2022MNRAS.512.6150S

The nebular spectra of Type Ia supernovae (100 d after explosion) consist mainly of emission lines from singly and doubly ionized Fe-group nuclei. However, theoretical models for many scenarios predict that non-thermal ionization leads to multiply ionized species whose recombination photons ionize and deplete Fe⁺, resulting in negligible [Fe II] emission. We investigate a method to determine the collisional excitation conditions from [Fe II] line ratios independently from the ionization state and find that it cannot be applied to highly ionized models due to the influence of recombination cascades on Fe⁺ level populations. When the ionization state is artificially lowered, the line ratios (and excitation conditions) are too similar to distinguish between explosion scenarios. We investigate changes to the treatment of non-thermal energy deposition as a way to reconcile overionized theoretical models with observations and find that a simple work function approximation provides closer agreement with the data for sub-M_ch models than a detailed Spencer-Fano treatment with widely used cross-section data. To quantify the magnitude of additional heating processes that would be required to sufficiently reduce ionization from fast leptons, we artificially boost the rate of energy loss to free electrons. We find that the equivalent of as much as an eight times increase to the plasma loss rate would be needed to reconcile the sub-M_ch model with observed spectra. Future studies could distinguish between reductions in the non-thermal ionization rates and increased recombination rates, such as by clumping.