Diaxial hindrance

This can be achieved only in favorable cases, e.g., by differences in stcric hindrance at both reactive sites in compound 1 or by annulating the epoxide to a cyclohexane chair with locked conformation (principle of diaxial attack4). 

These results can be readily interpreted if the rate-determining step of these reactions is the formation of the tetrahedral intermediates 241 and 245. Indeed, there is a direct relationship between the relative ease of formation of these two intermediates (which depends upon their respective steric hindrance) and that of the acetamidoalcohol products 243 and 247 (R=H or CHj). In intermediate 241, when R=CH3, there is a strong 1,3-diaxial steric interaction between the N-CH3 group and the axial phenyl group. Such-steric interaction does not exist when R=H. Consequently, the formation of acetamidoalcohol 243 (via 241) should occur with ease only when R=H. This conclusion agrees with the experimental results. In the case of intermediate 245, when R=CH3, there is no 1,3-diaxial steric interaction caused by the N-methyl group. On that basis, intermediate 245 (R CHg) should be readily formed. However in this intermediate, there is a strong 1,3-diaxial steric... 

We are now in a position to interpret the reactions of steroid amines, particularly with respect to the dominant retention of configuration in substitution and the considerable or exclusive elimination of axial amino groups. Shoppee has pointed out [3] that the ratio, elimination/substitution, for a series of axial amines follows the order expected on the basis of steric hindrance to the formation of the axial alcohol. In simple mono- and bi-cyclic systems, as well as in acycUc amines, the proportion of olefinic products is far lower (often ca. 30% than for many steroids. This agrees with the concept that very high yields of olefins result from steric frustration of cis substitution. A semi-quantitative correlation of olefin yield with the number and nature of the sy -diaxial interactions at the reaction centre is even possible (Table 27). 

The effectiveness of conformational control of halogen addition is indicated in the formation of the diaxial 3a,4)3-dibromide from 5a-cholest-3-ene (97%) and the 2 3,3a-di-bromide from 5a-cholest-2-ene ca, 88%) [loi]. The slightly lower yield of the diaxial product in the latter case may result from steric hindrance by the C i9) methyl group to approach of the Br ion towards the 2j3-position, so that "parallel attack at the 3jS position could become relatively a little more effective. The possibility has not been excluded, however, that the diequatorial (2a,3j3) by-product arises by partial... 

It is sometimes possible to estimate the extent of hindrance by an examination of a molecular model of the substrate. While the sites that promote hydrogenation reactions are the comer atoms or adatoms on the catalyst siuface, models of such surface species are not commonly available. The classic procedure for determining the relative steric hindrance to adsorption has been to place a model of the substrate on a flat surface with the It cloud of the alkene perpendicular to the surface as depicted in Fig. 14.3. In this case adsorption from side B is clearly favored. A similar conclusion can be drawn on examining the adsorption of the n cloud on a surface comer atom as depicted in Fig. 14.4. Here adsorption from the B direction is favored because of the interference of the 3,5 diaxial hydrogens to adsorption from side A, but the difference between the two modes of adsorption does not appear to be as great as that assumed from consideration of Fig. 14.3. 

To account for the stereochemical outcome, we propose a mechanism in Scheme 38 that involves a metal-5-fra j -diene cationic intermediate I. A near chairlike transition state structure is proposed to account for the stereochemistry. The states A and B represent the most likely transition states that have the most bulky complexed RCO(BF3) group situated at the equatorial site to avoid 1,3-diaxial steric hindrance. A further comparison between the two stractures indicates that A is the preferred structure because it avoids steric hindrance between two neighboring equatorial substituents, as in B. The presence of an electron-withdrawing group, such as carbomethoxyl, is required for this cyclization. No cycli-zation occurs for other T) -pentadienyl compounds, including that having the substituent CH=CH2 at the C(2)-allylic carbon position. 

For this reason, the equilibrium between the two chair conformations will generally favor the conformation with the equatorial substituent. The exact equilibrium concentrations of the two chair conformations will depend on the size of the substituent. Larger groups will experience greater steric hindrance resulting from 1,3-diaxial interactions, and the equilibrium will more strongly favor the equatorial substituent. For example, the equilibrium of r rt-butylcyclohexane almost completely favors the chair conformation with an equatorial r rr-butyl group ... 

SUBSTITUENT STERIC HINDRANCE FROM 1,3-DIAXIAL INTERACTIONS (kj/mol) AXIAL-EQUATORIAL RATIO (at equilibrium)... 

The equilibrium will favor the chair conformation with the substituent in the equatorial position, because an axial substituent generates 1,3-diaxial interactions, a form of steric hindrance. 

Steric hindrance at the a-position has a considerable impact on the preferred conformation of the l,4-pentadien-3-one. While the dienone can adopt three in-plane conformations, only one of them is properly aligned for the Nazarov reaction to occur (Scheme 3.2, Eq. (3.1)). If R and R have small A-values, the divinyl ketone prefers the unreactive s-cisis-cis conformer, it being the most stable by avoiding 1,3-diaxial interactions between the allyl groups. Conversely, a-substituents possessing large A-values favor the formation of the reactive s-trans/s-trans conformer. Thus, the increased prevalence of the reactive conformer... 

---