Alkylation of benzene with short-chain olefins over MCM-22 zeolite: Catalytic behaviour and kinetic mechanism

Article abstract of DOI:10.1006/jcat.2000.2849

Cumene and ethylbenzene are important compounds in the petrochemical industry for the production of phenol and styrene, respectively. Cumene and ethylbenzene are produced by benzene alkylation with propylene and ethylene. Benzene alkylation with ethylene

Full text of DOI:10.1006/jcat.2000.2849

Journal of Catalysis 192, 163–173 (2000)
doi:10.1006/jcat.2000.2849, available online at http://www.idealibrary.com on
Alkylation of Benzene with Short-Chain Olefins over MCM-22 Zeolite:
Catalytic Behaviour and Kinetic Mechanism
A. Corma,�,1
�
V. Mart ı´ nez-Soria,† and E. Schnoeveld
�
Instituto de Tecnolog ´ı a Qu ´ı mica, CSIC-UPV, 46022 Valencia, Spain; and †Departament d’Engenyeria Quimica,
Universitat de Valencia, 46100, Burjassot, Spain
Received October 1, 1999; revised December 13, 1999; accepted February 4, 2000
chalice or cup type structures, with 12-ring openings and an
Benzene alkylation with ethene and propene has been carried out approximated depth of 7
A˚ (12) (see Fig. 1).
In this paper we discuss the catalytic behaviour of MCM-
2 for the benzene alkylation with short-chain olefins
under liquid-phase reaction conditions over zeolites MCM-22, Beta,
and ZSM-5. MCM-22 seems to be a good catalyst for benzene alky-
lation especially with propene, showing high activity and stability
and good selectivity. Kinetic experiments show that alkylation with
2
(
ethene and propene) and compare it with those of Beta
and ZSM-5. A mechanistic kinetic model has been devel-
oped for the alkylation of benzene with propene in order
to achieve not only a kinetic expression useful for reactor
design, but also a better understanding of the molecular
events occurring during alkylation on zeolite MCM-22.
propene follows an Eley–Rideal type mechanism.
�c 2000 Academic
Press
Key Words: benzene alkylation with MCM-22 zeolite; benzene
alkylation on Beta zeolite; cumene and ethylbenzene production on
zeolites.
EXPERIMENTAL
INTRODUCTION
Materials
Cumene and ethylbenzene are important compounds in
the petrochemical industry for the production of phenol sised using hexamethylenimine (HM) as template follow-
and styrene, respectively (1). Cumene and ethylbenzene ing the procedure reported in Ref. (13), and starting from
MCM-22 samples with Si/Al = 15 and 50 were synthe-
gels having the following compositions:
are primarily produced by benzene alkylation with propene
and ethene, and in the commercial processes the reactions
have conventionally been catalysed by mineral acids (e.g.,
solid phosphoric acid) and Friedel–Crafts systems (2) (e.g.,
AlCl3). However, economic, engineering, and environmen-
tal factors have prompted the development of new tech-
nologies in which solid acids, such as zeolite-based cata-
lysts, are used to catalyse the direct alkylation of benzene
with propene or ethene (3). In this way, several commer-
cial processes have been developed in the past few years
for the production of cumene and ethylbenzene based on
zeolite catalysts (4–7). The most recent one claims the use
of MCM-22 or SSZ-25 as alkylating catalysts (8–10).
Si/Al = 15 : 0.5 HM : 44.9 H2O : 0.18 Na : 0.11 OH : 0.033 Al2O3 : SiO2
Si/Al = 50 : 0.5 HM : 44.9 H2O : 0.18 Na : 0.16 OH : 0.010 Al2O3 : SiO2
Commercial samples of zeolite Beta (CP811, Si/Al = 13)
and zeolite ZSM-5 (CBV3020, Si/Al = 18) were used as
supplied by P.Q. Corp. The nature and purity of the as-
synthesized MCM-22 sample were checked by X-ray pow-
der diffraction (Philips 1060 diffractometer). The acidity
was determined by infrared spectroscopy with adsorption
of pyridine and desorption at different temperatures in a
Nicolet 710 FTIR apparatus following the experimental
procedure described in Ref. (14). Quantitative determina-
tion of the amount of Brønsted and Lewis acid sites was
derived from the intensities of the IR bands at ca. 1450
MCM-22 and SSZ-25 are isostructural zeolites with the
IZA code MWW. The structure is formed by two indepen-
dent system of pores (11). One of the pore systems is si-
nusoidal with 10 MR opening and the second is defined by
large 12 MR cavities (7.1 A˚ diameter, 18.2 A˚ height) con-
� 1
and 1545 cm , respectively, and using the extinction molar
coefficients given by Emeis (15). Textural properties of ze-
olites were determined from the N2 adsorption–desorption
experiments at 77 K on a Micromeritics ASAP-200 appa-
ratus. The main characteristics of the zeolites are given in
Table 1.
nected by a single 10 MR window. These cavities are opened
to the exterior at the termination of crystallites forming
1
In order to determine the influence of the external sur-
face acidity on catalytic isopropylation of benzene over
To whom correspondence should be addressed. E-mail: acorma@itq.
upv.es. Fax: 346-3877809.
163
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

´
CORMA, MARTINEZ-SORIA, AND SCHNOEVELD
1
64
charged with the chosen amount of catalyst always within
a particle size of 0.25–0.42 mm diameter, diluted with
carborundum (SiC/zeolite weight ratio = 4). The reaction
conditions for catalytic test were the following: pressure,
�
� 1
3
.5 MPa; temperature, 160–240 C; WHSV, 1.5–640 h (re-
ferred to olefin); benzene/olefin mole ratio, 3.0–37. Under
these conditions the reaction takes place in the liquid phase.
Benzene was supplied by Prolabo (A.R.). Ethene (pu-
rity >99.8%) and propene (purity >99.8%) were supplied
by Abello Linde S.A. All these reagents were used without
further manipulations. Preliminary alkylation experiments
were carried out to ascertain that under these conditions
the process was not controlled by mass transfer limitations
(
external and intraparticle diffusion). A small flow of nitro-
gen (usually <10 ml/min STP) was fed in order to improve
FIG. 1. MCM-22 structure, large cavities opened to the exterior at the the pressure control.
termination of crystal (chalices).
Before the alkylation reaction was started, the previously
�
�
calcined (500 C) catalyst was treated at 160 C and atmo-
spheric pressure in 100 ml/min flow of N2 for 150 min. Then,
MCM-22, a sample of MCM-22 (Si/Al = 15) was treated
with 2,6-di-tert-butylpyridine (DTBPy) following the ad-
sorption procedure described in Ref. 16. The presence of
adsorbedDTBPyonthetreatedzeolitesamplewaschecked
by means of IR analysis. Previous results show that DTBPy
cannot enter in the channels of MCM-22 (16), since the
molecular size of DTBPy (6.3 � 8.0 A˚ ) is too large for the
�
the catalytic bed was cooled below 60 C, and benzene was
fed at 2000 �l/min, increasing the pressure and tempera-
ture to the desired reaction conditions. When the desired
pressure(3.5MPa)andtemperaturewerereached, benzene
flow was adjusted to that desired for each particular exper-
iment, and then the olefin was fed. Propene was pumped as
a liquefied gas (1.5 MPa), while ethene was fed as a gas. The
liquid effluent samples were collected periodically in a cold
trap and analysed offline in a Varian 3350 GC equipped
with an FID detector and a Supelcowax capillary column.
molecule diffuse through 10 MR windows and channels. On
this basis DTBPy was selected to passivate the external sur-
face of MCM-22 without poisoning the internal acid site of
this zeolite.
(
60 m long and 0.20 mm o.d. with a 0.20 �m thick film of
stationary phase).
The MCM-22 (Si/Al = 15) sample with adsorbed DTBPy
Catalytic Experiments
Alkylation of benzene with propene or ethene was was also tested in the alkylation of benzene with propene,
performed in a computer-controlled down-flow fixed-bed in order to compare the results obtained with the same
stainless steel tubular microreactor, 4.1 mm i.d, 172 mm sample without surface deactivation. Moreover, MCM-22
long, equipped with a 1.6 mm o.d. axial thermowell and (Si/Al = 15) samples with and without DTBPy were tested
heated by a two-zone electrical furnace. The reactor was for the catalytic isomerization of 1-hexene, in order to make
TABLE 1
Acidic and Textural Properties of the Catalysts
a
Acidity, �mol of pyridine/g
Textural propertiesc
Brønsted
Lewis
Si/Al
ratio
BET area,
Micropore area,
Pore volume,
Micropore volume,
Crystal size,d
�
b
�
b
�
b
�
b
2
2
3
3
250 C
350 C
250 C
350 C
m /g
m /g
cm /g
cm /g
�m
e
MCM-22
MCM-22
Beta
50
15
13
15
25
45
36
47
15
24
17
28
13
19
47
12
11
16
40
7
461
453
509
380
355
325
295
228
0.52
0.61
0.67
0.29
0.17
0.15
0.13
0.12
0.3–0.5
e
0.3–0.5
0.1–0.2
1.5–2.5
ZSM-5
a
Calculated using the extinction molar coefficients given by Emeis (1995).
Desorption temperature.
Textural properties determined using N2 adsorption–desorption isotherms.
Crystal size estimated from the SEM pictures.
b
c
d
e
Hexagonal plates morphology. Average diameter.

BENZENE ALKYLATION WITH OLEFINS OVER MCM-22
165
sure that the acid sites within the micropores were not de- zeolite, which is usually related to the presence of highly dis-
activated by the DTBPy adsorbed. The catalytic isomer- persed extraframework aluminium (EFAL) species formed
ization of 1-hexene was carried out in the gas phase under during thermal treatment (25), supports this assumption.
�
atmospheric pressure at 180 C using the reactor system de-
The differences in textural properties, i.e., lower micro-
scribed above. 1-Hexene, supplied by Fluka (purity >98%), pore volume and larger mesopore volume in Beta, cannot
�
1
was fed (2.5 ml h ) together with N2 (N2/hexene mole be explained only by taking into account zeolite structure.
ratio = 30). The catalyst particle size was 0.25–0.42 mm The lower microporosity and higher mesoporosity of Beta
mesh and the products were analysed by online GC.
zeolite can be attributed to the agglomeration of very small
crystallite size as well as a partial destruction of the zeolite
lattice, which may occur during dealumination.
RESULTS AND DISCUSSION
The catalytic activity of MCM-22 and Beta zeolites
(
Si/Al = 15) for the alkylation of benzene with propene was
Alkylation of Benzene with Propene
measured at different WHSV, temperature, and benzene-
Electrophilic alkylation of aromatics by olefins is com- to-propene mole ratios (Fig. 2). It can be seen that even
monly considered as proceeding via a carbenium ion if both zeolites show high activity, at least at relatively
type mechanism. In this way, propene is protonated by a short time on stream (TOS), Beta zeolite appears to have
Brønsted acid site and a carbenium ion is formed. Finally a slightly higher initial activity. Since alkylation reactions
this ion is attacked by a free or weakly adsorbed benzene are catalysed by the acid sites of zeolites (26), we could
or isopropylbenzene (17, 18) molecule, producing cumene expect that the higher the concentration of Brønsted acid
or diisopropylbenzene (DIPB), respectively. However, sites in a zeolite the higher the alkylation activity should
the carbenium ion formed can also react with another be. However, in this case, although MCM-22 presents a
propene molecule, producing oligomers, which are respon- higherconcentrationofacidsites, itsactivityisslightlylower
sible for the deactivation of the catalyst. n-Propylbenzene (Fig. 2a). This result can be explained by taking into account
is produced in a minor amount in the reaction by secondary that in the case of MCM-22 only the sites pointing to the
isomerization of cumene, which would occur via bimolec- external surface, i.e., at the top and bottom of the crystal-
ular transalkylation between cumene and benzene (19), lites (see Fig. 1), could allow the cumene to be formed and
and/or a primary product by direct alkylation of benzene to diffuse out without great difficulty. Indeed, it has been
with propene (20). The formation of n-propylbenzene is un- reported (24) that the diffusion of cumene in the pores of
desired since the isomer is difficult and costly to separate MCM-22 is hampered by a high energy barrier. Moreover,
from the cumene stream. Cumene can also suffer a con- it has been presented (27) by means of molecular dynamic
secutive alkylation reaction, yielding diisopropylbenzenes simulation that even benzene presents a low diffusivity in
(
DIPBs). However, the DIPBs are not considered as waste either of the two pore systems of the MWW structure. On
products since they can be recovered by transalkylation the other hand, the framework of Beta zeolite, with 12-
with benzene (21). membered ring channels, presents lower steric hindrance
It has been shown that Beta zeolite performs well in and both benzene and cumene can diffuse without great dif-
liquid-phase alkylation with both ethene and propene (22) ficulty. Nevertheless, if the crystal size is too large, cumene
and excellent efficiencies in benzene alkylation have been diffusion can control, at least partially, the rate of the global
reported (23). Similar results have been claimed for MCM- process even in zeolites with 12 MR pores like Beta (23).
2
2 zeolite in patent literature (8, 9) but there is little infor-
Under some reaction conditions, it can be seen (Fig. 2)
mation available in the open literature. A recent paper (24) that Beta deactivates faster than MCM-22, and this deac-
compares catalytic activity of MWW with Beta and other tivation is faster if the B/P ratio decreases from 7.2 to 3.5
zeolites. Beta zeolite appears to be one of the most effi- (Fig. 2d). Zeolite MCM-22 does not show deactivation, at
cient catalysts and MCM-22 also shows a surprisingly good least under the reaction conditions studied here. The de-
behaviour on the basis of a single experiment.
activation in Beta zeolite can be related to the formation
In our case, when we compare the textural and acidic of propene oligomers in the channels. These species can
properties of Beta and MCM-22 (Si/Al = 15) zeolites, it is remain strongly adsorbed and can be responsible for the
possible to see (Table 1) that the Beta zeolite used here has observed deactivation. Since it was claimed that only the
a much lower Brønsted acidity, in terms of both number and “external” acid sites are active in the case of MCM-22,
strength, than the MCM-22 sample, although their nominal the formation of oligomers in the 10 MR cavities should
Si/Al ratio is very similar. The differences cannot be ex- not affect the activity. Similar results have been obtained
plainedbyconsideringonlythecrystallinityandstructureof on SSZ-25 (10), which is isostructural with MCM-22, and
the samples, but it has to be assumed that a higher dealumi- it was concluded that MWW zeolite appears to be clearly
nation took place with the Beta sample during calcination. superior to Beta in terms of both activity, and especially
The presence of a higher amount of Lewis acid sites in Beta stability, when alkylating benzene with propene in a light

´
CORMA, MARTINEZ-SORIA, AND SCHNOEVELD
1
66
FIG. 2. Propene conversion with Beta (᭹) and MCM-22 (᭿) zeolites under the following reaction conditions: total pressure, 3.5 MPa;
�
�
�
(
a) temp, 180 C; WHSVpropene = 2.6, B/P mole ratio, 7.2; (b) temp, 180 C, WHSVpropene = 5.6, B/P mole ratio, 7.2; (c) temp, 220 C, WHSVpropene = 5.6,
�
B/P mole ratio, 7.2; (d) temp, 200 C, WHSVpropene = 11.5, B/P mole ratio, 3.5.
reformate feedstock, but unfortunatelly a clear explanation ternal acid sites of this zeolite (16), while leaving untouched
for the observation was not given.
the acid sites within the 10 MR micropores that cannot be
In order to explain the performance of MCM-22 we reached from the external surface. In agreement with pre-
should assume that a significant number of acid sites are vious work (16), the IR results show that DTBPy was in-
�
1
placed in the large “half cavities” or “chalices” opened to deed adsorbed with IR bands at 3370, 1616, and 1530 cm
+
the exterior at the termination of the crystallites (Fig. 1). which are characteristic of the DBTPyH ion. It should
In these cavities the diffusion of the products, as well as be taken into account that the external acid sites that can
the diffusion of the coke precursors, is easier and faster and be reached by DTBPy are only a rather small fraction of
therefore the deactivation is slower.
the total acid sites presented in MCM-22 (16). Then, when
In order to prove the role of the external acid sites on the sample containing adsorbed DTBPy was used to catal-
the catalytic performance of MCM-22, the sample with yse the alkylation of benzene with propene, it can be seen
Si/Alratioof15 wasdeactivatedby 2,6-di-tert-butylpyridine (Fig. 3a) that a very large decrease in activity has been pro-
(DTBPy) which has been shown to adsorb only on the ex- duced due to the neutralisation of the most external acid
FIG. 3. Catalytic activity of MCM-22 (Si/Al = 15) without (᭹) and with (᭺) poisoning with DBTPy. (a) Alkylation of benzene with propene under
�
the following reaction conditions: total pressure, 3.5 MPa, temp, 180 C, WHSVpropene = 6.2, B/P mole ratio, 7.0. (b) Isomerization of 1-hexene under
�
the following reaction conditions: temp, 180 C, WHSV1-hexene = 6.6, N2/1-hexene mole ratio, 30.

BENZENE ALKYLATION WITH OLEFINS OVER MCM-22
167
sites by the bulky DTBPy. On the other hand, and in or-
Although the selectivities on Beta and MCM-22 are quite
der to show that the internal acid sites were still present, similar, and no great differences are observed, it has to be
the DTBPy deactivated sample was used to catalyse the pointed out that MCM-22 zeolite appears to be slightly less
isomerization of 1-hexene. This reactant was chosen be- selective to the most undesirable products, oligomers and
�
cause of the easy diffusion through the 10 MR windows, and n-propylbenzene. At the lowest temperature, 180 C, MCM-
also because the double bond isomerization requires low 22 is more selective to the formation of diisopropylben-
�
temperatures within the range of those used for the alky- zenes, while at the highest temperature tested, 220 C, Beta
lation of benzene. During the isomerization of 1-hexene, presents a higher diisopropylbenzene selectivity. This fact
the only products detected were (cis + trans)-2-hexene and could be related with differences in activation energy for
(
cis + trans)-3-hexene, inagreementwiththereportedwork the transalkylation reaction between the two zeolites. On
ontheisomerizationof1-hexeneonZSM-5(41). Theresults the other hand, despite the diisopropylbenzene selectivity
presented in Fig. 3b show that the fresh MCM-22 sample being similar for both zeolites, there are some differences
hasonlyaslightlyhigheractivityfor1-hexeneisomerization in the distributions of the different isomers as can be seen in
than the sample in which the external acid sites were de- Table 2, where is shown that the para-to-meta DIPB ratio
activated by DTBPy. These results are in agreement with is higher with MCM-22.
the previous work (16) that shows that less than 10% of
the total acid sites present in an MCM-22 sample can be
reached from the external surface by using DTBPy. If this
Alkylation of Benzene with Ethene
is the case, it has to be concluded that the alkylation of ben-
In Fig. 4, ethene conversion versus TOS for MCM-22,
zene by propene with MCM-22 takes place predominantly Beta, and ZSM-5 catalysts has been plotted. Since ethene
on external acid sites and/or those closer to the external is less reactive than propene, it was necessary to increase
surface of the crystallites.
severity, i.e., temperature and/or contact time, in order to
The selectivity to the different products, as well as the obtain conversion values similar to the ones obtained dur-
distribution of the diisopropyl isomers obtained under dif- ing propene alkylation. For the TOS considered here, de-
ferent reaction conditions, is given in Table 2. The se- activation is not observed with Beta and MCM-22 (Si/Al =
lectivity to cumene is quite high for both zeolites, about 15), indicating that oligomerization is much lower for
9
0% referred to propene (>95% referred to benzene). The ethene than for propene.
main secondary products observed are diisopropyl ben-
ZSM-5 zeolite appears to be a poor catalyst for the alky-
zene isomers, n-propylbenzene, and propene dimerization– lation of benzene with ethene in the liquid phase, and deac-
polymerization products, mainly hexenes (oligomers in tivates fast owing to the formation of oligomers. However,
Table 2). Often a negligible amount of other compounds, when the reaction is carried out in the gas phase ZSM-5 is
mainly alkyl aromatics, is observed. The selectivity to oneofthebestcatalystsforethylbenzeneproduction(7, 29).
cumene and to oligomers decreases with increasing con-
Comparing Beta and MCM-22 zeolites we can conclude
version and temperature, while the opposite occurs with in general terms that the results are quite similar to those
respect to the selectivity of diisopropylbenzene and n- obtained in alkylation with propene; i.e., Beta is more active
propylbenzene, in good agreement with previous work (28). than MCM-22 for ethylbenzene production.
TABLE 2
Selectivity and Diisopropylbenzene Distribution Products at Different Temperature and Propene Conversions
b
c
Selectivity, %
DIPBs distribution (%)
Temperature,
Xpropene,a
Iso/n-propyl
benzene ratio
�
DIPBsb
Catalyst
C
%
Cumene
Oligomers
ortho
meta
para
MCM-22
180
20
180
20
76.05
92.12
90.56
90.78
89.54
7.34
9.03
8.84
9.60
0.32
0.27
0.18
0.11
1650
830
790
460
10
8
7
30
32
33
38
60
60
60
57
9
7.97
91.70
6.28
76.25
7.34
89.90
8.34
2
9
5
Beta
92.16
90.76
89.34
88.67
6.96
8.33
10.07
10.58
0.41
0.25
0.21
0.15
920
900
720
460
6
5
5
3
42
44
46
51
52
51
49
46
9
2
9
Note. Benzene alkylation with propene with MCM-22 and Beta zeolites. Reaction conditions: total pressure, 3.5 MPa; B/P mole ratio, 7.2. Catalyst
with Si/Al ratio about 15.
a
The different propene conversions were achieved by changing the WHSV.
Selectivity referred to propene.
DIPBs: diisopropylbenzene isomers (ortho + meta + para).
b
c

´
CORMA, MARTINEZ-SORIA, AND SCHNOEVELD
1
68
TABLE 3
Product Selectivity in Benzene Alkylation with Ethene
a
Selectivity, %
Catalyst,
Si/Al � 15
Xethene,
%
EBa
DEBsb
ButylBc
TEBsd
MCM-22
Beta
ZSM-5
85.70
87.15
31.30
88.30
88.9
42.8
11.05
9.95
3.48
0.40
0.57
51.31
0.005
0.08
0.00
�
Note. Reaction conditions: 240 C, total pressure, 3.5 MPa, benzene/
ethene mole ratio, 9.
a
Selectivity referred to propene.
b
DEBs: diethylbenzene isomers (ortho + meta + para).
c
ButylB: butylbenzene.
d
TEBs: triethylbenzene isomers.
with propene (Fig. 5a) and ethene (Fig. 5b). As we expected,
according to the concentration of acid sites (Table 1), a
decrease in catalytic activity is observed when the amount
of framework aluminum is decreased.
Tables 4 and 5 report the selectivities obtained in the
experiments of alkylation of benzene with propene and
ethene, respectively. Selectivity is expressed as mole ratio
of monoalkylated cumene or ethylbenzene versus the dif-
ferent products. As can be seen the ratio of monoalkylated
to dialkylated products is very similar for the two MCM-22
samples, indicating that at the levels of Si/Al ratio stud-
ied here the zeolite framework composition does not af-
fect much the selectivity between mono- and dialkylated
products. Nevertheless, the highest Si/Al ratio produces
more oligomers (C6 olefin mainly) in the alkylation with
propene (Table 4) and more butylbenzene in the alkylation
with ethene (Table 5). These trends suggest that the low
aluminum content favours the oligomerization versus the
alkylation reaction and this together with the lower concen-
tration of acid sites on the sample with lower Al content
causes this sample to deactivate faster, as can be seen in
Fig. 6. Similar results were observed for Beta zeolite (23).
Finally, it is worth noting that the MCM-22 sample with
FIG. 4. Ethene conversion of the different zeolites. Reaction condi-
�
tions: total pressure, 3.5 MPa; (a) temp, 240 C, WHSVethene = 1.2, B/E
�
mole ratio, 8; (b) temp, 200 C, WHSVethene = 6.5, B/E mole ratio, 3.0.
Beta (᭹), MCM-22 (᭿), and ZSM-5 (᭡).
The selectivity to the different products is shown in
Table 3. MCM-22 and Beta zeolites present high ethylben-
zene selectivity (� 90% referred to ethene and >95% re-
ferred to benzene), while ZSM-5 presents a high proportion
of butylbenzene arising from the alkylation of benzene with
butene, formed by ethene dimerization. The high amount
of butylbenzene formed indicates that oligomerization has
proceeded quite extensively and this is responsible for the
fast deactivation observed with ZSM-5.
Influence of MCM-22 Si/Al Ratio on Alkylation
of Benzene with Propene and Ethene
Figure 5 shows the catalytic activity of MCM-22 samples
with Si/Al ratios of 15 and 50 for the alkylation of benzene the highest Si/Al mole ratio gives the highest amount of
FIG. 5. Olefin conversion of the MCM-22 samples, Si/Al = 50 (᭹) and Si/Al = 15 (᭿), as a function of contact time. (a) Propene conversion at
� �
.5 MPa, 220 C, and B/P mole ratio 9. (b) Ethene conversion at 3.5 MPa, 240 C, and B/E mole ratio 8.
3

BENZENE ALKYLATION WITH OLEFINS OVER MCM-22
169
TABLE 4
Influence of MCM-22 Si/Al Ratio on Product Selectivity in Benzene Alkylation with Propene
Temperature,
B/P
mole ratio
Conversion,
%
Si/Al
ratio
iPB/DIPBsa
mole ratio
iPB/nPBa
mole ratio
iPB/oligomers
mole ratio
�
C
2
00
20
4
7
� 80
� 90
15
6.0
6.1
850
620
370
240
50
2
15
10.1
10.3
790
450
500
340
50
Note. Total pressure, 3.5 MPa.
a
iPB, isopropylbenzene; DIPBs; diisopropylbenzenes; nPB, n-propylbenzene.
n-propylbenzene in the alkylation of benzene with propene
In order to determine the kinetic mechanism for the
(Table 4). An explanation for these results can be found alkylation of benzene with propene on MCM-22, we have
in the differences of acid site strength distributions in the carried out a kinetic study by working with a differen-
two MCM-22 samples (Table 1). Although the sample with tial reactor and calculating initial rates at different ben-
Si/Al = 15 has more acid sites, the proportion of acid sites zene/propene (B/P) ratios. For all these experiments the
�
�
with medium to high strength (350 C versus 250 C pyri- sample used was the one with an Si/Al ratio of 15. How-
dine desorption temperature in Table 1) is higher for the ever, since low conversion favours high olefin concentra-
sample with the lower amount of aluminum. The higher tion and consequently a faster catalyst deactivation, we
the strength of the acid sites, the longer the average life- have measured conversion at different TOS, and from this
time of the adsorbed species will be and consequently the conversion at TOS = 0 could be determined (Fig. 7).
secondary reactions, such as the isomerization of cumene The experiments were performed at four different temper-
to n-propylbenzene, will be favoured.
atures, varying WHSV and benzene/propene mole ratio.
The propene conversion curves as a function of contact
Kinetics of the Alkylation of Benzene with Propene
time (1/WHSV) with different benzene/propene ratios at
�
2
20 C reaction temperature are shown in Fig. 8. Similar
There has been some controversy over the kinetic mech-
anism which can better explain the alkylation of aromat-
ics with light olefins over acidic zeolites. While Venuto
et al. (26) and Weitkamp (30) suggested an Eley–Rideal
curves were obtained at the other temperatures studied.
As the experiments were carried out with an excess of ben-
zene and, usually, at low conversions we can make the ap-
proximation that the benzene concentration (Cb) remains
practically constant.
Then if one considers the LH model and assumes that the
controlling step is the propene adsorption or the chemical
reaction between the adsorbed propene and benzene the
(ER) type mechanism for the alkylation of benzene with
ethene over faujasite and ZSM-5, other authors have con-
cluded that alkylation of benzene with short olefins follows
a Langmuir–Hinselwood (LH) model (31, 32). More re-
cently, it has been reported that the size of the pores of the
zeolite can determine the alkylation mechanism (33).
TABLE 5
Influence of MCM-22 Si/Al Ratio in Product Selectivity
on Benzene Alkylation with Ethene
Temperature,
Conversion,
%
Si/Al
ratio
EB/DEBa
mole ratio
EB/BBa
mole ratio
�
C
2
40
� 70
15
11
13
15
16
475
200
375
110
5
0
�
60
15
5
0
2
20
� 60
40
15
17
18
19
22
330
95
300
75
5
0
�
15
5
0
FIG. 6. Propene conversion of the MCM-22 samples as a function of
time on stream, Si/Al = 50 (᭹) and Si/Al = 15 (᭿), under the following
Note. Reaction conditions: total pressure, 3.5 MPa, benzene/ethene
�
mole ratio, 8.
reaction conditions: 3.5 MPa, 220 C, B/P mole ratio 9, and WHSVpropene =
a
EB, ethylbenzene; DEB; diethylbenzene; BB, butylbenzene.
1.5 and 5 for Si/Al = 50 and 15, respectively.

´
CORMA, MARTINEZ-SORIA, AND SCHNOEVELD
1
70
On the other hand, if one considers that the reaction
follows an Eley–Rideal (ER) kinetic model either with or
without competitive adsorption of benzene, the following
result are found:
1
) Without competitive adsorption of benzene,
0
k KaCa
o
ra =
.
[3]
[4]
(
1 + KaCa)
2
) With competitive adsorption of benzene,
0
k KaCa
o
ra =
FIG. 7. Alkylation of benzene with propene over MCM-22 at low
(1 + K C + K C )
a
a
b
b
propene conversion. Reaction conditions: total pressure, 3.5 MPa, temp,
�
where constants have the same meaning as above.
2
20 C, WHSVpropene = 478, B/P mole ratio, 10.
By linearizing Eqs. [3] and [4] we arrive at the following
equation,
rate of alkylation is given (34, 35) by the following equation,
0
k KaCa
1
o
ra =
,
[1]
= A4 R + A5,
[5]
(
1 + KaCa + KbCb)2
ra
0
where k = ko if propene adsorption is the controlling step, where
o
0
or k = koCb if the chemical reaction is the controlling step.
o
1
Kb
0
k Ka
o
Ca and Cb are the concentrations of propene and benzene,
respectively.
A4 =
A4 =
A5 =
+
if there is benzene competitive
adsorption and
0
k KaCb
o
If Eq. [1] is linearized it becomes
1
without benzene competitive
adsorption.
0
k KaCb
o
1
A2
=
A1 R +
+ A3,
[2]
ra
R
1
0
k
o
where R = Cb/Ca is the benzene/propene concentration
ratio, and
It is well known that short olefins adsorb fast and easily.
Hence, what determines the kinetic behaviour is the state of
benzene when it reacts with the adsorbed alkylating agent.
Then if benzene reacts in an adsorbed form, the LH mech-
anism applies and this corresponds with Eq. [2]. On the
other hand, if benzene does not react in an adsorbed form
the process follows an ER kinetic expression and Eq. [5]
applies.
1
Kb
2
A1 =
A2 =
A3 =
+
+
0
0
2
0
k Ka KbC
k Ka
k KaCb
o
o
b
o
Ka
0
k Kb
o
2
2
+
.
0
0
k KbCb
o
k
o
Taking into account the continuity equation for a fixed-
bed plug flow reactor (34),
xa=0
Z
W
dxa
ra
=
,
[6]
Fao
xa=xa
and that at low conversion
W
xai
�
,
[7]
[8]
Fao
rai
the following results:
1
W
=
.
rai
Fa xai
o
�
By working with a differential reactor we have obtained
the initial reaction rates from the tangent in the origin
FIG. 8. Initial propene conversion at 220 C over MCM-22 as a func-
tion of contact time for different B/P mole ratios.

BENZENE ALKYLATION WITH OLEFINS OVER MCM-22
171
TABLE 6
TABLE 7
�
1
� 1
Initial Rate Values (ro, mol h gcat) Obtained in Alkylation of
Benzene with Propene over MCM-22 (Si/Al = 15) at Different Re-
action Temperatures and Benzene/Propene Mole Ratios
Kinetic and Adsorption Constants and Standard Deviation
for the Different Mechanisms
�
Temperature, C
Reaction Temperature
B/P mole
Constantsa
Mechanism
RE
160
180
200
220
�
�
�
�
ratio
160 C
180 C
200 C
220 C
� 1
ko (h
)
14.4
36.7
0.036
31.3
15.0
0.053
97.9
3.14
0.071
166.7
KiCb
2.86
0.049
177.6
1084.3
369.2
�
5
7
0
0
7
20
20
18
17
15
38
32
28
22
16
68
53
40
26
17
75
65
50
33
18
�
�
LH
ko (h 1)
KaCb
KbCb
�
85.4
14.1
3.9
105.1
78.8
25.5
130.6
75.3
28.2
1
2
3
0.050
0.067
0.086
0.062
p
a
2
�
=
(Xexperimental � Xestimated) /(N � P), where N is the number
�
on the curves given in Fig. 8 and the values are given in
Table 6. If the benzene alkylation reaction follows an ER
kinetic mechanism then the inverse of the initial reaction
rate should vary lineally with the benzene/propene ratio (R
parameter). OntheotherhandiftheLHmechanismapplies
a convex curve should be obtained when the inverse of the
initial reaction rate versus R is plotted. As can be seen in
Fig. 9, the results show a good lineal correlation between
of points and p is the number of model parameters. Cb, benzene concen-
tration.
criteria, one can see that the evolution of the adsorption
constants with the temperature also supports the ER model
as more suitable. Indeed, the results from Table 7 clearly
show that the adsorption constant values obtained from
the LH and ER models increase and decrease respectively
when the reaction temperature increases. Since adsorption
is, generally, anexothermicprocess(36), takingintoaccount
the thermodynamic criteria (37), we should also favour the
ER versus the LH mechanism for alkylation of benzene
with propene on MCM-22 zeolite. Unfortunately, with the
experimental data available we cannot determine if a com-
petitive adsorption of benzene does or does not take place.
However, taking into account the working temperatures
and reported adsorption experiments (17, 38), it should be
possible to say that benzene does indeed adsorb on acid
sites, but the adsorption constant for benzene is much lower
than the corresponding value for propene. Therefore we
1/ra and R, under the reaction conditions studied here,
favouring therefore a kinetic mechanism based on an ER
model. It has to be remarked that the reaction has not been
studied at lower B/P ratios because (1) olefin oligomeriza-
tion will dominate, (2) the reaction will not occur in the
liquid phase, and (3) the assumptions made to develop the
kinetic expression would not be valid anymore.
The parameter values of Eqs. [1] and [3], which are rep-
resentatives of the two different mechanisms, were deter-
mined by a nonlinear optimisation method (Marquardt),
integrating Eq. [6], and are given in Table 7. The quality of
the fit is given by the standard deviation (�), which is sub-
stantially lower for the ER mechanism expression for all the
temperatures studied here. We can see this more clearly in
Fig. 10 where it becomes apparent that ER fits better the
experimental results than the LH model. Besides the fitting
FIG. 9. Variation of initial rate with B/P ratio at different tempera-
tures: (᭿) 160 C, (᭹) 180 C, (᭡) 200 C, (᭢) 220 C.
FIG. 10. Experimental (᭿, ᭹) and calculated (lines) propene conver-
sion at 220 C over MCM-22. (���) ER model and (—) LH model.
�
�
�
�
�

´
CORMA, MARTINEZ-SORIA, AND SCHNOEVELD
1
72
4
5
. Chen, J., Worldwide Solid Acid Process Conference, Session 3, Cata-
lyst Consultant Inc., Houston, TX, 14–16 November 1993.
. Meima, G. R., van der Aalst, M. J. M., Samson, M. S. V., Vossenbeerg,
N. J., Adiansens, R. J. O., Garces, J. M., and Lee, J. G., Worldwide
Solid Acid Process Conference, Session 3, Catalyst Consultant Inc.,
Houston, TX, 14–16 November 1993.
6. Chem. Eng. News, 18 December, 12 (1995). Hydrocarbon Processes
October, 40 (1996).
7
8
. Chen, N. Y., and Garwood, W. E., Catal. Rev.-Sci. Eng. 28, 185 (1986).
. Kresge, C. T., Le, Q. N., Roth, W. J., and Thompson, R. T., U.S. Patent
5,259,565, 1993.
9. Chu, P., Landis, M. E., and Le, Q. N., U.S. Patent 5,334,795, 1994.
1
1
0. Holtermann, D. L., and Innes, R., U.S. Patent 5,149,894, 1992.
1. Leonowicz, M. E., Lawton, I. A., Lawton, S. L., and Rubin, M. K.,
Science 264, 1910 (1994). Beaumont, L. R., Vartuli, J. C., Roth, W. J.,
Leonowicz, M. E., Kresge, C. T., Schmitt, K. D., Chu, C. T.-W., Olson,
D. H., Sheppard, E. W., McCullen, S. B., Higgins, J. B., and Schenker,
J. A., J. Am. Chem. Soc. 114, 10834 (1992).
FIG. 11. Arrhenius plot for kinetic constant of the Eley–Rideal
mechanism.
1
2. Lawton, S. L., Leonowicz, M. E., Partridge, R. D., Chu, P., and Rubin,
M. K., Microporous Mesoporous Mater. 23(1–2), 109 (1998).
can conclude that while benzene is adsorbed, the product
KbCb is negligible compared to the product KaCa, and the 13. Corma, A., Corell, C., Mart ı´ nez, A., and P e´ rez-Pariente, J., Appl.
adsorption term of the expression rate, Eq. [4], is governed
mainly by the propene adsorption.
In Fig. 11 the Arrhenius plot for the kinetic rate constant
is given. As we can see the fit of the experimental values to
Catal. A 115, 121 (1994). Corma, A., Corell, C., and P e´ rez-Pariente, J.
Zeolites 15, 2 (1995).
4. Corma, A., Forn e´ s, V., Mart ´ı nez, A., and Orchill e´ s, A. V., ACS Symp.
Ser. 368, 542 (1988).
1
1
5. Emeis, C. A., J. Catal. 141, 347 (1993).
the Arrhenius equation is good, and an activation energy of 16. Corma, A., Forn e´ s, V., Forni, L., M a´ rquez, F., Mart ´ı nez-Triguero, J.,
�
1
� 1
7
7 kJ mol has been obtained. Values of 42 kJ mol over
and Mascotti, D., J. Catal. 179, 451 (1998).
17. Venuto, P. B., and Landis, P. S., Adv. Catal. 18, 259 (1968).
18. Coughlan, B., and Keane, M. A., J. Catal. 138, 164 (1992).
9. Derouane, E. G., He, H., Derouane-AbdHami, S. B., and Ivanova,
I. I., Catal. Lett. 58, 1 (1999). Wichterlov a´ , B., and Cejka, J., J. Catal.
146, 523 (1994).
�
1
an HM zeolite (32), 75 kJ mol over an Fe-FMI zeolite
39), and 50–67 kJ mol (40) for CaY and LaY zeolites
have been reported for benzene isopropylation to cumene.
The value obtained in the present work was close to the
highest value of activation energy reported.
�
1
(
1
20. Bentham, M. F., Gadja, G. J., Jensen, R. H., and Zinnen, H. A., in
Proceeding of the DGMK Conference, Catalysis on Solid Acids and
“
Bases, Berlin, Germany, March 14–15, 1996” (J. Weitkamp and B.
L u¨ cke, Eds.), p. 155, 1996.
CONCLUSION
2
2
1. Pradhan, A. R., and Rao, B. S., Appl. Catal. A 106, 143 (1993).
2. Cavani, F., Arrigoni, V., and Bellussi, G., Eur. Pat. Appl. 432,814
A1, 1991. Innes, R. A., Zones, S. I., and Nacamuli, G. J., U.S. Patent
4,891,458, 1990. Smith, L. A., Jr., PCT Int. Appl. WO 9809929 A1,
MCM-22 zeolite is a good catalyst to carry out ben-
zene alkylation with short olefins. Comparing with a com-
mercial Beta zeolite, MCM-22 shows similar activity and
selectivity but better stability. The alkylation of benzene
on MCM-22 takes place mainly on the external surface,
which in this material is perfectly structured and formed by
1
998. Hendriksen, D. E., Lattner, J. R., and Janssen, M. J. G., PCT Int.
Appl. WO9809928 A2, 1998. Gadja, G. J., and Gajek, R. T., U.S. Patent
,723,710, 1995.
5
2
3. Bellussi, G., Pazzuconi, G., Perego, C., Girotti, G., and Terzoni, G.,
J. Catal. 157, 227 (1995).
�
0.7 � 0.7 nm cups where active sites are located. It appears
then that the higher the amount of those chalices, i.e., the 24. Perego, C., Amarilli, S., Millinni, R., Bellussi, G., Girotti, G., and
Terzoni, G., Microporous Mater. 6, 395 (1996).
5. Corma, A., Forn e´ s, V., Mart ´ı nez, A., and Sanz, J., ACS Symp. Ser. 446,
higher the external surface area, the better should be the
catalyst for carrying out the alkylation of benzene by either
propene or ethene.
From a full kinetic study it was possible to conclude that
the alkylation of benzene with propene on MCM-22 follows
an Eley–Rideal type mechanism in which benzene can also
compete by the active sites but its coverage is much lower
than that of propene.
2
17 (1988).
2
2
6. Venuto, P. B., Hamilton, L. A., and Landis, P. S., J. Catal. 5, 484 (1966).
7. Sastre, G., Catlow, C. R. A., and Corma, A., J. Phys. Chem. B 103, 5187
(1999).
2
2
3
3
8. Panmig, J., Qingying, W., Chao, Z., and Yanhe, X., Appl. Catal. A 91,
125 (1992).
9. Itoh, H., Miyamoto, A., andMurakami, Y., J. Catal. 64, 284(1980). Itoh,
H., Hattori, T., Suzuki, K., and Murakami, Y., J. Catal. 79, 21 (1983).
0. Weitkamp, J. (Int. Symp. in Zeolite Catalysis, Siofok, Hungary), Acta
Phys. Chem. 217 (1985).
1. Morita, Y., Matsumoto, H., Kimura, T., Kato, F., and Takayasu, M.,
Bull. Jpn. Pet. Inst. 15(1), 37 (1973).
REFERENCES
1
2
3
. Pujado, P. R., Salazar, J. R., and Berger, C. V., Hydrocarbon Process
5
5, 91 (1976).
32. Becker, K. A., Karge, H. G., and Streubel, W.-D., J. Catal. 28, 403
. Bentham, M. F., Gajda, G. J., Jensen, R. H., and Zinnen, H. A., Erd o¨ l
Erdgas Kohle 113, 84 (1997).
. Corma, A., Chem. Rev. 95, 559 (1995).
(1973).
33. Panagiotis, G. S., and Ruckenstein, E., Ind. Eng. Chem. Res. 43, 1517
(1995).

BENZENE ALKYLATION WITH OLEFINS OVER MCM-22
173
3
4. Froment, G. F., and Bischoff, K. B., “Reactor Analysis and Design.”
Wiley, New York, 1974.
38. Flego, C., Kiricsi, I., Perego, C., and Bellussi, G., Stud. Surf. Sci. Catal.
94, 405 (1995).
3
3
5. Corrigan, T. E., Chem. Eng. January, 199 (1955).
6. Adamson, A. W., and Gast, A., “Physical Chemistry of Surfaces.”
Wiley, New York, 1997.
39. Parikh, P. A., Subrahmanyam, N., Bath, Y. S., and Halgeri, A. B., Can.
J. Chem. Eng. 71, 756 (1993).
40. Allakherdieva, D. T., Romanovskii, B. V., and Topchieva, V. M.,
Neftekhimiya 9, 373 (1968).
3
7. Rudzinski, W., and Everett, D. H., “Adsorption of Gases
on Heterogeneous Surfaces.” Academic Press, London, 41. Abbot, J., Corma, A., and Wojciechowski, B. W., J. Catal. 92, 398
992. (1985).
1