Searching for Hidden Descriptors in the Metal–Ligand Bond through Statistical Analysis of Density Functional Theory (DFT) Results

Abstract

A statistical treatment of the DFT-computed heterolytic bond dissociation energies (BDE) between a diverse variety of metal fragments and ligands leads to the identification of five hidden descriptors that best characterize the bonding ability per moiety, and of a simple mathematical formula able to obtain from these hidden descriptors a BDE estimation within a few kcal/mol from the DFT value. A simple extension of this treatment beyond the original set of metal fragments and ligands is also presented. The first two hidden descriptors can be associated with the well-known concepts of σ-donation and π-effects, with the next two associated with cis influence and degree of covalency. The procedure can be easily extended to additional ligands and metal fragments, and it opens the way to an improved understanding of fundamental concepts of chemical bonding.