Is the Hartree-Fock prediction that the chemical potential µ of non-relativistic neutral atoms is equal to minus the ionisation potential I sensitive to electron correlation?

Abstract

Numerical estimates of the chemical (µ) and ionisation (I) potentials from natural orbital functional (NOF) theory of one of us are presented for neutral atoms ranging from H to Kr, and compared with the corresponding experimental ionisation energies. The semi-empirical treatment by Vela et al. relates µ approximately to the measured I. By a physical argument based on the concept of the Pauli potential, we thereby derive in the fully correlated ground states of atoms. The predicted vertical I by means of the extended Koopmans' theorem is in good agreement with the corresponding experimental data. However, the NOF theory of µ lowers the experimental values considerably though oscillatory behaviour is again in evidence. This comparison has prompted us to reopen the Hartree-Fock prediction by Alonso and March that .