Enhancing shape selectivity without loss of activity-novel mesostructured ZSM5 catalysts for methylation of toluene to p-xylene

Article abstract of DOI:10.1039/c3cc46197a

An increase in p-xylene selectivity was observed without losing the catalytic activity over novel mesoporous nano-sized ZSM5 crystals covered with an external SiO₂ overlayer created by deposition of tetraethyl orthosilicate.

Full text of DOI:10.1039/c3cc46197a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Enhancing shape selectivity without loss of
activity – novel mesostructured ZSM5 catalysts
for methylation of toluene to p-xylene†
Cite this: Chem. Commun., 2013,
49, 10584
Received 13th August 2013,
John H. Ahn,a Robin Kolvenbach,a Sulaiman S. Al-Khattaf,b Andreas Jentysa and
Johannes A. Lercher*a
Accepted 20th September 2013
DOI: 10.1039/c3cc46197a
www.rsc.org/chemcomm
An increase in p-xylene selectivity was observed without losing the zeolite crystals with particular shapes, thus, tailoring the diffu-
catalytic activity over novel mesoporous nano-sized ZSM5 crystals sivity by selectively increasing the transport path length of the
covered with an external SiO2 overlayer created by deposition of crystals10–12
tetraethyl orthosilicate.
We report here a novel approach towards tailored nano-
structured zeolites that are able to produce p-xylene selectively
p-Xylene is an important intermediate for the production of without losing catalytic activity compared to the parent zeolite
polyethylene terephthalate with a current capacity of 26 Â by combining the sequence of well known post-treatment pro-
106 tons per year, growing at 6–8% annually.1 It is primarily cedures as follows: (i) enhancing the transport rates of mole-
produced from steam cracking of naphtha and catalytic reforming, cules in the zeolite by generating mesopores and reducing the
along with benzene, toluene and other xylene isomers.1 Among effective diffusion length via desilication (DS),13 (ii) increasing
these aromatic molecules, toluene is formed in excess and the transport rates even further by selectively removing the
p-xylene in deficit relative to the market demand,2 and, thus, extra-framework aluminum species from pore entrances of the
the routes to convert toluene selectively to p-xylene, such as the modified nanocrystals (DS) via dealumination (DS–DA), and
acid catalyzed toluene disproportionation and methylation,3 (iii) compensating the loss of the shape selectivity by increasing
have been the focus of industrial and academic research for the the diffusivity of p-xylene and decreasing that of m-/o-xylene via
last few decades.
modification with a mesoporous SiO2 overlayer (SM).9,14
The zeolite modification sequence is shown in Scheme 1 and
Medium pore acid zeolites, such as H-ZSM5, are well-proven
shape-selective catalysts,4 because their micropore dimensions the chemical composition and the textural properties are compiled
are similar to the size of the aromatic reactants and products in Table 1. The external surface areas and mesopore volumes
(B0.55 nm). It has been shown that the selectivity of p-substituted increased significantly after desilication, but only marginally after
products is enhanced (from B25 to >80%) by impregnating the subsequent dealumination. The mesopore volume also increased
zeolites with boron or phosphorus compounds5 as well as by the deposition of a SiO2 overlayer, but was accompanied by a
chemical vapor6 or liquid deposition7,8 of tetraethyl orthosili- modest decrease in the micropore volume.
cate (TEOS). The modification procedures partially block pore
openings and enhance the shape selectivity by increasing and
decreasing the diffusivity of p- and o-/m-xylenes, respectively.9
These methods, however, suffer from a lower catalyst effici-
ency induced by the decrease in diffusivity of bulky aromatic
products, and lower concentration of Brønsted acid sites.5–9
Alternatively, the selectivity can be enhanced by synthesizing
a
¨
¨
Department of Chemistry, Technische Universitat Munchen, Lichtenbergstraße 4,
85748 Garching, Germany. E-mail: Johannes.Lercher@ch.tum.de;
Fax: +49 89 289 13544; Tel: +49 89 289 13540
b Center of Research Excellence in Petroleum Refining and Petrochemicals,
King Fahd University of Petroleum & Minerals, 31261 Dhahran, Saudi Arabia.
E-mail: skhattaf@kfupm.edu.sa; Fax: +966 3 860 4509; Tel: +966 3 860 2029
† Electronic supplementary information (ESI) available: Experimental details,
concentration of acid sites, in situ IR spectra obtained during toluene methylation
and estimation of surface coverage. See DOI: 10.1039/c3cc46197a
Scheme 1 Zeolites prepared from H-ZSM5 by desilication (DS), subsequent
dealumination (DS–DA), and deposition of the SiO2 overlayer with tetraethyl
orthosilicate from H-ZSM5 (SM), DS (DS–SM) and DS–DA samples (DS–DA–SM).
c
10584 Chem. Commun., 2013, 49, 10584--10586
This journal is The Royal Society of Chemistry 2013

View Article Online
Communication
ChemComm
Table
1
Chemical composition, textural and diffusional properties of
the samples
H-ZSM5 DS DS–DA SM DS–SM DS–DA–SM
Si/Al ratio
36
57
0.12
0.03
6.1
10
27
96
39
99
42
24
36
41
47
84
0.08
0.06
15
Sext (m2 gÀ1
)
a
b
Vmi (cm3 gÀ1
)
0.12 0.12
0.05 0.05
23
11
0.09 0.06
0.08 0.08
1.8 8.6
Vmeso (cm3 gÀ1
)
b
Do-x (10À18 m2 sÀ1
)
)
63
23
c
Dp-x (10À18 m2 sÀ1
17
18
26
d
a
b
Sext = external surface area. Vmi and Vmeso correspond to the micro-
c
and meso-pore volume of the zeolite, respectively. Do-x = apparent
diffusion coefficient of o-xylene at 403 K, from infrared spectroscopy
d
uptake rates.
Dp-x = apparent diffusion coefficient of p-xylene at 403 K,
from zero length column measurements.
Fig. 1 Selectivity of p-xylene within xylenes vs. toluene consumption rate during
toluene methylation at 723 K (ptoluene = 6.0 kPa, pmethanol = 1.5 kPa, 10 mg of
catalyst, C1 (MeOH and dimethyl ether) conversion = 82–89%, total flow rate =
2.3 cm3 sÀ1). The filled symbols represent samples with deposition of the SiO2
overlayer (SM (’), DS–SM (m) and DS–DA–SM (K)) and unfilled symbols without
(parent HZSM5 (&), DS (n) and DS–DA (J)). The arrow highlights the overall
The p-xylene selectivity as a function of the toluene con-
sumption rate during the toluene methylation at 723 K is shown
in Fig. 1. The rates are normalized to the net weight of the
zeolite phase after modifications. Desilication and subsequent
dealumination increased the toluene consumption rates by 6 effect of the modification.
and 16% relative to parent H-ZSM5, respectively. The fact that
these modifications led to higher reaction rates in parallel to an
increase in the diffusivity of xylenes (Table 1) suggests that the methanol (see Fig. 1, legend). Thus in summary, the SiO2
reaction is diffusion limited (see also ESI† and below). Note that overlayer causes the reduction of the toluene methylation rate
the apparent diffusion coefficients of p-xylene in Table 1 are by a lower coverage of activated methanol and by a lower
underestimated compared to the o-xylene values and were listed probability of toluene methylation in the presence of increasing
only to show the relative trends (see ESI†).
concentrations of bulky aromatic molecules (vice versa for DS
The deposition of a mesoporous SiO2 overlayer on these and DS–DA samples because of the lower concentration of
materials led to a decrease in the toluene consumption rate by bulky aromatic products in the pores).
approximately 12–18%, mainly because the concentration of
Desilication and subsequent dealumination decreased the
Brønsted acid sites decreased by approximately 10% after these p-xylene selectivity (from 64 to 52%), while it increased signifi-
modifications (ESI†). The apparent diffusion coefficient of bulky cantly with the deposition of a SiO2 overlayer (to 78–85%).
aromatic molecules such as o-xylene (Table 1) also decreased During the methylation of toluene, the selectivity of p-xylene is
after deposition of the SiO2 overlayer, leading in turn to a determined by the rates of three different reaction pathways,
significant increase in the concentration of these molecules i.e., toluene methylation, xylene isomerization and dealkylation
in the pores. The higher concentration is deduced from the of highly-methylated aromatic molecules.16 The rate of the first
fact that the intensity of the band of the bridging OH groups at reaction increases with increasing diffusivity, albeit to a relatively
3610 cmÀ1 (Brønsted acid sites) decreased by a factor of two modest degree. As methanol is certainly not diffusion limited, we
more strongly during toluene methylation with the SM materials conclude that its concentration adsorbed on Brønsted acid sites
compared to parent H-ZSM5 (ESI†). The higher concentration changes at least by 25% (from SM to DS–DA samples). Interest-
of more substituted products leads to a higher probability in ingly, the rate of p-xylene formation was constant for all three
the methylation of these aromatic products than of toluene. samples with the SiO2 overlayer (SM, DS–SM and DS–DA–SM) and
Assuming that the coverages are in the linear portion of the the three without (H-ZSM5, DS and DS–DA), with the former
isotherm, the ratio of toluene to all other higher substituted being larger by ca. 30% (Fig. 2). This strongly suggests that the
aromatics decreased from 5/95 with parent H-ZSM515 to changes in the selectivity observed are primarily caused by the
2.5/97.5 for the SM sample. The competition between toluene relative rates of diffusion of the bulkier molecules. The higher
and the various substituted aromatic molecules causes its rate of p-xylene formation with the SiO2 overlayer indicates that
conversion rate to decrease in proportion to its relative concen- the combination of dealkylation and isomerization provides a
tration. The high abundance of bulky aromatic products15 higher contribution to the rate of p-xylene formation in the three
should in turn lead also to a lower fraction of chemisorbed surface modified than the unmodified samples.
methanol (which also should account for a decrease of a factor
In order to determine the significance of these reaction
of two, because of the similar adsorption constants of the two pathways in the xylene selectivities, methanol was co-fed with
reactants) because of competitive adsorption on Brønsted acid toluene (represents toluene methylation) or 1,2,4-trimethylbenzene
sites. This should proportionally reduce the overall rate of (TriMB; represents dealkylation reaction) at low (573 K) and high
methylation to all aromatic species in the pores. Both effects (723 K) reaction temperatures. The product distribution at lower
combined would account for an approximate decrease by a reaction temperature represents the reaction controlled products,
factor of four in the forward rate. The much lower impact of the as the reaction is operated more at the reaction limited regime and
modification on the rate is attributed to the high conversion of vice versa, the diffusion controlled products at high temperatures
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 10584--10586 10585

View Article Online
ChemComm
Communication
that maintain methanol reactivity despite the severe product
diffusion limitations. With the conventional ZSM-5 catalyst, the
rate decreases after deposition of the SiO2 overlayer, because
the higher concentration of bulky aromatic products leads to
the lower availability of acid sites to activate methanol and
because the high abundance of aromatic products lowers the
probability to methylate toluene. As the size of the primary
crystal domains is decreased from 122 to 68 and 45 nm for the
parent, desilicated and subsequently dealuminated samples
(calculated from the diffusion coefficients of o-xylene, Table 1),
respectively, the concentration of the aromatic molecules retained
decreased. This suggests that the relative rates of transport out of
the pores is maintained using the small crystal domains, while the
lower concentration of aromatic molecules in the pores allows for
higher alkylation rates. We would like to emphasize that such
effects can only be obtained for nano-sized zeolite crystals, where
the rate for the pore entrance step is part of the rate determining
step due to the short diffusional path length of the pores and the
significant fraction of the external surface. While the current
system is certainly not yet optimized, it highlights a new strategy
for enhancing the selectivity without losing the catalytic activity
and these modification concepts should be applicable to many
reactions that are diffusion limited and require transport induced
shape selectivity.
Fig. 2 p-Xylene formation rate vs. apparent diffusion coefficient of o-xylene
during toluene methylation at 723 K (ptoluene = 6.0 kPa, pmethanol = 1.5 kPa, 10 mg
of catalyst, C1 (MeOH and dimethyl ether) conversion = 82–89%, total flow rate =
2.3 cm3 sÀ1). The filled symbols represent samples with deposition of the SiO2
overlayer (SM (’), DS–SM (m) and DS–DA–SM (K)) and unfilled without (parent
HZSM5 (&), DS (n) and DS–DA (J)).
The authors are thankful for the financial support from the
King Fahd University of Petroleum and Minerals, the Deutsche
Forschungsgemeinschaft (DFG Project JE 260-10/1) and the
‘‘Fonds der Chemischen Industrie’’ (R.K.). We also acknowl-
´
edge Gary Haller, Mirek Derewinski, Oliver Gutierrez and
Xianyong Sun for helpful technical discussions, Xaver Hecht
for N2 physisorption experiments and Martin Neukamm for
AAS measurements.
Notes and references
Fig. 3 Selectivity of xylene isomers within xylenes at 573 and 723 K with
1 S. Kulprathipanja, Zeolites in Industrial Separation and Catalysis,
Wiley-VCH, Weinheim, 2010.
2 T. Tsai, S. Liu and I. Wang, Appl. Catal., A, 1999, 181, 355–398.
3 W. Vermeiren and J. P. Gilson, Top. Catal., 2009, 52, 1131–1161.
H-ZSM5. The reaction of toluene and methanol (ptoluene = 6.0 kPa, pmethanol
1.5 kPa, 4–10 mg) or with 1,2,4-TriMB and methanol (p1,2,4-TriMB = 1.2 kPa,
pmethanol
0.3 kPa, 9–25 mg) with total flow rates of 1.2–2.3 cm3 sÀ1
=
=
,
C1 (methanol and DME) conversion = 47–55%, toluene conversion = 4–7% and
1,2,4-TriMB conversion = 2–3%. Black, grey and white represent p-, m- and
o-xylene selectivities.
¨
4 G. Ertl, H. Knozinger, F. Schu¨th and J. Weitkamp, Handbook of
Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2008.
5 W. W. Kaeding, C. Chu, L. B. Young, B. Weinstein and S. A. Butter,
J. Catal., 1981, 67, 159–174.
¨
¨
¨
6 H. P. Roger, M. Kramer, K. P. Moller and C. T. O’Connor, Micro-
porous Mesoporous Mater., 1998, 21, 607–614.
as the reaction becomes more diffusion limited. Fig. 3 shows
that reaction controlled products (573 K) were quite different
along these two reaction pathways, i.e., 28 vs. 49% m-xylene
selectivity. Despite these differences, however, similar xylene
selectivities were observed at high reaction temperatures and
this suggests that at relatively high temperatures (723 K) the
xylene selectivity is most significantly controlled by xylene iso-
merization, in agreement with ref. 17.
Thus, we show here that by adjusting the size of the primary
crystal domain via desilication–dealumination and using a
mesoporous overlayer to modify the pore entrance, it is possible
to increase the selectivity to p-xylene considerably (from 64 to
78%) without significant loss in the rate of toluene consump-
tion. This has been made possible because the insight into the
ˇ
ˇ
´
´
7 J. Cejka, N. Zilkova, B. Wichterlova, G. Eder-Mirth and J. A. Lercher,
Zeolites, 1996, 17, 265–271.
8 S. Zheng, H. R. Heydenrych, H. P. Roger, A. Jentys and J. A. Lercher,
Top. Catal., 2003, 22, 101–106.
¨
9 S. Zheng, A. Jentys and J. A. Lercher, J. Catal., 2006, 241, 304–311.
10 O. Mikkelsen, P. O. Rønning and S. Kolboe, Microporous Mesoporous
Mater., 2000, 40, 95–113.
11 Y. S. Bhat, J. Das, K. V. Rao and A. B. Halgeri, J. Catal., 1996, 159,
368–374.
12 Y. Liu, X. Z. Zhou, X. M. Pang, Y. Y. Jin, X. J. Meng, X. M. Zheng,
X. H. Gao and F. S. Xiao, ChemCatChem, 2013, 5, 1517–1523.
13 F. C. Meunier, D. Verboekend, J. P. Gilson, J. C. Groen and J. Perez-
´
´
Ramırez, Microporous Mesoporous Mater., 2012, 148, 115–121.
14 S. J. Reitmeier, O. C. Gobin, A. Jentys and J. A. Lercher, J. Phys. Chem. C,
2009, 113, 15355–15363.
15 G. Mirth and J. A. Lercher, J. Catal., 1991, 132, 244–252.
16 J. H. Ahn, R. Kolvenbach, S. S. Al-Khattaf, A. Jentys and J. A. Lercher,
ACS Catal., 2013, 3, 817–825.
mechanistic details of methylation allowed designing catalysts 17 G. Mirth and J. A. Lercher, J. Catal., 1994, 147, 199–206.
c
10586 Chem. Commun., 2013, 49, 10584--10586
This journal is The Royal Society of Chemistry 2013