3394
COMMUNICATIONS TO THE EDITOR
VOl. 84
should be formed. In view of this we have now from C4 t o C1 (probably via a cyclic intermediate) ; carried out such experiments with ketene and found (2) hydrogen migration from C2 to C1 (1,2-shift of that triplet methylene is formed, as shown by its hydrogen); (3) hydrogen migration from C2to C3 nonstereospecific reactions with b ~ t e n e - 2 ~ ~and ~ ~ 7 (1,2-shift ~8 of hydrogen) ; (4) methyl migration from lack of random CH insertion reactions. C 2 to C1 or to C 3 (1,2-shift of methyl, occurring We have studied reactions of triplet methylene perhaps partly or entirely externally). generated in this manner with cis- and trans-buteneThe products from these migrations in the case of 2 and isobutene. I n all experiments the ratio of cis- or trans-butene-2 would be (1) 3-methylbuteneketene to oiefiii was 1:lO. At lower pressures the 1, (2) 2-methyl-butene-2, (3) 2-methylbutene-1 product composition varied as a result of secondary and (4) cis- and trans-pentene-2 and 3-methylbureactions of the non-stabilized "hot" adduct, as is tene-1, and in the case of isobutene, (1) 2-methylalso observed in the reactions of singlet m e t h ~ l e n e . ~butene-1, ( 2 ) 2-methylbutene-8 and (3) 3-methylSteady values were obtained in the case of cis- and butene-1. Examination of the product composition trans-butene-2 a t approximately 50 cm. and, in the shows that the probability of these reactions appears case of isobutene, a t about 20 cm. The composition to be in the order (4) > (1) > (2) > (3) for cis- and of the products in the high pressure region, ex- trans-butene-2 and (1) > (2) > (3) for isobutene. pressed as percentage of the total Cb compounds, In the case of isobutene (4) is structurally not possiis shown in the table. ble. This rearrangement scheme is consistent for the three olefins so far examined, but should of TABLE I course be verified on a much larger number of exIlirnethylamples. Further studies of the reactions of triplet cq clopropane Pentene-2 232inethylene are in progress. Cis- l l ' Q l l S hie" bleb hleC Reactant lrnifs-Butene-2 ci5-Butene-2 cis-Butene-Zd Isobutene
1,2 13.5 21.6 28.1 ,.
1,2
1,I
cis
trans
51.9 31.3 27.1
.. .. ..
5.6 19.3
18.7 0 . 7 12.3 0.7
17.0
16.5
..
..
01.0
..
B-I
B-1
B-2
..
6.8 2.8 8 1 3.7 10.5 4 . 0
6.5
0.5 2.3
2-Metliylbutene-I. b 3-Methy-lbutene-1. 2-Methylbutene-2. The values obtained from diazomethane photolysis in the presence of excess inert gas (ref. 7). a
Unlike singlet methylene, the triplet is seen to react nonstereospecifically with butene-2 and to give far more dimethyl-cyclopropane and 3-methylbutene-2. In the case of cis-butene-2 the results can be compared with those obtained from diazomethane photolysis in the presence of excess inert gas,I with which they agree reasonably well. In the isobutene reaction, comparison with singlet methylenegVl0 shows a large increase in the amount of 1,ldimethylcyclopropane with corresponding reduction in the yields of other products. The suggested mechanism for reaction of triplet methylene with olefins involves initial formation of a triplet addition complex (biradical I), in which partial rotation around the original carbon-carbon double bond is permitted and is responsible for the nonstereospecific formation of addition products. K,
I1
H
In the biradical I, lil = I3 and Rz = CHI for cis- and trans-butene-2, and R, = CH3and Rz = H for isobutene. Closure of the C1C2C3ring gives the corresponding dimethylcyclopropane. Since there is no evidence for random insertion of triplet methylene into CH bonds, the olefinic Cg products must be formed by rearrangement. This could occur by niigration of H,and perhaps of CH3 as well, in the biradical, ie. (1) hydrogen migration ( 8 ) K . R. Kopecky, G S. Hammond and P. A. Leermakers, J . Am. Chem. Soc., 84, 1015 (1902). (9) H. M.Frey, Proc. R o y . SOL.(London), A260, 409 (19.59). (10) J. H. Knox. A. F. Trotman-Dickenson a n d C. H. J. Wells, J . Chem. Soc., 2897 (1958).
(11) National Research Council Postdoctorate Fcllow.
F. J. UUKCAN" I)IVISIONO F APPLIED CHEMISTRY NATIONAL RESEARCH COUNCIL OF CANADA K. J. CVETANOVI~: OTTAWA, ONTARIO RECEIVED JULY 16, 1962 A NEW CARBOCYCLIC STEROID STRUCTURE : 12,lS-CYCLOSTEROIDS
Sir: The steroid nlolecule has been subjected to the most varied chemical and photochemical transformations. Among the most interesting recent changes have been those involving the angular methyl groups,' resulting in the inclusion of the C-19 methyl group in cyclopropanolj and cyclobutano'j structures and the C-1s methyl group in cyclobutanold and cyclopentano" structures. Phyllanthol, cycloartenol, cyclolaudenol, and cycloeucaleno13 possess the angular methyl groups in the cyclopropano arrangement. We wish to present herein the synthesis of the iiiteresting pentacyclic 12, IWehydrosteroidal ring system and the iiitcrconvertibility of the latter with the conaninc structure. (1) ( a ) IS. J. Coreg and N . R . IIertlei, . I . .4m,Chem. Snc., SO, 2903 (19;s); 8 1 , 5200 ( 1 9 5 9 ) ; (b) P. Riich,;chncher, J. Kalvoda, D. Arigoni and 0. Jeger, ibid., 80, 290.; (1!)58), ( c ) F. Greter, J. Kalvoda and 0. Jeger, Proc. Chcin. > o r . , 3-19 ( 1 9 i R ) ; (d) P. Buchschacher, hl. Cerephetli, H. Wehrli, K . Schaffner and 0. Jeger, f Z f l u . Chim. A c l n , 42, 2122 (19.59); N . C. V a n g and D. 11. Yane, Telrnhedron L e t f e r r . n o . 4 , I O (l9tiO); (e) G. Cainelli, hl. Lj. hlihnilnvic, I ) . Arigoni and 0. Jcxer, Ilelu. Chiin. A c t a . , 42, 1124 (1959); ( f ) C h . Meystre, K. Heusler, J . Kalvoda, P. Wieland, G. Anner and A. Wettstein, E x p r r i e a f i n , 11, 47,i (19lll\: (g) 1). H. R. Barton. J. hl. Beaton, I,. E:. Geller and RI. hl. Pechet, J . A m . C h r m .Tor., 82, 2610 (1900); r). H. R. Bart(m and J. M. Beaton. i b r d , 82, 20-11 (1960); A. L. Nussbaum. F. E. Carlon, E. P. Oliveto, E. Townley, P. Kabasakalian, and D. H. R. Barton, ibid., 82, 2973 (1000): D.H. R. Barton, J. M. Beaton, L. E. Geller and M. M. Pechet, ibid., 82, 4076 (1961) ; D. H. R. Barton and J. M. Beaton, ihid., 83,4083 (19G1); M. Akhtar and D. H. R. Barton, i b i d . , 84, 1496 (1962); (h) M Akhtar and D. H. R . Barton, ibid., 83, 2213 (1961); ii) D. H. R . Barton and L. R. Morgan, P r o c . Chem. Soc., 20ci (1901); (j) H. Wehrli, hI. S. Helbr, R. Schaffner and 0. Jeger, H c h . Chim. A c ~ Q44, , 2162 (1961); (k) far a review article, see K. Schaffner, D. Arigoni and 0. Jeger, Experienfia, 16, 169 (1960). (2) D. H . R Barton and P. deMayo, J . Chem. Soc., 2178 (1953). (3) For a summary of evidence, see L. F. Fieser and M. Fieser,
"Steroids," Reinhold Publishing Carp., New York, N. Y., 1959, pp. 386-389.
COMMUNICATIONS TO THE EDITOR
Sept. 20, 1962
3505
Corroboration for the 12,18-cyclo structure was obtained from further transformations : treatment of I I I b with HBr-HOAc (room temp.) cleaved the cyclopropyl ring with the production of the corresponding 18-bromo derivativelo ( X ~ ~ x 5.88 o M, elemental analysis confirming), which was readily converted (aq. RZCOa) to the conanine IIb. Cyclopropyl ring formation could also be effected I1 I by operating on the 11-ketoconanine IIb. The latter was transformed via the sequence : quaternization, Hofmann elimination to an 18-dimethylamino-20-pregnene, quaternization, and intramolecular alkylation (NaOCHSIDMF) of the 18trimethylammonium salt to the 12,123-cyclopregnene IVb (normal C=O absorption in all intermediates, but shifted to 6.0 pl1 in IVb). Compound IVb also could be generated from IIIb by a Hofmann elimination sequence. I1 I IV A representative substance in the A4-3-keto 239 mp ( E = 17,800); :"A,: 3.00, series V (A:".," 3.26, 3.30, 5.96-6.00, 6.20 p ) was readily available from the treatment of 18-chloro-20a-methylamino4-pregnene-3,ll-dione trifluoroacetate with alcoholic potassium hydroxide, whereas, treatment with potassium carbonate afforded 4-conanene3,ll-dione (VI) ":A(: 239 mp ( E = 18,000); VI Xz:;5.87, 5.98, 6.20 k ) . Ri Rz 5H It is interesting to observe that cyclopropyl ring a, CYOH O= formation competes, in media favoring enolization, p b, pOH O= 01 with a very facile pyrrolidine ring closure, even N C, @OH Hz though the 11-keto steroid structure is conspicuous In the transformations of 18-chloro-amines for its lack generally of reactivity transmittal I a and Ib4 with alkali, the conanines I I a and I I b toward C-12. Since under mild alkaline condiobtained were accompanied by considerable yields tions conanine formation predominates, and since of secondary amine by-products, which were iso- with acylated amine functions in I 12,18-ring lable as the neutral amides. The initially dis- formation results quantitatively, either ring system tinguishing feature of these substances was their may be generated by choice. 11-keto carbonyl absorption a t 5.98 p (in addition SMITH,KLISE& FRENCH LABS. 1'. G E O R C I A N ~ ~ to amide a t 6.10 p ) , indicating a modicum of conju- PHILADELPHIA J. F. KERWIN 1, PENNA., AND M. E. WOLFF LABORATORIES, INC. gation for this carbonyl when contrasted with JULIAN F. F. OWINGS PARK,ILLISOIS normal absorption a t 5.88 p in the cases of the FRANKLIN RECEIVED JUNE26, 1962 conanines I1 and the parent amine precursors to I. Intense absorption a t 202 mp ( E = 5,600),6 R. Kocsis, P. G. Ferrini, D . Arigoni and 0. Jegzr, Hclo. Chim. lack of same in the 240 mp region, and sharp A c(9) f a . 43, 2178 (1960), report a complex signal a t ca. 10.0 for 4n,5weak bands a t 3.27 p and 3.31 p6 (cyclopropyl methylene cholestane, and R. McCrindle and C . Djerassi, Chem. and methylene C-H stretching)' pointed to an 11- Ind., 1311 (1961), report doublets a t 20 and 35 c./s. for cyclopropyl keto-l2,18-cyclo structure 111. The n.m.r. spec- hydrogens of t h e cycloartane nucleus. See also L. M. Jackman, of Nuclear Magnetic Resonance Spectroscopy in Organic trum of I I I b was uninformative,* but definite "Applications Chemistry," Pergamon Press, Ltd., London, 1969, p. 52. evidence for the 12,lS-cyclo structure was ob(10) 'The direction of t h e cyclopropyl cleavage of I I I b with H B r tained from the 11-desoxo derivative IIIc (Wolff- is interesting. T h e usual mode of scission would have resulted in the Kishner reduction of IIIb), whose spectrum con- generation of a methyl group by attachmcnt of the proton on the methylene leading t o a secondary or tertiary carbonium ion on a ring tained two complex signals, each equivalent to position. cf. D. H. R . Barton, J. E. Page and E . W. Warnhoff, J . one proton, centered a t 7 0.67 and lO.OSe (CHC13 Chem. Soc., 2715 (1954), in the case of cycloartenol. T h a t bromide ion has in fact become attached t o methylene with the formation of reference).
CII
1
I
CK3 I
/
1
/
T
(4) J. F.Rerwin, M. E. Wolff, F. F. Owings. B. E. Lewis, B. Blank, A. Magnani, C. Karash, a n d V. Georgian, manuscript submitted for publication. ( 5 ) For references t o cyclopropyl-carbonyl conjugation in ultraviolet ahsorption, see E. M . Kosower, Pvoc. Chem. Soc., 25 (1962), and J . A m . Chcm. Soc., 80, 3261 (1958), and ref. 9. ( 6 ) Spectrum determined with a Perkin-Elmer Model 137-G grating spectrophotometer. (7) S. A. Liebman a n d B. J. Gudzinowicz. Anal. Chem., 33, 931 (1961). (8) Cyclopropyl methylene is lowered t o T ca. 8.9 b y conjugation with carbonyl as shown in t h e case [0.1,4]bicycloheptanone-2, G. Stork and J. Finici, J . A m . Chcm. Soc., 83, 4678 (1961). With our compounds general methylene a n d methyl absorption obscure this region and prevent identification of cyclopropyl proton resonance peaks.
18-bromomethyl indicates t h a t t h e 11 carbonyl has participated in a conjugative fashion a s in this mechanistic scheme. vie.
(11) There are many instances recorded of carbonyl infrared absorption displaced toward longer wave lengths when conjugated with cyclopropyl. Among others, see R . N. Jones and F. Herlink, J . Org. Chem., 19, 1252 (1954); A. Sandoval, G. Rosenkrantz and C. Djerassi, J . A m . Chem. Soc., 1 3 , 2383 (1931); €I. L, Slates and S . L. Wendler, i b i d . , 81,5472 (1959), and ref. 9. (12) Department of Chemistry, T u f t s Uui rersity, h!edfurd 55, Massachusetts.
