One of the major outstanding questions in galaxy evolution concerns the interplay between star formation and AGN activity. Many theoretical models have proposed that AGN feedback is responsible for regulation of star formation, either by heating or removing the cold gas in their host galaxies (e.g., Silk & Rees 1998; Di Matteo et al. 2005; Springel et al. 2005; Hopkins et al. 2008). In this picture, galaxy mergers are thought to be a possible fueling mechanism for AGN activity wherein black holes grow rapidly close to the Eddington limit, until the AGN feedback is sufficient to blow out the surrounding gas, suppressing star formation. On the other hand, stochastic low-luminosity AGNs, triggered by secular processes, maintain a hot gas halo to prevent further star formation in galaxies (e.g., Bower et al. 2006; Croton et al. 2006). However, the observational evidence for the existence and nature of this feedback is highly incomplete, and the extent to which AGN activity affects the star formation process is still a matter of debate.
The main focus of this dissertation research is to systematically study the influence of AGN on the growth of galaxies by examining the possible connection between AGN activity and star formation in galaxies. The primary targets are selected in the X-ray, which is the most efficient way to select AGN over a wide range of luminosities and redshifts because these galaxies are less affected by obscuration, and the contamination from non-nuclear
emission (mainly due to star-formation) is far less significant than in optical- or infrared- selected samples (e.g., Donley et al. 2008; Lehmer et al. 2012; Stern et al. 2012). My approach is to combine multi-wavelength photometry and spectroscopy to characterize the AGN and their host galaxy properties, examining (1) black hole accretion at z ∼ 1 − 2, when SMBH growth is at its peak, to explore the growth of black holes, and (2) the AGN activity and its connection with star formation in AGN host galaxies. Below I highlight key results from this dissertation research.
5.1.1 AGN accretion and Growth of Black Holes
The assembly of galaxies and SMBHs appears to follow a “downsizing” trend; i.e. the AGN luminosity function and the SFRs show a violent phase with the most luminous AGNs and powerful star-forming galaxies at z ∼ 2 − 3, and more moderate activity at later cosmic times (z < 1; see e.g., Hasinger et al. 2005). The Eddington ratio, the ratio of the AGN bolometric and the Eddington luminosity, is particularly interesting because it provides an observational constraint on the efficiency of gas accretion during the active phases of a black hole’s life. Thus, to quantify a growth rate requires the independent measurement of AGN bolometric luminosity and black hole mass. Unfortunately, the detailed study in the redshift interval z ∼ 1 − 2, where the downsizing appears and where a significant part of the accretion growth of black holes takes place, has been difficult because of the lack of emission line diagnostics in the optical wavelength range, which is why the range z ∼ 1 − 2 is often referred to as the redshift desert.
Using high-quality optical and near-infrared spectroscopy from Keck/DEIMOS and Subaru/FMOS observations, I determine black hole masses and Eddington ratios via spectral-line fitting in the key redshift interval z = 1 − 2 using the Balmer lines that are the same lines for which the black hole masses are calibrated at low redshift. AGNs with similar black hole masses show a broad range of bolometric luminosities, which are calculated from X-ray luminosities, indicating that the accretion rate of black holes is distributed over a wide range. This suggests that the “AGN cosmic downsizing” phenomenon can be explained
by a substantial fraction of massive black holes accreting significantly below the Eddington limit at z < 2, in contrast to what is generally found for luminous AGNs at high redshift. This can be interpreted as the fraction of AGNs radiating close to the Eddington limit is decreasing after their peak activity phases (z ∼ 2 − 3), suggesting that the dominant fueling mechanism for the growth of black holes might change through the cosmic time (Suh et al. 2015).
5.1.2 AGN activity and Star Formation in AGN host galaxies
To examine whether AGN activity can significantly enhance or quench star formation in galaxies, I focused on the X-ray selected AGNs in the Chandra COSMOS Legacy Survey (CCLS; Civano et al. 2016). This study greatly improves upon the statistics of previous works, as it contains one of the largest samples of X-ray selected AGNs (4016 sources). In order to derive the physical properties of AGN host galaxies, I develop a multi-component SED fitting technique which allows one to disentangle the nuclear emission from the stellar light, and derived host galaxy properties using the Bayesian statistics. Specifically, I decompose the entire SED into separate components with nuclear AGN emission, the host galaxy’s stellar populations, and a starburst contribution in the FIR using the existing multi- wavelength photometric data (from NUV through the FIR) available in the COSMOS field. This technique of SED decomposition is crucial for estimating reliable physical properties of the host galaxies, such as stellar mass and star formation rate (SFR).
Combining with black hole masses estimated via virial method, I investigate the evolution of MBH− Mstellar relation up to z ∼ 2.5. The MBH− Mstellar distribution for
the moderate-luminosity AGNs shows a broadly consistent with the local scaling relation, suggesting that there is no significant evolution with redshift. The MBH− Mstellar ratios
seem to depend on the Eddington ratio in the sense that AGNs with higher Eddington ratio have smaller black hole masses for a given stellar mass, lying below the local MBH− Mstellar
relation. I also explore the distribution of AGN host galaxies on the SFR−Mstellar diagram
the SFR and their stellar mass, commonly referred to as the main sequence (MS) of star formation (e.g., Noeske et al. 2007; Daddi et al. 2007; Elbaz et al. 2007). Overall, AGN host galaxies seem to have SFRs that lie on the star-forming MS, but with much broader dispersions. While the “flattening” in the star-forming MS at high masses could be interpreted as a consequence of quenching the star formation, the SFRs of AGN host galaxies are consistent with those expected from normal star-forming galaxies in most stellar mass bins up to z ∼ 3, indicating no clear signature for enhanced or suppressed SFRs compared to normal star-forming galaxies (Suh et al. 2017). The correlation between LIR due to star
formation and X-ray luminosities show broadly consistent with being flat relationship, which can be explained by the AGN variability along with the broad Eddington ratio distribution. All of these results imply that nuclear activity and star formation seem to co-exist fueling black hole accretion and star formation simultaneously. There seems to be a lack of significant evolution in the stellar masses of X-ray selected AGN host galaxies, which are typically massive (> 1010 M
⊙) since z∼3, indicating that they might have already
experienced substantial growth at higher redshift. This could be suggests that the secular evolution may play an important role in growing SMBHs and the bulge formation in massive galaxies for the majority of AGN host galaxies at later cosmic time (z < 3).