Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
72Si
HOE ~ljS 020
The invention relates in the first place to a
process for oxidatively dehydrogenating dehydrogenatable
organic compounds to prepare styrene.
A known process which involves the thermal de-
hydrogenation of ethylbenzene (EB) as an equilibriumreaction affords conversions of EB of only 35 - 40% while
the styrene selectivities are high at 85 - 95% [E.D. Haus-
mann and C.J. Kiwo, Ind.Eng.Chem., Fundamentals 5 295
(1966)]. The oxidative dehydrogenation of EB to give
styrene as a~ alternative process to thermal dehydrogena-
tion was initially only successful when using promoters
such as S02 [G.S. Schaffel and R.L. Marcell, paper pre-
sented at the 71st AICHE Meeting9 Dallas, Texas, February
20-23, lg723, H2S lM. Vadekar, J.S. Pasternak and N.J.
Gaspar, Can.J.Chem. Eng. 52 (1974) 788] and halogens such
as I2 IJ.H. Raley, R.D. Mullineaux and C.W. Bittner,
J.Am.Chem. Soc., 85 3174 (1963)~. The use of oxygen as
sole oxidation agent and of known catalysts lC.R. Adams,
H.H. Voge, C.Z. Morgan and W.E. Armstrong, J. Catalysis 3
379 ~1964)~ led to uncontrolled combustion and crackin
reactions.
Although the use of alkaline earth/nickel phosphate
catalysts, as described in U.S. Patent 31953,126, produced
high conversions and selectivities for the reaction with 2 as
sole oxidation agent, the space-time yield ol` styrene was
only moderate.
$
In contrast, the object of the invention is to
achieve a higher space-time yield of styrene (relative to
the amount of catalyst used) and to produce chiefly C0 and
C0? as by-products.
This object is achieved by reacting dehydrogenat-
able organi~ compounds in the vapor phase with oxygen at
a temperature above 350C in the presence of a new cata-
lyst based on zirconium phosphate. In this reaction oxy
gen is used as sole oxidation agent. Suitable dehydrogen-
atable organic compounds are saturated alkylaromatics,
above all ethylbenzene, but also, for example, diethyl-
benzene or ethylnaphthalene. The catalyst used is prefer-
ably zirconium phosphate which has been treated with
orthophosphoric acid. The catalyst has a zeolite struc-
ture and the properties of an inorganic ion exchangematerial. It therefore preferably also contains exchanged
ions of the groups Ia, IIa, IVb, Vb, VIb, VIlb and VIII
of the periodic system of the elements, namely generally
in a concentration of 1-20% by weight. Suitable examples
are the K, Cs, Mg, Ba, Cr, Mn, Fe and Ni ions, in particu-
lar K, Cs, Ba~ Fe or Ni ions.
Conversion and selectivity are particularly favor-
able when a mixture of an oxygen-containing gas and a
dehydrogenatable compound, with or without dilution by an
inert subs-tance, in the dilution case the ratio of the
diluent to the dehydrogenatable compound being at most 10/~,
is passed over the zirconium phosphate catalyst at tem-
peratures of 350C - 550C and residence times of 0.5 to
5~ preferably 0.7 to 3.2, in particular 1.0 - 1.8, grams
72~
of gas mixture per gram o~ catalyst per hour. The de-
hydrogenatable organic compound preferably used is ethyl-
benzene 9 in which case the temperature range is from about
- 380C to 530C.
The invention also relates to an agent for carry-
in~ out the~abovementioned process, in the form of a zir-
conium phosphate catalyst as described above.
Zirconium phosphate is prepared, for example~by
the method described by Clearfield and Stynes in J.Inorg.
Nucl.Chem. 26 117 (1964).
In this method, a gel-like amorphous precipitate is
first obtalned~by adding an excess of phosphoric acid or of
a soluble phosphate to a soluble zirconium salt. This can
- be carried out, for example~by mixing an 0.5 M zirconium
(IY) dichloride oxide solution with a 3 M orthophosphoric
acid in a ratio which is such that the atomic ratio of
zirconium to phosphorus is 1:2.
This precipitate is then converted into a crystal-
line form, namely by impregnating the gel-like precipitate
with an a-t least 10 M, preferably 10 - 15 M9 orthophos-
phoric acid; preferably by adding an excess, at least l
liter, of orthophosphoric acid of the abovementioned con-
centration to each 100 g of the gel-like precipitate and
maintaining the resulting mix-ture for 10 to 14 hours, with
stirring, at temperatures around -the boiling point. The
impregnating step is completed by then leaving the pre-
cipitate to stand in the orthophosphoric acid for 30 to 40 hours.
The catalyst preferably has a surface area of 10 -
50 rn2/g, in particular of 25 - 40 rn2/g. This surface area
2~i
can be obtained by establishing a suitable acid strength
~at leas-t 10 M) and a sufficiently strong turbulence in
the impregnating vessel.
The resulting crystalline zirconium phosphate is
~iltered off from excess phosphoric acid and then washed
with wa-ter until the water draining off has a pH of 2 - 3.
The zirconium phosphate is dried in air at tem-
peratures between 120C and 250~C, preferably between 120C
and 200C, for a period of 24 - 48 hours.
The ratio of P04 to Zr- becomes established in
the preparative method described above at values between
2 and 3. The favorable incorporation mentioned of addi-
tional metal ions is effected by ion exchange in which the
zirconium phosphate catalys-t, after the drying step, is
stirred for 50 to 80 hours in a corresponding metal salt
solution.
Because of their size, Cs ions can be incorporated
into the crystal latice only by adding a corresponding
solution before the impregnation with orthophosphoric acid
and the drying. For example, a 1 N solution of CsCl is
added to the gel-like zirconium phosphate precipitate and
12 M orthophosphoric acid is then added for the impreg-
nation. The other steps in the procedure thereafter
correspond to those used in the preparation of undoped
crystalline zirconium phosphate.
The dried product is comminuted and either pellet--
ized or classified by screening.
The process according to the invention is distin-
guished by t}le following process data.
-- 6 --
The rnolar ratio of oxygen to alkylaromatics can
be 0.5-2 moles cf 2 per mole of alkylaromatic, the prefer-
able range being 0.7 to 1.~, in particular 0.9-1.2, moles of 2
p-er mole of alkylaromatlc. The oxygen can be puie oxygen
or air. The a~ditional diluents used can be noble gases,
nitrogen, C~2 or steam. These gases can be used in
amoun-ts of up to 10 moles, preferably of 4 to 10 moles,
in particular of 4 to 7 mcles per mole of alkylàromatic.
The pressure under which the reaction is carried
out is generally within a range of 0.5-20 bar, preferably
0.5-3 bar~in particular 1 to 2 bar.
Thetzirconium phosphate catalyst has a stable long-
term behavior with conversions of 40-70% at styrene selec-
tivities of 80 to 95%. The catalyst is regenerated after
a relatively long operating period through only incomplete
burning-off, preferably at 400 - 540C, of carbon-contain-
ing deposits by means of oxygen or gases containing mole-
cular oxygen. Regeneration is complete when the deposi-
tion on the catalyst has gone down to 7-15% by weight of
carbon-containing deposits. --
The exarr,ples which follow are intended to illus~
trate the invention in more detail but not to restrict it
in any way.
Example 1
Quantities of 0.5 M zirconium chloride solution
and 3 M orthop}losphoric acid were added together which
were such that the atomic ratio of zirconium to phosphorus
WclS 1:2. The resulting gel-like zirconium phosphate was
added in an amount of 100 g to an excess, at least 1 liter,
3 7~5i
of 12 hl orthophosphoric acid, and the resulting mix-ture
was heated under reflux up to the boil with stirring for
12 hours. Thereafter, the zirconium phosphate was left
for a f~rther 38 l,ours in this strong phosphoric acid.
The resulting crystalline zirconium phosphate was
washed with H20 until the H20 draining off had a pH of
'~
..
The material was dried for 24 hours in an air stream
at a temperature of 120C.
Two samples of catalyst, namely Zr-P-(1) and Zr-P-
(3), were prepared in accordance with theseinstructions.
. The use of only 1/3 of the amount of 12 M ortho-
phosphoric acid but otherwise the same treatment as indi-
cated above produced catalyst Zr-P-(2).
The activities of the catalysts were investigated
on an integral micro-reactor~ The latter can have, for
example, a height o~ packing of 100 mm and a diameter of
12 mm. An integral mlcro-reactor of this type can also
be used in the remaining examples. The amount of ca-talyst
used was in each case 11 cm3~ Screened material of
particle size 0.63 - 1 mm was used. The reaction tempera-
ture was set at 450C. The other operating conditions
and experimental results ob-tained by means of the three
samples of catalyst are shown in the table below. In that
table, and in the following examples, RT denotes residence
time, C denotes conversion, S denotes selectivity and
EB denotes ethylbenzene.
~L~17~5i
~T~s] EB/N2 EB/02 Zr-P-(1) Zr-P-(2) Zr-P-(3)
lmole/ [mole/ C S C S C S
mole] mole] ~%] [%] [/0] [%] [%] l%]
0.71 1/7 1/1 39.8 86.1 9.5 82.1 37.0 89.1
1.5 1/7 1/0.7 39.2 86.4 16.3 76.2 40.6 89.1
The results show the very strong influence of the
phosphoric acid treatment on catalys-t activity.
Example 2
To demonstrate the particularly favorable product
spectrum obtained by means of the zirconium phosphate
catalysts, the results of two performed experiments are
shown. In experiment A~ pure ethylbenzene (EB) was used
as starting hydrocarbon. The RT of the reactants was
0.71 [s].
In experiment B, an equimolar ethylbenzene/styrene
mixture was used (RT likewise being 0.71 [s]). The cata-
lyst used was sample Zr-P-(1) which was used in the sa~e
amount as in Example 1. Oxygen and N2 ~as inert material)
were added in an amount which was such that the molar
ratios of HC/02 = 1~1 and HC/N2 = 1~7 were produced (HC =
hydrocarbons = EB plus styrene).
The following results table shows that the only
by-products observed in relatively large amounts were CO
and C02, while toluene and benzene (indicatecl in the table
as o.HC = other hydrocarbons) were found only in traces.
MOI.E %
Experim~n-t C0 C02 EB STYRENE 2 H20 0 . HC C [/OJ S [/0]
A 6 . 09 12 . 50 24 . 46 15 . 46 12 . 04 29 . 40 0 . 05 42 . 7 84 . 8
B 6 .19 10. 79 20. 55 37. 39 8. 66 16 . 41 0. 02 31. 42 7~3. 8
Exar,nple B also shows at the same time that still
high selectivities can be obtained when reacting a mixture
having a high styrene content. This is particularly
important when consumed oxygen is to be replaced in an
industrial reactor by intermediate feeding~in of addi-
tional oxygen whereby further reaction then becomespossible. -~ To clarify this issue, a theoretical
model for a multi-stage reactor with additional
oxygen between the reactor stages was cornputed on the
basis of kinetic equations. The projection was carried
out for a production level of 70,000 tonnes per annum of
styrene and use of the catalyst Zr-P-(lj. At the entry
of the 1st reactor stage 2 is added in the form of air.
Between the reactor stages pure oxygen is added so that
the EB/02 ratios shown in the table below are produced.
The table shows the conversions and selectivities which
can be obtained in each reactor stage.
72~
Reactor Length Oxygen TIN ToUT C S
stage [m3 at entrY [C~ [C] 1%~ l~/ol
lmole/
mole]
1 0.57I/1 400 480 22 83
2 0.681/1 415 ~0 39.3 8Z.5
3 0.651/1.5 425 480 53.5 81
4 1.4~1/1.5 425 ~0 66.6 79.4
10 5 1.661/2 ~25 442 71 79.1
.
Tln denotes the temperature of the mixture on entry, and
ToUt denotes the temperature of the mixture on leaving the
particular stage.
Example 3
To prepare a zirconium phosphate catalyst doped
with Cs ions, 50 ml o~ 0.2 M Cs2S04 solution were added
to 400 ml of the washed gel-like zirconium phosphate and
the mixture was stirred for 32 hours. Further treat-
- ment was as indicated in Example 1.
An activity investigation on the integral micro-
reactor (amount of catalyst 11 cm3) produced -the conver-
sions and selectivities indicated in the table. The reac-
tion temperature in the experiments was 450C.
RT [s3 ~B/~2 EB/02 C [%~ S 1%]
[mole/mole] [mole/mole3
0.71 1/7 1/1 36.9 ~9.9
1.5 1/7 1/0.7 ~0.2 90.7
-- 11 .
Example ~
Exyeriments to investigate the long-term behavior
were carried ou~ on the catalyst Zr-P-(1) using different
residence times (A : 2.47 s, B : 0.71 s) of the reactan-ts
on the catalyst. The reaction temperature in both experi-
ments was 450C. The molar ratios set were EB/02 = 1/1
and EB/inert 1/7.
The results are shown in the following table.
Long-term experiment A Long-term experiment B
Duration Conver- Selec- ~uration Conver- Selec-
o~ sion tivity ofsion tivity
experi- l%~ [%] experi- [%] [%]
ment ment-
lh] ~h]
0.3 63.~ 83.2 1.0 28.087.~
1.0 67.2 91.2 8.0 4~.088.3
1.5 65.2 90.7 9.5 48.087.1
1.8 61.1 90.0 13.5 50.185.0
2.5 5~.1 86.2 18.~ 50.484.7
5.0 51;2 84.0 26.1 49.883.6
g.5 52.2 83~0 33.6 48.982.5
16 50.4 80 37.8 47.g~1.7
24 50.3 80.5 41.0 48.181.2
4~ 50.2 80.4 80 47.980.8
The data inclicated in the table show that there
is an activity maximum after a short operating period.
On operating for longer periods an almost constant activity
value is then obtained. It ~as possible to maintain this
activi-ty or a period of 160 hours - 200 hours. Only thcn
J~
- 12
did the lowering of the activity necessitate regeneration
of the catalyst. It was possible to effect this by simply passing
over N2/02 in the form of air at a temperature of 450C.
In this step, it was not necessary to burn off the car-
bonization products until the catalyst was completelycarbon-free. It was sufficient to regenerate down to a
deposition on the catalyst of 7 ~ 15% by weight of carbon-
containing deposits. This means that the regenerating
period is considerably shortened from thP usual 6.5 hours to about
1.5 hours. After a regeneration of this type, as was
necessary after about 160 hours in the long-term experi-
merit B, conversions of 48% and sel~ctivities of 81.5% were
then again obtainable. The catalyst Zr-P-(1) was used for
a total of 3,500 hours in the integral micro-reactor.
1~ During this period the catalyst was regenerated àbout 30
times and even toward the end of the investigation period
it was possible to obtain experimental results as had
already been obtained at tne start.
Example 5
To investigate the effect of the concentration of
the inert material, experiments were carried out using
various EB/N2 ratios. Again catalyst Zr-P-(1)
(amount~ 11 cm3) was used at a temperature of 450C in
the integral micro-reactor. On entry into the reactor
oxygen was added in an amount such that a EB/02 ratio
of ljl was produced. Ethylbenzene was added at a rate of
3.2 g/h. The results are shown in the following table.
- 13 -- ,
EB/N2 ~mcle/mole~ C 1%] S 1%]
1/4 60.0 70.0
1/7 53.0 75.0
~/10 52.5 7~.0
Example 6
The influence of the concentration of oxygen was
investigated at a content of inert materlal (inert mater-
ial N2) of EBfN2 - 1/7. In these experiments, ethyl-
ben~ene was added at a rate of 2.9 g/h (catalyst Zr-P (1) :
10 11 cm3 ). The results of these experiments have been
collated in the following table.
T [CI EB/02 lmole/mole] C l%~ S [%]
430 1/0 0
430 1/0.5 40.5 95.0
43C 1/1.0 S0.3 88.9
430 1/1.5 58.1 8~.1
430 . lt2.0 60.3 ~1.8
435 1/2.3 61.6 ~1.1
460 1/2.7 68.8 76.2
S00 1~3.0 76.5 70~4
540 1/3.2 79.4 ~4,0
Example 7
To investigate the influence of the reaction tem-
perature, experiments were carried out at differing tem-
perature but otherwise identical reaction conditions.
- 14 -
Catalyst Zr-P~ was used in an amount Or 11 cm3. Ethyl-
benzene was added at a rate of 11.1 g/h. The molar feed
ratios EB/N2 and EB/02 were set a-t the values 1/4 ancl 1/1
respectively (N2 and 2 in the form of air). The conver-
sions and selectivities obtained were as shown in thefollowing table.
T [CI C l%] S 1%]
400 21.1 86.2
420 26.3 8~.4
430 29.8 84.8
450 39.3 85.4
460 42.'l 84.0
485 48.6 84.1
500 50O1 83.5
