Insights into the doping effect of rare-earth metal on ZnAl2O4 supported PtSn catalyzed isobutane dehydrogenation

Article abstract of DOI:10.1016/j.cattod.2020.04.016

Isobutane dehydrogenation is a vital route for the production of isobutene, an important substance for methyl tert-butyl ether. However, the reaction is typically performed at relatively low pressure and high temperature, resulting in a facilitated coke formation. Here, we used rare-earth metals (Y, La, Ce) as dopants to modify the ZnAl₂O₄ support and studied their effects on Pt-Sn catalyzed dehydrogenation of isobutane. Combining the experimental and theoretical results, it is demonstrated that while Y and La tend to incorporate into the matrix of ZnAl₂O₄, separate CeO₂ phase could be easily formed on ZnAl₂O₄ surface, leading to a decrease in both amount and strength of the Lewis acid sites. And for the La-ZnAl₂O₄, because of the large local deformation, oxygen vacancy can be readily formed, and results in a lot acid sites in the subsurface layer available for reactions. Deactivation rates of the catalysts in isobutane dehydrogenation is found to linearly correlate with the Lewis acid amounts over the modified supports. Compared with the catalysts of Pt-Sn/ZnAl₂O₄, Pt-Sn/La-ZnAl₂O₄ and Pt-Sn/Y-ZnAl₂O₄, Pt-Sn/Ce-ZnAl₂O₄ exhibits superior catalytic performance due to the low coke contents and high Pt dispersion. These results may provide additional insights on the design and optimization of isobutane dehydrogenation catalysts by tailoring the composition and structure of oxide supports.

Full text of DOI:10.1016/j.cattod.2020.04.016

Journal Pre-proof
Insights into the Doping Effect of Rare-Earth Metal on ZnAl2O4
Supported PtSn Catalyzed Isobutane Dehydrogenation
Jianfeng Liu (Investigation) (Writing - original draft), Wei Zhou
(Validation) (Writing - original draft) (Writing - review and editing),
Dongyu Jiang (Conceptualization) (Methodology), Dong Wang
(Software)<ce:contributor-role>Data Curation), Wenhai Wu
(Visualization) (Investigation), Yue Wang (Writing - review and
editing) (Funding acquisition), Xinbin Ma (Supervision)
PII:
S0920-5861(20)30204-2
DOI:
https://doi.org/10.1016/j.cattod.2020.04.016
CATTOD 12795
Reference:
To appear in:
Catalysis Today
Received Date:
Revised Date:
Accepted Date:
20 January 2020
1 April 2020
5 April 2020
Please cite this article as: Liu J, Zhou W, Jiang D, Wang D, Wu W, Wang Y, Ma X, Insights into
the Doping Effect of Rare-Earth Metal on ZnAl2O4 Supported PtSn Catalyzed Isobutane
Dehydrogenation, Catalysis Today (2020), doi: https://doi.org/10.1016/j.cattod.2020.04.016

This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Insights into the Doping Effect of Rare-Earth Metal on
ZnAl2O4 Supported PtSn Catalyzed Isobutane
Dehydrogenation
Jianfeng Liu,†,‡ Wei Zhou,† Dongyu Jiang,*,‡ Dong Wang,‡ Wenhai Wu,‡ Yue Wang, *,† and
Xinbin Ma†
†Key Laboratory for Green Chemical Technology of Ministry of Education, Collaborative
Innovation Centre of Chemical Science and Engineering, School of Chemical Engineering
and Technology, Tianjin University, Tianjin 300072, China
‡State Key Laboratory of Green Chemical Engineering and Industrial Catalysis, Sinopec
Shanghai Research Institute of Petrochemical Technology, Shanghai 201208, China
*Corresponding author: Dr. Yue Wang, Dr. Dongyu Jiang
E-mail: yuewang@tju.edu.cn, jiangdy.sshy@sinopec.com
Fax: +86-22-87401818
Tel: +86-22-27409248
1

Graphical Abstract
Highlights
 Y and La were incorporated into ZnAl2O4 matrix, while Ce was existed on
ZnAl2O4 surface as CeO2.
 The addition of La and Ce increased Pt dispersion, thus improved the
initial conversion of isobutane.
 Deactivation rates of the catalysts in isobutane dehydrogenation is found
to linearly correlate with the Lewis acid amounts over the modified
supports.
 The addition of Y and Ce weakened the surface acidity, thus enhanced
the catalyst stability.
2

Abstract
Isobutane dehydrogenation is a vital route for the production of isobutene, an
important substance for methyl tert-butyl ether. However, the reaction is typically
performed at relatively low pressure and high temperature, resulting in a facilitated
coke formation. Here, we used rare-earth metals (Y, La, Ce) as dopants to modify the
ZnAl2O4 support and studied their effects on Pt-Sn catalyzed dehydrogenation of
isobutane. Combining the experimental and theoretical results, it is demonstrated that
while Y and La tend to incorporate into the matrix of ZnAl2O4, separate CeO2 phase
could be easily formed on ZnAl2O4 surface, leading to a decrease in both amount and
strength of the Lewis acid sites. And for the La-ZnAl2O4, because of the large local
deformation, oxygen vacancy can be readily formed, and results in a lot acid sites in
the subsurface layer available for reactions. Deactivation rates of the catalysts in
isobutane dehydrogenation is found to linearly correlate with the Lewis acid amounts
over the modified supports. Compared with the catalysts of Pt-Sn/ZnAl2O4,
Pt-Sn/La-ZnAl2O4 and Pt-Sn/Y-ZnAl2O4, Pt-Sn/Ce-ZnAl2O4 exhibits superior
catalytic performance due to the low coke contents and high Pt dispersion. These
results may provide additional insights on the design and optimization of isobutane
dehydrogenation catalysts by tailoring the composition and structure of oxide
supports.
Keywords: Dehydrogenation; Rare-earth metal; Isobutane; Lewis acidity;
PtSn/ZnAl2O4
3

1. Introduction
Isobutane dehydrogenation over supported Pt catalysts is a vital route for the
production of isobutene, an important substance for methyl tert-butyl ether (MTBE)
and ethyl tert-butyl ether (ETBE) [1-4]. Isobutane dehydrogenation is typically
performed at a relatively low pressure and high temperature (above 500 °C) because it
is highly endothermic and equilibrium-limited reaction. However, these reaction
conditions usually facilitate coke formation, resulting in a decrease of isobutane
conversion, isobutene selectivity and catalyst stability [5-7]. Therefore, it is very
important to design catalysts with high activity, high selectivity and less coke
formation for dehydrogenation reactions.
Bimetallic PtSn catalysts supported on γ-Al2O3 have been extensively studied
and used in dehydrogenation processes [8-10]. Recently, Using ZnAl2O4 and
MgAl2O4 as supports have gained increasing attention due to their low acidic
characteristics and high thermal stabilities, resulting in the improved selectivity of
alkane dehydrogenation and the reduced cracking and coke deposition. Miura et al.
[11] reported that Sn deposited Pt to form Pt-Sn, could promote activity and Sn
deposited on ZnAl2O4 could promote isobutene selectivity for the isobutane
dehydrogenation on Pt/ZnAl2O4. However, PtSn/Al2O3 and PtSn/ZnAl2O4 catalysts
are not stable during reaction [12]. It was found that the stability of supported Pt-Sn
alloy catalyst was affected by the coke transfer from Pt sites to the support.
Aguilar-Rios et al. [13] revealed that Sn could reduce the cracking reactions and
thereby improve the stability of Pt/MgAl2O4 in isobutane dehydrogenation.
4

PtSn/MgAl2O4 catalyst showed better stability than PtSn/ZnAl2O4 [14], which could
be ascribed to the different effects of the reaction-regeneration cycles on the structure
of the Pt phase. On the other hand, the co-feed of H2 was usually applied to enhance
the performance of dehydrogenation catalysts [11-17] because it could decrease the
coke deposition rates. Compared with H2, adding steam into the feed could increase
the alkane equilibrium conversion, decrease coke formation through water-gas shift
reaction and assist desorption of alkene to improve the selectivity. Moreover, it could
reduce the temperature drop of dehydrogenation reactors [18].
The supports in heterogeneous catalysts are far more than an inert platform of
active components. Different kinds of species, such as Lewis acid-base sites [19],
Brönsted acid-base sites [20], and lattice oxygen [21] on the surface of oxide supports
may be active sites of the main/side reactions or have influence on the active phases.
Rare earth elements have been used to promote catalytic performance of Pt-Sn on
Al2O3. They are usually related to the thermal stabilization of Al2O3 [22, 23] and
strong metal -rare earth interaction by which the structure and electronic properties of
the noble metal are modified [24]. It is reported that the formation and the stability of
Pt-Sn alloy are greatly influenced by Re doping [16]. The influence of support surface
properties on the catalytic performance of Pt-Sn is not yet fully understood.
This work focuses on distinct changes in the geometry structure and catalyst
acidity of ZnAl2O4 caused by doping of different metals. Surface properties and bulk
structure of the catalysts were characterized by the combinations of Density
5

Functional Theory (DFT) calculations and a series of physical and chemical
characterizations, such as H2 chemisorption, X-ray photoelectron spectroscopy (XPS),
NH3 temperature programmed desorption (NH3-TPD), pyridine-adsorbed Fourier
transform
infrared
spectroscopy
(Py-FTIR),
thermo-gravimetric
(TG),
temperature-programmed oxidation (TPO), X-ray diffraction (XRD), and Nuclear
Magnetic Resonance (NMR). The performance of Pt-Sn on the modified ZnAl2O4 was
investigated in isobutane dehydrogenation under steam atmosphere. Structure-activity
relationship of PtSn/RE-ZnAl2O4 was further studied.
2. Experimental
2.1 Catalyst preparation
ZnAl2O4 support, named as ZA in this paper, was prepared by a co-precipitation
method using Zn(NO3)2·6H2O and Al(NO3)3·9H2O as metal precursors (the mole ratio
of Al and Zn was 2:1). And pH of the liquid was controlled by NH4OH. The resulting
precipitate was dried at 120 °C overnight and then calcined at 650 °C for 8 h. RE-ZA
were prepared in the same way using Y(NO3)3·6H2O, Ce(NO3)2·6H2O and
La(NO3)2·6H2O, respectively (the mole ratio of Al and RE is 99:1). Simultaneous
impregnation was applied to prepare PtSn/ZA and PtSn/RE-ZA catalysts. These
supports were co-impregnated in H2PtCl6 and SnCl2 solutions. After drying, the
samples were calcined at 650 °C for 8 h. Pt and Sn loadings were 0.4 wt% and 1.0
wt%, respectively.
2.2 Catalyst characterization
BET surface area was measured using a Micromeritics ASAP 2020 apparatus.
6

Prior to the adsorption, the samples were pretreated at 300 °C under vacuum for 4 h.
The multi-point method and BJH method were used to calculate the BET surface area
and the pore structural data of catalysts.
X-ray diffraction (XRD) was performed on a Bruker D8 Advanced
Diffractometer to analyze the crystalline phases of the catalysts. The instrument was
equipped with a radiation of Cu Kα (λ=1.542 nm, 40 kV, 40 mA) and the
measurement was within a 2θ range of 5°-80°.
27Al solid-state MAS NMR spectra were recorded on a VARIAN VNMRS-
400WB spectrometer under cross polarization conditions. The spectra were measured
at a frequency of 104.26 MHz, a spinning rate of 8.0 kHz, and a recycle delay of 1 s
with 1 mol/LAl(NO3)3 solution as the chemical shift reference.
H2 chemisorption was applied to characterize the Pt dispersion of catalyst, which
was carried out on AutoChem II 2920. Firstly, the sample was reduced in H2 (10
mL/min) at 550 °C for 2 h. Secondly, H2 was purged by flowing Ar (99.99%, 50
ml/min) at 580 °C for 1 h. Finally, the sample was cooled to 45 °C in Ar gas. Pt
dispersion was obtained from exposed Pt atoms divided by total Pt atoms, assuming
the mole ratio of Pt/H is 1:1.
After reduction in a hydrogen flow at 500°C for 2 h, the chemical states of
reduced catalysts were analyzed by XPS on a Perkin-Elmer PHI 5000C ESCA
instrument equipped with Al Kα radiation. Before the measurement, the catalysts
were dried in pure Ar gas. The C1s peak at 284.8 eV was considered as the reference
of binding energy.
7

The surface acidity of the catalysts was determined by temperature programmed
desorption of ammonia (NH3-TPD). Catalyst sample (0.15 g) was pretreated at 600 °C
for 1 h, and then cooled to 30 °C in flowing He (30 mL/min). Then, NH3 adsorption
was performed at 100 °C for 0.5 h. Then the system was purged in He (30 mL/min) at
100 °C for 1.5 h. After that, the temperature was raised to 600 °C at a rate of
10 °C/min. The NH3 desorption signals were continuously recorded with a thermal
conductivity detector (TCD).
The Brønsted acid and Lewis acid sites were characterized by pyridine-adsorbed
Fourier transform infrared spectroscopy (Py-FTIR) on a Nicolet 380 FTIR
Spectrometer. 0.02 g catalyst was pressed into a regular wafer (R =13 mm) and then
placed in an infrared cell. Before the testing, the sample chamber was programmed to
heat up to 400 °C for 2 h and then cooled to 200 °C under vacuum, wherein a
background was taken. Then it was cooled to 30 °C and the pyridine adsorption was
performed for 30 min. After that, the sample was purged at 200 °C and the FTIR
spectra were obtained.
Coke deposited on the used catalyst was characterized using the thermal
gravimetric (TG) test and temperature programmed oxidation (TPO) experiment. For
TG analysis, 0.02 g catalyst was treated in air flow (30 mL/min) with TA-STDQ600
(TA Instruments Inc., United States) from 25 °C to 700 °C with a rate of 20 °C/min in
flowing air. For the TPO experiment, about 0.05 g sample was introduced into the
reactor and heated up to 700 °C with a rate of 10 °C /min in a 5% O2/He mixture (40
ml/min).
8

2.3 Catalytic test
Isobutane dehydrogenation reaction was carried out in a conventional quartz
tubular micro-reactor. The catalyst was previously reduced by H2 at 550 °C for 2h.
The reaction was carried out at 550 °C under atmospheric pressure. The feed gas
mixture was H2O/i-C4H10 with a molar ratio of 2.0 and the mass space velocity was
4.0 h-1. The reaction products were analyzed with an online Agilent GC-7890
equipped with an activated alumina packed column and a flame ionization detector
(FID). The isobutane conversion and isobutene selectivity are defined as the following
equation (1) and (2), where Xisobutane is the conversion of isobutane, Sisobutene is the
selectivity of isobutene.
moles of isobutane reacted
Xisobutane
=
=
100%
(1)
(2)
moles of isobutane fed
moles of isobutene formed
100%
Sisobutene
mols of all products produced (excpt H )

2
2.4 Theoretical calculations
Spin polarized density functional theory (DFT) calculations were performed
using the VASP program [25, 26]. The DFT functionality was utilized at the
Perdew-Burke-Ernzerhof (PBE) level. The project-augmented wave (PAW) method
was used to represent the core-valence electron interaction. The valence electronic
states were expanded in plane wave basis sets with energy cut-off at 450 eV [27, 28].
The ionic degrees of freedom were relaxed using the conjugate gradient scheme until
all forces on each relaxed atom were less than 0.05 eV/Å. With the above settings, the
theoretically optimized lattice constant of bulk ZnAl2O4 (a = b = c = 8.095 Å) agrees
9

well with the experimental values (a = b = c = 8.087 Å), demonstrating the feasibility
of our calculation method in dealing with this kind of systems.
For the ZnAl2O4 (110) surface, a six-Al-layer p(1×1) slab (containing 48 O, 24
Al, and 12 Zn atoms) with a vacuum thickness of 12 Å between slabs was constructed,
and a 2×1×1 k-point mesh was used during the calculation. The upmost four layers
were allowed to relax while the bottom two layers were fixed during the structural
optimization. For other oxides, their stable surfaces, namely the La2O3 (001), Y2O3
(100), CeO2 (110) [29, 30] were considered in estimating the thermal stability of
metal dopants in ZnAl2O4 lattice. It should be noted that the DFT+U method with an
effective U value of 5 eV for Ce was used in simulating the CeO2 system, as
suggested in literature reports. [31, 32]
The vacancy formation energy of metal ion (Evac,M) or lattice oxygen (Evac,O) is
defined as [33]:
Evac,M = E(surface-Mvac) + E(M-bulk) - E(surface) (3)
Evac,O = E(surface-Ovac) + 0.5E(O2) - E(surface)
(4)
where E(surface), E(surface-Mvac), E(surface-Ovac), E(M-bulk) and 0.5E(O2) are the
energies of the oxide surface, surface with a metal vacancy (Mvac), surface with an
oxygen vacancy (Ovac), one metal atom in bulk phase (M-bulk), and O2 in gas phase ,
respectively. The more positive the Evac,M (or Evac,O) is, the more difficulte it is to
create a vacancy on the surface .
The adsorption energy of molecule X (Ead,X) is defined as:
Ead,X = E(X/surface) - E(surface) - E(X)
(5)
10

where E(surface), E(X) and E(X/surface) are the energies of the oxide surface,
molecule X in the gas phase and X adsorbed on the surface, respectively. The more
negative the Ead,X is, the more strongly the X binds on surface.
3. Results and discussion
3.1. Textural properties
Textural characteristics of different catalysts are illustrated in Table S1. SBET, VP
and Dpore of PtSn/ZA catalysts are 84.9 m2/g, 0.14 cm3/g and 6.9 nm, respectively.
Different cations have different influences on the textural properties of the catalysts.
The surface area of PtSn/ZA decreases with the addition of three rare-earth metals.
But the pore diameter of the PtSn/ZA after metal doping increases in different degrees.
Compared with the PtSn/ZA sample, the pore volume of PtSn/Ce-ZA and Y-ZA are
slightly increased while that of PtSn/La-ZA is decreased slightly.
3.2 Structural properties
The thermal stability (in terms of Evac,M) either in the ZnAl2O4 lattice as dopants
or in their corresponding oxides was investigated via DFT calculations. The
separation energy (ΔEsep) was defined as transferring a doping metal atom from
ZnAl2O4 (denoted as Mdop-ZnAl2O4) to a vacancy site on the corresponding oxide
surface (Mvac-MOn), which is calculated according to the following equation.
Mdop-ZnAl2O4 + Mvac-MOn  Mvac-ZnAl2O4 + MOn
(6)
As one can deduce that the ΔEsep value equals to the Evac,M difference between
surfaces of the Mdop-ZnAl2O4 and the corresponding oxide of M ions (i.e. MOn),
11

respectively. This parameter allows us to estimate, from the view of thermodynamics,
whether Mdop prefers to stay in the ZnAl2O4 lattice (ΔEsep > 0), or separates from
ZnAl2O4 and forms an extra oxide phase (ΔEsep < 0).
Table 1 DFT calculated separation energy ΔEsep, vacancy formation energy of lattice
O2c and O3c (Evac,O2c and Evac,O3c), and adsorption energy of ammonia molecule on the
doping site (Ead,NH3) for different models.
Model
ΔEsep / eV
/
Evac,O2c / eV
3.16
Evac,O3c / eV
4.30
Ead,NH3 / eV
-1.01
ZnAl2O4
Y-ZnAl2O4
Ce-ZnAl2O4
La-ZnAl2O4
-0.25
-2.81
-0.85
1.16
3.04
0.97
3.26
4.54
2.48
-1.00
-0.66
-0.67
From Table 1, compared to Y, La dopants, the separation energy of Ce shows a
large endothermic ΔEsep of (-2.81 eV), implying the formation of CeO2 phase on
ZnAl2O4 is highly favored in thermodynamics. In addition, DFT calculations indicate
that the incorporation of Y and/or La into ZnAl2O4 might not be easy, owing to their
slightly exothermic separation energies. However, considering the well-known high
barriers in solid phase transition [34] yet small thermal driving forces, we deduced
that Y and La would remain as dopants in the ZnAl2O4 lattice, despite of being less
stable.
Geometrically, we observed distinct expansion in local surface structures after
doping of rare-earth (RE) elements (Fig.S1) as a consequence of mismatching in atom
radius. Both the two- and three-coordinated oxygen (O2c, O3c) nearby the dopant are
pushed outward along the M-O bond, leading to decreased Evac,O2c and Evac,O3c relative
to pure ZnAl2O4 surface (see Table 1). Particularly for the case of La-ZnAl2O4, all
Evac,O are ~2 eV lower than that in the pure surface, corresponding with the largest
12

local deformations among the three systems, and thus it is expected that Ovac could
readily exist on La-doped surface.
Fig.1 shows XRD patterns of different samples and the standard patterns of ZA
and three rare earth oxides. The characteristic peaks of ZA at 2θ of 31.2°, 36.8°, 44.8°,
49.1°, 55.6°, 59.3° and 65.2° are presented for all samples, indicating ZA phase is not
changed after doping.
According to the standard patterns, the strongest diffraction peak of rare earth
oxides doesn’t overlap the characteristic peaks of the ZA. However, there is no
diffraction peak of Y2O3 and La2O3 of for the Y-ZA and La-ZA, respectively.
Generally, no corresponding peak on the XRD profile could be ascribed to two
reasons: 1) highly dispersed or poorly crystallized on the support (below XRD
detection limitation); 2) incorporated into the support lattice. For the Ce-ZA, the
diffraction patterns included peaks attributed to CeO2 (111), indicating some Ce3+ are
oxidized into Ce4+ and deposits on the surface of the support during the calcination,
which is well consistent with ΔEsep data in Table 1.
Compared with the pure ZA sample, the diffraction peaks of Y-ZA, Ce-ZA and
La-ZA are slightly shifted to lower value in the order La > Ce ≈ Y, as shown in
Fig.1 inlet. It implies an expansion of the unit cell of ZA, which is in accordance with
the results of Motloung et al [35]. The ionic radius of Y3+ (0.90 Å), Ce3+ (1.06 Å) and
La3+ (1.03 Å) are bigger than that of Al3+ (0.54 Å) [36-38]. When they are
incorporated into ZA lattice and replaces Al3+, an expansion of the lattice occurs,
resulting in the increase of the cell parameter.
13

La
0.13°
Ce
Y
0.07°
0.05°
PtSn/La-ZA
PtSn/Ce-ZA
35
36
37
38
39
40
41
PtSn/Y-ZA
PtSn/ZA
ZnAl2O4 PDF#05-0669
La2O3 PDF#73-2141
CeO2 PDF#34-0394
Y2O3 PDF#41-1105
10
20
30
40
50
60
70
80
2-Theta(degree)
Fig.1 XRD patterns of the catalysts
27Al NMR spectra show that pure and doped ZA mainly display signals around 8
and 65 ppm (Fig. 2), which are assigned to Al species with octahedral (Od) and
tetrahedral symmetries (Td), respectively [39-41]. After the doping of the RE elements,
the resonance becomes broader and has a lower symmetry (La> Y> Ce), suggesting
the Al species existed in changeable micro-environments. As shown in Table S2, the
variation of Od/Td demonstrates that the number of Od (Al) sites decreases with La, Y
and Ce doping. It is indicated that more La ions are incorporated into the ZA lattice
compared to Y and Ce.
As the results above discussed, with a smaller radius, most of metallic Y and La
ions are integrated into the ZA lattice and some oxides may be highly dispersed on the
support. In the case of Ce, the ionic radius is too big to be incorporated into the lattice.
More Ce species exist on the surface as CeO2 phase, a few Ce species are integrated
into the ZA lattice.
14

Al(VI)
ZA
Y-ZA
Ce-ZA
La-ZA
Al(IV)
150
100
50
0
-50
-100
-150
Chemical Shift (ppm)
Fig.2 27Al MAS NMR spectra of the catalysts
3.3 Surface acidity analysis
DFT calculations were performed to investigate the influence of surface
modification on support acidity. The adsorption energy of NH3 (Ead, NH3) was used to
evaluated the surface acidity of modified support (listed in Table 1). It was found that
doping of Ce and La reduced the NH3 adsorption strength at the doping site
(-0.66,-0.67 Vs. -1.01 eV). CeO2 phase forms on the surface of the carrier. It can
cover some acid sites of ZnAl2O4. The surface of as-formed CeO2 phase shows an
even weaker Ead,NH3 of -0.42 eV than that of support. Consequently, a decrease in both
the amount and the strength of acid site can be predicted in the Ce-ZnAl2O4 system. In
the case of doping La, incorporation into Al framework benefits the Ovac formation
and makes more acid sites in the subsurface layer available for surface reactions (e.g.
15

Alsub, see Fig. S2). Therefore, the acid content of La-doped catalyst increased, yet the
acid strength might be reduced (Ead,NH3 = -0.67 eV) compared with pure ZnAl2O4.
For the Y-doped ZA system, the acid strength doesn’t change evidently according
to the calculated result. It is deduced that not all Y dopants are doped in the ZA
system. Highly dispersed Y spices may be existed on the surface to cover some acid
centers, resulting in the decline of the surface acidity. Two adsorption sites are
considered and the NH3 adsorption strength is generally weak on the Y2O3 surface
(Ead,NH3 = -0.66 and -0.54 eV) .
FTIR spectra of pyridine adsorption on the catalysts were carried out to
characterize the surface acidity of the support and catalyst. As shown in Fig.3, two
peaks centered at 1450 cm-1 and 1610 cm-1 are found on all samples, which are
ascribed to Lewis acid sites [42, 43]. The Brønsted acid sites with the peak at 1540
cm-1 are barely detected, which means only Lewis acid sites exist on the ZA support
and PtSn/RE-ZA catalysts. The peak area of PtSn/La-ZA is apparently larger than
those of other samples, indicating more acid sites exist on the catalyst surface.
Fig.3 Pyridine adsorbed FTIR spectra of the catalysts
16

The strength distribution of Lewis acid could be obtained from the NH3-TPD
results in Fig. 4. Three peaks at about 190°C (peak I), 270°C (peak II) and 380 °C
(peak III) are presented for all samples, which can be assigned to weak, medium and
strong acid sites, respectively [44, 45]. Gaussian deconvolution method was used in
the semi-quantitative analysis of the acid strength distribution. The results containing
total acidity and acidity strength distribution are listed in Table S3. The peak
temperature and the peak area are related to the acid strength and acid quantity,
respectively. The addition of Y and Ce significantly affects the acidity of PtSn/ZA
catalysts. When Y and Ce are introduced to PtSn/ZA, the acid strength is weakened
and the total acid sites decrease. And the PtSn/Ce-ZA has less total acid sites,
especially strong acid sites. With the La addition in the catalyst, the total amount of
acid sites increased slightly and the strong acid strength (peak III) decrease. It is
deduced that not all Y dopants are doped in the ZA system. A minority of highly
dispersed Y spices (Ead,NH3 = -0.66 and -0.54 eV) may be existed on the surface,
resulting in the decline of the surface acidity. It testifies the results of Py-FTIR and
DFT calculation.
Fig. 4 NH3-TPD of the catalysts.
17

3.4 Pt and Sn states
Pt and Sn components were impregnated on the RE-ZA support. The XRD
diffraction peaks of these species are not detected, indicating that they are highly
dispersed on the support [46]. Many studies have shown that the catalytic
performance of Pt-Sn catalysts remarkably relays on the interaction of Pt and Sn,
while the structure of Pt-Sn bimetallic is complicated and remains a debate [8-12].
Isobutane molecular mainly activates and reacts on the Pt surface [47]. The
dispersion of Pt is the important parameter to evaluate the performance of Pt-Sn
catalyst. H2 chemisorption was used to measure Pt dispersion and the results are listed
in Table 2. On PtSn/ZA catalyst, Pt dispersion is 40.2 %. It is interesting to note that
Pt dispersion and Pt surface area of PtSn/ZA catalyst increase on both La and Ce
samples. It has been suggested that the interaction between Pt and CeO2 can improve
the dispersion of Pt [48]. PtSn/La-ZA shows the highest Pt dispersion and Pt surface
area. However on PtSn/Y-ZA, the Pt metal dispersion is slightly decreased, indicating
that the addition of rare-earth metals has effect on the Pt particles.
Table 2 H2 chemisorption results for the catalysts
Pt dispersion
（%）
40.2
Pt surface
area(m2/g)
99
Pt particle size
(nm)
Samples
PtSn/ZA
PtSn/Y-ZA
PtSn/Ce-ZA
PtSn/La-ZA
2.81
39.2
97
2.88
42.3
104
2.68
44.1
108
2.56
18

(a)
(b)
Fig.5 Pt 4f (a) and Sn3d5/2 (b) XPS spectra of the catalysts.
19

The chemical states of surface species of the catalyst after reduction is
characterized by XPS. Although Pt-Sn alloy with different Pt/Sn atomic ratios on
catalyst surface might be the active phase such as PtSn, Pt3Sn, PtSn2, Pt2Sn3 [15, 16].
It could only be observed by XRD in the case of high Pt, Sn loadings (>1 wt%)
[30-32, 49]. In present work, Pt 4f7/2 was used to characterize Pt chemical state
illustrated in Fig. 5(a) and Table 4S. No deviation of the peak at ca. 71.6 eV was
detected. Pt chemical state has no change for all the samples, which is the same
results as reported even with a higher Pt loadings and Pt-Sn alloy formation [12].
The XPS spectra and fitting results of Sn3d5/2 on ZA, Y-ZA, Ce-ZA, and La-ZA
are illustrated in Fig. 5(b) and Table S4. The peaks correspond to metallic tin (Sn0
,
485.2 eV) and oxides tin (SnOx) (II, IV) can be distinguished. However, it is difficult
to discriminate between Sn2+ and Sn4+ as their binding energies are very close [50]. It
has been reported that both metallic state and non-metallic state of Sn were suggested
as the active spieces [11, 12, 51]. As shown in Table S4, the peak area ratio of metallic
tin on PtSn/ZA, PtSn/Y-ZA, PtSn/Ce-ZA and PtSn/La-ZA catalysts are 29.2%, 33.7%,
28.6% and 26.9%, respectively. Metal Y doping decreases the SnOx content and La,
Ce doping increases the SnOx content. Notably, the Pt dispersion were enhanced with
the increase of amount of SnOx, which may divide the Pt ensembles into small
clusters [52-54].
3.5 Catalytic activity
20

Fig.6 (a) Isobutane conversion and (b) isobutene selectivity of the catalysts.
Fig.7 Correlations between (a) isobutane conversion and Pt surface area; (b) the
relationship of conversion decreasing rates and the total acid sites.
Fig. 6 shows the reaction results of the catalysts which possess different Pt
surface area. On PtSn/ZA, the initial and final conversion of isobutane is 50.6% and
31.0% (Fig.6a). For PtSn/Ce-ZA and PtSn/La-ZA, isobutane conversion can reach
52.7% and 56.5%, respectively. In the case of PtSn/Y-ZA, the initial conversion is
only 48.1%. The variations of initial conversion could be well correlated with the
dispersion of Pt. Fig. 7a shows the linear relationship between initial activity and Pt
surface area. Isobutane initial conversion raises with the increase of Pt dispersion,
suggesting the Pt surface area plays a key role in determining the catalytic
21

performance of PtSn/RE-ZA catalysts.
According to Fig. 9b, the selectivity of isobutene on PtSn/ZA increases from
93.7% to 95.4% after 8 h. And the selectivity is increased with the addition of Ce and
Y, because the PtSn/RE-ZA catalysts have an important density of hydrogenolysis and
cracking sites. These sites are quickly poisoned by coke disposition, which can
decrease side reaction, thus improving the selectivity of isobutene [55]. The highest
isobutene selectivity can reach 96% after 8 h reaction.
Meanwhile, the stability of the catalyst is enhanced after the introduction of Ce
and Y. PtSn/Ce-ZA has the best stability, i.e. the final conversion is still > 43% even
after 8 h. However, the addition of La leads to fast decrease of the catalytic activity.
Fig. 7b shows the relationship between decline rate of conversion and total acid sites.
It was found that the stability of PtSn/RE-ZA catalyst drops with the acid content
increases. This phenomenon can be attributed to more amounts of acid sites on the
catalyst surface. As the reaction time prolonged, Pt sites can be blocked more easily,
resulting in the higher activity drop. Conversely, better stability can be attained due to
less acidity sites on PtSn/Ce-ZA surface. Besides, SnOx can keep the Pt sites clean
from coke deposits (‘drain-off’ effect) thus increasing the catalytic stability [60].
More Sn species existing as SnOx state on PtSn/Ce-ZA is another reason for the
enhanced stability.
Continuous coke deposition on the catalyst surface during reaction causes
catalyst deactivation [56]. The typical TPO profiles of four used catalysts are shown
in Fig. 8, in which three successive peaks representing different carbon deposits are
22

observed. The carbon deposits for the first peak at about 410 °C are mainly poorly
polymerized ones that cover the Pt metal, while the peaks at higher temperature about
475 °C and 525 °C represent the highly polymerized coke located on the support
surface [57].
As listed in Table S5, with the Ce and Y addition, the coke formation decreases
from 1.48 wt% to 1.34 wt% and 1.37 wt% respectively. However, when La is added,
the coke amount increases up to 2.32 wt%. It is obvious that the main combustion
peak of PtSn/Ce-ZA is the lowest. The lower acidity of PtSn/Ce-ZA and PtSn/Y-ZA
relieve the coke formation and improve the selectivity. In addition, the Ce addition
evidently suppress the coke formation process. Vu et al. [58, 59] also reported the
Pt-Sn/CeO2-Al2O3 had higher oxygen mobility than Pt-Sn/Al2O3, resulting in a lower
coke combustion temperature during the regeneration process.
Fig. 8 TPO profiles of used catalysts
23

4. Conclusions
Modification of ZnAl2O4 using rare earth elements was investigated for Pt-Sn
catalyzed isobutane dehydrogenation. A combination of DFT calculations and XRD,
NMR revealed that Y and La could be successfully incorporated into ZnAl2O4 matrix,
while Ce constituted a separate phase on ZnAl2O4 support as the form of CeO2. The
formed CeO2 covers some Lewis acid sites and leads to a decrease in both amount and
strength of the acid sites. Due to the largest local deformation of La-ZnAl2O4，the Ovac
readily forms and makes more acid sites in the subsurface layer available for surface
reactions. PtSn/La-ZnAl2O4 exhibited a higher surface acidity, which was detrimental
to the reaction. Deactivation rate of the catalysts during isobutane dehydrogenation
was found to be proportional to the number of surface Lewis acid sites. Compared
with Pt-Sn/ZnAl2O4, Pt-Sn/La-ZnAl2O4 and Pt-Sn/Y-ZnAl2O4, Pt-Sn/Ce-ZnAl2O4
showed superior catalytic performance due to the lower coke contents and high Pt
dispersion.
Declaration-of-competing-interests
None
Acknowledgement
We gratefully acknowledge support from Tianjin Natural Science Foundation
(18JCQNJC06100) and National Postdoctoral Program for Innovative Talents of
China (BX20180211).
24

CRediT authorship contribution statement
Jianfeng Liu: Investigation, Original draft preparation
Wei Zhou: Validation, Original draft editing
Dongyu Jiang: Conceptualization, Methodology
Dong Wang: Software, Data Curation
Wenhai Wu: Visualization, Investigation
Yue Wang: Writing- Reviewing and Editing, Funding acquisition
Xinbin Ma: Supervision
References
[1] S. De Rossi, G. Ferraris, S. Fremiotti, V. Indovina, A. Cimino, Appl. Catal. A:
Gen.106 (1993) 125-141.
[2] C. Wei, F. Xue, C. Miao, Y. Yue, W. Yang, W. Hua, Z. Gao, Chin. J. Chem. 35
(2017) 1619-1627.
[3] S.M. Stagg, C.A. Querini, W.E. Alvarez, D.E. Resasco, J. Catal. 168 (1997) 75-94.
[4] M. Farsi, A. Jahanmiri, M.R. Rahimpour, J. Energy Chem. 22 (2013) 633-638.
[5] G.J. Siri, M.L. Casella, G.F. Santori, O.A. Ferretti, Ind. Eng. Chem. Res. 36 (1997)
4821-4826.
[6] L. Xu, Z. Wang, H. Song, L. Chou, Catal. Commun 35 (2013) 76-81.
[7] R.D. Cortright, J.M. Hill, J.A. Dumesic, Catal. Today 55 (2000) 213-223.
[8] S.A. Bocanegra, S.R. de Miguel, I. Borbath, J.L. Margitfalvi, O.A. Scelza, J. Mol.
Catal. A-Chem. 301 (2009) 52-60.
25

[9] B. Barghi, M. Fattahi, F. Khorasheh, Pet. Sci. Technol. 32 (2014) 1139-1149.
[10] F.T. Zangeneh, S. Sahebdelfar, M. Bahmani, Chin. J. Chem. Eng. 21 (2013)
730-735.
[11] H. Miura, T. Itoh, React. Kinet. Catal. Lett. 66 (1999) 189-194.
[12] B.K. Vu, M.B. Song, I.Y. Ahn, Y.W. Suh, D.J. Suh, W.I. Kim, H.L. Koh, Y.G.
Choi, E.W. Shin, Appl. Catal. A-Gen. 400 (2011) 25-33.
[13] G. Aguilar-Rios, P. Salas, M.A. Valenzuela, H. Armendariz, J.A. Wang, J.
Salmones, Catal. Lett. 60 (1999) 21-25.
[14] S.A. Bocanegra, A. Guerrero-Ruiz, S.R. de Miguel, O.A. Scelza, Appl. Catal.
A-Gen. 277 (2004) 11-22.
[15] Y. Zhang, Y. Zhou, J. Shi, S. Zhou, Z. Zhang, S. Zhang, M. Guo, Fuel Process.
Technol. 111 (2013) 94-104.
[16] B.K. Vu, M.B. Song, I.Y. Ahn, Y.W. Suh, D.J. Suh, W.-I. Kim, H.L. Koh, Y.G.
Choi, E.W. Shin, Catal. Today 164 (2011) 214-220.
[17] Z. Ma, J. Wang, J. Li, N. Wang, C. An, L. Sun, Fuel Process. Technol. 128 (2014)
283-288.
[18] Y. Shan, Y. Zhu, Z. Sui, D. Chen, X. Zhou, Catal. Sci. Technol. 5 (2015)
3991-4000.
[19] J. Sun, R.A.L. Baylon, C. Liu, Mei D., K. J. Martin, P. Venkitasubramanian Y.
Wang, J. Am. Chem. Soc. 138 (2016) 507-517.
[20] J. Sun, K. Zhu, F.Gao, C. Wang, J. Liu, C. H. F. Peden, Y. Wang, J. Am. Chem.
Soc. 133 (2011) 11096-11099.
26

[21] T. Kiyokawa, N. Ikenaga, Appl. Catal. A: Gen. 536 (2017) 97-103.
[22] A. Trovarelli, Catal. Rev. 38 (1996) 439-520.
[23] A. Piras, A. Trovarelli, G. Dolcetti, Appl. Catal. B: Environ. 28 (2000) 77-81.
[24] G. Del Angel, A. Bonilla, Y. Peña, J. Navarrete, J.L.G. Fierro, D.R. Acosta, J.
Catal. 219 (2003) 63-73
[25] G. Kresse, J. Furthmüller, Comp. Mater. Sci. 6 (1996) 15-50.
[26] G. Kresse, J. Hafner, Phys. Rev. B 49 (1994) 14251-14269.
[27] D. Wang, Z.P. Liu, W.M. Yang, ACS Catal. 8 (2018) 7270-7278.
[28] H.Yuan, J. Chen, H. Wang, P. Hu, ACS Catal. 8 (2018) 10864-10870.
[29] D. Wang, J. Jiang, H.F. Wang, P. Hu, ACS Catal. 6 (2015) 733-741.
[30] Y. Hou, D. Wang, X. H. Yang, W. Q. Fang, B. Zhang, H. F. Wang, G. Z. Lu, P. Hu,
H. J. Zhao, H. G. Yang, Nat. Commun. 4 (2013) 1583-1590.
[31] H.F. Wang, Y.L. Guo, G.Z. Lu, P. Hu, Angew. Chem. Int. Ed. 48 (2009)
8289-8292.
[32] J. Xing, H.F. Wang, C. Yang , D. Wang, H.J. Zhao, G.Z. Lu, P. Hu, H.G. Yang,
Angew. Chem. Int. Ed. 51 (2012) 3611-3615.
[33] J.S. Elias, K.A. Stoerzinger, W.T. Hong, M. Risch , L. Giordano, A.N. Mansour,
Y. Shao-Horn, ACS Catal. 7 (2017) 6843-6857.
[34] S.H. Tolbert, A.P. Alivisatos , Sci. 265 (1994) 373-376.
[35] S.V. Motloung, K.G. Tshabalala, R.E. Kroon, T.T. Hlatshwayo, M. Mlambo, S.
Mpelane, J. Mol. Struct. 1175 (2019) 241-252.
[36] M.A. Gomes, A.C. Brandão-Silva, J.F.M. Avila, M.A.R.C. Alencar,J. J.
27

Rodrigues, Z.S. Macedo, J. Lumin. 200 (2018) 43-49.
[37] B.P. Gangwar, S.C. Maiti, S. Sharma, J. Solid State Chem. 256 (2017) 109-115.
[38] Y. Xue, R. Liu, Y. Liu, H. Xu, G. Chen, X. Zhang, W. Gao, Z. Jiang, J. Rare Earth
2018. In press.
[39] C. E. A. Kirschhock, R. Ravishankar, F. Verspeurt, P. J. Grobet,P. A. Jacobs and J.
A. Martens, J. Phys. Chem. B 103 (1999) 4960-4964.
[40] E. J. M. Hensen, Q. Zhu, R. A. J. Janssen, P. C. M. M. Magusin, P. J. Kooyman
and R. A. V. Santen, J. Catal. 233 (2005) 123-135.
[41] O. H. Han, C.-S. Kim and S. B. Hong, Angew. Chem., Int. Ed. 41 (2002)
469-472.
[42] J. Lavalley, M. Daturi, V. Montouillout, G. Clet, C.O. Areán, M.R. Delgado, A.
Sahibed-Dine, Phys. Chem. Chem. Phys., 5 (2003) 1301-1305.
[43] A. Vimont, J. Lavalley, A. Sahibed-Dine, C. Otero Areán, M. Rodríguez Delgado,
M. Daturi, J. Phys. Chem. B, 109 (2005) 9656-9664.
[44] P. Iengo, M. Di Serio, V. Solinas, D. Gazzoli, G. Salvio, E. Santacesaria, Appl.
Catal.A -Gen. 170 (1998) 225-244.
[45] H. Seo, J.K. Lee, U.G. Hong, G. Park, Y. Yoo, J. Lee, H. Chang, I.K. Song, Catal.
Commun., 47 (2014) 22-27.
[46] P. Iengo, M. Di Serio, A. Sorrentino, V. Solinas, E. Santacesaria, Appl. Catal.
A-Gen. 167 (1998) 85-101.
[47] L. Lin, W. Yang, J. Jia, Z. Xu, T. Zhang, Y. Fan, Y. Kou, J. Shen, Sci. China Ser.
B, 42 (1999) 571-580.
28