1977 2163An Improved Synthesis of 13-epi-Androstanes and of 13-epi-OestranesBy Robin 6. Boar," Francis K. Jetuah, James F. McGhie, and Michael S. Robinson, Department of Chem-istry, Chelsea College, London SW3 6LXDerek H. R. Barton, Department of Chemistry, Imperial College, London SW7 2AYTreatment of a 17-hydroxyimino-steroid with refluxing acetic anhydride and pyridine affords an equilibrium mixtureof 13-epi-enimide and 13-epi-enamide. A free-radical mechanism is proposed. Acidic hydrolysis gives the 17-oxo-1 3-epi-steroid in high overall yield.THE generally greater stability of steroids with CD cis-as opposed to trans-stereochemistry permits the syn-thesis of 13-e$i-derivatives by reactions which involvecleavage and subsequent re-formation of the 13,17bond.Thus, U.V. irradiation of 17-oxo-steroidsgenerates, via a 13,17-diradical, an equilibrium mixtureof the normal and 13-e$i-isomer~.~-~ Separation of thismixture then affords the 13-e$i-compound in acceptable,if not high, yield. Alternatively, treatment of 20-0x0-pregnanes with antimony pentafluoride-hydrogenfluoride affords, via a 17-carbocation, a major amount ofthe 13a,17a-i~omer.~ We now report details of ourrecently described enamide synthesis 7 as applied to thepreparation in high yield of pure 17-e$i-androstanes and-oes t ranes .Reduction of 3~-acetoxy-17-hydroxyiminoandrost-5-ene8 (1) in acetic anhydride with chromium(I1) acetategave the expected enamide (2), m.p. 229-232",[a], -15". The structure of this product is well estab-l i ~ h e d .~ When the oxime (1) was treated with refluxingacetic anhydride and pyridine, two products resulted.These were easily separated by chromatography onsilica gel. Comparison of the n.m.r. spectrum and themolecular rotation of the more polar product, m.p.160-163', [a], -102", with those of the enamide (2)strongly suggested that the former was the 13-e$i-enamide (3; R = H). The less polar product, m.p.118-121", [a], -99", was evidently the 13-e$i-enimide(3; R = Ac), since it was converted into the more polarproduct by chromatography on an alumina column.7The enamide (3; R = H) and the enimide (3; R = Ac)are apparently in equilibrium in refluxing aceticanhydride and pyridine.To confirm the formulation of the above products aslS-e$i-androstanes, the enamide (3; R = H) was sub-jected to acidic hydrolysis. The ketone thus obtainedwas identical with authentic 3P-hydroxy-13a-androst-5-en-17-one (4).3By the above method an overall yield of ca.65% isreadily obtained for the conversion of 17-oxo-steroidinto 17-oxo-13-e$i-steroid.M. Hanack, ' Conformation Theory,' Academic Press, NewA. Butenandt and L. Poschmann, Ber., 1944, 77, 394, andJ . P. L. Bots, Rec. Trav. chim., 1958, 77, 1010.H. Wehrli and K. Schaffner, Helv. Chim. Acta, 1962, 45, 385.J.-C. Jacquesy, R. Jacquesy, and J.-F. Patoiseau, Bull. SOC.R. B. Boar, F. K. Jetuah, J . F. McGhie, M. S. Robinson, andYork, 1965, pp. 176 el seq.references therein.chim. France, 1974, 1959.D. H. R . Rarton, J .C . S . ChPm.. (7om.m.. 1975, 748.In an analogous series of experiments, 17-hydroxy-imino-3-methoxyoestra-1,3,5( 10)-triene lo (5) was con-verted with similar efficiency into the 13-e$-enimide (6 ;A0Me0 & '( 7 )R = Ac) and the 13-epi-enamide (6; R = H). Hydro-lysis of the enamide gave 3-methoxy-13a-oestra-1,3,5( lO)-trien-l7-one (lumioestrone methyl ether) (7),m.p. 130-133", fa], -27.5' (1it.,l1 m.p. 129-130',In addition to providing an efficient, non-photo-chemical route to 13-e$i-steroids, the reaction of 17-oximes with refluxing acetic anhydride and pyridineR. B. Boar, J. F. McGhie, M. Robinson, I>. H. R. Barton,D. C. Horwell, and R. V. Stick, J.C.S. Perkin I , 1975, 1237.R. Anliker, M. Miiller, J . Wohlfahrt, and H.Heusser, Helv.Chim. Acta, 1955, 38, 1404.G. Rosenkranz, 0. Mancera, F. Sondheimer, and C. Djerassi,J . Org. Chem., 1956, 21, 520.lo B. M. Regan and F. N. Hayes, J . Amer. Chem. SOL, 1956,78,639.A. Butenandt, A. Wolff. and P. Karlson, Ber., 1941,74, 1308.[a], -28')2164also provides compelling evidence that, as proposed inour original publication,' enamide formation involves afree-radical pathway (Scheme). The alternative ofAc+IAcINOH .N - OAc N- OAc. H HJA cIN *H AcIHAcI1HIt wc2 HSCHEMEring opening and re-closure via a carbocation species isnot feasible under these (basic) reaction conditions.EXPERIMENTALN.m.r. data (90 MHz) are for solutions in deuterio-chloroform with tetramethylsilane as internal reference.Acetic anhydride and pyridine were of AnalaR or equivalentgrade.T.1.c. plates were prepared using Merck silica gelGF,,,. Laporte type 0 alumina and Merck 60 silica gelwere employed.17-Acetamidoandrosta-5,16-dien-3P-y2 Acetate (2) .-17-Hydroxyiminoandrost-5-en-3P-yl acetate 8 (1 g) in dryNN-dimethylformamide (20 ml) was stirred under nitrogen,and acetic anhydride (25 ml) was added. After 1 h freshlyprepared anhydrous chromium(I1) acetate l2 (8 g) wasadded. The mixture was stirred overnight a t 40 "C. Thesolvent was removed under reduced pressure, N-sodiumcarbonate solution (100 ml) was added, and the mixture wasextracted with ethyl acetate. The combined extracts werewashed with water, dried, and evaporated. The residuewas chromatographed on a silica gel column to afford the13P-enamide (2) (65%), m.p.229-232", [a], -15" (c 1.2 inethanol) (lit.,9 m.p. 237-240", [a], - 18")' T 3.5br (1 H,exchanged with deuterium oxide, -NH-), 4.0br (1 H,16-H). 4.60br ( 1 H, d, 6-H). 5.4 ( 1 H. m, 3a-H), 7.93 andJ.C.S. Perkin I7.99 (each 3 H, s, Ac), and 8.95 and 9.11 (each 3 H, s, 18-and 19-H,).Action of Acetic Anhydride-Pyridine on 17-Hydroxyimino-com~ounds.-l7-Hydroxyiminoandrost-5-en-3~-yl acetate 8(1 g) in pyridine (50 ml) and acetic anhydride (30 ml) wasstirred under nitrogen and heated on an oil-bath so thatgentle refluxing occurred. The mixture rapidly developeda dark colour. Overheating afforded excessive amounts ofa black deposit and lower yields of the required products.When no oxime or oxime acetate remained (t.1.c.control;ca. 12 h), the mixture was evaporated to dryness. Theblack, oily residue was treated with ether (80 ml), andN-sodium carbonate solution (80 ml) was added. Themixture was triturated thoroughly for several minutes,and then filtered through a thin pad of Celite. Thegranular black residue was washed thoroughly with severalportions of ether. The combined ethereal extracts werewashed with water, dried, and evaporated. T.1.c. of theresultant, pale brown oil indicated the presence of twoproducts, one slightly less polar and one more polar than theoriginal oxime.Chromatography on a dry silica gel column, eluted with30% ethyl acetate in light petroleum gave 17-diacetylamino-13a-androsta-5,16-dien-3P-yZ acetate (3; R = Ac), m.p.(from hexane) 118-121", [a], -99" ( G 0.7 in chloroform),vmx.(Nujol) 1730, 1705, 1240, and 1020 cm-l, T 4.35br(1 H, t, 16-H), 4.60br (1 H, d, 6-H), 5.4 (1 H, m, 3a-H), 7.62(6 H, s, NAc,), 7.98 (3 H, s, OAc), and 8.90 and 9.04 (each3 H, s) (Found: C, 72.65; H, 8.7; N, 3.4. C,5H,5N0,requires C, 72.6; H, 8.6; N, 3.4%). Further elution with50% ethyl acetate in light petroleum afforded 17-acetamido-13a-androsta-5,16-dien-3P-yl acetate (3; R = H), m.p. (fromacetonitrile) 160-163", [a], - 102" (c 1.8 in chloroform),vmax. (Nujol) 3 370, 1725, 1690, 1540, 1370, 1240, and1 030 cm-1, T 3.15br (1 H, s, exchanged with deuteriumoxide, -NH-), 4.00br (1 H, s, 16-H), 4.6br (1 H, d, 16-H),5.4 (1 H, m, 3a-H), 7.91 and 7.99 (each 3 H, s, Ac), and 9.01and 9.12 (each 3 H, s) (Found: C, 74.3; H, 8.9; N, 3.75.C,,H,,NO, requires C, 74.4; H, 8.95; N, 3.8%).Alternatively, chromatography of the total product on analumina column yielded the enamide (3; R = H) (85%) asthe sole product.Identical treatment of 17-hydroxyimino-3-methoxy-oestra-1,3,5( 10)-triene 10 (5) with refluxing acetic anhydrideand pyridine afforded 17-diacetylamino-3-methoxy- 13a-oestra-1,3,5(10),16-tetraene (6; R = Ac), m.p.(from methanol)114-116", [a], +32" (c 0.5 in chloroform), T 2.86 (1 H, d,J 9 Hz, 1-H), 3.2-3.45 (2 H, m, 2- and 4-H), 4.31 (1 H, m,16-H), 6.26 (3 H, s, OMe), 7.63 (6 H, s, NAc,), and 8.85(3H, s, 18-H,) (Found: C, 75.3; H, 8.0; N, 3.7. C~&&QNO,requires C, 75.2; H, 7.95; N, 3.8%) and 17-acetamido-3-methoxy-13a-oestra-1,3,5( lO),lB-tetraene (6; R = H), agum, [a], + 11" (c 1.2 in chloroform), z 2.86 (1 H, d, J 9 Hz,1-H), 3.12br (1 H, s, exchanged with deuterium oxide,-NH-), 3.2-3.45 (2 H, m, 2- and 4-H), 4.0br (1 H, s,16-H), 6.27 (3 H, s, OMe), 7.93 (3 H, s, Ac), and 8.93 (3 H,s, 18-H,).Alternatively, chromatography of this mixtureon a column of alumina gave solely the enamide (6 ; R = H)R = H) and ( 6 ;R = H).-The enamide (3; R = H) (200 mg) in methanol(20 ml) containing B~-hydrochloric acid (6 ml) was heatedunder reflux for 1 h. The mixture was evaporated to half(80%).Hydrolysis of the 13a-Enamides (3;l2 J. R. Hanson. Synthesis, 1974, 1 1977 2165;volume, poured into water, and extracted with ether. 13a-oestra-1,3,5( lO)-trien-17-one (7), m.p. 130-133", [a],Crystallisation from hexane gave 3P-hydroxy-13a-androst- - 27.5" (G 0.5 in chloroform) (lit. ,11 m.p. 129-130", [a],5-en-17-one (a), m.p. 180-184", [a], -162" (c 0.6 in -28").ethanol) (lit.,3 m.p. 187-189", [a], - 162.9'). The acetatehad m.p. 135-139", [a], -152" (C 0.5 in ethanol) (lit.,3m.p. 143-144", [m], - 149').We thank the Schering-Plough Corporation for gifts ofSimilarly, the enamide (6; R = H) afforded 3-methoxy- [7/602 Received, 5th April, 19771chemicals