Functionalization of saturated fluorocarbons with and without light

Article abstract of DOI:10.1016/j.jfluchem.2006.06.005

Photochemical transformation of saturated fluorocarbons into tetrabutylammonium enolates has been improved, and a method employing ketyls as reductants has been developed that accomplishes the same chemistry without light. Enolates have been isolated as enol methyl ethers, from which they can be efficiently regenerated with tetrabutylammonium iodide. In other cases, enolates have been isolated as the corresponding ketone or stable enol. Fluorocarbon LUMO energies correlate with their reactivity and serve as a guide to the choice of ketyl. Use of this chemistry for fluoropolymer surface modification is discussed.

Full text of DOI:10.1016/j.jfluchem.2006.06.005

Journal of Fluorine Chemistry 127 (2006) 1158–1167
www.elsevier.com/locate/fluor
Functionalization of saturated fluorocarbons with and without light
*
Xudong Chen, David M. Lemal
Department of Chemistry, Dartmouth College, Hanover, NH 03755, USA
Received 21 March 2006; received in revised form 2 June 2006; accepted 2 June 2006
Available online 9 June 2006
Abstract
Photochemical transformation of saturated fluorocarbons into tetrabutylammonium enolates has been improved, and a method employing
ketyls as reductants has been developed that accomplishes the same chemistry without light. Enolates have been isolated as enol methyl ethers,
from which they can be efficiently regenerated with tetrabutylammonium iodide. In other cases, enolates have been isolated as the corresponding
ketone or stable enol. Fluorocarbon LUMO energies correlate with their reactivity and serve as a guide to the choice of ketyl. Use of this chemistry
for fluoropolymer surface modification is discussed.
# 2006 Elsevier B.V. All rights reserved.
Keywords: Fluorocarbon functionalization; Photoreduction; Ketyls; Enolates; Enol ethers; b-Diketonates; Fluoropolymer modification
1. Introduction
intermediates in its reduction (Table 1). The values descend
from 1 to perfluoro-1-methylcyclohexene (2) to perfluoro-2-
methylcyclohexa-1,3-diene (3) (and its 1-methyl isomer 3a,
which may also form), indicating that reduction becomes
progressively easier until the ring is aromatized as perfluor-
otoluene (4). Then reduction becomes enough slower that it can
be interrupted if conditions are right.
Saturated fluorocarbon liquids enjoy a wide variety of uses
ranging from vapor phase soldering and heat exchange media to
breathing liquids and oxygen carriers in blood substitutes [1,2].
A common requirement of most of their applications is
chemical inertness, for which they are famous. Their spectrum
of uses could potentially be expanded in new directions if
functional groups were introduced [3,4]. The challenge to
functionalization that their inertness poses is enhanced by the
fact that conditions vigorous enough to attack them often
suffice to bring about catastrophic destruction. Because they are
highly resistant to acids, bases and oxidizing agents, powerful
electron donors are required to attack them, and the
intermediates generated in the course of reaction are more
susceptible to reduction than the starting materials. Thus, a
destructive cascade ensues with wholesale loss of fluoride ions,
as illustrated for perfluoromethylcyclohexane (1) in Scheme 1.
In the case of fluorocarbons containing six-membered rings,
like 1, the cascade can be stopped under certain conditions at
the stage when the ring is aromatized because of the relatively
strong resistance of benzene (and naphthalene) derivatives to
further reduction [5–7]. These results are understandable
in terms of the LUMO energies [8] calculated for 1 and
Although functionalization without aromatization has been
achieved in other ways [3,7,9], several years ago we developed
the only general method that introduces a functional group in
good yields and without extensive loss of fluoride into saturated
fluorocarbons having tertiary sites [10]. Its secret of success is
interception by hydroxide ion of an early intermediate in the
reduction cascade, with the formation of an enolate ion. The
negative charge on the ion protects the molecule from further
attack by electrons, as illustrated for the functionalization of
fluorocarbon 1 in Scheme 2. The reaction is carried out by
irradiating a THF solution of the fluorocarbon containing
tetrabutylammonium iodide (TBAI). Iodide ion undergoes a
charge-transfer-to-solvent (CTTS) transition [11], producing a
solvated electron for the reduction plus an iodine atom.1 The
reaction is regioselective, as the initial loss of fluoride ion
occurs exclusively at a tertiary position.
1
Other applications of photoreduction by iodide ion include the photo-
Finkelstein reaction [12] and iododefluorination of perfluoroacid esters [13].
Triiodide ion produced in all these reactions may also serve as a source of
electrons upon irradiation.
* Corresponding author. Tel.: +1 603 646 2989; fax: +1 603 646 3946.
E-mail address: david.m.lemal@dartmouth.edu (D.M. Lemal).
0022-1139/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2006.06.005

X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
1159
Scheme 1.
In that work, we found that the presence of too much water
lowered yields dramatically by hydrolyzing the enolate ion, 5
for example, to the corresponding b-diketonate ion 7 (Scheme
3). By hydrogen bonding to a fluorine in the g-position, a water
molecule assists loss of fluoride ion to give a,b-unsaturated
ketone 6, which is followed by attack of hydroxide ion and loss
of HF. For that reason, the tetrabutylammonium iodide was
dried before use, but it was assumed that reduction of residual
water was the source of the hydroxide ion necessary for making
the enolate.
2. Results and discussion
2.1. Photochemical functionalization
The first task in the present investigation was to establish
definitively the source of the oxygen in the enolate product.
Integration of the 1H NMR spectrum of a solution of TBAI in
D2O revealed that its water content was too low to provide the
necessary hydroxide ion. On the chance that air was not
excluded rigorously enough in the earlier work, and that it was
the source of the enolate oxygen (e.g., via superoxide ion), dry
Table 1
LUMO energiesa (kcal/mol) of perfluoromethylcyclohexane and degradation
products
ꢀ14.75
ꢀ51.34
ꢀ62.40
Scheme 2.
ꢀ64.86
ꢀ35.05
a
Calculated at the B3LYP/6-31G* level of theory.
Scheme 3.

1160
X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
Scheme 4.
air was bubbled into irradiated THF solutions of 1 and dried
TBAI. Yields of enolate were only 5–10%, and even that
amount may have resulted from adventitious moisture.
extrusion are shown in bold. While the bold fluorine in front has
three close neighbors, that in back has only two:
In the earlier work, nominally anhydrous magnesium sulfate
had been added to assure sufficient dryness to avoid b-diketonate
formation. We now discovered that this salt actually contained
sufficient water to account for the enolate formed! Addition of
‘‘anhydrous’’ magnesium sulfate to otherwise dry reaction
mixtures resultedin enolate yields upto 80% byNMR. Thewater
content varied even in samples of the salt from the same bottle,
and sometimes therewas enough to make b-diketonate the major
product. Obviously, a better controlled source of water was
needed. We found that magnesium sulfate monohydrate served
that purpose well, giving an 88% NMR yield of enolate 5 at 91%
conversion of fluorocarbon 1. However, the simple expedient of
adding a somewhat less than stoichiometric amount of water
directly to the reaction mixture was the best solution, affording
enolate 5 in 91% yield by NMR at complete conversion.
Enolate 5 reacted nearly quantitatively at RT with methyl
triflate to yield enol ether 8 [10,14] (Scheme 4), which was
isolated by preparative GC in 52% yield based on fluorocarbon
1. In similar fashion, perfluoromethylcyclopentane (9) was
transformed in high yield into enolate 10 [10], which was
methylated to give enol ether 11 [15]. Interestingly, b-
diketonate ion 7 failed to react at RT with methyl triflate,
revealing it to be a remarkably weak nucleophile.
The 19F NMR spectrum of 14 comprised nine sets of peaks: d
ꢀ59.1 (s, 3F, CF3); ꢀ70.4 (s, 3F, CF3); ꢀ98.5, ꢀ105.4 (ABq,
J = 307 Hz, 2F); ꢀ105.0 (dd, J = 24, 301 Hz, IF); ꢀ119.5 (d,
289 Hz, IF);ꢀ125.1 (d, 301 Hz, IF);ꢀ135.9 (d, 289 Hz, IF);
ꢀ183.9 (s, IF, C4). Thus, the three pairs of CF2 fluorines,
distinguished by their coupling constants, appeared at ꢀ98.5
and ꢀ105.4, ꢀ105.0 and ꢀ125.1, and ꢀ119.5 and ꢀ135.9.
Assignments were made with the help of NOE measurements
and are displayed in Fig. 1, which depicts the structure of 14
calculated at the B3LYP/6-31G* level of theory. Irradiation of
the tertiary F revealed interaction with one F in each of the CF2
groups, as well as the geminal CF3. This measurement sufficed
As was found in the earlier work [10], perfluoro-1,3-
dimethylcyclohexane (12) underwent functionalization only
once despite the presence of two tertiary sites. Apparently the
enolate’s negative charge suffices to protect the ion from attack at
the other vulnerable position. Surprisingly, the reaction is
regiospecific, yielding only enolate 13, which was methylated to
give 14. The only abundant species available to assist loss of
fluoride ion to form the intermediate alkene is the tetrabuty-
lammonium ion, which may be sufficiently bulky to discriminate
on steric grounds between the two possible reaction sites. The
situation is depicted in 15, where most of the fluorines have been
omitted for clarity, and where the two axial candidates for
Fig. 1. Calculated structure of l-methoxyperfluoro-2,4-dimethylcyclohexene
(14) with 19F NMR assignments.

X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
1161
to complete the assignments, and they were confirmed by the
NOEs resulting from irradiation of the two CF3 groups.
yield at 85% conversion (Eq. (2)). Before this reaction was
optimized, a better process was developed, as described below:
2.2. Functionalization without light
There are potential advantages in being able to accomplish
the chemistry described above ‘‘thermally.’’ The photoreac-
tions are quite slow, and scale-up by a large factor would be far
more practical if the reactions could be carried out without the
need for light. We therefore sought electron donors that could
substitute for the solvated electrons generated in the
photoreactions. Because light from an arc is effectively a
dilute reagent, producing solvated electrons slowly, there is
time for intermediate alkenes to be trapped by hydroxide ion
before suffering further reduction. In the presence of a
powerful electron donor in substantial concentration, however,
the lifetimes of those alkenes might be too short for trapping.
Our plan, therefore, was to find reducing agents with barely
enough power to attack the fluorocarbon, so that reaction
would be slow enough for efficient capture when the first
double bond appears.
(2)
The use of tetrabutylammonium hydroxide (TBAH) is proble-
matic for a number of reasons. Available as a 40% aqueous
solution, it can be dried by azeotropic distillation [17], but the
process consumes much time and solvent. TBAH is difficult to
obtain in completely anhydrous form. Moreover, several equiva-
lents of it are required in the functionalization reactions because
the potassium counterion of the ketyl precipitates hydroxide ion
from THF as KOH. Both the use of TBAH and the precipitation
problem could be circumvented if the ketyl counterion were
tetrabutylammonium and the necessary hydroxide ion were
generated in situ from reaction of this ketyl with water.
Pez et al. found that sodium benzophenone ketyl in THF
reduces perfluoromethylcyclohexane (1) to carbon and sodium
fluoride [6]. When we carried out the reaction of the potassium
ketyl with 1 in the presence of dry tetrabutylammonium
hydroxide (TBAH), capture of an intermediate was successful,
but at too late a stage. Only potassium perfluoro-p-cresolate
(16) was obtained, showing that the hemorrhaging of fluoride
ions was not halted until the ring had become aromatized
(Eq. (1)):
Tetrabutylammonium fluorenone ketyl (TBAFK) was easily
prepared by addition of tetrabutylammonium bromide to a THF
solution of the potassium ketyl, and the potassium bromide by-
product was removed by filtration (Scheme 5). When water was
added to a THF solution of the ketyl, its purple color disappeared
in about 5 min, showing the practicality of in situ generation of
hydroxide ion. The stoichiometry of the fluorocarbon functio-
nalization requires 1 equiv. of water and 4 equiv. of ketyl, since
2 equiv. are need to reduce the fluorocarbon and 2 equiv. serve as
a base to deprotonate the water and enol. Slow addition of the
ketylin THF to fluorocarbon1 in moist THFafforded enolate5 in
79% yield by NMR at complete conversion.
(1)
As noted above, methylation of the enolate with methyl
triflate is nearly quantitative at RT, but separation of the enol
ether from by-products and residual reagent is not easy. Though
Some commercially available aryl ketones (17–21) with their
half-wave potentials are shown in Table 2 in order of decreasing
reducing power of their radical anions. Having found the ketyl
from 17 too potent, our next attempt was with that from benzil
(21). No reaction with 1 was observed even at elevated tem-
peratures. With the requisite reducing power bracketed, the
potassium ketyl of 4-benzoylbiphenyl (18) was tried, and a
small amount of the desired enolate (5) was obtained. Finally,
the potassium ketyl of fluorenone (19) gave 5 in 69% NMR
Table 2
Half-wave reduction potentials of aryl ketones [16]
Ketone
E1/2 (V vs. SCE)
Solvent
Benzophenone (17)
4-Phenylbenzophenone (18)
Fluorenone (19)
ꢀ1.97
ꢀ1.35
ꢀ1.29
ꢀ1.20
ꢀ1.04
DMSO
46% EtOH
DMF
4,40-Dimethoxybenzil (20)
DMSO
DMSO
Benzil (21)
Scheme 5.

1162
X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
it reacts much more slowly and requires gentle heating,
dimethyl sulfate is the better choice for methylation because
product isolation is easier, not to mention that it is cheaper and
less poisonous. Volatile by-products, notably methanol, can be
removed in vacuo before the enolate reacts, and being
nonvolatile, excess reagent does not interfere with product
purification. Methyl ether 8 was isolated in yields up to 71%
(based on starting fluorocarbon) after reaction of enolate 5,
prepared as in Scheme 5, with dimethyl sulfate.
Direct isolation of a tetrabutylammonium enolate such a 5 is
not feasible because the reaction mixture contains large
amounts of tetrabutylammonium fluoride, etc., and because the
enolate is extremely susceptible to hydrolysis to b-diketonate.
Enol ether 8 is a convenient storage form for 5 from which the
enolate is quantitatively regenerated by treatment with
tetrabutylammonium iodide at RT (Eq. (3)). The ether is
easily purified and stable to water:
Scheme 7.
(3)
hyperconjugation with the five fluorines b to C2 and by the
inductive and field effects of those and other fluorines.
Reaction of enolate 5 with trimethylsilyl iodide or triflate
gave the TMS enol ether 24 [10], but trimethylsilyl chloride
(TMSCl) afforded instead two products, chloroketone 25 and
b-diketonate 7 (Scheme 7). The contrast may be understood in
terms of the weaker electrophilicity and thus greater selectivity
of TMSCl. Since hydroxide in the reaction mixture is far more
nucleophilic than the enolate, the reagent was apparently
largely hydrolyzed. This provided chloride ion and a source of
protons (trimethylsilanol or water from deprotonation of the
silanol by hydroxide) to assist loss of fluoride ion from 5
(Scheme 8). Attack on the resulting a,b-unsaturated ketone 6
by chloride ion gave 25, and by hydroxide ion gave 7 via
addition–elimination reactions. The weak nucleophilicity of
enolate 5 is further revealed in its failure to react with the good
electrophiles a-bromoacetophenone and chloroacetone.
Enolate 5 reacts readily with triflic anhydride to give enol
triflate 22 (Scheme 6). We anticipated that this molecule would
serve to introduce other functional groups at C1 via addition–
elimination reactions, given the excellence of triflate ion as a
leaving group. However, treatment of 22 with TBAI yielded,
not iodide 23, but enolate 5, testimony to the ability of the
enolate to function as a leaving group. As a quintessentially soft
nucleophile, iodide was expected to attack the relatively soft
center C1, not the hard sulfur atom. Clearly, the negative charge
on the enolate ion is very well stabilized, both by negative
The ketyl TBAFK was allowed to react with perfluor-
omethylcyclopentane (9), giving enolate 10 in 80% yield by
NMR (Eq. (4)). Reduction of this substrate was notably faster
than that of 1, suggesting a lower LUMO energy. At the B3LYP/
6-31G* level of theory, it is in fact lower by 5.3 kcal/mol
(Table 3). Methylation of enolate 10 with dimethyl sulfate
afforded methyl ether 11 in 55% isolated yield, based on 9:
(4)
Scheme 6.

X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
1163
Scheme 8.
Surprisingly, reaction of 1,3-dimethylcyclohexane (12) with
TBAFK in moist (or dry) THF gave a very different result from
the photochemical functionalization of 12 described above. The
product was the tetrabutylammonium salt of allylic anion 26, a
stable, weakly basic species, the origin and nature of which has
been discussed elsewhere (Eq. (5)) [18]:
(5)
Table 3
Acyclic substrate perfluoro-2-methylpentane (27) was not
reduced by TBAFK. Perfluorocycloalkanes are well known
to have higher electron affinities than their acyclic counterparts
[19], and indeed the LUMO energy calculated for 27,
ꢀ3.28 kcal/mol, lies well above those of both 1 and 9. We
therefore attempted to prepare the tetrabutylammonium ketyl
of 4-benzoylbiphenyl (18), which would have more reducing
power than TBAFK (Table 1). Unfortunately, the ketyl anion
was apparently basic enough to bring about Hoffmann elim-
ination on the tetrabutylammonium ion, as the deep blue color
of the potassium ketyl faded when tetrabutylammonium bro-
mide was introduced. Functionalization of fluorocarbon 27 was
successfully accomplished nonetheless using the potassium
ketyl (KBBK) of 18. TBAH was added to a THF solution of
the fluorocarbon, and the ketyl was introduced slowly using a
syringe pump. Enolate 28 was obtained in 73% yield by NMR,
and acidification with concentrated sulfuric acid gave ketone 29
[20], isolated in 62% yield based on 27 (Scheme 9). Monitoring
LUMO energies (kcal/mol) of some saturated fluorocarbons calculated at the
B3LYP/6-31G* level of theory
ꢀ14.75
ꢀ20.05
ꢀ3.28
ꢀ10.09
ꢀ16.63
ꢀ17.77
Scheme 9.

1164
X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
of the reaction revealed that the enolate precipitated initially as
the potassium salt, only to slowly return to solution as the
tetrabutylammonium salt. The cation exchange was presum-
ably driven by formation of potassium fluoride.
methods have been developed to improve adhesion. These
techniques include treatments with plasmas [23], X-ray [24],
ion beam [25], and UV irradiation [26]; electrochemical reduc-
tion [27]; wet chemical etching with solvated electrons in liquid
ammonia [28], sodium naphthalene [29], and benzoin dianion
[30]. All these methods damage the surface in non-selective
ways.
That different substrates require different reducing power is
clear from the above examples, and the array of ketyls that are
available gives the method flexibility. Since the calculated
LUMO energies of the fluorocarbon substrates we have studied
correlate nicely with their reactivity under reducing conditions,
one should be able to predict the reactivity of other substrates
with tertiary fluorines and choose a ketyl accordingly.
Fluorocarbons having tertiary sites and thus relatively weak
C–F bonds are easier to reduce than those with only primary
and secondary fluorines, indicating that bond strength as well as
LUMO energy influence the ease of reduction. LUMO energies
may well correlate with reactivity for fluorocarbons lacking
tertiary fluorine, such as the series perfluorocyclohexane (30),
perfluorocyclopentane (31) and perfluorocyclobutane (32)
(Table 3), but with a different calibration than for fluorocarbons
with tertiary sites. Thanks to its very low lying LUMO,
perfluorocyclobutane was readily reduced by TBAFK (Eq. (6)).
Protonation of the resulting enolate (33) gave the enol (34) [21],
which we found some years ago to be lower in energy than its
keto form [22]. Under our conditions, perfluorocyclohexane
(30) was reduced very slowly if at all even by potassium
benzophenone ketyl.
Teflon FEP is a random copolymer of tetrafluoroethylene
and hexafluoropropene, typically containing ꢁ8.5 mol% of the
latter component [31]. There is a tertiary fluorine at each of the
branch points, and our selective reduction chemistry should be
capable of introducing b-diketonate functionality precisely at
these sites either photochemically or without light (Scheme 10).
The resulting polymer with an ionic surface might find some
applications. Its wettability and affinity for various coating
materials would be enhanced, but the prospect of exploiting the
specific functionality that has been introduced is more
interesting. In principle, the b-diketonate moieties would be
capable of chelating an array of metals. Depending upon the
choice of metal and other ligands, it should be possible to
modify the properties of the surface in various ways. One could
imagine making the surface fluoresce or function as a catalyst,
for example.
Teflon FEP film was treated with potassium 4-benzoylbi-
phenyl ketyl (KBBK) in the presence of TBAH, resulting in a
dramatic increase in its wettability. Teflon FEP powder, treated
similarly, displayed new strong, broad infrared absorption at
ꢁ1600 cmꢀ1 and new absorptions peaking at 2969 cmꢀ1
.
Bands in the latter region are attributable to the tetrabuty-
lammonium ion, the counterion of the b-diketonate groups,
which are responsible for the ꢁ1600 cmꢀ1 absorption. For
comparison, the IR spectrum of tetrabutylammonium b-
diketonate 7 showed absorptions at 2971 and 1606 cmꢀ1
(strong and broad).
(6)
Preliminary attempts to prepare metal chelates from the
treated film and powder have been unavailing, however.
Fluorinated b-diketonates are known to form complexes with a
variety of metals [32], but perhaps the polymer-bound b-
diketonate moieties exist primarily in the W (35) or skew (36)
configurations instead of the U-shape (37) that is required for
The inertness and low surface energy of fluoropolymers make
them resistant to bonding with other substances, so a variety of
Scheme 10.

X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
1165
chelation. We believe that selective surface functionalization of
fluoropolymers by our reduce-and-capture method, or a
variation thereof, merits further investigation.
The whole assembly was immersed in a large beaker, into
which cooling water was introduced continuously. A large pan
with a draining hose caught the overflow. The mixture was
irradiated for 48 h, while it was stirred with a horseshoe magnet
attached to a stirring motor. The magnet was mounted outside
the beaker, close to the quartz tube. Formation of perfluoro-2-
methylcyclohex-1-enolate (5) was observed and quantitated by
19F NMR. Yield: 91%.
Tetramethylenesulfone (5 mL) was added to the mixture,
and THF and other volatile materials were removed in vacuo.
Methyl triflate (300 mL, 2.65 mmol) was added, together with
CaCO3 (0.50 g, 5 mmol). The mixture was stirred for 0.5 h,
then methyl ether 8 was vacuum transferred to a U-trap cooled
by liquid nitrogen. The crude product was purified by
preparative GC. Yield: 0.217 g, 52% based on fluorocarbon
3. Conclusions
The photochemical method developed earlier in our
laboratory for functionalizing saturated fluorocarbons as
tetrabutylammonium enolates has been refined, and derivatives
of the enolates have been isolated. The enolates are highly
susceptible to hydrolysis, forming b-diketonates, but their
methyl ethers serve as stable storage forms from which they can
be regenerated cleanly in excellent yield. The same functio-
nalization reactions have been accomplished without light,
utilizing ketyls with variable reducing power. Fluorocarbon
LUMO energies correlate with ease of reaction, and guide the
choice of ketyl to be employed with a particular substrate.
Infrared spectra indicate that b-diketonate ions have been
introduced on the surface of Teflon FEP using this chemistry.
1
1. H NMR (CDCl3): d 4.18 (t, J = 1.8 Hz, CH3). 19F NMR
(CDCl3): d ꢀ59.2 (s, 3F, CF3), ꢀ108.3 (s, 2F, C3), ꢀ114.9 (s,
2F, C6), ꢀ134.7 (s, 2F, C4 or 5), ꢀ134.9 (s, 2F, C4 or 5). IR
(neat, cmꢀ1): 2977, 1657, 1470, 1386, 1255, 1198, 1030, 997.
4.1.2. Without light
The following operations were done under nitrogen
protection, and glassware was dried in an oven at 120 8C
overnight. Tetrabutylammonium bromide (3.87 g, 12 mmol),
fluorenone (2.16 g, 12 mmol), potassium (0.48 g, 12 mmol),
and THF (40 mL) were added to a 60 mL dropping funnel with
a fritted filter plate. The mixture was magnetically stirred
overnight. Then nitrogen pressure was applied to force
tetrabutylammonium fluorenone ketyl solution through the
filter plate into a 50 mL flask. Potassium bromide was thus
removed by filtration. The deep red ketyl solution was
transferred to a 50 mL syringe. Water (50 mL, 2.8 mmol)
and THF (10 mL) were placed in a 100 mL two-neck round-
bottom flask, and one-fourth of the ketyl solution (10 mL) was
added. After the mixture had been stirred for 0.5 h,
perfluoromethylcyclohexane (1) (0.70 g, 2.0 mmol) was added
to the mixture. The remaining ketyl was slowly added to the
above flask over the course of 8 h, controlled by a syringe
pump. Formation of perfluoro-2-methylcyclohex-1-enolate (5)
was observed and quantitated by NMR. Yield: 79%.
4. Experimental
NMR spectra were taken on Varian Unity Plus 300 and
500 MHz machines. 19F NMR spectra were recorded at 282.2
and 470.3 MHz, using an internal standard of trichlorofluor-
omethane. NMR yields were determined using a,a,a-trifluor-
otoluene as an internal area standard. 13C NMR spectra were
recorded at 125.7 MHz. GC was performed on a Hewlett-
Packard 5750 gas chromatograph, and GC/MS on a Hewlett-
Packard 5890A gas chromatograph with a Hewlett-Packard
5971 series mass detector. IR spectra were measured on a
Perkin-Elmer 599 FT-IR instrument. The UV source for
photoreaction was a Vycor-filtered 450W Hanovia medium-
pressure mercury arc contained in a quartz water jacket. The
high resolution mass spectrum was obtained from the Mass
Spectrometry Center, University of Massachusetts, Amherst,
MA.
Dimethyl sulfate (10 mL) was added to the mixture. THF
and other volatile materials were removed by vacuum transfer.
The mixture was then heated to 45 8C, and stirred for 10 h under
vacuum. Methyl ether 8 was collected in a U-trap cooled by
liquid nitrogen. This crude product was further purified by
preparative GC. Yield: 0.36 g, 56% based on fluorocarbon 1.
THF was distilled from potassium benzophenone ketyl, and
acetonitrile from calcium hydride. Preparation of perfluoro-2,4-
dimethylcyclohex-1-enolate (13) and its methyl ether (14) is
described elsewhere [18].
4.2. Perfluoro-2-methylcyclohex-1-enolate (5) from methyl
ether 8
4.1. Tetrabutylammonium perfluoro-2-methylcyclohex-1-
enolate (5) and 1-methoxyperfluoro-2-methylcyclohexene
(8) [10]
2-Methoxyperfluoro-1-methylcyclohexene (8) (30 mg,
0.093 mmol), tetrabutylammonium iodide (35 mg, 1 equiv.),
and THF (0.5 mL) were added to a flame-dried NMR tube.
Tetrabutylammonium perfluoro-2-methylcyclohex-1-enolate (5)
was generated quantitatively, as determined by 19F NMR. 19F
NMR (THF): d ꢀ55.6 (s, 3F, CF3), ꢀ93.8 (s, 2F, C3), ꢀ122.0 (s,
2F, C6), ꢀ133.5 (s, 2F, C4 or 5), ꢀ135.7 (s, 2F, C4 or 5).
4.1.1. Photochemically
Perfluoromethylcyclohexane (1) (0.77 g, 2.2 mmol), water
(25 mg, 1.4 mmol), tetrabutylammonium iodide (2.0 g,
5.4 mmol), and THF (10 mL) were added to a 25 mL quartz
tube. It was sealed with a septum and attached to the UV source.

1166
X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
4.3. 2-Trifluoromethanesulfonyloxyperfluoro-1-
methylcyclohexene (22)
4.6. Perfluoro-2-methylpent-2-en-3-olate (28) and
2H-Perfluoro-2-methyl-3-pentanone (29) [10]
To the above solution in the NMR tube was added triflic
anhydride (27 mg, 0.095 mmol). Triflate derivative 22 was
formed in 88% yield, determined by NMR. 19F NMR (THF):
d ꢀ59.5 (s, 3F, CF3C), ꢀ79.2 (s, 3F, CF3S), ꢀ108.2 (s, 2F,
C3), ꢀ115.2 (s, 2F, C6), ꢀ135.3 (s, 2F, C4 or 5), ꢀ135.7
(s, 2F, C4 or 5).
Under nitrogen protection, potassium (0.80 g, 20 mmol),
THF (40 mL), and p-phenylbenzophenone (5.16 g, 20 mmol)
were placed in a 100 mL round-bottom flask. After the
mixture had been stirred overnight, it was transferred to a
50 mL syringe. Tetrabutylammonium hydroxide solid
(2.08 g, 8.0 mmol, prepared from the commercial 40%
aqueous solution by azeotropic concentration using benzene,
3 ꢂ 10 mL), perfluoro-2-methylpentane (27) (0.67 g,
2.0 mmol), and THF (10 mL) were added to a 100 mL
round-bottom flask. The ketyl solution was added dropwise
via syringe pump to the above mixture over the course of 8 h.
Formation of perfluoro-2-methylpent-2-en-3-olate (28) was
quantitated by NMR. Yield: 81%.
Tetrabutylammonium iodide (70 mg, 0.19 mmol) was added
to the above solution, and periluoro-2-methylcyclohex-1-
enolate (5) was regenerated. 19F NMR (THF): d ꢀ55.6 (s,
3F, CF3), ꢀ93.8 (s, 2F, C3), ꢀ121.9 (s, 2F, C6), ꢀ133.4 (s, 2F,
C4 or 5), ꢀ135.6 (s, 2F, C4 or 5).
4.4. 3-Chloroperfluoro-2-methylcyclohex-2-enone (25)
Freshly distilled triglyme (10 mL) was added to the
mixture, and THF was removed under vacuum. Concentrated
sulfuric acid (10 mL) was added, and the resulting 2H-
perfluoro-2-methyl-3-pentanone (29) was vacuum transferred
to a U-trap cooled by liquid nitrogen. Yield: 0.37 g, 62% based
on 27. 1H NMR (CDCl3): d 4.71 (m, 1H). 19F NMR (CDCl3): d
ꢀ63.4 (narrow m, 6F, gem CF3s), ꢀ81.8 (s, 3F, C5), ꢀ122.7
(s, 2F, C4).
Prepared as above from methyl ether 8, perfluoro-2-
methylcyclohex-1-enolate (5, ca. 0.32 mmol) in 10 mL THF
was added to a 25 mL oven-dried round-bottom flask.
Tetramethylenesulfone (4 mL) was added, and then THF was
removed under vacuum. Trimethylsilyl chloride (1 mL, freshly
distilled) was added, and the mixture was stirred for 0.5 h.
Volatile materials were vacuum transferred to a U-trap cooled
by liquid nitrogen. The resulting crude 3-chloroperfluoro-2-
methylcyclohex-2-enone (25) was purified by preparative GC.
19F NMR (CDCl3): d ꢀ60.2 (s, 3F, CF3), ꢀ112.2 (s, 2F),
ꢀ127.5 (s, 2F), ꢀ133.6 (s, 2F). 13C NMR (CDCl3): d 173.5
(C1); 149.5 (C3); 129.7 (C2); 119.6 (CF3); 108.2, 108.0, 105.5
(C4-6). IR (neat, cmꢀ1): 1751, 1610, 1333, 1291, 1258, 1179,
1068, 992, 902, 835. HRMS: 305.9486 found, 305.9494
calculated for C7F9ClO.
4.7. Perfluorocyclobut-1-enol (33)
The following operations were done under nitrogen, and
glassware was dried in an oven at 120 8C overnight.
Tetrabutylammonium bromide (3.87 g, 12 mmol), fluorenone
(2.16 g, 12 mmol), potassium (0.48 g, 12 mmol), and THF
(40 mL) were added to a 60 mL dropping funnel with a
fritted filter plate. The mixture was magnetically stirred
overnight, then nitrogen pressure was applied to force the
tetrabutylammonium fluorenone ketyl solution through the
filter plate into a 50 mL flask. The deep red ketyl solution
was transferred to a 50 mL syringe. Water (50 mL, 2.8 mmol)
was placed in a 100 mL two-neck round-bottom flask
together with THF (10 mL) and one-fourth of the ketyl
solution (10 mL). After the mixture had been stirred for
0.5 h, it was cooled to ꢀ31 8C in a bromobenzene/Dry Ice
bath. Perfluorocyclobutane (32, 0.38 g, 1.9 mmol) was
condensed into the mixture from a balloon, the quantity
determined by the weight difference before and after the
condensation. The remaining ketyl was slowly added to the
above flask via syringe pump over the course of 5 h.
Formation of perfluorocyclobut-1-enolate was observed by
NMR.
4.5. Perfluoro-2-methylcyclopent-1-enolate (10) and 1-
methoxyperfluoro-2-methylcyclopentene (11)
4.5.1. Photochemically
According to the procedure for fluorocarbon 1, perfluor-
omethylcyclopentane (9) (0.500 g, 1.67 mmol) was photo-
reduced to perfluoro-2-methylcyclopent-1-enolate (10) [10] in
80% yield, measured by NMR. Perfluoro-2-methylcyclopent-1-
enol methyl ether (11) was obtained by treating 10 with methyl
triflate, as described for enolate 5, and the crude product (with
ca. 10% mechanical loss) was purified by preparative GC.
1
Yield: 0.176 g, 43% based on 9. H NMR (CDCl3): d 4.19
(narrow m, CH3). 19F NMR (CDCl3): d ꢀ59.1 (s, 3F, CF3),
ꢀ106.9 (s, 2F, C3), ꢀ114.2 (s, 2F, C5), ꢀ130.2 (s, 2F, C4). IR
(neat, cmꢀ1): 1678, 1480, 1416, 1296, 1155, 1059, 1005, 931,
875.
1,2,4-Trichlorobenzene (40 mL) was added, then THF and
other volatile materials were removed by vacuum transfer.
Concentrated sulfuric acid (10 mL) was added to the mixture.
With vigorous stirring, perfluorocyclobut-1-enol was vacuum
transferred into a series of two traps. The first trap at ꢀ13 8C
(ethylene glycol/Dry Ice bath) collected solvent; the second
trap, cooled to ꢀ78 8C (acetone/Dry Ice), collected the enol
(0.21 g, 69%). 19F NMR (CDCl3): d ꢀ116.2 (m, 2F), ꢀ120.3
(m, 2F), ꢀ141.8 (m, IF).
4.5.2. Without light
Perfluoromethylcyclopentane (9) (0.60 g, 2.0 mmol) was
functionalized as described for 1, and the formation of
perfluoro-2-methylcyclopent-1-enolate (10) was quantitated
by NMR. Yield: 80%. Treatment with dimethyl sulfate at 35 8C
according to the procedure for 5 gave pure methyl ether 11.
Yield: 0.30 g, 55% based on 9.

X. Chen, D.M. Lemal / Journal of Fluorine Chemistry 127 (2006) 1158–1167
chemistry, vol. 1, CRC Press, Cleveland, OH, 1977, pp. 554–555;
1167
Acknowledgement
For 20: L. Meites, P. Zuman, CRC Handbook Series in Organic Electro-
chemistry, vol. 1, CRC Press, Cleveland, OH, 1977, pp. 684–685;
For 21: L. Meites, P. Zuman, CRC Handbook Series in Organic Electro-
chemistry, vol. 1, CRC Press, Cleveland, OH, 1977, pp. 606–607.
[17] T. Hasegawa, H. Yamamoto, Synlett (1999) 84–86.
[18] X. Chen, D.M. Lemal, J. Org. Chem. 69 (2004) 8205–8208.
[19] J.F. Liebman, J. Fluorine Chem. 3 (1973) 27–33.
[20] T. Martini, C. Schumann, J. Fluorine Chem. 8 (1976) 535–540.
[21] R.A. Bekker, V.Ya. Popkova, I.L. Knunyants, Dokl. Akad. Nauk SSSR 229
(1976) 870;
The National Science Foundation is gratefully acknowl-
edged for support of this research.
References
[1] R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.), Organofluorine Chemistry,
Principles and Commercial, Applications, Plenum, New York, 1994.
[2] D.M. Lemal, J. Org. Chem. 69 (2004) 1–11.
R.A. Bekker, V.Ya. Popkova, I.L. Knunyants, Dokl. Akad. Nauk SSSR,
Engl. Transl. 229 (1976) 514.
[3] J. Burdeniuc, W. Chupka, R.H. Crabtree, J. Am. Chem. Soc. 117 (1995)
10119–10120.
[22] P.E. Lindner, R.A. Correa, J. Gino, D.M. Lemal, J. Am. Chem. Soc. 118
(1996) 2556–2563.
[4] J. Burdeniuc, B. Jedlicka, R.H. Crabtree, Chem. Ber./Recueil 130 (1997)
145–154;
[23] Y. Yamada, T. Yamada, S. Tasaka, N. Inagaki, Macromolecules 29 (1996)
4331–4339;
G. Saunders, Angew. Int. Ed. 35 (1996) 2615–2617.
[5] K. Sung, R.J. Lagow, J. Chem. Soc., Perkin Trans. I (1998) 637–638;
J. Burdeniuc, R.H. Crabtree, Science 271 (1996) 340–341;
J.L. Kiplinger, T.G. Richmond, J. Am. Chem. Soc. 118 (1996) 1805–1806;
J.L. Kiplinger, T.G. Richmond, C.E. Osterberg, Chem. Rev. 94 (1994)
373–431;
J.P. Badey, E. Espuche, D. Sage, B. Chabert, Polymer 37 (1996) 1377–
1386;
D. Youxian, H.J. Griesser, A. Mau, R. Schmidt, Polymer 32 (1991) 1126–
1130.
[24] R.R. Rye, G.W. Arnold, Langmuir 5 (1989) 1331–1334.
[25] S.K. Koh, S.C. Park, S.R. Kim, W.K. Choi, H.J. Jung, K.D. Pae, J. Appl.
Polym. Sci. 64 (1997) 1913–1921;
B. Gething, C.R. Patrick, M. Stacey, J.C. Tatlow, Nature 183 (1959) 588–
589.
[6] J.A. Marcella, A.G. Gilicinski, A.M. Coughlin, G.P. Pez, J. Org. Chem. 57
(1992) 2856–2860.
C.A. Chang, J.E. Baglin, A.G. Schrott, K.C. Lin, Appl. Phys. Lett. 51
(1987) 103–105;
[7] J. Burdeniuc, R.H. Crabtree, J. Am. Chem. Soc. 118 (1996) 2525–2526.
[8] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R.
Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E. Stratman, J.C.
Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain,
O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala,
Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B.
Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith,
M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challa-
combe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, M.
Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian ’98 Revision A.5,
Gaussian, Inc., Pittsburgh, PA, 1998.
C.A. Chang, C.J. Chan, F. Jones, Appl. Phys. Lett. 59 (1991) 1069–1071.
[26] C. Girardeaux, Y. Idrissi, J.J. Pireaux, R. Caudano, Appl. Surf. Sci. (1996)
586–590;
H. Niino, A. Yabe, Appl. Surf. Sci. (1996) 550–557;
J. Heitz, H. Niino, A. Yabe, Appl. Phys. Lett. 68 (1996) 2648–2650;
K. Allmer, A.E. Feiring, Macromolecules 24 (1991) 5487–5488;
I. Noh, K. Chittur, S.L. Goodman, J.A. Hubbell, J. Polym. Sci. A 35 (1997)
1499–1514.
[27] C. Amatore, C. Combellas, F. Kanoufi, C. Sella, A. Thiebault, L. Thouin,
Chem. Eur. J. (2000) 820–835;
L. Kavan, J. Klima, M. Pserdlova, Carbon 23 (1985) 45–49;
Z. Pelzbauer, J. Baldrian, J. Jansta, F.P. Dousek, Carbon 17 (1979) 317–
322.
[9] J. Burdeniuc, R.H. Crabtree, Organometallics 17 (1998) 1582–1586;
N.A. Kaprinidis, N.J. Turro, Tetrahedron Lett. 37 (1996) 2373–2376.
[10] N.S. Stoyanov, N. Ramchandani, D.M. Lemal, Tetrahedron Lett. 40
(1999) 6549–6552.
[28] K.B. Brace, C. Combellas, E. Dujardin, A. Thiebault, M. Delamar, F.
Kanoufi, M. Shanahan, Polymer 38 (1997) 3295–3305;
N. Chakrabarti, J. Jacobus, Macromolecules 21 (1988) 3011–3014.
[29] R.R. Rye, J. Polym. Sci. B, Polym. Phys. 26 (1988) 2133–2144.
[30] H. Sakurai, Y. Kubo, M. Shiotani, H. Yahiro, Y. Okuda, J. Appl. Polym.
Sci. 74 (1999) 286–289;
[11] M.F. Fox, E. Hayon, Trans. Faraday Soc. 1 72 (1976) 1990–1996;
M.J. Blandamer, M.F. Fox, Chem. Rev. 70 (1970) 59–93.
[12] Y. Zhang, J.R. Smith, D.M. Lemal, J. Am. Chem. Soc. 118 (1996) 9454–
9455.
C. Costello, T.J. McCarthy, Macromolecules 20 (1987) 2819–2828;
Z. Iqbal, D.M. Ivory, J.S. Szobota, R.L. Elsenbaumer, R.H. Baughman,
Macromolecules 19 (1986) 2992–2996;
[13] X. Chen, D.M. Lemal, J. Fluorine Chem., in press.
[14] P.A. Carter, C.R. Patrick, J.C. Tatlow, J. Fluorine Chem. 21 (1982) 407–411.
[15] V.F. Snegirev, K.N. Makarov, Izv. Akad. Nauk SSSR, Ser. Khim. (1986)
1331–1340.
C. Costello, T.J. McCarthy, Macromolecules 18 (1985) 2087–2088;
C. Costello, T.J. McCarthy, Macromolecules 17 (1984) 2940–2942.
[31] W.W. Schmiegel, in: M. Hudlicky, A.E. Pavlath (Eds.), Chemistry of
Organic Fluorine Compounds, vol. II, American Chemical Society,
Washington, DC, 1995, pp.1101–1118.
[16] For 17: L. Meites, P. Zuman, CRC Handbook Series in Organic Electro-
chemistry, vol. 1, CRC Press, Cleveland, OH, 1977, pp. 564–565;
For 18: L. Meites, P. Zuman, E.B. Rupp, CRC Handbook Series in Organic
Electrochemistry, vol. 3, CRC Press, Cleveland, OH, 1977, pp. 306–307;
For 19: L. Meites, P. Zuman, CRC Handbook Series in Organic Electro-
[32] B. Croxtall, J. Fawcett, E.G. Hope, A.M. Stuart, J. Fluorine Chem. 119
(2003) 65–73.