DOI: 10.1002/chem.201104007
Organotin-Catalyzed Highly Regioselective Thiocarbonylation of
Nonprotected Carbohydrates and Synthesis of Deoxy Carbohydrates in a
Minimum Number of Steps
Wataru Muramatsu,* Satoko Tanigawa, Yuki Takemoto, Hirofumi Yoshimatsu, and
Osamu Onomura[a]
The selective functionalization of polyols remains one of
the most fundamental challenges for achieving the efficient
synthesis of building blocks for natural product synthesis or
new drug development.[1] The direct regioselective function-
alization of the secondary hydroxy group in carbohydrates
with nonenzymatic catalysts is of particular interest due to
the difficulty in functionalizing one specific hydroxy group
of the multiple present in carbohydrates. Over the last sever-
al decades, progress has been made in the catalytic regiose-
lective acylation,[2] alkylation,[3] sulfonylation,[4] and glycosyl-
ation[5] of monosaccharides in the presence of an organome-
tal catalyst or organocatalyst. Additionally, our group re-
cently reported the regioselective acylation of monosaccha-
rides by using an organotin catalyst.[2g] These catalytic
methods are useful techniques for the protection or func-
tionalization of a hydroxy group in carbohydrates in a mini-
mum number of steps. However, some of these methods are
not applicable to the monosaccharides with a nonprotected
primary hydroxy group,[2j–k,3–5] and the resulting monofunc-
tionalized carbohydrates are usually not the ideal precursors
for further functionalization.[2–3] To resolve such problems,
we began an investigation of the catalytic regioselective
thiocarbonylation of nonprotected monosaccharides. This
reaction would afford monothiocarbonates suitable for use
as electrophiles in the synthesis of biologically interesting
deoxy monosaccharides through the venerable Barton–
McCombie deoxygenation[6] and new drug candidates by
phenoxythiocarbonate in 83% yield with 94% regioselectiv-
ity.[8] However, this method is limited to the transformation
of a select few monosaccharides (a-d-Glc, b-d-Glc, a-d-Xyl,
and b-d-Xyl), and requires a stoichiometric amount of
Bu2SnO (1.5 equiv) under harsh conditions for the prepara-
tion of the tin acetal intermediate. In addition, the deoxy-
genation of nonprotected monothiocarbonates for the syn-
thesis of deoxy saccharides has never been reported even by
Tsuda.[9] Miller and co-workers also reported impressive or-
ganocatalytic thiocarbonylation of carbohydrates with good
selectivity. However, this method requires a protecting
group on primary hydroxy groups of carbohydrates for the
selectivity.[10] Herein, we report the first catalytic regioselec-
tive thiocarbonylation of nonprotected monosaccharides by
using organotin dichloride under mild conditions, and the
synthesis of representative deoxy monosaccharides in a mini-
mum number of steps through the Barton–McCombie de-
oxygenation.
After a series of optimization studies, we found that the
À
selective thiocarbonylation at C(2) OH of methyl a-d-glu-
copyranoside proceeded efficiently in the presence of
10 mol% of Oc2SnCl2, 10 mol% of tetrabutylammonium
iodide (TBAI), 1.3 equiv of phenyl chlorothionoformate,
and 1.5 equiv of 1,2,2,6,6-pentamethylpiperidine (PEMP)[10]
as a less nucleophilic base in THF at 208C (Table 1, entry 1;
98% yield, no regioisomers or other derivatives). Table 1
provides information about the effect of a number of reac-
tion parameters on the efficiencies of thiocarbonylation at
À
C C bond formation or halogenation under radical reaction
conditions.[7]
C(2) OH of methyl a-d-glucopyranoside. In the absence of
À
Tsuda, a pioneer in the field of regioselective thiocarbon-
ylation of carbohydrates, and co-workers developed a useful
approach for the selective introduction of a phenoxy-
thiocarbonyl group at a secondary hydroxy group of methyl
a-d-glucopyranoside by the use of Bu2SnO, which gave 2-O-
Oc2SnCl2 or PEMP, essentially no reaction was observed
(Table 1, entries 2 and 3). In the absence of TBAI, the yield
was slightly lower due to failure to activate the phenyl chlo-
rothionoformate as an electrophile (Table 1, entry 4; 93%
yield). The use of other alkyl or aryl tin dichlorides instead
of Oc2SnCl2 gave the desired product 1 in lower yield
(Table 1, entries 5–7), as did the use of an alkyl tin oxide as
a catalyst or a smaller amount of catalyst (Table 1, entries 8
and 9). Nucleophilic catalysts, such as N,N-dimethylamino-
pyridine (DMAP) and N-methylimidazole (NMI), deactivat-
ed the electrophile (Table 1, entries 10 and 11). The addition
[a] Prof. Dr. W. Muramatsu, S. Tanigawa, Y. Takemoto, H. Yoshimatsu,
Prof. Dr. O. Onomura
Graduate School of Biomedical Sciences
Nagasaki University
1-14 Bunkyo-machi, Nagasaki 852-8521 (Japan)
[10]
of FeCl3 as a co-catalyst had no significant effect under
the present reaction conditions (Table 1, entry 12; 40%
yield). Dimethoxyethane (DME) can be used instead of
Supporting information for this article is available on the WWW
4850
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4850 – 4853