in benzene solution was achieved only with the coupling constants of opposite signs. The spacings between the lines shown in Fig. 1 at 241.7 and 243.7 cps. and 221.9 and 220.7 cps. (Fig. 1) were reduced to 1.4 and 0.1 cps., respectively, a t 40 Mc, IVith coupling constants of the same sign, the computed spacings for these lines were 0.S and 0.9 cps., respectively. On the basis of these results, we believe that the relative signs of JAC and J B C must be taken as being indeed different from JAB. Interestingly, this conclusion could only have been reached by treatment of the spin system as 9BC3, not as the ABX3 approximation in which a change in the sign of J A B relative to J.kc and J ~ does c not affect the theoretical spectrum. The results can be regarded as disquieting in two connections. First, the difference in sign between the geminal and vicinal H-H couplings runs couiiter to the theoretical predictions of Karplus arid Gutowskylo for substances with normal bond angles. Second, in analogy with the findings of Castellano and Ji'aughll for ABC systems, rather different sets of spin-coupling and chemical-shift parameters were found to give nearly identical theoretical spectra for diethyl sulfite in benzene and each of these corresponds amazingly well t o the experimental spectrum a t GO 11c. Indeed, the previously mentioned excellent agreement between theoretical and experimental spectra obtained for acetaldehyde diethyl acetal now has been duplicated with V B - v.4 = 9.23 cps., vc - V A = 146.3 cps., J l ~ c= 9.30 cps. and J A Ceqiinl to J B C = 7.03 cps. with J A Bopposite in sign to JAC and J B C . Clearly, caution and judgment must be exercised in use of the iterative method of obtaining spinspin coupling constants since there seems to be no assurance t h a t a "final" solution is truly Acknowledgment.-11-e are deeply indebted to Dr. Stanley L. Manatt and the Computing Center of the Jet Propulsion Laboratory for the I B I I 7090 calculations of the theoretical spectra. (IO) €1. S. Gutowsky, M . Karplus and D. hI. Grant, J. Chcnz. P h y r . , 31, 1278 (lL69); hI.Karplus, ibid., 30, 11 (19.59) (11) S. Castellano a n d J. S. Waugh, ibid., 34, 295 (1961).
-3 ,& --C":,
& J -
H0
(-,
'143
lI>IL.S
\
110
I
11, R'=H III? K ' = C H 3
progesterone by Oppenauer oxidation. Only about 20% of the "normal" product, the corresponding I 7a,20-dimethyl-5-pregnene-3P,20-diol, carbinol was formed and conveniently separated by chromatography. Analogously, the 3a,Cia epoxide obtained by peracetic acid treatment of I, m.p. 1TO-1'71", [ a ] D - 67.2, gave in similar yields 3fl,3a-dihydroxy-G@,lia-dimethylpregnan-2O-one, m.p. lSS-192",which was directly oxidized by chromic acid to the 3,20dione, n1.p. 230-233", and converted by niethanolic sodium hydroxide to 6 a ,17-dimethylprogesterone, ~ Am,, 239 mp, log E 4.2. 1n.p. 137-140°, [ a ]+90.3, The latter substance, and the corresponding B-dehydro derivative, m.p. 143-14(j0, [ @ I D +87.1, Amax 292 mp, log e 4.4, obtained by chloranil dehydrogenation, were orally a t least as active, in the Clauand berg test, as 19-nor-lia-ethynyltestosterone, twenty times more active than 170-methylprogesterone. KOTEADDEDIS P~uo~.-ij,l7-Dimethy~-6-dehJ.droprogesterone is as active orally as Ba-rneth~-l-17-acetoxyprogesterone. It is devoid of any androgenic properties.
Treatment of I with ethylmagnesium iodide in anisole gave lTa,21-dimethylpregnenolone 111, m.p. 145-14So,with no detectable amount of carbinol being formed. Oppenauer oxidation gave 1ia,21-dimethylprogesterone,m.p. 1S7-l~59°, [ a ] D +1Oi.i, A,, 2-1-1 m p , log E 4.2' in 705; over-all yield. Further transformations of these methylated steroids together with a report on their G-halogenated derivatives3 will be presented in a more detailed paper. .ail new compounds ( 2 ) All rotations in 1'; chloroiurm solution had satisfactory elemental analysis We are indebted t o l)r. G. Papineau-Couture and his associates for the analytical data and t o Drs C . Chappel and C.Kevesz for the bioassays. f ~ 3 )Prepared by Dr. Y. Lefebvre of these laboratories
CONTRIBUTION KO.2738 FREDKAPLAN GATESAND CRELLIN LABORATORIES OF CHEMISTRY .\"ERST RESEARCH LABORATORIES CALIFORNIA INSTITUTE O F TECHNOLOGY l i . I)EGHESGHI PASADESA, CALIFORNIA JOHX L). ROBERTS I). 0. Box 6115 R . GAUDRY MONTREAL, P. Q., CANADA RECEIVED AUGCST9, 1961 RECEIVED SEPTEMBER 20, 1961
STERICALLY CONTROLLED GRIGNARD REACTIONS. A NEW SIMPLE ROUTE TO METHYLATED STEROID ANALOGS
17a-
Say:
INERTIAL EFFECTS OF SUBSTITUENTS ON HOT ATOM CAPTURE'
Sir:
We wish to report a convenient new synthesis of Hot hydrogen atoms can replace hydrogen and 17a-methylated steroid derivatives which has made other atoms and groups with high collision effipossible the preparation of highly active proges- ciency.2 The course of these reactions has been terone homologs When the readily available shown to be largely controlled by steric factors.' (I) was However certain results, particularly on alkyl remethyl 3~-hydroxy-l7a-methyl-5-etienate refluxed in anisole for one hour in presence of placement, have not found explanation. This excess niethyhagnesium bromide, the correspond- letter reports a remarkable new effect in the hot ing sterically hindered ketone, 17a-methylpregnenol- hydrogen displacement of halogen atoms from one 11' was obtained in 70yoyield. The latter sub(1) Studies supported by the U. S . Atomic Energy Commissivll. stance was converted to the known' 17a-methyl- Contribution S o . lBS1 from the Sterling Chemistry Laboratory. ( 1 ) PI. 4 Plattner, H H r u s v r P Th H e r z i g , 32, 270 ( l Y l %
Hell
Chrin A r l o
(2) D. Urch and K. Tb-olfgang.,J . d i n . Chem Soc., 83, 2982 (1961). This paper contains references t o t h e earlier literature