The development of bimetallic aluminium-salen complexes [{Al(salen)}(2)O] as catalysts for the synthesis of cyclic carbonates (including the commercially important ethylene and propylene carbonates) from a wide range of terminal epoxides in the presence of tetrabutylammonium bromide as a cocatalyst is reported. The bimetallic structure of one complex was confirmed by X-ray crystallography. The bimetallic complexes displayed exceptionally high catalytic activity and in the presence of tetrabutylammonium bromide could catalyse cyclic carbonate synthesis at atmospheric pressure and room temperature. Catalyst-reuse experiments demonstrated that one bimetallic complex was stable for over 60 reactions, though the tetrabutylammonium bromide decomposed in situ by a retro-Menschutkin reaction to form tributylamine and had to be regularly replaced. The mild reaction conditions allowed a full analysis of the reaction kinetics to be carried out and this showed that the reaction was first order in aluminium complex concentration, first order in epoxide concentration, first order in carbon dioxide concentration (except when used in excess) and unexpectedly second order in tetrabutylammonium bromide concentration. Further kinetic experiments demonstrated that the tributylamine formed in situ was involved in the catalysis and that addition of butyl bromide to reconvert the tributylamine into tetrabutylammonium bromide resulted in inhibition of the reaction. The reaction kinetics also indicated that no kinetic resolution of racemic epoxides was possible with this class of catalysts, even when the catalyst was derived from a chiral salen ligand. However, it was shown that if enantiomerically pure styrene oxide was used as substrate, then enantiomerically pure styrene carbonate was formed. On the basis of the kinetic and other experimental data, a catalytic cycle that explains why the bimetallic complexes display such high catalytic activity has been developed.
