2006ApJ...639..740B

As a result of deep hard X-ray observations by Chandra and XMM-Newton, a significant fraction of the CXRB has been resolved into individual sources. These objects are almost all AGNs, and optical follow-up observations find that they are mostly obscured type 2 AGNs, have Seyfert-like X-ray luminosities, and peak in redshift at z∼0.7. Since this redshift is similar to the peak in the cosmic star formation rate, this paper proposes that the obscuring material required for AGN unification is regulated by star formation within the host galaxy. We test this idea by computing CXRB synthesis models with a ratio of type 2 to type 1 AGNs that is a function of both z and 2-10 keV X-ray luminosity, L_X. The evolutionary models are constrained by parameterizing the observed type 1 AGN fractions from the recent work by Barger et al. The parameterization that simultaneously best accounts for Barger's data, the CXRB spectrum, and the X-ray number counts has a local, low-L_Xtype 2/type 1 ratio of 4 and predicts a type 2 AGN fraction that evolves as (1+z)^0.3. This particular evolution predicts a type 2/type 1 ratio of 1-2 for logL_X>44, and thus the deep X-ray surveys are missing about half the obscured AGNs with these luminosities. These objects are likely to be Compton thick. Overall, these calculations show that the current data strongly support a change to the AGN unification scenario in which the obscuration is connected with star formation in the host galaxy rather than a molecular torus alone. The evolution of the obscuration implies a close relationship between star formation and AGN fueling.