Protection of ketones against reduction

Ketonic carbonyl groups are commonly encountered in steroids and their reduction is facile, even in the absence of an alcohol. The lithium-ammonia reduction of androsta-l,4-diene-3,17-dione affords androst-4-ene-3,17-dione in 20% yield but concurrent reduction of the C-17 ketone results in formation of testosterone in 40% yield, even though the reduction is performed rapidly at —40 to —60° and excess lithium is destroyed with solid ammonium chloride. Similar reduction of the C-17 carbonyl group has been observed in other compounds. In the presence of an alcohol, a ketone is complete- 

A carbonyl group cannot be protected as its ethylene ketal during the Birch reduction of an aromatic phenolic ether if one desires to regenerate the ketone and to retain the 1,4-dihydroaromatic system, since an enol ether is hydrolyzed by acid more rapidly than is an ethylene ketal. 1,4-Dihydro-estrone 3-methyl ether is usually prepared by the Birch reduction of estradiol 3-methyl ether followed by Oppenauer oxidation to reform the C-17 carbonyl function. However, the C-17 carbonyl group may be protected as its diethyl ketal and, following a Birch reduction of the A-ring, this ketal function may be hydrolyzed in preference to the 3-enol ether, provided carefully controlled conditions are employed. Conditions for such a selective hydrolysis are illustrated in Procedure 4. 

The application of the Birch reduction to ethers of estradiol by A. J. Birch opened up the area of 19-norsteroids to intensive research. The major Birch reduction product is an enol ether which affords either a 3-keto-A -or a 3-keto-A -19-norsteroid depending upon the hydrolysis conditions. Various 19-norsteroids have been found to have useful clinical activity compounds (30), (31), and (32) are oral contraceptive agents and compound (33) has been used as an oral anabolic agent. Several of these compounds were prepared on an industrial scale for a number of years by the Birch reduction of estradiol derivatives. 

The term Birch reduction was originally applied to the reduction of aromatic compounds by alkali metals and an alcohol in ammonia. In recent years many chemists have used the term to include all metal-ammonia reductions, whether an alcoholic proton source is present or not. The author prefers to use the term Birch reduction to designate any reduction carried out in ammonia with a metal and a proton donor as or more acidic than an alcohol, since Birch customarily used such a proton donor in his extensive pioneering work. The term metal-ammonia reduction is best reserved for reductions in which ammonia is the only proton donor present. This distinction in terminology emphasizes the importance of the acidity of the proton donor in the reduction process. 

The Birch reduction of a benzenoid compound involves the addition of two electrons and two protons to the ring. The order in which these additions occur has been the subject of both speculation and study. Several reviews of the subject are available and should be consulted for details. The present discussion is concerned with summarizing data that is relevant to understanding the reaction from the preparative point of view. For convenience, reaction intermediates are shown without indicating their solvation by liquid ammonia. This omission should not obscure the fact that such solvation is largely responsible for the occurrence of the Birch reduction. 

---