In our ongoing program on biological compound mimics,
bromofluoroolefins are key intermediates. We tried first to
obtain bromofluoroolefins via the more general method
which is the Wittig reaction (Scheme 1). Using standard
the reaction between triphenylphosphine and tribromofluo-
romethane. In the presence of a second equivalent of
triphenylphosphine, the phosphonium salt can be debromi-
nated to give the ylide. This step is an equilibrium, which
lies to the left, but in the presence of a trapping agent, i.e.,
an aldehyde, the equilibrium is shifted to the right. An
alternative procedure with activated zinc to debrominate the
phosphonium salt gives a stable solution of a metal-stabilized
ylide. Phosphine and zinc are the two reagents used to
debrominate the phosphonium salt derived from tribromo-
fluoromethane, to form the key ylide intermediate.
Scheme 1. Synthesis of Bromofluoroolefins
Recently it was reported that the zinc carbenoid could be
produced in situ from tribromofluoromethane treated with
diethylzinc. The carbenoid was then added to the appropriate
aldehydes or ketones to give the corresponding alcohols.14
Another interesting paper reports that the zinc carbenoid
derived from diethylzinc and chloroiodomethane could react
with tetrahydrothiophene to give the corresponding sulfonium
ylide.15 In regard to these papers, we wondered whether
diethylzinc could act as a debrominating agent of phospho-
nium derived from tribromofluoromethane to generate the
reacting ylide. This might constitute an alternative to the
classical condition.
In the first assay, we added dropwise diethylzinc (1 equiv)
to a THF solution of triphenylphosphine, 3-phenylpro-
pionaldehyde, and tribromofluoromethane (3/1/1 equiv,
respectively) at room temperature. Using these conditions,
the bromofluorovinyl Wittig product could be isolated in an
encouraging 65% yield, the best ever obtained for that
particular aldehyde. By optimizing the reaction conditions,
changing the solvent (THF, CH2Cl2, Et2O, diglyme), the
temperature (from -78 to 100 °C), the nature of the
phosphine (PPh3, P(NMe2)3), and the number of equivalents,
and employing the phosphonium salt instead of forming it
in situ, the yields could be increased to 88% by using THF
as solvent at room temperature with a 3-phenylpropion-
aldehyde/Et2Zn/CBr3F/PPh3 ratio of 1.0/1.2/1.2/1.2. (see the
typical procedure in ref 16).
conditions, described by Burton,8 we studied the Wittig
reaction of 3-phenylpropionaldehyde (1.0 equiv) with tri-
phenylphosphine(3.0 equiv) and fluorotribromomethane (1.0
equiv) in the presence of activated zinc powder (3.0 equiv)
in dry THF at room temperature under argon. The desired
product was never observed in the reaction mixture, even
after 24 h. Burton demonstrated in his review8 that the yields
of the Wittig reaction for nonactivated ketones or aliphatic
aldehydes are often low. Therefore, we tested various Lewis
acids to activate the substrate under the above reaction
conditions. The results were very discouraging, and most of
the Lewis acids such as BF3‚OEt2, AlCl3, ZnCl2, TiCl4, TiCl-
(OiPr)3, Ti(OiPr)4, AlMe3, and Et2AlCl had only a small
effect on the reaction. The best result was obtained with use
of Cp2TiCl2 as an activating agent, and 1-bromo-1-fluoro-
4-phenylbutene was isolated in 65% yield when 3-phenyl-
propionaldehyde was treated with the mixture of PPh3/CBr3F/
Zn/Cp2TiCl2 (3.0/1.0/3.0/1.0 equiv) in anhydrous THF under
Ar at room temperature overnight. Unfortunately, the im-
provement obtained with titanocene chloride was not general
and replacing 3-phenylpropionaldehyde by 2-acetylnaphtha-
lene, a ketone, led to a lower yield (20%). In addition, the
phosphonium salt could rarely be observed by 19F NMR.
The mechanism of the Wittig reaction in the absence or
presence of activated zinc has been published8,11-13 (Scheme
2).
Then, various aldehydes were subjected to these optimized
reaction conditions, giving the corresponding bromofluo-
roolefins with high yields (Table 1). Under the above
conditions, even nonactivated aldehydes could be converted
to the corresponding bromofluoroalkenes (Table 1; entries
1-3). The reaction is general and can tolerate various
functional groups such as ester, nitro, protected alcohol, and
others with yields always around 80-90%.
Scheme 2. Formation of the Ylide with or without Zinc
Unfortunately, whatever the conditions and the aldehydes
were, the stereoselectivity was never better than 70:30. At
(14) Hata, T.; Kitagawa, H.; Shimizu, M.; Hiyama, T. Bull. Chem. Soc.
Jpn. 2000, 73, 1691-1695.
(15) Aggarwal, V. K.; Ali, A.; Coogan, M. P. J. Org. Chem. 1997, 62,
8628-8629.
The first step in the above mechanism is the formation of
the phosphonium salt, through the rearrangement of the
methide ion and the bromophosphonium cation obtained from
(16) To a solution of triphenylphosphine (2.4 mmol, 1.2 equiv), tribro-
mofluoromethane (2.4 mmol, 1.2 equiv), and an appropriate aldehyde or
ketone (2.0 mmol, 1.0 equiv) in anhydrous THF (30-40 mL) was added a
solution of diethylzinc in hexanes or toluene (2.4 mmol, 1.2 equiv) dropwise
via a syringe pump over 30 min at room temperature under argon. The
mixture was stirred at room temperature for 30 min. The resulting solution
was then quenched with methanol (10 mL), stirred for 30 min, and
concentrated under reduced pressure. The residue was then chromatographed
on silica gel (eluent: cyclohexane/ethyl acetate), affording the desired
bromofluoroolefins.
(11) Burton, D. J. J. Fluorine Chem. 1983, 23, 339-357.
(12) Vanderhaar, R. W.; Burton, D. J.; Naae, D. G. J. Fluorine Chem.
1971/1972, 1, 381-383.
(13) Naae, D. G.; Kesling, H. S.; Burton, D. J. Tetrahedron Lett. 1975,
3789-3792.
2102
Org. Lett., Vol. 6, No. 13, 2004
Lei, Xinsheng
