Equatorial ketones

Reduction by that smallest reagent, electrons, gives the all-equatorial product. Since the ratic > products is the same as the ratio of starting materials (87 13), the reduction is totally stereoselectc The all-equatorial ketone gives 100% all-equatorial alcohol and the other isomer must give uoe other diastereoisomer (we cannot say which). 

An interesting aspect of this reaction is the contrasting stereoselective behaviour of the dimethyisulfonium and dimethyloxosuifonium methylides in reactions with cyclic ketones (E.J. Corey, 1963 B, 1965 A C.E. Cook, 1968). The small, reactive dimethyisulfonium ylide prefers axial attack, but with the larger, less reactive oxosulfonium ylide only the thermodynamically favored equatorial addition is observed. 

The reduction of an asymmetric cyclohexanone (e.g. a steroidal ketone) can lead to two epimeric alcohols. Usually one of these products predominates. The experimental results for the reduction of steroidal ketones with metal hydrides have been well summarized by Barton and are discussed in some detail in a later section (page 76) unhindered ketones are reduced by hydrides to give mainly equatorial alcohols hindered ketones (more accurately ketones for which axial approach of the reagent is hindered " ) are reduced to give mainly axial alcohols. 

The relation of rates of reduction with NaBH4 to variations in structure in a wide variety of monocyclic and bridged bicyclic compounds has also been discussed for example, a methyl a to a ketone slows the rate of reduction. Brown ° stated that reactions should not be discussed in terms of axial and equatorial attack, since the rates simply reflect differences in the energies of the possible transition states and not enough is known about the transition state to analyze it. He accepted th concepts of SAC and PDC, but preferred to call them steric strain contrpl and product stability control. ... 

By a suitable choice of conditions (metal hydrides or metal/ammonia) ketones at the 1-, 2-, 4-, 6-, 7-, 11-, 12- and 20-positions in 5a-H steroids can be reduced to give each of the possible epimeric alcohols in reasonable yield. Hov/ever, the 3- and 17-ketones are normally reduced to give predominantly their -(equatorial) alcohols. Use of an iridium complex as catalyst leads to a high yield of 3a-alcohol, but the 17a-ol still remains elusive by direct reduction. 

Because of the presence of alkali in Raney nickel, ketones are hydrogenated over this catalyst to yield the more stable, equatorial alcohol e.g. 59) as the predominant product, Similar results can be expected with platinum in basic media or with platinum oxide in an alcoholic solvent since this catalyst also contains basic impurities. 

The optical rotatory dispersion curves of steroidal ketones permit a distinction to be made between the conformations, and assignment of configuration is possible without resorting to chemical methods (see, e.g. ref. 36) which are often tedious. The axial halo ketone rule and, in the more general form, the octant rule summarize this principle and have revealed examples inconsistent with the theory of invariable axial attack in ketone bromination. 2-Methyl-3-ketones have been subjected to a particularly detailed analysis. There are a considerable number of examples where the products isolated from kinetically controlled brominations have the equatorial orientation. These results have been interpreted in terms of direct equatorial attack rather than initial formation of the axial boat form. 

In the absence of steric factors e.g. 5 ), the attack is antiparallel (A) (to the adjacent axial bond) and gives the axially substituted chair form (12). In the presence of steric hindrance to attack in the preferred fashion, approach is parallel (P), from the opposite side, and the true kinetic product is the axially substituted boat form (13). This normally undergoes an immediate conformational flip to the equatorial chair form (14) which is isolated as the kinetic product. The effect of such factors is exemplified in the behavior of 3-ketones. Thus, kinetically controlled bromination of 5a-cholestan-3-one (enol acetate) yields the 2a-epimer, (15), which is also the stable form. The presence of a 5a-substituent counteracts the steric effect of the 10-methyl group and results in the formation of the unstable 2l5-(axial)halo ketone... 

Bromination of 5j5-3-ketones yields the equatorial 4 -bromo compounds (22) as the thermodynamic or kinetic products,although the presence of a considerable amount of 2-bromo isomer has been reported in bromination with phenyltrimethylammonium bromide-perbromide. This is in keeping with other evidence that enolization of 5j5-3-ketones is not specifically directed to C-4. Cleaner results would probably be obtained via thermodynamic enol acelylation. ... 

The chemical reduction of enamines by hydride again depends upon the prior generation of an imonium salt (111,225). Thus an equivalent of acid, such as perchloric acid, must be added to the enamine in reductions with lithium aluminum hydride. Studies of the steric course (537) of lithium aluminum hydride reductions of imonium salts indicate less stereoselectivity in comparison with the analogous carbonyl compounds, where an equatorial alcohol usually predominates in the reduction products of six-membered ring ketones. 

Stereoselective epoxidation can be realized through either substrate-controlled (e.g. 35 —> 36) or reagent-controlled approaches. A classic example is the epoxidation of 4-t-butylcyclohexanone. When sulfonium ylide 2 was utilized, the more reactive ylide irreversibly attacked the carbonyl from the axial direction to offer predominantly epoxide 37. When the less reactive sulfoxonium ylide 1 was used, the nucleophilic addition to the carbonyl was reversible, giving rise to the thermodynamically more stable, equatorially coupled betaine, which subsequently eliminated to deliver epoxide 38. Thus, stereoselective epoxidation was achieved from different mechanistic pathways taken by different sulfur ylides. In another case, reaction of aldehyde 38 with sulfonium ylide 2 only gave moderate stereoselectivity (41 40 = 1.5/1), whereas employment of sulfoxonium ylide 1 led to a ratio of 41 40 = 13/1. The best stereoselectivity was accomplished using aminosulfoxonium ylide 25, leading to a ratio of 41 40 = 30/1. For ketone 42, a complete reversal of stereochemistry was observed when it was treated with sulfoxonium ylide 1 and sulfonium ylide 2, respectively. ... 

---