Oxidative dehydrogenation of n-butene to buta-1,3-diene with novel iron oxide-based catalyst: Effect of iron oxide crystalline structure

Article abstract of DOI:10.1016/j.mcat.2021.111560

We investigate the effect of the crystalline structure of iron oxide catalysts on oxidative dehydrogenation (ODH) of n-butene. ODH reactions of n-butene were carried out with a fixed-bed flow reactor at 450 °C under a but-1-ene (1-C₄H₈) or cis-but-2-ene (cis-2-C₄H₈)(mL/min)/O₂(mL/min) flow ratio of 5/2.5. Of the various iron oxide-based catalysts (α-, β-, γ-, ε-Fe₂O₃, Fe₃O₄, ZnFe₂O₄), ε-Fe₂O₃ showed the highest ODH activity. To the best of our knowledge, ε-Fe₂O₃ has never been used for the ODH of n-butene. Moreover, the catalytic performance of ε-Fe₂O₃ was improved by adding SiO₂, which is related to the maintenance of its structure and improves its redox property. A high BD selectivity of 65 % and BD yield of 18 % were then obtained for 4 h without deactivation. This catalyst can be applied to the ODH of cis-2-C₄H₈ and is proposed as the noble high catalytic performance catalyst for the ODH of n-butene.

Full text of DOI:10.1016/j.mcat.2021.111560

Molecular Catalysis 507 (2021) 111560
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.journals.elsevier.com/molecular-catalysis
Oxidative dehydrogenation of n-butene to buta-1,3-diene with novel iron
oxide-based catalyst: Effect of iron oxide crystalline structure
Takayasu Kiyokawa*, Naoki Ikenaga
Graduate School of Science and Engineering, Kansai University, 3-3-35 Yamate, Suita, Osaka 564-8680, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
We investigate the effect of the crystalline structure of iron oxide catalysts on oxidative dehydrogenation (ODH)
Epsilon iron oxide
Oxidative dehydrogenation
n-Butene
◦
of n-butene. ODH reactions of n-butene were carried out with a fixed-bed flow reactor at 450 C under a but-1-
ene (1-C
Of the various iron oxide-based catalysts (
activity. To the best of our knowledge, -Fe
catalytic performance of -Fe was improved by adding SiO
2
4
H
8
) or cis-but-2-ene (cis-2-C
4
H
8
)(mL/min)/O
-, β-, γ-, -Fe
has never been used for the ODH of n-butene. Moreover, the
, which is related to the maintenance of its
2
(mL/min) flow ratio of 5/2.5.
α
ε
2
O
3
, Fe , ZnFe ), -Fe
3
O
4
2
O
4
ε
2 3
O showed the highest ODH
Buta-1,3-diene
ε
2 3
O
ε
2
O
3
structure and improves its redox property. A high BD selectivity of 65 % and BD yield of 18 % were then obtained
for 4 h without deactivation.
This catalyst can be applied to the ODH of cis-2-C
4
H
8
and is proposed as the noble high catalytic performance
catalyst for the ODH of n-butene.
1
. Introduction
o
C
4
H
8
+ 4O
2
→ 4CO + 4H
2
O
ΔH = -1409.0 kJ/m ol
(3)
One of the principal intermediate products in the petrochemical in-
In order to inhibit complete oxidation, the ODH of n-butene has been
carried out under an O flow with a large amount of steam [1]. However,
dustry is buta-1,3-diene (BD). The demand for BD is increasing, as it is as
a major building block in the industry and can be mostly produced by
the steam cracking of naphtha. However, this process must be operated
2
since the latent heat of steam is large, this requires a lot of energy. The
ODHs of n-butene and propane have also been performed using the
lattice oxygen of a catalyst in order to prevent the complete oxidation of
the substrates (Eqs. (4) and (5)) [2,3]. However, this approach has not
been suitable for the continuous BD production over a long duration
because it requires the regeneration of the catalyst used.
◦
at temperatures over 700 C, and the endothermic reaction thus requires
a lot of energy. In addition, the steam cracking of naphtha shows the low
BD selectively because light olefins such as ethene and propylene were
produced. Due to the shale gas revolution, ethane can be cheaply sup-
plied; the substrate used in cracking to produce ethylene transfers from
naphtha to ethane derived from the shale gas. Therefore, in terms of a
near-future supply of BD, an alternative production process is needed.
Recently, the oxidative dehydrogenation (ODH) of C4 fraction (Eq.
C
4
H
8
+ MO
x
→ C
4
H
6
+ H
2
O + MOx-1
(4)
(5)
MOx-1 + 1/2O → MO
2
x
For effective and continuous BD production, it would be preferable to
(
1)), which is an exothermic reaction, has attracted attention. This re-
2
conduct the ODH of n-butane and n-butene under O flow without
action is very useful from the viewpoint of energy conservation, and it
has been proposed as a process for the efficient production of BD.
steam.
Various effective catalysts such as Bi-Mo composite oxide [4–13],
ferrite type [14–27], and V-containing [28–30] have been proposed for
the ODH of n-butene and n-butane. Among these catalysts, the Bi-Mo
composite oxide and ferrite type catalysts have been widely studied. It
has been reported that the ODH reaction with these catalysts proceeds
o
C
4
H
8
+ 1/2O
2
→C
4
H
6
+ H
2
O
ΔH = -132.1 kJ/m ol
(1)
However, ODH has a serious problem in that the complete oxidation
of reactant and product to CO
2
proceeds easily (Eqs. (2) and (3)).
o
C
4
H
8
+ 6O → 4CO + 4H
2
2
2
O
ΔH = -2540.8 kJ/m ol
(2)
*
Corresponding author. Present address: Chemical Research Group, Research Institute of Innovative Technology for the Earth (RITE), 9-2 Kizugawadai, Kizugawa,
Kyoto 619-0292, Japan.
E-mail address: t.kiyokawa@rite.or.jp (T. Kiyokawa).
https://doi.org/10.1016/j.mcat.2021.111560
Received 30 September 2020; Received in revised form 19 March 2021; Accepted 30 March 2021
Available online 12 April 2021
2
468-8231/© 2021 Elsevier B.V. All rights reserved.

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
through the Mars-van Krevelen mechanism [4,5]. This mechanism is a
redox cycle with the lattice oxygen in the metal oxide, and the reactivity
of lattice oxygen in the metal oxide during the ODH reaction is a crucial
factor in determining the catalytic performance.
water, n-octane, and butan-1-ol. CTAB at H
was added to the mixed solution containing Fe(NO
until dissolved (Solution 1). Next, the solution was changed from Fe
(NO O to 28 % NH3.aq, prepared with the same method as So-
2
O/CTAB = 30/1 (mol/mol)
3
)
3
⋅9H O, and stirred
2
3
)
3
⋅9H
2
On the other hand, the effect of the crystalline structure of the
catalyst on the ODH has often been discussed [7,17,19]. The Bi-Mo
lution 1 (Solution 2). Solution 2 was added drop-wise to Solution 1, and
the slurry obtained was stirred for 30 min. TEOS was added to the slurry
complex oxide has
α
-Bi
2
Mo
3
O
12, β-Bi
2
Mo
2
O
9
, and γ-Bi
2
MoO
6
. Gener-
to yield a weight ratio of (Fe
2
O
3
)/(SiO
2
+Fe
2
O
3
) = 0.1, 0.2, and 0.5, and
ally, the catalytic activity for the ODH of n-butene is on the order of
stirred overnight. After stirring, slurries were separated by centrifuga-
tion. Obtained solids were washed with a mixed solvent of acetone and
γ>
α
[7]. β-Bi
2
Mo
2
O
9
is therefore inappropriate for the ODH, because it
◦
◦
decomposes to
α
and γ at the reaction temperature of 400–550 C [6]. It
methanol (volume ratio = 1/1), and then dried at 60 C in an oven for
◦
is suggested that this order is consistent with lattice oxygen reactivity
31]. In the case of an iron oxide catalyst, it is reported that ferrite type
catalysts with a spinel structure can exhibit high ODH activity, and the
presence of -Fe leads to low ODH performance [17,19].
5 h. The solid was calcined at 1050 C for 4 h in air. The different
[
Fe
(NO
content (wt%). Pure
in -Fe (50)-SiO
2 3
O -containing catalysts were prepared by varying the amount of Fe
3
)
3
⋅9H
2
O. Here, the notation is
ε-Fe
2
O
3
(X)-SiO
2
, where X is Fe
2
O
3
2
α
2
O
3
ε
-Fe
2
O
3
was obtained by the following process. SiO
◦
Thus, extensive study of the effect of the crystallite structure on the
ODH is important to the development of a catalyst with high catalytic
ε
2
O
3
2
was dissolved in 1 mol/L NaOH solution at 60
C
overnight, and the solid was filtered, washed with a large amount of
◦
performance. Iron oxide has various crystalline structures, including
α
-,
pure water, and dried at 110 C overnight in an oven.
β-, γ-, and
ε
-Fe , and each the single phase can be easily synthesized.
2
O
3
ZnFe
2
O
4
was prepared as previously reported [19]. Briefly, Fe
O and Zn(NO O were dissolved in 100 mL of pure
⋅6H
Meanwhile, detailed investigation of the crystalline phase of iron oxide
in various ODH reactions has not yet been conducted.
(NO
3
)
3
⋅9H
2
3
)
2
2
water. The mixed solution was stirred at room temperature for 1 h. After
stirring, 3 mol/L of NaOH solution was added to the solution under
vigorous stirring to reach pH 9. After stirring at room temperature for
12 h, it was aged overnight. The resultant precipitate was separated by
centrifugation, the solid was washed with a large amount of pure water
Therefore, in this study we examined the effect of the crystalline
structure of iron oxide on the ODH of n-butene. We were able to propose
an effective iron oxide catalyst for the ODH reaction. In order to achieve
efficient BD production, the ODH of butenes was carried out under O
2
◦
flow without steam, and catalytic performance stability during the ODH
was also investigated.
until pH was 7–8, and then it was dried at 175 C for 16 h. The solid was
◦
calcined at 650 C for 2 h in air.
2
. Experimental
2.3. Catalyst characterization
2
.1. Materials
X-ray diffraction (XRD) patterns of iron oxide catalysts were ob-
tained by the powder method using a Rigaku RINT-TTRIII diffractom-
Fe(NO
3
)
3
⋅9H
3
2
O (assay = min. 99.0 %), FeSO
4
⋅7H
⋅6H
solution, n-octane
2
O (assay = 99–102
eter with monochromatic CuK
α
radiation under the following
%
), Fe (SO
2
4
)
⋅nH
2
O (assay = 60–80 %), Zn(NO
3
)
2
2
O (assay = min.
conditions: tube voltage 40 kV, tube current 30 mA, scan step 0.02 de-
9
9.0 %), NaOH (assay = min. 97.0 %), 28 %NH
3
grees, scan region 10–80 degrees, and scan speed 4.0 degrees/min.
(
(
(
assay = min. 98 %), butan-1-ol (assay = min. 99 %), acetone
assay = min. 99.5 %), methanol (assay = min. 99.8 %), NaCl
assay = 99.5 %), cetyltrimethylammonium bromide (CTAB), tetraethyl
2.4. Catalyst test
orthosilicate (TEOS) (assay = min. 95.0 %), and Fe
3
O
4
were purchased
2.4.1. ODH of n-butene with iron oxide catalyst under O
2
atmosphere
from Wako Pure Chemical Industries, and used for the preparation of
iron oxide catalysts. But-1-ene (1-C
) (assay = min. 99.0 %) was
supplied by Sumitomo Seika Chemical. Cis-but-2-ene (cis-2-C
assay = min. 99.0 %) was supplied by Takachiho Chemical Industrial
Co., Ltd.
The ODH of 1-C
4
H
8
was carried out using a fixed-bed flow quartz
◦
4
H
8
reactor at 450 C under atmospheric pressure. After 200 mg of the
4
H
8
)
catalyst was placed in the reactor, it was preheated to a reaction tem-
◦
(
perature of 450 C under 22.5 mL/min (STP) of Ar flow. Oxygen, 1-C
4
H
8
or cis-2-C
-C
/Ar = 2.5/5/22.5 mL/min (STP). Total gas flow rate was fixed at
30 mL/min.
The C4 fractions (1-C
analyzed using a flame ionization detector (FID) gas chromatograph
(Shimadzu GC14B, column: Unicarbon A-400). CH , CO, and CO were
also analyzed with a thermal conductivity detector (TCD) gas chro-
matograph (column: activated carbon). O and H were analyzed by
4 8 2 4 8
H , and Ar were introduced at a flow rate of O /1-C H or cis-
2
4 8
H
2
.2. Preparation of various iron oxide catalysts
4 8 4 8 4 8 4 6
H , cis-2-C H , trans-2-C H , and C H ) were
α
-Fe
2
O
3
was prepared by the precipitation method. Fe(NO
3
)
3
⋅9H
2
O
(
10 mmol) was dissolved in 100 mL of pure water. The solution was
4
2
stirred at room temperature for 1 h. After stirring, 1 mol/L of NaOH
solution was added to the solution under vigorous stirring until the pH
reached 10. After stirring for 1 h, the resultant precipitate was separated
by centrifugation, and the solid was washed with a large amount of pure
2
2
TCD gas chromatograph (Shimadzu GC8A, column: molecular sieves
5A).
◦
water until the pH reached 7ꢀ 8. It was then dried at 110 C overnight in
◦
an oven. The solid was calcined at 500 C for 2 h in air. γ-Fe
2
O
3
was
2.4.2. ODH of but-1-ene with lattice oxygen in iron oxide catalyst
prepared by the same method. FeSO
4
⋅7H
2
O (10 mmol) was used as a
The ODH of 1-C
4
H
8
was carried out using the fixed-bed flow quartz
◦
◦
precursor, and the calcination was carried out at 400 C.
reactor at 450 C under atmospheric pressure. After 200 mg of the
β-Fe
(SO
flow. Fe
2
O
3
was prepared as described in previous reports [32,33]. First,
catalyst was placed in the reactor, it was preheated to a reaction tem-
◦
◦
Fe
2
4
)
3
was prepared by calcining Fe
(SO
2
(SO
4
)
3
⋅nH
2
O at 400 C under N
2
perature of 450 C under 25 mL/min (STP) of Ar flow. Then, 5 mL/min
)
3
and NaCl at a molar ratio of 1:2 were mixed in a mortar
4 8
of 1-C H and 25 mL/min of Ar were introduced for 5 min. Reoxidation
2
4
◦
for 30 min. The mixed material was calcined at 550 C for 2 h in air. The
calcined product was added to pure water, and stirred at room tem-
perature overnight. After stirring, the solid was filtered, washed with a
was carried out after the reaction under 5 mL/min of O
2
and 25 mL/min
◦
of Ar for 10 min at 450 C. Quantification of the products was carried out
using the same equipment as in Section 2.4.1.
◦
large amount of pure water, and dried at 110 C overnight in an oven.
ε
-Fe
34–36]. That is,
method. Fe(NO
2
O
3
-SiO
2
was first prepared with reference to previous reports
-Fe -SiO was prepared by the reverse micelle
⋅9H O was dissolved in a mixed solution of pure
[
ε
2
O
3
2
3
)
3
2
2

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
Fig. 3. Effect of crystal phase of iron oxide on ODH of but-1-ene with lattice
oxygen.
Catalyst: 200 mg.
ODH: 30 mL/min (1-C4H8 /Ar = 5/25 mL/min).
◦
ODH temperature: 450 C, ODH time: 5 min.
Table 1
Specific sureface area of various iron oxide catalysts.
Catalyst
Specific surface are
2
m /g
α
-Fe
β-Fe
γ-Fe
-Fe
2
O
3
3
3
38
2
O
O
n.d.
129
67
2
ε
2 3
O
Fig. 1. XRD patterns of various iron oxides.
more than 10 %. In the case of
-C
other iron oxides; the highest BD yield of 17.1 % was also obtained.
ε
-Fe
2
O
3
, this catalyst showed the highest
1
4 8 2
H conversion of 43.5 % and lower CO selectivity of 16.6 % than
Moreover,
ε
-Fe
that were already reported here (Fig. S1). Among various
-Fe exhibited the highest ODH activity.
To the best of our knowledge, this is the first successful result
showing the high activity of -Fe for the ODHs of various substrates,
2 3 3 4
O showed higher catalytic activity compared to Fe O
and ZnFe
2
O
4
iron oxide catalysts,
ε
2 3
O
ε
2 3
O
such as propane and but-1-ene. In order to understand the effect of the
crystalline structure of iron oxide on the ODH in the absence of molec-
ular O
presents our results.
and low BD selectivity of 15.6 %. While γ-Fe
version of 24.4 %, this catalyst exhibited high CO
Therefore, γ-Fe progressed the complete oxidation. Although
β-Fe indicated low conversion of 13.8 %, it produced the highest BD
selectivity of 56.9 %. Surprisingly, -Fe showed the highest conver-
sion of 46.4 %, and the high BD selectivity of 49.9 %.
The order of BD yield was thus: -Fe
(23.2 %) > β-Fe
γ-Fe (5.5 %) > -Fe (1.6 %). These results clarified that the lattice
oxygen in -Fe was effectively used for the ODH of 1-C . The
unique performance of -Fe appeared to be related to the ODH of 1-
in the presence of O . When the specific surface area of various iron
oxide catalysts was measured (Table 1), the order of specific surface area
was following: γ-Fe -Fe -Fe . Generally, the specific sur-
face area depends on the particle size or the controlled pore structure.
The order of particle size may be -Fe -Fe because the
>γ-Fe
pore structure is not controlled. On the other hand, the order of ODH
activity was -Fe -Fe . Therefore, it is considered that
>γ-Fe
2
, we carried out ODH with lattice oxygen in iron oxide. Fig. 3
α-Fe
2
O
3
showed a low 1-C
4
H
8
conversion of 9.9 %
2 3
O showed a high con-
2
selectivity of 22.1 %.
2 3
O
Fig. 2. Effect of crystal phase of iron oxide on ODH of but-1-ene under O2 flow.
Catalyst: 200 mg.
2 3
O
ε
2 3
O
Flow rate: 1-C4H8/O2/Ar = 5/2.5/22.5 mL/min.
◦
Reaction temperature: 450 C.
Reaction time: 30 min.
ε
2
O
3
2 3
O (7.9 %) >
2
O
3
α
3
2 3
O
ε
2
O
4 8
H
3
. Results and discussion
ε
2 3
O
C
4
H
8
2
3
.1. Effect of crystal phase of iron oxide on the ODH
O
3
>ε
2
O
3
>α
2 3
O
Fig. 1 presents the XRD patterns of various iron oxide catalysts.
2
α
-Fe
2
O
3
and γ-Fe
2
O
3
showed
α
-Fe
2
O
3
and γ-Fe
-Fe
2
O
3
diffraction peaks,
and β-Fe in the
α
2
O
3
>ε
2
O
3
2 3
O
respectively (Fig. 1(a), (d)). Formation of
ε
2
O
3
2 3
O
single phase could be confirmed with diffraction peaks matching those
reported in the literature (Fig. 1(b), (c)) [33,36,37].
ε
2
O
3
2
O
3
>α
2 3
O
the specific surface area of the catalyst is not related to ODH activity.
This fact indicated that the ODH activity strongly depends on the crys-
tallite structure of iron oxide rather than the specific surface area and
particle size.
Fig. 2 shows the effect of the crystallite structure of various iron
oxides on the ODH of 1-C
showed a high O conversion of more than 95 % (data not shown in
Fig. 2), we give the different catalytic performances in each iron oxide
catalyst. The -Fe catalyst mainly progressed the complete oxidation,
because a high CO selectivity of 47.6 % and the lowest BD yield of 2.8 %
were given. When γ- and β-Fe were used for the ODH, these catalysts
exhibited a high 1-C conversion of more than 20 %, and BD yields of
4 8
H . Although all the iron oxide catalysts
2
On the other hand, it is reported that the
catalyst showed the high ODH performance. According to the research,
n-butene was interacted with ZnFe and ZnAl which have the
spinel structure, and the activated butene reacted with the lattice
2 3 2 4
α-Fe O /ZnFe O biphasic
α
2 3
O
2
2
O
4
2 4
O
2 3
O
4 8
H
3

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
Fig. 4. Stability test of -Fe2O3 catalyst during ODH reaction.
ε
(
(
a) ODH under O2 flow.
b) XRD patterns of fresh and used catalyst.
Fig. 5. Evaluation of redox property for
ε
-Fe2O3.
a) Repeated ODH of but-1-ene and re-oxidation
with -Fe2O3.
b) XRD patterns of
(
ε
(
ε
-Fe2O3 after the ODH with
lattice oxygen and re-oxidation with O2.
Catalyst: 200 mg.
ODH: 30 mL/min (1-C4H8 /Ar = 5/25 mL/
min).
◦
ODH temperature: 450 C, ODH time: 5 min.
Re-oxidation: O2/Ar = 5/25 mL/min.
◦
Re-oxidation temperature: 450 C, Re-oxidation
time: 10 min.
oxygen in
cated that iron-based catalyst combined corundum structure and spinel
structure is more effective for the ODH of n-butene. When -Fe was
focused, -Fe has the crystalline structure that combine corundum
structure ( -Fe ). Therefore, it is
) and spinel structure (γ-Fe
catalyst could show the particularly high ODH
α
-Fe
2
O
3
to produce BD [38]. In other words, this result indi-
On the other hand, Wan et al. reported that catalytic activity for the
ODH of n-butene is related to the oxidation and reduction cycle (redox
cycle) through the lattice oxygen in the metal oxide [11]. Hence, the
ε
2 3
O
ε
2
O
3
ODH with the lattice oxygen in the catalyst and regeneration with O
were carried out using -Fe . The crystallite phase of -Fe after the
ODH with lattice oxygen and regeneration with O were then also
2
α
2
O
3
2
O
3
ε
2
O
3
ε
2 3
O
considered that
activity.
ε
-Fe
2
O
3
2
analyzed by XRD. The results are shown in Fig. 5. In the first ODH re-
action (Fig. 5(a)), a high BD yield of 23 % was obtained. Although the
1
-C
4 8
H conversion (53.7 %) of the second reaction was higher than that
3
.2. Stability of
ε
-Fe
2
O
3
catalyst for ODH
(
46.4 %) of the first reaction, low BD selectivity of 19.2 % and BD yield
of 10.3 % were observed.
Catalytic performance stability for the ODH reaction is an important
factor for the efficient production of BD. Therefore, we conducted a
stability test of -Fe . Fig. 4 shows the result of ODH and XRD analysis
before and after the reaction. Although -Fe showed the high BD
yield of 17.1 % for 30 min, after 60 min, the 1-C conversion
decreased from 34.3 to 27.8 %, while CO selectivity increased from
0.5 to 25.5%. Unfortunately, the BD yield decreased from 17.1 % to
These results indicated that the lattice oxygen in
ε
-Fe
2
O
3
could be
. Thus, BD
yield decreased in the second ODH. According to XRD analysis of
-Fe after the ODH using the lattice oxygen and the regeneration
using O (Fig. 5(b)), diffraction peaks attributable to Fe were shown
in the catalyst (Fig. 5(b-2)), and those of γ-Fe in the catalyst after
regeneration with O (Fig. 5(b-3)). This result clearly indicates that
-Fe was reduced to Fe , and could not be regenerated to -Fe
by molecular O
This suggested that keeping the redox property of the lattice oxygen
of the -Fe phase is very important to maintaining ODH activity
under O flow.
used to produce BD, but could not be regenerated with O
2
ε
2 3
O
ε
2 3
O
ε
2 3
O
4 8
H
2
3 4
O
2
2 3
O
2
1
2
3.7 % (Fig. 4(a)).
ε
2
O
3
3
O
4
ε
2 3
O
To examine the reason for the decrease in the ODH activity, XRD
2
.
analyses of the catalyst before and after the ODH were conducted. As
shown in Fig. 4(b), -Fe diffraction peaks observed before the ODH
disappeared after the reaction, and Fe diffraction peaks appeared
after the reaction for 2 h. In the ODH of 1-C , it has been reported that
ε
2 3
O
ε
2
2 3
O
3 4
O
4 8
H
the ODH activity of the iron oxide-based catalyst declined due to the
formation of divalent iron oxide species during the reaction [26,27].
3.3. Effect of addition of SiO
2
to the catalyst on the ODH
-Fe catalyst showed the highest ODH
Actually, the Fe
3 4
O catalyst showed low ODH activity (Fig. S1). There-
fore, it is considered that the catalytic performance was decreased due to
As mentioned above, the
ε
2 3
O
it changing from
the ODH.
ε
-Fe
2
O
3
to a reduced iron oxide such as Fe
3
O
4
during
activity among the iron oxide catalysts and was proposed as a novel
catalyst. However, it was deactivated due to changes in the crystalline
4

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
Fig. 6. Stability test of -Fe2O3(50)-SiO2 catalyst during ODH reaction.
ε
(
(
a) ODH under O2 flow.
b) XRD patterns of fresh and used catalyst.
Fig. 7. Evaluation of Redox property for
ε
-Fe2O3 containing SiO2.
a) Repeated ODH of but-1-ene and re-oxidation
with -Fe2O3(50)-SiO2.
b) XRD patterns of -Fe2O3(50)-SiO2 after the
(
ε
(
ε
ODH with lattice oxygen and re-oxidation with
O2.
Catalyst: 200 mg.
ODH: 30 mL/min (1-C4H8 /Ar = 5/25 mL/
min).
◦
ODH temperature: 450 C, ODH time: 5 min.
Re-oxidation: O2/Ar = 5/25 mL/min.
◦
Re-oxidation temperature: 450 C, Re-oxidation
time: 10 min.
structure during the reaction and failure of the redox cycle. Therefore,
we attempted to improve catalytic performance stability.
catalyst was not related to the catalytic activity and the stability of
catalyst. When XRD analyses of the
after the ODH were conducted, -Fe
in the catalyst even after the ODH (Fig. 6(b)). Moreover, from the result
of FT-IR, although all -Fe phase in pure -Fe catalyst was con-
verted to Fe phase after reaction, -Fe (50)-SiO catalyst con-
tained only phase without the formation of Fe (Fig. S2).
Thus, this result is indicated that the amorphous iron oxide species other
than -Fe phase, which cannot confirm with XRD analysis, did not
exist in the used -Fe (50)-SiO
Unlike pure -Fe , we found that the stability of the
could be improved by SiO , and that maintenance of the
led to increased ODH activity and stability during the reaction.
We also examined the redox property of the lattice oxygen in the
-Fe (50)-SiO catalyst through repeated ODH. The results are shown
in Fig. 7. The low 1-C conversion of 22.4 %, high BD selectivity of
ε
-Fe
2
O
3
(50)-SiO
2
catalyst before and
Studies have proposed Fe
catalytic performance for high-temperature sulfuric acid decomposition
37,39]. According to the literature, the redox property of metal oxide is
an important factor in determining the reaction rate, and it has been
reported that Fe /SiO showed the better redox property among
various iron oxide-based catalysts. It was also reported that Fe
containing Al , CaO, and SiO showed high redox cycle stability for
the cyclic water gas shift process. Thus, SiO seems to be related to in-
hibition of iron species sintering by physical blocking [40]. We therefore
2
O
3
/SiO
2
as a good catalyst with high
ε
2 3
O diffraction peaks were observed
[
ε
2
O
3
ε
2 3
O
3
O
4
ε
2
O
3
2
2
O
3
2
ε
-Fe
2
O
3
3 4
O
2 3
O
2
O
3
2
ε
2 3
O
2
ε
2
O
3
2
.
ε
2
O
3
ε
-Fe
2
O
3
phase
phase
expected that the redox property could be improved by including SiO
the iron oxide.
2
in
2
ε
-Fe
2
O
3
On the other hand,
γ-Fe ) coated with SiO
catalytic activity during the ODH, we investigated the effect of
containing SiO on the ODH of 1-C under O flow. -Fe
)-containing SiO catalyst -Fe (50)-SiO was used for the
reaction.
Fig. 6(a) shows the results of ODH with an
Although the -Fe (50)-SiO catalyst showed a high 1-C
ε
2 3
-Fe O can be formed by calcining iron oxide
◦
(
2
O
3
2
at 1050 C [36]. Hence, to maintain the
-Fe
(50 wt
ε
2
O
3
2
ε
3
2
O
3
4 8
H
2
4
H
8
2
ε
2
O
49.2 %, and low BD yield of 11 % were obtained in the first ODH re-
action (Fig. 7(a)). The result showed lower lattice oxygen reactivity than
%
2
(
ε
2
O
3
2
)
that of pure
could be used for the ODH was probably half that of pure
Meanwhile, when the ODH with the lattice oxygen in the reoxidized
-Fe (50)-SiO was carried out, the 1-C conversion of 23.3 %, BD
selectivity of 45.7 %, and BD yield of 10.7 % were similar to the results
in the first ODH. This indicated that the used lattice oxygen in -Fe
containing SiO could be easily regenerated by molecular O
XRD analysis (Fig. 7(b)) revealed that -Fe diffraction peaks were
not evident in -Fe (50)-SiO after the reaction with the lattice oxy-
gen, and the small diffraction peaks related to Fe were also observed
(Fig. 7(b-2)). Surprisingly, after reoxidation with molecular O , the
catalyst showed -Fe diffraction peaks (Fig. 7(b-3)). Therefore, after
ε-Fe
2
O
3
. This is because the amount of lattice oxygen that
ε
-Fe
2
O
3
(50)-SiO
2
catalyst.
ε
-Fe
2 3
O .
ε
2
O
3
2
H conver-
4 8
sion of 22.5 %, and a BD yield of 13.0 % was obtained in the reaction for
0 min, after 60 min, this catalyst showed a higher BD yield (ca.18 %)
than that of pure -Fe . In addition, no deactivation was observed
during the reaction for 4 h. Therefore, the -Fe containing SiO
catalyst had a higher catalytic performance than pure -Fe
In order to understand the effect of adding SiO on the ODH reaction,
adsorption, XRD and FT-IR analyses were conducted. From N
adsorption, the both -Fe and -Fe -SiO catalysts showed almost
the same value (Table S1). This result indicated that pore property in
ε
2
O
3
2
4 8
H
3
ε
2
O
3
ε
2 3
O
ε
2
O
3
2
2
2
.
ε
2
O
3
.
ε
2 3
O
2
ε
2
O
3
2
N
2
2
3 4
O
ε
2
O
3
ε
2
O
3
2
2
ε
2 3
O
5

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
3
.4. Effect of
ε
-Fe
2
O
3
content on the ODH
-Fe in SiO
The usefulness of
Therefore, we investigated the effect of
tested -Fe (10, 20, 50)-SiO and pure
analyses were carried out, and the results are shown in Fig. 8. The
-Fe diffraction peaks appeared in all the catalysts, and amorphous
SiO peaks were observed in -Fe (10, 20, 50)-SiO
Fig. 9 shows the results of the ODH using -Fe
As mentioned in Sections 3.2 and 3.3, the pure -Fe
high 1-C conversion of 34.3 % and high BD yield of 17.1 % for
0 min. However, after 30 min, a decrease in BD selectivity and increase
ε
2
O
3
2
for the ODH was suggested.
ε
-Fe
2
O
3
content on the ODH. We
ε
2
O
3
2
ε
-Fe for the ODH. XRD
2 3
O
ε
2 3
O
2
ε
2
O
3
2
.
ε
2
O
3
(10, 20, 50)-SiO
2
.
ε
2 3
O catalyst gave the
4 8
H
3
in CO
2
selectivity were observed, and the BD yield declined to 13.7 %.
-Fe (50)-SiO catalyst showed lower 1-
conversions and lower BD yield than pure -Fe until 30 min,
after 60 min, the BD yield increased to about 18.0 % over 4 h. In the
ODH with the -Fe (20)-SiO catalyst, the CO selectivity of about 14
was lower than that of -Fe (50)-SiO . In addition, the high BD
yield of ca.18 % was obtained, and the ODH activity of the -Fe (20)-
SiO catalyst was maintained during the reaction for 4 h, similar to the
-Fe (50)-SiO catalyst. The -Fe (10)-SiO catalyst showed the
lowest CO selectivity of 11 % in this study. Low O conversion of about
0 % (data not shown in Fig. 9) and 1-C conversion of 22 % were
exhibited, producing a lower BD yield of ca.14 % compared to the other
-Fe containing SiO catalysts. However, the catalytic performance
was the most stable for the -Fe (10)-SiO catalyst.
As shown in these results, all -Fe containing SiO
stability compared to pure -Fe . Although the BD yield improved
with increasing -Fe content, the BD yield plateaued over 20 wt% of
-Fe . In addition, the CO selectivity did not increase in the reaction
with the catalyst containing -Fe over 20 wt%. Therefore, we
determined that 20 wt% of -Fe was the best content percentage.
Finally, cis- or trans-but-2-ene is also important as a raw material for
producing BD by the ODH. When the ODH of cis-2-C with
-Fe (20)-SiO catalyst was carried out (Fig. 10), high ODH activity
was obtained similar to the case of ODH of 1-C . In addition, the ODH
activity was again maintained for 4 h. These results indicated that cis-2-
Meanwhile, although the
ε
2
O
3
2
C
4
H
8
ε
2 3
O
ε
2
O
3
2
2
%
ε
2
O
3
2
ε
2 3
O
2
ε
2
O
3
2
ε
2
O
3
2
2
2
6
4 8
H
ε
2
O
3
2
ε
2
O
3
2
ε
2
O
3
2
exhibited high
Fig. 8. XRD patterns of -Fe2O3 catalysts containing SiO2.
ε
ε
2 3
O
ε
2 3
O
the ODH with the lattice oxygen,
to the -Fe phase by O . Results of the XRD analyses and repeated
ODH showed that SiO improved the redox property of the lattice oxy-
gen in -Fe compared to pure -Fe
As above the results, the fact, that the other iron oxide species such as
in -Fe (50)-SiO after the ODH under O flow were not seen,
2 3 2
ε-Fe O (50)-SiO could be restructured
ε
2
O
3
2
ε
2
O
3
2
ε
2 3
O
2
ε
2 3
O
ε
2
O
3
ε
2 3
O .
4 8
H
Fe
3
O
4
ε
2
O
3
2
2
ε
2
O
3
2
seems to be related to improvement of the redox property. In addition,
improvement of redox property led to the high catalytic activity and
4 8
H
2
stability during the ODH under O flow.
Fig. 9. Effect of -Fe2O3 containing level on ODH of but-1-ene under O2 flow.
ε
6

T. Kiyokawa and N. Ikenaga
Molecular Catalysis 507 (2021) 111560
online version, at doi:https://doi.org/10.1016/j.mcat.2021.111560.
References
[
[
1] J.H. Park, C.H. Snin, Appl. Catal. A Gen. 495 (2015) 1–7.
2] K. Fukudome, N. Ikenaga, T. Miyakke, T. Suzuki, Catal. Sci. Technol. 1 (2011)
9
87–998.
[
3] S. Gong, S. Park, W.C. Choi, H. Seo, N.Y. Kang, M. Wan, Han, Y.K. Park, J. Mol.
Catal. A: Chem. 391 (2014) 19–24.
[
[
4] M.F. Portela, Top. Catal. 15 (2001) 241–245.
5] J.C. Jung, H. Lee, H. 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.
[
6] J.C. Jung, H. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.K. Song,
J. Mol. Catal. A Chem. 264 (2007) 237–240.
[
[
7] J.C. Jung, H. Lee, I.K. Song, Catal. Surv. Asia 13 (2009) 78–93.
8] J.C. Jung, H. Lee, H. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.
K. Song, J. Mol. Catal. A Chem. 271 (2007) 261–265.
[
9] J.H. Park, H. Noh, J.W. Park, K. Row, K.D. Jung, C.H. Shin, Appl. Catal. A: Gen.
4
31–432 (2012) 137–143.
[
[
10] J.C. Jung, H. 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.
11] C. Wan, D. Cheng, F. Chen, X. Zhan, Chem. Eng. Sci. 135 (2015) 553–558.
Fig. 10. ODH of cis-but-2-ene under O2 flow with
ε
-Fe2O3(20)-SiO2.
[12] C. Wan, D. Cheng, F. Chen, X. Zhan, J. Chem. Technol. Biotechnol. 91 (2016)
353–358.
Catalyst: 200 mg, Flow rate: 30 mL/min (cis-2-C4H8 /O2/Ar = 5/2.5/22.5).
◦
[13] H. Armendariz, G. Aguilar-Rios, P. Salas, M.A. Valenzuela, I. Schifter, H. Arriola,
N. Nava, Appl. Catal. A: Gen. 92 (1992) 29–38.
Reaction temperature: 450 C.
[
[
[
14] J.A. Toledo, N. Nava, M. Martinez, X. Bokhimi, Appl. Catal. A: Gen. 234 (2002)
137–144.
C
4
H
8
can also be used as a new material for BD with the
ε
-Fe
2 3
O (20)-
15] H. Lee, J.C. Jung, H. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.
K. Song, Catal. Lett. 131 (2009) 344–349.
SiO
2
catalyst. We propose the -Fe catalyst containing SiO
ε
2
O
3
2
as a novel
iron oxide catalyst that shows high catalytic performance for the ODH of
n-butene.
16] Y.M. Chung, Y.T. Kwon, T.J. Kim, S.J. Lee, S.H. Oh, Catal. Lett. 130 (2009)
417–423.
[
[
17] H. Lee, J.C. Jung, I.K. Song, Catal. Lett. 133 (2009) 321–327.
18] H. Lee, J.C. Jung, H. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.
K. Song, Korean J. Chem. Eng. 26 (2009) 994–998.
4
. Conclusions
[
[
19] H. Lee, J.C. Jung, H. Kim, Y.M. Chung, T.J. Kim, S.J. Lee, S.H. Oh, Y.S. Kim, I.
K. Song, Catal. Commun. 9 (2008) 1137–1142.
In this study, we investigated the effect of the crystalline structure of
20] J.A. Toledo, P. Bosch, M.A. Valenzuela, A. Montoya, N. Nava, J. Mol. Catal. 125
iron oxide catalyst on the ODH of 1-C
4
H
8
ε
2 3
. The -Fe O catalyst showed
(
1997) 53–62.
the highest ODH activity (BD yield of 17.1 %) among various iron oxide
catalysts, and we therefore present it as a novel and excellent catalyst.
[21] H. Armendariz, J.A. Toledo, G. Aguilar-Rios, M.A. Valenzuela, P. Salas, A. Cabral,
H. Jimenez, I. Schifter, J. Mol. Catal. 92 (1994) 325–332.
[
[
22] R.J. Rennard, W.L. Kehl, J. Catal. 21 (1971) 282–293.
Moreover, when
this reaction, the catalyst showed high BD yield, and deactivation was
not seen over 4 h. We found that SiO improved the maintenance of the
crystalline structure and the redox property of lattice oxygen in -Fe
The optimum -Fe content was 20 wt%, because a high BD yield of 18
for 4 h was obtained. This catalyst could also be used for the ODH of
cis-2-C with the same BD yield (17 %) and high stability as for 1-
. This suggested that cis- and trans-2-C could be recycled, which
ε
-Fe
2
O
3
containing SiO
2
(
ε
2 3 2
-Fe O -SiO ) was used for
23] B.J. Liaw, D.S. Cheng, B.L. Yang, J. Catal. 118 (1989) 312–326.
[24] W.Q. Xu, Y.G. Yin, G.Y. Li, S. Chen, Appl. Catal. A: Gen. 89 (1992) 131–142.
[
[
25] F.Y. Qiu, L.T. Weng, E. Sham, P. Ruiz, B. Delmon, Appl. Catal. 51 (1989) 235–253.
2
26] Y.M. Chung, Y.T. Kwon, T.J. Kim, S.J. Lee, S.H. Oh, Catal. Lett. 131 (2009)
ε
2 3
O .
5
79–586.
ε
2 3
O
[27] M.A. Gibson, J.W. Hightower, J. Catal. 41 (1976) 431–439.
ˆ ˆ ˆ
[28] A. Dejoz, J.M. LoApez Nieto, F. MaArquez, M.I. VaAzquez, Appl. Catal. A: Gen. 180
%
(
1999) 83–94.
4 8
H
[
29] J.K. Lee, H. Lee, U.G. Hong, J. Lee, Y.-J. Cho, Y. Yoo, H.S. Jang, I.K. Song, J. Ind.
C
H
4 8
4 8
H
Eng. Chem. 18 (2012) 1096–1101.
means the substantial selectivity of BD is higher than those indicated in
the Tables and Figures in this report.
[30] J.M. Lopez Nieto, P. Concepci, A. Dejoz, H. Knozinger, F. Melo, M.I. Vazquez,
J. Catal. 189 (2000) 147–157.
[
[
31] E. Ruckenstein, R. Krishnan, K.N. Rai, J. Catal. 45 (1976) 270–273.
32] R. Zboril, M. Mashlan, D. Krausova, Mossbauer Spectrosc. Mater. Sci. (1992)
49–56.
CRediT authorship contribution statement
[
[
[
[
[
[
33] Y. Ikeda, M. Takano, Y. Bando, Bull. Inst. Chem. Res., Kyoto Univ. 64 (1986)
2
49–258.
Takayasu Kiyokawa: Conceptualization, Validation, Investigation,
Writing - original draft, Visualization. Naoki Ikenaga: Resources, Su-
pervision, Writing - review & editing.
34] S. Ohkoshi, S. Sakurai, S. Sasaki, K. Sato, J. Shimoyama, Japan patent,
JP5124825B2.
35] S. Ohkoshi, S. Kuroki, S. Sakurai, K. Matumoto, K. Sato, S. Sasaki, Angew. Chem.
Int. Ed. 46 (2007) 8392–8395.
36] V.N. Nikolic, V. Spasojevic, M. Panjan, L. Kopanja, A. Mrakovic, M. Tadic, Ceram.
Int. 43 (2017) 7497–7507.
Declaration of Competing Interest
37] A. Nadar, A.M. Banerjee, M.R. Pai, S.S. Meena, R.V. Pai, R. Tewari, S.M. Yusuf, A.
K. Tripathi, S.R. Bharadwaj, Appl. Catal. B: Environ. 217 (2017) 154–168.
38] Y. Benqun, L. Li, Z. Guojun, L. Xu, Z. Hailin, X. Shan, J. Catal. 381 (2020) 70–77.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[39] A. Nadar, A.M. Banerjee, M.R. Pai, R.V. Pai, S.S. Meena, R. Tewari, A.K. Tripathi,
Int. J. Hydrogen Energy 43 (2018) 37–52.
[
40] P. Datta, Mater. Res. Bull. 48 (2013) 4008–4015.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
7