XPS study of the silica-supported Fe-containing catalysts for deep or partial H2S oxidation

Article abstract of DOI:10.1016/S1381-1169(00)00085-6

Catalytic properties of silica-supported Fe-containing catalysts in H₂S oxidation have been correlated with their surface composition characterized by XPS. This allows us to show that iron sulfate supported on silica is active in sulfur product

Full text of DOI:10.1016/S1381-1169(00)00085-6

Ž
.
Journal of Molecular Catalysis A: Chemical 158 2000 251–255
www.elsevier.comrlocatermolcata
XPS study of the silica-supported Fe-containing catalysts for
deep or partial H2S oxidation
G.A. Bukhtiyarova), V.I. Bukhtiyarov, N.S. Sakaeva, V.V. Kaichev, B.P. Zolotovskii
BoreskoÕ Institute of Catalysis, LaÕrentieÕa prospekt, 5, NoÕosibirsk, 630090, Russia
Abstract
Catalytic properties of silica-supported Fe-containing catalysts in H2S oxidation have been correlated with their surface
composition characterized by XPS. This allows us to show that iron sulfate supported on silica is active in sulfur production
with 100% selectivity, whereas the decrease in selectivity to sulfur is accompanied by the appearance of iron disulfide phase
in the catalyst. q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
Keywords: X-ray photoelectron spectroscopy XPS ; Fe-containing catalysts; H2 S oxidation
1. Introduction
In spite of long-standing interest of re-
searchers to both reactions, the structure of ac-
tive centers governed by partial or deep oxida-
tion of H2S is still a disputable point. Recently,
a great deal of attention is turned to Fe-contain-
ing supported systems. Thus, a number of au-
Oxidation of hydrogen sulfide to elemental
sulfur:
H2Sq1r2 O2 s1rn Sn qH2O
1
Ž .
is of practical importance being a widely-used
method for purification of technological gases,
such as Claus tail gas, natural gas, etc. Along
with selective oxidation, the deep oxidation of
hydrogen sulfide to sulfur dioxide may occur:
w
x
thors 2,3 have suggested that the selectivity
decreases due to iron sulfide, which catalyzes
Ž .
Eq. 2 , formed as a result of the catalyst modi-
fication by reaction medium. On the other hand,
the activity of catalysts containing Fe, Cu, Zn,
Co, or Ni sulfides in oxidation of hydrogen
H2Sq3r2 O2 sSO2 qH2O
2
Ž .
w x
sulfide to sulfur has been concluded in Ref. 4 .
to decrease the overall sulfur yield. On the other
To tackle this problem, we tried to prepare a
number of silica-supported iron-containing cata-
lysts for H2S oxidation using various precursors
and to select those which exhibit an essentially
different level of selectivity towards sulfur.
Then, their surface composition has been char-
acterized using X-ray photoelectron spec-
Ž .
hand, Eq. 2 is of significance itself in those
cases when SO2 produced is converted to valu-
able chemical products or fed back to the Claus
w x
plant 1 .
)
Ž
.
troscopy XPS .
Corresponding author.
1381-1169r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 00 00085-6

(
)
252
G.A. BukhtiyaroÕa et al.rJournal of Molecular Catalysis A: Chemical 158 2000 251–255
2. Experimental
and particle size effect indicated no diffusion
limitations under these conditions for the reac-
tion studied. The composition of the inlet gas
was: 2 vol. % of H2S, 0.5–3 vol.% of O2, 30
vol. % of H2O, and He to balance. Catalytic
properties of the samples were characterized by
the steady-state rate of H2S oxidation, and by
selectivity to sulfur at the concentrations in the
circulation loop being 1 vol.% H2S, 0.5 vol.%
O2. Concentrations of O2, H2S and SO2 were
measured using a chromatograph with TCD.
Before unloading the samples from the reactor
after catalytic testing, they were cooled in flow-
ing helium to room temperature.
Samples were prepared by impregnation of
Ž
.
SiO2 KSK-1, 0.25–0.5 mm with an aqueous
Ž
.
solution of iron III nitrate or with an aqueous
solution of iron II sulfate. After impregnation,
Ž .
the samples were dried at ambient conditions
and then the former sample was calcined at
Ž
.
5008C denoted as N500 and the latter one —
Ž
.
at 5508C S550 .
The concentration of Fe analyzed with a Baird
atomic-emission spectrometer using plasma
atomic-absorption method was about 10 wt.% in
both samples. To achieve this concentration in
the case of the S550 sample, multiple impregna-
tion was used.
X-ray photoelectron spectra were measured
using a VG ESCALAB HP electron spectrome-
3. Results and discussion
Ž
.
ter with MgKa irradiation hns1253.6 eV
Catalytic testing of H2S oxidation on the
catalysts reveals unstable catalytic properties of
and were calibrated against Si 2p spectrum with
w x
Eb s103.5 eV used as internal reference 5 .
Atomic element ratios to silicon were calculated
according to the following equation:
Ž
.
the sample prepared using Fe III nitrate aque-
Ž
.
ous solution N500 : concentration of SO2
among the reaction products increases con-
stantly in time. After 12 h, when the activity
and the selectivity reach the steady-state level,
SO2 is a predominant product of H2S oxidation
Ii
ni
ASFi
ISi
s
nSi
Ž
.
Table 1 . In contrast to N500, the sulfate-
ASFSi
Ž
.
originated catalyst S550 provides the stable
where ASFi and ASFSi are the atomic sensitiv-
100% selectivity of H2S oxidation to sulfur.
w x
ity factors taken from Ref. 5 , and Ii, ISi are
Fig. 1 shows the Fe 2p3 and S 2s core level
r2
Ž
the intensities of XPS lines for the analyzed
spectra for the fresh and treated after achieving
Ž .
.
elements i and silicon, respectively.
steady-state catalytic properties samples of the
Catalytic properties were tested at normal
pressure and 2508C in an external recirculation
reactor, using catalyst fraction of 0.25–0.5 mm.
Additional checking of circulation multiplicity
N500 catalyst, XPS spectra for the treated sam-
ple being measured two times: immediately af-
ter catalytic testing and after storage in the
atmosphere air. Choice of the S 2s core level
Table 1
Ž
.
Activity and selectivity of N500 and S550 samples in H2 S oxidation reaction temperature is 2508C
Sample
Pre-treatment conditions
Fe content,
mass %
Selectivity of H2 S
conversion to Sn
Reaction rate, H2 S
molecules gy1 sy1
N500
S550
S550
12 h in reaction mixture at 2508C
8 h in reaction mixture at 2508C
4 h in reaction mixture at 3008C and
2 h in reaction mixture at 2508C
10.3
10.1
11
100
9.0=1018
8.3=1018
9.4=1018
10.1
84

(
)
G.A. BukhtiyaroÕa et al.rJournal of Molecular Catalysis A: Chemical 158 2000 251–255
253
also into account the calcination temperature
and full absence of N1s signal, we suggest the
formation of Fe2O3.
The treatment of the fresh N500 sample in
reaction conditions for 12 h gives rise to Fe
2p3 signal at 707.4 eV and S 2s signal at
r2
226.8 eV, whereas the Fe 2p3 line with Eb s
r2
711.4 disappeared. Using the difference in S 2s
Ž
.
and S 2p binding energies see above we have
Ž
.
calculated the Eb S 2p value which is equal to
162.8 eV. The comparison of this value with the
y
2
y
table data for SO42
169.1"0.3 eV , SO3
Ž
.
2
y
Ž
.
Ž
.
Ž
162.9
166.8"0.2 eV , Sn 164.0 eV , S2
2
y
.
Ž
.
eV , S
161.9 eV allows us to assign the
observed S 2s line to sulfide ions, most proba-
bly to S22y. The Fe 2p3 binding energy of
r2
707.4 eV can be also assigned to iron disulfide.
Indeed, only metallic iron exhibits similar low
w x
value of binding energy 5 , while much higher
Eb values are characteristic of iron ions, for
example Fe 2p3 spectrum for iron sulfide,
r2
w x
FeS, is characterized by the line at 710.1 eV 5 .
The conclusion about the formation of FeS2 is
confirmed additionally by the quantitative XPS
data. FerSi atomic ratio, determined from XPS
intensities of the corresponding lines, is equal to
0.45, which is close to the stoichiometric ratio
for FeS2–0.5.
Ž .
Fig. 1. Fe 2p3r2 and S 2s spectra for the fresh
1
and treated
Ž
.
2,3 samples of N500 catalysts. The spectra of the treated sample
Ž .
2 and after
were recorded immediately after catalytic testing
Ž .
storage in the atmosphere air 3 for 2 months.
It should be noted, however, that FeS2 pro-
duced as a result of the reaction mixture treat-
ment is unstable. This conclusion is based on
the analysis of the XPS spectra for the air-stored
spectrum for analysis instead of the most inten-
sive sulfur line, S 2p, is explained by the over-
Ž
.
lapping of the S 2p region Eb ;160–169 eV
with one of the levels of the support — Si 2s
sample: Fe 2p3 signal is shifted to 711.4 eV
r2
Ž
.
Eb ;155–165 eV . To recalibrate the S 2s
and S 2s signal — to 233.7 eV. The calculation
Ž
.
binding energy value to the S 2p one, we have
of Eb S 2p by means of the procedure de-
measured both spectra for the FeSO4rAl2O3
scribed above gives the value of 169.7 eV char-
Ž
.
w x
sample and determined the difference Eb S 2p
acteristic of sulfate ions 5 . The shift of the Fe
Ž
.
yEb S 2s s64.0"0.2 eV. This recalibration
2p3 spectrum by 4.1 eV to higher binding
r2
was necessary to compare our data with the
energies is also in accordance with the forma-
tion of iron sulfate. Unfortunately, other meth-
ods, but not XPS, are necessary to study in
more detail the nature of the sulfate formed.
The same variations in the spectra are observed
if the treated sample is unloaded from the cat-
alytic reactor without cooling to room tempera-
ture.
w
x
reference ones 5,6 .
It is seen from Fig. 1 that the fresh sample is
characterized by a wide Fe 2p3
line with
r2
Eb s711.3"0.2 eV and the absence of S 2s
signal. The comparison of Fe 2p3 binding
r2
w x
energy with the reference data 5 allows us to
Ž
.
assign the Fe 2p3 line to Fe III ion. Taking
r2

(
)
254
G.A. BukhtiyaroÕa et al.rJournal of Molecular Catalysis A: Chemical 158 2000 251–255
XPS spectra of the same core levels for the
S550 catalyst recorded before and after its treat-
ments under the reaction conditions at different
Ž
.
temperatures 2508C and 3008C are compared
in Fig. 2. Positions of both spectra for the fresh
Ž
sample reflect its sulfate origin: both Eb Fe
.
Ž
.
Ž
.
2p3 s711.3 eV and Eb S 2p sEb S 2s y
r2
64.0s233.2y64.0s169.2 eV are in rather
good agreement with the data for FeSO4 P7H2O
w x
measured by Wagner et al. 6 . Thus, as op-
the lower is the value of FerSi atomic ratio
determined from XPS data. Then, the fresh
S550 sample contains the active phase particles
which are smaller in size than those of the
N500.
The treatment of the fresh S550 sample in
reaction conditions seems to result in the en-
largement of the iron-containing particles due to
sintering. This suggestion is based on a triple
decrease in the FerSi atomic ratio for the treated
sample which achieves the level of the corre-
posed to the N500 fresh sample, the calcination
of the impregnated S550 sample at 5508C does
not decompose the phase of the precursor. This
fact suggests that the silica-supported iron sul-
fate is more stable than the iron nitrate phase.
Another difference of the fresh S550 sample
from N500 is a higher FerSi atomic ratio calcu-
lated from XPS intensities in the former case:
Ž
.
sponding N500 sample 0.04 . At the same time,
the positions of the Fe 2p3 and S 2s spectra
r2
Ž
do not change in the reaction course Fig. 2,
.
curves 2 determining again the stability of the
sulfate phase. Only the elevation of the tempera-
ture of the reaction mixture treatment causes the
appearance of Fe 2p3 line at 707.5 eV and S
r2
2s line at 226.7 eV due to formation of the iron
Ž
.
disulfide phase Fig. 2, curves 3 . The formation
of FeS2 is accompanied by the appearance of
SO2 among the reaction products. As a conse-
quence, the selectivity of H2S oxidation to sul-
fur decreases, as seen from the data of Table 1.
Thus, XPS characterization of the catalysts
with essentially different levels of selectivity
towards sulfur shows that iron sulfate provides
the activity of the iron-containing catalyst to
selective oxidation of hydrogen sulfide to sul-
fur, whilst the appearance of SO2 among the
products of H2S oxidation is accompanied by
the formation of the iron disulfide phase on the
catalyst surface. FeS2 formed at the reaction
conditions is unstable. This is an important point
when Fe-containing catalysts for hydrogen sul-
fide oxidation are ex situ characterized by phys-
ical methods.
Ž .
Fig. 2. Fe 2p3r2 and S 2s spectra for the fresh
1
and treated
2,3 samples of S550 catalysts. The catalyst were treated at
Ž . Ž . Ž
Ž
.
.
different temperatures: 2508C 2 and 3008C 3 see Table 1 .

(
)
G.A. BukhtiyaroÕa et al.rJournal of Molecular Catalysis A: Chemical 158 2000 251–255
255
w x
4
O.N. Kovalenko, N.N. Kundo, V.M. Novopashina, D.P.
Ivanov, in: Proceed. Intern. Symp. Molecular Aspects of
Catalysis by Sulfides, St-Petersburg,1998, p. 113.
References
w x
Ž
.
1
w x
2
Keeping abreast of the regulations, Sulphur 231 1994 33.
P.H. Berben, A. Scholten, M.K. Titulaer, N. Brahma, W.J.J.
w x
5 Practical Surface Analysis, 2nd edn., D. Briggs, M.P. Seah
Ž
.
Van der Wal, J.W. Geus, Stud. Surf. Sci. Catal. 34 1987
Ž
.
Eds. , Auger X-ray Photoelectron Spectrosc. Vol. 1 Wiley,
303.
1990.
w x
3
A. Pieplu, A. Janin, O. Saur, in: Proceed. 1st World Congr. on
Environ. Catal. for Better World and Life, Pisa,1995, p. 303.
w x
6 C.D. Wagner, L.H. Gale, R.H. Raymond, Anal. Chem. 51
Ž
.
1979 466.