Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the methanol-to-olefins reaction

Article abstract of DOI:10.1016/j.cattod.2014.02.057

Several approaches to improve the catalytic performance of SSZ-13 and SAPO-34 for application as acid catalysts in the methanol-to-olefins (MTO) reaction were explored. Silylation of mesoporous SSZ-13 with a Si/Al ratio of 20 zeolite resulted in increased lifetime in the MTO reaction. Lowering the acidity of SSZ-13 by increasing the Si/Al ratio to 50 also increased the lifetime. The generation of additional mesoporosity in SSZ-13 with a Si/Al ratio of 50 by use of the organosilane octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride (TPOAC) only resulted in a minor improvement of the lifetime. Attempts to synthesize mesoporous SSZ-13 at high Si/Al ratios by use of (C ₂₂H₄₅N⁺(CH₃)₂C ₄H₈N⁺(CH₃)₂C ₄H₉)Br₂ (C_22-4-4Br₂) were unsuccessful, and instead ZSM-5 zeolite was obtained. Similarly, SAPO-34 could not be made hierarchical by using C_22-4-4Br₂ as a mesoporogen. In this case, other AlPO-phases were obtained. Mesoporous SAPO-34 was synthesized by using TPOAC in the synthesis gel. The additional intracrystalline mesoporosity did not lower the deactivation rate of SAPO-34 as was earlier observed for SSZ-13. The total methanol conversion capacity per acid site for microporous and mesoporous SAPO-34 were however comparable. The lower acidity of the acid sites in SAPO-34 led to the complete utilization of the micropore space. This is to be contrasted to SSZ-13 zeolite, for which the increased rate of coke formation results in more extensive coking deactivation and underutilization of the micropore space.

Full text of DOI:10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for
the methanol-to-olefins reaction
Leilei Wu, Emiel J.M. Hensen∗
Laboratory of Inorganic Materials Chemistry, Schuit Institute of Catalysis, Eindhoven University of Technology, PO Box 513, Eindhoven, 5600 MB The
Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Several approaches to improve the catalytic performance of SSZ-13 and SAPO-34 for application as acid
catalysts in the methanol-to-olefins (MTO) reaction were explored. Silylation of mesoporous SSZ-13 with
a Si/Al ratio of 20 zeolite resulted in increased lifetime in the MTO reaction. Lowering the acidity of SSZ-13
by increasing the Si/Al ratio to 50 also increased the lifetime. The generation of additional mesoporosity in
SSZ-13 with a Si/Al ratio of 50 by use of the organosilane octadecyl-(3-trimethoxysilylpropyl)-ammonium
chloride (TPOAC) only resulted in a minor improvement of the lifetime. Attempts to synthesize meso-
porous SSZ-13 at high Si/Al ratios by use of (C22H45 N+(CH3)2 C4H8 N+(CH3)2 C4H9)Br2 (C22-4-4Br2)
were unsuccessful, and instead ZSM-5 zeolite was obtained. Similarly, SAPO-34 could not be made hier-
archical by using C22-4-4Br2 as a mesoporogen. In this case, other AlPO-phases were obtained. Mesoporous
SAPO-34 was synthesized by using TPOAC in the synthesis gel. The additional intracrystalline mesoporo-
sity did not lower the deactivation rate of SAPO-34 as was earlier observed for SSZ-13. The total methanol
conversion capacity per acid site for microporous and mesoporous SAPO-34 were however comparable.
The lower acidity of the acid sites in SAPO-34 led to the complete utilization of the micropore space.
This is to be contrasted to SSZ-13 zeolite, for which the increased rate of coke formation results in more
extensive coking deactivation and underutilization of the micropore space.
Received 14 November 2013
Received in revised form 19 January 2014
Accepted 21 February 2014
Available online xxx
Keywords:
SAPO-34
SSZ-13
Methanol-to-olefins
Stability
Hierarchical zeolites
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
of carbonaceous deposits. This requires intermittent regeneration
of the catalyst in commercial operation. In China, the MTO pro-
Light olefins are key chemical intermediates in the petrochem-
ical industry, which are usually produced through non-catalytic
cracking of petroleum feedstock. With the growing importance of
alternative feedstocks, most notably cheap natural gas and coal,
there is now growing interest in methanol as a chemical interme-
diate of the synthesis of olefins. The methanol-to-olefins (MTO)
reaction has been intensively studied in the last decades [1–5].
HSAPO-34, a silicoaluminophosphate having the chabazite (CHA)
pore topology is the commercial catalyst for the MTO reaction. The
relatively low acidity of silicoaluminophosphates and the combi-
cess is being commercialized on a large scale as part of chemical
plants that convert cheap coal via gasification to syngas followed
by methanol manufacture, light olefin and polymer production [6].
Catalyst deactivation is delt with by carrying out the MTO reaction
in a fluidized bed with a regenerator in which part of the coke is
burnt. This capital-intensive process would benefit from catalysts
with a longer lifetime, because the size of the regeneration section
could be reduced [7].
Recently, we have been shown that the introduction of meso-
pores in HSSZ-13 crystals results in a substantial decrease of the
negative effect of coke formation [8,9]. Further characterization
of mesoporous and conventional HSSZ-13 showed that the better
accessibility of the micropores results in more complete utilization
of the micropore space for the MTO reaction. For the synthesis of
mesoporous HSSZ-13, we were inspired by the approaches devel-
oped by the Ryoo group [10–12]. We employed a combination of
a structure-directing agent (SDA) for CHA formation (trimethyl-
adamantanammonium hydroxide) with a mesopore-generating
template C22H45 N+(CH3)2 C4H8 N+(CH3)2 C4H9)Br2, identified
by predictive computational modeling. Our approach was recently
˚
nation of large cavities with a diameter of 9.4 A, which stabilize
the methylated benzene catalytic intermediates, with small 8-ring
˚
pore openings (3.8 A) result in high selectivity to light olefins,
typically above 80%. An inherent drawback of HSAPO-34 as MTO
catalyst, which is typical for methanol-to-hydrocarbons (MTH)
conversion reactions in zeolites, is deactivation due to formation
∗
Corresponding author. Tel.: +31 40 2475178; fax: +31 40 2455054.
E-mail address: e.j.m.hensen@tue.nl (E.J.M. Hensen).
http://dx.doi.org/10.1016/j.cattod.2014.02.057
0920-5861/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
2
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
shown to be more generally applicable for a wide range of zeo-
lites [13]. When applied in the MTO reaction the presence of
interconnected micro- and mesoporosity in HSSZ-13 substantially
increases the total methanol conversion capacity [8,9]. The meso-
porous HSSZ-13 zeolites exhibited a much greater lifetime than
conventional HSSZ-13 in the MTO reaction at nearly similar light
olefins yield. The increased lifetime is explained by better utiliza-
tion of the micropore space. There are two competing effects of
the introduction of mesoporosity: (i) increased accessibility of the
micropore space and, accordingly, better utilization of the micro-
pore space and (ii) increased rate of coke formation with increasing
external surface (increasing mesopore volume). Due to the fast coke
built-up in HSSZ-13, most of the coke forms in the external region of
the zeolite. Maximum catalyst lifetime was obtained at a relatively
low mesoporogen/SDA ratio.
In this work, we investigate further issues related to mesoporous
CHA zeolites for the MTO reaction. First, we explore the possibility
to further improve the catalytic performance of HSSZ-13 zeolite for
the MTO reaction. Two approaches were followed, i.e. decreasing
the Al content of the zeolite framework so as to lower the density of
Brønsted acid sites and the rate of coking [14–16] and by silylating
the acid groups at or close to the external zeolite surface. Sec-
ond, we attempted to synthesize mesoporous HSAPO-34 using the
mesoporogen successfully employed for synthesis of mesoporous
HSSZ-13, C22H45 N+(CH3)2 C4H8 N+(CH3)2 C4H9)Br2. For com-
parison, we also used an amphiphlic organosilane, which has been
shown to induce mesopores for AlPO-5 and AlPO-11 molecular
sieves [12].
for 10 h in static air. The proton form of the zeolite was obtained
by triple ion exchange of the calcined zeolite with 1 M NH4NO3 at
70 ◦C for 2 h followed by calcination in static air at 550 ◦C for 4 h.
Meso-Zeo(50, C22-4-4): A zeolite with Si/Al = 50 was synthe-
sized in the same manner as SSZ-13 by combining C22-4-4Br2 and
TMAdOH as templates. The starting molar gel composition was
10 TMAdOH: 5C22-4-4Br2: 10 Na2O: Al2O3: 100 SiO2: 5000 H2O,
which was subjected to crystallization in a Teflon-lined stainless-
steel autoclave at 160 ◦C for 6 days. After crystallization, the solid
product was collected by filtration, washed with deionized water,
and dried at 110 ◦C. The zeolites were finally calcined at 550 ◦C for
10 h in static air. The proton forms of the zeolites were obtained by
triple ion exchange of the calcined form with 1 M NH4NO3 at 70 ◦C
and calcination at 550 ◦C for 4 h in static air.
Meso-Zeo(50, TPOAC): A mesoporous zeolite was synthesized
by using octadecyl-(3-trimethoxysilylpropyl)-ammonium chloride
(TPOAC, ABCR) as a mesopore-directing organosilane surfactant.
The gel molar composition was 4 TPOAC: 20 TMdAOH: 10 Na2O:
Al2O3: 100 SiO2: 4400 H2O. The mixture was stirred further at room
temperature until a homogeneous gel was obtained. The resulting
gel was transferred into a Teflon-lined stainless steel autoclave and
kept at 160 ◦C for 6 days. Thereafter, the solid material was recov-
ered by filtration. The catalysts were finally calcined at 550 ◦C for
10 h in static air. The proton form of the zeolite was obtained by
triple ion exchange of the calcined form with 1 M NH4NO3 at 70 ◦C
and calcination at 550 ◦C for 4 h in static air.
Meso-SSZ-13(C22-4-4
, 0.17)-sil: A mesoporous SSZ-13 with
Si/Al = 20 synthesized by C22-4-4Br2 and TMAdOH as templates was
treated with TEOS to deactivate the external surface [18]. To this
purpose, 1 g dehydrated SSZ-13(C22-4-4, 0.17) [8] in its proton form
was suspended in a mixture of 50 ml n-hexane and 0.15 ml TEOS at
50 ◦C for 1 h. The silylated product was centrifuged, dried at 110 ◦C,
and calcined in air in two steps: the temperature was increased
to 120 ◦C at the rate of 5◦ min−1 for 2 h, and went to 550 ◦C at
0.2 ◦C min−1 for 4 h.
2. Experimental
2.1. Synthesis of materials
2.1.1. Template synthesis
C
22-4-4Br2:Thesurfactant(C22H45 N+(CH3)2 C4H8 N+(CH3)2
C4H9)Br2 was synthesized following a published procedure [9].
First, 4.1 g (0.01 mol) 1-bromodocosane (Aldrich, 96%) was
dissolved in 20 ml toluene and added dropwise into the 20 ml solu-
tion of 10.3 g (0.07 mol) N,N,Nꢀ,Nꢀ-tetramethyl-1,4-butanediamine
(Aldrich, 98%) in acetonitrile. The resulting solution was stirred
for 3 h at room temperature and then mixed at 70 ◦C under reflux
overnight. After cooling to room temperature, the solution was
kept in a refrigerator at 4 ◦C for 1 h, filtered and washed with diethyl
ether. The resulting solid was dried in a vacuum oven at room tem-
perature. The product was (C22H45 N+(CH3)2 C4H8 N(CH3)2)Br.
Second, 3.7 g (0.007 mol) (C22H45 N+(CH3)2 C4H8 N(CH3)2)Br
and 1.96 g (0.014 mol) 1-bromobutane (Aldrich, 98%) were
dissolved in 110 ml of acetonitrile and then stirred in a reflux con-
denser at 70 ◦C overnight. Next, the solid product was quenched in
refrigerator at 4 ◦C for 1 h, filtered, washed with diethyl ether and
dried in a vacuum oven at room temperature. The resulting product
was (C22H45 N+(CH3)2 C4H8 N+(CH3)2 C4H9)Br2 (denoted as
C22-4-4Br2).
SAPO-34: SAPO-34 was synthesized according to a procedure
described in a patent assigned to the Union Carbide Corporation
[19]. Example 35 was followed in this work. To this end, aluminum
isopropoxide, Ludox HS-30, orthophosphoric acid, and tetraethyl
ammonium hydroxide (TEAOH) were used to obtain a homoge-
neous gel with the composition 2 TEAOH: 0.3 SiO2: Al2O3: P2O5:
50 H2O. This gel was placed in a Teflon-lined autoclave and kept
in oven at 200 ◦C for 5 days. Afterwards, the solid material was
recovered by filtration.
Meso-SAPO: Mesoporous SAPO zeolites were synthesized
by addition of octadecyl-(3-trimethoxysilylpropyl)-ammonium
chloride (TPOAC, ABCR) as a mesopore-directing organosilane
surfactant to the synthesis gel of SAPO-34. Two samples were pre-
pared. In the first one, the silica source was stoichiometrically
replaced by TPOAC so that a molar gel composition of 0.3 TPOAC:
2 TEAOH: Al2O3: P2O5: 50 H2O was used. This sample is denoted
by meso-SAPO(TPOAC, ∞), the infinity sign indicating the molar
TPOAC/Si ratio. The other sample denoted by meso-SAPO(TPOAC,
0.04) was obtained by addition of a small amount of TPOAC to the
synthesis of SAPO-34. The gel composition was 0.012 TPOAC: 2
TEAOH: 0.3 SiO2: Al2O3: P2O5: 50 H2O.
Meso-SAPO(C22-4-4, n): In this case, 50% and 100% of the TEAOH
compound in the standard SAPO-34 synthesis was replaced by
C22-4-4Br2. The gel compositions were C22-4-4Br2: 0.3 SiO2: Al2O3:
P2O5: 50 H2O: 2NH3·H2O and TEAOH: 0.5 C22-4-4Br2: 0.3 SiO2:
Al2O3: P2O5: 50 H2O: NH3·H2O, respectively. The synthesis pro-
cedure for these two zeolites is similar to conventional SAPO-34.
The zeolites are denoted as meso-SAPO(C22-4-4, n), where n refers
to the fraction of TEAOH replaced by C22-4-4Br2 in the synthesis gel.
All the materials were finally calcined at 550 ◦C for 5 h in static air
before further use.
2.1.2. Synthesis of molecular sieves
SSZ-13(50): SSZ-13 was synthesized as described in litera-
ture [17]. An amount of 2 g of a 1 M NaOH solution, 4 g 0.5 M
N,N,N-trimethyl-1-adamantanammonium hydroxide (TMAdOH,
SACHEM, 25%) and 2 g deionized water were mixed together. 0.05 g
aluminum hydroxide (Sigma Aldrich) was added to this solution
under vigorous stirring. After 30 min, 0.24 g fumed silica (Sigma)
was added. The resulting mixture was stirred at room temperature
to obtain a homogeneous gel with the composition 20 TMAdOH:
10 Na2O: Al2O3: 100 SiO2: 4400 H2O, which was then transferred
into a Teflon-lined autoclave and kept in an oven at 160 ◦C for 4
days. The zeolite product was filtered, dried and calcined at 550 ◦C
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
3
2.2. Physicochemical characterization
2.2.1. Basic characterization
X-ray diffraction patterns were recorded on a Bruker D4
Endeavor diffractometer using Cu K␣ radiation in the 2Â range
of 5–60◦. Elemental analyses were carried out by ICP-OES (Spec-
tro Ciros CCD ICP optical emission spectrometer with axial plasma
viewing). To extract the metals, the catalysts were dissolved in
1.5 ml of an acid mixture of HF/HNO3/H2O. Argon adsorption exper-
iments were determined at 87.6 K on a Micromeritics ASAP 2020
instrument in static mode. The samples were outgassed at 400 ◦C
for 8 h prior to the sorption measurements. The Langmuir adsorp-
tion isotherm model was used to determine the total surface area
(SL) in the p/p0 range between 0.05 and 0.20. The mesopore volume
(Vmeso) and mesopore size distribution was calculated from the
adsorption branch of the isotherm by the Barrett–Joyner–Halenda
(BJH) method. The micropore volume (Vmicro) was determined by
the t-plot method (thickness range 0.34–0.40 nm).
2.2.2. Electron microscopy
Scanning electron microscopy (SEM) pictures were taken on a
FEI Quanta 200F scanning electron microscope at an accelerating
voltage of 3–5 kV. The catalysts were coated with gold prior to mea-
surements. Transmission electron microscopy (TEM) pictures were
taken on a FEI Tecnai 20 at 200 kV. The catalysts were suspended
in ethanol and dispersed over a carbon coated holey Cu grid with a
film prior to measurements.
Fig. 1. XRD patterns of as-synthesized (a) SSZ-13(50), (b) meso-Zeo(50, C22-4-4), and
(c) meso-Zeo(50, TPOAC).
on TGA/DSC 1 STAR system of Mettler Toledo. The temperature
was increased to 850 ◦C at a rate of 5 ◦C min−1 under flowing air
(50 ml min−1).
2.2.3. Vibrational spectroscopy
3. Results and discussion
FTIR spectra of CO adsorbed to the zeolite samples were
recorded in the range of 4000–400 cm−1 by a Bruker Vertex V70 v
instrument. The spectra were acquired at a 2 cm−1 resolution and
averaged over 20 scans. The samples were prepared as thin self-
supporting wafers of 5–10 mg cm−2 and placed inside a controlled
environment infrared transmission cell, capable of heating and
cooling, gas dosing and evacuation. Prior to CO adsorption, the cat-
alyst wafer was heated to 550 ◦C at a rate of 2 ◦C min−1 in an oxygen
atmosphere. Subsequently, the cell was outgassed at the final tem-
perature until the residual pressure was below 5 × 10−5 mbar. The
sample was then cooled to 80 K. CO was introduced into the sample
cell via a sample loop (5 ␮l) connected to a Valco six-port valve.
3.1. Synthesis and characterization
Fig. 1 shows the XRD patterns of the (aluminosilicate) zeolites
synthesized at a gel Si/Al ratio of 50. The patterns of SSZ-13(50)
and meso-Zeo(50, TPOAC) are consistent with predominant for-
mation of crystalline zeolite with the CHA topology. The pattern
of meso-Zeo(50, C22-4-4) is that of MFI zeolite. The XRD patterns of
the calcined zeolites (not shown) are very similar to those of the
as-synthesized zeolites. SEM images show relatively large pseudo-
cubic crystals for the bulk SSZ-13 zeolite (Fig. 2). These crystals
are more rounded than the crystals synthesized at a Si/Al ratio
of 20 [8]. TEM images of meso-Zeo(50, C22-4-4) evidence that the
crystalline domains of the ZSM-5 zeolite are relatively small, sug-
gesting that the mesoporogen limited the crystal growth also in
this case, although no CHA zeolite had formed. The zeolite pre-
pared with TPOAC has the same morphology as a sample prepared
at a Si/Al ratio of 20 [8], however, with less well-defined cubic crys-
tals (Fig. 2c). This material is well-crystallized and made up from
many small crystals.
2.2.4. NMR spectroscopy
Nuclear Magnetic Resonance (NMR) spectra were recorded on
a Bruker DMX-500 NMR spectrometer. For the 27Al Magic Angle
Spinning (MAS) NMR a standard Bruker MAS probehead was used
with rotor diameter of 2.5 mm, at a spinning rate of 20 kHz. The 27Al
chemical shift is referenced to a saturated Al(NO3)3 solution.
2.3. Catalytic activity measurements
Fig. 3 shows Ar physisorption isotherms and BJH pore size
distributions of the calcined zeolites. The textural properties
derived thereof are summarized in Table 1. The isotherm of SSZ-
13(50) is type I, which is characteristic for microporous materials.
The isotherms of the two zeolites synthesized with mesopore-
directing agents have typical type IV isotherms, representative
of mesoporous materials. All three materials have high Langmuir
surface areas. The total pore volume of meso-Zeo(50, C22-4-4) is
substantially lower than for the other two zeolites, consistent
with dominant formation of ZSM-5 zeolite, which usually exhibits
pore volumes around 0.15 cm3 g−1. The physisorption isotherm
signifies that uniform mesopores have formed in this material. The
use of TPOAC in SSZ-13 synthesis led to a relatively small increase
of the mesopore volume as compared to the use of C22-4-4Br2 for
SSZ-13 with a Si/Al ratio of 20 [8,9]. The pore size distribution
curves in Fig. 3 do not provide evidence for formation of ordered
mesoporosity.
Catalytic activity measurements were carried out in a quartz
tubular fixed-bed reactor. First, the zeolites were pressed, crushed
and sieved in a particle size fraction between 250 and 500 ␮m.
Second, 50 mg of the shaped catalyst was placed in a quartz tube
(inner diameter 4 mm) between two quartz-wool plugs. Prior to
the reaction, the catalyst was activated at 550 ◦C in artificial air
(30 ml min−1) for 2 h. The methanol-to-olefins reaction was per-
formed at 350 ◦C for SSZ-13 zeolites and 450 ◦C for SAPO zeolites.
Methanol (Merck, 99%) was introduced to the reactor by flow-
ing He through a saturator kept at −17 ◦C with the flow rate
30 ml min−1. The WHSV was kept at 0.8 g g−1 h−1 and the efflu-
ent was analysized online by gas chromatography (Compact GC
Interscience equipped with TCD and FID detectors with RT-Q-Bond
and Al2O3/KCl columns). The amount of coke deposited during
the reaction was determined by thermogravimetric analysis (TGA)
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
4
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
Fig. 2. SEM (a–c) and TEM (d and e) images of (a) SSZ-13(50), (b and d) meso-Zeo(50, C22-4-4), and (c and e) meso-Zeo(50, TPOAC).
Fig. 3. Ar physisorption (left) and pore size distribution (right) of (a) SSZ-13(50), (b) meso-Zeo(50, C22-4-4), and (c) meso-Zeo(50, TPOAC). The isotherms were vertically
offset by equal intervals of 150 cm3 g−1. The pore size distribution was calculated via BJH algorithm using the adsorption branch and vertically offset by equal intervals of
0.003 cm3 g−1 nm−1
.
27Al MAS NMR spectra of these three zeolites (not shown) are
dominated by a feature around 59 ppm, which is characteristic for
tetrahedrally coordinated Al atoms in the zeolite framework. A
small feature around 0 ppm is assigned to octahedral Al species.
Deconvolution of 27Al MAS NMR spectra shows that more than 90%
of the Al species is built into the zeolite framework for all three
samples.
Fig. 4 shows XRD patterns for SAPO-34, meso-SAPO(TPOAC)
and meso-SAPO(C22-4-4). The meso-SAPO(TPOAC) material has the
CHA structure similar to SSZ-13 and SAPO-34. Compared to the
Table 1
Textural properties of the calcined SSZ-13 and meso-SSZ-13 materials determined by Ar physisorption.
b
c
d
e
Sample
Si/Al ratioa
SL (m2 g−1
)
Vtot (cm3 g−1
)
Vmeso (cm3 g−1
)
Vmicro (cm3 g−1
)
SSZ-13(50)
meso-Zeo(50, TPOAC)
meso-Zeo(50, C22-4-4
SAPO-34
meso-SAPO(TPOAC, 0.04)
meso-SAPO(C22-4-4, 100)
54
61
51
–
–
–
599
792
558
735
778
247
0.22
0.31
0.24
0.30
0.36
0.55
0.01
0.04
0.06
0.03
0.11
0.51
0.20
0.22
0.14
0.22
0.21
0
f
)
a
Si/Al ratio of aluminosilicate zeolites determined by ICP elemental analysis.
SL is the Langmuir surface area obtained in the relative pressure range (p/p0) of 0.05–0.20.
Vtot is the total pore volume at p/p0 = 0.97.
Vmeso is the mesopore volume calculated from BJH method.
Vmicro is the micropore volume calculated from t-plot method.
Pedominantly ZSM-5.
b
c
d
e
f
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 4. XRD patterns of as-synthesized (a) SAPO-34, (b) meso-SAPO(TPOAC, 0.04), (c) meso-SAPO(TPOAC, ∞), (d) SAPO-34, (e) meso-SAPO(C22-4-4, 50), and (f) meso-
SAPO(C22-4-4, 100) (* and + indicate reflections of CHA and AFI phases).
volume is substantially higher (0.11 cm3 g−1 vs. 0.03 cm3 g−1 for
SAPO-34). It can be seen that the micropore volumes of SAPO-34
and meso-SAPO(TPOAC, 0.04) are very similar to the micropore vol-
SAPO-34 reference, the reflections for meso-SAPO(TPOAC, 0.04)
are, however, substantially lower and broader, which is indicative
of the smaller size of the crystalline domains for this mesoporous
material. When all Si of the Ludox HS-30 silica source in the syn-
thesis gel has been replaced by TPOAC, AlPO-5 (AFI topology) was
obtained with only a small amount of SAPO-34 (Fig. 4c). The AlPO-5
structure contains one-dimensional 12-membered ring channels.
Similar to the successful synthesis of mesoporous SSZ-13 [8], we
also attempted to use the diquaternary ammonium-type surfactant
ume of SSZ-13(50). The textural properties of meso-SAPO(C22-4-4
,
100) show that this layered material does not contain micropores.
Summarizing, we have found that the use of TPOAC results
in a small amount of mesopores in high-silica (Si/Al = 50) SSZ-13.
Attempts to employ C22-4-4Br2 as mesoporogen to synthesize hier-
archical SSZ-13 at this relatively high Si/Al ratio were unsuccessful.
Instead, the more stable ZSM-5 zeolite formed. Similar observations
were made when these strategies were employed to SAPO-34 syn-
thesis. Addition of small amounts of TPOAC to the synthesis gel of
SAPO-34 led to generation of significant mesoporosity in SAPO-34.
An increase of the TPOAC content led to nearly exclusive formation
of AlPO-5 instead of SAPO-34. The use of C22-4-4Br2 during SAPO-34
synthesis resulted in formation of other phases than SAPO-34.
Before discussing the catalytic activity of some of these mate-
rials in the MTO reaction, we will first compare the acidity of
SAPO-34 and meso-SAPO(TPOAC, 0.04). Fig. 7 displays the FTIR
spectra of increasing doses of CO adsorbed on calcined SAPO-
34 and meso-SAPO(TPOAC, 0.04). The high-frequency (HF) band
at 3630 cm−1 and the low-frequency (LF) band at 3600 cm−1 are
assigned to Brønsted acid sites. The intensities of these bands for
meso-SAPO(TPOAC, 0.04) are lower than those for SAPO-34, which
indicates that the mesoporous zeolite has a smaller number of
Brønsted acid sites. Upon CO adsorption, the bands due to Brønsted
acid sites shift to 3360 cm−1. The equal red-shift of the hydroxyl
frequency, ꢀꢁ(OH) = −270 cm−1, demonstrates that the strength
of the acid sites in these molecular sieves is very similar. The shift
is smaller than typically observed for the aluminosilicate zeolites,
which are indeed usually considered to be stronger Brønsted acids
[22,23]. Upon increasing the CO pressure, the HF band erodes faster
than the LF band, which points to stronger acidity or better accessi-
bility of the former hydroxyl groups. The spectra representing the
CO stretching are also shown. The band at 2171 cm−1 appearing
upon CO exposure is ascribed to CO adsorbed on Brønsted acid sites.
Consistent with the view put forward by Almutairi et al., the polar-
ization of the CO bond is not an adequate indicator of Brønsted
acidity, as the blue shift of perturbed CO to 2171 cm−1 is similar to
that for the stronger acidic HZSM-5 zeolite [14,24].
C22-4-4Br2 to obtain mesoporous SAPO-34. Two different synthe-
sis gel compositions were used in which either half or all of the
TEAOH in a typical gel for SAPO-34 synthesis was replaced by
C22-4-4Br2 (equivalent amount of N atoms). The XRD patterns of
meso-SAPO(C22-4-4, 50) and meso-SAPO(C22-4-4, 100) are shown in
Fig. 4e and 4f, respectively. In neither case, SAPO-34 was formed.
Meso-SAPO(C22-4-4, 50) turned out to have the AlPO-15 structure
[20,21], whilst the structure of the other material could not be
matched with patterns from known databases.
Electron microscopy shows that SAPO-34 consists of small
particles intergrown into larger agglomerates (Fig. 5). The crys-
tals of the mesoporous SAPO-34, meso-SAPO(TPOAC, 0.04), are
slightly smaller than those of SAPO-34. The AlPO-5 sample con-
sists of large intergrown particles. The SEM image of AlPO-15
(meso-SAPO(C22-4-4, 50)) shows large crystals with a square prism
morphology without any sign of mesoporosity. The scanning and
transmission electron micrographs for meso-SAPO(C22-4-4, 100) are
very different and show the formation of agglomerates of a layered
material. Its layered nature is consistent with the presence of a low-
angle diffraction peak in Fig. 4f. The significant background in this
XRD pattern around 23◦ and the observation of globular particles
by TEM (Fig. 2i) indicates that a significant amount of amorphous
silica was also present. All this suggests that crystallization of this
novel material is far from complete.
The Ar physisorption isotherms of SAPO-34, meso-SAPO(TPOAC,
0.04) and meso-SAPO(C22-4-4, 100) are shown in Fig. 6 and the text-
ural properties are listed in Table 1. The isotherm of SAPO-34 is of
type I. The isotherm of meso-SAPO(TPOAC, 0.04) is a combination of
type I and IV, which is characteristic of porous materials containing
micro- as well as mesopores. Meso-SAPO(TPOAC, 0.04) has a similar
surface area and micropore volume as SAPO-34, but its mesopore
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
6
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
Fig. 5. Electron microscopy images of (a) conventional SAPO-34, (b and d) meso-SAPO(TPOAC, 0.04), (c and e) meso-SAPO(TPOAC, ∞), (f) meso-SAPO(C22-4-4, 50), (g–i)
meso-SAPO(C22-4-4, 100).
Fig. 6. Ar physisorption (left) and PSD (right) of (ꢀ) SAPO-34, (ꢁ) meso-SAPO(TPOAC, 0.04) and (ꢀ) meso-SAPO(C22-4-4, 100). The pore size distribution was calculated via
BJH algorithm using the adsorption branch.
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
7
Fig. 7. FTIR spectra of CO adsorbed on SAPO-34 (a and b) and meso-SAPO(TPOAC, 0.04) (c and d) at liquid N2 temperature: (left) OH stretching region, and (right) CO stretching
region (trace 1 is the spectrum of dehydrated zeolite before CO adsorption; traces 2–4 after introduction of 0.55, 0.85, 1.20, 1.85 ␮mol CO in the IR cell; the spectra normalized
by sample weight).
3.2. Catalytic activity measurements
of mesoporous SSZ-13 by lowering the rate of coke formation on
or close to the external surface.
We first discuss the performance of silylated mesoporous
SSZ-13 in the MTO reaction. The rationale of silylating the most
promising mesoporous SSZ-13 zeolite, meso-SSZ-13(C22-4-4, 0.17)
with a Si/Al ratio of 20 [8,9] was to evaluate whether deactivation
of acid sites at or close to the external surface would have a benefi-
The catalytic activities of the SSZ-13 zeolites with a Si/Al ratio
of 50 at a WHSV of 0.8 g g−1 h−1 are shown in Fig. 9 and the
corresponding performance data are listed in Table 2. Initially,
all catalysts show complete conversion of methanol to products
with carbon–carbon bonds. For SSZ-13(50), the lifetime in MTO
reaction is about 7 h, which is significantly higher than that of
SSZ-13(20) (2.8 h) [8,9]. This shows that a lower Brønsted acid site
density results in a lower rate of deactivation. When additional
mesoporosity is introduced in SSZ-13(50) by use of TPOAC (meso-
SSZ-13(50, TPOAC)), the lifetime of the catalyst is further improved.
The increase in lifetime upon mesopore introduction in SSZ-13(50)
is however much lower than earlier found for SSZ-13(20) [8,9]. This
should be due to the lower rate of coke formation of SSZ-13(50),
cial effect on catalyst stability. Treatment of meso-SSZ-13(C22-4-4
,
0.17) with TEOS did not significantly affect the crystallinity of the
mesoporous SSZ-13 zeolite (Fig. 8a), yet it lowered the silanol
density. Fig. 8b shows the performance in the MTO reaction with
reaction time. Clearly, the rate of coke formation was decreased.
The product distributions (mainly C2 and C3 olefins) for both
zeolites were similar (Table 2). These results indicate that there is
room for improvement of the total methanol conversion capacity
Fig. 8. (left) XRD patterns and (right) catalytic performance in the MTO reaction (WHSV = 0.8 g g−1 h−1; T = 350 ◦C) of calcined (a) meso-SSZ-13(C22-4-4, 0.17) and (b) meso-
SSZ-13(C22-4-4, 0.17)-sil.
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
8
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
Table 2
Lifetime and product distributiona of zeolite catalysts for the MTO reaction (WHSV = 0.8 g g−1 h−1; T = 350 ◦C) after 1 h time on stream.
b
Zeolite
t50 (h)
CH4 (%)
C2 (%)
C2 (%)
C3 (%)
C3 (%)
Aliphatic C4–C6 (%)
SSZ-13(50)
7
9
8
1.4
1.2
1.0
1.3
52.6
51.4
45.8
48.5
0.2
0.2
0.4
0.3
41.5
44.1
47.7
45.7
2.7
1.6
2.9
2.9
1.5
1.4
2.2
1.3
Meso-Zeo(50, TPOAC)
Meso-Zeo(C22-4-4, 0.17)c
Meso-Zeo(C22-4-4, 0.17)-sil
10
a
Trace amounts of CO and CO2 not taken into account.
Catalyst lifetime defined as the time passed to reach a conversion of 50%.
[8].
b
c
Fig. 9. Catalytic performance in the MTO reaction (WHSV = 0.8 g g−1 h−1; T = 350◦C)
of (᭹) SSZ-13(50), (ꢁ) meso-Zeo(50, C22-4-4) and (ꢀ) meso-Zeo(50, TPOAC).
Fig. 10. Catalytic performance in the MTO reaction (WHSV = 0.8 g g−1 h−1; T = 450◦C)
of (ꢀ) SAPO-34, (᭹) meso-SAPO(TPOAC, 0.04), (ꢁ) meso-SAPO(TPOAC, ∞), (♦) meso-
SAPO(C22-4-4, 50) and (ꢁ) meso-SAPO(C22-4-4, 100).
which results in better utilization of the micropore space. The most
stable zeolite is meso-Zeo(50, C22-4-4), which had the MFI topology.
This zeolite showed high stability in the methanol conversion reac-
tion. As we were primarily interested in CHA zeolite in this study,
we did not determine the stability of the ZSM-5 zeolite in further
detail.
generating mesoporosity in SAPO-34. If this would be the case, the
lower stability of mesoporous SAPO-34 should be due to the lower
concentration of Brønsted acid sites. A rough estimation based on
the difference in the intensity of the CO stretching band around
2171 cm−1 shows that meso-SAPO(TPOAC, 0.04) contains about
40% less acid sites than SAPO-34. This difference coheres very well
with the shorter lifetime, so that the total methanol conversion
capacity of the two zeolites per Brønsted acid site is comparable.
To fortify this line of reasoning, Table 4 shows the textural proper-
ties and coke content of SAPO-34 and meso-SAPO(TPOAC, 0.04) as
the function of the time on the stream. After a reaction time of 1 h,
already a large part of the high Langmuir surface area of the parent
samples (>700 m2 g−1) has become inaccessible by coke. It is also
seen that the micropore volume has decreased below 0.10 cm3 g−1
(initial value ∼0.21 cm3 g−1). With increasing time on stream the
surface area of the two materials decreases in a very similar manner.
The coke build-up is slightly slower for meso-SAPO(TPOAC, 0.04),
which may be attributed to the lower acidity of this sample. Yang
et al. recently reported that nanosizing SAPO-34 crystals led to a
decrease in coke formation and, concomitant remarkable increase
of the lifetime in the MTO reaction [25]. This discrepancy may be
caused by the higher Brønsted acid intensity and the larger particle
Fig. 10 shows the catalytic properties of the SAPO materials in
the MTO reaction. The lifetime and the product distribution are
given in Table 3. Meso-SAPO(TPOAC, ∞) shows some activity with
a product distribution typical for ZSM-5 zeolite. The methanol con-
version for meso-SAPO(C22-4-4, 50) and meso-SAPO(C22-4-4, 100)
is very low. Both microporous SAPO-34 and meso-SAPO(TPOAC,
0.04) initially show complete conversion of methanol with good
selectivity to C2 and C3 olefins. Deactivation of the mesoporous
SAPO-34 zeolite sets in earlier than for the conventional SAPO-34
one. The lifetimes of SAPO-34 and meso-SAPO(TPOAC, 0.04) are
3 and 1.5 h, respectively. Thus, whilst the performance of SSZ-13
can be improved by introducing mesoporosity, such improvement
could not be brought about for SAPO-34. It is well known that the
Brønsted acid sites in SAPO-34 are much weaker than those in
SSZ-13 [22,23]. Accordingly, it is expected that coking in SSZ-13
proceeds at a higher rate than in SAPO-34. The lower rate of coke
formation for SAPO-34 may thus lead to more efficient utilization
of the micropore space. As a consequence, there is no benefit by
Table 3
Lifetime and product distributiona of SAPO-34 and meso-SAPO(TPOAC, 0.04) zeolite catalysts for the MTO reaction (WHSV = 0.8 g g−1 h−1; T = 450 ◦C) after 1 h time on stream.
Zeolite
Lifetimeb (h)
CH4 (%)
C2 (%)
C2 (%)
C3 (%)
C3 (%)
Aliphatic C4–C6 (%)
SAPO-34
3.0
1.5
2.0
2.1
3.8
4.5
41.9
40.7
16.1
0.2
0.2
0.2
39.5
37.9
34.9
0
0
0.5
16.3
17.4
43.8
meso-SAPO(TPOAC, 0.04)
meso-SAPO(TPOAC, ∞)
a
Trace amounts of CO and CO2 not taken into account.
Catalyst lifetime defined as the time passed to reach a conversion of 50%.
b
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057

G Model
CATTOD-8994; No. of Pages9
ARTICLE IN PRESS
L. Wu, E.J.M. Hensen / Catalysis Today xxx (2014) xxx–xxx
9
Table 4
Textural properties of spent SAPO-34 and meso-SAPO(TPOAC, 0.04) catalysts as a function of the time on stream determined by Ar physisorption.
TOSa (h)
SAPO-34
Vmeso (cm3 g−1
Meso-SAPO(TPOAC, 0.04)
SL (m2 g−1
)
Vmicro (cm3 g−1
)
)
Cokeb (wt.%)
SL (m2 g−1
)
Vmicro (cm3 g−1
)
Vmeso (cm3 g−1
)
Cokeb (wt.%)
1
4
12
48
347
295
164
149
0.09
0.07
0.04
0.04
0.08
0.10
0.07
0.06
3.5
11.7
12.8
13.7
300
284
191
141
0.08
0.07
0.05
0.07
0.09
0.11
0.08
0.10
3.6
6.9
8.0
10.0
a
Time on stream.
Coke content based on TGA.
b
size of their reference sample, and the much higher space velocity
of the MTO reaction employed by these authors.
University of Technology and the SACHEM company for providing
N,N,N-trimethyl-1-adamantanammonium hydroxide.
4. Conclusion
References
[1] C.D. Chang, Catal. Rev. Sci. Eng. 26 (1984) 323–345.
[2] M. Stocker, Microporous Mesoporous Mater. 29 (1999) 3–48.
[3] J.F. Haw, W.G. Song, D.M. Marcus, J.B. Nicholas, Acc. Chem. Res. 36 (2003)
317–326.
[4] F.J. Keil, Microporous Mesoporous Mater. 29 (1999) 49–66.
[5] U. Olsbye, S. Svelle, M. Bjorgen, P. Beato, T.V.W. Janssens, F. Joensen, S. Bordiga,
K.P. Lillerud, Angew. Chem. Int. Ed. 51 (2012) 5810–5831.
[6] J. Liang, H.Y. Li, S. Zhao, W.G. Guo, R.H. Wang, M.L. Ying, Appl. Catal. 64 (1990)
31–40.
[7] Z.M. Liu, C.L. Sun, G.W. Wang, Q.X. Wang, G.Y. Cai, Fuel Process. Technol. 62
(2000) 161–172.
[8] L. Wu, V. Degirmenci, P.C.M.M. Magusin, N.J.H.G.M. Lousberg, E.J.M. Hensen, J.
Catal. 298 (2013) 27–40.
[9] L. Wu, V. Degirmenci, P. Magusin, B.M. Szyja, E.J.M. Hensen, Chem. Commun.
48 (2012) 9492–9494.
[10] M. Choi, H.S. Cho, R. Srivastava, C. Venkatesan, D.H. Choi, R. Ryoo, Nat. Mater. 5
(2006) 718–723.
[11] M. Choi, K. Na, J. Kim, Y. Sakamoto, O. Terasaki, R. Ryoo, Nature 461 (2009)
246–250.
[12] M. Choi, R. Srivastava, R. Ryoo, Chem. Commun. (2006) 4380–4382.
[13] C. Jo, J. Jung, H.S. Shin, J. Kim, R. Ryoo, Angew. Chem. Int. Ed. 52 (2013)
10014–10017.
[14] S.M.T. Almutairi, B. Mezari, E.A. Pidko, P.C.M.M. Magusin, E.J.M. Hensen, J. Catal.
307 (2013) 194–203.
[15] M. Guisnet, L. Costa, F.R. Ribeiro, J. Mol. Catal. A: Chem. 305 (2009) 69–83.
[16] Y. Kumita, J. Gascon, E. Stavitski, J.A. Moulijn, F. Kapteijn, Appl. Catal., A: Gen.
391 (2011) 234–243.
[17] L.T. Yuen, S.I. Zone, Verified Syntheses of Zeolitic Materials, in: H. Robson (Ed.),
Synthesis Commission of the International Zeolite Association, 2nd ed., Elsevier,
2001.
[18] S.R. Zheng, H.R. Heydenrych, A. Jentys, J.A. Lercher, J. Phys. Chem. B 106 (2002)
9552–9558.
Several approaches to improve the catalytic performance of
SSZ-13 and SAPO-34 for application as acid catalysts in the
methanol-to-olefins (MTO) reaction were explored. Silylation of
a mesoporous SSZ-13 zeolite with a Si/Al ratio resulted in minor
improvement of the lifetime in the MTO reaction. Lowering the Al
content of SSZ-13 to a Si/Al ratio of 50 resulted in lower rate of
deactivation compared to SSZ-13 with a Si/Al ratio of 20 due to
the lower density of Brønsted acid sites. Additional mesoporosity
in SSZ-13(50) by use of an organosilane mesoporogen only results
in a minor improvement of the lifetime. This is because the lower
rate of coke formation for SSZ-13(50) as compared to SSZ-13(20)
results in better utilization of the micropore space. It was not pos-
sible to synthesize mesoporous SSZ-13 at high Si/Al ratios with the
C22-4-4Br2 template. Instead, ZSM-5 zeolite was obtained.
Attempts to synthesize SAPO-34 by use of C22-4-4Br2 as meso-
porogen were also not successful and other AlPO phases were
obtained. By use of TPOAC a mesoporous SAPO-34 was obtained.
The intracrystalline mesoporosity did not lower the deactivation
rate. The total methanol conversion capacity per acid site for micro-
porous and mesoporous SAPO-34 was comparable. These findings
are interpreted in terms of complete utilization of the micro-
pore of SAPO-34 in the MTO reaction. The very different findings
for SSZ-13, whose micropore space remains underutilized unless
mesopores are introduced, relate to the much lower acidity of sili-
coaluminophosphates. Accordingly, there is no benefit of additional
mesoporosity for SAPO-34.
[19] B.M. Lok, C.A. Messina, R.L. Patton, R.T. Gajek, T.R. Cannon, E.M. Flanigen, U.S.
Patent 4440871, 1984.
[20] J.J. Pluth, J.V. Smith, J.M. Bennett, J.P. Cohen, Acta Crystallogr. Sect. C: Cryst.
Struct. Commun. 40 (1984) 2008–2011.
[21] J.B. Parise, Acta Crystallogr Sect. C: Cryst. Struct. Commun (1984) 1641–1643.
[22] S. Bordiga, L. Regli, D. Cocina, C. Lamberti, M. Bjorgen, K.P. Lillerud, J. Phys. Chem.
B 109 (2005) 2779–2784.
[23] F. Bleken, M. Bjorgen, L. Palumbo, S. Bordiga, S. Svelle, K.P. Lillerud, U. Olsbye,
Top. Catal. 52 (2009) 218–228.
[24] S.M.T. Almutairi, B. Mezari, G.A. Filonenko, P. Magusin, M.S. Rigutto, E.A. Pidko,
E.J.M. Hensen, ChemCatChem 5 (2013) 452–466.
[25] G. Yang, Y. Wei, S. Xu, J. Chen, J. Li, Z. Liu, J. Yu, R. Xu, J. Phys. Chem. C 117 (2013)
8214–8222.
Acknowledgments
Financial support from the Technology Foundation STW, the
applied science division of The Netherlands Organization for
Scientific Research (NWO) and the Programme Strategic Alliances
between China and the Netherlands is acknowledged. We also
thank the Soft Matter Cryo-TEM Research Unit of Eindhoven
Please cite this article in press as: L. Wu, E.J.M. Hensen, Comparison of mesoporous SSZ-13 and SAPO-34 zeolite catalysts for the
methanol-to-olefins reaction, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.057