NLTE spectral modelling for a carbon-oxygen and helium white-dwarf merger as a Ca-rich transient candidate

TL;DR

This work tests whether a CO WD + He WD merger can reproduce Ca-rich transient signatures by performing full NLTE, non-thermal radiative transfer on a 1D ejecta model derived from a 3D merger simulation. Using the ARTIS code, it tracks photospheric and nebular phases with detailed energy deposition and expanded atomic data, revealing a nebular spectrum dominated by [Ca II] and Caii emissions and a peak optical spectrum with strong Hei and Ca II features. The results support CO+He WD mergers as a plausible Ca-rich transient channel but reveal tensions: optical Hei lines and Ti II induce a red SED and [Ca II] line widths are broader than observed, indicating limitations of the 1D approach. The study underscores the need for 3D NLTE radiative transfer and broader binary configurations to capture observed diversity and viewing-angle effects in Ca-rich transients.

Abstract

We carry out NLTE (non local thermodynamic equilibrium) radiative transfer simulations to determine whether explosion during the merger of a carbon-oxygen (CO) white dwarf (WD) with a helium (He) WD can reproduce the characteristic Ca II/[Ca II] and He I lines observed in Ca-rich transients. Our study is based on a 1D representation of a hydrodynamic simulation of a 0.6 $M_{\odot}$ CO + 0.4 $M_{\odot}$ He WD merger. We calculate both photospheric and nebular-phase spectra including treatment for non-thermal electrons, as is required for accurate modelling of He I and [Ca II]. Consistent with Ca-rich transients, our simulation predicts a nebular spectrum dominated by emission from [Ca II] 7291, 7324 angstrom and the Ca II near-infrared (NIR) triplet. The photospheric-phase synthetic spectrum also exhibits a strong Ca II NIR triplet, prominent optical absorption due to He I 5876 angstrom and He I 10830 angstrom in the NIR, as is commonly observed for Ca-rich transients. Overall, our results therefore suggest that CO+He WD mergers are a promising channel for Ca-rich transients. However, the current simulation overpredicts some He I features, in particular both He I 6678 and 7065 angstrom and shows a significant contribution from Ti II, which results in a spectral energy distribution that is substantially redder than most Ca-rich transients at peak. Additionally the Ca II nebular emission features are too broad. Future work should investigate if these discrepancies can be resolved by considering full 3D models and exploring a range of CO+He WD binary configurations.

Figures (5)

• Figure 1: Bottom panel: model ejecta composition at 160 s after explosion for key species in the simulation. Top panel: energy deposition profiles for the three epochs we show spectra for. Overplotted for reference are the density profiles of He and key radioisotopes that we track the decay chain energy deposition for in the simulation.
• Figure 2: Optical spectroscopic evolution of our model from peak until the early nebular phase. The last species with which packets interacted before they left the simulation are indicated with different colours. Contributions to emission are plotted on the positive axis with contributions to absorption plotted on the negative axis. For reference, key lines are labelled along with their rest wavelength given in Å. The dashed lines indicate the wavelengths of absorptions from the prominent optical Hei lines.
• Figure 3: Same as Figure \ref{['fig:optical_emission_absorption']} for NIR spectroscopic evolution of the model.
• Figure 4: Spectroscopic comparisons around peak between our simulated spectra along with the observed Ca-rich transients SN 2012hn valenti2014a, SN 2019ehk jacobson-galan2020anakaoka2021a, SN 2010et kasliwal2012a, SN PTF11kmb lunnan2017a and SN iPTF16hgs de2018a. The model epochs are relative to the Sloan r-band peak and have been chosen to match the epochs of each observed spectrum relative to either Sloan r-band or Bessel R-band peak. For reference the strong optical Hei features, attributed to the Hei 5876, 6678 and 7065 Å lines, are highlighted in cyan and the Caii NIR triplet is highlighted in pink.
• Figure 5: Nebular phase spectroscopic comparisons between our simulated spectra at 45 d post r-band peak compared to the observed Ca-rich transients SN 2021gno jacobson-galan2022aertini2023a, SN 2022oqm yadavalli2024a, SN 2019ehk jacobson-galan2020anakaoka2021a and SN PTF12bho lunnan2017a at similar epochs. The epochs of the observations are relative to Sloan r-band or Bessel R-band peak. For reference, the grey dashed line is plotted between the rest wavelengths of the key [Caii] 7291, 7324 Å lines, the Fei emission feature predicted by the simulation is highlighted in grey and the location of the [Oi] 6300, 6364 Å feature commonly observed for Ca-rich transients is highlighted in green.