What Makes a Good (Computed) Energy Profile?

Abstract

A good meal cannot be defined in an absolute manner since it depends strongly on where and how it is eaten and how many people participate. A picnic shared by hikers after a challenging climbing is very different from a birthday party among a family or a banquet for a large convention. All of them can be memorable and also good. The same perspective applies to computational studies. Required level of calculations for spectroscopic properties of small molecular systems and properties of medium or large organic or organometallic, polymetallic systems are different. To well-specified chemical questions and chemical systems, efficient computational strategies can be established. In this chapter, the focus is on the energy profile representation of stoichiometric or catalytic reactions assisted by organometallic molecular entities. The multiple factors that can influence the quality of the calculations of the Gibbs energy profile and thus the mechanistic interpretation of reactions with molecular organometallic complexes are presented and illustrated by examples issued from mostly personal studies. The usual suspects to be discussed are known: representation of molecular models of increasing size, conformational and chemical complexity, methods and levels of calculations, successes and limitations of the density functional methods, thermodynamics corrections, spectator or actor role of the solvent, and static vs dynamics approaches. These well-identified points of concern are illustrated by presentation of computational studies of chemical reactions which are in direct connection with experimental data. Even if problems persist, this chapter aims at illustrating that one can reach a representation of the chemical reality that can be useful to address questions of present chemical interest. Computational chemistry is already well armed to bring meaningful energy information to numerous well-defined questions.