GRMHD Simulations of Neutron-star Mergers with Weak Interactions: r-process Nucleosynthesis and Electromagnetic Signatures of Dynamical Ejecta

Abstract

Abstract Fast neutron-rich material ejected dynamically over ≲10 ms during the merger of a binary neutron star (BNS) can give rise to distinctive electromagnetic counterparts to the system’s gravitational-wave emission that serve as a “smoking gun” to distinguish between a BNS and an NS–black hole merger. We present novel ab initio modeling of the kilonova precursor and kilonova afterglow based on 3D general-relativistic magnetohydrodynamic simulations of BNS mergers with nuclear, tabulated, finite-temperature equations of state (EOSs), weak interactions, and approximate neutrino transport. We analyze dynamical mass ejection from 1.35–1.35 M ⊙ binaries, consistent with properties of the first observed BNS merger GW170817, using three nuclear EOSs that span the range of allowed compactness of 1.35 M ⊙-neutron stars. Nuclear reaction network calculations yield a robust second-to-third-peak r-process. We find few ×10−6 M ⊙ of fast (v > 0.6c) ejecta that give rise to broadband synchrotron emission on ∼years timescales, consistent with tentative evidence for excess X-ray/radio emission following GW170817. We find ≈2 × 10−5 M ⊙ of free neutrons that power a kilonova precursor on ≲ hours timescale. A boost in early UV/optical brightness by a factor of a few due to previously neglected relativistic effects, with enhancements up to ≲10 hr post-merger, is promising for future detection with UV/optical telescopes like Swift or ULTRASAT. We find that a recently predicted opacity boost due to highly ionized lanthanides at ≳70,000 K is unlikely to affect the early kilonova based on the obtained ejecta structures. Azimuthal inhomogeneities in dynamical ejecta composition for soft EOSs found here (“lanthanide/actinide pockets”) may have observable consequences for both early kilonova and late-time nebular emission.