Feeding the black hole with condensing accretion flows: radiatively efficient and radiatively inefficient cases

We study the accretion flow of a hot gas captured by the black hole gravity in the presence of a thin cold accretion disk. Such geometrical arrangement is expected in Active Galactic Nuclei (AGN) and in galactic X-ray binary systems because both hot and cold gases are present in the black hole vicinity. Previous astrophysical literature concentrated on the evaporation of the cold disk in the classical heat conduction limit. Here we consider the inverse process, i.e. condensation of the hot gas onto the cold disk. We find two distinct condensation regimes. (i) In the classical thermal conduction limit, the radiative cooling in the hot gas itself force condensation above a certain critical accretion rate. Most of the flow energy in this case is re-emitted as X-ray radiation. (ii) Below a certain minimum accretion rate, the hot electrons are collisionless and the classical heat flux description becomes invalid. We use the ``non-local'' heat flux approach borrowed from the terrestrial laser heated plasma experiments. Due to their very large mean free path, the hot particles penetrate deep into the cold disk where the radiative losses are significant enough to enable condensation. In this case the hot flow energy is inconspicuously re-radiated by the transition layer in many UV and especially optical recombination lines (e.g., Ly$\alpha$, $H\alpha$, H$\beta$) as well as via the optically thick disk emission. We derive an approximate analytical solution for the dynamics of the hot condensing flow. If the cold disk is inactive, i.e. accumulating mass for a future accretion outburst, then the two-phase flows appear radiatively inefficient. These condensing solutions may be relevant to \sgra, low luminosity AGN, and transient binary accreting systems in quiescence.