One of the major goals of chemistry is to control chemical reactions with the purpose of generating new compounds with useful properties. Control of a chemical reaction implies a detailed understanding of its mechanism as it results from the breaking and forming of chemical bonds. In practice, it is rather difficult to get a detailed mechanistic and dynamical description of even the simplest chemical reactions. This has to do with the fact that apart from reactants, products, and possible stable intermediates, all other molecular forms encountered during a reaction have such a short lifetime that standard experimental means are not sufficient to detect and describe them. Progress in modern laser spectroscopy seems to provide an access to transient species with lifetimes in the pico‐ to femtosecond region; however, computational investigations utilizing state‐of‐the art methods of quantum chemistry, in particular ab initio methods, provide still the major source of knowledge on reaction mechanism and reaction dynamics. The reaction path Hamiltonian model has proven as a powerful tool to derive the dynamics of a chemical reaction by following the reacting species along the reaction path from reactants to products as traced out on the potential energy surface. In this article, the original reaction path Hamiltonian will be reviewed, extensions and applications over the past decades will be summarized, and a new perspective, namely to use it in form of the unified reaction valley approach to derive a deep and systematic insight into the mechanism of a chemical reaction will be introduced. © 2011 John Wiley & Sons, Ltd. WIREs Comput Mol Sci 2011 1 531–556 DOI: 10.1002/wcms.65This article is categorized under:
Electronic Structure Theory > Ab Initio Electronic Structure Methods
