Organic Letters
Letter
Figure 3. Preparation of fluorinated mimics of drug-like compounds. (i) LiOH, MeOH/H2O, rt, 24 h; (ii) (COCl)2, DMF (0.1 equiv), CH2Cl2,
rt, 3 h.
In general, cobalt-catalyzed coupling reaction of alkyl
halides was proposed to form an alkyl radical via single-
electron-transfer (SET) of an alkyl halide to the cobalt
complex.4 To probe whether a free alkyl radical was generated
in the current protocol, we conducted two sets of experiments
(eq 1). In the first experiment, the reaction of compound 1a
reaction via a benzylic radical intermediate, even though
further detailed mechanistic studies are required to further
elucidate the mechanism of the reaction.
Finally, to demonstrate the applicability of the current
protocol, we applied this methodology to synthesize a few
fluorine-substituted mimics of potentially drug-like com-
pounds (Figure 2). Compound 5, which is a fluorinated
mimic of an inhibitor for the histone lysine methyltransferase
enhancer of zeste homologue 2 (EZH2),20 was prepared in
53% overall yield with 95:5 e.r. via a four-step transformation
from easily available methyl 4-(bromofluoromethyl)benzoate
1a (eq 2 in Figure 3). Likewise, a fluoride-substituted
compound 6, which is a mimic of histamine H3 receptor,21
was synthesized from the same starting material after four
steps in 64% overall yield and 95:5 e.r. (eq 3 in Figure 3).
Furthermore, we applied this method to synthesize compound
7, a key intermediate of a fluoride-substituted analog of Lilly’s
mGlu2 receptor potentiators, a compound for the acute
treatment of migraine headaches,22 in 84% yield with 92:8 e.r.
(eq 4 in Figure 3). These examples clearly showed the
potential of the current protocol in the preparation of enantio-
enriched fluorinated drug analogs.
In summary, we developed a cobalt-catalyzed asymmetric
cross-coupling reaction of α-bromo-α-fluorotoluene deriva-
tives with a variety of aryl zincates that were in situ generated
from lithium aryl n-butyl pinacol boronates with ZnBr2 under
mild conditions, which may serve as a versatile, efficient, and
convenient approach for the rapid access of optically pure α-
fluorinated diarylmethane derivatives. Furthermore, this
reaction, which exhibits some advantages over nickel catalysis
in some respects, greatly strengthens the arsenal of the cobalt-
catalyzed asymmetric cross-coupling reactions of secondary
alkyl halides. Studies to extend this protocol to cobalt-
catalyzed asymmetric cross-coupling reactions of other
secondary alkyl halides and nucleophiles are currently
underway in our laboratory.
with lithium phenyl n-butyl boronate 2a was conducted in the
presence of 1.0 equiv of radical inhibitor (2,2,6,6-tetramethyl-
piperidin-1-yl)oxyl (TEMPO) or butylated hydroxytoluene
(BHT) under otherwise identical conditions. It was found that
reaction in the presence of TEMPO was completely shut
down, while the yield of the reaction in the presence of BHT
was decreased significantly to 20%. In addition, a doublet peak
at −104.80 ppm in 19F NMR spectroscopy showed up for the
reaction with TEMPO, which was further characterized by a
high resolution mass spectrometer (ESI) to be TEMPO-R
(MS(ESI): 324.4) wherein the R group was the benzylic
radical generated from compound 1a. In a second experiment,
the same reaction was conducted in the presence of 1.0 equiv
of an SET inhibitor 1,4-dinitrobenzene. Again, the reaction
was fully shut down. Furthermore, in several cases, the dimers
of the benzylic radical that was supposed to be generated from
the benzylic bromides were observed. These results suggest
that the cobalt-catalyzed asymmetric cross-coupling of the
fluorinated secondary benzylic bromides likely involves a free
alkyl radical intermediate.
To assess whether the enantioselectivity of the reaction
resulted from kinetic resolution of the α-bromo-α-fluoroto-
luene derivatives, we monitored the enantio-excess of
compound 1a at a different period of the reaction by
quenching an aliquot of the reaction mixture with an acid and
subsequently analyzing the organic layer by HPLC. It was
found that compound 1a remained racemic at 2.0, 5.0, or 12
h, respectively. These results clearly ruled out the possibility of
a kinetic resolution process.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
The above-mentioned initial mechanistic experiments
indicate that the current cobalt-catalyzed asymmetric cross-
coupling reaction is a catalyst-controlled enantioselective
1H, 19F, 13C NMR and HPLC spectra of compounds
3a−3ai, 5−7; X-ray structure of 3q (PDF)
D
Org. Lett. XXXX, XXX, XXX−XXX