Study on the Preparation of Cr-based Catalysts Doped by Zn using Sol–Gel Auto-Combustion Method and Its Application for Synthesis of 1-Chloro-2,3,3,4,4,5,5-heptafluorocyclopentene

Article abstract of DOI:10.1002/jccs.201700121

High-surface-area chromium-based catalysts in the presence of a small amount of zinc were prepared via a sol–gel auto-combustion method using chromic nitrate, zinc nitrate, and citric acid. First, the auto-combustion behavior of the dried gel was investigated by derivative thermogravimetry and (DTG)-TG and infrared (IR) techniques. The results revealed that the dried gel exhibited self-propagating combustion properties. Second, the as-burnt powders were characterized by IR, X-ray diffraction (XRD), Brunauer–Emmett–Teller analysis (BET), and scanning electron microscopy (SEM). The findings showed that the gels were directly converted into CrZn-O nanoparticles with high surface area during the auto-combustion process. Third, the pre-fluorination Cr-Zn catalysts were characterized by XRD, BET, SEM, X-ray photoelectron spectroscopy (XPS), and Fourier transform (FT)-IR spectroscopy of pyridine adsorption techniques. It was found that the presence of zinc led to significant structural changes in the catalyst, the particle size was smaller, the surface area became larger, and more active sites appeared. Finally, the catalytic activities of the samples were tested for the fluorination of 1,2-dichlorohexafluorocyclopentene (1,2-F6) with anhydrous hydrogen fluoride. The obtained results indicated that the pre-fluorination activated Cr-Zn catalysts prepared by this sol–gel auto-combustion method exhibited high efficiency in the synthesis of cyclic hydrofluorocarbons.

Full text of DOI:10.1002/jccs.201700121

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Article
Study on the Preparation of Cr-based Catalysts Doped by Zn using Sol–Gel Auto-
Combustion Method and Its Application for Synthesis of 1-Chloro-2,3,3,4,4,5,5-
heptafluorocyclopentene
a
a
a
b*
Yajun Han, Jinwei He, Zhenhai Wu and Xiaomeng Zhou
a
College of Environmental Science and Engineering, Nankai University, Tianjin 300071, China
Economics and Management College, Civil Aviation University of China, Tianjin, China
b
(Received: April 5, 2017; Accepted: May 18, 2017; DOI: 10.1002/jccs.201700121)
High-surface-area chromium-based catalysts in the presence of a small amount of zinc were pre-
pared via a sol–gel auto-combustion method using chromic nitrate, zinc nitrate, and citric acid.
First, the auto-combustion behavior of the dried gel was investigated by derivative thermogravi-
metry and (DTG)-TG and infrared (IR) techniques. The results revealed that the dried gel exhib-
ited self-propagating combustion properties. Second, the as-burnt powders were characterized by
IR, X-ray diffraction (XRD), Brunauer–Emmett–Teller analysis (BET), and scanning electron
microscopy (SEM). The findings showed that the gels were directly converted into CrZn-O nano-
particles with high surface area during the auto-combustion process. Third, the pre-fluorination
Cr-Zn catalysts were characterized by XRD, BET, SEM, X-ray photoelectron spectroscopy
(XPS), and Fourier transform (FT)-IR spectroscopy of pyridine adsorption techniques. It was
found that the presence of zinc led to significant structural changes in the catalyst, the particle size
was smaller, the surface area became larger, and more active sites appeared. Finally, the catalytic
activities of the samples were tested for the fluorination of 1,2-dichlorohexafluorocyclopentene
(1,2-F6) with anhydrous hydrogen fluoride. The obtained results indicated that the pre-fluorination
activated Cr-Zn catalysts prepared by this sol–gel auto-combustion method exhibited high effi-
ciency in the synthesis of cyclic hydrofluorocarbons.
Keywords: Catalytic fluorination; Sol–gel auto-combustion method; 1-Chloro-2,3,3,4,4,5,5-
heptafluorocyclopentene; Chromium oxide; Zinc.
INTRODUCTION
Chlorofluorocarbons (CFCs), hydrochlorofluoro-
carbons (HCFCs), and hydrofluorocarbons (HFCs)
were widely used as refrigerants, foaming agents, clean-
ing agents, and etching agents in the past few years.
However, because of their ozone depletion potential
Nowadays, the most common method to synthe-
size HFCs is vapor-phase catalytic fluorination of
chlorinated hydrocarbons. However, the easily formed
oligomers and cokes for c-HFCs make a more rigorous
requirement for catalysts compared to HFCs. The cur-
rently known methods to produce c-HFCs commonly
occur in the liquid phase, which is polluting and uneco-
nomical. Thus, it is necessary for the industry to
develop appropriate catalysts to produce c-HFCs by
vapor-phase catalytic fluorination successfully.
(ODP) and high global warming potential (GWP), the
use of these haloalkanes has been limited by the Mont-
1–6
real Protocol and its subsequent amendments.
Recently, cyclic HFCs (c-HFCs), especially five-carbon
rings, were identified as ideal alternatives to these
because of their short atmospheric lifetimes, zero ODP,
For the fluorinated catalysts, the acid sites, surface
area, particle size, and composition of the catalysts are
important factors for their catalytic activity. Hence,
nanoscale materials with high surface area and tunable
surface properties have the potential to be used in
7,8
and low GWP. However, due to the ring molecular
structure and even unsaturated bonds, it is a serious
challenge to achieve an efficient synthetic route for the
c-HFCs.
9
c-HFCs synthesis. Besides, chromium-based materials
*Corresponding author. Email: yajunhan110@163.com
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Article
Han et al.
are well-known catalysts for the F/Cl exchange
10–12
reactions.
However, due to their high melting
point, oxidation resistance, and poor sintering ability,
nano-sized chromium materials are very difficult to pre-
pare, and only a few patents have been reported up to
13
now. So the search for low-cost, high-efficiency nano-
chromium-based catalyst production methods is still
going on.
To produce nanoscale catalysts, many approaches
1
4
15
such as plasma fluorination, the sol–gel technique,
1
6
17
hydrothermal method, and micro-emulsion synthesis
have been discussed. Compared to these methods, sol–
gel auto-combustion method has many advantages such
18,19
as cheap precursors and a simple preparation route,
Fig. 1. TG and DTG thermograms of the dried
Many superfine metal oxide powders with superior sur-
citrate–nitrate gel CrZn(0) under air at the
face and bulk properties have been prepared through
ꢀ
heating rate of 10 C/min.
2
0–24
this method in the materials science field.
There-
fore, the preparation of nano-scale chromium-based
fluorination catalyst through sol–gel auto-combustion
method has great value for industrial applications.
In this work, we prepared high-surface-area
chromium-based catalysts in the presence of a small
amount of zinc via the sol–gel auto-combustion method
using chromic nitrate, zinc nitrate, and citric acid. Then,
the dried gels, the as-burnt powders, and the activated cat-
alyst were characterized by thermogravimetry-differential
thermal analysis (TG-DTA), infrared (IR) spectroscopy,
X-ray diffraction (XRD), Brunauer–Emmett–Teller (BET)
analysis, scanning electron microscopy (SEM), X-ray pho-
toelectron spectroscopy (XPS), and Fourier transform
infrared (FT-IR) spectroscopy, Finally, the catalytic per-
formance of the prepared catalysts was evaluated through
the fluorination of 1,2-dichlorohexafluorocyclopentene
ꢀ
the temperature range 160–250 C, which is attributable
to the auto-catalytic anionic oxidation–reduction reac-
tion between the nitrate and citrate. The decomposition
of NH NO produced by pH adjustment of the solution
also took place in this temperature range. During the
oxidation–reduction reaction, the nitrate ions provide
an in situ oxidizing environment for the decomposition
of the organic components, lowering the decomposi-
tion temperature. As expected, abundant amounts of
gases such as CH , H , H O, and CO were released.
The peak of the DTG curve in the temperature range
60–250 C was relatively sharp, and the weight change
associated with this reaction was ~25%. The final step
in weight loss, which occurred between 250 and 500 C,
signified the burning of the residual combustible mate-
rial and decomposition of the carbonate complex to
chromium oxide. There was a sharp peak in the DTG
curve with a weight loss of ~40% in TG. As the temper-
ature was further raised, there was no further weight
loss and the DTG was zero. The weight loss for the
entire process was ~77%. Notably, the thermal behavior
of the other three gel precursors [CrZn(5), CrZn(15),
and CrZn(25)] was basically the same as that of CrZn
4
3
25
26
4
2
2
2
ꢀ
1
ꢀ
(1,2-F6) with anhydrous hydrogen fluoride.
RESULTS AND DISCUSSION
Auto-combustion behavior
Experimental observation showed that the dried
gel exhibited self-propagating combustion behavior.
The curves of thermal analysis (TG and DTG) were
plotted to study the auto-catalytic combustion process
of the nitrate–citrate gel.
(
0), which can be seen in Table 1. This is because the
burning of the gels involved the nitrate and citrate ions
but not the metal cation.
From Figure 1, it can be seen that the whole
pyrolysis process of the dried gel can be divided into
three stages. The first stage of weight loss in the
Characterization of gels and as-burnt powders
ꢀ
50–160 C range corresponds to the removal of residual
Figure 2 shows the IR spectra of the dried gels in
−1
moisture. The second stage of weight loss occurred in
the range 400–4000 cm . The dried gel showed a
2
www.jccs.wiley-vch.de
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Chin. Chem. Soc. 2017

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Catalytst Produced by Sol–Gel Auto-Combustion
Table 1. TG of the CrZn(x) gels
−1
CrZn(15)-O and to 516 cm for CrZn(25)-O, which
may be due to the production of ZnCr O . Remarka-
2
4
Auto-catalytic
anionic oxidation–
Burning of the
residual
−
1
bly, the band at ~950 cm , which is due to the stretch-
x−
Dehydration
reduction
reaction ( C)
combustible
ing vibration of Cr–O–Cr in CrO4 (x = 2, 3, or 4),
ꢀ
ꢀ
ꢀ
Samples
( C)
material ( C)
indicates that Cr with high valence (>3) was pres-
32,33
ent.
The IR results show that the citrate ions and
CrZn(5)
CrZn
50–155
50–155
155–245
155–245
>245
>245
the nitrate ions fade away, and the Cr O and chromite
2
3
(
15)
CrZn
25)
phases gradually appear during the combustion process.
Combining the TG-DTG and IR results, we can
conclude that the clathrate compound turned into inor-
ganic substances after the combustion process. The
organic functional groups relevant to carboxylate spe-
cies and ions took part in the oxidation–reduction reac-
50–150
150–245
>245
(
–1
characteristic band at ~1384 cm , corresponding to the
antisymmetric stretching vibration. The “steamed
bread” peak appearing at 3400–2200 cm represented
the O–H group. The band at 1720 cm in the spectrum
corresponds to the terminal free carboxyl groups of cit-
ric acid, which means that there was citric acid
remaining in the gel. The characteristic bands at
1620 cm represents the asymmetric stretching vibra-
tions of the carboxyl ions (COO ), while the band at
–1
ꢀ
tion when the temperature was above 160 C. A large
−
1
quantity of gas was released during the burning process.
This left many tiny pores and resulted in a burnt sample
with a high surface area. Finally, the chromium oxides
to be used as catalyst precursors were obtained.
2
7
−
1
~
−
Characterization of catalysts
−1
1
(
550 cm is attributed to the asymmetric stretches of
OCO) in the carboxylate species.
2
8–31
XRD of as-burnt powders and pre-fluorinated catalysts
-
Powder XRD studies were carried out on the burnt
powders and the pre-fluorinated catalysts. No diffrac-
tion peaks related to citric acid or citrate were observed,
implying that the combustion proceeded completely. In
Figure 4(a), it can be seen that for the CrZn(0)-O and
CrZn(5)-O samples, only diffraction peaks due to crys-
talline Cr O are observed and no diffraction peaks of
Figure 3 shows the IR curves of the as-burnt pow-
der. The characteristic bands of the carboxyl group and
the ion in the IR spectrum appeared weak after com-
−1
bustion. The bands at ~560 and 620 cm are associ-
ated with the vibration mode of O–Cr–O bond,
1
3
confirming the formation of the Cr O phase. The
2
3
−
1
−1
2
3
characteristic band at 560 cm shifted to 538 cm for
zinc or zinc oxide are found, which indicates that the
Fig. 2. IR spectra of the dried gels.
Fig. 3. IR spectra of the as-burnt powders.
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jccs.wiley-vch.de
3

Article
Han et al.
Zn species are well dispersed on the Cr O surface as
On comparing Figure 4(a) and (b), it can be seen
that the diffraction peaks of the pre-fluorination cata-
lysts and the as-burnt powder are almost identical,
which suggests that the phase structure and composi-
tion remain stable during the fluorination. The crystal-
line phase CrF ꢁ3H O was found in the CrZn(5)-F
2
3
an amorphous or microcrystalline phase at low concen-
trations. There were both Cr O and ZnCr O phases
2
3
2
4
in the CrZn(15)-O, and only ZnCr O phase with a spi-
2
4
nel structure was found in CrZn(25)-O. This indicates
that crystalline ZnCr O is easily formed for high Zn
2
4
3
2
content. The diffraction peaks became relatively broad
and weak with increased addition of Zn, which con-
catalyst. It suggests that a small amount of Zn helps in
reducing the crystallinity of Cr O and forms unstable
2
3
3
4
firms the viewpoint that an appropriate dopant causes
Cr species which are easily fluorinated. The diffraction
ꢀ
a decrease in the crystallinity of the host lattice.
peak at ~26.4 (2θ) is due to the graphite that was used
in the mold. CrF and CrF ꢁ3H O are difficult to form
3
3
2
because the free reaction enthalpy (5 G = −68 kJ/mol)
R
for the formation of CrF from the reaction of Cr O
with HF is only slightly less than zero.
3
2
3
35
Morphology of as-burnt powders and pre-fluorinated
catalysts
SEM micrographs (Figure 5) show the
pre-fluorinated catalyst as well as the as-burnt powder,
which are composed of much smaller irregular-shaped
particles. The particle size of the as-burnt powder was
less than 100 nm, and it became ~200 nm after the pre-
fluorination process. It can also be seen from the SEM
images that all catalysts had a mesoporous structure
and there were many micropores of irregular shape.
High-surface-area materials have unusual proper-
ties in heterogeneous catalysis, where the activity of the
36
catalyst depends largely on its surface area. The con-
tact of the reactant gas with this high-surface catalyst
would be enhanced, thus improving the reaction effi-
ciency. The BET surface area, pore volume, pore diam-
eter, and elemental analysis data of the fluorinated
catalyst together with the data of the as-burnt powder
are summarized in Table 2.
The surface area of the single-component oxide
2
CrZn(0)-O is 124.8 m /g. As the Zn content increases,
the surface area of the CrZn(x)-O samples increases
gradually, which suggests that the addition of Zn facili-
tates the high dispersion of the catalyst crystallites and,
as a result, increases the BET surface area of the chro-
37,38
mium oxide-based catalyst.
The surface area of the
CrZn(x)-F catalysts first increases and then decreases
upon further increasing the Zn, and CrZn(15)-F, with
2
the largest surface areas of up to 139.4 m /g, is the most
accessible. The surface area was reduced by 41.75,
4
7.63, 12.59, and 23.94% for CrZn(0)-O, CrZn(5)-O,
Fig. 4. XRD patterns of (a) burnt powders and
(
b) pre-fluorinated catalysts.
CrZn(15)-O, and CrZn(25)-O, respectively. The
4
www.jccs.wiley-vch.de
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Chin. Chem. Soc. 2017

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Catalytst Produced by Sol–Gel Auto-Combustion
37
According to the literature, the peak at 577.5 eV is
attributed to the formation of CrO F species, which
x
y
40–42
was deemed to be the active sites of fluorination.
Wagner indicated that the Cr O and ZnCr O had
43
2
3
2
4
the same Cr 2p3/2 BE, which is 575.5 eV in this study.
Table 3 lists the BEs of Cr 2p3/2 and CrO F sur-
x
y
face contents of the CrZn(x)-F catalysts. The surface
CrO Fy content was calculated by the peak area of
x
CrO F divided by the total peak area of the Cr species.
x
y
As can be seen in Table 3, with the addition of Zn, a
shift toward a higher BE of the CrO Fy peak was
x
observed, which could be attributed to the interaction
of Zn with Cr. Additionally, the surface CrO F con-
x
y
tent in the CrZn(x)-F catalysts varied depending on the
Zn addition. CrZn(15)-F had the highest content of sur-
face CrO F species (71.2%), while the surface CrO F
x y
x
y
content of the CrZn(0)-F catalyst was the low-
est (64.3%).
The Zn 2p3/2 XPS spectra of the pre-fluorination
catalysts were analyzed as well. Figure 8 shows that
after the treatment by HF, the zinc in the CrZn(5)-O
sample was nearly completely fluorinated to ZnF . The
2
presence of the mixed phase of ZnCr O and ZnF was
2
4
2
also detected in the CrZn(15)-F catalyst. However, only
the ZnCr O phase was found in the CrZn(25)-F sam-
Fig. 5. SEM images of (a) CrZn(0)-O and (b) CrZn
2
4
(
0)-F catalysts.
ple. According to Ref. 37 Zn at a relatively low concen-
tration is more easily fluorinated. For a high Zn
content, a simultaneous increase in the formation of
aggregates of crystalline ZnCr O makes it difficult to
decrease in the specific surface area could correspond to
the formation of fluorides or oxyfluorides.
3
9
2
4
be fluorinated. Thus, when the amount of Zn on the
catalyst surface is too low or too high, the formation of
Cr O or ZnCr O , which are difficult to fluorinate, is
A pore size between 2 and 50 nm, called mesopor-
ous, was observed for all of the samples. Figure 6 shows
the N2 sorption isotherms of the oxides and pre-
fluorination catalysts. All samples displayed type IV
isotherms, indicating that they are porous materials.
Indeed, as can be seen in Table 2, the pre-fluorinated
catalysts had a narrower pore size distribution than the
catalyst precursors. For all CrZn(x)-O samples, the spe-
cific surface and pore volume decreased after the fluori-
nation step. The pore volume was reduced by 28.6,
2
3
2
4
favored and the concentration of other species such as
CrO F decreases.
x
y
Surface acidity
Lewis acids are commonly used as
catalysts in fluorination reactions. It has been proposed
that fluorination and disproportionation catalysts must
44
possess Lewis acidity for their activities. The FT-IR
photoacoustic spectra of adsorbed pyridine were used
for characterizing and identifying the type of acid sites
in post-fluorinated catalysts. The results (Figure 9) con-
firm the presence of strong Lewis (L-Py) acid sites, as
40, 15.4, and 14.3%.
Surface composition
XPS experiments were conducted
to further investigate the Cr species on the catalysts sur-
face. The Cr 2p XPS spectra of the pre-fluorination cata-
lysts are shown in Figure 7. For all catalysts, the broad
peak of Cr 2p3/2 could be de-convoluted into two peaks
with binding energies (BEs) of 577.5 and 575.5 eV.
−
1
shown by the vibrational bands at 1454 and 1612 cm .
−
1
Brønsted acid sites (B-Py) at 1540 cm
were also
−
1
found. Simultaneously, the band at ~1490 cm can be
assigned to both Lewis and Brønsted acid sites. Patil
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jccs.wiley-vch.de
5

Article
Han et al.
Table 2. Textural properties of the oxides and pre-fluorination catalysts
CrZn(x)-O
CrZn(x)-F
a
2
b
3
c
a
2
b
3
c
Addition of Zn (mol%)
SBET (m /g)
V (cm /g)
dp (Å)
SBET (m /g)
V (cm /g)
dp (Å)
0
5
1
2
a
124.8
146.7
159.5
178.7
0.14
0.15
0.13
0.14
55
50
37
35
72.7
76.8
139.4
135.9
0.1
58
47
35
38
0.09
0.11
0.12
5
5
BET surface area.
b
c
BJH cumulative desorption pore volume.
Mean pore diameter = 4V/SBET
.
of Lewis acid sites in CrZn(25)-F was the highest, and
that of CrZn(0)-F was the lowest. However, the number
of Brønsted acid sites first increased and then decreased
upon increasing the Zn content, and the number of
Brønsted acid sites in CrZn(15)-F catalyst was distinctly
less than that in the other tested samples
A series of characterization experiments were car-
ried out to further investigate the structure and proper-
ties of the catalysts prepared by the sol–gel auto-
combustion method. It was found that the prepared cat-
alysts had a much smaller particle size than those pre-
pared by precipitation, which had micrometer-sized
particles. For all CrZn(x)-F catalysts, the BET surface
areas were higher than those of the Cr-based catalysts
prepared conventionally by precipitation from aqueous
Fig. 6. Adsorption–desorption isotherms of the oxide
samples and pre-fluorination catalysts:
(
(
(
(
a1) CrZn(0)-O, (a2) CrZn(0)-F; (b1) CrZn
5)-O, (b2) CrZn(5)-F; (c1) CrZn(15)-O,
c2) CrZn(15)-F; (d1) CrZn(25)-O, (d2) CrZn
25)-F.
2
media, exhibiting surface areas in the range 10–60 m /
46
g. Chromium-based catalyst with 5 mol% zinc was
prepared as well by the precipitation method in our lab-
oratory. Post fluorination, the catalyst precursor was
transformed into chromium fluoride, and only the
CrF ꢁ3H O phase, which showed very low catalytic
3
2
28
activity, was found in the XRD. The content of
CrO F was ~66.4%, which is lower than that of the
x
y
CrZn(5)-F catalyst prepared by sol–gel auto-combus-
tion. The characterization results showed that the
x y
Table 3. Binding energy of Cr 2p3/2 and CrO F surface
content of CrZn(x)-F catalysts
Cr 2p3/2/eV
Cr O /
Surface CrO Fy
content (%)
2
3
x
Fig. 7. XPS spectra of the Cr 2p level for CrZn(x)-F
Catalysts
CrO Fy
ZnCr O4
x
2
catalysts.
CrZn(0)-F
CrZn(5)-F
CrZn(15)-F 578.1
CrZn(25)-F 578.2
577.5
578.3
575.5
575.9
575.6
575.9
64.3
70.8
71.2
68.5
4
5
−1
et al. noted that the peak at ~1490 cm is indicative
of Brønsted sites only if its intensity exceeds that of the
peak at ~1450 cm . As seen from Figure 9, the number
−1
6
www.jccs.wiley-vch.de
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Chin. Chem. Soc. 2017

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Catalytst Produced by Sol–Gel Auto-Combustion
certain degree of catalytic activity for the fluorination
of 1,2-F6. Especially, the conversion of 1,2-F6 reached
ꢀ
3
1.7% on the CrZn(15)-F catalyst at 430 C and its
selectivity was 85.2%.
From Table 4, it can also be seen that the addition
of Zn has a great influence on the catalytic activity. An
increase of the activity for the synthesis of F7 was
noticed when a small amount of zinc was added to the
chromium-based catalyst; the conversion of 1,2-F6 and
the selectivity to F7 reached maximum values at the Zn
content of 15%. However, with further increase in the
zinc content, the activity of CrZn(25)-F decreased obvi-
ously. On the other hand, the conversion of 1,2-F6
increased with increase in the reaction temperature for
all the prepared catalysts. Yet, the selectivity of F7
declined with increase in the reaction temperature, and
more of the by-product F8E was observed. When the
Fig. 8. XPS spectra of the Zn 2p3/2 level for catalysts
after treatment with HF.
ꢀ
reaction temperature reached 430 C, the selectivity of
F7 was only 69.7% for CrZn(25)-F and the oligomers
were detected.
catalysts prepared by the sol–gel auto-combustion
method had smaller particle size, higher surface area,
and higher content of surface active sites, all of which
indicated that the catalyst prepared by this novel
method should possess higher catalytic activity.
Figure 10 shows the stability of the CrZn(x)-F cat-
ꢀ
alysts at 400 C under reaction conditions. It can be seen
that the conversion of 1,2-F6 slightly fluctuated in the
initial period and then reached steady state after 3 h.
The selectivity to F7 was below 80% during the first
Performance of catalysts
To investigate the catalytic
3
h, but immediately increased to ~90% and stabilized
activity of the as-prepared CrZn(x)-F catalysts, the
fluorination of 1,2-F6 to synthesize F7 was studied. As
shown in Table 4, it is obvious that all catalysts show a
for the CrZn(15)-F catalyst, which indicates that cata-
lysts were quite stable during the reaction.
The unreacted 1,2-F6 was separated from the pro-
ducts by distillation and then recycled into the reactor
system as raw materials. By this means, a method to
produce F7 with a high yield was obtained. This
demonstrates that the catalysts prepared by this novel
auto-combustion method have a great value in indus-
trial application.
An increase in the activity for the formation of F7
was noted when a small amount of zinc was added to
the chromium-based catalyst. This is because the phase
composition of the as-burnt powder CrZn(15)-O is
complex and metastable, which is beneficial to the for-
mation of the Cr active site CrO F (71.2% for CrZn
x
y
(
15)-F catalyst) after pre-fluorination. The BET surface
2
area of the CrZn(15)-F catalyst is the largest (139.4 m /
g). Besides, the number of Brønsted acid sites in CrZn
(15)-F is less than that in the other prepared catalysts,
Fig. 9. FTIR photoacoustic spectra of pyridine
which means that it has a higher relative number of
Lewis acid sites. Thus the conversion of 1,2-F6 and the
chemisorbed on CrZn(x)-F samples.
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jccs.wiley-vch.de
7

Article
Han et al.
Table 4. Effect of reaction temperature to 1,2-F6 conversions and F7 selectivity on CrZn(x)-F catalysts
ꢀ
Temperature ( C)
3
50
380
400
430
Sample
C (%)
S (%)
C (%)
S (%)
C (%)
S (%)
C (%)
S (%)
CrZn(0)-F
CrZn(5)-F
CrZn(15)-F
CrZn(25)-F
3.7
4.6
7.1
2.9
90.7
92.5
94.3
85.6
8.6
11.6
19.6
9.8
87.4
88.7
89.2
79.3
16.9
20.7
24.9
16.3
85.5
86.7
90.3
76.5
21.7
27.2
31.7
21.9
79.8
82.3
85.2
69.7
Reaction conditions: HF/F6 molar ratio of 10 with contact time 12 s.
C, conversion in %; S, selectivity in %.
selectivity to F7 reach a maximum at the Zn content of
5%. However, upon further increasing the Zn content,
the activity of CrZn(25)-F catalyst clearly decreased.
This implies that when the amount of Zn on the cata-
lyst surface is too high, the formation ZnCr O , which
1
2
4
is difficult to fluorinate, is favored and the Cr active
sites on CrO F decreases. The number of Lewis acid
x
y
sites in CrZn(25)-F was the largest. But olefin molecules
could be easily adsorbed on the surface acid sites of the
33,34
catalyst,
and the high surface acidity of the catalyst
would result in rapid coke formation that blocks the
35
active sites, leading to a decrease in the specific rate
of the reaction.
EXPERIMENTAL
Chemicals
Hexachlorocyclopentadiene was obtained from J&K
Chemical Technology. Cr(NO ) ꢁ9H O (>99%) and
3
3
2
Zn(NO ) ꢁ6H O (>99%) were obtained from Yunli Chem-
3
2
2
ical Industry Co., Ltd., Shanxi, China. C H O ꢁH O
6
8
7
2
(
>99.5%) and NH ꢁH O (25–28%) were from Beijing
3
2
Chemical Works Co., Ltd. AHF (>99%) was obtained
from Sinochem Group.
Catalysts preparation
The nano-size chromium-based oxide was pre-
pared by a unique combination of the chemical sol–gel
process and the combustion process. First, stoichiomet-
ric amounts of chromic nitrate nonahydrate (Cr
NO ) ꢁ9H O) and zinc nitrate hexahydrate (Zn
(
(
3
3
2
NO ) ꢁ6H O) powders were dissolved in distilled
3
2
2
water. The total molarity of the metal ions was
0
1
0
3+
2+
Fig. 10. Relationship between the conversion of 1,2-
.1 mol/L in solution (the mole rate of Cr :Zn is
00:0, 95:5, 85:15, 75:25 respectively). Then,
.1 mol/L citric acid solution was added dropwise to
F6 and the selectivity to F7 over CrZn(x)-F
a
ꢀ
catalysts at 400 C and time on stream.
8
www.jccs.wiley-vch.de
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Chin. Chem. Soc. 2017

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Catalytst Produced by Sol–Gel Auto-Combustion
the above solution till the molar ratio of metal ions/cit-
min) and AHF (50 mL/min) was passed through reac-
tor at 250 C for 8 h. Then, the N2 flow rate was
ꢀ
rate ions reached 1:1.2. NH ꢁH O was added to the
3
2
nitrate–citric solution, the pH value was adjusted to
decreased to 50 mL/min and AHF flow rate increased
ꢀ
3, and the mixture solution was prepared.
to 100 mL/min at 250 C. This mixture was fed into the
The mixture solution was thoroughly stirred for
2 h at 70 C, and water in the solution was vaporized
reactor for 10 h. Subsequently, N flow was stopped
and the sample was heated at 350 C for 10 h in AHF
2
ꢀ
ꢀ
1
gradually, when the solution become sticky. This sticky
liquid was transferred into a drying oven and heated for
at a flow rate of 150 mL/min. Finally, the pre-
fluorination catalysts, denoted as CrZn(0)-F, CrZn(5)-
F, CrZn(15)-F, and CrZn(25)-F, respectively, were
formed.
ꢀ ꢀ
24 h at 80 C and 4 h at 120 C to obtain a dried gel.
The dried gel was ignited and burned in a self-
propagating combustion manner until it entirely chan-
ged to a loose powder. A schematic diagram of the
preparation process is shown in Figure 11. Various
Characterization of catalysts and gels
The characteristics and kinetics of the decomposi-
tion of the gel precursors were examined by a Perkin-
3
+
2+
samples with different mole rates of Cr /Zn was
obtained by varying the relative proportions of
Cr(NO ) ꢁ9H O and Zn(NO ) ꢁ6H O. For mole rates
ꢀ
Elmer TGA-2 instrument at the heating rate of 10 C/
min under air and nitrogen. The sample weight was
~5 mg, and the gas flow rate was kept at 100 mL/min.
The IR spectra of the as-burnt powder were
recorded on a Bomem DA 3.002 spectrometer from
3
+
3
2
3 2
2
3
2+
of Cr :Zn 100:0, 95:5, 85:15, and 75:25, the corre-
sponding dried gels and burned dried gel were denoted
as CrZn(0), CrZn(5), CrZn(15), CrZn(25), and CrZn
−
1
(
0)-O, CrZn(5)-O, CrZn(15)-O, and CrZn(25)-O,
respectively.
Before the reaction, pre-fluorination was carried
400 to 4000 cm by the KBr pellet method.
The structural characterization of the powders
was performed by XRD using a Siemens D500 diffrac-
ꢀ
out to activate all catalyst precursors. A 5 g sample of
tometer with Cu Kα radiation in the 10–80 2θ angle
ꢀ
the precursors was packed into the reactor. N2
range at the scan rate of 0.3 /min. Phase composition
ꢀ
47
(
150 mL/min) was passed through reactor at 250 C for
and cell parameter were calculated by the Rietveld
4
h to remove the water, which was followed by heating
method using the MID-JADE 6.5 software.
The specific surface areas were determined by the
classical BET procedure using a Coulter SA 3100
apparatus.
ꢀ
at 350 C for 4 h to decompose the residual organics in
catalyst precursors. Then, a mixture of N (100 mL/
2
The microstructure of the resulting ceramic was
observed by SEM (Hitachi X-650).
XPS analyses were carried out using an M-probe
apparatus (Surface Science Instruments). The accuracy
of the reported BEs was estimated to be ꢂ0.2 eV.
FT-IR photoacoustic pyridine adsorption spec-
troscopy was carried out according to the following
ꢀ
procedure. A sample (80 mg) was pretreated at 150 C
under a nitrogen flow of 35 mL/min for 15 min. Then,
6
0 mL pyridine was injected into the sample tube. The
sample was flushed with nitrogen for 30 min to remove
the physisorbed pyridine. Spectra of the sample were
recorded at room temperature using an MTEC cell and
FT-IR system 2000 (Perkin-Elmer).
The apparatus for the vapor-phase fluorination
reaction was composed of a pump for transferring the
Fig. 11. Diagram of the preparation of nano-sized
liquid-phase reactant, HF, and N mass flow control-
2
CrZn(x)-O powders.
lers, and an electrically heated tubular Inconel reactor
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jccs.wiley-vch.de
9

Article
Han et al.
the fluorination of 1,2-F6, with the catalyst maintaining
high activity and stability. Besides, the zinc content of
the catalyst had a great influence on the structure and
activity of the chromium-based catalyst, which illus-
trated that the addition of the promoter Zn should be
moderate. It can therefore be concluded that the sol–gel
auto-combustion method is very suitable for the prepa-
ration of high-surface-fluorination catalysts, and zinc is
a very good doping agent for chromium-based catalysts
used in the vapor-phase Cl/F exchange.
Fig. 12. Schematic diagram of the synthesis of 1-
chloro-2,3,3,4,4,5,5-heptafluorocyclopentene
F7).
(
REFERENCES
1
2
. M. J. Molina, F. S. Rowland, Nature 1974, 249, 810.
. R. J. Salawitch, S. C. Wofsy, M. B. McElroy, Planet.
Space Sci. 1988, 36, 213.
(
14 mm in diameter and 300 mm in length) equipped
with an inner Inconel tube for the insertion of type-K
thermocouples with 1 mm diameter. The thermocouple
enters the reactor through a Monel-type fitting and
extends into the catalyst bed to measure the tempera-
ture changes at different positions along the reactor.
3
. S. C. Wofsy, M. B. McElroyp, N. D. Sze, Science 1975,
1
87, 535.
4. H. Yu, E. M. Kennedy, A. Uddin, S. P. Sullivan,
B. Z. Dlugogorski, Int. J. Chem. Kinet. 2005, 37, 134.
5
. Y. Hu, Y. Xu, T. Wang, C. Wang, S. Li, J. Anal. Appl.
Pyrolysis 2011, 90, 27.
The synthesis of F7 from 1,2-F6
As shown in Figure 12, 1,2-F6 was prepared by
the vapor-phase fluorination of hexachlorocyclopenta-
diene. The details are given in the patent application
filed by our laboratory (CN201210254955.0).
Next, to prepare F7 and F8E, 1,2-F6(10 mL/min)
6. T. J. Wallington, W. F. Schneider, D. R. Worsnop,
O. J. Nielsen, J. Sehested, W. J. Debruyn, J. A. Shorter,
Environ. Sci. Technol. 1994, 28, 320A.
7
8
. A. McCulloch, J. Fluorine Chem. 1999, 100, 163.
. X. M. Zhou, W. W. Chen, M. Y. Chao, G. X. Liao,
J. Fluorine Chem. 2013, 153, 101.
and AHF (100 mL/min) were fed into a mixer at
9
. S. Wuttke, S. M. Coman, J. Krohnert, F. C. Jentoft,
E. Kemnitz, Catal. Today 2010, 152, 2.
ꢀ
1
5
50 C, and then passed through a reactor packed with
ꢀ
g of the catalyst at 350–430 C. The products were
1
0. H. D. Quan, Z. Li, Z. X. Zhao, H. E. Yang, J. Lu, Appl.
Catal. B: Environ. 1996, 8, 209.
collected in an ice bath and washed by a dilute solution
of KOH solution, and then the water was removed by
anhydrous sodium sulfate and molecular sieves (4). The
obtained products were analyzed by gas chromatogra-
11. T. Yoshimura,Y. Homoto, Y. Yamada, T. Tsuda,
T. Shibanuma, Process for producing pentafluoroethane
and tetrafluorochloroethane U.S. Patent 6,011,185, 2000.
12. B. Cheminal, E. Lacroix, A. Lants, Process for fluorination
of perchloroethylene or of pentachloroethane U.S. Patent
1
9
phy, GC-mass spectrometry, and F NMR.
5
,932,776, 1999.
CONCLUSIONS
1
3. P. Gibot, L. Vidal, J. Eur. Ceram. Soc. 2010, 30, 911.
From the above experimental results, it is con-
cluded that the dried nitrite–citrate gel can burn in a
self-propagating combustion manner. Superfine metal
oxide powders were directly synthesized via this sol–gel
auto-combustion method. Then, the high-surface-area
fluorination catalysts were obtained after the activation
process, and particle size of the catalysts prepared by
this method reached the nanoscale. The content of the
catalytically active species and was high. The prepared
catalysts showed commendable reaction efficiency for
14. J. L. Delattre, P. J. Chupas, C. P. Grey and A. M. Stacy,
J. Am. Chem. Soc. 2001, 123, 5364.
15. S. Rudiger, U. Groß, E. Kemnitz, J. Fluorine Chem.
2
007, 128, 353.
1
6. F. Zhang, D. Y. Zhao, ACS Nano 2009, 3, 159.
1
7. M. H. Cao, C. W. Hu, E. B. Wang, J. Am. Chem. Soc.
2
003, 125, 11196.
1
8. U. Kurtan, R. Topkaya, A. Baykala and M. S. Toprak,
Ceram. Int. 2013, 39, 6551.
19. K. H. Wu, T. H. Ting, M. C. Li and W. D. Ho, J. Magn.
Magn. Mater. 2006, 298, 25.
1
0
www.jccs.wiley-vch.de
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Chin. Chem. Soc. 2017

JOURNAL OF THE CHINESE
CHEMICAL SOCIETY
Catalytst Produced by Sol–Gel Auto-Combustion
2
0. D. Kovacheva, H. Gadjov, K. Petrov, S. Mandal, M. G.
Lazarraga, L. Pascual, J. M. Amarilla, R. M. Rojas, P.
Herrero, and J. M. Rojo, J. Mater. Chem. 2002, 12, 1184.
1. A. Saberi, F. Golestani-Fard, H. Sarpoolaky, M. Willert-
Porada, T. Gerdes, R. Simon, J. Alloys Compd. 2008,
35. B. Adamczyk, O. Boese, N. Weiher, S. L. M. Schroeder,
E. Kemnits, J. Fluorine Chem. 2000, 101, 239.
36. G. Ertl, H. Knözinger, J. Weitkamp Eds., Handbook
of Heterogeneous Catalysis, Wiley-VCH, Weinheim,
1997.
2
4
62, 142.
37. A. Loustaunau, R. Fayolle-Romelaer, S. Celerier, Catal.
Lett. 2010, 138, 215.
38. Y. X. Cheng, J. L. Fan, Z. Y. Xie, J. Q. Lu and M. F.
Luo, J. Fluorine Chem. 2013, 156, 66.
39. C. Cochon, S. Celerier, A. Riviere, K. Vigier, J. D. Com-
parot, F. Metz and S. Brunet, Catal. Commun. 2010,
12, 151.
40. H. Lee, H. D. Jeong, Y. S. Chung, H. S. Lee,
M. J. Chung, J. Catal. 1997, 169, 307.
2
2
2
2. F. T. Li, Y. Liu, Z. M. Sun, R. H. Liu, C. G. Kou,
Y. Zhao, D. S. Zhao, Mater. Lett. 2011, 65, 406.
3. R. Ianos, I. Lazau, C. Pacurariu, P. Barvinschi, Mater.
Res. Bull. 2008, 43, 3408.
4. S. Habibzadeh, A. Kazemi-Beydokhti, A. A. Khodadadi,
Y. Mortazavi, S. Omanovic, M. Shariat-Niassar, Chem.
Eng. J. 2010, 156, 471.
2
2
2
2
2
5. K. H. Wu, C. H. Yu, Y. C. Chang and D. N. Horng,
J. Solid State Chem. 2004, 177, 4119.
6. Z. X. Yue, J. Zhou, L. T. Li, H. G. Zhang and Z. L.
Gui, J. Magn. Magn. Mater. 2000, 208, 55.
7. L. Shi, K. Tao, R. Q. Yang, F. Z. Meng, C. Xing and N.
Tsubaki, Appl. Catal. A: Gen. 2011, 401, 46.
8. R. Q. Yang, Y. L. Fu, Y. Zhang, N. Tsubaki, J. Catal.
41. E. F. Andrew, TaF.sub.5 and NbF.sub.5 as fluorination
catalysts U.S. Patent 4,258,225, 1981.
42. B. G. Lee, H. S. Kim, H. Kim, M. J. Chung, Method for
concurrently producing 1,1,1,2-tetrafluoroethane and pen-
tafluoroethane U.S. Patent 5,723,700, 1998.
43. C. D. Wagner, In Practical Surface Science, Vol. 1,
D. Briggs, M. P. Seah Eds., Wiley, New York, 1993.
44. M. Guisnet, J. Barrault, C. Bouchoule, D. Duprez and
C. Montassler Eds., Heterogeneous Catalysis and Fine
Chemicals, Elsevier, Amsterdam, 1988, p. 257.
45. P. T. Patil, A. Dimitrov, H. Kirmse, W. Neumann
and E. Kemnitz, Appl. Catal. B: Environ. 2008,
78, 80.
2
004, 228, 23.
9. R. Srivastava, D. Srinivas, P. Ratnasamy, J. Catal. 2005,
33, 1.
2
30. A. Yee, S. J. Morrison, H. Idriss, J. Catal. 1999, 186, 279.
31. A. Yee, S. J. Morrison, H. Idriss, J. Catal. 2000, 191, 30.
32. R. B. Fahim, R. M. Gabr, M. J. Zaki and S. A. A. Man-
sour, J. Colloid Interface Sci. 1981, 81, 468.
3
3. P. Ratnasamyand, A. J. Leonard, J. Phys. Chem. 1972,
46. J. He, G. Q. Xie, J. Q. Lu, L. Qian, X. L. Zhang, P.
Fang, Z. Y. Pu and M. F. Luo, J. Catal. 2008, 253, 1.
47. R. A. Young, The Rietveld Method, Oxford University
Press, Oxford, 1993.
7
6, 1838.
4. B. Adamczyk, A. Hess, E. Kemnitz, J. Mater. Chem.
996, 6, 1731.
3
1
J. Chin. Chem. Soc. 2017
© 2017 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jccs.wiley-vch.de
11