Investigations of AGN populations through X-ray and optical surveys

The identification of Active Galactic Nuclei (AGN) and their classification as obscured or unobscured are important parameters for the explanation of the X-ray background and its synthesis models. AGN detection and classification is expected to be sensitive to its environment (obscuration, host galaxy starlight, star-formation) and to its intrinsic properties such as the Eddington ratio. Using the 3XMM-DR4 X-ray catalogue, and the SDSS-DR9 spectroscopic catalogue, we have obtained a new sample of X-ray selected narrow emission line galaxies. A mismatch between the optically-based and X-ray-based classifications has been found for 6% of the X-ray selected AGN, named the elusive Seyfert 2 (Sy2). The absence of an optical signature can be understood by the intrinsic weakness of the AGN which can be associated with dilution from the host galaxy. In addition, some of the sources are good candidates to be "True Sy2" AGN as they are very likely to lack the broad line region. By extending the work to higher redshift using the Chandra-COSMOS sample, which allows fainter X-ray sources to be detected, evolutionary effects can be examined. The study of elusive Sy2 among this sample have shown similar explanations for the absence of an optical AGN signature, i.e. weak AGN with low accretion rates in addition to, in some cases, dilution by galaxy starlight. Some unobscured Sy2 have also been confirmed to intrinsically lack broad-lines. Importantly, it seems that the fraction of elusive Sy2 increases with redshift. The fraction of unobscured Sy2 among X-ray selected and/or optically selected AGN from the 3XMM-SDSS sample is about 60% which drops to 47% after subtraction of the contaminants (narrow-line Sy1 and Compton-thick AGN). The main explanations for the absence of optical broad-lines are variability, host galaxy dilution, dust reddening and intrinsic absence of the broad line region.