Scheme 2. Synthesis of Hydroxy Mesylate 2
Scheme 3. Stereoselectivity of 1,4-Addition of Alcohols (as
the solvent/nucleophile) to Compound 2, in the Presence of
t-BuOK
new glycosylation procedure. Moreover, to the best of our
knowledge, no procedure of this type and no epoxy glycal
such as 1 have been described before in literature. Prompted
by these considerations, thinking that the presence of the
nucleophile as the solvent in the previous protocol, besides
not being practical and/or suitable in some cases, undoubtedly
had a detrimental effect on the stereoselectivity of the
addition reaction, we elaborated a one-pot procedure to make
the cyclization process of the hydroxy mesylate 2 independ-
ent from the glycosylation process.
with TrCl (1 equiv) to give the trityl derivative 5, O-
monoprotected on the less hindered primary hydroxyl. To
introduce the mesyl group on the C(4) regioselectively, it
was necessary to protect the C(3) position. Actually, due to
the decidedly different steric hindrance around the alcoholic
functionality on C(3) and C(4), the treatment of 5 with
TBDMS-Cl (1 equiv) afforded the mono C(3)-O-TBS
derivative 6, which was then treated with MsCl to afford
the 3,6-di-O-protected mesylate 7. The treatment of 7 with
TBAF in THF readily deprotects the C(3) position, affording
the desired hydroxy mesylate 2.
The transformation of compound 2 into epoxide 1 was
conducted using t-BuOK in anhydrous benzene. We were
unable to isolate the epoxide, but conducting the reaction in
C6D6 showed that compound 1 was formed cleanly within 5
min (1H NMR). This observation led us to leave aside the
isolation of epoxide 1 and to utilize a solution of compound
2 in a solvent (benzene, an alcohol, or MeCN, vide infra) in
the presence of t-BuOK as a source of the epoxide 1, which
is actually formed in situ. In this way, it was possible to
check the reactivity of this oxirane system even if it was
not effectively isolated.
In the modified protocol, t-BuOK (1 equiv) was added to
a solution of compound 2 in anhydrous benzene at room
temperature (Scheme in Table 1). The formation in situ of
epoxide 1 was followed by TLC (disappearance of the
starting material), and MeOH (the glycosyl acceptor, 3 equiv)
was rapidly added to afford the 2-unsaturated methyl
â-glycoside 8â as practically the only reaction product (the
1
corresponding R-anomer 8R was less than 3%, H NMR)
(Table 1, entry 1). Glycosyl acceptors other than MeOH were
utilized in coupling with glycal-derived vinyl oxirane 1 under
the same protocol. The results obtained indicate that primary
(EtOH and BnOH), secondary (i-PrOH), and even tertiary
alcohols (t-BuOH) and phenol are glycosylated to afford a
good yield of the corresponding â-glycosides 9-13â in a
completely stereoselective way (the corresponding R-anomers
1
were absent, H NMR) (Table 1, entries 2-6).1a,c,6-8 The
use of diacetone D-glucose (15) as the glycosyl acceptor
At first we checked the chemical behavior of 1 in the
presence of oxygenated nucleophiles such as alcohols, used
as the solvent. In this way, a solution of compound 2 in
MeOH in the presence of t-BuOK (1 equiv) at room
temperature afforded a crude product consisting of an almost
1:1 mixture of anomeric 2-unsaturated methyl R- and
â-glycosides 8R and 8â, indicating that, in this case,
nucleophilic attack had occurred regioselectively in a con-
jugate fashion (1,4-addition pathway, Scheme 1), with
inversion of configuration on C(4) with respect to 2, but
unfortunately devoid of any stereocontrol at the anomeric
C(1) carbon (Scheme 3).
Use of more hindered alcohols as the solvent/nucleophile
(Scheme 3) led to an increased C(1)-â-selectivity. In fact,
with EtOH and i-PrOH, the corresponding R/â ratios were
25:755 and 5:95, respectively. The reaction appeared to be
interesting to us, because, after all, it turned out to be like a
demonstrated that our protocol is also useful for the
(6) The structures of glycosides 8-14 have been demonstrated by
comparison of their 1H and 13C NMR spectra with previously reported data
for the same compound, as in the case of 9R (see ref 5), and/or closely
related compounds (R and â-anomers). See: (a) Wieczorek, E.; Thiem, J.
J. Carbohydr. Chem. 1998, 17, 785. (b) Nguefack, J.-F.; Bolitt, V.; Sinou,
D. J. Org. Chem. 1997, 62, 1341. (c) Baer, H. H.; Hanna, Z. S. Can. J.
Chem. 1981, 59, 889.
(7) A confirmation of the assigned â-glycosidic structure of 8-14â was
obtained by hydrogenation (Pd/C, AcOEt) of the O-ethyl derivatives 9â
and 9R to the corresponding saturated compounds 9â-H and 9R-H,
respectively. The 1H NMR spectrum of 9â-H shows the anomeric H-1 proton
(δ 4.50) as a doublet of doublets (J ) 8.6, 2.1 Hz), whereas 9R-H shows
the same proton (δ 4.89) as a broad singlet (W1/2 ) 6.5 Hz) to indicate its
axial (9â-H) and equatorial orientations (9R-H). See: Chini, M.; Crotti,
P.; Gardelli, C.; Macchia, F. J. Org. Chem. 1994, 59, 4131 and references
therein.
(5) Ethyl R-glycoside 9R has been previously described. See: Moufid,
N.; Chapleur, Y.; Mayon, P. J. Chem. Soc., Perkin Trans. 1 1992, 991.
3696
Org. Lett., Vol. 4, No. 21, 2002