OMe
OH
O
O
O
HO
HO
i, ii
i
7
i, ii
O
OH
OH
O
OH
O
OH
O
10
11(±)-pinitol
Br
O
O
Br
O
Br
O
O
Scheme 4 Reagents and conditions: i, K2CO3, MeOH, heat 2 h, 83%; ii,
Al2O3, MeOH, heat 24 h; iii, H2O, THF, HCl (cat), 60% from 10
+
+
O
O
OH
O
O
O
3-O-methyl ether of chiro-inositol, with both enantiomers
occurring in various plant sources. (+)-Pinitol has been shown
to possess significant hypoglycemic and antidiabetic activity in
diabetic albino mice.5 Previous syntheses of pinitol, either in
racemic form6 or of either antipode,7,8 have relied on the use of
cis-dienediols available by microbial oxidation of arenes using
Pseudomonas putida.
7
8
9
neo
chiro
Scheme 3 Reagents and conditions: i, OsO4 (cat.), hn, NaBrO3 (0.22 m),
15 °C; ii, acetone, TsOH, H2O, (7) 8%, (8) 6.7%, (9) 2.5%
Thus, as shown in Scheme 4, ring closure of the bromohydrin
7 to epoxide 10 was smoothly accomplished using K2CO3 in
MeOH (83%). The regiospecificity of the subsequent ring-
opening of the epoxide with alumina in MeOH is governed by
the necessity for a trans opening anti to the neighbouring
isopropylidene groups and finds precedent in Hudlicky’s
elegant enantiodivergent synthesis of both antipodes.8 An acidic
work up to this reaction led to in situ deprotection to afford
(±)-pinitol 11 in 60% isolated yield from 10. The overall
sequence therefore requires three operations and five discrete
steps from benzene.
In summary, the present study has shown that the influence of
temperature in the catalytic photoinduced charge transfer
osmylation of benzene may be usefully controlled, either in the
reaction using chlorate anion as oxidant to suppress tertiary
osmylation at lower temperature and hence favourer conduritol
E production, or to influence the stereochemical outcome in the
bromate reaction leading to neo and chiro deoxybromoinosi-
tols.
slower at lower temperatures and that the deoxybromoinositols
2 and 3 were formed by addition of hypobromous acid to
conduritol E 6 during the work up stage, which involved a
reductive quench of bromate residues with sodium metabisul-
fite. Indeed, this very combination has recently been shown to
be an excellent method for bromohydrin formation.2 Some
support for this hypothesis came from an experiment in which
a vast excess of cyclohexene was added as a sacrificial alkene
following a photoosmylation at 2 °C, and prior to the addition of
metabisulfite. In this instance, the relative proportion of the
isolated bromohydrin acetates was significantly lower than
before [2 + 3; 6.2%, compare 12.6% (Table 1)], and counterbal-
anced by the corresponding isolation of conduritol E (tetra-
acetate, 8.6%).
While the results presented in Table 1 clearly demonstrate
that diastereoisomeric crossover has occurred, it should be
noted that the strongly ionising conditions employed in the
acetylation step almost certainly promote neighbouring group
participation.3 This aspect was highlighted when the products
from a bromate driven photoosmylation reaction at 15 °C were
isolated as their isopropylidene derivatives by treatment of the
crude reaction mixture with acetone and toluene-p-sulfonic
acid. (Scheme 3).
We thank the EPSRC for the award of a postdoctoral
fellowship (P. M. J. J.) and B. P. for the award of a studentship
(A. S. W.). We also wish to acknowledge stimulating and
helpful discussions with Dr H. A. J. Carless.
The supposition that the isolated ratio of deoxybromocycli-
tols from this experiment [7 + 8 (neo, 14.7%):9 (chiro, 2.5%);
5.9:1] is probably a more appropriate reflection of the initial
product distribution prior to derivatisation was also supported
by a further blank experiment in which reaction of conduritol E
6 with sodium bromage–sodium metabisulfite was followed by
neutralisation with sodium hydroxide and subsequent acetyla-
tion under less ionising acidic conditions with acetic anhydride
in acetic acid to give 2 and 3 (72%, ratio 2:3, 6.3:1).
From a preparative standpoint, irrespective of the fact that
two competing pathways are operating, it is apparent that
experimental conditions can be selected to favour either the neo
or the chiro series as the major diastereoisomer. Moreover, in
practical terms, a single recrystallisation of a mixture of 2 and 3
from EtOH at room temperature gives pure 3, while advantage
can be taken of the known relative insolubility of neo-inositol4
to precipitate the parent bromocyclitol from 2 by simple acid
catalysed hydrolysis using HCl in MeOH.
Footnote
* E-mail: w.b.motherwell@ucl.ac.uk
References
1 W. B. Motherwell and A. S. Williams, Angew. Chem., Int. Ed. Engl.,
1995, 34, 280.
2 H. Masuda, K. Takase, M. Nishio, A. Hasegawa, Y. Nishiyama and
Y. Ishii, J. Org. Chem., 1994, 59, 5550.
3 For examples in the cyclitol area, see G. E. McCasland and E. C.
Horswill, J. Am. Chem. Soc., 1953, 75, 4020 and references cited
therein.
4 S. J. Angyal and D. C. Craig, Carbohydr. Res., 1994, 263, 149.
5 C. R. Narayanan, D. D. Joshi, A. M. Mujumdar and V. V. Dhekne, Curr.
Sci., 1987, 56, 139.
6 S. V. Ley, F. Sternfeld and S. Taylor, Tetrahedron Lett., 1987, 28, 225;
H. A. J. Carless, J. R. Billinge and O. Z. Oak, Tetrahedron Lett., 1989, 30,
3113.
7 S. V. Ley, F. Sternfeld and S. Taylor, Tetrahedron, 1989, 45, 3463.
8 T. Hudlicky, J. D. Price, F. Rulin and T. Tsunoda, J. Am. Chem. Soc.,
1990, 112, 9439.
The ready availability of these protected building blocks in a
single one-pot operation offers a variety of possibilities for
further synthetic manipulation in the important cyclitol area. By
way of illustration, we elected to prepare (±)-pinitol 11, the
Received in Cambridge, UK, 6th May 1997; Com. 7/03071A
1284
Chem. Commun., 1997