The origin, properties and fate of the first black holes in the Universe

Observations of the Universe’s earliest quasars, less than 1 Gyr after the Big Bang, open the door to many questions. They are found to host supermassive black holes (SMBHs) with MBH = 10^9 - 10^10 Msun (Fan et al., 2001, 2004; Mortlock et al., 2011), and BH formation models need to explain their existence and evolution in such a short time. In the first part of this original work, we introduce the cosmological, semi-analytic code GAMETE/SuperQSOdust, which reconstructs several hierarchical merger histories of high-z bright quasars, following the time evolution of central BHs together with the mass of stars, gas, metals and dust. With this tool, we have studied the relative importance of different accretion regimes for the formation of the first quasars, with particular attention to accretion events occurring over the classical luminosity threshold - the so-called Eddington limit. We find that 80 % of the final SMBH mass is grown by super-Eddington accretion, which can be sustained down to z ~ 10 in dense, gas-rich environments, and the average BH mass at z ~ 20 is MBH ~ 10^4 Msun, comparable to that of direct collapse BHs. However, stellar feedback from BH seed progenitors and winds from BH accretion disks may decrease BH accretion rates. Therefore, we studied the impact of these physical processes on the formation of z ~ 6 quasars, including new physical prescriptions in the model. We find that the feedback produced by the first stellar progenitors on the surrounding environment does not play a relevant role in preventing the SMBH formation. In order to grow the z 6 SMBHs, the accreted gas must efficiently lose angular momentum. Moreover, disk winds, easily originated in the super-Eddington accretion regime, can strongly reduce duty cycles, producing a decrease in the active fraction among the progenitors of z ~ 6 bright quasars and thus reducing the probability to detect them. From an observational point of view, no convincing candidates of faint progenitors of luminous high-z quasars have been selected in X-ray surveys (Treister et al., 2013; Weigel et al., 2015; Cappelluti et al., 2016). In order to interpret this lack of detections, we have modeled the spectral energy distribution of accreting BHs. This modeling has been applied to a sample of simulated z ~ 6 SMBH progenitors, also taking into account the photon trapping effect which plays an important role at high accretion rates. The results show that faint progenitors are still luminous enough to be detected with current X-ray surveys. Even accounting for a maximum obscuration effect, the number of detectable BHs is reduced at most by a factor of 2. In our simulated sample, observations of faint BHs are mainly limited by their very low active fraction (fact ~ 1 percent), which is the result of short, supercritical growth episodes. We suggest that to detect high-z SMBH ancestors, large area surveys with shallower sensitivities, such as COSMOS Legacy and XMM-LSS+XXL, should be preferred with respect to deep surveys probing smaller fields, such as Chandra Deep Field South. An alternative way of constraining the early growth of BHs is to compare theoretical models with observations of massive BHs (MBH ~ 10^5 Msun) in local dwarf galaxies. To this aim, in the last part of this work, we introduced GAMESH, a simulation following the formation of a Milky Way-like halo in a well resolved cosmic volume of (4 cMpc)^3. This model allows to follow the star formation and chemical enrichment histories of all the galaxies in the simulation box. In the near future, we plan to extend the model including a self-consistent evolution of BHs and their feedback onto the host galaxies. This will allow us to compare results obtained by different BH seeding and accretion models with observations of BH masses hosted by the Milky Way and dwarf galaxies. Here, we present a preliminary study, where we have post-processed the simulation output to analyze the mass and redshift distribution of BH seeds formed as remnants of Pop III stars, and the BH occupation fraction at z = 0. Our preliminary results have been obtained under the assumption that gas accretion gives a negligible contribution to BH mass growth and, hence, provide a lower limit to the mass of nuclear BHs found at z = 0. We compare our results with recent studies carried out by means of cosmological hydrodynamical simulations (Marinacci et al. 2014; Bonoli et al. 2016), and - given the quiescent history experienced by the Milky Way-like halo - we conclude that either (i) light BH remnants of Pop III stars are able to rapidly grow their masses soon after their formation, or (ii) that the MilkyWay nuclear BH originates from more massive BH seeds, with masses comparable to the ones that characterize direct collapse BHs. In our future study, we will be able to analyze each of these two possibilities using the detailed treatment of chemical and radiative feedback effects allowed by GAMESH. This thesis is divided into four main parts. In the first part, we introduce some basic theoretical tools for understanding the most important features of the formation of galaxies and black holes: in Chapter 1, we present the CDM Cosmological Model and some fundamental properties of our Universe, including Large Scale Structures and galaxy formation, and in Chapter 2 we briefly describe the main characteristics of black holes and gas accretion disks orbiting around these compact objects. The second part of this work is dedicated to the high-z BHs: Chapter 3 is an extract from the review Valiante R., Agarwal B., Habouzit M., Pezzulli E., 2017, PASA 34, 31. In Chapter 4 we discuss the results obtained in the manuscript Pezzulli E., Valiante R., Schneider R., 2016, MNRAS, 458, 3047, and we also introduce the cosmological, semi-analytic model used for the study on the occurrence of different accretion regimes for the formation of high-z QSOs. In Chapter 5 we discuss on the sustainability of super-Eddington accretion in a cosmological context, including some prescriptions for the two negative feedback mechanisms introduced above. The results have been published in Pezzulli E., Volonteri M., Schneider R., Valiante R., 2017, MNRAS, 471, 589. Possible solutions for the current lack of faint, high-z AGNs observations are reported in Chapter 6, as investigated in Pezzulli E., Valiante R., Orofino M.C., Schneider R., Gallerani S., Sbarrato T., 2017, MNRAS, 466, 2131. In the third part of the thesis, we turn our attention to the Local Universe and to the constraints that can be put on the evolution of nuclear BHs and their hosts from observations of the Milky Way and local dwarf galaxies. In Chapter 7 we present the results of our preliminary study on the mass and redshift distribution of BH seeds and their impact on the z = 0 BH occupation fraction. Finally, in Part IV, we summarize our main conclusions.