Benzene alkylation with propane over Pt-modified MFI zeolites

Article abstract of DOI:10.1006/jcat.2000.2910

Benzene alkylation with propane was studied on acidic H-MFI zeolites with SiO₂/Al₂O₃ ratios of 51, 104, and 325, mixed (H-MFI + Pt/CeO₂) catalytic systems, and Pt/H-MFI bifunctional catalysts with Pt content of 0.02 and 0.3%. Interaction of benzene with propane over zeolite catalyst proceeded via three main reaction pathways leading to propylbenzenes and dihydrogen, ethylbenzene and methane, and toluene and ethane. At > 573 K, propylbenzenes were very unstable and were converted into benzene and propylene or interacted with benzene to yield ethylbenzene and toluene. The increase in the number of strong Broensted sites in zeolite led to the increase of propane conversion and to the decrease of the selectivity to target products, which was due to the cracking of propane and propylbenzenes dealkylation and dismutation. Incorporation of platinum into the catalytic system enhanced propane dehydrogenation at the expense of its cracking and resulted in significant improvement of propane conversion and selectivity to propylbenzenes. Addition of an H₂ scavenger allowed a shift of the thermodynamic equilibria and an increase in propane and benzene conversion. The best result was observed on the (Pt(II)/H-MFI-300 + Zr₂Fe) catalytic system, on which the yield of propylbenzenes reached 96% with respect to equilibrium and selectivity to propylbenzenes was ~ 60%.

Full text of DOI:10.1006/jcat.2000.2910

Journal of Catalysis 194, 266–277 (2000)
doi:10.1006/jcat.2000.2910, available online at http://www.idealibrary.com on
Benzene Alkylation with Propane over Pt-Modified MFI Zeolites
Andrei V. Smirnov, Evgenii V. Mazin, Valentina V. Yuschenko, Elena E. Knyazeva, Sergei N.
Nesterenko, Irina I. Ivanova, ,1 Leonid Galperin,† Robert Jensen,† and Steven Bradley†
Laboratory of Kinetics and Catalysis, Department of Chemistry, Moscow State University, Vorob’evy Gory, 117234 Moscow, Russia;
and †UOP LLC, 25 East Algonquin Road, P.O. Box 5017, Des Plaines, Illinois 60017-5017
Received January 14, 2000; revised April 17, 2000; accepted April 17, 2000
reactions with other organic molecules, for example, in the
alkylation of aromatic hydrocarbons.
Benzene alkylation with propane has been studied on acidic H-
MFI zeolites with SiO2/Al2O3 ratios of 51, 104, and 325, mixed
(H-MFI + Pt/CeO2) catalytic systems, and Pt/H-MFI bifunctional
catalysts with Pt content of 0.02 and 0.3%. In some of the ex-
periments, Zr2Fe intermetallic compound was added to the cata-
lytic system to ensure H2 removal from the reaction zone. The
reaction was carried out at atmospheric pressure, in the tem-
perature range of 573–773 K, C3H8/C6H6 = 1, and WHSV =
0.2–7 h 1. The conversion of propane and selectivity to propyl-
benzenes was found to increase in the following order of
catalytic systems: H-MFI < (H-MFI + Pt/CeO2) < Pt/H-MFI < (Pt/
H-MFI + Zr2Fe). The increase of the number of strong Brønsted
sites in zeolite led, on one hand, to the increase of propane con-
version, but on the other hand, to the decrease of the selectivity
to target products, which was due to the cracking of propane and
propylbenzenes dealkylation and dismutation. Incorporation of Pt
into the catalytic system enhanced propane dehydrogenation at the
expense of its cracking and resulted in significant improvement of
propane conversion and selectivity to propylbenzenes. Addition of
an H2 scavenger allowed a shift of the thermodynamic equilibria
and an increase in propane and benzene conversion. The best result
was obtained on the (Pt/H-MFI + Zr2Fe) catalytic system with an
SiO2/Al2O3 ratio of 325 and a Pt content of 0.22%, on which the
yield of propylbenzenes reached 96% with respect to equilibrium
and selectivity to propylbenzenes was about 60%. The mechanisms
of benzene alkylation with propane operating on various catalysts
c
The use of both light alkanes and aromatics as raw ma-
terials in a common process can lead to valuable products
and intermediates in petrochemical synthesis. In particu-
lar, alkylation of benzene or toluene with alkanes would
be a new potential way to synthesize many important alky-
laromatic products, such as ethylbenzene, xylenes, cumene,
ethyltoluene, etc.
The catalytic synthesis of alkylaromatics from alkanes
and benzene was reported by Schmerling and Vesely (2)
for the Friedel–Crafts catalyst systems and then studied by
Olah et al. (3) in superacidic media. The above catalysts
showed high catalytic activity and selectivity. The major re-
action products resulted from direct alkane addition to aro-
matics. However, these catalysts suffered from numerous,
mainly environmental and corrosion drawbacks.
These drawbacks can be overcome by using solid cata-
lysts. In view of this, over the past few years, there have
been some efforts aimed at creating the heterogeneous sys-
tems for this process. In particular, zeolite-type solid acids
were reported to produce alkylbenzenes from aromatics
and light paraffins (4–8). However, these catalysts showed
rather low selectivity toward the products of direct addi-
tion; instead, alkylaromatic products were formed from
the fragments of alkane molecules and aromatic substrate
(5–7). The reaction pathway leading to alkylaromatics in
the presence of zeolites was found to be much more com-
plicated than in the case of conventional acids (6, 7).
Moreover, the reactivity of alkanes over zeolites was re-
ported to be very low with respect to superacids and
therefore high temperatures were required for this process
(5, 7).
It has been demonstrated that zeolite reactivity in alkane
activation can be improved significantly by modification
with Pt, Ga, or Zn (9–11). Another possible way to increase
conversion in the processes leading to H2 release is to shift
thermodynamic equilibria by continuous removal of hydro-
gen from the reaction zone. This could be achieved by ad-
dition of H2 scavengers, such as hydride-forming materials,
as demonstrated for the case of alkane dehydrogenation
are discussed.
2000 Academic Press
Key Words: alkylation with alkanes; propylbenzenes; propane
activation; MFI zeolites; Pt-supported catalysts; H2 scavengers.
INTRODUCTION
Light alkanes such as methane, ethane, propane, and
butanes are among the most abundant and inexpensive hy-
drocarbons available today. Until now there has been no
efficient commercial process for light alkanes functional-
ization, and therefore such a route would represent supe-
rior useful technology for the utilization of natural gas and
similar refinery-derived light hydrocarbon streams (1). One
possible way to upgrade light alkanes is to involve them in
1 To whom correspondence should be addressed.
266
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
267
(12, 13), alcohol decomposition (14), and aromatization
(14, 15) reactions.
Pt/CeO2 and Pt/Al2O3 catalysts were prepared by an in-
cipient wetnessimpregnation method usinga [Pt(NH3)4]Cl2
In this paper, we report on benzene alkylation with aqueous solution. The samples were dried, calcined, and re-
propane on Pt-supported zeolites and oxides in the pres- duced as above.
ence of an H2 acceptor. The effects of the nature of sup-
Intermetallic compound Zr2Fe was prepared from
port, its acidity, metal distribution, and addition of hydro- 99.99% pure zirconium and iron components by noncon-
gen scavenger are examined.
EXPERIMENTAL
sumable electrode arc melting under a purified argon at-
mosphere. Sample was remelted five times to ensure its
homogeneity. The hydrogen absorption capacity found for
the sample prepared was 1.6 wt% , which is close to the data
reported previously (16).
1. Materials
The MFI zeolite catalysts with SiO2/Al2O3 ratios of 51,
104, and 325, designated in text as MFI-50, MFI-100, and
MFI-300, were provided by UOP. CeO2 and Al2O3 were
obtained from Rhoˆne Poulenc.
The characteristics of initial and modified catalysts are
given in Table 1.
2. Characterization
The acidityofthe catalystswasmeasured bytemperature-
programmed desorption (TPD) of ammonia. TPD NH3 ex-
periments were performed in a locally constructed appara-
tus. The catalyst (150 mg, 45 mesh particles) was loaded in
a stainless steel microreactor connected online with a TC
detector. The typical procedure for sample pretreatment
before NH3 adsorption included in situ activation at 773 K
in a flow of dry air for 1 h and then in a flow of dry nitrogen
or 1 h. Pt-containing samples were pretreated in a flow of
high-purity Ar (99.998% ) at 723 K for 1 h. NH3 adsorption
was carried out at ambient temperature in a flow of NH3
diluted with N2 (1 : 1) and purified on a column with granu-
lated NaOH. The complete saturation of the samples with
ammonia wasreached within 30min. Afterward, the weakly
bound NH3 was removed in a flow of dry He at 373 K for
30 min and the reactor was cooled to ambient temperature.
The typical TPD experiment was carried out in a flow of
dry He (30 ml/min), the temperature was increased linearly
from 295 to 973 K, and the rate of heating was 8.5 K/min.
H
forms and Na, H forms of MFI zeolites were ob-
tained by complete and partial ion exchange with NH4Cl
aqueous solution, respectively. Ion exchange was followed
by washing, drying, and calcination at 823 K for 5 h.
Pt/zeolite catalysts were prepared by ion exchange of
NH4 forms of corresponding zeolites with [Pt(NH3)4]Cl2
aqueous solution. After ion exchange, samples were
washed, dried at 333 K, calcined in the dry air flow at 773 K
for 5 h, and reduced in the hydrogen flow at 723 K for 3 h.
For the MFI-300 sample with low ion exchange capac-
ity, this procedure led to low Pt contents (Table 1). To
obtain higher degrees of Pt incorporation in that sam-
ple, the Na form of MFI-300 zeolite was exchanged with
[Pt(NH3)4](OH)2 aqueous solution. The sample was then
washed, dried, calcined, and reduced as above.
TABLE 1
Characteristics of the Catalysts
Pt sites
Acid sites concentration
(TPD NH3 data), mol g
a
1
Average particle size, A
Pt content,
wt%
a
a
a
Catalyst
H-MFI-50
H-MFI-100
Na,H-MFI-100
H-MFI-300
CeO2
SiO2/Al2O3
Na2O/Al2O3
a
a
a
Total
STEM
H2 adsorption
51
104
104
325
—
—
51
104
325
325
—
0.02
<0.01
0.83
0.09
—
153
168
133
61
n/db
n/d
168
90
n/d
27
n/d
n/d
369
120
151
73
n/d
n/d
230
111
n/d
48
334
141
12
856
429
296
212
33
—
—
—
—
—
—
—
—
—
—
—
9
n/d
n/d
n/d
n/d
9
—
—
—
—
—
—
9
10
n/d
12
14
9
78
n/d
n/d
188
153
n/d
30
Al2O3
—
77
—
Pt/H-MFI-50
Pt/H-MFI-100
Pt/(I)/H-MFI-300
Pt(II)/H-MFI-300
Pt/CeO2
<0.01
<0.01
<0.01
<0.01
—
586
354
n/d
105
90
0.29
0.28
0.02
0.22
0.30
0.30
n/d
n/d
n/d
n/d
Pt/Al2O3
—
—
77
a Amount of acid sites, corresponding to
, , and peaks in NH3 TPD profiles of the samples ( : E < 105 kJ/mol; :
105 kJ/mol < E < 145 kJ/mol, : E < 145 kJ/mol).
b Not determined.

268
SMIRNOV ET AL.
The acid sites distribution as a function of activation ener-
The selectivity to a specific reaction product was deter-
gies was determined according to the procedure described mined as the percentage of this product in total feed-free
elsewhere (17).
Pt contents were determined by AE spectroscopy using
a microwave plasma generator “Chromatron-1.”
products.
The yield of propylbenzenes was determined as experi-
mentalconversion ofreactantsinto propylbenzenesdivided
The average sizes of Pt particles were estimated by STEM by equilibrium conversion of reactants into propylben-
and hydrogen adsorption measurements. STEM images zenes. Equilibrium conversion wascalculated for the system
were obtained on Vacuum Generators HB 601. Adsorp- including benzene, propane, cumene, n-propylbenzene, and
tion measurements were carried out on a standard static dihydrogen.
volumetric adsorption apparatus. The dispersion (D) of Pt
particles was determined as D = Ns/Ntot, where Ns corre-
RESULTS
sponds to the fraction of the metal atoms exposed at the
1. Nature and Distribution of Acidic
and Dehydrogenation Sites
surface and Ntot to the total number of Pt atoms in the sam-
ple. Ntot was obtained from chemical analysis data, and Ns
from isotherms of hydrogen adsorption using the following
procedure. The sample was first reduced in H2 at 700 K,
evacuated at that temperature, and then cooled to ambient
temperature in a vacuum. The isotherm of hydrogen ad-
sorption was then measured at ambient temperature. The
linear part of the isotherm was extrapolated to zero pres-
sure to obtain the sum of physisorbed and chemisorbed
hydrogen. Physisorbed hydrogen was then evacuated at
room temperature for 0.5 h and the back-sorption isotherm
was measured. Subtraction of the second isotherm from
the first gave a number of H2 molecules chemisorbed on
the metal particles. Taking the stoichiometry of chemisorp-
tion as 1 H atom per 1 atom of metal (18), the number of
chemisorbed hydrogen atoms gave the fraction of metal
atoms exposed at the surface (Ns). The size of the Pt parti-
cle was then calculated as described in (19).
The materials selected for the study involved three
types of solids: neutral (CeO2), slightly acidic (Al2O3), and
strongly acidic (MFI zeolites). The acidity was character-
ized by TPD of ammonia according to the procedure de-
scribed above. The TPD profiles obtained and the acid sites
distribution as a function of activation energies are pre-
sented in Figs. 1 and 2 and Table 1, respectively.
Cerium oxide which has been selected as a model of
the neutral material is indeed characterized by the lowest
amount of adsorption sites (Fig. 1, Table 1). The adsorption
of ammonia on this sample is probably due to structural
defects, which are always present on ceria.
On aluminum oxide, the amount of adsorption sites is a
factor of 2 higher with respect to ceria. The origin of NH3
adsorption in this case is most probably the Lewis acid sites
of alumina.
Interestingly, the modification of CeO2 and Al2O3 with
Pt results in different consequences. The TPD NH3 pro-
file of Pt/Al2O3 is similar to those observed for Al2O3
(Fig. 1). In contrast, Pt deposition on CeO2 results in signif-
icant changes: a narrow peak at 480 K appears (Fig. 1) and
the total amount of sites increases by a factor of 3 (Table 1).
This observation suggests that Pt modifies the adsorption
3. Catalytic Tests
Catalytic tests were carried out in a continuous-flow cata-
lytic unit. In a typical experiment, 0.8 g of the catalyst was
placed in an up-flow, fixed-bed stainless steel reactor and
pretreated in a flow of dry Ar (30 ml/min) at 723 K for 1 h;
the system was then cooled to the reaction temperature.
Ar flow was passed through the MnO/alumina indicator to
verify the absence of oxygen in the stream.
When the catalytic system included two components
(H-MFI + Pt/CeO2 or Pt/MFI + Zr2Fe), zeolite was thor-
oughly mixed with the second component, and the mixtures
were pressed into pellets, crushed, and size separated to give
catalyst grains between 0.25 and 0.50 mm in diameter. In
the case of the H-MFI + Pt/CeO2 system, 0.8 g of zeolite
was mixed with 0.8 g of Pt/CeO2. For the preparation of
Pt/MFI + Zr2Fe catalyst, 0.8 g of Pt/MFI was ground with
2.4 g of intermetallic compound.
Benzene alkylation with propane was carried out at at-
mospheric pressure, in the temperature range of 573–773 K.
The C3H8/C6H6 molar ratio was 1, and the weight hourly
space velocity (WHSV) of the feed was varied from 0.2 to
7 h 1. The reaction products were analyzed online by GC
using 50-m SE-30 and 50-m KCl/Al2O3 capillary columns.
FIG. 1. TPD NH3 profiles for oxide samples.

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
269
FIG. 2. TPD NH3 profiles for zeolite samples.
a
sites of ceria, while the Lewis sites of alumina remain prac- 14 A. STEM images suggest that Pt distribution is rather
tically unchanged. Further experiments are needed to give uniform. Moreover, on zeolite samples, Pt particles are
an explanation for this phenomenon.
shown to be inside of the channels (Fig. 3). Based on the
TPD NH3 curves for H-MFI zeolites (Fig. 2a) are typical latter observation and taking into account the dimensions
a
a
for this class of materials (20, 21). Three peaks at 350–450, of Pt particles (9 A) and zeolite channels (5.5 A), one can
450–600, and 600–900 K, corresponding to weak, interme- conclude that Pt particles are most probably located in ze-
diate, and strong acid sites, are resolved. According to the olite channel intersections, where more space is available.
classification proposed by Topsoe et al. (20), these peaks Interestingly, the Brønsted acid sites of MFI are also lo-
can be assigned to
, , and sites, respectively. The in- cated in zeolite channel intersections (22). Consequently,
tensity of the peak and the temperature of its maximum the dehydrogenation and acidic sites should be very close
increase with the number of aluminum atoms in the frame- to each other in the MFI samples. This conclusion supports
work (Table 1, Fig. 2a), in line with previous observations our interpretation of the effect of Pt on the acidity of H-MFI
(20, 21). The highest concentration of the strongest sites zeolites.
was observed on the H-MFI-50 sample.
2. Benzene Alkylation with Propane over Acidic Zeolites
On the Na, H-MFI-100 sample, the peak gives the main
contribution to the overall acidity, and the peak is less
essential, while the peak is absent in the TPD profile with propane was performed over H-MFI-50 catalyst. The
(Fig. 2a). Consequently, this sample exhibits mainly weak variations of propane and benzene conversions with tem-
The study of the main regularities of benzene alkyation
acid sites.
perature and contact time are shown in Fig. 4.
Pt deposition decreases the total amount of adsorbed am-
The reaction starts at 573 K, the conversion of reactants
monia on H-MFI samples by a factor of 1.5–2 but does being about 0.3% . In the range of 573–673 K, the conver-
not change significantly the acid sites distribution (Table 1, sion remains rather low. A further increase of the reaction
Fig. 2b). This effect could be either due to the partial block- temperature leads to a significant increase of activity and
age of zeolite channels by Pt particles or due to the interac- at 773 K, conversions of benzene and propane reach 37 and
tion of Pt with the Brønsted acid sites of zeolites. Since the 63% , respectively (Fig. 4a).
pore volume of the samples, measured by N2 adsorption,
The variations of propane and benzene conversions with
does not change upon Pt deposition, the second hypothe- contact time show linear dependences at 623 K (Fig. 4b).
sis is more realistic. Hence, protonic sites of zeolites and The maximum conversions reached at a contact time of 5 h
aprotonic sites of alumina show different behavior upon Pt were 7 and 11% for benzene and propane, respectively. It
deposition.
should be mentioned that at low temperatures (<673 K)
The amount and size distribution of dehydrogenation and contact times (<1 h), the conversion of propane is
sites on Pt-containing samples is given in Table 1 and Fig. 3. close to benzene conversion, suggesting that the main re-
For the most of the samples studied, Pt content is close to action pathway is alkylation, while at higher temperatures
0.3% . For H-MFI-300, two samples with different Pt con- and contact times conversion of neat propane contributes
tents of 0.02 and 0.22% , respectively, were prepared.
According to STEM and H2 adsorption data (Table 1),
significantly to the overall reaction.
The major products observed include C1–C4 gases,
the average Pt particle sizes do not change significantly toluene (T), ethylbenzene (EB), isopropylbenzene (IPB),
within the range of samples studied and vary from 9 to and n-propylbenzene (NPB). Others products detected

270
SMIRNOV ET AL.
FIG. 3. STEM image for Pt/H-MFI-50 sample.
in smaller amounts involve alkyl- and polyalkylbenzenes, of propylbenzenes. Ethylbenzene arises from both primary
naphthalene, and its derivatives. The influence of temper- and secondary reactions which could be as follows:
ature and contact time on the distribution of the main re-
=
C3H8 + C6H6 CH4 + C6H5C2H5
[1]
action products is shown on Fig. 5. At low temperatures
(<623 K) and contact times (<0.5 h) propylbenzenes are
the major reaction products. At higher temperatures and
=
C6H5C3H7 + C6H6 C6H5C2H5 + C6H5CH3.
[2]
contact times, ethylbenzene and methane become the main Toluene is most probably formed only via the secondary
products. A further increase of temperature to 700 K re- route [2] as no ethane is observed as a primary product.
sults in a decrease of ethylbenzene yields and an increase
of conversion into toluene and ethane.
The comparison of the results obtained over the catalysts
with different acidity is presented in Table 2. CeO2, Al2O3,
The analysisofthe initialpartsofkineticcurvespresented and Na,H-MFI-100 materials do not show any activity in
in Fig. 5a suggests that propylbenzenes are the primary re- benzene alkylation with propane at 623 K. The activity of
action products; however, they undergo rapid dealkylation H-MFI zeolites increases with the number of strong acid
or dismutation into ethylbenzene and toluene under our siteson these samples. Indeed, almost linear correlation was
reaction conditions. Methane is also a primary product but obtained between propane conversion and the amount of
the rate of its formation is much lower than in the case sites (Fig. 6a). In contrast, no correlation was observed with
FIG. 4. Variation of benzene and propane conversions on H-MFI-50 with temperature at WHSV = 1.3 h 1 (a) and with contact time at 623 K (b).

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
271
FIG. 5. Variation of product selectivity on H-MFI-50 with temperature at WHSV = 1.3 h 1 (a) and with contact time at 623 K (b).
and sites as well as with the total amount of sites. These
Pt/CeO2 is inactive in benzene alkylation (Table 3). The
observations suggest that only strong Brønsted sites can main process which takes place over this catalyst is propane
activate propane, in line with the previous findings (9–11). dehydrogenation, the conversion ofpropane beinglessthan
It should be mentioned that an increasing amount of 0.2% . Pt/Al2O3 gives 2% propane conversion, but the main
sites favors methane and ethylbenzene formation at the reaction product is also propene. Only small amounts of
expense of propylbenzenes, which leads to decreased selec- propylbenzenes are formed on this catalyst. This result con-
tivity in the alkylation process (Fig. 6b).
Consequently, on one hand, activation of propane needs
firms that Lewis acid sites are not active in alkylation.
MixingofPt/CeO2 with acidic zeolitesresultsin a tremen-
strong acid sites, while on the other hand, these sites dous increase of propane conversion (Table 3). The higher
favor cracking at the expense of alkylation and lead to the acidity of the zeolite, the higher is the improvement. The
low selectivity toward the target process. To improve the maximum propane conversion achieved on Pt/CeO2 + H-
catalyst activity and selectivity, the propane activation step MFI-50 catalyst is 11.8% , which is 65 times higher than on
should be modified. This could be achieved by addition of neat Pt/CeO2 and 14 times higher than on neat H-MFI-50.
a dehydrogenation component and coupling of alkylation This result points to a complementary synergy between
process with dehydrogenation, as was done previously for the actions of Pt and acidic sites in the overall alkylation
alkane isomerization (23, 24) and aromatization (9–11).
process.
A comparison of the main products distribution on
H-MFI-50 and mixed Pt/CeO2 + H-MFI-50 catalysts at the
same conversion level ( 11% ) is presented in Fig. 7. The
following differences occur due to the addition of a dehy-
3. Benzene Alkylation with Propane over
Pt-Supported Catalysts
Platinum metal was selected as the dehydrogenation drogenation component. First, the reaction route leading
component in our study. The results obtained on Pt- to methane and ethylbenzene is suppressed in the presence
supported catalysts are presented in Tables 3 and 4.
of Pt/CeO2. Second, mixing with Pt/CeO2 results in a 5-fold
FIG. 6. Variation of propane conversion (a) and selectivity to the main products (b) with the amount of sites in MFI catalysts (T = 623 K,
WHSV = 1.3 h 1, TOS = 1 h).

272
SMIRNOV ET AL.
TABLE 2
Benzene Alkylation with Propane over the Catalysts with Different Acidity (WHSV = 1.3 h 1, 623 K, TOS = 1 h)
CeO2
Al2O3
H-MFI-300
Na,H-MFI-100
H-MFI-100
H-MFI-50
Total propane
0
0
0.17
0
0.32
0.86
conversion,%
Selectivity to
hydrocarbons, mol%
CH4
C2H6
C3H6
C4–C5
T
EB
IPB
NPB
C9–C11
Others
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
13.61
—
2.80
—
8.08
17.66
13.07
22.12
5.83
16.83
20.36
0.99
1.22
2.88
15.17
35.52
5.79
10.04
—
—
8.04
3.12
17.60
30.90
31.60
8.74
8.03
TABLE 3
1
Benzene Alkylation with Propane over Neat Pt-Supported Oxides and Mixed with Acidic Zeolites (WHSV = 1.3 h
623 K, TOS = 1 h)
,
Pt/CeO2 +
Pt/CeO2 +
Pt/CeO2 +
Pt/CeO2 +
Pt/CeO2
0.18
Pt/Al2O3
1.91
H-MFI-300
Na,H-MFI-100
H-MFI-100
H-MFI-50
Total propane
conversion, %
Selectivity to
hydrocarbons, mol%
1.26
0.77
2.15
11.84
CH4
C2H6
C3H6
C4–C5
T
EB
IPB
NPB
C9–C11
Others
14.14
—
85.86
—
—
—
—
—
—
—
13.31
5.79
72.35
—
—
—
6.90
1.65
—
3.50
0.40
4.83
—
2.20
3.54
29.68
40.74
15.11
—
5.10
0.46
37.71
—
1.71
2.49
22.86
10.30
8.81
4.06
2.77
6.33
3.01
15.98
0.59
26.80
21.54
11.40
4.93
6.61
1.67
7.47
3.73
14.61
17.39
13.75
25.33
3.36
—
10.56
8.67
TABLE 4
Benzene Alkylation with Propane over Pt-Supported H-MFI Catalysts (WHSV = 1.3 h 1, 623 K,
TOS = 1 h)
Pt(I)/H-MFI-300
Pt(II)/H-MFI-300
Pt/H-MFI-100
Pt/H-MFI-50
Total propane
0.45
6.62
22.99
44.56
conversion, %
Selectivity to
hydrocarbons, mol%
CH4
C2H6
C3H6
C4–C5
T
EB
IPB
NPB
C9–C11
Others
—
—
—
0.38
4.91
13.62
6.85
1.27
0.88
23.28
36.60
3.86
8.46
13.03
2.23
9.18
4.29
2.28
8.54
16.90
21.52
13.57
4.39
27.83
0.35
10.26
19.58
1.37
1.58
3.24
7.73
23.67
—
2.56
2.56
35.90
48.72
—
10.26
8.35

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
273
The positive effect depends on the amount of Pt deposi-
ted on zeolite. The increase of Pt content from 0.02 (Pt(I)/
H-MFI-300) to 0.22 (Pt(II)/H-MFI-300) leads to an or-
der of magnitude increase of propane conversion (Table 4).
As in the case of mixed catalysts, direct Pt deposition
suppresses methane formation and results in an increase of
selectivity to propylbenzenes. It should be mentioned, how-
ever, that at high conversions, the process is accompanied
by formation of significant amounts of polyalkylbenzenes,
naphthalene, and its derivatives.
To conclude, both methods of Pt deposition lead to signif-
icant improvement of zeolite activity and selectivity. How-
ever, the best effect is obtained on Pt/H-MFI catalysts, on
which Pt and acidic sites should be very close to each other
and may compose a single bifunctional center.
FIG. 7. Effect ofmixingwith Pt/CeO2 on the product distribution over
H-MFI-50 (T = 623 K, propane conversion = 11% ).
4. Benzene Alkylation with Propane on Pt-Supported
Catalysts in the Presence of an H2 Acceptor
It has been demonstrated that thermodynamic equilibria
in reactions leading to H2 release (dehydrogenation, arom-
atization, etc.) can be shifted by addition of hydrogen scav-
engers, such as O2, CO, or CO2 (25, 26), or intermetallic
compounds [12–15]. In this study, we applied a similar ap-
proach to benzene alkylation with propane.
Zr2Fe intermetallic compound was used as a hydrogen
scavenger. The reaction was carried out at 623 K, WHSV =
1.3 h 1. The results are presented in Table 5 and Fig. 8.
The following observations can be made:
increase of the selectivity to target products. Hence, both
the activity and the selectivity are increased upon Pt
addition.
When Pt was deposited directly on the zeolite com-
ponent the effect was even more pronounced (Table 4).
Propane conversion increases 40–250 times with respect to
neat Pt/CeO2, 40–70 times compared to pure acidic zeolites,
and 4–10 times with respect to mixed Pt/CeO2 + H-MFI
systems. Propane conversion on the Pt/H-MFI-50 sample
reaches 12% at 573 K and 44% at 623 K. On the H-MFI-50
catalyst, the same conversions can be reached at 700 and
—Propane conversion increases 2–8 times upon addition
750 K, respectively. Thus, addition of Pt to zeolite catalyst of Zr2Fe; the effect is more pronounced on the catalysts
allows a decrease in the reaction temperature by more than with lower acidity (Tables 4 and 5).
100 K, which is in line with the results obtained by Bragin
et al. (4, 5).
—The catalysts which exhibit lower activity (Pt(I)/H-
MFI-300 and Pt(II)/H-MFI-300) show stable behavior
TABLE 5
Quad Benzene Alkylation with Propane over Pt-Supported H-MFI Catalysts in the Presence of H2 Acceptor
(WHSV = 1.3 h 1, 623 K, TOS = 1 h)
Pt(I)/H-MFI-300 + Zr2Fe
Pt(II)/H-MFI-300 + Zr2Fe
Pt/H-MFI-100 + Zr2Fe
Pt/H-MFI-50 + Zr2Fe
Total propane
3.78
14.36
38.11
76.20
conversion, %
Selectivity to
hydrocarbons, mol%
CH4
C2H6
C3H6
C4–C5
T
EB
IPB
NPB
C9–C11
Others
0.37
0.74
9.23
1.48
0.74
2.21
33.21
50.18
—
0.24
3.69
12.31
11.83
1.04
5.16
13.20
4.59
20.50
4.66
3.21
8.30
17.86
12.06
10.46
5.72
29.08
0.77
9.67
11.66
4.74
1.89
3.75
9.70
23.02
1.18
23.11
35.65
8.63
1.84
2.32

274
SMIRNOV ET AL.
FIG. 8. Variation of propane conversion with time on stream over
Pt/H-MFI catalysts mixed with Zr2Fe (T = 623 K, WHSV = 1.3 h 1, cata-
lyst : H2 acceptor = 1 : 3).
FIG. 9. Reaction network.
during the time course of the reaction, while on more active
catalysts (Pt/H-MFI-100 and Pt/H-MFI-50), the positive ef-
fect decreases with time on stream (Fig. 8). The latter obser-
vation could be due to saturation of hydrogen scavenger or
due to fast deactivation of the catalytic system at high con-
versions. The estimation of the amounts of H2, which evolve
at the highest conversion levels under typical reaction con-
ditions, have demonstrated that the H2 capacity of Zr2Fe
could be exhausted after 2–3 h on stream. Consequently,
the first hypothesis seems to be more probable.
formed from alkane molecules is more favorable due to the
contribution of the reaction pathways which allow cracking
of alkane.
On nonmodified acidic zeolites, formation of ethylben-
zene and methane was indeed the main reaction pathway
(Table 2, Fig. 5). These catalysts showed the lowest overall
conversion and selectivity to the products of direct addition.
Incorporation of Pt in the catalysts resulted in a signifi-
cant increase of overall conversion and selectivity to propy-
lbenzenes (Tables 3 and 4, Fig. 7). It should be emphasized
that the effect was more pronounced on Pt/H-MFI systems
than on Pt/CeO2 + H-MFI mixed systems, suggesting that
the mean distance between Pt and the acid site is very
—Addition of an H2 acceptor does not change signifi-
cantly the product distribution and selectivity to propyl-
benzenes remains the same as on neat catalyst (Tables 4
and 5).
DISCUSSION
The results presented above and obtained in our previ-
ous studies (6, 7) suggest that interaction of propane with
benzene over zeolite catalysts occurs via three main reac-
tion pathways (Fig. 9) leading to (1) propylbenzenes and
dihydrogen, (2) ethylbenzene and methane, (3) toluene and
ethane. Under our experimental conditions (T > 573 K),
propylbenzenes are very unstable intermediate products,
and convert into benzene and propene or react with
benzene to give ethylbenzene and toluene. Apart from
being involved in alkylation reactions, propane can also
be dehydrogenated to propene which can undergo further
oligomerization, isomerization, cracking alkylation, and
aromatization reactions leading to long-chain oligomers,
various alkylbenzenes, naphthalene, and its derivatives.
Thermodynamic calculations show (Fig. 10) that direct
alkylation is not a favored reaction within the temperature
range studied. In contrast, alkylation with the fragments
FIG. 10. Equilibrium constants as
a
function of temperature
==
==
for the following reactions: (1) C3H8 C3H6 + H2; (2) C3H8 + C6H6
==
C6H5 C3H7 + H2; (3) C3H8 + C6H6 C6H5 C2H5 + CH4; and (4) C3H8
+
==
C6H6 C6H5CH3 + C2H6.

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
275
important. The shorter the distance, the more active and dehydrogenation and acidic centers are very close to each
selective is the catalyst. It appears that dehydrogenation other and can act in synergy as combined bifunctional cen-
and acidic sites work in synergy for propane activation and ters.
further alkylation.
(IV) Bifunctional catalysts with an H2-acceptor (Pt/H-
Finally, addition of an H2 acceptor allows the shift of MFI + Zr2Fe), where the later works as a scavenger for hy-
thermodynamic equilibria and an increase in propane and drogen molecules formed during the reaction.
benzene conversion at the same level of selectivity.
The following mechanistic pathways can be proposed to
In Fig. 11, selectivity is plotted versus conversion for
account for the results obtained for each group of catalysts:
all the catalysts studied. Interestingly, all the data fell on
four distinct curves, suggesting that all the catalysts can be
—On acidic catalysts (group I) selectivity to propylben-
subdivided into four groups according to the mechanisms zenes is very low. High amounts of CH4 observed at low
operating on them:
propane conversion point to a carbonium ion mechanism of
propane activation on such catalysts [7, 27–29]. This mecha-
nism involves propane protonation on the strong Brønsted
sites of H-MFI catalysts and the formation of a carbonium
ion type transition state, which may then evolve in three
different ways leading to propyl-, ethyl-, and methylcarbe-
nium ions, which further alkylate benzene to give propyl-,
ethyl-, and methylbenzenes, respectively:
(I) Pure acidic catalysts (H-MFI).
(II) Mixed catalysts (Pt/CeO2 + H-MFI), which consist
of metal supported on neutral material and an acidic com-
ponent. These catalysts include both dehydrogenation and
acidic centers but the mean distance between these centers
is rather long.
(III) Bifunctional catalysts (Pt/H-MFI), which involve Pt
supported directly on acidic material. On these catalysts,
According to our data, the second route leading to ethyl-
benzene and methane is predominant on this type of
catalysts.
—On mixed catalytic systems (group II) selectivity to
methane decreases and propene appears in products in sig-
nificant amounts. These observations suggest that on such
catalysts propane activation preferentially occurs via its de-
hydrogenation to propene and dihydrogen. Metal (Pt) sites
are most probably responsible for this step. Propene is fur-
ther protonated on acid sites to give propenium ion which
can alkylate benzene leading to propylbenzenes:
Propene formed on the first step can also participate in side
reactions (oligomerization, aromatization, etc.) leading to
a variety of by-products (Fig. 9). Propylbenzenes can un-
dergo secondary reactions giving benzene and propene or
ethylbenzene and toluene.
—On bifunctionalcatalysts(group III) propane dehydro-
genation and protonation may take place on a metal-acidic
center, which results in direct benzene alkylation:
FIG. 11. Selectivity versus conversion (a) and yield versus selec-
tivity (b) plots: (1) H-MFI-300; (2) H-MFI-100; (3) H-MFI-50; (4) H-
MFI-300 + Pt/CeO2; (5) H-MFI-100 + Pt/CeO2; (6) H-MFI-50 + Pt/CeO2;
(7) Pt(I)/H-MFI-300; (8) Pt(II)/H-MFI-300; (9) Pt/H-MFI-100; (10) Pt/H-
MFI-50; (11) Pt(I)/H-MFI-300 + Zr2Fe; (12) Pt(II)/H-MFI-300 + Zr2Fe;
(13) Pt/H-MFI-100 + Zr2Fe; (14) Pt/H-MFI-50 + Zr2Fe.
As in the previous case, the products of direct addition can
undergo dealkylation and dismutation reactions leading to
benzene and propylene or toluene and ethylbenzene, re-
spectively. Of course, one cannot exclude the existence of

276
SMIRNOV ET AL.
separated acidic and Pt sites on Pt-MFI catalysts along with also formed in smallamounts. The reactionsofC2H+5 , C3H7+,
bifunctional metal–acidic centers. On the separated sites, and CH+3 carbenium ions with benzene lead to ethylben-
a mechanism similar to those proposed for mixed catalytic zene, propylbenzenes, and toluene, respectively.
systems should be realized.
Incorporation of Pt into the catalysts results in a signif-
—Finally, when the reaction is carried out on bifunctional icant increase of the overall conversion and selectivity to
catalysts in the presence of an H2 acceptor (group IV), the propylbenzenes. It is demonstrated that on Pt-containing
reaction mechanism is the same as in the previous case but systems, the reaction mechanism is bifunctional: both de-
thermodynamic equilibrium is shifted toward formation of hydrogenation and acidic sites act in synergy for propane
propylbenzenes:
activation, leading preferentially to H2 and C3H+7 and fur-
ther to propylbenzenes. The effect is more pronounced on
Pt/H-MFI systems than on Pt/CeO2 + H-MFI mixed sys-
tems, due to the shorter mean distance between the acidic
and dehydrogenation sites on the former.
Addition of an H2 scavenger shifts the thermodynamic
equilibrium of the reaction and allows for further im-
provement of activity and selectivity. The best result is ob-
served on the (Pt(II)/H-MFI-300 + Zr2Fe) catalytic system,
on which the yield of propylbenzenes reaches 96% with re-
spect to equilibrium and selectivity to propylbenzenes is
about 60% .
The results show that an H2 acceptor (Zr2Fe) influences
only the initial reaction step and does not participate in
the other reactions. As a result, the product distribution
remains as on the neat catalyst, while conversion increases
2–8 times.
To compare the activity of various catalysts and to select
the best catalytic systems among those studied, the yields
of propylbenzenes were determined as described in the
Experimental part and plotted versus selectivity. The re-
sults are presented in Fig. 11b. The highest yields of 66–44%
are observed on Pt/H-MFI-100and Pt(II)/H-MFI-300, cata-
lysts. Addition of Zr2Fe shifts the thermodynamic equilib-
rium and allows for a 2-fold increase of the yield. The high-
est yield of 114% with respect to equilibrium is observed
on (Pt/H-MFI-100 + Zr2Fe); however, selectivity to propy-
lbenzenes is rather low (26% ) on this catalytic system. On
(Pt(II)/MFI-300 + Zr2Fe), the yield is only slightly lower
(96% ), however, selectivity to propylbenzenes increases 2
times and reaches 60% .
REFERENCES
1. Shilov, A. E., “Activation of Saturated Hydrocarbons by transition
Metal Complexes,” p. 304. Reidel, Dordrecht, 1984.
2. Schmerling, L., and Vesely, J. A., J. Org. Chem. 38, 312 (1973).
3. Olah, G. A., Schilling, P., Staral, J. S., Halpern, Yu., and Olah, J. A.,
J. Am. Chem. Soc. 97, 6807 (1975).
4. Bragin, O. V., Vasina, T. V., Isaev, S. A., and Minachev, Kh. M., Izv.
Akad. Nauk USSR, Ser. Khim. 7, 1682 (1987).
5. Isaev, S. A., Vasina, T. V., Bragin, O. V., Izv. Akad. Nauk USSR, Ser.
Khim. 10, 2228 (1991).
6. Ivanova, I. I., Blom, N., and Derouane, E. G., J. Mol. Catal. 109, 157
(1996).
7. Ivanova, I. I., and Fajula, F., Proc. 12th Int. Zeolite Conf. IV, 2273
(1998).
8. Kato, S., Nakagawa, K., Ikenaga, N., and Suzuki, T., Chem. Lett., 207
(1999).
CONCLUSION
9. Guisnet, M., Gnep, N. S., and Alario, F., Appl. Catal. 89, 1 (1992).
10. Ono, Y., Catal. Rev. Sci. Eng. 34, 179 (1992).
Interaction of benzene with propane over zeolite cat-
alysts proceeds via three main reaction pathways lead- 11. Derouane, E. G., Abdul Hamid, S. B., Ivanova, I. I., Blom, N., and
Hojlund-Nielsen, P. E., J. Mol. Catal. 86, 371 (1994).
12. Fanelli, A. J., Maeland, A. J., Rosan, A. M., and Crissey, R. K., J. Chem.
Soc., Chem. Commun. 8 (1985).
13. Imai, H., Tagawa, T., and Kuraishi, M., Mater. Res. Bull. 20, 511
ing to (1) propylbenzenes and dihydrogen, (2) ethylben-
zene and methane, and (3) toluene and ethane. Under
our experimental conditions (T > 573 K), propylbenzenes
are very unstable and are converted into benzene and
(1985).
propene or interact with benzene to give ethylbenzene and 14. Imai, H., Tagawa, T., and Nakamura, K., Appl. Catal. 62, 348
(1990).
toluene.
15. Chetina, O. V., Vasina, T. V., Lunin, V. V., and Bragin, O. V., Catal.
Today 13, 639 (1992).
16. Van Essen, R. M., and Buschow, K. H. J., J. Less Common Met. 64, 277
On acidic H-MFI zeolites, the major reaction route is
conversion into methane and ethylbenzene. The products
of direct addition are observed only in small amounts at low
(1979).
temperatures (<650 K), when the conversion is very low 17. Yushchenko, V. V., Russ. J. Phys. Chem. 71, 547 (1997).
18. O’Rear, D. J., Loffler, D. G., and Boudart, M., J. Catal. 121, 131
(<2% ). The results observed are accounted for on the basis
of carbonium ion mechanistic pathway including propane
protonation on strong Brønsted sites, and formation of pro-
ponium ion, which is further decomposed predominantly
(1990).
19. Scholten, J. J. F., in “Preparation of Catalysts II” (B. Delmon,
P. Grange, P. Jacobs, and G. Poncelet, Eds.), p. 685. Elsevier Scien-
tific, New York, 1979.
into CH4 and C2H+; H2 and C3H+ or C2H6 and CH+ are 20. Topsoe, N. Y., Pedersen, K., and Derouane, E. G., J. Catal. 70, 41(1981).
7
3
5

BENZENE ALKYLATION WITH PROPANE OVER ZEOLITES
277
21. Meshram, N. R., Hegde, S. G., Kulkarni, S. B., and Ratnasamy, P., Appl. 26. Meitzner, G. D., Iglesia, E., Baumgartner, J. E., and Huang E. S.,
Catal. 8, 359 (1983).
J. Catal. 140, 209 (1993).
22. Jacobs, P., von Ballmoos, R., J. Phys. Chem. 86, 3050 (1982).
23. Maxwell, L. E., and Stork, W. H. G., Stud. Surf. Sci. Catal. 58, 575
(1991).
27. Haag, W. O., and Dessau, R. M., in “Proceedings, 8th International
Congress on Catalysis, Berlin, 1984,” p. 305. Dechema, Frankfurt-am-
Main, 1984.
24. Yashima T., Wang, Z. B., Kamo, A., Yoneda, T., and Komatsu, T., Catal. 28. Krannila, H., Haag, W. O., and Gates, B. C., J. Catal. 135, 115 (1992).
Today. 29, 279 (1981).
29. Ivanova, I. I., Pomakhina, E. B., Rebrov A. I., and Derouane, E. G.,
25. Iglesia, E., and Baumgartner, J. E., Catal. Lett. 21, 55 (1993).
Topics Catal. 6, 49 (1998).