2662
V. R. MATTOX, B. F. MCKENZIE AND E. C. KENDALL
Vol. 70
[ CONTRIBUrION FROM THE DIVISION OF BIOCHEMISTRY, MAYOFOUNDATION]
Hemihydrohalides of 3 (a)-Hydroxy Steroids' BY VERNON R. MATTOX, BERNARD F. MCKENZIE AND EDWARD C. KENDALL The formation of a crystalline compound by in an ice-bath. It was hoped that a single solvent treatment of 3( a)-hydroxy-l2-keto-Ae~ll-cholenic or a mixture of solvents would be satisfactory for acid in ether with hydrogen bromide has been ob- the isolation of all of the hemihydrohalides; served.2 The experimental details for preparation however, it was found that this was not possible. of this product have recently been published,8and For several of the methyl esters methanol was a the procedure was suggested as a method for sepa- satisfactory solvent; acetone was used for some ration of the cholenic acid. However, no struc- of the acids and chloroform-petroleum ether was ture was suggested for the product and no analysis suitable in other cases. The hemihydrohalides were dried under reduced was given. In the presence of hydrogen bromide or hydro- pressure over sodium hydroxide or in air a t room gen chloride, crystalline products have been pre- temperature for one to two days. Methyl 3(a)pared in this laboratory with the already men- hydroxy-12(a)-bromo-Ag-"-cholenate hemihydrotioned cholenic acid and also with methyl 3(a)- bromide retained 0.5 molecule of hydrogen brohydroxy-12-ketocholanate. With both of these mide after it had been dried a t 0.1 mm. pressure a t steroids it was found that the crystalline com- room temperature for six hours. Most of the pounds separated in combination with 0.5 mole- hemihydrohalides melt over a range of several decule of hydrogen halide. grees and the melting point is frequently deSubsequently it was observed that, when pendent on the rate of heating. Their physical methyl 3(a)-hydroxy-12(a)-methoxy- Aesl l-cholen- constants are given in Table I. ate was treated with hydrogen halides, the 12(a)The halogen content of the compounds was dehalogen derivative which was produced separated termined by distribution between water and an in crystalline form and contained in each instance immiscible solvent and titration of the halogen in 0.5 molecule of halogen acid. the aqueous phase by the method of Volhard. In The diverse nature of the material treated with 6 experiments the amount of acid in the aqueous hydrogen halide and the fact that only 0.5 mole- phase was determined. In each instance the concule of halogen acid was combined with the crys- centration agreed with that of the halogen ion. talline product indicated that neither the double That no deep-seated change of the steroid molebond a t Cs-Cll nor the substituent a t Clz was cule had occurred was shown by isolation of the essential. It occurred to us that it was desirable starting material from the organic phase after reto determine what structure of the steroid was moval of hydrogen halide. Two of the reaction necessary for the formation of the hemihydro- products, methyl 3 (a)-hydroxy-12(a)-bromo- A9t11halide derivatives. Such an investigation has re- cholenate hemihydrobromide and methyl 3(a)vealed the fact that all derivatives of cholane, hydroxy-l2(a)-chloro-Ag,ll-cholenatehemihydroAg*"-cholene and All-cholene which have been chloride, were prepared by treatment of methyl studied (Table I) in which there was an (a)-hy- 3(a)-hydroxy-12(a)-methoxy- Aesll-cholenatewith droxyl group a t Cat form crystalline products and the appropriate hydrogen halide. These two separate with 0.5 molecule of halogen acid. If were reconverted into methyl 3(a)-hydroxythe 3(a)-hydroxyl group is esterified, as with the 12(a)-methoxy-AS*"-cholenateby treatment with acetyl group, addition products are not f ~ r m e d , ~methanol. as shown in the last three lines of the table. No It is thought that these hemihydrohalides of the attempt has been made to prepare hemihydro- 3(a)-hydroxy steroids are oxonium compounds, halides from S(P)-hydroxysteroids or from steroids Favorskii' has prepared oxonium compounds from which have hydroxyl groups in other positions. the aliphatic alcohols and reports that they are exThe general method of preparation consisted of tremely hygroscopic. However, the oxonium the introduction of a stream of dry hydrogen halide into a solution containing from l to 3 milli- compounds of the steroids reported in this paper moles of the compound in an appropriate solvent are not hygroscopic. In addition, Favorskii also) prepared from 2,2-dimethylpentanol-3 both a 696
(1) This paper was presented at the Fifteenth Midwest Regional Meeting of the American Chemical Society, Kansas City, Missouri, June 24, 1947. (2) Reported by DI. E. S. Wallis at a conference of the Committee on Synthesis of Adrenal Hormones held under auspices of the National Research Council, Washington. D. C.. 1942. (3) Hicks, Berg and Wallis, J . B i d . Chcm., la,633 (1946). (4) Since only the 3(n)-hydroxyl group appears to be necessary for formation of hemihydrohalides, separation of B(n)-hydroxy-lZketo- A*v11-cholenic acid from 3(a)-hydmry-l2-ketocholpnic acid cannot be made satisfactorily. For a discussion see McKenzie, Maltox, Engel and Kendall, J . B i d . Chcm., lTS, 271 (1948).
(5) Mattox, Turner, Engel, McKenzie, McGuckin and Kendall, J . B i d . Chem.. 164, 569 (1946). (6) When a solution of methyl 3(u)-hydroxy-l2(n)-bromo-Ae,l1cholenate hemihydrobromide in chloroform is repeatedly concentrated under reduced pressure and diluted with petroleum ether, methyl 3(u)-hydroxy-12(~)-bromo-A~~~~-cholenate is obtained; however, because of its somewhat variable melting point, it is not satisfactory for identification. For this reason the 12-halogen compounds were converted into the 12(u)-methoxy compound, which has characteristic physical properties. (7) Favorskii, Cheni. Abstr., 8, 493 (1914).
HEMIHYDROHALIDES OF 3 (a)-HYDROXY STEROIDS
Aug., 1948
PHYSICAL
TABLE I CONSTANTS OF THE HEMIHYDROHALIDES Halogen acid
Starting material
2663
Methyl 3(a)-hydroxy-l2(a)-bromo-Ag*ll-cholenatec HBr Methyl 3 ((1)-hydroxy-12 (a)-chloro- AgJ1-cholenatec HCl HBr 3(a)-Hydroxy-12-ketocholanicacid HC1 HBr Methyl 3((1)-hydroxy-12-ketocholanate
Solvent
CHCla-P. Ed CHCla-P. E. Ether CHCls CHJOH
M;g.,*
yield %
Addition product % of 0.5 HX prepared As 45 After daysb
100 97*-' 94 go'*' 54 103 8 6 8 3 62 55 98 78 99 55 E 97 98 3( a)-Hydroxy-l2-keto- A9s11-cholenicacid 40 95 CHaOH 65 96 Methyl 3(a)-hydroxy-12-keto-A9*11-cholenate CHaOH 86 95 36 HBr Acetone 39 98 90 3(a)-Hydroxy- A1l-cholenic acid HCl CHCL 96 74 Methyl 3 (a)-hydrox y- A "-cholenate 93-98 81 94 80 115-122 45 92 Methyl 3(a)-hydroxycholanate 105-106 75 30 10 3(a)-Hydroxy-11-keto-24,24-dipheny1-Az3-cholenek HBr Acetone 143-150 93 99 93' Methyl 3(a)-acetoxy-12(a)-chloro-A9t1l-cholenate HCl CHCla-P. E. 0 3(a)-Acetoxy- 12-ketocholanic acid HC1 CHClo-P. E. 0 3(a)-Acetoxy-12-keto- ASv1l-cholenicacid HCl CHCla-P. E. 0 a All melting points were determined on the Fisher-Johns apparatus. Samples were exposed t o atmosphere. c This compound was prepared from methyl 3(n)-hy?roxy-12(a)-methoxy-A9,11-cholenate. The conversion of the 12(a)P. E. is petroleum methoxy compound to the 12(a)-halogen derivative has been shown to be nearly quantitatiw.8 ether. e Analysis by combustion. This figure is based on the halogen found in excess of that calculated for 1 atom of halogen a t Cia. 0 For halogen determination the sample was heated in 1 N methanolic sodium hydroxide for ten minutes and the chloride ion was titrated by the Volhard metho+. M. p. 112-115" when heated rapidly. Crystals separated when the solution was about 2 N with HCl; as the solution became more concentrated with HCl the crystals dissolved. I M. p. 90-93' when heated rapidly. Prepared from methyl 3,9-epoxy-ll-ketocholanateB by treatment with phenylmagnesium bromide and dehydration of the resulting carbinol with acetic acid. The 3,g-epoxy linkage was opened with hydrogen bromide and the halogen was removed from C12with zinc in acetic acid. 1 After 45 days in desiccator over solid sodium hydroxide.
[ig. : ! : E
{
137-138 131-137 128-133 115-117 127-130 108-109h 134-l3? 119-122 117-123 85-90' 118-122
'
hemihydrohalide, (C7HI60H)2.HBr,and a hydrohalide, C7HI6OH.HBr. Only the hemihydrohalides of the 3(a)-hydroxy steroids could be isolated. Favorskii represented the structure of the hemihydrohalide of the aliphatic alcohols by H H R-O-O-R. H X H b:g:H]
related to the intrinsic strengths of the two acids. The steroid hemihydrohalidesare unstable compounds and slowly lose halogen acid under atmospheric cogditions. A comparison of the last two columns in the table shows the amount of halogen acid lost in forty-five days.
Summary
However, Whitmore* suggested ++
[Br]-[OR]- as being more probable.
'Ontent Of Some Of the oxonium The compounds was found to be less than the retical amount for a hemihydrohalide. This was observed more frequently with the hydrochlorides than with the hydrobromides and may be (8) Whitmore, "Organic Chemistry," D. Van Nostrand Co., Piew York, K. Y., 1937, p. 139.
Crystalline oxonium compounds which contain of hydrogen are by treatment of 3(a)-hydroxysteroids with hydrogen halides. Acetylation of the hydroxyl group prevents formation of the oxonium compound. The preparation of sixteen of these hemihydrohalides is described in this paper. a half
ROCHESTER, MI".
RECEIVED FEBRUARY 24, 1948
(9) Turner. Mattox, Engel, McKenzie and Kendail, J . B i d . Chcm., 166, 345 (1946)
