The Infrared Nuclear Emission of Seyfert Galaxies on Parsec Scales: Testing the Clumpy Torus models

Abstract

We present subarcsecond resolution mid-infrared (mid-IR) photometry in the wavelength range from 8 to 20 micron of eighteen Seyfert galaxies, reporting high spatial resolution nuclear fluxes for the entire sample. We construct spectral energy distributions (SEDs) that the AGN dominates adding near-IR measurements from the literature at similar angular resolution. The IR SEDs of intermediate-type Seyferts are flatter and present higher 10 to 18 micron ratios than those of Seyfert 2. We fit the individual SEDs with clumpy torus models using the in-house-developed BayesClumpy tool. The models reproduce the high spatial resolution measurements. Regardless of the Seyfert type, even with high spatial resolution data, near- to mid-IR SED fitting poorly constrains the radial extent of the torus. For the Seyfert 2, we find that edge-on geometries are more probable than face-on views, with a number of clouds along equatorial rays of N = 5-15. The 10 micron silicate feature is generally modeled in shallow absorption. For the intermediate-type Seyferts, N and the inclination angle of the torus are lower than those of the Seyfert 2 nuclei, with the silicate feature appearing in weak emission or absent. The columns of material responsible for the X-ray absorption are larger than those inferred from the model fits for most of the galaxies, which is consistent with X-ray absorbing gas being located within the dust sublimation radius whereas the mid-IR flux arises from an area farther from the accretion disc. The fits yield both the bolometric luminosity of the intrinsic AGN and the torus integrated luminosity, from which we derive the reprocessing efficiency of the torus. In the models, the outer radial extent of the torus scales with the AGN luminosity, and we find the tori to be confined to scales less than 5 pc.