You are on page 1of 7

Kinetics of tetrachloromethane ¯uorination by hydrogen

¯uoride in the presence of antimony pentachloride
Egbert Meissner, Eugeniusz Milchert
*
Department of Organic Technology, Technical University of Szczecin, PL 70-310 Szczecin, Poland
Received 27 February 2002; received in revised form 23 July 2002; accepted 24 September 2002
Abstract
Contradictory information concerning the kinetics of tetrachloromethane ¯uorination by anhydrous hydrogen ¯uoride in the presence of
antimony pentachloride has been explained. The present studies were performed using a stainless steel autoclave. The exchange of the ®rst
chlorine atomwas carried out in the temperature range of 35±95 8C, with the molar ratio of HF/CCl
4
varied within the range of 1.2±1.7 and at a
constant molar ratio of SbCl
5
/Cl
2
equal to 1.9. The degree of conversion of CCl
4
to CCl
3
F and the reaction rate constant as a function of
temperature ®t to an Arrhenius straight line. The Arrhenius constants (pre-exponential factor) and the activation energy were determined.
Studies concerning the exchange of the second chlorine atom, i.e. the conversion of CCl
3
F to CCl
2
F
2
have been performed. The results of
kinetic investigations were created in the same way. Based on these results it was found that the Arrhenius equation is ful®lled also in the
second stage of the ¯uorination. The same slope of the straight lines con®rms that the ¯uorination stages proceed according to the same
mechanism with the same activation energy but with different pre-exponential factors.
Several kinetic problems associated with the performance of the reactor can be solved on the basis of the equations derived. However,
knowledge of CCl
4
conversion as a function of time, temperature, catalyst concentration and the type of the reaction is required for this
purpose. Such calculations enable the selection of the optimal process parameters.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Kinetics; Fluorination; Chloro¯uorocarbons
1. Introduction
The ¯uorination of tetrachloromethane by anhydrous
hydrogen ¯uoride in the presence of antimony pentachloride
was the principal reaction in the synthesis of chloro¯uor-
ocarbons (CFCs): CCl
3
F (CFC-11) and CCl
2
F
2
(CFC-12)
which were utilized previously as refrigerants.
The ¯uorine±chlorine exchange reaction is of considerable
practical importance in the synthesis of hydrochloro¯uoro-
carbons (HCFC) and completely safe hydro¯uorocarbons
(HFC), e.g. CF
3
CH
2
F (HFC-134a). The latter compound
[1,2] is most often used as a replacement for CFC-11 and -12.
Literature reports concerning the way in which SbCl
5
participates in ¯uorine exchange between HF and CCl
4
provide no intrinsic discrepancy [1±4]. There is convincing
evidence, con®rming that antimony(V) chloro¯uorides with
a general formula of SbCl
5÷n
F
n
can be used as ¯uorinating
agents. They are formed in the following reaction
SbCl
5
÷nHF ÷SbCl
5÷n
F
n
÷nHCl (1)
where n = 1 to 4.
There is also consistency in the opinions that the reaction
of antimony(V) chloro¯uoride with CCl
4
proceeds accord-
ing to the following way
SbCl
4÷n
F
n÷1
÷CCl
4
÷SbCl
5÷n
F
n
÷CCl
3
F (2)
SbCl
4÷n
F
n÷1
÷CCl
3
F ÷SbCl
5÷n
F
n
÷CCl
2
F
2
(3)
SbCl
4÷n
F
n÷1
÷CCl
2
F
2
÷SbCl
5÷n
F
n
÷CClF
3
(4)
SbCl
4÷n
F
n÷1
÷CClF
3
÷SbCl
5÷n
F
n
÷CF
4
(5)
Antimony ¯uorides were the subject of many studies
performed by Kolditz and coworkers [4±7]. Although their
objective has been focused mainly on the analytical pro-
blems, they shed some light on the process of ¯uorine
exchange between HF and SbCl
5
. The results con®rm
that the exchange of ¯uorine with the ®rst and second
chlorine atoms in SbCl
5
proceeds readily even at low
temperatures (from ÷70 to ÷40 8C). Complete exchange
of the third atom occurs with a suf®cient excess of HF
in the temperature range from ÷40 to 20 8C, whereas the
fourth atom is exchanged in the temperature range from
÷25 to 0 8C. However, the exchange of the last chlorine
atom for ¯uorine does not proceed by this route. The use
of even a considerable excess of HF (HF/SbCl
5
~20 mol),
Journal of Fluorine Chemistry 119 (2003) 89±95
*
Corresponding author. Tel.: ÷48-91-449-4185; fax ÷48-91-449-4365.
E-mail address: milch@mailbox.tuniv.szczecin.pl (E. Milchert).
0022-1139/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S0 0 2 2 - 1 1 3 9 ( 0 2 ) 0 0 2 4 1 - 5
with annealing carried out for more than 2 h at temperature
of 85 8C and distillation of HF, does not lead to the
formation SbF
5
but to formation of SbClF
4
HF. Further
distillation of SbClF
4
HF causes its decomposition with
the evolution of chlorine, and simultaneous reduction of
Sb(V) to Sb(III).
2. The kinetic model
To our knowledge no paper has been published on the
kinetics of ¯uorine exchange between SbCl
5
and HF in the
existing literature. Simakov and Korobochko [8] have
reported that the HF absorption is accompanied by the
chemical reaction, the role of which can be given by the
kinetic equation
R = k[SbCl
5
[
2
[HF[ (1)
where [SbCl
5
]
2
and [HF] are molar concentrations, k the
observed rate constant of reaction, estimated at room tem-
perature as k = 1 ×10
3
dm
6
mol
÷2
s
÷1
.
From this value it can be concluded that it describes the
rate of exchange of the ®rst chlorine atom in SbCl
5
. How-
ever, one cannot evaluate whether further stages of
exchange in (1) for n = 2; 3; 4 will proceed with the same
reaction rate. When the less advantageous variant is
assumed and in the reaction (1), only SbCl
4
F (n = 1) is
formed, SbCl
5
is regenerated in the reactions (2)±(5) and is
almost instantaneously ¯uorinated to SbCl
4
F. Thus, as long
as HF is present in the reaction, the principal process of
CCl
4
¯uorination proceeds at a constant concentration of
¯uorinating agent.
The authors of the papers related to the kinetic aspect
[4,9,10] are in agreement that at temperature higher than 20±
26 8C the reaction is (Eq. (6)) irreversible
SbCl
4
F ÷CCl
4
÷CCl
3
F ÷SbCl
5
(6)
Kolditz and Schultz [7] believed that the effect of tem-
perature can be attributed to the deactivation of SbCl
4
F. This
compound forms a stable complexes (SbCl
4
F)
n
SbCl
5
at low
temperatures in which all ¯uorine atoms are involved in
bridge bonds, thereby they cannot participate in the
exchange reaction with CCl
4
.
The papers discussing the kinetic equation of the reaction
lead to contradictory conclusions. The results of studies
performed by Chekmarev et al. [9], indicate that the rate of
Eq. (6) reaction at the temperature range between 20 and
50 8C can be described by the kinetic equation
r = k[SbCl
4
F[
2
[CCl
4
[ (2)
Kolditz and Schultz [7] claimed that at a temperature range
between 0 and 26 8C the following equation is more appro-
priate
r = k[SbCl
4
F[[CCl
4
[ (3)
Different values of the activation energy obtained by both
authors suggest that a change of the reaction mechanismtake
places with change in temperature.
The reaction between antimony(V) chloro¯uorides and
CCl
4
was also presented as (7).
SbCl
2
F
3
÷3CCl
4
÷3CCl
3
F ÷SbCl
5
(7)
However, the authors have restricted their discussion only
to a comparison of the kinetic curves of the reactions (6) and
(7). On the basis of the form of Eq. (7) they have assumed
that SbCl
2
F
3
is three times more active than SbCl
4
F during
the exchange of ¯uorine. In accordance with this assumption
they compared the kinetic curves for solutions with the
initial composition: 1 mol SbCl
4
F ÷nCCl
4
and 1/3 mol
SbCl
2
F
3
÷nCCl
4
.
However, the authors did not take into account that the
exchange abilities of SbCl
4
F and SbCl
2
F
3
equalize with
increase in reaction temperature. By recalculation of the
number of moles of CCl
3
F formed in the reactions (6) and
(7) on 1 mol of SbCl
5÷n
F
n
instead of 1/n mol, the initial rates
of both reaction are the same at 26 8C. The kinetic curve for
the reaction (7) is characteristic for a system of irreversible
consecutive reactions with the rate constants being identical
or very close to each other. Thus, it can be assumed that
SbCl
2
F
3
participates in the three consecutive steps of reac-
tions (2), by exchanging consecutively with CCl
4
by one
¯uoride atom according to
SbCl
2
F
3
÷CCl
4
÷CCl
3
F ÷SbCl
3
F
2
(8)
SbCl
3
F
2
÷CCl
4
÷CCl
3
F ÷SbCl
4
F (9)
SbCl
4
F ÷CCl
4
÷CCl
3
F ÷SbCl
5
(6)
SbCl
2
F
3
÷3CCl
4
÷3CCl
3
F ÷SbCl
5
(7)
Assuming, that the above scheme describes the real reaction
mechanism and the rate constants of the reactions (6), (8)
and (9) are identical, then for a signi®cant excess of CCl
4
,
the kinetic equations are linear
d[SbCl
2
F
3
[
dt
= ÷k
ef
[SbCl
2
F
3
[
d[SbCl
3
F
2
[
dt
= k
ef
[SbCl
2
F
3
[ ÷[SbCl
3
F
2
[
d[SbCl
4
F[
dt
= k
ef
[SbCl
3
F
2
[ ÷[SbCl
4
F[
9
>
>
>
>
>
>
=
>
>
>
>
>
>
;
(4)
where [SbCl
x
F
x
] is the molar concentration of component, t-
reaction time, k
ef
the effective rate constant of reaction
determined as: k
ef
= kb, where k the rate constants of
reactions (6), (8), (9), b the initial concentration of CCl
4
.
The rate of (6) reaction can be given by the equation
d[SbCl
4
F[
dt
= ÷k
ef
[SbCl
4
F[ (5)
Denoting by X and Y the yields of reactions (6) and (7)
respectively, and determining them as the ratio of the mole
90 E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95
number of CCl
3
F to the initial mole number of SbCl
5÷n
F
n
,
respectively for n = 1 and 3, then we can obtain
X = 1 ÷e
÷k
ef
t
(6)
By the solution of the equation system (4) and for a short
reaction time, it can be demonstrated that the relationship
Y = f (t) is approximately linear
Y = k
ef
t (7)
The results of work performed by Kolditz and Schultz [7] are
consistent with this derived equation. The experimental
results of Kolditz are presented in Fig. 1 together with
the values calculated on the basis of Eqs. (6) and (7) for
k
ef
= 0:194 h
÷1
.
The proposed kinetic scheme indicates that in each ele-
mentary reaction one chlorine atom undergoes exchange,
according to reaction (2). The rate of the ®rst stage of
reaction describes the kinetic equation
r
n
= k[SbCl
5÷n
F
n
[[CCl
4
[ (8)
with the reaction rate constant k independent of n.
Our studies reported here permit the explanation of
literature reports concerning the kinetics and mechanism
of this complex process.
3. Results and discussion
3.1. The exchange reaction of fluorine between
antimony(V) chlorofluorides and CCl
4
(reaction (2))
The kinetic constant of reaction was calculated with the
assumption, that the reaction is second order and proceeds in
the liquid phase, thus
k =
2:303
t(a ÷b)
log
b(a ÷x)
a(b ÷x)
(9)
where a is initial concentration of HF, b the initial concen-
tration of CCl
4
, x the concentration of CCl
3
F after time t.
The average reaction rate constants for particular tem-
peratures fromthe range under investigation are summarized
in Table 1. Due to the excess of hydrogen ¯uoride it can be
expected that under the conditions investigated, the rate of
conversion of CCl
4
to CCl
3
F is dependent only on the CCl
4
concentration and the initial concentration of SbCl
5
, i.e. the
kinetic equation is ful®lled
r = kc
SbX
[CCl
4
[ (10)
This equation is valid when the total amount of SbCl
5
will be
transformed into the active form, i.e. when SbCl
5
= 0.
Taking into consideration the rate of reaction given by
r =
÷d[CCl
4
[
dt
(11)
after the integration of Eqs. (10) and (11) we can obtain
a = 1 ÷e
÷kc
SbX
t
(12)
Fig. 1. The yield of CCl
3
F in reaction (6) (^) and (7) (&) as a function of the reaction time.
Table 1
The kinetic constants of reaction in CCl
4
fluorination to CCl
3
F
Number Temperature (8C) Kinetic constant of reaction
k × 10
4
(dm
3
mol
÷1
s
÷1
)
1 0 0.33
a
2 13 0.69
a
3 26 1.67 ± 0.39
a
4 35 4.22 ± 0.39
5 45 4.47 ± 0.46
6 55 8.51 ± 0.71
7 65 15.3 ± 2.2
8 75 27.0 ± 3.5
9 85 34.3 ± 4.9
10 95 36.4 ± 5.6
a
Based on literature data.
E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 91
where a is the degree of CCl
4
conversion. This equation can
be written in a more convenient form for testing as
÷ln (1 ÷a) = kc
SbX
t (13)
Because of the uncertainly in the rate constant shown in
Eq. (9), testing of Eq. (13) separately for individual mea-
surements will give a wide scatter of results and it would not
be possible to show unequivocally that Eq. (13) correctly
describes the kinetics of transformation of CCl
4
to CCl
3
F.
For this reason it was veri®ed that
v the reaction rate constants calculated on the basis of
Eq. (13) comply with the Arrhenius equation;
v the values of the constants determined from this equation
(pre-exponential factor and the activation energy) are
consistent with the literature data.
The calculation were carried out in the following way:
v the concentration of c
SbX
catalyst was evaluated;
v for each measuring point (a, t) the value of reaction rate
constant, k, was calculated from Eq. (13);
v at each temperature after several reaction times, the value
of the constant, k, was averaged and the standard deviation
was calculated.
These magnitudes are summarized in Table 1 and they were
compared with literature data [7]. In order to determine the
catalyst concentration, the volume of the liquid phase was
calculated. It was assumed that the volume is an additive
function
v =
X
i
n
i
V
i
(14)
where n
i
is the number of mole component i of the liquid
phase, V
i
molar volume of component i. The molar volume
of particular components was calculated from the equation
V
i
=
M
i
d
i
(15)
where M
i
is the molecular weigh of component i, d
i
the
density of component i (HF, CCl
4
, CCl
3
F, SbCl
5
, SbClF
4
).
The numbers of moles of the respective reagents were
calculated with the following assumptions:
v the reaction between SbCl
5
and HF proceeds instanta-
neously and the total amount of SbCl
5
is transformed to
SbClF
4
;
v introduction of Cl
2
and formation of HCl occur in the
gaseous phase;
v the pressure in the reactor is so high that the remaining
components, i.e. HF, CCl
4
, CCl
3
F and SbClF
4
are in the
liquid phase.
From the mass balance calculation the catalyst concen-
trations c
SbX
are
c
SbX
=
1
0:968 ÷1:891(m=a) ÷0:19
(16)
c
SbX
is the catalyst concentration (mol dm
÷3
), m the mass of
mixture of CCl
4
, SbCl
5
, Cl
2
(g); a the degree of CCl
4
conversion.
The catalyst concentrations in the particular tests calcu-
lated from Eq. (16) vary from 0.8 to 0.9 mol dm
÷3
. Taking
into account simpli®cations, the average value of
c
SbX
= 0:85 mol dm
÷3
was assumed.
Based on the results summarized in Table 1, Fig. 2, line 1
was plotted as
ln k = f
1
T

(17)
Fig. 2. Verification of the Arrhenius equation for the reaction CCl
4
÷CCl
3
F (1) and CCl
3
F ÷CCl
2
F
2
(2) (~), (&) the result of our studies; (*) data from
work of Kolditz and Schultz.
92 E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95
This chart convincingly validates the assumed mechanismof
CCl
4
¯uorination. Most of the data points forma straight line
according to Arrhenius equation
ln k = ln k
0
÷
E
RT
(18)
These results are consistent with those obtained by Kol-
ditz et al. (Fig. 2, long line 1) [7]. The constants of the
Arrhenius equation determined from the graph are: pre-
exponential factor k
0
= 2:08 ×10
4
dm
3
mol
÷1
s
÷1
, and
the activation energy E = 46 kJ mol
÷1
.
3.2. Exchange reaction of fluorine between antimony(V)
chlorofluorides and CCl
3
F (reaction (3))
The second stage of CCl
4
¯uorination (reaction (3)) is
practically not recognised. It is known from the literature
[3,7] that the exchange of the second ¯uorine atom proceeds
with signi®cantly more dif®culty than the ®rst. From the
literature also, the only reaction product of CCl
4
with
SbCl
4
F (reaction (2)) at atmospheric pressure and tempera-
ture below the boiling point of CCl
4
(76.6 8C) is CCl
3
F. It
was found that a small amount of CCl
2
F
2
occurs at tem-
perature 77 8C after several hours of the reaction [7].
The results of the kinetic studies of CCl
3
F ¯uorination
were created similarly to the results of the ®rst stage
(¯uorination of CCl
4
). The following parameters were cal-
culated: the catalyst concentration (c
SbX
= 0:9 mol dm
÷3
),
the reaction rate constant, assuming that the second stage
proceeds in excess of hydrogen ¯uoride according to the
same mechanism as the ®rst. Thereby, the rate of R-3
reaction can be described by the kinetic equation
r = kc
SbX
[CCl
3
F[ (19)
The results of calculations of the reaction rate constants
and the degrees of conversions are summarized in Table 2.
After the calculation of the averages, the following values
of the rate constants were obtained:
k
95
·
C
= 2:49 ±0:73 ×10
÷4
dm
3
mol
÷1
s
÷1
k
85
·
C
= 1:39 ±0:26 ×10
÷4
dm
3
mol
÷1
s
÷1
k
75
·
C
= 1:20 ±0:41 ×10
÷4
dm
3
mol
÷1
s
÷1
The results are presented in Fig. 2 (line 2). Fromthis ®gure
results that the Arrhenius equation is also ful®lled for reaction
(3) (stage 2). The same slope of the straight line may indicate
that both stages of CCl
4
¯uorination proceed according to the
same mechanismwith the activation energy E = 46 kJ mol
÷1
.
The pre-exponential factor of reaction (3) is signi®cantly
smaller and amounts to k
0
= 8:01 ×10
÷12
dm
3
mol
÷1
s
÷1
.
As a result of the rapid reaction of SbCl
5
with HF, a
mixture of SbCl
5÷n
F
n
(1 _ n _ 4) is formed, the composi-
tion of which is dependent on both temperature and the
amount of HF. In previous work [7,11] it has been concluded
that SbCl
4
F and SbCl
2
F
3
play a signi®cant role in the
replacement of chlorine in CCl
4
and consecutive chloro-
¯uorocarbons. Our studies reveal that the rate of replace-
ment of the ®rst chlorine atom in CCl
4
does not depend on
the value of n in SbCl
5÷n
F
n
. At signi®cantly high tempera-
ture, this compound is SbClF
4
. Thus the course of ¯uorina-
tion can be presented as follows:
CCl
4
÷SbClF
4
÷CCl
3
F ÷SbCl
2
F
3
(10)
CCl
3
F ÷SbClF
4
÷CCl
2
F
2
÷SbCl
2
F
3
(11)
SbCl
2
F
3
÷HF ÷SbClF
4
÷HCl (12)
Reaction (12) proceeds so rapidly that the total amount of
SbCl
5
introduced occurs in the active form as SbClF
4
, i.e.
[SbClF
4
[ = [SbCl
5
[
0
= c
SbX
(20)
The rates reactions (10) and (11) which determine the ®rst
and the second stage of CCl
4
¯uorination can be described
by the following kinetic equations:
r
I
= k
I
c
sbx
[CCl
4
[ (21)
r
II
= k
II
c
SbX
[CCl
4
[ (22)
The reaction rate constants ful®ll the Arrhenius equation
k = k
0
e
÷E=RT
(23)
The activation energies of both reactions are the same and
the pre-exponential factors are given by, respectively
k
0I
= 2:08 ×10
4
dm
3
mol
÷1
s
÷1
k
0II
= 8:00 ×10
2
dm
3
mol
÷1
s
÷1
Based on Eqs. (20)±(23) any kinetic problem associated
the operation of the reactor for the synthesis may be solved.
In the case of a closed batch reactor, a fundamental kinetic
approach leads to the solution of the system of equations
d[CCl
4
[
dt
= ÷kc
SbX
[CCl
4
[
d[CCl
3
F[
dt
= kc
SbX
[CCl
4
[ ÷k
II
c
SbX
[CCl
3
F[
d[CCl
2
F
2
[
dt
= k
II
c
SbX
[CCl
3
F[
9
>
>
>
>
>
>
=
>
>
>
>
>
>
;
(24)
Table 2
Rate constants and degree of CCl
3
F conversion to CCl
2
F
2
Temperature
(8C)
Reaction
time (min)
Degree of
conversion (%)
Rate constant k ×10
4
(dm
3
mol
÷1
s
÷1
)
75 13 11.8 1.79
16 13.4 1.40
19 13.5 1.41
85 10 10.1 1.97
13 10.0 1.66
16 10.9 0.83
19 10.6 1.09
95 10 11.3 2.22
13 13.7 2.03
16 15.2 1.91
19 16.1 1.81
E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 93
The solution of system(24) for the isothermal process can be
presented in the analytical form
a
I
=
k
I
k
I
÷k
II
(e
÷k
II
c
SbX
t
÷e
÷kc
SbX
t
) (25)
a
II
= 1 ÷
k
II
k
I
÷k
II
e
÷kc
SbX
t
÷
k
I
k
I
÷k
II
e
÷k
II
c
SbX
t
(26)
a = 1 ÷e
÷k
SbX
t
(27)
where a is the total conversion of CCl
4
; a
I
, a
II
the conversion
of CCl
4
to CCl
3
F and CCl
2
F
2
, respectively.
The results of calculation for the conditions: temperature
95 8C, catalyst concentration 0.6 mol dm
÷3
(the closed, batch
reactor) are presented in Fig. 3. Such the calculations for
various catalyst concentrations, different types of the reactor,
enable the optimal selection of the process parameters.
4. Experimental
4.1. Raw materials
Hydrogen ¯uoride from a steel cylinder, purity 99.9 wt.%,
was the product of Lubon Chemical Works, Poland. Tetra-
chloromethane, purity 99.9 wt.%, after distillation was
obtained from Anwil, Poland. Antimony pentachloride,
purity _99 wt.%, from Fluka. Dichlorine from a steel
cylinder, purity 99.9 wt.% was the product ``Rokita'' Che-
mical Works, Poland. Chloro¯uorocarbons, CCl
3
F, CCl
2
F
2
;
purity 99.8 wt.% were obtained in Department of Organic
Technology, Technical University of Szczecin.
4.2. Apparatus and procedure
All studies have been performed in a 7 cm
3
stainless steel
autoclave which was charged by means of a syringe with a
mixture of CCl
4
or CCl
3
F with the catalyst SbCl
5
and dichlor-
ine. The molar ratio of SbCl
5
/Cl
2
was constant. After weigh-
ing, the autoclave was cooled to ca. ÷10 8C, anhydrous
hydrogen ¯uoride was introduced, and its amount determined
by weighing. The autoclaves were placed in a shaker and were
then submerged in oil bath. After a de®ned time the autoclaves
were taken out, cooled in a water bath with ice and were
unloaded. The gaseous content of the reactor was discharged
to a polyethylene tank under the water surface using a needle
valve. Above the tank was placed a polyethylene absorber,
which was sprinkled by cooled water. When the ¯ow of gases
was terminated while the needle valve was kept open, water
was sucked into the autoclave. The autoclave was opened, the
content was discharged to re¯ux from the absorber and the
concentration of the F
÷
ion was determined potentiometricly.
Measurements were performed with ion-selective electrode in
OP-108 apparatus (Radelkis). Cl
2
concentration was deter-
mined by iodometric titration. CCl
4
and CFCs were deter-
mined by gas chromatography. The mass balance of the F
÷
ion
and organic raw material was performed.
4.3. Gas chromatography conditions
A Chrom 5 apparatus equipped with 2:5 m×4 mm steel
column packed with Porapak Q (80±100 mesh) was
employed. The apparatus was equipped with a ¯ame ioniza-
tion detector (FID) with N
2
¯ow rate 20 cm
3
min
÷1
(carrier
gas), air ¯ow rate 300 cm
3
min
÷1
, H
2
¯ow rate
30 cm
3
min
÷1
. The temperature were: thermostat 170 8C,
detector and sample injector 250 8C.
4.4. General procedure for the synthesis
Experiments were carried out at a temperature range
between 35 and 95 8C with variation of the temperature
in 10 8C intervals, at a molar ratio of HF/CCl
4
varied over a
Fig. 3. The degree of CCl
4
conversion as a function of the time: a
I
, degree of CCl
4
conversion to CCl
3
F; a
II
, degree of CCl
4
conversion to CCl
2
F
2
; a, total
conversion.
94 E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95
range of 1.7÷1.2. The composition of raw material was as
follows: CCl
4
, 82 wt.%; SbCl
5
, 16 wt.%; Cl
2
, 2 wt.%. The
degree of conversion of CCl
4
to CCl
3
F was calculated using
the following de®nition
a =
n
0
HF
÷n
HF
n
0
CCI
4
×100% (28)
where n
0
HF
and n
HF
the number of moles of hydrogen ¯uoride
before and after the reaction, n
0
CCI
4
: the number of moles of
CCl
4
before the reaction.
References
[1] S. Brunet, C. Batiot, J. Barrault, M. Blanchard, J. Fluorine Chem. 59
(1992) 33±39.
[2] S. Brunet, R. Requieme, E. Colnay, J. Barrault, M. Blanchard, Appl.
Catal. B Environ. 5 (1995) 305±317.
[3] S. Brunet, C. Batiot, J. Barrault, M. Blanchard, J.M. Coustard, J.
Fluorine Chem. 63 (1993) 227±232.
[4] L. Kolditz, H Daunicht, Z. Anorg. Allg. Chem. 302 (1959) 230±
233.
[5] L. Kolditz, Z. Anorg. Allg. Chem. 289 (1957) 128±133.
[6] L. Kolditz, W. von Lieth, Z. Anorg. Allg. Chem. 310 (1961) 236±
242.
[7] L. Kolditz, S. Schultz, J. Fluorine Chem. 5 (1975) 141±150.
[8] B.V. Simakov, N.A. Korobochko, Teor. Osn. Chim. Techn. 3 (1980)
117±140, CA v.91, 63317h (1979).
[9] P.M. Chekmarev, B.V. Simakov, A.G. Oshuev, Raboty po, termodi-
nam. i kinetike chim. procesov 62 (1974) 99±102, CA v.83, 198336u
(1975).
[10] B.V. Siriakov, N.A. Kazakevicz, Modelirovane. i Upr. Chim.-Techn.
Procesami, Chimija, Moskva, 1981, pp. 42±48.
[11] S. Brunet, C. Batiot, P. Moriceau, N. Thybaud, J. Mol. Cat. A 142
(1999) 183±186.
E. Meissner, E. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 95

Further distillation of SbClF4ÁHF causes its decomposition with the evolution of chlorine. By recalculation of the number of moles of CCl3F formed in the reactions (6) and (7) on 1 mol of SbCl5ÀnFn instead of 1/n mol. From this value it can be concluded that it describes the rate of exchange of the ®rst chlorine atom in SbCl5. Simakov and Korobochko [8] have reported that the HF absorption is accompanied by the chemical reaction. (8) and (9) are identical. Thus. then for a signi®cant excess of CCl4. by exchanging consecutively with CCl4 by one ¯uoride atom according to SbCl2 F3 ‡ CCl4 3 CCl3 F ‡ SbCl3 F2 SbCl3 F2 ‡ CCl4 3 CCl3 F ‡ SbCl4 F SbCl4 F ‡ CCl4 3 CCl3 F ‡ SbCl5 SbCl2 F3 ‡ 3CCl4 3 3CCl3 F ‡ SbCl5 (8) (9) (6) (7) Kolditz and Schultz [7] believed that the effect of temperature can be attributed to the deactivation of SbCl4F. On the basis of the form of Eq.9. When the less advantageous variant is assumed and in the reaction (1). The kinetic model To our knowledge no paper has been published on the kinetics of ¯uorine exchange between SbCl5 and HF in the existing literature. E.10] are in agreement that at temperature higher than 20± 26 8C the reaction is (Eq. where k the rate constants of reactions (6). Meissner. kef the effective rate constant of reaction determined as: kef ˆ kb. This compound forms a stable complexes (SbCl4F)nÁSbCl5 at low temperatures in which all ¯uorine atoms are involved in bridge bonds. and simultaneous reduction of Sb(V) to Sb(III). (6) reaction at the temperature range between 20 and 50 8C can be described by the kinetic equation r ˆ k‰SbCl4 FŠ2 ‰CCl4 Š (2) Assuming. the initial rates of both reaction are the same at 26 8C. dt (4) where [SbClxFx] is the molar concentration of component. k the observed rate constant of reaction. only SbCl4F (n ˆ 1) is formed. [9]. Thus. SbCl2 F3 ‡ 3CCl4 3 3CCl3 F ‡ SbCl5 (7) where [SbCl5]2 and [HF] are molar concentrations. The rate of (6) reaction can be given by the equation d‰SbCl4 FŠ ˆ Àkef ‰SbCl4 FŠ dt (5) Kolditz and Schultz [7] claimed that at a temperature range between 0 and 26 8C the following equation is more appropriate r ˆ k‰SbCl4 FŠ‰CCl4 Š (3) Denoting by X and Y the yields of reactions (6) and (7) respectively. the authors did not take into account that the exchange abilities of SbCl4F and SbCl2F3 equalize with increase in reaction temperature. The results of studies performed by Chekmarev et al. In accordance with this assumption they compared the kinetic curves for solutions with the initial composition: 1 mol SbCl4 F ‡ nCCl4 and 1/3 mol SbCl2 F3 ‡ nCCl4 . it can be assumed that SbCl2F3 participates in the three consecutive steps of reactions (2). and determining them as the ratio of the mole . as long as HF is present in the reaction. does not lead to the formation SbF5 but to formation of SbClF4ÁHF.90 E. indicate that the rate of Eq. the role of which can be given by the kinetic equation R ˆ k‰SbCl5 Š2 ‰HFŠ (1) Different values of the activation energy obtained by both authors suggest that a change of the reaction mechanism take places with change in temperature. (9). one cannot evaluate whether further stages of exchange in (1) for n ˆ 2. SbCl5 is regenerated in the reactions (2)±(5) and is almost instantaneously ¯uorinated to SbCl4F. The reaction between antimony(V) chloro¯uorides and CCl4 was also presented as (7). The authors of the papers related to the kinetic aspect [4. (6)) irreversible SbCl4 F ‡ CCl4 3 CCl3 F ‡ SbCl5 (6) However. treaction time. However. 2. the principal process of CCl4 ¯uorination proceeds at a constant concentration of ¯uorinating agent. 3. 4 will proceed with the same reaction rate. that the above scheme describes the real reaction mechanism and the rate constants of the reactions (6). The papers discussing the kinetic equation of the reaction lead to contradictory conclusions. However. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 with annealing carried out for more than 2 h at temperature of 85 8C and distillation of HF. b the initial concentration of CCl4. (7) they have assumed that SbCl2F3 is three times more active than SbCl4F during the exchange of ¯uorine. thereby they cannot participate in the exchange reaction with CCl4. the kinetic equations are linear 9 d‰SbCl2 F3 Š > > ˆ Àkef ‰SbCl2 F3 Š > > > dt > = d‰SbCl3 F2 Š ˆ kef ‰SbCl2 F3 Š À ‰SbCl3 F2 Š > dt > > > > d‰SbCl4 FŠ > ˆ kef ‰SbCl3 F2 Š À ‰SbCl4 FŠ . The kinetic curve for the reaction (7) is characteristic for a system of irreversible consecutive reactions with the rate constants being identical or very close to each other. estimated at room temperature as k ˆ 1  103 dm6 molÀ2 sÀ1. (8). the authors have restricted their discussion only to a comparison of the kinetic curves of the reactions (6) and (7).

71 2.4 Æ 0.0 Æ 34. .3 Æ 36. the rate of conversion of CCl4 to CCl3F is dependent only on the CCl4 concentration and the initial concentration of SbCl5.5 4.6 Based on literature data. The yield of CCl3F in reaction (6) (^) and (7) (&) as a function of the reaction time.51 Æ 15. Due to the excess of hydrogen ¯uoride it can be expected that under the conditions investigated.9 5. The exchange reaction of fluorine between antimony(V) chlorofluorides and CCl4 (reaction (2)) The kinetic constant of reaction was calculated with the assumption. The proposed kinetic scheme indicates that in each elementary reaction one chlorine atom undergoes exchange. the kinetic equation is ful®lled r ˆ kcSbX ‰CCl4 Š (10) This equation is valid when the total amount of SbCl5 will be transformed into the active form. it can be demonstrated that the relationship Y ˆ f …t† is approximately linear Y ˆ kef t (7) The results of work performed by Kolditz and Schultz [7] are consistent with this derived equation. number of CCl3F to the initial mole number of SbCl5ÀnFn. 1.1.67 Æ 4. 1 together with the values calculated on the basis of Eqs. The experimental results of Kolditz are presented in Fig. according to reaction (2). Meissner. (10) and (11) we can obtain a ˆ 1 À eÀkcSbX t Table 1 The kinetic constants of reaction in CCl4 fluorination to CCl3F Number 1 2 3 4 5 6 7 8 9 10 a (12) Temperature (8C) 0 13 26 35 45 55 65 75 85 95 Kinetic constant of reaction k  104 (dm3 molÀ1 sÀ1) 0. i. b the initial concentration of CCl4.39a 0.69a 1. x the concentration of CCl3F after time t. The average reaction rate constants for particular temperatures from the range under investigation are summarized in Table 1.2 3. then we can obtain X ˆ 1 À eÀkef t (6) By the solution of the equation system (4) and for a short reaction time.3 Æ 27. thus kˆ 2:303 b…a À x† log t…a À b† a…b À x† (9) where a is initial concentration of HF.47 Æ 8.e.E. i. respectively for n ˆ 1 and 3. that the reaction is second order and proceeds in the liquid phase. Our studies reported here permit the explanation of literature reports concerning the kinetics and mechanism of this complex process. The rate of the ®rst stage of reaction describes the kinetic equation rn ˆ k‰SbCl5Àn Fn Š‰CCl4 Š (8) with the reaction rate constant k independent of n. 3.46 0. (6) and (7) for kef ˆ 0:194 hÀ1.33a 0. E.22 Æ 4. Taking into consideration the rate of reaction given by rˆ Àd‰CCl4 Š dt (11) after the integration of Eqs. when SbCl5 ˆ 0.39 0. Results and discussion 3. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 91 Fig.e.

was calculated from Eq. the value of the constant. (13). Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 where a is the degree of CCl4 conversion. This equation can be written in a more convenient form for testing as Àln …1 À a† ˆ kcSbX t (13) of particular components was calculated from the equation Vi ˆ Mi di (15) Because of the uncertainly in the rate constant shown in Eq. Based on the results summarized in Table 1. line 1 was plotted as   1 ln k ˆ f (17) T Fig.  the pressure in the reactor is so high that the remaining components.  the values of the constants determined from this equation (pre-exponential factor and the activation energy) are consistent with the literature data. (*) data from work of Kolditz and Schultz. (16) vary from 0. From the mass balance calculation the catalyst concentrations cSbX are cSbX ˆ 1 0:968 ‡ 1:891…m=a† À 0:19 (16) where ni is the number of mole component i of the liquid phase. Fig. was averaged and the standard deviation was calculated.e. The calculation were carried out in the following way:  the concentration of cSbX catalyst was evaluated. The molar volume cSbX is the catalyst concentration (mol dmÀ3).  introduction of Cl2 and formation of HCl occur in the gaseous phase. 2. Meissner. E. (9). The numbers of moles of the respective reagents were calculated with the following assumptions:  the reaction between SbCl5 and HF proceeds instantaneously and the total amount of SbCl5 is transformed to SbClF4.8 to 0. SbClF4). For this reason it was veri®ed that  the reaction rate constants calculated on the basis of Eq. These magnitudes are summarized in Table 1 and they were compared with literature data [7]. k. . (13) separately for individual measurements will give a wide scatter of results and it would not be possible to show unequivocally that Eq. The catalyst concentrations in the particular tests calculated from Eq.9 mol dmÀ3. 2. (13) comply with the Arrhenius equation.  for each measuring point (a. SbCl5. i. Cl2 (g).  at each temperature after several reaction times. k. SbCl5. (&) the result of our studies. the volume of the liquid phase was calculated. HF. CCl4.92 E. di the density of component i (HF. CCl3F. the average value of cSbX ˆ 0:85 mol dmÀ3 was assumed. a the degree of CCl4 conversion. Vi molar volume of component i. CCl4. It was assumed that the volume is an additive function X vˆ n i Vi (14) i where Mi is the molecular weigh of component i. In order to determine the catalyst concentration. Verification of the Arrhenius equation for the reaction CCl4 3 CCl3 F (1) and CCl3 F 3 CCl2 F2 (2) (~). testing of Eq. m the mass of mixture of CCl4. t) the value of reaction rate constant. (13) correctly describes the kinetics of transformation of CCl4 to CCl3F. CCl3F and SbClF4 are in the liquid phase. Taking into account simpli®cations.

1 Rate constant k  10 (dm3 molÀ1 sÀ1) 1.83 1. the rate of R-3 reaction can be described by the kinetic equation r ˆ kcSbX ‰CCl3 FŠ (19) The results of calculations of the reaction rate constants and the degrees of conversions are summarized in Table 2. ˆ kII cSbX ‰CCl3 FŠ dt 85 95 .0 10.5 10.3 13.66 0. Meissner.11] it has been concluded that SbCl4F and SbCl2F3 play a signi®cant role in the replacement of chlorine in CCl4 and consecutive chloro¯uorocarbons.91 1. the composition of which is dependent on both temperature and the amount of HF.1 10. 2. a fundamental kinetic approach leads to the solution of the system of equations 9 d‰CCl4 Š > > ˆ ÀkcSbX ‰CCl4 Š > > > dt > = d‰CCl3 FŠ (24) ˆ kcSbX ‰CCl4 Š À kII cSbX ‰CCl3 FŠ > dt > > > > d‰CCl2 F2 Š > .7 15. respectively k0I ˆ 2:08  104 dm3 molÀ1 sÀ1 k0II ˆ 8:00  102 dm3 molÀ1 sÀ1 Based on Eqs.40 1. The same slope of the straight line may indicate that both stages of CCl4 ¯uorination proceed according to the same mechanism with the activation energy E ˆ 46 kJ molÀ1. The constants of the Arrhenius equation determined from the graph are: preexponential factor k0 ˆ 2:08  104 dm3 molÀ1 sÀ1.81 4 k85  C ˆ 1:39 Æ 0:26  10À4 dm3 molÀ1 sÀ1 k75  C ˆ 1:20 Æ 0:41  10À4 dm3 molÀ1 sÀ1 The results are presented in Fig. From the literature also. ln k ˆ ln k0 À 3. (Fig.2. a mixture of SbCl5ÀnFn (1 n 4) is formed. ‰SbClF4 Š ˆ ‰SbCl5 Š0 ˆ cSbX (20) The rates reactions (10) and (11) which determine the ®rst and the second stage of CCl4 ¯uorination can be described by the following kinetic equations: rI ˆ kI csbx ‰CCl4 Š rII ˆ kII cSbX ‰CCl4 Š k ˆ k0 eÀE=RT (21) (22) The reaction rate constants ful®ll the Arrhenius equation (23) The activation energies of both reactions are the same and the pre-exponential factors are given by. Exchange reaction of fluorine between antimony(V) chlorofluorides and CCl3F (reaction (3)) The second stage of CCl4 ¯uorination (reaction (3)) is practically not recognised. It was found that a small amount of CCl2F2 occurs at temperature 77 8C after several hours of the reaction [7].03 1. It is known from the literature [3.6 11. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 93 This chart convincingly validates the assumed mechanism of CCl4 ¯uorination. this compound is SbClF4. The pre-exponential factor of reaction (3) is signi®cantly smaller and amounts to k0 ˆ 8:01  10À12 dm3 molÀ1 sÀ1.97 1. 2 (line 2).e. Our studies reveal that the rate of replacement of the ®rst chlorine atom in CCl4 does not depend on the value of n in SbCl5ÀnFn. long line 1) [7]. E. Thereby. The results of the kinetic studies of CCl3F ¯uorination were created similarly to the results of the ®rst stage (¯uorination of CCl4). and the activation energy E ˆ 46 kJ molÀ1. i.2 16.6 8C) is CCl3F. the reaction rate constant.4 13. In the case of a closed batch reactor. Thus the course of ¯uorination can be presented as follows: CCl4 ‡ SbClF4 3 CCl3 F ‡ SbCl2 F3 CCl3 F ‡ SbClF4 3 CCl2 F2 ‡ SbCl2 F3 (10) (11) SbCl2 F3 ‡ HF 3 SbClF4 ‡ HCl (12) Reaction (12) proceeds so rapidly that the total amount of SbCl5 introduced occurs in the active form as SbClF4. the only reaction product of CCl4 with SbCl4F (reaction (2)) at atmospheric pressure and temperature below the boiling point of CCl4 (76.41 1.22 2. The following parameters were calculated: the catalyst concentration (cSbX ˆ 0:9 mol dmÀ3). At signi®cantly high temperature.7] that the exchange of the second ¯uorine atom proceeds with signi®cantly more dif®culty than the ®rst.8 13.79 1. (20)±(23) any kinetic problem associated the operation of the reactor for the synthesis may be solved. From this ®gure results that the Arrhenius equation is also ful®lled for reaction (3) (stage 2). assuming that the second stage proceeds in excess of hydrogen ¯uoride according to the same mechanism as the ®rst.E. As a result of the rapid reaction of SbCl5 with HF.9 10.09 2. the following values of the rate constants were obtained: k95  C ˆ 2:49 Æ 0:73  10À4 dm3 molÀ1 sÀ1 Table 2 Rate constants and degree of CCl3F conversion to CCl2F2 Temperature (8C) 75 Reaction time (min) 13 16 19 10 13 16 19 10 13 16 19 Degree of conversion (%) 11. In previous work [7. Most of the data points form a straight line according to Arrhenius equation E (18) RT These results are consistent with those obtained by Kolditz et al. After the calculation of the averages.

anhydrous hydrogen ¯uoride was introduced. aII. degree of CCl4 conversion to CCl3F. from Fluka. CCl3F. was the product of Lubon Chemical Works. 4. The apparatus was equipped with a ¯ame ionization detector (FID) with N2 ¯ow rate 20 cm3 minÀ1 (carrier gas). air ¯ow rate 300 cm3 minÀ1. Such the calculations for various catalyst concentrations. General procedure for the synthesis Experiments were carried out at a temperature range between 35 and 95 8C with variation of the temperature in 10 8C intervals.9 wt. The results of calculation for the conditions: temperature 95 8C. Apparatus and procedure All studies have been performed in a 7 cm3 stainless steel autoclave which was charged by means of a syringe with a mixture of CCl4 or CCl3F with the catalyst SbCl5 and dichlorine.94 E. À10 8C. 4. at a molar ratio of HF/CCl4 varied over a .%. Technical University of Szczecin. Poland. The degree of CCl4 conversion as a function of the time: aI. cooled in a water bath with ice and were unloaded. catalyst concentration 0.8 wt. Antimony pentachloride. and its amount determined by weighing.6 mol dmÀ3 (the closed. Chloro¯uorocarbons.%.3.9 wt. purity 99. 4. The molar ratio of SbCl5/Cl2 was constant. which was sprinkled by cooled water. CCl4 and CFCs were determined by gas chromatography. The autoclave was opened. Tetrachloromethane. enable the optimal selection of the process parameters.%. E. 3. The temperature were: thermostat 170 8C. aI. Cl2 concentration was determined by iodometric titration. When the ¯ow of gases was terminated while the needle valve was kept open. Raw materials Hydrogen ¯uoride from a steel cylinder.4. Poland. batch reactor) are presented in Fig. water was sucked into the autoclave.1. CCl2F2. 3. The mass balance of the FÀ ion and organic raw material was performed. respectively. a. purity 99. degree of CCl4 conversion to CCl2F2. the autoclave was cooled to ca. after distillation was obtained from Anwil.% were obtained in Department of Organic Technology. purity 99. The autoclaves were placed in a shaker and were then submerged in oil bath. the content was discharged to re¯ux from the absorber and the concentration of the FÀ ion was determined potentiometricly. The solution of system (24) for the isothermal process can be presented in the analytical form aI ˆ kI …eÀkII cSbX t À eÀkcSbX t † kI À kII kII kI eÀkcSbX t À eÀkII cSbX t kI À kII kI À kII (25) (26) (27) aII ˆ 1 ‡ a ˆ 1 À eÀkÁSbX t where a is the total conversion of CCl4. After weighing. purity !99 wt. Above the tank was placed a polyethylene absorber. Gas chromatography conditions A Chrom 5 apparatus equipped with 2:5 m  4 mm steel column packed with Porapak Q (80±100 mesh) was employed. 4. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 Fig. Experimental 4.2. Dichlorine from a steel cylinder. aII the conversion of CCl4 to CCl3F and CCl2F2. purity 99. H2 ¯ow rate 30 cm3 minÀ1. Poland. different types of the reactor. After a de®ned time the autoclaves were taken out.% was the product ``Rokita'' Chemical Works. Meissner.9 wt. detector and sample injector 250 8C. The gaseous content of the reactor was discharged to a polyethylene tank under the water surface using a needle valve. Measurements were performed with ion-selective electrode in OP-108 apparatus (Radelkis). total conversion.

Simakov. Brunet. i Upr. H Daunicht.%. The composition of raw material was as follows: CCl4.G.M. N. J. Meissner. B Environ. A. 3 (1980) 117±140. E. 5 (1995) 305±317. Blanchard. 2 wt. Batiot. Anorg. [5] L. J. N. Barrault. [8] B. W. Requieme. [4] L.M. Schultz. S. J. Z. Fluorine Chem.%. 1981. Siriakov. C. J. i kinetike chim. [11] S. Moriceau.V. Chem. Catal. J.V. 302 (1959) 230± 233. Blanchard. Colnay. 42±48.2. Modelirovane. Mol. Korobochko. 5 (1975) 141±150. Brunet.83. Kolditz. [3] S. C. [10] B. Simakov. Cat. Anorg. E. Anorg.A. R. A 142 (1999) 183±186. 59 (1992) 33±39. Chekmarev. pp. Z. procesov 62 (1974) 99±102. Fluorine Chem.E. References [1] S. Batiot. Osn. Chem. Chim.7À1.%. Raboty po. Allg. Cl2. Coustard. [9] P. Z. [2] S. Kolditz. SbCl5. Blanchard. P. Barrault. Techn. 310 (1961) 236± 242. Brunet. 63317h (1979). Chem. Moskva. n0 4 : the number of moles of CCI CCl4 before the reaction. Kolditz. .91. M. Chim. Barrault. Milchert / Journal of Fluorine Chemistry 119 (2003) 89±95 95 range of 1. J. [7] L. Chimija. CA v.A. Procesami. B. Teor. 82 wt. von Lieth. CA v. Brunet. 63 (1993) 227±232. C. Fluorine Chem. M. Allg. Kolditz. Appl. The degree of conversion of CCl4 to CCl3F was calculated using the following de®nition aˆ n0 À nHF HF  100% n0 4 CCI (28) where n0 and nHF the number of moles of hydrogen ¯uoride HF before and after the reaction.V. M. 289 (1957) 128±133. J. Batiot. termodinam. Oshuev. N. 198336u (1975). [6] L. J. Thybaud. Allg. Kazakevicz. 16 wt.-Techn.