Pd/meso-CoO derived from: In situ reduction of the one-step synthesized Pd/meso-Co3O4: High-performance catalysts for benzene combustion

Article abstract of DOI:10.1039/c9nj03039e

The chemical state of Pd plays an important role in the catalytic combustion of volatile organic compounds (VOCs). In this work, we adopted a novel one-step modified KIT-6-templating strategy with nitrates of cobalt and palladium as the metal source to successfully synthesize the three-dimensionally ordered mesoporous Co₃O₄-supported Pd nanoparticles (0.85 wt% Pd/meso-Co₃O₄, denoted as 0.85Pd/meso-Co₃O₄). The 0.93 wt% Pd/meso-CoO (denoted as 0.93Pd/meso-CoO) and 1.08 wt% Pd/meso-Co-CoO (denoted as 1.08Pd/meso-Co-CoO) samples were prepared via in situ reduction of 0.85Pd/meso-Co₃O₄ in a H₂ flow at 200 and 350 °C, respectively. Among these samples, 0.93Pd/meso-CoO exhibited the highest catalytic activity for benzene combustion (T_50% = 167 °C and T_90% = 189 °C at a space velocity of 40000 mL (g h)^-1). The chemical state of Pd on the 0.93Pd/meso-CoO surface was metallic Pd⁰, which favored oxygen activation to active adsorbed oxygen (O_ads) species, hence rendering this sample to possess the largest desorption of O_ads species below 400 °C. The intermediates of formate, acetate, maleate, and phenolate were generated via the interaction of benzene and O_ads species. We conclude that the excellent catalytic performance of 0.93Pd/meso-CoO was related to the mainly formed Pd⁰ species, good oxygen activation ability, and high surface area.

Full text of DOI:10.1039/c9nj03039e

NJC
View Article Online
View Journal
PAPER
Pd/meso-CoO derived from in situ reduction of
the one-step synthesized Pd/meso-Co3O4: high-
performance catalysts for benzene combustion†
a
Cite this:DOI: 10.1039/c9nj03039e
Xingtian Zhao,a Ran Zhang,b Yuxi Liu,a Jiguang Deng,
Peng Xu,ac Sijie Lv,a
Shuang Li,a Wenbo Pei,a Kunfeng Zhanga and Hongxing Dai
*
a
The chemical state of Pd plays an important role in the catalytic combustion of volatile organic
compounds (VOCs). In this work, we adopted a novel one-step modified KIT-6-templating strategy with
nitrates of cobalt and palladium as the metal source to successfully synthesize the three-dimensionally
ordered mesoporous Co3O4-supported Pd nanoparticles (0.85 wt% Pd/meso-Co3O4, denoted as
0.85Pd/meso-Co3O4). The 0.93 wt% Pd/meso-CoO (denoted as 0.93Pd/meso-CoO) and 1.08 wt%
Pd/meso-Co–CoO (denoted as 1.08Pd/meso-Co–CoO) samples were prepared via in situ reduction of
0.85Pd/meso-Co3O4 in a H2 flow at 200 and 350 1C, respectively. Among these samples, 0.93Pd/meso-
CoO exhibited the highest catalytic activity for benzene combustion (T50% = 167 1C and T90% = 189 1C
at a space velocity of 40 000 mL (g h)À1). The chemical state of Pd on the 0.93Pd/meso-CoO surface
was metallic Pd0, which favored oxygen activation to active adsorbed oxygen (Oads) species, hence
rendering this sample to possess the largest desorption of Oads species below 400 1C. The intermediates
of formate, acetate, maleate, and phenolate were generated via the interaction of benzene and Oads
species. We conclude that the excellent catalytic performance of 0.93Pd/meso-CoO was related to the
mainly formed Pd0 species, good oxygen activation ability, and high surface area.
Received 17th June 2019,
Accepted 9th July 2019
DOI: 10.1039/c9nj03039e
rsc.li/njc
Co3O4, one of the most popular spinel-type metal oxides, is a
catalytic material in a number of total oxidation of pollutants
1. Introduction
Recently, the legislation has become more strict with regard to (e.g., CO and CH4 as well as benzene, toluene, and xylene (BTX))
emission of industrial volatile organic compounds (VOCs), result- because of its good redox property and low cost.9–11 It has been
À
ing in an increased attention to removal of VOCs.1 Benzene, one proven that Co2+ species contribute to generation of active O2
2À
of the widely utilized aromatic VOCs with high toxicity and and O2 species, which favor the enhancement in catalytic
significant carcinogenic effect, is harmful to the environment performance.12 Since a catalytic reaction occurs on the catalyst
and human health.2,3 Among the various methods employed for surface, a rise in surface area could improve catalytic activity.
VOC elimination, catalytic combustion is one of the most effec- Recently, three-dimensionally (3D) ordered mesoporous Co3O4
tive and environmentally friendly pathways due to its advantages has attracted much attention.12–14 Such material can be synthesized
of good low-temperature removal efficiency and no secondary using a facile nanocasting method. The highly ordered mesoporous
pollutant production.4–8 Hence, it is needed to investigate highly structure favors dispersion of the noble metals as well as adsorption
active and stable catalysts for the removal of VOCs.
and diffusion of the reactant molecules. Up to now, many studies
have been focused on supported noble metal catalysts, in which
Pd-based materials exhibited excellent catalytic activities for organics
combustion.15–19 The loading of Pd on the surface of Co3O4 could
generate the noble metal–metal oxide interfaces that were beneficial
for improving oxygen activation ability and hence the catalytic
activity.20–23 It has been reported that the PdO-doped mesoporous
Co3O4 catalyst derived from an in situ nanocasting route showed
better activity for o-xylene oxidation than the PdO-loaded meso-
porous Co3O4 catalyst obtained by the post-impregnation method,
which was possibly due to the presence of a large amount of the
well-dispersed PdO species in the former sample.21
a Beijing Key Laboratory for Green Catalysis and Separation, Key Laboratory of
Beijing on Regional Air Pollution Control, Key Laboratory of Advanced Functional
Materials, Education Ministry of China, Laboratory of Catalysis Chemistry and
Nanoscience, Department of Chemistry and Chemical Engineering, College of
Environmental and Energy Engineering, Beijing University of Technology,
Beijing 100124, China. E-mail: hxdai@bjut.edu.cn
b Petrochina Petrochemical Research Institute, Beijing 102206, China
c CAS Key Laboratory of Standardization and Measurement for Nanotechnology,
National Center for Nanoscience and Technology, Beijing 100190, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9nj03039e
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
Pretreatment of a Pd-based catalyst in a reducing (H2) or 120 1C for 12 h, a brown powder was obtained which was
oxidative (O2 or air) atmosphere can induce a change in the thermally treated in a muffle furnace at a heating rate of 1 1C minÀ1
chemical state of the Pd species. It has been demonstrated that from room temperature to 450 1C and kept at 450 1C for 4 h. The
the reduction pretreatment led to a positive effect on the silica template was removed using a hot (80 1C) sodium hydroxide
activity of the Pd-based catalysts for acetylene hydrogenation aqueous solution (2.0 mol LÀ1). After filtering, washing with
and VOC oxidation.16,24 For example, Wang et al.25 attributed deionized water and ethanol three times, and drying at 70 1C for
the superior activity of the Al2O3-, TiO2-, SiO2-, CeO2-, MnO2- or 12 h, the Pt/meso-Co3O4 sample was obtained. The Pt/meso-Co3O4
Co3O4-supported catalyst for o-xylene combustion to formation sample was pretreated in a 3.5 vol% H2/96.5 vol% Ar flow of
of the metallic Pd0 species. For VOC combustion, the supported 30 mL minÀ1 at 200 or 350 1C for 1 h. The X-ray diffraction
Pd catalysts with metallic Pd0 species were more active than (XRD) patterns demonstrate that the cobalt oxide support was
the ones with PdO species.20 Metallic Pd0 with a strong ability reduced at 200 and 350 1C to CoO and Co–CoO, respectively. We
to adsorb o-xylene was the main phase on mesoporous CoO used the inductively coupled plasmaÀatomic emission spectro-
(meso-CoO), resulting in a high catalytic activity (o-xylene scopic (ICPÀAES) technique to measure the actual Pd contents
conversion = 90% at 173 1C).20 However, there is a controversial in the samples, and the results showed that the actual Pd
viewpoint on the effect of the Pd chemical state on VOC loadings in the three samples were 0.85, 0.93, and 1.08 wt%,
combustion. Some authors recognized that the catalysts with respectively, and they were denoted as 0.85Pd/meso-Co3O4,
a mixture of PdO and Pd0 performed better than the ones 0.93Pd/meso-CoO, and 1.08Pd/meso-Co–CoO, respectively.
with only Pd0.21,26 Our previous work also found that catalytic
activity was associated with the increased amount of Pd2+
2.2. Catalyst characterization
species.27 Therefore, it is important to investigate active sites
Physicochemical properties of the samples were characterized
of the Pd catalysts supported on meso-Co3O4.
using the ICPÀAES, XRD, TEM, HAADF-STEM, EDX, BET, XPS,
In one of our previous works,20 we investigated the role of
Raman, H2-TPR, O2-TPD, and in situ DRIFTS techniques. The
the interface between Pd and meso-CoO, and found that the
highly active Pd–CoO interface was responsible for the excellent
detailed characterization procedures can be found in the ESI.†
catalytic performance of the Pd/meso-CoO sample for o-xylene
oxidation. In the present work, however, we focus our attention
to the in situ reduction in H2 at different temperatures of the
one-step prepared Pd/meso-Co3O4 (nominal Pd loading was 1 wt%)
to Pd/meso-CoO and catalytic benzene combustion mechanisms by
the in situ diffuse reflection infrared Fourier transform spectro-
scopic (DRIFTS) technique, and the reaction mechanisms are
discussed and correlated with the chemical state of Pd. It should
be pointed out that compared with the Pd/meso-Co3O4 and
Pd/meso-CoO samples derived from the traditional impregna-
tion method, those prepared by the one-step approach
possessed better Pd dispersion that would enhance the catalytic
activity of the as-obtained samples. Furthermore, the one-step
approach simplified the preparation procedure, as compared
with the traditional two-step fabrication method.
2.3. Catalytic activity evaluation
We evaluated the catalytic activities of the samples in a continuous-
flow fixed-bed quartz tubular microreactor (i.d. = 6.0 mm). A gas
mixture containing 1000 ppm benzene + 20 vol% O2 + N2 (balance)
at a total flow of 33.3 mL minÀ1 was fed into the microreactor, in
which 0.5 g of the catalyst (40–60 mesh) was diluted with 1.5 g of
quartz sand (40–60 mesh), giving rise to a space velocity (SV) of
ca. 40 000 mL (g h)À1. The 1000 ppm benzene gas was generated
using a N2 flow that was bubbled in a saturator filled with liquid
benzene at 6 1C. The reactants and products were online
analyzed on a Shimadzu GC-14C apparatus. The detailed proce-
dures are described in the ESI.†
3. Results and discussion
3.1. Catalytic performance
The catalytic activities of the samples for benzene combustion
are shown in Fig. 1. We used the temperatures T10%, T50%, and
T90% (which corresponded to benzene conversions of 10, 50,
2. Experimental
2.1. Catalyst preparation
3D ordered mesoporous silica (KIT-6) was synthesized using the and 90%) to evaluate the activities of the samples. As revealed
method reported in the literature.28 The 1 wt% Pd/3D ordered by the data in Fig. 1 and Table 1, catalytic activity dropped in
mesoporous Co3O4 (1.0Pd/meso-Co3O4) sample was prepared the order of 0.93Pd/meso-CoO 4 0.85Pd/meso-Co3O4
using a modified KIT-6-templating strategy, by which formation 1.08Pd/meso-Co–CoO. The 0.93Pd/meso-CoO sample showed
of meso-Co3O4 and loading of Pt nanoparticles (NPs) were the highest activity (T10% = 142 1C, T50% = 167 1C, and T90%
4
=
achieved concurrently. 1.0 g of KIT-6 was outgassed in a 189 1C), while the 1.08Pd/meso-Co–CoO sample exhibited the
vacuum oven for 3 h and then dispersed in 50 mL of toluene. lowest activity (T10% = 187 1C, T50% = 246 1C, and T90% = 275 1C)
After the above solution was ultrasonically treated at room at SV = 40 000 mL (g h)À1. The above results demonstrate that
temperature for 0.5 h and stirred at 65 1C for 0.5 h, a well the 0.85Pd/meso-Co3O4 sample obtained after reduction at
ground mixture of 2.0 g of Co(NO3)2Á6H2O and 14 mg of 200 1C was much more active than the unreduced sample or
Pd(NO3)2Á2H2O was added to the above solution at 65 1C, and the sample obtained after reduction at 350 1C. We detected the
the mixture was stirred for 3 h. After filtering and drying at products of benzene combustion, and the CO2 yields were
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. 1 (A) Benzene conversion as a function of temperature, (B) T10%, T50%, and T90%, (C) CO2 yield over 0.85Pd/meso-Co3O4, 0.93Pd/meso-CoO, and
1.08Pd/meso-Co–CoO for benzene oxidation, and (D) benzene conversion as a function of on-stream reaction time over 0.93Pd/meso-CoO for
benzene oxidation at 170 1C and SV = 40 000 mL (g h)À1
.
Table 1 Catalytic activities over the samples at SV = 40 000 mL (g h)À1
was less than 20%, and found that 5-hexen-1-ol, phenol, and
malic acid were the main byproducts of the partial oxidation of
benzene; in the meanwhile some unidentified organics with
high molecular weights were also detected. These unidentified
organics tended to form carbon deposit, leading to higher
temperatures required for deep oxidation of benzene. In addi-
tion, no significant loss in activity was detected over 0.93Pd/
meso-CoO after 10 h of on-stream benzene oxidation (Fig. 1D),
indicating that this sample was catalytically stable. A piece of
evidence for such a deduction was from one of our previously
Benzene oxidation
temperature
CO2 yield temperature
T10% T50% T90% T10%
T50%
(1C)
T90%
(1C)
Sample
(1C)
(1C)
(1C)
(1C)
0.85Pd/meso-Co3O4
0.93Pd/meso-CoO
1.08Pd/meso-Co–CoO 187
160
142
197
167
246
214
189
275
162
147
200
200
168
250
220
192
280
calculated, as shown in Fig. S1 and Fig. 1C (ESI†). CO2 yield reported works, in which the in situ XRD results demonstrated
rose with increasing temperature, which was similar to the that CoO was not oxidized to Co3O4 after CoO was pretreated
tendency in benzene conversion. As shown in Table 1, lower in air at 270 1C for 1 h.12 That is to say, the phase of CoO
temperature was required for benzene conversion than for CO2 in 0.93Pd/meso-CoO was stable in an oxidative atmosphere
yield, which was consistent with the results reported in the below 270 1C.
literature.29,30 Liu et al.29 analyzed the compositions of the
The TOFPd (TOFPd = xC0/nPd, where x, C0 (mol sÀ1), and
effluent gases on a GC–MS apparatus when benzene conversion nPd (mol) are the benzene conversion at 160 1C, initial benzene
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
concentration per second, and Pd molar amount, respectively) carbonate species formed on the surface of the catalyst resulted
and specific reaction rates were calculated according to the in a loss in activity, and such a deactivation induced by CO2 was
activity data and Pd molar amounts, as summarized in reversible. In the case of simultaneous addition of 3.0 vol%
Table S1 (ESI†). Obviously, 0.93Pd/meso-CoO showed a catalytic H2O and 5.0 vol% CO2, the changing trend in activity was rather
activity (TOFPd = 1.46 Â 10À3 sÀ1 and specific reaction rate = similar to that in the presence of 5.0 vol% CO2. Therefore, the
13.74 mmol (gPd s)À1) much higher than that over 0.85Pd/ above results reveal that addition of H2O and/or CO2 to the
meso-Co3O4 (TOFPd = 0.62 Â 10À3 sÀ1 and specific reaction feedstock induced a negative impact on benzene combustion.
rate = 5.85 mmol (gPd s)À1) and 1.08Pd/meso-Co–CoO (TOFPd
=
0.07 Â 10À3 sÀ1 and specific reaction rate = 0.62 mmol (gPd s)À1),
respectively. Summarized in Table S1 (ESI†) are the activities of
the supported noble metal catalysts reported in the literature.
3.2. Crystal structure and textural property
The XRD, Raman, N2 adsorptionÀdesorption, HRTEM, and
HAADF–STEM techniques were used to characterize the crystal
phase compositions, morphologies, and textural properties of
the samples. The small-angle XRD pattern (Fig. S2 (ESI†)) of
0.85Pd/meso-Co3O4 showed significant diffraction peaks in the
region of 2y = 0.8–2.81 assignable to the (211) and (332) crystal
planes, indicative of formation of a good-quality ordered meso-
porous structure. There were well-resolved reflections in the
wide-angle XRD pattern (Fig. 3) of the 0.85Pd/meso-Co3O4
sample, which were assignable to the cubic Co3O4 phase
(JCPDS 74-1657). After 0.85Pd/meso-Co3O4 was pretreated in
H2 at 200 1C, the obtained 0.93Pd/meso-CoO sample displayed
a decreased ordering of the mesoporous structure due to the
disappearance of the (322) crystal phase in the low-angle XRD
pattern. Meanwhile, the wide-angle pattern of 0.93Pd/meso-
CoO showed two weak diffraction peaks at 2y = 36.61 and 42.51,
indicating the formation of the cubic CoO or CoOx phase
(JCPDS 48-1719).12 For the 1.08Pd/meso-Co–CoO sample, new
peaks at 2y = 44.21 and 75.81 appeared, which were assignable
to the metallic Co0 phase (JCPDS 15-0806), although the CoO or
CoOx phase was still detected. No significant signals were
recorded in the low-angle XRD pattern of 1.08Pd/meso-Co–
CoO, suggesting that most of the ordered mesoporous structure
was damaged after treatment in H2 at 350 1C. We detected five
characteristic bands at 193, 477, 516, 617, and 687 cmÀ1 in the
Raman spectrum (Fig. S3 (ESI†)) of 0.85Pd/meso-Co3O4, which
were attributable to the F2g(1), Eg, F2g(2), F2g(3), and A1g
symmetries of Co3O4,11 respectively. Obviously, the reduction
treatment exerted a significant impact on the Raman spectrum
It can be observed that the TOFPd (1.46 Â 10À3
s
À1) and
specific reaction rate (13.74 mmol (gPd s)À1) at 160 1C over
0.93Pd/meso-CoO were significantly higher than those over
6.5Au/meso-Co3O4 (0.33 Â 10À3 sÀ1 and 1.68 mmol (gPd s)À1),31
0.53Pd/Co3O4 (0.30 Â 10À3 sÀ1 and 2.82 mmol (gPd s)À1),32 and
K/Ag-OMS-40 (1.28 Â 10À3 sÀ1 and 11.90 mmol (gPd s)À1),5
but lower than those over 1.0Pd/Al2O3 (2.72 Â 10À3 sÀ1 and
24.37 mmol (gPd s)À1).5
In practical applications, the effects of H2O and CO2 on the
activity of a catalyst should be taken into account. Benzene
combustion over 0.93Pd/meso-CoO was carried out when 3.0 or
5.0 vol% H2O was introduced into the reaction system. It is seen
from Fig. 2 that addition of 3.0 or 5.0 vol% H2O to the feedstock
at 185 1C led to a decrease of 5–8% in benzene conversion.
Nevertheless, benzene conversion was recovered to its original
levels after water vapor was cut off. It has been reported that
there was a reversible inhibition on VOC oxidation by water,
and the significant drop in activity was induced by competitive
adsorption of H2O and VOC molecules.4,33,34 After 5.0 vol% CO2
was introduced into the reaction system, a slow decrease
in benzene conversion from 86 to 75% was observed. Even if
CO2 was removed from the feed stream, benzene conversion
was maintained at 77%. Wang et al.35 pointed out that the
Fig. 2 Effects of (a) 3.0 or 5.0 vol% H2O, (b) 5.0 vol% CO2, and (c) 3.0 vol%
H2O + 5.0 vol% CO2 on benzene conversion over the 0.93Pd/meso-CoO
sample at 185 1C and SV = 40 000 mL (g h)À1
.
Fig. 3 Wide-angle XRD patterns of the 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
of the sample. After reduction of 0.85Pd/meso-Co3O4, the NPs were well distributed on the surface of each sample and
Raman signals were weakened. There was a significant red no significant Pd NP agglomeration was observed.
shift (ca. 12 cmÀ1) in the A1g symmetry of the 0.93Pd/meso-CoO
sample (at 675 cmÀ1), which was much lower than that of the
3.3. Surface property
0.85Pd/meso-Co3O4 sample. The results demonstrate that lat- Illustrated in Fig. 6 are the Co 2p3/2, Pd 3d, and O 1s XPS spectra
tice distortion and residual stress of the spinel crystal were of the samples, and their surface parameters are listed in
generated during the reduction process.11,36 Theoretically, a Table 2. The asymmetrical Co 2p3/2 XPS spectrum (Fig. 6A) of
more amount of oxygen vacancies can be formed due to more each sample contained three components at BE = 779.8 eV
lattice defects, which would favor activation of O2 to the active (assignable to the surface Co3+ species), 781.8 eV (attributable
oxygen adspecies and hence greatly enhance the activity of to the surface Co2+ species), and 786.2 eV (the shake-up satellite
a catalyst.
of Co2+).27 After reduction of 0.85Pd/meso-Co3O4 at 200 or
The 0.85Pd/meso-Co3O4 and 0.93Pd/meso-CoO samples dis- 350 1C, the surface Co2+/Co3+ molar ratio rose from 0.58 to
played an IV-type N2 adsorptionÀdesorption isotherm and a 0.86, i.e., the surface Co2+ concentration increased. That is
H1-type hysteresis loop (Fig. 4), suggesting formation of an to say, compared with the 0.85Pd/meso-Co3O4 sample, the
ordered mesoporous structure. The hysteresis loop of the reduced samples showed a more amount of oxygen vacancies
0.93Pd/meso-CoO sample was slightly shifted to the higher on the surface. The asymmetrical Pd 3d XPS spectrum of the
relative pressure, a result possibly due to partial sintering of 0.85Pd/meso-Co3O4 sample (Fig. 6B) was decomposed into four
the mesoporous structure.12 The average pore diameter of the components at BE = 335.3 and 340.5 eV (attributed to the
0.93Pd/meso-CoO sample was 5.0 nm, which was similar to that surface Pd0 species) and 336.9 and 342.2 eV (assigned to the
(4.8 nm) of the 0.85Pd/meso-Co3O4 sample. Summarized in surface Pd2+ species).16,19,27 Only Pd0 species, however, were
Table 2 are textural parameters of the samples. BET surface detected on the surface of 0.93Pd/meso-CoO and 1.08Pd/meso-
areas of the 0.85Pd/meso-Co3O4, 0.93Pd/meso-CoO, and 1.08Pd/ Co–CoO. The above results confirm that the H2 pretreatment
meso-Co–CoO samples were 91.2, 88.5, and 52.6 m2 gÀ1, respec- exerted a significant influence on the oxidation state of the
tively. The results indicate that the reduction treatment induced surface metals. The asymmetrical O 1s XPS spectrum of each
an obvious change in the mesoporous structure.
sample (Fig. 6C) was deconvoluted into three components at
TEM images (Fig. S4 (ESI†)) of 0.85Pd/meso-Co3O4, 0.93Pd/ BE = 529.0, 530.9, and 532.7 eV, which could be assignable to
meso-CoO, and 1.08Pd/meso-Co–CoO show highly ordered the surface lattice oxygen (Olatt), adsorbed oxygen (Oads, e.g.,
2À
mesopore arrays. A rise in reduction temperature to 350 1C O2À, O2 or OÀ), and hydroxyl and/or carbonate species,27
resulted in significant agglomeration of Pd nanoparticles (NPs) respectively. The Oads/Olatt molar ratio increased in the
and enlargement in mesopore diameter. It should be noted that sequence of 0.85Pd/meso-Co3O4 (0.85) o 0.93Pd/meso-CoO
no distinct Pd NPs could be seen in the TEM image of the (1.52) o 1.08Pd/meso-Co–CoO (2.68) (Table 2). It was reported
0.85Pd/meso-Co3O4 or 0.93Pd/meso-CoO sample. Fig. 5 shows that the more the surface oxygen vacancy amount, the easier
HAADF-STEM images and element mappings of the 0.85Pd/ the activation of O2, leading to better performance for VOC
meso-Co3O4 and 0.93Pd/meso-CoO samples. Apparently, Pd combustion.19,20 The higher Oads concentration is an indication
Fig. 4 (A) Nitrogen adsorption–desorption isotherms and (B) pore-size distributions of (a) 0.85Pd/meso-Co3O4, (b) 0.93Pd/meso-CoO, and (c) 1.08Pd/
meso-Co–CoO.
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
Table 2 BET surface area, pore size, pore volume, Pd content, surface element composition, H2 consumption, and oxygen desorption of the samples
Surface element
composition
H2 consumption Oxygen desorption
Textural parameter
(mmol gÀ1
)
(mmol gÀ1
)
Surface area Pore size Pore volume Pd content Co2+/Co3+
O
ads/Olatt
Below
400 1C
Above
400 1C
Sample
(m2 gÀ1
)
(nm)
(cm3 gÀ1
)
(wt%)
molar ratio molar ratio Above 200 1C
0.85Pd/meso-Co3O4
0.93Pd/meso-CoO
1.08Pd/meso-Co–CoO 52.6
91.2
88.5
4.8
5.0
10.0
0.17
0.15
0.13
0.85
0.93
1.08
0.58
0.70
0.86
0.85
1.51
2.68
16.12
15.19
8.49
7.0
37.1
17.4
33.8
10.6
7.4
Fig. 5 HAADF-STEM images and elemental mappings of (a–d) 0.85Pd/meso-Co3O4 and (e–h) 0.93Pd/meso-CoO.
that the reduction favored formation of the active oxygen required for reduction from Co3+ to Co2+ or from Co2+ to Co0 in
species, hence enhancing the catalytic activity of the sample. 0.93Pd/meso-CoO and 1.08Pd/meso-Co–CoO. Dacquin et al.37
As demonstrated above, the 1.08Pd/meso-Co–CoO sample reported the reduction of Co3+ to Co2+ at ca. 200 1C due to the
exhibited the highest molar ratios of surface Co2+/Co3+ (0.86) H2 spillover effect. In the present work, the H2 spillover effect
and Oads/Olatt (2.68). However, the 0.93Pd/meso-CoO sample could also exist in the reduction of Co3+ to Co2+ or Co2+ to Co0.
performed the best for benzene combustion. This fact clearly Quantitative analysis of the reduction peaks above 200 1C in the
suggests that other factors might govern the activity of the H2-TPR profiles can obtain H2 consumption of the samples
sample except for oxygen activation ability and Oads species (Table 2). Theoretically speaking, the H2 consumption is
concentration. In our present work, the 350 1C-reduction 13.33 and 18.07 mmol gÀ1 supposing that the Co ions in cobalt
pretreatment of 0.85Pd/meso-Co3O4 resulted in the 1.08Pd/ oxide are present in the form of only Co2+ and only Co3+ and are
meso-Co–CoO sample with the lowest surface area, which reduced to Co0, respectively. Apparently, the H2 consumption
was possibly associated with its worst activity for benzene corresponding to the reduction peaks above 200 1C was
combustion.
13.33–18.07 mmol gÀ1 (Table 2). This result suggests that cobalt
ions in these samples were present in mixed oxidation states
(i.e., Co3+ and Co2+), which was in good consistency with
3.4. Redox property
Illustrated in Fig. 7A are the H2-TPR profiles of the as-prepared the reported results.38 More importantly, H2 consumption
samples. The 0.85Pd/meso-Co3O4 sample showed three peaks (16.12 mmol gÀ1) of 0.85Pd/meso-Co3O4 was higher than that
at 130 1C (peak I), 300 1C (peak II), and 375 1C (peak III), which (15.19 mmol gÀ1) of 0.93Pd/meso-CoO, revealing that there was
were the reduction of PdO to Pd0, Co3+ to Co2+, and Co2+ to Co0, a less amount of Co3+ in the latter than in the former. H2
respectively. The H2 reduction treatment brought about a consumption (8.49 mmol gÀ1) of the 1.08Pd/meso-Co–CoO
significant change in reduction behavior. Firstly, the peak at sample, however, decreased dramatically, indicating that part
130 1C owing to PdO reduction in 0.93Pd/meso-CoO and of Co3+ and/or Co2+ were reduced to Co0. On the basis of the
1.08Pd/meso-Co–CoO disappeared, indicating that Pd was pre- H2-TPR results, we can realize that PdO could be reduced to
sent in metallic Pd0. Secondly, the lower temperatures were metallic Pd0 in 3.5 vol% H2/96.5 vol% Ar below 200 1C, but most
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. 6 (A) Co 2p3/2, (B) Pd 3d, and (C) O 1s XPS spectra of (a) 0.85Pd/meso-Co3O4, (b) 0.93Pd/meso-CoO, and (c) 1.08Pd/meso-Co–CoO.
Fig. 7 (A) H2-TPR and (B) O2-TPD profiles of (a) 0.85Pd/meso-Co3O4, (b) 0.93Pd/meso-CoO, and (c) 1.08Pd/meso-Co–CoO.
of the Co3+ and Co2+ species were retained in the 0.93Pd/meso- adsorption, activation, consumption, and replenishment in
CoO sample. Some authors proposed that the samples with VOC combustion over a catalyst. The active oxygen species were
high amounts of the Co2+ species and high Oads species con- vital in the oxidation of intermediates on the catalyst surface.13,27,39
centrations possessed more amounts of oxygen vacancies, Hence, the more the Co2+ and Oads species in/on the sample,
which were beneficial for the activation of gas-phase oxygen the better the catalytic activity for the redox-involving reactions
to active Oads species. It is known that oxygen is involved in (e.g., benzene combustion). The 1.08Pd/meso-Co–CoO sample
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
contained a large amount of inactive Co0 species that resulted anti-symmetric vibration of C–O in the acetate species,12,41,42
in a decrease of catalytic activity. Moreover, the poor meso- the ones at 1532 and 1307 cmÀ1 were due to the maleate
porous structure and low surface area were also accountable for species,42–44 the one at 1400 cmÀ1 (which was slightly shifted
the drop in the catalytic activity of 1.08Pd/meso-Co–CoO.
to 1422 cmÀ1) was attributed to the symmetric vibration of C–O
To further investigate the nature and amounts of different in the acetate species,12,42 the ones at 1481–1493 and 1443 cmÀ1
oxygen species, we recorded the O2-TPD profiles of the samples were ascribable to the skeletal C–C vibration,40,44,45 and the one
(Fig. 7B). According to the literature,9,11,39 the peaks below at 1469 cmÀ1 were attributable to the C–H deformation vibration
400 1C were due to desorption of the Oads species, the ones in of the aromatic ring,40,44,45 and the one at 1242 cmÀ1 was
the range of 400–500 1C were ascribed to desorption of the assignable to the surface phenolate species.42 It is clearly seen
coordinately unsaturated surface Olatt species, and the ones that the intensity of these absorption bands of the 0.93Pd/meso-
above 700 1C were assigned to desorption of the coordinately CoO sample increased monotonically with the exposure time,
saturated Olatt species. For the 0.85Pd/meso-Co3O4 sample, the indicating that the organic byproducts were accumulated on
peaks at 165 and 280 1C were assignable to desorption of the the sample surface. Partially oxidized products (e.g., acetate,
Oads species, whereas the one at 810 1C was attributable to maleate, formate, and phenolate species) were formed during
desorption of the Olatt species in Co3O4. Differently from the the adsorption treatment process in the absence of O2, suggest-
0.85Pd/meso-Co3O4 sample, 0.93Pd/meso-CoO showed three ing involvement of the Oads species in the benzene combustion
strong oxygen desorption signals at 160, 305, and 348 1C, process. Up to now, there have been a number of research
assignable to desorption of the Oads species in the different works on the role of surface oxygen species in the catalytic
coordination environments. In addition, a desorption peak oxidation of organic compounds in the literature.39,42,46–48
appeared at 439 1C, which was assignable to desorption of Some authors attributed the cleavage of the benzene ring to
lattice oxygen in CoO. We observed similar oxygen desorption the amount of the electrophilic Oads (O2À, O22À or OÀ) species,
profiles in the 1.08Pd/meso-Co–CoO sample. It should be and thought that oxidation of small carboxylate molecules was
pointed out that no apparent signals attributable to desorption associated with the oxidation ability of the nucleophilic Olatt
of lattice oxygen at 810 1C were recorded in the 0.93Pd/meso- species.39,48 As discussed in the XPS characterization results,
CoO and 1.08Pd/meso-Co–CoO samples, indicating that the H2 reduction pretreatment increased the amount of Oads
Co3O4 phase was converted to the CoO and/or Co phase. We species in the 0.93Pd/meso-CoO sample, and the enhanced
quantitatively analyzed the oxygen desorption of the samples in Oads species could efficiently attack the adsorbed benzene
a certain temperature range (Table 2). Desorption of the Oads molecules to form the products in the absence of gas-phase
species below 400 1C significantly increased after the H2 oxygen. It should be pointed out that the surface Olatt species
reduction pretreatment, and the changing trend decreased in with highly negative charge and nucleophilic property could
the sequence of 0.93Pd/meso-CoO (37.1 mmol gÀ1) 4 1.08Pd/ favor the oxidation of carboxylates.
meso-Co–CoO (17.4 mmol gÀ1) 4 0.85Pd/meso-Co3O4 (7.0 mmol gÀ1).
When the atmosphere was switched from benzene to N2 and
The desorbed oxygen species could be ascribed to the Oads after N2 purging for 1 h, the absorption bands attributable to
species bound to the Co3+ sites since the Co2+ species were first the gaseous benzene molecules and aromatic ring disappeared.
easily oxidized into the Co3+ species by gas-phase O2 and then Acetate, maleate, and phenolate species were retained in the
the surface oxygen species were bound to the Co3+ sites.
flow of N2 (Fig. S7 (ESI†)). The same process of benzene
adsorption on the 0.85Pd/meso-Co3O4 sample was carried out
and the in situ DRIFTS spectra were recorded. It is seen that
there were absorption bands attributable to the acetate, maleate
3.5. Possible intermediates formed during the benzene
oxidation process
The in situ DRIFTS characterization was performed to deter- and phenolate species after N2 purging, similar to the in situ
mine the intermediate species generated on the surface of the DRIFTS spectra of the 0.93Pd/meso-CoO sample. Hence, it can
samples, so that benzene combustion mechanisms could be be concluded that benzene underwent the same activation path-
investigated. In situ DRIFTS spectra (Fig. S5 and S6 (ESI†)) of way on the surface of the two samples.
the samples were collected as a function of time after 1000 ppm
The benzene-adsorbed 0.93Pd/meso-CoO sample was then
benzene + N2 (balance) was introduced into the reaction cell at consecutively exposed to a flow of O2 to testify the reactivity of
150 1C. With the introduction of benzene, numerous absorp- the formed intermediates at different temperatures. As shown
tion bands were recorded. The absorption bands cannot be in Fig. 8A, there were slight decreases in the intensity of the
distinguished precisely in the region of 1800–1100 cmÀ1 since a in situ DRIFTS signals (especially in the intensity of the signal
large number of different species have overlapping absorption due to the acetate species). This observation suggests that the
bands in this region. On the basis of the assignments reported Oads species generated from O2 activation reacted with the
in the literature (Table S2 (ESI†)), we can tentatively assign all of intermediates, and the reactivity of the acetate species was
the absorption bands of the samples. The bands at 3657–3705, higher than that of the other intermediates.45 Moreover, no new
1700–2000, and 600–1100 cmÀ1 were due to the adsorbed absorption bands were recorded, indicating that the intermediates
gaseous benzene (not shown),13 the one at 1596 cmÀ1 was had already been generated via the interaction of benzene and
attributed to the anti-symmetric vibration of C–O in the formate Oads species. When the temperature was increased to 180 or
species,40 the ones at 1555–1565 cmÀ1 were ascribed to the 200 1C, these intermediate bands became weak in intensity or
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. 8 In situ temperature-dependent DRIFTS spectra of (A) 0.93Pd/meso-CoO and (B) 0.85Pd/meso-Co3O4 in a 20 vol% O2 (N2 as balance) flow after
benzene pre-adsorption at 150 1C.
even disappeared. The decreasing trend in the intensity of the The results demonstrate that the intermediate species were
species in the 1800–1100 cmÀ1 region suggests that these species stable below 200 1C, and could react with the Oads species at
were the main intermediates in the catalytic combustion of benzene, 250 1C to generate CO2 and H2O.
and that they were further catalytically oxidized into CO2 and H2O.
Furthermore, the integral DRIFTS area of the representative
This finding coincided with that reported by other researchers.12 acetate species centered at 1422–1400 cmÀ1 was plotted as a
Comparatively, the intermediates formed on the 0.85Pd/meso-Co3O4 function of time, as shown in Fig. 10 and 11. After making a
sample disappeared at a higher temperature (Fig. 8B). This result quantitative analysis (Fig. 10A), one can know that the acetate
well explained the changing trend in the catalytic activity of the formation rate was higher over 0.93Pd/meso-CoO than that over
samples. Therefore, we conclude that the 0.93Pd/meso-CoO sample 0.85Pd/meso-Co3O4, implying that the metallic Pd0 species
favored the dissociative adsorption of O2 at a lower temperature, as possessed a stronger benzene activation ability than the oxi-
demonstrated by the O2-TPD and XPS results.
dized Pd species. It is reasonably inferred from Fig. 10B and 11
In order to distinguish the impacts of Oads and gas-phase O2, that the atmosphere exerted a little effect on the formation of
in situ DRIFTS experiments on the 0.93Pd/meso-CoO and the acetate species at 150 1C, but further introducing O2
0.85Pd/meso-Co3O4 samples after benzene pre-adsorption at resulted in a decrease in the intensity of the acetate species
150 1C were carried out by elevating the temperature in a N2 due to their reaction. The integral DRIFTS area of each sample
atmosphere (Fig. 9). It is noteworthy that the two samples also declined with the elevated temperature (Fig. 11A and B).
displayed similar DRIFTS spectra with only slightly decreased Finally, the band at 1422–1400 cmÀ1 disappeared over 0.93Pd/
intensity at 180 and 200 1C, as compared with that at 150 1C. meso-CoO at 200 1C and over 0.85Pd/meso-Co3O4 at 250 1C,
At 250 1C, however, all of the absorption bands disappeared. indicating that the acetate species were completely eliminated.
Fig. 9 In situ temperature-dependent DRIFTS spectra of (A) 0.93Pd/meso-CoO and (B) 0.85Pd/meso-Co3O4 in a N2 flow after benzene pre-adsorption
at 150 1C.
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 Time-dependent trends of integral DRIFTS areas corresponding to the acetate species (at 1400–1422 cmÀ1) on the samples in a flow of
(A) 1000 ppm benzene + N2 (in the range of 1–60 min) and (B) N2 (in the range of 61–120 min).
Fig. 11 Time-dependent trends of integral DRIFTS areas corresponding to the acetate species (at 1400–1422 cmÀ1) on (A) 0.93Pd/meso-CoO and
(B) 0.85Pd/meso-Co3O4 in a flow of 20 vol% O2 + N2 at different temperatures.
By combining the above in situ DRIFTS results as well as the 0.85Pd/meso-Co3O4 sample in H2 at 200 and 350 1C,
those reported in the literature, we propose the five-step respectively. The H2 reduction treatment led to partial mesos-
catalytic benzene oxidation mechanism over the cobalt oxide- tructure destroying in meso-Co3O4, decrease in surface area,
supported Pd samples: (i) benzene adsorption, (ii) formation of increase in pore size, and big Pd NPs in 1.08Pd/meso-Co–CoO.
phenolate, (iii) breaking of the benzene ring due to the attack of The 0.93Pd/meso-CoO sample with the surface Pd0 being the
the active surface species, (iv) generation of the oxygenate dominant species exhibited the highest benzene oxidation
species (carbonate bidentate, formate, acetate, and maleate), activity (T10% = 142 1C, T50% = 167 1C, and T90% = 189 1C at a
and (v) further reaction of the intermediate species with the SV of 40 000 mL (g h)À1). A higher temperature was required for
Oads species to produce CO2 and H2O.
CO2 yield than for benzene conversion. Carbonate bidentate,
formate, acetate, maleate, and phenolate were the main inter-
mediate species during the benzene combustion process, and
the 0.93Pd/meso-CoO sample exhibited the strongest oxygen
activation ability among all of the samples. It is concluded that
4. Conclusions
The 0.85Pd/meso-Co3O4 sample was synthesized via a one-step the outstanding catalytic performance of 0.93Pd/meso-CoO was
KIT-templating nanocasting route, and the 0.93Pd/meso-CoO related to the mainly formed Pd0 species, good oxygen activa-
and 1.08Pd/meso-Co–CoO samples were prepared by reducing tion ability, and high surface area.
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
23 T. J. Hua, Y. Wang, L. N. Zhang, T. Tang, H. Xiao, W. W.
Chen, M. Zhao, J. F. Jia and H. Y. Zhu, Appl. Catal., B, 2019,
243, 175–182.
Conflicts of interest
There are no conflicts of interest to declare.
24 S. C. Kim and W. G. Shim, Appl. Catal., B, 2009, 92, 429–436.
25 Y. F. Wang, C. B. Zhang and H. He, ChemCatChem, 2018, 10,
998–1004.
Acknowledgements
26 H. Tan, J. Wang, S. Yu and K. Zhou, Environ. Sci. Technol.,
2015, 49, 8675–8682.
27 S. H. Xie, J. G. Deng, S. M. Zang, H. G. Yang, G. S. Guo,
H. Arandiyan and H. X. Dai, J. Catal., 2015, 322, 38–48.
28 F. Kleitz, S. H. Choi and R. Ryoo, Chem. Commun., 2003,
2136–2137.
This work was supported by the National Natural Science
Foundation of China (Grant No. 21677004 and 21876006).
References
1 M. S. Kamal, S. A. Razzk and M. M. Hossain, Atmos. Environ., 29 L. Z. Liu, J. X. Li, H. B. Zhang, L. Li, P. Zhou, X. L. Meng,
2016, 140, 117–134.
2 M. Hakim, Y. Y. Broza, O. Barash, N. Peled, M. Phillips,
M. M. Guo, J. P. Jia and T. H. Sun, J. Hazard. Mater., 2019,
362, 178–186.
A. Amann and H. Haick, Chem. Rev., 2012, 112, 5949–5966. 30 J. Chen, X. Chen, W. J. Xu, Z. Xu, H. P. Jia and J. Chen, Appl.
3 A. L. Bolden, C. F. Kwiatkowski and T. Colborn, Environ. Sci.
Technol., 2015, 49, 5261–5276.
4 L. F. Liotta, Appl. Catal., B, 2010, 100, 403–412.
Catal., B, 2018, 224, 825–835.
31 Y. X. Liu, H. X. Dai, J. G. Deng, S. H. Xie, H. G. Yang, W. Tan,
W. Han, Y. Jiang and G. S. Guo, J. Catal., 2014, 309, 408–418.
5 H. Deng, S. Y. Kang, J. Z. Ma, C. B. Zhang and H. He, Appl. 32 J. Q. Li, W. H. Li, G. Liu, Y. Z. Deng, J. Yang and Y. F. Chen,
Catal., B, 2018, 239, 214–222. Catal. Lett., 2016, 146, 1300–1308.
6 Z. Y. Chen, J. X. Mao and R. X. Zhou, Appl. Surf. Sci., 2019, 33 Y. J. Luo, K. C. Wang, J. C. Zuo, Q. R. Qian, Y. X. Xu,
465, 15–22.
X. P. Liu, H. Xue and Q. H. Chen, Mol. Catal., 2017, 436,
7 H. X. Dai, Sci. Bull., 2015, 60, 1708–1710.
259–266.
8 H. B. Huang, Y. Xu, Q. Y. Feng and D. Y. C. Leung, Catal. Sci. 34 R. Gholami, M. Alyani and K. J. Smith, Catalysts, 2015, 5,
Technol., 2015, 5, 2649–2669. 561–594.
9 Z. Wang, W. Z. Wang, L. Zhang and D. Jiang, Catal. Sci. 35 H. Wang, W. Yang, P. H. Tian, J. Zhou, R. Tang and S. J. Wu,
Technol., 2016, 6, 3845–3853. Appl. Catal., A, 2017, 529, 60–67.
10 F. G. Wang, L. J. Zhang, L. L. Xu, Z. Y. Deng and W. D. Shi, 36 Q. Liu, L. C. Wang, M. Chen, Y. Cao, H. Y. He and K. N. Fan,
Fuel, 2017, 203, 419–429. J. Catal., 2009, 263, 104–113.
11 W. X. Tang, W. Xiao, S. B. Wang, Z. Ren, J. Ding and 37 J. P. Dacquin, C. Dujardin and P. Granger, J. Catal., 2008,
P. X. Gao, Appl. Catal., B, 2018, 226, 585–595. 253, 37–49.
12 S. H. Xie, Y. X. Liu, J. G. Deng, J. Yang, X. T. Zhao, Z. Han, 38 Z. X. Wu, J. G. Deng, Y. X. Liu, S. H. Xie, Y. Jiang, X. T. Zhao,
K. F. Zhang and H. X. Dai, J. Catal., 2017, 352, 282–292.
13 X. Y. Wang, Y. Liu, T. H. Zhang, Y. J. Luo, Z. X. Lan,
J. Yang, H. Arandiyan, G. S. Guo and H. X. Dai, J. Catal.,
2015, 332, 13–24.
K. Zhang, J. C. Zuo, L. L. Jiang and R. H. Wang, ACS Catal., 39 W. Jiang, Y. J. Pang, L. L. Gu, Y. Yao, Q. Su, W. J. Ji and
2017, 7, 1626–1636. C. T. Au, J. Catal., 2017, 349, 183–196.
14 J. Gonzalez-Prior, R. Lopez-Fonseca, J. I. Gutierrez-Ortiz and 40 C. P. O’Brien and I. C. Lee, J. Catal., 2017, 347, 1–8.
´
´
´
B. D. Rivas, Appl. Catal., B, 2018, 222, 9–17.
15 S. Zhao, K. Z. Li, S. Jiang and J. H. Li, Appl. Catal., B, 2016,
181, 236–248.
41 X. Y. Wang, W. T. Zhao, X. W. Wu, T. H. Zhang, Y. Liu,
K. Zhang, Y. H. Xiao and L. L. Jiang, Appl. Surf. Sci., 2017,
426, 1198–1205.
16 Y. B. Li, C. B. Zhang, J. Z. Ma, M. Chen, H. Deng and H. He, 42 W. Deng, Q. G. Dai, Y. J. Lao, B. B. Shi and X. Y. Wang, Appl.
Appl. Catal., B, 2017, 217, 560–569. Catal., B, 2016, 181, 848–861.
17 C. B. Zhang, Y. B. Li, Y. F. Wang and H. He, Environ. Sci. 43 X. L. Liu, J. L. Zeng, W. B. Shi, J. Wang, T. Y. Zhu and
Technol., 2014, 48, 5816–5822. Y. F. Chen, Catal. Sci. Technol., 2017, 7, 213–221.
18 C. H. Liu, C. Y. Lin, J. L. Chen, K. T. Lu, J. F. Lee and 44 K. Yang, Y. X. Liu, J. G. Deng, X. T. Zhao, J. Yang, Z. Han,
J. M. Chen, J. Catal., 2017, 350, 21–29. Z. Q. Hou and H. X. Dai, Appl. Catal., B, 2019, 244, 650–659.
19 S. Zhao, F. Y. Hu and J. H. Li, ACS Catal., 2016, 6, 45 X. C. Sun, J. Lin, H. L. Guan, L. Li, L. Sun, Y. H. Wang,
3433–3441.
S. Miao, Y. Su and X. D. Wang, Appl. Catal., B, 2018, 226,
20 S. H. Xie, Y. X. Liu, J. G. Deng, J. Yang, X. T. Zhao, Z. Han,
575–584.
K. F. Zhang, Y. Wang, H. Arandiyan and H. X. Dai, Catal. Sci. 46 Y. B. Li, X. Y. Chen, C. Y. Wang, C. B. Zhang and H. He, ACS
Technol., 2018, 8, 806–816. Catal., 2018, 8, 11377–11385.
21 Y. F. Wang, C. B. Zhang, F. D. Liu and H. He, Appl. Catal., B, 47 Y. S. Wu, S. S. Yuan, R. Feng, Z. C. Ma, Y. Z. Gao and
2013, 142–143, 72–79. S. T. Xing, Mol. Catal., 2017, 442, 164–172.
22 G. Ercolino, P. Stelmachowski, G. Grzybek, A. Kotarba and 48 V. P. Santos, M. F. R. Pereira, J. J. M. Orfao and
S. Specchia, Appl. Catal., B, 2017, 206, 712–725. J. L. Figueiredo, Appl. Catal., B, 2010, 99, 353–363.
´
˜
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.