• DOI:10.3847/2041-8213/aa991c
• Corpus ID: 119342447

Constraining the Maximum Mass of Neutron Stars from Multi-messenger Observations of GW170817

@article{Margalit2017ConstrainingTM,
  title={Constraining the Maximum Mass of Neutron Stars from Multi-messenger Observations of GW170817},
  author={Ben Margalit and Brian D. Metzger},
  journal={The Astrophysical Journal},
  year={2017},
  volume={850}
}

• B. Margalit, B. Metzger
• Published 16 October 2017
• Physics
• The Astrophysical Journal

We combine electromagnetic (EM) and gravitational wave (GW) information on the binary neutron star (NS) merger GW170817 in order to constrain the radii $R_{\rm ns}$ and maximum mass $M_{\rm max}$ of NSs. GW170817 was followed by a range of EM counterparts, including a weak gamma-ray burst (GRB), kilonova (KN) emission from the radioactive decay of the merger ejecta, and X-ray/radio emission consistent with being the synchrotron afterglow of a more powerful off-axis jet. The type of compact remnant produced in the immediate merger aftermath, and its predicted EM signal, depend sensitively on the high-density NS equation of state (EOS). For a soft EOS which supports a low $M_{\rm max}$, the merger undergoes a prompt collapse accompanied by a small quantity of shock-heated or disk wind ejecta, inconsistent with the large quantity $\gtrsim 10^{-2}M_{\odot}$ of lanthanide-free ejecta inferred from the KN. On the other hand, if $M_{\rm max}$ is sufficiently large, then the merger product is a rapidly-rotating supramassive NS (SMNS), which must spin-down before collapsing into a black hole. A fraction of the enormous rotational energy necessarily released by the SMNS during this process is transferred to the ejecta, either into the GRB jet (energy $E_{\rm GRB}$) or the KN ejecta (energy $E_{\rm ej}$), also inconsistent with observations. By combining the total binary mass of GW170817 inferred from the GW signal with conservative upper limits on $E_{\rm GRB}$ and $E_{\rm ej}$ from EM observations, we constrain the likelihood probability of a wide-range of previously-allowed EOS. These two constraints delineate an allowed region of the $M_{\rm max}-R_{\rm ns}$ parameter space, which once marginalized over NS radius places an upper limit of $M_{\rm max} \lesssim 2.17M_{\odot}$ (90\%), which is tighter or arguably less model-dependent than other current constraints.