Communications
the starting material and formation of a new product. To our
generating the diol 2 in situ,[13] which then should undergo a
double intramolecular hydroalkoxylation reaction of the
triple bond. Indeed, we found after some experimentation
that 2mol% PtCl 4 in acetic acid led to complete conversion of
the starting material into the bicyclic acetal. Although other
Lewis acids were able to cleave the BDA group and form the
diol 2 as well, only PtCl4 and AuCl3 could catalyze the
subsequent hydroalkoxylation reaction. Other solvent sys-
tems were not as effective for the domino sequence. In
particular, the addition of excess water (5% v/v) leads to
complete deactivation of the catalyst and no conversion of the
starting material was detected. Addition of trifluoroacetic
acid to the reaction mixture resulted in faster conversion;
however, we observed the formation of numerous decom-
position products. Acetic acid alone did not give any
conversion of the starting material, clearly demonstrating
that the platinum catalyst is necessary for both steps of the
domino reaction.
The cycloisomerization step of the domino reaction
proceeded with the same regioselectivity as already observed,
that is, terminal alkynes lead to [3.2.1]bicyclic acetals 1 by a 6-
exo pathway, whereas aryl-substituted alkynes give rise to
[4.2.1]bicycles 6 through a 7-endo cyclization (Table 2,
entries 1–8). The yields of the isolated products for the
domino process were generally good and comparable to the
cycloisomerizations of the corresponding alkyne diols. The
versatility of the BDA-protected building blocks allowed us
to easily access a number of propargylic ethers 3 (X = O) with
substituents at different positions (Table 2, entries 9–17).
These substrates also underwent the domino deprotection-
cyclization sequence in good yield.
surprise, the major isolated product in most cases was the
[4.2.1]bicyclic acetal 6, which arises from an initial 7-endo
cyclization, instead of the [3.2.1]bicyclic acetals 1, formed by
an initial 6-exo cyclization.[8] While the competition between
6-endo and 5-exo cycloisomerization reactions has been
extensively studied,[4e,9] to the best of our knowledge, only
one example of a 7-endo cyclization of an alkyne diol bearing
a dialkyl-substituted triple bond has been reported.[4d,10] The
structures of 6c, 6d, 6g–6j were confirmed by single-crystal
X-ray diffraction analysis.[11] The reaction proceeded in good
yield for substrates without a strong electron-withdrawing
substituent on the aromatic ring (2c–i), and gave the
homochiral bicyclic acetals 6c–i resulting from a 7-endo
cycloisomerization reaction with high selectivity (Table 1,
entries 3–9). However, the presence of an electron-withdraw-
ing substituent in the para position of the aromatic ring
diminishes the reactivity and changes the regioselectivity of
the cycloisomerization reaction. Thus, in the case of p-CF3
(2j) we observed a 1:2.5 ratio of the 6-exo product 1j to the 7-
endo product 6j (Table 1, entry 10), which could be separated
by column chromatography and their structures were con-
firmed by X-ray analysis. When the crude reaction mixture
was analyzed by NMR spectroscopy after three hours, the
enol ethers 7j and 8j were detected exclusively (Scheme 2).
The ratio between the 6-exo-dig enol ether 7j and the 7-endo-
dig enol ether 8j was the same as for the final products (1:2.5),
A plausible mechanism for the cycloisomerization-hydro-
alkoxylation reaction is depicted in Scheme 3. Coordination
of the platinum catalyst to the alkyne 2 provides the p-
complex 9 in which the triple bond is activated towards an
intramolecular nucleophilic attack by one of the hydroxy
groups.[14] This step proceeds in such a way that the metal
migrates to the sterically less encumbered position, and
nucleophilic attack occurs at the end of the triple bond where
the developing positive charge is best stabilized.[3b] Thus,
terminal alkynes cyclize by the 6-exo pathway, whereas the
cyclization of arylalkynes proceeds almost exclusively by the
7-endo pathway. Subsequent proton transfer leads to the enol
ether 7 or 8, respectively. In one case we were able to observe
the initial formation of both enol ethers, which strongly
supports their role as intermediates in the reaction.[15] Finally,
the corresponding fused bicyclic acetals 1 or 6 are formed by a
proton or Lewis acid catalyzed intramolecular hydroalkox-
ylation. In the cases where BDA-protected diols 3 are
employed, the described sequence is preceded by a Lewis
acid catalyzed deprotection of the substrate.
Scheme 2. The 6-exo-dig versus the 7-endo-dig cyclization reaction of
diol 2j.
which supports the hypothesis that these enol ethers are
indeed intermediates in the formation of the fused bicyclic
acetals. Both enol ethers could be isolated after column
chromatography.[12] The strongly electron-withdrawing nature
of the p-NO2 group in 2k resulted in the reaction proceeding
much more slowly and, notably, the major isolated product
was the fused bicyclic acetal 1k arising from a 6-exo
cyclization. However, the formation of several unidentified
side products was observed in this reaction.
In the present study we have described for the first time
the synthesis of [4.2.1]- and [3.2.1]-fused bicyclic acetals by an
intramolecular double alkoxylation of alkyne diols. The
course of the reaction depends on the substitution of the
triple bond. Terminal alkynes give the [3.2.1]bicyclic product
by a 6-exo pathway, whereas arylalkynes undergo a 7-endo
cyclization to the [4.2.1]bicycles.
Encouraged by these results we decided to explore
whether the platinum catalyst could directly convert the
BDA-protected substrate 3 into the desired bicyclic acetal
product by
a
domino deprotection-hydroalkoxylation
sequence. Since acetal protecting groups can generally be
removed by acid treatment, we reasoned that the Lewis acid
catalyst should be capable of cleaving the BDA group and
210
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 209 –212