Core polarization and oscillator strength ratio anomaly in potassium, rubidium and caesium

Abstract

The problem of the anomalous ratios of the oscillator strengths in the heavy alkali principal series, though known for almost 70 years, still attracts the attention of both theorists and experimentalists. Our present fully relativistic model potential calculations demonstrate that the spin-orbit interaction cannot be solely responsible for the observed anomalies in potassium, rubidium and caesium and that they result from the interplay of spin-orbit interaction, core-valence electron correlation (core polarization) and cancellation observed in transition integrals. The role of core polarization is carefully studied. It is shown that in the case of the empirically adjusted model potential the most important correction is not the inclusion of polarization potential but the allowance made for the additional induced dipole moment in the dipole moment operator of transition. Also the form employed to represent the correction to the dipole moment is important, if one wishes to avoid any adjustment in core polarization corrections. It is also shown that the employment of J-dependent effective model potentials (J = total angular momentum) is essential in reproducing relativistic effects in f-value ratios.