An optimal envelope ejection efficiency for merging neutron stars

We use the rapid binary stellar evolution code BINARY_C to estimate the rate of merging neutron stars with numerous combinations of envelope ejection efficiency and natal kick dispersion. We find a peak in the local rate of merging neutron stars around $\alpha \approx 0.3$$-$$0.4$, depending on the metallicity, where $\alpha$ is the efficiency of utilising orbital energy to unbind the envelope. The peak height decreases with increasing electron-capture supernova kick dispersion $\sigma_\mathrm{ECSN}$. We explain the peak as a competition between the total number of systems that survive the common-envelope phase increasing with $\alpha$ and their separation, which increases with $\alpha$ as well. Increasing $\alpha$ reduces the fraction of systems that merge within a time shorter than the age of the Universe and results in different mass distributions for merging and non-merging double neutron stars. This offers a possible explanation for the discrepancy between the Galactic double neutron star mass distribution and the observed massive merging neutron star event GW190425. Within the $\alpha$$-$$\sigma_\mathrm{ECSN}$ parameter space that we investigate, the rate of merging neutron stars spans several orders of magnitude up to more than $1\times 10^{3} \, \mathrm{Gpc}^{-3}\,\mathrm{yr}^{-1}$ and can be higher than the observed upper limit or lower than the observed lower limit inferred thus far from merging neutron stars detected by gravitational waves. Our results stress the importance of common-envelope physics for the quantitative prediction and interpretation of merging binary neutron star events in this new age of gravitational wave astronomy.