What zeroth-order Hamiltonian for CASPT2 adiabatic energetics of Fe(II)N(6) architectures?

Complete active space self-consistent field (CASSCF) calculations and subsequent second-order perturbative treatment (CASPT2) have been carried out on a series of metallic complexes based on a Fe(II)N(6) core. These compounds play a determinant role in the elaboration of potential bistable architectures such as spin-crossover materials. The adiabatic energies between the high-spin (HS) (S=2) and low-spin (LS) (S=0) states are evaluated with respect to the value of the shift ionization potential-electronic affinity (IPEA shift) recently introduced in the zeroth-order Hamiltonian [Ghigo et al., Chem. Phys. Lett. 396, 142 (2004)]. Based upon a series of experimental data, it is concluded that the commonly applied IPEA shift value (0.25 a.u.) is not satisfactory to properly discriminate the open-shell HS and closed-shell LS states. We suggest that a 0.50-0.70 a.u. value would be preferable for these specific adiabatic gap calculations.