A new sample of X-ray selected narrow emission-line galaxies

A sample of X-ray and optically selected narrow emission-line galaxies (769 sources) from the 3XMM catalogue cross-correlated with SDSS (DR9) catalogue has been studied. Narrow-emission line active galactic nuclei (AGN; type-2) have been selected on the basis of their emission line ratios and/or X-ray luminosity. We have looked for X-ray unobscured type-2 AGN. As X-ray spectra were not available for our whole sample, we have checked the reliability of using the X-ray hardness ratio (HR) as a probe of the level of obscuration and we found a very good agreement with full spectral fitting results, with only 2% of the sources with apparently unobscured HR turning out to have an obscured spectrum. Despite the fact that type-2 AGN are supposed to be absorbed based on the Unified Model, about 60% of them show no sign or very low level of X-ray obscuration. After subtraction of contaminants to the sample, that is Narrow-Line Seyfert 1 and Compton-thick AGN, the fraction of unobscured Sy2 drops to 47%. For these sources, we were able to rule out dust reddening and variability for most of them as an explanation of the absence of optical broad emission-lines. The main explanations remaining are the dilution of weak/very broad emission-lines by the host galaxy and the intrinsic absence of the broad-line region (BLR) due to low accretion rates (i.e. True Sy2). However, the number of True Sy2 strongly depends on the method used to verify the intrinsic lack of broad lines. Indeed using the optical continuum luminosity to predict the BLR properties gives a much larger fraction of True Sy2 (about 90% of the unobscured Sy2 sample) than the use of the X-ray 2 keV luminosity (about 20%). Nevertheless the number of AGN we securely detected as True Sy2 is at least three times larger that the previously confirmed number of True Sy2.