Sensitive Herschel far-infrared observations can break degeneracies that were inherent to previous studies of star formation in high-z AGN hosts. Combining PACS 100 and 160 μm observations of the GOODS-N field with 2 Ms Chandra data, we detect ∼20% of X-ray AGN individually at >3σ. The host far-infrared luminosity of AGN with L 2−10 keV ≈ 10 43 erg s −1 increases with redshift by an order of magnitude from z = 0 to z ∼ 1. In contrast, there is little dependence of far-infrared luminosity on AGN luminosity, for L 2−10 keV < ∼ 10 44 erg s −1 AGN at z > ∼ 1. We do not find a dependence of far-infrared luminosity on X-ray obscuring column, for our sample which is dominated by L 2−10 keV < 10 44 erg s −1 AGN. In conjunction with properties of local and luminous high-z AGN, we interpret these results as reflecting the interplay between two paths of AGN/host coevolution. A correlation of AGN luminosity and host star formation is traced locally over a wide range of luminosities and also extends to luminous high-z AGN. This correlation reflects an evolutionary connection, likely via merging. For lower AGN luminosities, star formation is similar to that in non-active massive galaxies and shows little dependence on AGN luminosity. The level of this secular, non-merger driven star formation increasingly dominates over the correlation at increasing redshift.

Models of galaxy evolution assume some connection between the AGN and star formation activity in galaxies. We use the multi-wavelength information of the CDFS to assess this issue. We select the AGNs from the 3 Ms XMM-Newton survey and measure the star-formation rates of their hosts using data that probe rest-frame wavelengths longward of 20 μm, predominantly from deep 100 μm and 160 μm Herschel observations, but also from Spitzer-MIPS-70 μm. Star-formation rates are obtained from spectral energy distribution fits, identifying and subtracting an AGN component. Our sample consists of sources in the z ≈ 0.5−4 redshift range, with star-formation rates SFR ≈ 10 1 −10 3 M yr −1 and stellar masses M ≈ 10 10 −10 11.5 M . We divide the star-formation rates by the stellar masses of the hosts to derive specific star-formation rates (sSFR) and find evidence for a positive correlation between the AGN activity (proxied by the X-ray luminosity) and the sSFR for the most active systems with X-ray luminosities exceeding L x 10 43 erg s −1 and redshifts z 1. We do not find evidence for such a correlation for lower luminosity systems or those at lower redshifts, consistent with previous studies. We do not find any correlation between the SFR (or the sSFR) and the X-ray absorption derived from high-quality XMM-Newton spectra either, showing that the absorption is likely to be linked to the nuclear region rather than the host, while the star-formation is not nuclear. Comparing the sSFR of the hosts to the characteristic sSFR of star-forming galaxies at the same redshift (the so-called "main sequence") we find that the AGNs reside mostly in main-sequence and starburst hosts, reflecting the AGN-sSFR connection; however the infrared selection might bias this result. Limiting our analysis to the highest X-ray luminosity AGNs (X-ray QSOs with L x > 10 44 erg s −1 ), we find that the highest-redshift QSOs (with z 2) reside predominantly in starburst hosts, with an average sSFR more than double that of the "main sequence", and we find a few cases of QSOs at z ≈ 1.5 with specific star-formation rates compatible with the mainsequence, or even in the "quiescent" region. Finally, we test the reliability of the colour-magnitude diagram (plotting the rest-frame optical colours against the stellar mass) in assessing host properties, and find a significant correlation between rest-frame colour (without any correction for AGN contribution or dust extinction) and sSFR excess relative to the "main sequence" at a given redshift. This means that the most "starbursty" objects have the bluest rest-frame colours.

We have determined the relation between the AGN luminosities at rest-frame 6 µm associated to the dusty torus emission and at 2-10 keV energies using a complete, Xray flux limited sample of 232 AGN drawn from the Bright Ultra-hard XMM-Newton Survey. The objects have intrinsic X-ray luminosities between 10 42 and 10 46 erg sand redshifts from 0.05 to 2.8. The rest-frame 6 µm luminosities were computed using data from the Wide-Field Infrared Survey Explorer and are based on a spectral energy distribution decomposition into AGN and galaxy emission. The best-fit relationship for the full sample is consistent with being linear, L 6 µm ∝L 0.99±0.03 2−10 keV , with intrinsic scatter, ∆ log L 6 µm ∼0.35 dex. The L 6 µm /L 2−10 keV luminosity ratio is largely independent on the line-of-sight X-ray absorption. Assuming a constant X-ray bolometric correction, the fraction of AGN bolometric luminosity reprocessed in the mid-IR decreases weakly, if at all, with the AGN luminosity, a finding at odds with simple receding torus models. Type 2 AGN have redder mid-IR continua at rest-frame wavelengths <12 µm and are overall ∼1.3-2 times fainter at 6 µm than type 1 AGN at a given X-ray luminosity. Regardless of whether type 1 and type 2 AGN have the same or different nuclear dusty toroidal structures, our results imply that the AGN emission at rest-frame 6 µm is not isotropic due to self-absorption in the dusty torus, as predicted by AGN torus models. Thus, AGN surveys at rest-frame ∼6 µm are subject to modest dust obscuration biases.

Many theoretical models require powerful active galactic nuclei (AGNs) to suppress star formation in distant galaxies and reproduce the observed properties of today's massive galaxies. A recent study based on Herschel-SPIRE submillimeter observations claimed to provide direct support for this picture, reporting a significant decrease in the mean star formation rates (SFRs) of the most luminous AGNs (L X > 10 44 erg s −1 ) at z ≈ 1-3 in the Chandra Deep Field-North (CDF-N). In this Letter, we extend these results using Herschel-SPIRE 250 μm data in the COSMOS and Chandra Deep Field-South fields to achieve an order-of-magnitude improvement in the number of sources at L X > 10 44 erg s −1 . On the basis of our analysis, we find no strong evidence for suppressed star formation in L X > 10 44 erg s −1 AGNs at z ≈ 1-3. The mean SFRs of the AGNs are constant over the broad X-ray luminosity range of L X ≈ 10 43 -10 45 erg s −1 (with mean SFRs consistent with typical star-forming galaxies at z ≈ 2; SFRs ≈ 100-200 M yr −1 ). We suggest that the previous CDF-N results were likely due to low number statistics. We discuss our results in the context of current theoretical models.