Effect of calcination temperature and fluorination treatment on NiF2-AlF3 catalysts for dehydrofluorination of 1, 1, 1, 2-tetrafluoroethane to synthesize trifluoroethylene

Article abstract of DOI:10.1016/j.apcata.2018.12.001

High-performance NiF₂-AlF₃ fluoride catalysts for the catalytic dehydrofluorination of 1, 1, 1, 2-tetrafluoroethane (CF₃CH₂F) were prepared by impregnation and fluorination methods. The effect of calcination temperature and vapor-phase fluorination on the properties of NiF₂-AlF₃ catalysts were investigated by BET, SEM, XRD, UV-DRS, Raman, IR, TG and XPS. By increasing the calcination temperature, the NiO species diffused from the surface into the inside bulk phase of the alumina support and the inverse spinel NiAl₂O₄ was formed at a calcination temperature up to 600 °C. Vapor-phase fluorination can improve the stability of the catalytic dehydrofluorination. The unfluorinated NiAl₂O₄ affected the surface area and acidity of NiF₂-AlF₃ catalyst. The acid sites of catalysts were investigated by py-IR, disclosing that the affinity of Lewis acid sites toward activity of catalysts. In addition, it was found that the weak and medium Lewis acid sites derived from NiF₂-AlF₃ complex phase are active centers for catalyzing dehydrofluorination of CF₃CH₂F to trifluoroethylene (CF₂=CHF). The highest activity was obtained over a fluorinated catalyst calcined at 500 °C, with a reaction rate of 2.13 mmol min^?1 g_cat.^?1 and a trifluoroethylene selectivity of 99%, highlighting a good prospect for the commercial application.

Full text of DOI:10.1016/j.apcata.2018.12.001

Accepted Manuscript
Title: Effect of calcination temperature and fluorination
treatment on NiF2-AlF3 catalysts for dehydrofluorination of 1,
1, 1, 2-tetrafluoroethane to synthesize trifluoroethylene
Authors: Wenzhi Jia, Yanfeng Chen, Min Liu, Xing Liu,
Xinhua Liu, Juanjuan Yuan, Xiaoju Lu, Zhirong Zhu
PII:
S0926-860X(18)30590-8
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.12.001
APCATA 16904
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
6 June 2018
19 November 2018
1 December 2018
Please cite this article as: Jia W, Chen Y, Liu M, Liu X, Liu X, Yuan J, Lu X, Zhu
Z, Effect of calcination temperature and fluorination treatment on NiF2-AlF3 catalysts
for dehydrofluorination of 1, 1, 1, 2-tetrafluoroethane to synthesize trifluoroethylene,
Applied Catalysis A, General (2018), https://doi.org/10.1016/j.apcata.2018.12.001
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. 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.

Effect of calcination temperature and fluorination treatment on NiF2-AlF3
catalysts for dehydrofluorination of 1, 1, 1, 2-tetrafluoroethane to
synthesize trifluoroethylene
Wenzhi Jiaa#, Yanfeng Chenb#, Min Liuc#, Xing Liud#, Xinhua Liue, Juanjuan Yuand,
Xiaoju Lua, Zhirong Zhud*
a
School of Chemistry and Chemical Engineering, Hubei Polytechnic University,
435003, Hubei, China
b Shanghai 3F New Material Co., Ltd, Shanghai 200092, China
c Department of Chemical Engineering, The University of Melbourne, Parkville, 3052
d
School of Chemical Science and Engineering, Tongji University, Shanghai 200092,
China
e Dyson School of Design Engineering, Imperial College London, London SW7 2AZ,
UK
* Corresponding author. Tel./fax: +86-21-65982563.
E-mail: zhuzhirong@tongji.edu.cn (Z.R Zhu)
#These authors contributed equally to this work.

Graphic Abstract
NiAlF-5
NiAlF-6
2.0
1.5
1.0
0.5
0.0
NiAlF-7
NiAlF-4
0.0
0.2
0.4
0.6
0.8
1.0
W-M Lewis acid sites / mmol g-1
It suggested that weak and medium Lewis acid are mainly active sites of catalysts for
dehydrofluorination of CF3CFH2 to CF2=CHF, the strong Lewis acid sites resulted in
the carbon depositon. The highest activity was obtained over a fluorinated catalyst
calcined at 500 ◦C, with 2.13 mmol min-1 g-1 of the reaction rate and selectivity of
over 99 %.
Highlights



Calcination temperature and fluorination treatment have impact on catalysts
properties
o
o
The NiF2-AlF3 catalyst calcined at 500 C showed 20 % yield of TrFE at 450 C
during 300 h
The weak and medium (weak&medium) Lewis acid were the active site for the
catalytic dehydrofluorination of CF3CH2F
2

Abstract: High-performance NiF2-AlF3 fluoride catalysts for the catalytic
dehydrofluorination of 1, 1, 1, 2-tetrafluoroethane (CF3CH2F) were prepared by
impregnation and fluorination methods. The effect of calcination temperature and
vapor-phase fluorination on the properties of NiF2-AlF3 catalysts were investigated by
BET, SEM, XRD, UV-DRS, Raman, IR, TG and XPS. By increasing the calcination
temperature, the NiO species diffused from the surface into the inside bulk phase of
the alumina support and the inverse spinel NiAl2O4 was formed at a calcination
o
temperature up to 600 C. Vapor-phase fluorination can improve the stability of the
catalytic dehydrofluorination. The unfluorinated NiAl2O4 affected the surface area
and acidity of NiF2-AlF3 catalyst. The acid sites of catalysts were investigated by
py-IR, disclosing that the affinity of Lewis acid sites toward activity of catalysts. In
addition, it was found that the weak and medium Lewis acid sites derived from
NiF2-AlF3 complex phase are active centers for catalyzing dehydrofluorination of
CF3CH2F to trifluoroethylene (CF2=CHF). The highest activity was obtained over a
o
-1
fluorinated catalyst calcined at 500 C, with a reaction rate of 2.13 mmol min-1 gcat.
and a trifluoroethylene selectivity of 99%, highlighting a good prospect for the
commercial application.
Keywords: 1, 1, 1, 2-tetrafluoroethane, Trifluoroethylene, Dehydrofluorination,
NiF2-AlF3 catalysts, Weak and medium Lewis acid sites
1. Introduction
1, 1, 1, 2-tetrafluoroethane (CF3CH2F, HFC-134a) was considered as a promising
alternative to dichlorodifluoromethane (CF2Cl2) due to its zero ozone depletion
potential (ODP), and has been widely used in the refrigeration industry [1]. However,
according to the Kyoto Protocol, CF3CFH2 with high Global Warming Potential
(GWP, > 1,300) contributes greatly to the greenhouse effect and should be largely
phased out [2]. CF3CFH2 has been replaced eventually by the 2, 3, 3,
3-tetrafluoro-propene (HFO-1234yf). Consequently, the transformation of CF3CFH2
are attracting increasing attention [3].
Numerous methods for the dispose of CF3CFH2 have been reported, such as
combustion, photodissociation, water plasmas and catalytic dehydrofluorination
3

[4,5,6,7]. Among these methods, the dehydrofluorination of CF3CFH2 to
trifluoroethylene (CF2=CHF) is a promising process due to the high additional value
of CF2=CHF for fluoropolymer. The routes to catalytic dehydrofluorination of
hydrofluorocarbons are normally used Mg2P2O7 and AlF3, Fe2O3-CdO-Al2O3 and
Pd/AlF3 as catalysts [8, 9, 10, 11, 12]. Teinz et al found that AlF3 was exclusively
selective toward the dehydrohalogenation of 3-chloro-1, 1, 1, 3-tetrafluorobutane,
owing to the presence of strong Lewis acid sites [13]. Okazaki et al disclosed that the
product distribution was dependent on the acid strength in the CF3CH3
dehydrofluorination over AlF3 catalyst [8]. In addition, Li et al reported that the
strength effect of Lewis acid sites of phosphate catalysts, and found that the weak
Lewis acid sites were the only active sites for the dehydrofluorination of CF3CH3 to
CF2=CH2, while the strong Lewis acid sties resulting in the formation of undesirable
coke [9, 10]. A previous work reported by our group found that the NiO doped Al2O3
catalyst was a good candidate for the dehydrofluorination of CF3CFH2 to CF2=CHF
[14]. These investigations reveal that the catalytic dehydrofluorination performance
significantly depend on the acidity of the applied catalysts. To date, the effect of
Lewis acid sites distribution and activity of NiF2-AlF3 catalysts for
dehydrofluorination of CF3CFH2 are not clearly investigated.
Herein, NiF2-AlF3 catalysts were prepared and characterized to investigate the
relationship between acid sites and catalytic activity for the dehydrofluorination of
CF3CFH2 to CF2CHF. BET, SEM, XRD, UV-DRS, Raman and XPS characterization
techniques were employed to disclose the effect of calcination temperature and
fluorination on the structure and acidity of the resultant NiF2-AlF3 catalysts. The
surface acid sites of the obtained catalysts were characterized by Py-IR. The Raman
and TG techniques were used to analyse the carbon deposition during the catalytic
process. The evolution of acid active sites of NiF2-AlF3 catalyst correlated to the
catalytic performance for dehydrofluorination reaction was also discussed.
2. Experimental
2.1 Catalyst preparation
The Al2O3 supports were impregnated with aqueous solution of Ni(NO3)2·6H2O
and dried at 100 °C overnight. The impregnated supports were calcined at 400, 500,
600 and 700 °C for 3 h in air, respectively. The resultant samples were denoted as
NiAlO-4, NiAlO-5, NiAlO-6 and NiAlO-7. The molar ratio of Al/Ni was 9. Prior to
4

use, the prepared NiAlO-x catalysts were subject to the fluorination treatment. The
fluorination was carried out in a stainless-steel tubular reactor with a diameter of 3 cm
and a length of 40 cm. Specially, 2 g of the prepared NiAlO-x was loaded into the
reactor and dried at 260 °C for 1 h. Then, the NiAlO-x was fluorinated at 260 °C and
400 oC for 1 h and 3 h under the mixed gas atmosphere of HF and N2 ( HF/N2 = 4 ),
respectively. The finally obtained catalysts were denoted as NiAlF-4, NiAlF-5,
NiAlF-6 and NiAlF-7, respectively.
2.2 Characterization
Surface areas of the catalysts were determined by the modified BET method
from the N2 sorption isotherms at 77 K on an automatic NOVA 2200e apparatus.
SEM (HitachiS-4700) was taken on a Hitachi S-4800, catalyst power was directly
imbedded in conductive tapes, with an acceleration voltage of 20 kV. XRD patterns
were collected with a Focus D8 powder diffractometer (40 kV, Cu-Kα). The pattern
for structure refinement was taken in the 2θ range of 10-70o with a step width of 0.02o.
Ultraviolet–visible spectra were collected in the range of 200-800 nm using a
Varian Cary 5G UV–vis spectrophotometer working at reflectance diffuse mode
(DRS). Raman spectra were collected on a Renishaw RM1000 confocal microscope
with an exciting wavelength of 514 nm. X-ray photoelectron spectroscopy (XPS)
experiments were carried on an RBD upgraded PHI-5000C ESCA system (Perkin
Elmer) with a Mg K radiation (h=1253.6 eV). The X-ray anode was run at 250W
and the high voltage was kept at 14.0 kV with a detection angle of 54. The pass
energy was fixed at 46.95 or 93.90 eV to ensure sufficient sensitivity. The base
pressure of the analyzer chamber was about 510-8 Pa. Binding energies were
calibrated by using the containment carbon (C1s = 284.6 eV). TG experiments were
performed using a Netzsch STA 449C instrument.
o
Py-IR was conducted at 200 and 400 C under a high vacuum. IR spectrum on a
self-supported wafer of the sample was collected in the spectral range of 2000–1200
cm-1 with 64 scans.
2.3 Dehydrofluorination of CF3CFH2
The dehydrofluorination of CF3CFH2 was carried out in a continuous-flow
fixed-bed metal reactor (15 mm i.d.) and operated in the down flow mode at
atmospheric pressure. The reaction temperature was determined with a thermocouple
in contact with 2 g catalyst bed. The flow rates of the gas CF3CFH2 and N2 were
5

carefully regulated by mass flow controllers. The molar ratio of N2/CF3CFH2 was 10
with a total flow rate of 22.5 ml min-1 and the gas hourly space velocity (GHSV) was
675 h-1. To remove the product of HF, the reaction effluent was passed through an
aqueous KOH solution. Then, it was analyzed by a gas chromatograph (HP-5890)
using a HP-PLOT/Al2O3 column (I. D. 0.53 mm × 50 m) with an FID detector.
3. Results and Discussion
3.1 Structural characterization
Table 1 lists the analytical results of the surface areas and composition of the
prepared NiAlO and NiAlF catalysts. The surface area of these catalysts decreased as
increasing the calcination temperature. For NiAlF catalyst, its surface area was much
smaller than corresponding NiAlO catalyst. For example, the NiAlO-5 catalyst
calcined at 500 oC had a surface area of 171 m2 g-1, while only 72 m2 g−1 was obtained
on the NiAlF-5 catalyst. The surface element composition of the prepared catalysts
was obtained by XPS characterization (Table 1). The amount of carbon was detected
by ex-situ XPS measurement. This is because the coverage of a few organics or
carbon on the surface of the samples in air. After the fluorination treatment, the
surface oxygen concentration of NiAlO decreased sharply and the corresponding
fluorine concentration of the resultant NiAlF catalysts significantly increased.
NiAlF-5 has much lower surface oxygen concentration (1.52 %) than those of
NiAlF-4 (20.94%), NiAlF-6 (4.76%) and NiAlF-7 (5.84%), resulting in the highest
fluorine concentration. The oxygen removal from the NiAlO catalysts was via the
pre-fluorination process:
Al2O3 + 6HF  2AlF3 + 3H2O
NiO + 2HF  NiF2 + H2O
(1)
(2)
However, NiAlO-4 catalyst presented much lower fluorination degree than other
samples under the same fluorination conditions. Such relatively lower fluorination
degree is probably due to the surface state NiO, which covered catalyst surface, and
result in the formation of thin NiF2 film after fluorination treatment [34, 35].The
formed NiF2 film restrained the further fluorination process.
Table 1. Surface area and surface composition analysis by XPS of the NiAlO
and NiAlF catalysts.
Element / %
S
/ m2 g-1
6

temperature
/ oC
NiAlO
Al
NiAlF
O
NiAlO NiAlF
Ni
O
C
Ni
Al
F
C
400
500
600
700
175
171
149
115
115
72
2.03 15.8 39.28 42.89
2.62 16.9 38.46 42.02
1.92 14.2 41.57 42.31
1.74 19.8 38.32 40.14
4.94 11.88 20.94 11.87 50.37
5.96 17.13 1.52
3.24 14.38 4.76
4.59 13.69 5.84
29.75 45.64
25.09 51.53
26.07 49.81
35
25
Fig. 1 shows the N2 adsorption isotherms at 77 K and pore size distribution of the
NiAlO-5 and NiAlF-5 catalysts. The NiAlO-5 show a type IV isotherm with an
obvious hysteresis loop, while the NiAlF-5 presents a type II isotherm, according to
the IUPAC classification. Type H3 attributed to the NiAlO-5 catalyst, and its pore
size is centered on 10 and 15 nm (Fig. 1 inset). The presence of disordered mesopores
in NiAlO-5 can be contributed to the gaps between catalyst particles. However, the
shape of hysteresis loop is a superposition of H3 and H4 in the NiAlF-5 catalyst [15],
and two kinds of pore sizes are centred on 3 and 20 nm. Overall, these results indicate
that the fluorination treatment significantly changed the structure and physical
properties of the NiAlO precursors.
450
400
NiAlO-5
NiAlF-5
350
0.04
300
0.02
250
200
0.00
0
10 20 30 40 50 60
150
100
50
D / nm
0
0.0
0.2
0.4
0.6
0.8
1.0
P / P0
Fig. 1 N2 adsorption isotherms at 77 K and pore size distributions of the
catalysts (inset).
7

o
The microstructures of the representative catalysts calcined at 500 C are shown
in Fig. 2. As seen from Fig. 2a), the NiAlO-5 shows partial slit-shape and the surface
porous structure, might be covered with the few disordered and flocculated clusters.
Moreover, the morphology properties of NiAlF-5 are shown in Fig. 2b). From Fig. 2b)
the surface floccule disappeared and the partial bulk became smooth. The less
slit-shape and the surface pores led to decreases in surface area and pore volume of
the NiAlF-5 catalyst.
Fig. 2 SEM images of a) NiAlO-5 and b) NiAlF-5 catalysts.
Fig. 3 shows XRD patterns of the NiAlO and NiAlF catalysts. Only diffraction
peak due to γ-Al2O3 for the individual NiAlO catalyst are detected in Fig. 3a. All of
NiAlO precursors only shows the typical γ-Al2O3 diffraction peaks and no
characteristic peak of the Ni species were detected, suggesting that small grains of
nickel oxide are below detection limit of XRD measurements [16]. Two characteristic
peaks due to the formation of AlF3 (PDF#: 44-0231) and NiF2 (PDF#: 76-0605) were
observed in the NiAlF catalysts (Fig. 3b). These fluorides were derived from the
hydrofluorination of corresponding metal oxide. For the NiAlF-7 catalyst, a main
peak at 27 o attributed to crystalline NiF2 is weaker than other NiAlF catalysts, which
indicates that a relatively lower concentration of the NiF2 presents in the NiAlF
catalyst with calcination temperature of 700 oC.
8

a)
b)
AlF3
 NiF2
Al2O3





NiAlO-7
NiAlO-6




NiAlF-7



NiAlF-6
NiAlF-5
NiAlF-4
NiAlO-5
NiAlO-4
10
20
30
40
50
60
70
10
20
30
40
50
60
70
2 Theta / o
2 Theta / o
Fig. 3 XRD patterns of a) the NiAlO and b) NiAlF catalysts.
UV-DRS results at the band ranging from 200 to 800 nm were recorded for the
NiAlO and NiAlF catalysts(Fig. 4). Due to the absence of the band gap absorption of
the white Al2O3 or AlF3, the band-gap absorption mainly originated from Ni species
for both the NiAlO or NiAlF catalysts [17]. As shown in Fig. 4a, the band in the first
region of 220-320 nm is attributed to the transition characteristic of the tetrahedral
Ni2+ ion in NiO, whereas the absorbance at 345 nm can be related to d-d transitions of
bulk NiO [18]. Preda et al. classified NiO in two states (surface and bulk)deposited on
highly ordered pyrolitic graphite (HOPG), according to an in situ XPS and ex-situ
atomic force microscopy (AFM) studies on the growth of NiO [19]. The band at 345
nm gradually became strong as increasing the calcination temperature, indicating the
growth of the bulk NiO species. For the NiAlO-7 catalyst, a broad absorption peak at
ca. 510 nm is assigned to υ2 (3A2g3T1g) transitions of Ni2+ ions in an octahedral
symmetry [20]. The absorbance band at 614 nm is assigned to Ni2+ ions in tetrahedral
coordination, deriving from the spinel-like structure of the NiAl2O4 [21, 22, 23]. The
formation of NiAl2O4 was ascribed to the solid-state reaction between NiO and Al2O3
at elevated temperature [24, 25]. As can be seen in Fig. 4b, the bands in the first
region of 220-340 nm and the weak band at 490 nm are both associated with Ni (II)
species (NiF2 and/or NiO). The characteristic bands of NiF2 and NiO are difficult to
be distinguished by UV-DRS. Besides, the weak band at 610 nm is attributed to the
NiAl2O4 in NiAlF-7 catalyst, indicating that NiAl2O4 was difficult to be fluorinated
into NiF2 and AlF3 at high calcination temperature.
9

b) NiAlF
a) NiAlO
NiAlF-7
NiAlF-6
NiAlO-7
NiAlO-6
NiAlF-5
NiAlF-4
NiAlO-5
NiAlO-4
200 300 400 500 600 700 800
200 300 400 500 600 700 800
Wavelength / nm
Wavelength / nm
Fig. 4 UV-diffuse reflection spectra of the a) NiAlO and b) NiAlF catalysts.
Raman experiments were conducted to further investigate the Ni species of the
NiAlO and NiAlF catalysts. Fig. 5 shows Raman spectra of the NiAlO (a) and NiAlF
(b) catalysts. No Raman band is observed for all the NiAlO due to the presence of
γ-Al2O3, which agrees well with the previous report [26]. A broad band at 548 cm-1 is
attributed to the longitudinal optical (1LO) vibration modes of the surface NiO
(dispersed on surface of supports) (Fig. 5a) [24, 27]. In-situ Laser Raman of NiO were
investigated by Chan and Wachs in detail, it can support that nickel oxide on Al2O3 at
room temperature exhibits a Raman band at ~550 cm-1 [25]. In addition, Zhang et al
reported the Raman band at 546 cm-1 was due to the cubic NiO singer-crystals [16], in
terms of Raman characterization of the metal-oxide-modified NiO/Al2O3 catalysts for
the methanation of CO. Their findings about Raman band of NiO are similar to our
results. Furthermore, the intensity of broad Raman peak at 548 cm-1 gradually
declined as increasing the calcination temperature, implying that the amount of
surface NiO decreased. However, for NiAlO-6 and NiAlO-7 catalysts, the weak broad
peak at 805 cm-1 attributed to the TO + LO mode of small bulk NiO [28], became
stronger as increasing the calcination temperature. In addition, two weak bands at 375
and 615 cm-1 corresponding to the NiAl2O4 were found in NiAlO-7 [29]. After the
fluorination of the NiAlO, γ-Al2O3 was transformed into the crystalline AlF3 (Fig. 3b).
However, no characteristic Raman peak was observed for AlF3 (Fig. 5b). The four
10

main bands at 150, 200, 250 and 300 cm-1 are attributed to the rutile-structure
antiferromagnets NiF2 [31, 31]. As for the NiAlF-7 catalyst, the weak band at 810 nm
is attributed to the NiAl2O4 phase, further indicating the difficulty in pre-fluorination
of NiAl2O4 species, in agreement with the UV-DRS results (Fig. 4b).
b) NiAlF
a) NiAlO
NiAlF-7
NiAlO-7
NiAlF-6
NiAlF-5
NiAlF-4
NiAlO-6
NiAlO-5
NiAlO-4
300
600
900
1200
1500
200
400
600
800
Raman shift / cm-1
Raman shift / cm-1
Fig. 5 Raman spectra of the a) NiAlO and b) NiAlF catalysts.
Overall, the UV-DRS (Fig. 4a) and Raman spectra (Fig. 5a) collectively
confirmed the presence of the surface and bulk phase NiO on the Al2O3 support.
There is no UV-vis and Raman band of bulk NiO in the NiAlO-4 catalyst, indicating
the whole NiO species occurs in the form of the surface phase. As can be seen from
Fig. 5a, the content of surface NiO in the NiAlO-5 catalyst is less than that of the
NiAlO-4 catalyst. The bulk phase NiO is the dominant phase for the NiAlO-7 catalyst.
The chemical composition and the distribution of Ni and Al species for the
NiAlO-5 and the NiAlF catalysts were investigated by means of XPS. Fig. 6 shows
the XPS spectra in Ni2p and Al2p region. The phases of Ni and Al species in the
NiAlF catalysts were evaluated by deconvolution of the Ni2p3/2, Ni3p and Al2p
region spectra, employing the fitting software of XPS peak 4.1. The BE (binding
energy) values of corresponding peaks obtained from the deconvolution and the
surface element ratio are summarized in Table 2.
Table 2 Binding energy of the Ni2p3/2, Ni3p and Al2p region and surface element ratio for
the NiAlO and NiAlF samples.
Ni2p3/2 / BE, eV
Ni3p / BE, eV
Al2p / BE, eV
Catalysts
11

NiO
853.4
853.2
—
NiF2
—
NiAl2O4
—
NiO NiF2 NiAl2O4
Al2O3
74.0
74.2
—
AlF3
—
NiAlO-5
NiAlF-4
NiAlF-5
NiAlF-6
NiAlF-7
68.1
—
—
—
858.1
858.2
858.4
858.6
—
67.9 71.6
77.7
77.9
77.8
77.5
—
—
—
—
71.2
71.7
71.8
—
—
855.1
854.9
69.3
69.4
—
—
—
As revealed from the analysis of the spectra in Fig. 6a, the Ni 2p3/2 peak of the
NiAlO-5 catalyst is resolved into two peaks at 853.4 and 859.3 eV, which were
assigned to NiO [32, 33]. After fluorination treatment, the environment of Ni species
was the characteristic of NiF2, demonstrated by the Ni 2p3/2 peak at a BE of ca. 858.1
eV, and an accompanying shake-up satellite peak at ca. 864.1 eV [34, 35, 36, 37, 38].
Whereas, for the NiAlF-4 catalyst, the NiO at a binding energy of ca. 853.2 eV was
still observed, with a high oxygen concentration of 20.9% (Table 2). It is proved that
the NiAlO-4 was not fully fluorinated, which is in accordance with the above
characterization results. For the NiAlF-5 catalyst, only two peaks at 858.2 and 864.1
eV were ascribed to the NiF2 phase, with the highest F concentration (Table 2). For
NiAlF-6 and NiAlF-7 catalysts, the BEs at 858.3 and 867.3 eV are both attributed to
NiF2, and a BE at 854.4 eV is related to Ni ions in the spinel NiAl2O4 [23, 33]. The
Al2p and Ni3p spectra are shown in Fig. 6b. For the NiAlO-5, the peaks at around
74.0 eV (Al2p) and 68.1 eV (Ni3p) are attributed to Al2O3 and NiO, respectively [39,
40, 41]. For the NiAlF-4 catalyst, the Al2p peak is resolved into two peaks at 74.2 and
77.5 eV. The former peak is assigned to the unfluorinated Al2O3 and the latter is
assigned to crystalline AlF3 [42]. The partial fluorination of the NiAlF-4 catalyst is
probably due to the presence of large amounts of surface phase NiO, which formed a
thin NiF2 film under the pre-fluorination process [34, 35], and restrained the further
fluorination of Ni and Al oxides. Furthermore, the low BE at 71.6 eV is due to the
NiF2 species. In the NiAlF-5 catalyst, the two peaks at 71.2 and 77.9 eV, are due to
the characteristic of NiF2 and AlF3, respectively. For the NiAlF-6 and NiAlF-7
catalysts, the AlF3 characteristic peak around 77.5 eV in the Al2p region, and the peak
at 71.7 and 69.3 eV are ascribed to NiF2 and NiAl2O4 in the Ni3p region, respectively.
Bolt et al investigated interfacial reaction between NiO and Al2O3 to spinel NiAl2O4 by
12

X-ray photoelectron spectroscopy. The Ni3p BE value at 69.1 eV is ascribed to the
inverse spinel NiAl2O4, which is very close to our results (69.3 eV) [39]. In addition, the
concentration of NiAl2O4 in NiAlF-7 (O: 5.84%) is higher than that of NiAlF-6 (O:
4.76%), in terms of the oxygen atom percentage(Table 1).
a)
b)
Al2p
Ni2p1/2
Ni2p3/2
Ni3p
NiAlF-7
NiAlF-6
NiAlF-7
NiAlF-6
NiAlF-5
NiAlF-5
NiAlF-4
NiAlF-4
NiAlO-5
NiAlO-5
890 880 870 860 850 840
85
80
75
70
65
Binding energy / eV
Binding energy / eV
Fig. 6 XPS spectra of the a) Ni2p and b) Al2p of NiAlO and NiAlF catalysts.
In summary, UV-DRS (Fig. 4b), Raman (Fig. 5b) spectra and XPS results (Fig.
6b) indicated that the inverse spinel NiAl2O4 still existed in the NiAlF-6 and NiAlF-7
catalysts after fluorination treatment. Besides, it was observed that the appearance of
NiAlF-6 and NiAlF-7 catalysts were light blue-green and bright azure blue,
respectively, demonstrating the presence of nickel aluminate [43]. The NiAl2O4
crystals in the cubic system and belongs to Fd-3 m space group. It is well-known that
inverse spinel NiAl2O4 phase formed under high calcination temperature is the
oxidation resistant and nonporous material with a stable structure [16, 44]. For the Ni
species of NiO/Al2O3 system, spinel NiAl2O4 with the Ni2+ in the form of tetrahedral
coordination is difficult to be reduced [33, 45, 46]. It was revealed by Priecel et al that
the spinel NiAl2O4 (Ni(Td)) was difficult to convert into sulfide form after sulfidation
[47]. Boukha et al reported that the inverse spinel NiAl2O4 was used as a stable
catalyst for methane steam reforming (SRM) under the high reaction temperature [23],
with the small surface area of ca. 50 m2 g-1. Therefore, the inverse spinel NiAl2O4 in
the NiAlO-7, is difficult to be fluorinated under the fluorination treatment, and has a
small surface area.
13

3.2 Surface Acidity
100 oC
(b)
0.02
300 oC
(a)
0.03
NiAF-7
NiAF-6
NiAF-7
NiAF-6
NiAF-5
NiAF-4
NiAF-5
NiAF-4
2000
1800
1600
1400
1200
2000
1800
1600
1400
1200
Wavenumber / cm-1
Wavenumber / cm-1
Fig. 7 Infrared spectra of adsorbed Py after desorption at 100 oC (a) and 300 oC (b) for
NiAlF catalysts.
Pyridine adsorption is an effective technique to measure the acidity, Lewis-
and/or Brönsted-acid of the acidic sites on the NiAlF catalysts. The IR spectroscopic
results of the previous reports [48, 49], three types of acid sites (weak, medium and
strong) existed in the NiAlF catalyst. To further understand the distribution of weak
sites and medium acid sites, the catalyst samples of pyridine adsorption at room
temperature were outgassed at 100 or 300 oC prior to py-IR tests. Py-IR spectra of the
NiAlF catalysts are shown in Fig. 7. It is well known that the wavenumber of the two
characteristic bands at 1450 and 1608 cm-1 (ν19b and ν8a modes, respectively, of
coordinated Py species) account for the Lewis acid sites, and the ν19b band at 1540
cm-1 is attributed to the Brönsted acid sites [50, 51]. Obviously, the presence of Lewis
acid sites at ν19b = 1450 cm-1 on the NiAlF catalysts is observed in Fig. 7 (a) and (b).
However, the Brönsted acid sites at ν19b = 1540 cm-1 cannot be detected by py-IR
characterization. Therefore, it can be concluded that the acid site on the surface of the
NiAlF catalysts are dominantly the Lewis acidic sites. Besides, the NiAlF-5 catalyst
o
has the strongest IR band at 1450 cm-1 (weak acid sites: 100 C; medium acid sites:
300 oC), relating to the largest number of weak and medium Lewis acid sites.
14

The species and acid distributions of the NiAlF catalysts are shown in Table 3.
The sequence of weak and medium Lewis acidity of the NiAlF catalysts is as follows:
NiAlF-5 > NiAlF-6 > NiAlF-7 > NiAlF-4 (Table 3). The NiAlF-4 catalyst composed
of AlF3, NiF2, Al2O3 and NiO, has strong Lewis acid sites (Table 3), which derived
from the Al2O3 phase (possessed strong acid sites) [52]. However, the NiAlF-5, -6 and
-7 catalysts do not have strong Lewis acid sites. The NiAlF-7 has the lowest total
acidic amount(0.521 mmol g-1) among the all fluoride catalysts, while weak and
medium (weak&medium) Lewis acidity is higher than those of NiAlF-4 (0.493 mmol
g-1). Comparatively, NiAlF-6 and NiAlF-7 catalysts made up of NiF2, AlF3 and
NiAl2O4, show the less weak and medium Lewis acid sites than those of NiAlF-5
catalyst. This is probably attributed to the influence of inverse spinel NiAl2O4, which
has a very weak acidity.
Table 3 Species, acidic properties, reaction rate and TOF of the NiAlF catalysts.
Weak&medium Total acidic Rate
Lewis acid sites amount
/ mmol min-1 g-1
TOF§
/ h-1
Catalysts Species
/ mmol g-1
/ mmol g-1
NiAlF-4
NiAlF-5
NiAlF-6
NiAlF-7
AlF3, NiF2, Al2O3, NiO
0.493
0.678
0.906
0.829
0.521
0.82
2.13
1.81
1.06
0.195
0.254
0.231
0.230
AlF3, NiF2
0.906
0.829
0.521
AlF3, NiF2, NiAl2O4
AlF3, NiF2, NiAl2O4
 The reaction rates measured based on their conversion, under the same reaction conditions: N2/CF3CH2F
= 10, 450 oC, GHSV = 675 h-1, 2 g catalysts.
§ The TOF measured based on their conversion, under the same reaction conditions: N2/CF3CH2F = 10,
400 oC, 2 g catalysts.
3.3. Catalytic testing
Fig. 8 shows the conversion of CF3CFH2 over NiAlF catalysts and the
corresponding selectivity to trifluoroethylene at different temperatures. It can be seen
that calcination temperature has a great influence on the catalytic activity. The activity
of NiAlF catalyst declined gradually with increasing calcination temperature from 500
o
to 700 C. Furthermore, the NiAlF-5 catalyst showed the highest activity among all
the NiAlF catalysts due to the high weak and medium active acid sites. For example,
CF3CH2F conversion of 32.8 % was obtained on the NiAlF-5 catalyst at 480 oC, while
conversion of 15.2 % was obtained on the NiAlF-4 catalyst.
15

35
30
25
20
15
10
5
100
95
90
85
80
75
70
NiAlF-4
NiAlF-5
NiAlF-6
NiAlF-7
NiAlF-4
NiAlF-5
NiAlF-6
NiAlF-7
b)
a)
400
420
440
460
480
400
420
440
460
480
Temperature /oC
Temperature / oC
Fig. 8 a) the conversion of CF3CFH2 and b) selectivity to trifluoroethylene at
different temperatures for NiAlF catalysts. Error bars are estimated from
standard deviations from three repeats of experiments where available.
Fig. 9 demonstrates the stability of AlF3, NiAlO-5 and NiAlF-5 catalysts for
o
dehydrofluorination of CF3CFH2 at 450 C under the same reaction conditions. The
conversion of the unfluorinated NiAlO-5 (metal oxides) declined drastically with time
on stream, indicating that it is not stable enough for dehydrofluorination of CF3CFH2.
This may be the strong acidity (1.426 mmol g-1) of the NiAlO-5 catalyst determined
by ammonia adsorption resulted in coking on the catalyst surface. In contrast, the
NiAlF-5 catalyst (metal fluorides) possess a high stability. Generally, metal fluorides
are more stable than metal oxides, and metal oxide cannot be applied as catalysts
and/or supports for reactions in which hydrogen fluoride is involved [53]. The
gas-phase fluorination with HF gas is a typical method to activate the catalysts [52],
especially for the metal oxide catalyst, because it can improve the rate and stability of
catalysts. Interestingly, the conversion of NiAlF-5 catalyst is much higher than that of
AlF3 catalyst (Fig. 9), which is attributed to the addition of Ni species. It could be
seen that the conversion reduced gradually, and then reached a steady state after 80 h.
In addition, statistics summary of the input and output materials were cumulatively
calculated during 50 h, and listed in Table 4. It was found that the loss of input
reactant (N2 and CF3CH2F) from 2.948 to 2.867 mol, the materials balance was break
up due to the coke transformation of CF3CH2F or CF2=CHF on the catalyst surface.
16

100
80
60
40
20
0
100
80
60
40
20
0
NiAlO-5
NiAlF-5
AlF3
0
50
100 150 200 250 300
Time on stream / h
Fig. 9 The activity of catalyst for dehydrofluorination of CF3CFH2. Reaction
conditions: N2/CF3CH2F = 10, 450 oC, GHSV = 675 h-1.
Table 4 Statistics summary of input and output materials in 50 h.
Phase
Cumulative amount / mol
Total amount / mol
2.948a
Input gas:
N2
2.680
0.268
CF3CH2F
N2
Output gas:
2.867b
CF3CH2F
CF2=CHF
Others
Output liquid*: HF
0.060
a Total amount of input gas includes N2 and CF3CH2F.
b Total amount of output gas included N2, CF3CH2F, CF2=CHF, and other byproducts.
* The amount of HF in tail gas absorber is calculated by the chemical titration.
The reaction conditions: N2/CF3CH2F=10, 450 oC, 1 bar, GHSV = 675 h-1, 2 g catalysts.
In order to investigate the state of spent catalyst under reaction conditions, XRD
were carried out and shown in Fig. S1 (supplementary information). The peaks
associated to AlF3 and NiF2 of the NiAlF-5 catalyst increase slowly during the 300 h.
17

The crystalline size of AlF3 and NiF2 increased, suggesting that the NiAlF-5 catalyst
might be slightly sintered. Raman spectra and TG curves of the NiAlF-5 catalyst are
shown in Fig. S2 (supplementary information). Two intensive broad bands at 1318
and 1593 cm-1 in Raman spectra (Fig. S2 a) are attributed to the A1g and the E2g
vibrations of the carbon, respectively [54], indicating that carbon deposited on the
surface of the NiAlF-5 catalyst (Fig. S2 a). TG experiments were conducted to further
illustrate the coke formation of NiAlF-5 catalyst. An obvious weight loss below 100
◦C for all the used catalyst was observed in Fig. S2 b, due to the loss of physical
adsorption H2O. In addition, a main weight loss was observed around the broad range
o
of 500-590 C, assigning to the combustion of the deposited carbon on the used
catalysts. It is found that the weight loss of the used catalyst increased gradually as
reaction time increased, which further confirmed the coke formation on the NiAlF
catalyst surface, agrees well with the Raman results (Fig. S2 b).
The reproducibility of catalyst is a significant feature for industrial applications
from the stand point of decreasing process costs. Therefore, the reproducibility of the
NiAlF-5 catalyst in the dehydrofluorination of CF3CH2F was investigated. The spent
NiAlF-5 catalyst was regenerated, and reused directly for the next cycle. Fig. S3
(supplementary information) shows the reproducibility of catalytic behaviour. The
conversion of CF3CH2F and selectivity to trifluoroethylene were obtained in
reaction-regeneration cycles with reaction step at 450 °C. As seen in Fig. S3, the
NiAlF-5 catalyst has a good catalytic performance for dehydrofluorination of
CF3CH2F to synthesize trifluoroethylene.
3.2 Active sites for dehydrofluorination reaction
The relationship between Lewis acidity and catalyst activity for
dehydrofluorination reaction is necessary to study. As seen in Table 3, the NiAlF-4
catalyst behaves a lower reaction rate (0.82 mmol min-1 g-1) than the NiAlF-7 catalyst
(1.06 mmol min-1 g-1), although it has less weak and medium Lewis acid sites.
Moreover, for the NiAlF-5, -6 and -7 catalyst without strong Lewis acid sites, their
activities are closely related to weak and medium Lewis acid sites. Fig. 10 shows the
relationship between weak and medium Lewis acid sites and reaction rate. It is
interesting that the reaction rate for the NiAlF catalysts, increases as the amount of
weak and medium Lewis acid sites increasing.
18

NiAlF-5
NiAlF-6
2.0
1.5
1.0
0.5
0.0
NiAlF-7
NiAlF-4
0.0
0.2
0.4
0.6
0.8
1.0
W-M Lewis acid sites / mmol g-1
Fig. 10 Relationship between weak and medium Lewis acid sites and the
reaction rate.
The prepared NiAlF-5 catalyst has the excellent catalytic performance for
dehydrofluorination of hydrofluorocarbons (HFCs) compared to the catalysts reported
previously. Okazaki and coworkers reported dehydrofluorination of CF3CH3 over
various metal fluorides catalysts [8], and found that AlF3 catalyst with Lewis acid
sites showed the highest activity, with the conversion of 30 %, at a very high
o
temperature of 700 C. In another group, Li et al. prepared a Mg2P2O4 catalyst and it
o
appears that CF3CH3 conversion became a steady state of 50 % over 100 h at 450 C
(W/F=0.067 g min cm-3, feed gas compositions: CF3CH3 2.0 mol %, H2 15 mol %,
and balance N2) [9, 10]. However, the initial selectivity of CF2CH2 was only around
75 %. In order to compare our normalized rates quantitatively with other reported
catalysts, several catalysts were prepared according to the literatures and used for the
dehydrofluorination of CF3CFH2. Under the same reaction conditions: N2/CF3CH2F =
o
10, 450 C, GHSV = 675 h-1, 2.0 g catalysts, the space time yield of the proposed
Mg2P2O7, AlF3, Fe2O3-CdO-Al2O3 and Pd/AlF3 catalyst are 0.0153, 0.0571, 0.0326,
0.0617 mol h-1 gcat.-1, respectively. In our work, the NiAlF-5 catalyst possessed the
highest space time yield of 0.127 mol h-1 gcat.-1, compared with the reported
dehydrofluorination catalysts presently. The NiAlF-5 catalyst can achieve a high
reaction rate, which can be attributed to its large weak and medium Lewis acidic
amount.
19

Turnover frequency (TOF) of the reaction calculated based on the ratio of rate
and weak and medium Lewis acid sites is shown in Table 3. It was found that TOF
firstly increased and then decreased with increasing calcination temperature, and the
highest TOF is of NiAlF-5 catalyst because of the most weak and medium Lewis acid
sites. It was reported by Li et al that weak Lewis acid sites were the only active sites
for dehydrofluorination of CF3CH3 to CF2=CH2 [9, 10]. Our finding supports the
reported conclusion because the activity closely correlates with the weak and medium
Lewis acid sites. In our previous works, the relationship between acidity and activity
for the dehydrofluorination performance of catalysts were investigated [14, 55]. Only
the weak Lewis acid was considered as the reaction active site for dehydrofluorination
of CF3CFH2 to CF2CHF. Herein, the acidic distribution of NiAlF catalysts were
investigated in detail. It demonstrates that both the weak and medium Lewis acid sites
are closely related to the catalytic performance. In conclusion, the weak and medium
Lewis acid sites were mainly active centres of catalysts for dehydrofluorination of
CF3CFH2 to trifluoroethylene. Furthermore, the new finding in this study is that the
origin of the weak, medium and strong Lewis acid sites of NiF2-AlF3 catalyst
identified. That is, the well Ni and Al oxide species on the catalyst surface could be
transformed into isolated NiF2 and AlF3, which act as the weak and medium Lewis
acid sites for the dehydrofluorination of CF3CFH2.
4. Conclusion
A series of NiAlF catalysts derived from the fluorination of NiO-Al2O3 at
different calcination temperatures, were used for the dehydrofluorination of CF3CH2F
to synthesize CF2CHF. The highest activity was obtained on a fluorinated catalyst
-1
calcined at 500 ◦C, with a reaction rate of 2.13 mmol min-1 gcat. at 450 ◦C.
Calcination temperature of the NiAlO catalysts and fluorination treatment have great
influence on the structure and acidity of the obtained catalysts. With the increasing
calcination temperature, NiO species transformed from the surface into bulk phase,
o
accompanied by the formation of inverse spinel NiAl2O4 over 600 C. Furthermore,
the effect of surface Lewis acid sites on the activity for dehydrofluorination was
investigated. It is found that both the weak and medium Lewis acid sites in AlF3-NiF2
complex were the active centres for catalyzing dehydrofluorination of CF3CH2F to
trifluoroethylene.
20

Acknowledgements
This work was financially supported by National Natural Science Foundation of
China (No. 21603069 and 91534115), Education Foundation of Hubei Province (No.
Q20184502) and the Science Foundation from Hubei Polytechnic University (No.
16xjz05R). Min Liu acknowledges the support from China Scholarship Council -
University of Melbourne Research Scholarship (No. 201606260063).
21

References
[1] Z.T. Liu, L. Liu, J. Wu, J. Lu, P. Sun, L.P. Song, Z.W. Liu, W.S. Dong, Z.W. Gao,
Ind. Eng. Chem. Res. 46 (2007) 22-28.
[2] K.P. Shine, W.T. Sturges, Science, 315 (2007) 1804-1805.
[3] S. Henne, D.E. Shallcross, S. Reimann, P. Xiao, D. Brunner, S. O’Doherty, B.
Buchmann, Environ. Sci. Technol., 46 (2012) 1650-1658.
[4] L.B. Qin, J. Han, L. Liu, X.L. Yang, H.J. Kim, F. Yu, Fresenius Environ. Bull. 22
(2013) 1919-1923.
[5] C.Q. Huang, B. Yang, R. Yang, L.X. Wei, S.S. Wang, X.B. Shan, F. Oi, Y.W.
Zhang, L.S. Sheng, L.Q. Hao, S.K. Zhou, Z.Y. Wang, Chem. J. Chinese Universities
26 (2005) 2314-2318.
[6] Narengerile, H. Saito, T. Watanabe, Plasma Chem. Plasma Process. 30 (2010)
813-829.
[7] H. Serge, L Andre, FR Patent 2710054 (1995), to Leboulenger Jean Elf Atochem
S. A.; H. Serge, FR Patent 2729136 (1997), to Elf Atochem S. A. Societe Anonyme;
R. Powell, A. Sharratt. US Patent 5856593 (1999), to Imperial Chemical Industries
PLC.
[8] S. Okazaki, S. Toyota, Nikkashi, 1972, 1615-1617.
[9] G.L. Li, H. Nishiguchi, T. Ishihara, Y. Moro-oka, Y. Takita, Appl. Catal. B, 16
(1998) 309-317.
[10] G.L. Li, T. Ishihara, H. Nishiguchi, Y. Moro-oka, Y. Takita, Chem. Lett. 7
(1996) 507-508.
[11] T.S. Sirlibaev, A. Akramkhodzhaev, K. U. Usmanov, J. Appl. Chem. USSR 58
(1985) 1541-1543.
22

[12] M. Tojo, S. Fukuoka, H. Tsukube, Bull. Chem. Soc. Japan. 84 (2011) 333-340.
[13] K. Teinz, S. Wuttke, F. Börno, J. Eicher, E. Kemnitz, J. Catal. 282 (2011)
175-182.
[14] W.Z. Jia, M. Liu, X.W. Lang, C. Hu, J.H. Li, Z.R. Zhu, Catal. Sci. Technol. 6
(2015) 3103-3106.
[15] X.Wen, R.S. Li, Y.X. Yang, J.L. Chen, F.Z. Zhang, Appl. Catal. A: General 468
(2013) 204-215.
[16] H. Zhang, Y.Y. Dong, W.P. Fang, Y.X. Lian, Chin. J. Catal. 34 (2013) 330-335.
[17] X.L. Fu, W.M. Tang, L. Ji, S.F. Chen, Chem. Eng. J. 180 (2012) 170-177.
[18] J. Escobar, J. A. De Los Reyes, T. Viveros, Appl.Catal. A: General 253 (2003)
151-163.
[19] I. Preda, R.J.O. Mossanek, M. Abbate, L. Alvarez, J. Méndez, A. Gutiérrez, L.
Soriano, Surface Science , 606 (2012) 1426-1430.
[20] B. Scheffer, J.J. Heijeinga, J.A. Moulijn. J. Phys. Chem. 91 (1987) 4752-4759.
[21] S. Damyanova, B. Pawelec, K. Arishtirova, J.L.G. Fierro, Int. J. Hydrogen
Energy 37 (2012) 15966-15975.
[22] C. Ragupathi , J. Judith Vijaya, P. Surendhar, L. John Kennedy, Polyhedron 72
(2014) 1-7.
[23] Z. Boukha, C. Jiménez-González, B. de Rivas, J.R. González-Velasco, J.I.
Gutiérrez-Ortiz, R. López-Fonseca, Appl. Catal. B: Environmental 158–159 (2014)
190-201.
[24] M.A. Vidales-Hurtado, A. Mendoza-Galván, Solid State Ionics 179 (2008)
2065-2068.
[25] S.S. Chan, I.E. Wachs, J. Catal. 103 (1987) 224-227.
23

[26] A. Aminzadeh, H. Sarikhani-Fard, Spectrochim. Acta Part A 55A (1999)
1421-1425.
[27] G.T. Liu, H.F. Chen, G.M. Lin, P.P. Ye, X. P. Wang, Y.Z. Jiao, X.Y. Guo, Y.
Wen, H.F. Yang, Biosens. Bioelectron. 56 (2014) 26-32.
[28] S.K.Yadav, P. Jeevanandam, J. Alloys Compd. 610 (2014) 567-574.
[29] A.V. Ghule, K. Ghule, T. Punde, J.Y. Liu, S.H. Tzing, J.Y. Chang, H. Chang,
Y.C. Ling, Mater. Chem. Phys. 119 (2010) 86-92.
[30] E. Meloche, M. G. Cottam, D. J. Lockwood, Phys. Rev. B 76 (2007)
104406-104410.
[31] V.N. Bukhmarina, A.Yu. Gerasimov, Y.B. Predtechenskii, V.G. Shklyarik, Opt.
Spektrosk. 65 (1988) 518-521.
[32] R. Molina, G. Poncelet, J. Phys. Chem. B 103 (1999) 11290-11296.
[33] J.L. Chen, Q. Ma, T. E. Rufford, Y.D. Li, Z.H. Zhu, Appl. Catal. A: General 362
(2009) 1-7.
[34] G.G. Totir, G.S. Chottiner, C.L. Gross, W. Ves Childs, D.A. Scherson, J.
Electrochem. Soc. 147 (2000) 4212-4216.
[35] G.G. Totir, G.S. Chottiner, C.L. Gross, D.A. Scherson, J. Electrochem. Soc. 148
(2001) 262-266.
[36] M.B. Chan-Park , J.X. Gao, A.H. L. Koo, J. Adhesion Sci. Technol. 17 (2003)
1979-2004.
[37] A.Tasaka, T. Osada, T. Kawagoe, M. Kobayashi, A. Takamuku, K. Ozasa, T.
Yachi, T. Ichitani, K. Morikawa, J. Fluorine Chem. 87 (1998) 163-171.
[38] A. Tasaka, E. Morimoto, A. Mimoto, A. Inoue, M. Inaba, ECS Transactions 3
(2007) 529-542.
[39] P.H. Bolt, E. ten Grotenhuis, J.W. Geus, F.H.P.M. Habraken, Surf. Sci. 329
(1995) 227-240.
24

[40] K. Ali, K.H. Choi, J. Jo, Y.W. Lee, Mater. Lett. 136 (2014) 90-94.
[41] L. Bloch, Y. Kauffmann, B. Pokroy, Cryst. Growth Des. 14 (2014) 3983-3989.
[42] A. Makarowicz, C.L. Bailey, N. Weiher, E. Kemnitz, S.L.M. Schroeder, S.
Mukhopadhyay, A. Wander, B.G. Searle, N.M. Harrison, Phys. Chem. Chem. Phys.
11 (2009) 5664-5673.
[43] D. Dissanayake, M.P. Rosynek, K.C.C. Kharas, J.H. Lunsford, J. Catal. 132
(1991) 117-127.
[44] E. Loginova, F. Cosandey, T.E. Madey, Surf. Sci. 601 (2007) 11-14.
[45] Y. Chen, L.F. Zhang, Catal. Lett. 12 (1992) 51-62.
[46] G. Poncelet, M.A. Centeno, R. Molina, Appl. Catal. A: General. 288 (2005)
232-242.
[47] P. Priecel, D. Kubičk, L. Čapek, Z. Bastl, P. Ryšanek, Appl. Catal. A: General
397 (2011) 127-137.
[48] C. Morterra, G. Magnacca, Catal. Today 27 (1996) 497-532.
[49] M.Stekrova, N. Kumar, P. Mäki-Arvela, A. Aho, J. Linden, K.P. Volcho, N.F.
Salakhutdinov, D.Yu. Murzin, React. Kinet. Mech. Cat. 110 (2013) 449-458.
[50] G. Ramis, L. Yi, G. Busca, Catal. Today 28 (1996) 373-380.
[51] G. Busca, Phys. Chem. Chem. Phys. 1 (1999) 723-736.
[52] T.K. Phung, C. Herrera, M.Á. Larrubia, M. García-Diéguez, E. Finocchio, L.J.
Alemany, G. Busca, Appl. Catal. A: General 483 (2014) 41-51.
[53] Y. Takita, T. Ohkuma, H. Nishiguchi, K. Nagaoka, T. Nakajo, Appl. Catal. A:
General 283 (2005) 47-52.
[54] Y.Xia, W.K. Zhang, Z. Xiao, H. Huang, H.J. Zeng, X.R. Chen, F. Chen, Y.P.
Gan, X.Y. Tao, J. Mater. Chem. 22 (2012) 9209-9215.
25

[55] W.Z. Jia, Q. Wu, X.W. Lang, C. Hu, G.Q. Zhao, J.H. Li, Z.R. Zhu, Catalysis
Letters, 2015, 145, 654-661.
26