2025ApJ...980..181C

Binary neutron stars (BNSs) detected in the Milky Way have total masses distributing narrowly around ∼2.6-2.7M_⊙, while the BNS merger GW190425 detected via a gravitational wave has a significantly larger mass (∼3.4M_⊙). This difference is not well understood, yet. In this paper, we investigate the BNS spin evolution via an improved binary star evolution model and its effects on the BNS observability, with the implementation of various relevant astrophysical processes. We find that the first-born neutron star component in low-mass BNSs can be spun up to millisecond pulsars by the accretion of Roche-lobe overflow from its companion and its radio lifetime can be comparable to the Hubble time. However, most high-mass BNSs have substantially shorter radio lifetimes than low-mass BNSs, and thus a smaller probability of being detected via radio emission. Adopting the star formation and metal enrichment history of the Milky Way given by observations, we obtain the survived Galactic BNSs with pulsar components from our population synthesis model and find that their distributions on the diagrams of the spin period versus the spin period time derivative ($P-\dot{P}$) and the orbital period versus the eccentricity (P_orb-e) can well match those of the observed Galactic BNSs. The total mass distribution of the observed Galactic BNSs can also be matched by the model. A significant fraction (∼19%-22%) of merging BNSs at redshift z ∼ 0 have masses ≳3M_⊙, which seems compatible with the GW observations. Future radio observations may detect many more Galactic BNSs, which will put strong constraints on the spin evolution of BNSs during their formation processes.