The role of neutron star mergers in the chemical evolution of the Galactic halo

Abstract

Context. The dominant astrophysical production site of the r-process elements has not yet been unambiguously identified. The suggested main r-process sites are core-collapse supernovae and merging neutron stars. Aims. We explore the problem of the production site of Eu. We also use the information present in the observed spread in the Eu abundances in the early Galaxy, and not only its average trend. Moreover, we extend our investigations to other heavy elements (Ba, Sr, Rb, Zr) to provide additional constraints on our results. Methods. We adopt a stochastic chemical evolution model that takes inhomogeneous mixing into account. The adopted yields of Eu from merging neutron stars and from core-collapse supernovae are those that are able to explain the average [Eu/Fe]-[Fe/H] trend observed for solar neighbourhood stars, the solar abundance of Eu, and the present-day abundance gradient of Eu along the Galactic disc in the framework of a well-tested homogeneous model for the chemical evolution of the Milky Way. Rb, Sr, Zr, and Ba are produced by both the s-and r-processes. The r-process yields were obtained by scaling the Eu yields described above according to the abundance ratios observed in r-process rich stars. The s-process contribution by spinstars is the same as in our previous papers. Results. Neutron star binaries that merge in less than 10 Myr or neutron star mergers combined with a source of r-process generated by massive stars can explain the spread of [Eu/Fe] in the Galactic halo. The combination of r-process production by neutron star mergers and s-process production by spinstars is able to reproduce the available observational data for Sr, Zr, and Ba. We also show the first predictions for Rb in the Galactic halo. Conclusions. We confirm previous results that either neutron star mergers on a very short timescale or both neutron star mergers and at least a fraction of Type II supernovae have contributed to the synthesis of Eu in the Galaxy. The r-process production of Sr, Zr, and Ba by neutron star mergers -complemented by an s-process production by spinstars -provide results that are compatible with our previous findings based on other r-process sites. We critically discuss the weak and strong points of both neutron star merging and supernova scenarios for producing Eu and eventually suggest that the best solution is probably a mixed one in which both sources produce Eu. In fact, this scenario reproduces the scatter observed in all the studied elements better.

COMPLETE ELEMENT ABUNDANCES OF NINE STARS IN THE r-PROCESS GALAXY RETICULUM II*

“…This site is responsible for the bulk of r-process material in RetII. The existence of multiple r-process sites has been suggested several times before (e.g., Wasserburg et al 1996;Qian & Wasserburg 2007Tsujimoto & Shigeyama 2014b;Cescutti et al 2015;Wehmeyer et al 2015). However, the offset in Figure 5 between UFD stars and most halo stars suggests that the bulk of neutron-capture elements are not synthesized by the common but inefficient r-process site.…”

The Intermediate r-process in Core-collapse Supernovae Driven by the Magneto-rotational Instability

“…However, there exist several unsolved problems with considering NSMs as only r-process sources (e.g., Argast et al 2004). Wehmeyer et al (2015) and Cescutti et al (2015) explained the chemical evolution of r-process nuclei, based on multiple sources, including NSMs and CC-SNe/MR-SNe. Tsujimoto & Nishimura (2015) showed that MR-SNe can explain the early growth of Eu in dwarf spheroidal galaxies for [ ] < -Fe H 2 by assuming an event rate of about 0.5% of CC-SNe, while NSMs have problems with doing so.…”

Nucleosynthesis in Supernovae, Hypernovae/Gamma-ray Bursts and Compact Binary Mergers

“…Such approaches have recently been undertaken by a number of authors [2,9,22,64,75,79]. While [2,9,79] came to the conclusion that the contribution of neutron star mergers with respect to heavy r-process elements shows up too late, i.e only at metallicities around [Fe/H]=-2, different from observations which see them already for [Fe/H]<-3, [22,64,75] argue otherwise. The question is how the delayed binary evolution and explosion enters into the interstellar medium, after the related supernovae producing the two neutron stars already ejected ample amounts of Fe.…”

“…If e.g. one assumes that only the typical mass swept up by a Taylor-Sedov blast wave is polluted (of the order a few 10 −5 M ⊙ ), then the results of [2,9,79] are reproduced (but see also in the high resolution run of [75]). There are two ways to avoid such conclusions (and there is of course an uncertainty related to them): (1) permitting large scale mixing (like turbulent mixing as argued in [64,75]), or (2) proposing that our Galaxy results from a mixture of initially different galactic substructures with different star formation rates which can shift [Eu/Fe] as a function of [Fe/H] to smaller and larger values [22].…”

“…There are two ways to avoid such conclusions (and there is of course an uncertainty related to them): (1) permitting large scale mixing (like turbulent mixing as argued in [64,75]), or (2) proposing that our Galaxy results from a mixture of initially different galactic substructures with different star formation rates which can shift [Eu/Fe] as a function of [Fe/H] to smaller and larger values [22]. (1) is either an artifact of too large star particles in SPH codes, which automatically assume a well mixed abundance distribution in each one of such star particles, while smaller masses of star particles in high resolution runs of [75] actually reproduce the results of [2,9,79] or Nature actually provided such large-scale turbulent mixing, a feature which we still need to understand. There is no doubt that compact binary mergers are a highly important or the dominant r-process site.…”