r-Process enrichment in the Galactic halo characterized by nucleosynthesis variation in the ejecta of coalescing neutron star binaries

TL;DR

This paper models the variation in r-process element abundances in the Galactic halo by considering different ejecta types from neutron star mergers, explaining observed abundance patterns and estimating event rates consistent with gravitational-wave data.

Contribution

It introduces a model linking nucleosynthesis in various ejecta from neutron star mergers to observed r-process element variations in halo stars, aligning with gravitational-wave event rates.

Findings

Reproduces [Eu/Fe] variation using ejecta mass differences.

Explains [Y/Eu] trend through combined nucleosynthesis in ejecta.

Estimates neutron star merger rates consistent with gravitational-wave observations.

Abstract

A large star-to-star variation in the abundances of r-process elements, as seen in the [Eu/Fe] ratio for Galactic halo stars, is a prominent feature that is distinguishable from other heavy elements. It is, in part, caused by the presence of highly r-process enriched stars, classified as r-II stars ([Eu/Fe]>+1). In parallel, halo stars show that the ratio of a light r-process element (Y) to Eu is tightly correlated with [Eu/Fe], giving the lowest [Y/Eu] ratio that levels off at r-II stars. On the other hand, recent hydrodynamical simulations of coalescing double neutron stars (cNSNSs) have suggested that r-process sites may be separated into two classes providing different electron-fraction distributions: tidally-driven dynamical ejecta and (dynamical or postmerger) non-tidal ejecta. Here, we show that a widely spanning feature of [Eu/Fe] can be reproduced by models that consider the different masses of tidally-driven dynamical ejecta from both cNSNSs and coalescing black hole/neutron star binaries (cBHNSs). In addition, the observed [Y/Eu] trend is explained by the combined nucleosynthesis in two kinds of ejecta with varying mass asymmetry in double NS systems. Our scenario suggests that massive tidally-driven dynamical ejecta accompanied by massive non-tidal part from cNSNSs or cBHNSs could alone accommodate r-II abundances, including an actinide boost in some cases. The event rate for cNSNSs estimated from our study agrees with the latest result of ~1000 (90% confidence interval of 110-3840) Gpc$^{-3}$yr$^{-1}$ by gravitational-wave detection, and a few events per Gpc$^3$ per year of cBHNSs associated with r-process production are predicted to emerge.