Accretion powers the strong X-ray emission from compact objects in binary systems. The same process is thought to be at work in Active Galactic Nuclei (AGNs), where the mass of the compact object is, however, much higher (M~10^6 - 10^9 Mo). In many of these systems there are evidences that the accretion process is mediated by an accretion disk. This geometry however is not unique: mass transfer is dominated by the radial component of the velocity when the angular momentum is not sufficiently high. Moreover, the motion of the accreting matter can be perturbated by the presence of a magnetosphere (as happens in many binary systems containing an X-ray pulsar), channeling the matter onto the magnetic poles.
Accretion disk are thought to be present in most of the galactic binary systems containing a low mass companion (LMXRBs) and in some with a high mass companion; in AGNs there are only indirect evidences in favour of the presence of a disk.
The conditions for a spherically symmetric accretion process to take place till the magnetospheric radius or even till the compact object, can be satisfied by weakly or not magnetized object in binary systems, capturing the stellar wind of a binary companion, or by isolated objects, accreting matter from the interstellar medium or from the densest regions of the Galaxy, such as molecular clouds.
In this Ph. D. thesis we have studied some problems about accretion (spherically symmetric as well as mediated by a disk) onto compact objects. In the following we describe the main results.
To calculate the line profile produced by a Keplerian disk orbiting a Schwarzschild black hole, we have used a model which accounts for all the relativistic effects. In particular, we have developed two limiting geometries describing the central source: a point-like one and an extended (spherical) one. This second is more likely in describing the strong X-ray reflected component, observed in many AGNs, which can not be originated from a point-like central source surrounded by a geometrically thin disk, due to the too small radiation intercepted by the disk.
With these hypothesis we have described the response of the disk to a rapid variation of the luminosity at the central source, describing the temporal evolution of the line intensity, of the line centroid energy and of the width of the line. Moreover, we have constructed the first 2D-transfer functions (which describe the response of the system to an arbitrary source variation) for a relativistic accretion disk. Besides the theoretical importance of this analysis, we have pointed out the possibility to constrain disk parameters and to measure the mass of the central object by studying the temporal evolution of the line centroid energy and the line width with X-ray instruments of the next generation.
With the available data, we have studied the X-ray continuum flux of the rather unique Seyfert galaxy NGC 6814. This source was observed by X-ray satellite Ginga to decrease by a factor of two in 50 s. The intensity of the iron K alpha line at ~ 6.5 keV and the X-ray continuum was determined to vary virtually simultaneously, with an upper limit to the delay of less then 250 s. By using line parameters (stationary line centroid energy and width) and a timescale for the variability (flux variation or line delay) we can estimate an inner radius of the line emitting region between 6-30 gravitational radii (GM/c2), a disk inclination i < 40 and a central black hole mass 8 104 L43 < M < 3.9 106 Mo (where L43 is the accretion luminosity in units of 1043 erg/s). Alternative accretion disk scenarios have also been discussed. The energy conversion efficiency implied by the fastest X-ray continuum variations, eta ~ 0.1, is suggestive of a storage mechanism of gravitationally released energy or the presence of a rotating black hole.
Though the physical conditions of the diffuse interstellar medium are not suitable for efficient accretion onto isolated galactic black hole, it has been suggested that molecular clouds, with their higher densities, could offer the environment for spherically accreting objects to produce detectable fluxes. Calculating the birth rate of isolated black holes and following their evolution in the galactic potential, we estimate their total number within the molecular cloud layer. By assuming the model by Ipser & Price for describing spherically accreting black holes, we derive that detectable fluxes at optical-infrared wavelengths as well as a marginally detectable X-ray flux, can be achieved only from slowly moving sources which result to be ~ 0.5-7 per molecular cloud within 1 kpc from the Sun (depending on the unknown initial velocity dispersion).
The identification of these sources is complicated by the presence of many other strong infrared sources in molecular clouds, such as T Tauri stars, therefore we need a criterion to discriminate accreting black holes. Isolated, faint optical sources with, probably, no emission lines which exhibit rapid intensity fluctuations on timescale down to 10^-5 - 10^-2 s with no periodic components and long term variability (~ 1 yr), caused by black hole sampling of different regions in the cloud, could be accreting black holes. Small cometary tails, due to reprocessing of high energy photons, and a substantial proper motion for the nearest ones, can strengthen the identification. Other features that could sign the presence of accreting black holes, but which are much more model dependent, are color indexes which are different from the ones expected from T Tauri stars. A detailed multifrequency analysis of the sources with K magnitude brighter than 14 present in molecular clouds within 1 kpc, is a necessary step in order to identify accreting black holes.
Because the greater number of neutron star in respect to black holes and the presence of a hard surface which enhance the accretion efficiency, we have studied the possibility of detecting ONSs, accreting either from the interstellar medium or molecular clouds, with the ROSAT satellite.
The perspective of detecting solitary neutron stars in the ROSAT deep fields considered is rather poor. In the 10 square degrees covered so far by ROSAT deep exposures, the number of detectable ONSs should be at most ~ 5.
Favourable sites for the detection of ONSs are molecular clouds. We indeed expect to be able to detect few ONSs in most of the clouds and in some cases several. These ONSs should appear within the level of sensitivity of the ROSAT survey and therefore obviously in pointed observations. The perspective of observing ONSs in molecular clouds is exciting both for testing our expectations on ONSs, which is a field essentially based only on theoretical surmises and for studying the physics of molecular clouds.