Identifying the Active Silver Species in Carbonylation of Dimethyl Ether over Ag?HMOR

Article abstract of DOI:10.1002/cctc.202000533

As one of the key steps in ethanol production from syngas, dimethyl ether (DME) carbonylation to methyl acetate (MA) catalyzed by zeolite has drawn much attention. H-mordenite (HMOR) modified with metal has been continuously developed for improving catalytic performance. Here, we investigated the role of Ag species through three series of Ag?HMOR catalysts prepared through altering Ag loading, varying reduction temperature and additional ion exchange. TEM, XPS, CO FTIR and UV-Vis spectroscopy were employed to identify the nature and the amount of Ag species in xAg?HM with different Ag loading. Taken together with catalytic performance evaluation, 5Ag?HM with the moderate size of Ag⁰ species (Ag_n^δ+ and Ag_m clusters) was proved to have greatest promotion effect on DME carbonylation. This was also evidenced by further improved MA formation rate over catalysts via elevating reduction temperature or via additional NH₄Cl ion exchange, which contained more Ag⁰ and fewer Ag⁺ species respectively. These provide insight into metal-modified HMOR, inspiring design and fabrication of zeolite catalysts with multi-active sites.

Full text of DOI:10.1002/cctc.202000533

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Identifying the Active Silver Species in Carbonylation of Dimethyl
Ether over Ag-HMOR
Authors: Shiyue Li, Kai Cai, Ying Li, Shuaipeng Liu, Man Yu, Yue
Wang, Xinbin Ma, and Shouying Huang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202000533
Link to VoR: https://doi.org/10.1002/cctc.202000533
01/2020

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Identifying the Active Silver Species in Carbonylation of Dimethyl
Ether over Ag-HMOR
Shiyue Li,[a] Kai Cai,[a] Ying Li,[a] Shuaipeng Liu,[a] Man Yu,[a] Yue Wang,[a] Xinbin Ma,[a] and Shouying
Huang*[a]
[a]
S. Li, K. Cai, Dr. Y. Li, S. Liu, M. Yu, Dr. Y. Wang, Prof. Dr. X. Ma, Dr. S. Huang
Key Laboratory for Green Chemical Technology of Ministry of Education
Collaborative Innovation Center of Chemical Science and Engineering
School of Chemical Engineering and Technology
Tianjin University
Tianjin 300072 (P. R. China)
E-mail: huangsy@tju.edu.cn
Abstract: As one of the key steps in ethanol production from syngas,
dimethyl ether (DME) carbonylation to methyl acetate (MA) catalyzed
by zeolite has drawn much attention. H-mordenite (HMOR) modified
with metal has been continuously developed for improving catalytic
performance. Here, we investigated the role of Ag species through
three series of Ag-HMOR catalysts prepared through altering Ag
loading, varying reduction temperature and additional ion exchange.
TEM, XPS, CO FTIR and UV-Vis spectroscopy were employed to
identify the nature and the amount of Ag species in xAg-HM with
different Ag loading. Taken together with catalytic performance
evaluation, 5Ag-HM with the moderate size of Ag0 species (Agnδ+ and
Agm clusters) was proved to have greatest promotion effect on DME
carbonylation. This was also evidenced by further improved MA
formation rate over catalysts via elevating reduction temperature or
via additional NH4Cl ion exchange, which contained more Ag0 and
fewer Ag+ species respectively. These provide insight into metal-
modified HMOR, inspiring design and fabrication of zeolite catalysts
with multi-active sites.
impregnated with noble metals such as Pt, Pd, Ru also improved
the efficiency of DME carbonylation in the dual-stage ethanol
production, and the combination of Pt/HMOR+Cu/ZnO showed
the highest ethanol yield.[8] Furthermore, Cu-Zn and Cu-Pt bimetal
modified HMOR raised the reaction stability as well as the
catalytic activity.[9]
Determining the active metal species is of great significance for
fabricating metal-decorated zeolite catalysts and improving
reaction efficiency. However, identification of active metal species
and their role in catalytic mechanism are still controversial in DME
carbonylation. Blasco et al.[5] reported Cu+ cations adjacent to
Brønsted acid sites formed Cu+-CO and promoted the MA
formation. Ma’s group [6a,10] proposed that the rate of MA formation
was positively correlated with the amount of Cu0 species through
experiments and DFT calculations. This was because Cu0 site
and neighboring Brønsted acid functioned as synergistic catalytic
site, which changed the mode of DME dissociation and reduced
the apparent activation energy. This was also evidenced by solid-
state NMR results of Deng’s group,[11] which demonstrated that
Cu+-CO adsorption was too strong to participate in MA or acetic
acid formation. Wang et al.[7b] claimed that the oxidized metal ions
(e.g. Cu2+, Ni2+, Zn2+) as Lewis acid sites facilitated CO adsorption
and activation. Besides, regarding bimetallic catalysts such as
Cu-Zn/HMOR, Zn played a role in dispersing and stabilizing Cu
species, which increased active sites and raised sintering
resistance.[9a]
Introduction
Ethanol production, integrating dimethyl ether (DME)
carbonylation to methyl acetate (MA) and MA hydrogenation to
ethanol, is an environment-friendly, economical and highly
efficient route.[1] DME carbonylation is the critical step in this
process, which can be catalyzed by solid acids, such as zeolites
with Brønsted acid (e.g. HMOR, HFER, EU-12)[2] and
heteropolyacids.[3] HMOR possesses unique 8-membered ring (8-
MR) side pockets parallel to the b axis that have a spatial
confinement effect, giving the high MA selectivity of 99% with
considerable activity under mild conditions.[4] The exploit of zeolite
catalysts, particularly HMOR makes the process more promising
in industry.
In this work, Ag with simple valence is selected as modifier from
Group IB. Various treatments were used to adjust the content and
properties of Ag species. The affiliation of species and their
impact on catalytic performance were discussed in combination
with TEM, N2 physical adsorption and desorption, XPS, CO-IR,
and UV-Vis, providing the possibility to clarify the metal promotion
mechanism.
There have been studies of transition metal modified HMOR in
DME carbonylation. The catalytic performance can be
significantly improved as transition metals (e.g. Cu, Zn, Co, Pt, Pd,
etc.) were introduced into HMOR system. Originally, using IR
operando spectroscopy and in-situ magic-angle spinning NMR
spectroscopy, the promotion effect of Cu-HMOR was elucidated
in methanol carbonylation.[5] The similar enhancement in activity
caused by Cu modification was then also found in DME
carbonylation.[6] Co, Zn, Ni-exchanged HMOR improved the
catalytic activity and stability, which was attributed to the
preferential locations of metal cations at different sites.[7] HMOR
Results and Discussion
Catalytic properties
The initial activity and selectivity (time on stream of 1 h) on the
catalysts with different Ag loading are presented in Figure 1,
because the zeolite catalysts show an induction period and suffer
a rapid deactivation in DME carbonylation that has been reported
in previous studies. It is evidently shown that as Ag content grows,
the STY of MA firstly increases and reaches the highest value at
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
393 g‧kgcat-1‧h-1 on 5Ag-HM sample, which is significantly
increased by 45% comparing with that of the parent HM sample.
Then the MA yield drops down to 330 g‧kgcat-1‧h-1 on 7Ag-HM but
still higher than HM. Thanks to the spatial confinement effect of
unique channels in MOR,[13] the outstanding selectivity of MA on
all samples are observed, suggesting Ag modification has
negligible impact on selectivity.
The changes in conversion of DME and selectivity of MA and
byproducts over all the catalysts with time on stream were
presented in Supporting Information (SI), Figure S1 and S2.
Both activity and selectivity decrease with time, but the Ag
modification shows very limited effect on deactivation rate and
selectivity. We also investigated the 5Ag-HMcal, which was not
reduced in comparison with 5Ag-HM. As shown in Figure S3,
5Ag-HMcal exhibits much lower activity than 5Ag-HM even HM,
confirming the reduction is prerequisite for improving activity.
Given that the amount of Brønsted acid sites in 8-MR channels
linearly correlates with the rate initial MA formation in pure H-
MOR,[13-14] we quantified the number of Brønsted acid sites
located within 8-MR and 12-membered ring (12-MR) to explicitly
elucidate the function of Ag species.
Figure 1. Yield and selectivity of MA on HM and xAg-HM at time on stream of
1 h. Reaction conditions: 6000 h-1, 473 K and 1.5 MPa.
Table 1. The quantity of acid sites, activity and composition of samples with different Ag loading.
Sample
Btotal
(μmol‧g-1)[a]
B12-MR
(μmol‧g-1)[b]
B8-MR
(μmol‧g-1)[c]
L12-MR
(μmol‧g-1)[b]
STY
(g‧kgcat-1‧h-1)
FRB
(h-1)[d]
Si/Al
(molar ratio)[e]
Ag loading
(wt.%)[e]
HM
1263
1229
1219
1124
1104
572
538
531
525
512
691
691
688
599
591
46
31
33
31
27
271
286
362
393
330
5.3
5.6
7.1
8.9
7.5
5.8
6.0
6.1
6.4
6.4
——
2.1
3.2
5.3
7.2
2Ag-HM
3Ag-HM
5Ag-HM
7Ag-HM
[a], [b] Calculated based on IR spectra of NH3 adsorption and pyridine adsorption, respectively. [c] B8-MR=Btotal-B12-MR. [d] Calculated via Equation (1). [e] Measured
by ICP-OES.
The sizes of 12-MR and 8-MR channels in MOR are 0.67 × 0.70
Particularly, the downward trend becomes more pronounced for
nm and 0.26 × 0.57 nm respectively. In terms of spatial constrains, the B8-MR when Ag loading is up to 5% and 7%. Previous studies
NH3 as a probe molecule is used to access all the acid sites in the
HM and xAg-HM (x=2, 3, 5, 7) catalysts, while pyridine with a
larger kinetic diameter of 0.48 nm is used to selectively detect the
acid sites in 12-MR.[15] According to integral peak areas of the
bands at 1430 cm-1 in NH3-adsorption IR and 1540 cm-1 in
pyridine-adsorption IR (Figure S4), we calculated the amounts of
total Brønsted acid sites (Btotal) and Brønsted acid sites in 12-MR
(B12-MR) by using the respective extinction coefficients in previous
have claimed that the strength of Brønsted acid in 8-MR is
stronger than that in 12-MR using ammonia as a probe.[20] This
implies that the negative framework Al atoms located in 8-MR
might be more readily attracted by H+ in solution, resulting in
severer dealumination. In addition, the particle Ag loading is a little
higher than theoretical value, which supports the partial resolution
of MOR matrix. To eliminate the influence of active site amounts,
we normalized the MA formation rate based on the numbers of B8-
MR in HM and xAg-HM samples. As shown in Table 1 (column 7,
FRB), there is a positive correlation between FRB and Ag content
except 7Ag-HM. It means that introduction of Ag species
promotes catalytic performance of H-MOR, while excess Ag
loading causes the STYMA and FRB to drop at the same time,
which will be explained in detail later.
studies.[16] The number of Brønsted acid sites within 8-MR (B8-MR
)
is obtained by subtracting B12-MR from Btotal. As summarized in
Table 1, an obvious decrease of B12-MR with maintaining B8-MR is
obtained on 2Ag-HM, suggesting that the Ag+-Si(O)Al species still
exist after reduction and they preferentially locate within 12-MR.
This phenomenon is in line with the previous work on Cu-
exchanged MOR.[17] Further increasing Ag loading, both B8-MR and
B12-MR decrease as a result of migration of Ag species from 12-
MR to 8-MR channels after reduction, which is also discussed in
the literature.[6a,18] Note that AgNO3 ion-exchange causes slight
dealumination of MOR framework, as evidenced by the
decreased Si/Al molar ratio (Table 1, ICP results), which is
probably caused by the weak acidity of AgNO3 aqueous solution.
This provides another explanation about the decrease of acid
amounts both in 12-MR and 8-MR from HM to xAg-HM.[19]
State and nature of Ag species in MOR
In the XRD patterns (Figure 2), characteristic diffraction peaks of
the as-prepared catalysts (xAg-HM) are sharp and symmetrical,
confirming the high crystallinity for these samples. The diffraction
peaks at 2θ of 38° and 44.5° are attributed to the Ag (111) and
(200) crystal plane[21] and their intensities significantly raise as the
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Ag loading increases, which is indicative of gradual agglomeration
of Ag nanoparticles.
Table 2. Textural properties of HM and xAg-HM samples.
Sample
Specific surface area (m2‧g-1)
Pore volume (cm3‧g-1)
SBET
602
522
521
503
505
Smicro
588
515
511
491
493
Sexternal
14
Vtotal
Vmicro
0.222
0.194
0.193
0.186
0.187
Vmeso
0.027
0.017
0.021
0.023
0.023
HM
0.249
0.211
0.214
0.209
0.210
2Ag-HM
3Ag-HM
5Ag-HM
7Ag-HM
7
10
12
12
Figure 2. XRD profiles of HM and xAg-HM samples.
Figure 3. TEM images of 2Ag-HM (a), 5Ag-HM (b), 7Ag-HM (c) and the lattice fringe of Ag0 in 5Ag-HM (d).
The nitrogen adsorption–desorption isotherms (Figure S5) are
type I according to IUPAC classification. It indicates that the
micropore structure of zeolites are well preserved, which is in
agreement with XRD results. A tiny hysteresis loop occurs when
the relative pressure P/P0 is greater than 0.5, indicating the
presence of a small amount of mesopore. Compare to HM, the N2
adsorption quantity of xAg-HM reduces obviously at rather lower
relative pressure P/P0. Correspondingly, the calculated specific
surface area and micropore volume have a distinct drop, as
collected in Table 2, and this reveals Ag species locate in the
micropore of H-MOR. Further increasing Ag loading, the
micropore volume and specific surface area decline slightly,
which implies severer aggregation probably occurred on external
surface. TEM images of xAg-HM exhibit two different populations
of Ag nanoparticles (Figure 3). As shown in the inserted images,
some large Ag nanoparticles with size from 10 nm to 150 nm
appear, which mainly situate on the external surface of zeolite
crystals. This provides an explanation for the significant peaks of
Ag0 in XRD patterns. Meanwhile, the uniform-dispersed Ag
particles could be clearly observed on the section of the samples,
endorsing they exist inside the zeolite pores. Particle size analysis
was performed by measuring more than 200 nanoparticles, and
the mean sizes of xAg-HM (x=2, 5, 7) are 1.26 nm, 2.18 nm and
2.24 nm severally. The lattice fringe seen in Figure 3 (d) reveals
that the nanoparticles mainly settle as metallic Ag rather than
AgxOy, which is consistent with XRD results. Besides, some much
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Table 3. XPS results and modified Auger parameters.
Sample
BE Ag 3d5/2 (eV)
KE Ag M4N45N45 (eV)
α[a]
́
2Ag-HM
3Ag-HM
5Ag-HM
7Ag-HM
368.6
368.5
368.5
368.5
355.9
357.3
357.7
357.6
724.5
725.8
726.2
726.1
[a] Modified Auger parameter ά =BE �Ag 3d5 2�+KE (Ag M4N45N45).
⁄
has been significantly improved and Ag exists mainly as Ag0
species in xAg-HM(x=3, 5, 7).
CO, as a type of probe molecule, is widely used for detecting
metal species.[24] IR studies of CO adsorption on the reduced
samples have been given in Figure 5 (a). From 2Ag-HM to 7Ag-
HM, the peak at about 2189 cm-1 has shown a downward trend
opposite to incremental Ag content. Using CO-IR approach,
researchers have been able to confirm that Ag+(CO)
monocarbonyls band of Ag-exchanged ZSM-5 centers at 2190
cm-1.[25] Besides, Ag+(CO) bands were reported at 2184 and 2170
cm-1 in AgX catalyst that were originated from S and S sites
Ⅱ
Ⅲ
and mainly at 2175 cm-1 in AgY catalyst respectively.[26] However,
there are quite few pieces of literature for Ag/H-MOR catalyst.
Thus, we present the spectra of xAg-HMoxy as the reference, in
which all Ag species are proposed exclusively as oxidized Ag.[25]
Obviously, an intense band at about 2189 cm-1 appears in all xAg-
HMoxy samples, related to Ag+(CO) in MOR (Figure 5 (c)).
Accordingly, we can confirm that the peak at 2189 cm-1 in xAg-HM
is attributed to the Ag+(CO), excluding formation of stable Ag0
(CO) species. We calculated the integral intensities of the
Ag+(CO) band upon xAg-HMoxy samples and established a linear
relationship between peak area and Ag content (see Figure 5
(d)).Based on this correlation, we can quantify the amounts of Ag+
and Ag0 species in xAg-HM catalyst, which are listed in Table 4.
It is noteworthy that the number of Ag+ ions decreases sharply
from 2Ag-HM to 3Ag-HM and gradually becomes steady with
further increasing Ag content, which accords with XPS results.
This indicates a small number of Ag species possess strong
interaction with framework oxygen atoms that inhibits their
reduction to metallic clusters, while excess Ag are easily to be
reduced and aggregated because of weaker affinity with zeolite
framework. Taking these findings together with MA formation rate
(FRB results, Table 1), we propose that Ag+ species have no
contribution to promoting H-MOR catalytic activity. Additionally,
7Ag-HM with the most Ag0 species shows a lower reaction activity,
probably due to the agglomeration and migration of large Ag
nanoparticles. For Ag/ZSM-5 and Ag/USY zeolites, only a very
weak band at about 2160 cm-1 with a predominant π-character is
formed between Ag0 and CO at room temperature.[25-26] For
Ag/SiO2 system, the band at 2162 cm-1 at 153 K is assigned to
Agnδ+-CO species, but there is no band in the region of 2020-2060
cm-1 that is assigned to CO adsorbed on metallic Ag such as Ag
foil.[27] Generally, the Ag0-CO is not stable and is mostly detected
at 90 K.[28] Here, we extend the region of CO-IR spectra and add
Figure S6 into the SI. Only an extremely weak band appears at
around 2160 cm-1 for 3Ag-HM zeolite, while this band is hardly
distinguished for other xAg-HM. It indicates that the amount of
Figure 4. XPS spectra of xAg-HM: Ag 3d core level transitions (a) and Ag MNN
transitions (b).
larger Ag particles are observed on the catalyst due to readily
agglomeration of Ag species in H2 reduction. Given the essence
of space-confined and Brønsted acid sites catalyzed mechanism,
we think that it is reasonable to neglect the teeny contribution of
large Ag particles outside the pores to MA formation.
For the purpose of investigating the chemical states of Ag species
in MOR structure, the Ag 3d photoelectron spectra and Ag MNN
Auger spectra of xAg-HM were obtained and presented in Figure
4. In the Ag 3d region the peaks at 374.5 eV and 368.5 eV
correspond to Ag 3d3/2 and Ag 3d5/2 (Figure 4 (a)). Because of the
small difference between Ag0 and Ag+ in the binding energies,[22]
it is pretty difficult to discriminate the Ag electronic states only via
analyzing Ag 3d transitions. Taken account into a relatively larger
shift of Ag Auger transitions, we plotted the Auger spectra in the
Ag MNN region as a function of kinetic energy (KE) and calculated
the modified Auger parameter (ά ), according to ref [22b]. In case
of the lowest Ag loading, the peak of Ag M4N45N45 of 2Ag-HM is
355.9 eV. With increasing the Ag loading, this peak position
apparently shifts towards higher binding energy, indicating the
changes in Ag valence state. The band positions and the
calculation of parameters ά are summarized in Table 3. Previous
studies have shown the parameter of around 724.0 eV for Ag+
ions and that of 726.0 eV for Ag0 species.[23] Therefore, the
calculation value of α
there are still considerable Ag+ ions in 2Ag-HM even after
reduction. As Ag loading increases, the α value moves from 724.5
eV to around 726.0 eV suggesting that the degree of Ag reduction
́ is 724.5 eV for 2Ag-HM suggesting that
δ+
Agn cluster in 3Ag-HM is more than that in others, which is
consistent with UV-Vis results. Besides, no other significant IR
́
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Figure 5. CO-IR spectra of xAg-HM (a), UV-Vis profiles of xAg-HM (b), CO-IR spectra of xAg-HMoxy (c) and the calibration of Ag+(CO) band in CO-IR spectra (d).
to Agnδ+ clusters and Agm clusters.[29] Meanwhile, the broad band
Table 4. The results of xAg-HM and 5Ag-HM-A samples from CO-IR.
at 340-480 nm is associated with metallic Ag, which is a plasma
Sample
Total Ag
content
Integral
Ag+ content Ag0 content
Ag0/
resonance absorption band.[30] It can be clearly observed that four
kinds of Ag species (i.e. Ag+, Agnδ+ clusters, Agm clusters and Ag
nanoparticles) co-existed in xAg-HM samples (Figure 5 (b)), as
evidenced by four bands at 219 nm, 264 nm, 307-320 nm and
350-450 nm. Compare to 2Ag-HM, 3Ag-HM shows a more intense
intensity of (μmol‧g-1)[c] (μmol‧g-1)[c] (Ag++Ag0)
(μmol‧g-1)[a] Ag+(CO)
(%)
(cm-1)[b]
2Ag-HM
3Ag-HM
5Ag-HM
7Ag-HM
195
301
499
671
3.15
1.07
1.09
0.62
0.69
103
35
36
20
23
92
47
88
93
97
95
δ+
band at 264 nm, indicative of more Agn species in 3Ag-HM.
Further increasing Ag content, the band from 307 nm shifts toward
320 nm and the intensity becomes higher, accompanied by a
decline in intensity of the band at 264 nm. This observation
apparently suggests that Ag species tend to agglomerate with
loading, which is consistent with the increased average particle
size noticed in Figure 3. As mentioned above, 5Ag-HM performs
the best activity in DME carbonylation, therefore a certain amount
266
463
651
449
5Ag-HM-A 472
δ+
of moderately agglomerated Ag clusters such as Agn and Agm
clusters have the most significant benefit for DME conversion.[29b]
[a] Measured by ICP-OES. [b] Measured by CO-IR. [c] Ag+=Integral intensity of
Ag+(CO)/(2.84 cm-1×MAg)×104, MAg=107.9 g‧mol-1; Ag0=Total Ag content-Ag+
content.
Supportive evidence for the role of Ag
feature about Ag0 (CO) is observed at room temperature in the
range of 2020-2060 cm-1.
UV-Vis spectroscopy is widely used to describe the states of Ag
species namely ions, clusters and nanoparticles by means of
characteristic bands. The bands at 190-230 nm are attributed to
4d10-4d9s1 transitions in Ag+ ions, which is overlapped with the
band of MOR framework. The bands at 240-326 nm are assigned
Considering that Ag clusters are the most probably active species
and there is still a certain portion of Ag+-Si(O)Al species in xAg-
HM after reduction at 673 K, we conducted two experiments to
further demonstrate the role of different Ag species. Firstly,
increasing the reduction temperature was applied as described in
the Experimental section. The catalytic performance and the
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Figure 6. Yield and selectivity of MA on samples at time on stream of 1 h. Reaction conditions: 6000 h-1, 473 K and 1.5 MPa. Small graph in Figure 6 (b) is CO-IR
spectra of 5Ag-HM and 5Ag-HM-A.
Table 5. The quantity of acid sites and activity of HM and 5Ag-HM samples with different treatment.
Sample
Btotal (μmol‧g-1)[a]
B12-MR (μmol‧g-1)[b]
B8-MR (μmol‧g-1)[c]
L12-MR (μmol‧g-1)[b]
STY (g‧kgcat-1‧h-1)
FRB (h-1)[d]
HM
1263
1124
1148
1190
1151
1105
572
525
536
564
563
470
691
599
612
626
588
635
46
31
28
33
32
39
271
393
473
492
474
429
5.3
5Ag-HM
8.9
5Ag-HM(723)
5Ag-HM(748)
5Ag-HM(773)
5Ag-HM-A
10.4
10.6
10.9
9.1
[a], [b] Calculated based on IR spectra of NH3 adsorption and pyridine adsorption, respectively. [c] B8-MR=Btotal-B12-MR. [d] Calculated via Equation (1).
amount of Brønsted acid sites are shown in Figure 6 (a), Figure
S7 and Table 5. As the reduction temperature is elevated, the
Likewise, another method that is replacing residual Ag+ species
by ion exchange with NH4Cl, was also applied to elucidate the role
of different Ag species (Figure 6 (d)). Figure 6 (b) provides a
catalytic graph of 5Ag-HM with and without NH4Cl treatment and
the former displays better activity. After once reduction of the Ag-
exchanged HMOR catalyst at 673 K, in order to further drop Ag+
species left over, Cl- from NH4Cl aqueous solution was to
precipitate Ag+ linked to Si(O-)Al. As shown in Figure 6 (b) and
Table 4, the value of Ag+(CO) in 5Ag-HM-A obviously declines
after NH4Cl treatment. Even so, 5Ag-HM-A is still shown similar
FRB in the condition of descending Ag+ species, which further
excludes the Ag+ species from role as activity promoter.
amounts of Brønsted acid sites (including total, B8-MR and B12-MR
)
increase gradually in comparison with that of 5Ag-HM. This
demonstrates that the improved reduction degree of Ag species
is accompanied by the regeneration of Brønsted acid sites. At
highest reduction temperature (773 K), the density of acid sites
decline again because the severe agglomeration of Ag would
cover the acid sites or block the microporous channels that reduce
the accessibility of probe molecules to the sites. Meanwhile, the
activity increases at first and then has a decrease at 773 K. To
exclude the influence of the changes in acidity, we normalized the
value of FRB and found that there is a positive correlation between
FRB and reduction temperature, indicating that more reduced Ag
species work with B8-MR synergistically with increasing reduction
degree. This ensures the identity of Ag clusters as active species.
Conclusion
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Nitrogen adsorption–desorption isotherms were carried out at 77 K on a
Micromeritics ASAP 2460 device. The porous structure information
including specific surface area, volume and pore size were calculated
using Brunauer-Emmett-Teller (BET) method, t-plot method and Horvath-
Kawazoe equation. Prior to adsorption, all samples were reduced at 673
K for 3 h and were outgassed at 623 K for 24 h.
Promotion of Ag in HMOR catalyzed DME carbonylation was
investigated on three series of Ag modified HMOR catalysts,
which are Ag-HM with different loading (xAg-HM), at different
reduction temperature (5Ag-HM(y)) and with an additional ion
exchange (5Ag-HM-A). It has been proved that all of the Ag
modified HMOR catalysts behave considerably enhanced activity
performance on DME carbonylation, in comparison with HMOR.
N2 adsorption and desorption and TEM demonstrate that
considerable Ag species locate both inside micropore and
mesopore, despite of existence of Ag agglomerates on external
surface. The results of XPS, CO-IR and UV-Vis demonstrate that
as Ag loading increases Ag species gradually change from Ag+ to
Ag0 with different particle sizes. Considering that on DME
carbonylation the Brønsted acid sites in 8-MR of HMOR are able
to catalyze DME carbonylation solely, basic molecules with
different size were used in-situ FTIR to exclude the effect of
changes in acidity. The results confirm that xAg-HM contain fairly
similar B8-MR with HM and what’s more critical is that the formation
rate of MA per B8-MR (FRB) shows a volcanic shape. Combining
The Ag content of samples as well as molar Si/Al ratio was determined by
Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES,
Varian Vista-MPX). Before measurements, all the samples were digested
into the HF solution and complexed by super-saturated H3BO3 solution and
later were diluted with constant volume method. Ag characteristic peak is
at 328.068 nm.
Transmission electron microscopy (TEM) images were collected by a JEM-
2100F electron microscope. All the reduced samples were sliced with
Leica EM UC7 Ultramicrotome and were dispersed ultrasonically in
ethanol for 30 min at room temperature and subsequently dropped on the
lacey-carbon-coated copper grids three drops of suspension per grid.
X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy
(AES) tests were measured by a Thermo Fisher Scientific ESCALAB Xi+
with monochromatic Al Kα (hν=1486.6 eV) X-ray radiation source. The
obtained binding energy was normalized by an internal reference of the Si
2p peak at 103.5 ± 0.1 eV.
δ+
with the above characterization results, Ag clusters (Agn and
Agm) with suitable size are identified to facilitate DME
carbonylation. This is further evidenced by improved activity on
5Ag-HM(y) and 5Ag-HM-A, which contain more Ag0 and fewer
Ag+ species respectively. This work enriches understanding about
metal-promoted zeolite for DME carbonylation and provides clues
for zeolite catalyst design. In future study, the fine structure of
metal cluster as well as the promotion mechanism should be
investigated.
Diffuse reflectance ultraviolet-visible (UV-Vis) measurements were
recorded on a UV-Vis Spectrometer (Shimadzu UV-2600) with the
intermediate speed in the range of 200–800 nm employing BaSO4 as
reference.
In-situ Fourier transform infrared (FTIR) spectra were conducted on
Thermo Scientific Nicolet 6700 equipment with a MCT/A detector. Alkaline
probe molecules (e.g. pyridine and NH3) and CO were employed to
determine the amounts of Brønsted acid sites and Ag+ species,
respectively. Prior to measurements, the xAg-HMcal sample (~20 mg) was
pretreated in the in-situ cell at 673 K in vacuum or reduced at 673 K with
a flow of 10% H2/Ar. Specially, the xAg-HMcal oxidized under the treatment
of 20% O2/Ar at 673 K was labeled as xAg-HMoxy. After that, the in-situ cell
was cooled to 423 K. IR spectra were obtained by 32 scans per spectrum
Experimental Section
Catalyst preparation
The commercial parent NH4-mordenite (NH4-MOR) zeolite was provided
by Yangzhou Zhonghe Petrochemical Institute Co., Ltd. Ag-exchanged
NH4-MOR was prepared by ion exchange method. Typically, 4 g NH4-MOR
mixed with 0.189 g AgNO3 and 100 mL deionized water was transferred
into the eggplant-shaped flask. After stirred at 303 K for 24 h, Ag-NH4-
MOR was filtered, washed with distilled water, dried overnight at 383 K and
calcined at 773 K for 4 h with a heat ramp of 2 K/min. The entire
preparation process was protected from light. The obtained catalyst was
labeled as 5Ag-HMcal. Before reaction and characterizations, 5Ag-HMcal
was reduced at 673 K for 3 h using 10% H2/Ar and was entitled 5Ag-HM.
Other samples with different loading (named xAg-HM, x=2, 3, 7) were
prepared in the same procedure as well. To investigate the effect of
reduction temperature, 5Ag-HM(y) (y=723, 748, 773) was pretreated by
10% H2/Ar at 723 K, 748 K and 773 K.
at a scanning resolution of 4 cm−1
.
Catalyst performance tests
The catalyst performance tests were carried out in a fixed-bed reactor
whose inner diameter was 8 mm. Basically, the samples (~500 mg, 40–60
mesh) were loaded into the tubular reactor with silica sand filled above and
below. Initially, the catalyst was pretreated at 673 K for 3 h with a flow of
100 mL/min 10% H2/Ar or N2 (only for HM sample) and then was exposed
to CO/DME reactant gases (CO/DME=1/49, vol./vol.) under the reaction
conditions of 473 K and 1.5 MPa. The products were dissected online
through
a coupled gas chromatograph (Shimadzu GC-2014C Gas
Chromatograph), provided with thermal conductivity detector (TCD) and
flame ionization detector (FID). The reactant conversion (XDME), product
selectivity (SMA) as well as product space-time yield (STYMA) was
confirmed as the former calculations,[6a, 12] while the MA formation rate on
per Brønsted acid sites in 8-MR (named FRB) was determined on the basis
of the following Equation (1):
5Ag-HM-A was prepared by the ion exchange with 3 g 5Ag-HM and 75 mL
0.05 mol/L NH4Cl aqueous solution at room temperature for 12 h.
Following this treatment, the catalyst was filtered, washed with plenty of
distilled water and dried overnight. Then 5Ag-HM-A was obtained by
calcination at 773 K for 4 h. Prior to undertaking the catalytic performance,
5Ag-HM-A was reduced at 573 K for 2 h with 10% H2/Ar.
-1
MA formation rate (mol∙h
)
FRB�h-1�= m
(1)
Characterization
-1
g ×Numberof Brønsted acid site in 8-membered ring (mol∙g
( )
)
cat
Powder X-ray diffraction (XRD) patterns were performed on a Rigaku
D/max-2500 diffractometer equipped with a Cu Kα (λ=0.154 nm) radiation
source at 40 kV and 200 mA. All patterns were operated from 3° to 90°
(2θ) with a scanning pace of 8°/min.
Acknowledgements
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
[29] a) A. N. Pestryakov, A. A. Davydov, J. Electron. Spectrosc. Relat.
Phenom. 1995, 74, 195-199; b) J. Shibata, J. Catal. 2004, 222, 368-376.
[30] K. Kubo, H. Iida, S. Namba, A. Igarashi, Appl. Catal. A: Gen. 2015, 489,
272-279.
We acknowledge the financial support by National Natural
Science Foundation of China (No. 21978209, 21325626) and
Postdoctoral Science Foundation of China (No. 2019M661021)
Keywords: Zeolites • Carbonylation• Heterogeneous catalysis •
Silver modification • Active species
[1]
[2]
E. S. Zhan, Z. P. Xiong, W. J. Shen, J. Energy Chem. 2019, 36, 51-63.
a) P. Cheung, A. Bhan, G. J. Sunley, E. Iglesia, Angew. Chem. Int. Ed.
2006, 45, 1617-1620; b) P. Feng, G. Zhang, X. Chen, K. Zang, X. Li, L.
Xu, Appl. Catal. A: Gen. 2018, 557, 119-124; c) X. Feng, J. Yao, H. Li, Y.
Fang, Y. Yoneyama, G. Yang, N. Tsubaki, Chem. Commun. 2019, 55,
1048-1051.
[3]
[4]
a) M. V. Luzgin, M. S. Kazantsev, G. G. Volkova, W. Wang, A. G.
Stepanov, J. Catal. 2011, 277, 72-79; b) H. Shen, Y. Li, S. Huang, K. Cai,
Z. Cheng, J. Lv, X. Ma, Catal. Today 2019, 330, 117-123.
a) M. Boronat, C. Martínezsánchez, D. Law, A. Corma, J. Am. Chem.
Soc. 2008, 130, 16316-16323; b) Y. Chu, A. Y. Lo, C. Wang, F. Deng, J.
Phys. Chem. C 2019, 123, 15503-15512.
[5]
[6]
T. Blasco, M. Boronat, P. Concepcion, A. Corma, D. Law, J. A. Vidal-
Moya, Angew. Chem. Int. Ed. 2007, 46, 3938-3941.
a) H. Zhan, S. Huang, Y. Li, J. Lv, S. Wang, X. Ma, Catal. Sci. Technol.
2015, 5, 4378-4389; b) Y. Zhang, C. Ding, J. Wang, Y. Jia, Y. Xue, Z.
Gao, B. Yu, B. Gao, K. Zhang, P. Liu, Catal. Sci. Technol. 2019, 9, 1581-
1594.
[7]
a) M. Ma, E. Zhan, X. Huang, N. Ta, Z. Xiong, L. Bai, W. Shen, Catal.
Sci. Technol. 2018, 8, 2124-2130; b) S. Wang, W. Guo, L. Zhu, H. Wang,
K. Qiu, K. Cen, J. Phys. Chem. C 2014, 119, 524-533.
[8]
[9]
P. Lu, G. Yang, Y. Tanaka, N. Tsubaki, Catal. Today 2014, 232, 22-26.
a) A. A. C. Reule, N. Semagina, ACS Catal. 2016, 6, 4972-4975; b) H.
Sheng, H. Ma, W. Qian, N. Fei, H. Zhang, W. Ying, Energ. Fuel. 2019,
33, 10159-10166.
[10] Y. Li, S. Huang, Z. Cheng, S. Wang, Q. Ge, X. Ma, J. Catal. 2018, 365,
440-449.
[11] L. Zhou, S. Li, G. Qi, Y. Su, J. Li, A. Zheng, X. Yi, Q. Wang, F. Deng,
Solid State Nucl. Magn. Reson. 2016, 80, 1-6.
[12] M. Wang, S. Huang, J. Lü, Z. Cheng, Y. Li, S. Wang, X. Ma, Chinese. J.
Catal. 2016, 37, 1530-1537.
[13] A. Bhan, A. D. Allian, G. J. Sunley, D. J. Law, E. Iglesia, J. Am. Chem.
Soc. 2007, 129, 4919-4924.
[14] B. Li, J. Xu, B. Han, X. Wang, G. Qi, Z. Zhang, C. Wang, F. Deng, J.
Phys. Chem. C 2013, 117, 5840-5847.
[15] T. He, X. Liu, S. Xu, X. Han, X. Pan, G. Hou, X. Bao, J. Phys. Chem. C
2016, 120, 22526-22531.
[16] C. A. Emeis, J. Catal. 1993, 141, 347-354.
[17] S. Grundner, W. Luo, M. Sanchez-Sanchez, J. A. Lercher, Chem.
Commun. 2016, 52, 2553-2556.
[18] Y. Wang, H. Wu, Q. Zhang, Q. Tang, Microporous Mesoporous Mater.
2005, 86, 38-49.
[19] A. A. C. Reule, J. A. Sawada, N. Semagina, J. Catal. 2017, 349, 98-109.
[20] M. Niwa, K. Suzuki, N. Katada, T. Kanougi, T. Atoguchi, J. Phys. Chem.
B 2005, 109, 18749-18757.
[21] N. Popovych, P. Kyriienko, S. Soloviev, R. Baran, Y. Millot, S. Dzwigaj,
Phys. Chem. Chem. Phys. 2016, 18, 29458-29465.
[22] a) S. G. Aspromonte, R. M. Serra, E. E. Miró, A. V. Boix, Appl. Catal. A:
Gen. 2011, 407, 134-144; b) Z. Qu, X. Zhang, F. Yu, X. Liu, Q. Fu, J.
Catal. 2015, 321, 113-122.
[23] A. M. Ferraria, A. P. Carapeto, A. M. Botelho do Rego, Vacuum 2012,
86, 1988-1991.
[24] C. Lamberti, A. Zecchina, E. Groppo, S. Bordiga, Chem. Soc. Rev. 2010,
39, 4951-5001.
[25] K. Tarach, K. Góra-Marek, M. Chrzan, S. Walas, J. Phys. Chem. C 2014,
118, 23751-23760.
[26] K. Góra-Marek, K. A. Tarach, Z. Piwowarska, M. Łaniecki, L. Chmielarz,
Catal. Sci. Technol. 2016, 6, 1651-1660.
[27] Z. Qu, S. Zhou, W. Wu, C. Li, X. Bao, Catal. Lett. 2005, 101, 21-26.
[28] X. D. Wang, R. G. Greenler, Surf. Sci. 1990, 226, L51-L54.
8
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000533
ChemCatChem
FULL PAPER
Entry for the Table of Contents
Three series of Ag modified MOR all outperform pure H-MOR in dimethyl ether carbonylation to methyl acetate, in which Ag0 species
were identified active by combining multiple complementary characterizations (e.g. FTIR, CO-adsorption, XPS, UV-Vis).
9
This article is protected by copyright. All rights reserved.