6474
E. Quesada, R. J. K. Taylor / Tetrahedron Letters 46 (2005) 6473–6476
The Bestmann–Ohira reagent 3 was easily prepared from
commercially available dimethyl (2-oxopropyl)-
phosphonate in high yield (>97% on a 5–10 g scale)
using KoskinenÕs diazo-transfer procedure.10 We ini-
tially explored the TOP sequence illustrated in Eq. 2,
in which the alcohol, MnO2 and the Bestmann–Ohira
reagent 3 were mixed together so that the aldehyde 2
would be trapped as soon as it was generated. Using p-
nitrobenzyl alcohol 1a (1 equiv), MnO2 (5 equiv), Best-
mann–Ohira reagent 3 (1.2 equiv) and K2CO3 (2 equiv)
in a mixture of THF–MeOH (1:1) at room temperature
for 18 h we were delighted to find that the desired termi-
nal alkyne 4a was obtained in 89% isolated yield. We
also established that the presence of methanol, which
deacetylates the Bestmann–Ohira reagent, is crucial
for success—reactions in THF alone failed to generate
any alkyne. In addition, attempts to replace MeOH by
other alcohols failed; when iso-propanol was employed
as co-solvent the p-nitrobenzyl alcohol was acetylated
by the Bestmann–Ohira reagent. Unfortunately, the
presence of methanol reduces the activity of the manga-
nese dioxide and the only other satisfactory substrate for
this TOP sequence was found to be 4-carbomethoxy-
benzyl alcohol 1b, which gave alkyne 4b in 78% yield.
Other benzyl alcohol derivatives with electron-withdraw-
ing substituents (e.g., p-bromobenzyl alcohol) reacted
only partially, while benzyl alcohol itself, and derivatives
containing electron-donating substituents, did not give
any observable alkyne product, even under forcing con-
ditions. Thus, only electron-deficient benzylic alcohols
such as 1a and 1b, which undergo rapid oxidation, are
suitable substrates for this tandem process.
The results in Table 1 clearly show that the two-step
sequence proceeds efficiently (59–99% isolated yields
after chromatography) and that it has general applic-
ability. Thus, excellent yields were achieved using
benzylic alcohols with electron-withdrawing groups
(entries i–iii), electron-donating groups (entry v) and
with benzyl alcohol itself (entry iv). In certain cases
(entries i and ii) the reactions were so efficient that
the products were pure (by NMR spectroscopy) after
extractive work-up and further chromatographic purifi-
cation was not required. Chromatography was also dif-
ficult when benzyl alcohol was employed (entry iv), due
to the volatile nature of phenyl acetylene: in this exam-
ple direct distillation from the crude reaction mixture
gave the product alkyne in 87% yield, although this
procedure needed to be carried out on a 10 mmol scale.
Success was also achieved using naphthalene–1-metha-
nol (entry vi), a heteroaromatic alcohol (entry vii) and
a propargylic alcohol (entry viii). It should be noted
that allylic alcohols cannot be employed in this meth-
odology because, as originally reported by Bestmann
and co-workers,5 methanol adds to the intermediate
a,b-unsaturated aldehydes.
The times needed for the oxidation step varied from 3
to 24 h, with 2,4-dimethoxybenzyl alcohol, pyridine
2-methanol and 3-phenylpropargyl alcohol taking the
longest (8, 10 and 24 h, respectively). This method does
not appear to be over-sensitive to steric factors as the
ortho-substituted examples, dimethoxybenzyl alcohol
and naphthalene–1-methanol, underwent oxidation–alk-
ynylation in good to excellent yields (entries v and vi).
MnO , K CO
3
2
2
THF-MeOH (1:1)
18 h, rt
a, X = NO2, 89%
OH
b, X = CO2Me, 78%
4
1
ð2Þ
X
O
O
X
P(OMe)
2
3
N
2
Given the above observations, we decided to develop a
sequential one-pot procedure in which the oxidation
was carried out using MnO2/THF before the addition
of the Bestmann–Ohira reagent in methanol. Thus,
the oxidations were accomplished using 5 equiv of
MnO2 in THF at room temperature. Once all of the
alcohol had been converted into the intermediate alde-
hyde 2 (TLC monitoring, 3–24 h), methanol was added
followed by K2CO3 (2 equiv) and the Bestmann–Ohira
reagent 3 (1.2 equiv). After, stirring for a further 12 h,
the terminal alkynes were obtained in good to excellent
yield. It is worth noting that the Bestmann–Ohira
alkynylation proceeds smoothly in the presence of the
unreacted MnO2. This procedure was successful with
a range of activated alcohols as can be seen from
Table 1.
Finally, the efficient preparation of 4-bromophenylacet-
ylene (entry iii) is noteworthy: transition metal-mediated
cross-coupling methods have been used to prepare this
and related compounds, but these processes require
anhydrous conditions and expensive catalysts, require
alkyne protection, and lead to product mixtures.13
In conclusion, we have developed a very mild and
straightforward sequential one-pot method for the con-
version of a variety of benzylic, heterocyclic and propar-
gylic alcohols into their corresponding homologated
terminal alkynes in good to excellent yield. In addition,
a tandem process has been developed for use with benzyl
alcohols containing highly electron-withdrawing substit-
uents (nitro, ester) on the aromatic ring. We are currently
applying this methodology in target molecule synthesis.
Quesada, Ernesto
