Catalytic role of different pore systems in MCM-49 zeolite for liquid alkylation of benzene with ethylene

Article abstract of DOI:10.1016/j.jcat.2011.07.004

The properties of three types of pore systems in MCM-49 zeolite have been investigated by precoking to block the supercages and poisoning the surface pockets. Most Lewis acid sites (ca. 73%) are located within the supercages, while 38% of Broensted acid s

Full text of DOI:10.1016/j.jcat.2011.07.004

Journal of Catalysis 283 (2011) 68–74
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Catalytic role of different pore systems in MCM-49 zeolite for liquid alkylation
of benzene with ethylene
Kefeng Liu a,b, Sujuan Xie a, Shenglin Liu a, Guoliang Xu a, Ningning Gao a,b, Longya Xu a,
⇑
a State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
b Graduate University of Chinese Academy of Science, Beijing 100049, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 May 2011
Revised 4 July 2011
Accepted 8 July 2011
Available online 30 August 2011
The properties of three types of pore systems in MCM-49 zeolite have been investigated by precoking to
block the supercages and poisoning the surface pockets. Most Lewis acid sites (ca. 73%) are located within
the supercages, while 38% of Brönsted acid sites are located in the surface pockets, 28% in the sinusoidal
pores and only 34% in the supercages, respectively. Alkylation of benzene with ethylene mainly occurs in
the surface pockets, and the contribution from the sinusoidal pores is limited. The generation of heavy
aromatics in supercages results in the deactivation of MCM-49 at the early reaction stage. A moderate
alkali-treatment process has been demonstrated to enlarge the opening of supercages without affecting
the sinusoidal pores, resulting in improved diffusion of big molecules and hence better catalytic perfor-
mance for liquid alkylation of benzene with ethylene.
Keywords:
MCM-49 zeolite
Liquid alkylation
Benzene
Ó 2011 Elsevier Inc. All rights reserved.
Ethylene
Pore system
1. Introduction
sinusoidal 10 membered ring (MR) pore (0.41 ꢀ 0.51 nm) and the
large cylindrical supercage defined by 12MR (0.71 nm o.d. ꢀ 1.82
Zeolites have been widely used in the petrochemical processes
because of their unique advantages of highly selective, less toxic,
environmentally friendly and readily reproducible in catalytic
reactions [1]. The morphology, textural properties and chemical
composition of zeolites significantly affect their catalytic perfor-
mances. It is well known that many zeolites contain two or more
kinds of pore systems. The nature and amounts of the acid sites
vary in different pore systems, or even in the same type of pore
in zeolites with different chemical compositions (Si/Al molar ratio),
leading to diverse and sometimes conflicting results for the same
catalytic reaction [2–5]. A number of methods, such as selective
ion exchange [6] and poisoning by basic molecule adsorption
[7,8], have been attempted to investigate the roles of acid sites in
different locations or with different strengths in catalytic reactions.
These results indicate that the acidity of zeolite can be tuned to
some extent to favor the formation of targeted products.
nm height) with 10MR opening (0.41 ꢀ 0.54 nm). Furthermore,
there are half-supercages (denoted as surface pocket) scattering
over the surface (0.71 nm o.d. ꢀ 0.90 nm height) [12,13]. The acid
sites in the pockets of MCM-22 zeolite can be characterized
by poisoning method with bulky alkaline molecules like di-
tert-butylpyridine [14], collidine [15] and 2,6-dimethylquinoline
[8,16]. Moreover, the supercage pore system of MCM-22 can be
selectively blocked by precoking through the m-xylene transforma-
tion [17–19]. Therefore, the identification of the acid sites in three
different pore systems (pocket, supercage and sinusoidal pore) of
MCM-22 has been realized, paving the path for the studies of their
roles in catalytic reactions. Although MCM-49 zeolite possesses
nearly the same framework topology as the calcinated MCM-22
(only the average unit cell c-parameter is ꢁ0.02 nm longer in
MCM-49) [20], it is noted that MCM-49 zeolite exhibits much high-
er selectivity of ethylbenzene in the alkylation of benzene with eth-
ylene [21]. Therefore, it is important to investigate the effect of acid
properties and pore structures of MCM-49 zeolite on the liquid
alkylation of benzene with ethylene.
Liquid alkylation of benzene with ethylene is an important pet-
rochemical process to produce ethylbenzene. Some zeolites, such as
USY, Beta and MCM-22, exhibit promising performance in this reac-
tion [1,9–11]. Especially MCM-22 zeolite attracts remarkable atten-
tion due to the higher selectivity of ethylbenzene [10]. This has
been attributed to its special pore systems: the two-dimensional
In this work, we have studied the roles of different pore systems
in MCM-49 zeolite (pocket, supercage and sinusoidal pore) for the
liquid alkylation of benzene with ethylene. Possible reaction path-
ways in different pore systems are proposed. Furthermore, the
effects of alkali treatment on different pore systems of MCM-49
have been studied to elucidate the primary cause of the improved
catalytic performance for the alkylation of benzene.
⇑
Corresponding author. Fax: +86 411 84693292.
E-mail address: lyxu@dicp.ac.cn (L. Xu).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.07.004

K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
69
2 h (1 ꢀ 10ꢂ2 Pa) in IR cell. After the sample was cooled down to
the room temperature, a spectrum was recorded as background.
Subsequently, the wafers were exposed to pyridine vapor for
20 min at 0 °C and then outgassed at 150 °C for 30 min. The spectra
were collected at room temperature. The distributions of B and L
acid sites in three pore systems of HM, 0.025AT and 0.3AT were
determined by a subtraction method with the bands at 1543 and
1450 cmꢂ1 using the values of the molar extinction coefficients
2. Experimental
2.1. Catalyst preparation
The as-received Na-MCM-49 zeolite (Si/Al molar ratio of 10.2,
Tieling Catalyst) was calcinated at 540 °C for 6 h to remove the
template, followed by three repeated ion exchanges in 0.8 N
NH4NO3 solution at 80 °C for 2 h. H-form MCM-49 zeolites was
obtained by the calcination of its NHþ4 form counterpart, and the
sample was named as HM. The alkali-treated samples were pre-
pared in the NaOH solutions with different concentrations at the
given temperatures for a certain period and transformed into H-
form. The resulting samples were denoted as 0.025AT (0.025 N
NaOH solution, 25 °C and 15 min) and 0.3AT (0.3 N NaOH solution,
75 °C and 120 min), respectively. All the H-form samples were
pressed into tablets, crushed and sieved into 20–40 mesh particles.
(1.13 and 1.28 cm
ple, in sample HM, the amounts of B and L acid sites in sinusoidal
pores were calculated with the integrated spectrum of HMxy-DMQ
l
molꢂ1) determined by Guisnet [23]. For exam-
,
and the distribution of B and L acid sites in the surface pockets
and supercages was acquired by the difference in results between
HMxy and HMxy-DMQ and between HM and HMxy, respectively. The
OH stretching spectra were determined after activating the sam-
ples in situ in an IR cell in vacuum (1 ꢀ 10ꢂ3 Pa) at 250 °C for 5 h.
Samples with the precoked supercages, denoted as HMxy
,
2.3. Catalyst evaluation
0.025ATxy and 0.3ATxy respectively, were prepared following a
reported procedure [22]: HM, 0.025AT and 0.3AT were pretreated
at 500 °C for 1 h and then cooled down to 350 °C under N2 flow.
Subsequently, the transformation of m-xylene (>99%, J&K) was car-
ried out at 350 °C, atmospheric pressure and 13 hꢂ1 of m-xylene
WHSV with N2 flow (32.5 ml/min) for 10 h. Samples with only
sinusoidal pore accessible (named as HMxy-DMQ, 0.025ATxy-DMQ
and 0.3ATxy-DMQ), for the IR characterization, were prepared by
dropwise adding mixture of 2,4-dimethylquinoline (2,4-DMQ)
(97%, TCI) and dichloromethane to the HMxy, 0.025ATxy and 0.3ATxy
wafers.
Liquid alkylation of benzene (>99%, J&K) with ethylene was car-
ried out in a stainless steel fixed bed reactor. One gram of catalyst
was loaded in the center of the reactor and pretreated at 500 °C for
1 h in N2. After the reactor was cooled down to 220 °C, benzene
was pumped in to fully fill the reactor before ethylene was intro-
duced. The reactions were carried out at 3.5 MPa, 220 °C, 1.5 hꢂ1
of weight hourly space velocity (WHSV) of ethylene, 12/1 of ben-
zene/ethylene molar ratio. The samples with precoked supercages
were washed in the liquid benzene (300 ml hꢂ1 gꢂ1) for 30 min to
remove carbon on the surface of catalysts before ethylene was
introduced. The sample with the precoked supercage and poisoned
surface pockets, such as HMxy-DMQ, was obtained by poisoning the
external surface of HMxy by means of 2,4-DMQ adsorption. The poi-
soning process was carried out by adding a given amount of 2,4-
DMQ into benzene. When the liquid alkylation of benzene with
ethylene was performed, the flow rate of 2,4-DMQ was 207
2.2. Characterization
X-ray diffraction (XRD) patterns were collected on an X Pert Pro
X-ray diffractometer operated at 40 kV and 40 mA using Cu K
a
radiation. The relative crystallinities (RC) of the samples were esti-
mated by comparing the areas of their characteristic peaks (10.0°,
14.3°, 16.0°, 22.7° and 26.0° of 2h) with those of 0.025AT, which
was assumed as 100%.
l
molꢃ(g cat)ꢂ1 hꢂ1. The ethylene conversion and selectivities of
products within different pore systems were calculated using the
same method as that in Section 2.2 for the distributions of B and
L acid sites.
The area of characteristic peaks of the sample
The area of characteristic peaks of 0:025AT
RC ¼
ꢀ 100%
3. Results and discussion
The chemical composition of samples was analyzed on a Philip
Magix 601X X-ray fluorescence (XRF) spectrometer.
3.1. Characterization of the precoked and poisoned samples
Ammonia temperature-programmed desorption (NH3-TPD) was
performed in a conventional U-shaped stainless steel micro-reactor
(i.d. = 4 mm) with helium (He) as the carrier gas. The effluent was
monitored by an online gas chromatograph (Shimadzu GC-8A)
equipped with a TCD detector. Typically, the sample (0.14 g) was
pretreated at 600 °C for 1 h, then cooled down to 150 °C and satu-
rated with NH3. After a stable baseline had been obtained, the sam-
ple was heated from 150 to 600 °C at a ramp rate of 18.8 °C/min.
The stability of the coke generated in the transformation of m-
xylene was analyzed by the thermogravimetry and differential
As shown in Fig. 1, HM possesses fully crystalline MWW struc-
ture [24]. After HM is precoked through m-xylene transformation
scanning calorimetry (TG-DSC) analysis with
a Netzsch STA
449F3-DSC 204HP analyzer. The process was performed from 40
to 400 °C at a ramp rate of 10 °C/min under N2 flow. The effluent
was analyzed by an Ametek Dymaxion mass spectrometer (MS).
N2 adsorption was carried out at ꢂ196 °C with an automatic
Micromeritics ASAP 2020 apparatus. The fresh samples were pre-
treated under vacuum at 350 °C for 6 h. However, the precoked
samples were pretreated only at 100 °C for 2 h to prevent coke
from being removed [22].
The Brönsted (B) and Lewis (L) acid sites of samples were deter-
mined by pyridine adsorption followed by infrared (Py-IR) mea-
surements on a Bruker 70 IR spectrometer. Samples were pressed
into self-supported wafers, followed by evacuation at 250 °C for
Fig. 1. XRD patterns of MCM-49 zeolite before and after alkali-treatment.

70
K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
(HMxy), the surface area reduces by 294 m2/g, which is similar to its
reduced area of micropore (263 m2/g) (Table 1), indicating that
most coke is located in micropores. The remaining micropore vol-
ume (0.053 m3/g) of HMxy is 31.2% of that in HM. This is perfec-
tively in agreement with the molecular modeling results that the
volume of the sinusoidal pores takes up 30% of the total micropore
volume in MWW zeolites [25]. This result suggests that the super-
cages can be successfully blocked by the coke generated in the
m-xylene transformation reaction while the sinusoidal pore is
preserved.
The mass loss determined by TG curve below 200 °C (spectra
are not shown here) is generally regarded as the evaporation of
the adsorbed water and corresponds to the endothermic peak in
the DSC curve. Moreover, the mass loss between 200 and 400 °C
is not significant. No coke signal except N2 and H2O is detected
by mass spectra in the temperature range of 40–400 °C, suggesting
that coke located in the supercages is stable below 400 °C.
2,4-DMQ adsorption has been proved to be an effective method
to poison the surface pockets of HMCM-22 zeolite. Furthermore,
the physisorbed and hydrogen-bonded 2,4-DMQ can be eliminated
over 200 °C [16]. So, the precoked and poisoned samples should be
characterized under proper conditions.
(HMxy), the band at 3668 cmꢂ1 is hardly to be observed, indicating
that most extraframework aluminum species are located in the
supercages. The band of external silanol group seems not to be af-
fected by coke, but that of the internal silanol group shifts to the
lower wave number region (ꢁ3712 cmꢂ1) with a weaker intensity.
This is possibly caused by the interaction between the silanol
groups and the coke in the supercages. The band intensity of bridg-
ing hydroxyl groups also obviously decreases, which is consistent
with the results on MCM-22 catalysts obtained by Matias et al.
[2]. This indicates that the bridging hydroxyl groups and the extra-
framework aluminum species in the supercages, which are widely
regarded as the origin of B and L acid sites, have been covered by
coke generated in the m-xylene transformation reaction. Compared
with HMxy, HMxy-DMQ exhibits much lower band intensity at
ꢁ3745 cmꢂ1, and this is in agreement with the results of Ayrault
et al. [16]. It confirms that 2,4-DMQ can effectively poison the
active sites in the pockets of HMxy
.
The acid properties of MCM-49 samples were studied by Py-IR
and NH3-TPD techniques. In the Py-IR spectra (Fig. 3), the bands
at 1450 and 1543 cmꢂ1 are representative of L and B acid sites,
respectively [28,29]. The band at 1490 cmꢂ1 is related with both
L and B acid sites. Compared with HM, the band intensity at
1450 cmꢂ1 of HMxy exhibits a much more notable decrease than
the one at 1543 cmꢂ1, indicating that the percentage of L acid sites
is higher than that of B acid sites in the supercages. The distribu-
tion of L and B acid sites in different pores of HM is calculated with
Py-IR spectra of HM, HMxy and HMxy-DMQ and the results are listed
in Table 2. The data clearly show that most L acid sites (ca. 73%) are
located in the supercages in HM, while 34% of B acid sites are lo-
cated in the supercages, 38% in the pockets and 28% in the sinusoi-
dal pores.
Fig. 2 shows the FT-IR spectra of HM, HMxy and HMxy-DMQ in the
–OH stretching region (3800–3500 cmꢂ1). Four –OH bands are
observed in the spectrum of HM. The bands at about 3745 and
3722 cmꢂ1 are related with the external and the internal silanol
groups, respectively. The band at 3618 cmꢂ1 is attributed to bridg-
ing hydroxyl groups. Another band at 3668 cmꢂ1 is ascribed to
Al–OH groups, where Al is connected with the zeolite framework
by one or two chemical bonds [26,27]. For the precoked sample
Table 1
3.2. Roles of acid sites in different pores of MCM-49 for catalytic
ethylene conversion
N2 adsorption and desorption results of fresh and coked MCM-49 before and after
alkali-treatment.
SBET (m2/g)
Smicro (m2/g)
Vtotal (cm3/g)
Vmicro (cm3/g)
As shown in Fig. 4A, ethylene conversion exhibits a very fast
decrease at the early period over HM and then tends to be stable.
In comparison, within 12 h of time on stream (TOS), ethylene con-
HM
HMxy
0.025AT
0.025ATxy
0.3AT
486
378
0.49
0.39
0.42
0.33
0.55
0.42
0.17
0.05
0.17
0.05
0.13
0.05
192 (ꢂ294)
115 (ꢂ263)
458
348
version is stable at about 70% and 8% over HMxy and HMxy-DMQ
,
197 (ꢂ261)
442
194 (ꢂ248)
118 (ꢂ230)
284
114 (ꢂ170)
respectively. Since the supercages of HMxy are blocked, and both
the supercages and the surface pockets of HMxy-DMQ are poisoned,
ethylene conversion within different pores vs. TOS can be obtained
from Fig. 4A. As shown in Fig. 4B, there is not obvious decrease for
ethylene conversion within the surface pockets and sinusoidal
pores during the TOS investigated. At TOS of 1 h, 26.2% of ethylene
is converted in the supercages, and it decreases gradually to only
0.3ATxy
Data listed within the parenthesis are the decreased value of the precoked sample
when compared with the fresh sample.
Fig. 2. The FT-IR spectra of HM, HMxy and HMxy-DMQ in the –OH stretching region.
HM: H-type MCM-49 zeolite. HMxy: H-type MCM-49 zeolite with the blocked
supercage. HMxy-DMQ: H-type MCM-49 zeolite with the blocked supercage and
poisoned surface pockets.
Fig. 3. IR spectra of pyridine adsorbed HM, HMxy and HMxy-DMQ
.

K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
71
Table 2
be relatively strong. Additionally, Li et al. found that more oligomer
products and coke were produced with the increase in L acid sites
by comparing the parent with the hydrothermally treated MCM-22
zeolites in the alkylation of benzene with diethyl carbonate [35].
Since most L acid sites are located within the supercages of HM
(Table 2), the oligomerization and coking reaction of ethylene will
cover some of strong acid sites, resulting in the deactivation of HM.
In addition, when zeolites are synthesized or post-treated, the non-
crystalline raw materials or amorphous silica–alumina generated
in the calcination process is usually included [36–38]. These amor-
phous silica–alumina species might deposit on the surface of
MCM-49 zeolite and cause partial blockage of the 10MR opening
of the supercages [39]. The resulting hindrance for the diffusion
of the bulky by-product molecules might also cause deactivation
of the supercages.
Calculated distributions of B and L acid sites, ethylbenzene, diethylbenzene and C10
aromatics yields in different pores of HMCM-49.
+
Supercage
Pocket
Sinusoidal
B (%)
L (%)
Ethylbenzene (%)
Diethylbenzene (%)
34
73
19.5
40.7
80.4
38
10
72.5
54.9
19.5
28
17
8.0
4.4
0.1
C
10+ aromatics (%)
Reaction conditions: 3.5 MPa, 220 °C, 12 of B/E molar ratio, 1.5 hꢂ1 of WHSV (eth-
ylene), 3 h of TOS.
Liquid alkylation of benzene with ethylene was mainly cata-
lyzed by strong acid sites, while weak acid sites were almost inef-
fective for the reaction [7]. Therefore, it is understandable for the
low ethylene conversion within the sinusoidal pores (most acid
sites are relatively weak) of MCM-49 [34].
3.3. Distribution of products in different pores of MCM-49
Liquid alkylation of benzene with ethylene is a complicated pro-
cess, which mainly includes the following reaction routes: alkyl-
ation between ethylene and benzene produces ethylbenzene
(process A), further alkylation of ethylbenzene with ethylene gen-
erates diethyl- or polyethylbenzene [11,40] (process B), oligomer-
ization of ethylene [41,42] (process C) and the transformation of
alkylbenzene (polyethylbenzene or products of alkylation between
benzene with ethylene oligomers) (process D) yield molecule with
bigger size [19].
Within 12 h of TOS, the distribution of products over HM, HMxy
and HMxy-DMQ exhibits indistinctive change. The yields of the
representative products, ethylbenzene, C10+ aromatics (heavy aro-
matic products with ten and more carbon atoms) and diethylben-
zene within different pore systems of HM can be calculated with
the values at 3 h of TOS in Table 3. The yield of ethylbenzene in
the supercages and in the sinusoidal pores is much lower than that
in the surface pockets (Table 2). Moreover, more than half of dieth-
ylbenzene is yielded in the pockets, indicating that the surface
pockets are the main locations for the alkylation of benzene (pro-
cesses A and B). Additionally, the yield of C10+ aromatics in the
supercages is the highest among these in three different pore sys-
tems (process D).
Since the supercage in MCM-49 zeolite has large internal space
and a relatively small 10MR opening, it facilitates the reactions
that produce relatively large molecules, such as processes B and
D. Compared with the smaller molecule like ethylbenzene, the
Fig. 4. (A) Ethylene conversion vs. time on stream over HM, HMxy and HMxy-DMQ. (B)
Ethylene conversion vs. time on stream in different pores of HM. Reaction
conditions: 3.5 MPa, 220 °C, 12 of B/E molar ratio and 1.5 hꢂ1 of WHSV (ethylene).
6% at TOS of 12 h. Accordingly, the overall conversion of ethylene
over HM decreases from 96% (TOS = 1 h) to 73.5% (TOS = 12 h).
Therefore, the deactivation mainly takes place in the supercages
of HM.
Table 3
The selectivities of products in the alkylation of benzene with ethylene over different
samples.
Catalysts
HM
HMxy
69.8
HMxy-DMQ 0.3AT 0.025AT
Fast decrease in conversion at the early reaction stage is also
observed over MCM-22 zoelite when it is applied in xylene isomer-
ization [17], n-heptane cracking [30], toluene disproportionation
[8,31], methylcyclohexane transformation [2], alkylation reaction
[32] and butane isomerization [33]. Matias et al. attributed the fast
deactivation of the methylcyclohexane transformation over MCM-
22 to the formation of coke in the supercages [2]. Corma et al. sug-
gested that the fast decrease in strong acid sites resulted in the
deactivation of MCM-22 zeolite catalysts in the alkylation of isobu-
tane and 2-butene [32]. According to the theoretical modeling by
Sastre [34], the acid sites located within the supercages tend to
Ethylene conversion (mol%) 87.8
Product selectivity (mol%)
8.1
92.1
97.2
Ethylbenzene
m-Diethylbenzene
p-Diethylbenzene
o-Diethylbenzene
Toluene
96.4
0.8
0.9
1.1
0.09
0.06
0.57
0.08
97.5
0.6
0.6
0.9
0.14 5.40
0.02
84.1
0.2
1.1
0.1
96.0
1.2
0.9
96.2
1.1
0.9
1.4
1.1
0.09
0.13
C9 aromatics
0
0.04
0.27
0.10
0.07
0.41
0.09
C10+ aromatics
0.14 <0.01
0.10 9.10
Paraffins
Reaction conditions: 3.5 MPa, 220 °C, 12 of B/E molar ratio, 1.5 hꢂ1 of WHSV (eth-
ylene), 3 h of TOS.

72
K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
diffusion of heavy aromatic products out of the supercages is sig-
nificantly inhibited. Simultaneously, as the alkylbenzene has high-
er electronegativity than benzene molecule, further alkylation of
alkylbenzene with carbocations (process B) will happen if the
space in the pore is large enough. Furthermore, cracking, cycliza-
tion and transformation of alkylbenzene will produce aromatics
with condensed rings and block the supercages [19,22] (process
D). In contrast, ethylbenzene produced in the surface pockets
would readily diffuse out due to its limited diffusion constraints,
and further reactions could be inhibited.
78.7% for HM to 100% for 0.025AT and then decreases to 61.4%
for 0.3AT (Fig. 1). Due to the extraction of Si atoms from the frame-
work of zeolites, the Si/Al molar ratio decreases from 10.2 for HM
to 7.9 for 0.3AT [48,49].
The effect of the alkali treatment on the distribution of acid sites
within different pore systems of the MCM-49 zeolite is also inves-
tigated. As seen from Table 4, the ratio between the amount of B
acid site in the surface pockets and in the supercages (B(poc/sup)
)
increases over samples upon alkali treatment under more harsh
conditions. According to the results before [45], the opening of
supercages can be expanded by selectively extracting Si atoms on
T1 and T3 sites upon alkali treatment in NaOH solution. Within
expanded supercages, the diffusion of relatively large molecules
in/out of the supercages is improved. Part of acid sites in the super-
cages will not be effectively covered by the precoking process and
will be regarded as the acid sites similar to those in the surface
pockets, resulting in an increased B(poc/sup). In a word, the alkali-
treatment process introduces some pores similar to the surface
pockets, which is much favorable for the alkylation of benzene
with ethylene. However, the proportion of B acid sites in the sinu-
soidal pores exhibits little change under different alkali-treatment
conditions (Table 4), indicating that the alkali treatment has less
significant influence on the acid sites within the sinusoidal pores.
As shown in Fig. 5, bands related to the stretching vibration of
In the sinusoidal pores, the relatively weak acid sites [34] and
narrow channel do not favor the processes A, B and D that need
strong acid sites [7] or produce molecules with bigger size. HMxy-
with only sinusoidal pore accessible shows lower selectivities
DMQ
toward ethylation products and only trace quantity of C10+ is de-
tected (Table 3). The small amount of diethylbenzene produced
in sinusoidal pores is predominantly para-product. However, the
formation of paraffin by oligomerization of ethylene (process C)
and toluene by paring of ethylbenzene (process E) are slow involv-
ing relatively complex mechanisms. Compared with supercage and
pocket, the zigzag and longer channel of sinusoidal pore is a more
proper location for these kinds of reactions. Moreover, these two
reactions can be effectively catalyzed by relatively weak acid sites
[43,44], so the selectivities of toluene and paraffin (Table 3) on
HMxy-DMQ are distinctively higher than those over HM and HMxy
.
Al–OH (3668 cmꢂ1) and bridging hydroxyl groups (3618 cmꢂ1
)
In conclusion, the generation of ethylbenzene mainly occurs in
the surface pockets, and heavy aromatics produced in the super-
cages are responsible for the deactivation of MCM-49 zeolite.
Few opportunities for the alkylation reactions within the sinusoi-
dal pores can be attributed to the lack of strong acid sites and
the space restriction inside those small pores.
show little change over 0.3AT. On the other hand, the intensity of
peak at 3745 cmꢂ1, corresponding to the external Si–OH, increases
sharply upon more harsh conditions. This could be attributed to
the breakage of Si–O–Si bonds in basic solution [45]. Another inter-
esting observation is that 0.025ATxy and 0.3ATxy show similar Vmicro
as HMxy, and this Vmicro nearly equals to the volume of the sinusoi-
dal pores in the parent MCM-49 zeolite (Table 1), suggesting that
the destroyed micropores in NaOH solutions are mainly the super-
cages, and the structure of sinusoidal pores is preserved.
3.4. Effects of the alkali treatment on different pore systems of
MCM-49 zeolite
Fig. 4B shows that ethylene conversion within the surface pock-
ets and sinusoidal pores of HM is stable with TOS. Similar results
are also observed over 0.025AT and 0.3AT. Ethylene conversion
within the supercages of HM, 0.025AT and 0.3AT reduces by
31.3%, 12.1% and 22.5% with TOS from 1 to 3 h, respectively (Table
4). These results suggest that alkali treatment can prevent the
supercages in MCM-49 zeolite from deactivation.
In liquid alkylation of benzene with ethylene, the coke content
can be estimated from the change of pore volume by N2 adsorption
measurement (Table 4) [50,51]. Apart from the micropores consist-
Recently, the alkali treatment was found to be an effective
method to enhance catalytic performance of MCM-49 zeolite in
liquid alkylation of benzene with ethylene [45]. Herein, the reasons
for this will be elucidated in more detail with the methods men-
tioned in Section 3.1.
It has been reported that the RC of zeolite can be effectively
improved with the mild alkali-treatment conditions [38] and then
decreases with harsh treatment conditions [46,47]. We also ob-
serve a similar phenomenon. For example, the RC increases from
Table 4
The distribution of ethylene conversion, B acid sites in different channels, B(poc/sup), Si/Al molar ratio, amount of acid and amount of coke over HM, 0.025AT and 0.3AT.
Ba (%)
B(poc/sup)
Si/Alb
(molar ratio)
Acid amountc
(mmol/g)
Coketotal
Cokemicro
d
d
Catalysts
Pore
Eth. con. (%)
(cm3/g Catalyst)
(cm3/g Catalyst)
1 (h)
3 (h)
HM
Supercage
Pocket
Sinusoidal
26.2
61.4
8.4
18.0 (31.3%)
61.7 (ꢂ0.5%)
8.1 (3.6%)
34
38
28
1.12
1.52
1.92
10.2
10.5
7.9
0.79
0.79
0.80
0.18
0.17
0.11
0.13
0.14
0.08
0.025AT
0.3AT
Supercage
Pocket
Sinusoidal
21.5
69.8
8.5
18.9 (12.1%)
70.1 (ꢂ0.4%)
8.2(3.5%)
29
44
27
Supercage
Pocket
Sinusoidal
24.4
63.6
9.7
18.9 (22.5%)
63.9 (0.5%)
9.3 (4.1%)
25
48
27
Data listed within the parenthesis are the decreased percent of ethylene conversion at the TOS of 3 h when compared with the value at 1 h of TOS.
B
(poc/sup): the ratio of B acid site located in the surface pockets and in the supercages.
Reaction conditions: 3.5 MPa, 220 °C, 12 of B/E molar ratio, 1.5 hꢂ1 of WHSV (ethylene).
a
B acid site distribution, measured by Py-IR.
Measured by XRF.
Measured by NH3-TPD.
b
c
d
Amount of coke (generated in the liquid alkylation reaction) measured by N2 adsorption (Coketotal: the total amount of coke over samples, Cokemicro: coke located within
the micropores).

K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
73
contribution to the production of ethylation products, but it is
favorable to the oligomerization of ethylene.
Further studies show that the catalytic performance of MCM-49
zeolite for liquid alkylation of benzene with ethylene can effec-
tively improved by moderate alkali treatment due to the genera-
tion of some pores resembling the surface pockets.
Acknowledgment
We acknowledge the National Basic Research Program of China
(No. 2009CB623501) for financial support.
References
[1] C. Perego, P. Ingallina, Catal. Today 73 (2002) 3.
[2] P. Matias, J.M. Lopes, S. Laforge, P. Magnoux, P.A. Russo, M.M.L. Ribeiro Carrott,
M. Guisnet, F. Ramôa Ribeiro, J. Catal. 259 (2008) 190.
[3] J. Rigoreau, S. Laforge, N. Gnep, M. Guisnet, J. Catal. 236 (2005) 45.
[4] B. Onida, F. Geobaldo, F. Testa, F. Crea, E. Garrone, Micropor. Mesopor. Mater.
30 (1999) 119.
[5] D. Meloni, S. Laforge, D. Martin, M. Guisnet, E. Rombi, V. Solinas, Appl. Catal. A:
Gen. 215 (2001) 55.
Fig. 5. The FT-IR spectra of HM, 0.025AT and 0.3AT in the –OH stretching region.
HM: H-type MCM-49 zeolite. 0.025AT: H-type MCM-49 zeolite alkali-treated in
0.025 N NaOH solution for 15 min at the room temperature. 0.3AT: H-type MCM-49
zeolite alkali-treated in 0.3 N NaOH solution for 120 min at 75 °C.
[6] G.G. Juttu, R.F. Lobo, Micropor. Mesopor. Mater. 40 (2000) 9.
[7] X.D. Sun, Q.X. Wang, L.Y. Xu, S.L. Liu, Catal. Lett. 94 (2004) 75.
[8] P. Wu, T. Komatsu, T. Yashima, Micropor. Mesopor. Mater. 22 (1998) 343.
[9] A. Corma, V. Martínez-Soriab, E. Schnoevelda, J. Catal. 192 (2000) 163.
[10] J. Cheng, T. Degnan, J. Beck, Y. Huang, M. Kalyanaraman, J. Kowalski, C. Loehr,
D. Mazzone, Stud. Surf. Sci. Catal. 121 (1999) 53.
[11] G. Bellussi, G. Pazzuconi, C. Perego, G. Girotti, G. Terzoni, J. Catal. 157 (1995)
227.
[12] S. Lawton, M. Leonowicz, R. Partridge, P. Chu, M. Rubin, Micropor. Mesopor.
Mater. 23 (1998) 109.
[13] A. Corma, C. Corell, J. Pérez-Pariente, Zeolites 15 (1995) 2.
[14] A. Corma, V. Fornés, L. Forni, F. Márquez, J. Martínez-Triguero, D. Moscotti, J.
Catal. 179 (1998) 451.
[15] H. Du, D. Olson, J. Phys. Chem. B 106 (2002) 395.
[16] P. Ayrault, J. Datka, S. Laforge, D. Martin, M. Guisnet, J. Phys. Chem. B 108
(2004) 13755.
[17] S. Laforge, D. Martin, M. Guisnet, Appl. Catal. A: Gen. 268 (2004) 33.
[18] S. Laforge, D. Martin, M. Guisnet, Micropor. Mesopor. Mater. 67 (2004) 235.
[19] S. Laforge, D. Martin, J.L. Paillaud, M. Guisnet, J. Catal. 220 (2003) 92.
[20] S. Lawton, A. Fung, G. Kennedy, L. Alemany, C. Chang, G. Hatzikos, D. Lissy, M.
Rubin, H. Timken, S. Steuernagel, J. Phys. Chem. 100 (1996) 3788.
[21] J. Cheng, C. Smith, D. Walsh, US Patent 5493065, 1996 (to Mobil Oil
Corporation).
[22] P. Matias, J. Lopes, S. Laforge, P. Magnoux, M. Guisnet, F. Ramoa Ribeiro, Appl.
Catal. A: Gen. 351 (2008) 174.
[23] M. Guisnet, P. Ayrault, C. Coutanceau, M. Fernanda Alvarez, J. Datka, J. Chem.
Soc. Faraday Trans. 93 (1997) 1661.
[24] J.M. Bennett, C.D.Chang, S.L. Lawton, M.E., Leonowicz, D.N. Lissy, M.K. Rubin, B.
Cynwyd, US Patent 5236575, 1993 (to Mobil Oil Corporation).
[25] E. Besset, D. Meloni, D. Martin, M. Guisnet, L. Schreyeck. In: B. Delmon, G.F.
Froment (Eds.), Catalyst Deactivation 1999, Proc. 8th Int. Symp., 10–13
October 1999, vol. 126, p. 171.
[26] A. Corma, C. Corell, V. Fornés, W. Kolodziejski, J. Pérez-Pariente, Zeolites 15
(1995) 576.
[27] M. Maache, A. Janin, J.C. Lavalley, J.F. Joly, E. Benazzi, Zeolites 13 (1993) 419.
[28] E.P. Parry, J. Catal. 2 (1963) 371.
ing of sinusoidal pores and supercages in MCM-49 zeolite, there
are also mesopores originating from the aggregation of MCM-49
sheets and extra mesopores introduced in alkali treatment with
certain NaOH solutions. After liquid alkylation of benzene with
ethylene for 12 h, compared with HM, the total coke content of
0.025AT exhibits limited change, and only the proportion of coke
deposited in micropores is a bit higher. The amount of coke over
0.3AT, especially that in micropores, is obviously lower than that
over HM. Moreover, the selectivities of most products with rela-
tively small molecule size are not influenced by the alkali-
treatment process (Table 3), but the selectivity of C10+ aromatics
decreases from 0.57% over HM to only 0.27% over 0.3AT. Since
the mild alkali-treatment process can destroy part of aggregated
mesopores, the coke deposit in the mesopores of 0.025AT
decreases, accompanying with the increase in coke deposition in
micropores. Over 0.3AT, the diffusion condition is improved with
the introduction of mesopores by alkali-treatment, and the diffu-
sion path in the micropores is shortened by desilication [52–54].
Reactions that produce larger molecules such as processes B and
D will be inhibited, resulting in the decrease in the C10+ aromatics
as well as coke deposition in micropores. In conclusion, alkali treat-
ment is an efficient method to improve the catalytic performances
of MCM-49 zeolite by introducing mesopores into the zeolite, as
well as expanding the opening of the supercages.
[29] K. Song, J. Guan, S. Wu, Y. Yang, B. Liu, Q. Kan, Catal. Lett. 126 (2008) 8.
[30] D. Meloni, D. Martin, M. Guisnet, Appl. Catal. A: Gen. 215 (2001) 67.
[31] X. Ren, J. Liang, J. Wang, J. Porous Mater. 13 (2006) 353.
[32] A. Corma, A. Martinez, C. Martinez, Catal. Lett. 28 (1994) 187.
[33] H.J. Jung, S.S. Park, C.H. Shin, Y.K. Park, S.B. Hong, J. Catal. 245 (2007) 65.
[34] G. Sastre, V. Fornes, A. Corma, J. Phys. Chem. B 104 (2000) 4349.
[35] Y. Li, B. Xue, X. He, J. Mol. Catal. A: Chem. 301 (2009) 106.
[36] A. Albuquerque, L. Marchese, L. Lisi, H.O. Pastore, J. Catal. 241 (2006) 367.
[37] A. Corma, A. Martinez, C. Martinez, Appl. Catal. A: Gen. 134 (1996) 169.
[38] E. Roland, P. Kleinschmit, A. Kiss, F. Heindl, US Patent 5223240, 1993, (to
Degussa Aktiengesellschaft, Fed. Rep. of Germany).
[39] H. Darmstadt, C. Roy, S. Kaliaguine, T. Kim, R. Ryoo, Chem. Mater. 15 (2003)
3300.
[40] A. Geatti, M. Lenarda, L. Storaro, R. Ganzerla, M. Perissinotto, J. Mol. Catal. A:
Chem. 121 (1997) 111.
[41] J.P. Lange, A. Gutsze, J. Allgeier, H.G. Karge, Appl. Catal. 45 (1988) 345.
[42] K.A. Becker, H.G. Karge, W.D. Streubel, J. Catal. 28 (1973) 403.
[43] A. Feller, I. Zuazo, A. Guzman, J.O. Barth, J.A. Lercher, J. Catal. 216 (2003) 313.
[44] P.D. Hopkins, J. Catal. 12 (1968) 325.
[45] K. Liu, S. Xie, G. Xu, Y. Li, S. Liu, L. Xu, Appl. Catal. A: Gen. 383 (2010) 102.
[46] L. Mokrzycki, B. Sulikowski, Z. Olejniczak, Catal. Lett. 127 (2009) 296.
[47] L.L. Su, L. Liu, J.Q. Zhang, H.X. Wang, Y.G. Li, W.J. Shen, Y.D. Xu, X.H. Bao, Catal.
Lett. 91 (2003) 155.
4. Conclusions
Through two-step process of precoking and 2, 4-DMQ adsorp-
tion, distributions of acid sites in three different types of pore sys-
tems of MCM-49 zeolite have been determined. The ratio of B acid
sites in the surface pockets, in the supercages and in the sinusoidal
pores is about 1.4:1.2:1.0. Most L acid sites (ca. 73%) exist within
supercages.
Under the selected conditions for the liquid alkylation of
benzene with ethylene, more than 60% of ethylene conversion oc-
curs in the surface pockets, and less than 30% and 10% takes place
in the supercages and the sinusoidal pores, respectively. The roles
of pores have been proposed. The surface pockets are the main
positions for ethylbenzene production, and they also yield more
than half of diethylbenzene. The majority of C10+ aromatics are
produced in the supercages, and coke deposition within the super-
cages is the main reason for the fast deactivation of MCM-49
zeolite at the early reaction stage. The sinusoidal system has a little
[48] Y. Tao, H. Kanoh, L. Abrams, K. Kaneko, Chem. Rev. 106 (2006) 896.

74
K. Liu et al. / Journal of Catalysis 283 (2011) 68–74
[49] X. Wei, P.G. Smirniotis, Micropor. Mesopor. Mater. 97 (2006) 97.
[50] D.M. Bibby, N.B. Milestone, J.E. Patterson, L.P. Aldridge, J. Catal. 97 (1986) 493.
[51] J. Kim, M. Choi, R. Ryoo, J. Catal. 269 (2010) 219.
[52] J.C. Groen, J.A. Moulijn, J. Pérez-Ramírez, Micropor. Mesopor. Mater. 87 (2005)
153.
[53] J.C. Groen, W. Zhu, S. Brouwer, S.J. Huynink, F. Kapteijn, J.A. Moulijn, J. Pérez-
Ramírez, J. Am. Chem. Soc. 129 (2007) 355.
[54] J.C. Groen, J.A. Moulijn, J. Pérez-Ramírez, J. Mater. Chem. 16 (2006) 2121.