Chapter 6: Computational Tools for Nitroxide Design

Computational tools for modelling nitroxide radical chemistry are outlined and illustrated through a series of case studies. Different types of application raise different problems. Herein we outline and assess the computational methods available for modelling high-spin systems, predicting EPR hyperfine coupling constants, redox potentials, alkoxyamine dissociation equilibria and photochemical processes. While high-spin systems and excited states are best modelled with complex multireference methods incorporating static and dynamic correlation, there is a growing family of more economical approaches that can often achieve an acceptable level of accuracy. Redox potentials and alkoxyamine dissociation equilibria require only single reference methods, albeit at a high level of theory, but the accurate treatment of solvation remains an ongoing challenge. Modelling EPR requires a range of specialist techniques and basis sets to accurately capture core correlation. Through the case studies presented here we not only show that accurate and useful quantitative predictions are possible, but introduce a wide range of complementary tools for enhancing our qualitative understanding of structure–reactivity trends.