Synthesis of α-fluorocarboxylates from the corresponding acids using acetyl hypofluorite

Article abstract of DOI:10.1021/jo010677k

α-Fluorocarboxylic esters and acids were synthesized in good yields. The corresponding esters and acids were converted to their ketene acetals, and these enol derivatives reacted with AcOF made directly from fluorine. This route circumvents the problems associated with nucleophilic fluorinations such as various eliminations and rearrangements. α- and β-branched carboxylic acid derivatives that cannot be directly fluorinated gave by this electrophilic fluorination the corresponding α-fluoro derivatives in good yield. Both the fluorination reaction and the preparation of AcOF are fast and suitable for [18]F incorporation into acids and esters needed for working with PET. α-Fluoroibuprofen (20) and methyl 2-fluoro-3,3,3-triphenylpropionate (32) are two examples of this general reaction.

Full text of DOI:10.1021/jo010677k

7464
J . Org. Chem. 2001, 66, 7464-7468
Syn th esis of r-F lu or oca r boxyla tes fr om th e Cor r esp on d in g Acid s
Usin g Acetyl Hyp oflu or ite
Shlomo Rozen,* Aviv Hagooly, and Rinat Harduf
School of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel-Aviv University,
Tel-Aviv 69978, Israel
rozens@post.tau.ac.il
Received J uly 5, 2001
R-Fluorocarboxylic esters and acids were synthesized in good yields. The corresponding esters and
acids were converted to their ketene acetals, and these enol derivatives reacted with AcOF made
directly from fluorine. This route circumvents the problems associated with nucleophilic fluorinations
such as various eliminations and rearrangements. R- and â-branched carboxylic acid derivatives
that cannot be directly fluorinated gave by this electrophilic fluorination the corresponding R-fluoro
derivatives in good yield. Both the fluorination reaction and the preparation of AcOF are fast and
suitable for [18]F incorporation into acids and esters needed for working with PET. R-Fluoroibu-
profen (20) and methyl 2-fluoro-3,3,3-triphenylpropionate (32) are two examples of this general
reaction.
Sch em e 1
The biological importance of fluorine-containing com-
pounds is already common knowledge and underlined in
practically any introduction of the relevant papers.
Fluorocarboxylic acids are no exception. The interest in
these compounds first arose in the 1950s when it was
found that straight-chain ω-fluoro acids with an even
number of carbons are almost as toxic as fluoroacetic acid
itself.1 Since that time, many papers appeared dealing
with the synthesis and biological importance of R-fluo-
rocarboxylic acids. The main method of preparation was
based on the displacement of a leaving group R to the
carboxylate moiety by a nucleophilic fluoride usually in
a form of KF (Scheme 1).2 Because of the basicity of the
reagent, however, such an approach results, along with
the desired R-fluoroacids, in elimination reactions or
various rearrangements. These two types of side reac-
tions become dominant whenever R- or â-branched car-
boxylic acids are the target molecules. Such limitations
encouraged alternative, multistep routes, such as open-
ing epoxides,3 reacting R-hydroxy acids with DAST4 or
Ishikawa reagent,5 reacting alkenes with [BrF] agents,
prepared from HF and a source of electrophilic bromine,6
or building the target R-fluoro acid by combining two
fragments, one of which already contains the fluorine
atom (Scheme 2).7
Sch em e 2
* To whom correspondence should be addressed. Fax: +972 3
6409293.
(1) (a) Pattison, F. L. M.; Norman, J . J . J . Am. Chem. Soc. 1957,
79, 2311. (b) Goldberg, R.; Pierron, A. P. M.; Catesson, A. M.; Czaninski,
Y.; Francesch, C.; Rolando, C. Phytochemistry 1988, 27, 1647. (c)
Leeper, F. J .; Rock, M. J . Fluorine Chem. 1991, 51, 381. (d) PCT Int
patent application WO 99 67,199; Chem. Abstr. 2000, 132, 49751d.
(2) (a) Bergmann, E. D.; Blank, I. J . Chem. Soc. 1953, 3786. (b)
Shahak, I.; Rozen, S.; Bergmann, E. D. Isr. J . Chem. 1970, 8, 763. (c)
Langhals, E. F. Tetrahedron: Asymmetry 1994, 5, 981.
(3) (a) Keul, H.; Pfeffer, B.; Griesbaum, K. Chem. Ber. 1984, 117,
2193. (b) Amanetoullah, A. O.; Chaabouni, M. M.; Baklouti, A. Synth.
Commun. 1996, 26, 1155.
(4) Schlosser, M.; Michel, D. Tetrahedron 1996, 52, 8257.
(5) Watanabe, S.; Fujita, T.; Sakamoto, M.; Endo, H. J . Fluorine
Chem. 1990, 47, 187.
(6) (a) Pattison, F. L. M.; Buchanan, R. L.; Dean, F. H. Can. J . Chem.
1965, 43, 1700. (b) Goj, O.; Kotila, S.; Haufe, G. Tetrahedron 1996,
52, 12761.
Fluorination on anionic centers R to a carboxylate,
using several “NF” reagents such as N-fluorobis[(trifluo-
10.1021/jo010677k CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/04/2001

Synthesis of R-Fluorocarboxylates
J . Org. Chem., Vol. 66, No. 22, 2001 7465
romethyl)sulfonyl]imide,8 N-fluorosultams,9 and Select-
fluor,10 have also been carried out. Such reactions are still
very few and require anhydrous conditions. The reagents,
prepared from F2, have to be isolated and purified in
order to function properly. What is more, because of the
relatively long fluorination reaction times, as well as the
time required for the reagent preparation, none of the
above methods is suitable for the incorporation of the
[18]F isotope in biologically interesting molecules needed
for work with positron-emitting tomography (PET).
Sch em e 3
Clearly, an additional method that does not suffer from
the secondary reactions associated with the basic fluoride
is desirable. Such a route should also shorten the multi-
step construction of both reagents and starting materials
needed to obtain the R-fluoro acids.
Acetyl hypofluorite (AcOF) was developed by us 20
years ago.11 Solutions of this reagent are easily prepared
by passing commercially available nitrogen-diluted fluo-
rine through a suspension of sodium acetate in either
trichlorofluoromethane or acetonitrile. The reagent solu-
tion is stable for a few hours at 0-20 °C and does not
need any isolation or purification prior to its use.12 In a
sense, it is a mild carrier of F2. Because AcOF possesses
an electrophilic fluorine, all the disadvantages associated
with nucleophilic fluorination are eliminated. Since this
reagent reacts smoothly with most kinds of double
bonds,13 and since we have had a successful experience
with electrophilic hydroxylations of trimethylsilyl ketene
acetals using HOF‚CH3CN complex,14 we decided to
investigate the reaction of these enol derivatives with
AcOF.
Methyl 2-fluorophenylacetate (3) is easily obtained by
several methods (see, for example, refs 2c and 3b) so we
chose it as a test case. Methyl phenylacetate (1) was
converted to its known trimethylsilyl ketene acetal (2)15
and reacted with AcOF for a few minutes forming 3 in
90% yield. Nonbenzylic carboxylic acids are also suitable,
and methyl octanoate (4) or methyl 4-cyclohexylbu-
tanoate (7) can serve as examples. Their trimethylsilyl
ketene derivatives (5) and (8) are easy to make, and
consecutive reactions with AcOF produce methyl 2-fluo-
rooctanoate (6)16 and the unknown methyl 2-fluoro-4-
cyclohexylbutanoate (9) in high yields (Scheme 3). It is
worth noting that the enolic double bond of 8 reacts much
faster than the tertiary hydrogen, which is a potential
reaction center whenever electrophilic fluorine is in-
volved.17
fluorination.2b By using AcOF, we synthesized in 90%
yield methyl 2-propyl-2-fluoropentanoate (12) from meth-
yl 2-propylpentanoate (10), which was first converted to
its trimethylsilyl ketene acetal 11. These types of com-
pounds seem to be very promising anticonvulsants and
antiepileptics.1d Methyl cycloheptanoate (13) was also
reacted in this manner, and the reaction between its
trimethylsilyl ketene acetal 14 and AcOF produced the
novel methyl R-fluorocycloheptanoate (15). Similar re-
sults were obtained with methyl 2-phenylbutanoate (16),
whose trimethylsilyl ketene acetal (17) was reacted with
acetyl hypofluorite to form methyl 2-fluoro-2-phenylbu-
tanoate (18).
Ibuprofen, 2-(4-isobutylphenyl)propionic acid (19), is
a powerful analgetic drug. Schlosser prepared its R-fluoro
derivative 20 by a multiple-step reaction, starting with
4-isobutylphenyl methyl ketone, which was cyanohy-
drated, and after a few additional steps the fluorine atom
was introduced with DAST.4 Our approach needed only
two steps, so methyl 2-fluoro-2-(4-isobutylphenyl) propi-
onate (23) was obtained in 70% yield by reacting AcOF
with the trimethylsilyl ketene acetal 22, easily made from
the parent methyl 2-(4-isobutylphenyl) propionate (21)
(Scheme 4).
Apart from R-branched carboxylic acids that cannot be
fluorinated at the R position by conventional nucleophilic
fluorination, compounds branched at the â-position are
equally resistant because steric effects prevent the typical
formation of the SN2 transition state. Thus, for example,
derivatives of R-bromoisovaleric (24), R-bromo-â,â,â-
triphenylpropionic (25), and R-bromo-â,â,â-trimethylpro-
pionic acid (26) were all reacted in the past with dry KF
in diethyleneglycol at 125 °C for 2.5 h as described in ref
2b, resulting in typical carbonium ion reactions. These
led to HBr elimination in the case of 24, HBr elimination
accompanied with phenyl rearrangement in the case of
25, and no reaction when such a rearrangement is not
very favorable as in the case of 26 (Scheme 5).18
These difficulties can easily be avoided with the
electrophilic fluorination method using AcOF. Methyl
2-norbornaneacetate (27) is similar to isovaleric acid in
exhibiting a strong steric hindrance on the R position. It
was easily converted to the corresponding trimethylsilyl
ketene acetal 28, reacted with AcOF, and formed in
minutes the novel methyl 2-fluoro(2-norbornane)acetate
(29). Even more impressive results were achieved with
The strength of this method is demonstrated with
branched carboxylic acids that cannot be converted to
their corresponding R-fluoro derivatives by nucleophilic
(7) Thenappan, A.; Burton, D. J . J . Org. Chem. 1990, 55, 2311.
Wang, Y.; Yang, Z. Y.; Burton, D. J . Tetrahedron Lett. 1992, 33, 2137.
(8) Resnati, G.; DesMarteau, D. D. J . Org. Chem. 1991, 56, 4925.
(9) Differding, E.; Lang, R. W. Tetrahedron Lett. 1988, 29, 6087.
(10) Wnuk, S. F.; Rios, J . M.; Khan, J .; Hsu, Y. L. J . Org. Chem.
2000, 65, 4169.
(11) Rozen, S.; Lerman, O.; Kol, M. J . Chem. Soc., Chem. Commun.
1981, 443.
(12) (a) Rozen, S.; Bareket, Y.; Kol M. J . Fluorine Chem. 1993, 61,
141. (b) Hebel, D.; Lerman, O.; Rozen, S. J . Fluorine Chem. 1985, 30,
141. (c) Appelman, E. H.; Mendelsohn, M. H.; Kim, H. J . Am. Chem.
Soc. 1985, 107, 6515.
(13) (a) Lerman, O.; Rozen, S. J . Org. Chem. 1983, 48, 724. (b) Rozen,
S.; Lerman, O.; Kol, M.; Hebel, D. J . Org. Chem. 1985, 50, 4753.
(14) Dayan, S.; Bareket, Y.; Rozen, S. Tetrahedron 1999, 53, 3657.
(15) Ainsworth, C.; Chen, F.; Kue, Y.-N. J . Organomet. Chem. 1972,
46, 59.
(16) Zhi, C.; Chen, Q.-Y. J . Chem. Soc., Perkin Trans. 1 1996, 1741.
(17) Rozen, S.; Gal, C. J . Org. Chem. 1987, 52, 2769.
(18) These side reactions occurred with many types of derivatives
including various amides. Esters, in addition to these typical SN1
reactions, suffered also from trans esterefication problems. Rozen, S.
Unpublished Results.

7466 J . Org. Chem., Vol. 66, No. 22, 2001
Rozen et al.
Sch em e 4
Sch em e 6
In the past, we have stated that AcOF is a relatively
mild carrier of elemental fluorine and its reactions better
controlled. In part, this is due to the fact that, unlike F2,
acetyl hypofluorite is completely soluble in organic
solvents so its concentration is controllable at any given
time during the reaction. The present study confirms this
statement once again. When we passed dilute fluorine
through silyl enol derivatives described above, the yields
of the R-fluoroesters dropped to around 20%. Although
the preparation of this hypofluorite seems at first to be
an extra step, it really is not. The preparation is fast,
the reagent does not need any manipulations such as
isolation or purification, and essentially, it is a one-pot
reaction.
Sch em e 5
Exp er im en ta l Section
1H NMR spectra were recorded using a 200 MHz machine
with CDCl3 as a solvent and Me4Si as an internal standard.
The 19F NMR spectra were measured at 188.1 MHz and are
reported in parts per million upfield from CFCl3 serving as
an internal standard. The proton broad band decoupled 13C
NMR spectra were recorded at 50.2 MHz. Here too, CDCl3
served as a solvent with TMS as an internal standard. MS
and GC-MS spectra were usually measured under EI condi-
tions. Chemical ionization (CI) protocol was used for more
sensitive samples. In extreme cases where neither EI nor CI
methods could detect the molecular ion, we have used a
supersonic GC-MS developed in our department.19 IR spectra
were recorded as neat films, in CHCl3 solution on an FTIR
spectrophotometer. The spectral properties of all products
presented in this work are in excellent agreement with their
structure, and where relevant, with the ones described in the
literature.
methyl 3,3,3-triphenylpropionate (30) and methyl 3,3,-
dimethylbutyrate (33), which were converted to their
respective trimethylsilyl ketene acetals 31 and 34 and
reacted with AcOF. The unknown methyl 2-fluoro-3,3,3-
triphenylpropionate (32) and methyl 2-fluoro-3,3,-dim-
ethylbutyrate (35) were produced, both in excellent yields
(Scheme 6).
All reactions described above involved the ester moiety,
but it is as easy to react free acids. The only difference
is using twice as much base and Me3SiCl in order to
convert the acids into ketene bis(trimethylsilyl acetals).
It should be pointed out, however, that the yields of these
enol derivatives are somewhat lower (around 70-80%)
than the ones obtained from esters. Thus, the ketene bis-
(trimethylsilyl acetal) 37 of cycloheptyl carboxylic acid
(36) was reacted with AcOF to produce the R-fluorocy-
cloheptyl carboxylic acid (38) in higher than 75% yield.
Similarly, ibuprofen (19) itself was converted to the
2-fluorinated free acid 20 in 60% yield (Scheme 4). The
free R-fluorocarboxylic acids could also be obtained by
basic hydrolysis (5% NaOH in EtOH/H2O at 70 °C for
about 1 h) in nearly quantitative yields.
Gen er a l P r oced u r e for Wor k in g w ith F lu or in e. Fluo-
rine is a strong oxidant and a corrosive material. In organic
chemistry, it is mostly used after dilution with nitrogen or
helium. Such dilution can be achieved by using an appropriate
vacuum line constructed in a well-ventilated area. A detailed
description for such a setup has appeared lately.20 Despite
hundreds of papers found in the literature dealing with its
reactions with organic compounds, working with F2 is still
viewed by some as something extraordinary that should be
avoided almost by any cost. The situation, however, is much
simpler, and if elementary precautions are taken, work with
F2 is simple and can be carried in regular glassware. We have
(19) Dagan, S.; Amirav, A. J . Am. Soc. Mass. Spectrom. 1995, 6,
120. Gordin, A.; Tzanani, N.; Amirav, A. Rapid. Com. Mass Spectrom.
2001, 15, 811.
(20) Dayan, S.; Kol, M.; Rozen, S. Synthesis 1999, 1427.

Synthesis of R-Fluorocarboxylates
J . Org. Chem., Vol. 66, No. 22, 2001 7467
run numerous reactions with it and never had any problem.
For the occasional user, a prediluted 10% fluorine in nitrogen
solution is commercially available, making the whole process
very simple indeed.
3.49 (3 H, s), 1.91-0.87 (14 H, m), 0.2 ppm (9 H, s); 13C NMR
150, 98, 56.33, 29.40, 20.88, 13.69, 0.50 ppm. Compound 11
was then reacted with AcOF as described above to give 12 in
90% yield: oil; IR 1763 cm-1; 1H NMR 3.78 (3 H, s), 1.80-1.2
(8 H, m) 0.9 ppm (6 H, t, J ) 7 Hz); 19F NMR -166.7 ppm (m);
13C NMR 172.16 (d, J ) 23 Hz), 97.93 (d, J ) 186 Hz), 52.09,
P r ep a r a tion of AcOF a n d Its Rea ction w ith th e En ol
F or m of th e Ester s a n d Acid s. A mixture of 10-15% F2 in
N2 was bubbled into a cold (-45 °C) suspension of 2 g of
AcONa‚AcOH dispersed in 100 mL of CH3CN and 10 mL of
AcOH placed in a standard glass vessel. The solvated salt could
be made by leaving anhydrous AcONa over AcOH in a closed
desiccator for at least 24 h. The amount of the AcOF thus
obtained could be easily determined by reacting aliquots of the
reaction mixture with aqueous KI solution and titrating the
liberated iodine. A concentration of up to 0.35 M (60 mmol)
AcOF was thus achieved in about 1.5-2 h.
Syn th esis of Tr im eth yl Silyl En ol Der iva tives. All enol
derivatives (25 mmol) were prepared by standard procedures
such as reacting the corresponding esters or acids with
commercial LDA followed by trimethylsilyl chloride.15 The
resulting ketene acetals were purified by vacuum distillation,
but no attempt was made to obtain analytically pure samples.
These silyl enol derivatives can be stored indefinitely under
nitrogen at 0 °C. Some spectral properties of the ketene acetals
that are not described in the literature are given. Yields of
70% and higher were easily achieved by this method.
Gen er a l Syn th esis of R-F lu or o Ester s. A stirred CHCl3
solution (10 mL) of the ketene acetal (usually about 1.3 mmol),
obtained as described above, was cooled to -45 °C. AcOF (3.9
mmol) solution was added to it in one portion. After a few
minutes, the reaction was terminated by pouring it into 500
mL of NaHCO3 solution followed by water until neutral, drying
the organic layer over MgSO4, and finally evaporating the
solvent. The products were usually purified by flash chroma-
tography (Merck silica gel 60H) using petroleum ether-
chloroform as eluent.
39.63 (d, J ) 23 Hz), 16.67, 13.98 ppm; HRMS calcd for C9H17
-
-
FO2 176.1213 (M)+, found 176.1214. Anal. Calcd for C9H17
FO2: C, 61.34; H, 9.72; F, 10.78. Found: C, 60.90; H, 9.48; F,
10.58.
Meth yl R-F lu or ocycloh ep ta n oa te (15) a n d R-F lu or o-
cycloh ep tyl Ca r boxylic Acid (38). Methyl cycloheptanoate
(13) was converted as described above to its trimethylsilyl
1
ketene acetal 14: oil; bp(0.1mm) 50 °C; H NMR 3.45 (3 H, s),
2.12-2.08 (4 H, m), 1.6-1.4 (8 H, m), 0.16 ppm (9 H, s); 13C
NMR 149.72, 100.71, 56.55, 29.90, 29.52, 29.00, 28.65, 27.94,
27.80, 1.39 ppm. Compound 14 was then reacted with AcOF
as described above to give 15 in 60% yield: oil; IR 1742 cm-1
;
1H NMR 3.75 (3 H, s), 2.10-2.05 (2 H, m) 1.99-1.94 (2 H, m),
1.67-1.58 (4 H, m), 1.24 (2 H, m), 0.85-0.82 ppm (2 H, m);
19F NMR -149.8 ppm (m); 13C NMR 173.71 (d, J ) 26 Hz),
97.46 (d, J ) 185 Hz), 54.55, 36.77 (d, J ) 24 Hz), 29.29, 22.10
ppm; HRMS (CI) calcd for C9H15FO2 175.11357 (MH)+, found
175.11343. Anal. Calcd for C9H15FO2: C, 62.05; H, 8.68.
Found: C, 62.25; H, 8.50.
The bis(trimethylsilyl acetal) 37 was obtained from cyclo-
1
heptyl carboxylic acid (36): oil; bp(0.6mm) 75 °C; H NMR 2.12
(4 H, m), 1.48 (8 H, m), 0.18 ppm (18 H, s); 13C NMR 144.75,
98.03, 30.09, 28.47, 0.47 ppm. It was then reacted with AcOF
to produce 38 in 75% yield: oil; IR 1723 cm-1; 1H NMR 8.7 (1
H, bs), 2.13-2.05 (4 H, m) 2.05-2.00 (2 H, m), 1.71-1.64 ppm
(6 H, m); 19F NMR -149.7 ppm (quint, 24 Hz); 13C NMR 178.71
(d, J ) 27 Hz), 96.93 (d, J ) 183 Hz), 36.47 (d, J ) 23 Hz),
29.20, 22.04 ppm (d, J ) 4 Hz); HRMS (CI) calcd for C8H13
-
FO2 161.09778 (MH)+, found 161.09648. Compound 38 was
also obtained in 90% yield by heating 40 mg of 15 with 5%
aqueous NaOH (6 mL) and 1.5 mL of EtOH at 70 °C for 1 h.
Meth yl 2-F lu or o-2-p h en ylbu ta n oa te (18). Methyl 2-phe-
nylbutanoate (16) was converted as described above to its
Meth yl 2-F lu or op h en yla ceta te (3).3b Methyl phenylac-
etate (1) was converted as described above to its trimethylsilyl
ketene acetal 215 (E and Z mixture): oil; bp(0.3mm) 78 °C; 1H
NMR (all peak pairs are due to the E and Z isomers) 7.47-
7.03 (5 H, m), 4.7 and 4.62 (1 H, s), 3.72 and 3.70 (3 H, s),
0.35 and 0.29 ppm (9 H, s). Compound 2 was then reacted with
1
trimethylsilyl ketene acetal 17: oil; bp(0.4mm) 42 °C; H NMR
(all pairs for Z and E isomers) 3.60 and 3.42 (3 H, s), 2.36 (2
H, m), 0.92 (3 H, m), 0.27 and 0.00 ppm (9 H, s); 13C NMR (all
pairs for Z and E isomers) 151 and 150, 139-125.08, 103.0
and 102.0, 56.76 and 55.62, 23.32 and 23.01, 13.95 and 13.00,
-0.14 and -0.48 ppm. Compound 17 was then reacted with
AcOF as described above to give 18 in 70% yield: oil; IR 1734
cm-1; 1H NMR 7.52-7.24 (5 H, m), 3.75 (3 H, s), 2.39-2.09 (2
H, m), 0.95 ppm (3 H, t, J ) 7 Hz); 19F NMR -167.73 ppm
(dd, J 1 ) 27 Hz, J 2 ) 21 Hz); 13C NMR 171.00 (d, J ) 27 Hz),
138.59 (d, J ) 60 Hz), 127.94, 124.81, 124.62, 97.48 (d, J )
189 Hz), 52.61, 31.55 (d, J ) 22 Hz), 7.44 ppm; HRMS calcd
for C11H13FO2 196.08995 (M)+, found 196.08949. Anal. Calcd
for C11H13FO2: C, 67.33; H, 6.68; Found: C, 67.78; H, 7.03.
Meth yl 2-F lu or o-2-(4-isobu tylp h en yl)p r op ion a te (23)4
a n d 2-F lu or o-2-(4-isobu tylp h en yl)p r op ion ic Acid (20).4,6b
Methyl 2-(4-isobutylphenyl)propionate (21) was converted as
described above to its trimethylsilyl ketene acetal 22: oil;
bp(0.2mm) 92 °C; 1H NMR (for both E and Z isomers) 7.31-7.26
(4 H, m), 3.62 and 3.51 (3 H, s), 2.43 (2 H, d, J ) 8 Hz), 2.10-
1.81 (1 H, m), 1.98 and 1.94 (3 H, s), 0.92-0.87 (6 H, m), 0.28
and 0.03 ppm (9 H, s). Compound 22 was then reacted with
AcOF as described above to give 23 in 70% yield: oil; IR 1760
cm-1; 1H NMR 7.4 (2 H, d, J ) 8 Hz), 7.16 (2 H, d, J ) 8 Hz),
3.77 (3 H, s), 2.47 (2 H, d, J ) 7 Hz) 1.94 (3 H, d, J ) 22 Hz),
1.86 (1 H, sept, J ) 7 Hz), 0.9 ppm (6 H, d, J ) 7 Hz); 19F
NMR -150.67 ppm (q, J ) 22 Hz); HRMS calcd for C14H19FO2
238.1369 (M)+, found 238.1370.
AcOF as described above to give 3 in 90% yield: oil; bp(15
mm)
100 °C; IR 1760 cm-1; H NMR 7.40 (5 H, m), 5.80 (1 H, d, J
1
) 47 Hz), 3.78 ppm (3 H, s); 19F NMR -180.07 ppm (d, J ) 47
Hz); HRMS calcd for C9H9FO2 168.0586 (M)+, found 168.0582.
Meth yl 2-F lu or oocta n oa te (6).16 Methyl octanoate (4) was
converted as described above to its trimethylsilyl ketene acetal
5 (mixture of E and Z): oil; bp(0.1mm) 53 °C; 1H NMR 3.68 (1 H,
t, J ) 7 Hz), 3.50 and 3.47 (for E and Z) (3 H, s), 1.93 (2 H,
m), 1.26 (8 H, m), 0.87 (3 H, t, J ) 7 Hz), 0.20 and 0.18 ppm
(9 H, s). Compound 5 was then reacted with AcOF as described
above to give 6 in 80% yield: oil; IR 1765 cm-1; 1H NMR 4.90
(1 H, dt, J 1 ) 50 Hz, J 2 ) 6 Hz), 3.79 (3 H, s), 1.88 (2 H, m)
1.35 (8 H, m), 0.87 ppm (3 H, t, J ) 7 Hz); 19F NMR -192.5
ppm (dt, J 1 ) 50 Hz, J 2 ) 25 Hz); 13C NMR 170 (d, J ) 22.5
Hz), 88.48 (d, J ) 183 Hz), 51.59, 31.95 (d, J ) 20.5 Hz), 31.07,
28.29, 23.85 (d, J ) 2 Hz), 22.03, 13.45 ppm.
Met h yl 2-F lu or o-4-cycloh exylb u t a n oa t e (9). Methyl
4-cyclohexylbutanoate (7) was converted as described above
to its trimethylsilyl ketene acetal 8: oil; bp(0.2mm) 60 °C; 1H
NMR 3.67 (1 H, t, J ) 7 Hz), 3.44 (3 H, s), 2.1-0.75 (15 H, m),
0.18 ppm (9 H, s); 13C NMR 153, 85.05, 54.12, 38.07, 36.69,
32.95, 26.41, 26.10, 21.32, 0.22 ppm. Compound 8 was then
reacted with AcOF as described above to give 9 in 70% yield:
1
oil; IR 1756 cm-1; H NMR 4.9 (1 H, ddd, J 1 ) 49 Hz, J 2 ) 7
Hz, J 3 ) 5 Hz), 3.78 (3 H, s), 2.10-1.5 (5 H, m) 1.5-1.0 (8 H,
m), 0.87 ppm (2 H, m); 19F NMR -192.6 ppm (dt, J 1 ) 49 Hz,
J 2 ) 25.2 Hz); 13C NMR 170.57 (d, J ) 24 Hz), 89.28 (d, J )
183 Hz), 52.22, 37.17, 33.14, 32.95, 31.76, 29.91 (d, J ) 21
The bis(trimethylsilyl acetal) of 19 was obtained from the
acid: oil; bp(0.3mm) 67 °C; 1H NMR 7.26 (2 H, d, J ) 8 Hz), 7.01
(2 H, d, J ) 8 Hz), 2.42 (2 H, d, J ) 8 Hz) 1.88 (3 H, s), 1.86-
1.76 (1 H, m), 0.88 (6 H, d, J ) 6 Hz) 0.26 (9 H, s), -0.06 ppm
(9 H, s). It was then reacted with AcOF to produce 20 in 60%
yield.6
Hz), 26.50, 26.20 ppm; HRMS (CI) calcd for
C11H19FO2
203.145042 (MH)+, found 203.144733. Anal. Calcd for C11H19
FO2: C, 65.32; H, 9.47. Found: C, 65.49; H, 9.38.
-
Meth yl 2-P r op yl-2-flu or op en ta n oa te (12). Methyl 2-pro-
pylpentanoate (10) was converted as described above to its
Meth yl 2-F lu or o-(2-Nor bor n a n e)a ceta te (29). Methyl
2-norbornaneacetate (27) was converted as described above to
1
trimethylsilyl ketene acetal 11: oil; bp(0.2mm) 45 °C; H NMR

7468 J . Org. Chem., Vol. 66, No. 22, 2001
Rozen et al.
335.14514 (MH)+, found 335.14473. Anal. Calcd. for C22H19
FO2: C, 79.01; H, 5.73, F, 5.69. Found: C, 78.79; H, 5.91; F,
5.49.
its trimethylsilyl ketene acetal 28: oil; bp(0.7mm) 86 °C; 1H
NMR 3.61 (1 H, d, J ) 9 Hz), 3.51 (3 H, s), 2.29-2.17 (2 H,
m), 1.91 (1 H, m), 1.61-1.05 (8 H, m), 0.2 ppm (9 H, s); 13C
NMR 151.61, 92.34,54.16, 43.33, 39.87, 37.24, 36.11, 35.41,
29.18, 28.15, -0.73 ppm. Compound 28 was then reacted with
AcOF as described above to give two diastereoisomers of 29
-
Meth yl 2-F lu or o-3,3,-d im eth ylbu tyr a te (35). Methyl 3,3-
dimethylbutyrate (33) was converted as described above to its
1
trimethylsilyl ketene acetal 34: oil; bp(12mm) 60 °C; H NMR
1
in 70% yield: oil; IR 1755 and 1711 cm-1; H NMR (for the
3.72 (1 H, s), 3.49 (3 H, s), 1.05 (9 H, s), 0.21 ppm (9 H, s); 13
C
two diastereoisomers) 4.74 (1 H, dd, J 1 ) 50 Hz, J 2 ) 6 Hz),
4.54 (1 H, dd, J 1 ) 50 Hz, J 2 ) 8 Hz), 3.78 (3 H, s), 2.40-1.8
(4 H, m), 1.6-1.0 ppm (7 H, m); 19F NMR (for the two
diastereoisomers) -198.69 (dd, J 1 ) 50 Hz, J 2 ) 28 Hz),
-188.46 ppm (dd, J 1 ) 50 Hz, J 2 ) 20 Hz); 13C NMR (for both
diastereoisomers) 169.85 and 169.79 both (d, J ) 22 Hz), 91.60
and 91.36 both (d, J ) 184 Hz), 52.01, 44.19 and 44.07 both
(d, J ) 20 Hz), 41.15 and 41.10, 39.20 and 39.17, 36.63 and
NMR 152.41, 96.67, 54.09, 31.01, 29.67, -0.4 ppm. Compound
34 was then reacted with AcOF as described above to give 35
in 83% yield: oil; IR 1755 cm-1; 1H NMR 4.55 (1 H, d, J ) 49
Hz), 3.79 (3 H, s), 1.03 ppm (9 H, s); 19F NMR -194.18 ppm
(d, J ) 49 Hz); 13C NMR 168.87 (d, J ) 25 Hz), 95.46 (d, J )
187 Hz), 51.24, 34.32 (d, J ) 19 Hz), 24.96 ppm; MS regular
measurements such as EI and CI with commercial instruments
fail to show any molecular ion peak. Repeating the measure-
ment with an unique supersonic GC-MS developed in our
department, using electron ionization of the sample compound
while it is vibrationally cooled in a supersonic molecular beam,
reveals the molecular ion peak at m/e ) 148 (M)+. The
vibrational cooling considerably enhance the relative abun-
dance of the molecular ion.19
36.16, 35.63, 32.33 and 31.31 ppm; HRMS (CI) calcd for C10H15
FO2 187.11362 (MH)+, found 187.11343. Anal. Calcd for C10H15
FO2: C, 64.50; H, 8.12. Found: C, 64.83; H, 8.40.
-
-
Meth yl 2-F lu or o-3,3,3-tr ip h en ylp r op ion a te (32). Methyl
3,3,3-triphenylpropionate (30) was converted as described
above to its trimethylsilyl ketene acetal 31: oil that could not
be distilled because of decomposition at bp; 1H NMR 7.3-7.13
(15 H, m), 4.87 (1 H, s), 2.86 (3 H, s), 0.27 ppm (9 H, s). The
crude 31 was then reacted with AcOF as described above to
Ack n ow led gm en t. This research was supported by
the USA-Israel Binational Science Foundation (BSF),
J erusalem, Israel and by the Israel Science Foundation
founded by the Israel Academy of Science and Human-
ities.
1
give 32 in 80% yield: oil; IR 1747 cm-1; H NMR 7.34-7.15
(15 H, m), 6.41 (1 H, d, J ) 47 Hz), 3.39 (3 H, s); 19F NMR
-187.82 ppm (d, J ) 47 Hz); 13C NMR 168.24 (d, J ) 24 Hz),
142.44, 129.47, 127.52, 126.47, 91.76 (d, J ) 194 Hz), 61.23
(d, J ) 19 Hz), 51.75 ppm; HRMS (CI) calcd for C22H19FO2
J O010677K