An evaluation of the density functional approach in the zero order regular approximation for relativistic effects: magnetic interactions in small metal compounds

The performance of the density functional approach in the relativistic zero order regular approximation for the evaluation of electron spin resonance(ESR) parameters in small metal compounds has been evaluated critically by comparison with experimental data and available theoretical results for 22 linear molecules, characterized by a 2Σ electronic ground state. For most of the molecules studied the calculated magnetic parameters are in good (Atensors) or reasonable (gtensors) agreement with experiment. Effects of spin-orbit coupling and spin polarization on the calculated hyperfine interaction are investigated. These two effects can only be evaluated separately, since the present method does not allow us to take spin-polarization effects into account in spin-orbit coupled density functional calculations. However, while spin-polarization effects are important for all the molecules investigated, spin-orbit effects are non-negligible only for the molecules containing heavier metal atoms. The ESR parameters, evaluated using different “standard” exchange-correlation potentials, have only shown little dependence on the specific functional. Direct relativistic contributions to the hyperfine parameters are often large, especially for the heavier metals, but also “secondary” contributions to the ligand hyperfine parameters can be large if the ligand is bound to a heavy element.