Effects of crystallinity of ZSM-5 zeolite on para-selective tert-butylation of ethylbenzene

Article abstract of DOI:10.1016/S1872-2067(11)60482-8

Highly crystalline ZSM-5 zeolites are important for para-selective alkylation of alkyl aromatics, because they carry few external acid sites for isomerization of p-dialkyl products. Such zeolites (Si/Al = 25, 50, and 75) were synthesized in a fluoride medium between pH 4 and 6. Their crystallinities, crystal sizes, and surface areas were higher than those of a commercial ZSM-5 zeolite. Their para selectivities in alkylation were tested for vapor-phase tert-butylation of ethylbenzene between 200 and 400 °C. As expected, all the catalysts showed more than 90% para selectivity. At 300 °C, ethylbenzene conversion decreased in the order ZSM-5(25, commercial) > ZSM-5(25) > ZSM-5(50) > ZSM-5(75). The catalysts had weak, medium, and strong acid sites, but all the acid sites of ZSM-5(75) were weaker than those of ZSM-5(25) and ZSM-5(50). The high activity of commercial ZSM-5 was caused by its strong acid sites being stronger than those of the synthesized zeolites. Although the activity of the commercial catalyst was higher than those of the present catalysts, the selectivity for 4-t-butylethylbenzene (4-t-BEB) was low. The optimum feed ratio (ethylbenzene:t-butyl alcohol) was 2:1 and the feed rate was 1.65 h^-1 for high ethylbenzene conversion and 4-t-BEB selectivity. Time-on-stream studies showed slow catalyst deactivation. Highly crystalline ZSM-5 zeolites are therefore better than a commercial zeolite for para-selective alkylation of alkyl aromatics. They do not require much post-modification for high para selectivity. A fluoride medium is therefore better than an alkaline medium for obtaining highly crystalline para-selective ZSM-5 zeolites.

Full text of DOI:10.1016/S1872-2067(11)60482-8

Chinese Journal of Catalysis 34 (2013) 294–304
ava i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e /c h n j c
Article
Effects of crystallinity of ZSM‐5 zeolite on para‐selective
tert‐butylation of ethylbenzene
PUSHPARAJ Hemalathaa, MANI Ganesha, MUTHIAHPILLAI Palanichamya,
VELAYUTHAM Murugesanb, PARK Yong‐Kic, CHOI Won Choonc, JANG Hyun Taea,*
a Department of Chemical Engineering, Hanseo University, Seosan‐si 356 706, South Korea
b Department of Chemistry, Anna University, Chennai 600 025, India
c Green Chemistry Research Division, Korea Research Institute of Chemical Technology, Jang‐dong 100, Daejeon 305 600, South Korea
A R T I C L E I N F O
A B S T R A C T
Article history:
Highly crystalline ZSM‐5 zeolites are important for para‐selective alkylation of alkyl aromatics,
because they carry few external acid sites for isomerization of p‐dialkyl products. Such zeolites
(Si/Al = 25, 50, and 75) were synthesized in a fluoride medium between pH 4 and 6. Their crystal‐
linities, crystal sizes, and surface areas were higher than those of a commercial ZSM‐5 zeolite. Their
para selectivities in alkylation were tested for vapor‐phase tert‐butylation of ethylbenzene between
200 and 400 °C. As expected, all the catalysts showed more than 90% para selectivity. At 300 °C,
ethylbenzene conversion decreased in the order ZSM‐5(25, commercial) > ZSM‐5(25) > ZSM‐5(50)
> ZSM‐5(75). The catalysts had weak, medium, and strong acid sites, but all the acid sites of
ZSM‐5(75) were weaker than those of ZSM‐5(25) and ZSM‐5(50). The high activity of commercial
ZSM‐5 was caused by its strong acid sites being stronger than those of the synthesized zeolites.
Although the activity of the commercial catalyst was higher than those of the present catalysts, the
selectivity for 4‐t‐butylethylbenzene (4‐t‐BEB) was low. The optimum feed ratio (ethylben‐
zene:t‐butyl alcohol) was 2:1 and the feed rate was 1.65 h−1 for high ethylbenzene conversion and
4‐t‐BEB selectivity. Time‐on‐stream studies showed slow catalyst deactivation. Highly crystalline
ZSM‐5 zeolites are therefore better than a commercial zeolite for para‐selective alkylation of alkyl
aromatics. They do not require much post‐modification for high para selectivity. A fluoride medium
is therefore better than an alkaline medium for obtaining highly crystalline para‐selective ZSM‐5
zeolites.
Received 18 July 2012
Accepted 13 November 2012
Published 20 February 2013
Keywords:
ZSM‐5
Fluoride medium
P‐selectivity
Ethylbenzene
Tert‐butylation
Zeolite
© 2013, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
Published by Elsevier B.V. All rights reserved.
1. Introduction
produce p‐methylstyrene, the monomer for poly(p‐methyl‐
styrene) [4,5]; p‐diethylbenzene is used to produce divi‐
The alkylation of alkyl aromatics to dialkylated compounds
via simple alkylation, trans alkylation, or disproportionation is
commercially important [1], because of the uses of the dialkyl
products; for example, p‐xylene is used to obtain terephthalic
acid, a monomer for terylene [2,3]; p‐ethyltoluene is used to
nylbenzene, a monomer for cross‐linked polystyrene [6]; and
other p‐alkylated ethylbenzenes are used to produce
p‐alkylated styrenes, the monomers for poly(p‐alkylated sty‐
renes), which have high glass‐transition temperatures (Tg)
compared with polystyrene [7]. Zeolites have been widely used
* Corresponding author. Tel.: +82‐41‐6601423; Fax: +82‐41‐6881343; E‐mail: htjang@hanseo.ac.kr
This work was supported by grants from Korea CCS R&D Centre, funded by the Ministry of Education, Science and Technology of the Korean Gov‐
ernment.
DOI: 10.1016/S1872‐2067(11)60482‐8 | http://www.sciencedirect.com/science/journal/18722067 | Chin. J. Catal., Vol. 34, No. 2, February 2013

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
as catalysts for alkylation of alkyl aromatics [8]. Dialkylation
Tetraethyl orthosilicate (TEOS; 98%), tetrapropylammo‐
nium bromide (98%), aluminum sulfate octadecahydrate
(98%), ammonia (28%), ammonium fluoride (99%), and so‐
dium aluminate (50–56% Al2O3) (all obtained from Aldrich), as
well as orthophosphoric acid (85%) and hydrofluoric acid
(48%) (both obtained from Merck), were used in the syntheses.
LUDOX‐ammonium‐stabilized colloidal silica (AS silica; 40 wt%
suspension in water), LUDOX‐sodium‐stabilized colloidal silica
(HS silica; 40 wt% suspension in water), fumed silica, and so‐
dium silicate (~26.5% SiO2) (all obtained from Aldrich) were
used as silicon sources in the syntheses. Potassium hydroxide
(90%; Aldrich) was used in potentiometric titrations. The
chemicals used in the catalytic reactions, i.e., ethylbenzene and
t‐butyl alcohol, were of Merck grade.
was reported over X and Y zeolites partly loaded with rare
earths [9]. In many alkylation processes, the meta isomer, ra‐
ther than the ortho and para isomers, is formed, but pa‐
ra‐selective alkylation has been attempted in most alkylations.
para‐Selective alkylation was first demonstrated by Mobil re‐
searchers for the production of p‐xylene [10,11].
Generally, in an alkylation, alkyl cations attack the 2‐, 4‐, and
6‐positions of alkyl aromatics, as alkyl groups are ortho‐ and
para‐directing. If the alkyl groups are bulky, para substitution
becomes predominant over ortho substitution. However, at
high temperatures, 1,4‐dialkyl products isomerize to
1,3‐dialkyl products on the Brönsted acid sites of zeolites. Since
isomerization requires a large amount of space, it is more prev‐
alent in large‐pore zeolites. External acid sites also contribute
to isomerization and such sites arise as a result of low crystal‐
linity. Hence, for para‐selective alkylation, a highly crystalline,
medium‐pore ZSM‐5 zeolite is a suitable choice. It can provide
selective constrained access to, and egress from, the in‐
tra‐crystalline free spaces to molecules of size less than the
pore size of about 0.55 nm. It is a unique characteristic of
ZSM‐5 to exhibit para selectivity.
For commercial ZSM‐5 zeolites, various modification meth‐
ods have been reported [12–14]. In all of these methods, pas‐
sivation of the external surface acidity and narrowing of the
pore openings [15] were used to enhance the para selectivity.
However, such modifications drastically reduced conversion. In
this context, ZSM‐5 zeolites synthesized in fluoride media are
convenient substitutes for commercial zeolites. Such zeolites
are large crystals, have high crystallinity, and are defect‐free,
with high surface areas [16–18]. High crystallinity and large
crystal size enhance para selectivity, and an inverse relation‐
ship between the rate of isomerization and crystal radius has
been reported in the literature [19–21].
In this study, tert‐butylation of ethylbenzene was carried
out in the vapor phase over ZSM‐5 catalysts synthesized in a
fluoride medium. 4‐tert‐Butylation of ethylbenzene is im‐
portant as its dehydrogenated product, 4‐t‐butylvinylbenzene
(4‐t‐BVB), is an important monomer for the commercial pro‐
duction of poly(4‐t‐butylstyrene), which has a high Tg value.
The same reaction was also studied over Al‐MCM‐41, and high
para selectivity was reported [22]. Bidart et al. [23] studied
tert‐butylation of ethylbenzene over FeY zeolite and reported
mainly monoalkylation. There have been no reports of the
same reaction over ZSM‐5 synthesized in a fluoride medium.
A major stumbling block in the synthesis of ZSM‐5 in a fluo‐
ride medium is the reduced amount of aluminum transport
from the gel to the framework as a result of the formation of
AlFx species. In our laboratory, this problem was solved by us‐
ing phosphate as a co‐complexant to fluoride. Syntheses of
ZSM‐5 were also carried out using different silicon sources to
select the best source for high crystallinity.
2.2. Optimization of fluoride content for ZSM‐5 syntheses
The following gel composition was used [16] for the ZSM‐5
syntheses in a fluoride medium: 0.07 C12H28NBr:1.0 SiO2:0.006
Al2O3:x NH4F:80 H2O.
The syntheses were carried out initially with TEOS and so‐
dium aluminate as the sources for silicon and aluminum, re‐
spectively. Tetrapropylammonium bromide was used as a
structure‐directing template. To optimize the fluoride content
for the syntheses, the F/Si ratio was altered using NH4F; the
weights are given in Table 1.
In a typical synthesis, sodium aluminate (0.022 g), an ap‐
propriate amount of ammonium fluoride, tetrapropylammo‐
nium bromide (0.462 g), and water (35 g) were mixed in a 100
ml beaker and stirred for 5 h. TEOS (5.16 g) was added
drop‐wise. Stirring was continued for 3 h. Finally, the pH of the
mixture was reduced to 7.0 by adding two or three drops of
hydrofluoric acid, and stirring was continued for 1 h. The re‐
sulting gel was autoclaved and hydrothermal crystallization
was carried out at 170 °C for 5 d in a static oven. The crystal‐
lized product was recovered by filtration, washed repeatedly
with water, dried at 110 °C in the oven for 12 h, and calcined at
650 °C in air for 8 h.
2.3. Syntheses of ZSM‐5 in fluoride medium with phosphate
Sodium aluminate was replaced by aluminum sulfate.
Phosphate was used as well as fluoride. Tetrapropylammonium
bromide was used as a structure‐directing agent and the gel
composition was as follows: 0.07 C12H28NBr:1 SiO2:0.02
Al2O3:1.3 NH4F:0.02 P2O5:80 H2O.
In a typical synthesis, aluminum sulfate (0.330 g), ortho‐
phosphoric acid (0.056 g), tetrapropylammonium bromide
Table 1
Weights of NH4F used in ZSM‐5 syntheses.
F/Si ratio
0.3
Weight of NH4 F (g)
0.27
0.54
0.90
1.17
1.44
1.80
0.6
1.0
1.3
1.6
2. Experimental
2.1. Materials
2.0

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
(0.462 g), and water were mixed in a 100 ml beaker and
steps of 0.002 nm, with a count time of 5 s at each point. For the
crystallinity calculations, the intensities of the peaks from
(−101), (011), (051), and (−501) were compared with those
from a commercial catalyst. The specific surface areas of the
samples were measured by nitrogen adsorption at −196 °C
using a Belsorp Mini II sorption analyzer. Prior to measure‐
ments, the samples were heated for 4 h in vacuum at 400 ºC.
The surface areas were determined by the Brunau‐
er‐Emmett‐Teller (BET) method. The morphologies and size
changes of the prepared zeolite samples were studied using
scanning electron microscopy (SEM; JEOL JSM 5600). The Si/Al
ratios of the samples were determined using an ener‐
gy‐dispersive X‐ray (EDS) analyzer in the SEM chamber.
stirred. Ammonium fluoride (1.17 g) was then added to the
mixture. TEOS (5.16 g) was added and stirring was continued
for 3 h. The resulting gel was transferred to a Teflon‐lined
stainless‐steel autoclave and hydrothermal crystallization was
carried out at 170 °C for 5 d in a static oven. The crystallized
product was recovered by filtration, dried at 110 °C for 12 h,
and calcined at 650 °C in air for 8 h. The same procedure was
repeated with appropriate quantities of aluminum sulfate to
obtain ZSM‐5 with Si/Al = 50 and 75.
2.4. Syntheses of ZSM‐5 in fluoride medium with different
silicon sources
Temperature‐programmed desorption of ammonia
(NH3‐TPD) was performed using a BELCAT‐M catalyst analyzer
equipped with a thermal conductivity detector. The catalyst
sample (0.2 g) was placed in the microreactor and treated at
430 °C for 3 h under a He flow, followed by sorption of dry
ammonia at 100 °C in a static system for 90 min and 1.5 bar
ammonia pressure. Stripping was then performed for 40 min at
100 °C under a He flow. In this way, weakly physically adsorbed
ammonia in the sample was avoided. Desorption of ammonia
was carried out at a rate of 10 °C/min from 100 °C up to 650 °C
under an He flow (50 ml/min).
Solid‐state nuclear magnetic resonance (NMR) spectra of
ZSM‐5 were obtained under magic‐angle spinning (MAS) con‐
ditions, at ambient temperature, on a Bruker MSL 500 spec‐
trometer. The 27Al MAS‐NMR spectra were recorded at a reso‐
nance frequency of 130.32 MHz, with short 2.1 μs pulse and a
recycle delay of 0.1 s. The typical rotational frequency was ca.
10 kHz, using a zirconia rotor of diameter 4 mm, and 12500
scans were necessary to obtain a well‐resolved spectrum. 27Al
chemical shifts were reported in relation to a liquid solution of
aluminum nitrate. 29Si MAS‐NMR spectra were recorded at a
resonance frequency of 99.361 MHz on a Bruker MSL 500
spectrometer, with a short pulse of 4 μs, a recycle delay of 30 s,
and 2500 scans. The typical rotational frequency was about 4.5
kHz, using a zirconia rotor of diameter 7 mm. Chemical shifts
were given in ppm from standard tetramethylsilane.
Syntheses of ZSM‐5 in a fluoride medium were performed
using different sources of silicon (TEOS, sodium silicate, AS
silica, HS silica, and fumed silica). The following gel composi‐
tion was used: 0.07 C12H28NBr:1.0 SiO2:0.02 Al2O3:1.3
NH4F:0.02 P2O5:80 H2O.
In a typical procedure, aluminum sulfate (0.330 g), ortho‐
phosphoric acid (0.056 g), tetrapropylammonium bromide
(0.462 g), and water (35 g) were mixed in a 100 ml beaker and
stirred for 2 h. Ammonia (1:1) was added to give a faint per‐
manent precipitate. After 30 min of stirring, ammonium fluo‐
ride (1.17 g) was added. After 5 h of stirring, an appropriate
amount of silica source was added and stirring was continued
for another 5 h. The resulting gel was transferred to a Tef‐
lon‐lined stainless‐steel autoclave, and hydrothermal crystalli‐
zation was carried out at 170 °C for 5 d in a static oven. The
crystallized product was recovered by filtration, dried in an
oven at 110 °C for 12 h, and calcined at 650 °C in air for 8 h.
2.5. Potentiometric titrations
To establish complexation between phosphate and Al3+
during the ZSM‐5 synthesis, potentiometric titrations were
carried out. Potentiometric titrations of orthophosphoric acid
(0.02 mol/L) versus potassium hydroxide (0.6 mol/L), and
orthophosphoric acid plus Al3+ (0.0475 mol/L) versus potas‐
sium hydroxide, were carried out using a pH meter (Toshni‐
wal) provided with a combined glass electrode. The titrations
were carried out in a double‐walled 100 ml beaker provided
with an inlet and an outlet for water circulation. The tempera‐
ture was maintained at 30 °C, with an ionic strength of 0.5
mol/L. The equilibrium pH was measured for each addition of
potassium hydroxide. The measured pH was plotted against the
volume of alkali added to establish complex formation between
phosphate and Al3+.
2.7. Catalytic activity measurements
Tert‐butylation of ethylbenzene with t‐butyl alcohol as the
alkylating agent was carried out on a fixed‐bed vertical
down‐flow reactor of internal diameter 2 cm. About 0.5 g of the
catalyst was placed in the reactor and supported on either side
by a thin layer of quartz wool and ceramic beads. The reactor
was heated to the requisite temperature using a tempera‐
ture‐programmed furnace. The catalyst was activated at 500 °C
for 6 h in a controlled stream of air. The reactants were fed into
the reactor using a syringe infusion pump. The products col‐
lected in the receiver flask were analyzed in a Shimadzu 17‐A
gas chromatograph equipped with a DB column and a flame
ionization detector. The products were also identified using gas
chromatography‐mass spectrometry (Perkin Elmer Clarus
500).
2.6. Catalyst characterization
The synthesized products were identified and their crystal‐
linities were determined by powder X‐ray diffraction (XRD)
analysis. The XRD patterns were acquired on a PANalytical
X’pert PRO diffractometer equipped with a Cu Kα (0.154 nm)
radiation source and a liquid‐cooled germanium solid‐state
detector. The samples were scanned from 5° to 40° (2θ) in

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
(7)
(6)
(5)
(6)
(5)
(4)
(4)
(3)
(2)
(1)
(3)
(2)
(1)
5
10
15
20
2/(o )
25
30
35
40
5
10
15
20
25
30
35
40
2/()

Fig. 1. XRD patterns of commercial ZSM‐5(25) (1) and ZSM‐5 synthe‐
sized with F/Si ratios of 0.3 (2), 0.6 (3), 1.0 (4), 1.3 (5), 1.6 (6), and 2.0
(7).
Fig. 3. XRD patterns of commercial ZSM‐5(25) (1) and ZSM‐5 synthe‐
sized with TEOS (2), HS silica (3), sodium silicate (4), AS silica (5), and
fumed silica (6) as silicon sources.
ZSM‐5(25). Its XRD pattern, shown in Fig. 2(2), displayed high
crystallinity for all reflections.
3. Results and discussion
The XRD patterns of ZSM‐5 zeolites synthesized with dif‐
ferent silicon sources are shown in Fig. 3. All the patterns
showed the peaks characteristic of ZSM‐5 reported in the liter‐
ature [26,27]. The pattern derived from TEOS yielded more
intense reflections than those from other silicon sources,
therefore it was chosen for the bulk synthesis of ZSM‐5 (Si/Al =
25, 50, and 75). Ethanol, released from the hydrolysis of TEOS,
reduced the dielectric constant of the medium, and enhanced
the formation of highly crystalline zeolites. Sano et al. [28] re‐
ported the synthesis of large mordenite crystals in the presence
of aliphatic alcohols. Slow dissolution of a gel and release of
silicon and aluminum species enhanced the crystal size. The
intensities of the peaks from (051) in Fig. 3(2), (3), and (4)
were higher than those in Fig. 3(5) and (6). A low dielectric
constant or sodium content in the medium was therefore sug‐
gested to play a characteristic role in enhancing crystallinity
during synthesis. In addition, it was also seen to aid enhanced
growth of (011). Therefore the sodium content, as well as the
alcohol content, enhanced growth along the (051) and (011)
directions compared with the other directions. Figure 3(6)
shows more intense reflections than Fig. 3(5). The ammonium
content of the synthesis gel with fumed silica was lower than
that of the gel with AS silica, hence ammonium was also im‐
portant in enhancing intensity. The synthesis gel with AS silica,
which has a higher ammonium content than fumed silica,
yielded low crystallinity ZSM‐5 crystals, but a high ammonium
content was seen to aid growth of (−101) compared with (011).
So, from this analysis, it was verified that a synthesis gel with a
low dielectric constant, and a high sodium content or low am‐
monium content, yielded highly crystalline ZSM‐5 zeolites.
The XRD patterns of the ZSM‐5(25), (50), and (75) synthe‐
sized in a fluoride medium are presented in Fig. 4. The pattern
of commercial ZSM‐5 is also shown for comparison. Based on
the intensities of the dominant XRD peaks, the crystallinities of
all the ZSM‐5 samples were calculated, and the values are pre‐
3.1. Catalyst characterization results
3.1.1. XRD
The powder XRD patterns of ZSM‐5 zeolites, synthesized in
a fluoride medium with different fluoride contents, are pre‐
sented in Fig. 1. The XRD pattern of commercial ZSM‐5 is also
presented in the same figure for comparison. The reflections in
the XRD patterns of the synthesized samples were higher than
those of the commercial sample. The fluoride medium was
therefore proved to enhance the crystallinity of ZSM‐5 [24,25].
The interactions of fluoride with both silicon and aluminum,
facilitating their slow release for zeolite construction, are sug‐
gested as the main cause of the high crystallinity. Based on the
intensity of the patterns, the optimum F/Si ratio for crystalliza‐
tion was in the range 1.0 to 1.6. A similar range was also re‐
ported by Louis and Kiwi‐Minsker [24].
In fluoride‐mediated synthesis, the acidity has been report‐
ed to be low as a result of formation of AlFx species [25]. To
suppress formation of such AlFx species, a co‐complexant to
fluoride, namely phosphate, was used during the synthesis of
(1)
(2)
5
10
15
20
25
30
35
40
2/(o )
Fig. 2. XRD patterns of commercial ZSM‐5(25) (1) and synthesized
ZSM‐5(25) (2).

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
250
ZSM-5(50)
ZSM-5(25)
ZSM-5(75)
(4)
200
(3)
150
(2)
100
Adsorption
Desorption
(1)
40
50
5
10
15
20
2/()
25
30
35
0
0.0
0.2
0.4
0.6
0.8
1.0
Relative pressure (p/p0)
Fig. 4. XRD patterns of commercial ZSM‐5(25) (1) and synthesized
ZSM‐5(25) (2), ZSM‐5(50) (3), and ZSM‐5(75) (4).
Fig. 5. N2 adsorption‐desorption isotherms of the synthesized zeolites.
sented in Table 2. For the calculations, the crystallinity of the
commercial ZSM‐5 was taken as 100%. As expected, the syn‐
thesized samples showed higher crystallinities than the com‐
mercial ZSM‐5. This was also verified by SEM analysis, as dis‐
cussed below.
phate‐aluminum complex.
As shown in Eqs. (1) and (2), complex formation between
Al3+ and phosphate and hydrolysis of aluminum phosphate (Eq.
2) in the pH region between 2.20 and 3.35 released a larger
number of protons than free orthophosphoric acid, and re‐
duced the pH.
H3PO4 ⇌ H2PO4 + H+
(1)
(2)
3.1.2. BET analysis
Al3+ + H2PO4 + 2H2O ⇌ [(OH)2AlH2PO4] + 2H+
The N2 adsorption‐desorption isotherms of the synthesized
zeolites are shown in Fig. 5. All the isotherms showed type I
adsorption. The adsorption isotherms matched the desorption
isotherms in all three cases. The knee portion of the ZSM‐5(75)
isotherm was lower than those of ZSM‐5(25) and ZSM‐5(50),
indicating its lower surface area, as shown in Table 2. The iso‐
therms also showed a lower micropore volume for ZSM‐5(75)
than for the others. The isotherms also showed a slightly higher
surface area and pore volume for ZSM‐5(50) than for
ZSM‐5(25) and ZSM‐5(75).
When the pH was increased to 6.73–7.5, a second dissocia‐
tion occurred (the second buffer region), according to Eqs. (3)
and (4). If the complex [(OH)2AlH2PO4] has the same acidity as
–
H2PO4 , then the buffer region in the curve of orthophosphoric
acid and that of orthophosphoric acid plus Al3+ coincide. How‐
ever, a decrease in pH was observed for the latter compared
with the former. The dissociation constant of [(OH)2AlH2PO4]
–
may therefore be higher than that of H2PO4 . Delocalization of
the oxygen lone pair of the Al–O–P linkage in the aluminum
vacant 3p orbital resulted in a higher dissociation constant than
3.1.3. Potentiometric investigation of phosphate complexation
of aluminum
–
that of H2PO4 .
H2PO4 ⇌ HPO2 + H+
(3)
(4)
4
The pH‐volume curves for the titration of orthophosphoric
acid versus potassium hydroxide, and orthophosphoric acid
plus Al3+ versus potassium hydroxide, are shown in Fig. 6. In
the titration of orthophosphoric acid, three buffer regions were
identified, and the pKa values of 2.12 (pKa1), 7.21 (pKa2), and
12.30 (pKa3) derived from them were comparable to those
reported in the literature [29]. The titration curve of ortho‐
phosphoric acid plus Al3+ versus potassium hydroxide followed
nearly the same trend as that of orthophosphoric acid versus
potassium hydroxide, except that a fourth buffer region oc‐
curred at around pH 9.7, and there was a slight shift of the
curve to lower pH. This established the formation of a phos‐
[(OH)2AlH2PO4] ⇌ [(OH)2AlHPO4]– + H+
When the pH was increased to 9.60–10.19, a third disso‐
ciation occurred (third buffer region), according to Eq. (5). This
14
(1)
(2)
12
10
8
6
4
Table 2
Percentage crystallinities and BET results for ZSM‐5 zeolites.
2
Crystallinity
(%)
Surface area
(m2/g)
Pore volume
(cm3/g)
0
Catalyst
1
2
3
4
5
6
7
ZSM‐5(25)
(Commercial)
ZSM‐5(25)
ZSM‐5(50)
ZSM‐5(75)
100
430
0.163
Volume of KOH (ml)
117
124
120
750
790
634
0.259
0.280
0.214
Fig. 6. Plot of pH versus volume of alkali for orthophosphoric acid ver‐
sus potassium hydroxide (1) and orthophosphoric acid plus Al3+ versus
potassium hydroxide (2).

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
(a)
(b)
(c)
Fig. 7. SEM images of the synthesized ZSM‐5(25) (a), ZSM‐5(50) (b), and ZSM‐5(75) (c).
buffer region did not exist in the titration of orthophosphoric
acid itself, and was observed only in the titration of ortho‐
phosphoric acid plus Al3+. The dissociation of [(OH)2AlHPO4]–
species was confirmed by the formation of a clear solution. This
was noted in the titration of Al3+ plus orthophosphoric acid, but
not in the titration of orthophosphoric acid itself.
sized zeolites were slightly lower than those of the mother liq‐
uids. Complete transportation of aluminum from the mother
liquid to the framework and a slightly reduced amount of sili‐
con in the framework were therefore observed. In contrast to
the previous report of formation of AlFx, which reduced the
framework Al3+ content [25], the use of phosphate increased
the content. These results therefore clearly established the
influence of phosphate on the AlFx equilibria and on the trans‐
portation of aluminum from the gel to the framework. The re‐
sults also showed the absence of phosphate in the zeolite
framework, even though the potentiometric titrations estab‐
lished bonding between aluminum and phosphate across the
entire pH range. Phosphate remains in the coordination sphere
of aluminum even if the latter is bonded to two silicate units, as
shown in Scheme 1. When aluminum acquires a fourth coordi‐
nation in the zeolite framework, phosphate will be immediately
expelled, leaving aluminum in the trigonal planar state again.
Tetrahedral coordination is therefore only allowed by chelating
silicate units, and not by monodentate ligands such as phos‐
phate.
[(OH)2AlHPO4]– ⇌ [O–(OH)AlHPO4]2– + H+
(5)
If the pH was further increased to 11.30–12.22, a fourth
dissociation occurred (fourth buffer region), as shown in Eq.
(6). Here again, the buffer region for the titration of Al3+ plus
orthophosphoric acid occurred at a lower pH than that of or‐
thophosphoric acid. This again established complexation be‐
tween Al3+ and phosphate.
[O–(OH)AlHPO4]2– ⇌ [O–(OH)AlPO4]3– + H+
(6)
These results established the formation of a stable phos‐
phate complex with Al3+. Formation of such a complex is the
reason for transportation of aluminum from the gel to the
framework.
3.1.4. SEM and EDS analyses
The SEM images of ZSM‐5(25), (50), and (75) are shown in
Fig. 7. They show elongated prismatic crystals. The crystals of
ZSM‐5(75) were slightly smaller than those of the others. This
verified that the Si/Al ratio in the gel influences the crystal di‐
mensions. As a gel with a high silica content takes longer to
mineralize with fluoride, the crystallization time would also be
longer than those for gels with low silica contents. This resulted
in smaller crystal sizes for ZSM‐5(75) than for the others. The
crystal terraces in all the images were smooth, without any
deposits. The tiny crystallites of ZSM‐5(75) were the result of
incomplete crystallization. Similar images were also obtained
by Arichi and Louis [30].
3.1.5. NH3‐TPD
The NH3‐TPD results are shown in Fig. 8. In the evaluation of
zeolite acidity by NH3‐TPD, three types of acid sites were sug‐
gested, based on the temperature range of ammonia desorp‐
tion: weak acid sites, when the desorption occurs between 150
and 250 °C, medium acid sites for desorption between 250 and
350 °C, and strong acid sites for desorption between 350 and
O
P
O
P
OH
OH
OH
OH
-
O
O
O
O
pH = 4 6
pH = 3 HO
HO
Al3+
Al
H3PO4
Al
O
OH
Framework
Si
To compare the Si/Al ratios of the synthesized zeolites with
those of their mother liquids, EDS analyses were carried out.
The results are given in Table 3. The Si/Al ratios of the synthe‐
O
OH
Framework Si
O
P
OH
H
O
OH
O
O
O
O
O
O
Al
Si
Al
Si
Table 3
Results of EDS analyses.
H3PO4
O
O
O
O
O
O
Framework Si Framework Al
Framework
Si/Al ratio
Atomic content by EDS (%)
Catalyst
Mother liquid By EDS
Si
Al
O
P
H
H+
O
O
O
Si
O
O
ZSM‐5(25)
(Commercial)
ZSM‐5(25)
ZSM‐5(50)
ZSM‐5(75)
25
—
—
—
—
—
O
O
O
O
O
O
-
Si
Al
Si
Al
Si
O
O
O
O
O
O
O
O
O
25
50
75
22.0
38.0
52.7
14.7 0.7 84.6
21.0 0.6 78.5
20.0 0.4 79.5
0.0
0.0
0.0
Framework Si
Scheme 1. Phosphate complexation of aluminum.

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
and strong acid sites were lower than those for the synthesized
ZSM‐5(25). Here, too, the peaks from the medium and strong
acid sites overlapped, but desorption of ammonia extended
above 400 °C. The strong acid sites of the commercial catalyst
might therefore be stronger than those of the synthesized
ZSM‐5(25) and (50) catalysts. This indicates a high density of
acid sites in the channel intersections in the commercial cata‐
lyst. The absence of desorption was the result of adsorp‐
tion‐desorption equilibrium of ammonia. The desorption peak
from the weak acid sites was more intense than those of the
synthesized zeolites. The desorption also extended above 200
°C, hence these acid sites were more acidic than those of the
synthesized zeolites. As the weak acid sites are located along
the channels, the density might be high enough to increase
their strength. So, it was expected that the commercial zeolite
might be more active than the synthesized zeolites in the
tert‐butylation of ethylbenzene.
(2)
(1)
(3)
(4)
100
200
300
400
500
Temperature (C)
Fig. 8. NH3‐TPD profiles of commercial ZSM‐5(25) (1) and the synthe‐
sized ZSM‐5(25) (2), ZSM‐5(50) (3), and ZSM‐5(75) (4).
450 °C [31−33]. The peak between 100 and 210 °C for the weak
acid sites was clearly resolved for the synthesized zeolites, but
those of the medium and strong acid sites between 210 and
450 °C overlapped significantly. The quantity of weak acid sites
was lower than those of medium and strong acid sites. An addi‐
tional role of the fluoride medium is therefore to create larger
numbers of acid sites in the channel intersections, the regions
for strong acid sites, than along the channels, the regions for
weak acid sites. It has also been reported that in zeolites, 75%
of total acid sites were present in channel intersections and the
rest along the channels [34]. This was verified for the synthe‐
sized zeolites in this study, because there were more medium
and strong acid sites than weak acid sites. The NH3‐TPD pro‐
files also illustrated that with increasing Si/Al ratio, the quanti‐
ty of strong acid sites, corresponding to desorption between
350 and 450 °C, decreased. In addition, the desorption shifted
from 450 to 400 °C with increasing Si/Al ratio, and this was
clearly seen for ZSM‐5(75). It is difficult to understand why this
was not also observed for ZSM‐5(25) and (50), for which only
the densities were affected, not the strengths. The Si/Al ratio
alone is therefore not the controlling factor for the strengths of
the acid sites; the densities of acid sites located in the channel
intersections are also important. If the Si/Al ratio of the zeolite
is above 50, the acid sites are weak/medium, irrespective of
their location, therefore the higher the Si/Al ratio (particularly
above 50), the fewer the number of acid sites, even in the
channel intersections, and the greater the decrease in the
strengths of such sites. Such a decrease was also noted for the
weak acid sites; again, this was clearly observed for ZSM‐5(75).
So, with increasing Si/Al ratio, the acid sites are located far off,
and they are weaker than those of zeolites with lower Si/Al
ratios. Again, the weak acid sites of ZSM‐5(25) and (50) showed
only a decrease in the density, not in the strength, unlike the
medium and strong acid sites. Diffuse reflectance infrared Fou‐
rier transform (DRIFT) is a technique used to study acidity,
using pyridine and other bulky amines. Bulky amines are im‐
portant in determining the external surface acidity, which con‐
firm the crystallinity, but DRIFT facilities were not available.
The results for the commercial ZSM‐5(25) zeolite are also
shown in Fig. 8. The intensities of the peaks from the medium
3.1.6. 27Al MAS‐NMR and 29Si MAS‐NMR
The 27Al MAS‐NMR and 29Si MAS‐NMR spectra of ZSM‐5(25)
is shown in Fig. 9. In Fig. 9(a), the intense sharp peak centered
at δ = 52 arose from the tetrahedrally coordinated framework
aluminum [35]. As there was no peak close to δ = 0,
non‐framework alumina was proved to be absent. Dislodging of
aluminum from the framework during calcination was there‐
fore ruled out. The result also confirmed the thermal stability of
ZSM‐5 synthesized in a fluoride medium.
The peak at δ = −106 in Fig. 9(b) was from Q3 species [Si–
(OSi)3(OAl)] and that at δ = −112 was from Q4 species [Si(OSi)4]
[17,36]. As there was no peak below δ = −106 (close to δ =
−103), there were not signiꢀicant numbers of defective Si–OH
groups. This is important evidence of high crystallinity of the
zeolite, as discussed above.
3.2. Catalytic studies
Vapor‐phase tert‐butylation of ethylbenzene with t‐butyl
alcohol was studied over commercial ZSM‐5(25) and the syn‐
thesized ZSM‐5(25), ZSM‐5(50), and ZSM‐5(75) between 200
and 400 °C. The reaction parameters were optimized for max‐
imum ethylbenzene conversion and high selectivity for 4‐t‐BEB.
The effects of temperature, feed ratio, weight hourly space ve‐
(a)
100
-80
50
0
-50
-100
-140

(b)
-90
-100
-110

-120
-130
Fig. 9. 27Al MAS‐NMR (a) and 29Si MAS‐NMR (b) spectra of ZSM‐5(25).

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
Table 4
Effects of temperature on ethylbenzene conversion and product selectivity.
Selectivity (%)
3‐t‐BEB
10.0
19.8
17.2
18.8
22.9
5.1
Temperature
(°C)
200
Ethylbenzene
conversion (%)
39.1
Catalyst
1,2‐DEB
4.3
3.0
1.6
1.5
1.4
—
—
—
—
—
—
—
—
—
1.3
—
—
—
2.2
3.2
1,3‐DEB
2.2
1.1
2.6
3.9
6.0
—
—
—
—
—
—
—
—
—
4.2
—
—
—
5.1
6.9
1,4‐DEB
6.8
8.3
11.9
12.0
11.7
—
—
—
—
—
—
—
—
—
2.0
—
—
—
4.2
5.1
4‐t‐BEB
74.1
64.0
61.3
22.2
5.2
94.9
93.3
92.4
91.0
88.7
92.8
91.7
89.9
86.8
73.6
91.7
89.6
87.7
69.5
59.7
4‐t‐BVB
—
—
Benzene
2.6
3.8
5.4
12.1
24.9
—
ZSM‐5(25)
(Commercial)
250
300
350
400
200
250
300
350
400
200
250
300
350
400
200
250
44.3
55.9
49.4
32.2
33.5
41.7
49.3
44.1
28.0
30.3
38.2
41.6
39.9
30.4
25.3
31.7
—
29.5
27.9
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
ZSM‐5(25)
ZSM‐5(50)
ZSM‐5(75)
6.7
7.6
9.0
11.3
7.2
—
—
—
—
—
—
—
—
4.0
—
—
—
4.5
7.3
8.3
10.1
13.2
14.9
8.3
10.4
12.3
14.5
17.8
300
350
400
38.3
36.5
33.8
Reaction conditions: feed ratio ethylbenzene:tert‐butyl alcohol = 1:1; WHSV = 1.65 h–1.
locity (WHSV), and time‐on‐stream on conversion and product
selectivity were examined; the results are discussed below.
sites of ZSM‐5 to form benzene and ethyl cations. Electrophilic
reaction between ethyl cations and ethylbenzene yields dieth‐
ylbenzene isomers, i.e., 1,2‐DEB, 1,3‐DEB, and 1,4‐DEB. The
ethyl cations formed by de‐ethylation of ethylbenzene attack at
the ortho‐ and para‐positions of ethylbenzene to form 1,2‐DEB
and 1,4‐DEB, respectively. As the meta‐position of ethylben‐
zene is not favored for electrophilic attack, 1,3‐DEB has to be
formed by isomerization of 1,4‐DEB or 1,2‐DEB on the external
surface of the zeolite.
Since the kinetic diameter of 1,2‐DEB is 0.68 nm, it can be
formed in the channel intersections, but it cannot diffuse out of
the crystals. 1,4‐DEB can be formed either in the channel inter‐
sections or along the channels, and it can freely diffuse out, as
its kinetic diameter is 0.58 nm. 1,3‐DEB, with a kinetic diameter
of 0.68 nm, can be formed by isomerization of 1,4‐DEB or
1,2‐DEB on the external surfaces or in the channel intersec‐
tions. The selectivity for 1,2‐DEB was very low, as it was
formed on the external surfaces, and some of it isomerized to
1,3‐DEB. The slight decrease in selectivity for 1,2‐DEB from 200
to 400 °C was the result of increased steric hindrance for sub‐
stitution at the ortho position as a result of free rotation of the
ethyl group about the C–C bond. Such free rotation of alkyl
groups in alkyl aromatics has been reported previously [38]. At
low temperatures, the rotation would be slow, but at high
temperatures it would be rapid, thus significantly suppressing
substitution at both ortho positions. The formation of free ben‐
zene also supported disproportionation of ethylbenzene. The
3.2.1. Effects of temperature
tert‐Butylation of ethylbenzene with t‐butyl alcohol was
carried out over commercial ZSM‐5(25) at 200, 250, 300, 350,
and 400 °C. The products were 1,2‐diethylbenzene (DEB),
1,3‐DEB, 1,4‐DEB, 3‐t‐BEB, 4‐t‐BEB, 4‐t‐BVB, and benzene. The
conversions of ethylbenzene and the selectivities for different
products are presented in Table 4. The conversion increased
from 200 to 300 °C, but decreased at 350 and 400 °C as a result
of blocking of active sites by polybutene [37] and/or coke. Coke
was also observed between 200 and 300 °C, but to a lesser ex‐
tent than between 350 and 400 °C. A plausible pathway for
tert‐butylation of ethylbenzene is shown in Scheme 2. t‐Butyl
alcohol is chemisorbed on the Brönsted acid site of zeolites to
form t‐butyl cations. The electrophilic reaction between
ethylbenzene and t‐butyl cations yields 4‐t‐BEB.
In addition to tert‐butylation of ethylbenzene, dispropor‐
tionation of ethylbenzene was also observed, but was minimal.
The disproportionation of ethylbenzene can occur as shown in
Scheme 3. Ethylbenzene is chemisorbed on the Brönsted acid
ZSM-5
OH
+
Scheme 2. Formation of 4‐t‐BEB.
H+
O
H+
O
Al-
Disproportionation
-
Al
Si
+
Si
C2H5
+
+
+
zeolite
H
DEB isomers
Benzene
H
Scheme 4. Isomerization of 4‐t‐BEB to 3‐t‐BEB.
Scheme 3. Disproportionation of ethylbenzene.

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
H
CH2
CH
+
H3C
2
CH3
HC
+
H3C
H+
O
H+
Si
+
+
Al-
Scheme 5. Carbenium/carbonium ion mechanism.
selectivity for 4‐t‐BEB decreased with increasing temperature
as a result of isomerization to form 3‐t‐BEB.
for 3‐t‐BEB increased with increasing temperature over all the
catalysts. ZSM‐5(75) showed a slightly higher selectivity for
3‐t‐BEB than the others, but a lower selectivity than commer‐
cial ZSM‐5. Since it is more hydrophobic than the others, it can
better retain 4‐t‐BEB and promote isomerization on the exter‐
nal surfaces. Although commercial ZSM‐5 is more hydrophilic
than the others, its strong acidity and low crystallinity result in
high selectivity for 3‐t‐BEB. The optimum temperature was 300
°C, based on high conversion, and ZSM‐5(25) performed better
than the others.
The synthesized zeolite ZSM‐5(25) showed lower conver‐
sion than the commercial ZSM‐5, although both have the same
Si/Al ratio. This is because the strong acid sites of the commer‐
cial zeolite are of higher strength than those of the synthesized
zeolite.
The isomerization of 4‐t‐BEB proceeds as shown in Scheme
4. 4‐t‐BEB is chemisorbed on the Brönsted acid sites of ZSM‐5.
Protonation of 4‐t‐BEB occurs on the carbon that carries the
t‐butyl group. Reorganization of bonds results in the migration
of the t‐butyl group to the meta position to form 3‐t‐BEB. Such
a mechanism has already been reported [22]. The meta isomers
are thermodynamically more stable than the para isomers, and
this becomes the driving force for isomerization, so the selec‐
tivity for 3‐t‐BEB increases with increasing temperature. The
molecular size of 3‐t‐BEB (0.68 nm) is higher than the zeolite
pore size. Isomerization of 4‐t‐BEB to 3‐t‐BEB can therefore
only occur on the external surface of the zeolite. An unexpected
product is 4‐t‐BVB, so there is dehydrogenation of the ethyl
group of 4‐t‐BEB to 4‐t‐BVB. Such dehydrogenation has been
reported to result from the presence of non‐framework alumi‐
na in mesoporous AlMCM‐41 [22], so the catalyst might have
non‐framework alumina. In the absence of it, dehydrogenation
is suggested to occur by a carbenium/carbonium ion mecha‐
nism (Scheme 5). A proton is transferred to the methylene
group of 4‐t‐BEB to form a pentacoordinated transition state.
This decomposes to form a carbenium ion, which subsequently
releases an α‐hydrogen to form 4‐t‐BVB. However, such dehy‐
drogenation was not observed for ethylbenzene, as styrene was
not observed. Hence, the para t‐butyl group of 4‐t‐BEB might
activate formation of 4‐t‐BVB. Since the t‐butyl group is induc‐
tively electron‐repelling, it enhances the electronic cloud on the
methylene group and facilitates proton transfer to the meth‐
ylene from the zeolite.
3.2.2. Effects of WHSV
Since the isomerization of 4‐t‐BEB occurs on the few
Brönsted acid sites present on the external surface, it could be
avoided, or at least minimized, by increasing the WHSV. The
results of the effects of WHSV on ethylbenzene conversion and
product selectivity are presented in Table 5. The conversion
decreased with increasing WHSV. As expected, the isomeriza‐
tion activity decreased with increasing WHSV, thus confirming
our deductions. The optimum WHSV was 1.65 h−1. The selectiv‐
ity for 4‐t‐BEB at 6.61 h−1 was higher than those at the other
WHSVs, but the conversion was low, so it was not the optimum
value.
3.2.3. Effects of feed ratio
The same reaction was also studied over synthesized
ZSM‐5(25), ZSM‐5(50), and ZSM‐5(75). The results are also
presented in Table 4. These catalysts showed variations in the
conversions of ethylbenzene similar to those of the commercial
catalyst. The conversion increased from 200 to 300 °C and de‐
creased thereafter. ZSM‐5(25) showed a higher conversion
than others as a result of a higher density of acid sites. Although
the acidity of ZSM‐5(50) was 50% less than that of ZSM‐5(25),
the conversion at 300 °C was not 50% lower. This is because of
slow diffusion of reactants along the channels in ZSM‐5(50) as a
result of its higher hydrophobicity. An important observation in
the reactions over ZSM‐5 catalysts synthesized in a fluoride
medium was the complete absence of dehydrogenation and
disproportionation at 300 °C. Disproportionation of ethylben‐
zene can occur either inside or outside the pores, but acid sites
must be available. Since t‐butyl alcohol is better chemisorbed
than ethylbenzene on Brönsted acid sites, disproportionation of
ethylbenzene is avoided over all the catalysts. The selectivity
The feed ratio was altered to study its influence on
ethylbenzene conversion and product selectivity; the results
are presented in Table 6. The conversion decreased with in‐
creasing t‐butyl alcohol content in the feed. The ZSM‐5 chan‐
nels could be filled with large numbers of t‐butyl cations, and
these could offer high steric hindrance to diffusion of
tert‐butylated ethylbenzene. A high alcohol content in the feed,
particularly above a feed ratio of 1:1, might therefore not be
Table 5
Effects of WHSV on ethylbenzene conversion and product selectivity
over ZSM‐5(25).
Selectivity (%)
3‐t‐BEB 4‐t‐BEB
Feed rate
(ml/h)
WHSV
(h–1)
1.65
3.31
4.96
6.61
Ethylbenzene
conversion (%)
49.3
1
2
3
4
7.6
5.7
2.5
1.5
92.4
94.3
97.5
98.5
43.5
40.2
30.9
Reaction conditions: temperature 300 °C, feed ratio 1:1.

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
does not provide mush space for coke formation inside the
Table 6
pores. The selectivity for 3‐t‐BEB decreased slightly with in‐
creasing time‐on‐stream as a result of blocking of the external
surface acid sites by coke. As coke formation is very much fa‐
vored on such external acid sites, these sites were gradually
deactivated and isomerization also gradually decreased. The
decrease in the selectivity for 3‐t‐BEB and the increase in the
selectivity for 4‐t‐BEB confirmed that the latter is the precursor
of the former. In addition, the deactivation of external acid sites
may be more rapid than deactivation of sites lying inside the
channels, as the medium pore size may not encourage coke
formation, which requires a lot of space [15], as stated above.
2‐t‐BEB was not obtained during the entire period of
time‐on‐stream. The time‐on‐stream study illustrates that if the
few external acid sites are completely blocked by coke for‐
mation, the catalyst can give 100% selectivity for 4‐t‐BEB.
Effects of feed ratio on ethylbenzene conversion and product selectivity
over ZSM‐5(25).
Selectivity (%)
3‐t‐BEB 4‐t‐BEB
WHSV
(h–1)
1.65
1.62
1.61
1.68
1.69
Ethylbenzene
conversion (%)
49.3
Feed ratio
1:1
1:2
1:3
2:1
3:1
7.6
5.5
3.9
6.8
9.2
92.4
94.5
96.1
93.2
90.8
42.7
36.2
51.5
48.8
Reaction conditions: temperature 300 °C, feed rate 1 ml/h.
conducive to high conversion. However, when the ratio was
changed to 2:1, the conversion increased as there may not be
many free t‐butyl cations along the channels to offer hindrance
to diffusion of 4‐t‐BEB. In other words, close adsorption of
t‐butyl cations prevents a high rate of diffusion of 4‐t‐BEB.
Hence, optimizing the ethylbenzene content in the feed may be
better than other alternatives for enhancing conversion; 2:1
was chosen as the optimum feed ratio. Although a large crystal
size is generally good for high para selectivity, as discussed in
the introduction, the present reaction does not depend on the
shape‐selective features of zeolites. As the ortho product is
avoided as a result of steric hindrance, and the meta product is
kinetically unfavorable, the reaction gives the para product
preferentially. Even if the meta product is formed in the chan‐
nel intersections, it cannot diffuse out of the pores. The meta
product can therefore only be formed by isomerization of the
para product on the external surface. If the external surface
carries few weak acid sites, these sites can be blocked by
post‐modification for complete suppression of isomerization.
Since ZSM‐5 synthesized in a fluoride medium was highly crys‐
talline, its isomerization activity was minimum. In addition,
disproportionation was also greatly suppressed.
4. Conclusions
ZSM‐5 synthesized in a fluoride medium is better than other
zeolites for para‐selective alkylation of aromatics or alkyl aro‐
matics, as trans alkylation, disproportionation, and isomeriza‐
tion are suppressed. The fluoride medium provides large
ZSM‐5 crystals of high crystallinity, significantly improving the
para selectivity of the catalyst. In ZSM‐5, coke formation is also
suppressed, because of its medium pore size. The number of
external acid sites is minimized because of the high crystallini‐
ty. Even if there are few external acid sites, their activity for
isomerization is rapidly lost as a result of coke deposition. By
blocking such acid sites by simple post‐modification, it is possi‐
ble to ensure 100% para selectivity. A fluoride medium is
therefore better than an alkaline medium for the commercial
production of defect‐free highly crystalline para‐selective
ZSM‐5 catalysts.
3.2.4. Effects of time‐on‐stream
References
The effects of time‐on‐stream on ethylbenzene conversion
and product selectivity were studied at 300 °C with a feed ratio
of 2:1; the results are shown in Fig. 10. The conversion de‐
creased with increasing time‐on‐stream, but the decrease was
not rapid. This was because of the medium pore size, which
[1] Ichioka R, Yamakawa S, Okino H, Kato H, Iwayama K, Konta H,
Kitamura A. US Patent 6 040 490. 2000
[2] Zuo X, Niu F, Snavely K, Subramaniam B, Busch D H. Green Chem,
2010, 12: 260
[3] Chavan S A, Srinivas D, Ratnasamy P. J Catal, 2001, 204: 409
[4] Chu C C. US Patent 4 496 662. 1985
100
[5] Burress G T. US Patent 4 565 899. 1986
[6] Kaeding W W, Yannich P J, Klosek J M, Young L B. EP Patent
0 217 492. 1987
[7] Drobny J G. Handbook of Thermoplastic Elastomers. 1st Ed. New
York: William Andrew Inc, 2007. 175
[8] Danilina N, Payrer E L, van Bokhoven J A. Chem Commun, 2010, 46:
1509
[9] Wise J J. US Patent 3 251 897. 1966
Conversion of ethylbenzene
Selectivity for 3-t-BEB
Selectivity for 4-t-BEB
80
60
40
20
0
[10] Haag W O, Lago R M, Weisz P B. Faraday Discuss Chem Soc, 1981,
72: 317
[11] Olson D H, Haag W O. ACS Symp Ser, 1984, 248: 275
[12] Halgeri A B. Bull Catal Soc India, 2003, 2: 184
[13] Kim J H, Namba S, Yashima T. Zeolites, 1991, 11: 59
[14] Kim J H, Namba S, Yashima T. Appl Catal A, 1993, 100: 27
[15] Chen N Y, Garwood W E, Dwyer F G. Shape Selective Catalysis in
Industrial Applications. 1st Ed. New York: Dekker, 1989. 26
1
2
3
4
5
Time (h)
Fig. 10. Effects of time‐on‐stream on ethylbenzene conversion and
product selectivity over ZSM‐5(25).

PUSHPARAJ Hemalatha et al. / Chinese Journal of Catalysis 34 (2013) 294–304
[16] Guth J L, Kessler H, Higel J M, Lamblin J M, Patarin J, Seive A,
lites, 1997, 19: 288
Chezeau J M, Wey R. ACS Symp Ser, 1989, 398: 176
[17] Axon S A, Klinowski J. Appl Catal A, 1994, 111: 29
[18] Kessler H, Patarin J, Schott‐Darie C. Stud Surf Sci Catal, 1994, 85:
75
[19] Nayak V S, Riekert L. Appl Catal, 1986, 23: 403
[20] Paparatto G, Moretti E, Leofanti G, Gatti F. J Catal, 1987, 105: 227
[21] Arsenova‐Härtel N, Bludau H, Schumacher R, Haag W O, Karge H G,
Brunner E, Wild U. J Catal, 2000, 191: 326
[28] Sano T, Wakabayashi S, Oumi Y, Uozumi T. Microporous Mesopo‐
rous Mater, 2001, 46: 67
[29] Jeffery G H, Bassett J, Mendham J, Denney R C. Vogel’s Textbook
of Quantitative Chemical Analysis. 5th Ed. UK: Longman, 1989.
833
[30] Arichi J, Louis B. Cryst Growth Des, 2008, 8: 3999
[31] Kim J H, Park M J, Kim S J, Joo O S, Jung K D. Appl Catal A, 2004,
264: 37
[22] Umamaheswari V, Palanichamy M, Arabindoo B, Murugesan V.
Proc Indian Acad Sci Chem Sci, 2002, 114: 203
[32] Vishwanathan V, Jun K W, Kim J W, Roh H S. Appl Catal A, 2004,
276: 251
[23] Bidart A M F, Borges A P S, Chagas H C, Nogueira L, Lachter E R,
Mota C J A. J Braz Chem Soc, 2006, 17: 758
[33] Kim S D, Baek S C, Lee Y J, Jun K W, Kim M J, Yoo I S. Appl Catal A,
2006, 309: 139
[24] Louis B, Kiwi‐Minsker L. Microporous Mesoporous Mater, 2004, 74:
171
[34] Bortnovsky O, Sazama P, Wichterlova B. Appl Catal A, 2005, 287:
203
[25] Gougeon R D, Brouwer E B, Bodart P R, Delmotte L, Marichal C,
Chezeau J M, Harris R K. J Phys Chem B, 2001, 105: 12249
[26] Treacy M M J, Higgins J B. Collection of Simulated XRD Powder
Patterns for Zeolites. 4th Ed (Rev). Amsterdam: Elsevier, 2001.
239
[35] Seiler M, Wang W, Hunger M. J Phys Chem B, 2001, 105: 8143
[36] Grieken R V, Sotelo J L, Menéndez J M, Melero J A. Microporous
Mesoporous Mater, 2000, 39: 135
[37] Liu Z, Moreau P, Fajula F. Appl Catal A, 1997, 159: 305
[38] Schaefer T, Chan W K, Sebastian R, Schurko R, Hruska F E. Can J
Chem, 1994, 72: 1972
[27] Woolery G L, Kuehl G H, Timken H C, Chester A W, Vartuli J C. Zeo‐
Graphical Abstract
Chin. J. Catal., 2013, 34: 294–304 doi: 10.1016/S1872‐2067(11)60482‐8
Effects of crystallinity of ZSM‐5 zeolite on para‐selective tert‐butylation of ethylbenzene
PUSHPARAJ Hemalatha, MANI Ganesh, MUTHIAHPILLAI Palanichamy, VELAYUTHAM Murugesan, PARK Yong‐Ki, CHOI Won Choon,
JANG Hyun Tae*
Hanseo University, South Korea; Anna University, India; Korea Research Institute of Chemical Technology, South Korea
A fluoride medium offers defect‐free, highly crystalline ZSM‐5 crystals. High crystallinity confers high para selectivity (> 90%) in
tert‐butylation of ethylbenzene. A fluoride medium is better than an alkaline medium for the commercial production of para‐selective
ZSM‐5 catalysts.