Not just winds: why models find binary black hole formation is metallicity dependent, while binary neutron star formation is not

Both detailed and rapid population studies alike predict that binary black hole (BHBH) formation is orders of magnitude more efficient at low metallicity than high metallicity, while binary neutron star (NSNS) formation remains mostly flat with metallicity, and black hole-neutron star (BHNS) mergers show intermediate behavior. This finding is a key input to employ double compact objects as tracers of low-metallicity star formation, as spectral sirens, and for merger rate calculations. Yet, the literature offers various (sometimes contradicting) explanations for these trends. We investigate the dominant cause for the metallicity dependence of double compact object formation. We find that the BHBH formation efficiency at low metallicity is set by initial condition distributions, and conventional simulations suggest that about \textit{one in eight interacting binary systems} with sufficient mass to form black holes will lead to a merging BHBH. We further find that the significance of metallicities in double compact object formation is a question of formation channel. The stable mass transfer and chemically homogeneous evolution channels mainly diminish at high metallicities due to changes in stellar radii, while the common envelope channel is primarily impacted by the combined effects of stellar winds and mass-scaled natal kicks. Outdated giant wind prescriptions exacerbate the latter effect, suggesting BHBH formation may be much less metallicity dependent than previously assumed. NSNS formation efficiency remains metallicity independent as they form exclusively through the common envelope channel, with natal kicks that are assumed uncorrelated with mass. Forthcoming GW observations will provide valuable constraints on these findings.