Tungsten vs Selenium as a potential source of kilonova nebular emission observed by Spitzer

Infrared emission lines arising from transitions between fine structure levels of heavy elements are expected to produce kilonova nebular emission. For the kilonova in GW170817, strong emission at 4.5 ${\rm \mu m}$ at late times was detected by the Spitzer Space Telescope but no source was detected at 3.6 ${\rm \mu m}$. This peculiar spectrum indicates that there exist strong line emitters around 4.5 ${\rm \mu m}$ and the absence of strong lines around 3.6 ${\rm \mu m}$. To model the spectrum we prepare a line list based on the selection rules in LS coupling from the experimentally calibrated energy levels in the NIST database. This method enables to generate the synthetic spectra with accurate line wavelengths. We find that the spectrum is sensitive to the abundance pattern whether or not the first r-process peak elements are included. In both cases, the synthetic spectra can match the observed data, leading to two possible interpretations. If the first peak elements are abundant a Se III line dominates the flux. If otherwise, W III with Os III, Rh III, and Ce IV can be the main sources. Observing nebular spectra for the future kilonovae in a wider wavelength range can provide more conclusive elemental identification.
Comment: 5 pages, 3 figures, submitted to MNRAS