Electronic and steric factors determining the asymmetric epoxidation of allylic alcohols by titanium-tartrate complexes (the Sharpless epoxidation)

Abstract

The structure and epoxidation properties of titanium-tartrate asymmetric epoxidation catalysts have been studied by using the frontier orbital approach. It is suggested that an important factor determining the dimeric structure of these titanium-tartrate epoxidation catalysts is electronic, as one of the LUMOs, located at the titanium atom, is oriented so it facilitates nucleophilic trans coordination to the titanium of the carbonyl group in the tartrate. Coordination of a peroxide to titanium-tartrate is analyzed. From the frontier orbitals at the equatorial peroxygen, four possible orientations of the allylic alcohol are possible. Analysis of the preferred orientation of a hydroxyl and a methoxy group at the equatorial site at the titanium atom, and of an alkene around the peroxygen, led to a spiro orientation of the alkene part of the allylic alcohol at the peroxygen as the most probable. The preferred orientation of the allylic alcohol at the titanium atom is discussed in relation to electronic as well as steric interactions with the tartrate. The orientation and reactivity of the alkene part of the allylic alcohol can be traced to two two-electron interactions: one is the peroxygen lone pair electron interaction with the T* orbital of the alkene part of the allylic alcohol and the other is the interaction of the titanium-peroxygen antibonding orbital with the K orbital of the alkene. Epoxidation of alkenes, and especially asymmetric epoxidation, is a fundamental and important organic reaction type. The asymmetric epoxidation was pioneered by Herbst, who used chiral monoperoxycamphoric acid to produce chiral epoxides with an enantiomeric excess of 5% or less;' later Pirkle and Rinaldi were able to improve the enantiomeric excess to 9%.* Catalytic epoxidation with hydrogen peroxide and tert-butyl hydrogen peroxide catalyzed by chiral phase-transfer agents such as benzyl guinidinium salts was investigated by Hummelen and Wynberg, with moderate ~ u c c e s s . ~ The first transition metal catalyzed asymmetric epoxidation was reported independently by Sharpless et aL4 and Yamada et al.,5 the first with vanadium and the second molybdenum complexes. Further examples of molybdenumcatalyzed epoxidations were reported by Otsuka et a1.,6 who described the treatment of squalene with a mixture of tert-butyl hydroperoxide, chelated molybdenum oxide, and the chiral inducing agent (+)-diisopropyl tartrate. By this method it was possible to obtain (S)-2,3-squalene epoxide in 14% enantiomeric excess. The real breakthrough in asymmetric epoxidation came when Sharpless et al. treated a mixture of commercially available titanium tetraisoproxide, tert-butyl hydroperoxide, and (+)or (-)-diethyl tartrate with allylic alcohol^.^ With (-)-diethyl tartrate the oxidant approaches the allylic alcohol from the top side of the plane shown in 1, whereas the bottom side is open for the (+)-diethyl tartrate reagent, giving rise to the corresponding optically active epoxy alcohols, 2. This asymmetric epoxidation, now known as the Sharpless epoxidation, has already shown its power in the synthesis of natural products.8 ' Aarhus University. 1 Baker Laboratory. o ( ) D i e t h y l t a r t r a t e