A magnetar origin for the kilonova ejecta in GW170817

TL;DR

This paper proposes that the blue kilonova ejecta in GW170817 originated from a neutrino-heated, magnetically-accelerated wind from a hypermassive neutron star remnant, explaining observed velocities, mass, and composition.

Contribution

It introduces a novel model attributing the blue kilonova ejecta to a magnetar wind from the HMNS, contrasting with previous dynamical ejecta explanations.

Findings

HMNS with magnetic field ~10^14 G can produce observed ejecta properties.

The inferred HMNS lifetime is close to its Alfven crossing time.

Shocks in the ejecta can homogenize composition and produce low lanthanide fraction.

Abstract

The neutron star (NS) merger GW170817 was followed over several days by optical-wavelength ("blue") kilonova (KN) emission likely powered by the radioactive decay of light r-process nuclei synthesized by ejecta with a low neutron abundance (electron fraction Ye ~ 0.25-0.35). While the composition and high velocities of the blue KN ejecta are consistent with shock-heated dynamical material, the large quantity is in tension with the results of numerical simulations. We propose an alternative ejecta source: the neutrino-heated, magnetically-accelerated wind from the strongly-magnetized hypermassive NS (HMNS) remnant. A rapidly-spinning HMNS with an ordered surface magnetic field of strength B ~ 1-3e14 G and lifetime t_rem ~ 0.1-1 s can simultaneously explain the velocity, total mass, and electron fraction of the blue KN ejecta. The inferred HMNS lifetime is close to its Alfven crossing time, suggesting global magnetic torques could be responsible for bringing the HMNS into solid body rotation and instigating its gravitational collapse. Different origins for the KN ejecta may be distinguished by their predictions for the emission in the first hours after the merger, when the luminosity is enhanced by heating from internal shocks; the latter are likely generic to any temporally-extended ejecta source (e.g. magnetar or accretion disk wind) and are not unique to the emergence of a relativistic jet. The same shocks could mix and homogenizes the composition to a low but non-zero lanthanide mass fraction, X_La ~ 1e-3, as advocated by some authors, but only if the mixing occurs after neutrons are consumed in the r-process on a timescale >~ 1 s.