It is commonly believed that the optical/UV and X-ray emissions in luminous AGN are produced in an accretion disk and an embedded hot corona respectively. The inverse Compton scattering of disk photons by hot electrons in the corona can effectively c
ool the coronal gas if the mass supply is predominantly via a cool disk like flow as in BHXRBs. Thus, the application of such a model to AGNs fails to produce their observed X-ray emission. As a consequence, a fraction of disk accretion energy is usually assumed to be transferred to the corona. To avoid this assumption, we propose that gas in a vertically extended distribution is supplied to a supermassive black hole by the gravitational capture of interstellar medium or stellar wind material. In this picture, the gas partially condenses to an underlying cool disk as it flows toward the black hole, releasing accretion energy as X-ray emission and supplying mass for the disk accretion. Detailed numerical calculations reveal that the X-ray luminosity can reach a few tens of percent of the bolometric luminosity. The value of $alpha_{rm ox}$ varies from 0.9 to 1.2 for the mass supply rate ranging from 0.03 to 0.1 times the Eddington value. The corresponding photon index in the 2-10 keV energy band varies from 1.9 to 2.3. Such a picture provides a natural extension of the model for low luminosity AGN where condensation is absent at low mass accretion rates and no optically thick disk exists in the inner region.
The magnitude of the viscosity and magnetic field parameters in hot accretion flows is investigated in low luminosity active galactic nuclei (LLAGNs). Theoretical studies show that a geometrically thin, optically thick disk is truncated at mass accre
tion rates less than a critical value by mass evaporated vertically from the disk to the corona, with the truncated region replaced by an advection dominated accretion flow (ADAF). The critical accretion rate for such a truncation is a function of the viscosity and magnetic field. Observations of X-ray photon indices and spectral fits of a number of LLAGNs published in the literature provide an estimate of the critical rate of mass accretion and the truncation radius respectively. By comparing the observational results with theoretical predictions, the viscosity and magnetic field parameters in the hot accretion flow region are estimated. Specifically, the mass accretion rates inferred in different sources constrain the viscosity parameter, whereas the truncation radii of the disk, as inferred from spectral fits, further constrain the magnetic field parameter. It is found that the value of the viscosity parameter in the corona/ADAF ranges from 0.17 to 0.5, with values clustered about 0.2-0.3. Magnetic pressure is required by the relatively small truncation radii for some LLAGNs and is found to be as high as its equipartition value with the gas pressure. The inferred values of the viscosity parameter are in agreement with those obtained from the observations of non-stationary accretion in stellar mass black hole X-ray transients. This consistency provides support for the paradigm that a geometrically thin disk is truncated by means of a mass evaporation process from the disk to the corona at low mass accretion rates.
The truncation of an optically thick, geometrically thin accretion disk is investigated in the context of low luminosity AGN (LLAGN). We generalize the disk evaporation model used in the interpretative framework of black hole X-ray binaries by includ
ing the effect of a magnetic field in accretion disks surrounding supermassive black holes. The critical transition mass accretion rate for which the disk is truncated is found to be insensitive to magnetic effects, but its inclusion leads to a smaller truncation radius in comparison to a model without its consideration. That is, a thin viscous disk is truncated for LLAGN at an Eddington ratio less than 0.03 for a standard viscosity parameter ($alpha = 0.3$). An increase of the viscosity parameter results in a higher critical transition mass accretion rate and a correspondingly smaller truncation distance, the latter accentuated by greater magnetic energy densities in the disk. Based on these results, the truncation radii inferred from spectral fits of LLAGN published in the literature are consistent with the disk evaporation model. The infrared emission arising from the truncated geometrically thin accretion disks may be responsible for the red bump seen in such LLAGN.
The disk corona evaporation model extensively developed for the interpretation of observational features of black hole X-ray binaries (BHXRBs) is applied to AGNs. Since the evaporation of gas in the disk can lead to its truncation for accretion rates
less than a maximal evaporation rate, the model can naturally account for the soft spectrum in high luminosity AGNs and the hard spectrum in low luminosity AGNs. The existence of two different luminosity levels describing transitions from the soft to hard state and from the hard to soft state in BHXRBs, when applied to AGNs, suggests that AGNs can be in either spectral state within a range of luminosities. For example, at a viscosity parameter, alpha, equal to 0.3, the Eddington ratio from the hard to soft transition and from the soft to hard transition occurs at 0.027 and 0.005 respectively. When the Eddington ratio of the AGN lies below the critical value corresponding to its evolutionary state, the disk is truncated. With decreasing Eddington ratios, the inner edge of the disk increases to greater distances from the black hole with a concomitant increase in the inner radius of the broad line region, $R_{BLR}$. The absence of an optically thick inner disk at low luminosities gives rise to region in the size of borad line-luminosity plane for which the relation $R_{BLR} propto L^{1/2}$ inferred at high luminosities is excluded. As a result, a lower limit to the accretion rate is predicted for the observability of broad emission lines, if the broad line region is associated with an optically thick accretion disk. Thus, true Seyfert 2 galaxies may exist at very low accretion rates/luminosities. The differences between BHXRBs and AGNs in the framework of the disk corona model are discussed and possible modifications to the model are briefly suggested.
