Speaker
Description
Comprehensive datasets of electron-impact excitation (EIE) collision strengths for near-neutral lanthanides and other r-process elements remain notably incomplete, despite their central role in setting the level populations of neutron-star merger ejecta during the late phase (non local thermodynamic equilibrium, more commonly known as non-LTE, regime), of kilonova emission [1, 2]. A long-recognised difficulty in this context is the treatment of resonant excitation (RE), the indirect channel in which electron capture populates an intermediate autoionising state that subsequently stabilises radiatively onto a bound excited level. At the nebular electron temperatures characteristic in kilonovae (2000 to 5000 K), the density of low-lying autoionising resonances in complex open-shell ions is such that RE can exceed the direct distorted-wave (DW) contribution over a broad range of transitions and temperatures [3, 4]. The computational expense of accounting for these channels, together with the structural complexity of open-f-shell systems, has nevertheless driven their usual omission from astrophysical applications.
In this work we report systematic calculations of EIE collision strengths for a large set of singly and doubly ionised lanthanides and other astrophysically important r-process ions, performed within the fully relativistic distorted-wave method implemented in the Flexible Atomic Code (FAC) [5]. Resonant contributions are incorporated through the Independent Process, Isolated Resonance Distorted-Wave (IPIRDW) approximation [6], with atomic structures constructed from tailored central potentials [7] and calibrated level energies [8]. Where recent R-Matrix results are available, we find that including RE systematically reduces the discrepancy between the DW and close-coupling effective collision strengths across the full temperature range of astrophysical interest [9].
The resulting datasets quantify, across a broad range of r-process ions, the magnitude of the resonant contribution that has been absent from virtually all non-LTE kilonova modelling to date. These data are ready for direct use in radiative transfer simulations of both lanthanide-rich ejecta and second r-process peak emission features in late-time kilonova spectra.
References
[1] Pognan et al. MNRAS 510, 3806 (2022)
[2] Gillanders et al. MNRAS 529, 2918 (2024)
[3] McCann et al. MNRAS 538, 537 (2025)
[4] Mulholland et al. JQSRT 345, 109545 (2025)
[5] Gu, Can. J. Phys. 86, 675 (2008)
[6] Li et al. Chin. Phys. B 24, 113401 (2015)
[7] Ferreira da Silva et al. Phys. Rev. A 112, 012802 (2025)