Stereocontrol in organic synthesis using silicon-containing compounds. A synthesis of nonactin

Abstract

The enantiomeric purity of (1R)-1-(1′-naphthyl)ethanol 7 was raised by Horeau’s method by separating its oxalate 11 from its diastereoisomer by crystallisation. The alcohol 7 was used to open the anhydride of (3RS,4SR)-3,4-bis[dimethyl(4-methylphenyl)silyl]hexane-1,6-dioic acid with selectivity of 96∶4 for one of the enantiotopic carbonyl groups, allowing the synthesis of (3R,4S)-3,4-bis[dimethyl(4-methylphenyl)silyl]hexane-1,6-dioic acid 6-(2-trimethylsilylethyl) ester 10. This acid was converted into the allylsilane methyl (E)-(3S,4R)-3,4-bis[dimethyl(4-methylphenyl)silyl]-7-(2-methyldioxolan-2-yl)hept-5-enoate 15, the carboxylic acid derived from which underwent epoxidation with unexpected silyl migration to give (3S,4S,5S,6R)-3,5-bis[dimethyl(4-methylphenyl)silyl]-6-hydroxy-7-(2-methyldioxolan-2-yl)heptano-4-lactone 17. Desilylative elimination and hydrogenation then gave the alcohol (3R,6R)-3-[dimethyl(4-methylphenyl)silyl]-6-hydroxy-7-(2-methyldioxolan-2-yl)heptanoic acid 19, in which the relative and absolute configuration at C-3 and C-6 have been controlled. The relative configuration at C-8 was controlled by anti-selective reduction of a 6-hydroxy-8-ketone using Evans’ method, and at C-2 by anti-selective enolate methylation of the β-silyl lactone 20. Silyl-to-hydroxy conversion with retention at C-3 and displacement of the tosylate with inversion at the same centre gave the correct relative and absolute configuration, completing a synthesis of methyl (+)-nonactate 4. The relative configuration at C-8 was controlled in the opposite sense by syn-selective reduction of a 6-hydroxy-8-ketone using Prasad’s conditions, and at C-2 in the opposite sense by anti-selective enolate methylation of the open-chain β-silyl ester 22. Silyl-to-hydroxy conversion with retention at C-3 and displacement of the tosylate with inversion at C-6 gave the correct relative and absolute configuration completing a synthesis of the pseudo-enantiomer, benzyl (–)-nonactate 25. Some protecting group changes and coupling of these two fragments gave the “dimers” 28 and 29, coupling of which gave the “tetramer” 30. Ring closure of this material using Yamaguchi’s method gave nonactin in 73% yield, substantially better than in any previous synthesis.