Zinc enolates: C- or O-metallation?

Abstract

Two methods have been used for the generation of zinc enolates: the reaction of EtZnOMe with enol acetates, and that of lithium enolates with zinc chloride. Most of the zinc compounds prepared proved to be very reactive towards carbonyl functions, and so they cannot be isolated from the EtZnOMe/enol acetate system. The final products of these reactions are polymerisation and self-condensation products and β-diketonates, the latter being formed by condensation reactions of the zinc enolates with an acetate molecule. The structure of [EtZnOMe·Zn(Pac) 2] 2 (HPac = pivaloylacetone, (CH 3) 3CCOCH 2COCH 3), isolated in 20% yield from the reaction of EtZnOMe with CH 3COOC(t-Bu)CH 2, was determined by X-ray diffraction analysis. The compound forms monoclinic crystals, space group P2 1/ c, with two dimers in a cell of dimensions a 11.677(4), b 18.299(9) and c 12.719(5) Å and β 117.26(3)°. The structure closely resembles that of the known complex [PhZnOPh·Zn(Pac) 2] 2. The complications involving reactions of zinc enolates with enol acetates were avoided by treating lithium enolates with zinc chloride. Polymerization and self-condensation could be prevented by using the very bulky enolate LiOC(t-Bu)CMe 2. In this way, the corresponding stable zinc enolate RZnCl·THF was obtained as a dissociating dimer. No replacement of the second chlorine atom by an enolate group occurred even when a large excess of lithium enolate was used. The reactivity of the zinc enolates suggests that they contain both zinccarbon and zincoxygen bonds. They are assumed to have a cyclic structure which resembles that of the Reformatsky reagent.