A new technique to efficiently select Compton-thick AGN

Abstract

We present a new efficient diagnostic method, based on mid-infrared and X-ray data, to select local (z<0.1) Compton-thick (CT) AGN with the aim of estimating their surface and space density. We define a region in the X-ray/IR vs. HR plane associated to CT AGN, i.e. F(2-12keV)/F25nu25<0.02 and HR>-0.2. We build up a sample of 43 CT AGN candidates using data from IRAS-PSC and 2XMM catalogue. In order to test the efficiency of the proposed method in selecting CT AGN we use the results of the X-ray spectral analysis performed on all the sources of our sample. After taking into account the different selection effects, we have estimated the number of CT in the local Universe and their density down to the IRAS flux limit of F25=0.5Jy. We find that the diagnostic plot proposed here is an efficient method to select Compton-thick AGN in the nearby Universe since ~84% of the sources populating the proposed CT region are actually CT AGN. Twenty percent are newly-discovered CT AGN. We then estimate the surface density of CT AGN down to the IRAS PSC catalogue flux limit (F25=0.5Jy) that turns out to be ~3e-3 src deg-2. After estimating an equivalent IR-hard X-ray limiting flux, we compare our result with those found with SWIFT-BAT. We find that the surface density derived here is a factor 4 above the density computed in the hard X-ray surveys. This difference is ascribed, at least in part, to a significant contribution (~60-90%) of the star-forming activity to the total 25 mic emission for the sources in our sample. By considering only the 25 mic AGN emission, we estimate a surface density of CT AGN which is consistent with the results found by hard X-ray surveys. Finally, we estimated the co-moving space density of CT AGN with intrinsic LX>1e43 erg s-1 (0.004<z<0.06): ~3.5e-6 Mpc-3. The prediction for CT AGN based on the synthesis model of XRB in Gilli et al.(2007) is consistent with this value.