Table 1 Epoxidation of allylic alcohols with H2O2–metal catalyst in
aqueous neutral solutions of glycosidic additives at + 2 °C
the epoxidation reaction. Some significant results for the
epoxidation of representative cyclic and acyclic allylic alcohols
are listed in Table 2. The data indicate that acyclic alcohols 4–8
gave mixtures of erythro/threo epoxides. Nevertheless, epox-
idation in glycosidic aqueous media generates stereoselectively
erythro enriched epoxy alcohols. This outcome is the same as
for metal catalysed systems in organic solvents but is opposite
to that obtained using peroxyacids in organic media.16
Con-
Metal Reaction version ketone
Entry Alcohol Additive catalyst time (h) (%) ratio
Epoxide: Isolated
yield
(%)
1
2
3
4
5
6
7
8
9
10
11
12b
13
14
15
16
4
5
1
Mo+6
Mo+6
Mo+6
W+6
24
72
72
24
72
72
48
48
24
48
48
48
72
24
48
24
100
78
30
100
92
97
94
98
97
100
100
100
100
70
> 99+1 75
> 99+1a 72
> 99+1 28
95+5 83
> 99+1 83
> 99+1 88
> 99+1 92
> 99+1 92
90+10 80
97+3 86
97+3 85
98+2 84
83+17 80
87+13 66
60+40 68
85+15 67
1
5
—
1
In the case of cyclohex-2-enols 11–13 the epoxidation was
highly stereoselective when compared with the results obtained
by a peracid epoxidation (Table 2). In particular, allylic alcohols
11–12 gave exclusively the product where epoxide and OH
functions are cis to each other when the reactions were
6
7
1
Mo+6
Mo+6
Mo+6
Mo+6
Mo+6
W+6
8
2
9
1
10
11
11
11
11
12
13
13
13
1
performed in aqueous solutions of -fructose derivatives 3a or
D
2
3b in the presence of Mo+6 or W+6 catalyst. Once again,
comparison of cis+trans ratio underlines the role played by
arabinose as additive (entries 7 and 8).
3a
3b
3b
1
W+6
W+6
W+6
The predictable sense of the addition to cyclic olefins can be
rationalised by means of a model where hydrophobic inter-
actions are probably involved between the most hydrophobic
face of the cyclic alcohol and that of sugar additive. These
interactions may account for the enhanced stereoselective
hydroxy directed epoxidation of cyclic allylic alcohols observed
herein.
2
—
3a
Mo+6
Mo+6
Mo+6
72
70
a Oxirane oxidation in CH2Cl2 gave 51+49.14b Recycled aqueous solution
from entry 11.
In conclusion, it was shown that dilute hydrogen peroxide
together with molybdenum(VI) or tungsten(VI) salts is an
efficient system for the chemo-, regio- and stereoselective
epoxidation of allylic alcohols in aqueous solutions of carbohy-
drates characterised by a facial amphiphilicity.
but highly soluble in 1 M sucrose solutions. A site-selective
epoxidation occurred at the least electron-rich allylic double
bond and pure 2,3-epoxy-geraniol and -nerol were isolated in
high yields (entries 7, 8). Unfortunately, epoxidation of geraniol
9 in the presence of sucrose gave 2,3-epoxygeraniol with
modest enantioselectivity (ee = 10% determined by polari-
metry). Our results obtained in neutral media therefore compare
favourably with previous work performed with peracids in
strongly alkaline media13 or with transition metal catalysed
epoxidations in organic solvents.15,16
Notes and references
1 For a review, see for instance: A. S. Rao, in Comprehensive Organic
Synthesis, ed. B. M. Trost and I. Fleming, Pergamon, Oxford, 1991, Vol.
7, pp. 357–387.
Oxidations of six-membered cyclic substrates (Table 1,
entries 9–15) indicated that enone formation competed with
epoxidation. The ratio of oxirane to ketone showed a pro-
nounced dependence on both substrates and reaction conditions
as exemplified by entries 14 and 15. The best chemoselectivities
were attained by using tungsten(VI) catalyst in the presence of
sugar additives. Moreover, it is important to mention that
solutions containing fructosides and catalyst could be recycled
for further experiments without loss of conversion yield,
chemoselectivity and stereoselectivity (entry 12).
The results presented here suggest OH-assisted epoxidation
of allylic alcohols, presumably related to the formation of an
intermediate in which the olefinic alcohol is most likely
coordinated to the metal via the hydroxy group. Such an
intermediate was previously suggested for tungsten catalysed
epoxidations in protic media.16,17 The next step was therefore to
ascertain whether sugar–substrate and/or intermediate inter-
actions were strong enough to promote stereodiscrimination in
2 C. J. Li, Chem. Rev., 1993, 93, 2023; A. Lubineau, J. Augé, H.
Bienaymé and Y. Queneau, Synthesis, 1994, 741; A. Lubineau, Chem.
Ind., 1996, 123.
3 C. Denis, B. Laignel, D. Plusquellec, J. Y. Le Marouille and A. Botrel,
Tetrahedron Lett., 1996, 37, 53.
4 G. Strukul, in Catalytic Oxidation with Hydrogen Peroxide as Oxidant,
Kluwer Academic Publishers, 1992 R. A. Sheldon, in Applied
Homogeneous Catalysis with Organometallic Compounds, ed. B.
Cornils and W. A. Hermans, VCH Weinheim, 1996J. Brinksma, R.
Hage, J. Kerschner and B. L. Feringa, Chem. Commun., 2000, 537.
5 K. Sato, M. Aoki, M. Ogawa, T. Hasimoto and R. Noyori, J. Org.
Chem., 1996, 61, 8310.
6 S. Immel and F. W. Lichtenthaler, Liebigs Ann., 1995, 1925; S. B
Engelsen, C. Hervé du Penhoat and S. Pérez, J. Phys. Chem., 1995, 99,
13334.
7 J. C. Hanson, L. C. Sieker and L. H. Jensen, Acta Crystallogr., Sect. B,
1973, 29, 797; G. M. Brown and H. A. Levy, Acta Crystallogr., Sect. B,
1973, 29, 790.
8 J. C. Christofides and D. B. Davies, Chem. Commun., 1985, 1553.
9 MAD V23, Oxford Molecular Ltd, Magdalen Centre, Oxford Science
Park, Oxford OX4 4GA, England.
Table 2 Diastereoselective epoxidations of allylic alcohols in aqueous
neutral solutions of carbohydrates 1–3
10 F. W. Lichtenthaler and S. Immel, Int. Sugar J., 1995, 97, 12.
11 S. A. Galena, M. J. Blandamer and J. B. F. N. Engberts, J. Org. Chem.,
1992, 57, 1995.
12 V. Ferrières, T. Benvegnu, M. Lefeuvre, D. Plusquellec, G. Mackenzie,
M. J. Watson, J. A Haley, J. W. Goodby, R. Pindak and M. K. Durbin,
J. Chem. Soc., Perkin Trans. 2, 1999, 951.
dra
mCPBA or
Metal
catalyst (cis+trans)
erythro+threo analogues
Entry
Substrate Additive
(lit.)
13 F. Fringuelli, R. Germani, F. Pizzo, F. Santinelli and G. Sanelli, J. Org.
Chem., 1992, 57, 1198; M. Nakaruma, N. Tsutsui and T. Takeda,
Tetrahedron Lett., 1984, 25, 3231.
1
2
3
4
4
5
1
Mo+6
Mo+6
Mo+6
Mo+6
W+6
68+32
68+32
30+7013
38+6214
36+6414
(95+5)18
(95+5)18
1
8
1
2
62+38
14 B. E. Rossiter, T. R. Verhoeven and K. B. Sharpless, Tetrahedron Lett.,
1979, 4733; W. Adam, F. Prechtl, M. J. Richter and A. K. Smerz,
Tetrahedron Lett., 1993, 34, 8427.
11
11
12
12
13
13
(99+1)
3a or 3b
( > 99+1)
(95+5)
5
6
7
8
1
W+6
15 K. B. Sharpless and R. C. Michaelson, J. Am. Chem. Soc., 1973, 95,
3a or 3b
W+6
Mo+6
Mo+6
( > 99+1)
(94+6)
6136.
2
—
(90+10)
16 K. B. Sharpless and T. R. Verhoeven, Aldrichimica Acta, 1979, 12,
63.
(89+11)
17 D. Prat and R. Lett, Tetrahedron Lett., 1986, 27, 707; D. Prat, B.
Delpech and R. Lett, Tetrahedron Lett., 1986, 27, 711.
18 P. Chantemps and J. L. Pierre, Tetrahedron, 1976, 32, 549.
a dr: diastereoisomeric ratio were determined by GLC. erythro/threo refers
to acyclic substrates whereas cis/trans to cyclic products.
Chem. Commun., 2001, 2460–2461
2461