SUPER: III. Broad line region properties of AGNs at z ∼2

Aims. The SINFONI survey for Unveiling the Physics and Effect of Radiative feedback (SUPER) was designed to conduct a blind search for AGN-driven outflows on X-ray-selected AGNs at redshift z∼2 with high (∼2 kpc) spatial resolution, and to correlate them with the properties of their host galaxy and central black hole. The main aims of this paper are: (a) to derive reliable estimates for the masses of the black holes and accretion rates for the Type-1 AGNs in this survey; and (b) to characterise the properties of the AGN-driven winds in the broad line region (BLR). Methods. We analysed rest-frame optical and UV spectra of 21 Type-1 AGNs. We used Hα, Hβ, and MgII line profiles to estimate the masses of the black holes. We used the blueshift of the CIV line profile to trace the presence of winds in the BLR. Results. We find that the Hα and Hβ line widths are strongly correlated, as is the line continuum luminosity at 5100 Å with Hα line luminosity, resulting in a well-defined correlation between black hole masses estimated from Hα and Hβ. Using these lines, we estimate that the black hole masses for our objects are in the range Log (MBH/M·) = 8.4-10.8 and are accreting at λEdd = 0.04-1.3. Furthermore, we confirm the well-known finding that the CIV line width does not correlate with the Balmer lines and the peak of the line profile is blueshifted with respect to the [OIII]-based systemic redshift. These findings support the idea that the CIV line is tracing outflowing gas in the BLR for which we estimated velocities up to ∼4700 km s-1. We confirm the strong dependence of the BLR wind velocity on the UV-to-X-ray continuum slope, the bolometric luminosity, and Eddington ratio. We infer BLR mass outflow rates in the range 0.005-3 M· yr-1, revealing a correlation with the bolometric luminosity consistent with that observed for ionised winds in the narrow line region (NLR), and X-ray winds detected in local AGNs, and kinetic power ∼10-7-10-4 × LBol. The coupling efficiencies predicted by AGN-feedback models are much higher than the values reported for the BLR winds in the SUPER sample; although it should be noted that only a fraction of the energy injected by the AGN into the surrounding medium is expected to become kinetic power in the outflow. Finally, we find an anti-correlation between the equivalent width of the [OIII] line and the CIV velocity shift, and a positive correlation between this latter parameter and [OIII] outflow velocity. These findings, for the first time in an unbiased sample of AGNs at z∼2, support a scenario where BLR winds are connected to galaxy-scale detected outflows, and are therefore capable of affecting the gas in the NLR located at kiloparsec scale distances.