JOURNAL OF CATALYSIS 10, 1-12 (1968)

The Mechanism and the Kinetics of Sulfur Dioxide Oxidation

on Catalysts Containing Vanadium and Alkali Oxides

I’. MARS* AND J. G. H. MAESSEN

From the Central Laboratory, Staatsmijnen/DSM, Geleen, Netherlwds

Received June 22, 1967; revised September 18, 1967

In continuation of a previous study on a catalyst containing potassium and vanadium

oxides, it is shown that the oxidation of SO, on a sodium-potassium-vanadium catalyst
also proceeds via rapid establishment of the equilibrium, SO* -+- 2V5+ + O*- ti SOa +
2V4+ (equilibrium constant K) and that the reaction of oxygen with V4+ is the rate-deter-
mining step. On the basis of this finding it has been derived that t.he reaction rate can be
formulated as follows

~psohsos
1: = kpo,
11 + wPS0,1Ps0,)“*12
By means of isothermal and differential kinetic measurements on both types of catalysts
mentioned abve this expression has been checked and found correct. The value of the
equilibrium constant K calculated from the kinetic experiments is in good agreement
with that derived from analytical determinations.

I. I~YTRODUCTION absence of SOS,however (thus, at the start

Since the time when the platinum cata- of the reaction), SO2 has only a minor
lysts in the sulfuric acid industry were effect on the reaction rate. This is evident
superseded by vanadium-containing masses from the study by Boreskov, Ritter, and
(1920-1930), there has been a continuous Volkova (5), whose results are illustrated
stream of publications dealing with the in Kg. 1. These workers recorded the
kinet,ics of the process. The object of these temperature at which a given slight heat
studies was, on the one hand, to place the generation took place in a reactor during
design of sulfuric acid plants on a more the passageof SOZ-02 mixtures of strongly
exact basis and to provide a key to the varying concentration over a potassium-
optimization of industrial reactor per- vanadium catalyst.
formance (atmospheric pollution) and, on All investigators share the opinion that the
the other, to gather information on the reaction rate is slowed down by even small
mechanism of the reaction. Unfortunately, amounts of SOS. As this effect is already
many of the papers that have so far been noticeable far from chemical equilibrium,
published on these measurements again its cause must obviously be sought in the
and again disclosed new kinetic formula- reaction mechanism.
tions. In Table 1 t,he most important As appears from Table 1, there is quite
results are summarized. The result’s show some diversity of opinion as to t,he details of
that most authors consider the reaction the kinetics. One of the principal causes is
to be approximately first, order (3, 4, 7, 10, undoubtedly the insufficient control of the
1.2) with respect to oxygen; with respect to catalyst temperature. In the fast and
SOz the reaction order is in bet,ween highly exothermic reactions involved here,
one-half and one (1, F?,3, ‘7, 11). In the this temperature may considerably exceed
that of the gas or the reactor wall. The kind
* Present address: Technical Trniversity Twente, of the error in the reaction rates found
Enshede, Netherlands. depends on the location of the point where
I
@ 1968 by Academic Press Inc.
MARS AND MAESSEN
 TABLE! 1 (Coontinued)
.4utbor and reference Experimental data and catalyst Kinetic data Notes

RI. Goldman, L. N. Gas flows are passed at various rates ~POdJBO*

Canjar, and R. B. and at temperatures of 360-426”C v = (1 + hpo, I- bzpso, f b3pso,!*
Beckmann (8) through fluidised and fixed catalyst,
beds in a 33-mm diameter reactor.
Catalyst: 10% VZO~ + 33%
K&IO4 on silica gel.
M. S. Zakar’evski and 390-475”C; V20j containing potassium v = kpso,0.3po,0~“-‘.~ The exponents in the rate equation vary $
Man-Cheng Chang (9) and sodium. with the reaction temperature in the
way indicated.
P. B. Rzaev, V. A. Flow reactor with diameter 18 mm; 21= kpo, (PSOJPSO,?.~
Roiter, and 0. P. particle size 1.5-2.0 mm. Catalyst:
. .
Korneichuk (10) VZOS contammg barium and
aluminum.
B. Davidson and At 240-383°C in a flowing system; .4t 340°C: These results are somewhat remarkable 6
G. Todes (11) conversion <0.22; po, and pso, are v = k(psor0.87/polD.6) compared with the others; this may be 9
varied; no precooversion; catalyst: At 384°C: due to the very low reaction 2
10% KzO + 8.570 V205 on kieselguhr. v = kp~,~pso,~.~~ temperatures.
G. V. Boreskov, Differentially working circulation 8
v = h-po,
R. A. Buyanov, and reactor, 485°C; PO,and pso, have been
4. A. Ivanov (18) varied; potassium-vanadium oxide
catalyst.

w
 4 MARS AND MAESSEN

1.0 Recently Boreskov, Buyanov, and Ivanov

25%SO* (12) described their results from differential
p&4 .
3 / measurements with the formula
t Pm
us01 = kpo, (2)
pso* + &so,
This formula reflects the above-mentioned
regularities.
Eklund (7) did measurements under
differential conditions in a reactor operated
O.lC ,_ under adequately isothermal conditions. He
arrived at approximately the same formula-
tion as Boreskov, but introduced a correc-
tion for the ratk of the reverse reaction
(Table 1).
/
In a previous publication on the mecha-
nism of the reaction involved (1s) the
present authors showed that the equilibrium
103
between the vanadium ions dissolved in the
----+T pyrosulfate melt and the partial pressuresof
0.01 1.40I 1.42 1.44 146 1.48 150 152 the sulfur oxides is always established
I I 1
4OOT 375oc
425'=C so1 + 2v+ + 02- -i SOS + 2V’f (3)
FIG. 1. Temperatures at which the reaction rate The rate of the reaction between oxygen
on the K and V catalyst becomes perceptible, as a and V4+ is rate-determining for the process.
function of the SO2 and oxygen contents in the gas
mixture (total pressure 1 atm). Measured by On the basis of this mechanism the deriva-
Boreskov, Ritter, and Volkova (6). tion has been given of the kinetic formula
KpsozI~soa
the temperature is measured: near the &OS = ‘cpo2[l + Kp~o,/pso,)““]~
(4)
reactor wall, in the surrounding furnace
space, or in the center of the catalyst bed. The present publication starts with a dis-
Once the heat transfer coefficient can be cussion of the mechanism mentioned above.
estimated, the difference between the aver- Next follows a report of kinetic measure-
age bed temperature and the wall tempera- ments on two catalysts and a confrontation
ture can be calculated from the average of the results with the kinetic expression (4)
heat generation. In many casesthis differ- as well as with the values of K derived from
ence proves to be considerable (i.e., >25”C). analytical determinations.
Among the literature deserving further
consideration the publications by Eklund II. EXPERIMENTAL
(7), Calderbank (6), and Boreskov, Buyanov, The catalytic activity was measuredunder
and Ivanov (12) relate to experiments in differential conditions in a B-mm diameter
which the temperature distribution can be reactor tube (volume about 20 ml) placed in
regarded as sufficiently homogeneous(AT < a heating bath of fluidized Carborundum
lo-15°C). Besides these, the papers by particles. Diameters of catalyst particles
Boreskov should be mentioned here. This were approx. 0.5 mm. For the analytical
author derived in 1937 the frequently used procedure we refer to an earlier publication
kinetic formula (13). The maximum shift in conversion was
0.15. Minor corrections (<lo%) were
vsos= ~PO,(PSO*IPSOl)n (1) applied to the rate of SO2conversion for the
where n = 0.8. [According to Raaev and extent to which the temperature of the
co-workers (IO), it would be more correct to catalyst particle may exceed that of the
take n equal to 0.4.1 fluidized bed, and for minor deviations in
 VANADIUIk-.\LKALI OXIDE CATALYSTS

the temperature adjustment. The excess One arrives via

temperature, calculated by means of the
customary heat transfer relations, was K _ Psos [v4+l”
found to be less than 10°C in virtually all Pso,[v5+12
cases. The activation energy used for these at formula (4), provided the following
corrections was taken as 30 kcal/mole. assumptions are made: (1) aor- is inde-
Estimation of the diffusion influence of pendent of the activity of the other com-
reactants in gas-filled pores by means of the ponents; (2) the activity coefficients of V4+
Thiele method showed that no diffusional and V5+ are equal; (3) the activity of V4+
retardation of the rate can be expected. at the surface is proportional to that in the
Diffusion of reactants in the catalytically melt. Assumptions (2) and (3) are not
active liquid film of potassium pyrosulfate unlikely; assumption (1) needs some
(wherein vanadium oxides are dissolved) explanation.
does not influence the reaction rate either, One might conceive that this assumption
because separate experiments showed that is justified in view of the large excess of
the activity of catalysts containing less than potassium pyrosuIfate present. Professor
6% or 79& Vz05 is practically proportional to Wagner (G6ttingen), however, brought to
the percentage of active material. The our notice that the activity of the oxygen
compositions of the catalysts used, kindly ions may quite well be a function of pso,
supplied by three catalyst manufacturers, owing to the establishment of the following
are given in Table 2. equilibria :
The analytical method used for deter-
s2op ts sop + sot
mining the equilibrium constant K was
analogous to that described previously (1.9) : and sop * SOS + 02-
After conditioning of the catalyst in a gas of SO szo,*- ti 02- + 2so3 (6)
known composition, and rapid cooling, the
concentrations of V4f and V5f in the catalyst Combination with (3) gives
are determined as well as the concentrations
so2 + 2v5+ + s207*- * 3SOa + 2v+ (7)
of SOz and SOa in the gas phase.
III. THE MECHAISISM OF THE SULFUR Allowance being made for assumptions (2)
and (3), this yields
DIOXIDE OXIDATION ON ALKALI-
PROMOTED VANADIUM CATALYSTS K, = psos3
--- [v4+l” 1
PSO~ W5+12as:,o,f- (8)
Equation (3), which depicts the equi-
librium between the vanadium ions in the The problem arises whether us1012-or ao,-
melt and the sulfur oxides in the gas phase, may be regarded as constant. We did an
can be formulated as follows: extensive investigation to see if the con-
centration of S2072- in the melt is not
strongly dependent on 17~0~.To that end we

TXBLE 2
COMPOSITION OF THE CATALYSTS”

Vanadium oxide
calculated as VZOS
Catalyst (% by wt.) (% b: wt.1 Other compounds

1 8.2 17 2.8 5.10j0 Ca, 0.9% Na

2 7.1 9 1.65 0 .OS% Ca,
0.15% Na
3 6.8 IO 2.5 0.0770 Ca,
2.9% Na

0 The snpport of all the catalysts is silica.

 MARS AND MAESSEN

14

12

1.0

0.8

0.6

10 20 30 40 50 60 8.0 90

FIG. 2. Chemical composition of the catalyst; the sulfur/alkali ratio as

ratio as function
function of
of the partial so,pressure.
Temperature: 450%. The pso, /pso, ratio varies from l/40 to 12.

determined the ratio between the total In view of (S), one might expect
amounts of sulfur and alkali after rapid
cooling of catalyst samples that had been K,, = pso13
__ [v4+12
~-
in contact with an S03-S02-02 mixture for pso,w+12
-20 hr. Figures 2 and 3 illustrate the to be constant. However, experiments show
observations made on three industrial catal- (see Table 3) that this is not so: calculations
ysis. They show that the above-mentioned from test results obtained under non-
ratio remains very close to unity and varies equilibrium * conditions and very widely
very slightly with temperature. Hence, the varying values of psol yielded strongly
composition is virtually identical to that of different values for K”, whereas K [formula
pyrosulfate, so that the assumption of (5b)] remained fairly constant.
as,o,+ being constant would seem correct. These analytical measurements show that
Studies by Fraser and Kirkpatrick (14), as demonstrated by earlier measurements
Tops@e and Nielsen (15), and Tandy (16) the ratio between the V*+ and V6+ con-
also reveal that the composition of the melt centrations very largely adapts itself to
is identical to that of pyrosulfate. that between p801 and PSO* in the way
described in Eq. (5b) (see Table 3). Allow-
I I I I I 1 ance being made for the very large variations
in the pso,/p~o~ ratio and for the experi-
mental difficulties, the variations in K are
relatively small.

IV. THE KINETIC MEASUREMENTS ON

Two ALKALI-PROMOTED CATALYSTS

1. Influence of the Partial Pressure of

06 Oxygen on the Reaction Rate
---p tomporature CT)
The results of our experiments, in which
350 400 450 500
par was varied are illustrated in Fig. 4.
Fm. 3. Chemical composition of the catalyst; the Here the reaction rate is expressed in the
sulfur/alkali ratio as function of the temperature. function v/w, where v is the reaction rate
The 3.4% of Ca present in Catalyst 1 is assumed to and w the term
be bounded to CaSOI; the sulfur content found has
been corrected accordingly; pso, varies from 0.002 * That is, there is no equilibrium between SOP, 02,
to 0.10 atm, psO,/p~O, from l/40 to 12. and SO,.
 VANADIUM-ALKALI OXIDE CATALYSTS

TABLE 3
RESULTS OF MEASUREMENTS ON CATALYST 3 REGAI~DING THE EQUILIBRWM BETWEEN
THE SULFUR OXIDES AND VANADIUM IONS”

[V4+] PSO(
-
PBOJ PSO,
(10-2 atIn) (lo-‘&II) [va+] PSO,

3.97 0.26 3.52 0.0655 0.81 ri.6 X lO-B

1.57 2.72 1.02 1.73 1.82 1.35 x 10-X
0.68 3.49 0.605 5.14 1.88 2.3 X lO-B
3.59 0.43 2.78 0.121 0.93 1.7 x lo-’
0.67 3.57 0.495 5.33 1.30 1.65 X 1OP
0.29 3.95 0.313 13.6 1.33 2.09 x 10-a

a 802, SOS, and 02 are not in chemical equilibrium with each other. Temperature 450°C.

K(1 - 5)/L+ will therefore be zero order with respect to

{l + [K(l - 5)/?5]“2)‘? SOS. Since the SOSformed may, of course,
reduce the reaction rate, a correction is
This function gives the reaction rate after
made for this effect by means of the term
correction for the retarding effect of SOS
w given above. The catalytic activity is
[compare formula (4)].
expressed in terms of the reaction rate
The figure also shows the reaction
constant k, i.e., the constant that is first
temperature, the average degree of con-
order with respect to oxygen under condi-
version 5, and the values of n from the
tions where no SO3is present,
relation v = kpo,” relating to each line. It
was found that below 20% 02 n approxi-
mately equals unity; over 20% 02 n is
much lower, i.e., 0.3-0.4. As shown in
Table 1, n = 1 is in good agreement with
the results obtained by most investigators. Table 4 shows that the influence of PSO,on
The cause of the deviation at higher oxygen vso, is small (provided PO, has not been
concentrations is not clear, however. It is changed) and that that on k is not notice-
worthwhile stating here that Eklund also able. This is in agreement with the experi-
found n to deviate from unity above 20% ments by Boreskov, Ritter, and Volkova (5)
02, albeit that he reports a value of ~2. referred to above (seeFig. 1).

2. Rate of Reaction as a Function of psol 3. The Injluence of SO3on the Reaction Rate
without SOSin the Initial Gas
In order to perform a check on the
At psoa<( PSO*,Eq. (4) will change into validity of kinetic formula (4) the results of
vso, = kpo,. In SOS-free gas the reaction the differential measurements on Catalysts

TABLE 4
INFLUENCE OF pao1 ON THE INITIAL REACTION RATE

Tempe$ure PO1 (PSO,h “SOJ lc

Catalyst O (atm) (atm) L11 (g mole/set g) w (g mole/set g at.)

3 395” 0.187 0.084 0.090 2.08 x 10-e 0.51 2.19 x 10-S

395” 0.197 0.040 0.154 1.68 X 1O-6 0.42 2.02 x 10-S
3 454” 0.187 0.087 0.062 1.49 x 10-b 0.745 1.07 x lo-’
455” 0.183 0.087 0.169 1.17 x 10-S 0.61 1.05 x 10-d
456” 0 046 0.042 0.125 3.15 x 10-a 0.65 1.05 x 104
2 446” 0.187 0.082 0.090 2.42 X lO-B 0.82 1.58 x 10”
446” 0.195 0.041 0.187 2.51 X 1O-6 0.75 1.71 x 10-s
446” 0.200 0.022 0.390 2.78 X lo+ 0.65 2.14 X lO-b
8 MARS AND MAESSEN

16' lo-' 1.0

FM. 4. Reaction rate as function of the partial oxygen pressure. Catalyst; paoz ahead of the reactor,
0.06-0.10 atm.

900- d YpO2
(,d"*. s.,“2, 2’12.gmo,“~2
)
I
aoo-
f

700 _

600
1
x 415-Z
500
q 450-c CO1

0 480-c
400
f--l A 48O‘T '1 COll

300

M4
100 /-
.
,..
I /

-2.5 -20 -1'5 -1.b -015 0'5 lb 1; 2'0 215

FIG. 5. Resulta of kinetic measurements on Catalysts 1 and 2 plotted according to Eq. (9). Differential
measurements were performed at various levels of conversion formed by variation of the preconversion;
pal * 0.19 atm.
 VANADIUM-ALKALI OXIDE CATALYSTS 9

1 and 2 have been plotted (Fig. 5) in a distinct deviation from the proposed
conformity with the linearized formula kinetic formula.
The results of the measurements per-
formed on Catalyst 3 have been plotted in
(g = (3’” + (-g2 (Pg’ (9)
the same way (see Fig. 6). The equilibrium
position (5b) for this catalyst with its high
The partial pressures before and after the sodium content not being known as a
catalyst have been averaged.* Figure 5 function of temperature, it was determined
by analysis of the V4+/V5+ ratio in the
shows that the measurements at higher
catalyst and of the S02/S0, ratio in the gas
temperatures (450°C and upwards) are
phase. The straight line drawn through the
indeed described by his formula. The
experimental points lies under that of
values for k and K can be calculated from
Catalyst 1 (Fig. 7).
this figure.
From the results obtained from the
The question arises whether the K values
kinetic measurements on Catalyst 3 (see
derived from such kinetic data are in con-
Fig. 6)* it appears that the kinetic views
formity with those calculated, by means of
suggested above are also applicable to this
(5b), from chemical analyses of [V4+]/[V6+]
in the catalyst and p&p~o~
in the gas catalyst
(above
under
420°C).
high-temperature
Further, the
conditions
values found
phase [see ref. (IS)]; the latter will further
prove to lie on a straight line, which inter-
be denoted as Kansl. It appears quite
feasible to draw a line through the experi- determined by the analytically determined 0]
sects the axis in a point [- (K8,a1)1'2,
mental points, starting from [ - (Kanal)1'2,01.
value of K. Hence, the formula (4) indeed
This shows that the agreement is satisfactory describes the rate of the SOS formation at
between the mechanism and the derived
elevated temperatures. A deviating trend
kinetics of this process. At temperatures
is again noted upon a decrease in tempera-
below 45O”C, however, there appears to be
ture. Now that the basic pattern of the
* The reverse reaction has been neglected here. *The reverse reaction has been neglected here.
At high degrees of conversion the value of the At high degrees of conversion the value of the
ordinate may be at most 7’% too high. ordinate may be at most 7% too high.

r
900

800
I
700

600 1

-2'5 -2'0 -1 5 -10 -d5 0 05 10 15 20 25

FIG. 6. Results of kinetic measurements on Catalyst 3 according to Eq. (9). Differential measurements
were performed at various levels of conversion formed by variation of the preconversion.
10 MARS AND MAESSEN

a. The deviation from normal kinetic

behavior at lower temperatures. (See Figs.
5 and 6.) From the sequence of the experi-
ments it can’be concluded that this effect is
not caused by insufficient establishing of
stationary conditions. This phenomenon
may be formulated as follows: At low
temperatures the rate-retarding effect of
SO3 is stronger than would follow from
formula (2). This effect is stronger in
Catalyst 3 than in Catalyst 1. It cannot
be due to the formation of polysulfate ions
(S3O10)2- (see the ratio S/(Na + K) in
Fig. 4), but, perhaps, it is to be sought in
rio li0 140 150 the formation of sulfovanadates. Such
-----, ‘q!z! sulfo groups might enclose and stabilize the
550 5bo 450 460 V4+. Another possibility is that at lower
TIrnP (“C) - temperatures the desorption of SOS in-
FIG. 7. Values of the
equilibrium constant fluences the reaction rate.
K = [Vr+]zp~~,/[V~+]2p~~, on the two types of Our experimental information does not
catalyst as a function of the temperature. enable us to decide which of these possible
causes has the strongest effect.
kinetics has been established, some further b. k as function of the temperature.
points remain to be discussed: (a) the Does the Arrhenius curve contain a break?
deviating trend at lower temperatures; If the influence of SOI on the forward
(b) k as a function of the temperature. Is reaction can be quantitatively described in
there a break in the Arrhenius curve? the way outlined above, the value of the

. Catalyst 3,from flgure6

110 1.20 1.30 1.40 1.50 1.60

FIG. 8. Calculated rate constant as function of the temperature (measurements on small catalyst grains:
0.547 mm).
 VANADIUM-ALKALI OXIDE CATALYSTS

0.50

Q25--p---p

-1.6 -16 -14 -1 2 -1 0 -00 -0

FIG. 9. Results of Eklund’s measurements (?‘), represented according to Eq. (9).

reaction rate constant Ic can be calculated

wherein n = 1. Then, the lines through the
as a function of the temperature. If the
experimental points in Figs. 5 and 6 are not
theory is correct, the Arrhenius plot will not
drawn through (-K 112,0) but through the
develop a break as is generally observed origin.
upon application of other reaction rateIn Fig. 9 Eklund’s experimental results
constants (1S,17). Included in Fig. 8 are the
have been plotted in the same way as in
k values found by extrapolating the experi-
Figs. 5 and 6, and the lines have been drawn
mental results illustrated in Figs. 5 and 6
through the [ - (Kansl)1/2, 0] values corre-
towards a = 0, and also those calculated sponding to the various temperatures. The
from the results of some individual experi-
points relating to temperatures of 455’C,
ments at other temperatures by means of
and above, nearly all give a very close fit
Eq. (4). This figure shows that application
and Eklund quite understandably states
of the derived kinetics does not produce a
that his reaction rate constant (see Table 1)
break in the Arrhenius curve. varies systematically with the degree of SO2
conversion. Consequently, Eq. (4) is, in
4. Comparison with the Literature fact, more appropriate than his own. It is
The kinetics examined in this study remarkable that Eklund’s results obtained
represent the slowdown caused by SO8 at lower temperatures (437°C) also deviate
somewhat differently than other formulas. from those found at higher temperatures,
At very low K values our formula (4) in the sameway as was noted by us.
changes into the Boreskov formula (l), The new formula of Boreskov et al. (12),
12 MARS AND MAESSEN

FIO. 10. Results of experiments of Boreskov et al. (I,@, represented according to Eq. (9)

i.e., (2), resembles very much our formula Sulphur Dioxide with a Commercial Vana-
(4). In Fig. 10 the experimental results dium Catalyst.” Thesis, Stockholm, 1956.
8. GOLDMAN, M., CANJAR, L. N., AND BECKMANN,
which form the basis for (2) are plotted ac-
R. B., J. Appl. Chem. (London) 7, 275
cording to the linearized form of formula (4). (1957).
This figure shows that also these results are 9. ZAKAREVSKI, M. S., AND CHANG, MAN-CHENG,
in accordance with our kinetic expression; Chem. Abstr. 53, 13755 (1959) ; Chang, Man-
moreover, the value of the constant K is Cheng c.s., Chem. Abstr. 57, 16127 (1962).
identical with that derived from analytical 10. RZAEV, P. B., ROITER, V. A., AND KORNEICHUK,
determination. G. P., Ulcr. Khim. Zh. 26, 161 (1960).
11. DAVIDSON, B., AND TODES, G., AJ.Ch.E. J. IO,
REFERENCES 568 ( 1964).
1. BORESKOV, G. K., AND PLIGUNOV, V. P., J. 1% BORESKOV, G. K., BUYANOV, R. A., AND IVANOV,
Appl. Chem. USSR (Zh. P&lad. Khim.) A. A., Kinetika i Kataliz 8, 153 (1967).
6, 785 (1933). 13. MARS, P., AND MAESSEN, J. G. H., Proc. Intern.
9. BOBESKOV, G. K., Chem. Abstr. 29, 6371 (1935). Congr. Catalysis, Srd, Amsterdam, 1964 1,
3. BORESKOV, G. K., AND S~KOLOVA, F. J., J. Chem. 266.
Ind. USSR 14, 1241 (1937). 14. FRASER, J. H., AND KIRKWIWCK, W. J., J. Am.
4. BRODOVICH, K. J., J. Appl. Chem. USSR 14, Chem. Sot. 62, 1659 (1940).
894 (1941); Chem. Abstr. 39, 3997 (1945). 16. TOPS@E, H. F. A., AND NIELSEN, A., Trans.
6. BORESKOV, G. K., RITTEX, L. G., AND VOLKOVA, Danish Acad. of Technical Sciences, No. 1,
E. J., Chem. Abstr. 43, 8251 (1949); J. p. 3 (1948).
Appl. Chem. USSR 22, 250 (1949). 16. TANDY, G. H., J. Appl. Chem. (London) 6,
6. CALDERBANK, P. H., J. AppZ. Chem. (London) 68 (1956).
2, 482 (1952). 17. GERBURT-GEIBOWITSCH, E. V., AND BORESKOV,
7. EKLUND, R.-B., “The Rate of Oxidation of G. K., .I. Phys. Chem. USSR 30, 1801 (1956).