Selective production of propylene from methanol: Mesoporosity development in high silica HZSM-5

Article abstract of DOI:10.1016/j.jcat.2008.06.019

The mesoporosity development in high silica HZSM-5 was carried out by alkaline desilication treatment and soft template method, and its relationship with the catalytic performance of the modified catalysts in methanol-to-propylene reaction was studied. High propylene selectivity (42.2%) and propylene/ethylene ratio (10.1) were observed on the high silica HZSM-5 catalyst modified by alkaline desilication. The enhanced catalytic performance can be attributed to the newly created open mesopores on the surface of the zeolite crystals together with the low Bronsted acidity. A greater amount of mesoporosity could be readily formed in HZSM-5 crystals via the soft template route. However, the mesopores formed in this method are randomly distributed in the zeolite crystals and play a minor role in the molecular transport of the reaction, so the improvement in propylene selectivity and propylene/ethylene ratio of the catalyst is less evident. The higher propylene selectivity and propylene/ethylene ratio on modified HZSM-5 (especially by alkaline treatment) than unmodified HZSM-5 could be also related to different contributions of the methylaromatics route and olefins methylation/cracking route in MTP reaction.

Full text of DOI:10.1016/j.jcat.2008.06.019

Journal of Catalysis 258 (2008) 243–249
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Selective production of propylene from methanol: Mesoporosity development
in high silica HZSM-5
a,c,1
b,1
a
a
a
a
a,∗
Changsong Mei
, Pengyu Wen , Zhicheng Liu , Hongxing Liu , Yangdong Wang , Weimin Yang , Zaiku Xie ,
Weiming Hua , Zi Gao c
c,∗
a
Shanghai Research Institute of Petrochemical Technology, SINOPEC, 1658 PuDong Beilu, Shanghai 201208, PR China
School of Chemistry and Molecular Engineering, East China University of Science and Technology, Shanghai 200237, PR China
Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Fudan University, Shanghai 200433, PR China
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
The mesoporosity development in high silica HZSM-5 was carried out by alkaline desilication treatment
and soft template method, and its relationship with the catalytic performance of the modified catalysts in
methanol-to-propylene reaction was studied. High propylene selectivity (42.2%) and propylene/ethylene
ratio (10.1) were observed on the high silica HZSM-5 catalyst modified by alkaline desilication. The
enhanced catalytic performance can be attributed to the newly created open mesopores on the surface
of the zeolite crystals together with the low Brønsted acidity. A greater amount of mesoporosity could
be readily formed in HZSM-5 crystals via the soft template route. However, the mesopores formed in
this method are randomly distributed in the zeolite crystals and play a minor role in the molecular
transport of the reaction, so the improvement in propylene selectivity and propylene/ethylene ratio of
the catalyst is less evident. The higher propylene selectivity and propylene/ethylene ratio on modified
HZSM-5 (especially by alkaline treatment) than unmodified HZSM-5 could be also related to different
contributions of the methylaromatics route and olefins methylation/cracking route in MTP reaction.
Received 14 April 2008
Revised 13 June 2008
Accepted 15 June 2008
Available online 17 July 2008
Keywords:
HZSM-5
Methanol-to-propylene process
Propylene selectivity
Propylene/ethylene ratio
Mesoporosity
Diffusion
©
2008 Elsevier Inc. All rights reserved.
1
. Introduction
Much attention has been paid to MTO process employing ze-
olite catalysts [1,2,4–7]. So far, there are few reports concerning
MTP process [3,8]. One-pass selectivity to propylene and P/E ratio
are still low in the current researches on MTP reaction [8,9], while
high P/E ratio is vital to producing propylene from methanol em-
ploying recycling technics [8,10]. High P/E ratio can be achieved
under the condition of low methanol partial pressure resulted from
use of excessive carrier gas or feed with high water/methanol ratio.
However, too low methanol partial pressure is uneconomic from
the practical point of view. Chang et al. reported that decreasing
Brønsted acidity of HZSM-5 zeolite could improve its selectivity to
propylene in MTO reaction [11]. Prinz and Riekert observed that
the selectivity for formation of ethylene and propylene as well as
P/E ratio increased with decreasing HZSM-5 crystal size during
the catalytic conversion of methanol to hydrocarbons [12], indicat-
ing that shortening of the diffusion path in HZSM-5 zeolite may
increase the P/E ratio.
In the present work, two kinds of mesopore-modified high sil-
ica HZSM-5 zeolites were prepared by alkaline treatment and soft
template method. These samples were well characterized by vari-
ous methods and used in MTP reaction. Our main purpose is to get
a better insight into the influence of mesoporosity development on
diffusion and catalytic performance of the zeolite catalysts in MTP
reaction through comparison of the structural and catalytic prop-
erties of samples obtained by different modification methods.
Propylene is one of the most important commodity petro-
chemicals, as a raw material for the production of polypropy-
lene, polyacrylonitrile, acrolein and acrylic acid. Nowadays, propy-
lene is mainly produced as a by-product of petroleum refining
and of ethylene production by the naphtha steam-cracking pro-
cess. Due to the growing demand for propylene and the shortage
of petroleum resource in the future, new processes with high-
yield of propylene are required. Methanol-to-olefins (MTO) and
methanol-to-propylene (MTP) processes are promising alternative
ways for the production of propylene instead of petroleum route,
since methanol can be easily produced from natural gas and coal.
So, these two new routes have attracted significant attention [1–3].
As compared to the traditional MTO process, where ethylene is
the main product, high propylene to ethylene (P/E) ratio is the
most distinguishable feature for MTP process that enables the high
propylene yield in the recirculation process.
* Corresponding authors. Faxes: +86 21 68462283 (Z. Xie), +86 21 65641740
W. Hua).
(
(
E-mail addresses: xiezk@mail.sript.com.cn (Z. Xie), wmhua@fudan.edu.cn
W. Hua).
1
Changsong Mei and Pengyu Wen contributed equally to this work.
0
021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.06.019

244
C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
of the absorbance peaks at 1540 and at 1450 cm 1, respectively.
The bulk Si/Al molar ratios of the samples were determined by in-
ductively coupled plasma (ICP) spectroscopy.
−
2
. Experimental
2.1. Catalyst preparation
Parent ZSM-5 zeolites with different Si/Al molar ratios were
2.3. Catalytic test
synthesized hydrothermally from the corresponding aluminosili-
cate gels. Tetrapropylammonium hydroxide (TPAOH) was used as
the template. The silicon source was colloidal silica, and the alu-
minum source was NaAlO2. Both colloidal silica and NaAlO2 were
from Sinopharm Chemical Reagent Co., Ltd. The molar composition
of the synthesis mixture was xSiO2:1Al2O3:8TPAOH:800H2O, where
◦
The MTP reaction was carried out at 470 C in a flow-type
fixed-bed microreactor under atmospheric pressure. The catalyst
load was 3 g and the weight hourly space velocity (WHSV) for
methanol was 1 h . Water vapor was used as diluent and the par-
tial pressure of methanol was 0.5 atm. The reaction products were
analyzed by an on-line gas chromatograph with a 50-m Poraplot Q
capillary column and a flame ionization detector (FID).
−1
◦
x was 20, 80, 140 and 220. After being stirred for 7 h at 25 C, the
◦
gel was transferred into an autoclave and crystallized at 180 C for
◦
5
days. The product was filtered, washed, dried at 110 C overnight
◦
3. Results
and then calcined at 550 C for 5 h to remove the template. The
parent ZSM-5 zeolites were turned into the H-form by three con-
secutive ion exchanges in 1 M HCl solution with a solution/zeolite
ratio of 10 cm /g at 90 C for 9 h. The obtained HZSM-5 samples
were labeled as S1, S2, S3 and S4, of which the Si/Al molar ratio of
the starting gel was 10, 40, 70 and 110, respectively. The S3 sample
The XRD patterns of the samples are shown in Fig. 1. It is ev-
ident that all the samples have a typical MFI structure [18]. The
intensity of the diffraction peaks of S5 sample is a bit lower than
the others, showing a slight decrease of crystallinity after desili-
cation. With S3 as the reference, the crystallinity loss of S5 was
calculated to be about 10% (see Table 1), suggesting that the frame-
work structure of HZSM-5 was reasonably retained after alkaline
treatment with aqueous solution of Na2CO3 and then acid treat-
ment with HCl. The decrease in XRD crystallinity of S5 is most
likely caused by the formation of framework defects through de-
silication, as demonstrated in 27Al MAS NMR spectra presented in
the following. The high relative crystallinity (97%) of the S6 sam-
ple indicates that use of starch as soft template for the generation
of mesopores in HZSM-5 crystals is rather successful.
3
◦
was subjected to mesoporosity generation by alkaline treatment in
3
0
.45 M Na2CO3 solution with a solution/zeolite ratio of 15 cm /g
◦
at 75 C for 30 h [13–17]. The obtained product was then filtered,
washed, dried and then exchanged in 1 M HCl solution for three
3
◦
times with a solution/zeolite ratio of 10 cm /g at 90 C for 9 h
to get the final mesopore-modified HZSM-5 which was designated
as S5. For comparison, another kind of mesoporous HZSM-5 with
Si/Al molar ratio of 76 (labeled as S6) was synthesized by the soft
template route. Synthesis of S6 was in the same way as S3, except
that 5 wt% starch was added into the gel with molar composition
of 140SiO2:1Al2O3:8TPAOH:800H2O.
27
Al MAS NMR has been shown to be a powerful technique in
the characterization of the local coordination environment of alu-
minum atoms in zeolites. It can discriminate between framework
aluminum atoms in tetrahedral coordination and extra-framework
aluminum atoms in octahedral coordination or penta-coordination
2.2. Catalyst characterization
X-ray powder diffraction (XRD) patterns were recorded on a
(
EFAL) [19–22]. 27Al MAS NMR spectra of S3, S5 and S6 samples are
Rigaku D/max-1400 diffractometer with Ni-filtered CuKα radiation
operated at 40 kV and 100 mA. The relative crystallinity of the
sample was determined by measuring the intensity of its peaks
at d = 3.851, 3.823, 3.798, 3.732 and 3.711 nm and comparing
the sum of the intensities with that of the reference sample S3.
The N2 adsorption/desorption isotherms of the samples were mea-
sured by using a Micromeritics TriStar 3000 system at liquid-N2
temperature. The Brunauer–Emmett–Teller (BET) surface area was
calculated from the adsorption isotherm, and the mesopore size
distribution was derived from the adsorption isotherm according
to the BJH model. The surface morphology and crystallite size of
the samples were determined by scanning electron microscopy
(SEM) recorded on a JEOL JSM-35C microscope. Transmission elec-
tron microscopy (TEM) was recorded on a JEM-CX-II instrument.
The samples were prepared by dispersing the powder products as
a slurry in ethanol, which was then deposited and dried on a holey
27
carbon film on a Cu grid. Al magic-angle spinning nuclear mag-
netic resonance (MAS NMR) spectra were recorded on a Varian 400
spectrometer. A resonance frequency of 101.19 MHz, a recycle delay
of 0.6 s, pulse widths of 0.4 μs and a spinning rate of 5 kHz were
Fig. 1. XRD patterns of unmodified and mesopore-modified HZSM-5 zeolites.
Table 1
applied. 2 Al chemical shifts were reported relative to Al(H2O)6
7
3+
Physicochemical properties of the catalysts
.
Catalyst Si/Al
SBET
Vmicro
Vmeso
(cm
B acidb L acidb Crystallinity
The NMR analysis was carried out with the samples prehydrated in
a wet desiccator for several days. The acidic properties (Brønsted
and Lewis acid sites) of the samples were investigated by Fourier
transform infrared (FTIR) spectra of adsorbed pyridine in an in
situ cell with CaF2 windows. Self-supporting wafers of the samples
ratio (m2
a
g
−1
)
(cm
3
g
−1
)
3
−1
g )
(a.u.)
(a.u.)
(%)
S1
S2
S3
S4
S5
S6
11
42
72
114
78
76
330
329
347
339
344
367
0.12
0.12
0.12
0.12
0.11
0.11
0.11
3.4
2.3
1.9
1.1
1.2
1.4
2.6
2.2
1.8
1.2
2.8
1.7
96
0.10
0.10
0.10
0.27
0.47
100
100
100
90
◦
−2
Pa and then cooled
were pretreated at 400 C for 2 h under 10
◦
◦
to 200 C. After the pyridine adsorption at 200 C for 15 min and
evacuation at 200 C for 30 min, IR spectra were recorded on a
Bruker IFS 88 spectrometer with a resolution of 4.0 cm . Brøn-
sted and Lewis acidities were quantified with the integrated areas
97
◦
a
Si/Al molar ratio determined by ICP analysis.
−1
b
Brønsted and Lewis acidities were quantified according to the integrated areas
−1
of the absorbance peaks at 1540 and at 1450 cm , respectively.

C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
245
Fig. 3. Pore size distributions of unmodified and mesopore-modified samples.
Fig. 2. 27Al MAS NMR spectra of unmodified and mesopore-modified samples.
shown in Fig. 2. The intense signal at 54 ppm is assigned to tetra-
hedrally coordinated framework aluminum in HZSM-5 zeolite, and
the weak signal at 0 ppm can be attributed to octahedral extra-
framework aluminum. There are almost no EFAL atoms in the S3
sample. A weak resonance at 0 ppm observed for the S5 sample
suggests the existence of a small amount of EFAL atoms in this
sample. This shows that the removal of framework Si in the zeolite
upon alkaline treatment is probably accompanied by the cleavage
of some Si–O–Al bonds, although the Si–O–Si bonds in the absence
of neighboring Al ions are easier to hydrolyze and dissolve [23,24].
As a consequence, a small quantity of four-coordinated framework
Al atoms are transformed into six-coordinated extra-framework
ones in S5 sample after alkaline treatment. In the meantime, the
number of EFAL atoms in S6 is significantly lower than that in S5.
This is in agreement with the XRD crystallinity results.
The textural properties of the samples are listed in Table 1.
The BET surface area of the untreated HZSM-5 (S3 sample) is
2
−1
3
47 m g . After the alkali-treatment there is almost no change
2
−1
for S5 sample). The S6 sample
in the BET surface area (344 m g
exhibits a somewhat larger surface area among all the samples.
HZSM-5 zeolites with different Si/Al ratios in the 10–114 range
3
−1
(
S1–S4 samples) exhibit mesopore volumes of ca. 0.10 cm g
,
which are contributions from intercrystalline voids present be-
tween the nanophase crystalline particles [25]. After treatment
of HZSM-5 in aqueous Na2CO3, the micropore volume decreases
3
−1
slightly from 0.12 to 0.11 cm g , which is in agreement with the
XRD results, whereas the mesopore volume increases significantly,
3
−1
i.e. from 0.10 to 0.27 cm g . The pore size distribution curve in
Fig. 3 demonstrates that the diameter of the newly created meso-
pores is in the range of 20–55 nm, centered around 35 nm. The
significant increase in the mesoporosity of ZSM-5 zeolites upon
alkaline treatment with aqueous NaOH was reported by several
groups [13–17]. Our result is similar to the reported ones, although
a milder and more controllable alkaline solution of Na2CO3 is used.
3
−1
An even larger mesopore volume of 0.47 cm g
is observed for
the S6 sample (Table 1), with diameter in the range of 10–50 nm,
centered around 22 nm (Fig. 3).
Fig. 4 shows the surface morphology of representative samples.
SEM images of S1, S2, S3 and S4 samples are similar. The parti-
cles of all the samples are sphere-like, and the distribution of the
particle size seems to be rather uniform. The average particle size
is ca. 500 nm in diameter for S1–S5 samples. The S6 sample ex-
hibits a smaller average particle size of ca. 200 nm. The external
surface of the as-synthesized HZSM-5 zeolites is clear and smooth.
After treatment in alkaline media, the external zeolite surface be-
comes rough and rugged (Fig. 4b). It can be seen clearly that there
are many nanoscale open holes on the exterior surface of S5. Fur-
thermore, TEM image of the S5 sample (Fig. 5b) shows that the
open holes are distributed on the surface without penetrating into
Fig. 4. SEM images of representative samples: (a) S3; (b) S5; (c) S6.
the crystals, i.e., the desilication takes place mainly on the surface
and the interior of the zeolite crystals is almost unaffected upon
alkaline treatment. The existence of these surface open holes is
probably the main cause for the facilitation of molecular transport

246
C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
Fig. 6. Product selectivity of S5 as a representative catalyst for MTP reaction as a
function of time: (Q) C3H6; (!) C2H4; (a) C4H8; (") aromatics; (2) C1–C4 satu-
rated hydrocarbons; (1) C5 and higher hydrocarbons excluding aromatics. Reaction
◦
−1
conditions: T = 470 C, WHSV = 1 h , PCH3OH = 0.5 atm, H2O:CH3OH = 1:1.
Fig. 7. Catalytic stability of unmodified and mesopore-modified HZSM-5 zeolites for
MTP reaction: (2) S1; (!) S2; (Q) S3; (e) S4; (1) S5; (") S6. Reaction conditions:
◦
−1
T = 470 C, WHSV = 1 h , PCH3OH = 0.5 atm, H2O:CH3OH = 1:1.
is distinct, and its effect on the catalyst performance in MTP reac-
tion is worthy to be further investigated.
Pyridine adsorption was followed by infrared spectroscopy to
identify the number and nature of acid sites in the catalysts. FTIR
◦
spectra of the catalysts after pyridine adsorption at 200 C and
◦
subsequent evacuation at 200 C exhibit two characteristic bands
−1
at about 1540 and 1450 cm , which are ascribed to pyridinium
ions chemisorbed on Brønsted acid sites and coordinatively bound
pyridine on Lewis acid sites, respectively [26–28]. Brønsted and
Lewis acidities have been quantified according to the integrated
areas of the peaks at 1540 and at 1445 cm , respectively, and
the data are summarized in Table 1. The number of Brønsted and
Lewis acid sites present over S1–S4 catalysts follows the order of
S1 > S2 > S3 > S4, as the same as that of the Si/Al ratio of these
catalysts. A comparison of the acidity data for S3 and S5 demon-
strates that alkaline treatment of the zeolite leads to a decrease
in Brønsted acidity and an increase in Lewis acidity. This is due
to the transformation of a part of four-coordinated framework Al
atoms into six-coordinated extra-framework ones, as evidenced by
−1
Fig. 5. TEM images of representative samples: (a) S3; (b) S5; (c) S6.
in MTP reaction and the improvement of catalytic performance of
the HZSM-5 catalyst. In contrast, the outer surface of the crys-
tals for the S6 sample is more or less smooth, as illustrated in
Fig. 4c. However, there are many nanoscale mesopores randomly
distributed in the bulk of the zeolite crystals, as evidenced by the
TEM image (Fig. 5c). These mesopores are distributed inside the
zeolite crystals and connected to the reaction system only through
the narrow zeolite microporous channels. The morphological dif-
ference of these two types of mesopore-modified HZSM-5 catalysts
27
Al MAS NMR analysis (Fig. 2). The S6 catalyst has slightly more
Brønsted acid sites and less Lewis acid sites, which is consistent
with its higher relative crystallinity.
The changes in product selectivities with time on stream for S5
as a representative catalyst are presented in Fig. 6. At the initial
stage the selectivities to propylene and butylene increase slightly,
whereas the selectivities to ethylene, saturated hydrocarbons and
aromatics decrease slightly. After 10 h on stream the reaction

C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
247
Table 2
MTP reaction over unmodified and mesopore-modified HZSM-5 catalystsc
Catalyst
Conv.
%)
Selectivity (C-mol%)
P/E
(
C1–4a
+b
C2H4
C3H6
C4H8
C
Aromatics
5
S1
S2
S3
S4
S5
S6
99.6
99.8
99.6
99.8
99.6
99.8
20.5
10.0
9.55
8.77
5.72
8.82
15.3
11.1
10.1
9.90
4.18
9.92
26.7
35.7
37.0
36.1
42.2
38.8
4.25
17.3
20.5
21.1
21.4
21.8
16.6
16.3
17.1
18.9
23.9
17.0
16.7
9.59
5.75
5.23
2.60
3.66
1.75
3.22
3.66
3.65
10.1
3.91
Scheme 1. The reaction pathway for methanol conversion.
4. Discussion
a
The catalytic conversion of methanol to olefins and hydrocar-
bons over acidic HZSM-5 zeolite consists of three categories of re-
actions: fast equilibrium of methanol (MeOH) with dimethyl ether
C1–C4 saturated hydrocarbons.
b
c
C5 and higher hydrocarbons excluding aromatics.
Reaction conditions: T = 470 ◦C, WHSV = 1
h
−1
,
PCH3OH = 0.5 atm, H2O:
CH3OH = 1:1.
(DME) over Brønsted acid sites, and thus MeOH and DME can be
treated as a single kinetic species or lump [1,11,29,30]; the equilib-
rium mixture of MeOH and DME is then converted to the primary
product of ethylene and propylene [1,31–33]; the subsequent con-
version of the primary products into a mixture of higher olefins,
paraffins and aromatics that can be lumped together (referred to
as others) [1,32]. The simplified reaction pathway is shown in
Scheme 1, where k1 and k2 are formation rate constants for ethy-
lene and propylene respectively, and k3 and k4 are consumption
constants for ethylene and propylene respectively.
The disappearance of oxygenates (mixture of MeOH and DME)
and olefins (C2H4 and C3H6) is assumed to be first order [11,
32], then Eqs. (1), (2) and (3) are obtained, where kobs is the
rate constant of conversion of oxygenates. The parameter kobs can
be determined from conventional first-order semilog plots of oxy-
genates vs τ. If Eqs. (2) and (3) are divided by (1), Eqs. (4) and (5)
are obtained after integrating, where XMeOH/DME is the conversion
of oxygenates. A series of experimental data, i.e. [C2H4], [C3H6],
reaches steady state. The methanol conversion over S5 catalyst
drops obviously after on stream for 100 h, as shown in Fig. 7.
However, the product selectivities are similar to those at steady
state. The other catalysts display the same variation trend in prod-
uct selectivities with time on stream. Table 2 shows the results
of MTP reaction at steady state with water vapor as diluent gas
◦
on various HZSM-5 catalysts at 470 C and atmospheric pressure.
For the unmodified S1–S4 series catalysts, the propylene selectivity
first increases and then decreases with an increase in Si/Al ra-
tio. A maximum of 37.0% selectivity to propylene is achieved on
S3 catalyst. The ethylene selectivity decreases gradually from 15.3
to 9.90%, as the Si/Al ratio is increased. As a result, the P/E ra-
tio initially increases with Si/Al ratio and then levels off at a P/E
ratio of ca. 3.7. In addition, the selectivity to aromatics decreases
with increasing Si/Al ratio. Since S1–S4 series catalysts have similar
particle sizes and textural properties, their difference in Brønsted
acidity could be the main cause for the change in product dis-
tribution of the catalysts in MTP reaction. A decrease in Brønsted
acidity of the catalysts favors the formation of propylene in the
reaction and brings about an increase in P/E ratio from 1.75 to
[MeOH/DME] and XMeOH/DME can be acquired by changing the cat-
alyst load, namely, the contact time. Fitting the experimental data
using Eqs. (4) and (5) gives the values of k1 − k4.
d[MeOH/DME]
−
= kobs[MeOH/DME],
(1)
(2)
(3)
3.66. However, as the Si/Al ratio increases to above 72, the effect of
dτ
Brønsted acidity becomes less important. Meanwhile, the influence
of mesopore-modification on product selectivity of the catalyst ex-
ceeds that of the acidity. For S5 catalyst, the propylene selectivity
increases from 37.0 to 42.2% after modification, while the ethylene
selectivity decreases from 10.1 to 4.18%, leading to a substantial
increase of P/E ratio from 3.66 to 10.1. The increment of P/E ra-
tio can be mainly attributed to the increase in mesoporosity of
the HZSM-5 zeolite resulting from alkaline treatment, which facil-
itates the molecular transport in the MTP reaction exceedingly. It
is noticed that the selectivity to propylene and P/E ratio are also
improved on S6 catalyst. However, the improvement is obviously
less striking than that on S5 catalyst. The probable reason is that
the mesopores or cavities generated inside the zeolite crystals of
S6 catalyst by using starch as soft template are not as efficient as
the opened mesopores on the surface of S5 for the enhancement
of diffusion of the gas molecules in the reaction process.
d[C2H4]
= k1[MeOH/DME] − k3[C2H4],
dτ
d[C3H6]
= k [MeOH/DME] − k [C H ],
2
4
3
6
dτ
[
C2H4]
k1/kobs[1 − (1 − XMeOH/DME)k3/kobs−1]
=
=
,
.
(4)
(5)
k3
[MeOH/DME]
− 1
kobs
[
C3H6]
k2/kobs[1 − (1 − XMeOH/DME)k4/kobs−1]
k4
kobs
[
MeOH/DME]
− 1
For kinetic measurements, the contact time in this work was
changed in the range of 0.04 to 1.28 s. The formation and con-
sumption rate constants of ethylene and propylene were calculated
through fitting the experimental data with Eqs. (4) and (5). The
effect of Si/Al ratio of HZSM-5 zeolite on the rate constants is
shown in Fig. 8. As the Si/Al ratio increases, both k1 and k2 de-
cline, but k2 is always greater than k1, indicating that propylene
could be the dominant product on this type of catalysts. On the
other hand, both k3 and k4 decrease with increasing Si/Al ratio,
but k4 drops much quicker than k3, suggesting that the selectiv-
ity to propylene is more sensitive to the Brønsted acidity of the
catalysts. Increasing the Si/Al ratio and reducing the Brønsted acid-
ity of the catalysts, the consumption rate of propylene decreases
more sharply than that of ethylene, hence the P/E ratio of the
product is increased. However, the k4 curve levels off at Si/Al ra-
tio above 72, and the increase in selectivity to propylene and P/E
ratio becomes insignificant. Therefore, adjusting the Si/Al ratio and
Brønsted acidity of HZSM-5 catalysts can improve the product dis-
tribution in MTP reaction, but the amplitude of improvement is
The catalytic stability of unmodified and mesopore-modified
◦
HZSM-5 zeolites for MTP reaction was tested at 470 C, and the re-
sults are shown in Fig. 7. It can be seen that the catalytic stability
of S1–S4 series catalysts follows the order of S4 > S3 > S2 > S1.
Compared with S3 catalyst, the mesopore-modified catalysts (S5
and S6) exhibit similar stability. This indicates that base treatment
does not improve the stability of the catalyst, although the propy-
lene selectivity and P/E ratio are significantly increased. As shown
in Fig. 7, the methanol conversion drops abruptly after on stream
for some period. Coke deposition is the reason for catalyst deacti-
vation [1]. When the coke accumulates to some amount, it would
block the micropores of HZSM-5 and subsequently results in a
sharp deactivation of the catalyst.

248
C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
on this novel insight, in addition to improved diffusion, another
cause for the increase in propylene selectivity upon the creation
of mesopores in the zeolite crystals could be that the contribution
of the olefins route to the propylene formation increases on the
mesopore-modified HZSM-5 catalysts. The increased selectivity to
C5 and higher saturated hydrocarbons observed for S5 catalyst is
an indirect evidence for improved contribution of the olefins route
in MTP reaction over this catalyst, according to suggested dual cy-
cle concept by Bjørgen and co-workers [42]. The lower ethylene
selectivity observed for S5 catalyst may be a consequence of de-
creased contribution of the methylaromatics route in MTP reaction
over this catalyst. In combination with higher propylene selectivity
for S5 catalyst, higher P/E ratio was achieved on the high silica
HZSM-5 catalyst modified by alkaline treatment.
5
. Conclusions
Fig. 8. Effect of Si/Al ratio of unmodified HZSM-5 zeolites on formation and con-
sumption rate constants of propylene and ethylene: (2) k1; (") k2; (1) k3; (!) k4.
Two kinds of mesopore-modified high silica HZSM-5 zeolites
were prepared by alkaline treatment and soft template method. Af-
ter alkali treatment, the structure of zeolite framework is scarcely
damaged, and open holes with diameter of 20–55 nm are created
on the zeolite crystal surface. These newly created open mesopores
enhance the diffusion of the primary olefin products, in partic-
ular propylene and butylene, and inhibit undesirable secondary
reactions. The propylene selectivity and P/E ratio of the HZSM-
limited due to reaction kinetic aspects. This explains that the max-
imum P/E ratio reached for the S1–S4 series catalysts in this work
is less than 4.
The creation of mesopores in the zeolite crystals shortens the
diffusion path of the primary olefin products and facilitates the
removal of the olefins, in particular propylene and butylene with
larger molecular sizes, from the reactive acid sites on the cata-
lyst. As a result, the reaction equilibrium shifts to the formation
of propylene and butylene, and also the probabilities that these
olefins further form higher olefins, paraffins, aromatics and naph-
thenes via various secondary reactions on the acid sites of the
catalysts are reduced. This leads to the increased selectivities to
propylene, butylene and P/E ratio for S5 catalyst as compared
with S3 catalyst (see Table 2). It is also interesting to note that
different mesopore-modification methods of HZSM-5 zeolite cata-
lyst may produce different aftereffects in MTP reaction. Besides the
quantity of the mesopores created and the integrity of the zeolite
crystals after treatment, a more important criterion of the modifi-
cation is the openness of the mesopores. The mesopores in S6 cat-
alyst created via the soft template route are randomly distributed
in the zeolite particles and surrounded by the narrow microporous
channels, so they are more diffusion limited than the open meso-
pores on the surface of S5 catalyst. This explains our experimental
results that the improvement in selectivity to propylene and P/E
ratio for S6 catalyst is not so evident as that for S5 catalyst. The
lower selectivity to aromatics observed for S5 and S6 catalysts is
a consequence of decreased probabilities for secondary reactions
of propylene and butylene on the acid sites of the catalysts, be-
cause the creation of mesopores in the zeolite crystals facilitates
the removal of these olefins. Although the soft template method
is more simple and popular for mesopore development in practice,
the present work proves that to create open mesopores in high sil-
ica HZSM-5 zeolite by alkaline treatment is probably a better way
to fulfill the selectivity requirements in MTP reaction.
5
catalyst prepared by alkaline treatment reached 42.2% and 10.1,
respectively. This P/E ratio is about thrice as large as that of or-
dinary high silica HZSM-5 catalyst. A large amount of mesopores
with diameter of 10–50 nm are formed in HZSM-5 prepared by
using starch as soft template. However, these mesopores locate in-
side the zeolite body and play a limited role in the diffusion of gas
molecules, so the change in product selectivity for MTP reaction
is insignificant on this type of modified HZSM-5 catalyst. The in-
crease in propylene selectivity and P/E ratio upon the creation of
mesopores in the zeolite crystals, especially by alkaline treatment,
could be also due to different contributions of the methylaromat-
ics route and olefins methylation/cracking route in MTP reaction
on the modified and unmodified HZSM-5 catalysts. The present
work demonstrates that the creation of open mesopore cavities or
channels in HZSM-5 zeolite is as important as the right adjust-
ment of its Brønsted acidity in the design and preparation of a
good HZSM-5 catalyst for MTP reaction.
Acknowledgments
This work was financially supported by the Major State Basic
Research Development Program of China (Grants 2003CB615801,
2003CB615802 and 2006CB806103).
References
[1] M. Stöcker, Microporous Mesoporous Mater. 29 (1999) 3.
[
2] J.Q. Chen, A. Bozzano, B. Glover, T. Fuglerud, S. Kvisle, Catal. Today 106 (2005)
03.
1
[3] M. Hack, U. Koss, P. König, M. Rothaemel, H.D. Holtmann, US Patent 7 015 369
B2, 2006, to MG Technologies AG.
Recently, a hydrocarbon-pool mechanism has been suggested to
explain the formation of light olefins for the MTO process under
steady-state conditions on acidic zeolite catalysts [34–39]. Accord-
ing to this mechanism, polymethylbenzenes or their protonated
counterparts, which are referred to as the hydrocarbon pool, serve
as platforms to which methanol can bind and from which pri-
mary olefin products can dissociate. More recently, it has been
shown that, besides the methylaromatics route, which is respon-
sible for both ethylene and propylene formation, an additional
olefins methylation/cracking route accounts for part of propylene
formed on HZSM-5 zeolite catalyst. On the other hand, ethylene
production is little influenced by the olefins route [40–42]. Based
[4] F.A. Wunder, E.I. Leupold, Angew. Chem. Int. Ed. Engl. 19 (1980) 126.
[
[
[
5] W.W. Kaeding, S.A. Butter, J. Catal. 61 (1980) 155.
6] C.D. Chang, Catal. Rev. Sci. Eng. 26 (1984) 323.
7] P.T. Barger, B.V. Vora, US Patent 6 534 692, 2003, to UOP LLC.
[8] T. Zhao, T. Takemoto, N. Tsubaki, Catal. Commun. 7 (2006) 647.
9] T. Zhao, T. Takemoto, Y. Yoneyama, N. Tsubaki, Chem. Lett. 34 (2005) 970.
10] S. Michael, S. Friedrich, B. Goetz, B. Henning, M. Friedrich-Wilhelm, Eur. Patent
448000, 1991, to Sued Chemie AG and Metallgesellschaft AG.
[
[
0
[11] C.D. Chang, C.T.W. Chu, R.F. Socha, J. Catal. 86 (1984) 289.
[12] D. Prinz, L. Riekert, Appl. Catal. 37 (1988) 139.
[13] M. Ogura, S. Shinomiya, J. Tateno, Y. Nara, M. Nomura, E. Kikuchi, M. Matsukata,
Appl. Catal. A 219 (2001) 33.
[14] L. Su, L. Liu, J. Zhuang, H. Wang, Y. Li, W. Shen, Y. Xu, X. Bao, Catal. Lett. 91
(2003) 155.

C. Mei et al. / Journal of Catalysis 258 (2008) 243–249
249
[
[
[
15] J.C. Groen, L.A.A. Peffer, J.A. Moulijn, J. Pérez-Ramírez, Colloid Surf. A 241 (2004)
3.
16] J.C. Groen, J.A. Moulijn, J. Pérez-Ramírez, Microporous Mesoporous Mater. 87
2005) 153.
17] Y. Tao, H. Kanoh, K. Kaneko, Adsorption 12 (2006) 309.
[27] N.Y. Topsoe, K. Pedersen, E.G. Derouane, J. Catal. 70 (1981) 41.
[28] C.A. Emeis, J. Catal. 141 (1993) 347.
5
[29] J.R. Anderson, T. Mole, V. Christov, J. Catal. 61 (1980) 477.
[30] C.D. Chang, Chem. Eng. Sci. 35 (1980) 619.
[31] T.R. Forester, R.F. Howe, J. Am. Chem. Soc. 109 (1987) 5076.
[32] F.J. Keil, Microporous Mesoporous Mater. 29 (1999) 49.
[33] T.Y. Park, G.F. Froment, Ind. Eng. Chem. Res. 40 (2001) 4172.
[34] M. Seiler, W. Wang, A. Buchholz, M. Hunger, Catal. Lett. 88 (2003) 187.
(
[
[
18] R.J. Argauer, G.R. Landolt, US Patent 3 702 886, 1972, to Mobil Oil Corp.
19] A.P. Kentgens, M.K. Scholl, W.S. Veeman, J. Phys. Chem. 87 (1983) 4357.
[
20] J.B. Nagy, Z. Gabelica, G. Debras, E.G. Derouane, J.P. Gilson, Zeolites 4 (1984)
1
33.
[35] J.F. Haw, W. Song, D.M. Marcus, J.B. Nicholas, Acc. Chem. Res. 36 (2003) 317.
[36] M. Bjorgen, U. Olsbye, S. Svelle, S. Kolboe, Catal. Lett. 93 (2004) 37.
[37] U. Olsbye, M. Bjorgen, S. Svelle, K.P. Lillerud, S. Kolboe, Catal. Today 106 (2005)
108.
[
[
[
21] F. Deng, Y. Du, C. Ye, J. Wang, T. Ding, H. Li, J. Phys. Chem. 99 (1995) 15208.
22] W. Liu, Y. Xu, S. Wong, J. Qiu, N. Yang, J. Mol. Catal. A 120 (1997) 257.
23] T. Sano, Y. Nakajima, Z.B. Wang, Y. Kawakami, K. Soga, A. Iwasaki, Microporous
Mater. 12 (1997) 71.
[38] J.F. Haw, D.M. Marcus, Top. Catal. 34 (2005) 41.
[
24] J.C. Groen, L.A.A. Peffer, J.A. Moulijn, J. Pérez-Ramírez, Chem. Eur. J. 11 (2005)
[39] W. Wang, Y. Jiang, M. Hunger, Catal. Today 113 (2006) 102.
[40] S. Svelle, F. Joensen, J. Nerlov, U. Olsbye, K.P. Lillerud, S. Kolboe, M. Bjørgen,
J. Am. Chem. Soc. 128 (2006) 14770.
[41] S. Svelle, U. Olsbye, F. Joensen, M. Bjørgen, J. Phys. Chem. C 111 (2007) 17981.
[42] M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino, L. Palumbo,
S. Bordiga, U. Olsbye, J. Catal. 249 (2007) 195.
4
983.
[
25] R. Ravishankar, C. Kirschhock, B.J. Schoeman, P. Vanoppen, P.J. Grobet, S. Storck,
W.F. Maier, J.A. Martens, F.C. De Schryver, P.A. Jacobs, J. Phys. Chem. B 102
(1998) 2633.
[26] E.P. Parry, J. Catal. 2 (1963) 371.