Free radical chemistry. Part 10.1 Addition of acyclic and cyclic alkanes to hexafluoropropene

Article abstract of DOI:10.1039/a909777e

γ-Ray or peroxide initiated reactions of acyclic and cyclic alkanes with hexafluoropropene are described. A variety of perfluorocarbons and polyfluorinated alkenes, di-enes and a tetra-ene were synthesised from these polyfluoroalkylated alkanes. Eliminati

Full text of DOI:10.1039/a909777e

View Article Online / Journal Homepage / Table of Contents for this issue
Free radical chemistry. Part 10.1 Addition of acyclic and cyclic
alkanes to hexafluoropropene
Richard D. Chambers,*a Robert W. Fuss,a Robert C. H. Spink,a Martin P. Greenhall,b
Alan M. Kenwright,a,c Andrei S. Batsanova,d and Judith A. K. Howarda,d
1
a Department of Chemistry, University of Durham, South Road, Durham, UK DH1 3LE
b F2 Chemicals, Lea Lane, Lea Town, Preston, Lancashire, UK PR4 0XJ
c NMR Service, University of Durham, South Road, Durham, UK DH1 3LE
d Crystallographic Unit, University of Durham, South Road, Durham, UK DH1 3LE
Received (in Cambridge, UK) 13th December 1999, Accepted 8th March 2000
Published on the Web 25th April 2000
γ-Ray or peroxide initiated reactions of acyclic and cyclic alkanes with hexafluoropropene are described. A variety
of perfluorocarbons and polyfluorinated alkenes, di-enes and a tetra-ene were synthesised from these polyfluoro-
alkylated alkanes. Elimination of hydrogen fluoride gives fluorinated alkenes, with unusual kinetic control of product
formation and further fluorinations gave novel perfluorocarbons. X-Ray crystallography established structures of
some isomers and NMR spectroscopy was used principally to establish the structures of others.
carried out at room temperature where greater selectivity can be
achieved. It is worth emphasising, with regard to propa-
Introduction
We have a continued interest in the carbon–hydrogen bond as a
gation, that radicals derived from hydrocarbons are relatively
functional group for reactions with unsaturated fluorine-
nucleophilic and fluorinated alkenes are, of course, extremely
containing compounds because, in principle, this is a poten-
electrophilic in character and, as a consequence of these polar
tially versatile route to a variety of fluorine containing
influences, we have very favourable circumstances for reac-
intermediates and here we apply this approach to hydrocarbons.
tivity. [Perhaps more appropriately, we should refer to radicals
with high SOMO energies, rather than nucleophilic radicals,
Many workers have directed considerable effort toward the gen-
eral process of functionalising hydrocarbons,2–4 with varying
reacting with alkenes of low LUMO energy, rather than elec-
success, but here we will demonstrate that formation of carbon–
carbon bonds, by reactions of hydrocarbons with fluorinated
alkenes, is an effective process, Scheme 1.
trophilic alkenes, etc.11 However, we have used ‘philicity’ ter-
minology, represented in Schemes by δϩ, δϪ etc., throughout,
merely for simplicity of presentation.] Indeed, the importance
of radical polarity in influencing reactivities is being increas-
ingly recognised12 and we have drawn attention to the
importance of such effects in earlier parts of this series.13
Dolbier has also drawn attention to the importance of polar
effects in reactions of fluorinated radicals with alkenes.14
Scheme 1
Premature chain termination can, of course, be a feature that
The area is still undeveloped, although some work has been
reported by Haszeldine and co-workers5–7 and by workers in the
former Soviet Union.8–10 Reactions that we will describe,
involve the familiar radical chain processes, Scheme 2, where
limits yields and trace amounts of impurities (sometimes not
easily detectable) may act as inhibitors. This makes separate
qualitative comparisons of reactivity based on yields very
capricious and our approach to circumventing this problem,
is to carry out competition reactions between various sub-
strates (see later).
We have used hexafluoropropene (HFP) especially because it
does not readily homopolymerise and therefore, the reactions
described in this paper provide a route to polyfluoro-n-propyl
derivatives and consequently to the further chemistry that can
be performed with these derivatives, which we will describe here
and in forthcoming publications.
Results and discussion
Starting with acyclic systems, various reactions are sum-
marised in Table 1; we were unable to effect reaction with
methane, even at 140 ЊC, but propane reacted readily even at
room temperature, showing a strong preference for the second-
ary site. However, it is interesting to note that n-butane gave the
di-adduct 10 as the principal product and that the second addi-
tion occurred exclusively at the primary site; we will return to
this point later. It is important to emphasise that using γ-rays
for initiation, the conversions noted in Table 1 are quite
modest but this is not significant because yields are essentially
Scheme 2
for initiation we have employed both peroxide and γ-ray
induced procedures and these give very similar results, for
reactions that are conducted at the same temperatures
although, for much of our work γ-ray induced reactions were
DOI: 10.1039/a909777e
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
This journal is © The Royal Society of Chemistry 2000
1623

View Article Online
Table 1 Addition of acyclic hydrocarbons to HFP
Major products (RFH = CF2CFHCF3) and conversions
(calculated based on alkane)
Alkane
Alkane–HFP
Initiator
CH4
1:1
DTBP (di-tert-butyl peroxide)
No reaction
1
1:1.2
1:1.5
γ-rays
19%
75%
1%
3%
DTBP
1:2
γ-rays
42%
80%
1:1.3
DTBP
3%
1:2
1:2
γ-rays
6%
11%
DTBP
quantitative and, of course, conversions may be increased with
time and/or temperature.
of HFP occurs, in most cases, although 1–3% of the opposite
isomers have been detected, depending on the system and the
reaction conditions.
Formation of the di-adduct 10 shown in Table 1, involves
preferential attack at the primary site for introduction of the
second group, and constitutes clear evidence for what is now a
well-established sequence in free-radical reactions.13–15 That
is, that radical 9a, Scheme 3, formed by addition of HFP to
The majority of the product from propane with HFP arises
from addition to the methylene carbon rather than to the
methyl sites. This is confirmed by the 1H NMR spectrum of the
product 3 which displayed three resonances in a 6:1:1 ratio
having chemical shifts appropriate to methyl, methine and CFH
hydrogens respectively. Similarly, addition of 2-methylpropane
5 to HFP proceeds mainly via the methine carbon to give the
major product 6, confirmed by the two resonances in its simple
1H NMR spectrum, in a ratio of 9:1 which have chemical shifts
consistent with methyl and CFH hydrogens respectively. The
mono-addition product 9 arises from addition of butane to
1
HFP via the methylene carbon, confirmed by the H NMR
spectrum of 9 which has six resonances. The two large reson-
ances at high field are characteristic of the two methyl groups
which have remained intact. The CFH and methine groups are
multiplets shifted to low field, as expected, and separate reson-
ances are observed for the two methylene hydrogens due to their
diastereotopic nature. The structure of the di-adduct 10 was
confirmed by its 13C NMR spectrum. There are four resonances
in the high field region, indicating that the second addition step
occurred via the terminal methyl, rather than a methylene
group, which would have led to a product with fewer reson-
1
ances than are observed. In comparison with the H NMR
spectrum of the mono-adduct 9, which has two methyl reson-
ances, the 1H NMR spectrum of 10 has only one methyl reson-
ance. Similar logic was used to deduce the structures of other
systems.
Scheme 3
n-butane, has the opportunity to abstract a hydrogen from the
chain in a ‘back-biting’ process, via a strain-free six membered
ring transition state; this is obviously in preference to a five
membered ring transition state and appears to be the basis for
preferential attack at the primary site in the intramolecular
step, leading to the formation of 10. It is useful to emphasise
that this intramolecular process obviously proceeds readily
whereas, in contrast, we shall note later that introduction of one
hexafluoro-n-propyl group deactivates the system to further
reaction with hexafluoropropene to form di-adducts by a
bi-molecular process.
In competition reactions we explored the effect of contain-
ment in a ring on the reactivity of C–H bonds and the results
are shown in Table 2. Relative reactivities are statistically cor-
rected for the number of sites available and we may conclude
that containment in a ring has no major effect on overall
reactivity. However, the intramolecular processes e.g. as illus-
trated in Scheme 3, that lead to multi-substitution products, are
probably absent in the cyclic systems.
Returning to preparative-scale reactions, products from
cyclohexane are shown in Table 3, where very high yields of
products are recovered and the ratio of mono- to di-addition
products depends on the ratio of HFP to cyclohexane, and on
temperature. Points to note are that a diastereomer of the 1,4-
di-adduct 20 crystallised from the di-adduct mixture on cooling
and a single crystal X-ray structure of this compound was
The structures of these and subsequent systems followed
principally from NMR spectra, although some examples of
single-crystal X-ray studies place our assignments on a firm
basis. Regioselective addition to the difluoromethylene group
1624
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
Table 2 Competition reactions between acyclic and cyclic hydrocarbons
Hydrocarbona
ratio after
reaction (R1)
Hydrocarbona
ratio prior to
reaction (R2)
Overall
relative
reactivities
Relative
reactivities
per CH2 site
Hydrocarbon
R1 Ϫ R2
47.07
44.60
48.42
38.50
38.81
10.5
8.57
1.48
1
0.37
0.2
5.79
37.92
36.86
1.03
0.26
48.56
11.7
1
0.17
a GLC peak areas.
Table 3 Addition of cyclic hydrocarbons to HFP
Conversion
of R–H (%)
Major products and yields (%)
(RFH = CF2CFHCF3)
R–H
R–H:HFP
Initiator
1:1.6
1:2
γ-rays
54
86
57
9
DTBP
100
23
1:1.5
1:2
γ-rays
79
99
90
39
4
DTBP
53
1:2
γ-rays
5
—
1:1.5
DTBP
79
39
51
27
1:2
1:2
γ-rays
5
—
DTBP
56
33
1:1.3
1:1.4
γ-rays
65
80
45
12
37
DTBP
100
1:1.2
1:7
DTBP
DTBP
92
60
59
19
36
100
determined (see later). Also the 1,3-:1,4- di-adduct ratio is ca.
2:1 and this statistical distribution implies that the deactivating
effect of the hexafluoropropyl group decreases rapidly with dis-
tance in this system, because no 1,2-di-adduct was detected.
Reactions with cis- and trans-decalins (decahydronaph-
thalene) gave mixtures containing isomeric mono- and di-
adducts in each case, but we were unable to separate these
mixtures and the spectra were too complex to form firm con-
clusions. However, elimination of hydrogen fluoride from the
mono-adduct mixture (see later) led to a considerable simplifi-
cation of the 13C NMR spectra, which enabled us to conclude
that negligible attack had occurred at the tertiary sites
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1625

View Article Online
Table 4 Competition reactions between cyclic hydrocarbons
Hydrocarbona
ratio prior to
reaction (R1)
Hydrocarbona
ratio after
reaction (R2)
Relative
reactivity per
C–H site
Overall
reactivities
Hydrocarbon
R1 Ϫ R2
47.82
52.18
53.41
43.69
43.12
49.75
4.13/18
1
1
9.06/18
3.66/18
2.19
2.19
(0.06) 1
1
46.58
45.13
54.76
34.46
36.94
44.54
12.12/12
8.19/10
3.31
1
(0.28) 4.7
(0.2) 1
10.22/12
1.24
(0.21) 1.05
a GLC peak areas.
Table 5 Competition reactions between hydrocarbons and ethers
(γ-rays, RT)
and, most obviously, this is a manifestation of steric effects.
Although we were unable to detect impurities in either sample
of cis- or trans-decalin, under the same conditions conversion
to products with HFP varied considerably (cis-:trans- = 39:79)
whereas in a competition reaction, there was little discernible
difference in reactivity (see Table 4). The notable conclusion
from the other results from competition experiments contained
in Table 4, is that there is comparatively little difference in
relative reactivity for these various ring systems.
Ratio
Ratio
aftera
reaction
(R2)
Relative
prior toa
reaction
(R1)
reactivity
per CH site
attacked
(R1 Ϫ R2)
58.92
44.38
14.54
1.21
Attack on norbornane (bicyclo[2.2.1]heptane) gave mainly
2-substitution, i.e. at the secondary site, 28, 29 Table 3, and
there was no evidence for attack at the tertiary site. Even the di-
adducts 29 contained no products arising from attack at a ter-
tiary site and the ratio of the 2,5-:2,6-isomers in 29 is ca. 2:1.
These results parallel other processes involving radical attack
on norbornane,16 but we are unaware of convincing arguments
advanced to account for the absence of radical attack at the
tertiary site, although an obvious point is the effect of strain.
Increased ring strain in the norbornyl system will lead to a
greater amount of s-character in the bridgehead carbon–
hydrogen bond, than in unstrained sp3 systems and this extra
s-character (less nucleophilic) could account for the reluctance
to react via a bridgehead position.
(12 Sites)
41.08
29.50
11.58
2.9
(4 sites)
No reaction
Reacted exclusively
a GLC peak areas.
In contrast to norbornane, adamantane reacts almost
exclusively at the tertiary sites, giving 31–34, and the product
distribution depends on conditions. The bridgehead position
in adamantane is unstrained and therefore is more favourable
to hydrogen abstraction than in the norbornyl system, indeed,
remarkably so and even a tetra-adduct is formed. Furthermore,
this tetra-adduct 34 was easily obtained pure because it crys-
tallised from the adduct mixture when dissolved in hexane or
chloroform.
We have compared reactivities of hydrocarbons with those
of corresponding ethers, Table 5, where adjacent oxygen is
known to encourage the formation of adjacent radical centres,
Scheme 4. Surprisingly, there is little overall difference, although
Scheme 4
attack occurs exclusively adjacent to oxygen in the ethers and
these sites are probably slightly activated, with respect to cyclo-
hexane. However, we can also conclude that the sites remote
from the oxygen in the ether are deactivated with respect to the
corresponding hydrocarbon. A comparison between tetrahy-
drofuran 37 and the mono-adduct 36 demonstrates clearly the
deactivating effect of the hexafluoro-n-propyl group on all of
the remaining sites in 36, including the position adjacent to
1626
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
oxygen. This is undoubtedly due to polar effects, arising from
the introduction of the electron withdrawing hexafluoro-n-
propyl group which makes it more difficult for an electrophilic
radical to abstract a hydrogen atom, from the more electrophilic
centre. Therefore both of the propagation steps k1, k2, Scheme
5, are inhibited after introduction of the first fluorinated group.
Fig. 1 Possible transition states for E1cB elimination.
Scheme 5
Eliminations of hydrogen fluoride
An obvious way of further functionalisation of these systems is
the elimination of hydrogen fluoride, giving new fluorinated
alkene derivatives, Scheme 6. These are, of course potential
Fig. 2 13C NMR chemical shifts (ppm) of pentafluoropropenylcyclo-
pentane systems 38 and 39.
Scheme 6
‘building blocks’ for incorporation of fluorine into more com-
plex molecules and after exploring a variety of approaches, we
adopted the procedure of adding the reactant to potassium tert-
butoxide in ether, while cooling. Yields of various products are
shown in Table 6 but, what is quite remarkable, is that the
Z-isomer, i.e. with the fluorine atoms syn- in the alkene, are
formed preferentially. This is especially surprising in the case of
the adamantane adduct where, at 0 ЊC, the Z-isomer is obtained
almost exclusively, while the E-isomer is obtained at higher
temperatures. Obviously, kinetic control is leading to the
Z-isomer and we have confirmed that the Z-isomer is converted
to the E-isomer by heating with caesium fluoride in tetraglyme
at 200 ЊC in a sealed tube, Scheme 7.
Scheme 8
ing in the carbanion-like transition-state. We can only probe this
situation by calculations, and these are proceeding in the hands
of other workers.17
Structure elucidation of cyclic systems
Elimination of hydrogen fluoride at low temperatures to give
only the Z-isomer in many cases, not only greatly simplifies the
product mixtures, by eliminating a chiral centre, but also allows
fuller characterisation of hexafluoropropyl adducts because of
their simplified NMR spectra.
Chemical shifts of Z-pentafluoropropenylcyclopentane 38
were used as a model for the structure elucidation of the analo-
gous di-enes 39a,b. Three components were found in the di-ene
mixture 39 and by comparing the broad band proton decoupled
13C NMR spectrum of the mixture with the mono-ene 38, two
of the components were assigned as the 1,3-substituted system
39a and the other as the 1,2-substituted cyclopentane 39b, Fig.
2. However, it was not possible to assign either of isomers
39a,b, as cis- and trans- with respect to the cyclopentane.
Z-Pentafluoropropenylcyclohexane 40 was investigated fully
due to its applicability as a model compound for many of the
other systems. The 19F and 1H NMR spectra of dehydrofluorin-
ated product 40 show antiperiplanar F-3–H-1 coupling, which
is quite large, giving a doublet ca. 32 Hz, Fig. 3. This coupling is
also evident in the cyclopentyl system 38, the norbornyl system
44 and the trans- 42 and cis-decalin 43a and 43b systems,
demonstrating that these systems are also formed by attack on
HFP via radicals derived from secondary sites in the corre-
sponding hydrocarbons. Conversely, the absence of analogous
coupling is diagnostic of products that are derived from attack
through tertiary sites in the hydrocarbon, as in the adamantyl
systems 46–48.
Scheme 7
The question then arises as to what the factors might be that
lead to the Z-isomer. It is probable that the mechanism of
elimination is either E1cB or E1cB-like and, indeed, we have
demonstrated that in reaction with 31, recovered starting
material contained ca. 9% deuterium, 31a, when the reaction
was carried out with tBuOD, after ca. 50% conversion to prod-
uct 46b, Scheme 8. anti-Elimination is most likely, and therefore
we need to account for why a transition-state that resembles (A)
is more favourable than one that resembles (B) (Fig. 1). Clearly,
for this situation to arise, there must be some energy-lowering
interaction between R and trifluoromethyl, e.g. hydrogen bond-
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1627

View Article Online
E
Table 6 Dehydrofluorination of cyclic hexafluoropropyl adducts (see Scheme 6, for RFZ, RF
)
R–RFH
Conditions
Yields (%)
KOtBu, Ϫ78 ЊC
75%
KOtBu, Ϫ78 ЊC
KOtBu, 0 ЊC
60%
92%
KOtBu, Ϫ10 ЊC
KOtBu, Ϫ10 ЊC
67%
30%
57%
KOtBu, Ϫ10 ЊC
KOtBu, Ϫ10 ЊC
82%
85%
KOtBu, Ϫ10 ЊC
KOtBu, Ϫ10 ЊC
90%
93%
KOtBu, RT
trace
85%
90%
KOtBu, Ϫ10 ЊC
trace
KOtBu, RT
78%
1
Further analysis of the vicinal coupling to H-1 in the H
NMR spectrum of Z-pentafluoropropenylcyclohexane 40
gives additional useful stereochemical information which can
be applied to other similar systems. If, as expected, the
fluoroalkenyl group occupies an equatorial position about
the cyclohexyl ring, H-1–H-2 antiperiplanar vicinal coupling
of ca. 12 Hz should be evident (see Fig. 3) which would not
occur if the substituent were axial. In fact, we observe this
coupling as a triplet (JHH 12 Hz) due to the two equivalent
H-2 hydrogens coupling to H-1 in this manner. We observe
similar coupling in other systems containing six-membered
rings such as the mono-ene derivatives of trans-, 42a, 42b
and cis-decalin 43a and 43b and the di-ene derivatives of
cyclohexane 41a, Table 6, confirming that the fluoroalkenyl
group also occupies equatorial positions in each of these
systems.
1628
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
Fig. 3 Antiperiplanar coupling in Z-pentafluoropropenylcyclohexane
40.
Fig. 5 Dihedral angles of exo- and endo-substituted norbornanes.
Fig. 4 13C NMR chemical shifts (ppm) of the sites β to fluoroalkyl-
ation in trans- and cis-decalins.
Fig. 6 13C NMR chemical shift assignments in the mono- and di-enes
of norbornane.
The di-enes derived from cyclohexane could be separated by
crystallisation. The crystalline trans-1,4-bis(pentafluoroprop-1-
enyl)cyclohexane 41a was identified by its simple NMR spectra,
due to its two-fold symmetry and a crystal structure confirmed
these findings. The structure of cis-1,3-bis(pentafluoroprop-
1-enyl)cyclohexane 41b was established via a broad band
proton decoupled 13C NMR experiment. Four resonances were
observed at high field which were attributable to the cyclo-
hexane ring and this confirmed the ring to be 1,3-substituted as
1,2- or 1,4-substitution would give three or two resonances in
the high field region, respectively.
firmed which hydrogens were associated with these positions.
Vicinal coupling (J ca. 6–8 Hz) between the tertiary hydrogen
(H-1) and the hydrogen (H-2) at the adjacent position, Fig. 5,
would be evident if the norbornyl system were endo-substituted,
due to their dihedral angle of 30Њ, but if the product were exo-
substituted no coupling would be observed, because of the
1
dihedral angle being ca. 90Њ. In fact, a H–1H COSY NMR
experiment determined that no coupling was evident between
H-1 and H-2, confirming the product 44, to be exo-substituted.
The NMR spectra of the norbornyl derivative 44 were used
as models to elucidate the structures of the di-enes 45. The 19
F
The regiochemistry of Z-pentafluoropropenyl derivatives
42a, 42b, of trans-decalin, was defined by the broad band pro-
ton decoupled 13C NMR and 13C DEPT NMR spectra of the
mixture; specifically, by analysis of the β-positions to the site of
fluoroalkylation, Fig. 4. The 13C DEPT NMR spectrum identi-
fied three methylene carbons, at 35.6, 29.5 and 28.5 ppm, as
doublets (3JCF ca. 3 Hz). The doublet at lowest field was
assigned to the 1-position of the 2-isomer 42b due to the
increased deshielding effect of the adjacent bridgehead posi-
tion, whereas the doublet at lowest field was assigned to the
3-position, due to its remoteness from the bridgehead site. The
remaining doublet was therefore assigned to the 2-position of
the 1-isomer 42a and the bridgehead carbon at lowest field, 42.5
ppm, was assigned to the other β-site, due to the deshielding
effect of the fluoroalkenyl group. A similar method was used to
identify the regiochemistry of the two major components of
(Z-pentafluoropropenyl)-cis-decalin 43a, 43b, Fig. 4.
It was confirmed that norbornane attacks HFP through a
secondary site, i.e. at the 7- or 2-position, due to the observation
of antiperiplanar vicinal coupling between its vinylic F-3 and a
methine H-2, akin to the cyclohexane derivative 40 discussed
earlier, and various NMR techniques were used to identify the
stereochemistry of the product 44 more precisely. Seven reson-
ances were observed in the high field region of the broad band
proton decoupled 13C NMR spectrum of the alkene 44, elimin-
ating the possibility of attack at the 7-position, due to the high
symmetry. To determine whether the norbornyl system was
endo- or exo-substituted at the 2-position, initially a 13C DEPT
NMR experiment was carried out, to identify the bridgehead
carbons, and then a 13C–1H HETCOR NMR experiment con-
NMR spectrum of the two component mixture revealed two
sets of syn-fluorine signals, in a ratio of 1.7:1, suggesting that
each di-ene contained equivalent Z-pentafluoropropenyl
groups and therefore addition had only occurred at the exo-
positions of the methylene groups in norbornane. The broad
band proton decoupled 13C NMR spectrum exhibited nine
resonances in the 0–50 ppm region and a 13C DEPT NMR
spectrum identified five of them as methine carbons. Of these,
two doublets at 38.0 (2JCF = 21 Hz) and 39.9 ppm (2JCF = 20 Hz)
were assigned as the carbons attached to the perfluoroalkenyl
groups and the remaining three as bridgehead carbons. A large
doublet (2JCF = 2.6 Hz) at 41.9 ppm was assigned as the bridge-
head carbon in the di-ene 45a, Fig. 6, because of its similar
chemical shift to the analogous bridgehead of the mono-ene
44. Singlets at 48.2 and 36.2 ppm were assigned as bridgehead
carbons of di-ene 45b. The lower field singlet was attributed
to the bridgehead carbon between the two perfluoroalkenyl
groups, because of the large deshielding from them and the
higher field singlet was assigned to the other bridgehead
carbon. The relative sizes of the 13C NMR resonances indicate
that the major component of the mixture was exo-2,5-bis[(1Z)-
pentafluoroprop-1-enyl]norbornane 45a.
Fluorination
Perfluorinated fluids are of increasing interest, for a variety of
uses that include the ‘fluoros-biphase’ approach of Horvath18
and co-workers and perfluorocarbons are made on the indus-
trial scale by the use of cobalt trifluoride,19 Scheme 9, because
reaction with elemental fluorine directly, is usually too vigorous
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1629

View Article Online
Table 7 Cobalt trifluoride fluorinations (RFH = CF2CFHCF3; RF =
CF2CF2CF3)
Temp./ЊC
Recovery
87%
Major product (yield)
375
60%
Fig. 7 X-Ray molecular structure of trans-1,4-bis(1,1,2,3,3,3-hexa-
fluoropropyl)cyclohexane 20a. Primed atoms are symmetrically related
via an inversion centre.
375
375
77%
86%
71%
63%
375
400
85%
77%
90%
65%
375
83%
58%
Fig. 8 X-Ray molecular structure of 1,3,5,7-tetrakis(1,1,2,3,3,3-hexa-
fluoropropyl)adamantane 34: one of the possible isomers (RSSS), the
disorder is not shown. Primed atoms are symmetrically related via a
mirror plane.
order. The molecule is located on a crystallographic mirror
plane, passing through the atoms C(1), C(2), C(3) and C(6), but
all four independent RFH groups in the structure are disordered
in different ways in the modes shown in Fig. 9. Thus all possible
isomers may share the same crystallographic site, and since the
space group is centrosymmetric (Pnma), the inversion equiv-
alent of each isomer is also present. (An attempted structure
solution in space group Pna21 revealed the same disorder, obvi-
ously conforming to the mirror symmetry). Crystals of 34 show
small but significant variations of unit cell parameters, e.g. at
room temperature, a = 11.513(2) and 11.486(2), b = 15.237(4)
and 15.217(2), c = 15.172(4) and 15.139(2) Å, respectively, for
two specimens from different crystallisations. The difference
may be due to uneven distribution of isomers.
Scheme 9
and therefore, difficult to control. We have earlier demonstrated
the benefit of introduction of a polyfluoroalkyl group because
this moderates reactivity towards cobalt trifluoride, so that con-
trolled reaction of both cobalt trifluoride and fluorine with
ethers20 and polyethers,1 to obtain perfluorinated systems is
possible. Here, we demonstrate a similar approach to synthesis
of perfluorocarbons showing that a range of new systems
may be obtained in this way. Examples of perfluorination are
contained in Table 7.
Molecule 41a (Fig. 10) possesses a crystallographic 2/m
symmetry, with the C(1), C(3), C(4), C(5), F(1), F(2), F(4) and
their equivalents lying in the mirror plane and the twofold axis
passing through the midpoints of the C(2)–C(2Љ) and C(2Ј)–
C(2ٞ) bonds. The cyclohexane ring adopts a chair conform-
X-Ray crystallographic studies
The crystal and molecular structures of trans-1,4-bis(1,1,2,3,3,3-
hexafluoropropyl)cyclohexane 20a, 1,3,5,7-tetrakis(1,1,2,3,3,3-
hexafluoropropyl)adamantane34,andtrans-1,4-bis[(1Z)-penta-
fluoroprop-1-enyl]cyclohexane 41a, were determined by single
crystal X-ray diffraction and confirmed unambiguously our
assignments, some of which would be controversial without this
complete evidence.
Molecule 20a (see Table 6) is situated at a crystallographic
inversion centre, Fig. 7, and is therefore a meso-isomer by its
two asymmetric atoms C(5) and C(5Ј). The cyclohexane ring
adopts a chair conformation with equatorial RFH substituents.
Molecule 34 (Fig. 8) contains four asymmetric carbon atoms
in the RFH side-chains and therefore can exist in five different
isomeric forms: SSSS and RRRR, RSSS and SRRR, and
RRSS. Because of the similar sizes of hydrogen and fluorine
atoms and conformational flexibility of the RFH group, mole-
cules (groups) with different configurations can occupy the
same crystallographic site, giving rise to a sophisticated dis-
Z
ation with equatorial orientation of the RF substituents. The
configuration around the C(3)᎐C(4) bond is cis. The CF group
᎐
3
is disordered between two orientations (differing by a 180Њ
rotation) with a 2:1 occupancy ratio. In either orientation, the
F(4) atom lies in the mirror plane, while two other fluorine
atoms, F(3) and F(3Ј), are related by this plane. However, high
displacement parameters of these atoms may indicate further
disorder.
Experimental
All starting materials were obtained commercially (Aldrich)
and all solvents were dried using literature procedures. NMR
spectra were recorded on either a Bruker AC 250, a Varian
Gemini 200, Varian VXR400S or a Bruker AMX 500 spec-
1630
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
carried out using either a Fischer Spahltroh MMS255 micro-
distillation apparatus or for higher boiling materials, a Buchi
kugelrohr GKR-51 apparatus. Boiling points were recorded
during the distillation or using the Siwoloboff method and
superscript numbers given as part of boiling point data indi-
cate the pressure (in mmHg) during measurement. Gas liquid
Chromatography (GC) analyses were carried out using a
Hewlett Packard 5890A gas liquid chromatograph equipped
with a 25 m cross-linked methyl silicone capillary column.
Preparative GC was performed on a Varian Aerograph Model
920 (catharometer detector) gas liquid chromatograph with
packed columns, which was mainly a 3 m 10% SE 30. Gamma
ray irradiations were performed in a purpose built shielded
chamber with a cobalt-60 source (500 C original activity).
Radical addition reactions
General procedure. The reactions were carried out using
sealed Pyrex tubes (for γ-ray initiation) or 150 ml and 250 ml
autoclaves (for peroxide initiation). Liquid reagents were added
to the tube/autoclave and degassed, then gaseous reagents were
transferred in vacuo, using normal vacuum line techniques. The
hydrocarbon substrate was usually used in a 1:1.5 deficiency to
HFP. The tube/autoclave was sealed in vacuo, while cold (liquid
air). For γ-initiation, the tube was irradiated at a fixed distance
for varying periods, to give a total dose in the range 7–12 Mrad
at room temperature. For thermal initiation using DTBP
(di-tert-butyl peroxide) (at 1–5% w/w concentration) the auto-
clave was placed in a thermostatically controlled furnace for
appropriate periods. The tube/autoclave was opened while the
contents were cold (liquid air) and gaseous components were
transferred in vacuo to a trap cooled in liquid air. Remaining
liquid was distilled to give excess hydrocarbon substrate and
products.
Methane (peroxide). Methane 1 (8.0 g, 0.5 mol), HFP (75.0 g,
0.5 mol) and DTBP (1.0 g, 6.8 mmol) were contained in a rock-
ing autoclave at 140 ЊC for 24 h. Gaseous products were
recovered to leave a small amount of liquid (ca. 3 ml), which on
analysis by GC-MS gave no indication of the existence of the
desired mono-adduct.
Fig. 9 Modes of disorder in RFH groups of 34: (a) the two positions
are related via a mirror plane; (b) the same, with each site shared by R
and S configurations (A and B positions of F(9) have occupancies of
0.15 and 0.35, respectively); (c) symmetrically unrelated positions A (S)
and B (R) have occupancies 0.7 and 0.3 respectively, with an additional
rotational disorder of the former (position C, occupancy 0.2).
Propane (ꢀ-rays). Propane 2 (2.3 g, 52 mmol) and HFP (9.0 g,
60 mmol) were irradiated for 8 days with γ-rays (12 Mrad) at
room temperature. Gaseous components (9.0 g) were recovered
to leave a colourless liquid, identified as an isomeric mixture
of 4-methyl-1,1,1,2,3,3-hexafluoropentane 3 (1.9 g, 19%) and
1,1,1,2,3,3-hexafluorohexane 4 (0.1 g, 1%) bp 81–82 ЊC (Found:
C, 37.4; H, 4.2. C6H8F6 requires C, 37.1; H, 4.1%); 4-methyl-
1,1,1,2,3,3-hexafluoropentane 3 δH (250 MHz; CDCl3; Me4Si)
1.12 (6 H, m, CH3), 2.34 (1 H, m, CHCF2), 4.82 (1 H, dm, 2JHF
44, CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ74.5 (3 F, m, CF3),
Ϫ117.7 and Ϫ121.8 (2 F, AB, JAB 265, CF2), Ϫ212.0 (1 F, d,
2JHF 27, CFH); m/z (EIϩ) 175 (0.5%), 93 (46), 65 (72), 43 (100),
41 (86); 1,1,1,2,3,3-hexafluorohexane 4 δH (250 MHz; CDCl3;
Me4Si) 1.00 (3 H, m, CH3), 1.58 (2 H, m, CH2), 1.90 (2 H, m,
CH2CF2), 4.82 (1 H, dm, 2JHF 44, CHF); δF (235 MHz; CDCl3;
CFCl3) Ϫ74.5 (3 F, m, CF3), Ϫ108.0 and Ϫ111.1 (2 F, AB, JAB
269, CF2), Ϫ211.0 (1 F, br s, CFH); m/z (EIϩ) 175 (2%), 93 (54),
69 (58), 65 (100), 43 (94).
Fig. 10 X-Ray molecular structure of trans-1,4-bis[(1Z)-pentafluoro-
prop-1-enyl]cyclohexane 41a. The CF3 group is disordered over two
positions with occupancies 2/3 (solid) and 1/3 (dashed). Primed atoms
are related to the reference ones by m plane, double-primed by axis 2,
triple-primed by the inversion centre.
trometer with tetramethylsilane and trichlorofluoromethane
as standards. Spectral assignments were made with the aid of
Propane (peroxide). Propane 2 (2.4 g, 55 mmol), HFP (12.7 g,
82 mmol) and DTBP (0.6 g, 4 mmol) were contained in a rock-
ing autoclave at 140 ЊC for 24 h. Gaseous products (6.3 g) were
recovered to leave a pale yellow liquid, which on fractional
distillation (81–82 ЊC) gave an isomeric mixture of 4-methyl-
1,1,1,2,3,3-hexafluoropentane 3 (7.9 g, 75%) and 1,1,1,2,3,3-
hexafluorohexane 4 (0.4 g, 3%).
data collected by 13C DEPT, H–13C HETCOR and H–1H
COSY experiments and coupling constants are given in Hz. In
19F NMR spectra, upfield shifts are quoted as negative. Mass
spectra were recorded on either a VG 7070E spectrometer or a
Fisons VG Trio 1000 spectrometer coupled with a Hewlett
Packard 5890 series II gas chromatograph. Elemental analyses
were obtained using either a Perkin-Elmer 240 or a Carlo
Erba Elemental Analyser. Melting points were recorded at
atmospheric pressure and are uncorrected. Distillations were
1
1
2-Methylpropane (ꢀ-rays). 2-Methylpropane 5 (5.8 g, 0.1 mol)
and HFP (30.0 g, 0.2 mol) were irradiated with γ-rays for 7 days
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1631

View Article Online
(10 Mrad) at room temperature. Gaseous components (26.5 g)
were recovered to leave a colourless liquid identified as 4,4-
dimethyl-1,1,1,2,3,3-hexafluoropentane 6 (8.8 g, 42%) bp 103–
104 ЊC (Found: C, 40.1; H, 4.5. C7H10F6 requires C, 40.4; H,
4.8%); δH (250 MHz; CDCl3; Me4Si) 1.13 (9 H, s, CH3), 4.91
(1 H, ddq, JHF 44, JHF 20, JHF 5.9, CFH); δF (235 MHz;
CDCl3; CFCl3) Ϫ74.6 (3 F, m, CF3), Ϫ117.6 and Ϫ126.1 (2 F,
AB, JAB 270, CF2), Ϫ206.9 (1 F, dm, 2JHF 40, CFH); m/z (EIϩ)
173 (5%), 65 (78), 57 (100), 41 (67).
Cyclopentane (ꢀ-rays). Cyclopentane 15 (7.1 g, 0.1 mol) and
HFP (23.6 g, 0.16 mol) were irradiated with γ-rays for 5 days
(7.5 Mrad) at room temperature. HFP (15.4 g) was recovered
and a colourless liquid obtained. Cyclopentane (3.2 g) was
removed by distillation, further fractional distillation gave
two fractions. The first fraction was identified as (1,1,2,3,3,3-
hexafluoropropyl)cyclopentane 16 (10.3 g, 86%) bp 134–135 ЊC
(Found: C, 43.5; H, 4.6. C8H10F6 requires C, 43.6; H, 4.6%);
δH (250 MHz; CDCl3; Me4Si) 1.63 (8 H, m, CH2), 2.52 (1 H, m,
2
3
3
2
3
3
CHCF2), 4.74 (1 H, ddq, JHF 44, JHF 21, JHF 6.2, CFH);
δF (235 MHz; CDCl3; CFCl3) Ϫ74.8 (3 F, m, CF3), Ϫ114.4 and
Ϫ118.8 (2 F, AB, JAB 266, CF2), Ϫ211.5 (1 F, d, 2JHF 38, CFH);
δC (100 MHz; CDCl3; Me4Si) 25.1 (br s, C-4), 25.6 (s, C-3), 25.8
(s, C-5), 26.1 (br s, C-2), 42.5 (t, 2JCF 22, C-1), 86.0 (ddqd, 1JCF
195, 2JCF 37, 2JCF 34, 2JCF 30, CFH), 120.2 (ddd, 1JCF 251, 1JCF
249, JCF 24, CF2), 120.9 (qd, JCF 282, JCF 26, CF3);
m/z (EIϩ) 199 (2%), 69 (100), 42 (78), 41 (83), 39 (47); and
the second fraction was identified as an isomeric mixture of
1,x-bis(1,1,2,3,3,3-hexafluoropropyl)cyclopentane (x = 2, 3) 17
bp15 80–81 ЊC (1.9 g, 9%) (Found: C, 36.0; H, 2.9. C11H10F12
requires C, 35.7; H, 2.7%); δH (250 MHz; CDCl3; Me4Si) 1.97
(6 H, m, CH2), 2.64 (2 H, m, CHCF2), 4.78 (2 H, m, CHF);
δF (235 MHz; CDCl3; CFCl3) Ϫ74.9 (6 F, m, CF3), Ϫ116.6 (4 F,
m, CF2), Ϫ211.2 (2 F, m, CFH); m/z (EIϩ) 331 (19%), 219 (36),
199 (100), 77 (96), 69 (70).
2-Methylpropane (peroxide). 2-Methylpropane 5 (2.5 g, 43
mmol), HFP (8.2 g, 55 mmol) and DTBP (0.6 g, 4 mmol)
were contained in a rocking autoclave at 140 ЊC for 24 h.
Gaseous products (3.4 g) were recovered to leave a pale yel-
low liquid which on fractional distillation gave an isomeric
mixture of 4,4-dimethyl-1,1,1,2,3,3-hexafluoropentane 6 (7.1
g, 80%) and 1,1,1,2,3,3-hexafluoro-5-methylhexane 7 (0.3 g,
3%) bp 103–104 ЊC (Found: C, 40.5; H, 4.9. C7H10F6 requires
C, 40.4; H, 4.8%); δH (250 MHz; CDCl3; Me4Si) 0.88 (6 H, d,
3JHH 4.5, CH3), 1.39 (1 H, br s, CH), 1.61 (2 H, m, CH2),
2
1
2
2
4.91 (1 H, dm, JHF 44, CFH); δF (235 MHz; CDCl3; CFCl3)
Ϫ75.5 (3 F, br s, CF3), Ϫ105.9 and Ϫ109.5 (2 F, AB, JAB 265,
CF2), Ϫ210.3 (1 F, br s, CFH); m/z (EIϩ) 173 (6%), 47 (33),
43 (100), 41 (25).
n-Butane (ꢀ-rays). n-Butane 8 (5.8 g, 0.1 mol) and HFP (30.0
g, 0.2 mol) were irradiated with γ-rays for 5 days (7.5 Mrad)
at room temperature. Gaseous components (29.8 g) were
recovered to leave a liquid (5.4 g) which was separated by pre-
parative scale GC into 4-methyl-1,1,1,2,3,3-hexafluorohexane 9
(6%, calculated by GC) (Found: C, 40.1; H, 4.4. C7H10F6
requires C, 40.4; H, 4.8%); δH (250 MHz; CDCl3; Me4Si) 0.97
(3 H, m, CH3), 1.12 (3 H, m, CH3), 1.30 (1 H, m, CH2), 1.73
(1 H, m, CH2), 2.06 (1 H, m, CHCF2), 4.84 (1 H, dm, 2JHF 44,
CFH); δF (235 MHz; CDCl3; CFCl3) Ϫ74.6 (3 F, m, CF3),
Ϫ116.4 and Ϫ119.3 (2 F, AB, JAB 266, CF2), Ϫ211.8 (1 F, d,
2JHF 44, CFH); m/z (EIϩ) 160 (4%), 65 (18), 59 (17), 57 (100);
and 4-methyl-1,1,1,2,3,3,7,7,8,9,9,9-dodecafluorononane 10
(11%, calculated by GC) (Found: C, 33.6; H, 2.5. C10H10F12
requires C, 33.5; H, 2.8%); δH (250 MHz; CDCl3; Me4Si) 1.08
(3 H, m, CH3), 1.5–2.0 (5 H, m, CH2 and CHCF2), 4.76 (2 H,
m, CFH); δC (100 MHz; CDCl3; Me4Si) 11.3 (br s, 4-Me),
22.7 (s, C-5), 32.9 (s, C-6), 36.7 (br s, C-4), 84.2 (m, CFH),
118.6 (m, CF2), 134.6 (m, CF3); δF (235 MHz; CDCl3; CFCl3)
Ϫ74.7 (3 F, m, CF3), Ϫ75.0 (3 F, m, CF3), Ϫ107.7 and
Ϫ111.9 (2 F, AB, JAB 269, CF2), Ϫ115.8 and Ϫ120.3 (2 F,
AB, JAB 271, CF2), Ϫ210.6 (1 F, br s, CFH), Ϫ211.7 (1 F, d,
2JHF 45, CFH); m/z (EIϩ) 319 (2%), 187 (57), 77 (53), 65 (98),
61 (100).
Cyclopentane (peroxide). Cyclopentane 15 (7.0 g, 0.1 mol),
HFP (29.5 g, 0.2 mol) and DTBP (0.6 g, 4 mmol) were con-
tained in a rocking autoclave at 140 ЊC for 24 h. HFP (15.0 g)
was recovered and a pale brown liquid obtained. Fractional
distillation gave two fractions. The first fraction was identified
as (1,1,2,3,3,3-hexafluoropropyl)cyclopentane 16 (12.5 g,
57%) and the second fraction was identified as 1,x-bis(1,1,2,3,
3,3-hexafluoropropyl)cyclopentane (x = 2, 3) 17 (8.3 g, 23%).
Cyclohexane (ꢀ-rays). Cyclohexane 18 (8.4 g, 0.1 mol) and
hexafluoropropene (HFP) (22.9 g, 0.15 mol) were irradiated
with γ-rays for 5 days (7.5 Mrad). HFP (10.9 g) was recovered
and a colourless liquid obtained. Cyclohexane (1.8 g) was
removed by distillation, further fractional distillation gave
two fractions. The first fraction was identified as (1,1,2,3,3,3-
hexafluoropropyl)cyclohexane 19 (16.6 g, 90%); bp 154–155 ЊC
(Found: C, 46.4; H, 5.5. C9H12F6 requires C, 46.2; H, 5.2%);
δH (250 MHz; CDCl3; Me4Si) 1.24 (5 H, m, CH2), 1.84 (6 H, m,
2
3
3
CH2 and CHCF2), 4.83 (1 H, ddqd, JHF 41, JHF 14, JHF 7.0,
3JHF 6.6, CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ74.8 (3 F, m,
CF3), Ϫ114.4 and Ϫ118.8 (2 F, AB, JAB 266, CF2), Ϫ212.3 (1 F,
d, 2JHF 41, CFH); δC (100 MHz; CDCl3; Me4Si) 24.0 (t, 3JCF 4.5,
C-6), 25.3 (s, C-4), 25.5 (s, C-5), 25.6 (s, C-2), 25.8 (s, C-3), 41.6
2
1
2
2
(t, JCF 21, C-1), 84.8 (ddq, JCF 195, JCF 37, JCF 34, CFH),
119.9 (ddd, 1JCF 252, 1JCF 248, 2JCF 24, CF2), 121.2 (qd, 1JCF 282,
2JCF 26, CF3); m/z (EIϩ) 234 (Mϩ, 1%), 195 (11), 83 (100), 55
(76); by comparison with data in the literature10 the second
fraction was identified as an isomeric mixture of 1,x-bis(1,1,2,
3,3,3-hexafluoropropyl)cyclohexane (x = 2–4) 20 (1.2 g, 4%)
bp15 105–106 ЊC (Found: C, 37.6; H, 3.3. C12H12F12 requires
C, 37.5; H, 3.1%); δH (250 MHz; CDCl3; Me4Si) 1.33 (4 H,
m, CH2), 1.69 (6 H, m, CH2), 2.09 (4 H, m, CH2), 4.84 (2 H,
n-Hexane (peroxide). n-Hexane 11 (8.6 g, 0.1 mol), HFP (30.0
g, 0.2 mol) and DTBP (0.6 g, 4 mmol) were contained in a
rocking autoclave at 140 ЊC for 24 h. Gaseous products (23.3 g)
were recovered to leave a liquid (6.5 g) which was separated by
preparative scale GC into isomers of (1,1,2,3,3,3-hexafluoro-
propyl)-n-hexane 12 (1% by GC), δH (250 MHz; CDCl3; Me4Si)
0.9–1.9 (13 H, m, CH3, CH2 and CHCF2), 4.47 (1 H, m, CFH);
δF (235 MHz; CDCl3; CFCl3) Ϫ74.7 (3 F, m, CF3), Ϫ111.3 (2 F,
m, CF2), Ϫ211.0 (1 F, br s, CFH); m/z (EIϩ) 187 (1%), 151 (3),
85 (80), 57 (100); bis(1,1,2,3,3,3-hexafluoropropyl)-n-hexane 13
(8% by GC), δH (250 MHz; CDCl3; Me4Si) 1.1–2.2 (12 H, m,
CH3, CH2 and CHCF2), 4.83 (2 H, m, CFH); δF (235 MHz;
CDCl3; CFCl3) Ϫ74.8 (3 F, m, CF3), Ϫ117.3 (2 F, m, CF2),
Ϫ211.0 (1 F, br s, CFH); m/z (EIϩ) 265 (4%), 235 (59), 195 (46),
155 (51), 61 (100); tris(1,1,2,3,3,3-hexafluoropropyl)-n-hexane
14 (5% by GC); δH (250 MHz; CDCl3; Me4Si) 1.2–2.2 (11 H, m,
CH3, CH2 and CHCF2), 4.90 (3 H, m, CFH); δF (235 MHz;
CDCl3; CFCl3) Ϫ75.2 (3 F, m, CF3), Ϫ113.9 (2 F, m, CF2),
Ϫ211.0 (1 F, br s, CFH); m/z (EIϩ) 477 (6%), 385 (40), 345 (64),
173 (68), 159 (72), 77 (90), 65 (100).
2
dm, JHF 41, CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ77.0 (6 F,
m, CF3), Ϫ120.9 (4 F, m, CF2), Ϫ213.8 (1 F, m, CFH),
Ϫ214.6 (1 F, m, CFH); m/z (EIϩ) 345 (12%), 233 (100), 213
(77), 77 (47).
Cyclohexane (peroxide). Cyclohexane 18 (8.3 g, 0.1 mol),
HFP (30.0 g, 0.2 mol) and DTBP (0.6 g, 4 mmol) were con-
tained in a rocking autoclave at 140 ЊC for 24 h. HFP (7.4 g) was
recovered and a yellow liquid obtained. Cyclohexane (0.1 g)
was removed by distillation, further fractional distillation of the
liquid gave two fractions. The first fraction was identified as
(1,1,2,3,3,3-hexafluoropropyl)cyclohexane 19 (9.0 g, 39%), and
1632
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
the second fraction was identified as mixture of isomers of 1,x-
bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane (x = 2–4) 20 (20.1
g,53%),fromwhich2R,2ЈS-trans-1,4-bis(1,1,2,3,3,3-hexafluoro-
propyl)cyclohexane 20a (3.4 g, 9%) crystallised out on standing
as a white solid; mp 80–81 ЊC (from MeOH) (Found: C, 37.3;
H, 3.0. C12H12F12 requires C, 37.5; H, 3.1%); δH (250 MHz;
CDCl3; Me4Si) 1.36 (4 H, m, CH2), 2.07 (6 H, m, CH2 ϩ CH),
4.83 (2 H, ddq, 2JHF 44, 3JHF 21, 3JHF 6.0, CFH); δF (235 MHz;
CDCl3; CFCl3) Ϫ74.2 (3 F, m, CF3), Ϫ117.1 and Ϫ118.4 (2 F,
AB, JAB 268, CF2), Ϫ211.2 (1 F, d, 2JHF 44, CFH); δC (100 MHz;
CDCl3; Me4Si) 22.8 (t, 3JCF 4.4, C-3), 24.1 (t, 3JCF 2.9, C-2), 40.6
(t, 2JCF 22, C-1), 85.0 (ddqd, 1JCF 196, 2JCF 38, 2JCF 34, 2JCF 31,
CFH), 119.3 (ddd, 1JCF 253, 1JCF 249, 2JCF 24, CF2), 121.2 (qd,
1JCF 282, 2JCF 26, CF3); m/z (EIϩ) 345 (12%), 233 (100), 213 (87).
decahydronaphthalene (x = 1, 2, 9) 25 (6.3 g, 56%) bp20 120–
121 ЊC (Found: C, 54.2; H, 6.1. C13H18F6 requires C, 54.2; H,
6.3%); δH (250 MHz; CDCl3; Me4Si) 1.31 to 1.70 (16 H, over-
lapping m, CH ϩ CH2), 2.08 to 2.20 (1 H, m, CHCF2), 4.80
(1 H, m, CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ74.5, Ϫ74.8
(3 F, br s, CF3), Ϫ111.4 to Ϫ118.5 (2 F, overlapping m, CF2),
Ϫ206.4, Ϫ212.1 (1 F, m, CFH); m/z (EIϩ) 288 (Mϩ, 64%), 246
(64), 137 (47), 95 (100); and the second fraction was identified
as an isomeric mixture of x,y-bis(1,1,2,3,3,3-hexafluoropropyl)-
decalin (x = 1, y = 2–10; x = 2, y = 3–10) 26 (5.6 g, 33%), bp20
152–153 ЊC (Found: C, 43.6; H, 4.0. C16H18F12 requires C,
43.8; H, 4.1%); δH (250 MHz; CDCl3; Me4Si) 1.39 to 2.05 (14 H,
overlapping m, CH ϩ CH2), 2.23 to 2.65 (2 H, overlapping m,
CHCF2), 4.80, 4.98 (2 H, overlapping m, CHF); δF (235 MHz;
CDCl3; CFCl3) Ϫ74.2, Ϫ74.6 (6 F, br s, CF3), Ϫ111.4 to Ϫ119.4
(4 F, overlapping m, CF2), Ϫ204.1, Ϫ206.2, Ϫ212.1 (2 F, m,
CFH); m/z (EIϩ) 400 (1%), 287 (72), 245 (100), 231 (88).
trans-Decahydronaphthalene (ꢀ-rays). trans-Decahydro-
naphthalene 21 (6.9 g, 0.05 mol), HFP (15.5 g, 0.1 mol) and dry
acetone (8 ml) were irradiated with γ-rays at room temperature
for 5 days (7.5 Mrad). Nearly all of the HFP (14.7 g) was
recovered. The liquid product mixture (7.5 g) was analysed by
GC/MS and 19F NMR which identified traces of x-(1,1,2,3,3,3-
hexafluoropropyl)-trans-decahydronaphthalene (x = 1, 2) 22 (ca.
5% by GC) and no further workup was performed.
Bicyclo[2.2.1]heptane (ꢀ-rays). Bicyclo[2.2.1]heptane 27 (7.2
g, 75 mmol), HFP (15.7 g, 0.1 mol) and dry acetone (8 ml) were
irradiated with γ-rays at room temperature for 5 days (7.5
Mrad). HFP (5.5 g) was recovered and a colourless liquid
obtained. Sublimation gave recovered bicyclo[2.2.1]heptane
(2.5 g), and subsequent fractional distillation under reduced
pressure gave two fractions. The first fraction was identified
as two diastereomers of exo-2-(1,1,2,3,3,3-hexafluoropropyl)-
bicyclo[2.2.1]heptane 28 (9.6 g, 80%) bp20 80 ЊC (Found: C,
48.8; H, 4.8. C10H12F6 requires C, 48.8; H, 4.9%); δH (250 MHz;
CDCl3; Me4Si) 1.26 (3 H, overlapping m, CH2), 1.61 (4 H, over-
lapping m, CH2), 1.75 (1 H, m, CH2), 2.14 (1 H, m, CHCF2),
2.37 (1 H, br s, CH), 2.45 (1 H, br s, CH), 4.81 (1 H, m, CHF);
δF (235 MHz; CDCl3; CFCl3) Ϫ74.7 (3 F, br s, CF3), Ϫ113.0
trans-Decahydronaphthalene (peroxide). trans-Decahydro-
naphthalene 21 (13.9 g, 0.1 mol), HFP (23.1 g, 0.15 mol), DTBP
(0.8 g, 5.5 mmol) and dry acetone (2 ml) were contained in a
rocking autoclave at 140 ЊC for 24 hours. HFP (4.1 g) was
recovered and a yellow liquid (33.3 g) obtained. Acetone and
trans-decahydronaphthalene (2.9 g) were removed by distil-
lation, further fractional distillation of the liquid under reduced
pressure gave two fractions. The first fraction was identified as
an isomeric mixture of x-(1,1,2,3,3,3-hexafluoropropyl)-trans-
decahydronaphthalene (x = 1, 2) 22 (11.5 g, 51%), bp5 80–82 ЊC
(Found: C, 54.2; H, 6.1. C13H18F6 requires C, 54.2; H, 6.3%);
δH (250 MHz; CDCl3; Me4Si) 1.00 to 1.69 (16 H, overlapping m,
CH ϩ CH2), 2.10 (1 H, overlapping m, CHCF2), 4.82 (1 H, m,
CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ74.0, Ϫ74.5 (3 F, br s,
CF3), Ϫ110.2, Ϫ118.6 (2 F, overlapping m, CF2), Ϫ211.1, 211.9
(1 F, m, CFH); m/z (EIϩ) 288 (Mϩ, 57%), 246 (43), 137 (37), 95
(100), 81 (81), 67 (75), 41 (80); and the second fraction was
identified as an isomeric mixture of x,y-bis(1,1,2,3,3,3-hexa-
fluoropropyl)-trans-decahydronaphthalene (x = 1, y = 2–10;
x = 2, y = 3–10) 23 (9.5 g, 27%), bp5 130–135 ЊC (Found: C,
43.6; H, 4.0. C16H18F12 requires C, 43.8; H, 4.1%); δH (250 MHz;
CDCl3; Me4Si) 1.00 to 1.70 (14 H, overlapping m, CH ϩ CH2),
2.10 (2 H, overlapping m, CHCF2), 4.83 (2 H, m, CHF); δF (235
MHz; CDCl3; CFCl3) Ϫ73.8, Ϫ74.2 (6 F, br s, CF3), Ϫ118.0
(4 F, overlapping m, CF2), Ϫ211.1 (2 F, overlapping m, CFH);
m/z (EIϩ) 438 (Mϩ, 7%), 287 (60), 246 (89), 231 (84), 81 (100).
2
and Ϫ118.8 (2 F, AB, JAB 265, CF2), Ϫ211.2 (1 F, d, JHF 44,
CFH); δC (100 MHz; CDCl3; Me4Si) 28.0 (br s, C-6), 30.4 (s,
C-5), 31.4 (s, C-3), 35.8 (s, C-4), 36.9 (br s, C-7), 37.6 (d, 4JCF 5,
2
1
2
2
C-1), 44.9 (t, JCF 22, C-2), 86.2 (ddqd, JCF 196, JCF 37, JCF
34, 2JCF 31, CFH), 120.0 (ddd, 1JCF 251, 1JCF 249, 2JCF 22, CF2),
121.3 (qd, JCF 282, JCF 29, CF3); m/z (EIϩ) 246 (Mϩ, 5%), 95
(57), 68 (87), 67 (100); and the second fraction was identified as
an isomeric mixture of 2,x-bis(1,1,2,3,3,3-hexafluoropropyl)-
bicyclo[2.2.1]heptane (x = 5, 6) 29 (2.3 g, 12%) bp20 120 ЊC
(Found: C, 39.4; H, 3.0. C13H12F12 requires C, 39.4; H, 3.1%);
δH (250 MHz; CDCl3; Me4Si) 1.00 (3 H, overlapping m, CH2),
1.50 (5 H, overlapping m, CH2), 2.15 (1 H, m, CHCF2), 2.32,
2.39 (1 H, br s, CH), 2.54, 2.80 (1 H, br s, CH), 4.75 (1 H, m,
CHF); δF (235 MHz; CDCl3; CFCl3) Ϫ74.7 (6 F, overlapping
m, CF3), Ϫ114.1 (4 F, overlapping m, CF3), Ϫ210.8 (2 F, over-
lapping m, CFH); m/z (EIϩ) 245 (9%), 77 (22), 67 (100), 41 (14).
1
2
Bicyclo[2.2.1]heptane (peroxide). Bicyclo[2.2.1]heptane 27
(6.1 g, 64 mmol), HFP (14.3 g, 90 mmol) and DTBP (0.6 g, 4
mmol) were contained in a rocking autoclave for 24 hours at
140 ЊC. No HFP was recovered and a pale yellow liquid (19.5 g)
was obtained. Fractional distillation of the liquid gave two frac-
tions, the first consisted of exo-2-(1,1,2,3,3,3-hexafluoropropyl)-
bicyclo[2.2.1]heptane 28 (7.1 g, 45%) and the second consisted
of 2,x-bis(1,1,2,3,3,3-hexafluoropropyl)bicyclo[2.2.1]heptane
(x = 5, 6) 29 (9.5 g, 37%).
cis-Decahydronaphthalene (ꢀ-rays). cis-Decahydronaph-
thalene 21 (6.9 g, 0.05 mol), HFP (15.3 g, 0.1 mol) and dry
acetone (8 ml) were irradiated with γ-rays at room temperature
for 5 days (7.5 Mrad). Nearly all of the HFP (14.5 g) was
recovered. The colourless liquid product mixture (7.6 g) was
analysed by GC/MS and 19F NMR which identified traces
of x-(1,1,2,3,3,3-hexafluoropropyl)-cis-decahydronaphthalene
(x = 1, 2, 9) 22 (ca. 5% by GC), but no further workup was
performed.
Adamantane (peroxide). Adamantane 30 (13.8 g, 0.1 mol),
HFP (17.6 g, 0.12 mol) and DTBP (0.7 g, 5 mmol) were con-
tained in a rocking autoclave for 24 hours at 140 ЊC. No HFP
was recovered and adamantane (1.1 g) crystallised out of the
liquid product (28.0 g). Fractional distillation under reduced
pressure gave two fractions, the first fraction was identified as
1-(1,1,2,3,3,3-hexafluoropropyl)adamantane 31 (16.3 g, 60%),
bp9 99–101 ЊC (Found: C, 54.4; H, 5.5. C13H16F6 requires C,
54.5; H, 5.6%); δH (250 MHz; CDCl3; Me4Si) 1.76 (12 H, m,
cis-Decahydronaphthalene (peroxide). cis-Decahydronaph-
thalene 24 (13.8 g, 0.1 mol), HFP (29.1 g, 0.2 mol), DTBP (0.6
g, 4 mmol) and dry acetone (2 ml) were contained in a rocking
autoclave at 140 ЊC for 24 hours. HFP (19.8 g) was recovered
and a yellow liquid (23.3 g) obtained. Acetone and cis-
decahydronaphthalene (8.4 g) were removed by distillation,
further fractional distillation of the liquid under reduced pres-
sure gave two fractions. The first fraction was identified as
an isomeric mixture of x-(1,1,2,3,3,3-hexafluoropropyl)-cis-
2
3
3
CH2), 2.07 (3 H, s, CH), 4.93 (1 H, ddq, JHF 44, JHF 20, JHF
6.4, CFH); δF (235 MHz; CDCl3; CFCl3) Ϫ74.3 (3 F, br s, CF3),
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1633

View Article Online
Ϫ122.6 and Ϫ130.0 (2 F, AB, JAB 274, CF2), Ϫ206.9 (1 F, d,
2JHF 41, CFH); δC (100 MHz; CDCl3; Me4Si) 27.5 (s, CH), 34.6
(q, JCF 3.4, CH2), 36.4 (s, CH2), 40.0 (t, JCF 21, CCF2), 83.6
decreased from 48.56% to 11.7% and the peak integration of
n-hexane decreased from 48.42% to 10.5%.
5
2
1
2
2
2
(ddqd, JCF 197, JCF 41, JCF 33, JCF 26, CFH), 119.5 (ddd,
Competition between cyclopentane and n-hexane. A Carius
tube was charged with cyclopentane 15 (7.0 g, 0.1 mol),
n-hexane 11 (8.6 g, 50 mmol) and HFP (4.6 g, 30 mmol) and
then irradiated with γ-rays for 10 days. The Carius tube was
opened and any remaining HFP recovered. The GC traces
from before and after the reaction showed the peak integra-
tion of cyclopentane decreased from 44.60% to 38.81% and
the peak integration of n-hexane decreased from 47.07% to
38.50%.
1JCF 261, JCF 247, JCF 22, CF2), 121.3 (qd, JCF 283, JCF 26,
CF3); m/z (EIϩ) 151 (1%), 135 (100), 79 (43), 41 (35); MS and
NMR data are consistent with those contained in the liter-
ature;8 and the second fraction was identified as 1,3-bis-
(1,1,2,3,3,3-hexafluoropropyl)adamantane 32 (7.9 g, 19%), bp9
124–126 ЊC (Found: C, 43.7; H, 3.8. C16H16F12 requires C, 44.0;
H, 3.7%); δH (250 MHz; CDCl3; Me4Si) 1.80 (12 H, m,
CH ϩ CH2), 2.28 (2 H, br s, CH2), 4.93 (2 H, ddq, 2JHF 44, 3JHF
20, 3JHF 6.4, CFH); δF (235 MHz; CDCl3; CFCl3) Ϫ74.3 (6 F, br
s, CF3), Ϫ121.7 and Ϫ129.4 (4 F, AB, JAB 275, CF2), Ϫ207.1
1
2
1
2
Competition between cis- and trans-decahydronaphthalene. An
autoclave (150 ml) was charged with cis-decahydronaphthalene
24 (6.9 g, 50 mmol), trans-decahydronaphthalene 21 (6.9 g, 50
mmol), HFP (2.3 g, 15 mmol) and DTBP (0.3 g, 2 mmol) and
then rocked at 140 ЊC for 24 h. The autoclave was opened
and HFP (0.1 g) was recovered. The GC traces from before
and after the reaction showed the peak integration of cis-
decahydronaphthalene decreased from 52.18% to 43.12% and
the peak integration of trans-decahydronaphthalene decreased
from 47.82% to 43.69%.
2
(2 F, d, JHF 37, CFH); δC (100 MHz; CDCl3; Me4Si) 27.0
(s, CH), 31.8 (br s, CH2), 33.8 (s, CH2), 35.3 (s, CH2), 40.6
(t, 2JCF 21, CCF2), 83.8 (ddqd, 1JCF 197, 2JCF 42, 2JCF 34, 2JCF 26,
CFH), 119.1 (ddd, 1JCF 261, 1JCF 247, 2JCF 23, CF2), 121.1 (qd,
1JCF 283, 2JCF 26, CF3); m/z (EIϩ) 397 (1%), 285 (100), 243 (11),
229 (12), 55 (16); MS and NMR data are consistent with those
contained in the literature.21
Adamantane (peroxide). Adamantane 30 (2.7 g, 20 mmol),
HFP (20.8 g, 140 mmol) and DTBP (0.5 g, 4 mmol) were con-
tained in a rocking autoclave for 24 hours at 140 ЊC. HFP (8.9
g) was recovered and a waxy liquid (12.7 g) obtained. Kugelrohr
distillation (175 ЊC, 1 mmHg) gave a waxy mixture which was
then mixed with chloroform, at which point a white solid pre-
cipitated out. Filtration of the solid, followed by removal of the
solvent gave a liquid identified as 1,3,5-tris(1,1,2,3,3,3-hexa-
fluoropropyl)adamantane 33 (6.9 g, 59%), bp9 143–145 ЊC
(Found: C, 39.1; H, 2.5. C19H16F18 requires C, 38.9; H, 2.7%);
δH (250 MHz; CDCl3; Me4Si) 1.83 (12 H, m, CH2), 2.15 (1 H, br
s, CH), 4.95 (2 H, ddq, 2JHF 41, 3JHF 19, 3JHF 5.3, CFH); δF (235
MHz; CDCl3; CFCl3) Ϫ74.2 (9 F, br s, CF3), Ϫ120.7 and
Ϫ128.7 (6 F, AB, JAB 274, CF2), Ϫ207.3 (3 F, d, 2JHF 29, CFH);
δC (100 MHz; CDCl3; Me4Si) 26.7 (s, CH), 31.5 (s, CH2), 33.2 (s,
CH2), 41.3 (t, 2JCF 22, CCF2), 84.1 (ddqd, 1JCF 197, 2JCF 42, 2JCF
34, 2JCF 26, CFH), 119.1 (ddd, 1JCF 262, 1JCF 248, 2JCF 26, CF2),
121.0 (qd, 1JCF 283, 2JCF 26, CF3); m/z (EIϩ) 547 (5%), 435 (100),
277 (15), 243 (12), 69 (34); MS and NMR data are consistent
with those contained in the literature;18 and the white solid
was identified as 1,3,5,7-tetrakis(1,1,2,3,3,3-hexafluoropropyl)-
adamantane 34 (5.3 g, 36%) mp 110–112 ЊC (from MeOH)
(Found: C, 36.0; H, 2.2. C22H16F24 requires C, 35.9; H, 2.2%);
δH (250 MHz; CDCl3; Me4Si) 2.07 (12 H, m, CH2), 5.99 (4 H,
Competition between cyclohexane and trans-decahydronaph-
thalene. An autoclave (150 ml) was charged with cyclohexane 18
(4.2 g, 50 mmol), trans-decahydronaphthalene 21 (6.9 g, 50
mmol), HFP (2.4 g, 16 mmol) and DTBP (0.3 g, 2 mmol) and
then rocked at 140 ЊC for 24 h. No HFP was recovered and the
GC traces showed the peak integration of cyclohexane
decreased from 53.41% to 49.75% and the peak integration of
trans-decahydronaphthalene decreased from 46.58% to 34.46%.
Competition between cyclohexane and cyclopentane. An auto-
clave (150 ml) was charged with cyclohexane 18 (4.2 g, 50
mmol), cyclopentane 15 (3.5 g, 50 mmol), HFP (2.4 g, 16 mmol)
and DTBP (0.3 g, 2 mmol) and then rocked at 140 ЊC for 24 h.
No HFP was recovered. The GC traces showed the peak
integration of cyclohexane decreased from 54.76% to 44.54%
and the peak integration of cyclopentane decreased from
45.13% to 36.94%.
Competition between cyclohexane and tetrahydropyran. A
Carius tube was charged with cyclohexane 18 (8.4 g, 0.1 mol),
tetrahydropyran 35 (8.6 g, 50 mmol) and HFP (4.6 g, 30 mmol)
and then irradiated with γ-rays for 10 days. The Carius tube was
opened and any remaining HFP recovered. The GC traces from
before and after the reaction showed the peak integration of
cyclohexane decreased from 58.92% to 44.38% and the peak
integration of tetrahydropyran decreased from 41.08% to
29.50%.
2
3
3
ddq, JHF 42, JHF 20, JHF 6.4, CFH); δF (235 MHz; CDCl3;
CFCl3) Ϫ74.1 (12 F, br s, CF3), Ϫ121.1 and Ϫ127.6 (8 F, AB,
2
JAB 276, CF2), Ϫ207.1 (4 F, d, JHF 29, CFH); δC (100 MHz;
CDCl3; Me4Si) 30.1 (s, CH2), 41.6 (t, 2JCF 22, CCF2), 83.1 (ddq,
1JCF 194, 2JCF 39, 2JCF 34, CFH), 119.0 (ddd, 1JCF 261, 1JCF 249,
2JCF 22, CF2), 121.4 (qd, 1JCF 282, 2JCF 26, CF3); m/z (EIϩ) 717
(1%), 697 (4), 585 (100), 277 (33), 151 (55), 55 (65).
Competition between (1,1,2,3,3,3-hexafluoropropyl)tetra-
hydrofuran and tetrahydrofuran. A Carius tube was charged
with (1,1,2,3,3,3-hexafluoropropyl)tetrahydrofuran 36 (11.1 g,
50 mmol), tetrahydrofuran 37 (3.6 g, 50 mmol) and HFP (2.3 g,
15 mmol) and then irradiated with γ-rays for 10 days. The
Carius tube was opened and any remaining HFP recovered.
The GC traces from before and after the reaction showed the
peak integration of (1,1,2,3,3,3-hexafluoropropyl)tetrahydro-
furan increased from 67.51% to 76.51% whilst the peak
integration of tetrahydropyran decreased from 31.49% to
22.56%.
General procedure for competition reactions
Competition reactions were performed using either DTBP or
γ-ray initiation using the usual experimental procedure
described previously. A 0.15 molar deficiency of HFP to
hydrocarbon was used. The reactions were followed by the dis-
appearance of hydrocarbons from capillary GC traces (Flame
ionisation detector) before and after the reaction, therefore
eliminating any differences in detector responses.
Competition between cyclohexane and n-hexane. A Carius
tube was charged with cyclohexane 18 (8.4 g, 0.1 mol), n-hexane
11 (8.6 g, 50 mmol) and HFP (4.6 g, 30 mmol) and then irradi-
ated with γ-rays for 10 days. The Carius tube was opened and
any remaining HFP recovered. The GC traces from before and
after the reaction showed the peak integration of cyclohexane
General procedure for elimination of hydrogen fluoride from
hexafluoropropene adducts
The hexafluoropropene adduct was added dropwise to a stirred
solution/suspension of the alkoxide in its solvent. The reaction
was monitored by the 19F NMR spectra of the reaction mixture
1634
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
and when elimination was complete the mixture was poured
into water, neutralised with 10% HCl, the organic layer was
separated and dried over MgSO4 and the products purified by
distillation.
δH (250 MHz; CDCl3; Me4Si) 1.69 (4 H, m, H-2/3/5/6), 1.89
3 3
(4 H, d, JHH 7.2, H-2/3/5/6), 2.56 (2 H, dm, JHF 32, H-1/4);
δF (235 MHz; CDCl3; CFCl3) Ϫ65.9 (3 F, br s, CF3), Ϫ132.2
3
(1 F, dm, JHF 32, CHCF), Ϫ159.2 (1 F, br s, CFCF3); δC (100
3
MHz; CDCl3; Me4Si) 27.8 (s, C-2/3/4/5), 35.5 (d, JCF 21, C-1/
4), 120.2 (qdd, 1JCF 270, 2JCF 35, 3JCF 9.6, CF3), 135.3 (dqd, 1JCF
Eliminations from:
3
3
1
3
252, JCF 40, JCF 24, CFCF3), 155.3 (dd, JCF 266, JCF 10,
CHCF); m/z (EIϩ) 344 (Mϩ, 9%), 158 (83), 108 (83), 95 (100), 54
(93).
(1,1,2,3,3,3-Hexafluoropropyl)cyclopentane. Addition of
(1,1,2,3,3,3-hexafluoropropyl)cyclopentane 16 (4.4 g, 20 mmol)
to sodium tert-butoxide (5.0 g, 40 mmol) in diethyl ether (40 ml)
at Ϫ78 ЊC for 30 minutes gave (1Z)-pentafluoroprop-1-enyl-
cyclopentane 38 (3.0 g, 75%) bp 119–121 ЊC (Found: C, 47.9; H,
4.5. C8H9F5 requires C, 48.0; H, 4.5%); δH (250 MHz; CDCl3;
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane (x ؍ 2–4).
Addition of 1,x-bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane
20 (16.0 g, 40 mmol) in diethyl ether (30 ml) to sodium tert-
butoxide (15.9 g, 0.14 mol) in diethyl ether (50 ml) at Ϫ10 ЊC for
30 minutes gave a fraction which was identified as a mixture of
cis-1,3-bis[(1Z)-pentafluoroprop-1-enyl]cyclohexane 41b (8.2 g,
57%) bp20 92–93 ЊC (Found: C, 41.7; H, 2.9. C12H10F10 requires
C, 41.9; H, 2.9%); δH (250 MHz; CDCl3; Me4Si) 1.42 (1 H, qm,
3
Me4Si) 1.63 (9 H, m, ring H), 2.52 (1 H, dm, JHF 32, CHCF);
δF (235 MHz; CDCl3; CFCl3) Ϫ66.2 (3 F, br s, CF3), Ϫ133.9
3
(1 F, d, JHF 32, CHCF), Ϫ160.1 (1 F, br s, CFCF3); δC (100
2
MHz; CDCl3; Me4Si) 26.2 (s, C-3), 29.5 (s, C-2), 36.8 (d, JCF
1
2
3
22, C-1), 120.4 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 134.7
(dqd, 1JCF 250, 2JCF 40, 2JCF 24, CFCF3), 155.3 (ddq, 1JCF 264,
2JCF 10, 3JCF 3.5, CHCF); m/z (EIϩ) 200 (Mϩ, 13%), 158 (60), 68
(98), 42 (100), 41 (89), 39 (82).
2,3
2,3
3
J
12, H-5), 1.57 (2 H, qd,
J
12, JHH 3.6, H-4/6), 1.77
HH
HH
2,3
(3 H, m, H-4/5/6), 1.88 (1 H, q,
J
13, H-2), 2.10 (1 H, d,
HH
3
2JHH 13, H-2), 2.64 (2 H, dt, JHF 31, JHH 12, H-1/3); δF (235
MHz; CDCl3; CFCl3) Ϫ66.3 (3 F, br s, CF3), Ϫ132.2 (1 F, d,
3JHF 31, CHCF), Ϫ159.1 (1 F, br s, CFCF3); δC (100 MHz;
3
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclopentane (x ؍ 2, 3).
Addition of 1,x-bis(1,1,2,3,3,3-hexafluoropropyl)cyclopentane
17 (7.4 g, 20 mmol) to sodium tert-butoxide (9.0 g, 80 mmol) in
diethyl ether (50 ml) at Ϫ78 ЊC for 30 minutes gave an isomeric
mixture of 1,x-bis[(1Z)-pentafluoroprop-1-enyl]cyclopentane
(x = 2, 3) 39 (4.0 g, 60%) bp21 76–77 ЊC (Found: C, 43.9; H, 2.6.
C11H8F10 requires C, 43.7; H, 2.4%) including 1,3-bis[(1Z)-
pentafluoroprop-1-enyl]cyclopentane 39a δH (250 MHz; CDCl3;
Me4Si) 1.65 (6 H, m, CH2), 2.50 (2 H, m, CHCF); δF (235 MHz;
CDCl3; CFCl3) Ϫ66.9 (6 F, br s, CF3), Ϫ135.0 (2 F, br s,
CHCF), Ϫ159.3 (2 F, br s, CFCF3); δC (100 MHz; CDCl3;
3
CDCl3; Me4Si) 24.6 (s, C-5), 27.4 (d, JCF 2.3, C-4/6), 30.0 (s,
C-2), 35.6 (d, 2JCF 21, C-1/3), 119.8 (qdd, 1JCF 270, 2JCF 35, 3JCF
9.6, CF3), 135.1 (dqd, 1JCF 253, 2JCF 41, 2JCF 24, CFCF3), 154.5
(ddq, JCF 266, JCF 11, JCF 3.5, CHCF); m/z (EIϩ) 344 (Mϩ,
8%), 211 (45), 158 (94), 54 (100), 41 (85); and trans-1,4-bis[(1Z)-
pentafluoroprop-1-enyl]cyclohexane 41a (4.3 g, 30%) which
precipitated out on cooling in an acetone–carbon dioxide bath
(Ϫ78 ЊC).
1
2
3
2
x-(1,1,2,3,3,3-Hexafluoropropyl)-trans-decahydronaphthalene
(x ؍ 1, 2). x-(1,1,2,3,3,3-Hexafluoropropyl)-trans-decahydro-
naphthalene (x = 1, 2) 22 (5.8 g, 20 mmol) was added to sodium
tert-butoxide (4.5 g, 40 mmol) in dry diethyl ether (20 ml),
cooled to Ϫ10 ЊC, and stirred for 30 minutes. Fractional distil-
lation gave an isomeric mixture of x-[(1Z)-pentafluoroprop-1-
enyl]-trans-decahydronaphthalene (x = 1, 2) 42 (4.4 g, 82%) bp6
75–77 ЊC (Found: C, 58.1; H, 6.6. C13H17F5 requires C, 58.2; H,
6.3%); 2-[(1Z)-pentafluoroprop-1-enyl]-trans-decahydronaph-
thalene 42b δH (250 MHz; CDCl3; Me4Si) 1.02 to 1.72 (16 H,
overlapping m, CH ϩ CH2), 2.58 (1 H, dm, 3JHF 32, CHCF); δF
(235 MHz; CDCl3; CFCl3) Ϫ66.3 (3 F, br s, CF3), Ϫ131.7 (1 F,
Me4Si) 29.9 (s, C4/5), 31.4 (s, C-2), 36.6 (dm, JCF 22, C-1/3),
1
2
3
1
120.0 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 135.2 (dqd, JCF
252, 2JCF 40, 2JCF 24, CFCF3), 153.6 (ddq, 1JCF 264, 2JCF 11, 3JCF
3.4, CHCF); m/z (EIϩ) 330 (Mϩ, 13%), 197 (100), 158 (65), 51
(62); and 1,2-bis[(1Z)-pentafluoroprop-1-enyl]cyclopentane
39b δH (250 MHz; CDCl3; Me4Si) 1.65 (6 H, m, CH2), 2.50 (2 H,
m, CHCF); δF (235 MHz; CDCl3; CFCl3) Ϫ66.9 (6 F, br s,
CF3), Ϫ135.9 (1 F, d, 3JHF 31, CHCF), Ϫ156.4 (1 F, s, CFCF3);
δC (100 MHz; CDCl3; Me4Si) 25.1 (s, C-4), 28.9 (s, C-3/5), 40.3
2
1
1
(dm, JCF 23, C-1/2), 119.8 (qm, JCF 270, CF3), 136.5 (d, JCF
250, CFCF3), 151.4 (dm, 1JCF 268, CHCF); m/z (EIϩ) 330 (Mϩ,
1%), 158 (100), 103 (40), 69 (54).
3
d, JHF 32, CHCF), Ϫ161.5 (1 F, br s, CFCF3); δC (100 MHz;
CDCl3; Me4Si) 26.4 (s, C-6), 26.5 (s, C-7), 28.5 (d, 5JCF 2.7, C-3),
32.8 (s, C-4), 33.5 (s, C-5), 33.6 (s, C-8), 35.6 (d, 5JCF 2.6, C-1),
36.4 (dm, 2JCF 21, C-2), 42.2 (s, C-10), 42.3 (s, C-9), 120.0 (qdd,
(1,1,2,3,3,3-Hexafluoropropyl)cyclohexane. Addition of
(1,1,2,3,3,3-hexafluoropropyl)cyclohexane 19 (70.2 g, 0.3 mol)
to sodium tert-butoxide (50.5 g, 0.45 mol) in dry diisopropyl
ether (100 ml) at 0 ЊC for 30 min gave (1Z)-pentafluoroprop-1-
enylcyclohexane 40 (59.1 g, 92%) bp 139–140 ЊC (Found: C,
2
3
1
2
1JCF 271, JCF 35, JCF 9.6, CF3), 135.5 (dqd, JCF 250, JCF 40,
2JCF 24, CFCF3), 156.0 (ddq, JCF 266, JCF 9.4, JCF 3.4,
CHCF); m/z (EIϩ) 268 (Mϩ, 17%), 135 (71), 67 (65), 41 (100);
1-[(1Z)-1,2,3,3,3-pentafluoroprop-1-enyl]-trans-decahydro-
naphthalene 42a δH (250 MHz; CDCl3; Me4Si) 1.02 to 1.72,
1
2
3
50.6; H, 5.2. C9H11F5 requires C, 50.5; H, 5.2%); δH (250 MHz;
2,3
CDCl3; Me4Si) 1.20 (1 H, qt,
J
12, 3JHH 3.2, H-4), 1.31 (2 H,
HH
2,3
2,3
qm,
J
12, H-3/5), 1.55 (2 H, qd,
J
12, 3JHH 3.2, H-2/6),
HH
HH
3
1.72 (3 H, dm, 3JHH 12, H-2/4/6), 1.83 (2 H, d, 4JHF 12, H-3/5),
(16 H, overlapping m, CH ϩ CH2), 2.23 (1 H, dm, JHF 33,
3
3
CHCF); δF (235 MHz; CDCl3; CFCl3) Ϫ65.7 (3 F, br s, CF3),
2.52 (1 H, dt, JHF 32, JHH 12, H-1); δF (235 MHz; CDCl3;
CFCl3) Ϫ66.2 (3 F, br s, CF3), Ϫ131.4 (1 F, d, 3JHF 32, CHCF),
Ϫ161.7 (1 F, br s, CFCF3); δC (100 MHz; CDCl3; Me4Si) 25.6 (s,
3
Ϫ133.1 (1 F, d, JHF 33, CHCF), Ϫ160.0 (1 F, br s, CFCF3);
δC (100 MHz; CDCl3; Me4Si) 25.3 (s, C-6), 26.2 (s, C-7), 26.3 (s,
C-3), 29.5 (s, C-2), 30.1 (s, C-5), 33.3 (s, C-4), 33.5 (dm, 2JCF 21,
C-1), 34.0 (s, C-8), 42.4 (s, C-10), 42.5 (br s, C-9), 120.0 (qdd,
3
2
C-4), 25.9 (s, C-3/5), 28.9 (d, JCF 2.2, C-2/6), 36.7 (d, JCF 21,
1
2
3
C-1), 120.4 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 134.9 (dqd,
2
3
1
2
1JCF 271, JCF 35, JCF 9.6, CF3), 135.8 (dqd, JCF 250, JCF 40,
2JCF 24, CFCF3), 154.9 (ddq, 1JCF 266, 2JCF 9.4, 3JCF 3.4, CHCF);
m/z (EIϩ) 268 (Mϩ, 17%), 95 (51), 81 (100), 67 (92), 41 (47).
1JCF 250, JCF 40, JCF 24, CFCF3), 156.6 (ddq, JCF 264, JCF
2
2
1
2
9.6, JCF 3.4, CHCF); m/z (EIϩ) 214 (Mϩ, 40%), 158 (91), 100
3
(88), 82 (79), 56 (77), 41 (100).
trans-1,4-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane. Addi-
tion of trans-1,4-bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane
20a (3.5 g, 9 mmol) to sodium tert-butoxide (2.1 g, 18 mmol) in
dry diisopropyl ether (25 ml) at Ϫ10 ЊC, using a salt–ice bath,
for 30 min gave trans-1,4-bis[(1Z)-pentafluoroprop-1-enyl]-
cyclohexane 41a (2.1 g, 67%) mp 101–102 ЊC (from MeOH)
(Found: C, 41.8; H, 2.9. C12H10F10 requires C, 41.9; H, 2.9%);
x-(1,1,2,3,3,3-Hexafluoropropyl)-cis-decahydronaphthalene
(x ؍ 1, 2, 9). x-(1,1,2,3,3,3-Hexafluoropropyl)-cis-decahydro-
naphthalene (x = 1, 2, 9) 25 (5.8 g, 20 mmol) was added to
sodium tert-butoxide (4.5 g, 40 mmol) in dry diethyl ether (40
ml), cooled to Ϫ10 ЊC, and stirred for 30 minutes. Fractional
distillation gave an isomeric mixture of 2-[(1Z)-pentafluoro-
prop-1-enyl]-cis-decahydronaphthalene (x = 1, 2, 9) 43 (4.6 g,
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1635

View Article Online
85%), bp20 138–140 ЊC (Found: C, 58.1; H, 6.3. C13H17F5
requires C, 58.2; H, 6.3%); 2-[(1Z)-pentafluoroprop-2-enyl]-cis-
decahydronaphthalene 43b δH (250 MHz; CDCl3; Me4Si) 1.29
to 1.84 (16 H, overlapping m, CH ϩ CH2), 2.65 (1 H, dm, 3JHF
36, CHCF); δF (235 MHz; CDCl3; CFCl3) Ϫ66.2 (3 F, br s,
1.86 (2 H, m, CH2), 2.46 (1 H, m, CH2), 2.53 (1 H, m, CH), 2.63
(2 H, dm, 3JHF 36, CHCF); δF (235 MHz; CDCl3; CFCl3) Ϫ66.0
(3 F, br s, CF3), Ϫ128.7 (1 F, d, 3JHF 36, CHCF), 158.6 (1 F, br s,
CF712;CF3); δC (100 MHz; CDCl3; Me4Si) 33.1 (s, C-3/5), 36.2 (s,
C-4), 36.5 (s, C-7), 41.9 (dm, 2JCF 20, C-2/6), 48.2 (s, C-1), 120.4
(qdd, 1JCF 269, 2JCF 35, 3JCF 9.4, CF3), 135.7 (dqd, 1JCF 252, 2JCF
3
CF3), Ϫ131.3 (1 F, d, JHF 36, CHCF), Ϫ161.3 (1 F, br
2
1
2
3
s, CFCF3); δC (100 MHz; CDCl3; Me4Si) 20.8 (s, C-6), 25.6 (s,
40, JCF 25, CFCF3), 154.8 (ddq, JCF 267, JCF 11, JCF
3.4, CHCF); m/z (EIϩ) 356 (Mϩ, 3%), 198 (80), 139 (49), 129
(100).
5
2
C-7), 28.1 (d, JCF 2.7, C-3), 28.9 (s, C-5), 31.3 (dm, JCF 21,
C-2), 32.2 (s, C-4), 34.9 (d, 5JCF 2.3, C-1), 35.2 (s, C-10), 35.3 (s,
1
2
3
C-9), 120.4 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 134.9 (dqd,
1JCF 248, JCF 40, JCF 24, CFCF3), 156.7 (ddq, JCF 266, JCF
1-(1,1,2,3,3,3-Hexafluoropropyl)adamantane. 1-(1,1,2,3,3,3-
Hexafluoropropyl)adamantane 31 (11.4 g, 40 mmol) was added
to sodium tert-butoxide (7.9 g, 70 mmol) in dry diisopropyl
ether (50 ml) at room temperature, and stirred for 30 minutes.
Fractional distillation gave 1-[(1E)-pentafluoroprop-1-
enyl]adamantane 46b (9.6 g, 90%) bp 219–221 ЊC (Found: C,
58.6; H, 5.9. C13H15F5 requires C, 58.6; H, 5.6%); δH (250 MHz;
CDCl3; Me4Si) 1.77 (6 H, br s, CH2), 1.97 (6 H, br s, CH2), 2.06
(3 H, br s, CH); δF (235 MHz; CDCl3; CFCl3) Ϫ67.6 (3 F, d,
4JF-F 22, CF3), Ϫ149.3 (1 F, dq, 3JFF 131, 4JFF 22, CCF), Ϫ175.6
(1 F, d, 3JFF 131, CFCF3); δC (100 MHz; CDCl3; Me4Si) 27.8 (s,
CH2), 36.3 (s, CH), 37.8 (m, CH2), 119.5 (qdd, 1JCF 273, 2JCF 36,
2
2
1
2
9.6, JCF 3.4, CHCF); m/z (EIϩ) 268 (Mϩ, 12%), 135 (67), 95
3
(98), 41 (100), 39 (77); 1-[(1Z)-pentafluoroprop-1-enyl]-cis-
decalin 43a δH (250 MHz; CDCl3; Me4Si) 1.29 to 1.84 (16 H,
3
overlapping m, CH ϩ CH2), 2.58 (1 H, dm, JHF 32, CHCF);
δF (235 MHz; CDCl3; CFCl3) Ϫ66.2 (3 F, br s, CF3), Ϫ132.2
3
(1 F, d, JHF 32, CHCF), Ϫ161.5 (1 F, br s, CFCF3); δC (100
MHz; CDCl3; Me4Si) 20.9 (s, C-6), 23.1 (d, 5JCF 2.6, C-2), 24.9
(s, C-7), 26.1 (s, C-3), 27.0 (s, C-5), 32.1 (s, C-4), 35.7 (s, C-10),
35.9 (s, C-9), 37.3 (dm, 2JCF 21, C-1), 120.4 (qdd, 1JCF 270, 2JCF
3
1
2
2
35, JCF 9.6, CF3), 134.8 (dqd, JCF 248, JCF 40, JCF 24,
1
2
3
CFCF3), 156.5 (ddq, JCF 265, JCF 9.6, JCF 3.4, CHCF); m/z
(EIϩ) 268 (Mϩ, 14%), 95 (42), 81 (100), 67 (49), 41 (52); 9-[(1Z)-
pentafluoroprop-1-enyl]decalin 43c δH (250 MHz; CDCl3;
Me4Si) 1.29 to 1.84 (17 H, overlapping m, CH ϩ CH2); δF (235
MHz; CDCl3; CFCl3) Ϫ65.7 (3 F, m, CF3), Ϫ134.2 (1 F, d, J 30,
CHCF), Ϫ160.1 (1 F, s, CFCF3); δC (100 MHz; CDCl3; Me4Si)
20.6 (s, C-3/6), 26.7 (s, C-2/7), 29.0 (s, C-4/5), 32.5 (br s, C-1/8),
34.0 (dm, 2JCF 21, C-9), 36.1 (s, C-10), 120.4 (m, CF3), 136.0 (m,
CFCF3), 155.9 (m, CCF); m/z (EIϩ) 268 (Mϩ, 13%), 95 (50), 81
(87), 67 (100), 41 (88).
1
2
2
3JCF 3.5, CF3), 138.2 (dqd, JCF 242, JCF 54, JCF 39, CFCF3),
160.2 (ddq, JCF 260, JCF 36, JCF 2.3, CCF); m/z (EIϩ) 266
1
2
3
(Mϩ, 80%), 94 (98), 93 (100), 79 (89), 41 (95).
1-(1,1,2,3,3,3-Hexafluoropropyl)adamantane. 1-(1,1,2,3,3,3-
Hexafluoropropyl)adamantane 31 (3.8 g, 13 mmol) was added
to sodium tert-butoxide (3.2 g, 28 mmol) in dry diisopropyl
ether (50 ml) at Ϫ10 ЊC, and stirred for 30 minutes. Fractional
distillation gave 1-[(1Z)-pentafluoroprop-1-enyl]adamantane
46a (3.0 g, 85%) bp 219–221 ЊC (Found: C, 58.7; H, 5.4.
C13H15F5 requires C, 58.6; H, 5.6%); δH (250 MHz; CDCl3;
Me4Si) 1.77 (6 H, br s, CH2), 1.94 (6 H, br s, CH2), 2.07 (3 H, br
s, CH); δF (235 MHz; CDCl3; CFCl3) Ϫ59.9 (3 F, br s, CF3),
Ϫ125.2 (1 F, br s, CCF), Ϫ154.7 (1 F, br s, CFCF3); δC (100
MHz; CDCl3; Me4Si) 27.9 (s, CH2), 36.2 (s, CH), 37.1 (dm, 2JCF
21, CCF), 38.0 (q, 3JCF 2.3, CH2), 120.1 (qdd, 1JCF 270, 2JCF 36,
exo-2-(1,1,2,3,3,3-Hexafluoropropyl)bicyclo[2.2.1]heptane.
exo-2-(1,1,2,3,3,3-Hexafluoropropyl)bicyclo[2.2.1]heptane 28
(7.5 g, 30 mmol) was added to sodium tert-butoxide (6.8 g,
60 mmol) in dry diisopropyl ether (20 ml) cooled to Ϫ10 ЊC,
and stirred for 1 hour. Fractional distillation gave exo-2-[(1Z)-
pentafluoroprop-1-enyl]bicyclo[2.2.1]heptane 44 (6.2 g, 90%)
bp20 65–67 ЊC (Found: C, 53.1; H, 4.9. C10H11F5 requires C,
53.1; H, 4.9%); δH (250 MHz; CDCl3; Me4Si) 1.25 (3 H, m,
CH2), 1.57 (4 H, m, CH2), 1.71 (1 H, m, CH2), 2.35 (2 H, s, CH),
1
2
2
3JCF 8.4, CF3), 137.8 (dqd, JCF 249, JCF 43, JCF 30, CFCF3),
160.7 (dd, 1JCF 260, 2JCF 13, CHCF); m/z (EIϩ) 266 (Mϩ, 18%),
94 (100), 93 (97), 79 (65), 41 (57).
3
2.54 (1 H, dm, JHF 36, CHCF); δF (235 MHz; CDCl3; CFCl3)
Ϫ67.1 (3 F, br s, CF3), Ϫ130.7 (1 F, d, 3JHF 36, CHCF), Ϫ162.2
(1 F, br s, CFCF3); δC 27.9 (s, C-6), 30.3 (s, C-5), 34.7 (s, C-3),
36.1 (s, C-4), 37.4 (s, C-7), 38.8 (dm, 2JCF 21, C-2), 42.0 (d, 5JCF
1,3,5,7-Tetrakis(1,1,2,3,3,3-hexafluoropropyl)adamantane.
1,3,5,7-Tetrakis(1,1,2,3,3,3-hexafluoropropyl)adamantane 34
(3.7 g, 5 mmol) in dry diethyl ether (20 ml) was added to sodium
tert-butoxide (4.5 g, 40 mmol) in dry diethyl ether (30 ml) and
stirred at room temperature. Kugelrohr distillation gave an
isomeric mixture of 1,3,5,7-tetrakis(pentafluoroprop-1-enyl)-
adamantane 47, 48 (2.6 g, 78%) bp1 175 ЊC (Found: C, 40.2;
H, 1.8. C22H12F20 requires C, 40.3; H, 1.8%); δH (250 MHz;
CDCl3; Me4Si) 2.18 (6 H, br s, CH2), 2.19 (6 H, br s, CH2);
δF (235 MHz; CDCl3; CFCl3) Ϫ60.4 (3 F, br s, CF3), Ϫ68.5
(9 F, br s, CF3), Ϫ126.7 (1 F, br s, CCF), Ϫ148.6 (1 F, br s,
1
2
3
3.2, C-1), 120.2 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 134.5
(dqd, 1JCF 250, 2JCF 40, 2JCF 25, CFCF3), 156.5 (ddq, 1JCF 265,
2JCF 9.6, 3JCF 3.4, CHCF); m/z (EIϩ) 226 (Mϩ, 4%), 139 (10), 68
(100), 67 (35).
exo-2,x-Bis(1,1,2,3,3,3-hexafluoropropyl)bicyclo[2.2.1]-
heptane (x ؍ 5, 6). exo-2,x-Bis(1,1,2,3,3,3-hexafluoropropyl)-
bicyclo[2.2.1]heptane (x = 5, 6) 29 (4.6 g, 12 mmol) was added to
sodium tert-butoxide (4.5 g, 40 mmol) in dry diisopropyl ether
(15 ml), cooled to Ϫ10 ЊC, and stirred for 30 minutes. Frac-
tional distillation gave an isomeric mixture of exo-2,x-bis[(1Z)-
pentafluoroprop-1-enyl]bicyclo[2.2.1]heptane (x = 5, 6) 45 (3.8
g, 93%) bp20 101–103 ЊC (Found: C, 43.6; H, 2.8. C13H10F10
requires C, 43.8; H, 2.8%); exo-2,5-bis[(1Z)-pentafluoroprop-1-
enyl]bicyclo[2.2.1]heptane 45a δH (250 MHz; CDCl3; Me4Si)
1.65 (2 H, m, CH2), 1.83 (2 H, m, CH2), 2.50 (2 H, m, CH), 2.61
(2 H, m, CHCF); δF (235 MHz; CDCl3; CFCl3) Ϫ66.0 (3 F,
3
CFCF3), Ϫ150.6 (3 F, m, CCF), Ϫ171.6 (3 F, d, JFF 135,
CFCF3); m/z (EIϩ) 656 (Mϩ, 10%), 145 (100), 95 (66), 69
(51); from which 1,3,5,7-tetrakis[(1E)-pentafluoroprop-1-enyl]-
adamantane 47 δH (250 MHz; CDCl3; Me4Si) 2.21 (12 H, br
4
s, CH2); δF (235 MHz; CDCl3; CFCl3) Ϫ67.6 (12 F, d, JFF
3
4
23, CF3), Ϫ149.7 (3 F, dq, JFF 134, JFF 23, CCF), Ϫ170.3
3
3
(3 F, dq, JFF 134, JFF 9.4, CFCF3); δC (100 MHz; CDCl3;
1
2
3
Me4Si) 36.4 (s, CH2), 118.8 (qdd, JCF 273, JCF 36, JCF 3.4,
1
2
2
CF3), 139.7 (dqd, JCF 248, JCF 52, JCF 40, CFCF3), 156.1
3
1
2
m, CF3), Ϫ130.5 (1 F, d, JHF 32, CHCF), Ϫ159.3 (1F, br s,
(dd, JCF 261, JCF 39, CCF); m/z (EIϩ) 656 (Mϩ, 5%), 145
(100), 95 (73), 69 (56) crystallised out when the mixture was
dissolved in chloroform and cooled in an acetone–carbon
dioxide bath (Ϫ78 ЊC).
CFCF3); δC (100 MHz; CDCl3; Me4Si) 35.2 (s, C-3/6), 37.0 (s,
C-7), 38.0 (dd, 2JCF 21, 3JCF 2.6, C-2/5), 41.9 (d, 3JCF 2.6, C-1/4),
1
2
3
1
120.5 (qdd, JCF 270, JCF 35, JCF 9.6, CF3), 135.3 (dqd, JCF
251, 2JCF 40, 2JCF 25, CFCF3), 155.2 (ddq, 1JCF 266, 2JCF 11, 3JCF
3.4, CHCF); m/z (EIϩ) 356 (Mϩ, 3%), 198 (85), 139 (47), 129
(100); exo-2,6-bis[(1Z)-pentafluoroprop-1-enyl]bicyclo[2.2.1]-
heptane 45b δH (250 MHz; CDCl3; Me4Si) 1.65 (4 H, m, CH2),
Conversion of 1-[(1Z)-pentafluoroprop-1-enyl]adamantane. A
Carius tube was charged with caesium fluoride (7.6 g, 50
mmol), dry tetraglyme (10 ml) and 1-[(1Z)-pentafluoroprop-2-
1636
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638

View Article Online
Table 8 Crystal data
Compound
20a
34
41a
47
Formula
Formula weight
T/K
Crystal system
a/Å
b/Å
c/Å
C12H12F12
384.2
293
C22H16F24
736.4
150
orthorhombic
11.442(8)
14.992(14)
15.011(10)
90
C12H10F10
344.2
150
monoclinic
15.006(13)
7.217(4)
6.223(5)
90
102.74(7)
90
C22H12F20
656.3
293
monoclinic
15.758(3)
15.951(3)
19.896(4)
90
93.21(1)
90
triclinic
6.365(4)
6.829(3)
8.949(5)
71.54(4)
82.58(6)
89.22(5)
365.7(3)
α/Њ
β/Њ
90
90
γ/Њ
V/Å3
2575(3)
Pnma1 (#62)
4
657.4(9)
C2/m (#12)
2
4993(2)
I2/a (#15)
8
¯
Space group
Z
P1 (#2)
1
λ/Å
0.71073
0.21
1.745
655
40
578
335
0.026
79
0.118
0.037
0.15, Ϫ0.21
0.71073
0.23
1.90
10933
51
2348
1558
0.040
250
0.357
0.71073
0.20
1.74
1042
55
816
672
0.014
88
1.54184
µ/mmϪ1
Dcalc/g cmϪ3
Reflections total
Max. 2θ/Њ
Unique refls.
Refls. with I > 2σ(I)
Rint
No. of variables
wR(F2), all data
R[I > 2σ(I)]
∆ρmax,min/e ÅϪ3
0.114
0.044
0.26, Ϫ0.23
0.119
0.62, Ϫ0.63
enyl]adamantane 46a (2.7 g, 10 mmol). The tube was cooled
(liquid air) and sealed under vacuum. It was allowed to warm to
room temperature and then it was heated in a rotating oil bath
at 200 ЊC for 50 h. On completion the tube was opened and a
sample of the product mixture was transferred into an NMR
tube and a 19F NMR spectrum was recorded, which showed
almost complete conversion to 1-[(1E)-pentafluoroprop-1-
enyl]adamantane 46b (ca. 97% by 19F NMR).
a cylinder, via FEP tubing through the heated reactor
(280 ЊC) until the soda-lime trap, attached to the exit, became
detectably warm.
(1,1,2,3,3,3-Hexafluoropropyl)cyclopentane.
(1,1,2,3,3,3-
Hexafluoropropyl)cyclopentane 16 (3.9 g, 18 mmol), was
passed through the cobalt trifluoride reactor, at 375 ЊC, to give
a colourless liquid (6.1 g). Continuous extraction followed by
preparative GC gave perfluoropropylcyclopentane 49 (3.7 g,
60%) bp 103–105 ЊC; δF (235 MHz; CDCl3; CFCl3) Ϫ81.2 (3 F,
s, CF3), Ϫ116.1 (2 F, s, CF2), Ϫ123.0 and Ϫ128.3 (4 F, AB, JAB
268, CF2), Ϫ125.1 (2 F, s, CF2), Ϫ129.0 and Ϫ132.5 (4 F, AB,
JAB 259, CF2), Ϫ185.2 (1 F, s, CF); m/z (EIϩ) 381 (1%), 281 (32),
181 (33), 131 (70), 69 (100).
1-(1,1,2,3,3,3-Hexafluoropropyl)adamantane.
A
sealable
NMR tube, cooled in liquid air, was charged with sodium tert-
butoxide (0.1 g, 0.6 mmol), tert-butanol (D) (2-methylpropan-
2-ol, ca. 1 ml) and 1-(1,1,2,3,3,3-hexafluoropropyl)adamantane
31 (0.3 g, 1 mmol). The tube was allowed to warm to room
temperature and then left for a further 15 minutes at 25 ЊC. A
19F NMR spectrum was then recorded on the contents, which
indicated 2D-hexafluoropropyladamantane 31a (9% by 19F
NMR spectroscopy); δF (235 MHz; CDCl3; CFCl3) Ϫ207.5 (br
s, CFD).
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclopentane, (x ؍ 2, 3).
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclopentane (x = 2, 3) 17
(5.0 g, 14 mmol), was passed through the cobalt trifluoride
reactor, at 375 ЊC, to give a colourless liquid (5.7 g). Continuous
extraction followed by preparative GC gave an isomeric mixture
of perfluoro-1,x-di-n-propylcyclopentane (x = 2,3) 50 (4.0 g,
71%) bp 159–161 ЊC; δF (235 MHz; CDCl3; CFCl3) Ϫ83.31
(12 F, m, 4CF3), Ϫ114.2 to Ϫ134.4 (28 F, overlapping m,
14CF2), Ϫ183.9 (2 F, s, 2CF), Ϫ185.9 (2 F, s, 2CF); m/z (EIϩ)
531 (5%), 431 (13), 281 (70), 181 (68), 169 (67), 131 (87), 69
(100).
General procedure for fluorinations with cobalt trifluoride
The apparatus for fluorination with cobalt trifluoride con-
sisted of a nickel tube with inlet and outlet pipes and nickel
paddles attached to a rod situated along the axis of the tube
and rotated by an electric motor. Cobalt trifluoride (440 g)
was contained in the tube and heated to the required tem-
perature, using an electric heating tape wrapped around the
tube, with continuous stirring. Dry nitrogen (30 ml minϪ1)
was passed through the reactor for 10 min, prior to use.
Starting materials were fed into the reactor at a rate of 0.1 ml
minϪ1 in a nitrogen flow (30 ml minϪ1). After all the com-
pound for fluorination was added, the reactor was flushed
with nitrogen for 30 min. Products were collected in a trap
cooled with liquid air, which was then detached from the
reactor and left to warm up in a fume cupboard. A con-
denser containing anhydrous soda-lime was attached to the
trap, so as to remove any hydrogen fluoride evolved and the
product was pipetted out. Any hydrogen-containing prod-
ucts were removed by continuous extraction with acetone
and pure samples of the perfluorinated product were isolated
by preparative GC. After use, the cobalt fluoride system was
regenerated by passing 50% fluorine gas (95 ml minϪ1), from
(1,1,2,3,3,3-Hexafluoropropyl)cyclohexane.
(1,1,2,3,3,3-
Hexafluoropropyl)cyclohexane 19 (3.5 g, 15 mmol) was passed
through the cobalt trifluoride reactor, at 375 ЊC, to give a col-
ourless liquid (6.0 g). Continuous extraction followed by pre-
parative GC gave perfluoropropylcyclohexane 51 (3.8 g, 63%)
bp 132–134 ЊC; δF (235 MHz; CDCl3; CFCl3) Ϫ80.6 (3 F, s,
CF3), Ϫ112.3 (2 F, s, CF2), Ϫ118.2 and Ϫ128.9 (4 F, AB, JAB
302, CF2), Ϫ121.9 and Ϫ139.9 (4 F, AB, JAB 287, CF2), Ϫ123.9
and Ϫ142.2 (4 F, AB, JAB 289, CF2), Ϫ128.4 (2 F, s, CF2),
Ϫ185.6 (1 F, s, CF); m/z (EIϩ) 431 (3%), 181 (15), 169 (14), 131
(49), 69 (100) in agreement with the literature.22
[(1Z)-Pentafluoro-1-propenyl]cyclohexane.
[(1Z)-Penta-
fluoro-1-propenyl)cyclohexane 40 (4.0 g, 19 mmol) was passed
through the cobalt trifluoride reactor, at 375 ЊC, to give a
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638
1637

View Article Online
colourless liquid (7.1 g). Preparative GC gave perfluoropropyl-
cyclohexane 51 (6.4 g, 90%).
electron density map is in agreement with the expected
molecular structure. Crystal data and experimental param-
eters are listed in Table 8. CCDC reference number 207/408.
See http://www.rsc.org/suppdata/p1/a9/a909777e for crystallo-
graphic files in .cif format.
trans-1,4-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane. trans-
1,4-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane 20a (2.4 g, 6
mmol) was passed through the cobalt trifluoride reactor, at
400 ЊC, to give a colourless liquid (2.9 g). Preparative scale
GC gave a mixture of cis- and trans-isomers of perfluoro-
1,4-dipropylcyclohexane 52 (1.9 g, 65%); cis-perfluoro-1,4-
dipropylcyclohexane δF (235 MHz; CDCl3; CFCl3) Ϫ81.0 (6 F,
s, CF3), Ϫ113.4 (4 F, s, CF2), Ϫ119.4 (4 F, s, CF2), Ϫ124.8 (8 F,
br s, CF2), Ϫ183.0 (2 F, br s, CF); m/z (EIϩ) 481 (4%), 331 (20),
181 (21), 169 (100), 131 (27), 119 (37), 69 (46); trans-perfluoro-
1,4-dipropylcyclohexane mp 80–81 ЊC (from CFCl3); δF (235
MHz; CDCl3; CFCl3) Ϫ81.2 (6 F, br s, CF3), Ϫ117.7 and
Ϫ126.7 (8 F, AB, JAB 297, CF2), Ϫ119.4 (4 F, s, CF2), Ϫ126.7
(4 F, s, CF2), Ϫ186.7 (2 F, s, CF); m/z (EIϩ) 481 (4%), 331 (16),
181 (21), 169 (100), 131 (21), 119 (26), 69 (69).
Acknowledgements
We thank BNFL (R. C. H. S.), F2 Chemicals Ltd. (M. P. G.),
and the European Community (R. W. F) for financial support
and Dr J. Hutchinson for valuable discussion.
References
1 Part 9. R. D. Chambers, J. Hutchinson, N. M. Kelly, A. K. Joel,
A. J. Rees and P. T. Telford, Isr. J. Chem., 1999, 39, 133.
2 D. Barton and W. D. Ollis, Comprehensive Organic Chemistry,
Pergamon, Oxford, 1979.
3 R. C. Larock, Comprehensive Organic Transformations, John Wiley,
New York, 1989.
4 G. A. Olah and A. Molnar, Hydrocarbon Chemistry, Wiley-
Interscience, New York, 1995.
5 R. N. Haszeldine and R. Rowland, USP 3,816,286, 1974 (Chem.
Abstr., 1974, 81, 49245v).
6 R. N. Haszeldine and R. Rowland, USP 3,927,129, 1975 (Chem.
Abstr., 1976, 84, 121095c).
7 T. Davies, R. N. Haszeldine, R. Rowland and A. E. Tipping,
J. Chem. Soc., Perkin Trans. 1, 1983, 109.
8 A. T. Podkhalyuzin, V. A. Morozov and A. Z. Yankelevich, Dokl.
Akad. Nauk SSSR, 1976, 228, 609.
9 V. A. Morozov, R. A. Zamyslov and A. T. Podkhalyuzin, J. Appl.
Chem. USSR (Engl. Transl.), 1980, 53, 534–537.
10 A. T. Podkhalyuzin and M. P. Nazarova, Khim. Vys. Energ., 1979,
13, 130.
11 See I. Fleming, Frontier Orbitals and Organic Chemical Reactions,
Wiley-Interscience, London, 1976, p. 185.
12 W. B. Motherwell and D. Crich, Free Radical Chain Reactions in
Organic Synthesis, Academic Press Ltd., London, 1992, and
references therein.
13 R. D. Chambers, B. Grievson and N. M. Kelly, J. Chem. Soc., Perkin
Trans. 1, 1985, 2209.
14 W. R. Dolbier, in Organofluorine Chemistry, Fluorinated Alkenes and
Reactive Intermediates, ed. R. D. Chambers, Springer, Berlin,
1997.
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane (x ؍ 3, 4).
1,x-Bis(1,1,2,3,3,3-hexafluoropropyl)cyclohexane (x = 3, 4) 20
(5.0 g, 13 mmol), was passed through the cobalt trifluoride
reactor, at 375 ЊC, to give a colourless liquid (6.5 g). Continuous
extraction followed by preparative scale GC gave an isomeric
mixture of perfluoro-1,x-dipropylcyclohexane 53 (x = 3, 4) (3.8
g, 58%) bp 167–169 ЊC; δF (235 MHz; CDCl3; CFCl3) Ϫ80.9
(24 F, m, CF3), Ϫ109.7 to Ϫ139.4 (64 F, overlapping m, CF2),
Ϫ182.52 (2F, s, 2CF), Ϫ182.81 (2F, s, 2CF), Ϫ184.35 (2 F, s,
2CF), Ϫ184.25 (2 F, s, 2CF); m/z 481 (5%), 331 (17), 181 (30),
169 (100), 131 (42), 119 (44), 69 (87).
Crystallographic studies
Single-crystal X-ray diffraction experiments for 20a, 41a and
47 were carried out on a Rigaku AFC6S 4-circle diffrac-
tometer, using graphite-monochromated Mo-Kα radiation or
(for 47) Cu-Kα radiation. Data were collected in a 2θ/ω scan
mode (17a), ω scans with Lehmann–Larsen profile analysis
(41a) or ω scan mode (47). 20a and 47 showed a significant
decay of intensities during data collection, by 16% and 9%
respectively. For 34, the experiment was performed with a
SMART 1K CCD area detector mounted on a 3-circle dif-
fractometer (Mo-Kα radiation). Four sets of ω scans, each at
different φ and/or 2θ angles, nominally covered over a hemi-
sphere of reciprocal space. A Cryostream (Oxford Cryosys-
tems) open-flow N2 gas cryostat was used for low temperature
experiments (34, 41a). The structures were solved by direct
methods and refined by full-matrix least squares against F2 of
all data, using SHELXTL software.23 All crystals gave exceed-
ingly weak high-angle reflections and were prone to cracking
on cooling, hence rather low resolution. Compound 47 did
not give satisfactory data for atomic resolution, although the
15 H. Muramatsu, H. Kimoto and K. Inukai, Bull. Chem. Soc. Jpn.,
1969, 42, 1155.
16 J. C. Walton, Chem. Soc. Rev., 1992, 105.
17 M. Wilson, personal communication, 1999.
18 I. T. Horvarth, Acc. Chem. Res., 1998, 31, 641.
19 G. B. Barlow and J. C. Tatlow, J. Chem. Soc., 1952, 4695.
20 R. D. Chambers, B. Grievson, F. G. Drakesmith and R. L. Powell,
J. Fluorine Chem., 1985, 29, 323.
21 A. T. Podkhalyuzin, V. V. Vikulin, V. A. Morozov, M. P. Nazarova
and I. V. Vereshchinskii, Radiat. Eff., 1977, 32, 9.
22 W. H. Lin and R. J. Lagow, J. Fluorine Chem., 1990, 50, 345.
23 G. M. Sheldrick, SHELXTL, Version 5, Bruker Analytical X-Ray
Systems, Madison, WI, USA, 1995.
1638
J. Chem. Soc., Perkin Trans. 1, 2000, 1623–1638