Accurate determinations of the extent to which the SE2' reactions of allyl-, allenyl- and propargylsilanes are stereospecifically anti.

The allylsilanes, (R)-E- and (R)-Z-4-trimethylsilylpent-2-ene 16, were prepared in essentially an enantiomerically and geometrically pure state (er >99.95 : 0.05, E : Z and Z : E >99.95 : 0.05) by, successively, conjugate addition of lithium dimethylcuprate to N-[(E)-3'-trimethylsilylpropenoyl]-(7S)-10,10-dimethyl-4-aza-5-thiatricyclo[5.2.1.0(3,7)]decane 5,5-dioxide 13, to give N-[(E)-(3'R)-3'-trimethylsilylbutanoyl]-(7S)-10,10-dimethyl-4-aza-5-thiatricyclo[5.2.1.0(3,7)]decane 5,5-dioxide, removal of the chiral auxiliary with bromomagnesium benzyloxide, aldol reaction with acetaldehyde, and decarboxylative elimination, to give either the Z- or E-isomer. Both the E- and Z-allylsilanes 16 reacted with the adamantyl cation to give mixtures of E- and Z-4-adamantylpent-2-enes 17. The E-allylsilane gave the E- and Z-products in a ratio of 40 : 60, and the Z-allylsilane gave the E- and Z-products in a ratio of 99.8:0.02. The enantiomer ratio was >99:1 for the reaction of the E-allylsilane giving the Z-product, 90:10 for the E-allylsilane giving the E-product, and 95 : 5 for the Z-allylsilane giving the E-product, showing that the reactions were stereospecific to a high degree, but not always quite completely so. The allenylsilane, 2-trimethylsilylpenta-2,3-diene 29, was prepared enantiomerically highly enriched (er 99:1) by copper-catalysed reaction of methylmagnesium chloride with (S)-4-trimethylsilylbut-3-yn-2-yl camphor-10-sulfonate 28. The allenylsilane 29 reacted with the adamantyl cation to give (S)-4-adamantylpent-2-yne (S)-30 with the same level of enantiomeric purity, showing that the reaction was, as accurately as can be measured, completely stereospecific. The allenylsilane 29 also reacted with isobutanal in the presence of titanium tetrachloride to give 2,4-dimethylhept-5-yn-3-ol as a mixture of diastereoisomers, syn 31 and anti 32, in a ratio of 95:5, with the major diastereoisomer present as a mixture of enantiomers (4R,5R):(4S,5S) in a ratio of 99:1, showing that the reaction was, as accurately as can be measured, completely stereospecific in the anti sense. The corresponding propargylsilane, 4-trimethylsilylpent-2-yne 37, reacted with the adamantyl cation to give dienes assigned the structures 2,3-diadamantyl-1,3-pentadiene 42 and 2,4-diadamantyl-1,3-pentadiene 43, and reacted with isobutanal in the presence of titanium tetrachloride to give 2-(1-hydroxy-2-methylpropyl)-3-trimethylsilylpenta-1,3-dienes 45 and 2,4-dimethyl-5-trimethylsilylhept-5-en-3-one 46. The enantiomerically enriched propargylsilane (R)-1,3-bis(trimethylsilyl)but-1-yne (er >99.7:0.3) was prepared from the sultam 13, by removal of the chiral auxiliary with lithium ethoxide, reduction of the ethyl ester to give (R)-3-trimethylsilylbutanal 60, enol triflate formation, beta-elimination and C-silylation. The propargylsilane reacted with 2,4-dinitrobenzaldehyde in the presence of titanium tetrachloride to give the allenes, 1-(2,4-dinitrophenyl)-2-trimethylsilylpenta-2,3-dienols 63-66, as two diastereoisomers in a ratio of 2 : 1, each of which was a pair of enantiomers in a ratio of approximately 3:1, showing that there was considerable loss of stereospecificity, but that what there was was in the anti sense. A similar reaction with isobutanal gave a similar set of four allenes, 2-methyl-4-trimethylsilylhepta-4,5-dien-3-ol 73-76, but with a negligible degree of stereospecificity.