Rational design of MgF2 catalysts with long-term stability for the dehydrofluorination of 1,1-difluoroethane (HFC-152a)

Article abstract of DOI:10.1039/c9ra04250d

In this study, three different approaches, i.e. the sol-gel method, precipitation method and hardlate method, were applied to synthesize MgF₂ catalysts with improved stability towards the dehydrofluorination of hydrofluorocarbons (HFCs); the in situ XRD technique was employed to investigate the relationship between the calcination temperature and the crystallite size of precursors to determine optimal calcination temperature for the preparation of the MgF₂ catalysts. Moreover, the physicochemical properties of MgF₂ catalysts were examined via BET, XRD, EDS and TPD of NH₃ and compared. Undoubtedly, the application of different methods had a significant influence on the surface properties and catalytic performances of MgF₂ catalysts. The surface areas of the catalysts prepared by the precipitation method, sol-gel method and template method were 120, 215 and 304 m² g^-1, respectively, upon calcination at 200 °C. However, the surface area of the MgF₂ catalysts decreased significantly when the calcination temperatures of 300 and 350 °C were applied. The catalytic performance of these catalysts was evaluated via the dehydrofluorination of 1,1-difluoroethane (HFC-152a). The MgF₂ catalyst prepared by the precipitation method showed the lowest catalytic activity among all the MgF₂ catalysts. When the calcination temperature was above 300 °C, the MgF₂ catalysts prepared via the template method demonstrated the highest catalytic conversion rate with catalytic activity following the order: MgF₂-T (template method) > MgF₂-S (sol-gel method) > MgF₂-P (precipitation method). The conversion rate generally agreed with the total amount of acid on the surface of the catalysts, which was measured by the NH₃-TPD technique. The MgF₂-T catalysts were further examined for the dehydrofluorination of HFC-152a for 600 hours, and a conversion rate greater than 45% was maintained, demonstrating superior long-term stability of these catalysts.

Full text of DOI:10.1039/c9ra04250d

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Rational design of MgF2 catalysts with long-term
stability for the dehydrofluorination of 1,1-
difluoroethane (HFC-152a)†
Cite this: RSC Adv., 2019, 9, 23744
a
b
Haodong Tang, Mingming Dang,b Yuzhen Li,a Lichun Li,
Wenfeng Han,a
*
*
Zongjian Liu,b Ying Li
and Xiaonian Lia
a
In this study, three different approaches, i.e. the sol–gel method, precipitation method and hard-template
method, were applied to synthesize MgF2 catalysts with improved stability towards the dehydrofluorination
of hydrofluorocarbons (HFCs); the in situ XRD technique was employed to investigate the relationship
between the calcination temperature and the crystallite size of precursors to determine optimal calcination
temperature for the preparation of the MgF2 catalysts. Moreover, the physicochemical properties of MgF2
catalysts were examined via BET, XRD, EDS and TPD of NH3 and compared. Undoubtedly, the application of
different methods had a significant influence on the surface properties and catalytic performances of MgF2
catalysts. The surface areas of the catalysts prepared by the precipitation method, sol–gel method and
template method were 120, 215 and 304 m2 ꢀ1, respectively, upon calcination at 200 ꢁC. However, the
g
surface area of the MgF2 catalysts decreased significantly when the calcination temperatures of 300 and
350 ꢁC were applied. The catalytic performance of these catalysts was evaluated via the dehydrofluorination
of 1,1-difluoroethane (HFC-152a). The MgF2 catalyst prepared by the precipitation method showed the
lowest catalytic activity among all the MgF2 catalysts. When the calcination temperature was above 300 ꢁC,
the MgF2 catalysts prepared via the template method demonstrated the highest catalytic conversion rate
with catalytic activity following the order: MgF2-T (template method) > MgF2-S (sol–gel method) > MgF2-P
(precipitation method). The conversion rate generally agreed with the total amount of acid on the surface
of the catalysts, which was measured by the NH3-TPD technique. The MgF2-T catalysts were further
examined for the dehydrofluorination of HFC-152a for 600 hours, and a conversion rate greater than 45%
was maintained, demonstrating superior long-term stability of these catalysts.
Received 6th June 2019
Accepted 11th July 2019
DOI: 10.1039/c9ra04250d
rsc.li/rsc-advances
rate and improve the yield of the target product. Currently, the
commonly used catalysts for the dehydrouorination of
1 Introduction
hydrouorocarbons (HFCs) are Lewis acid catalysts including
chromia5–7 and alumina5,8-based catalysts. However, the chro-
mia and alumina-based catalysts oen suffer from short life-
time as they get deactivated from coking on the strong Lewis
acid offered by the catalysts. Zhu et al. have reported the
application of porous alumina (Al2O3) as a catalyst for the
dehydrouorination of 1,1,1,2-tetrauoroethane and demon-
strated that strong Lewis acid sites are responsible for the
deactivation of catalysts due to coke formation.8 As reported by
Luo et al.,6 the catalytic dehydrouorination of 1,1,1,3,3-penta-
uoropropane was investigated using Cr2O3 as a catalyst. The
results showed that the conversion rate of the Cr2O3 catalyst
signicantly decreased from 68% to 23% aer a 10 hour test,
and the addition of NiO beneted the stability of the catalyst.
Due to its high strength, the Lewis acid certainly acts as an
active site and offers high conversion rate; however, it also
increases the chances of catalyst deactivation.
Polyvinyl uoride (PVF) is a widely used polymeric material
because of its outstanding resistance to weathering, chemically
inert nature, suitable mechanical strength, low permeability
and light transmittance.1,2 PVF has been commonly applied in
packaging, glazing and electrical applications; the production
of vinyl uoride (VF) is rather important as it is the raw chemical
material for the synthesis of PVF. The dehydrouorination of
1,1-diuoroethane (HFC-152a) is the most broadly used chem-
ical process for the production of VF. The dehydrouorination
of hydrouorocarbons (HFCs) is generally accepted to be ther-
modynamically and kinetically hindered.3,4 Hence, it is essential
to employ highly effective catalysts to accelerate the reaction
aInstitute of Industrial Catalysis, Zhejiang University of Technology, Hangzhou,
310014, PR China. E-mail: tanghd@zjut.edu.cn; lichunli@zjut.edu.cn
bCollege of Chemical Engineering, Zhejiang University of Technology, Hangzhou,
310014, PR China
On the other hand, magnesium uoride (MgF2) nano-
particles have been widely applied as supporting materials and
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9ra04250d
23744 | RSC Adv., 2019, 9, 23744–23751
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
heterogeneous catalysts because they contain medium-strength (Hangzhou Wanjing Chemical Reagent Co., Ltd), and HF
Lewis acids, which are less likely to get deactivated due to (40 wt%, Quzhou Juhua Reagents Co., Ltd.) were used as ob-
coking;5,9–11 MgF2 typically exists in the rutile form where each tained without further purication. 1,1-Diuoroethane (HFC-
Mg2+ ion is surrounded by six Fꢀ ions and each Fꢀ ion is sur- 152a, 99.0%) was purchased from Zhejiang Juhua Group
rounded by three Mg2+ ions.12,13 A typical MgF2 crystal does not Company.
contain any acid property and exists in the form of a neutral
solid. Therefore, the manipulation of the synthesis method
offers an intelligent way to modify the surface property and
Preparation of catalysts
create a surface disordered MgF2 material to produce medium-
Precipitation method. The prepared Mg(AC)2 solution was
strength Lewis acid sites on the surface. This can be explained mixed with the NH4F solution under rigorous stirring for 1 hour
by the Tanabe model where the surface acidity of MgF2 is to ensure complete precipitation. The molar ratio of F : Mg was
generated from the excessive positive charge caused by the controlled to be 2 : 1. The resulting gel solution was aged for 3
coordinated unsaturated Mg2+ ions on the surface;14 the hours before ltration for three times. The generated residue
ꢁ
generation of high surface area is oen accompanied by the was dried in an oven for 12 hours at 100 C. The dry paste was
production of large amount of acid sites;15 therefore, signicant then calcined for 4 hours at 200, 300 and 350 C under an air
ꢁ
research efforts have been focused on the synthesis of high- atmosphere. The obtained product was labelled as MgF2-P.
surface area MgF2 (HS-MgF2).16–23 Kemnitz et al.17 have
Sol–gel method. Mg turnings were added to excess amount of
successfully synthesized amorphous nano-sized MgF2 with the anhydrous methanol to form magnesium methoxide slurry under
high surface area of 150–350 m2 gꢀ1 via a non-aqueous uo- rigorous stirring. The obtained mixture was then uorinated by
rolytic sol–gel method. The results of the analysis of adding a certain amount (F/Mg ¼ 2 : 1) of 40 wt% HF solution
temperature-programmed desorption of ammonia (NH3-TPD) followed by stirring for 1 hour until a clear gel was formed. The
showed that the strong acid sites disappeared when the cata- resulting gel mixture was subjected to aging for 3 hours at room
lysts were pre-treated at 300 ꢁC, and above this temperature, the temperature and then dried overnight under vacuum at 70 C.
ꢁ
structure of HS-MgF2 would probably be affected. The micro- The dried samples were calcined for 4 hours at 200, 300 and
emulsion route was also applied as another effective method to 350 ꢁC. The obtained product was labelled as MgF2-S.
synthesise HS-MgF2.22,23 Saberi et al.13 have prepared HS-MgF2
Template method. SP-15 (SiO2) was rst soaked in the
with the surface area of 190 m2 gꢀ1 and the average crystallite Mg(AC)2 solution before the addition of the NH4F solution to
size of 9–11 nm via the microemulsion route. However, it was ensure the uorination process. Rigorous stirring was applied
challenging to remove the microemulsion agent.
for one hour to produce a gel solution. The gel solution was
Despite all these promising efforts towards the synthesis of aged for three hours before ltration. The residue was then
HS-MgF2, to date, the MgF2 catalysts still suffer from deactiva- dried at 100 ꢁC for 12 hours and calcined for 4 hours at 200, 300
tion caused by the aggrandizement of particle size at evaluated and 350 ꢁC under an air atmosphere. The calcined product was
temperatures. The main purposes of this study were to ratio- transferred into a 40 wt% HF solution to remove the SP-15
nally design and manufacture the surface properties of MgF2 as template. The obtained product was labelled as MgF2-T.
they govern the catalytic performance of the catalyst. In this
study, three different synthesis methods, i.e. the precipitation
method, aqueous sol–gel method and hard template method,
Characterisation
were attempted to synthesize MgF2. The in situ XRD technique The surface area of the catalysts was examined using N2 gas
was rst applied to examine the effect of calcination tempera- sorption via the Micromeritics ASAP 2000 instrument at
ture on the MgF2 size. Moꢁreover, three calcination tempera- ꢀ196 ꢁC. The samples were degassed at 120 ꢁC for 6 hours
tures, i.e. 200, 300 and 350 C, were selected and used in each before each measurement. The surface area was evaluated by
synthesis method to further evaluate the inuence of calcina- the Brunauer–Emmett–Teller (BET) method using the adsorp-
tion temperature on catalyst properties. The nine MgF2 samples tion data. Pore size distributions were obtained from the
prepared using three synthesis methods and three calcination adsorption branches via a nonlocal density functional theory
temperatures were examined via Brunauer–Emmett–Teller (NLDFT) method. The pore volume was approximated at the
(BET), X-ray powder diffraction (XRD) and NH3-TPD techniques relative pressure P/P0 of 0.99.
to reveal the surface properties of these samples. The dehy-
X-ray powder diffraction (XRD) characterization was con-
drouorination of 1,1-diuoroethane (HFC-152a) was employed ducted using the X'pert PRO diffractometer with Cu Ka radia-
as a probe reaction to test the catalytic performance of the MgF2 tion (40 kV and 100 mA) in the range of 10ꢁ # 2q # 80ꢁ. The as-
catalysts. The surface acidic property was further employed to prepared MgF2 catalysts were evaluated before and aer the
rationalize the catalytic performance of the MgF2 catalysts.
dehydrouorination reaction of HFC-152 using the XRD
technique.
The in situ XRD characterization was performed using the
same instrument under the same conditions with temperature
ranging from 25 to 400 C, which was controlled via a thermo-
2 Materials and methods
Chemicals
ꢁ
Mg(AC)2 (AR grade, Sinopharm Chemical Reagent Co., Ltd), couple in the presence of a H2 atmosphere. The MgF2 samples
NH4F (AR grade, Sinopharm Chemical Reagent Co., Ltd), SP-15 prepared by the precipitation, sol–gel, and template methods
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 23744–23751 | 23745

View Article Online
RSC Advances
Paper
before calcination were examined viaꢁthe in situ XRD technique
at temperatures between 25 and 400 C.
The temperature-programmed desorption of ammonia (NH3-
TPD) was used to reveal the distribution and strength of the acid
sites on the catalysts. The MgF2 sample (about 0.15 g) was rst
degassed by heating at 300 ꢁC for 30 minutes under an argon
atmosphere. Aer being gently cooled down to room tempera-
ture, the catalyst was exposed to a gas mixture containing 10% of
NH3 and 90% of He for 30 minutes. The TPD program was started
ꢀ1
ꢁ
ꢁ
(10 C min up to 850 C) aer the excess amount of NH3 was
ushed out under N2 at 100 ꢁC. The amount of desorbed NH3 was
continuously monitored via online mass spectrometry.
Energy dispersive spectrometry (EDS) analysis was per-
formed using the HITACHI 7700 microscope at the accelerating
voltage of 100 kV.
Catalytic performance
The MgF2 catalysts were applied for the dehydrouorination of
HFC-152 to evaluate the catalytic performance of the catalysts
prepared using different synthesis methods and calcination
temperatures. The dehydrouorination reaction of HFC-152 is
shown in eqn (1).
CH3CHF2 / CH2]CHF + HF
(1)
In a typical experiment, about 2.0 mL of the catalyst was
packed in the Nickel-iron-chromium alloy tube ow reactor with
the owing mixed gases of HFC-152 (ꢂ25 mL minꢀ1) and N2 (20
mL minꢀ1). The reaction was performed at 200, 300 and 350 ꢁC
under atmospheric conditions. The component of the gas phase
was monitored by a gas chromatography instrument (GC16900)
equipped with the TCD analyzer.
3 Results and discussion
In situ XRD
Fig. 1 shows the in situ XRD patterns of the pre-calcined MgF2-S,
MgF2-T, and MgF2-P catalysts obtained under a H2 atmosphere
with temperature ranging from 25 to 450 ꢁC. The distinct
advantage of the in situ XRD measurement is that it eliminates
the multiple XRD measurements of the MgF2 samples calcined
at different temperatures, which are time and effort consuming;
herein, the MgF2 samples examined via the in situ XRD
measurements are not exactly the same as the actual catalysts as
they have typically undergone a calcination process at a certain
temperature for 4 hours under a N2 atmosphere. The purpose of
the in situ XRD measurements was to rationalize the relation-
ship between the calcination temperature and the size of the
MgF2 crystallite to conrm that temperature played an impor-
tant role in the crystallite size and select optimal temperature
for the synthesis of the MgF2 catalysts.
Fig. 1 In situ XRD patterns of the MgF2 catalysts MgF2-P (a), MgF2-S
(b), and MgF2-T (c) catalysts at calcination temperatures varying from
25 to 450 ꢁC.
this indicates the growth of MgF2 crystallites. To further
investigate the effect of temperature on crystallite size, the
Scherrer equation was employed to calculate the average crys-
tallite size of MgF2 based on the diffraction peak at 40.12ꢁ
representing the (111) plane of the MgF2 crystallite in the in situ
XRD patterns shown in Fig. 1. The average MgF2 crystal size as
a function of temperature is plotted in Fig. 2.
As can be concluded from Fig. 2, there is no signicant
change in the average crystal size of MgF2 when the treatment
temperature is below 300 ꢁC, whereas the crystal size increases
ꢁ
signicantly when the temperature increases up to 300 C and
above. Therefore, it can be concluded that the crystal size is very
sensitive to the treatment temperature; this is of signicant use
for the rational optimization of the calcination temperature
during the synthesis of the MgF2 catalysts. Hence, three calci-
nation temperatures were selected and used for the preparation
The different planes of the MgF2 crystallite with corre-
sponding diffraction peaks are shown in Fig. 1(a)–(c). This
validated that all the preparation methods could successfully
produce MgF2 catalysts.
It is also clear that with the increasing temperature, all the
diffraction peaks become sharp at full-width-at-half maxima;
ꢁ
of MgF2 catalysts: 200, 300 and 350 C. In addition, the crystal
23746 | RSC Adv., 2019, 9, 23744–23751
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
and H1 hysteresis loop, indicating the existence of a meso-
porous structure. All the MgF2 samples possess inter-particular
voids with an exception of MgF2-S-200 ꢁC that possesses an
intra-particular mesoporous property. The surface area and the
average pore size of all the catalysts are listed in Table 1. The
surface area generally decreased with an increase in calcination
temperature for all three different preparation methods. When
the calcination temperature was 200 ꢁC, high-surface area MgF2
(HS-MgF2) catalysts were successfully prepared with the surface
area following the trend: MgF2-T (304 m2 gꢀ1) > MgF2-S (215 m2
gꢀ1) > MgF2-P (120 m2 gꢀ1). With an increase in calcination
temperature, the surface area decreased dramatically in all
cases. For example, the surface area of the MgF2 catalysts
prepared using the hard template method decreased from 304
m2 gꢀ1 to 24 m2 gꢀ1 when the calcination temperature was
Fig.
2 Average crystal size of MgF2 crystallites of the catalysts
prepared via three different methods as a function of atmospheric
temperature calculated from the diffraction peak at 40.12ꢁ repre-
senting the (111) plane of the MgF2 crystallite.
ꢁ
increased from 200 to 350 C. This observation shows a good
size generally followed the trend MgF2-T > MgF2-P > MgF2-S at
all temperatures.
agreement with the open literature that calcination at high
temperatures causes a decrease in the surface area of MgF2.17
The average pore size of all the MgF2 catalysts is mainly in the
range from 5 to 33 nm, indicating the mesoporous nature of all
catalysts. Moreover, the MgF2 crystallite size increases notably
with an increase in the calcination temperature; this agrees well
with the in situ XRD results. This veries the application of the
in situ XRD technique for the intelligent selection of catalysts
calcination temperature.
Characterization of the MgF2 catalysts
The MgF2 catalysts synthesized via different methods using
three different calcination temperatures certainly exhibited
different surface properties including surface area, pore struc-
ture, crystal size as well as surface acid properties. The N2
adsorption/desorption isotherms together with the pore size
distribution for the nine MgF2 samples are shown in Fig. 3.
All the MgF2 catalysts synthesized using different methods
and calcination temperatures belong to the type IV isotherm
NH3-TPD analysis
The temperature-programmed desorption of ammonia (NH3-
TPD) was employed to reveal the surface acidic properties of the
MgF2 catalysts. In general, the NH3 desorption peak is
a measure of the strength of the acid sites in the catalyst,
whereas the area of the desorption peak is an indication of the
amount of acid. The NH3-TPD results of the MgF2-S, MgF2-T,
and MgF2-P catalysts calcined at 200 ꢁC (a), 300 ꢁC (b) and
ꢁ
350 C (c) are shown in Fig. 4.
To distinguish between the medium and high-strength acid
sites in the catalysts, the desorption of NH3 at a temperature
less than 350 ꢁC is attributed to the presence of medium-
strength acid sites, whereas the desorption of NH3 at a temper-
ꢁ
ature greater than 350 C is ascribed to the presence of high-
Table 1 Surface area and crystal size of the MgF2 catalysts
Surface areaa
Catalyst
(m2 gꢀ1
)
Pore sizea (nm) Crystal sizeb (nm)
MgF2-P 200 ꢁC 120
30
18
12
19
28
18
5
82
142
226
163
191
217
67
300 ꢁC
350 ꢁC
63
20
MgF2-T 200 ꢁC 304
300 ꢁC
350 ꢁC
97
24
MgF2-S 200 ꢁC 215
300 ꢁC
350 ꢁC
46
22
33
10
91
196
Fig. 3 N2 adsorption/desorption isotherms (a) together with the pore
size distribution (b) for samples: MgF2-P-200 ꢁC, MgF2-P-300 ꢁC,
MgF2-P-350 ꢁC, MgF2-S-200 ꢁC, MgF2-S-300 ꢁC, MgF2-S-350 ꢁC,
MgF2-T-200 ꢁC, MgF2-T-300 ꢁC and MgF2-T-350 ꢁC.
a
b
Calculated from the nitrogen physical desorption isotherm.
Calculated from the XRD results employing the Debye–Scherrer
equation.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 23744–23751 | 23747

View Article Online
RSC Advances
Paper
catalysts, the NH3 desorption peaks appeared at ꢂ350 ꢁC upon
calcination at 200 ꢁC, whereas the NH3 desorption peaks
appeared at ꢂ450 ꢁC upon calcination at 300 and 350 ꢁC. MgF2-
P exhibited lowest amount of acid because it did not have
enough surface defects to provide acid sites; moreover, the
relatively stable structure of MgF2-P with low surface defects did
not experience a signicant change at different calcination
temperatures, which contributed to the relatively uniform
amount of acid at different calcination temperatures. It can also
be observed from Table 2 that the lowest total amount of acid
can be observed on the surface of MgF2 prepared via the
precipitation method. The total amount of acid follows the
order: MgF2-S > MgF2-T > MgF2-P at 200 and 300 ꢁC and MgF2-T
ꢁ
> MgF2-S > MgF2-P at 350 C.
Catalytic reactivity
The MgF2 catalysts synthesized via three different approaches at
three calcination temperatures were applied for the production
of VF via the dehydrouorination of HFC-152a.
The temperature employed for the dehydrouorination of
HFC-152a was kept the same as the calcination temperature of
the catalyst. The dehydrouorination reaction of CF2HCH3
(HFC-152a) in this study exhibited almost 100% selectivity
towards the target product CH2]CHF under the reaction
conditions applied herein with the MgF2 catalysts. This obser-
vation agrees well with the results reported in the open litera-
ture where the reaction of CH2FCF3 (ref. 8) and HFC-152a (ref. 7)
over Lewis acid catalysts shows almost 100% selectivity towards
the dehydrouorination product. As can be observed from
Fig. 5, the conversion rate generally increases with the
increasing temperature, which agrees well with liteꢁrature.7 The
conversion rate was between 8% and 20% at 200 C, between
ꢁ
ꢁ
35% and 55% at 300 C, and between 55% and 85% at 350 C.
The conversion rate at 350 ꢁC was comparable to that obtained
using chromia-based catalysts (Cr2O3) in the dehydrouoriation
of HFC-152a in our previous study.7
Fig. 4 NH3-TPD results of MgF2-S, MgF2-T, and MgF2-P catalysts
calcined at 200 ꢁC (a), 300 ꢁC (b) and 350 ꢁC (c).
strength acid sites on the catalysts. The amount of acid des-
orbed at 350 ꢁC for the MgF2-P-200 ꢁC sample was normalized to
1. Following this, the amount of the medium, strong and total
amount of acid sites on the surface of the MgF2 catalysts was
calculated, and the results are presented in Table 2.
Note that the total amount of acid decreases when the
calcination temperature increases from 300 to 350 ꢁC; this
conrms the conclusion reported by Kemnitz et al.17 that the
acidity of MgF2 gradually diminishes upon heating; however,
for the MgF2 samples prepared by the precipitation method, the
amount of acid does not change signicantly with only minor
changes in the NH3 desorption temperature. For the MgF2-P
MgF2 prepared via the precipitation method showed the
lowest conversion rate at all three temperatures. At 200 C, the
ꢁ
conversion rate follows the order: MgF2-S > MgF2-T > MgF2-P.
This agrees well with the trend of the total amount of acid on
the surface of the catalysts as surface acids contribute to the
active sites for the dehydrouorination reaction. For the reac-
tion temperature above 300 ꢁC, MgF2 prepared via the hard
template method demonstrates the best conversion rate with all
the conversion rates following the order: MgF2-T > MgF2-S >
MgF2-P. Note that the conversion rate obtained using the MgF2
catalysts synthesized using the hard template method out-
performed that obtained using the MgF2 catalysts synthesized
Table 2 Amount of medium, high strength and total acid on the MgF2 catalysts
MgF2-P
200 ꢁC
MgF2-T
MgF2-S
200 ꢁC
300 ꢁC
350 ꢁ
C
200 ꢁ
C
300 ꢁ
C
350 ꢁ
C
300 ꢁ
C
350 ꢁ
C
Med.
Str.
Total
0.0
1.1
1.1
0.0
1.1
1.1
0.0
1.0
1.0
1.2
1.8
3.0
4.2
4.3
8.5
1.4
1.1
2.4
4.8
0.0
4.8
1.4
2.6
4.0
1.0
2.8
3.8
23748 | RSC Adv., 2019, 9, 23744–23751
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
using the sol–gel method when the calcination temperature was Fig. 5(c). This discrepancy is likely due to the fact that the MgF2-
greater than 300 ꢁC. For the conversion rate at 300 ꢁC, the order S catalyst deactivates very quickly when exposed to high
matches with the amount of acid on the surface of the MgF2 temperatures. The conversion rates of the MgF2-T catalyst in the
catalysts prepared by different preparation methods. It can also initial three hours were determined and are shown in Fig. S1,
be observed in Fig. 5(b) that the MgF2-T catalyst demonstrates ESI,† demonstrating its intense deactivation rate with
the best stability with no obvious decrease in the conversion a decrease in the conversion rate from ꢂ81.8% to 77.5% in the
rate aer 25 hours of reaction. However, with same time span, initial 2.5 hours; aer reaction for ꢂ30 hours, the conversion
there is somehow a slight decrease in the conversion rate in the rate upon the MgF2-T catalyst still remains at ꢂ80%, which is
case of both MgF2-S and MgF2-P catalysts. In other words, MgF2- only about 5% less than the initial conversion rate. Note that
T demonstrates the best stability among all three catalysts at this catalytic performance is comparable to that obtained using
ꢁ
300 C.
the chromia-based Lewis acid catalysts reported in open liter-
For the catalytic performance at 350 ꢁC, as shown in Fig. 5(c), ature.7 However, the conversion rate of both MgF2-S and MgF2-P
the MgF2-T catalyst exhibits the best conversion rate and catalysts declined substantially to about 57% during the reac-
stability among all three MgF2 catalysts. However, as shown in tion within same time span. To further investigate the reason
Table 2, there is more relative amount of acid on the surface of for the deactivation of the MgF2 catalysts, XRD and EDS tech-
the MgF2-S catalyst than that on the MgF2-T catalyst; this is not niques were employed to measure the crystal size and surface
exactly consistent with the overall conversion rate shown in carbon content of the MgF2-T, MgF2-S, and MgF2-P catalysts
ꢁ
before and aer reaction for ꢂ30 hours at 350 C. The results
are listed in Table 3. As can be observed from Table 3, there are
no major changes between the surface carbon contents of the
as-prepared and post-reaction MgF2 catalysts. Thus, coking can
be excluded from the main reasons accounting for the deacti-
vation of the MgF2-S and MgF2-P catalysts. On the other hand,
the crystal size of the MgF2 synthesised via the sol–gel method
and precipitation method increased signicantly from below
10 nm to ꢂ20 nm. This is consistent with the in situ XRD results
stating that the size of the MgF2 crystal increases drastically
with the increasing temperature. This is probably the main
reason why the MgF2-S and MgF2-P catalysts easily deactivate
than MgF2-T at high temperatures.
Stability performance
The MgF2-T-300 ꢁC catalysts were further examined for about
600 hours for the dehydrouorination reaction of HFC-152a to
test the stability and lifetime of the catalysts, which are shown
in Fig. 6(a). The conversion rate of the MgF2-S catalyst shows
good stability as it remains greater than 45% at the reaction
time of 600 hours. To the best of our knowledge, the longest
reaction time reported in the open literature is 100 hours using
Cr2O3-based catalysts.7 The MgF2 catalysts reported in this study
have certainly demonstrated superior stability as catalysts for
the dehydrouorination of HFCs.
The XRD patterns were obtained for the MgF2-T-300 ꢁC
catalyst before and aer the 600 hour test to evaluate the
Table 3 Crystal size and surface carbon content of the MgF2-T, MgF2-
S, and MgF2-P catalysts before and after reaction at 350 ꢁC
Catalysts
Crystal size nma Surface carbonb%
MgF2-T-350 ꢁC Fresh
16
3.88
3.45
4.07
4.36
2.91
3.64
Post-reaction 21
Fresh
Post-reaction 19
MgF2-P-350 ꢁC Fresh
Post-reaction 22
MgF2-S-350 ꢁ
C
6
8
Fig. 5 Conversion rate as a function of time on stream for three MgF2
catalysts, including MgF2-S, MgF2-T, and MgF2-P at 200 ꢁC (a), 300 ꢁC
(b) and 350 ꢁC (c).
a
Calculated from the XRD results employing the Debye–Scherrer
equation. b From the EDS measurements.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 23744–23751 | 23749

View Article Online
RSC Advances
Paper
ꢁ
the calcination temperature of 200 C, high-surface area MgF2
catalysts were generated via all preparation methods, whereas at
a calcination temperature greater than 200 ꢁC, the surface area
decreased signicantly; this agreed well with the in situ XRD
results. The catalytic performance of the catalysts followed the
order: MgF2-S > MgF2-T > MgF2-P at 200ꢁꢁC, whereas it was
MgF2-T > MgF2-S > MgF2-P at 300 and 350 C. The order of the
catalytic performance generally agreed with the total amount of
acid measured by the NH3-TPD technique. Among the MgF2
catalysts prepared via different methods, the MgF2 catalysts
synthesized via the hard template method demonstrated best
stability for the dehydrouorination of HFC-152a. Moreover,
the MgF2-T-300 ꢁC catalyst was successfully applied for the
dehydrouorination of HFC-152a for 600 hours, and the
conversion rate remained greater than 45%, showing the
superior long-term stability of this catalyst.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
The authors acknowledge the nancial support received from
the Zhejiang Natural Science Foundation, LY19B060009.
Fig. 6 (a) Conversion rate of the MgF2-T catalyst as a function of
reaction time at 300 ꢁC and (b) XRD patterns of the MgF2-T catalyst
before and after the reaction.
References
changes in the crystal size of the catalyst, as shown in Fig. 6(b).
As can be observed in Fig. 6(b), there are no signicant changes
in the diffraction peaks before and aer the reaction with only
a minor increase in the intensity of the peaks. The crystal size in
the (111) plane of MgF2 was calculated applying the Scherrer
equation. The crystal size increased from ꢂ19.1 nm to
ꢂ22.9 nm aer the 600 hour test. This minor increase in the
crystal size is probably the reason why the MgF2 catalysts
prepared via the template method demonstrate superior long-
¨
¨
1 J. K. Murthy, U. Groß, S. Rudiger, E. Unveren and E. Kemnitz,
J. Fluorine Chem., 2004, 125, 937–949.
2 S. Ebnesajjad and L. G. Snow, Poly(vinyl uoride), in
Encyclopedia of Chemical Technology, ed. J. I. Kroschwitz,
Wiley, New York, 4th edn, 1994, vol. 11, pp. 683–694.
¨
3 A. W. Baker, D. Bonniface, T. M. Klapotke, I. Nicol, J. D. Scott,
W. D. S. Scott, R. R. Spence, M. J. Watson, G. Webb and
J. M. Wineld, J. Fluorine Chem., 2000, 102, 279–284.
4 A. Kohne and E. Kemnitz, J. Fluorine Chem., 1995, 75, 103–
110.
5 G.-L. Li, H. Nishiguchi, T. Ishihara, Y. Moro-oka and
Y. Takita, Appl. Catal., B, 1998, 16, 309–317.
6 J.-W. Luo, J.-D. Song, W.-Z. Jia, Z.-Y. Pu, J.-Q. Lu and
M.-F. Luo, Appl. Surf. Sci., 2018, 433, 904–913.
7 W. Han, X. Li, H. Tang, Z. Wang, M. Xi, Y. Li and H. Liu, J.
Nanoparticle Res., 2015, 17, 365.
8 W. Jia, W. Qian, X. Lang, H. Chao, G. Zhao, J. Li and Z. Zhu,
Catal. Lett., 2015, 145, 654–661.
ꢁ
term stability throughout the 600 hour test at 300 C. In addi-
tion, the surface carbon content of the MgF2-T catalyst aer the
600 hour reaction was examined via the EDS technique. The
surface carbon content of the MgF2-T catalyst aer the 600 h
reaction was revealed to be 4.74%, demonstrating only minor
changes from the initial carbon content before reaction
(3.88%). This indicates that the MgF2-T catalyst has high
resistance towards coking when applied as a catalyst for the
ꢁ
dehydrouorination of HFC-152a at 350 C.
9 M. Wojciechowska, J. Bartoszewicz and J. Goslar,
ChemInform, 1995, 26, 2207–2211.
10 A. Malinowski, W. Juszczyk, J. Pielaszek, M. Bonarowska,
4 Conclusions
´
To manipulate the surface property and rationally design the
catalytic performance of MgF2 catalysts, three synthesis
Z. Karpinski and M. Wojciechowska, Chem. Commun.,
1999, 8, 685–686.
methods were used and compared in the current study. The in 11 M. Wojciechowska, M. Pietrowski, S. Lomnicki and
situ XRD results showed that the crystal size increased
B. Czajka, Catal. Lett., 1997, 46, 63–69.
dramatically when the calcination temperature was greater than 12 W. H. Baur and A. A. Khan, Acta Crystallogr., Sect. B: Struct.
ꢁ
ꢁ
ꢁ
ꢁ
300 C. Moreover, three temperatures, i.e. 200 C, 300 C and
Crystallogr. Cryst. Chem., 2010, 27, 2133–2139.
350 C, were investigated to further evaluate the effect of calci- 13 A. Saberi, Z. Negahdari, S. Bouazza and M. Willert-Porada, J.
nation temperature on the surface properties of the catalysts. At
Fluorine Chem., 2010, 131, 1353–1355.
23750 | RSC Adv., 2019, 9, 23744–23751
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
14 E. Kemnitz, Y. Zhu and B. Adamczyk, J. Fluorine Chem., 2002, 19 A. B. D. Nandiyanto, O. Takashi and O. Kikuo, ACS Appl.
114, 163–170. Mater. Interfaces, 2014, 6, 4418–4427.
15 E. Kemnitz and S. Rudiger, High Surface Area Metal Fluorides 20 M. Chen, J. M. Jin, S. D. Lin, Y. Li, W. C. Liu, L. Guo, L. Li and
as Catalysts, 2010. X. N. Li, J. Fluorine Chem., 2013, 150, 46–52.
16 I. Agirrezabal-Telleria, F. Hemmann, C. Jager, P. L. Arias and 21 J. Noack, K. Teinz, C. Schaumberg, C. Fritz, S. Rudiger and
¨
¨
¨
E. Kemnitz, J. Catal., 2013, 305, 81–91.
E. Kemnitz, J. Mater. Chem., 2010, 21, 334–338.
´
¨
17 J. Krishna Murthy, U. Groß, S. Rudiger, E. Kemnitz and 22 M. Pietrowski, M. Zielinski and M. Wojciechowska, Pol. J.
J. M. Wineld, J. Solid State Chem., 2006, 179, 739–746.
Chem. Technol., 2014, 16, 63–68.
18 I. Agirrezabal-Telleria, Y. Guo, F. Hemmann, P. L. Arias and 23 A. Saberi, Z. Negahdari, S. Bouazza and M. Willert-Porada, J.
E. Kemnitz, Catal. Sci. Technol., 2014, 4, 1357–1368.
Fluorine Chem., 2010, 131, 1353–1355.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 23744–23751 | 23751