Dimethyl ether carbonylation to methyl acetate over highly crystalline zeolite seed-derived ferrierite

Article abstract of DOI:10.1039/c8cy00329g

Gas-phase carbonylation of dimethyl ether (DME) to methyl acetate (MA) was investigated on ferrierite (FER) zeolite having different Si/Al molar ratios of 10.4-12.5 as well as high crystallinity synthesized by using various zeolite seed materials such as MOR, ZSM-5 and USY. The enhanced crystallinity of the FER prepared by simply using FER seeds (denoted as FER@FER) having newly formed mesopore structures was responsible for an increased amount of active Br?nsted acid sites, which resulted in a higher MA productivity of 2.94 mmol g_cat h^-1 with MA selectivity of above 99%. The highly crystalline FER@FER revealed the suppressed deposition of aromatic coke precursors due to the presence of fewer defect sites. Compared to other zeolite seed-derived FER zeolites, the lesser amount of defect sites (extra-framework Lewis acid Al species, EFAL) on the FER@FER was successfully controlled through a recrystallization process. The active Br?nsted acid sites for the DME carbonylation reaction mainly originated from the preferential formation of stable tetrahedral Al sites (especially the T2 sites of the Al-O-Si-O-Al framework of FER) on the 8- and 10-membered ring channels of the FER@FER. On those stable T2 sites having proper acid strength, the adsorbed methyl intermediates formed by the dissociation of DME can be transformed to acetyl adsorbates by a relatively faster CO insertion rate on the vicinal Br?nsted acid sites, which results in a high catalytic stability and activity of the highly crystalline FER@FER.

Full text of DOI:10.1039/c8cy00329g

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Kim, H. Ham, H.
S. Jung, Y. Wang, Y. He, N. Tsubaki, S. Cho, G. Y. Han and J. W. Bae, Catal. Sci. Technol., 2018, DOI:
10.1039/C8CY00329G.
This is an Accepted Manuscript, which has been through the
Volume 6 Number 1 7 January 2016 Pages 1–308
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
Accepted Manuscripts are published online shortly after
www.rsc.org/catalysis
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
ISSN 2044-4753
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
PAPER
Qingzhu Zhang et al.
Catalytic mechanism of C–F bond cleavage: insights from QM/MM
analysis of fluoroacetate dehalogenase
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/catalysis

Page 1 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
Submitted to the Catalysis Science & Technology
Dimethyl ether carbonylation to methyl acetate over highly crystalline
zeolite-seed derived ferrierite
Jihyeon Kima,1, Hyungwon Hama,1, Hyun Seung Junga, Yang Wangb, Yingluo Heb,
Noritatsu Tsubakib, Sung June Choc,*, Gui Young Hana, Jong Wook Baea,**
aSchool of Chemical Engineering, Sungkyunkwan University (SKKU), 2066 Seobuꢀro,
Janganꢀgu, Suwon, Gyeonggiꢀdo, 16419, Republic of Korea
bDepartment of Applied Chemistry, School of Engineering, University of Toyama, Gofuku 3190,
Toyama, 930ꢀ8555, Japan
cDepartment of Chemical Engineering, Chonnam National University, Gwangju, 500ꢀ757,
Republic of Korea
1The authors of J. Kim amd H. Ham were equally contributed.
ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ
*Corresponding author (S.J. Cho): Tel.: +82ꢀ62ꢀ530ꢀ1902; Fax: +82ꢀ62ꢀ530ꢀ1819; Eꢀmail
address: sjcho@chonnam.ac.kr
**Corresponding author (J.W. Bae): Tel.: +82ꢀ31ꢀ290ꢀ7347; Fax: +82ꢀ31ꢀ290ꢀ7272; Eꢀmail
address: finejw@skku.edu
1

Catalysis Science & Technology
Page 2 of 36
View Article Online
DOI: 10.1039/C8CY00329G
ABSTRACT
Gasꢀphase carbonylation of dimethyl ether (DME) to methyl acetate (MA) was investigated
on the ferrierite (FER) zeolite having different Si/Al molar ratios of 10.4 – 12.5 as well as
high crystallinity synthesized by using various zeolite seed materials such as the MOR, ZSMꢀ
5 and USY. The enhanced crystallinity of the FER prepared by simply using the FER seed
(denoted as FER@FER) having newly formed mesopore structures was responsible for an
increased amount of the active Brønsted acid sites, which resulted in a higher MA
productivity of 2.94 mmol/(gcat·h) with MA selectivity above 99%. The highly crystalline
FER@FER revealed the suppressed depositions of aromatic coke precursors due to the less
presence of defect sites. Compared to other zeolite seedꢀderived FER zeolites, the less
amount of defect sites (extraꢀframework Lewis acidic Al species, EFAL) on the FER@FER
was successfully controlled through a recrystallization process. The active Brønsted acid sites
for DME carbonylation reaction were mainly originated from the preferential formations of
the stable tetrahedral Al sites (especially, T2 sites of the AlꢀOꢀSiꢀOꢀAl framework of FER)
on the 8 and 10ꢀmembered ring channels of the FER@FER. On those stable T2 sites having
proper acid strengths, the adsorbed methyl intermediates formed by the dissociation of DME
can be transformed to acetyl adsorbates by a relatively faster CO insertion rate on the vicinal
Brønsted acid sites, which are resulted in a high catalytic stability and activity of the highly
crystalline FER@FER.
Keywords: Carbonylation; Dimethyl ether (DME); Methyl acetate (MA); Ferrierite (FER);
Brønsted acid sites.
2

Page 3 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
1. Introduction
Liquid or gasꢀphase carbonylation reactions of methanol or dimethyl ether (DME) to acetic
acid or methyl acetate (MA) has been intensively investigated to produce valueꢀadded useful
chemical intermediates,1-4 where MA can be further transformed to an alternative clean fuel
of ethanol by subsequent hydrogenation.1,5-7 The MA has been widely used for petrochemical
productions such as an acetic anhydride as well as acetyl cellulose.4 Even though a lot of
environmentallyꢀbenign solidꢀacid zeolites were applied for DME carbonylation with CO to
selectively produce MA, DME conversion on those heterogeneous solidꢀacid catalysts was
relatively lower and deactivated significantly. These phenomena were much more significant
on the MFI or BEA zeolites having a less amount of 8 memberedꢀring (8ꢀMR) channels.
Therefore, mordenite (MOR) or ferrierite (FER) zeolites having a lot of acidity in the 8ꢀMR
channels have been reported to be effective ones for a gasꢀphase carbonylation of DME,3,6,8-10
where DME can be produced by CO (or CO2) hydrogenation using bifunctional CuꢀZnOꢀ
Al2O3/zeolites.10 The MOR zeolite has been well reported to have superior characteristics
such as the structures connected by 8ꢀMR of side pocket and 12 memberedꢀring (12ꢀMR),
where the former is known to be active sites and the latter is pathway of reactant and product
transports for a higher DME conversion and MA selectivity.6,9 The number of OH groups in
the 8ꢀMR channels on the MOR has been well related with DME carbonylation activity since
it can easily generate methyl groups by dissociate adsorption of DME resulted in forming
surface acetyl (CH3COꢀ) intermediates through a facile insertion of the adsorbed CO.3,6,9
However, the MOR zeolite showed severe deactivation phenomena by forming heavy coke
precursors on the 12ꢀMR channels selectively.6,9 Therefore, many studies have been focused
on solving the coke depositions on the MOR by using postꢀtreatments such as desilication,
dealumination and metal ion exchanges such as copper, nickel, zinc or iron and so on.11-14
Recently, the FER having structural characteristics of the side pockets of 8ꢀMR channels
3

Catalysis Science & Technology
Page 4 of 36
View Article Online
DOI: 10.1039/C8CY00329G
connected with the 10ꢀMR channels, which is smaller than the 12ꢀMR channels of the MOR,
revealed a higher MA selectivity and stability due to a suppressed coke formation at a
relatively lower activity than the MOR.5,15,16 Therefore, based on our previous research
results,15 the gasꢀphase DME carbonylation was investigated by using a labꢀmade pristine
FER with different Si/Al ratios to effectively prohibit the coke depositions, and the results
revealed that a higher crystallinity of FER is responsible for a less coke deposition with a
longꢀterm stability and higher selectivity to MA during DME carbonylation.
In the present investigation, we tried to find new methods to shorten the preparation period
and to further increase the crystallinity of asꢀprepared FER through a simple surface coating
method using various zeolite seeds such as MOR, ZSMꢀ5 and USY, which are known to be
active for DME carbonylation as well. The higher crystallinity of the FER coated with
previously synthesized FER seed zeolite showed a significantly improved catalytic activity
and stability even for a less duration of FER synthesis period required. The enhanced
catalytic performance on the highly crystalline FER was well related with its crystallinity and
amount of active Brønsted acid sites, which were precisely analyzed by defining surface
acidity and intermediates with the fine structural analyses through Rietveld refinement
analysis of XRD and solidꢀstate NMR.
2. Experimental section
2.1. Preparation procedures of seed-derived FER zeolites and activity measurements
The FER seed zeolite was previously synthesized through hydrothermal method at s fixed
Si/Al molar ratio of 10.4 as described in our previous works.15 For more details, after mixing
the separate precursors of silica and alumina such as a fumed silica (SiO2) and sodium
aluminate (NaAlO2) with a structure directing agent of piperidine (hexahydropyridine,
C5H11N) at a fixed molar ratio of SiO2/NaAlO2/piperidine/NaOH/DIW = 1:0.096:0.9:0.14:40,
4

Page 5 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
o
the solution mixtures were kept in a hydrothermal reactor with stirring for 7 days at 160 C.
The obtained gel after washing several times was subsequently calcined at 550 oC for 6 h, and
asꢀprepared zeolite was denoted as the FER. For the preparation of the seedꢀderived FER
zeolites coated by using various zeolite seeds such as commercially available MOR (Si/Al =
15), ZSMꢀ5 (Si/Al = 25) and USY (Si/Al = 2.3), the same mixed solutions for a preparation
of the FER seed was stirred for 1 h in the slurry of each zeolite seed before a hydrothermal
treatment. However, while using the labꢀmade FER seed, the mixed solutions was added in a
slurry of the labꢀmade FER seed without using the structure directing agent of piperidine due
to their same surface structures of the FER seed. After a complete mixing again, the slurry
o
was kept in a hydrothermal reactor for 4 days at 160 C After the hydrothermal synthesis, the
FER prepared from each zeolite seed was filtered, washed again using deionized water (DIW)
followed by drying overnight. Asꢀprepared white powder was calcined under air environment
at 550 oC for 6 h. To convert Naꢀform zeolite to Hꢀform ones, asꢀprepared seedꢀderived FER
zeolites were stirred in 1 M NH4NO3 solution at 80 oC and washed with DIW for more than 6
o
times at the same temperature. The sample was further calcined at 550 C for 3 h, and asꢀ
prepared FER zeolites using different zeolite seeds were separately denoted as FER@FER,
MOR@FER, ZSMꢀ5@FER and USY@FER with respect to the different types of zeolite
seeds. The weight ratio of the various zeolite seed material was fixed to 24 wt% based on
total weight of the final asꢀprepared FER zeolites.
Catalytic activity measurements for a gasꢀphase carbonylation of DME to MA on the seedꢀ
derived FER zeolites using various zeolite seeds were performed in a fixed bed tubular
reactor having an inner diameter of 9 mm. For the activity measurement, the catalyst of 0.4 g
was previously loaded into the reactor and it was pretreated at 500 °C under N2 environment
to remove any physisorbed contaminants. The gasꢀphase carbonylation reaction was carried
out at T = 220 °C, P = 1.0 MPa, and space velocity (SV) of 2000 L/(kgcat·h) for a reaction
5

Catalysis Science & Technology
Page 6 of 36
View Article Online
DOI: 10.1039/C8CY00329G
time of 100 h using mixed gas reactants of DME/CO/N2(mol%) = 5/45/50, where N2 used as
an internal standard gas at a higher N2 concentration for a les coke deposition. The effluent
gases from the reactor were analyzed using an onꢀline gas chromatography (Younglin GC,
YL6100) installed with a flame ionization detect (FID) with DBꢀWAX capillary column to
analyze the hydrocarbons, methanol, DME, MA and other byproducts. Conversion of DME
and selectivity of MA with other byproducts were carefully calculated based on total carbonꢀ
balance due to an insignificant coke formation on the used catalysts for a long reaction
duration of 100 h. The production rate (space time yield, STY) of MA was calculated using
an averaged mole of MA produced [mmol/(gcat·h)] at the reaction duration of 80 ꢀ 100 h, and
turnꢀover frequency (TOF, sꢀ1) was defined as the MA produced at an initial reaction duration
per Brønsted acid site and second [mmol/(mmol(B sites)·s)]. The deactivation rate (RD, %/h)
of the seedꢀderived FER zeolites was also calculated by the equation of [(maximum
conversion of DME – conversion of DME at 100 h)/(reaction duration), which was further
normalized using total Brønsted acid sites as well.
2.2. Characterization of the FER prepared by using various zeolite seeds
Wideꢀangle powder Xꢀray diffraction (XRD) patterns of the fresh seedꢀderived FER zeolites
were obtained using a Bruker Xꢀray diffractometer (D8 Advance) with a Cu Kα radiation of
0.15406 nm at a scanning rate of 4o/min operated at 40 kV and 100 mA. The characteristic
diffraction peaks of zeolites were measured in the range of 2θ = 5 ꢀ 50o. The relative
crystallinity of the seedꢀderived FER zeolites with various zeolite seeds was calculated by
comparing the integrated areas of most intense two diffraction peaks of 2θ = 9.3 and 25.2o,
which are the characteristic diffraction peaks of FER, by assuming 100% crystallinity of the
most highly crystalline FER@FER. The Si/Al molar ratios and Na contents of the fresh seedꢀ
derived FER zeolites prepared from various zeolites seeds were measured by Xꢀray
6

Page 7 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
fluorescence (XRF) using a Bruker S4 instrument operating at 60 kV and 150 mA. N2
adsorptionꢀdesorption analysis of the seedꢀderived FER zeolites with various zeolite seeds
was carried out by using a Micromeritics instrument (Tristar II) to characterize specific
surface area (BrunauerꢀEmmettꢀTeller (BET) method), pore volume (tꢀplot method) and
median pore diameter (BarrettꢀJoynerꢀHalenda (BJH) method from desorption branch) at
liquid nitrogen temperature of ꢀ196 oC.
To verify the surface morphologies and crystalline sizes of the seedꢀderived FER zeolites,
Fieldꢀemission scanning electron microscopy (FESEM) analysis was carried out by a JEOL
scanning electron microscopy (JSMꢀ7500F) operated at 1.4 nm guaranteed at 1 kV and at 0.6
nm guaranteed at 30 kV. Before the image analysis, the sample was previously coated with a
conductive platinum metal on a conductive carbon tape substrate. Highꢀresolution
transmission electron microscopy (HRTEM) analysis was performed by a JEOL transmission
electron microscopy (JEMꢀ2100F) operated at 200 kV on the selected fresh FER and
FER@FER only. To elucidate the detailed fine structures of the seedꢀderived FER zeolites,
Rietveld refinement analysis from XRD data and solidꢀstate magic angle spinning nuclear
magnetic resonance (MAS NMR) analysis were further carried out on the fresh FER zeolites.
27Al and 29Si MAS NMR analysis on all the fresh seedꢀderived FER zeolites were performed
by using a 500 MHz solid state NMR (a Varian Unity INOVA instrument) at a resonance
frequency of 130.26 MHz using a 3.2 mm Chemagnetics MAS probe head at a spinning rate
of 10 kHz with a reference material of Al(NO3)2 and a delay time (D1) of 1 s at a number of
scans of 4000 and π/2 pulse of 1.0 ꢁs. The Rietveld refinement analysis of the seedꢀderived
FER zeolites from the XRD data was performed by using a GSAS package and EXPGUI as
an interface with a starting model in P42 space group with the isotropical refinements of all
atoms, and the refined XRD peak shapes were described with a pseudoꢀVoigt function.
7

Catalysis Science & Technology
Page 8 of 36
View Article Online
DOI: 10.1039/C8CY00329G
Temperature programmed desorption of NH3 (NH3ꢀTPD) analysis was performed using 30
mg powder of the fresh seedꢀderived FER zeolites at the temperature range of 100 ꢀ 700 °C
after the sample pretreatment at 350 oC for 1 h under He flow using a BELCATꢀM instrument
equipped with a thermal conductivity detector (TCD). Fourier transformed infrared spectra of
pyridine (PyꢀIR) analysis were carried out using 30 mg sample, which was previously treated
o
at 500 C for 1 h under a vacuum condition to remove any contaminants and water adsorbed
on the surfaces of the thin FER pellet. The PyꢀIR spectra were obtained after the adsorption
of pyridine molecules for 15 min on the sample using a Nicolet 6700 FTꢀIR spectrometer
with 4 cmꢀ1 resolution using a MCT detector at a fixed temperature of 220 oC to maximize the
pyridine adsorption capacity on the small pores16 of the seedꢀderived FER zeolites. The
formed surface intermediates and adsorption strengths of DME reactant on the fresh seedꢀ
derived FER zeolites prepared with different zeolite seeds were characterized by inꢀsitu
Fourier transform infrared (FTꢀIR) analysis under a flow of reactant of DME and CO at
o
various temperatures. Prior to the analysis, the sample was pretreated at 500 C for 1 h under
a vacuum condition to remove any contaminants adsorbed. After the DME and CO reactants
were flowed into the labꢀmade FTꢀIR cell for 20 min, the physisorbed DME and CO
molecules were removed at a vacuum evacuation for 10 min. The FTꢀIR spectra of the
adsorbed chemical species were obtained at each temperature stages from 180 to 260 oC at an
ambient pressure with an interval of 20 oC using a Perkin Elmer Frontier FTꢀIR spectrometer
equipped with 4 cmꢀ1 resolution using a MCT detector as well. In addition, the successive
o
adsorption characteristics of DME followed by CO molecules at 220 C using a same
instrument were further verified to confirm the different CO insertion rates on the selected
FER and FER@FER.
Temperatureꢀprogrammed surface reaction with H2 (TPSR) analysis was carried out at the
temperature range of 100 ꢀ 700 °C using a BELCATꢀM instrument to confirm the types of
8

Page 9 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
coke precursors on the used seedꢀderived FER zeolites after the reaction for 100 h. The
methane (m/z = 15), which can be formed by hydrogenation of coke precursors during the
TPSR, was monitored by using a quadruple mass spectrometer (MS Pfeiffer QMS 200).
o
Before the TPSR analysis, all samples were pretreated at 250 C for 1 h under a He flow to
remove any physically adsorbed contaminants. Thermogravimetric analysis (TGA) of the
used seedꢀderived FER zeolites was conducted by a TA Instrument (Q50) to quantitatively
identify the amount of coke precursors formed during the DME carbonylation for 100 h. In
addition, for the further analysis of chemical species of the deposited cokes, the used catalyst
was completely dissolved in a hydrofluoric acid (HF) solution, and it was subsequently
dissolved in a 4 ml of dichloromethane solvent to completely extract the deposited cokes
from the dissolved zeolite components. After a complete layer separation of the above mixed
solution, the dichloromethane including coke precursors formed during DME carbonylation
reaction for 100 h was analyzed by using GC/MS (a YoungLin instrument 6500 GC with
YL6900 GC/MS system) equipped with a HPꢀ5 capillary column having a specification of 15
m length x 0.25 um inner diameter x 0.25 mm coating thickness.
3. Results and discussion
3.1. Physicochemical properties of the FER prepared by various zeolite seeds
The wellꢀsynthesized FER structures with their higher crystallinity on all the fresh seedꢀ
derived FER zeolites prepared by various zeolite seeds such as a pristine FER, MOR, ZSMꢀ5
and USY were supported from the observed characteristic XRD peaks of FER15,17,18 without
showing any characteristic XRD patterns of each seed zeolites as shown in Figure 1(A). A
relative crystallinity calculated from the most intense peaks of 2θ = 9.3 and 25.2o by
assuming 100 % crystallinity of the highly crystalline FER@FER was summarized in Table
1. Compared with the FER seed, a relative crystallinity of the FER@FER was slightly
9

Catalysis Science & Technology
Page 10 of 36
View Article Online
DOI: 10.1039/C8CY00329G
increased from 94.7 to 100 % even for much shorter synthesis duration for 4 days. However,
the crystallinity of the FER synthesized using other zeolite seeds such as MOR, ZSMꢀ5 and
USY was found to be lower in the range of 76.8 – 83.7 %. These lower crystallinities, except
for the FER@FER, seem to be attributed to different crystalline phases of each zeolite seeds
and its surface acidity due to different Si/Al ratios, which seems to prevent an easy growth of
the FER structures. The different crystallinity of the seedꢀderived FER zeolites can generate
different amounts and types of acidic sites with various morphologies, which resulted in a
different DME carbonylation activity. In general, the postꢀsynthesis of the FER for making a
microꢀmesopore zeolite structure largely altered surface properties by changing the defect site
densities originated from Al(OH)3 species (mainly responsible for Lewis acid sites).19-22 In
addition, the Na content in the fresh seedꢀderived FER zeolites were found to be less than
0.07wt% on all the seedꢀderived FER zeolites from XRF analysis as summarized in Table 1,
and the Si/Al molar ratios on the seedꢀderived FER zeolites were found to be similar in the
range of 10 – 12.5 even though the original seed zeolites have various Si/Al ratios in the
range of 2.3 to 25 due to the excess amounts of the coated FER zeolites on the zeolite seeds.
We believe that a higher crystallinity of the FER@FER compared with other zeolites can be
attributed to the reduced amounts of defect sites by the recrystallization process of the defect
sites with insignificant contributions of Na residue and Si/Al ratios on the seedꢀderived FER
zeolites even though the catalytic activity on the virgin zeolite seeds themselves were largely
affected by their Si/Al ratios (higher initial activity on the USY zeolite having a smaller Si/Al
ratio of 2.3) as shown in supplementary Figure S3. In addition, a lower crystallinity on the
seedꢀderived FER zeolites prepared by MOR, ZSMꢀ5 and USY seed can be also induced
from the shortened intraꢀparticular domain lengths from the different morphologies of seed
zeolites with locally formed amorphous phases.23
10

Page 11 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
The surface acidity with its types on the fresh seedꢀderived FER zeolites was verified by
NH3ꢀTPD and PyꢀIR, and these results are displayed in Figure 1(B) and Figure 1(C). The
desorption peaks of the adsorbed NH3 molecules at ~ 200 and 450 °C as shown in Figure 1(B)
can be assigned to weak and strong acid sites, which are similar with the typical desorption
patterns of many zeolites. As summarized in Table 1, the amounts of weak acid sites
o
(assigned to the peak below 300 C from NH3ꢀTPD) were found to be in the range of 0.40 –
0.66 mmol/g with a larger value on the FER@FER. However, the amounts of strong acid
o
sites (assigned to the peak at 350 ~ 550 C from NH3ꢀTPD) were found to be in the wide
range of 0.44 – 0.70 mmol/g with a maximum value on the FER@FER. Therefore, the total
acid sites were maximized on the FER@FER with the value of 1.36 mmol/g, and the acid
strengths were modified by its selective depositions on the seed zeolites. In general, too
stronger acid sites on the zeolites can be responsible for a preferential deposition of coke
precursors by a facile decomposition of DME to hydrocarbons.10,15,18 The hypothesis was
supported by maximum temperature shifts of Tmax on the various seedꢀderived FER zeolites
o
as shown in Figure 1(B). Compared with the FER seed with Tmax = 204 C, the Tmax was
o
shifted to lower ones such as Tmax = 200 C on the FER@FER and MOR@FER, and 191 oC
on the ZSMꢀ5@FER and USY@FER. Therefore, the types of those acid sites were further
measured by PyꢀIR analysis, and a relative concentration of Brønsted and Lewis acid sites
was reevaluated using total acid sites measured by NH3ꢀTPD. As shown in Figure 1(C), FTꢀ
o
IR peaks of adsorbed pyridines at 220 C were observed with the characteristic absorption
peaks at 1542 cmꢀ1 for Brønsted acid sites (assigned to B) and at 1454 cmꢀ1 for Lewis acid
sites (assigned to L) with the combined acid sites at around 1480 cmꢀ1.24 On the seedꢀderived
FER zeolites, Brønsted acid sites originated 8ꢀMR channels of the FER structures are known
to be the most active sites for DME carbonylation.9,25 Even though the larger size of pyridine
molecules seems not to be effective to quantitatively measure the acidic sites of 8ꢀMR
11

Catalysis Science & Technology
Page 12 of 36
View Article Online
DOI: 10.1039/C8CY00329G
channels of the FER due to the spatial size restriction effect of pyridine,16(b) they can be
possibly adsorbed on all the Brønsted acid sites through an increased lattice flexibility as well
o
as kinetic energy of the pyridine at higher temperatures above 200 C.25(b) As summarized in
Table 1, the ratio of Brønsted to Lewis acid site (B/L) on the outer zeolite surfaces was
maximized on the FER@FER with that of 84.1, which means a larger amount of Brønsted
acid sites with 1.34 mmol/g. A relatively smaller outer surface B/L ratio in the range of 10.5
– 16.4 with smaller amount of Brønsted acid sites of 0.79 – 0.85 mmol/g were observed on
the MOR@FER, ZSMꢀ5@FER and USY@FER. These Brønsted acid sites were known to be
active through an easy formation of surface acetyl intermediates (CH3COꢀ).3,6,9,25 Therefore,
the FER@FER having abundant Brønsted acid sites by a selective removal of defect sites
with a higher crystallinity seems to be the best catalyst, which was also supported by TEM
images in Figure 2 by showing a higher crystallinity on the FER@FER than the FER seed.
As shown by SEM images in Figure 2, the surface morphologies of all the seedꢀderived
FER zeolites clearly showed the characteristic plateꢀlike crystals with an irregular sizes and
shapes of the typical FER morphology. Although the characteristic XRD patterns of the
zeolite seeds on the seedꢀderived FER zeolites were not clearly observed in Figure 1(A) due
to the partial disintegrations of the intrinsic structures of seed zeolites possibly, the intrinsic
morphologies of each seed zeolites were maintained with the present FER morphology
mainly. Interestingly, the FER@FER showed much smaller grain sizes of FER crystallites
with an average crystallite size of ~ 1 ꢁm compared to the FER seed. The MOR@FER, ZSMꢀ
5@FER and USY@FER showed their original characteristic morphology of each seed
zeolites covered with small plateꢀlike FER crystallites. These smaller crystallite sizes on the
FER@FER were responsible for the newly formed mesopores with its high crystallinity,
which is beneficial for a higher DME conversion through an easy mass transport. The
specific surface area, average pore diameter above mesopore size and pore volume are
12

Page 13 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
summarized in Table 1. All the seedꢀderived FER zeolites showed similar mesopore
characters in the range of 249 – 282 m2/g, 0.16 – 0.17 cm3/g and 5.6 – 8.5 nm, respectively.
As shown in supplementary Figure S1, N2 adsorptionꢀdesorption isotherms on all catalysts
showed the characteristic type IV isotherms with similar hysteresis patterns. A little bit larger
average pore size and smaller surface area on the FER@FER seem to be originated from the
newly formed interꢀparticular mesopores between the smaller plateꢀlike FER crystallites
(SEM images in Figure 2).21-23 Interestingly, the mesopores above 3 nm in size on the seedꢀ
derived FER zeolites were responsible for the larger mesopore surface areas in the range of
20.5 ꢀ 32.0 m2/g with the maximized area of 32.0 m2/g on the highly crystalline FER@FER
due to its wellꢀdeveloped interꢀparticular mesopores. The smaller surface area of 261 m2/g
and larger acidic sites of 1.36 mmol NH3/g on the FER@FER than the pristine FER with the
respective values of 282 m2/g and 1.10 mmolNH3/g can be attributed to the abundant
presence of the newly formed interꢀparticular mesopores with its higher crystallinity.
3.2. Catalytic activity and product distribution for DME carbonylation reaction
The production rate of MA and DME conversions with product distributions on the labꢀ
made seedꢀderived FER zeolites and commercial FER zeolite are summarized in Table 1 and
supplementary Table S1. The catalytic activity at the selected reaction temperature of 220 oC
was compared, which showed an insignificant initial deactivation rate based on the activity of
the commercial FER (CFER, Si/Al ratio of 10.4 as shown in Table S1). The space time yield
(STY, mmol/gcat·h) with time on stream (h) on the seedꢀderived FER zeolites are also
displayed in Figure 1(D). Even though all the seedꢀderived FER zeolites showed a slight
deactivation at the present reaction conditions of T = 220 °C, P = 1.0 MPa, and space velocity
(SV) of 2000 L/(kgcat·h), a higher DME conversion was kept for a longer time on the highly
crystalline FER@FER compared to other seedꢀderived FER zeolites. The calculated
13

Catalysis Science & Technology
Page 14 of 36
View Article Online
DOI: 10.1039/C8CY00329G
production rate (STY) of MA at a steadyꢀstate was found to be in the range of 0.82 – 2.94
mmol/(gcat·h) with the maximum value on the FER@FER and minimum one on the
USY@FER, which was similar with the previous work.6 A much higher STY at around
reaction durations of 80 – 100 h was found to be 2.94 mmol/(gcat·h) (equal to DME
conversion of 71.5 %) with the MA selectivity of 99.2 % on the FER@FER (higher MA yield
of 70.9%), which was much higher than the FER seed itself with the STY of 1.99
mmol/(gcat·h), which can be attributed to the newly formed interꢀparticular mesopore
structures and larger Brønsted acid sites. The relatively lower STYs in the range of 0.82 ꢀ
1.34 mmol/(gcat·h) were observed on the MOR@FER, ZSMꢀ5@FER and USY@FER, where
those catalysts possessed a smaller amount of Brønsted acid sites with a lower crystallinity.
On all the seedꢀderived FER zeolites, the byproducts such as methanol and hydrocarbons
(mainly, CH4) were found to be below 4 %, except for the ZSMꢀ5@FER since it has been
well known for the active catalyst for methanol to gasoline (MTG) process by carbon pool
mechanisms by decomposition or hydrogenolysis reactions of methanol and DME.26,27 In
addition, the initial TOF (sꢀ1) (calculated from the active sites on the fresh seedꢀderived FER
zeolites) was also found to be a higher on the FER@FER with the value of 4.90 x 10ꢀ4 sꢀ1
compared to other seedꢀderived FER zeolites in the range of 0.87 – 2.56 x 10ꢀ4 sꢀ1. The
variations of DME conversion and product distributions with time on stream with small
initial deactivation rate are displayed in supplementary Figure S2 and the product
distributions were stabilized after 30 h on stream without any significant changes on all the
seedꢀderived FER zeolites. The deactivation rate (%/h) per the unit Brønsted acid sites on the
seedꢀderived FER zeolites was calculated and summarized in Table 1. Even at an initially
higher STY on the FER@FER compared to other seedꢀderived FER zeolites, the deactivation
rate was found to be 0.07 on the FER@FER, which revealed much smaller deactivation than
that of the MOR@FER of 0.23 from the very beginning of the reaction. Therefore, the types
14

Page 15 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
of zeolite seeds revealed a significantly different surface properties and catalytic activity
according to the crystallinity and density of Brønsted acid sites. These surface properties on
the seedꢀderived FER zeolites synthesized with various zeolite seeds largely altered the
deactivation mechanisms and adsorption properties as well. In addition, this coating
procedure using the FER seed by recrystallization step of defect sites can significantly
enhance the STY of MA with its selectivity without any significant byproduct formations
rather than other zeolite seeds themselves such as MOR, ZSMꢀ5 and USY as shown in
supplementary Figure S3. All commercially available zeolites such as the MOR, ZSMꢀ5 and
USY themselves except for the labꢀmade FER seed without coating processes showed a fast
deactivation with a lower MA selectivity with time on stream (supplementary Figure S3),
which also revealed less significant contributions of the zeolite seeds themselves for altering
the catalytic activity and stability by their physical mixing effects with the pristine FER.
These modified zeolite surfaces by decorating with the FER structures, especially on the
FER@FER can enhance the catalytic performances significantly by adjusting the surface
properties. Therefore, we carried out the precise surface characterizations by using FTꢀIR and
TPSR analysis and the fine structural analysis using XRD and solidꢀstate NMR, which are to
elucidate the behaviors of coke depositions and adsorption characteristics of DME as well as
to verify the origins of Brønsted acid sites.
3.3. Adsorption properties of DME and CO with coke analysis on the synthesized FER
The adsorbed surface intermediates from DME/CO reactants were verified by using inꢀsitu
o
FTꢀIR analysis at adsorption temperature ranges of 180 ꢀ 260 C. The FTꢀIR spectra in the
wavenumber ranges of 2800 – 3100 for the adsorbed DME intermediates are displayed in
o
Figure 3 and supplementary Figure S4 for one selected temperature at 200 C. On the
various types of Brønsted acid sites such as SiOH, AlOH and SiOHAl surface groups on the
15

Catalysis Science & Technology
Page 16 of 36
View Article Online
DOI: 10.1039/C8CY00329G
zeolites, the absorption bands appeared at 3007, 2968, 2949 and 2840 cmꢀ1 can be possibly
assigned to the hydrogen bonded DME on these Brønsted acid sites without showing any CO
adsorption peaks. The absorption bands at 2968 and 2864 cmꢀ1 observed at a higher
o
temperature above 220 C can be possibly assigned to asymmetric and symmetric stretching
vibration modes of the dissociatively adsorbed methyl (CH3ꢀ) groups from the adsorbed
DME,6,9,28-34 which can be fast reacted with CO molecules to selectively generate acetyl
intermediate. In addition, the absorption bands appeared at 2949 and 2864 cmꢀ1 were
attributed to the hydrogenꢀbonded DME on the defected surface sites such as SiOH and
AlOH, respectively,31 which were found to be much larger intensity compared with that at
2968 cmꢀ1 on the lower crystalline zeolites such as the ZSMꢀ5@FER and USY@FER rather
than the highly crystalline pristine FER and FER@FER. Interestingly, the relatively smaller
intensities of these peaks on the MOR@FER seem to be originated from a fast deactivation at
the very beginning of reaction as shown in Figure 1(D) as well. In addition, the much weaker
absorption band intensity of 2855 cmꢀ1 on the FER@FER in all temperature ranges compared
with other seedꢀderived FER zeolites was mainly attributed to its higher crystallinity, which
can be assigned to methyl groups adsorbed on the external silanol groups existing in the form
of SiOCH3.32 Therefore, the stable formations of hydrogenꢀbonded DME and methyl
intermediates on the highly crystalline Brønsted acidic sites (SiOHAl) can enhance an easy
insertion of CO molecules into the surface methyl groups to form surface acetyl groups
selectively. Generally, the CO insertion step has been known to be kinetically controlled and
possible rate determining step for DME carbonylation to MA.6,9,28,29 Furthermore, the extra
framework Al species (EFAL) can act as Lewis acid sites,35,36 which can preferentially form
the byproducts such as hydrocarbons through methanolꢀtoꢀolefin reaction mechanisms by the
wellꢀknown decomposition reaction of methanol and DME.27,37 Therefore, the highly
crystalline FER@FER having larger amount of Brønsted acid sites instead of less active
16

Page 17 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
Lewis acid sites (defect EFAL sits) seems to be responsible for a higher STY and MA
selectivity due to its optimal bond strengths of the adsorbed methyl intermediates appeared at
2968 cmꢀ1 for further fast CO insertion reaction. The observed lower activity on the less
crystalline zeolites such as the MOR@FER, ZSMꢀ5@FER and USY@FER, which revealed
much larger absorption peak intensities at the absorption bands at 2949 and 2855 cmꢀ1
(strongly interacted methyl groups on the defect sites) can be explained by a relatively slow
CO insertion rate to form acetyl groups due to too stronger adsorption characters of the
methyl intermediates on the defect sites. These hypotheses were further supported by the
characteristic absorption peaks of water molecules formed by methanol dehydration or
hydrocarbon formation reactions during DME activation step on the Brønsted acid sites as
shown in supplementary Figure S5 and Figure S6. The water absorption bands assigned to
the separate peaks at 1640 and 3610 cmꢀ1 on the Brønsted acid sites originated from the 8 and
12ꢀMR channels were found to be much smaller on the most active FER@FER and their
intensities were decreased with an increase of adsorption temperatures due to the selective
absorption natures of hydrogenꢀbonded DME molecules on those active Brønsted acidic OꢀH
groups assigned to the acidic sites of the 8 and 12ꢀMR channels6,9,29,38,39 or possible
significant coke deposition such as polymerized alkenes38 on the fast deactivated MOR@FER.
Even though the characteristic absorption bands of C=O in acetyl groups at 1704 cmꢀ1 was
not clearly observed due to a lower experimental pressure condition (0.1 MPa),6,9,29,39 the
absorption bands appeared at 3660 cmꢀ1 and that at 1680 cmꢀ1, which can be respectively
assigned to the defect sites of the EFAL29 and C=O absorption band in aldehyde groups, was
found to be larger on the less crystalline seedꢀderived zeolites, especially on the ZSMꢀ
5@FER and USY@FER. Therefore, we believe that the water formed by methanol
dehydration reaction to DME or hydrocarbons9 on the Brønsted acid sites can be
competitively adsorbed on the same acid sites,40 which can prevent an easy DME activation
17

Catalysis Science & Technology
Page 18 of 36
View Article Online
DOI: 10.1039/C8CY00329G
as well as CO insertion reaction to form acetyl groups. These competitive adsorptions of
water molecules on the zeolite surfaces can largely alter the activity of methanol dehydration
to DME, where the reaction was well correlated with the surface weak acid sites in two
separate regions on the solid acid heterogeneous catalysts10,40 as shown in supplementary
Figure S7.
In terms of the catalyst deactivation, the polyaromatic hydrocarbons have been considered
as the inactive coke precursors to irreversibly deactivate solidꢀacid heterogeneous catalysts
by being formed inside channels or external surfaces of zeolites.41 As shown in Figure 4(A)
and Table 1, the weight loss (Wloss, %) measured by TGA can be correlated with deactivation
rates since it was related with the total amounts of fully dehydrogenated graphitic or
amorphous carbons.42 The deposited amount of coke precursors corresponding to the weight
o
loss above ~400 C, which can be possibly originated from the decomposition of DME
through carbon pool mechanisms of methanol to gasoline process on the defect sites of the
seedꢀderived FER zeolites, was found to be in the following orders of MOR@FER > ZSMꢀ
5@FER ≈ USY@FER > FER >> FER@FER with their respective values of 6.3, 5.4, 5.3, 4.7
and 3.6 % with the similar thermostability of the seedꢀderived FER zeolites. The more
detailed analysis of the types of coke precursors was carried out by TPSR and GC/MS
experiments, and the results are displayed in Figure 4(B) and Figure 4(C). The coke
precursors deposited on the used zeolites can be categorized as the characteristic
hydrogenation peaks with α, β, and γ coke precursors, which can be assigned to the TPSR
peaks at ~ 350, 450, and 550 °C, respectively. The αꢀform coke precursor can be originated
from the hydrogenation of dehydrogenated surface carbidic carbons, the βꢀform coke for
amorphous hydrocarbons, and γꢀform coke for inactive surface carbons, which can be
removed hardly due to its natures of heavy cyclic polyaromatic hydrocarbons.43 The relative
peak intensities on the used seedꢀderived FER zeolites revealed the smallest intensity on the
18

Page 19 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
FER@FER, and the extent of coke depositions follows the orders of MOR@FER ≈ ZSMꢀ
5@FER > USY@FER > FER >> FER@FER. Interestingly, the peaks of the much harder
coke precursors (assigned to γꢀform) were found to be larger on the less active ZSMꢀ5@FER
and USY@FER. These deposited coke precursors were found to be the wellꢀknown aromatic
heavy hydrocarbon species such as the tetramethyl benzene and tirmethyl benzene as
characterized by GC/MS analysis as shown in Figure 4(C), which were found to be more
dominant on the MOR@FER showing a faster initial deactivation nature. Therefore, a facile
coke deposition and lower catalytic activity with fast deactivation seems to be mainly
attributed to the lower crystallinity and unstable Brønsted acid site (originated from defected
Al sites) by selectively forming dehydrogenated heavy coke precursors instead of active
acetyl intermediates.3,6,9,25,44
3.4. Analyses of fine structures of FER zeolites and proposed reaction mechanisms
The more detailed analyses of crystalline structures were carried out by Rietveld refinement
techniques45,46 from XRD data of the fresh seedꢀderived FER zeolites prepared by using
various zeolite seeds. The distributions of tetrahedral sites (T sites) and their bond valences
(BVs) with its bond energies were calculated, and the final results are displayed in Figure
5(A) and supplementary Figure S8 with the summarized bone valences of T2 sites in Table 1.
The more detailed results of BVs and crystallographic parameters are also summarized in
Table S2 and Table S3. The crystallographic parameters were found to be similar on all the
fresh seedꢀderived FER zeolites with a space group of Immm, where they revealed the values
of a = 1.88 nm, b = 1.41 nm and c = 0.74 nm with an acceptable R values below 10 %47,48 as
summarized in supplementary Table S2. Although the similar BVs of T1, T3 and T4 sites
were observed on all the seedꢀderived FER zeolites, the smaller BVs of T2 sites with the
value of 3.76 was clearly observed on the FER@FER compared with other FER zeolites as
19

Catalysis Science & Technology
Page 20 of 36
View Article Online
DOI: 10.1039/C8CY00329G
summarized in supplementary Table S3 and Table 1. This observation strongly suggests the
stable presence of T2 sites in the 8 and 10ꢀMR channels on the FER@FER, which are in line
the observations of the previous work due to their hard removal characteristics from the
framework tetrahedral Al species,47 which was schematically summarized in Figure 5(A). In
general, the number of Brønsted acid sites in the 8ꢀMR channels of the FER has been
reported to relate with DME carbonylation activity by easily generating the surface acetyl
(CH3COꢀ) intermediates via a facile insertion of the CO molecules.3,6,9 Therefore, the most
stable T2 sites assigned to Brønsted acid sites of the AlꢀOꢀSi main frameworks6,9,28-34 on the
FER@FER seem to be responsible for an increased activity and stability compared with other
seedꢀderived FER zeolites. In addition, solidꢀstate MAS NMR analysis of 27Al and 29Si on the
fresh seedꢀderived FER zeolites was further performed to verify the structural characteristics
and its environments. As shown in Figure 5(B), the 27Al NMR peaks showed the
characteristic chemical shifts of aluminum such as at ~ 54 ppm associated with the tetrahedral
(4ꢀcoordinated) Al atoms (Al(OSi)4) in the main framework of the FER and at ~ 0 ppm
associated with the octahedral (6ꢀcoordinated) Al atoms, which is known as the extraꢀ
framework Al species (EFAL) as possible defect sites. A shoulder peak at ~ 37 ppm indicates
the pentaꢀcoordinated Al atoms or distorted tetrahedral EFAL species as well.49-51 On the
FER@FER having a highly crystalline structure with larger Brønsted acid sites at the stable
T2 sites, the peak intensity of Al(OSi)4 appeared at ~54 ppm was noticeably larger than other
FER zeolites due to the stochastically abundant presences of the stable T2 sites with less
defect sites in the 8 and 10ꢀMR channels on the FER@FER. The 29Si solidꢀstate MAS NMR
spectra on the seedꢀderived FER zeolites are also displayed in Figure 5(B) (supplementary
Figure S9 for all synthesized FER zeolites) with the characteristic peaks at the chemical
shifts of silicon of ꢀ99, ꢀ105, ꢀ111 and ꢀ115 ppm with a broad peak centered at ꢀ110 ppm. The
peaks appeared at ꢀ99 and ꢀ105 ppm can be respectively related to the Si atoms bonded to two
20

Page 21 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
Al atoms (Si(2Al)) or one Al atom (Si(1Al)) with a smaller peak intensity of Si(2Al), and the
peaks at ꢀ111 and ꢀ115 ppm can be attributed to the Si atoms without any vicinal Al atoms
(Si(0Al)) with the silanol groups (SiOH) on the FER surfaces appeared at ꢀ99 ppm.52-54 The
chemical shift at ꢀ105 ppm assigned to the Si(1Al) was decreased on the FER@FER with a
concomitant increase of the peak intensity at ꢀ 111 ppm compared to those of the FER. This
observation can be possibly originated from the wellꢀdeveloped T2 sites with the
homogeneous distribution of Al atoms on the FER@FER framework by preferentially
forming the stable AlꢀOꢀSiꢀOꢀAl structures.54 In addition, the smaller presence of the stable
Si(0Al) sites seems to be related with the catalytic stability by showing a much fast
deactivation on the MOR@FER having the smallest value (76 %) of (Si(0Al)) and largest
amount of SiOH nest groups (7 %) as shown in supplementary Figure S9. It also suggests
that the highly crystalline FER@FER has much smaller defect sites with the abundant
presence of T2 sites assigned to Brønsted acid sites in the 8 and 10ꢀMR channels of the FER
structures compared to other seedꢀderived FER zeolites, which is responsible for its higher
intrinsic catalytic performance since the activities were measured under the conditions of
insignificant mass transfer limitations (more detailed calculations of those mass and heat
transport limitations are further included in supplementary material) as well as far below the
equilibrium conversion55 as shown in Figure S10.
Finally, as shown in Figure 5(C) (scheme S1), the initial step of DME adsorption on the
Brønsted acid sites of the seedꢀderived FER zeolites can easily occur to form surface methyl
groups with a concomitant formation of methanol intermediate. However, the rate
determining step of acetyl group formation by CO insertion reaction can be largely altered by
the crystallinity (stable T2 sites) of the seedꢀderived FER zeolites by competing with its water
adsorption natures formed by methanol dehydration to DME or hydrocarbons.9 Since the
higher crystallinity of the FER@FER with less amounts of defect sites, which can be
21

Catalysis Science & Technology
Page 22 of 36
View Article Online
DOI: 10.1039/C8CY00329G
originated from the isolated SiOH or AlOH groups or EFAL sites, can possibly generate
larger amounts of the active Brønsted acid sites assigned to the stable T2 sites in the 8 and
10ꢀMR channels of the FER@FER through their facile recrystallizations during the seedꢀ
oriented preparation step. In addition, these stable Brønsted acid sites can enhance the
formation rate of acetyl intermediates through a fast rate of CO insertion to the stable methyl
surface groups. The fast CO insertion rate on the highly crystalline FER@FER having a
larger amount of Brønsted acid sites instead of the pristine FER was further confirmed by
comparing the variation of FTꢀIR peak intensity of the adsorbed methyl intermediates
appeared at 2968 cmꢀ1 to the strongly interacted methyl groups on the defect sites for the band
at 2949 cmꢀ1 (I2968/I2949) by a successive adsorption of DME followed by CO reactant as
shown in supplementary Figure S11. As summarized in Figure 5(D), the observed much
larger ratios of the integrated FTꢀIR peak area (I2968/I2949) on the FER@FER compared to
pristine FER strongly revealed the more facile CO insertion on the highly crystalline active
Brønsted acid sites assigned to the stable T2 sites in the 8 and 10ꢀMR channels of FER
structures. Finally, the formation MA can be finished by reacting with DME to regenerate the
surface methyl groups as well as schematically summarized in Figure 5(C). Much strongly
interacted methyl groups assigned to the absorption peaks at 2949 and 2855 cmꢀ1 in the FTꢀIR
can be responsible for a coke and undesired byproduct formation, which can irreversibly
deactivate the seedꢀderived FER zeolites such as ZSMꢀ5@FER, USY@FER and MOR@FER.
The possible rate determining step of acetyl formation by CO insertion step can be
significantly increased on the highly crystalline FER@FER having a large amount of
Brønsted acid sites (T2 sites) with fewer amounts of defect sites such as the EFAL. The
newly formed large mesopore structures on the FER@FER with its higher crystallinity
having a nature of Brønsted acidity can also suppress the formations of inactive heavy coke
22

Page 23 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
precursors irreversibly, and the crystallinity of the FER zeolite can be easily adjusted by
using the previously synthesize FER seed even with a shorter zeolite synthesis duration.
4. Conclusions
The seedꢀderived FER zeolites with its Si/Al molar ratios of 10.0 – 12.5 prepared by simply
using the various zeolite seed materials such as FER, MOR, ZSMꢀ5 and USY were
investigated for a gasꢀphase carbonylation of DME to MA to verify the effects of the
crystallinity of the seedꢀderived FER zeolites with the number of active Brønsted acid sites.
The largely enhanced crystallinity of the FER@FER through its possible recrystallization was
mainly responsible for an increased amount of Brønsted acid sites assigned to stable T2 sites
in the 8 and 10ꢀMR channels of the FER@FER, which also showed higher STY and TOF of
MA production with a higher selectivity above 99 %. The highly crystalline FER@FER
revealed a lower deactivation rate due to the suppressed depositions of heavy aromatic coke
precursors as well. The decreased amounts of defect sites assigned to an extraꢀframework Al
species were observed on the highly crystalline FER@FER by the facile recrystallization
natures compared with other seedꢀderived FER zeolites prepared from other zeolite seeds.
The active Brønsted acid sites assigned to the stable T2 sites on the FER@FER effectively
played as the important active sites for a selective formation of hydrogenꢀbonded DME
intermediates by overcoming a competitive absorption of the water generated from
dehydration of methanol on the Brønsted acid sites. These preferential formations of active
Brønsted acidic sites were responsible for an enhanced activity of DME carbonylation with
less coke depositions by easy transformations of the dissociatively adsorbed DME molecules
to active acetyl adsorbates due to a relatively faster CO insertion rate on the vicinal Brønsted
acid sites on the highly crystalline FER@FER.
23

Catalysis Science & Technology
Page 24 of 36
View Article Online
DOI: 10.1039/C8CY00329G
Acknowledgments
The authors would like to acknowledge the financial support from the National Research
Foundation of Korea (NRF) grant funded by the Korea government sincerely (NRFꢀ
2018M3D3A1A01018009 and NRFꢀ2017R1D1A1B03028214). This work was also
supported by “Next Generation Carbon Upcycling Project” (2017M1A2A2043135) through
the National Research Foundation (NRF) funded by the Ministry of Science and ICT,
Republic of Korea.
Supplementary Information
Additional information such as pore size distribution from N2 sorption isotherms, FTꢀIR of
the adsorbed intermediates, detailed catalytic activities including DME conversion with MA
selectivity with time on stream (h), original XRD data for Rietveld refinement analysis and
27Al and 29Si NMR spectra on all the seedꢀderived FER zeolites are also included in
supplementary material.
References
1
(a) H. Xue, X. Huang, E. Zhan, M. Ma and W. Shen, Catal. Commun., 2013, 37, 75ꢀ79;
(b) Q. Qian, M. Cui, J. Zhang, J. Xiang, J. Song, G. Yang and B. Han, Green Chem.,
2018, 20, 206ꢀ213.
2
3
J. Liu, H. Xue, X. Huang, P. H. Wu, S. J. Huang, S. B. Liu and W. Shen, Chinese J.
Catal., 2010, 31, 729ꢀ738.
B. Li, J. Xu, B. Han, X. Wang, G. Qi, Z. Zhang, C. Wang and F. Deng, J. Phys. Chem. C,
2013, 117, 5840ꢀ5847.
4
5
X. F. Wu and H. Neumann, ChemCatChem., 2012, 4, 447ꢀ458.
J. Liu, H. Xue, X. Huang, Y. Li and W. Shen, Catal. Lett., 2010, 139, 33ꢀ37.
24

Page 25 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
6
A. Bhan, A. D. Allian, G. J. Sunley, D. J. Law and E. Iglesia, J. Am. Chem. Soc., 2007,
129, 4919ꢀ4924.
7
8
9
G. A. FloresꢀEscamilla and J. C. FierroꢀGonzalez, Catal. Sci. Technol., 2015, 5, 843ꢀ850.
H. Ham, J. Kim, S. Park and J. W. Bae, Adv. Por. Mater., 2016, 4, 200ꢀ205.
P. Cheung, A. Bhan, G. J. Sunley and E. Iglesia, Angew. Chem. Int. Ed., 2006, 45, 1617ꢀ
1620.
10 J. W. Jung, Y. J. Lee, S. H. Um, P. J. Yoo, D. H. Lee, K. W. Jun and J. W. Bae, Appl.
Catal. B: Environ., 2012, 126, 1ꢀ8.
11 S. Wang, W. Guo, L. Zhu, H. Wang, K. Qiu and K. Cen, J. Phys. Chem. C, 2015, 119,
524ꢀ533.
12 H. Zhou, W. Zhu, L. Shi, H. Liu, S. Liu, S. Xu, Y. Ni, Y. Liu, L. Li and Z. Liu, Catal. Sci.
Technol., 2015, 5, 1961ꢀ1968.
13 H. Zhan, S. Huang, Y. Li, J. Lv, S. Wang and X. Ma, Catal. Sci. Technol., 2015, 5, 4378ꢀ
4389.
14 A. A. C. Reule and N. Semagona, ACS Catal., 2016, 6, 4972ꢀ4975.
15 (a) H. Ham, J. Kim, J. H. Lim, W. C. Sung, D. H. Lee and J. W. Bae, Catal. Today, 2018,
303, 93ꢀ99; (b) S. Y. Park, C. H. Shin and J. W. Bae, Catal. Commun., 2016, 75, 28ꢀ31.
16 (a) X. Li, X. Liu, S. Liu, S. Xie, X. Zhu, F. Chen and L. Xu, RSC Adv., 2013, 3, 16549ꢀ
16557; (b) W. Chu, X. Li, X. Zhu, S. Xie, C. Guo, S. Liu, F. Chen and L. Xu, Micropor.
Mesopor. Mater., 2017, 240, 189ꢀ196.
17 D. E. W. V. Karl and G. Strohmaier, J. Am. Chem. Soc., 2003, 125, 16035ꢀ16039.
18 H. M. Koo, H. S. Roh, Y. W. Suh and J. W. Bae, Reac. Kin. Mech. Catal., 2014, 112, 499ꢀ
510.
19 M. S. Holm, S. Svelle, F. Joensen, P. Beato, C. H. Christensen, S. Bordiga and M.
Bjørgen, Appl. Catal. A: Gen., 2009, 356, 23ꢀ30.
25

Catalysis Science & Technology
Page 26 of 36
View Article Online
DOI: 10.1039/C8CY00329G
20 X. Zhu, L. L. Lobban, R. G. Mallinson and D. E. Resasco, J. Catal., 2010, 271, 88ꢀ98.
21 C. Fernandez, I. Stan, J. P. Gilson, K. Thomas, A. Vicente, A. Bonilla and J. Perezꢀ
Ramirez, Chem. Eur. J., 2010, 16, 6224ꢀ6233.
22 D. Verboekend, R. CaicedoꢀRealpe, A. Bonilla, M. Santiago and J. PerezꢀRamirez, Chem.
Mater., 2010, 22, 4679ꢀ4689.
23 A. Bonilla, D. Baudouin and J. PerezꢀRamirez, J. Catal., 2009, 265, 170ꢀ180.
24 T. Gott and S. T. Oyama, J. Catal., 2009, 263, 359ꢀ371.
25 (a) D. B. Rasmussen, J. M. Christensen, B. Temel, F. Studt, P. G. Moses, J. Rossmeisl, A.
Riisager and A. D. Jensen, Angew. Chem. Int. Ed., 2015, 54, 1ꢀ5; (b) J. A. Z. Pieterse, S.
VeefkindꢀReyes, K. Seshan, L. Domokos and J. A. Lercher, J. Catal., 1999, 187, 518ꢀ520.
26 N. Y. Chen and W. J. Reagan, J. Catal., 1979, 59, 123ꢀ129.
27 (a) J. W. Park and G. Seo, Appl. Catal. A: Gen., 2009, 356, 180ꢀ188; (b) W. Zhao, B.
Zhang, G. Wang and H. Guo, J. Energy Chem., 2014, 23, 201ꢀ206.
28 A. Bhan and E. Iglesia, Acc. Chem. Res., 2008, 41(4), 559ꢀ567.
29 P. Cheung, A. Bhan, G. J. Sunley, D. J. Law and E. Iglesia, J. Catal., 2007, 245, 110ꢀ123.
30 J. G. Chen, P. Basu, T. H. Ballinger and J. T. Yates, Jr. Langmuir, 1989, 5(2), 352ꢀ356.
31 H. Zhou, W. Zhu, L. Shi, H. Liu, S. Liu, Y. Ni, Y. Liu, Y. He, S. Xu, L. Li and Z. Liu, J.
Mol. Catal. A: Chem., 2016, 417, 1ꢀ9.
32 T. R. Forester and R. F. Howe, J. Am. Chem. Soc., 1987, 109, 5076ꢀ5082.
33 W. Wang, A. Buchholz, M. Seiler and M. Hunger, J. Am. Chem. Soc., 2003, 125 15260ꢀ
15267.
34 W. Wang, M. Seiler and M. Hunger, J. Phys. Chem. B, 2001, 105, 12553ꢀ12558.
35 A. Zecchina and C. O. Arean, Chem. Soc. Rev., 1996, 25, 187ꢀ197.
36 L. Benco, T. Bucko, J. Hafner and H. Toulhoat, J. Phys. Chem. B, 2004, 108, 13656ꢀ
13666.
26

Page 27 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
37 D. Lesthaege, V. Van Speybroeck, G. B. Marin and M. Waroquier, Chem. Phys. Lett.,
2006, 417, 309ꢀ315.
38 Y. X. Wei, D. Z. Zhang, Z. M. Liu and B. L. Su, J. Catal., 2006, 238, 46ꢀ57.
39 O. Kresnawahjuesa, R. J. Gorte and D. White, J. Mol. Catal. A: Chem., 2004, 208, 175ꢀ
185.
40 K. Saravanan, H. Ham, N. Tsubaki and J. W. Bae, Appl. Catal. B: Environ., 2017, 217,
494ꢀ522.
41 M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino, L. Palumbo, S.
Bordiga and U. Olsbye, J. Catal., 2007, 249, 195ꢀ207.
42 B. Sanchez, M. S. Gross, B. D. Costa and C. A. Querini, Appl. Catal. A: Gen., 2009, 364,
35ꢀ41.
43 J. Xu, W. Zhou, J. Wang, Z. Li and J. Ma, Chinese J. Catal., 2009, 30, 1076ꢀ1084.
44 M. V. Luzgin, M. S. Kazantsev, W. Wang and A. G. Stepanov, J. Phys. Chem. C, 2009,
113, 19639ꢀ19644.
45 B. Toby, J. Appl. Crystallogr., 2001, 34, 210ꢀ213.
46 W. Kraus and G. Nolze, J. Appl. Crystallogr., 1996, 29, 301ꢀ303.
47 P. Norby, F. I. Poshni, A. F. Gualtieri, J. C. Hanson and C. P. Grey, J. Phys. Chem. B,
1998, 102, 839ꢀ856.
48 R. Baran, Y. Millot, T. Onfroy, J. M. Krafft and S. Dzwigaj, Micropor. Mesopor. Mater.,
2012, 163, 122ꢀ130.
49 X. F. Li, R. Prins and J. A. van Bokhoven, J. Catal., 2009, 262, 257ꢀ265.
50 M. Muller, G. Harvey and R. Prins, Micropor. Mesopor. Mater., 2000, 34, 281ꢀ290.
51 A. A. C. Reule, J. A. Sawada and N. Semagina, J. Catal., 2017, 349, 98ꢀ109.
52 Y. P. Khitev, Y. C. Kolyagin, I. I. Ivanova, O. A. Ponomareva, F. ThibaultꢀStarzyk, J. P.
Gilson, C. Fernandez and F. Fajula, Micropor. Mesopor. Mater., 2011, 146, 201ꢀ207.
27

Catalysis Science & Technology
Page 28 of 36
View Article Online
DOI: 10.1039/C8CY00329G
53 J. B. Nagy, Z. Gabelica, G. Debras, P. Bodart and E. G. Derouane, J. Mol. Catal., 1983,
20, 327ꢀ336.
54 J. Dedecek, S. Sklenak, C. Li, F. Gao, J. Brus, Q. Zhu and T. Tatsumi, J. Phys. Chem. C,
2009, 113, 14454ꢀ14466.
55 S. T. Oyama, X. Zhang, J. Lu, Y. Gu and T. Fujitani, J. Catal., 2008, 257, 1ꢀ4.
28

Page 29 of 36
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY00329G
Figure Captions
Figure 1. Bulk and surface properties and catalytic activity of the seedꢀderived FER zeolites
synthesized by using various zeolite seeds; (A) XRD patterns, (B) NH3ꢀTPD profiles, (C) Pyꢀ
IR spectra and (D) Space time yield (STY, mmol/(gcat·h)) with time on stream (h) at T =
220 °C, P = 1.0 MPa, and space velocity (SV) of 2000 L/(kgcat·h) using a mixed gas reactant
of DME/CO/N2(mol%) = 5/45/50
Figure 2. SEM images of the seedꢀderived FER zeolites synthesized by using various zeolite
seeds such as the FER, MOR, ZSMꢀ5 and USY and TEM images of the fresh FER and
FER@FER (upper middle figures)
Figure 3. FTꢀIR spectra of the adsorbed DME in the range of 2800 – 3100 cmꢀ1 at different
temperatures on the fresh seedꢀderived FER zeolites synthesized from various zeolite seeds
with the analyzing temperature profiles
Figure 4. Characterization of the seedꢀderived FER zeolites synthesized from various zeolite
seeds; (A) TGA patterns, (B) TPSR patterns by measuring the methane formed (m/z = 15),
and (C) Coke analysis by GC/MS
Figure 5. (A) Tetrahedral sites at 8 and 10ꢀMR channels on the seedꢀderived FER zeolites
analyzed by Rietveld refinement analysis from XRD data, (B) NMR spectra of the selected
fresh FER and FER@FER with 27Al MAS NMR spectra and 29Si MAS NMR spectra, (C)
Proposed reaction mechanisms on the highly crystalline FER@FER, and (D) variations of
characteristic FTꢀIR peaks of the adsorbed methyl intermediates at 2968 cmꢀ1 to the strongly
interacted methyl groups on the defect sites assigned to the band at 2949 cmꢀ1 (denoted as
I2968/I2949) by a successive adsorption of DME followed by CO reactant for 10 minutes
29