In situ DRIFT study of dimethyl ether carbonylation to methyl acetate on H-mordenite

Article abstract of DOI:10.1016/j.molcata.2016.02.032

In situ diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy was used to study the reaction of dimethyl ether (DME) carbonylation to methyl acetate (MAc). Methoxyl and acetyl groups were observed in situ, confirming the reaction mechanistic proposal. DME adsorption bands on Si-OH, Al-OH and SiOHAl (Br?nsted acid sites) groups were identified respectively. The effect of temperature and pressure on acetyl formation and side reactions were systematically studied. It was found that methoxyls formed by DME chemisorbed on Br?nsted acid sites could be observed above 433 K. High pressure and appropriate temperature interval promoted the formation of acetyls. Side reactions, DME transformed into olefins and then into cokes, were favored at low pressure and high temperature. Carbonylation reactions on catalysts with acid sites only in 8MR side pockets and 12MR main channels of mordenite (samples noted as HMOR-SP and HMOR-MC) were conducted by in situ DRIFT spectroscopy and catalytic testing. Pyridine molecules adsorbed in 12MR main channels resulted in the low utilization of acid sites in 8MR side pockets. The synthesis rate of MAc on HMOR-SP was stable but lower. HMOR-MC presented the lowest MAc synthesis rate with the production of coke, which was also observed in DRIFT spectra.

Full text of DOI:10.1016/j.molcata.2016.02.032

Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
In situ DRIFT study of dimethyl ether carbonylation to methyl acetate
on H-mordenite
Hui Zhoua,b,c, Wenliang Zhua,b, Lei Shia,b, Hongchao Liua,b, Shiping Liua,b,c
,
Youming Nia,b, Yong Liua,b, Yanli Hea,b, Shuliang Xua,b, Lina Lia,b,c, Zhongmin Liua,b,∗
a National Engineering Laboratory for Methanol to Olefins, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, P.O. Box 110, 116023 Dalian,
PR China
b Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, PR China
c Graduate University of Chinese Academy of Sciences, Beijing 100049, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 January 2016
Received in revised form 25 February 2016
Accepted 26 February 2016
Available online 28 February 2016
In situ diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy was used to study the reac-
tion of dimethyl ether (DME) carbonylation to methyl acetate (MAc). Methoxyl and acetyl groups were
observed in situ, confirming the reaction mechanistic proposal. DME adsorption bands on Si OH, Al OH
and SiOHAl (Brønsted acid sites) groups were identified respectively. The effect of temperature and pres-
sure on acetyl formation and side reactions were systematically studied. It was found that methoxyls
formed by DME chemisorbed on Brønsted acid sites could be observed above 433 K. High pressure and
appropriate temperature interval promoted the formation of acetyls. Side reactions, DME transformed
into olefins and then into cokes, were favored at low pressure and high temperature. Carbonylation
reactions on catalysts with acid sites only in 8MR side pockets and 12MR main channels of mordenite
(samples noted as HMOR-SP and HMOR-MC) were conducted by in situ DRIFT spectroscopy and catalytic
testing. Pyridine molecules adsorbed in 12MR main channels resulted in the low utilization of acid sites
in 8MR side pockets. The synthesis rate of MAc on HMOR-SP was stable but lower. HMOR-MC presented
the lowest MAc synthesis rate with the production of coke, which was also observed in DRIFT spectra.
© 2016 Elsevier B.V. All rights reserved.
Keywords:
In situ DRIFT
Dimethyl ether
Methyl acetate
Carbonylation
Mordenite
1. Introduction
porous zeolites have made the largest contribution to society so far,
and the field is still developing rapidly [7,8]. The synthesis of porous
Ethanol, one of the most important renewable fuels, can be used
as fuel additive, hydrogen carrier, chemical feedstock, and so on.
The production and transformation of ethanol have been triggering
wide research interest [1–4]. Traditional commercialized route by
hydration of ethylene to ethanol employs non-renewable oil as raw
material, whereas limited production capacity of ethanol by fer-
mentation of biomass materials cannot meet large-scale demands
in various fields [3]. The synthesis of ethanol by DME carbonylation
to MAc and MAc hydrogenation is a potential way from synthesis
gas (CO + H2, syngas) which can be derived readily from natural gas,
coal and biomass [5,6].
materials by templates have been well reviewed in literature [9].
Researches on DME carbonylation by heterogeneous catalysts have
been gaining much attention and great progress have been made
on the active sites [10–12], the stability of catalysts [13] and the
positive enhancement of metal atoms [14–16] since Fujimoto first
reported that HMOR was effective for the catalysis of DME to MAc
in 1984 [17].
The intermediates and elementary steps of DME carbonylation
reaction on zeolites have been proposed by Iglesia et al. [10,18,19].
They suggested that DME first adsorbed on the Brønsted acid sites of
zeolite to form methanol and a chemisorbed methyl group CH3* at
zeolite exchange sites, and then CO inserted the C-O between the
chemisorbed methyl group CH3* and the oxygen atom of zeolite
framework to produce CH3CO*. This crucial intermediate CH3CO*
reacted with another DME molecule to form MAc and the surface
methyl group was regenerated. Therefore, the direct observation of
CH3* and CH3CO* under real reaction conditions would be prereq-
uisite to confirm this mechanism proposal.
Porous materials are important in widespread processes such
as catalysis, molecular separations, energy and health, in which
∗
Corresponding author at: National Engineering Laboratory for Methanol to
Olefins, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, P.O. Box
110, 116023 Dalian, PR China.
E-mail address: liuzm@dicp.ac.cn (Z. Liu).
http://dx.doi.org/10.1016/j.molcata.2016.02.032
1381-1169/© 2016 Elsevier B.V. All rights reserved.

2
H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
Cheung et al. had tried to prove this mechanism by infrared spec-
troscopy. They observed the surface acetyl CH3CO* by dissociative
adsorption of acetic anhydride on HMOR zeolite at 438 K [19]. So,
the acetyl came from the reaction of material containing carbonyl
group and the O-H of zeolite, not from the reaction of DME and
CO under the real carbonylation reaction conditions. In addition as
described in article, acetic anhydride could only react with Brønsted
acid sites in twelve-member ring (12MR) main channels but not
access those in side-pockets of mordenite, which were considered
as the real active centers of DME carbonylation reaction [10,11].
Andrzej Malek reported the observation of acetyl group in order
to explore the source of the first C-C species in MTO reaction [20].
The systematic study about the mechanism of DME carbonylation
reaction, especially the observation and the variation of intermedi-
ates, methoxyl and acetyl, under real DME carbonylation reaction
conditions, would help us to understand the reaction in depth.
In order to relate the reaction products and the adsorbed species,
in situ information is essential. In this sense, in situ DRIFT spec-
troscopy can be regarded as an ideal technique. DRIFT spectroscopy
is an increasingly popular technique in catalysis research since
Ko˝ rtu˝ um and Griffiths proved its fundamental principle theoreti-
cally in 1970s. This technique has been widely used to characterize
the catalyst and surface species prior to and under reaction con-
ditions and thus can provide vital information about reactions.
Moreover, it permits in-situ observation of the reactor bed in pow-
dered or granular form without any catalyst manipulation, not in
wafer with the negative effect of the diffusion of reactants and
products. Therefore, DRIFT spectroscopy is an invaluable method
for studying powders, rough-surfaced solids and the reactions
on powders, which is close to the actual state of the true reac-
tions [21,22], and this method has been successfully used to study
the mechanism of many reactions, such as the low-temperature
methanol synthesis mechanism using ethanol promoter [23], the
selective catalytic reduction of NOx with NH3 [24–26], the ethanol
steam-reforming reaction [27], water-gas shift reaction [28] and CO
oxidation at room temperature on Au/SiO2 [29]. In this study, we
used in situ DRIFT spectroscopy to study DME carbonylation reac-
tion and observed the intermediates of methoxyl and acetyl groups
on HMOR by DME reacting with CO under the real reaction condi-
tions. Systematic study was carried out about DME adsorption on
the HMOR, the effect of temperature and pressure on the acetyl
formation and side reactions and the distinct effect of acid sites at
different channels of HMOR on DME carbonylation reaction.
Fig. 1. In situ DRIFT spectrum of DME on ␥-Al2O3 at 403 K.
MCT detector. The catalyst powder was contained in the diffuse
reflectance infrared cell with ZnSe window which could work at
high temperatures and high pressures. In situ absorbance spectra
were obtained by collecting 16 scans at 4 cm−1 resolution.
DME used in experiments was 3% in volume and N2 was the
equilibrium gas. The degree of purity of CO and purging N2 was
99.999%, respectively.
Before the adsorption of reactants, the sample was first cal-
cined in situ in N2 stream flowing 30 cm3 min−1 at 723 K (523 K
for HMOR-SP) for 120 min. Subsequently the catalyst was cooled
down to adsorption temperature in N2 stream, and the spectrum
of catalyst was recorded as reference.
The spectra could be obtained continuously during the adsorp-
tion and reaction processes thus we could study the variation
of surface intermediates over time. The spectra were obtained
according to the following procedure after the catalyst pretreat-
ment: the activated catalyst was exposed to DME at a total rate
of 10 cm3 min−1 at reaction temperature for 5 min, followed by
sweeping in N2 for 30 min to yield the adsorbed DME species on
catalyst surface. As gaseous CO did not have effect on the obser-
vation and intensity of acetyl bands, so the spectra of acetyl were
obtained continuously as CO passed through the in situ cell at dif-
ferent temperatures and pressures. Then DME was carried into the
cell again. Gaseous products were observed by online mass spec-
trometer (Omnistar).
2. Experiment
2.1. Catalysts preparation
3. Results and discussion
The materials, synthesis processes and characterization results
of HMOR were presented elsewhere [15]. The sample marked as
HMOR-MC (with acid sites only in 12MR main channels and the
counter ions in 8MR side pockets were Na+) was obtained by the ion
3.1. Adsorption of DME on aluminum hydroxyls, silanols and
Brønsted acid sites
In situ DRIFT spectra of DME adsorption on aluminum hydroxyl
(AlOH), silanol (SiOH) and HMOR are shown in Figs. 1–4 . The
adsorptions of DME on AlOH and SiOH were conducted on ␥-
Al2O3 and SiO2, mainly with the AlOH and SiOH at 3660 and
3740 cm−1, respectively. There are four bands, 3033, 2960, 2884
and 2845 cm−1, in the spectrum of DME adsorption on AlOH in
Fig. 1. Assignment of the bands was made by analogy with the
spectrum of known compounds and by comparison with pub-
lished literatures. The bands at 3033 and 2960 cm−1 were the
two asymmetric stretching modes of CH3, respectively. The band
at 2845 cm−1 was the symmetric stretching mode of CH3. The
2884 cm−1 was assigned to the overtone of the deformation mode
of CH3. Those bands indicated the formation of methoxyl groups of
DME adsorption on AlOH at 403 K [31].
exchange of NaMOR with NH4NO3 aqueous solution (0.1 mol L−1
,
1 g of the solid in 10 mL of the solution) at room temperature for
0.5 h. After filtered and washed with deionized water, the solid was
dried and calcined in flowing dry air. Na content left in HMOR-MC
was 2.5 wt% analyzed by XRF, about 53% of the total Na content
in NaMOR 4.7 wt%. HMOR-SP (with acid sites in 8MR side pockets
only) was prepared by adsorption of pyridine molecules to neutral-
ize the acid sites in 12MR main channels of HMOR [13,30].
2.2. In situ DRIFT spectroscopy study
In situ DRIFT spectra were collected on a Bruke Tensor 27 instru-
ment supplied with a diffuse reflectance attachment and with a

H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
3
The spectrum of DME adsorption on SiOH of SiO2 at 403 K is
shown in Fig. 2. The four bands in the spectrum are at 2995, 2946,
2899 and 2834 cm−1. DME adsorption on HZSM-5 (Si/Al = 121) with
mainly SiOH at 423 and 473 K presented four bands at 2993, 2949,
2897 and 2836 cm−1, assigned to the molecularly adsorbed DME,
which interacted with silanol groups in zeolite through hydrogen
bonding [32]. Therefore, the four bands in Fig. 2 were assigned to
bands of molecularly adsorbed DME to the SiOH groups by hydro-
gen bonding.
Fig. 3 shows the DRIFT spectra of DME adsorption on HMOR
at different temperatures. The saturated adsorption of DME on
HMOR was conducted at 313 K first and then the equilibrium tem-
peratures increased to certain point progressively and held for
20 min in N2 stream. The spectra were obtained by deducting the
zeolite skeleton as reference. A series of broad bands could be
observed at around 3000, 2400 and 1700 cm−1 below 433 K, los-
ing intensity gradually with the increase of temperature. These
three broad bands may come from the H-bonding complex and be
attributed to the perturbed OH vibration, because they were quite
similar to the A-B-C triplet bands of the hydroxyl groups reported
previously [33,34]. Usually, for the H-bonding complex, the A-
B-C pattern with broad and intense absorbance at around 2800,
2400, and 1700 cm−1 was caused by a Fermi resonance between
out-of-plane deformation vibration of bridged OH group and the
perturbed hydrogen-bound OH group due to O-H· · ·X-H formation.
The appearance of A-B-C triplet indicated that the DME inter-
acted with the Brønsted acid sites by hydrogen bonds as a neural
molecule [32]. The disappearance of A-B-C triplet with the temper-
ature increase implied that the hydrogen bonded DME started to
decompose.
Fig. 2. In situ DRIFT spectrum of DME on SiO2 at 403 K.
In situ DRIFT spectra in O-H and C-H regions of DME adsorption
on HMOR at different temperatures are presented in Fig. 4(a) and
(b) in detail. The adsorption of DME induced the negative bands at
3745, 3730, 3660 and 3610 cm−1 at 313 K. The presence of negative
bands in O-H region signified that DME interacted with these kinds
of hydroxyls. The band at 3610 cm−1 was attributed to the Brønsted
acid sites, combination of Brønsted acid sites in 12MR and 8MR
channels of mordenite [10]. The band at 3660 cm−1 was associated
with the AlOH formed by the extra-framework Al species dropped
from the tetrahedral sites in mordenite framework. The bands at
3730 and 3745 cm−1 could be assigned to the internal and terminal
or external SiOH groups. All the O-H peak positions of these bands
shifted towards low wave number slightly with the increase of tem-
perature. The intensity of negative O-H, especially those of Si-OH
Fig. 3. In situ DRIFT spectra of DME on HMOR from 313 K to 473 K.
Fig. 4. In situ DRIFT spectra in O-H and C-H regions of DME adsorption on HMOR at different temperatures (a: O-H region, b: C-H region).

4
H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
Table 1
infrared absorption is related to the concentration of functional
groups and the chemical environment i.e. the attached chemical
groups. Therefore caution is recommended when spectra are com-
pared quantitatively. In this reaction, the intensity and shape of
CH3 were also influenced by other side reactions, such as the trans-
formation of methoxyl with the products of the high-polymerized
alkenes, (C C)n or coke species that were signed by the bands at
∼1600 and 1490 cm−1 [36]. The above mentioned phenomena were
in good accordance with the result that the bands in C H region also
became broader than those of methoxyls due to the formation of
a small amount of aliphatic hydrocarbon pool accompanied with
the reaction of methoxyls into olefins [20]. More importantly, the
acetyl was observed under the real carbonyaltion reaction condi-
tions by the reaction of CO and methoxyl on HMOR. The intensity of
acetyl band increased slowly in 1 h, after then it had little change,
implying the end of the reaction of CO with methoxyl to acetyl.
Then the gas phase carbon monoxide was flushed by N2 from the
cell. DME was introduced into the cell again to react with the acetyl.
The band of C O in acetyl changed almost instantaneously as DME
contacted with acetyl. The change of acetyl with the time is pre-
sented in Fig. 5(c). The bands at 1776 and 1758 cm−1 were ascribed
to the C O stretches in gaseous MAc [13,19], signifying that there
was MAc product after the acetyl reacted with DME, which was also
observed by the on line mass spectrometer (m/e = 74) in Fig. 6. The
intensity of band (m/e = 74) increased also instantaneously when
DME passed through the cell, and then started to diminish gradually
after the peak value.
Vibration wave numbers of adsorbed species on HMOR.
Species
Frequency (cm−1
)
DME hydrogen-bonded to SiOH
DME hydrogen-bonded to SiOHAl
AlOCH3
2995, 2946, 2899, 2834
3008, 2966, 2947, 2839
3033, 2960, 2884, 2845
2978, 2868
1704
1600,1490
SiOCH3Al
C
O in acetyl
(C C)n
Adsorbed water
1640
and Al-OH, decreased with the increase of temperature, inferring
that DME desorbed from acid sites and the interaction between
DME and Si-OH and Al-OH groups was weak.
We have attempted to assign the bands in C H region in spectra
of DME adsorption on HMOR above, which contained SiOH, AlOH
and SiOHAl, at different temperatures in Fig. 4(b). Because of the
relative intensity of bands and the overlapping one another, DME
molecules on silica and methoxy species on alumina presented in
Figs. 1 and 2 were part, but not all, observed in Fig. 4(b). Accord-
ing to the literature and DME adsorption on Brønsted acid sites,
silanol and aluminum hydroxyl investigated above, the bands at
2947, 2899 and 2832 cm−1 were assigned to the DME adsorption
on SiOH groups by hydrogen bonds at 313 K. As the temperature
increased to 403 K, the spectrum was different from that of 313 K.
The bands at 3008 and 2947 cm−1 lost intensity and a new shoul-
der band at 2966 cm−1 appeared, the band at 2832 cm−1 shifted to
2839 cm−1, the band at 2899 cm−1 almost disappeared, the bands
at 3008, 2966 and 2839 cm−1 were assigned to the DME interaction
with the Brønsted acid sites by hydrogen bond. The bands of DME
adsorption on AlOH would be underly by hydrogen bonded DME.
The adsorbed DME started to decompose as the temperature above
423 K. Only two bands at 2978 and 2868 cm−1 were observed at
473 K, which was the asymmetric and symmetric stretching modes
of CH3 [19], indicating the formation of methoxyl groups by reac-
tion of DME and Brønsted acid sites. The intensity of methoxyl
bands decreased with the increase of temperature. Assignment of
bands in Fig. 4(b) and in the following text is concluded in Table 1.
The C-H region in spectra after N2 flush for 30 min is shown in
Fig. 5(d), the methoxyl groups formed again.
The direct observation of acetyl under real reaction conditions
by CO reaction with methoxyl demonstrated the intermediate
effect of acetyl. It was noteworthy that the formation rate of acetyl,
1 h for the complete reaction, was slower than that of methoxyl,
which formed on acid sites in several minutes and also slower than
the following acetyl reaction with DME (shown in Fig. 5(c) and
Fig. 6).
The observation of all of the intermediates also proved that the
designed in situ DRIFT experimental process was practical and fea-
sible for the study of DME carbonylation.
3.2. DME carbonylation by combination of in situ DRIFT
spectroscopy and on line MS
3.3. Effect of reaction pressure on the formation of acetyl
DME carbonylation reaction was conducted in the in situ DRIFT
cell, and gaseous fluent was monitored by on line mass spectrom-
eter. The saturated adsorption of DME on HMOR was finished in
several minutes at 473 K, accompanied by the observation of a burst
of methanol and water by the mass spectrometry when the zeolite
was first contacted with DME (not shown) and then DME in cell
was flushed by N2 at reaction temperature to remove the gaseous
and physisorbed DME. Spectrum collected as N2 flushed the cell
A detailed DRIFT spectroscopy study was performed to gain a
better understanding of the effect of pressure on the formation of
acetyl. CO was carried into in situ cell under different pressures
after DME was absorbed on activated HMOR at 473 K. The spec-
tra of adsorbed species on catalysts under 0, 0.5, 1.5 and 3.0 MPa
for 10, 30 and 60 min were recorded and are shown in Fig. 7. The
Beer-Lambert Law is the base for the quantitative spectrometric
analysis. There are many chemical and instrument factors to influ-
ence the intensity of infrared absorption so only semi-quantitative
analysis is possible from the infrared spectrum. As for the DRIFT
spectroscopy, the catalyst sample was placed in the crucible, and
the gaseous reactant reacted with the acid sites on the surface
of catalyst to produce the adsorbed species. Therefore, it is more
difficult for the quantitative analysis by DRIFT spectrum. How-
ever, the intensity of band in DRIFT spectrum for the functional
groups in the identical chemical environment from one sample is
proportional to the concentration of the studied functional group.
Therefore it is possible to study the variation trend of functional
group over time by the in situ DRIFT spectroscopy in such circum-
stances. From Fig. 7(a), the C O in acetyl (CH3CO) at ∼1704 cm−1
was not observed when the carbonylation reaction was conducted
at atmosphere pressure for 10 min, meaning that acetyl was not
formed by the reaction CO and methoxyl. However, the band at
∼1600 cm−1, attributed to the high polymerized alkenes (C C)n or
is shown in Fig. 5(a). There are two bands, 2978 and 2868 cm−1
,
left in the end which are ascribed to the stretching modes of CH3,
signifying the formation of surface methoxyl groups.
Then, CO at 1.5 MPa was carried into the cell to react with
methoxyls. As shown in Fig. 5(b), a new band at ∼1704 cm−1
,
which was ascribed to the ꢀ(C O) stretch of the acetyl-zeolite
intermediate [19,35], was observed after the introduction of CO,
demonstrating the carbonylation reaction of CO with methoxyl
occurred. Another proof was that the intensity of bands in C
H
stretching region also changed along with the formation of acetyls
as shown in Fig. 5(b), because it is well known that the intensity of
CH3 would decrease significantly when CH3 groups are connected
with electronegative groups C O.
The intensity of infrared absorption depends on the magni-
tude of the transition probability and the change of the dipole
moment of the vibration. To be more specific, the intensity of

H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
5
Fig. 5. In situ DRIFT spectra of DME carbonylation conducted in the in situ cell (a: DME adsorption on HMOR at 473 K, b: Acetyl formation at 473 K and 1.5 MPa for 1 h, c:
MAc formation at 473 K, d: Infrared spectrum in C-H region after carbonylation reaction at 473 K).
Fig. 6. The intensity of m/e at 74 in MS after DME introduced into the cell.
coke species, was obvious, indicating that methoxyl on the sur-
face of HMOR transformed into olefin and then into cokes through
polymerization and hydrogen transfer reactions [36]. Therefore,
the reaction into olefin was favored in comparison with the DME
carbonylation reaction under atmospheric pressure. The band of
C
O in acetyl at ∼1704 cm−1 was observed as the pressure up to
0.5 MPa at 10 min. The intensity of the bands at ∼1704 cm−1 con-
tinued to increase with the pressure up to 1.5 MPa and 3.0 MPa at

6
H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
Fig. 7. In situ DRIFT spectra of acetyl at 473 K under different pressures ((a): 10 min, (b): 30 min, (c): 60 min).
10 min, therefore high pressure could promote the insertion reac-
tion. The C O in acetyl at ∼1704 cm−1 in spectrum was observed
under atmosphere pressure after CO was carried into the cell in situ
for 30 min, indicating that the carbonylation reaction could occur
under atmosphere pressure though the side reactions into cokes
were favored. The intensity of the acetyl for the 0.5, 1.5 and 3.0 MPa
became stronger as the time was prolonged from 10 to 30 min.
Moreover, the increase rate of band at ∼1704 cm−1 was quicker
than that of ∼1600 cm−1. The intensity of all of the bands contin-
ued to increase at the reaction time for 60 min. It was noticeable
that the effect of pressure on the formation of acetyl declined with
the extended reaction time by comparing the intensity of acetyl in
Fig. 7(c) and (a). Andrzej Malek compared the acetyl under high
and low CO pressure and found that only under high CO pressure
could the acetyl group accumulate to substantial amount for spec-
troscopic observation; on the contrary, other competing reactions
would either consume the active Brønsted methyl groups or react
with the acetyl groups at low CO pressure. Therefore, the acetyl
could not be observed by the IR [20]. However, it was clear in our
study that the spectrum of acetyl could be observed at atmosphere
pressure even though it would take longer time to accumulate for
observation.
Fig. 8. In situ DRIFT spectra of DME adsorption on HMOR at different temperatures.
Therefore, the carbonylation and cokes formation reactions
were competitive under high pressure, and methoxyl tended to
transform into cokes under atmosphere pressure in comparison
with the carbonyaltion. DME carbonyaltion reaction should be con-
ducted under high pressure.
3.4. Effect of reaction temperature on DME carbonylation
The carbonylation reactions by in situ DRIFT spectroscopy were
conducted at different temperatures in order to investigate the
effect of temperature on the formation of intermediates and side
reactions. After DME adsorbed on the catalysts with saturation
loading at different reaction temperatures, nitrogen was intro-
duced into the in situ cell for 30 min to degas the gaseous DME. The
DRIFT spectra in C-H region after the adsorption of DME are shown
in Fig. 8. The result was repeatable in comparison with that of Fig. 4,
indicating the reliability of designed experiment process. The bands
at 2978 and 2868 cm−1, attributing to the asymmetric and symmet-
ric stretching vibrations of methyl group, were not present at 403 K,
and could be observed at 433 K though there were also other bands
with different adsorbed forms of DME on acid sites. The methoxyl
groups were formed completely above 473 K. The bands gradually
lost intensity with the increase of temperature, methoxyl desorbed
or transformed into other species.
Fig. 9. In situ DRIFT spectra of acetyl on HMOR at different temperatures and
1.5 MPa.
meaning that no conversion occurred at 403 K. There was no
methoxyl and acetyl observed at 403 K, maybe inferring that the
formation of methoxyl was necessary for the insertion reaction of
CO. The C O in acetyl at ∼1704 cm−1 was observed as CO contacted
with adsorbed species on the surface of HMOR at 433 K and 1.5 MPa
in Fig. 9, suggesting that insertion of CO into methoxyl occurred.
The interesting and relevant changes appeared starting at 433 K,
so the conversion may start from this temperature. The conversion
The DRIFT spectra of C O in acetyl after carbon monoxide
reacted with DME adsorbed on surface for 60 min at different
temperatures and 1.5 MPa is shown in Fig. 9. Reaction performed
at 403 K for 60 min produced no significant change in spectrum,

H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
7
Fig. 11. The synthesis rate of MAc (gMAc/(gCat h)) over HMOR, HMOR-SP and HMOR-
MC catalysts. Reaction conditions: 473 K, 2.0 MPa, 5%DME, 35%CO, and 60%H2,
1500 mL/(gCat h).
Fig. 10. In situ DRIFT spectra in O-H region of HMOR, HMOR-SP and HMOR-MC at
473 K.
was enhanced comparing with that at lower reaction temperature
as the vibration band of (C O) (1704 cm−1) at 473 K became more
remarkable. The band intensity of (C O) gradually decreased with
the increasing of temperature to 493 and 513 K, maybe because
methoxyls and acetyls on the surface of zeolite transformed into
other species at high temperatures.
Three main bands are observed in the region of 1800–1550 cm−1
in Fig. 9, the band at ∼1704 cm−1 assigned to C O in acetyl, the band
at ∼1600 cm−1 attributed to the high polymerized alkenes (C C)n,
or coke species [36], the band at around 1640 cm−1 belonged
to water accompanied with the formation of hydrocarbons at
∼1600 cm−1. The appearance of band at ∼1600 cm−1 demonstrated
the occurrence of alkene polymerization or coke formation when
the conversion was carried out at relatively high temperature. On
a further temperature increase to 513 K, the band at ∼1600 cm−1
became more prominent with enhanced hydrogen transfer and
polymerization reactions. It could be obtained that more coke was
formed on the surface of HMOR as the temperature increased.
In conclusion, high temperature was in favor of the formation of
hydrocarbons in comparison with DME carbonylation.
Fig. 12. In situ DRIFT spectra of DME adsorption on HMOR, HMOR-MC and HMOR-SP
at 473 K.
3.5. Study of DME carbonylation on acid sites at different
channels of HMOR
main channels and HMOR-SP with acid sites only in side pockets.
The quantitative calculation of the amounts of acid sites in HMOR,
Bhan and his co-workers [10] reported that the insertion of CO
into chemisorbed methyl groups occurred selectively within 8MR
zeolite channels. Boronat et al. [11] further pointed out that it was
the T3-O33 position effective for the insertion reaction because of
the unusual orientation of the methoxy group with respect to the
8MR channel (parallel to the cylinder axis) by quantum-chemical
methods. Here, we studied the respective effect of acid sites in 8MR
and 12MR channels of mordenite on DME carbonylation by in situ
DRIFT spectroscopy.
The spectra in O H region of catalysts treated in N2 stream
within the in situ cell then cooled to 473 K and pyridine adsorbed in
HMOR-SP are shown in Fig. 10 and Fig. S1. The bands of O H were
at ∼3601, 3608 and 3587 cm−1 for HMOR, HMOR-MC, HMOR-SP,
respectively. The antisymmetric O H stretch spectrum in H-MOR
can be deconvoluted into two types of O-H species, with bands
at ∼3610 cm−1 for H+ in 12MR main channels and at ∼3590 cm−1
for H+ sites in 8MR side pockets [10,37,38]. Therefore, the acid
sites in HMOR-MC and HMOR-SP had been distinguished from each
other, meaning the HMOR-MC with the acid sites only in 12MR
HMOR-SP, HMOR-MC was conducted according to the results of 1
H
MAS NMR spectra (Fig. S2), 47% in 12MR main channels and 53% in
8MR side pockets of mordenite. The distribution of acid sites was
similar with the results in literatures [12,13], and was the base to
study DME carbonylation reaction with specific active sites.
Change in synthesis rate of MAc (gMAc/(gCat h)) with time-on-
stream (TOS) for HMOR, HMOR-SP, HMOR-MC catalysts is shown
in Fig. 11. For HMOR, synthesis rate increased quickly after a short
induction period and then declined soon, indicating that the cata-
lyst started to deactivate. But for HMOR-SP with acid sites in 8MR
side pockets, the induction period prolonged and more obviously,
the synthesis rate was stable for 16 h on stream without deacti-
vation. The synthesis rate for HMOR-MC was extremely low in
comparison with that of HMOR and HMOR-SP. The difference of
synthesis rate among HMOR, HMOR-SP and HMOR-MC catalysts
probably indicated the distinct effect of acid sites in different chan-
nels of mordenite on DME carbonylation reaction.
The DRIFT spectra in C H region after adsorption of DME on
three samples at 473 K are presented in Fig. 12, and the inset shows

8
H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
SP shown in Fig. 12. The band at around 1640 cm−1 assigned to the
adsorbed water was not observed on HMOR-SP, illustrating that
there was no water and coke formed as the carbonylation reaction
proceeded in 8MR side pockets of mordenite, which was consistent
with the better stability of HMOR-SP shown in Fig. 11. However, the
acetyl formed on HMOR-MC was very weak in comparison with that
of HMOR, although the intense methoxyl species was observed on
HMOR-MC in Fig. 12, illustrating that the insertion reaction of CO
into the methoxyl was almost unavailable owing to the lack of space
confinement. On the contrary, the band at 1640 cm−1 along with the
formation of water was strong in comparison with that on HMOR-
SP. Therefore, the acetyl could form on the acid sites in 12MR main
channels of mordenite but the yield was very low. Li et al. [12]
observed the acetaldehyde, not the acetyl groups, considered as
the intermediate to MAc in 12MR main channels of mordenite by
in-situ solid state NMR spectroscopy.
The in situ DRIFT spectroscopy results were illustrative of DME
catalytic test consequences in Fig. 11. The stability of HMOR-SP
was improved greatly with relative low synthesis rate of MAc in
comparison with that on HMOR because of the blocked diffusion
of reactants and products by adsorbed pyridine molecules in 12MR
main channels. The synthesis rate was lowest on the acid sites in
12MR main channels accompanied with the formation of coke.
Fig. 13. In situ DRIFT spectra of C O in acetyl for HMOR, HMOR-MC and HMOR-SP.
the corresponding O H spectra. The methoxyl was formed com-
pletely on HMOR, and the negative O H was at around 3603 cm−1
at 473 K. However, the intensity of methoxyl on HMOR-SP was rela-
tively weak in comparison with that on HMOR, even the adsorption
was completed at high pressure for long time, far less than 53% of
the intensity of methoxyl on HMOR, which was the concentration
of H+ in 8MR side-pockets characterized by the 1H MAS NMR tech-
nique. The acid sites in 12MR main channels of HMOR-SP were
deactivated by the basic pyridine molecules. It was reasonable that
the pore volume and channels of 12MR main channels were blocked
to certain extent. The 8MR side pockets were connected with 12MR
main channels and the reactants and products diffused into and
out zeolite through 12MR main channels. Therefore the pyridine
molecules adsorbed in 12MR main channels blocked the diffusion
of DME into 8MR side pocket and resulted in the lower utilization
rate of acid sites in side pockets of mordenite. The C H region spec-
trum after DME saturated adsorption on HMOR-MC is also shown
in Fig. 12. The bands at 2978 and 2868 cm−1 ascribed to CH3 were
very strong, thus there were methoxyls formed on HMOR-MC.
The bands of C O in acetyl groups, formed by the reaction of CO
with methoxyl on HMOR, HMOR-SP and HMOR-MC are shown in
Fig. 13. The acetyl on HMOR was intense but that on HMOR-SP was
weak maybe because of the weak intensity of methoxyl on HMOR-
4. Conclusions
Base on the observation above by in situ diffuse reflectance
infrared Fourier transform (DRIFT) spectroscopy, a possible mech-
anism for the DME carbonylation and side reactions is offered as
a scheme of Fig. 14. The methoxyl, chemisorbed DME on the acid
sites, was observed above 433 K. Reaction of methoxyl and CO gen-
erated acetyl and then acetyl reacted with DME to produce MAc
and to regenerate methoxyl again. The formation of acetyl took
longer time than those of the saturation adsorption of DME and
the reaction of acetyl with DME, implying that the insertion of CO
into methoxyl was the rate-controlling step in DME carbonylation.
The reaction of DME into olefins was favored at low pressure and
high temperature, whereas high pressure promoted the DME car-
bonylation reaction. The in situ carbonylation reactions in different
channels of mordenite demonstrated that the acid sites both in
side pockets and 12MR main channels of mordenite were active for
the formation of acetyl. Meanwhile, acid sites in side pockets were
more active for carbonylation than those in 12MR main channels
which favored the formation of coke.
Fig. 14. Proposed mechanism for the DME carbonylation and side reactions.

H. Zhou et al. / Journal of Molecular Catalysis A: Chemical 417 (2016) 1–9
9
Acknowledgements
[15] H. Zhou, W.L. Zhu, L. Shi, H.C. Liu, S.P. Liu, S.T. Xu, Y.M. Ni, Y. Liu, L. Li, Z.M. Liu,
Catal. Sci. Technol. 5 (2015) 1961–1968.
[16] S.R. Wang, W.W. Guo, L.J. Zhu, H.X. Wang, K.Z. Qiu, K.F. Cen, J. Phys. Chem. C
119 (2015) 524–533.
[17] K. Fujimoto, T. Shikada, K. Omata, H. Tominaga, Chem. Lett. (1984) 2047–2050.
[18] P. Cheung, A. Bhan, G.J. Sunley, E. Iglesia, Angew. Chem. Int. Ed. 45 (2006)
1617–1620.
[19] P. Cheung, A. Bhan, G.J. Sunley, D.J. Law, E. Iglesia, J. Catal. 245 (2007) 110–123.
[20] X.Y. Chen, M.L. Neidig, R. Tuinstra, A. Malek, J. Phys. Chem. Lett. 1 (2010)
3012–3015.
[21] M. Milosevic, S.L. Berets, Appl. Spectrosc. Rev. 37 (2002) 347–364.
[22] B. Klingenberg, M.A. Vannice, Appl. Catal. B: Environ. 21 (1999) 19–33.
[23] R. Yang, Y. Fu, Y. Zhang, N. Tsubaki, J. Catal. 228 (2004) 23–35.
[24] L. Chen, J. Li, M. Ge, Environ. Sci. Technol. 44 (2010) 9590–9596.
[25] D. Wang, L. Zhang, K. Kamasamudram, W.S. Epling, ACS Catal. 3 (2013)
871–881.
We acknowledge the financial support from the China Postdoc-
toral Science Foundation (Grant No: 2014M560224)
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2016.02.
032.
References
[26] Z. Liu, S. Zhang, J. Li, L. Ma, Appl. Catal. B: Environ. 144 (2014) 90–95.
[27] J. Llorca, N. Homs, P. Ramirez de la Piscina, J. Catal. 227 (2004) 556–560.
[28] E. Chenu, G. Jacobs, A.C. Crawford, R.A. Keogh, P.M. Patterson, D.E. Sparks, B.H.
Davis, Appl. Catal. B: Environ. 59 (2005) 45–56.
[29] Z. Wu, S. Zhou, H. Zhu, S. Dai, S.H. Overbury, J. Phys. Chem. C 113 (2009)
3726–3734.
[30] H.F. Xue, X.M. Huang, E.S. Zhan, M. Ma, W.J. Shen, Catal. Commun. 37 (2013)
75–79.
[31] J.G. Chen, P. Basu, T.H. Ballinger, J.T. Yates, Langmuir 5 (1989) 352–356.
[32] S.M. Campbell, X.Z. Jiang, R.F. Howe, Microporous Mesoporous Mater. 29
(1999) 91–108.
[1] C.A. Cardona, O.J. Sanchez, Bioresour. Technol. 98 (2007) 2415–2457.
[2] M.H. Youn, J.G. Seo, P. Kim, I.K. Song, J. Mol. Catal. A: Chem. 261 (2007)
276–281.
[3] V. Subramani, S.K. Gangwal, Energy Fuels 22 (2008) 814–839.
[4] H. Zheng, R. Zhang, Z. Li, B. Wang, J. Mol. Catal. A: Chem. 404–405 (2015)
115–130.
[5] G. Larsen, E. Lotero, M. Marquez, H. Silva, J. Catal. 157 (1995) 645–655.
[6] X.G. San, Y. Zhang, W.J. Shen, N. Tsubaki, Energy Fuels 23 (2009) 2843–2844.
[7] A.G. Slater, A.I. Cooper, Science 348 (2015) aaa8075.
ˇ
[8] J. Cejka, G. Centi, J. Perez-Pariente, W.J. Roth, Catal. Today 179 (2012) 2–15.
[9] V. Malgras, Q. Ji, Y. Kamachi, T. Mori, F.-K. Shieh, K.C.W. Wu, K. Ariga, Y.
Yamauchi, Bull. Chem. Soc. Jpn. 88 (2015) 1171–1200.
[10] A. Bhan, A.D. Allian, G.J. Sunley, D.J. Law, E. Iglesia, J. Am. Chem. Soc. 129
(2007) 4919–4924.
[11] M. Boronat, C. Martinez-Sanchez, D. Law, A. Corma, J. Am. Chem. Soc. 130
(2008) 16316–16323.
[12] B.J. Li, J. Xu, B. Han, X.M. Wang, G.D. Qi, Z.F. Zhang, C. Wang, F. Deng, J. Phys.
Chem. C 117 (2013) 5840–5847.
[13] J.L. Liu, H.F. Xue, X.M. Huang, P.H. Wu, S.J. Huang, S.B. Liu, W.J. Shen, Chin. J.
Catal. 31 (2010) 729–738.
[33] A.G. Pelmenschikov, J.H.M.C. Vanwolput, J. Janchen, R.A. Vansanten, J. Phys.
Chem. 99 (1995) 3612–3617.
[34] A.G. Pelmenschikov, R.A. Vansanten, J. Phys. Chem. 97 (1993) 10678–10680.
[35] O. Kresnawahjuesa, R.J. Gorte, D. White, J. Mol. Catal. A: Chem. 208 (2004)
175–185.
[36] Y.X. Wei, D.Z. Zhang, Z.M. Liu, B.L. Su, J. Catal. 238 (2006) 46–57.
[37] M. Maache, A. Janin, J.C. Lavalley, E. Benazzi, Zeolites 15 (1995) 507–516.
[38] J. Datka, B. Gil, A. Kubacka, Zeolites 17 (1996) 428–433.
[14] T. Blasco, M. Boronat, P. Concepcion, A. Corma, D. Law, J.A. Vidal-Moya,
Angew. Chem. Int. Ed. 46 (2007) 3938–3941.