The effect of 4beta and 19 ester functionalities on some electrophilic addition reactions of delta5-steroids.

Article abstract of DOI:10.1016/S0039-128X(99)00062-8

The reactions of 3beta-acyloxyandrost-5-enes with bromine/silver acetate (Petrow reaction) and mercury(II) trifluoroacetate (modified Treibs oxidation) have been used previously to effect allylic oxidation on these substrates en route to biologically active compounds. In both these reactions, which involve electrophilic addition to the delta5-bond, the 3-acyloxy substituent plays a significant role. In this report, the effect of introducing other substituents proximate to the delta5-bond has been studied by using derivatives of 3beta-acetoxyandrost-5-en-17-one (1), namely, 3beta,4beta-diacetoxyandrost-5-en-17-one (13), 3beta,19-diacetoxyandrost-5-en-17-one (14), 3beta-acetoxyandrost-5-ene-7,17-dione (15), and 3beta-acetoxy-4,4-dimethylandrost-5-en-17-one (17). Our results indicate that in both sets of reactions the effect of the introduced functional groups was pronounced. In the Petrow reaction, electrophilic addition rather than allylic oxidation on the diacetates was observed. With the Treibs reaction, allylic oxidation on the diacetates occurred. The 7-keto and 4,4-dimethyl steroids proved to be poor substrates in both reactions.

Full text of DOI:10.1016/S0039-128X(99)00062-8

Steroids 64 (1999) 812–819
The effect of 4␤ and 19 ester functionalities on some electrophilic
addition reactions of ⌬5-steroids
Peter L.D. Ruddock, Paul B. Reese*
Department of Chemistry, University of the West Indies, Mona, Kingston 7, Jamaica
Received 8 July 1998; received in revised form 4 March 1999; accepted 6 April 1999
Abstract
The reactions of 3␤-acyloxyandrost-5-enes with bromine/silver acetate (Petrow reaction) and mercury(II) trifluoroacetate (modified
Treibs oxidation) have been used previously to effect allylic oxidation on these substrates en route to biologically active compounds. In both
these reactions, which involve electrophilic addition to the ⌬5-bond, the 3-acyloxy substituent plays a significant role. In this report, the
effect of introducing other substituents proximate to the ⌬5-bond has been studied by using derivatives of 3␤-acetoxyandrost-5-en-17-one
(1), namely, 3␤,4␤-diacetoxyandrost-5-en-17-one (13), 3␤,19-diacetoxyandrost-5-en-17-one (14), 3␤-acetoxyandrost-5-ene-7,17-dione
(15), and 3␤-acetoxy-4,4-dimethylandrost-5-en-17-one (17). Our results indicate that in both sets of reactions the effect of the introduced
functional groups was pronounced. In the Petrow reaction, electrophilic addition rather than allylic oxidation on the diacetates was observed.
With the Treibs reaction, allylic oxidation on the diacetates occurred. The 7-keto and 4,4-dimethyl steroids proved to be poor substrates in
both reactions. © 1999 Elsevier Science Inc. All rights reserved.
Keywords: ⌬5-Steroid; Treibs; Petrow; Bromine; Silver acetate; Mercury(II) trifluoroacetate
1. Introduction
[11,12] leads to the formation of 3␤-acetoxy-6␤-hy-
droxyandrost-4-en-17-one (9) (43% yield) and 3␤-acetoxy-
6-chloromercuriandrost-5-en-17-one (12) (6%) [4]. The
cholesterol analog of the latter has been implicated indi-
rectly for use in the treatment of Addison’s disease [13], a
potentially fatal affliction of the adrenal glands. A small
quantity of the ⌬4-6-ketone 10 (1.5%) was also formed
presumably via oxidation of 6␤-alcohol 9, with the rest of
the steroid apparently going to a polymer that remains on
the silica gel column. The first intermediate in the reaction
is thought to be the 5␣,6␣-mercurinium ion (6). Loss of the
4␤-proton [3] and opening of the mercurinium ion would
lead to the allylic organomercurial compound 7, which after
nucleophilic attack by trifluoroacetate on the ␤ face would
form the allylic ester 8. Hydrolysis of the trifluoroacetate on
workup would give the 6␤-alcohol 9 (Scheme 2). Alter-
nately, mercurinium ion 6 could suffer loss of the 6␤-
proton, giving the 6-trifluoroacetoxymercuri-⌬5-steroid 11,
which would undergo ligand exchange with the solvent to
form the chloromercury steroid 12 (Scheme 3) [4].
Our studies in the electrophilic substitution of 3-substi-
tuted-⌬5-steroids by bromine/silver acetate (Petrow reac-
tion) [1,2] and mercury(II) trifluoroacetate (modified Treibs
oxidation) [3–5] have led us to examine the effect of struc-
tural modification of the substrate on these reactions.
The treatment of 3␤-acetoxy-⌬5-steroids, e.g. 1, with
bromine at Ϫ78°C followed by silver acetate in pyridine
[6,7] produces mainly 4␤-acetoxy-3␤-hydroxy-⌬5-deriva-
tives, e.g. 5, in good yield. Several biologically active ste-
roids possess a C-4 oxygen function, some of which selec-
tively inhibit estrogen synthases in tumor cells [8–10]. The
mechanism of the Petrow reaction on 1 is believed to in-
volve the formation of the 5␣,6␤-dibromide 2, which loses
hydrogen bromide to give the ⌬4 compound 3. The latter
undergoes an anchimeric assisted SN2Ј reaction to give 5 via
the dioxolenium ion 4 (Scheme 1) [2].
The addition of dehydroepiandrosterone acetate (1) to a
solution of mercury(II) trifluoroacetate in dichloromethane
We decided to examine the effects of insertion of an
acetoxyl moiety at C-4␤ or C-19 on the product(s) of the
two above-mentioned reactions. In view of this, we also
* Corresponding author. Tel.: ϩ876-927-1910; fax: ϩ876-977-1835.
E-mail address: pbreese@uwimona.edu.jm (P.B. Reese)
0039-128X/99/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(99)00062-8

P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
813
Scheme 3.
in chloroform with a Perkin–Elmer 241 mc polarimeter
(Norwalk, CT, USA). Infrared spectra were recorded on a
Perkin–Elmer 1600 FT-IR and 735b spectrometers with
KBr disks being the primary medium employed. Mass spec-
tral data (high-resolution electron impact ionization [EI] and
chemical ionization [CI]) were determined at an ionizing
voltage of 70 eV on Kratos A.E.I. MS-50 and MS-12 spec-
trometers (Kratos Analytical Instruments, Chestnut Ridge,
NY, USA), respectively. 1H NMR spectra were recorded on
Bruker AM 300 and AC 200 instruments (Bruker, Billerica,
MA, USA) operating at 300 MHz and 200 MHz, respec-
tively, in the indicated solvent and were referenced to in-
ternal tetramethylsilane. 13C NMR spectra were determined
on the same instruments but at 75 MHz and 50 MHz,
respectively. The 13C data were compared where possible to
those published by Blunt and Stothers [14]. Petrol refers to
the petroleum fraction of boiling range 60–70°C.
Scheme 1. Reagents: (i) Br2/CHCl3. (ii) AgOAc/py Ϫ78°C.
sought to observe the route of the reactions on a 7-keto- (15)
and a 4,4-dimethyl-steroid (17).
2. Experimental
Melting points were determined on a Thomas–Hoover
melting point apparatus by using open-ended capillary tubes
or on a Ko¨fler hot stage instrument. Ultraviolet spectra were
recorded on a Pye Unicam PU8800 spectrophotometer. Ro-
tations were measured, unless otherwise stated, for solutions
3␤,4␤-Diacetoxyandrost-5-en-17-one (13) [15], 3␤,19-di-
acetoxyandrost-5-en-17-one (14) [16] and 3␤-acetoxyandrost-
5-ene-7,17-dione (15) [17] were prepared by literature meth-
ods. 3␤-Acetoxy-4,4-dimethylandrost-5-en-17-one (17) is a
new compound.
2.1. 3␤-Acetoxy-4,4-dimethylandrost-5-en-17-one (17)
(Scheme 4)
A solution of 17,17-ethylenedioxy-3␤-hydroxy-4,4-di-
methylandrost-5-ene (16) [18] (540 mg, 1.5 mmol) in eth-
anol (30 ml) containing dilute aqueous sulfuric acid (4 ml,
8.5% v/v) was refluxed for 1 h. The solvent was removed
under reduced pressure and the precipitated steroid was
redissolved in ethyl acetate. This solution was washed with
water and saturated sodium hydrogen carbonate solution
and was dried. Removal of the solvent on a rotary evapo-
rator afforded a white crystalline solid (410 mg), which was
dissolved in pyridine (5 ml) and treated with acetic anhy-
dride (2.5 ml, 26.5 mmol). The mixture was stirred over-
night at room temperature. Ethyl acetate was added and the
diluted solution was washed with cold dilute hydrochloric
Scheme 2. Reagents: (i) Hg(O2CCF3)2/CH2Cl2. (ii) aq NaHCO3.

814
P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
acid, water, and saturated sodium hydrogen carbonate so-
lution. The organic layer was dried and the solvent was
removed in vacuo to give a white crystalline residue, which
was chromatographed. Elution with 15% ethyl acetate in
petrol gave 3␤-acetoxy-4,4-dimethylandrost-5-en-17-one
(17) (402 mg, 1.22 mmol, 81%), which crystallized from
acetone as needles, m.p. 180–182°C, [␣]D Ϫ12° (c ϭ 0.09);
(C-8), 38.3 (C-1), 43.7 (C-9), 45.7 (C-10), 47.8 (C-13), 50.6
(C-14), 56.2 (C-6), 67.4 (C-4), 70.7 (C-3), 76.1 (C-5), 169.9
(CH3CO2-4␤)††, 170.2 (CH3CO2-3␤)††, 219.5 (C-17).
(*,†,†† interchangeable assignments).
Further elution gave 3␤,4␤-diacetoxy-5␣-bromo-6␤-hy-
droxyandrostan-17-one (18) (259 mg, 0.534 mmol, 23%)
which crystallized from ethyl acetate as needles, m.p. 166–
167°C (dec.), [␣]D Ϫ5° (c ϭ 0.10);
IR ␯
1740, 1725, 1252 cmϪ1
;
max
MS (EI) m/z (%) 358.2508 (Mϩ, 50), 298.2292 (47),
283.2061 (100), 280.1671 (38). (C23H34O3 requires
358.2508);
IR ␯
3400, 1730, 1240 cmϪ1
;
max
Combustion analysis: Found: C, 56.83%; H, 6.82%.
(C23H33BrO6 requires C, 56.91%; H, 6.85%);
MS (CI NH3) m/z (%) 504 (13), 502 (Mϩ, 14), 422 (14),
406 (37), 286 (23), 268 (100). (C23H3379BrO6NH4 requires
502);
1H NMR (CDCl3) ␦ 0.90 (3H, s, H-18), 1.05 (3H, s,
H-19), 1.15 (3H, s, CH3-4␤), 1.17 (3H, s, CH3-4␣), 2.08
(3H, s, CH3CO2), 4.50 (1H, dd, J ϭ 7, 10 Hz, H-3␣), 5.58
(1H, dd, J ϭ 3, 4 Hz, H-6);
1H NMR (CDCl3) ␦ 0.93 (3H, s, H-18), 1.64 (3H, s,
H-19), 2.09 (3H, s, CH3CO2-3␤)*, 2.10 (3H, s, CH3CO2-
4␤)*, 4.45 (1H, t, J ϭ 2 Hz, H-6␣), 5.68 (1H, t, J ϭ 3 Hz,
H-4␣), 5.78 (1H, m, w/2 ϭ 18 Hz, H-3␣). (* interchangeable
assignments);
13C NMR (CDCl3) ␦ 13.5 (C-18), 19.8 (C-11), 21.2
(C-19), 21.3 (CH3CO2), 21.7 (C-15), 23.7 (C-2), 24.9 (CH3-
4␤), 27.1 (CH3-4␣), 30.4 (C-8), 31.3 (C-7), 31.3 (C-12),
35.8 (C-16), 36.1 (C-1), 36.7 (C-10), 40.3 (C-4), 47.4 (C-
13), 50.8 (C-9), 52.1 (C-14), 79.1 (C-3), 119.9 (C-6), 149.3
(C-5), 170.7 (CH3CO2), 220.9 (C-17).
13C NMR (CDCl3) ␦ 13.9 (C-18), 18.4 (C-19), 19.9
(C-11), 21.1 (CH3CO2-4␤)*, 21.6 (C-15), 21.7 (CH3CO2-
3␤)*, 22.2 (C-2), 30.8 (C-8), 31.2 (C-12), 32.2 (C-1), 34.7
(C-7), 35.7 (C-16), 40.3 (C-10), 47.7 (C-13), 49.4 (C-9),
50.7 (C-14), 71.9 (C-6), 73.9 (C-4), 75.6 (C-3), 83.5 (C-5),
169.5 (CH3CO2-4␤)†, 170.5 (CH3CO2-3␤)†, 220.3 (C-17).
(*,† interchangeable assignments).
2.2. Petrow reaction on 3␤,4␤-diacetoxyandrost-5-en-17-
one (13)
3␤,4␤-Diacetoxyandrost-5-en-17-one (13) (0.92 g, 2.37
mmol) was dissolved in dichloromethane (10 ml) and the
solution was cooled to Ϫ78°C in the dark. Bromine (0.22
ml, 4.27 mmol) was added with stirring to the cold solution
followed by a suspension of silver acetate (0.99 g, 5.93
mmol) in pyridine (1.5 ml). The reaction mixture was stirred
in the dark for 10 h and allowed to attain room temperature.
The yellow-green suspension was filtered through celite and
the filtrate was washed with dilute hydrochloric acid, water,
and saturated sodium hydrogen carbonate solution and was
dried. The organic solution was evaporated to dryness and
the off-white residue was chromatographed (25% ethyl ac-
etate in petrol). 3␤,4␤-Diacetoxy-6␣-bromo-5␤-hydroxyan-
drostan-17-one (19) (213 mg, 0.439 mmol, 19%) was ob-
tained and crystallized from ethyl acetate as very small
needles, m.p. 203–205°C (dec.), [␣]D 62° (c ϭ 0.12);
2.3. Petrow reaction on 3␤,19-diacetoxyandrost-5-en-17-
one (14)
A solution of 3␤,19-diacetoxyandrost-5-en-17-one (14)
(1.75 g, 4.50 mmol) in dichloromethane was prepared and
cooled to Ϫ78°C in the dark. Bromine (0.30 ml, 5.95 mmol)
was added with stirring to the cold solution followed by a
suspension of silver acetate (2.26 g, 13.5 mmol) in pyridine
(3.5 ml). The reaction mixture was stirred overnight in the
dark and allowed to attain room temperature. The silver
salts were removed by filtration through celite and the
filtrate was washed with dilute hydrochloric acid, water, and
saturated sodium hydrogen carbonate solution, and was
dried. The resultant dark-brown foam was chromatographed
(25% ethyl acetate in petrol) to give starting material (167
mg, 0.430 mmol, 9.6%). Further elution afforded 3␤,19-
IR ␯
3525, 1730, 1250 cmϪ1
;
max
MS (CI NH3) m/z (%) 504 (89), 502 (Mϩ, 100), 422 (26),
345 (11), 329 (18), 285 (22), 269 (8). (C23H3379BrO6NH4
requires 502);
diacetoxy-6␣-bromo-5␤-hydroxyandrostan-17-one
(23)
MS (EI) m/z (%) 424.1230 (1), 405.2267 (4), 328.2030
(5), 285.1862 (38), 283.2062 (100). (C23H3379BrO6–
CH3CO2H requires 424.1249);
(437 mg, 0.900 mmol, 20%), which resisted crystallization,
IR ␯
3513, 1730, 1250 cmϪ1
;
max
MS (CI NH3) m/z (%) 504 (100), 502 (Mϩ, 99), 422 (40),
362 (12), 327 (45), 267 (27). (C23H3379BrO6NH4 requires
1H NMR (CDCl3) ␦ 0.86 (3H, s, H-18), 1.10 (3H, s,
H-19), 2.13 (3H, s, CH3CO2-3␤)*, 2.14 (3H, s, CH3CO2-
4␤)*, 4.50 (1H, d, J ϭ 9 Hz, H-6␤), 5.26 (1H, d, J ϭ 3 Hz,
H-3␣), 5.66 (1H, d, J ϭ 3 Hz, H-4␣). (* interchangeable
assignments);
502);
m
MS (EI)
/
(%) 486.1438 (3), 484.1453 (Mϩ, 3),
z
345.2063 (67), 326.1879 (55), 285.1846 (100).
(C23H3379BrO6 requires 484.1461);
1H NMR (CDCl3) ␦ 0.85 (3H, s, CH3-18), 2.10 (3H, s,
CH3CO2-19)*, 2.11 (3H, s, CH3CO2-3␤)*, 3.19 (1H, bs,
OH), 4.38 (1H, d, J ϭ 3.5 Hz, H-6␤), 4.61 (1H, d, J ϭ 3.8
13C NMR (CDCl3) ␦ 13.7 (C-18), 16.8 (C-19), 20.8
(C-11), 21.3 (CH3CO2-3␤)*, 21.6 (CH3CO2-4␤)*, 21.9 (C-
15), 23.7 (C-7)†, 26.3 (C-2)†, 31.2 (C-12), 35.7 (C-16), 37.0

P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
815
Scheme 4. Reagents: (i) aq H2SO4. (ii) Ac2O/py.
Hz, H-19), 4.68 (1H, d, J ϭ 3.8 Hz, H-19), 5.27 (1H, bs, w/2
ϭ 7 Hz, H-3␣). (* interchangeable assignments);
13C (CDCl3) ␦ 13.7 (C-18), 21.1 (C-11), 21.4 (CH3CO2-
3␤), 21.4 (CH3CO2-19), 21.6 (C-15), 23.8 (C-7), 31.4 (C-2),
31.8 (C-12), 35.6 (C-4, C-16), 36.6 (C-8), 38.8 (C-1), 42.9
(C-9), 45.9 (C-10), 47.7 (C-13), 51.5 (C-14), 60.8 (C-6),
65.2 (C-19), 69.3 (C-3), 74.8 (C-5), 169.3 (CH3CO2-19)*,
170.3 (CH3CO2-3␤)*, 219.5 (C-17). (* interchangeable as-
signments).
dark green and the amount of mercury(I) salts increased.
The mixture was filtered through celite and the filtrate was
washed with saturated sodium hydrogen carbonate solution
and was dried. Evaporation of the solvent in vacuo gave a
brown foam (1.13 g), which was chromatographed. Elution
with 30% ethyl acetate in petrol produced 3␤-acetoxy-6␤-
hydroxyandrost-4-en-17-one (9) (170 mg, 0.491 mmol,
9.5%) which was identified by comparison with an authen-
tic sample. Further elution gave 6␤-acetoxy-3␤-hydroxyan-
drost-4-en-17-one (30) [20] (102 mg, 0.294 mmol, 5.7%) as
a gum, which resisted crystallization,
Elution with 30% ethyl acetate in petrol afforded 3␤,19-
diacetoxy-5␤,6␤-epoxyandrostan-17-one (26) (502 mg,
1.241 mmol, 28%), which crystallized from ethyl acetate as
flat prisms, m.p. 119–120°C, [␣]D 14° (c ϭ 0.10), lit [19].
m.p. 133–134°C, [␣]D 26°;
IR ␯
3425, 1730, 1250 cmϪ1
;
max
1H NMR (CDCl3) ␦ 0.93 (3H, s, H-18), 1.18 (3H, s,
H-19), 2.06 (3H, s, CH3CO2), 4.15 (1H, dt, J ϭ 3, 8 Hz,
H-3␣), 5.30 (1H, t, J ϭ 4 Hz, H-6␣), 5.73 (1H, s, H-4);
13C NMR (CDCl3) ␦ 13.6 (C-18), 19.9 (C-11), 20.5
(C-19), 21.5 (CH3CO2), 21.5 (C-15), 28.6 (C-2), 30.4 (C-8),
31.1 (C-12), 35.6 (C-16), 35.8 (C-7), 36.6 (C-10), 36.7
(C-1), 47.5 (C-13), 50.8 (C-14), 53.8 (C-9), 67.4 (C-3), 75.1
(C-6), 132.4 (C-4), 141.2 (C-5), 170.1 (CH3CO2), 220.7
(C-17).
IR ␯
1735, 1240, 910, 840 cmϪ1
;
max
1H NMR (CDCl3) ␦ 0.88 (3H, s, H-18), 2.03 (3H, s,
CH3CO2-19), 2.11 (3H, s, CH3CO2-3␤), 3.05 (1H, d, J ϭ 2
Hz, H-6␣), 4.06 (1H, d, J ϭ 12 Hz, H-19), 4.55 (1H, d, J ϭ
12 Hz, H-19), 4.84 (1H, m, w/2 ϭ 23 Hz, H-3␣);
13C NMR (CDCl3) ␦ 13.4 (C-18), 21.2 (CH3CO2-3␤),
21.2 (CH3CO2-19), 21.3 (C-11), 21.7 (C-15), 27.3 (C-2),
30.1 (C-8), 31.2 (C-7), 31.6 (C-12), 32.3 (C-4), 35.6 (C-16),
38.1 (C-1), 38.3 (C-10), 47.4 (C-13), 50.0 (C-9), 51.3 (C-
14), 60.57 (C-6), 60.64 (C-5), 65.3 (C-19), 70.5 (C-3), 170.5
(CH3CO2-3␤)*, 170.7 (CH3CO2-19)*, 220.3 (C-17). (* in-
terchangeable assignments).
Elution with 40% ethyl acetate in petrol afforded 6␤-
acetoxy-3␣-hydroxyandrost-4-en-17-one (32) (107 mg,
0.309 mmol, 6.0%), which crystallized from acetone-petrol
as needles, m.p. 168–169°C, [␣]D 179° (c ϭ 0.11);
IR ␯
3450, 1720, 1250 cmϪ1
;
max
MS (CI NH3) m/z (%) 364 (Mϩ, 12), 329 (100), 286 (81),
2.4. Treibs oxidation of 3␤,4␤-diacetoxyandrost-5-en-17-
one (13)
269 (71) (C21H30O4NH4 requires 364);
m
MS (EI)
/
(%) 328.2035 (10), 286.1930 (100),
z
271.1697 (15). (C21H30O4–H2O requires 328.2039);
1H NMR (CDCl3) ␦ 0.92 (3H, s, H-18), 1.08 (3H, s,
H-19), 2.04 (3H, s, CH3CO2), 2.64 (1H, b s, OH), 4.13 (1H,
m, w/2 ϭ 14 Hz, H-3␤), 5.36 (1H, d, J ϭ 4 Hz, H-6␣), 5.88
(1H, d, J ϭ 5 Hz, H-4);
Mercury(II) oxide (2.79 g, 12.9 mmol) and trifluoroace-
tic anhydride (2.50 ml, 11.7 mmol) were stirred together in
dichloromethane (200 ml). 3␤,4␤-Diacetoxyandrost-5-en-
17-one (13) (2.00 g, 5.15 mmol) was added and the solution
was stirred for 4 days. Thin layer chromatographic analysis
of the orange colored reaction mixture suggested that the
starting material had not reacted, but small amounts of
mercury(I) deposits were seen in the reaction vessel. The
mixture was refluxed for 24 h, during which time it became
13C NMR (CDCl3) ␦ 13.6 (C-18), 19.5 (C-19), 20.4
(C-11), 21.5 (CH3CO2), 21.5 (C-15), 27.3 (C-2), 30.1 (C-8),
31.2 (C-12), 32.3 (C-1), 35.5 (C-7)*, 35.6 (C-16)*, 36.7
(C-10), 47.5 (C-13), 50.8 (C-14), 53.1 (C-9), 63.0 (C-3),

816
P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
75.4 (C-6), 129.4 (C-4), 143.5 (C-5), 169.9 (CH3CO2),
220.7 (C-17). (* interchangeable assignments).
3␤,19-diacetoxy-6␤-hydroxyandrost-4-en-17-one (33) [19]
(212 mg, 0.524 mmol, 12%), which crystallized from ethyl
acetate as amorphous crystals, m.p. 165–166°C, [␣]D 46°
(c ϭ 0.120);
2.5. Treibs oxidation of 3␤,19-diacetoxyandrost-5-en-17-
one (14)
IR ␯
3475, 1730, 1250 cmϪ1
;
max
Combustion analysis: Found: C, 68.66%; H, 8.22%.
(C23H32O6 requires C, 68.28%; H, 7.98%);
Mercury(II) oxide (2.44 g, 11.2 mmol) and trifluoroace-
tic anhydride (2.0 ml, 14.0 mmol) were stirred together in
dichloromethane (175 ml). 3␤,19-Diacetoxyandrost-5-en-
17-one (14) (1.75 g, 4.50 mmol) was added and the solution
was stirred for 49 h. The reaction mixture was filtered
through celite to remove the deposits of mercury(I) triflu-
oroacetate. The clear yellow filtrate was washed with satu-
rated sodium hydrogen carbonate solution and was dried.
Removal of the solvent on a rotary evaporator gave a dark
brown foamy gum (2.02 g), which was chromatographed.
Elution with 20% ethyl acetate in petrol gave unreacted
starting material (14) (249 mg, 0.641 mmol, 14%). Further
elution with 30% ethyl acetate in petrol gave 3␤-acetoxy-
19-hydroxyandrost-5-en-17-one (35) (27 mg, 0.078 mmol,
1.7%) which crystallized from ethyl acetate-petrol as plates,
m.p. 153–155°C, [␣]D 8° (c ϭ 0.10), lit [21]. m.p. 157–
158°C, [␣]D 7°;
MS (CI NH3) m/z (%) 422 (Mϩ, 60), 387 (100), 346 (41),
327 (14), 285 (11). (C23H32O6NH4 requires 422);
1H NMR (CDCl3) ␦ 0.90 (3H, s, H-18), 2.10 (6H, s,
CH3CO2-3␤,19), 4.32 (1H, s, H-6␣), 4.40 (1H, d, J ϭ 12
Hz, H-19), 4.60 (1H, d, J ϭ 12 Hz, H-19), 5.25 (1H, m, w/2
ϭ 27 Hz, H-3␣), 5.73 (1H, s, H-4);
13C NMR (CDCl3) ␦ 13.9 (C-18), 20.7 (C-11), 21.3
(CH3CO2-3␤), 21.3 (CH3CO2-19), 21.6 (C-15), 24.9 (C-2),
30.3 (C-8), 31.7 (C-12), 32.6 (C-1), 35.7 (C-16), 37.8 (C-7),
40.0 (C-10), 47.7 (C-13), 51.3 (C-14), 54.3 (C-9), 68.1
(C-19), 69.6 (C-3), 73.2 (C-6), 127.5 (C-4), 144.1 (C-5),
170.8 (CH3CO2-3␤)*, 170.9 (CH3CO2-19)*, 220.3 (C-17).
(* interchangeable assignments).
3. Results and discussion
IR ␯
3475, 1730, 1250 cmϪ1
;
max
1H NMR (CDCl3) ␦ 0.93 (3H, s, H-18), 2.03 (3H, s,
CH3CO2), 3.64 (1H, d, J ϭ 12 Hz, H-19), 3.90 (1H, d, J ϭ
3.1. Preparation of the substrates
w
12 Hz, H-19), 4.64 (1H, m, /2 ϭ 27 Hz, H-3␣), 5.80 (1H,
The preparations of the substrates were done according
to literature procedures [1–2,18,22–30].
d, J ϭ 5 Hz, H-6);
13C NMR (CDCl3) ␦ 13.9 (C-18), 20.9 (C-11), 21.3
(CH3CO2), 21.7 (C-15), 28.0 (C-2), 30.1 (C-12), 31.7 (C-7),
32.9 (C-8), 33.2 (C-1), 35.8 (C-16), 38.1 (C-4), 41.7 (C-10),
47.9 (C-13), 50.4 (C-9), 52.5 (C-14), 62.8 (C-19), 73.2
(C-3), 127.5 (C-6), 134.9 (C-5), 170.5 (CH3CO2), 220.9
(C-17).
3.2. Petrow reaction on substrates
Treatment of the 3␤,4␤-diacetate 13 with bromine pro-
duced two new compounds, 3␤,4␤-diacetoxy-5␣-bromo-
6␤-hydroxyandrostan-17-one (18) (23%) and 3␤,4␤-diace-
toxy-6␣-bromo-5␤-hydroxyandrostan-17-one (19) (19%),
which remained unchanged on addition of the silver salt.
The 4␤-acetoxyl moiety of the bromonium ion intermediate
successfully competed with the bromide nucleophile to af-
ford attack at C-5␤ and C-6␤. The resulting five- and six-
membered dioxolenium ions were solvolyzed to give the
respective products (Scheme 5).
The 3␤,19-diacetate 14 provides an interesting substrate
in that the proposed intermediary 6␤-bromo-⌬4-steroid
would have two potential functionalities to assist in anchi-
meric expulsion of the bromide ion. However, attack at both
the 5␤- and 6␤-positions only by the 19-acetyl group to
quench the bromonium ion occurred [31] (Scheme 6). The
5␤,19-acetoxonium ion (21) was hydrolyzed to give the
previously unreported 3␤,19-diacetoxy-6␣-bromo-5␤-hy-
droxyandrostan-17-one (23) (20%). The 5␤,6␤-epoxide 26
(28%), the other product, was possibly formed from the
6␤,19-acetoxonium ion (24). Some starting material (10%)
was also recovered.
Elution with 35% ethyl acetate in petrol gave 3␤,19-
diacetoxy-6-chloromercuriandrost-5-en-17-one (34) (71
mg, 0.114 mmol, 2.5%), which resisted crystallization,
IR ␯
1735, 1720, 1250 cmϪ1
;
max
MS (CI NH3) m/z (%) 646 (1), 645 (2), 644 (6), 643 (5),
642 (Mϩ, 14), 641 (8), 640 (10), 639 (7), 638 (4), 406 (100),
327 (2), 268 (10). (C23H3135Cl202HgO5NH4 requires 642);
MS (EI) m/z (%) 505.1149 (3), 504.1137 (9), 503.1132
(5), 267.1748 (100). (C23H3135Cl202HgO5–2 ϫ CH3CO2H
requires 504.1136);
1H NMR (CDCl3) ␦ 0.90 (3H, s, H-18), 2.03 (3H, s,
CH3CO2-19)*, 2.06 (3H, s, CH3CO2-3␤)*, 3.97 (1H, d, J ϭ
12 Hz, H-19), 4.62 (1H, d, J ϭ 12 Hz, H-19), 4.65 (1H, m,
w/2 ϭ 27 Hz, H-3␣). (* interchangeable assignments);
13C NMR (CDCl3) ␦ 13.6 (C-18), 20.9 (C-11), 21.1
(CH3CO2-3␤)*, 21.2 (CH3CO2-19)*, 21.7 (C-15), 27.7 (C-
2), 31.4 (C-12), 33.8 (C-1), 34.3 (C-8), 35.7 (C-16), 39.8
(C-7), 43.7 (C-10), 44.4 (C-4), 47.6 (C-13), 49.7 (C-9), 52.0
(C-14), 64.5 (C-19), 72.5 (C-3), 141.4 (C-5), 150.2 (C-6),
170.3 (CH3CO2-3␤), 170.3 (CH3CO2-19), 220.0 (C-17). (*
interchangeable assignments).
The 7-ketone (15) was inert to the reaction conditions
and was recovered unchanged. The 4,4-dimethylsteroid (17)
reacted readily under Petrow conditions giving a very com-
Further elution (40% ethyl acetate in petrol) produced

P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
817
Scheme 5. Reagents: (i) Br2/CHCl3. (ii) AgOAc/py Ϫ78°.
plex mixture of products, none in quantities sufficient for
characterization.
3.3. Modified Treibs oxidation on substrates
In view of the proposed mechanism for the Treibs oxi-
dation, the 3␤,4␤-diacetate (13) would provide an interest-
ing substrate, because there can be no 4␤-proton loss en-
Scheme 7. Reagents: (i) Hg(O2CCF3)2/CH2Cl2. (ii) aq NaHCO3.
route to the allylic trifluoroacetate. Loss of the 6␤-hydrogen
would produce the vinylic chloromercury steroid exclu-
sively. Under modified Treibs conditions, this substrate (13)
was converted to the two ⌬4-3␤,6␤-diol monoacetates 30
and 9, as well as the novel 6␤-acetoxy-3␣-hydroxyandrost-
4-en-17-one (32) (in 6%, 10%, and 6% yields, respectively).
A key intermediate in this reaction is proposed to be the
3␤,4␤-dioxolenium cation (4) formed by mercury mono(tri-
fluoroacetate) ion assisted removal of the 3␤-acetoxy group
(Scheme 7) [5]. The dioxolenium ion (4) could then be
quenched at C-6␤ by trifluoroacetate [20,32] with migration
of the double bond and collapse of the acetate bridge to give
the 6-trifluoroacetate (8). Work-up would afford the 6-alco-
hol 9. Alternately, attack on the dioxolenium ion (4) by the
more nucleophilic acetate ion (originally lost from C-3)
would produce the ⌬4-3␤,6␤-diacetate (27). Removal once
more of the less hindered 3-acetoxy group by mercury(II)
Scheme 6. Reagents: (i) Br2/CHCl3. (ii) AgOAc/py Ϫ78°.

818
P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
more electronegative than mercury, makes the carbon to
which it is attached susceptible to nucleophilic attack by the
neighboring acetoxy groups. It is apparent that the deacti-
vation and steric hindrance of the ⌬5-bond by the C-7
ketone and dimethyl substituents of 15 and 17, respectively,
are likely to have been responsible for the lack of reactivity
shown by these substrates.
Acknowledgments
P.L.D.R. thanks the University of the West Indies for the
granting of a Postgraduate Scholarship and a Tutorial As-
sistantship. The authors would also like to thank Professor
John C. Vederas (University of Alberta) for mass spectral
and analytical data as well as for helpful discussions.
This work was supported in part by funds secured under
a Canadian International Development Agency/Natural Sci-
ences and Engineering Research Council Research Fellow-
ship.
Scheme 8. Reagents: (i) Hg(O2CCF3)2, CH2Cl2. (ii) aq NaHCO3.
ion would give the allylic cation (28), which would be
attacked by trifluoroacetate ion either from the ␣ or ␤ face.
Finally, hydrolysis of the resulting epimeric 3-esters would
give the 3␣- and 3␤-alcohols 32 and 30.
Interest in the 3␤,19-diacetoxy-⌬5-steroid 14 as a sub-
strate stemmed from the possibility of neighboring group
participation at C-5 or C-6 by the 19-acetoxy group in the
decomposition of the initially formed 5␣,6␣-mercurinium
ion. However, the products of the reaction were those of the
normal Treibs reaction, i.e. the 6␤-alcohol 33 (12%) and the
novel chloromercury steroid 34 (2.5%). Also found was
3␤-acetoxy-19-hydroxyandrost-5-en-17-one 35 (1.7%) and
unreacted starting material (14%). Surprisingly, the 19-ester
functionality does not seem to directly participate in nucleo-
philic attack on the intermediary mercurinium ion; however,
it appears to shield the top face of the molecule thereby
reducing the yields of expected products (Scheme 8). It is
proposed that the 19-alcohol 35 was formed by slow mer-
cury mono(trifluoroacetate) assisted removal of the acetate
[5].
References
[1] Hanson JR, Reese PB, Wadsworth HJ. Allylic acetoxylation of ⌬5
steroids at C-4. J Chem Soc Perkin Trans I 1984;2941–4.
[2] Hanson JR, Reese PB. Neighbouring group participation in the allylic
oxidation of a ⌬5 steroid. J Chem Soc Perkin Trans I 1985;647–9.
[3] Broad RA, Hanson JR, Reese PB. The stereochemistry of the Treibs
allylic oxidation of ⌬5 steroids. J Chem Res (S) 1987;172–3.
[4] Ruddock PL, Reese PB. An unexpected product from the reaction of
mercury(II) trifluoroacetate with 3-substituted ⌬5 steroids. J Chem
Res (S) 1994;442–3.
[5] Ruddock PL, Williams DJ, Reese PB. The scope and limitations of
the reaction of ⌬5 steroids with mercury(II) trifluoroacetate. Steroids
1998;63;650–64.
[6] Petrow VA. Steroids and related compounds. Part I. Isomeric cho-
lestenediols. J Chem Soc 1937;1077–81.
[7] Petrow VA, Rosenheim O, Starling WW. Acyl migration in steroids.
J Chem Soc 1943;135–9.
[8] Glotter E, Kirsen I, Lavie D, Abraham A. The withanolides—a group
of steroids. Bioorg Chem 1978;2:57–97.
[9] Nittala SS, Lavie D. Withanolides of Acnistus breviflorus. Phyto-
chemistry 1981;20:2735–9.
[10] Brodie AMH, Garrett WM, Hendrickson JR, Tsai-Morris C-H, Mar-
cotte PA, Robinson CH. Inactivation of aromatase in vitro by 4-hy-
droxy-4-androstene-3,17-dione and 4- acetoxy-4-androstene-3,17-di-
one and sustained effects in vivo. Steroids 1981;38:693–702.
[11] Treibs W. Acetylating oxidation of ketones and olefins with mercuric
trifluoroacetate. Naturwissenschaften 1948;35:125.
[12] Massiot G, Husson H-P, Potier P. Allylic substitution of steroidal
olefins by a modified Treibs reaction. Synthesis 1974;722–3.
[13] Flanagan RJ, Charleson FP, Synnes EI, Weibe LI, Theriault YX,
Nakashima TT. Radiolabeling with organomercury compounds. Part
1. The synthesis and structure of 6-halocholest-5-en-3␤-ols. Can
J Chem 1985;63:2853–60.
It was thought that the presence of the 7-ketone in com-
pound 15 would increase the acidity of the 6␤-hydrogen in
the intermediary 5␣,6␣-mercurinium ion, thus facilitating
its loss over that at C-4␤ and directing the course of the
reaction toward the formation of the chloromercury steroid.
In a similar vein, because there is no 4␤-proton in 3␤-
acetoxy-4,4-dimethylandrost-5-en-17-one (17), it was ex-
pected that this would force the formation of the ⌬5-6-
chloromercury steroid. However, both substrates remained
unaffected by the reagent.
In summary, the substituents proximate to the ⌬5-bond in
the substrates had a pronounced influence on electrophilic
reactions, and this resulted in the isolation and character-
ization of several new steroids. The diacetates 13 and 14
reacted with bromine and silver acetate (Petrow reaction) to
give addition products exclusively. Mercury(II) trifluoroac-
etate, on the other hand, effected mainly allylic oxidation.
The products reflected the differences in the nature of the
C-Br and the C-Hg bonds in the bromonium and mer-
curinium intermediates, respectively. Bromine, being much
[14] Blunt JW, Stothers JB. 13C N.m.r. spectra of steroids—a survey and
commentary. Org Magn Reson 1977;9:439–64.
[15] Baldwin D, Hanson JR, Holtom A. Aromatization of some steroidal
enediols. J Chem Soc Perkin Trans I 1973;1703–7.
[16] Heusler K, Kalvoda J, Meystre Ch, Ueberwasser H, Wieland P, Anner
G, Wettstein A. Ein neues Verfahren zur Herstellung von 19-Nors-
teroiden. Experientia 1962;18;464–6.

P.L.D. Ruddock, P.B. Reese / Steroids 64 (1999) 812–819
819
[17] Billeter JR, Meischer K. Uber steroide. Abbauprodukte der sterinoxy-
dation. IV. Isolierung von ⌬3,5-androstadien-dion-(7,17). Helv Chim
Acta 1948;31:629–32.
[18] Adams WJ, Patel DK, Petrow V, Stuart–Webb IA, Sturgeon B.
4,4-Dimethylsteroids. II. Androstane and pregnane derivatives.
J Chem Soc 1956;4490–5.
[19] Mastalerz H, Morand P. 19-Hydroxy-steroids. Part 7. Boron trifluo-
ride-catalysed reactions of 19-hydroxy- and 19-acetoxy-5,6-epoxy-
steroids. J Chem Soc Perkin Trans I 1981;154–60.
11␤,19,21-trihydroxypregn-4-ene-3,20-dione. J Chem Soc Perkin
Trans I 1983;2945–51.
[25] Heusler K, Wettstein A. Uber Steroide. Herstellung von 7,9,11-
dienen der Androstan-Reihe. Helv Chim Acta 1952;35:284–94.
[26] Marshall CW, Ray RE, Laos I, Riegel B. 7-Keto Steroids. II. Steroi-
dal 3␤-hydroxy-⌬5-7-ones and -⌬3,5-7-ones. J Am Chem Soc 1957;
79:6308–13.
[27] Julian PL, Meyer EW, Ryden I. Sterols. X. 17␣-Hydroxyprogester-
one. J Am Chem Soc 1950;72:367–70.
[20] Hanson JR, Reese PB. A nuclear magnetic resonance study of the
conversion of 4␤-acetoxy-3␤-hydroxy-⌬5-steroids into 3␤,6␤-diace-
toxy-⌬4-steroids. J Chem Soc Perkin Trans I 1985;331–4.
[21] Tadanier J. Preparation and solvolysis of 6␤,19-oxido-17-ethyl-
enedioxy-3␣,5␣-cycloandrostane. J Org Chem 1963;28:1744–51.
[22] Rodewald WJ, Jaszczynski JR, Sicinski RR. 19-Functionalized ste-
roids. Part 1. An improved synthesis of androst-5-ene-3␤,17␤,19-triol
derivatives. Pol J Chem 1978;52:501–9.
[23] Heusler K, Kalvoda J. Selective functionalization of the angular
methyl group and further transformation to 19-norsteroids. In: Fried
J, Edwards JA, editors. Organic Reactions in Steroid Chemistry (Vol.
2). New York: Van Nostrand Reinhold, 1972. p. 251.
[28] Fieser LF, Fieser M. Reagents for Organic Synthesis. New York:
John Wiley & Sons, 1967. p. 35.
[29] Djerassi C. Oppenauer oxidation. Org React 1951;6:207–72.
[30] Ringold HJ, Rosenkranz G. Steroids. LXXXV. Synthesis of 4-methyl
and 4,4-dimethyl hormone analogs. J Org Chem 1957;22:602–5.
[31] Kocovsky P, Stary I, Zajicek J, Turecek F, Vasickova. Stereo- and
regio-control of electrophilic additions to cyclohexene systems by
neighbouring groups: participation of allylic and homoallylic ester
groups in hypobromous acid addition to some 5-unsaturated choles-
tane derivatives. J Chem Soc Perkin Trans I 1988;2297–303.
[32] Hanson JR, Reese PB. Neighbouring group participation in the rear-
rangement of 4␤-acetoxy-⌬5-steroids to 6␤-acetoxy-⌬4-steroids. Tet-
rahedron Lett 1983;24:303–6.
[24] Kirk DN, Yeoh BL. New syntheses of 19,21-dihydroxypregn-
4-ene-3,20-dione, 21-hydroxy-19-norpregn-4-ene-3,20-dione and