Supernovae Producing Unbound Binaries and Triples

TL;DR

This study investigates the fraction of stars in binaries or triples at the time of supernova, using Gaia data to identify unbound systems, and finds most stars are not in binaries when they explode.

Contribution

The paper provides new statistical constraints on the fractions of supernova progenitors in binaries or triples, based on Gaia data and analysis of supernova remnants.

Findings

72.0% of supernova progenitors are not in binaries at explosion

13.9% produce bound binaries after supernova

12.5% produce unbound binaries

Abstract

The fraction of stars which are in binaries or triples at the time of stellar death and the fraction of these systems which survive the supernova (SN) explosion are crucial constraints for evolution models and predictions for gravitational wave source populations. These fractions are also subject to direct observational determination. Here we search 10 supernova remnants (SNR) containing compact objects with proper motions for unbound binaries or triples using Gaia EDR3 and new statistical methods and tests for false positives. We confirm the one known example of an unbound binary, HD 37424 in G180.0-01.7, and find no other examples. Combining this with our previous searches for bound and unbound binaries, and assuming no bias in favor of finding interacting binaries, we find that 72.0% (52.2%-86.4%, 90% confidence) of SN producing neutron stars are not binaries at the time of explosion, 13.9% (5.4%-27.2%) produce bound binaries and 12.5% (2.8%-31.3%) produce unbound binaries. With a strong bias in favor of finding interacting binaries, the medians shift to 76.0% were not binaries at death, 9.5% leave bound and 13.2% leave unbound binaries. Of explosions that do not leave binaries, <18.9% can be fully unbound triples. These limits are conservatively for M>5Msun stars, although the mass limits for individual systems are significantly stronger. At birth, the progenitor of PSR J0538+2817 was probably a 13-19Msun star, and at the time of explosion it was probably a Roche limited, partially stripped star transferring mass to HD 37424 and then producing a Type IIL or IIb supernova.