Effects of iron content on bismuth molybdate for the oxidative dehydrogenation of n-butenes to 1,3-butadiene

Article abstract of DOI:10.1016/j.apcata.2012.04.028

BiMoFe_x oxide catalysts (x = 0-1.00) were prepared by co-precipitation and their catalytic activities in the oxidative dehydrogenation of n-butenes were tested. X-ray diffraction (XRD) and Raman spectroscopy showed that the main solid phases were composed of Bi₃Mo ₂FeO₁₂ as oxygen acceptor and Fe₂(MoO ₄)₃ as oxygen donor and the mixing of these phases enhanced catalytic activity. XRD patterns showed that Fe₂(MoO ₄)₃ was reduced to FeMoO₄ during the reaction. The peak temperature of programmed reduction of 1-butene and successive oxidation (TPRO) was dependent on Fe contents in BiMoFe_x oxide catalysts and was minimized at x = 0.65, which showed the greatest oxygen mobility. The peak position in the low temperature region of TPRO profiles could be correlated with butene conversion and BD yield.

Full text of DOI:10.1016/j.apcata.2012.04.028

Applied Catalysis A: General 431–432 (2012) 137–143
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Effects of iron content on bismuth molybdate for the oxidative dehydrogenation
of n-butenes to 1,3-butadiene
a
b
b
b
c
a,∗
Jung-Hyun Park , Hyeryeung Noh , Ji Won Park , Kyoungho Row , Kwang Deog Jung , Chae-Ho Shin
a
Department of Chemical Engineering, Chungbuk National University, Chungbuk 361-763, Republic of Korea
Process Solution Team, Kumho Petrochemical R&BD Center, Daejeon 305-348, Republic of Korea
Clean Energy Center, Korea Institute of Science and Technology, Seoul 136-791, Republic of Korea
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
BiMoFe oxide catalysts (x = 0–1.00) were prepared by co-precipitation and their catalytic activities in the
oxidative dehydrogenation of n-butenes were tested. X-ray diffraction (XRD) and Raman spectroscopy
x
Received 8 November 2011
Received in revised form 31 March 2012
Accepted 19 April 2012
showed that the main solid phases were composed of Bi3Mo2FeO12 as oxygen acceptor and Fe2(MoO4)3
as oxygen donor and the mixing of these phases enhanced catalytic activity. XRD patterns showed that
Fe2(MoO4)3 was reduced to FeMoO4 during the reaction. The peak temperature of programmed reduction
of 1-butene and successive oxidation (TPRO) was dependent on Fe contents in BiMoFex oxide catalysts
and was minimized at x = 0.65, which showed the greatest oxygen mobility. The peak position in the low
temperature region of TPRO profiles could be correlated with butene conversion and BD yield.
Available online 26 April 2012
Keywords:
Bismuth molybdenum iron oxide
Oxidative dehydrogenation
n-Butenes
©
2012 Elsevier B.V. All rights reserved.
1
.3-butadiene
Oxygen mobility
1
. Introduction
,3-Butadiene (BD) is an important raw material for the manu-
not fully elucidated. However, it is generally accepted that oxygen
mobility of bismuth molybdate catalysts affects the ODH of butenes,
likely related to the Mars-van Krevelen mechanism [11–13,17].
Ruckenstein et al. [22] reported that oxygen diffusivity and cat-
alytic activity could be varied with phase, increasing in the order:
␥ > ␤ > ␣. Jung et al. [18] also reported that ␥-Bi MoO showed good
catalytic activity in the ODH of butenes due to facile oxygen mobil-
ity. Multicomponent oxides based on bismuth molybdate for ODH
reaction have been extensively studied to enhance catalytic activity
by adding oxides such as P, Co, Fe, and Ni, etc. [7,20,23–28].
This work reports the comparison of catalytic activities for the
ODH of 1-butene to BD over BiMoFex oxide catalysts with different
iron contents (x = 0–1.00) prepared by co-precipitation. The Bi/Mo
molar ratio was set at 1.0 and the various concentrations of iron
were tested to maximize the conversion and selectivity to BD in
the ODH of 1-butene. BiMoFex oxide catalysts with different Fe
contents were characterized by using N2 sorption, X-ray diffrac-
tion (XRD), temperature-programmed reduction (TPR)-successive
oxidation (TPRO), 1-butene temperature-programmed desorption
(TPD), Raman spectroscopy, and elemental analysis.
1
facture of such as styrene butadiene rubber, poly-butadiene rubber,
and acrylonitrile–butadiene–styrene resin and demand for it is
increasing. It is mainly produced by extracting C4 raffinates from
naphtha cracking, though the limited supply of C4 raffinates has
encouraged the use of alternative intermediates such as butenes
and ethanol for BD production. The oxidative dehydrogenation
2
6
(
ODH) of butenes can efficiently produce BD in high yields [1,2].
Tested transition metal oxide catalysts for the ODH of 1-butene
include manganese oxide molecular sieves [3], K-doped VOx/Al O3
2
[
4], zinc ferrites [5–9], and multicomponent oxides based on bis-
muth molybdate [10–19]. Bismuth molybdates can be classified
into three phases – ␣-Bi Mo O , ␤-Bi Mo O , and ␥-Bi MoO
–
6
2
3
12
2
2
9
2
according to Bi/Mo molar ratio and its structural effects have been
investigated for the ODH of butenes [10–15], the ammoxidation of
propylene [20] and the selective oxidation of propane to acrolein
[
21]. Matsuura et al. [15] have compared the catalytic activities of
bismuth molybdate catalysts with different molar ratios of Bi/Mo.
The catalytic activity of ␥- and ␣-type bismuth molybdate has been
less than ␤-type or mixed phases of ␥- and ␣-type, though the
structural effects of bismuth molybdate on its catalytic activity are
2. Experimental
2.1. Preparation of bismuth molybdenum iron catalysts
∗ Corresponding author. Tel.: +82 43 261 2376; fax: +82 43 269 2370.
E-mail address: chshin@chungbuk.ac.kr (C.-H. Shin).
BiMoFex oxide catalysts (x = 0–1.00) were prepared by con-
ventional co-precipitation. The desired amount of bismuth
0
926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.04.028

1
38
J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
nitrate (Bi(NO ) ·5H O, Junsei Chemical, 98%) was dissolved in
of Bi Mo FeO and Fe (MoO ) (9:1 in wt%) were performed
3 2 12 2 4 3
at 420 C. Each 0.1 g catalyst was placed in the U-shape quartz
3
3
2
◦
de-ionized water acidified with 10% nitric acid and added drop-
wise into ammonium molybdate solution ((NH ) Mo7O ·4H O,
reactor (I.D. 12 mm) with densely packed quartz wool. All cata-
4
6
24
2
◦
◦
Sigma–Aldrich, 99%) under vigorously stirring at 60 C. Ferric
lysts were pretreated at 420 C for 1 h under flowing 5% H /Ar
2
3
−1
nitrate (Fe(NO ) ·9H O, Samchun, 98%) in de-ionized water was
(50 cm min ) and subsequently purged for 0.5 h under flowing
3
3
2
3
−1
then added. The mixture was adjusted to pH 5 by aqueous 35 wt%
He (50 cm min ). After that, pulse O2 adsorption experiments
were carried out by contacting fixed amounts of reactant gases with
He as the carrier gas. The amount of O2 in each pulse, which was
introduced using a volumetric sample loop of 2 cm3 (71 ␮mol) and
NH OH and aged at room temperature for 4 h. Excess water was
4
◦
then removed at 60 C using a rotary evaporator. The resulting pow-
◦
◦
der was dried at 100 C overnight and calcined at 550 C for 2 h in
an air flow. The catalysts were labeled BiMoFex (x = 0–1.0). For the
comparison of catalytic activities of BiMoFex oxide catalysts, bis-
muth molybdate with different molar ratios of Mo/Bi = 3/2, 2/2, and
the pulse was repeated seven times with 10 min interval. The mass
•
signal of m/z = 32 ( O ) was detected.
2
1
/2 were prepared by co-precipitation method.
2.3. Oxidative dehydrogenation (ODH) of 1-butene
The ODH of 1-butene was performed at atmospheric pres-
2
.2. Characterization
sure in a conventional continuous-flow microreactor. Prior to the
reaction, the catalyst was routinely activated under flowing N
Elemental analysis was carried out with a Jarrell-Ash Polyscan
1E inductively coupled plasma spectrometer with a PerkinElmer
000 atomic emission spectrophotometer (ICP-AES). Powder XRD
2
3
−1
◦
◦
(
50 cm min ) at 500 C for 2 h. The reaction temperature was
6
5
fixed at 420 C and 1-butene, air and water reagents with a 1:5.75:5
3
−1
molar ratios were used at a constant flow rate of 100 cm min
.
patterns were collected on a Siemens D5005 diffractometer with Cu
K␣ radiation (30 kV and 50 mA). Raman spectra were recorded on
a Bruker Optic GMBH FRA 106/S with an Nd:YAG laser (300 mW,
The water feed was continuously vaporized in a pre-heating zone
◦
at 180 C. The reaction products were analyzed by on-line gas
chromatography (Varian 3800) on an Al O /KCl column of 50 m
5
00 scan). Raman spectra of pure Bi Mo FeO and Fe (MoO )
2
3
3
2
12
2
4 3
length and 0.32 mm diameter with a flame ionization detector (FID)
to analyze hydrocarbons and a Porapak Q packed column of 2 m
length and 1/8 in. diameter with a thermal conductivity detector
synthesized by co-precipitation were recorded for reference. BET
surface areas and total pore volumes were determined from N2
◦
adsorption isotherms (Micromeritics ASAP 2020) at −196 C. The
(
TCD) to analyze CO and CO . The reaction products consisted of
relationship between catalytic activity and Fe content was assessed
through the temperature-programmed reduction of 1-butene and
subsequent oxidation (TPRO) recorded on fixed-bed, flow-type
apparatus attached to a Balzers QMS200 quadruple mass spec-
2
dehydrogenation products (BD), isomerization products (trans and
cis-2-butene) and combustion products (CO and CO ). Cracking
2
products (CH , C H , C H , C H and C H ) were almost negligible.
4
2
4
2
6
3
6
3
8
The isomerization products of 1-butene to cis- and trans-2-butene
were considered as reactants. The conversion of butenes was calcu-
lated as the molar ratio of the butenes reacted to the 1-butene fed.
The selectivity and yield in BD were calculated on the basis of mass
balance and the yield was obtained by multiplying the conversion
and the selectivity of BD.
trometer. For the temperature programmed reduction by 1-butene,
◦
0
.3 g catalyst was pretreated at 420 C for 3 h in flowing 1-butene
3
−1
(
20 cm min ), held for 1 h at that temperature and then cooled
3
−1
to room temperature in flowing He (20 cm min ). Subsequent
temperature programmed oxidation was performed in flowing
5
1
tus. 1-Butene was adsorbed at 50 C and purged under flowing
He (50 cm min ) for 0.5 h before TPD from 50 to 800 C at
3
−1 ◦
% O /N (20 cm min ) from room temperature to 500 C at
2 2
◦
−1
0 C min . 1-Butene TPD was carried out on the same appara-
◦
3. Results and discussion
3
−1
◦
◦
−1
•
•
•
1
0 C min . Mass signals of m/z = 18 ( H O), 44 ( CO ), 54 ( C H ),
3.1. Characterization of the BiMoFex oxide catalysts
2
2
4
6
•
and 56 ( C H ) were detected.
4
8
In order to elucidate the concept of oxygen acceptor and donor
Table 1 lists BET surface area, total pore volume, elemental anal-
in BiMoFex oxide catalysts, Bi Mo FeO and Fe (MoO ) which
were the main component in BiMoFex oxide catalysts were inde-
yse and solid phases of the fresh and used BiMoFex oxides. BET
3
2
12
2
4 3
◦
surface area and total pore volume of samples calcined at 550 C
2
−1
pendently prepared and the mechanical mixture of Bi Mo FeO
increased with increasing Fe content in the range 1.4–3.8 m g
.
3
2
12
and Fe (MoO ) was also prepared by using ultrasonic waves in
Elemental analysis showed that the samples maintained atomic
ratios of each element similar to their nominal compositions.
XRD patterns of samples were recorded before and after the
ODH of 1-butene at 420 C for 14 h (Fig. 1). The XRD patterns of
bismuth molybdate with different molar ratios of Bi/Mo before
and after ODH reaction are also shown in Table SI 1 and Fig. SI 1.
2
4 3
hexane as solvent. The composition of the mixture of Bi Mo FeO
3
2
12
and Fe (MoO ) was 9:1 in wt%. After that, the hexane was removed
2
4 3
◦
◦
by using a rotary evaporator and dried at 100 C for 12 h. The
catalysts were used without heat treatment. Pulse O2 adsorp-
tion experiments of Bi Mo FeO , Fe (MoO ) and the mixture
3
2
12
2
4 3
Table 1
Atomic ratio, BET surface area, total pore volume and XRD phases of BiMoFex oxide catalysts with different Fe contents.
Atomic ratioa
SBET (m2
g
−1
)
Pore volume
XRD phases calcinedb
After reactionc
Catalyst
3
−1
)
(
cm
g
Bi
Mo
Fe
BiMoFe0.20
BiMoFe0.35
BiMoFe0.50
BiMoFe0.65
BiMoFe0.75
BiMoFe0.85
BiMoFe1.00
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.98
0.97
0.96
0.97
0.95
0.98
1.04
0.21
0.34
0.50
0.65
0.74
0.87
1.07
2.6
2.8
3.1
3.6
3.8
3.5
3.4
0.008
0.009
0.012
0.014
0.023
0.027
0.030
␤-Bi2Mo2O9 > Bi3Mo2FeO12 > ␣-Bi2Mo3O12
␤-Bi2Mo2O9 > Bi3Mo2FeO12 > ␣-Bi2Mo3O12
Bi3Mo2FeO12 > Fe2(MoO4)3
Bi3Mo2FeO12 > Fe2(MoO4)3
Bi3Mo2FeO12 > Fe2(MoO4)3
␤-Bi2Mo2O9 > Bi3Mo2FeO12 > ␣-Bi2Mo3O12
␤-Bi2Mo2O9 > Bi3Mo2FeO12 > ␣-Bi2Mo3O12 > FeMoO4
Bi3Mo2FeO12 > FeMoO4
Bi3Mo2FeO12 > FeMoO4
Bi3Mo2FeO12 > FeMoO4
Bi3Mo2FeO12 > Fe2(MoO4)3
Bi3Mo2FeO12 > Fe2(MoO4)3
Bi3Mo2FeO12 > FeMoO4
Bi3Mo2FeO12 > FeMoO4
a
Determined by ICP-AES.
Calcined at 550 C for 2 h in air flow.
XRD patterns were obtained after 14 h on stream in ODH of 1-butene at 420 C for BiMoFex oxide catalysts.
b
c
◦
◦

J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
139
◦
Fig. 1. Powder XRD patterns of BiMoFex oxide catalysts with different Fe contents before (A) and after 14 h reaction at 420 C (B); (a) mixture of ␤-Bi2Mo2O9 and ␥-Bi2MoO6,
b) x in BiMoFex oxide = 0.2, (c) 0.35, (d) 0.5, (e) 0.65, (f) 0.75, (g) 0.85, and (h) 1.00.
(
783 cm 1 and MoO3 and MoO4 modes at 283 cm and 355 cm
−1
,
−
−1
The solid phases of ␣-Bi Mo O and ␥-Bi MoO catalyst were
2
3
12
2
6
◦
not changed after ODH of 1-butene at 420 C for 6 h. However, ␤-
respectively [33]. MoO contains six oxygen atoms in an octahedral
structure; with five of the oxygen atoms shared in distorted MoO6
structures and the remainder as unshared Mo O. BiMoFe0.2 oxide
catalyst showed low-intensity absorption bands at 240–450 cm
and 900 cm ascribed to the ␣ phase [30]. The band at 958 cm
3
◦
Bi Mo O catalyst, which was stable after calcination at 350 C, was
2
2
9
partly decomposed to ␤-Bi Mo O and ␥-Bi MoO6 by the calcina-
2
2
9
2
◦
−1
−1
tion at 550 C. In addition, the peak intensity of ␥-Bi MoO phase
2
6
−
1
increased after ODH of 1-butene, while that of ␤-Bi Mo O phase
2
2
9
−
1
decreased. This means the addition of H O in reactants is favorable
was attributed to the ␤ phase; that at 816 cm to the ␥ phase [15].
2
to decompose ␤-Bi Mo O to ␣- and ␥-phase.
Raman spectra of BiMoFe oxide catalysts with x > 0.65 were almost
2
2
9
x
The solid phases of the fresh BiMoFex oxides were consisted
of ␤-Bi Mo O , Bi Mo FeO , ␣-Bi Mo O , and Fe (MoO ) . Bis-
consistent with that of pure Bi Mo FeO [26], supporting the XRD
pattern.
3
2
12
2
2
9
3
2
12
2
3
12
2
4 3
muth molybdate can be formed in three phases: ␣-Bi Mo O12,
Many parameters have been proposed for the enhancement of
catalysis in the ODH of 1-butene to favor BD production. They can
be classified into five groups related to lattice oxygen ions, cations,
2
3
␤
-Bi Mo O and ␥-Bi MoO6 [10–15]. At x < 0.35, fresh samples
2 2 9 2
showed mainly ␤-Bi Mo O and minor phases of Bi Mo FeO and
2
2
9
3
2
12
␣
-Bi Mo O ␤-Bi Mo O with Bi/Mo = 1/1 is thermally unstable
2 3 12. 2 2 9
◦
at 420 C and was decomposed to a mixture of ␣- and ␥-phases
15]. However, the ␤-Bi Mo O in BiMoFe oxide was conserved
[
2
2
9
0.35
after 14 h ODH of 1-butene and a new solid phase of FeMoO4 was
observed by the partial decomposition of Bi Mo FeO12. BiMoFex
3
2
oxide with x > 0.5 showed crystalline phases of Bi Mo FeO and
3
2
12
Fe (MoO ) . Fe (MoO ) could be partially transformed to FeMoO
2
4
3
2
4
3
4
after 14 h ODH of 1-butene, changing the oxidation state of Fe from
3 to +2. The formation FeMoO4 was maximized at x = 0.65. After
4 h reaction, ␣-Bi Mo O in BiMoFex decreased with increasing
+
1
2
3
12
Fe content and was not observed in BiMoFe0.65 oxide. Stoi-
chiometrically, BiMoFe0.65 oxide should comprise equal parts of
Bi Mo FeO and FeMoO . However, XRD patterns showed mainly
3
2
12
4
Bi Mo FeO with a minor FeMoO4 phase after 14 h reaction.
3
2
12
Raman spectroscopy was used to investigate the composi-
tions and structures of catalysts (Fig. 2). Pure Bi Mo FeO and
3
2
12
Fe (MoO ) , verified by XRD and elemental analysis, were used
2
4 3
as references. Optical signals representing the vibrational modes
−
1
of MoO4 have been reported at 700–900 cm and its bending at
−1
−1
3
00–400 cm . BiO bands have been recorded at 400–600 cm
[
15,29–31]. Hoefs et al. [30] suggested that the bands at 720
−1
and 680 cm
were due to the dioxo bridge connected to the
molybdenum atom, which shows lattice oxygen. Pure bismuth iron
molybdate, Bi Mo FeO which could be noted as Bi FeO (MoO )
3
2
12
3
4
4 2
Fig. 2. Raman spectra of BiMoFex oxide catalysts with different Fe contents calcined
at 550 C for 2 h; (a) pure Bi3Mo2FeO12, (b) x in BiMoFex oxide = 0.2, (c) 0.5, (d) 0.65,
(e) 0.75, (f) 0.85, (g) 1.00 and (h) pure Fe2(MoO4)3.
−
1
showed bands at 875, 783, 715, 323, 283, 133 cm , in good agree-
ment with reported data [31]. Fe (MoO ) showed a main peak at
◦
2
4
3

1
40
J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
•
Fig. 4. MS signals of 1-butene TPD of BiMoFe0.65 oxide catalyst. m/z = 18 ( H2O), 44
•
• •
CO2), 54 ( C4H6) and 56 ( C4H8) were recorded during 1-butene TPD analysis.
(
between the oxygen mobility and catalytic performance of BiMoFex
oxide catalysts in the ODH of n-butenes to BD. The catalytic activity
of the BiMoFex oxide catalysts with different Fe content was related
to the oxygen mobility. Yield in BD increased with decreasing peak
temperature. This result means that the catalytic activity could be
increased with increasing oxygen mobility of the catalyst. The oxy-
gen mobility showed a linear plot with respect to BD yield and
the peak temperature of TPRO profiles was minimized at x = 0.65,
indicating that BiMoFe0.65 oxide catalyst had the highest oxygen
mobility.
The adsorption of 1-butene on the catalysts was assessed by
TPD (Figs. 4 and 5). Fig. 4 shows 1-butene TPD profiles from the
•
most active catalyst, BiMoFe0.65 oxide. The m/z = 56 ( C H ) mass
4
8
signal progressively decreased with increasing desorption tem-
•
perature. The mass signal m/z = 54 ( C H ), possibly produced by
4
6
the dehydrogenation of 1-butene, showed maxima at 70–140 and
Fig. 3. (A) TPRO profiles of partially reduced BiMoFex oxide catalysts and (B) correla-
tion between catalytic performance and oxygen mobility of BiMoFex oxide catalysts;
(
a) x in BiMoFex oxide = 0.2, (b) 0.5, (c) 0.65, (d) 0.85, and (e) 1.00.
electron mobility, cation-oxygen bonds and crystallographic struc-
tures [32]. The most widely applied parameter is oxygen mobility,
which can be verified by TPRO and TPD analyses. In order to exam-
ine the role of lattice oxygen in the ODH of n-butenes to BD, TPRO
analysis of BiMoFex oxide catalysts was performed (Fig. 3). Prior
to TPRO analysis, catalysts were reduced under flowing 1-butene
◦
at 420 C for 3 h. It is generally accepted that peak position in low
temperature region of TPRO profile, that is related with the oxygen
mobility, could be correlated with the catalytic activities for the
ODH reaction of n-butenes [10,18]. Two main peaks were observed
for each catalyst, implying the BiMoFex oxide had diverse oxygen
◦
vacancy sites. The first peak is observed in the 130–230 C and the
◦
other one is in the 320–450 C. The first peak of TPRO profiles could
be contributed to the oxidation of bismuth species partially reduced
and the second peak is assigned to the oxidation process of molyb-
denum and iron partially reduced [34]. Fig. 3(B) shows a correlation
•
Fig. 5. MS signals of m/z = 54 ( C4H6) recorded during 1-butene TPD of BiMoFex
oxide catalyst with different Fe contents; (a) Bi3Mo2FeO12, (b) Fe2(MoO4)3, (c) x in
BiMoFex oxide = 0.2, (d) 0.5, (e) 0.65, (f) 0.75, (g) 0.85, and (h) 1.00.

J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
141
mixed in weight. The amount of consumed O2 on the mix-
ture of Bi Mo Fe O + Fe (MoO ) oxide was much larger
3
2
1
12
2
4 3
than that of the only Bi Mo FeO
and Fe (MoO ) cat-
3
2
12
2 4 3
alyst. The sequence of O2 consumption is Bi Mo FeO +
3
2
12
Fe (MoO ) > Bi Mo FeO > Fe (MoO ) . This means that the
2
4
3
3
2
12
2
4 3
adsorption capability of mixed Bi Mo FeO + Fe (MoO ) cat-
3
2
12
2
4 3
alyst is higher than that of pure Bi Mo FeO and Fe (MoO )
3
2
12
2
4 3
catalyst, meaning that the addition of Fe (MoO ) phase in pure
2
4 3
Bi Mo FeO phase could promote the O2 adsorption. Weng and
3
2
12
Delmon [32] reported that the synergy of multicomponent oxide
catalysts containing several oxide phases is attributed to the
remote association of the donor and acceptor of oxygen or elec-
trons in allylic oxidation and ODH. In such a mechanism, a donor
phase dissociates oxygen to form a surface mobile species that can
move to the acceptor phase, which is potentially the active phase.
Stable Bi Mo FeO has been shown to be an active phase (Table 2).
3
2
12
Therefore, the less active redox couple of Fe (MoO ) /FeMoO
2
4
3
4
could be a donor of oxygen. Chang et al. [33] reported that the
presence of Fe (MoO ) phase in multicomponent BiMoFeCoP
2
4 3
oxide promoted the mobility of surface oxygen or electrons formed
by the transformation of Fe3+ to Fe . Thus, in our system, it can be
2+
deduced that the phase of Bi Mo FeO and Fe (MoO ) could be
3
2
12
2
4 3
acceptor and donor of oxygen, respectively.
.2. ODH of 1-butene
Based on bismuth molybdate catalyst with Bi/Mo = 1/1, cat-
3
alytic activities over BiMoFex oxide catalysts (x = 0–1.00) which
were butene conversion, selectivities to BD and CO2 and yield
in BD are summarized in Table 2 and Fig. 7. For comparison,
the catalytic performance of ␣-Bi Mo O , ␤-Bi Mo O , and ␥-
2
3
12
2
2
9
Bi MoO6 catalyst was also tested and the results were shown
2
in Table 2 and SI 2. Although initial conversion of butenes on
␥
-Bi MoO6 catalyst was superior to the other bismuth molyb-
2
dates catalysts, the rapid deactivation was observed. Among the
bismuth molybdates catalysts used, the catalytic activity of bis-
muth molybdate with Bi/Mo = 1/1 was stable and showed the best
catalytic activity after 6 h on stream in the ODH reaction. The
main products were BD and CO , with negligible by-products of
2
C1–C3 compounds formed by C C bond cleavage of 1-butene.
Pure solid phase Bi Mo FeO catalyst was stable during the reac-
3
2
12
tion and showed the highest initial yields of BD and CO . BD
2
yields of the BiMoFex oxide catalysts were much higher than
that of pure bismuth molybdates and iron molybdates indi-
◦
Fig. 6. Pulse O2 adsorption profiles at 420 C for (A) partially reduced catalysts
of (a) Fe2(MoO4)3, (b) Bi3Mo2FeO12, and (c) Bi3Mo2FeO12/Fe2(MoO4)3 = 9/1 which
was mechanically mixed in weight, and (B) the quantity of total O2 adsorbed.
Pulse O2 adsorption experiments were repeated seven times with 10 min interval
cating that Bi Mo FeO was an active phase in the ODH of
3
2
12
1-butene. After 14 h reaction, BiMoFe0.65 oxide catalyst consist-
ing of Bi Mo FeO and FeMoO4 showed the highest butene
◦
at 420 C and O2 consumption was calculated by integrating each pulse signal of
3
2
12
O2.
conversion and BD yield. Conversion on BiMoFex oxides with
x < 0.65 did not increase greatly with time on stream. BD yield
on BiMoFex with x > 0.65 decreased, indicating that isolated Fe
lowered BD selectivity. The low conversion and selectivity of
Fe2(MoO4)3 indicate that the redox couple of Fe2(MoO4)3/FeMoO4
was not directly related to the catalytic reaction. Overall these
results showed that enhanced catalysis by BiMoFex oxide in the
ODH of 1-buene was due to the coexistence of Bi3Mo2FeO12 and
Fe2(MoO4)3. Water is widely used in the dehydrogenation pro-
cesses and it is added to help in the removal of the heat released
in the dehydrogenation reaction [35]. When water was not fed
in reactant stream, the conversion of n-butenes and the selectiv-
ity to CO2 were higher than that of while the water-cofeeding
[11]. However, the heat of reaction by the formation of CO2 is
∼20 times higher comparing to that for the formation of BD.
The higher heat of reaction may cause hot spot during the ODH
reaction. Therefore, it is necessary that the water is supplied for
the suppression of CO2 formation and safe operation in the ODH
reaction.
◦
4
00–550 C, respectively, due to weakly or strongly adsorbed 1-
butene with lattice oxygen of the catalyst.
-Butene TPD profiles from the samples with different Fe con-
1
tents show two desorption peaks (Fig. 5). The low temperature
peak (␣ position) was related to oxygen donor sites; that at higher
temperature (␤ position) was related to the active sites for dehydro-
genation. The temperatures of both ␣ and ␤ positions were lowest
at x = 0.65 in BiMoFex oxide catalysts. Desorbed BD was also max-
imized at x = 0.65. Pure solid phase Fe (MoO ) did not show ␤
position, suggesting that the ␣ position was related to oxygen donor
sites. For the better understanding the role of oxygen acceptor and
donor of BiMoFex oxide catalysts we performed the experiments of
2
4 3
pulse O2 adsorption.
◦
Fig.
6
shows pulse O2 adsorption profiles at 420 C for
partially reduced catalysts of Fe (MoO ) , Bi Mo FeO12,
and Bi Mo FeO /Fe (MoO ) = 9/1 which was mechanically
2
4
3
3
2
3
2
12
2
4 3

1
42
J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
Table 2
Catalytic performance of BiMoFex oxide catalysts with different Fe contents.
Catalyst
Conversion (%)
Selectivity (wt%)
1,3-BD
Yield (wt%)
1,3-BD
t = 0
t = 0
14 h
CO2
t = 0
14 h
t = 0
14 h
14 h
␣
␤
␥
-Bi2Mo3O12a
15.1
48.9
47.8
40.9
41.7
43.7
50.3
42.4
49.3
49.5
70.5
16.4
11.8
16.7
43.6
10.9
40.0
46.0
52.4
68.6
67.2
65.1
57.7
70.8
22.9
16.4
64.3
81.0
75.4
90.5
82.1
86.9
85.5
85.5
79.7
72.1
77.5
61.4
31.8
77.8
84.1
38.8
91.4
92.0
92.7
91.4
90.7
87.6
79.6
79.2
69.5
39.9
29.8
18.5
10.1
9.2
17.7
12.3
14.3
14.6
17.3
19.6
22.2
37.1
67.1
13.0
15.7
52.0
7.6
7.5
7.3
9.0
39.6
36.4
37.0
34.2
38.0
43.0
36.3
39.3
35.7
54.7
10.0
3.8
12.6
36.7
3.8
a
-Bi2Mo2O9
-Bi2Mo1O6a
BiMoFe0.20
BiMoFe 0.35
BiMoFe 0.50
BiMoFe 0.65
BiMoFe 0.75
BiMoFe 0.85
BiMoFe 1.00
Bi3Mo2FeO12
Fe2(MoO4)3
FeMoO4
36.6
42.3
48.6
62.7
61.0
57.0
45.9
56.1
15.9
6.5
8.5
9.3
12.3
18.1
18.8
26.4
57.6
a
Conversion, selectivity, and yield for bismuth molybdate catalysts were obtained after 6 h on stream.
4. Conclusions
BiMoFex (x = 0–1.00) oxide catalysts were prepared by co-
precipitation and used in the ODH of 1-butene to BD. Their main
solid components confirmed by XRD and Raman spectroscopy was
Bi Mo FeO and Fe (MoO ) . Their conversion of 1-butene and
3
2
12
2
4 3
selectivity to BD depended on their Fe content due to synergy
between the Bi Mo FeO and Fe (MoO ) phases and was maxi-
3
2
12
2
4 3
mized at x = 0.65 in BiMoFex oxide catalysts. The peak temperature
of TPRO profiles was dependent on Fe contents in BiMoFex oxide
catalysts and was minimized at x = 0.65, which showed the greatest
oxygen mobility. The peak position in the low temperature region
of TPRO profiles could be correlated with butene conversion and
BD yield.
Acknowledgments
The authors acknowledge the financial support of the Korea
Energy Management Cooperation (KEMCO). This work is part of
the Energy Technology Innovation Project (ETI) under the Energy
Resources Technology Development Program.
Appendix A. Supplementary data
Supplementary data associated with this article can
be found, in the online version, at http://dx.doi.org/
1
0.1016/j.apcata.2012.04.028.
References
[
[
1] H.H. Kung, Ind. Eng. Chem. Prod. Res. Dev. 25 (1986) 171–178.
2] M.M. Bhasin, J.H. McCain, B.V. Vora, T. Imai, P.R. Pujado, Appl. Catal. A 221 (2001)
397–419.
[
[
3] V.V. Krishnan, S.L. Suib, J. Catal. 184 (1999) 305–315.
4] J.M.L. Nieto, P. Concepcion, A. Dejoz, F. Melo, H. Knozinger, M.I. Vazquez, Catal.
Today 61 (2000) 361–367.
[
[
[
[
[
5] J.A. Toledo-Antonio, N. Nava, M. Martinez, X. Bokhimi, Appl. Catal. A: Gen. 234
(
2002) 137–144.
6] H.W. Lee, J.C. Jung, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Lett. 122 (2008) 281–286.
7] J.C. Jung, H.W. Lee, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Lett. 123 (2008) 239–245.
8] H.W. Lee, J.C. Jung, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Lett. 124 (2008) 364–368.
9] H.W. Lee, J.C. Jung, I.K. Song, Catal. Lett. 133 (2009) 321–327.
[
[
[
10] J.C. Jung, H.W. Lee, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, J. Mol. Catal. A: Gen. 271 (2007) 261–265.
11] J.C. Jung, H.S. Kim, Y.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, I.K. Song, Appl.
Catal. A: Gen. 317 (2007) 244–249.
12] J.C. Jung, H.W. Lee, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Lett. 124 (2008) 262–267.
Fig. 7. Catalytic performance of BiMoFex oxide catalysts as a function of Fe content
in the ODH of 1-butene; (A) conversion of 1-butene and yield in BD and (B) selectivity
to BD and CO2. The data were chosen after 14 h on stream.

J.-H. Park et al. / Applied Catalysis A: General 431–432 (2012) 137–143
143
[
[
13] J.C. Jung, H.W. Lee, D.R. Park, J.G. Seo, I.K. Song, Catal. Lett. 131 (2009) 401–405.
14] A.P.V. Soarew, L.D. Dimitrov, M.C.R.A. Oliveria, L. Hilaire, M.F. Portela, R.K. Gras-
selli, Appl. Catal. A 253 (2003) 191–200.
15] I. Matsuura, R. Schut, K. Kirakawa, J. Catal. 63 (1980) 152–166.
16] J.C. Jung, H.W. Lee, H.S. Kim, D.R. Park, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S.
Kim, I.K. Song, Catal. Commun. 9 (2008) 2059–2062.
17] J.C. Jung, H.W. Lee, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Commun. 9 (2008) 943–949.
18] J.C. Jung, H.W. Lee, H.S. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K.
Song, Catal. Commun. 8 (2007) 625–628.
19] J.C. Jung, H.S. Kim, A.S. Choi, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, I.K. Song, J.
Mol. Catal. A 259 (2006) 166–170.
20] A. Batist, J.F.H. Bouwens, G.C.A. Shuit, J. Catal. 25 (1973) 1–11.
21] J.N. Wu, H.P. Yang, Y.N. Fan, B.L. Xu, Y. Chen, J. Fuel Chem. Technol. 35 (2007)
[22] E. Ruckenstein, R. Krishnan, K.N. Rai, J. Catal. 45 (1976) 270–273.
[23] C. Daniel, G.W. Keulks, J. Catal. 29 (1973) 475–478.
[24] T. Notermann, G.W. Keulks, A. Skliarov, Y. Maximov, L.Y. Margolis, O.V. Krylov,
J. Catal. 39 (1975) 286–293.
[
[
[25] W.J. Linn, A.W. Sleight, J. Catal. 41 (1976) 134–139.
[26] P.L. Villa, A. Szabo, F. Trifiro, M. Carbucicchio, J. Catal. 47 (1977) 122–133.
[27] P. Forzatti, P.L. Villa, N. Ferlazzo, F. Jones, J. Catal. 76 (1982) 188–207.
[28] T.S.R. Prasada Rao, K.R. Krishnamurthy, J. Catal. 95 (1985) 209–219.
[29] W.M. Sears, W.J. Keeler, Appl. Spectrosc. 46 (1992) 1898–1903.
[30] E.V. Hoefs, J.R. Moinner, G.W. Keulks, J. Catal. 57 (1979) 331–337.
[31] I. Wachs, L. Briand, PCT WO99/52630.
[32] L.-T. Weng, B. Delmon, Appl. Catal. A 81 (1992) 141–213.
[33] T.S. Chang, S.G. Lee, C.H. Shin, Y.K. Lee, S.S. Yun, Catal. Lett. 68 (2000) 229–234.
[34] H. Miura, Y. Morikawa, T. Shirasaki, J. Catal. 39 (1975) 22–28.
[35] B.L. Yang, D.S. Cheng, S.B. Lee, Appl. Catal. 70 (1991) 161–173.
[
[
[
[
[
6
84–690.