Thanks very much for your comments. In answer to your first question
concerning the osmylation, we have considered the Sharpless oxidation. We
were put off slightly by the comments of a colleague who was very close to
the Sharpless work who said that cyclic cis olefins often are poor
substrates in the system, but the main reason was one of expediency, i.e.
the student had to finish lab and write up his thesis. We'd like to go
back and look at the area; another suggestion which has been made is to
protect the existing diol with a carbonate prior to osmylation. Some
colleagues at Bristol have observed some changes in osmylation selectivity
on going from bicyclic lactone-cyclopentene substrates (double bond in the
carbocycle) to the corresponding lactols. The lactones give sterically
apparently anomalous endo-diol formation, whilst the lactol gives the more
expected exo-diol.
Concerning your second question, we did look briefly at reacting the
enolate of the fully protected (i.e. two acetonides) methyl ester with
oxidising agents such as MoOPH and Davis' oxaziridine. We even tried
making the silyl ketene acetal and reacting it with peracid, all to no
avail. Again, we were hampered by lack of material and time. The
synthesis has just been accepted as a Tetrahedron Letter, and should appear
in the next few weeks.
You may be interested to know that we have begun to explore the synthetic
utility of sulfonyl glycals such as the one which we reacted with peracid
to make the hydroxylactol prior to thioglycosidation in the 2-deoxyKDO
synthesis. We can alkylate with almost complete stereoselectivity (anti to
the side-chain next to the ring oxygen) and completely regioselectively
alpha- to the sulfonyl group (MeI and allyl bromide so far). Also,
sulfonylglycals are effective C-glycoside precursors by reaction with
intermolecular nucleophiles. If we treat them with soft carbon
nucleophiles (enol ethers, allyl silanes and the like) in the presence of
Lewis acids (SnCl4 appears to be the best so far), we get loss of sulfone,
double bond transposition, and attachment of the nucleophile in what was
the anomeric position. These reactions appear to be completely regio- and
stereoselective as well, but we don't yet know whether the nucleophile
attaches syn or anti to the side-chain.
With best wishes.
Donald Craig