Structure and phase analysis of one-pot hydrothermally synthesized FePO4-SBA-15 as an extremely stable catalyst for harsh oxy-bromination of methane

Article abstract of DOI:10.1016/j.apcata.2012.12.022

FePO₄-SBA-15 (OP) was directly synthesized via a one-pot hydrothermal technique, using Fe(NO₃)₃ and H ₃PO₄ as the precursors. FePO₄/SBA-15 (IMP) was also prepared as a reference, using an impregnation method and commercially available SBA-15 as the support. The yielding samples were employed to catalyze the harsh oxy-bromination of methane (OBM) reaction, showing similar initial catalytic performances. The fresh and spent samples after catalytic reaction were thoroughly characterized by N₂-physisoption, inductively coupled plasma, wide- and small-angle X-ray diffraction, transmission electron microscopy, diffuse reflectance UV-vis spectroscopy, temperature-programmed oxidation, and room temperature ⁵⁷Fe M?ssbauer spectroscopy. It was found that the FePO₄ was in good crystalline in the OP sample while it was in amorphous for the IMP catalyst. Despite this difference, both FePO₄ phases in the fresh samples were transformed into Fe ₇(PO₄)₆ and Fe₂P₂O ₇ in the spent ones. Furthermore, the OP catalyst showed excellent stability in a period of 1000 h time-on-stream performance without apparent deposition of cokes. The losses of P and Fe after the catalytic evaluation were only 9.5% and 15.5%, respectively, while the ratio of P/Fe remained close to 1.0. N₂-adsorption and TEM observations confirmed that the mesoporous pores were extremely stable under the harsh reaction ambience, which might play a crucial role in the stability test.

Full text of DOI:10.1016/j.apcata.2012.12.022

Applied Catalysis A: General 453 (2013) 235–243
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Structure and phase analysis of one-pot hydrothermally synthesized
FePO4-SBA-15 as an extremely stable catalyst for harsh
oxy-bromination of methane
Runqin Wanga,c, Ronghe Lina, Yunjie Dinga,b,∗, Jia Liua,c, Junhu Wangb,d, Tao Zhangb,d
a Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
b State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
c University of Chinese Academy of Sciences, Beijing 100049, PR China
d Mössbauer Effect Data Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
FePO4-SBA-15 (OP) was directly synthesized via a one-pot hydrothermal technique, using Fe(NO3)3 and
H3PO4 as the precursors. FePO4/SBA-15 (IMP) was also prepared as a reference, using an impregna-
tion method and commercially available SBA-15 as the support. The yielding samples were employed
to catalyze the harsh oxy-bromination of methane (OBM) reaction, showing similar initial catalytic
performances. The fresh and spent samples after catalytic reaction were thoroughly characterized by N2-
physisoption, inductively coupled plasma, wide- and small-angle X-ray diffraction, transmission electron
microscopy, diffuse reflectance UV–vis spectroscopy, temperature-programmed oxidation, and room
temperature 57Fe Mössbauer spectroscopy. It was found that the FePO4 was in good crystalline in the
OP sample while it was in amorphous for the IMP catalyst. Despite this difference, both FePO4 phases in
the fresh samples were transformed into Fe7(PO4)6 and Fe2P2O7 in the spent ones. Furthermore, the OP
catalyst showed excellent stability in a period of 1000 h time-on-stream performance without apparent
deposition of cokes. The losses of P and Fe after the catalytic evaluation were only 9.5% and 15.5%, respec-
tively, while the ratio of P/Fe remained close to 1.0. N2-adsorption and TEM observations confirmed that
the mesoporous pores were extremely stable under the harsh reaction ambience, which might play a
crucial role in the stability test.
Received 23 August 2012
Received in revised form
18 November 2012
Accepted 1 December 2012
Available online xxx
Keywords:
SBA-15
Iron phosphate
Methane
Oxy-bromination
Stability
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
intrinsic problems associated with the direct conversion routes,
such as poor selectivity and severe corrosion problem. Halo-
Methane is the major component of nature gas which is one
of the most abundant fossil fuel resources. Conversion of methane
to value-added chemicals is a promising approach to solve energy
crisis which is a matter of great urgency nowadays, but methane
activation is difficult due to the high energy of the C H bond
(439 kJ/mol). Many efforts have been devoted to activate methane
molecule. Traditionally, people have transformed methane into
syngas by steam reforming, followed by CO hydrogenation process
to produce methanol, oxo chemicals and their derivatives, such as
urea fertilizer and synthetic hydrocarbon fuels. However, methane
steam reforming is an endothermic and energy-intensive process;
thus, the direct conversion of methane to value-added chemicals
is highly desirable and becomes a hot topic in catalysis. But up
to now, the progress in this field is limited since there are some
genation of methane is also a cold and effective approach for
methane activation. With methyl halide as an intermediate prod-
uct, methane could be converted to various products which offer
an alternative to the CO hydrogenation process. Photochemical or
thermal halogenation of hydrocarbon with halogen is known as a
free radical chain process and lacks selectivity. Many interesting
works [1–4] have been carried out to improve the selectivity of
methyl halide. However, halogens are highly corrosive to equip-
ment and dangerous to handle which present a big trouble for
the industrialization. In view of halogen corrosion, methane oxida-
tive bromination [5–8] and chlorination [9] were studied which
use hydrobromic acid or HCl as reactants instead of halogens
to reduce corrosion problem. Zhou et al. [5,6,10] firstly reported
methane oxidative bromination (OBM) reaction and developed
Ru- and Rh-containing catalysts which showed high selectivity for
CH3Br and excellent stability, but the uses of noble metals raise the
production cost. We also reported FePO4/SiO2 [8] for the OBM reac-
tion, producing nearly equal-molar CH3Br and CO which could be
used as feed stock for producing acetic acid. However, FePO4/SiO2
∗
Corresponding author at: 457 Zhongshan Road, Dalian, Liaoning 116023, PR
China. Fax: +86 411 84379143.
E-mail address: dyj@dicp.ac.cn (Y. Ding).
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.12.022

236
R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
gradually deactivated after an induction period due to coke
deposition [11]. Recently, He et al. [12] studied modified CeO2
nanocrystals for the oxyhalogenations of methane with a relatively
high selectivity of CH3X (X = Cl, Br), but long-term (>100 h) stability
test was needed. It still remains a challenge to develop an efficient
and stable catalyst that could stand the severe chemical ambience
under the OBM reaction.
SBA-15 is a silica-based mesoporous material and has an
ordered mesopore structure, high surface area and large pore vol-
ume, presenting potential applications in many scientific fields.
People try to incorporate metal ions into the framework of SBA-
15 during the synthesis of this material because of their promising
catalytic performance. However, introducing metal ions into SBA-
15 is challenging work due to the difficulty of the formation of
was transferred into an autoclave, aged for 24 h at 90 ◦C. The resul-
tant solid was filtered, washed with deionized water, and dried
at 60 ◦C for 12 h in air. The calcined procedure consisted of two
steps based on heating: first at 250 ◦C for 3 h, and then at 600 ◦C
for 4 h with a heating rate at 1 ◦C/min. The catalyst was shaped
to 20–40 mesh before evaluation. The molar composition of initial
solutions presented was as follows: 1.0 TEOS:0.017 P123:0.04415
Fe:1.5 H3PO4:208 H2O. The synthesized FePO4-SBA-15 materials
were labeled as OP-X (X = 90 or 130), where X denoted the aged tem-
perature in ◦C. The spent catalysts were labeled as OP-X-Y, where
Y denoted the reaction time in hour.
Another FePO4/SBA-15 was prepared by an impregnation
method with Fe(NO3)3·9H2O and H3PO4 as precursors and the
synthesis procedure was described previously [11]. SBA-15 was
purchased from Chang Chun Jilin University High Tech. Co. Ltd. This
catalyst was labeled as IMP and the spent catalyst was labeled as
IMP-Z, where Z denoted the reaction time in hour.
metal
O Si bonds under strongly acidic conditions of synthesis. A
series of methods have been surveyed to introduce V, Mo, Al, Fe and
Co into SBA-15 [13–17], mostly with HCl aqueous solution as acid
source. H3PO4 was also used in some cases as acid source for the
synthesis of SBA-15 [18]. To the best of our knowledge, the simul-
taneous incorporation of Fe and P atoms into SBA-15 has not been
reported yet. Hydrothermal stability is also very important for cat-
alysts, and hydrothermal stable catalysts are favored in industrial
applications, especially when used under harsh conditions. Sev-
eral methods such as carbon propping [19] and the addition of salt
[20] have been employed to improve the hydrothermal stability
of the mesoporous silicates. For the OBM reaction, the presence of
corrosive HBr/H2O ambience and high reaction temperature up to
590 ◦C present a formidable challenge to the structure stability of
catalysts.
Based on the above consideration, we try to introduce Fe into
SBA-15 under hydrothermal conditions, using mediate strong acid
H3PO4 as the acid source. The H3PO4 here functions in two ways:
first, providing an acidic environment necessary for the synthe-
sis of SBA-15; second, providing excess phosphorous species for
the in situ formation of FePO4. The as-synthesized sample was
used for the OBM reaction, and another FePO4/SBA-15 sample
prepared by an incipient impregnation method was used as a
reference. Compared with the reference sample, the one yielded
by novel hydrothermal method showed relatively higher activ-
ity and distinctively stable catalytic performance even up to
1000 h. The textual and chemical properties of the novel cata-
lysts before and after the OBM test were thoroughly characterized
by N2-physisoption, transmission electron microscopy (TEM), X-
ray diffraction (XRD), diffuse reflectance UV–vis spectra and 57Fe
Mössbauer spectroscopy, and the superior catalytic performance
was also discussed according to the characterization results. It
is worthy to note the hydrothermally synthesized FePO4-SBA-15
catalyst with distinctive crystalline FePO4 phase even in a low
loading of 10 wt.% showed excellent structure stability during the
1000 h harsh OBM reaction. Hopefully, these kinds of novel mate-
rials would greatly widen the traditional application fields of iron
phosphates.
2.2. Characterization of catalysts
The structural properties of the samples were determined by
N2-adsorption–desorption method, using a physical adsorption
instrument (Quantachrome, USA). Specific surface area and pore
volumes were determined by N2 adsorption at 77 K. Transmission
electron microscopy (TEM) measurements were taken on a Tecnai
G2 Spirit instrument with an accelerating voltage of 120 kV. The
elemental contents of the fresh and spent catalysts were analyzed
by inductively coupled plasma (ICP) on an ICPS-8100 apparatus.
Powder X-ray diffraction (XRD) measurements of the samples
were performed on a PANalytical X’Pert-Pro powder X-ray diffrac-
tion spectrometer equipped with Cu K˛ (40 kV, 40 mA) irradiation,
covering 2ꢀ angles of 0.5–6◦ and 10–90◦ for small- and wide-angle
scanning, respectively.
Diffuse reflectance UV–vis spectra were recorded on a Cary 5000
UV-VIS-NIR spectrophotometer. The spectra were collected in the
range of 200–600 nm referenced to BaSO4.
57Fe Mössbauer spectra of the catalysts were taken on a Topo-
logic 500A spectrometer and a proportional counter at room
temperature. 57Co(Rh) moving in a constant acceleration mode was
used as radioactive source. All of the spectral analyses were con-
ducted assuming a Lorentzian lineshape for computer folding and
fitting.
Hydrogen temperature-programmed reduction (H2-TPR) mea-
surements were carried out on a Micromeritics Autochem II 2920
apparatus equipped with a TCD detector. Before each measure-
ment, the sample was treated in a flow of He at 120 ◦C for 0.5 h,
and cooled at room temperature. TPR was then performed by heat-
ing at a rate of 10 ◦C/min to 900 ◦C in a flow of 10 vol.% H2/Ar
(50 sccm).
Temperature-programmed oxidation (TPO) measurements
were carried out on a Micromeritics Autochem II 2920 appara-
tus equipped with a mass detector. Before each measurement, the
sample was treated in a flow of He at 350 ◦C for 1 h. After the tem-
perature of the sample returned to room temperature, it was heated
to 900 ◦C in a flow of 2 vol.% O2/He (30 sccm) at a ramp of 10 ◦C/min.
2. Experimental
2.1. Catalyst preparation
The synthesis procedure of FePO4-SBA-15 catalyst is described
as follows: a certain amount of Fe(NO3)3·9H2O and 8.2 mL of
tetraethyl orthosilicate (TEOS) were hydrolyzed in 10 mL deionized
water for 30 min (Solution A). 4 g of nonionic triblock copolymer
surfactant EO20PO70EO20 (P123) was dissolved in 85 wt.% H3PO4
and deionized water and then stirred at 35 ◦C for 2 h to obtain Solu-
tion B. Solution A was dropped to Solution B under stirring and
it was subsequently stirred vigorously for 20 h at 35 ◦C. Then it
2.3. OBM reaction evaluation
The OBM reaction was performed in a quartz-tube micro-
reactor. The 40 wt.% HBr/H2O solution was fed by a syringe. Catalyst
was loaded in the center of the reactor, with quartz sand filling in
both sides of the catalyst to obtain a flat-temperature zone. The
sampling and analysis methods have been described elsewhere [7].

R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
237
Fig. 1. N2-adsorption–desorption isotherms and corresponding pore size distributions of the fresh and spent catalysts: (a) OP-90, (b) OP-90-1000, (c) IMP, (d) IMP-100, (e)
OP-130 and (f) OP-130-30.
3. Results and discussion
as-prepared samples were all very close to stoichiometric value of
1.0. This might suggest that most of the P atoms in the obtained
products were coordinated with Fe atoms during the synthesis
process, and the excess P atoms were completely removed after
filtering and washing.
3.1. Physicochemical properties of catalysts
3.1.1. Structure and morphology of catalysts
Two hydrothermally synthesized catalysts and an impregnated
one as a reference were thoroughly compared both in physico-
chemical properties and catalytic performances. The elemental
analysis was performed by ICP and the results are summarized in
Table 1. As mentioned earlier, H3PO4 was used to manipulate the
acid synthesis environment and provide the phosphorus source.
The molar ratio of P/Fe in raw materials (∼34.0) was far beyond
the stoichiometric value of FePO4. However, the P/Fe values in the
Fig. 1 depicts the N2-adsorption–desorption isotherms and the
corresponding pore size distributions of the samples. The structural
characteristics of these samples obtained from N2-adsorption are
summarized in Table 1. All these isotherms are of type IV with a
H1 hysteretic loop, indicative of the mesoporous structure of the
catalysts. Besides, the pore size distributions of these samples were
narrow, and the pore size of both OP-90 and IMP centered at 6.6 nm.
For OP-90, another peak at 3.8 nm appeared that might be due to the

238
R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
Table 1
Physical properties of the fresh and spent catalysts.
Sample
Element
P/Fe (mol/mol)
Surface area (m2/g)
Pore volume (cm3/g)
Average pore size (nm)
D(FePO4) (nm)b
concentration
(wt.%)a
Fe
P
IMP
IMP-100
OP-90
OP-90-1000
OP-130
OP-130-30
3.70
3.50
3.36
3.04
2.00
1.84
2.06
1.74
0.98
0.95
1.11
1.03
489
304
769
365
553
522
0.90
0.72
1.01
0.74
1.32
1.31
7.4
9.5
5.3
8.1
9.5
A
21.8
20.2
A
10.0
c
FePO4/SiO2
3.71
3.00
2.04
1.41
0.99
0.85
FePO4/SiO2-100c
a
Elemental composition was characterized by ICP.
The crystallite size (D) of FePO4 was calculated by Scherrer equation.
Ref. [11].
b
c
‘A’ means iron phosphate was in amorphous.
tensile strength effect (TSE) of the adsorbed phase [21]. According
to Table 1, the OP-90 possesses the highest specific surface area
and the smallest average pore size as compared to the other two
samples. Generally, the one-pot prepared samples showed higher
surface areas than the IMP, and crystallization temperature played
a crucial role in the structure of the yielded catalysts, i.e., lower
crystallization temperature led to higher surface area and smaller
average pore size.
The mesoporous structure characteristics of the synthesized cat-
alysts were further characterized by small- and wide-angle X-ray
diffraction as shown in Fig. 2. Fig. 2a shows the wide-angle XRD
patterns of the samples. It is interesting that strong and distinctive
diffraction patterns of crystal FePO4 phase were clearly observed
for the two one-pot derived catalysts while the impregnated one
only presented a broad and amorphous peak related to silica. For
the impregnated FePO4 catalyst, high dispersion of loaded FePO4
is very common possibly due to the relatively large surface area of
SBA-15, leaving the XRD patterns of corresponding samples no dis-
tinctive peak of FePO4 phase, especially when the loading is lower
than 40 wt.% [22]. It is worthy to note that the nominal loading of
FePO4 in the one-pot synthesized samples was only 10 wt.%. It is
the first time that crystal FePO4 in such a low loading was loaded
on SBA-15 with a surface area as high as 769 m2/g. We hypothe-
size the crystallization process of FePO4 might be strengthened in
the unique hydrothermal synthesis environment. Furthermore, the
intensities of diffraction were different for OP-90 and OP-130. The
crystal sizes for OP-90 and OP-130 were 21.8 and 20.2 nm, respec-
tively, calculated by Scherrer equation based on the most intensive
peak of 25.7◦, strongly indicating that the crystallization temper-
ature exerted an influence on the crystal size of FePO4. Big FePO4
granules with diameters in microns were also observed in the SEM
image and elemental components were convincingly confirmed
by EDS (Figs. S1a,b, see Supporting information). The small-angle
XRD patterns of the samples were given in Fig. 2b, with only one
clear diffraction peak for each sample located at 0.5–1.0◦ that can
be attributed to the diffraction of (1 0 0) facets of SBA-15 (space
group P6mm). However, the peaks corresponding to (1 1 0) and
(2 0 0) facets were missing in the small-angle XRD patterns. Fur-
thermore, TEM images also showed that OP-90 was of hexagonal
array of mesoporous channels (Fig. 3a,b) which are typical for SBA-
15. These results confirmed the as-obtained samples were of highly
ordered mesoporous structure, which are in accordance with the
N2-adsorption observations.
To learn the locations of FePO4 in the FePO4-SBA-15, the diffuse
reflectance UV–vis spectra were collected on the fresh samples and
the results are presented in Fig. 4. The UV–vis spectrum of OP-90
exhibits two kinds of absorption bands. One centers at 300–370 nm,
which could be assigned to the octahedral Fe ions in oligomeric
clusters [23]. Another one is in the range of 220–250 nm which is
characteristic of Fe in the bulk silica [24]. The diffuse reflectance
Fig. 2. Wide- (a) and small-angle (b) X-ray diffraction patterns of catalysts. (ꢀ) FePO4; (ꢁ) Fe7(PO4)6; (♦) Fe2P2O7.

R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
239
Fig. 3. TEM images of (a,b) OP-90 and (c,d) OP-90-1000.
Fig. 5. H2-TPR profiles of catalysts.
Fig. 4. UV–vis spectra of OP-90 and IMP samples.

240
R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
Fig. 6. Room temperature 57Fe Mössbauer spectra of the fresh and spent catalysts: (a) IMP, (b) IMP-100, (c) OP-90, (d) OP-90-1000, (e) OP-130 and (f) OP-130-30.
UV–vis spectra strongly suggest that FePO4 species exist in two
different ways in the FePO4-SBA-15: as bulk FePO4 and in the
framework of SBA-15. On the contrary, the UV–vis spectrum of IMP
exhibited only one strong band at 250–290 nm, which is assigned
to isolated extra-framework species in an octahedral or pseudo-
tetrahedral environment, implying that FePO4 species were located
on the outer surface of SBA-15 as isolated species in the IMP sample
[24]. These results are highly coincident with the X-ray diffraction
results, and further illustrate the novelty of the one-pot synthesized
samples.
prepared catalysts to ascertain the redox property. Fig. 5 shows
the H2-TPR profiles of the fresh catalysts. As a reference, the H2-
TPR profile of bulk FePO4 was also presented. It is seen that the
first H2 consumption peak of all the samples appeared at a tem-
perature lower than 590 ◦C which is roughly lower than the OBM
reaction temperature. These peaks can be attributed to the reduc-
tion of FePO4 to Fe2P2O7, as illustrated in our previous report [11].
In general, the first reduction temperatures were in the order of
FePO4 > OP-90 > OP-130 > IMP. It has been revealed in literatures
[22,25] that the reduction temperatures can be greatly reduced
when iron phosphate was loaded on a mesoporous material such
as SBA-15 and MCM-41 via an impregnation method. Our exper-
iments further confirmed that the reduction temperature of IMP
was 112 ◦C lower than that of the bulk catalyst. One possible
3.1.2. Redox property and the nature of iron species
Since the OBM reaction is largely dominated by the oxidation–
reduction pair of Fe2+/Fe3+, we further conducted H2-TPR on the

R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
241
Table 2
57Fe Mössbauer superfine parameters of the fresh and spent catalysts at room temperature.
Entry
IMP
Component
ıa (mm/s)
ꢂb (mm/s)
2ꢃ c (mm/s)
Fd (%)
Attributed phase
Fe3+
Fe3+
0.36
0.39
1.13
0.67
0.44
0.42
40
60
Highly dispersed FePO4
FePO4-quartz
IMP-100
Fe3+
Fe2+
Fe2+
0.46
1.18
1.25
0.68
2.47
2.85
0.33
0.28
0.18
51
39
10
Fe7(PO4)6
Fe7(PO4)6
Fe2P2O7
OP-90
Fe3+
0.28
0.62
0.30
100
FePO4-quartz
OP-90-1000
Fe3+
Fe2+
Fe2+
0.42
1.13
1.24
0.68
2.53
2.89
0.37
0.28
0.28
61
26
13
Fe7(PO4)6
Fe7(PO4)6
Fe2P2O7
OP-130
Fe3+
Fe3+
0.31
0.36
0.67
0.33
0.30
0.32
58
42
FePO4-quartz
FePO4-tridymite
OP-130-30
Fe3+
Fe2+
Fe2+
0.45
1.18
1.26
0.68
2.53
2.72
0.37
0.27
0.68
45
28
27
Fe7(PO4)6
Fe7(PO4)6
Fe2P2O7
a
Isomer shift, relative to ˛-Fe.
Electric quadrupole splitting.
Full linewidth at half maximum.
b
c
d
Relative resonance areas of different components of the absorption patterns.
explanation is due to the small crystal size and the high disper-
sion state of FePO4 over the support. Surprisingly, we found the
reduction behaviors of OP catalysts were extremely similar to that
of the bulk one, irrespective of the very large surface area of the
support and the relatively low loading of FePO4. These reduction
behaviors were possibly related to the crystal phase of iron phos-
phate. Wide-angle XRD results indicated the crystalline property of
the OP catalysts while the IMP catalyst was in an amorphous state,
highly in coincidence with the H2-TPR results.
conversions only at 12.1% for CH4 and 14.8% for O2 under the typical
OBM reaction conditions (Table 3). However, all these SBA-15-
supported catalysts showed similar high catalytic activities under
the same reaction conditions, with the IMP catalyst giving the high-
est methane conversion of 46.4% and the lowest CH3Br selectivity
of 26.6%. For OP-90, the methane conversion and CH3Br selectivity
were 45.7% and 32.2%, respectively. The lowest methane conver-
sion and CH3Br selectivity were observed over OP-130. Although
the physicochemical properties of these catalytical materials were
different in many aspects, the initial catalytic performances among
them were not huge. Thus, the IMP and OP-90 catalysts were further
chosen for stability test. The 100 h time-on-stream (TOS) perfor-
mance of the IMP catalyst is shown in Fig. 7a. During the stability
survey, the methane conversion slightly increased from 45.7% to
ca. 50.0%. The selectivity of CH3Br also increased slightly, but only
from 26.6% to ca. 30.0%. Fig. 7b shows the 1000 h TOS performance
of OP-90. A relatively stable methane conversion of 45.0% was
observed over the catalyst, and CH3Br selectivity kept increasing
slightly. On contrary to the IMP catalyst, the CH2Br2 selectivity
increased rapidly in the initial reaction stage. CH2Br2 is known
as an unwanted byproduct of the OBM reaction. In order to sup-
press the formation of this byproduct, the reaction temperature was
increased by 10 ◦C for two times from 570 to 580 ◦C and to 590 ◦C,
and kept at 590 ◦C for the rest of the evaluation process. Even so,
the methane conversion was rarely influenced and the CH3Br selec-
tivity even witnessed an increase from 40.0% to 45.0%. After 900 h,
the CH3Br selectivity further increased to ca. 50.0% while CO selec-
tivity showed a little drop. The CO2 selectivity was always very low
(ca. 2.0%) although it fluctuated during 550–750 h, possibly because
of the malfunction of the 40 wt.% HBr/H2O feeding syringe. Gener-
ally speaking, the OP-90 catalyst presented very stable catalytic
performance during the 1000 h TOS evaluation.
The nature of iron species of the as-prepared samples was char-
acterized by X-ray diffraction, but it was not sensitive enough to
explore the phase information of the IMP catalyst. In this sec-
tion, we employed room temperature57Fe Mössbauer spectroscopy
to analyze the iron species in the fresh samples, and the results
are presented in Fig. 6 and Table 2. Fig. 6a shows the Mössbauer
spectrum of the IMP catalyst with two unsymmetric peaks which
can be fitted into two set of doublets: one with ı = 0.39 mm/s and
ꢂ = 0.67 mm/s can be attributed to quartz FePO4; another one with
much higher ꢂ value might be ascribed to highly dispersed FePO4
with strong interaction with the support. These results were also
observed over FePO4/SiO2 catalyst, but the ı value (0.27 mm/s) was
a little lower over the latter one, resulting from the differences
of different support materials. Fig. 6c shows the Mössbauer spec-
trum of OP-90, with a highly symmetric doublet (ı = 0.28 mm/s,
ꢂ = 0.62 mm/s), characteristic of quartz FePO4. The ı value was also
lower over OP-90 than IMP, but close to that over silica. The Möss-
bauer spectrum of OP-130 is presented in Fig. 6e, showing two
unsymmetric peaks which can be fitted into two set of doublets.
Besides the common quartz FePO4 phase, tridymite FePO4 phase
was detected with similar ı value and lower ꢂ one (0.33 mm/s).
From the 57Fe Mössbauer spectroscopic analysis, we can conclude:
(i) impregnation method usually yielded highly dispersed FePO4
component while one-pot hydrothermal synthesis produced highly
crystallized FePO4 phase; (ii) crystallization temperature played
an important role in the final crystal phase of FePO4, and higher
crystallization temperature led to mixed FePO4 phases.
In a previous report, we found FePO4/SiO2 was subject to coke
deposition even after 100 h TOS experiment, which would cause
gradual catalyst deactivation and increasing CO2 selectivity [8].
So, we conducted TPO experiments over the two catalysts after
stability test. The results (Fig. S2, see Supporting Information)
suggested that no coke was deposited over the IMP and OP-
90 catalysts. N2-adsorption results showed that the structure of
FePO4/SiO2 was seriously destroyed after the 300 h OBM reaction,
with the mesopores sharply declined (Fig. S3, see Supporting Infor-
mation). Considering similar activities among the tested catalysts
and FePO4/SiO2, the superior anti-coking ability of the IMP and
3.2. Catalytic performance and stability test
The three prepared catalysts, together with commercially
obtained SBA-15, were employed in the OBM reaction, and the cat-
alytic performances are summarized in Table 3. Compared with
SBA-15-supported catalysts, SBA-15 alone was rather inactive with

242
R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
Table 3
Performances of the OBM reaction over catalysts with different synthesis methods.
Catalysts
Conversion (%)
CH4
Selectivity (%)
CH3Br
O2
14.8
100.0
97.7
CH2Br2
CO
CO2
SBA-15
IMP
OP-90
OP-130
12.1
46.4
45.7
45.1
56.3
26.6
32.2
25.5
2.1
0.5
1.3
0.8
41.6
70.6
64.6
70.7
0
2.4
2.0
3.0
100.0
Reaction conditions: T 590 ◦C, CH4 10 mL/min, O2 5 mL/min, CH4:O2 = 2:1 (v:v), 40 wt.% HBr/H2O 3.0 mL/h, catalyst 2.0 g.
OP-90 catalysts might be related to the high hydrothermal stability
of the mesoporous supports. To gain a deep insight into the above
hypothesis, the spent catalysts after stability tests were thoroughly
characterized and the results were discussed in the next section.
were more tolerant to the OBM reaction ambience than previously
reported FePO4/SiO2. One intuitive explanation could be due to the
much larger specific surface area of SBA-15 than SiO2 (363 m2/g).
In addition, the molar ratio of P/Fe for both samples dropped a little
for the loss was slower for Fe than P. As discussed earlier, the input
mole of phosphorus was largely in excess in the raw materials and
most of the excess P was removed after hydrothermal treatment
by washing. The lost P might come from the phases uncoordinated
with Fe atoms for iron phosphate phases were suggested to be very
stable in the OBM ambience [8].
3.3. Characterization of spent catalysts
3.3.1. Elemental and structural analyses
The loss of active components is common for the OBM reac-
tion [7,26] since bromine can be involved in the catalytic cycle for
methane activation, which presents a huge challenge for the long-
term stability of catalysts. FePO4/SiO2 was found to be a relatively
stable catalyst, after 100 h TOS performance, the relative losses
of P and Fe were estimated to be 19.1% and 30.9%, respectively.
For comparison purpose, these values were also calculated for the
spent catalysts as shown in Table 1. The relative losses of P and Fe
for OP-90-1000 and IMP-100 were 9.5% and 15.5%, and 5.4% and
8.0%, respectively. It is apparent that both the OP and IMP catalysts
We further used N2-adsorption to identify the structural
changes between the fresh and spent catalysts, with the results
summarized in Table 1 and Fig. 1. The hysteresis loops were still
observed in the relative pressure (P/P0) range of 0.6–0.8, indicative
of the maintenance of the mesoporous structure of catalysts. The
pore size distributions of catalysts before and after reactions were
also very close (insets in Fig. 1). Small-angle XRD patterns (Fig. 2b)
of IMP-100 and OP-90-1000 even showed enhanced diffraction
intensity of the (1 0 0) crystal plane. These observations indicated
the excellent structure stability of the catalysts during the reac-
tion. From Table 1, we can still see obvious reduction in specific
surface area and total pore volume for the spent catalysts. The
reduction of surface area was near 50% for OP-90-1000, mean-
while the average pore size increased from 5.3 to 8.1 nm. Similar
cases were found for the IMP-100 and OP-130-30. The morphology
of the spent catalysts was shown in Fig. 3c,d. Hexagonal array of
mesoporous channels were well maintained for OP-90-1000 which
implies that OP-90 is a rather stable catalyst for the harsh OBM
reaction. Several approaches have been employed to improve the
hydrothermal stability of SBA-15, and the mesoporous structures
of these SBA-15 materials maintain well after being treated in boil-
ing water or steamed at 600–800 ◦C [19,20]. In our experiment,
the mesoporous structures of OP-90 kept well after undergoing the
critical OBM reaction ambience with corrosive HBr/H2O and high
temperature for 1000 h. It suggested that OP-90 had rather high
hydrothermal stability and exhibited great advantages in severe
reaction conditions.
3.3.2. Phase analysis of iron phosphate
For supported iron phosphate samples with low loadings
(<40 wt.%), X-ray diffraction was almost impossible to gain the
phase information, especially when the supports possess a high
surface area. For the OP samples, corresponding to a FePO4 load-
ing of ca. 10 wt.%, strong and clear diffraction peaks were observed,
enabling us to identify the phase information of the spent catalysts.
As shown in Fig. 2a, several peaks were detected for the OP-90-
1000, corresponding to the diffraction of Fe2P2O7 and Fe7(PO4)6.
For the other two spent samples, no strong diffraction peaks were
observed, probably due to the detection limit. Thus, 57Fe Möss-
bauer spectroscopy was employed to identify the phases of the
used catalysts, with the results summarized in Fig. 6 and Table 2.
As can be seen, the spectra of all the spent catalysts showed
three peaks; one located at ca. 2.5 mm/s, convincingly demon-
strating the reduction from Fe3+ to Fe2+. Take OP-90-1000 for
example, the obtained spectrum was further fitted into three
Fig. 7. OBM reaction over IMP (a) and OP-90 (b) catalysts. Reaction conditions:
CH4 10 mL/min, O2 5 mL/min, CH4: O2 = 2:1 (v:v), 40 wt.% HBr/H2O 3.0 mL/h, catalyst
2.0 g, T 590 ◦C for (a). (᭹) CH4; (ꢁ) CH3Br; (ꢀ) CH2Br2; (ꢂ) CO; (ꢃ) CO2.

R. Wang et al. / Applied Catalysis A: General 453 (2013) 235–243
243
doublets; one with ı = 1.24 mm/s, ꢂ = 2.89 mm/s was attributed
Acknowledgments
to Fe2P2O7 with two Fe2+ in similar coordination environment
[27], the other two doublets with different hyperfine interaction
parameters were both ascribed to Fe7(PO4)6 that has four similar
Fe3+ atoms (ı = 0.42 mm/s, ꢂ = 0.68 mm/s) and three similar Fe2+
atoms (ı = 1.13 mm/s, ꢂ = 2.53 mm/s) [27,28]. These results were
in agreement with XRD observations. Similarly, both Fe2P2O7 and
Fe7(PO4)6 phases were found in the OP-130-30 and IMP-100, which
might explain the similar catalytic performances in the OBM reac-
tion of these catalysts. One should note that both ˛-Fe3(P2O7)2
and Fe2P2O7 were also found in the spent catalyst of FePO4/SiO2
in a previous report by 57Fe Mössbauer spectroscopy [8]. Like ˛-
Fe3(P2O7)2, Fe7(PO4)6 is also a kind of iron phosphates of mixed
valence irons (Fe3+:Fe2+ = 2:1). In this sense, a similar redox reac-
tion route predominantly governed by the redox pair of Fe3+/Fe2+
might be applied to the OP catalysts.
For both impregnated and one-pot hydrothermally synthesized
iron phosphate catalysts with SBA-15 as the support, the initial
FePO4 phases (either quartz or tridymite) were partially reduced,
yielding Fe7(PO4)6 and Fe2P2O7 during the OBM reaction. Elemen-
tal analysis has revealed that these active phases were extremely
stable under corrosive reaction ambience and well maintained even
after 1000 h TOS evaluation, which might be a main reason for the
long-term stability of catalysts even under such harsh reaction con-
ditions. Besides, both N2-adsorption and TEM analysis indicated
the excellent structural stability of mesoporous pores under OBM
reactions, which might be another important reason for the good
anti-coking ability.
We thank Dr. Tao Liu for help in TEM test and analysis. We also
appreciate visiting professor Alexandre I. Rykov and Ms. Yan Zhang
for help in 57Fe Mössbauer spectroscopy test and analysis. This
work was financially supported by the National Natural Science
Foundation of China (No. 21103170).
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.
2012.12.022.
References
[1] V. Degirmenci, A. Yilmaz, D. Uner, Catal. Today 142 (2009) 30–33.
[2] G.A. Olah, B. Gupta, M. Farina, J.D. Felberg, W.M. Ip, A. Husain, R. Karpe-
les, K. Lammertsma, A.K. Melhotra, N.J. Trivedi, J. Am. Chem. Soc. 107 (1985)
7097–7105.
[3] S.G. Podkolzin, E.E. Stangland, M.E. Jones, E. Peringer, J.A. Lercher, J. Am. Chem.
Soc. 129 (2007) 2569–2576.
[4] I. Bucsi, G.A. Olah, Catal. Lett. 16 (1992) 27–38.
[5] H.F. Xu, K.X. Wang, W.S. Li, X.P. Zhou, Catal. Lett. 100 (2005) 53–57.
[6] K.X. Wang, H.F. Xu, W.S. Li, X.P. Zhou, J. Mol. Catal. A: Chem. 225 (2005)
65–69.
[7] R. Lin, Y. Ding, L. Gong, J. Li, W. Chen, L. Yan, Y. Lu, Appl. Catal. A 353 (2009)
87–92.
[8] R. Lin, Y. Ding, L. Gong, W. Dong, J. Wang, T. Zhang, J. Catal. 272 (2010)
65–73.
[9] E. Peringer, S.G. Podkolzin, M.E. Jones, R. Olindo, J.A. Lercher, Top. Catal. 38
(2006) 211–220.
[10] F. Yang, Z. Liu, W.S. Li, T.H. Wu, X.P. Zhou, Catal. Lett. 124 (2008) 226–232.
[11] R. Lin, Y. Ding, L. Gong, W. Dong, W. Chen, Y. Lu, Catal. Today 164 (2011)
34–39.
4. Conclusions
[12] J. He, T. Xu, Z. Wang, Q. Zhang, W. Deng, Y. Wang, Angew. Chem. Int. Ed. 51
(2012) 2438–2442.
FePO4-SBA-15 was directly synthesized via
a
one-pot
[13] F. Ying, J. Li, C. Huang, W. Weng, H. Wan, Catal. Lett. 115 (2007) 137–142.
[14] E. Briot, J.Y. Piquemal, J.M. Bregeault, New J. Chem. 26 (2002) 1443–1447.
[15] Z. Lou, R. Wang, H. Sun, Y. Chen, Y. Yang, Micropor. Mesopor. Mater. 110 (2008)
347–354.
[16] Y. Li, Z.C. Feng, Y.X. Lian, K.Q. Sun, L. Zhang, G.Q. Jia, Q.H. Yang, C. Li, Micropor.
Mesopor. Mater. 84 (2005) 41–49.
[17] Y. Li, W.H. Zhang, L. Zhang, Q.H. Yang, Z.B. Wei, Z.C. Feng, C. Li, J. Phys. Chem. B
108 (2004) 9739–9744.
[18] M. Colilla, F. Balas, M. Manzano, M. Vallet-Regí, Solid State Sci. 10 (2008)
408–415.
[19] F.Q. Zhang, Y. Yan, H.F. Yang, Y. Meng, C.Z. Yu, B. Tu, D.Y. Zhao, J. Phys. Chem. B
109 (2005) 8723–8732.
[20] C. Li, Y. Wang, Y. Guo, X. Liu, Y. Guo, Z. Zhang, Y. Wang, G. Lu, Chem. Mater. 19
(2007) 173–178.
[21] J.C. Groen, J. Perez-Ramirez, Appl. Catal. A 268 (2004) 121–125.
[22] Y. Wang, X.X. Wang, Z. Su, Q. Guo, Q.H. Tang, Q.H. Zhang, H.L. Wan, Catal. Today
93-5 (2004) 155–161.
[23] Y. Li, Z. Feng, H. Xin, F. Fan, J. Zhang, P.C.M.M. Magusin, E.J.M. Hensen, R.A. van
Santen, Q. Yang, C. Li, J. Phys. Chem. B 110 (2006) 26114–26121.
[24] Y. Li, Z. Feng, R.A. van Santen, E.J.M. Hensen, C. Li, J. Catal. 255 (2008)
190–196.
hydrothermal method, with Fe(NO3)3 and H3PO4 as the iron
and phosphorous precursors, respectively. Although the molar
ratio of P/Fe was greatly beyond the stoichiometric value of FePO4
in the raw synthesis material, this ratio was close to 1.0 in the final
yielding sample; besides, the FePO4 was in good crystallinity even
when the nominal loading was as low as ca. 10 wt.%, which was
different from the impregnated sample of FePO4/SBA-15 (FePO4 in
amorphous at a loading lower than 40 wt.%). Both the OP and IMP
samples were used to catalyze the OBM reaction, giving similar
catalytic performances. 57Fe Mössbauer spectroscopy analysis
confirmed that despite the differences in the initial active phases
in OP and IMP catalysts, the final active components of the spent
catalysts were consisted of Fe7(PO4)6 and Fe2P2O7. Compared
with previously reported FePO4/SiO2 catalyst, the OP catalyst
showed much superior stability performance. The relative losses of
P and Fe were only 9.5% and 15.5%, respectively, after 1000 h TOS
evaluation, while the ratio of P/Fe remained close to 1.0. Besides,
the mesoporous pores were extremely stable under the harsh
reaction ambience, which might play a crucial role in the stability
test.
[25] X.X. Wang, Y. Wang, Q.H. Tang, Q.A. Guo, Q.H. Zhang, H.L. Wan, J. Catal. 217
(2003) 457–467.
[26] Z. Liu, W. Li, X. Zhou, J. Nat. Gas Chem. 19 (2010) 522–529.
[27] J.M.M. Millet, Catal. Rev. 40 (1998) 1–38.
[28] M.M. Gadgil, S.K. Kulshreshtha, J. Solid State Chem. 111 (1994) 357–364.