Confinement effect and synergistic function of H-ZSM-5/Cu-ZnO-Al 2O3 capsule catalyst for one-step controlled synthesis

Article abstract of DOI:10.1021/ja101882a

Dimethyl ether (DME) is an industrially important intermediate, as well as a promising clean fuel, but the effective production through traditionally consecutive steps from syngas to methanol and then to DME has been hindered by the poorly organized structure of the conventional physical mixture catalyst. Here, a novel zeolite capsule catalyst possessing a core-shell structure (millimeter-sized core catalyst and micrometer-sized acidic zeolite shell) was proposed initially through a well-designed aluminum migration method using the core catalyst as the aluminum resource and for the first time was applied to accomplish the DME direct synthesis from syngas. The selectivity of the expected DME on this zeolite capsule catalyst strikingly exceeded that of the hybrid catalyst prepared by the traditional mixing method, while maintaining the near-zero formation of the unexpected alkanes byproduct. The preliminary methanol synthesis reaction on the core catalyst and the following DME formation from methanol inside the zeolite shell cooperated concertedly and promoted mutually. This zeolite capsule catalyst with a synergetic confinement core-shell structure can be used to efficiently realize the combination of two and more sequential reactions with many synergistic effects.

Full text of DOI:10.1021/ja101882a

Published on Web 05/19/2010
Confinement Effect and Synergistic Function of H-ZSM-5/
Cu-ZnO-Al2O3 Capsule Catalyst for One-Step Controlled
Synthesis
Guohui Yang,† Noritatsu Tsubaki,*,†,‡ Jun Shamoto,† Yoshiharu Yoneyama,† and
Yi Zhang†
Department of Applied Chemistry, School of Engineering, UniVersity of Toyama, Gofuku 3190,
Toyama 930-8555, Japan, and JST, CREST, Sanbancho 5, Chiyoda-ku, Tokyo 102-0075, Japan
Received March 5, 2010; E-mail: tsubaki@eng.u-toyama.ac.jp
Abstract: Dimethyl ether (DME) is an industrially important intermediate, as well as a promising clean fuel, but
the effective production through traditionally consecutive steps from syngas to methanol and then to DME has
been hindered by the poorly organized structure of the conventional physical mixture catalyst. Here, a novel
zeolite capsule catalyst possessing a core-shell structure (millimeter-sized core catalyst and micrometer-sized
acidic zeolite shell) was proposed initially through a well-designed aluminum migration method using the core
catalyst as the aluminum resource and for the first time was applied to accomplish the DME direct synthesis
from syngas. The selectivity of the expected DME on this zeolite capsule catalyst strikingly exceeded that of
the hybrid catalyst prepared by the traditional mixing method, while maintaining the near-zero formation of the
unexpected alkanes byproduct. The preliminary methanol synthesis reaction on the core catalyst and the following
DME formation from methanol inside the zeolite shell cooperated concertedly and promoted mutually. This
zeolite capsule catalyst with a synergetic confinement core-shell structure can be used to efficiently realize the
combination of two and more sequential reactions with many synergistic effects.
Introduction
However, previous studies paid more attention to the simply
mechanical mixing of the two types of catalysts, improving
Alternative energy sources have moved into the spotlight in
recent years with soaring oil prices and dwindling resources.1
In addition to utilization as an important chemical intermediate,
DME is also very promising as a new energy source, being
available for various purposes such as an LPG alternative, a
fuel for power generation, synthetic diesel, and hydrogen
resource for fuel cells.2-4 Traditionally, DME can be produced
through two methods. One is methanol dehydration on a single
dehydration catalyst, where the methanol is separately made
from syngas,5 and another is a couple of consecutive reactions
for DME direct synthesis from syngas on a hybrid catalyst,
where the hybrid catalyst is a simple mixture of methanol
synthesis catalyst and methanol dehydration catalyst.6 The latter
method is more thermodynamically favorable than the former
because methanol synthesis is generally severely limited by
thermodynamics and in situ conversion of the formed methanol
to DME can enhance the maximum syngas conversion.6,7
catalyst lifetime, performance, or activity.8-10 The effective
structure combination of two active catalysts was not deeply
studied until now.
In heterogeneous catalysis, the active chemical composition
of a catalyst can work more efficiently only in combination with
a well-organized structure.11-13 A catalyst having a finely
designed structure usually exhibits better catalytic perfor-
mance.14 Recently, a material study on nano- or micro-sized
particles with a special core-shell structure has attracted
tremendous interest.15 The core-shell particles have broad
promising application in many fields, such as drug delivery,16
photonic devices,17 and catalytic and separation processes.18
(7) Mao, D.; Yang, W.; Xia, J.; Zhang, B.; Song, Q.; Chen, Q. J. Catal.
2005, 230, 140–149.
(8) Lewnard, J. J.; Hsiung, T. H.; White, J. F.; Brown, D. M. Chem. Eng.
Sci. 1990, 45, 2735–2741.
(9) Sofianos, A. C.; Scurrell, M. S. Ind. Eng. Chem. Res. 1991, 30, 2372–
2378.
† University of Toyama.
‡ JST, CREST.
(10) Xia, J. C.; Mao, D. S.; Zhang, B.; Chen, Q. L.; Tang, Y. Catal. Lett.
2004, 98, 235–240.
(1) Farrell, A. E.; Plevin, R. J.; Turner, B. T.; Jones, A. D.; O’Hare, M.;
Kammen, D. M. Science 2006, 311, 506–508.
(2) San, X.; Zhang, Y.; Shen, W.; Tsubaki, N. Energy Fuels 2009, 23,
2843–2844.
(11) Ueda, W.; Oshihara, K.; Vitry, D.; Hisano, T.; Kayashima, Y. Catal.
SurV. Jpn. 2002, 6, 33–44.
(12) Takehira, K.; Shishido, T.; Shoro, D.; Murakami, K.; Honda, M.;
Kawabata, T.; Takaki, K. Catal. Commun. 2004, 5, 209–213.
(13) Zhang, Y.; Liu, Y.; Yang, G. H.; Sun, S. L.; Tsubaki, N. Appl. Catal.,
A 2007, 321, 79–85.
(3) Cheung, P.; Bhan, A.; Sunley, G. J.; Iglesia, E. Angew. Chem., Int.
Ed. 2006, 45, 1617–1620.
(4) Faungnawakij, K.; Tanaka, Y.; Shimoda, N.; Fukunaga, T.; Ka-
washima, S.; Kikuchi, R.; Eguchi, K. Appl. Catal., A 2006, 304, 40–
48.
(14) Zhang, Y.; Yoneyama, Y.; Tsubaki, N. Chem. Commun. 2002, 1216–
1217.
(5) Xu, M. T.; Lunsford, J. H.; Goodman, D. W.; Bhattacharyya, A. Appl.
Catal., A 1997, 149, 289–301.
(15) Peng, X.; Schlamp, M. C.; Kadavanich, A. V.; Alivisatos, A. P. J. Am.
Chem. Soc. 1997, 119, 7019–7029.
(6) Fujimoto, K.; Asami, K.; Shikada, T.; Tominaga, H. Chem. Lett. 1984,
13, 2051–2054.
(16) Ferrari, M. Nat. ReV. Cancer 2005, 5, 161–171.
(17) Mitzi, D. B. Chem. Mater. 2001, 13, 3283–3298.
9
10.1021/ja101882a  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 8129–8136 8129

A R T I C L E S
Yang et al.
coprecipitation method.24 An ethanol solution of oxalic acid was
added rapidly to a mixed ethanol solution of copper, zinc, and
aluminum nitrates (Cu/Zn/Al in molar ratio 45/20/10) at room
temperature under vigorous stirring. The gel-like precipitate was
aged at room temperature for 24 h and then separated by centrifuge.
The drying process was conducted at 393 K for 6 h followed by
calcination in air at 643 K for 1 h. Before the following utilization,
the raw CZA catalyst was granulated into a new grain with the
size of 0.85-1.70 mm.
Zeolite Capsule Catalysts Preparation. In this report, two
different zeolite membrane hydrothermal synthesis methods were
adopted to directly construct two types of H-type ZSM-5 zeolite
shell onto the CZA catalysts. One was the traditional acidic
H-ZSM-5 zeolite membrane synthesis method with the aluminum
resource in the precursor solution. Another was the close-to-neutral
Silicalite-1 zeolite membrane preparation method without an
aluminum resource in the synthesis solution; here, the core CZA
catalyst acted as the support and Al resource for H-ZSM-5 zeolite
membrane growth simultaneously.
Usually, the core-shell structure is made up of two or more
varied materials, in which the different materials (organic and/
or inorganic) are spatially separated in different locations,
promoting mutually at the same time. In our previous reports,
the zeolite capsule catalysts with a supported metallic core
catalyst and a zeolite shell exhibited excellent properties for
isoparaffin direct synthesis that are significantly different from
their physical mixture analogy, which makes them better than
the conventional hybrid catalyst usually used for the consecutive
reactions.19,20 However, it is difficult to prepare the defect-free
H-type zeolite shell on the pure metallic catalyst, coprecipitated
catalyst, or alloy catalyst without sodium hydroxide in the
precursor synthesis solution, protecting the core catalyst from
damage by hydrothermal treatment at the same time. Especially,
the coprecipitated catalyst used here as the core generally has
mechanical and chemical strength much weaker than that of a
supported catalyst and can very easily be decomposed or
dissolved in the hydrothermal synthesis process of zeolite
membrane.
In preparation, the structure directed reagent TPAOH (tetrapro-
pylammonium hydroxide, 10% in water), silica resource TEOS
(teraethyl orthosilicate, 95%), and aluminum resource (aluminum
nitrate, 99.9%) were the Wako Co. products. Dehydrated ethanol
(99.5%) and nitric acid (69%) were both purchased from Kanto
Chemical Co. The first H-ZSM-5 zeolite shell was prepared on the
CZA core catalyst using a traditional acidic H-ZSM-5 zeolite
synthesis recipe of 0.48 TPAOH:2 TEOS:8 EtOH:120 H2O:
0.025Al2O3. The second H-ZSM-5 zeolite shell was prepared by
the close-to-neutral Silicalite-1 zeolite synthesis recipe of 0.48
TPAOH:2 TEOS:8 EtOH:120 H2O:0.24 HNO3. All reagents were
mixed in a Teflon container with vigorous stirring at room
temperature for 6 h. Then the CZA pellets (0.85-1.70 mm) were
added into the precursor solutions, and the Teflon container was
placed into the stainless steel autoclave for hydrothermal synthesis.
Zeolite shell synthesis was performed in a rotation oven with the
rotation rate of 2 rpm at 453 K for 72 h. Here, the rotation synthesis
effectively prevented the cementation of catalyst pellets during the
hydrothermal synthesis process, improving the integrity of the
zeolite shell simultaneously. Without the rotation, for example, using
a static autoclave instead, it was difficult to enwrap all core catalyst
surfaces completely. The final samples were separated from the
mother liquid and then were calcined at 773 K in air for 5 h,
removing the organic template stored in the zeolite pores. The
zeolite capsule catalyst obtained by this process using different
synthesis methods were named as CZA-Z and CZA-S respectively,
where the “Z” stands for the preparation using the H-ZSM-5 zeolite
synthesis recipe and “S” means the synthesis utilizing the Silicalite-1
zeolite synthesis recipe. It is noted that no aluminum resource was
adopted for CZA-S preparation. Moreover, it is very important that
commonly used reagent containing NaOH, KOH, Cl-, or Br- cannot
be used here for zeolite shell synthesis, as they will deactivate the
Cu/ZnO/Al2O3 core catalyst severely. For the same reason, Na-
type or NH3-type zeolite membrane is not permitted for our zeolite
capsule catalyst as ion-exchange to H-type zeolite will deactivate
core catalyst. We must synthesize H-type capsule catalyst in one
step.
Herein, we report a millimeter-sized zeolite capsule catalyst
possessing a special core-shell structure (a tricomponent core
catalyst enwrapped by one layer of H-ZSM-5 zeolite shell) with
a coprecipitated catalyst as core. Different from the traditional
zeolite membrane preparation methods, such as in situ or seeding
method using the aluminum-containing precursor solution for
zeolite layer growth on various supports,21-23 here an acidic
H-ZSM-5 zeolite shell was directly prepared successfully
through an unreported way of aluminum migration from the
core catalyst body. The aluminum-containing core catalyst acted
as the substrate for zeolite membrane growth and at the same
time as the sole aluminum resource to construct the zeolite
framework in the total synthesis process, resulting in the defect-
free covering and tightly enwrapping of the zeolite shell on its
surface. For the first time, this zeolite capsule catalyst was
applied to accomplish a successive reaction, DME direct
synthesis from syngas. The selectivity of the expected DME
on this zeolite capsule catalyst strikingly exceeded that of the
traditional hybrid catalyst (the simple blending of core catalyst
and zeolite powder), while maintaining zero formation of the
unexpected alkane byproducts. The encouraging results proved
that this novel method for the preparation of H-type zeolite shell
on aluminum-containing metallic catalyst was feasible and very
reliable, effectively protecting the vulnerable metallic core
catalyst from damage by the hydrothermal synthesis process
and realizing the controlled synthesis of the target product on
a single catalyst. The zeolite capsule catalyst preparation method
reported here together with its application in DME controlled
synthesis from syngas have the potential to inspire not only the
zeolite membrane synthesis but also heterogeneous catalysis.
Experimental Section
Core Catalyst Preparation. The core catalyst, tricomponent Cu/
ZnO/Al2O3 (CZA), was prepared by the conventional oxalate
Hybrid Catalyst Preparation. Self-made H-ZSM-5 (Si/Al )
163, atomic ratio) zeolite powder was used for the preparation of
the physically mixed catalyst. This zeolite powder was physically
well mixed with the CZA catalyst with a weight ratio of 1:10 and
then granulated to 0.85-1.70 mm at 60 MPa pressure. This catalyst
was named as CZA-M, where the “Z” means H-ZSM-5 zeolite and
“M” means the mechanical mixture of H-ZSM-5 with CZA.
Catalysts Characterization. X-ray diffraction (XRD) patterns
were collected using a Rigaku RINT 2400 X-ray powder diffrac-
tometer equipped with a Cu KR radiation source at 40 KV and 40
(18) Ren, N.; Yang, Y. H.; Shen, J.; Zhang, Y. H.; Xu, H. L.; Gao, Z.;
Tang, Y. J. Catal. 2007, 251, 182–188.
(19) Yang, G. H.; He, J. J.; Yoneyama, Y.; Tan, Y. S.; Han, Y. Z.; Tsubaki,
N. Appl. Catal., A 2007, 329, 99–105.
(20) Bao, J.; He, J.; Zhang, Y.; Yoneyama, Y.; Tsubaki, N. Angew. Chem.,
Int. Ed. 2008, 47, 353–356.
(21) Lai, R.; Yan, Y.; Gavalas, G. R. Microporous Mesoporous Mater.
2000, 37, 9–19.
(22) Yan, Y.; Davis, M. E.; Gavalas, G. R. Ind. Eng. Chem. Res. 1995,
34, 1652–1661.
(23) Yang, G.; Zhang, X.; Liu, S.; Yeung, K. L.; Wang, J. J. Phys. Chem.
Solids 2007, 68, 26–31.
(24) Sun, Q.; Zhang, Y.-L.; Chen, H.-Y.; Deng, J.-F.; Wu, D.; Chen, S.-
Y. J. Catal. 1997, 167, 92–105.
9
8130 J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010

H-ZSM-5/Cu-ZnO-Al2O3 Capsule Catalyst
A R T I C L E S
was adopted to directly coat a H-ZSM-5, not Na-ZSM-5, shell
onto the CZA core catalyst surface. The obtained catalyst,
denoted as CZA-Z, was separated from the mother liquid,
washed, and calcined, removing the template settled in the
zeolite pores. The XRD pattern of this CZA-Z zeolite capsule
catalyst in Figure 1 identified the classic X-ray diffraction peaks
of H-ZSM-5 zeolite in the ranges of 2θ ) 5-10 and 21-25,
indicating that H-ZSM-5 zeolite had crystallized on the CZA
core catalyst successfully.25
The surface SEM image and EDS analysis of CZA-Z in
Figure 3a and b gives the surface morphology and elemental
composition of zeolite shell. After this zeolite shell preparation,
the surface of CZA core catalyst had been homogeneously
covered by one layer of randomly oriented H-ZSM-5 zeolite
crystals. The average size of zeolite crystals of the zeolite shell
was about 1 µm. The surface EDS analysis result in Figure 3b
presented that the Si/Al ratio of zeolite shell was about 32. The
signals of Cu and Zn on the zeolite shell surface were zero,
which proved that this H-ZSM-5 zeolite shell was compact and
integrated, indicating that this zeolite shell preparation method
was successful. Figure 3c and d shows the cross-section SEM
image and EDS line analysis results of CZA-Z zeolite capsule
catalyst, respectively. A well-intergrown zeolite shell had
enwrapped the CZA core catalyst perfectly. The related EDS
line analysis along the white line in the SEM image exhibited
the change of element signals from the zeolite shell to the
catalyst core part. The Si KR signal increased in the zeolite
shell and then deceased sharply at the interface between zeolite
shell and core catalyst, indicating the change from zeolite shell
to core catalyst, as well as identifying that the thickness of
zeolite shell was about 3.9 µm. The Al KR signal kept a stable
level in both the zeolite shell and the core catalyst. For Cu KR
and Zn KR, they were close to zero in the zeolite shell, which
suggested that there was no obvious weight loss for the active
species of core catalyst during this hydrothermal synthesis
process.
Figure 1. XRD patterns of CZA, CZA-Z, CZA-S, CZA-M, and pure
H-ZSM-5 zeolite.
mA. A JEOL JSM-6360LV scanning electron microscope (SEM)
and JED-2300 energy dispersive spectroscopy (EDS) were used
for SEM imaging and EDS analyzing on the catalyst surface and
cross-section.
Reaction and Analysis of Products. Before the reaction, the
catalyst was first reduced in situ at 493 K by a 5% H2 in nitrogen
flow (60 mL/min) for 10 h. Reactions were carried out in a down
flow fixed-bed stainless steel reactor. The reaction conditions were
523 K, 5.0 MPa, WCZA catalyst/Fsyngas ) 10 g ·h/mol, and syngas (H2
59.22%, CO 32.60%, CO2 5.16%, Ar 3.02%). The effluent products
from the heated line were first analyzed by an online gas
chromatograph (Shimadzu, TCD, GC-8A) for CO and CO2, and
then the tail gases were analyzed online using another gas
chromatograph (Shimadzu, FID, GC-8A) for other products. All
analysis lines and valves were heated to prevent possible condensa-
tion of the products before entering the gas chromatograph, ensuring
reliable materials balance.
For the second type of zeolite shell preparation, the traditional
Silicalite-1 zeolite synthesis method was utilized to construct
zeolite shell on CZA core catalyst. No aluminum in the precursor
synthesis solution was the specific feature of this preparation
process. The Al used to construct the framework of H-ZSM-5
zeolite shell here would be extracted from the CZA core catalyst
during the hydrothermal synthesis process, as designed. The
obtained zeolite capsule catalyst by this preparation method was
named as CZA-S, where the “S” means this Silicalite-1 zeolite
synthesis method. The XRD pattern of CZA-S zeolite capsule
catalyst is presented in Figure 1. This sample exhibited the MFI
structure with major peaks located at 2θ ) 7.9, 8.9 and the
characteristic triplet at 2θ ) 23.5, confirming the formation of
zeolite with MFI structure on the CZA core catalyst. In addition,
some comparable peaks of CZA-S catalyst to that of the bare
CZA catalyst were observed in the range of 2θ ) 30-40,
indicating that the CZA core catalyst was still stable after this
hydrothermal synthesis treatment.
Results and Discussion
Capsule Catalysts Preparation and Characterization. The
tricomponent catalyst Cu/Zn/Al2O3 (CZA) made by the copre-
cipitation method is a conventional catalyst usually used for
methanol synthesis. In this report, it was selected as the core
catalyst for preparation of the zeolite capsule catalysts. The bare
CZA catalyst without hydrothermal treatment had a pellet size
of 0.85-1.70 mm. For its X-ray diffraction (XRD) analysis,
no crystalline peaks of H-ZSM-5 zeolite were observed, as
shown in Figure 1. The surface scanning electron microscope
(SEM) and energy dispersive spectroscopy (EDS) analysis
results in Figure 2 indicated that this CZA catalyst had the molar
composition of Cu/Zn/Al ) 61.90:24.50:13.60. It was very
similar to its original bulk ratio in this catalyst preparation recipe.
The hybrid catalyst named as CZA-M was prepared by mixing
CZA catalyst with pure H-ZSM-5 (Si/Al ) 163) zeolite powder.
The XRD peaks assigned to CZA or H-ZSM-5 zeolite can be
distinguished by comparing the XRD patterns of CZA-M with
pure H-ZSM-5 zeolite or bare CZA core catalyst independently
in Figure 1.
Zeolite shell surface morphology and elemental composition
of CZA-S zeolite capsule catalyst are shown in Figure 4a and
4b, respectively. The average size of zeolite crystal constructed
this zeolite shell was much larger than that of CZA-Z zeolite
capsule catalyst as above, probably because of the special zeolite
synthesis solution for CZA-S preparation without aluminum
In this report, two different methods for zeolite membrane
preparation were adopted to construct two types of H-ZSM-5
zeolite shell directly on the millimeter-sized CZA core catalyst.
For the first type of zeolite shell preparation, the traditional
acidic H-ZSM-5 zeolite synthesis method using the aluminum-
containing precursor solution, but without sodium hydroxide,
(25) Olson, D. H.; Kokotailo, G. T.; Lawton, S. L.; Meier, W. M. J. Phys.
Chem. 1981, 85, 2238–2243.
9
J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010 8131

A R T I C L E S
Yang et al.
Figure 2. (a) Surface SEM image and (b) EDS analysis of the bare CZA core catalyst.
Figure 3. (a) Surface SEM image and (b) surface EDS analysis; (c) cross-section SEM image and (d) EDS line analysis of the zeolite capsule catalyst
CZA-Z.
resource, in conjunction with the lower pH value.26,27 In
addition, there were no pinholes and cracks on the zeolite shell,
which indicated that this hydrothermal synthesis process was
successful. Surface EDS analysis on the zeolite shell gave zero
signals of Zn and Cu, suggesting that this zeolite shell was
compact and well integrated as well. The most interesting finding
was the Al content in the zeolite shell. Actually, it was not zero
despite the close-to-neutral Silicate-1 zeolite synthesis method
for zeolite shell preparation. Nonzero Al signal, together with
the XRD pattern of CZA-S in Figure 1, confirmed that the
zeolite shell of the CZA-S catalyst had the MFI zeolite structure
and belonged to a H-ZSM-5 zeolite membrane. The surface Si/
Al ratio of this zeolite shell was 221, as listed in Table 1, much
higher than that of the conventional H-ZSM-5 zeolite powder
usually used for DME direct synthesis from syngas. However,
the total acid amount of this zeolite capsule catalyst was similar
to that of hybrid catalyst CZA-M (see Figure S1, Table S1 in
Supporting Information). These should be attributed to the
aluminum-containing CZA core catalyst used here. It acted
as the substrate and aluminum resource simultaneously for
zeolite shell growth, which could lead to a concentration
gradient of Al in the zeolite shell.22 The cross-section SEM
(26) Persson, A. E.; Schoeman, B. J.; Sterte, J.; Otterstedt, J. E. Zeolites
1994, 14, 557–567.
(27) Persson, A. E.; Schoeman, B. J.; Sterte, J.; Otterstedt, J. E. Zeolites
1995, 15, 611–619.
9
8132 J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010

H-ZSM-5/Cu-ZnO-Al2O3 Capsule Catalyst
A R T I C L E S
Figure 4. (a) Surface SEM image and (b) surface EDS analysis; (c) cross-section SEM image and (d) EDS line analysis of the zeolite capsule catalyst
CZA-S.
Table 1. Catalyst Properties and Catalytic Performance in STD
also form H-ZSM-5 zeolite shell with lower aluminum
content on aluminum-containing supports.
Reactiona
selectivity (%)
Both of the zeolite shell preparation methods, acidic zeolite
zeolite
(%)
CO conversion
(%)
synthesis and close-to-neutral zeolite synthesis, can realize the
formation of a complete, well-intergrown acidic zeolite shell
on the CZA core catalyst. Here, the realization of core catalyst
enwrapped tightly by one layer of zeolite shell should be
attributed to the aluminum-containing CZA core catalyst. The
aluminum migrated from the aluminum supports and was
incorporated into the MFI zeolite framework.22,28 At the same
time, a transition layer of zeolite crystals with aluminum-
containing supports could be formed at the interface area
between the zeolite shell and CZA core catalyst. These factors
ensured the integrity of zeolite capsule catalyst even if undergo-
ing calcination and the following catalysis reaction. In addition,
the zeolite shells prepared by two different methods had different
structural properties, not only in regard to the element composi-
tion but also the shell thickness, as well as zeolite loading
amount and acidic sites distribution shown in NH3-TPD curves
(Supporting Information). Therefore, only by an ordinary
method, with/without the initial aluminum resource in precursor
solution, the physical and chemical properties of zeolite capsule
catalyst could be dominated well.
catalyst
Si/Alb
MeOH
DME
others
CZA
44.02
5.59
30.40
58.07
98.38
3.41
21.43
57.29
0.50
96.59
78.57
40.51
1.12
0.00
0.00
2.20
CZA-Zc
CZA-Sd
CZA-Me
32
221
163
7.60
9.63
10.0
a Reaction conditions: 523 K, 5.0 MPa, WCZA/Fsyngas ) 10 g·h·mol-1
.
Syngas: H2/CO/CO2/Ar ) 59.22/32.60/5.16/3.02. b Si/Al ratio of zeolite
shell or zeolite powder. c “Z” means the H-ZSM-5 synthesis method.
d “S” stands for the Silicalite-1 synthesis method. e “M” means the
physical mixing of CZA with H-ZSM-5 zeolite powder.
image and EDS line analysis of CZA-S zeolite capsule
catalyst are shown in Figure 4c and 4d, respectively. In the
SEM image, the zeolite shell could be distinguished easily
and was also in a good state, enwrapping the CZA core
catalyst. The signal intensity of Si KR, Cu KR, and Zn KR
in the EDS line analysis result gave a change similar to that
of CZA-Z zeolite capsule catalyst along the direction from
the zeolite shell to the core catalyst, but the thickness of the
zeolite shell was 5 µm here. It is noteworthy that the Al KR
signal in the zeolite shell also was not zero, even if the
synthesis solution for the CZA-S preparation did not contain
any aluminum resource. For the aluminum constructing the
final H-ZSM-5 zeolite shell, it must come from the core
catalyst, the sole aluminum resource in this total hydrothermal
synthesis process. Accordingly, it can be concluded that the
hydrothermal synthesis solution without Al resources could
Capsule Catalysts Activity. The direct conversion from syngas
to dimethyl ether (STD reaction) was the application of zeolite
capsule catalysts CZA-Z and CZA-S. The bare CZA catalyst
and hybrid catalyst CZA-M were also performed under the same
(28) Beving, D. E.; McDonnell, A. M. P.; Yang, W.; Yan, Y. J.
Electrochem. Soc. 2006, 153, B325–B329.
9
J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010 8133

A R T I C L E S
Yang et al.
Figure 5. Products distribution of the CZA catalyst, zeolite capsule catalyst CZA-Z and CZA-S, and the hybrid catalyst CZA-M.
reaction conditions as references. The catalyst activity is
summarized in Table 1. The hybrid catalyst CZA-M consisted
of two types of catalysts, CZA and zeolite catalyst, for different
reactions, respectively. In the STD reaction on this hybrid
catalyst, the formed methanol on the CZA catalyst was randomly
converted to DME on the zeolite catalyst via its dehydration
reaction, which broke the thermodynamic equilibrium of the
reaction from syngas to methanol on CZA catalyst, enhancing
the CO conversion from 44.02% (on single CZA catalyst) to
58.07%. It is well-known that the reaction combination of
methanol synthesis with its in situ dehydration to DME is
thermodynamically more favorable than either occurring solely.6
The results obtained on this hybrid catalyst CZA-M also proved
it.
considered. In the hydrothermal process, part of the synthesis
solution entered the core catalyst pores and then crystallized
into zeolite over some active sites of core catalyst, and the ease
of Cu species reduction can essentially control the catalyst
activity.29 However, increasing metal content of the core catalyst
or extending contact time of the reactant with zeolite capsule
catalyst can enhance CO conversion very easily.
Reaction Products Distribution. Table 1 also gives the product
selectivity on the bare CZA catalyst, hybrid catalyst CZA-M,
and zeolite capsule catalysts CZA-Z and CZA-S. The corre-
sponding products distribution patterns are compared in Figure
5. For CZA catalyst, the highest methanol selectivity of 98.38%
was obtained, and DME was only a tiny product with a
selectivity of 0.50%. Light hydrocarbons, such as methane and
C
2+, could be found in the final products. In the case of CZA-
For the CZA-Z zeolite capsule catalyst, its CO conversion
was 5.59%, much lower than that of the bare CZA catalyst.
The CZA-Z zeolite capsule catalyst was prepared by the acidic
zeolite synthesis solution with aluminum resource. Possibly, the
lower catalytic activity should be attributed to some damage
on the core CZA catalyst during the hydrothermal synthesis
process, such as dissolution of Cu or Zn, by the formation of a
coordinated compound between TPAOH and Cu or Zn, or
weakening interaction between Cu and Zn by TPAOH when
the aluminum resource existed in the precursor solution, which
led to the change of composition and structure of the core CZA
catalyst. However, as a new method to overcome the problem
of CZA-Z synthesis, the CZA-S zeolite capsule catalyst prepared
by a close-to-neutral Silicalite-1 zeolite synthesis method, where
HNO3 was added to lower the pH value of TPAOH-containing
synthesis solution, without aluminum resource, stopping forma-
tion of coordinated compounds, exhibited remarkably enhanced
catalytic activity compared with the CZA-Z zeolite capsule
catalyst, as shown in Table 1. The CO conversion of CZA-S
reached 30.40%, which indicated that the close-to-neutral
Silicalite-1 zeolite synthesis method could effectively reduce
the damage from synthesis solution to core catalyst.
M, the selectivity of DME reached 40.51% along with a sharp
decrease of methanol selectivity compared with the single CZA
catalyst. However, there was only random contact between
methanol and the active sites of zeolite catalyst in the hybrid
catalyst CZA-M during STD reaction, which was decided by
hybrid catalyst structure (simply mechanical blend of CZA with
zeolite catalyst). As a result, only a fairly ordinary selectivity
for DME was obtained, accompanyied by the existence of
unexpected alkanes in the final products.
Different from the physically mixed catalyst, zeolite capsule
catalysts had a special core-shell structure in which the zeolite
shell enwrapped the core catalyst perfectly, which offered an
inevitable way, passing through the zeolite shell, for methanol
to escape from the zeolite capsule catalyst. Here, all of the
methanol must contact the H-ZSM-5 zeolite shell. Capsule
catalyst CZA-Z was prepared by the acidic H-ZSM-5 zeolite
synthesis method. Its zeolite shell had higher acidic intensity
but slightly lower zeolite loading amount compared with another
zeolite capsule catalyst CZA-S. In the STD reaction on CZA-Z
catalyst, the DME selectivity reached 96.59% together with zero
alkanes byproduct, probably benefiting from the high acidic
intensity and thinner zeolite shell thickness of capsule catalyst.
In addition, for the lower catalytic activity of zeolite capsule
catalysts compared with bare CZA catalyst and hybrid catalyst
CZA-M, another factor that some active sites of the core catalyst
were partially covered by zeolite crystals should also be
(29) Prasad, P. S. S.; Bae, J. W.; Kang, S. H.; Lee, Y. J.; Jun, K. W. Fuel
Process. Technol. 2008, 89, 1281–1286.
9
8134 J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010

H-ZSM-5/Cu-ZnO-Al2O3 Capsule Catalyst
A R T I C L E S
Scheme 1. (a) Schematic Diagram Showing DME Direct Synthesis
from Syngas on a Single Zeolite Capsule Catalyst, (b) Schematic
Diagram of the Bare Catalyst with Zeolite Layer Loading in
Reactor in the Form of a Dual Layer, and (c) Zeolite Capsule
Catalysts in Reactor
The strongly acidic sites ensured that almost all methanol formed
on the core catalyst could be converted into DME, while the
thinner zeolite shell thickness offered a short residence time
for DME within the zeolite shell, inhibiting the formation of
alkanes by DME further dehydration. Another zeolite capsule
catalyst CZA-S was prepared using a close-to-neutral zeolite
synthesis method that exerted less damage to core catalyst, and
the CZA core catalyst acted in situ as the aluminum resource
during hydrothermal synthesis process. The zeolite shell of
CZA-S had a lower zeolite loading amount and similar acid
amount compared with the physically mixed catalyst CZA-M
(Supporting Information) but higher weight increment and
greater thickness (5 µm) than zeolite capsule catalyst CZA-Z.
In the STD reaction on the CZA-S zeolite capsule catalyst, there
was still zero selectivity for the alkane byproducts as shown in
Table 1, but the DME selectivity reached 78.57%, nearly twice
that of the hybrid catalyst CZA-M even if the latter was prepared
using the H-ZSM-5 zeolite with similar acid amount and larger
weight proportion. Also, it was only slightly lower than that of
the CZA-Z zeolite capsule catalyst, possibly due to the low
acidic intensity and smaller acid amount together with the
insufficient distance offered by the zeolite shell for methanol
dehydration, leading to the partial conversion of the methanol
to DME. It is referred that increasing the acidic sites intensity,
density, and thickness of the zeolite shell of CZA-S, which will
be easily realized only by controlling the hydrothermal synthesis
conditions or recipe, can further reduce the selectivity of
methanol, thereby enhancing the selectivity of DME.
Discussion
The STD reactions on the zeolite capsule catalyst CZA-Z
and CZA-S exhibited a completely different product distribution
compared with that on the bare CZA and hybrid CZA-M
catalyst, realizing an exclusive DME selectivity, very low
methanol selectivity, and zero byproduct formation simulta-
neously, as shown in Figure 5. The outstanding capability of
the capsule catalyst should be attributed to its specific core-shell
structure. For hybrid catalyst CZA-M, there was no spatially
confined effect between the methanol synthesis reaction and
methanol dehydration reaction, demonstrating that these two
consecutive reactions occurred independently and randomly. As
a result, a part of the methanol formed on the methanol synthesis
catalyst could escape from the catalyst bed directly without
undergoing dehydration on zeolite catalyst. In addition, the
random nature of the process also demonstrated that some DME
may get more opportunities to be further dehydrated on the
zeolite catalyst or transformed to alkanes or alkenes. However,
the zeolite capsule catalysts with the core-shell structure might
control the occurrence of the two reactions in a consecutive
order. As illustrated in Scheme 1a, the feed gas first passed
through the zeolite shell to enter the core catalyst, where the
methanol was produced on the core catalyst. Then the methanol
formed in the core catalyst must go through the zeolite shell to
escape, which ensured that all of the methanol had enough
opportunities to contact the active sites of the H-type zeolite
shell, to be converted into the expected products. As a result,
the selectivity for DME was increased sharply. Moreover, the
zeolite shell had a fixed position, uniform thickness, and
homogeneous property, which made it an equal opportunity
provider to each methanol molecular that diffused from the core
catalyst, improving the conversion from methanol to DME
concertedly, while suppressing its further dehydration to generate
the alkane or alkene byproducts simultaneously. From the
viewpoint of energy efficiency, it should be mentioned that the
reaction heat from the exothermic methanol synthesis on the
core catalyst could be in situ recycled inside the zeolite capsule
by the DME synthesis from methanol on the zeolite shell to
protect the catalyst lifetime from overheating.
It should also be noted that this capsule method, with a zeolite
shell enwrapping the millimeter-sized core catalyst, provided a
new method to avoid the difficulty of large-area membrane
preparation due to the easy formation of pinholes or cracks.
Especially, the ratio of membrane surface area to catalyst bed
volume of the zeolite capsule catalyst was extremely high
compared to that of a dual-layer method or other membrane
reactor configurations, providing excellent catalysis and separa-
tion since each core catalyst is enwrapped by the zeolite shell,
as illustrated in Scheme 1b and c.
Conclusions
Two direct preparation methods were demonstrated to
fabricate two types of H-ZSM-5 zeolite capsule catalyst with a
tailor-made millimeter-sized catalyst core and microsized zeolite
shell. One method was the traditional acidic H-ZSM-5 zeolite
synthesis, and another was a close-to-neutral Silicate-1 zeolite
synthesis using the core catalyst as the aluminum provider. The
latter method explored a facile, reliable synthetic method for
acidic zeolite membrane growth on an aluminum-containing
support and effectively reduced the damage from zeolite
precursor solution to substrates during the hydrothermal syn-
thesis process. In the consecutive reactions of DME direct
synthesis from syngas, zeolite capsule catalysts exhibited an
extreme selectivity for DME compared with the conventional
hybrid catalyst, suppressing the further dehydration of DME to
form alkane/alkene byproducts. This concept of a zeolite capsule
9
J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010 8135

A R T I C L E S
Yang et al.
on Priority Areas “Chemistry of Concerto Catalysis” from Ministry
of Education, Culture, Sports, Science and Technology, Japan.
catalyst with the effect of an equilibrium shift, shape selectivity,
and synergetic confinement is therefore promising to be widely
used for many consecutive reactions such as multiple-step
organic synthesis. In addition, it is clear that the original
preparation method of the acidic H-ZSM-5 zeolite shell using
a close-to-neutral Silicalite-1 zeolite synthesis method on an
aluminum-containing substrate can give a novel inspiration to
the science of zeolite membranes.
Supporting Information Available: Further characterization
and detailed discussion on the concentration and strength of
acid sites of bare CZA catalysts, zeolite capsule catalyst CZA-
S, CZA-Z, and hybrid catalyst CZA-M by the temperature
programmed desorption of ammonia (NH3-TPD). This material
is available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. This work was supported by Nippon Oil
Corporation and in part by a Grant-in-Aid for Scientific Research
JA101882A
9
8136 J. AM. CHEM. SOC. VOL. 132, NO. 23, 2010