Effects of density on the oxygen ionisation equilibrium in collisional plasmas

TL;DR

This study improves the modeling of oxygen ionization in collisional plasmas by incorporating metastable level effects and density-dependent processes, leading to more accurate plasma diagnostics and line intensity predictions.

Contribution

It introduces new electron impact ionization cross sections for oxygen, integrates them into collisional radiative models, and demonstrates enhanced accuracy in ionization equilibrium predictions at higher densities.

Findings

15-40% improved agreement with observed line intensities

Ionization equilibrium shifts with density and metastable population

Enhanced modeling accuracy for solar transition region diagnostics

Abstract

The ion populations most frequently adopted for diagnostics in collisional plasmas are derived from the density independent, coronal approximation. In higher density, lower temperature conditions, ionisation rates are enhanced once metastable levels become populated, and recombination rates are suppressed if ions recombine into Rydberg levels. As a result, the formation temperatures of ions shift, altering the diagnostics of the plasma. To accurately model the effect of ionisation from metastable levels, new electron impact, ionisation cross sections have been calculated for oxygen, both for direct ionisation and excitation--auto-ionisation of the ground and metastable levels. The results have been incorporated into collisional radiative modelling to show how the ionisation equilibrium of oxygen changes once metastable levels become populated. Suppression of dielectronic recombination has been estimated and also included in the modelling, demonstrating the shifts with density in comparison to the coronal approximation. The final results for the ionisation equilibrium are used in differential emission measure modelling to predict line intensities for many lines emitted by O II-VI in the solar transition region. The predictions show improved agreement by 15-40% for O II, O VI and the inter-combination lines of O III-V, when compared to results from coronal approximation modelling. While there are still discrepancies with observations of these lines, this could, to a large part, be explained by variability in the observations.