NOTES
Oct. 20, 1954
5171
solved in 40 cc. of ethyl acetate was hydrogenated over 0.4 of the catalytic hydrogenation product of Ia (m.p. 166168'; Marker, et al.,3* give m.p. 120') and the infrared g. of a 5% palladium-charcoal catalyst at room temperaspectra were identical. The diacetate differed from chloro- ture and atmospheric pressure. Crystallization of the genin diacetate (m.p. 155-156') as evidenced by a depres- product from ether-pentane afforded 1.81 g. (90%) of the sion in m.p. on admixture and differences in the infrared saturated triacetate V with m.p. 137-139', [a]D +23", n o appreciable absorption in the ultraviolet. Identity wtth spectra. the sample prepared by method (a) was established through A~6-Allopregnene-3p,6p-diol-20-one Diacetate (III).A mixture of 10 g. of the diacetate I I b and 50 cc. of acetic mixture m.p. determination and infrared comparison. anhydride was heated in an autoclave a t 1857190' for 8 S.A. hours, poured into water, extracted with ether, washed well SYNTEX 2679 with sodium carbonate solution, dried and evaporated. APARTADO The resulting oily "furosten" was oxidized with chromium MEXICO,D. F. trioxide and the resulting "diosone" subjected to saponification as described previously.'* Chromatographic purification on alumina and crystallization from ether-pentane Steroids. L1X.l Ring D Rearrangement of 17a,afforded 5.11 g. (6370) of the fllopregnene derivative 111 2 1-Dihydroxy-20-ketosteroids with m.p. 164-165', [a]D -18 , Amsx 238 mp, log e 4.03, umsI 1718 and 1660 an.-' (reported8 m.p. 233-235'). BY E. BATRES,G. ROSENKRANZ A N D FRANZ SONDHEIMER Anal. Calcd. for ClbHsaOr: C, 72.08; H, 8.71. Found: RECEIVED APRIL28, 1954 C, 72.38; H, 8.58. Allopregnane3j3,6P-diol-20-oneDiacetate (IV) .-The AI6The so-called "D-homo" rearrangement of 17acompound I11 (1.20 g.) dissolved in 100 cc. of ethyl acetate hydroxy-20-ketosteroids of the pregnane series was hydrogenated over 0.4 g. of 5% palladium-charcoal (type Ia) to the corresponding 17aa-hydroxya t atmospheric pressure and room temperature. After 2 17a~-methyl-l7-keto-~-homoandrostane derivatives hours, the catalyst and solvent were removed and the residue was crystallized from ether-pentane. The resulting (type IIa) has been effected stereospecifically allopregnane-3j3,613-diol-20-onediacetate (0.96 9.) showed through heat treatment and through reaction with m.p. 175-177', [a]D+18", no appreciable absorption in the aluminum alkoxides and with boron trifluorideultraviolet, urnax 1718 and 1700 cm.-'. Anal. Calcd. for Cg~H3806: C, 71.74; H, 9.15. Found: acetic acid-acetic anhydride (to yield the 17aaacetates) .23 Moreover this isomerization may be C, 71.45; H , 8.89. A1e~20-Allopregnadiene-3~,6j3,20-triol Triacetate (Enol brought about by the action of potassium hyAcetate of III).-A solution of 8.0 g. of the AIB-derivative droxide, which however yields the l7a-isomer of I I a I11 and 1.5 g. of p-toluenesulfonic acid in 180 cc. of isoproThe D-homo rearrangement penyl acetate was slowly distilled during the course of 10 as the major p r o d ~ c t . ~ hours (120 cc. of distillate collected). Addition of water, has not been carried out previously with compounds followed by extraction with ether, washing with sodium car- of type Ib, possessing the 17a-hydroxy-21-acetoxybonate, drying and evaporation left a residue which on crys- 20-keto side chain characteristic of cortisone acetallization from ether-pentane yielded 7.1 g. (81%) of the tate. We were interested in performing the isoenol acetate with m.p. 169-172'. The analytical sample merization with compounds of this series so as to showed m.p. 179-181°, [a]D +36', Am,, 238 mp, log e 4.18, make available reference substances which could be urnax 1736 and 1718 cm.-'. Anal. Calcd. for C W H S ~ OC, ~ :70.71; H, 8.35. Found: compared with microbiological transformation C, 70.98; H, 8.53. prod~cts.~ A1B-Allopregnene-3p,6P,2I-triol-20-one Triacetate (VIb) 17a-Hydroxydesoxycorticosterone (Reichstein's The above enol acetate (6.9 g.) mas heated with 6.9 g. of Nsubstance S) 21-acetate (Ib) was subjected to iodosuccinimide in 40 cc. of dioxane a t 80" for 2 hours. Addition of water, extraction with ether (ether layer washed Oppenauer oxidation conditions (boiling with with sodium thiosulfate and water) and crystallization from aluminum isopropoxide and cyclohexanone in methanol afforded 6.7 g. of the iodoketone VIa with m o p .ca. toluene). The product, isolated in 46% yield, was 200' (dec., varies with rate of heating), A, 250 mp. log assigned the 17a/3-acetoxymethyl-l7aa-hydroxyB 3.94, umSx 1718 and 1660 ern.-'. 17-keto-~-homoandrostane structure IIb, since Anal. Calcd. for C Z S H ~ ~ OC, S I 55.33; : H , 6.51. Found: analysis proved i t to be isomeric with the starting C, 55.01; H , 6.60. The iodoketone was refluxed with 20 g. of potassium ace- material, and since similar conditions (aluminum tate in 200 cc. of acetone for 6 hours, and was then poured t-butoxide in benzene with or without acetone) into water. Ether extraction and crystallization from in the 21-desoxy series had led to the corresponding ether-pentane furnished 4.8 g. (54% over-all based on 111) 17a~-methyl-l7aa-hydroxy-17-keto compounds of the unsaturated triacetate VIb with m.p. 137-138", Amax 240 mp, log B 4.20. (type IIa).2 I n agreement with this formulation, Anal. Calcd. for C Z ~ H ~ ~C, OT 68.33; : H, 8.07. Found: the rearranged acetate I I b gave a negative reaction C, 68.48; H, 8.24. with triphenyltetrazolium chloride,s although the Allopregnane-3p,6~,2l-triol-2O-oneTriacetate (V). ( a ) saponification product IIc reacted weakly posiFrom Allopregnane-3j3,6p-diol-20-one Diacetate (IV).-A solution of 0.80 g. of the saturated diacetate IV and 1.5 g. tively. of lead tetraacetate (Arapahoe Chemicals, Boulder, Colo.; (1) Steroids. LVIII, J. Romo, G. Rosenkranz and F. Sondheimer, ca. 90% pure) in 20 cc. of glacial acetic acid was heated on THIS JOURNAL, 76, 5169 (1954). the steam-bath for 6 hours and then left a t room tempera(2) Infer al., P. Hegner and T . Reichstein, Hclv. Chim. A d a , 24, 828 ture overnight. The solution was poured into water, the (1941); R . B. Turner, THIS JOURNAL, 76, 3484 (1953). product was isolated with ether and chromatographed on (3) J. v. Euw and T . Reichstein, Hclv. Chim. Acta, 24, 879 (1941). 30 g. of neutral alumina. Crystallization of the fractions (4) Cf.J. Fried, R . W. Thoma, J. R. Gerke, J. E. Herz, M. N . Donin eluted with hexane-benzene from ether-pentane yielded and D. Perlman, THISJOURNAL, 74,3962 (1952). Added in proof: V. 0.44 g. (48%) of the triacetate V with m.p. 13&140', [ a ] ~Georgian and N. Kundu [Ch~mislrrand Industry, 431 (1954)l have +25O, umax 1736 and 1718 cm.-'. now described the D-homo rearrangement of Substance S Acetate and Anal. Calcd. for C27H400,: C, 68:04; H, 8.46. Found: of cortisone acetate with boron trifluoride-acetic acid-acetic anhydride. C, 68.10; H, 8.12. (5) This test in the steroid series has been found so far t o be specific (b) From A16-Allopregnene-3,c3,6j3,21-triol-20-one Triacetate (VIb).-The unsaturated triacetate VIb (2.0 g.) dis- for compounds containing the 21-hydroxy-20-keto function either in
.-
(12) Cf.c. Djerassi, J . Romo and G . Rosenkranz, J . Org. Chcm.. 16, 754 (1951).
the free or esterified form (cf. A. Zaffaroni, "Recent Progress in Hormone Research," Academic Press. Inc., New York, N. Y., Vol. VIII, 1853. p. 77,
5172
NOTES
Similarly cortisone acetate (111) and hydrocortisone (17a-hydroxycorticosterone) acetate (V) were isomerized to the corresponding D-homo compounds IV and VI. The latter two substances could be inter-related, for oxidation of the l l p hydroxy group of VI with chromic acid yielded the cortisone acetate rearrangement product IV. It was to be expected that the llp-hydroxy function would be unaffected during the rearrangement of hydrocortisone acetate '(V) to VI, since i t is known that the llp-hydroxy grouping is not attacked to any appreciable extent under the Oppenauer oxidation conditions employed.6
Yol. Tli
shift in the molecular rotation. This effect is in contrast to the rather large negative shift observed in the 21-desoxy series on passing from 17a-hydroxy-20-ketones to D-homo compounds of type I I a (Table I, lower part).
Experimental' 17ap-Acetoxyrnethyl-~-homo-A~-androsten-l 7aa-01-3,17dione (IIb) .--4 solution of 5 g. of Reichstein's substance S 31-acetate ( I h ) and 1.25 g. of aluminum isopropoxide in 200 cc. of carefully dried toluene was refluxed with 40 cc. of cyclohexanone for 1 hour, moisture being excluded. The cooled mixture was poured into water, the organic solvents mere removed by steam distillation, and the product was extracted with ethyl acetate. Washing the organic estract with water, drying and evaporation yielded a clear oil, CHiR which slowly crystallized. It was purified through chro1 matography on 250 g. of neutral alumina. The fractions C=O eluted with benzene-ether on crystallization from acetoneether furnished 2.28 g. (046%) of the rearrangement product I I b with m.p. 188-191 . The analytical sample was obtained by further crystallization from the same solvent pair, +134", Amax 240 mp and showed m.p. 194-196', [,.]'OD (log e 4.22), vmaX 1736, 1700 and 1670 cm.? and free hydroxyl band. The reaction with triphenyltetrazolium chloride was negative. Anal. Calcd. for C23H3205: C, 71.10; H , 8.30. Found: C, 71.27: H, 8.60. H H IIa, R = H, R' =
