Solvent effects in the fluorination of 1,2-dichloro-1,1-difluoroethane (R-132b) to 2-chloro-1,1,1-trifluoroethane (R-133a)

Article abstract of DOI:10.1016/j.jfluchem.2006.08.002

Trichloroethylene has been found to act as a rate enhancing co-factor in the liquid phase, tantalum (V) halide catalyzed, fluorine-for-chlorine exchange reaction of 1,2-dichloro-1,1-difluoroethane (R-132b) to 2-chloro-1,1,1-trifluorethane (R-133a). Several trifluoromethyl substituted benzenes have also been found to be rate-enhancing solvents.

Full text of DOI:10.1016/j.jfluchem.2006.08.002

Journal of Fluorine Chemistry 127 (2006) 1606–1610
www.elsevier.com/locate/fluor
Short communication
Solvent effects in the fluorination of 1,2-dichloro-1,1-difluoroethane
R-132b) to 2-chloro-1,1,1-trifluoroethane (R-133a)
(
*
Randolph K. Belter , Nanaji K. Bhamare
Department of Chemistry and Physics, Southeastern Louisiana University, Hammond, LA 70402, United States
Received 11 July 2006; received in revised form 28 July 2006; accepted 8 August 2006
Available online 17 August 2006
Abstract
Trichloroethylene has been found to act as a rate enhancing co-factor in the liquid phase, tantalum (V) halide catalyzed, fluorine-for-chlorine
exchange reaction of 1,2-dichloro-1,1-difluoroethane (R-132b) to 2-chloro-1,1,1-trifluorethane (R-133a). Several trifluoromethyl substituted
benzenes have also been found to be rate-enhancing solvents.
#
2006 Elsevier B.V. All rights reserved.
Keywords: Trichloroethylene; 1,2-Dichloro-1,1-difluoroethane; 2-Chloro-1,1,1-trifluorethane; TaClF
4
1
. Introduction
Mechanistically, the first reaction can be broken down into
three individual steps (see Scheme 1, Reaction 1). First, a
molecule of HF adds across the TCE double bond to produce
1,1,2-dichloro-1-fluoroethane (R-131a). Second, direct fluor-
ine-for-chlorine exchange converts R-131a to 1,2-dichloro-1,
1-difluoroethane (R-132b) [7–9]. The third reaction is another
fluorine-for-chlorine exchange that converts R-132b to R-133a.
In the normally vapor phase second reaction, R-133a is
converted to R-134a (as the fourth and last individual step) (see
Scheme 1, Reaction 2).
For decades chlorofluorocarbons have been useful chemicals
for refrigeration, solvent, foam manufacture and firefighting
applications. The refrigerant R-12 (CF Cl ) was the standard
2
2
refrigerant and found widespread use in automotive air
conditioners. The discovery of the harmful nature of chloro-
fluorocarbons towards the Earth’s protective ozone layer led to
the outlawing of the manufacture and use of most of these
chemicals in the 1989 Montreal Protecol. The most popular
non-ozone depleting replacement for R-12 for use in
automotive air conditioning units has been R-134a (CF3–
In the liquid phase reaction of TCE to R-133a, many
transition metal halide Lewis acid catalysts have been reported
to be effective. Antimony (V) halides are the benchmark
catalyst for chlorine-for-fluorine exchange (Swartz reaction)
but suffer from reduction to Sb (III) at temperatures above
80 8C [10]. The most current patent literature indicates that
tantalum (V) halides and niobium (V) halides are the best
choices for TCE to R-133a conversion [11–14]. In repetitions of
the literature experiments, we found that tantalum (V) halides
were superior to niobium (V) halides in converting TCE to R-
133a.
CH F). There are several manufacturers of R-134a in the US [1]
2
as well as many overseas.
The production of R-134a generally begins with trichloro-
ethylene (TCE) as feedstock. The processes are typically
performed as two distinct reactions. First, TCE is fluorinated
under catalytic conditions to 2-chloro-1,1,1-trifluoroethane (R-
1
33a). This can be done in the liquid or vapor phase. In the
second reaction, R-133a is further fluorinated to R-134a [2]. As
this reaction is more difficult, it is most successfully performed
as a high temperature vapor phase reaction over an alumina or
chromia catalyst [3–6].
Procedurally, these reactions were run at 140 8C. A blank
run was performed without TCE, cycling catalyst and HF
through the reaction warm-up and heating sequence in order to
fluorinate TaCl to TaF . This would ensure that the catalyst
5
5
was fluorinated for the first run to the same degree that it would
be for subsequent runs. For each run, the conversion rate
decreased with time as TCE concentration decreased and HCl
*
Corresponding author. Tel.: +1 985 549 3529; fax: +1 985 549 5126.
E-mail address: rbelter@selu.edu (R.K. Belter).
0022-1139/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2006.08.002

R.K. Belter, N.K. Bhamare / Journal of Fluorine Chemistry 127 (2006) 1606–1610
1607
investigate this possibility, experiments were repeated on
commercial R-132b with commercial TaF . The procedure of
Feiring [16], who reports the reaction of R-132b with HF and
5
2
% TaF at 100 8C, was followed closely. However, after
5
heating to 100 8C followed by 20 h at 140 8C only 25%
conversion to R-133a had occurred. Three hours at 160 8C did
not change this result. This contradiction to the reported
literature led us to consider the possibility that our catalyst was
too fluorinated. Certainly, in the antimony (V) literature, it is
well known that mixed chlorofluorides like SbCl F are active
Scheme 1.
pressure increased. While the reaction could be driven to
completion with the venting of HCl pressure, reactions were
considered complete when the TCE concentration was <5%.
Without venting HCl, this TCE level remained unchanged with
additional reaction time. Reactions were ‘‘complete’’ after
2
3
and highly soluble species [17]. Perhaps the reaction of TCE to
R-132b is facile enough to generate an initiating concentration
of HCl, which through equilibria, would prevent some or all of
the catalyst from fluorinating completely to (apparently
inactive) TaF5.
1
.5 h at reaction temperature.
For every liquid phase fluorination reaction, a certain
amount of oligomerization occurs, evidenced by the formation
of higher boiling liquid by-products. Such by-products are
referred to as ‘‘tars’’. It was observed that for 1-mol batch
reactions of TCE to R-133a catalyzed by 4 mol% tantalum (V)
halide, tar levels were around 9% (w/w). While the catalyst
remained active through five consecutive batches, it is clear that
in a continuous process the build-up of tars would eventually
deactivate the catalyst.
Two sets of experiments were undertaken to test this theory.
First, an analysis was performed on the fluoride content of the
tantalum (V) halide that was produced by treating TaCl with
5
HF at 140 8C. Invariably, the fluoride content of samples
throughout this study were found to be 80% of theroretical for
TaF ! This result indicates the empirical formula of ‘‘fully
5
fluorinated’’ tantalum (V) to beTaClF [18,19]. Also indicated
4
is that this mixed halide is just as inactive as TaF without TCE
5
In an effort to minimize the contact of tars with the
expensive tantalum (V) catalyst, the reaction of TCE to R-133a
was separated into its individual reactions. The working
hypothesis was that tar generation occurred principally from
oligomerization of the olefinic TCE. Conversion of R-131a to
R-132b and R-132b to R-133a do not occur through an olefinic
intermediate and so are expected to give lower tars. As such,
one could potentially perform the HF addition to TCE with an
as a co-reagent.
The second set of experiments were to run R-132b + HF
reactions with either TaF or TaClF under the pressure of at
least 1 equiv. of anhydrous HCl gas. If TaCl F , etc. is the active
5
4
2
3
species and that species is generated by fortuitous HCl from the
preceding reaction steps, then this addition of HCl to the system
should make for a reactive system. It did not. We observed no
enhancement in the rate of R-133a formation.
inexpensive and expendable Lewis acid catalyst (i.e. TiCl ) and
4
perform the more demanding, but cleaner fluorine-for-chlorine
exchanges with the tantalum catalyst.
2.2. Reactivity of R-131b
The poor reactivity of R-132b under the tantalum halide
catalyzed conditions made us question the mechanistic precept
that R-131a ! R-132b ! R-133a proceeds by direct fluorine-
for-chlorine exchange. While no olefinic intermediates were
observed on GC chromatograms during the reactions of TCE, it
was wondered whether saturated species were not necessarily
the mechanistic intermediates on the pathway to fluorination.
Possibly R-131a and R-132b are only observed because they
are mechanistic dead-ends that have to revert back to an olefin
(if not back to TCE) in order to fluorinate further. As such,
R-131a would be expected to be relatively unreactive, similar to
R-132b. Experimentally, the results were quite the opposite. At
110 8C, R-131a began reacting and within 1.5 h was completely
consumed. The product was 22% R-133a and 74% R-132b.
Interestingly, additional reaction time at 140 8C did not
complete the conversion to R-133a. Again, R-132b was
resisting fluorination under tantalum catalysis! In retrospect,
this was not really surprising. In the reactions of TCE, R-132b
concentrations build and then convert to R-133a while the R-
131a concentration remains low and constant. Obviously the
rate of conversion for R-131a to R-132b is faster than that of R-
132b to R-133a. If the good reactivity of R-131a is attributed to
its ability to eliminate to a mechanistically active olefin
2
. Results and discussion
2
.1. Non-reactivity of R-132b
It was decided to test the hypothesis on the fluorination of
R-132b to R-133a (rather than R-131a). The R-132b was
isolated from the products of earlier TCE runs. After a blank,
catalyst fluorinating run, the R-132b was added and heated to
1
40 8C precisely like the reference TCE reaction. Surprisingly,
the reaction was extremely sluggish. After 1.5 h the conversion
to R-133a was only 36%. A subsequent run had only 11%
conversion and a third run, 3% conversion, even at high
temperature (150 8C). What little activity the catalyst had,
decreased quickly. The first instinct was to question the quality
of ‘‘homemade’’ R-132b. A commercial sample was purchased
and the experiment repeated, again first running a fluorinating
run on the TaCl . Results were again poor, in fact worse. A
5
conversion of 22% for the first run was followed conversions of
1
% for runs at 160 8C for as long as 12 h!
Next the catalyst came under suspicion. Perhaps the
prefluorination runs were not converting TaCl to TaF , though
5
5
the literature indicated that this was a facile process [11,15]. To

1608
R.K. Belter, N.K. Bhamare / Journal of Fluorine Chemistry 127 (2006) 1606–1610
(
CFCl CHCl) then the question still remains as to why R-132b
solubility of HF in TCE is much lower than that of HF in R-
132b. As one might expect, the more fluorinated species
solubilizes HF to the greater extent. Further, the catalyst is
almost exclusively dissolved in the HF, and the resulting
ionic solution even less miscible with the organic phase.
The possibility that TCE was forming a soluble complex
with the tantalum halide catalyst was explored by running
reactions with the close TCE analogs, 1,2-dichloroethylene and
tetrachloroethylene. In short, neither chloroolefin enhanced
the reaction. Reactions of R-132b with up to an equal volume
of t-1,2-dichloroethene exhibited only 8% conversion after
23 h. Similarly, after 50 h in the presence of tetrachloro-
ethylene, only 33% conversion of R-132b to R-133a had
occurred. Both versions of the reaction could be ‘‘rescued’’ and
taken to high conversion with the addition of TCE. It appeared,
then, that if TCE were acting as a complexing agent at all, the
structural requirements for activity were very specific. (It
should be noted that tetrachloroethylene and dichloroethylene
are themselves rather unreactive to tantalum/HF conditions
[16] and maintained a good concentration throughout the
course of the reaction trials.)
does not do the same. The elimination of HCl from R-131a
becomes thermodynamically favorable above 58 8C, while the
elimination of HCl from R-132b becomes favorable above
8
those temperatures.
2 8C [7]. In either case, reaction conditions have been above
2
.3. Catalytic activity of TCE and R-131b
One thing that had been evident in the original TCE
experiments was that the reaction ceased to progress when the
TCE levels fell to ꢀ5%. Originally, the simple assumption was
made that the pressure of by-product HCl was restraining the
equilibrium. However, even with venting of HCl, it was difficult
to react out the last of the R-132a. It was now surmised that
reactant TCE was possibly playing a role in accelerating the
reaction of R-132b to R-133a. As such, this idea was tested by
running R-132b to R-133a reactions spiked with TCE.
A prefluorinating reaction of HF with 4% TaCl (versus
5
anticipated R-132b) was run, followed by attempted reaction of
R-132b without TCE in order to establish that the system was
non-reactive. To this non-reactive system was added 10 mol%
TCE and the reaction resumed. Within 15 min the conversion to
R-133a was 87% (with 60% of the TCE remaining). At 1.75 h
conversion was >99% (with only 40% of the TCE remaining).
The reaction was stopped, vented and only R-132b was
recharged. Upon reheating, the reaction again initiated.
Conversion at 1.75 h was 99%. Upon reaching reaction
temperature, TCE was down to 0.4% and was nearly
undetectable throughout the reaction. Batch reactions could
be run to completion indefinitely by charging 15% TCE into
every other run.
Solvent effect. At this point, the simple solvent effect as
cause for the reaction enhancing properties of TCE or R-131a
was begging to be investigated. As such, several alternative
solvents were screened to see if simple solvation could enhance
the reaction rate. Solvents were chosen on their ability to
withstand HF/Lewis acid conditions and their potential for use
on industrial scale (some were even chosen for their potential
complexing ability). Each solvent was scoped for activity by
testing it with R-132b and preformed TaClF . Exploratory
4
reactions were performed where the solvent was used in equal
volume to the R-132b substrate as well as at the 15% level
where TCE had good enhancing effect.
As has been noted above, the reaction of pure R-131a to R-
33a proceeds, but mostly produces R-132b. We noted at this
1
The straight-forward solvent perfluorononane showed no
enhancement activity. Nor did the straight chain, but polar
perfluorooctanesulfonyl fluoride, nor the aromatic and polar
benzenesulfonyl fluoride. R-133a was tested as an obvious
choice, but also showed no activity. Success came quickly with
the use of m-(bis)-trifluoromethylbenzene (see Fig. 1). Truth-
fully, this solvent was chosen by hypothesizing on a bidentate
point that the R-131a level had decreased to 2%. It seemed
possible that the reaction had ‘‘hung-up’’ because, like TCE, R-
1
31 was a component that could make for an active system. In
practice, this appears to be true as a non-reactive R-132b
reaction, as prepared above, became reactive with the addition
of 15 mol% R-131a. The reaction appears to be a bit slower
than the TCE activated reaction until one realizes that 90% of
the R-131a had reacted as well as the R-132b. As such, at 1.75 h
the conversion was 99%.
Lewis base, one with –CF pincers. While the reactivity was
3
welcome, the pincer theory was quickly discounted when
p-(bis)-trifluoromethylbenzene was found to be equally active
as a rate enhancing agent. However, now it was surmised that
strong solvation of the catalyst was coming as a result of
catalyst complexation by the aromatic ring [20]. Follow-up
2
.4. Mode of activity of TCE and R-131a
With the strong evidence for the critical role of TCE and/or
R-131a in the fluorine-for-chlorine substitution reaction of R-
32b to R-133a, several possible modes of activity were
1
considered. They were, chloride source, catalyst complexing
agent and solvent effect.
Chloride source. The concept of TCE or R-133a acting as
chloride ion sources was explored above when the mixed halide
nature of the tantalum chlorofluoride catalyst was investigated.
That role was thus discounted.
Complexing agent. A simple solvent effect is an unlikely
cause for reaction enhancing properties of TCE or R-131a. The
Fig. 1. R-132b ! R-133a conversion with equal volume of solvent.

R.K. Belter, N.K. Bhamare / Journal of Fluorine Chemistry 127 (2006) 1606–1610
1609
readily occur in liquid hydrogen fluoride with tantalum (V)
halide catalysis. A co-reagent of trichloroethylene or 1,1,2-
trichloro-1-fluoroethane (R-131a) is required in 10–15% molar
concentration to affect full conversion. Alternatively, a solvent
of CF -, F-, Cl- or NO -substituted benzotrifluoride can also
3
2
effect the conversion.
The nature and action of these and other co-reagents is under
investigation and will be reported when sufficient data is
collected to define the mechanism of action.
Fig. 2. R-132b ! R-133a conversion with 15% volume of solvent.
4. Experimental
4.1. General
Anhydrous hydrogen fluoride was Matheson-Trigas CP
grade. Trichloroethylene was PPG Fluorocarbon Grade TR/119
and was used without removal of inhibitor. Trifluoromethyl-
and fluoro-substituted benzenes, perfluorononane, perfluor-
ooctane sulfonyl fluoride and 97% R-132b were from Synquest
Labs. R-133a was from Halocarbon. Tantalum (V) chloride was
NOAH Technologies 99.99%. Tantalum (V) fluoride, t-
dichloroethylene, tetrachloroethylene, chlorobenzene and dini-
trobenzenes were Aldrich. Benzene sulfonyl fluoride was
prepared in situ from J.T. Baker benzene sulfonyl chloride.
Reactions were performed in a Parr 300 mL hastelloy mini-
reactor. GC was performed on an HP 5730A spectrometer with
a 60 M DB-1 capillary column.
Fig. 3. R-132b ! R-133a conversion with equal volume of a nitro-benzotri-
fluoride.
testing was performed to see how much or how little electron-
withdrawing effect was allowable to make for an activating
aromatic system.
A single –CF or a single –Cl substituent is insufficient
3
to empart reaction enhancing ability. Benzotrifluoride
Cautionary note. Anhydrous HF causes instantaneous
severe burns to the skin and mucous membranes. HF should
be handled with full PPE protection. An ample supply of HF
antidote gel should be kept on hand before handling HF. See
reference for burn treatment procedures [21].
(
trifluoromethylbenzene) and chlorobenzene offered no
enhancement. However the paired combination of –CF and
3
–Cl did. Both m-chlorobenzotrifluoride and p-chlorobenzotri-
fluoride were rate-enhancing solvents. Negative results for m-
difluorobenzene indicates that two fluorine substituents are not
enough to impart activity. However, pairing a –CF group with
4.1.1. Example 1. Hydrofluorination of R-132b in the
presence of 15% trichloroethylene enhancer
Seven hundred and twenty milligrams (0.004 mol) of
3
a single F, as m-fluorobenzotrifluoride, afforded another rate
enhancing solvent.
Interestingly, at equal volume, all three ‘‘mixed’’ aromatics
lagged behind the (bis)-trifluoromethylbenzenes. However, at
the 15% level, the reactivities were on par (see Fig. 2).
The pairing of a nitro substituent with a trifluoromethyl
group also makes for an activating aromatic solvent, but only in
one of the three cases. m-Nitrobenzotrifluoride enhanced the
rate of R-132b to R-133a conversion, but to a lesser extent than
any other active solvent. o- and p-Nitrobenzotrifluoride showed
no enhancement activity. Interestingly, m-nitrobenzotrifluoride
exhibited an induction period of about 1 h where it showed zero
activity (see Fig. 3). Once the enhancement began, the reaction
proceeded with a linear increase in conversion from then on.
This curious result is under active investigation.
TaCl was charged to the reactor. The reactor was evacuated
5
and cooled with ice. Fifty grams (2.5 mol) of anhydrous
hydrogen fluoride was charged. The solution was heated with
stirring to 140 8C for 1 h. The reactor was again cooled with
ice. A mixture of 13.4 g (0.1 mol) 1,2-dichloro-1,1-difluor-
oethane and 1.97 g (0.15 mol) trichloroethylene was injected
and the reactor heated to 140 8C. Samples were drawn from
the reactor headspace and the reaction was monitored by GC.
See Fig. 2 for results.
4.1.2. Example 2. Hydrofluorination of R-132b in the
presence of 15% 1-fluoro-1,2,3-trichloroethane (R-131a)
enhancer
Pairing two nitro groups was not sufficient to make an
activating solvent. No reaction was observed with m-
dinitrobenzene in the reaction mixture.
Seven hundred and twenty milligrams (0.004 mol) of TaCl5
was charged to the reactor. The reactor was evacuated and
cooled with ice. Fifty grams (2.5 mol) of anhydrous hydrogen
fluoride was charged. The solution was heated with stirring to
140 8C for 1 h. The reactor was again cooled with ice. A
mixture of 13.4 g (0.1 mol) 1,2-dichloro-1,1-difluoroethane
and 2.2 g (0.15 mol) 1-fluoro-1,2,3-trichloroethane was
injected and the reactor heated to 130 8C. Samples were drawn
3
. Conclusion
The conversion of 1,2-dichloro-1,1-difluoroethane to 2-
chloro-1,1,1-trifluoroethane (R-132b ! R-133a) does not

1610
R.K. Belter, N.K. Bhamare / Journal of Fluorine Chemistry 127 (2006) 1606–1610
from the reactor headspace and the reaction was monitored by
GC. Conversion at 1 h was 87.5%.
with ice. Fifty grams (2.5 mol) of anhydrous hydrogen fluoride
was charged. 13.4 g (0.1 mol) 1,2-dichloro-1,1-difluoroethane
was injected. Then 3.6 g (0.1 mol) anhydrous hydrogen
chloride was injected and the reactor heated to 140 8C for
3 h. No reaction occurred. The reactor was cooled with ice and
an additional 5.8 g (0.16 mol) of anhydrous hydrogen chloride
was injected, pressurizing the reactor to 200 psi. The reaction
was heated to 140 8C for 3.5 h. No reaction was observed.
4.1.3. Example 3. Hydrofluorination of R-132b in the
presence of equal volume solvent enhancer
Seven hundred and twenty milligrams (0.004 mol) of TaCl5
was charged to the reactor. The reactor was evacuated and
cooled with ice. Fifty grams (2.5 mol) of anhydrous hydrogen
fluoride was charged. The solution was heated with stirring to
1
40 8C for 1 h. The reactor was again cooled with ice. A
Acknowledgement
mixture of 13.4 g (0.1 mol) 1,2-dichloro-1,1-difluoroethane
and 10 mL of the chosen solvent was injected and the reactor
heated to 140 8C. Samples were drawn from the reactor
headspace and the reaction was monitored by GC. See Fig. 1 for
results.
This project was performed under University contract to
Norphlet Chemical, Norphlet AR with consultation by B.
Boyce & Associates, Houston, TX.
References
4.1.4. Example 4. Hydrofluorination of R-132b in the
presence of 15% volume solvent enhancer
[
[
[
1] DuPont & Co., Ineos, Ltd., Arkema Inc., Honeywell, Inc.
2] L. Manzer, US Patent 5,185,482 (1993).
Seven hundred and twenty milligrams (0.004 mol) of TaCl5
was charged to the reactor. The reactor was evacuated and
cooled with ice. Fifty grams (2.5 mol) of anhydrous hydrogen
fluoride was charged. The solution was heated with stirring to
3] S. Brunet, B. Requieme, E. Colnay, J. Barrault, M. Blanchard, Appl. Catal.
B 5 (1995) 305.
[4] J. Lu, H. Yang, S. Chen, L. Shi, J. Ren, H. Li, S. Peng, Catal. Lett. 41
1996) 221.
5] S. Brunet, B. Boussand, A. Rousset, D. Andre, Appl. Catal. A 168 (1998)
7.
6] Y. Chung, H. Lee, H. Jeong, Y. Kim, H. Lee, H. Kim, S. Kim, J. Catal. 175
1998) 220.
[7] A. Kohne, E. Kemnitz, J. Fluor. Chem. 75 (1995) 103.
(
1
40 8C for 1 h. The reactor was again cooled with ice. A
mixture of 13.4 g (0.1 mol) 1,2-dichloro-1,1-difluoroethane and
.5 mL the chosen solvent was injected and the reactor heated
[
[
5
1
to 140 8C. Samples were drawn from the reactor headspace and
the reaction was monitored by GC. See Fig. 2 for results.
(
[
8] A. Dmitiriev, I. Trukshin, P. Smykalov, Russ. J. Appl. Chem. 75 (2002)
71.
9] E. Kemnitz, K.-U. Niederson, A. Kohne, Indian J. Chem. A and B 36
1977) 485.
7
4
.1.5. Example 5. Preparation and analysis of TaClF5
01.0 g (0.28 mol) TaCl was charged to the reactor. The
[
1
5
(
reactor was cooled with dry ice/acetone. One hundred and
twenty five grams (6.25 mol) of anhydrous hydrogen fluoride
was charged. The solution was heated 3 h. The reaction was
cooled in a water bath to 30 8C and the HCl pressure vented.
The reactor was held then under aspirator vacuum for 60 min.
[
10] L. Manzer, V. Rao, Advances in Catalysis, vol. 39, Academic Press, 1993,,
p. 329.
[
[
11] W. Gumprecht, W. Schindel, V. Felix, US Patent 4,967,024 (1990).
12] J. Eicher, K. Fazniewscy, W. Rudolph, H.-W. Swidersky, US Patent
5
,214,223 (1993).
[
13] J. Eicher, K. Fazniewscy, M. Rieland, H.-W. Swidersky, US Patent
5,336,817 (1994).
The vacuum was broken with N and the reactor opened in a
2
glove box. 74.25 g (91%) of yellow-brown solid was isolated.
One hundred and forty five milligrams of solid was dissolved
in 500 mL 15% sodium acetate buffer solution and analyzed
[14] A. Feiring, US Patent 4,258,225 (1981).
[
15] G. Bauer (Ed.), Handbook of Preparative Inorganic Chemistry, vol. 1, 2nd
ed., Cornell Academic Press, 1963, p. 255, and references therein.
16] A. Feiring, J. Fluor. Chem. (1979) 7.
[
[
ꢁ
versus standard 100 ppm F with a fluoride selective electrode.
The analysis was 81 ppm. A check sample of 145 mg TaF had
an analysis of 100 ppm.
17] A. Henne, Org. React. 2 (1944) 49.
5
[18] It is not presumed that TaClF is necessarily a pure compound. The bulk
4
composition may be a mixture of chlorofluorides. For simplicity, this
discussion will use TaClF
19] Bauer [17] reports a 65% yield of TaF
spheric distillation. The low yield opens the possibility that TaF
produced from disproportionation of TaCl
4
to describe this possibly complex material
from HF and TaCl after atmo-
was
[
5
5
4
.1.6. Example 6. Attempted hydrofluorination of R-132b
with additional hydrogen chloride
5
x 5ꢁx
F
.
Five hundred and forty milligrams (0.002 mol) of TaF was
5
[20] G. Olah, S. Kuhns, J. Am. Chem. Soc. 80 (1958) 6541.
[21] D. Peters, R. Methchen, J. Fluor. Chem. 79 (1996) 161.
charged to the reactor. The reactor was evacuated and cooled