A high Eddington-ratio, true Seyfert 2 galaxy candidate: implications for broad-line region models

A bright, soft X-ray source was detected on 2010 July 14 during an XMM–Newton slew at a position consistent with the galaxy GSN 069 (z = 0.018). Previous ROSAT observations failed to detect the source and imply that GSN 069 is now ≥240 times brighter than it was in 1994 in the soft X-ray band. Optical spectra (from 2001 to 2003) are dominated by unresolved emission lines with no broad components, classifying GSN 069 as a Seyfert 2 galaxy. We report here results from a ∼1 yr monitoring with Swift and XMM–Newton, as well as from new optical spectroscopy. GSN 069 is an unabsorbed, ultrasoft source in X-rays, with no flux detected above ∼1 keV. The soft X-rays exhibit significant variability down to time-scales of hundreds of seconds. The UV-to-X-ray spectrum of GSN 069 is consistent with a pure accretion disc model which implies an Eddington ratio λ ≃ 0.5 and a black hole mass of ≃ 1.2 × 10⁶ M_⊙. A new optical spectrum, obtained ∼3.5 months after the XMM–Newton slew detection, is consistent with earlier spectra and lacks any broad-line component. The lack of cold X-ray absorption and the short time-scale variability in the soft X-rays rule out a standard Seyfert 2 interpretation of the source. The present Eddington ratio of GSN 069 exceeds the critical value below which no emitting broad-line region (BLR) forms, according to popular models, so that GSN 069 can be classified as a bona-fide high Eddington-ratio true Seyfert 2 galaxy. We discuss our results within the framework of two possible scenarios for the BLR in AGN, namely the two-phase model (cold BLR clouds in pressure equilibrium with a hotter medium), and models in which the BLR is part of an outflow, or disc-wind. Finally, we point out that GSN 069 may be a member of a population of super-soft active galactic nuclei (AGN) whose spectral energy distribution is completely dominated by accretion disc emission, as it is the case in some black hole X-ray binary transients during their outburst evolution. The disc emission for a typical AGN with black hole mass of 10⁷–10⁸ M_⊙ does not enters the soft X-ray band, so that GSN 069-like objects with larger black hole mass (i.e. the bulk of the AGN population) are missed by current X-ray surveys, or misclassified as Compton-thick candidates. If the analogy between black hole X-ray binary transients and AGN holds, the lifetime of these super-soft states in AGN may be longer than 10⁴ years, implying that the actual population of super-soft AGN may not be negligible, possibly contaminating the estimated fraction of heavily obscured AGN from current X-ray surveys.