3974
M. Carballido et al. / Tetrahedron Letters 42 (2001) 3973–3976
O
O
O
R1O
HO
R1O
HO
O
H
OsO4 cat
H
+
O
O
R1O
H
OR2
OH
R3O
H
11
10
OR2
OR2
5: R1= TBS; R2= Bz
6: R1= H; R2= Bz
8: R1= TBS; R2= H
9: R1= TBS; R2= Bn
a: R1= TBS; R2= Bz; R3= H
b: R1= TBS; R2= H; R3= H
c: R1= TBS; R2= Bn; R3= H
d: R1= H; R2= Bz; R3= H
e: R1= TBS; R2= H; R3= Bz
f: R1= H; R2= H; R3= Bz
a: R1= TBS; R2= Bz
b: R1= TBS; R2= H
c: R1= TBS; R2= Bn
d: R1= H; R2= Bz
Scheme 3.
In this communication, we report the synthesis of
carba-sugars 3 and 4 starting from a cheap and com-
mercially available (−)-quinic acid (1) as a chiral tem-
plate,3 by cis-hydroxylation of the 5-cyclohexene
derivative 2 (Scheme 1).
Kishi noted that osmium tetroxide approaches prefer-
entially to the face of the double bond opposite to the
pre-existing allylic hydroxyl or alkoxyl group.8 How-
ever, if there is more than one allylic substituent, the
stereochemical results are not so easily predicted. In our
case, the cis-hydroxylation also depends on the reaction
conditions. Intrigued by this result, we turned our
attention to study the osmylation reaction with a series
of alkenes and with different co-oxidants for the regen-
eration of the osmium tetroxide.
The cyclohexene derivative 5 was obtained in three
steps and 47% overall yield from (−) quinic acid (1), as
described by Bartlett et al. (Scheme 2).4 Treatment of
the benzolactone 5 with sodium methoxide or potas-
sium carbonate in methanol afforded, in both cases, a
mixture of the methyl ester 75 and the hydroxycarbolac-
tone 8.5 The main product during the reaction course
was the methyl ester 7; however, after work-up, the
reaction mixture was equilibrated to an approximate
1:1 mixture of 7 and 8. Treatment of diol 7 with an
equivalent of sodium hydride yielded lactone 8 in 83%
yield.
The major product arising from the diprotected ben-
zoylalkene 5 arose from Si-hydroxylation when NaIO4
or NMO were the co-oxidants (entries 1 and 2, Table
1), whereas OsO4/M3NO oxidation gave predominantly
Re-hydroxylation. In the latter case, compound 12 was
also formed. This is an intermediate in the equilibrium
process in acid media between lactones 11a and 13
(Scheme 4). TBS deprotection of 5 afforded hydroxy-
alkene 6, which was subjected to osmylation and gave
an approximately 1:1 mixture of Re- and Si-hydroxyla-
tion in low yield (entries 4 and 5, Table 1). Benzylation
of 8 afforded 9. This compound was not hydroxylated
with OsO4/NaIO4, but afforded mainly the Re-hydroxy-
lation product using NMO as co-oxidant (entry 11,
Table 1). Finally, removal of the lactone bridge, as in 7,
The key step of the synthesis of 3 and 4 is the introduc-
tion of the hydroxyl groups in the 2- and 3-positions
using osmium tetroxide (Scheme 3). This paper
describes a study of various conditions and reagents.
The results are summarized in Table 1.6 Different co-
oxidants for the regeneration of osmium tetroxide were
investigated (potassium ferrocyanide, N-methylmor-
pholine oxide, N-trimethylamine oxide and sodium
periodate). No reaction was observed when potassium
ferrocyanide, ADmix-a or ADmix-b was used. Sodium
periodate gave cleaner reaction mixtures than both
N-methylmorpholine oxide and N-trimethylamine
oxide.
Table 1. Osmium tetroxide oxidationa
Entry
Substrate
Method
Hydroxylation
Si (%) Re (%)
Hydroxylation of the allylic alcohol 8 (entries 7, 8 and
9, Table 1) gave a mixture of two hydroxylated prod-
ucts, 10b and 11b, resulting from the reaction of the Si
and Re faces, respectively. When the addition occurs
from the Re face, migration of the carbolactone from
the 1,5- to the 1,3-position was also observed. Surpris-
ingly, the proportional ratio of Si and Re cis-hydroxy-
lation was very dependent on the co-oxidant.
Hydroxylated lactone 11b was the major diastereoiso-
mer in the OsO4/NaIO4 oxidation (entry 7, Table 1)
proceeding from the sterically more congested face of
the molecule, which has been demonstrated to be the
more sterically hindered one.7 In contrast, hydroxylated
lactone 10b was the major diastereoisomer in the OsO4/
NMO oxidation (entry 8, Table 1).
1
2
3
4
5
6
7
8
5
5
5
6
6
6
8
8
8
9
9
A
B
C
A
B
C
A
B
C
A
B
71
64
19
22
23
–
26
B5
57
19
24
–
b
b
12
61
36
–
55
18
18
–
9
10
11
c
c
B5
80
a Ratio determined from isolated compounds. Method A: NaIO4; B:
NMO; C: Me3NO.
b Complex mixture of products.
c No reaction.