Modelling supernova line profile asymmetries to determine ejecta dust masses: SN 1987A from days 714 to 3604

TL;DR

This study models supernova line profiles to estimate dust masses, revealing significant dust formation in SN 1987A over years, with implications for understanding dust evolution in supernovae.

Contribution

Introduces DAMOCLES, a new code for modeling dust effects on supernova line profiles to determine ejecta dust masses, considering clumping and grain properties.

Findings

Dust masses <3 x 10^-3 M_sun at day 714 for most grain types

Large dust mass (>0.1 M_sun) formed by day 3604

Majority of dust formed years after explosion

Abstract

The late time optical and near-IR line profiles of many core-collapse supernovae exhibit a red-blue asymmetry as a result of greater extinction by internal dust of radiation emitted from the receding parts of the supernova ejecta. We present here a new code, DAMOCLES, that models the effects of dust on the line profiles of core-collapse supernovae in order to determine the masses of newly formed dust. As noted by Lucy et al. (1989), the presence of an extended red scattering wing in late-time line profiles can also indicate dust formation. We find that dust-affected line profiles need not necessarily be flux-biased towards to the blue, although the profile peak will always be blue-shifted. We have collated optical spectra of SN 1987A from a variety of archival sources and have modelled the evolution of the H$\alpha$ line from days 714 to 3604, as well as that of the [OI] 6300,6363A doublet between days 714 and 1478. A variety of evidence points to the presence of clumping and we find that our clumped dust models require significantly higher dust masses than smoothly distributed dust models. Our line profile fits imply day 714 dust masses of <3 $\times$ 10$^{-3}$ M$_{\odot}$ for all grain types apart from very high albedo pure magnesium silicates, for which up to 0.07M$_{\odot}$ can be accommodated. Large grain radii (>0.6$\mu$m) are generally required to fit the line profiles even at the earlier epochs. We find that a large dust mass (>0.1M$_{\odot}$) had formed by day 3604 and infer that the majority of the present dust mass must have formed after this epoch. Our findings agree with recent estimates from SED fits for the dust mass evolution of SN 1987A and support the inference that the majority of SN 1987A's dust formed many years after the initial explosion.