Unification of QSOs via black hole and accretion properties
Although the orientation-based AGN unification scheme can successfully explain many QSO observational phenomena, orientation does not address all the object-to-object differences in QSOs. Physical differences of the underlying engine, such as luminosity, black hole mass (MBH) and Eddington ratio (L/LEdd), are crucial to our understanding of QSO central engines. Broad Absorption Line (BAL) QSOs are a particularly interesting type of QSO that exhibits both orientation and intrinsic property-related observational features. In this thesis, I studied a large QSO sample, including 16 BAL QSOs at z ∼ 2, with new spectroscopy data for the Hβ region. This sample covers a luminosity range substantially wider than similar studies in the past and hence enables us to differentiate luminosity from other underlying mechanisms driving QSO observational properties. I found that overall, QSOs accrete at close to Eddington rate. Due to the narrow range of L/LEdd, the QSO luminosity is almost directly proportional to the MBH. The slight increase of L/LEdd at high luminosity suggests that the QSO MBH distribution has a high mass cut-off near 109M¯. Compared with radio quiet QSOs, radio loud QSOs tend to have higher MBH for the same luminosity. The [O iii] versus Fe ii anti-correlation discovered from low luminosity QSOs (BGEV1) extends to high luminosity objects with BAL QSOs at the weak [O iii] strong Fe ii end of the trend, and radio loud QSOs at strong [O iii] weak Fe ii end of the trend. Both [O iii] and Fe ii strengths are well correlated with L/LEdd over the entire luminosity range, indicating that L/LEdd is the physical driver behind the BGEV1 correlations. Although BAL QSOs have higher L/LEdd than most QSOs, they do not stand out when compared with high luminosity non-BAL QSOs. One interpretation is that [O iii] and Fe ii are indirectly linked to L/LEdd via the availability of accretion fuel. Even with the expanded luminosity coverage, I could not confirm the existence of an Hβ Baldwin Effect. An [O iii] ”Baldwin Effect” is observed, suggesting a limited amount of [O iii] NLR gas in all QSO systems.