Two-component Relativistic Effective Core Potential Calculations for Molecules

Abstract

Spin-orbit and other relativistic effects important for the reliable description of valence states of atoms and molecules can be represented as two-component relativistic effective core potentials (RECPs) derived from Dirac-Coulomb Hamiltonian based all-electron calculations of atoms. Using spin-orbit RECPs, we have implemented and tested a series of two-component methods for molecular electronic structure calculations starting from a two-component Kramers restricted Hartree-Fock (KRHF) method for the polyatomic molecules with closed-shell configurations. The KRHF method utilizes RECPs with effective one-electron spin-orbit operators at the Hartree-Fock level in a variational manner and produces molecular spinors obeying the double group symmetry. Electron correlations are treated at various levels including the coupled-cluster level of theory with and without spin-orbit interactions. Spin-orbit effects on many molecules containing sixth-row p-block elements (T1∼Rn), transactinide d-block elements (Rf, Db, and Sg), and p-block elements (element 113∼118) are evaluated and discussed. In the present work, the spin-orbit effect is defined by the difference between the results of one- and two-component RECPs. The one-component RECP, which is derived by a potential average scheme, is found to be useful for describing spin-free molecular properties even for the transactinide molecules. The potential average scheme is proposed as the consistent definition of spin-orbit effects in any RECP scheme when spin-orbit effects are compared among various methods.