Effects of M-promoter (M = Y, Co, La, Zn) on Cr2O3 catalysts for fluorination of perchloroethylene

Article abstract of DOI:10.1016/j.jfluchem.2013.08.016

The vapor phase fluorination of perchloroethylene (PCE) to synthesize 2,2-dichloro-1,1,1-trifluoroethane (HCFC-123), 1-chloro-1,2,2,2- tetrafluoroethane (HCFC-124) and pentafluoroethane (HFC-125) was carried out on M-Cr₂O₃ catalysts with different promoters (M = Y, Co, La, Zn). The catalysts were characterized by X-ray diffraction (XRD), hydrogen temperature-programmed reduction (H₂-TPR), Raman spectrum, Ammonia temperature-programmed desorption (NH₃-TPD) and X-ray photoelectron spectroscopy (XPS) techniques. It was found that in the pre-fluorination process CrO_x (x ≥ 1.5) in M-Cr₂O₃ catalysts could be transformed into CrO_xF_y species. The highest activity was obtained on La-Cr₂O₃(F) catalyst with 90.6% of PCE conversion and 93.7% to total selectivity (HCFC-123 + HCFC-124 + HFC-125) at 300 C. The decline in surface acid sites density of the catalyst could improve the specific reaction rate, and the formation of surface CrO_xF _y species could enhance the selectivities to HCFC-123, HCFC-124 and HFC-125 for gas phase fluorination of PCE. Copyright - 2013 Published by Elsevier B.V. All rights reserved.

Full text of DOI:10.1016/j.jfluchem.2013.08.016

Journal of Fluorine Chemistry 156 (2013) 66–72
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Effects of M-promoter (M = Y, Co, La, Zn) on Cr O catalysts
2
3
for fluorination of perchloroethylene
Yong-Xiang Cheng, Jing-Lian Fan, Zun-Yun Xie, Ji-Qing Lu, Meng-Fei Luo *
Key Laboratory of the Ministry of Education for Advanced Catalysis Materials, Institute of Physical Chemistry, Zhejiang Normal University,
Jinhua 321004, China
A R T I C L E I N F O
A B S T R A C T
Article history:
The vapor phase fluorination of perchloroethylene (PCE) to synthesize 2,2-dichloro-1,1,1-trifluor-
oethane (HCFC-123), 1-chloro-1,2,2,2-tetrafluoroethane (HCFC-124) and pentafluoroethane (HFC-125)
Received 16 May 2013
Received in revised form 29 August 2013
Accepted 30 August 2013
Available online 8 September 2013
was carried out on M-Cr
characterized by X-ray diffraction (XRD), hydrogen temperature-programmed reduction (H
Raman spectrum, Ammonia temperature-programmed desorption (NH -TPD) and X-ray photoelectron
spectroscopy (XPS) techniques. It was found that in the pre-fluorination process CrO
(x ꢀ 1.5) in M-
species. The highest activity was obtained on La-
(F) catalyst with 90.6% of PCE conversion and 93.7% to total selectivity (HCFC-123 + HCFC-
2
O
3
catalysts with different promoters (M = Y, Co, La, Zn). The catalysts were
2
-TPR),
3
x
Keywords:
Cr
Cr
2
2
O
O
3
3
x y
catalysts could be transformed into CrO F
2 3
M-Cr O
Fluorination
PCE
1
24 + HFC-125) at 300 8C. The decline in surface acid sites density of the catalyst could improve the
specific reaction rate, and the formation of surface CrO species could enhance the selectivities to
HCFC-123, HCFC-124 and HFC-125 for gas phase fluorination of PCE.
Crown Copyright ß 2013 Published by Elsevier B.V. All rights reserved.
x y
F
HCFC-123
HCFC-124
HFC-125
1
. Introduction
continuously [9]. Catalysts for gas phase Cl/F exchange reactions
are mainly chromium compounds such as unsupported chromium
Since 1970s, Molina and Rowland found that the ozone has
oxide [10], CrF
supported on Al
chromium [15,16].
Among HFCs, it is well known that pentafluoroethane (HFC-
125) has ODP of zero and is a promising candidate for the
3
supported on AlF
3
[11], chromium species
[13,14], doped
been depleted constantly, which is closely related to the
chlorofluorocarbon (CFC) emission [1]. Following the United
Nations Environmental Protection Protocol for CFC regulations
adopted in Montreal [2,3], the usage of CFCs have been prohibited,
which are usually replaced by hydrochlorofluorocarbons (HCFCs)
in some applications such as solvents, detergents and refrigerants.
However, the values of ozone-depleting potential (ODP) of HCFCs
is higher than zero. Therefore, hydrofluorocarbons (HFCs) with
ODP values of zero are more promising as the alternatives of
ozone-depleting substances (ODS).
Nowadays, the most common method to synthesize ODS
substitutes is catalytic fluorination of chlorinated hydrocarbons.
For example, fluorination of trichlorethylene to synthesize 1,1,1,2-
tetrafluoroethane (HFC-134a) [4], fluorination of perchloroethy-
lene (PCE) to synthesize 1,1,1,2,2-pentafluoroethane (HFC-125)
2
O
3
[12], CrF supported on MgF
3
2
2
replacement of CHF Cl (HCFC-22) in refrigerate systems, blend
component of numerous refrigerants, blowing agent, solvent and
propellant. Therefore, many routes for the production of HFC-125
have been developed [5,6,17,18]. One of the popular routes is
catalytic fluorination of PCE. The reaction includes three steps:
CCl
2
¼ CCl
CHCl
þ HF ! CF
CHClF þ HF ! CF
2
þ 3HF ! CF
CHClF þ HCl
CHF
þ HCl
3
CHCl
2
þ 2HCl
(1)
(2)
(3)
CF
CF
3
3
2
3
3
2
[
2
5,6], fluorination of 1,1,2,3-tetrachloropropene to synthesize
,3,3,3-tetrafluoropropene (HFO-1234yf) [7,8]. These Cl/F ex-
Note that the first step is usually carried out in liquid phase in
industrial production using SbF as the catalyst [19]. Due to the
5
change reactions in the gas phase appear to be more advantageous
than the traditional liquid phase process since the gas
phase reactions produce less pollution and could be operated
advantages of gas phase method, gas phase reaction is chosen for
this step. The reported catalysts for the reaction include mainly
chromium compounds such as CrO
x
[20], Cr
2
O
3
3
/AlF [6,21]. But the
effect of promoters on the performance of Cr
rarely reported.
2 3
O catalysts has been
*
Corresponding author. Tel.: +86 0579 82283910; fax: +86 0579 82282595.
An understanding of the active sites on the chromium catalysts
and the reaction mechanism is helpful in the design of highly
E-mail address: mengfeiluo@zjnu.cn (M.-F. Luo).
0022-1139/$ – see front matter. Crown Copyright ß 2013 Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfluchem.2013.08.016

Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
67
(
a)
Cr O3
(b)
Cr O3
2
2
2
-1
2
-1
BET surface area(m g )
BET surface area(m g )
1
17
2
2
2
Zn-Cr O3
110
39
Zn-Cr O (F)
2
3
4
6
La-Cr O3
La-Cr O (F)
2
3
6
5
Co-Cr O3
5
9
Co-Cr O (F)
2
3
5
4
51
56
Y-Cr O3
Y-Cr O (F)
2 3
2
6
3
Cr O3
Cr O (F)
2 3
2
10
20 30 40 50 60 70 80 90
10 20 30 40 50 60 70 80 90
2 ( )
2
( )
2 3 2 3
Fig. 1. XRD patterns of (a) M-Cr O and (b) M-Cr O (F) catalysts.
efficient catalyst systems for Cl/F exchange reaction. Concerning
the active sites, many studies show that the CrO species are the
active site of the fluorination reaction. For example, Lee et al. [13]
prepared the Cr/MgF and Cr catalysts and they found the
catalytic activity of the Cr/MgF catalyst was higher than that of the
Cr catalyst. They concluded that the active sites are CrO
instead of CrF . Adamczyk et al. [22] found that all chromium
oxyfluorides catalysts exhibited significantly higher catalytic
activity than the CrF and Cr catalyst for the fluorination of
CF CH Cl (HCFC-133a). It has also been reported that CrO
catalyst shows catalytic activity 20 times higher than CrF
catalyst for the fluorination of HCFC-133a [23]. In addition, Quan
et al. [24] found that Cr species with higher oxidation state shows
higher catalytic activity than Cr(III). Besides, it was found that the
catalyst activity also strongly depended on the surface acidity of
the catalysts [25,26].
2
and ZnF [31]. However, note that the content of the promoter M is
x
F
y
only 1 mol%, so the M fluorides in the M-Cr
not be observed.
2
O
3
(F) catalysts could
2
2
O
3
The surface areas of the Cr
Zn-Cr are 63, 54, 65, 46 and 117 m g , respectively. The
surface areas of the Cr (F), Y-Cr (F), Co-Cr (F) and La-
Cr (F) are 56, 51, 59 and 39 m g , respectively, indicating that
the surface areas of the catalysts before and after pre-fluorination
2
O
3
, Y-Cr
2
O
3
, Co-Cr
2
O
3
, La-Cr and
2 3
O
2
ꢂ1
2
2 3
O
2
O
3
x
F
y
2
O
3
2
ꢂ1
O
3
2 3
O
2
3
2 3
O
3
2
O
3
hardly change. However, the surface area of the Zn-Cr
2
O
O
3
3
(F)
(F)
2
ꢂ1
3
2
x
F
y
catalysts is 110 m g . The higher surface area of the Zn-Cr
catalyst compared to that of the Cr
2
3
ꢁ4H
2
O
2
O
3
(F) suggests that the addition
of Zn helps to increase the surface area of Zn-doped Cr
catalyst. The increased surface area of the Zn-doped Cr
2
2
O
O
3
3
(F)
(F)
catalyst is probably due to the decline in Cr
this sample, which could be confirmed by the XRD results (Fig. 1b).
It is found that the diffraction peaks of Cr phase in the Zn-
Cr (F) are much broader than those in the Cr (F) catalyst,
indicating smaller crystallite size of Cr in the former sample.
Further analysis using Scherrer equation and the most intensive
diffraction peak in Fig. 1b (36.28), the crystallite size of Cr in the
Zn-Cr (F) is 18 nm, while that in the Cr (F) is 24 nm.
2 3
O crystallite size in
2 3
O
In this paper, a series of Cr
and Zn were prepared, and the effect of the doping elements on the
performance of the Cr catalyst for fluorination of perchloro-
2
O
3
catalysts modified with Y, Co, La
2
O
3
2 3
O
2 3
O
2 3
O
ethylene was studied. Meanwhile, the effects of the surface acidity
and Cr species in the catalysts on the catalytic performance of these
catalysts were discussed.
2 3
O
2
O
3
2 3
O
2 3
2.2. Reduction properties of M-Cr O catalysts
2
. Results and discussion
Fig. 3 shows the H -TPR profiles of the M-Cr O catalysts. For
2
2 3
2
.1. Phase analysis of M-Cr
2
O
3
and M-Cr
2
O
3
(F) catalysts
the Cr
2
O
3
catalyst, only one reduction peak at 290 8C was observed.
Fig. 1 shows the XRD patterns of the M-Cr
(F) catalysts. As shown in Fig. 1a, diffraction peaks due to
were clearly observed in all the M-Cr
2 3
O and the M-
Cr O
F/Cr=0.1
2
3
Cr
2
O
3
Cr
O
crystalline Cr
2
O
3
2
O
3
catalysts. However, no diffraction peak related to M species could
be detected by XRD measurement because of the very low content
of promoters in the catalyst. After pre-fluorination, no obvious
changes were found in the M-Cr
to those of the M-Cr samples. Fig. 2 shows the EDX spectra of
the Cr (F) catalyst. It was found that the molar fraction of F/Cr in
the catalyst was 0.1, which is much lower than the nominal value
in CrF (3/1). This indicates that crystalline Cr is difficult to be
2 3
O (F) catalyst (Fig. 1b) compared
2 3
O
2 3
O
3
2 3
O
fluorinated in HF atmosphere, which is consistent with the results
reported in literature [27]. It was reported [22] that the free
reaction enthalpy of the reaction between Cr
CrF is a small negative value (
G = ꢂ68 kJ mol ). Therefore, the
maximum degree of fluorination of Cr is very small even after
hundreds of hours of contact time. In contrast, Y , Co , La
and ZnO could be easily fluorinated to YF [28], CoF [29], LaF [30]
2 3
O with HF to form
ꢂ1
3
D
R
2 3
O
0
.5
0.7
0.9
2
O
3
3
O
4
2 3
O
2 3
Fig. 2. EDX spectra of the Cr O (F) catalyst.
3
2
3

6
8
Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
2 2 3
Fig. 3. H -TPR profiles of M-Cr O catalysts.
3 2 3
Fig. 4. NH -TPD profiles and the amounts of acid sites of the M-Cr O (F) catalysts.
Table 1
Hydrogen consumption and average valence of Cr in M-Cr
2
O
3
catalysts.
Valence of Cr
of the surface acid sites of the Cr
Cr (F)and Zn-Cr (F) were calculated to be 1.55, 1.50, 1.65, 1.44
and 2.29 mol m , respectively.
2 3 2 3 2 3
O (F), Y-Cr O (F), Co-Cr O (F), La-
Catꢂ1
2
3
2 3
O
ꢂ2
Catalyst
Cr
H
2
consumption (mmol g
)
O
m
2
O
3
0.21
0.34
0.29
0.17
0.13
3.07
3.11
3.09
3.05
3.04
Y-Cr
2
O
3
Co-Cr
2
O
3
2 3
2.4. Raman spectra of M-Cr O (F) catalysts
La-Cr
2 3
O
Zn-Cr
2 3
O
Fig. 5 shows the Raman spectra of the M-Cr
2
O
3
and M-Cr (F)
2
O
3
catalysts. From Fig. 5a, four Raman bands at 302 (
a
1
), 350 (a ), 550
2
ꢂ1
(
a
3
) and 610 cm
4
(a ) were observed, all these Raman bands
Considering that crystalline phase Cr
below 800 8C in H atmosphere [32], the reduction peak of Cr
catalyst is probably attributed to the easily reducible Cr(VI)
species. So, for the M-Cr catalysts, the reduction peaks at 290 8C
are also attributed to the easily reducible Cr(VI) species. For the Co-
Cr catalyst, two weaker reduction peaks at 350 8C and 400 8C
were observed, which could be attributed to the reduction of Co
33].
Table 1 lists the hydrogen consumption of M-Cr
and average valence of Cr. The hydrogen consumptions of the M-
Cr catalysts are calibrated by the known amount of CuO
powder. Regarding the reduction product of Cr species expressed
as Cr , the average valence of Cr is calculated according to the
amount of hydrogen consumption of M-Cr catalysts. From
2
O
3
could not be reduced
could be attributed to Cr(III) species (crystalline Cr
2
O
3
) [34].
ꢂ1
2
2
O
3
Besides, a weak Raman band at 990 cm
(
g
) assigned to Cr(VI) was
and La-Cr catalysts, a
) corresponding to polymeric
also observed [35,36]. For the Y-Cr
2
O
3
2 3
O
ꢂ1
2
O
3
new Raman band at 850 cm
(b
chromate was observed [37]. Furthermore, for the Co-Cr
2
O
3
ꢂ1
2
O
3
catalyst, a new Raman band at 691 cm
corresponding to
3
O
4
Co
Co
3
O
O
4
4
was detected [38], indicating that Co species exist as
phase in the Co-Cr catalyst. For the M-Cr (F) catalysts
), 550 ( ) and
[
3
2
O
3
2 3
O
2
O
3
catalysts
(Fig. 5b), only four Raman bands at 302 (
610 cm
a
1
), 350 (
a
2
a
3
ꢂ1
4 2 3
(a ) were observed, indicating that the crystalline Cr O
2
O
3
is difficult to be fluorinated in HF atmosphere, which is consistent
with the XRD results. Besides, for the M-Cr (F) catalysts, the
and bands corresponding to polymeric chromate and Cr(VI)
2
O
3
b
2
O
3
g
2
O
3
species disappeared, indicating that these high valent Cr species
could be easily fluorinated, which is consistent with the results
Table 1, it can be seen that promoter of M has little effect on
average valence of Cr and the amount of high valent Cr species such
as Cr(VI) in the catalyst is very limited.
reported in literatures [26]. Compared with the Co-Cr
2
O
3
catalyst,
(F)
transforms into Co fluoride.
ꢂ1
the Raman band at 691 cm
2 3
disappeared for the Co-Cr O
catalyst, which indicates that Co
3
O
4
2 3
2.3. Surface acidity of M-Cr O (F) catalysts
2 3
2.5. XPS analysis of M-Cr O (F) catalysts
Fig. 4 shows the NH
of the M-Cr (F) catalysts. It can be seen that only one broad NH
desorption peak was observed in the M-Cr (F) catalysts,
indicating wide distribution of the surface acid sites on these
catalysts. Also, it is clearly shown that the dopants have
remarkable effects on the amount of surface acid sites. For
3
-TPD profiles and the amounts of acid sites
2
O
3
3
2 3
Fig. 6 shows the XPS spectra of the Cr 2p level for M-Cr O (F)
2
O
3
catalysts. The broad peak of Cr 2p3/2 could be deconvoluted to two
peaks with binding energies of 577.5 and 575.5 eV, which could be
assigned to CrO
results indicate that surface Cr species in the M-Cr
were a mixture of Cr and CrO
Table 2 lists the binding energy of Cr 2p3/2 and surface contents
of CrO in the M-Cr (F) catalysts. The surface CrO content
was calculated by the peak area of CrO divided by the total peak
area of Cr species. As can be seen in Table 2, with the addition of M,
a shift toward higher binding energy of the CrO peak was
observed, which could be attributed to the interaction of promoter
M with Cr . Also, the surface CrO contents in the M-Cr (F)
x
F
y
and Cr
2
O
3
, respectively, [31,39,40]. These
2 3
O
(F) catalysts
example, the La-doped catalyst (La-Cr
2
O
3
(F)) had the lowest
2
O
3
x y
F .
ꢂ1
surface acid sites (44.6
m
mol g ) while the Y- and Co-doped
samples had similar surface acid sites as the pure Cr
2
O
3
(F).
x
F
y
2
O
3
x y
F
However, for the Zn-Cr (F) catalysts, it had much more surface
2
O
3
x y
F
acid sites compared to the pure Cr
2
O
3
(F), with surface acid sites of
ꢂ1
2
52.5
m
mol g . This suggests that the addition of Zn is conducive
x y
F
to increase the acidity of the catalyst. Based on the amount of
surface acid sites and the surface area of the catalysts, the densities
2
O
3
x
F
y
2 3
O

Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
69
2 3 2 3
Fig. 5. Raman spectra of M-Cr O and M-Cr O (F) catalysts.
catalysts were varied depending on the dopants. For example, the
improved the PCE conversion, and the doping of Y, Co, La or Zn in
the Cr catalysts was beneficial to the total selectivity. The
highest conversion and selectivity were obtained on the La-
Cr (F) catalyst, with a PCE conversion of 90.6% and a total
Zn-Cr
species (48.6%), and the La-Cr
21.6%). This indicates that the promoter has remarkable effect on
2
O
3
(F) catalyst had the highest content of surface CrO
x
F
y
2 3
O
2 3
O
(F) catalyst had the lowest value
(
2 3
O
the formation of surface CrO
x
F
y
species in the catalyst.
selectivity of 93.7%. As for the specific rate of reaction based on the
surface area of the catalyst, Table 3 also shows that the promoter M
had remarkable effects on the specific rates of reactions. The La-
2
.6. Catalytic activity of M-Cr
2
O
3
(F) catalysts
Cr
(0.778
lowest value (0.267
Since it has been reported that the Cl/F exchange reaction
strongly depends on the surface acidity of the catalyst [24,41] and
the surface areas of the catalysts employed in the current work are
different, the effect of surface acidity on the catalytic activity of the
catalyst was analyzed. Fig. 8 shows the relationship between the
specific reaction rate and the density of surface acid sites in the M-
2 3
O (F) catalyst had the highest specific reaction rate
ꢂ1
ꢂ2
Fig. 7 shows the catalytic behavior of the La-Cr
2
O
3
(F) catalyst
m
mol min
m
) while the Zn-Cr
2 3
O (F) catalyst had the
ꢂ1
ꢂ2
).
for the fluorination of PCE at 300 8C. It is known that the stability of
catalysts is critical in practical applications. As shown in Fig. 7, the
conversion of PCE and selectivity to HCFC-123, HCFC-124 and HFC-
25 declined slightly in the first 3 h, the conversion of PCE slightly
declined from 94.9% to 90% and the selectivity dropped from 96.4%
to 93.6%. After that, the conversion of PCE and selectivity remained
m
mol min
m
1
2 3
stable, indicating good stability of the La-Cr O (F) catalyst.
However, at the early stage of the reaction, the decline in the
activity was probably due to the coverage of organic compounds on
the catalyst surface which could block the active sites.
Table 3 lists conversion of PCE, specific reaction rate and
selectivities to HCFC-123, HCFC-124 and HFC-125 over the M-
2 3
Cr O (F) catalysts. It can be seen that the specific rate of reaction
decreases with the increase in surface acid sites density. This
indicates that too large amount of surface acid sites may be not
beneficial for the reaction. This is because olefin molecules could
be easily adsorbed on the surface acid site of the catalyst [42,43],
and high surface acidity of the catalyst would result in a rapid coke
formation and consequently block of active sites [44], which leads
to the decrease in the specific rate of reaction.
2 3
Cr O (F) catalysts. It was found that the promoter in the catalyst
Fig. 9 shows the relationship between the PCE conversion and
contents of surface Cr species in the catalysts. It can be seen that
the surface CrO
of reaction (Fig. 9(a)). The PCE conversion changed little with
increasing surface CrO content. The La-Cr (F) catalyst with
the lowest content of CrO had the highest conversion (90.6%),
while the Cr (F) catalyst with medium content of CrO had the
lowest conversion (80.2%). This indicates that the presence of
surface CrO species in the catalyst merely promoted the
catalytic activity. The finding is in contrast with those reported
x y
F content exerts little influence on the conversion
x
F
y
2 3
O
x y
F
2
O
3
x y
F
x y
F
x y
in literature, in which it was proposed that CrO F species are
Table 2
Binding energy of Cr 2p2/3 and CrO
Catalyst Cr2p3/2/eV
CrO
x
F
y
surface content of M-Cr
2
O
3
(F) catalysts.
x y
Surface CrO F content (%)
x
F
y
2 3
Cr O
Cr
Y-Cr
Co-Cr
2
O
3
(F)
(F)
(F)
(F)
(F)
577.5
577.7
577.7
578.4
577.7
575.5
575.4
575.5
575.5
575.5
39.2
37.4
44.4
21.6
48.6
2
O
3
2
O
3
La-Cr
Zn-Cr
2
O
3
2
O
3
2 3
Fig. 6. XPS spectra of the Cr 2p level for M-Cr O (F) catalysts.

7
0
Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
Table 3
Conversion of PCE, specific reaction rate and selectivity to HCFC-123, HCFC-124, HFC-125 over M-Cr
2 3
O (F) catalysts (obtained at 10 h reaction).
ꢂ1 ꢂ2
Catalyst
Conversion (%)
Specific rate (
mmol min
m
)
Selectivity (%)
HCFC-123
a
HCFC-124
HFC-125
Total (%)
Cr
Y-Cr
Co-Cr
2
O
3
(F)
(F)
(F)
(F)
(F)
80.2
84.0
82.5
90.6
87.9
0.507
0.551
0.468
0.778
0.267
42.4
44.1
43.4
56.0
46.5
36.9
31.2
31.1
27.6
31.6
4.5
18.0
18.9
10.1
7.9
83.8
93.3
93.4
93.7
86.0
2
O
3
2
O
3
La-Cr
Zn-Cr
2
O
3
2
O
3
a
HCFC-123 + HCFC-124 + HFC-125.
1
00
0
0
0
0
0
0
0
.8
.7
.6
.5
.4
.3
.2
La-Cr O (F)
2
3
PCE conversion
8
6
4
2
0
0
0
0
0
Sel. to HFC-125
Sel. to HCFC-124
Sel. to HCFC-123
Total selectivity
Y-Cr O (F)
2
3
Cr O (F)
2
3
Co-Cr O (F)
2
3
Zn-Cr O (F)
2
3
1
.4 1.6 1.8 2.0 2.2 2.4 2.6
-2
Surface acid density (µmol m )
0
2
4
6
8
10
12
Fig. 8. Relationship between the specific reaction rate and the density of surface
acid sites in the M-Cr (F) catalysts.
Time on stream (h)
2 3
O
Fig. 7. Catalytic behavior of the La-Cr
00 8C.
2 3
O (F) catalyst for the fluorination of PCE at
3
active sites of fluoridation [13,17,18], but the relationship between
the surface CrO species content and catalytic activity of catalysts
has not been quantitatively analyzed yet. In the current work, since
the CrO species in the catalyst has little effect on the catalytic
performance, it may imply that these species may not be the active
species for fluorination reaction, at least for the fluorination of PCE.
Meanwhile, the PCE conversion is also independent on the content
exerts more pronounced influence on the catalytic behavior (Fig. 9)
than the surface Cr species do.
x y
F
In order to better understand the role of CrO
reaction, a Cr catalyst without pre-fluorination was tested, and
the results are shown in Fig. 10. Interestingly, the Cr catalyst
x y
F species in the
x
F
y
2 3
O
2 3
O
without pre-fluorination also shows quite good activity. It can be
seen that the PCE conversion slightly increased from 72.0% to 80.2%
in the first 2 h reaction, and reached a steady state with a PCE
conversion of 80.2%. Concerning the selectivity, the selectivity to
HCFC-123 gradually declined with the reaction, accompanied by a
slight increase in the selectivity to HFC-125. Meanwhile, the
2 3
of Cr O in the catalyst as shown in Fig. 9(b). These results indicate
that the active sites for this reaction are more complicated than we
expect, and more detailed investigation is needed. Nevertheless,
from our results, it seems that the surface acidity of the catalyst
100
100
(
a)
(b)
8
6
4
2
0
0
0
0
0
80
60
40
20
0
2
0
25 30 35 40 45 50
50 55 60 65 70 75 80
CrO
x
F
y
content (%)
Cr
2
O content (%)
3
x y 2 3
Fig. 9. Relationship between surface contents of (a) CrO F and (b) Cr O in the catalysts and PCE conversion.

Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
71
100
100
a) Cr O without pre-fluorination
b) Cr O with pre-fluorination
2
3
2
3
80
60
40
20
0
80
60
40
20
0
PCE conversion
Sel. to HFC-125
Sel. to HCFC-124
Sel. to HCFC-123
Total selectivity
PCE conversion
Sel. to HFC-125
Sel. to HCFC-124
Sel. to HCFC-123
Total selectivity
0
2
4
6
8
10
0
2
4
6
8
10 12
Time on stream (h)
Time on stream (h)
2 3
Fig. 10. Comparison of catalytic performance of Cr O catalyst with or without pre-fluorination treatment.
selectivity to HCFC-124 remarkably increased from 10% to 40% in
2 3
1/99 in M-Cr O catalyst. A detailed process was as follows:
5
h reaction. Therefore, a total selectivity of 83.6% was obtained at
steady state. Note that the Cr catalyst without pre-fluorination
may be gradually fluorinated by HF following the reaction
CrO + H O, because it contained some
(x ꢀ 1.5) + HF ! CrO
Cr(VI) species as evidenced by the Raman spectroscopy (Fig. 5).
Thus, the main role of CrO species in the catalyst is likely
to promote the further fluorination of HCFC-123 to HCFC-124
and HFC-125 as shown in the following reactions: CF CHCl
FCFC-123) + HF ! CF CHClF (HCFC-124) + HCl, CF CHClF (HCFC-
24) + HF ! CF CHF (HFC-125) + HCl. On the other hand, as the
PCE conversion slightly increased from 72% to 80% on this catalyst,
the formation of such CrO species may also enhance the activity.
In contrast, the pre-fluorinated Cr (F) catalyst shows more
Ammonia (10% aqueous solution, Lanxi Yongli Chemical Co., LTD.,
ꢂ1
2
O
3
China) was added to a chromic nitrate (Cr(NO
3
)
3
, 1 mol L
,
analytical grade, Sinopharm Chemical Reagent Co., Ltd., China) or
the mixed aqueous solution of Cr(NO and M(NO (M = Y, Co, La,
x
x
F
y
2
3
)
3
3 x
)
Zn, analytical grade, Sinopharm Chemical Reagent Co., Ltd., China)
under stirring until a precipitated slurry was obtained. The pH
value of the suspension was controlled at around 8.0. The resulting
slurry was aged for 2 h and then separated by centrifugation from
the mother liquid, washed several times with deionized water and
dried at 120 8C overnight. Finally, it was calcined at 500 8C for 4 h in
x y
F
3
2
(
1
3
3
3
2
N
2
to obtain the catalysts and the catalysts were denoted M-Cr
2 3
O
x
F
y
(include all catalysts including Cr ).
2 3
O
2
O
3
Before reaction, pre-fluorination was carried out in order to
activate the catalyst. The pre-fluorination of the oxide sample (M-
stable catalytic behavior, as the PCE conversion and selectivities to
HCFC-123, HCFC-124 and HFC-125 were constant during the
reaction. It can be seen by comparing these two catalyst that very
close catalytic results (PCE conversion and total selectivity) were
obtained on the two catalysts at steady state, indicating the final
properties of these two catalysts are similar. This again confirms
that the presence of CrO
promotes the formation of HCFC-124 and HFC-125, rather than to
improve the activity.
2 3
Cr O ) was carried out in a stainless steel tubular reactor (10 mm
(i.d.) ꢃ 300 mm), equipped with an electric heater. 3 g of M-Cr
2 3
O
was loaded into the reactor and dried at 300 8C for 2 h in N
2
with a
ꢂ1
flow of 30 ml min . Then, the N
2
flow was stopped and mixture of
ꢂ1
ꢂ1
HF (40 ml min ) and N
2
(10 ml min ) was introduced at 260 8C
catalysts
(F) (include
x
F
y
species in the catalyst are mainly to
for 1 h and subsequently at 400 8C for 3 h. The M-Cr
2 3
O
after the activation process were denoted as M-Cr O
2 3
2 3
all catalysts including Cr O ).
3
. Conclusions
4.2. Fluorination reaction
A series of M-Cr
2
O
3
catalysts with different promoters M (M = Y,
The fluorination reaction was carried out in the same reactor
after the catalyst activation at 300 8C under atmospheric pressure.
The mole ratio of PCE/HF (Hydrogen fluoride) was 1/10. The total
Co, La, Zn) were prepared for vapor-phase fluorination of PCE to
synthesize HCFC-123, HCFC-124 and HFC-125. During the fluorina-
ꢂ1
tion process, well dispersed CrO
crystalline Cr was difficult to be fluorinated. The highest activity
was obtained over the La-Cr (F) catalyst at 300 8C, with a PCE
3
transformed into CrO
F
x y
and the
flow rate was 22 ml min . To remove the HF and HCl, the reaction
O
2 3
effluent passed an aqueous KOH solution, then, products were
analyzed by a gas chromatograph (Shimadzu GC-2014) equipped
with a flame ionization detector (FID) and a GS-GASPRO capillary
column (60 m ꢃ 0.32 mm).
2 3
O
conversion of 90.6% and a total selectivity to HCFC-123, HCFC-124
and HFC-125 of 93.7%. The enhanced specific reaction rates obtained
on the La- and Y-doped catalysts could be attributed to the declines
in the density of surface acid sites. The presence of surface CrO
species helped to promote the selectivity.
F
x y
4.3. Catalyst characterizations
X-ray diffraction (XRD) patterns of the catalysts were recorded
4
. Experimental
using a Philips PW3040 diffractometer with Cu K
a radiation
(l
= 0.1542 nm) operating at 40 kV and 40 mA. The patterns were
4.1. Preparation of catalyst
collected in a 2
u
range from 10 to 908, with a scanning step of
ꢂ1
0.158 s
.
The Cr
2
O
3
and M-Cr
2
O
3
(M = Y, Co, La, Zn) catalysts were
Surface areas of the catalysts were determined by the modified
BET method from the N adsorption isotherms at liquid nitrogen
prepared by a co-precipitation method, the molar ratio of M/Cr was
2

7
2
Y.-X. Cheng et al. / Journal of Fluorine Chemistry 156 (2013) 66–72
[
[
4] H.D. Quan, Z. Li, Z.X. Zhao, H.E. Yang, J. Lu, Appl. Catal. B: Environ. 8 (1996) 209–
15.
5] T. Yoshimura,Y. Homoto, Y. Yamada, T. Tsuda, T. Shibanuma, US Patent 6011185
(2000).
temperature (ꢂ195.7 8C) on an NOVA 40000e Surface Area & Pore
Size Analyzer. Before the measurement, the samples were out-
gassed at 300 8C for 4 h under vacuum.
2
[6] B. Cheminal, E. Lacroix, A. Lants, US Patent 5932776 (1999).
[7] V.N. Mallikarjuna Rao, Allen C. Sievert, WO Patent 2008060614 (2008).
[8] S. Bektesevic, R.C. Johnson, D.C. Merkel, K.A. Pokrovski, H.S. Tung, EP Patent
2103587 (2009).
The reduction properties of M-Cr
by hydrogen temperature programmed reduction (H
was carried out in a fixed-bed (i.d. = 6 mm) reactor containing
2
O
3
catalysts were measured
2
-TPR), which
2
0 mg of catalyst. The sample was heated in a flow of N
2
to 300 8C
[9] C. Alonso, A. Morato, F. Medina, Y. Cesteros, P. Salagre, Appl. Catal. B: Environ. 40
(2003) 259–269.
10] S. Brunet, B. Requieme, E. Colnay, J. Barrault, M. Blanchard, Appl. Catal. B: Environ.
ꢂ1
at a rate of 10 8C min , and kept at 300 8C for 30 min. After cooling
[
down to 100 8C, the sample was heated from 100 to 500 8C with a
5
(1995) 305–317.
ꢂ1
heating rate of 10 8C min
under a mixture of 5%H
consumption was measured by a
2
-95%N
2
[11] S. Brunet, B. Requieme, E. Matouba, M. Blanchard, J. Catal. 152 (1995) 70–74.
[12] D.H. Cho, Y.G. Kim, M.J. Chung, J.S. Chung, Appl. Catal. B: Environ. 18 (1998) 251–
ꢂ1
(20 ml min ). The amount of H
2
2
61.
13] H. Lee, H.D. Jeong, Y.S. Chung, H.S. Lee, M.J. Chung, J. Catal. 169 (1997) 307–316.
[14] H. Kim, H.S. Kim, B.G. Lee, H. Lee, S. Kim, J. Chem. Soc. Chem. Comm. 23 (1995)
383–2384.
15] J. Krishna Murthy, U. Gross, S. R u¨ diger, E. U¨ nveren, W. Unger, Appl. Catal. A: Gen.
82 (2005) 85–91.
[16] D.W. Bonniface, J.D. Scott, M.J. Watson, J.R. Fryer, P. Landon, Green Chem. 1 (1999)
–11.
gas chromatograph with a thermal conductivity detector (TCD),
which was calibrated by the quantitative reduction of a known
amount of CuO powder.
[
2
[
Surface acidity of the catalysts were measured by ammonia
2
temperature-programmed desorption (NH
carried out in fixed-bed reactor (i.d. = 6 mm) containing
00 mg of catalyst. The sample was heated in a flow of N to
3
-TPD), which was
9
a
[
[
17] E. Feiring Andrew, US Patent 4258225 (1981).
18] B.G. Lee, H.S. Kim, H. Kim, M.J. Chung, US Patent 5723700 (1998).
[19] S. Brunet, C. Batiot, M. Calderon, J. Mol. Catal. A: Chem. 108 (1996) 11–14.
20] M. Nose, K. Takahashi, T. Shibanuma, US Patent 20100268002 (2010).
21] D. Carmello, G. Guglielmo, EP Patent 0282005 (1988).
22] B. Adamczyk, O. Boese, N. Weiher, S.L.M. Schroeder, E. Kemnits, J. Fluorine Chem.
101 (2000) 239–246.
1
3
2
ꢂ1
00 8C at a rate of 10 8C min , and kept at 300 8C for 60 min, and
ꢂ1
[
[
[
then was cooled down to 50 8C. Then a flow of NH
3
(20 ml min
)
was introduced to the reactor for 30 min. The gaseous or physical
ꢂ1
adsorption of NH
3
was removed by purging N
2
flow (20 ml min
)
at 50 8C for 90 min. Then the sample was heated from 50 to 600 8C
at a rate of 10 8C min , and the profile was recorded using a gas
[23] Y.S. Chung, H. Lee, H.D. Jeong, Y.K. Kim, H.G. Lee, J. Catal. 175 (1998) 220–225.
[24] H.D. Quan, M. Tamura, Y. Matsukawa, J. Mizukado, T. Abe, J. Mol. Catal. A: Chem.
ꢂ1
219 (2004) 79–85.
chromatograph with a TCD detector.
[25] A. Hess, E. Kemnitz, A. Lippitz, W.E.S. Unger, D.H. Menz, J. Catal. 149 (1994) 270–
Raman spectra were collected by a Renishaw RM1000 confocal
microprobe under ambient conditions. The wavelength of the
280.
[
[
26] S. Brunet, B. Boussand, J. Barrault, Stud. Surf. Sci. Catal. 101 (1996) 379–385.
27] L.Q. Xing, Q.Y. Bi, Y.J. Wang, M. Guo, J.Q. Lu, M.F. Luo, J. Raman Spectrosc. 42 (2011)
excitation laser was 514 nm. The scanning range was 200–
1095–1099.
ꢂ1
1
200 cm
.
[28] L. Qian, L.Q. Xing, Q.Y. Bi, H.F. Li, K.F. Chen, X.L. Zhang, J.Q. Lu, M.F. Luo, Acta. Phys.
Chim. Sin. 25 (2009) 336–340.
X-ray photoelectron spectroscopy (XPS) experiments were
carried out on a VGESCALAB MK2 system with Al-K radiation
hv = 1486.6 eV). Binding energies were calibrated by using the
[
29] S.H. Jeon, Y.S. Kim, C.H. Jung, Plasma Chem. Plasma Process. 28 (2008) 617–
28.
[30] Q.Y. Bi, J.Q. Lu, L.Q. Xing, M. Guo, M.F. Luo, Chin. J. Chem. Phys. 23 (2010) 89–94.
a
6
(
[
31] A. Loustaunau, R. Fayolle-Romelaer, S. Celerier, A. D’Huysser, L. Gengembre, Catal.
Lett. 138 (2010) 215–223.
contaminant carbon (C1s = 284.6 eV).
[32] X. Wang, Y.C. Xie, Appl. Catal. B: Environ. 35 (2001) 85–94.
Acknowledgements
[33] S. Tuti, F. Pepe, Catal. Lett. 122 (2008) 196–203.
[34] H.C. Barshilia, K.S. Rajam, Appl. Surf. Sci. 255 (2008) 2925–2931.
[35] D.S. Kim, J.M. Tatibouet, I.E. Wachs, J. Catal. 136 (1992) 209–212.
[36] E. Groppo, A. Damin, F. Bonino, F. Bonino, A. Zecchina, S. Bordiga, C. Lamberti,
Chem. Mater. 17 (2005) 2019–2027.
This work was supported by the National Natural Science
Foundation of China (No. 20873125) and the Program for
Changjiang Scholars and Innovative Research Team in Chinese
Universities (IRT0980).
[37] B.M. Weckhuysen, I.E. Wachs, J. Phys. Chem. 100 (1996) 14437–14442.
[38] V.G. Hadjiev, M.N. Iliev, I.V. Vergilov, J. Phys. C 21 (1988) Ll99–L201.
[39] M. Kang, C.H. Lee, Appl. Catal. A: Gen. 226 (2004) 163–172.
[40] J.M. Rao, A. Sivaprasad, P.S. Rao, B. Narsaiah, S.N. Reddy, J. Catal. 184 (1999) 105–
111.
References
[
[
[
41] E. Kemnitz, A. Hess, G. Rother, S. Troyanov, J. Catal. 159 (1996) 332–339.
42] J.N. Kondo, K. Domen, J. Mol. Catal. A 199 (2003) 27–38.
43] S.H. Kang, J.W. Bae, P.S. Sai Prasad, S.J. Park, K.J. Woo, Catal. Lett. 130 (2009) 630–
[
[
1] M.J. Molina, F.S. Rowland, Nature 249 (1974) 810–812.
2] Programme for alternative fluorocarbon toxicity testing, Report of PAFT-V: HFC-
6
36.
44] M.T. Xu, J.H. Lunsford, D. Wayne Goodman, A. Bhattacharyya, Appl. Catal. A: Gen.
49 (1997) 289–301.
3
2, September 1992.
[
[
3] United Nations Environmental Programme, Report of the Ad Hoc technical
advisory committee on ODS destruction technologies, May 1992.
1