Spin adapted versus broken symmetry approaches in the description of magnetic coupling in heterodinuclear complexes

Abstract

The performance of a series of wave function and density functional theory based methods in predicting the magnetic coupling constant of a family of heterodinuclear magnetic complexes has been studied. For the former, the accuracy is similar to other simple cases involving homodinuclear complexes, the main limitation being a sufficient inclusion of dynamical correlation effects. Nevertheless, these series of calculations provide an appropriate benchmark for density functional theory based methods. Here, the usual broken symmetry approach provides a convenient framework to predict the magnetic coupling constants but requires deriving the appropriate mapping. At variance with simple dinuclear complexes, spin projection based techniques cannot recover the corresponding (approximate) spin adapted solution. Present results also show that current implementation of spin flip techniques leads to unphysical results.