Rod-, flower-, and dumbbell-like MnO2: Highly active catalysts for the combustion of toluene

Article abstract of DOI:10.1016/j.apcata.2012.05.016

The rod-like tetragonal α-MnO₂, flower-like hexagonal ε-MnO₂, and dumbbell-like tetragonal β-MnO₂ were prepared using the hydrothermal or water-bathing method under different conditions. It is shown that the α-MnO₂, ε-MnO ₂, and β-MnO₂ catalysts possessed a surface area of ca. 53, 30, and 114 m²/g, respectively. The oxygen adspecies concentration and low-temperature reducibility decreased in the order of α-MnO₂ > ε-MnO₂ > β-MnO₂, coinciding with the sequence of their catalytic activities for toluene combustion. The well-defined morphological MnO₂ catalysts performed much better than the bulk counterpart. At a space velocity of 20,000 mL/(g h), the temperature for 90% toluene conversion was 238, 229, and 241 °C over α-MnO₂, ε-MnO₂, and β-MnO₂, respectively. The apparent activation energies of α-MnO₂, ε-MnO₂, and β-MnO₂ were in the range of 20-26 kJ/mol. It is concluded that higher oxygen adspecies concentrations and better low-temperature reducibility were responsible for the good catalytic performance of the α-MnO₂, ε-MnO₂, and β-MnO ₂ materials.

Full text of DOI:10.1016/j.apcata.2012.05.016

Applied Catalysis A: General 433–434 (2012) 206–213
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Rod-, flower-, and dumbbell-like MnO2: Highly active catalysts for the
combustion of toluene
Fengjuan Shia , Fang Wanga , Hongxing Daia,∗ , Jianxing Daib , Jiguang Denga , Yuxi Liua , Guangmei Baia ,
Kemeng Jia, Chak Tong Auc,∗∗
a Laboratory of Catalysis Chemistry and Nanoscience, Department of Chemistry and Chemical Engineering, College of Environmental and Energy Engineering, Beijing University of
Technology, Beijing 100124, China
b Office 7143, Nanchang Military Academy, Nanchang 330103, Jiangxi Province, China
c Department of Chemistry, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China
a r t i c l e i n f o
a b s t r a c t
Article history:
The rod-like tetragonal ␣-MnO2, flower-like hexagonal ␧-MnO2, and dumbbell-like tetragonal ␤-MnO2
were prepared using the hydrothermal or water-bathing method under different conditions. It is shown
that the ␣-MnO2, ␧-MnO2, and ␤-MnO2 catalysts possessed a surface area of ca. 53, 30, and 114 m2/g,
respectively. The oxygen adspecies concentration and low-temperature reducibility decreased in the
order of ␣-MnO2 > ␧-MnO2 > ␤-MnO2, coinciding with the sequence of their catalytic activities for toluene
combustion. The well-defined morphological MnO2 catalysts performed much better than the bulk coun-
terpart. At a space velocity of 20,000 mL/(g h), the temperature for 90% toluene conversion was 238, 229,
and 241 ◦C over ␣-MnO2, ␧-MnO2, and ␤-MnO2, respectively. The apparent activation energies of ␣-MnO2,
␧-MnO2, and ␤-MnO2 were in the range of 20–26 kJ/mol. It is concluded that higher oxygen adspecies con-
centrations and better low-temperature reducibility were responsible for the good catalytic performance
of the ␣-MnO2, ␧-MnO2, and ␤-MnO2 materials.
Received 14 January 2012
Received in revised form 5 April 2012
Accepted 14 May 2012
Available online 22 May 2012
Keywords:
Hydrothermal preparation method
Water-bathing preparation method
Well-defined morphology
Manganese oxide catalyst
Toluene combustion
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
past years, manganese oxides with different crystal structures
and morphologies have been prepared using various methods.
Most of volatile organic compounds (VOCs) emitted from indus-
trial and transportation activities are pollutants to the atmosphere
and human health. Catalytic combustion is an effective pathway to
remove the VOCs. The important aspect of such a technology is the
availability of high-performance catalysts. Although the precious
metal catalysts perform excellently for the total oxidation of VOCs
(e.g., toluene) at low temperatures [1,2], the high cost limits their
wide applications. Cheap metal oxide (e.g., Co3O4 [3] and MnOx
[4])-based materials show good catalytic activities for VOCs com-
bustion at high temperatures [5–7]. Therefore, it is highly desired to
develop base metal oxide catalysts that exhibit good performance
at low temperatures for the removal of VOCs.
Manganese oxides (MnOx) are among the most active oxide
catalysts in VOC oxidation [8–12]. The crystal structure of MnOx
has a great influence on its catalytic performance. For exam-
ple, the ␥-MnO2 phase was reported to be more active than
␤-MnO2 or ␣-Mn2O3 in the oxidation of ethanol [10–12]. In the
For example, Torres et al. [13] prepared mesoporous MnOx cat-
alysts and observed their good catalytic activities for the total
oxidation of formaldehyde. Lamaita et al. [10] pointed out that
␣-Mn2O3 and ␥-MnO2 were more active than ␤-MnO2 in the
catalytic oxidation of ethanol. Santos et al. [14] claimed that ␣-
Mn2O3 and ␤-MnO2 showed good catalytic performance for the
total oxidation of ethyl acetate. He and coworkers [15] fabri-
cated porous cellulose-supported MnO2 nanosheet catalysts, and
observed that such morphological MnO2 materials exhibited good
catalytic activities for the oxidative removal of formaldehyde. After
investigating flower-like manganese oxide nanospheres for the
oxidation of formaldehyde, Zhou et al. [16] found that the ␣-
MnO2 outperformed the other crystalline manganese oxides below
120 ◦C. Working on the preparation and catalysis of Mn3O4, ␣-
Mn2O3, and ␤-MnO2 materials for toluene or benzene combustion,
Kim et al. [17] concluded that the catalytic activity decreased in
the sequence of Mn3O4 > ␣-Mn2O3 > ␤-MnO2. To the best of our
knowledge, however, there have so far been rare studies on the cat-
alytic application of MnO2 with different crystal phases and various
well-defined morphologies for the combustion of toluene.
∗
Corresponding author. Tel.: +86 10 6739 6118; fax: +86 10 6739 1983.
Corresponding author. Tel.: +852 3411 7067; fax: +852 3411 7348.
E-mail addresses: hxdai@bjut.edu.cn (H. Dai), pctau@hkbu.edu.hk (C.T. Au).
By using the ordered mesoporous silica (KIT-6- or SBA-16)-
nanocasting approach, our group have prepared a number of
∗∗
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.05.016

F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
207
mesoporous transition-metal oxides (e.g., chromia [18,19], iron
oxide [20], manganese oxide [21], and cobalt oxide [21,22]), and
investigated their catalytic properties for the removal of formalde-
hyde, acetone, methanol, and toluene. Recently, we adopted the
hydrothermal or water-bathing strategy to generate numerous
transition-metal oxides with well-defined morphologies. In this
paper, we report the preparation and catalytic performance of ␣-,
␧- or ␤-MnO2 with a rod-, flower- or dumbbell-like morphology for
the combustion of toluene.
(a)
2. Experimental
(b)
(c)
2.1. Catalyst preparation
The ␣-MnO2, ␧-MnO2, and ␤-MnO2 catalysts with various mor-
phologies were prepared using the hydrothermal or water-bathing
method. The detailed procedures are as follows:
(i) 8 mmol of KMnO4 and 3 mmol of MnSO4·H2O were dissolved
in 80 mL of deionized water, after stirring for 1 h the mixed
solution was transferred to a 100-mL autoclave for thermal
treatment at 160 ◦C for 12 h [23];
(ii) 2 mmol of KMnO4 was dissolved in 40 mL of deionized water,
and 2 mL of 98 wt.% H2SO4 solution was dropwise added to the
KMnO4 solution under stirring, and then a piece of copper plate
(2 cm × 8 cm in dimension) was added to the mixed solution,
which was water-bathed at 60 ◦C for 8 h [24];
(iii) 4 mmol of KMnO4 was dissolved in 80 mL of deionized water
under stirring, and 16 mmol of glucose, 1 mL of ethylenedi-
amine (EDA), and 1.4 mL of 37 wt.% HCl solution were added to
the KMnO4 solution, which was then transferred to a 100-mL
autoclave for thermal treatment at 120 ◦C for 12 h [25].
10
20
30
40
50
60
70
80
2 Theta (Deg.)
Fig. 1. Wide-angle XRD patterns of (a) rod-like ␣-MnO2, (b) flower-like ␧-MnO2,
and (c) dumbbell-like ␤-MnO2.
(hv = 1253.6 eV) was used as excitation source and the instrumental
resolution was 0.5 eV. After treatment in O2 (flow rate = 20 mL/min)
at 500 ◦C for 1 h (for removal of adsorbed water and surface car-
bonate), the samples were cooled to RT and transferred into the
spectrometer under helium (GLOVE BAG, Instruments for Research
and Industry, USA). The samples were then degassed in the prepara-
tion chamber (10−5 Torr) for 0.5 h and introduced into the analysis
chamber (3 × 10−9 Torr) for spectrum recording. The C 1s signal at
BE = 284.6 eV was taken as reference for BE calibration.
After being filtered, washed with deionized water three times
and absolute ethanol two times, and dried at 60 ◦C for 24 h, the sam-
ple precursor derived from (i), (ii) or (iii) was calcined in a muffle
furnace at a ramp of 1 ◦C/min from RT to 300, 300 or 350 ◦C and
kept at this temperature for 4 h, thus obtaining the rod-, flower-
, and dumbbell-like MnO2 catalysts with different crystal phases,
respectively.
The bulk MnO2 sample (denoted as bulk-MnO2) with a surface
area of 10.1 m2/g and the other chemicals (A.R. in purity) were pur-
chased from Beijing Chemical Company and used without further
purification.
Hydrogen temperature-programmed reduction (H2-TPR) was
carried out in the 50–600 ◦C range on a Micromeritics AutoChem
II 2920 apparatus. 0.1 g of the sample (40–60 mesh) was treated
in a helium flow of 25 mL/min at 350 ◦C for 1 h and then cooled in
the same atmosphere to 50 ◦C. The sample was then reduced in a
flow of 5% H2–95% Ar (v/v) at a ramp of 10 ◦C/min. The outlet gases
were analyzed on-line using a thermal conductivity detector (TCD).
The TCD responses were calibrated against that of the complete
reduction of a known standard CuO sample (Aldrich, 99.995%).
2.3. Catalytic evaluation
The catalytic activity was evaluated with the sample charged in
a continuous flow fixed-bed quartz microreactor (i.d. = 4 mm). To
minimize the effect of hot spots, the catalyst (0.1 g, 40–60 mesh)
was diluted with 0.3 g of quartz sands (40–60 mesh). The reactant
feed (flow rate = 33.3 mL/min) was 1000 ppm toluene + O2 + N2 (bal-
ance) with a toluene/O2 molar ratio of 1/400. The space velocity
(SV) was 20,000 mL/(g h). The outlet gases were analyzed online by
a gas chromatograph (GC-2010, Shimadzu) equipped with a flame
ionization detector and a TCD, using a Chromosorb 101 column for
toluene and a Carboxen 1000 column for permanent gas separation.
The balance of carbon throughout the investigation was estimated
to be 99.5%.
2.2. Catalyst characterization
Powdered X-ray diffraction (XRD) patterns of the samples
were recorded on a Bruker/AXS D8 Advance diffractometer oper-
ated at 35 kV and 35 mA using Cu K˛ radiation and a Ni filter
(ꢀ = 0.15406 nm). Crystal phases were identified by referring the
diffraction lines to those of the powder diffraction files, 1998 ICDD
PDF database. The Brunauer–Emmett–Teller (BET) surface areas of
the samples were measured via N2 adsorption at −196 ◦C on an
ASAP 2020 adsorption analyzer (Micromeritics) with the samples
being outgassed at 250 ◦C for 2 h under vacuum prior to mea-
surements. The scanning electron microscopic (SEM) images of
the samples were recorded on a Gemini Zeiss Supra 55 appa-
ratus operating at 10 kV. By means of a JEOL-2010 instrument
(operated at 200 kV), transmission electron microscopic (TEM) and
high-resolution TEM images as well as the selected-area electron
diffraction (SAED) patterns of the samples were obtained. X-ray
photoelectron spectroscopy (XPS, VG CLAM 4 MCD analyzer) was
employed to determine the Mn 2p, O 1s, and C 1s binding ener-
gies (BEs) of surface species of the as-prepared samples; Mg K˛
3. Results and discussion
3.1. Crystal structure
Fig. 1 shows the XRD patterns of the as-prepared MnO2 catalysts.
By referring to the XRD patterns of the standard ␣-MnO2 (JCPDS
PDF# 44-0141), ␧-MnO2 (JCPDS PDF# 30-0820), and ␤-MnO2

208
F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
Fig. 2. (a–f) SEM and (g–l) TEM images as well as the SAED patterns (insets) of (a, b, g and h) rod-like ␣-MnO2, (c, d, i and j) flower-like ␧-MnO2, and (e, f, k and l) dumbbell-like
␤-MnO2.
(JCPDS PDF# 24-0735), one can deduce that the rod-, flower-, and
dumbbell-like manganese oxide samples were single-phase and
of tetragonal ␣-MnO2, hexagonal ␧-MnO2, and tetragonal ␤-MnO2
crystal structures, respectively. All of the diffraction peaks could
be well indexed, as indicated in Fig. 1(a)–(c). The discrepancies
in intensity and width of XRD peaks suggest the difference in
crystallinity of these samples, with the ␣-MnO2 sample being
better than the ␧- and ␤-MnO2 samples in crystallinity. Therefore,
the XRD results reveal that the crystal structure and crystallinity
of the manganese oxide sample were strongly dependent on the
preparation method and condition.
with KMnO4 as Mn source in the presence of glucose, ethylenedi-
amine, and HCl, most of the particles showed a dumbbell-like shape
with the ends being flower-like (Fig. 2(e) and (f)), and the dumb-
bell was roughly 2–4 ␮m in diameter. From Fig. 2(g)–(l), one can
clearly see that the ␣-MnO2 and ␧-MnO2 samples were nanorod-
like and nanoflower-like spherical in morphology (Fig. 2(g) and
(i)), respectively, whereas the dumbbell ends of the ␤-MnO2 sam-
ple were composed of a large number of aggregated nanoparticles
(Fig. 3(k)). It is observed from Fig. 3(h), (j), and (l) that there were a
large number of well-resolved lattice fringes. The lattice spacings (d
values) of the (2 1 1), (1 0 0), and (1 0 1) crystal planes of the rod-like
␣-MnO2, flower-like ␧-MnO2, and dumbbell-like ␤-MnO2 samples
were respectively 0.239, 0.241, and 0.240 nm, rather close to the
corresponding d values (0.23950, 0.24200, and 0.24070 nm) of the
standard ␣-MnO2 (JCPDS PDF# 44-0141), ␧-MnO2 (JCPDS PDF# 30-
0820), and ␤-MnO2 (JCPDS PDF# 24-0735) samples. Furthermore,
the recording of linearly aligned bright electron diffraction spots
in the SAED pattern (inset of Fig. 3(b)) suggests that the rod-like
␣-MnO2 was single-crystalline; in the SAED patterns of the flower-
like spherical ␧-MnO2 and dumbbell-like ␤-MnO2 samples, the
appearance of multiple bright electron diffraction rings indicates
the formation of polycrystalline ␧-MnO2 and ␤-MnO2.
3.2. Morphology and textural property
Fig. 2 shows the SEM and TEM images as well as the SAED
patterns of the as-prepared manganese oxide samples. As can
be clearly seen from Fig. 2(a)–(f), the ␣-MnO2 sample derived
hydrothermally with KMnO4 and MnSO4·H2O as metal source dis-
played a rod-like morphology (Fig. 2(a) and (b)), the diameter and
length of the rods were mostly ca. 80 nm and 3–6 ␮m, respec-
tively. The ␧-MnO2 sample derived from the water-bathing process
with KMnO4 and copper plate as metal source was flower-like in
morphology, which was aggregated by a number of nanoplates
(Fig. 2(c) and (d)), the diameter of the flower-like sphere was typi-
cally ca. 1.5 ␮m. In the ␤-MnO2 sample fabricated hydrothermally
As can be seen from Table 1, the crystallite sizes of the rod-like
␣-MnO2, flower-like ␧-MnO2, and dumbbell-like ␤-MnO2 sam-
ples were 23.1, 5.8, and 6.0 nm, respectively. The BET surface area

F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
209
(A) Mn 2p2/3
641.1
(B) O 1s
529.0
642.5
531.7
644.8
(c)
(b)
(c)
(b)
(a)
(a)
636
638
640
642
644
646
648
526
528
530
532
534
536
Binding energy (eV)
Binding energy (eV)
Fig. 3. (A) Mn 2p3/2 and (B) O 1s XPS spectra of (a) rod-like ␣-MnO2, (b) flower-like ␧-MnO2, and (c) dumbbell-like ␤-MnO2.
of the rod-like ␣-MnO2 sample was 53.1 m2/g, higher than that
(30.3 m2/g) of the flower-like ␧-MnO2 sample. The dumbbell-like
␤-MnO2 sample, however, showed a BET surface area of 113.5 m2/g,
markedly higher than those (30.3–53.1 m2/g) of the rod-like ␣-
MnO2 and flower-like ␧-MnO2 samples.
and the other at BE = 531.7 eV, the former was ascribable to the sur-
face lattice oxygen (Olatt) species, whereas the latter to the surface
adsorbed oxygen (Oads) species [26–29]. After making quantitative
analysis (Table 1), one can see that the surface Oads/Olatt molar ratio
decreased in the sequence of rod-like ␣-MnO2 (1.28) > flower-like
␧-MnO2 (0.96) > dumbbell-like ␤-MnO2 (0.91) > bulk-MnO2 (0.78),
coinciding with the order of Mn3+/Mn4+ molar ratio of these sam-
ples.
3.3. Surface species and reducibility
Reducibility is an important factor influencing the catalytic
activity of a material. Fig. 4(A) illustrates the H2-TPR profiles of
the MnO2 samples. The bulk-MnO2 sample displayed only one
reduction peak centered at ca. 510 ◦C, corresponding to a H2 con-
sumption of 6.2 mmol/g (Table 2). For the rod-like ␣-MnO2 sample,
there were two reduction peaks centered at 245 and 352 ◦C, cor-
responding to a H2 consumption of 6.0 and 3.1 mmol/g (Table 2),
respectively; for the flower-like ␧-MnO2 sample, two reduction
peaks centered at 250 and 311 ◦C appeared and the correspond-
ing H2 consumptions were 5.0 and 2.9 mmol/g (Table 2); for the
dumbbell-like ␤-MnO2 sample, however, only one reduction signal
at 296 ◦C was recorded, and the H2 consumption was 9.3 mmol/g
(Table 2). According to the literature [13,17], the reduction pro-
cess of MnO2 could be reasonably divided into two steps: (i)
MnO2 → Mn3O4 and (ii) Mn3O4 → MnO, which corresponds to a
H2 consumption of 7.67 and 3.83 mmol/g (totally 11.50 mmol/g).
By comparing the theoretical and experimental H2 consumption
data (Table 2), one can see that 69–81% Mn4+ and Mn3+ in rod-
like ␣-MnO2, flower-like ␧-MnO2, and dumbbell-like ␤-MnO2 were
reduced to Mn2+ below 400 ◦C, whereas less half of Mn4+ and Mn3+
in bulk-MnO2 was reduced to Mn2+ below 550 ◦C. Furthermore, the
Fig. 3 illustrates the Mn 2p3/2 and O 1s XPS spectra of the
manganese oxide samples. It is observed from Fig. 3(A) that
the asymmetrical Mn 2p3/2 XPS signal at BE = ca. 642 eV could
be decomposed into three components at BE = 641.1, 642.5, and
644.8 eV, assignable to the surface Mn3+ and Mn4+ species and the
satellite [17,26–28], respectively. A quantitative analysis on the Mn
2p3/2 XPS spectra of the samples gives rise to the surface Mn3+/Mn4+
molar ratios, as summarized in Table 1. It is clear that the prepa-
ration method had an influence on the surface Mn3+/Mn4+ molar
ratio of the product, with the rod-like ␣-MnO2 sample possess-
ing the highest surface Mn3+/Mn4+ molar ratio (0.60) whereas the
bulk-MnO2 sample showing the lowest surface Mn3+/Mn4+ molar
ratio (0.34). Usually, oxygen molecules are adsorbed at the oxy-
gen vacancies of an oxide material. Therefore, we believe that the
oxygen adspecies locate at the surface oxygen vacancies of MnO2.
Based on the principle of electroneutrality, we deduce that the
surface oxygen vacancy density would be the highest on the rod-
like ␣-MnO2 surface but the lowest on the bulk-MnO2 surface.
Such a deduction was substantiated by the results of O 1s XPS
investigations. As shown in Fig. 3(B), the asymmetrical O 1s sig-
nal could be deconvoluted to two components: one at BE = 529.0 eV
Table 1
Preparation conditions, crystallite sizes, BET surface areas, and surface element compositions of the as-prepared MnO2 samples.
Sample
Preparation
method
Mn source
Calcination
condition
Crystallite size
(nm)
BET surface
area (m2/g)
Surface element
molar ratio
a
Mn3+/Mn4+
Oads/Olatt
Bulk-MnO2
Rod-like ␣-MnO2
–
–
–
–
23.1
10.1
53.1
0.34
0.60
0.78
1.28
Hydrothermal
treatment at 160 ◦C
for 12 h
KMnO4 and MnSO4
300 ◦C, 4 h
Flower-like ␧-MnO2
Water-bathing
treatment at 60 ◦C
for 8 h
Hydrothermal
treatment at 120 ◦C
for 12 h
KMnO4
KMnO4
300 ◦C, 4 h
350 ◦C, 4 h
5.8
6.0
30.3
0.56
0.50
0.96
0.91
Dumbbell-like ␤-MnO2
113.5
a
The data were obtained according to the Scherrer equation using the FWHM of the (1 0 1) line for rod-like ␣-MnO2, the (1 0 0) line for flower-like ␧-MnO2, and the (2 1 1)
line for dumbbell-like ␤-MnO2.

210
F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
10.00
(B)
(A)
296
(b)
8.00
6.00
4.00
2.00
0.00
(d)
(c)
250
245
311
(c)
352
(b)
(a)
(d)
510
500
(a)
600
0
100
200
300
400
1.25
1.45
1.65
1.85
2.05
Temperature (oC)
1000/ (K-1)
T
Fig. 4. (A) H2-TPR profiles and (B) initial H2 consumption rates of (a) bulk-MnO2, (b) rod-like ␣-MnO2, (c) flower-like ␧-MnO2, and (d) dumbbell-like ␤-MnO2.
first reduction temperatures (245–296 ◦C) of the rod-like ␣-MnO2,
flower-like ␧-MnO2, and dumbbell-like ␤-MnO2 samples were
much lower than that (510 ◦C) of the bulk-MnO2 sample, suggest-
ing that the three well-defined morphological MnO2 samples were
more reducible than the bulk-MnO2 sample at low temperatures.
In other words, the rod-like ␣-MnO2, flower-like ␧-MnO2, and
dumbbell-like ␤-MnO2 samples exhibited better low-temperature
reducibility than the bulk-MnO2 sample, which was confirmed by
the initial H2 consumption rate of these samples. In ordered to bet-
ter compare the low-temperature reducibility of the manganese
oxide samples, we calculated the initial H2 consumption rate of the
first reduction peak (where less than 25% oxygen in the sample
was removed [30,31]) of each sample, and the results are shown
in Fig. 4(B). Obviously, the initial H2 consumption rate decreased
in the order of rod-like ␣-MnO2 > flower-like ␧-MnO2 > dumbbell-
like ␤-MnO2 > bulk-MnO2. That is to say, the low-temperature
reducibility of these samples fellow such a sequence, which might
be intimately associated with their crystal structures, Mn3+/Mn4+
molar ratios, and morphologies.
over our catalysts. The reaction temperatures T50% and T90% (corre-
sponding to the toluene conversion = 50 and 90%) over the MnO2
catalysts were listed in Table 2. It is observed that the rod-like ␣-
MnO2 sample was better at lower temperatures (170–220 ◦C) than
but slightly inferior above 220 ◦C to the flower-like ␧-MnO2 sample
in catalytic performance; both the rod-like ␣-MnO2 and flower-like
␧-MnO2 catalysts outperformed the dumbbell-like ␤-MnO2 cata-
lyst. Shown in Fig. 5(B) is the effect of SV on the catalytic activity of
the rod-like ␣-MnO2 sample. As expected, the toluene conversion
increased with the drop in SV at the same reaction temperature,
a result due to the prolongation of contact between the reactants
and the catalyst. In order to examine the catalytic stability of the
rod-like ␣-MnO2 sample, we monitored the toluene conversion at
238 ◦C within 48 h of on-stream reaction and the result was shown
in Fig. 5(C). It is seen that no significant activity losses were detected
during the on-stream reaction process. Therefore, the rod-like ␣-
MnO2 material was catalytically durable. Furthermore, the H2-TPR
profile of the used rod-like ␣-MnO2 sample was similar to that of
the fresh counterpart (Fig. S2 of the Supplementary material).
The combustion of toluene has been studied over various
catalysts in the past years. The rod-like ␣-MnO2, flower-
like ␧-MnO2, and dumbbell-like ␤-MnO2 samples prepared
in the present study showed good catalytic performance
(T50% = 220–231 ◦C and T90% = 229–241 ◦C at SV = 20,000 mL/(g h)).
Apparently, such catalytic performance was much better than that
(T50% = 245 ◦C and T90% = 270 ◦C at SV = 15,000 mL/(g h)) over Mn3O4
[17], that (T50% = 280 ◦C and T90% = 295 ◦C at SV = 15,000 mL/(g h))
over ␣-Mn2O3 [17], that (T50% = 340 ◦C and T90% = 375 ◦C at
SV = 15,000 mL/(g h)) over ␤-MnO2 [17], that (T50% = 254 ◦C and
T90% = 295 ◦C at SV = 178 h−1) over LaMnO3 [32], that (T50% = 270 ◦C
and T90% = 300 ◦C at SV = 186 h−1) over 5 wt.% Au/CeO2 [33], and
that (T90% = 250 ◦C at SV = 18,000 mL/(g h)) over 0.5 wt.% Pd/LaMnO3
[34].
3.4. Catalytic performance
When only quartz sands were loaded into the microreactor, we
detected 10% conversion of toluene at 450 ◦C (Fig. S1 of the Sup-
plementary material). This result indicates that no homogeneous
reactions took place under the adopted reaction conditions (below
400 ◦C). Fig. 5(A) shows the catalytic performance of the as-
prepared MnO2 and bulk-MnO2 catalysts for the combustion of
toluene. Under the conditions of toluene concentration = 1000 ppm,
toluene/O2 molar ratio = 1/400, and SV = 20,000 mL/(g h), toluene
conversion increased monotonously with the rise in reaction
temperature; the rod-like ␣-MnO2, flower-like ␧-MnO2, and
dumbbell-like ␤-MnO2 catalysts showed much better activities
than the bulk-MnO2 catalyst. It is worth pointing out that CO2
and H2O were the only products and the estimated carbon balance
was around 99.5%. Therefore, toluene could be completely oxidized
It is well accepted that the oxidation of organic molecules
over the transition metal oxide catalysts involves a Mars–van
Krevelen mechanism, where the organic molecule is oxidized by
Table 2
Reduction temperatures, H2 consumption, catalytic activities (at SV = 20,000 mL/(g h)), and kinetic parameters of the MnO2 samples.
Sample
Reduction temperature (◦C) H2 consumption (mmol/g)
Catalytic activity (◦C)
Kinetic parameter
Peak 1
Peak 2
Peak 1
Peak 2
Total
T50%
T90%
A (s−1
)
Ea (kJ/mol)
R2
Bulk-MnO2
510
245
250
296
–
352
311
–
6.2
6.0
5.0
9.3
–
6.2
9.1
7.9
9.3
284
220
221
231
317
238
229
241
8.06 × 106
74.3
19.6
26.1
23.6
0.982
0.913
0.957
0.956
Rod-like ␣-MnO2
Flower-like ␧-MnO2
Dumbbell-like ␤-MnO2
3.1
2.9
–
29.1
1.30 × 102
58.0

F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
211
100
80
60
40
20
0
100
(A)
(B)
80
α
Rod-like -MnO2
SV (mL/(g h))
10,000
20,000
Ⴘ
႒
Ⴃ
Ⴠ
ε
Flower-like -MnO2
Ⴘ
႒
Ⴃ
β
Dumbbell-like -MnO2
60
40
20
0
Bulk-MnO2
40,000
50
100
150
200
250
300
350
50
100
150
200
250
300
350
Temperature (oC)
Temperature (oC)
100
(C)
90
80
70
60
SV = 20,000 mL(g h)
Toluene/O2 molar ratio = 1/400
0
10
20
30
40
50
On-stream reation time (h)
Fig. 5. (A) Toluene conversion as a function of reaction temperature over the catalysts under the conditions of toluene concentration = 1000 ppm, toluene/O2 molar
ratio = 1/400, and SV = 20,000 mL/(g h); (B) effect of SV on the catalytic activity of rod-like ␣-MnO2 at toluene/O2 molar ratio = 1/400; and (C) toluene conversion as a function
of on-stream reaction time over the rod-like ␣-MnO2 catalyst at 238 ◦C.
the lattice oxygen of metal oxides, the latter being re-oxidized
by gas-phase oxygen [35–40]. Santos et al. related the lattice
oxygen of manganese oxides to reactivity of manganese oxides
towards the oxidation of VOCs [37]. Wang et al. suggested that
higher reactivity of manganese oxide octahedral molecular sieve
catalysts can be assigned to the improvement of oxygen activa-
tion, which led to higher reactivity of ethanol and acetaldehyde
[38]. In the case of MnO2, lattice oxygen of the catalyst can
be consumed by reaction with toluene and then be replenished
by oxygen from the gas phase [41–43]. Thus, lattice oxygen
of manganese oxides can be of great importance in the cat-
alytic oxidation of toluene. The low-temperature reducibility
reflects the reactivity of lattice oxygen in MnO2. As shown in
Fig. 4(A) and (B), the lattice oxygen reactivity decreased in the
order of rod-like ␣-MnO2 > flower-like ␧-MnO2 > dumbbell-like ␤-
MnO2 > bulk-MnO2, in basic agreement with the sequence of the
catalytic performance of these samples (flower-like ␧-MnO2 > rod-
like ␣-MnO2 > dumbbell-like ␤-MnO2 > bulk-MnO2). The formation
of surface oxygen vacancies has also an important role to play.
These defective sites can be the active centers in the catalytic
reaction since oxygen of defective oxides tends to be readily
released and transferred, which would enhance the reactivity of
the catalyst in oxidation of VOCs [44,45]. It is known that oxygen
vacancy, reducibility, and surface area of a transition metal oxide
are important factors influencing its catalytic activity [10,13–17].
The presence of oxygen vacancies favors the activation of oxy-
gen molecules to active oxygen adspecies. A good reducibility of
the catalyst can provide a facile redox process that would lead
to an enhanced catalytic performance [10,13,46]. The surface area
(ca. 132 m2/g) of the dumbbell-like ␤-MnO2 sample was much
higher than those (30–53 m2/g) of the rod-like ␣-MnO2 and flower-
like ␧-MnO2 samples, but the catalytic performance of the former
was inferior to that of the latter two. This result indicates that
surface area was not the key factor governing the catalytic per-
formance. That is to say, other factors should be responsible for
the better performance of the rod-like ␣-MnO2 and flower-like
␧-MnO2 catalysts. As revealed from the XPS and H2-TPR studies,
the rod-like ␣-MnO2 and flower-like ␧-MnO2 catalysts possessed
higher oxygen adspecies concentrations (relevant to the surface
oxygen vacancy densities) and better low-temperature reducibil-
ity than the dumbbell-like ␤-MnO2 sample, in good agreement
with their catalytic activity sequence. Therefore, we conclude
that the higher oxygen adspecies concentrations and better low-
temperature reducibility mainly accounted for the better catalytic
performance of the rod-like ␣-MnO2 and flower-like ␧-MnO2 sam-
ples for toluene combustion.
3.5. Kinetic parameters
In the past years, the kinetics of catalytic oxidation of VOCs has
been investigated. After studying the oxidation of butyl acetate over
the AgZSM-5 catalyst, Wong et al. believed that it was first-order
toward butyl acetate concentration and zero toward oxygen con-
centration [47]. By supposing a first-order kinetics with respect to
toluene and a zero-order kinetics with respect to oxygen, Alifanti
et al. obtained good linear Arrhenius plots for toluene oxidation

212
F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
␣-MnO2 and flower-like ␧-MnO2 catalyst outperforming the
dumbbell-like ␤-MnO2 catalyst. Under the conditions of toluene
concentration = 1000 ppm, toluene/O2 molar ratio = 1/400, and
SV = 20,000 mL/(g h), the T90% value was 238, 229, and 241 ◦C
over rod-like ␣-MnO2, flower-like ␧-MnO2, and dumbbell-like
␤-MnO2, respectively. The apparent activation energies of the
rod-like ␣-MnO2, flower-like ␧-MnO2, and dumbbell-like ␤-MnO2
catalysts were in the range of 20–26 kJ/mol. It is concluded that
the higher oxygen adspecies concentrations and better low-
temperature reducibility were the main factors responsible for the
better catalytic performance of the rod-like ␣-MnO2, flower-like
␧-MnO2, and dumbbell-like ␤-MnO2 materials for the combustion
of toluene.
-1
-2
-3
-4
-5
-6
ulk MnO2
E
ꢀ
rod-like -MnO
α
2
Ⴗ
Ⴃ
flower-like -MnO
ε
2
dumbbell-like -MnO
β
2
Acknowledgments
1.5
2
2.5
3
3.5
The work described above was supported by the NSF of Beijing
Municipality (Grant No. 2102008), the NSF of China (Grant Nos.
20973017 and 21077007), the National High-Tech Research and
Development (863) Program of China (Grant No. 2009AA063201),
the Creative Research Foundation of Beijing University Technol-
ogy (Grant Nos. 00500054R4003 and 005000543111501), and the
Funding Project for Academic Human Resources Development
in Institutions of Higher Learning under the Jurisdiction of Bei-
jing Municipality (Grant Nos. PHR201007105 and PHR201107104).
CTAU thanks the Foundation of the Hong Kong Baptist University
(Grant No. FRG2/09-10/023) for financial support. We also thank
Mrs. Jianping He (State Key Laboratory of Advanced Metals and
Materials, University of Science & Technology Beijing) for doing the
SEM analysis.
1000/T (K-1)
Fig. 6. Arrhenius plots for total oxidation of toluene on the rod-like ␣-MnO2, flower-
like ␧-MnO2, dumbbell-like ␤-MnO2, and bulk-MnO2 catalysts.
over the ceria-zirconia-supported LaCoO3 catalyst [48]. Therefore,
it is reasonable to assume that the oxidation of toluene in the pres-
ence of excess oxygen (toluene/O2 molar ratio = 1/400) would obey
a first-order reaction mechanism with respect to toluene concen-
tration (c):
ꢀ
ꢁ
−Ea
r = −kc = −A exp
c
RT
where r is the reaction rate (mol/s), k the rate constant (s−1), A
the pre-exponential factor, and Ea the apparent activation energy
(kJ/mol). The k values could be calculated from the reaction rates
and reactant conversions at different SV and reaction temperatures.
Fig. 6 shows the Arrhenius plots for the oxidation of toluene at
toluene conversion <30% (at which the reaction temperature range
was 50–220 ◦C for the rod-like ␣-MnO2, flower-like ␧-MnO2, and
dumbbell-like ␤-MnO2 catalysts and 150–240 ◦C for the bulk-MnO2
catalyst). According to the slopes of the Arrhenius plots, one can
calculate the rate constants, pre-exponential factors, and appar-
ent activation energies of toluene oxidation over these catalysts,
as summarized in Table 2. The Ea value (74.3 kJ/mol) of the bulk-
MnO2 catalyst was much higher than those (19.6–26.1 kJ/mol) of
the ␣-, ␧-, and ␤-MnO2 catalysts, with the rod-like ␣-MnO2 catalyst
exhibiting the lowest Ea value (19.6 kJ/mol). The Ea values obtained
over the ␣-, ␧-, and ␤-MnO2 catalysts for toluene oxidation were
much lower than those (73–89 kJ/mol) over the MxFe3−xO4 (M = Ni,
Mn; x = 0.5–0.65) catalysts [49], those (51–79 kJ/mol) over the
10–20 wt.% LaCoO3/Ce1−xZrxO2 (x = 0–0.2) catalysts [48], and that
(62 kJ/mol) over the 7 wt.% Pt/16 wt.% Ce0.64Zr0.15Bi0.21O1.895/␥-
Al2O3 catalyst [1]. All of the results explain why the ␣-, ␧-, and
␤-MnO2 catalysts performed excellently in catalyzing the complete
oxidation of toluene at low temperatures.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.apcata.2012.05.016.
References
[1] T. Masui, H. Imadzu, N. Matsuyama, N. Imanaka, J. Hazard. Mater. 176 (2010)
1106–1109.
[2] V.P. Santos, S.A.C. Carabineiro, P.B. Tavares, M.F.R. Pereira, J.J.M. Órfão, J.L.
Figueiredo, Appl. Catal. B 99 (2010) 198–205.
[3] B. Solsona, T.E. Davies, T. Garcia, I. Vázquez, A. Dejoz, S.H. Taylor, Appl. Catal. B
84 (2008) 176–184.
[4] C. Lahousse, A. Bernier, E. Gaigneaux, P. Ruiz, P. Grange, B. Delmon, Stud. Surf.
Sci. Catal. 110 (1997) 777–785.
[5] T.K. Tseng, H. Chu, H.H. Hsu, Environ. Sci. Technol. 37 (2003) 171–176.
[6] H. Rotter, M.V. Landau, M. Herskowitz, Environ. Sci. Technol. 39 (2005)
6845–6850.
[7] F. Wyrwalski, J.F. Lamonier, S. Siffert, A. Aboukaïs, Appl. Catal. B 70 (2007)
393–399.
[8] C. Lahousse, A. Bernier, P. Grange, B. Delmon, P. Papaefthimiou, T. Ioannides, X.
Verykios, J. Catal. 178 (1998) 214–222.
[9] M.A. Peluso, J.E. Sambeth, H.J. Thomas, React. Kinet. Catal. Lett. 80 (2003)
241–248.
[10] L. Lamaita, M.A. Peluso, J.E. Sambeth, H.J. Thomas, Appl. Catal. B 61 (2005)
114–119.
[11] L. Lamaita, M.A. Peluso, J.E. Sambeth, H.J. Thomas, G. Mineli, P. Porta, Catal.
Today 107/108 (2005) 133–138.
4. Conclusions
[12] M.A. Peluso, E. Pronsato, J.E. Sambeth, H.J. Thomas, G. Busca, Appl. Catal. B 78
(2008) 73–79.
[13] J.Q. Torres, J.-M. Giraudon, J.-F. Lamonier, Catal. Today 176 (2011) 277–280.
[14] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, Top. Catal. 52 (2009)
470–481.
[15] L. Zhou, J.H. He, J. Zhang, Z.C. He, Y.C. Hu, C.B. Zhang, H. He, J. Phys. Chem. C 115
(2011) 16873–16878.
[16] L. Zhou, J. Zhang, J.H. He, Y.C. Hu, H. Tian, Mater. Res. Bull. 46 (2011) 1714–1722.
[17] S.C. Kim, W.G. Shim, Appl. Catal. B 98 (2010) 180–185.
[18] Y.S. Xia, H.X. Dai, L. Zhang, J.G. Deng, H. He, C.T. Au, Appl. Catal. B 100 (2010)
229–237.
The rod-like tetragonal ␣-MnO2, flower-like hexagonal ␧-MnO2,
and dumbbell-like tetragonal ␤-MnO2 catalysts could be prepared
using the hydrothermal or water-bathing method. Although the
rod-like ␣-MnO2 and flower-like ␧-MnO2 samples displayed a
lower surface area (30–53 m2/g) than the dumbbell-like ␤-MnO2
(ca. 132 m2/g), the form two samples possessed higher oxygen
adspecies concentrations and better low-temperature reducibil-
ity than the latter sample. The rod-like ␣-MnO2, flower-like
␧-MnO2, and dumbbell-like ␤-MnO2 catalysts showed much
better activities than the bulk-MnO2 catalyst, with the rod-like
[19] Y.S. Xia, H.X. Dai, H.Y. Jiang, J.G. Deng, H. He, C.T. Au, Environ. Sci. Technol. 43
(2009) 8355–8360.

F. Shi et al. / Applied Catalysis A: General 433–434 (2012) 206–213
213
[20] Y.S. Xia, H.X. Dai, H.Y. Jiang, L. Zhang, J.G. Deng, C.T. Au, J. Hazard. Mater. 186
(2011) 84–91.
[21] J.G. Deng, L. Zhang, H.X. Dai, Y.S. Xia, H.Y. Jiang, H. Zhang, H. He, J. Phys. Chem.
C 114 (2010) 2694–2700.
[22] Y.S. Xia, H.X. Dai, H.Y. Jiang, L. Zhang, Catal. Commun. 11 (2010) 1171–1175.
[23] S.H. Liang, F. Teng, G. Bulgan, R.L. Zong, Y.F. Zhu, J. Phys. Chem. C 112 (2008)
5307–5315.
[24] B.X. Li, G.X. Rong, Y. Xie, L.F. Huang, C.Q. Feng, Inorg. Chem. 45 (2006)
6404–6410.
[25] W. Xiao, D.L. Wang, X.W. Lou, J. Phys. Chem. C 114 (2010) 1694–1700.
[26] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, Handbook of X-ray photo-
electron spectroscopy, in: J. Chastain, R.C. King Jr. (Eds.), Physical Electronics,
Eden Prairie, 1995, p. 78.
[27] J.C. Carver, G.K. Schweitzer, J. Chem. Phys. 57 (1972) 973–982.
[28] S. Ponce, M.A. Pen˜a, J.L.G. Fierro, Appl. Catal. B 24 (2000) 193–205.
[29] S.D. Gardner, G.B. Hoflund, M.R. Davidson, D.R. Schryer, J. Schryer, B.T.
Upchurch, E.J. Kielin, Langmuir 7 (1991) 2135–2139.
[30] K.D. Chen, S.B. Xie, A.T. Bell, E. Iglesia, J. Catal. 198 (2001) 232–242.
[31] H.X. Dai, A.T. Bell, E. Iglesia, J. Catal. 221 (2004) 491–499.
[32] M.P. Irusta, M. Pina, J. Menéndez, Santamaría, J. Catal. 179 (1998) 400–412.
[33] S. Scirè, S. Minicò, C. Crisafulli, C. Satriano, A. Pistone, Appl. Catal. B 40 (2003)
43–49.
[35] Y.F. Chang, J.G. McCarty, Catal. Today 30 (1996) 163–170.
[36] B. Grbic, N. Radic, B. Markovic, P. Stefanov, D. Stoychev, Ts Marinova, Appl. Catal.
B 64 (2006) 51–56.
[37] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, Appl. Catal. B 99 (2010)
353–363.
[38] R.H. Wang, J.H. Li, Environ. Sci. Technol. 44 (2010) 4282–4287.
[39] V. Roche, L. Mazri, A. Boréave, M.H. Ta, L. Retailleau-Mevel, A. Giroir-Fendler,
P. Vernoux, J.P. Deloume, Appl. Catal. A 385 (2010) 163–169.
[40] Y.S. Wu, Y.X. Zhang, M. Liu, Z.C. Ma, Catal. Today 153 (2010) 170–175.
[41] M. Baldi, E. Finocchio, F. Milella, G. Busca, Appl. Catal. B 16 (1998) 43–51.
[42] C. Cellier, V. Ruaux, C. Lahousse, P. Grange, E.M. Gaigneaux, Catal. Today 117
(2006) 350–355.
[43] J. Luo, Q. Zhang, J. Garcia-Martinez, S.L. Suib, J. Am. Chem. Soc. 130 (2008)
3198–3207.
[44] X. Chen, Y.F. Shen, S.L. Suib, C.L. O’Young, J. Catal. 197 (2001) 292–302.
[45] Q.H. Tang, T. Liu, Y.H. Yang, Catal. Commun. 9 (2008) 2570–2573.
[46] A.R. Gandhe, J.S. Rebello, J.L. Figueiredo, J.B. Fernandes, Appl. Catal. B 72 (2007)
129–135.
[47] C.T. Wong, A.Z. Abdullah, S. Bhatia, J. Hazard. Mater. 157 (2008) 480–489.
[48] M. Alifanti, M. Florea, S. Somacescu, V.I. Pârvulescu, Appl. Catal. B 60 (2005)
33–39.
[49] M. Florea, M. Alifanti, V.I. Pârvulescu, D. Mihaila-Tarabasanu, L. Diamandescu,
M. Feder, C. Negrila, L. Frunza, Catal. Today 141 (2009) 361–366.
[34] A. Musialik-Piotrowska, H. Landmesser, Catal. Today 137 (2008) 357–361.