1752
COMMUNICATIONS TO THE EDITOR
of bromodifluoromethane, chlorodifluoromethane and difluoroiodimethane are concerted one-step reactions2 but t h a t the a-dehydrohalogenations of chloroform and most other haloforms are stepwise processes involving the formation of small concentrations of reactive intermediate carbanions.laS3 We wish now t o report that ethyl orthothioformate and methyl orthothioformate, with potassium amide in liquid ammonia, undergo essentially complete transformation to tris-(alkylthio) . methyl anions, which, more slowly, lose thioalkoxy anions to give intermediate bis-(alkylthio)-methylenes, some of which are transformed to tetrakis-(alkylthio)-ethylenes. Both methyl and ethyl orthothioformate react with potassium amide in ammonia to give green solutions, which are similar in color to those that may be prepared from potassium amide and bis(methylthi~)-methane,~ but which acquire a brownish color on standing for several hours. Such a solution, prepared from 11.7 g. (76 mmoles) of methyl orthothioformate and 150 mmoles of potassium amide in about 150 ml. of ammonia, was allowed to reflux for an hour before the ammonia was allowed to evaporate. From the residue was isolated 1.8 g. (12 mmoles) of methyl orthothioformate and 2.7 g. (13 mmoles-40% based on unrecovered orthothioformate) of tetrakis-(methy1thio)-ethylene, m.p. 59-60', reported m.p. 61.5°;6 sole n . m ~band . ~ a t 7.63 T (in carbon tetrachloride). Tetrakis-(ethylthio) -ethylene, prepared analogously from ethyl orthothioformate, melted a t 50-53' and showed no melting-point depression when mixed with tetrakis-(ethy1thio)ethylene prepared from tetrachloroethylene and sodium thioethoxide.8 In at least one previously reported a-elimination the concentration of intermediate carbanion was large enough to observe d i r e ~ t l y . ~In the present case there are two observations that show that the reactant is transformed almost quantitatively to its conjugate base. After the addition of one equivalent of potassium amide to solutions of orthothioformates the green color is made no more intense by the addition of further amide. Treatment of a freshly prepared solution of the tris-(methy1thio)-methylanion with methyl iodide leads to the formation of 75%, of methyl orthothioacetate, b.p. 9S.5-100' (11 mm.), d2".j41.1197; molecular refractivity n 2 4 . 51.5680, ~ calcd. 49.21, found 49.22; sharp n . m . ~ -absorption .~ band a t S.20 T and one with three times the integrated intensity a t 7.92 T (neat). Anal. Calcd. ( 2 ) J. Hine and P. B. Langfurd, J . A m . C h e m . Soc., 79, 5497 (1957); J, Hine and U . C. I h f f e y , ibid., 81, 1131 (1959); J, Hine and A . U. Ketley, J . O r g . Chewz.. 2 5 , 606 (1'360). ( 3 ) J. Hine and S. J. Ehrenson. J . A m . Cheat. Soc., 80, 824 (1958). (4) We have alkylated such a solution of bis-(methylthio)-methane with propyl bromide and obtained t h e dimethyl mercaptal of butyraldehyde. T h e first report of such a reaction was b y Arens. Frohling and I'rOhling,5 who did not s t a t e what specific compounds they studied. ( 3 ) J. F. Arens, M. Frohling and A . Frohling, Rec. li'av chim., 1 8 , O f i 3 (1'3.5'3). (0) B. Fetkenheuer, H. Fetkenheuer and H. Lecus, B e y . , 6 0 , 2535 (1927). (7) A Varian A-60 instrument was used. (8) P. Claesson, J. puakt. Chem., 121 16,193 (1877). (9) C . G . Swain and E. G. Thornton, J . A m C h e m S o c . , 83, -1033 (1961).
Vol. 84
for C6H12S3: C, 35.67; H , 7.10; S,57.14. Found: C , 35.49; H , 7.42; S,57.31. The fact t h a t tris-(alkylthio)-methyl anions may be generated in high concentrations is to be attributed to the relative ease of removal of a proton from orthothioformateslO.lland t o the difficulty of displacing thioalkoxy anions from saturated carbon atoms. The tetrakis- (alkyl thio) -ethylenes formed in the reactions studied need not arise from the dimerization of bis-(alkylthio)-methylenes. They may just as plausibly be ascribed to the reaction of such a methylene with the much more abundant carbanions as shown below. (RS)3CH
+ SHp- +(RS),C- --+ RS-C-SI< + IZS(RS)2C=C(SR)y
+ RS-
f-
(RS)3C-C(SK),-
In view of the difficulty of displacing thioalkoxy anions from saturated carbon, an alternate mechanism involving the formation of the intermediate pentakis- (alkylthio)-ethane by the nucleophilic attack of tris-(alkylthio)-methyl anions on orthothioformate molecules seems improbable. Acknowledgment.-\Ye wish to acknowledge the support of this investigation by the U. S. -1rmy Research Office (Durham), (10) Cj-. L . H. Slaugh and E. Bergman, J . 07.8. C h e m . , 26, 3158 (1961); S. Oae, W. Tagaki and A . Ohno, J . A m Chenz. Soc., 83, 5036 (1961). (11) When triphenylmethane is added t o a green solution containing two equivalents of bis-(methy1thio)-methane per equivalent of potassium amide t h e sdlutidn becomes r e d . T h e red colo. is not discharged by t h e addition of excess triethyl orthothioformate. We therefore conclude t h a t both of t h e sulfur compounds are weaker acids t h a n triphenylmethane. JACK HINE SCHOOL OF CHEMISTRY RAPMOSD P. BAYER GEORGIA INSTITUTE OF TECHSOLOGT GARYG. HAMMER -4TLANTA 13, GEORGI.4 RECEIVED MARCH7, 1962
MICROBIOLOGICAL TRANSFORMATIONS. 11. THE AROMATIZATION OF RING A OF STROPHANTHIDIN Sir :
-1 recent paper described the conversion of strophanthidin (I) to the anhydrostrophanthidone I1 by Chaetomium g1obosum.l We report herewith the transformation of I to I1 and thence to the new phenol I11 by -Vocardia restrictus.
I11
IV
(1) C . J. Sih, S. M. K u p c h a n , 0. E1 T a y e b and A . Afonso, J . .Wed Pharix. C h c m . , in press.
May 5 , 1902
COMMUNICATIONS TO THE EDITOR
Fermentation of strophanthidin by the general procedure previously described, with Nocardia restrictus, produced two principal transformation products. Paper chromatography* revealed that both were less polar than strophanthidin. Separation was effected by chromatography upon silica gel and fractional crystallization. The product with higher Rfwas characterized as the anhydrostrophanthidone II.133The second product was a new phenolic compound (III), m.p. 263-265' dec., [ o 1 I z 7 ~+'iso ( c 0.76, pyr.) ; A&. 217 mp ( E 25,600), 2S0 mp ( E 2,200); '::A; 2 88 p , 3.03 p j 5.77 p , 6.22 p , 6.30 p , G.68 p . 4 Acetylation of 111 yielded a monoacetate, m.p. 206-20s0, [cy]*'D +6G0 (c 0.70, chf.); "::A: 215 mp ( E 22,100), 2 G i mp ( E 680), 275 mp ( E 6-10); 2 SS p ; 5.60 p , 5.72 p , 6.15 p , (3.30 p , 6.70 p. Catalytic hydrogenation of I11 with platinum (hydrogen consumption: 1 mole equivalent) yielded the dihydro derivative V, characterized by direct comparison with a
1753
formations. Work is currently under way to evaluate the hypothesis that biological conversion of I1 to I11 may proceed via AI-dehydrogenation to a 19-0xo-D~~~-dienone and then aromatization accompanied by liberation of a C1 unit.' This work was supported by grants from the Research Committee of the Graduate School from funds supplied by the Wisconsin Alumni Research Foundation. (7) Formaldehyde has recently been identified as appearing in stoichiometric ratio t o estrune after placental micrusomal aromatization of A'-androstenedione and 19-hydroxyandrostenedione ( H . Breuer and P. Grill, Z . physiol. Chenz., 324, 254 (1861)). However, in view of t h e facile biological interconversion of formaldehyde and formic acid (6. F. R I . Huennekens and 11.J . Osborn, Ado. iii Enzymology, 21, 369 (1969)), it i s possible t h a t t h e formaldehyde m a y have arisen b y teduction of a formic acid-equivalent.
DEPARTMENTS OF PHARMACEUTICAL S. MORRISKUPCHAN CHEMISTRY A N D PHARMACOGSOSY CHARLES J. SIH UNIVERSITY OF ~ ' I S C O S S I N K . KATSUI MADISOS6, WISCOSSIS 0. EL TAYEB RECEIVED FEBRUARY 24, 1962
SYNTHESIS OF BORON NITRIDE
Sir: A broad resemblance of boron nitride to carbon
sample prepared from dihydrostrophanthidin by the procedure of Turner and Xeschino.6 Catalytic oxidation of I in the presence of platinum and oxygen, followed directly by treatment with alkali a t Uo, yielded the 19-noranhydrostrophanthidone IV, m.p. 240-241°, [ c Y ] * ~ D+33' (c 2.67, chf.); A$:, 2.85 p , 5.130 p , 5.72 p 6.01 p , 6.15 p ; AEt,"," 230 mp ( E 24,700). Dehydrogenation of I V over palladium black in refluxing ethanol gave phenol 111. When anhydrostrophanthidone I1 was exposed to ,V.restrictus, a rapid and high-yield conversion to phenol 111 was observed. Fermentation of the 19noranhydrostrophanthidone IV under the same conditions resulted in exceedingly slow conversion to phenol 111. These findings are in good agreement with the recent studies of Morato, et ~ l . on the biosynthesis of estrogens from androgens with human placental microsomes. The latter authors' observations support the sequence A4androstenedione + 19-hydroxyandrostenedione + 19-oxo-androstenedione -+. . . ' > estrone for the steps in biological estrogen formation. Furthermore, i t was found t h a t 19-norandrostenedione was a poor substrate for estrogen formation. The parallel findings in the respective studies that the 19-nor-steroids are less effective phenol-precursors than the corresponding 19-oxo-steroids suggest that the stepwise sequence may be similar for the microbiological and placental microsomal trans(2) H. R . Urscheler, Ch. T a m m a n d T. Reichstein, Hrlu. C h i m . A d a , 38, 897 (1955). ( 3 ) A . K a t z , ibid., 40, 831 (1937). (4) Satisfactory analytical d a t a were obtained for t h e new compounds reported herein. ( 5 ) R . B. T u r n e r and J. A. hleschino, J . A m . Chem. Soc., 80, 4862 (1938). (6) T . hlorato, M. H a y a n o , R . I. Ilorfman and I,. R. Axelrod, Hiorhviri B i o p h y s . Res. Coinm., 6, 334 (1961).
in its various structural forms,'J combined with certain physical characteristics, makes boron nitride a material of considerable theoretical interest and of growing importance in nuclear, refractory, electrical and lubrication technologies. M'hile the synthesis of boron nitride has been studied fairly e~tensively,3,~-* current synthetic methods are beset with considerable technical difficulties. Xccordingly, a straightforward synthesis adaptable to laboratory or large-scale production of boron nitride is of interest. Orthoboric acid (1 mole) and urea ( 2 moles) are mixed and gradually heated. At about GOo the reactants melt with the evolution of water which is removed from the system in a stream of nitrogen or, preferably, under reduced pressure. As water is removed, the initially limpid liquid residue increases in viscosity and sets to a glass. Above about 165' (subsequent heating is most conveniently effected in a nitrogen stream) the glass , undergoes ~ decomposition, evolving a volatile sublimate and leaving a white, cinder-like residue (residue a t GOOo is lS.45 wt. 70urea and boric acid reactants.
