The enhancement effects of BaX2 (X?=?F, Cl, Br) on SnO2-based catalysts for the oxidative coupling of methane (OCM)

Article abstract of DOI:10.1016/j.cattod.2020.04.010

In this study, the promotional effects of different barium halides (BaF₂, BaCl₂ and BaBr₂) on SnO₂ for OCM reaction have been investigated. It is observed that the addition of all the barium halides can improve the reaction performance of SnO₂, following the sequence of BaBr₂ > BaCl₂ > BaF₂. Raman and XRD results have substantiated that adding barium halides to SnO₂ can increase the amount of surface vacancies/defects, hence creating more abundant surface OCM reactive oxygen sites by anion substitution. At the same time, the amount of the moderate basic sites contributing to OCM reaction is also improved. During OCM reaction, the amount of the OCM reactive oxygen species as well as the moderate basic sites can be further increased. As proved by In situ Raman and XPS results, on BaF₂:SnO₂ = 1:1 and BaCl₂:SnO₂ = 1:1, O₂^? is the OCM reactive oxygen sites. In contrast, on BaBr₂:SnO₂ = 1:1, O₂^2- is the OCM reactive oxygen sites, which is induced by the formation of BaSnO₃ perovskite phase. Moreover, O₂^2- anions are reported to be favourable for the direct formation of C₂H₄ through carbene mechanism, which explains the high ethylene selectivity on BaBr₂-modified catalyst. The amount and properties of the surface active oxygen species and basic sites are of great importance for OCM reaction, and their concerted interaction determines the reaction performance of BaX₂/SnO₂ catalysts. BaBr₂:SnO₂ = 1:1 contains the apporiate amount of both types of active sites, hence displaying the best OCM performance among all the catalysts.

Full text of DOI:10.1016/j.cattod.2020.04.010

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
The enhancement effects of BaX2 (X = F, Cl, Br) on SnO2-based catalysts for
the oxidative coupling of methane (OCM)
Rong Xi1, Junwei Xu1, Yan Zhang, Zhixuan Zhang, Xianglan Xu, Xiuzhong Fang, Xiang Wang*
Key Laboratory of Jiangxi Province for Environment and Energy Catalysis, College of Chemistry, Nanchang University, Nanchang, 330031, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
In this study, the promotional effects of different barium halides (BaF2, BaCl2 and BaBr2) on SnO2 for OCM
reaction have been investigated. It is observed that the addition of all the barium halides can improve the
reaction performance of SnO2, following the sequence of BaBr2 > BaCl2 > BaF2. Raman and XRD results have
substantiated that adding barium halides to SnO2 can increase the amount of surface vacancies/defects, hence
creating more abundant surface OCM reactive oxygen sites by anion substitution. At the same time, the amount
of the moderate basic sites contributing to OCM reaction is also improved. During OCM reaction, the amount of
the OCM reactive oxygen species as well as the moderate basic sites can be further increased. As proved by In situ
Oxidative coupling of methane
SnO2-based catalysts
Barium halide promoters
BaSnO3
O2-and O22-active sites
−
Raman and XPS results, on BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1, O2 is the OCM reactive oxygen sites. In
2-
contrast, on BaBr2:SnO2 = 1:1, O2 is the OCM reactive oxygen sites, which is induced by the formation of
2-
BaSnO3 perovskite phase. Moreover, O2 anions are reported to be favourable for the direct formation of C2H4
through carbene mechanism, which explains the high ethylene selectivity on BaBr2-modified catalyst. The
amount and properties of the surface active oxygen species and basic sites are of great importance for OCM
reaction, and their concerted interaction determines the reaction performance of BaX2/SnO2 catalysts.
BaBr2:SnO2 = 1:1 contains the apporiate amount of both types of active sites, hence displaying the best OCM
performance among all the catalysts.
1. Introduction
to form C2H6, which is then converted into C2H4 via dehydrogenation
[2,5,6].
Methane accounts for more than 95 % of natural gas, and it also
occupies a relatively high proportion in shale gas [1,2]. Over the past
three decades, numerous studies have investigated direct or indirect
conversion of methane to value-added products including n-paraffins,
ethylene, hydrogen, methanol and dimethyl ether [1,2]. The indirect
conversion ways include dry reforming, steam reforming and partial
oxidation of methane, which first convert methane into syngas, and
then use syngas to produce more valuable products via Fischer-Tropsch
reaction. The direct conversion ways include non-oxidative methane
conversion into aromatics and oxidative coupling of methane into C2
hydrocarbons, which have not yet been industrialized due to the lack of
highly efficient catalysts [3,4]. Oxidative coupling of methane into C2
hydrocarbons has aroused great interest because of the attractive pro-
spect to transform directly inexpensive methane into ethylene, a critical
platform product in petrochemical industry. The process generally in-
cludes the activation of CH4 on the catalyst surface active sites to
generate CH3· radicals, followed by gas-phase coupling of CH3· radicals
Up to now, the most frequently studied catalyst systems are based
on Li/MgO and Mn/Na2WO4/SiO2. Li/MgO exhibits good initial reac-
tion performance, but suffers from an inherent instability due to the
vaporization of Li+ cation, which is believed to be a structure modifier
of the catalyst [7,8]. In contrast, Mn/Na2WO4/SiO2 exhibits not only
better reaction performance, but also long-term stability, which is re-
garded as the catalyst with the highest potential for industrial appli-
cation [9–11]. Whereas, over this catalyst, only 25–27% C2 product
yield can be obtained at a high temperature ∼ 850 °C [9–11]. Fur-
thermore, the nature of the active sites in Mn/Na2WO4/SiO2 have not
yet been understood very clearly due to the complex of the catalyst
[12]. Therefore, there is still strong incentive to seek for improved
catalysts that can be operated at relatively lower temperature with
enough C2 yield to industrialize this important reaction.
Since Keller and Bhasin first reported their work on OCM in early
1980s [13], more than 1800 catalysts used for this reaction have been
explored and reported [14]. By analyzing a large amount of literature
⁎ Corresponding author.
E-mail address: xwang23@ncu.edu.cn (X. Wang).
1 These authors contributed equally.
https://doi.org/10.1016/j.cattod.2020.04.010
Received 6 January 2020; Received in revised form 20 March 2020; Accepted 4 April 2020
0920-5861/©2020ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:RongXi,etal.,CatalysisToday,https://doi.org/10.1016/j.cattod.2020.04.010

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 1. XRD patterns of the catalysts, (a) fresh, (b) spent.
data, Takahashi et al. concluded that three kinds of materials are fa-
vorable for constructing OCM catalysts with considerable C2 yield,
which include alkaline earth metals, lanthanide metals and IIIA, VA, via
and VIIA group metals [14]. In the early stage, alkaline earth metal
oxides modified by alkali metals and some other un-modified or mod-
ified metal oxides have been adopted as typical catalysts to investigate
the mechanism of OCM reaction. It was found that C2 selectivity is
intimately related to the basicity of the catalysts, thus basic metal
oxides are regarded as appropriate catalysts for the reaction [3,15]. On
the other hand, pure or composite lanthanide oxides possessing surface
or bulk oxygen vacancies are reported to be beneficial to generate re-
active oxygen species for OCM reaction [16,17]. For instances, La2O3,
Sm2O3, Y2O3 and CeO2 have been investigated for the reaction, which
display certain selectivity to C2 products [16,17]. Moreover, on com-
posite lanthanide oxides such as ABO3 perovskites or A2B2O7 pyro-
chlores, abundant oxygen vacancies are present, which can induce the
generation of a large amount of surface active oxygen sites favorable for
OCM reaction [18–20]. It was formerly reported that using alkali metal
and alkali earth metal halides as catalyst promoter can improve the C2
selectivity because the halogen anions have a negative effect on me-
thane deep oxidation [21–23].
the modified catalysts exhibit significantly improved performance for
the reaction, due to the concerted interaction between the surface basic
sites, electrophilic oxygen species and halogen anions. A catalyst with a
BaBr2:SnO2 molar ratio of 1:1 displays the best promotional effects
among all of the catalysts, on which the highest C2 product yield of 18.0
% can be obtained, also with a high C2H4/C2H6 ratio of 4/1. By using
different characterization means, the relationship between the reaction
performance and the structure of the catalysts has been elucidated in
depth.
2. Experimental
2.1. Catalyst synthesis
A series of SnO2-based catalysts modified by BaX2 (X = F, Cl, Br)
with a BaX2: SnO2 molar ratio of 1:1 were synthesized with a physical
mixing method, using BaF2 (99.8 %), BaCl2.2H2O (99.99 %),
BaBr2.2H2O (99.8 %) and CVD-SnO2 as the precursors. In detail, the
mixture of BaX2 and SnO2 was moistened with a certain amount of
deionized water and ground in a mortar until it became a hard paste.
After drying at 120 ℃ overnight, the paste was ground into powder, and
then calcined in air at 800 ℃ for 4 h.
On the basis of the above details, it can be seen that a good OCM
catalyst requires usually high thermal stability, suitable basicity to ac-
tivate CH4 molecules, and certain amount of oxygen vacancies to gen-
erate selective oxygen species for the target product formation.
Moreover, the addition of halogen anions to some catalysts can ob-
viously improve the C2 selectivity and yield by avoiding methane deep
oxidation.
2.2. Catalyst characterization and reaction performance evaluation
OCM reaction has been used to evaluate the performance of the
catalysts. The bulk and surface properties of the catalysts have also
been explored by different means. The equipment models, operation
conditions, experimental procedures and parameters are listed in detail
in the supporting information.
As an n-type semiconductor with a high melting point at 1630 ℃,
SnO2 has been reported to be a very good oxidation catalyst due to its
abundant oxygen vacancies and reducible lattice oxygen species
[24,25]. But SnO2 related catalysts have been rarely used for OCM re-
action except that several publications are reported to adopt SnO2 as a
catalyst additive [26,27]. When using different alkali metal oxides to
modify SnO2 for OCM reaction, our group has found that a Li-modified
catalyst with a Sn/Li molar ratio of 5:5 exhibits the best reaction per-
formance, which depicts much better low temperature reaction per-
formance than Mn/Na2WO4/SiO2 due to the synergistic effects of the
abundant electrophilic oxygen species and moderate basic sites [25].
Chou et al. have investigated the mechanism of OCM reaction over
LiCl/SnO2, and found that adding Cl− anions can improve its reaction
performance due to the formation of higher concentration of surface
anion vacancies, more rapid oxygen mobility and improved redox
ability of Sn [28].
3. Result and discussion
3.1. XRD and N2-BET measurement of the catalysts
XRD techniques have been employed to identify the phase compo-
sitions of the fresh and spent catalysts, with the patterns displayed in
Fig. 1. For comparative study, the individual fresh and spent barium
halides were also analyzed, with the patterns exhibited in Fig. S1.
Compared to fresh barium halides, the spent ones exhibit no crystalline
phase change, indicating that they are stable during OCM process.
For all the fresh BaX2/SnO2 catalysts, tetragonal rutile SnO2 and the
corresponding barium halide diffraction peaks can be observed sepa-
rately, testifying that no solid phase reaction between SnO2 and barium
halides was taken place during the high temperature calcination pro-
cess. Compared to the fresh catalysts, after OCM reaction, while the
BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1 samples exhibit no any detect-
able change, a new BaSnO3 phase was detected for BaBr2:SnO2 = 1:1,
as indicated by the intensive diffraction peaks at 30.6, 43.8 and 54.4°.
Inspired by these work, with the objective to develop more efficient
catalysts, and to investigate the influence of alkaline earth metal halides
and halogen anions on the reaction performance of SnO2, BaF2, BaCl2
and BaBr2 have been employed to modify SnO2 to prepare catalysts for
OCM reaction. It is revealed that in comparison with pure SnO2, all of
2

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Table 1
Specific surface areas, lattice parameters and Raman quantification results of the fresh and spent catalysts.
Samples
Surface area (m2 g−1
)
Lattice parametera
A1/A1g
α=β=γ = 90°, a = b≠c (Å)
Fresh
Spent
Fresh
Spent
Fresh
Spent
SnO2
30
9
28
7
a = b = 4.735
a = b = 4.740
a = b = 4.765
a = b = 4.782
c = 3.178
a = b = 4.733
a = b = 4.739
a = b = 4.806
–
c=3.176
c=3.175
c=3.204
–
0
0
BaF2:SnO2 = 1:1
BaCl2:SnO2 = 1:1
BaBr2:SnO2 = 1:1
c = 3.177
c = 3.186
c = 3.193
0.06
0.08
0.17
0.10
0.20
–
7
5
4
3
a
Calculated using the Scherrer equation.
The initial BaBr2 and SnO2 phases can hardly be observed, indicating
that solid phase reaction occurred between SnO2 and BaBr2 during the
high temperature OCM process. To verify the formation of BaSnO3, a
pure BaSnO3 sample was also prepared intentionally by the same
method described in the experimental section, but using Ba(NO3)2 and
SnO2 as the precursors. XRD results in Fig. S1 (c) have proved that
BaSnO3 phase has indeed formed in BaBr2:SnO2 = 1:1 catalysts.
Compared with pure SnO2, the 2θ diffraction peaks of the SnO2
phase in the fresh BaX2/SnO2 catalysts shift more or less to lower an-
gles, which accompany the increase of the side lengths of the unit cell in
the order of BaF2 < BaCl2 < BaBr2, as listed in Table 1. This indicates
that the addition of barium halides has enlarged the unit cell volume of
SnO2 phase in the modified samples, especially for BaCl2 and BaBr2
modified catalysts. This enlargement effect of SnO2 cell may be a result
of anion substitutions. It is known that Sn4+ in SnO2 lattice is 6-fold
coordinated with a radius of 0.69 Å, and the 6-fold coordinated Ba2+
has a radius of 1.35 Å. Therefore, it is impossible for Ba2+ to substitute
Sn4+ in SnO2 lattice to form a solid solution structure [24,25].
Whereas, the radius of O2− in SnO2 lattice is 1.40 Å, and the radius of
F-, Cl- and Br- is 1.33, 1.81 and 1.96 Å, respectively, in the range to be
able to replace partially the lattice O2−. Therefore, It is rational to
propose that during the calcination process, a certain amount of O2−
anions in the SnO2 lattice could be replaced by the X- anions with larger
radii, hence expanding the SnO2 cell volume and side lengths [23]. As a
side effect, surface vacancies could be formed to maintain electric
neutrality, which is beneficial to generate surface mobile oxygen sites.
For BaCl2:SnO2 = 1:1 catalyst, Its side lengths of SnO2 phase become
evidently bigger after OCM reaction, indicating that the anion sub-
stitution continued during the reaction process [23].
BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1 catalysts, no obvious phase
change can be observed after OCM reaction, well consistent with the
XRD results. Nevertheless, it is apparent that significant phase change is
detectable for the spent BaBr2:SnO2 = 1:1 catalyst. To clarify this, the
Raman spectroscopy of the pure BaSnO3 compound was collected and
compared with the spent BaBr2:SnO2 = 1:1 catalyst in Fig. S2(c).
It was previously reported that in a perfect cubic BaSnO3 perovskite
structure with space group of Pm3m, three IR active modes with F1u
symmetry can be detected, but no Raman active modes can be observed
[29,30]. As hown in Fig. S2(c), the Raman bands observed for BaSnO3
and the spent BaBr2:SnO2 = 1:1 catalyst are likely to present defects in
the samples, which may locally lower the internal symmetry of the
perovskite phase, thus leading to unexpected Raman modes. A sharp
peak at 1060 cm−1 is observed, which is ascribed to the BaCO3 phase
[29,30]. Because the defective cubic BaSnO3 perovskite structure is an
A-site Schottky type disorder, which possesses an A-site vacancy cou-
pled with an oxygen vacancy [29]. In this case, the elimination of the
structure Ba cations happens, which would eventually convert into
BaCO3 in contact with the atmosphere CO2 or OCM byproduct CO2.
Except for the Raman peaks of BaSnO3, some Raman peaks belonging to
BaBr2 can be observed for the spent BaBr2:SnO2 = 1:1 catalyst, sug-
gesting that solid phase reaction between SnO2 and BaBr2 is not very
complete during the high temperature OCM process.
Compared to pure SnO2, all the fresh BaX2/SnO2 catalysts exhibit a
new peak at 575 cm−1, which can be attributed to interface or surface
phonon modes corresponding to lattice disorder and oxygen vacancies
[24,25]. We hence quantified the integrated area ratios (A1/A1g) of the
fresh and spent samples in Table 1. For the fresh catalysts, the A1/A1g
ratios
follows
the
order
of
BaBr2:SnO2 = 1:1 >
The specific surface areas of the catalysts were analyzed with N2-
BET. As displayed in Table 1, the addition of barium halide has de-
creased the specific surface area of the modified catalysts in comparison
with the individual SnO2. Compared with the fresh catalysts, after OCM
reaction, all the samples have only insignificant surface area decrease,
indicating that these SnO2-based samples are stable during the high
temperature reaction.
BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 > SnO2, indicating that the addi-
tion of barium halide to SnO2 can improve the vacancy amount on the
catalyst surface. For the spent catalysts, except for BaBr2:SnO2 = 1:1, it is
obvious that the A1/A1g ratios are larger than the corresponding fresh
catalysts
and
follow
the
order
of
BaCl2:SnO2 = 1:1 >
BaF2:SnO2 = 1:1 > SnO2. This supports the XRD results to testify that
anion substitution occurs to the modified catalysts, which also continues
during the OCM process. The presence of surface vacancies/defects are
believed to be favorable for the generation of surface active sites, such as
electrophilic oxygen species [16,17].
3.2. Raman spectra characterization of the catalysts
Raman technique was also used to investigate the compositions and
properties of the catalysts. For comparison, the spectra of individual
barium halides and BaSnO3 were also collected and shown in Fig. S2.
Compared with fresh barium halides, the Raman spectra of the spent
ones have no any change in Raman modes, testifying that they are
chemically stable during OCM process, well consistent with the XRD
results.
3.3. Reaction performance of the catalysts
The reaction performance of the catalysts has been evaluated for
OCM, with the results displayed in Fig. 3 and Table 2. For comparison,
the related pure barium halides were also tested for the reaction. Fig. S3
displays that all of them have poor reaction performance.
As formerly reported, SnO2 with a rutile crystalline phase belongs to
As depicted in Fig. 3, pure SnO2 exhibits constant CH4 conversion
around 10 % from 600 to 800 ℃, but has low selectivity to C2 products.
In contrast, all the BaX2/SnO2 catalysts exhibit evidently lower activity
than pure SnO2 below 650 ℃. But starting from 700 ℃, the reaction
performance of these modified catalysts begins to output it. Above this
temperature, the CH4 conversion, C2 product selectivity and yield im-
prove quickly with the increasing of the temperature. Obviously, the
D144h space group, whose active Raman modes can be distributed as B1g
,
Eg, A1g and B2g according to the group theory [24,25]. For the fresh
BaX2/SnO2 catalysts, as shown in Fig. 2(a), three typical Raman vi-
bration modes at 478 cm−1, 633 cm−1 and 774 cm−1 corresponding to
Eg, A1g and B2g modes can be obviously observed, which indicates that
SnO2 is the major phase in all the catalysts [24,25]. For spent
3

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 2. Raman spectra of the catalysts, (a) fresh, (b) spent.
catalytic performance of the catalysts follows the order of
BaBr2:SnO2 = 1:1 > BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 > SnO2.
BaBr2:SnO2 = 1:1 possesses the best OCM reaction performance among
all the catalysts. Particularly, it is noted here that this catalyst displays
also the highest C2H4/C2H6 ratio in all the samples (Fig. 3(d) and
Table 2), demonstrating that it owns also the surface active sites fa-
cilitating the formation of ethylene. The C2H4 selectivity, C2H4 yield
Table 2
The OCM performance of different catalysts at 800 °C.
Catalysts
Conversion (%) Selectivity (%)
Yield (%)
C2H4/
C2H6
C2
C2H4
C2
C2H4
0.4
SnO2
11.6
22.5
11.8
56.4
63.9
62.6
49.3
3.0
1.4
0.3
0.9
1.5
4.0
1.6
BaF2:SnO2 = 1:1
26.2
38.3
50.1
30.3
12.7 5.9
17.9 10.7
18.1 14.5
14.0 8.6
and
C2H4/C2H6
ratios
follow
the
sequence
of
BaCl2:SnO2 = 1:1 28.0
BaBr2:SnO2 = 1:1 28.9
BaBr2:SnO2 = 1:1 > BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 > SnO2 at
800 °C.
BaSnO3
28.4
As evidenced by XRD and Raman results, in the spent
BaBr2:SnO2 = 1:1, a large amount of BaSnO3 has been formed during
OCM reaction, which could play a critic role for the reaction perfor-
mance and selectivity to ehtylene. Therefore, the OCM performance of
pure BaSnO3 has also been evaluated. Fig. 3 shows that BaSnO3 has
much better OCM performance than SnO2, which is close to
BaF2:SnO2 = 1:1. However, its performance is still much lower than
BaBr2:SnO2 = 1:1. In this sample, beside the large amount of BaSnO3, a
small amount of BaBr2 is still present. This seems to indicate that the
concerted interaction between BaSnO3 and BaBr2 is necessary to get an
improved OCM reaction performance.
To get deeper understanding on the reaction process, both of CO2
and CO selectivity over the catalysts are also shown in Fig. S4. Fig. S4
(A) demonstrates that the CO2 selectivity on the catalysts decreases
Fig. 3. The OCM performance of the catalysts. (a) CH4 conversion, (b) C2 selectivity, (c) C2 yield, (d) C2H4/C2H6 ratio. Reaction conditions: CH4/O2 = 4:1 (nitrogen
as balance gas), WHSV = 18,000 mL−1 h−1 g−1
.
4

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 4. Long-term stability test of the catalysts at 800 °C. Reaction conditions: CH4/O2 = 4:1, (nitrogen as balance gas), WHSV = 18,000 mL−1 h−1 g−1
.
with the sequence of SnO2 > BaF2:SnO2 = 1:1 > BaCl2:SnO2
= 1:1 > BaBr2:SnO2 = 1:1, which is opposite to their C2 selectivity
trend. For the convenience of comparison, the added COx selectivity is
displayed in Fig. S4 (C). In comparison with the un-modified SnO2, with
the addition of different barium halides, the COx selectivity over all the
modified catalysts is significantly dropped above 650 °C. This indicates
that the presence of surface halogen anions can evidently inhibit the
deep oxidation of methane and C2 products, especially at elevated
temperature, which is favorable for the formation of coupling products.
Among the three kinds of barium halides, BaBr2 possesses the strongest
ability to restrict CO2 formation, which might explain its best OCM
reaction performance in all the samples.
obviously observed, testifying that BaF2 and BaCl2 re-dispersed on the
catalyst surface during the OCM reaction, and their interaction with
SnO2 could be altered. On the profile of BaBr2:SnO2 = 1:1 and BaSnO3,
the reduction becomes negligible, proving that the Sn4+ cations in the
lattice of BaSnO3 becomes non-reducible. The small peak observed for
BaBr2:SnO2 = 1:1 is attributed to the reduction of the residual amount
of remaining SnO2 in the catalyst.
3.5. O2-TPD analysis of the catalysts
Since surface oxygen sites are of great importance for CH4 activation
and C2 product formation during an OCM process, O2-TPD experiments
were conducted for the fresh and spent catalysts. Fig. 6 demonstrates
that two or three distinct desorption peaks can be observed for all the
samples, which can be classified into three groups.
The spent catalysts have been used to estimate their reaction sta-
bility at 800 °C. Fig. 4 exhibits that during 50 h’s tests, all the catalysts
possess very stable reaction performance, indicating that these catalysts
can stand high temperature, which is important for an OCM catalyst.
As formerly reported, the peaks below 400 °C are attributed to the
desorption of loosely bonded oxygen species, and the peaks at
400∼600 °C can be ascribed to the desorption of reactive oxygen spe-
cies beneficial to CH4 activation and C2 product formation [31–33]. The
peaks above 600 °C can be attributed to the desorption of some lattice
oxygen species, which might not be important for OCM reaction. After
reaction, the part of OCM reactive oxygen species obviously increases,
as evidenced by the enlarged peaks at 400∼600 °C in Fig. 6 (b).
Therefore, the oxygen desorption amount on the fresh and spent cata-
lysts are quantified in Table 3.
3.4. H2-TPR analysis of the catalysts
H2-TPR was used to investigate the redox properties of the fresh and
spent catalysts. As shown in Fig. 5(a), pure SnO2 depicts an un-
symmetrical reduction peak, which is attributed to the stepwise re-
duction of SnO2 via Sn4+ → Sn2+ → Sn0 [24,25,28]. In contrast, though
all the BaX2/SnO2 catalysts still display the major reduction peak at the
same temperature region, it is obviously broadened. On the profiles of
fresh BaCl2:SnO2 and BaBr2:SnO2 = 1:1, two seperated reduction peaks
can be evidently discerned. This indicates that barium halide addition
can influence the redox behavior of SnO2. However, in the fresh sam-
ples, no solid phase reaction occurred between BaX2 and SnO2 during
the calcination process, in good accordance to the XRD and Raman
results.
For the fresh catalysts, the addition of barium halides significantly
improves the amount of the OCM reactive oxygen sites as well as the
total amount of the desorbed oxygen species, which indicates that
barium halides can interact and modify the SnO2 surface property.
Raman results in Table 1 have testified that the addition of barium
halides can create more surface vacancies, thus being favorable for the
generation of surface active oxygen sites. For the fresh catalysts, the
Fig. 5(b) displays the H2-TPR profiles of all the spent catalysts. on
the profiles of BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1 catalysts, two
reduction peaks corresponding to the stepwise reduction process are
OCM
reactive
oxygen
amount
follows
the
order
of
BaBr2:SnO2 = 1:1 > BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 > SnO2,
5

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 5. H2-TPR profiles of the catalysts, (a) fresh, (b) spent.
matching very well the reaction performance of the catalysts. After
reaction, no significant change is observed for the O2-TPD profile of
pure SnO2, together with the amount of its OCM reactive oxygen spe-
cies. For the modified catalysts, the amount of the OCM reactive oxygen
species together with the total oxygen desorption amount became evi-
dently larger, indicating that the surface active sites experienced an
adjustment during the OCM reaction process, possibly due to the re-
dispersion of the BaX2 at high temperature. The oxygen desorption
amount obeys also the sequence of BaBr2:SnO2 = 1:1 >
BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 for the spent catalysts, which is
consistent to the reaction performance.
Table 3
O2-TPD Quantification Results of the fresh and spent catalysts.
Samples
O2 Desorption Amount (μmol/g)
Fresh
Spent
Loosely
OCM
Total Loosely
bounded
OCM
Total
bounded
reactive
reactive
SnO2
22.0
30.6
4.4
26.4
45.1
47.3
35.1
22.7
15.1
18.3
7.8
5.2
27.9
44.4
65.0
67.7
BaF2:SnO2 = 1:1
14.5
20.7
26.2
29.3
46.7
59.9
BaCl2:SnO2 = 1:1 27.3
BaBr2:SnO2 = 1:1 8.9
In brief, O2-TPD results have substantiated that the amount of sur-
face active oxygen sites plays a crucial role for OCM reaction on BaX2/
SnO2 catalysts. During the OCM reaction, the amount of the reactive
oxygen sites can be further improved.
the peak above 600 °C can be ascribed to strong basic sites
[19,25,31–33].
According to this classification, the CO2 desorption amount on the
fresh and spent catalysts is quantified in Table 4. Again, it is observed
that the addition of barium halides into SnO2 can significantly improve
the amount of the moderate basic sites, following the sequence of
BaBr2:SnO2 = 1:1 > BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 for both the
fresh and spent catalysts, no significant change is observed for the pure
SnO2 sample. In comparison with the modified fresh catalysts, the
amount of the moderate basic sites beneficial to the reaction is also
improved evidently for all the spent catalysts.
3.6. CO2-TPD analysis of the catalysts
It has been reported that C2 selectivity is closely related to the basic
properties of an OCM catalyst [14,25,31–33]. Therefore, CO2-TPD ex-
periments were performed to analyze the basic sites on the fresh and
spent catalysts, with the profiles shown in Fig. 7.
Some researchers pointed out that those active basic sites beneficial
to C2 selectivity are generally inactive at low temperature, because they
could be blocked by CO2 adsorption during the OCM reaction. By in-
creasing the reaction temperature, these active basic sites start to be
progressively available for methane activation due to desorption of
CO2. On the other hand, the strong basic sites can capture CO2 easily to
form stable carbonates, which also contribute negligibly to OCM reac-
tion [31,34,35]. Therefore, surface basic sites with moderate strength
could devote mostly to C2 product selectivity. Based on former reports,
the CO2 desorption peak below 300 °C can be attributed to weak basic
sites, the peak at 300∼600 °C corresponds to moderate basic sites and
It has been generally accepted that surface basic sites can be at-
−
tributed to the chemisorbed electrophilic oxygen species such as O2
,
O22-, O− and surface defects [31,34]. For example, by doping Li+ into
MgO lattice, Li+O− electron pair defects can be formed, which con-
tribute either to the basicity of the catalyst [31,33,36,37]. Therefore, it
believed that the improved basicity of the BaX2/SnO2 catalysts comes
from the formation of more OCM reactive oxygen species and surface
oxygen vacancies, as evidenced by the Raman and O2-TPD results. The
anion substitutions continued during the OCM process, thus resulting in
Fig. 6. O2-TPD analysis of the catalysts, (a) fresh, (b) spent.
6

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 7. CO2-TPD analysis of the catalysts, (a) fresh catalysts, (b) spent catalysts.
Table 4
on its Raman profile is assigned to micro crystalline BaCO3 instead of
O2-. In the following section, the in situ Raman results will prove that
this peak is inert to CH4, providing extra evidence to testify this as-
signment. Whereas, on the profile of the spent BaBr2:SnO2 = 1:1, a new
peak at 862 cm-1 is observed, which can be ascribed to active surface
CO2-TPD Quantification Results of the fresh and spent catalysts.
Samples
CO2 Desorption Amount (μmol/g)
Fresh
Weak
Spent
2-
O2 species [41,42,44], possibly induced by the formation of the new
Moderate
Total
Weak
Moderate
Total
BaSnO3 phase.
SnO2
0.10
0.08
0.05
0.12
0.24
0.50
0.62
0.92
0.34
0.58
0.67
1.04
0.07
0.22
0.08
0.07
0.30
0.71
1.16
1.29
0.37
0.93
1.24
1.36
BaF2:SnO2 = 1:1
BaCl2:SnO2 = 1:1
BaBr2:SnO2 = 1:1
3.7.2. The reactivity of the surface oxygen sites on the spent BaX2/SnO2
catalysts probed by in situ Raman
To further identify the active surface oxygen species and probe its
reactivity toward OCM reaction, another series of in situ Raman ex-
periments have been designed and performed with the spent catalysts.
Typically, a catalyst was first treated in a 30 mL/min O2 flow at 800 ℃
for 1 h and then flushed briefly in a 30 mL/min He flow. After these
steps, the feed gas was switched to a 30 mL/min CH4 flow. As shown in
the formation of more oxygen vacancies and improved basicity in the
spent catalysts.
3.7. In situ Raman analysis of the catalysts
Fig.
9
(a) and (b), over the spent BaF2:SnO2 = 1:1 and
BaCl2:SnO2 = 1:1 catalysts, the O2 peak at ∼1060 cm-1 decreases
gradually with the increase of the exposing time, and disappears com-
pletely after 15 min. After flushing briefly in a 30 mL/min He flow, the
feed was switched to a 30 mL/min CH4:O2 = 4:1 gas mixture flow for
−
Former publications have indicated that O2− superoxide species can
be stabilized at 200 °C on MgO-based catalysts and is the active oxygen
site for OCM [38]. Over CaO/Y2O3 OCM catalyst, Osada et al. found
−
−
5 min. It is apparent that the ∼1060 cm-1 Raman peak of O2 species
that O2 can even be stabilized up to 750 °C [39]. Moreover, over
2−
Na2O2, SrO2, and BaO2 catalysts, Otsuka et al. proved that O2
per-
can be restored. These results have indicated strongly that the surface
O2− is reactive to CH4 and is the reactive sites for OCM reaction. Under
reaction condition, it can be consumed and regenerated easily by the
gas phase O2.
oxide could be another type of reactive oxygen species for OCM reac-
2−
tion [40]. Lunsford et al. have provided the first evidence that O2 is
the OCM reactive species over Ba/MgO, La2O3 and Sr/La2O3 by using in
situ Raman spectroscopy [41,42]. Wan et al. have systematically in-
vestigated active oxygen species of the fuoride-containing rare earth
and alkaline earth-based catalysts for OCM by using in situ FTIR and
For the spent BaBr2:SnO2 = 1:1 catalyst, the species corresponding
to the ∼1060 cm−1 band is obviously inert to CH4 at 800 ℃, as sub-
stantiated by Fig. 9 (c). Therefore, this confirms that it should be as-
−
2-
Raman techniques [43]. They concluded that O2 species can be con-
cribed to the stable BaCO3 instead of surface O2- sites. Whereas, the O2
oxygen species corresponding to the peak at 862 cm−1 is obviously
reactive to CH4, and which can be regenerated by the gas phase O2.
In summary, In situ Raman results have proved strongly on BaX2/
SnO2 catalysts modified by different barium halides, varied surface
active oxygen sites contributing to OCM reaction have been generated.
sidered as OCM reactive oxygen species over the fuoride-containing
rare earth and alkaline earth metal oxide based catalysts [43].
3.7.1. Identifying surface oxygen species on the fresh and spent BaX2/SnO2
catalysts with in situ Raman
−
On BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1 catalysts, superoxide O2
To discern the active oxygen species over the fresh and spent cat-
alysts for OCM reaction, in situ Raman experiments have thus been
performed. All the samples were pretreated in a 30 mL/min O2 gas flow
at 800 ℃ for 1 h before taking the Raman spectra. Fig. 8(a) displays that
a peak at ∼ 1050 cm−1 can be evidently observed for all the fresh
catalysts, which is attributed to O2- oxygen species [43,44]. As mani-
fested in Fig. 8(b), on the spent BaF2:SnO2 = 1:1 and BaCl2:SnO2 = 1:1
catalysts, this O2- Raman peak can still be observed. As testified by XRD
and Raman results in above sections, during OCM reaction, no solid
phase reaction between SnO2 and BaF2/BaCl2 occurred, thus these two
samples are composed majorly of SnO2 phase. However, for the spent
BaBr2:SnO2 = 1:1 catalyst, a solid phase reaction between SnO2 and
BaBr2 happened during the OCM reaction to form defect BaSnO3, which
accompanies the generation of BaCO3. Therefore, the 1060 cm-1 peak
anions are formed as the active oxygen sites, but on BaBr2:SnO2 = 1:1
catalyst, peroxide O22- anions are formed as the active oxygen sites. The
2-
generation of active O2 species may come from the interaction be-
tween BaBr2 and BaSnO3, which was formed by the solid phase reaction
between SnO2 and BaBr2 during the high temperature OCM reaction
process.
3.8. XPS analysis of the oxygen properties on the surface of the fresh and
spent catalysts
It has been reported that surface electrophilic oxygen species such
as O2−, O22- and O- are beneficial to C2 product formation, but surface
lattice O2- could contribute to deep oxidation of CH4 and the OCM
7

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 8. In situ Raman spectra of the catalysts treated in O2 at 800℃ for 1 h, (a) fresh, (b) spent.
−
products [19,25,31,32]. The ratio of electrophilic oxygen species (O2
and O22-) to lattice (O2-) could be used to estimate the reaction per-
formance of an OCM catalysts, which can be quantified by XPS O1s
spectra [19,25,31,32]. Therefore, the surface oxygen properties of the
fresh and spent catalysts were particularly analyzed with XPS to get
supplementary information to other characterizations.
SnO2 < BaF2:SnO2 = 1:1 < BaCl2:SnO2 = 1:1 < BaBr2:SnO2 = 1:1. By
adding barium halides, the O2-/ O2- ratios of the fresh catalysts was
obviously increased, which could be a result of the anionic substitution
between halogen anions and lattice O2-, as observed by XRD. As a
consequence, a larger amount of oxygen vacancies can be formed, as
testified by Raman. The O1s analysis results of the fresh samples are
well consistent with the Raman, H2-TPR and O2-TPD results, and ex-
plain well the reaction performance.
3.8.1. XPS analysis of the O1s peaks for the fresh catalysts
−
Over the fresh catalysts, in situ Raman results have proved that O2
is the OCM reactive oxygen species. As depicted in Fig. 10, the O1s
spectra of the fresh catalysts can be deconvoluted into three peaks
corresponding to different oxygen species by assuming a single peak
having 80 % Gaussian plus 20 % Lorentzian peak shape, and fixing the
FWHM (full widths at half maximum) of a single XPS oxygen peak
between 1.8 and 2.0 eV. In sequence, they are O2- anions at 533.1 ∼
533.2 eV, oxygen species of the −OH groups on the SnO2 surface at
532.2 ∼ 532.4 eV and surface lattice O2- anions at 530.5∼ 530.8 eV
[19,25,29,30]. On the basis of this, the O2-/ O2- ratios of all the fresh
catalysts are quantified in Table 5, which obeys the order of
3.8.2. XPS analysis of the O1s peaks for the spent catalysts
The deconvoluted and curve-fitted O1s spectra of the spent catalysts
are exhibited in Fig. 11. According to Raman results, defect BaSnO3
compound is detected on the spent BaBr2:SnO2 = 1:1 catalyst as a
major component. For comparison, the O1s spectra of the pure defect
BaSnO3 was thus also included in the figure. With the same method
described as above, the O1s spectra of the spent catalysts are also de-
convoluted and fitted, with the O22−/O2− or O2-/O2− ratios quantified
in Table 6. Notably, five kinds of surface oxygen anions with different
chemical environments are totally detected on the spent catalysts
Fig. 9. In situ Raman spectra of the spent catalysts. (a) BaF2:SnO2 = 1:1, (b) BaCl2:SnO2 = 1:1, (c) BaBr2:SnO2 = 1:1.
8

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 10. XPS analysis of the O1s peaks for the fresh catalysts. (a) SnO2, (b) BaF2:SnO2 = 1:1, (c) BaCl2:SnO2 = 1:1, (d) BaBr2:SnO2 = 1:1.
for a 30 mol% BaCl2/ Gd2O3 catalyst which eventually induces the
formation of OCM reactive O2- species [23]. Therefore, it is rational to
believe that the same phenomenon occurs to the BaX2/SnO2 catalysts in
this study, as O2- species is detected on all the fresh catalysts.
Table 5
Curve-fitting and quantification results of XPS O1s spectra for the fresh cata-
lysts.
Catalysts
O1s B.E., FWHM (eV) and relative amount (%)
O2−/O2-
During the high temperature OCM reaction, no solid phase reaction
happened between BaF2/BaCl2 and SnO2 for BaF2:SnO2 = 1:1 and
BaCl2:SnO2 = 1:1 catalysts, due to the high bonding energies of Ba-F
(487 kJ/mol) and Ba-Cl (444 kJ/mol). Therefore, in comparison with
the fresh samples, the catalyst compositions have no evident change,
−
O2
OH−
O2−
SnO2
533.1/2.0/1.0 532.2/1.9/27.3 530.7/1.3/71.7 0.01
533.2/1.9/4.9 532.4/1.8/28.4 530.6/1.7/78.1 0.05
BaF2:SnO2 = 1:1
BaCl2:SnO2 = 1:1 533.2/1.8/5.6 532.2/1.8/15.3 530.7/2.0/79.1 0.07
BaBr2:SnO2 = 1:1 533.1/2.0/8.0 532.2/1.8/16.4 530.7/1.8/75.6 0.11
−
and O2 species is still the major OCM reactive sites on the two spent
catalysts. However, due to the much lower bonding energy of Ba-Br
(370 kJ/mol), a significant solid phase reaction took place between
BaBr2 and SnO2 during the OCM reaction to form defect BaSnO3, hence
[19,25,31–33], as listed in Table 6.
2-
In comparison with the corresponding fresh catalysts, all the spent
catalysts have improved O22−/O2− or O2-/O2− ratios, testifying the
further formation of the OCM reactive and selective oxygen sites during
the reaction process, possibly due to the continued substitution of the
lattice O2- by halogen anions to generate more oxygen vacancies. For
the spent BaBr2:SnO2 = 1:1, it is obviously observed that the amount of
lattice oxygen was decreased accompanying the amount increase of the
leading to the formation of more reactive O2 species on the spent
BaBr2:SnO2 = 1:1 catalyst. With the concerted interaction between the
left BaBr2 and BaSnO3, this catalyst displays the best reaction perfor-
mance among all the catalysts.
It has been commonly accepted that OCM reaction involved two
typical steps, that is, a heterogeneous step to activate CH4 molecules
into gas-phase CH3· radicals on the catalyst surface, and a homogeneous
gas-phase step to combine two CH3· radicals to form an C2H6 molecule
[2,5,6]. The production of C2H4 in OCM reaction can go through two
ways. One way can be attributed to the dehydrogenation of the formed
C2H6, and another way can be attributed to coupling of carbene species
generated by active oxygen species [46–48]. Actually, in most of the
cases, these two ways exist simultaneously during OCM reaction
[46–48]. As formerly reported, both O2− and O22- anions are the active
oxygen species contributing to C2H6 dehydrogenation [49]. However,
2-
active peroxide O2 species. The similar phenomenon is observed on
the pure defect BaSnO3 sample. This indicates that the presence of
BaBr2 and BaSnO3 phase in the sample contributes significantly to the
2-
generation of more amount of active O2 species. Based on molecular
2-
orbital theory, the bond order of O2 and O2- are 1.0 and 1.5 in se-
2-
quence. Therefore, surface O2 anions are more reactive than O2- an-
ions, which might be an important cause that BaBr2:SnO2 = 1:1 has the
best reaction performance among all the catalysts.
2-
O2 anions can active CH4 to form carbene species, which can be di-
rectly coupled into form C2H4 [48,49]. As mentioned above, O2- anions
are the active oxygen sites on both BaF2:SnO2 = 1:1 and
BaCl2:SnO2 = 1:1 catalysts, thus the way of C2H4 formation on these
two catalysts may preferentially go through the dehydrogenation of
3.9. A brief discussion
It was reported that exchange equilibrium between lattice O2− and
F- in the lattice of some fluoro-oxide can take place, thus leading to the
formation of O2- species and anion vacancies, which is beneficial to C2
selectivity for OCM reaction [45]. In addition, the diffusion of Cl- into
the Gd2O3 matrix to substitute part of the lattice O2- was also observed
2-
C2H6. In contrast, on BaBr2:SnO2 = 1:1 catalyst, O2 anions is the
predominant active oxygen sites, which were generated during the
OCM process due to the formation of defect BaSnO3. Therefore, the way
9

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 11. XPS analysis of the O1s peaks for the spent catalysts. (a) BaF2:SnO2 = 1:1, (b) BaCl2:SnO2 = 1:1, (c) BaBr2:SnO2 = 1:1 catalysts and (d) BaSnO3.
Table 6
Curve-fitting and quantification results of XPS O1s spectra for the spent catalysts.
or O2-/O2−
2−
Catalysts
O1s B.E., FWHM (eV) and relative amount (%)
O2
ratio
O2
OH/CO3
O2
O2−
−
2−
2−
BaF2:SnO2 = 1:1
BaCl2:SnO2 = 1:1
BaBr2:SnO2 = 1:1
BaSnO3
533.2/1.8/4.2
532.4/1.8/17.7
532.3/1.9/26.3
532.1/2.0/23.7
532.1/1.9/12.0
–
–
530.6/1.4/66.7
530.7/1.3/67.9
529.5/1.8/15.0
529.4/1.8/37.7
0.07
0.09
4.08
1.33
533.2/1.8/5.8
–
–
531.2/1.7/61.3
531.1/1.8/50.3
of C2H4 formation may preferentially go through the coupling of car-
bene species, but include also the dehydrogenation of the formed C2H6.
As a result, BaBr2:SnO2 = 1:1 exhibits extremely high C2H4/C2H6 ratio
in comparison with other catalysts.
larger amount of surface O2− sites contributing to OCM reaction on
the fresh samples, as evidenced by O2-TPD, in situ Raman and XPS
results.
(3) During the reaction process, the amount of O2− sites can be further
improved on all the spent catalysts except for BaBr2:SnO2 = 1:1. On
2-
4. Conclusion
this catalyst, a new type of more active oxygen sites, peroxide O2
anions, are generated due to the solid phase reaction between BaBr2
2-
Aiming to design and develop improved catalysts for OCM reaction,
SnO2 modified by different barium halides have been purposely de-
signed and prepared as catalysts in this work. By using different phy-
sical chemical means, the surface and bulk properties of the catalysts
have been characterized thoroughly, and correlated with the reactivity.
and SnO2 to form BaSnO3 perovskite phase. Moreover, O2 anions
are also favourable for the direct formation of C2H4 through car-
bene mechanism, which explains that an extremely high C2H4/C2H6
ratio is obtained on this catalyst.
(4) The amount and properties of the surface active oxygen species and
basic sites are important for OCM reaction, and their concerted
interaction determines the reaction performance of BaX2/SnO2
catalysts. BaBr2:SnO2 = 1:1 contains the apporiate amount of both
types of active sites, hence displaying the best OCM performance
among all the catalysts.
(1) In comparison with pure SnO2, all the catalysts modified by dif-
ferent barium halides exhibit evidently improved OCM reaction
performance above 650 °C, following the sequence of
BaBr2:SnO2 = 1:1 > BaCl2:SnO2 = 1:1 > BaF2:SnO2 = 1:1 > SnO2.
BaBr2 displays the best promotional effects among all the halides.
Over BaBr2:SnO2 = 1:1, the highest C2 product yield of 18.1 % can
be obtained at 800 ℃, which accompanies an extremely high C2H4/
C2H6 ratio of 4/1.
CRediT authorship contribution statement
Rong Xi: Data curation, Formal analysis, Investigation,
Methodology, Visualization, Writing - original draft. Junwei Xu: Data
curation, Formal analysis, Investigation, Methodology, Visualization,
Writing - original draft. Yan Zhang: Investigation, Formal analysis,
(2) XRD and Raman results have testified that the addition of different
barium halides into SnO2 can generate more surface vacancies/
defects due to anion substitution effect, thus producing obviously
10

R. Xi, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Methodology, Validation. Zhixuan Zhang: Investigation, Formal ana-
lysis. Xianglan Xu: Project administration, Resources, Software,
Funding acquisition. Xiuzhong Fang: Project administration,
Resources. Xiang Wang: Conceptualization, Data curation, Formal
analysis, Funding acquisition, Validation, Project administration,
Supervision, Writing - review & editing.
[14] K. Takahashi, I. Miyazato, S. Nishimura, J. Ohyama, ChemCatChem 10 (2018)
3223–3228.
[15] N.G. Maksimov, G.E. Selyutin, A.G. Anshits, E.V. Kondratenko, V.G. Roguleva,
Catal. Today 42 (1998) 279–281.
[16] T. Levan, M. Che, M. Kermarec, C. Louis, J.M. Tatibouet, Catal. Lett. 6 (1990)
395–400.
[17] S. Sugiyama, Y. Matsumura, J.B. Moffat, J. Catal. 139 (1993) 338–350.
[18] D.V. Ivanov, L.A. Isupova, E.Y. Gerasimov, L.S. Dovlitova, T.S. Glazneva,
I.P. Prosvirin, Appl. Catal. A Gen. 485 (2014) 10–19.
[19] A. Farsi, S. Ghader, A. Moradi, S.S. Mansouri, V. Shadravan, J. Nat. Gas Chem. 20
(2011) 325–333.
Declaration of Competing Interest
[20] J. Xu, L. Peng, X. Fang, Z. Fu, W. Liu, X. Xu, H. Peng, R. Zheng, X. Wang, Appl.
Catal. A Gen. 552 (2018) 117–128.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
[21] H. Wan, Z. Chao, W. Weng, X. Zhou, J. Cai, K. Tsai, Catal. Today 30 (1996) 67–76.
[22] R. Long, Y. Huang, W. Weng, H. Wan, K. Tsai, Catal. Today 30 (1996) 59–65.
[23] C.T. Au, K.D. Chen, C.F. Ng, Appl. Catal. A Gen. 171 (1998) 283–291.
[24] C. Rao, J. Shen, F. Wang, H. Peng, X. Xu, H. Zhan, X. Fang, J. Liu, W. Liu, X. Wang,
Appl. Surf. Sci. 435 (2018) 406–414.
Acknowledgement
[25] L. Peng, J. Xu, X. Fang, W. Liu, X. Xu, L. Liu, Z. Li, H. Peng, R. Zheng, X. Wang, Eur.
J. Inorg. Chem. 2018 (2018) 1787–1799.
The authors acknowledge deeply the financial supporting by the
National Natural Science Foundation of China (21962009, 21567016,
21666020), the Natural Science Foundation of Jiangxi Province
(20181ACB20005),the Key Laboratory Foundation of Jiangxi Province
for Environment and Energy Catalysis (20181BCD40004), which are
greatly acknowledged by the authors.
[26] K. Nagaoka, T. Karasuda, K.-i. Aika, J. Catal. 181 (1999) 160–164.
[27] Z. Wang, G. Zou, X. Luo, H. Liu, R. Gao, L. Chou, X. Wang, J. Nat. Gas Chem. 21
(2012) 49–55.
[28] F. Cheng, J. Yang, L. Yan, J. Zhao, H. Zhao, H. Song, L. Chou, React. Kinet. Mech.
Catal. 125 (2018) 675–688.
[29] J. Cerda, J. Arbiol, G. Dezanneau, R. Diaz, J.R. Morante, Sens. Actuators B Chem. 84
(2002) 21–25.
[30] W.F. Zhang, J.W. Tang, J.H. Ye, J. Mater. Res. 22 (2007) 1859–1871.
[31] J. Xu, Y. Zhang, X. Xu, X. Fang, R. Xi, Y. Liu, R. Zheng, X. Wang, ACS Catal. 9 (2019)
4030–4045.
Appendix A. Supplementary data
[32] Y.H. Hou, W.C. Han, W.S. Xia, H.L. Wan, ACS Catal. 5 (2015) 1663–1674.
[33] P. Huang, Y. Zhao, J. Zhang, Y. Zhu, Y. Sun, Nanoscale 5 (2013) 10844–10848.
[34] F. Papa, P. Luminita, P. Osiceanu, R. Birjega, M. Akane, I. Balint, J. Mol. Catal. A
Chem. 346 (2011) 46–54.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.cattod.2020.04.010.
[35] T.W. Elkins, B. Neumann, M. Bäumer, H.E. Hagelin-Weaver, ACS Catal. 4 (2014)
1972–1990.
References
[36] D.J. Driscoll, W. Martir, J.X. Wang, J.H. Lunsford, J. Am. Chem. Soc. 107 (1985)
58–63.
[37] S. Bernal, G. Blanco, A. El Amarti, G. Cifredo, L. Fitian, A. Galtayries, J. Martín,
J.M. Pintado, Surf. Interface Anal. 38 (2006) 229–233.
[1] A. Holmen, Catal. Today 142 (2009) 2–8.
[2] J.H. Lunsford, Catal. Today 63 (2000) 165–174.
[3] S. Gu, H.-S. Oh, J.-W. Choi, D.J. Suh, J. Jae, J. Choi, J.-M. Ha, Appl. Catal. A Gen.
562 (2018) 114–119.
[38] M. Iwamoto, J.H. Lunsford, J. Phys. Chem. 84 (1980) 3079–3084.
[39] Y. Osada, S. Koike, T. Fukushima, S. Ogasawara, T. Shikada, T. Ikariya, Appl. Catal.
59 (1990) 59–74.
[4] J.-Z. Zhang, M.A. Long, R.F. Howe, Catal. Today 44 (1998) 293–300.
[5] A.A. Latimer, A.R. Kulkarni, H. Aljama, J.H. Montoya, J.S. Yoo, C. Tsai, F. Abild-
Pedersen, F. Studt, J.K. Nørskov, Nat. Mater. 16 (2016) 225.
[6] K.D. Campbell, E. Morales, J.H. Lunsford, J. Am. Chem. Soc. 109 (1987)
7900–7901.
[40] K. Otsuka, T. Komatsu, Y. Shimizu, T. Inui (Ed.), Stud. Surf. Sci. Catal. Elsevier,
1989, pp. 43–50.
[41] J.H. Lunsford, X.M. Yang, K. Haller, J. Laane, G. Mestl, H. Knozinger, J. Phys. Chem.
97 (1993) 13810–13813.
[42] G. Mestl, H. Knozinger, J.H. Lunsford, J. Chem. Phys. 97 (1993) 319–321.
[43] H.L. Wan, X.P. Zhou, W.Z. Weng, R.Q. Long, Z.S. Chao, W.D. Zhang, M.S. Chen,
J.Z. Luo, S.Q. Zhou, Catal. Today 51 (1999) 161–175.
[7] T. Ito, J.X. Wang, C.H. Lin, J.H. Lunsford, J. Am. Chem. Soc. 107 (1985)
5062–5068.
[8] L. Luo, Y. Jin, H. Pan, X. Zheng, L. Wu, R. You, W. Huang, J. Catal. 346 (2017)
57–61.
[44] H.L. Wan, Z.S. Chao, W.Z. Weng, X.P. Zhou, J.X. Cai, K.R. Tsai, Catal. Today 30
(1996) 67–76.
[9] P. Wang, G. Zhao, Y. Wang, Y. Lu, Sci. Adv. 3 (2017) e1603180.
[10] N. Hiyoshi, K. Sato, Fuel Process. Technol. 151 (2016) 148–154.
[11] Z. Zhang, S. Ji, Comput. Chem. Eng. 90 (2016) 247–259.
[12] Y. Gambo, A.A. Jalil, S. Triwahyono, A.A. Abdulrasheed, J. Ind. Eng. Chem. 59
(2018) 218–229.
[45] X. Zhou, W. Zhang, H. Wan, K. Tsai, Catal. Lett. 21 (1993) 113–122.
[46] C.T. Au, H. He, S.Y. Lai, C.F. Ng, J. Catal. 163 (1996) 399–408.
[47] G.J. Hutchings, M.S. Scurrell, J.R. Woodhouse, Chem. Soc. Rev. 18 (1989) 251–283.
[48] C.T. Au, T.J. Zhou, W.J. Lai, C.F. Ng, Catal. Lett. 49 (1997) 53–58.
[49] H.X. Dai, C.F. Ng, C.T. Au, Appl. Catal. A Gen. 202 (2000) 1–15.
[13] G.E. Keller, M.M. Bhasin, J. Catal. 73 (1982) 9–19.
11