Obscuring fraction of active galactic nuclei implied by supernova and radiative feedbacks

TL;DR

This study models the obscuring structures of circumnuclear disks in AGNs, highlighting how supernova and radiative feedbacks influence the obscuring fraction depending on black hole mass and luminosity.

Contribution

It introduces a combined supernova and radiative feedback model to explain the obscuring fraction in AGNs, aligning with observations across different black hole masses.

Findings

Obscuring fraction peaks at ~10% of Eddington luminosity.

Maximal obscuring fraction is ~0.6 for SMBHs with less than 10^8 solar masses.

For more massive SMBHs, the obscuring fraction is below 0.2 and less dependent on luminosity.

Abstract

We study the obscuring structure of circumnuclear disks (CNDs) by considering supernova (SN) feedbacks from nuclear starburst and the effect of anisotropic radiative pressure from AGNs. We suppose that the mass accretion onto a central supermassive black hole (SMBH) is triggered by SN-driven turbulence within CNDs, and we explore how the structures of CNDs depend on the BH mass ($M_{\rm BH}$) and AGN luminosity ($L_{\rm AGN}$). We find that the obscuring fraction ($f_{\rm obs}$) peaks at $\sim10\%$ of the Eddington luminosity ($L_{\rm Edd}$), and its maximal value is $f_\mathrm{obs} \sim 0.6$ for less massive SMBHs (e.g., $M_{\rm BH} < 10^{8}M_{\odot}$). This is because the scale height of CNDs is determined by the SN-driven accretion for a smaller $L_{\rm AGN}$, while the dusty molecular gas in CNDs is blown away by the radiation pressure from AGNs beyond the critical luminosity. On the other hand, for massive SMBHs (e.g., $M_{\rm BH} > 10^{8}M_{\odot}$), $f_{\rm obs}$ is always smaller than $0.2$, and it is almost independent of $L_{\rm AGN}$ because the scale height of CNDs is mainly controlled by the maximal star-formation efficiency ($C_{\rm *, max}$) in CNDs. By comparison with the obscuring fractions suggested from the mid-infrared observations of nearby AGNs, the SN plus radiative feedback model with $C_{\rm *, max} = 10^{-7}\, {\rm yr}^{-1}$ well reproduces the observations for $M_\mathrm{BH} = 10^8 M_\odot$. We also find that the intense starburst or the existence of dust-free absorbers inside CNDs are necessary, to explain X-ray observations.