Addition of some unreactive fluoroalkanes to tetrafluoroethylene.: Direct catalytic synthesis of F-butene-2

Article abstract of DOI:10.1016/S0022-1139(99)00279-1

Condensation of trifluoromethanes, CF₃X (X=H, Cl, Br, I), with tetrafluorethylene to form the corresponding F-n-propyl adducts have been carried out with aluminum chlorofluoride as catalyst. Yields of C₃F₇I and C₃F₇Br are especially good, making these useful perfluoropropyl intermediates readily available. Details of the reactions, especially the presence of low percentages of perfluoroisopropyl iodide and bromide in the products, are accounted for by proposed mechanisms involving halonium intermediates. Longer-chain primary iodides can also be added to tetrafluoroethylene, but the final products are predominantly fluoroolefins, with pentafluoroethyl iodide as a byproduct. In the case of the addition of C₂F₅I to tetrafluoroethylene, conditions for an efficient, low temperature dimerization of terafluoroethylene to F-butene-2 catalyzed by a combination of C₂F₅I/aluminum chlorofluoride have been defined. Evidence for an unusual transfer of I⁺ from the iodonium derivative of F-butene-2 to tetrafluoroethylene is presented.

Full text of DOI:10.1016/S0022-1139(99)00279-1

Journal of Fluorine Chemistry 102 (2000) 199±204
Addition of some unreactive ¯uoroalkanes
to tetra¯uoroethylene.
Direct catalytic synthesis of F-butene-2
Viacheslav A. Petrov*, Carl G. Krespan
DuPont Central Research and Development, Experimental Station, P.O.Box 80328, Wilmington, DE 19880-0328, USA1
Received 24 June 1999; received in revised form 12 August 1999; accepted 8 September 1999
Dedicated to Professor Paul Tarrant on the occasion of his 85th birthday and in appreciation for
his substantial and varied contributions to ¯uorine chemistry
Abstract
Condensation of tri¯uoromethanes, CF3X (X ˆ H, Cl, Br, I), with tetra¯uorethylene to form the corresponding F-n-propyl adducts have
been carried out with aluminum chloro¯uoride as catalyst. Yields of C3F7I and C3F7Br are especially good, making these useful
per¯uoropropyl intermediates readily available. Details of the reactions, especially the presence of low percentages of per¯uoroisopropyl
iodide and bromide in the products, are accounted for by proposed mechanisms involving halonium intermediates. Longer-chain primary
iodides can also be added to tetra¯uoroethylene, but the ®nal products are predominantly ¯uoroole®ns, with penta¯uoroethyl iodide as a
byproduct. In the case of the addition of C2F5I to tetra¯uoroethylene, conditions for an ef®cient, low temperature dimerization of
tera¯uoroethylene to F-butene-2 catalyzed by a combination of C2F5I/aluminum chloro¯uoride have been de®ned. Evidence for an unusual
‡
transfer of I from the iodonium derivative of F-butene-2 to tetra¯uoroethylene is presented. # 2000 Elsevier Science S.A. All rights
reserved.
Keywords: Fluoroalkanes; F-butene-2
1. Introduction
Both aluminum chloride and the more recently discovered
aluminum chloro¯uoride (ACF) [9] are potent Lewis acid
catalysts with no oxidizing capability to cause undesirable
side-reactions. ACF, in particular, has such high reactivity
that it catalyzes condensation reactions of substrates resis-
tant to the action of SbF5 [1,2,10]. Sensitivity to protic
impurities is a characteristic of ACF which necessitates
particular care in excluding impurities such as water and
hydrocarbons and which limits the range of useful substrates
to those containing low amounts of hydrogen, i.e., those
leading to negligible HF elimination. Within these limits,
however, its range of usable substrates is broad, indeed
[1,2,11,12].
Additions of haloalkanes to ¯uoroole®ns using Friedel±
Crafts catalysts constitute an established and very useful
synthetic method [1±3]. A number of these reactions invol-
ving simple alkyl ¯uorides or chlorides are effectively
catalyzed by a solution of antimony penta¯uoride in anhy-
drous HF [4±6], since this mixture provides a superacid
system in which the pronounced oxidizing power of SbF5
has been obviated.
Fluoride substrates requiring a stronger catalyst, CH3CF3
[7], CH2F2 [8], and CH3F [8] being examples, are success-
fully added to ¯uoroole®ns such as tetra¯uoroethylene
(TFE) with neat SbF5 as catalyst. Antimony penta¯uor-
ide-catalyzed additions of chloro¯uoroalkanes are limited
to readily activated molecules such as ClCF2CFCl2 [3], since
they can be conducted at lower temperatures, while ¯uor-
oalkyl bromides and CF3I are not operable; higher iodides
perform poorly (vide infra).
2. Results and discussion
Tri¯uoromethyl halides and ¯uoroform have proved to be
resistant to condensation with ¯uoroole®ns via cationic
intermediates. Attempted additions to TFE of CF3H [8]
have been unsuccessful with SbF5 as catalyst. The case of
CF3H/SbF5 is instructive, since this combination does allow
* Corresponding author.
1 Publication No. 7956.
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 9 9 ) 0 0 2 7 9 - 1

200
V.A. Petrov, C.G. Krespan / Journal of Fluorine Chemistry 102 (2000) 199±204
condensation reactions at elevated temperatures if a sub-
strate more resistant to direct reaction with SbF5 than TFE is
provided [13]. Thus, alkylation of penta¯uorobenzene
occurs in good yield to give the main product shown in
Eq. (1).
(1)
Attempted additions of CF3Br to TFE in the presence of
SbF5 carried out in this work were unsuccessful, while CF3I
suffers oxidative replacement of iodine with ¯uorine even at
low temperature, in sharp contrast to higher per¯uoroalkyl
iodides, which are stable to the action of SbF5 at ambient
temperature.
In contrast to the ®ndings with SbF5, ACF does catalyze
additions of different tri¯uoromethanes to ¯uoroole®ns, as
is illustrated with TFE [11] Eqs. (2) and (3), con®rming its
increased range of catalytic activity over that of SbF5.
50ꢀC
XCF3 ‡ CF2 ˆˆCF2 ! XCF2CF2CF3;
Scheme 1.
ACF
1a; XˆˆH; 29%; 1b; XˆˆCl; 4%
(2)
50ꢀC
YCF3 ‡ CF2 ˆˆCF2 ! YCF2CF2CF3 ‡ ꢁCF3†2CFY;
presumably reversibly. The carbocationic moieties in such
complexes cannot isomerize and will disproportionate only
under extreme conditions, making them potentially useful
promoters for ACF-catalyzed reactions. These cationic
intermediates not only add to TFE, leading to product,
but apparently also are capable of inducing the isomeriza-
tions noted earlier with ACF and a promoter such as HFP
[16]. Low reaction temperatures limit the extent of isomer-
izations promoted by 2a0 and 2b0 via proposed intermediates
4 and 5 (Scheme 1).
ACF
2a; YˆˆBr; 63%; 3a; YˆˆBr; 4%
(3)
2b; YˆˆI; 80%;
3b; YˆˆI; 4%
The reaction provides a simple, direct route to poly¯uori-
nated propanes and is especially useful for the preparation
of F-n-propyl iodide (2b). This precursor to per¯uorinated
straight-chain derivatives having an odd number of carbon
atoms has generally been prepared by treatment of
C3F7C(O)OAg with I2 [14], an expensive route, or by
heating F-butyroyl chloride with KI at >2008C [15], which
requires equipment for operation either continuously or
under pressure.
The formation of detectable amounts of rearrangement
products 3 along with the normal addition products 2
suggests the occurrence of a secondary reaction similar
to that previously observed directly with F-n-propyl bro-
mide and iodide [16]. The isomerizations of F-n-propyl
bromide and iodide to the corresponding F-isopropyl
halides were either quite slow or did not occur with ACF
only, but proceeded with a co-catalyst such as hexa¯uor-
opropene (HFP), suggesting further that a carbocationic
species is required to promote rearrangement. In the present
case, reactive carbocationic complexes 2a0 and 2b0 between
tri¯uoromethanes and the insoluble ACF [17]2 are formed
As the low yield from CF3Cl attests, labilization of a-F
in this compound is much less pronounced than with
¯uorine substituents in the a-position to bromine or iodine.
Furthermore, ACF-catalyzed isomerization of ¯uorinated
vic-dihalides is markedly more dif®cult with dichlorides
than with dibromides, although evidence of anchimeric
assistance from bridging to adjacent halogen has been
adduced even for dichloride cases [16]. Our proposal of a
type of anchimeric assistance (i.e., in 4; Y ˆ Br) to account
for partial isomerization of BrCF2CF2CF3 induced by
‡
[BrCF2 ÁACF2 ] during addition of CF3Br to CF2=CF2
is in accord with the catalyzed isomerizations known to
occur with primary F-alkyl bromides, but not occur with the
corresponding chlorides [16].
The heightened reactivity of CF3I and CF3Br prompted a
study of ACF-catalyzed reactions of higher primary F-alkyl
iodides. In sharp contrast to n-C3F7Br (which does not
react with TFE in the presence of ACF catalyst) the reaction
of CF3CF2I with CF2=CF2 produced, surprisingly, F-
butene-2 (6) along with small amounts of iodide adduct
6a (Eq. (4)).
2 This paper provides evidence that the formation of polyfluoromethyl
‡
carbocations XCF2 (X ˆ H,F,Cl, Br) in solution is uphill thermodyna-
mically due to pronounced destabilizing effect of multiple fluorine
substituents in carbocations of this type.

V.A. Petrov, C.G. Krespan / Journal of Fluorine Chemistry 102 (2000) 199±204
201
easy to scale up and as carried out in a stirred autoclave, it
was used for the preparation of several pounds of F-butene-2
(see Section 3). Typical yields of F-butene-2 are in the range
60±75%. Detailed examination of a reaction mixture shows
that the main process (formation of 6) is accompanied by the
formation of a noticeable amount of isomeric F-hexenes,
surprisingly, with F-3-methylpentene-2 (10a and 10b,
trans- and cis-isomers, respectively) being the major com-
ponents of the mixture, containing also F-hexene-2 and
F-hexene-3 (11a, b and 12a, b, respectively; mixture of
trans- and cis-isomers) and secondary iodide 6a (Eq. (7)).
(4)
Interestingly, the reaction of C2F5I and TFE in the pre-
sence of SbF5 as a catalyst proceeds differently under
similar conditions. It is much slower and does not produce
butene (6); compound 6a is the major product formed in low
yield (10%) along with a small amount of telomeric per-
¯uoroalkyl iodides C2F5(CF2CF2)nI (n ˆ 1±4, calculated
yield 5%). Telomeric iodides C2F5(CF2CF2)nI probably,
result from the well-known radical telomerization reaction
between C2F5I and TFE initiated by the oxidizer (SbF5) in a
process similar to that recently described in [18] (Eq. (5)).
25 30ꢀC
2CF2 ˆˆCF2
!
6 ‡ 6a ‡ 10a; b ‡ 11a; b ‡ 12a; b
ACF=C2F5I
(7)
As was demonstrated in separate experiments, the for-
mation of hexenes 11 and 12 is the result of condensation
reactions of F-butene-2 and TFE. It is similar to ACF-
catalyzed condensation of ¯uoroole®ns such as F-pentene-2
and TFE [19], although the reaction of F-butene-2 is sig-
ni®cantly slower (Eq. (8)).
(5)
Formation of the 4-carbon products in both reactions
indicates that cation 7 is formed and that addition to TFE
occurs, but in the case of the more powerful Lewis acid
(ACF) the adduct 6a is capable of transferring the elements
of IF to another molecule of TFE under the reaction con-
ditions. We propose that the driving force leading to for-
mation of F-butene-2 is the reactive intermediate iodonium
ion 8 arising from the anchimerically assisted loss of
¯uoride depicted by structure 9 (Scheme 2).
As is predicted by this scheme, CF3CF2I is reformed by
net transfer of IF to TFE and can therefore serve as part of a
catalyst system (along with ACF) to dimerize TFE to
F-butene-2 (Eq. (6)).
(8)
We believe that the formation of branched ole®ns 10,
however, is the result of a competitive process involving
condensation of secondary iodide 6a with TFE. Transfer of
IF from intermediate 13 to TFE leads to C2F5I and ole®n 10
as a ®nal products (Scheme 3).
Predominant formation of ole®n 10 (see Section 3) is an
indication of the higher rate of the process represented by
Scheme 3 compared to the rate of the reaction 6 and TFE
(Eq. (8)).
Longer-chain primary alkyl iodides such as F-n-butyl
iodide can also be induced to add to TFE using ACF catalyst,
forming the corresponding ole®ns extended by two carbon
atoms (mixture of ole®ns 11 and 12). However, effective
transfer of IF from both starting iodide and product to TFE
produces major amounts of ole®n 6 and C2F5I (Eq. (9)).
(6)
Electrophilic dimerization of tetra¯uoroethylene proceeds
under mild condition and produces F-butene-2 in moderate
yield, providing easy excess to this material. The reaction is
(9)
Scheme 2.
Scheme 3.

202
V.A. Petrov, C.G. Krespan / Journal of Fluorine Chemistry 102 (2000) 199±204
Conversion of long-chain iodides to ole®ns can be accom-
plished more cleanly with HFP as the receptor for `IF', as
indicated by the example in Eq. (10).
3.1. Reaction of trifluoromethyl iodide with
tetrafluoroethylene (TFE)
A Hastelloy 400 ml shaker tube is ¯ushed with N2 and
loaded with 3 g of ACF inside a nitrogen bag. The vessel is
closed, cooled down, evacuated and loaded with 60 g
(0.31 mol) of CF3I and 15 g (0.15 mol) of TFE. It is kept
on a shaker at 25±308C for 16 h. The reactor is cooled to 08C
and unloaded. Excess of CF3I was distilled out of the crude
product using a low temperature distillation column afford-
ing 28 g of starting iodide. The residue was distilled giving
38 g of material with b.p. 35±408C. According to GC and
19F NMR this material contains 95% of CF3CF2CF2I and
5% of (CF3)2CFI. Total yield on both isomers is 84.4%.
(10)
Reaction of n-C6F13I and HFP proceeds in similar fash-
ion, producing a mixture of F-hexenes 11a, b and 12a, b and
(CF3)2CFI in high yield.
The reaction patterns of ¯uoroalkyl chlorides, bromides
and iodides found in this work and in our earlier study of
halide rearrangements [16] leads to some useful general
conclusions. Isomerizations of primary chlorides via chlor-
onium ions have substantial energy barriers and have only
been observed directly at high temperature, as in the special
case of CHFClCF2CF2Cl at 1608C [9]. Primary F-alkyl
bromides rearrange at more modest temperatures via bro-
monium ion intermediates, which can be generated by
reaction with various ¯uorocarbocations. No evidence has
been found so far for transferring Br from n-C3F7Br or n-
C4F9Br to HFP even at 1308C [16]. Primary F-alkyl iodides
isomerize more readily than the bromides [16]. In this case,
evidence has been found that transfer of I from an iodo-
nium cation to a terminal ¯uoroole®n (speci®cally, TFE or
HFP) also occurs readily.
3.2. Reaction of CF3Br with TFE
Using 45 g (0.3 mol) of CF3Br, 35 g (0.35 mol) of TFE,
1 g of ACF and 50 ml of dry cyclic dimer of hexa¯uoro-
propene (mixture of isomeric F-1,2-dimethyl- and F-1,3-
dimetylcyclobutanes) as a solvent, after 16 h at 25±308C,
50.1 g of material with b.p. 12±138C is isolated. According
to 19F NMR it is a mixture of 93% of CF3CF2CF2Br, 6% of
(CF3)2CFBr and 1% of solvent. Total yield of C3F7Br was
67%.
‡
‡
3.3. Reaction of CF3Cl with TFE
Vic-dihalides rearrange via halonium ions more readily
than monohalides, apparently due to the ease of removal of
F- from a carbon bearing the vic-halogen rather than a
second ¯uorine substituent. Vic-dichlorides can be isomer-
ized to gem-dichlorides with the catalyst combination ACF/
¯uoroole®n [16]. In the sterically favorable case of F-1,2-
dichloro-cyclopentane, ACF alone is suf®ciently active to
cause isomerization. Vic-dibromides in general isomerize
readily to gem-dibromides with ACF only as catalyst. We
have not examined the available, but relatively unstable
ICF2CF2I.
Using 42 g (0.4 mol) of CF3Cl, 30 g (0.3 mol) TFE, 2 g of
ACF and 50 ml of CDHFP, after 16 h at 508C followed by
low temperature distillation of crude product 10 g of a
fraction b.p. 108C 208C is isolated containing, according
to 19F NMR, 30% of CF3CF2CF2Cl and 70% of solvent.
Calculated yield of C3F7Cl is 4.2%. 19F NMR: 80.62 (3F,
t; 8.7 Hz), 125.30 (2F, s), 69.83 (2F, q; 8.7 Hz) ppm.
22 g of polytetra¯uoroethylene (PTFE) is also isolated from
this reaction.
3.4. Reaction of CF3H with TFE
Using 21 g (0.3 mol) of CF3H, 30 g (0.3 mol) TFE, 2 g of
ACF and 30 ml of CDHFP, after 16 h at 508C, 14.6 g of
material with b.p. 178C to 108C (main fraction 16 to
158C), identi®ed by 19F and 1H NMR as CF3CF2CF2H, is
isolated. The yield is 29.4%. 15 g of PTFE was also isolated
from this reaction.
3. Experimental section
1
19F and H NMR spectra were recorded on a QE-300
(General Electric, 200 MHz) or Brucker DRX-400 instru-
ments (400.5524 and 376.8485 MHz, respectively) using
CFCl3 as internal standard and d-chloroform or d-acetone as
a lock solvent. IR spectra were recorded on Perkin±Elmer
1600 FT spectrometer in a liquid ®lm. CF3H, CF3Cl, CF3Br,
CF3I (Aldrich), C2F5I, C4F9I (DuPont) are commercially
available and were used without further puri®cation. ACF is
prepared by literature method [9,19], using reaction of
AlCl3 and CFCl3, stored and handled inside a dry box.
Compounds 1a, b, 2a, b, 3a, b and 6 are identi®ed by
comparison with an authentic sample. Compounds 6a [20],
10a, b [21], 11a, b, 12a, b [22] by comparison of NMR data
to reported values.
3.5. Reaction of C2F5I and TFE catalyzed by SbF5
A mixture of 100 g (0.41 mol) of C2F5I, 5 g (0.023 mol)
of SbF5 and 40 g (0.4 mol) of TFE is kept in a shaker tube
for 18 h at 25±308C. The reactor is cooled to 08C, vented,
and the reaction mixture is fractionated to give 78.7 g of
C2F5I (b.p. 10±128C). The residue, 4.5 g of liquid, contained
based on GC and NMR data 15% of C2F5I, 80% 6a and 5%
of C2F5(CF2CF2)nI (n ˆ 1±4). Calculated yield of 6a is 3%.

V.A. Petrov, C.G. Krespan / Journal of Fluorine Chemistry 102 (2000) 199±204
203
3.6. Preparation of F-butene-2 (6)
3.9. Reaction of F-butene-2 with TFE
A 240 ml Hastelloy shaker tube is loaded with 5 g of
aluminum chloro¯uoride (ACF) inside a nitrogen bag,
cooled, evacuated and loaded with 40 g (0.16 mol) of
C2F5I and 40 g (0.4 mol) of TFE. The reaction vessel
was kept on a shaker at 25±308C. Signi®cant pressure drop
(about 150 psi) is observed in the ®rst 2 h. After 18 h at
ambient temperature the shaker tube is unloaded, and the
product (55 g) is collected in a cold ( 788C) trap. Accord-
ing 19F NMR the crude reaction mixture contains 41% of
C2F5I and 59% of 6 (mixture of trans and cis-isomers, ratio
3 : 2). The calculated yield of per¯uorobutene-2 is 82.5%.
Using 100 g (0.5 mol) of 6, 10 g of ACF and 50 g
(0.5 mol) of TFE, after 16 h at 508C the reaction mixture
is fractionated. 70 g of 6 is recovered. The residue (30 g)
according to GC and NMR data is a mixture of 25% 6,
14.5% 11a, 1.2% 11b, 53% 12a and 7.2% 12b. Conversion
of 6 is 22.5%, yield of ole®ns 11 and 12 67% on converted
6.
3.10. Reaction of n-C4F9I and hexafluoropropene
After 34.6 g (0.1 mol) of n-C4F9I, 15 g (0.1 mol) of
hexa¯uoropropene (HFP) and 3.5 g of ACF are reacted at
808C for 24 h, the reaction mixture is transferred into
another vessel containnig 3.5 g of ACF and it is charged
with 7.5 g (0.05 mol) of HFP. The vessel is kept at 808C
another for another 16 h. Crude product according to GC
and 19F NMR data is a mixture of 6 and 3b. Fractionation of
crude product gives 10 g of the material with b.p. 3±28C,
which is identi®ed by 19F NMR as 97% pure per¯uorobu-
tene-2 (trans : cisˆ81 : 19). The residue (33 g) according to
19F NMR contains 71% of (CF3)2 CFI, 13% of per¯uor-
obutene-2 and 16% of starting iodide. Calculated yield of
F-butene-2 is 84% and (CF3)2 CFI 80%.
3.7. Preparation of 6 in stirred autoclave
The reaction is scaled up in a 1 l Hastelloy stirred
autoclave which is ¯ushed with N2, loaded with 20 g
ACF and 100 g of C2F5I. TFE is added at 30±358C in 20
to 30 g increments. After addition of 400±600 g of TFE, the
clave is unloaded and a new portion (100 g) of C2F5I is
added and the cycle is repeated. Using 20 g of ACF, 200 g of
C2F5I and 980 g of TFE, 1150 g of crude product is isolated.
The distillation of crude product gives 650 g of azeoptropic
mixture containing 86% of 6 and 14% of C2F5I (b.p. 0±18C),
and 220 g of material having higher b.p. The yield of 6 is
57%.
The separation of CF3CF=CFCF3 and C2F5I is achieved
by distillation of the mixture from dry dimethylacetamide.
A ®rst distillation [350 g of mixture of CF3CF=CFCF3/
C2F5I (86 : 14), 200 ml (CH3)2NC(O)CH3)] affords 253 g
of CF3CF=CFCF3 containing 2% of C2F5I. Relatively pure
F-butene-2 [210 g of mixture trans- and cis-isomers in ratio
4 : 1, containing 0.2% of C2F5I (GC)] is isolated after a
second distillation using 300 g of 98 : 2 mixture of
CF3CF=CFCF3/C2F5I and 100 ml of (CH3)2NC(O)CH3).
220 g of the product with higher b.p. obtained in this
reaction is distilled giving 143 g of material with b.p. 50±
528C and 24 g of residue. According to 19F NMR of the
main fraction is a mixture containing: trans-CF3CF=
3.11. Reaction of n-C6F13I with HFP
4.5 g (0.01 mol) of n-C6F13I and 0.5 g of ACF is placed in
a heavy wall Pyrex sample tube equipped with a Te¯on
stopcock. The sample tube is evacuated at 1968C and 1.5 g
(10 mmol) of HFP is loaded through the vacuum line. The
sample tube is kept behind a shield at 758C for 10 h. After
that it is cooled to 788C, opened and the reaction mixture
is poured into water. The organic (lower) layer is separated,
dried over P2O5. to give 5.8 g of a mixture containing 49%
(CF3 )2 CFI, 48% of a mixture of F-hexene-2 and F-hexene-
3 (ratio 23 : 77) and a small amount (2±3%) of unidenti®ed
product (probably, C6F13CF(I)C2F5). Calculated yield of
F-hexenes is 93%, (CF3)2 CFI-95%.
C(CF3)C2F5 ± 38%, cis- 9.5%; trans-CF3CF=CFC3F7
7%, cis- 2%; trans-C2F5CF=CFC2F5 ± 29%, cis- 5%;
±
CF3CFICF2CF3 ± 9.5%.
4. Conclusions
ACF catalysis of additions of tri¯uoromethanes CF3X to
TFE has made ¯uoropropanes CF3CF2CF2X readily avail-
able. F-n-propyl iodide is a particularly useful intermediate
obtainable from this synthetic method. Reactivity patterns
of a series of halides in their condensation and isomerization
reactions have been clari®ed further.
Electrophilic dimerisation of TFE catalyzed by the C2F5I/
ACF catalytic system provides a novel, practical route to F-
butene-2, making this ¯uoroole®n readily available for
synthetic chemists.
3.8. Preparation of F-butene-2, using n-C4F9I
From 20 g (0.06 mol) of n-C4F9I, 50 g (0.5 mol) TFE and
5 g of ACF after 16 h at 25±308C. Crude reaction mixture
(65 g) contains 55% of 6, 14% C2F5I, 21% of 12a, b, 4% of
11a, b and 6% of C4F9I (starting material and 6a). After
fractionation 40 g of a fraction with b.p. 0 108C is isolated,
containing, according to 19F NMR, 90% of CF3 CF=CFCF3
(trans- and cis-isomers, ratio 3 : 2) and 10% of C2F5I. The
yield of 6 is 72% (based on TFE).

204
V.A. Petrov, C.G. Krespan / Journal of Fluorine Chemistry 102 (2000) 199±204
[10] V.A. Petrov, C.G. Krespan, B.E. Smart, J. Fluorine Chem. 77 (1996)
References
139.
[11] C.G. Krespan, V.A. Petrov (to DuPont) PCT US WO 97, 02227,
1997, CA 126 (1997) 171292w.
[1] C.G. Krespan, V.A. Petrov, Chem. Rev. 96 (1996) 3269.
[2] V.A. Petrov, V.V. Bardin, in: R.D. Chambers (Ed.), Topics in Current
Chemistry, Springer, Berlin, 1997, Vol. 192, pp. 39±95.
[3] G.G. Belen'kii, L.S. German, in: M.E. Volpin (Ed.), Soviet Scientific
Reviews/Sec. B. Chemistry Reviews, Harwood Academic Publishers
Gmb H, New York, 1984, Vol. 5, p. 183.
[12] C.G. Krespan, V.A. Petrov (to DuPont) PCT US WO 97, 02226,
1997, CA 126 (1997) 171293x.
[13] V.V. Brovko, V.A. Sokolenko, G.G. Yakobson, Zh. Org. Khim. 10
(1974) 300.
[14] J.H. Simons, T.G. Brice, in: J.H. Simons (Ed.), Fluorine Chemistry,
New York, 1954, Vol. 2, p. 367.
[4] G.G. Belen'kii, G.I. Savicheva, L.S. German Izv. Akad Nauk SSSR
Ser. Khim. (1978) 1433.
[15] C.G. Krespan, J. Org. Chem. 23 (1958) 2016.
[16] V.A. Petrov, C.G. Krespan, B.E. Smart, J. Fluorine Chem. 89 (1998)
125.
[5] V.A.Petrov, G.G. Belen'kii, L.S. German Izv. Akad Nauk SSSR Ser.
Khim. (1980) 1099.
[6] V.A.Petrov, G.G. Belen'kii, L.S. German Izv. Akad Nauk SSSR Ser.
Khim. (1980) 2821.
[17] V.A. Petrov, F. Davidson, J. Fluorine Chem. 95 (1999) 5.
[18] C.G. Krespan, V.A. Petrov, B.E. Smart, US Patent (to DuPont)
5,574,193 (1996) CA 124 (1996) 177222.
[7] V.A.Petrov, G.G. Belen'kii, L.S. German Izv. Akad Nauk SSSR Ser.
Khim. (1980) 2117.
[19] C.G. Krespan, D.A. Dixon, J. Fluorine Chem. 77 (1996) 117.
[20] C.G. Krespan, J. Org. Chem. 27 (1962) 1813.
[21] P.L. Coe, A. Sellars, J.C. Tatlow, J. Fluorine Chem. 23 (1983) 103.
[22] T.I. Filaykova, M.I. Kodess, N.V. Peschanskii, A.Ya. Zapevalov, I.P.
Kolenko, Zh. Org. Khim. 23 (1987) 1858.
[8] G.G. Belen'kii, V.A. Petrov, P.R. Resnick (to DuPont) PCT WO 98,
42645, 1998, CA 129 (1998) 246869.
[9] A.C. Sievert, F.G. Weigert, C.G. Krespan, US Patent 5157171,
1992