Selective oxidation of H2S over V2O5 supported on CeO2-intercalated Laponite clay catalysts

Article abstract of DOI:10.1039/c3cy00431g

A series of V₂O₅ supported on CeO₂- intercalated clay catalysts (named V₂O₅/Ce-Lap catalysts) with mesoporous structure and high specific surface area were prepared. The structural characteristics and physicochemical properties were studied in detail. These catalysts were evaluated for the H₂S selective oxidation. It was revealed that all V₂O₅/Ce-Lap catalysts showed high catalytic activities in the temperature range of 120-220 °C due to the redox reaction of highly dispersed V⁵⁺ species, and the catalytic mechanism obeyed a redox process, i.e. a stepwise mechanism. Additionally, the chemically absorbed oxygen also played an important role in H₂S selective oxidation. Among them, the 5% V₂O ₅/Ce-Lap catalyst presented the best catalytic activity and excellent regenerability at 180 °C. Finally, the catalyst deactivation mechanism was explored.

Full text of DOI:10.1039/c3cy00431g

Catalysis
Science & Technology
View Article Online
View Journal
PAPER
Selective oxidation of H S over V O supported on
2
2
5
CeO -intercalated Laponite clay catalysts
2
Cite this: DOI: 10.1039/c3cy00431g
Xin Zhang, Guangyu Dou, Zhuo Wang, Jie Cheng, Hailin Wang, Chunyan Ma and
Zhengping Hao*
A series of V
mesoporous structure and high specific surface area were prepared. The structural characteristics and
physicochemical properties were studied in detail. These catalysts were evaluated for the H S selective
oxidation. It was revealed that all V /Ce-Lap catalysts showed high catalytic activities in the temperature
2 5 2 2 5
O supported on CeO -intercalated clay catalysts (named V O /Ce-Lap catalysts) with
2
2 5
O
5+
Received 24th June 2013,
Accepted 26th July 2013
range of 120–220 1C due to the redox reaction of highly dispersed V species, and the catalytic
mechanism obeyed a redox process, i.e. a stepwise mechanism. Additionally, the chemically absorbed
2 2 5
oxygen also played an important role in H S selective oxidation. Among them, the 5% V O /Ce-Lap
DOI: 10.1039/c3cy00431g
catalyst presented the best catalytic activity and excellent regenerability at 180 1C. Finally, the catalyst
www.rsc.org/catalysis
deactivation mechanism was explored.
Pillared interlayered clays (PILCs) supported vanadium species
1. Introduction
1,3,5,6
catalysts, such as Ti, Zr, Fe PILC supported V O catalysts,
2
5
2
Nowadays, a large amount of toxic H S is released by combustion have been of great research interest in the past few decades
1
of fossil fuels owing to energy demand. Thus, the conversion of because PILCs can be tailored to particular applications by
H S to elemental sulfur is desired to reduce its impact on the varying the size and separation of the pillars as well as their
2
environment. The most popular currently utilized technology for composition. Unfortunately, their microporous and moderate
2
the production of elemental sulfur from H S-containing gas is the porosity structure limits their further application (a pore volume
2
3
ꢀ1
7
Claus process. However, because of the thermodynamic limita- of 0.15–0.40 cm g ). This means that when PILCs are used as
tions, it is difficult to remove H S completely and 1% H S is still catalyst supports, the loaded active component will block inter-
left in Claus tail gas. In order to remove the residual H S, various layer pores, as a result, the active component becomes inacces-
2
2
3
2
processes of additional purification are applied. Among them, sible to most reactants. On the other hand, for vanadium based
the most well-known is H S selective catalytic oxidation technol- catalysts, the catalyst deactivation should be seriously considered
2
ogy; in this process, H S is directly oxidized to elemental sulfur by owing to their possible application in practical industries. Thus,
2
oxygen over a suitable catalyst. The irreversible reaction equa- it is desirable to design, explore and develop suitable catalysts
tions are as follows (eqn (1) is the main reaction, and eqn (2) and with high catalytic performances.
4
(
3) are side reactions):
Recently, a new approach was developed to synthesize alumina-
intercalated Laponite clay composites with mesoporous structure
and high specific surface area in the presence of surfactants.
In this process, the surfactants were first intercalated into
the galleries and form micelles, acting as templates, then,
alumina nanoparticles were embedded surrounding the micelles.
Finally, the composites were obtained after removal of the
(
1/n)S
n
+ H
2
O - H
+ O - SO
- SO
2
S + 1/2O
2
(1)
(2)
(3)
8
1
/nS
n
2
2
H
2
S + 3/2O
2
2
2
+ H O
It is noted that SO can be formed via the deep oxidation
2
2
surfactants by heating. Using this method, CeO -intercalated
of H S or the further oxidation of elemental sulfur, causing a
decrease in sulfur yield. Therefore, the conversion efficiency
strongly depends on the catalyst activity.
9
2
clay composites were synthesized successfully. As expected
the good redox property and the high mobility of the capping
10–12
oxygen of CeO2
make CeO -intercalated clay composites
2
suitable and attractive catalyst supports to load and modify
vanadium species.
Department of Environmental Nano-materials, Research Center for
Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085,
P. R. China. E-mail: zpinghao@rcees.ac.cn; Fax: +86-10-62923564;
Tel: +86-10-62923564
2
In this study, a series of CeO -intercalated clay composites
supported V
2
O
5
catalysts with mesoporous structure and high
This journal is c The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
specific surface area were prepared and investigated for H
2
S
Reduction temperature was increased from room temperature
ꢀ
1
selective oxidation. It is desirable that the catalytic performances to 900 1C at a rate of 10 1C min . Hydrogen concentration
could be enhanced. Moreover, the physicochemical properties in the effluent was continuously monitored using a thermo-
of catalysts were characterized by using various techniques. conductivity detector. Prior to each TPR run, the catalyst was
Besides, the catalytic mechanism and catalyst deactivation were pre-heated in He flow from room temperature to 300 1C and
explored.
held for 30 min.
temperature-programmed desorption (O
out using the same apparatus as in H -TPR experiments (Micro-
O
2
2
-TPD) was carried
2
2. Experimental
meritics Chemisorb 2720). The sample was heated from room
2.1 Catalyst preparation
ꢀ1
temperature to 900 1C at a rate of 10 1C min in pure He.
Laponite clay was supplied by Fernz Specialty Chemicals, Before each TPD test, 100 mg of the sample was pre-heated in
Australia. Nonionic PEO surfactant Tergitol type 15-S-9 was He flow from room temperature to 300 1C and held for 30 min.
procured from the Sigma Chemical Company. The CeO
2
-
After cooling to room temperature, 2% O /N was fed into
2 2
ꢀ1
intercalated Laponite clay composites were synthesized according the reactor at 50 mL min for 30 min, then, pure He was fed
9
ꢀ1
to the procedures described by Li et al. The prepared CeO
2
-
into the reactor at 50 mL min for 30 min to purge away any
intercalated Laponite clay composites were denoted as Ce-Lap. residual O
The V O loaded Ce-Lap catalysts containing 3, 5 and 8% V O
2
.
2
5
2 5
2
.3 Catalytic performance evaluation
were prepared by the wetness impregnation method using
NH VO as the active component precursor. The impregnated All catalytic performance evaluations were performed in a
materials were dried at 100 1C overnight, and then were calcined continuous flow fixed-bed quartz reactor at atmospheric pres-
at 400 1C for 4 h. The samples were denoted as n% V /Ce-Lap, sure. A catalyst (0.6 g, 20–40 mesh) was placed in the central
4
3
2 5
O
where n% refers to the V
on the support.
2
O
5
loading content (weight percentage) section of the reactor. The mixture gas containing 5000 ppm
H S, 2500 ppm O and balance gas (N ) was fed into the reactor
2
2
1
2
ꢀ
ꢀ1
at 200 mL min of the total gas flow rate (GHSV = 7000 h
)
2.2 Catalyst characterization
and reacted in the temperature range of 120–220 1C. The effluent
X-ray diffraction (XRD) patterns were recorded on a PANalytical stream was detected using a gas chromatograph (GC126) equipped
0
X Pert PRO powder diffraction system using Cu Ka radiation with a FPD and TCD. A condenser was located at the bottom of the
(
0
l = 0.15418 nm) in the 2y range of 10–801 (at a scanning rate of reactor to trap the sulfur gas in effluent stream. Instantaneous
ꢀ
1
.51 min ).
fractional conversion of H S, sulfur selectivity and sulfur yield
2
UV-vis diffuse reflectance spectra (UV-vis DRS) were recorded were defined as:
ꢀ
1
4
in the air against BaSO in the region of 12 500–24 000 cm on
a Hitachi UV-3000 spectrometer.
ðH2SÞ ꢀðH2SÞ
in
out
H2S conversion ¼
ðH2SÞ
in
The chemical composition of samples was determined by
X-ray fluorescence (XRF) using an XRF1800 spectrometer, and
an Rh tube as the excitation source.
ðH2SÞ ꢀðH2SÞ ꢀðSO2Þ
in
out
out
Sulfur selectivity ¼
ðH2SÞ ꢀðH2SÞ
in
out
Thermal stability was investigated by thermogravimetry
(
TG-DSC, Seteram Labsys). Typically, around 30 mg of the
sample was heated in an Al crucible at a constant heating
rate of 10 1C min from 25 to 1000 1C, with air purging at a
Sulfur yield ¼ ðH2S conversionÞ ꢁ ðSulfur selectivityÞ
2 3
O
ꢀ
1
ꢀ
1
flow rate of 30 mL min
.
3
3
. Results and discussion
.1. Structural and textural properties of V O /Ce-Lap
2 5
X-ray photoelectron spectra (XPS) were recorded on a
Thermo Electron Escalab250 instrument using Al Ka radiation.
ꢀ8
catalysts
The base pressure was 5 ꢁ 10 Pa. The binding energies were
calibrated using the C 1s peak of contaminant carbon (BE = 285 eV) Nitrogen adsorption–desorption isotherms and the BJH pore
size distribution calculated from the desorption branch (insert)
as standard, and quoted with a precision of ꢂ0.2 eV.
BET surface areas and textural properties of catalysts were of all catalysts and supports are shown in Fig. 1, and the textural
determined by nitrogen adsorption–desorption isotherms using properties are listed in Table 1. It is significant to observe that
a gas sorption analyzer NOVA1200. Prior to N
measurements, the samples were degassed at 300 1C for 3 h.
FTIR spectra of these V /Ce-Lap catalysts were collected modification, they increase to 580 m g and 1.12 cm g
on the Bruker Tensor 27 spectrometer with 256 scans at a On the other hand, all catalysts possess type IV isotherms,
2
adsorption the specific surface area and total volume of host clay are
2
ꢀ1
3
ꢀ1
322 m
g
and 0.24 cm g , respectively. After the CeO
2
2
ꢀ1
3
ꢀ1
2
O
5
.
ꢀ
1
13
resolution of 4 cm
.
a typical characteristic of mesoporous materials. Therefore,
H2 temperature-programmed reduction (H -TPR) experi- for the larger specific surface area, typical isotherms associated
2
ments were conducted using a Micromeritics Chemisorb 2720 with the BJH pore size distribution indicate that Ce-Lap com-
apparatus. TPR profiles were obtained by passing a 5% H
flow (50 mL min ) through the pretreated catalyst (100 mg). and the mesoporous structure remained after the V addition.
2
/He posites with mesoporous structure were successfully obtained,
ꢀ1
2 5
O
Catal. Sci. Technol.
This journal is c The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Fig. 2 XRD patterns of V
2
O
5
/Ce-Lap catalysts: (a) Laponite, (b) Ce-Lap, (c) 3%
Fig. 1 Nitrogen adsorption–desorption isotherms and pore size distribution calcu-
V
2
O
5
/Ce-Lap, (d) 5% V O /Ce-Lap and (e) 8% V O /Ce-Lap.
2 5 2 5
lated from the desorption branch (insert) of V
2
O
5
/Ce-Lap catalysts: (a) Laponite,
/Ce-Lap and (e) Ce-Lap.
(
b) 8% V /Ce-Lap, (c) 5% V /Ce-Lap, (d) 3% V O
2
O
5
2
O
5
2 5
Table 2 Chemical composition (wt%) of the samples investigated
Table 1 Textural properties of V
2
O
5
/Ce-Lap catalysts
Catalyst
SiO
2
MgO
V
2
O
5
CeO
2
a
2
ꢀ1
b
3
ꢀ1
c
Catalyst
S
BET (m g
)
V
p
(cm g
)
D
p
(nm) 3%V O /Ce-Lap
59.9
58.92
59.47
23.24
19.56
17.69
3.37
5.32
8.32
12.4
13.5
13.28
2
2
2
5
5
5
5
8
%V
%V
O
O
/Ce-Lap
/Ce-Lap
Lap
Ce-Lap
322
580
378
367
284
0.24
1.12
0.48
0.36
0.32
2.9
7.7
5.0
3.9
4.7
3%V
5%V
8%V
2
2
2
O
5
O
5
O
5
/Ce-Lap
/Ce-Lap
/Ce-Lap
7
than that in host clay and the content decreases with the
rise of V O loading. As is known, the Laponite consists of an
a
b
2 5
BET specific surface area calculated at P/P
0
= 0.05–0.25. Total pore
c
octahedral MO layer sandwiched between two tetrahedral SiO4
volume estimated at P/P
0
= 0.99. BJH pore diameter calculated from
6
the desorption branch.
layers. Magnesium atoms are at the center of the MO octahedra.
6
When the magnesium is leached out in the acidic environment,
the Si residue is more likely in the form of amorphous silicate
particles. Thus, the loss of magnesium implies the damage of
clay layers. As a result, more amorphous silicate particles are
Additionally, the high specific surface area and mesoporous
structure are favorable for H S selective oxidation reaction, as
they allow V to be well dispersed on the support. However,
2
2 5
O
formed with the rise of V
2 5
O loading.
the nitrogen adsorption capacity decreases slightly with the rise
of V O loading, attributed to the pore blockage by the loaded
Fig. 3 exhibits the FTIR spectra of all catalysts; the spectra
reveal intense bands at around 1200 cm , which correspond to
the Si–O vibration of SiO4 tetrahedra. Apparently, the inten-
2
5
ꢀ
1
V O . Meanwhile, the steps of hysteresis loops shift towards
2
5
lower relative pressure, indicating that relatively small sized
1
3
2 5
sities of the bands weakened with the rise of V O loading due to
pores are formed, which is confirmed by the pore size data in
Table 1.
XRD was employed to further investigate the structure of
V O /Ce-Lap catalysts. As shown in Fig. 2, for Ce-Lap, it can be
2 5
noticed that the characteristic diffraction peaks of host clay are
maintained. Meanwhile, new peaks corresponding to CeO2
appear at 28.21, 47.71 and 56.41. This demonstrated that CeO2
was successfully intercalated into the galleries. This is in
accordance with the results of nitrogen adsorption–desorption
isotherms. Meanwhile, for all catalysts, no peak for V
observed, implying that V is highly dispersed on Ce-Lap.
2 5
However, CeO is better crystallized with the rise of V O
2 5
O was
2 5
O
2
loading, which is perhaps due to the interaction of vanadium
species with cerium species. However, compared with host clay,
the structure of V O /Ce-Lap is seriously damaged. This can be
2
5
explained by the fact that the acid properties of the precursor
metal salt) solutions cause the leaching and damage of the clay
(
7
layer. Such an argument is supported by the XRF analysis. As
shown in Table 2, the MgO content in all catalysts is smaller
Fig. 3 FTIR spectra of V
/Ce-Lap, (d) 5% V
2
O
5
/Ce-Lap catalysts: (a) Laponite, (b) Ce-Lap, (c) 3%
V
2
O
5
2
O
5
/Ce-Lap and (e) 8% V /Ce-Lap.
2 5
O
This journal is c The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 4 TG patterns of V
2
O
5
/Ce-Lap catalysts: (a) Laponite, (a) Ce-Lap, (b) 3%
V
2
O
5
/Ce-Lap, (c) 5% V
2
O
5
/Ce-Lap and (d) 8% V /Ce-Lap.
2 5
O
Fig. 5 UV-vis DRS spectra of V
2 5 2 5
O /Ce-Lap catalysts: (a) 3% V O /Ce-Lap, (b) 5%
V
2
O
5
/Ce-Lap and (c) 8% V /Ce-Lap.
2 5
O
the content of SiO
4 2 5
tetrahedra decreasing with the rise of V O
loading. This was confirmed by XRF analysis. Meanwhile, the
bands appearing at around 1660 cm are attributed to the
bending vibration of adsorbed water and the broad bands at
5
+
ꢀ1
260 nm is related to the isolated V in a 4-fold coordination
state. While the band at 350 nm corresponds to the polymeric
5
+
ꢀ
1
ꢀ1
2 5
V O chains, which are formed by the isolated V species with
3
–
3
200 cm –3600 cm are generally assigned to the vibration of
1
,5,18,19
ꢀ1
the formation of a V–O–V bridge,
which is in good correla-
OH groups. In addition, the bands located at 3678 cm and
ꢀ
1
tion with the FTIR result. Furthermore, it is worth mentioning
that no bands appear in the range of 600–800 nm, which relate
O species. Therefore, the UV-vis DRS
2 5
analysis suggests that vanadium species are mainly present in
742 cm correspond to the vibration of –OH groups and
ꢀ1
silanol groups, respectively. Moreover, a band at 818 cm was
clearly observed, which is supposed to be related to the defor-
mation of V–O–V bridges arising from V
the formation of V
to the crystalline V
1
4
2
O
5
.
Thus, it confirms
5
+
a highly dispersed state of isolated V and polymeric V O .
2
5
2 5
O .
Fig. 6 presents the V 2p XPS spectra of a represented catalyst
5% V /Ce-Lap). For the fresh catalyst, the spectrum is pre-
dominated by a single peak that appeared at 517.8 eV, which is
the typical characteristic of V . Thus, the XPS analysis con-
firms the unique existence of V . In order to further explore
2 5
the reaction mechanism, the XPS spectra of Ce 3d for 5% V O /
Ce-Lap are displayed in Fig. 7. For the fresh catalyst, the peaks
labeled m and n refer to 3d2/5 and 3d2/3 spin–orbit states,
respectively. Moreover, the bands labeled m and n correspond
Fig. 4 provides the thermogravimetric (TG) analysis patterns
of all samples. In the case of Ce-Lap, the curve shows two
distinct weight losses; the first one can be assigned to the
removal of adsorbed and hydration water molecules on the surface
(
2 5
O
5
+
5
+ 20
9
of clay layers, at temperatures below 150 1C. While the second
weight loss observed in the temperature range of 600–900 1C
1
,15
corresponds to the dehydroxylation of clay structure.
In this
process, an inflection point is found at around 800 1C due to
I
I
1
6
the collapse of clay structure. Additionally, the slight weight
loss that occurred at 150–600 1C is attributed to the removal of
interlayer water. For the V O /Ce-Lap catalysts, it is noticed that
2
5
a new weight loss appears at around 200 1C, which can be
1
attributed to the dehydroxylation of V
amount of the two steps increases with the rise of V
This can be attributed to the desorption of adsorbed water on
2
O
5
.
The weight loss
2 5
O
loading.
1
7
2
the CeO surface as well as the easier and larger amount
removal of structural –OH due to the damage of layer structure.
Meanwhile, the second weight loss shifts towards the lower
temperature range. Likewise, the inflection point is found at
around 450 1C, indicating that the thermal stability of the
catalysts is reduced because the layer structure is damaged.
3.2 Surface species analysis
UV-vis DRS was employed to characterize the chemical status of
the loaded vanadium species. As shown in Fig. 5, the neat
spectra reveal the presence of two bands at 260 nm and 350 nm
2 5
Fig. 6 V 2p XPS spectra of (a) the fresh 5% V O /Ce-Lap catalyst and (b) the
2 5
as the main features of the catalysts. As reported, the band at used 5% V O /Ce-Lap catalyst.
Catal. Sci. Technol.
This journal is c The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
2
6
probably associated with the reduction of bulk CeO
2
to Ce
2 3
O .
3
+
However, because of the low content of Ce , there are no peaks
for Ce species in XRD and H -TPR patterns. Nevertheless, for
3
+
2
V O /Ce-Lap catalysts, the two peaks at 440 1C and 570 1C are
2
5
replaced by a new single peak that appeared in the range of
5
+
5
70–600 1C, which can be assigned to the reduction of V
4
+ 3,5
to V . Moreover, the reduction temperature of V
2 5
O /Ce-Lap
is lower than that of V /Lap at around 630 1C. The reduction
2 5
O
can be attributed to the weakening of the vanadyl group, which
is caused by the significant interaction of vanadium species
2
7
with cerium species. Meanwhile, the new peak shifts towards
higher temperature with the rise of V O loading. This can be
2
5
explained by the fact that with the rise of V O loading, more
2
5
2 5
Fig. 7 Ce 3d XPS spectra of (a) the fresh 5% V O /Ce-Lap catalyst and (b) the
surface vanadium atoms interact with the internal vanadium
layer and increase the bond strength of the vanadyl group.
Moreover, the lower reduction temperature reveals that vanadium
2 5
used 5% V O /Ce-Lap catalyst.
28–30
3
+
II
III
species is highly dispersed on Ce-Lap.
with the UV-vis DRS analysis.
This is in agreement
to the presence of Ce . While the peaks labeled m, m , m , n,
II
III
4+ 20,21
n and n are related to the existence of Ce .
reveals the partial existence of Ce (area percentage is 22%).
XPS analysis
3
+
3
+
2
3.4 O -TPD analysis of catalysts
As is known Ce can result in a charge imbalance, and the
vacancies and the unsaturated chemical bonds on the catalyst The ability to capture and release oxygen of all samples was
22–24
surface are favorable for the formation of chemisorbed oxygen,
clarified by O -TPD measurements. As shown in Fig. 9, there
2
which is the most active oxygen and plays an important role in exist three oxygen desorption peaks for Ce-Lap, the one that
22,25
oxidation reactions.
appeared at 120 1C corresponds to the removal of ordinarily
chemically adsorbed oxygen. While the one located at 800 1C is
ascribed to the removal of lattice oxygen. Finally, the one located
3.3 Reductive properties of catalysts
The reducibility of all catalysts was investigated by H -TPR at between these two temperatures (450 1C) can be attributed to
2
experiments. Their H
clearly observed that only one main peak appears at 300 1C for vacancies.
host clay, corresponding to the slight reduction of Laponite. Ce-Lap, a distinctive feature is observed: the curves are dominated
Apparently, after the CeO
modification, a poor resolved peak by two peaks located at 370 1C and 470 1C, respectively. Both of
2
-TPR profiles are shown in Fig. 8. It is the desorption of oxygen chemically adsorbed on the oxygen
3
1,32
However, after the addition of V O into the
2 5
2
appears at 460 1C, which acts as a shoulder and overlapped with them correspond to the desorption of oxygen chemically adsorbed
a broad peak with maximum at 570 1C. The two peaks can be on the oxygen vacancies. Meanwhile, the peak that appeared at
attributed to the removal of surface oxygen from CeO2 and higher temperature becomes obscured and could be ignored.
2 5
formation of non-stoichiometric cerium oxides, CeO from CeO , Among all the catalysts studied, the 5% V O /Ce-Lap catalyst
x
2
respectively. Meanwhile, the peak that appeared at 820 1C is exhibits the strongest peak intensity, indicating that it possesses
the largest amount of oxygen vacancies.
Fig. 8
H
2
-TPR profiles of V
2
O
5
/Ce-Lap catalysts: (a) Laponite, (b) Ce-Lap, (c) 3%
/Ce-Lap and (f) 5% V /Lap.
Fig. 9
O
2
-TPD profiles of V
2
O
5
/Ce-Lap catalysts: (a) Ce-Lap, (b) 3% V
/Ce-Lap.
2 5
O /Ce-Lap,
V
2
O
5
/Ce-Lap, (d) 5% V
2
O
5
/Ce-Lap, (e) 8% V
2
O
5
2
O
5
(c) 8% V
2
O
5
/Ce-Lap and (d) 5% V O
2 5
This journal is c The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
4
. Catalytic performances
As is shown the Ce-Lap exhibits the maximum sulfur yield
(
60%) at 200 1C. Thus, it can be deduced that the highly
4.1 Effect of reaction temperature
5
+
dispersed V species and the enhanced physicochemical
Fig. 10 shows the effect of temperature on the catalytic activities
of V
from Fig. 10(A) that the H
the rise of reaction temperature up to 180 1C, but slightly falls
upon further increasing the temperature to 220 1C. Besides, an
properties play important roles in H S selective oxidation.
2
2
O
5
/Ce-Lap catalysts for the oxidation of H
2
S. It can be seen
Furthermore, the slight decrease of catalytic activity for the 8%
2
S conversion initially increases with
2 5
V O /Ce-Lap catalyst is mainly due to the decrease of specific
surface area.
increase in H
2
S conversion is found when V
2
O
5
loading raised from 4.2 Catalytic mechanism
3
% to 5%. Nevertheless, H S conversion decreases with the rise of
2
The V 2p and Ce 3d XPS spectra of the used 5% V
catalyst are presented in Fig. 6 and 7 respectively. For V 2p, a
2 5
O /Ce-Lap
V O loading up to 8%. The maximum H S conservation (98%) is
obtained at 180 1C for the 5% V O /Ce-Lap catalyst. Fig. 10(B)
shows the relationship between sulfur yield and reaction tempera-
ture for these catalysts. Similar trends are observed for sulfur yield,
because of the high and constant sulfur selectivity (almost 99%).
Likewise, the maximum sulfur yield (98%) is obtained at 180 1C for
2
5
2
2
5
4+
new peak that appeared at 516.8 eV corresponds to the V
18,20,34
state
(the area percentage is about 40%). Meanwhile, no
new peak can be detected for Ce 3d. Nevertheless, the propor-
3+
tion of Ce increases from 22% to 35%. Thus, it indicates that
5+
4+
both V and Ce participate in the oxidation reaction and the
catalytic mechanism obeys the Mars–van Krevelen or redox
1
,3,5,6,33
the 5% V
2
O
5
/Ce-Lap catalyst. As reported,
the reaction
temperature for H S selective oxidation over pillared clay
supported metal oxide catalysts is higher than 200 1C and the
maximum sulfur yield is about 95%. Thus, the V O /Ce-Lap
catalysts can not only reduce the reaction temperature but also
improve the catalytic activities significantly.
12
2
process, i.e. the stepwise mechanism. Based on this, the
catalytic oxidation process can be tentatively proposed as follows:
2
5
5+
4+
H S firstly was oxidized to elemental sulfur by V and Ce , and
2
4+
3+
then the V and Ce were re-oxidized by the gas phase oxygen in
a separate step. Moreover, the high catalytic activities of V
2 5
O /
The catalytic activity of Ce-Lap is investigated to determine
5+
Ce-Lap catalysts are mainly due to the redox reaction of V .
And the redox reaction of cerium species is responsible for the
poor catalytic activity of Ce-Lap.
2
the influences of vanadium species on H S selective oxidation.
In order to further explore the catalytic mechanism, the O 1s
XPS spectra of 5% V O /Ce-Lap are provided in Fig. 12. For the
2
5
fresh catalyst, the spectrum is dominated by the peak that
appeared at 532 eV, which is related to the oxygen chemically
adsorbed on the oxygen vacancies. While a small shoulder peak
that appears at around 530 eV should be attributed to the lattice
2
1
oxygen. The two peaks remain after reaction. However, the
intensity of the peak at 532 eV decreases obviously. Moreover,
it is particularly important to observe that the variation trend of
H S conversion and sulfur yield is extraordinarily consistent
2
with that of the chemically absorbed oxygen on oxygen vacancies
(
Fig. 9), when V O loading increases from 3% to 8%. Thus, it
2 5
can be concluded that the large amount of chemically absorbed
oxygen on oxygen vacancies, which is caused by the interaction
of vanadium species with cerium species, plays an important
2
role in H S selective oxidation.
The S 2p XPS spectrum of used 5% V
2 5
O /Ce-Lap is presented
to study the catalyst deactivation mechanism. As shown in
Fig. 13, two peaks at around 164 eV and 169 eV are detected, which
2
ꢀ
35
can be assigned to the presence of SO
and S respectively.
n
4
ꢀ
2
Moreover, the area percentage of SO
4
and S
n
is 65% and 35%,
respectively. This demonstrates that the surface sulfur species
mainly exist in the form of sulfate species and perhaps are
3
6
4 4 2 2 4 3
VOSO , Ce(SO ) and Ce (SO ) . As is known, the sulfate species
associated with elemental sulfur deposited on the surface can
cause the deactivation of the catalyst. However, the actual content
of each sulfur species is unknown. According to the stepwise
mechanism, the slower oxidation rate of the catalyst active phases
2
by oxygen as compared to the reduction rate by H S also leads to
10
catalyst deactivation. Therefore, the detailed reason for catalyst
deactivation is not very clear.
Fig. 10 Effect of temperature on (A) conversion of H
/Ce-Lap catalysts.
2
S and (B) sulfur yield for
2 5
V O
Catal. Sci. Technol.
This journal is c The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
4
.3 Regeneration experiments
In order to further explain the deactivation mechanism and
investigate the regenerability of V /Ce-Lap catalysts, a set of
successive deactivation–regeneration experiments were carried
out on the 5% V /Ce-Lap catalyst at 180 1C. In the experi-
2 5
O
2 5
O
ment, after each deactivation step (reaction for 7 h, the sulfur
yield reduces to 85%), prior to regeneration, the catalyst was
purged in an N atmosphere at 300 1C for 30 minutes, then 2%
2
O was fed for 90 minutes. The profiles of regeneration experi-
2
2
ments are shown in Fig. 11. It can be seen that the H S
conservation decreases with reaction time, but it is still higher
than 96%. Nevertheless, the sulfur selectivity remains constant
2
at 99%. Thus, similar trends are observed for H S conversion
and sulfur yield, even the cycle is repeated three times. This
suggests that the catalytic activities are highly reproducible and
the catalysts possess good regenerability.
Fig. 12 O 1s XPS spectra of (a) the fresh 5% V
2
O
5
/Ce-Lap catalyst and (b) the
used 5% V O /Ce-Lap catalyst.
2 5
As is known to all both vanadium sulfate and cerium sulfate
4
,18
are hardly oxidized because VOSO
4
is less active
4 2
and Ce(SO )
4
+
and Ce (SO can disrupt the redox reaction between Ce
and Ce due to their high thermal stability. Thus, the good
regenerability reveals that the slower oxidation rate of the
2
4 3
)
3
+
37
2
catalyst by oxygen as compared to the reduction rate by H S
2 5
Fig. 13 The S 2p XPS spectrum of the used 5% V O /Ce-Lap catalyst.
is the main reason for catalyst deactivation. However, the
3
+
increase of Ce content can create more oxygen vacancies,
which is favorable for the oxidation reaction. The durability
behavior of the 5% V O /Lap catalyst is established to deter-
2
5
3
+
mine the influence of Ce . As shown in Fig. 11(A), for the 5%
/Lap catalyst, the H S conversion decreases sharply with
reaction time. After reaction for 5 h, the sulfur selectivity
remains at 100%. However, the H S conversion drops to 90%,
which is much lower than that of the 5% V /Ce-Lap catalyst,
indicating that the 5% V O /Lap catalyst is deactivated faster
V
2
O
5
2
2
2 5
O
2
5
3
+
than the 5% V O /Ce-Lap catalyst. Thus, the existence of Ce
2
5
can provide resistance to the catalyst deactivation obviously.
5. Conclusions
A series of V O /Ce-Lap catalysts with mesoporous structure
2
5
and high specific surface area were prepared. The structure
of V /Ce-Lap catalysts was partially damaged due to the
collapse of the clay layer. However, the surface oxygen property
2 5
O
Fig. 11 Conversion of H
2 2 5
S (A) and sulfur selectivity (B) obtained with 5% V O /
Ce-Lap at 180 1C in 3 cycles with successive regenerations (reactant composition
ꢀ1
in vol%: H
2
S/O
2
/N
2
= 5/2.5/92.5 at 7000 h of GHSV).
was enhanced and favorable for H
2
S selective oxidation.
This journal is c The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Moreover, vanadium species are mainly presented in a highly 13 C. He, Q. Li, P. Li, Y. F. Wang, X. Y. Zhang, J. Cheng and
5
+
4+
dispersed state of isolated V and polymeric V
2
O
5
, while Ce
Z. P. Hao, Chem. Eng. J., 2010, 162, 901–909.
3
+
and Ce coexisted in the catalyst. The catalysts exhibited high 14 B. M. Reddy, I. Ganesh and B. Chowdhury, Catal. Today,
catalytic activities and sulfur selectivity in the temperature range
of 120–220 1C due to the redox reaction of highly dispersed V
species. And the catalytic mechanism obeyed a redox process, i.e.
1999, 49, 115–121.
15 J. Arfaoui, L. K. Boudali, A. Ghorbel and G. Delahay, J. Phys.
Chem. Solids, 2008, 69, 1121–1124.
5
+
a stepwise mechanism. Furthermore, the chemically absorbed 16 B. G. Mishra and G. R. Rao, Microporous Mesoporous Mater.,
oxygen also plays an important role in H S selective oxidation.
2004, 70, 43–50.
Among them, the 5% V
/Ce-Lap catalyst presented the best 17 E. A. Khudaish, Sens. Actuators, B, 2008, 131, 223–229.
catalytic activity and excellent regenerability at 180 1C. It is worth 18 M. D. Soriano, J. Jimenez-Jimenez, P. Concepcion, A. Jimenez-
2
2 5
O
4
+
mentioning that Ce also participated in the oxidation reaction.
The catalyst deactivation was mainly due to the slower oxidation
Lopez, E. Rodriguez-Castellon and J. M. L. Nieto, Appl.
Catal., B, 2009, 92, 271–279.
rate of the catalyst active phases by oxygen as compared to the 19 E. P. Reddy, L. Davydov and P. G. Smirniotis, J. Phys. Chem.
3
+
reduction rate by H
2
S. The existence of Ce could provide
B, 2002, 106, 3394–3401.
resistance to the catalyst deactivation obviously.
20 Q. Li, X. Hou, H. Yang, Z. Ma, J. Zheng, F. Liu, X. Zhang and
Z. Yuan, J. Mol. Catal. A: Chem., 2012, 356, 121–127.
2
1 S. Damyanova, C. A. Perez, M. Schmal and J. M. C. Bueno,
Appl. Catal., A, 2002, 234, 271–282.
Acknowledgements
This work was financially supported by the National High 22 L. Chen, J. Li and M. Ge, J. Phys. Chem. C, 2009, 113,
Technology Research and Development Program of China
21177–21184.
No. 2012AA063101), the National Basic Research Program of 23 L. Chen, J. Li and M. Ge, Environ. Sci. Technol., 2010, 44,
(
China (No. 2010CB732300), National Natural Science Founda-
9590–9596.
tion (21007082) and Science Promotion Program of Research 24 L. Chen, J. Li, M. Ge and R. Zhu, Catal. Today, 2010, 153, 77–83.
Center for Eco-environmental Sciences, CAS (YSW2013B05).
25 L. Jing, Z. Xu, X. Sun, J. Shang and W. Cai, Appl. Surf. Sci.,
001, 180, 308–314.
2
2
6 S. Pitk ¨a aho, L. Matejova, K. Jiratova, S. Ojala and
R. L. Keiski, Appl. Catal., B, 2012, 126, 215–224.
References
1
2
3
4
5
6
7
K. V. Bineesh, D. K. Kim, D. W. Kim, H. J. Cho and 27 F. Klose, T. Wolff, H. Lorenz, A. Seidel-Morgenstern,
D. W. Park, Energy Environ. Sci., 2010, 3, 302–310.
V. V. Shinkarev, A. M. Glushenkov, D. G. Kuvshinov and
G. G. Kuvshinov, Carbon, 2010, 48, 2004–2012.
K. V. Bineesh, D. K. Kim, M. I. Kim, M. Selvaraj and
D. W. Park, Dalton Trans., 2011, 40, 3938–3945.
Y. Suchorski, M. Pi o´ rkowska and H. Weiss, J. Catal., 2007,
247, 176–193.
28 A. Gervasini, P. Carniti, J. Ker ¨a nen, L. Niinist ¨o and
A. Auroux, Catal. Today, 2004, 96, 187–194.
29 C. T ´e llez, M. Abon, J. A. Dalmon, C. Mirodatos and
J. Santamar ´ı a, J. Catal., 2000, 195, 113–124.
M. Y. Shin, D. W. Park and J. S. Chung, Appl. Catal., B, 2001,
3
0, 409–419.
30 S. Besselmann, C. Freitag, O. Hinrichsen and M. Muhler,
Phys. Chem. Chem. Phys., 2001, 3, 4633–4638.
31 L. Zhu, J. Yu and X. Wang, J. Hazard. Mater., 2007, 140, 205–210.
K. V. Bineesh, D.-K. Kim, M.-I. L. Kim and D.-W. Park, Appl.
Clay Sci., 2011, 53, 204–211.
K. V. Bineesh, S. Y. Kim, B. R. Jermy and D. W. Park, J. Ind. 32 Z. Zhao, X. Yang and Y. Wu, Appl. Catal., B, 1996, 8, 281–297.
Eng. Chem., 2009, 15, 207–211.
H. Y. Zhu, J. C. Zhao, J. W. Liu, X. Z. Yang and Y. N. Shen,
Chem. Mater., 2006, 18, 3993–4001.
H. Y. Zhu and G. Q. Lu, Langmuir, 2001, 17, 588–594.
J. J. Li, Z. Jiang, Z. P. Hao, X. Y. Xu and Y. H. Zhuang, J. Mol.
Catal. A: Chem., 2005, 225, 173–179.
33 K. V. Bineesh, S. Y. Kim, B. R. Jermy and D. W. Park, J. Mol.
Catal. A: Chem., 2009, 308, 150–158.
34 B. Fleutot, H. Martinez, B. Pecquenard, J. B. Ledeuil,
A. Levasseur and D. Gonbeau, J. Power Sources, 2008, 180,
836–844.
35 G. A. Bukhtiyarova, V. I. Bukhtiyarov, N. S. Sakaeva,
V. V. Kaichev and B. P. Zolotovskii, J. Mol. Catal. A: Chem.,
2000, 158, 251–255.
8
9
1
1
1
0 S. Yasyerli, G. Dogu and T. Dogu, Catal. Today, 2006, 117,
271–278.
1 S. S. Lin, C. L. Chen, D. J. Chang and C. C. Chen, Water Res., 36 N. M. A. Coelho, G. M. da Cruz and R. Vieira, Catal. Lett.,
002, 36, 3009–3014. 2012, 142, 108–111.
2 A. A. Davydov, V. I. Marshneva and M. L. Shepotko, Appl. 37 W. Q. Xu, H. He and Y. B. Yu, J. Phys. Chem. C, 2009, 113,
Catal., A, 2003, 244, 93–100. 4426–4432.
2
Catal. Sci. Technol.
This journal is c The Royal Society of Chemistry 2013