Title:Dust masses for a large sample of core-collapse supernovae from optical emission line asymmetries: dust formation on 30-year timescales

Abstract:Modelling the red-blue asymmetries seen in the broad emission lines of core-collapse supernovae (CCSNe) is a powerful technique to quantify total dust mass formed in the ejecta at late times ($>5$ years after outburst) when ejecta dust temperatures become too low to be detected by mid-IR instruments. Following our success in using the Monte Carlo radiative transfer code DAMOCLES to measure the dust mass evolution in SN~1987A and other CCSNe, we present the most comprehensive sample of dust mass measurements yet made with DAMOCLES, for CCSNe aged between four and sixty years after outburst. Our sample comprises of multi-epoch late-time optical spectra taken with the Gemini GMOS and VLT X-Shooter spectrographs, supplemented by archival spectra. For the fourteen CCSNe that we have modelled, we confirm a dust mass growth with time that can be fit by a sigmoid curve which is found to saturate beyond an age of $\sim30$ years, at a mass of 0.23$^{+0.17}_{-0.12}$ M$_\odot$. An expanded sample including dust masses found in the literature for a further eleven CCSNe and six CCSN remnants, the dust mass at saturation is found to be 0.42$^{+0.09}_{-0.05}$~M$_\odot$. Uncertainty limits for our dust masses were determined from a Bayesian analysis using the affine invariant Markov Chain Monte Carlo ensemble sampler emcee with DAMOCLES. The best-fitting line profile models for our sample all required grain radii between 0.1 and 0.5 $\mu$m. Our results are consistent with CCSNe forming enough dust in their ejecta to significantly contribute to the dust budget of the Universe.