Abstract

The absence of high-Eddington-ratio, obscured active galactic nuclei (AGNs) in local (z less than or similar to 0.1) samples of moderate-luminosity AGNs has generally been explained to result from radiation pressure on the dusty gas governing the level of nuclear (less than or similar to 10 pc) obscuration. However, very high accretion rates are routinely reported among obscured quasars at higher luminosities and may require a different feedback mechanism. We compile constraints on obscuration and Eddington ratio for samples of X-ray, optical, infrared, and submillimeter selected AGNs at quasar luminosities. Whereas moderate-luminosity, obscured AGNs in the local universe have a range of lower Eddington ratios (f(Edd) similar to 0.001-0.1), the most luminous (L-bol greater than or similar to 10(46) erg s(-1)) IR/submillimeter-bright, obscured quasars out to z similar to 3 commonly have very high Eddington ratios (f(Edd) similar to 0.11). This apparent lack of radiation-pressure feedback in luminous, obscured quasars is likely coupled with AGN timescales, such that a higher fraction of luminous, obscured quasars are seen because of the short timescale for which quasars are most luminous. When adopting quasar evolutionary scenarios, extended (similar to 10(2-3) pc) obscuration may work together with the shorter timescales to explain the observed fraction of obscured, luminous quasars, while outflows driven by radiation pressure will slowly clear this material over the AGN lifetime.