Luminous buried AGNs as a function of galaxy infrared luminosity revealed through Spitzer low-resolution infrared spectroscopy

We present the results of Spitzer IRS infrared 5-35 micron low-resolution spectroscopic energy diagnostics of ultraluminous infrared galaxies (ULIRGs) at z > 0.15, classified optically as non-Seyferts. Based on the equivalent widths of polycyclic aromatic hydrocarbon emission and the optical depths of silicate dust absorption features, we searched for signatures of intrinsically luminous, but optically elusive, buried AGNs in these optically non-Seyfert ULIRGs. We then combined the results with those of non-Seyfert ULIRGs at z < 0.15 and non-Seyfert galaxies with infrared luminosities L(IR) < 10^12Lsun. We found that the energetic importance of buried AGNs clearly increases with galaxy infrared luminosity, becoming suddenly discernible in ULIRGs with L(IR) > 10{12}Lsun. For ULIRGs with buried AGN signatures, a significant fraction of infrared luminosities can be accounted for by detected buried AGN and modestly-obscured (Av < 20 mag) starburst activity. The implied masses of spheroidal stellar components in galaxies for which buried AGNs become important roughly correspond to the value separating red massive and blue, less-massive galaxies in the local universe. Our results may support the widely-proposed AGN-feedback scenario as the origin of galaxy downsizing phenomena, where galaxies with currently larger stellar masses previously had higher AGN energetic contributions and star-formation-originating infrared luminosities, and have finished their major star-formation more quickly, due to stronger AGN feedback.

💡 Research Summary

This study exploits low‑resolution (R ≈ 60–120) Spitzer‑IRS spectra covering 5–35 µm to search for intrinsically luminous but optically hidden (buried) active galactic nuclei (AGNs) in galaxies that are classified as non‑Seyfert in the optical. The sample consists of two redshift bins of ultraluminous infrared galaxies (ULIRGs, L_IR > 10¹² L_☉) – 30 objects at z > 0.15 and 45 objects at z < 0.15 – together with a control set of ∼60 infrared‑bright galaxies with L_IR < 10¹² L_☉. The authors employ two well‑established mid‑infrared diagnostics: (1) the equivalent widths (EWs) of the 6.2 µm and 11.3 µm polycyclic aromatic hydrocarbon (PAH) features, which are strong in star‑forming regions but suppressed by the hard radiation fields of AGNs, and (2) the optical depth of the 9.7 µm silicate absorption (τ_9.7), which quantifies the column of intervening dust and thus the degree of obscuration. Objects with EW < 0.15 µm (6.2 µm) and τ_9.7 > 2.0 are flagged as buried‑AGN candidates.

The analysis shows a striking luminosity dependence. Only ∼10 % of the sub‑ULIRG control galaxies meet the buried‑AGN criteria, whereas roughly 60 % of the ULIRGs do. Moreover, the transition occurs sharply at L_IR ≈ 10¹² L_☉: above this threshold the fraction of buried AGNs jumps, indicating that the energetic importance of heavily obscured nuclei becomes dominant in the most infrared‑luminous systems. For ULIRGs with buried‑AGN signatures, the combined contribution of the AGN (re‑processed dust emission) and modestly obscured starbursts (A_V < 20 mag) can account for 30–70 % of the total infrared output, implying that a substantial portion of the bolometric luminosity is AGN‑driven even though the galaxies appear optically inactive.

The authors interpret these results in the context of the AGN‑feedback scenario for galaxy “downsizing”. The stellar masses of the spheroidal components inferred for galaxies where buried AGNs become significant (∼10¹⁰.⁵ M_☉) correspond to the empirical mass divide between red, massive galaxies and blue, less‑massive systems in the local universe. The implication is that galaxies that are now massive experienced stronger AGN activity in the past, which quenched star formation more efficiently and accelerated their transition to the red sequence. This observational evidence thus supports models in which AGN‑driven winds, radiation pressure, or jets heat or expel the interstellar medium, curtailing further star formation.

The paper also discusses limitations. High τ_9.7 can arise from dense star‑forming regions rather than an AGN torus, and PAH suppression may be caused by intense UV fields unrelated to AGN activity. The modest spectral resolution of IRS hampers detailed line profile analysis and the separation of blended features. The authors suggest that forthcoming JWST/MIRI observations, with superior sensitivity and resolution, will enable more precise decomposition of AGN and starburst components, better constraints on dust geometry, and a clearer picture of how buried AGNs influence galaxy evolution.

In summary, by combining PAH equivalent widths and silicate absorption depths, the study demonstrates that buried AGNs are rare in modest‑luminosity infrared galaxies but become prevalent and energetically important in ULIRGs. This luminosity‑driven emergence of hidden AGNs aligns with theoretical expectations of AGN feedback driving the rapid quenching of massive galaxies, offering a compelling observational link between infrared diagnostics and the broader narrative of galaxy formation and downsizing.