbon on the surface of
75.2
CeO because of a significant oxidizing power of the
2
latter. For example, published survey spectra [37] indiꢀ
cate that carbon deposits were almost absent on the
surface of a Ceꢀcontaining oxide. Thus, another signal
suitable to serve as an internal standard should be
found in the spectrum. It is most convenient to use the
u
''' signal for this purpose; in the complex Ce 3d specꢀ
1
trum, this signal is located separately from other sigꢀ
nals and has a stable binding energy value of 917.6 eV
[32, 34, 37, 38]. Taking into account this binding
60
65
70
75
80
85
Binding energy, eV
energy Eb
(
u
'''), we obtain the correction
Δ
C
= 4.9 eV for
the CeO component. Taking into account this correcꢀ
2
Fig. 2. XPS spectra of the Pt/(BaCO + CeO ) model catꢀ
tion, we calculated binding energies for all “strange”
photoemission signals (the corresponding values are
given in Table 1, column 4). It is evident that, after the
3
2
alyst measured in the Pt 4
f region (1) in the initial state and
(
2
) after its reaction with SO (260 Pa SO + 2600 Pa O +
x
2
2
5
25 Pa H O) at 300
°
C and the subsequent regeneration in a
2
correction for
positions. Moreover, a weak doublet signal observed as
a shoulder in the Ba 3 spectrum acquires a value of
= 778.8 eV after correction for ; this value is typꢀ
Δ
C, the signals occupy the expected
mixture of 2100 Pa CO + 525 Pa H O at ( ) 500 or ( ) 700 C.
3
4
°
2
d
characteristic of inorganic carbonates, in particular,
BaCO3 [24, 28–30]. In the O1 region, the spectrum is
E
b
Δ
C
2+
ical of barium in the oxidation state Ba . On this
basis, this signal can be ascribed to barium cations in
contact with CeO2. A similar state of barium was found
s
also the superposition of two signals with an intensity
ratio of ~1 : 1 (Table 1). After correction for charging,
the binding energy of one of the signals becomes
31.3 eV, which is close to the values published for barꢀ
ium carbonate (530.7–531.2 eV) [24, 28–30]. In this
case, the second signal becomes 528.5 eV, which is
earlier in the2 Pt/BaCO /Al O model catalyst and
3
2
3
+
referred to Ba cations located at the BaCO –Al O
5
3
2
3
interface [24].
After the establishment of the fact that two compoꢀ
substantially lower than the value expected for both nents (BaCO3 and CeO2) that acquire different positive
BaCO3 and CeO2 (530.4 eV) [31]. In the Ba 3 and potentials in the course of XPS measurements occur in
Ba 4 spectral regions, which characterize barium, the test model catalyst, let us discuss the spectra that
signals are observed the binding energies Eb(Ba 3d5/2) characterize platinum supported onto the surface of
and Eb(Ba 4d5/2 of which after correction for
these components. Figure 2 (curve ) shows the Pt 4f
d
d
)
Δ
1
B
acquire the values of 778.9 and 89.0 eV respectively, spectrum of the initial catalyst. It is difficult to sepaꢀ
which will agree well with those published for barium rate the individual signals that belong to platinum supꢀ
carbonate [24, 28–30]. Thus, the use of the correction ported onto BaCO and CeO from the overall outline
3
2
Δ
= 6.7 eV, which was obtained from the position of of the Pt 4
f
spectrum. Nevertheless, it is evident that
signal of amorphous carbon, for determining the shape of the Pt 4
spectrum cannot be described by
true binding energies gives reasonable values for the one Pt 4f7/2–Pt 4f5/2 doublet. We attempted to decomꢀ
C 1 (289.2 eV), O 1 (531.3 eV), Ba 3d5/2 (778.9 eV), pose this spectrum into two doublets (Figure 2, specꢀ
B
the C 1
s
f
s
s
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

USE OF THE DIFFERENTIAL CHARGING EFFECT IN XPS TO DETERMINE
599
trum
1
), one of which (more intense with a smaller
Δ
binding energy) can be related to platinum supported
on CeO2 and the other, which is less intense with a
larger binding energy, to platinum on BaCO3. In this
C
B
Pt CeO2
BaCO3
Δ
approach, the intensity ratio between the Pt 4
that characterize platinum on CeO2 and BaCO3 is
.53, which practically coincides with the Ce : Ba
atomic ratio (2.57) determined from Ce 3 and Ba 3
signal intensities. In Fig. 2, the spectra are shifted by
the CeO charging along the axis of Eb; therefore,
f signals
Al O
2
3
2
Ta foil
d
d
Δ
C
2
the Pt 4
corresponds to the true binding energy Eb(Pt 4f7/2
which is 71.9 eV for the initial sample (Table 1). Taking
into account the charging for the second signal, we
f signal position of platinum supported on CeO2
),
Fig. 3. Schematic diagram of the Pt/(BaCO + CeO )
3
2
model catalyst;
Δ
and
Δ
are positive charges acquired by
B
C
Δ
B
the Pt/BaCO and Pt/CeO subsystems, respectively, as a
3 2
obtained a similar value of Eb(Pt 4f7/2
)
= 72.0 eV.
result of electron photoemission in the course of XPS meaꢀ
surements.
Although the occurrence of platinum particles in a
metal state is expected based on a vacuum evaporation
procedure, the found binding energy Eb(Pt 4f7/2) is
Intensity
2
–
higher than the corresponding value for bulk platiꢀ
num—71.2 eV [35, 36]. The observed difference can
be explained by the manifestation of the final state
effect in the process of electron photoemission from
the small (about several nanometers) particles of the
SO /BaCO
4
3
2–
SO /CeO
4
2
Ba 4p3
/2
supported metal [39–41]. Similar values of Eb(Pt 4f7/2
)
we obtained earlier upon platinum evaporation onto
the surfaces of Al O [42] and BaCO /Al O [24].
2
3
3
2
3
3
2
1
Table 1 (the last column) summarizes the assignꢀ
ment of all of the observed signals based on the corꢀ
rected values of binding energies. From the standpoint
of XPS, the model catalyst sample, which is schematꢀ
ically depicted in Fig. 3, consist of two subsystems:
Pt/BaCO3 (subsystem B (based on ВаСО3)) and
Pt/CeO2 (subsystem C (based on СеО2)). Based on the
intensities of photoemission signals with considerꢀ
ation for atomic sensitivity factors [35], we estimated
the chemical composition of both of the subsystems:
S 2p3/2
S 2p1/2
160
165
170
175
180
185
Binding energy, eV
Pt/BaCO3 (subsystem B)
Pt/CeO2 (subsystem C)
Ba : C : O = 1 : 2.1 : 7.0,
Ce : O = 1 : 2.5.
Fig. 4. XPS spectra of the Pt/(BaCO
alyst measured in the S2 region ( ) in the initial state and after
its reaction with SO (260 Pa SO + 2600 Pa O + 525 Pa
+ CeO ) model catꢀ
3
2
p
1
x
2
2
The O : Ce ratio in CeO2 higher than the stoichioꢀ
metric ratio can be explained by the partial hydroxylaꢀ
tion of the oxide surface. As judged from the atomic
H O) at (
2) 150 or (3) 300 C.
°
2
ratio between the elements in subsystem B, it is most Ba 4 3 signal, which is also shown in the spectra in
p
/2
likely that its surface layer consists of barium bicarꢀ Fig. 4, was located closely to the S 2p signals. The posiꢀ
bonate, Ba(HCO3)2. The platinum contents of subꢀ tion of this signal remains almost unchanged after
systems B and C evaluated from the intensities of the treatment in SOx; therefore, the accuracy of the deterꢀ
corresponding Pt 4
expressed in terms of atomic ratios are the following:
f, Ce 3d, and Ba 3d5/2 signals and
mination of the binding energy of S 2
additionally controlled with respect to Eb(Ba 4p3/2
After the interaction of the catalyst with SOx at 150
a signal appeared in the S 2
ited noticeable asymmetry because it is the S 2p3/2
p signals can be
).
°С,
subsystem B
subsystem C
Pt : Ba = 0.6 : 1,
Pt : Ce = 0.7 : 1.
p
region; this signal exhibꢀ
–
S 2p1/2 doublet with a spin–orbital splitting of 1.2 eV
and an intensity ratio of 2 : 1 [35]. After correction for
Interaction of Pt/(BaCO + CeO ) with SO
x
3
2
ΔB, the binding energy Eb(S 2p) acquires a value of
Figure 4 shows the S 2
CeO2) catalyst recorded (
ment in a mixture of SOx at (
p
1
spectra of the Pt/(BaCO + 168.5 eV or, taking into account decomposition into
3
) before and after its treatꢀ spin–orbital components, Eb(S 2p3/2) = 169.1 eV, which
2) 150 and (
3
)
300°С. The is typical of sulfur in inorganic sulfates. In particular,
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

600
SMIRNOV et al.
can correspond to sulfate ions as the constituents of
the Ceꢀcontaining component (subsystem C).
Intensity
The Ce 3 spectra recorded in the course of treatꢀ
d
3
2
ment of the model catalyst in SO (Fig. 5) suggest that
x
the discussed S 2
fate nature. Spectrum
p
signal belongs to particles of the sulꢀ
of the catalyst in the initial
1
state has a shape characteristic of cerium dioxide. The
spectrum has a complex structure formed by the
u
'
superposition of three Ce 3d5 –Ce 3d3/2 doublet sigꢀ
/2
4+
nals [32, 38, 43] inherent to Ce cations, which
appear because of the fact that cerium cations can
relax into different final states after electron photoeꢀ
v
'
u
'''
mission from the Ce 3
the initial catalyst in Fig. 5, these signals are desigꢀ
nated by the symbols '', and ''' for Ce 3 5 or ''
and ''' for Ce 3 3/2, respectively, in accordance with
the designations proposed by Burroughs et al. [43].
The doublet marked by the symbols v' and ' was
d level [44]. In the spectrum of
v
'''
1
u
u
''
v
,
v
v
d
u, u ,
/2
u
d
v
v
''
u
3+
related to electron photoemission from the Ce catꢀ
3+
ion [32–34, 37, 38]. The Ce content of the initial
catalyst was 15%, as determined from the intensity
ratio between signals that form the Ce 3d spectrum;
this value is consistent with estimations published for
3+
cerium oxides [45]. The detection Ce in an amount
of 10–20% in the XPS spectra of CeO2 was usually
related to the partial reduction of the CeO2 under the
action of Xꢀradiation or as a result of the reaction with
residual gases (CO and H2) [34, 46]. After the interacꢀ
870
880
890
900
910
920
930
Binding energy, eV
Fig. 5. XPS spectra of the Pt/(BaCO + CeO ) model catꢀ
tion with SOx at 150 С, the structure of the Ce 3d specꢀ
°
3
2
alyst measured in the Ce 3
d region (1) in the initial state
trum did not crucially change; the contribution of
and after its reaction with SO (260 Pa SO + 2600 Pa O +
3+
x
2
2
Ce remained at almost the same level (~18%). After
5
25 Pa H O) at (
2
) 150 or (
3
) 300 C.
°
2
the interaction at 300 С, the spectrum changed draꢀ
°
matically to acquire a shape typical of Ce(III) comꢀ
pounds [31, 34]. The change in the structure of the
Christie et al. [28] published the value of Eb(S 2
68.4 eV for BaSO4. If we determine Eb(S 2p3/2
rected for in subsystem C (Pt/CeO2), we obtain a
value of 171.1 eV, which is too high to be ascribed to an
oxygen compound of sulfur. Thus, the treatment of the
p) =
) corꢀ
Ce 3
Ce(IV) to Ce(III) in parallel with the appearance of
the S 2 signal with a binding energy characteristic of
sulfates in the Pt/CeO2 subsystem, is a convincing arguꢀ
d spectrum, which suggests the reduction of
1
Δ
C
p
model catalyst in SOx at 150
°С
leads to the predomiꢀ ment in favor of the formation of Ce(III) sulfate at
nant formation of barium sulfate.
300 . Previously, we obtained cerium sulfate by the
°С
interaction of the CeO2 [38] and Pt/CeO2 [47] model
As the interaction temperature was increased to
00 С, the intensity of the S 2p signal increased. The
systems with SO and a mixture of SO + O2 in the temꢀ
2
2
3
°
perature range of 200–400°С.
simultaneous signal broadening and the appearance of
a shoulder on the side of smaller Еb values indicate the
In Fig. 2, spectrum 2 in the Pt 4f region was
formation of at least two Sꢀcontaining compounds. obtained after the interaction of the catalyst with SOx
After decomposing spectrum in Fig. 4 into two comꢀ at 300 . A considerable broadening of signals was
ponents and shifting it by the differential charging observed, as compared with the spectra of samples in
3
°С
Δ
B
along the axis of binding energies, we can see that one the initial state. For this reason, it becomes impossible
of the components should be attributed to barium sulꢀ to separate two doublets that characterize platinum
fate. The S 2
increases in comparison with the signal intensity after theless, we can reliably state that Eb(Pt 4f7/2
treatment in SOx at 150 . The binding energy strongly change after treatment in SOx; that is, platiꢀ
Eb(S 2p3/2 of the second component acquires a value of num remained in a metal state. Indeed, the oxidation
67.9 or 169.1 eV with the chosen correction or C, of supported platinum particles upon the interaction
respectively. The former value can correspond to of the Pt/BaCO /Al O model catalysts with NOx
p
signal intensity of BaSO4 substantially supported separately onto BaCO3 and CeO2. Neverꢀ
)
did not
°С
)
1
Δ
B
Δ
3
2
3
sulfite ions in the composition of the Baꢀcontaining resulted in a shift of the Pt 4f spectrum to the region of
component (subsystem B), whereas the latter value higher binding energies by 1–1.5 eV [24, 42], which
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

USE OF THE DIFFERENTIAL CHARGING EFFECT IN XPS TO DETERMINE
601
Intensity
Intensity
Ba 4p3/2
v
'''
u
'''
u
v
u''
4
3
2
4
3
v
''
BaSO4
Ce (SO )
2
4 3
u
'
v
'
2
1
1
870
880
890
900
910
920
930
1
55 160 165 170 175 180 185
Binding energy, eV
Binding energy, eV
Fig. 6. XPS spectra of the Pt/(BaCO + CeO ) model catꢀ
Fig. 7. XPS spectra of the Pt/(BaCO + CeO ) model catꢀ
3 2
3
2
alyst measured in the S 2
p
region (
1
) after its reaction with
alyst measured in the Ce 3
d region (1) after its reaction
SO (260 Pa SO + 2600 Pa O + 525 Pa H O) at 300
°
C
with SO (260 Pa SO + 2600 Pa O + 525 Pa H O) at
x
2
2
2
x
2
2
2
and the subsequent regeneration in a mixture of 2100 Pa
CO + 525 Pa H O at ( ) 500, ( ) 600, or ( ) 700 C.
300
°
C and the subsequent regeneration in a mixture of
2
3
4
°
2100 Pa CO + 525 Pa H O at (
2
) 500, ( ) 600, or ( ) 700 C.
3
4
°
2
2
was not the case in the spectra depicted in Fig. 2. It decomposition of barium sulfate occurred to an insigꢀ
should be noted that changes in Eb(Pt 4f7/2 were also nificant degree. The decomposition of cerium sulfate
not observed in a study of the interaction of platinum was accompanied by the appearance of signals characꢀ
)
supported on Al O or CeO2 with SO (SO + O ) in the teristic of the state Ce(IV) in the Ce 3
d
spectrum and a
decrease in the intensity of signals due to Ce(III) (Fig. 7,
spectrum ). Table 2 summarizes the S : Ba and S : Ce
2
3
x
2
2
temperature range of 30–400
°
С
[47].
2
atomic ratios calculated for sulfate ions bound in Ba
and Ce sulfates, respectively, and the contribution (on
a percentage basis) of the Ce(III) state to the Ceꢀconꢀ
taining component. The conversion of Ce(III) into
Ce(IV) under the action of a reducing atmosphere
Regeneration of the Pt/(BaCO + CeO ) Catalyst
3
2
after Sulfatization
Figure 6 shows the S 2p spectra recorded (1) after
the interaction of the Pt/(BaCO + CeO ) catalyst with
3
2
(
CO) seems unusual; nevertheless, this observation
SOx at 300°С and in the course of the subsequent
can be explained taking into account the fact that the
regeneration in a mixture of CO and H O at (2) 500,
2–
2
3+
S
O4 anions can serve as an oxidant for both Ce and
(3
) 600, and (
ous section, barium and cerium sulfates were present
on the catalyst surface after sulfatization. Two S 2 sigꢀ
nals spaced at ~2 eV on the binding energy scale in
spectrum correspond to the above sulfates because of
4
) 700 C. As demonstrated in the previꢀ
°
carbon monoxide:
Ce (SO ) + 2CO
→
2CeO + 3SO + 2CO .
2 2 2
2
4 3
p
The possibility in principle for the occurrence of this
reaction at elevated temperatures follows from the
estimated isobaricꢀisothermal potential of the reaction
determined from thermodynamic parameters [48, 49]
with no regard for the temperature dependence of heat
0
2
the different charging of the Pt/BaCO3 and Pt/CeO2
subsystems. After the treatment of the sulfated catalyst
in a mixture of CO and H O at 500 С, the major porꢀ
°
2
tion of cerium sulfate disappeared, whereas the capacities. At room temperature (298 K), ΔG2
≈
98
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

602
SMIRNOV et al.
Table 2. S : Ba and S : Ce atomic ratios and the percentage concentration of Ce(III) as a constituent of a catalyst sulfated
at 300°C in the course of its regeneration in a mixture of CO and H O at different temperatures
2
2
–
2–
Sample
S
O4 /Ba
S
O4 /Ce
Ce(III), %
15.5
Initial catalyst
–
–
After interaction with SO at 300°C
0.97
0.80
(Ce O )
x
2
3
After interaction with CO + H O
2
at 500
at 600
at 700
°
C
C
0.79
0.41
0.11
0.12
–
34
11
10
°
°
C
–
1
54 kJ/mol, whereas this value becomes negative at (spectra
3and 4), on the contrary, upon the interaction
0
of the catalyst with a mixture of CO + H O at 500 and
2
500
°С
(773 K): ΔG7
≈
–87 kJ/mol.
73
7
00 С, the binding energy Eb(Pt 4f7/2) decreased by
°
An increase in the reaction temperature to 600 and ~0.5 eV, as compared with that of the initial sample, to
then to 700 led to the complete decomposition of acquire a value close to that of the state of bulk platiꢀ
cerium sulfate (Fig. 6, Table 2) and, as judged from the num metal. Thus, in the case under consideration, the
Ce 3 spectrum (Fig. 7), the regeneration of cerium in formation of platinum sulfide under the conditions of
the Ce(IV) state. Barium sulfate is more stable: a conꢀ reducing regeneration is not confirmed. When the
siderable decrease in the intensity of its S 2 signal NSR catalyst contained CeO2 and the treatment of the
occurs in the course of interaction over the temperaꢀ catalyst in SOx led to the formation of cerium(III) sulꢀ
tures range of 600–700 (Fig. 6, Table 2). Starting fate, reducing procedures can lead to the formation of
with a temperature of 600 , an additional signal in Ce(III) oxo sulfide, Ce O S [54]. Spectra
and in
the region of smaller binding energies appeared in the Fig. 7 and data given in the last column of Table 2 indiꢀ
S 2 spectrum. With consideration for the correction cate that, in the case under consideration, regeneraꢀ
for charging, Eb(S 2 was 162.0 eV for the interacꢀ tion almost completely converted cerium into the
or 160.8 eV for 700 . With Ce(IV) state; therefore, the conversion of cerium sulꢀ
consideration for the correction C, Eb(S 2 became fate into sulfide did not occur. Thus, based on the
63.1 eV for 600 or 162.2 eV for 700 . All of the above facts, being it is most reasonable to ascribe the
obtained values of binding energy fall within the range S 2
signal with a binding energy of 161–162 eV, which
°С
d
p
°С
°С
3
4
2
2
p
Δ
B
p)
С
tion temperature of 600
°
°С
Δ
p)
1
°С
p
characteristic of bulk metal sulfides. In particular, the appeared after the reducing regeneration of the catalyst
values of 161.5–162.0 and 163.6 eV were reported for sample at high temperatures, to the barium sulfide BaS.
BaS [50, 51] and PtS [52], respectively. The value of
Eb(S 2p) in the range of 162.2–162.5 eV was obtained
Thus, using the Pt/(BaCO + CeO ) model NSR
3
2
catalyst as an example, we demonstrated the applicaꢀ
bility of the differential charging effect in the measureꢀ
ment of Xꢀray photoelectron spectra to the identificaꢀ
tion of compounds formed on the surface of a multiꢀ
component catalyst upon its interaction with a
reaction atmosphere. We found that barium sulfate
was predominantly formed in the reaction of the cataꢀ
for surface platinum sulfide, which was formed upon
the interaction of supported platinum catalysts with
SO2 or H2S [47, 53]. Thus, we can convincingly state
that sulfides are formed in the process of the reducing
regeneration of the sulfated catalyst. However, at the
same time, it is very difficult to unambiguously idenꢀ
tify the resulting sulfides (BaS, PtS) based on only the
lyst with SO (SO + O + H O) at 150
in the reaction temperature to 300 С
°
С
. An increase
resulted in the
additional formation of cerium sulfate, which was
x
2
2
2
binding energies Eb(S 2p) even taking into account difꢀ
ferential charging.
°
Previously, it was reported that the conversion of accompanied by the reduction of Ce(IV) to Ce(III).
sulfates into barium sulfide occurs in the process of the Supported platinum particles retained their metal
reducing regeneration of NSR catalysts [8, 10, 12, 13]. state in the course of reaction with SOx. The regeneraꢀ
At the same time, in a number of cases, the formation tion of the catalyst sulfated at 300
°С in a mixture of
of platinum sulfide was also assumed [4, 10, 14, 15]. It CO and H O caused the decomposition of cerium(III)
2
is well known that the conversion of platinum into the sulfate at temperatures of
≤
500 С. In spite of the presꢀ
°
PtS sulfide is accompanied by an increase in the bindꢀ ence of a strong reducing agent (CO) in the reaction
ing energy Eb(Pt 4f7/2 by ~1.5 eV, as compared with atmosphere, the decomposition of cerium sulfate was
that of platinum metal [53]. As can be seen in Fig. 2 accompanied by the oxidation of surface Ce(III) catꢀ
)
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

USE OF THE DIFFERENTIAL CHARGING EFFECT IN XPS TO DETERMINE
603
ions to CeO2. At treatment temperatures of
the major portion of barium sulfate was retained. The
≤
500
°
С
,
22. Corbos, E.C., Courtois, X., Bion, N., Marecot, P., and
Duprez, D., Appl. Catal., B, 2008, vol. 80, p. 62.
decomposition of BaSO4 was observed in the temperaꢀ 23. Phatak, A.A., Koryabkina, N., Rai, S., Ratts, J.L.,
tures range of 600–700
the formation of an amount of the sulfide BaS.
°
С
, and it was accompanied by
Ruettinger, W., Farrauto, R.J., Blau, G.E., Deglass, W.N.,
and Ribeiro, F.H., Catal. Today, 2007, vol. 123, p. 234.
2
4. Smirnov, M.Yu., Kalinkin, A.V., Dubkov, A.A.,
Vovk, E.I., Sorokin, A.M., Nizovskii, A.I., Carberry, B.,
and Bukhtiyarov, V.I., Kinet. Katal., 2008, vol. 49,
p. 876 [Kinet. Catal. (Engl. Transl.), vol. 49, p. 831].
REFERENCES
1. Matsumoto, S., Cattech, 2000, vol. 4, p. 102.
2
5. Bukhtiyarov, V.I., Prosvirin, I.P., and Kvon, R.I.,
2
. Epling, W.S., Campbell, L.E., Yezerets, A.,
Currier, N.W., and Parks, J.E., Catal. Rev., 2004,
vol. 46, p. 163.
J. Electron Spectrosc. Relat. Phenom., 1996, vol. 77, p. 7.
26. Bukhtiyarov, V.I., Prosvirin, I.P., Kvon, R.I., Gonꢀ
charova, S.N., and Bal’zhinimaev, B.S., J. Chem. Soc.,
Faraday Trans., 1997, vol. 93, p. 2323.
3
4
5
6
. Breen, J.P., Marella, M., Pistarino, C., and
Ross, J.R.H., Catal. Lett., 2002, vol. 80, p. 123.
2
2
2
3
3
3
7. Jaramillo, A., Spurlock, L.D., Young, V., and Brajterꢀ
. Sedlmair, C., Seshan, K., Jentis, A., and Lercher, J.A.,
Toth, A., Analyst, 1999, vol. 124, p. 1215.
Catal. Today, 2002, vol. 75, p. 413.
8. Christie, A.B., Lee, J., Sutherland, I., and Walls, J.M.,
. Liu, Z. and Anderson, J.A., J. Catal., 2004, vol. 228,
Appl. Surf. Sci., 1983, vol. 15, p. 224.
p. 243.
9. Teterin, Yu.A. and Sosulnikov, M.I., Physica C, 1993,
. Rohr, F., Peter, S.D., Lox, E., Kogel, M., Sassi, A.,
Juste, L., Rigaudeau, C., Belot, G., Gelin, P., and
Primet, M., Appl. Catal., B, 2005, vol. 56, p. 201.
vol. 212, p. 306.
0. Miot, C., Husson, E., Proust, C., Erre, R., and Couꢀ
tures, J.P., J. Mater. Res., 1997, vol. 12, p. 2388.
7. Wei, X., Liu, X., and Deeba, M., Appl. Catal., B, 2005,
1. Mullins, D.R., Overbury, S.H., and Huntley, D.R.,
vol. 58, p. 41.
Surf. Sci., 1998, vol. 409, p. 307.
8
. Elbouazzaoui, S., Corbos, E.C., Courtois, X., Mareꢀ
cot, P., and Duprez, D., Appl. Catal., B, 2005, vol. 61,
p. 236.
2. Romeo, M., Bak, K., El Fallah, J., Le Normand, F.,
and Hilaire, L., Surf. Interface Anal., 1993, vol. 20,
p. 508.
9. Anderson, J.A., Liu, Z., and Garcia, M.F., Catal.
3
3
3
3. Hardacre, H., Roe, G.M., and Lambert, R.M., Surf.
Today, 2006, vol. 113, p. 25.
Sci., 1995, vol. 326, p. 1.
1
0. Abdulhamid, H., Fridell, E., Dawody, J., and Skoꢀ
4. Park, P.W. and Ledford, J.S., Langmuir, 1996, vol. 12,
glundh, M., J. Catal., 2006, vol. 241, p. 200.
p. 1794.
1
1. Strehlau, W., Leyrer, J., Lox, E.S., Kreuzer, T.,
5. Moulder, J.F., Stickle, W.F., Sobol, P.E., and
Bomben, K.D., Handbook of XꢀRay Photoelectron
Spectroscopy, Eden Prairie, Minn.: PerkinElmer, 1992.
Hori, M., and Hoffmann, M., SAE Technical Paper
,
1996, no. 962047.
1
2. Poulston, S. and Rajaram, R.R., Catal. Today, 2003,
3
6. Nefedov, V.N., Rentgenoelektronnaya spektroskopiya
khimicheskikh soedinenii (XꢀRay Photoelectron Specꢀ
troscopy of Chemical Compounds), Moscow: Khimiya,
vol. 81, p. 603.
1
3. Kim, D.H., Scanyi, J., Kwak, J.H., Szailer, T.,
Hanson, J., Wang, C.M., and Peden, C.H.F., J. Phys.
Chem. B, 2006, vol. 110, p. 10 441.
1984.
37. Smirnov, M.Yu. and Graham, G.W., Catal. Lett., 2001,
1
4. Limousy, L., Mahzoul, H., Brilhae, J.F., Garin, F.,
Maire, G., and Gilot, P., Appl. Catal., B, 2003, vol. 45,
p. 169.
vol. 72, p. 39.
3
8. Smirnov, M.Yu., Kalinkin, A.V., Pashis, A.V.,
Sorokin, A.M., Noskov, A.S., Kharas, K.C., and
Bukhtiyarov, V.I., J. Phys. Chem. B, 2005, vol. 109,
p. 11712.
1
5. Fridell, E., Amberntsson, A., Olsson, L., Grant, A.W.,
and Skoglundh, M., Top. Catal., 2004, vols. 30–31,
p. 143.
39. Huizinga, T., van’t Blik, H.F.J., Vis, J.C., and Prins, R.,
1
1
1
1
2
2
6. Ji, Y., Toops, T.J., and Crocker, M., Catal. Lett., 2009,
Surf. Sci., 1983, vol. 135, p. 580.
vol. 127, p. 55.
4
0. Parmigiani, F., Kay, E., Bagus, P.S., and Nelin, C.J.,
J. Electron Spectrosc. Relat. Phenom., 1985, vol. 36,
p. 257.
7. Matsumoto, S., Ikeda, Y., Suzuki, H., Ogai, M., and
Miyoshi, N., Appl. Catal., B, 2000, vol. 25, p. 115.
8. Yamazaki, K., Suzuki, T., Takahashi, N., Yokota, K., 41. Steinrück, H.ꢀP., Pesty, F., Zhang, L., and Madey, T.E.,
and Sugiura, M., Appl. Catal., B, 2001, vol. 30, p. 459.
Phys. Rev. B: Condens. Matter, 1995, vol. 51, p. 2427.
9. Fanson, P.T., Horton, M.R., Deglass, W.N., and Lautꢀ 42. Smirnov, M.Yu., Kalinkin, A.V., and Bukhtiyarov, V.I.,
erbach, J., Appl. Catal., B, 2003, vol. 46, p. 393.
Zh. Strukt. Khim., 2007, vol. 48, p. 1120 [J. Struct.
Chem. (Engl. Transl.), vol. 48, p. 1053].
0. Basile, F., Fornasari, G., Grimandi, A., Livi, M., and
Vaccari, A., Appl. Catal., B, 2006, vol. 69, p. 58.
43. Burroughs, P., Hamnett, A., Orchard, A.F., and Thornꢀ
ton, G., J. Chem. Soc., Dalton Trans., 1976, p. 1686.
1. Corbos, E.C., Elbouazzaoui, S., Courtois, X.,
Bion, N., Marecot, P., and Duprez, D., Top. Catal.
007, vols. 42–43, p. 9.
,
44. Fujimori, A., Phys. Rev. B: Condens. Matter, 1983,
2
vol. 28, p. 2281.
KINETICS AND CATALYSIS Vol. 52
No. 4
2011

604
SMIRNOV et al.
4
5. Shyu, J.Z., Weber, W.H., and Gandhi, H.S., J. Phys. 50. Rohr, F., Gobel, U., Kattwinkel, P., Kreuzer, T.,
Chem., 1988, vol. 92, p. 4964.
Muller, W., Philipp, S., and Gelin, P., Appl. Catal., B
007, vol. 70, p. 189.
,
2
46. Paparazzo, E., Surf. Sci., 1990, vol. 234, p. L253.
4
7. Smirnov, M.Yu., Kalinkin, A.V., Pashis, A.V., 51. Stakheev, A.Yu., Gabrielsson, P., Gekas, I.,
Sorokin, A.M., Noskov, A.S., Bukhtiyarov, V.I., Khaꢀ
ras, K.S., and Rodkin, M.A., Kinet. Katal., 2003, vol. 44,
p. 629 [Kinet. Catal. (Engl. Transl.), vol. 44, p. 575].
Teleguina, N.S., Bragina, G.O., Tolkachev, N.N., and
Baeva, G.N., Top. Catal., 2007, vols. 42–43, p. 143.
5
2. Visser, J.P.R., Groot, C.K., van Oers, E.M., de
48. Svoistva neorganicheskikh soedinenii: Spravochnik
Beer, V.H.J., and Prins, R., Bull. Soc. Chim. Belg.
984, vol. 93, p. 813.
,
(
Properties of Inorganic Compounds: A Handbook),
1
Efimov, A.I., Belorukova, L.P., Vasil’kova, I.V., and
Chechev, V.P., Eds., Leningrad: Khimiya, 1983.
5
3. Wang, T., Vazques, A., Kato, A., and Schmidt, L.D.,
J. Catal., 1982, vol. 78, p. 306.
49. Khimicheskaya entsiklopediya (Encyclopedia of Chemꢀ
istry), Moscow: Bol’shaya Rossiiskaya Entsiklopediya, 54. Luo, T., Vohs, J.M., and Gorte, R.J., J. Catal., 2002,
1995, vol. 4.
vol. 210, p. 397.
KINETICS AND CATALYSIS Vol. 52
No. 4
2011