Mihail Atanasov: Ab initio Ligand Field Theory

Ligand field theory (LFT) was originally designed as a model for a given class of compounds -transition metal complexes with open d-shells in not too low and not too high formal oxidation states II and III (Werner type complexes).¹ Since then, LFT has been a tool for the interpretation/rationalization rather than a tool for prediction of magnetic and spectroscopic properties long before ab initio and DFT methods became broadly used. In this talk I will present the so called ab initio ligand field theory (AILFT)^2-4 –a computational protocol, allowing one to extract ligand field parameters from correlated wave functions in a much more sophisticated way than spectroscopic methods (even in high resolution) usually do allow. Ligand field parameters deduced in such a way turned out to be useful in three different directions: i) they offer the unique possibility to interpret numerical results from correlated calculation using a language familiar to chemists – one electron ligand field energy spitting diagrams and effective atomic like parameters of inter electronic repulsion and spin-orbit coupling modified in a complex; ii) such parameters derived using correlated ab initio Complete Active Space Self Consistent Field (CASSCF) wave functions and N-Electrons Valence Perturbation Theory to second order (NEVPT2) can be used in a predictive way to allow rational design of spectroscopically and magnetic interesting molecules with transition metals, lanthanides and actinides prior to their synthesis and experiment. iii) From a comparison between ligand field parameters derived separately, from experiment and by ab initio quantum chemistry, improvement of correlated electronic structure methods can be done in a systematic and well controlled way. Illustrative applications will cover classical 3dⁿ-metal, lanthanide (4fⁿ) and actinide (5fⁿ) complexes.