We present results from model fitting to the Spectral Energy Distribution (SED) of a homogeneous sample of Seyfert II galaxies drawn from the $12mu$m Galaxy Sample. Imaging and nuclear flux measurements are presented in an accompanying paper (Videla
et al., 2013). Here we add IRS Spitzer observations to further constrain the SEDs after careful subtraction of a starburst component. We use the library of CLUMPY torus models from Nenkova et al.~(2008ab) and also test the two-phase models recently produced by Stalevski et al.~(2012). We find that photometric and spectroscopic observations in the mid-IR (>5mu) are crucial to properly constrain the best-fit torus models. About half of our sources show clear near-IR excess of their SEDs above the best fit models. This problem can be less severe when using the Stalevski et al.~(2012) models. It is not clear what is the nature of this emission since best fitted black body temperatures are very high (~1700-2500 K) and the Type II classification of our sources would correspond to a small probability to peer directly into the hottest regions of the torus. Crucially, the derived torus parameters when using CLUMPY models are quite robust,, independently of whether the sources require an additional black body component or not. Our findings suggest that tori are characterized by N_0 > 5, sigma > 40, tau < 25, i > 40 degrees, Y < 50 and A_v^los ~ 100-300. From these we can determine that typical torus sizes and masses of 0.1-5.0 pc and 10^{4-6} M_odot. We find tentative evidence that those nuclei with a detected Hidden Broad Line Regions are characterized by lower levels of extinction than those without one. Finally, we find no correlation between the torus properties and the presence of circumnuclear or more global star-formation.
We present near and mid--IR observations of a sample of Seyfert II galaxies drawn from the $12mu$m Galaxy Sample. The sample was observed in the J, H, K, L, M and N bands. Galaxy Surface Brightness Profiles are modeled using nuclear, bulge, bar (when
necessary) and disk components. To check the reliability of our findings the procedure was tested using {em Spitzer/} observations of M,31. Nuclear Spectral Energy Distributions (SEDs) are determined for 34 objects, and optical spectra are presented for 38, including analysis of their stellar populations using the STARLIGHT spectral synthesis code. Emission line diagnostic-diagrams are used to discriminate between genuine AGN and HII nuclei. Combining our observations with those found in the literature, we have a total of 40 SEDs. It is found that about 40% of the SEDs are characterized by an upturn in the near-IR, which we have quantified as a NIR slope $alpha < 1$ for an SED characterized as $lambda f_{lambda} propto lambda^{alpha}$. Three objects with an HII nucleus and two Seyfert nuclei with strong contamination from a circumnuclear starburst, also show an upturn. For genuine AGN this component could be explained as emission from the accretion disk, a jet, or from a very hot dust component leaking from the central region through a clumpy obscuring structure. The presence of a very compact nuclear starburst as the origin for this NIR excess emission is not favored by our spectroscopic data for these objects.
