Green synthesis of cavity-containing manganese oxides with superior catalytic performance in toluene oxidation

Article abstract of DOI:10.1016/j.apcata.2019.06.005

Manganese oxide prepared by a simple hydrothermal method is highly active and stable in both lab and quasi industrial conditions for the total oxidation of toluene, achieving total conversion at only 175 °C in the absence of water and at 150 °C in the presence of 4 mol % water. Moreover, this performance is stable for at least 6 cycles. The enhanced catalytic performance has been related to the presence of highly reactive oxygen surface species. This species are probably promoted by Mn₃O₄ nanoparticles with Mn³⁺ at the tetragonal distorted cationic sites, likely compensated by structural water at the anionic sites. The presence of structural water can be related to the formation of internal cavities/vesicles in the Mn₃O₄ crystalline structure.

Full text of DOI:10.1016/j.apcata.2019.06.005

Accepted Manuscript
Title: Green synthesis of cavity-containing manganese oxides
with superior catalytic performance in toluene oxidation
´
´
´
´
´
Authors: Tomas Garcıa, Jose Manuel Lopez, Alvaro Mayoral,
´
Yaping Zhang, Raul Arenal, Daniel Alonso-Domınguez,
´
´
´
Marıa Pilar Pico, Marıa Luisa Lopez, Ana Dejoz, Inmaculada
´
´
Alvarez-Serrano, Rut Sanchis, Benjamın Solsona
PII:
S0926-860X(19)30259-5
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2019.06.005
APCATA 17107
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
13 February 2019
5 June 2019
8 June 2019
´
´
´
Please cite this article as: Garcıa T, Lopez JM, Mayoral A, Zhang Y, Arenal R,
´
´
´
Alonso-Domınguez D, Pico MP, Lopez ML, Dejoz A, Alvarez-Serrano I, Sanchis
R, Solsona B, Green synthesis of cavity-containing manganese oxides with superior
catalytic performance in toluene oxidation, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.06.005
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Green synthesis of cavity-containing manganese
oxides with superior catalytic performance in
toluene oxidation
Tomás García *1,a, José Manuel López a, Álvaro Mayoral b, Yaping Zhangb, Raul Arenal
e
c,d, Daniel Alonso-Domínguez e, María Pilar Pico f, María Luisa López , Ana Dejoz g,
Inmaculada Álvarez-Serrano *2,e, Rut Sanchis g, Benjamín Solsona*3,g
a.
b.
Instituto de Carboquímica (ICB-CSIC), C/Miguel Luesma 4, 50018 Zaragoza, Spain;.
Centre for High-Resolution Electron Microscopy (CћEM), School of Physical Science
and Technology, ShanghaiTech University, 393 Middle Huaxia Road, Pudong, Shanghai,
201210, China.
c.
Laboratorio de Microscopías Avanzadas, Instituto de Nanociencia de Aragꢀn,
Universidad de Zaragoza, 50018 Zaragoza, Spain
d.
e.
ARAID Foundation, 50018 Zaragoza, Spain.
Departamento de Química Inorgánica, Facultad de Ciencias Químicas, Universidad
Complutense de Madrid; 28040 Madrid, Spain;+
f. Sepiolsa, Avda. del Acero, 14-16, Pol. UP-1 (Miralcampo), 19200 Azuqueca de
Henares, Spain.
g.
Departament d’Enginyeria Química, ETSE, Universitat de València, Av. Universitat,
46100 Burjassot, Valencia, Spain.
* Corresponding Authors: *1 tomas@icb.csic.es , *2 ias@ucm.es , *3 benjamín.solsona@uv.es
Graphical abstract

Manganeseoxide
cavities
O = C = O
Highlights


Green synthesis of cavity-containing manganese oxides by hydrothermal
method
High catalytic performance in toluene oxidation with total conversion at <
200ºC


Highly reactive oxygen surface species at the tetragonal distorted cationic sites
Presence of structural water and internal cavities in the Mn3O4 crystalline
structure
Abstract
Manganese oxide prepared by a simple hydrothermal method is highly active and stable
in both lab and quasi industrial conditions for the total oxidation of toluene, achieving
total conversion at only 175ºC in the absence of water and at 150 ºC in the presence of 4
mol % water. Moreover, this performance is stable for at least 6 cycles. The enhanced

catalytic performance has been related to the presence of highly reactive oxygen surface
species. This species are probably promoted by Mn3O4 nanoparticles with Mn3+ at the
tetragonal distorted cationic sites, likely compensated by structural water at the anionic
sites. The presence of structural water can be related to the formation of internal
cavities/vesicles in the Mn3O4 crystalline structure.
Keywords: manganese oxide; cavities; structural water; toluene; VOCs oxidation.
1. Introduction
In the last decades, oxides of non-noble transition metals have been extensively
investigated as catalysts for the total oxidation of volatile organic compounds (VOCs),
due to the high price of the conventional catalysts which are based on noble metals
(NM) [1,2]. Moreover, these NM-based catalysts permit to operate at temperatures
ranging between 100 and 200 ºC. Among the non-NM oxides based catalysts; we can
highlight pure and supported metal oxides of cobalt, copper, manganese or iron [3-8].
Manganese oxide catalysts have demonstrated to be highly active in the total oxidation
of one-ring aromatics [9-11] and other subsets of VOCs [12-15].
Different synthetic routes, e.g. solid state, hydrothermal, molecular sieve techniques,
lead to manganese oxide-based catalysts with high stability and great catalytic
performance in the oxidation of aromatic compounds [16,17], e.g. a solvothermal
technique combined with a calcination at 500ºC, reported as the most adequate for

obtaining Mn(III) oxides-based catalysts [18]; and hydrolysis driven redox reactions
leading to manganese oxide catalysts highly tolerant to moisture [19,20].
Hence, hydrothermal processes both in continuous and batch regimes have been
employed for synthesizing inorganic particles in a wide variety of research fields, from
functional ceramics and electronics to catalysis, biomass processing and energy storage
devices [21-23]. Taking into account simplicity, low-cost and controllability criteria,
solution-based processes stand out other synthetic strategies as they present saving
energy benefits together with innocuous manipulation and tiptop control over particle
size and morphology [24]. Hydrothermal routes are based on the chemical reactions and
solubility changes of substances in a sealed heated aqueous solution above ambient
temperature and pressure to grow nanocrystals [25]. Hence, in recent literature these
environmentally benign hydrothermal techniques have been successfully employed for
obtaining a great variety of materials with enhanced response for different specific
fields; i.e. photocatalytic properties of MoS2 [26], pseudocapacitive properties of NiS
[27], photoelectrochemical performance of Sn-doped iron oxides [28], catalytic activity
of (Ti/Sn)O2 [29], electrochemical efficiency of organometal halide perovskites [30],
supercapacitor response of hybrid electrodes [31], etc. In particular, manganese oxides
with improved catalytic and electrochemical response [32] or enhanced adsorption
performance [33] could be obtained by hydrothermal routes. Besides, solvothermal
processes also permit to obtain nanoparticles with different morphologies and tuneable
nanoarchitectures of functional materials [34].
Thus, it would be highly interesting to synthesize manganese oxides in a simple,
sustainable and straightforward way maintaining the high reactivity and excellent
stability. In this sense, hydrothermal synthesis, involving water as the only solvent, is
revealed as one of the most simple, promising and green ways to prepare catalytically

active materials [35]. Among the different manganese oxide crystalline phases, Kim et
al. [14] reported that Mn3O4 is the most active phase in the total oxidation of benzene
and toluene, which was associated to the best redox properties and the highest oxygen
mobility. Later, other authors pointed out the high efficiency of -MnO2 catalyst [11],
interpreted on the basis of its three-dimensional macroporous and mesoporous
morphology. In fact, VOC oxidation over these catalysts usually occurs through a Mars
and van Krevelen (MVK) mechanism [36], where the hydrocarbon molecule is oxidised
by lattice oxygen from the manganese oxide. The lattice oxygen is then replenished by
the reduction of gaseous oxygen. Therefore, the availability and reactivity of surface
oxygen species affect the activity of manganese oxides to a great extent, being the key
parameter to develop high performance catalysts [37].
Herein, we report that a cavity-containing manganese oxide catalyst obtained by facile,
low-cost and environmentally friendly hydrothermal (HT) process shows an outstanding
catalytic activity in the removal of toluene by total oxidation. It is worth underlining the
interest of the green synthetic route chosen, i.e. a hydrothermal process in which all the
reactions take place in aqueous solution, just in one step and at low or moderate
temperatures. The remarkable activity has been associated to the presence of internal
cavities where water molecules are occluded, leading to very reactive surface oxygen
species. Besides, this catalyst is compared to another one, which is synthesized by a
solvothermal method (ST). Although the as-prepared manganese oxide (ST) shows
good catalytic activity, its performance is inferior to that displayed by the material
obtained by a HT process
2. Experimental
2.1. Preparation method

Manganese oxide samples Mn-HT (hydrothermal) and Mn-ST (solvothermal) were
prepared from 0.1 M manganese (II) acetate, Mn(C2H3O2)2•4H2O (99%, Aldrich),
solutions. The solvent was pure aqueous or ethylene glycol-water in 1:1 volume ratio,
respectively. Ammonia was then added to adjust the pH value at 11 and the
precipitation was facilitated. The suspensions were further introduced in autoclaves and
treated at 180 ºC for 24 hours. The obtained precipitates were then filtered and dried in
a stove at 100 ºC.
2.2. Catalyst characterization
Nitrogen adsorption measurements were performed on a Micromeritics ASAP 2010
physisorption analyzer at 77K. Brunauer–Emmett–Teller (BET) and Barret–Joyner–
Halenda (BJH) methods were used for the surface area and pore size distribution
determination using N2 adsorption data. Samples were previously outgassed for at least
4 hours at 120 °C.
X-ray powder diffraction (XRD) patterns were registered at room temperature with a
PAN analytical X'PERT POWDER diffractometer using Cu (K) radiation with
=1.5418 Å. Fullprof program was used for Rietveld refinement and pattern matching
analysis.
X-ray photoelectron spectroscopy (XPS) measurements were made on a Kratos Axis
ultra DLD photoelectron spectrometer using a non-monochromatized Mg Kα X-ray
source (hν = 1253.6 eV). Analyser pass energy of 50 eV was used for survey scans and
20 eV for detailed scans. Binding energies were referenced to the C1s peak from
adventitious carbonaceous contamination, assumed to have a binding energy of 284.8
eV. XPS data were analysed using CasaXPS software. All the peaks of the corrected
spectra were fitted with a Gaussian–Lorentzian shape function to peak fit the data.

Iterations were performed using the Marquardt method. Relative standard deviations
were always lower than 1.5%.
Temperature programmed reduction (TPR) analyses were carried out in a Micromeritics
Autochem 2920 instrument equipped with a thermal conductivity detector, operated
under a 50 mL min-1 10% H2/Ar flow at temperatures between -50 and 800 °C, with a
heating rate of 10 °C min-1.
Spherical aberration corrected (Cs-corrected) Scanning Transmission Electron
Microscopy (STEM) observation have been carried out in a Cold FEG JEOL Grand
ARM 300, with double corrector, allowing a point resolution of 0.6 Å, equipped with
the JEOL SDD EDS system (100mm2 x 2), a dual EELS Gatan 965 GIF Quantum ER;
and in a FEI-Titan Low-Base microscope equipped with a Cs probe corrector, a
monochromator and an ultra-bright X-FEG electron source. All measurements were
performed at 300 kV. Prior observations, the samples were highly dispersed in ethanol
for few minutes and few drops of the suspension formed were deposited onto holy
carbon copper grids.
The SEM analyses were recorded in JEOL 7800 Prime. In order to acquire the best
resolution and allow the observation of the surface of the crystallites, samples were not
coated with either gold or carbon. Instead, we adopted gentle beam mode to decrease
the chromatic aberration. Namely, the gun voltage applied was 5.5 keV, with a
decelerating voltage -5 keV on the sample holder, giving a final landing energy of 0.5
keV on the samples. In order to avoid any misinterpretation, the samples observed in the
SEM were the same ones as those observed in the TEM. Therefore, they were supported
on carbon coated TEM Cu grids.

2.3. Catalytic tests
The catalytic activity was determined in a fixed bed micro-reactor with the reaction
temperature controlled by employing a thermocouple placed in the catalyst bed. In most
of the experiments the amount of toluene used was maintained in 200 vppm (0.02%
molar) and the gas hourly space velocity was fixed in GHSV = 6,000 h-1. In the standard
experiments 99.98% of the feed consisted of synthetic gas (oxygen and helium in a
21/79 molar ratio) to give a C7H8/O2/He = 0.02/21.0/79.0 molar ratio. In the
experiments with water in the feed two amounts, 4 and 10 molar %, were used. Thus the
feed was C7H8/H2O/O2/He = 0.02/4.0/20.2/75.8 or 0.02/10.0/18.9/71.1 molar ratios. In
the experiments with CO2, 2 molar % of CO2 was added to the feed maintaining the
O2/He ratio constant. Thus, the feed employed was C7H8/CO2/O2/He =
0.02/2.0/20.6/77.4 molar ratios whereas in the experiment with H2O and CO2 was
C7H8/H2O/CO2/O2/He = 0.02/10.0/2.0/18.5/69.5 molar ratios. Some experiments were
undertaken with very high gas hourly space velocity GHSV = 240,000 h-1 maintaining
the C7H8/O2/He = 0.02/21.0/79.0 molar ratio. The reactants and reaction products were
analyzed using an online gas chromatograph with TCD and FID detectors. Two
chromatographic columns were used: (i) Porapak Q (for hydrocarbons and CO2) and (ii)
Molecular Sieve 5A (for oxygen and carbon monoxide). The reaction temperature was
employed in the 100–350 ºC range. Conversion was calculated by two methods, through
the differences between the inlet and outlet concentrations of toluene and by using the
area of the reaction products (CO2) and that of the reactant without reacting. These two
procedures lead us to adjust the carbon balance with an accuracy of ±5%. Two analyses
were done at each temperature after steady state was attained (ca. 30 min before the first
analysis) and the results were averaged. A blank run was conducted in an empty reactor
without catalyst until 350ºC, showing negligible (<1%) conversion.

3. Results and discussion
The catalytic performance has been studied by plotting the toluene conversion with the
reaction temperature using low space velocities but in the industrial range. Figure 1
shows the evolution of the toluene conversion with the reaction temperature for the two
manganese oxide catalysts synthesised. For our samples, 50 % conversion is achieved at
125 and 160 oC for Mn–HT and Mn-ST, respectively. Moreover, the 80 % conversion is
o
achieved at 150 and 200 C for Mn–HT and Mn-ST, respectively. It is noteworthy the
o
low temperature, 175 C, required for total toluene conversion using the most active
catalyst, Mn-HT. In order to check the stability of the most active material, several
catalytic cycles were conducted on Mn-HT. Figure 1 also shows the corresponding
curves for six consecutive catalytic cycles and no appreciable differences were observed
among the cycles, pointing out that this catalyst is highly stable.
Figure 1. Variation of the toluene conversion with the reaction temperature for manganese oxide
catalysts. 200 ppm in synthetic air. GHSV = 6000 h-1.

Table 1 shows the comparison in terms of activity between our Mn-HT and Mn-ST
materials and some catalysts reported in the literature, probing that the both catalysts
prepared in this work, especially under hydrothermal conditions, have an outstanding
good catalytic performance in the catalytic oxidation of toluene.
Table1. Comparison of the catalyst activity between this work and literature.
Catalyst
T50
(oC)
125
165
250
265
256
220
230
240
275
345
T80
(oC)
150
200
260
280
--
Toluene
concentration
(ppm)
GHSV
(h-1)
Ref.
Mn-HT
Mn-ST
Mn-HT
Mn-ST
Mn2O3
Mn3O4
Mn3O4
Mn3O4
Mn2O3
MnO2
200
200
6000
This work
This work
This work
This work
[18]
6000
200
240000
240000
--
200
1000
1000
1000
2000
2000
2000
230
240
248
280
353
19100
180000
15000a
15000a
15000a
[38]
[39]
[14]
[14]
[14]
a in ml h-1 g-1
catalyst
Manganese oxide hydrothermally prepared (Mn-HT) can totally remove toluene using
low space velocities at very low temperature (175ºC). This behaviour is stable and after
6 cycles similar light-off curve is observed. This stable performance is likely due to the
fact that the temperature this material has been previously heat-treated (180ºC) is higher
than that required for total conversion (175ºC) and then the structural and surface
modifications after the catalytic tests are not important. However, using high space
velocities the temperatures required for total conversion already exceed 180ºC and then
a drop in the reactivity is observed. This drop has been demonstrated undertaking an

experiment at low space velocity with catalyst previously used at high space velocity.
Then it is observed an important activity fall. For example, using low space velocity
50% conversion is obtained at 125ºC with the fresh or catalysts used until 175ºC;
however, 50% conversion is obtained at ca. 175ºC with the catalyst previously used at
high space velocity (this experiment required ca. 275ºC for total conversion).
As Mn-HT is the most active catalyst, different reaction conditions were studied. In the
industrial effluents, VOCs are usually emitted in streams containing other components,
mainly water and carbon dioxide. Accordingly, the performance of Mn-HT catalyst was
evaluated at 125 ºC in the presence of different concentrations of these compounds
(Figure 2). In the case of the addition of water to the feed, increasing amounts of H2O %
up to 10 mol% were used. Interestingly, decreasing activity is not observed in any case.
Even at low H2O concentrations a slight positive effect is attained. Additionally, a
complete cycle was also run using 4 wt.% water in the feed (curve not shown).
Interestingly, total conversion is observed at only 150ºC which is lower than that
temperature at which total conversion is achieved in the absence of water (175ºC) (see
Figure 1).

60
55
50
45
40
60
55
50
45
40
4% H2O
10% H2O
Without
water
Without
water
water is
fed
water is
fed
0
2
4
6
8
10
12
0
2
4
6
8
10
Time on line (h)
Time on line (h)
ReTacimtioen Toenmplienraetu(rhe)(ºC)
Without
water
Without
water
60
55
50
45
40
60
55
50
45
40
2% CO2
10% H2O/2%CO2
CO2 is fed
water and CO2 are fed
0
2
4
6
8
10
12
0
2
4
6
8
10
Time on line (h)
Time on line (h)
Reaction Temperature (ºC)
Time on line (h)
Figure 2. Toluene conversion under different reaction feeds. Toluene concentration: 200 ppm.
Reaction temperature: 125 oC. GHSV: 6000 h-1.
Moreover, the addition of CO2 to the feed was not deleterious as only a subtle decrease
of conversion is observed. The simultaneous addition of water and carbon dioxide did
neither have an influence on the catalytic performance. Therefore, Mn-HT shows an
outstanding activity and stability in relevant conditions.
In order to explain the remarkable catalytic activity of these catalysts a detailed
characterization has been conducted. Firstly, textural characteristics of the samples were
analysed by N2 adsorption/desorption isotherm measurements. Data are shown in Figure
3. Unexpectedly, Mn-HT displays a remarkably lower SBET value than Mn-ST, 29 vs.
108 m2 g-1, ruling out the relevant role of this parameter on the catalyst performance.
The isotherms recorded for these materials correspond to Type IV isotherms usual for

mesoporous samples. In all isotherms the hysteresis loop is rather narrow, making the
denomination of the hysteresis type not easy. We use the adsorption branch of the
isotherm to determine the mesopore size distribution, eluding the nitrogen adsorption
artefact due to the steep desorption at P/P0 of 0.4-0.5. It is noteworthy that the mesopore
volume for the sample obtained under solvothermal conditions (Mn-ST) shows a
remarkable contribution of the pores lower than 10 nm, whereas this contribution is
marginal for the Mn oxide prepared under hydrothermal conditions.
b)
a)
1.6x10-3
1.4x10-3
1.2x10-3
160
140
120
1.0x10-3
8.0x10-4
6.0x10-4
4.0x10-4
100
80
60
40
20
0
Mn-ST
Mn-ST
Mn-HT
0.8
2.0x10-4
0.0
Mn-HT
1
10
100
0.0
0.2
0.4
0.6
1.0
Pore Width / nm
Relative pressure / P/P
0
Figure 3. a) N2 adsorption/desorption isotherms and b) pore-size distributions for manganese
catalysts.
Therefore, the synthesis under solvothermal conditions favours the formation of
intraparticle porosity, which is hardly created under hydrothermal conditions, as
corroborated later by microscopy analyses.
The main crystalline phase observed in both samples is Mn3O4 which is reported to be
the most active crystalline phase of the manganese oxides for total oxidation reactions
[14,38]. XRD patterns (Figure 4) permit to identify unambiguously that the two
manganese catalysts present the spinel (AB2O4, hausmannite) type structure belonging
to the tetragonal crystal system and I41/amd space group. The unit cell comprises 32

oxide ions and there are 8 Mn2+ ions occupying the tetrahedral (A) sites and 16 Mn3+
ions located in the octahedral (B) ones.
Mn-ST
Mn-HT
Mn-HT
Mn-ST
bc
a
f
d
e
10
20
30
40
50
60
70
80
o
2  ( )
2 θ (º)
Figure 4. XRD patterns of manganese oxide catalysts. In both cases the only phase detected is
Mn3O4. The small diffractions a, b, c, d, e and f correspond to (204), (312), (303), (402), (305) and
(413), respectively.
Average crystallite sizes calculated from XRD patterns applying Double-Voigt
approach reveals that the solvothermal synthesis produces materials with smaller
crystallite size than the hydrothermal one, 9 and 19 nm respectively, which is also in
line with the highest surface area of the Mn-ST sample. The higher particle size under
hydrothermal conditions can be due to the fact that ethylene glycol is a less polar
solvent than water, thus the solvation energy of the oxide/hydroxide phase is lower than
in pure water. This leads to a higher initial energy of the oxide/hydroxide before
nucleation and a lower energy barrier has to be overcome for crystallization.
Consequently, the nucleation rate is higher in solvent-rich solutions, preventing the
growth of bigger crystallites.

Moreover, structural analyses were made applying Rietveld analyses for both materials.
The excellent results of refinement (Table 2) indicate the great influence of the solvent
used during the synthesis procedure on the tetragonal cell parameters, a contraction of
the unit cell occurs in “a” direction and elongation in the direction of “c” axis take place
by increasing the mean particle size <d>. Both samples show a defect in its structure. In
fact, in the catalysts studied both cationic and anionic vacancies have been detected in
the spinel structure. The refinements were done assuming that the octahedral sites are
fully occupied by Mn3+ cations. SOF (site occupation factors) values for Mn2+ and
oxygen atoms (Table 2) point out a defect in the cationic and anionic sites, respectively,
which could explain the relevant activity of both catalysts. For the Mn-ST sample, 10%
deficiency approximately exists in cationic sites containing Mn2+ that can arise from
partial oxidation of Mn2+ to Mn3+ leading to formation of cationic vacancies. These are
mainly compensated by the deficiency in the anionic sites (~8%) through formation of
anionic vacancies and in fewer amounts via presence of interstitials (e.g., OH¯). The
calculated density for such sample from refinement data reveals a defect spinel
structure, with a value of 4.63 g cm-3 in contrast with 4.86 g cm-3 for the ideal spinel
structure (bulk material).
Table 2. Rietveld analyses.
Sample
Mn-ST
Mn3O4
Mn-HT
Mn3O4
Phase name
Space group (No.) I 41/a m d (141) I 41/a m d (141)
RBragg 0.315 0.234

a = b (Å)
c (Å)
5.765
9.425
1.14
5.763
9.444
1.16
c/√2 a
ρ (g/cm3)
V (Å3)
4.630
313.31
0.85
4.737
313.75
0.92
Rexp
Rp
0.80
0.88
Rwp
1.01
1.12
*GOF
1.18
1.21
**s.o.f. / Mn2+
**s.o.f. / Mn3+
**s.o.f. / O2-
0.89
0.93
1.00
1.00
0.92
0.97
*GOF: goodness of fit. It compares Rwp with Rexp. GOF=1 means the model is as good as possible. GOF
values less than 4 are acceptable.
**s.o.f.: site occupancy factor. s.o.f.=1 means every equivalent position for xyz is occupied by that atom
and s.o.f.<1 means some of the sites are vacant.
On the other hand, Mn-HT sample showed a less defective spinel structure with a
calculated density of 4.74 g cm-3, in line with its higher mean crystallite size. However,
a higher activity is observed for this catalyst. Significantly, 7% deficiency
approximately exists in cationic sites containing Mn2+ but, contrary to what happened
for the Mn-ST catalyst, defects are not mainly compensated by the deficiency in the
anionic sites (~2%) through formation of anionic vacancies but mainly by the presence
of higher amounts of interstitials, likely structural water or interstitial OH¯. Therefore,
the insertion into the crystalline structure of these species during the hydrothermal
synthesis could be the driving factor for the better catalytic performance of this material.
This fact could also explain the higher activity found in presence of water since the

possible loss of structural water at those temperatures needed to reach full toluene
conversion could be reduced under humid conditions.
The mixed oxidation state of Mn, typical of hausmannite spinel, was also confirmed by
the XPS results (Figure 5a) were the Mn 3s splitting value of 5.7 eV corresponds to the
mixed valence state of Mn2+ and Mn3+ with characteristic splitting values of 5.9 eV and
5.5 eV, respectively [40, 41]. The Mn 2p3/2 peak (Figure 5b) is centered at 641.3 eV and
the Mn 2p1/2 peak at 653.1 eV, with a splitting of 11.8 eV, which is in good agreement
with literature values [40, 41]. Moreover, deconvolution of Mn 2p spectral region
allowed calculating a Mn2+: Mn3+ ratio 1:1.8 which is in good agreement with the
stoichiometric value (1:2) but showing no differences among Mn-HT and Mn-ST
catalysts. The slight discrepancies between the Mn2+: Mn3+ ratio determined by XPS
and the bulk can be related to the presence of oxygen/cationic vacancies in the surface.
The asymmetrical O 1s signal can be deconvoluted into two components (Figure 5c).
The peak at 529.5 eV is attributed to the surface lattice oxygen O2- whereas the broad
shoulder at a higher BE (531.3 eV) assignment is more difficult, as this feature could
either be the result of hydroxyl groups or alternatively might be due to the presence of
oxygen vacancies, surface adsorbed oxygen, or carbonate species.

Mn 2p3/2
Mn 2p1/2
a)
b)
Mn-HT
Mn-HT
11.8 eV
5.7 eV
Mn 3s
Mn3+
Mn 3s (s)
Mn2+
Mn-ST
Mn-ST
100
95
90
85
80
75
660
655
650
645
B.E. (eV)
640
635
B.E. (eV)
c)
Mn-HT
Olatt
Oads
Mn-ST
536
534
532
530
528
526
524
B.E. (eV)
Figure 5. XPS spectra of the (a) Mn 3s region, (b) Mn 2p region and (c) O 1s region of the
manganese oxide catalysts.
The reactivity of the oxygen species was investigated by H2-TPR analysis. The TPR
profile of the Mn3O4 samples (Figure 6) shows a low-temperature band around 200-
300ºC, which can be attributed to the reduction of Mn3+ ions located in the tetrahedral
sites of the hausmannite lattice, and a wide signal in the 300–500ºC temperature range,
due to the hausmannite reduction (Mn3O4→MnO). Since MnO is the final reduction
o
product (terminating at ca. 700 C), the two peaks should reveal the reduction of two
different Mn3+-O bonds (to MnO) that are available in Mn3O4 [38]. The H2 consumption
(4.3 mmol g−1), related to the Mn3O4→MnO step, is in close agreement with the

theoretical data. Although both Mn-catalysts present a similar profile, some relevant
differences can be clearly observed in the low temperature reduction band. In this
temperature range, it is clearly seen that Mn3+ in the tetrahedral sites of the hausmannite
lattice are in two different chemical environments, being these species reduced around
200 ºC and 250 ºC, respectively. In line with Rietveld refined data, these two different
sites could be tentatively related to the presence of cationic vacancies compensated by
either interstitials or anionic vacancies, respectively, being the reducibility of the former
higher than that observed for the latter. Therefore, it can be observed in Figure 6 that
there is a higher proportion of highly reactive oxygen species associated to cationic
vacancies compensated by interstitials in the Mn-HT sample, which is also in line with
its higher activity.
In the present article it has been observed that the maximum of the main reduction
bands of manganese oxides in the TPR experiments arise at ca. 350ºC whereas toluene
can be converted on manganese oxide catalysts at reaction temperatures lower than
200ºC. However, if we see in detail the TPR profiles, we can notice that the onset
temperatures of the first reduction bands are observed at temperatures below 140ºC. In
any case, the conditions used in the TPR experiments are very different to those of our
toluene oxidation reaction and then the reduction of the catalyst in both environments
can be very different.

Mn-HT
Mn-ST
100
200
300
400
500
600
Temperature / oC
Figure 6. H2-TPR for the manganese oxide catalysts.
Electron microscopy observations show that the material named as Mn-ST exhibits a
predominant phase composed of nanoparticles of around 10-40 nm (Figures 7a, 7b and
S1). In all cases, the particles exhibited very good crystallinity in good agreement with
the XRD data obtained and they can be indexed assuming I41/amd symmetry and lattice
constants similar to the bulk Mn3O4, a ≈ 5.7 Å and c ≈ 9.4 Å. Figure 7b corresponds to a
representative nanoparticle observed along the [111] orientation; a closer look at the
atomic distribution is shown in Figure 7c with the model superimposed which allows
the identification of all atoms in the structure. In this model, Mn is represented in green
while O appears in red. Considering that all data was recorded using a High Angle
Annular Dark Field Detector (HAADF) the contrast observed is dependent on the
atomic number and thickness of the material. Therefore, the contrast variations observed
in the nanoparticles analyzed (Figure 7a) are owed to the presence of cavities/roughness
at the surface of the material, which is in line with the presence of pores lower than 5
nm, as observed by N2 adsorption data.
The material obtained by hydrothermal synthesis, Mn-HT, presents a similar shape and
a slightly higher particle size than the Mn-ST sample. Figure 7d depicts a low

magnification image confirming the similar morphology of the material obtained under
hydrothermal conditions composed of nanoparticles with various sizes. The crystallinity
for every particle observed was also found to be very good as was the case for Mn-ST,
see Figure 7e, in perfect agreement with the tetragonal symmetry I41/amd. The high-
resolution data, Figure 7e and 7f, allow the visualization of the atomic distribution of
the atoms along the [011] orientation, where the Mn atoms appear as green spheres
while O are in red. Surprisingly, cavities/roughness are again apparent, even in a higher
amount (see Fig. S1d-S1f), despite the fact that the presence of surface mesopores is
hardly observed by N2 adsorption data. Therefore, the internal cavities/vesicles instead
of surface mesopores are predominantly created during the hydrothermal synthesis,
where water molecules could be occluded leading to highly active oxygen surface
species, according to TPR data.
Figure 7. Cs-corrected STEM-HAADF data. a) Low-magnification image showing several Mn-ST
nanoparticles. b) High-resolution image of a Mn-ST isolated nanoparticle along the [111] zone axis.
c) A closer look of the structure of the Mn-ST nanoparticle with the model superimposed. d) Low-
magnification image of Mn-HT material. e) An isolated Mn3O4-HT nanoparticle along the [011]

orientation. f) Enlarged image of the particle shown in e) with the model superimposed. Green
spheres correspond to the Mn while O atoms appear as red spheres.
The chemical composition for both samples was corroborated by means of Energy
dispersive X-ray Spectroscopy (EDS) and Electron Energy Loss Spectroscopy (EELS),
see Figure 8, concluding that the Mn and O are equally distributed with no detectable
compositional variations in any case. Figure 8a corresponds to the EDS map recorded
over several nanoparticles proving the homogeneous composition for every particle.
The EELS spectrum image appears in Figure 8b. For this case, an isolated nanoparticle
was analyzed in order to observe any variations along the nanoparticle (if any). A
homogeneous distribution was obtained for both samples.

Figure 8. Study by TEM of Mn-HT catalyst. a) EDS mapping analysis of several nanoparticles
which confirms the presence of Mn (green map) and O (red map). b) EEL spectrum image of a Mn-
HT nanoparticle showing the perfect concordance between the Mn and O. c) EEL spectra of the
signals extracted corresponding to Mn and O edges respectively.
Considering the similarities between the products obtained by means of both synthetic
methods, we believe that the differences observed in their catalytic performance can be
explained in terms of the type of cavities created during the preparation method. As
already mentioned, all NPs show contrast variations in the images that are attributed to

the presence of a rough surface, as the chemical composition has been found to be
homogeneous with the same Mn:O ratio. With the intention of gaining more
information about this assumption SEM analyses were performed (Figure 9) showing
that for Mn-ST there are more interparticle voids, resulting in a higher surface area. For
both samples, Mn-ST and Mn-HT, the morphology of the materials is very similar
observing three main products sphere-like particles, elongated particles with plate
morphology and nano-fibers type formations, which could be attributed to manganese
hydrated phases [42], see Figure 9. Figures 9a and 9b present the SEM observations at
low and at high magnification for the Mn-ST material. It is possible to observe the
presence of spherical-like particles all over the sample. In the high-resolution SEM data
it is possible to visualize the channels opened to the surface of these nanoparticles,
which make them look like if they were agglomerated entities but from the STEM
analysis it can be inferred that they are single individual crystals.
After observation of the Mn-HT at low-magnification Figure 9c, it is apparent a more
heterogeneous material, with the plates and nano-fibers also present. These structures
were also observed in Mn-ST but into a less extent. By looking closer at the
nanoparticles formed by this synthetic method, Figure 9d, it is also possible to visualize
a similar morphology as observed for Mn-ST, but in this case less porosity appears,
making the material more compact exposing less open pores opened to the surface
giving as a result a lower adsorption capability.
Meanwhile, despite the similarities, for Mn-HT the particles are more compact and
exposing less accessible pores opened to the surface. By analysing several nanoparticles
and in line with XRD, H2-TPR and N2 adsorption data, the Mn-HT contains more
internal cavities where water molecules can be retained, leading to highly active surface

oxygen vacancies. These sites could act as reaction sites increasing the catalytic
performance.
Figure 9. SEM observations of Mn-ST, a) Low-magnification and b) High-magnification images.
And Mn-HT c) Low-magnification and d) High-magnification images.
Manganese oxide synthesized by an easy hydrothermal method show very high catalytic
activity and using low space velocities can completely remove toluene at very low
temperatures (150-175ºC). This performance does not vary and after several cycles the
same behaviour is observed. This stable behaviour is likely due to the fact that the
temperature this material has been heat-treated (180ºC) is higher than that required for
total conversion and then the structural and surface modifications after the catalytic tests
are not important. According to the high catalytic stability observed, the characteristics

of the used catalyst have changed little compared to the fresh catalyst. Deconvoluted
XPS spectra of the used Mn-HT catalyst (Figure S2) shows a slight shift of ca. 0.2 eV to
higher binding energies in the Mn2p region probably due to a slight oxidation of Mn3+
species to Mn4+. Moreover, the Mn2+ : Mn3+ ratio increases from 1:1.80 to 1:1.85
attributed to a slightly lower presence of oxygen/cationic vacancies in the surface. This
fact is also confirmed by the deconvolution in the O1s region which shows a value of
1.6 for the Olatt/Oads ratio compared with 1.4 value in the fresh catalyst. In the XRD
patterns (Figure S3) the same crystalline phase has been observed (Mn3O4) with just a
slight decrease in the width of the peaks and consequently a slight increase in the mean
crystallite size after use. It has been also observed a small fall in the surface area for the
spent catalyst (from 29 to 25 m2/g). The electron microscopy observations prove the
good crystallinity of the reacted material, same of the as-synthesized one, and also the
chemical analysis performed by means of EEL spectral imaging shows a homogeneous
distribution of Mn and O after reaction (Figure S4). From electron microscopy
perspective, it is difficult to assure that any significant transformation took place after
use, as both materials were very similar in terms of composition, particle size and
structure. Nevertheless, we cannot discard the possibility of a slight increment of the
initial good crystallinity in accordance with the XRD observations, as a result, possibly,
of the thermal treatments that the catalyst underwent. It is important to keep in mind
that electron microscopy is a more local technique while XRD gives a more averaged
information of the materials (Figure S5).
Interestingly, the hydrothermally prepared catalyst presents a catalytic activity higher
than that synthesized by a solvothermal method which presents much higher surface
area and lower crystallite size. The extremely high catalytic activity of the Mn3O4
catalyst prepared by the hydrothermal method can be due to the presence of oxygen

surface species which are highly reactive. The presence of Mn3O4 nanoparticles with a
high concentration of Mn3+ at the tetragonal distorted cationic sites does likely play an
important role. Interestingly, there are internal cavities/vesicles in the Mn3O4 structure
which induces the presence of structural water.
4. Conclusions
Highly active manganese oxide catalyst has been synthesised using a simple and green
synthesis method employing a hydrothermal procedure. A catalyst with remarkably
higher activity is obtained by this method in spite of the fact that the surface area of the
hydrothermal catalyst is ca. 4 times lower than that of the catalyst prepared by a
solvothermal method. This catalyst has been tested in the total oxidation of toluene
showing total conversion at only 175ºC in the absence of water and, more remarkably,
at 150ºC using low concentrations of water (4 mol%). This performance hardly varies
and after 6 catalytic cycles similar results are observed. In view of the characterization
of these catalysts and in line with literature data, the outstanding catalytic activity of the
Mn3O4 catalyst prepared by the hydrothermal method can be ascribed to the presence of
highly reactive oxygen surface species. This species are likely promoted due to the
existence of Mn3O4 nanoparticles with a relevant amount of Mn3+ at the tetragonal
distorted cationic sites, likely compensated by structural water at the anionic sites.
Interestingly, the presence of structural water seems to be a consequence of the
formation of internal cavities/vesicles in the Mn3O4 crystalline structure.
Acknowledgements
AM and YZ acknowledge The Centre for High-resolution Electron Microscopy
(CħEM), supported by SPST of ShanghaiTech University under contract No.

EM02161943. Electron microscopy analyses were performed at Centre for High-
Resolution Electron Microscopy (CћEM), ShanghaiTech University. AM acknowledges
the National Natural Science Foundation of China through the projects NSFC NFSC-
21850410448 and NSFC- 21835002. Authors from UCM thank Santander-UCM
institution (PR41/17-20951 project) and MINECO (MAT2017-84118-C2-2-R project)
and UCM CAI centers (XRD and EM). TG and JML thank Regional Aragon
Government (DGA) under the research groups support programme. Some of the STEM
analyses were conducted at the Laboratorio de Microscopias Avanzadas, Universidad de
Zaragoza, Spain. R. A. acknowledges support from Spanish MINECO grant MAT2016-
79776-P (AEI/FEDER, UE). Authors from UV thank MINECO (MAT2017-84118-C2-
1-R project) for funding.
References
1
2
3
L. F. Liotta, Appl. Catal. B: Environ. 100(3-4) (2010) 403–412.
Z. Chen, J. Ma, R. Zhou, Appl. Surf. Sci. 465 (2019) 15–22.
R. Balzer, L. F. D. Probst, V. Drago, W. H. Schreiner and H. V. Fajardo, Braz. J.
Chem. Eng. 31 (2014) 757-769.
4
5
6
7
B. Solsona, T.E. Davies, T. García, I. Vázquez, A. Dejoz and S.H. Taylor, Appl.
Catal. B: Environ. 84 (2008) 176–184.
T. García, D. Sellick, F. Varela, I. Vázquez, A. Dejoz, S. Agouram, S. H. Taylor,
and B. Solsona, Appl. Catal. A: Gen. 450 (2013) 169–177.
M. Popova, A. Ristic, M. Mazaj, D. Maučec, M. Dimitrov and N. Novak Tušar.
ChemCatChem. 6 (2014) 271–277.
M. Popova, Á. Szegedi, Z. Cherkezova-Zheleva, A. Dimitrova and I. Mitov. Appl.
Catal. A: Gen. 381 (2010) 26–35.

8
9
S. Boycheva, D. Zgureva, M. Václavíková, Y. Kalvachev, H. Lazarova and M.
Popova. J. Hazardous Materials, 361 (2019) 374-382.
J. Luo, Q. Zhang, A. Huang and S. L. Suib, Micropor. Mesopor. Mater. 35-36
(2000) 209–217.
10 C. Lahousse, A. Bernier, P. Grange, B. Delmon, P. Papaefthimiou, T. Ioannides and
X. Verykios, J. Catal. 178 (1998) 214–225.
11 W. Si, Y. Wang, Y. Peng, X. Li, K. Lim, J. Li, Chem. Commun. 51 (2015) 14977-
14979.
12 S. S. T. Bastos, J. J. M. Orfao, M. M. A. Freitas, M. F. R. Pereira and J. L.
Figueiredo, Appl. Catal. B: Environ. 93 (2009) 30–37.
13 S. S. T. Bastos, S. A. C. Carabineiro, J. J. M. Órfao, M. F. R. Pereira, J. J. Delgado
and J. L. Figueiredo, Catal. Today, 180 (2012) 148–154.
14 S. C. Kim and W. G. Shim, Appl. Catal. B: Environ., 98 (2010) 180–185.
15 H.Pan, Y. Jian, C. Chen, C. He, Z. Hao, Z. Shen, H. Liu, Environ. Sci. Technol. 51
(2017) 6288-6297.
16 H. Sun, S. Chen, P. Wang and X. Quan, Chem. Eng. J., 178 (2011) 191–196.
17 Y. S. Ding, X. Shen, S. Sithambaram, S. Gomez, R. Kumar, M. B. Vincent and S.
L. Suib, Chem. Mater. 17 (2005) 5382–5389.
18 Z. Sihaib, F. Puleo, J. M. Garcia-Vargas, L. Retailleau, C. Descorme, L. F. Liotta, J.
L. Valverde, S. Gil and A. Giroir-Fendler, Appl. Catal. B: Environ. 209 (2017) 689-
700.
19 J. Chen, X. Chen, W. Xu, Z. Xu, J. Chen, H. Jia, J. Chen, Chem. Eng. J. 330 (2017)
281-293.
20 Y. Wang, L. Zhang, L. Guo, ACS Appl. Nano Mater. 1 (2018) 1066−1075.