J.C.S. Perkin I Steric Course of Reduction with Sodium Borohvdride of Steroidal a@-Epoxy-ketones By Martin Weissenberg,' Department of Organic Chemistry, The Weizmann Institute of Science, Rehovot, Israel Erwin Glotter,' Faculty of Agriculture, The Hebrew University of Jerusalem, Rehovot 76-1 00, P.O.Box 12,. Israel The formation of cis-and trans-epoxy-alcohols by treatment of steroidal epoxy-ketones with sodium borohydride is analysed in terms of the direction of attack on the corresponding unsubstituted ketones, and the steric hindrance exerted by the epoxide ring. ALTHOUGHup-epoxy-alcohols are easily obtained lq~ see Table, led to the view 293 that the preferential reduction of the Corresponding epoxy-ketones with formation of trans-epoxy-alcohols is a general feature sodium borohydride, there are relatively few reports on of this reduction.The stereochemistry of these, as stereocheinical aspects of this reaction. Earlier results well as of other similar reductions of related unhindered obtained by reduction of steroidal ap-epoxy-ketones, steroidal epoxy-ketones compounds (lc), (2a, c, and d), mainly 4,5-epoxy-3-ones (Zc and d) and (4b and c); 192 1. hl. Coxon, M. P. Hartshorn, and D. N. Kirk, I. Ch~m. R. Camerino and D. Cattapan. Farmnaco, Ed. Sci., 1958, 13, S0c.,~1964, 2461. 39. E. Toromanoff, Compt. rend.. 1967, 264, 1881. and (4a-c) has been evaluated by T~romanoff,~ who suggested that the direction of the kinetic addition of the hydride ion is determined by electronic factors and should take place preferentially from the side of the oxygen atom (or of the methylene group, in the case of ap-methylene ketones) of the three-membered ring.The predominant products would therefore be trans-epoxy- alcohols (or trans-methylene-alcohols) . More recently, Chautemps and Pierre5 proposed a different interpretation of the results obtained by reduction of ap-epoxy-ketones with sodium borohydride, based on steric and electronic factors. Whenever one 5a-3-ones and of 5a-4-ones gives mainly 3p- (87-94y0) and 4p-alcohols i.e. the major attack of the reduc- ing agent takes place from the rear side of the molecule, irrespective of the presence or absence of the epoxide ring. However, 3p- and 4p-alcohols are obtained from the unsubstituted ketones in higher yields than are trans-epoxy-alcohols from the corresponding epoxy-ketones (2) and (3).Similarly, reduction of 53-3-ones and of 5P-4-ones yields mainly 3a- and 4cc-alcohols,1deg; the steric course of reduction being the same as for the corresponding epoxy-ketones (4) and (5). The tendency of a one in the 5a-series to give by Reduction of steroidal ap-epoxy-ketones with sodium borohydride Substrate Choles tane Androstan-l7p-01Androstan-17-one Cholestane 17p-AcetoxyandrostanePregnan-20-one4-MethylcholestaneCholes tane AB-CiS (44 Choles tane (4b) Pregnan-20-one4-Me thylcholes tane (5) Choles tane Epoxide Ketone Epoxy-alcohol trans (yo) cis (Yo) Ref. 1a, 2a- la,2a-la,2a-4a.5~- 3-3-3-3- 60 Major 60 75 40 13 11 12 a 4a,5a-4a,5u-4a,5a-5a, 6u- 3-3-3-4- 63 76 58 72 12 b 1 2 c 4p,5p-4p,5p-4p, 5p- 5p, 6P- 3-3-3-4- 85 75 70 90 a 1 2 d a I).J.Cpllins, J. Chem. SOC., 1959, 3919. D. Baldwin and J. R. Hanson, J.C.S. Perkin I, 1972, 1889. D. Lavie, Y. Kash- man, and I*..Glotter, Tetrahedron, 1966, 22,1102. S. Greenfield, E. Glotter, D. Lavie, and Y. Kashman, J. Chem. SOC.(C),1967, 1460. of the lone pairs of the epoxidic oxygen and the carbonyl can assume a amp;-relationship, a chelate with the cation of the complex hydride might be formed (' cyclic model ') ; consequently the reducing species would attack the c-arbonyl from the less hindered face, leading largely to a trans-epoxy-alcohol. If the formation of such a cyclic.model is sterically difficult, mixtures of cis-and trans-epoxy-alcohols would be formed ; their formation could also be rationalised by use of Felkin's A re-examination of the results in the Table leads to the conclusion tliat only tri-and tetra-substituted 4,5-epoxy- and 5,6-epoxy-ketones give predominantly trans- epoxy-alcoliols, whereas disubstituted 1a,2a-epoxy-ketones show only a slight preference for the format ion of such derivatives. In 4a, 5a-epoxy-3-ones (2) and in 5r,6a-epoxy-4-ones (3), rings A and B are trans-fused and the spatial arrangement of the molecule is almost the same as in the unsubstituted 3-and 4-ones (5a-series). In fact, reduction with borohydride of IS. Toromanoff, Bull. SOC.chim. France, 1968, 2457.P. Chautcmps and J.-L. Pierre, Tetrahedron, 1976, 32,549. M. Chcrest and H. Felkin, Tetrahedron Letters, 1968, 2205; 1971, 383. 7 For a pertinent discussion on the factors controlling axial or equatorial attack on cyclohexanones, see J. Huet, Y. Maroni-Barnaud, Nxuyen Trong Anh, and J. Seyden-Penne, Tetrahedron Letters, 1976, 159. * L). M. S. Wheeler and M. M. Wheeler, in ' Organic Reactions in Steroid Chemistry,' eds. J. Fried and J. A. Edwards, Van Nostrand, New York, 1972, vol. 1, pp. 61, 78, and references cited tlirrci ti. hydride reduction the equatorial 3p-01 is diminislied in the reduction of 1a,2a-epoxy-3-ones. In the experi- ments with compounds (lb) l1 and (1c),12 only the trans- epoxy-alcohols are mentioned ; in the experiment with compound (la)l3 a substantial amount of the cis-epoxy-alcohol has been isolated.The behaviour of these compounds suggests that electronic factors alone cannot satisfactorily accommo-date the results given above and that steric factors should be considered as well. In our view, the outcome of the hydride reduction of ap-epoxy-ketones in a rigid framework may be interpreted as follows: (i) the main direction of attack of the reducing agent is the same as in the corresponding unsubstituted ketone ; and (ii) the epoxide ring may sterically interfere with this attack, leading to various amounts of the stereoisomeric epoxy- alcohol. A case in point is that of 1P,ZP-epoxy-5a-cholestan-one (6).14 By analogy with the behaviour of several ap-methylene-ketones, Toromanoff predicted that D.C. Ayres, D. N. Kirk, and R. Sawdaye, J. Chem. SOC.(B),1970, 505. lo M. Weissenberg and E. Glotter, J.C.S. Perkin I, 1977, 988. l1 W. R. Benn, F. Colton, and R. Pappo, J. Amer. Chem. SOC., 1957, 79,3920. 12 B. Pelc, J. Hodkova, and J. Holubek, CoZZ. Czech. Chem. Comun., 1966, 31,1363. l3 M. Weissenberg, D. Lavie, and E. Glotter, Tetrahedron, 1973, 29, 353. 14 (a) R. Albrecht and Ch. Tamm, Helv. Chim. Acta, 1957, 40, 2216; (b) H. B. Henbest and I.A. L. Wilson, J. Chem. SOC., 1957. 1958. reduction with borohydride of a 5a-steroidal 1p,Zp-methylene-3-one and of the analogous epoxy-ketone should yield predominantly trans-methylene- and epoxy- alcohols, respectively.In fact, only the cis-epoxy-alcohol (7) is obtained by reduction of the epoxy-ketone (6). In our opinion, the factors outlined above concur to secure the stereospecificity of this reaction: the ten- dency of the one to give the equatorial 3p-01 by rear- side attack is supported by the hindrance of the p-face of the molecule (epoxide ring, in addition to the angular methyl group), precluding ' top '-side attack. The isomeric 2a ,3a-epoxy-5a-cholestan- 1-one (8) was prepared by mild oxidation of the epoxy-alcohol (9).15 Reduction of compound (8) is stereospecific leading to the cis-epoxy-alcohol (9). The result agrees with our views, since reduction of 5a-cholestan-l-one with sodium borohydride yields only 1a-hydroxy-5a-cholestane.10 In compound (8), the tendency, if any, for rear-side attack of the reducing agent is counterbalanced by the hindrance due to the epoxide ring.H H 02x3' H H 0 H (16 1 Electronic and steric factors should have been anti- cipated to concur in the reduction of 2p,3p-epoxy-5a- cholestan-l-one (10),16 leading mainly to the trans-epoxy-alcohol (11). In fact a 2 : 1 mixture of trans- (1 1) and cis-(12) epoxy-alcohols is obtained.16 The outcome of this reduction is consistent with the hindrance exerted l5 E. Glotter, P. Krinsky, M. Rejto, and M. Weissenberg, J.C.S. Perkin I, 1976,1442. l6 E. Glotter and P. Krinsky, J.C.S. Perkin I, 1978,413. J.C.S. Perkin I by the p-oriented epoxide ring, diminishing the pro- pensity for approach of the reducing agent from the ' top ' of the molecule. Reduction of 1P,Z~-epoxy-5p-cholestan-3-one(13)l7 afforded a mixture of trans- (14) and cis-(15) epoxy-alcohols, roughly in the same ratio as in the reduction of the stereoisomeric epoxy-ketone (la).Although the orientation of the hydroxy-group in the major product (14) is the same as in the alcohol obtained by reduction of Sp-cholestan-3-one, the propensity to a-attack by the reducing agent leading to more than 40 of the cis-epoxy-alcohol (15) is remarkable for a ketone in the 5p-series, in which the folding of rings A and B admittedly precludes the rear-side approach of the reagent at C-3. In 2~,3~-epoxy-5~-cholestan-l-one(16),l* front-side approach of the reducing agent is difficult owing to hindrance by the epoxide ring, and rear-side approach at C-1 is difficult owing to the folding of rings A and B.In fact compound (16) remained mostly unchanged on attempted reduction with sodium borohydride. EXPERIMENTAL M.p.s were taken with a Fisher-Johns apparatus. Optical rotations were recorded with an automatic Perkin- Elmer polarimeter and refer to solutions in chloroform. N.m.r. spectra were determined with a Varian A-60 instru- ment for solutions in deuteriochloroform. T.1.c. was carried out on chromatoplates of silica gel G (Merck) and spots were developed with iodine vapour. Column chro- matography was performed on silica gel 60 (Merck; 70-230mesh). Analyses were performed in the niicroanalytical laboratory of the Weizinann Institute, under the direction of Mr. R.Heller. 2a,3a-E~oxy-5a-choZestan-l-one (8).-To a solution of 2a,3a-epoxy-la-hydroxy-5a-cholestane(9)l5 (300 mg) in acetone (50ml), Jones reagent was added dropwise at 5-10 "C. The excess of oxidising agent was then destroyed with a few drops of methanol and most of the solvent was removed under reduced pressure. The product was extracted with ether; the extract was washed with water, dried, and evaporated. The crude product ('280mg) was homogeneous on t.1.c.; m.p. 95-97 "C (from methanol), a, t26.5"(c 1.1); 6 3.42(3P-H,ni, I.Vj 8Hz), 3.11 (2P-H, d, 3.5 Hz), and 0.99 (19-H, s) (Found: C, 80.7; H, 11.1. C2amp;amp;2 requires C, 80.95;H, 11.05). General Procedure for Reduction of Steroidal ap-Epoxy-ketones with Sodium Borohydride.-To a solution of the epoxy-ketone (100mg) in methanol (15-25 ml), sodium borohydride (100mg) was added over a few minutes.The solution was stirred for 2 h at room temperature, then neutralised with dilute hydrochloric acid; most of the sol- vent was removed under reduced pressure, water was added, and the crude product was filtered off or extracted with ether, washed with water, and dried. Further purification was carried out as required, by direct crystallisation or by chromatography. 1~,2~-Epoxy-5a-cholestan-3-one(6) l4 afforded (after chromatography; elution with hexane-ether, 4 : 6), 1@,2@-epoxy-3~-hydroxy-5a-cholestane(7)l4 (95 mg), m.p. and mixed m.p.174-1763 "C (from methanol) (lit.,14a 175-l7 E. Glotter, M. Weissenberg, and D. Lavie, Tetrahedron, 1970,26,3857. M. Weissenberg, unpublished results. 1978 176 "C); 6 3.98 (3a-HI m, Wt 18 Hz), 3.25 (la- and 2a-HI narrow signal), and 0.90 (19-H, s). 2a, 3a-Epoxy-Sa-cholestan- 1-one (8) gave quantitatively 2~,3a-epoxy-la-hydroxy-5a-cholestane(9),15 m.p. and mixed m.p. 93-95 "C (from methanol) (lit.,15 94-95 "C). 1~,2~-Ep0xy-5~-cholestan-3-one( 13) l7 gave a crude product which was separated by chromatography into two components. Elution with hexane-ether (7.5 : 2.5) afforded lp,2P-epoxy-5B-choZesfan-3p-~Z(15) (45 nig), m.p. 140-142 "C (from methanol), aI1, +6l" (c. 0.6); 6 4.11 (3a-H, m, U71 10 Hz), 3.3 (la-and Za-H, partial overlap), and 1.18 (19-H, s) (Found: C, 80.6; H, 11.4. C27H4602 requires C, 80.55; H, 11.5). Further elution with the same solvent gave a mixture of (14) and (15) (5mg), followed by l~,2~-e~oxy-5~-choZestan-3a-oZ(14) (50 mg), m.p. 141- 143 "C (from methanol), aID +37" (c 0.2); 6 4.02 (3P-H, m), 3.10 (la- and 2a-H, narrow signal), and 1.14 (19-H, s) (Found: C, 80.4; HI 11.6. C2,H4,02 requires C, 80.55; H, 11.5y"). 2P,3P-Epoxy-5P-cholestan-I-one ( 16),18 1ii.p. 98-99 "C (from ethanol-chloroform), a,, +6.5" (c O.G), remained practically unchanged under similar conditions (t.1.c. and n.ni.r. evidence). 7/1555 Received, 31st August, 39771
展开▼
