Interaction of Al2O3 and CeO2 Surfaces with so2 and so2 + O2 studied by X-ray photoelectron spectroscopy

Article abstract of DOI:10.1021/jp0508249

The interaction of Al₂O₃ and CeO₂ thin films with sulfur dioxide (2.5 mbar) or with mixtures of SO₂ with O₂ (5 mbar) at various temperatures (30-400 ?°C) was studied by X-ray photoelectron spectroscopy (XPS). The analysis of temperature-induced transformations of S2p spectra allowed us to identify sulfite and sulfate species and determine the conditions of their formation on the oxide surfaces. Sulfite ions, SO₃^2-, which are characterized by the S2p_3/2 binding energy (BE) of a??167.5 eV, were shown to be formed during the interaction of the oxide films with pure SO₂ at temperatures a??200 ?°C, whereas sulfate ions, SO₄^2-, with BE (S2p_3/2) a??169 eV were produced at temperatures a?￥300 ?°C. The formation of both the sulfite and sulfate species proceeds more efficiently in the case of CeO₂. The addition of oxygen to SO ₂ suppresses the formation of the sulfite species on both oxides and facilitates the formation of the sulfate species. Again, this enhancement is more significant for the CeO₂ film than for the Al₂O ₃ one. The sulfation of the CeO₂ film is accompanied by a reduction of Ce(IV) ions to Ce(III) ones, both in the absence and in the presence of oxygen. It has been concluded that the amount of the sulfates on the CeO₂ surface treated with the SO₂ + O₂ mixture at a?￥300 ?°C corresponds to the formation of a 3D phase of the Ce(III) sulfate. The sulfation of Al₂O₃ is limited by the surface of the oxide film. ? 2005 American Chemical Society.

Full text of DOI:10.1021/jp0508249

11712
J. Phys. Chem. B 2005, 109, 11712-11719
Interaction of Al2O3 and CeO2 Surfaces with SO2 and SO2 + O2 Studied by X-ray
Photoelectron Spectroscopy
Mikhail Yu. Smirnov,* Alexander V. Kalinkin, Andrei V. Pashis, Alexei M. Sorokin,
†
Alexander S. Noskov, Karl C. Kharas, and Valerii I. Bukhtiyarov
BoreskoV Institute of Catalysis SB RAS, LaVrentieVa aVe, 5, NoVosibirsk 630090, Russia, and Delphi,
Tulsa, Oklahoma 74158-0970
ReceiVed: February 16, 2005; In Final Form: April 22, 2005
The interaction of Al
5 mbar) at various temperatures (30-400 °C) was studied by X-ray photoelectron spectroscopy (XPS). The
analysis of temperature-induced transformations of S2p spectra allowed us to identify sulfite and sulfate species
2 3 2 2 2
O and CeO thin films with sulfur dioxide (2.5 mbar) or with mixtures of SO with O
(
2
-
and determine the conditions of their formation on the oxide surfaces. Sulfite ions, SO
characterized by the S2p3/2 binding energy (BE) of ∼167.5 eV, were shown to be formed during the interaction
of the oxide films with pure SO at temperatures e200 °C, whereas sulfate ions, SO , with BE (S2p3/2
169 eV were produced at temperatures g300 °C. The formation of both the sulfite and sulfate species
proceeds more efficiently in the case of CeO . The addition of oxygen to SO suppresses the formation of the
sulfite species on both oxides and facilitates the formation of the sulfate species. Again, this enhancement is
more significant for the CeO film than for the Al one. The sulfation of the CeO film is accompanied by
a reduction of Ce(IV) ions to Ce(III) ones, both in the absence and in the presence of oxygen. It has been
concluded that the amount of the sulfates on the CeO surface treated with the SO + O mixture at g300 °C
is limited by the
3
, which are
2
-
2
4
)
∼
2
2
2
O
2 3
2
2
2
2
corresponds to the formation of a 3D phase of the Ce(III) sulfate. The sulfation of Al
surface of the oxide film.
2 3
O
1
. Introduction
of 600 °C and above.2,3,5 The oxidation of the sulfite species
with oxygen or the treatment of alumina with SO2 + O2 mixtures
at T ≈ 400-700 °C have been shown to result in the formation
Studying the interaction of SO2 with surfaces of alumina and
ceria is of importance, first of all, because of their catalytic
application. High surface area Al2O3 is used as a catalyst for
the Claus process in which SO2 is reduced to sulfur by hydrogen
sulfide. Furthermore, both Al2O3 and CeO2 are the main
components of monolith catalysts for the purification of exhaust
gases emitted from car engines. Sulfur dioxide, which is one
of the pollutants released to the atmosphere as a result of fuel
combustion, can react with components of the emission control
catalysts, reducing the activity of the catalysts in CO, NOx, and
2
,3,6,8,11-14
of sulfate species.
In contrast, a concurrent formation
of the sulfite and sulfate species on γ-Al2O3 after its interaction
with SO2 at 80 °C was reported on the basis of X-ray
1
5
photoelectron spectroscopy (XPS) data. The amounts of the
sulfite and sulfate species produced under the interaction of SO2
14
-2 7,10,13,16
or SO2 + O2 with alumina did not exceed 2 × 10 cm ,
that is, the interaction with Al2O3 was most probably limited
only by the surface area. Different structures with various types
of coordination have been proposed for sulfite and sulfate
1
hydrocarbon abatement. Therefore, the interaction of SO2 with
Al2O3 and CeO2 surfaces has been studied widely in the past.
1
1,12,17-19
species on the surface of Al2O3.
2-25
The interaction of SO2 with CeO2 at elevated temperatures
and the mbar SO2 partial pressure range has also been studied
However, the results of these investigations are often contradic-
tory. These contradictions concern, first of all, the nature and
properties of reaction products and, second, the conditions of
their appearance.
It has been shown that the adsorption of SO2 on Al2O3
surfaces in a wide range of temperatures (-100 to 500 °C) and
partial pressures of g10 mbar results in the formation of several
weakly and strongly bonded species, which, on the basis of
1
0,20-22
22
extensively.
As concluded from gravimetric and el-
1
0
emental analysis measurements, the accumulation of sulfur
species occurs mainly in the form of sulfite and hydrosulfite
1
4
2
and does not exceed ∼5 × 10 S atoms per cm , when the
adsorption proceeds in the temperature range of 30-350 °C.
2
2
According to an IRS study, the treatment of sulfite in a large
excess of oxygen at 400 °C leads to the formation of both
surface and bulk sulfates. The heating of the sulfated ceria up
to 600 °C results in decomposition of the bulk sulfate, whereas
infrared (IR) spectra, have been identified as sulfite species
SO32
- 2-8
.
It has been established that the sulfite species are
4
adsorbed on basic sites, which, for instance, can be terminal
O
higher the strength of the Lewis basicity of an oxide the greater
the amount of SO2 strongly held in the form of sulfite
species.
2
2
2
-
the surface sulfate persists even at 700 °C. At the same time,
XPS studies performed in UHV have revealed that sulfite and
even sulfate species were produced under SO2 adsorption on a
ions on the alumina surface. It has been found that the
23-25
7
-10
CeO2 film at room temperature or below.
The total amount
The sulfite species were found to decompose
of the sulfite and sulfate species accumulated did not exceed a
completely after annealing the sample in vacuum at temperatures
monolayer (ML) after a dose of SO2 in 5 L (1 L (Langmuir) )
-
6
1
.3 × 10 mbar × sec) and became larger than 1 ML after
*
Corresponding author. E-mail: smirnov@catalysis.nsk.su.
Delphi.
†
24
treatment in 1.7 mbar of SO2 at 30 °C for 5 min. It has been
1
0.1021/jp0508249 CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/20/2005

Interaction of Al2O3 and CeO2 Surfaces
J. Phys. Chem. B, Vol. 109, No. 23, 2005 11713
Figure 1. The S2p photoemission spectra recorded after the treatment of the alumina (a) and ceria (b) films in 2.5 mbar SO
temperatures indicated. For the case of Al , the spectra are multiplied by factor 2.
2
for 20 min at the
2 3
O
concluded that the interaction of SO2 with CeO2 does not change
the oxidation state of cerium Ce(IV), and, therefore, the
formation of SO42 was assumed to take place via dispropor-
tionation of SO2 into Ce(IV) sulfate and SO.2 In contrast, sulfate
formation detected by XPS/TPD measurements under treatment
of a CeO2 film in mixture SO2 + O2 (1:1) at total pressure of
from XPS spectra. Finally, the alumina foil was annealed for
-
4
10 min at P(O2) ) 10 mbar and T ) 600 °C. Afterward, the
binding energies and intensity ratio of Al2p and O1s lines were
found to correspond to the Al2O3 composition.
-
4
A thin film of CeO was obtained via subsequent UHV
2
evaporation of metallic cerium over the Al O film followed
2
3
-5
-5
8
× 10 mbar and 25 °C was accompanied by partial reduction
by oxidation at P(O ) ) 2.5 × 10 mbar and T ) 600 °C. The
2
25
of Ce(IV) into Ce(III). The sulfite species have a moderate
thermal stability and decompose releasing SO2,g in the temper-
ature range from -70 to 330 °C. The temperature interval of
the sulfate species decomposition determined in different works
varies from -70 to 330 °C24 to 530-730 °C.
role of the intermediate alumina film was to prevent diffusion
of very mobile oxygen from cerium oxide into the tantalum
foil, thereby maintaining the stoichiometry of the CeO film.
2
As a consequence, Ce3d, Ce4d, and O1s lines taken from the
25
film were typical of CeO . Again, repetitive evaporation cycles
2
As evidenced by the foregoing survey, the results obtained
in UHV and in the millibar pressure range led to the conflicting
conclusions concerning the nature of the products and their
thermal stability. One can suggest that these contradictions are
most probably determined by the effect of temperature, pressure,
and the content of the gas phase on the chemical nature of sulfur-
containing species produced during reactions of SO2 with Al2O3
or with CeO2. To bridge the UHV and real catalytic conditions
and to obtain more information about the sulfation mechanisms
of oxide surfaces, a comparative XPS study of the interaction
of SO2 and SO2 + O2 with Al2O3 and CeO2 was undertaken in
this work.
were applied to produce the desirable thickness of the ceria
films. Based on the attenuation of the Al2p intensity, the
thickness of the CeO2 film was estimated at about 20-30 Å.
The treatments in SO (2.5 mbar) or SO + O (2.5 mbar +
2
2
2
2.5 mbar) were performed in a preparation chamber of the
spectrometer for 20 min at various temperatures: 30, 100, 200,
300, and 400 °C. The sample was heated resistively by passing
a current through the tantalum foil. The temperature of the
sample was measured with a chromel-alumel thermocouple
spot-welded to the Ta foil. After cooling to room temperature
and evacuation of the reaction mixture, the sample was
transferred into the analyzer chamber, where survey XP spectra
as well as narrow spectra of Al2p, S2p, O1s, and Ce3d regions
were measured. The binding energies and intensities were
calculated after subtraction of a Shirley-type background from
raw photoemission spectra. Complex spectra in the Ce3d region
were resolved into eight components: six for the Ce(IV) and
two for the Ce(III) oxidation states (see below), according to a
2
. Experimental Section
All of the experiments were performed using a VG ESCA 3
-
9
spectrometer with the base pressure <5 × 10 mbar. X-ray
photoelectron spectra were taken using nonmonochromatized
Al KR irradiation (hν ) 1486.6 eV). Before taking measure-
ments, we calibrated the spectrometer against the Au4f7/2 binding
energy (BE) of 84.0 eV and the Cu2p3/2 BE of 932.6 eV. The
original data acquisition system was used for XPS signal
detection and control of the sample temperature.26
2
7
fitting procedure described elsewhere. The BE (Al2p) )
7
4.5 eV and BE (Ce3d u′′′) ) 916.7 eV were used as references
2
7,28
for the cases of Al2O3 and CeO2, respectively.
The atomic
ratios of S/Al and S/Ce were calculated from S2p, Al2p, and
Ce3d areas using the corresponding atomic sensitivity factors
A thin film of Al2O3 was prepared via evaporation of metallic
2
9
(ASF) taken in.
-
4
aluminum in the presence of oxygen (P(O2) ) 10 mbar) on
a clean surface of a tantalum foil. The necessary thickness of
the alumina film was achieved via a number of repetitive
evaporation cycles and was larger than ∼100 Å. This estimation
was based on the complete disappearance of the tantalum lines
3
. Results
Figure 1 presents two sets of S2p spectra taken from Al2O3
(a) and CeO2 (b) films after their interaction with sulfur dioxide

11714 J. Phys. Chem. B, Vol. 109, No. 23, 2005
Smirnov et al.
TABLE 1: Values of S2p3/2 BE for Various
in Figures 1 and 2 indicates that the addition of O2 to an SO2
flow changes the composition of the overlayers. This conclusion
also follows from the data of Table 3, which shows S2p3/2
binding energies and S/Al, S/Ce atomic ratios determined from
the spectra of Figure 2. First of all, suppression of the sulfite
formation for both oxides at T < 200 °C should be mentioned.
One can see that the intensity of the S2p spectra measured from
the Al2O3 surface at T e 200 °C is comparable with the XPS
sensitivity limit (Figure 2a). Higher values of the S2p3/2 binding
energies registered in this temperature range, 168.6 eV (Table
Sulfur-Containing Compounds and Adsorbed Species
sulfides
sulfur
sulfites
sulfates
Na2S PtS
S Na2SO3 PbSO3 Na2SO4 PbSO4 Fe2(SO4)3
BE, eV 161.8 163.6 ∼164 166.5 167.8 168.7 169.4
ref 30 31 30, 32 30 33 30 33
169.1
30
TABLE 2: S2p3/2 Binding Energies and S/Al and S/Ce
Atomic Ratios Measured after the Treatment of Al
CeO with SO at Various Temperatures
2
O
3
and
2
2
BE, eV
S/Al and S/Ce atomic ratios
3
), compared to those typical of sulfite (167.5-167.7 eV) is
temperature,
another difference of the SO2 + O2 mixture treatment. This shift
can be explained by the formation of the sulfate ions admixed
with the sulfite species. Unfortunately, the low intensity of S2p
signals does not allow us to carry out a more detailed analysis
of these spectra. Similar to the alumina surface, the interaction
of CeO2 with the SO2 + O2 mixture decreases the amount of
sulfite produced at room temperature and shifts the sulfite-
originated S2p feature to higher binding energies at 100 °C
(Figure 2b and Table 3). The complete transformation of the
sulfite overlayer to the sulfate one at 200 °C seems to prove
that the intermediate position of the S2p spectra at lower
temperatures originates from coadsorption of the sulfite- and
sulfate-originated features.
°
C
Al
2
O
3
CeO
2
Al CeO
2
O
3
2
3
0
167.6
167.8
167.7
168.4
169.2
167.6
167.4
167.7
168.9
169.2
0.041
0.038
0.032
0.032
0.046
0.13
0.18
0.19
0.19
0.22
1
2
3
4
00
00
00
00
TABLE 3: S2p3/2 Binding Energies and S/Al and S/Ce
Atomic Ratios Measured after the Treatment of Al
CeO with SO + O at Various Temperatures
2 3
O and
2
2
2
BE, eV S/Al and S/Ce atomic ratios
temperature,
°
C
Al
2
O
3
CeO
2
Al
2
O
3
CeO
2
A more pronounced accumulation of the sulfates at T >
3
0
168.6
168.6
168.6
169.2
169.2
167.7
168.1
168.9
169.1
169.1
0.012
0.017
0.016
0.036
0.060
0.063
0.072
0.19
0.29
0.45
200 °C (Figure 2b, Table 3) is another result of the O2 influence
1
2
3
4
00
00
00
00
on the SO2 interaction with the alumina and ceria surfaces. One
can see that for both oxides, the sulfate yield enhances with
the reaction temperature (Tables 2 and 3). It is also noteworthy
that efficient sulfate formation on the ceria film occurs at a
temperature about 100 °C below that of the reaction with pure
SO2.
(
see Experimental Section) at various temperatures. One can
see that the treatment of the oxides by SO2 results in an
appearance of S2p features, both their intensities and positions
being dependent on temperature. The asymmetrical shape of
the features is explained by the spin-orbital splitting of the
S2p spectrum on two components, S2p3/2 and S2p1/2. To analyze
these spectra in more detail and to apply literature data for
comparison, we deconvoluted the S2p lines into these compo-
nents. As seen from Table 1, which lists literature data for
various sulfur-containing compounds, namely, the S2p3/2, bind-
ing energies were usually applied to elucidate the chemical state
of sulfur. Despite a structureless shape of the S2p spectra, such
deconvolution is straightforward because both the spin-orbital
splitting of 1.2 eV and the component ratio S2p3/2/S2p1/2 ) 2:1
are well known from previous papers, which have applied
synchrotron radiation as a source of X-rays.24
Figure 3 compares the yields of the sulfate species produced
in the reactions of the Al2O3 and CeO2 with SO2 and SO2 + O2
at various temperatures. The sulfate yields were expressed as
S/Al and S/Ce atomic ratios calculated from the intensities of
the corresponding photoemission lines. One can see that the
formation of the sulfate species occurs more efficiently on the
CeO2 film. The diagrams in Figure 3 also show that the
efficiency of sulfation increases with temperature. In the
presence of oxygen, the sulfation emerges at a lower temperature
and proceeds with a larger yield of sulfate species. The influence
of oxygen and temperature on the sulfation is more pronounced
in the case of CeO2.
An essential difference in the sulfation of Al O and CeO is
2
3
2
also exhibited in the behavior of the Al2p and Ce3d lines. The
parameters of the Al2p line change negligibly under the reaction,
indicating the permanency in the chemical state of aluminum
ions. Contrary to alumina, a part of the Ce(IV) ions transform
into the Ce(III) ions as a result of the interaction of the CeO2
surface with pure SO2 or with the SO2 + O2 mixture. This
conclusion is based on the data of Figure 4, which demonstrates
the variation of the Ce3d spectra after the treatment of CeO2
with SO2 (a) or with SO2 + O2 (b) at various temperatures.
The Ce3d spectra of the initial CeO2 surfaces are also shown
for comparison.
The S2p3/2 binding energies as well as the S/Al and S/Ce
atomic ratios determined from the spectra of Figure 1 are
collected in Table 2. One can see that similar behaviors of the
S2p spectra are observed for both oxides. Being located at
1
67.5 ( 0.2 eV at temperatures below 200 °C, the S2p3/2
features are shifted to higher BE as the sample temperature rises
to 300 °C and achieve the BE value of 169.2 eV at 400 °C
(Figure 1, Table 2). The comparison with literature data (Table
1
) allows unambiguous assignment of these S2p3/2 features to
sulfite (∼167.5 eV) and sulfate (∼169.2 eV) species. The data
presented in Table 2 also show that the CeO2 surface contains
a higher amount of sulfur-containing species in both temperature
ranges. Furthermore, the abundance of sulfur on the Al2O3
surface is weakly dependent on the reaction temperature,
whereas for CeO2, the abundance of sulfur has a tendency to
increase with the reaction temperature.
The S2p3/2 spectra taken from the Al2O3 (a) and CeO2 (b)
surfaces after their interaction with SO2 + O2 mixtures at various
temperatures are shown in Figure 2. A comparison of the spectra
As mentioned in the Experimental Section, the Ce3d spectra
have a complex structure composed of three Ce3d5/2/Ce3d3/2
doublets from Ce(IV) ions and one doublet from Ce(III)
2
7,28,34
ones.
The Ce3d5/2 lines, characteristic of the Ce(IV)
oxidation state, are located at 882.4, 889.0, and 898.3 eV.
2
7,28
Following convention,
these features are designated as V,
V′′, and V′′′. The corresponding u, u′′, and u′′′ features of the
Ce3d3/2 lines appear at 900.8, 907.4, and 916.7 eV. This

Interaction of Al2O3 and CeO2 Surfaces
J. Phys. Chem. B, Vol. 109, No. 23, 2005 11715
Figure 2. The S2p photoemission spectra recorded after the treatment of the alumina (a) and ceria (b) films in 2.5 mbar SO
0 min at the temperatures indicated. For the case of Al , the spectra are multiplied by factor 2.
2
2
+ 2.5 mbar O for
2
2 3
O
their deconvolution into the eight above-mentioned components
belonging to the Ce(IV) and Ce(III) states. The fraction of
Ce(III) ions in the CeO2 films was estimated as a ratio of the
total intensity of the V′ and u′ components to the total intensity
under the entire Ce3d contour. This estimation indicated that
the initial ceria films, before the treatment in SO2-containing
mixtures, comprises about 15% of Ce(III) ions. The XPS
detection of the Ce(III) oxidation state in the clean and
preoxidized ceria samples has been reported frequently in
2
7,36-38
literature.
There are several reasons for the observation
of Ce ions in the initial CeO2 sample. First, the ceria surface
3
+
39
can be reduced under the influence of X-rays. Second, because
our CeO2 films were deposited on the alumina surface, their
annealing in a reducing medium (see Experimental Section) can
transform the ceria/alumina interface into very stable cerium
(
III) aluminate CeAlO3, which can be reoxidized successfully
Figure 3. The S/Al and S/Ce atomic ratios calculated for the sulfate
species produced on the alumina and ceria films treated in 2.5 mbar
3
6-38
in air only at very high temperatures (>600 °C).
reactions of CeO2 with such background gases as CO and H2
can also cause the reduction of the Ce ions. The detailed
analysis of the reasons was not made in this work because the
Finally,
2 2 2
SO or the mixture of 2.5 mbar SO and 2.5 mbar O at the temperatures
4
+
indicated.
complicated structure of the Ce3d spectrum is determined by
the electronic structure of cerium ions in the initial and final
shape of the initial spectrum (Figure 4) was close to CeO and
the determined fraction of the Ce(III) oxidation state (15%) did
not exceed the value reported for ceria in the literature.
2
4
+
states. As a result of a strong hybridization between the Ce
and O2 valence orbitals, the initial electronic state of cerium
-
Analysis of the Ce3d spectra shown in Figure 4 indicates
that the CeO2 surface is reduced during the reaction with SO2,
with the extent of the reduction being increased gradually with
temperature. Addition of O2 to the SO2 flow enhances signifi-
cantly the degree of reduction when the temperature is 300 °C
or above, coinciding with the formation of the sulfate species.
Moreover, the comparison of the S2p (Figures 1b and 2b) and
Ce3d (Figure 4) spectra allows us to conclude that the sulfation
and reduction of ceria proceed in parallel. To confirm this
conclusion quantitatively, we plotted the amount of the sulfate
produced under the reaction of the CeO2 film with SO2 + O2
mixtures in Figure 5 as a function of the Ce(III) fraction, which
has been corrected for the Ce(III) content in the original CeO2
films. This figure reveals a linear relationship between these
two parameters. An attempt to plot similar parameters for the
reaction with pure SO2 led to less obvious dependence because
0
1 -1
ions in CeO2 is a mixture 4f and 4f L configurations, where
-1
34
L
denotes a hole in the valence band. According to literature
data,3 the V′′′/u′′′ doublet corresponds to the photoionization
5,36
0
into the 4f final state, whereas two other doublets, V′′/u′′ and
1
2
V/u, correspond to the final state in which the 4f and 4f
configurations are mixed. The Ce3d5/2/Ce3d3/2 lines at 885.6
and 904.0 eV designated in Figure 4 as V′ and u′ features arise
from the Ce3 ions. In general, the spectrum of the pure
+
27
Ce(III) oxidation state is characterized additionally by another
doublet, designated as V0/u0. However, the position and intensity
of these minor lines allows us to exclude them from consider-
ation. Indeed, their addition to the V′/u′ lines leads only to a
slight broadening and asymmetry of the entire Ce3d spectrum
from Ce(III) ions. Therefore, the contribution of the V0/u0 doublet
into the Ce3d spectra is not taken into account in this work. As
a consequence, Figure 4 shows the Ce3d spectra together with

11716 J. Phys. Chem. B, Vol. 109, No. 23, 2005
Smirnov et al.
Figure 4. The Ce3d photoemission spectra recorded for the CeO
2
film treated in 2.5 mbar SO
2
2 2
(a) or 2.5 mbar SO and 2.5 mbar O (b) at the
temperatures indicated. The bottom spectra in panels a and b characterize the initial state of the CeO
2
film before the treatments.
reducible ceria suggest different mechanisms of the formation
of the sulfate species on these oxides.
2
-
4
.1. SO3 Formation. Sulfite-like species are the major
product when pure SO2 interacts with alumina or ceria surfaces
at 200 °C or below. This conclusion is based on the appearance
of the S2p features with the S2p3/2 binding energies of
∼
167.6 eV that is much less than the value 169.2 eV observed
at higher temperatures (Figure 1). It is obvious that the highest
value of the binding energies arises from the sulfate-like species
(Table 1). The most probable mechanism of the sulfite formation
on an oxide surface is a molecular adsorption of SO2 on Lewis
2
-
4
basic sites, such as Os surface anions, via the following
surface reaction:
SO + Os f SO32-
2-
(1)
2
However, our study contains some experimental results that
cannot be rationalized in the terms of this simple model. Among
them are the much lower amount of the sulfite species on the
Al2O3 surface as compared with the CeO2 surface (Table 2)
and the suppression of the sulfite formation under the influence
of O2 (Table 3). A more detailed analysis of the literature allows
us to propose some reasonable explanations of these facts.
First of all, it should be taken into account that according to
literature, adsorbed SO2 on the surfaces of oxides can exist in
more than one state. Even in 1978, Chang characterized two
species formed on the γ-alumina surface at room temperature
Figure 5. The yield of sulfate produced under the reaction of the CeO
film with 2.5 mbar SO + 2.5 mbar O at various temperatures as a
function of the Ce(III) fraction corrected for the Ce(III) content in the
2
2
2
2
initial CeO film. The yield of sulfate is evaluated by the S/Ce atomic
ratio.
the yield of the sulfate per one Ce(III) ion is much smaller in
this case (see reactions 5 and 7 in Section 4).
-
1
-1 2
by IR bands at 1060 cm and 1326-1140 cm . Karge and
Dalla Lana showed by selective poisoning with Lewis acid BF3
that the first species, called sulfite species, were chemisorbed
4. Discussion
2
-
4
The systematic XPS study of the temperature-dependent
on O basic sites. They also revealed that the IR bands at
-
1
adsorption of pure SO2 and the reaction of SO2 + O2 mixtures
with Al2O3 and CeO2 films in the millibar pressure range
provides a relatively clear picture of the transformations of the
sulfur-containing species on these oxide surfaces. It has been
shown unambiguously that sulfite and sulfate species are
produced both in the absence and in the presence of oxygen,
with their population being dependent not only on temperature
but also on the nature of the oxide surfaces (Al2O3 or CeO2).
Differences in redox properties of nonreducible alumina and
1330-1150 cm were blocked by water or ammonia during
preadsorption. The authors have concluded that these species
represent a physisorbed form of SO2 weakly bonded to Lewis
acid sites (coordinatively unsaturated aluminum ions). The
amount of the weakly bonded SO2,ads at T e 150 °C was several
times higher than the concentration of the strongly bonded
8
species. Lavalley et al. also suggested the occurrence of bisulfite
species resulting from the interaction between SO2 and the most
9
basic OH groups of alumina.

Interaction of Al2O3 and CeO2 Surfaces
J. Phys. Chem. B, Vol. 109, No. 23, 2005 11717
This consideration indicates that the relative population of
the surface by these species should depend on the termination
species on the alumina surface followed from data of Ziolek et
al., who had shown that 71% of SO2 is adsorbed reversibly
1
0
(
Al or O ions) and degree of hydroxylation of the alumina
on γ-Al2O3. Contrary to this, only 7% of adsorbed SO2 could
1
0
surface under study. The formation of weakly bonded SO2,ads
requires the presence of anion vacancies on the alumina surfaces,
whereas the amount of strongly bonded SO2,ads species will be
enhanced with the increase in the basicity of surface Os ions.
The latter supposition is confirmed by Mitchell, Sheinker, and
be desorbed from the CeO2 surface. At the same time, the
above-discussed explanation of this shift in S2p binding energies
due to admixing of sulfate formed in the presence of oxygen
(see Results) cannot be avoided completely.
2
-
2
-
4.2. SO4 Formation on Al2O3. The interaction of Al2O3
with pure SO2 or with SO2 + O2 mixtures at temperatures above
300 °C results in the formation of sulfate-like species. This
conclusion is based on the appearance of the S2p3/2 feature at
169.1 ( 0.2 eV in the corresponding experiments (Figures 1
and 2). This implies that the process of sulfation is activated.
The sulfate formation in the presence of molecular oxygen at
high temperatures is well-known and has been reported many
8
White, who showed that an introduction of sodium to high-
purity γ-alumina increases the amount of chemisorbed sulfite
by more than 10 times as compared to the undoped sample.
Namely, the high Lewis basicity of CeO2 provided a predomi-
nant contribution of strongly bonded sulfite species. And at
last, an enhanced concentration of hydrosulfite species should
be observed in the case of the hydroxylated surface of high-
surface area oxides.
1
0
2
,3,6,8,11-14
times.
Comparing the initial rates of adsorption and
6
oxidative adsorption of SO2, Nam and Gavalas have shown
that SO2 adsorbed on the alumina surface can be oxidized to
the sulfate ions
The surface composition of the Al2O3 and CeO2 films used
in this study are determined by the procedure of their prepara-
tion. The UHV deposition of the corresponding metal followed
by annealing in an O2 atmosphere inside the spectrometer
chamber suggests that we deal with high-purity, completely
dehydroxylated surfaces of alumina and ceria with a low surface
area. According to calculations of the electronic structure and
energetics of the R-Al2O3 (0001) surface, the most stable
configuration of the clean dehydroxylated sample is terminated
3
Al O + 3SO + / O f Al (SO )
(2)
2
3
2
2
2
2
4 3
which remain at the original SO2 adsorption sites. Although
aluminum sulfate is a thermodynamically stable compound, the
authors did not observe the bulk sulfate phase in the IR spectra.
They concluded that only surface sulfate could be produced at
350 °C and P(SO2) ) 8 mbar and P(O2) ) 100 mbar, that is,
under the conditions similar to those used in our study. Based
on IRS studies, the following sulfate structures on alumina
40
by an Al layer, which provides adsorption sites of the first
type, Als3+. This theoretically predicted structure was later
experimentally confirmed by ion scattering spectroscopy (ISS)
and reflection high-energy electron diffraction (RHEED). At
the same time, O ions located in the second layer are also
accessible, providing sites of the second type for SO2 adsorption.
Furthermore, the polycrystalline nature of our samples supposes
4
1
1
1,17-19,44
2
-
surfaces were supposed.
that other planes, which are terminated by an oxygen layer, for
example, (1 1h 02),4
0,42
can also be present on the surface. The
If the formation of surface aluminum sulfate via the reaction
accessibility of the Ce(IV) ions for adsorbates follows from
of the Al2O3 films with SO2 + O2 mixtures is expected (Figure
43
recent data obtained with the use of elevated-temperature STM.
They have concluded that the unreduced CeO2 (001) surface is
stable if the oxygen coverage of the topmost layer is reduced
to 50%.43 The oxygen vacancies formed in this case provide
the electroneutrality of the surface and the accessibility of Ce
ions located in the second layer.
2
), then the alumina sulfation under the influence of SO2 in the
absence of oxygen (Figure 1) looks strange. Indeed, the majority
of the previous works, which have studied the adsorption of
pure SO2 on powdery alumina, have not reported the sulfate
4+
2
-8
formation under comparable conditions. To our knowledge,
there was only one work where the reaction of SO2 with γ-Al2O3
at 80 °C resulted in the parallel formation of sulfite and sulfate
species characterized by the XPS S2p3/2 lines with binding
These considerations allow us to propose a reasonable
explanation of the above-mentioned experimental facts. The
more efficient adsorption of SO2 on the CeO2 surface than that
on the Al2O3 surface (Figures 1 and 2, Tables 2 and 3) is
explained by the higher Lewis basicity strength of ceria as
compared to alumina. Moreover, one can suggest that the low
Lewis basicity of Al2O3 decreases considerably the concentration
of chemisorbed SO2 (bonded with Os2 ions) so that the
15
energies of 167.5 and 169.3 eV, respectively. At the same
time, the formation of the sulfate species was detected by XPS
and XANES methods after small doses (several Langmuirs) of
SO2 adsorption on such nonreducible oxides as MgO and ZnO
45-49
at room temperature or even below.
-
Two mechanisms of the formation of sulfate species on the
surfaces of alumina and other nonreducible oxides might be
proposed. The first mechanism includes a disproportionation
of sulfur dioxide into sulfate and SO species with a subsequent
3
+
contribution of physisorbed SO2 (bonded with Als ions)
becomes predominant. This suggestion is supported by the
suppression of the SO2 adsorption in the presence of O2 in the
gas phase (compare Figures 1a and 2a). Note that a similar result
was observed when water was coadsorbed with SO2 (the spectra
are not shown in this paper). Because both molecular oxygen
24,48
desorption into the gas phase
Al O + 6SO f Al (SO ) + 3SO
(3)
2
3
2
2
4 3
3
+
and water are adsorbed on Lewis acid sites, only Als ions,
2
-
but not Os ones, were blocked in these experiments. As a
consequence, one can suggest that the low intensive S2p features
observed after the interaction of Al2O3 with the SO2 + O2
mixture at T < 300 °C (Figure 2a, Table 3) can be attributed to
sulfite-like species adsorbed on Os2- ions, which are not affected
by O2 or H2O adsorption. This could explain the small shift of
the S2p lines to higher binding energies. A similar conclusion
about the predominant contribution of the weakly bonded SO2,ads
However, evaluation of the thermodynamic parameters of this
reaction for Al O using the standard data tabulated in refs 50
2
3
and 51 leads to values ∆H298° ≈ +29 kJ/mol and ∆S ° ≈
298
-635 J/mol‚K, which suggest that reaction 3 is hardly thermo-
dynamically possible.
Another possible mechanism of the sulfation is based on an
2
-
assumption that SO4 surface species are formed upon the
adsorption of SO2 molecules on undersaturated oxygen ions,

11718 J. Phys. Chem. B, Vol. 109, No. 23, 2005
Smirnov et al.
the existence of which has been shown recently for the (1 1h 02)
sulfate ions to Ce(III) ions estimated from these data ranges
from 1.3 to 1.6, which is very close to the stoichiometric ratio
of the Ce(III) sulfate (3:2). As a consequence, the sulfation in
the presence of O2 can be described formally by the following
reaction:
42
surface by the crystal truncation rod (CTR) diffraction method.
It has been revealed that the most probable structure of this
plane is terminated by additional oxygen ions situated above
the theoretically predicted Al topmost layer. Such a kind of
undersaturated oxygen species are less negatively charged as
compared to usual O2- anions in the bulk of alumina and,
therefore, they could be sites for strong adsorption of SO2
molecules in the form of the surface sulfate species. Assuming,
for simplicity, that the undersaturated oxygen anions are singly
charged, the sulfate formation could be described formally by
the following reaction:
2CeO + O + 3SO f Ce (SO )
4 3
(7)
2
2
2
2
In the presence of oxygen, continuous growth of the
concentration of the sulfate ions with temperature (Figure 3)
suggests that sulfation of the CeO2 proceeds not only on the
surface but also spreads in the bulk. Earlier, the formation of
two types of Ce(III) sulfates (surface and bulk-like) has been
SO + 2O f SO42-
-
(4)
2
s
22
reported by Waqif et al., who have studied the sulfation of
CeO2 samples with different textures by IRS. They have shown
that the relative amount of each type of sulfate species and the
temperature of their formation depend on a composition of the
reaction mixture (SO2 versus SO2 + O2) and on the texture of
the samples.
It should be recalled that the polycrystalline structure of the
alumina film prepared in this work allows a description of the
surface by a set of different crystallographic faces separated by
interplanar boundaries. The consequence of this mechanism of
sulfation is the fact that the maximum possible sulfate yield
cannot be equal to or exceed one monolayer. This conclusion
is in agreement with the literature data that the amounts of
sulfate species produced on alumina samples after their interac-
tion with SO2 even in the presence of oxygen were below 2 ×
5
. Summary
The interaction of Al2O3 and CeO2 thin films with sulfur
dioxide at a partial pressure of 2.5 mbar and T < 300 °C leads
to the formation of sulfite species. This process is more effective
for the CeO2 surface, which is explained by a higher Lewis
basicity strength of ceria as compared to alumina. The low Lewis
basicity of alumina decreases the concentration of chemisorbed
14
2 7,10,13,16
1
0
atoms/cm .
-
4
.3. SO42 Formation on CeO2. The redox properties of
CeO2 make the sulfation of the CeO2 surfaces under the reaction
of ceria with pure SO2 reasonable. The reduction of Ce(IV) ions
to Ce(III) ones can be considered as a coupled semireaction,
which provides the trapping of electrons that is necessary for
the oxidation of SO2 to surface sulfate or oxosulfate
2
-
sulfite species (bonded with Os ions) on the Al2O3 surface,
and the contribution of physisorbed SO2 (bonded with Als3
ions) becomes predominant. The dual mechanism of SO2
adsorption is confirmed by suppression of the sulfite species
formation if the oxides interact with SO2 in the presence of
oxygen.
+
6
CeO + 3SO f Ce (SO ) +
2 2 2 4 3
2
Ce O {or 2CeO + SO f Ce O (SO )} (5)
2 3 2 2 2 2 4
An increase in the reaction temperature to 300 °C results in
the formation of sulfate species. The formation of the sulfate
species is accompanied by the reduction of ceria to cerium (III)
sulfate or oxosulfate. The formation of sulfate species on
alumina might be presented as adsorption of SO2 molecules on
the undersaturated surface oxygen anions.
Indeed, the shift of S2p features to higher binding energies
was registered after the interaction of the CeO2 films with SO2
at T g 300 °C as compared to the features measured at lower
temperatures (Figure 1). This observation indicates the activated
formation of the sulfate ions. The formation of Ce(III) sulfate
was reported earlier by Diwell et al.20 after the treatment of
CeO2 with 1 vol % SO2 in nitrogen at 550 °C. Parallel sulfate
formation and ceria reduction under annealing of ceria surfaces
with adsorbed SO2 at 550 °C or above was also concluded
Addition of oxygen to SO facilitates the sulfation. In the
2
case of CeO2, the formation of the sulfate ions is observed at a
temperature 100 °C lower than that in the case of pure SO2.
Moreover, the addition of oxygen to SO2 increases the yield of
recently from XPS data by Ferrizz and coauthors.2 Another
mechanism of the sulfate formation without reduction of the
CeO2 surface
5
sulfate on CeO by a factor of 1.5-2. At elevated temperatures,
2
the sulfation of ceria in the presence of oxygen results in the
beginning of the formation of a 3D phase of cerium (III) sulfate.
The effect of oxygen on the sulfation of Al2O3 is essentially
smaller.
CeO + 4SO f 2SO + Ce(SO )
4 2
(6)
2
2
has been proposed by Rodriguez et al., who have concluded
from XPS data (S2p) the appearance of the sulfate ions after
SO2 adsorption (5 L) on the surface of a ceria film at ambient
temperatures.24 It should be noted, however, that mechanism 6
was not supported by an analysis of the Ce3d spectra. This
analysis was made in the present paper. The spectra presented
in Figure 4a demonstrate unambiguously that the reduction of
Ce(IV) into Ce(III) occurs when the sulfation proceeds under
the influence of SO2 at temperatures g300 °C.
Even the presence of oxygen in the SO2 + O2 mixtures during
sulfation, which decreases the temperature of sulfation and
increases the sulfate yield, does not prevent the reduction of
Ce(IV) ions to Ce(III) ones (Figure 4b). Moreover, a quantitative
correlation between the sulfate yields and the fraction of the
Ce(III) ions is observed (Figure 5). The atomic ratio of the
Acknowledgment. V.I.B. gratefully acknowledges the Rus-
sian Science Support Foundation, grant for young talented
scientists.
References and Notes
(1) Truex, T. J. SAE Paper 1999, 1999-01-1543.
(2) Chang, C. C. J. Catal. 1978, 53, 374.
(3) Babaeva, M. A.; Tsyganenko, A. A.; Filimonov, V. N. Kinet. Katal.
984, 25, 787 (in Russian).
1
(
(
4) Karge, H. G.; Dalla Lana, I. G. J. Phys. Chem. 1984, 88, 1538.
5) Datta, A.; Cavell, R. G.; Tower, R. W.; George, Z. M. J. Phys.
Chem. 1985, 89, 443.
(
6) Nam, S. W.; Gavalas, G. R. Appl. Catal. 1989, 55, 193.
(7) Waqif, M.; Saad, A. M.; Bensitel, M.; Bachelier, J.; Saur, O.;
Lavalley, J. C. J. Chem. Soc., Faraday Trans. 1992, 88, 2931.
8) Mitchell, M. B.; Sheinker, V. N.; White, M. G. J. Phys. Chem.
1996, 100, 7550.
(

Interaction of Al2O3 and CeO2 Surfaces
J. Phys. Chem. B, Vol. 109, No. 23, 2005 11719
(
9) Saad, A. B. M.; Saur, O.; Wang, Y.; Tripp, C. P.; Morrow, B. A.;
Lavalley, J. C. J. Phys. Chem. 1995, 99, 4620.
10) Ziolek, M.; Kujawa, J.; Saur, O.; Aboulayt, A.; Lavalley, J. C. J.
Mol. Catal. A: Chem. 1996, 112, 125.
11) Saur, O.; Bensitel, M.; Saad, A. B. M.; Lavalley, J. C.; Tripp, C.
P.; Morrow, B. A. J. Catal. 1986, 90, 104.
12) Waqif, M.; Saur, O.; Lavalley, J. C.; Perathoner, S.; Centi, G. J.
Phys. Chem. 1991, 95, 4051.
13) Pieplu, A.; Saur, O.; Lavalley, J. C.; Pijolat, M.; Legendre, O. J.
Catal. 1996, 159, 394.
14) Uy, D.; Dubkov, A.; Graham, G. W.; Weber, W. H. Catal. Lett.
000, 68, 25.
15) Gergely, B.; Guimon, C.; Gervasini, A.; Auroux, A. Surf. Interface
Anal. 2000, 30, 61.
16) Dalla Lana, I. G.; Karge, H. G.; George, Z. M. J. Phys. Chem.
993, 97, 8005.
17) Dosumov, K.; Umbetkaliev, A. K.; Popova, N. M. Zh. Prikl. Khim.
994, 67, 1961 (in Russian).
18) Dunn, J. P.; Jehng, J. M.; Kim, D. S.; Briand, L. E.; Stenger, H.
G.; Wachs, I. E. J. Phys. Chem. B 1998, 102, 6212.
19) Khadjivanov, K.; Davydov, A. Kinet. Katal. 1988, 29, 398
in Russian).
(30) Lindberg, B. J.; Hamrin, K.; Johansson, G.; Gelius, U.; Fahlmann,
A.; Nordling, C.; Siegbahn, K. Phys. Scr. 1970, 1, 286.
(31) Visser, J. P. R.; Groot, C. K.; van Oers, E. M.; de Beer, V. H. J.;
Prins, R. Bull. Soc. Chim. Belg. 1984, 93, 813.
(
(
(
(
(
32) Brion, D. Appl. Surf. Sci. 1980, 5, 133.
33) Manocha, A. S.; Park, R. L. Appl. Surf. Sci. 1977, 1, 129.
34) Burroughs, P.; Hamnett, A.; Orchard, A. F.; Thornton, G. J. Chem.
(
Soc., Dalton Trans. 1976, 1686.
35) Fujimori, A. Phys. ReV. B 1983, 28, 2281.
(36) Shyu, J. Z.; Weber, W. H.; Gandhi, H. S. J. Phys. Chem. 1988, 92,
4964.
(
(
(
2
(
(37) Shyu, J. Z.; Otto, K.; Watkins, W. L. H.; Graham, G. W.; Belitz,
R. K.; Gandhi, H. S. J. Catal. 1988, 114, 23.
(38) Shyu, J. Z.; Otto, K. J. Catal. 1989, 115, 16.
(
1
1
(
(
(
39) Paparazzo, E. Surf. Sci. 1990, 234, L253.
40) Guo, J.; Ellis, D. E.; Lam, D. J. Phys. ReV. B 1992, 45, 13647.
41) Suzuki, T.; Hishita, S.; Oyoshi, K.; Souda, R. Surf. Sci. 1999, 437,
(
(
2
89.
(42) Trainor, T. P.; Eng, P. J.; Brown, G. E.; Robinson, I. K.; De Santis,
(
M. Surf. Sci. 2002 496, 238.
(43) N o¨ renberg, H.; Harding, J. H. Surf. Sci. 2001, 477, 17.
(44) Bensitel, M.; Saur, O.; Lavalley, J. C.; Morrow, B. A. Mater. Chem.
Phys. 1988, 19, 147.
(45) Chaturvedi, S.; Rodriguez, J. A.; Jirsak, T.; Hrbek, J. J. Phys. Chem.
B 1998, 102, 7033.
(
(
(
20) Diwell, A. F.; Hallett, C.; Taylor, J. R. SAE Paper 1987, 872163.
21) Beck, D. D.; Krueger, M. H.; Monroe, D. R. SAE Paper 1991,
9
10844.
22) Waqif, M.; Bazin, P.; Saur, O.; Lavalley, J. C.; Blanchard, G.;
Touret, O. Appl. Catal. B 1997, 11, 193.
23) Overbury, S. H.; Mullins, D. R.; Huntley, D. R.; Kundakovic, Lj.
J. Phys. Chem. B 1999, 103, 11308.
24) Rodriguez, J. A.; Jirsak, T.; Freitag, A.; Hanson, J. C.; Larese, J.
Z.; Chaturvedi, S. Catal. Lett. 1999, 62, 113.
25) Ferrizz, M.; Gorte, R. J.; Vohs, J. M. Catal. Lett. 2002, 82, 123.
26) Kaichev, V. V.; Sorokin, A. M.; Timoshin, A. I.; Vovk, E. I.
Instrum. Exp. Tech. 2002, 45, 58.
27) Romeo, M.; Bak, K.; El Fallah, J.; Le Normand, F.; Hilaire, L.
Surf. Interface Anal. 1993, 20, 508.
(
(
(
46) Rodriguez, J. A.; Jirsak, T.; Chaturvedi, S.; Kuhn, M. Surf. Sci.
999, 442, 400.
47) Rodriguez, J. A.; Jirsak, T.; Freitag, A.; Larese, J. Z.; Maiti, A. J.
Phys. Chem. B 2000, 104, 7439.
48) Rodriguez, J. A.; Jirsak, T.; Perez, M.; Chaturvedi, S.; Kuhn, M.;
Gonz a´ lez, L.; Maiti, A. J. Am. Chem. Soc. 2000, 122, 12362.
49) Rodriguez, J. A.; P e´ rez, M.; Jirsak, T.; Gonz a´ lez, L.; Maiti, A. Surf.
1
(
(
(
(
(
(
(
Sci. 2001, 477, L279.
(50) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC
Press: Boca Raton, FL, 1992-1993.
(
(
28) Park, P. W.; Ledford, J. S. Langmuir 1996, 12, 1794.
29) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D.
Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer: Eden
Prairie, MN, 1992.
(51) Thermodynamic Properties of IndiVidual Substances; Glushko, V.
P., Ed.; Nauka: Moscow, 1978.