Computational and notably DFT studies are currently carried out for determining reaction mechanisms, resulting in a deeper understanding of reaction pathways. These studies have been increasingly successful in recent years even though difficulties (e.g. solvent effects, free energies) are known. In addition, a proper mechanistic study should include calculations of several competing pathways. It is sometimes not easy to draw definite conclusions on the preferred paths. Despite these difficulties, combining computations and experiment always leads to a better understanding of the chemistry.

We will illustrate these points by showing results on the computational study of bond energies and some consequences on cross coupling reactions. We will also present the results of some recent studies in the field of hydrogen transfer reactions.