50 J. CHEM. RESEARCH (S), 1998
The Stereochemistry of Epoxidation of D5-Steroids with
Sodium Perborate and Potassium Permanganate†
James R. Hanson,* Nicolas Terry and Cavit Uyanik
J. Chem. Research (S),
1998, 50–51†
School of Molecular Sciences, University of Sussex, Brighton, Sussex, UK, BN1 9QJ
Sodium perborate, with potassium permanganate as a catalyst, has been shown to be a novel reagent for the epoxidation
of steroidal 5-enes with the attack occurring predominantly on the b-face.
Table 1 Epoxidation of steoridal D5-alkenes
The epoxidation of steroidal 5-enes with peracids takes place
predominantly from the a-face of the molecule to afford the
5a,6a-epoxides.1 Recently there has been an effort to pre-
pare2,3 the biologically interesting but relatively inaccessible
5b,6b-epoxides. A number of groups4–9 have shown that these
epoxides can be obtained from the 5-enes using the biphasic
systems involving potassium permanganate and transition
metal nitrates or sulfates. Many years ago, it was shown–10
that potassium permanganate in acetic acid would epoxidize
3b-acetoxyandrost-5-en-17-one (1) although at the time the
stereochemistry of the epoxides was unknown. We have now
repeated this work and shown that the major product was the
5b,6b-epoxide (2:1; b-epoxide:a-epoxide). The epoxides may
be clearly distinguished by the position of the 6-H resonance
in the 1H NMR spectrum (dH 2.87, a-epoxide; dH 3.07 b-epox-
ide).11 In this paper we report the catalytic use of potassium
permanganate in forming the b-epoxides.
Sodium perborate in glacial acetic acid provides an epoxi-
dizing agent for alkenes.12 With 1 it slowly gave a mixture of
the 5a,6a- and 5b,6b-epoxides, containing predominantly the
5a,6a-epoxides (ca. 4:1; a:b-epoxides) paralleling the stereo-
chemical results obtained with other peracids.1 However, in
the presence of catalytic amounts of potassium permanga-
nate, the reaction was much faster and the stereoselectivity
was reversed with the b-epoxide now predominating. A
number of steroidal 5-enes were examined, including some
with b-substituents at C-4. The ratios of the epoxides that
were formed are given in Table 1. In the case of 1 some
cleavage of the epoxide and allylic oxidation also took place.
A similar oxidation has been reported with the permanga-
nate–periodate reagent in pyridine.13 Interestingly, the
5a,6a-epoxide, identical to the product of peracid oxidation,
was obtained from the B-nor steroid, 3b-acetoxy-7-nor-
androst-5-en-17-one (2).
Compound
Ratio a:b-epoxide
Cholesteryl acetate
1:5
1:4
3b-Acetoxyandrost-5-en-17-one (1)
3b-Acetoxyandrost-5-ene
b-Epoxide only
b-Epoxide only
1:4
1:6.5
a-Epoxide only
3b,17b-Diacetoxyandrost-5-ene
3b-Acetoxy-4b-hydroxyandrost-5-en-17-one
4b-Acetoxy-3b-hydroxyandrost-5-en-17-one
3b-Acetoxy-7-norandrost-5-en-17-one (2)
form an epoxide in the second step was facilitated by the
metal sulfate.5,7 In the six-membered ring B of the steroids,
the axial position at C-6 is b-oriented whilst in the five-mem-
bered 7-nor series the pseudo-axial position is a-oriented.
This interpretation has been challenged8 and the alternative
view has been proposed that prior complexation by the metal
sulfate on the less hindered face of the alkene occurs, directly
the permanganate to the more hindered face of the molecule.
Although we also considered this7 it has difficulty in explain-
ing why the 7-nor steroid affords the same epoxide with both
peracid and permanganate. In the case of the perborate sys-
tem the role of the perborate/acetic acid (peracetic acid) is to
re-oxidize the manganese. This reaction is faster than the
peracetic acid epoxidation.
In conclusion the potassium permanganate/sodium per-
borate/glacial acetic acid reagent is a novel, cheap epoxidiz-
ing system that in this instance has afforded epoxides, albeit
in moderate yield, that differ in their stereochemistry from
those formed by conventional peracids.
Experimental
Experimental details have been described previously.5
General Experimental Procedure.•Sodium perborate (1.1 g) was
dissolved in glacial acetic acid (15 cm3) with gentle warming
ꢀ50 °C. Potassium permanganate (80 mg) in water (1 cm3) was
added to a solution of the steroid (900 mg) in glacial acetic acid (10
cm3). The sodium perborate solution was then added in portions
(2.5 cm3) over a period of 1 h. The mixture was left to stand at room
temperature overnight. It was poured into aqueous sodium hydro-
gen carbonate and the products were recovered in ethyl acetate.
The extract was washed with aqueous sodium sulfite, aqueous
sodium hydrogen carbonate and water, and dried over sodium
sulfate. The solvent was evaporated to give a gum, which was
O
O
AcO
AcO
1
2
1
assayed by H NMR for its epoxide content [ratio of signals at dH
These results confirm the earlier observations10 that the
epoxidation on the b-face of a steroidal 5-ene occurs with
potassium permanganate and show that the b-epoxidation
does not have an absolute requirement for a metal sulfate.
We have suggested previously that the stereochemical-deter-
mining feature of the potassium permanganate–metal sulfate
epoxidation is the kinetically preferred pseudo-axial attack of
the electron-deficient manganese, in a Markownikov sense,
on the alkene to form a manganate, the collapse of which to
2.87 (a) to 3.07 ppm (b)] and separated by chromatography on
silica by elution with increasing concentrations of ethyl acetate in
light petroleum (bp 60–80°C). The epoxides were identified by
their mps and 1H NMR spectra.
3b-Acetoxycholest-5-ene (900 mg) gave the starting material (85
mg), 3b-acetoxy-5b,6b-epoxycholestane (295 mg)14 and 3b-acetoxy-
5a,6a-epoxycholestane (62 mg).14
3b-Acetoxyandrost-5-en-17-one (1) (900 mg) gave the starting
material (79 mg), 3b-acetoxy-5b,6b-epoxyandrostan-17-one (292
mg),15 3b-acetoxy-5a,6a-epoxyandrostan-17-one (67 mg),15 3b-acet-
oxyandrost-5-ene-7,17-dione (30 mg)16 and 3b,6b-diacetoxy-
5a-hydroxyandrostan-17-one (73 mg).10
3b-Acetoxyandrost-5-ene (900 mg) gave the starting material
(105 mg) and 3b-acetoxy-5b,6b-epoxyandrostane (345 mg).
3b,17b-Diacetoxyandrost-5-ene (500 mg) gave the starting
material (53 mg) and 3b,17b-diacetoxy-5b,6b-epoxyandrostane
(104 mg).17
*To receive any correspondence.
†This is a Short Paper as defined in the Instructions for Authors,
Section 5.0 [see J. Chem. Research (S), 1998, Issue 1]; there is there-
fore no corresponding material in J. Chem. Research (M).