2007A&A...463..903D

We present an analysis of all BeppoSAX observations of 1. This source was observed 5 times from 1996 to 2001 with durations ranging from a day to four days. The intrinsic continuum (described as a cut-off power law) is absorbed at low energies by a complex system: a cold patchy absorber plus a warm uniform screen photoionised by the central continuum. We find that this ``dual absorber'' is the main driver of the observed variability, up to a factor of eight, at 3keV. In particular the covering fraction of the cold absorber changes on time scales of the order of a day, supporting its association with the broad-line region. The column density of the warm gas varies on a longer time scale (months to year). Some of the small amplitude spectral variability above 10keV can be explained with an intrinsic variation (with variation of the photon index ΔΓ∼0.2). The flux below 1keV remains constant confirming an extended origin. Its spectrum is reproduced by a combination of a thermal component (with temperature kT=0.15keV) and a power law with the same slope as the intrinsic continuum, but with an intensity of a few percent. A Compton reflection component is significantly detected in 1996 (averaged value of Ω/2π∼0.4, with the solid angle Ω covered by the reflecting medium), with intensity decreasing on a time scale of a year, and it desappears in 2000 and 2001. The long time scale of variations argues for an association with an optically thick torus at a distance of a few light years. An iron line was detected in all spectra. Its energy is consistent with fluorescence by cold iron. We find that the line is variable. Its behaviour is reproduced by a variable component proportional to the level of the reflection flux plus a constant component. The flux of the latter is consistent with the extended line emission observed by Chandra. We conclude that the first component likely arises from the torus and the second is produced in the extended narrow-line region.