Verifying the dominant catalytic cycle of the methanol-to-hydrocarbon conversion over SAPO-41

Article abstract of DOI:10.1039/c3cy00740e

In the present work, the mechanism of the methanol-to-hydrocarbon (MTH) conversion over the silicoaluminophosphate SAPO-41 with one-dimensional 10-ring pores has been investigated. For this purpose, catalytic experiments with co-feeding of reactants, in situ FTIR and UV/Vis spectroscopy, and ¹H MAS NMR spectroscopy of used catalysts upon loading of ammonia were performed. Using these methods, only a low content of aromatics could be observed on the working SAPO-41 catalyst. These findings and the characteristic changes in the ethene selectivity during the co-feeding experiments indicate that alkylaromatics can be excluded as active hydrocarbon pool compounds on SAPO-41 applied as MTH catalysts. On the other hand, dienes and enylic carbenium ions were detected using FTIR and UV/Vis spectroscopy and the formation of amines by reaction of alkenes with ammonia was detected via¹H MAS NMR spectroscopy. These results are experimental evidence that large olefins partially existing in their carbenium state dominate the catalytically active hydrocarbon pool on the working SAPO-41. Due to this dominant alkene-based reaction mechanism and the limited formation of aromatics, SAPO-41 is a suitable catalyst for the C₃+C₄ olefin as well as gasoline production.

Full text of DOI:10.1039/c3cy00740e

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Verifying the dominant catalytic cycle of the
methanol-to-hydrocarbon conversion
over SAPO-41
Cite this: Catal. Sci. Technol., 2014,
4, 688
a
a
a
a
a
Xin Wang, Weili Dai, Guangjun Wu, Landong Li,* Naijia Guan
and Michael Hunger*b
In the present work, the mechanism of the methanol-to-hydrocarbon (MTH) conversion over the
silicoaluminophosphate SAPO-41 with one-dimensional 10-ring pores has been investigated. For this
purpose, catalytic experiments with co-feeding of reactants, in situ FTIR and UV/Vis spectroscopy, and
1
H MAS NMR spectroscopy of used catalysts upon loading of ammonia were performed. Using these
methods, only a low content of aromatics could be observed on the working SAPO-41 catalyst. These
findings and the characteristic changes in the ethene selectivity during the co-feeding experiments
indicate that alkylaromatics can be excluded as active hydrocarbon pool compounds on SAPO-41
applied as MTH catalysts. On the other hand, dienes and enylic carbenium ions were detected using
FTIR and UV/Vis spectroscopy and the formation of amines by reaction of alkenes with ammonia was
Received 29th September 2013,
Accepted 21st November 2013
1
detected via H MAS NMR spectroscopy. These results are experimental evidence that large olefins partially
existing in their carbenium state dominate the catalytically active hydrocarbon pool on the working SAPO-41.
Due to this dominant alkene-based reaction mechanism and the limited formation of aromatics, SAPO-41
DOI: 10.1039/c3cy00740e
www.rsc.org/catalysis
3 4
is a suitable catalyst for the C +C olefin as well as gasoline production.
2
,4
the topics of numerous controversy debates. The generally
discussed routes are based on the formation of a hydrocarbon
pool occluded in the cages or pores of the catalysts, to which
methanol and dimethyl ether (DME) are added and from
1
. Introduction
Since the world oil reserves are declining, the utilization of
alternative feedstocks for the production of petrochemicals is
attracting more and more interest. Methanol can be easily
produced from synthesis gas obtained by gasification of coal
or other sustainable feedstocks. One of the most important
routes for methanol conversion is the methanol-to-hydrocarbon
1
4–16
which alkenes are split in closed cycles.
aromatic-based hydrocarbon pool mechanism is widely
For SAPO-34, an
1
7–19
accepted for the MTH conversion.
In the case of H-ZSM-5,
a similar mechanism involving methylation and subsequent
cracking of higher alkenes is proposed in parallel with the
aromatic-based hydrocarbon pool mechanism, i.e. the dual-
(MTH) reaction on microporous catalysts with high or moderate
acid strength. This is an oil-free process to obtain hydrocarbons
which has drawn significant attention since it was disclosed
in the 1970s.
7
,20–22
cycle mechanism shown in Scheme 1.
This dual-cycle
1
–4
A variety of microporous catalysts have been
mechanism has significantly improved the general under-
standing of the catalytic properties of microporous catalysts
with different pore architectures and of specific catalysts as a
function of time-on-stream (TOS).
5
–13
explored for MTH conversion.
Amongst these catalysts, the
silicoaluminophosphate SAPO-34 and aluminosilicate-type
H-ZSM-5 zeolites have already been commercialized for the
methanol-to-olefin (MTO) and methanol-to-gasoline (MTG)
conversion, respectively.
The MTH conversion is known as an extremely complex
reaction and the understanding of the mechanisms is a great
challenge. In the past decades, the origin of the first C–C bond
and the detailed reaction mechanism of MTH conversion were
a
Key Laboratory of Advanced Energy Materials Chemistry (Ministry of Education),
College of Chemistry, Nankai University, Tianjin 300071, PR China.
E-mail: lild@nankai.edu.cn; Fax: +86 22 23500341; Tel: +86 22 23500341
b
Institute of Chemical Technology, University of Stuttgart, 70550 Stuttgart,
Scheme 1 Concept of the dual-cycle mechanism for the conversion
Germany. E-mail: michael.hunger@itc.uni-stuttgart.de; Fax: +49-711-685-64081
of methanol-to-hydrocarbons over H-ZSM-5, adapted from ref. 4.
688 | Catal. Sci. Technol., 2014, 4, 688–696
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Accordingly, the topology of the microporous catalysts
significantly influences the reaction intermediates and the
reaction mechanism. In SAPO-34 and H-Beta containing
cages and crossing intersections (crossing 12-ring pores with
diameters of 0.66 nm × 0.67 nm and 0.56 nm × 0.56 nm),
respectively, hexamethylbenzene is formed and its further
chemical composition of the synthetic mixture of SAPO-41 is
1.2Al O :H PO :2H PO : 4DPA : 0.15SiO :50H O. The synthesis
2
3
3
3
3
4
2
2
mixture was hydrothermally treated in a Teflon-lined stainless-
steel autoclave at 200 °C for 4 days. After crystallization, the
solid phase was separated by centrifugation, washed several
times with deionized water and subsequently dried at 80 °C
for 12 h. The as-synthesized sample was calcined in flowing
synthetic air at 600 °C for 6 h.
2
3
methylated cation, i.e. the heptamethylbenzenium cation
+
(
heptaMB ) is found to be the dominating intermediate in
7
,17
the formation of light alkenes.
For H-ZSM-5 possessing
crossing intersections of 10-ring pores with diameters of
2
3
2.2 Sample characterization
0
.56 nm × 0.53 nm and 0.55 nm × 0.51 nm, lighter poly-
methylbenzenes (such as xylene, triMB, and tetraMB) are proven
to be the main aromatic-based hydrocarbon pool species. On
the other hand, for ZSM-22 characterized by one-dimensional
The X-ray diffraction (XRD) pattern of the sample was
7
recorded on a Bruker D8 diffractometer using CuKα radiation
−
1
(λ = 1.5418 Å) from 5–50° with a scan speed of 2θ = 6.0° min .
A HITACHI S-4700 Scanning Electron Microscope (SEM)
was used to study the morphology of the SAPO-41 particles.
The chemical composition of the calcined sample was
determined using inductively coupled plasma atomic emission
spectroscopy (ICP-AES, IRIS Advantage). The surface area of
the calcined sample was determined by means of nitrogen
adsorption at −196 °C on a Quantachrome iQ-MP gas adsorp-
tion analyzer. Before the nitrogen adsorption, the sample was
dehydrated at 200 °C for 2 h. The total surface area was calcu-
lated using the Brunauer–Emmett–Teller (BET) equation.
The temperature-programmed desorption of ammonia
1
0-ring pores having a diameter of 0.57 nm × 0.46 nm, the
aromatic-based hydrocarbon pool mechanism is inhibited,
olefin chains grow via step by step carbon addition, and the
2
4
product mixture is rich in higher alkenes. Hence, cages,
crossing intersections or large pores are necessary to accom-
modate polymethylbenzenes (polyMBs), and the size of the
pores influences the specific polyMB species, wherein narrow
pores support the alkene-based mechanism.
On the other hand, the acid strength of the catalyst also
greatly influences the reaction mechanism of the MTH con-
version. It has been reported that the primary dehydration
reaction and oligomerization of lower olefins requires moderate
(NH -TPD) was carried out in a quartz U-shaped reactor
3
25
acid sites, while the aromatization requires strong acid sites.
and monitored using an online chemisorption analyzer
(Quantachrome ChemBet 3000). The sample with a mass
of ca. 0.1 g was pretreated at 600 °C for 1 h in flowing He
For example, the dual-cycle mechanism is also present in
SAPO-5 characterized by large one-dimensional 12-ring pores
with a diameter of 0.73 nm × 0.73 nm, but the lower acid
strength compared to HZSM-5 shifts the reaction mechanism
−
1
(30 ml min ), cooled to 100 °C, and then saturated with
5% NH /He. After purging with He for 30 min to eliminate
3
2
6
to the alkene-mediated cycle.
the physically adsorbed ammonia, the NH -TPD measure-
3
−
1
In previous studies, the silicoaluminophosphate SAPO-41,
characterized by narrow one-dimensional 10-ring pores with
a diameter of 0.70 nm × 0.43 nm and a moderate Brønsted
ment was performed in flowing He (30 ml min ) at 100 to
−
1
600 °C with a heating rate of 10 °C min .
Solid-state NMR experiments were performed on a Bruker
Avance III spectrometer at resonance frequencies of 400.1,
1
1
acidity, has been proven to be a suitable MTH catalyst. The
present work focuses on the mechanism of the MTH conver-
sion over SAPO-41, and the obtained results are discussed in
1
27
29
31
104.3, 79.5, and 161.9 MHz for H, Al, Si, and P, respec-
tively. The experimental conditions are as follows: single
2
7
1
31
27
comparison with those of SAPO-34 studied earlier. Both
microporous silicoaluminophosphates, SAPO-41 and SAPO-34,
are characterized by Brønsted sites with moderate acid strength.
In contrast to SAPO-34 that contains chabazite cages, which
are large enough for the formation of polyalkylaromatics and
polycyclic aromatics, SAPO-41 has neither cages nor crossing
intersections. Therefore, SAPO-41 is an interesting model cata-
lyst for clarifying the effect of its pore architecture on the reac-
tion mechanism and the nature of the hydrocarbon pool
compounds responsible for the MTH conversion property of
this material.
pulse excitation of π/2 for H and P, π/6 for Al, and π/4 for
2
9
1
29
Si, with repetition times of 20 s for H and Si, 0.5 s for
Al, and 30 s for P MAS NMR spectroscopy. The H, Al,
and P MAS NMR spectra were recorded with a sample spin-
ning rate of 8 kHz, while the Si MAS NMR spectra were
obtained at 4 kHz. The Al, Si, and P MAS NMR measure-
ments were performed using hydrated samples. To avoid
damage of the silicoaluminophosphate framework by long
term hydrolysis, the sample material was exposed to an atmo-
2
7
31
1
27
3
1
2
9
2
7
29
31
3 2
sphere that was saturated with vapours of a Ca(NO ) solu-
tion at ambient temperature to be fully hydrated one day
1
before the NMR investigations. Before the H MAS NMR studies
of the fresh SAPO-41 sample (TOS = 0 min), the sample was
−
2
dehydrated at 450 °C at a pressure below 10 Pa for 12 h.
After dehydration, the sample was sealed and kept in glass
tubes before being filled into the MAS NMR rotor in a glove
box purged with dry nitrogen gas. The sample treatment
2
. Experimental
2
.1 Sample preparation
SAPO-41 was synthesized via the hydrothermal route
following the procedure described in literature.
2
8
1
The
utilized in the case of H MAS NMR investigations on used MTH
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 688–696 | 689

View Article Online
Paper
Catalysis Science & Technology
1
catalysts is described in section 2.5. Quantitative H MAS NMR
measurements were performed by a comparison of the signal
intensities of the samples under study with that of an external
standard (dehydrated zeolite H,Na-Y with an ion exchange
degree of 35%). The decomposition and simulation of NMR
spectra were carried out using the Bruker software WINNMR
and WINFIT.
of 10 s, and a 2.5 mm MAS NMR probe with the sample
spinning rate of 25.0 kHz were used. The MTH reaction
performed at 400 °C was quenched by stopping the methanol
flow and cooling down to room temperature under flowing
nitrogen before taking the catalyst samples. Subsequently, the
catalyst samples were transferred from the fixed-bed reactor
to MAS NMR rotors without contact with air in a glove box
purged with dry nitrogen gas. The determination of the number
of accessible Brønsted acid sites was performed by adsorption
of ammonia at room temperature. After the ammonia loading,
the samples were evacuated at 180 °C for 2 h to eliminate
2
.3 Catalytic investigations
The MTH reaction was performed in a fixed-bed reactor at atmo-
spheric pressure. Typically, 0.4 g of the sample (sieve fraction,
.25–0.5 mm) was placed in a stainless steel reactor (5 mm i.d.)
and activated under flowing N at 450 °C for 0.5 h, after which
the temperature was decreased to the desired temperature.
Pure methanol or methanol with a co-feeding agent was
1
physisorbed ammonia. Quantitative H MAS NMR measure-
ments and their evaluation were performed as described in
section 2.2.
0
2
−
1
pumped in at 0.5 mL h , corresponding to the weight hourly
−
1
3. Results and discussion
3.1 Results of the catalyst characterization
space velocity (WHSV) of 1 h . The products were analyzed
using an online gas chromatograph equipped with a flame
ionization detector and a capillary column DB-624 to separate
the products. The temperature of the column was maintained
at 40 °C for 7 min and then increased to 200 °C at a ramping
The XRD pattern of the as-synthesized SAPO-41 sample in Fig. 1
shows typical diffraction lines corresponding to the AFO struc-
28
ture, which indicate that pure SAPO-41 was obtained. As
−
1
rate of 10 °C min and maintained at 200 °C for 4 min.
shown in the SEM image (see inset) of the as-synthesized
SAPO-41 material, the particles appear as hexagonal prisms
with dimensions of ca. 2 μm diameter and ca. 6 μm height.
The n /(n + nAl + n ) of calcined SAPO-41 determined
2
.4 In situ UV/Vis and FTIR studies of the MTH conversion
on SAPO-41
Si
Si
P
The nature of organic compounds formed on the catalysts
during the MTH reaction was in situ monitored by FTIR and
UV/Vis spectroscopy. FTIR spectra were recorded in diffuse
reflection mode on a Bruker Tensor 27 spectrometer equipped
using ICP-AES is 0.03. The surface area of the calcined sample
was determined to be 268 m g , indicating the intactness
and accessibility of the pore system.
2
−1
The acidic properties of the activated and fresh SAPO-41
with a liquid N
2
cooled high sensitivity MCT detector and
3
material (TOS = 0 min) were characterized by NH -TPD (Fig. 2).
an in situ reaction chamber. The catalyst samples weighing
ca. 20 mg were finely ground and placed in the chamber. Prior
to each experiment, the samples were activated in flowing He
at 450 °C for 1 h and cooled to 400 °C to record the back-
ground spectrum. Then, methanol was fed into the chamber
The obtained curve consists of a low-temperature ammonia
desorption peak centered at 210 °C attributed to weak acid
sites and a medium-temperature ammonia desorption peak
centered at 320 °C assigned to moderate acid sites. According
2
9
to earlier studies of Meriaudeau et al., this NH -TPD curve is
3
−1
−1
at 0.025 mL h (WHSV = 1.0 h ) and the time-resolved spectra
typical for SAPO-41 and most of the other pure SAPO materials.
−
1
were recorded with a resolution of 4 cm and an accumula-
tion of 128 scans.
Diffuse reflectance UV/Vis spectra were recorded using an
AvaSpec-2048 fiber optic spectrometer with an AvaLight-DH-S
deuterium light source manufactured by Avantes and a glass
fiber reflection probe HPSUV1000A manufactured by Oxford
Electronics. Before the MTH reaction, the glass fiber reflection
probe was placed in the fixed-bed reactor on the top of the
catalyst with a gap of ca. 1.0 mm. Reference UV/Vis spectra of
catalysts were recorded at reaction temperature prior to starting
the methanol flow. The in situ UV/Vis spectra were recorded
in the range of 200–600 nm during the MTH reaction.
1
2
.5 Solid-state H NMR characterization of surface sites
The Brønsted acid sites and occluded organic compounds
formed on the used MTH catalysts were characterized by means
1
of H MAS NMR spectroscopy. For this purpose, a Bruker
Avance III 400WB spectrometer at the resonance frequency of
Fig. 1 XRD pattern and SEM image of the as-synthesized SAPO-41
4
00.1 MHz with π/2 single pulse excitation, the repetition time
material.
690 | Catal. Sci. Technol., 2014, 4, 688–696
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
spectrum of the calcined and rehydrated SAPO-41 material is
assigned to tetrahedrally coordinated framework aluminum
atoms, while the signal at −14 ppm is due to octahedrally coordi-
nated aluminum atoms formed through a reversible coordina-
tion of two water molecules to tetrahedrally coordinated framework
aluminum atoms or caused by extra-framework aluminum
3
0
31
species. The signal at −30 ppm in the P MAS NMR spectrum
is due to tetrahedrally coordinated phosphorus atoms in the
aluminophosphate regions of the SAPO framework. The narrow
30
2
9
Si MAS NMR signal at −90 ppm indicates incorporation of
silicon atoms on tetrahedral framework positions with four
neighboring aluminum atoms at T positions. The Si MAS
NMR spectrum is clear evidence that most of the silicon atoms
are incorporated into the SAPO-41 framework in a highly isolated
manner since no Si MAS NMR signals of silicon islands
ca. −110 ppm) occurred. This leads to the formation of a
3
0
29
2
9
(
3
Fig. 2 NH -TPD profile of the calcined SAPO-41 material.
corresponding number of bridging OH groups. The above-
mentioned results of the solid-state NMR characterization
indicate that the SAPO-41 material under study maintains its
Brønsted acid sites and good textural properties after calcination.
For a better understanding of the acidic and textural prop-
erties of SAPO-41 after calcination, multi-nuclear solid-state
NMR spectroscopy was employed to analyze the Brønsted
acid sites and the bonding states of silicon, phosphorus and
aluminum in the SAPO-41 framework (Fig. 3). The signals
3.2 Catalytic investigation of SAPO-41 applied in the
1
MTH conversion
occurring at 3.6 and ca. 4.7 ppm in the H MAS NMR spec-
trum of the calcined and dehydrated material, both accompa-
nied by spinning sidebands with similar relative intensities
Fig. 4 shows the time-on-stream (TOS) behavior of the
methanol conversion and product selectivity of the SAPO-41
catalyst at reaction temperatures of 350 to 450 °C. At 350 °C,
100% methanol conversion is maintained up to TOS = 4 h,
and then gradually decreased to 70% at TOS = 16 h. With the
(not shown), are due to bridging OH groups (Si(OH)Al) acting
as Brønsted acid sites. Specifically, signals at 3.6 ppm indicate
the presence of non-interacting Si(OH)Al groups in pores and
1
cages, while H MAS NMR signals at 4.7 ppm are assigned to
2 4
progress of the MTH reaction, the selectivity to C –C olefins
Si(OH)Al groups involved in interactions with neighboring
framework oxygen atoms, such as in small structural building
gradually decreased and dropped from the initial 50 to 25%
at TOS = 16 h. With increasing reaction temperature up to
450 °C, 100% methanol conversion could be kept for more
than 16 h, and the initial selectivity to C –C olefins was
3
0
1
units. The weak H MAS NMR signals at 1.6 and 2.6 ppm are
generally explained by hydroxyl groups at the outer crystal
surface and framework defect sites, i.e. by SiOH, AlOH and
2
4
increased to 75%. These findings and the formation of
methane can be explained by an increasing tendency toward
cracking of heavier hydrocarbons to lighter fragments at
3
1
27
POH groups. The Al MAS NMR signal at 38 ppm in the
3
2
high temperatures. In addition, the selectivity to aromatics
mainly polymethylbenzenes, e.g. toluene, xylene, and durene)
(
decreases obviously with increasing reaction temperature.
This may be additional support for an increased tendency
toward cracking of large organic compounds. On the other
hand, oligomers desorb from the catalyst before they can
3
3
form aromatics at high reaction temperatures. Both these
assumptions are in line with the longer lifetime of the
SAPO-41 catalyst at 450 °C observed in comparison with its
catalytic properties at 350 °C. Moreover, we could not exclude
the possibility that the longer lifetime is simply due to the
higher reaction rates at higher temperatures, as suggested
3
4
by Olsbye et al. from kinetic modelling. Interestingly, more
than 50% of the gasoline-range hydrocarbons (C –C alkanes
5
8
and alkenes) can be obtained over the SAPO-41 catalyst at
the relatively low reaction temperature of 400 °C. From
this point of view, SAPO-41 can be applied as an eligible cata-
lyst for the gasoline production via the environment-friendly
MTH process.
1
27
29
31
Fig. 3 H, Al, Si, and P MAS NMR spectra of the calcined SAPO-41
1
material recorded in the dehydrated state ( H) and upon rehydration
Al, Si, and P). Asterisks indicate spinning sidebands.
27
29
31
(
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 688–696 | 691

View Article Online
Paper
Catalysis Science & Technology
Fig. 5 In situ UV/Vis spectra recorded during the MTH reaction on
SAPO-41 at 400 °C.
3
5
aromatics, respectively. The bands at 370 and 440 nm may
hint at the formation of dienylic and trienylic carbenium
3
5
36
ions, respectively, or different polycyclic aromatics.
For SAPO-34 applied as MTO catalysts, a strong UV/Vis
band was observed in the spectral range of 350 to 450 nm,
2
7
which strongly increased with increasing TOS. For SAPO-34,
however, this band occurs at 400 nm and is significantly
narrower than those at 370 and 440 nm in the case of SAPO-41.
The band at 400 nm in the UV/Vis spectra of working SAPO-34
was assigned to polycyclic aromatics, which agrees well with
the formation of aromatic compounds in the chabazite cages
of this silicoaluminophosphate. Considering the different
spectral properties of the UV/Vis band at 400 nm for SAPO-34
and those of the bands at 370 and 440 nm for SAPO-41, the
latter bands are explained by the presence of polyenylic
carbenium ions.
In Fig. 6, the in situ FTIR spectra recorded during
the MTH conversion over SAPO-41 performed at 400 °C
and recorded up to TOS = 180 min are shown. The weak
bands of C–H stretching vibrations occurring in the range of
−
1
2
825–2955 cm are difficult to assign to specific organic
−
1
compounds, while the weak bands at 1345 and 1460 cm are
typical for skeletal CC vibrations of aromatics and bending
3
7
vibrations of the C–H bond. The intensities of these bands
that slightly increase with the progress of the MTH reaction,
however, are weak in comparison with those observed in
the same spectral ranges in in situ FTIR studies of SAPO-34
Fig. 4 Methanol conversion and product selectivity during the MTH reaction
on SAPO-41 at different reaction temperatures.
27
applied as MTO catalysts. The most intensive and characteristic
3
.3 In situ UV/Vis and FTIR studies of organic compounds
formed on SAPO-41 by the MTH conversion
The nature of organic compounds formed during the MTH
conversion on SAPO-41 was monitored using in situ UV/Vis
and FTIR spectroscopy. Fig. 5 shows the UV/Vis spectra of the
species formed on SAPO-41 under steady-state conditions at
4
2
00 °C with TOS up to 180 min. After TOS = 20 min, bands at
80, 370 and 440 nm occur, while for further increased TOS,
an additional band at 240 nm is formed with a slightly lower
intensity compared with the band at 280 nm. With the prog-
ress of the MTH reaction, the band at 370 nm is more and
more overlapped with the band at 440 nm. Generally, the
bands at 240 and 280 nm are assigned to dienes and
Fig. 6 In situ FTIR spectra recorded during the MTH reaction on
SAPO-41 at 400 °C.
692 | Catal. Sci. Technol., 2014, 4, 688–696
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
FTIR bands observed for working SAPO-41 are those appearing
(TOS = 30 to 240 min) indicate the formation of a few AlOH
groups probably formed by the reaction of water and frame-
work aluminum species. Typically, these signals are charac-
−
1
at 3015 and 1305 cm , which are due to the C–H stretching
3
8
31
and methyl bending vibrations of dienes. The intensities
of these bands strongly increase during the first minutes of
the MTH reaction and are stable during the further progress
of the MTH conversion indicating continuous formation of
dienes on SAPO-41. This finding agrees well with the observed
properties of the UV/Vis band of dienes at 240 nm (Fig. 5).
terized by weak spinning sidebands (not shown). Similarly,
the weak signals at 3.9 to 4.0 ppm are also due to surface
POH groups formed from the hydrolysis of –Al–O–P– bonds
in the SAPO-41 framework. The assignment of these POH
groups is supported by FTIR studies of SAPO-41 conducted
2
9
by Meriaudeau et al. These authors observed bands of POH
−
1
stretching vibrations at 3677 cm near the band of bridging
3
.4 Solid-state NMR studies of surface sites and organic
−1
OH groups at 3627 cm .
compounds on SAPO-41
For a better quantification of the number of accessible
bridging OH groups acting as Brønsted acid sites in SAPO-41
applied as MTH catalysts, ammonia was loaded on the sam-
ples leading to the formation of ammonium ions with a
In order to investigate the fate of Brønsted acid sites as a
function of the reaction time, samples of the SAPO-41 catalyst
were taken after TOS = 0 to 240 min. These samples were
1
40
investigated using H MAS NMR spectroscopy without and
chemical shift of 6.8 ppm (Fig. 7, right). Evaluation of the
after loading of ammonia (Fig. 7, left and right, respectively).
The H MAS NMR spectra of the SAPO-41 catalysts shown on
signal intensities at 6.8 ppm gave the densities of accessible
Brønsted acid site to be 0.36 to 0.19 mmol g for SAPO-41
1
−1
the left-hand side of Fig. 7 consist of the signals of bridging
hydroxyl groups Si(OH)Al at 3.6 and ca. 4.7 ppm already
discussed in section 3.1. With increasing TOS, additional
signals of organic compounds occur in the aliphatic region
at ca. 0 to 2 ppm and in the olefinic and aromatic region
at ca. 5 to 7 ppm. These spectral ranges are also typical for
non-branched and branched dienes, such as 2,4-hexadiene
catalysts applied in the MTH conversion at TOS = 0 to 240 min
(Table 1, column 2). For SAPO-34, it was observed that the
bridging OH groups are rapidly covered by aromatics and
other larger organic species causing a rapid catalyst deacti-
vation due to their blocking. In the case of SAPO-41, the
one-dimensional 10-membered ring channels permit the dif-
fusion also of larger alkanes and alkenes as well as aromatics,
so that the apparent consumption and blocking of Brønsted
acid sites in SAPO-41 are much slower than that found for
SAPO-34 used as MTH catalysts. Accordingly, SAPO-41 has
a much longer lifetime in the methanol conversion under
similar reaction conditions.
2
7
39
(
δ1H = 1.7 ppm and 5.5 to 5.9 ppm), 2,3-dimethyl-1,3-butadiene
1H = 1.9 ppm and 4.9 to 5.0 ppm) etc., which supports the
3
9
(δ
assignment of dienes in the in situ FTIR spectra shown in Fig. 6.
1
The weak signals appearing at 0.2 ppm in the H MAS
NMR spectra of the SAPO-41 materials used as MTO catalysts
Upon loading of the SAPO-41 samples used as MTH cata-
1
lysts with ammonia, additional H MAS NMR signals appeared
at 2.8 ppm (Fig. 7, right). The signals hint at the formation
of mono-substituted secondary amines by reaction of ammonia
with alkenes that are probably in their carbenium state. Possi-
ble examples are ethylamine (δ1H = 2.36 ppm), propylamine
(
δ1H = 2.59 ppm) or secondary NH groups in larger alkane
2
3
9
chains (δ1H = 2.50 ppm). It is interesting to note that
the amines formed on aromatics and alkylaromatics can be
excluded from contributing to the signals at 2.8 ppm due
to their significantly higher chemical shifts (e.g. δ1H = 3.49
to 3.40 ppm for mono- to pentamethylphenylamine). Also,
protonated NHx groups have much higher chemical shift
values than 2.8 ppm.
Table 1 Number of accessible Si(OH)Al groups, nSiOHAl, and amines, namine
,
formed by reaction of ammonia with reactive organic compounds on
calcined SAPO-41 (TOS = 0 min) and on samples used as MTH catalysts
at 400 °C for TOS = 30 to 240 min, determined using quantitative
1
H MAS NMR spectroscopy
−1
−1
TOS/min
0
n
SiOHAl (mmol g
)
n
amine (mmol g )
0.36
0.30
0.27
0.22
0.19
0
30
0.003
0.006
0.009
0.013
1
Fig. 7 H MAS NMR spectra of SAPO-41 samples used as MTH catalysts
60
120
240
at 400 °C and for TOS = 0 to 240 min, recorded before (left) and after
ammonia loading (right).
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 688–696 | 693

View Article Online
Paper
Catalysis Science & Technology
The observed amines must be formed already at low tem-
dominating MTH reaction mechanism for this silicoalumino-
phosphate. Also, co-feeding of ethanol leads to a strong increase
in the ethene selectivity from 6.4% without and 32.7%
with co-feeding (Table 2). This finding can be explained by
direct formation of ethene via dehydration of ethanol over
silicoaluminophosphates and also at weak Brønsted acid sites
perature (180 °C), which indicates the high reactivity of the
organic compounds reacting with the adsorbed ammonia.
Probably, these organic compounds are the enylic or polyenylic
carbenium ions responsible for the UV/Vis bands occurring at
3
70 and 440 nm in the in situ UV/Vis spectra shown in Fig. 5.
1
41
Evaluation of the H MAS NMR signals at 2.8 ppm gave the
number of amine molecules summarized in column 4 of Table 1.
near the outer surface of the SAPO-41 particles. Meanwhile,
there is a significant decrease in the selectivities to C –C
5
8
in the case of ethanol co-feeding. This finding indicates
that the ethene molecules, which are formed by dehydration
of ethanol, cannot be converted to higher olefins via the
methylation route, as supported by the high reaction barrier
of 0.79 eV from theoretical calculations. In this case, the
ethene formed via dehydration of ethanol can rapidly desorb
from the catalyst particles and be detected in the reac-
tion outlet.
3
.5 Co-feeding of aromatics and large alcohols during the
MTH conversion
4
2
Considering the aromatic-based cycle of the hydrocarbon pool
mechanism shown in Scheme 1, right-hand side, toluene or
xylene introduced as co-feeds may contribute to the MTH
reaction. In this case, they would be responsible for the for-
mation of small reaction products, such as ethene. On the
other hand, co-feeding of ethanol and longer-chain alcohols
may help to understand the role of olefins in the MTH con-
version since these alcohols are rapidly dehydrated to their
corresponding olefins. Therefore, co-feeding experiments are
interesting tools for clarifying or supporting reaction mecha-
nisms of the MTH conversion under study.
For co-feeding of isopropanol and 1-butanol, no obvious
increase in the corresponding olefin formation could be observed,
while a strong increase in C –C formation is observed. Since
5
8
the dehydration of isopropanol and 1-butanol readily takes
place in the presence of acid sites, this finding may hint
at the further conversion of dehydration product olefins.
It has been reported that the reaction barriers for the methyla-
4
2
Table 2 shows the results of the co-feeding experiments
during MTH conversion over SAPO-41 and the initial activities
at TOS = 0.2 h are used for subsequent discussion. Without
co-feeding, the SAPO-41 catalyst has a much lower selectivity
tion of propene and butene are 0.35 eV and 0.20 eV, respec-
tively, which are much lower than that of ethene (0.79 eV).
Under a reaction temperature of 400 °C, the methylation of
C
3
+ olefins readily takes place, which results in the increase
1
1,27
to ethene (6.4%) in comparison with SAPO-34 (25–30%).
in C –C formation.
5
8
Already this finding hints at the small contribution of the
aromatic-based cycle in the MTH conversion over SAPO-41
3.6 Mechanism of the MTH conversion over SAPO-41
(Scheme 1, right-hand side) since this catalytic cycle is respon-
sible for the ethene formation. Upon co-feeding of toluene
and xylene, on the other hand, the ethene selectivity is
strongly enhanced from 6.4% to about 30%, while the selectivity
toward propene and gasoline-range hydrocarbons (C –C )
decreases accordingly. These results indicate that (i) toluene
and xylene can act as the hydrocarbon pool species of the
MTH conversion over SAPO-41. Without co-feeding of aromatics,
however, (ii) a very small number of aromatics is formed in
the pores of SAPO-41 to make the aromatic-based cycle the
Investigations of the MTH conversion on numerous zeolite
catalysts performed in the past decades led to the dual-cycle
concept (Scheme 1) as the best explanation for the different
product selectivities obtained for zeolites with various pore
systems and in specific states of their catalyst lifetime and for
materials with strong or moderate Brønsted acid sites. The
aim of the present study on SAPO-41 was to clarify whether the
aromatic- as well as alkene-based catalytic cycles contribute
to the MTH conversion or a single catalytic cycle dominates.
Furthermore, the variety of possible hydrocarbons contributing
to the catalytically active hydrocarbon pool in SAPO-41 had
to be limited to e.g. a group of similar organic compounds.
A very interesting finding of the catalytic experiments with
co-feeding of methylaromatics was the strong increase in the
ethene selectivity indicating that methylaromatics in principle
can work as active hydrocarbon-pool compounds in SAPO-41.
In the case of the conversion of pure methanol over SAPO-41,
however, a very low number of aromatics is formed and exists
in the SAPO-41 pores during the MTH conversion, leading
to the observed low selectivity to ethene. This explanation is
supported by in situ UV/Vis and FTIR spectroscopy results,
hinting at a much lower content of aromatic deposits on
SAPO-41 in comparison with SAPO-34. Furthermore, while
5
8
Table 2 Product distribution in MTH conversion over SAPO-41 at 400 °C,
TOS = 0.2 or 1 h, with different co-feedings of reactants
Selectivity %
TOS Conc.



4
Co-feeding
None
(h)
(v/v%)
C
2
C
3
C
C
5
–C
8
Aromatics
0.2
1
0.2
1
0
6.4 37.2 17.8 34.9
4.0 28.6 16.5 42.9
32.7 22.6 14.6 24.1
31.6 19.5 13.2 26.1
2.5 28.4 19.0 44.4
1.4 22.4 18.0 46.9
2.2 25.1 19.7 47.0
1.4 21.9 18.6 51.7
28.8 22.3 16.8 26.5
26.9 22.0 15.4 29.2
31.2 21.3 15.9 25.2
29.6 19.8 15.2 28.6
1.1
1.7
4.5
7.8
4.1
7.2
5.4
5.7
4.9
5.5
5.6
6.4
Ethanol
16.7
16.7
16.7
2.0
Isopropanol 0.2
1
1-Butanol
0.2
1
0.2
1
0.2
1
Toluene
Xylene
1
H MAS NMR studies of SAPO-34 applied as MTH catalysts
gave evidence for the presence of benzene-based carbenium
2.0
694 | Catal. Sci. Technol., 2014, 4, 688–696
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
4
0
ions, these species could not be detected for the SAPO-41
catalysts treated under similar conditions. All these findings
indicate that the aromatic-based cycle is not the dominating
reaction mechanism of the MTH conversion over SAPO-41.
Already, the high selectivities to propene and butenes as
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (21303089), 111 Project (B12015), the
Ministry of Education of China (NCET-11-0251) and the
Collaborative Innovation Center of Chemical Science and
Engineering (Tianjin). Furthermore, M.H. wants to acknowledge
the financial support of Deutsche Forschungsgemeinschaft.
5 8
well as the high yield of gasoline (C –C ) in the MTH conver-
sion on SAPO-41 without co-feeding of other reactants hint at
a reaction mechanism based on large olefins as active hydro-
carbon pool compounds. Support for this explanation was
obtained from the detection of dienes and polyenyl carbenium
ions during the MTH conversion via in situ UV/Vis and FTIR
spectroscopy. Interestingly, the adsorption of ammonia on
SAPO-41 samples applied as MTH catalysts led to the forma-
tion of amines by the reaction of ammonia with alkenes that
are probably in their carbenium state, but not with aromatics
or with their benzenium ions. Also this finding is an indica-
tion that large olefins and not aromatics are the most active
organic compounds in SAPO-41.
References
1 C. D. Chang, Catal. Rev. Sci. Eng., 1983, 25, 1–118.
2 M. Stöker, Microporous Mesoporous Mater., 1999, 29, 3–48.
3 J. F. Haw, W. G. Song, D. M. Marcus and J. B. Nicholas,
Acc. Chem. Res., 2003, 36, 317–326.
4 U. Olsbye, S. Svelle, M. Bjørgen, P. Beato, T. V. W. Janssens,
F. Joensen, S. Bordiga and K. P. Lillerud, Angew. Chem., Int. Ed.,
2012, 51, 2–24.
An important advantage of SAPO-41 is the relative low con-
sumption rate of Brønsted acid sites in comparison with
SAPO-34 under identical reaction conditions. This property
of SAPO-41 is due to the absence of pore blocking by aromatic
deposits, improving the reactant diffusion and desorption
also for large reaction products, such as gasoline compounds
5 A. T. Aguayo, A. G. Gayubo, R. Vivanco, M. Olazar and J. Bilbao,
Appl. Catal., A, 2005, 283, 197–207.
6 A. G. Gayubo, A. T. Aguayo, A. Alonso, A. Atutxa and J. Bilbao,
Catal. Today, 2005, 106, 112–117.
7 S. Svelle, U. Olsbye, F. Joensen and M. Bjørgen, J. Phys.
Chem. C, 2007, 111, 17981–17984.
2
7
(
C –C ) and aromatics.
8 M. Castro, S. J. Warrender, P. A. Wright, D. C. Apperley,
Y. Belmabkhout, G. Pirngruber, H. K. Min, M. B. Park and
S. B. Hong, J. Phys. Chem. C, 2009, 113, 15731–15741.
5
8
4
. Conclusion
9
J. W. Park, S. J. Kim, M. Seo, S. Y. Kim, Y. Sugi and G. Seo,
By means of catalytic investigations of the MTH conversion over
SAPO-41 without and with co-feeding of aromatics and large
alcohols, in situ UV/Vis and FTIR spectroscopy, and H MAS NMR
spectroscopy of used catalyst samples upon loading of ammonia,
the dominant catalytic cycle of the methanol conversion was
investigated. Interestingly, a strong increase in the ethene
selectivity was observed during co-feeding of methylaromatics
Appl. Catal., A, 2008, 349, 76–85.
10 J. H. Lee, M. B. Park, J. K. Lee, H. K. Min, M. K. Song and
S. B. Hong, J. Am. Chem. Soc., 2010, 132, 12971–12982.
11 W. Dai, X. Wang, G. Wu, N. Guan, M. Hunger and L. Li,
ACS Catal., 2011, 1, 292–299.
12 M. A. Djieugoue, A. M. Prakash and L. Kevan, J. Phys. Chem. B,
2000, 104, 6452–6461.
1
(
toluene, xylene) indicating that ethene formation according
13 S. Teketel, W. Skistad, S. Benard, U. Olsbye, K. P. Lillerud,
P. Beato and S. Svelle, ACS Catal., 2012, 2, 26–37.
14 I. M. Dahl and S. Kolboe, Catal. Lett., 1993, 20, 329–336.
15 I. M. Dahl and S. Kolboe, J. Catal., 1994, 149, 458–464.
16 I. M. Dahl and S. Kolboe, J. Catal., 1996, 161, 304–309.
17 B. Arstad and S. Kolboe, J. Am. Chem. Soc., 2001, 123, 8137–8138.
18 B. Arstad and S. Kolboe, Catal. Lett., 2001, 71, 209–212.
19 W. G. Song, J. F. Haw, J. B. Nicholas and C. S. Heneghan,
J. Am. Chem. Soc., 2000, 122, 10726–10727.
20 S. Svelle, F. Joensen, J. Nerlov, U. Olsbye, K. P. Lillerud,
S. Kolboe and M. Bjørgen, J. Am. Chem. Soc., 2006,
128, 14770–14771.
21 M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe,
F. Bonino, L. Palumbo, S. Bordiga and U. Olsbye, J. Catal.,
2007, 249, 195–207.
to the aromatic-based pathway (Scheme 1, right-hand side) is
in principle possible in the narrow 10-ring pores of SAPO-41.
However, the low ethene selectivity obtained for the conver-
sion of pure methanol as well as the bands of aromatic
deposits detected using UV/Vis and FTIR spectroscopy hint to
a low content of aromatics on SAPO-41 under steady state con-
ditions of the MTH conversion. All these findings led to an
exclusion of the aromatic-based cycle as the dominant mecha-
nism of the MTH conversion over SAPO-41.
On the other hand, in situ UV/Vis and FTIR spectroscopy
evidenced the presence of dienes and polyenylic carbenium
ions on SAPO-41 under reaction conditions. Furthermore,
1
H MAS NMR investigations of SAPO-41 samples applied as
MTH catalysts and loaded with ammonia indicated the
formation of amines with alkenes probably in their carbenium
state, but not with aromatics or benzenium ions. Also these
observations support the presence of large olefins, which are
the active hydrocarbon pool compounds of the MTH conver-
sion over SAPO-41 and make the alkene-based cycle (Scheme 1,
left-hand side) the dominating mechanism.
22 S. Llias, R. Khare, A. Malek and A. Bhan, J. Catal., 2013,
303, 135–140.
23 Ch. Baerlocher, L. B. McCusker and D. H. Olson, Atlas of
Zeolites Framework Types, Elsevier, Amsterdam, 6th edn, 2007.
24 Z. M. Cui, Q. Liu, Z. Ma, S. W. Bian and W. G. Song,
J. Catal., 2008, 258, 83–86.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 688–696 | 695

View Article Online
Paper
Catalysis Science & Technology
2
2
2
2
2
3
3
3
3
5 K. K. Ramasamy and Y. Wang, J. Energy Chem., 2013,
2, 65–71.
6 M. W. Erichsen, S. Svelle and U. Olsbye, J. Catal., 2013,
98, 94–101.
7 W. Dai, G. Wu, L. Li, N. Guan and M. Hunger, ACS Catal.,
013, 3, 588–596.
8 Y. Ma, N. Li, X. Ren, S. Xiang and N. Guan, J. Mol. Catal. A:
Chem., 2006, 250, 9–14.
9 P. Meriaudeau, V. A. Tuan, V. T. Nghiem, S. Y. Lai, L. N. Hung
and C. Naccache, J. Catal., 1997, 169, 55–66.
34 T. V. W. Janssens, S. Svelle and U. Olsbye, J. Catal., 2013,
308, 122–130.
35 Y. Jiang, J. Huang, V. R. R. Marthala, Y. Ooi, J. Weitkamp
and M. Hunger, Microporous Mesoporous Mater., 2007,
105, 132–139.
36 W. Wang, Y. J. Jiang and M. Hunger, Catal. Today, 2006,
113, 102–114.
37 J. W. Park and G. Seo, Appl. Catal., A, 2009, 356, 180–188.
38 H. Binder, A. Anikin and B. Kohlstrunk, J. Phys. Chem. B,
1999, 103, 450–460.
39 H NMR Predictor, Product Version 9.08, Advanced Chemistry
Development Inc., 2006.
40 W. Dai, M. Scheibe, N. Guan, L. Li and M. Hunger,
ChemCatChem, 2011, 3, 1130–1133.
41 W. Dai, W. Kong, G. Wu, N. Li, L. Li and N. Guan, Catal.
Commun., 2011, 12, 535–538.
42 C. M. Wang, Y. D. Wang and Z. K. Xie, J. Catal., 2013,
301, 8–19.
2
2
2
0 A. Buchholz, W. Wang, M. Xu, A. Arnold and M. Hunger,
Microporous Mesoporous Mater., 2002, 56, 267–278.
1 Y. Jiang, J. Huang, W. Dai and M. Hunger, Solid State Nucl.
Magn. Reson., 2011, 39, 116–141.
2 N. Rahimi and R. Karimzadeh, Appl. Catal., A, 2011,
3
98, 1–17.
3 A. J. Marchi and G. F. Froment, Appl. Catal., A, 1991,
1, 139–152.
7
696 | Catal. Sci. Technol., 2014, 4, 688–696
This journal is © The Royal Society of Chemistry 2014