Determination of pairing matrix elements from average single particle level densities

Abstract

A simple and efficient method to treat nuclear pairing correlations within a simple Hartree-Fock--plus-BCS description is proposed and discussed. It relies on the fact that the intensity of pairing correlations depends crucially on level densities around the Fermi surface ($ ρ(e_F)$) and that any fitting of nuclear energies as functions of the nucleon numbers is akin of a semi-classical average, smoothing out their quantal structure. A particular attention has been paid to two points generally ignored in previous similar approaches. One is a correction advocated by Möller and Nix [Nucl. Phys. A 536, 20 (1992)] taking into account the fact that the data included into the fit correspond to $ ρ(e_F)$ values systematically lower than average. The second is due to a systematic overestimation of the proton sp level density at the Fermi surface resulting from the local Slater approximation of the Coulomb terms in use in most microscopic descriptions. Our approach is validated by the agreement with data of corresponding calculated moments of inertia of well and rigidly deformed rare-earth nuclei, evaluated according to the Inglis-Belyaev ansatz with some crude Thouless-Valatin corrections. Indeed, the agreement which is found, is at least of the same quality as what results from a specific fit of the pairing intensities to these particular pieces of data. While this approach is currently limited to the very simple seniority force pairing treatment, it may serve as a starting point to define pairing residual interactions from averaged odd-even mass differences data, using merely average sp level densities associated to calculated canonical basis.