Effect of reducing agent on the structure and activity of manganese oxide octahedral molecular sieve (OMS-2) in catalytic combustion of o-xylene

Article abstract of DOI:10.1016/j.cattod.2016.05.024

Manganese oxide octahedral molecular sieve (OMS-2) with different microstructures were synthesized via simple wet chemical methods of KMnO₄/benzyl alcohol and KMnO₄/Mn(NO₃)₂ at room-temperature (the products were denoted as B-OMS-2 and N-OMS-2, respectively). The physicochemical properties of the materials were characterized using numerous analytical techniques. The catalytic activities of the catalysts were evaluated for the complete catalytic oxidation of a typical volatile organic compound (VOCs), o-xylene. It was found that B-OMS-2 presented a loose structure, and contained almost 100% Mn⁴⁺, while N-OMS-2 possessed a mixture of Mn⁴⁺ and Mn³⁺. H₂-TPR and O₂-TPO analyses showed that B-OMS-2 exhibited good low-temperature reducibility and high oxygen exchange ability with gas phase oxygen. The microstructure difference was caused by different reducing reagents used in the synthesis. Benzyl alcohol might adsorb on the surface of MnO₂ nuclei acting as a ligand and/or structure-directing agent, and the desorption of the organic compound led to the formation of bulk oxygen vacancy in B-OMS-2. B-OMS-2 could convert 100% o-xylene into CO₂ at 190 °C at a space velocity of 8000 h^?1, 50 °C lower than N-OMS-2. The excellent catalytic performance of B-OMS-2 might be caused by its bulk oxygen vacancy, potent reducibility and high re-oxidation ability. It is believed that B-OMS-2 is a promising catalyst for the elimination of VOCs from air.

Full text of DOI:10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Effect of reducing agent on the structure and activity of manganese
oxide octahedral molecular sieve (OMS-2) in catalytic combustion of
o-xylene
a
a
b
a
a
a,∗
Yinsu Wu , Rui Feng , Chaojie Song , Shengtao Xing , Yuanzhe Gao , Zichuan Ma
a
College of Chemistry and Material Science, Hebei Normal University, Shijiazhuang 050024, China
Energy, Mining and Environment, National Research Council Canada, 4250 Wesbrook Mall, Vancouver, BC, V6T 1W5, Canada
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Manganese oxide octahedral molecular sieve (OMS-2) with different microstructures were synthesized
via simple wet chemical methods of KMnO4/benzyl alcohol and KMnO4/Mn(NO3)2 at room-temperature
Received 25 December 2015
Received in revised form 22 March 2016
Accepted 5 May 2016
(the products were denoted as B-OMS-2 and N-OMS-2, respectively). The physicochemical properties of
the materials were characterized using numerous analytical techniques. The catalytic activities of the
catalysts were evaluated for the complete catalytic oxidation of a typical volatile organic compound
Available online xxx
(
VOCs), o-xylene. It was found that B-OMS-2 presented a loose structure, and contained almost 100%
Keywords:
OMS-2
Oxygen vacancy
o-Xylene
Catalytic combustion
4
+
4+
3+
Mn , while N-OMS-2 possessed a mixture of Mn and Mn . H2-TPR and O2-TPO analyses showed that
B-OMS-2 exhibited good low-temperature reducibility and high oxygen exchange ability with gas phase
oxygen. The microstructure difference was caused by different reducing reagents used in the synthesis.
Benzyl alcohol might adsorb on the surface of MnO2 nuclei acting as a ligand and/or structure-directing
agent, and the desorption of the organic compound led to the formation of bulk oxygen vacancy in B-OMS-
◦
−
1
◦
2
. B-OMS-2 could convert 100% o-xylene into CO2 at 190 C at a space velocity of 8000 h , 50 C lower than
N-OMS-2. The excellent catalytic performance of B-OMS-2 might be caused by its bulk oxygen vacancy,
potent reducibility and high re-oxidation ability. It is believed that B-OMS-2 is a promising catalyst for
the elimination of VOCs from air.
©
2016 Elsevier B.V. All rights reserved.
1
. Introduction
decreasing the particle size or increasing the surface area [11], dop-
ing K+ in OMS-2 channels [5], substituting part of the Mn with V,
Cu, and Fe ions [12], controlling the morphology of OMS-2 [10],
and increasing the oxygen vacancy defect concentration in OMS-2
[7,13].
It is commonly accepted that OMS-2 catalyzes VOCs oxida-
tion via the Mars-van Krevelen mechanism. Organic molecules
adsorbed on the catalyst surface are oxidized by surface lat-
tice oxygen, and the resultant oxygen vacancies are subsequently
replenished by gas-phase O2 [4]. According to this mechanism, the
reducibility or activity of lattice oxygen and re-oxidation ability
of OMS-2 appear to be two critical parameters determining their
catalytic performance for VOC oxidation.
Volatile organic compounds (VOCs) such as benzene, toluene,
and xylene (BTX) are harmful to the environment. It is highly desir-
able to remove these pollutants [1–3]. Complete catalytic oxidation
techniques converting organic compound contaminants into CO2
and H O are an effective pathway to control VOC emissions [4].
2
Manganese oxide, such as MnO2 has been considered as an
efficient catalyst for complete catalytic oxidation of VOCs [5–8].
The cryptomelane-type octahedral molecular sieve (OMS-2), MnO2
with a one-dimensional tunnel structure composed of 2 × 2 edge
shared MnO octahedral chains, has received appreciable attention
6
due to its excellent catalytic performance [9]. However, at low tem-
perature, OMS-2 has lower catalytic activity toward benzene type
materials’ oxidation than precious metals [10]. To improve OMS-2’s
catalytic activity, several strategies have been proposed including
In the OMS-2 synthesis, reactants usually influence OMS-2
microstructure, and oxygen vacancy concentration and types,
which affect the mobility of lattice oxygen, the re-oxidation abil-
ity of OMS-2 and interface between VOCs, gas oxygen and OMS-2
[
13,14]. The microstructure of MnO2 is mainly determined by the
rate of nuclei growth (governing the formation of building blocks
for desired structures) and the extent to which the building blocks
∗ Corresponding author.
E-mail addresses: mazc@vip.163.com, mazc@hebtu.edu.cn (Z. Ma).
http://dx.doi.org/10.1016/j.cattod.2016.05.024
0
920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
2
aggregate or assemble. The aggregation process plays an important
role in tailoring the features of the morphology and porous struc-
ture in the product. Usually, the growth of primary structures is
modulated by preferential adsorption of anions, solvent molecules,
and reaction conditions such as temperature etc. [15]. Usually,
OМS-2 materials are prepared by reducing KMnO4 with various
inorganic compounds such as Mn(II) salts [16–18]. Recently, OMS-
a
310
211
110 200
411
521
301
541
220
600
002
2
with porous and/or three-dimensional frameworks was prepared
by reducing KMnO4 with some oxygen-containing organic com-
pounds, such as benzyl alcohol and ascorbic acid [19,20]. In a
b
KMnO /organic reducing agent system, the organic reducing agent
4
probably not only result in a two-phase synthesis process to control
the reaction rate, but also can act as structure-directing to control
the growth of nuclei and further influence its microstructure. So far
a comparison between the two methods has not been studied yet.
In this paper, we studied the effect of reducing agent (organic
compound, and inorganic compound,) on OMS-2 structure and
catalytic activity toward VOCs complete oxidation. The OMS-2 syn-
thesized via benzyl alcohol at room temperature was denoted as
B-OMS-2, and that prepared by Mn(NO3)2 at the same condition
was denoted as N-OMS-2. Their properties were characterized and
catalytic activities toward the complete catalytic oxidation of o-
xylene were tested. The correlation between catalytic performance
and microstructure was analyzed.
1
0
20
30
40
50
60
70
80
2 (degree)
Fig. 1. XRD patterns of (a) B-OMS-2; (b) N-OMS-2.
(HRTEM). XRD measurements were carried out in a Bruker AXS-
D8 -ADVANCE using Cu K␣ radiation and operated at 40 kV and
40 mA. The nitrogen adsorption–desorption isotherm was obtained
◦
at −196 C using a Quantachrome NOVA 4000e instrument.
Brunauer–Emmett–Teller (BET) method was used to calculate the
surface area. The pore diameter distributions were calculated
from desorption branches using the Barrett–Joyner–Halenda (BJH)
method. Thermal stability and phase transitions were analyzed by
2
. Experimental
◦
TGA using a STA449F3 thermogravimetric analyzer (25–820 C) at
a heating rate of 10 C/min in air atmosphere. The oxidation state of
2.1. Synthesis of B-OMS-2 and N-OMS-2
◦
Mn in the product was measured by XPS using a SHIMADZU ESCA-
2
0 mmol benzyl alcohol was added dropwise to a 20 mmol
3
400 spectrometer (Mg K␣ radiation). The C1 s (284.6 eV) binding
KMnO4 solution. After the mixture was vigorously stirred at room
temperature for 24 h, a black product was obtained. The product
was washed several times with deionized water and absolute alco-
hol to remove any possible residual reactants, dried at 110 C for
about 12 h, and then calcined in air at 400 C for 6 h. The product
energy (BE) was used as internal reference. The binding energy
regions of Mn2p and O1s were 630–670 eV, 522–542 eV respec-
tively. The morphology of the material was observed using SEM
on Hitachi S-4800 microscope. The product was further examined
with transmission electron microscopy (TEM; H-600) and high-
resolution transmission electron microscopy (HRTEM; JEM-2100,
JEOL, Japan) operated at 200 kV.
◦
◦
was denoted as B-OMS-2.
For comparison, OMS-2 was also prepared by reducing KMnO4
with Mn(NO3)2 at the same conditions as that used in KMnO4-
benzyl alcohol. The product was denoted N-OMS-2.
2
.4. H -TPR and O -TPO
2 2
2.2. Catalytic activity evaluation
H2 temperature-programmed reduction (H -TPR) and O2
2
temperature-programmed oxidation (O -TPO) were conducted on
2
Catalytic activity was measured in a 4 mm i.d. quartz tubular
a PCA-1200 multifunctional adsorption apparatus equipped with a
TCD detector. 0.04 g of the OMS-2 sample was loaded in a quartz
reactor. A silica gel particle trap was placed in front of the detector
to adsorb water. Before the TPR analysis, the OMS-2 samples were
reactor. 0.50 g of catalyst supported by quartz wool was placed in
the center of the reactor. Simulated air (20 vol.% O , 80 vol.% N )
2
2
containing 500 ppm of o-xylene in continuously passed through the
−
1
catalyst bed at a flow rate of 50 mL/min (W/F = 0.60 g s mL ), corre-
◦
pretreated at 300 C in Ar at a flow rate of 40 mL/min for 1 h. TPR
−1
sponding to a GHSV of 8000 h . Here, W/F is defined as the catalyst
weight divided by the gas flow rate. CO2 was the only detectable
C-containing product, which initially passed through the TDX-01
stainless steel packed column, and then converted to methane in
a reformer furnace; no significant amount of any partial oxidation
product was detected in the effluent. The reactant and product were
analyzed with an on-line gas chromatograph equipped with two
was performed by heating the pretreated OMS-2 sample at a rate
◦
◦
of 10 C/min to 600 C in 10 vol% H /Ar at a flow rate of 30 mL/min.
2
◦
After the TPR analysis, the OMS-2 sample was treated in Ar at 25 C
for 0.5 h. Then TPO analysis was performed by heating the OMS-2
◦
◦
sample at a rate of 10 C/min to 300 C (avoiding the decomposition
of O2 in the OMS-2) in a 5 vol% O /He at a flow rate of 30 mL/min.
2
flame ionization detectors (FID) in series. The yield of CO2 (YCO2
was calculated according to the following formula:
)
3. Results and discussion
YCo2 = 100 × Co2 out/8/o − xylenein
3.1. Structural and textural properties of B-OMS-2 and N-OMS-2
Powder XRD analysis (Fig. 1) shows that both B-OMS-2 and
N-OMS-2 materials display typical manganese oxide octahedral
molecular sieve (OMS-2) crystal phase (JCDPS 44-0141, tetragonal,
I4/m, a = b = 0.978 nm, c = 0.286 nm). The pattern of the OMS-2 can
be indexed to the planes of {110}, {200}, {220}, {310}, {211}, {411},
2
.3. Materials characterization
The materials were characterized by powder X-ray diffraction
XRD), N2 sorptometry, thermogravimetric analysis (TGA), X-ray
photoelectron spectrometry (XPS), scanning electron microscopy
SEM) and high resolution transmission electron microscopy
(
◦
◦
◦
◦
{600}, {521}, {002} and {541} at 2ꢀ = 12.8 , 18.1 , 25.8 , 37.6 ,
◦ ◦ ◦ ◦ ◦
(
41.9 , 49.8 , 60.1 , 65.7 and 69.4 respectively [16]. The intensity of
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
3
zyl alcohol, which gave a chance for benzyl alcohol molecules to
adsorb on the MnOx. And the adsorbate hindered the aggregation of
MnOx nuclei. When benzyl alcohol was removed in the calcination
process, large pore was formed.
0
0
0
0
0
.7
.6
.5
.4
.3
1
20
00
—ƾ— a
Ʒ— b
—
1
8
6
4
2
0
0
0
0
0
0.2
The nitrogen adsorption isotherms and the pore size distribu-
tions are shown in Fig. 2. It can be concluded that N-OMS-2 is
typically type-IV with hysteresis loops, which means that the mate-
rial presents a mesoporous structure, with a pore size distributions
concentrated at (50.0 nm). Whereas, the B-OMS-2 exhibited large
mesopores or macropores with a widepore size distributions con-
centrated at 50–100 nm.
Fig. 3 showed the SEM images of the N-OMS-2 and B-OMS-2.
The texture and morphology of the two materials are significantly
different. The N-OMS-2 consists of large and closely packed spher-
ical particles with particle size of 200–500 nm (Fig. 3b left). The
B-OMS-2 presents a loose structure, which is composed of a series
of unequal-sized nanocubes. (Fig. 3a left). EDX spectra (Fig. 3a,b
right) reveal that the two materials are composed of Mn, O, as well
as K. B-OMS-2 has higher K contents than N-OMS-2.
0
0
.1
.0
0
50 100 150 200 250 300
Pore diameter(nm)
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure(P/P0)
Fig. 2. N2 adsorption–desorption isotherms and Barret–Joyner–Halenda (BJH) pore
size distribution plots (inside) of (a) B-OMS-2; (b) N-OMS-2.
the {200} plane of B-OMS-2 was slightly higher than that of {110}
plane, indicating more {200} plane exposure, which is consistent
with the HRTEM result (see below in Fig. 4b).
TEM (Fig. 4) further confirmed the packed nanoparticles mor-
phology of N-OMS-2 and loose nanocubes morphology of the
B-OMS-2. The HRTEM images showed that the N-OMS-2 nanopar-
ticles maintained two kinds of interlayer distance of 0.49 nm and
The BET surface area and pore volume of B-OMS-2 are
2
−1 −1
3
4
0.65 m g and 0.17 cm g , respectively, similar to those of N-
0
.69 nm, respectively. These lattice distances agree well with the
2
−1
3
−1
OMS-2 (40.98 m g and 0.17 cm g ). However, the pore size of
N-OMS-2 is only 46.52 nm, significantly smaller than that of B-
OMS-2 (73.98 nm). These results indicated that different reducing
agent led to a difference in pore size. Chen et al. found that the
aggregation process plays an important role in tailoring the fea-
tures of the porous structure in the product [21]. For B-OMS-2, the
formation of MnOx nuclei is slow due to the hydrophobicity of ben-
lattice spacing of the {200} and {110} crystal planes. While, B-OMS-
2
nanocubes showed only {200} crystal plane.
Chemical species located in the region near the catalysts sur-
face is characterized by XPS. The deconvolution of Mn 2p3/2 and Mn
2
p1/2 signal is known to be useful to distinguish the oxidation states
of Mn␦ and the results are shown in Fig. 5. The binding energies
+
Fig. 3. SEM images and EDX patterns of (a) B-OMS-2; (b) N-OMS-2.
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
4
Table 1
XPS and AOS results of B-OMS-2 and N-OMS-2 samples.
Samples
BE(eV)
Mn3+
Mn4+/(Mn3+ + Mn4+)(%)
BE(eV)
O␣/(O␣+O␤+O␥)
AOSa
Mn4+
O␣
O␤
O␥
B-OMS-2
N-OMS-2
0
642.42
642.23
100
61.84
529.70
529.35
531.45
531.83
–
78.96
32.28
3.96
3.55
641.73
533.40
a
The AOS is obtained by the results of chemical titration.
of Mn 2p3/2 at 639.9–640.3 eV, 640.5–641.9 eV and 642.0–642.5 eV
2
+
3+
4+
can be ascribed to Mn , Mn and Mn ions, respectively [7,10].
Table 1 summarized the percentage of different Mn oxidation
states, and different oxygen types. It can be seen that the two mate-
4+
3+
rials have significantly different Mn /Mn and Olatt/Osurf ratios,
which was caused by the different reducing agent used, For B-
4
+
OMS-2, only Mn can be detected on its surface. Although the
shape of Mn2p 3/2 signal for both samples is very similar, the
signal intensity of N-OMS-2 is weaker than that of B-OMS-2. More-
over, for N-OMS-2, there is an obvious shift of the Mn2p3/2 peak
towards lower binding energy (from 642.0 eV to 641.6 eV), which
4
+
3+
indicates that Mn and Mn species are coexisted in the sample.
The Mn /Mn + Mn3+ ratio is shown in Table 1. From chemical
4
+
4+
titration, the average oxidation state (AOS) of B-OMS-2 and N-OMS-
2
were calculated to be 3.55 and 3.96, respectively. Based on the
principle of electroneutrality, it can be deduced that there is some
surface oxygen vacancy on the N-OMS-2 surface. According to the
previous report [22], this kind of oxygen vacancy was defined as
3+
Mn -oxygen vacancy associates. The result is agreement with the
following O1 s analysis.
The chemical environment of oxygen in metal oxide catalysts
played an important role in their catalytic properties. Fig. 5(B)
shows the XPS spectra of O1 s in B-OMS-2 and N-OMS-2. For N-
OMS-2, three peaks were observed, attributing to three types of
oxygen species: the peak at BE of 529–530 eV was ascribed to
the lattice oxygen in Mn-O-Mn (denoted as O␣), the peak at BE
of 531–532 eV was ascribed to surface adsorbed oxygen, surface
oxygen ions with low coordination Mn-O-H and oxygen vacancies
Fig. 4. TEM and HRTEM images of (a) B-OMS-2; (b) N-OMS-2.
formation of more labile oxygen species in the sample. Therefore,
the BE of some lattice oxygen might be close to surface adsorbed
oxygen. In addition, it is found that the binding energy of Mn 2p3/2
and the lattice oxygen shift to higher level compared to those of
N-OMS-2. Qu et al. investigated the effects of Ag on Mn/SBA-15
catalysts, and found that Ag enters into MnO2 phase and caused
the presence of a defect in manganese, leading to an increase in the
binding energy of O1s [7]. Therefore, it can be deduced that there
are some structural defects in B-OMS-2.
(
denoted as O␤), and the peak at a higher BE (above 533 eV) cor-
responded to adsorbed molecular water (denoted as O␥) [23]. The
relative amount of O␣, O␤ and O␥ in N-OMS-2 were about 1/3 for
each respectively. As for B-OMS-2, it possesses almost 100% Mn4+
.
However, the peak deconvolution of O1 s indicate that both O␣ and
O␤ species exist in it. The relative amount of O␣ in B-OMS-2 is
8.96% (see Table 1). 100% Mn4 content generally corresponds to
+
3.2. Thermal stability of B-OMS-2 and N-OMS-2
7
the absence of O␤. The reason may be that the bond of Mn-O is weak
due to the existence of bulk oxygen vacancy. It might lead to the
The TGA profiles of N-OMS-2 and B-OMS-2 in the temperature
range of 25 and 800 C in air are presented in Fig. 6. Usually, thermal
◦
A
Mn 2p
B
3/2
Mn 2p
1/2
a
b
a
O
4+
Mn
O
O
3+
Mn
4+
Mn
3+
Mn
b
6
40
645
650
Bind energy(eV)
655
660
5
28
530
532
534
536
Bind energy(eV)
Fig. 5. Mn 2p XPS spectra (A) and O 1s XPS spectra (B) of B-OMS-2 (a) and N-OMS-2 (b).
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
5
◦
It is worth noting that a weight increase occurred after 450 C for
1
00
B-OMS-2. Fig. 6 shows that B-OMS-2 achieved its maximum weight
9
8
increase (0.2%) at 690 ◦C due to an oxygen gain, and a quick weight
a
◦
loss due to oxygen evolution subsequently occurred at 710 C. N-
9
6
OMS-2 did not exhibit any weight increase during the whole TGA
profile; only weight losses of water and oxygen were observed. The
results indicated that B-OMS-2 might have high oxygen exchange
ability in air due to the existence of bulk oxygen vacancies, which is
also supported by the subsequent O -TPO measurement result (see
Fig. 7B below). This is in agreement with Zhang and Suib’s results
9
4
93%
9
2
9
0
b
2
8
8
[
26].
87%
8
6
1
00 200 300 400 500 600 700
3
.3. H -TPR and O -TPO of B-OMS-2 and N-OMS-2
2
2
Temperature(
)
Fig. 6. TGA curves of OMS-2 treatment under air: (a) B-OMS-2; (b) N-OMS-2.
H2 temperature-programmed reduction (H2-TPR) has been used
extensively in the literature to characterize the surface and bulk
oxygen reducibility of oxide materials.
According to the Mars-van Krevelen mechanism, the redox
analysis of manganese oxides (MnO )in air presents three thermal
property of MnO2 is also related to its re-oxidation ability. In this
2
◦
features: (a) loss of physically adsorbed molecular water in the tem-
work, we interrupted the H -TPR at 600 C, the samples were re-
2
◦
◦
perature range 100–200 C; (b) loss of chemically bound water and
oxidized with 5% O /He at 25 C, and then O -TPO analysis was
2
2
◦
a change of crystalline phase between 200 and 400 C; and (c) the
performed.
◦
transformation of MnO2 to Mn O3 around 550 C. The weight loss
above 800 C is attributed to the phase transformation of MnO2 to
Fig. 7A showed the H -TPR profiles of B-OMS-2 and N-OMS-
2
2
◦
2. The reduction curve of N-OMS-2 is the typical feature of two
Mn O [24,25]. The total (physically and chemically bound) water
step reduction of MnO . The low-temperature reduction peak cen-
3
4
2
◦
losses for B-OMS-2 and N-OMS-2 were 2.0% and 5.0%, respectively.
The water losses of N-OMS-2 were much larger than that of B-
OMS-2, in accordance with the XPS results that show abundant
water molecules exist in N-OMS-2. Moreover, in the temperature
tered at ca. 273 C represented the reduction of MnO to Mn O3
2
2
◦
and the high-temperature reduction peak at 343 C referred to the
further reduction of Mn O3 to MnO [14]. However, for B-OMS-2,
2
◦
only one main reduction band centered at 315 C along with a wide
◦
◦
◦
range of 400–800 C, N-OMS-2 exhibited a more obvious crystalline
shoulder peak ranging from 230 C to 270 C was observed. The
shoulder peak reveals the presence of the labile oxygen species with
weak Mn–O bond strength. This may support the XPS result, which
indicates the existence of some surface oxygen species. The main
phase change from MnO2 to Mn O3 and then to Mn O4 than that
2
3
of B-OMS-2.
Theoretically, the weight loss for the phase transformation of
MnO2 to Mn O is 13%. This is what was observed for N-OMS-2.
reduction curve indicates the direct reduction of MnO to MnO [14].
3
4
2
However, only 7% weight loss was observed for B-OMS-2, 1/2 of
the theoretic value. The lower weight loss suggested that there are
lower valence state Mn ions and/or oxygen vacancies in B-OMS-2.
In order to prove this hypothesis, B-OMS-2 was calcined at tem-
TPR results show that N-OMS-2 has a relatively high reducibility
in the low temperature, which may be due to its large amount of
surface oxygen. Whereas, the reduction rate of B-OMS-2 is faster
than that of N-OMS-2. TPR analysis shows that the redox behavior
of the manganese oxide in B-OMS-2 was potent, indicating a high
reducibility or lattice oxygen mobility.
It is known that the catalytic activity is also closely related to
the oxygen adsorption capacity and refreshment of lattice oxygen
of the catalyst. To reveal the re-oxidation capability, a series of O2-
TPO tests were carried out. Fig. 7B showed the TPO profiles of B-
◦
peratures of 500, 600, and 650 C for 4 h to investigate their phase
transformation. XRD results confirmed that B-OMS-2 is stable even
◦
above 650 C. The results of XPS and chemical titration also showed
4
+
that only Mn exist sin B-OMS-2. These results indicated there may
be some bulk oxygen vacancies in the B-OMS-2 lattice. Luo et al.
also reported that, for OMS-2, the evolution of oxygen could lead
to the formation of framework oxygen vacancies without destroy-
ing its structure [8]. Based on XPS and TGA results, we suggested
that some oxygen vacancies exist in B-OMS-2.We denoted this kind
of oxygen defects as bulk oxygen vacancy in order to distinguish it
OMS-2 and N-OMS-2 after H -TPR reduction. For B-OMS-2, there
2
are one strong O consumption peak and two small peaks at 170,
2
◦
◦
215 C and 253 C, respectively. According to the H -TPR analysis,
the peaks at 170, 215 C may be assigned to two-step reoxidation for
2
◦
3
+
from the Mn -oxygen vacancy associates in N-OMS-2.
B-OMS-2 pre-reduced by H : Mn O to MnO2 and MnO to Mn O ,
2
3
4
3
4
315
A
B
170
215
253
a
110
343
a
273
140
122
b
b
100
200
300
400
500
600 50
100
150
200
250
300
Temperature/
Temperature/
Fig. 7. H2-TPR (A) and O2-TPO(B) profiles of (a) B-OMS-2 and (b) N-OMS-2. The signal intensities of O2-TPO are enlarged 10 times.
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
6
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
100
a
a
b
80
60
40
20
0
2
9232853
1395 1055
b
4
000 3500 3000 2500 2000 1500 1000 500
-1
Wavenumber(cm
)
160
180
200
220
240
260
Temperature(
)
◦
◦
Fig. 9. FT-IR spectra of B-OMS-2: (a) dried at 110 C, (b) calcined at 400 C.
Fig. 8. CO2 yield over B-OMS-2 (a) and N-OMS-2 (b). Reaction conditions: o-xylene
−
1.
5
00 ppm, 20% O2/N2 balance, total flow rate 50 mL/min, W/F = 0.60 g s mL
activity for the catalytic combustion of o-xylene [28]. Because the
relative amounts of Mn ion can influence the surface properties
◦
4+
respectively [5]. The weak peak at 253 C indicates that there are
gas phase oxygen adsorption/exchange sites in B-OMS-2. There is
and composition of MnO2, such as the surface acid-base property,
the contents of lattice oxides and surface hydroxy group. In order to
exclude the effect of relative amounts of Mn4+ ion on the B-OMS-2
and N-OMS-2, we compared the catalytic performance of B-OMS-2
and RP-MnO2 (our previous product, which possessed almost 100%
◦
also a small surface O spillover peak appeared at 110 C, indicating
2
surface oxygen release without reaction in H -TPR.
2
◦
◦
For N-OMS-2, it starts to be re-oxided at ∼122 C and 142 C.
The reoxidation temperature is much lower than that of B-OMS-2,
and the O2 consumption peak area was smaller. These phenom-
4+
Mn ion on their surfaces [28]). The results showed that the cata-
ena demonstrate that the O adsorption properties and the oxygen
lyst (RP-MnOx (8-400)) showed 100% o-xylene conversion to CO2
at 210 C, 20 C higher than that of B-OMS-2. In view of the differ-
ence of the surface area of B-OMS-2 and RP-MnO2, the areal rates
2
◦
◦
exchange ability of B-OMS-2 and N-OMS-2 are different at tem-
◦
peratures less than 300 C. The larger O consumption on B-OMS-2
2
−
8
means that the interaction between B-OMS-2 and the gas phase
oxygen is stronger than that of N-OMS-2. That is to say, B-OMS-2
exhibits stronger adsorption properties and higher exchange ability
with gas phase oxygen than N-OMS-2 does.
of the two samples were calculated. The data were 2.90 × 10 and
−
9
−2
−1
◦
1.27 × 10 mol m min
for B-OMS-2 and RP-MnO2 at 170 C.
4+
The results indicate that Mn ion species was also not the main fac-
tor for its catalytic activity. There may be other factors determining
the B-OMS-2 activity.
Tang et al. found that the insertion of vanadium in the frame-
work of OMS-2 may create surface defect sites in the manganese
oxide structure, which enhances the activity of OMS-2 for HCHO
catalytic oxidation [29]. Qu et al. also reported that doping Ag in
manganese oxides resulted in the presence of some defects in man-
ganese, and these defective sites can be the active centers in the
catalytic reaction since the oxygen species of the defective oxides
tends to be easily released and transferred, enhancing the reactivity
of catalysts in oxidation of VOCs [7]. This result indicates that any
processing condition that favors the formation of more desirable
oxygen vacancies/defects will result in enhanced redox ability of
manganese oxide.
3
.4. o-Xylene catalytic oxidation over B-OMS-2 and N-OMS-2
The catalytic performance of B-OMS-2 and N-OMS-2 are eval-
uated in the o-xylene catalytic oxidation and the results are
presented in Fig. 8. It is clearly seen that B-OMS-2 performed much
better than the N-OMS-2 catalyst, indicating that the microstruture
significantly influences the catalytic activity of OMS-2 for o-xylene
◦
oxidation. 100% CO2 yield was observed at 190 C for B-OMS-2,
◦
◦
while for N-OMS-2, 100% CO2 yield was observed at 240 C, 50 C
higher than B-OMS-2. The performance of B-OMS-2 is comparable
to some precious metals [3]. To examine the catalytic stability of B-
OMS-2 sample, we carried out the on-stream reaction experiment
◦
According to our XPS and TGA results, there are some bulk
at 190 C (the result is not shown here). No significant decrease in
oxygen vacancies in B-OMS-2. The H -TPR and O -TPO analyses
catalytic activity was observed within 60 h of on-stream reaction.
Hence, we believe that B-OMS-2 was catalytically durable.
In the past years, a number of materials have been used as cat-
alysts for the oxidative removal of o-xylene. It was reported that
for the combustion of o-xylene, the T90% value was 249 C over the
Pd/Co O catalyst at WHSV = 60 000 mL/(g h) [3], 240 C over the
Mn-Ce oxides catalyst at GHSV = 8000 h [27], 220 C over ␣-MnO2
catalyst at GHSV = 8000 h
value was 180 C) catalyst outperformed the above-mentioned cat-
alysts for the combustion of o-xylene.
2
2
revealed that B-OMS-2 possesses higher lattice oxygen mobility
and stronger re-oxidation ability than N-OMS-2. The results are in
good agreement with the sequence of the catalytic performance of
the two OMS-2 samples. The more surface oxygen content generally
corresponds to higher activity in VOCs oxidation reaction. However,
it is known that the catalytic activity is also closely related to the
oxygen adsorption capacity and refreshment of lattice oxygen of
the catalyst. In spite of B-OMS-2 having lower surface oxygen con-
tent, it showed high activity because it possesses more bulk oxygen
vacancy. The presence of bulk oxygen vacancy enhances the reac-
tivity of OMS by modifying the rate of vacancy exchange between
◦
◦
3
4
−
1
◦
−
1
[28]. Apparently, the B-OMS-2 (T90%
◦
It is well-known that the catalytic activity of manganese oxide
4+
3+
is associated with several factors, such as surface area, Mn /Mn
the bulk and oxide surface. H -TPR and O -TPO results suggest that
molar ratio, oxygen vacancy, and reducibility. For the combustion of
organic compounds, the catalyst with a higher surface area would
show a better catalytic activity. The surface area of B-OMS-2 is
similar to that of N-OMS-2. However, B-OMS-2 significantly outper-
formed N-OMS-2, indicating that surface area was a minor factor
influencing the catalytic performance.
2
2
the bulk oxygen vacancy engenders higher lattice oxygen mobility
and faster adsorption, diffusion and dissociation of gaseous oxy-
gen at B-OMS-2 surface sites, which are able to explain its high
o-xylene conversion activity. Therefore, we speculate that bulk oxy-
gen vacancies in B-OMS-2 may be an important factor for its high
catalytic activity. In addition, the loose structure and high K ion
content may also play some roles.
In our previous studies, we have found that the surface concen-
4
+
tration of Mn ion plays an important role for its high catalytic
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024

G Model
CATTOD-10216; No. of Pages7
ARTICLE IN PRESS
Y. Wu et al. / Catalysis Today xxx (2016) xxx–xxx
7
3.5. The possible formation mechanism of bulk oxygen vacancy in
ciates. H -TPR and O -TPO results showed that B-OMS-2 exhibited
2 2
B-OMS-2
high reducibility and oxygen exchange ability with gas phase oxy-
gen. The activity test results showed that B-OMS-2 could convert
◦
−1
,
The formation of bulk oxygen vacancy is related to the
100% o-xylene into CO2 at 190 C at a space velocity of 8000 h
◦
microstructure of OMS-2. Usually, the growth of primary struc-
tures is modulated by preferential adsorption of anions, solvent
molecules, and different reaction conditions. Although the exact
growth mechanism of the B-OMS-2 is not completely understood,
we suggest that benzyl alcohol, the organic oxygen-containing
functional group may act as anchoring sites or nucleation sites
controlling the growth of B-OMS-2 nuclei. The formation mech-
anism of bulk oxygen vacancy may be as follows: (1) a water and
oil two-phase system is formed when benzyl alcohol is dropwise
50 C lower than N-OMS-2. It is concluded that the bulk oxygen
vacancy, high reducibility and re-oxidation ability may be respon-
sible for B-OMS-2’s excellent catalytic performance.
Acknowledgements
This work was financially supported by the National Natural
Science Foundation of China (No. 21147004, No. 21201054), the
Natural Science Foundation of Hebei Province (No. B2013205100,
No. 2013205136), and the Natural Science Foundation of Hebei
Education Department (No. ZD2015114).
added to KMnO solution. (2) MnOx nuclei are formed slowly as the
4
reaction going on. At the same time, benzyl alcohol might adsorb
on the MnOx nuclei, and the adsorbate may function as a struc-
ture directing agent that is capable of retarding the aggregation of
the building blocks. (3) A rapid and uniform evolution of oxygen-
containing functional groups may occur as the sample is calcined
References
[
[
1] W.X. Tang, X.F. Wu, S.D. Li, X. Shan, G. Liu, Y.F. Chen, Appl. Catal. B 162 (2015)
110–121.
2] S.J. Jiang, E.S. Handberg, F. Liu, Y.T. Liao, H.Y. Wang, Z. Li, S.Q. Song, Appl. Catal.
B 160–161 (2014) 716–721.
◦
at 400 C, which lead to the formation of bulk oxygen vacancy. Suib
and co-workers also found that some of the oxygen in the OMS-
2
framework can easily dislodge from the framework and evolve
[3] Y.F. Wang, C.B. Zhang, Y.B. Yu, R.L. Yue, H. He, Catal. Today 242 (2015)
94–299.
2
into air without destroying the skeleton of the parent cryptome-
lane structure when the sample is heated in air at as high as 650 C.
The evolution of oxygen results in the formation of framework oxy-
gen vacancies, which are catalytically active sites for the oxidation
reactions. Our results may agree with the literatures [8,26].
[
[
4] J. Spivey, Ind. Eng. Chem. Res. 26 (1987) 2165–2175.
5] J.T. Hou, L.L. Liu, Y.Z. Li, M.Y. Mao, H.Q. Lv, X.J. Zhao, Environ. Sci. Technol. 47
(2013) 13730–13736.
◦
[
[
[
6] H. Sun, S. Chen, P. Wang, X. Quan, Chem. Eng. J. 178 (2011) 191–196.
7] Z.P. Qu, Y.B. Bu, Y. Qin, Y. Wang, Q. Fu, Appl. Catal. B 132–133 (2013) 353–362.
8] J. Luo, Q.H. Zhang, A. Huang, S.L. Suib, Microporous Mesopor. Mater. 35–36
(2000) 209–217.
Fig. 9 showed the FT-IR spectra of the product of KMnO /benzyl
4
◦
◦
[9] W.N. Li, J.K. Yuan, S. Gomez-Mower, S. Sithambaram, S.L. Suib, J. Phys. Chem. B
10 (2006) 3066–3070.
alcohol dried at 110 C and calcined at 400 C respectively. The
product dried at 110 C showed a clear C-H stretching and bend
vibration bands at 2923, 2853, and 1395 cm . Moreover, there
is also a C-O stretching vibration band appear at 1055 cm . The
1
◦
[
10] F. Wang, H.X. Dai, J.G. Deng, G.M. Bai, K.M. Ji, Y.X. Liu, Environ. Sci. Technol. 46
(2012) 4034–4041.
[11] S. Dharmarathna, C.K. Kingondu, W. Pedrick, L. Pahalagedara, S.L. Suib, Chem.
Mater. 24 (2012) 705–712.
12] C.K. Kingondu, N. Opembe, C.H. Chen, K. Ngala, H. Huang, A. Iyer, H.F. Garces,
S.L. Suib, Adv. Funct. Mater. 21 (2011) 312–323.
13] J.T. Hou, Y.Z. Li, L.L. Liu, L. Ren, X.J. Zhao, J. Mater. Chem. A 1 (2013) 6736–6741.
14] H.C. Genuino, S. Dharmarathna, E.C. Njagi, M.C. Mei, S.L. Suib, J. Phys. Chem. C
−
1
−
1
−1
−1
broadband at ca. 3400 cm
and 1638 cm correspond to the
[
stretching and bending vibrations of OH-groups, respectively. The
IR result indicates the existence of aliphatic alcohol. The result
supports the hypothesis that there are some organic compounds
absorbed on the surface of MnO2 nuclei in advance, which decom-
pose to form small aliphatic alcohol in the subsequent heating
process. When the product was calcined at 400 C, the absorbed
aliphatic alcohol disappeared. B-OMS-2 presents a clear signature
at 527, 578 and 705 cm , which are ascribed to the vibrations of
the MnO6 octahedral framework of OMS-2 as previously reported
[
[
116 (2012) 12066–12078.
[15] S.K. Meher, G.R. Rao, J. Phys. Chem. C 117 (2013) 4888–4900.
[16] C.K. King’ondu, N. Opembe, C.H. Chen, K. Ngala, H. Huang, A. Iyer, H.F. Garcés,
S.L. Suib, Adv. Funct. Mater. 21 (2011) 312–323.
◦
[
17] B.X. Hu, C.H. Chen, S.J. Frueh, L. Jin, R. Joesten, S.L. Suib, J. Phys. Chem. C 114
2010) 9835–9844.
[18] J.H. Li, R.H. Wang, J.M. Hao, J. Phys. Chem. C 114 (2010) 10544–10550.
(
−
1
[
19] H. Tian, J.H. He, X.D. Zhang, L. Zhou, D.H. Wang, Microporous Mesopor. Mater.
38 (2011) 118–122.
1
[
19].
[
[
[
20] Y.T. Wang, A.H. Lu, H.L. Zhang, W.C. Li, J. Phys. Chem. C 115 (2011) 5413–5421.
21] W. Xiao, D. Hu, C. Peng, G.Z. Chen, Appl. Mater. Interfaces 3 (2011) 3120–3129.
22] X.W. Liu, K.B. Zhou, L. Wang, B.Y. Wang, Y.D. Li, J. Am. Chem. Soc. 131 (2009)
4
. Conclusion
3140–3141.
[
[
[
23] V.P. Santos, M.F.R. Pereira, J.J.M. Órfão, J.L. Figueiredo, Appl. Catal. B 99 (2010)
353–363.
24] L. Lamaita, M.A. Peluso, J.E. Sambeth, H.J. Thomas, Appl. Catal. B 61 (2005)
B-OMS-2 and N-OMS-2 samples were prepared via KMnO4 as
oxidizing agent and benzyl alcohol and Mn(NO ) as reducing agent
3
2
114–119.
respectively. The OMS-2 samples were used as catalysts for o-
xylene removal. XRD, BET, FT-IR, SEM, TEM, HRTEM, XPS and TGA
techniques revealed that there was a direct relationship between
the reduced agent, the microstructure of the OMS-2 materials and
their reactivity toward o-xylene removal. B-OMS-2 presented a
25] H. Sun, S. Chen, P. Wang, X. Quan, Chem. Eng. J. 178 (2011) 191–196.
[26] Q.H. Zhang, S.L. Suib, Chem. Mater. 11 (1999) 1306–1311.
[
[
27] Y.S. Wu, Y.X. Zhang, M. Liu, Z.C. Ma, Catal. Today 153 (2010) 170–175.
28] Y.S. Wu, Y. Lu, C.J. Song, Z.C. Ma, S.T. Xing, Y.Z. Gao, Catal. Today 201 (2013)
32–39.
[29] X.F. Tang, J.H. Li, J.M. Hao, Catal. Commun. 11 (2010) 871–875.
4+
loose structure and contained almost 100% Mn . XPS and TGA
results showed that there was some bulk oxygen vacancy in B-
OMS-2. FT-IR verified that the adsorption and desorption of benzyl
alcohol on the MnO2 nuclei may result in the formation of the bulk
4
+
oxygen vacancy. Whereas, N-OMS-2 presented a mixture of Mn
3
+
3+
and Mn , indicating there exist small Mn -oxygen vacancy asso-
Please cite this article in press as: Y. Wu, et al., Effect of reducing agent on the structure and activity of manganese oxide octahedral
molecular sieve (OMS-2) in catalytic combustion of o-xylene, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.05.024