Skeletal isomerization of 1-butene on MCM-22 zeolite catalyst

Article abstract of DOI:10.1006/jcat.1996.0054

The skeletal isomerization of 1-butene has been studied on two MCM-22 zeolite catalysts synthesized with different Si/A1 ratios, i.e., 15 and 47, under a wide range of operating conditions. Product yields are a function of n-butene conversion, and a maximum of isobutene yield is found at ca. 50% conversion. Besides isobutene, propylene and pentenes, which are formed by dimerization-cracking reactions, are the main products formed on MCM-22, while C₉ and heavier hydrocarbons are not observed in the reaction products. Isobutene also takes part, together with the n-butenes, in the formation of by-products through dimerization and/or codimerization processes, which would also lead to a lower isobutene selectivity. Moreover, the yield of isobutene increases with increasing reaction temperature and decreasing 1-butene partial pressure, while the contrary holds for the formation of by-products. On the basis of these results the mechanism of formation of isobutene and the different by-products on MCM-22 is discussed. Finally, the selectivity to isobutene is strongly improved by increasing the Si/A1 ratio of the zeolite, which is explained by a decrease of the undesired consecutive reactions when decreasing the density of Bronsted acid sites of the zeolite.

Full text of DOI:10.1006/jcat.1996.0054

JOURNAL OF CATALYSIS 158, 561–569 (1996)
ARTICLE NO. 0054
Skeletal Isomerization of 1-Butene on MCM-22 Zeolite Catalyst
M. A. Asensi, A. Corma, and A. Martınez1
´
Instituto de Tecnolog´ıa Qu´ımica, UPV - CSIC, Universidad Polite´cnica de Valencia, Avda. de los Naranjos s/n, 46022 Valencia, Spain
Received April 12, 1995; revised September 20, 1995; accepted September 28, 1995
(2). Among them, halogenated aluminas and crystalline
The skeletal isomerization of 1-butene has been studied on microporous materials (zeolites and zeotypes) were the
two MCM-22 zeolite catalysts synthesized with different Si/Al
ratios, i.e., 15 and 47, under a wide range of operating condi-
tions. Product yields are a function of n-butene conversion, and
a maximum of isobutene yield is found at ca. 50% conversion.
Besides isobutene, propylene and pentenes, which are formed
by dimerization-cracking reactions, are the main products
formed on MCM-22, while C9 and heavier hydrocarbons are
not observed in the reaction products. Isobutene also takes part,
together with the n-butenes, in the formation of by-products
through dimerization and/or codimerization processes, which
would also lead to a lower isobutene selectivity. Moreover, the
most active and selective. In the case of zeolites and zeo-
types, and owing to a high concentration of reactants inside
the micropores and cavities, oligomerization of the n-bu-
tenes is a serious competing reaction that may decrease
the selectivity to isobutene and can promote catalyst deac-
tivation through hydrogen transfer and coking reactions.
The extent of such undesired bimolecular processes can
be decreased by decreasing the partial pressure of n-bu-
tenes in the feed and/or by using high reaction tempera-
tures (2). However, higher temperatures can promote ex-
yield of isobutene increases with increasing reaction tempera- tensive cracking and limit the yield of isobutene due to
ture and decreasing 1-butene partial pressure, while the con-
trary holds for the formation of by-products. On the basis of
these results the mechanism of formation of isobutene and the
different by-products on MCM-22 is discussed. Finally, the
selectivity to isobutene is strongly improved by increasing the
Si/Al ratio of the zeolite, which is explained by a decrease of the
undesired consecutive reactions when decreasing the density of
thermodynamic considerations.
Furthermore, besides the above factors, there are other
parameters that may affect the activity and selectivity
of zeolites during the isomerization of n-butenes, i.e.,
the acidity of the catalyst and the geometrical restrictions
imposed by the particular framework topology of these
materials. In the first case, it has been reported that
the presence of strong acid sites favor the undesired
oligomerization-cracking reactions, whereas weak acid
sites are more selective for isomerization (3). In this
sense, boron-substituted zeolites (4–6) and aluminophos-
Brønsted acid sites of the zeolite.
1996 Academic Press, Inc.
1. INTRODUCTION
The addition of higher amounts of oxygenated com- phates of the SAPO and MeAPO type (7, 8), having
pounds, such as methyl tert-butyl ether (MTBE), tert-amyl acid sites of weak-medium strength have been studied
methyl ether (TAME), or ethyl tert-butyl ether (ETBE) as isomerization catalysts, although they usually require
to the gasoline pool is being adopted in many refineries temperatures above 500ЊC to be active and selective,
worldwide (1) in order to reduce toxic emissions from with the corresponding detrimental effect on isobutene
vehicle exhausts, as well as to boost the octane of lead- equilibrium yield. By contrast, zeolites (aluminosilicates)
free gasoline. As a consequence, the growing demand for having strong acid sites are capable of performing the
isobutene, which is a raw material for the production of reaction at lower temperatures, but in this case the
MTBE and ETBE, can not be satisfied by the FCC (fluid second factor, i.e., the pore structure of the zeolite,
catalytic cracking) and steam cracking units which are the becomes a key parameter to achieve high yields and
main isobutene sources in the actual refinery schemes. In selectivities to isobutene. In this sense, medium pore
this respect, skeletal isomerization of the much more abun- 10MR zeolites were predicted to be more selective than
dant n-butenes appears to be a very attractive alternative large pore 12MR structures (9) by preventing extensive
for producing the required isobutene.
oligomerization inside their narrower pores. Furthermore,
A number of solid acids have been studied as potential there appear to be large differences in product selectivity
catalysts for the skeletal isomerization of linear butenes between 10MR zeolites depending on the particular topol-
ogy of the zeolite structure. For instance, unidirectional
1 To whom correspondence should be addressed.
10MR zeolites, such as Theta-1, ZSM-22, and ZSM-23
561
0021-9517/96 $12.00
Copyright 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

´
ASENSI, CORMA, AND MARTINEZ
562
are more selective than zeolite ZSM-5, the latter having 2.2. Catalytic Experiments
a bidirectional system of intersecting 10MR channels (2).
The isomerization of 1-butene (Ͼ99% pure) was per-
Moreover, ferrierite was found to be even more selective
than the above zeolites, affording high yields of isobutene
with high selectivity at relatively low temperatures (ca.
350ЊC) and without the need of using any diluent in the
feed (10). This peculiar catalytic behavior of ferrierite
has been related to its particular pore topology having
interconnected 10MR and 8MR channels (11). However,
whether or not this behavior is unique for ferrierite is
still an open question that merits testing on new zeolite
structures. In this way, the recently synthesized MCM-
22 zeolite (12) has been shown to have a peculiar structure
formed by two independent 10MR multidimensional sys-
tem of channels, one of them sinusoidal and the other
formed at atmospheric pressure in a conventional continu-
ous flow fixed-bed reactor having 10 mm inner diameter
and 48 cm length. The zeolite powder was pressed, crushed,
sieved to 0.59–0.84 mm particle size, and mixed with inert
silica up to a constant bed volume of 2.3 cm3 before being
introduced into the reactor. The catalyst was activated at
400ЊC for 1 h under 270 cm3/min of nitrogen before starting
a reaction run. Then, a mixture of 1-butene and nitrogen
(total flow rate above 300 cm3/min to avoid external diffu-
sion) is passed through the reactor (considered as time
zero) at the desired reaction temperature, with the 1-bu-
tene and nitrogen flow rates adjusted to obtain the desired
1-butene partial pressure in the feed. The 1-butene space
˚
containing large supercages defined by 12MR of 7.1 A
1
velocity, WHSV, was varied between 4.5 and 50 hϪ in
˚
inner diameter and 18.2 A height (13), which gives it
order to obtain different conversion levels, either by chang-
ing the 1-butene flow rate or the amount of zeolite (typi-
cally between 0.05–1.00 g) in the catalyst bed.
The unconverted n-butenes and the reaction products
were analyzed on line in a gas chromatograph (HP 5890
Series II) equipped with a KCl/Al2O3 column (50 m length,
0.32 mm internal diameter) and a flame ionization detector
(FID), with the first analysis being done after 30 min on
stream.
catalytic properties intermediate between 10MR and
12MR zeolites (14).
In this work we have investigated the catalytic perfor-
mance of zeolite MCM-22 for the branching isomerization
of 1-butene under a wide range of operating conditions,
i.e., temperature, 1-butene space velocity, 1-butene partial
pressure in the feed, and time on stream. Moreover, the
influence of zeolite Si/Al ratio on activity and selectivity
has also been studied. On the basis of these results the
mechanism of formation of isobutene and by-products is
also discussed.
3. RESULTS AND DISCUSSION
3.1. Physicochemical Characteristics of MCM-22
Catalysts
2. EXPERIMENTAL
2.1. Synthesis and Characterization of Zeolite MCM-22
The XRD patterns of the calcined samples A and B
Two samples of zeolite MCM-22 with different Si/Al (Fig. 1) showed the presence of a pure crystalline MCM-
ratios were synthesized following the procedure described 22 phase according to the intensity and position of the
in Ref. (15) and using hexamethyleneimine (HM) as tem- diffraction peaks previously reported (16).
plate. Gels of the molar composition 2.7 Na2O : Al2O3 :
Table 1 shows the chemical composition, and the acidity
x SiO2 : y HM : 1347 H2O were prepared from NaOH (Pro- of the two samples as measured by IR spectroscopy with
labo), NaAlO2 (Carlo Erba, 56% Al2O3 , 37% Na2O), SiO2 adsorbed pyridine and desorbed at different temperatures.
(Aerosil 200, Degussa), and HM (Aldrich), where x ϭ 30 The amount of Brønsted and Lewis sites given in Table 1
1
and y ϭ 10.5 for sample A, and x ϭ 100 and y ϭ 50 for was calculated from the intensity of the 1450 and 1550 cmϪ
sample B.
IR bands, respectively, and using the extinction coefficients
Crystallization of the gels was carried out in Teflon- given in Ref. (17). It is seen that MCM-22 sample A, syn-
lined stainless steel 60-ml autoclaves rotated at 60 rpm thesized with a lower Si/Al ratio, possesses a higher con-
at temperatures of 150 and 135ЊC for samples A and B, centration of Brønsted acid sites than sample B, in
respectively. After the autoclaves were cooled the product agreement with a higher aluminum content in the former
was filtered, washed with deionized water, and dried at sample. Both samples show a relatively high proportion
80ЊC overnight. Finally, the solids were calcined in air at of strong Brønsted sites retaining pyridine after desorption
580ЊC for 3 h, exchanged with a 2 M NH4Cl aqueous solu- at 400ЊC. The observed Lewis acidity can be related with
tion (80ЊC for 1 h), and calcined again at 500ЊC for 3 h in extraframework Al species, indicating that some dealumi-
order to obtain the acid form of the zeolites.
nation took place during the calcination steps. Thus, Table
The final catalysts were characterized by XRD (Phillips 1 indicates that the extent of dealumination was lower in
PW 1830, CuKͰ radiation), IR spectroscopy with adsorbed the case of sample B synthesized with a higher Si/Al ratio.
pyridine (Nicolet FTIR 710), and scanning electron micros-
copy (SEM).
Finally, the SEM micrographs of MCM-22 showed crys-
tals having a hexagonal platelet morphology with an aver-

1-BUTENE ON MCM-22 ZEOLITE CATALYST
563
FIG. 2. n-Butene conversion (᭡), isobutene yield (᭹), and isobutene
1
selectivity (*) obtained on sample A at 450ЊC, 28 hϪ WHSV, and 0.1
atm 1-butene, as a function of time on stream (TOS).
version is defined here as the sum of the yields of all
compounds other than n-butenes, while selectivity is given
by the yield divided by the total n-butene conversion.
Yields to the different reaction products are calculated
according to the following equation:
FIG. 1. X-ray diffraction patterns of calcined MCM-22: (a) sample
A, (b) sample B.
age diameter of 0.3–0.5 Ȑm. No large differences in mor-
phology and size of the crystallites of samples A and B
could be appreciated in the SEM micrographs.
mole of product i formed
mole of 1-butene fed
% yield (i) ϭ
ϫ 100.
[1]
The influence of time on stream (TOS), 1-butene space
velocity, reaction temperature, and 1-butene partial pres-
sure on the activity and selectivity of MCM-22 sample
A will be discussed in the following sections. Then, the
influence of zeolite Si/Al ratio will be studied by carrying
out isomerization experiments on MCM-22 sample B.
3.2. Catalytic Isomerization of 1-Butene
Under the whole range of reaction conditions studied
the 1-butene was seen to rapidly isomerize on the MCM-
22 catalysts to the cis- and trans-2-butene isomers, giving
a mixture of n-butenes in the product with a composition
very close to the predicted thermodynamic equilibrium.
Then, for simplicity we have grouped all the n-butenes to
calculate conversion and product selectivities. Thus, con-
A. Influence of time on stream (TOS). The evolution
of n-butene conversion, isobutene yield and isobutene se-
1
lectivity with TOS measured at 450ЊC, WHSV of 28 hϪ ,
and 0.1 atm of 1-butene in the feed are presented in Fig.
2. It can be seen there that conversion slightly decreases,
while both isobutene yield and selectivity slightly increase
during the period of 10 h on stream studied. The increase
of isobutene yield and selectivity with TOS has also been
observed during 1-butene isomerization on other 10MR
zeolites, such as ZSM-22 (18) and ZSM-23 (19). This effect
was explained considering that coking occurred first on the
TABLE 1
Physicochemical Properties of MCM-22 Catalysts
Chemical
analysis
Acidity (Ȑmol/g)
Brønsted
Si/Al Na2O
Lewis
Sample
ratio (wt%)
250ЊC 350ЊC 400ЊC 250ЊC 350ЊC 400ЊC strongest acid sites responsible for the formation of by-
products through oligomerization and cracking reactions,
while the sites of lower acid strength still remained active
for isomerization. Besides that, it was also suggested that
A
B
15.2
46.8
0.026
0.001
75
28
49
17
28
6
37
12
27
10
23
6

´
ASENSI, CORMA, AND MARTINEZ
564
of isobutene first increases rapidly, goes through a maxi-
1
mum at ca. 30 hϪ , and then decreases at higher WHSV.
Thus, low space velocities (high contact times) produce
low isobutene selectivities by favoring undesired secondary
reactions, such as oligomerization, cracking, and hydrogen
transfer leading to by-product formation. Besides isobu-
tene, C2 to C8 unsaturated and saturated hydrocarbons
were also formed during the reaction. Among these, pro-
pylene and pentenes were the major by-products formed
on the MCM-22 catalyst.
In order to obtain information on the different reactions
occurring during the isomerization of n-butenes on zeolite
MCM-22 the product yields have been presented as a func-
tion of conversion in Fig. 4. It can be seen that isobutene,
C3 , C5 , and saturated butanes (n-butane and isobutane)
are all formed at low conversions and can be considered
as primary products (Fig. 4a). The proportion of propane
and pentanes in the C3 and C5 fragments is initially very
low, and increases with conversion as it occurs with the
butanes. On the other hand, C2 and C6–C8 hydrocarbons
are secondary products as they are formed at higher con-
versions (Fig. 4b). No hydrocarbons higher than C8 are
observed in the reaction products in the whole range of
conversions studied.
Figure 4a shows that the yield of isobutene first increases,
passes a maximum at ca. 50% conversion, and then declines
at higher conversion levels. Above 70% conversion the
concentration of isobutene in the products becomes deter-
mined by the thermodynamic equilibrium between the bu-
tenes (dashed line in Fig. 4a), and dimerization (oligomer-
ization)-cracking become the predominant reactions. It is
also seen in Fig. 4a that the C3/C5 ratio is very close to
unity at low conversions, indicating that C3 and C5 are
initially formed from a common intermediate, i.e., C8 di-
mers which are rapidly cracked by ͱ-scission on the strong
acid sites of MCM-22. However, the C3/C5 ratio increases
at higher conversions, suggesting that some of the C3 is
also formed by cracking of hydrocarbons other than C8 .
At very high conversions (above 80%) the rate of formation
of C3 and C5 decreases (Fig. 4a), while that of C6 and C7
hydrocarbons strongly increases (Fig. 4b).
FIG. 3. Influence of 1-butene space velocity on conversion (᭡), isobu-
tene yield (᭹), and isobutene selectivity (*) on MCM-22 sample A.
Reaction conditions: 350ЊC, 0.1 atm 1-butene, and 30 min TOS.
coke deposits could reduce the free space around the acid
sites thus decreasing the effective pore size of the zeolite
and favoring isomerization against the more steric de-
manding oligomerization. Moreover, poisoning of external
surface acid sites, which would be more selective for oligo-
merization, may also contribute to the increased isobutene
selectivity as the catalyst ages. These effects could also
explain why ‘‘fresh’’ ferrierite produces initially a very high
concentration of by-products, and only when the amount
of coke deposited on the zeolite reaches a certain level
does it become very selective for producing isobutene (11).
However, in the case of MCM-22 sample A with a relatively
high Al content and having a high proportion of very strong
acid sites, there is another factor that may explain the
increase of isobutene yield with TOS, and this is the de-
crease of acid site density as the zeolite becomes poisoned.
Decreasing the acid site density would decrease the extent
of secondary consecutive reactions, such as oligomeriza-
tion, while favoring isomerization. This will be further dis-
cussed when studying the influence of zeolite Si/Al ratio
during 1-butene isomerization.
The formation of C6 and C7 may be explained consider-
ing the cracking of C12 or even larger oligomers as
i
C+8 + C4
[i C12]+
B. Influence of 1-butene space velocity (WHSV). The
effect of 1-butene WHSV on conversion, isobutene yield,
and selectivity at 350ЊC and 0.1 atm of 1-butene in the feed
2 i C6
+ i C7
[2]
[3]
ͱ-scission
H+
i
C5
is shown in Fig. 3. Unless otherwise specified, and owing while further recracking of the C6 and C7 formed could
to the low deactivation rate observed for the MCM-22 account for the increase of the C3/C5 ratio as conversion in-
catalysts, the catalytic data given in this work will corre- creases.
spond to the values obtained after 30 min TOS. Under
Recently (19), recombination of C3 and C4 fragments
these conditions conversion decreases and isobutene selec- has been considered in order to explain the formation of
tivity increases with increasing WHSV, whereas the yield hexenes and heptenes during the isomerization of n-bu-

1-BUTENE ON MCM-22 ZEOLITE CATALYST
565
FIG. 4. Product yields obtained on sample A as a function of n-butene conversion. (a) Primary products: C3 (᭿), n-butane ϩ isobutane (*),
isobutene (᭹), C5 (᭜). The dashed line represents the isobutene equilibrium yield. (b) Secondary products: C2 (᭡), C6 (᭞᭝), C7 (᭢), C8 (X). Reaction
conditions are as in Fig. 3.
tenes on ZSM-23 zeolite. This zeolite has unidimensional ments on the MCM-22 catalyst, feeding pure 1-butene,
˚
10MR channels of 4.5 ϫ 5.2 A diameter (20), and therefore, pure isobutene, and 1-butene/isobutene mixtures under
formation of branched C12 or larger oligomers inside its identical reaction conditions, i.e., 0.1 atm partial pressure
channels should be strongly hindered. In the case of MCM- of olefin, 350ЊC reaction temperature, and WHVS of 28
1
22, oligomerization may well take place in the large 12MR hϪ based on total butene fed. The molar composition of
cavities, although it should be more impeded in the tortu- the reaction products obtained after 30 min TOS is given
ous 10MR system of MCM-22. However, even if the oligo- in Table 2. Several general conclusions can be derived from
mers can be formed in the large cages, they should diffuse these experiments. First, a very close product distribution,
out through 10MR windows, thus favoring their cracking with the obvious exception of n-butenes and isobutene, is
into lighter fragments. This could also explain the absence obtained when feeding either 1-butene or isobutene, in
of C9 and higher hydrocarbons and the relatively low con- agreement with the results reported in Ref. (19) on a ZSM-
centration of octenes in the reaction products. Neverthe- 23 zeolite. Furthermore, we also observed a similar distri-
less, dimerization and/or codimerization of C3–C5 cannot bution of by-products when feeding the 1-butene/isobu-
be ruled out in order to explain the distribution of by- tene mixtures. These results may indicate that both n-
products at very high conversions.
butenes and isobutene participate in the formation of by-
The complex product distribution observed during 1-
butene isomerization clearly shows the difficulty in estab-
lishing a simple reaction mechanism, since, besides the
skeletal isomerization, a large number of consecutive and
parallel reactions also take place. Indeed, there are contro-
TABLE 2
Product Distribution Obtained on MCM-22 Sample A when
Feeding 1-Butene, Isobutene, and 1-butene/Isobutene Mixtures
versial opinions in the literature concerning the mechanism at 350؇C, 0.1 atm of Olefin, 28 h؊1 WHSV, and 30 Min TOS
of formation of isobutene and by-products. In the latter
case, some authors claimed that by-products are formed
Composition of the
olefin feed (%)
Product distribution (mol%)
nC4ϭ iC4ϭ
C5 C6
exclusively from the n-butenes (4), while others stated that
both n-butenes and isobutene can lead to the formation
of by-products (19). The latter assumption was based on
the observation that a similar product distribution was
obtained when feeding n-butenes or isobutene on zeolite
ZSM-23 under the same experimental conditions.
1-Butene Isobutene
C2
C3
C4Њ
ϩ
100
83
67
0
0
17
33
0.22 10.20 2.12 57.39 17.93 8.80 3.34
0.18 9.39 1.74 53.32 23.60 8.61 3.16
0.18 9.76 1.82 47.81 28.55 8.77 3.10
0.17 8.74 1.67 18.41 57.11 8.42 5.49
100
In our case we have carried out isomerization experi-

´
566
ASENSI, CORMA, AND MARTINEZ
TABLE 3
Influence of Reaction Temperature on Conversion and Product Yieldsa
Yield (mol%)b
i-C4ϭ
% i Ϫ C4ϭ
with respect to
equilibrium
Conversion
(mol%)
T (ЊC)
C2
C3
C4Њ
C5
C6
ϩ
250
300
350
400
450
500
11.9
22.9
52.3
51.7
55.1
49.8
0.0
0.1
0.4
0.6
1.5
2.1
1.1
2.9
13.0
12.8
15.1
11.2
0.6
1.1
2.9
1.9
2.3
1.3
4.4
9.2
21.5
25.7
26.5
29.6
2.2
4.2
10.4
8.4
7.9
5.2
3.6
5.4
4.2
2.3
1.7
0.5
10
23
71
85
95
100
a Reaction conditions: WHSV ϭ 28 hϪ1, 0.1 atm of 1-butene, TOS ϭ 30 min.
b C4Њ ϭ isobutane ϩ n-butane. The other fractions include both saturated and unsaturated hydrocarbons.
products, and that the dimerization-cracking process, tenes are the predominant by-products in the 350–500ЊC
range, owing to a higher extent of cracking reactions at
high temperatures.
the main secondary reaction leading to propylene and
pentenes, is very similar whether the C8 intermediates are
formed from dimerization of n-butenes, codimerization of
n-butenes with isobutene, or dimerization of isobutene.
The decrease of the total yield of by-products above
350ЊC probably occurs because the dimerization of the
butenes becomes thermodynamically unfavored at high
reaction temperatures. The different behavior observed
for isobutene strongly suggests that this product and the by-
products are really formed on MCM-22 through different
mechanisms. Usually, it is assumed that isobutene is pro-
duced from n-butenes by a monomolecular mechanism
involving the formation of a very unstable primary carbo-
cation.
n
C4 + n C4
C4 + i C4
C4 + i C4
[4]
[5]
[6]
cracking
n
[i C8
]
C3 + C5
i
Moreover, by comparing the results obtained when feed-
ing pure 1-butene and pure isobutene it can be concluded
that the skeletal isomerization of n-butenes is a reversible
process with both isomerization of n-butenes to isobutene
and that of isobutene to n-butenes occurring at similar
reaction rates.
Reactions [5] and [6], in which isobutene participate,
would be increasingly important as the concentration of
isobutene in the reaction products increases, and would
explain the decrease in isobutene yield observed in Fig.
4a in the 50–70% conversion range, before the isobutene
yield starts to be limited by the thermodynamic equi-
librium.
C
C
C
C
C
C
C
C
+
C
+
+
H
C
C
C
H+
+
C
H+
C
C
C
C
C
C
C
C
C
C
C
However, in some cases it has also been suggested that
isobutene can be formed by a bimolecular mechanism in-
volving the dimerization of n-butenes, rearrangement of
the C8 intermediate, and further cracking.
2 n C4
[i C8
]
2 i C4
[7]
C. Influence of reaction temperature. The effect of tem- Indeed, this mechanism has been proposed to occur on
perature on conversion and product yields is shown in the ferrierite zeolite in order to explain the high selectivity
1
Table 3 at 28 hϪ WHSV and 0.1 atm partial pressure. of this catalyst (11), and also on modified aluminas where
Under these conditions the n-butene conversion rapidly in some cases isobutene concentrations above the thermo-
increases up to 350ЊC and then remains practically constant dynamic equilibrium were observed (21, 22).
at higher temperatures. The yield of isobutene continu-
In the case of MCM-22 catalyst the results of Table 3
ously increases and reaches the predicted equilibrium yield allow us to advance that, although reaction [7] cannot be
at ca. 500ЊC. On the other hand, the total amount of by- completely excluded, this is not the main mechanism pro-
products first increases, passes a maximum at ca. 350ЊC ducing isobutene, since otherwise the yield of isobutene
and then decreases upon further increase of reaction tem- and by-products, which are really formed through a dimer-
perature. At temperatures below 350ЊC C6–C8 hydrocar- ization-cracking pathway, should follow the same trends
bons are the major by-products, while propylene and pen- with reaction temperature. Further support to this will be

1-BUTENE ON MCM-22 ZEOLITE CATALYST
567
TABLE 4
Effect of 1-Butene Partial Pressure on
Product Selectivities at ca. 50% Conversion
at 350؇C Reaction Temperature and 30
Min TOS
1-Butene
partial
Selectivities (mol%)
pressure
(atm)
C2ϭ
C3ϭ
i-C4ϭ
C5ϭ
C6
ϩ
0.1
0.5
1.0
0.8
0.7
0.9
24.4
22.2
21.7
41.0
31.7
26.6
18.8
18.9
18.7
8.0
19.4
23.7
opposite is observed for the by-products. Moreover, the
selectivity to propylene and pentenes remains almost con-
stant whereas that of C6–C8 compounds increases with
partial pressure. These results indicate a strong increase
of the bimolecular dimerization-oligomerization processes
with increasing 1-butene partial pressure, and strongly sug-
gest that isobutene is mainly formed on MCM-22 through
a monomolecular mechanism, and not via dimerization-
cracking as is the case of propylene and pentenes.
FIG. 5. Influence of 1-butene partial pressure for MCM-22 sample
A on conversion (᭡), isobutene yield (᭹), and total by-products yield
(᭿). The dashed line represents the isobutene equilibrium yield. Reaction
1
conditions: 350ЊC, 28 hϪ WHSV, and 30 min TOS.
given in the following section which discusses the effect of
1-butene partial pressure.
E. Influence of the Si/Al ratio of MCM-22 zeolite.
Figure 6 compares the product yields obtained on the
MCM-22 samples A (Si/Al ϭ 15) and B (Si/Al ϭ 47) at
D. Influence of 1-butene partial pressure. The partial
pressure of 1-butene in the feed is expected to have a larger
influence on bimolecular than on unimolecular processes.
Figure 5 shows the yield to isobutene and total by-products
as a function of 1-butene partial pressure at 350ЊC and 28
1
350ЊC, 0.1 atm of 1-butene, and WHSV of 28 hϪ . The rates
of formation of the different reaction products on the two
MCM-22 samples under the above reaction conditions are
also given in Table 5. It can be seen in Fig. 6 that the
isobutene yield remains practically constant while that of
by-products strongly decreases with an increasing Si/Al
ratio of the zeolite. According to the IR-pyridine data
1
hϪ WHSV. It is clearly seen that isobutene and by-prod-
ucts follow different trends when increasing the olefin par-
tial pressure. The yield of isobutene first increases, goes
through a maximum at ca. 0.1 atm, and then continuously
decreases with the 1-butene partial pressure. Above ca.
0.5 atm the isobutene yield practically follows the predicted
equilibrium yield (dashed line in Fig. 5). On the contrary,
the yield of by-products strongly increases when increasing
the 1-butene partial pressure as a consequence of the in-
crease of bimolecular processes, such as dimerization-
oligomerization of the butenes. The decrease of isobutene
yield observed above 0.1 atm may be attributed to the
increase of dimerization and codimerization processes (re-
actions [5] and [6]), and also to the limited equilibrium
yield of isobutene at high n-butene conversions.
Since the n-butene conversion increases with 1-butene
partial pressure (Fig. 5), and product yields are a function
of conversion (Fig. 4), the effect of partial pressure can be
better seen when comparing the product selectivities at
constant conversion and different 1-butene pressures (Ta-
ble 4). It is clearly seen that isobutene selectivity continu-
FIG. 6. Influence of the Si/Al ratio of MCM-22 zeolite on product
ously decreases with 1-butene partial pressure, while the yields obtained at 350ЊC, 0.1 atm 1-butene, 28 hϪ1 WHSV, and 30 min TOS.

´
ASENSI, CORMA, AND MARTINEZ
568
TABLE 5
plain the higher isobutene selectivity obtained on the
MCM-22 sample with higher Si/Al ratio.
Rates of Formation (mmol/h gcat)
of the Different Reaction Products
on MCM-22 Samples A (Si/Al ؍ 15)
and B (Si/Al ؍ 47) at 350؇C, 0.1
atm of 1-Butene, 28 h؊ WHSV, and
30 Min TOS
The fact that the isobutene yield remains at the same
level while that of by-products is considerably reduced
upon increasing the zeolite Si/Al ratio suggests again that
isobutene should be mainly formed on MCM-22 via direct
isomerization of n-butenes, since monomolecular pro-
cesses would be much less affected than bimolecular ones
when changing the acid site density in the zeolite.
Product
Sample A
Sample B
Furthermore, Table 6 shows the effect of reaction condi-
tions on n-butene conversion and product yields for the
high Si/Al ratio sample B. The same general trends that
were already discussed for sample A can be applied for
sample B. Increasing the reaction temperature increases
both the yield and selectivity to isobutene, while increasing
the 1-butene partial pressure favors undesired dimeriza-
tion-oligomerization reactions thus decreasing the yield
and selectivity to isobutene. Then, at nearly optimum op-
erating conditions, a selectivity to isobutene of ca. 80% at
ca. 40% conversion is obtained on MCM-22 sample B.
Methane
Ethane
Ethylene
Propane
Propylene
iso-Butane
n-Butane
iso-Butene
iso-Pentane
n-Pentane
Pentenes
0.0
0.0
2.1
0.0
0.0
0.2
0.1
16.1
1.6
1.4
102.8
0.1
0.1
17.8
11.1
1.0
64.2
10.7
3.6
107.8
2.7
0.3
49.3
19.4
C6
ϩ
4. CONCLUSIONS
already shown in Table 1 there are not big differences in
the acid strength distributions of the Brønsted sites in the
Zeolite MCM-22 has been shown to be an active and
two samples, but sample B has a much lower acid site stable catalyst for the skeletal isomerization of 1-butene.
density owing to its lower Al content. This indicates that, The activity and selectivity were seen to depend on the
besides the possible influence of acid strength distribution, reaction conditions, i.e., temperature, 1-butene partial
in the case of zeolites with strong Brønsted acid sites, such pressure, and olefin space velocity. A maximum of isobu-
as MCM-22, the acid site density, which in turn depends tene yield was observed at ca. 50% n-butene conversion.
on the framework Si/Al ratio of the zeolite, is a key param- At low conversions, isobutene, propylene, and pentenes,
eter to control the selectivity to isobutene. This is explained the latter two compounds formed by dimerization of the
by a decrease of the extent of consecutive reactions leading butenes followed by cracking, are the primary products
to by-products as the acid sites become more and more obtained on MCM-22. In addition, some n-butane and
isolated. On top of that, the decrease of adsorption proper- isobutane are also initially formed through hydrogen trans-
ties of the zeolite upon increasing the framework Si/Al fer reactions. At high conversions the concentration of C6
ratio (23), which will decrease the concentration of olefins and C7 compounds strongly increases while that of isobu-
inside the zeolite pores thus decreasing the extent of bimo- tene decreases due to the increase of secondary consecutive
lecular processes, may also be considered in order to ex- reactions and to the limited equilibrium yield of isobutene
TABLE 6
Influence of Reaction Conditions on Conversion and Product Yields Obtained at 30 min
TOS During the Isomerization of 1-Butene on the MCM-22 Sample B (Si/Al ؍ 47)
Reaction Conditions
Yield (mol%)
Pp (1 Ϫ C4ϭ)
WHSV
Conversion
(mol%)
1
T (ЊC)
(atm)
(hϪ
)
C2
C3
C4Њ
i Ϫ C4ϭ
C5
C6
ϩ
350
350
400
450
350
350
0.1
0.1
0.1
0.1
0.5
1.0
14
28
28
28
28
28
46.3
30.1
36.3
36.3
60.3
76.8
0.1
0.1
0.1
0.2
0.3
0.5
6.2
3.2
4.5
3.5
9.5
1.0
0.6
0.6
0.5
3.3
7.3
29.4
26.1
26.1
28.4
22.2
15.8
6.4
4.1
4.1
3.1
11.6
16.7
3.2
0.9
0.9
0.6
13.4
25.8
10.7

1-BUTENE ON MCM-22 ZEOLITE CATALYST
569
in the product stream. Formation of C6 and C7 probably
REFERENCES
occurred by cracking of C12 or even larger oligomers that
could be formed in the 12MR large cavities present in
MCM-22. However, the concentration of C8 was relatively
low in the whole range of conversions, and no C9 and higher
hydrocarbons were observed in the reaction products. This
was ascribed to the high cracking activity of the strong
acid sites of MCM-22 (24), and also to a strong limitation
for the large oligomers to diffuse out of the 12MR cages
through the 10MR windows.
1. Maxwell, I. E., and Naber, J. E., Catal. Lett. 12, 105 (1992).
2. Butler, A. C., and Nicolaides, C. P., Catal. Today 18, 443 (1993).
3. Choudary, V. R., Chem. Ind. Dev. 8, 32 (1974).
4. Bianchi, D., Simon, M. W., Nam, S. S., Xu, W.-Q. Suib, S. L., and
O’Young, Ch. L., J. Catal. 145, 551 (1994).
5. Simon, M. W., Xu, W.-Q., Suib, S. L., and O‘Young, Ch. L., Micropo-
rous Mater. 2, 477 (1994).
6. O’Young, Ch. L., Xu, W.-Q., Simon, M. W., and Suib, S. L., ‘‘Studies
in Surface Science and Catalysis,’’ Vol. 84, p. 1671. Elsevier, Amster-
dam, 1994.
Experiments feeding isobutene and 1-butene/isobutene
mixtures demonstrated that the skeletal isomerization is a
reversible process, and that isobutene also intervenes in
the formation of by-products, probably by dimerization
and codimerization reactions, leading to a distribution of
by-products closely resembling that obtained when feeding
pure 1-butene.
The yield of isobutene increased with increasing reaction
temperature and decreasing 1-butene partial pressure,
while the opposite behavior was observed for the formation
of by-products. These results strongly suggested different
reaction mechanisms operating for both isobutene and by-
product formation. Thelatter are formed through dimeriza-
tion, oligomerization, and cracking reactions, while isobu-
tene may be mainly formed on MCM-22 via direct isomer-
ization of n-butenes through a monomolecular mechanism.
Finally, the selectivity to isobutene was strongly im-
proved when the Si/Al ratio of the zeolite increased from
15 to 47. This was ascribed to a reduction of the bimolecular
consecutive reactions with decreasing acid site density of
the zeolite. Thus, a selectivity to isobutene of ca. 80% at
ca. 40% conversion was obtained on the high Si/Al ratio
MCM-22 sample at near optimum operating conditions.
7. Zubowa, H. L., Richter, M., Roost, V., Parlitz, B., and Fricke, R.,
Catal. Lett. 19, 67 (1993).
8. Yang, S. M., Guo, D. H., Lin, J. S., and Wang, G. T., ‘‘Studies in
Surface Science and Catalysis,’’ Vol. 84, p. 1677. Elsevier, Amster-
dam, 1994.
9. Thomas, J. M., Sci. Am. April, 112 (1992).
10. Grandvallet, P., de Jong, K. P., Mooiweer, H. H., Kortebeek, A. G.
T. G., and Kraushaar-Czarnetzki, B., European Patent 501.577 (1992).
11. Mooiweer, H. H., de Jong, K. P., Kraushaar-Czarnetzki, B., Stork,
W. H. J., and Krutzen, B. C. H., ‘‘Studies in Surface Science and
Catalysis,’’ Vol. 84, p. 2327. Elsevier, Amsterdam, 1994.
12. Rubin, M. K., and Chu, P., U.S. Patent 4,954,325 (1990).
13. Leonowicz, M. E., Lawton, J. A., Lawton, S. L., and Rubin, M. K.,
Science 264, 1910 (1994).
14. Corma, A., Corell, C., Llopis, F., Martınez, A., and Perez-Pariente,
´
´
J., Appl. Catal. A 115, 121 (1994).
15. Corma, A., Corell, C., and Perez-Pariente, J., Zeolites 15,
2
´
(1995).
16. Abril, R. P. L., Bowes, E., Green, G. J., Marler, D. O., Shihabi,
D. S., and Socha, R. F., U.S. Patent 5,085,762 (1992).
17. Hughes, T. R., and White, H. M., J. Phys. Chem. 71, 2192 (1967).
18. Simon, M. W., Suib, S. L., and O’Young, Ch. L., J. Catal. 147, 484
(1994).
19. Xu, W.-Q., Suib, S. L., and O’Young, Ch. L., J. Catal. 150, 34 (1994).
20. Wright, P. A., Thomas, J. M., Millward, G. R., Ramdas, S., and Barri,
S. A., J. Chem. Soc. Chem. Commun., 1117 (1985).
21. Cheng, Z. X., and Ponec, V., Appl. Catal. A 118, 127 (1994).
22. Cheng, Z. X., and Ponec, V., J. Catal. 148, 607 (1994).
23. Corma, A., Faraldos, M., Martınez, A., and Mifsud, A., J. Catal. 122,
´
ACKNOWLEDGMENT
230 (1990).
24. Corma, A., Mart´ınez-Alfaro, V., and Orchille´s, A. V., Appl. Catal.,
Financial support by the Comision Interministerial de Ciencia y Tecno-
´
logıa of Spain (Project MAT 94-0166) is gratefully acknowledged.
in press.
´