Formation of Supermassive Black Hole Binaries and Massive Seed SMBHs in Gas-Rich Mergers

Abstract

We review the results of multi‐scale, hydrodynamical simulations of major mergers between galaxies with or without central supermassive black holes (SMBHs) to investigate the orbital decay of SMBH pairs in galactic nuclei and the formation of massive SMBH seeds via direct gas collapse. Both SPH simulations and AMR simulations are carried out. The complex balance between heating and cooling is modeled via an effective EOS with varying adiabatic index γ apparopriate for the conditions of an intense nuclear starburst such as that expected during and after the merger. Prominent gas inflows due to tidal torques produce nuclear disks at the centers of merger remnants whose properties depend sensitively on the details of gas thermodynamics. In parsec‐scale resolution simulations starting with two SMBHs originally at the centers of the two galaxies, a SMBH binary forms very rapidly, less than a million years after the merger of the two galaxies, owing to the drag exerted by the surrounding gaseous nuclear disk. Binary formation is significantly suppressed if heating, by e.g. radiative feedback from the accreting SMBHs, renders cooling negligible.The nuclear disk rearranges its mass distribution in response to a second, internal gas inflow occurring while the binary sinks. The inflow is driven by spiral instabilities imprinted by the final collision between the two galactic cores. In simulations with 0.1 pc resolution, the gas inflow continues all the way down to the center and peaks at >104 M⊙/yr, producing a Jeans‐unstable supermassive central cloud (with mass a few times 108 M⊙) only 105 yr after the merger. If the collapse continues, the cloud could form a massive black hole seed (Mseed>105 M⊙) after prior formation of a supermassive star or quasi‐star. The massive SMBH seed can grow up to a billion solar masses in less than a billion years by accreting the surrounding nuclear gas. If the gas‐rich merger occurs at z>8, this is then a new, attractive way to explain the rapid emergence of the bright QSOs discovered by the Sloan Digital Sky survey at z>6, which does not require the assumption of primordial gas composition in order to suppress cooling below 104 K and star formationas in models starting from unstable, isolated protogalactic disks. If there is a pre‐existing pair of SMBHs their orbital decay stalls at parsec scales because, as a result of the formation of the supermassive cloud, the nuclear disk density decreases outside the cloud, yielding much weaker dynamical friction. We envision a new scenario in which direct formation of massive black hole seeds and SMBH binary formation are mutually exclusive; if a SMBH is already present in the nuclear disk it can stabilize it and weaken the secondary inflow via its energetic feedback, maintaining a high enough density where the SMBHs are located and assisting their sinking.