Systematic Infrared 2.5—5μm Spectroscopy of Nearby Ultraluminous Infrared Galaxies with AKARI

Abstract

We report on the results of systematic infrared 2.5-5 µm spectroscopy of 45 nearby ultraluminous infrared galaxies (ULIRGs) at z < 0.3 using the IRC infrared spectrograph onboard the AKARI satellite. This paper investigates whether the luminosities of these ULIRGs are dominated by starburst activity, or alternatively, whether optically elusive buried active galactic nuclei (AGNs) are energetically important. Our criteria include the strengths of the 3.3 µm polycyclic aromatic hydrocarbon (PAH) emission features and the optical depths of absorption features at 3.1 µm due to icecovered dust grains and at 3.4 µm from bare carbonaceous dust grains. Because of the AKARI IRC's spectroscopic capability in the full 2.5-5 µm wavelength range, unaffected by Earth's atmosphere, we can apply this energy diagnostic method to ULIRGs at z > 0.15. We estimate the intrinsic luminosities of extended (several kpc), modestly obscured (A V < 15 mag) starburst activity based on the observed 3.3 µm PAH emission luminosities measured in AKARI IRC slitless spectra, and confirm that such starbursts are energetically unimportant in nearby ULIRGs. In roughly half of the observed ULIRGs classified optically as non-Seyferts, we find signatures of luminous energy sources that produce no PAH emission and/or are more centrally concentrated than the surrounding dust. We interpret these energy sources as buried 1 AGNs. The fraction of ULIRGs with detectable buried AGN signatures is higher in ULIRGs classified optically as LINERs than HII-regions, and increases with increasing infrared luminosity. Our overall results support the scenario that luminous buried AGNs are important in many ULIRGs at z < 0.3 classified optically as non-Seyferts, and that the optical undetectability of such buried AGNs occurs merely because of a large amount of nuclear dust, which can make the sightline of even the lowest dust column density opaque to the ionizing radiation of the AGNs.

Luminous Buried AGNs in Ultraluminous Infrared Galaxies and the Origin of Galaxy Downsizing

“…Figures 2, 3, and 4 present Subaru, AKARI, and Spitzer infrared spectra of selected ULIRGs at z < 0.3, respectively (Imanishi et al, 2006;2008).…”

THE 3.3 μm POLYCYCLIC AROMATIC HYDROCARBON EMISSION AS A STAR FORMATION RATE INDICATOR

“…The difference between these two correlations supports AGN activity contributing to L IR more than has thus far been considered to be the case. Also, considering the fact that there is no sign of departure for the ULIRG sample of Imanishi et al (2008) from the correlation between L PAH3.3 and L PAH6.2 (Figure 6), it is more likely to be L IR which causes the break. Recently, Yamada et al (2012) have found that the correlation between L PAH3.3 / L IR and L IR has a break around L IR ∼ 10 12 L from another MP of AKARI, MSAGN.…”

“…Therefore, we look for the correlation between the combined sample of AMUSES and the literature samples. The large filled circles represent samples of AMUSES objects, while the small symbols represent samples from the literature (Rodríguez-Ardila & Viegas 2003;Imanishi et al 2008Imanishi et al , 2010Sajina et al 2009;Lee et al 2012). Note that L IR comes from Wu et al (2010).…”

“…Imanishi et al (2008) obtained AKARI Infrared Camera (IRC; Onaka et al 2007) spectra of 45 nearby ultraluminous infrared galaxies (ULIRGs) in order to investigate the energy source of ULIRGs. They detected the 3.3 μm PAH emission from 40 ULIRGs.…”

Unbiased Large Spectroscopic Surveys of Galaxies Selected by SPICA Using Dust Bands

“…Therefore the equivalent widths of the PAH features enable us to roughly estimate the star-formation contribution to the total L IR of a galaxy (Moorwood 1986;Roche et al 1991;Genzel et al 1998;Armus et al 2007;Imanishi et al 2007Imanishi et al , 2008Imanishi et al , 2010Veilleux et al 2009;Nardini et al 2008Nardini et al , 2009Nardini et al , 2010Pope et al 2008;Menéndez-Delmestre et al 2009;Coppin et al 2010;Stierwalt et al 2013Stierwalt et al , 2014, and more reliably estimate it when they are normalized with other spectral indicators (such as H 2 or [NeII] 12.8 µm line fluxes) or the slope of the IR continuum (Tommasin et al 2010). This approach is complementary to the one that uses spectral energy distribution (SED) fitting (e.g., Gruppioni et al 2016;Delvecchio et al 2014).…”