Short Communication
Aust. J. Chem. 2002, 55, 83–85
A New Catalyst for the Reductive Elimination of
Acylated Glycosyl Bromides to Form Glycals
Robert V. Stick,A,B Keith A. Stubbs,A D. Matthew G. TilbrookA and Andrew G. WattsA
A Department of Chemistry, The University of Western Australia, 35 Stirling Highway, Crawley, W.A. 6009,
Australia.
B Author to whom correspondence should be addressed (e-mail: rvs@chem.uwa.edu.au).
C
H
0201
R
.
V
e
t
c
k
a
l
s
t
for
t
he
R
d
u
c
t
v
E
i
m
i
a
t
o
n
o
f
A
c
yl
a
e
d
G
l
yoc
s
y
l
r
m
i
d
s
Ethylene-N,N´-bis(salicylideneiminato)oxovanadium(IV) {VO(salen)} has been developed as a useful catalyst for
the reductive elimination of acylated glycosyl bromides to form glycals, both in the pyranose and furanose series,
using zinc/ammonium chloride/methanol or zinc/acetic acid/acetonitrile.
Manuscript received: 8 January 2002.
Final version: 7 February 2002.
Glycals constitute a particular class of unsaturated
monosaccharide, namely that having the double bond
between C1 and C2 of the pyranose or furanose form of a
2-deoxy anhydro alditol. As such, glycals are useful starting
materials for the synthesis of glycosides (the Ferrier
reaction[1]), amino glycosides (Lemieux’ azido nitration[2])
and oligosaccharides (Danishefsky’s glycal assembly[3]).The
first synthesis of a glycal (tri-O-acetyl-D-glucal) was by
Fischer and Zachs in 1913 and was achieved by the reductive
elimination of an acetylated glycosyl bromide, with activated
zinc, in aqueous acetic acid at low temperature.[4] Since then,
various improvements have been made to the process,[5]
including a one-pot procedure,[6] a reduction under alkaline
conditions,[7] and also a catalytic method.[8] Substrates apart
from glycosyl bromides are also convertible into glycals,
using a variety of reagents.[9]
Another stable compound of vanadium(IV) is VO(salen)
(3),* easily prepared from VO(acac)2.[10] Treatment of the
bromide (1) with VO(salen), again in the presence of zinc
and ammonium chloride in methanol, gave the D-glucal (2)
in an apparently quantitative yield (by TLC), and in a process
even more rapid than that using VO(acac)2. The eventual
isolated yield of (2) was 91%. In the absence of the
VO(salen), the reaction does not proceed at any appreciable
rate.
We have found VO(salen) to be an effective catalyst for
the preparation of glycals (Table 1) on a multi-gram scale.
The catalyst itself is green in colour and changes to brown as
the successful reaction progresses. There appears to be a
limitation to the method, in that some precursor acetylated
glycosyl bromides tend to react with the solvent (methanol)
before reduction can occur. This encouraged us to modify the
reaction conditions.
We were interested in the synthesis of various glycals and
chose to use the catalytic method, namely
zinc/methanol/ammonium chloride/vitamin B12.[8] This
redox system works well but has one drawback in that the
catalyst, vitamin B12, is expensive (Aldrich, US$50 g–1). One
of us (A.G.W.), whilst dallying in the undergraduate
Treatment of the bromide (1) with zinc and a slight molar
excess of acetic acid in acetonitrile containing VO(salen)
gave the D-glucal (2). This modified procedure was slower
than the original (using methanol/ammonium chloride) but
appeared just as general (Table 1).
laboratory,
noted
the
easy
synthesis
of
The synthesis of furanoid glycals proved to be a more
difficult task. Treatment of the reactive D-ribofuranosyl
bromide (4) under the original reaction conditions
[Zn/VO(salen)/methanol/ammonium chloride] resulted in
the formation of appreciable amounts of the methyl
D-ribofuranosides. Using the modified conditions
[Zn/VO(salen)/acetonitrile/acetic acid], the bromide (4) was
converted into a mixture of the D-ribofuranosyl acetates.
However, by increasing the amounts of both the zinc and
ammonium chloride in the methanol procedure, a good yield
bis(acetylacetonato)oxovanadium(IV) {VO(acac)2}, which
is also commercially available from Aldrich. Treatment of
tetra-O-acetyl-α-D-glucopyranosyl bromide (1) (Table 1)
with zinc and ammonium chloride in methanol containing
one mole percent of VO(acac)2 gave tri-O-acetyl-D-glucal
(2), rapidly and apparently ‘quantitatively’ (by thin-layer
chromatography (TLC)). The isolation of the D-glucal was
hampered somewhat by the instability of the vanadium
complex on silica gel during flash chromatography.
*Ethylene-N,N´-bis(salicylideneiminato)oxovanadium(IV).
© CSIRO 2002
10.1071/CH02011
0004-9425/02/01083