Ultraluminous infrared galaxies and the origin of quasars

Abstract

An evolutionary connection between ultraluminous infrared galaxies and quasars is deduced from the observations of all 10 infrared galaxies with luminosities L(8-1000 µm) ~ 10 12 L 0 , taken from a flux-limited sample of infrared bright galaxies. Images of the infrared galaxies show that nearly all are strongly interar•ing merger systems with exceptionally luminous nuclei. Millimeter-wave CO observations show that these objects typically contain 0.5-2 x 10 10 M 0 of H2. Optical spectra indicate a mixture of starburst and active galactic nucleus (AGN) energy sources, both of which are apparently fueled by the tremendous reservoir of molecular gas. It is proposed that these ultraluminous infrared galaxies represent the initial, dust-enshrouded stages of quasars. Once these nuclei shed their obscuring dust, allowing the AGN to visually dominate the decaying starburst, they become optically selected quasars. The origin of quasars through the merger of molecular gasrich spiral galaxies can account for both the increased number of high-luminosity quasars at large redshift, when the universe was smaller and gas supplies less depleted, and the observed "redshift-cutoff" of quasars which represents the epoch after galaxy formation when the first collisions occur.

Dusty star-forming galaxies at high redshift

“…The observation that high infrared luminosity correlated with high star formation rates and major mergers led to the widely accepted evolutionary picture for extreme luminosity systems first proposed by Sanders et al (1988). A schematic diagram of this evolutionary picture is shown in Figure 2, here reproduced from Hopkins et al (2008).…”

“…Kennicutt, 1998a).While 8-1000µm is inclusive of all dusty emission, the drawback is that such a wide range captures multiple types of dust emission processes, from cold diffuse dust which dominates the long-wavelength portion of the SED to hot-dust and PAH emission in the mid-infrared (described more in § 5.7) to non-star formation driven heating, like AGN heating ( § 5.6). Non-star forming processes can come close to dominating the 8-1000µm infrared luminosity, particularly for optically or X-ray identified AGN at rest-frame wavelengths < ∼ 40µm (Sanders et al, 1988;. As a result, some works have pushed a narrower range of integration limits to restrict the computation of L IR to star-formation-driven emission only.…”

Star formation in mergers with cosmologically motivated initial conditions

“…Sanders et al 1988;Treister et al 2012;Bessiere et al 2014). There is increasing evidence that most extreme local star forming galaxies, the Ultra Luminous InfraRed Galaxies (ULIRGS), are triggered by mergers (e.g.…”

Elliptical Galaxies with Emission Lines from the Sloan Digital Sky Survey

“…Major mergers are the mergers of two disc galaxies of comparable mass, using the numerical simulations, Mihos & Hernquist (1996) have shown that starburst activities during the merger process could be two orders of magnitude higher than that in isolated galaxies and can be sustained from several 10 7 yr to ∼ 2 × 10 8 yr after the collision. It could be the case of many infrared luminous galaxies, which are found to show morphological peculiarities indicative of encounters, such as multiple nuclei, tidal tails, loops, and shells (Sanders et al 1988;Sanders 1992). Though major mergers trigger the most powerful starburst, they are less common than minor mergers with the satellite less than 10% of the galaxy's mass.…”