We present results from a spectral analysis of a sample of high-redshift (z>3) X-ray selected AGN in the 4 Ms Chandra Deep Field South (CDF-S), the deepest X-ray survey to date. The sample is selected using the most recent spectroscopic and photometr
ic information available in this field. It consists of 34 sources with median redshift z=3.7, 80 median net counts in the 0.5-7 keV band and median rest-frame absorption-corrected luminosity $L_{2-10 rmn{keV}}approx1.5times10^{44}rmn{erg} rmn{s^{-1}}$. Spectral analysis for the full sample is presented and the intrinsic column density distribution, corrected for observational biases using spectral simulations, is compared with the expectations of X-ray background (XRB) synthesis models. We find that $approx57$ per cent of the sources are highly obscured ($N_H>10^{23}rmn{cm^{-2}}$). Source number counts in the $0.5-2rmn{keV}$ band down to flux $F_{0.5-2 rmn{keV}}approx4times10^{-17}rmn{erg} rmn{s^{-1}cm^{-2}}$ are also presented. Our results are consistent with a decline of the AGN space density at z>3 and suggest that, at those redshifts, the AGN obscured fraction is in agreement with the expectations of XRB synthesis models.
We present the analysis of four XMM-Newton observations of the narrow-line quasar PG 1543+489 at z=0.400 carried out over a rest-frame time-scale of about three years. The X-ray spectrum is characterized by a broad, relativistic iron K_alpha emission
line and a steep photon index, which can be both explained by a ionized reflection model, where the source of X-ray photons is presumably very close to the black hole. If this were the case, strong light-bending effects are expected, and actually they provide the most plausible explanation for the large equivalent width (EW=3.1+/-0.8 keV in the source rest frame) of the iron line. Although the light-bending model provides a good description of the X-ray data of PG 1543+489, it is not possible to rule out an absorption model, where obscuring matter partially covers the X-ray source. However, the apparent lack of variations in the properties of the absorber over the time-scale probed by our observations may indicate that this model is less likely.
