We have obtained new solutions and methods for the process of thermal Comptonization. We modify the solution to the kinetic equation of Sunyaev & Titarchuk to allow its application up to mildly relativistic electron temperatures and optical depths > ∼ 1. The solution can be used for spectral fitting of X-ray spectra from astrophysical sources. We also have developed an accurate Monte Carlo method for calculating spectra and timing properties of thermal-Comptonization sources. The accuracy of our kinetic-equation solution is verified by comparison with the Monte Carlo results. We also compare our results with those of other publicly available methods. Furthermore, based on our Monte Carlo code, we present distributions of the photon emission times and the evolution of the average photon energy for both up and downscattering.

We have recently improved our model of active galactic nucleus (AGN) by attaching the supermassive black hole (SMBH) to a massive nuclear star cluster (NSC). Here we study the effects of this new model in massive, gas-rich galaxies with several simulations of different feedback recipes with the hydrodynamics code RAMSES. These simulations are compared to a reference simulation without any feedback, in which the cooling halo gas is quickly consumed in a burst of star formation. In the presence of strong supernovae (SN) feedback, we observe the formation of a galactic fountain that regulates star formation over a longer period, but without halting it. If only AGN feedback is considered, as soon as the SMBH reaches a critical mass, strong outflows of hot gas are launched and prevent the cooling halo gas from reaching the disk, thus efficiently halting star formation, leading to the so-called "quenching". If both feedback mechanisms act in tandem, we observe a non-linear coupling, in the sense that the dense gas in the supernovae-powered galactic fountain is propelled by the hot outflow powered by the AGN at much larger radii than without AGN. We argue that these particular outflows are able to unbind dense gas from the galactic halo, thanks to the combined effect of SN and AGN feedback. We speculate that this mechanism occurs at the end of the fast growing phase of SMBH, and is at the origin of the dense molecular outflows observed in many massive high-redshift galaxies.

We introduce a new model for the formation and evolution of supermassive black holes (SMBHs) in the code using sink particles, improving over previous work the treatment of gas accretion and dynamical evolution. This new model is tested against a suite of highresolution simulations of an isolated, gas-rich, cooling halo. We study the effect of various feedback models on the SMBH growth and its dynamics within the galaxy. In runs without any feedback, the SMBH is trapped within a massive bulge and is therefore able to grow quickly, but only if the seed mass is chosen larger than the minimum Jeans mass resolved by the simulation. We demonstrate that, in the absence of supernovae (SN) feedback, the maximum SMBH mass is reached when Active Galactic Nucleus (AGN) heating balances gas cooling in the nuclear region. When our efficient SN feedback is included, it completely prevents bulge formation, so that massive gas clumps can perturb the SMBH orbit, and reduce the accretion rate significantly. To overcome this issue, we propose an observationally motivated model for the joint evolution of the SMBH and a parent nuclear star cluster (NSC), which allows the SMBH to remain in the nuclear region, grow fast and resist external perturbations. In this scenario, however, SN feedback controls the gas supply and the maximum SMBH mass now depends on the balance between AGN heating and gravity. We conclude that SMBH/NSC co-evolution is crucial for the growth of SMBH in high-z galaxies, the progenitors of massive ellipticals today.

Stars in globular clusters (GCs) lose a non-negligible amount of mass during their post-main-sequence evolution. This material is then expected to build up a substantial intracluster medium (ICM) within the GC. However, the observed gas content in GCs is a couple of orders of magnitude below these expectations. Here, we follow the evolution of this stellar wind material through hydrodynamical simulations to attempt to reconcile theoretical predictions with observations. We test different mechanisms proposed in the literature to clear out the gas such as ram-pressure stripping by the motion of the GC in the Galactic halo medium and ionization by UV sources. We use the code ramses to run 3D hydrodynamical simulations to study for the first time, the ICM evolution within discretized multimass GC models including stellar winds and full radiative transfer. We find that the inclusion of both ram pressure and ionization is mandatory to explain why only a very low amount of ionized gas is observed in the core of GCs. The same mechanisms operating in ancient GCs that clear the gas could also be efficient at younger ages, meaning that young GCs would not be able to retain gas and form multiple generations of stars as assumed in many models to explain ‘multiple populations’. However, this rapid clearing of gas is consistent with observations of young massive clusters.