Autoxidation of cholest-5-en-3-one, and its accompanying isomerization, in acetic acid

Abstract

The rearrangement of cholest-5-en-3-one to cholest-4-en-3-one in acetic acid in the dark is of the first kinetic order and shows the characteristics expected for catalysis by acetic acid molecules. If no special precautions are taken to avoid the presence of oxygen and traces of metal ions, a smooth autoxidation, also of the first kinetic order, accompanies this rearrangement and gives 6α- and 6β-hydroperoxycholest-4-en-3-one together with cholest-4-ene-3,6-dione and other products of oxidation. The rate of autoxidation is not increased by five-fold increase in the concentration of oxygen nor by a 1 000-fold increase in the concentration of copper(II) ions; it can be reduced by the inclusion of free-radical chain inhibitors and stopped by the inclusion of ethylenediaminetetra-acetic acid, when the rate of isomerization is the same as that in the absence of oxygen. When autoxidation is proceeding, this isomerization is reduced in rate but is not completely eliminated. Similar reactions in ethanol or in carbon tetrachloride are much slower. These results show that isomerization proceeds by at least two heterolytic pathways, and that the rate of formation of an intermediate concerned in one of these can control the rate of an autoxidation by molecular oxygen catalysed by traces of metal ions. Study of primary (4-D) and solvent isotope effects shows that both mechanisms for isomerization are intermolecular, and involve stereoselective removal of 4β-H and introduction of 6β-D. The first may be formulated as a synchronous syn-S_E2′ process. The second involves an intermediate which can be intercepted by oxygen in the presence of trace amounts of metal ions. The kinetic measurements show that in the absence of oxygen this intermediate gives more rapidly the unexchanged starting material than the rearranged enone; so it cannot be the free enol (which is known to be protonated much more rapidly at the 6- than at the 4-position), nor the enolate ion (which would give exchange at the 4-position). Isotope effects suggest that this intermediate should be formulated as an ion-pair, and establish also that exchange of both 4β- and 4α-hydrogen atoms with the solvent occurs at a rate which is significant but slower than those of the other reactions occurring in this solvent.