Multiple regimes and coalescence timescales for massive black hole pairs ; the critical role of galaxy formation physics (1703.00661v1)

Abstract: We discuss the latest results of numerical simulations following the orbital decay of massive black hole pairs in galaxy mergers. We highlight important differences between gas-poor and gas-rich hosts, and between orbital evolution taking place at high redshift as opposed to low redshift. Two effects have a huge impact and are rather novel in the context of massive black hole binaries. The first is the increase in characteristic density of galactic nuclei of merger remnants as galaxies are more compact at high redshift due to the way dark halo collapse depends on redshift. This leads naturally to hardening timescales due to 3-body encounters that should decrease by two orders of magnitude up to $z=4$. It explains naturally the short binary coalescence timescale, $\sim 10$ Myr, found in novel cosmological simulations that follow binary evolution from galactic to milliparsec scales. The second one is the inhomogeneity of the interstellar medium in massive gas-rich disks at high redshift. In the latter star forming clumps 1-2 orders of magnitude more massive than local Giant Molecular Clouds (GMCs) can scatter massive black holes out of the disk plane via gravitational perturbations and direct encounters. This renders the character of orbital decay inherently stochastic, often increasing orbital decay timescales by as much as a Gyr. At low redshift a similar regime is present at scales of $1-10$ pc inside Circumnuclear Gas Disks (CNDs). In CNDs only massive black holes with masses below $10^7 M_{\odot}$ can be significantly perturbed. They decay to sub-pc separations in up to $\sim 10^8$ yr rather than the in just a few million years as in a smooth CND. Finally implications for building robust forecasts of LISA event rates are discussed