Activity enhancement of Nafion resin: Vapor-phase carbonylation of dimethoxymethane over Nafion-silica composite

Article abstract of DOI:10.1016/j.apcata.2015.03.010

A kind of composite materials consisting of nano-sized Nafion resin homogeneously dispersed in a high-surface silica matrix was prepared by sol-gel method and used as catalysts in the vapor-phase carbonylation reaction of dimethoxymethane for production of methylmethoxy acetate. These composite catalysts showed much higher catalytic activity in comparison with the original pure Nafion resin catalyst. The remarkable improvement of catalytic performance was due to the increased accessibility of acid sites of Nafion resin in the composite catalysts. The contents of Nafion-H resin and silica precursors were found to have a great influence on the structure of the composite, resulting in different catalytic activity during DMM carbonylation reactions. Furthermore, the comparison of the composite catalysts with high-silica H-Y and H-Beta zeolite catalysts demonstrated that the excellent catalytic performance is closely related to the higher acid strength and larger pores of Nafion-silica composite materials.

Full text of DOI:10.1016/j.apcata.2015.03.010

Applied Catalysis A: General 497 (2015) 153–159
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Activity enhancement of Nafion resin: Vapor-phase carbonylation of
dimethoxymethane over Nafion-silica composite
Shiping Liua,b, Wenliang Zhua, Lei Shia, Hongchao Liua, Yong Liua, Youming Nia,
Lina Lia,b, Hui Zhoua,b, Shutao Xua, Yanli Hea, Zhongmin Liua,∗
a National Engineering Laboratory for Methanol to Olefins, Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics,
Chinese Academy of Sciences, Dalian 116023, PR China
b 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:
A kind of composite materials consisting of nano-sized Nafion resin homogeneously dispersed in a
high-surface silica matrix was prepared by sol–gel method and used as catalysts in the vapor-phase
carbonylation reaction of dimethoxymethane for production of methylmethoxy acetate. These compos-
ite catalysts showed much higher catalytic activity in comparison with the original pure Nafion resin
catalyst. The remarkable improvement of catalytic performance was due to the increased accessibility of
acid sites of Nafion resin in the composite catalysts. The contents of Nafion-H resin and silica precursors
were found to have a great influence on the structure of the composite, resulting in different catalytic
activity during DMM carbonylation reactions. Furthermore, the comparison of the composite catalysts
with high-silica H-Y and H-Beta zeolite catalysts demonstrated that the excellent catalytic performance
is closely related to the higher acid strength and larger pores of Nafion-silica composite materials.
© 2015 Elsevier B.V. All rights reserved.
Received 23 December 2014
Received in revised form 7 March 2015
Accepted 10 March 2015
Available online 18 March 2015
Keywords:
Dimethoxymethane
Carbonylation
Ethylene glycol
Methyl methoxy acetate
Nafion-silica composite
1. Introduction
resulting in server environmental and safety problems. In the pre-
vious studies, more environmentally friendly solid acids, including
Ethylene glycol (EG), an important automotive antifreeze and
an intermediate for polyester resins and fibers [1], has a world-
wide capacity of about 23 million tons per year, and the demand
for this chemical is expected to grow in the future [2]. Presently, EG
is mainly produced through the hydration of ethylene oxide syn-
thesized by the partial oxidation of petroleum-derived ethylene.
The depleting of limited oil reservoirs has made it necessary and
pressing to consider alternative starting materials for EG synthesis.
Syngas, which is a mixture of CO and H2 and can be readily gener-
ated from non-oil-based feedstocks such as the relatively abundant
coal, natural gas, and renewable biomass, is a potential candidate
for EG synthesis. One of attractive routes using syngas as the star-
ing material is the carbonylation of formaldehyde catalyzed by
acids to glycolic acid followed by esterification and hydrogena-
tion [3,4]. However, acid-catalyzed carbonylation of formaldehyde,
the key step during EG synthesis, has been exclusively conducted
with strong liquid mineral acids, such as H2SO4 and HF, in liquid
phase under high pressure [3,5]. One major concern about this pro-
cess is the use of highly corrosive acids as homogenous catalysts,
acidic resins [6], acidic zeolites [7], and heteropolyacids [8], have
been explored to replace strong homogenous catalysts to solve the
corrosion problem, but the catalytic performance of these hetero-
geneous catalysts is inferior to that of homogeneous systems. On
the other hand, the solubility of CO in most solvents is poor [9] and
high pressure is required to promote CO concentration in liquid
phase. Up to now, the development of a highly selective and envi-
ronmentally benign process for the carbonylation of formaldehyde
is still a challenge.
Recently, the carbonylation of dimethoxymethane (DMM), a
derivative of formaldehyde, was reported with acidic zeolite cat-
alysts in gas phase [10]. Cliek et al. reported that the selectivity of
aimed MMAc reached 79% with H-Faujasite catalyst at low pressure
(a CO pressure of 3 atm). The results were comparable to those of
formaldehyde carbonylation catalyzed by solid acid resins in liq-
uid phase at CO pressure of 238 atm [11]. Therefore, it is inherently
advantageous to perform the carbonylation of DMM in gas phase.
Based on theoretical and kinetic studies, Bell and co-workers sug-
gested that the acid-catalyzed carbonylation of DMM occurred by
the Koch-type mechanism [12,13]. For Koch carbonylation reac-
tions, the acid strength of catalysts generally plays a vital role at
the catalytic activity. In some cases, only strong acids can effec-
tively catalyze the carbonylation reactions [14,15]. Further studies
∗
Corresponding author. Tel.: +86 411 84685510; fax: +86 411 84685510.
E-mail address: liuzm@dicp.ac.cn (Z. Liu).
http://dx.doi.org/10.1016/j.apcata.2015.03.010
0926-860X/© 2015 Elsevier B.V. All rights reserved.

154
S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
by Bell and co-workers showed that the zeolites framework types
have a significant influence on the catalytic performance, and high-
silica H-Y zeolites are more selective for the carbonylation reaction
than other zeolite catalysts because the disproportion of DMM, the
side reaction, is suppressed within the relatively large supercages
of H-Y zeolites [16]. Thus, according to the present findings, it is of
interest to develop new type of catalysts with strong acid sites and
large pores for the carbonylation of DMM.
20 mL NaOH solution (0.1 M) was mixed with 5 g of 5 wt% Nafion
solution (solution B). Then, the resulting clear solution A was
rapidly added to the solution B under vigorous stirring. Within a few
seconds, the gel was formed. After aging at ambient temperature
for 4 h, the solid gel was dried for two days in flowing nitrogen at
95 ◦C, and changed into glass-like solid. The resulting material was
gently grounded into smaller particles, treated with 25 wt% nitric
acid solution overnight at 75 ◦C. Finally, the composite catalyst was
obtained after drying in a vacuum oven at 95 ◦C for 24 h. Analo-
gously, the contents of Nafion resin in the samples can be adjusted
by varying the amounts of Nafion resin solution. The obtained sam-
ples were denoted as TEOS/Nafion-X, where X is referred as the
predetermined loadings of Nafion resin.
In addition, other two silica precursors including sodium silicate
(“water glass”), and colloidal silica were also used for prepar-
ing the Nafion-silica composite in the present work. In the case
of using sodium silicate as the silica source, the preparing pro-
cedure was modified from a literature method [28]. In a typical
procedure, 6.5 mL of HCl solution was added to the 17 g of 2.5 wt%
Nafion solution under stirring condition. Then the mixture was
rapidly added to the 25 g of a solution of sodium silicate (10 wt%
SiO2) and the gelation immediately occurred. When using col-
loidal silica as the silica source, 15 g Nafion solution was added
to 16.7 g of colloidal silica solution, then, HCl solution (1 M)
was added into the mixture to induce the gelation. The dry-
ing and re-acidification steps were carried out in the same way
as described for the TEOS-based samples. The obtained samples
using sodium silicate and colloidal silica as the silica sources
were sighed as Na-silicate/Nafion-15 and Col-silica/Nafion-15,
respectively.
Nafion-H resin, a cross-linked perfluorinated polymeric sulfonic
acid resin, appears to be an ideal candidate for the vapor-phase
carbonylation of DMM, since the acid strength of Nafion resin is
markedly high due to the high electron-withdrawing effect of ␣-CF2
moiety attached to the sulfonic acid groups [17]. Also, in con-
trast with the acid sites located within the channels of zeolites,
the acid sites of Nafion resin have no steric constraint of the pore
walls. Unfortunately, one major drawback of commercially avail-
able resin is the very low surface area, and most active sites are
buried in the bulk material. Therefore, they are not accessible for
reactants, especially in gas-phase reactions. In order to improve the
catalytic efficiency of Nafion resin catalyst, several preparing meth-
ods have been adopted to increase the accessibility of acid sites
within Nafion resin. For example, Nafion resin has been dispersed
onto the high surface area supports, such as highly porous SiO2
[18], siliceous SBA-15 or MCM-41 materials [19,20] by impregna-
tion method. These supported Nafion catalysts showed improved
performance in various reactions, but would suffer from leach-
ing problem in the solvent [21]. An alternative strategy which has
attracted much attention is to trap the small Nafion resin particles
within highly porous silica matrix by using the sol–gel technique
[22]. The obtained composite materials ingeniously have high sur-
face areas and highly available strong acid sites of Nafion resin. The
composite have been used as catalysts for various acid-catalyzed
reactions, such as acylation [20], alkylation [23], Friedel-Crafts
benzylation [24,25], esterification [26], and showed much higher
reactivity than the pure Nafion resin. Here, we first reported to
employ the composite materials as catalysts for vapor-phase car-
bonylation of DMM. In this work, Nafion-silica composite catalysts
with different Nafion resin loadings were prepared, characterized,
and evaluated for the vapor-phase carbonylation of DMM. The
same type of results obtained with high-silica acidic zeolites and
pure Nafion resin were included for comparison. In addition, the
influence of silica sources used for preparation of the composite
catalysts on the catalytic activity for DMM carbonylation was dis-
cussed as well.
For comparison, supported Nafion resin catalyst was prepared
by impregnating 4.5 g of SiO2 support into 7 mL of Nafion-H solution
containing 10% Nafion-H resin (concentrated from 5 wt% Nafion
solution), followed by drying at room temperature for 12 h, then
overnight under vacuum condition at 95 ◦C. Before testing, the
sample must be treated with nitric acid solution as described
above. The SiO2-supported Nafion catalyst was sighed as SiO2/
Nafion-15.
2.3. Catalytic test
The tests of DMM carbonylation reaction were carried out in
a continuous-flow fixed-bed stainless steel reactor. Typically, an
appropriate amount of catalyst was loaded into a reactor tube
(8 mm internal diameter). The sample was then heated to 120 ◦C
for an hour under nitrogen atmosphere (30 mL min−1) to remove
residual water and then cooled to the reaction temperature. DMM
was carried into reactor by carbon monoxide, which was passed
through a stainless steel saturator containing liquid DMM isother-
mally held at 20 ◦C. The reaction effluent was analyzed by on line gas
chromatography (Agilent 7890) equipped with a flame ionization
detector. The conversion and selectivity were calculated as follows
[10]:
2. Experimental
2.1. Reagents
All the chemicals were obtained commercially and used without
any further purification. Nafion NR50 resin, Nafion resin solu-
tion (D520, 5.0–5.4 wt% Nafion), tetraethyl orthosilicate (TEOS),
colloidal silica and sodium silicate were obtained from Aldrich.
High-silica H-Y zeolite (Si/Al = 40) and high-silica H-Beta (Si/Al = 31)
zeolite were purchased from Zeolyst and Nankai University catalyst
Co., Ltd., respectively.
ꢀ
ꢁ
1 − 3CDMM
DMM conversion =
× 100%
(1)
3CDMM + 2CDME + 2CMF + Cmethanol + 3CMMAc
ꢀ
ꢁ
2.2. Catalysts preparation
3CMMAc
MMAc selectivity =
× 100%
2CDME + 2CMF + Cmethanol + 3CMMAc
In this study, a series of Nafion-silica composite samples with
different loadings of Nafion resin were prepared. When using
TEOS as the silica source, the composite samples were prepared
according to a procedure described elsewhere [22,27]. First, 4 mL
deionized water and 1 mL HCl solution (0.1 M) were mixed with
15 g TEOS and stirred for about 1 h (solution A). In another flask,
(2)
where Ci was the molar concentration of compound i in the
reaction effluent and n was the number of carbon derived from
DMM.

S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
155
2.4. Characterization of the catalysts
of the hysteresis loop is observed between samples with different
types of silicon sources, i.e. sodium silicate and colloidal silica. The
former samples show a hysteresis loop in the P/P0 rang of 0.6–0.85,
whereas the hysteresis loops of the latter samples shift toward
higher relative pressures (0.85–0.98), indicating the presence of
larger pores. The structural parameters, such as BET surface area,
pore volume, and pore diameter, are also summarized in Table 1. It
can be seen that the specific surface area of TEOS/Nafion-5 is about
350 m2 g−1, considerably higher than that of pure Nafion resin (less
than 0.02 m2 g−1) [17]. Here, it should be reminded that both the
silica matrix and Nafion resin contribute to the high surface area
of Nafion-silica composite material, and the effective surface area
of Nafion resin alone in the composite is lower than the apparent
value. Harmer and co-workers estimated the effective surface area
of Nafion resins to be about 100 m2 g−1 [22]. With the increase of
the Nafion resin loadings, the surface area of Nafion-silica compos-
ite begins to reduce because too much Nafion resin will disrupt
the continuity of the porous silica network or lead to the blockage
of the pores. Additionally, though the loadings of Nafion resin are
very similar, the surface area of TEOS/Nafion-15 sample is much
higher than that of Na-silicate/Nafion-15 and Col-silica/Nafion-15,
suggesting that the type of silica sources has a pronounced effect
on the microstructure of the synthesized composite samples.
Fig. 1a is an SEM micrograph of the surface of TEOS/Nafion-
15 sample. As displayed, the sample appears to be formed by
agglomeration of smaller particles. Furthermore, the porosity of the
material is confirmed by close examination of the microstructure.
Fig. 1b is the TEM micrograph of the TEOS/Nafion-15 sample, again
showing the particulate substructure and the porosity indicated by
the white areas.
Fig. 2 shows the EDX elemental mapping images of TEOS/Nafion-
15 sample. In the images, the bright dots indicate the present of
Si, O, F and S and no area enriched entirely in S or F is observed,
indicating that Nafion resin is homogeneously distributed within
the silica matrix. Indeed, Harmer suggested that Nafion resin and
silica can be well intermixed at the nanometer level [22]. It can
be expected that the effective surface of Nafion resin in compos-
ite material is considerably higher than that of pure Nafion resin,
resulting in the great improvement of accessibility of the reactants
toward the active sites.
In order to confirm the accessibility of acid sites in the compos-
ite catalysts, IR spectra of TEOS/Nafion-15 sample after pyridine
adsorption are shown in Fig. 3. It can be seen that three bands
at 1545, 1490, and 1445 cm−1, characteristic of bands of Brönsted
and Lewis acid, obviously demonstrate the presence of both Brön-
sted and Lewis acid sites. The band at 1540 cm−1 corresponds to
the pyridinium ions resulting from the reaction of pyridine with
proton sites, and the band at 1445 cm−1 indicates the presence of
Lewis acid sites. After evacuating at 80 ◦C for 30 min, the inten-
sity of band at 1445 cm−1 decreased significantly, and disappeared
after the temperature increased to 150 ◦C, indicating that the acid-
ity of Lewis acid is rather weak. However, the band at 1544 cm−1
remained almost unchanged, even after evacuating at 150 ◦C These
results, therefore, clearly confirm the accessibility of the strong sul-
fonic acid groups of Nafion resin for the gaseous pyridine molecules
and its stability under high temperature condition.
Nitrogen adsorption–desorption isotherms were obtained using
a Micromeritics 2020 apparatus. Prior to the measurements, the
samples were degassed at 120 ◦C at high vacuum for 3 h. BET model
was used to estimate the surface areas of the samples. The meso-
pore size was calculated from BJH model using the adsorption
branches of isotherms. The pore volumes of samples were calcu-
lated by t-plot method.
Scanning electron microscope (SEM) and energy-dispersive X-
ray (EDX) microanalyses were obtained using a Hitachi SU8020
apparatus operated at 20 kV. Transmission electron micrographs
(TEM) were obtained using a JEM-2100 microscope operated at
200 kV.
IR spectra of adsorbed pyridine (Py-IR) were collected on BRUK-
ERTENSOR 27 FT-IR spectrophotometer. The powder sample was
pressed to self-supported wafer, which was then placed into an IR-
cell connected to a vacuum system. The sample was outgassed at
150 ◦C for 1 h under vacuum condition and subsequently exposed
to the pyridine vapor for 5 min after cooling down to room temper-
ature. Then, the spectra were recorded at room temperature after
evacuating at different temperatures for 30 min.
31P MAS NMR spectrum was recorded on a Bruker AvanceIII
600 spectrometer using a single pulse sequence under the follow-
ing conditions: ␲ inpulse-width, 2.25 ␮s; recycle delay, 4 s; sample
spinning rate, 12 kHz. Aqueous 85% H3PO4 solution was used as
external reference for the 31P NMR chemical shift. Prior to the
adsorption of trimethylphosphine oxide (TMPO) molecule, Nafion-
silica sample was treated in a flask under vacuum at 120 ◦C for 12 h
to remove the adsorbed water molecule. Then, the dehydrated sam-
ple was transferred into glovebox and an appropriate amount of
TMPO dissolved in CH2Cl2 was added to the flask containing the
sample. The obtained mixture was stirred overnight to ensure the
efficient reaction between acid sites and TMPO molecules. Subse-
quently, the solvent was removed in flowing N2 at 40 ◦C. Finally,
the TMPO-loaded sample was transferred into a NMR rotor for NMR
test.
The acid site concentrations of samples were determined using
aqueous NaCl solutions and titrated by dilute aqueous NaOH. In a
typical experiment, 0.5 g of solid was added to 20 mL of NaCl (2 M)
solution, and vigorously stirred at room temperature overnight.
After removing the solids by centrifugation, the obtained solution
was titrated by a drop wise addition of 0.05 M NaOH solution.
The weight loss curves were recorded on SDT Q600 thermal
analyzer from ambient temperature to 700 ◦C at a ramping rate of
10 ◦C min−1 in nitrogen atmosphere to determine the real contents
of Nafion resin in the samples (Fig. S1, see supplementary informa-
tion).
3. Results and discussion
3.1. Catalyst characterization
Compared with the pure Nafion resin, Nafion-silica compos-
ite materials prepared by sol–gel method have properties of
mesoporous silica matrix with high surface area. Nafion-silica
composites presented the advantages of both Nafion resin and
mesoporous silica matrix. Here, the detailed information about
surface area and porosity of porous materials were obtained by
analysis of the nitrogen adsorption/desorption isotherms, which
are shown in Fig. S2 (see supplementary information). As can be
seen from Fig. S2, all the samples display type-IV isotherms with
clear H1-type adsorption–desorption hysteresis [29], indicating the
good development of mesoporosity due to the closely arranged
spherical particles [30]. A remarkable difference in the location
The previous studies have demonstrated that 31P MAS NMR
spectroscopy with adsorbed trimethylphosphine oxide (TMPO)
is suitable to characterize the acid strength of solid acids [31],
becausethe 31P NMR chemical shift of TMPOmoleculesadsorbedon
Brønsted acid sites can increase with the increase of acid strength
[32]. Fig. 4 shows the 31P MAS NMR spectrum of TMPO adsorbed
on the TEOS/Nafion-15 sample. As displayed, three peaks at 84, 53
and 36 ppm are observed. The weak peak at 36 ppm is ascribed to
crystalline TMPO, and the peak at 53 ppm is attributed to TMPO
adsorbed on Lewis acid sites, consistent with the analysis from

156
S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
Table 1
Textural and chemical properties of the Nafion-silica samples.
b
d
e
Entry
Sample
Nafion-H (wt%)a
BET surface (m2 g−1
)
Dpore (nm)c
Vpore (cm3 g−1
)
H+ capacity (mmol g−1
)
1
2
3
4
5
6
7
8
TEOS/Nafion-5
5.7
15.2
31.0
43.7
60.5
18.0
14.5
21.8
354
349
264
127
30
370
123
107
8.4
6.5
7.3
10.2
13.3
5.4
26.0
21.3
0.84
0.73
0.54
0.36
0.09
0.61
0.77
0.55
0.05
0.14
0.24
0.38
0.58
0.18
0.14
0.17
TEOS/Nafion-15
TEOS/Nafion-30
TEOS/Nafion-40
TEOS/Nafion-60
SiO2/Nafion-15
Na-silicate/Nafion-15
Col-silica/Nafion-15
a
Weight loss of samples in the temperature range of 200–600 ◦C.
BET surface area.
BJH adsorption average pore diameter.
Single-point pore volume at P/P0 = 0.975.
Determined by acid–base titration method.
b
c
d
e
Fig. 1. SEM micrograph (a) and TEM micrograph (b) of the TEOS/Nafion-15 sample.
Fig. 2. EDX-mapping images of TEOS/Nafion-15.

S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
157
100
Methanol
MF
DME
Conversion
MMAc
90
d 150
c 80
80
70
60
50
40
30
b RT
a Without Py
20
L
B+L
B
10
0
1580
1560
1540
1520
1500
1480
1460
1440
Pure Nafion
TEOS/Nafion-15
SiO2/Nafion-15
Wavenumbers (cm-1)
Fig. 5. The catalytic performance DMM carbonylation over different Nafion cat-
alysts at 2 h TOS (reaction conditions: reaction pressure = 30.0 atm, DMM partial
pressure = 0.42 atm, CO stream = 85 mL min−1).
Fig. 3. FT-IR spectra of TEOS/Nafion-15 sample (a) evacuated at 150 ◦C before pyri-
dine adsorption, after pyridine adsorption and evacuation (b) at room temperature,
(c) at 80 ◦C (d) 150 ◦C.
the reaction conditions, including reaction temperatures, gas space
velocities, reaction pressure and partial pressures of all reactants
in all three runs, were nearly the same. In the case of the test with
pure Nafion resin, the catalyst was physically mixed with inert silica
particles. Clearly, the TEOS/Nafion-15 sample is the most active cat-
alyst among these catalysts, whereas the pure Nafion sample shows
the lowest DMM conversion. The results unambiguously demon-
strate that catalyst activity of the Nafion catalysts correlates very
well with their accessibility of acid sites, i.e., the higher accessibil-
ity of acid sites in Nafion resin results in higher catalyst activity. It
is also found that Nafion-silica composite sample shows a slightly
higher conversion of DMM than silica-supported Nafion catalyst
with identical selectivity of MMAc. This difference in catalytic activ-
ity may be ascribed to the larger number of surface acid sites on
Nafion-silica composite sample compared to the silica-supported
Nafion catalyst. It is consistent with the results report by Olah and
co-workers [18]. Additionally, as reported in the literature, Nafion-
silica composite catalyst also catalyzes the disproportionation of
DMM to produce dimethyl ether (DME) and methyl formate (MF)
[10], and a small amount of methanol attributed to the decompo-
sition of MF is detected [16].
84
53
*
36
140 120 100 80
60
40
20
0
-20 -40
ppm
Fig. 4. 31P MAS NMR spectrum of TMPO adsorbed on TEOS/Nafion-15 (asterisk
Fig. 6 shows the effect of Nafion loadings on the catalytic
performance of the composite catalysts for the DMM carbony-
lation reaction. It can be seen that increasing in the loadings of
Nafion resin leads to the improvement of catalytic performance,
and the conversion of DMM reaches a maximum value, up to 95%
when the amount of Nafion resin is 40%. Apparently, the excel-
lent performance of TEOS/Nafion-40 catalyst is attributed to the
higher amount of accessible strong acid sites in the composite
sample. Though the activity of Nafion-silica catalysts increases
with the increase of Nafion contents, the opposite trend occurs
when the activity per acid site is considered. The rate of MMAc
synthesis is 12.3 mol (mol H+)−1 h−1 for TEOS/Nafion-5, but only
7.2 mol (mol H+)−1 h−1 for the TEOS/Nafion-40 sample. When the
Nafion resin content increases to 60%, the conversion of DMM
decreases some, but the rate of MMAc synthesis has a sharp reduc-
tion, which is attributed to the decrease in the number of the
accessible acid sites, evidenced by the low specific surface area
(only 30 m2 g−1, seen Table 1).
denote spinning sidebands).
Py-FTIR spectra. The strong peak at 84 ppm is associated with TMPO
adsorbed on Brönsted acid sites. The value is very close to the
threshold value for 31P chemical shift of TMPO adsorbed on solid
superacid [32], suggesting that the acid strength of Nafion-silica
composite sample is very high.
Additionally, several other parameters related to the acidic
properties of Nafion-containing samples are also listed in Table 1.
The loadings of Nafion resins on the samples were determined by
the weight loss between 200 and 600 ◦C through thermo decom-
position of Nafion resin, and the actual loadings of Nafion resin are
well consistent with the acid capacities of samples measured by
acid–base titration.
3.2. Catalytic studies
Results of vapor-phase carbonylation of DMM over three differ-
ent kinds of Nafion catalysts are depicted in Fig. 5. In order to obtain
the experimental data that are indicative of intrinsic catalysts, the
amounts of Nafion resin in these three tests were equivalent, and
In Fig. 7, MMAc selectivity and DMM conversion versus TOS
for Nafion-silica composite catalysts prepared from different sil-
ica precursors have been plotted. It is interesting to note that there
is a slight difference in the conversion of DMM between TEOS and

158
S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
100
15
12
9
100
90
80
70
60
50
40
30
20
10
0
Conversion
Selectivity
90
80
70
60
50
40
30
20
10
0
Rate of MMAc synthesis
Methanol
MF
DME
6
MMAc
Conversion
3
60
5
Nafion contents in Nafion-silica catalysts (%)
30
40
15
H-Beta
TEOS/Nafion-15
H-Y
Fig. 6. The effect of loadings of Nafion on the catalytic performance of DMM
carbonylation at 2 h TOS (reaction conditions: catalyst weight = 0.4 g, reaction pres-
sure = 30.0 atm, DMM partial pressure = 0.42 atm, CO stream = 85 mL min−1).
Fig. 8. Carbonylation of DMM over different acid catalysts (reaction conditions: cat-
alyst weight = 0.4 g, reaction pressure = 30.0 atm, DMM partial pressure = 0.42 atm,
CO stream = 85 mL min−1).
100
Selectivity to MMAc
90
80
,
,
,
Na-silicate/Nafion-15
TEOS/Nafion-15
Silica/Nafion-15
70
60
50
40
30
20
10
Conversion of DMM
1
2
3
4
5
6
7
8
9
10 11 12
Time-on-stream (h)
Fig. 7. The carbonylation of DMM over Nafion-silica catalysts derived from
different silica sources (reaction conditions: catalyst weight = 0.4 g, reaction pres-
sure = 30.0 atm, DMM partial pressure = 0.42 atm, CO stream = 85 mL min−1).
sodium silicate derived Nafion-silica samples, but the influence on
the selectivity of MMAc can be negligible. A plausible explanation
for the results is that the silica sources could alter the state of aggre-
gation of Nafion resin within the porous silica matrix, which in turn
had a subtle influence on the catalytic reactivity of Nafion-silica
catalysts [21]. In addition to the high selectivity, a good catalytic
stability is observed on Nafion-silica composite catalyst. For exam-
ple, the DMM conversion over TEOS/Nafion-15 catalyst decreases
from 50% at 2 h on stream to 48% at 12 h on stream.
Fig. 8 provides a comparison of the catalytic performance
between the Nafion-silica composite and zeolites catalysts for the
vapor-phase carbonylation of DMM. As displayed, Nafion-silica
sample shows higher conversion than high-silica H-Y catalyst, but it
presents a lower conversion than high-silica H-Beta catalyst. Here,
it should be pointed out that the acid density of TEOS/Nafion-15
sample is much lower than that of the acidic zeolites. Thus, the spe-
cific rate of the formation of MMAc for TEOS/Nafion-15 is around
9.8 mol (mol H+)−1 h−1, which is higher than those for H-Y and H-
Beta zeolite catalysts (2.6 and 2.9 mol molAl−1 h−1, respectively).
The superior catalytic activity of Nafion-silica catalyst is closely
related to the strong acid strength of Nafion resin, which was
Scheme 1. Proposed mechanism for acid-catalyzed DMM carbonylation.
indicated by the high 31P chemical shift of TMPO adsorbed on the
composite catalyst (84 ppm, shown in Fig. 4), while the reported
31P chemical shift of TMPO adsorbed on H-Beta and H-Y zeo-
lites were located at 72 and 65 ppm [33,34], respectively. The
observed influence of acid strength on reactivity well reflects the
carbocationic nature of acid-catalyzed DMM carbonylation reac-
tion. Based on the previous studies [12,35], a mechanism for
acid-catalyzed DMM carbonylation was proposed. As displayed in
Scheme 1, the acid-catalyzed DMM carbonylation reaction involves
the following elementary steps: DMM are initially protonated by
the protons from Brönsted acids to form methoxymethyl species,
which are then nucleophilically attacked by CO molecules, form-
ing methoxyacetyl species. The methoxyacetyl species further react
with another DMM to produce MMAc and the methoxymethyl
species are simultaneously regenerated. Like the typical Koch-
type carbonylation reactions (i.e. acid-catalyzed carbonylation

S. Liu et al. / Applied Catalysis A: General 497 (2015) 153–159
159
reactions), the reaction between methoxymethyl species and CO
affording the methoxyacetyl species has been proved to be the
rate-determining step in DMM carbonylation reaction. According
to the literatures, the transition state for the carbonyltion step can
be described as an association of CO with the methoxymethyl car-
bocation [10,36]. And, it is well known that increasing the acid
strength is beneficial for the stabilization of carbenium ions (e.g.
methoxymethyl carbocations), which leads to the decrease in the
activation barrier for the carbonylation step [37–39]. Although the
confinement effect of zeolite channels has been demonstrated to
stabilize the carbenium ions and hence promote the catalytic reac-
tivity of acid-catalyzed carbonylation reactions [13,40–42], the acid
strength of Nafion-silica catalyst is considerably higher than that of
zeolite catalysts, making the acid strength be the dominant factor
governing the reactivity of DMM carbonylation. As a result, Nafion-
silica catalyst with higher acid strength presents higher specific
catalytic activity for the DMM carbonylation reaction compared to
zeolite catalysts. In addition to high activity, it should be noted
that Nafion-silica catalyst exhibits the highest selectivity of MMAc
among these catalysts, nearly 90%, but the selectivities of MMAc
over H-Y and H-Beta zeolites are 80% and 65%, respectively. The
difference in the selectivity of MMAc is ascribed to the structures
of different types of catalysts. Bell and co-workers have demon-
strated that the selectivity of MMAc is significantly influenced by
the framework of zeolites, because the steric constraint of small
pore channel would promote the disproportion of DMM [16]. As a
result, high-silica H-Beta zeolite shows much lower MMAc selectiv-
ity than H-Y zeolite which possesses large super cages. In contrast
with zeolite catalysts, the pore diameter of Nafion-silica catalyst is
in the mesoporous range, much larger than that of zeolite catalysts.
Thus, the disproportion of DMM is heavily suppressed, enabling to
the highly selective formation of MMAc.
promise for producing MMAc and its derivatives, such as EG, 2-
methoxyethanol and so on.
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.apcata.
2015.03.010.
References
[1] H. Yue, Y. Zhao, X. Ma, J. Gong, Chem. Soc. Rev. 41 (2012) 4218–4244.
[2] http://www.ptq.pemex.com/productosyservicios/eventosdescargas/
Documents/Foro%20PEMEX%20Petroqu%C3%ADmica/2012/pci%20oxido%
20de%20etileno.pdf
[3] D.J. Loder, US Patent 2,152,852 (1939).
[4] A.T. Larson, US Patent 2,153,064 (1939).
[5] S. Suzuki, US Patent 5,72,780 (1975).
[6] S.Y. Lee, J.C. Kim, J.S. Lee, Y.G. Kim, Ind. Eng. Chem. Res. 32 (1993) 253–259.
[7] S.A.I. Barri, D. Chadwick, Catal. Lett. 141 (2011) 749–753.
[8] Y. Sun, H. Wang, J.H. Shen, H.C. Liu, Z.M. Liu, Catal. Commun. 10 (2009) 678–681.
[9] S.B. Dake, R.V. Chaudhari, J. Chem. Eng. Data 30 (1985) 400–403.
[10] F.E. Celik, T.-J. Kim, A.T. Bell, Angew. Chem. Int. Ed. 48 (2009) 4813–4815.
[11] D.E. Hendriksen, Abstr. Pap. Am. Chem. Soc. 185 (1983) 176–190.
[12] F.E. Celik, T. Kim, A.N. Mlinar, A.T. Bell, J. Catal. 274 (2010) 150–162.
[13] V. Shapovalov, A.T. Bell, J. Phys. Chem. C 114 (2010) 17753–17760.
[14] B.L. Booth, T.A. El-Fekky, J. Chem. Soc. Perkin Trans. 1 (1979) 2441–2446.
[15] S.D. Pirozhkov, A.S. Stepanyan, T.N. Myshenkova, M.B. Ordyan, A.L. Lapidus,
Bull. Acad. Sci. USSR Div. Chem. Sci. 31 (1982) 1852–1858.
[16] F.E. Celik, T.-J. Kim, A.T. Bell, J. Catal. 270 (2010) 185–195.
[17] M.A. Harmer, Q. Sun, Appl. Catal. A 221 (2001) 45–62.
[18] B. Török, I. Kiricsi, Á. Molnár, G.A. Olah, J. Catal. 193 (2000) 132–138.
[19] S. Wang, J.A. Guin, Energy Fuels 15 (2001) 666–670.
[20] F. Martínez, G. Morales, A. Martín, R. Van Grieken, Appl. Catal. A 347 (2008)
169–178.
[21] M.A. Harmer, Q. Sun, A.J. Vega, W.E. Farneth, A. Heidekum, W.F. Hoelderich,
Green Chem. 2 (2000) 7–14.
[22] M.A. Harmer, W.E. Farneth, Q. Sun, J. Am. Chem. Soc. 118 (1996) 7708–7715.
[23] P. Botella, A. Corma, J. López-Nieto, J. Catal. 185 (1999) 371–377.
[24] Q. Sun, M.A. Harmer, W.E. Farneth, Ind. Eng. Chem. Res. 36 (1997) 5541–5544.
[25] A. Heidekum, M. Harmer, W. Hoelderich, Catal. Lett. 47 (1997) 243–246.
[26] M.J. Climent, A. Corma, S. Iborra, S. Martínez-Silvestre, A. Velty, ChemSusChem
6 (2013) 1224–1234.
4. Conclusion
We have demonstrated that Nafion-silica composite catalysts
prepared by sol–gel method were very efficient and selective
for catalyzing DMM carbonylation to produce MMAc in vapor-
phase. Comparing with the pure Nafion resin, the excellent activity
of Nafion-silica composite materials is apparently attributed to
the increased accessibility of active site, ensured by the porous
structure and large surface area of the composite material. The
mesoporous structure of composite material not only enhances
the catalytic efficiency of Nafion resin but also helps suppress the
disproportion of DMM, resulting in a higher selectivity of MMAc
compared to the microporous acid zeolite catalysts. Furthermore,
the catalytic performance of composite catalysts was significantly
influenced by the amount of Nafion resin loading. When increasing
the loading of Nafion resin to 40%, a nearly complete conversion of
DMM (95%) was observed. Additionally, the type of silica sources
was found to influence the catalytic reactivity. Sodium silicate
derived Nafion-silica sample exhibited a higher conversion than
others. The excellent catalytic performance in the DMM carbony-
lation reaction shows that Nafion-silica composite has significant
[27] M.A. Harmer, Q. Sun, US Patent 5,824,622 (1999).
[28] A. Lassoued, C. Lalo, J. Deson, P. Batamack, J. Fraissard, M. Harmer, D. Corbin,
Chem. Phys. Lett. 303 (1999) 368–372.
[29] R. Pierotti, J. Rouquerol, Pure Appl. Chem. 57 (1985) 603–619.
[30] H. Lim, M. Yarmo, N. Huang, P. Khiew, W. Chiu, J. Phys. Sci. 20 (2009) 23–36.
[31] A. Zheng, S.-J. Huang, Q. Wang, H. Zhang, F. Deng, S.-B. Liu, Chin. J. Catal. 34
(2013) 436–491.
[32] A. Zheng, S.-J. Huang, W.-H. Chen, P.-H. Wu, H. Zhang, H.-K. Lee, L.-C.d. Ménor-
val, F. Deng, S.-B. Liu, J. Phys. Chem. A 112 (2008) 7349–7356.
[33] J. Guan, X. Li, G. Yang, W. Zhang, X. Liu, X. Han, X. Bao, J. Mol. Catal. A: Chem.
310 (2009) 113–120.
[34] S.-J. Huang, Y.-H. Tseng, Y. Mou, S.-B. Liu, S.-H. Huang, C.-P. Lin, J.C. Chan, Solid
State Nucl. Magn. Reson. 29 (2006) 272–277.
[35] P. Cheung, A. Bhan, G. Sunley, D. Law, E. Iglesia, J. Catal. 245 (2007) 110–123.
[36] H. Hogeveen, The Reactivity of Carbonium Ions Towards Carbon Monoxide,
Elsevier, Amsterdam, 1973.
[37] A.J. Jones, R.T. Carr, S.I. Zones, E. Iglesia, J. Catal. 312 (2014) 58–68.
[38] R.T. Carr, M. Neurock, E. Iglesia, J. Catal. 278 (2011) 78–93.
[39] D.A. Simonetti, R.T. Carr, E. Iglesia, J. Catal. 285 (2012) 19–30.
[40] A. Bhan, A.D. Allian, G.J. Sunley, D.J. Law, E. Iglesia, J. Am. Chem. Soc. 129 (2007)
4919–4924.
[41] A. Bhan, E. Iglesia, Acc. Chem. Res. 41 (2008) 559–567.
[42] H.J. Fang, A.M. Zheng, J. Xu, S.H. Li, Y.Y. Chu, L. Chen, F. Deng, J. Phys. Chem. C
115 (2011) 7429–7439.