Lanthanide Impact on the Infra-Red Spectra of Nebular Phase Kilonovae

TL;DR

This work evaluates how lanthanide elements influence the nebular-phase infrared spectra of kilonovae using NLTE radiative transfer on a grid of ejecta models drawn from neutron star merger simulations. The results show lanthanide effects are strongest at higher densities and primarily affect wavelengths below about $4\,\mu$m, with Te III at $2.1\,\mu$m and Se III at $4.5$ and $5.7\,\mu$m acting as key spectral signatures; MIR emission is dominated by non-lanthanide lines, and the study argues that line opacity alone cannot reproduce the smooth continuum seen in AT2023vfi, suggesting optically thin IR at these epochs. The findings highlight Te III, Nd II, Ce III, Se III, and Ni III as major contributors in the IR, provide guidance for interpreting JWST observations, and indicate that NIR data are most diagnostic of lanthanide content while MIR spectra probe non-lanthanide emission. The work also emphasizes the need for improved atomic data and diverse ejecta models to robustly interpret KN spectra across epochs.

Abstract

Nebular phase kilonovae (KNe) have significant infra-red (IR) emission thought to be mostly forbidden emission lines from rapid neutron capture (r-process) species in neutron star merger ejecta. Lanthanide elements in particular have complex atomic structures with many IR transitions. Using non-local thermodynamic equilibrium (NLTE) radiative transfer simulations, we explore the impact of lanthanides on the IR spectra of KNe in the nebular phase, exploring a parameter space of ejecta mass and lanthanide fraction. We find that lanthanide impact is greater at higher densities, corresponding to earlier epochs and greater ejecta masses. The wavelengths most affected are found to be $λ\lesssim 4~μ$m, with the species Ce\,\textsc{iii} and Nd \textsc{ii} being the most important contributors to spectral formation. We also find significant emission from species proposed in observations, notably Te\,\textsc{iii} at 2.1 $μ$m, and Se\,\textsc{iii} at 4.5 and 5.7 $μ$m, while W\,\textsc{iii} is subdominant at 4.5 $μ$m. The Te\,\textsc{iii} feature at 2.1 $μ$m is always blended, particularly with Zr\,\textsc{ii}, Ce\,\textsc{iii}, and Nd\,\textsc{ii}. We do not reproduce the smooth blackbody-like continua observed in AT2023vfi. Based on our results, we argue that line opacity alone is likely insufficient to produce optically thick continua in the nebular phase, even in the case of lanthanide/actinide-rich ejecta, as our models are optically thin in the IR at these epochs. Given that lanthanide contributions are dominant below 4 $μ$m, we suggest that NIR observations best probe these elements, while MIR spectroscopy with \textit{JWST} can reliably probe non-lanthanide emission even in relatively lanthanide-rich cases.

Figures (15)

• Figure 1: Model compositions (top panel) and associated energy depositions (bottom panel) including thermalisation efficiency, where the solid lines have $M_{\rm{ej}} = 0.01M_{\odot}$, the dashed lines have $M_{\rm{ej}} = 0.05M_{\odot}$, and the dashed-dotted lines have $M_{\rm{ej}} = 0.005M_{\odot}$. In the top panel, the points show the elements included in the model compositions and their mass-fractions. Note that the mass-fractions of chosen elements are slightly increased due to renormalisation such that the sum of mass-fractions is always unity.
• Figure 2: The evolution of temperature (left panels) and ionisation degree (electron fraction $x_e$, right panels) by zone of the $f_{\rm{dyn}} = 0.05$ model, with masses $M_{\rm{ej}} = 0.005, 0.01, 0.05 \, M_{\odot}$ in the top, middle and bottom rows respectively
• Figure 3: Spectra of the $f_{\rm{dyn}} = 0.05$, $M_{\rm{ej}} = 0.01\, M_{\odot}$ model from 10 to 75 days post-merger. Elemental compositions are shaded. Note that the contributions are summed (i.e. stacked), and that the black-coloured lanthanide contribution excludes lanthanide species already present otherwise.
• Figure 4: Fraction of Te iii flux in the range of 1.9 -- 2.3 $\mu$m.
• Figure 5: Spectra of the $f_{\rm{dyn}} = 0.05$ model at select epochs with $M_{\rm{ej}} = 0.05\, M_{\odot}$ in the left-hand panels, and $M_{\rm{ej}} = 0.005\, M_{\odot}$ in the right-hand panels.