The Mechanism of The Catalytic Oxidation of Hydrogen Sulfide
The Mechanism of The Catalytic Oxidation of Hydrogen Sulfide
The kinetics of the catalytic oxidation of hydrogen sulfide by molecular oxygen have been
studied in the temperature range 20-250°C. The primary reaction product is sulfur which may
undergo further oxidation to SO9 at temperatures above 200°C.
From the kinetics of this autocatalytic reaction we derived an oxidation-reduction mecha-
nism. The two rate influencing steps are the chemisorption of oxygen and the reaction between
dissociatively chemisorbed HsS and chemisorbed oxygen.
The high activation energy for the formation of SO2 (120 kJ mol-1) explains the high
selectivity towards sulfur, although SO2 is thermodynamically the most favored product. At
temperatures above 3OO”C, where the formation of SO2 occurs readily, the SO2 may be an in-
termediate in the reaction of H2S with 02 leading to S and HzO.
87
TABLE 1
Survey of Literature Data on the Kinetics of HIS Oxidation
3. Puri et al. (6) a. Active charcoal a. T, 12&240 a. I.+ = 25.1 a + b. A part of the sulfur
SBET = 30&400 m*/g HtS, 5* no2 = 0.1 formed is bound irreversi-
02, 5-80 bly by the carbon
b. Carbon black b. T, 120-240 b. Ea = 12.6-16.8
SBET = 25-45 ma/g HzS, 5b
02, 30
0 Energy of activation calculated from reaction rates and not from rate constants.
b The concentration is varied around this value.
c We calculated that the rates were largely determined by the diffusion rate of HxS through the hangdown tube of the thermo-
balance to the catalyst sample.
150°C. It is known, however, that under under near differential conditions ; the
Cariaso’s conditions Ss is the main S-species highest conversions of H&S or O2 were 0.25.
(8). Summarizing, it can be concluded that Also the sulfur, deposited during reaction,
from the literature data it is not possible was equally distributed over the catalyst
to prefer one mechanism. bed. The uniform yellow color of the bed in
The present paper describes experiments the case of a zeolite catalyst pointed to this
on concentration and temperature depen- equal distribution of sulfur. Temperature
dences of the H&S oxidation rate on different effects were negligible because the maxi-
porous materials. A mechanism is proposed mum (i.e., adiabatic) increase in tempera-
which covers most of the kinetic data. The ture amounts to only 3°C in the case of 10%
oxidation of sulfur was studied to investi- conversion of 0.5 ~01% HZS.
gate the consecutive reaction leading to SO2
and SO3. The role of SO2 as an intermediate Catalysts
is also discussed. In Table 2 the catalysts used are listed
and some properties relating to their pore
METHODS
structure are given. The surface area was
Apparatus
calculated by applying the BET equation
The fixed-bed reactor system and the to argon adsorption data obtained at
thermobalance have already been described -196°C. The pore volume was calculated
(1). All our experiments were carried out from the difference between the apparent
HYDROGEN SULFIDE OXIDATION. II 89
TABLE 2
Catalyst Data
QThis catalyst was extracted with dilute sulfuric acid to remove the inorganic impurities.
catalyst volumes in mercury and methanol unfilled and from the outer surface of the
at room temperature. part’icles. In case .a material possesses a
Carbon molecular sieve (CWS) is a very high intrinsic activity compared to the
pure carbon with a pore radius mainly in autocatalytic activity, deposition of sulfur
the range of lo-12 A. The material is pre- lowers the activity and branch A is not
pared by carbonization of vinylidene chlo- observed.
ride polymer. For experiments in the regions A or B
where the activity did not attain a station-
RESULTS ary level as a function of reaction time, the
Kinetic experiments were carried out not rate data are always compared at the same
only on different materials (Table 2) but sulfur content. The sulfur content is cal-
also at different sulfur content,s of these culated by integration of the HZS oxidation
porous materials. As we already pointed out rate in the fixed-bed experiments. The de-
in Part I (I) the oxidation rate of H2S can sorption rate of sulfur is in this case much
be strongly dependent on the amount of smaller than the oxidation rate.
sulfur deposited in the micropores. For At a stationary activity level (region C),
materials which have a low intrinsic oxida- the sulfur content of the material does not
tion activity (= initial oxidation rate), the change further. For some materials it is
characteristic course of the activity as a even possible to vary concentrations and
function of sulfur uptake by the material temperature in a limited region without
is given in Fig. 1. There are three distinct
ranges, A, B, and C. In range A the activity
increase wit.h reaction time is due to the
autocatalytic activity of the sulfur de-
posited in the pores. In range B a decrease
of the activity is observed because pores
become filled up with sulfur which results
in a considerable decrease of the specific
surface area of sulfur. A stat,ionary activity
level is reached in range C, when the
amount of sulfur formed equals the amount FIG. 1. Catalytic activity as a function of the
of sulfur desorbed from t’he large pores left amount of sulfur deposited in the porous material.
90 STEIJNS ET AL.
TABLE 3
Rate Dependence on the Oxygen Concentration
Norit RBWI 0.12 180 1.0 0.15 -0.63 70 02 > 0.35, 0.0
% 02 < 0.35, 0.85
Carbon (sugary 0.70s 130-150 > 0.30 0.65 -1.76 0.23 f 0.10
0.7w 170-190 1.68 0.07 -0.57 0.79 f 0.10
changing the amount of adsorbed sulfur. the conversions of oxygen and hydrogen
Such a material is molecular sieve 13X, sulfide. At temperatures below 200°C the
where the amount of sulfur adsorbed cor- selectivity towards sulfur is always above
responds with the volume filling of super- 90% assuming that the main by-product is
cages in the aluminosilicate framework. In SO&
a separate experiment we found that the
relative sulfur pressure (P& can be varied Oxygen Concentration Dependence of the Rate
in the range Prel = 0.0054.25 without
changing the amount of adsorbed sulfur A part of the results is summarized in
significantly. Table 3. The amount of sulfur adsorbed on
In the case of the fixed-bed experiments the. catalyst does not affect the reaction
the kinetic parameters can be derived from order significantly. The oxidation rate of
H2S is, however, a function of the sulfur
r -- charge of the material.
The oxidation rate on zeolite 13X shows
a first order dependence on the concentra-
tion over a wide concentration and tempera-
ture range. Only at a high excessof oxygen
(l&SO times the stoichiometric amount)
can a lower order be observed.
For the active carbons the observed re-
d P ( N m?)
02 action order varies from zero to one. As a
typical example the rate on activated
FIQ. 2. Oxygen concentration dependence of the
rate on activated charcoal prepared from sugar. charcoal prepared from sugar is plotted in
Catalyst at stationary level; T = 170°C; Pn,s Fig. 2 as a function of oxygen concentra-
=6,8 X lo2 N rnM2;sulfur content: 0.7 g (g carbon)-‘. tion. The transition of first order to zero
HYDROGEN SULFIDZ OXIDATION. II 91
TABLE 4
Rate Dependence on the Hydrogen Sulfide Concentration
order was also observed for other types of the observed differences in reaction orders
activated charcoal. The activated charcoal are noted, the conclusion can be drawn that
prepared from sugar was free of inorganic the mechanism is not, the same for all the
impurities and available in batches of uni- materials. In Part, I (1) we reported on the
form composition. The autocatalytic effect autocatalytic effect of sulfur in different
of sulfur was very pronounced for this materials. Therefore, it is very likely that
material, which made it especially attrac- sulfur adsorbed in the micropores of the
tive for kinetic studies. 13X molecular sieve catalyzes the oxidation
reaction in the same manner as sulfur ad-
Hydrogen Sulfide Concentration Dependence sorbed in porous carbons. Any mechanism
of the Rate proposed should be in accordance with the
Table 4 gives a survey of the results. A observed variation of reaction orders, and
zero order dependence was observed for the an oxidation-reduction mechanism can
oxidation rate on zeolite 13X and r-A1203.
For the active carbons the order varies
from approximately 0.5 at low H2S concen-
trations to zero at high concentrations.
In Fig. 3 the oxidation rate on activated
carbon prepared from sugar is plotted
versus the square root of the hydrogen
sulfide pressure.
DISCUSSION
Oxidation-Reduction Mechanism for the
Oxidation of HsS FIG. 3. H& oxidation rate on activated sugar
charcoal as a function of the square root of the
When the results of the kinetic experi- partial pressure of HB. T = 190°C; PO, = 2.2
ments (Tables 3 and 4) are compared, and X lo-* N rnA2; sulfur content: 0.7 g (g carbon)-‘.
92 STEIJNS ET AL.
satisfy this condition. This mechanism of OsHcan be related to the partial pressure of
two chemical resistances in series has been HzS if there is-an equilibrium between H,S
observed for many different oxidation reac- in the gas phase and dissociatively ad-
tions, e.g., on vanadium oxide (9a, b). The sorbed H,S :
two reactions involved in the oxidation of +s
hydrogen sulfide could be : H$(g) G Hkb, % 2&H. (4)
hydrogen sulfide + oxidized site + The equilibrium constant (K) for reaction
products + reduced site, (1) (4) is:
&Hz
reduced site + oxygen -+ K=
oxidized site. (2) P H2S.(1 - es&
If these two chemisorption steps occur in If &H << 1 and the equilibrium is estab-
series then both reaction orders should vary lished then :
between zero and one, depending on the 6s~ = K *PH,S”~5 (5)
partial pressures of oxygen and H$. Our
experiments show that this is only true for Substituting Eq. (5) in (3) :
the order in oxygen but not for the order in
HzS. The maximum half order in H,S ob- dpHzs
--= k red ‘.P HtSo’5’ e0. (6)
served indicates that H2S is not directly dt
involved in the reduction step [Eq (a)].
A dissociative adsorption of H2S is likely. For the oxidation step the rate of reoxida-
We launched this idea in Part I (1). From tion of sites reduced by H2S is proportional
the literature it is known that the reaction to the oxygen partial pressure and the
of H2S with sulfur forming polysulfides fraction of sites not covered by oxygen, i.e.,
(H&L, z 1 2) proceeds very fast already at
room temperature (10). Golyland et al. (10) -dPo,
___ = koxvoz~ (1 - e,). (7)
observed a high rate of isotopic exchange dt
between radioactive H2S and sulfur de-
posited in the pores of active carbon: For the oxidation of one mole of H&3 to
sulfur a half mole of oxygen is required, so
Saolid+ HzS* + H$S* ti HZS + Ssolid*, that:
asterisk = radioactive. -*.51p,, - -dpoz. (8)
The maximum half order in H2S is in dt dt
accordance with a reduction step which is
a surface reaction between adsorbed oxygen In the steady state :
and dissociatively adsorbed hydrogen sul-
o.5~k~,d’~PH,,D~5~&,
fide. Although the formation of water from
adsorbed oxygen occurs in several steps, = koxvoz. (1 - e,). (9)
only one of these steps is the rate-determin- Solving Eq. (9) for e. and substituting in
ing reduction step. Assuming that the rate (6) gives:
of this reduction reaction is proportional to
the degrees of+ occupation of the specific - dPHzS
sites on the sulfur surface with oxygen and dt
hydrogen sulfide, respectively, 00 and ASH,
then 1
=
df’m
--= k red I, .BSH’f30. (3) (l/&d’. PH,s”.y + (1/2Jc,,‘~PO,)’
dt (10)
HYDROGEN SIJLFIDE OXIDATION. II
is not the only reaction which may occur at formation of SO2 is a slow react.ion. The
the catalyst surface. Especially at high main reason is the relatively high energy of
temperatures (2’ > 200°C) and high excess activation for sulfur oxidation. Wiewio-
of oxygen a consecutive reaction occurs : rowski and Slater (11) assumed that this
activation energy is related to the dissocia-
$S,+ Ot-tS02 400K: Kp
tion energy of a terminal sulfur-sulfur bond
= 4 x 1039. (13) in a sulfur chain molecule. This assumption
From thermodynamic calculations we points to a reaction of oxygen with the ends
learned that at 400 K this SOa can be of sulfur chains, which contain the unpaired
further oxidized to SOL We will not con- electrons.
sider this reaction because in practice this Physically adsorbed SOZis postulated as
compound is only formed in traces. The an intermediate in H2S oxidation on
SO2formed according to Eq. (13) will react -r-A1203 at 100°C by Prettre and Sion (4).
with unconverted H,S via the fast reaction: The following reactions should occur :
2HzS + SO2F? 3% 2HzS + 302 +
+ 2HzO 400 K : Kp = 6 X 10’. (14) 2SOz(ads) + 2HzO (slow), (15)
From the equilibrium constants calculated
2H.8 + SOz(ads) e
for the different reactions it can be con-
eluded that the possibility exists that SO2 $S8+ 2Hz0 (fast). (16)
is an intermediate.
Here a direct oxidation of H&S to SO2
To check this possibility we performed a
occurs, something which is not very likely
few experiments on the oxidation of ad-
at 100°C. At excessoxygen and a high con-
sorbed elemental sulfur by molecular oxy-
version level SOz should appear in the off
gen. In the temperature region lOO-200°C
gases. Prettre and Sion reported, however,
we observed that the formation rate of
the complete absence of SO, in the gas
sulfur dioxide was at least a factor 100
leaving the reactor even with OS/H8 = 3.
smaller than the oxidation rate of hydrogen
For this reason we doubt that there is
sulfide. At temperatures above 3OO’C the
physisorbed SO2 present at the surface of
order of magnitude of the rates is equal.
alumina which is an intermediate.
This is because the energy of activation for
There still remains the question as to
S oxidation (125 f 10 kJ mol-I) found by
what is the nature of the product of the
us is much higher than that for HzS oxida-
reaction between sulfur and oxygen. We
tion (-40 kJ mol-I). The energy of activa-
discuss the nature of the oxidized catalyst
tion for oxidation of adsorbed sulfur has the
site S,O, in Part III (13).
same value as for pure liquid sulfur, i.e.,
literature values: 120 kJ mol-’ (11) and
CONCLUSIONS
116 kJ mol-l (12).
Combining the thermodynamic data with 1. Kinetic data for the oxidation of H8
the kinetic data on sulfur and hydrogen on various materials agree with the kinetics
sulfide oxidation it can be concluded that of an oxidation-reduction model.
at temperatures above 300°C SO2 may be 2. The oxidation step is a reaction be-
an intermediate in the reaction. Below tween molecular oxygen and adsorbed sul-
200°C the rate of sulfur oxidation is too low fur. The reduction step is a reaction be-
to make a contribution to H&S oxidation. tween dissociatively adsorbed H&S and a
The high selectivity of H&S oxidation to- chemisorbed oxygen, yielding water and
wards sulfur at temperatures below 200°C sulfur. The kinetics are formulated in Eq.
and even excess oxygen implies also that (11).
HYDROGEN SULFIDE OXIDATION. II 95
3. Below 200°C the formation of SO2 is 3. Brodsky, M., and Pagny, P., Bull. Sot. Chim.
slow compared with the oxidation of H2S Fr. 15, 582 (1948).
4. Prettre, M., and Sion, R., Z. Elektrochem. 63,
into sulfur. This causes the high selectivity
166 (1959).
towards sulfur. Above 300°C cleavage of 5. Puri, B. R., Kumar, G., and Kalra, K. C.,
S-S bonds occurs readily and an alternative Indian J. Chem. 9, 970 (1971).
route wherein SO2 acts as an intermediate 6. Sreeramamurthy, R., PhD thesis, Andhra Univ.,
begins to contribute to the overall reaction 1972.
rate. 7. Cariaso, O., PhD thesis, Pennsylvania State
Univ., 1972.
ACKNOWLEDGMENTS 8. Gamson, B. W., and Elkins, R. H., Chem. Eng.
We thank Mr. A. Peppelenbos for carrying out Prog. 49, 203, No. 4 (1953).
part of the experimental work. We gratefully 9. (a) Mars, P., PhD thesis, Delft, 1958 (in
acknowledge NORIT Laboratories (Amersfoort,, Dutch); (b) Dixon, J. K., and Longfield,
The Netherlands) for the preparation of some acti- J. E in “Catalysis,” (P. H. Emmett, Ed.),
vated carbon samples. The investigations were sup- Vol. ‘;. Reinhold, New York, 1960.
ported by Chemical Research Netherlands (SON) 10. Golyland, S. M., Krapivina, T. K., and Lazarev,
with the financial aid of the Dutch Organization for V. I., Russ. J. Phys. Chem. 36, 700 (1966).
Pure Research (Z WO) . 11. Wiewiorowski, T. K., and Slater, A. L., J.
Phys. Chem. 71, 3014 (1967).
REFERENCES 12. Norrish, R. G. W., and Rideal, E. K., J. Chem.
1. Steijns, M., and Mars, P., J. Catal. 35, 11 (1974). Sot. 123, 502 (1923).
2. Sreeramamurthy, R., and Menon, P. G., J. 13. Steijns, M., Koopman, P., Nieuwenhuijse, B.,
Catal. 37, 287 (1975). and Mars, P., J. Catal. 42, 96 (1976).