Crystallization and catalytic properties of molecular sieve SAPO-34 by a vapor-phase transport method

Article abstract of DOI:10.1039/c4ta06775d

Formation of molecular sieve SAPO-34 has been examined under vapor-phase transport (VPT) conditions. Powder XRD, SEM/EDX, and solid-state NMR were used to follow the evolution of gel samples as a function of crystallization time. The results indicate that the initial VPT dry gel is a mixture of an amorphous material (major component) and a dense phase (minor component). The amorphous material is aluminophosphate in nature, but its structure is not suitable for transformation into SAPO-34 since it only has octahedral Al. Introducing morpholine (the structure directing agent) onto the surface of the dry gel via a vapor phase leads to changes in the structure of the amorphous phase as is evidenced by the conversion of octahedral Al into the tetrahedral Al sites. The dense phase is transformed into the tridymite phase. Upon further heating under VPT conditions, the equilibrium between tridymite and amorphous phases shifts to the amorphous material which is subsequently reorganized to form SAPO-34. The ²⁹Si NMR results show that Si was incorporated into the framework at a later stage of crystallization via the formation of aluminosilicates. Catalytic tests indicate that the SAPO-34 product obtained by the VPT method had a longer lifetime for 100% methanol conversion and similar product selectivity to the SAPO-34 prepared by the normal hydrothermal synthesis. This journal is

Full text of DOI:10.1039/c4ta06775d

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Crystallization and catalytic properties of
molecular sieve SAPO-34 by a vapor-phase
Cite this: J. Mater. Chem. A, 2015, 3,
transport method†
4522
*
Lu Zhang and Yining Huang
Formation of molecular sieve SAPO-34 has been examined under vapor-phase transport (VPT) conditions.
Powder XRD, SEM/EDX, and solid-state NMR were used to follow the evolution of gel samples as a function
of crystallization time. The results indicate that the initial VPT dry gel is a mixture of an amorphous material
(major component) and a dense phase (minor component). The amorphous material is aluminophosphate
in nature, but its structure is not suitable for transformation into SAPO-34 since it only has octahedral Al.
Introducing morpholine (the structure directing agent) onto the surface of the dry gel via a vapor phase
leads to changes in the structure of the amorphous phase as is evidenced by the conversion of
octahedral Al into the tetrahedral Al sites. The dense phase is transformed into the tridymite phase. Upon
further heating under VPT conditions, the equilibrium between tridymite and amorphous phases shifts to
the amorphous material which is subsequently reorganized to form SAPO-34. The 29Si NMR results show
that Si was incorporated into the framework at a later stage of crystallization via the formation of
aluminosilicates. Catalytic tests indicate that the SAPO-34 product obtained by the VPT method had a
longer lifetime for 100% methanol conversion and similar product selectivity to the SAPO-34 prepared
by the normal hydrothermal synthesis.
Received 10th December 2014
Accepted 15th January 2015
DOI: 10.1039/c4ta06775d
www.rsc.org/MaterialsA
high selectivity for ethene and propene from the conversion of
methanol (MTO reactions).18–20 The catalytic properties of SAPO-
Introduction
34 are closely related to its framework compositions, which are
inuenced by the synthesis conditions as well as the choice of
the structure-directing agent (SDA). SAPO-34 is usually prepared
using the hydrothermal synthesis method (HTS). A variety of
SDAs have been used for the HTS synthesis.21–28 Among them,
morpholine is a good choice for preparing SAPO-34 with high
thermal and hydrothermal stability, high crystallinity, and high
Si content at a relatively low cost. To optimize the catalytic
properties of SAPO-34, understanding its formation mechanism
as well as the role of the SDA during crystallization is very
critical to controlling the Si incorporation and distribution as
well as the crystal size and the morphology. There are several
studies using morpholine as a SDA to examine the formation of
SAPO-34 under HTS conditions.28–31 However, SAPO-34 crystal-
lization with morpholine as a SDA is still not fully understood at
an atomic level.
An alternative synthetic method to HTS is vapor-phase
transport (VPT) which was rst reported by Xu et al.32 VPT is an
approach where a SDA solution placed at the bottom of an
autoclave is separated from the dry gel powder (for the reaction
vessel used for this method, see ESI, Scheme S1†). Upon heat
treatment, the SDA molecules transported via the vapor phase
will be in contact with the dry gel and help convert the amor-
phous dry gel powder into a crystalline sieve. The VPT method
has been used to synthesize various zeolites and AlPO4s.33–38
Since the discovery of aluminophosphate (AlPO4)-based mate-
rials by Union Carbide in 1982,1 much work has gone into this
type of molecular sieve. The structures of many AlPO4-n are
similar to those of zeolites, but their primary building units are
formed by Al–O–P linkages rather than by Si–O–Si or Si–O–Al
bridges in zeolites.2–4 The neutral framework of AlPO4s limits
their application as catalysts. The incorporation of silicon
atoms into their framework introduces negative charges, and
consequently the resulting materials, silicoaluminophosphates
(SAPOs), can be used as acid catalysts.3 Among SAPOs, SAPO-34
is recognized as one of the most important molecular sieves
because of its intermediate acidity and small pore size (around
0.38 nm).5 It has a CHA topology analogous to natural zeolite
chabazite. The framework is composed of CHA cages with eight-
membered ring (8R) windows.6 SAPO-34 has various applica-
tions including ion-exchange, heat storage, gas adsorption and
separation as well as catalysis,7–17 but it is best known for its
Department of Chemistry, The University of Western Ontario, London, Ontario, N6A
5B7, Canada. E-mail: yhuang@uwo.ca; Tel: +001 519 661-2111 ext. 86384
† Electronic supplementary information (ESI) available: One scheme and ve
gures: the diagram of the reaction vessel used for the VPT method; 27Al{31P}
REDOR and (B) 31P{27Al} REDOR spectra of the unwashed initial VPT dry gel;
powder XRD patterns, 31P MAS, and 27Al MAS spectra of selected washed VPT
dry gel samples; powder XRD pattern, 29Si MAS NMR spectrum, and SEM
images of SAPO-34 synthesized by the HTS method. See DOI: 10.1039/c4ta06775d
4522 | J. Mater. Chem. A, 2015, 3, 4522–4529
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
SAPO-34 has been previously synthesized by the VPT part was rst washed with distilled water in a beaker and then
method.35,39,40 For example, using diethylamine as a SDA, SAPO- dried in the same beaker in air without isolation from the liquid
34 crystallizes from a semi-crystalline layered precursor.40
phase. This ensured that no solid particles were washed away.
Since the initial “dry” gel powder contains 20–40% water by All dried samples were kept in sealed vials for further analysis.
weight, it is now generally recognized that the fundamental To compare the catalytic performance, a SAPO-34 sample
mechanisms underlying the crystallization in hydrothermal was also prepared hydrothermally according to the method
synthesis and VPT may be similar.41,42 The reaction is, however, reported in the literature.29,31 The initial gel composition
slower under the VPT conditions compared to the HTS due to (Al2O3 : P2O5 : SiO2 : morpholine : H2O) was 1.0 : 1.0 : 1.0 : 2.1 :
the lack of bulk liquid, which may allow the intermediates to
60. Specically, 15.7 g of Catapal B, 23.1 g of H3PO4, and 42.5 g of
be captured. Thus, examining the evolution of the dry gel water were mixed and stirred vigorously for 10 min, to which a
phase may provide additional information on understanding solution containing 20 g of colloidal silica and 21.3 g of water was
the crystallization in the conventional hydrothermal process. added. Separately, 18.3 of morpholine and 21.2 g of water were
Furthermore, as mentioned earlier, the SDA molecules are mixed. The resulting solution was then added slowly to the above
slowly brought into contact with the dry gel powder; it can be prepared alumina–phosphoric acid–silica mixture under stirring.
employed to probe the role of the SDA during the synthesis.
Once homogeneity was reached, the wet gel was placed into auto-
In this paper, the formation of SAPO-34 was examined under claves and heated in an oven at 473 K for 48 h. The solid product
the VPT conditions using morpholine as a SDA with attention was separated from the liquid phase by centrifugation. It was then
being paid to Si incorporation. For the reaction intermediates dried in air at room temperature and stored in sealed vials.
obtained during crystallization, the long-range ordering and the
local environments of P, Al, and Si were probed by powder X-ray
Characterization
diffraction (PXRD) and various solid-state NMR techniques
including 31P, 27Al, and 29Si magic-angle spinning (MAS), 27Al PXRD patterns were recorded on a Rigaku diffractometer using
triple-quantum MAS (3QMAS) NMR, 27Al{31P} and 31P{27Al} Co Ka radiation (l ¼ 1.7902 A). For the SEM image, the repre-
˚
rotational-echo double-resonance (REDOR). The combination sentative samples were rst coated with 3 nm osmium metal
of these techniques provides insights into the formation of using a Filgen OPC-80T instrument, and their morphology and
SAPO-34. Furthermore, the catalytic performance of the SAPO- elemental composition were then examined using a FIB/SEM
34 synthesized by the VPT method in MTO reactions was also 1540XB microscope equipped with an energy-dispersive X-ray
tested and compared with that of the SAPO-34 prepared by spectrometer for elemental analysis.
hydrothermal synthesis.
All the NMR experiments were carried out on a Varian/
Chemagnetics Innityplus 400 WB spectrometer equipped with
three rf channels operating at a eld strength of 9.4 T. The
Experimental
Sample preparation
1
Larmor frequencies of H, 31P, 27Al, and 29Si were 399.5, 161.7,
104.1, and 79.4 MHz, respectively. The magic angle was set
The synthesis of SAPO-34 was performed according to the general using the 79Br resonance of KBr. The chemical shis of 31P, 27Al,
VPT technique described elsewhere.35,39 The reagents used were and 29Si were referenced to 85% H3PO4 (0 ppm), 1 M Al(NO3)3
pseudo-boehmite (Catapal-B, ca. 65 wt% Al2O3, Vista), H3PO4 aqueous solution (0 ppm), and external tetrakis(trimethylsilyl)-
(85%, EM Science), morpholine (Aldrich), colloidal silica (Ludox silane (–SiMe3, ꢀ9.9 ppm), respectively. Depending on the
LS-30, Aldrich), and distilled water. The initial gel composition requirements of the individual experiment, three NMR probes
(Al2O3 : P2O5 : SiO2 : H2O) was 1.0 : 1.0 : 1.0 : 60.
(a Varian/Chemagnetics 7.5, a 4.0 mm H/X/Y triple-tuned T3
A typical procedure for the preparation of the initial VPT dry MAS probe, and a 5.0 mm H/F/X/Y triple-tuned MAS probe) were
31
gel was as follows: 31.4 g of Catapal B, 46.1 g of H3PO4, and 85.0 used. For P MAS experiments, a 45ꢁ pulse was typically used
g of distilled water were combined and subjected to vigorous and the recycle delay was 60 s. The proton-decoupling eld was
stirring for 15 min before adding another 42.5 g of distilled about 60 kHz. The 27Al spectra were acquired using a small
water. Separately, 40 g of colloidal silica and 42.5 g of distilled pulse angle with a pulse delay of 1 s. For the 29Si MAS experi-
water were mixed. This solution was added slowly to the ments, a 45ꢁ pulse was used with a pulse delay of 60 s. For 1H to
above alumina and phosphoric acid mixture under stirring for 29Si cross-polarization (CP) experiments, the 1H 90ꢁ pulse length
homogeneity. Aer being stirred at room temperature for was 5 ms and the Hartmann–Hahn condition was determined
30 min, the wet gel was dried at 353 K with constant stirring to using TTMSS. 31P{27Al} and 27Al{31P} REDOR experiments were
allow the evaporation of water until solids formed. The solid carried out using the 5.0 mm probe with a spinning speed of 8
sample was then ground into a ne powder (which is hereaer kHz The details of REDOR experiments were described in
referred to as the initial dry gel). A series of intermediates were previous studies.43,44 27Al 3QMAS experiments were performed
synthesized by placing 1.0 g of initial dry gel powder into small using the 4.0 mm probe and the spinning speed was 10 kHz.
Teon cups. Each cup was placed in a 23 mL Teon-lined The spectra were obtained by utilization of a three-pulse z-lter
autoclave with 1.2 g of 25% aqueous morpholine solution at the sequence.45 The rf strengths of the rst two hard pulses and
bottom and heated in an oven at 473 K. The reactions were third so pulse were optimized individually, and the optimized
quenched in cold water. For comparison, each solid sample was pulse lengths were 5.7, 2.0, and 15.0 ms for the three consecutive
divided into two parts. One part was dried in air and the other pulses.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 4522–4529 | 4523

View Article Online
Journal of Materials Chemistry A
Paper
5 h resulted in the emergence of several sharp peaks, whose
positions correspond to (101), (021), (211), (104), (401), and
(214) reections of trigonal SAPO-34. This indicates that when
the dry gel is in contact with morpholine for a longer time, it
gradually transformed into SAPO-34. This also implies that one
of the roles played by morpholine is to adjust the pH values of
the dry gel, thus promoting the transformation of dense phases
into SAPO-34.42 Further heating the VPT dry gel led to sharp-
ening of the peaks due to SAPO-34. The powder pattern of the
5-day sample only contains the peaks solely from SAPO-34.6
To further understand the role of the SDA under the VPT
conditions, two gel samples obtained aer VPT treatment for 2
and 5 h were put back in the Teon cups and reheated at 473 K
in the presence of a small amount of pure water at the bottom of
the autoclave. Interestingly, aer heating the sample for 5 days,
the 5 h sample (which is a mixture of the amorphous phase,
tridymite and a small amount of SAPO-34) almost completely
transformed into SAPO-34 (Fig. 2). The 2 h sample (which
contains mainly the amorphous material and a small amount of
dense phase), however, was only converted into pure tridymite.
It appears that for gel samples heated for 2 h or less, there were
not enough morpholine molecules occluded in the solids to
direct the reorganization of the gel to form SAPO-34. Therefore,
without a supply of additional morpholine molecules, resumed
heating can only produce the tridymite phase. The 5 h sample,
on the other hand, contains sufficient SDA molecules. Conse-
quently, no extra morpholine molecules are required to form
the CHA structure. The fact that the tridymite phase seen in the
XRD pattern of the 5 h sample disappeared aer reheating
implies that there exists an equilibrium between the tridymite
phase and the amorphous material, and the presence of suffi-
cient morpholine molecules shis the equilibrium towards
the amorphous phase, leading to subsequent conversion to
SAPO-34.
Catalytic tests
MTO reactions were carried out with a xed-bed reactor at
atmospheric pressure. 0.3 g of SAPO-34 catalyst (40–60 mesh)
was loaded into the reactor. The sample was preheated in
nitrogen ow at 773 K for 1 h, and then the temperature of the
reactor was decreased to 723–730 K. The methanol and nitrogen
mixture was then pumped into the reactor. The weight hourly
space velocity (WHSV) was 2.9 hꢀ1. The products were analyzed
on-line using a Varian GC3800 gas chromatograph equipped
with a FID detector and a PoraplotQ-HT capillary column.
Results and discussion
To follow the evolution of the long-range ordering of the
intermediate gel samples, the PXRD patterns of gel samples
were recorded as a function of crystallization time (Fig. 1A). The
powder pattern of the initial VPT dry gel presents a relatively
strong peak at about 24.4ꢁ, corresponding to a dense phase
with some long-range ordering. Generally, synthesis of SAPO
molecular sieves needs weakly acidic or neutral conditions.46
Thus, the existence of the dense phase in the initial dry gel is
not surprising as the lower pH value of the initial dry gel due
to the absence of morpholine favors the formation of dense
materials. In addition to the strong reection at about 24.4ꢁ, the
presence of several broad reections suggests the co-existence
of amorphous materials. In fact, as indicated by 31P MAS NMR
(shown later), the vast majority of the samples are actually
amorphous materials.
Aer heating the dry gel for 2 h, there appear several rela-
tively sharp peaks at around 23.7ꢁ, 25.0ꢁ, and 26.3ꢁ in the
powder pattern. The corresponding d spacings agree with those
of the dense AlPO4 form of tridymite.47,48 Heating the dry gel for
31P and 27Al MAS NMR spectra were obtained to characterize
the local environments of P and Al in the intermediate gel
samples. Fig. 3A shows the 31P MAS spectra of unwashed VPT
dry gel samples. The 31P spectrum of the initial dry gel exhibits a
strong peak at ꢀ20 ppm, a weak resonance at ꢀ30 ppm, as well
as several small peaks in the region from ꢀ8 to ꢀ12 ppm. The
ꢀ20 ppm peak has been observed in the dry gel phases of many
AlPO4-and SAPO-based molecular sieves and assigned to the not
fully condensed HOP(OAl)3 environment in the amorphous
materials.43,49–53 The weak peak at ꢀ30 ppm is due to the fully
Fig. 1 Powder XRD patterns of unwashed VPT dry gel samples. Fig. 2 Powder XRD pattern of selected unwashed VPT dry gel samples
* indicates the tridymite phase.
reheated with a small amount of water.
4524 | J. Mater. Chem. A, 2015, 3, 4522–4529
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
Heating the dry gel for 1 h results in the appearance of a new
peak at around 38 ppm which can be ascribed to tetrahedral Al.
Upon increasing the heating time, the intensity of the signal
due to tetrahedral Al grew at the expense of the octahedral Al
peak, suggesting that reorganization of the gel structure has
occurred under the inuence of the SDA. The 27Al MAS spec-
trum of the 5 day sample agrees with those of SAPO-34 reported
in the literature.6 However, since the resolution of the spectrum
is low, it is unclear how many peaks are present. The low-
resolution is due to the fact that for quadrupolar nuclei such as
27Al (I ¼ 5/2), the spinning sample at the magic angle cannot
completely average out a second-order quadrupolar interaction
that broadens the central transition. To achieve high resolution
and identify individual Al species in the nal 5 day product, 27Al
3QMAS experiments were carried out. 3QMAS (or MQMAS) is a
2D technique that provides a high-resolution spectrum in the
second (F1) dimension free of the residual quadrupolar inter-
action and separates the overlapping signals. The spectrum of
the SAPO-34 product (the 5 day sample) is illustrated in Fig. 4,
which clearly exhibits four signals. The broad peak at around
38 ppm in the MAS spectra is actually composed of two
components with isotropic chemical shi values of about 48
and 58 ppm. Based on the previous work,50,58 the 48 ppm peak
can be assigned to the tetrahedral Al with the Al(OP)4 environ-
ment and the 58 ppm peak to the tetrahedral Al in alumino-
silicate domains with Si as the nearest neighbor.56,59 The peak at
6 ppm in the MAS spectrum appears in the F1 dimension with
an isotropic chemical shi value of around 12 ppm. As dis-
cussed earlier, it is due to unreacted alumina. There is also a
peak with an isotropic chemical shi of ꢀ12 ppm, originating
from the octahedral Al in the framework.55,56
Fig. 3 31P MAS (A) and 27Al MAS (B) spectra of the unwashed VPT dry
gel sample. Asterisks indicate spinning sidebands.
condensed P(OAl)4 species, which is from the dense phase
represented by the reection at 24.4ꢁ in the corresponding XRD
patterns. The 31P MAS NMR spectra conrm the results of XRD
that the dry gel is indeed a mixture of amorphous materials and
a dense phase and their relative intensities further indicate that
the amount of dense phase is actually small. Aer 5 h of the VPT
treatment, a new peak at ꢀ28 ppm started emerging, indicating
the formation of the P(OAl)4 environment in SAPO-34.54 This is
consistent with the XRD results. The spectrum also shows that
the previously obtained relatively broad peak at ꢀ30 ppm
sharpens signicantly coinciding with the formation of tridy-
mite whose 31P MAS spectrum also exhibits a single peak at
ꢀ30 ppm.54 It is also noticed that there are a signicant
amount of amorphous peaks with chemical shis greater than
ꢀ20 ppm, which conrms that at a longer heating time, the
equilibrium between tridymite and amorphous materials shis
in favor of the amorphous phase. The spectrum of the nal
product (5 day sample) has a single sharp peak at ꢀ28 ppm and
is consistent with that of pure trigonal SAPO-34 reported in the
literature.55
Previous work has shown that the structures of the inter-
mediate phases formed during the crystallization of several
AlPO- and SAPO-based molecular sieves under VPT conditions
are held by weak non-covalent bonding interactions.37,40,60 To
investigate the nature of the bonding in VPT dry gel samples,
selected intermediates were washed with water. Upon washing,
both the XRD patterns (Fig. S2†) and the NMR spectra (Fig. S3†)
exhibit little change, implying that in the present case, the
structures of the intermediate phases are held by covalent
bonds.
The 27Al MAS spectra convey similar information. The
spectrum of the initial VPT dry gel contains a strong peak at
ꢀ12 ppm and a much weaker one at 6 ppm (Fig. 3B). The
ꢀ12 ppm peak indicates the existence of octahedral Al in the
amorphous phase.56 The 6 ppm peak is due to unreacted
alumina.31,57 There is no peak due to tetrahedral Al in the
spectrum of the initial dry gel, indicating that the presence of
the SDA is critical to the formation of tetrahedral Al in the gel.
To conrm the amorphous material, which is the major
component of the initial dry gel, is AlPO in nature, we carried
out 31P{27Al} and 27Al{31P} REDOR experiments. REDOR exper-
iments are based on the 31P–27Al dipolar interaction. In a 31P
{27Al} REDOR difference spectrum, only the P sites with Al in
close proximity can be observed. Fig. S1† shows that the 31P
{27Al} REDOR difference spectrum displays the ꢀ20 ppm peak
seen in the 31P MAS spectrum and that the 27Al{31P} REDOR
difference spectrum has the ꢀ12 ppm peak seen in the 27Al MAS
spectrum.
Fig. 4 27Al 3QMAS spectrum of the 5 day sample.
J. Mater. Chem. A, 2015, 3, 4522–4529 | 4525
This journal is © The Royal Society of Chemistry 2015

View Article Online
Journal of Materials Chemistry A
Paper
To examine the morphological evolution as a function of
crystallization time, several representative gel samples were
chosen for SEM analysis (Fig. 5). The image of the initial VPT dry
gel shows the materials with amorphous nature. The image of
the 5 h sample clearly indicates the beginning of the formation
of SAPO-34 crystals (Fig. 5B, labeled with an arrow). The picture
of the 1 day sample shows that more SAPO-34 crystals emerge
from the gel (Fig. 5C). Aer 5 days of heating, a highly crystalline
SAPO-34 product was formed. The trend agrees with the XRD
and NMR results well.
To follow the Si incorporation, 29Si MAS spectra were
obtained (Fig. 6A). 1H / 29Si cross-polarization (CP) experi-
ments were also carried out on the VPT dry gel samples (Fig. 6B).
CP is a technique that utilizes the H–29Si dipolar interaction.
1
Fig. 6 29Si MAS (A) and 1H / 29Si CP (B) spectra of unwashed VPT dry
Thus, it can be used to selectively enhance the intensity of 29Si
nuclides with protons nearby. The 29Si MAS spectra of the VPT
dry gel samples heated for less than 5 h exhibit a broad peak at
around ꢀ111 ppm. It is assigned to unreacted amorphous silica
with Si(OSi)4 environments.6,61 This observation implies that the
majority of silica in these gel samples has not started reacting.
In the corresponding CP spectra, besides the peak at ꢀ111 ppm,
there exists another stronger signal at ꢀ101 ppm, which is not
visible in the MAS spectra. This signal mainly results from the
(SiO)3SiOH species on the surface of silica particles. Aer 5 h
of heating, the CP spectrum shows two new broad peaks at
ꢀ85 and ꢀ90 ppm. The former represents (AlO)3SiOH and/or
(SiO)(AlO)Si(OH)2 and the latter Si(OAl)4, respectively.62 It
appears that the Si is incorporated into SAPO-34 via the
formation of a small amount of aluminosilicate particles.40 The
MAS spectrum of the SAPO-34 nal product contains a sharp
peak at ꢀ90 ppm, which is due to the isolated Si(OAl)4 species in
SAPO-34.26,61,63–65 There is also a weak one at ꢀ95 ppm and a
broad one at ꢀ111 ppm. They are due to a small amount of
Si(OAl)3(OSi)1 and Si(OSi)4 species in the framework.59
gel samples with a contact time of 0.5 ms.
4) environments are produced when a combination of SM II and
III occurs. In AlPO4-based molecular sieves the P/Al ratio is strictly
1 : 1. Upon Si incorporation into AlPO4s, the P/Al ratio will change
and the actual value reects the Si incorporation mechanisms. To
obtain semi-qualitative information on elemental compositions,
EDX analysis was performed on the 5 day sample. The results were
the average data obtained from three different SAPO-34 crystals.
The Si content in the as-formed SAPO-34 is about 19 mol%, which
is much higher than the proposed upper threshold value (11%) of
the SM II mechanism in SAPO-34.67,68 Further, the (Si + P)/Al ratio
(1.27) is greater than 1. The EDX data thus indicate that Si is
incorporated via both SM II and III with SM II being the main
pathway. These results are consistent with the 29Si NMR spectrum
of SAPO-34 (Fig. 6A), which exhibits a sharp and strong signal at
ꢀ90 ppm due to the dispersed Si(OAl)4 environments and two
weaker ones at ꢀ95 and ꢀ111 ppm originating from the Si species
in Si islands.
Flanigen et al. have proposed three Si substitution mecha-
nisms.66 The rst is Si incorporation into an Al site (SM I). The
second is Si incorporation into a P site (SM II). This gives rise to an
isolated Si(OAl)4 species, therefore a Brønsted acid site. The third
one involves replacing an Al–P pair by two Si atoms (SM III). Si
islands with various Si species in Si(OAl)n(OSi)4ꢀn (n ¼ 0, 1, 2, 3, or
Vistad et al.30 and Yan et al.31 previously studied the crystal-
lization of SAPO-34 by the HTS method with morpholine as
the SDA. They reported that SAPO-34 was formed from the
Fig. 5 SEM images of selected unwashed VPT dry gel samples: (A) Fig. 7 Methanol conversion vs. time-on-stream for different SAPO-34
initial dry gel, (B) 5 h, (C) 1 day, and (D) 5 day samples.
products.
4526 | J. Mater. Chem. A, 2015, 3, 4522–4529
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
Table 1 Product selectivity of MTO reaction over different SAPO-34 products
Selectivity (wt%)
Samples
VPT
Time (min)
CH4
C2H4
C2H6
C3H6
C3H8
C4
C5
C6
C2] + C3]
4
157a
4
1.7
1.5
1.4
1.5
31.6
45.6
37.6
45.1
1.3
1.0
0.7
1.1
34.7
36.6
38.1
35.8
12.5
2.5
5.6
3.4
13.0
9.6
13.3
10.2
4.2
2.6
2.7
2.5
0.9
0.6
0.5
0.5
66.3
82.1
75.7
80.9
HTS
143a
a
The last data point before the catalyst became deactivated (methanol conversion > 99.99 wt%).
transformation of an amorphous phase aer several hours of effect of SAPO-34 on its lifetime in MTO reactions was reported
heating. Apparently, under VPT conditions, crystallization of recently by Liu et al.20
SAPO-34 takes a much longer time. This is because, due to the
absence of morpholine molecules and consequently a lower pH
value, the initial VPT dry gel has a structure not readily trans-
Conclusions
formed into the CHA structure. Once the SDA molecules were
brought into contact with the dry gel via the vapor phase, the gel
underwent structural reorganization and eventually evolved
into SAPO-34. When the SDA molecules were rst brought to the
surface of the dry gel, their distribution over the entire solid gel
may not be homogeneous. Due to the lack of the apparent liquid
phase, they diffused through the solid sample slowly.
Formation of SAPO-34 was examined by the VPT method.
Several techniques including PXRD, SEM/EDX, and solid-state
NMR were utilized to characterize the intermediate gel samples.
Under the VPT conditions, the initial dry gel contains amor-
phous materials which are not readily transformed into the
CHA structure due to the lack of the SDA in the gel phase as well
as a small amount of the dense phase. There is an equilibrium
between the dense phase and amorphous materials. Once the
morpholine molecules were gradually brought onto the surface
of the solid dry gel, the equilibrium shis to the amorphous
phase that subsequently reorganizes its own structure towards
the CHA structure. Si incorporation involves the formation of a
small amount of aluminosilicates rst and then incorporation
of isolated Si species and Si islands into the SAPO-34 frame-
work. The VPT-synthesized SAPO-34 shows a moderately longer
catalytic lifetime than the conventional HTS-synthesized SAPO-
34 in MTO reactions. The VPT sample shows comparable MTO
performances to the HTS sample.
To check the catalytic properties of the SAPO-34 prepared by
the VPT method, catalytic tests on MTO reactions were carried
out on the 5 day sample. A SAPO-34 sample synthesized by the
normal HTS method was also tested under the same reaction
conditions for comparison. The PXRD pattern of the HTS
sample shown in Fig. S4A indicates that it is pure trigonal SAPO-
34 with high crystallinity. Fig. 7 exhibits the plots of the meth-
anol conversion versus time-on-stream for the two samples. The
results show that the VPT sample has a longer lifetime as it
maintained a 100% conversion within 157 min, whereas
methanol conversion was kept at 100% within 143 min of time-
on-stream over the HTS sample that deactivates faster. Table 1
illustrates the selectivity towards different hydrocarbons. At the
initial stage of the reaction (4 min of time-on-stream), the
production of light olens (C2H4 and C3H6) was more favoured
on the HTS sample, whereas the selectivity for propane was
higher for the VPT sample. At a longer reaction time (above
140 min), the two samples show comparable product selectivity.
Acidity and particle size are two important factors that
inuence the catalytic performance of the SAPO-34 catalyst.69
The 29Si NMR spectrum of the HTS sample (Fig. S4B†) is
dominated by a very strong signal at ꢀ90 ppm. The EDX data
indicate that the HTS sample has a Si content of about 13 mol%,
which is less than 19 mol% in the VPT sample. It is reported
that at the borders of the Si islands exist acid centres stronger
than the isolated Brønsted acid sites.70 The relatively low light
olen selectivity and high propane selectivity over the VPT
sample aer 4 min of time-on-stream can be attributed to its
stronger acidity, which promoted hydrogen transfer reaction of
the olens to form alkanes.71 Its moderately longer lifetime may
be related to its smaller average crystal size (40 mm, Fig. 5D)
than the HTS sample (around 60 mm, Fig. S5†). A similar size
Acknowledgements
Y. H. thanks the Natural Science and Engineering Research
Council of Canada for a Discovery grant and a Discovery
Accelerator grant. Funding from the Canada Research Chair
program is also acknowledged. The authors sincerely thank
Drs Peng Tian and Zhongmin, Liu (Laboratory of Applied
Catalysis, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, China) for their kind assistance in cata-
lytic measurements.
Notes and references
1 S. T. Wilson, B. M. Lok, C. A. Messina, T. R. Cannan and
E. M. Flanigen, J. Am. Chem. Soc., 1982, 104, 1146–1147.
2 W. Yan, J. Yu, R. Xu, G. Zhu, F. Xiao, Y. Han, K. Sugiyama and
O. Terasaki, Chem. Mater., 2000, 12, 2517–2519.
3 G. Sastre, D. W. Lewis, C. Richard and A. Catlow, J. Phys.
Chem. B, 1997, 101, 5249–5262.
4 J. Yu and R. Xu, Acc. Chem. Res., 2003, 36, 481–490.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 4522–4529 | 4527

View Article Online
Journal of Materials Chemistry A
Paper
5 Y. Hirota, K. Watanabe, Y. Uchida, Y. Egashira, K. Yoshida, 33 M. H. Kim, H. X. Li and M. E. Davis, Microporous Mater.,
Y. Sasaki and N. Nishiyama, J. Membr. Sci., 2012, 415–416,
176–180.
6 A. M. Prakash and S. Unnikrishnan, J. Chem. Soc., Faraday
Trans., 1994, 90, 2291–2296.
7 M. Zamadics, X. Chen and L. Kevan, J. Phys. Chem., 1992, 96,
5488–5491.
8 S. Li, J. L. Falconer and R. D. Noble, J. Membr. Sci., 2004, 241,
121–135.
9 S. Li, J. L. Falconer and R. D. Noble, Adv. Mater., 2006, 18,
2601–2603.
10 M. Hong, S. Li, H. F. Funke, J. L. Falconer and R. D. Noble,
Microporous Mesoporous Mater., 2007, 106, 140–146.
11 S. Li, J. L. Falconer and R. D. Noble, Microporous Mesoporous
Mater., 2008, 110, 310–317.
12 M. A. Carreon, S. Li, J. L. Falconer and R. D. Noble, J. Am.
Chem. Soc., 2008, 130, 5412–5413.
13 S. Li, M. A. Carreon, Y. Zhang, H. H. Funke, R. D. Noble and
J. L. Falconer, J. Membr. Sci., 2010, 352, 7–13.
14 G. Li, J. Yang, J. Wang, W. Xiao, L. Zhou, Y. Zhang, J. Lu and
D. Yin, J. Membr. Sci., 2011, 374, 83–92.
1993, 1, 191–200.
34 Y. Wang, Y. Tang, A. Dong, X. Wang, N. Ren and Z. Gao, J.
Mater. Chem., 2002, 12, 1812–1818.
35 J. Yao, H. Wang, S. P. Ringer, K.-Y. Chan, L. Zhang and N. Xu,
Microporous Mesoporous Mater., 2005, 85, 267–272.
36 A. MacIntosh and Y. Huang, Microporous Mesoporous Mater.,
2013, 182, 40–49.
37 B. Chen, C. W. Kirby and Y. Huang, J. Phys. Chem. C, 2009,
113, 15868–15876.
38 M. Matsukata, N. Nishiyama and K. Ueyama, Microporous
Mater., 1996, 7, 109–117.
39 L. Zhang, J. Yao, C. Zeng and N. Xu, Chem. Commun., 2003,
2232–2233.
40 L. Zhang, J. Bates, D. Chen, H. Y. Nie and Y. Huang, J. Phys.
Chem. C, 2011, 115, 22309–22319.
41 M. Matsukata, M. Ogura, T. Osaki, P. Raja, H. P. Rao,
M. Nomura and E. Kikuchi, Top. Catal., 1999, 9, 77–92.
42 M. H. Kim, H. X. Li and M. E. Davis, Microporous Mater.,
1993, 1, 191–200.
43 Y. Huang, R. Richer and C. W. Kirby, J. Phys. Chem. B, 2003,
107, 1326–1337.
15 I. M. Dahl and S. Kolboe, Catal. Lett., 1993, 20, 329–336.
16 B. Vora, J. Q. Chen, A. Bozzano, B. Glover and P. Barger, 44 T. Gullion and J. Schaefer, J. Magn. Reson., 1989, 81, 196–200.
Catal. Today, 2009, 141, 77–83.
45 J. Amoureux, C. Fernandez and S. Steuernagel, J. Magn.
Reson., Ser. A, 1996, 123, 116–118.
¨
17 J. Janchen, D. Ackermann, E. Weiler, H. Stach and
¨
W. Brosicke, Thermochim. Acta, 2005, 434, 37–41.
46 G. Liu, P. Tian and Z. Liu, Chin. J. Catal., 2012, 33, 174–182.
18 Q. Sun, N. Wang, D. Xi, M. Yang and J. Yu, Chem. Commun., 47 Y. Xiao, R. J. Kirkpatrick and Y. J. Kim, Phys. Chem. Miner.,
2014, 50, 6502–6505. 1995, 22, 30–40.
19 B. P. C. Hereijgers, F. Bleken, M. H. Nilsen, S. Svelle, 48 Y. Xiao and R. J. Kirkpatrick, J. Mater. Res., 1995, 10, 2586–
K.-P. Lillerud, M. Bjorgen, B. M. Weckhuysen and
U. Olsbye, J. Catal., 2009, 264, 77–87.
2591.
49 B. Chen and Y. Huang, Microporous Mesoporous Mater., 2011,
20 G. Yang, Y. Wei, S. Xu, J. Chen, J. Li, Z. Liu, J. Yu and R. Xu, J.
Phys. Chem. C, 2013, 117, 8214–8222.
21 K.-H. Schnabel, R. Fricke, I. Girnus, E. Jahn, E. Loffler,
143, 14–21.
50 Y. Huang, B. A. Demko and C. W. Kirby, Chem. Mater., 2003,
15, 2437–2444.
B. Parlitz and C. Peuker, J.Chem. Soc., Faraday Trans., 1991, 51 Y. Huang, C. W. Kirby and R. Richer, Stud. Surf. Sci. Catal.,
87, 3569–3574. 2004, 154B, 1471–1477.
22 M. Yang, P. Tian, C. Wang, Y. Yuan, Y. Yang, S. Xu, Y. He and 52 Y. Huang, Z. Yan and R. Richer, Chem. Mater., 2005, 17,
Z. Liu, Chem. Commun., 2014, 50, 1845–1847. 6545–6554.
23 S. M. Hendrik van Heyden and T. Bein, Chem. Mater., 2008, 53 B. Chen and Y. Huang, J. Am. Chem. Soc., 2006, 128, 6437–
20, 2956–2963.
6446.
24 M. M. Mertens and B. Engels, US Pat., 6 685 905, 2004.
25 H. Zhou, Y. Wang, F. Wei, D. Wang and Z. Wang, Appl.
Catal., A, 2008, 341, 112–118.
26 G. Liu, P. Tian, Y. Zhang, J. Li, L. Xu, S. Meng and Z. Liu,
Microporous Mesoporous Mater., 2008, 114, 416–423.
27 E. Dumitriu, A. Azzouz, V. Hulea and H. Kessler, Microporous
Mater., 1997, 10, 1–12.
54 S. Qiu, W. Pang, H. Kessler and J. L. Guth, Zeolites, 1989, 9,
440–444.
55 A. M. Prakash and S. Unnikrirhnan, J. Chem. Soc., Faraday
Trans., 1994, 90, 2291–2296.
56 Z. Yan, J. Zhuang, X. Lei, X. Han, X. Liu and X. Bao, Chin.
Chem. Lett., 2003, 14, 87–90.
57 B. Zhang, J. Xu, F. Fan, Q. Guo, X. Tong, W. Yan, J. Yu,
F. Deng, C. Li and R. Xu, Microporous Mesoporous Mater.,
2012, 147, 212–221.
28 Ø. B. Vistad, D. E. Akporiaye and K. P. Lillerud, J. Phys. Chem.
B, 2001, 105, 12437–12447.
29 Ø. B. Vistad, E. W. Hansen, D. E. Akporiaye and 58 D. Zhou, J. Xu, J. Yu, L. Chen, F. Deng and R. Xu, J. Phys.
K. P. Lillerud, J. Phys. Chem. A, 1999, 103, 2540–2552. Chem. B, 2006, 110, 2131–2137.
30 Ø. B. Vistad, D. E. Akporiaye, F. Taulelle and K. P. Lillerud, 59 A. M. Prakash, S. Unnikrishnan and K. V. Rao, Appl. Catal., A,
Chem. Mater., 2003, 15, 1639–1649. 1994, 110, 1–10.
31 Z. Yan, B. Chen and Y. Huang, Solid State Nucl. Magn. Reson., 60 B. Chen and Y. Huang, J. Phys. Chem. C, 2007, 111, 15236–
2009, 35, 49–60. 15243.
32 W. Xu, J. Dong, J. Li, J. Li and F. Wu, J. Chem. Soc., Chem. 61 S. Ashtekar, S. V. V. Chilukuri and D. K. Chakrabarty, J. Phys.
Commun., 1990, 755–756.
Chem., 1994, 98, 4878–4883.
4528 | J. Mater. Chem. A, 2015, 3, 4522–4529
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
62 C. Doremieux-Morin, C. Martin, J. M. Bregeault and 67 R. Vomscheid, M. Briend, M. J. Peltre, P. P. Man and
J. Fraissard, Appl. Catal., 1991, 77, 149–161. D. Barthomeuf, J. Phys. Chem., 1994, 98, 9614–9618.
63 J. Tan, Z. Liu, X. Bao, X. Liu, X. Han, C. He and R. Zhai, 68 D. Barthomeuf, J. Phys. Chem., 1993, 97, 10092–10096.
´
´
´
˜
Microporous Mesoporous Mater., 2002, 53, 97–108.
64 G. Liu, P. Tian, J. Li, D. Zhang, F. Zhou and Z. Liu,
Microporous Mesoporous Mater., 2008, 111, 143–149.
65 L. Xu, A. Du, Y. Wei, Y. Wang, Z. Yu, Y. He, X. Zhang and
Z. Liu, Microporous Mesoporous Mater., 2008, 115, 332–337.
66 E. M. Flanigen, R. L. Patton and S. T. Wilson, Stud. Surf. Sci.
Catal., 1988, 37, 13–27.
69 T. Alvaro-Munoz, C. Marquez-Alvarez and E. Sastre, Catal.
Today, 2012, 179, 27–34.
70 G. Sastre and D. W. Lewis, J. Chem. Soc., Faraday Trans.,
1998, 94, 3049–3058.
71 Y. Wei, D. Zhang, Y. He, L. Xu, Y. Yang, B. Su and Z. Liu,
Catal. Lett., 2007, 114, 30–35.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 4522–4529 | 4529