AGN at low and high redshifts: physical properties and co-evolution of galaxies and black holes

Abstract

Active galactic nuclei (AGN) are the most luminous astronomical objects in the sky. Different from normal galaxies, AGN are characterized by a strong luminous nucleus which can not be associated with its stars. In this work we study AGN at low and high redshift. We use low redshift AGN in order to study in detail the physical properties of the broad line region (BLR), in particular its FeII emission. Moving to the high redshift universe we present a new discovery of the most distant radio galaxy known and we use high redshift quasars in order to study their central supermassive black holes (SMBH) and how they grow. We start this thesis by studying low redshift Seyfert galaxies and quasars to probe their broad line region (BLR). Our primary focus here is to use near-infrared (NIR) spectra to study the FeII emission and its role in the BLR. In order to analyse FeII emission in AGN we present a sample of 25 AGN observed with the SpeX at the Infrared Telescope Facility (IRTF) in the NIR. We modeled the FeII emission in these AGN using the only NIR template available. We tested the NIR FeII template for our entire sample. The results obtained show that the NIR FeII template successfully reproduces the NIR FeII emission in AGN, independent of their luminosity and FeII intensity. We detected the FeII bump at 9200Ang in all spectra of our sample, as predicted by theoretical models. This result confirms Ly-alpha fluorescence as one of the excitation mechanisms responsible for the production of the FeII emission. We analysed the FeII emission in the NIR through the 9200Ang bump and the 1 micron lines and find a strong correlation between these features and the optical FeII emission, proving that Ly-alpha fluorescence plays an important role in the FeII production. We estimate an inferior limit of 18% for the contribution of this mechanism to the production of the optical FeII emisison. The analysis of the line profiles of the FeII 10502Ang, OI 11287Ang, CaII 8664Ang and Pa-Beta lines shows that the first three ones have similar widths. On the other hand, FeII is systematically narrower by a factor of 0.7 than Pa-Beta. Assuming the emitting clouds are virialized, this result implies that the FeII is formed in a region that is twice more distant than Pa-Beta, ranging 8.5 to 198.2 light-days. This is important for constraining future photoionization models aimed to study the FeII emission. The second main result of this thesis is the discovery of a new radio galaxy, TGSSJ1530+1049, at z=5.72 which is the most distant radio galaxy to be discovered since TNJ0924-2201 at z=5.2 in 1999. This galaxy was discovered through the follow up of ultra-steep spectrum galaxies selected at 150MHz. The analysis of the spectral index at 150Mhz and 1.4GHz shows an alpha(150 MHz)(1.4 GHz)=-1.4, and images with the Very Large Array (VLA) at 1.4GHz show an unresolved compact structure. The spectroscopic follow up was carried out with the GEMINI telescope using the GEMINI Multi-Object Spectrograph (GMOS) and revealed the presence of a single emission line at the observed wavelength of lambda=8165Ang. The asymmetric shape of the line, absence of other optical lines, and high equivalent width lead us to conclude that this line is Lyman-alpha, at z=5.72. We measured a line luminosity of 5.7x10^42erg/s, a FWHM of 370km/s, a deconvolved physical extent of 3.5kpc, and a radio luminosity of L{150Mhz}=29.1$W/Hz. Although the line luminosity and full width at half maximum (FWHM) are comparable to those of other non-radio Lyman alpha galaxies the size and radio luminosity are common to those of other z>4 radio galaxies. NIR imaging in J and K with Large Binocular Telescope (LBT) using LBT Utility with CAM and Integral FIELD (LUCI) did not detect the host galaxy at the 3-sigma detection limit of J>24.4 and K>22.4 mag. These limits are consistent with those expected for a radio galaxy at z>5 in accordance with the radio galaxy K-z relation. We estimate the stellar mass limit using simple stellar population models finding M_{stars}<10^{10.5} Msol, which suggests a radio galaxy in an early phase of evolution. The third maing result of this thesis relates to high redshift quasars and their supermassive black holes (SMBH). We used a sample of 35 high redshift radio loud quasars (2.2<z<3.5) from the Clusters Around Radio Loud AGN} (CARLA) survey observed with the Very Large Telescope (VLT) using the Spectrograph for INtegral Field Observations in the Near Infrared (SINFONI). We used the H-alpha line for accurate estimates of the BH masses and compare with previous results using CIV. Our results show a reduced scatter of 0.27dex of CIV compared to H-alpha. We found that the Eddington ratio strongly correlates with H-alpha and L(5100) and only moderately with M_{BH}. On the other hand, the radio power does not correlate with the main observables. We estimate the growth time of the BH and found a median below the cosmic age at the redshift of the quasars, but with a significantly fraction of them having longer times. Our results suggest that the BH mass of these quasars experienced a near-Eddington growth in the past, followed by a slower accretion phase at the measured L/L_{Edd} until z~2. We finish this chapter revisiting the correlation between the M_{BH} and environment as measured by the CARLA survey and discuss our results in the context of the local M-sigma relation.