1840
LOUISF. FIESERAND HUANG-MINLON [CONTRIBUTION FROM
THE
Vol. 71
CHEMICAL LABORATORY OF HARVARD UNIVERSITY]
The Serini Reaction BY LOUISF. FIESER AND HUANG-MINLON
A reaction introduced to the steroid field by Serini and eo-workers1 that has found service in the partial synthesis of cortical steroids consists in the elimination of the elements of acetic acid from a 17-hydroxy-20-acetoxy compound (I) by treatment with zinc dust. Fieser and Fieser2 pointed out that although in most of the reported in-
without experimental documentation) that the isopregnenolone acetate (V) could be obtained from either diacetate or from the unseparated mixture of the two resulting from the synthetic operations. If this is indeed the case, then one of the reactions proceeds without inversion. In a repetition of Butenandt’s work we found the synthesis of the isomers I11 and 1V to be a teR R dious process because the initial addition of ethylI I CO CHOAc magnesium halide (or of ethyllithium, as we found) to dehydroepiandro~terone~ is attended with extensive reduction of the carbonyl group. The re-AcOH maining steps of dehydration and osmium tetrox/-/ ide hydroxylation of the cis-trans mixture of 17I I1 ethylenes proceed satisfactorily, and the two triol stances the side chain a t C17 suffers inversion from diacetates I11 and IV were prepared in pure form the a- to the @-orientation,or the reverse, an ap- and each was submitted to the Serini reaction. parent exception to the rule is recorded by Buten- This reaction also is Iess smooth than would be andt.3 At a time when the configurations of the inferred from the literature and actually requires compounds a t C I ~were completely unknown, repeated reprocessing. A6-Pregnene-3P,17a,20PButenandt had conducted a synthesis that af- triol,3,20-diacetate (111) afforded 17-isopregnenoforded two isomers that have been assigned the lone acetate (V), the product isolated by Butenconfigurations shown in formulas I11 and IV on andt. 17-Iso- A5-pregnene-3/3,17P,20P-triol-3,20the basis of evidence, theoretical inferences, and a diacetate (IV), however, yielded pregnenolone proposed convention for representation of abso- acetate (VI). Therefore Butenanclt’s statement lute configurations a t Cz0presented in the article is in error and both reactions proceed with inverby Fieser and Fieser.2 Butenandt recorded a full sion as in all other examples of the Serini reaction, experimental description of the conversion of one including the case of a 20a-acetoxy compound reof the triol diacetates (111 or IV) into 17-iso-A6- cently reported by Shoppee.5 Fieser and Fieser2noted that the Serini reaction pregnene-3/3-01 acetate (V) but did not indicate which diacetate had been employed. He stated, may possibly proceed through a A1’-enol-20-acetate, a 17,20-oxide,or a cyclic ortho ester. MarshCH8 CHI all, et al., have found that A17-enol-20-acetates I I AcOCH AcOCH can be prepared from 20-ketones by the action of acetic anhydride and p-toluenesulfonic acid and in cHffl-OH one instance they isolated the cis and trans isomers. We have now prepared a further pair of such isomers in order to see if they behave like /\I intermediates in the Serini reaction. Treatment of pregnenolone with acetic anhydride and an acid A c O N catalyst and extensive chromatography of the reaction mixture afforded two isomeric enol diaceI11 tates, m. p. 147 and 172’. That the double bond Zn in both isomers is a t the 17,20- and not the 20,21CHa position was established by bromination, when both compounds gave the same tribromo derivative identical with 3/3-acetoxy-20-keto-5,6,17-tribromopregnane (IX), prepared by bromination of pregnenolone acetate according to Plattner and The substances thus appear to be co-~orkers.~
i l
___f
iA
‘ ‘U .1
v
VI
f1) Serini, Logemann and Hildebrand, Bcr., 74,391 (1Y3Y). (2) Fieser and Fieser, Ex9cricnfia, 4, 285 (1948) (3) Butenandt, Schmidt-Thom.6 and Paul, Bcr., 74, 1112 (1999)
(4) Revised nomenclature proposed by Fieser and Fieser, ref. 2, and in “Natural Products Related to Phenanthrene,” 3rd edition, 1949. Reinhold Publ. Corp., New York, N. Y., ( 5 ) Shoppee, Expcricnfio, 4, 418 (1948). (6) Marshall, Kritchevsky, Lieberman and Gallagher, THIS JOURNAL, 70, 1837 (1948). (7) Plattner. Heusser and Boyce. Hclo. Chim. A c t a . 81, 603 (1948).
THESERINI REACTION
May, 1919
the A"-enol-20-acetates VI1 and VIII, but we cannot as yet distinguish between the two formu-
oxide ring is 0 (f), then the rearrangement produces a ketone of the 17-normal series (g). R
I
VI
~
1
1
i 1 KHCO3
CHs
[
A-OH A.0CHOAC Zn
i KHCOI
1841
+
~ H O A C ] 1/2 Zn++
-AcO-
IC".
AcOCH
HCOAc
A d , ArS03H R
I
CHOAc
A- OH 1
AcO
AcO VI1
VI11
1
IX
las. T4e isomer with the acetoxyl group cis to the ring system (VII) would be expected to be hydrolyzed with less ease than that of trans structure (VIII), but both substances are hydrolyzed rapidly by potassium bicarbonate in aqueous methanol a t room temperature and any difference was not discernible under these conditions. If such an enol acetate were an intermediate in the Serini reaction, one isomer or the other should yield the ketone of the 17-iso series on hydrolysis. Both enol acetates, however, afforded the 17-normal pregnenolone as the sole product, and hence the Serini reaction must proceed by some other mechanism. Shoppee6 is of the opinion that the reaction proceeds through a 17,20-oxide, and this seems to us the most likely course. Shoppee's view that oxide formation is initiated by an attack by zinc a t Czo with removal of an acetate ion appears less plausible, because an acetoxyl group a t C S , and sometimes also one a t (221, escapes attack, and because the reaction can be effected with the free alcohol as well as with the acetate.8 The initiating reaction more likely involves attack by zinc of the 17-hydroxyl group, which is always present, with removal of hydrogen (b). Oxide formation can then occur (c) and then, as noted by Shoppee, if the oxide ring extends to the rear (c) migration of hydrogen from Czo results in inversion of the side-chain from the 0-to the a-orientation. If the starting material is a 17-is0 compound (e) and the (8) Slotta and Niesser, B r r . , 71, 2342 (1938).
-
R
R
I
;p,1+ co
k
