COMMUNICATION
A convergent approach to c-carbonyl vinyl boronates
Markus R. Heinrich, Lisa A. Sharp and Samir Z. Zard*
Received (in Cambridge, UK) 25th February 2005, Accepted 19th April 2005
First published as an Advance Article on the web 11th May 2005
DOI: 10.1039/b502885j
?
to give alkyl radical R . The severity of the competition will vary
c-Carbonyl vinyl boronates can be prepared by a visible light
induced radical chain addition of an S-acyl dithiocarbonate
(xanthate) to the pinacol ester of vinyl boronic acid, followed by
treatment with base.
with the nature of the initial carboxylic acid 1, since the rate of the
6
?
fragmentation depends largely on the stability of R .
Irradiation with a tungsten-halogen lamp of a 1,2-dichloro-
methane solution of xanthate 2a derived from 3,3-dimethylbuta-
noic acid and a three-fold excess of vinyl boronate 3 produced the
corresponding addition product 4a, which was observed by NMR
but which was not isolated. It was simply treated in the crude state
with triethylamine and methyl iodide to give the desired boronate
5a in 67% overall yield (Table 1).{ The presence of the methyl
iodide allows a cleaner elimination by irreversibly intercepting the
xanthate salt and any other sulfur nucleophile that could be
generated in the medium. In its absence, complex mixtures were
formed.
The palladium catalysed Suzuki–Miyaura cross coupling and
related reactions of vinyl- and aryl-boronates have emerged in
recent times as exceptionally powerful and useful carbon–carbon
bond forming processes.1 This has, in consequence, created a need
for new methods allowing access to variously substituted vinyl
boronates.2 Vinyl boronates possessing a ketone group in the
c-position represent a rare class for which only two synthetic
routes have been reported. The earliest, devised by Jehanno and
Vaultier,3 consists of a multistep sequence proceeding through the
intermediacy of c-boronyl acrolein. The second, developed by
Miyaura et al.,4 relies on a palladium catalysed coupling of enol
triflates derived from cyclic diketones with bis(pinacolato)diboron.
We now describe a radical based, expedient approach to members
of this family, some of which would be difficult to obtain by
existing approaches.
It appeared therefore possible to capture the acyl radical derived
from a primary carboxylic acid using the mildly reactive vinyl
boronate trap.7 This was further confirmed by converting the
cholic acid xanthate 2b into the corresponding boronate 5b, albeit
in lower overall yield. The use of the more hindered O-neopentyl
xanthate (neoPn 5 neopentyl) is often beneficial in view of the
tendency of acyl xanthates to undergo decomposition through an
ionic chain reaction. In the case of xanthate 2c derived from the
secondary 4-piperidinecarboxylic acid, decarbonylation of the acyl
radical is much faster and becomes a serious complication causing
the yield of the desired product to drop dramatically.
Our concept is summarised in Scheme 1. It hinges on the
possibility of adding an acyl radical, derived from the correspond-
ing S-acyl dithiocarbonate (xanthate) 2, to the commercially
available vinyl boronate 3.5 The key property of this process is
that the xanthate group is transferred to the product 4 in a position
b- to the carbonyl group and becomes therefore susceptible to
elimination with base, leading thus to the desired unsaturated
boronate 5. The major side reaction that could limit the scope of
this approach is the premature decarbonylation of the acyl radical
Acyl radicals arising from aromatic carboxylic acids do not
normally lose carbon monoxide and can therefore be successfully
employed in the above sequence. This is illustrated by the synthesis
of boronates 5d,e,f starting from the respective xanthates 2d,e,f
(Table). Substituted aromatic and heteroaromatic motifs can thus
be introduced into the stucture. Cyclopropylacyl radicals is the last
category of acyl radicals that do not decarbonylate easily that we
examined,8 as shown by the transformation of 2g into boronate 5g
in 46% yield (PhthN 5 phthalimido). Such densely functionalised
boronates would not be trivial to obtain by more conventional
routes.
In principle, the same approach could be applied to the synthesis
of the analogous vinyl stannanes. These are also valuable
intermediates in Stille’s palladium catalysed cross-coupling reac-
tion.9 However, when acyl xanthate 2a was irradiated in the
presence of vinyl tributyltin 6, only a low yield (24%) of the
addition product 7a was obtained (Scheme 2). The main source of
difficulty was uncovered when the reaction was repeated using
xanthate 2e, as this compound allows a more convenient
analysis of the reaction mixture. In addition to the expected
product 7e (31%), we isolated ethyl ester 8 in 48% yield. Such
ester formation was not observed to any significant extent with
the boronate reactions discussed above. Organotin derivatives
are known to catalyse trans-esterification reactions10 and
Scheme 1 A radical based approach to vinyl boronates.
*zard@poly.polytechnique.fr
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3077–3079 | 3077
Heinrich, Markus R.
