PNNL

Abstract

A combination of classical force field molecular dynamics and electronic structure calculations were used to identify the low-lying, gas phase conformations of isolated 15-crown-5 and complexes it forms with a single alkali cation (Li+, Na+, K+, Rb+, and Cs+). The computed binding enthalpies are compared with recent gas phase, collision-induced dissociation measurements. While agreement is reasonably good for sodium and potassium, it worsens for the larger cations, where theory and experiment differ by as much as 20 kcal/mol. Tests of the sensitivity of the predicted binding enthalpies to the size fo the basis sets were conducted with large, diffuse-function-augmented, correlation consistent basis sets. Despite the increase in size and flexibility of the basis sets, they resulted in relatively little change in the computed binding energies. A possible explatation of these differencies, first proposed for M+/12c4, is discussed. It assumes that the experiment measures dissociation from higher-lying metal/ether complexes, rather than the global minimum which serves as the reference point for the theoretical values. A cluster-based metal exchange reaction, which crudely models the binding preference in aqueous solution, reproduces the low selectivity of 15-crown-5 relative to that of 18-crown-6 in solution.