Applied Catalysis, 26 (1986) 65-79 65

Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

ROLES OF POTASSIUM IN POTASSIUM-PROMOTED IRON OXIDE CATALYST FOR DEHYDROGENATION

OF ETHYLBENZENE

Takenori HIRANO
Kasukabe Laboratory, Nissan Girdler Catalyst Co. Ltd., Ushijima 573, Kasukabe-shi,
Saitama-ken, Japan.

(Received 9 October 1985, accepted 28 May 1980)

ABSTRACT
Experimental kinetic data on the dehydrogenation of ethyl benzene over potassium-
promoted iron oxide catalysts containing various amounts of potassium were obtained
in the temperature range 550-620 C. Unimolecular reactions for the formations of
styrene, benzene and toluene on the different sites, where potassium reduced the
activation energies, were considered on the basis of Langmuir-Hinshelwood type
mechanism. Using a method of in situ catalyst deactivation with CO potassium
2,
ferrite formed in the surface of the catalyst under reaction was suspected to take
part in the formation of the active sites.

INTRODUCTION
Styrene monomer has been commercially produced by the dehydrogenation of ethyl-
benzene over potassium-promoted iron oxide catalysts in the presence of steam. The
most widely used industrial catalysts comprise iron oxide, potassium carbonate and
promoters: such as Cr203, Ce0 2, Mo0 3, V205 and so on [1]. Potassium-promoted iron
oxide catalyst has been uniquely superior to other known catalysts in the presence
of steam; potassium increases the activity of iron oxide over an order of magnitude
[2]. The promoting effect of potassium is complicated and not completely understood,
though various concepts have been proposed. lee [3] suspected that various alkali
and alkaline earth promoters increased catalytic activity by encouraging electron
transfer at the solid-gas interface. Vijh [4] theoretically argued that the catalyst
had active sites located on or around the points on the catalyst where the promoter
atoms were embedded. He showed a good correlation between the ionicity of the pro-
moter oxide and its promoting effect. Murakami [5] suspected that potassium reduced
the oxidation state of iron oxide to increase the selectivity to styrene especially
by decreasing the benzene formation. Considerable amounts of industrial research
have been carried out to understand and improve the catalyst because the styrene
production is industrially of importance; however basic study on this catalyst is
not suffi ci ent.
On the other hand, potassium content in the dehydrogenation catalyst has gradu-
ally been increased for the purpose of being used under a low steam to ethyl benzene

01669834/86/$03.50 1986 Elsevier Science Publishers B.V.

66

ratio from a stand point of saving energy; an industrial catalyst of Shell 105
contained KZO below 5 wt% when it was first used [6J, however this catalyst recently
contains 10 wt% of KZO and Shell Z05 is said to contain more than 10 wt% [7J. It is
considered that potassium enhances water-gas reaction to remove carbonaceous deposit
[8,9J which may more easily be produced by lowering the steam to ethyl benzene ratio.
The purpose of this study is to investigate the roles of potassium and the effects
of the potassium content on the properties of potassium-promoted iron oxide catalyst.

EXPERIMENTAL
Catalysts
The catalysts examined were prepared as follows: iron oxide (~-FeZ03' 23 m2g-1)
powder and a suitable amount of potassium carbonate (K 2C03) powder were mixed and
then extruded to pellets of 3.Z mm dia. and 4-6 mm length. The pellets were dried
at 120C overnight followed by a calcination at 600C for 4 h; which were crushed
into 16/Z0 mesh granules before being used for the tests to eliminate the influ-
ence of diffusion. Four catalysts of different compositions were prepared: (1) 80
wt% FeZ03-20 wt% K2C03 (80Fe-20K), (2) 75 wt% Fe203-25 wt% K2C03 (75Fe-25K), (3)
72 wt% Fe203-28 wt% K2C03 (72Fe-28K), (4) 69 wt% Fe203-31 wt% K2C03 (69Fe-31K).

Apparatus and reaction

The activity tests were carried out in the temperature range 550-620 C in the
presence of steam: steam to ethyl benzene molar ratio of 11.8 was maintained under
atmospheric pressure. Occasionally, purified helium or carbon dioxide was added to
the feed as a diluent or an inhibitor of the catalyst. A differential type reactor
consisted of stainless-steel (20 mm i.d.) containing 3 g catalyst was used with
total SV of 39100 ml g-1 h-1; operation in a low conversion region was practiced.
The products were condensed in an annular condenser; in addition to ethyl benzene
(EB) and water, the condensate contained styrene (ST) and small amounts of benzene
(Bz) and toluene (Tol), which were analyzed by gas-chromatography. Main products in
the vent gas were hydrogen and minor amounts of carbon dioxide, carbon monoxide,
methane and ethylene were also detected. The selectivity to styrene was defined as
follows:

% selectivity = ST/(ST + Bz + Tol), x 102

because the amount of output aromatics was almost equal to the amount of input
ethyl benzene [10J.

Analysis of catalyst
The X-ray diffraction measurements were carried out on Rigaku X-ray diffracto-
meter with Co Ka radiation. The deposited amounts of carbon on the used catalysts
were determined by trapping the CO 2 released in an oxidation treatment at 450C for
 67

TABLE 1
The formations of styrene, benzene and toluene in the ethyl benzene dehydrogenation a
over Fe203-K2C03 catalysts at 580C

Catalyst min , x10 5

Rate/mol g-1-1 Ea/kcal mol-1 Selec.
ST Bz Tal ST Bz Tal to ST/%
100Fe 3.2 0.12 0.04 34.2 39.7 57.2 95.4
80Fe-20K 33.8 0.69 0.44 30.2 35.4 53.3 96.6
75Fe-25K 29.6 0.58 0.44 31.0 36.2 50.1 96.9
72Fe-28K 26.6 0.50 0.36 30.0 36.2 50.9 96.9
69Fe-31K 24.7 0.46 0.30 30.2 35.8 50.1 97.0

aSteam/ethylbenzene (molar ratio) = 11.8, total SV 39100 ml g-1 h-1

2 h after evacuation at 250C for 30 min. The surface areas of the catalysts were
measured by the BET method (N 2 adsorption).

RESULTS AND DISCUSSION

Ethylbenzene dehydrogenation over Fe2Q~2C03 catalysts
When a mixture of ethyl benzene and steam was passed through the potassium-promoted
iron oxide catalyst, the activity increased gradually at the initial stage owing to
the reduction of iron oxide as has already been reported [11,12J, and then reached
the steady state within 20 h. All experimental data were obtained at the steady
state of the performance test in the temperature range 550-620 C.
As shown in Table 1, both the activity and selectivity of the potassium-promoted
catalysts were remarkably superior to those of the pure iron oxide. It is noticeable
that the activation energies for the styrene, benzene and toluene formation on the
promoted catalysts were lower than those on the pure iron oxide. From these facts,
it is presumed that the addition of potassium might increase the number of active
sites participating in the rate-determining steps for each product to reduce the
activation energies. There is little difference in the activation energies for each
reaction over these promoted catalysts containing different amounts of potassium;
this fact indicates that each reaction proceeded by the same mechanism over the
various catalysts.
The specific surface area (SA) of fresh (before the reaction) and used (after
the reaction) catalysts are listed in Table 2. The surface area of the fresh catalyst
is correlated with the content of iron oxide, i.e., the surface area decreases with
the decrease in Fe content, although the surface of iron oxide may be covered
203
with potassium compounds to reduce the surface area. The surface area of the used
catalyst is larger than that of fresh catalyst, in contract to the case of the
pure iron oxide. By analysis of CO contents in the fresh and used catalysts using
2
68

TABLE 2
Surface area of fresh and used catalysts and deposited amount of carbon on the
used catalysts

SA/m 2g- 1 Deposited % C/r EBa

Catalyst
Fresh Used carbon/wt% Iwt% min g mol -1 ,x10 -2
100Fe 15.0 12.0 0.359 107
80Fe-20K 7.2 7.5 0.050 1.4
75Fe-25K 6.7 7.3 0.031 1.0
72F e-28K 6.3 6.9 0.029 1.1
69Fe-31K 4.7 5.2 0.024 0.9

ar EB: rate of ethyl benzene consumption at 580C.

alkali-meter, it was found that a part of potassium carbonate was decomposed into
other compounds; the decomposed extents in the fresh and used catalysts were in the
range of 27-36 % and 12-23 %, respectively.
Doposited amounts of carbon on the catalysts used for the dehydrogenation tests
are given in Table 2. The carbon deposition seems to be more depressed on the
catalyst containing higher amount of potassium. The carbonaceous deposit might be
removed by water-gas reaction catalyzed by potassium compound according to the
proposed mechanism [13J; potassium carbonate might behave as the active species [14J.
The carbonaceous deposit was sometimes considere-dto deactivate the catalyst [15J,
however such deactivating effect of carbon was not observed in this study; there was
a tendency that higher amount of carbon was deposited on more active catalyst, then
the amounts of carbon per the rate of ethyl benzene consumption, represented as % CI
rEB' were obtained as listed in the table. The % C/r EB value of the potassium-
promoted iron oxide catalyst is markedly less than that of the pure iron oxide,
however little difference in the value is observed among the potassium-promoted
catalysts containing various amounts of potassium. From these results, it is sus-
pected that the carbonaceous deposit on the potassium-promoted iron oxide might be
produced via reaction intermediates and might not exist on the active sites to
inhibit the activities.

Reaction kinetics
The dependence of the rates of styrene, benzene and toluene formation upon the
partial pressure of ethyl benzene over the catalysts containing various amounts of
potassium was investigated at 580C. As shown in Figure 1, the rate of styrene
formation, i.e., the main reaction, was independent of the partial pressure. On the
other hand, the rates of benzene and toluene formation, i.e., the side reactions,
were dependent of the partial pressure as shown in Figures 2a and 2b; where the
 69

'1!x 40
'IC 30
rl
...-E
101 20
.--l
o
E
'f-l0
......
V) 1~0--~----1._--'-_---1._-""""'''''''''''
5

FIGURE 1 Dependence of the rate of styrene formation at 580C upon the partial
pressure of ethyl benzene; over 80Fe-20K (0), 75Fe-25K (rn), 72Fe-28K (.), and
69Fe-31K (E1).

l.!")
l.!") 0.6 / 0
...- 0.6
~ x
x e-r- b
e-r- I
c
1
rl
C 0.4 .-E
rl 0.4
...-E I
1 01
01 .-l
0
.-l
0.2 E 0.2
~
<,
.--l
N 0
I-
co
...... ......

0 0.02 0.04 0 0.02 0.04

PEB/atm PEB/atm

FIGURE 2 Dependences of the rates of benzene (a) and toluene (b) formation upon
the partial pressure of ethyl benzene at 580C; over 80Fe-20K (0,0), 75Fe-25K (~,),
72Fe-28K (e,e), and 69Fe-31K (e,G).

dependences of both reactions seem to be different from each other.

Different amounts of styrene were added to the reactant feed to investigate its
influence on the reaction. The rate of styrene formation was largely inhibited by
the addition of styrene as shown in Figure 3. However, no inhibition effects on the
rates ofbenzene and toluene formation were observed; it can also be understood
from this figure that the formations of both by-products might mainly be produced
from ethyl benzene, not from styrene, without being inhibited by the added styrene.
When the partial pressure of steam in the feed was changed while maintaining
that of ethyl benzene constant, all formation rates of styrene, benzene and toluene
were not affected as shown in Figure 4; the steam added as a diluent is not
70

1.5 30
,....., 5T
0
:--.
f-
1.0 ~ 20 f-
:--. Bz IJ)
:--.
0
N 0.5 10
co
:--. Tal 0
0 0.5 1.0 1.5
Added styrene/atm,x10 2

FIGURE 3 Effects of added styrene upon the rates (mol g min

-1 -1 x 105) of styrene,
benzene and toluene formation at 580C; over 72Fe-28K.

0.5
,....., 5T - 30
0
f- 0.4 f--G 0 0 0-
:--.
f-
:--.
0 Bz :--.
IJ)

N 0.3 f- 0
0----0 0- 20
co
:--. ~
0.2 Tal
I

0.3 0.4 0.5

PH o/atm
2
FIGURE 4 The formation rates (mol g-l mi n-l, x 105) of styrene, benzene and toluene
against the partial pressure of steam at 580C; over 72Fe-28K; PEB: 0.028 atm.

considered to participate in the rate-determining steps of these reactions.

The reaction kinetics of the dehydrogenation have been studied by a few authors
[16-18J in the past, using tubular reactors. Carra and Forni [16J studied the rate
of styrene production on Shell 105 in the absence of internal diffusion and far
from equilibrium conditions, and finally proposed the following rate expression of
the Langmuir-Hinshelwood type employing a kinetic model in which it was assumed
that the reaction is unimolecular on the catalytic surface:

(1)

where r ST' kST' KEB and KST represent the rate of styrene formation, the rate con-
stant, the equiribrium constant of the ethyl benzene adsorption and that of the
styrene adsorption, respectively. The adsorption of benzene and toluene were neg-
lected because they were present only in small amounts. The adsorption of hydrogen
was also neglected because of the high operating temperature.
 71

t"l
I
8
0
..-
x
.-
I
...-;
0
6
E
c
.......
E
CJ)
4
<,
f-
(/)
~
<,
..-
2
0 O. 1 0.2
PST/PEB

FIGURE 5 The inhibition effect of added styrene on the rate of styrene formation
at 580C; over 80Fe-20K (0), 75Fe-25K (rn), 72Fe-28K (.), and 69Fe-31K (8).

In the present study, it was found that the styrene formation rate was independent
of both the partial pressures of steam and ethyl benzene and that the amount of
styrene in the feeding mixture strongly affected the reaction rate; these results
indicate that the proposed rate equation is applicable to the examined catalysts
containing higher amounts of potassium. In order to confirm the rate equation, it is
rearranged to give

(2)

Using the data on the inhibiting effect of styrene on the styrene formation rate,
the reciprocal of the rates of the styrene formation is plotted against PST/PEB in
Figure 5, where straight lines are found as expected in the equation (2). This
result indicates that styrene is produced by the proposed surface unimolecular
reaction of ethyl benzene on the sites for the styrene formaion which are occupied
by ethyl benzene and styrene.
The values of the rate constants of kST at 580C on the various catalysts were
obtained from the values of l/r ST at PST/PEB = 0 in Figure 5, the ordinates being
equivalent to the reciprocal of kST' The obtained kST and the ratio of KST/K EB,
which means the relative adsorption strength of styrene to ethyl benzene on the
dehydrogenation sites, are listed in Table 3. As expectea from the rates obtained
under low conversion, the kST increases as the potassium content in the catalyst
decreases. It is of interest that the KST/K ratio is dependent of the potassium
EB
content, i.e., the ratio decreases as the content increases. In other words,
potassium may prevent the adsorption of styrene onto the dehydrogenation sites
compared to that of ethyl benzene.
 -l
TABLE 3 ~

The rate constants and the equilibrium constants of ethyl benzene adsorption at 580C in the equations for the styrene,
benzene and toluene formations a

5
kBz x 10 (KEB)Bz kTol x 105 (KEB)Tol kST x 105 KST/K EB
Catalyst
I mol 9-1-1
min latm-1 Imol 9-1 min -1 latm-1 Imol 9-1 min -1

80Fe-20K 1.20 18.2 0.57 54.9 34.0 8.37

75Fe-25K 1.05 16.0 0.53 31.9 28.6 3.03
72Fe-28K 0.97 14.0 0.49 32.2 26.7 1.81
69Fe-31K 0.87 14.1 0.43 29.7 25.8 1.61

aEquation (1) for styrene formation, equation (5) for benzene formation and equation (6) for toluene formation.
 73

o 20 40 60 80 100 o 20 40 60 80 100
-1
1/PEB/ atm-1 1/P EB
/atm

FIGURE 6 1/r (g min mol- 1, x10- 5) vs. 1/PEB plots for benzene formation (a) and
toluene formation (b) at 580 C; over 80Fe-20K (0,0), 75Fe-25K (~,~), 72Fe-28K (e,e),
and 69Fe-31K (e,e).

Carra and Forni also considered the side reactions of benzene and toluene forma-
tion, and finally lumped together both reactions into the rate equations as follows:

(3 )

and

(4)

where r Sz' r Tol' kSz and kTol represent the rate of benzene formation, that of
toluene formation, the rate constant of benzene formation and that of toluene forma-
tion, respectively. It must be noted that both equations are the same function of the
partial pressures of ethyl benzene and styrene; no consideration on the difference
in the sites for the benzene and toluene formation were taken into.
The results obtained in this study, however, indicate that both the rates of
benzene and toluene formation are independent of the partial pressure of styrene
and the dependences upon the partial pressure of ethyl benzene seem to be different
for each reaction as seen in Figures 2a and 2b; these facts indicating that the
adsorption of styrene can be neglected and different adsorption constants of ethyl-
benzene must be used for each reaction. Consequently, assuming that the reactions
are the unimolecular as the styrene formation, equations (3) and (4) will be modified
to be expressed as follows:

(5)
74

and

kTol(KEB)TolPEB (6)
1 + (KEB)To'pEB'

where (KEB)Bz and (KEB)Tol are the equilibrium constant of ethyl benzene adsorption
onto the benzene formation sites and that onto the toluene formation sites, respec-
tively.
Modifying equations (5) and (6) as

(7)

and

(8)

1/r and 1/r were plotted against 1/PEB in Figure 6, where linear correlations
Bz Tol
were found.
The values of the rate constants, kBz and kTol' were obtained from the values of
1/r at 1/P = 0 in Figure 5 according to the equations (7) and (8), and the values
EB
of the equilibrium constants of ethyl benzene adsorption onto the benzene and toluene
formation sites, (KEB)Bz and (KEB)To1' were obtained from the slopes of the straight
lines; the obtained values are listed in Table 3. Both rate constants decrease with
the increase in the potassium content. The ~quilibrium constants of the ethyl benzene
adsorption do not remarkably change as the potassium content increases. The satura-
tions with ethyl benzene of the sites for the benzene and toluene formation are
0.52-0.59 and 0.70-0.81, respectively. Consequently, it is confirmed that the forma-
tions of benzene and toluene are expressed by the equations (5) and (6), respec-
tively.
Being based on the Langmuir-Hinshelwood type reaction mechanism, consequently,
different reaction equations being different functions of the partial pressure of
ethyl benzene and styrene are considered to be applied for the formations of styrene,
benzene and toluene over the potassium-promoted iron oxide catalyst. From the
obtained results of the equilibrium adsorption constants of ethyl benzene for each
product being different from each other, it is suspected that the sites for each
reaction concerned with the rate-determining steps may be different from each other.

In situ catalyst deactivation with CO 2

When CO 2 was admitted to the reactant feed gas at the steady state of the reac-
tion, all of the formation rates of styrene, benzene and toluene were largely
 75

2.0 0.54%C02 50
5T CO2 removal
t 40
.--i 1.5 : \ 0.95%(0
f-
0 2
l-< 30
l-<
0 LO Bz
....
t f-
Vl
l-<
N
CD
20
l-<
0.5 Tol 10
0 0
24 26 28 30 32 34
Tlne/h

-1 . -1
FIGURE 7 The influence of CO 2 addition to the feed upon the rates (mol g mln
x 105) of styrene, benzene and toluene formation over 72Fe-28K at 600C.

reduced and the extents of these reductions increased as the partial pressure of
CO 2 added in the feed increased; a typical result of the deactivation in situ with
CO 2 at 600 C is illustrated in Figure 7. In spite of the large reduction, the
activity gradually recovered after removing the CO 2 from the feed gas. This deactiva-
tion with CO 2 was observed on every catalyst examined. Though the reduction was also
observed on a pure iron oxide, the extent was negligibly small. Consequently, CO 2
inhibited the formations of styrene, benzene and toluene in the dehydrogenation
over the potassium-promoted iron oxide catalysts.
From the chemically analyzed amounts of CO 2 contained in the various catalysts,
it was presumed that most of potassium existed as potassium carbonate in the used
catalysts, however a part of potassium might exist as other compounds: such as
potassium oxide or potassium ferrite, even though potassium had initially been added
as potassium carbonate. It has been confirmed that pure potassium carbonate does not
decompose at temperatures up to 750C even in the presence of steam, measured on
TGA. However, the decomposition occurred in the potassium carbonate-iron oxide
system, i.e., potassium-promoted iron oxide. Ohtsubo et al. [19] studied the reac-
tion between potassium carbonate and iron oxide, and found that potassium oxide at
first produced on the surface of iron oxide and then potassium ferrite was formed;
the formation of potassium ferrite finished when iron oxide had completely been
converted into ferrite in the presence of excess amount of potassium carbonate. On
the other hand, it is known that potassium can easily migrate on the surface of
iron oxide under dehydrogenation conditions [2,20]. As the potassium carbonate and
potassium ferrite are stable compounds, it can be presumed that the potassium may
migrate as potassium oxide.
In this study, it was found that CO 2 inhibited the reactions of styrene, benzene
and toluene formation over potassium-promoted iron oxide. Since a part of potassium
76

might exist as potassium oxide or potassium ferrite, it is suspected that CO 2 might

react with these potassium compounds to form potassium carbonate. X-ray analysis
showed that potassium ferrite rapidly decomposed to potassium carbonate and iron
oxide on exposure to CO 2 at any temperature from 550 to 650C. It can consequently
be assumed that the potassium ferrite, which is formed after the decomposition of
potassium carbonate in the presence of steam, may form an active phase for the
react ions.
The mechanism for the formation of potassium ferrite on the surface at a steady
state of the dehydrogenation may be expressed as follows:

(9 )

(10)

(11 )

(12)

and

(13)

where [ ], [K 20] and NK represent the vacant, occupied and total amount of sites
for the decomposition of potassium carbonate to form potassium oxide on the surface
of iron oxide, respectively, and (Fe203), (K 2Fe204) and N represent the iron oxide,
F
potassium ferrite and total amount of them which can form the active sites of
potassium ferrite in the surface of iron oxide. K2C0 in equations (9) and (11)
3
represents the particles of potassium ferrite being in contact with the surface of
iron oxide. As the amount of K2C03 was excess and the partial pressure of CO was
2
low, the amount of [K 20] may be almost equal to NK. Using a steady state method,
following equation is obtained:

(14)

where C1 is a constant of NK/N F ratio, because potassium carbonate can decompose

only on the surface of iron oxide, i.e., the sites for the decomposition are
presumed to be proportional to the amount of surface iron oxide. (K in this
2Fe204)
equation represents both the potassium ferrite and its amount, i.e., the number of
 77

,.,..., 10 10 10
I
l..C\ '"I"0
.....
0
I
~
.....
x
x x
.....I .....I .....I
.--;
.--; .--; 0
0
E ~ E
c
c
....... 5 5 c;
....... ....... 5
E
E E
Ol
<,
c c <,
<, .--;
t- N 0
~
I./) co t-
~ ~

.....
'- <,
<- .....
<,

0
0 0.5 1.0 1.5 0 0.5 1.0 1.5
Added Peo /atm,x102 Added Peo /atm,x102
2 2

FIGURE 8 l/r vs . PCO added in feed; over 80Fe-20K (0,0,0), 75Fe-25K (m,CD,([l), 72Fe-
28K (_,e,e) and 69Fe~31K (8,9,9).

active sites. As the rate is proportional to the amount of active sites, i.e., r =
a ( K Fe 0 ) , next equation can be obtained:
1 2 2 4

(15)

where a = a 1C1k/ k4 . Using Fe203 content (wt fraction) in a catalysttnstead of NF:

2
NF = C2(Fe203 in cat.), equation (15) is rearranged to give

(16)

where (Fe203 in cat.), S1 and S2 represent the Fe203 content in a catalyst, a1C2
and a2C~, respectively. The reciprocal of the rate of styrene formation is plotted
against PCO in feed gas in Figure 8a, where linear correlations are found as
expected fr6m equation (16). From this figure the dependence order of the slope of
the straight lines on the Fe203 content was determined to be -2.1; this order is
approximately -2 which is expected from the equation (16). In the cases of benzene
and toluene formation, the rates and Peo had the same relations as the styrene
formation, as shown in Figures 8a and 8bf the dependence orders of the benzene and
78

toluene formation were -1.9 and -2.1, respectively, and these orders are also near
to the expected value of -2. These results substantiate the equation (14), i.e.,
the proposed mechanism for the formation of active sites. Furthermore, from the
slope of the straight lines 8 values for the styrene, benzene and toluene formation
2
were determined to be 5.7 x 10-6 ,2.0 x 10-7 and 3.2 x 10-8 mol atm g-1 mln
. -1
,respec-
tively; since the 82 value is expressed as 02C~ and the C2 means the population of
active iron oxide capable to form the potassium ferrite per unit content of iron
oxide, i.e., NF = C in cat.), it can be presumed that the 82 is a function
2(Fe203
of the population of the active iron oxide.
Consequently, it is considered that an interaction between potassium oxide and
iron oxide, which forms the potassium ferrite, participates in the site formaions
for the reactions of styrene, benzene and toluene formation.

CONCLUSIONS
The results indicate that the potassium takes part in the formation of the active
sites to reduce the activation energies for the reactions of styrene, benzene and
toluene formation. From the ~netic analysis based on the Langmuir-Hinshelwood type
mechanism, different equilibrium constants of ethyl benzene adsorption were obtained
for each reaction of styrene, benzene and toluene formation; this result being
suggestive of the differences in the active sites for each reaction, even though
futher investigation might be necessary to conclude the existence of the different
sites. Since the interaction between potassium oxide and iron oxide is considered
to form the active sites of potassium ferrite from the result of in situ catalyst
deactivation with CO 2, it is suspected that different sites of iron oxide capable
to react with potassium may form the different active sites of potassium ferrite
for the formations of styrene, benzene and toluene.
As has been known, potassium carbonate showed the property to reduce the amount
of carbonaceous deposit. Furthermore, the obtained data on the relation between
the deposit amount and the activity of the potassium-promoted iron oxide suggest
that the carbonaceous deposit may be produced via reaction intermediates, although
the deposit is not considered to exist on the acitive sites to inhibit the acitv-
ities.

REFERENCES
1 E.P. Davies and F.T. Eggerson, U.S. Patent 2,460,311 (1949), 2,461,147 (1949),
C.L. Gutzeil, U.S. Patent 2,408,140 (1948), W.G. Gutman, U.S. Patent 3,361,683
(1968), F.J. O'Hara, U.S. Patent 3,904,552 (1975).
2 K.K. kearby, U.S. Patent 2,428,829 (1947), K. Shibata and T. Kiyoura, Bull. Chern.
Soc. Japan, 42 (1969) 871.
3 E.H. Lee, Catal. Rev., 8 (1973) 285.
4 A.K. Vijh, J. Chern. Phys., 72 (1975) 5.
5 T. Hattori, Y. Murakami, M. Ii and H. Uchida, Kogyo Kagaku Zasshi, 72 (1969)
2188.
6 F.T. Eggerson and H.H. Vonge, U.S. Patent 2~414,585 (1947).
7 Y. Ohta, Shokubai Koza, Vol. 7, p.188, Chijin Shokan (1964).
 79

8 R.R. wenner and E.C. Dybdal, Chem. End. Prog., 44 (1948) 275.
9 C.L. Gutzeil, U.S. Patent 2,449,295 (1948).
10 T. Hattori, M. Ii and Y. Murakami, J. Japan Petrol. Inst., 23 (1980) 152.
11 W.R. Price, U.S. Patent 2,971,927 (1961).
12 P. Courty and J.F. LePage, Preparation of Catalyst II, p293, Elsevier(1979).
13 C.A. Mims, J.J. Chludzinski, J.J.K. Pabst and R.T.K. Baker, J. Catal., 88 (1984)
97.
14 D.W. Mackee, Carbon, 20 (1982) 59.
15 J. Mason and B. Delmon, Proc. 5th Int. Congo Catal., Amsterdam, 1972, Paper 6-1.
16 S. Carra and L. Forni, Ind. Eng. Chem., Process Des. Dev., 4 (1965) 281.
17 A.A. Balandin and A.A. Tolstopyatova, Zh. Obshch. Khim., 17 (1947) 2182.
18 N.N. Lebedev, G.V. Odabashayan, V.V. Lebedev and M.G. Makorov, Kinet. Katal.,
18 (1977) 1441.
19 Y. Ohtsubo and K. Yamaguchi, Nihon Kagaku Zasshi, 82 (1961) 676.
20 B.D. Herzog and H.F. Rase, Ind. Eng. Chem., Prod. Res. Dev., 23 (1984) 187.