Towards an observationally motivated AGN dusty torus model – II. The roles of density distribution and chemical composition of the dust

ABSTRACT Several models of nuclear dust in active galactic nuclei (AGNs) have been proposed to determine its physical and geometrical properties, usually assuming the dust density distribution as the main aspect producing differences in the mid-infrared (MIR) emission of AGNs. We present a study of the MIR emission of nearby AGNs by exploring the effects of dust distribution and chemical composition on the spectral energy distributions (SEDs) using radiative transfer simulations. Our model grid includes smooth, clumpy, and two-phase dust distributions, combined with two dust compositions: the interstellar medium (ISM) dust composition including large grains (up to $\rm {10 \ \mu m}$), and the oxide/silicate-based composition obtained in our previous work. A synthetic SED library was generated and analysed both on a model-to-model basis and with observed MIR spectra from 68 AGNs. We found that both dust distribution and composition significantly influence the spectral shape and silicate features at 10 and $\rm {18 \ \mu m}$, especially at edge-on orientations. Smooth distribution produces stronger and broader absorption features, while clumpy distribution enhances silicate emission. Two-phase distributions exhibit intermediate characteristics depending on clumpiness ($f_{\text{cl}}$) and filling factor ($f_{\text{fill}}$). The ISM composition with large grains better reproduces observations, particularly with Type-2 SEDs, independently of dust distribution. The oxide/silicate-based composition provides more good fits with Type-1 SEDs for $f_{\text{cl}}\le 0.5$, and with Type-2 SEDs for $f_{\text{cl}}\ge 0.9$. This work shows that none of the considered combinations of dust distribution and composition reproduces all observed spectra.