Black holes in galactic nuclei: seed formation from stellar mass black holes and massive black hole pairing in galaxy mergers.

Black holes (BHs) are a very important class of astrophysical objects. They are the most compact objects in the Universe, hence they represent the most extreme sources of gravity. BHs come in two flavours: the stellar mass BHs (SBHs) relic of young massive stars (1−20M⊙) and the massive BHs (MBHs), with masses of 106−109M⊙, dwelling in the nuclei of the most massive galaxies. While the formation mechanisms of SBHs are well understood, no clear consensus exists about MBH formation. According to the Soltan arguments (Soltan, 1982), MBHs gain the largest fraction of their mass via radiative efficient accretion of gas. As a consequence, we expect that MBH formed early in the Universe as smaller mass seeds. Recently, observations of high redshift quasars (e.g.; Mortlock et al., 2011; Fan et al., 2006) showed that MBHs with masses above 109M⊙ were already in place when the Universe was less than 1 Gyr old and posed tight constraints on the models for the formation and growth of MBHs. Two main scenarios have been developed for MBH seed formation: the light seed scenario, where seeds formed as relic of the first generation of stars with masses of up to few hundred solar masses (Madau & Rees, 2001), and the heavy seed scenarios, where seeds formed from the direct collapse of massive gas clouds in primordial haloes with masses of up to few 105M⊙ (Haehnelt & Rees, 1993). Despite the large number of studies about MBH formation models, each model still has its own caveats, which make the study of MBH formation worth of further investigations. According to the -CDM cosmology, galaxies form when gas cools down within dark matter haloes, which assembly in a hierarchical fashion from small density perturbations. Galaxies grow via accretion and mergers, and the central MBHs evolve in the same way. So, when a galaxy merger occur, the MBHs hosted in the nucleus of the galaxy progenitors can sink towards the centre of the merger remnant, forming a MBH binary (MBHB). Despite galaxy mergers are usually observed, no clear detections of MBHBs exist to date. The formation and evolution of MBHBs is a complex process, since it occurs in a rapidly varying environment where gas, star formation and SNa feedback play a pivotal role. Several studies have been performed to date, but a clear understanding of the whole process is still far from being reached. In this thesis I cover both aspects of MBH formation and evolution. In the first study I consider an alternative route for seed BH formation. Using two different codes, the AMR code RAMSES (Teyssier, 2002) and the meshfree code GIZMO (Hopkins, 2015), I studied the evolution of a single massive circum-nuclear gaseous disc embedding a population of SBHs. The disc was subject to radiative cooling, star formation and supernova feedback and becomes unstable to fragmentation, which led to the formation of clumps as massive as 104 − 105M⊙. My simulations showed that during the disc evolution, some SBHs can be gravitationally captured by a clump. Within the clumps, such BHs can experience episodes of super-critical accretion, which make them grow up to 103 − 104M⊙ in few Myr. Thanks to the very low radiative efficiency associated to the slim accretion disc (Abramowicz et al., 1988), the energy released to the surrounding gas is too small to halt the accretion flow, hence BHs can accrete almost unimpeded until one of these events occur: the clump is totally accreted by the BH, the clump is consumed by star formation or the clump is destroyed by supernova explosions. In the second study, instead, I consider the intermediate stages of a galaxy merger, when the MBHs originally dwelling in the centre of their own progenitor galaxies reach few hundred separations in the nucleus of the merger remnant. I assumed that each MBH was embedded in a self-gravitating circumnuclear gaseous disc. With the code RAMSES I studied the evolution of the MBHs and their surrounding discs, including physical processes like radiative cooling, star formation and supernova feedback, which are implemented in the code as sub-grid recipes. First, I implemented a new refinement prescription aimed at improving the orbital evolution of massive particles, an already known major issue in AMR codes, like observed by Gabor & Bournaud (2013); Dubois et al. (2014). Secondly, I evolved the discs assuming different sub-grid recipes to study how the MBH and gas dynamics could be affected by the different choices. I found that the MBH dynamics is almost independent of the physical modelling, if one assumes that no previous star formation occurred in the discs, while the gas evolution and its final distribution can be significantly affected. On the other side, if one assumes that star formation was already ongoing, even the BH dynamics can be modified, if supernovae are powerful enough to disrupt gas clumps forming in the discs. A general introduction to the work is reported in Chapter 1. In Chapter 3 I discuss the first study about an alternative model for seed BH formation. In Chapter 4, instead, I describe the second study concerning the evolution of the MBH pair in the intermediate stages of a galaxy merger. The reader interested in the main results of the work can directly move to Chapters 3 and 4. Finally, Chapter 5 reports my conclusions.