Total oxidation of naphthalene using bulk manganese oxide catalysts

Article abstract of DOI:10.1016/j.apcata.2012.10.029

Several Mn₂O₃ catalysts have been synthesized using different preparation methods and tested for the total oxidation of naphthalene, a model polycyclic aromatic compound. The catalysts have been characterized by several physico-chemical techniques such as XRD, TPR, XPS, EDX and TEM. The surface area of the catalyst seems to be of paramount importance, since the mass normalized activity of catalysts increases as the surface area of the Mn ₂O₃ catalysts increases. Consequently, a high surface area ordered mesoporous Mn₂O₃ catalyst, obtained through a nanocasting route using mesoporous KIT-6 silica as a hard template, was the most efficient catalyst for the deep oxidation of naphthalene. In addition, this catalyst also exerts the highest surface area normalized catalytic activity, which can be related to the highest reducibility and mobility of its lattice oxygen. It was also observed that the Mn₂O₃ crystalline phase presents a higher intrinsic activity than MnO₂.

Full text of DOI:10.1016/j.apcata.2012.10.029

Applied Catalysis A: General 450 (2013) 169–177
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Total oxidation of naphthalene using bulk manganese oxide catalysts
Tomas Garciaa, David Sellickb, Francisco Varelab, Isabel Vázquezc, Ana Dejozc, Said
Agouramd, Stuart H. Taylorb,∗, Benjamin Solsonac,∗
a Instituto de Carboquímica (ICB-CSIC), C/Miguel Luesma 4, 50018 Zaragoza, Spain
b Cardiff Catalysis Institute, Cardiff University, School of Chemistry, Main Building, Park Place, Cardiff, CF10 3AT, United Kingdom
c Departament d’Enginyeria Química, Universitat de València, Avda. de la Universitat s/n, 46100 Burjassot, Valencia, Spain
d Departament de Física Aplicada i Electromagnetisme, Universitat de València, C/Dr. Moliner 50, 46100 Burjassot, Valencia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 July 2012
Received in revised form 10 October 2012
Accepted 11 October 2012
Available online 9 November 2012
Several Mn2O3 catalysts have been synthesized using different preparation methods and tested for the
total oxidation of naphthalene, a model polycyclic aromatic compound. The catalysts have been charac-
terized by several physico-chemical techniques such as XRD, TPR, XPS, EDX and TEM. The surface area
of the catalyst seems to be of paramount importance, since the mass normalized activity of catalysts
increases as the surface area of the Mn2O3 catalysts increases. Consequently, a high surface area ordered
mesoporous Mn2O3 catalyst, obtained through a nanocasting route using mesoporous KIT-6 silica as a
hard template, was the most efficient catalyst for the deep oxidation of naphthalene. In addition, this
catalyst also exerts the highest surface area normalized catalytic activity, which can be related to the
highest reducibility and mobility of its lattice oxygen. It was also observed that the Mn2O3 crystalline
phase presents a higher intrinsic activity than MnO2.
Keywords:
Naphthalene
Polycyclic aromatic compound
PAH
Manganese oxide
Total oxidation
Nanocasting
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
Only more recently, there have been specific studies focussing on
the total oxidation of PAHs. In most of these articles naphthalene
Polycyclic aromatic hydrocarbons (PAHs) are a specific type of
volatile organic compound (VOC) formed by at least two fused aro-
matic rings, which are emitted to the atmosphere as a result of the
combustion of a wide range of hydrocarbons and other fossil fuels
in various applications, such as vehicles and industrial exhausts
[1]. Examples of common PAHs are naphthalene, phenanthrene and
pyrene. Apart from the deleterious effects of VOCs in general, such
as ozone depletion and formation of ground level ozone, PAHs are
known to be highly carcinogenic and their emission is related to
other serious health hazards [1,2]. Hence, abatement of PAH emis-
sions is important, and one of the most efficient and promising
methods is catalytic total oxidation. A comprehensive review has
been recently published by Ntainjua and Taylor [3], showing that
the majority of studies related to the catalytic combustion of PAHs
are concentrated in the last few years.
has been the hydrocarbon selected as the feed, since it is the sim-
plest and one of the least toxic PAHs, and its catalytic behaviour can
be extrapolated to more complex PAHs.
In one of the earlier studies on the catalytic oxidation of naph-
thalene Zhang et al. [7] employed a series of metal catalysts
supported on ␥-Al2O3. The most effective catalysts proved to be
those based on precious metals, especially those containing Pt.
Thus, initial studies focused on precious metals, showing that they
were generally more active than metal oxide catalysts [7,8]. Ntain-
jua et al. [9] found that, similarly to other VOCs such as alkanes and
monoaromatics, Pt/␥-Al2O3 catalytic performance can be enhanced
by means of the addition of vanadium oxide. In another study, it
was found that the alumina support was not the best support for
platinum, with the best results obtained using silica as a support
[10].
Most of the first catalytic studies dealing with the oxidation of
PAHs were conducted using gaseous mixtures of VOCs, in which a
model PAH was included [4–6]. Additionally, in some of these stud-
ies CO was also co-fed together with the range of different VOCs.
Generally, noble metals are more active than metal oxides for
total oxidation of VOCs, although the latter are cheaper and more
resistant to poisoning by sulphur [11]. Hence, it is desirable to
identify metal oxide based catalysts which show activity compa-
rable to noble metal-based catalysts. In previous work published
by our group [12] Co3O4, Mn2O3, ZnO, CuO, Fe2O3 and CeO2 cata-
lysts prepared by a standard precipitation method were tested for
the total oxidation of naphthalene. Among these catalysts Mn2O3
was the most active of the whole series, even more so than Co3O4
∗
Corresponding authors. Tel.: +34 963543735; fax: +34 963544898.
E-mail addresses: taylorsh@cardiff.ac.uk (S.H. Taylor), benjamin.solsona@uv.es
(B. Solsona).
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.10.029

170
T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
which is recognized as being highly active for VOC oxidation.
Therefore, it seems that Mn2O3 could be a potential catalyst for
naphthalene abatement and it would be interesting to explore the
performance of catalysts with an expanded structure and high sur-
face area since these parameters have been demonstrated to be of
paramount importance for other catalytic systems used for the total
oxidation of naphthalene [3]. One preparation method for obtain-
ing these characteristics is the “nanocasting route” that allows
the preparation of highly ordered metal oxides with high surface
areas [13–15]. Mesoporous manganese oxide replicas have been
synthesized previously using the nanocasting method employing
mesoporous silica as a hard template [16,17].
Manganese oxide catalysts have demonstrated appreciable
activity for the oxidation of a range of VOCs, such as alkanes
[18–20], alcohols [21–23], monoaromatics [24,25] or even VOC
mixtures [26]. The high activity of manganese oxides has often
been explained in terms of the characteristics of the oxygen species.
Thus Figueiredo et al. [22], studying the total oxidation of ethanol
on manganese oxides, synthesized by exotemplating using carbon,
found a strong correlation between the catalytic activity and the
ability of the catalyst to donate lattice oxygen. Baldi et al. [18]
proposed that the oxygen vacancies, present in manganese oxides,
were the species responsible for the good catalytic performance. In
other work, Kim and Shim [27] also related the high catalytic activ-
ity for the oxidation of benzene and toluene to the oxygen mobility
of a manganese oxide catalyst.
of tetraethyl orthosilicate (TEOS, 98%, Sigma–Aldrich) were added
and stirred for 24 h at the same temperature. The mixture was aged
for 24 h at 80 ◦C, under static conditions. For hydrothermal treat-
ment the gel was then transferred into a stainless-steel autoclave
and heated at 130 ◦C for 72 h under static conditions. The white
solid formed was recovered by filtration, washed with deionized
water and dried for 24 h in an oven. Thereafter, the products were
calcined in flowing air at 600 ◦C for 6 h at a heating rate of 10 ◦C/min
[28,29].
The manganese oxide was prepared by dispersing the siliceous
KIT-6 in ethanol with manganese nitrate. After 30 min of stir-
ring, the ethanol was removed by evaporation through heating the
mixture for 16 h at 120 ◦C in an oven. The resulting powder was
heated in a ceramic crucible in an oven at 350 ◦C for 6 h to com-
pletely decompose the nitrate species. The impregnation step was
repeated with a further 5.0 mL of the metal salt in ethanol solution
in order to achieve higher manganese loadings. After evaporation
of the solvent, the resulting material was calcined at 500 (sample
called Mn3-500) and 600 ◦C (Mn3-600) for 6 h. Finally, the silica
template was removed by treatment using 2 M NaOH aqueous solu-
tion at 80 ◦C. The NaOH etching of the silica template was repeated
3 times, each time a fresh 5 mL portion of NaOH solution was used
for 2 h. The MnOx catalyst was recovered by centrifugation, washed
with water and finally dried at 200 ◦C.
According to the chemical analysis it was determined that the
Mn3-500 and Mn3-600 samples contained residual silica, although
to a relatively low extent (1.9 wt.%). In order to remove the residual
silica 1 g of Mn3-600 was placed in a beaker with 100 mL of 2 M
NaOH solution and stirred at 80 ◦C for 48 h. As a consequence of
evaporation distilled water was continuously added to keep the
volume of solution constant. The resulting material was recovered
by centrifugation, washed again with abundant water and finally
dried at 200 ◦C. This catalyst was denoted as Mn3-600B.
Against this background, and considering the fact that not all
VOCs behave in the same way, as catalysts which are very efficient
for the oxidation of one type of VOC can present low activity for
another subset, it is interesting to study further the total oxidation
of naphthalene using manganese oxide catalysts. In the present
study a range of Mn2O3 oxides have been synthesized, among
them two manganese oxides prepared by a nanocasting procedure,
and tested for total oxidation of naphthalene. The influence of the
preparation method on catalytic activity, as well as the calcination
temperature of the manganese oxide, has also been studied.
2.2. Catalyst characterization
Catalyst samples were characterized by N2 adsorption at
−196 ◦C, using a Micromeritics ASAP 2020 apparatus. Samples were
degassed at 150 ◦C prior to analysis. From these data, the follow-
ing textural parameters were determined: Total pore volumes were
calculated using the adsorbed volume at a relative pressure of 0.97.
2. Experimental
2.1. Preparation of catalysts
Multipoint Brunauer–Emmet–Teller (BET) surface area [30], (SBET
)
Series 1: The first series of manganese oxides can be considered
as a reference series, as these samples were prepared by calcining
manganese nitrate in static air at 300, 500 and 600 ◦C for 3 h. These
catalysts have been named as Mn1-300, Mn1-500 and Mn1-600.
Series 2: The second series involved the preparation using oxalic
acid acting as an organic template. Manganese oxide catalysts were
prepared through the evaporation at 60 ◦C of a stirred ethanol solu-
tion of manganese nitrate (Mn(NO3)2·6H2O, Sigma–Aldrich). Oxalic
acid was added to the solution with an oxalic acid/Mn ratio (molar)
in all cases of 1:1 for consistency. The solids were dried for 16 h
at 120 ◦C, and finally calcined in static air for 3 h at 300, 500 and
600 ◦C. The catalysts were named as Mn2-300, Mn2-500 and Mn2-
600 respectively.
Series 3: The third series of manganese oxides were prepared by
a nanocasting procedure using mesoporous silica KIT-6 as a hard
template, which presents a three-dimensional Ia3d cubic arrange-
ment of pores. Such a pore structure is known to yield mesoporous
morphologies, with highly crystalline walls and good thermal sta-
bility, after the silica template is removed. KIT-6 was synthesized in
acidic conditions using a mixture of Pluronic P123 triblock copoly-
mer (EO20PO70EO20) as template and butanol. 6 g of P123 were
added to a solution of 220 g of distilled water and 12 g of concen-
trated HCl (35%). The mixture was stirred for 6 h at 35 ◦C and then
6 g of butanol were added and allowed to stir for 1 h. Finally, 12.48 g
was estimated from the relative pressure range from 0.05–0.25.
The pore size distribution of the mesoporous materials was ana-
lyzed using the Barrett–Joyner–Halenda (BJH) method using the
desorption branch of the N2 adsorption isotherms [31].
Powder X-ray diffraction (XRD) was used to identify the crys-
talline phases present in the catalysts. A Bruker D8 Advance
diffractometer with monochromatic Cu K␣ source operated at
40 KV and 40 mA was used. The experimental patterns were cal-
ibrated against a silicon standard and the crystalline phases were
identified by matching the experimental patterns to the JCPDS pow-
der diffraction files [32].
Temperature Programmed Reduction (TPR) studies were per-
formed using a Micromeritics Autochem II instrument with a
thermal conductivity detector. In all the experiments, 10% vol. H2
in Ar, at a constant flow rate of 50 cm3/min, was used as reducing
gas. A temperature range of 30–800 ◦C and a constant heating rate
of 10 ◦C/min were used. In each experiment, 0.30 g of powdered
catalyst was analyzed.
XPS measurements were made on an Omicron ESCA+ photo-
electron spectrometer using a non-monochromatized Mg K␣ X-ray
source (hꢀ = 1253.6 eV). An analyser pass energy of 50 eV was used
for survey scans and 20 eV for detailed scans. Binding energies are
referenced to the C 1s peak from adventitious carbonaceous con-
tamination, assumed to have a binding energy of 284.5 eV.

T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
171
Table 1
Physicochemical and catalytic properties of manganese oxide catalysts.
Catalyst
SBET (m2/g) Phases by XRD
T10a (◦C) T50a (◦C) T90a (◦C)
Mn1-600
Mn1-500
Mn1-300
Mn1-300
Mn1-500
Mn1-600
Mn2-300
Mn2-500
Mn2-600
Mn3-500
Mn3-600
Mn3-600B
2
<1
<1
42
16
7
108
78
25
MnO2
MnO2, Mn2O3
Mn2O3
MnO2 < Mn2O3
Mn2O3
Mn2O3
280
285
280
235
245
250
205
205
225
310
315
310
260
265
270
230
215
255
> 315
> 315
> 315
280
285
290
260
250
270
MnO2 ꢀ Mn2O3
Mn2O3
Mn2O3
a
T10, T50 and T90, reaction temperatures for 10, 50 and 90% conversions to CO2,
respectively.
20
30
40
50
60
70
2θ
Morphological and structural characterization of the samples
was performed by Transmission Electron Microscopy (TEM), High
Resolution TEM (HRTEM) and Selected Area Electron Diffraction
(SAED) by using a FEI Field Emission Gun (FEG) TECNAI G2 F20 S-
TWIN microscope operated at 200 kV. The synthesized manganese
oxide powder samples were treated by sonicating in absolute
ethanol for few minutes, and a drop of the resulting suspension
was deposited onto a holey-carbon film supported on a copper grid,
which was subsequently dried.
Mn2-600
Mn2-500
Mn2-300
2.3. Catalyst performance
20
30
40
50
60
70
Catalytic activity tests for naphthalene oxidation were carried
out in a fixed bed laboratory micro-reactor. Blank tests were con-
ducted by passing 450 vppm of naphthalene (Fluka >99%) through
an empty reactor that was heated from 100 ◦C to 350 ◦C at a rate of
10 ◦C/min. Catalysts (pelletized to 0.1–0.2 mm particle size) were
tested using a 3/8” o.d quartz tube as the reactor. The reaction
feed consisted in all cases of ca. 450 vppm naphthalene in a mix-
ture of 20 vol. % oxygen and 80 vol. % helium. A total flow rate of
50 cm3/min was used and the catalysts occupied a constant vol-
ume, giving a GHSV of ca. 75,000 h−1 for all the catalysts. Analyses
were performed by an on-line gas chromatograph with thermal
conductivity and flame ionization detectors. The catalytic activ-
ity was measured over the temperature range of 100–350 ◦C, in
incremental steps of 25 ◦C, and temperatures were controlled by
a thermocouple, placed in the catalyst bed. Data were obtained at
each temperature after naphthalene adsorption equilibrium was
accomplished and steady state activity attained. Four consistent
analyses were made at each temperature and average values were
calculated. The reaction temperature was increased and the same
procedure followed to determine each data point.
2θ
Mn3-600
Mn3-500
20
30
40
50
60
70
2θ / degrees
Fig. 1. XRD patterns for the different manganese oxides prepared through different
preparation procedures and calcination temperatures. Crystalline phases: (᭹) ␣-
MnO2, (ꢀ) Mn2O3.
The estimated crystallite sizes by the Scherrer equation [33] vary,
as expected, inversely to the surface area of the catalysts.
The reduction profiles (TPR experiments under H2 flow) are
shown in Fig. 2. Two intense bands are observed for all these
catalysts, although some other reduction features of low intensity
can also be observed. The TPR profiles for these catalysts are typical
of the Mn2O3 phase with two main features which correspond to
Mn2O3 → Mn3O4 and Mn3O4 → MnO transitions [34]. Total hydro-
gen consumptions together with those obtained for each transition
are shown in Table 2. It can be observed that these values are quite
close to their theoretical H2 consumption, which are 2.47 and
3.83 mmol/g for Mn2O3 → Mn3O4 and Mn3O4 → MnO transitions,
respectively. These results indicate that all catalysts present H2
consumption in the 6.0–6.8 mmol/gcat range. These values increase
in the sequence of Mn3-600 > Mn2-600 > Mn1-600 > Mn3-600B.
This fact is in agreement with XPS data as it can be seen in Table 3,
where the Mn2+/Mn3+ ratio decreases in the same order.
3. Results and discussion
3.1. Influence of the preparation method on the catalytic
performance (catalysts calcined at 600 ◦C)
The surface areas of the catalysts calcined at 600 ◦C are shown in
Table 1. The catalyst prepared by simple calcination of manganese
nitrate, Mn1-600, has a surface area lower than 1 m2/g, whereas
the one prepared using oxalate, Mn2-600, has 7 m2/g. The sample
prepared by a nanocasting route, Mn3-600, presents a much higher
surface area of 78 m2/g, and the catalyst post-treated to remove
silica further, Mn3-600B, has a reduced surface area of 25 m2/g.
Fig. 1 shows the XRD patterns for the manganese oxide catalysts
synthesized in this work and a comparative figure of those calcined
at 600 ◦C is shown in the supplementary Information (Fig. S1). In all
the samples calcined at 600 ◦C only diffraction peaks corresponding
to Mn2O3 (JCPDS: 24-0508) have been detected although the pres-
ence of some amorphous phase cannot be completely ruled out.
The maximum of the main peak is located between 447 and
489 ◦C (Fig. 2) and there is a relationship indicating that the higher
the surface area of the catalyst the lower the temperature for the
reduction maximum. Thus the maximum for Mn3-600 appears at

172
T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
Table 2
Temperature programmed reduction results for Mn oxide catalysts calcined at 600 ◦Ca.
Tonset (◦C)
Tmax (◦C)
Mn2O3 → Mn3O4
Mn3O4 → MnO
H2
(mmolH /gcat
)
(mmolH /gcat
)
consumption
2
2
(mmolH /gcat
)
2
Mn1-600
Mn2-600
Mn3-600
Mn3-600B
265
260
158
198
489
480
447
467
2.8
2.4
2.4
2.9
3.9
3.8
3.6
3.9
6.7
6.2
6.0
6.8
a
Theoretical consumption values are 2.47 and 3.83 mmol/g for Mn2O3 → Mn3O4 and Mn3O4 → MnO transitions, respectively.
Table 3
XPS characteristics of Mn oxide catalysts calcined at 600 ◦C.
Catalyst
O 1s
OI
Mn 2p
OII
OII/OI peak ratios
2p1/2 (eV)
2p3/2 (eV)
Mn2+/Mn3+
Peak ratios
Mn3+/Mn4+
Peak ratios
BE (eV)
at. (%)
BE (eV)
at. (%)
Mn1-600
Mn2-600
Mn3-600
Mn3-600B
529.75
529.60
529.25
529.55
61
66
67
66
531.4
531.1
531.0
531.5
26
27
24
24
0.43
0.41
0.36
0.36
653.30
653.25
653.30
653.80
641.60
641.55
641.60
641.85
0.21
0.29
0.30
0.09
0.96
1.09
1.49
0.64
447 ◦C and at 489 ◦C for Mn1-600. Table 2 also shows the onset
temperature for the reduction of the catalysts and a marked varia-
tion of the onset of reduction with the variation of surface area is
observed. Onset temperature has been estimated from the initial H2
consumption rates before the occurrence of phase transformation
(where the initial H2 consumption of the first reduction band of the
catalyst is lower than 25% [35]). Initial H2 consumption has been
plotted vs 1000/T and linearly fitted. The Tonset is then calculated by
extrapolating that curve for a H2 consumption value of 0. Thus the
reduction of Mn1-600 begins at 265 ◦C, whereas for Mn3-600 the
onset was only 158 ◦C. The first peak for Mn2-600 and Mn3-600B
begins at 260 and 198 ◦C respectively.
Characteristics of the surface of the manganese oxide catalysts
have been determined by means of XPS survey spectra. The rela-
tive contribution of Mn2+, Mn3+, and Mn4+ can be calculated from
the Mn 2p XPS spectra. Accordingly, the asymmetrical signal at
BE = ca. 641 eV for the manganese oxide, can be decomposed to
three components at BE = 640.6, 641.6, and 642.8 eV. The com-
ponents were attributable to the surface Mn2+, Mn3+, and Mn4+
species, respectively. A quantitative analysis on the Mn 2p3/2 XPS
spectra of the samples let us determine the surface Mn3+/Mn4+ as
well as Mn2+/Mn3+ ratios. Table 3 shows that Mn3-600 and Mn2-
600 catalysts have the most reduced surface. Fig. 3a shows the XPS
spectra of Mn 2p3/2 for the manganese oxide catalysts calcined
at 600 ◦C. In all cases a band at 641–643 eV has been observed.
According to reference [27] the binding energies for Mn 2p3/2
(Mn 2p1/2) for MnO2, Mn2O3 and Mn3O4 are located in the 642.6
(653.2), 642.2 (654.2) and 641.0 (654.4) eV regions respectively. For
our catalysts, in spite of the fact that only one unique crystalline
phase, Mn2O3, has been detected by XRD, the position of the XPS
peaks seems to be located between the binding energies typical
of Mn2O3 and Mn3O4. These observations most likely indicate a
lower oxidation state of manganese on the surface than in the bulk
of these catalysts, which can be related to the presence of oxygen
vacancies.
In the case of photoemission from oxygen, the O 1s signal shows
three different surface oxygen species as can be seen from Fig. 3B,
where the asymmetrical O 1s signal was deconvoluted to three
components. The low binding energy peak (OI: 529.2–529.8 eV),
which is ascribed to lattice oxygen, the medium binding energy
peak (OII: 531.0–531.5 eV), assigned to surface adsorbed oxygen,
OH groups, carbonate groups (a small C peak is also detected at
approximately 288.5 eV) and oxygen vacancies; and finally the high
binding energy peak (ca. 533.0 eV), likely to be associated with
adsorbed molecular water [36], (OIII).
As shown in Table 3, the relative amount of OII species,
expressed as OII/OI ratio, is the lowest in Mn3-600 (the catalyst
with the highest surface area) and the highest in Mn1-600 (the cat-
alyst with the lowest surface area). Additionally it can be seen that
the position of the OI binding energy differs depending on the cat-
alyst. Thus the binding energy for OI is the lowest for Mn3-600
and varies according to the order: Mn3-600 < Mn3-600B < Mn2-
600 < Mn1-600. The energy shift occurred when the surface area
increases is similar to that observed when potassium is added to
the Mn2O3 [37].
To confirm the crystallinity of the manganese oxide samples
transmission electron microscopy combined with SAED were car-
ried out. TEM images and SAED analysis of Mn1-600 and Mn2-600
catalysts are shown in Fig. 4. Both samples have similar mor-
phology, which consists of an array of connected grains with
sizes ranging from 20 to 100 nm. Also it is of interest to empha-
size that some areas of the material consist of domains of ca.
Mn3-600B
Mn3-600
Mn2-600
Mn1-600
100 200 300 400 500 600
Temperature / ºC
Fig. 2. Temperature programmed reduction (TPR-H2) profiles for manganese oxides
calcined at 600 ◦C.

T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
173
Fig. 3. Mn 2p XPS (a) and O 1s XPS spectra (b) of representative manganese oxide catalysts.
1 ␮m2 (Fig. 4a and d). Selected area electron diffraction pat-
terns recorded from both samples (Fig. 4b and e) show at least
six diffraction rings. The measured interplanar distance deter-
mined from the electron diffraction pattern from the centre to
the outer ring are as follows: 3.8639, 2.7107, 2.3663, 1.9342,
other phases were not present. Direct measurement of the spacing
in between the crystal fringes for the Mn1-600 sample visualized
in the HRTEM micrograph is 2.73 A (Fig. 4c), which corresponds to
˚
the (2 2 2) lattice spacing of Mn2O3. Fig. 4f shows, for the Mn2-600
˚
catalyst, a lattice spacing of 3.85 A between adjacent lattice planes
˚
which corresponds to the (2 1 1) planes of Mn2O3.
1.7409, 1.6772 and 1.4558 A respectively. These correspond to
Fig. 5 shows representative TEM micrographs of the Mn3-600
sample and HRTEM images. Additionally the pore size distribution
of this catalyst has also been included. TEM micrographs obtained
for Mn3-600 indicate that the structure of this material consists of
an ordered network pore structure. TEM images with higher mag-
nifications of the mesoporous Mn2O3 revealed that the material
was made of ordered crystalline nanoparticles with sizes ran-
the planes (2 1 1), (2 2 2), (4 0 0), (2 2 4), (6 2 1), (0 4 4) and (1 4 5)
which are indexed to orthorhombic Mn2O3 (JCPDS: 24-0508)
with space group Pcab (61), and are in good agreement with the
XRD measurements (see Fig. 4b for SAED diffraction rings assign-
ment).
The most interesting finding from XRD and electron diffraction
analyses is that characteristic features arising from impurities or
Fig. 4. Representative TEM micrographs (a, d) and their corresponding SAED patterns (b, e) and HRTEM images (c, f) for Mn1-600 (a, b, c) and Mn2-600 (d, e, f) samples.

174
T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
Fig. 5. Representative TEM micrographs (a, b) and HRTEM images (d, e) for Mn3-600 catalyst. The pore size distribution for Mn3-600 (c) and a TEM image of Mn3-600B
catalyst (f) are also included.
ging from 5 to 7 nm. In this sample nanoparticles are connected
by nanobridges, with the nanoparticles being in a body centred
arrangement, although the silica template is not perfectly repli-
cated since open pores can be clearly observed at the edges of
the nanoparticles. However, it can be concluded that the ordered
arrangements of nanoparticle domains are evident, and the ordered
structures, which are prevalent, coexist with low amounts of amor-
phous domains. From selected area electron diffraction patterns
(not shown here) only orthorhombic Mn2O3 was detected. Fig. 5e
shows the pore size distribution of Mn3-600. A broad pore size
distribution between 3 and 10 nm was obtained. This pore size dis-
tribution shows that although the replication of the mesoporous
silica has been accomplished, as can also be corroborated by the
TEM images, the formation of microporous bridges is not fully
achieved, leading to either wider pores in the nanoparticle frame-
work or open pores at the nanoparticle edges [38]. Finally, the TEM
images of Mn3-600B catalyst (Fig. 5f) were comprised of compact
blocks in which the ordered structure has been completely lost,
which is in agreement with the drastic decrease in the surface area
compared to Mn3-600.
Fig. 6 shows the variation of naphthalene conversion to CO2 as
a function of reaction temperature for the manganese oxide cata-
lysts calcined at 600 ◦C. It can be seen that the catalyst prepared
by nanocasting, Mn3-600, presents the highest catalytic activity.
In general, the catalytic activity increases with the surface area
of the catalysts. Thus 50% conversion to CO2 is obtained at 215 ◦C
with Mn3-600 whereas for Mn3-600B, Mn2-600 and Mn1-600 the
same conversion is attained at higher temperatures of 255, 270 and
315 ◦C respectively.
The objective of the process is not only the total removal of
naphthalene, but also its total transformation into CO2, as other
compounds more toxic than naphthalene could be formed. In these
experiments, apart from CO2, some other reaction compounds
have been detected in the reactor effluent stream. The main addi-
tional products detected were naphthenol, dimethyl phthalate and
benzaldehyde. Fig. 7 shows the variation of the selectivity to CO2
obtained over the various manganese oxide catalysts as a function
100
80
60
40
20
0
100
80
Mn3-600
60
Mn3-600B
Mn2-600
40
20
0
Mn1-600
0
20
40
60
80
100
325
175
200
225
250
275
300
Reaction Temperature / ºC
Np conversion (%)
Fig. 6. Evolution of naphthalene conversion into CO2 with the reaction tempera-
ture for manganese oxide catalysts calcined at 600 ◦C. Reaction conditions in text.
Catalysts: (ꢁ) Mn3-600, (ꢀ) Mn3-600B, (᭹) Mn2-600, (ꢂ) Mn1-600.
Fig. 7. Variation of the selectivity to CO2 with naphthalene conversion for rep-
resentative manganese oxide catalysts. Reaction conditions in text. Catalysts: (ꢁ)
Mn3-600, (ꢀ) Mn3-600B, (᭹) Mn2-600, (ꢂ) Mn1-600.

T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
175
of naphthalene conversion. The selectivity to CO2 was determined
as follows: Np conversion to CO2 divided by Np conversion (to
CO2 and also to the other reaction products). As can be seen, the
selectivity to CO2 increases with the naphthalene conversion, from
ca. 60% at 10% conversion to 90–100% at conversions higher than
90%. The trend observed is similar for all these catalysts in spite
of the fact that this graph has been depicted using different reac-
tion temperatures (same conditions as in Fig. 6). Therefore, the
preparation method does not seem to have any significant influ-
ence on the selectivity to CO2, but it does strongly influence catalyst
activity.
Focussing on the best Mn3-600 catalyst, an experiment to
assess initial stability was carried out. Employing the usual reac-
tion conditions and fixing the reaction temperature at 225 ◦C
catalyst performance was monitored for 24 h. Initially the naph-
thalene conversion to CO2 was 69% and within experimental
error the same conversion was maintained for the whole
run.
oxygen and hence the lowest OII/OI ratio. Therefore, the presence
of OII species, for which assignment is not clear, as they can be
indicative of oxygen vacancies but also of hydroxyl and carbonate
groups, does not seem to be the determining factor for the catalytic
activity.
The mechanism for oxidation of some VOCs, such as benzyl alco-
hol [40] or C3 hydrocarbons [41] by manganese oxide catalysts
has been related to the active participation of the lattice oxygen,
which means that oxidation takes place via a redox mechanism.
A correlation between the redox properties and the catalytic per-
formance of manganese oxides, and that lattice oxygen is involved
in the VOC oxidation mechanism, has also been shown previously
[36]. Moreover, the total oxidation of naphthalene by a cerium
oxide catalyst has been demonstrated to proceed via a Mars-van
Krevelen mechanism [42]. Thus it is reasonable to assume that
naphthalene oxidation on a manganese oxide catalyst also pro-
ceeds via a Mars-van Krevelen mechanism although it cannot be
ruled out that surface adsorbed species can also play a role, so
a Langmuir–Hinshelwood mechanism could partially contribute.
Therefore, the activity of these manganese oxide catalysts must be
mainly determined by both the reduction and the re-oxidation rates
of the active sites. The reducibility can be estimated through the H2
TPR experiments and it can be seen that ease of reduction increases
with increasing surface area of the manganese oxides. Accordingly
a correlation between catalytic activity (both per gram or per m2)
and the reducibility is clearly obtained for these catalysts (Fig. 8b).
In this sense, the characteristics of the lattice oxygen, which is used
for oxidation must be important. As observed by XPS the propor-
tion of surface lattice oxygen is highest in the most active catalysts
and lowest in the least active catalyst. However the high surface
concentration is not necessarily an indication of the reactivity of
the catalysts as we also have to consider the mobility of this lattice
oxygen. Thus if this oxygen can be easily removed from the lattice
the formation of CO2 will probably be more facile. The O 1s signal
has the lowest binding energies (see Table 3) for the most active
catalyst and it increases as the catalytic activity of the catalysts
decreases. These lower binding energies can be linked to a higher
lattice oxygen mobility [43], which we have also confirmed from
TPR, and they will have a direct influence on the reactivity of these
catalysts.
The rate of CO2 formation at 225 ◦C normalized per catalyst mass
(intrinsic activity) and catalyst surface area (specific activity) are
plotted as a function of catalyst surface area in Fig. 8a. An appro-
priate comparison of intrinsic and specific activity should be made
under differential conditions using conversions not much larger
than 5–10%. Unfortunately the reactivity of these catalysts is very
different, so we have selected for comparison a temperature at
which all catalysts are active. Therefore these data, although signif-
icant, must be taken with caution as there can be some contribution
from mass transfer. The surface area of the catalysts seems to be of
paramount importance in the determination of the catalytic activ-
ity, as the highest activity per gram of catalyst has been achieved by
Mn3-600 and increases as the surface area of the catalyst increases.
Additionally, if the activity is normalized by the surface area the
greatest values are also obtained by samples with the highest
surface areas, following the same trend. Thus the enhanced perfor-
mance obtained by high surface area catalysts is not only the result
of the increased surface area, and therefore other factors have to be
considered.
The activity of manganese oxides for the total oxidation of
ethanol has been related to the ease to which lattice oxygen
can be donated [22]. In another study for the total oxidation of
benzene and toluene it was found that defective oxygen species
or hydroxyl-like groups, which have higher mobility than lat-
tice oxygen species, were more abundant. This is evidenced by
the higher OII/OI ratio detected by XPS in the most active man-
ganese oxide catalysts, and this was tentatively assigned to be
the species responsible for the enhancement of activity [27,39].
Conversely, in the present work it has been observed by XPS that
the most active catalyst presents the highest proportion of lattice
The mechanism for naphthalene oxidation on a n-type oxide,
such as Mn2O3, is thought to involve strong adsorption of the naph-
thalene at an anionic oxygen site on the oxide lattice leading to the
formation of an activated complex [11], that will decompose, most
likely in a number of steps, to yield CO2. For this mechanism to take
place, a high degree of oxygen mobility on the surface is necessary,
with the successive decomposition of the reaction intermediates
and desorption of complete oxidation products [44]. Accordingly,
a
b
400
300
200
100
0
400
300
200
100
0
400
300
200
100
0
400
300
200
100
0
0
10 20 30 40 50 60 70 80
Surface area, m2/g
150 175 200 225 250 275
Tonset-TPR (ºC)
Fig. 8. Variation of the specific (ꢁ) and intrinsic STY (ꢁ) rates as a function of: (a) catalyst surface area and (b) onset temperature in the TPR-H2 experiments. Reaction
conditions in text and a reaction temperature of 225 ◦C.

176
T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
this high oxygen mobility is what we have observed for the most
active Mn3-600 catalyst.
demonstrated by TPR and (iii) the mobility of the active lattice oxy-
gen, which favours the CO2 formation, is higher than for the rest
of the catalysts. On the other hand, it was also observed that the
Mn2O3 crystalline phase presents a higher intrinsic activity than
MnO2.
3.2. Effect of the calcination temperature (catalysts treated at
300, 500 and 600 ◦C)
The XRD of the catalysts calcined at 300, 500 and 600 ◦C synthe-
sized through different preparation methods are shown in Fig. 1,
and phases identified along with surface areas summarized in
Table 1. It has been observed that the nature of the MnOx produced
is dependent of the preparation procedure and the calcination tem-
perature employed as it has been previously reported [45].
In the case of series 1 catalysts, the catalyst calcined at 300 ◦C
presented diffraction peaks typical from ␣-MnO2 (JCPDS: 44-0141),
whilst calcination at 600 ◦C only showed diffraction corresponding
to Mn2O3. The catalyst calcined at an intermediate temperature,
Mn1-500, presents peaks corresponding to both crystalline phases,
MnO2 and Mn2O3. Therefore the increase in the calcination temper-
ature leads to the formation of less oxidized manganese oxides. The
surface area of these catalysts was very low, and it only exceeded
2 m2/g for Mn1-300.
Acknowledgements
B.S, A.D. and I.V. thank the financial support from DGICYT in
Spain (Project CTQ-2009-14495) and T.G. to the Ministry of Science
and Innovation (Spain) and Plan E through project ENE2009-11353.
SCSIE (Universitat de València) is acknowledged for the TEM mea-
surements.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2012.10.029.
References
In the case of the series 2 catalysts, Mn2O3 was the predominant
crystalline phase in all the catalysts. Only in the catalyst calcined
at 300 ◦C were some peaks from MnO2 apparent. As can be seen
this second method leads to the formation of more reduced bulk
crystalline phases. These catalysts also present a surface area much
higher than those of the first series (7–42 versus <2 m2/g). Therefore,
the oxalic acid present in the synthesis procedure exerts multiple
roles, as a template increasing the surface area and as a reducing
agent hindering the formation of the more oxidized MnO2 phase.
In the catalysts prepared by nanocasting only two calcination
temperatures, 500 and 600 ◦C, have been studied. In both cases
Mn2O3 was the predominant phase, although low intensity diffrac-
tion peaks resulting from MnO2 could be observed in the sample
calcined at 500 ◦C. This preparation method leads to high surface
areas that, for Mn3-500, exceed 100 m2/g.
Table 1 (and Fig. S2) compares the activity of the different cat-
alysts synthesized by considering the T10, T50 and T90 values
(reaction temperatures required for naphthalene conversions to
CO2 of 10, 50 and 90% respectively). Considerable differences of
activity were observed between the different preparation meth-
ods, whilst the effect of the calcination temperature appears to
be less important. For all the series of catalysts it can be seen
that an increase in the calcination temperature leads to a reduc-
tion of the surface area and also an increase in the relative
proportion of Mn2O3 to MnO2. Therefore the Mn2O3 crystalline
phase seems to present a higher intrinsic activity than MnO2,
as the higher catalytic activity expected for the higher surface
areas is compensated for by the higher proportion of the less
active MnO2 phase. The observation of higher activity of Mn2O3
is also in agreement with other published studies dealing with
the total oxidation of pollutants such as CO [46] and aromatics
[18,28].
[1] Luxemburg: Office for Official Publications of the European Communities, PAH
Position Paper, 2001, ISBN92-894-2057-X.
[2] J.I. Levy, E.A. Houseman, J.D. Spengler, P. Loh, L. Ryan, Environ. Health Prospect
109 (2001) 341.
[3] E. Ntainjua, S.H. Taylor, Top. Catal. 52 (2009) 528–541.
[4] A.K. Neyestanaki, L.E. Lindfors, T. Ollonqvist, J. Vayrynen, Appl. Catal. A: Gen.
196 (2000) 233–246.
[5] F. Klingstedt, A. Kalantar Neyestanaki, L.-E. Lindfors, T. Salmi, T. Heikkila, E.
Laine, Appl. Catal. A: Gen. 239 (2003) 229–240.
[6] A.K. Neyestanaki, L.-E. Lindfors, Fuel 77 (1998) 1727–1734.
[7] X.W. Zhang, S.C. Shen, L.E. Yu, S. Kawi, K. Hidajat, K.Y. Simon Ng, Appl. Catal. A:
Gen. 250 (2003) 341–352.
[8] J.-L. Shie, C.-Y. Chang, J.-H. Chen, W.-T. Tsai, Y.-H. Chen, C.-S. Chiou, C.-F. Chang,
Appl. Catal. B: Environ. 58 (2005) 289–297.
[9] N.E. Ntainjua, T. García, S.H. Taylor, Catal. Lett. 110 (2006) 125–128.
[10] N.E. Ntainjua, A.F. Carley, S.H. Taylor, Catal. Today 137 (2008) 362–366.
[11] J.J. Spivey, Ind. Eng. Chem. Res. 26 (1987) 2165–2180.
[12] T. García, B. Solsona, S.H. Taylor, Appl. Catal. B: Environ. 66 (2006) 92–99.
[13] A.H. Lu, D. Zhao, Y. Wan, Nanocasting: A Versatile Strategy for Creating Nano-
structured Porous Materials, RSC Nanoscience and Nanotechnoogy No. 11,
2010.
[14] M. Tiemann, Chem. Mater. 20 (2008) 961–971.
[15] T. Valdés-Solís, A.B. Fuertes, Mater. Res. Bull. 41 (2006) 2187–2197.
[16] W.B. Yue, W.Z. Zhou, J. Mater. Chem. 17 (2007) 4947–4952.
[17] B.Z. Tian, X.Y. Liu, H.F. Yang, S.H. Xie, C.Z. Yu, B. Tu, D.Y. Zhao, Adv. Mater. 15
(2003) 1370–1374.
[18] M. Baldi, V. Sanchez-Escribano, J.M. Gallardo-Amores, F. Milella, G. Busca, Appl.
Catal. B: Environ. 17 (1998) L175–L182.
[19] J. Hu, W. Chu, L.J. Shi, Nat. Gas Chem. 17 (2008) 159–164.
[20] M.I. Zaki, M.A. Hasan, L. Pasupulety, N.E. Fouad, H. Knözinger, New J. Chem. 23
(1999) 1197–1202.
[21] F.N. Aguero, A. Scian, B.P. Barbero, L.E. Cadus, Catal. Today 133–135 (2008)
493–501.
[22] S.S.T. Bastos, J.J.M. Orfao, M.M.A. Freitas, M.F.R. Pereira, J.L. Figueiredo, Appl.
Catal. B: Environ. 93 (2009) 30–37.
[23] S.S.T. Bastos, S.A.C. Carabineiro, J.J.M. Órfão, M.F.R. Pereira, J.J. Delgado, J.L.
Figueiredo, Catal. Today 180 (2012) 148–154.
[24] J. Luo, Q. Zhang, A. Huang, S.L. Suib, Micropor. Mesopor. Mater. 35–36 (2000)
209–217.
[25] C. Lahousse, A. Bernier, P. Grange, B. Delmon, P. Papaefthimiou, T. Ioannides, X.
Verykios, J. Catal. 178 (1998) 214–225.
[26] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, J. Hazard. Mater. 185
(2011) 1236–1240.
[27] S.C. Kim, W.G. Shim, Appl. Catal. B: Environ. 98 (2010) 180–185.
[28] E. Rossinyol, J. Arbiol, F. Peiró, A. Cornet, J.R. Morante, B. Tian, T. Bob, D. Zhao,
Sens. Actuators B 109 (2005) 57–63.
4. Conclusions
[29] B. Puertolas, B. Solsona, S. Agouram, R. Murillo, A.M. Mastral, A. Aranda, S.H.
Taylor, T. Garcia, Appl. Catal. B: Environ. 93 (2010) 395–405.
[30] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[31] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380.
[32] JCPDS – International Centre for Diffraction Data, JCPDS (24-0508, 44-0141).
[33] P. Scherrer, Göttinger Nachrichten Math. Phys. 2 (1918) 98–100.
[34] L. Christel, A. Pierre, D.A.M.R. Abel, Thermochim. Acta 306 (1997) 51–59.
[35] F. Wang, H. Daí, J. Deng, G. Bai, K. Ji, Y. Liu, Environ. Sci. Technol. 46 (2012)
4034–4041.
The surface area of Mn2O3 catalysts appears to be one of
the most important factors controlling naphthalene total oxida-
tion activity. The use of high surface area Mn2O3 prepared by a
nanocasting route significantly increases the catalytic activity for
the total oxidation of naphthalene. The nanocast catalyst demon-
strates the greatest surface area normalized activity, and this is
due to a combination of the following factors: (i) the number of
catalytically active sites increase as the surface area increases,
(ii) the reducibility of the manganese oxides sites is easier as
[36] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, Appl. Catal. B: Environ.
99 (2010) 353–363.
[37] M.I. Zaki, M.A. Hasan, L. Pasupulety, K. Kumari, New J. Chem. 22 (1998) 875–882.

T. Garcia et al. / Applied Catalysis A: General 450 (2013) 169–177
177
[38] E. Rocchini, A. Trovarelli, J. Llorca, G.W. Graham, W.H. Weber, M. Maciejewski,
A. Baiker, J. Catal. 194 (2000) 461–476.
[43] E.C. Njagi, C.-H. Chen, H. Genuino, H. Galindo, H. Huang, S.L. Suib, Appl. Catal.
B: Environ. 99 (2010) 103–110.
[39] H. Chen, A. Sayari, A. Adnot, F. Larachi, Appl. Catal. B: Environ. 32 (2001)
195–204.
[40] V.D. Makwana, Y.-C. Son, A.R. Howell, S.L. Suib, J. Catal. 210 (2002) 46–52.
[41] M. Baldi, E. Finocchio, F. Milella, G. Busca, Appl. Catal. B: Environ. 16 (1998)
43–51.
[44] J.E. Germain, Intra-Sci. Chem. Rep. 6 (3) (1972) 101.
[45] M.I. Zaki, A.K.H. Nohman, C. Kappenstein, T.M. Wahdan, J. Mater. Chem. 5 (1995)
1081–1088.
[46] K. Ramesh, L. Chen, F. Chen, Y. Liu, Z. Wang, Y.F. Han, Catal. Today 131 (2008)
477–482.
[42] B. Solsona, T. García, R. Murillo, A.M. Mastral, E.N. Ntainjua, C.E. Hetrick, M.D.
Amiridis, S.H. Taylor, Top. Catal. 52 (2009) 492–500.