Insight into the mechanisms of the ethylbenzene disproportionation: Transition state shape selectivity on zeolites

Article abstract of DOI:10.1021/ja8042849

The direct experimental evidence shows that ethylbenzene disproportionation is a transition state shape selective reaction on zeolites: a bimolecular reaction mechanism via diphenylethane-mediated pathway on large-pore zeolites X and Y (ca. 0.74 nm) and a monomolecular reaction mechanism on medium-pore zeolites ZSM-5 (ca. 0.56 nm) via the ethoxy-mediated intermolecular ethyl group transfer. The lifetime of bulky diphenylethane species was prolonged by a fine-tune of FAU-zeolites, which makes this transition state detectable by ¹³C MAS NMR spectroscopy. Due to tunable catalytic properties and pore shapes, zeolites are promising catalysts toward emulating the efficiency and selectivity in desired reactions. Copyright

Full text of DOI:10.1021/ja8042849

Published on Web 08/29/2008
Insight into the Mechanisms of the Ethylbenzene Disproportionation:
Transition State Shape Selectivity on Zeolites
Jun Huang, Yijiao Jiang, V. R. Reddy Marthala, and Michael Hunger*
Institute of Chemical Technology, UniVersity of Stuttgart, 70550 Stuttgart, Germany
Received June 24, 2008; E-mail: michael.hunger@itc.uni-stuttgart.de
Scheme 1
Chemists and chemical engineers seek to utilize zeolites “under
control” in heterogeneously catalyzed reactions.1 With the ability
to finely tune zeolite catalysts by synthesis in a wide range of Si/
Al ratios, dealumination, and ion exchange, it is possible to obtain
various properties of multiple catalytic sites. Brønsted and Lewis
acid sites on zeolites initiate catalytic reactions via proton transfer
or hydride abstraction,2,3 and negatively charged framework oxygen
can stabilize formed hydrocarbon cations.4 These multiple active
centers of zeolites can work together to promote a desired reaction.
By choosing specific zeolite structures, channel and cage geometry
can be varied, which can strongly influence the activity and
selectivity in chemical transformations.4-6 Generally, three mech-
anisms are described for the shape selective catalysis on zeolites:
Reactant shape selectivity, product shape selectivity, and transition
state shape selectivity.1c,d In contrast to reactant and product shape
selectivity, it is difficult to give direct experimental evidence for
the transition state shape selectivity. Recently, Clark et al.7 used
theoretical methods to investigate the transition state shape selectiv-
ity in the m-xylene disproportionation on zeolites. This theoretical
study proposed a diphenylmethane-mediated mechanism for FAU-
type zeolite catalysts and a methoxide-mediated pathway in more
restricted MFI- and MOR-type zeolite catalysts. In the present work,
FAU- and MFI-type zeolites with pore sizes of 0.74 and 0.56 nm,8
respectively, were utilized to examine the transition state shape
selectivity in a catalyzed reaction. In addition, these zeolite catalysts
have been prepared with suitable catalytic properties, which make
the reaction intermediates detectable by solid-state NMR spectroscopy.
Disproportionation of aromatic hydrocarbons is an important
petrochemical process.6,7,9 Ethylbenzene disproportionation has
attracted much attention in recent years due to the production of
valuable diethylbenzene and as a standard reaction for acidity
characterization by the International Zeolite Association (IZA).9
As shown by earlier studies, this reaction strongly depends on the
acidic properties and pore shape of zeolites.9
Scheme 2
peak at 29 ppm due to the 13C-enriched methylene group of
ethylbenzene. Two weak signals at 15 and 129 ppm are assigned
to the nonenriched carbon atoms of the methyl group and the
aromatic ring, respectively.11 In contrast to the 13C MAS NMR
signal of the enriched methylene group of A adsorbed on zeolite
H,Na-Y/91 (Figure 1a), these signals are much broader in the spectra
of A adsorbed on zeolites Al,Na-Y/63 and Al,Na-X/61 (Figure 1b
and 1c). Zeolites Al,Na-Y/63 and H,Na-Y were prepared from the
same parent material. Lercher and co-workers3a reported that 3-fold
charged extra-framework lanthanum cations in zeolite X act as
strong adsorption sites and can polarize the C-H bonds and even
activate alkanes by hydride abstraction under near-ambient condi-
tions. The extra-framework aluminum species of the aluminum-
exchanged zeolites X and Y in the present work, therefore, should
cause comparable interactions with the side chain of ethylbenzene.
These interactions result in a heterogeneity of the adsorbate
complexes formed by ethylbenzene and cause a chemical shift
distribution responsible for the broadening of the 13C MAS NMR
signals of ethyl groups. Quantum chemical calculations indicated
that zeolites can initiate the m-xylene transalkylation by the hydride
Most of the mechanistic studies of ethylbenzene disproportion-
ation have been carried out in liquid Friedel-Crafts or superacidic
catalysts.10 There are two proposed reaction pathways: (i) the
bimolecular diphenylethane-mediated reaction pathway (Scheme 1)
and (ii) the monomolecular ethyl-transfer reaction pathway (Scheme
2).10 Up to now, the reaction intermediates have not been observed.
The present study gives experimental evidence for two mechanistic
pathways caused by the transition state shape selectivity during the
ethylbenzene disproportionation on zeolites. The bulky transition
state (diphenylethane species) is directly observed by in situ solid-
state NMR spectroscopy on large-pore zeolites upon a “fine-tune”
of the acidity (see Supporting Information). However, this inter-
mediate is too large to be formed on medium-pore zeolites ZSM-
5, and the ethoxy-mediated pathway was verified for the latter
material.
After adsorption of ethyl[R-13C]benzene (A) on FAU-type
zeolites, the 13C MAS NMR spectra in Figure 1 show a dominating
9
12642 J. AM. CHEM. SOC. 2008, 130, 12642–12644
10.1021/ja8042849 CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
the reaction of B with another A to form C rather than the
recombination with a hydride ion back to A. In addition, the SiO-Al
sites may stabilize the benzenium-type carbenium ions C and E,
which would contribute to the signal at 46 ppm in Figure 1.
For the transalkylation of m-xylene, Clark et al.7 proposed a direct
migration of the additional proton from one aromatic ring to the
other in the intermediate diphenylmethane and a subsequent
scission. In the present work, this corresponds to a direct step from
C to E, which is split into F and G. In this case, no formation of
D would occur in the catalytic cycle. Due to the high energy barriers
for the hydride abstraction and transfer (F to B), the rate of the
formation of diphenylethane species would be much lower than
that of the scission into F and G. The diphenylethane species would
be rapidly split and could not be the compound observed at 46
ppm in the 13C MAS NMR spectra (Figure 1). Therefore, it must
be assumed that the additional proton of C migrates to the nearby
SiO-Al site leading to species D. Due to the small number of
Brønsted acid sites in the supercages (0.275 and 0.15 SiOHAl/s.c.
for Al,Na-Y/63 and Al,Na-X/61, respectively, with 8 s.c./u.c.), the
second aromatic ring of species D can not be rapidly protonated to
form species E. Therefore, species D exist with a lifetime that allows
observation by 13C MAS NMR spectroscopy (Figure 1b and 1c).
In zeolite H,Na-Y/91 with the much higher concentration of
Brønsted acid sites of 2.5 SiOHAl/s.c., the species D are directly
protonated to species E on nearby unoccupied Brønsted acid sites
and E is rapidly split into F and G. In this case, the short lifetime
of D makes it undetectable by 13C MAS NMR spectroscopy (Figure
1a). However, the experimental observation of diphenylethane (D)
in the ethylbenzene disproportionation on zeolites Al,Na-X and
Al,Na-Y provides the undoubted evidence that this reaction occurs
via the bimolecular reaction pathway on large-pore zeolites (Scheme
1). Because of the significantly smaller pores of zeolite ZSM-5 (0.56
nm) compared to zeolites X and Y (0.74 nm), bulky diphenyl-
methane-type transition states are too large to be formed on
ZSM-5.
Figure 1. 13C MAS NMR spectra of ethyl[R-13C]benzene conversion on
zeolites H,Na-Y/91 (a), Al,Na-Y/63 (b), and Al,Na-X/61 (c). On all these
zeolites, two molecules per supercage were loaded. Asterisks indicate
spinning sidebands.
abstraction at one of the methyl groups leading to a carbenium ion.7
In a recent in situ pulsed-flow 1H MAS NMR-UV/vis study, it was
shown that extra-framework aluminum species of dealuminated
H-Y acting as Lewis acid sites play a crucial role in the formation
of sec-ethylphenyl carbenium ions (B) by hydride abstraction.12
The formation of this carbenium ion on zeolite H,Na-Y was found
to require higher reaction temperatures or longer heating times.12
Upon heating zeolite Al,Na-Y/63 at 443 K for 24 h, a new 13C
MAS NMR signal occurred at 46 ppm (Figure 1b), which is due
to diphenylethane (D). This species was also observed in the 13C
CP MAS NMR spectra of ethylbenzene disproportionation on
zeolite Al,Na-X/61 at 453 K (Figure 1c). Heating of zeolite H,Na-
Y/91 at the same temperatures did not result in a signal of species
D (Figure 1a), but a signal of diethylbenzene (H) appeared at 27
ppm.11 Benzene (G) was generated at the same time. However, it
is difficult to be observed in 13C MAS NMR spectra due to its
nonenriched 13C atoms. Species H should also be produced on
zeolites Al,Na-Y/63 and Al,Na-X/61, but this signal might be
overlapped by the broad signal at 29 ppm. As described in the
Supporting Information, zeolites Al,Na-Y/63 and Al,Na-X/61 were
prepared with 11.9 and 16.9 extra-framework aluminum species
per unit cell and 2.2 and 1.2 bridging OH groups in the supercages
per unit cell, respectively. Furthermore, these materials contain ca.
10.7 and 20.6 SiO-Al/u.c., respectively, acting as base sites.13
Increasing the number of highly charged Al3+ cations can enhance
the polarizing effect on the C-H bonds of the methylene group of
A and the further formation of B by hydride abstraction.3 The
hydride ions are stabilized by the extra-framework aluminum
cations, and the SiO-Al sites stabilize the nearby B. This promotes
Early studies of the alkylbenzene disproportionation in HF-BF3
solution indicated that the alkyl group maintains its configuration
when it transfers from one ring to another via a transition state
containing a partial bond between the alkyl group and the second
alkylbenzene.10b The content of ca. 90% of ethylbenzene was
disproportionated via this pathway at room temperature. At higher
temperatures (439-553 K), however, the protonated alkylbenzene
was split into a benzene molecule and an alkylcarbonium ion. This
carbonium ion alkylated a second alkylbenzene to complete the
disproportionation or further reacted by polymerization or hydride
abstraction. Hence, the alkylbenzene disproportionation by the
monomolecular mechanism is accompanied by side reactions
leading to complex products.10b
Figure 2 shows 13C MAS NMR spectra of ethyl[R-13C]benzene-
loaded zeolite Al,Na-ZSM-5 before and after thermal treatments
at different temperatures. Before heating, 13C CP MAS signals occur
at 15, 29, and 129 ppm, which are assigned to the nonenriched-
methyl group, 13C-enriched methylene group, and nonenriched
aromatic ring of ethylbenzene, respectively.11 Upon heating zeolites
ZSM-5 at 483 and 503 K for 0.5 h, the reaction starts as indicated
by 13C CP MAS NMR signals at 14, 73, and 85-90 ppm. Based
on recent studies of the adsorption of 13C-1-ethanol on zeolite
H,Na-Y at 453 K, the weak signal at 73 ppm can be assigned to
surface 13C-1-ethoxy groups (J).14a This active surface compound
can combine with a second A to form diethylbenzene (H). Due to
the restricted pore size of zeolite ZSM-5, p-diethylbenzene is the
most preferred product in the disproportionation and only few
m-diethylbenzenes are formed.9 The 13C MAS NMR signals of p-
9
J. AM. CHEM. SOC. VOL. 130, NO. 38, 2008 12643

C O M M U N I C A T I O N S
states of ethylbenzene disproportionation and results in different
reaction mechanisms. On large-pore zeolites X and Y (ca. 0.74 nm),
the bulky diphenylethane species are produced, but this type of
transition state is too large to be formed on medium-pore zeolite
ZSM-5 (ca. 0.56 nm). The present work shows a bimolecular
reaction mechanism of ethylbenzene disproportionation on large-
pore zeolites and a monomolecular reaction pathway on medium-
pore zeolites via the ethoxy-mediated intermolecular ethyl group
transfer. The latter pathway is accompanied by side reactions. Upon
“fine-tune” of zeolites X and Y (decrease of Brønsted acid sites,
increase of Lewis acid sites and base sites) it is possible to promote
the formation of diphenylethane species and reduce their subsequent
scission. On these zeolite catalysts, diphenylethane species exist
with a lifetime that allows their observation by solid-state NMR
spectroscopy.
Figure 2. 13C MAS NMR spectra of ethyl[R-13C]benzene conversion on
Al,Na-ZSM-5/52 (1 molecule per SiOHAl group). Asterisks indicate
spinning sidebands.
Acknowledgment. Financial support by the Deutsche For-
schungsgemeinschaft is gratefully acknowledged.
and m-diethylbenzene occur at 28.8 and 27.9 ppm, respectively.11
The signals of these products are difficult to distinguish from the
strong signal of A at 29 ppm.
Supporting Information Available: Sample preparation and char-
acterization, ethylbenzene disproportionation on H,Na-ZSM-5. This
material is available for free of charge via the Internet at http://
pubs.acs.org.
After decomposition of the surface 13C-1-ethoxy species at 523
K, Wang et al.14a observed a weak and broad signal at 78-89 ppm
and a strong signal at 13 ppm. Stepanov et al.14b detected signals
at 89 and 14 ppm after adsorption of 13C-1-ethylene on H-ZSM-5
at 296 K. They assigned the signal at 89 ppm to oligomeric alkoxy
groups, which was supported by the signal of terminal methyl
carbons of alkoxy groups at 13 to 14 ppm. Accordingly, the signals
at 14 and 85-90 ppm in the present work were assigned to terminal
methyl carbons of oligomeric alkoxy groups and carbon atoms of
these oligomeric alkoxy groups bound to framework oxygen atoms,
respectively. Upon thermal treatments of ethylbenzene-loaded
zeolites Al,Na-ZSM-5 and H,Na-ZSM-5 at 503 K for 0.5 h and at
573 K for 0.5 or 1 h, respectively, the oligomeric species are further
transferred into aromatics (increasing the intensity of peak at 129
ppm), toluene (130, 20 ppm), butane (23 ppm), propane (15), and
ethane (6 ppm). These byproducts were also observed upon ethylben-
zene disproportionation in the HF-BF3 solution and are considered
as a hint for the monomolecular reaction mechanism.10b Therefore,
the compared reaction on large-pore zeolites is a clean process
without side reactions, which is interesting for industrial applications.
Based on the above-mentioned experimental findings, the eth-
ylbenzene disproportionation on the medium-pore zeolite ZSM-5
is proposed to be a monomolecular reaction via the ethoxy-mediated
pathway as summarized in Scheme 2. After adsorption of A on
ZSM-5, the aromatic ring is rapidly protonated to form the
ethylcyclohexadienyl carbenium ion I even at room temperature
with the low activation energy of 29 kJ/mol.15 The proton is added
to the carbon atom holding the ethyl group to form a σ-complex.10b
At higher temperatures, the σ-complex I is very unstable and splits
off an ethyl cation from the aromatic ring to produce the benzene
molecule G. The alkyl cation is stabilized by a nearby SiO-Al site
and forms a more stable ethoxy group J.7 Subsequently, J combines
with a second A to yield the diethylbenzylium cation with an
additional proton at the aromatic ring (K). This proton can migrate
back to the nearby SiO-Al site to complete the catalytic cycle,
and diethylbenzene H is produced as the final product.
References
(1) (a) Rees, L. V. C. Nature 1984, 309, 583. (b) Hupp, J. T.; Poeppelmeier,
K. R. Science 2005, 309, 2008–2009. (c) Sto¨cker, M. Microporous
Mesoporous Mater. 2005, 82, 257–292. (d) Smit, B.; Maesen, T. L. M.
Nature 2008, 451, 671–678.
(2) (a) Mildner, T.; Freude, D. J. Catal. 1998, 178, 309–314. (b) van Santen,
R. A.; Kramer, G. J. Chem. ReV. 1995, 95, 637–660. (c) Corma, A. Chem.
ReV. 1995, 95, 559–614.
(3) (a) Sievers, C.; Onda, A.; Guzman, A.; Otillinger, K. S.; Olindo, R.; Lercher,
J. A. J. Phys. Chem. C 2007, 111, 210–218. (b) Li, S.; Zheng, A.; Su, Y.;
Zhang, H.; Chen, L.; Yang, J.; Ye, C.; Deng, F. J. Am. Chem. Soc. 2007,
129, 11161–11171.
(4) (a) Bhan, A.; Iglesia, E. Acc. Chem. Res. 2008, 41, 559–567. (b) Farneth,
W. E.; Gorte, R. J. Chem. ReV. 1995, 95, 615–635.
(5) (a) Jones, C. W.; Tsuji, K.; Davis, M. E. Nature 1998, 393, 52–54. (b)
Corma, A.; Fornes, V.; Pergher, S. B.; Maesen, Th. L. M.; Buglass, J. G.
Nature 1998, 396, 353–356. (c) Weisz, P. B.; Haag, W. O.; Rodewald,
P. G. Science 1979, 206, 57–58.
(6) (a) Jones, C. W.; Zones, S. I.; Davis, M. E. Microporous Mesoporous Mater.
1999, 28, 471–481. (b) Jones, C. W.; Zones, S. I.; Davis, M. E. Appl. Catal.,
A 1999, 181, 289–303.
(7) (a) Clark, L. A.; Sierka, M.; Sauer, J. J. Am. Chem. Soc. 2004, 126, 936–
947. (b) Clark, L. A.; Sierka, M.; Sauer, J. J. Am. Chem. Soc. 2003, 125,
2136–2141.
(8) Meier, W. M.; Olson, D. H.; Baerlocher, Ch. Atlas of Zeolite Structure
Types, revised 4th ed.; Elsevier: New York, 1996.
(9) (a) De Vos, D. E.; Ernst, S.; Perego, C.; O’Connor, C. T.; Sto¨cker, M.
Microporous Mesoporous Mater. 2002, 56, 185–192. (b) Weiss, U.; Weihe,
M.; Hunger, M.; Karge, H. G.; Weitkamp, J. Stud. Surf. Sci. Catal. 1997,
105, 973–980. (c) Karge, H. G.; Ernst, S.; Weihe, M.; Weiss, U.; Weitkamp,
J. Stud. Surf. Sci. Catal. 1994, 84, 1807–1812. (d) Arsenova, N.; Bludau,
H.; Haag, W. O.; Karge, H. G. Microporous Mesoporous Mater. 2000,
35-36, 113–119. (e) Weihe, M.; Hunger, M.; Breuninger, M.; Karge, H. G.;
Weitkamp, J. J. Catal. 2001, 198, 256–265.
(10) (a) Streitwieser, A.; Reif, L. J. Am. Chem. Soc. 1960, 82, 5003–5005. (b)
McCaulay, D. A.; Lien, A. P. J. Am. Chem. Soc. 1953, 75, 2411–2413. (c)
Lien, A. P.; McCaulay, D. A. J. Am. Chem. Soc. 1953, 75, 2407–2410.
(11) (a) Philippou, A.; Anderson, M. W. J. Catal. 1997, 167, 266–272. (b)
Pradhan, A. R.; Lin, T.; Chen, W.; Jong, S.; Wu, J.; Chao, K.; Liu, S. J.
Catal. 1999, 184, 29–38.
(12) Huang, J.; Jiang, Y.; Marthala, V. R. R.; Ooi, Y. S.; Hunger, M.
ChemPhysChem 2008, 9, 1107–1109.
(13) Huang, J.; Jiang, Y.; Marthala, V. R. R.; Thomas, B.; Romanova, E.;
Hunger, M. J. Phys. Chem. C 2008, 112, 3811–3818.
(14) (a) Wang, W.; Jiao, J.; Jiang, Y.; Ray, S. S.; Hunger, M. ChemPhysChem
2005, 6, 1467–1469. (b) Stepanov, A. G.; Luzgin, M. V.; Romannikov,
V. N.; Sidelnikov, V. N.; Paukshtis, E. A. J. Catal. 1998, 178, 466–477.
(15) Huang, J.; Jiang, Y.; Marthala, V. R. R.; Wang, W.; Sulikowski, B.; Hunger,
M. Microporous Mesoporous Mater. 2007, 99, 86–90.
In summary, this work provides NMR spectroscopic evidence
that the pore shape of zeolite catalysts strongly affects the transition
JA8042849
9
12644 J. AM. CHEM. SOC. VOL. 130, NO. 38, 2008