Highly improved acetone oxidation activity over mesoporous hollow nanospherical Mn:XCo3- xO4 solid solutions

Article abstract of DOI:10.1039/c9cy01791g

Hierarchical bimetal oxide solid solutions with mesoporous hollow-interior architecture have greatly promising applications in environmental catalysis due to their exceptional catalytic performances, while facile synthesis of such a unique architecture is still challenging. Herein, mesoporous hollow nanospherical (MHS) Mn_xCo_3-xO₄ solid solutions with tunable Mn/Co molar ratios in the spinel-type cubic crystal phase, fabricated by a facile solvothermal alcoholysis approach, have been developed to catalyze acetone oxidation for the first time. Experimental results confirm that MHS Mn_1.20Co_1.80O₄ gave the optimal acetone oxidation activity, achieving 100% acetone conversion and 100% CO₂ selectivity at 140 °C under the reaction conditions of acetone concentration = 1000 ppm, 20%O₂/N₂ and WHSV = ca. 93000 mL g_cat^-1 h^-1. This superior activity of MHS Mn_1.20Co_1.80O₄ can be mainly ascribed to the strong synergetic effect of the unique mesoporous hollow nanospherical feature, abundant oxygen vacancies and surface active oxygen species, the enhanced chemical nature resulting from the presence of more Mn⁴⁺, Mn³⁺ and Co³⁺ reactive sites, and the improved redox capacities. Kinetics studies provide direct evidence to validate the finding that the strong synergetic effect rather than individual factors played the critical role in determining the acetone oxidation activity. Meanwhile, a plausible catalytic mechanism of acetone oxidation over MHS Mn_xCo_3-xO₄ solid solutions has been proposed and elaborated in detail by elucidating the structure-activity relationship. Furthermore, MHS Mn_1.20Co_1.80O₄ also presented excellent long-term stability and good water tolerance due to its highly stable crystal phase and robust morphological structure, showing good potential in eliminating volatile organic compounds.

Full text of DOI:10.1039/c9cy01791g

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Highly improved acetone oxidation activity over
mesoporous hollow nanospherical MnxCo3−xO4
solid solutions†
Cite this: Catal. Sci. Technol., 2019,
9, 6379
*
Jinguo Wang,
Chi Zhang, Shuaifeng Yang, Hao Liang and Yong Men
Hierarchical bimetal oxide solid solutions with mesoporous hollow-interior architecture have greatly prom-
ising applications in environmental catalysis due to their exceptional catalytic performances, while facile
synthesis of such a unique architecture is still challenging. Herein, mesoporous hollow nanospherical (MHS)
MnxCo3−xO4 solid solutions with tunable Mn/Co molar ratios in the spinel-type cubic crystal phase, fabri-
cated by a facile solvothermal alcoholysis approach, have been developed to catalyze acetone oxidation
for the first time. Experimental results confirm that MHS Mn1.20Co1.80O4 gave the optimal acetone oxidation
activity, achieving 100% acetone conversion and 100% CO2 selectivity at 140 °C under the reaction condi-
tions of acetone concentration = 1000 ppm, 20%O2/N2 and WHSV = ca. 93000 mL gcat−1 h−1. This superior
activity of MHS Mn1.20Co1.80O4 can be mainly ascribed to the strong synergetic effect of the unique meso-
porous hollow nanospherical feature, abundant oxygen vacancies and surface active oxygen species, the
enhanced chemical nature resulting from the presence of more Mn4+, Mn3+ and Co3+ reactive sites, and
the improved redox capacities. Kinetics studies provide direct evidence to validate the finding that the
strong synergetic effect rather than individual factors played the critical role in determining the acetone
oxidation activity. Meanwhile, a plausible catalytic mechanism of acetone oxidation over MHS MnxCo3−xO4
solid solutions has been proposed and elaborated in detail by elucidating the structure–activity relationship.
Furthermore, MHS Mn1.20Co1.80O4 also presented excellent long-term stability and good water tolerance
due to its highly stable crystal phase and robust morphological structure, showing good potential in elimi-
nating volatile organic compounds.
Received 4th September 2019,
Accepted 10th October 2019
DOI: 10.1039/c9cy01791g
rsc.li/catalysis
and catalytic oxidation, have been widely explored to reduce
VOC emissions,1–6 among which catalytic oxidation is gener-
1. Introduction
Volatile organic compounds (VOCs), as one group of the most
important contributors to atmospheric pollution,1–3 usually
come from various industrial processes, automobiles and mu-
nicipal incineration, which seriously threaten public health
and environmental quality due to their extremely toxic
essence.1–6 Driven by the increasingly rigorous environmental
legislation for VOC emissions, several end-of-pipe techniques,
mainly including adsorption, condensation, non-thermal
plasma, thermal incineration, photocatalysis, biotechnology
ally believed to be one of the most efficient and cost-effective
pathways for VOC elimination owing to its high efficiency,
low operating temperature and lack of secondary
pollutants.5–9 Nonetheless, the major challenge of such a
pathway is still how to acquire highly-efficient catalysts.7–9
For alleviating this major challenge, two effective ways have
been developed to acquire highly-efficient catalysts for VOC
elimination: one is to design catalysts with special morpho-
logical features and the other is to improve the catalyst's in-
trinsic activity by modulating its active components.
During the last few decades, hierarchical materials with
special morphological features, e.g. nanotubes, yolk–shells,
porous networks, spherical particles and others, have
attracted more and more attention in the fields of adsorp-
tion, optical devices, drug delivery, energy storage and hetero-
geneous catalysis due to their exceptional structure–activity
relationships.9–17 Among various hierarchical morphologies,
mesoporous hollow nanospheres (MHS), usually integrating
the hollow-interior void with mesoporous shells of various
dimensions into one special structure, have received much
College of Chemistry and Chemical Engineering, Shanghai University of
Engineering Science, Shanghai 201620, P. R. China.
E-mail: Jinguowang1982@sues.edu.cn; Tel: +86 21 6787 4046
† Electronic supplementary information (ESI) available: Schematic flow chart of
the reaction system for acetone oxidation, XPS surveys of different catalysts,
FESEM images of crushed Mn1.20Co1.80O4, acetone conversion curves and the
produced CO2 concentration curves over Mn1.20Co1.80O4 and crushed
Mn1.20Co1.80O4, and XRD pattern and FESEM image of Mn1.20Co1.80O4 after
reacting for 60 h at 140 °C with the effect of H2O vapor and activity data of dif-
ferent catalysts for acetone oxidation according to previous studies. See DOI:
10.1039/c9cy01791g
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6379

View Article Online
Paper
Catalysis Science & Technology
interest in heterogeneous catalysis owing to their distinct ad-
vantages of high surface area, large hollow-interior void and
mesoporous channels within porous shells.9,15–17 The high
surface area can afford abundant active sites for the adsorp-
tion and activation of reactant molecules, and the large
hollow-interior void not only benefits the improved adsorp-
tion of reactant molecules but also increases the contact fre-
quency of the reactant molecule–catalyst when reactant mole-
cules diffuse in and out of the large hollow-interior void.15,16
In addition, the mesoporous channels within the porous
shells favor the decrease in mass transfer blockage, which in
turn facilitates the adsorption and diffusion of reactant mole-
cules.16,17 Although MHS have considerably distinct merits,
facile synthesis of such architecture is still challenging. Cur-
rently, the most adopted strategy for synthesis of MHS is the
template approach, mainly including soft-templates (e.g. sur-
factants, macromolecules, vesicles, microemulsions, gas bub-
bles or oil droplets) and hard-templates (e.g. silica, carbon,
polymers, metals or metal oxides).15–19 However, the template
approach requires the sacrifice of templates during the prep-
aration process, inevitably increasing the complexity and un-
certainty. Thus, template-free strategies such as solvothermal
and hydrothermal methods are much more attractive.
that MnCo2O4.5 gave the optimal CO oxidation activity be-
cause of its unique mesoporous structure.33 Binh et al. re-
vealed that nano-spinel MeCo2O4 (Me: Mn, Ni, Cu) powders
demonstrated dramatically enhanced toluene oxidation activ-
ity and their activities varied in the order CuCo2O4
>
MnCo2O4 > NiCo2O4 owing to their distinct reduction capa-
bilities.34 Jiang et al. studied ordered mesoporous ACo2O4 (A:
Cu, Ni, Mn) spinel catalysts and found that ordered meso-
porous MnCo2O4 showed the highest benzene oxidation activ-
ity, mainly ascribable to its higher surface Mn concentration
and more active oxygen species,35 while Kim et al. also inves-
tigated ordered mesoporous MCo2O4 (M: Cu, Zn, Ni) spinel
catalysts and disclosed their remarkably improved methane
oxidation activity as a result of the increased number of their
surface Co3+ cations.36 Considering the above reports, it can
be concluded that the intrinsic activity of spinel Co3O4 can
be greatly improved by incorporating other transition-metal
cations into its lattice to form solid solutions, especially the
incorporation of Mn cations. Although much effort has been
done to improve the intrinsic activity of spinel Co3O4 for VOC
elimination, MHS MnxCo3−xO4 solid solutions fabricated by a
template-free strategy with the aim of catalyzing VOC oxida-
tion have not yet been reported so far.
Besides morphological engineering, the intrinsic activity
of catalysts is also considered as a decisive factor regarding
the catalytic efficiency of VOC elimination.20–26 In recent de-
cades, two major kinds of representative catalysts, mainly
including noble-metal-based catalysts (e.g. Pt, Pd, Rh, Au and
Ag) and transition-metal-based catalysts (e.g. V, Cr, Mn, Co,
Ni and Cu), have been extensively studied for VOC
elimination.1,6,20–26 It is generally admitted that noble-metal-
based catalysts are catalytically more active at low tempera-
ture as compared with metal oxides,20,21 but their high cost
and rapid deactivation significantly limit their large-scale ap-
plications. Accordingly, great efforts have been devoted to de-
veloping highly-efficient catalysts containing transition-metal
oxides.22–26 Among various transition-metal oxides, spinel
Co3O4 is usually considered to be one of the most effective
candidates for VOC elimination, and has received extensive
attention for its widespread application in catalytic oxidation
of formaldehyde, 1,2-dichloroethane, benzene, acetone, meth-
ane, carbon monoxide and soot particles because of its high
intrinsic activity at low temperature.22–31 Intrinsically, in
spinel Co3O4, usually expressed in the formula CoCo2O4, it is
typical that the Co2+ and Co3+ cations occupy the tetrahedral
and octahedral sites,29,30 respectively. Recent studies prove
that the intrinsic activity of spinel Co3O4 can be further im-
proved by incorporating other transition-metal cations (e.g.
Zn2+, Cu2+, Ni2+ and Mn2+) into its lattice to form solid
solutions.9,32–36 For instance, Gao et al. synthesized ordered
mesoporous MCo2O4 (M: Cu, Mn, Ni) spinel catalysts by
using SBA-15 as the hard template and experimental results
indicated that MnCo2O4 showed the highest CO oxidation ac-
tivity owing to its higher surface area,32 while Shi et al. also
prepared mesoporous MCo2O4IJ4.5) (M: Mn, Ni, Fe, Cu) spinel
catalysts by the pyrolysis of complex oxalates and discovered
Herein, the MHS MnxCo3−xO4 solid solutions with tunable
Mn/Co molar ratios in the spinel-type cubic crystal phase,
successfully prepared by a facile solvothermal alcoholysis ap-
proach, have been developed to catalyze acetone oxidation
for the first time. Results confirmed that MHS Mn1.20Co1.80O4
presented the optimal acetone oxidation activity, achieving
100% acetone conversion and 100% CO2 selectivity at 140 °C
under reaction conditions of acetone concentration = 1000
ppm, 20%O2/N2 and WHSV = ca. 93 000 mL gcat h−1. This
−1
superior actconversion curves (a) and correspondingvity of
MHS Mn1.20Co1.80O4 mainly resulted from the strong syner-
getic effect achieved by combining the unique mesoporous
hollow nanospherical feature, abundant oxygen vacancies
and surface active oxygen species, the enhanced chemical na-
ture stemming from the presence of more Mn4+, Mn3+ and
Co3+ reactive sites, and the improved redox capacities. Mean-
while, MHS Mn1.20Co1.80O4 also exhibited strong long-term
stability and good water tolerance due to its highly stable
crystal phase and robust morphological structure, showing
great promise in practical applications. Moreover, a plausible
catalytic mechanism of acetone oxidation has been proposed
and elaborated in detail by elucidating the structure–activity
relationship. We believe that the findings from this study can
deliver some new insights for guiding the rational design of
highly-efficient catalysts applied in the field of VOC
elimination.
2. Experimental section
2.1. Catalyst preparation
The chemical reagents employed in this study were of analyti-
cal grade and used as received without further treatment.
MHS MnxCo3−xO4 solid solutions with tunable Mn/Co molar
6380 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
ratios in the spinel-type cubic crystal phase were fabricated
by a facile solvothermal alcoholysis strategy with minor modi-
fications according to previous studies.9,37 Typically, the
precursors CoIJNO3)2·6H2O and MnIJNO3)2·4H2O with a total
amount fixed at 0.375 mmol were added to a mixture solu-
tion containing 40 mL isopropanol and 8.0 mL glycerol under
magnetic stirring until a transparent solution formed, which
was then poured into a 100 mL autoclave and maintained at
180 °C for 6.0 h. Afterwards, the precipitates were filtered,
thoroughly washed with absolute ethanol and distilled water
several times, then dried at 90 °C for 24 h in a vacuum oven,
and finally calcined at 500 °C for 4.0 h in an air atmosphere.
The products were labeled as MnxCo3−xO4, in which x was de-
termined by the different molar ratios of MnIJNO3)2·4H2O and
CoIJNO3)2·6H2O adopted in the preparation process. For com-
parison, crushed Mn1.20Co1.80O4 was acquired by grinding
intact Mn1.20Co1.80O4 for 20 min in an agate mortar by hands.
2.3. Activity evaluation
Activity measurements for acetone oxidation were performed
in a continuous flow fixed-bed tubular quartz reactor (ϕ = 6.0
mm) by temperature-programmed oxidation (TPO) using 100
mg catalyst (40–60 mesh) at atmospheric pressure as shown
in Fig. S1.† Each TPO reaction run from 50 °C to 230 °C with
a K-thermocouple placed inside the catalyst bed to ensure
that the temperature increases at a ramp of 2.0 °C min−1
.
Feed gases (e.g. 1000 ppm acetone, 20% O2 and N2 as the bal-
ance gas) flowed through the catalyst bed with a total stream
of 155 mL min−1, i.e., a weight hourly space velocity (WHSV)
of 93 000 mL gcat h−1. For studying the effect of WHSV on
−1
catalytic activity, WHSVs of about 120 000 mL gcat h−1 and
−1
150 000 mL gcat h−1 were acquired by changing the total
−1
streams of feed gases to 200 mL min−1 and 250 mL min−1, re-
spectively. For exploring the effect of water on catalytic activ-
ity, 8.0 vol% and 16 vol% H2O vapor were separately intro-
duced into the reaction system together with feed gases,
which were produced via pumping distilled water into a gasi-
fication chamber at 180 °C. The concentrations of effluent
gas products were analyzed using an online Shimadzu GC-
2014C apparatus equipped with a TCD and FID. The catalytic
performances of all the MnxCo3−xO4 catalysts were evaluated
by using T10, T50 and T90 representing the reaction tempera-
tures required to achieve acetone conversion (ηacetone) of
10%, 50% and 90%, respectively, where ηacetone was calcu-
lated by using the equation ηacetone = [(Cacetone-in − Cacetone-out)/
Cacetone-in] × 100%.
2.2. Catalyst characterization
The morphological features captured by high-resonance
transmission electron microscopy (HRTEM) and field emis-
sion scanning electron microscopy (FESEM) were taken using
a JEOL JEM-2100 and Hitachi-S4800 with accelerating volt-
ages of 200 kV and 5.0 kV, respectively. N2 sorption isotherms
measured on a Micromeritics ASAP 2460 at a temperature of
−196 °C were employed to calculate the pore volume (VP),
pore diameter (DP) and specific surface area (SBET) by using
the classic Barrett–Joyner–Halenda (BJH) and Brunauer–
Emmett–Teller (BET) models from the desorption branches.
Electron paramagnetic resonance (EPR) analysis was
conducted using a JEOL JES-FA200 at a temperature of −196
°C. Raman spectra were recorded on a HORIBA LabRAM HR
Evolution using an excitation wavelength of 532 nm at room
temperature. X-ray diffraction (XRD) patterns were determined
on a BRUKER D2 PHASER with CuKα radiation in the 2θ value
range of 15–70° operating at 40 kV and 40 mA. X-ray photo-
electron spectroscopy (XPS) was carried out on a PerkinElmer
PHI-5000C with AlKα radiation and all the binding energies of
Mn, Co and O elements were charge-corrected with reference
to C1s = 284.6 eV from contaminant carbon. The oxygen
temperature-programmed desorption (O2-TPD) experiment was
accomplished using a Micromeritics Auto-chemisorbed II 2920
with a TCD to quantify the desorption amount of oxygen spe-
cies in the 50–500 °C temperature range. 50 mg of catalyst was
placed into a U-shaped quartz tube; it was initially pretreated
at 350 °C for 1.0 h in a 10% O2/He stream (30 mL min−1) and
subsequently cooled to 50 °C, which was finally raised up to
600 °C with a 10 °C min−1 ramp in a pure He atmosphere. The
hydrogen temperature-programmed reduction (H2-TPR) test
was achieved in the same O2-TPD apparatus to quantify the
amount of H2 consumption in the 50–550 °C temperature
range; 50 mg of catalyst was firstly purged at 300 °C for 1.0 h
in a highly pure Ar atmosphere, then cooled to 50 °C, and fi-
nally heated up to 600 °C with a 10 °C min−1 ramp in a 30 mL
min−1 stream of 10% H2/Ar.
3. Results and discussion
3.1. Physicochemical characteristics
The FESEM images in Fig. 1 show that all the MnxCo3−xO4
catalysts presented uniform monodispersed nanospherical
features with an average diameter of about 300 nm except for
Mn1.50Co1.50O4 with a larger diameter of approximately 360
nm. Clearly, the nanosphere surface was roughly textured
and constructed from a great number of closely packed nano-
particle units, and abundant mesopores existed on the nano-
sphere surface due to the aggregated voids by the accumula-
tion of the closely packed nanoparticle units. Meanwhile, the
Fig. 1 FESEM images of different catalysts.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6381

View Article Online
Paper
Catalysis Science & Technology
corresponding TEM images in Fig. 2 demonstrate the consis-
tent morphological feature and nanospherical diameter dis-
tribution of all the MnxCo3−xO4 catalysts as characterized by
FESEM images. Notably, the sharp contrast between the po-
rous shell and the interior void in their TEM images reveal
that the structure of all the MnxCo3−xO4 catalysts was well-
defined hollow nanospheres. The average size of the hollow-
interior voids and the average thickness of the porous shells
for all the MnxCo3−xO4 catalysts were calculated to be approxi-
mately 200 nm and 50 nm, respectively, except for
Mn1.50Co1.50O4 with a larger hollow-interior void of about 260
nm. Moreover, the distinct reflection of the d-spacing value
of ca. 0.24 nm in the HRTEM images of all the MnxCo3−xO4
catalysts was assigned to the principal (311) plane of spinel-
type cubic metal oxides,9,29–31 which not only confirmed that
all the MnxCo3−xO4 catalysts were present in the spinel-type
cubic crystal phase, but also revealed that their porous shells
were constructed from the well-crystallized nanoparticle
units. Furthermore, the attached SAED images in their
HRTEM images indicated that all the MnxCo3−xO4 catalysts
possessed polycrystalline structures as evidenced by their
bright diffraction rings, suggesting that all the MnxCo3−xO4
catalysts had a high crystallization degree.
Fig.
3
XRD patterns (a) and the magnified 35–38° range (b) of
different catalysts.
planes of spinel-type face-centered cubic metal oxides,30,38 re-
spectively. Meanwhile, it can be seen that the characteristic
diffraction peaks of all the MnxCo3−xO4 catalysts apparently
weakened and broadened with increasing Mn/Co molar ratio,
suggesting the gradual decrease in the crystallization degree.
Based on the principal (311) peak of all the MnxCo3−xO4 cata-
lysts, the crystallite size has been calculated by using
Scherrer's equation. As listed in Table 1, it is obvious that the
crystallite size of all the MnxCo3−xO4 catalysts decreased grad-
ually from 18.5 nm to 9.50 nm with the gradual increase in
the Mn/Co molar ratio, indicating the gradual decrease in the
crystallization degree, which can be mainly ascribed to the in-
corporation of Mn cations into the Co3O4 lattice inhibiting
the crystal growth of Co3O4. This speculation can be
evidenced by the magnified principal (311) peak as shown in
Fig. 3b. Clearly, it can be noted that the principal (311) peak
position of all the MnxCo3−xO4 catalysts slightly shifted to
lower angles with increasing Mn/Co molar ratio as compared
with pure Co3O4 owing to the bigger ionic radii of Mn2+
(0.080 nm) and Mn3+ (0.072 nm) than those of Co2+ (0.072
nm) and Co3+ (0.069 nm),38–40 which revealed that the incor-
poration of Mn cations into the Co3O4 lattice formed
MnxCo3−xO4 solid solutions rather than composite oxides like
as MnOx/Co3O4. Moreover, the incorporation of Mn cations
into the Co3O4 lattice will inevitably result in lattice distor-
tion and formation of defective sites in all the MnxCo3−xO4
catalysts due to the differences in their ionic and electronic
states.
The XRD patterns in Fig. 3a show that all the MnxCo3−xO4
catalysts had the typical spinel-type face-centered cubic crys-
tal phase with the space group Fd3m (JCPDS Card No. 43-
1003),9,31–33 and no diffraction peaks assigned to other impu-
rities such as CoO, Co2O3 and MnOx were found, signifying
the high crystal phase purity. The characteristic diffraction
peaks of all the MnxCo3−xO4 catalysts at 2θ values of 18.9°,
31.2°, 36.8°, 44.8°, 55.6°, 59.3° and 65.2° can be well indexed
to the (111), (220), (311), (400), (422), (511) and (440) lattice
Table 1 Physical structural parameters of different catalysts
SBET
VP
DP
(nm)
Crystallite size
(nm)
Catalyst
(m2 g−1
)
(cm3 g−1
)
Co3O4
18
41
51
65
65
41
0.11
0.12
0.14
0.12
0.14
0.092
23
18.5
16.3
15.9
11.2
10.8
9.50
Mn0.32Co2.68O4
Mn0.64Co2.36O4
Mn1.00Co2.00O4
Mn1.20Co1.80O4
Mn1.50Co1.50O4
8.3
7.3
5.1
6.2
6.3
Fig. 2 TEM, HRTEM and SAED (inset) images of different catalysts.
6382 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
To identify the formation of defective sites in all the
MnxCo3−xO4 catalysts, the Raman technique has been applied
to characterize all the catalysts as shown in Fig. 4a. Notably,
it can be found that pure Co3O4 possessed five major peaks
centered at around 181 cm−1, 467 cm−1, 514 cm−1, 602 cm−1
and 667 cm−1 corresponding to the Raman-active variation
The N2 adsorption–desorption isotherms in Fig. 5a reveal
that all the MnxCo3−xO4 catalysts exhibited typical type-IV N2
sorption isotherms coupled with an H3 hysteresis loop in the
relative pressure (P/P0) range of 0.50–1.0, signifying the pres-
ence of typical mesoporous structures according to the classic
IUPAC definitions.9,31,47 Clearly, no saturated adsorption of
all the MnxCo3−xO4 catalysts can be achieved until P/P0 = 1.0,
implying the presence of some larger mesoporous struc-
tures.31,47,48 This speculation can be verified by their respec-
tive pore size distribution curve (see Fig. 5b) in which a rela-
tively wide pore size distribution can be found. Obviously,
the mesopores centered at around 3.8 nm were mainly lo-
cated at the porous shells due to the aggregated voids by the
accumulation of the closely packed nanoparticle units, and
the larger mesopores above 3.8 nm were associated with the
hollow-interior voids and interparticle voids formed by the
packing of the hollow nanospheres. Based on the N2 adsorp-
tion–desorption isotherms, the specific surface area (SBET),
pore volume (VP) and pore size (DP) of all the MnxCo3−xO4 cat-
alysts have been acquired by using the classic BET and BJH
models calculated from the desorption branches, respec-
tively. From Table 1, one can see that the SBET of all the
MnxCo3−xO4 catalysts increased remarkably from pure Co3O4
to Mn1.20Co1.80O4, showing an inverse trend to the crystallite
size as shown by XRD analysis. This finding can be attributed
to the fact that the incorporation of Mn cations into the
Co3O4 lattice inhibited the crystal growth of Co3O4, which in
turn facilitated the increase in SBET. However, Mn1.50Co1.50O4
gave a lower SBET than Mn1.20Co1.80O4 although it had the
smallest crystallite size, mainly owing to the larger average
size of its nanosphere diameter and hollow-interior void.
The surface chemical compositions and chemical valence
states of all the MnxCo3−xO4 catalysts have been studied by
using the XPS technique as shown in Fig. 6. The survey spec-
tra in Fig. S2† reveal that all the MnxCo3−xO4 catalysts
contained only Mn, Co and O elements and no other
1
2
3
modes F2g, Eg, F2g, F2g and A1g for pure spinel-type cubic
Co3O4,9,41,42 respectively. Compared with pure Co3O4, it can
be observed that the peaks of all the MnxCo3−xO4 catalysts
presented a slight shift to lower wavenumbers and became
broader and weaker with the increase in the Mn/Co molar ra-
tio, signifying the increased number of defective sites.27,30
This appearance can be associated with the fact that the in-
corporation of Mn ions into the Co3O4 lattice resulted in lat-
tice distortion and expansion owing to the larger size of Mn
ions with respect to the corresponding Co ions, which corre-
spondingly caused the formation of defective sites and
inhibited the grain growth rate,24 agreeing well with the XRD
results. Moreover, the lattice distortion and expansion could
lead to the surface atoms being bound by weaker forces due
to a lack of adjacent atoms, which might induce the forma-
tion of oxygen vacancies. To confirm the formation of oxygen
vacancies in all the MnxCo3−xO4 catalysts, the EPR technique
has been employed to characterize all the MnxCo3−xO4 cata-
lysts. Clearly, the results of EPR spectroscopy in Fig. 4b gave
direct evidence to verify the formation of oxygen vacancies in
all the MnxCo3−xO4 catalysts. Compared with pure Co3O4, all
the MnxCo3−xO4 catalysts exhibited a much higher EPR signal
peak centered at around g = 1.995, which was mainly linked
to the Zeeman effect of the unpaired electrons trapped at the
sites of oxygen vacancies,43,44 distinctly validating the forma-
tion of oxygen vacancies. According to previous studies,45,46
the number or concentration of oxygen vacancies can be pro-
portional to the peak intensity of the EPR signal. Obviously,
it can be noted that the peak intensity of EPR signals in-
creased gradually from pure Co3O4 to Mn1.50Co1.50O4, demon-
strating the increased number or concentration of oxygen
vacancies.
Fig.
5 N2 adsorption–desorption isotherms (a) and pore size
Fig. 4 Raman spectra (a) and EPR spectra (b) of different catalysts.
distribution curves (b) of different catalysts.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6383

View Article Online
Paper
Catalysis Science & Technology
oxygen vacancies can be reflected by the peak intensity of sur-
face chemisorbed oxygen (Oads) species. Clearly, the peak
intensity of Oads in Fig. 6c increased slightly from pure Co3O4
to Mn1.50Co1.50O4, signifying the increased number of oxygen
vacancies. Based on the XPS analysis, the quantified surface
chemical compositions and chemical valence states of all the
MnxCo3−xO4 catalysts have been calculated. As listed in
Table 2, one can see that the molar ratio of (Mn4+ + Mn3+)/
Mn2+ for all the MnxCo3−xO4 catalysts decreased gradually
from Mn0.32Co2.68O4 to Mn1.50Co1.50O4 although the Mn spe-
cies with higher oxidation states were more conductive to oxi-
dation reactions over a Mn-based catalyst,24 mainly due to
the fact that the increase in the amount of incorporated Mn
species into the Co3O4 lattice extremely exceeded the
electrons transferring from Mn to Co. However, the molar ra-
tio of Co3+/Co2+ increased gradually from pure Co3O4 to
Mn1.50Co1.50O4, suggesting the increased number of Co3+ cat-
ions. Moreover, the molar ratio of Oads/Olatt also increased
gradually from Co3O4 to Mn1.50Co1.50O4 corresponding to the
gradual increase in the number of oxygen vacancies, which
matched well with the results revealed by EPR analysis. This
finding was attributed to the incorporation of Mn cations
into the Co3O4 lattice resulting in lattice expansion/or distor-
tion and producing more defective sites, thereby leading to
increased formation of oxygen vacancies.
Fig. 6 XPS spectra of different catalysts: Mn2p (a), Co2p (b) and O1s (c).
impurities were observed, suggesting their high purity. The
high-resolution spectrum of the Mn2p region in Fig. 6a
shows two spin–orbit doublets with binding energies at about
642.3 eV and 653.8 eV corresponding to Mn2p1/2 and
24
Mn2p3/2
,
respectively. Meanwhile, it can be seen that the
binding energies of Mn2p1/2 and Mn2p3/2 slightly shifted to
higher binding energies with increasing Mn/Co molar ratio,
indicating the strong interaction between Mn and Co species,
which was attributed to the fact that Co has a much stronger
electrophilicity (1.88) than Mn (1.55) and resulted in partial
electron transfer from Mn to Co. The signals of Mn4+, Mn3+
and Mn2+ cations were found after the deconvolution of the
Mn2p spectra by using Gaussian fitting methods. Two dis-
tinct peaks with binding energies at around 643.9 eV and
655.3 eV were associated with typical Mn4+ cations, and two
major peaks with binding energies at approximately 642.6 eV
and 653.7 eV were linked to characteristic Mn3+ cations,
while other two small peaks with binding energies at about
641.5 eV and 652.5 eV were assigned to Mn2+ cations.49 The
high-resolution spectrum of the Co2p region in Fig. 6b is
composed of two obvious peaks with binding energies at
The reducibility of all the MnxCo3−xO4 catalysts has been
evaluated by H2-TPR experiments as depicted in Fig. 7.
Clearly, pure Co3O4 exhibited only two distinct reduction
peaks below 500 °C, signifying that the reduction of pure
Co3O4 included two stepwise processes Co3+ → Co2+ and Co2+
→ Co0. The first reduction peak located at around 287 °C
corresponded to the reduction of Co3+ → Co2+ while the re-
duction peak centered at about 357 °C represented the reduc-
tion of Co2+ → Co0.9,30,31 However, the reduction profiles of
all the MnxCo3−xO4 catalysts can be resolved into three reduc-
tion peaks below 500 °C. The first reduction peak at approxi-
mately 201 °C was assigned to the reduction of Mn4+
→
Mn2O3, the second reduction peak located at around 278 °C
was associated with the simultaneous reductions of Mn2O3 →
Mn3O4 and Co3+ → Co2+, and the third reduction peak cen-
tered at about 392 °C belonged to the simultaneous reduc-
tions of Mn3O4 → MnO and Co2+ → Co0.22,52 Notably, the re-
duction peak of Co3+ → Co2+ shifted to lower temperature
from 287 °C to 278 °C for all the MnxCo3−xO4 catalysts as
compared with pure Co3O4, implying that the oxygen mobility
was dramatically improved owing to the strong synergy be-
tween Mn and Co species. According to a previous report,52
the strong synergy can cause severe deformation of the struc-
ture which in turn generates abundant oxygen vacancies and
thus, promotes oxygen diffusion on the surface of catalysts
which penetrates into the bulk, which undoubtedly will facili-
tate the improved oxidation reaction activity. The amount of
H2 consumption below 550 °C has been determined by quan-
titatively calculating the reduction peaks in H2-TPR profiles
of all the MnxCo3−xO4 catalysts. From Table 3, it can be no-
ticed that the amount of H2 consumption increased linearly
around 795.3 eV and 780.1 eV corresponding to Co2p1/2 and
9,30,31
Co2p3/2
,
respectively, along with a small shake-up satel-
lite peak centered at approximately 790.1 eV, implying the ex-
istence of two cobalt species.9 The signals of Co2+ and Co3+
cations can be observed after the decomposition of the Co2p
spectra. The two major peaks with binding energies at about
780.4 eV and 795.8 eV were indexed to characteristic Co3+ cat-
ions, while the other two small peaks with binding energies
at approximately 781.4 eV and 797.6 eV represented typical
Co2+ cations.9,30,31 The existence of Mn2+ and Co2+ cations is
usually connected to the formation of oxygen vacancies. To
elucidate the origin of oxygen vacancies, the high-resolution
O1s spectrum of all the MnxCo3−xO4 catalysts in Fig. 6c has
been fitted into two major peaks at binding energies of 530.1
eV and 531.8 eV, corresponding to the lattice oxygen (Olatt
)
species and surface chemisorbed oxygen (Oads) species (e.g.
O22−, O2 , OH− or O−),30,31,50 respectively. According to previ-
−
ously reported studies,48,51 the number or concentration of
6384 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Table 2 Surface elemental compositions and chemical states of different catalysts
Mn2p envelope
Mn4+ + Mn3+ (%)
Co2p envelope
O1s envelope
Oads (%)
Catalyst
Mn2+ (%)
(Mn4+ + Mn3+)/Mn2+
Co3+ (%)
Co2+ (%)
Co3+/Co2+
Olatt (%)
77.6
Oads/Olatt
Co3O4
—
—
—
46.9
57.2
58.4
61.5
61.8
62.2
53.1
42.8
41.6
38.5
38.2
37.8
0.88
1.33
1.41
1.60
1.62
1.65
22.4
26.3
29.6
31.8
34.0
35.2
0.29
0.36
0.42
0.47
0.52
0.54
Mn0.32Co2.68O4
Mn0.64Co2.36O4
Mn1.00Co2.00O4
Mn1.20Co1.80O4
Mn1.50Co1.50O4
87.9
86.1
85.5
85.4
85.4
12.1
13.9
14.5
14.6
14.6
7.26
6.19
5.90
5.85
5.85
73.7
70.3
68.2
66.0
64.8
from pure Co3O4 to Mn1.20Co1.80O4, indicating the great in-
crease in reducibility. This fact can be mainly ascribed to the
following factors. On the one hand, the increase in Co3+ cat-
ions from pure Co3O4 to Mn1.20Co1.80O4 played a crucial role
in increasing the amount of H2 consumption owing to the re-
duction of Co3+ → Co0 requiring much more H2 than the re-
duction of Co2+ → Co0, regardless of the decrease in the
Mn4+ and Mn3+ cations.9,30 On the other hand, the increase
in oxygen vacancies from pure Co3O4 to Mn1.20Co1.80O4 fa-
vored the enhanced adsorption of active oxygen species and
also increased the amount of H2 consumption, agreeing well
with the results of Raman, EPR and XPS analysis. However,
Mn1.50Co1.50O4 showed a lower amount of H2 consumption
than Mn1.20Co1.80O4 although it contained more Co3+ cations
and oxygen vacancies, possibly due to its lower surface area
which decreases the contact between H2 and the catalyst.
The adsorption and activation behaviors of all the
MnxCo3−xO4 catalysts towards different oxygen species have
been explored by O2-TPD experiments as shown in Fig. 8.
Generally, the surface oxygen species in redox reactions often
mately 88 °C, 153 °C and 408 °C, corresponding to the re-
lease of surface physically adsorbed oxygen species (O2), sur-
−
face chemisorbed active oxygen species (O2 ) and surface
lattice oxygen species (O2−),31,47,53 respectively. However, only
two desorption peaks for all the MnxCo3−xO4 catalysts can be
found. The first desorption peak below 200 °C can be linked
to the simultaneous liberations of surface physically
adsorbed oxygen species (O2) and surface chemisorbed active
−
oxygen species (O2 ), while the second desorption peak at
around 408 °C was attributed to the removal of the surface
lattice oxygen species (O2−).31,47,53 Accordingly, the desorption
amount of surface active oxygen species below 500 °C has
been obtained by quantitatively calculating the desorption
peaks in O2-TPD profiles of all the MnxCo3−xO4 catalysts. As
listed in Table 3, it can be clearly seen that the desorption
amount of surface active oxygen species increased gradually
from pure Co3O4 to Mn1.20CO1.80O4, suggesting that
Mn1.20CO1.80O4 had the highest amount of surface active oxy-
gen species, which was in good agreement with the variation
in the Oads/Olatt molar ratio as revealed by XPS analysis. This
result can be ascribed to their different amounts of oxygen
vacancies as evidenced by Raman and EPR characterizations.
However, Mn1.50Co1.50O4 gave a lower amount of surface ac-
tive oxygen species than Mn1.20Co1.80O4 although it contained
more oxygen vacancies, which might be mainly due to its
lower surface area hampering the adsorption and activation
of surface oxygen species.
−
undergo the following pathways: O2IJads) → O2 (ads) →
O−(ads) → O2−(latt), and the former three oxygen species are
usually recognized as the active oxygen species in redox reac-
tions.31,47 As shown in Fig. 8, it was notable that pure Co3O4
demonstrated three desorption peaks located at approxi-
3.2. Catalytic performance
The catalytic performances of all the MnxCo3−xO4 catalysts
have been evaluated by using acetone oxidation as a model
reaction at atmospheric pressure in the temperature range
Table 3 Chemical structural parameters and special activities of different
catalysts
Consumption
(mmol g−1
Special activity
(°C)
)
RS at 140 °C
Catalyst
H2
O2
T10
T50
T90
(mol g−1 s−1
)
Co3O4
9.12
0.062
0.110
0.166
0.173
0.185
0.119
141
119
115
110
101
103
163
139
135
131
122
124
184
159
148
139
130
136
0.97 × 10−7
5.92 × 10−7
7.23 × 10−7
10.2 × 10−7
11.1 × 10−7
10.6 × 10−7
Mn0.32Co2.68O4
Mn0.64Co2.36O4
Mn1.00Co2.00O4
Mn1.20Co1.80O4
Mn1.50Co1.50O4
10.4
10.9
11.2
11.9
11.1
Fig. 7 H2-TPR profiles of different catalysts.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6385

View Article Online
Paper
Catalysis Science & Technology
produced CO2 concentrations over all the MnxCo3−xO4 cata-
lysts also increased gradually in the temperature range from
50 °C to 230 °C regardless of their different degrees. Combin-
ing the normalization of acetone conversion and produced
CO2 concentration, the molar ratio of CO2/acetone was calcu-
lated to be about 2.98, which was close to the theoretical
value of 3.0, indicating that acetone was completely oxidized
into CO2 and H2O under the present conditions. Meanwhile,
the reaction temperatures T10, T50 and T90 corresponding to
achieved acetone conversions of 10%, 50% and 90%, respec-
tively, were selected and used as the evaluation criteria for ac-
etone oxidation. From Table 3, it can be obviously seen that
the T10, T50 and T90 of all the MnxCo3−xO4 catalysts, varying
in the same order, firstly decreased gradually from pure
Co3O4 to Mn1.20Co1.80O4 and then increased slightly from
Mn1.20Co1.80O4 to Mn1.50Co1.50O4. Hence, Mn1.20Co1.80O4 was
undeniably considered as the optimal catalyst for acetone oxi-
dation in this study. In addition, the acetone conversion rate
of all the MnxCo3−xO4 catalysts at 140 °C has been acquired
by using the acetone conversion normalized to per unit gram
of catalyst per unit second of time according to the activity
data in Fig. 9a. As listed in Table 3, one can see that the ace-
tone conversion rate initially increased remarkably from 0.97
× 10−7 mol g−1 s−1 to 11.1 × 10−7 mol g−1 s−1 corresponding to
pure Co3O4 to Mn1.20Co1.80O4, and then declined slightly from
11.1 × 10−7 mol g−1 s−1 to 10.6 × 10−7 mol g−1 s−1 correspond-
ing to Mn1.20Co1.80O4 to Mn1.50Co1.50O4, showing that
Mn1.20Co1.80O4 possessed the highest acetone conversion rate
under the present conditions and its acetone conversion rate
was about 11.4 times higher than that of pure Co3O4. There-
fore, the above results confirmed that Mn1.20Co1.80O4 acted as
the optimal catalyst for acetone oxidation in this study. More-
over, from Table S1,† it should be noted that the catalytic ac-
tivity of Mn1.20Co1.80O4 was also much higher than many pre-
viously reported catalysts for acetone oxidation, e.g. noble
metal-based, cobalt-based, iron-based, copper-based, ceria-
based, titanium-based and vanadium-based catalysts together
with some multi-component metal oxides, further confirming
the excellent acetone oxidation activity of Mn1.20Co1.80O4.
Importantly, it must be pointed out that the activity data
in Fig. 9 were obtained by conducting acetone oxidation over
all the MnxCo3−xO4 catalysts based on the same catalyst mass
basis rather than simultaneously on the same surface area
basis. Thus, the effect of the surface area on the acetone oxi-
dation activity must be considered owing to the fact that ace-
tone oxidation as a typical gas–solid heterogeneous reaction
generally takes place at the binary-phase contact boundary or
points.7,9 Accordingly, the effect of the surface area on the ac-
etone oxidation activity has been explored and excluded by
performing kinetics studies. According to previous reports,7,9
it has been accepted that acetone oxidation in the presence
of enough oxygen gas follows the first-order reaction mecha-
nism related to the acetone concentration with the following
equation: r = −kc = [−A expIJ−Ea/RT)]c, in which r, k, c, Ea and A
refer to the reaction rate (μmol m−2 s−1), rate constant (s−1),
acetone concentration (mol), apparent activation energy (kJ
Fig. 8 O2-TPD profiles of different catalysts.
from 50 °C to 230 °C under the reaction conditions of ace-
tone concentration = 1000 ppm, 20%O2/N2 and WHSV = ca.
−1
93 000 mL gcat h−1. It should be pointed out that no appre-
ciable acetone oxidation was found to happen without cata-
lysts in the examined temperature range, and only CO2 and
H2O as the final products could be detected over all the
MnxCo3−xO4 catalysts under the experimental conditions
used. From Fig. 9a, it can be clearly observed that the acetone
conversion over all the MnxCo3−xO4 catalysts increased gradu-
ally with the rise in the reaction temperature, although they
had significantly different degrees. Complete acetone conver-
sion can be achieved at reaction temperatures of 230 °C, 180
°C, 162 °C, 153 °C, 140 °C and 150 °C for pure Co3O4,
Mn0.32Co2.68O4, Mn0.64Co2.36O4, Mn1.00Co2.00O4, Mn1.20Co1.80O4
and Mn1.50Co1.50O4, respectively. The produced CO2 concen-
trations over all the MnxCo3−xO4 catalysts have been also
monitored and quantified as shown in Fig. 9b. Notably, the
Fig.
9 Acetone conversion curves (a) and the produced CO2
concentration curves (b) of different catalysts.
6386 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
mol−1) and pre-exponential factor, respectively. Hence, the ap-
parent activation energy (Ea) values of all the MnxCo3−xO4 cat-
alysts have been calculated using Arrhenius-type plots
according to the activity data in Fig. 9a under lower acetone
conversion below 15% in the kinetic regime due to the fact
that the effect of feed compositions in the inlet and outlet of
the entire catalyst bed and the thermal effect can be
neglected under such conditions.9,30,54–56 As shown in
Fig. 10, it can be seen that the Ea values of all the
MnxCo3−xO4 catalysts decreased gradually from 97.4 kJ mol−1
to 63.1 kJ mol−1 corresponding to pure Co3O4 to
Mn1.20Co1.80O4, and then increased slightly from 63.1 kJ
mol−1 to 66.1 kJ mol−1 corresponding to Mn1.20Co1.80O4 to
Mn1.50Co1.50O4, showing an inverse trend to the acetone oxi-
dation activity order and Mn1.20Co1.80O4 with the lowest Ea
value. To the best of our knowledge, the lower the Ea value is,
the easier it is to completely oxidize acetone.9,30 Thus, the
variation in the Ea value demonstrated an inverse trend to
the acetone oxidation activity order. The above results further
confirmed the highest acetone oxidation activity of
Mn1.20Co1.80O4 and also excluded the effect of the surface
area although it played an important role in influencing the
acetone oxidation activity. Therefore, this superior activity of
Mn1.20Co1.80O4 can be associated with other influencing fac-
tors, e.g. the unique mesoporous hollow nanospherical fea-
ture, abundant oxygen vacancies and surface active oxygen
species, the enhanced chemical nature resulting from the
presence of more Mn4+, Mn3+ and Co3+ cations and the im-
proved redox properties. Firstly, the unique mesoporous hol-
low nanospherical structure had a large hollow-interior void,
which not only benefitted the enhanced adsorption of ace-
tone molecules but also increased the contact frequency of
the acetone molecule–catalyst when acetone molecules dif-
fused in and out of the hollow-interior void. Meanwhile, the
mesoporous channels within the porous shells facilitated the
decrease in mass transfer blockage which in turn promoted
the enhanced adsorption and diffusion of acetone molecules
and thus, improved the acetone oxidation activity. This inter-
pretation can be verified by conducting acetone oxidation
over crushed Mn1.20Co1.80O4 (see Fig. S3†) due to the fact that
the crushed Mn1.20Co1.80O4 gave lower acetone oxidation ac-
tivity than Mn1.20Co1.80O4 (see Fig. S4†). Secondly,
Mn1.20Co1.80O4 contained relatively more oxygen vacancies as
evidenced by Raman, EPR and XPS characterizations, which
favored the improved adsorption and activation of oxygen
species and thus, resulted in the formation of more surface
active oxygen species to participate in acetone oxidation,
thereby enhancing the acetone oxidation activity. However,
Mn1.50Co1.50O4 had the highest amount of oxygen vacancies
and surface active oxygen species while its acetone oxidation
activity was lower than that of Mn1.20Co1.80O4, possibly due to
its lower surface area hindering the adsorption of acetone
molecules and their contact with catalysts. Thirdly,
Mn1.20Co1.80O4 possessed abundant Mn4+, Mn3+ and Co3+ cat-
ions as disclosed by XPS analysis. Generally, the more the
high-valence metal cations, the higher the acetone oxidation
activity.9,24 Meanwhile, both theoretical calculations and ex-
perimental results have proved that high-valence metal cat-
ions, especially Mn4+ and Co3+ cations, are usually considered
as the reactive sites for many oxidation reactions, e.g. VOC
oxidation, CO oxidation and soot oxidation, but low-valence
metal cations like Co2+ and Mn2+ are nearly inactive.7,9,29–31
Finally, the highly improved redox properties of
Mn1.20Co1.80O4 also promoted acetone oxidation, mainly as-
cribed to the highly improved redox properties favoring the
adsorption and activation of oxygen species to produce sur-
face active oxygen species to participate in acetone oxidation
and thus, improving the acetone oxidation activity. All the
above factors simultaneously contributed to Mn1.20Co1.80O4
having the optimal acetone oxidation activity in this study.
The effect of WHSV on the acetone oxidation activity over
Mn1.20Co1.80O4 has been studied as shown in Fig. 11. Notably,
the acetone conversion over Mn1.20Co1.80O4 decreased gradu-
ally and a higher reaction temperature was required for com-
plete acetone conversion with increasing WHSV value from
93 000 mL gcat h−1 to 160 000 mL gcat h−1. In other words,
the increase in the WHSV value led to a lower acetone conver-
sion at the same reaction temperature, which can be further
verified by the evidence that the reaction temperatures T10,
T50 and T90 for acetone conversion were elevated to higher
temperatures with increasing WHSV value, distinctly increas-
ing by about nearly 2.0 °C for T10, 13 °C for T50 and 30 °C for
−1
−1
−1
T90 at the WHSV value of 160 000 mL gcat h−1 as compared
with the WHSV value of 93 000 mL gcat h−1. This distinct ac-
−1
etone oxidation activity of Mn1.20Co1.80O4 related to the
WHSV value was mainly caused by the shorter retention time
of acetone molecules in the catalyst bed under higher WHSV
values,54–56 which significantly reduced the contact time be-
tween the acetone molecule and catalyst and thus, decreased
the acetone oxidation activity.
Catalyst deactivation is a major challenge that urgently re-
quires addressing for practical applications. Meanwhile, it's
Fig. 10 Arrhenius-type plots of different catalysts.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6387

View Article Online
Paper
Catalysis Science & Technology
vapor into the reaction system. The experimental results in
Fig. 12b demonstrated that the acetone conversion, at the re-
action temperature of 140 °C, apparently decreased gradually
from 100% to nearly 92% for 12 h of reaction when 8.0 vol%
H2O vapor was introduced, and then the acetone conversion
greatly decreased to approximately 82% when 16 vol% H2O
vapor was introduced. This fact obviously demonstrated that
H2O vapor had a detrimental effect on the acetone oxidation
activity of Mn1.20Co1.80O4, mainly due to the existence of com-
petitive adsorption between the acetone molecule and H2O
vapor on the catalyst's surface.9,54–56 Interestingly, it can be
noticed that the acetone conversion could immediately re-
cover back to almost 100% in several hours once the H2O
vapor was interrupted, indicating that Mn1.20Co1.80O4 had
good tolerance for H2O vapor at a certain content. This
excellent long-term stability and good water tolerance of
Mn1.20Co1.80O4 can be associated with its highly stable crystal
phase and robust morphological feature since the crystal
phase and morphology of the used Mn1.20Co1.80O4 remained
almost unchanged as evidenced by the XRD pattern and
FESEM image in Fig. S5.†
Considering the above results and discussion, a plausible
catalytic mechanism of acetone oxidation over MnxCo3−xO4
solid solutions has been tentatively proposed by combining
the structure–activity relationship and the classic Mars–Van
Krevelen (MVK) mechanism as depicted in Scheme 1. Gener-
ally, the MVK mechanism has been accepted and widely ap-
plied in the field of VOC oxidation over various metal oxide
catalysts.57–60 The classic MVK mechanism is observed in the
entire catalytic oxidation reaction between VOC molecules
and active oxygen species adsorbed on the catalyst's surface,
mainly including two stepwise stages as follows: the oxygen
gas coming from feed gases can be adsorbed and activated
by catalysts to form oxidized catalysts and surface active
oxygen species, subsequently the surface active oxygen spe-
cies react with VOC molecules adsorbed on the catalyst's sur-
face, and finally the oxidized catalysts can be reduced by
hydrocarbons.58–60 Informed by our present results, the plau-
sible catalytic mechanism of acetone oxidation over
MnxCo3−xO4 solid solutions can be briefly depicted as follows:
Fig. 11 Acetone conversion curves (a) and corresponding T10, T50 and
T90 (b) over Mn1.20Co1.80O4 at different WHSVs.
well-known that water is often generated during the VOC oxi-
dation process and also has a negative effect on the catalytic
activity.54–56 Hence, the long-term stability and water toler-
ance of catalysts for VOC oxidation are generally recognized
as important evaluation criteria for their potential in practi-
cal applications. Accordingly, the long-term stability of
Mn1.20Co1.80O4 for 48 h of testing was measured under the re-
action conditions of acetone concentration = 1000 ppm, 20%
O2/N2 and WHSV = ca. 93 000 mL gcat h−1 at 120 °C and 140
−1
°C corresponding to the low and high acetone conversions
for acetone oxidation, respectively. As shown in Fig. 12a, one
can clearly see that Mn1.20Co1.80O4 exhibited excellent long-
term stability at the two selected reaction temperatures, with
acetone conversions being almost unchanged and stable at
nearly 100% for 140 °C and about 36% for 120 °C in the en-
tire catalytic oxidation test. The effect of H2O vapor on the ac-
etone oxidation activity over Mn1.20Co1.80O4 was also investi-
gated by separately introducing 8.0 vol% and 16 vol% H2O
Fig. 12 Long-term stability test (a) and the effect of H2O vapor on the
Scheme 1 The plausible catalytic mechanism of acetone oxidation
over MnxCo3−xO4.
catalytic activity (b) of Mn1.20Co1.80O4 at WHSV = 93 000 mL gcat−1 h−1
.
6388 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
firstly, the acetone molecules (CH3COCH3) in feed gases dif-
fused onto the catalyst's surface, and then adsorbed onto the
reactive sites (e.g. Mn4+, Mn3+ and Co3+ cations) to produce
the chemisorbed acetone molecules (*CH3COCH3), and the
oxygen gas coming from feed gases adsorbed onto the oxygen
vacancies to generate the surface active oxygen species (Oact),
and finally the chemisorbed acetone molecules (*CH3COCH3)
strong long-term stability and good water tolerance owing to
its highly stable crystal phase and robust morphological
structure, showing great promise in practical applications.
We believe that the findings from this study can provide
some new insights for guiding the rational design of highly-
efficient catalysts applied in VOC elimination.
reacted with the surface active oxygen species (Oact) to Conflicts of interest
produce the final products CO2 and H2O. During the entire
There are no conflicts to declare.
catalytic oxidation process, the reactive sites Mn4+, Mn3+
and Co3+ cations acted as the chemisorption sites of acetone
Acknowledgements
molecules to form the chemisorbed acetone molecules (*
CH3COCH3) according to previous studies.9,29,49,60–62 For in-
This work is supported by the Shanghai Talent Development
stance, our recent study disclosed that the Co3+ cations
served as the adsorption and reactive sites for acetone oxida-
tion,9 Shen et al. revealed that the Co3+ cations served as the
adsorption and reactive sites in low-temperature CO oxida-
tion,29 and Hao et al. attested that the Co3+ cations func-
tioned as the adsorption and reactive sites for formaldehyde
oxidation.60 Zhan et al. reported that the Mn4+, Mn3+ and
Co3+ cations were actively engaged as the adsorption and re-
active sites for activating C–H in selective cyclohexane oxida-
tion,49 Zhang et al. found that the Mn4+ and Co3+ cations
acted as the adsorption and reactive sites for low-temperature
formaldehyde oxidation,61 and Zhao et al. reported that the
Mn4+ and Mn3+ cations were considered as the adsorption
sites for NH3 and –NH2 in low-temperature SCR of NOx with
NH3,62 respectively. Meanwhile, the oxygen vacancies served
as the adsorption and reactive sites of oxygen gas to produce
the surface active oxygen species (Oact). The improved redox
properties of catalysts could also promote the enhanced
adsorption and activation of oxygen gas to form the surface
active oxygen species (Oact). All the above factors simulta-
neously worked together to complete the entire acetone
oxidation reaction, which further confirmed that the highest
acetone oxidation activity of Mn1.20Co1.80O4 was attributed to
the strong synergetic effect achieved by combining abundant
oxygen vacancies and surface active oxygen species, the
enhanced chemical nature resulting from the presence of
more Co3+, Mn4+ and Mn3+ cations, and the improved redox
properties.
Foundation (Grant No. 2017076), National Natural Science
Foundation of China (Grant No. 21503133), Shanghai Key
Laboratory of Rare Earth Functional Materials (Grant No.
2018No.01), Shanghai Municipal Education Commission
(Grant No. ZZGCD15031), Shanghai Government (Grant No.
17SG44) and the Talent Program of Shanghai University of
Engineering Science.
Notes and references
1 C. He, J. Cheng, X. Zhang, M. Douthwaite, S. Pattisson and
Z. P. Hao, Chem. Rev., 2019, 119, 4471–4568.
2 Y. Y. Zhang, Z. N. Tan, X. Y. Wang, Y. Y. Zhan, Y. H. Xiao, C.
Au and L. L. Jiang, Nanoscale, 2019, 11, 4794–4802.
3 J. Chen, X. Chen, X. Chen, W. J. Xu, Z. Xu, H. P. Jia and J.
Chen, Appl. Catal., B, 2018, 224, 825–835.
4 F. Dong, W. L. Han, H. J. Zhao, G. D. Zhang and Z. C. Tang,
Nanoscale, 2019, 11, 9937–9948.
5 P. Yang, S. K. Fan, Z. Y. Chen, G. F. Bao, S. F. Zuo and C. Z.
Qi, Appl. Catal., B, 2018, 239, 114–124.
6 J. N. Li, X. Y. Li, L. B. Zeng, S. Y. Fan, M. M. Zhang, W. B.
Sun, X. Chen, M. O. Tadé and S. M. Liu, Nanoscale, 2019, 11,
3877–3887.
7 X. C. Zhu, S. Zhang, X. N. Yu, X. B. Zhu, C. H. Zheng, X.
Gao, Z. Y. Luo and K. F. Cen, J. Hazard. Mater., 2017, 337,
105–114.
8 Y. S. Xia, H. X. Dai, H. Y. Jiang, L. Zhang, J. G. Deng and
Y. X. Liu, J. Hazard. Mater., 2011, 186, 84–91.
9 C. Zhang, J. G. Wang, S. F. Yang, H. Liang and Y. Men,
J. Colloid Interface Sci., 2019, 539, 65–75.
10 Z. F. Bian, J. Zhu, F. L. Cao, Y. N. Huo, Y. F. Lu and H. X. Li,
Chem. Commun., 2010, 46, 8451–8453.
11 Z. M. Chen, J. G. Wang, G. J. Zhai, W. An and Y. Men, Appl.
Catal., B, 2017, 218, 825–832.
12 W. L. He, X. H. Guo, J. Zheng, J. L. Xu, T. Hayat, N. S.
Alharbi and M. Zhang, Inorg. Chem., 2019, 58, 7255–7266.
13 J. Zheng, M. Zhang, T. Miao, J. X. Yang, J. L. Xu and M.
Wakeel, Mater. Chem. Front., 2019, 3, 224–232.
14 J. G. Wang, X. R. Li, X. Li, J. Zhu and H. X. Li, Nanoscale,
2013, 5, 1876–1881.
15 X. J. Wang, J. Feng, Y. C. Bai, Q. Zhang and Y. D. Yin, Chem.
Rev., 2016, 116, 10983–11060.
16 Y. S. Li and J. L. Shi, Adv. Mater., 2014, 26, 3176–3205.
4. Conclusions
In summary, this work developed a facile approach to fabri-
cate mesoporous hollow nanospherical MnxCo3−xO4 solid so-
lutions with tunable Mn/Co molar ratios in the spinel-type
cubic crystal phase and explored their catalytic performances
for acetone oxidation. Results indicated that Mn1.20Co1.80O4
presented the optimal acetone oxidation activity, mainly as-
cribed to the strong synergetic effect achieved by integrating
the unique mesoporous hollow nanospherical feature, abun-
dant oxygen vacancies and surface active oxygen species, the
enhanced chemical nature originating from the presence of
more Mn4+, Mn3+ and Co3+ reactive sites, and the improved
redox properties. Meanwhile, Mn1.20Co1.80O4 also presented
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6379–6390 | 6389

View Article Online
Paper
Catalysis Science & Technology
17 K. Fujiwara, Y. Kuwahara, Y. Sumida and H. Yamashita,
Langmuir, 2017, 33, 288–295.
40 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751–767.
18 Y. Zeng, X. Wang, H. Wang, Y. Dong, Y. Ma and J. N. Yao,
Chem. Commun., 2010, 46, 4312–4314.
41 L. Liu, J. J. Shi and R. Y. Wang, J. Nanosci. Nanotechnol.,
2018, 18, 7775–7785.
19 V. Malgras, P. Jood, Z. Q. Sun, S. X. Dou, Y. Yamauchi and
J. H. Kim, Chem. – Eur. J., 2014, 20, 10451–10455.
20 J. H. Guo, C. X. Lin, C. J. Jiang and P. Y. Zhang, Appl. Surf.
Sci., 2019, 475, 237–255.
21 E. Genty, J. Brunet, C. Poupin, S. Ojala, S. Siffert and R.
Cousin, Appl. Catal., B, 2019, 247, 163–172.
42 Z. K. Zhao, J. H. Lin and G. R. Wang, AIChE J., 2015, 61,
239–252.
43 J. G. Hou, S. Y. Cao, Y. Z. Wu, F. Liang, Y. F. Sun, Z. S. Lin
and L. C. Sun, Nano Energy, 2017, 32, 359–366.
44 Y. Liu, B. Wei, L. L. Xu, H. Gao and M. Y. Zhang,
ChemCatChem, 2015, 7, 4076–4084.
22 Y. J. Luo, Y. B. Zheng, J. C. Zuo, X. S. Feng, X. Y. Wang, T. H.
Zhang, K. Zhang and L. L. Jiang, J. Hazard. Mater.,
2018, 349, 119–127.
23 X. Chen, J. J. Li, X. Chen, S. C. Cai, E. Q. Yu, J. Chen and
H. P. Jia, ACS Appl. Nano Mater., 2018, 1, 2971–2981.
24 T. Chang, Z. X. Shen, Y. Huang, J. Q. Lu, D. X. Ren, J. Sun,
J. J. Cao and H. X. Liu, Chem. Eng. J., 2018, 348, 15–25.
25 J. H. Li, X. Liang, S. C. Xu and J. M. Hao, Appl. Catal., B,
2009, 90, 307–312.
26 T. Ataloglou, J. Vakrosa, K. Bourikas, C. Kordulisa and A.
Lycourghiotis, Appl. Catal., B, 2005, 57, 299–312.
27 J. González, R. López, J. Ortiz and B. Rivas, Appl. Catal., B,
2016, 199, 384–393.
28 Y. L. Zheng, W. Z. Wang, D. Jiang and L. Zhang, Chem. Eng.
J., 2016, 284, 21–27.
45 H. Li, J. G. Shi, K. Zhao and L. Z. Zhang, Nanoscale, 2014, 6,
14168–14173.
46 W. C. Huo, X. Dong, J. Y. Li, M. Liu, X. Y. Liu, Y. X. Zhang
and F. Dong, Chem. Eng. J., 2019, 361, 129–138.
47 S. F. Yang, J. G. Wang, W. Chai, J. Zhu and Y. Men, Catal.
Sci. Technol., 2019, 9, 1699–1709.
48 J. G. Wang, Z. M. Chen, G. J. Zhai and Y. Men, Appl. Surf.
Sci., 2018, 462, 760–771.
49 M. Z. Wu, W. C. Zhan, Y. L. Guo, Y. Guo, Y. S. Wang, L.
Wang and G. Z. Lu, Appl. Catal., A, 2016, 523, 97–106.
50 J. G. Wang, P. H. Rao, W. An, J. L. Xu and Y. Men, Appl.
Catal., B, 2016, 195, 141–148.
51 J. G. Wang, H. Liang, C. Zhang, B. Jin and Y. Men, Appl.
Catal., B, 2019, 256, 117874.
52 Y. R. Shi, X. L. Tang, H. H. Yi, F. Y. Gao, S. Z. Zhao, J. G.
Wang, K. Yang and R. C. Zhang, Ind. Eng. Chem. Res.,
2019, 58, 3606–3617.
29 X. W. Xie, Y. Li, Z. Q. Liu, M. Haruta and W. J. Shen, Nature,
2009, 458, 747–749.
30 G. J. Zhai, J. G. Wang, Z. M. Chen, W. An and Y. Men, Chem.
Eng. J., 2018, 337, 488–498.
53 Q. H. Zhang, X. H. Liu, W. Q. Fan and Y. Wang, Appl. Catal.,
B, 2011, 102, 207–214.
31 G. J. Zhai, J. G. Wang, Z. M. Chen, S. F. Yang and Y. Men,
J. Hazard. Mater., 2019, 363, 214–226.
54 J. Chen, X. Chen, W. J. Xu, Z. Xu, J. Z. Chen, H. P. Jia and J.
Chen, Chem. Eng. J., 2017, 330, 281–293.
32 J. K. Zhu and Q. M. Gao, Microporous Mesoporous Mater.,
2009, 124, 144–152.
55 X. Chen, S. C. Cai, J. Chen, W. J. Xu, H. P. Jia and J. Chen,
Chem. Eng. J., 2018, 334, 768–779.
33 G. N. Li, L. Li, J. J. Shi, Y. Y. Yuan, Y. S. Li, W. R. Zhao and
J. L. Shi, J. Mol. Catal. A: Chem., 2014, 390, 97–104.
34 N. T. Binh, G. Cedric, C. Renaud, T. H. Lucette, N. M. Son,
L. T. Son and N. P. Xuan, Chim. Oggi, 2013, 31, 51–54.
35 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.
36 T. H. Lim, S. B. Park, J. M. Kim and D. H. Kim, Mol. Catal.,
2017, 426, 68–74.
37 L. F. Shen, L. Yu, H. B. Wu, X. Y. Yu, X. G. Zhang and X. W.
Lou, Nat. Commun., 2015, 6, 6694–6702.
56 X. Chen, E. Q. Yu, S. C. Cai, H. P. Jia, J. Chen and P. Liang,
Chem. Eng. J., 2018, 344, 469–479.
57 N. Radic, B. Grbic and A. Terlecki-Baricevic, Appl. Catal., B,
2004, 50, 153–159.
58 H. J. Wu, L. D. Wang, J. Q. Zhang, Z. Y. Shen and J. H. Zhao,
Catal. Commun., 2011, 12, 859–865.
59 P. Liu, H. P. He, G. L. Wei, X. L. Liang, F. H. Qi, F. D. Tan, W.
Tan, J. X. Zhu and R. L. Zhu, Appl. Catal., B, 2016, 182, 476–484.
60 C. Y. Ma, D. H. Wang, W. J. Xue, B. J. Dou, H. L. Wang and
Z. P. Hao, Environ. Sci. Technol., 2011, 45, 3628–3634.
61 Y. C. Huang, K. H. Ye, H. B. Li, W. J. Fan, F. Y. Zhao, Y. M.
Zhang and H. B. Ji, Nano Res., 2016, 9, 3881–3892.
62 Y. P. Wan, W. R. Zhao, Y. Tang, L. Li, H. J. Wang, Y. L. Cui,
J. L. Gu, Y. S. Li and J. L. Shi, Appl. Catal., B, 2014, 148–149,
114–122.
38 W. T. Zhao, Y. Y. Zhang, X. W. Wu, Y. Y. Zhan, X. Y. Wang,
C. T. Au and L. L. Jiang, Catal. Sci. Technol., 2018, 8, 4494–4502.
39 M. C. Cabanas, G. Binotto, D. Larcher, A. Lecup, V. Giordani
and J. M. Tarascon, Chem. Mater., 2009, 21, 1939–1947.
6390 | Catal. Sci. Technol., 2019, 9, 6379–6390
This journal is © The Royal Society of Chemistry 2019