Zeolite silylation for the enhancement of para-selectivity in toluene alkylation with ethylene

Article abstract of DOI:10.1135/cccc19970337

The effect of zeolite silylation using tetraethyl and tetramethyl orthosilicate on the enhancement of para-selectivity and of yields of p-ethyltoluene in toluene alkylation with ethylene was investigated for different Si loadings. It was found that zeolite silylation caused a substantial increase in the selectivity and yield of p-ethyltoluene. Data on changes in sorption capacity, in the number of surface and inner strong acid OH groups, in the rate of uptake of p-ethyltoluene together with the catalytic activity of zeolite and its deactivation indicate that the silylation is restricted to surface and subsurface zeolite layers. The enhancement of para-selectivity and yield of p-ethyltoluene are controlled predominantly by changes in geometry of pore openings and thus by transport rates of ethyltoluenes.

Full text of DOI:10.1135/cccc19970337

Zeolite Silylation
337
ZEOLITE SILYLATION FOR THE ENHANCEMENT OF para-SELECTIVITY
IN TOLUENE ALKYLATION WITH ETHYLENE
1,
2
Jan KRTIL, Jiri CEJKA * and Blanka WICHTERLOVA
J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences of the Czech Republic,
182 23 Prague 8, Czech Republic; e-mail: 1 cejka@jh-inst.cas.cz, 2 wichterl@jh-inst.cas.cz
Received October 31, 1996
Accepted November 28, 1996
Dedicated to Dr Karel Mach on the occasion of his 60th birthday.
The effect of zeolite silylation using tetraethyl and tetramethyl orthosilicate on the enhancement of
para-selectivity and of yields of p-ethyltoluene in toluene alkylation with ethylene was investigated
for different Si loadings. It was found that zeolite silylation caused a substantial increase in the se-
lectivity and yield of p-ethyltoluene. Data on changes in sorption capacity, in the number of surface
and inner strong acid OH groups, in the rate of uptake of p-ethyltoluene together with the catalytic
activity of zeolite and its deactivation indicate that the silylation is restricted to surface and subsur-
face zeolite layers. The enhancement of para-selectivity and yield of p-ethyltoluene are controlled
predominantly by changes in geometry of pore openings and thus by transport rates of ethyltoluenes.
Key words: Zeolite silylation; Toluene alkylation; Ethylene; Tetraethyl orthosilicate; Transport of
products; para-Shape-selectivity.
p-Dialkylaromatic hydrocarbons, namely p-xylene, p-ethyltoluene, p-diethylbenzene
and p-propyl- and p-isopropyltoluenes, represent important starting materials for pro-
duction of various polymers, pesticides, fungicides, flavours etc.1–3. When the alkyla-
tion of toluene or ethylbenzene is carried out on amorphous alumosilicates an
enrichment in ortho- and para-isomers in gaseous products, due to the higher reactivity
of the respective positions relative to thermodynamic equilibrium is observed1,3. On the
other hand, when the alkylation takes place in the pores of zeolites of MFI structure
(ZSM-5) possessing medium pores of the size 5.1 × 5.5 and 5.4 × 5.6 Å in three-dimen-
sional structure, high selectivity with respect to para-isomers is achieved1,4,5
.
It was found, by 13C MAS NMR study that in alkylation of toluene with methanol to
xylenes the alkylation step inside the ZSM-5 zeolite pores results, in preferential formation
of p- and o-xylenes6, similarly as on an open reaction space of the amorphous alumo-
* The author to whom correspondence should be addressed.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

338
Krtil, Cejka, Wichterlova:
silicates. Moreover, “in situ” FTIR studies of the alkylation of toluene with methanol to
xylenes in ZSM-5 zeolite indicated an enrichment of ortho- and particularly meta-
isomers in the adsorbed phase, along with a considerable enrichment of the para-
isomer among the gaseous products7,8. It implies that after the alkylation step of the
reaction there proceeds a positional isomerization of xylenes, followed by a fast release
of the para-isomer from the zeolite inner volume into the gaseous phase. It means that
the resulting product composition is predominantly affected by differences in the rate
of diffusion of individual dialkylbenzene isomers. It was found that the ratio of diffusi-
vities of individual xylene isomers in the temperature range of 373–473 K is about p- :
o- : m- = 1 000 : 10 : 1 (ref.7).
Although para-selectivity can be improved by reaction conditions, its main enhance-
ment can be achieved by tailoring the crystal structure of zeolite catalysts. Employment
of large crystals of ZSM-5 or of zeolites with narrow pore diameter (as ZSM-22 with
pore sizes – 4.5 × 5.5 Å, ZSM-23 – 4.5 × 5.6 Å, ZSM-48 – 5.3 × 5.6 Å) yields higher
selectivities to para-isomers9,10. However, these zeolites are less resistant against deac-
tivation by coke, formed via subsequent alkylation, isomerization, transalkylation and
hydrogen transfer reactions leading to polyaromatic hydrocarbons. These side reactions
result from a longer contact time in the narrow or long pores particularly of bulkier
molecules, which are released with difficulty into the gaseous phase.
To improve zeolite para-selectivity, it was also suggested to block the acid sites
located on the surface of zeolite crystals, on which the alkylation reaction (as well as
isomerization of para-isomers leaving the inner zeolite pores) can proceed without any
space limitations. Poisoning of the surface acid sites by bulky strong bases (e.g. by
quinoline) which cannot enter zeolite pores was employed and resulted in an increase
of para-selectivity11,12. However, organic bases are easily desorbed into a reactant
stream during the reaction time-on-stream, and thus this procedure is not applicable in
a large scale. Another approach which can block the surface acid sites is their silylation
by bulky organosilicates, which cannot enter the zeolite pores and thus block the inner
acid sites1,13,14. There are several papers concerned with application of this procedure
either in liquid15 or gaseous phase (so called chemical vapour deposition, CVD)16–18
,
resulting in an enhancement of zeolite para-selectivity.
In this paper the liquid phase silylation of the crystal surface of ZSM-5 zeolites by
different organosilicates is described in detail, and its impact on the number of acid
sites on the zeolite surface and in its pores, pore volume modification, uptake rate of
p-ethyltoluene (p-ET), and on the function of zeolite catalyst in alkylation of toluene
with ethylene to ethyltoluenes is reported.
EXPERIMENTAL
The ZSM-5 zeolite with Si/Al ratio 22.5 was supplied by the Research Institute for Oil and Hydro-
carbon Gases, Bratislava. The acid form of this zeolite was prepared by ion exchange of its sodium
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Zeolite Silylation
339
form with 0.5 M nitric acid. The silylation procedure was following: ZSM-5 zeolite was calcined at
770 K in a dry oxygen stream for 2 h. Then the sample was cooled to ambient temperature in a dry
nitrogen, a solvent (usually hexane or toluene) was added followed by a known amount of silylation
agent. Tetramethyl- (TMOS) – ZSM-5/B1-B4 and tetraethyl- (TEOS) – ZSM-5/A1-A7 orthosilicates
(Fluka) were used (Tables I–IV). All these steps were carried out with exclusion of air moisture. The
resultant mixture was stirred for 4 h at ambient temperature and than for 7 h at 390 K. The mother
liquid was then evaporated in vacuum, and the silylated sample was calcined in an oxygen stream at
770 K for 8 h. X-Ray diffraction and IR spectra of zeolite skeletal vibrations indicated that silylation
does not change the structure of ZSM-5 zeolite.
The zeolite inner void volume was estimated from the sorption capacity for argon measured in a
static apparatus at 13.33 kPa of argon pressure at 80 K on samples pretreated at 720 K in a vacuum
of 10–4 kPa for 16 h. The number of strong acid sites was calculated from the high temperature peak
of the temperature-programmed desorption (20 K/min) of ammonia (TPDA) performed in a helium
stream (5 ml/s) on zeolites (0.3 g) pretreated at 770 K for 1 h. The zeolites were equilibrated in an
ammonia stream (10% ammonia in nitrogen) at 373 K for 0.5 h and then in a helium stream at the
same temperature for 0.5 h.
The number of strong acid bridging Si–OH–Al groups was also determined by FTIR spectra of
hydroxyl region (Nicolet, Magna 550). Parent H-ZSM-5 zeolite exhibited 0.69 mmol/g of strong acid
groups estimated from the high-temperature peak of TPD of ammonia, and the intensity of IR band
at 3 612 cm–1 corresponding to structural bridging Si–OH–Al groups. The number of strong Lewis
sites estimated from pyridine adsorption (IR band at 1 445 cm–1) was lower than 5% relative to the
number of bridging OH groups. In addition, the zeolite contained also some terminal Si–OH groups
(IR band at 3 745 cm–1), roughly 10% of the bridging OH groups, of low acidity, which do not take
part in acid-catalyzed reactions. All the zeolite characteristics including toluene conversions given in
Tables I–IV are related to the weight of parent ZSM-5 zeolite and not to the weight of the catalyst.
TABLE I
Characteristics of ZSM-5 zeolite (Si/Al = 22.5) modified with TEOS
Sorption capacity (mmol g–1) for
Si–OH–Al
groupsa
TIPB
conversionb
%
Si added
wt.%
p-ET
rel. rate
Zeolite
mmol g–1
argonc
p-ETc
H-ZSM-5
0
0.69
0.57
0.67
0.63
0.66
0.63
0.65
0.74
82.5
5.26
4.77
5.05
4.91
5.08
4.84
4.85
4.46
0.63
0.58
0.62
0.61
0.61
0.59
0.62
0.60
0.27
0.20
0.22
0.23
0.23
0.22
0.20
0.13
H-ZSM-5/A1
H-ZSM-5/A2
H-ZSM-5/A3
H-ZSM-5/A4
H-ZSM-5/A5
H-ZSM-5/A6
H-ZSM-5/A7
0.1
0.5
1.0
1.5
2.0
4.3
8.4
61.5
d
d
d
55.3
20.0
<1.0
a
b
c
The number of OH groups determined from TPD of ammonia; after 15 min of T-O-S; sorption
capacities for argon and p-ET are related to 1 g of parent zeolite (not to the weight of silylated ze-
d
olite); not determined.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

340
Krtil, Cejka, Wichterlova:
The sorption of p-ethyltoluene (Fluka) over parent and silylated samples was performed to follow
the effect of silylation on transport rates of p-ethyltoluene. Zeolite sample was evacuated in vacuum
of 1.3 . 10–6 kPa for 3 h at 700 K. The adsorption of p-ethyltoluene was carried out at 373 K at
p-ethyltoluene partial pressure of 85 Pa. The adsorption was followed at least 6 h to reach the equili-
brium amount of adsorbed p-ethyltoluene (p-ET). Time intervals of 1–5 min were used to estimate
the initial rate of the adsorption (Tables I and III). Within this period no isomerization of p-ET to
meta-isomer occurred. The uptake rate was measured gravimetrically and calculated according to the
following equation
r5–1 = (Q5 – Q1)/Q∞
.
(1)
Q1 and Q5 denote the uptakes at given time (in min), Q∞ the uptake after equilibrium.
A relative number of external surface strong acid sites of ZSM-5 crystals was checked by the de-
composition of 1,3,5-triisopropylbenzene (TIPB, Fluka), the kinetic diameter of which is too large to
penetrate inside the ZSM-5 channel system. TIPB was charged into a stream of dry nitrogen via syr-
inge pump at the concentration of 1.0 vol.%, at temperature of 570 K and WHSV of 1.0 h–1. The first
analysis of the reaction products was made after 15 min of time-on-stream (T-O-S), similarly as in
the case of alkylation and reaction.
The toluene alkylation with ethylene was carried out in a vapour phase continuous down-flow
microreactor at atmospheric pressure. Nitrogen, used as a carrier gas, was saturated with toluene at
334 K to perform kinetic runs at toluene concentration of 18.5 vol.% and WHSV of 10.0 h–1. The
stream of nitrogen and toluene was mixed with a stream of ethylene to reach the toluene to ethylene
molar ratio of 3.8. A catalyst (weight of 0.1–0.4 g) in a granulated form (0.3–0.7 mm) was located
in a microreactor (i.d. 10 mm) and pretreated in an oxygen stream at 770 K for 1 h. The reaction
TABLE II
Toluene conversion and p-ET selectivity in toluene alkylation with ethylene in dependence on T-O-S
for H-ZSM-5 (Si/Al = 22.5) modified with TEOS
Toluene
p-ET
Toluene
p-ET
Si added
wt.%
p-ET yielda
p-ET yieldb
Zeolite
conversiona selectivitya
conversionb selectivityb
%
%
%
vol.%
%
vol.%
H-ZSM-5
0
96.6
74.3
77.9
74.5
74.9
72.4
75.4
59.8
33.4
52.6
45.6
74.8
82.3
93.7
94.8
95.3
32.3
39.1
35.5
55.7
61.6
67.8
71.5
56.7
93.4
74.5
78.2
73.3
65.9
54.4
43.8
38.4
34.0
56.3
52.9
83.4
86.8
95.8
96.0
97.1
31.7
41.9
41.4
61.1
57.2
52.1
42.0
37.3
H-ZSM-5/A1
H-ZSM-5/A2
H-ZSM-5/A3
H-ZSM-5/A4
H-ZSM-5/A5
H-ZSM-5/A6
H-ZSM-5/A7
0.1
0.5
1.0
1.5
2.0
4.3
8.4
a
b
After 15 min of T-O-S; after 135 min of T-O-S.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Zeolite Silylation
341
products were analyzed by an “on-line” gas chromatograph (Hewlett–Packard, 5890 Series II),
equipped with mass spectrometric (5971A) and flame ionization detectors, using Supelcowax 10 (Su-
pelco) high resolution capillary column (30 m length, i.d. 0.2 mm, phase thickness 0.2 µm).
RESULTS AND DISCUSSIONS
The silylation, depending on the amount and type of silylating agent (TEOS, TMOS),
was accompanied by changes in the overall number of strong acid sites (inner and on
external surface), in sorption capacity of p-ET, an inner zeolite volume (reflected in the
sorption capacity of argon) and in relative number of strong acid sites located on the
surface of silylated zeolite crystals with respect to the parent zeolite, as estimated by
the degree of dealkylation of 1,3,5-triisopropyltoluene (see Tables I and III).
It was found that with increasing amount of TEOS added to the zeolite, the number
of strong acid OH groups as well as the sorption capacity of p-ET decreases by 5–7
rel.%. The same trend was also observed for adsorbed argon. Similarly to increasing
amount of Si in the zeolite, the sorption capacity and relative rate of p-ET uptake
decrease by ca 10–20%. Interestingly, all these values are lower compared to the parent
zeolite, but do not differ substantially in increasing Si loading (with the exception of
sample ZSM-5/A7). The exception in the series is the zeolite silylated with the Si
amount equal to 0.1 wt.%, which shows greater decrease in values for all the zeolite
characteristics studied.
Similar results have been obtained also for the series of silylated zeolites when
TMOS was employed. However, the higher decrease in all the above characteristics
was observed for zeolite with added 0.1 and 1.0 wt.% Si, while at the higher Si loadings
(2.0 and 4.3 wt.% Si), comparable values of the corresponding characteristics of the
zeolite silylated with TEOS (at loadings higher than 1.0 wt.% Si) were found.
TABLE III
Characteristics of ZSM-5 zeolite (Si/Al = 22.5) modified with TMOS
Sorption capacity (mmol g–1) for
Si–OH–Al
groups
TIPB
conversiona
%
Si added
wt.%
p-ET
rel. rate
Zeolite
mmol g–1
argonc
p-ET
H-ZSM-5
0
0.69
0.52
0.59
0.66
0.58
82.5
b
5.26
4.78
4.79
5.17
5.05
0.63
0.53
0.61
0.63
0.64
0.27
0.19
0.17
0.22
0.22
H-ZSM-5/B1
H-ZSM-5/B2
H-ZSM-5/B3
H-ZSM-5/B4
0.1
1.0
2.0
4.3
b
81.6
60.7
a
b
After 15 min of T-O-S; not determined.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

342
Krtil, Cejka, Wichterlova:
Surprisingly, the conversion of triisopropylbenzene (TIPB) indicated that not all
strong acid bridging OH groups on the surface of the zeolite crystals were consumed by
silylation up to such high loadings as 4.3 wt.% of Si. This implies that hydrolysis of
TEOS or TMOS does not take place preferentially on the strong acid OH groups, but
also low acid (Si–OH) sites are involved. Moreover, formation of some SiO2 phase
outside the zeolite crystals can be expected.
Toluene conversion and selectivity to p-ethyltoluene (together with p-ET yield) in
toluene alkylation with ethylene at times-on-stream of 15 and 135 min (to estimate
zeolite deactivation) are compared for parent samples and the zeolites silylated by
TEOS and TMOS in Tables II and IV, respectively. It is seen that there is a dramatic
decrease in conversion (compared to the parent zeolite) for 0.1 wt.% of added Si for
TEOS and 0.1–1.0 wt.% of added Si for TMOS. Then, depending on Si loading, the
conversion slightly increases (0.5 wt.% Si for TEOS and 2.0 wt.% for TMOS), being
followed by further conversion decrease with increasing Si loading, particularly at the
highest value of Si loading (H-ZSM-5/A7). For all the zeolites investigated, some deac-
tivation in T-O-S took place which was caused by zeolite coking. It was much higher
compared to the parent zeolite for Si loadings above 1.5 wt.% Si (TEOS), but for the
zeolites silylated by TMOS the deactivation was nearly not observed. However, some
coking did take place, as the selectivity and yield were higher at T-O-S for 135 min in
comparison with those for 15 min.
The p-ET yields together with conversions within T-O-S for 135 min indicate that for
the zeolites silylated by TEOS up to ca 1.5 wt.% Si no dramatic deactivation occurs,
and the yields of p-ET are substantially increased. It is necessary to note that with Si
loadings higher than 2.0 wt.%, a comparison of selectivity (yield) for the silylated zeo-
lites is not straightforward as the selectivity reaches values between 95 and 100%, and
TABLE IV
Toluene conversion and p-ET selectivity in toluene alkylation with ethylene depending on T-O-S for
H-ZSM-5 (Si/Al = 22.5) modified with TMOS
Toluene
p-ET
Toluene
p-ET
Si added
wt.%
p-ET yielda
p-ET yieldb
Zeolite
conversiona selectivitya
conversionb selectivityb
%
%
%
vol.%
%
vol.%
H-ZSM-5
0
96.6
70.3
64.2
79.7
75.0
33.4
55.0
80.1
54.9
76.6
32.3
38.7
51.4
43.8
57.5
93.4
71.6
64.9
80.1
74.6
34.0
61.9
85.2
65.7
85.2
31.8
44.3
55.3
52.6
63.6
H-ZSM-5/B1
H-ZSM-5/B2
H-ZSM-5/B3
H-ZSM-5/B4
0.1
1.0
2.0
4.3
a
b
After 15 min of T-O-S; after 135 min of T-O-S.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Zeolite Silylation
343
therefore it is no more sensitive to further changes of this property. For the zeolites
silylated by TMOS also a decrease in conversion compared to the parent zeolite was
observed. However, the increase in the p-ET yield at the same amount of added Si was
lower compared to the zeolites silylated by TEOS.
It follows that silylation of H-ZSM-5 zeolite results in a substantial increase in the
selectivity and the yield of p-ET. Its yields increase with Si loadings. Simultaneously,
the decrease in conversion and particularly the increase in the zeolite deactivation by
coke at higher Si loadings indicate changes in the number of strong acid sites and/or
considerable changes in pore openings.
For the zeolites silylated by TEOS and TMOS (0.1 and 0.1–1.0 wt.% Si, respec-
tively) a decrease in overall number of strong acid sites is unexpectedly high and, by
contrast, a decrease in the number of surface acid centres is very low (cf. Tables I and III).
The number of surface sites for crystal sizes around 1 µm was estimated not to be
greater than 5% of the total number of strong acid Si–OH–Al groups. This indicates that
silylation inside pores has to take place. As the kinetic diameter of TEOS and TMOS
molecules (ca 8.9 Å) is greater compared to channel openings (5.1 × 5.5 and 5.4 × 5.6 Å
for ZSM-5) it is supposed that dealkylation of silylating agents proceeds to some ex-
tent, and the rest of molecules penetrate inside the pores particularly during the zeolite
calcination. Also the sorption capacities for argon and p-ET confirm this suggestion.
On the other hand, at higher loadings of Si it is supposed that silylation proceeds
exclusively on the external surface or subsurface layers, and the interior of the channels
is not affected. The total number of strong acid OH groups does not change signifi-
cantly by the zeolite silylation with TEOS (above 0.1 wt.% Si) and TMOS (above 1.0
wt.% Si). This follows from their estimation by ammonia desorption and equilibrium
sorption of p-ET, which both reflect the number of strong acid sites. Similarly, the
sorption capacity for argon indicates that the inner volume of zeolite channels is not
significantly decreased by Si.
There are different views in the literature on the effect of silylation on the zeolite
structure and catalytic activity. From one side it is considered that surface strong acid
Si–OH–Al sites, which are assumed to play a dominant (negative) role in para-selecti-
vity, are blocked by silylation15. The others prefer the effect of narrowing of the chan-
nel pore openings by silylation16–18, and, thus the changes in diffusivities, or better,
increased differences among transport rates of para-, meta- and ortho-isomers.
In our previous paper19 it has been shown that the surface sites can hardly play a key
role in the zeolite para-selectivity as the “coked” zeolite possessing no free surface
sites (evidenced by a negligible TIPB conversion) did not exhibit higher p-ET yield
compared to the parent zeolite. It is reported here that with silylated zeolites a substan-
tial increase in both the selectivity and yield to p-ET were obtained although the
samples contained a high number of strong acid sites on the zeolite surface, reflected in
a high degree of dealkylation of TIPB relative to the parent zeolite (cf. Table II, 0.1–2.0
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

344
Krtil, Cejka, Wichterlova:
wt.% Si added and Table IV, whole range of Si added). This indicates that surface sites
do not contribute substantially to the isomerization of p- to m-ethyltoluene or to nonse-
lective alkylation. Therefore, their function in the catalytic reaction can hardly be domi-
nant.
On the other hand, the conversion and selectivity data indicate that the reaction is
affected by transport processes which undergo changes by silylation. Both the selectiv-
ity and yield to p-ET are dramatically improved, however, at very high Si loadings
zeolite deactivation by coke takes place. We have shown by FTIR spectroscopy that at
higher loadings of Si the diffusivity of the bulkier ethyltoluene molecules is decreased,
and thus significantly longer contact times and higher coverage of strong acid bridging
OH groups inside the zeolite channels by ortho- and meta-isomers are obtained20. This
results in the enhancement of the rate of isomerization reflected in a higher para-selec-
tivity and in the decreased rate of the alkylation, because a significant part of catalyti-
cally active bridging OH groups is blocked by adsorbed ethyltoluene molecules. The
application of TMOS has been shown to provide silylated zeolites which exhibited
better stability of toluene conversion with T-O-S without substantial deactivation and
producing high yields of p-ET. Zeolites silylated with TEOS reached at higher Si load-
ings the selectivity to p-ET above 90%. However, they exhibited faster deactivation. As
surface strong acid sites do not affect significantly the catalyst selectivity, the contribu-
tion of transport properties changed by zeolite silylation is suggested to play a domi-
nant role for both zeolite para-selectivity and stability in conversion.
There is still lack of information on the structure of grown SiO2 phase on the zeolite.
Some information concerning the silylation of silica with methyl-(trimethoxy)silane
was already reported21,22, suggesting mono- and bifunctional interaction with Si–OH
groups of silica. This is in line with our assumption of the simultaneous interaction of
TMOS or TEOS molecules with both strong acid Si–OH–Al groups and terminal Si–OH
groups. The presence of a considerable number of strong acid sites on the surface of
silylated zeolites indicates that TEOS and TMOS do not interact preferentially with
these sites, but their hydrolysis takes place also via terminal Si–OH groups. Further
step can be polymerization of the alkoxide. On the other hand, TMOS preferable inter-
acts with surface OH groups and its hydrolysis is faster than that of TEOS, which
prevent formation of polymeric layer. The kinetic diameter of TMOS molecule is about
8.9 Å (ref.23) being significantly larger compared to ZSM-5 channel openings. How-
ever, during the subsequent calcination at 770 K, methyl and ethyl groups can be deal-
kylated, and some SiO2 particles can penetrate into the inner volume. This process
seems to occur with low concentration of TEOS (0.1 wt.% Si) and TMOS (0.1–1.0
wt.% Si), which is reflected in a decrease in the number of OH groups, sorption capa-
cities, the rate of p-ET adsorption as well as in the toluene conversion (cf. Tables I–IV).
Although Niwa et al.23 suggested the use of TMOS for fine control of narrowing of
ZSM-5 channel openings, according to our results we prefer the blocking or plugging
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Zeolite Silylation
345
of zeolite openings and prolonging, thus, diffusion pathways inside the zeolite crystals.
However, at higher concentrations of alkoxides the polymerization is probably much
faster and only surface, and subsurface zeolite layers are attacked during the silylation
process.
CONCLUSIONS
The results of ZSM-5 zeolite silylation with tetraethyl and tetramethyl orthosilicates
can be summarized as follows:
1. A significant increase in p-ET selectivity and yield of ZSM-5 zeolite has been
found after the zeolite silylation.
2. This increase in p-ET selectivity is probably caused by differences in diffusivities
of individual ethyltoluene isomers. Moreover, after silylation the bulkier isomers spend
longer time inside the zeolite pores and can isomerize to p-ET, increasing thus the p-ET
selectivity.
3. A higher p-ET selectivity and p-ET yield were found with silylation in the range
of Si loadings from 0.1 to 4.3 wt.%, despite the presence of a substantial amount of
active sites on the “external” surface of zeolite crystals. This indicates that the blocking
of active sites can hardly play a dominant role in the enhancement of the para-selecti-
vity.
4. It can be assumed that under silylation conditions tetramethyl orthosilicate tends
to hydrolysis much more easily than tetraethyl orthosilicate exhibiting a higher tend-
ency to polymerization. This results in penetration of individual SiO2 species into the
inner zeolite volume during the calcination, which is particularly evident at low Si
loading.
5. The detailed investigation of the effect of zeolite silylation on the enhancement of
para-selectivity confirms our previous results emphasizing the positive role of diffusiv-
ity on the para-selectivity. However, a mutual effect of blocking of active sites on the
“external” surface cannot be fully excluded.
This work was supported by the Academy of Sciences of the Czech Republic (Grant No. A4040707).
REFERENCES
1. Kaeding W. W., Young L. B., Chu C.-C.: J. Catal. 89, 267 (1984).
2. Paparatto G., Moretti E., Leofanti G., Gatti F.: J. Catal. 105, 227 (1987).
3. Fraenkel D., Levy M.: J. Catal. 118, 10 (1989).
4. Cejka J., Bednarova S., Wichterlova B.: Appl. Catal., A 79, 215 (1991).
5. Wichterlova B., Cejka J.: Catal. Lett. 16, 421 (1992).
6. Ivanova I., Corma A.: Stud. Surf. Sci. Catal. 87, 27 (1995).
7. Mirth G., Cejka J., Lercher J. A.: J. Catal. 139, 24 (1993).
8. Mirth G., Cejka J., Nusterer E., Lercher J. A.: Stud. Surf. Sci. Catal. 83, 287 (1994).
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

346
Krtil, Cejka, Wichterlova:
9. Kumar R., Ratnasamy P.: J. Catal. 116, 440 (1989).
10. Kumar R., Ratnasamy P.: J. Catal. 118, 68 (1989).
11. Nunan J., Cronin J., Cunningham J.: J. Catal. 87, 77 (1984).
12. Lonyi F., Engelhardt J., Kallo D.: Stud. Surf. Sci. Catal. 49, 1357 (1989).
13. Kaeding W. W., Chu C., Young L. B., Weinstein B., Butter S. A.: J. Catal. 67, 159 (1981).
14. Kaeding W. W.: J. Catal. 95, 512 (1985).
15. Chen N. Y.: J. Catal. 114, 17 (1988).
16. Hibino T., Niwa M., Kawashima Y., Murakami Y.: Stud. Surf. Sci. Catal. 60, 151 (1991).
17. Hibino T., Niwa M., Murakami Y.: J. Catal. 128, 551 (1991).
18. Hibino T., Niwa M., Murakami Y.: Zeolites 13, 518 (1993).
19. Cejka J., Zilkova N., Wichterlova B., Mirth-Eder G., Lercher J. A.: Zeolites 17, 265 (1996).
20. Mirth G., Cejka J., Krtil J., Lercher J. A.: Stud. Surf. Sci. Catal. 88, 241 (1994).
21. Hertl W.: J. Phys. Chem. 72, 1248 (1968).
22. Hertl W.: J. Phys. Chem. 72, 3993 (1968).
23. Niwa M., Kato M., Hattori T., Murakami Y.: J. Phys. Chem. 90, 6233 (1986).
Collect. Czech. Chem. Commun. (Vol. 62) (1997)