Catalytic pyrolysis of CHF3 over activated carbon and activated carbon supported potassium catalyst

Article abstract of DOI:10.1016/j.jfluchem.2010.03.002

The catalytic activity of activated carbon (AC) and activated carbon supported potassium for the decomposition of CHF₃ was investigated at temperatures between 873 and 1173 K and at a space velocity of 4300 h ^-1. It is found that activated carbon supported potassium shows high and relatively stable activity during the pyrolysis of CHF₃ under the conditions studied. Compared with the gas phase reaction, the conversion of CHF₃ increases by up to 10 times between 873 and 1123 K, with the major products being C₂F₄ and C₃F₆. Selectivities as high as 55% to C₂F₄ and 35% to C ₃F₆ are achieved under optimum conditions. The main byproduct HF readily reacts with K₂O in the catalyst, converting the catalyst from K2O/AC into KF/AC. Selectivity to the major products remains relatively constant following this transformation.

Full text of DOI:10.1016/j.jfluchem.2010.03.002

Journal of Fluorine Chemistry 131 (2010) 698–703
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Catalytic pyrolysis of CHF3 over activated carbon and activated carbon supported
potassium catalyst
b
b
c
Wenfeng Han a, Eric M. Kennedy a, , Huazhang Liu , Ying Li , Adesoji A. Adesina ,
*
John C. Mackie a, Bogdan Z. Dlugogorski a
a Process Safety and Environment Protection Research Group, School of Engineering, The University of Newcastle, Callaghan, NSW 2308, Australia
b Institute of Catalysis, State Key Laboratory Breeding Base of Green Chemistry-Synthesis Technology, Zhejiang University of Technology, Hangzhou 310032, PR China
c Reactor Engineering and Technology Group, School of Chemical Sciences and Engineering, University of New South Wales, Sydney, New South Wales 2052, Australia
A R T I C L E I N F O
A B S T R A C T
Article history:
The catalytic activity of activated carbon (AC) and activated carbon supported potassium for the
decomposition of CHF3 was investigated at temperatures between 873 and 1173 K and at a space
velocity of 4300 hꢀ1. It is found that activated carbon supported potassium shows high and relatively
stable activity during the pyrolysis of CHF3 under the conditions studied. Compared with the gas phase
reaction, the conversion of CHF3 increases by up to 10 times between 873 and 1123 K, with the major
products being C2F4 and C3F6. Selectivities as high as 55% to C2F4 and 35% to C3F6 are achieved under
optimum conditions. The main byproduct HF readily reacts with K2O in the catalyst, converting the
catalyst from K2O/AC into KF/AC. Selectivity to the major products remains relatively constant following
this transformation.
Received 2 February 2010
Received in revised form 1 March 2010
Accepted 10 March 2010
Available online 18 March 2010
Keywords:
Trifluoromethane
Greenhouse gas
Tetrafluoroethylene
Potassium
ß 2010 Elsevier B.V. All rights reserved.
Activated carbon
Catalytic pyrolysis
1. Introduction
are active and stable catalysts for the destruction of HFC-23 in the
presence of O2 and steam at relatively low temperatures [3,4]. A
CHF3 (HFC-23) has limited use, generally for specific applica-
tions in refrigeration or for dry etching in the semiconductor
industry and has limited applications as a fire inhibition agent.
More importantly, it is formed as a byproduct of HCFC-22 (CHClF2)
production. It has recently been reported that its emission and rate
of accumulation into the atmosphere has increased very rapidly
[1]. CHF3 has a large global warming potential (GWP), roughly
11,700 times that of CO2 and its emission is now being regulated
by the Kyoto Protocol. Hence, an effective treatment method for
CHF3 is required to minimize its emission into the receiving
environment. Until very recently, the treatment of this gas has
received very limited research attention.
Thermal oxidation is an established technology which is
sometimes used for the destruction of HFC-23 and its use to
destroy CHF3 has been ratified by the United Nations Framework
Convention on Climate Change [2]. This process uses steam, O2 and
natural gas as reactants to destroy the CHF3 at the temperatures
around 1473 K. It is also reported that phosphates and ZrO2–SO4
major problem associated with these oxidative processes is that
fluorine is removed in the form of HF from the exhaust gas stream
and is then recycled or disposed of as a fluoride salt. It would be
environmentally more sustainable to use waste HFC-23 as a feed
stocks for production of other valuable chemicals, rather than
simply destroying it.
In previous work, we reported that CHF3 can be converted to
vinyl difluoride (CH255CF2), a highly valuable feedstock, through its
reaction with CH4, although the conversion of CH4 and subse-
quently the yield of the targeted product was low [5]. The gas
phase homogeneous pyrolysis of CHF3 is reported to be a potential
route to the production of tetrafluoroethylene (TFE, C2F4) and
hexafluoropyropylene (HEP, C3F6) [6]. TFE and HEP are valuable
chemicals and are used as monomers for the manufacture of
polytetrafluoroethylene (PTFE), which is usually prepared by the
gas phase pyrolysis of CHClF2 (HCFC-22) [7]. The introduction of a
solid catalyst to enhance the yield of C2F4 and C3F6 during the
decomposition of either CHF3 or CHClF2 is desirable, as this will
allow the process to operate at low temperatures. However,
catalytic materials employed in these processes suffered severe
deactivation in the highly corrosive acid gas (HCl and HF)
environment.
* Corresponding author. Tel.: +61 2 4985 4422; fax: +61 2 4921 6893.
E-mail address: eric.kennedy@newcastle.edu.au (E.M. Kennedy).
0022-1139/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2010.03.002

W. Han et al. / Journal of Fluorine Chemistry 131 (2010) 698–703
699
In this contribution, we report a novel catalyst, K2O/AC
(activated carbon), for the decomposition of CHF3 into C2F4
(TFE) and C3F6 (HEP). Activated carbon is chosen because it is stable
in strongly acid and basic environments. Catalysts supported on
various silica and metal oxides are reportedly unstable because of
the degradation of supports in the presence of corrosive HF. The
catalytic reaction was carried out in a fixed-bed reactor and the
effect of temperature on the reaction, as well as catalyst stability,
were investigated.
2. Experimental
2.1. Catalyst preparation
Activated carbon (1.0–1.4 mm) used as catalyst support,
originated from coconut shell, was provided by Hainan Activated
Carbon Co., China, with BET surface area of 993 m2 gꢀ1 and pore
volume of 0.51 cm3 gꢀ1. Catalysts were prepared by traditional
impregnation methods with KNO3 (Aldrich, 99%) as precursor.
Prior to impregnation, activated carbon was treated in 30% HNO3
solution at 363 K for 5 h in order to eliminate ash content in the
carbon. The nominal K loading was 10 wt%. The decomposition
temperature of KNO3 is approximately 673 K.
Fig. 1. Conversion of CHF3, carbon balance and formation rate of products as a
function of time-on-stream over activated carbon at 1073 K and
4300 hꢀ1
a GHSV of
.
(GHSV) of 4300 hꢀ1 to investigate the effect of support. The
support has little influence on the reaction compared to K2O/AC
(shown in Section 3.2), in particular after several hours on stream.
CF4 is the dominant carbon containing product with C2F6 being a
minor product. C2F4 and C3F6 were detected only in trace amounts.
Fig. 1 shows a typical run of CHF3 pyrolysis at temperature of
1073 K. The rate of formation of CF4 was relatively high in the first
hour and dropped rapidly until it reached a stable value.
Consensus exists in the literature that the initial step of
homogeneous gas phase pyrolysis of CHF3 is the elimination of HF
at temperatures above 1023 K leading to formation of CF2, which
rapidly dimerises to C2F4 which is the major product of the reaction
[10,11].
2.2. Temperature-programmed desorption-mass spectrometry
(TPD-MS) experiment of activated carbon
TPD-MS experiment was carried out on AutoChem 2910
(Micromeritics Co.) attached to a QMS 200 (Omnistar) mass
spectrometer. A 50 mg of sample was heated to 373 K in the
sample tube and held that temperature for 0.5 h in flowing helium
gas to remove absorbed water and other impurities on the surface
of the sample. After cooling down to room temperature, the sample
was heated to 1273 K at a ramp rate of 10 K minꢀ1 in He with a flow
rate of 40 mL minꢀ1. The effluent gas was monitored by a thermal
conductivity detector and mass spectrometer detector.
CHF3 ! CF2 þ HF
(1)
2.3. Catalytic experiments
A tubular high purity (99.99%) alumina reactor (i.d. 7.0 mm)
was employed for all experiments. Flow rates of CHF3 (>98%, Core
Gas) and diluting gas N2 (BOC gases, 99.99%) were controlled by
mass flow controllers (Brooks) to give a total flow rate of
220 mmol hꢀ1, with CHF3 accounting for 10% of the total volume
flow. The catalytic pyrolysis of CHF3 was carried out at a reaction
During decomposition of CHF3 over activated carbon, however,
only trace amounts of C2F4 are detected. This indicates that the
decomposition of CHF3 over activated carbon follows a mechanism
which is different from the homogeneous mechanism. Recently,
Yang et al. [12–14] studied the pyrolysis of CHF3 over activated
carbon and in their experiments, C2F4 was not detected. They
suggest that the reactions leading to formation of CF4 and C2F6 is
likely to be via the steps as illustrated in Scheme 1. However, this
seems to contradict the results of their efforts to capture the
difluorocarbene intermediate with H2 and 2-methyl-2-butene,
where no CH2F2 or gem-difluorotrimethylcyclopropane was
captured. Furthermore, when CHClF2 was introduced instead of
CHF3, virtually no CF4 or C2F6 were found in the products [15].
CHClF2 is believed to form CF2 radicals more readily via
dehydrochlorination compared with the elimination of HF from
CHF3. Apparently, an increase of CF2 does not facilitate the
formation of CF4 and C2F6, and the reaction steps proposed in
Scheme 1 likely play only a relatively minor role in the overall
reaction.
temperature between 873 and 1173 K, space velocity of 4300 hꢀ1
,
and at atmospheric pressure. HF formed during reaction was
trapped by a caustic scrubber (NaOH solution) before the reactor
effluent reached an online micro gas chromatograph. Carbon
containing products were identified by a GC/MS (Shimadzu
QP5000) equipped with an AT-Q column, and quantified with a
micro GC (Varian CP-2003) equipped with molecular sieve 5A and
PoraPLOT Q columns. Relative molar response (RMR) factors of
fluorocarbons for TCD detection were experimentally obtained
from standard gas mixtures, and quantification of halogenated
hydrocarbons was performed with diluted halogenated hydro-
carbons in nitrogen. Quantification of other species was estimated
from published correlations [8–10].
3. Results and discussion
3.1.2. Effect of surface oxygen groups
For the purpose of destruction of chlorofluorocarbons (CFCs),
Burdeniuc and Crabtree [15] discovered that the reaction of
sodium oxalate (Na2C2O4) with CFCs enables the complete
conversion of these environmentally hazardous species at
temperatures as low as 543 K. It is important to note that the
existence of oxygen groups on the surface of activated carbon,
3.1. CHF3 pyrolysis over activated carbon
3.1.1. Decomposition of CHF3 over activated carbon
Pyrolysis of CHF3 was conducted over activated carbon at
temperature range of 873–1173 K and gas hourly space velocity

700
W. Han et al. / Journal of Fluorine Chemistry 131 (2010) 698–703
whose decomposition temperatures were between 873 K and
1173 K in the CO evolution [21–25].
The effect of these surface groups involved in the reaction with
CHF3 is unclear. However, the attack of a hydroxyl radical on the C–
H bonds in CHF3 was proposed in a number of studies both
experimentally and theoretically [10]
CHF3 þ OH ! CF3 þ H2O
(2)
There are numerous OH groups on the surface of carbon as
shows in the TPD experiment (Fig. 2). It is likely that reaction (2)
occurs on the surface of the activated carbon, as confirmed by the
appearance of C2F6, which forms as a result of the dimerization of
CF3. As shown in Fig. 1, the rate of formation of C2F6 is rather low
compared with that of CF4. Hence, we suggest that CF3 and CF2
species adsorbed on the surface of carbon, dismutation and/or
decomposition of CF3 over carbon surface dominate the reaction
mechanism rather than dimerization via the following reactions:
Scheme 1. Proposed mechanism for the formation of C2F6 and CF4.
which have the similar structures as that of oxalate, is well known
[16–18]. It has been found that large amounts of oxygen-
containing species exist on the surface of carbon, and these groups
have often been attributed to influence its chemical and catalytic
properties [16,18–20]. TPD experiments can provide information
to quantify and characterize the nature of these surface oxygen
groups. During thermal desorption, these groups decompose
primarily as CO and CO2, at different temperatures. Fig. 2 shows
the CO and CO2 evolution profiles obtained with the activated
carbon used in this study. It can be seen that the carbon-based
material treated with HNO3 contains large amounts of carboxyl
(–COOH), carboxylic anhydride (–C2O3–) and lactonic (–COO–)
groups evolving as CO2 (decomposition temperatures are 473–
673 K and above 673 K respectively). Also, numerous phenolic
(–C5H5-OH), quinine and carbonyl (–CO–) groups were detected,
2CF3ðsÞ ! CF4ðsÞ þ CF2ðsÞ
(3)
(4)
(5)
(6)
(7)
(8)
(9)
2CF2ðsÞ ! CF3ðsÞ þ CF
ðsÞ
CF ! CðsÞ þ F
ðsÞ
ðsÞ
CF2ðsÞ ! CFðsÞ þ F
ðsÞ
CF3ðsÞ ! CF2ðsÞ þ F
ðsÞ
CF2ðsÞ þ F ! CF3ðsÞ
ðsÞ
CF3ðsÞ þ F ! CF4ðsÞ
ðsÞ
To confirm the effect of the surface oxygen groups on pyrolysis
of CHF3, activated carbon treated in H2 flow at 1073 K for 3 h was
examined. As shown in Fig. 2, following H2 treatment, only trace
amounts of CO2 and CO desorption are observed during He-TPD
experiments which suggests that most of these oxygen groups can
be removed via reaction with H2 at 1073 K.
Fig. 3 presents the catalytic performance of the H2 treated
activated carbon material for the pyrolysis of CHF3 as a function of
time-on-stream at 1073 K. As expected, in the absence of the
surface oxygen groups, the rate of formation of CF4 and C2F6 drops
below 0.10 mmol hꢀ1 during the first hour of time-on-stream,
almost 70 times lower than the rates shown in Fig. 1. Although the
conversion of CHF3 follows the similar trend as the reaction over
HNO3-treated activated carbon (Fig. 1), its initial conversion level
is only 9% and drops to almost zero in less than 1 h. Comparatively,
for the reaction over HNO3-treated activated carbon, the initial
conversion level is close to 100% and takes roughly 3 h to reach
zero. It is noted that the rate of formation of C2F6 increases with
Fig. 2. He-TPD-MS spectrums of activated carbon (heating rate: 10 K minꢀ1). (A)
CO2: m/e = 44 and (B) CO: m/e = 28. (1) Activated carbon treated in H2 at 1073 K for
3 h and (2) activated carbon treated in 30% HNO3 solution at 363 K for 5 h.
Fig. 3. Conversion of CHF3, rate of formation of products as a function of time-on-
stream over activated carbon (treated in H2 at 1073 K for 3 h) at 1073 K and a GHSV
of 4300 hꢀ1
.

W. Han et al. / Journal of Fluorine Chemistry 131 (2010) 698–703
701
Table 1
CHF3 conversion, rate of formation of products and elemental balance for the
catalytic pyrolysis of CHF3 over HNO3-treated activated carbon.
TOS
(min) (%)
Conversion Product profile (mmol hꢀ1
)
Element balance (%)c
CF4
C2F6
C2F4
0
HF
C
H
F
47
80
82.1
4.77 0.085
9.66 67.9
98.6
94.5
79.6
84.6
87.8
99.3
61.7
27.5
15.4
3.64 0.056 0.004 6.31 75.8
3.38 0.061 0.009 3.87 78.3
2.60 0.072 0.017 2.73 85.0
1.48 0.079 0.043 1.43 96.0
100.1
96.0
97.9
96.9
131
178
238
0.64
time-on-stream. We suggest that this trend can be attributed to the
gradual diffusion of oxygen groups from the small pores in
activated carbon since this material has well developed porosity. It
is evident that the oxygen groups on the surface of activated
carbon play a major role during the catalytic pyrolysis of CHF3.
With these results in mind, it becomes clear why the presence of O2
is crucial during the synthesis of CF3I via the reaction of CHF3 with
I2 over activated carbons [26]. Similar to the effect of HNO3
treatment, O2 is also likely to produce large surface oxygen groups,
which then catalyze the formation of CF3 and ultimately CF3I.
Fig. 4. Conversion of CHF3 and selectivity of products as a function of time-on
stream over K2-xO/AC at 1073 K and a GHSV of 4300 hꢀ1
.
Table 2, the conversion of CHF3 and selectivity to HFP in the
catalytic pyrolysis is significantly higher than that in non-catalytic
case, except the selectivity to TFE which by contrast is much lower
than the non-catalytic conversion. However, both the yields of TFE
and HFP in catalytic pyrolysis are significantly higher compared
with the non-catalytic process due to the higher conversion of
CHF3 in the catalytic reaction. In addition to the major products,
trace amounts of CF4, C2F6 and C3F8 were formed during the
reaction over catalyst. During the catalytic pyrolysis process, we
speculate that the effect of activated carbon is to enhance the
localized surface concentration of the key intermediate, CF2
carbene, which arises following the elimination of HF from
absorbed CHF3. The rate of polymerization of CF2 on the surface
is enhanced, and is higher than that occurring during the gas phase
reaction. As shown in Table 2, both the selectivity and yield of HFP
increase dramatically following the introduction of a catalyst.
3.1.3. Mass balance of pyrolysis of CHF3 over activated carbons
Conversion of CHF3 and elemental balances for the pyrolysis of
CHF3 over HNO3-treated activated carbon are illustrated in Table 1.
Generally, satisfactory H and F balance levels are achieved; a
maximum level of 96% F is recovered. Much lower carbon balances
were observed, although they gradually increase with time-on-
stream. For instance, at 47 min time-on-stream, the carbon balance
is roughly 68%. During the reaction of sodium oxalate with CFCs,
Burdeniuc and Crabtree [15] found the stoichiometric formation of
carbon from CFCs and F and Cl were detected in the form of mineral
acids, namely HF and HCl. Catalyzed by the oxygen groups, we
suggest that similar reactions could also take place on the surface
of carbon and consequently are responsible for the carbon loss.
This explanation is consistent with the observations that the
surface area of the used catalyst drops to roughly 50% of the fresh
material, presumably due to the deposition of carbonaceous
materials (coke) on the catalyst [26].
3.2.2. Catalytic pyrolysis of CHF3 as a function of TOS
Fig. 4 shows the conversion of CHF3 and selectivity of products
as a function time-on-stream (TOS) over the catalyst at 1073 K and
space velocity of 4300 hꢀ1. The conversion of CHF3 decreases
during an initial period of catalyst deactivation, gradually reaching
steady state. Over the potassium catalyst, we suggest that the
initial step in CHF3 pyrolysis is the elimination of HF [26] as shown
in reaction (1). This byproduct readily reacts with K2O, which is
present on the catalyst, converting the catalyst from K2O/AC into
KF/AC. Consuming HF by K2O, the reaction of HF with K2O enhances
the rate of decomposition of CHF3 during the initial time-on-
stream. Following the transformation of K2O into KF, the
conversion level of CHF3 gradually reduces to a stable state level.
Nagasaki et al. investigated similar catalysts to synthesize CF3I
from CHF3 and I2 in the presence of O2, and found that using
various anions had little effect on the steady-state reaction
selectivity or overall rate. Furthermore, they discovered that all
the anions present in the catalysts they studied were converted to
fluoride, presumably via reaction with hydrogen fluoride [26]. As
shown in Fig. 4, the reaction selectivity remains relatively constant
once the reaction reaches this pseudo-steady-state. The transfor-
3.2. CHF3 pyrolysis over activated carbon supported potassium
catalyst
3.2.1. Comparison of catalytic and gas phase pyrolysis of CHF3
Similar to the non-catalytic pyrolysis of CHF3, the main reaction
products detected over potassium catalyst are C2F4 (TFE) and C3F6
(HFP), as well as trace quantities of CF4, C2F6 and C3F8. Table 2
compares the conversion of CHF3, selectivity and yield between
catalytic and non-catalytic pyrolysis. Catalytic pyrolysis reactions
are carried out at a reaction temperature of 1073 K, space velocity
of 4300 hꢀ1 and at atmospheric pressure. Non-catalytic pyrolysis
was conducted under the same temperature and pressure
conditions, with a reaction residence time of 0.5 s. As shown in
Table 2
Product distribution for the pyrolysis of CHF3 over activated carbon supported potassium catalyst and non-catalytic reaction at temperature of 1073 K.
Entry
CHF3 conversion (%)
Selectivity (%)a
Yield (%)a
TFE
TFE
HFP
CF4
C2F6
C3F8
HFP
1.21
TFE + HFP
Non-catalyticb
Catalyticc
15.1
61.1
55.9
33.3
8.03
–
–
–
8.45
16.1
9.66
34.7
23.6
0.23
4.30
5.05
14.4
a
Based on the conversion of CHF3.
b
c
At residence time of 0.5 s, and pressure of 1 bar.
At space velocity of 4300 hꢀ1 and pressure of 1 bar.

702
W. Han et al. / Journal of Fluorine Chemistry 131 (2010) 698–703
As shown in Fig. 6, the rate of formation of products increases
steadily with temperature except for TFE. With increasing tempera-
ture, the maximum rate of formation of TFE, 1.8 mmol hꢀ1 is
achieved at about 1073 K before it drops gradually. This phenom-
enon implies that CF2 carbenes formed on the surface of catalyst
undergo disproportionation in a parallel pathway to dimerization
[11,27]. However, unlike pyrolysis over activated carbon, saturated
fluorocarbons, C2F6 and C3F8 are formed rather than CF4. The carbon
balance remains between 98 and 100% at reaction temperatures
below 973 K, although with increasing temperature, the carbon
balance drops dramatically. This change is consistent with the
disproportionation mechanism for CF2. The decreasing carbon
balance at high temperatures implies that large amounts of coke
are deposited on the catalyst during the reaction. We suggest that
reaction (5), (6) and (10) are the primary reactions which are
responsible for the formation of coke
CHF3ðsÞ ! CF2ðsÞ þ HF
CF ! CðsÞ þ F
(10)
(5)
ðsÞ
ðsÞ
Fig. 5. Conversion of CHF3 during the catalytic and non-catalytic pyrolysis as a
function of temperature. For the catalytic reaction, GSHV is 4300 hꢀ1 and residence
time is 0.5 s for the non-catalytic case.
CF2ðsÞ ! CFðsÞ þ F
(6)
ðsÞ
Through these reactions, CF3 can readily form and result in
formation of gaseous C2F6 and C3F8 as well as C3F6.
mation of catalyst precursor does not affect the selectivity of
products except TFE and CF4, where selectivity to TFE increases
gradually before it reaches steady state. Conversely, selectivity of
CF4 drops steadily until it is below detection limits. As discussed,
although the catalyst was calcined at 973 K, we presume some
oxygen groups remain on the surface of carbon, and therefore, a
proportion of CF2 species were consumed by these surface oxygen
species, forming CF4 through disproportion of CF2 carbene before
these groups completely disappear.
CF2ðsÞ þ F ! CF3ðsÞ
(8)
ðsÞ
However, the exact surface mechanism is yet unclear. At low
temperatures, the high dissociation energy of carbon and fluorine
bond inhibits the above reactions in the gas phase (Fig. 5). Clearly,
the catalyst plays a major role for the activation of C–F, although, at
temperature above 1023 K, gas phase reactions also contribute to
the overall product scheme (Fig. 5).
For the gas phase reaction, it has been suggested that HEP might
form through the following reactions [6]:
3.2.3. Catalytic pyrolysis of CHF3 as a function of temperature
The conversion of CHF3 and selectivity towards various
products as a function of temperature are shown in Figs. 5 and
6. Conversion of CHF3 during non-catalytic pyrolysis is also
included in Fig. 5 for comparison. As expected, conversion of CHF3
increases with temperature in both cases. To achieve a similar
conversion level, the non-catalytic reaction requires a temperature
2C2F4 ! C4F8
(11)
(12)
(13)
(14)
(15)
(16)
2C2F4 ! c-C4F8
C4F8 ! C3F6 þ CF2
c-C4F8 ! C3F6 þ CF2
C2F4 þ CF2 ! c-C3F6
c-C3F6 ! C3F6
approximately 100 K higher than that when
a catalyst is
introduced. Furthermore, the gas phase decomposition of CHF3
commences at 973 K, but in the presence of catalyst, the reaction
occurs at temperatures below 873 K. Generally, following the
introduction of catalyst, CHF3 conversion levels increase by 2–10
times at temperatures between 873 and 1123 K.
However, high activation energies were found for all the above
reactions and consequently a low HFP yield is observed at
moderate temperatures. With the introduction of catalyst, the
activation energy for the cleavage of C–F bond is lowered with
respect to the gas phase reaction, and as a result, both HFP and
other saturated fluorocarbons can be formed more readily and
directly. However, further investigation needs to be performed to
clarify the surface mechanism.
3.2.4. Yield of TFE and HFP
Recently, Sung et al. [6] reported 28.7% yield of TFE through
pyrolysis of CHClF2 over a surface of metal fluorides at 923 K, a
GHSV of 15,000 hꢀ1 and atmospheric pressure. The gas phase
reaction of a mixture of CHF3 and TFE, resulted in a selectivity of
28.5% to HFP at 1145 K and under optimized conditions [11]. Under
similar conditions and as shown in Fig. 7, the yield of HFP reached
20.2% at 1113 K for the catalytic pyrolysis of CHF3 alone. At 1073 K,
total yield of TFE and HFP is about 35%. However, at higher
temperature, the yield of TFE does not increase with temperature,
probably due to the polymerization of TFE into higher polymers
than HFP.
Fig. 6. Selectivity of products during the catalytic pyrolysis as a function of
temperature.

W. Han et al. / Journal of Fluorine Chemistry 131 (2010) 698–703
703
Acknowledgements
The authors wish to thank the Australian Research Council for
financial support of this project. W.H. is indebted to the
Department of Education, Science and Training (DEST) of the
Australian Government and the University of Newcastle, Australia
for postgraduate scholarships.
References
[1] D.E. Oram, W.T. Sturges, S.A. Penkett, A. McCulloch, P.J. Fraser, Geophys. Res. Lett.
25 (1998) 35–38.
[2] AM0001, Incineration of HFC 23 Waste Streams, UNFCCC, 2006, http://cdm.unfccc.
int/methodologies/index.html.
[3] W.B. Feaver, J.A. Rossin, Catal. Today 54 (1999) 13–22.
[4] H. Onoda, T. Ohta, J. Tamaki, K. Kojima, Appl. Catal. A 288 (2005) 98–103.
[5] H. Yu, E.M. Kennedy, J.C. Mackie, B.Z. Dlugogorski, Environ. Sci. Technol. 40 (2006)
5778–5785.
Fig. 7. Yield of TFE and HFP during the catalytic pyrolysis as a function of
temperature.
[6] D.J. Sung, D.J. Moon, S. Moon, J. Kim, S.-I. Hong, Appl. Catal. A 292 (2005) 130–137.
[7] M.J. Height, E.M. Kennedy, B.Z. Dlugogorski, J. Chromatogr. A 841 (1999) 187–195.
[8] K.P. Schug, H.G. Wagner, F. Zabel, Berichte der Bunsen-Gesellschaft 83 (1979)
167–175.
[9] S.F. Politanskii, V.U. Shevchuk, Kinet. Katal. 9 (1968) 496–503.
[10] R. Romelaer, V. Kruger, J.M. Baker, W.R. Dolbier, J. Am. Chem. Soc. 123 (2001)
6767–6772.
4. Conclusion
[11] D.J. Moon, M.J. Chung, H. Kim, Y.S. Kwon, B.S. Ahn, Ind. Eng. Chem. Res. 41 (2002)
2895–2902.
[12] G.C. Yang, S. Lei, R.M. Pan, H.D. Quan, J. Fluorine Chem. 130 (2009) 231–235.
[13] G.-C. Yang, X.-Q. Jia, R.-M. Pan, H.-D. Quan, J. Mol. Catal. A: Chem. 309 (2009) 184–
188.
[14] G.-C. Yang, X.-Q. Jia, R.-M. Pan, H.-D. Quan, J. Fluorine Chem. 130 (2009) 985–988.
[15] J. Burdeniuc, R.H. Crabtree, Science 271 (1996) 340–341.
[16] F. Rodriguez-Reinoso, Carbon 36 (1998) 159–175.
[17] C. Moreno-Castilla, F. Carrasco-Marin, F.J. Maldonado-Hodar, J. Rivera-Utrilla,
Carbon 36 (1998) 145–151.
[18] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Orfao, Carbon 37 (1999) 1379–
1389.
[19] J.M. Solar, F.J. Derbyshire, V.H.J. de Beer, L.R. Radovic, J. Catal. 129 (1991) 330–342.
[20] H.-H. Tseng, M.-Y. Wey, Chemosphere 62 (2006) 756–766.
[21] M. Schwartz, P. Marshall, R.J. Berry, C.J. Ehlers, G.A. Petersson, J. Phys. Chem. A 102
(1998) 10074–10081.
[22] T.V. Albu, S. Swaminathan, Theor. Chem. Acc. 117 (2007) 383–395.
[23] N.K. Srinivasan, M.-C. Su, J.V. Michael, S.J. Klippenstein, L.B. Harding, J. Phys. Chem.
A 111 (2007) 6822–6831.
[24] L.J. Medhurst, J. Fleming, H.H. Nelson, Chem. Phys. Lett. 266 (1997) 607–611.
[25] J. Korchowiec, J. Phys. Org. Chem. 15 (2002) 524–528.
[26] N. Nagasaki, Y. Morikuni, K. Kawada, S. Arai, Catal. Today 88 (2004) 121–126.
[27] H. Yu, E.M. Kennedy, W.H. Ong, J.C. Mackie, W. Han, B.Z. Dlugogorski, Ind. Eng.
Chem. Res. 47 (2008) 2579–2584.
Activated carbon shows significant effect on the pyrolysis
of CHF3, with a high rate of formation of CF4 and relatively
small amounts of C2F6 and C3F8. We suggest that the oxygen
groups on the surface of carbon play a major role for the
decomposition of CHF3, probably CF3 and CF2 as well. This
speculation was confirmed by experimental results that show
that a much lower conversion level of CHF3, rates of formation of
CF4, C2F6 and C3F8 were observed when these oxygen groups
were removed by pre-treating the activated carbon in H2 at
1073 K for 3 h.
It is found that activated carbon supported potassium catalyst
has high catalytic activity for the pyrolysis of CHF3 into TFE and
HFP. Activated carbon as a support material is relatively stable
during reaction, except for its surface oxygen groups which can
react with CHF3 and adsorbed reaction intermediates, such as CF3
and CF2. Compared with non-catalytic pyrolysis cleavage of C–F
appears to be enhanced by the carbon support which results higher
yields of TFE and HFP.