Low-valence or tetravalent cation doping of manganese oxide octahedral molecular sieve (K-OMS-2) materials for nitrogen oxide emission abatement

Article abstract of DOI:10.1039/c9cy01016e

Substitutionally doped oxide catalysts with abundant oxygen vacancy defects (OVDs) can effectively improve the catalytic activity efficiency. Here, transition metal (Zn, Fe, Zr, V, Mo, or W)-doped cryptomelane manganese oxide (K-OMS-2) catalysts were synthesized via a reflux method. The characterization and density functional theory plus U (DFT+U) results demonstrated that low-valence or tetravalent metals could be substituted for the Mn in the framework of K-OMS-2 catalysts, facilitating the production of OVDs. In contrast, higher-valence metals (Mo and W) compared to tetravalent manganese could not produce the corresponding Mo-K-OMS-2 and W-K-OMS-2, but only MoO₃ and WO₃, respectively. Substitutionally doping the Mn in the framework of K-OMS-2 with Zn, Fe, and Zr enhanced the selective catalytic reduction with NH₃ (NH₃-SCR) of nitrogen oxides, while dopants such as V, Mo, and W hindered the NH₃-SCR activity. Moreover, the higher average oxidation state (AOS) of manganese and the stronger activity of OVDs in low-valence or tetravalent metal-doped K-OMS-2 were two crucial factors in determining NH₃-SCR.

Full text of DOI:10.1039/c9cy01016e

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Low-valence or tetravalent cation doping of
manganese oxide octahedral molecular sieve (K-
OMS-2) materials for nitrogen oxide emission
abatement†
Cite this: DOI: 10.1039/c9cy01016e
Xiaomin Wu, Xiaolong Yu, Ziyi Chen, Zhiwei Huang and Guohua Jing
*
Substitutionally doped oxide catalysts with abundant oxygen vacancy defects (OVDs) can effectively im-
prove the catalytic activity efficiency. Here, transition metal (Zn, Fe, Zr, V, Mo, or W)-doped cryptomelane
manganese oxide (K-OMS-2) catalysts were synthesized via a reflux method. The characterization and den-
sity functional theory plus U (DFT+U) results demonstrated that low-valence or tetravalent metals could be
substituted for the Mn in the framework of K-OMS-2 catalysts, facilitating the production of OVDs. In con-
trast, higher-valence metals (Mo and W) compared to tetravalent manganese could not produce the corre-
Received 24th May 2019,
Accepted 5th July 2019
sponding Mo-K-OMS-2 and W-K-OMS-2, but only MoO3 and WO , respectively. Substitutionally doping
3
the Mn in the framework of K-OMS-2 with Zn, Fe, and Zr enhanced the selective catalytic reduction with
NH
Moreover, the higher average oxidation state (AOS) of manganese and the stronger activity of OVDs in
low-valence or tetravalent metal-doped K-OMS-2 were two crucial factors in determining NH -SCR.
3 3 3
(NH -SCR) of nitrogen oxides, while dopants such as V, Mo, and W hindered the NH -SCR activity.
DOI: 10.1039/c9cy01016e
rsc.li/catalysis
3
1
3
one. In addition, doping Ce with La, Gd, Pr, or Nb could
1
. Introduction
1
4
lower the activity toward n-butane oxidation. Therefore, the
design of doped oxides should be a promising direction for
enhancing catalytic reactions but not lowering the activity.
It is natural to inquire which dopants are most effective in
activating catalytic reactions. Principally, there are two kinds
of dopants that affect the catalytic activity. (1) Substitutional
dopants, which have a lower-valence or equivalent to their
parent oxide compared to the oxidation state of the cation in
Recent studies on substitutionally doped oxide catalysts have
proven that catalyst performance can be improved by replac-
ing a fraction of the cation with another cation. Indeed,
some dopants can weaken the bond between the oxygen
atoms in the surface layer and the metal crystal lattice, facili-
tating the production of oxygen vacancy defects (OVDs). The
oxygen atoms that are more easily extracted near the dopants
can be more reactive and lower the energy of OVD formation.
1
,2
3
the host oxide creating effective OVDs at the lattice position,
4
15,16
Thus, OVDs are good descriptors of photocatalysis, selective
can accelerate the catalytic efficiency.
(2) In contrast, dop-
catalytic reduction of NO combustion with NH3 (NH3-
ants, with a higher-valence than the cation they replaced in
the host oxide, have effective negative charges, resulting in
x
5
–8
9
SCR), and partial oxidation. For example, CO reacts much
more facilely with an oxygen atom of doped oxides than in
the absence of the dopant: doping enhances the Mars–van
1
7,18
excess electrons over the host oxide.
Moreover, higher-
valence cations tend to adsorb another O molecule from the
2
1
0,11
Krevelen mechanism for oxidation.
Fe-doped manganese
gas phase when replacing a cation in the host oxide (e.g., Zn
1
9
oxide octahedral molecular sieve (K-OMS-2) catalysts exhibited
much higher photocatalytic activity for benzene than pure K-
OMS-2, ascribed to the significant enhancement in the oxygen
oxide doped with Al), thus requiring much more energy.
Thus, it is helpful to explore and synthesize lower- or
equivalent-valence cation substitutionally doped metal oxide
catalysts, which can provide much more effective OVDs at lat-
tice positions.
1
2
activity of K-OMS-2 induced by Fe doping. However, doping
is not always beneficial. For example, Ce doped with La, Nb,
Sm, Gd, and Ta exhibited slower CO oxidation than cerium al-
In this article, we will focus on K-OMS-2 doped with Zn,
Fe, Zr, V, Mo and W to evaluate the NH -SCR of NO . We are
3
x
interested in K-OMS-2 because it has been widely used as a
Department of Environmental Science & Engineering, Huaqiao University, Xiamen
2
0
21
catalyst in degradation of ciprofloxacin, CO oxidation, re-
361021, P. R. China. E-mail: zhoujing@hqu.edu.cn
23
moval of NO (ref. 22) and VOCs. Furthermore, as a porous
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
x
c9cy01016e
manganese oxide with a composition of KMn O and a
8
16
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
2
4
structure consisting of 4.6 Å × 4.6 Å tunnels, K-OMS-2 has a
and 6.0 M HNO solution were mixed and reacted under re-
3
4
−
mixed-valence manganese framework due to the coexistence
flux at ∼100 °C for 24 h. The molar ratio between MnO and
4
+
3+ 25
2+
of Mn and a minor amount of Mn . This framework pro-
vides an opportunity for transition metal dopants, especially
for low-valence or tetravalent dopants, to be incorporated
into the Mn in the framework, subsequently promoting the
Mn was fixed at 0.72–0.76. After the reaction, the sample
was washed with deionized water thoroughly, and then dried
at 100 °C overnight. The synthesized samples were denoted
as K-OMS-2, Zn-K-OMS-2, Fe-K-OMS-2, Zr-K-OMS-2, V-K-OMS-
2, MoO3 and WO . Here, MoO and WO3 were represented
NH -SCR activity.
3
3
3
In a previous study, V-K-OMS-2 was reported in low-
because only MoO and WO phases formed after the addi-
3 3
temperature NO reduction with NH
3
, and the 2% V-K-OMS-2
tion of the Mo and W precursors during the synthesis pro-
cess, respectively.
catalyst showed the highest catalytic activity, because the se-
lected doping amount of V provided more Lewis acid sites
and maintained excellent surface redox ability for the adsorp-
2.2 Catalyst characterization
2
6
tion of NH3. However, the key properties of OVD activation
and the relationship between Mn and OVDs have not been
thoroughly studied. In our recent work, we found that a Ce-
doped K-OMS-2 catalyst presented good performance in low-
X-ray diffraction (XRD) was performed on a Bruker D8 Advance
diffractometer to detect the crystallographic structures. The
intensity data were collected in a 2θ range of 10–90° at 40 kV
and 30 mA. The morphology of the catalysts was observed by
field-emission scanning electron microscopy (FESEM, Hitachi
SU8020) and high-resolution transmission electron microscopy
2
7
temperature NH -SCR. Moreover, Ce-doped K-OMS-2 in the
3
framework with OVDs substantially accelerated the low-
temperature NH -SCR. These results inspired us to introduce
3
(
HRTEM, FEI Tecnai G20). The Brunauer–Emmett–Teller (BET)
transition metals, including Zn, Fe, Zr, V, Mo and W, into the
framework of K-OMS-2. This study will provide insight into
the role of OVD formation, and the relationship between Mn
and OVDs in transition metal-doped K-OMS-2. Thus, further
investigation of transition metal-doped K-OMS-2 for low-
surface areas were calculated from the adsorbed nitrogen vol-
ume using an automatic volumetric apparatus. Inductively
coupled plasma optical emission spectroscopy (ICP-OES) was
conducted with an Optima 7000 DV. X-ray photoemission
spectroscopy (XPS) experiments were carried out on a Thermo
Scientific Escalab 250Xi using the adventitious carbon (C 1s =
temperature NH -SCR may be possible. However, to the best
3
of our knowledge, implementing this strategy in low-
temperature NH -SCR is still in a rudimentary state.
3
Inspired by the above studies, we first incorporated vari-
ous transition metals, including Zn, Fe, Zr, V, Mo and W,
into the framework of K-OMS-2 materials. Characterization
techniques were employed to elucidate the dopant factors
2 2
284.6 eV). Temperature-programmed reduction of H (H -TPR)
was performed on a chemisorption analyzer equipped with a
thermal conductivity detector. Typically, prepared catalysts
(
0.10–0.20 g) were pretreated at 300 °C for 60 min in a flowing
stream of high-purity helium. After cooling down to room tem-
perature and introducing the reduction agent 5% H
flow rate of 50 mL min , the programming temperature was
2
/Ar at a
promoting the NH
3
-SCR activity by establishing potential re-
−1
lationships between the properties (e.g. structure, morphol-
ogy, composition, and reducibility) and catalytic perfor-
mance. We used density functional theory plus U (DFT+U)
and experiments to demonstrate the lower-valence or equiva-
lent cations, but not higher-valence cations, and doping of K-
OMS-2 can accelerate the catalytic activity. Furthermore, the
relationship between the Mn and OVDs, and the role of OVDs
controlled from room temperature to 600 °C at a rate of 10 °C
min .
−1
2
.3 Catalytic activity testing
The NH -SCR of NO was performed in a fixed-bed quartz re-
actor (i.d. = 6 mm) under atmospheric pressure. The catalyst
0.56 mL, 40–80 mesh) was charged for activity evaluation.
3
x
(
in the NH -SCR reaction were explored. After that, we will ex-
3
−
1
The feed gas (600 mL min ) was composed of 1000 ppm NO,
plicate the enhancement of doped K-OMS-2 catalytic activity
and the mechanism for NH -SCR.
3
1
000 ppm NH , 6.0% O , 150 ppm SO (when used), 10 vol%
3
2
2
H
2
O (when used) and balanced N
2
. The gas hourly space ve-
−
1
locity (GHSV) was 64 000 h , and the reaction temperature
2
. Materials and methods
was programmed from 60 °C to 350 °C. The concentrations
2
.1 Preparation of catalysts
2 x
of NO/NO /NO gases were determined using a flue gas ana-
lyzer (Testo 350). Finally, NO conversion was calculated by
the following equation: NO conversion = (NOin − NOout)/NOin
2
× 100% (NO and NO were the only nitrogen containing ox-
The transition metal-doped K-OMS-2 materials (denoted as
M-K-OMS-2, where M = Zn, Fe, Zr, V, Mo, or W) were pre-
2
8
pared by using a reflux method. During the synthesis of
transition metal-doped K-OMS-2, the nominal molar ratio of
transition metal (Zn, Fe, Zr, V, Mo, or W) to Mn for each sam-
ple was controlled at 0.12. The precursors of the dopants
were ZnIJCH COO) ·2H O, FeIJNO ) ·9H O, ZrIJNO ) ·5H O,
ides detected).
2.4 Computational details
To date, there are few studies on the effort of combining DFT
+U calculations with experimental techniques to thoroughly
reveal the effect of substituting transition metals with various
cation valences into K-OMS-2 on the catalyst surface
3
2
2
3 3
10
2
3 4
2
NH VO
4 3
, (NH
)
4 6
Mo
7
O
24, and (NH
4
)
W
12
O41·xH
2
O, respec-
tively. In a typical Zn-K-OMS-2 preparation procedure, a solu-
tion of 0.4 M KMnO , 1.7 M MnSO , 2.0 M ZnIJNO ) ·6H O
4
4
3 3
2
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
properties. K-OMS-2 was modeled with a primitive cell be-
2
9
longing to the I4/m space group as follows. The ground-
state magnetic order of K-OMS-2, according to Cockayne's
3
0,31
work,
in the K-OMS-2 framework. A transition metal-doped K-OMS-
model (1 × 1 × 2, K 32) was built. According to Zhou's
described the complex coupling between Mn atoms
2
2
Mn16O
3
2
report, the (100) surfaces of K-OMS-2 facilitate the adsorp-
tion/activation of O , and atomic slabs separated by a vacuum
2
layer of 15 Å were used to simulate the surface structure of K-
OMS-2. The Mn in the (100) surfaces of K-OMS-2 was re-
placed with different transition metals to construct the M-K-
OMS-2 (M = Zn, Fe, Zr, or V) model and form oxygen vacan-
cies, as shown in Fig. 1(a and b). The calculation process in-
cluded geometry optimization and energy calculation.
First-principles calculations were carried out using CASTEP,
based on the plane wave basis set and the ultrasoft pseudo-
Fig. 2 XRD patterns of different M-K-OMS-2 (M = Zn, Fe, Zr, V), MoO
and WO catalysts.
3
33
potential. The generalized gradient approximation (GGA) of
the Perdew–Burke–Ernzerhof (PBE) functional was used to ad-
dress the exchange correlation interactions. The energy cutoff
was 400 eV when doping with Fe, Zr, or V, and 550.0 eV when
the dopant was Zn. k-Point sampling was performed, using the
Monkhorst–Pack scheme with a 1 × 3 × 1 sampling grid based
on convergence studies. A Gaussian smearing of 0.1 eV was
used for Fermi surface broadening. To overcome the limita-
tions of the conventional GGA, for example the underestima-
tion of the band gap, Coulomb U corrections of 2.5 eV (Mn
3
of the diffraction lines were in agreement with the reported
data of the pure cryptomelane phase (KMn 16, JCPDS 29-
8
O
2
1
1020, space group I4/m). After doping with Zn, Fe, Zr and V,
only peaks that belong to the cryptomelane phase were
detected, which meant that the doped metals (Zn, Fe, Zr and
V) were incorporated into or well-dispersed in the K-OMS-2
structure. No additional peaks appeared in the M-K-OMS-2
(M = Zn, Fe, Zr, and V) materials, indicating that these cat-
ions were incorporated into the framework of K-OMS-2, as
evidenced by the corresponding Raman spectra (Fig. S1†).
The diffraction peaks of V-K-OMS-2 were broadened and de-
creased in intensity in comparison to those of the K-OMS-2
material, as evidenced by XRD. Such observations might indi-
cate the loss of long-range order, reduced crystallinity, and
decreased particulate sizes. In contrast, additional diffraction
3
1
34
35
36
3
3
d), 6.5 eV (Zn 3d), 4.0 eV (Fe 3d), 5.0 eV (Zr 4d), and
3
7
.0 eV (V 3d) electrons were induced.
3
. Results and discussion
3
.1 Crystal structure, oxidation state and BET surface areas
Fig. 2 illustrates the XRD patterns of the as-prepared M-K-
OMS-2 catalysts. In the case of pure K-OMS-2, the positions
peaks corresponding to crystalline MoO
3
(JCPDS 21-0569,
space group P) and WO (JCPDS 33-1387, space group P6/
3
mmm) phases appeared while the K-OMS-2 cryptomelane
phase was not formed during the preparation of Mo- or
W-doped K-OMS-2. This was because the dopants of
hexavalent molybdenum and tungsten possess higher valence
than the tetravalent manganese in K-OMS-2, and thus, much
more energy is required to produce Mo- or W-doped K-OMS-
2. Therefore, Mo-K-OMS-2 and W-K-OMS-2 were not easily
synthesized and the hexavalent cations of Mo and W could
not easily be substituted for Mn in the framework of K-OMS-
2.
The bulk mole% M, K/(Mn + M) and M/(Mn + M) and BET
surface area of the M-K-OMS-2 (M = Zn, Fe, Zr, or V) catalysts,
listed in Table 1, were determined by ICP and N adsorption
2
techniques. The bulk mole% Zn of Zn-K-OMS-2 was low
(
0.49) despite the initial molar ratio of Ce in the reactants be-
ing higher than 0.12. This was probably because the ionic ra-
2
+
n+
3+
dius of Zn (r = 0.90 Å) is larger than that of Mn (Mn
0
=
Fig. 1 The constructed (100) surface of K-OMS-2 (a), and M-K-OMS-2
4
+
.58 Å and Mn = 0.53 Å). The bulk mole% M (M = Fe, Zr, or
(
b) for calculation; purple (Mn), blue (Mn), red (O), pink (K), and grey-white
V) of the doped K-OMS-2 slightly deviated from the values
presented in the initial reaction mixtures. Interestingly, the
(M = Zn, Fe, Zr and V). The purple spheres stand for the Mn atoms in spin
up states and the blue spheres for the Mn atoms in spin down states.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 1 Quantitative data from the analysis of M-K-OMS-2 samples
ICP
Total pore
BET surface
area (m g
volume
(cm g
Average pore
size (nm)
a
b
2
−1
3
−1
Sample
Mole% M
K/(Mn + M)
Mn/(Mn + M)
Formula
AOS
)
)
K-OMS-2
0
0.08
0.09
0.09
0.09
0.08
—
1
K
K
K
K
K
0.64Mn
0.72Zn0.04Mn7.96
0.72Fe0.74Mn7.26
0.72Zr0.67Mn7.33
8
O
16
3.92
3.92
4.00
3.90c
3.79
—
82.0
113.0
101.8
290.7
199.1
255.2
130.7
0.79
0.48
0.38
0.36
0.23
0.31
0.32
38.7
28.6
15.0
4.9
4.7
5.0
Zn-K-OMS-2
Fe-K-OMS-2
Zr-K-OMS-2
V-K-OMS-2
0.49
8.62
7.50
10.02
—
0.995
0.907
0.916
0.892
—
O
16
O
16
O
16
0.64
V
0.86Mn7.14
O
16
MoO
3
—
—
WO
3
—
—
—
—
4.9
a
b
Determined by the ICP-AES analysis using the same strategy in ref. 26. AOS of Mn: calculated by the corresponding formula, (AOS = average
c
+
5+
2−
oxidation state). The AOS of manganese of V-K-OMS-2 by assuming that the K , V and O ion present in the V-K-OMS-2 samples.
bulk mole% V of the V-K-OMS-2 material calculated from the
ICP data reached 10.02, reflecting that the vanadium cations
incorporated into K-OMS-2 were both tetravalent and penta-
valent. Obviously, much more energy was required to form V-
K-OMS-2 if the pentavalent vanadium cations replaced the
tetravalent manganese cations of the host K-OMS-2 than if
tetravalent vanadium cations were used. As presented in
Table 1, the pure K-OMS-2 sample with the largest average
pore size (38.7 nm) showed a low specific surface area (82.0
with Zn, the morphology had been little affected because the
bulk mole% Zn of Zn-K-OMS-2 was low (mole% Zn = 0.49).
However, the nanorod length of the materials was sharply re-
duced after the addition of Fe, Zr, and V (Fig. 3(e–j)). Clusters
were also found among these short fibrous rods. This could
be due to the distortion of octahedral units after abundant
Fe, Zr and V substitution and the subsequent introduction of
4
0
OVDs. In contrast, the morphology of MoO , which had a
3
needle structure with more agglomeration, was markedly dif-
ferent from that of the other M-K-OMS-2 catalysts.
The morphology of the materials was confirmed by
HRTEM, as illustrated in Fig. 4. There was no lattice fringe in
2
−1
m g ). The surface area increased from 82.0 to 113.0, 101.8,
2
−1
2
90.7, and 199.1 m g with the addition of Zn, Fe, Zr and V,
respectively, indicating that the substitution of these metals
was beneficial to the increase of surface areas. Moreover, the
addition of these metals resulted in the formation of OVDs,
3 3
the morphology of MoO and WO (Fig. 4(f and g)), which
meant that the cryptomelane phase of K-OMS-2 did not form
during the hydrothermal synthesis process. This was in
agreement with the experimental observations in the XRD re-
sults. The dopants, including Zn, Fe, Zr and V, were
substituted for Mn in the framework of K-OMS-2 during the
nucleation and growth processes, and well-defined lattice
fringes were observed. Typically, the fringe distance of pure
K-OMS-2 was approximately 0.50 nm, which corresponds to
the interplane distance of the (200) planes of K-OMS-2. The
nanorod morphology did not change and the lattice fringe
was also 0.50 nm after Zn doping (see Fig. 4(b1–b3)). How-
ever, the nanorods of Fe-K-OMS-2, Zr-K-OMS-2, and V-K-OMS-
2 became short and aggregated in comparison to those in K-
OMS-2 and Zn-K-OMS-2. Furthermore, the lattice fringes be-
came complex and obscure with an increase in the valence
2
7
leading to a lower pore size and total pore volume.
The average oxidation state (AOS) of M-K-OMS-2 can be
2
6
determined by ICP-AES analysis or by XPS analysis of Mn
3
8,39
3
s.
Herein, the AOS was calculated from the formula by
using the results from ICP analysis, as depicted in Table 1.
The AOS value of the pure K-OMS-2 (3.92) sample in our
3
8
study was in agreement with that (3.9) of the reported,
suggesting that our method was reasonable.
The dopants, including Zn, Fe, and Zr, had little impact
on the AOS values of the catalysts, and the AOS of manganese
ranged in a narrow window of ca. 3.9–4.0. The AOS value of
3
+
Fe-K-OMS-2 increased slightly after doping with Fe ions.
However, the AOS of manganese in V-K-OMS-2 was decreased
5
+
4+
due to cycling between V and V . In this study, we calcu-
lated the AOS of manganese in V-K-OMS-2 by assuming that
2
6
state of dopants, suggesting the formation of OVDs. From
the XRD, ICP, FESEM and HRTEM experiments, it could be
+
5+
2−
the AOS values of the M-K-OMS-2 catalyst K , V and O
2
6
ions are represented in the samples. Thus, it could be con-
cluded that a high AOS of the various M-K-OMS-2 (M = Zn,
Fe, or Zr) materials was produced when transition cations
were introduced into the framework of K-OMS-2.
6
inferred that if the framework MnO octahedral units of K-
OMS-2 were substituted by low-valence or tetravalent cations,
2
+
3+
4+
4+
including Zn , Fe , Zr , and V , the relationship between
the dopant and Mn would be promoted, subsequently
resulting in the production of OVDs. In contrast, the high-
valence cations of hexavalent molybdenum and tungsten
would not generate the corresponding Mo-K-OMS-2 and W-K-
OMS-2, but only MoO and WO , respectively. During the hy-
3
.2 Morphologies and oxygen vacancy defects (OVDs)
The morphology of the as-prepared M-K-OMS-2 catalysts was
characterized by FESEM, and the images are shown in Fig. 3.
The pure K-OMS-2 catalyst displayed a typical fibrous mor-
phology with a length from 1 to 1.5 μm and a diameter of ap-
proximately 10–30 nm, as shown in Fig. 3(a). After doping
3
3
drothermal synthesis process, low-valence or tetravalent cat-
ions could be substituted for Mn in the framework of K-OMS-
4
+
2, while higher-valence cations compared to Mn could not
replace the Mn in framework of K-OMS-2.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Fig. 3 FESEM images of the various catalysts (a and b: K-OMS-2; c and d: Zn-K-OMS-2; e and f: Fe-K-OMS-2; g and h: Zr-K-OMS-2; i and j: V-K-
OMS-2; k and l: MoO ; m and n: WO ), 200 nm in (a, c, e, g i, k and n) and 1 μm in (b, d, f, h, j, l and m).
3
3
XPS spectra were employed to gain further insight into the
surface properties of the M-K-OMS-2 materials, and the re-
sults are shown in Fig. 5 and Table 2. Fig. 5(a) illustrates the
corresponding O 1s XPS spectra of the M-K-OMS-2 (M = Zn,
Fe, Zr, or V) materials. The asymmetrical O 1s spectra were
deconvoluted to two peaks. The peak at ∼529.8 eV could be
attributed to the lattice oxygen species (Olatt), and the other
at ∼531.4 eV corresponds to the surface adsorbed oxygen
uted to Zn being slightly incorporated into the framework of
K-OMS-2. This sequence indicated that the low-valence or tet-
ravalent cation doping of K-OMS-2 could effectively enhance
the composition of the higher active Oads species, which
would exhibit potential for OVD formation.
The Mn 2p 3/2 spectra of the M-K-OMS-2 catalysts are
shown in Fig. 5(b). The Mn 2p spectra can be deconvoluted
into two peaks located at ∼644.1 and ∼642.3 eV, ascribed to
4
1
4+
3+
(O
ads
).
O
species have a critical effect on catalytic NH -
Mn , and Mn , respectively. With the addition of low-
valence or tetravalent cations, the relative concentration of
ads
3
42
SCR, due to their higher mobility than Olatt species. In the
SCR reaction, Oads was combined to the gas phase and
extracted from the surface of the catalyst, leaving resultant
OVDs. The surface Oads/Olatt molar ratios of the M-K-OMS-2
4
+
3+
Mn /Mn increased as depicted by the peak-fitting result.
4+
3+
The increase in the ratio of Mn /Mn for M-K-OMS-2 indi-
cated that doping with low-valence or tetravalent cations in-
4
3,44
(M = Zn, Fe, Zr, or V) materials were calculated based on the
creased the oxygen vacancy defects.
peak-fitting results and are listed in Table 2. The Oads/Olatt
molar ratio followed the order Zr-K-OMS-2IJ35.9%) > Fe-K-
OMS-2IJ31.2%) > V-K-OMS-2IJ27.9%) > K-OMS-2IJ24.7%) ≈ Zn-
3
.3 Redox properties
K-OMS-2IJ24.4%). There was no difference in the surface O
/
ads
The H -TPR technique was applied to examine the reducibil-
2
Olatt molar ratio between Zn-K-OMS-2 and K-OMS-2, attrib-
ity of the M-K-OMS-2 (M = Zn, Fe, Zr, or V), MO and WO
3
3
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 5 XPS spectra of M-K-OMS-2 (M = Zn, Fe, Zr, V) catalysts over
the spectral regions of O 1s (a) and Mn 2p (b).
2 3 4 3 4
ment with transition of MnO to Mn O , and then Mn O to
4
5
MnO. The reducibility of the samples decreased in the se-
quence Zn-K-OMS-2 ∼ Zr-K-OMS-2 > V-K-OMS-2 ∼ K-OMS-2
>
Fe-K-OMS-2. The addition of Fe into K-OMS-2 caused the
reduction peaks to shift to high temperature, which might be
4
6
related to Fe promoting the reduction of Mn species. The
reducibility demonstrated by H -TPR could reflect the oxygen
Fig. 4 HRTEM images of the various catalysts (a1–a3: K-OMS-2; b1–
b3: Zn-K-OMS-2; c1–c3: Fe-K-OMS-2; d1–d3: Zr-K-OMS-2; e1–e3: V-
2
mobility in the catalysts, subsequently leading to OVD
formation.
3 3
K-OMS-2; f1–f3: MoO ; g1–g3: WO ).
catalysts, as shown in Fig. 6. For the addition of Mo and W,
there were no reduction peaks of MnO to MnO with Mn
and Mn O as intermediates, which implied that K-OMS-2
did not form during the hydrothermal synthesis process.
However, for the dopants Zn, Fe, Zr and V, the reduction of
all the catalysts occurred in a narrow temperature range of
Table 2 The XPS results of M-K-OMS-2 (M = Zn, Fe, Zr, V) about O1s
and Mn2p
2
2 3
O
3
4
Mn 2p 3/2
O 1s BE (eV)
BE (eV)
4
+
3+
4+
3+
Sample
O
latt
O
ads
Oads/Olatt%
Mn
Mn
Mn /Mn %
K-OMS-2
529.9 531.6 24.7
644.1 642.3 61.3
644.0 642.1 63.6
644.4 642.4 68.4
643.9 642.3 66.9
644.4 642.5 69.1
2
70–320 °C. Two principal peaks (peak I and II) were inferred
Zn-K-OMS-2 529.7 531.2 24.4
Fe-K-OMS-2 529.8 531.4 31.2
Zr-K-OMS-2 529.9 531.5 35.9
from the TPR analysis (see Fig. 6). For all the M-KOMS-2 ma-
terials, the ratio of peak I to peak II was approximately 2 : 1,
which is a typical feature of the reduction of MnO , in agree-
V-K-OMS-2
529.9 531.4 27.9
2
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
2 3
Fig. 6 The H -TPR profiles of M-K-OMS-2 (M = Zn, Fe, Zr, V), MoO
and WO
3
.
3
.4 Evaluation of the catalytic NH -SCR performance
3
Fig. 7(a) shows the activity of the as-prepared catalysts during
the NH -SCR reaction depending on temperature, and the GHSV
was controlled at 64 000 h . And the N
(M = Zn, Fe, and Zr) catalysts is presented in Fig. S2 .† As
3
−1
2
selectivity of M-K-OMS-
2
presented in Fig. 7(a), pure K-OMS-2 exhibited a satisfactory
low-temperature activity, which resulted in a NO conversion of
approximately 50% at 100 °C and achieved a 100% NO conver-
sion in the temperature range of 180–230 °C. Further increasing
the temperature led to a slight decrease in NO conversion at
Fig. 7 Catalytic performance of NH
3
-SCR over M-K-OMS-2 catalysts
(
M = Zn, Fe, Zr, V), MO and WO (a), and the effect of H O and SO
3
3
2
2
over Zr-K-OMS-2 at 150 °C (b). (Reaction conditions: NO = 1000 ppm,
NH3 = 1000 ppm, O2 = 6.0%, 10% H O (when used), 150 ppm SO
2
2
−
1
2
(when used), N to balance, and GHSV = 64 000 h ).
320 °C. After transition metal doping, the M-K-OMS-2 (M = Zn,
Fe, or Zr) catalysts exhibited the best NO conversion of 100% at
a lower temperature compared to pure K-OMS-2. However, for
V-K-OMS-2, the catalytic activity of NO decreased, which could
4+
3+
3.5 Computational details
be because the cycling between Mn and Mn was hindered
by the high-valence state of V. In contrast, the MoO and WO
3
3
Suib et al. performed computational calculations to study the
catalysts demonstrated completely different characteristics from
K-OMS-2, which exhibited poor low-temperature activity with no
more than 45% NO conversion below 320 °C.
effect of Nb doping and the energy of CO adsorption on the
4
3
active sites of K-OMS-2. However, the study did not men-
tion the importance and effect of OVD formation during Nb
substitution on the K-OMS-2 material. To date, structures
with dopants placed as the nearest-neighbor (NN) of K-OMS-2
have been shown to be energetically favorable, with similar
The influences of H O and SO on the catalytic perfor-
2
2
mance were also studied. Fig. 7(b) illustrates the effect of
O and SO on the performance of NH -SCR over the Zr-K-
H
2
2
3
47
OMS-2 catalyst at 150 °C. In the presence of 10% H O, the ac-
total energies.
2
tivity was decreased slightly (NO conversion > 85%). Upon
Taking into account these findings, we consider the for-
mation of OVDs at the NN site of the K-OMS-2 (100) surface
when introducing the transition metal dopant cations. In this
article, the incorporation of transition metal (Zn, Fe, Zr, or V)
into the (100) surface of K-OMS-2 was studied, and the energy
cutting off H
2
O, the activity could be recovered, indicating
that the influence of H O is reversible. However, when SO
2
2
(
150 ppm) was added or SO
added, the activity gradually becomes worse. In addition, SO
H O would accelerate the poisoning effects of the catalyst
2 2
(150 ppm) + H O (10%) were
2
+
of oxygen vacancy formation (E ) was calculated by the follow-
2
v
compared to SO
2
alone. After cutting off SO
2
or SO
2
+ H
2
O,
ing equation:
the catalytic activity could not be recovered. Therefore, the
addition of H O can prolong the NH -SCR activity of the cata-
lysts to a certain degree. However, SO resulted in irreversible
2
2
3
ΔE = (E
v
vac-M-K-OMS-2(100)
+ 1/2E ) − E
O
M-K-OMS-2(100)
2
deactivation because of the consumption of the active sites of
Zr-K-OMS-2 as NH HSO was blocked or metal sulfates
where EM-K-OMS-2IJ100), Evac-M-K-OMS-2IJ100) and EO2 are the energies
of the relaxed (100) surface slabs before and after introducing
4
4
formed.
a surface oxygen vacancy and the energy of O in a vacuum,
2
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
−
1
respectively. This equation indicates that oxygen atoms move
ture even at a high GHSV (64 000 h ). Obviously, low-valence
or tetravalent transition metal doping generally enhanced the
catalytic activity of K-OMS-2. It is a broad consensus that
away from the surface to form O molecules and oxygen va-
2
v
cancies. With our definition, a lower value of ΔE indicates
that the process is energetically preferable.
4
+
3+
mixed-valence (Mn and Mn ) in M-K-OMS-2 (M = Zn, Fe,
or Zr) catalysts is important for the catalytic activity of NOx
combustion, especially owing to the higher molar ratio of
Mn /Mn . Moreover, cycling between different valences of
Mn and Mn was beneficial for NH -SCR.
the AOS of manganese of M-K-OMS-2 (M = Zn, Fe, or Zr)
maintained a steady high value of ca. 3.9–4.0 after the addi-
tion of Zn, Fe, or Zr. Thus, this result implied that M-K-OMS-
2 (M = Zn, Fe, or Zr) had a mixed-valence manganese frame-
Moreover, since transition metal doping into K-OMS-2
usually induces structural deformations, the differences in
doping energy also needed to be considered. The dopants
were doped into the K-OMS-2 material in the form of ions.
The doping energy for the various transition metal dopants
at K-OMS-2 (100) was calculated by the following equation:
4
+
3+ 49
4
+
3+
50,51
In our work,
3
ΔE
d
= (EM-K-OMS-2(100) + EMn) − (EK-OMS-2(100) + E
M
)
4
+
work due to the coexistence of a major amount of Mn and
3
+
where EK-OMS-2IJ100) and EM-K-OMS-2IJ100) are the total energy of
the relaxed (100) surface slabs before and after introducing a
transition metal dopant, respectively. EMn and EM are the en-
ergy of one isolated Mn or dopant (M = Zn, Fe, Zr, or V)
atom, respectively.
a minor amount of Mn , similar to pure K-OMS-2. Previous
studies have reported that higher manganese valence could
lead to higher CO oxidation rates over the Cu(Co, Ag)-OMS-2
5
2
catalysts.
Ru/K-OMS-2 has shown potential for partial NO oxidation
Table 3 reports the substitution energy (E ) and the oxy-
into NO in view of fast SCR applications, attributed to the
higher reducibility and manganese valence of this catalyst.
d
2
5
3
gen vacancy formation energy (E
v
) when a dopant atom was
substituted for Mn in the framework of K-OMS-2. The doping
energy increased with increasing doping ion radius. The dop-
ing energy followed the order Zn-K-OMS-2 (6.55 eV) > Zr-K-
OMS-2 (2.39 eV) > Fe-K-OMS-2 (2.13 eV). The doping energy
for substituting a Mn atom in the oxide lattice by a Zn atom
Santos et al. also reported that K-OMS-2 with the highest AOS
presented the best catalytic activity for the oxidation of ethyl
3
8
acetate. Therefore, it could be concluded that the AOS
values of manganese of the M-K-OMS-2 (M = Zn, Fe, or Zr)
catalysts were in correlation with the corresponding catalytic
(
6.55 eV) was considerably larger than that for Fe, Zr and V.
3 x
NH -SCR of NO .
2
+
This was probably because the ionic radius of Zn (r = 0.90
Moreover, abundant Oads present on the surface of the
catalysts could favor the adsorption of gaseous NO and
4
+
Å) is remarkably larger than that of Mn (r = 0.53 Å). Fur-
thermore, the calculation of EV-K-OMS-2IJ100) failed even if the
energy cutoff was higher than 700 eV. This might be because
5
4
3
NH molecules. The Oads was extracted from the surface
of the catalyst, leaving resultant OVDs. After that, the
OVDs were subsequently replenished by O2 molecules.
Zhang et al. had reported that the decoration by TiO
4
+
the framework of V-K-OMS-2 has a mixed-valence of V and
5
+
V , causing the calculation system to not reach the ground
state. On the other hand, the oxygen vacancy formation en-
ergy was sharply decreased while Zn and Zr were substituted
for Mn in the framework of the K-OMS-2 catalyst. However,
for the Fe atom, the oxygen vacancy formation energy de-
creased slightly, which was mainly attributed to the magnetic
enhancement brought by Fe. Therefore, it could be deduced
from our calculation that low- and equivalent-valence cations,
including Zn , Fe , and Zr , could be substituted for Mn in
the framework of K-OMS-2, leaving resultant OVDs.
2
fur-
ther improved the redox properties by increasing the high
4
+
ratio of Mn and OVDs on the surface of a highly active
2
2
2
(OMS-5IJH)@TiO ) catalyst. Therefore, the higher surface
OVDs represented by the Oads/Olatt molar ratio could con-
tribute to the high NO emission abatement on M-K-OMS-
2 (M = Zn, Fe, or Zr). In this study, the framework MnO
6
octahedral units of K-OMS-2 substituted by low-valence or
tetravalent metals, including Zn, Fe and Zr, could have
higher OVDs, as evidenced by ICP, XPS, FESEM, HRTEM
and DFT+U. This abundant Oads could enhance the fur-
ther adsorption and reaction of reactants (NH3 and/or
NO), subsequently resulting in the production of OVDs.
The catalytic cycle is completed after the OVDs were
x
3
1
2
+
3+
4+
3
.6 Correlation between the structure and catalytic activity
As presented in Fig. 7, the M-K-OMS-2 (M = Zn, Fe, or Zr) cat-
alysts can achieve better NO conversion at a lower tempera-
replenished by O molecules.
2
3 x
For NH -SCR of NO over the M-K-OMS-2 (M = Zn, Fe, Zr)
catalysts at low-temperature, essentially, the higher AOS values
of manganese and the stronger activity of OVDs were two key
influencing factors in determining the catalytic activity. Besides,
all of the samples possessed basically high and similar intensi-
ties of weak acidic sites and moderate acidic sites, as evidenced
Table 3 The doping energy E
d
and oxygen vacancy formation energy E
v
(in eV) calculated via the DFT+U method
Substitution
ions
4
8
Model
Dopant radius r (Å)
E
d
(eV)
E
v
(eV)
3
+
4+
K-OMS-2
Zn-K-OMS-2 Zn
Fe-K-OMS-2 Fe
Zr-K-OMS-2 Zr
0.58IJMn ), 0.53IJMn
)
2.36
1.34
2.24
1.33
—
by ammonia temperature programmed desorption (NH
(Fig. S3†). Under the NH -SCR reaction conditions, the high
AOS of manganese on the surface of the catalysts reacted first
3
-TPD)
2
+
0.90
0.78
0.72
6.55
2.13
2.39
—
3
+
3
4
+
4
+
5+
4+
5+
V-K-OMS-2
V
, V
0.72IJV ), 0.54IJV
)
with NH
3
and/or NO, and the active Oads was consumed, leaving
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Funds of Huaqiao University (605-52418050), and the Scien-
tific Research Funds of Huaqiao University (605-50Y17071).
We also thank the Instrumental Analysis Center of Huaqiao
University for analysis support and the Subsidized Project for
Postgraduates' Innovative Fund in Scientific Research of
Huaqiao University.
Notes and references
1
2
3
4
R. Jacot, R. Moré, R. Michalsky, A. Steinfeld and G. R.
Patzke, J. Mater. Chem. A, 2017, 5, 19901–19913.
B. Li and H. Metiu, J. Phys. Chem. C, 2010, 114,
1
2234–12244.
H. L. Guo, Q. Zhu, X. L. Wu, Y. F. Jiang, X. Xie and A. W. Xu,
Nanoscale, 2015, 7, 7216–7223.
L. J. Liu, H. L. Zhao, J. M. Andino and Y. Li, ACS Catal.,
Fig. 8 Scheme of the reaction mechanism.
2
012, 2, 1817–1828.
resultant OVDs. The OVDs could activate the gaseous O
2
to sup-
5 J. Liu, X. Y. Li, Q. D. Zhao, J. Ke, H. N. Xiao, X. J. Lv, S. M.
Liu, M. Tade and S. B. Wang, Appl. Catal., B, 2017, 200,
297–308.
ply the consumed Oads. Therefore, the redox coupling between
4
+
3+
3
Mn /Mn and Oads/OVD was very important to the NH -SCR
activity. Then, the activated NO reacted with the neighboring
NH species to form N and H O (Fig. 8).
6 K. Lin, L. P. Han, J. B. He, H. R. Li, T. T. Yan, G. R. Chen,
J. P. Zhang, L. Y. Shi and D. S. Zhang, Environ. Sci. Technol.,
3
2
2
2
019, 53, 938–945.
4
. Conclusions
7 L. P. Han, M. Gao, C. Feng, L. Y. Shi and D. S. Zhang,
Environ. Sci. Technol., 2019, 53, 5946–5956.
Transition metal (Zn, Fe, Zr, or V)-doped K-OMS-2 catalysts
were successfully synthesized via a reflux method and charac-
terized, while only MoO and WO were formed after the ad-
dition of Mo and W precursors during the hydrothermal syn-
thesis process. Doping allowed the mixed oxides to be
activated, facilitating the OVD formation, which is a key prop-
erty in applications such as catalysis. Moreover, the higher
AOS of manganese in the doped K-OMS-2 was another
influencing factor in determining the catalytic activity, as
8
9
L. P. Han, M. Gao, J. Y. Hasegawa, S. X. Li, Y. J. Shen, H. R.
Li, L. Y. Shi and D. S. Zhang, Environ. Sci. Technol., 2019, 53,
3
3
6
462–6473.
N. J. Lawrence, J. R. Brewer, L. Wang, T. S. Wu, J. Wells-
Kingsbury, M. M. Ihrig, G. H. Wang, Y. L. Soo, W. N. Mei
and C. L. Cheung, Nano Lett., 2011, 11, 2666–2671.
1
0 H. Y. Kim, H. M. Lee, R. G. S. Pala, V. Shapovalov and H.
Metiu, J. Phys. Chem. C, 2008, 112, 12398–12408.
1 S. Chrétien and H. Metiu, Catal. Lett., 2006, 107, 143–147.
2 J. Chen, Y. Z. Li, S. M. Fang, Y. Yang and X. J. Zhao, Chem.
Eng. J., 2018, 332, 205–215.
1
1
2
shown by BET, TEM, H -TPR, and XPS characterization. Theo-
retical calculations (DFT+U) revealed that the lower energy re-
quired for OVD formation validated the activation of Oads by
the inclusion of dopants (Zn, Fe, or Zr) rather than other
transition metals (V, Mo, or W). Low-valence or tetravalent
metal (Zn, Fe, Zr, or V) doped K-OMS-2 catalysts, MoO and
3 3
WO were studied for the catalytic NH -SCR. Using a combi-
nation of theory and experiments, substitutionally doping the
Mn in the framework of K-OMS-2 with Zn, Fe or Zr was
3
shown to enhance the catalytic NH -SCR activity. The high
AOS of manganese on the surface of the catalysts reacted
with NH3 and NO molecules first, and the active Oads was
consumed, leaving resultant OVDs. The activated NO reacted
with neighboring NH species to form N and H O.
1
3 M. F. Wilkes, P. Hayden and A. K. Bhattacharya, J. Catal.,
2
003, 219, 286–294.
1
1
1
4 S. Zhao and R. J. Gorte, Appl. Catal., A, 2003, 248, 9–18.
5 M. Nolan, J. Phys. Chem. C, 2011, 115, 6671–6681.
6 A. K. Lucid, P. R. L. Keating, J. P. Allen and G. W. Watson,
J. Phys. Chem. C, 2016, 120, 23430–23440.
3
1
1
7 J. J. Carey, M. Legesse and M. Nolan, J. Phys. Chem. C,
2
016, 120, 19160–19174.
8 D. E. P. Vanpoucke, S. Cottenier, V. Van Speybroeck, I. Van
Driessche and P. Bultinck, J. Am. Ceram. Soc., 2014, 97,
2
58–266.
3
2
2
1
9 R. G. S. Pala, W. Tang, M. M. Sushchikh, J. N. Park, A. J.
Forman, G. Wu, A. Kleiman-Shwarsctein, J. P. Zhang, E. W.
McFarland and H. Metiu, J. Catal., 2009, 266, 50–58.
Conflicts of interest
2
2
0 L. L. Zhang, J. J. Tu, L. Lyu and C. Hu, Appl. Catal., B,
There are no conflicts to declare.
2
016, 181, 561–569.
1 V. P. Santos, S. A. C. Carabineiro, J. J. W. Bakker, O. S. G. P.
Soares, X. Chen, M. F. R. Pereira, J. J. M. Órfão, J. L.
Figueiredo, J. Gascon and F. Kapteijn, J. Catal., 2014, 309,
58–65.
Acknowledgements
This research was supported by the National Natural Science
Foundation of China (21806045), the Scientific Research
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
2
2 L. J. Yan, Y. D. Gu, L. P. Han, P. L. Wang, H. R. Li, T. T. Yan,
S. Kuboon, L. Y. Shi and D. S. Zhang, ACS Appl. Mater.
Interfaces, 2019, 11, 11507–11517.
39 V. R. Galakhov, M. Demeter, S. Bartkowski, M. Neumann,
N. A. Ovechkina, E. Z. Kurmaev, N. I. Logachevskaya, Y. M.
Mukovskii, J. Mitchell and D. L. Ederer, Phys. Rev. B:
Condens. Matter Mater. Phys., 2002, 65, 102–113.
40 Y. S. Wu, S. S. Yuan, R. Feng, Z. C. Ma, Y. Z. Gao and S. T.
Xing, Mol. Catal., 2017, 442, 164–172.
41 T. Boningari, D. K. Pappas, P. R. Ettireddy, A. Kotrba and
P. G. Smirniotis, Ind. Eng. Chem. Res., 2015, 54, 2261–2273.
42 T. Boningari, P. R. Ettireddy, A. Somogyvari, Y. Liu, A.
Vorontsov, C. A. McDonald and P. G. Smirniotis, J. Catal.,
2015, 325, 145–155.
2
2
3 J. T. Hou, Y. Z. Li, M. Y. Mao, Y. Z. Yue, G. N. Greaves and
X. J. Zhao, Nanoscale, 2015, 7, 2633–2640.
4 C. Calvert, R. Joesten, K. Ngala, J. Villegas, A. Morey,
X. F. Shen and S. L. Suib, Chem. Mater., 2008, 20,
6
382–6388.
2
5 Y. S. Ding, X. F. Shen, S. Sithambaram, S. Gomez, R. Kumar,
V. M. B. Crisostomo, S. L. Suib and M. Aindow, Chem.
Mater., 2005, 17, 5382–5389.
2
2
2
6 L. A. Sun, Q. Q. Cao, B. Q. Hu, J. H. Li, J. M. Hao, G. H. Jing
and X. F. Tang, Appl. Catal., A, 2011, 393, 323–330.
7 X. M. Wu, X. L. Yu, X. Y. He and G. H. Jing, J. Phys. Chem. C,
43 H. C. Genuino, M. S. Seraji, Y. T. Meng, D. Valencia and
S. L. Suib, Appl. Catal., B, 2015, 163, 361–369.
44 J. T. Hou, Y. Z. Li, L. L. Liu, L. Ren and X. J. Zhao, J. Mater.
Chem. A, 2013, 1, 6736–6741.
45 J. Z. Ma, C. X. Wang and H. He, Appl. Catal., B, 2017, 201,
503–510.
2
019, 123, 10981–10990.
8 M. Sun, W. P. Li, B. T. Zhang, G. Cheng, B. Lan, F. Ye, Y. Y.
Zheng, X. L. Cheng and L. Yu, Chem. Eng. J., 2018, 331,
6
26–635.
46 Y. B. Xu, X. L. Jia and X. H. Liu, Catal. Sci. Technol., 2018, 8,
1953–1970.
47 D. G. Pintos, A. Juan and B. Irigoyen, J. Phys. Chem. C,
2013, 117, 18063–18073.
48 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751–767.
2
9 J. Luo, Q. Zhang, J. Garcia-Martinez and S. L. Suib, J. Am.
Chem. Soc., 2008, 130, 3198–3207.
0 E. Cockayne and L. Li, Chem. Phys. Lett., 2012, 544, 53–58.
1 Y. P. Duan, Z. Liu, Y. H. Zhang and M. Wen, J. Mater. Chem.
C, 2013, 1, 1990–1994.
3
3
3
3
3
3
3
3
3
2 J. Zhou, L. F. Qin, W. Xiao, C. Zeng, N. Li, T. Lv and H. Zhu,
49 S. C. Deng, T. T. Meng, B. L. Xu, F. Gao, Y. H. Ding, L. Yu
and Y. N. Fan, ACS Catal., 2016, 6, 5807–5815.
50 G. S. Qi, R. T. Yang and R. Chang, Appl. Catal., B, 2004, 51,
93–106.
51 L. Qu, C. T. Li, G. M. Zeng, M. Y. Zhang, M. F. Fu, J. F.
Ma, F. M. Zhan and D. Q. Luo, Chem. Eng. J., 2014, 242,
76–85.
52 G. G. Xia, Y. G. Yin, W. S. Willis, J. Y. Wang and S. L. Suib,
J. Catal., 1999, 185, 91–105.
53 S. Adjimi, J. M. Garcia-Vargas, J. A. Diaz, L. Retailleau, S. Gil,
M. Pera-Titus, Y. Guo and A. Giroir-Fendler, Appl. Catal., B,
2017, 219, 459–466.
Appl. Catal., B, 2017, 207, 233–243.
3 Z. L. Song, Z. G. Yan, X. J. Yang, H. Bai, Y. H. Duan, B. Yang
and L. Leng, Chem. Phys. Lett., 2018, 695, 216–221.
4 D. Huang and C. Persson, Thin Solid Films, 2013, 535,
2
65–269.
5 Z. Jiang, W. H. Zhang, W. F. Shangguan, X. J. Wu and Y.
Teraokag, J. Phys. Chem. C, 2011, 115, 13035–13040.
6 H. F. Wang, H. Y. Li, X. Q. Gong, Y. L. Guo, G. Z. Lu and P.
Hu, Phys. Chem. Chem. Phys., 2012, 14, 16521–16535.
7 S. H. Chen, Z. R. Xiao, Y. P. Liu, P. H. Lee and Y. K. Wang,
J. Magn. Magn. Mater., 2011, 323, 176–179.
8 V. P. Santos, M. F. R. Pereira, J. J. M. Órfão and J. L.
Figueiredo, Top. Catal., 2009, 52, 470–481.
54 C. X. Wang, J. Z. Ma, F. D. Liu, H. He and R. D. Zhang,
J. Phys. Chem. C, 2015, 119, 23119–23126.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019