A theoretical study of the bonding in trivalent americium complexes

An account of the electronic structure of AmF3 and AmCl3 has been obtained by Complete Active Space calculations (CASSCF and CASPT2) using either a scalar Douglas–Kroll–Hess Hamiltonian or a relativistic effective core potential. It has been compared to the description obtained by density functional theory (DFT) methods combined with a quasi-relativistic ZORA approach, either in the scalar or two-component approach. Optimised geometries of the ground and first excited states (7A1 and 7A2 in C3v symmetry) given by both methodologies are similar, which let us conclude that monoconfigurational DFT methods are reliable tools at least for the geometrical and bonding analysis of Am(III) complexes.