2,3-Dimethyl-6-(1-methylethylidene)cyclohexanone

The regioisomeric problem upon fragmentation of the tosylhydrazone (i.e., production of mix 74/75) was overcome by increased substitution for increased selectivity. Therefore 3R-pulegone was used as a starting material for synthesis (Scheme VIII). The enolate of pulegone was generated with lithium isopropylcyclohexylamide (LICA) and alkylated with methyl iodide to furnish mainly 2,3-dimethyl-6-isopropylidene cyclohexanone 78 along with by-product 79, ##STR51## which was previously observed by Reusch et al., (Lee, R. A., McAndrews, C., Patel, K. M., and Reusch, W. Tetrahedron Lett. 965 (1973)) but 79 did not react in the following conversion of crude material. When the mixture containing the isopropylidene 78 was placed in acid and submitted to prolonged heating, acetone distilled prior to the water azeotrope of the epimeric mix of 80:81 (1:1.96, as determined by NMR) in 72% yield. The corresponding tosylhydrazone mix 82 from the mix of 80:81 was made as before and underwent n-butyllithium-effected fragmentation in TMEDA to a regioisomerically pure cyclohexenyl anion, which was capped with dimethylformamide to afford isomeric aldehydes 83 in 76% yield. The mix was treated with tris(trimethylsilyl)aluminum (III) etherate (L. Rosch and G. Altman, J. Organomet. Chem., 195, 47 (1980).It should be noted that this procedure has failed to work in our hands, as well as others: See B. Trost, J. Yoshida, and M. Lautens, J. Amer. Chem. Soc., 105, 4494 (1983). However, a procedure which works is provided in the experimental. The reagent can be titrated as described or by reaction with peperonal: L. Rosch, G. Altman, and W. Otto, Angew. Chem. Int. Ed. Engl., 20 581 (1981)) and followed by acetylation to provide a mixture of all possible diastereomers of 84. The lack of selectivity was surprising in contrast to the total synthesis of QHS, in which a synthetic intermediate substrate differs in the presence of a 6'-methyl instead of a larger alkyl chain. Regardless, the mixture 84 upon exposure to lithium diethylamide rearranged to a mixture of diastereomeric cyclohexylacetic acids, which upon rigorous chromatographic separation furnished geometric isomers of acid 85 in a 1:1 ratio by NMR, in 28% yield. The acid 85 was submitted to single-pot exposure to ozone and acidification to give trioxane 30 in 22% yield. Alternatively the acid 85 was methylated via the corresponding LDA-generated dianion to the propionic acid 86, which was subsequently reacted with ozone and acidified to provide trioxane 31. The two optically active trioxanes 30 and 31 have been assessed for their antimalarial activity. In addition, all three optically active trioxanes, 29, 30 and 31, display temperature-dependent NMR behavior. ##STR52##