A theoretical study of the Edward–Lemieux effect (the anomeric effect). The stereochemical requirements of adjacent electron pairs and polar bonds

Abstract

Many data, collected from the literature, support the postulation of two general rules which describe the static and dynamic results of having in a molecule or generating in an intermediate adjacent electron pairs and/or polar bonds. In disagreement with current chemical intuition, structures which contain the maximum number of gauche-interactions between lone electron pairs or polar bonds represent energy minima and, frequently, the lowest minima. The stereochemical implications of the phenomenon are discussed, along with possible physical explanations. It is suggested that partitioning of the total energy of the system into attractive-dominant and repulsive-dominant interactions provides the most helpful framework for the construction of a physical picture of the phenomenon. The total energy and its components can be obtained by ab initio molecular quantum mechanical calculations.

The only apparent exception to the phenomenon that has been found corresponds to a polar bond adjacent to two lone pairs (as in the case of two hetero-atoms attached to the same carbon atom). This exception, here termed the Edward–Lemieux effect, has been examined theoretically by an ab initio(Hartree–Fock) calculation using fluoromethanol as a model compound. The calculation has reproduced the Edward–Lemieux effect; the stable conformation has the C–F bond trans to one ‘electron pair’ and gauche to another, and the conformation in which the C–F bond bisects the ‘electron pairs’ is the energy maximum. The partitioning of the total energy into its components of attraction and repulsion and comparison of the results with other systems for which barriers to internal rotation have been obtained by ab initio methods reveals a similarity between fluoromethanol, hydrazine, hydroxylamine, and hydrogen peroxide. An interpretation of this result is provided in which an early suggestion by Lemieux and Chü is supported and the concept of ‘rabbit-ears’ is not.