Non-oxidative Coupling of Methane to Ethylene Using Mo2C/[B]ZSM-5

Article abstract of DOI:10.1002/cphc.201701001

Methane non-oxidative coupling to ethylene was investigated on Mo₂C/[B]ZSM-5 catalyst at 923 K and atmospheric pressure. In contrast to Mo₂C/[Al]ZSM-5 catalysts for methane aromatization, this material exhibits very high ethylene selectivity (>90 %) and low aromatics (benzene and naphthalene) selectivity. The much weaker Br?nsted acidity of [B]ZSM-5 leads to a slow rate of ethylene oligomerization. The stability of the catalyst is greatly enhanced with 93 % of the initial reaction rate remaining after 18 h of time on stream. In-situ UV/VIS spectra indicate that prior to carburization, mono/binuclear Mo oxides are initially well dispersed onto the zeolite support. Mo carbides clusters, formed during carburization with methane, appear similar to clusters formed in [Al]ZSM-5, as indicated by the X-ray Absorption Spectroscopy (XAS) data.

Full text of DOI:10.1002/cphc.201701001

Accepted Article
Title: Non-oxidative Coupling of Methane to Ethylene Using Mo2C/
[B]ZSM-5
Authors: Huibo F Sheng and Raul Francisco Lobo
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: ChemPhysChem 10.1002/cphc.201701001
Link to VoR: http://dx.doi.org/10.1002/cphc.201701001
A Journal of
www.chemphyschem.org

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
Non-oxidative Coupling of Methane to Ethylene Using
Mo C/[B]ZSM-5
2
Huibo Sheng, Edward P. Schreiner, Weiqing Zheng, and Raul F. Lobo*[a]
Abstract: Methane non-oxidative coupling to ethylene was
investigated on Mo C/[B]ZSM-5 catalyst at 923 K and atmospheric
pressure. In contrast to Mo C/[Al]ZSM-5 catalysts for methane
aromatization, this material exhibits very high ethylene selectivity
>90%) and low aromatics (benzene and naphthalene) selectivity.
achieved in the presence of Mo in the sample.[5] An induction
period was observed with MoO /HZSM-5, implying that the
actual active phase was formed in-situ from the MoO /HZSM-5
starting material.[6, 7] Further research has revealed that the
primary reaction of C–H bond dissociation occurs on Mo
nanoparticle surface.[8] At the surface, a molybdenum-carbene-
like intermediate, CH =Mo, is involved in the proposed
2
3
2
3
(
2
C
The much weaker Brønsted acidity of [B]ZSM-5 leads to a slow rate
of ethylene oligomerization. The stability of the catalyst is greatly
enhanced with 93% of the initial reaction rate remaining after 18 h of
time on stream. In-situ UV/VIS spectra indicate that prior to
carburization, mono/binuclear Mo oxides are initially well dispersed
onto the zeolite support. Mo carbides clusters, formed during
carburization with methane, appear similar to clusters formed in
2
mechanism. Ethylene is believed to be the primary product from
the coupling of the surface carbene species in these samples.[
Ethylene oligomerization reactions take place simultaneously
leading to the formation of aromatics species, the
8, 9]
thermodynamically favored products at the high reaction
temperature.[
10]
The formation of aromatics is catalyzed by
[Al]ZSM-5, as indicated by the X-ray Absorption Spectroscopy (XAS)
data.
Brønsted acid sites in the zeolite pores and the material
operates as a bi-functional catalysts.[
11]
The acidic OH groups of the zeolite catalyst, however, also
promote coke formation and fast catalyst deactivation at high
Introduction
temperature.[ By removal of external acid sites or of part of
12]
[13]
Effective utilization of methane is of great interest and
importance to the chemical industry because of the rapid growth
of shale gas production and the low cost of this raw material.[
The conversion of methane to higher hydrocarbons can be
categorized into indirect methods and direct methods. Indirect
methods use carbon monoxide and hydrogen—produced via
methane steam reforming, carbon dioxide reforming, or partial
oxidation—as intermediates.[2] Direct methods involve methane
coupling with or without the presence of an oxidant, known
respectively as oxidative coupling of methane (OCM) and non-
oxidative coupling of methane (NCM).[
[14]
the framework aluminum, the catalyst activity and stability can
be significantly improved. In the rich literature of methane
aromatization to benzene, the selectivity of ethylene has not
been investigated in much detail. Considering the nature of the
bi-functional Mo/HZSM-5 catalysts, one may assume that
ethylene could be selectively produced solely on molybdenum
carbide, where the methane C–H bond dissociation and coupling
1]
2
reactions occur. However, unsupported Mo C shows very limited
methane reactivity (0.1 ~ 0.3% conversion), has poor selectivity
(
hydrogen, carbon, and a minor amount of ethane are produced),
3]
[10]
and suffers from fast deactivation. In contrast, in this report we
describe an investigation of methane coupling to ethylene on
molybdenum carbide nanoparticles supported on non-acidic
One of the methane coupling products is ethylene and according
to a PwC(PricewaterhouseCoopers) report, the capital cost of
ethylene produced using NCM is estimate as $316/ton, while the
current ethylene price from crude oil steam cracking is
2
[B]ZSM-5. Different from Mo C/[Al]ZSM-5 catalysts for methane
aromatization reaction, this material exhibits very high ethylene
selectivity (>90%) instead of producing aromatics (benzene and
naphthalene). Due to the low acidity of the zeolite support, the
$
1717/ton.[4] Non-oxidative coupling of methane to ethylene is
thus a promising alternative to generate a carbon stream that
can be fed directly into the existing structure of the chemical
industry, with potentially less than one fifth of the current cost.
stability of the catalyst is also greatly improved with
deactivation rate of less than 0.5% h during an 18 h reaction
a
-1
test period, while Mo
few hours.[
2
C/[Al]ZSM-5 usually deactivates within a
11]
Wang and co-workers pioneered non-oxidative methane
dehydroaromatization reaction on Mo/ZSM-5. It is reported that
only 1.4% of methane was fully converted to benzene at 973 K
[5]
Results and Discussion
on HZSM-5 catalysts.
A higher conversion of 7.2% was
[
B]ZSM-5 was synthesized under hydrothermal crystallization
[
a]
H. Sheng, E. P. Schreiner, Dr. W. Zheng, Prof. Dr. R. F. Lobo
Center for Catalysis Science and Technology
Department of Chemical and Biomolecular Engineering
University of Delaware
conditions as indicated in the experimental section. The samples
were analyzed by Inductively Coupled Plasma (ICP)
measurements to quantify composition. A high Si/B ratio of 70 is
obtained by this technique. 2% of Mo is introduced to the zeolite
by wetness impregnation. The impregnated sample is then
calcined in air at 823 K for 4 h. Before the methane reaction, the
150 Academy Street, Newark, DE, 19716 (USA)
E-mail: lobo@udel.edu
Supporting information for this article is given via a link at the end of
the document\
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
calcined sample is carburized in 20% CH
4
/H
2
at 923K (detailed
per unit cell, this number decreased to 1.7 after calcination and
decreased further to 1 per unit cell after carburization.
information can be found below). Different types of
characterization techniques were applied to the above four
samples, namely [B]ZSM-5, impregnated, calcined, and
carburized 2% Mo/[B]ZSM-5.
Table 1. Unit cell volume (UCV) and parameters of 2% Mo/[B]ZSM-5
parameter
B]ZSM-5
%Mo/[B]ZSM-5
UCV / Å3
a / Å
b / Å
c / Å
[
5322
19.853
19.845
19.847
19.904
20.067
20.080
20.087
20.094
13.360
13.357
13.387
13.378
2
5323
5337
5350
(
impregnated)
%Mo/[B]ZSM-5
calcined)
%Mo/[B]ZSM-5
carburized)
2
(
2
(
*2%Mo/[B]ZSM-5 (carburized)*
Textural and micropore volume estimates of the catalyst were
measured using N adsorption isotherms (Table 2). Both BET
surface area and the micropore volume decreased after Mo was
loaded onto the zeolite, suggesting that the Mo species are
entering the channels of the zeolite in this step.
2
*
2%Mo/[B]ZSM-5 (calcined)
*
*
*2
%Mo/[B]ZSM-5 (impreg.)
*[B]ZSM-5
Table 2. BET surface areas and micropore volumes of 2% Mo/[B]ZSM-5
*
BET surface area
Micropore volume
Material
B]ZSM-5
2
3
m /g
cm /g
[
Al]ZSM-5 reference
30 35 40
[
403.8
0.136
0.112
0.101
0.099
5
10
15
20
25
45
50
2%Mo/[B]ZSM-5
3
91.1
58.5
2
Theta / °
(impregnated)
%Mo/[B]ZSM-5
calcined)
%Mo/[B]ZSM-5
carburized)
2
(
3
Figure 1. XRD patterns of Mo/[B]ZSM-5 (Peaks marked are from 10% wt. Si
internal standard), [Al]ZSM-5 reference: H0.32(Si95.68Al0.32O )
2
(
[15]
337.5
192
We carried out an in-situ UV/VIS investigation of molybdenum
oxides particles formed in the samples upon impregnation and
calcination. Oxomolybedenum species can give strong
absorption bands in the UV/VIS region due to the ligand-to-metal
As shown in Figure 1, the X-ray powder diffraction patterns of
the synthesized B-containing MFI-type zeolites samples have a
well-defined diffraction pattern consistent with the formation of
the pure MFI-type structure. There is no evidence for the
presence of any amorphous material in any of the samples.
Furthermore, in all the 2% Mo/[B]ZSM-5 samples, the peak
positions are slightly shifted towards lower angles as compared
2-
6+ [20]
charge transfer (O → Mo ). The bands are usually relatively
narrow and can be observed between 200 and 400 nm due to
the nature of π to π* charge-transfer transition.[ For the ligand-
to-metal charge transfer, the energy of the electronic transitions
21]
to silicalite, which indicates
a
decrease of the unit cell
have a strong dependence on the ligand field symmetry
dimensions of the boron-containing zeolite upon impregnation
and calcination of the Mo precursor (Figure S1). This is
consistent with the small size of B with respect to Si. No
diffraction peaks for molybdenum oxides are observed in any of
the samples ruling out the formation of (detectable) large Mo
particles in the exterior of the zeolite crystals. This is likely due
to the low Mo loading and is an indication of good dispersion of
the Mo precursors before and after the experiments.
surrounding the molybdenum center. In this work for oxo-ligands,
a more energetic transition is expected for tetrahedral Mo(VI)
than for octahedral Mo(VI).[ Therefore in the literature, most
bands for pure tetrahedral oxomolybdenum were reported at
260-280 nm, while bands for pure octahedral oxomolybdenum
were observed at 300-320 nm.[
21]
22]
In Figure 2, a broad band around 250-350 nm with a shoulder at
235 nm was observed for the impregnated Mo/[B]ZSM-5, while
To measure accurately the zeolite unit cell volume, 10 wt.% of
after a calcination a sharp band at 260 nm with a shoulder at
235 nm was formed. The broad band can be considered as the
overlapping of the bands at 260-280 nm and 300-320 nm,
suggesting the fact of polymeric structure of both tetrahedral and
pure Si powder was mixed with each sample and used as
internal standard.[
16]
Calculated unit cell volumes and u.c.
parameters are listed in Table 1. This table shows that zeolite
unit cell expanded after the calcination and carburization steps,
indicating the loss of framework boron during both reactions.
Deboronation is facilitated at the high temperature range and in
the presence of water generated in both calcination and
caburization steps. Unlike dealumination in aluminosilicate
zeolite, boron can be removed from the framework even under
mild conditions (e.g. stirring in water at room temperature), an in
the presence of high temperatures and water vapor this process
is accelerated.[ The amount of framework B can be estimated
on the basis of the size of the unit cell because unit cell
dimensions are linearly correlated with the boron content in the
octahedral molybdenum centers of the supported oxomolybdate
species.[ Since the precursor (NH
23]
4
)
6
Mo
7
O24 is only composed
of 7 MoO
6
octahedra linked by sharing oxo-edges, we believe
the tetrahedral molybdenum centers are formed during the
impregnation and the drying steps, since the blue shift of the
absorption bands is also reported during the impregnation step
in other metal oxide catalysts due to the near surface pH
changes.[ After a calcination, the material presents two bands
at 235 nm and 260 nm, implying the formation of tetrahedral
molybdednum centers and the disappearance of octahedral
24]
17]
centers. In fact these two bands are very typically observed for
unit cell.[
18, 19]
According to the published correlation, the as-
[19]
2-
CaMoO
4
or Na
2
MoO
4
, where MoO
4
anions are isolated from
2+
+ [25]
made zeolite B-MFI sample contained about 2.5 boron atoms
each other either by Ca or Na .
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
The UV/VIS spectra of Mo/[Al]ZSM-5 are also very well studied
in the literature by other authors. Vasenin et al.[ reported a
band at 330 nm for the initial 10% Mo/[Al]ZSM-5 after drying
26]
Edge energy : 3.6 eV 4.1 eV
8
6
4
2
0
3
73K and assign the band to the precursor (NH
4
)
6
Mo
7 24
O .
[27]
Martinez and Peris
only observed bands at 220-250 nm for
the calcined 3% Mo/[Al]ZSM-5 catalyst, indicating the
molybdenum mainly exists as highly dispersed tetrahedral
[
28]
oxomolybdate species. Ngobeni et al. also reported the bands
at 220-250 nm for 2% Mo/[Al]ZSM-5 and found the addition of
boron or silver as dopants does not influence the coordination
sphere of molybdenum species.
2%Mo/[B]ZSM-5
impregnated)
(
2%Mo/[B]ZSM-5
(calcinated)
2% Mo/[B]ZSM-5
(calcined)
3
.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0
Energy / eV
Figure 3. UV/VIS edge energies of 2%Mo/[B]ZSM-5
2% Mo/[B]ZSM-5
(impregnated)
After the formation of mono/binuclear molybdenum oxides, the
4 2
samples were carburization in a flow of 20% CH /H (20 sccm)
at 973 for h. temperature programmed oxidation
K
8
A
experiment (TPO) was carried out on these samples to quantify
the ratio of Mo and carbon after the carburization step (Figure 4).
2
50
300
350
400
450
850 The trace with negative values depicts the consumption of O
and the positive signals are due to the generation of CO and
CO . At the beginning of the experiment, in the low temperature
regime (400-600 K), CO is the favored oxidation product and as
the temperature increased (700-823 K), CO exceeded the CO
produced to become the main product. However Ding et al.[
reported that the reoxidation of the 4% Mo C/[Al]ZSM-5 occurs
as a single fast process with sharp evolutions of both CO and
CO at about 725 K. For bulk Mo the CO evolution
temperature is even higher at about 1000 K. The lowest CO
evolution temperature suggests the best reactivity and oxygen
dissociation ability of the Mo C/[B]ZSM-5. In addition, by
2
Wavelength / nm
2
Figure 2. UV/VIS diffused reflectance spectra of 2%Mo/[B]ZSM-5
2
8]
2
As the narrowing and the blue shift of the band are observed
when there is a decrease of the oxomolybdates cluster size or
the counter-cation numbers,[ the UV/VIS spectra suggest that
the surface oxomolybdates species have been well dispersed
and anchored onto the zeolite support, but how small the cluster
is remains unclear. Weber[ reported that the average number
of adjacent molybdenum cations in very small oxide clusters is
linearly correlated with the optical band gap energy of the bulk
oxide, and this property can be readily determined using UV/VIS
spectroscopy (Figure 3). According to the correlation the
polyoxomolybdates turn into mono/binuclear molybdenum
oxides after calcination implying not only that the molybdenum
migrates into the zeolite pores but also that these species are
anchored onto the silanol groups and/or the weak Brønsted acid
boron sites.
2
2
C
x
[8]
x
23]
2
comparing the oxygen consumed by carbon and by
molybdenum, a Mo/C ratio of 1.9 was estimated, indicating the
carburization process was almost completed with all the oxides
turning into carbides.
29]
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
a rate of 2 K/min. Growth of the pre-edge feature was observed,
that is, there is an increase of the tetrahedral symmetry of the
Mo center.[ During the second step (upper right), when the
temperature was increased to 923 K at 1 K/min, a plateau was
reached with no major structure change. In the third step (lower
left) the pre-edge feature became less and less intense as the
CO signal
1
CO signal
2
30]
0
temperature was increased to 973
disappearance of the pre-edge absorption feature - indicates
K at 1 K/min. The
-1
-2
-3
-4
that the Mo(VI)-oxo species were reduced and carbidic Mo
species are formed.[
8, 31]
The spectra did not change further at
973 K. The lower right figure shows the overall change of the K-
edge through the whole carburization process. The Mo-K
absorption edge shifts to lower energies (∆E = 4.5 eV) during
the carburization process, and the ∆E is about 5 eV with
Mo/[Al]ZSM-5 according to Ding et al.[ The final oxidation state
O signal
Treatment gas: 20 sccm 10% O
2
/He
8]
2
Temperature ramping rate: 10 K/min
Hold at 823 K for 3 h
2
of the Mo was close to the oxidation state of Mo in Mo C,
suggesting that the sample contains a mixture of molybdenum
carbide and molybdenum oxycarbide. This result is consistent
with other XANES analysis during the carburization process of
Mo/[Al]ZSM-5, where Mo center is anchored either onto the
silanol groups and/or Brønsted acid aluminum sites.[
3
00 400 500 600 700 800 900100011001200130014001500
Temperature / K
32, 33]
Figure 4. Temperature programmed oxidation of carburized 2% Mo/[B]ZSM-5
1
1
.2
.0
1.2
1.0
0.8
0.6
0.4
0.2
0.0
A
transmission electron microscope (TEM) image of the
carburized 2% Mo/[B]ZSM-5 is shown in Figure 5. In the image,
black dots of ~1 nm in diameter can be identified inside the
2
particle as well-dispersed Mo C particles. However a few larger
particles around 2 nm are also observed. This can be explained
by the “knock-on” effect observed during the electron beam
irradiation, where the Mo C particles are squeezed out from the
2
zeolite channel to the external surface (Figure S2).
0.8
0
0
0
0
.6
.4
.2
.0
2
5
7
98 K
26 K
23 K
723 K
840 K
923 K
Mo C
Mo C
2
2
1
9980 20000 20020 20040
19980 20000 20020 20040
Energy / eV
Energy / eV
1
1
0
0
0
0
0
.2
.0
.8
.6
.4
.2
.0
1.2
1.0
0.8
0.6
0.4
0.2
0.0
9
9
9
23 K
45 K
73 K
298 K
973 K
Mo C
Mo C
2
2
1
9980 20000 20020 20040
19980 20000 20020 20040
Figure 5. TEM image of 2% Mo/[B]ZSM-5 after carburization
Energy / eV
Energy / eV
Figure 6. X-ray absorption near edge structure (XANES) of 2% Mo/[B]ZSM-5
during in-situ carburization process
The structural coordination and electronic state of Mo can be
described with in-situ X-ray absorption fine structure (XAFS)
spectroscopy during the carburization process. The Mo K-edge
X-ray absorption near edge structure (XANES) was recorded
and is shown in (Figure 6). A three-step K-edge feature change
was observed during the carburization: The first step (upper left)
was performed ramping the temperature from 298 K to 723 K at
The Extended X-ray Absorption Fine Structure (EXAFS) of the
Mo in the 2% Mo/[B]ZSM-5 and standard compounds are
displayed in Figure 7. The sample calcined at 773 K showed
doublet peaks at ~1.2 Å while the intensity of all features beyond
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
3
Å are much weaker; these results reveal that MoO
3
crystallites
implying the substitution of the molybdenum oxides to the
protons of the exchangeable OH groups. The intensity further
decreased after a calcination at 773 K as a result of the
molybdenum oxides being better dispersed as smaller clusters
onto more anchoring sites, which is in agreement with the
previous results from UV/VIS and XAS.
are dissociated into smaller particles such as mono/binuclear
molybdenum oxides.[
33]
The peaks at ~1.2 Å indicate the
formation of Mo centers in tetrahedral coordination instead of the
initial octahedral coordination environment. This observation is
also in consistence with our previous discussion from the
4 2
UV/VIS results. After the carburization with 20% CH /H , the
molybdenum oxide species were reduced and carburized to
form molybdenum carbide. As shown in Figure 6, the newly
2.0
a. 2%Mo/[Al]ZSM-5
3610
b. [B]ZSM-5
x
formed MoC species give rise to two peaks at ~1 and 2 Å.
These peaks however cannot be described only using carbon
c. 2%Mo/[B]ZSM-5(impreg.)
d. 2%Mo/[B]ZSM-5(calcined)
e. 2%Mo/[B]ZSM-5(carburized)
1.6
neighbors in the first shell simulation, which usually only present
one strong peak at ~2.5 Å as reported by others.[
8, 13, 32]
Instead
they can be well described by one oxygen neighbor near 1.7 Å
and three carbon neighbors near 2 Å and was also reported by
3743
1
.2
.8
Li et al. when describing the mechanism of the formation of
x 2
MoC C, Mo has three
species on [Al]ZSM-5.[ In the bulk Mo
33]
3
724 3709
0
carbon neighbors at a distance of 2.1 Å and twelve Mo
neighbors at distances from 2.9 to 3.1 Å.[ The absence of Mo-
Mo pairs at the distance around 3 Å for the carburized sample
suggests that the Mo small species formed upon calcination do
not aggregate into larger clusters upon carburization.
3735
8]
a
0.4
b
c
d
e
0.0
3
800
3750
3700
3650
3600
-1
Wavenumber / cm
Mo Foil
Figure 8. FT-IR spectra of 2%Mo/[Al]ZSM-5 and 2%Mo/[B]ZSM-5
Bulk Mo C
2
Catalytic Tests
2% Mo/[B]ZSM-5 (carburized)
The catalytic properties were evaluated in a flow reactor under
conditions similar to conditions reported in the literature for [Mo]-
Al-ZSM-5 (923 K, 1 atm, 5% CH
Mo C/[B]ZSM-5). Typically, 50 mg of the calcined Mo/[B]ZSM-5
was loaded into the reactor followed by carburization in 20%
vol.) CH /H This step takes about 16 hours with the
temperature ramping from 298 K to 623 K at 2 K/min, then to
73 K at 1 K/min, and held at 973 K for 8 hours. After this step,
4
/He flow of 20 sccm, 50mg
2
2
% Mo/[B]ZSM-5 (calcined)
(
4
2
.
0
2
4
6
8
9
Radical Distance / Å
[35]
the system was cooled to the reaction temperature in He.
After the temperature was stabilized, a reaction gas consisting of
Figure 7. Radical structure function of Mo in standard compounds and in 2%
Mo/[B]ZSM-5 before and after in-situ carburization
4
5% (vol) CH /He was fed into the system and the products are
analysed with an online GC equipped with TCD and FID
detectors (Figure 9). Due to thermodynamic limitations, the
conversion of methane around 1% with a 90%+ selectivity to
ethylene instead of benzene. In terms of the catalyst stability of
the methane aromatization reactions on Mo/[Al]ZSM-5, the
Brønsted acid sites catalyzed oligomerization reaction may
cause carbon deposition, which further covers the active site
In the typical [Al]ZSM-5 catalysts for the methane aromatization
reactions the zeolite offers anchoring sites for the metal to
disperse, and as the co-catalyst, the Brønsted acid sites
[32]
catalyze the ethylene oligomerization reaction to aromatics.
[
11]
and leads to catalyst deactivation within
a few hours.
Diffused reflectance FT-IR experiments were carried out to
compare the role of [B]ZSM-5 to that of the [Al]ZSM-5 (Figure 8).
The 2% Mo/[Al]ZSM-5 showed two IR bands assigned to OH
According to this catalyst deactivation mechanism, Mo/[B]ZSM-5
is expected to have higher stability for methane activation
reaction due to the nature of low acidity. Our result shows that
the catalyst is very stable through the run, as the rate only
decreases to 93% of initial reaction rate after 18 h.
-
1
stretching vibrations: the band at 3743 cm
attributed
to
terminal silanol groups,[
34]
-1
and the band at 3610 cm is typical
for the acidic bridging OH of Si–OH–Al.[ For the [B]ZSM-5, the
17]
OH vibration signal from silanol groups is slightly red-shifted to
The noisy data of methane reaction rate in Figure 9, is
presented here because the rate is calculated from a subtraction
of two large numbers (methane concentration before and after
reaction). Therefore a linear regression was fitted with the rate of
methane as a function of time on stream. A null hypothesis H =0
0
of the slope and the intercept is tested to verify if there is
significant linear relationship between the methane rate and the
-
1
3
735 cm , and the OH vibration band from Si–OH⋯B is
-1 -1
observed at 3724 cm and 3709 cm . In a zeolite framework,
the boron atom is coplanar with three oxygens due to its small
atomic diameter, and the distance to the fourth oxygen is too
long to form strong bond.[ Therefore the acidity of the bridging
OH in [B]ZSM-5 is much weaker than that of [Al]ZSM-5. After the
Mo impregnation the intensity of the IR bands decreased,
34]
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
time on stream. The P-value for the slope is ~0.001 and for the Conclusions
intercept is 0. In both cases, the null hypothesis is rejected,
suggesting that both slope and intercept value does not equal to
zero. The details of the t-test can be found in the supporting
information.
A
Mo
highly selective methane non-oxidative coupling catalyst
2
C/[B]ZSM-5 was synthesized and characterized. Boron is
employed instead of aluminium in the zeolite framework, which
leads to a dramatic decrease in the zeolite acidity. 2% wt. of
molybdenum was loaded into the zeolite by wetness
impregnation. After calcination, mono and binuclear Mo oxide
species were anchored onto the zeolite surface, as a high
3
2
1
0
1.0
0.8
[
29]
UV/VIS band gap energy (4.1 eV) was observed.
Rate (CH4)
Molybdenum oxides were reduced and carburized to form well
dispersed and isolated molybdenum carbides, as proved by the
absence of Mo-Mo pairs in the EXAFS measurement. Due to the
low acidity of the support, the catalyst shows a high selectivity
towards ethylene (~90%) instead of aromatics. Ethylene
oligomerization reaction, which is responsible for heavy coke
formation, was thus inhibited. As a result, no significant catalyst
deactivation was observed during the 18-hour long test.
Selectivity (C H )
2
4
Selectivity (C H )
6
6
0
0
0
.6
.4
.2
Experimental Section
0.0
1
.0 Catalyst preparation
0
2
4
6
8
10
12
14
16
18
[
B]ZSM-5 was synthesized by the hydrothermal method using
Time on Stream / h
tetrapropylammonium bromide (TPABr) as the organic structure director.
Fumed silica was used as the silica source and boric acid as the boron
source. Typically, 0.3112
Figure 9. Catalytic rates and selectivity of methane activation on
%Mo/[B]ZSM-5 at 923 K. The average rate is r=0.3 mmol h
3 3 2
g H BO (Sigma Aldrich), 4.5479 g SiO
Cabosil, M5), 0.4540 g NaOH (Fisher), 2.6201 g TPABr (Sigma Aldrich)
-
1
-1
2
gcat at ~0.8%
(
conversion. Based on an equilibrium conversion of 2.4%, the forward rate is
-
1
-1
were dispersed in 60 mL DI water and stirred until it became
homogeneous. The resulting gel was transferred to a 125 mL Teflon-
lined stainless steel autoclave (Parr 4748) and heated to 423 K for 3
days under static conditions. The as-made product was filtered, washed
with DI water and dried in an oven at 353 K overnight. After calcination at
823 K and NH
ammonium heptamolybdate (Sigma Aldrich) was impregnated onto the
zeolite support. Before reaction, the catalyst was calcined at 773 K and
carburized at 973 K.
estimated as r = 0.34 mmol h g .
+
c
a
t
The Mo/[Al]ZSM-5 has been extensively studied since Wang[5]
+
first discovered its novel property of methane direct
4
4 3
ion (0.05 M NH NO , Sigma Aldrich) exchange, 2%
dehydroaromatization to benzene.[
3, 6-11, 14, 32, 36]
As shown in
2 2
Scheme 1, in comparison to Mo C/[Al]ZSM-5, Mo C/[B]ZSM-5
showed 90% selectivity towards ethylene, with only 7%
selectivity to benzene. In contrast Li[ has reported a similar
37]
selectivity for this reaction with Mo
Mo C/[Al]ZSM-5 (88% benzene and 3% ethylene at initial and
8% benzene and 24% ethylene after 120 min). The difference
2
C/[B]ZSM-5 to that with
2
.0 Catalyst evaluation
2
6
The catalyst was evaluated in
a quarter inch quartz tube reactor.
in selectivity between the results reported here and those of Li et
al is due to the presence of aluminum impurities in their boron
source, impurities that would introduce strong acid sites into
their samples. Note too that a different catalyst synthesis
Typically, 50 mg of sample will be loaded into the reactor between two
layers of quartz wool. The reactor was installed in a flow reactor system,
which also includes a gas flow system, a heating system, and products
analysis system. Multiple gases can be fed using Brooks mass flow
controller E5850. The quartz reactor is heated by a furnace and the gas
line after the reactor are properly wrapped with heating tapes. An online
gas chromatography (GC, Shimadzu 2014) and a mass spectroscopy
method was applied in their report where Mo precursor MoO
3
and [B]ZSM-5 were physically mixed and grinded, although both
impregnation and physical mixture synthesis showed similar
methane conversion and products selectivity in the case of
(Pfeiffer, Thermostar GSD 320T) are employed to collect and analyse the
[7]
products.
[Al]ZSM-5, as Borry and co-workers reported.
3
3
.0 Catalyst characterization
.1 X-ray diffraction (XRD) and Transmission electron microscopy
(
TEM)
XRD analysis were performed using Bruker D8 x-ray powder
diffractometer with monochromatic Cu Kα1 line (k = 1.540 Å). TEM image
was obtained by
2
a JEM-2010F. N adsorption experiments were
performed with 3Flex surface characterization analyzer. BET Surfaces
were calculated using the MicroActive for 3Flex software.
3
.2 UV/vis spectroscopy
Diffused reflectance UV/vis study was carried out using a flow cell and a
Jasco V550 UV/vis spectroscopy. The in-situ flow cell is made of U-
shaped quartz round tube with a UV window made of square quartz tube
Scheme 1. Methane activation products selectivity of Mo
Mo C/[B]ZSM-5
2
C/[Al]ZSM-5 vs.
(Quartz Plus, Inc.). Typically 100 mg of the catalyst sample was loaded
2
into the flow cell at the window. Other auxiliary systems (a heating
This article is protected by copyright. All rights reserved.

ChemPhysChem
10.1002/cphc.201701001
FULL PAPER
system and gas flow control system) are included to perform the in-situ
UV/vis study. All the spectra were taken at room temperature with inert
gas flowing through the sample bed.
[13] W. P. Ding, G. D. Meitzner, E. Iglesia, Journal of Catalysis 2002, 206,
14-22.
[
[
14] D. Ma, Y. Lu, L. L. Su, Z. S. Xu, Z. J. Tian, Y. Xu, L. W. Lin, X. H. Bao,
Journal of Physical Chemistry B 2002, 106, 8524-8530.
15] H. Vankoningsveld, J. C. Jansen, H. Vanbekkum, Zeolites 1990, 10,
3.3 Temperature programmed oxidation (TPO)
235-242.
TPO experiment was conducted with the quartz reactor as mentioned in
section 2.0. About 100 mg of the calcined catalyst was loaded into the
reactor and a carburization was performed before the TPO experiment.
After the catalyst bed was cooled down to room temperature in He flow, a
[16] D. T. On, S. Kaliaguine, L. Bonneviot, Journal of Catalysis 1995, 157,
235-243.
[17] H. Koller, C.-Y. Chen, S. I. Zones, Topics in Catalysis 2015, 58, 451-
479.
[
[
[
18] A. Cichocki, J. Datka, A. Olech, Z. Piwowarska, M. Michalik, Journal of
the Chemical Society-Faraday Transactions 1990, 86, 753-756.
19] B. L. Meyers, S. R. Ely, N. A. Kutz, J. A. Kaduk, E. Vandenbossche,
Journal of Catalysis 1985, 91, 352-355.
20% O2/He gas was introduced with a ramping rate of 10 K/min until 823
K. The gas products were collected and analysed by
spectroscopy (Pfeiffer, Thermostar GSD 320T).
a mass
20] L. Wang, W. K. Hall, Journal of Catalysis 1982, 77, 232-241.
3
.4 X-ray absorption spectroscopy (XAS)
[21] F. A. Cotton, G. Wilkinson, C. A. Murillo, M. Bochmann, Advanced
Inorganic Chemistry, 6 ed., Wiley, New York, 1999.
In-situ XAS studies were carried out in Argonne National Lab (ANL) at
beamline 11B.The impregnated 2% wt. Mo/[B]ZSM-5 was loaded into an
in-situ reaction cell. The sample was first calcined at 773 K and then
carburized with 20% CH /H at 973 K. Full EXAFS spectra were taken at
[
22] J. H. Ashley, P. C. Mitchell, Journal of the Chemical Society a -
Inorganic Physical Theoretical 1968, 2821-&.
[40]
[
23] M. Fournier, C. Louis, M. Che, P. Chaquin, D. Masure, Journal of
Catalysis 1989, 119, 400-414.
4
2
the beginning and the end of each process. XANES spectra were taken
every 5 min during the in-situ treatment.
[24] L. G. van de Water, J. A. Bergwerff, T. A. Nijhuis, K. P. de Jong, B. M.
Weckhuysen, Journal of the American Chemical Society 2005, 127,
5024-5025.
[
[
25] P. C. Mitchell, F. Trifiro, Journal of the Chemical Society a -Inorganic
Physical Theoretical 1970, 3183-&.
26] N. T. Vasenin, V. F. Anufrienko, I. Z. Ismagilov, T. V. Larina, E. A.
Paukshtis, E. V. Matus, L. T. Tsikoza, M. A. Kerzhentsev, Z. R.
Ismagilov, Topics in Catalysis 2005, 32, 61-70.
3
.5 In situ Fourier transform infrared (FT-IR) characterization
The FT-IR measurements were performed using an in situ diffused
reflectance infrared Fourier transform (DRIFT) cell with ZnSe windows
(15 x 2 mm, Harrick Sci.). The IR cell consists of a cell body, sample
holder and aluminium heating block. About 50 mg of the catalyst sample
was required to load into the sample holder as a pellet. Proper gases
could be fed through the sample by mass flow controllers (Brooks 5850E
series). IR spectra were collected using a Fourier transform infrared
spectrometer (Nicolet Nexus 470). The IR spectral data were analysed
by OMNIC software. The catalyst sample was pre-treated with He at 773
K (5 K/min ramping rate) for 4 hours to remove moisture and other
surface contaminates. After cooling down to room temperature, IR
spectra were collected in He flow.
[27] A. Martinez, E. Penis, Applied Catalysis a-General 2016, 515, 32-44.
[28] M. W. Ngobeni, A. F. Carley, M. S. Scurrell, C. R. Nicolaides, Journal of
Molecular Catalysis a-Chemical 2009, 305, 40-46.
[
[
29] R. S. Weber, Journal of Catalysis 1995, 151, 470-474.
30] S. B. Xie, K. D. Chen, A. T. Bell, E. Iglesia, Journal of Physical
Chemistry B 2000, 104, 10059-10068.
[
31] J. Bedard, D.-Y. Hong, A. Bhan, Journal of Catalysis 2013, 306, 58-67.
[32] S. T. Liu, L. Wang, R. Ohnishi, M. Ichikawa, Journal of Catalysis 1999,
181, 175-188.
[
[
[
33] W. Li, G. D. Meitzner, R. W. Borry, E. Iglesia, Journal of Catalysis 2000,
91, 373-383.
34] J. Datka, Z. Piwowarska, Journal of the Chemical Society-Faraday
Transactions I 1989, 85, 47-53.
1
35] A. Mehdad, R. E. Jentoft, F. C. Jentoft, Journal of Catalysis 2017, 347,
89-101.
Acknowledgements
[36] V. T. T. Ha, L. V. Tiep, P. Meriaudeau, C. Naccache, Journal of
Molecular Catalysis a-Chemical 2002, 181, 283-290; D. Ma, Y. Y. Shu,
M. J. Cheng, Y. D. Xu, X. H. Bao, Journal of Catalysis 2000, 194, 105-
114; R. Ohnishi, S. T. Liu, Q. Dong, L. Wang, M. Ichikawa, Journal of
Catalysis 1999, 182, 92-103; J.-P. Tessonnier, B. Louis, S. Rigolet, M. J.
Ledoux, C. Pham-Huu, Applied Catalysis a-General 2008, 336, 79-88.
We gratefully acknowledge the funding from US Air Force. The
authors thank Dr. Jason Loiland and Chen-Yu Chou for their
help in the XAFS data collection at Argonne National Laboratory.
[
37] S. Li, D. Ma, Q. B. Kan, P. Wu, Y. B. Peng, C. L. Zhang, M. S. Li, Y. H.
Fu, J. Y. Shen, T. H. Wu, X. H. Bao, Reaction Kinetics and Catalysis
Letters 2000, 70, 349-356.
Keywords: methane activation • C1 chemistry • ethylene •
[
[
38] S. Y. Xing, D. H. Zhou, L. Cao, X. Li, Chinese Journal of Catalysis 2010,
31, 415-422.
molybdenum carbide • boron ZSM-5
39] D. H. Zhou, D. Ma, Y. Wang, X. C. Liu, X. H. Bao, Chemical Physics
Letters 2003, 373, 46-51.
[
[
1]
2]
J. H. Lunsford, Catalysis Today 2000, 63, 165-174.
A. T. Ashcroft, A. K. Cheetham, M. L. H. Green, P. D. F. Vernon, Nature
[40] P. J. Chupas, K. W. Chapman, C. Kurtz, J. C. Hanson, P. L. Lee, C. P.
Grey, Journal of Applied Crystallography 2008, 41, 822-824.
1991, 352, 225-226.
[
[
[
[
[
[
[
[
[
[
3]
4]
5]
6]
7]
8]
9]
T. V. Choudhary, E. Aksoylu, D. W. Goodman, Catalysis Reviews-
Science and Engineering 2003, 45, 151-203.
PwC, Vol. 2016, http://www.pwc.com/us/en/industrial-
products/publications/shale-gas-chemicals-industry-potential.html, 2013.
L. S. Wang, L. X. Tao, M. S. Xie, G. F. Xu, J. S. Huang, Y. D. Xu,
Catalysis Letters 1993, 21, 35-41.
B. M. Weckhuysen, D. J. Wang, M. P. Rosynek, J. H. Lunsford, Journal
of Catalysis 1998, 175, 338-346.
R. W. Borry, Y. H. Kim, A. Huffsmith, J. A. Reimer, E. Iglesia, Journal of
Physical Chemistry B 1999, 103, 5787-5796.
W. P. Ding, S. Z. Li, G. D. Meitzner, E. Iglesia, Journal of Physical
Chemistry B 2001, 105, 506-513.
Y. D. Xu, S. T. Liu, L. S. Wang, M. S. Xie, X. X. Guo, Catalysis Letters
1995, 30, 135-149.
10] F. Solymosi, J. Cserenyi, A. Szoke, T. Bansagi, A. Oszko, Journal of
Catalysis 1997, 165, 150-161.
11] D. J. Wang, J. H. Lunsford, M. P. Rosynek, Journal of Catalysis 1997,
69, 347-358.
12] B. M. Weckhuysen, M. P. Rosynek, J. H. Lunsford, Catalysis Letters
998, 52, 31-36.
1
1
This article is protected by copyright. All rights reserved.