Applied Catalysis B: Environmental 282 (2021) 119560

Contents lists available at ScienceDirect

Applied Catalysis B: Environmental

journal homepage: www.elsevier.com/locate/apcatb

Generation of abundant defects in Mn-Co mixed oxides by a facile agar-gel T

method for highly efficient catalysis of total toluene oxidation
Peifen Wanga, Jing Wanga, Xiaowei Ana, Jin Shib,c, Wenfeng Shangguanc, Xiaogang Haod,
Guangwen Xue, Bing Tangf, Abuliti Abudulaa, Guoqing Guana,b,*
a
Graduate School of Science and Technology, Hirosaki University, 1-Bunkyocho, Hirosaki, 036-8560, Japan
b
Energy Conversion Engineering Laboratory, Institute of Regional Innovation (IRI), Hirosaki University, 2-1-3 Matsubara, Aomori, 030-0813, Japan
c
School of Mechanical Engineering, Shanghai Jiao Tong University, 800 Dongchuan RD. Minhang District, Shanghai, 200240, PR China
d
Department of Chemical Engineering, Taiyuan University of Technology, Taiyuan 030024, PR China
e
Institute of Industrial Chemistry and Energy Technology, Shenyang University of Chemical Technology, Shenyang University of Chemical Technology (SYUCT), Shenyang,
110142, PR China
f
School of Environmental Science and Technology, Guangdong University of Technology, Guangzhou, 510006, PR China

ARTICLE INFO ABSTRACT

Keywords: A facile agar-gel method was successfully applied to prepare a series of homogenous Mn-Co mixed oxides with
Mixed Mn-Co oxides different Mn/Co molar ratios for highly efficient catalysis of total toluene oxidation. The obtained Mn2Co1
Agar-gel method mixed oxide exhibited a high catalytic activity with excellent durability as well as high tolerability even under
Defect-rich structure 20 vol.% water vapor environment. Characterizations by XRF, XRD, N2 adsorption-desorption, SEM, TEM, H2-
Toluene oxidation
TPR, O2-TPD, XPS measurements with density functional theory (DFT) calculations indicated that the obtained
DFT calculation
optimum Mn-Co mixed oxide had high specific surface area, large amounts of Co3+, Mn3+, and Mn4+ species,
abundant adsorbed oxygen species, rich lattice defects, high reducibility at low temperature, and excellent O2
adsorption ability. Meanwhile, in situ DRIFTS spectra measurements confirmed the generated intermediates at
different environments during the toluene oxidation and revealed that both adsorbed oxygen and lattice oxygen
species could simultaneously participate in the toluene adsorption-oxidation process, especially the rate-con­
trolling step should be the C=C breakage of aromatic ring. It is expected that the present agar-gel method with
low cost and convenient operation could be widely applied for the designing of novel mixed metal oxides cat­
alysts for VOCs oxidation.

1. Introduction noble metals (e.g., Pt, Au, Pd, and Ag) and transition metal oxides [3].
The former ones commonly demonstrate excellent activities at low
Volatile organic compounds (VOCs) are harmful to the atmosphere temperatures [11,12], however, their widely practical applications are
environment and human health [1,2]. Among VOCs, toluene with toxic, limited owing to the rare resource and high cost [13,14]. In contrast,
mutagenic and malodorous characteristics, is always used as a model the latter ones have abundant resources and low cost, but the perfor­
VOC to evaluate the performances of potential VOC removal technol­ mances are always lower the former ones. Nowadays, various methods
ogies [3,4]. So far, various processes such as biodegradation [5], ad­ have been developed to tune the catalytic performance so that transi­
sorption and absorption [6], plasma-catalysis [7], photocatalytic oxi­ tion metal oxide based catalysts can achieve high activity as well as
dation [1], wet scrubbing [8], and catalytic oxidation [9,10] have been stability comparable to the noble metal based ones [15–17].
developed and applied for the toluene removal. Especially, the catalytic Among various transition metal oxide based catalysts for the total
oxidation is considered as a most economical way due to its high re­ toluene oxidation, manganese oxide exhibits excellent catalytic per­
moval efficiency, low energy consumption, and almost no harmful by­ formance due to the mixed valences of Mn (Mn2+, Mn3+ and Mn4+),
products [9,10]. However, development of high-performance catalysts open tunnel structure, high mobility of lattice oxygen, and abundant
for the total VOCs oxidation plays the key role in the removal effi­ oxygen species on the surface [18,19]. Meanwhile, cobalt oxide with
ciency. To date, the developed catalysts mainly include the supported modified morphology and more exposed crystalline planes also shows

⁎
Corresponding author at: Energy Conversion Engineering Laboratory, Institute of Regional Innovation (IRI), Hirosaki University, 2-1-3 Matsubara, Aomori 030-
0813, Japan.
E-mail address: [email protected] (G. Guan).

https://doi.org/10.1016/j.apcatb.2020.119560
Received 11 July 2020; Received in revised form 12 September 2020; Accepted 16 September 2020
Available online 20 September 2020
0926-3373/ © 2020 Elsevier B.V. All rights reserved.
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

high catalytic performance [4]. Moreover, it is found that combining of 100 mL of distilled water and then the calculated amount (0–8.61 g) of
manganese with cobalt and/or other metal oxides to create Mn con­ manganese nitrate hydrate (Mn(NO3)2∙6H2O, Wako, Japan) was added
taining mixed oxide catalyst by using a suitable method could achieve into the above solution based on the molar ratio of Mn : Co (0, 1, 2, and
higher activity for the total toluene oxidation than either single MnOx 3). Thereafter, the mixture solution was heated to 80 °C and magneti­
or single CoOx due to the synergistic effect. Zhan et al. [20] prepared cally stirred until the solution became transparent, then the solution
various bimetal M-Mn2O3 (M = Zn, Co, Ce) mixed oxide catalysts with was naturally cooled down to room temperature. The transparent so­
highly dispersed active sites and designable compositions and found lution was covered with a parafilm and aged at least 72 h. Finally, the
that the ratio of different valences of Mn species in the mixed oxides as-synthesized gel was calcined at 350 °C for 2.5 h with a heating rate of
could be adjusted by mixing with different metal species. They con­ 5 °C/min to obtain the final powder catalysts, which are labelled as Co,
firmed that the presence of more Mn2+ species is not benefit for the Mn1Co1, Mn2Co1, and Mn3Co1, respectively (Mn-based catalysts). For
oxidation reaction. Xu et al. [21] loaded Co-Mn mixed oxides on Ni comparison, 1 g of agar powder and 2.87 g of Mn(NO3)2∙6H2O were
foam support by hydrothermal method for the total toluene oxidation dissolved in 100 mL of distilled water and the calculated amount of Co
and certified that high concentrations of Mn3+ and Co3+ species with (NO3)2∙6H2O in the range of 0–8.73 g was also introduced into the so­
rich oxygen vacancy defects existed on the surface, which played the lution based on the molar ratio of Co : Mn (0, 2, and 3), and then fol­
key role in the improving of catalytic activity. Meanwhile, Luo et al. lowed the same steps as the above to obtain the final powder catalysts,
[22] synthesized the Mn-Co mixed oxides by pyrolysis of the metal which are labelled as Mn, Mn1Co2, and Mn1Co3 respectively (Co-based
organic framework (MOF) precursors, and found that the optimum one catalysts).
with excellent catalytic performance for the total toluene oxidation
contained more Mn4+ rather than Mn3+ with the Co3+ species and
abundant oxygen vacancy defects on the surface. It seems that the 2.2. Catalyst characterizations
generation of more Mn3+ as well as Mn4+ species on the surface of Mn-
based catalysts could result in higher catalytic activity for the total The average chemical contents in the catalysts were determined by
VOCs oxidation. In addition, the formation of abundant multi-phase using X-ray fluorescence (XRF) analysis (Energy Dispersive X-ray
interfaces by metal doping can enhance the catalytic activity for the spectrometer, EDX-800HS, Shimadzu). Morphology was observed by a
total toluene oxidation since more oxygen vacancies could be generated Hitachi SU8010 SEM at 15 kV. TEM image was collected by a JEM-2100
on the catalyst surface [23,24]. Thus, generation of abundant defects transmission electron microscope (JEOL) with an operating voltage of
with the appropriate oxygen vacancy defects in the Mn-Co mixed oxides 200 kV. XRD pattern was obtained by a Rigaku SmartLab X-Ray dif­
is necessity for the promoting of catalyst performance. fractometer using a Cu-Kα (λ = 1.5405 Å) radiation source. Raman
Many traditional strategies such as coprecipitation method [2,25], spectrum was measured using a JASCO NRS-5100 Raman spectrometer
sol-gel method [26,27], and electrodeposition method [28] have been at 532 nm. The Brunauer-Emmett-Teller (BET) surface area was mea­
proposed to increase the surface defects of various mixed metal oxides sured on an ASAP 2020 Plus Phys sorption instrument (Micromeritics,
for the total VOCs oxidation. Recently, some templating strategies were ASAP 2020, US) at the liquid N2 temperature (-196 °C). The surface
also applied to improve the catalyst performance of transition metal chemical composition and elemental valence states of catalysts were
oxide catalysts. For instance, Liu et al. [29] used a carboxy-modified acquired by an X-ray photoelectron spectrometer (XPS) with a VG
colloidal crystal templating method to prepare MxCo3-xO4 (M = Zn and Scientific ESCALab250i-XL unit equipped with an Al Kα radiation
Ni) mixed oxide catalysts and observed that the carboxy-group source.
(−COOH) in the precursor had great influence on the activity of final Hydrogen temperature programmed reduction analysis (H2-TPR)
catalysts. Fu et al. [30] applied citric acid (including −COOH and −OH was performed on a BELCAT catalyst analyzer equipped with a thermal
group) gel method to synthesize various MnOx-CeO2 mixed oxides, and conductivity detector (TCD). In brief, 50 mg of catalyst was pretreated
found that more oxygen vacancy defects were formed on the surface of at 300 °C in a He atmosphere with a flow rate of 50 cm3/min for 30 min
catalysts. In our previous study, a regenerated cellulose template was at first and then, cooled down to 30 °C at the same condition. H2-TPR
used to synthesize Co3O4 catalysts with rich scattered defects and dis­ analysis was performed in a 5% H2/Ar flow (50 cm3/min) from 30 to
order phases, and it is found that the template played an important role 900 °C with a heating rate of 10 °C/min. Oxygen temperature pro­
in the formation of surface defects, thereby improving the catalytic grammed desorption (O2-TPD) was tested on the same instrument. In
performance [31]. Inspired by these templating methods, herein, we brief, 50 mg of catalyst was pretreated under a O2 atmosphere (30 cm3/
developed an agar-gel synthesis method to prepare novel Mn-Co mixed min) at 120 °C for 60 min and cooled down to 30 °C at the same con­
oxide catalysts with controllable structure and optimum amount of dition. Then, the pretreated catalyst was heated from 30 to 900 °C in a
oxygen vacancy defects for the total toluene oxidation. The obtained He flow (30 cm3/min) with a heating rate of 10 °C/min [2].
catalysts were characterized by XRF, XRD, BET, SEM, TEM, H2-TPR, O2- All of the operando diffuse reflectance infrared FT spectroscopy
TPD, XPS analyses with density functional theory (DFT) calculations. (DRIFTS) studies were carried out through a Frontier FTIR spectrometer
The catalysis stability was tested in the absence and presence of water (PerkinElmer) equipped with a DRIFTS cell and a highly sensitive MCT
vapor. Furthermore, to understand the rate-controlling step as well as detector cooled by liquid N2 in the range of 650–4000 cm−1 with 32
the structure-activity relationship, the key intermediates during the scans. Firstly, the powder sample was pre-treated in pure N2 (100 cm3/
adsorption-oxidation of toluene over the catalysts were identified by in min) at 300 °C for 30 min to remove the adsorbed gas, and then the
situ DRIFTS analysis. It is expected to provide a facile but effective way temperature was cooled down to 236 °C to collect the background in N2
to coordinate the metal species homogeneously with controllable atmosphere. Subsequently, the reactant gas (1000 ppm toluene/N2)
structure and optimum amount of oxygen vacancy defects in the ob­ with a flow rate of 100 cm3/min was continuously introduced into the
tained catalysts for the total toluene oxidation. in situ reaction chamber at 236 °C, and the DRIFTS spectrum (4000–650
cm−1) was collected for 50 min. Thereafter, N2 purging was performed
2. Experimental section for about 1 h to remove the physical absorbed toluene. Then, air was
introduced into the cell and simultaneously the spectrum was started to
2.1. Preparation of MnxCoy mixed oxides by agar-gel method record. After removing the reactant gas from the cell using pure N2 gas,
a mixture of toluene/air/5 vol.% water vapor was introduced into the
Mn-Co mixed oxides were prepared by an agar-gel method for the system and simultaneously the DRIFTS spectrum (4000–650 cm−1) was
first time. Briefly, 1 g of agar powder (Wako, Japan) and 2.91 g of co­ collected.
balt nitrate hydrate (Co(NO3)2∙6H2O, Wako, Japan) were dissolved in

2
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

2.3. Catalytic performance test which decreased with the increase in the doping amount of Co species.
It is noteworthy that the measured Mn/Co ratios well matched the
Catalytic toluene oxidation was performed in a tubular fixed bed corresponding theoretical values (3, 2, 1, 0.5, 0.33) in MnxCoy mixed
reactor (i.d. = 8.0 mm), in which about 50 mg of catalysts (40–60 oxide catalysts. To describe conveniently, the theoretical doping
mesh) were mixed with quartz sand (SiO2, 40 mesh). Before activity amount in MnxCoy mixed oxide catalysts was used in the following
test, the reactor system was pretreated in a N2 flow at 100 °C with a context. In addition, the Mn-based catalysts were mainly discussed in
ramp rate of 10 °C/min for 60 min in order to avoid the possibility of the main context since one of the Mn-based catalysts, i.e., Mn2Co1,
over-estimation of toluene oxidation conversion. Subsequently, the exhibited the best catalysis performance with the lowest T100% as
activities from 100 °C to 300 °C with a temperature ramp of 10 °C/min shown in Section 3.6.
were investigated under a gas mixture contained 1000 ppm toluene in
20 vol.% O2 with a total gas flow rate of 50 cm3/min (N2 as balance
gas), corresponding to a weight hourly space velocity (WHSV) of 3.1. TEM and XRD characterization
60,000 cm3/(g h). Herein, a thermocouple was inserted into the center
of catalyst bed for monitoring the actual reaction temperature and it is SEM images of the MnxCoy mixed oxide catalysts with different Mn/
found that the actual temperature was about 12 °C higher than the Co molar ratios are shown in Fig. S1. Interestingly, for the single Mn or
setting temperature. To investigate the effect of WHSV, the amounts of Co oxide, an overall coral-like network structure was formed, but for
catalysts were adjusted to the demands of WHSVs = 30,000 and the MnxCoy mixed oxides, three-dimensional porous structure with
120,000 cm3/(g h). Moreover, the effect of water steam on the catalytic random pore size distribution was obviously observed. This mor­
toluene oxidation over the catalyst with the best performance in this phology may be caused by the strong interaction between Mn and Co
study was studied by introducing 5, 10 and 20 vol.% water vapors se­ species, which hinder the separately aggregations of either MnOx or
parately into the reaction system. CoOx crystals [32]. The nanostructures of single MnOx and the Mn2Co1
Inlet and outlet gases were analyzed by on-line gas chromatographs mixed oxide catalysts with the best catalytic performance as indicated
(GC) equipped with a flame ionization detector (FID, GC-2014, in Section 3.6 were further elucidated by TEM analyses. As shown in
Shimadzu) and a thermal conductivity detector (TCD, GC-8A, Figs. 1 and S2, the nanostructure of catalysts was formed by assembling
Shimadzu). Concentration of CO2 in the outlet gas was measured by of the nanoparticles. The distinct lattice fringes of the Mn oxide with
using an FT-IR gas analyzer (Horiba, FG-120). Toluene conversion the interplanar spacings of 0.20, 0.24, 0.28, and 0.49 nm corresponded
(Xtoluene) and CO2 selectivity (YCO2 ) were calculated respectively by the to the (220), (211), (103) and (101) planes of the Mn4O3 phase, re­
following two equations: spectively, and the Mn3O4 particle sizes were relatively uniform (Fig.
S2). The distinct lattice fringes of Mn2Co1 mixed oxide with the in­
CInlet COutlet terplanar spacings of 0.14, 0.20, 0.24, 0.27 nm indicated the presences
Xtoluene = ×100%
CInlet of (400), (220), (211), and (103) planes of the CoMn2O4, respectively
CCO2 (Fig. 1a). Compared with the lattice fringes of the single Mn oxide,
YCO2 = × 100% those of the Mn2Co1 mixed oxide were not obvious but still observed as
7(CInlet COutlet )
shown in Fig. 1a, indicating its much lower crystallinity, which are
where, COutlet and CInlet represent the outlet and inlet concentrations of consistent with the XRD results in Fig. 1c. Thus, the doping of Co
toluene, respectively, and CCO2 represents the outlet CO2 concentration. species decreased the crystallinity of mixed oxide and resulted in the
formation of more amorphous phases [22,32]. In this case, as shown in
2.4. Computational method Figs. 1b and S1, it is very difficult to measure the particle size based on
SEM or TEM images. Especially, abundant defects and edge dislocations
The toluene oxidation occurred on the main (211) facet of Mn-based were confirmed by the TEM image. As observed in Fig. 1(a-1), there
catalyst was investigated by DFT calculations, in which the Mn3O4 and were distinct lattice fringes. However, as shown in Fig. 1(a-2)–(a-4), the
CoMn2O4 crystals were selected as the calculation models in this study. crystalline fringes along the crooked edge were discontinuous, which
The computational method is described in details in the supplementary could provide abundant defect structures. Fig. 1(a-5) illustrate the
material. possible defect structures derived by the edge dislocation with the ad­
ditional active edge sites. It is well-known that crystal defects existed
3. Results and discussion ubiquitously in metal oxides can radically boost the catalytic perfor­
mance, and especially the oxygen vacancy defects on the catalysts could
The average element contents of MnxCoy mixed oxide catalysts in have great effect on the reducibility of heterogeneous catalysts, leading
bulk confirmed by XRF analysis are summarized in Tables 1 and S2. The to excellent catalytic activity for VOC oxidation at low temperature.
atomic ratios of Mn to Co (Mn/Co) were 2.91, 1.93, 0.99, 0.49 and 0.33 The crystal defects include interface and edge dislocations, crystalline
for Mn3Co1, Mn2Co1, Mn1Co1, Mn1Co2 and Mn1Co3, respectively, deformation and so on, and the oxygen vacancy defects usually appears

Table 1
Elemental contents based on XRF analysis, BET surface areas, and surface molar ratios based on XPS spectra of the MnxCoy mixed oxides (Mn-based catalysts).
a b c d
Sample Element content (at. %) BET (m²/g) D (nm) V (cm3/g) Surface molar ratio e

Mn Co Mn/Co Mn4+/Mn2+ Mn3+/Mn2+ Co3+/Co2+ Oads/Olatt

Mn 100 / / 35.5 10.05 0.11 0.87 2.36 / 0.72

Mn1Co1 49.8 50.2 0.99 21.9 10.63 0.096 2.68 3.42 0.88 0.64
Mn2Co1 65.9 34.1 1.93 85.4 5.81 0.17 3.14 4.43 1.00 0.87
Mn3Co1 74.4 25.6 2.91 53.6 7.09 0.12 2.83 4.01 0.87 0.83

a
Element content determined by XRF analysis.
b
BET specific surface area.
c
Average pore diameter calculated from the desorption branch of the isotherm using the BJH method.
d
Total pore volume measured at P/P0 = 0.99.
e
Calculated based on XPS spectra.

3
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Fig. 1. TEM images of the Mn2Co1 mixed oxide catalyst (a, b); (a-1), (a-2), (a-3) and (a-4) are different magnification areas in Fig. 1a. (a-5) Schematic illustration of
defect structure derived from edge dislocation with additional active edge sites generation. The edge dislocations are marked by yellow “T”. The defects are indicated
in Figs a-2 to a-4. Scale bar is 2 nm. XRD patterns of the as-prepared MnxCoy mixed oxides (Mn-based catalysts) (c). (For interpretation of the references to colour in
this figure legend, the reader is referred to the web version of this article).

along with these defects. In general, the oxygen vacancy defects should Fig. S4 shows Raman spectra of the prepared MnxCoy mixed oxide
be benefit for the promoting of the replenishment and reconstruction of catalysts. For the Mn-based ones, with the increase in the Co amount,
oxygen species, and enhance the cycling of the emergence-annihilation the most intensive peak at 645 cm−1 belonging to the symmetric
of oxygen vacancy defects during the VOC oxidation process at low stretching of Mn (Ⅲ, Ⅳ) decreased, and even disappeared in the case of
temperature [21,22]. Therefore, a large number of defects on the basal the Mn/Co molar ratio of 1:1 whereas the weak peak at 355 cm−1 (out-
surface could generate abundant oxygen vacancies, creating additional of-plane bending modes) disappeared soon after the doping of Co spe­
reaction sites to improve the catalytic activity [31]. Such defects were cies, indicating the strong interaction between Mn and Co species,
considered to be derived from agar-gel aging process, which should be which could weaken the Mn−O bond. These results are consistent with
benefit for the total toluene oxidation at low temperatures. the XRD analysis results [32]. Meanwhile, three weak Raman peaks at
Figs. 1c and S3 show XRD patterns of the MnxCoy mixed oxide the 190, 473, 515 cm−1 and a strong peak at 670 cm−1 were observed
catalysts. For the single Mn oxide, all diffraction peaks can be ascribed for the single Co oxide, where the strong Raman peak at 670 cm−1 was
to those of Mn3O4 (JCPDS 18-0803) (Fig. 1c), which is consistent with attributed to the Co3O4. With the addition of Mn species in the Co-based
the TEM observation results (Fig. S2). In contrast, for the MnxCoy mixed catalysts, even the strongest peaks became much broader, weaker and
oxides, the broad characteristic XRD peaks indicated the formation of even shifted to some extent or disappeared, also suggesting the strong
CoMn2O4-like phase (JCPDS 01-1126), which is also consistent with the interaction between Mn and Co species [22]. As such, the mobility of
TEM analysis results (Fig. 1a). Herein, the main diffraction peaks cor­ oxygen species could be enhanced so that abundant oxygen vacancy
responding to the CoMn2O4 phase were 18.3, 20.3, 29.9, 33.1, 36.2, defects could be more easily generated around the weakened Mn−O or
44.4, 54.2, 60.9, 64.7°. With the decrease in the percentage of Co Co−O bond, and the cycling of the emergence-annihilation of oxygen
species in mixed oxide catalyst, the intensity of diffraction peak got vacancy defects could be easily realized during the oxidation process,
weaker and weaker, indicating that the incorporation of cobalt species thereby improving the catalysis performance.
lowered the crystallinity of Mn3O4 oxides with the formation of more
amorphous-like structure. Meanwhile, the main (211) facet was ob­ 3.2. N2 adsorption-desorption measurements
served clearly but slightly right-shifted, which should be resulted from
the contraction of the unit cell due to the incorporation of cobalt into Figs. 2 and S5 show nitrogen adsorption-desorption isotherms of the
the Mn3O4 framework. Table S1 lists the calculated lattice parameters Mn-based and Co-based MnxCoy mixed oxide catalysts, respectively.
for the Mn3O4 and CoMn2O4 based on the DFT calculations. Compared One can see that all these catalysts had a type IV N2 isotherm but with
with the pristine Mn3O4, it is obvious that the introduction of Co species distinct H3-shaped hysteresis loops [34] at 0.4–1.0, 0.7–1.0, 0.45–1.0,
led to the lattice contractions in the a, b and c directions, resulting the 0.43–1.0 for the single Mn, Mn1Co1, Mn2Co1, and Mn3Co1 mixed
generation of amorphous phase. The existence of the mixed crystalline oxide catalysts respectively, which should be resulted from the ag­
and amorphous phases could lead to rich defect structures, thereby gregation of nano-particles to form slit-like mesoporous pores (as in­
improving the intrinsic catalytic activity. For example, Aguilera et al. dicated in the TEM images). Tables 1 and S2 summarize the corre­
and Zhang et al. [2,25] also reported that the coupling of Mn and Co sponding BET surface areas of the Mn-based and Co-Based MnxCoy
species could result in the generation of amorphous phases with the mixed oxide catalysts, respectively. One can see that the Mn2Co1 mixed
enhancement of redox properties, leading to superior VOC oxidation oxide catalyst had the highest BET surface area and too large fractions
activity. Herein, the similar effectiveness could be also expected for the of either Mn or Co were not beneficial for the improvement of surface
catalytic total toluene oxidation. In addition, as shown in Fig. 1c, no area since they cannot result in good three-dimensional mesoporous
typical broad peaks of amorphous carbon in the 2θ range of 20−30° structure. As such, the optimal Mo/Co ratio to achieve the best per­
and 40−45° were detected, demonstrating that the agar in MnxCoy formances (as indicated in Fig. 5c) existed for either Mn-based or Co-
mixed catalysts was completely removed in the annealing process and a based catalysts by using the agar-gel method. Furthermore, as observed
portion of the carbon maybe remained as the carbide state in the cat­ in Fig. S6, the doped-Co species in MnOx led to a wider pore size dis­
alysts [33]. tribution, which was also changed with the molar ratio of Mn/Co.

4
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

shown in Fig. 3b, an obvious reduction peak at 306 °C, and other two
large broad peaks at 491 and 591 °C were observed for the single Mn
oxide catalyst, which corresponded to the reductions of MnO2 →
Mn2O3, Mn2O3 → Mn3O4, and Mn3O4 → MnO, respectively [38,39]. In
comparison, for the Mn2Co1 and Mn3Co1 mixed oxides, two reduction
peaks corresponding to the reductions of MnO2 → Mn2O3 and Mn2O3 →
Mn3O4 at the similar peak positions became more obvious, implying
that the Mn4+ and Mn3+ species should be the main species in these
mixed oxides [38]. Thus, the incorporation of Co into MnOx also pro­
moted the reducibility of catalysts, which could enhance the total to­
luene oxidation at a lower temperature. Meanwhile, the peak area
corresponding to the reduction of Mn3O4 → MnO in the high tem­
perature region also became larger significantly, however, it should
have a minimal effect on catalytic activity at the low operating tem­
perature range in the toluene oxidation. In addition, for the Mn1Co1
Fig. 2. Nitrogen adsorption-desorption isotherms of the MnxCoy mixed oxides mixed oxide, all the reduction peaks shifted to high temperature di­
(Mn-based catalysts). rection when comparing with the single Mn oxide, indicating its rela­
tively poor reduction properties. Moreover, the peaks on the TPR at
Hence, it can be concluded that the addition of Co species had obvious high temperature range (above 600 °C) in the MnxCoy mixed oxide
influence on the pore structure of the MnxCoy mixed oxide catalysts due catalysts could be resulted from the reductions of the different local
to the strong interaction between Mn and Co species. environment phases species, such as carbide, Co-Co, Mn-Co, and Mn-
Mn [39], which also indicated the strong interaction between Co and
Mn species in the MnxCoy mixed oxide catalysts. As summarized in
3.3. O2-TPD and H2-TPR analyses Table 2, the total amounts of O2 consumptions were in the order of
Mn3Co1 > Mn2Co1 > Mn > Mn1Co1 whereas the total amounts of
O2-TPD analysis was applied for studying oxygen species and H2 consumptions were in the order of Mn2Co1 > Mn1Co1 >
oxygen desorption behavior of the MnxCoy mixed oxides and the results Mn3Co1 > Mn. Thus, considering the oxygen species and reducibility,
are shown in Figs. 3a, S7a and Table 2. Obviously, three O2 desorption as confirmed in Section 3.5, the Mn2Co1 mixed oxide had the best
peaks in the temperatures centered at 223 °C (α-species, O2, O2− and catalytic performance among all prepared catalysts for the total toluene
O22− adsorbed on O-vacancies), 414 °C (β-species, oxygen atoms bound oxidation.
to Mn(III)) and 561 °C (γ-species, oxygen atoms bound to Mn(IV)) were
observed in the Mn oxide, which are consistence with the reported
results [35]. For the Mn2Co1 and Mn3Co1 mixed oxide catalysts, the O2 3.4. XPS analysis
desorption peaks also appeared in the similar ranges of 150–300,
300–467, and 467−685 °C °C, respectively, but the corresponding The surface atomic concentrations and valence states of elements on
peaks shifted to lower temperatures and became obviously stronger, the catalysts were investigated by XPS analysis, and the results are
indicating that the oxygen desorption ability for the Mn2Co1 and shown in Figs. 4 and S8, and summarized in Tables 1 and S2. As shown
Mn3Co1 mixed oxides was significantly enhanced via the doping of Co in Fig. 4a, the Mn 2p spectra of the MnxCoy mixed oxide catalysts can be
species into the MnOx lattice by the agar-gel method, which should be decomposed into three main spin orbital lines with the binding energies
benefit for the deep toluene oxidation at a lower temperature [36]. In at 640.5, 641.8 and 643.5 eV, which are assigned to the Mn2+, Mn3+
comparison, only two peaks at a higher temperature range of and Mn4+ species, respectively [40]. The relatively high surface atomic
320−770 °C were observed for the Mn1Co1 mixed oxide, indicating no ratios of Mn4+/Mn2+ and Mn3+/Mn2+ indicated the abundance Mn3+
existence of O2, O2− and O22− species on it, which could lead to a low and Mn4+ species on Mn2Co1 oxide, which are consistence with the H2-
efficiency in the total toluene oxidation. Such differences between the TPR results. It should be noted that the Mn2+ amount on the Mn2Co1
single Mn oxide and the Mn-based MnxCoy mixed oxides should be oxide was lower than those on other Mn-based MnxCoy mixed oxides
related to the conversion of crystal phase, especially the appearance of but more Mn4+ and Mn3+ species were found on it, which could pro­
amorphous CoMn2O4 phase in the mixed oxides after the Co-doping mote the redox property [41]. Meanwhile, the asymmetrical Co 2p XPS
using the agar-gel method [2,37]. spectra featured at the binding energies of 780.9, 779.6 and 786.3 eV
H2-TPR analysis was used for studying the reducibility of catalysts were assigned to Co3+ and Co2+ species (Fig. 4b) whereas the shake-up
and the corresponding results are shown in Figs. 3b, S7b and Table 2. As satellites were related to Co2+ species [42]. As summarized in Table 1,

Fig. 3. (a) O2-TPD profiles and (b) H2-TPR profiles of the MnxCoy mixed oxides (Mn-based catalysts).

5
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Table 2
Peak positions, O2 consumptions in O2-TPD and H2 consumptions in H2-TPR for the Mn-based MnxCoy mixed oxide catalysts.
Catalyst O2-TPD (mmol/g) H2-TPR (mmol/g)

T (°C) O2-uptake T (°C) O2-uptake T (°C) O2-uptake Total uptake T (°C) H2-uptake T (°C) H2-uptake T (°C) H2-uptake Total uptake

Mn 223 0.179 414 0.592 561 0.613 1.384 306 3.313 491 5.092 591 3.715 12.12
Mn1Co1 / / 449 0.397 599 0.508 0.905 331 1.375 451 6.606 610 11.946 19.927
Mn2Co1 204 0.106 409 0.899 519 1.334 2.339 325 7.544 436 4.917 628 8.167 20.628
Mn3Co1 200 0.149 407 1.148 507 1.468 2.765 332 7.644 452 5.144 636 3.979 16.767

the surface molar ratio of Co3+/Co2+ decreased in the sequence of 3.5. Catalytic performances
Mn2Co1＞Mn3Co1≈Mn1Co1. Herein, more Co3+ species exposed on
the surface of the Mn2Co1 oxide could also lead to excellent perfor­ Toluene oxidations over all the prepared catalysts were performed
mance for the total toluene oxidation [22]. under a WHSV = 60,000 cm3/(g h) at first, and the toluene conversions
In Fig. 4c, the O 1s XPS spectrum of each Mn-based MnxCoy mixed as a function of reaction temperature over the Mn-based catalysts and
oxide catalyst can be divided into three peaks at the binding energies of Co-based catalysts are shown in Figs. 5 and S9. As displayed in Fig. 5a,
529.8, 531.2, and 533.5 eV, which are ascribed to the lattice oxygen the toluene conversion increased with the rise in temperature, but the
(Olatt, O22−), the surface adsorbed oxygen species with low coordina­ complete conversion temperatures had significant differences for dif­
tion such as OH groups, oxygen vacancies (Oads, O2− or O−), and the ferent MnxCoy mixed oxide catalysts. It is found that the Mn2Co1 mixed
oxygen species in the adsorbed molecular water, respectively [43]. oxide catalyst exhibited the best performance in the total toluene oxi­
Generally, the surface adsorbed oxygen species, especially oxygen va­ dation to CO2 and H2O. As summarized in Table 3, the temperatures for
cancy defects (O2− or O−), could contribute to a greater extent on the converting half amount of toluene (T50 %) and for converting com­
catalytic activity since they have higher mobility than the lattice pletely (T100 %) were in the order of Mn2Co1 (228 °C, 238 °C) <
oxygen [44]. As summarized in Table 1, the Mn2Co1 oxide had a Mn3Co1 (243 °C, 254 °C) < Mn (253 °C, 270 °C) < Mn1Co1 (256 °C,
highest ratio of Oads/Olatt with the sequence of Mn2Co1＞Mn3Co1＞Mn 274 °C). Compared with the previously reported Mn-Co mixed oxide
＞Mn1Co1, which is also consistence with the catalytic performances in catalysts, the prepared Mn2Co1 mixed oxide catalyst in this study also
the total toluene oxidation as shown in Section 3.5. Herein, it should be showed a higher catalytic activity. Meanwhile, according to the tem­
noted that the three surface molar ratios of Oads/Olatt, Mn4+/Mn2+, and perature of mineralization as shown in Fig. 5b, the Mn2Co1 mixed
Co3+/Co2+ on the Mn2Co1 mixed oxide catalyst were the highest ones oxide also exhibited a better performance than those over other Mn-
among all the samples (Table 1 and Fig. 4d). based catalysts. These results are well consistent with the

Fig. 4. XPS spectra of (a) Mn 2p, (b) Co 2p, and (c) O 1s for the Mn-based catalysts, (d) Comparison of the surface molar ratios of Oads/Olatt, Mn4+/Mn2+, and Co3+/
Co2+ on the surfaces of Mn-based catalysts.

6
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Fig. 5. (a) Toluene conversions as a function of reaction temperature over the Mn-based catalysts; (b) CO2 selectivity during the catalytic toluene oxidation; (c)
Comparison between Mn-based catalysts and Co-based catalysts in terms of T50% and T100% conversions.

Table 3
Comparison of the performances of the Mn-Co mixed oxide catalysts reported in the literature with the present work.
Catalyst Preparation Method WHSV cm3/(g h) Toluene conversion/ºC Stability time/h Ref.

T50% T90% T100%

CoMn/AC oxide Impregnation 66,000 244 246 263 11.7 [65]

Mn-Co oxides Pyrolysis 96,000 226 240 300 25 [22]
Co1Mn1 BHNCs Pyrolysis 60,000 232 248 300 35 [66]
Mn-Co (1:1)/CZ Co-precipitation. 30,000 222 242 250 50 [67]
Co-Mn composite Hydrothermal 30.000 263 267 270 30 [21]
Mn oxide Agar-gel 60,000 253 266 270 / This work
Mn1Co1 oxides Agar-gel 60,000 256 270 274 /
Mn2Co1 oxides Agar-gel 60,000 228 236 238 50
Mn3Co1 oxides Agar-gel 60,000 243 252 254 /

characterizations as stated above. Herein, the intrinsic catalytic activity 60,000 cm3/(g h) was higher than that of total toluene conversion.
of the Mn-Co mixed oxide catalysts prepared by the ager-gel method for Meanwhile, the WHSVs of 30,000 and 120,000 cm3/(g h) exhibited the
the toluene oxidation should be related to the oxygen vacancy defects mineralization temperatures similar to those of toluene conversion. It
on the catalysts. The strong interaction between Co and Mn species in indicated that the WHSV had obvious effect on the toluene oxidation
Mn-Co mixed oxide could result in abundant the oxygen vacancy de­ performance at a low reaction temperature. However, compared with
fects. Thus, the Mn2Co1 mixed oxide with the best catalytic perfor­ the case where the WHSV = 120,000 cm3/(g h), the performance of
mance for total toluene oxidation should be due to its appropriate catalytic toluene oxidation had a little difference when comparing with
molar ratio of Mn/Co, leading to higher surface molar ratios of Oads/ those in the cases of WHSVs = 30,000 and 60,000 cm3/(g h). Thus, to
Olatt, Mn4+/Mn2+, Mn3+/Mn2 and Co3+/Co2+ with more oxygen va­ reduce the cost and maintain efficiency for the catalytic total toluene
cancies than other ones. oxidation, the WHSV of 60,000 cm3/(g h) was selected as the optimum
In addition, Fig. 5c compare the performances of Mn-based catalysts one for the further study.
and Co-based catalysts (Table. S3) in terms of T50% and T100% conver­
sions along with the Mn/Co molar ratio. Interestingly, the plots of to­
3.7. Stability test
luene conversion temperature vs. the Mn/Co molar ratio presented an
inverted “M” shape with the increase of the Co ration. The Mn1Co1
To investigate the catalytic stability and durability, long-term ac­
mixed oxide catalyst exhibited a highest conversion temperature
tivity of the Mn2Co1 mixed oxide catalyst was evaluated in the absence
whereas the Mn2Co1 and Mn1Co2 mixed oxide catalysts showed the
of water vapor under different temperatures at first. As shown in
lowest conversion temperatures in the Mn-based and Co-based cata­
Fig. 7a, the complete oxidation of toluene maintained unchanged
lysts, respectively, which indicated that the optimal doping ratio ex­
during the long-term test at a temperature of 252 °C, indicating that the
isted for either Mn-bases or Co-based catalysts. In this study, Mn/Co of
catalyst was very stable at this temperature. After decreasing of the
2/1 (Mn2Co1 mixed oxide) and Mn/Co of 1/2 (Mn1Co2 mixed oxide)
reaction temperature to 247 °C, the toluene conversion over the
were the optimum balance ratio when the agar-gel method was used for
Mn2Co1 mixed oxide catalyst was reduced to ca. 30 % but also main­
the catalyst preparation.
tained long-term durability at this condition.
In order to further examine the effect of water vapor on the per­
3.6. Effect of WHSV formance of the Mn2Co1 mixed oxide catalyst, 5, 10, and 20 vol% of
water vapors were separately introduced into the reaction system with
WHSV is usually considered as a significant factor to affect the the air at a WHSV of 60,000 cm3/(g h). As shown in Fig. 7b, the toluene
conversion in the catalytic VOCs oxidation [45]. Herein, the toluene conversion remained at 100 % during the introducing of 5, 10 and
conversions over the Mn2Co1 mixed oxide catalyst under different 20 vol% H2O at 252 °C in 50 h test period. At 247 °C, the conversion of
WHSVs of 30,000, 60,000 and 120,000 cm3/(g h) were investigated. As toluene still maintained a stable value of ca. 30 % in the presence of
shown in Fig. 6a, the toluene conversion decreased gradually with the 5 vol% of water vapor. It is noteworthy that the further increasing of
increase in the WHSV. When the WHSV was 60,000 cm3/(g h), T50% water vapor content to 10 and even 20 vol% resulted in the increase of
(228 C) and T100% (238 °C) were obtained, which were only 13 and 6 °C conversion from ca. 30 % to ca. 40 %, which was higher at the same
higher than those under WHSV = 30,000 cm3/(g h), but 16 and 20 °C reaction temperature, indicating that the catalytic performance was
lower than those under 120,000 cm3/(g h), respectively. As shown in improved after the addition of moderate amounts of water vapor.
Fig. 6b, the mineralization temperature under the WHSV o;1;f Herein, since abundant hydroxyl-modified sites existed on the Mn-Co

7
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Fig. 6. Effect of WHSV on the catalytic performance of Mn2Co1 mixed oxide catalyst (a) Toluene conversion; (b) Selectivity towards CO2.

accepted especially for the VOCs oxidation over the transition metal
oxide catalysts. To understand the relationship between the surface
adsorbed species and catalytic reaction process of toluene oxidations
over the single Mn oxide and Mn2Co1 mixed oxide catalysts, in situ
DRIFTS analysis was performed. As shown in Fig. 8, the peak at
1302 cm−1 is attributed to C–H in-plane bending vibration of toluene
aromatic ring [46,47], the peaks at 1072 and 1087 cm−1 belong to the
CO− stretching vibrations of alkoxide, symbolizing the formation of
benzyl alcohol [48], the peak at 1212 cm−1 corresponds to the CO– –
stretching vibration of phenolate species [47], the peak at 1302 cm−1 is
assigned to CH2 deformation vibration of benzyl [49], the peaks at
around 1388 and 1551 cm−1 belong to the symmetric and antisym­
metric carboxylate C=O stretching vibration of benzoate, respectively,
and the peak at 1388 cm−1 may be also assigned to monodentate
carbonates [50–52]. Herein, the bridging structure of carboxylate spe­
cies can be confirmed by the distance (Δυ) between asymmetric and
symmetric carboxylate C=O stretching vibrations [53]. Meanwhile, the
peaks at 1452, 1496 and 1603 cm−1 are attributed to skeletal C=C
stretching vibration of the aromatic ring [51,52,54], the bands at 1725,
1795, 1863 and 1944 cm−1 are associated with the symmetric and
asymmetric C=O stretching vibrations of maleic anhydride [48,49],
the peaks at 2749 and 2888 cm−1 can be assigned to C=O stretching
vibrations of carbonyl, suggesting the formation of benzaldehyde spe­
cies [55], the bands at approximately 2936 and 3038 cm−1 are ascribed
to the symmetric and asymmetric CH− stretching vibrations of benzyl,
Fig. 7. (a) Reaction stability test for the toluene oxidation over the Mn2Co1 respectively [56,57], and the peak at 3086 cm−1 corresponds to the
mixed oxide catalyst; (b) Effect of water vapor on the catalytic stability and
phenylic C–H stretching vibration of −CH3 group [46,47]. Furthermore,
durability for the Mn2Co1 mixed oxide catalyst at 247 and 252 °C respectively
the final product of CO2 indicated by the peaks at around 2324 and
with a WHSV of 60,000 cm3/(g h).
2361 cm−1 were also observed obviously at the all in situ DRIFTS
spectra and the intensities were gradually intensified with the
mixed oxide catalysts, the introduction of H2O vapor into the reaction prolonging of detecting time [32,58]. In addition, the peak at around
system could activate those hydroperoxyl-like groups on the catalysts, 3236 cm−1 is assigned to O–H, suggesting the formation of water as the
which should be benefit to the activation O−O bond and enhancement important final product in the catalytic toluene oxidation.
of the O2 replenishment during the oxidation process, thereby im­ Fig. 8a displays a typical set of time-dependent DRIFTS spectrum
proving the performance [32]. In addition, more H2O vapor introducing during a toluene oxidation process over the Mn oxide catalyst at 236 °C
could more easily take away the generated intermediates from the in N2 atmosphere. It is observed that two weak peaks appeared at 1551
surface of catalyst, which is also beneficial to expose the active sites for and 1388 cm−1 at the first 2 min period, and the related position on the
the deep oxidation. To further understand the improvement effect, in Δυ (about 163 cm−1) of asymmetric and symmetric C=O stretching
situ DRIFTS study was performed as shown in the following Section 3.8. vibration was very close to the free ion values, suggesting that the
Moreover, after water vapor introduction was interrupted, the toluene carboxylate species with a bridging structure were coupled with Mn
conversion was recovered back to ca. 29 %. In addition, the morphology valence ions on the surface of catalyst and the adsorbed toluene could
of Mn2Co1 mixed oxide catalyst after the long-term stability test was interact with the oxygen species on the catalyst surface to produce
observed, and as shown in Fig. S10, the 3D porous structure was also benzoate with a rapid speed. However, many functional groups of in­
maintained, further indicating that the Mn2Co1 mixed oxide catalyst termediates were difficult to be observed at the first 5 min period be­
had good tolerance to the water vapor and retained excellent catalytic sides the aromatic ring. With extending of reaction time, the species
performance in the simulated real exhausting gas condition. adsorbed on the catalyst surface increased significantly and accumu­
lated to a maximum till 50 min. In this process, the peak at 1212 cm−1
appeared after 10 min, indicating that the CO- of phenolate species
3.8. In situ DRIFTS analysis
could be accumulated in the O2-free environment. With further ex­
tending of time, four peaks at 1725, 1795, 1863, and 1944 cm−1 were
The Mars-van Krevelen (MVK) mechanism is becoming widely

8
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Fig. 8. In situ DRIFTS spectra of toluene oxidation as a function of time exposed to (a) 1000 ppm toluene/N2 atmosphere and (b) 1000 ppm toluene/air atmosphere
over Mn oxide catalyst; (c) 1000 ppm toluene/N2 atmosphere, (d) 1000 ppm toluene/air atmosphere, and (e) 1000 ppm toluene/air atmosphere with 5 vol.% water
vapor over Mn2Co1 mixed oxide catalyst; (f) Proposed mechanism for toluene oxidation over Mn-based catalysts.

detected, which also enhanced with the extending time, indicating that respectively, demonstrating the mobility of oxygen species was greatly
the related intermediates could react with lattice oxygen in the bulk to improved on the surface of the Co-doped Mn-based catalyst to enhance
form anhydride even after exhausting the adsorbed oxygen species on the formation of the final product when compared with the single Mn
the surface or subsurface of catalyst. In contrast, for the Mn2Co1 mixed oxide catalyst. As stated above, the amorphous-like structure in the
oxide catalyst (Fig. 8c), all characteristic peaks appeared at the first Mn2Co1 mixed oxide catalyst could produce abundant surface oxygen
2 min period in the time-dependent DRIFTS spectra and their intensities species, which can enhance the oxidation of the intermediates into CO2
increased significantly with extending of time, indicating that the species. These results indicated that toluene can be rapidly adsorbed
Mn2Co1 mixed oxide catalyst accelerated the toluene oxidation rate and activated on the surface of Mn2Co1 mixed oxide catalyst, which is
more greatly than the pure Mn oxide catalyst. Moreover, it should be conducive to the catalytic toluene oxidation even under oxygen-free
noted that there was no peak of carboxylate group at around 1550 conditions due to the improvement of the migration rate of lattice
cm−1 appeared, instead, the characteristic peak at 1388 cm−1 of car­ oxygen as well as the oxygen storage capacity as confirmed by the O2-
bonates monotonously increased, confirming that the oxygen-free at­ TPD analysis.
mosphere prevented the decomposition of carbonates preferentially to Furthermore, in order to study the effect of O2 in the gas phase on
CO2 and H2O. Significantly, the weak peak at 1212 cm−1 was hardly the toluene oxidation over the single Mn oxide and Mn2Co1 mixed
observed, indicating that the formation of phenolate species was not oxide catalysts, air was introduced into the system at 236 °C for in situ
enhanced with time in the toluene oxidation process. Thus, the DRIFTS analysis on the toluene adsorption and oxidation processes. As
benzoate should be preferentially decomposed to carbonates, or the shown in Fig. 8b and d, all characteristic peaks of various intermediates
transformation between phenolate and maleic anhydride could be in­ appeared at the first 2 min period and their intensities gradually en­
terrupted rapidly on the surface of the Mn2Co1 mixed oxide catalyst. In hanced with the time increasing, indicating that the presence of O2
addition, two higher peaks at around 2360 and 3236 cm−1 than those accelerated the toluene oxidation rate. The relatively lower intensities
shown in the Fig. 8a belong to CO2 and crystal water species of IR bands corresponding to anhydride were observed in Fig. 8b and d,

9
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

indicating the presence of less maleic anhydride species under the air atmosphere at 236 °C based on durability test over the Mn2Co1 mixed
atmosphere since the consumable adsorbed oxygen species on the sur­ oxide catalyst under the condition with water vapor. As shown in
face or subsurface of catalyst can be replenished and reconstructed Fig. 8e, all intensities of bands were higher than those recorded in the
rapidly by O2 in the gas phase via the oxygen vacancy defects. Im­ cases without water vapor, indicating that the presence of water vapor
portantly, small differences were observed when compared with in situ could activate O–O bond, which should be benefit for the toluene ad­
DRIFTS spectra obtained in the N2 atmosphere. For the single Mn oxide sorption on the surface of catalyst [32,60]. Here, the intensities of the
catalyst as shown in Fig. 8b, the intensity of peak at 1555 cm−1 as­ bands at ca. 2324 and 2361 cm−1 ascribed to the adsorbed CO2, sug­
signing to benzoate obviously increased with the time due to the oxi­ gesting that the accumulation of CO2 occurred on the surface of Mn-Co
dation of the benzaldehyde to benzoate, manifesting that benzaldehyde mixed catalyst. Meanwhile, the bands also attributed to the continuous
was quickly transformed to benzoate in the presence of rich gaseous increasing of carbonate species. As such, the competition between ac­
oxygen. Meanwhile, a weak band at 1212 cm−1 corresponding to canyumulation and desorption of CO2 should exist on the active sites.
phenolate was observed, implying that C=C could not be quickly However, in the water vapor environment, it should be noted that the
broken to maleic anhydride or carbonates. In comparison, for the characteristic band at 1402 cm-1 disappeared quickly at the first 2 min
Mn2Co1 mixed oxide catalyst as shown in Fig. 8d, no increases in the period, manifesting that the carbonates were quickly and preferentially
characteristic peaks of benzoate, monodentate, and phenolate at around transformed to CO2 and H2O, leading to the regeneration of active sites,
1550, 1380, and 1212 cm−1 respectively with the time were observed, which is beneficial to the catalyst stability. The durability test as shown
confirming that the Co-doped Mn-based catalyst can effectively pro­ Fig. 7b also confirmed that the introduction of water vapor enhanced
mote the toluene oxidation. Therefore, it can be verified that the C=C the catalytic activity of toluene oxidation over the Mn2Co1 mixed oxide
breakage of the aromatic ring could be the rate-controlling step in the catalyst.
catalytic toluene oxidation over the Mn-based catalysts. Herein, the Based on the above DRIFTS spectra analyses, Table 4 summaries the
oxygen vacancies caused by the abundant crystal defects could promote important intermediates, and Fig. 8f illustrates the proposed me­
the C=C breakage of aromatic ring, which further accelerate the re­ chanism for toluene oxidation on the surface of Mn-based catalysts. As
action efficiency [59]. shown in Fig. 8f, toluene molecule attracted on the surface of catalyst
Furthermore, the bond length for O2 adsorbed on the surface of could be oxidized as the following order: benzyl → benzyl alcohol →
catalyst after migrating from gas phase to the oxygen vacancy defect benzaldehyde → benzoate. Then, the benzoate could be decomposed
site was investigated based on DFT calculations. In Fig. 9, the bond into either the short-chain carbonates by breaking C=C bond or phe­
length was defined as the distance between the O atom and metal atom, nolate by breaking C=O bond, and further oxidized to maleic anhy­
which was represented by the dotted lines, and showed that the oxygen dride and finally to CO2 and H2O. Herein, the fast C=C breaking of
molecule was adsorbed on the model surface. Herein, 2.137 Å and aromatic ring in benzoate could be the rate-controlling step. In addi­
1.875 Å were the distances of two different Mn atoms on the Mn3O4 tion, it should be noted that the typical time-dependent DRIFTS spectra
(211) plane in the single Mn oxide to O atom whereas 2.008 Å and exposed to the three environments of N2, air and water vapor were
1.868 Å were the distances of Mn-O and Co-O on the CoMn2O4 (211) similar, suggesting that both lattice oxygen and adsorbed oxygen in the
plane in the MnxCoy mixed oxide, respectively. The shorter distances Mn2Co1 mixed oxide catalyst could participate in the toluene oxidation
indicated that the oxygen adsorption ability was promoted by the since the rich defect sites in the catalyst could enhance the mobility of
doping of Co species on the Mn oxide. Moreover, the adsorption energy lattice oxygen species, thereby improving the activation. Meanwhile,
of O2 adsorbed on the CoMn2O4 (211) plane was -4.2726 eV, which was the consumable oxygen species would be replenished and reconstructed
also higher than that on the Mn3O4 (211) plane model, also demon­ rapidly in the presence of gas-phase oxygen molecules via the oxygen
strating that the doping of Co species on the Mn oxide enhanced the O2 vacancy defects.
adsorption ability on the (211) facet, which should be benefit for the
replenishing and reconstructing of the surface adsorbed oxygen though 3.9. Effect of agar in the process of catalyst preparation
O2 in the gas phase for intermediate oxidation via abundant defect sites
generated on amorphous region. As discussed above, it is found that abundant adsorbed oxygen
To further investigate the effect of water vapor on the formation of species are formed in the Mn2Co1 mixed oxide catalyst due to the rich
intermediates, 5 vol.% H2O was introduced with 1000 ppm toluene/air defects, which should be originally derived from the agar structure.
Agar is generally consisted of the polydisperse linear polymers of D-
galactose and 3,6-anhydro-L-galactose with few variations in the re­
petitive units (Fig. 10) [61], resulting in at least two conformations, i.e.,
ordered and helical conformations [62]. Furthermore, the structural
change is mainly resulted from a series of reactions regarding to the
hydrolysis and oxidation [61,63,64]. As shown in Fig. 10, the ag­
gregation of agar could occur primarily via intermolecular and in­
tramolecular hydrogen bonds or Vander Waals forces. After the in­
troduction of Mn2+ and Co2+ ions, the ionic bonds between the agar
and metal ions will be formed. Subsequently, the agar gelation takes
place with the generation of agarose, in which the divalent cations are
wrapped in the structure of agar [62]. The combining effect of both
agar conformations and chemical reactions should be taken into con­
sideration in the agar-gel aging process. The helical conformation could
result in the molecule rotation at random, where the rotational strength
always depends on the configuration of the polymer chains attached
with nearby metal ions like Mn2+ and Co2+ [62]. For example, a
fraction of the Co2+ and Mn2+ ions attached to the helical conforma­
tion section could lead to the crystalline deformation, leading to the
formation of edge dislocations and amorphous-like region in the final
Fig. 9. The models illustrating O2 adsorbed on the surfaces of Mn3O4 (211) and catalyst structure (as shown in Fig. 1a-2–a-4). Meanwhile, Mn2+ and
CoMn2O4 (211), respectively. Co2+ connected with the ordered conformation section of agar could

10
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

Table 4
Assignment of IR bands appearing in the process of toluene oxidation over the Mn oxide catalyst and Mn2Co1 mixed oxide catalyst at 236 °C.
Position (cm−1) Assignment Corresponding species

1032 C–H in-plane bending vibration Aromatic ring

1072, 1087 CO– stretching vibrations of alkoxide species Benzyl alcohol
1212 CO– stretching vibration Phenolate
–
1302 CH2 deformation vibration Benzyl
Symmetric and asymmetric C=O stretching vibrations of carboxylate group
1388, 1551 Benzoate
Skeletal C=C stretching vibrations
1452, 1496, 1603 Aromatic ring
Symmetric and asymmetric C=O stretching vibrations of cyclic anhydrides
1725, 1795, 1863, 1944 Maleic anhydride
2749, 2888 C=O stretching vibrations of carbonyl Benzaldehyde
2936, 3038 Symmetric and asymmetric CH– stretching vibrations of CH2 group Benzyl
3086 Phenylic C–H stretching vibration Aromatic ring
3236 O–H stretching vibrations Crystal water

lead to the formation of crystalline region (Fig. 1a-1). Simultaneously, 4. Conclusions

as shown in Fig. 10, the polymer chains become shorter owing to the
hydrolysis reaction and the generation of exfoliated small molecules, The MnxCoy mixed oxide catalysts with defect-rich structures were
which could cause abundant defects and edge dislocation structures and successfully synthesized using the agar-gel method. It is demonstrated
more catalytic sites will be exposed. The intermolecular oxidation could that the molar ratio of Mn/Co greatly affected the phase structure and
make the hydroxyl groups connected to the primary carbon atoms, textural properties of the as-prepared catalysts. Among the MnxCoy
resulting in the formation of carboxyl functional groups and the gen­ mixed oxide catalysts, the Mn2Co1 mixed oxide catalyst exhibited a
eration of a large amount of free H+ ions in the agar gel. According to highest catalytic performance with T50% of 228 C and T100% of 238 °C
the report by Yang et al. [48], the Mn-based catalyst can be treated by for the toluene oxidation. According to the characterizations of the
acid (H+) to generate a relatively high amount of surface Mn4+ ions Mn2Co1 mixed oxide catalyst, it is considered that the enhanced cat­
and probably lead to the formation of oxygen vacancy defects though alytic performance was contributed to the higher surface area, higher
the disproportionation of Mn3+, which could further contribute to contents of Mn4+, Mn3+ and Co3+ species with more oxygen vacancies
better oxygen mobility and lower temperature reducibility. Similarly, and better reducibility. The high stability and renewability of Mn2Co1
herein, it can be considered that the oxidation of hydroxyl groups mixed oxide catalyst were also confirmed by the activity and endur­
promotes the H+ ion generation, which can serve as leaching species to ability tests with 5, 10, and 20 vol% of water vapors. Furthermore, in
provide high concentration of oxygen vacancy defects in the final oxide situ DRIFTS analysis revealed that both adsorbed oxygen and lattice
catalysts. As such, the high concentration of defects is conducive to the oxygen species could simultaneously participate in the toluene ad­
formation of adsorbed oxygen species on the surface or subsurface of sorption-oxidation process, especially the rate-controlling step should
catalysts. be the C=C breakage of aromatic ring. It is expected that the present
agar-gel method with low cost and convenient operation can be widely

Fig. 10. Schematic diagram of the synthesis of MnxCoy mixed oxide catalysts using the agar-gel method.

11
P. Wang, et al. Applied Catalysis B: Environmental 282 (2021) 119560

applied for the designing of novel mixed metal oxides catalysts for [21] X. Jiang, W. Xu, S. Lai, X. Chen, RSC Adv. 9 (2019) 6533–6541.
VOCs oxidation. [22] Y. Luo, Y. Zheng, J. Zuo, X. Feng, X. Wang, T. Zhang, K. Zhang, L. Jiang, J. Hazard.
Mater. 349 (2018) 119–127.
[23] Z. Li, Q. Yan, Q. Jiang, Y. Gao, T. Xue, R. Li, Y. Liu, Q. Wang, Appl. Catal. B:
CRediT authorship contribution statement Environ. 269 (2020) 118827.
[24] C. Dong, Z. Qu, X. Jiang, Y. Ren, J. Hazard. Mater. 391 (2020) 122181.
[25] D.A. Aguilera, A. Perez, R. Molina, S. Moreno, Appl. Catal. B: Environ. 104 (2011)
Peifen Wang: Data curation, Formal analysis, Investigation, 144–150.
Writing - original draft. Jing Wang: Investigation. Xiaowei An: [26] X. Li, J. Zheng, S. Liu, T. Zhu, J. Colloid Interface Sci. 555 (2019) 667–675.
Investigation. Jin Shi: Investigation. Wenfeng Shangguan: [27] D.P. Debecker, V. Hulea, P.H. Mutin, Appl. Catal. A: Gen. 451 (2013) 192–206.
[28] J. Wang, A. Yoshida, P. Wang, T. Yu, Z. Wang, X. Hao, A. Abudula, G. Guan, Appl.
Investigation. Xiaogang Hao: Investigation. Guangwen Xu: Catal. B: Environ. 271 (2020) 118941.
Investigation. Bing Tang: Investigation. Abuliti Abudula: [29] M. Zhao, J. Deng, J. Liu, Y. Li, J. Liu, Z. Duan, J. Xiong, Z. Zhao, Y. Wei, W. Song,
Investigation, Project administration. Guoqing Guan: Y. Sun, ACS Catal. 9 (2019) 7548–7567.
[30] X. Lin, S. Li, H. He, Z. Wu, J. Wu, L. Chen, D. Ye, M. Fu, Appl. Catal. B: Environ. 223
Conceptualization, Funding acquisition, Methodology, Project admin­
(2018) 91–102.
istration, Resources, Supervision, Validation, Writing - review & [31] X. Li, X. Su, Y. Pei, J. Liu, X. Zheng, K. Tang, G. Guan, X. Hao, J. Mater. Chem. A 7
editing. (2019) 10745–10750.
[32] L. Zhao, Z. Zhang, Y. Li, X. Leng, T. Zhang, F. Yuan, X. Niu, Y. Zhu, Appl. Catal. B:
Environ. 245 (2019) 502–512.
Declaration of Competing Interest [33] A.A. Voevodin, S.V. Prasad, J.S. Zabinski, J. Appl. Phys. 82 (1997) 855–858.
[34] L. Yu, R. Peng, L. Chen, M. Fu, J. Wu, D. Ye, Chem. Eng. J. 334 (2018) 2480–2487.
[35] Y.G. Yin, W.Q. Xu, R. Deguzman, S.L. Suib, C.L. O’Young, Inorg. Chem. 33 (1994)
The authors declare that they have no known competing financial 4384–4389.
interests or personal relationships that could have appeared to influ­ [36] X. Xie, Y. Li, Z.Q. Liu, M. Haruta, W. Shen, Nature 458 (2009) 746–749.
ence the work reported in this paper. [37] L. Xue, C. Zhang, H. He, Y. Teraoka, Appl. Catal. B: Environ. 75 (2007) 167–174.
[38] W. Gac, Appl. Catal. B: Environ. 75 (2007) 107–117.
[39] S. Todorova, H. Kolev, J.P. Holgado, G. Kadinov, C. Bonev, R. Pereñíguez,
Acknowledgements A. Caballero, Appl. Catal. B: Environ. 94 (2010) 46–54.
[40] Z. Sihaib, F. Puleo, J.M. Garcia-Vargas, L. Retailleau, C. Descorme, L.F. Liotta,
J.L. Valverde, S. Gil, A. Giroir-Fendler, Appl. Catal. B: Environ. 209 (2017)
This work is supported by ZiQoo Chemical Co. Ltd., Japan. Ms P.
689–700.
Wang and Mr J. Shi gratefully acknowledge China Scholarship Council [41] Z. Ye, J.M. Giraudon, N. Nuns, P. Simon, N. De Geyter, R. Morent, J.F. Lamonier,
(CSC), P. R. China. Appl. Catal. B: Environ. 223 (2018) 154–166.
[42] H. Yi, Z. Yang, X. Tang, S. Zhao, F. Gao, J. Wang, Y. Huang, Y. Ma, C. Chu, Q. Li,
J. Xu, Chem. Eng. J. 333 (2018) 554–563.
Appendix A. Supplementary data [43] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, Appl. Catal. B: Environ.
99 (2010) 353–363.
Supplementary material related to this article can be found, in the [44] M.A. Peluso, L.A. Gambaro, E. Pronsato, D. Gazzoli, H.J. Thomas, J.E. Sambeth,
Catal. Today 133–135 (2008) 487–492.
online version, at doi:https://doi.org/10.1016/j.apcatb.2020.119560. [45] J. Chen, X. Chen, X. Chen, W. Xu, Z. Xu, H. Jia, J. Chen, Appl. Catal. B: Environ. 224
(2018) 825–835.
References [46] H. Sun, Z. Liu, S. Chen, X. Quan, Chem. Eng. J. 270 (2015) 58–65.
[47] Q. Zhang, S. Mo, J. Li, Y. Sun, M. Zhang, P. Chen, M. Fu, J. Wu, L. Chen, D. Ye,
Catal. Sci. Technol. 9 (2019) 4538–4551.
[1] V. Héquet, C. Raillard, O. Debono, F. Thévenet, N. Locoge, L. Le Coq, Appl. Catal. B: [48] X. Yang, X. Yu, M. Lin, X. Ma, M. Ge, Catal. Today 327 (2019) 254–261.
Environ. 226 (2018) 473–486. [49] J. Li, H. Na, X. Zeng, T. Zhu, Z. Liu, Appl. Surf. Sci. 311 (2014) 690–696.
[2] Q. Zhang, X. Liu, W. Fan, Y. Wang, Appl. Catal. B: Environ. 102 (2011) 207–214. [50] A. Lu, H. Sun, N. Zhang, L. Che, S. Shan, J. Luo, J. Zheng, L. Yang, D.-L. Peng, C.-
[3] H. Huang, Y. Xu, Q. Feng, D.Y.C. Leung, Catal. Sci. Technol. 5 (2015) 2649–2669. J. Zhong, B. Chen, ACS Catal. 9 (2019) 7431–7442.
[4] L.F. Liotta, H. Wu, G. Pantaleo, A.M. Venezia, Catal. Sci. Technol. 3 (2013) 3085. [51] S. Mo, Q. Zhang, Y. Sun, M. Zhang, J. Li, Q. Ren, M. Fu, J. Wu, L. Chen, D. Ye, J.
[5] L. Malhautier, G. Quijano, M. Avezac, J. Rocher, J.L. Fanlo, Chem. Eng. J. 247 Mater. Chem. A 7 (2019) 16197–16210.
(2014) 199–204. [52] C. Zhang, C. Wang, H. Huang, K. Zeng, Z. Wang, H.-p. Jia, X. Li, Appl. Surf. Sci. 486
[6] L. Li, S. Liu, J. Liu, J. Hazard. Mater. 192 (2011) 683–690. (2019) 108–120.
[7] F. Thevenet, L. Sivachandiran, O. Guaitella, C. Barakat, A. Rousseau, J. Phys. D: [53] X. Chen, X. Chen, E. Yu, S. Cai, H. Jia, J. Chen, P. Liang, Chem. Eng. J. 344 (2018)
Appl. Phys. 47 (2014) 224011. 469–479.
[8] C.J. Hennigan, M.M.H. El-Sayed, A. Hodzic, Atmos. Environ. 188 (2018) 12–17. [54] M.D. Hernández-Alonso, I. Tejedor-Tejedor, J.M. Coronado, M.A. Anderson, Appl.
[9] C. Zhang, C. Wang, W. Zhan, Y. Guo, Y. Guo, G. Lu, A. Baylet, A. Giroir-Fendler, Catal. B: Environ. 101 (2011) 283–293.
Appl. Catal. B: Environ. 129 (2013) 509–516. [55] X. Du, X.G. Hao, Z.D. Wang, G.Q. Guan, J. Mater. Chem. A 4 (2016) 6236–6258.
[10] Y. Wang, Y. Xue, C. Zhao, D. Zhao, F. Liu, K. Wang, D.D. Dionysiou, Chem. Eng. J. [56] M. Wang, F. Zhang, X. Zhu, Z. Qi, B. Hong, J. Ding, J. Bao, S. Sun, C. Gao, Langmuir
300 (2016) 300–305. 31 (2015) 1730–1736.
[11] G. Spezzati, A.D. Benavidez, A.T. DeLaRiva, Y. Su, J.P. Hofmann, S. Asahina, [57] J.-J. Li, E.-Q. Yu, S.-C. Cai, X. Chen, J. Chen, H.-P. Jia, Y.-J. Xu, Appl. Catal. B:
E.J. Olivier, J.H. Neethling, J.T. Miller, A.K. Datye, E.J.M. Hensen, Appl. Catal. B: Environ. 240 (2019) 141–152.
Environ. 243 (2019) 36–46. [58] S. Mo, Q. Zhang, J. Li, Y. Sun, Q. Ren, S. Zou, Q. Zhang, J. Lu, M. Fu, D. Mo, J. Wu,
[12] S. Xie, Y. Liu, J. Deng, X. Zhao, J. Yang, K. Zhang, Z. Han, H. Arandiyan, H. Dai, H. Huang, D. Ye, Appl. Catal. B: Environ. 264 (2020) 118464.
Appl. Catal. B: Environ. 206 (2017) 221–232. [59] X. Dong, W. Cui, H. Wang, J. Li, Y. Sun, H. Wang, Y. Zhang, H. Huang, F. Dong, Sci.
[13] H. Deng, S. Kang, J. Ma, C. Zhang, H. He, Appl. Catal. B: Environ. 239 (2018) Bull. 64 (2019) 669–678.
214–222. [60] A. Bongiorno, U. Landman, Phys. Rev. Lett. 95 (2005) 106102.
[14] K. Everaert, J. Baeyens, J. Hazard. Mater. 109 (2004) 113–139. [61] B. Mao, A. Bentaleb, F. Louerat, T. Divoux, P. Snabre, Food Hydrocoll. 64 (2017)
[15] W.B. Li, J.X. Wang, H. Gong, Catal. Today 148 (2009) 81–87. 59–69.
[16] M. Romero-Sáez, D. Divakar, A. Aranzabal, J.R. González-Velasco, J.A. González- [62] A.H. Clark, Food Polym. Gels Colloids 83 (1991) 322–338.
Marcos, Appl. Catal. B: Environ. 180 (2016) 210–218. [63] R. Armisén, F. Gaiatas, Handbook of Hydrocolloids, (2009), pp. 82–107.
[17] Z. Ren, Z. Wu, W. Song, W. Xiao, Y. Guo, J. Ding, S.L. Suib, P.-X. Gao, Appl. Catal. B: [64] J. Zhao, Z. Chen, R. Zou, X. Zhang, Food Hydrocoll. 77 (2018) 421–426.
Environ. 180 (2016) 150–160. [65] G. Zhou, X. He, S. Liu, H. Xie, M. Fu, J. Ind. Eng. Chem. 21 (2015) 932–941.
[18] L. Lamaita, M.A. Peluso, J.E. Sambeth, H. Thomas, G. Mineli, P. Porta, Catal. Today [66] Z. Weitao, Z. Yang-Yu, W. Xiangwei, Z. Yingying, W. Xiuyun, A. Chak-Tong,
107–108 (2005) 133–138. J. Lilong, Catal. Sci. Technol. 8 (2018) 4494–4502.
[19] L. Lamaita, M.A. Peluso, J.E. Sambeth, H.J. Thomas, Appl. Catal. B: Environ. 61 [67] J. Feng, Z.-Y. Hou, X.-Y. Zhou, H.-L. Zhang, T.-Q. Cheng, T. Lin, Y.-Q. Chen, Chem.
(2005) 114–119. Pap. 72 (2017) 161–172.
[20] B. Chen, X. Yang, X. Zeng, Z. Huang, J. Xiao, J. Wang, G. Zhan, Chem. Eng. J. 397
(2020) 125424.

12