Catalytic conversion of COS and CS2 under an excess of sulfur dioxide

Article abstract of DOI:10.1134/S1070427208100042

Conversion of organosulfur components (COS and CS₂) in catalytic processing of a real reduced sulfurous gas was experimentally studied.

Full text of DOI:10.1134/S1070427208100042

ISSN 1070-4272, Russian Journal of Applied Chemistry, 2008, Vol. 81, No. 10, pp. 1725–1728. © Pleiades Publishing, Ltd., 2008.
Original Russian Text © O.I. Platonov, L.Sh. Tsemekhman, 2008, published in Zhurnal Prikladnoi Khimii, 2008, Vol. 81, No. 10, pp. 1597–1600.
INORGANIC SYNTHESIS AND INDUSTRIAL
INORGANIC CHEMISTRY
Catalytic Conversion of COS and CS under an Excess
2
of Sulfur Dioxide
O. I. Platonov and L. Sh. Tsemekhman
Gipronikel’ Institute Limited Liability Company, St. Petersburg, Russia
Received May 19, 2008
2
Abstract—Conversion of organosulfur components (COS and CS ) in catalytic processing of a real reduced
sulfurous gas was experimentally studied.
DOI: 10.1134/S1070427208100042
At present, about 40 million tons of sulfur is an-
nually produced in the world in purification of dis-
charge gases by the Claus method in natural-gas- and
oil-processing industries [1]. It is prognosticated [2]
that, with the environment protection regulations be-
coming increasingly stringent and requiring higher
ecological payments, the volume of accompanying
sulfur used to manufacture fertilizers will grow via
utilization of metallurgical sulfurous gases. The only
working technology for processing of these gases into
sulfur is reduction with hydrocarbons, followed by
catalytic conversion by the Claus method [3]. Because
of the specific features of reduced sulfurous gas, its
catalytic conversion strongly differs from that in the
classical Claus technique [1, 4]. This circumstance de-
termines the topicality of its analysis with account of
parameters of a gas being processed.
The aim of our study was to experimentally ana-
lyze specific features of the conversion of organosulfur
components (COS and CS ) in catalytic processing of a
2
reduced sulfurous gas.
EXPERIMENTAL
We performed tests in the temperature range 220–
600°C on a pilot installation with a flow rate of the gas
3
–1
mixture equal to 0.6–0.7 standard m h at Gipronikel’
Institute Limited Liability Company. We studied two prom-
ising vanadium-containing catalysts: titanium oxide cata-
2
–1
lyst AOK-78-57 with a specific surface area of 145 m g
(extrudates 4–5 mm in diameter), manufactured by
Katalizator Limited Liability Company (Novosibirsk),
and a new VA-2 catalyst on an alumina support, devel-
oped at the Institute of Catalysis, Siberian Branch,
Russian Academy of Sciences, and manufactured by
Novomichurinsk Catalyst Plant Private Company in
the form of spherical grains 4–5 mm in diameter and
A characteristic feature of the reduced gas is the
increased content of organosulfur compounds (CS and
2
2
–1
active surface area of no less than 240 m g .
COS) formed from carbon introduced with the reduc-
ing agent, which predetermines the priority role of
their conversion, traditionally the most difficult stage
in catalytic processing by the Claus method [1, 4]. At
large contents of sulfur dioxide, characteristic of the
reduced sulfurous gas [5], the commonly taken into
The tests were made in a vertical stainless steel
reactor with external electric heating (catalyst-
containing cartridge with an inner diameter of 53 mm
and height of 175 mm) and a catalyst charge of
3
0.38 dm . The temperature was controlled and re-
account mechanism of COS and CS hydrolysis [1, 3]
experiences competition from the alternative channel
of conversion by reactions of sulfur dioxide reduction:
corded in the course of the experiments with a DT-600
digital data-acquisition device [Data Taker Pty. Ltd.
(Data Electronics)]. Metran 231-01 Chromel–Alumel
thermoelectric transducers (thermocouples) with a
sheath diameter of 3 mm were used as temperature
sensors. One of the thermocouples, buried to 4/5 of the
catalyst bed height, was also used to maintain the reac-
tor temperature; the electric power for heating the reac-
tor and gas heater was varied using TP 403 microproc-
2
CS
2
+ SO
2
= CO
2
+ 3/nS
n
,
(1)
(2)
2
COS + SO
2
= 2CO
2
+ 3/nS ,
n
whose contribution can be predominant under an ex-
cess of SO2.
1
725

1
726
PLATONOV, TSEMEKHMAN
Gas compositions before and after conversion on the VA-2
catalyst
(3)
Content, vol %
Gas
in
out
where c and c are the initial and final concentration
component
i
i
initial
at the outlet
in
of gas component i (i = H S, COS, SO ), and c and
c , the initial and final concentrations of nitrogen in
dried gas samples taken at, respectively, the inlet and
outlet of the catalytic reactor.
2
2
N
2
SO
CO
2
20.9 ± 6.5
<0.001
12.5 ± 8.0
0.00 ± 0.03
4.0 ± 3.2
21.5 ± 5.0
8.3 ± 7.0
1.2 ± 0.8
2.9 ± 2.6
3.8 ± 2.6
44.9 ± 6.1
out
N
2
H
2
3.7 ± 2.4
14.6 ± 1.9
6.1 ± 3.0
2.0 ± 0.97
6.2 ± 2.6
3.7 ± 2.7
43.8 ± 4.4
CO
2
2
H S
The gas composition was monitored at four points
of the pilot installation with an automated analytical
complex based on a gas analyzer, EMG 20–1 time-of-
flight mass spectrometer manufactured by METTEK
Limited Liability Company (St. Petersburg), which can
measure the volume concentrations of all the basic
components of the gas: H S, SO , SO , COS, CS , CO,
COS
CS
CH
2
4
N
2
essor heat-control units manufactured by Varta Re-
search and Production Company (St. Petersburg).
2
2
3
2
CO , H , CH , O , N , and Ar with a threshold sensitiv-
2
2
4
2
2
The coefficient (degree) of conversion of compo-
ity of 0.01–0.03%. The EMG 20-1 gas analyzer was
calibrated against reference gas mixtures prepared at
Scientific Instruments Open Joint-Stock Company and
Monitoring Limited Liability Company (both from
St. Petersburg).
nents of the gas mixture, η , was found using the for-
i
mula [6]:
In the course of the tests, a real sulfurous gas re-
duced in the homogeneous reactor of the pilot installa-
tion at a temperature of 1100°C in accordance with the
promising technology for processing of metallurgical
gases [7] was delivered to the inlet of the catalytic re-
actor (see table).
In all, the tests were made in 67 operation modes
for the VA-2 catalyst and 57 modes for the AOK-78-
5
7 catalysts. The modes differed in temperature at the
same initial composition of the gas (see table) and
–
1
volumetric flow rates W = 1500–1900 h .
The results obtained in the pilot tests show that
the high efficiency of conversion of organosulfur com-
ponents on the catalysts under study is high. For example,
this is indicated by comparison of the initial composition
of the gas with that at the outlet of the catalytic reactor,
found from a sample of measurements in 60 modes at a
3
–1
gas flow rate of 11 ± 1.1 standard dm min (W =
–
1
1
600–1900 h ), also represented in the table.
It can be seen that more than a half, on average, of
the organosulfur compounds (CS and COS) is con-
2
verted in catalytic processing of a reduced sulfurous
gas. In the process, the content of carbon dioxide no-
ticeably increases (nearly twice, on average), which
points to the decisive contribution of reactions (1) and
Fig. 1. Relative change in the conversion of (a) carbonyl
sulfide, ηCOS, and (b) carbon disulfide, ηCS2, during tests
for a time τ at different temperatures of the VA-2 ca-
talyst. (6–8) Conversion on the AOK-78-57 catalyst. T (°C):
(
2) of sulfur dioxide reduction to the overall conver-
(
1, 6) 310, (2) 340, (3, 4) 380, (5) 450, and (7, 8) 250.
sion.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 10 2008

CATALYTIC CONVERSION OF COS AND CS
2
UNDER AN EXCESS OF SULFUR DIOXIDE
1727
We observed a rapid variation of the activities of
both the catalysts, which is manifested in that the con-
version grows during the first hour after a catalyst is
put into operation in the on-line mode. This effect is
illustrated by Fig. 1, in which the conversion coeffi-
cients of sulfur-containing components, η , normalized
i
to the maximum conversion of a given component, are
arranged in order of increasing time τ elapsed from
the beginning of a run. Different symbols in the fig-
ure show results of measurements in different runs, on
different catalysts, and at different process tempera-
tures.
Fig. 2. Conversion η of (1) carbon disulfide and (2) car-
i
bonyl sulfide on the VA-2 catalyst vs. temperature T at
–
1
a volumetric flow rate of 1600–1800 h .
The activity variation is manifested to the maxi-
mum extent in conversion of COS, whose relative con-
version coefficient ηCOS on the alumina catalyst VA-2
changes from –100% to 100% in 1 h (Fig. 1a). The con-
version of sulfur dioxide and the overall conversion of
sulfur-containing components varies qualitatively simi-
lar to ηCS2 . The conversion of hydrogen sulfide shows
no dynamic effect.
tested is their short-term poisoning via sulfation in the
warm-up and startup of the installation with the oxy-
gen-containing sulfurous gas. As follows from the re-
sults of our tests, the catalyst activity in conversion of
organosulfur compounds, low at the beginning of a
run, is completely restored during the first hour of on-
line operation in the medium of a reduced gas, which
contains reducing components (H , CO, and H S), but
A similar dynamic effect was observed on the
titanium oxide catalyst AOK-78-57, on which the con-
version of carbonyl sulfide varied under identical proc-
ess conditions, e.g., at T ≈ 420°C, from –80 to 76 rel %
2
2
is free of oxygen. The differences in the degrees of
initial sulfation of a catalyst in each startup predeter-
mine the observed statistical scatter of the empirical
(
Fig. 1).
values in the dynamic dependences η (τ).
i
The statistical scatter of values in the dynamic
dependences of the conversion of both the organosul-
As follows from the results of our experiments,
CS interacts with SO at least in two stages, with car-
bonyl sulfide formed in the first of these. The only
possible other intermediate in the first stage is, accord-
ing to the stoichiometric balance of the overall reaction
2
2
fur components, η (τ), is due to uncontrollable specific
i
features of each run (duration of catalyst acidification)
at its beginning. Nevertheless, the data scatter σ ob-
served in different runs is noticeably smaller than the
variance δη (τ): 3σ < δη (τ), which demonstrates that
(
1), sulfur oxide S O, as suggested by the two-stage
2
i
i
scheme of the process [6]:
the dynamic effect is revealed rather reliably.
(
4)
2 2 2 2 8
CS + SO = COS + [SO ] = CO + 1/8S .
The activities of the catalysts in CS and COS
2
conversion, determined after 1 h and more of uninter-
rupted operation, nearly coincide to within the experi-
mental error, even though a tendency is observed for
ηCS2 to somewhat exceed the conversion of carbonyl
sulfide, ηCOS, under the same conditions.
In the initial period of operation of a catalyst after
its sulfation, carbonyl sulfide is generated, which is
manifested in large negative values of its conversion
ηCOS (Fig. 1a). Together with the simultaneously high
conversion of carbon disulfide (Fig. 1b), this circum-
stance suggests that the second stage of carbonyl sul-
fide conversion is the rate-determining stage, the most
sensitive to the catalyst sulfation.
For illustration, Fig. 2 shows the average, over
temperature intervals, conversions of carbonyl sul-
fide, ηCOS, and carbon disulfide, ηCS2 , on the VA-2
catalyst as functions of temperature in the catalyst bed.
The vertical portions correspond to the variances of
the averages, and the horizontal portions (defined as
variance) are proportional to the width ∆T of the aver-
aging interval.
An alternative explanation of the generation of
carbonyl sulfide, e.g., that by interaction on the cata-
lyst with sulfur vapor:
CO + 1/nS
n
= COS,
(5)
A possible reason for the appearance of the dy-
namic dependence of the activities of the catalysts
would not do because of the insufficient initial
concentration of carbon monoxide (see table). No con-
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 10 2008

1
728
PLATONOV, TSEMEKHMAN
version of methane (also present in the starting gas),
which could serve as a reactant for COS formation,
was observed in our experiments within the experi-
mental error limits.
oxide. An intermediate formed in oxidation of carbon
disulfide is carbonyl sulfide, whose further conversion
is hindered by sulfation of the active centers of a cata-
lyst. As a result, the conversion of carbon disulfide
under an excess of SO is more complete than that of
2
According to the two-stage mechanism of carbon
disulfide conversion by reaction (4), with the rate-
determining second stage, the conversion of carbon di-
sulfide should be more efficient than that of COS, and
this, indeed, was observed in our experiments (Fig. 2).
This makes it possible (at least as regards the process-
ing of a reduced sulfurous gas) to reconsider the
known paradigm [1, 4] of the more efficient conver-
sion of COS. According to the results of this study, the
conversion of organosulfur compounds under an ex-
cess of sulfur dioxide predominantly occurs via redox
reactions with sulfur dioxide, of which the most effi-
cient is that with CS2.
carbonyl sulfide, which predetermines the principal
characteristic feature of the catalytic processing of a
reduced sulfurous gas.
(2) The experimental data obtained directly con-
firm for the first time the two-stage mechanism of car-
bon disulfide conversion by reaction (4) to give COS
as an intermediate product of CS conversion.
2
REFERENCES
1
2
. Grunval’d, V.R., Tekhnologiya gazovoi sery (Technology
of Gas Sulfur), Moscow: Khimiya, 1992.
. Okura, T., Ueda, S., and Yamaguchi, K., Sohn Int. Symp.
on Sulfide Smelting, San-Diego, USA, Aug. 27–31, TMS,
The fact that the dynamic effect is observed on
catalysts of different types (alumina catalyst VA-2 and
titanium oxide catalyst AOK-78-57) means that the
conversion of carbonyl sulfide is not directly related to
the properties of the active surface, but occurs via ad-
sorption of COS on active basic centers. By analogy
with the hydrolysis process [8], the role of centers of
this kind can be played by hydroxy groups, which are
subject to sulfation in the presence of oxygen.
2
006, vol. 8, pp. 425–431.
3
4
5
6
. Ushakov, K.I., Eremin, O.G., Kalnin, E.I., et al., Tsvet.
Met., 1980, no. 2, pp. 34–37.
. Pearson, M.J., Hydrocarbon Proc., 1973, vol. 52, no. 2,
pp. 81–85.
. Platonov, O.I., Zh. Prikl. Khim., 2007, vol. 80, no. 9,
pp. 1440–1444.
. Clark, P.D., Dowling, N.I., and Huang, M., Appl. Catal.
B: Environ., 2001, vol. 31, pp. 107–112.
CONCLUSIONS
7. Diachenko, V.T., Kotukhov, S.B., Kozyrev, V.F., et al.,
Sulphur 2007,” Conf. papers, Montreal, Canada, 28–31
“
(
1) If an excess of SiO is present in the gas being
October 2007, London: CRU Publ. Ltd., 2007, pp. 75–82.
2
processed, the conversion of organosulfur compounds
predominantly occurs via their oxidation by sulfur di-
8. Bachelier, J., Aboulayt, A., Lavalley, J.C., et al., Catal.
Today, 1993, vol. 17, pp. 55–62.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 10 2008