No abstract

No abstract

Although a major goal of inorganic spectroscopy is to determine the energetics of the low-lying spin states of transition metal complexes, surprisingly little has been accomplished in this respect by means of accurate ab initio calculations. Against this context, we present ab initio multiconfiguration reference perturbation theory (CASPT2) calculations with large basis sets on the low-lying spin states of Fe(III)(P)Cl and [Fe(P)Cl](+) (P(2-)=porphinato). The CASPT2 results on the energetics of various low-lying spin states studied differ significantly, sometimes even dramatically, from those obtained from density functional theory calculations.

The equilibrium structure and bond energies of the transition metal complexes Ni(CO)x (x=1–4), Fe(CO)5, and Cr(CO)6 have been studied using the complete active space (CAS)SCF method and second-order perturbation theory (CASPT2). It is shown that the major features of the electronic structure are properly described by a CASSCF wave function based on an active space comprising the bonding and antibonding orbitals directly involved in the metal–ligand bond. Remaining correlation effects are dealt with in the second, CASPT2, step. The computed energies have been corrected for basis set superposition errors (BSSE) and relativistic corrections have been added. Resulting bond distances and bond energies are in agreement with experimental data, when available: Cr(CO)6, r(Cr–C)=1.91(1.91) Å, D0=148(153) kcal/mol; Fe(CO)5, rax(Fe–C) =1.79(1.81) Å, req(Fe–C)=1.80(1.83) Å, D0=133(137) kcal/mol; Ni(CO)4, r(Ni–C)=1.83(1.83) Å, D0=139(138) kcal/mol (experimental values within parentheses). Some excited states were computed for Fe(CO)5. The first charge transfer (CT) state was located at 4.8 eV in agreement with an intense band found at 5.0 eV.