The study of the magnetic coupling in binuclear complexes has been successfully performed using the Difference Dedicated Configuration Interaction (DDCI) method which allows, besides the calculation of the magnetic coupling constant (J), for an exhaustive analysis of the factors governing the magnetic coupling. The main difficulty with DDCI lies in its computational cost which prevents its application to large systems (large ligands, many magnetic centers or many electrons per magnetic center). Therefore, a less expensive computational strategy is highly desirable and Multi Reference Perturbation Theory (MRPT), applied to second order for the energy, is a natural candidate. This approach (with different partitions of the Hamiltonian) has been applied in the past and problems have been found. In particular, the MRPT2 approach only accounts for a fraction (40%-70%) of the effect of single and double excitations on the minimal CAS (2 electrons in 2 orbitals for two S_z=1/2 centers). In a first attempt to improve the quality of MRPT2, a strategy has been proposed to overcome the problem of the internal contraction (i.e. the fact that the ionic/neutral ratio is fixed at the CASCI level and not modified under the effect of the electronic correlation). Practical applications have shown that this approach, although improving the results, is far from being satisfactory. By using the paradigmatic binuclear Cu(d9) cases, in this talk we discuss the relevance of the starting orbitals in MRPT2 calculations, demonstrating that they play a key role for a correct calculation of J at this level. A simple (and computational inexpensive) strategy is proposed for the calculation of a set of magnetic orbitals almost identical to the (costly) natural DDCI orbitals.
Magnetic coupling in binuclear systems: the role of the magnetic orbitals for the calculation of J with multireference perturbation theory
ANGELI, Celestino
2011
Abstract
The study of the magnetic coupling in binuclear complexes has been successfully performed using the Difference Dedicated Configuration Interaction (DDCI) method which allows, besides the calculation of the magnetic coupling constant (J), for an exhaustive analysis of the factors governing the magnetic coupling. The main difficulty with DDCI lies in its computational cost which prevents its application to large systems (large ligands, many magnetic centers or many electrons per magnetic center). Therefore, a less expensive computational strategy is highly desirable and Multi Reference Perturbation Theory (MRPT), applied to second order for the energy, is a natural candidate. This approach (with different partitions of the Hamiltonian) has been applied in the past and problems have been found. In particular, the MRPT2 approach only accounts for a fraction (40%-70%) of the effect of single and double excitations on the minimal CAS (2 electrons in 2 orbitals for two S_z=1/2 centers). In a first attempt to improve the quality of MRPT2, a strategy has been proposed to overcome the problem of the internal contraction (i.e. the fact that the ionic/neutral ratio is fixed at the CASCI level and not modified under the effect of the electronic correlation). Practical applications have shown that this approach, although improving the results, is far from being satisfactory. By using the paradigmatic binuclear Cu(d9) cases, in this talk we discuss the relevance of the starting orbitals in MRPT2 calculations, demonstrating that they play a key role for a correct calculation of J at this level. A simple (and computational inexpensive) strategy is proposed for the calculation of a set of magnetic orbitals almost identical to the (costly) natural DDCI orbitals.I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
