Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous Al2O3 for Isobutane Dehydrogenation

Article abstract of DOI:10.1007/s10562-019-02686-x

Abstract: Ordered mesoporous Zn/OMA-Fe materials were easily prepared via one pot evaporation induced self-assembly (EISA) method in combination with incipient wetness strategy. Dehydrogenation of isobutane to isobutene were carried out on these materials, the isobutane conversion of 50.7% and the yield of 37.8% were obtained over 13Zn/OMA-10Fe catalyst at 580?°C with 300?h^?1 GHSV. The synthesized materials with large specific surface areas and uniform pore sizes were characterized by XRD, N₂ adsorption–desorption, TEM, XPS, H₂-TPR, M?ssbauer and NH₃-TPD. A portion of Fe species were highly dispersed on the support surface and others incorporated into Al₂O₃ frameworks, Zn species existed in the form of hexagonal ZnO phase. The total acidity of these catalysts was increased by the introduction of Zn, facilitating the conversion of isobutane. Moreover, the conversion of Fe species might play a major role in improving isobutene selectivity. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-019-02686-x

Catalysis Letters (2019) 149:1326–1336
https://doi.org/10.1007/s10562-019-02686-x
Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst
Based on Ordered Mesoporous Al O for Isobutane Dehydrogenation
2
3
Ming Cheng1,2 · Huahua Zhao · Jian Yang · Jun Zhao · Liang Yan · Huanling Song · Lingjun Chou
1
1
1
1
1
1,3
Received: 31 October 2018 / Accepted: 19 January 2019 / Published online: 25 February 2019
©
Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
Ordered mesoporous Zn/OMA-Fe materials were easily prepared via one pot evaporation induced self-assembly (EISA)
method in combination with incipient wetness strategy. Dehydrogenation of isobutane to isobutene were carried out on these
materials, the isobutane conversion of 50.7% and the yield of 37.8% were obtained over 13Zn/OMA-10Fe catalyst at 580 °C
−
1
with 300 h GHSV. The synthesized materials with large specific surface areas and uniform pore sizes were characterized
by XRD, N adsorption–desorption, TEM, XPS, H -TPR, Mӧssbauer and NH -TPD. A portion of Fe species were highly
2
2
3
dispersed on the support surface and others incorporated into Al O frameworks, Zn species existed in the form of hexago-
2
3
nal ZnO phase. The total acidity of these catalysts was increased by the introduction of Zn, facilitating the conversion of
isobutane. Moreover, the conversion of Fe species might play a major role in improving isobutene selectivity.
Graphical Abstract
Keywords Isobutane dehydrogenation · Zinc · Iron · Mesoporous alumina
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-019-02686-x) contains
supplementary material, which is available to authorized users.
2
*
Huanling Song
University of Chinese Academy of Sciences, Beijing 100049,
People’s Republic of China
songhl@licp.cas.cn
3
*
Lingjun Chou
Suzhou Research Institute of LICP, Chinese Academy
of Sciences, Suzhou 215123, People’s Republic of China
ljchou@licp.cas.cn
1
State Key Laboratory for Oxo Synthesis and Selective
Oxidation, Lanzhou Institute of Chemical Physics
(
LICP), Chinese Academy of Sciences, Lanzhou 730000,
People’s Republic of China
V
1 3
ol:.
(1234567890)

Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous…
1327
1
Introduction
afford high catalytic activity and stability on the green and
economical Fe-based catalysts is still a challenging work.
Furthermore, it is well known that promoters could
improve the catalytic property of catalysts in dehydrogena-
tion of light alkanes [17, 18]. ZnO can be acted as a pro-
moter in catalysis for the character of n-type semiconductor
and strong interaction with other metal [19, 20]. Zhang et al.
studied the catalytic performance on Zn-doped Pt/ZSM-5
catalyst in propane dehydrogenation, indicated that the intro-
duction of Zn increased the Pt dispersion and the interac-
tion of Pt with support, thus improved catalytic activity and
stability [21]. Silvestre-Albero reported that PtZn/MCM-41
catalysts showed a superior behavior in isobutane dehydro-
genation with 100% isobutene selectivity [22]. In addition,
the results from DFT calculations showed the addition of
Zn to Pt weakened C=C bond, which helped to increase
isobutene selectivity [23].
Isobutene is extensively used as chemical intermediate for
producing many kinds of industrial products, such as butyl
rubber, para-xylene, MTBE (methyl tert-butyl ether) and
methacrylonitrile [1, 2]. In the early industrialized pro-
duction stages, fluidized catalytic cracking and naphtha
steam cracking are two main ways to produce isobutene
[
3, 4]. However, along with the rapid development of pet-
rochemical industry, the growing demand for isobutene
follows, and the production came from such processes is
insufficient. In addition, the C hydrocarbon resources
4
from petroleum cracking increased greatly in the recent
years. These have resulted in considerable interest in direct
dehydrogenation of isobutane to isobutene because of its
lower energy demand, lower cost and higher selectivity
[
5, 6]. In this reaction, Cr and Pt based catalysts are the
most widely studied catalysts, which has been commercial-
ized in industrial production for decades [7–9]. However,
some problems are still remained such as the pollution of
Cr-based catalysts to environment or the higher cost of Pt-
based catalysts, which will limit their further applications
to some degree. Therefore, it is important for isobutane
dehydrogenation to exploit benign and cheap catalysts. By
comparison, the abundant and non-noble elements such as
iron, nickel, molybdenum, vanadium and cobalt are very
attractive candidates that can form various heterogeneous
catalysts. Tian et al. reported the yield of isobutene over
V-K/γ-Al O catalysts in the isobutane dehydrogenation
In our previous work, ordered mesoporous material
exhibited better catalytic stability than traditional support,
but the activity was not satisfied [24, 25]. Herein, a series of
Fe–Zn dual-sites catalysts based ordered mesoporous Al2O3
(OMA) were explored to catalyze the dehydrogenation of
isobutane. First of all, OMA-Fe materials with different Fe
content were facilely synthesized via evaporation induced
self-assembly (EISA) strategy, then the loading of Zn via
incipient-wetness impregnation method. The fresh and
some used catalysts were characterized by XRD, N2 adsorp-
tion–desorption, H2-TPR, Mӧssbauer and so on. Particular
emphasis was focused on the interaction of metals, the forms
of Fe (Zn) species and surface acidity of the catalysts. And
all the characterizations were associated with relevant cata-
lytic reactions. As a result, the inexpensive and pollution-
free catalyst of 13Zn/OMA-10Fe showed the promising
prospect for dehydrogenation of light hydrocarbon.
2
3
improved significantly after sulfidation [10]. Wang et al.
investigated the effects of support over Co-based catalysts
for isobutane dehydrogenation, silica-supported catalysts
with no acidity exhibited the best catalytic performance
[
11].
Recently, Fe-based catalysts are drawing great inter-
est in alkanes dehydrogenation, which are environmental-
friendly and inexpensive [12, 13]. These studies focused
on improving the catalytic activity of Fe-based catalysts by
exploring various supports, preparation methods and pre-
treatments [14–16]. Yun et al. employed different silicate
zeolites (ZSM-5, beta zeolite) as support for the dehydro-
genation of propane, and found that isolated framework
Fe sites played a key role in propane selectivity [14]. Xu
et al. compared Fe/Al O catalysts prepared by vapor dep-
2 Experimental
2.1 Catalyst Preparation
During the synthesis of Fe–Zn dual-sites catalysts, the
main reagents included ferric nitrate (Fe(NO3)3·9H2O,
≥ 99.0%), zinc nitrate (Zn(NO3)2·6H2O, ≥ 99.0%), alu-
minum isopropoxide (C9H21AlO3, ≥98.0%), Pluronic P123
(EO20PO70EO20, Mw=5800), nitric acid (HNO3, 67 wt%)
and absolute ethanol (C2H5OH). All the chemicals were
A.R. grade and directly used without purification.
2
3
osition with impregnation method. The former presented
a much higher activity in oxidative dehydrogenation of
ethane [15]. Sun et al. investigated the impact of hydro-
gen reduction on Fe O /Al O catalysts, which achieving
OMA-Fe materials were synthesized via EISA strategy
with fine control according to the previous literatures [26].
Typically, 1 g P123 was dissolved 20 ml absolute ethanol.
Then a mmol Fe(NO3)3·9H2O, b mmol aluminum isopropox-
ide (a+b=10 mmol) and 1.6 ml nitric acid was added. After
2
3
2
3
2
4% propane conversion and 83% propene selectivity [16].
However, these catalysts exhibited relatively low activity
or deactivated rapidly. None of these can be compared to
commercial catalysts at the reactivity levels. Therefore, to
6
h of stirring, the mixed solution was shifted to a culture
1
3

1328
M. Cheng et al.
dish, then transferred the culture dish into a 60 °C drying
oven for 2 days. Finally calcined at 600 °C with a ramp-
X-ray diffraction (XRD) was operated on X’Pert Pro dif-
fractometer system with Cu Kα radiation ranging from 10.0
to 80.0° and 0.6 to 5.0°.
−
1
ing rate of 1 °C min for 5 h. The catalysts were denoted
as OMA-xFe, where x refers to Fe/(Al+Fe) molar ratio in
initial precursor.
Transmission electron microscopy (TEM) was carried out
2
on the TECNAIG F20 instrument.
Ordered mesoporous Zn/OMA-Fe materials were synthe-
sis via incipient wetness method. The OMA-xFe materials
were impregnated with the relevant amount of zinc nitrate
solution and calcined at 600 °C for 5 h. The obtained cata-
lysts were denoted as yZn/OMA-xFe, where y stands for
wt% of ZnO.
X-ray photoelectron spectroscopy (XPS) were performed
on Thermon ESCALAB 250xi spectrometer. The binding
energies were calibrated against at 284.8 eV of C1s.
H temperature-programmed reduction (H -TPR) was
2
2
performed on ChemBET Pulsar Analyzer. About 100 mg
catalyst was loaded, pretreated at 600 °C in He flow for
45 min. After cooling to room temperature, then raised tem-
For comparison, ordered mesoporous alumina (OMA)
was prepared by the same via EISA strategy. And the 13Zn/
OMA material was prepared on the same incipient wetness
method, where 13 stands for wt% of ZnO.
perature to 800 °C in 10% H -Ar mixed flow.
2
NH temperature-programmed desorption was performed
3
on ChemBET Pulsar analyzer combined with mass spectros-
copy. After being pretreated at 600 °C in He flow, the cata-
lyst (150 mg) was adsorbed to saturation by NH at 120 °C.
3
2.2 Catalytic Dehydrogenation
The profile was detected from 50 to 600 °C.
57
The Fe Mӧssbauer were measured by a MA-260 (Bench
MB-500) Mӧssbauer spectrometer at room temperature and
the Doppler velocity was calibrated against α-Fe foils.
The experiments for isobutane dehydrogenation were con-
ducted in fixed-bed. Before the test, the catalyst (900 mg)
was sieved at 60–80 mesh, and loaded in quartz tube diluted
with quartz grain. Then catalyst bed was programmed up to
desire temperature (560–620 °C) under N flow, pure isob-
3 Results and Discussion
2
utane gas was switched in fixed flow at gas hourly space
−
1
velocity (GHSV) of 300 h . The reactants and products
were analyzed using gas chromatography equipped with a
flame ionization detector and thermal conductivity detectors.
3.1 Nitrogen Adsorption–Desorption Analysis
Nitrogen adsorption–desorption characterization is used to
study textural properties of OMA-xFe and yZn/OMA-10Fe
catalysts and the results are shown in Fig. 1. Each sample
exhibited IV type isotherm with H1 hysteresis loop, indi-
cated the presence of mesopore in these catalysts [27]. The
textural properties of above samples were listed in Table 1.
The specific surface areas and pore volumes of OMA-xFe
2.3 Characterization
The specific surface areas, pore volumes and average pore
diameters were determined by N adsorption and desorption
2
2
−1
on Autosorb-iQ analyzer.
materials were not clearly change (around 200 m g ) with
Fig. 1 Nitrogen adsorption–desorption analysis of the fresh catalysts: (a) OMA-3Fe; (b) OMA-5Fe; (c) OMA-7Fe; (d) OMA-10Fe; (e) 5Zn/
OMA-10Fe; (f) 7Zn/OMA-10Fe; (g) 10Zn/OMA-10Fe; (h) 13Zn/OMA-10Fe; (i) 20Zn/OMA-10Fe
1
3

Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous…
1329
Table 1 The textural data of the fresh catalysts
of well-ordered mesoporous structure [27, 31]. Moreover,
it should be noted that the diffraction peak around 0.8° and
SBET (m2 g−1)
V (cm3 g−1)
Samples
dP (nm)
P
1
.3° was still maintained after the introduction of Zn, sug-
OMA-3Fe
OMA-5Fe
OMA-7Fe
OMA-10Fe
209.0
207.1
203.4
205.3
186.7
184.6
190.4
183.7
178.8
0.33
0.32
0.31
0.32
0.31
0.29
0.35
0.30
0.36
9.5
9.5
9.5
9.5
9.6
9.5
9.5
9.5
9.6
gesting that the ordered mesoporous yZn/OMA-10Fe with
different Zn contents were successfully synthesized. The
wide-angle XRD patterns of abovementioned materials are
shown in Fig. 2B. No diffraction peaks for Al O or iron
2
3
5
7
1
1
2
Zn/OMA-10Fe
Zn/OMA-10Fe
0Zn/OMA-10Fe
3Zn/OMA-10Fe
0Zn/OMA-10Fe
species were visible in the OMA-xFe materials, implying Fe
species were highly dispersed on support surface or incorpo-
rated into amorphous Al O frameworks [22, 32]. However,
2
3
the hexagonal crystalline ZnO (ICDD No. 89-1397) was
apparently observed on the yZn/OMA-10Fe catalysts more
than 7%Zn and the diffracted intensity increased distinctly
with increasing Zn content.
the increment of iron content up to 10%, which implying
similar pore structural properties of these OMA-xFe materi-
als. It has been known larger surface area contributed to the
dispersity of Fe active sites, which could restrain the metal
sintering. Thus, the big specific surface areas of these fresh
OMA-xFe catalysts were conducive to catalytic stability [28,
3.3 TEM analysis
TEM characterization is one of the most effective techniques
to further confirm the ordered mesoporous structure of the
samples. OMA-10Fe and 13Zn/OMA-10Fe were chosen
as representatives of abovementioned OMA-xFe and yZn/
OMA-10Fe materials (Fig. 3). As we can see, these two
materials both presented one-dimensional parallel channels
along [1 1 0] (Fig. 3a, c) and hexagonal pores along [1 0 0]
(Fig. 3b, d), which directly verified the presence of ordered
2
9]. Furthermore, it is reasonable that the specific surface
areas of yZn/OMA-10Fe catalyst decreased slightly after
the introduction of Zn, illustrating that it was successfully
impregnated on the pore channel of OMA-10Fe materials
[
30].
meso-structure and accorded with SXRD and N adsorp-
2
3.2 XRD Analysis
tion–desorption analysis.
For characterizing the pore symmetry and crystalline
structure, the XRD patterns of all the fresh samples are
displayed in Fig. 2. The low-angle XRD patterns of OMA-
xFe (Fig. 2A) indicated that all these samples presented
an intense (100) peak at 0.8° and a very weak (110) peak
around 1.3°, which were the characteristics of P6mm hexag-
onal symmetry, indicating that these samples are all typical
3.4 XPS Analysis
The Fe (Zn) 2p XPS patterns of fresh catalysts are in
Fig. 4. Fe 2p peaks for all the OMA-xFe catalysts
3
/2
appeared at 711.7 eV with a satellite peak around 719 eV.
It demonstrated that Fe species were Fe(III), as proved by
13.5 eV between Fe 2p and 2p peaks in spin–orbital
1
/2
3/2
Fig. 2 XRD patterns of the fresh catalysts: (a) OMA-3Fe; (b) OMA-5Fe; (c) OMA-7Fe; (d) OMA-10Fe; (e) 5Zn/OMA-10Fe; (f) 7Zn/OMA-
0Fe; (g) 10Zn/OMA-10Fe; (h) 13Zn/OMA-10Fe; (i) 20Zn/OMA-10Fe
1
1
3

1330
M. Cheng et al.
Fig. 3 TEM images of the fresh catalysts: a and b OMA-10Fe; c and d 13Zn/OMA-10Fe
Fig. 4 Fe2p (A) and Zn2p (B) XPS spectra of the fresh catalysts: (a) OMA-3Fe; (b) OMA-5Fe; (c) OMA-7Fe; (d) OMA-10Fe; (e) 5Zn/OMA-
0Fe; (f) 7Zn/OMA-10Fe; (g) 10Zn/OMA-10Fe; (h) 13Zn/OMA-10Fe; (i) 20Zn/OMA-10Fe
1
splitting theory [33–35]. Moreover, the bands of Zn 2p3/2
were located at the BE = 1021.7 eV, in consistent with the
standard Zn(II) [36, 37]. Significantly, the Fe 2p peaks
shift, while the Zn 2p peaks exhibited a diminutive posi-
3/2
tive shift. It suggested the very weak electron-transport
between Fe and Zn species, and the two metals might may
function independently in dehydrogenation reaction.
3
/2
of yZn/OMA-10Fe catalysts exhibited a slight negative
1
3

Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous…
1331
3
.5 H ‑TPR Profiles
reduction of metal oxides deriving from the semi-conductor
feature of ZnO itself [41].
2
H -TPR experiment is an effective method for testing the
2
reduction behaviors of the heterogeneous catalysts. TPR pro-
files of OMA-xFe and yZn/OMA-10Fe catalysts are shown
in Fig. 5. As shown, all the OMA-xFe catalysts with differ-
ent Fe contents exhibited similar profiles, which presenting
three hydrogen reduction peaks. The low temperature region
in 445–480 °C belong to the reduction of Fe O to Fe O ,
3.6 Mӧssbauer Spectra
5
7
Figure S1 shows the Fe Mӧssbauer spectra of fresh and
spent catalysts. The associated fitting data are in Table 2. The
5
7
Fe Mӧssbauer spectra of OMA-10Fe catalyst displayed a
3
+
doublet peak: the IS of 0.33 mm/s was characteristic of Fe
2
3
3
4
the high temperature region around 570 °C was linked with
in octahedral site with content of 100%; Same phenomenon
was also found in 13Zn/OMA-10Fe [42]. These results were
well consistent with the conclusion in XPS analysis, the Fe
species on the fresh OMA-xFe and yZn/OMA-10Fe cata-
the reduction of Fe O to FeO, and the higher reduction peak
2
3
around 740 °C was the reduction to Fe [23, 38, 39]. After
introducing 5% ZnO to OMA-10Fe catalyst, the TPR pro-
file still exhibited three reduction peaks. However, it was
worth noting that only one reduction peak appearing around
3
+
lysts existed in a form of Fe . However, after reaction at
620 °C, the spent OMA-10Fe displayed two doublet peaks.
3+
4
60 °C in the profiles of 7~20Zn/OMA-10Fe catalysts with
A doublet peak with IS of 0.41 mm/s was ascribed to Fe in
octahedral site (54%), and another with IS of 1.06 mm/s was
the introduction of more zinc. Combining with the profile
of 13Zn/OMA, the only reduction peak in 561 °C assigned
2
+
assigned to Fe in octahedral site (46%), suggesting that
2
+
3+
to the reduction of Zn [40]. It is speculated that the reduc-
sectional Fe was reduced by the dehydrogenation products
3
+
tion of Fe in Zn/OMA-10Fe catalysts was easily to Fe O
during reaction [42, 43]. Besides, compared with the spent
3
4
2
+
after doping more Zn, because its existence can promote the
OMA-10Fe catalyst, the content of Fe in the spent 13Zn/
OMA-10Fe decreased, which resulted from the reduction of
3
+
Fe in yZn/OMA-10Fe catalysts was more prone to Fe O
3
4
rather than further reduction to FeO and Fe. These results
are accorded with the conclusion in H -TPR analysis.
2
3
.7 NH ‑TPD Analysis
3
The catalytic activity of the catalyst is strongly related to
surface acidic properties of the catalyst. The effects of Zn on
the amount of acid sites of the yZn/OMA-10Fe catalyst are
determined by NH -TPD and the corresponding profiles are
3
depicted in Fig. S2. As we can see, the catalysts presented a
broad asymmetric desorption peak at 50–600 °C, which indi-
cating the different kinds of acidic sites on these catalysts.
The peak located between 50 and 200 °C, 200–450 °C and
>
450 °C, could be accredited to weak, medium and strong
acidic sites respectively [28, 30, 44]. All the NH -TPD
3
results are de-convoluted using Gaussian functions and the
corresponding the amount of acidic sites are presented in
Table 3. The total acid amount of OMA-xFe catalysts with
different iron contents and pure OMA support was very sim-
ilar, even the weak, medium and strong acid amount was
Fig. 5 H -TPR profiles of the fresh catalysts: (a) OMA-3Fe; (b)
2
OMA-5Fe; (c) OMA-7Fe; (d) OMA-10Fe; (e) 5Zn/OMA-10Fe; (f)
7
2
Zn/OMA-10Fe; (g) 10Zn/OMA-10Fe; (h) 13Zn/OMA-10Fe; (i)
0Zn/OMA-10Fe; (j) 13Zn/OMA
Table 2 57Fe Mӧssbauer
parameters of the
Sample
IS (mm/s)
QS (mm/s)
LW (mm/s)
Area (%)
Remarks
catalysts before and after
dehydrogenation
3+
Fresh-OMA-10Fe
Fresh-13Zn/OMA-10Fe
Spent-OMA-10Fe
0.33
0.33
0.41
1.22
1.22
1.19
1.91
1.20
1.99
0.51
0.47
0.58
0.58
0.63
0.58
100
100
54
46
69
Fe in Oh
3
+
Fe in Oh
3
+
Fe in Oh
2
+
1
.06
0.41
.06
Fe in Oh
3+
Spent-13Zn/OMA-10Fe
Fe in Oh
2+
1
31
Fe in Oh
1
3

1332
M. Cheng et al.
Table 3 The acidic properties
and the results of isobutane
dehydrogenation over the
catalysts
The amount of acidic sites (mmol g−1
a
TOFb (h−1)
Samples
)
Activity
−1
mmol g min−1
(
)
Weak
Medium
Strong
Total
OMA
0.042
0.041
0.042
0.043
0.045
0.085
0.091
0.093
0.098
0.102
0.097
0.023
0.023
0.024
0.024
0.025
0.055
0.063
0.064
0.069
0.071
0.068
0.010
0.009
0.009
0.010
0.009
0.017
0.017
0.018
0.019
0.019
0.024
0.075
0.073
0.075
0.077
0.079
0.157
0.171
0.175
0.186
0.192
0.189
11.8
–
–
–
–
–
–
31.6
31.8
34.2
33.4
36.5
31.6
36.8
OMA-3Fe
OMA-5Fe
OMA-7Fe
OMA-10Fe
–
115.3
229.9
260.4
272.2
304.5
291.9
287.5
5Zn/OMA-10Fe
7Zn/OMA-10Fe
10Zn/OMA-10Fe
13Zn/OMA-10Fe
20Zn/OMA-10Fe
13Zn/OMA
a
Activity was defined as mmol of produced isobutene per gram of catalyst per min
TOF was defined as total moles of isobutane converted/total moles of acid sites per hour
b
−1
The activity and TOF was calculated over the isobutane dehydrogenation at T=580 °C, GHSV=300 h
almost identical, suggesting the introduction of iron didn’t
increase the acidity of OMA-xFe catalysts. The three types
of acid sites of OMA-xFe catalysts originated from surface
coordinatively unsaturated Al sites on pure OMA support
other words, the highly dispersed Fe O species were excel-
2 3
lent active sites for isobutane dehydrogenation. the optimal
conversion of isobutane reached 36.6% and the selectivity of
isobutene attained 64.8% at 620 °C over OMA-10Fe catalyst.
Furthermore, isobutane conversion and isobutene selectivity
increased gradually with time on stream, indicating the great
stability of the OMA-10Fe catalyst.
[
45]. However, the amount of acid sites of yZn/OMA-10Fe
catalysts significantly enhanced due to the introduction of
Zn. The three types acidic amount of 5Zn/OMA-10Fe cata-
lysts was almost twice that of OMA-10Fe catalysts. And
with the increasing of Zn loadings, the amount of acid
sites of these catalysts increased gradually. It is generally
accepted that Zn species can provide acid sites, combined
with XRD analysis, the increased acid sites of yZn/OMA-
Isobutane dehydrogenation is an endothermic reaction
in thermodynamics, which need relatively high temperature
to obtain excellent isobutene yield [48]. For exploring the
influence of reaction temperature, the dehydrogenation reac-
tion was carried out at 580–640 °C (Fig. 6b). The conversion
of isobutane was 18.7% and the selectivity of isobutene was
76.9% at 580 °C after 30 min. With increasing the reaction
temperature to 640 °C, the conversion of isobutane increased
to 49.1% and selectivity of isobutene decreased to 55.2%
at the same reaction time, which was very conformed to
the characteristic of alkane dehydrogenation reaction [49].
In consideration of raising reaction temperature, too much
undesired by-products followed. Furthermore, isobutane
would suffer serious homogeneous conversion after the reac-
tion temperature above 600 °C (Fig. S3). Thus, to minimize
homogeneous reaction and by-products, 580 °C was selected
to further determine the catalytic performance of yZn/OMA-
10Fe catalysts.
1
0Fe catalysts derived from ZnO crystal [46]. In addition, it
was well established that the presence of total acidity could
facilitate alkane conversion, which may be beneficial for the
catalytic dehydrogenation of isobutane [47]. Therefore, the
introduction of Zn on yZn/OMA-10Fe catalysts may have a
promoting effect on the catalytic performance of isobutane
dehydrogenation.
3
.8 Catalytic Performance in the Isobutane
Dehydrogenation
The reactivity of isobutane dehydrogenation over different
OMA-xFe catalysts is shown in Fig. 6a. As we can see that
the OMA-xFe catalysts with different Fe contents presented
similar initial isobutane conversion and different initial isob-
utene selectivity. However, with increasing reaction time, the
isobutane conversion of these catalysts slightly increased
and isobutene selectivity enhanced rapidly, which may result
The reactivity of isobutane dehydrogenation over dif-
ferent yZn/OMA-10Fe catalysts is shown in Fig. 6c. It
could be seen that the 5Zn/OMA-10Fe catalyst exhibited
a notably high isobutane conversion (37.2%) and isobu-
tene selectivity (about 77%) at 580 °C after 30 min. With
increasing the Zn contents, the isobutane conversion grad-
ually increased to the maximum, then slightly decreased
thereafter, however isobutene selectivity changed lit-
tle. Over the 13Zn/OMA-10Fe catalyst, the isobutane
3+
2+
from the partial conversion of Fe to Fe . Finally, the isob-
utane conversion and isobutene selectivity of OMA-xFe cat-
alysts was similar at 120 min time on stream suggesting that
isobutane dehydrogenation within the scope of this paper. In
1
3

Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous…
1333
Fig. 6 The reactivity of isobutane dehydrogenation over the fresh catalysts: a OMA-xFe; b OMA-10Fe; c yZn/OMA-10Fe; d OMA and 13Zn/
OMA
conversion reached 50.7%, the selectivity of isobutene
attained 74.7%. The isobutane conversion and isobutene
yield slowly declined with the increasing reaction time,
which was due to the coke formation or the change of
active species over the catalyst [50, 51].
In contrast to OMA, 13Zn/OMA showed better initial
activity of isobutane dehydrogenation, the conversion was
achieved about 55% and isobutene selectivity about 66%.
Thus, ZnO can be deduced as a kind of active sites for isobu-
tane dehydrogenation. In addition, the isobutane conversion
of 13Zn/OMA-10Fe catalyst was 2.7 times that of OMA-
10Fe catalyst at 580 °C after 30 min, indicating that ZnO was
an effective promoter to the isobutane catalytic conversion,
In order to further investigate what species were active
sites for isobutane dehydrogenation, we tested this reaction
at 580 °C over pure OMA support and 13Zn/OMA catalyst
as a contrast (Fig. 6d). It can be seen that the isobutane con-
version and isobutene yield over pure OMA support were
only 3.1% and 2.1%, respectively. It illustrated that ordered
mesoporous alumina (OMA) was inactive to the dehydro-
genation. When iron was introduced into OMA, the activity
dramatically enhanced. For example, the isobutane conver-
sion and isobutene yield over OMA-10Fe were 18.7% and
2
+
Zn was indeed identified as another active site. In order to
further understand the reactivity over different yZn/OMA-
10Fe catalysts, the TOF values were calculated and listed in
Table 3. TOF data indicated that the isobutane dehydrogena-
tion activity depends on the acid sites to some extent. The
−
1
maximum was obtained over 13Zn samples (36.5–36.8 h )
with higher total acid sites. Moreover, the activity decreased
−
1
1
4.4%. In association with the XPS characterization and
to 31.6 h on 20Zn sample with the most acid sites, which
implied that the acid sites from ZnO are not alone the active
sites. It should be noted that the level of reactivity is usu-
ally the result of combined influence of multiple factors in
heterogeneous catalytic system, which doesn’t depend on a
single factor, such as active component, catalyst acidity or
the interaction of metal. Thus, the concept of “Activity” may
3
+
Mӧssbauer spectra of fresh OMA-Fe catalysts, Fe species
was concluded as probably active sites. In addition, with
increasing reaction time, isobutene selectivity of OMA-xFe
catalysts enhanced rapidly, which implied the conversion of
3+
2+
Fe to Fe played an important role in improving isobutene
selectivity.
1
3

1334
M. Cheng et al.
3
+
2+
be more accurate to reflect catalytic performance (Table 3).
Compared with OMA support, the “Activity” of isobutene
on OMA-10Fe catalyst showed noticeable growth. With the
introduction of Zn, the “Activity” further improved up to the
max of 13% Zn contents. These results intuitively reflected
the level of catalyst performance over different catalysts.
indicating a part of Fe was transformed to Fe [52].
However, compared with the fresh 13Zn/OMA-10Fe cata-
lyst, the Zn 2p peak of the spent 13Zn/OMA-10Fe cata-
3
/2
lyst exhibited an obvious positive shift, which resulted from
the enhancement of the binding energy, the increase of the
electron cloud density surrounding Zn atom and suggest-
ing the reduction of the surface ZnO [53]. Meanwhile, in
the XPS spectra of O 1s, there is no distinction between
O1s XPS spectra of OMA-10Fe and the spent OMA-10Fe
(Fig. S5). However, the O1s binding energy of 13Zn/OMA-
3.9 The Characterizations of the Spent Catalysts
3.9.1 WXRD Analysis
1
0Fe catalyst exhibited a visible positive shift after reac-
3
+
The WXRD patterns of the fresh and corresponding spent
catalysts are presented in Fig. S4. Compared with the fresh
catalysts, the phase structure of spent OMA-10Fe and 13Zn/
OMA-10Fe catalyst was not changed. Moreover, there is
tion, which further verifying some Fe were transformed
2
+
to Fe and the presence of surface ZnO reduction. In order
to clarify the change of Fe, Zn species in the dehydrogena-
tion reaction, the XPS data was listed in Table 4. We can see
that the mole ratio of Fe/(Fe+Al) in the Fresh-OMA-10Fe
and Fresh-13Zn/OMA-10Fe catalyst surface was similar
and less than the total mole ratio value (10%), which was
due to sectional Fe species incorporated into Al O frame-
0
no other iron carbide or Zn phase, illustrating that phase
structure of Zn species was stable and the superficial Fe
species were also highly dispersed during this dehydrogena-
tion reaction.
2
3
works. These results are accorded with the result in wide-
angle XRD analysis. Furthermore, after the dehydrogenation
reaction, it was worth noting that the Fe content of support
surface increased, Zn content of support surface decreased,
indicating some Zn species might migrate into bulk phase
and slight Fe species in Al O frameworks move out to sup-
3
.9.2 XPS Analysis
The Fe 2p and Zn 2p XPS spectra of the fresh and cor-
responding spent catalysts are shown in Fig. 7. It can be
observed that the Fe 2p peaks for the spent catalysts
3
/2
2
3
was 711.3 eV, which shifted 0.4 eV to lower BE values,
port surface. The above results implied the surface Fe and Zn
Fig. 7 Fe2p (A) and Zn2p (B) XPS spectra of the catalysts before and after dehydrogenation: (a) OMA-10Fe; (b) the spent OMA-10Fe; (c)
3Zn/OMA-10Fe; (d) the spent 13Zn/OMA-10Fe
1
Table 4 The percentage of
Sample
Fe (%)
Zn (%)
Al (%)
O (%)
Mole ratio of Fe
and Fe+Al (%)
Fe, Zn, Al, O element and Fe/
(
Fe+Al) mole ratio
Fresh-OMA-10Fe
Fresh-13Zn/OMA-10Fe
Spent-OMA-10Fe
2.13
1.92
2.49
2.46
–
7.16
–
35.20
33.17
37.00
34.15
62.67
58.37
60.50
57.59
5.7
5.5
6.3
6.7
Spent-13Zn/OMA-10Fe
5.80
1
3

Synthesis and Catalytic Performance of a Dual-Sites Fe–Zn Catalyst Based on Ordered Mesoporous…
1335
species underwent reconstruction during dehydrogenation
such as reduction by the gas phase products.
seldom encountered oligomerization under this experimental
condition.
3.9.3 TG-DSC Analysis
4 Conclusions
As we all know, coking is usually the main reason for cata-
lyst deactivation in alkane dehydrogenation process. The
TG-DSC is a very efficient technique to analyze the cok-
ing amount and the characterization results are shown in
Fig. 8. The TG curves presented downtrend with the increase
of temperature, which was divided into two stages. In the
first stage, a slight weight losses in TG curves up to 300 °C,
ascribed to the loss of physically adsorbed water. The second
stage, located in the 350–500 °C range, the weight losses
belonged to the removal of coke. It was worth noting that the
coke content in spent OMA-10Fe was only 1.9%, but in the
spent 13Zn/OMA and 13Zn/OMA-10Fe catalysts was about
A series of Fe–Zn dual-sites catalysts based ordered
mesoporous Al O were easily prepared via one pot evapo-
2
3
ration induced self-assembly in combination with post-
impregnation method, while the pure OMA support and
13Zn/OMA catalyst were utilized for comparison. The
obtained materials possessed large specific surface areas,
big pore volumes and uniform pore sizes. A part of Fe spe-
cies were highly dispersed on support surface and the oth-
ers incorporated into Al O frameworks, Zn species existed
2
3
in the form of hexagonal ZnO phase. Mӧssbauer analysis
3
+
indicated that Fe species were in a form of Fe in octahe-
dral site in the fresh catalysts. During the reaction, sectional
3
+
2+
5
%. Combined with the stability of these three catalysts, the
Fe in octahedral site was reduced to Fe in octahedral
site by by-products of the dehydrogenation. The conversion
of Fe species played an important role in improving isobu-
tene selectivity. After the introduction Zinc, the total acid-
ity of the catalysts was also enhanced, which was benefit
reason of catalyst deactivation may caused by coke. In addi-
tion, the carbon balance value of the three spent catalysts
changed in the range of 94–98%, which suggested the lost
carbon was mainly existed in the form of coke and isobutene
Fig. 8 TG-DSC curves of the spent catalysts and carbon balance analysis: a the spent OMA-10Fe; b the spent 13Zn/OMA-10Fe; c the spent
3Zn/OMA, and d carbon balance over the three catalysts
1
1
3

1336
M. Cheng et al.
to promote the catalytic conversion of isobutane. Conse-
quently, the enhanced dehydrogenation performance could
be interpreted as a combined influence of Fe and Zn species.
Notably, the catalysts exhibited excellent activity with 50.7%
isobutane conversion and 37.8% isobutene yield, demon-
strated great potential for commercial applications.
23. Albero JS, Castillo MAS, He R, Escribano AS, Reinoso FR,
Dumesic JA (2001) Catal Lett 74:1–2
2
4. Zhao HH, Song HL, Yang J, Zhao J, Yan L, Chou LJ (2017) Catal
Sci Technol 15:3256–3267
25. Cheng M, Zhao HH, Yang J, Zhao J, Yan L, Song HL, Chou LJ
(2018) Microporous Mesoporous Mater 266:117–125
2
6. Morris SM, Fulvio PF, Jaroniec M (2008) J Am Chem Soc
1
30:15210–15216
2
7. Li L, Han W, Dong F, Zong L, Tang Z, Zhang J (2017) Micropo-
rous Mesoporous Mater 279:1–9
Acknowledgements The authors sincerely acknowledge the financial
support from the National Natural Science Foundation of China (No.
2
2
8. He Z, Jiao Y, Wang J, Chen Y (2018) Fuel 212:193–201
9. Jiao Y, Li S, Liu B, Du Y, Wang J, Lu J (2015) Appl Therm Eng
2
1573260 and No. 21403255), Youth Innovation Promotion Associa-
tion CAS (No. 2017460) and Jiangsu Science and Technology Program
No. BY2016048-01).
9
1:417–425
(
3
0. H HW, Jeong MH, Koo HM, Chung CH, Bae JW (2015) React
Kinet Mech Catal 116:173–189
Compliance with Ethical Standards
31. Chen J, Cai Z, Chen X, Zheng X, Zheng Y (2016) Mater Lett
1
68:214–217
3
2. Xiang X, Zhao HH, Yang J, Zhao J, Yan L, Song HL, Chou LJ
Conflict of interest All authors declare no conflicts of interest.
(
2016) Appl Catal A 520:140–150
3
3
3. Yamashita T, Hayes P (2008) Appl Surf Sci 254:2441–2449
4. Hawn DD, DeKoven BM (1987) Surf Interface Anal 10:63
References
35. Sun Y, Tao L, You T, Li C, Shan H (2014) Chem Eng J
2
44:145–151
3
3
6. Schoen G (1973) J Electron Spectrosc Relat Phenom 75:2
7. Velu S, Suzuki K, Gopinath CS, Yoshida H, Hattori T (2002) Phys
Chem Chem Phys 4:1990–1999
1
2
.
.
Centeno MA, Debois M, Grange P (2000) J Catal 192:296–306
Zhu G, Li P, Zhao F, Song H, Xia C (2016) Catal Today
2
76:105–111
3
3
8. Cao J, Wang Y, Yu X, Wang S, Wu S, Yuan Z (2008) Appl Catal
B 79:26–34
3
4
.
.
Bhasin MM, Mccain JH, Vora BV, Imai T, Pujadó PR (2001) Appl
Catal A 221:397–419
9. Munteanu G, Ilieva L, Andreeva D (1997) Thermochim Acta
Otroshchenko T, Radnik J, Schneider M, Rodemerck U, Linke D,
Kondratenko EV (2016) Chem Commun 52:8164–8167
Tian Y, Liu X, Rood MJ, Yan Z (2017) Appl Catal A 545:1–9
Dong A, Wang K, Zhu S, Yang G, Wang X (2017) Fuel Process
Technol 158:218–225
2
91:171
4
4
4
0. Wang G, Li C, Shan H (2016) Catal Sci Technol 6:3128–3136
1. Liang MS, Kang WK, Xie KC (2009) J Nat Gas Chem 18:110–113
2. Dyar MD, Agresti DG, Schaefer MW, Grant CA, Sklute EC (2006)
Annu Rev Earth Planet Sci 34:83–125
5
6
.
.
7
8
9
.
.
.
Cavani F, Koutyrev M, Trifirò F, Bartolini A, Ghisletti D, Iezzib
R, Santuccib A, Pieroc GD (1996) J Catal 158:236–250
Nawaz Z, Tang X, Zhang Q, Wang D, Fei W (2009) Catal Com-
mun 10:1925–1930
4
4
4
4
4
4
4
5
5
5
5
3. Mimura N, Takahara I, Saito M, Sasaki Y, Murata K (2002) Catal
Lett 78:1–4
4. Kang SH, Bae JW, Jun KW, Potdar HS (2008) Catal Commun
9
:2035–2039
Siri GJ, Bertolini GR, Casella ML, Ferretti OA (2005) Mater Lett
5. Rodemerck U, Kondratenko EV, Otroshchenko T, Linke D (2016)
Chem Commun 52:12222–12225
5
9:2319–2324
1
1
1
1
0. Tian Y, Liu Y, Liu X, Yan Z (2018) Catal Sci Technol
:5473–5481
1. Wang G, Zhu X, Zhang J, Sun Y, Li C, Shan H (2014) RSC Adv
6. Liu G, Liu J, He N, Miao C, Wang J, Xin Q, Guo H (2018) RSC
Adv 8:18663–18671
8
7. Shi Y, Li X, Rong X, Gu B, Wei H, Sun C (2017) RSC Adv
4
:57071–57082
7
:19841–19848
2. Michorczyk P, Kustrowski P, Chmielarz L, Ogonowski J (2004)
React Kinet Catal Lett 82:121–130
8. Weckhuysen BM, Schoonheydt RA (1999) Catal Today
5
1:223–232
3. Sugiyama S, Osaka T, Hirata Y, Kondo Y, Nakagawa K, Sotowa
K (2007) J Chem Eng Jpn 40:1064–1071
9. Mu J, France LJ, Liu B, Shi J, Long J, Lv L, Li X (2016) Catal Sci
Technol 6:8562–8570
1
1
4. Yun JH, Lobo RF (2014) J Catal 312:263–270
0. Gu B, He S, Rong X, Shi Y, Sun C (2016) Catal Lett
5. Xu L, Lin X, Xia Y, Lu X, Wang C, Liu C (2014) Mater Res Bull
1
46:1415–1422
5
9:254–260
1. Xia K, Lang W, Li P, Yan X, Guo Y (2015) RSC Adv
1
6. Sun Y, Wu Y, Tao L, Shan H, Wang G, Li C (2015) J Mol Catal
A 397:120–126
5
:64689–64695
2. Wang X, Zhang J, Chen J, Ma Q, Fan S, Zhao T (2018) Chin J
Chem Eng 26:761–767
1
1
1
7. Hill JM, Cortright RD, Dumesic JA (1998) Appl Catal A 168:9–21
8. Zhang S, Zhou Y, Zhang Y, Huang L (2010) Catal Lett 135:76–82
9. Aoun M, Chater M, Marecot P, Especel C, Lafaye G (2010) Stud
Surf Sci Catal 175:237–240
3. Li Z, Zhang Q, Liu X, Chen M, Wu L, Ai Z (2018) Appl Surf Sci
4
52:123–133
2
2
0. Yu C, Xu H, Ge Q, Li W (2007) J Mol Catal A 266:80–87
1. Zhang Y, Zhou Y, Huang L, Zhou S, Sheng X, Wang Q, Zhang C
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
(
2015) Chem Eng J 270:352–361
2
2. Albero JS, Ruiz JCS, Escribano AS, Reinoso FR (2008) Appl
Catal A 351:16–23
1
3