Fabrication of homogeneous and highly dispersed CoMn catalysts for outstanding low temperature catalytic oxidation performance

Article abstract of DOI:10.1039/c9nj03450a

A series of homogeneous and highly dispersed CoMnO_x bimetallic oxides with different ratios were prepared through pyrolysis of CoMn-MOF-71, which was applied to the catalytic oxidation of toluene. Notably, the effect of Co/Mn molar ratio on the catalytic activity of toluene oxidation was systematically investigated. After calcination, the samples consisted of many uniformly distributed nanoparticles, which could be seen from the TEM images. In particular, the Co₁Mn₁O_x alloy catalyst exhibited superior low temperature catalytic oxidation performance. The complete conversion temperature of toluene (T₁₀₀) was 180 °C, which is a relatively low temperature in the field of toluene oxidation. Through a series of characterizations, it was concluded that the different molar ratios of cobalt and manganese could obviously change the surface atomic ratio of Co³⁺/Co²⁺ and (Mn⁴⁺ + Mn³⁺)/Mnⁿ⁺. Simultaneously, we discovered that an appropriate amount of manganese improved the catalytic performance with an increase in the surface atomic ratio of (Mn⁴⁺ + Mn³⁺)/Mnⁿ⁺. In addition, lots of surface chemically adsorbed oxygen would speed up the catalytic oxidation of toluene. Furthermore, no significant decrease in the catalytic performance was observed over 14 h on the Co₁Mn₁O_x sample, which indicated that it exhibited excellent stability for toluene oxidation. Therefore, it showed potential as a metal-oxide catalyst in the elimination of toluene.

Full text of DOI:10.1039/c9nj03450a

NJC
View Article Online
View Journal
PAPER
Fabrication of homogeneous and highly dispersed
CoMn catalysts for outstanding low temperature
catalytic oxidation performance
Cite this:DOI: 10.1039/c9nj03450a
a
a
Weigao Han,ab Fang Dong,
Weiliang Hana and Zhicheng Tang
*
A series of homogeneous and highly dispersed CoMnOx bimetallic oxides with different ratios were
prepared through pyrolysis of CoMn-MOF-71, which was applied to the catalytic oxidation of toluene.
Notably, the effect of Co/Mn molar ratio on the catalytic activity of toluene oxidation was systematically
investigated. After calcination, the samples consisted of many uniformly distributed nanoparticles, which
could be seen from the TEM images. In particular, the Co1Mn1Ox alloy catalyst exhibited superior low
temperature catalytic oxidation performance. The complete conversion temperature of toluene (T100) was
180 1C, which is a relatively low temperature in the field of toluene oxidation. Through a series of
characterizations, it was concluded that the different molar ratios of cobalt and manganese could obviously
change the surface atomic ratio of Co3+/Co2+ and (Mn4+ + Mn3+)/Mnn+. Simultaneously, we discovered that
an appropriate amount of manganese improved the catalytic performance with an increase in the surface
atomic ratio of (Mn4+ + Mn3+)/Mnn+. In addition, lots of surface chemically adsorbed oxygen would speed up
the catalytic oxidation of toluene. Furthermore, no significant decrease in the catalytic performance was
observed over 14 h on the Co1Mn1Ox sample, which indicated that it exhibited excellent stability for toluene
oxidation. Therefore, it showed potential as a metal-oxide catalyst in the elimination of toluene.
Received 3rd July 2019,
Accepted 19th July 2019
DOI: 10.1039/c9nj03450a
rsc.li/njc
attracted great attention due to their good catalytic oxidation
of VOCs. Among them, Co-based catalysts have relatively high
1. Introduction
Volatile organic compounds (VOCs) are mainly emitted from catalytic activity, especially Co3O4. Ren et al.10 prepared different
transportation, the petrochemical industry and the use of morphologies of Co3O4 catalyst to remove VOCs. The results
organic solvents. Some of the VOCs damage public health. showed that the 3D hierarchical cube-stacked Co3O4 micro-
They have become a major source of air pollution.1,2 Compared spheres exhibited the best activity for total oxidation of VOCs,
with other control technologies, catalytic oxidation has the compared with the needle-stacked Co3O4 and plate-stacked
advantage of relatively low operating temperature and high Co3O4. Manganese oxide-based materials also exhibited great
efficiency. And it has attracted more and more attention. People potential. Piumetti et al. prepared three-dimensional meso-
also regard it as the most promising technology for the removal porous manganese oxide catalysts (Mn2O3, Mn3O4 and MnxOy)
of VOCs.3,4 Although the noble metal catalyst has a good and applied them to the oxidation of VOCs. The results
catalytic performance,5,6 it’s hard to apply it practically because showed that the Mn3O4-based catalyst exhibited a high catalytic
of its high price. Many efforts were made to replace the noble performance.11 Many efforts were also made to remove VOCs in
metal catalyst with a transition metal catalyst.7,8 But the transition the field of mixed metal catalysts. Gallegos et al. prepared ZnMnOx
metal catalysts don’t show a satisfactory performance especially at solid catalysts from wasted batteries to remove VOCs.12 Lin et al.
low temperatures.9
prepared metal (M = Cu, Co, Ni, Mn, Fe)-modified cerium oxide,
In the last decade, scientists have reported numerous transition which was supported on mesoporous silica particles.13 The results
metal oxides such as MnOx, Co3O4 and CuOx, and a large number showed that an appropriate amount of Mn could significantly
of mixed metal oxides. These transition metal oxides have promote the catalytic activity of Ce/Al-MSPs. Behar et al. prepared
Cu1.5Mn1.5O4 catalysts and studied the catalytic oxidation of
toluene at low concentrations.14 Zhang prepared LaMnO3 and
a State Key Laboratory for Oxo Synthesis and Selective Oxidation, National
LaB0.2Mn0.8O3 (B = Co, Ni, Fe) oxide catalysts and studied the
Engineering Research Center for Fine Petrochemical Intermediates,
oxidation of vinyl chloride.15 The results showed that the catalytic
Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences,
activity of the LaB0.2Mn0.8O3 sample was higher than that of the
pure LaMnO3. As is shown in the characterization results, the
Lanzhou 730000, P. R. China. E-mail: tangzhicheng@licp.cas.cn; Tel: +86 931 4968083
b University of Chinese Academy of Sciences, Beijing 100039, P. R. China
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
low-temperature reducibility of the metal sites greatly influ-
enced the activity. MnxCo3ꢀxO4 catalysts have the character-
istics of low cost, environmental friendliness and high activity
for the combustion of VOCs. The synergetic effect between
manganese and cobalt oxides can significantly promote the
catalytic abilities.
MOFs have the advantages of high surface area, and various
pore sizes and structures. And great efforts have been made to
prepare new MOFs and explore their applications. Metal oxides
(based on Co Mn and Cu) have been widely used in the catalytic
oxidation of VOCs.16,17 And in recent years, MOFs’ thermal
decomposition has become an effective way to prepare metal
oxides. Yu et al. prepared porous hybrid metal oxides derived
from MOF-74 and applied them to electrochemical energy
storage.18 Zhang et al. prepared Co3O4 hollow parallelepipeds
Scheme 1 Schematic showing the synthetic procedures of CoMnOx.
and Co3O4 microsheets by annealing a cobalt-based MOF at previously reported method.23 First, 750 mg of Co(NO3)2ꢁ6H2O
different temperatures. And the Co3O4 hollow parallelepipeds and 428 mg of terephthalic acid were dissolved in 48 mL DMF
exhibited a better electrochemical performance.19 Luo et al. took and 12 mL absolute EtOH. After thoroughly mixing, the solution
nanocube-like MOFs (Mn3[Co(CN)6]2ꢁnH2O) as the precursor to turned transparent. Then the solution was transferred into a
prepare Co–Mn oxides. They found that a moderate ratio of 100 mL Teflon-lined stainless steel autoclave and placed in an
Co/Mn (1 : 1) exhibited a good catalytic performance in toluene oven at 100 1C for 17.5 h. After the autoclave was cooled down to
oxidation.20 Han et al. synthesized one-dimensional (1D) nano- room temperature, the resulting crystals were collected by
particles, 2D hexagonal nanosheets and 3D rosettes of CeCoOx filtration and washed thoroughly with absolute ethanol. Finally,
by annealing Ce[Co(CN)6] at different temperatures.21 Above all, the pink powder of Co-MOF-71 was collected after drying in a
MOFs provide a way to prepare metal oxides. Although pyrolyz- vacuum oven at 60 1C for 12 h. MOF-71-Co1Mn0.33, MOF-71-Co1Mn1,
ing MOFs would reduce the surface area, they still had a bigger and MOF-71-Co1Mn3 were prepared as described above, using
surface area compared with oxides prepared in other ways. It Mn(NO3)2 and Co(NO3)2ꢁ6H2O with a molar ratio of 3 : 1, 1 : 1,
would also expand their pores to mesopores.22
Herein, we introduce a simple method to obtain mesoporous in Scheme 1.
well-dispersed Co3O4 and mixed metal oxides by pyrolyzing 2.1.2 Preparation of nanostructured Co/Mn catalyst. Co3O4
[Co(bdc)(DMF)] (MOF-71, bdc 1,4-benzenedicarboxylate, catalyst was prepared via a one-step calcination of cobalt based
and 1 : 3, respectively. The detailed synthesis process is shown
=
DMF = N,N-dimethylformamide) as a template. To improve its MOF-71 in a furnace under a flow of air. The temperature was
catalytic performance, we tried to replace a designated amount raised from room temperature to 350 1C at a ramping rate of
of cobalt with manganese. Briefly, there were two steps in the 2 1C minꢀ1, and then stabilized at 350 1C for 2 h. Finally, a black
synthetic route: (1) the preparation of Co-MOF-71 and Co/Mn-MOF- powder was obtained (Scheme 1). Co1Mn0.33Ox, Co1Mn1Ox and
71, (2) removal of the templates by annealing. The as-prepared Co1Mn3Ox were prepared using the same process.
samples were used to eliminate VOCs using carcinogenic
2.2 Characterization of materials
toluene as a representative. Moreover, the prepared Co1Mn1Ox
samples showed a good low-temperature catalytic performance. The morphology of the catalysts was evaluated using a transmission
The temperature of 90% conversion was 158 1C. Total conversion electron microscope (TEM). The TEM images and selected-area
was reached at 180 1C. And it also showed long-term stability. electron diffraction patterns were obtained under a microscope of
Furthermore, the influence of different molar ratios of cobalt and 200 kV using an FEI TECNAIG. Powder X-ray diffraction (XRD)
manganese on the catalysts’ structure and catalytic performance analysis was performed at 60 kV and 55 mA by using Cu Ka
was systematically investigated by scanning electron microscopy radiation with a scanning speed of 0.51 minꢀ1. The 2y of
(SEM), high-resolution transmission electron microscopy (HRTEM), wide-angle was from 101 to 801. Raman scattering was done by
X-ray diffraction patterns (XRD), X-ray photoelectron spectroscopy using a Laboratory Human Resources Evolution Raman spectro-
(XPS), the Brunauer–Emmett–Teller (BET) method and Raman meter (BX41). The XPS measurement was done using a VG
spectroscopy.
ESCALAB 210 Electron Spectrometer (Mg Ka radiation; hn =
1253.6 eV). The average pore diameter and the specific surface
area of these catalysts were measured by nitrogen adsorption
using a micromeritics ASAP2010 instrument.
For H2-TPR, a 50 g catalyst was pre-treated in an atmosphere
of nitrogen. The temperature was increased from room temperature
2. Experiments
2.1 Preparation of catalyst
2.1.1 Synthesis of MOF-71. The cobalt-based MOF-71 was to 350 1C with a heating rate of 10 1C minꢀ1. When it cooled
synthesized by a one-step hydrothermal method with different down to room temperature, the temperature was increased to
molar ratios and a subsequent calcination process following a 900 1C with a heating rate of 10 1C minꢀ1. A reducing gas which
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
contains 5% vol. H2 and 95% vol. N2 was passed through the
sample at the same time.
The inlet and outlet toluene concentrations in the feed
stream, toluene gas flow rate and surface area were denoted as
[C]inlet, [C]outlet, Ntoluene (mol sꢀ1) and SBET (m2 gꢀ1), respectively.
2.3 Catalytic activity measurements
The catalytic oxidation of toluene was evaluated by using a
fixed-bed flow reactor, which operates at a steady-state flow
mode. Then, 0.4 g catalyst (40–60 mesh) and 0.7 g quartz sand
(40–60 mesh) were mixed together. They were put onto the
reactor. The reaction gas which contains VOCs (3000 ppm) was
generated by mixing with air through a VOC saturator. Then
the mixing gas was passed through the reactor. The weight
3. Results and discussion
3.1 Characterization of catalysts
3.1.1 BET and TEM analysis. The specific surface areas,
pore volumes, and pore sizes of the samples are summarized in
Table 1. The specific surface areas of the Co3O4, Co1Mn0.33Ox,
Co1Mn1Ox and Co1Mn3Ox samples were 47.73, 84.78, 63.07 and
76.68 m2 gꢀ1, respectively. The surface areas were larger than
other Co–Mn oxides prepared in other ways. All N2 adsorption–
desorption isotherms of these samples exhibited type V at a
relatively high pressure. The results turned out that these
samples were mesoporous materials (Fig. 1a). The distribution
of the BJH pore size indicated that all these samples displayed a
20–40 nm mesoporous distribution (Fig. 1b).
As is known to us, the surface area had an important effect
on the catalytic performance. Compared with other technolo-
gies of preparing the Co/Mn oxides, the catalyst with a larger
BET surface area was obtained by deriving from MOF-71. The
Co1Mn1Ox samples owned a 63.07 m2 gꢀ1 surface area. For
example, the Co1Mn1Ox catalyst obtained by the co-precipitation
hourly space velocity (WHSV) was 30 000 mL gꢀ1 ꢀ1. The first
h
temperature was 100 1C. The activity was measured per 20 1C.
Before each test, it needed to stabilize for 2 h. The reactants
and products were analyzed using an online GC-6820 gas
chromatograph with a flame ionization detector, conversion
was defined as:
½Cꢂinlet ꢀ ½Cꢂoutlet
Ctoluene
¼
ꢃ 100%
½Cꢂinlet
The specific toluene reaction rate stoluene (mol (m2 s)ꢀ1
)
could be calculated by the following equation:
Ntoluene ꢃ Xtoluene
stoluene
¼
SBET
method was 18.7 m2 gꢀ1 24
.
The catalyst obtained by the high-
The TiO2-supported
energy ball milling method was 47 m2 gꢀ1 25
.
1
manganese and cobalt oxide catalysts owned a 56.7 m2 gꢀ1 surface
area.26 The catalyst derived from the Co–Mn–CO3 samples had a
Xtoluene ¼ log
Ztoluene
100
1 ꢀ
Table 1 Data of characterization of the Co3O4, Co3Mn1Ox, Co1Mn1Ox and Co1Mn3Ox samples
Samples
SBETa/m2 gꢀ1
Pore volume/m3 gꢀ1
Crystallite sizeb/nm
Co3+/Co2+
(Mn4+ + Mn3+)/Mnn+
Ob/(Oa + Ob + Og)
Co3O4
47.73
84.78
63.07
76.68
—
0.20
0.29
0.30
0.27
—
30.81
22.52
22.84
22.17
—
0.64
0.38
0.37
0.49
0.75
—
0.30
0.23
0.30
0.24
0.30
Co1Mn0.33Ox
Co1Mn1Ox
Co1Mn3Ox
Used-Co1Mn1Ox
0.95
0.92
0.88
0.84
a
b
BET surface area. Crystallite sizes of particles calculated from (311) with the Scherrer equation.
Fig. 1 Adsorption–desorption curves (a) and pore-size distributions (b) of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox and Co1Mn3Ox samples.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
14.9 m2 gꢀ1 surface area.27 The catalyst derived from the MIL-53
materials had a 29.42 m2 gꢀ1 surface area.28
Morphologies of these samples were characterized using
transmission electron microscopy (TEM). Fig. 2 shows the mor-
phology of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox, and Co1Mn3Ox
samples. From the TEM images, we saw granule-like structures.
The nanoparticles were stuck with each other and then formed a
sheet-like structure. After calcination, the Co3O4 and Co1Mn1Ox
samples showed a uniform distribution (Fig. 2). From the TEM
images, it’s obviously seen that one nanosheet consisted of many
individual single crystalline nanoparticles. We made a grain
distribution diagram for the four samples. From the grain
distribution diagram (Fig. 3), we also calculated the average
size of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox, and Co1Mn3Ox
samples, which was 21.12 nm, 15.11 nm, 11.11 nm and
10.12 nm, respectively. With the amount of Mnn+ increasing,
the particle size decreased. And the Co3O4 sample had the
biggest particle size. We figured out that the introduction of
Mnn+ played an important role in the particle size.
From the HRTEM image, there was no fringe crystal phase
corresponding to Mn oxides (Fig. 4a). This proved that no MnOx
species were formed, which was in accordance with the XRD
results. The HRTEM image of the Co1Mn1Ox sample exhibited
different lattice fringes. The interplanar distance was 0.226 nm,
0.246 nm and 0.291 nm, which corresponded to the (400), (311)
and (220) lattice planes.28,29 The SAED patterns showed that the
polycrystalline metal oxides consisted of individual single
crystalline nanoparticles (Fig. 4).
3.1.2 XRD and Raman analysis. Fig. 5 shows the XRD
patterns of the catalysts after calcination. The key 2y peaks at
31.271, 36.851, 44.811, 59.351 and 65.231 were related to the (220),
(311), (400), (511) and (440) lattice planes of Co3O4 (PDF#43-
1003). This illustrated the formation of Co3O4 oxides. Although
Fig. 3 TEM images of the Co3O4 (a), Co1Mn0.33Ox (c), Co1Mn1Ox (e), and
Co1Mn3Ox (g) samples and the corresponding grain distribution diagrams
of Co3O4 (b), Co1Mn0.33Ox (d), Co1Mn1Ox (f), and Co1Mn3Ox (h).
there was no detection of peaks ascribable to manganese oxides, doped onto the Co. Furthermore, the Co1Mn3Ox sample corre-
we saw that the intensity of the peaks decreased as the content of sponded to the Mn0.1Co0.9O4.5 phase (PDF#78-1991). From the
Mn increased. And they presented different degrees of peak shifts previous report,23 we knew that the Co–O–Mn solid solution was
compared with Co3O4. We could infer that Mn was successfully formed because of the introduction of Mnn+ into the crystal
lattice of Co3O4. This would cause the distortion of the lattice
and reduce its particle size. Such an effect could also be con-
firmed by the disappearance of the (200) and (400) peaks besides
the reducing intensity.30 Among the four samples, the highest
XRD signal intensity and biggest particle size were achieved for
Co3O4 (based on the analysis of the half width of the representa-
tive diffraction peak using Scherrer formula applied for (311)).
The crystal size of Co3O4 is 30.8 nm. The crystal sizes of the
Co1Mn0.33Ox, Co1Mn1Ox, and Co1Mn3Ox samples were 22.1 nm,
22.8 nm and 22.5 nm, respectively. This tendency was the same
as the results of TEM. From a previous report,30 the content of
manganese has a big influence on the structure of the Co–Mn
oxides. According to the XRD result, we know that the Co3O4’s
parameters of a, b and c lattice were 8.085 nm, 8.085 nm and
8.085 nm, respectively. It’s an orthorhombic structure. After the
insertion of the manganese, the parameters of the a, b and c
lattice changed significantly, and became 3.0096 nm, 3.0086 nm
and 6.00104 nm, respectively. And now, it became a monoclinic
structure. Because of the structure’s change, the crystallite size
Fig. 2 TEM images of the Co3O4 (a), Co1Mn0.33Ox (b), Co1Mn1Ox (c) and
Co1Mn3Ox (d) samples.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
Fig. 4 HRTEM images of the Co1Mn1Ox (a) sample and the SAED patterns of the Co1Mn1Ox (b).
Fig. 5 XRD patterns of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox and Co1Mn3Ox
samples.
Fig. 6 Raman spectra of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox and
Co1Mn3Ox samples.
had a big drop. The small crystallite size improved the dispersion
of Co. From these results, we confirmed the formation of the octahedral sites (CoO6). For other samples, the Eg symmetry’s
Co–O–Mn solid solution.
intensity decreased and it’s also hard to find the 2F22g phonon
All the samples were investigated by Raman scattering and symmetries because of the initiation of Mn. This might
the resulting spectra are shown in Fig. 6. Raman bands at illustrate that the intrusion of Mn caused a decrease in the
187, 467, 508, 605, and 668 cmꢀ1 could be attributed to the tetrahedral sites (CoO4).
Co3O4 samples, which revealed five Raman-active modes (A1g,
3.1.3 Thermogravimetry analysis. The structure of MOF-71
Eg and 3F2g). They could correspond to the tetrahedral sites consisted of infinite chains of corner-sharing CoO6 octahedra,
(CoO4) of the F12g symmetry, Eg symmetry, 2F22g phonon symme- with each chain linked to four parallel running chains via the
tries and octahedral sites (CoO6) with A1g species in the O7h bdc linkers.33 The thermal behavior of MOF-71 was evaluated
spectroscopic symmetry, respectively.31 The 668 cmꢀ1 band of by the thermogravimetry (TG) experiment. There were two
the Co3O4 samples was shifted slightly compared with other weight losses in the TG curves. The first weight drop was
samples and there were various degrees of blue shifts and red from room temperature to 190 1C. It resulted from the decom-
shifts with different molar ratios, which was the same as what’s position of the free disordered BDC and DMF in the pores of
reported in a previous study.32 It’s because of the intrusion of MOF-71.34 And MOF-71 also experienced a sharp weight loss
manganese. It could change the polyhedral distortion and the between 360 1C and 450 1C. The loss meant the release of a lot
cation–anion bond length in the form of a spinel lattice. This of CO2 and H2O. And it might come from the elimination of
also illustrated the shift of the lattice planes in the XRD image. BDC linkers. There was no significant loss seen with an
And its intensity decreased with the addition of manganese. increase in the temperature. We could conclude that the
The Raman bands also became broader with Mn amount. MOF-71 crystals were completely transformed into Co3O4, when
The intensity of the relative peak in the form of I(F12g)/I(A1g) the temperature was over 450 1C. And the Co3O4 catalyst
was used to estimate the ratio of tetrahedral sites (CoO4) and retained 28% of its original weight (Fig. 7). In our experiment,
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
The second step could be regarded as the reduction of Co3O4
and Mn3O4.
Mn3O4–Co3O4(s) + H2(g) - MnO–CoO(s) + H2O(g)
(2)
Pure Co3O4 and the low content of manganese samples had
two main reduction peaks, which is in accordance with other
reports.34,35 The peak which was lower than 300 1C could be
regarded as the reduction of Co3O4 and Mn3O4 into CoO and
MnO. We could also see that the peak shifted to a higher
temperature with the amount of manganese increasing. It’s
because Mn4+/Mn3+ replaced the Co3+ ions in the octahedral
sites. This also led to the result that the reducibility of these
samples decreased because of the formation of Co–O–Mn.
The third step could be regarded as the reduction of Co2+ to
metallic cobalt. Because the reduction of Mn2+ needed a high
temperature, the peak wasn’t detected.
Fig. 7 TG curves of the MOF-71 sample in air.
the calcination temperature we chose was 350 1C, which led to
the elimination of BDC linker incompletely. That might be the
reason why our catalysts owned a higher specific surface area.
To verify the existence of the linker, we made a test using FT-IR.
We will illustrate this in the following parts.
MnO–CoO(s) + H2(g) - MnO(s) + Co(s) + H2O(g)
(3)
The peaks between 350–400 1C could be regarded as the
reduction of Co3O4 nanoparticles. The third peak could be
regarded as the reduction of Co2+ to Co0.
3.1.4 H2-TPR analysis. H2-TPR experiments were carried to
explore the reducibility of the samples. The H2-TPR of the
samples is shown in Fig. 8. From Fig. 8, we can see that all of
the samples had low-temperature reducibility. We could also
see that there were three main reduction steps in the Co1Mn1Ox
and Co1Mn3Ox samples. As the addition of the manganese
increased, three typical zones could be seen. Compared with
the pure oxides, the reduction of Mn–Co bimetal oxides could
be summarized as three steps.35 The first one could be regarded
as the reduction of easily reducible species.
From Fig. 8, we could see that the first peak of the Co1Mn1Ox
samples shifted to a low temperature with the insertion of Mn
into Co3O4. This indicated that the synergistic effect of man-
ganese and cobalt improved the reducibility. From Fig. 8 and
10, we knew that the reducibility was related to the mole ratio of
Mn/Co. When the the mole ratio of Mn/Co was 3, the peak
shifted to a high temperature. The peaks above 400 1C were
regarded as the reduction of Co2+ to Co0. This indicated that the
process of Co2+ to Co0 became difficult in the Co/Mn solid
solution. From the XRD results, the insertion of Mn led to a
small crystallite size. It’s reported that a small crystallite size
was easy for the formation of reducible sites.35 These explained
why the Co1Mn1Ox samples had a good catalytic performance.
3.1.5 XPS analysis. The composition of element species
and the chemical states could be identified using X-ray photo-
electron spectroscopy. The XPS of the Mn 2p, Co 2p and O 1s is
shown in the Fig. 9. From Table 1, we saw that the Co/Mn ratio
had a significant change. The Co/Mn ratios of Co1Mn0.33Ox,
Co1Mn1Ox and Co1Mn3Ox were 3.4, 1.2 and 0.6, respectively and
were in accordance with the tendency of the increasing amount
of manganese.
MnxCoyOz(s) + H2(g) - Mn3O4–Co3O4(s) + H2O(g)
(1)
As for the Co1Mn1Ox samples, it had a low-temperature
peak. It could be attributed to the reduction of Mn2O3 to
Mn3O4 or CoMnO3. So, it might illustrate a good catalytic
performance at a low temperature.
And all Co 2p spectra could be divided into four peaks (two
main peaks: Co3+ and Co2+, and two satellite peaks: Sa1 and
Sa2). With the introduction of manganese, there was also an
obvious shift of the Co 2p3/2 peak (Fig. 9a). These peaks shifted
towards a higher binding energy, as the amount of manganese
increased. It’s because of the strong interaction of manganese
and cobalt. The deconvolution of Co 2p3/2 showed peaks at
779.4 ꢄ 0.3 and 780.6 ꢄ 0.2 eV. It could be considered as the
Co3+ and Co2+ species, respectively.36,37 Co3+ had a 2p3/2–2p1/2
splitting of 15.1 eV and Co2+ had a 2p3/2–2p1/2 splitting of
15.5 eV. As summarized in Table 1, the molar ratio of Co3+/Co2+
had a significant decrease, when the manganese was introduced
into the spinel structure. It was considered that Mn3+ and Mn4+
Fig. 8 H2-TPR curves of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox and Co1Mn3Ox
samples.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
interaction between Mn and Co. This changed the environ-
ment of oxygen coordination. The relative concentration of
Ob/(Oa + Ob + Og) is listed in Table 1. In previous reports,
people found that the surface absorbed oxygen had a higher
mobility compared with the lattice oxygen. So the surface absorbed
oxygen was more active and played an important role in oxidation
reactions. The amounts of the surface absorbed oxygen of Co3O4
(0.303) and Co1Mn1Ox (0.296) were higher than that of Co1Mn0.33Ox
(0.2284) and Co1Mn3Ox (0.2349). More surface absorbed oxygen
would lead to a better result of oxidation, which was the same as the
results of our toluene oxidation.
With the introduction of Mn into the spinel lattice, Mn
species were mainly composed of Mn2+, Mn3+ and Mn4+. Fig. 9c
shows the peaks of Mn 2p at 640.5 ꢄ 0.3, 641.5 ꢄ 0.3 and
642.0 ꢄ 0.3 eV, which was assigned to Mn2+, Mn3+ and Mn4+,
respectively.28,40 All three ions had a 2p3/2–2p1/2 splitting of
11.7 eV. The previous study28 reported that Mn4+ and Mn3+ were
more active for the ORR. It’s known that the high-valence metal
ions had a big influence on the chemical potential and reactivity
of adjacent oxygen. The higher the valence is, the more easily
the chemical potential and reactivity of the adjacent oxygen will
be promoted. The value of (Mn4+ + Mn3+) ratio decreased with
an increase in the amount of manganese (Table 1). Both
Co1Mn1Ox and Co1Mn0.33Ox had a high ratio. The proportion
of (Mn4+ + Mn3+) redox pairs of the Co1Mn1Ox and Co1Mn0.33Ox
samples reached 92% and 95%, respectively. The Co1Mn3Ox
exhibited a value of 88%, which was relatively low compared
with others. The previous study32 also reported that the Mn3+
ions were easy to transform to Mn4+ ions in the Co3O4 spinel
structure. This explained why the Co1Mn0.33Ox sample had a
high value of Mn4+.
3.2 The catalytic perfoamnce of CoMnOx samples in VOC
catalytic combustion
The activities of the samples were evaluated by the catalytic
oxidation of toluene. And the result is illustrated in Fig. 10a.
The conversion of toluene increased with an increase in the
temperature. From the picture, we saw that the performance
of the catalyst was in the order of Co1Mn1Ox 4 Co3O4
4
Co1Mn0.33Ox 4 Co1Mn3Ox catalysts. The total conversion of
toluene on these samples was 180 1C, 200 1C, 240 1C, and
340 1C, respectively. From this, we saw that the catalytic
performances were different because of the different molar
Fig. 9 XPS spectra of the Co3O4, Co1Mn0.33Ox, Co1Mn1Ox and Co1Mn3Ox
samples: Co 2p (a), O 1s (b), Mn 2p (c).
ions might replace the octahedral sites of Co3+. This might ratios. We had illustrated the catalytic performance of the
generate additional Co2+ ions. It’s reported that Co3+ was more Co1Mn3Ox samples at the H2-TPR part. In this part, we focused
active for the ORR. That might explain why the Co3O4 sample had on the catalytic performance of the Co1Mn1Ox samples. We saw
a good catalytic performance. But the (Mn4+ + Mn3+) pair was more that the Co1Mn1Ox samples showed an excellent catalytic
active and we will introduce this in the following part. It was performance at a low temperature. This is in accordance with
considered that other parts might play an important role in the the H2-TPR. The result of XPS also illustrated why the Co1Mn1Ox
ORR apart from the ratio of Co3+/Co2+.
sample had a good catalytic performance.
Fig. 9b showsthe peaks of O 1s at 529.6 ꢄ 0.2, 531.3 ꢄ 0.3
In Table 2, we compared the catalytic performance of our
and 533.0 ꢄ 0.0 eV. These peaks were assigned to the lattice samples with the previous reports. Liu et al.43 prepared CoAl
oxygen (Oa), surface absorbed oxygen (Ob) and adsorbed mole- mixed oxide which contained Pd. Its 90% conversion temperature
cular water or hydroxyl groups (Og).38,39 It’s obviously seen that is 272 1C. Qu et al.45 also prepared a kind of CoMn oxide. This
the peak of Oa has a shift toward to a high binding energy with nanocrystal-like catalyst had a high surface area and porosity.
the introduction of manganese. It might be the result of the It showed a relatively good activity with a 90% conversion
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Fig. 10 (a) Catalytic performance for toluene oxidation, (b) effect of weight hourly space velocity on toluene oxidation over the Co1Mn1Ox sample,
(c) the toluene conversion during three consecutive runs using the Co1Mn1Ox sample, and (d) stability tests on the Co1Mn1Ox sample.
Table 2 Comparison of different catalysts on the activity of toluene oxidation
Samples
VOC type
VOC conc. (ppm)
WHSV (mL gꢀ1 hꢀ1
)
T90/1C
Ref.
Cube-stacked Co3O4 microspheres
Nanocube CoMnOx
CeCoOx
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
1000
500
3000
1000
1000
800
48 000
96 000
30 000
20 000
40 000
30 000
60 000
30 000
20 000
30 000
60 000
96 000
248
240
168
260
260
272
253
249
260
154
232
261
10
20
21
41
42
43
44
45
Co3O4-HT
8.4Au/3DOM Co3O4
Pd/Co3AlO (WIE)
Ce–Co EM 1 : 2
Mn–Co (1 : 1)
6.55Au/Fe2O3
Co1Mn1Ox
260
1000
1000
3000
3000
3000
46
This work
This work
This work
Co1Mn1Ox
Co1Mn1Ox
temperature of 249 1C. Both of them are tested in the condition conversion temperature of CoAl mixed oxide containing Pd is
of WHSV = 30 000 mL gꢀ1 ꢀ1. In the condition of WHSV = 260 1C. Our catalyst showed a relatively good performance with a
20 000 mL gꢀ1 ꢀ1, the 90% conversion temperature of a 90% conversion temperature of 232 1C (WHSV = 60 000 mL gꢀ1 hꢀ1).
h
h
nanodisk-like Fe2O3 catalyst fabricated by gold is 260 1C. Our And the 85% conversion temperature was 180 1C, which was also a
catalyst showed a relatively good performance with a 90% relatively low temperature. Luo et al.20 prepared nanocube-like
conversion temperature of 154 1C (WHSV = 30 000 mL gꢀ1 ꢀ1). CoMnOx oxides derived from MOFs. This catalyst showed a very
h
Ren et al.10 prepared cube-stacked Co3O4 microspheres. They good catalytic performance. At a high WHSV (96 000 mL gꢀ1
h
ꢀ1),
showed a relatively good activity, whose 90% conversion tem- the 90% conversion temperature can reach 240 1C. At this condition,
perature is 240 1C in the condition of WHSV = 48000 mL gꢀ1 hꢀ1
.
our catalyst’s 90% conversion temperature is 261 1C. When the
Li et al.42 prepared a 3DOM Co3O4 oxide doped with Au. Its temperature is 240 1C, the conversion is about 75%. It seemed a little
90% conversion temperature is 260 1C in the condition of bad for our catalyst maybe because the concentration of toluene
WHSV = 60 000 mL gꢀ1
h
ꢀ1. At the same condition, the 90% is higher.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
The influence of WHSV over the Co1Mn1Ox sample was where F stands for toluene feeding rate (1.08 ꢃ 10ꢀ7 mol sꢀ1), W
studied and the result is shown in Fig. 10b. As expected, the stands for the weight of the catalyst (0.4 g), and x stands for the
catalytic performance of the Co1Mn1Ox sample decreased toluene conversion. So, the equation can be rewritten to:
with an increase in the WHSV. As the WHSV was 30 000 and
ꢀEa
60 000 mL gꢀ1 hꢀ1, the 90% conversion temperature of toluene
was 154 and 232 1C. The Co1Mn1Ox sample had a better
performance at low WHSV, which indicated that the WHSV
influenced the Co1Mn1Ox sample. And we also saw that the catalyst
still had a low-temperature catalytic activity even at a high WHSV.
The toluene conversion of three consecutive runs is shown in
Fig. 10c. There was a loss at the temperature of 140 1C in the
second run. Compared with the second, the third had a loss at the
temperature of 160 1C. But all of them had a high conversion
at a relatively high temperature of 180 1C. The durability of the
Co1Mn1Ox sample is shown in Fig. 10d. In a run of 14 h at 90%
conversion, there was no significant loss of catalytic activity. During
industrial applications, the thermal stability of the catalyst was paid
a lot of attention. In summary, we could conclude that the
Co1Mn1Ox catalyst had a good catalytic stability.
ln r ¼
þ B
RT
Here, B stands for a constant. Therefore, the activation
energy could be evaluated from the slope of linear ln r versus
1000/T (Fig. 11a). The Co1Mn1Ox sample had the lowest Ea
(Ea = 24.52 kJ molꢀ1), compared with the Co3O4 (Ea = 26.61 kJ molꢀ1),
Co1Mn0.33Ox (Ea = 29.93 kJ molꢀ1) and Co1Mn3Ox (Ea
=
65.68 kJ molꢀ1). From this, we could see that the catalytic
activities followed an inverse trend of the Ea values. This
illustrated that the Co1Mn1Ox sample had the best catalytic
performance. And the specific reaction rate of toluene combus-
tion was regarded as another way to evaluate catalysts. It was
used to evaluate the strength of the catalyst in per unit area. As
shown in Fig. 11b, the catalytic activity for toluene combustion
of the four catalysts was in the order Co1Mn1Ox 4 Co3O4
4
Co1Mn0.33Ox 4 Co1Mn3Ox. From this, we saw that the Co1Mn1Ox
and Co3O4 samples had a significant increase compared with the
others. This also indicated that the Co1Mn1Ox sample had the best
catalytic activity in the toluene combustion reaction.
As is known to us, the chemical reaction rate corresponds to
the activation energy. A low activation energy corresponds to a
high reaction rate. We can calculate the Arrhenius plots of the
four catalysts and the linear relations (ln r versus 1000/T) could
be well fitted by the equation (Fig. 11a).
3.3 Proposed reaction mechanism
It was reasonable to hypothesize that the oxidation of
toluene would follow a first-order reaction mechanism in the The Langmuir–Hinshelwood (L–H) mechanism is generally
presence of excessive oxygen. It’s influenced by the toluene used to illustrate the oxidation process of toluene. This model
concentration:
held the point that the adsorbed toluene and the chemical
oxygen of the catalyst rather than the oxygen in the gas phase
played an important role in the reaction. The model was
divided into two parts, which is shown in Scheme 2. In the
first step, we thought the adsorbed toluene combined with
oxygen and reacted in the catalyst, which lead to the reduction
of the metal oxide content. In the second step, the reduced
metal oxide was oxidized again by oxygen in the gas phase.
When the catalyst experienced the progress of reduction and
oxidization, it was called the reducing mechanism. We knew
that the oxidation–reduction sites played a significant role in
ꢀ
ꢀ
ꢁꢁ
ꢀEa
r ¼ ꢀkc ¼ ꢀA exp
c
RT
where r stands for the reaction rate (mol sꢀ1), k stands for rate
constant (sꢀ1), A stands for pre-exponential factor, c stands for
concentration of toluene in the feed stream (mol mꢀ3) and Ea
stands for apparent activation energy (kJ molꢀ1).
F
r ¼ ðꢀ lnð1 ꢀ xÞÞ
W
Fig. 11 (a) Arrhenius plots of toluene catalytic oxidation reaction, (b) specific reaction rate of toluene combustion over the Co3O4, Co3Mn1Ox, Co1Mn1Ox
and Co1Mn3Ox samples.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Scheme 2 Schematic of oxidation of toluene on Co1Mn1Ox samples.
the catalytic performance from the reaction mechanism. It
seemed that different ratios had a tremendous effect on the
compositional evolution of cobalt and manganese oxides and
the catalytic performances for toluene oxidation, which has
been illustrated by the following analysis.
There were many factors that have an influence on the
catalytic performance, such as particle size, the number of
surface oxygen-containing functional groups, oxygen vacancies
and so on. From the narrative in the prior part, we found that
the proportion of (Mn4+ + Mn3+) redox pairs and the surface
absorbed oxygen were the main factors which had the biggest
influence on the catalytic performance. In Fig. 12, the XPS
comparison of the fresh and used Co1Mn1Ox is shown. The
used catalyst was chosen, after a 14 hour test of stability. It’s
obviously seen that there was an obvious increase in the ratio of
Co3+/Co2+ and a decrease in the proportion of (Mn4+ + Mn3+)
redox pairs. We thought this was the result of bimetallic
synergy. And we also found that there was no significant change
of the surface absorbed oxygen ratio. This illustrated that the
catalyst had a good stability. In the prior part, we illustrated
that the (Mn4+ + Mn3+) redox pairs were more active. The
decreased ratio was in accordance with the catalytic perfor-
mance. The addition of manganese also caused a decrease in
the particle size. Small particle size and the nanosheet like
structure contributed to a larger specific surface area and more
exposed metal sites.
Fig. 12 XPS spectra of the used and fresh Co1Mn1Ox samples: Co 2p (a),
O 1s (b), and Mn 2p (c).
Fig. 13 displayed the FT-IR spectrum of the used and fresh
Co1Mn1Ox samples. Two bands located at around 658 cmꢀ1 and the result of TG. The broad peak at 3432 cmꢀ1 was assigned to
561 cmꢀ1 were observed. According to Shikha Dubey,47 the two the stretching vibration of the O–H group, which came from
bands at around at 660 cmꢀ1 and 570 cmꢀ1 were considered as molecular water and hydrogen-bound groups. The band at
the stretching vibration mode of Co–O bonds. According to 1414 cmꢀ1 was considered as the stretching vibration of the
Askarinejad and Morsali,48 the two bands at around at 600 cmꢀ1 C–O group of CO2. We could obviously see that the bands at
and 500 cmꢀ1 were considered as the stretching vibration mode 3432 cmꢀ1 and 1414 cmꢀ1 had a significant increase. This
of Mn–O. So the shift came from the introduction of manganese might be due to the adsorption of the carbon dioxide and water
into the cubic structure, which supported the XRD results. The molecule from the decomposition of the toluene during the
bands between 1430 cmꢀ1 and 1650 cmꢀ1 could be regarded as process of cooling down. This might illustrate the reason why
the skeleton vibration of benzene, which was in accordance with the three consecutive runs had a slight change.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
References
1 Y. Liao, L. Jia, R. Chen, O. Gu and M. Sakurai, Appl. Catal., A,
2016, 522, 32–39.
2 Y. Yan, L. Wang and H. Zhang, Chem. Eng. J., 2014, 255,
195–204.
3 L. Wang, M. Sakurai and H. Kameyama, J. Hazard. Mater.,
2008, 154, 390–395.
4 L. Lin and H. Bai, Chem. Eng. J., 2016, 291, 94–105.
5 Z. Abbasi, M. Haghighi, E. Fatehifar and S. Saedy, J. Hazard.
Mater., 2011, 186, 1445–1454.
6 H. Rotter, M. Landau and M. Herskowitz, Environ. Sci.
Technol., 2005, 39, 6845–6850.
7 P. Liu, H. He, G. Wei, D. Liu and X. Liang, Microporous
Mesoporous Mater., 2017, 239, 101–110.
Fig. 13 FT-IR spectra of the used and fresh Co1Mn1Ox.
8 S. Li, H. Wang, W. Li, X. Wu, W. Tang and Y. Chen, Appl.
Catal., B, 2015, 166–167, 260–269.
4. Conclusion
9 G. Li, C. Zhang, Z. Wang, H. Huang, H. Peng and X. Li, Appl.
Catal., A, 2018, 550, 67–76.
In conclusion, we prepared a series of cobalt or manganese
oxides with different ratios through pyrolysis of MOF-71 or
CoMn-MOF-71. From the TEM images, we can see that the
samples consisted of many uniformly distributed nano-
particles. Because of the addition of manganese, we could see
a significant decrease in the particle size. There was also an
obvious change in the lattice planes from the XRD result.
Furthermore, the toluene oxidation was significantly improved
by the introduction of manganese. Most of the samples showed
a good catalytic performance. In particular, the Co1Mn1Ox
samples showed a relatively low-temperature catalytic activity.
The complete conversion temperature was 180 1C. After a three
cycle run, the total conversion could also reach 220 1C. When
10 Q. Ren, S. Mo, R. Peng, Z. Feng, M. Zhang, L. Chen, M. Fu,
J. Wu and D. Ye, J. Mater. Chem. A, 2018, 6, 498–509.
11 M. Piumetti, D. Fino and N. Russo, Appl. Catal., B, 2015, 163,
277–287.
12 M. Gallegos, M. Peluso, E. Finocchio, H. Thomas, G. Busca
and J. Sambeth, Chem. Eng. J., 2017, 313, 1099–1111.
13 L. Lin and H. Bai, Chem. Eng. J., 2016, 291, 94–105.
´
´
´
14 S. Behar, N. Gomez-Mendoz, M. Gomez-Garcı, D. Swiercz´ynski,
F. Quignard and N. Tanchoux, Appl. Catal., A, 2015, 504,
203–210.
15 C. Zhang, C. Wang, W. Zhan, Y. Guo, Y. Guo, G. Lu, A. Baylet
and A. Giroir-Fendler, Appl. Catal., B, 2013, 129, 509–516.
16 P. Sun, W. Wang, X. Dai, X. Weng and Z. Wu, Appl. Catal., B,
2016, 198, 389–397.
the WHSV increased to 60 000 mL gꢀ1 ꢀ1, the temperature of
h
90% conversion was 232 1C. The catalyst also had a good
stability. It had no significant loss of activity within a 14 hour
run. From the analysis of the XPS result, we saw that the
catalytic activity was mainly related to the (Mn4+ + Mn3+) redox
pairs and surface chemically adsorbed oxygen. Therefore, this
novel strategy could open a door for the application of MOF
materials for the removal of toluene. It was anticipated that the
present study will motivate researchers to generate mixed metal
oxides for toluene oxidation applications.
17 S. Somekawa, T. Hagiwara, K. Fujii, M. Kojima, T. Shinoda,
K. Takanabe and K. Domen, Appl. Catal., A, 2011, 409–410,
209–214.
18 C. Yu, Y. Wang, J. Cui, D. Yu, X. Zhang, X. Shu, J. Zhang,
Y. Zhang, R. Vajtai, P. Ajayan and Y. Wu, J. Mater. Chem. A,
2018, 6, 8396–8404.
19 H. Zhang, L. Zhang, J. Deng, Y. Han and X. Li, J. Mater. Sci.,
2018, 53, 8474–8482.
20 Y. Luo, Y. Zheng, J. Zuo, X. Feng, X. Wang, T. Zhang,
K. Zhang and L. Jiang, J. Hazard. Mater., 2018, 349,
119–127.
Conflicts of interest
21 W. Han, H. Zhao, F. Dong and Z. Tang, Nanoscale, 2018, 10,
21307–21319.
There are no conflicts to declare.
22 Y. Han, P. Qi, S. Li, X. Feng and J. Zhou, Chem. Commun.,
2014, 50, 8057–8060.
23 N. Rosi, J. Kim, M. Eddaoudi, B. Chen, M. O’Keeffe and
O. Yaghi, J. Am. Chem. Soc., 2005, 127, 1504–1518.
24 X. Yao, Y. Li, Z. Fan, Z. Zhang, M. Chen and W. Shangguan,
Ind. Eng. Chem. Res., 2018, 57, 4214–4224.
25 A. Masi, M. Bellusci, M. Carlini, S. McPhail, F. Padella and
P. Reale, J. Am. Ceram. Soc., 2016, 99, 308–314.
26 H. Hu, S. Cai, H. Li, L. Huang, L. Shi and D. Zhang, ACS
Catal., 2015, 5, 6069–6077.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (21707145, 51808529), Key Science and
Technology Program of Lanzhou City (2017-4-111, 2018-RC-65),
Province Natural Science Foundation of GanSu (18JR3RA383,
17JR5RA317) and West Light Foundation of the Chinese Academy of
Sciences.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
27 L. Pi, N. Yang, W. Han, W. Xiao, D. Wang, Y. Xiong, M. Zhou, 38 X. Huang, G. Zhang, F. Dong and Z. Tang, Catal. Sci.
H. Hou and X. Mao, Chem. Eng. J., 2018, 334, 1297–1308. Technol., 2018, 8, 5604–5615.
28 L. Yao, W. Yang, H. Liu, J. Jia, G. Wu, D. Liu, T. Liu, T. Tan 39 Y. Zhao, F. Dong, W. Han, H. Zhao and Z. Tang, New
and C. Wang, Dalton Trans., 2017, 46, 15512–15519. J. Chem., 2017, 41, 12052–12060.
29 J. Zhao, W. Han, F. Dong, J. Zhang and Z. Tang, Chemis- 40 M. Wang, Y. Wei, Q. Zou, W. Zhang, A. Xu and X. Li,
trySelect, 2018, 3, 10408–10417. J. Colloid Interface Sci., 2019, 536, 271–280.
30 B. Faure and P. Alphonse, Appl. Catal., B, 2016, 180, 41 B. Bai and J. Li, ACS Catal., 2014, 4, 2753–2762.
715–725.
42 P. Li, C. He, J. Cheng, C. Ma, B. Dou and Z. Hao, Appl. Catal.,
B, 2011, 101, 570–579.
31 V. Hadjiev, J. Phys. C: Solid State Phys., 1988, 21, L199–L201.
32 T. Cai, H. Huang, W. Deng, Q. Dai, W. Liu and X. Wang, 43 S. Xie, H. Dai, J. Deng, Y. Liu, H. Yang, Y. Jiang, W. Tan,
Appl. Catal., B, 2015, 166–167, 393–405. A. Ao and G. Guo, Nanoscale, 2013, 5, 11207–11219.
33 D. Miles, D. Jiang, A. Burrows, J. Halls and F. Marken, 44 S. Carabineiro, X. Chen, M. Konsolakis, A. Psarras,
´
˜
Electrochem. Commun., 2013, 27, 9–13.
34 Y. Zhang, Q. Gao, Z. Lin, T. Zhang, J. Xu, Y. Tan., W. Tian
P. Tavares, J. Orfao, M. Pereira and J. Figueiredo, Catal.
Today, 2015, 244, 161–171.
and L. Jiang, Sci. Rep., 2014, 4, 1–6, DOI: 10.1038/srep04947. 45 Z. Qu, K. Gao, Q. Fu and Y. Qin, Catal. Commun., 2014, 52, 31–35.
35 W. Tang, W. Li, D. Li, G. Liu, X. Wu and Y. Chen, Catal. Lett., 46 W. Han, J. Deng, S. Xie, H. Yang, H. Dai and C. Au, Ind. Eng.
2014, 144, 1900–1910.
36 F. Dong, Y. Zhao, W. Han, H. Zhao, G. Lu and Z. Tang, 47 S. Dubey, J. Kumar, A. Kumar and Y. Sharma, Adv. Powder
J. Mol. Catal. A: Chem., 2017, 439, 118–127. Technol., 2018, 29, 2583–2590.
37 W. Zhao, Y. Zhang, X. Wu, Y. Zhan, X. Wang, C. Au and 48 A. Askarinejad and A. Morsali, Ultrason. Sonochem., 2009,
L. Jiang, Catal. Sci. Technol., 2018, 8, 4494–4502. 16, 124–131.
Chem. Res., 2014, 53, 3486–3494.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019