NOTES
Sept. 5 , 1952
imino-ether formation involving the resulting hydroxy nitrile ester has not been demonstrated. We observed, however, that the saturated compound, 2,2diethylvaleronitrile (obtained by hydrogenation of 2,2-diethyl-4-pentenenitrile over Pd:CaC03, b.p. 84-86' (20 mm.)), when subjected t o I11 IV, R C6Hb the same experimental conditions was recovered largely V, R CHI unchanged; only a small amount of the amide' (crystallized Experimental from petroleum ether, m.p. 68-69.5') was formed. Craig3 demonstrated that similar cyclizations of a-di8, p'-Thiodi-(propionthioamide) (111).-A mixture of substituted allylacetic acids or esters could be carried out 139.0 g. (4.1 moles) of hydrogen sulfide, 180.9 g. (1.43 moles) by the use of bromine. We became aware of this possibility of p-diethylaminopropionitrile and 700 cc. of absolute ethduring a study of the reactions of 2,3,4,5-bis-(AP-butenyl- anol was shaken a t 60' for 10 hours. The viscous, brownish residue which remained after removal of the low-boiling ene)-tetrahydrofuroic acid ( IV)8 (octahydrodibenzofuran4a-carboxylic acid). When either the acid or its methyl material weighed 150.0 g. (50.5% yield). Attempts to ester was treated with an excess of bromine in chloroform induce the thioamide to crystallize were unsuccessful. at 0'. there was obtained a neutral, colorless crystalline 6,S'-Di-( 4-phenyl-2-thiazolyl)-diethyl Sulfide (IV) .-A product (needles from acetic acid, m.p. and mixed m.p. solution of 17.5 g. (0.08 mole) of crude (111) in 175 cc. of 228-230' dec.) which contained but three atpms of bromine. absolute ethanol was treated with 20.0 g. (0.1 mole) of .4nal. Calcd. for Cl3Hl6O3Br8: Br, 52.2. Found: Br, phenacyl bromide and refluxed three hours. The cooled 51.8, 52.2. This we believed to be the tribromolactone mixture gave a greenish-white solid (21.3 g., m.p. 192-173') when a large volume of ether was added. The crude dihy(VI. drobromide was converted to the base (IV) with aqueous ammonia. After three crystallizations from hexane, 8.7 g. __f 2Br2 B rlf-)(\r' (43%) of pure (IV) resulted; m.p. 68.2-69.2' cor. (lit.' m.p. 68-69'). V \ O / Br/ \o/v Anal. Calcd. for C22H10N2S3: N, 6.86; S, 23.54; mol. wt., 408.6. Found:, N, 6.98; S, 23.48, 23.62; mol. wt., COOR co-0 420. Iv V p, p'-Di-(4-methyl-2-thiazolyl)-diethylSulSde (V).-The R = H , CHIreaction of 0.08 mole of crude (111) with 0.1 mole of chloroacetone in ethanol was run as indicated above. A crude More detailed study of the mechanism of the transformayield of 11.0 g. of dihydrochloride of (V), m.p. 195-197', tions involving allylic systems of this type may be expected to resulted; two crystallizations from ethanol-hexane gave contribute to the knowledge of the structure of the com- 7.4 g. (42%) of fine white needles, m.p. 215-216'. The pounds related t o IV and V. I t has been pointed out that base (V) was an oil. such a study has been undertaken in another laboratory.9 Anal. Calcd. for C12HlaN2S3.2HCl: N, 7.84; S, 26.91; C1-, 19.84. Found9 N,8.05; S, 26.97; C1-, 19.91. (7) C. L. Carter and S. N. Slater, J . Chcm. Soc., 130 (1946). (8) This structure was proposed by J. C. Hillyer, ef ol., Ind. Eng.
n--p
Chcm., 40,2216 (1948). An extensive series of compounds may be derived from the aldehyde formed by condensation of furfural with butadiene. We do not regard this structure as unequivocally established. (9) See Craig, ref. 3, footnote 6.
RESEARCH LABORATORIES THE XATIONAL DRUGCOMPANY PHILADELPHIA 41. PA.
(2) Analyses by Mr. M. E. Auerbach and staR of the Analytical Laboratories of this Institute.
STERLING-WINTHROP RESEARCH INSTITUTE RENSSELAER, NEWYORK
Polarographic Behavior of 12-Ketosapogenins
/3, B '-Di-(2-thiazoly1)-diethyl SuEdes
BY CONSTANTINE RICCIUTI, C. 0. WILLITS,M. E. WALLAND M. M. KRIDER RECEIVED DECEMBER 19, 1951
BY EDGAR A. STECKAND LYNNT.FLETCHER RECEIVED APRIL29, 1952
Steroidal compounds containing a,&unsaturated keto groups are polarographically reducible. EisenA recent publication of Dahlbom' prompts us to brand and Picher' and Sartori and Bianchi2 found record observations of similar nature in attempts that steroids such as testosterone, progesterone, to prepare 2- (P-diethylaminoethyl)-4-phenylthia- pregnenol-17-one-3 and desoxycorticosterone, are zole (I). I t was found that the action of hydrogen reducible a t the dropping electrode in aqueous sulfide on @-diethylaminopropionitrilehad not pro- ethanol solutions and give waves which are proporduced the expected thioamide (11),but, rather, an- tional to concentration. The polarographic other substance, since (I) did not result from the method was applicable only to the A4-unsaturatedreaction of the intermediate with phenacyl bromide. 3-ketosteroids, for their saturated analogs did not The intermediate, which we did not obtain in crys- give polarographic waves. Wolfe, Hershberg and talline form, was indicated to be P, @'-thiodi-(pro- Fieser3s4 investigated A'-cholestenone and found pionthioamide) (111) through analyses of the com- that i t was reducible. All of these reducible compounds (IV) and (V) formed by reaction with pounds contain an a,P-unsaturated keto group. phenacyl bromide and chloroacetone. Only the They also extended the polarographic method to 8,@'-di-(4-substituted-2-thiazolyl)-diethyl sulfide include 17-ketosteroids and 20-ketosteroids by type could be isolated, and no further study was reaction of these steroids with Girard Reagent T to done. Dahlbom' has given the problem more de- form polarographically reducible Girard derivatives. tailed consideration. There have been no previous reports on the polarographic behavior of 12-ketosapogenins which
I
I1
(1) R. Dahlbom, A c f a Chcm. Scand., 6 , 690 (1851).
(1) J. Eisenbrand and H. Picher, 2.p h y s i d . Chcm., 260, 83 (1939). (2) G.Sartori and E. Bianchi, Gaaz. chim. ital., 74,8 (1944). (3) 1. K. Wolfe, E. B. Hershberg and L. F. Pieser, 3 . B i d . C h o n . . 186,668 (1940). (4) J. K. Wolfe, E. €3. Hershberg and L. F. Fieser, ibrd.. 140, 215
(1941).
1701. 74
NOTES TABLE I [DENTIPICATION CHARACTERISTICS OF SAPOGENINS USED IN Melting point. 'Ca Specific rotation6 Genin Acetate Geum Aceta-
Sapogenin
- 70
THISSTUDY
- 71 -3 -80
Infrared absorption
Carbonyl absent t 7 Carbonyl max. at 1706 ern.-' 54 Carbonyl max. a t 1714 cm.-l - 2 -4.2 Carbonyl max. a t 1709 cm.'' Carbonyl max. at 1676 - 7.5 C=C max. a t 1602 cm.-' * Rotations determined a t 25', sodium lamp, concen0 All melting point determinations made with the Kofler block. 95% pure, as estimated from ultraviolet and infrared absorption, mith approximately trations between 8-10 mg./ml. 5% of a non-conjugated carbonyl sapogenin. 207-209 260-261 241-243 244-246 232-233
Tigogenin Hecogenin Kammogenin Manogenin 9,ll-Dehydromanogenin'
205-206 345-216 263-234 248-250 258-260
are important as precursors in steroid syntheses. The present method for the determination of polarographic behavior of these compounds allows the use of a non-aqueous medium consisting of a lithium chloride methanol-benzene electrolytic solution in which the 12-ketosapogenins are soluble. Experimental
A Sargent Model XXI Polarograph was used to obtain the current-voltage curves. The capillary had t and m
-
fusion current constant of 2.33. This diffusion current constant is similar to that found for other conjugated ketones in the non-aqueous electrolyte. Mesityl oxide, for example, has a diffusion current constant of 2.07. The reducibility of the 12-keto group of 9,11-dehydromanogenin in the nonaqueous electrolyte was expected because of the conjugated carbonyl group in this compound. Table 11 shows that the wave height of the 9,11dehydromanogenin is directly proportional to concentration in the range studied (10 to 40 mg./ 30 ni1.j. The use of the lithium chloride nonaqueous electrolyte made possible the direct polarographic analysis of these water-insoluble steroidal compounds. Acknowledgment.-The authors acknowledge the assistance of C. R. Eddy in obtaining the infrared data for the sapogenins.
values of 1.35 seconds and 3.587 mg. per sec., respectively, which gave a capillary constant of 2.46 mg.z/wec.-":. The m and 1 values were obtained in an open circuit, with the polarographic cell maintained at 25.0°, and with the capillary dipping into the non-aqueous electrolytic solution. The capillary constant a t 1.SO volts under the above conditions was 2.38. This value has been used to calculate the diffusion current constant of the 9,11-dehydromanogenin. The electrolytic cell was a modified Lingane H-cells with a saturated calomel reference electrode. This cell had an open circuit resistance of 1175ohms, and all half-wave potentials were corrected for I R drop. Half-wave potential readKEGIOSALKESEARCH LABORATORY' ings were made against the saturated calomel oelectrode, and EASTERN 18, PENXSYLVASIA The sapo- PHILADELPHIA the polarograms were obtained a t 25 rt 0.1 __ genins were isolated and characterized by methods developed (73 One of the laboratories of t h e Bureau of Agricultural and I n nt this Laboratory.6 Descriptive data for these compounds dustrial Chemistry, Agricoltural Kesearch Administration, United are presented in Table I. Thirty milliliters of the electro- States 12epartment of .-igriculturr Article not copyrighted. lytic solution, consisting of 0 . 3 Mlithium chloride in a 50-50 (by volume) mixture of absolute methanol-benzene, was measured into the sample arm of the H-cell. The solution was degassed with high-purity nitrogen, and a polarogram The Heat of Neutralization of Hydrogen-Bentonite was recorded. The sample was then added (10-40 mg.), and the solution was again degassed with nitrogen. A u\\\., H . SLABAUGH polarogram was recorded, and from t h e increase in wave height, the diffusion current of the reducible sapogenin w u RECEIVED APRIL 5 , 1952 calculated. Many clays undergo stoichiometric reactions TABLE 11 with various types of reagents. Reactions which POLAROGRAPHIC CHARACTERISTICS OF ~ , ~ ~ - D E H Y D R O M A N O -
-
.
GENIN W t . sample per 40 ml.
per liter
i d , pa.
rd/C
id / Cmz/sP/a
0.0121 ,0258 ,0390
0.000885 ,001887 ,003019
4.76 10.10 15.70
5516 5489 5643
2.32 2.31 2.37
Mole
Results and Discussion Tigogenin, which has no keto group, shows no polarographic reduction. Hecogenin and manogenin, both having a 12-keto group but no unsaturated linkage, show no reduction. Kammogenin, which has a 12-keto group and a A5-unsaturated linkage, also does not reduce. However, 9 , l l dehydromanogenin, which has both a 12-keto group and a conjugated unsaturated linkage, reduces a t the dropping electrode with a half-wave potential of -1.72 volts 8s. S.C.E. and has a dif(5) C. 0. Willits, C. Wcciuti, H. I). Knight and D. Swern, A u n t . Chem., in pree. (6) M . E , n'all. c l ni., J , B i d . C h r m i n prpss.
involve the base-exchange character of clays have been widely investigated and are adequately reviewed elsewhere.' I n previous worke there was good indication that the exchangeable hydrogen which was present in electrodialyzed bentonite systems was of two principal types, namely, hydrogen ions in the adsorbed layer of the colloidal micelles and hydrogen ions at the base-exchange sites. Further, these ions showed primary and secondary characteristics comparable to a typical diprotic acid. The present work which consists of a thermodynamic study of the neutralization of these hydrogen ions gives good indication that there is a fundamental difference between these types of hydrogen ions and that, during storage, there is a definite tendency for the absorbed ions to become more intimately associated with the clay micelles. (1) E . A. Hauser, Ciicnt. R E V S 87, . . 807 (1946); C. E, Marshall, "The Colloid Chemistry of the Silicate Minerals," Academic Press, I n < , S e a T o r k , N . Y , , 1940. ' ~ 2 )11.. 11. Sla!,:iugli a n , i j I C:iilherisoti, .I Pliys. Colloid Cheni.. 66, 7.14 I 1951)
